A

WOODS HOLE
OCEANOGRAPHIC INSTITUTIOi^

UBORATORY
BOOK COLLECTfON

PURCHASE ORDER HO.^ J 20l } »-f-

OCT 2 6 1954 S-i^^^^^^U

t:e>p^-

m

THE NUCLEIC ACIDS

Chemistry and Biology

Volume I

Contents of Volume II

16. The Nucleic Acid Content of Tissues and Cells by I. Leslie

17. Cytochemical Techniques for Nucleic Acids by Hewson H. Swift

18. The Isolation and Composition of Cell Nuclei and Nucleoli by Alexander
L. Bounce

19. The Deoxyribonucleic Acid Content of the Nucleus by R. Vendrely

20. Nucleic Acids in Chromosomes and Mitotic Division by B. Thorell

21. The Cytoplasm by George H. Hogeboom and Walter Schneider

22. Biosynthesis of Pentoses by Gertrude E. Glock

23. Biosynthesis of Purines and Pyrimidines by Peter Reichard

24. Biosynthesis of Nucleosides and Nucleotides by F. Schlenk

25. Biosynthesis of Nucleic Acids by G. B. Brown and P. M. Roll

26. The Metabolism of the Nucleic Acids by R. M. S. Smellie

27. The Biological Role of the Deoxypentose Nucleic Acids by R. D. Hotchkiss

28. The Biological Role of Pentose Nucleic Acids by J. Bracket

THE

C5^

V

NUCLEIC ACIDS ,i

Chemistry and Biology

Edited by

ERWIN CHARGAFF

Department of Biochemistry

Columbia University

New York, N. Y.

J. N. DAVIDSON

Department of Biochemistry

University of Glasgow

Glasgow, Scotland

Volume I

1955

ACADEMIC PRESS INC, PUBLISHERS

NEW YORK, N. Y.

Copyright 1955 by Academic Press Inc.

all rights reserved

no part of this book may be reproduced in any form

by photostat, microfilm, or any other means,

without written permission from the publishers.

ACADEMIC PRESS INC.

HI Fifth Avenue

New York 3, N. Y.

United Kingdom Edition

Published by

ACADEMIC PRESS INC. (London) Ltd.

Berkeley Square House, London W. 1

Library of Congress Catalog Card Number: 54-11055

• :f/

First printing 1955

Second printing 1960

Third Printing, 1962

PRINTED IN THE UNITED STATES OF AMERICA

Contributors to Volume I

J. Baddiley, Lister Institute of Preventive Medicine, London, England.

G. H. Beaven, Medical Research Council Spectrographic Unit, The Lon-
don Hospital, London, England.

Aaron Bendich, The Sloan-Kettering Institute for Cancer Research, New
York, N. Y.

D. M. Brown, University Chemical Laboratory, Cambridge, England.
Erwin Chargaff, Department of Biochemistry, College of Physicians and

Surgeons, Columbia University, New York, N.Y.
Waldo E. Cohn, Biology Division, Oak Ridge National Laboratory, Oak

Ridge, Tennessee.
J. N. Davidson, Department of Biochemistry, The University of Glasgow,

Glasgow, Scotland.
Zacharias Dische, Department of Biochemistry, College of Physicians

and Surgeons, Columbia University, New York, N.Y.

E. R. Holiday, Medical Research Council Spectrographic Unit, The Lon-
don Hospital, London, England.

E. A. Johnson, Medical Research Council Spectrographic Unit, The Lon-
don Hospital, London, England.

D. 0. Jordan,* Chemistry Department, The University, Nottingham,
England.

Hubert S. Loring, Department of Chemistry, Stanford University, Stan-
ford, California.

B. Magasanik, Department of Bacteriology and Immunology, Harvard
Medical School, Boston, Massachusetts.

W. G. OvEREND, Chemistry Department, The University of Birmingham,
Birmingham, England.

G. Schmidt, The Boston Dispensary, Boston, Massachusetts.

J. D. Smith, Agricultural Research Council, Plant Virus Research Unit,
Molteno Institute, University of Cambridge, Cambridge, England.

M. Stagey, Chemistry Department, The University of Birmingham,
Birmingham, England.

A. R. Todd, University Chemical Laboratory, Cambridge, England.

G. R. WYATT,t Laboratory of Insect Pathology, Sault Ste. Marie, Ontario,
Canada.

* Present address: Department of Chemistry, The University, Adelaide, South
Australia.

t Present address: Department of Biochemistry, Yale University, New Haven,
Connecticut.

Preface

There are many mansions in the house that the Natural Sciences have
been building in the course of 400 years; but it has remained one edifice.
Although its architectural style may not be of one piece, combining, as it
does, the labors of so many generations, the very consistency of the struc-
ture makes difficult the coherent description of one wing without a distor-
tion of proportions. An arbitrary line of demarcation must, however, be
drawn in a book of the kind we are presenting here.

The progress in our knowledge of nucleic acids has been unexpectedly
rapid in the past few years, and there is no reason to assume that it will
not continue. Until recently, the state of nucleic acid chemistry could be
compared to that in which protein chemistry found itself at the beginning
of this century. The principal constituents were known, at least in the two
nucleic acids that served as the prototypes; but a real insight into the order
and mode of their arrangement was lacking. All this has changed; the
nucleic acids are no longer regarded as mere pegs for the proteins. The large
number of publications, the multiplication of partial efforts, of fragmentary
reviews and symposia, make imperative an attempt at a detailed inventory
of our knowledge.

This treatise tries to collect all the information at present available into
a single comprehensive work. It is divided into two main parts, covering
the chemical and the more biological or biochemical aspects of the subject,
respectively. The chemistry of the hydrolysis products of nucleic acids is
covered in Chapters 2, 3, and 4, and their separation and estimation in
Chapters 5 to 9. The two main types of polynucleotides are discussed in
Chapters 10 and 11. The nature of the chemical bonds in nucleic acids is
considered in Chapter 12, their physical properties in Chapter 13, and their
optical properties in Chapter 14. The principal absorption spectra pro-
vided in Chapter 14 are also reproduced, on a metric scale, on two folded
sheets that will be found in a pocket on the inside back cover of Volume I.

The next chapter (15), which deals with the enzymes attacking nucleic
acids, forms the connecting link with the second main division of the book
which begins with a survey of the nucleic acid content of tissues, in Chap-
ter 16, and of cytochemical methods, in Chapter 17. The cell nucleus is
discussed in Chapters 18, 19, and 20, and the cytoplasm in Chapter 21.
The next four chapters cover the biosynthesis of nucleic acids and their
components. Nucleic acid metabolism is discussed in Chapter 26, and the
book ends with two chapters on the biological role of the two main types
of nucleic acid (Chapters 27 and 28).

A collaborative undertaking, the only one possible at the present time,

VIU PREFACE

entails well-known dangers. The Editors dare not hope that they have
evaded them, but in designing the book they have, as far as lay in their
power, made a determined effort to avoid producing a mere collection of
essays or review articles. Instead they have attempted to provide a con-
tinuous narrative in which the tale, although taken up by one storyteller
after another, nevertheless has a continuous and connecting thread running
through it from beginning to end. There may be gaps in the narrative, parts
of the story may be told more than once, but a sincere attempt has been
made to help the reader by linking the chapters with abundant cross refer-
ences. And it must not be forgotten that, even in a book of this size, the
story is far from complete, for additions to our knowledge of the nucleic
acids are made available with almost every issue of a host of scientific
journals.

The Editors owe a deep debt of gratitude to the many contributors who
have made the treatise possible and to the publishers for their unfailing
help and advice. They also wish to thank their secretaries, Mrs. Emmy
Bloch and Miss Mary Gilmour, for their patient and willing help.

If this book helps create an early need for a supplement to its present
content, it will have fulfilled one of its purposes.

Erwin Chargaff
J. N. Davidson

Contents

Contributors to Volume I v

Preface vii

1. Introduction by J. N. Davidson and Erwin Chargaff 1

I . The Early History 1

II. The Two Types of Nucleic Acid 3

III. Previous Literature 7

2. Chemistry of Ribose and Deoxyribose by W. G. Overend and M. Stacey . . 9

I . Introduction 9

II. Occurrence of D-Ribose and 2-Deoxy-D-Ribose 11

III. Chemistry of Ribose 16

IV. Chemistry of 2-Deoxyribose 48

V. Addendum 60

Appendix 65

3. Chemistry of Purines and Pyrimidines by Aaron Bendich 81

I. Introduction 81

II. General Properties of Purines and Pyrimidines 107

III. Synthetic Methods 125

Addendum 135

General References 136

4. Chemistry of Nucleosides and Nucleotides by J. Baddiley 137

I. Nucleosides 138

II. Nucleotides 160

III. Addendum 188

5. Hydrolysis of Nucleic Acids and Procedures for the Direct Estimation of
Purine and Pyrimidine Fractions by Absorption Spectrophotometry by

Hubert S. Loring 191

I. Hydrolysis of Nucleic Acids 191

II. Estimation of Purine and Pyrimidine Components in PNA 199

6. The Separation of Nucleic Acid Derivatives by Chromatography on Ion-Ex-
change Columns by Waldo E. Cohn 211

I. Ion Exchange 212

II. The Separation of Bases and Nucleosides 217

III. Separation of Nucleotides 221

IV. Separations Involving Sugar-Borate Complexing 235

V. Related Reviews 241

ix

X CONTENTS

7. Separation of Nucleic Acid Components by Chromatography on Filter Paper

BY G. R. Wyatt 243

I . Introduction 243

II. General Technique of Paper Chromatography 244

III. Detection of Purine and Pyrimidine Derivatives on Filter Paper 245

IV. Solvent Systems 248

V. Quantitative Estimation of the Nitrogenous Components of Nucleic

Acids 257

VI. Chromatography of Nucleic Acid Sugars 263

VII. Addendum 264

8. The Electrophoretic Separation of Nucleic Acid Components by J. D. Smith 267

I. Theory of the Electrophoretic Separation of Nucleic Acid Components 267

II. Apparatus and Techniques 274

III. Applications of the Method 276

9. Color Reactions of Nucleic Acid Components by Zacharias Dische 285

I. Color Reactions of Nucleic Acids and their Constituents Based on Re-
actions of Their Sugars 286

II. Determination of Purine and Pyrimidine Bases of Nucleic Acids by
Characteristic Color Reactions 303

10. Isolation and Composition of the Deoxypentose Nucleic Acids and of the
Corresponding Nucleoproteins by Erwin Chargaff 307

I. Introductory Remarks 308

II. Deoxypentose Nucleoproteins 309

III. Isolation of Deoxypentose Nucleic Acids 321

IV. Properties of Deoxypentose Nucleic Acids 333

V. Some Partial Degradation Products 340

VI. Constituents of Deoxypentose Nucleic Acids 345

VII. Composition of Deoxypentose Nucleic Acids 348

VIII. Fractionation of Deoxypentose Nucleic Acids , . 358

IX. Composition Studies and Structural Investigations 366

X. Correlations and Concluding Remarks 368

11. Isolation and Composition of the Pentose Nucleic Acids and of the Corre-
sponding Nucleoproteins by B. Magasanik 373

I . Introduction 373

II. Isolation of Pentose Nucleoproteins 374

III. The Nature of Pentose Nucleoprotein 379

IV. The Isolation of Pentose Nucleic Acids 384

V. The Nature of PNA •. 391

VI. The Nucleotide Composition of PNA 394

12. Evidence on the Nature of the Chemical Bonds in Nucleic Acids by D. M.
Brown and A. R. Todd 409

I. Introduction 409

II. Chemistry of the Ribonucleic Acids 413

III. Structure of the Deoxyribonucleic Acids 439

CONTENTS XI

13. The Physical Properties of Nucleic Acids by D. O. Jordan 447

I. Pyrimidines, Purines, Nucleosides, and Nucleotides 447

II. Nucleic Acids 461

14. Optical Properties of Nucleic Acids and Their Components by G. H. Beaven,

E. R. Holiday, and E. A. Johnson 493

I. Introduction 493

II. Bases, Nucleosides, and Mononucleotides 495

III. Nucleic Acids and Polynucleotides 514

IV. Ultraviolet Dichroism 532

V. Infrared Absorption Spectra 545

15. Nucleases and Enzymes Attacking Nucleic Acid Components by G. Schmidt 555

I . Introduction 556

II. Enzymes Catalyzing the Cleavage of Bonds Between Nucleotides. . . 558

III. Enzymes Catalyzing the Hydrolytic Cleavage of the Phosphoryl
Groups of Mononucleotides 590

IV. Nucleoside Kinases 594

V. Enzymes Acting on the Amino Groups of Purine and Pyrimidine Com-
pounds 595

VI. Enzymes Acting on the Linkages Between the Basic and the Carbo-
hydrate Groups of Nucleic Acid Derivatives 600

VII. Xanthine 0.\idase 609

VIII. Enzymes Involving the Opening of the Purine Ring 614

IX. Enzymes Involving the Opening of the Pyrimidine Ring 619

X. Some Data Concerning the Intracellular Distribution of Enzymes of

Nucleic Acid Metabolism 623

XI. Addendum 625

Author Inde.x 627

Subject Index 655

Errata, The Nucleic Acids, Vol. I

Page 124. Paragraph 3, line 4 should read: Table I.

Page 135. Center label of formula, line 2 should read: carboxamidine.

Page 138. Paragraph 1, line 10 should read: hydrolysis.

Page 138. Paragraph 3, line 1 should read: ribonucleic.

Page 145. Paragraph 1, line 4 should read: simultaneous.

Page 153. Line 1 should read: 6-glycosylaminopurines.

Page 188. Paragraph 5, line 1 should read: synthesis.

Page 197. Paragraph 2, line 2 should read: leads.

Page 208. For Pallade read: Palade. (Footnote 6, Line 1, and Reference

69.)

Page 208. Line 3 should read: Dounce.

Page 221. Paragraph 3, line 5 should read: orthophosphate.

Page 223. Line 2 should read: inosinic.

Page 231. Paragraph 2, line 11 should read: orthophosphate.

Page 247. Last line of text should read: Lactobacillus.

Page 262. Table IV, last column, last entry should read: 0.214.

Page 274. Last line of text should read: most separations.

Page 453. Formula XI:

I I

C C

/- \ /- \

for H read N

I I

HC HC

\ \

Page 463. Figure legend should read: nucleotide.

Page 509. Table II should read:

Thymidine^ — (fourth column): for 240.5 read 245.5
LTridine-'^ — (third column): for 8.5 read 7.4; (fourth column):
for 236.5 read 243; (fifth column): for 4.48 read 5.35.

Page 644. For Paladin read: Paladini.

Page 646. For Rush read: Rusch.

CHAPTER 1

Introduction

J. N. DAVIDSON AND ERWIN CHARGAFF

Page

I. The Early History 1

II. The Two Types of Nucleic Acid 3

III. Previous Literature 7

Die Wiirdigung Miescher's und seiner Arbeiten wird mit der Zeit nicht
abnehmen, sondern wachsen, und die von ihm gefundenen Thatsachen und
gedachten Gedanken sind Keime, denen noch eine fruchtbringende Zukunft
bevorsteht.

W. His. May, 1897

I. The Early History

The disco\'ery of the nucleic acids was the result of the work of Friedrich
Miescher (1844-1895), the founder of our knowledge of the chemistry of the
cell nucleus. As a pupil of Hoppe-Seyler in Tubingen in 1868-69, Miescher
became interested in the problem of isolating nuclear components, choosing
as his source of material the pus cells which he obtained from the surgical
bandages discarded in the nearby surgical clinics. By digesting the cells with
pepsin-hydrochloric acid and then shaking with ether, he was able to iso-
late the nuclei as a separate layer which settled to the bottom of the vessel
and could be filtered off. From the nuclear material he was able to prepare a
hitherto unknown compound which he called "nuclein." This substance was
acidic in nature, readily soluble in dilute alkali but insoluble in dilute acid,
and contained a high proportion of phosphorus. This last property alone
was sufficient to attract attention to the compound for at that time the only
known organic compound of phosphorus in the tissues was lecithin.

Miescher submitted an account of his results to Hoppe-Seyler, who found
them so surprising that he hesitated to publish them in his journal until he
had himself repeated the work, but in 1871 Miescher's original account to-
gether with Hoppe-Seyler's confirmation and supplementary papers by two
of his pupils appeared in Hoppe-ScTjIer's Medicinisch-chemische Untersuch-
1 in gen}

' F. Miescher, Hoppe-Seyler's Med. chevi. Unters. 1871, 441; P. Plosz, ibid. 1871,
461 ; N. Liibavin, ibid. 1871, 463; F. Hoppe-Seyler, ibid. 1871, 486; F. Miescher, ibid.
1871, 502.

1

2 J. N. DAVIDSON AND E. CHARGAFF

By this time Miescher had returned to Basel in his native Switzerland
where he found a more congenial and convenient source of nuclear material
in the sperm heads of the Rhine salmon. From these he isolated a high-
molecular-weight nuclein and a basic material less complex than known
proteins which he called "protamine" and which could be extracted from
the defatted sperm with dilute acid leaving the "nuclein" in the residue.
This nuclein had a phosphorus content of 9.59 % and gave analytical figures
corresponding to what is now known as nucleic acid. Indeed Miescher's
many preparations of nucleins from a variety of sources, carried out at low
temperatures under the most exacting conditions, would be regarded by
present day standards as highly satisfactory. In subsequent work he showed
that the nucleins of the salmon sperm were synthesized at the expense of
the musculature of the fish, which do not eat during the period of gonadal
development in fresh water.

After Miescher's death his friends, in particular His, organized the publi-
cation of a collected edition of his works including many of his letters.^
His work was continued by his successors. For example, Altmann,^ who first
used the term nucleic acid, developed methods for the preparation of pro-
tein-free nucleic acid from yeast as well as from animal tissues, while Kossel
and Neumann^ described a method for its preparation from thymus glands.
These methods of preparation were subsequently improved and developed
by Neumann.^

The discovery of the purine bases in nucleic acids was made by Piccard,®
who at Miescher's suggestion extracted salmon sperm with boiling hydro-
chloric acid and isolated guanine and hypoxanthine. But the most outstand-
ing work on the purine bases was carried out by Kossel,^ who for several
years from 1879 onwards was actively engaged in this field and was respon-
sible for the isolation of xanthine and adenine.

Miescher had isolated from the products of the hydrolysis of nucleic acid
a base which we now know to have been thymine, but this pyrimidine was
not properly identified until the work of Kossel and Neumann in 1894.^
Cytosine was isolated and identified in 1902-03 by Kossel and SteudeP

2 "Die histochemischen und physiologischen Arbeiten von Friedrich Miescher,"
2 vols. F. C. W. Vogel, Leipzig, 1897.

3 R. Altmann, Arch. Anat. u. Physiol, Physiol. Abt. 1889, 524.

* A. Kossel and A. Neumann, Ber. 27, 2215 (1894).

* A. Neuman, Arch. Anat. u. Physiol., Suppl. 1899, 552.
8 J. Piccard, Ber. 7, 1714 (1874).

7 A. Kossel, Z. physiol. Chem. 3, 284 (1879) ; 5, 152 (1881) ; 6, 422 (1882) ; 7, 7 (1882-3) ;
8, 404 (1883-4); 10, 248 (1886); 12, 241 (1888).

8 A. Kossel and H. Steudel, Z. physiol. Chem. 37, 177 (1902-3).

INTRODUCTION 3

and by Levene^ from thymus nucleic acid, while Ascoli*'' in 1900 isolated
uracil from yeast nucleic acid.

II. The Two Types of Nucleic Acid

The work of these early pioneers in nucleic acid chemistry, Miescher,
Kossel, Neuman, Steudel, O. Hammarsten, and others, is fully described
in the books by Jones," Feulgen,^^ and Levene and Bass,^^ and no more need
be said about it here. By 1930, however, a definite picture had emerged of
two definite types of nucleic acid. One of them, the nucleic acid from yeast,
on hydrolysis yielded adenine, guanine, cytosine, uracil, phosphoric acid,
and a sugar recognized by 0. Hammarsten as a pentose and identified by
Levene and Jacobs" as ribose. The other, the nucleic acid from thymus,
yielded adenine, guanine, cytosine, thymine, phosphoric acid, and a sugar
at first thought to be a hexose but later shown by Levene'^ to be a deoxy-
pentose and identified as deoxyribose.'^ These two nucleic acids, therefore,
came to be called ribonucleic acid and deoxyribonucleic acid, respectively,
and, since most nucleic acids of animal origin appeared to resemble that
from thymus, while the triticonucleic acid isolated from wheat embryo by
Osborne and Harris'^ was similar to that from yeast, the assumption was
made that pentose nucleic acids were characteristic of plants and deoxy-
pentose nucleic acids of animal tissues."' " The terms phytonucleic acid
and zoonucleic acid were suggested for these two groups, respectively. '^

This classification, however, was never free from objection since it had
been known from the end of the last century that pentose derivatives were
present in animal tissues. In 1894, for example, 0. Hammarsten'* prepared
from pancreas tissue a "i3-nucleoprotein" from which Bang'^ obtained a

9 P. A. Levene, Z. physiol. Chem. 37, 402 (1902-3).

1" A. Ascoli, Z. physiol. Chem. 31, 161 (1900-1901).

" W. Jones, "Nucleic Acids — Their Chemical Properties and Physiological Con-
duct," 2nd ed. Longmans Green & Co., London, 1920.

12 R. Feulgen, "Chemie und Physiologic dei Nukleinstoffe," Borntraeger, Berlin,
1923.

" P. A. Levene and L. W. Bass, "Nucleic Acids." Chemical Catalog Company, New
York, 1931.

1* P. A. Levene and W. A. Jacobs, Ber. 42, 2102, 2469, 2474, 2703 (1909).

>* P. A. Levene and E. S. London, /. Biol. Chem. 81, 711 (1929); 83, 793 (1929).

i« P. A. Levene, L. A. Mikeska, and T. Mori, /. Biol. Chem. 85, 785 (1930).

" T. B. Osborne and I. F. Harris, Z. physiol. Chem. 36, 85 (1902).

>8 O. Hammarsten, Z. physiol. Chem. 19, 19 (1894).

'« I. Bang, Z. physiol. Chem. 26, 133 (1898-99); 31, 411 (1900-1901).

4 J. N. DAVIDSON AND E. CHARGAFF

"guanylic acid" which could be precipitated by acetic acid. This material
was subsequently shown by the work of Feulgen,^'' of E. Hammarsten,^^
of Hammarsten and Jorpes,^^ and Jorpes^^"" to be a pentose polynucleotide.
From such pancreas material also, Jones and Perkins^® isolated, in crystalline
form, the pentose nucleotides of adenine, guanine, and cytosine and re-
ported the presence in spleen and liver of a substance resembling the
guanylic acid of the pancreas. It was at this time (1924) that the idea be-
gan to develop of a more widespread occurrence of pentose nucleic acids in
animal tissues than had previously been supposed.

It had been known for a long time that, in addition to the coenzyme nu-
cleotides, pentose nucleotides presumably derived from polynucleotides
were present in animal tissues such as chick embryo," mammary gland,^^' ^^
haddock^" and sea urchin^^ eggs, and even spleen and liver.^^' ^^ All this evi-
dence led to the suggestion made by Jones and Perkins-^ in 1924 and sup-
ported by Jorpes^* in 1928 that "the distinction between animal and plant
nucleic acid will in future not be so definitely drawn." Further support for
this view was provided by evidence for the presence of deoxyribonucleic
acids in plant tissues. ^^"'^

As the result of the development of Brachet's ribonuclease test [Chapter
17], the presence of pentose nucleic acids was demonstrated histochemically
in amphibia,"-^'' in the anterior pituitary of the growing rat and guinea pig,^'
and in toad's eggs.'*'^ At the same time the pioneer work of the Caspersson

2" A. Feulgen, Z. physiol. Chem. 108, 147 (1919-20).
" E. Hammarsten, Z. physiol. Chem. 109, 141 (1920).

22 E. Hammarsten and E. Jorpes. Z. physiol. Chem. 118, 224 (1922).

23 E. Jorpes, Biochem. Z. 151, 227 (1924).

2* E. Jorpes, Acta Med. Scand. 68, 253, 503 (1928).

25 E. Jorpes, Biochem. J. 28, 2102 (1934).

26 W. Jones and M. E. Perkins, /. Biol. Chem. 62, 290 (1924-25).

27 H. O. Calvery, /. Biol. Chem. 77, 489 (1928).

28 R. Odenius, Jahresber. Fortschr. Thierchem.. 30, 39 (1900).

29 J. A. Mandel and P. A. Levene, Z. physiol. Chem. 46, 155 (1905).

30 P. A. Levene and J. A. Mandel, Z. physiol. Chem. 49, 262 (1906).

31 K. C. Blanchard, J. Biol. Chem. 108, 251 (1935).

32 P. Thomas and C. Berariu, Compt. rend. soc. biol. 91, 1470 (1924).

33 R. Feulgen and H. Rossenbeck, Z. physiol. Chem. 135, 203 (1924).

34 A. Kiesel and A. N. Belozerski, Z. physiol. Chem. 229, 160 (1934).

35 A. N. Belozerski, Biokhimyia (U. S. S. R.) 1, 253 (1936); Compt. rend. acad. sci.
U. R. S. S. 25, 751 (19,39).

36 M. Behrens, Z. physiol. Chem. 253, 185 (1938).

37 J. Brachet, Arch. biol. (Liege) 44, 519 (1933).

38 J. Brachet, Arch. biol. (Liege) 48, 529 (1937).

39 J. Brachet, Arch. biol. (Liege) 51, 151, 167 (1940).

" J. Brachet, Compt. rend. soc. biol. 133, 88, 90 (1940).

" L. Desclin, Compt. rend. soc. biol. 133, 457 (1940).

« T. S. Painter and A. N. Taylor, Proc. Natl. Acad. Sci. U. S. 28, 311 (1942).

INTRODUCTION 5

school^^"*^ with the quantitative ultraviolet spectrophotometric technique
[Chapter 17] led to the detection of the presence of high concentrations of
pentose nucleotides or pentose polynucleotides in the cytoplasm of rapidly
proliferating cells such as sea urchin eggs, the spinach root-tip periblem cell,
the imaginal disks of larvae of Drosophila melanogaster, ^^' " embryonic
tissues,**'^* tumor tissues,^^ and the cells of actively secreting glands.^"
From all these results Caspersson"** concluded that a high concentration of
pentose nucleic acid was characteristic of cells in which rapid protein syn-
thesis was taking place, either for growth or for secretion.

These conclusions, based on histochemical and spectrophotometric meth-
ods, were confirmed for embryonic tissues in 1943 by Davidson and Way-
mouth,^' • ^2 who showed by chemical methods that pentose nucleic acid was
abundant in a large number of tissues in the sheep embryo and more abun-
dant than deoxypentose nucleic acid in many. The pentose nucleic acid,
however, was not peculiar to embryonic tissues for it was also found in a
corresponding series of adult tissues. Although the total nucleic acid was in
most cases higher in the embryo than in the adult, the amount of pentose
nucleic acid relative to deoxypentose nucleic acid varied from tissue to tis-
sue and was of the same order in embryonic as in the corresponding adult
tissue. The pentose nucleic acid of sheep liver was isolated in 1944 and
shown to be present in liver tissue in the surprisingly large proportion of 3
or 4 times as much pentose nucleic acid as deoxypentose nucleic acid.^'*
Its sugar was identified conclusively as D-ribose.*'' Since then, pentose nu-
cleic acids have been isolated from many animal tissues. Their properties
and the methods of isolation are discussed in Chapter 1 1 . The deoxypentose
nucleic acids are discussed in Chapter 10.

Although the nucleic acids were originally thought to be essentially nu-
clear constituents, the occurrence of the pentose type in the cytoplasm was
suspected as long ago as 1905." In deciding the location of the nucleic acids

" T. Caspersson, Skand. Arch. Physiol. 74, suppl. 8 (1936).

**T. Caspersson, J. Roy. Microscop. Soc. 60, 8 (1940).

" T. Caspersson, Naturwissenschaften 29, 33 (1941).

« T. Caspersson and J. Schultz, Nature 143, 602 (1939).

" T. Caspersson and J. Schultz, Proc. Natl. Acad. Sci. U. S. 26, 507 (1940).

^8 T. Caspersson and B. Thorell, Chromosoma 2, 132 (1941).

" T. Caspersson, C. Nystrom, and L. Santesson, Naturwissenschaften 29, 29 (1941);

Acta Radiol, Suppl. 46, (1942).
'" T. Caspersson, H. Landstrom-Hyden, and L. Aquillonius, Chromosoma 2, 111

(1941).
" J. N. Davidson and C. Waymouth, Nature 152, 47 (1943).
^^ J. N. Davidson and C. Waymouth, Biochem. J. 38, 39 (1944).
" J. N. Davidson and C. Waymouth, Biochem. J. 38, 379 (1944).
" J. N. Davidson and C. Waymouth, Biochem. J. 38, 375 (1944).
" S. P. Beebe and B. Shaffer, Am. J. Physiol. 14, 231 (1905).

6 J. N. DAVIDSON AND E. CHARGAFF

in the cell, histochemical tests have been of great value [Chapter 17]. For
example, the Feulgen nucleal reaction,^^ which is specific for deoxypentose
nucleic acid, has been used to demonstrate that this type of nucleic acid is
confined to the cell nucleus. Similarly the Brachet""^* histochemical test
with ribonuclease [Chapter 17] has demonstrated the presence of pentose
nucleic acid in the cell cytoplasm, in obviously abundant amounts in such
tissues as liver." In Caspersson's ultraviolet technique no distinction is
made between the two types of nucleic acid, but material which absorbs
ultraviolet light strongly and which is Feulgen-negative is considered to
be pentose polynucleotide.

The general conclusions from histochemical tests have been confirmed by
the procedure of cell disruption with separation of the morphological com-
ponents of the cells, as described in Chapters 18 and 21. The cytoplasmic
components contain pentose nucleic acid while the nuclei contain deoxypen-
tose nucleic acid and small amounts of pentose nucleic acid [Chapter 18].
We may therefore conclude that both types of nucleic acid are present
in all types of cell, both plant and animal, and that the main biological
distinction between pentose nucleic acid and deoxypentose nucleic acid is
that the former is mainly cytoplasmic while the latter is exclusively, or al-
most exclusively, nuclear. The designations "chromonucleic" and "plasmo-
nucleic acids" have been proposed for deoxypentose and pentose nucleic
acids, respectively," • ^^ but these names are not widely used.

The study of nucleic acids entered a new stage when the application of
paper chromatography to the separation of nucleic acid constituents (pu-
rines, pyrimidines, nucleosides, nucleotides) made possible precise analytical
investigations with very small amounts of materiaP^"^' [Chapters 7, 10, and
11]. The introduction of ion-exchange chromatography [Chapter 6] repre-
ss J. N. Davidson and C. Waymouth, Proc. Roy. Soc. Edinburgh B62, 96 (1944).
" A. W. Pollister and A. E. Mirsky, Nature 152, 692 (1943).
"8 A. W. Pollister and A. E. Mirsky, Nature 153, 711 (1944).
" E. Vischer and E. Chargaff, J. Biol. Chem. 168, 781 (1947).
«" E. Vischer and E. Chargaff, Federation Proc. 7, 197 (1948).
" R. D. Hotchkiss, /. Biol. Chem. 175, 315 (1948).

62 E. Vischer and E. Chargaff, /. Biol. Chem. 176, 703, 715 (1948) .

63 E. Chargaff, Experientia 6, 201 (1950); J. Cellular Comp. Physiol. 38, suppl. 1, 41
(1951); Federation Proc. 10, 654 (1951).

6« E. Chargaff and E. Vischer, Ann. Rev. Biochem. 17, 201 (1948).
66 J. N. Davidson, Ann. Rev. Biochem. 18, 155 (1949).
66 G. Schmidt, Ann. Rev. Biochem. 19, 149 (1950).
6' J. Baddiley, Ann. Rev. Biochem. 20, 149 (1951).

68 D. O. Jordan, Ann. Rev. Biochem. 21, 207 (1952).

69 G. B. Brown, Ann. Rev. Biochem. 22, 141 (1953).

76 J. N. Davidson, "The Biochenustry of the Nucleic Acids," 2nd ed. Methuen,
London, 1953.

INTRODUCTION 7

sented another important step. All these techniques were instrumental in
the development of current conceptions of nucleic acid structure [Chapters
10-12] and helpful in the rapid advance in synthetic procedures for the
preparation of nucleic acid constituents [Chapter 4].

It has now become abundantly clear that the names pentose nucleic acid
and deoxypentose nucleic acid are generic terms indicating groups of com-
pounds of similar composition. Chargaff and his colleagues*'^ were the first
to show that there are many nucleic acids differing in composition as regards
molar proportions of bases according to the biological source of the material
from which they are derived. There is even evidence of heterogeneity within
the cell, for the pentose nucleic acids of the cytoplasm appear to differ
slightly from those of the nucleus in the same tissue [Chapter 11] while the
deoxypentose nucleic acids of the nuclei of a single cell type have been sepa-
rated into fractions of different molar composition [Chapter 10].

The pentose sugar has been identified as ribose in the pentose nucleic acid
of yeast^'* and of liver*^ and has been shown to be chromatographically iden-
tical with ribose in the nucleic acids from a large number of other sources
[Chapter 11]. Consequently, such nucleic acids are frequently referred to as
ribonucleic acids (RNA) instead of pentose nucleic acids (PNA). Since
there is no evidence of the presence of any other pentose sugar, both terms
are legitimate, but for the sake of uniformity the contraction PNA will be
used in this book.

The sugar in thymus deoxypentose nucleic acid has been conclusively
proved to be D-2-deoxyribose^^ and the sugars in the corresponding nuclear
nucleic acids from a large number of other tissues have been shown to be
chromatographically identical with it [Chapter 10]. Consequently, these nu-
cleic acids are frequently referred to as deoxyribonucleic acids; but, whether
they be called deoxypentose nucleic acids or deoxyribonucleic acids, the
contraction DNA is conveniently and frequently used for both names.

III. Previous Literature

No major treatises on the chemistry of nucleic acids have appeared since
those of Jones," Feulgen,'- and Levene,'' nor has the biochemistry of nucleic

" R. Vendrely, Bull. soc. Chim. biol. 32, 427 (1950).

" F. Schlenk, Advances in Enzymol. 9, 455 (1949).

'' Nucleic acid, Symposia Soc. Exptl. Biol. 1 (1947).

''* Nucleic acids and nucleoproteins, Cold Spring Harbor Symposia Quant. Biol. 12

(1947).
'^Symposium on the biochemistry of nucleic acids, /. Cellular Comp. Physiol. 38,

suppl. 1 (1951).
'^ The chemistry and phj-siology of the nucleus, Exptl. Cell Research Suppl. 2 (1952).
" P. Boulanger and J. Montreuil, Bull. soc. chim. France, 1952, 844.
78 A. E. Mirsky, Sci. American 188, 47 (1953).

8 J. N. DAVIDSON AND E. CHARGAFF

acids and of their constituents been considered in detail. A large number of
monographs, reviews, and symposia has, however, been published, a selec-
tion of which is listed here.^'*'^^

^9 The chemistry and metabolism of nucleic acids, Phosphorus Metabolism 2, 301-439
(1952).

80 Symposium on nucleoproteins, Can. J. Med. Sci. 31, 222-302 (1953).

81 J. Brachet, "Le role des acides nucl^iques dans la vie de la cellule et de I'embryon."
Desoer, Li^ge, and Masson, Paris, 1952.

82 F. Egami, "Nucleic Acids and Nucleoproteins, Physics, Chemistry, Biology and
Medicine." 2 vols. Kyoritsu Pub., Tokyo, 1951.

CHAPTER 2

Chemistry of Ribose and Deoxyribose
W. G. OVEREND AND M. STAGEY

Page

I. Introduction 9

II. Occurrence of D-Ribose and 2-Deox3^-i)-ribose 11

III. Chemistry of Ribose 16

1. Preparation 16

2. Identification and Estimation 20

3. Properties and Derivatives 22

a. Physical Properties 22

b. 0-Glycosides 24

c. A''-Glycosides 27

d. Phosphates 33

(1) Ribose-1 -phosphate 36

(2) Ribose-2- and -3-phosphates 37

(3j Ribose-5-phosphate 39

(4) Other Pentose Phosphates 41

e. Ethers, Esters, Acetals and Anhydrides 42

(1) Ethers ,42

(2) Esters 43

(3) Acetals 46

(4) Anhydrides 46

f. Other Properties 47

IV. Chemistry of 2-Deo.\yribose 48

1. Preparation 48

2. Identification 53

3. Physical Properties 55

4. Properties and Reactions of Derivatives 55

a. 0-Glycosides 55

b. iV-Glycosides 56

c. Phosphates 58

d. Other Derivatives and Reactions 59

V. Addendum 60

Appendix. Tables of Physical Constants of Derivatives 65

I. Introduction

In recent years knowledge of the chemistry of the sugar components of
nucleic acids has been considerably broadened. Methods have been de-
veloped for the synthesis of important derivatives of these sugars, and the

10 W. G. OVEREND AND M. STACEY

properties and reactions of the compounds thereby obtained have been
investigated. In particular the general chemistry of 2-deoxyribose has been
intensively studied. Many experiments with D-ribose have had as objective
the preparation in good yield of intermediates suitable for use in projected
syntheses of naturally occurring ribo-nucleosides and -nucleotides. Several
excellent reviews have been published in recent years describing the chem-
istry of ribose/ of sugar phosphates including the phosphoric acid esters
of ribose and deoxyribose^ and of ribo(and deoxyribo)-nucleosides and
-nucleotides.^ In addition a recent description of the chemistry of 2-deoxy-
sugars* includes much of interest concerning 2-deoxyribose.

At the outset of this account attention must be directed to the nomen-
clature used for the class of compounds known as deoxysugars. A brief
perusal of the literature shows that it is extremely confusing. (See Overend
and Stacey^ for a full discussion of this problem.) For example, at various
times the sugar component of thymus nucleic acid has been referred to by
several different names. It is usual to name a deoxy-pentose or -hexose
after the parent sugar from w^hich it is prepared. The sugar moiety of
thymus nucleic acid can be prepared from D-ribose (I) by the glycal re-
action— which is used for the synthesis of 2-deoxysugars — and therefore
may be named 2-deoxy-D-ribose (II). However, due to the loss of asym-

CHO CHO CHO

HO— C— H

I
H- H— C— OH

H— C— OH

I
CH2OH

m

metry at carbon atom 2 during this reaction, II could be prepared from
D-arabinose (III) by the same series of reactions, and so could equally
well be referred to as 2-deoxy-D-arabinose. This latter name has been pre-
ferred by some chemists. Recently it has become possible to prepare II
from D-erythrose^ • ^ without using either D-ribose or D-arabinose as an
intermediate, so that at first sight this system of nomenclature might

1 R. W. Jeanloz and H. G. Fletcher, Jr., Advances in Carbohydrate Chem. 6, 135
(1951).

2 L. F. Leloir, Fortschr. Chem. org. Naturstoffe 8, 47 (1951).

3 G. W. Kenner, Fortschr. Chem. org. Naturstoffe 8, 96 (1951).

* W. G. Overend and M. Stacey, Advances in Carbohydrate Chem. 8, 45 (1953).

8 J. C. Sowden, J. Am. Chem. Soc. 71, 1897 (1949); 72, 808 (1950).

6 W. G. Overend, M. Stacey, and L. F. Wiggins, J. Chem. Soc. 1949, 1358.

H-

-C— OH

CH2

H-

-C— OH -^

Th-

1

-C— OH

H-

-C— OH

|H-

-C— OH

1

CH2OH

I .

CH2OH

I

n

CHEMISTRY OF RIBOSE AND DEOXYRIBOSE 11

appear to be outmoded. Besides minor variations, such as the use of both
"deoxyribose" and "desoxyribose," a trivial name — "thyminose" — has
been adopted by some authors. Although this trivial name is found rarely
nowadays in chemical literature, it persists to some extent in contemporary
publications relating to the biological sciences.

Attempts have been made to systematize the nomenclature of deoxy-
sugars. Bergmann and his co-workers^ introduced the ending "desose"
to denote a deoxysugar and referred to II as ribodesose. Probably the best
suggestion for systematizing the names of deoxysugars was made l)y Sow-
den,^ who proposed that the stereochemistry of the hydroxy 1 groups in
deoxy-pentoses and -hexoses should be denoted by a prefix. On this system
II is named D-er?/</?ro-2-deoxypentose. The dotted outline denotes the
portion of the sugar molecule to which the prefix relates.

Athough from the chemical ^•iewpoint this latter system of nomencla-
ture is probably the most satisfactory, it is not proposed to use it in this
review and instead II will be referred to as 2-deoxy-D-ribose. In this way
continuity and familiarity will be maintained with the nomenclature
adopted in the majority of papers relating to aspects of the chemistry of
II. In a survey summarizing the chemistry of ribose and deoxyribose, it
is considered to be important that changes in nomenclature from that
generally adopted by previous authors should be minimized.

Although this account will be concerned primarily with a description
of the chemistry of D-ribose and 2-deoxy-D-ribose, with particular emphasis
on those aspects of the subject which relate directly to nucleic acids, refer-
ence will be made to other pentose derivatives whenever they serve to
illustrate by analogy or comparison the main theme of this chapter.

II. Occurrence of D-Ribose and 2-Deoxy-D-Ribose

The only sugars found so far as components of nucleic acids are D-ribose
(I) and 2-deoxy-D-ribose (II).

Although KosseP recognized in 1891 that acidic hydrolysis of what is
now called ribonucleic acid liberates a carbohydrate derivative and three
years later Hammarsten^'' demonstrated that this sugar component was a
pentose, some fifteen years elapsed before Levene and Jacobs"- '- succeeded
in isolating the sugar in crystalline form and identifying it as D-ribose, at
that time an unkno\Mi sugar. Its identification was based on the facts that
it differed in its physical constants from the crystalline pentoses (L-arabi-

7 M. Bergmann, H. Schotte, and W. Lechinsky, Ber. 55, 158 (1922).

8 J. C. Sowden, J. Am. Chem. Soc. 69, 1047 (1947).

'A. Kossel, Arch. Anat. u. Physiol., Physiol. Abt. 1891, 181.
'0 O. Hammarsten, Z. physiol. Chem. 19, 19 (1894).
" P. A. Levene and W. A. Jacobs, Ber. 41, 2703 (1908).
'2 P. A. Levene and W. A. Jacobs, Ber. 42, 1198 (1909) ; 44, 746 (1911).

12 W. G. OVEREND AND M. STAGEY

nose, D-xylose and D-lyxose) known at that time: that its benzylphenyl-
hydrazone differed from that of arabinose;'^ but the melting point and
numerical value (though not the sign) of the specific rotation of its osazone
and p-bromophenylosazone were the same as for these derivatives of
L-arabinose: that on oxidation it gave D-ribonic acid, enantiomorphous
with that previously prepared synthetically by Fischer and Piloty,^'* and
that on further oxidation it yielded an inactive trihydroxyglutaric acid.
Final complete verification was achieved when Alberda van Eken stein
and Blanksma succeeded in synthesizing crystalline L-ribose'^ and later
D-ribose.^* The synthetic n-ribose was identical with the crystalline sugar
obtained by Levene and Jacobs from the purine nucleosides of yeast nucleic
acid.

Simultaneous hydrolysis and oxidation of cytidine with hydrobromic
acid and bromine affords bromouracil and D-ribonic acid.''' Since cytidine
is readily transformed into uridine by deamination,'^ it follows that the
pyrimidine nucleosides of yeast nucleic acid also have D-ribose as the sugar
component. Using benzimidazole derivatives Gulland et al.^^'^^ have re-
studied recently the problem of the identification of D-ribose in yeast
nucleic acid. In the naturally occurring ribosyl-purines and -pyrimidines
the sugar is present in the furanose form and is glycosidically linked in the
/3-configuration to position 9' of the purine bases or to position 3' of the
pyrimidine bases. (Compare some reviews summarizing the evidence lead-
ing to these conclusions.^^' -^) These conclusions, initially based on analytical
experiments, have been confirmed by synthesis studies^ • ^^'^^ and to some
extent by X-ray investigations. ""^^ In nucleic acids, hydroxyl groups in the
D-ribose portion of the molecule are concerned in phosphate ester inter-
im O. Ruff and G. Ollendorf, Ber. 32, 3234 (1899).

1* E. Fischer and O. Piloty, Ber. 24, 4214 (1891).

i^W. Alberda van Ekenstein and J. J. Blanksma, Chem. Weekblad 6, 373 (1909).

1* W. Alberda van Ekenstein and J. J. Blanksma, Chem. Weekblad 10, 664 (1913).

1' P. A. Levene and F. B. La Forge, Ber. 45, 608 (1912).

18 P. A. Levene and W. A. Jacobs, Ber, 43, 3150 (1910).

13 G. R. Barker and J. M. Gulland, J. Chem. Soc. 1943, 625.

20 G. R. Barker, Kathleen R. Cooke, and J. M. Gulland, J. Chem. Soc. 1944, 339.

" G. R. Barker, Kathleen R. Farrar, and J. M. Gulland, J. Chem. Soc. 1947, 21.

22 R. S. Tipson, Advances in Carbohydrate Chem. 1, 193 (1945).

23 D. O. Jordan, Ann. Rev. Biochem. 21, 209 (1952).

2" A. R. Todd, /. Chem. Soc. 1946, 647; Harvey Lectures, Ser. XX, 1-20, 1951-52.
2* J. Baddiley, Roy. Inst. Chem. (London), Lectures, Monographs Repts. No. 3, 1-17

(1950).
26 A. M. Michelson, J. Cellular Comp. Physiol. 38, Suppl. 1, 11 (1951).
2'S. Furberg, Nature 164, 22 (1949).

28 S. Furberg, Acta Cryst. 3, 325 (1950).

29 S. Furberg, Ada Chem. Scand. 4, 751 (1950).

CHEMISTRY OF RIBOSE AND DEOXYRIBOSE 13

nucleotide linkages. A discussion of these linkages forms the subject of
Chapter 12.

In addition to its occurrence as a component of ribonucleic acid and of
nucleotides like adenosine di- and triphosphate, D-ribose has been found in
combination with uric acid in the blood^" and united with 2-hydroxy-(i-
aminopurine (isoguanine) in the croton l)ean {Croton tiqlium L.).'"'^'-
2-Hydroxy-6-aminopurine-D-riboside (sometimes called crotonoside) can
be isolated from the croton seed by extraction with methanol, and Spies
and Drake''^ succeeded in obtaining from it crystalline D-ribose. The pentose
also occurs in certain vitamins and coenzymes. Acidic degradation of
vitamin B12 results in the formation of l-(Q;-D-ribofuranosyl)-5,6-dimethyl-
benzimidazole,^'*"^* and it has been shown that the ribofuranose moiety of
the vitamin is phosphorylated at either C-2 or C-3.^^ Coenzyme I (diphos-
phopyridine nucleotide) and coenzyme II (triphosphopyridine nucleotide)
isolated by Euler et alP'^^ and by Warburg et al}'^ from yeast and red
blood cells, respectively, coenzyme A''^ and the coenzyme of "galactowal-
denase" — uridine diphosphate glucose^^ — all have substituted ribose com-
ponents as part of their molecules. (See also Chapter 4 and Schlenk^-^ and
Romberg and Pricer^^ for further details.) Leloii- has suggested that
ribose-1 , 5-diphosphate is probably a coenzyme for the enzyme responsible
for the conversion of ribose-1 -phosphate to ribose-5-phosphate. Over a
quarter of a century ago Winter"*^ obtained a substance from certain animal

'"Alice R. Davis, Eleanor B. Newton, and S. R. Benedict, J. Biol. Cheyn. 54, 595
(1922)

31 E. Cherbuliez and K. Bernhard, Helv. Chirn. Acta 15, 464 (1932).

32 E. Cherbuliez and K. Bernhard, Helv. Chim. Acta 15, 978 (1932).

33 J. R. Spies and N. L. Drake, J. Avi. Chem. Soc. 57, 774 (1935).

3^N. G. Brink, F. W. Holly, C. H. Shunk, Elizabeth W. Peel, J. J. Cahill, and K.
Folkers, J. Am. Chem. Soc. 72, 1866 (1950).

35 N. G. Brink and K. Folkers, J. Am. Chem. Soc. 71, 2951 (1949); 72, 4442 (1950).

36 E. R. Holiday and V. Petrow, J. Pharm. and Pharmacol. 1, 734 (1949).

3' G. R. Beaven, E. R. Holiday, E. A. Johnson, B. Ellis, P. Mamalis, V. Petrow,
and B. Sturgeon, /. Pharm. and Pharmacol. 1, 957 (1949).

38 J. G. Buchanan, A. W. Johnson, J. A. Mills, and A. R. Todd, Chemistry d- In-
dustry 1950, 426; J. Chem. Soc. 1950, 2845.

33 H. V. Euler, H. Albers, and F. Schlenk, Z. physiol. Chem. 237, 1 (1935).

■"' H. V. Euler, and F. Schlenk, Z. physiol. Chem. 246, 64 (1937).

^1 H. V Euler, P. Karrer, and E. Usteri, Helv. Chim. Acta 25, 323 (1942).

« O. Warburg, W. Christian, and A. Griese, Biochem. Z. 282, 157 (1935).

"3 J. Baddiley, E. M. Thain, G. D. Novelli, and F. Lipmann, Nature 171, 76 (1953).

^* R. Caputto, L. F. Leloir, C. E. Cardini, and A. C. Paladini, J. Biol. Chem. 184,
333 (1950).

" F. Schlenk, J. Biol. Chem. 146, 619 (1942).

^« A. Romberg and W. E. Pricer, Jr., J. Biol. Chem. 186, 557 (1950).

"L. B. Winter, Biochem. J. 21, 467 (1927).

14 W. G. OVEREND AND M. STACEY

tissues which he considered to be an ethyl riboside, but confirmation of this
discovery has not been reported. Very recently, however, it has been claimed
that the browning occurring in moist fish flesh on exposure to high tempera-
tures is due to reactions of the Maillard type^* and that the sugar contrib-
uting the aldehyde groups is D-ribose.'*^

2-Deoxy-D-ribose is the only 2-deoxypentose believed to occur in nucleic
acids. Chargaff and his co-workers^"- ^^ have examined by chromatographic
techniques a range of nucleic acids isolated from animal, plant and bac-
terial sources and have found no trace of any other deoxysugar. In nucleic
acids 2-deoxy-D-ribose is linked glycosidically to either a purine or py-
rimidine base. As is the case with ribose, the purine bases are adenine and
guanine. The pyrimidine bases found united to the deoxypentose are
cytosine and thymine (in ribonucleic acids uracil replaces this latter base)
and in some cases 5-methylcytosine in small amounts. Although not a
normal component of deoxypentose nucleic acids, the deoxyriboside of
uracil has been obtained from an enzymic hydrolysis of a commercial
sample of herring sperm deoxypentose nucleic acid.*- It is considered to
have been formed by an enzyme (cytosine deoxyriboside deaminase^^)
introduced into the digest by bacterial contamination. In these nucleosides,
2-deoxy-D-ribose is linked to position 9' of the purine bases or to position
3' of the pyrimidine bases. ^"^ There is no conclusive evidence concerning
the stereochemical disposition of the deoxyribose-base linkage, although
Furberg-^ considers that deoxyribonucleotides may have the same shape
as ribonucleotides, in which he demonstrated by X-ray analysis that the
linkage was of the /3-D-type. Hammarsten and colleagues*^ have presented
evidence from biochemical experiments which can be interpreted*® as
indicating a /3-D-configuration for the sugar-base linkage in deoxyribonucleo-
tides. Until recently, the lactol ring structure of 2-deoxy-D-ribose in nucleic
acid had been proved conclusively only for the thymidine nucleoside com-
ponent and in this case it was furanose in form." Subsequently Brown
and Lythgoe*^ by apphcation of the periodate oxidation procedure to the

*8 H. L. A. Tarr, J. Fisheries Research Board Can. 8, 74 (1950).

" H. L. A. Tarr, Nature 171, 344 (1953).

^» E. Chargaflf, E. Vischer, R. Doniger, Charlotte Green, and F. Misani, J. Biol.

Cheni. 177, 405 (1949).
" E. Chargaff and Rakoma Lipshitz, J. Am. Chem. Sac. 75, 3658 (1953).
" C. A. Dekker and A. R. Todd, Nature 166, 557 (1950).

" T. P. Wang, H. Z. Sable, and J. O. Lampen, J. Biol. Chem. 184, 17 (1950).
" J. M. Gulland and L. F. Story, J. Chem. Soc. 1938, 259, 692.
" E. Hammarsten, P. Reichard, and E. Saluste, J. Biol. Chem. 183, 105 (1950).
56 J. Baddiley, Ann. Rev. Biochem. 20, 172 (1951).

" P. A. Levene and R. S. Tipson, Science 81, 98 (1935); J. Biol. Chem. 109, 623 (1935).
w D. M. Brown and B. Lythgoe, /. Chem. Soc. 1950, 1990.

CHEMISTRY OF RIBOSE AND DEOXYRIBOSE 15

2'-deoxyribosicles of guanine, hypoxanthine, cytosine and thymine, afforded
proof of the presence of a furanose sugar in each compound.

Besides 2-deoxy-D-ribose, other derivatives of D-ribose occur naturally. 3-Amino-
D-ribose has been shown to be one of the products of hydrolysis of Puromycin,^^' *"
and 5-thiomethyl-D-ribose occurs in the adenylthiomethylpentose^' which has been
isolated from yeast, crude vitamin Bi or crude cozymase. [Cf. Baddiley, Chapter 4.]
Reduction products of D-ribose are also found in Nature. For example ribitol (adoni-
tol) is isolable from Adonis vernalis L."' ^^ and the reduced ribityl residue forms part
of the riboflavin (vitamin B2 , 6,7-dimethyl-9-(D-ribityl)isoalloxazine) molecule
and of the coenzyme D-riboflavin-5'-phosphate (flavin mononucleotide). Despite the
absence of the ring structure of the sugar in these derivatives of ribitol, the stereo-
chemistry of the sugar residue is of great biological importance.*'' Synthesis of analo-
gous flavins containing other sugar residues gives rise to substances of quite different
biological activity. Flavin adenine dinucleotide, a mi.xed diphosphate ester of ribo-
flavin and adenosine, contains both ribofuranose and ribityl residues in its molecule.
This substance, which is the coenzyme chiefly involved in transporting electrons from
the reduced pyridine nucleotides to cytochrome, has been demonstrated in blood
cells,** liver, heart,** brain, kidnej^ intestine*' and muscle.*' It is probably a matter
of some significance that, in naturally occurring derivatives of ribose, the sugar is
present in the furanose form.

III. Chemistry of Ribose

1. Preparation

The first synthesis of ribose was achieved by Fischer and Piloty''' in
1891 by epimerization of L-arabonic acid to L-ribonic acid and subsequent
reduction of its lactone to syrupy L-ribose. Since that time many attempts
have been made to convert derivatives of glucose and arabinose into ribose.
Blanksma and Alberda van Ekenstein^^ repeated the work of Fischer and
Piloty and obtained a syrupy product which was contaminated with
ribitol. Subsequently, the crude ribose was purified through its p-bromo-
phenylhydrazone and crystalhne L-ribose was obtained for the first time.'^
The same workers repeated the synthesis in the D-series and by pyridine
treatment succeeded in converting D-arabonic acid (IV) into D-ribonic

" C. W. Waller, P. W. Fryth, B. L. Hutchings, and J. H. Williams, J. Am. Chem.

Soc. 75, 2025 (1953).
*» B. R. Baker and R. E. Schaub, J. Am. Chem. Soc. 75, 3864 (1953).
*' J. A. Mandel and E. K. Dunham, J. Biol. Chem. 11, 85 (1912).
*2 W. V. Podwissotzky, Arch, pharm. 227, 141 (1889); quoted by R. W. Jeanloz and

H. G. Fletcher, Jr., Advances in Carbohydrate Chem. 6, 145 (1951).
*3 E. Merck, Arch, pharm. 231, 129 (1893).
«^ E. L. Hirst, J. Chem. Soc. 1949, 522.

*5 J. R. Klein and H. I. Kohn, /. Biol. Chem. 136, 177 (1940).
** S. Ochoa and R. J. Rossiter, Biochem. J. 33, 2008 (1939).
«' A. V. Trufanov, Biokhimiya 6, 301 (1941).
68 A. V. Trufanov, Biokhimiya 7, 188 (1942).
*' J. J. Blanksma and W. Alberda van Ekenstein, Chem. Weekblad 5, 777 (1908).

16

W. G. OVEREND AND M. STAGEY

COOH
I
H-C-OH

COOH
I
HO-C-H

HOCH

H-OH

H-C-OH -»► H-C-OH

I

H-C-OH
I
CH2OH

IV

I

H-C-OH
I
CH2OH

V

H AcOyH-OAc

AcO

H-Br

COOH
I
HO-C-H
I
H-C-OH

I
H-C-OH

I
H-C-OH

CH2OH
XI

jj CH,— 0

HO

H

HO

C=0

HOH

H AcO
AcO Nj Y' ^^^

AcO H AcO H HO H

VIII IX X

acid (V). Thereafter by reduction with sodium amalgam and purification
via the p-bromophenylhydrazone, a crystalline sample of D-ribose (I)
was obtained.^® This method of synthesis was reinvestigated thoroughly by
Steiger''" in 1936. D-Arabinose (III) \Yas oxidized electrolytically'^^ to D-ara-
bonic acid (IV). IV was partly epimerized through the action of boiling
aqueous pyridine and D-ribonic acid (V) was isolated as its cadmium salt
and converted into D-ribonolactone (VI). Controlled reduction of the
lactone with 2.5% sodium amalgam gave D-ribose (I). More product could
be obtained from the mother liquors as the p-bromophenylhydrazone. By
this method D-ribose was obtained from D-arabinose in 17% overall yield.
Alberda van Ekenstein and Blanksma''^ claimed that L-arabinose may be
partly converted directly to L-ribose through the Lobry de Bruyn rearrange-
ment, but the process has no preparative value. Austin et alP failed to
confirm the findings of van Ekenstein and Blanksma,^^ since they were

'o Marguerite Steiger, Helv. Chim. Acta 19, 189 (1936).

" H. S. Isbell and Harriet L. Frush, /. Research Natl. Bur. Standards 6, 1145 (1931).
" W. Alberda van Ekenstein and J. J. Blanksma, Chem. Weekblad 10, 213 (1913).
"W. C. Austin, C. J. Smalley, and M. I. Sankstone, /. Am. Chem. Soc. 54, 1933
(1932).

CHEMISTRY OF RIBOSE AND DEOXYRIBOSE 17

unable to obtain crystalline ribose, its p-bromophenylhydrazone or crystal-
line derivatives (i.e., phenylhydrazide or cadmium salt) of ribonic acid
from syrups obtained by treating d- or L-arabinose with alkali. Alberda van
Ekenstein and Blanksma heated L-arabinose in N sodium hydroxide solu-
tion and then oxidized the product to a mixture of L-arabonic and L-ribonic
acids which were separated by fractional crystallization of their phenyl-
hydrazides. Except that they used calcium hydroxide (0.04 N) as alkali,
Austin et al. employed the same procedure. Explanations of the different
results reported by these two groups of workers have been offered in terms
of difTerent temperatures for epimerization, nature of the alkali, and con-
centration effects, but these are not entirely convincing.

Ribose has been obtained from arabinose by Gehrke and Aichner^"
by application of a method originally developed by Bergmann and Schotte"
during their researches with glucal. D-Arabinose (III) was converted via
its tetraacetate (VII) and 2,3,4-tri-O-acetyl-D-arabinopyranosyl bromide
(VIII) into 3 , 4-di-O-acetyl-D-arabinal (IX). (This compound can also be
named 3,4-di-O-acetyl-D-ribal.) After deacetylation, the product (X,
D-arabinal = D-ribal) in chloroform solution was hydroxylated at 0° with
perbenzoic acid to give a mixture pi D-arabinose (III) and D-ribose (I),
with the latter predominating. The ribose was purified via its crystalline
benzylphenylhydrazone. The sequence of reactions was also carried out
using L-arabinose as the initial material and in this case the syrupy product
was converted by oxidation into crystalline L-ribonolactone. The method
was extended and improved by Austin and HumoUer^® (cf. Neker and
Lewis" for related work), purification of the product being achieved by
direct crystallization and through formation of the p-bromophenylhydra-
zone. It was shown that treatment of L-arabinal with perbenzoic acid in
ethyl acetate at 0° yields 5 parts of L-ribose to 1 part of L-arabinose and
the overall yield of crystalline ribose, based on the arabinose used, was
nearly 10% of theoretical. Workers in other laboratories^^- "^ have also
employed this method to prepare ribose and frequently obtained the arabi-
nose needed as starting material from calcium D-gluconate by descent of
the sugar series. Hudson and Richtmyer*" have used this latter method to
convert the calcium salt of D-altronic acid (XI) (for the preparation of

7" M. Gehrke and F. X. Aiehner, Ber. 60, 918 (1927).
75 M. Bergmann and H. Schotte, Ber. 54, 450 (1921).
'SW. C. Austin and F. C. Humoller, J. Am. Chem. Soc. 54, 4749 (1932); 56, 1152

(1934).
" H. T. Neker and W. L. Lewis, J. Am. Chem. Soc. 53, 4411 (1931 ).
'8 R. Kuhn, K. Reinemund, F. Weygand, and R. Strobele, Ber. 68, 1765 (1935).
" P. Karrer, B. Becker, F. Beiiz, P. Frei, H. Salomen, and K. Schopp, Helv. Chim.

■Acta 18, 1435 (1935).
«» C. S. Hudson and N. K. Richtmyer, U. S. Pat. 2,162,721 (July 20, 1939).

18

W. G. OVEREND AND M. STACEY

calcium D-altronate see Richtmyer*^ into D-ribose (I) by the improved^^
Ruff^' degradation method for converting hexonic acids into pentoses.

Recently Sowden^ has introduced a method (cf. Fischer and Sowden*^)
whereby D-ribose can be obtained from D-glucose. 4,6-0-Benzylidene-D-
glucose (XII) was reduced catalytically to 4,6-0-benzylidene-D-glucitol
(XIII) which on oxidation with sodium metaperiodate yielded 2,4-0-
benzylidene-D-erythrose (XIV). Condensation of XIV with nitromethane
gave a mixture of epimeric, crystalline substituted C-nitro-alcohols, namely,
3,5-0-benzylidene-l-deoxy-l-nitro-D-arabitol (XV) and 3 , 5-0-benzylidene-
1-deoxy-l-nitro-D-ribitol (XVI).

0-CH,

HOH

CH.OH

I
H-C-OH
I
HO-C-H

H-C-0

H-C-OH

CHaO

XIII

HO

CH,NO,
-C-H

;CHPh

H-C-0.

I \
H-C-OH CHPh

I y

CHjO^
XV

CHO

I
H-C-0.

I \
H-C-OH CHPh

CHjO

XIV

CHjNO,
I
H-C-OH

& H-C-0.

H-C-OH XHPh

I y

CHjO
XVI

Separation of XV and XVI was achieved by virtue of a remarkable dif-
ference in their solubility in chloroform: XVI was very soluble in this sol-
vent. Hydrolysis of XVI afforded amorphous 1-deoxy-l-nitro-D-ribitol,
which in the form of its sodium salt was decomposed directly to D-ribose.
The pentose was isolated as the benzylphenylhydrazone. The purification
of syrupy samples of ribose can be achieved by the tedious procedure of
forming the pure crystalline p-bromophenylhydrazone^^- '^' ^° or benzyl-
phenylhydrazone^-''^ or p-toluenesulfonylhydrazone.^* A more convenient

81 N. K. Richtmyer, Advances in Carbohydrate Chem. 1, 37 (1945).

8* R. C. Hockett and C. S. Hudson, J. Am. Chem. Soc. 56, 1632 (1934).

83 O. Ruff, Ber. 32, 550 (1899) ; 35, 2360 (1902).

8< J. C.SowdenandH. O.L.Fischer, J. ^m. Chem. Soc. 6S, 1048 (1947).

80 D. G. Easterby, L. Hough, and J. K. N. Jones, J. Chem. Soc. 1951, 3416.

CHEMISTRY OF RIBOSE AND DEOXYRIBOSE 19

method has been introduced by Lee and Berger and their co-workers.*^- ^^
The technique is based upon the fact that arylamine-A^-ribosides (e.g.,
A^-phenyl-D-iibosylamine) prepared by dissolving the crude ribose in an
aqueous-alcohoHc solution of an arylamine, form "complex salts" with
the soluble salts of alkali metals. (The sugar-A^-glycoside forms a loose
combination with the inorganic salt.) The arylamine-A^-riboside may be
extracted- from this complex with dioxane and decomposed by boiling water
containing a trace (0.25 %) of acetic acid. Alternatively, aqueous hydrolysis
of the arylamine-A^-riboside, catalyzed by bases, or in the presence of alde-
hydes such as benzaldehyde, regenerates the original amine and D-ri-
bose.**'*'' After removal of the arylamine, either by steam distillation or
by forming the benzylidene derivative, pure crystalline ribose may be
obtained in yields of 70-90 %. Triacylribose derivatives can be prepared
by this procedure from arylamine A^-acylated ribosides. Thus, pure 2,3,4-
tri-0-acetyl-D-ribose was obtained by hydrolysis of the 2,3, 4-tri-O-acety 1-
A''-phenyl-a-D-ribosylamine^° which resulted from the anilination of crude
2,3,4-tri-O-acetyl-D-ribose and purification of the "anilide" via a "com-
plex salt."

The preparation of D-ribose from yeast nucleic acid has been extensively
investigated. Levene and Clark^' developed a method involving hydrolysis
of the nucleic acid with ammonia at an elevated temperature and pressure.
From the resultant mixture of nucleosides, adenosine and guanosine were
separated, purified and subjected to further hydrolysis to give D-ribose.
Levene^'- described a preparation of crystalline D-ribose by acidic hydrolysis
(0.05 A'^ sulfuric acid at the reflux temperature for 2 hours) of pure ash-free
guanosine. It was considered essential by Levene to work with a colorless
hydrolysate, if pure ribose is required. For the hydrolysis of nucleic acid,
Phelps^' replaced ammonia by magnesia and heated the mixture at 145°
for 4 hours. Magnesium-containing phosphates are removed by filtration
and then guanosine precipitates directly and adenosine is recovered from
the mother liquors as the picrate. Acidic hydrolysis of either guanosine
or adenosine picrate results in the liberation of D-ribose. (See also Bates
and associates'^ for preparative details of D-ribose from nucleic acid.)

8« J. Lee, U. V. Solmssen, and L. Berger, U. S. Pat. 2,384,103 (Sept. 4, 1945).
" L. Berger and J. Lee, J. Org. Chem. 11, 84 (1946).

8* L. Berger, U. V. Solmssen, F. Leonard, Ed. Wenis, and J. Lee, J. Org. Chem. 11,
91 (1946).

89 J. Lee, E. Fells, and L. Berger, U. S. Pat. 2,383,977 (Sept. 4, 1945).

90 J. Lee and L. Berger, U. S. Pat. 2.384,104 (Sept. 4, 1945).

" P. A. Levene and E. P. Clark, /. Biol. Chem. 46, 19 (1921).
92 P. A. Levene, J. Biol. Chem. 108, 419 (1935).
" F. P. Phelps, U. S. Pat. 2,152,662 (April 4, 1939).

9* F. J. Bates and associates, "Polarimetry, Saccharimetry and the Sugars," Natl.
Bur. Standards (U.S.) Circ. C440, 476 (1942).

20 W. G. OVEREND AND M. STACEY

Hydrolysis of aqueous solutions of yeast nucleic acid can be accomplished
by the agency of enzymic preparations from sweet almonds, lucerne seeds
or germinated peas.^^ Good yields of guanosine and adenosine are obtained,
and acidic hydrolysis of the former nucleoside yields D-ribose.^^- »« Details
for the isolation of D-ribose from yeast nucleic acid have been patented by
Laufer and Charney;^^ the method aims essentially at facilitating the
separation of nucleosides from a hydrolysate of the nucleic acid.

Addition of cuprous ions results in precipitation of the purine nucleosides
as cuprous salts. Acidic hydrolysis of the salts results in the formation of
D-ribose and the insoluble cuprous salts of adenine and guanine. Alter-
natively, the nucleic acid is hydrolyzed to a mixture of nucleotides which
are converted into their cuprous salts by addition of copper sulfate and
sodium bisulfite. The pyrimidine nucleotides remain in solution, but the
purine nucleotides are precipitated. After separation and washing these
purine nucleotides are subjected to hydrolysis with sulfuric acid. Insoluble
cuprous salts of adenine and guanine are removed from the hydrolysate by
filtration, leaving a solution containing D-ribose and sulfuric and phosphoric
acids. Addition to the solution of an alkaline earth hydroxide followed by
filtration gives a solution of practically pure D-ribose. »» Nowadays it is
more usual to separate nucleoside and nucleotide mixtures by ion-exchange
resin chromatography. ^^ [Cf. Cohn, Chapter 6.]

2. Identification and Estimation
A tentative identification of ribose may be achieved from the usual tests
for a reducing sugar and specific tests for pentoses. Determinations on
crystalline samples of the melting point, specific rotation and optical
crystallographic properties will serve to differentiate between the various
pentoses. Of the derivatives which -can be prepared for characterization,
substituted hydrazones have been preferred. At various times the p-bromo-
phenylhydrazone,i2. ,4. i5, 33. to, 76, 79. ioo-io3the benzylphenylhydrazone,!^. 74
the p-toluenesulf onylhydrazone^s • '"^ and to a lesser extent the diphenyl-

» H. Bredereck and G. Rothe, Ber. 71, 408 (1938).

»« H. Bredereck, M. Kothnig, and Eva Berger, Ber. 73, 956 (1940).

" L. Laufer and J. Charney, U. S. Pat. 2,379,913 (July 10, 1945).

98 L. Laufer and J. Charney, U. S. Pat. 2,379,914 (1945).

99 W. E. Cohn, Science 109, 377 (1949); /. Am. Chem. Soc. 71, 2275 (1949); /. Am.
Chem. Soc. 72, 1471 (1950).

looW. Alberda van Ekenstein and J. J. Blanksma, Chem. Weekblad 11, 182 (1914);
Brit. Abstr. 106 (i), 388 (1914).

101 P. A. Levene and R. S. Tipson, J. Biol. Chem. 115, 731 (1936).

102 C. W. Klingensmith and W. L. Evans, J. Am. Chem. Soc. 61, 3012 (1939).
"3 p. A. Levene and W. A. Jacobs, Ber. 42, 3247 (1909).

104 B. Helferich and H. Schirp, Chem,. Ber. 86, 547 (1953).

CHEMISTRY OF RIBOSE AND DEOXYRIBOSE 21

methane-dimethyl-dihydrazone^°^ have been used for this purpose. Ribose
may also be characterized by preparing 2-{'n6o-tetrahydroxybutyl)ben-
zimidazole (ribobenzimidazole)^"- ^'^ ^"^^ ^"^ from ribonic acid and o-phenyl-
enediamine. This latter method of characterization is not entirely un-
equivocal since Dimler and Link^"^ have demonstrated that the conversion
of ribose to ribonic acid by oxidation with alkaline hypoiodite results in
slight (5 %) epimerization and so the product is contaminated with arabonic
acid. Epimerization at this stage may be avoided by using the improved
procedure for preparing aldonic acids described by Hudson and Isbell,!"*
but there is a further risk of epimerization during the condensation of the
aldonic acid with o-phenylenediamine, particularly if there is insufficient
acid present. ^^'^^ (For a review^ of substituted benzimidazoles derived from
sugar acids see Richtmyer.^"^)

The j8-tetraacetate^^' ""^ and /3-tetrabenzoate^" and various mercap-
tals"^- "^ have frequently been used. Zinner"* recommends the di-n-propyl
and di-isobutyl mercaptals for the characterization of ribose. Chromato-
graphic methods have been used to identify ribose in mixtures of sugars
and to separate it from such mixtures. [Compare Chapters 6 and 7.] Various
reagents have been used to demonstrate the position of the sugar spots on
paper chromatograms : aniline hydrogen phthalate"*- "® in moist butanol
gives a bright red color with pentoses which may be differentiated from
aldohexoses, uronic acids and deoxysugars as these give various shades of
green and brown. It is possible with this reagent to detect 1 ng. of ribose.
Benzidine can be used as a developing agent"^ and pentoses are shown as
chocolate-browTi colored spots on a chromatogram. Jones et aZ."^ have
described experiences using various acidic spray reagents, as for example,
p-anisidine hydrochloride and aniline trichloroacetate and have reported
Rf values for ribose. Potassium permanganate with sodium carbonate has

"« J. V. Braun, Ber. 46, 3949 (1913).

»« N. K. Richtmyer and C. S. Hudson, J. Am. Chem. Soc. 64, 1612 (1942).

«7 R. J. Dimler and K. P. Link, J. Biol. Chem. 150, 345 (1943).

"8 C. S. Hudson and H. S. Isbell, /. Research Natl. Bur. Standards 3, 57 (1929).

"* N. K. Richtmyer, Advances in Carbohydrate Chem. 6, 175 (1951).

" P. A. Levene and R. S. Tipson, J. Biol. Chem. 92, 109 (1931).

" R. Jeanloz, H. G. Fletcher, Jr., and C. S. Hudson, J. Am. Chem. Soc. 70, 4052

(1948).
" E. Hardegger, E. Schreier, and Z. El Heweihi, Helv. Chim. Acta 33, 1159 (1950).
" H. Zinner, Chem. Ber. 86, 496 (1953).

1* (a) H. Zinner, Chem. Ber. 83, 275 (1950); (b) see also Chem. Ber. 86, 495 (1953).
15 S. M. Partridge, Nature 164, 443 (1949).

^' Judith Blass, M. Macheboeuf, and G. Nunez, Bull. soc. chim. biol. 32, 130 (1950).
" R. H. Horrocks, Nature 164, 444 (1949).
'8 L. Hough, J. K. N. Jones, and W. H. Wadman, J. Chem. Soc. 1950, 1702.

22 W. G. OVEREND AND M. STAGEY

been used'^^ successfully for indicating the position of D-ribose on paper
chromatograms.

Ribose may be determined by conversion into furfuraldehyde which
may be estimated as the phloroglucide'-"' '-' by colorimetric methods'^^' ^^^
or by separation on the paper-partition chromatogram and subsequent
estimation by sodium metaperiodate.^-'*' '-^ Under standard conditions the
formation of crystalline o-ribose p-toluenesulf onylhydrazone is quantitative
and may be used for the determination of D-ribose.*^ Solutions of the pentose
(1-4%) in methanol may be determined with an accuracy of ±1 %.

3. Properties and Derivatives

a. Physical Properties

Ribose (d-, l- and DL-forms) crystallizes without water of hydration.
Values reported by various workers for the melting point and specific
rotation are listed in the Appendix (Table VII) (p. 65). The refractive in-
dex and optical crystallographic properties of the D-isomer have been de-
scribed/-^ and Ellinghaus^" has reported the heat of combustion for the
sugar. From results of a polarographic study of D-ribose, Cantor and
Peniston^28 concluded that in aqueous solution 8.5 to 30 % of this sugar is
present in the aldehyde form. (The corresponding value for arabinose is
0.13 to 0.4 %.) The mutarotation of d- and L-ribose in aqueous solution at 1°
was studied by Isbell and his colleagues/-^ and was shown to be complex.
The mutarotation takes place rapidly and the direction of the change re-
verses after a few minutes so that the initial and final rotations are not very
different. On account of the rapidity w^ith which the reaction takes place,
the mutarotational changes are best observed at low temperatures. It was
concluded that in solution, equilibria probably exist involving both pyra-
nose and furanose forms. The initial mutarotational change of D-ribose
suggests that in the usual crystalline form it exists as the /3-anomer. Bred-
ereck et al.^^ examined the mutarotation of D-ribose (at 20°) and 5-0-trityl-
D-ribose (at 3°) in pyridine solution. Whereas the mutarotation of D-ribose

"9 E. Pacsu, T. P. Mora, and P. W. Kent, Science 110, 446 (1949).
"0 A. W. Schorger, Ind. Eng. Chem. 15, 742 (1923).

121 C. Dor^e, "Methods of Cellulose Chemistry," p. 381. Chapman & Hall, London,
1947.

122 Sonia Dunstan and A. E. Gillam, /. Chertl. Soc. 1949, S. 140.
>23 G. R. Barker, J. Chem. Soc. 1950, 1636.

124 E. L. Hirst and J. K. N. Jones, J. Chem. Soc. 1949, 1659.

12* E. L. Hirst, L. Hough, and J. K. N. Jones, /. Chem. Soc. 1949, 928.

i2« G. T. Keenan, J. Wash. Acad. Set. 16, 433 (1926).

1" J. Ellinghaus, Z. physiol. Chem. 164, 308 (1927).

128 S. M. Cantor and Q. P. Peniston, /. Am. Chem. Soc. 62, 2113 (1940).

123 F. P. Phelps, H. S. Isbell, and W. Pigman, /. Am. Chem. Soc. 56, 747 (1934).

CHEMISTRY OF RIBOSE AND DEOXYRIBOSE

23

Table I
Oxidation of Ribose with Bromine

Sugar

Av. value
of velo-
city const.
X 10'

Relative

reaction

rates ifsugar/

■'^a-D -glucose

Ratio of

rates for

a- and

/3-isomers

Mutarotation
constant at 0.2° db

0.2°C.
wi X 10^ mi X 10^

D-Ribose (crystalline) 196 6.1

;3-D-Ribose (from equilib- 1010 32

rated solution)

L-Ribose (crystalline) 195 6.1

/3-L-Ribose from equilib- 1456 45.5

rated solution)

5.2

6.87

54.0

Table II
R/? Values (corrected to 20°) of d-Ribose on Whatman No. 1 Filter Paper

Solvent"

Additions"

R/

Phenol
s-Collidine

n-BuOH (40%)
Acetic acid (10%)
Water (50%)

n-BuOH (40%)
Ethanol (10%)
Water (50%)

n-BuOH (45%)
Ethanol (5%)
Water (49%)

1-BuOH

Isobutyric acid
Methyl ethj-l ketone

NH3 (1% wt./vol.) HCN
None

None

NH3 (1%)

NH3 (1% wt./vol.)

NH3 (1% wt./vol.)

None

NH3 (1% wt./vol.)

0.59
0.56

0.31
0.285

0.210

0.180
0.220
0.165

Unless otherwise stated, all percentages are on a vol./vol. basis.

in this solvent is complex, that of the 5-trityl derivative, which cannot
exist in a pyranose form, is of the normal first-order type of reaction. The
explanation forwarded for the anomalous results with D-ribose was based on
furanose-pyranose interconversions. The bromine oxidation of ribose equili-
brated in aqueous solution has been investigated. ^^^^ '^° Initial rapid oxida-
tion is followed by a decrease in the reaction rate, a change ascribed to the
presence of a small quanity of some form which is oxidized more readily
than the crystalline form of the sugar. Table I illustrates the results ob-
tained.

"0 H. S. Isbell and W. Pigman, J. Research Natl. Biir. Standards 18, 141 (1937).

24 W. G. OVEREND AND M. STAGEY

The chromatographic behavior of ribose has been extensively studied
and, as stated, the technique of partition chromatography has been used
to identify the sugar, particularly when it is present in small quantities in
admixture with other substances. Partridge'^^ has reported Rf values for
D-ribose in a variety of solvent mixtures (see Table II) and has shown that
as little as 30 ng. of the sugar can be handled according to his procedure.

Using a column of powdered cellulose, separation of D-ribose from a four-
component mixture containing also galactose, rhamnose and arabinose,
was achieved successfully by Jones et al}^- Working on a 100-500 mg. scale,
crystalline ribose was recovered from this mixture in 94 % yield.

Kuhn'^^ has measured the infrared absorption of D-ribose from 8 to 15
microns.

h. 0 -Glycosides

As mentioned previously, Winter^^ claimed to have evidence supporting the exis-
tence of two pentose derivatives in animal (i.e., goats) tissues (e.g., liver and muscle),
one of which was probably an alkyl(ethyl) D-riboside. Confirmation of the discovery
has not yet been reported. Winter considered the possibility that the ethyl riboside
might have arisen during the process of tissue extraction, but dismissed the idea,
since arabinose failed to yield a glycoside under the conditions used for the isolation.
In view of subsequent work by Barker'^^ this conclusion was not justified, since ribose
rapidly forms a methyl ribofuranoside whereaa methyl arabofuranoside forms more
slowly. ^'^

H/CH'-\

^H-OMe

HO HO

XVII

l/^'~\

Ch Hy

>C=0

1 r

MeO MeO

XIX

H-OH

MeO MeO
XVIII

COjMe

I
H-C-OMe

I
H-C-OMe

I
H-C-OMe

COjMe
XX

1" S. M. Partridge, Nature 158, 270 (1946); Biochem. J. 42, 238 (1948).

"2 L. Hough, J. K. N. Jones, and W. H. Wadman, Nature 162, 448 (1948); /. Chem.

Soc. 1949, 2511.
"3 L. P. Kuhn, Anal. Chem. 22, 276 (1950).
"* G. R. Barker, J. Chem. Soc. 1948, 2035.
»"> Edna M. Montgomery and C. S. Hudson, J. Am. Chem. Soc. 59, 992 (1937).

CHEMISTRY OF RIBOSE AND DEOXYRIBOSE 25

■* H-C c(

O H-C

II
0

XXII

/«"-\ OMe

\6 o-c^ "h

/

\

H

Sr++

O

XXIII

Levene and Tipson"" warmed D-rihose with methanolic hydrogen chloride and
obtained a syrupy methyl D-riboside (XVII). Methylation'^' of this glycoside fol-
lowed by acidic hydrolysis afforded crystalline tri-0-methylribose (XVIII). Oxida-
tion of XVIII with bromine yielded a tri-O-methylribonolactone (XIX) which hy-
drolyzed at a rate characteristic for 5(1, 5) -lactones. When XVIII was oxidized with
nitric acid (d., 1.42) and the product esterified, the diester XX obtained was devoid of
optical activity and was in fact methyl (-trimethoxyglutarate. It was apparent from
this sequence of reactions that XVII had a pyranose ring structure and that XVIII
was 2,3,4-tri-O-methyl-D-ribose. Soon after publication of the above work, the prep-
aration of a crystalline methyl D-riboside was reported by Minsaas.'^^ This glycoside
was obtained by treatment of D-ribose with 0.25% methanolic hydrogen chloride, and
it exhibited a different specific rotation from that reported by Levene and Tipson"*
for their syrupy product (XVII). The structure of this crystalline compound was
subsequently investigated by Jackson and Hudson'^* and it was proved to be methyl
;8-D-ribopyranoside (XXI). For this work the glycoside was prepared by slight modi-
fication of Minsaas' procedure. Oxidation of the methyl riboside with periodic acid
or sodium metaperiodate resulted in the consumption of 2.02 moles of oxidant and
simultaneous formation of 1 mole of formic acid, thereby indicating that the material
was a glycopyranoside. The dialdehyde (XXII) also formed in this oxidation was
further oxidized with bromine to a dibasic acid, isolated as its strontium
salt (XXIII). The identity of XXIII with the strontium salts obtained when either
methyl /3-D-arabinopyranoside or methyl /3-D-xylopyranoside were subjected to the
same treatment'^' demonstrated that XXI had the/3-glycosidic configuration. Lattice
constants have been calculated'^" for this crystalline glycoside. The crystals were
rhombic and values obtained were a = 5.75 ± 0.02 A.;b = 6.39 db 0.02 A.; c = 19.90
± 0.03 A.

Barker'" has shown that the initial product of the action of methanol containing
1% hydrogen chloride on D-ribose is methyl D-ribofuranoside. The structure of this
glycoside was established by the classical method of methylation, hydrolysis and
oxidation to a lactone, which hydrolyzed at a rate characteristic of a 7 (1,4) -lactone.

'38 P. A. Levene and R. S. Tipson, /. Biol. Chcm. 93, 623 (1931).

'" J. Minsaas, Ann. 512, 286 (1934).

'38 E. L. Jackson and C. S. Hudson, J. Am. Chem. Soc. 63, 1229 (1941).

•" E. L. Jackson and C. S. Hudson, J. Am. Chem. Soc. 59, 994 (1937); 61, 1530 (1939).

'"» H. Brackken, Kgl. Norske Videnskab. Selskabs Fork. 9 184 (1936).

26 W. G. OVEREND AND M. STAGEY

Hence it was 2,3,5-tri-O-methyl-D-ribonolactone rather than the corresponding
2,3,4-derivative, and therefore the original glycoside must have had a furanose
structure.

Syntheses of alkyl ribosides by treatment of triacylribopyranosyl halides with
alcohols have been frequently reported. Levene and Tipson"" prepared a crystalline
methyl 2,3,4-tri-O-acetyl-D-riboside by reacting tri-0-acetyl-D-ribopyranosyl bro-
mide with methanol in the presence of silver carbonate. The tri-O-acetyl-o-ribo-
pyranosyl bromide was made by acting on crystalline D-ribose 1,2,3,4-tetra-O-acetate
with hydrogen bromide. The halide derivative was obtained in crystalline form and
showed a value of —209.3° in chloroform solution for its specific rotation, and so
presumably belongs to the /3-series. It is markedly unstable and is reconverted to
ribose tetra-0-acetate by shaking with silver acetate. The corresponding tri-0-acetyl-
/3-D-ribopyranosyl chloride has been prepared in crystalline form,'^^ but the ace-
tylated D-ribofuranosyl halides are very unstable, and although both the chlo-
i.i(jgi4i-i43 and the bromide"' ■ '^^ have been handled in various researches, neither has
been adequately characterized. The benzoylated D-ribopyranosyl halides apparently
are more stable than the acetyl analogues. Jeanloz et aZ.'" obtained 2,3,4-tri-O-
benzoyl-D-ribopyranosyl bromide in crystalline form through the action of hydrogen
bromide on /3-D-ribopyranose 1,2,3,4-tetra-O-benzoate in glacial acetic acid. The
compound could be stored at 5° indefinitely if kept over calcium chloride and caustic
potash. The j3-configurational assignment was based on optical rotation measure-
ments, and the pyranose structure follows from the observation that on reaction with
methanol in the absence of an acid-acceptor the ribose halide derivative is converted
to methyl 2,3,4-tri-0-benzoyl-/3-D-ribopyranoside,"* identical with the product
obtained by benzoylation of methyl /3-D-riboside (XXI) prepared according to Min-
saas, the structure of which was established as described above. Since tri-0-benzoyI-
/3-D-ribopyranosyl bromide may be obtained from D-ribose in 68% of the theoretical
yield and will react with methanol to give the benzoylated methyl riboside in high
yield (88%) and since debenzoylation can be carried out in nearly quantitative yield,
Jeanloz and Fletcher* claim that this route for the synthesis of alkyl ribopyranosides
is rather more attractive than that involving direct condensation of the sugar with
an alcohol, particularly as the latter procedure leads to mixtures from which it is
often difficult to obtain the required product in crystalline form. Furthermore it is
not absolutely necessary to isolate the lialogenated sugar in syntheses of alkyl ribo-
sides. Ethyl /8-D-ribopyranoside has been prepared by both routes.'" The pyranose
structure was confirmed by periodate titration.

Treatment of 2,3,4-tri-0-benzoyl-/8-D-ribopyranosyl bromide with methanol in
the presence of an acid-acceptor (i.e., silver carbonate) yielded an amorphous mixture
from which no crystalline material could be obtained.'" Reinvestigation of this
reaction and also of the reaction between /3-D-ribopyranose tetra-0-benzoate and
hydrogen bromide in glacial acetic acid, by Ness, Fletcher and Hudson,'" led to the
isolation in low yield (5.2%) of a new crystalline tri-O-benzoyl-D-ribopyranosyl

"' H. Zinner, Chem. Ber. 83, 153 (1950).

»« G. A. Howard, A. C. McLean, G. T. Newbold, F. S. Spring, and A. R. Todd, /.

Chem. Soc. 1949, 232.
1" J. Davoll, B. Lythgoe, and A. R. Todd, J. Chem. Soc. 1948, 967.
'" G. A. Howard, B. Lythgoe, and A. R. Todd, J. Chem. Soc. 1947, 1052.
'" R. Jeanloz, H. G. Fletcher, Jr., and C. S. Hudson, J. Am. Chem. Soc. 70, 4055

(1948).
'" R. K. Ness, H. G. Fletcher, Jr., and C. S. Hudson, J. Am. Chem. Soc. 73, 959 (1951).

CHEMISTRY OF RIBOSE AND DEOXYRIBOSE 27

bromide, isomeric with that previously known. It was dextrorotatory and was con-
sidered to be the a-isomer. The new a-isomer could also be obtained in very low yield
(1.3%) when tri-0-benzoyl-/3-D-ribopyranosyl bromide was treated with hydrogen
bromide in glacial acetic acid. Most of the starting material (i.e., /3-isomer) (96%)
could be recovered unchanged.

Treatment of /3-D-ribopyranose tetra-0-benzoate with titanium tetrachloride in
chloroform solution gives two crystalline products which are considered to be tri-O-
benzoyl-a- and -/3-D-ribopyranosyl chloride. The major product is the /3-anomer. Both
compounds react with methanol in the absence of an acid-acceptor to give methyl
tri-0-benzoyl-/3-D-ribopyranoside.

Further evidence in favor of the assigned configurations for the triacylribopy-
ranosyl halides was obtained by studying their reactions with methanol in the pres-
ence and absence of acid-acceptors. As expected the triacyl-;8-D-ribopyranosyl halides
reacted with the alcohol in the presence of acid-acceptors to yield acid-labile dex-
trorotatory syrups which most probably contained ortho-ester derivatives, whereas
the a-halides gave acylated derivatives of methyl /3-D-ribopyranoside in good yield.
The latter result is in conformity with present ideas on the role of neighboring groups
in replacement reactions (see Remick'") since the a-halides, having a halogen on
C-1 in a czs-position relative to the benzoyloxy group on C-2 might be expected to
react with methanol with simple inversion to give methyl /S-D-riboside tri-0-benzoate.
The j3-halides, on the other hand, having a /rans-relationship between the groups at
C-1 and C-2, react with methanol, in part at least, by a different mechanism. (Cf.
Jeanloz and Fletcher* for further details.) The /rans-halide reacted more rapidly
with methanol than its cis-isomer as shown in Table III."^ It wasshown'"*^ that tri-
0-benzoyl-;3-D-ribopyranosyl bromide in dry benzene-ether solution could be con-
verted in high yield to the corresponding chloride by treatment with active silver
chloride"*' "' according to the method developed originally by Haworth et al.^^°

Levene and Tipson"" found that tri-0-acetyl-/3-D-ribopyranosyl bromide reacts
with methanol in the presence of silver carbonate to give 3,4-di-O-acetyl-D-ribo-
pyranose methyl 1 ,2-orthoacetate. Klingensmith and Evans'"^ obtained an analogous
compound (i.e., di-O-acetyl-n-ribose l,2-ortho-3'-acetoxyacetonyl acetate) on con-
densing the same halide with dihydroxyacetone monoacetate. The structures of these
compounds have recently been discussed by Pacsu.'^'

c. N -Glycosides

A^-Ribosides occur naturally in important biological substances, as for
example ribonucleic acids, vitamin Bn and cozymase. Furthermore in the
ribityl derivatives which are found in Nature, C-1 of the reduced sugar
residue is linked to a nitrogen atom in a heterocyclic nucleus. Consequently
it is not surprising that a great deal of current interest in the chemistry of
D-ribose is centered on the A^-ribosides. The product from the condensation
of D-ribose (XXIV) with ammonia or a primary amine might be expected

1" A. E. Remick, "Electronic Interpretations of Organic Chemistry," 2nd ed., p.

339. J. Wiley & Sons, New York, 1949.
"8 H. H. Schlubach, Ber. 69, 840 (1926).
1" H. H. Schlubach and R. Gilbert, Ber. 63, 2292 (1930).
ISO W. N. Haworth, E. L. Hirst, and M. Stacey, J. Chem. Soc. 1931, 2864.
*^i E. Pacsu, Advances in Carbohydrate Chem. 1, 77 (1945).

28

W. G. OVEREND AND M. STAGEY

TABLE III

Reaction op Tri-O-benzoyl-D-ribopyranosyl Hamdes with

Dioxane-Methanol (1:9) at 20°

Compound

K X 10* Time of

min., logio "half -change,"
min.

Final Wr

Tri-0-benzoyl-a-D-ribopyranosyl

40

75

-68°

bromide

Tri-0-benzoyl-/3-D-ribopyranosyl

760

4.0

-53°

bromide

Tri-O-benzoyl-a-D-ribopyranosyl

0.62

4900

-65°

chloride

Tri-0-benzoyl-/3-D-ribopyranosyl

53

57

-49°

chloride

"Calculated on the assumption that each halide was quantitatively converted to methyl pentoside
tri-0-benzoate. Methyl ^-d ribopyranoside 2,3,4-tri-O-benzoateshows lalo" — 65.2° in dioxane -methanol (1:9).

to exist in a variety of tautomeric structures, since the Schiff base {syn-
or a7iti-io\m. of XXV) contains potentialities for isomerism and might
conceivably tautomerize to the anomeric pyranosides (XXVI and XXVII)
or the anomeric pair of furanosides (XXVIII and XXIX). These possible
changes have led to some confusion in explanations of the behavior of
A^-ribosides.

INH— C— H

CH=NR

H— C— NHR

H— C— OH

H-

-C— OH

H— C— OH

H— C— OH

^ H-

-C— OH

-^ H— C— OH

H— C— 0-^

1

H-

-C— OH

H— C— 0

1
CH2OH

CH2OH

CH2OH

XXVHI

XXV

XXIX

RNH— C— H

I
H— C— OH

H— C— OH

H— C— OH

CH2— 0^
XXVI

NH2R

CHO

H— C— OH

I

I
H— C— OH

H— C— OH

CH2OH
XXIV

H— C— NHR

I
H— C— OH

H— C— OH

H— C— OH

I
CH2

-O^

XXVII

CHEMISTRY OF RIBOSE AND DEOXYRIBOSE 29

The simplest type of compound of this class, namely that formed by
reaction between ammonia and D-ribose, was investigated by Levene and
La Forge. '^2 D-Ribose was treated with dry methanolic ammonia and the
nbosimine (XXX or one of its cyclic tautomers) obtained was further
treated with hydrogen cyanide in an attempted Kiliani-type synthesis
of two epimeric hexosaminic acids. When less rigorously anhydrous con-
ditions were employed, a diribosylamine was obtained, but neither of these
derivatives of D-ribose has been thoroughly studied. Interest in the syn-

H— C=NH

H— C— OH

H— C— OH

H— C— OH

I
CH2OH

XXX

thesis of vitamin B. led to the investigation of arylamine-N-ribosides. A
brief but adequate review of the products of reaction between D-ribose
and anihne was provided recently by Honeyman and Tatchell '^^ In 1937
Kuhn and Strobele'^" condensed D-ribose with 2-nitro-4,5-dimethylaniline
by heating in alcoholic solution in the presence of 2-3 % ammonium chlo-
ride. The product on acetylation yielded a triacetate and gave a monotrityl
derivative in high yield on tritylation and on the basis of these results
was considered to be iV-2-nitro-4,o-dimethylphenyl-D-ribofuranosylamine.
Owing to the possibility of isomerizations taking place, during both the
acetylation and tritylation reactions, the arguments in favor of the furanose
structure cannot be considered convincing. Moreover, it is well known that
trityl chloride will react with secondary hydroxyl groups in certain circum-
stances^^^ and is not a specific reagent for primary hydroxyl groups.

An extensive series of investigations concerning the reactions of arylamines, and
in particular aniline, with D-ribose was reported in 1945-46 by Berger, Lee and co-
workers. They concluded that condensation at low temperatures of arylamines with
D-ribose in alcoholic or aqueous alcoholic solution, with traces of acid as catalyst,
yields A^-aryl-a-n-ribopyranosylamine. If the reactants are condensed at the reflux
temperature of the solvent A^-aryl-a-o-ribofuranosylamines are formed. Configura-
tional assignments were based on mutarotation studies on the products. In view of
subsequent work by Barclay et al.,^^^ this is a doubtful criterion for distinguishing
between a- and ^-anomers of .V-glycosides. The A'-aryl-p-ribopyranosylamines could

1" P. A. Levene and F. B. La Forge, /. Biol. Chem. 20, 433 (1915).

'" J. Honeyman and A. R. Tatchell, /. Chem. Soc. 1950, 967.

"^ R. Kuhn and R. Strobele, Ber. 70, 773 (1937).

'" R. C. Hockett and C. S. Hudson, /. Am. Chem. Soc. 56, 945 (1934).

"« J. L. Barclay, A. B. Foster, and W. G. Overend, Chemistry & Industry 1963, 462.

30 W. G. OVEREND AND M. STAGEY

be converted quantitatively to the corresponding A'^-aryl-D-ribofuranosylamines
by heating in boiling alcoholic solution. Hydrogenation of both the pyranose and
furanose forms of the arylamine-A''-ribosides afforded the ribitylamine derivative.^'-
167, 168 Working with isomers of A^-phenyl-D-ribosylamine, Howard et aZ.'^' found that
the optical rotation of both isomers (i.e., pyranose and furanose isomers) is constant
in dry pyridine and that a trace of moisture is necessary for mutarotation. Under
these latter circumstances the two isomers do not come to the same end-value for
the specific rotation and it was considered likely that each is undergoing a,/3-isom-
erism. On the other hand, in solution in pyridine containing 10% acetic acid, the
isomers are claimed to mutarotate to the same end-point. It was thought that an
equilibrium involving change in ring size is established under these conditions and
acylation experiments supported this viewpoint. Acetylation of both A^-phenyl-
ribosylamines led to the same tri-0-acetyl-A^-phenyl-D-ribosylamine which upon
hydrolysis with methanolic ammonia afforded A^-phenyl-D-ribopyranosylamine.
Evidently the ribofuranosylamine had rearranged to the ribopyranosylamine during
acetylation. Acidic hydrolysis of tri-0-acetyl-A''-phenyl-D-ribopyranosylamine re-
moved the aniline residue, and upon further acetylation of the sugar moiety /3-d-
ribopyranose tetra-0-acetate was obtained. The mechanism of isomerization of amine
glycosides was discussed by Howard et al.^^^ Measurements of the changes in the
optical rotation of solutions of iV-phenyl-D-ribosylamines have also been reported
by Stacey and his colleagues.'*" In addition these workers also measured rates of
acidic hydrolysis of these compounds. Recent experiments by Barclay et al.^^^ sug-
gest that care must be exercised in attributing to true mutarotational phenomena
the changes in optical rotation that may occur in solutions of A^-glycosides without
prior ascertainment of the effect of water and pH alterations, since frequently the
changes in rotation are caused by hydrolysis rather than mutarotation.

Very recently the reaction between D-ribose and aniline has been reexamined
critically by Ellis and Honeyman.'*' It was established that water has an influence
in determining which isomer is obtained. Contrary to the suggestions of previous
workers, temperature is not important in influencing the nature of the product. For
example, the isomer (A) thought to be iV-phenyl-D-ribopyranosylamine was produced
either at room temperature or at the boiling point when aqueous ethanol was used
as solvent. The presence of moisture in the condensation mixture always resulted in
the production of this isomer and the supposed A^-phenyl-D-ribofuranosylamine (B)
was obtained only in anhydrous ethanol. Similarly, even a small amount of water
prevents the conversion of isomer (A) into isomer (B) in boiling ethanol, and this
probably explains some of the inconsistencies reported by Howard et al.^^^ Contrary
to workers in other laboratories, Ellis and Tatchell found no mutarotation for these
isomers when in pyridine solution, even on addition of a drop of water. It will be
recalled that it was largely on the basis of the mutarotation that Berger and Lee
considered the isomers to have different lactol rings.

A similar cycle of reactions was carried out with D-ribose and p-toluidine.'*' Until
the experiments of Ellis and Honeyman, the only known A'^-p-tolyl-D-ribosylamine
was that prepared and described by Berger and Lee.*' It was obtained by reaction at
room temperature, in a solvent of ethanol containing a trace of water. Ellis and
Honeyman obtained two isomers according to whether anhydrous or moist conditions

1" L. Berger and J. Lee, J. Org. Chem. 11, 75 (1946).

158 J. Lee, U. V. Solmssen, and L. Berger, U. S. Pat. 2,384,102 (Sept. 4, 1945).

1" G. A. Howard, G. W. Kenner, B. Lythgoe, and A. R. Todd, /. Chem. Soc. 1946, 855.

160 K. Butler, S. G. Laland, W. G. Overend, and M. Stacey, /. Chem. Soc. 1950, 1433.

i«i G. P. Ellis and J. Honeyman, Nature 167, 239 (1951); J. Chem. Soc. 1952, 1490.

CHEMISTRY OF RIBOSE AND DEOXYRIBOSE 31

of preparation were employed. The conditions of interconversion of these two isomers
followed exactly those found to be satisfactory for interconverting the A'-phenj'l-D-
ribosylamines. Likewise two isomers of A^-o-nitrophenyl-D-ribosylamine were sj'n-
thesized. The absorption of ultraviolet light by both isomers of A'-p-tolyl-D-ribo-
sylamine in methanolic solution was measured during the time mutarotation changes
were occurring. There was no appreciable change in the absorption corresponding
to the change in optical rotation which occurred. The absorption curve closely re-
sembled those obtained with simple aromatic amines, the main difference being a
slight lateral shift of the maxima [e.g., A^-p-tolylribosylamine (formed in the absence
of water) has Xmax. 2450 (log « 4.47) and 2900 A. (log « 3.72) whereas the other isomer
shows Xmax. 2500 (log e 5.62) and 2940 A. (log e 4.25). Aniline is reported'^^ to have
Xmax. 2300 (log ( 3.90) and 2800 A. (log « 2.30)]. Details of other arylamine-A^-ribosides
are listed in the Appendix (Table VII).- Reference has already been made to the
ability of arjiamine-A^-ribopyranosides to form "complex salts" with the soluble
salts of alkali metals.

Ribobenzimidazole [2- {p-riho- 1,2,3, 4-tetrahydroxybutyl)benzimida-
zole] has been used for the characterization of this pentose and as a means
of identifying the sugar component of yeast nucleic acid.'^"^^ These ex-
periments were described in Section III. (2). Recently interest in ribosyl-
benzimidazole derivatives has been stimulated by the finding that 5,6-
dimethylbenzimidazole-a-D-ribofuranoside (XXXI) is a component part
of the molecule of vitamin B12. Folkers and his colleagues^''' ^^ and workers
in other laboratories^^"^* succeeded in degrading vitamin B12 by acidic hy-
dryolsis to XXXI, the structure of which was confirmed by synthesis.^^
2-Nitro-4,5-dimethylaniline (XXXII) was condensed with 5-o-trityl-D-
ribofuranose (XXXIII) to give 2-nitro-4,5-dimethyl-A^-(5'-trityl-D-ribo-
furanosido)aniline (XXXIV) or one of its tautomers. Successive hy-
drogenation and condensation of XXXIV with ethyl formimino ether
hydrochloride and acid hydrolysis yielded 5,6-dimethylbenzimidazole-Q:-
D-ribofuranoside (XXXI), isolated as the crystalline picrate. When XXXIV
was acetylated prior to conversion to the benzimidazole derivative, the final
product obtained after the above sequence of reactions and deacetylation
was an isomer of XXXI. Optical rotation data suggested that it was the
)3-isomer. For both isomers the furanose structure for the sugar moiety was
confirmed by periodate titration. For convenience the dextrorotatory
isomer (XXXI) was termed a-ribazole and the levorotatory isomer was
referred to as /3-ribazole. Natural and synthetic a-ribazole had rat animal
protein factor activity of about J^oo of that displayed by vitamin B12 .^^^
At the level used j8-ribazole had about the same activity as the a-isomer.
Wacker and Weygand'" have studied /3-ribazole as an inhibitor of Lacto-
bacillus leichmannii 313.

162 Yor references see Ann. Rpts. on Progr. Chem. (Chem. Soc. London) 42, 124 (1945).
1" Gladys Emerson, F. W. Holly, C. H. Shunk, N. G. Brink, and K. Folkers, J. Am.

Chem. Soc. 73, 1069 (1951).
i«^ A. Wacker and F. Weygand, Z. Naturforsch. 6b, 130 (1951).

32

W. G. OVEREND AND M. STAGEY

H— C

H— C— OH
H— C— OH

H— C— O —

I
CH2OH

XXXI

CH3

CH3

NH,

NO2

XXXII

CHO
H— C— OH
+ H— C— OH
H— C— OH
CHiOTr

XXXIII

\

\

H— C

H— C— OH
H— C— OH

H— C— O— J

I
CHjOTr

-NH-

CH3
— CH3

NO5

XXXIV

Subsequently Folkers and co-workers^ ^^ prepared 5 , 6-dimethylbenzimi-
dazole-a- and -j3-D-ribopyranoside. The /3-pyranoside (XXXV) has also
been synthesized by Petrow et a/.^** The pyranose structure was confirmed
by periodate titration and the /S-configuration of the sugar-base linkage was
assumed from the method of synthesis, i.e., that 5 ,6-dimethylbenzimidazole
silver and a-acetobromoribose react with concomitant Walden inversion
to give the acetate of XXXV, which would thus have a j8-linkage.

XXXV

'«6F. W. Holly, C. H. Shunk, Elizabeth W. Peel, J. J. Cahill, J. B. Lavigne, and

K. Folkers, /. Am. Chem. Soc. 74, 4521 (1952).
'«8 G. Cooley, B. Ellis, P. Mamalis, V. Petrow, and B. Sturgeon, J. Phann. and

Pharmacol. 2, 579 (1950).

CHEMISTRY OF RIBOSE AND DEOXYRIBOSE 33

Petrow and his colleagues'" measured the pKa at 25 ± 1° of 5,6-di-
methylbenzimidazole-l-j8-D-ribopyranoside and found a value in water of
4.70 at a dilution of 490 liters per mole. Benzimidazole glycosides are
less basic than the corresponding free benzimidazoles and in general the
basicity falls in the series

l-/3-D-ribo- > 1-a-D-arabo- > 1-a-L-arabo- > l-/3-D-xylo-pyranose.

Various methods have been used in connection with syntheses of ribo-
flavin-type compounds to prepare A^-ribityl derivatives.'**'*^ A^-3,4-
Dimethylphenyl-D-ribosylamine affords directly on hydrogenation 3,4-
dimethylphenyl-D-ribamine.'^^ Ribitylbenzimidazole derivatives have been
prepared as possible carcinolytic compounds: l-(l'-D-ribityl)-5,6-dimethyI-
benzimidazole was ineffective in producing regression of established lym-
phosarcoma implants in mice.'^' Because of the great interest in the nat-
urally occurring nucleosides, studies of the synthesis of D-ribosides of
purine and pyrimidines have been numerous [cf. Baddiley, Chapter 4.].

d. Phosphates

Many biochemical reactions proceed by the agency of phosphorylated
derivatives of D-ribose (for reviews see Avison and Hawkins, '^^ and Foster
et al."^). [Cf. Glock, Chapter 22, and Schlenk, Chapter 24.] The present
account will be limited mainly to a description of the chemistry of the
monoorthophosphoric esters of ribose. Brief reference will be made to other
pentose phosphates, both for comparative purposes and because some were
suggested as components of nucleic acid, prior to the recognition of the
sugar moiety as D-ribose. For comparison of the properties and reactions of
pentose phosphates with those of hexose phosphates, disaccharide phos-
phates, etc., reference should be made to the excellent review of sugar
phosphates by Leloir.^

Several methods are available for differentiating between the various
phosphate esters of ribose. Like other aldose- 1 -phosphates, ribose- 1 -phos-
phate is very sensitive to acid treatment'^^ and in this respect resembles the
0-glycosides. Furthermore, like the glycosides, the /8-anomers of aldose-1-

'" M. T. Davies, P. Mamalis, V. Petrow, and B. Sturgeon, J. Phann. and Pharmacol .

3, 420 (1951).
'«« M. Tishler and J. W. Wellman, U. S. Pat. 2,261,608 (Nov. 4, 1941).
•»»R. Pasternack and E. V. Brown, U. S. Pat. 2,237,263 (April 1, 1941).
"» R. Kuhn and L. Birkofer, Ber. 71, 621 (1938).
'7' F. W. Holly, Elizabeth W. Peel, J. J. Cahill, and K. Folkers, J. Am. Chem. Soc.

73, 332 (1951).
'" A. W. D. Avison and J. D. Hawkins, Quart. Revs. (London) 5, 171 (1951).
'" A. B. Foster, W. G. Overend, and M. Stacey, Die Starke 5, 285 (1953).
'" H. M. Kalckar, J. Biol. Chem. 167, 477 (1947).

34

W. G. OVEREND AND M. STAGEY

TABLE IV
Rates of Acidic Hydrolysis of Some Pentose Phosphates

Compound

Normality of acid Temp., °C. K X 10'°

Ribose-1 -phosphate
Ribose-3-phosphate

Ribose-5-phosphate

Xylose-1 -phosphate
Xylose -5-phosphate

0.50
0.01
0.25
0.01
0.25
0.10
1.00

25
100
100
100
100

36
100

1200
-1.7
-4.5
-0.3
-0.5
6.21
3-4

Reference

174
175
176

175
176
177
178

1 a 1

"Constants are calculated from the formula K = - logio or more usually K = logi

t a — X h — ti

, when a is the initial concentration of the substance and the time (t) is expressed in minutes.

TABLE V
Specific Rotations of some Pentose Phosphate Esters

Compound

Salt

Solvent

[al

Reference

D-Ribose-3-phosphate

disodium

water

-9.7°

179

disodium

0.5 saturated boric
acid

+38°

179

D-Ribose-5-phosphate

free acid

water

-fl6.5°

175

barium

water

+5°

180

D-XyIose-1 -phosphate

barium

water

+65°

177

dipotassium

water

+76°

177

D-Xylose-5-phosphate

disodium

water

+3.2°

178

disodium

0.5 saturated
borax

+4°

178

D-Arabinose-5-phosphate

barium

water

-18.8°

181

brucine

50% aq. pyridine

-48.6°

181

phosphates are usually more acid-labile than the a-forms. Levene and
Stiller'^* demonstrated that a pentose-3-phosphate hydrolyzes much more
rapidly in acid solution than does the 5-isomer. For example, ribose-3-
phosphate hydrolyzes 5 to 9 times faster than ribose-5-phosphate under
identical conditions. This method of distinguishing between the 3- and 5-
phosphate esters of ribose is applicable not only to the compounds them-

175 p. A. Levene and E. T. Stiller, /. Biol. Chem. 104, 299 (1934).
"8 H. G. Albaum and W. W. Umbreit, /. Biol. Chem. 167, 369 (1947).
1" W. R. Meagher and W. Z. Hassid, /. Am. Chem. Soc. 68, 2135 (1946).
178 p. A. Levene and A. L. Raymond, /. Biol. Chem. 102, 347 (1933).
"9 P. A. Levene and S. A. Harris, J. Biol. Chem. 95, 755 (1932).

180 A. M. Michelson and A. R. Todd, /. Chem. Soc. 1949, 2476.

181 P. A. Levene and C. C. Christman, /. Biol. Chem. 123, 607 (1938).

CHEMISTRY OF RIBOSE AND DEOXYRIBOSE 35

selves, but also to substances containing either of them as a molecular
component.^'^ Table IV includes values calculated for the velocity con-
stant {K X 10^) for the acidic hydrolysis of ribose phosphates, and for,
comparison some xylose phosphates are included.

Details concerning the alkaline hydrolysis of pentose phosphates are
somewhat more scanty. Measurements of specific rotations have been used
for distinguishing between ribose-3- and -5-phosphates and other pentose
phosphates. In Table V values of the specific rotation for various pentose
phosphates are listed.

Klimek and Parnas'*- distinguished between adenylic acids possessing
ribose-3- and -5-phosphate moieties by a method based on the formation
of a blue soluble complex by adenosine-5'-phosphate in alkaline solution
in the presence of copper sulfate. Under the same conditions only an in-
soluble precipitate is formed by adenosine-3'-phosphate, which after
centrifuging leaves a clear colorless supernatant solution. The procedure
was standardized more completely l)y Berlin and Westerberg.'^^ Albaum and
Umbreit'"^ have developed a method for differentiating between ribose-3-
and -5-phosphates and compounds containing them, by means of the
orcinol pentose color reaction. The method is rapid and can be used on as
low an amount as 10 ^g. of phosphate ester. It cannot be used precisely on
crude plant and bacterial extracts containing polysaccharides, since these
alter the rate of color development. Paper chromatography provides a val-
uable micromethod for the identification of sugar-phosphate esters. '^''"'^^
Some Rf values of ribose phosphates and other pentose phosphates, as
quoted by Cohen and McNair Scott, '^^ are shown in Table VI.

The addition of boric acid to the solvents retards the movement of
ribose-5-phosphate compared with arabinose-5-phosphate, an effect at-
tributed to the combination of boric acid with the cfs-hydroxyl groups
attached to C-2 and C-3 in ribose. Ribose phosphates may be separated
from other sugar phosphates by ion-exchange resin chromatography.
[Cf. Cohn, Chapter 6.] Horecker and Smyrniotis'^* used Dowex 1 formate
for the separation of pentose phosphates resulting from the action of a
yeast enzyme on 6-phosphogluconate. Sugar phosphates have also been

'82 R. Klimek and J. K. Parnas, Biochem. Z. 252, 392 (1932).

183 H. Berlin and J. Westerberg, Z. phijsiol. Chan. 281, 98 (1944).

184 C. S. Hanes and F. A. Isherwood, Nature 164, 1107 (1949).
186 S. S. Cohen and D. B. McNair Scott, Science 111, 543 (1950).

186 R. S. Bandurski and B. Axelrod, J. Biol. Chem. 193, 405 (1951).

187 A. A. Benson, J. A. Bassham, M. Calvin, T. C. Goodale, V. A. Haas, and W.
Stepka, /. Am. Chem. Soc. 72, 1710 (1950).

'88 (a) B. L. Horecker and P. Z. Smyrniotis, Arch. Biochem. and Biophys. 29,
232 (1950) ; see also (b) B. L. Horecker, P. Z. Smyrniotis, and J. E. Seegmiller, J.
Biol. C/iem. 193, 383 (1951).

36

W. G. OVEREND AND M. STAGEY

TABLE VI
Rf Values of Pentose Phosphates

Solvent system

Compound

80% ethanol containing
0.8% acetate at pH 3.5

80% ethanol containing
0.64% boric acid

D-Ribose-5-phosphate

D-Arabinose-5-phosphate

D-Xylose-5-phosphate

D-Ribose-3-phosphate

D-Xylose-3-phosphate

0.50
0.64
0.55
0.50
0.53

0

0.25

0, 0.25

0, 0.19

0, 0.23

separated by ion exchange with the use of the borate complex, ^*^' '^° and
it is also possible that paper ionophoresis can be used. [Compare Chapter 8.]
(1) Ribose-1 -phosphate. Kalckar"''- '^i-i^^ discovered that enzymic phos-
phorolysis of some nucleosides (inosine, guanosine) leads to the production
of a pentose phosphate, isolable as its barium salt, which is most probably
D-ribose-1 -phosphate. [Cf. Schlenk, Chapter 24.] The yield of pentose phos-
phate generally was low (i.e., 7-10 mg. of the barium salt from 40-60 mg.
of inosine). Probably losses are due to acid hydrolysis, to nonspecific and
specific contaminant phosphatase action during the incubation with the
enzyme and to retention on the bulky barium phosphate precipitate during
the working up stage. Moreover the position of the equilibrium favors the
formation of hypoxanthine riboside. Although Kalckar'^^ was unable to
demonstrate that adenosine or xanthosine or the pyrimidine ribosides
would act in the system, it appears that the following relation might
hold true:

Ribose-1-purine + PO4 ^ Ribose-1 -phosphate -t- purine

Subsequent work has demonstrated that in addition to the examples stud-
ied by Kalckar mammalian purine nucleoside phosphorylase catalyzes the
synthesis from the respective base and ribose-1 -phosphate of xanthosine, >^*
8-azaguanine riboside, '^^ nicotinamide riboside'^''' and 4-amino-5-imidazole-

•89 J. X. Khym and W. E. Cohn, /. Ain. Chem. Soc. 75, 1153 (1953).

190 J. X. Khym, D. G. Doherty, E. Volkin, and W. E. Cohn, J. Ajh. Chem. Soc. 75,

1262 (1953).
'9' H. M. Kalckar, Federation Proc. 4, 248 (1945).
192 H. M. Kalckar, J. Biol. Chem. 158, 723 (1945).
'93 H. M. Kalckar, Symposia Soc. Exptl. Biol. 1, 38 (1947).
'94 H. M. Kalckar, Biochim. et Biophys. Ada 4, 232 (1950).
196 M. Friedkin, J. Am. Chem. Soc. 74, 112 (1952).
196 M. Friedkin, Federation Proc. 11, 216 (1952).
19V J, W. Rowen and A. Kornberg, J. Biol. Chem. 193, 497 (1951).

CHEMISTRY OF RIBOSE AND DEOXYRIBOSE 37

carboxamkle riboside. '^'^ Moreover Korii et a/.'^* working with purified
beef liver nucleoside phosphorylase obtained some evidence of reaction
between ribose-1 -phosphate and adenine. This result differs from that ob-
tained by Kalckar working with rat liver nucleoside phosphorylase.

An accumulation of evidence supports the designation of the pentose
phosphate as ribose-1 -phosphate. It is nonreducing and readily hydrolyzed
by acid to equimolecular quantities of phosphate ions and pentose. The
extreme acid lability supports a glycosidic phosphate ester linkage and
indeed the substance is sufficiently acid-labile to lose its phosphate group
by hydrolysis at the acidity employed in some of the methods available
for phosphate estimation. The ester is somewhat more acid-labile than
phosphocreatine but less so than acetyl phosphate. For example in 0.5 N
sulfuric acid the half-time for splitting for phosphocreatine, acetyl phos-
phate and ribose-1-phosphate is 4 miiuites, 30-40 seconds and 2.5 minutes,
respectively. Further support for the glycosidic linkage is afforded by the
enzymic conversion of the phosphate ester to purine ribosides which also
have a linkage to C-1 of the sugar moiety. The stereochemical configuration
of the phosphate ester linkage is unknown, but it is probably of the jS-type.
Natural nucleosides are considered to be /S-furanosides,'"*^' '^^ and, if nucleo-
side phosphorylase (like polysaccharide phosphorylase) produces no in-
versioUj^"" the ribose-1 -phosphate should also be of the /3-furanose type. The
furanose structure for the lactol ring is supported by the observation' ^'^
that synthetic ribopyranose-1 -phosphate is inactive as a substrate for
nucleoside phosphorylase.

Ribose-1 -phosphate can be converted enzymically by mutase action into
ribose-5-phosphate. [Cf. Glock, Chapter 22.] Klenow and Larsen'^"' have
shown that phosphoglucomutase preparations,^"^' ^"^ acting in conjunction
with glucose- 1 ,6-diphosphate (or possibly ribose-1 ,5-diphosphate) as
coenzyme, will bring about this change. The ratio of the enzymic activities
of phosphoglucomutase : phosphoribomutase was about 100:1. Reports by
Wajzer and Baron-"* indicate that liver contains an enzyme capable of
transforming ribose-1 -phosphate into the -5-phosphate.

(2) Rihose-2- and -3 -'phosphates. The location of the phosphate residue
in the first pair of isomeric nucleotides discovered and isolated by Cohn^^
(adenylic acids "a" and "b") has generally been regarded as at the 2'- and

»9» E. D. Korn, F. C. Charalampous, and J. M. Buchanan, J. Am. Chem. Soc. 75, 3610

(1953).
'99 J. Davoll, B. Lythgoe, and A. R. Todd, J. Chem. Soc. 1946, 833.
^oo Mildred Cohn, J. Biol. Chem. 180, 771 (1949).

^"^ Hans Klenow and B. Larsen, Arch. Biochem. and Biophys. 37, 488 (1952).
^o^ V. A. Najjar, ./. Biol. Chem. 175, 281 (1948).
2« E. W. Sutherland, J. Biol. Chem. 180, 1279 (1949).
20^ J. Wajzer and Frangoise Baron, Bull. soc. chim. biol. 31, 750 (1949).

38 W. G. OVEREND AND M. STAGEY

3'-positions, but not necessarily respectively.^"^- ^oe [Compare Chapters 4
and 12.] The structures of the subsequently isolated isomeric pairs of
guanylic,^^ cytidylic^*'^ and uridylic acids-"^ have been assumed to be the
same as the adenylic acid pair. The demonstrated acid-catalyzed migra-
tion of the phosphate group^''^-^"^ made difficult a decision as to which
nucleotide was 2'- and which 3'- in synthetic^"* and degradative^o^ ap-
proaches to the problem. The problem has been solved in elegant fashion
by Cohn et a/./^° who succeeded in identifying the isomeric adenylic acids
"a" and "b" as the adenosine-2'- and -3 '-phosphates, respectively. They
were able to hydrolyze catalytically the iV-glycoside linkage of the individ-
ual adenylic acid isomers with the hydrogen form of a polystyrene sulfonic
acid resin (Dowex 50) at a rate comparable to the rate of isomerization. The
advantage of this method of hydrolysis lies in the fact that the ribose phos-
phates are released from the resin at the time of formation (in contrast to
adenine and most of the adenylic acid) and, therefore, little or no iso-
merization takes place subsequent to their formation. The two ribose
phosphates obtained were separated by an ion-exchange procedure.'*^ The
ribose phosphate "a" (derived from adenyhc acid "a") could be converted
to a methyl phosphoribopyranoside which consumed one mole of periodate,
and to a ribitol phosphate with a marked optical activity which is en-
hanced by borate. The reverse properties (i.e., no periodate oxidation of the
methyl phosphoriboside, no optical activity of the ribitol phosphate with
or without borate) were noted for the "b" ribose phosphate. The possi-
bility of the 1- or 5-phosphate isomers arising was excluded by the ion-
exchange behavior of the substances, and the 4-phosphate ester is a 'priori
excluded by the furanoside structure of the parent nucleotide.^^" Hence it
follows that ribose phosphate "a" and *'b" are ribose-2- and -3-phosphate,
respectively. This is the first isolation of ribose-2-phosphate. Moreover
the work proved that Levene and his colleagues in their earlier structural
studies of the purine nucleotides were dealing with the "b" isomers which
would be expected to give rise to ribose-3-phosphate if no migration oc-
curred during isolation.

Initial attempts by Levene and Jorpes^^^ to prepare ribose-3-phosphate
by acidic hydrolysis of adenosine-3 '-phosphate, were unsuccessful as
cleavage of the basic and phosphate residues proceeded at about the same

206 D. M. Brown and A. R. Todd, J. Chem. Soc. 1952, 44, 52.

206 D. M. Brown, D. I. Magrath, and A. R. Todd, J. Chem. Soc. 1952, 2708.

207 W. E. Cohn, /. Am. Chem. Soc. 72, 2811 (1950).

208 D. M. Brown, L. J. Haynes, and A. R. Todd, J. Chem. Soc. 1950, 408.

209 D. G. Doherty, Abstracts Papers 118th Meeting Am. Chem. Soc. 56 (1950).

2>o p. A. Levene and R. S. Tipson, J. Biol. Chem. 94, 809 (1932); 97, 491 (1932); 101,

529 (1933).
2" P. A. Levene and E. Jorpes, J. Biol. Chem. 81, 575 (1929).

CHEMISTRY OF RIBOSE AND DEOXYRIBOSE 39

rate. However, Levene and his co-workers'^^- -'- were able to isolate a crys-
talline dibrucine salt of a ribose phosphate (not identical with the known
ribose-5-phosphate) from a hydrolysate of xanthylic acid. The xanthylic
acid was obtained by treatment of guanylic acid with nitrous acid: after
deamination the sugar-base linkage is found to be more labile. Reduction
of the ribose phosphate yielded an optically inactive ribitol phosphate,-'^
considered to be ribitol-3-phosphate, derived from ribose-3-phosphate. This
structural proof makes no account for possible migration of the phosphate
group during any of the experimental operations. Deamination of adenosine-
3'-phosphate to the corresponding inosinic acid, followed by hydrolysis,
also yields some ribose-3-phosphate.-'^

By methanolysis of yeast nucleic acid, Levene and Harris-'^ obtained a
crude sample of a methyl D-ribopyranoside-3-phosphate. Exhaustive meth-
ylation and subsequent dephosphorylation afforded methyl 2,4-di-O-
methyl-D-riboside, which on successive hydrolysis and catalytic reduction
yielded 2,4-di-O-methyl-D-ribitol. Although this compound is a meso
structure and would be expected to be optically inactive, the product did
exhibit some optical activity, and it was shown that the crude methyl d-
ribopyranoside-3-phosphate was contaminated with a furanoside isomer
which gave rise to optically active 2 , 5-di-O-methyl-D-ribitol as an impurity
in the 2 , 4-di-O-methyl derivative.

LePage and Umbreit-'^ have prepared ribose-3-phosphate by acidic hy-
drolysis of a pure adenosine triphosphate isolated from the autotrophic
bacterium Thiohacillus thio-oxydans.

(3) Ribose-5-phosphate. Ribose-5-phosphate was first obtained by Levene
and Jacobs'' by subjecting the barium salt of inosinic acid to acidic hy-
drolysis. The pentose phosphate was isolated as the crystalline hydrated
barium salt. Shortly afterwards the same workers'^ showed that oxidation
of the pentose phosphate with either bromine or nitric acid yields a phos-
phoribonic acid. If position C-5 had been unsubstituted, nitric acid oxida-
tion would have been expected to produce a trihydroxyglutaric acid, and
so the phosphate residue was considered to be located at C-5 of the ribose
molecule. Much later^'^ further evidence was forwarded which substan-
tiated this conclusion, since lactonization of the D-ribonic acid phosphate
proceeded very slowly, equilibrium being reached only after 150 hours.
This is the behavior expected of a pentonic acid substituted at C-5 and
unable to form other than a 7(1 ,4)-lactone.-'^' ^'^ Furthermore, reduction

212 P. A. Levene and A. Dmochovvski, J. Biol. Chcm. 93, 563 (1931).

213 p. A. Levene and S. A. Harris, J. Biol. Chem. 98, 9 (1932).

2>4 P. A. Levene and S. A. Harris, J. Biol. Chcm. 101, 419 (1933).
2'5 G. A. LePage and W. W. Umbreit, J. Biol. Chem. 148, 255 (1943).
2>« P. A. Levene and T. Mori, J. Biol. Chem. 81, 215 (1929).
217 p. A. Levene and H. S. Simms, /. Biol. Chem. 65, 31 (1925).

40 W. G. OVEREND AND M. STAGEY

of the pentose phosphate afforded an optically active phosphoribitol, and
so position C-3 of the sugar molecule was excluded as the site of esterifica-
tion. In solution in methanolic hydrogen chloride the pentose phosphate
underwent mutarotation in a manner characteristic of a sugar which
can only form a furanoside.-'** Moreover the behavior of the pentose phos-
phate isolated from natural sources and a synthetic sample of D-ribose-5-
phosphate was identical. The synthetic material was prepared'" from d-
ribose by condensation with acetone and methanol in the presence of
hydrogen chloride and anhydrous copper sulfate (or alternatively dilute
sulfuric acid) followed by phosphorylation of the resultant syrupy methyl
2,3,-0-isopropylidene-D-ribofuranoside (XXXVI) with phosphorus oxy-
chloride and pyridine at —40°. Hydrolysis of the isopropylidene and
glycosidic groups from methyl 2,3-0-isopropylidene-D-ribofuranoside-5-
phosphate (XXXVII) yielded ribose-5-phosphate (XXXVIII) (as barium
salt). The structure of the important intermediate XXVI was demon-
strated by methylation and hydrolysis to an amorphous monomethylribose
(XXXIX) which was converted to a crystalline p-bromophenylosazone,
identical with the p-bromophenylosazone of authentic 5-0-methyl-D-
ribose which had been previously prepared--" (see p. 42).

Using as phosphorylating agent dibenzylphosphorochloridate in drj^
pyridine solution at —40°, Michelson and Todd'^° considerably improved
the synthesis. Protecting groups were removed from the intermediate XL
by hydrogenation and hydrolysis to give XXXVIII in 86% yield.

An improved method for the preparation from muscle (horse, dog and
rabbit muscle are excellent sources) of inosinic acid and thence ribose-5-
phosphate has recently been described.--' Optimum conditions were deter-
mined for hydrolysis of the nucleotide to ribose phosphate which can be
obtained in a yield of 50-60% by use of this method. This phosphate
ester is also obtainable by acidic hydrolysis of cozymase.*^ Adenine and
nicotinamide are cleaved quantitatively while 20 % of the total phosphorus
is liberated. Adenine was removed as its silver salt and the ribose-5-phos-
phate was isolated as the barium salt. The isolation of this compound
served to identify the sugar moiety in cozymase and also the site at which
it was esterified by the phosphoric acid residue. Similar conclusions were
reached from less direct evidence by other investigators.^^ Ribose-5-phos-
phate has been obtained in a high degree of purity from adenosine tri-
phosphate.-^^ Purification was achieved by chromatography on ion-exchange

218 P. A. Levene and M. L. Wolfrom, J. Biol. Chem. 77, 671 (1928).

2" P. A. Levene, S. A. Harris, and E. T. Stiller, J. Biol. Chem. 105, 153 (1934).

220 P. A. Levene and E. T. Stiller, J. Biol. Chem. 102, 187 (1933).

221 J. Marmur, F. Schlenk, and R. N. Overland, Arch. Biochem. and Biophys. 34, 209
(1951).

222 D. P. Groth, G. C. Mueller, and G. A. LePage, J. Biol. Chem. 199, 389 (1952).

CHEMISTRY OF RIBOSE AND DEOXYRIBOSE

41

HO-CHj 'Os

H H>H-OMe

HO
0 = P0-CH2 ^0

H-OH

OH OH
XXXIX

OH OH
XXXVIII

A

PhCHsO,
PhCHsO'

:PO-CH

0

H-OMe

resin (monochloroacetate of Dowex 1). The overall yield of the pentose
phosphate from adenosine triphosphate was 63-70 %. As previously men-
tioned, ribose-5-phosphate can be obtained from ribose-1 -phosphate by the
mutase action of liver^"^ and yeast-"' extracts. D-Gluconic acid 6-phosphate
may be degraded enzymically to D-ribose-S-phosphate'**^- "*** and this
observation is relevant to any discussion on possible routes--^ for the
biogenesis of D-ribose. [Cf. Glock, Chapter 22.] Ribose-5-phosphate was
found to be vigorously metabolized by animal and yeast extracts,-'' • ^^^
but arabinose-5-phosphate and xylose-5-phosphate were only slightly
attacked.

(4) Other Pentose Phosphates. During the researches which led to the identification

223 I. A. Bernstein, J. Am. Chem. Soc. 73, 5003 (1951).

"^ F. Dickens, Biochem. J. 32, 1626, 1645 (1938).

2" F. Dickens and Gertrude E. Glock, Nature 166, 33 (1950).

42 W. G. OVEREND AND M. STACEY

of D-ribose as the sugar component of yeast nucleic acid, several other pentose phos-
phates were prepared. Interest in xylose phosphate stemmed from Robinson's"*
suggestion that xylose might be a primary constituent of nucleic acids and that
ribose might result by hydrolysis of xylose-3-phosphate with Walden inversion. This
possibility does not apply to ribose-5-phosphate isolated from inosinic acid because
a Walden inversion at the primary hydroxyl group would not affect the configuration
of the sugar. Attempts by Levene and Raymond"* to prepare xylose-3-phosphate for
testing the validity of Robinson's hypothesis, were unsuccessful. Phosphorylation
of xylose derivatives with only the hydroxyl group at C-3 free, .gave xylose-5-phos-
phate derivatives, and obviously migration occurred at some stage of the preparation
via an intermediate cyclic diester structure. The only 3-phosphate derivative which
they were able to isolate was l,2-0-isopropylidine-3-phosphate-5-0-methylxylose.
Owing to the difficulty of removing the methyl group, this derivative was of little
value for comparison with the phosphopentose derived from nucleic acid. The rate
of dephosphorylation for this compound was many times greater than that of xylose-
5-phosphate.

Xylopyranose-1 -phosphate (isolated as the barium or dipotassium salt) was pre-
pared by reacting bromoacetylxylose and trisilver phosphate and subsequent partial
hydrolysis."'

Generally pentose phosphate esters are stronger acids than free phosphoric acid
and the values of pKi' and pKi' are smaller. For example, the dissociation constants
of xylose-1 -phosphate calculated by means of Van Slyke's"' formula and the Hen-
derson-Hasselbach equation are pKi' = 1.25 and pK^' = 6.15. Comparative values
for phosphoric acid are pKi' = 1.95-2.00 and pKi' = 6.83-6.93.2"-23o

D-Arabinose-5-phosphate was synthesized by Levene and Christman.'*'

(e) Ethers, Esters, Acetals and Anhydrides.

(1) Ethers. Reference has already been made to the complete methyla-
tion of methyl D-ribf)side and to the fact that acidic hydrolysis of the
product yields crystalline 2,3,4-tri-O-methyl-D-ribose.^^^' ^^^ Proof of struc-
ture followed from nitric acid oxidation which gave f-trimethoxyglutaric
acid. An isomeric tri-0-methyl-D-ribose was obtained by Levene and
Tipson by subjecting either adenosine^'" or guanosine^^" to methylation
and subsequent hydrolysis. The amorphous product reacted more rapidly
with acidic methanol than the 2,3,4-isomer and could be converted to a
7-lactone. Nitric acid oxidation afforded i-dimethoxysuccinic acid, and on
the basis of these results the sugar was considered to be 2 , 3 , 5-tri-O-methyl-
D-ribose, a conclusion subsequently confirmed by synthesis studies.^^''
Methyl 2 , 3-0-isopropylidene-5-0-methyl-D-ribofuranoside was prepared by
successive acetonation and methylation of methyl D-ribofuranoside. Hy-
dolysis, further methylation and rehydrolysis furnished 2,3,5-tri-O-

226 R. A. Robinson, Nature 120, 44 (1927).

2" D. D. Van Slyke, J. Biol. Chem. 52, 525 (1922).

228 O. Meyerhof and J. Suranyi, Biochem. Z. 178, 427 (1926).

229 C. F. Cori, S. P. Colowick, and Gerty T. Cori, J. Biol. Chem. 121, 465 (1937).

230 H. T. S. Britton and R. A. Robinson, Trans. Faraday Soc. 28, 531 (1932).
2" P. A. Levene and J. Compton, J. Biol. Chem. 116, 169 (1936).

CHEMISTRY OF RIBOSE AND DEOXYRIBOSE 43

methyl-D-ribose, identical with that obtained from adenosine and guanosine.
The structure of the synthetic material was rigidly established by the usual
methods of carbohydrate chemistry. Further confirmation of the structure
has since been provided by Barker. '^'' Benzylation of methyl 2,3-0-iso-
propylidene-D-ribofuranoside afforded the 5-benzyl ether, which on hydroly-
sis was converted to syrupy 5-0-benzyl-D-ribose.-^^ Originally it was
reported by Bredereck et al.^^ that direct tritylation of D-ribose in pyridine
solution yields crystalline 5-0-trityl-a-D-ribose, and Barker and Lock^^^
demonstrated that acetylation without prior isolation of the trityl deriva-
tive tripled the yield of acetylated trityl ether. An improved alternative
preparation of S-O-trityl-cc-D-ribofuranose was reported by Zinner."^ In
a subsequent publication Bredereck and Greiner-^^ described methods for
the preparation of several trityl ethers (e.g., 1-, 5-, 1,3(2?)-, 1,5- and
1 ,3,5-) of D-ribose and of their acetyl and benzoyl derivatives.

(2) Esters. The acetates and benzoates of D-ribose have been extensively
investigated. Acetylation in pyridine solution at low or ordinary tem-
peratures results in the formation of crystalline 1 ,2,3,4-tetra-0-acetyl-(S-
D-ribopyranose.^•■^• 110 '235 When acetic anhydride and sodium acetate were
used as acetylating agents at higher temperature the product was 1,2,3,5-
tetra-0-acetyl-D-ribofuranose.'^' Zinner'^^ investigated the acetylation with
acetic anhydride of ribose in pyridine solution at various temperatures and
found that increase in the reaction temperature is accompanied by more
formation of the furanose isomer, so that at 100° the proportions of the
furanose and pyranose forms are approximately equal. The first synthesis
of crystalline D-ribofuranose tetraacetate was accomplished successfully
by Howard ef a/.'^^ by reductive detritylation of 1 ,2,3-tri-0-acetyl-5-0-
trityl-D-ribo furanose (XLI) and subsequent acetylation. The tetraacetate is

OAc

/
H— C

H— C— OAc

H— C— OAc

I
H— C— O

CH20CPh3
XLI

"2 G. W. Kenner, C. W. Taylor, and A. R. Todd, J. Chem. Soc. 1949, 1620.

2" G. R. Barker and M. V. Lock, J. Chem. Soc. 1950, 23.

"^ H. Bredereck and W. Greiner, Chem. Ber. 86, 717 (1953).

"5 H. Zinner, Chem. Ber. 86, 817 (1953).

44 W. G. OVEREND AND M. STAGEY

a useful intermediate in the synthesis of naturally occurring ribonucleosides
and this offers some confirmation for the structure assigned. The method was
improved, so that the product could be isolated directly by accomplishing
simultaneous detritylation and acetylation of XLI with acetyl bromide in
acetic anhydride.^^* 5-0-Benzyl-D-ribofuranose can also be used as an
intermediate for the preparation of this acetate. ^'^

Some confusion has arisen in the literature concerning D-ribofuranose tetra-0-
acetate. Thus, for this compound Howard et a/.'^^ report m.p. 58° and [a]^^ +20°
(in chloroform), Bredereck and Hoepfner"« state that the physical constants are
m.p. 56° and [afo —3.6° (in methanol), whereas Zinner'" describes a product of m.p.
82° and [a\o -12.6° (in chloroform) and —15.4° (in methanol). Zinner'". "6 referred
to his product as /3-tetra-O-acetylribofuranose, and it obviously differed from the
acetate prepared by Howard et al.^** and Bredereck and Hoepfner.^e Davoll et al.'^^''
have described some experiences with the preparation of this acetate by the Bredereck
method. The first three portions prepared had m.p. 56-58°, but the fourth and all
subsequent batches had m.p. 85° and [a]" —12° (in chloroform) and —13.5° (in meth-
anol), and were obviously identical with the product prepared by the method of
Zinner. After the isolation of the latter product the melting points of earlier samples
spontaneously changed to 85° and it became impossible to prepare samples of the
acetate of lower melting point. Similarly, samples of products of lower m.p. forwarded
to Davoll from other laboratories changed into the higher melting form. According
to Davoll et al.'^^'' the change was accompanied by a change in optical rotation. The
furanose structure of the material of higher melting point was proved by its conver-
sion through the aceto-chloro-compound to adenosine in fields comparable to those
obtained using tetra-O-acetyl-D-ribofuranose of m.p. 56-58° as the initial material.
Davoll and his colleagues were unable to determine the difference between the iso-
mers, but did not think that it was due to simple a,/3-isomerism, since according to
them measurements of optical rotation changes indicated that a complex process was
operating in which at least three molecular species were involved. They considered it
possible that one of the isomers had an orthoacetate anhj'dride structure, but this is
difficult to reconcile with the stability of both forms to water and ethanol. Farrar^"*
considers that speculations regarding possible structures of the two forms are un-
necessary, and all that is involved is a simple, but interesting, case of dimorphism,
since she found that isomerization proceeded without any significant change in
optical rotation. Obviously a contradiction exists between this result and the rotation
changes reported by Davoll et a/.,^'^ and the problem needs to be investigated further.

2 , 3 , 4 , 5-Tetra-0-acetyl-a/c/e/i?/£/o-D-ribose has been obtained crystalline
and can be prepared by several routes. Demercaptalation of 2,3,4,5-tetra-
0-acetyl-D-ribose diethyl thioacetal,-^* hydrogenolysis with Raney nickel
of ethylthio-D-ribonate tetra-O-acetate^^^ or Rosenmund reduction of
tetra-0-acetyl-D-ribonyl chloride, all yield this compound. ^^^ Attempts to

236 H. Bredereck and Eva Hoepfner, Chem. Ber. 81, 51 (1948).

2" J. Davoll, G. B. Brown, and D. W. Visser, Nature 170, 64 (1952).

""* Kathleen R. Farrar, Nature 170, 896 (1952).

238 H. Zinner, Che7n. Ber. 83, 418 (1950).

"9 M. L. Wolfrom and J. V. Karabinos, /. Am. Chem. Soc. 68, 724, 1455 (1946).

CHEMISTRY OF RIBOSE AND DEOXYRIBOSE 45

prepare 2,3,5-tri-O-acetyl-D-ribose by hydrolysis of arylamine derivatives
have afforded only crude amorphous samples of the ester.^- ^°' ^^^ Recently,
Zinner^'^ has reported the preparation of all the tetra-0-acetates of D-ribose.
Methods for obtaining tetra-O-acetyl-D-ribonic acid have been investi-
gated by Tishler et al.'^*° Acetylation of the acid according to the procedure
of Robbins and Upson^*' resulted in the production of some tetra-O-acetyl-
D-ribonic acid in low yield (15%), admixed with tri-0-acetyl-D-ribonolac-
tone (10%) and an intractable oil. A method was developed which gave
the required acetate in high yield. Cadmium D-ribonate was treated at 10°
with acetic anhydride and hydrogen chloride and the crystalline ribonic
acid tetra-0-acetate was obtained in 85% of the theoretical yield. The
cation is important in this acetylation, as shown by the varying yields
of product obtained when other salts were used. When made to react with

Cation Ba++ Ca++ K+ NH4+ Cd++

Yield of product (%)... 4 22 25 46 85

phosphorus oxychloride in chloroform, tetra-0-acetyl-D-ribonamide was
converted to tetra-0-acetyl-D-ribononitrile.^*°

Knowledge of the benzoates of D-ribose is due mainly to the researches
of Hudson, Fletcher and their co-workers. A crystalline tetra-0-benzoate
was obtained by benzoylation of the sugar in p3a-idine solution at low
temperature."^ This was considered to have the /3-configuration and was
demonstrated to have a pyranose structure by conversion into tri-0-
benzoyl-D-ribosyl bromide, which was condensed with the potassium salt of
2-thionaphthol to give 2'-naphthyl-l-thio-/3-D-riboside tri-0-benzoate.
DesuKurization with Raney nickel and subsequent debenzoylation af-
forded an anhydroribitol in high yield. This product resembled 1 , 5-an-
hydro-D-xylitoP^^ and 1 , 5-anhydro-D-arabitoP^^ in its solubility characteris-
tics. Neither the anhydro derivative nor its acetyl or benzoyl derivatives
exhibited optical activity. This would be expected if it had a meso structure,
as would be the case if the original tetra-0-benzoate had a pyranose(l ,5)-
lactol ring structure. Periodate oxidation on the anhydroribitol confirmed
this lactol ring structure. Hydrolysis of 2,3,4-tri-O-benzoyl-D-ribosyl
bromide with moist acetone in the presence of an acid-acceptor (Ag2C03)
yielded 2,3,4-tri-O-benzoyl-D-ribose.i^^' ^''^ The isomeric 2,3,5-tri-O-ben-
zoyl derivative has also been prepared in crystalline form.^^* Esters of

"0 K. Ladenburg, M. Tishler, J. W. Wellman, and R. D. Babson, /. Am.. Chem. Soc.

66, 1217 (1944).
"1 G. B. Robbins and F. W. Upson, /. Atn. Chem. Soc. 60, 1788 (1938).
«« H. G. Fletcher, Jr., and C. S. Hudson, J. Am. Chem. Soc. 69, 921 (1947).
"« H. G. Fletcher, Jr., and C-. S. Hudson, J. Am. Chem. Soc. 69, 1672 (1947).
*" R. K. Ness and H. G. Fletcher, Jr., J. Am. Chem. Soc. 75, 3289 (1953).

46

W. G. OVEREND AND M. STACEY

D-ribose containing both acetate and benzoate groupings are known.
Kenner et al}^^ synthesized 2,3,4-tri-0-acetyl-5-0-benzoyl-D-ribose, and
l-0-acetyl-2,3,5-tri-0-benzoyl-D-ribose has been prepared from guano-
sine.2^®

(3) Acetals. Two products are obtained when D-ribose is treated with
acetone in the presence of sulfuric acid and anhydrous copper sulfatej^^"
one of which is crystalHne and the other syrupy. The crystalhne material
is unreactive and is considered to be 2,3-0-isopropylidene-l ,5-anhydro-D-
ribofuranose. The syrupy material, which is the major product, can be
purified via its di-0-acetyl derivative and by the usual methods of classical
carbohydrate chemistry and has been shown to be 2,3-0-isopropylidene-
D-ribofuranose. If D-ribose is treated with acetone and methanol in the
presence of sulfuric acid and anhydrous copper sulfate, methyl 2,3-0-iso-
propylidene-D-ribofuranoside is obtained directly. ^^* This compound has
been used as an intermediate in syntheses of naturally occurring ribose
derivatives. Attempts to acetonate methyl D-ribopyranoside gave mixtures
of methyl mono-0-isopropylidene-D-ribo-furanoside and -pyranoside, and
obviously some rearrangement of the lactol ring occurred during the
reaction. ^^^

(4) Anhydrides. In an attempt to prepare 1 ,2,3-tri-0-acetyl-D-ribo-
f uranose by detritylation of 1,2, 3-tri-0-acetyl-5-0-trityl-D-ribof uranose
with hydrogen bromide in glacial acetic acid at 0°, Bredereck et al.^^ iso-
lated an acetate which on hydrolysis afforded an anhydride of D-ribose.
The structure assigned to this compound by these workers was D-ribo-
san(l,4) (3(1,5) (XLII). The anhydride gave a positive test for vicinal

CH

H— C— OH

H— C— OH

H— C— O—

— OCH2
XLII

CH-

H— C— OH

H— C— OH

H— C— 0-

-OCH2

CH2O

-O— C— F

HO— C— H

HO— C— H

:CH

XLIII

hydroxyl groups and was reducing to Fehling's solution after acidic hy-
drolysis. Barker and Lock^^^ and Jeanloz^"*^ reexamined the compound and
found that it was bimolecular. A crystalline tetra-0-methyl-di-D-ribose

^^^ G. W. Kenner, H. J. Rodda, and A. R. Todd, /. Chem. Soc. 1949, 1613.

"6 F. Weygand and W. Sigmund, Chem. Ber. 86, 160 (1953).

2" P. A. Levene and E. T. Stiller, /. Biol. Chem. 106, 421 (1934).

^^ R. W. Jeanloz, G. R. Barker, and M. V. Lock, Nature 167, 42 (1951).

CHEMISTRY OF RIBOSE AND DEOXYRIBOSE 47

anhydride was prepared which hydrolyzed to 2,3-di-O-methyl-D-ribose.
Oxidation of the anhydride with sodium metaperiodate was also investi-
gated, and all the evidence was consistent with the structure of the com-
pound being most probably 1 ,5'-5, 1'-diribofuranose anhydride (XLIII).
As already briefly mentioned, acetonation of ribose in the presence of
sulfuric acid and acetic anhydride leads to the production of a small
amount of a product assumed to be 1 ,5-anhydro-2,3-0-isopropylidene-D-
ri]>ofuranose. If hydrogen chloride is used as condensing agent, an isomeric
compound is obtained having a lower melting point. '°^ Neither compound
has been fully investigated. Methyl-2,3-anhydroribosides have been pre-
pared by treatment of 2-0-tosyl or 2-0-mesyl derivatives of methyl arab-
inoside with sodium methoxide.^''^-"" This anhydro compound has been
used as an intermediate in attempted syntheses of 2-deoxyribose as de-
scribed on pages 50-51. Hot aqueous sodium hydroxide solution converts
methyl 2,3-anhydro-/3-L-riboside into methyl ^S-L-xyloside (75%) and
methyl /3-L-arabinoside (25%).-^^ The synthesis and establishment of
structure of 1 ,5-anhydroribitol has already been described'" (p. 45).
When D-ribobenzimidazole is heated with zinc chloride and concentrated
hydrochloric acid at 180°, anhydrization of the sugar molecule occurs to
give 2- ( 1 ' , 4 '-anhydro-D-n6o-tetrahy droxybutyl)ben zimidazole .-^^

(/) Other Properties.

D-Ribose reacts very readily with alkyl- and aryl-thiols in the presence
of hydrochloric acid.'- "^ Mention has already been made to the conver-
sion of ribose into brom-acetyl (and benzoyl) ribose and to uses of the
compounds. (See also Baxter et a/."^.) Various references to the action of
oxidizing agents on this pentose have already been rjuoted during the
course of this review. In general the reactions seem to proceed as expected.
It has been reported^"*" that D-ribonic acid is somewhat unstable at room
temperature, as indicated by a lowering of the melting point by several
degrees after storage for 24 hours. The acid can be converted into a lac-
tone, the amide and esters. The tetra-O-acetate of D-ribonic acid when
acted on by phosphorus pentachloride gives D-ribonyl chloride tetra-O-
acetate. This latter compound can be converted into /ce/o-D-psicose penta-
0-acetate by reaction with diazomethane in ether and subsequent treat-

"' J. Honeyman, /. Chem. Soc. 1946, 990.

"0 S. Mukherjee and A. R. Todd, /. Chem. Soc. 1947, 969.

"1 P. W. Kent, M. Stacey, and L. F. Wiggins, /. Chern. Soc. 1949, 1232; Nature 161,

21 (1948).
"2 R. Allerton and W. G. Overend, /. Chem. Soc. 1951, 1480.
2" C. F. Huebner, R. Lohmar, R. J. Dimler, S. Moore, and K. P. Link, /. Biol. Chem.

159, 503 (1945).
2" R. A. Baxter, A. C. McLean, and F. S. Spring, /. Chem. Soc. 1948, 523.

48 W. G. OVEREND AND M. STACEY

ment of the 1-diazo-l-deoxy-fce/o-D-psicose tetra-0-acetate so formed, with
copper acetate and acetic acid.^^*

Soon after the first preparation of D-ribose, Fischer^^* succeeded in re-
ducing ribonolactone through the free sugar to a polyol (ribitol, adonitol).
Catalytic reduction with Raney nickel of aldehydo-D-rihose 2,3,4,5-tetra-
0-acetate afforded 2,3,4,5-tetra-0-acetylribitol.2" A method has been
patented for the hydrogenation of sugars to the corresponding polyol by
magnesium-activated Raney nickel and in this way ribose can be converted
to ribitol practically quantitatively."^

The condensation of nitromethane with D-ribose has been studied by
Sowden and Fischer,^ • ^^^ and the results have been discussed in a recent
extensive review.^^" Like other pentoses D-ribose is converted to furfural
when heated with dilute acid: quantitative aspects of the conversion have
been investigated.^*' The behavior of ribose in the Kiliani-Fischer synthesis
has been reviewed by Hudson^*^ (see also Richtmyer^^.

rv. Chemistry of 2-Deoxyribose

1. Preparation

The isolation of 2-deoxy-D-ribose from deoxyribonucleic acid has proved
to be very difficult. In early experiments it was usual to degrade the nucleic
acid by chemical methods to the constituent nucleosides, separate these,
and then hydrolyze the glycosidic linkage and isolate the sugar portion.
Separation of the nucleosides w^as a tedious procedure, and, since acidic
treatment for hydrolysis results irl the conversion of some of the deoxy-
pentose to levulinic acid, yields were poor. Following attempts by Thann-
hauser and Ottenstein,^®^ who employed picric acid for the hydrolysis of
thymus nucleic acid and obtained diphosphoric esters of pyrimidine deoxy-
ribosides, Levene and London^^ resorted to enzymic methods of degrada-
tion and isolated deoxyribonucleosides of guanine, hypoxanthine (arising
from deamination of adenine), cytosine and thjrmine. Very mild hydrolysis
of the guanine nucleoside gave the deoxysugar in crystalline form.^**- ^^^

2" M. L. Wolfrom, A. Thompson, and E. F. Evans, /. A7n. Chetn. Soc. 67, 1793 (1945).

266 E. Fischer, Ber. 26, 633 (1893).

2" H. H. Fox, /. Org. Chem. 13, 580 (1948).

"8 L. A. Flexser, U. S. Pat. 2,421,416 (June 3, 1947).

"9 J. C. Sowden and H. O. L. Fischer, U. S. Pat. 2,480,785 (Aug. 30, 1949).

26' J. C. Sowden, Advances in Carbohydrate Chem. 6, 291 (1951).

261 R. C. Hockett, A. Guttag, and M. E. Smith, J. Am. Chem. Soc. 66, 1 (1943).

262 C. S. Hudson, Advances in Carbohydrate Chem. 1, 1 (1945).

263 S. J. Thannhauser and B. Ottenstein, Z. physiol. Chem. 114, 17, 39 (1921).

264 p. A. Levene and E. S. London, /. Biol. Chem. 83, 793 (1929).
266 P. A. Levene and T. Mori, /. Biol. Chem. 83, 803 (1929).

266 p. A. Levene, L. A. Mikeska, and T. Mori, /. Biol. Chem. 86, 785 (1930).

CHEMISTRY OF RIBOSE AND DEOXYRIBOSE 49

Similar mild acidic hydrolysis of hypoxanthine deoxyriboside gave a solu-
tion with the same optical rotation as an equivalent amount of 2-deoxy-D-
ribose.^^ Attempts to isolate the sugar from the pyrimidine deoxyribonu-
cleosides were unsuccessful, as the sugar was immediately converted into
levulinic acid under the more drastic hydrolytic conditions necessary to
cleave the linkage between the pyrimidine base and the deoxysugar.^^''
Development of an improved enzymic hydrolysis procedure^" and of chro-
matographic and ion-exchange methods-®*'^" for the separation of deoxy-
ribo-nucleotides and -nucleosides has afforded the possibility of obtaining
these products in good yield in a high degree of purity. The writers and
colleague"^ have succeeded in developing a method for obtaining 2-deoxy-
D-ribose in fair yield by acidic hydrolysis of purine deoxyribonucleosides
which had been separated by ion-exchange resin chromatography. Enzymic
evidence has been provided"'* to show that deoxyribonucleotides isolated
by the above methods are esterified at carbon atom 5 of the sugar moiety.
Consequently, there is also the possibility that in the near future 2-deoxy-
D-ribose-5-phosphate will be obtainable from nucleic acid.

Kent"^^ demonstrated that mercaptanolysis of deoxyribonucleic acids
resulted in the liberation of the sugar which was isolated as the dibenzyl
mercaptal.

The best known and probably still the most direct method for the syn-
thesis of 2-deoxy-D-ribose is the "glycal" method, which is a general
method for the synthesis of 2-deoxysugars. In this reaction, the elements
of water are added to the olefinic linkage in a glycal by treatment with
dilute sulfuric acid at low temperature and in this way D-arabinal (X) has
been converted into 2-deoxy-D-ribose."^""^ Likewise the conversion has
been effected in the L-series.'*- ^^^' "''• ^^° If the glycal is treated with a 2-3 %
solution of hydrogen chloride in methanol instead of with dilute aqueous

2" W. Klein, Z. physiol. Chem. 218, 164 (1933).

"» O. Schindler, Helv. Chim. Acta 32, 979 (1949).

269 P. Reichard and B. Estborn, Acta Chem. Scand. 4, 1047 (1950).

"» E. Volkin, J. X. Khym, and W. E. Cohn, J. Am. Chem. Soc. 73, 1533 (1951).

"1 R. L. Sinsheimer and J. F. Koerner, Science 114, 42 (1951).

2" W. Andersen, C. A. Dekker, and A. R. Todd, /. Chem. Soc. 1952, 2721.

"3 S. G. Laland and W. G. Overend, Acta Chem. Scand., 8, 192 (1954).

"♦ C. E. Carter, J. Am. Chem. Soc. 73, 1537 (1951).

"<» P. W. Kent, Nature 166, 442 (1950).

"6 G. E. Felton and W. Freudenberg, J. Am. Chem. Soc. 57, 1637 (1935).

"« A. M. Gakhokidze, Zhur. Obshchel Khim. 15, 539 (1945).

2" R. E. Deriaz, W. G. Overend, M. Stacey, Ethel G. Teece, and L. F. Wiggins,

J. Chem. Soc. 1949, 1879.
"8 K. Ohta, J. Biochem. {Japan) 38, 31 (1951).
"9 K. Ohta and K. Makino, Science 113, 273 (1951).
"s" J. Meisenheimer and H. Jung, Ber. 60, 1462 (1927).

50 W. G. OVEREND AND M. STACEY

acid, then the methyl glycoside of the 2-deoxysugar is obtained. For
example, L-arabinal was converted into methyl 2-deoxy-/3-L-ribopyrano-
side.2^^ The reactions involved in the glycal synthesis have been studied in
considerable detail, especially in the conversion of arabinose to arabinal
and then into 2-deoxyribose, and the overall yield of this particular con-
version has been doubled by recently introduced improvements, but it is
still very low.^"

2-Deoxy-D-ribose has been prepared by using D-erythrose or its deriva-
tives as the initial material.^- ^ Sowden* used both 2,4-0-benzylidene-D-
erythrose and D-erythrose as initial materials. The former was condensed
with nitromethane and sodium methoxide to give a mixture of 3,5-0-
benzylidene-1-nitro-l-deoxy-D-ribitol and -D-arabitol; these were separated
by chloroform extraction. Hydrolysis and acetylation of the arabitol
derivative resulted in the formation of 2,3,4,5-tetra-O-acetyl-l-nitro-l-
deoxy-D-arabitol, which was converted into D-er?/^/iro-triacetoxy-l-nitro-
pentene by boiling under reflux in benzene solution with sodium bicarbon-
ate. Similarly this compound could be prepared from D-erythrose.^' ^
Reduction of the pentene derivative, followed by treatment with sulfuric
acid yielded 2-deoxy-D-ribose. Sowden^ purified the product by forming
the benzylphenylhydrazone and regenerated the deoxypentose by treating
with either benzaldehyde or formaldehyde. Overend and co-workers®
favored formation of the aniline derivative ("anilide") as a means of isola-
tion and purification. Deanilination was effected by treatment with 0.5%
oxalic acid in aqueous solution, and 2-deoxy-D-ribose was obtained in
crystalline form. Good yields are obtained at all stages of this synthesis,
and for preparative purposes Sowden* claims that the isolation of inter-
mediates is unnecessary. The method would be a valuable one for the
preparation of 2-deoxy-D-ribose, if D-erythrose was obtainable in a pure
state in large quantities.

Attempts have been made to convert methyl 2,3-anhydroriboside (XLIV) into
methyl 2-deoxy-D-riboside. The anhydro ring in XLIV can theoretically cleave in
two ways, giving rise to two different products, i.e., a 2-substituted derivative of
methyl D-arabinoside (XLV) and a 3-substituted derivative of methj^l D-xyloside
(XLVI). By using appropriate reagents for the cleavage it is possible to obtain de-
rivatives which can be converted directly to 2- or 3-deoxysugar derivatives. The
protection afforded by the glycosidic residue increases the possibility of improved
overall yields of the deoxysugars.

Methyl 2,3-anhydro-/3-L-ribopyranoside was caused to react with sodium methyl-
mercaptide and thereafter the product was boiled under reflux with Raney nickel
to effect catalytic desulfurization of the mixture of methyl 2-methylthio-2-deoxy-
/3-L-arabinoside (XLV, R = SMe, D-isomer) and methyl 3-methylthio-3-deoxy-/3-
L-xyloside (XLVI, R = SMe, D-isomer). The main deo.xypentoside obtained was

"1 R. E. Deriaz, W. G. Overend, M. Stacey, and L. F. Wiggins, J. Chem. Soc. 1949,
2836.

CHEMISTRY OF RIBOSE AND DEOXYRIBOSE

51

.CH2— O,

H-OMe

R-

CH2— 0.

HO

HO

H-OMe —

HO

j^ >H-OMe

CH2— O.

CH2 — O,

HyH-OMe

HO

H >H-OMe

HO

methyl 3-deoxy-/3-L-riboside (xyloside) (XLVI, R = H, D-isomer), and only traces
of methyl 2-deoxy-/3-L-riboside (XLV, R = H, D-isomer) were detected.^^o Change of
configuration at the glycosidic center had no effect in increasing the proportion of
the 2-deoxypentose derivative, since similar results were obtained with methyl
2,3-anhydro-a-L-ribopyranoside.^^'' Likewise change in the size of the lactol ring had
no effect on the orientation of the substituents.^*^

The action of halogen acids on alkyl 2,3-anhydroaldosides results in the formation
of an alkyl 2-halogeno- or 3-halogeno-deoxyaldoside which can readily be reduced
to give the corresponding alkyl 2- or 3-deoxyaldoside. Methyl 2,3-anhydro-/3-D-
riboside when treated with hydrobromic acid yielded mainly methyl 3-bromo-3-
deoxy-/3-D-xyloside (XLVI, R = Br) and only a small amount (10%) of methyl
2-bromo-2-deoxy-/3-D-arabinoside (XLV, R = Br). After separation of these isomers,
the methyl 2-bromo-2-deoxy-/3-D-arabinoside was subjected to catalytic hydrogena-
tion with Raney nickel in the presence of calcium hydroxide and afforded methyl 2-
deoxy-D-ribose. This on hydrolysis with dilute acetic acid yielded 2-deoxy-D-ribose
which was isolated as iV-phenyl-2-deoxy-D-ribosylamine. Allerton and Overend"*
showed that the action of hydrochloric acid on this anhydroriboside leads to a slightly
better yield (18.4%) of the methyl 2-halogeno-2-deoxy-D-arabinoside (XLV, R =
halogen), but these methods have no value for the large-scale synthesis of 2-deoxy-D
ribose.

Attempts to convert methyl anhydroriboside directly into methyl 2-deoxyriboside
have not been very successful. Reaction of methyl 2,3-anhydro-^-D-riboside with
lithium aluminum hydride gave mainly methyl 3-deoxy-/3-D-riboside and only a
small amount (14%) of methyl 2-deoxy-/3-D-riboside."2 On heating methyl 2,3-
anhydro-/3-L-riboside at 110° in hydrogen (at 100 atmospheres) with Raney nickel
cleavage of the anhydro ring occurred to yield predominantly methyl 3-deoxy-/3-L-
riboside, accompanied by only a small amount of the 2-deoxy-isomer."''

Recently Allerton and Overend^^^ succeeded in directly replacing by
hydrogen the p-toluenesulfonyloxy (or methanesulfonyloxy) group in

282 J. Davoll, B. Lythgoe, and S. Trippett, J. Chem. Soc. 1951, 2230.
^^^ R. Allerton and W. G. Overend, J. Chem. Soc, in press.

52 W. G. OVEREND AND M. STAGEY

methyl 2-0-p-toluenesulfonyl(or methanesulfonyl)-i3-L-arabinoside. The
complex mixture of products contained some methyl 2-deoxy-/3-L-riboside.
Since derivatives of 3-deoxy-D-glucose became available, efforts have been
directed towards the preparation of 2-deoxy-D-ribose from this deoxyhexose
and from either calcium or barium 3-deoxy-D-gluconate. It has been shown
in the writers' laboratory that when the modification of Ruff's method that
was introduced by Hockett and Hudson^- is employed, calcium 3-deoxy-D-
gluconate can be degraded to 2-deoxy-D-ribose. Richards^** has demon-
strated that the application of Ruff's method of degradation to calcium
c^ex^ro-metasaccharinate^*^ under the conditions described by Fletcher
et al."^^^ affords the 2-deoxy-D-ribose in satisfactory yield. Furthermore, it
is not necessary to separate a- from /3-metasaccharinic acid, since a mixture
of the two is equally effective. Hough^*^ has outlined a novel method for
the synthesis of 2-deoxy-D-ribose. An excess of allylmagnesium bromide was
allowed to react with 2,3-isopropylidene-D-glyceraldehyde and, after de-
composition of the resultant complex, syrupy 5,6-isopropylidene-l-
hexene-4 , 5 , 6-triol (XLVII) was obtained in excellent yield. This reaction
results in the formation of a new asymmetric center* at carbon atom 4. On
treatment of XLVII with a solution of hydrogen peroxide in tert-huty\
alcohol containing a little osmium tetroxide as catalyst, the double bond is
hydroxylated and a new center of asymmetry is produced at carbon atom 2.
From the mixture of products a fraction containing 5 , 6-0-isopropylidene-
3-deoxyhexitols was obtained. Oxidation of this fraction with sodium

CH2

II
CH

CH2

*CHOH

H— C— O CH3

\ /
C

/ \
H2C— O CH3

XLVII

metaperiodate destroyed the asymmetric center at carbon atom 2 and gave
a mixture of 4 , 5-isopropylidene-2-deoxypentoses which on acidic hydrolysis

28* G. N. Richards, Chemistry & Industry 1953, 1035.
28^ J. U. Nef, Ann. 376, 1 (1910).

286 H. G. Fletcher, Jr., H. W. Diehl, and C. S. Hudson, /. Am. Chem. Soc. 72, 4546
(1950).

287 L. Hough, Chemistry & Industry 1951, 406; J. Chem. Soc. 1953, 3066.

CHEMISTRY OF RIBOSE AND DEOXYRIBOSE 53

yielded the free sugars. The major component was 2-deoxy-D-ribose and this
was isolated as 7V-phenyl-2-deoxy-D-ribosylamine, from which the parent
sugar was easily regenerated. Another method for the preparation of 2-
deoxy-D-ribose, using 2,3-isopropylidene-D-glyceraldehyde as an initial
material has been outlined briefly by Overend and Stacey.^^^ The glycer-
aldehyde derivative was condensed with acetaldehyde in the presence of
anhydrous potassium carbonate and the products were subjected to mild
acidic hydrolysis to yield 2-deoxy-D-ribose and 2-deoxy-D-xylose. D-Arabi-
nose has been converted into 2-deoxy-D-ribose in 3 % overall yield by the
following sequence of reactions :"^^ by heating in pyridine D-arabinose was
converted into ribulose which was isolated as its nitrophenylhydrazone
and reduced as such with Raney nickel catalyst to the 2-amino-2-deoxy-
pentitols. The amino-alcohols were converted into the deoxy pentose by
treatment with nitrous acid. 2-Deoxy-D-ribose was isolated as its benzyl-
phenylhydrazone.

2. Identification

In addition to the usual methods of carbohydrate identification, and
the preparation of suitable derivatives (anilide, benzylphenylhydrazone,
etc.), several color tests are available to test for this deoxypentose. Methods
of identification of this sugar have recently been reviewed thoroughly by
the authors.'*

Furthermore, many of the color tests used to identify deoxyribonucleic
acid depend on the sugar component of the inicleic acid. These color tests
are described in Chapter 9. With the Dische-^" diphenylamine reagent, 2-
deoxyribose gives an intense blue coloration. The test is not specific for
2-deoxyribose, but is given by 2-deoxypentoses generally^^^ ■ ^^^ and depends
upon conversion of the 2-deoxysugar under acidic conditions into co-hy-
droxylevulinaldehyde, which reacts with diphenylamine to give a blue-
colored dyestuff.

3-Deoxy- and 2,3-dideoxyribose (but not 4-deoxyribose) also give faint
blue colors with the diphenylamine reagent, but in these cases it is neces-
sary to heat for a longer time than is required for 2-deoxyribose. The trypto-
phane reaction, introduced by Cohen-^^ for the detection and estimation of
deoxyribonucleic acids, is given equally well by 2-deoxyribose. In the rec-

2«8 W. G. Overend and M. Stacey, J. Sci. Food Agr. 1, 168 (1950).

289 Y. Matsushima and Y. Imanaga, Nature 171, 475 (1953); Bull. Chem. Soc. (Japan)

26, 506 (1953).
"oz. Dische, Mikrochemie 8, 4 (1930).
29' R. E. Deriaz, M. Stacey, Ethel G. Teece, and L. F. Wiggins, Nature 157, 740 (1946) ;

J. Chem. Soc. 1949, 1222.
292 W. G. Overend, F. Shafizadeh, and M. Stacey, J. Chem. Soc. 1950, 1027.
2" S. S. Cohen, J. Biol. Chem. 156, 691 (1944).

54 W. G. OVEREND AND M. STAGEY

ommended procedure, the material is heated at 100° for 10 minutes with
tryptophane and 30% (final concentration) perchloric acid, and in a
positive test there is a rapid development of a red color. Quantitative es-
timations can be carried out by measuring the intensity of the color de-
veloped in a photoelectric colorimeter with filters having a transmission
range of 485-550 m/x.

The intensities of the colors produced by normal hexoses and pentoses
and their 2-deoxy-analogues w^ith Schiff's reagent^^^ have been measured
quantitatively^^^ with strict control of temperature and air contamination.
It was found that 2-deoxyribose gave a much more intense color than
ribose under comparable conditions. Gurin and Hood^^® have demon-
strated that addition of a solution of carbazole to an ice-cold mixture of
sulfuric acid and 2-deoxyribose results in the formation of an intense yellow
color. This test, however, is not as suitable as the diphenylamine reaction
for the detection and estimation of 2-deoxyribose and is markedly un-
specific. The reaction of cysteine and sulfuric acid-^^ with 2-deoxyribose
has also been used for estimation purposes. [Cf. Dische, Chapter 9.] Only
aldehydes with an a-methylene group (i.e., R — CH2CHO) condense with
3 , 5-diaminobenzoic acid to form quinaldines.^^* Consequently, in the
carbohydrate series only 2-deoxysugars will react and hence they can be
differentiated from their normal parent sugars. Attempts have been made to
estimate 2-deoxyribose by the orcinol reaction for pentoses, but in the test
the deoxysugar is mainly converted into levulinic acid which gives no
color wdth the reagent. The Dreywood^^^ anthrone reagent, w^hich gives a
positive qualitative test for a large variety of carbohydrates, gives a nega-
tive test for 2-deoxyribose.^''^

Recently, color tests have been found by which deoxysugars may be
distinguished from other sugars or sugar derivatives on paper chromato-
grams.'''* Partridge'" has reported Rp values (corrected to 20°) of 2-deoxy-
D-ribose on Whatman No. 1 filter paper in various solvents. These were
as follows:

Solvent Phenol containing s-Collidine Isobutyric acid

NH3(l%wt./vol.):HCX
/2f value 0.73 0.60 0.32

23" H. Schiff, Ann. 140, 102 (1866).

298 W. G. Overend, J. Chem. Soc. 1950, 2769.

296 S. Gurin and Dorothy B. Hood, J. Biol. Chem. 131, 211 (1939); 139, 775 (1941).

2" Z. Dische, Proc. Soc. Exptl. Biol. Med. 55, 217 (1944).

298 L. Vellu, M. Pesez, and G. Amiard, Bull. soc. chim. France 15, 680 (1948).

299 R. Dreywood, Ind. Eng. Chem., A7ial. Ed. 18, 499 (1946).

SO" L. Sattler and F. W. Zerban, /. Am. Chem. Soc. 72, 3814 (1950).

">• J. T. Edward and Deirdre M. Waldron, /. Chem. Soc. 1952, 3631.

chemistry of ribose and deoxyribose 55

3. Physical Properties

Deriaz et alP"^ investigated the mutarotation in aqueous solution of 2-
deoxyribose. The velocity constants for the mutarotations of both the
D- and L-forms of this sugar were calculated, and the values obtained were
2 to 3 times greater than those for L-arabinose. In the presence of 0.01 A'^
hydrochloric acid or 0.01 A^ sodium hydroxide, 2-deoxyribose reached the
equilibrium value for the specific rotation immediately on dissolution.
Values of the physical constants of 2-deoxyribose and its derivatives are
listed in the Appendix (Table VIII.)

4. Properties and Reactions of Derivatives
a. 0 -Glycosides

Early in the development of deoxysugar chemistry it was noted that the rates of
formation and hydrolysis of deoxysugar glj^cosides greatly exceeded those of the
corresponding normal pentose or hexose. When 2-deoxj'-L-ribose was treated with 1%
methanolic hydrogen chloride it afforded a methyl 2-deoxyriboside mixture which
was separated into crystalline a- and /3-isomers (A and B, respectively).^*' If 0.1%
methanolic hydrogen chloride was used then a third methyl 2-deo.\ypentoside (C)
was obtained. The glycoside C was much more rapidly hydrolyzed by acids than
were the glycosides A and B,^*! thereby indicating that C was probably a glycofur-
anoside. The glycosides A, B and C were separately mechanically shaken with acetone
and anhydrous copper sulfate. Glycosides A and B readily formed mono-0-isopro-
pylidene derivatives whereas C was recovered unchanged. ^^^ Since it is usual for
acetone to condense with adjacent cis-hydroxyl groups, this would imply that A and
B had pyranose structures and C a furanose structure.

When the glycosides B and C were treated with p-toluenesulfonyl chloride in dry
pyridine, they both yielded a di-p-toluenesulfonyl derivative. That from the gly-
coside C readil}' underwent exchange with one mole of sodium iodide when heated
at 105-110° for 3 hours with excess sodium iodide in acetone, whereas the derivative
from B was unaffected by this treatment, indicating that only in glycoside C was
there a primary hydroxyl group. From this it follows that B had a pyranose and C a
furanose structure. Results of oxidation of the glycosides A, B and C with lead tetra-
acetate confirmed these conclusions.-*' Final proof of the structures of the glycosides
B and C was obtained by methylation, hydrolysis, oxidation and comparison of the
rates of hydrolysis of the lactones so obtained. From the glycoside B a 1,5-lactone
was obtained'*' whereas a 1,4-lactone was derived from C, showing that the glycoside
B had a 1,5-pyranose lactol ring and C a 1,4-furanose lactol ring. Since A and B were
a- and /3-anomers, it followed that A also had a pyranose structure.

Independent proof of the structure of B was furnished by the fact that it could
be obtained from methyl 2,3-anhydro-/3-D-ribopyranoside by a series of transforma-
tions which did not affect the lactol ring structure or configuration at carbon atom 1
of the initial material or intermediates in the conversion. 2^' Further, this synthesis
served to establish the a,)3-relationship in the pyranose series, since the product was
the optical enantiomorph of the glycoside B designated as methyl 2-deoxy-/3-L-ribo-
pyranoside. Davoll and Lythgoe'"^ have pointed out that configurational assignments

'"^ J. Davoll and B. Lythgoe, /. Chem. Soc. 1949, 2526.

56 W. G. OVEREND AND M. STAGEY

for methyl deoxyglycosides cannot be based only on optical rotation measurements
and application of Hudson's^"' system that the more dextrorotatory anomer of a
D-compound is termed a-D-, the other anomer being the /3-d form, since it is not cer-
tain that the contribution of the asymmetric center at carbon atom 1 will be of the
same sign in methyl 2-deoxyglycosides as in the related methyl glycosides. The
a,/3-relationship between methyl 2-deoxyglycosides was studied by Stacey and his
colleagues.^*' They showed that 1% methanolic hydrogen chloride separately con-
verts the a- and /3-isomers of methyl 2-deoxy-L-ribopyranoside into the a,|3-mixture
from which they were initially isolated.

An outstanding property of the 0-glycosides of 2-deoxysugars is the lability of
the glycosidic substituent towards acid. Calculations from the experimental data of
the velocity constant (K) for the hydrolysis of the methyl glycosides of D-arabinose
and 2-deoxy-L-ribose (i.e., 2-deoxy-L-arabinose), with 0.01 N hydrochloric acid at
100°, gave the following results:'"''

(1) D-Arabinose:

Methyl glycoside a-pyranoside /3-pyranoside a,/3-furanoside

K

1 , ^0 — ?■„ , .
= - log (t in minutes)

t r — r

0.00064 0.00064 0.0068

(2) 2-Deoxy-h-ribose:

Methyl glycoside a-pyranoside /3-pyranoside a,|3-furanoside

K 0.18 0.22 1.4

It was noted by Stacey and co-workers^*' that, on distillation of methyl 2-deoxy-
Qr,/3-L-ribofuranoside, polymeric material was formed whenever cautious superheat-
ing occurred.

h. N -Glycosides

A'^-Phenyl-2-deoxy-D-^ and -L-ribosylamine^" have been obtained in
crystalline form by the usual methods of synthesis, and are useful for the
isolation and characterization of this deoxypentose. It is thought that these
derivatives have a pyranose (l,5)-lactol ring structure; the stereochemical
configuration of the sugar-base linkage is unknown. A^-Glycosides of 2-
deoxysugars are unable to undergo the Amadori rearrangement, and it is
considered likely that the specificity of the reaction of deoxyribose with
secondary amines to yield colored products is a function of the inability of
2-deoxysugar derivatives to participate in this rearrangement.

Although deoxyribonucleosides have been isolated from natural sources,
attempts to synthesize them have been less successful. The reaction inves-
tigated was the coupling of an acetyl(or benzoyl)-l-bromo-2-deoxysugar
with the silver salt of a base. Attempts to prepare 3,4-di-0-benzoyl-2-
deoxy-D-ribopyranosyl bromide by treating syrupy 1,3, 4-tri-0-benzoyl-2-
deoxy-D-ribose with hydrogen bromide in acetic acid resulted in extensive
decomposition. Treatment of syrupy 1 ,3,4-tri-O-acetyl-D-ribose with
ethereal hydrogen chloride also resulted in decomposition, although in this

303 C. S. Hudson, J. Am. Chem. Soc. 31, 66 (1909).

304 T. Reichstein, Angew. Chem. 63, 412 (1951).

CHEMISTRY OF RIBOSE AND DEOXYRIBOSE 57

case condensation of the crude product with theophylUne silver gave a
very small quantity of 3',4'-di-0-acetyl-2'-deoxy-D-ribopyranosyltheo-
phylline (D).^''^

Diacetyl-D-arabinal was treated with hydrogen chloride in benzene and
theophylline silver added to the crude product. Two forms (3 and 23 %) of
3' ,4'-di-0-acetyl-2'-deoxy-D-ribopyranosyltheophylline were obtained.
[The isomer formed in the smaller amount, was identical with the product
(D)]. Deacetylation afforded two forms of 2'-deoxy-D-ribopyranosyltheo-
phylline, which were considered to be a- and /3-anomers. To avoid instabili-
ties inherent in the deoxysugar, Lythgoe and his colleagues-^- attempted
to defer its formation until the final stage of the synthesis by making 2' ,3'-
anhydro-5'-0-trityl-7-i9-D-ribofuranosyltheophylline and then converting
it to the deoxysugar derivative by the procedures available for the con-
version of sugar epoxide rings to deoxysugars, but a negligible amount of
the 2'-deoxyribose derivative was obtained. Some chemical properties of
the naturally occurring deoxyribonucleosides have been examined. Towards
acidic hydrolysis, cytosine deoxyriboside exhibited the greatest stability .^°^
As with the ribose nucleosides, it was found that purine deoxyribosides
rotate the plane of polarized light in a Jevo direction, whereas pyrimidine
nucleosides are dextroYota.tory }^^

Reaction between 5,6-dimethylbenzimidazole silver and 3,4-di-O-acetyl-
2-deoxy-D-ribopyranosyl chloride in xylene solution at 100° gave 5,6-
dimethylbenzimidazole- 1 - (3 ',4 '-di-O-acetyl-2 '-deoxy-D-ribopy ran oside) , hy-
drolysis of which furnished 5,6-dimethylbenzimidazole-l-(2'-deoxy-D-
ribopyranoside).'*^ In like manner, benzimidazole-l-(2'-deoxy-D-riboside)
was prepared.

As is the case with 0-glycosides, the A^-glycoside derivatives of 2-deoxy-
sugars are hydrolyzed by acid much more rapidly than the normal pentose
or hexose analogues. This has been demonstrated for .V-phenyl-D-ribosyl-
amine and -2-deoxy-D-ribosylamine.'^° Cohn and his colleagues"" state
that adenine-9-(2'-deoxy-D-ribofuranoside-5'-phosphate), when in 0.01 A^
hydrochloric acid at room temperature, liberated adenine at a rate of 2 %
per hour, whereas adenine-9-(i3-D-ribofuranoside-5'-phosphate) was unaf-
fected by this treatment. 5,6-Dimethylbenzimidazole glycosides are much
more stable than arylamine-A^-glycosides, but again the 2-deoxysugar de-
rivatives are more labile than the normal sugar derivatives. For example,
5,6-dimethylbenzimidazole-l-(2'-deoxy-D-ribopyranoside) is hydrolyzed by
heating in a sealed tube with 6 A^ hydrochloric acid for 12 hours at 100*^,
whereas the corresponding 5,6-dimethylbenzimidazole-l-(D-ribopyrano-
side) luidergoes no hydrolysis.' ^^

'"5 F. Bielschowsky and Marianne Siefken-Angermann, Z. physiol. Chem. 207, 210
(1932).

58 W. G. OVEREND AND M. STACEY

c. Phosphates

Manson and Lampen'"® reported that they obtained phosphorolysis and
arsenolysis of hypoxanthine deoxyriboside by enzyme preparations from
calf thymus gland and rat liver. An acid-stable phosphate was isolated as a
product of phosphorolysis. It was concluded that this ester was 2-deoxy-D-
ribose-5-phosphate and that it was formed from 2-deoxy-D-ribose-l -phos-
phate by mutase action. [Cf. Glock, Chapter 22, and Schlenk, Chapter 24.]
The same authors^ "^ obtained indications for the formation of 2-deoxy-D-
ribose-l -phosphate during the phosphorolysis of thymidine. Reverse reac-
tions have been demonstrated since it was shown that hypoxanthine deoxy-
D-riboside was formed enzymically from deoxy-D-ribose-1 -phosphate and
hypoxanthine.^^^ Similarly, guanine deoxyriboside could be formed by using
guanine in place of hypoxanthine.^"^ Manson and Lampen^^" also observed
nucleoside synthesis with the enzyme found in cell-free extracts of Es-
cherichia coli. Determination of the equilibrium constant indicates that the
synthesis rather than the splitting of hypoxanthine deoxy-D-riboside is
favored.^"*

By enzymic phosphorolysis of guanine deoxy-D-riboside Friedkin^"^ was
able to isolate the labile phosphate in the form of a crystalline cyclohexyl-
amine salt. This was a salt of the sugar phosphate designated above as 2-
deoxy-D-ribose-1 -phosphate and used in the experiments on nucleoside
synthesis. Several reasons are forwarded to support the structure assigned
to this compound. The ester, which contains one mole of 2-deoxy-D-ribose
per mole of phosphorus, is extremely acid-labile — 50 % of the phosphorus
present is released as inorganic phosphate within 10-15 minutes upon
hydrolysis at pH 4 at 23°. This lability of the phosphate ester linkage points
to carbon atom 1 as the site of esterification, and indeed a free aldehyde
group is also released upon hydrolysis at pH 4.

The phosphate resulting from the mutase action on 2-deoxy-D-ribose-l-
phosphate was isolated as the barium salt. Hydolysis with A^ hydrochloric
acid for 7 minutes liberated 45 % of the organic-bound phosphate. Since
the sugar phosphate was reducing and gave no formaldehyde on periodate
oxidation, it was, in view of its formation from hypoxanthine deoxy-D-
riboside, assumed to be 2-deoxy-D-ribose-5-phosphate.

Arsenolysis of hypoxanthine deoxy-D-riboside resulted in the formation
of hypoxanthine and free 2-deoxy-D-ribose.^°^ Probably the primary product

3°^ L. A. Manson and J. O. Lampen, Abstracts Papers 114th Meeting Am. Chem. Soc

53C (1948); /. Biol. Chem. 191, 95 (1951).
307 L. A. Manson and J. O. Lampen, Federation Proc. 8, 224 (1949).
SOS M. Friedkin, /. Biol. Chem. 184, 449 (1950).

309 M. Friedkin and H. M. Kalckar, /. Biol. Chem. 184, 437 (1950).
3'o L. A. Manson and J. 0. Lampen, Federation Proc. 9, 397 (1950).

CHEMISTRY OF RIBOSE AND DEOXYRIBOSE 59

is 2-deoxy-D-ribose-l-arsenate, which decomposes in aqueous media to
2-deoxy-D-ribose and arsenate ions.

Allerton et al.^^^ have succeeded in synthesizing some phosphoric acid
esters of 2-deoxy-L-ribose. Phosphorylation of methyl 2-deoxy-/3-L-ribo-
pyranoside with diphenylphosphorochloridate afforded the crystalline
3,4-62sdiphenyl phosphate. Hydrogenation of this compound in the
presence of Adam's catalyst afforded the syrupy 3,4-diphosphoric acid
derivative which was characterized by the formation of its acridine and
cyclohexylamine salts. Very mild acidic hydrolysis resulted in cleavage of
the glycosidic methyl group without simultaneous scission of the phosphate
residues and in this way 2-deoxy-L-ribose-3,4-diphosphate was obtained.
By a similar series of reactions 2-deoxy-L-ribose-3 , 5-diphosphate was pre-
pared from methyl 2-deoxy-a!,|3-L-ribofuranoside. Furthermore 2-deoxy-
L-ribose-3-phosphate and 2-deoxy-L-ribose-5-phosphate were prepared and
isolated as salts.

2-Deoxy-D-ribose-5-phosphate may be involved in the biosynthesis of
this deoxypentose. [Cf. Glock, Chapter 22.] The suggestion was made by
Hough and Jones^'^ that deoxypentoses may arise from the aldol-type
condensation of acetaldehyde and glyceraldehyde. Racker^'^ has deduced
evidence which indicates a similar route for the enzymic synthesis of deoxy-
pentose from triose phosphate and acetaldehyde. Extracts of E. coli, C.
diphtheriae and S. fecalis prepared by sonic disintegration of the bacterial
cells, or by grinding with alumina, are capable of catalyzing the reversible
reaction :

glyceraldehyde phosphate + acetaldehyde ^ deoxypentose phosphate

By combining the enzyme (phosphodeoxyriboaldolase) bringing about
this change with purified phosphoriboaldolase from yeast, Racker^'^'' was
able to demonstrate the long-sought conversion of D-ribose into deoxy-D-
ribose; e.g.

D-ribose-5-phosphate > deoxy-D-ribose-5-phosphate

It is apparent that a triose phosphate is the common intermediate between
D-ribose and 2-deoxy-D-ribose in metabolism.
d. Other Derivatives and Reactions

The action of acid on 2-deoxyribose results in the formation of co-hy-
droxylevulinaldehyde and then levulinic acid.^^^ Oxidation with bromine
water follows the normal course and affords 2-deoxyribonic acid."-"®
Similarly, oxidation can be achieved with barium hypoiodite in the presence

3" R. Allerton, W. G. Overend, and M. Stacey, Chemistry & Industry 1952, 952.

312 L. Hough and J. K. N. Jones, Nature 167, 180 (1951).

3'3 E. Racker, (a) Nature 167, 408 (1951); (b) J. Biol. Chem. 196, 347 (1952).

60 W. G. OVEREND AND M. STAGEY

of barium hydroxide.^^^ The acid can be converted to a lactone: in solution
2-deoxy-L-ribonolactone undergoes no change in optical rotation during
8 days."^ Likewise 3 , 5-di-0-methyl-2-deoxy-L-ribonolactone underwent
neghgible hydrolysis, but 3,4-di-0-methyl-2-deoxy-L-ribonolactone was
completely hydrolyzed in 96 hours.^^^ When di-O-acetyl-D-arabinal is treated
with chlorine in chloroform solution, it yields 1 ,2-dichloro-l ,2-dideoxy-D-
pentose 3 , 4-di-O-acetate, which on reacting with silver carbonate in ether
is converted to 2-chloro-2-deoxypentose 3, 4-di-O-acetate. Heating with
lead oxide in a mixture of chloroform and water results in simultaneous re-
arrangement and deacetylation of the latter compound and affords 2-de-
oxy-D-ribonolactone .2'^ ^

Esters (e.g., 1,3,4-tri-O-acetate and 1,3,4-tri-O-benzoate) of 2-deoxy-
D-ribose have been obtained crystalline.^^^ The deoxypentose forms a ben-
2yi.278,279 ^j^fj p-nitrophenyl-hydrazone,"^ but does not form osazones. It
affords mercaptal derivatives,"^ and methyP*^ and trityP^^ ethers of the
sugar are also known.

V. Addendum

It has been demonstrated recently that D-ribose occurs in natural ma-
terials additional to those already described. A study of the immunologi-
cally active type-specific substance of Hemophilus influenzae, type b, has
given indications that this material consists of a polyribophosphate chain
as it exists in pentose nucleic acids, in which the place of the purines and
pyrimidines is occupied by a second similar chain, linked to the first in
1 , l'"-glycosidic linkages.*^*

Several nucleosides and nucleotides have been isolated. Nebularine,
a natural product which was isolated*^^ from the mushroom, Agaricus
(Clitocybe) nehularis Batsch., and which gave purine and ribose on hy-
drolysis'^® has been shown to be 9-/3-D-ribofuranosylpurine.'" Marrian'^^
claims to have a new adenine nucleotide which is adenosine-5'-tetraphos-
phate, and the isolation from yeast of a nucleotide containing both ribose
and mannose is reported. '^^ The properties of this latter nucleotide are
consistent with those of a structure in which the terminal phosphate group
of guanosine-5'-pyrophosphate is joined to a mannosyl residue, (i.e., it is
guanosine diphosphate mannose).

^^* S. Zamenhof, Grace Leidy, Patricia L. Fitz Gerald, Hattie E. Alexander, and E.

Chargaff, Federation Proc. 11, 315 (1952); /. Biol. Chem. 203, 695 (1953).
'1* L. Ehrenberg, H. Hedstrom, N. Lofgren, and B. Takman, Svensk. Kem. Tidskr

58, 269 (1946).
3'« N. Lofgren, and B. Liining, Acta Chem. Scand. 7, 225 (1953).
"7 G. B. Brown and Virginia S. Weliky, /. Biol. Chem. 204, 1019 (1953).
"8 D. H. Marrian, Biochim. et Biophys. Acta 13, 278 (1954).
319 E. Cabib and L. F. Leloir, /. Biol. Chem. 206, 779 (1954).

CHEMISTRY OF KIBOSE AND DEOXYKIBOSE 61

Wyatt and Coheu^^'' have demonstrated that deoxyribonucleic acids of
bacteriophages T2, T4, and T6 of Escherichia colt contain no cytosine:
instead the pyrimidine base, 5-hydroxymethylcytosine, is a component of
these nucleic acids. Subsequently, Weed and Courtenay^-' succeeded in
isolating a deoxyribonucleotide containing hydroxymethylcytosine from
the deoxyribonucleic acid of Escherichia coli bacteriophage.

D-[l-C^'']-Ribose has been prepared.^- 2-Deoxy-D-ribose was obtained by
partial oxidation of 3-deoxy-D-glucose with sodium metaperiodate, fol-
lowed by deformylation of the product by rendering the reaction mixture
slightly alkaline. ^-^ The product was isolated as A^-phenyl-2-doxy-D-ribosyl-
amine.

The synthesis of A^-(3,4-dimethylphenyl)-D-ribosylamine by a rearrange-
ment method has been reported.'^* 9-,S-D-Ribofuranosylpurine ("nebular-
ine") has been synthesized^" by condensation of the chloromercuri-deriva-
tive of purine and tri-O-acetyl-o-ribofuranosyl chloride according to the
general method of Davoll and Lowy.^" Alternatively, 6-chloro-9-i3-D-ribo-
furanosylpurine was prepared and subjected to reductive dehalogenation to
afford nebularine. Further syntheses of 5,G-dimethylbenzimidazole-a-D-
ribofuranoside (a-ribazole) have been reported. 2-Nitro-4,5-dimethyl-A''-
(2',3'-di-0-acetyl-5'-0-trityl-D-ribofuranosido)aniline was reduced to the
(iorresponding 2-amino derivative, and this syrupy product in benzene
solution was stirred with carbon disulfide and a base (i.e., barium hydrox-
ide) to afford 5,6-dimethyl-2-sulfhydrylbenzimidazole-2',3'-di-0-acetyl-
5'-0-trityl-D-ribofuranoside. Desulfurization with nickel of this latter
compound and removal of the acetyl and trityl residues gave 5 , 6-dimethyl-
benzimidazole-a-D-ribofuranoside (isolated as the picrate) but none of the
j8-isomer.'2® Condensation of 5,6-dimethylbenzimidazole in dry dioxane
at 100° with tri-0-acetyl-D-ribofuranosyl chloride followed by deacetylation
of the reaction product afforded 5,6-dimethylbenzimidazole-Q;- and -/3-ribo-
furanoside.^^^ The yields of products were 2% and 10%, respectively, based
on the amount of tetra-0-acetyl-D-ribofuranose used for prepg^ration of
the acetochloro-ribose. a-Ribazole phosphate has been synthesized by
phosphorylation of the 5'-0-trityl derivative of a-ribazole with either

320 G. R. Wyatt and S. S. Cohen, Nahire 170, 1072 (1952); Ann. inst. Pasteur 84, 143
(1953) ; Biochem. J. 55, 774 (1953).

321 L. L. Weed and T. A. Courtenay, J. Biol. Chem. 206, 735 (1954).

522 Harriet L. Frush and H. S. Isbell, J. Research Natl. Bur. Standards 51, 307 (1953).

323 P. A. J. Gorin and J. K. N. Jones, Nature 172, 1051 (1953).

324 V. A. Konkova, Zhur. Obshchel Khim. 22, 1896 (1952).

325 J. Davoll and B. A. Lowy, J. Am. Chem. Soc. 73, 1650 (1951).

32« Dorothea Heyl, Edith C. Chase, C. H. Shunk, Marjorie U. Moore, Gladys A.

Emerson, and K.'Folkers, J. Am. Chem. Soc. 76, 1355 (1954).
327 A. W. Johnson, G. W. Miller, J. A. Mills, and A. R. Todd, J. Chem.. Soc, 1963, 3061.

62 W. G. OVEREND AND M. STACEY

diphenyl- or dibenzyl-phosphorochloridate and subsequent removal of pro-
tecting groups. This phosphate has approximately the same biological
activity as a-ribazole.^-^ Folkers and his colleagues^^® have prepared 5,6-
dimethylbenzimidazole-2'-deoxy-D-ribopyranoside (isolated as the picrate)
by condensation of 5 , 6-dimethylbenzimidazole silver and 3,4-di-O-acetyI-
2-deoxy-D-ribopyranosyl chloride.

The viewpoint that cytidylic acid "b" is cytidine-3 '-phosphate, based
on indirect evidence^-*' ^^° (i.e., solubility, ultraviolet absorption, and acid
strength measurements), has been reinforced by the results of a comparison
of the infrared spectra of the phosphates of cytidine and deoxycytidine.^^^
The spectra of the two 5'-phosphates are closely similar, as also are those
of deoxycytidine-3 '-phosphate (synthesized chemically^^O and one form of
cytidylic acid "b." The infrared spectrum of cytidylic acid "a" (in either
of its isomorphic modifications^^-) is quite different from any of these.
The conclusion that cytidylic acid "b" is cytidine-3'-phosphate is further
supported by a comparison of optical rotation and ultraviolet absorption
data in the cytidine and deoxy cytidine phosphates series.'*^ It follows that
uridylic acid "b" is uridine-3'-phosphate since it can be prepared from
cytidylic acid "b" by deamination under conditions which preclude phos-
phoryl migration. ^'^ In connection with work on the identification of
adenylic acids "a" and "b" as the 2'- and 3'-phosphoadenosines, respec-
tively, [see Baddiley, Chapter 4], Cohn and his colleagues^^'* have prepared
and characterized pure ribose-2- and -3-phosphates and have compared their
properties with the compounds described by Levene and Harris.''^^- "^^' ^^*
By hydrolysis at 100° for 4 minutes of adenylic acids with a poly sty rene-
sulfonic acid cation-exchange resin a mixture of ribose-2- and -3-phosphate
was obtained.^'^ This mixture, which consisted of 36 parts of the 2-iso-
mer and 64 parts of the 3-isomer, was separated by ion-exchange chro-
matography with borate complexing, and by fractional crystallization
of the brucine salts. Although separation by the ion-exchange method gave
pure ribose-2- and -3-phosphate, the crystallization method afforded only
pure dibrucine ribose-3-phosphate hexahydrate and the pure corresponding
salt of ribose-2-phosphate could not be obtained. The pure ribose phos-
phates were characterized by their optical and ion-exchange behaviors, in

^28 E. A. Kaczka, Dorothea Heyl, W. H. Jones, and K. Folkers, /. Am. Cheni. Soc.

74, 5549 (1952).
3" H. S. Loring, Myrtle L. Hammell, L. W. Levy, and H. W. Bortner, /. Biol. Chem.

196, 821 (1952).
"0 L. F. Cavalieri, /. Am. Chem. Soc. 74, 5804 (1952).
"1 A. M. Michelson and A. R. Todd, /. Che7n. Soc. 1954, 34.
"2 R. J. C. Harris, S. F. D. Orr, E. M. F. Roe, and J. F. Thomas, /. Chem. Soc. 1953,

489.

333 D. M. Brown, C. A. Dekker, and A. R. Todd, /. Chem. Soc. 1952, 2715.

334 J. X. Khym, D. G. Doherty, and W. E. Cohn, private communication.

CHEMISTRY OF RIBOSE AND DEOXYRIBOSE 63

the presence and absence of borate, by the differences in susceptibihty of
their methyl pyranosides to periodate oxidation, and by the differences
in their decomposition rates in alkah. The decreasing order of stabiUty to
alkaU (0.01 N XaOH at 22°) is ribose-2-phosphate— which is scarcely at-
tacked— ribose-3-phosphate, and ribose-5-phosphate- — which is the most
labile. Methods were described for the interconversion of the various salts
of these phosphates. Comparison of the properties of the salts of the pure
ribose-2- and -3-phosphates prepared by Cohn et aU^^ with those described
by Levene and Harris'" • -'^ • ''^ for preparations of ribose phosphates in-
dicates that the latter workers were handling mixtures composed of about
20 parts of ribose-2 -phosphate and 80 parts of ribose-3 phosphate. In the
light of present knowledge it is clear that the experimental conditions em-
ployed by Levene and Harris could not have resulted in retention of iso-
meric integrity, but would be expected to result in the formation of mix-
tures. Consequently, the samples prepared by Cohn and his colleagues are
the first pure preparations of ribose-2- and -3-phosphate. When ribose-2- or
-3-phosphate is heated for 2 hours with Dowex 50 (H+) resin or for 45
minutes with 0.1 iV sulfuric acid it forms ribose-4-phosphate.^^* The
yields ranged from 8 to 10% of theory, and not enough material was
isolated to permit characterization of a solid salt of the compound.
Khym ct alP^ prepared ribose-5-phosphate by treating adenosine-5'-phos-
phate with Dowex 50 (H+) resin at 100° for 4 minutes. Thereafter the
solution was cooled and more than 96 % of the adenine and adenosine com-
pounds were removed by filtration. The ribose phosphate {ca. 95 % yield)
was isolated in crystalline form as the barium salt. A fraction contain-
ing 70-80% of D-ribose-5-phosphate is obtained when xylose and ad-
enosine triphosphate are incubated with pentose phosphate isomerase
(from extracts of Lactobacillus -peniosus) }^^ A new simplified procedure for
the isolation of deoxyribose-1 -phosphate has been developed: it involves
the phosphorolysis of thymidine in the presence of dicyclohexylammonium
hydrogen phosphate, followed by a fractionation step with n-butanol-
diethyl ether, which yields crystalline dicyclohexylammonium deoxyribose-
1-phosphate after a single filtration.^^^ Reaction between thymine and
deoxyribose-1 -phosphate in the presence of mammalian thymidine phos-
phorylase gives thymidine which can be isolated in crystalline form.^^^

The preparation and properties of 2,3,4-tri-0-benzoyl-i3-D-ribose have
been described,^" and new benzoyl derivatives of D-ribofuranose and
aldehydo-i>-v\hos,e have been outlined. ^^^ A preliminary X-ray study is

"6 J. O. Lampen, J. Biol. Chem. 204, 999 (1953).

336 M. Friedkin and DeWayne Roberts, J. Biol. Chem. 207, 257 (1954).

337 H. G. Fletcher, Jr., and R. J. Ness, /. .4m. Chem. Soc. 76, 760 (1954).

338 R. K. Ness, H. W. Diehl, and H. G. Fletcher, Jr., J. Am. Chem. Soc. 76, 763 (1954).

64 W. G. OVEREND AND M. STAGEY

reported^^^ of the solid state transformation in tetra-O-acetyl-D-ribofur-
anose (cf. Davoli et alP'' and Farrar^"*).

The hydrolysis of tri-0-benzoyl-,S-D-ribofuranosyl bromide gives not
only 2,3,5-tri-0-benzoyl-/3-D-ribose as previously reported but also a
substance which is most probably 3,5-di-O-benzoyl-l ,2-0-(l-hydroxy-
benzylidine)-a-D-ribose.^''° The compound designated by Weygand and
Wirth^^^ as 2,3,5-tri-O-benzoyl-D-ribose has been shown to be identical
with 3 ,5-di-O-benzoyl-l ,2-0-(l-hydroxybenzylidine)-a-DTribose.

'" A. L. Patterson and Barbara P. Groshens, Nature 173, 398 (1954).

34" R. K. Ness and H. G. Fletcher, Jr., J. Am. Chem. Soc. 76, 1663 (1954).

3"' F. Weygand and F. Wirth, Chem. Ber., 85, 1000 (1952).

3« M. Viscontirii, R. Hochreuter, and P. Karrer, Helv. Chim. Acta 36, 1777 (1953).

3« W. A. van Ekenstein and J. J. Blanksma, Chem. Weekblad 4, 743 (1907).

3" K. Rehorst, Annalen 503, 143 (1933).

3« C. S. Hudson and S. Komatsu, J. Am. Chem. Soc. 41, 1141 (1919).

3« R. A. Weerman, Rec. Trav. Chem. 37, 16 (1917).

3" T. Brady, Biochem. J. 35, 855 (1941).

3« A. M. Gakhokidze, Zhur. Obschei Khim. 10, 497 (1940).

CHEMISTRY OF RIBOSE AND DEOXYRIBOSE

65

^ i^

•o

H

►^

fe

0

0

(73

Oi

0

fQ

a

<

tJ

<

>

<;

H

H

0

<;

tC

>

(H

5

K

w

Dh

P

tf

o

73

C
3
O

a

S
o
O

00 55 _a 5£

^ CO '-- <N

"* ^^ O '"'

^ o o c-

(M --H rt

^ s

I I

+

o '(p i5

I -2^.

w

« -H

^^

o

05 CO
I I

o

+

CO o>

ci

O- 00

fe s l^

°o 00 I
^ bC 00 t^

o ^ 2~ 22

Iq M >2 'c

CS 00 03 ^

_r; 00

^

o

^ ^0)

W CO

+

O

H

^ a

CO O O

3 c 3 c
'rt .2" .?

^ »3 -t^

co"

flj

0

a;

ylidene-

ri-0-acet;

ri-0-acet^

0

0
rt

0

a;

0
0

0

0
0

0

'fcn

0

0

">>

0
0

(SJ

C

^
4

6

^
-^^

0

00

-5

0 CO -^
D, (M~ co"

-1^

•4*

CO

J2

CO

J2

r^

CO

co"

T-T

-T <n"

C-i

c<r

CO

d

+

0 e

^

SiH

Ic

5 >.

0

0 Cl.

c E

■5 ^ H

■Z o c

■Sq

£ II II
« (ZJ <<
■^ -a <u

>• •- +3

m S o

IP 3 ■"

3 -c II

03 -Q

H o.

66

W. G. OVEREND AND M. STACEY

P^

O

CJ

^

t— 1

fN

,— ^

O

^^

o^

Xi

CO

fl)

o

+

-a,

,-H

.— 1

f-1

^-^

bC

,-H

^^

O

(•1

i-H

»~*

o

CT

*^^

u

1

Ci

3

O

T

>->

CO

^ Zl '>* ^ Q

Tt< ;i^ CO ^ ^

-C ^

CO -"^

7 ^

o o

1

s~

2

o o

-o

o

o

O

^ -H

. »v

(M

<N

• •» • -^

' — ^

■^

O

Si

§,

1

7

J

o

■^

s~

^-^

g

Tt; 0

o

S~

S"

1

S'

qS

l>^

CI

o

O

w

o

2

o

7 12^
^ 1 3

o

00

CO •-
+ <

+
T

1

00

7

lO

p

T

o
•^

t^

^ — N. ■*— ^

, s

CO

, — ^

00

> — ^

00

'5~

+

oi

^

+

o3

•1

7

c3

+

,-^

^^

":i

0)

3

p

i

Q

o^

1— I

r-l CO

CO

o ^

•-^ a

o

_>

_>

>

,

6

t-c

s

■5

^

o

s

3

3

J,

J,

»c

o

'a

T3

03

O

>>

'>. '>i

o

C

>,

o

c

0)

c

OJ

a

01

43
C
43

'>,

43 ,

s
o

c3

c3

c«

c3

03

'V,

0)

o o

S3 S]

-D

^

[H

-a

"H

CO

O

3 §

O

9

<o

O

6

03

o

O

&,

a a

O

6

">,

Ti

">,

+J

T>^

S
o
O

c4
03

O

*J,H

(1)

o3

CO
O
-Q

'u

4

Si

o

6 6

>>

a;

o

o

a

So

+J

-1-i

-tj

■*^

+J

<i>

N

•^ ,^

-k^

N

-*j

o

o

73

o

o

.2 o

Tt*

'^^

Q

lO

lO

a

lO

S

c

c

;-i u
-hJ -tJ

^

c

ic

CO

_in

CO

co"

CO

8

CO

CO

8

^^

"e

C3

-^ IC

co"

O

^

co"

o

6

o

o

43

9"?

o

IM

(M

O

(N

(N

O

CO

o

O

N

CO CO

(M

o

<N

CO

CO

O

CO

lO

CO lo

C<l C>? r-l lO

CHEMISTRY OF RIBOSE AND DEOXYRIBOSE 67

■^ "^ ry-j »— ' "^ CO

(M 1-1

-I

-O —I

3

-D

__ o

00

'7^ o

00

^

""^

^-^

■^

-f

■*

"ti

'^

'^

;?S

CO

CO

CO

CO

CO

CO

c^

<M

oq

(M

(>)

ca

(M ^

^-N "—I

-a ^.

X! CO

-D

O -c.

1 CO cr

^ ^

^^ oi oj

o «

e?r

1 -^

o

1 «^

C3

^

•" ,— V "*

T ^^

1

^7+

f^ .S ^

">!

>i

0

0^ s +w

Ah

p

^

1 ii

^ ^ ^^ ^~^

2q

00

+

00

+

o

+ _

o ^ O*

t^ °^ +

K

C^

6

O

S"6

•O Ph ^

^ r-i

q

~~^ "^ h c

~~^

~~-^

"-^

^^

~~-'

O "^

lO .. °

"^ ^ ^

° —' ^ .5

0

~ — o

1 r^ '^

1 +S

ffi

°^ + o^ a

(M

(M

^_

ca

'-I

o ■^.

w ^

'^

CO —

1-H

ci

t~-l

■rV

Cl

O -^

,-^ ■^— ^ 1-^

-' S ,. s ^^ ^

CO ^- --^ >c

CO

(M

CO

to

(M

00 C5

^ 1

S 03

+ ^

+ +

+ +

+

+ 1

o ^-

+ -

la '

1 '^ I Q I

S o ^ ^ »o

CO —itTi^ (M 1— i(M,— 1'^'~'00

ij;^ V'^--— OOi; lOiO coco CO (N"*!!^^^ .-H

|CO rJ-1.3 "^1 (NOO 00<M O O^Of-Li-^ajOJ

3 ^ c

f^-"^^

>.

03

u

CO

CO

_-,

G

m

03

>i

o
S
o

CO

>i

o

O
o3

o

c
o

O

N

o3

0)

c
o

im"

1 -1.^

Tl

03

03
m
O

a

£ £

o

O

>>6

03 A

C

s

<u

9 o

J,

^

'n

'C

^

-3

>i

o

6

A

a
o

■5 -3

G

>1

9 9

-S -3

oT

1 ^
-r o

o

6^

.J. El

-a 3

c3

tS]

O

£
o

o

£
o

EC

o

>. ^— '

t-l

o

O

+ lO

fr3

CO CO

o

o

iC

»o

■^"^

CO

lO CO

o

CO

CO

c

-9

JD

to

C^" CO

im" (rf

,-^

•o

^''

r^

_r

T^

rt"

A

A

^

i.

a

68

W. G. OVEREND AND M. STAGEY

O

o

o

+

r

+

+ I + I I

^

2 ^ CO

I J2

w

S o *

00 -^ ^ -—- ^

00

CO 00 CO CO

--H 03 C» 05 ^

Tf I I I I

rt iC I^ (M M

C2 00 c:n CO

o 2 '^

^ CO CO t^ -*!
rt rt CO ■<*" CO
(M (M rt (M 1-1

I I I I +

^ 4^

^2 3

r

c

'>,

o

-C

-fj

Lh

0)

T3

■5

£

c

a>

C
O

-5 c

S C3

C
o

03

3

■— 1 u

tn

C

3

o

aT3

^

a

P,

00

d

O

a

ex

a

S'

^

1

^

X!

^

,,^

CO

a,

3

3

'5

.2"

-a
'3

3

s

02

m'

a;

6

a

OJ

."2

73

*C/3

T

3

O

o

'33
o

'oi

o

J,

O

c

^

a;
C

S

CO

,

T3
'm

«

">.

C
03

>.

cS

'

">>

02

^— ^

•^

O

c3
3

o

o

">>

o

T3

',

^'

-c'

3

c

>>

c

a,

QJ

G

">,

>i

>-J

I.

o

QJ

a,

a>

o

O

V

o

^

J3

3

£

^

o

X!

J2

o3

^

0)

"o

'C

o

JO

O

"C

o o

a

s

£

o

1

X!

a

"C

Q

to.

'C

Q

'u,

'V,

o

1

6 o

'u

'>.

'C

'4-

lO

>:*<

to.

•4*

o

-5

■5

^

03

o

">>

>-.

o

N

CO

>i

CO

JS

CO

CO

CO

^

■*

CO

O

X!

c

-C

-tJ

<>)

IM

0)

<M

<N

(M

<M

(>)

IM

"*

02

«

W

§

§

s s

XJ

2, s

X! XJ

•7 'T -C

o o "S

CHEMISTRY OF RIBOSE AND DEOXYRIBOSE

69

O CO

r-( ^ Tfl

2^.

W -I

+

O

+

00:
a:

+ I I

^

"5 m

+

+

c3

+

+

+ •5

^ 1=

^ 7

Ti CJl rt

i

3 e

(U

g CO 2 ^

0

"?

0)

in

1

a?

."2

">.

0

s

(1

6

0

.^

CO

;-i

T^

*f

0

5S

10

0

'*

o

Q

6

2

<v o
-9 c O

9- rt

^ CO ^

-Q CO

Pi (m"

o .^

=3 .j.

c3

-►J

>> +3

£-1
3

>o

0 "*„

'o

CO

•S w

Si

c<r

Pi (m"

S

«

-S5

M CO
O ^.

o

J. -a
c 2 'S "3

j2 ^

o3 o3

Pi

Pi

70

W. G. OVEREND AND M. STAGEY

rt

CO CO «D
lO lO CO
IM C^ -H

00 -o

w

+

a; -^ to

03 CO .-

>, o S-
t °o, ? e^

o

o
;o

■ CO

+ d + T -
T 'S -° +

°c^ °^ ^

-H ^

X3 CO

P-i (M ^^ 5S

>- ~^ _1_ ._

>l GO

Pi '-'

CO '"'

T ^

a u

+ S

5, .- +

CO -— ^ *

7 S '

-f- ^-- O

feS
+

h + +

+

+

o

o

d,

00 lO '-^
(M ^ (M

t^ O

GO CO

lO .-1 (M

»0 I I

rt CO 05

00 c^

t^

-^

I-H

3

^^

t^

e=r:

t^ 1— 1

iM

1 T3

o ■"-'

^ ^

/ V 1—1

03

,_,

IM

S"

cA

■^■^

s

O

■—1

■<*<

1

73

00

^ — '

CO

-*

(£l CO

=8 i^ .«

O

CO

1 c

o >> a;
>- Q. _o
73 o =:
>,xi >>

fog

--. N ^ J, J3
>. " - - ""^

(U O) <I>

c3 03
6 6

CO CO 'O

(M (N J3

^ N
— H 03
>>T3
to .«

(=1 .5

03 N

3 a;

1 oj

° s

.8

IC O

o -:
t:^ Ph -e

9 S

— CO

aS

=« .-r

o -£3
" D-i

o

J2

>> J-
S S c

O >. 03

CHEMISTRY OF RIBOSE AND DEOXYRIBOSE 71

^ CO cc CO '-^■'

3 ^ ^ ^ ^

lO rt ^

d

Tt<

+

+

P~i ^ — -3- '^ ^ crs ^~^

jD .T" >> ^ ^ ii. i*> I S"

>>

00

(M

d
+

+

T

T

•,• -' - ; UN _ ^ . + + ^ ?H &5 I ^°

o o _LO o fQ I 1^^* ■ -Tl ..^ooo

«5 O + (N '>! ^ I +n-^ 00 M O.t^fO-cO

CO ^ C^ (M <£! Pnt^ r- OJ o oW-^-

+ + 3 + + + S 3""+ + 7 + +3^7 + +

O (-<1 '"^ ^r "-^ ^ — ^ , I ^ -^ ^^ r-^ ^ ^

rol I I •-'^ICOICOcOt^l

;.; -4 CO

9 ^>^' o 'r 'r ' <u 'S <u Qi ' i ojoj ^o c

^^ u ^ Q^ t:*? 2j^ Oj^ o't' o^^a^^c.^ O'^ o-^ o.-r£ eocoOrS
o s a. s,^-'^-- o CO CO ?5. o c^ il Pi (>r<M''ioDH

72

W. G. OVEREND AND M. STACEY

CO _ ^ rt rt

CO lO
^ (M

CO

CO ,-1

03

2 2 ^ ^ «) -^

'^ '-' -H _^ C^ ,-,

CO CO i>) o lO

-^ rt -H 'T (M
rH ^ r-H (M

o

H

+

+ +

+

Ah

BS

j3

+ I

H

+

66§5

o § <^. ^_, O ir

00

f^ ^ ^ ' c^ o6

CO CO (M ^ — . ^
I I I rt

I I

+

C^ I

fe5

u

o

o oo "-i

ut

T

cb

Ol

O

,-H

t^

CO

°

1

d

-"^

05

-O

s

s

T3

■«•

3

o

4<

^ s

- - ^ «

^ 00 O Zh aj

I i 00 I J-

2 d :°. 2 G

^

00 -< j-j

S <='^

1 cs a _

73
O

a,

O

N

>>

•^

^ ,

>>

a

a

TO

, 1

>>

ii^i

.•^

X

c

C

■fcn

c3

■+J

cS

flj

cd

O

;h

0)

^ 9

o

s

0)

9

03

9

03

S
o

03

0) lO

lO

"?

•—1

OJ

"?

r— <

(— (

o

s

a;

iC

03

«— <

u

03

«3 >>

">>

Tj

o

s

'>>

a,

c3

'>,

S

">>

o3

o

03

■a

O "S

QJ

a

0)

o

Si

o

9

3

">>

QJ

6

C

6

J "

O

O

<u

0)

o

Oh

O

O

03

N

O

03

o

2 o3

^9

o3

6

o

J2

6

,

J3

9

,

H=

c3
s-<

Oi

s

^

9

c3

03

Si

9

,

2

e3

03
03

CO CO

o

tS]

CO

co^

c
O

CO

0)

O
CO

O
X5

CO

9

o

-3

03

o

CO

CO

O
Si

03
CO

O

CO

CO

O

03

CO

9

3

C3
03

'S
o

Xi

CO

(m" (n"

(m"

(m"

in

(m"

to

S

c^"

lO

(5 p?

cf

5 P?

iM

iM~

im"

U5

S tf

(n"

Q

a

a

O

O

a

CHEMISTRY OF RIBOSE AND DEOXYRIBOSE 73

a c

^ ^ ^ -^ ^^ ^ X!

lO ' -* C5 lO ^^ ^~^

lO O CO O I^ iC o

CO o

25 +

^""^^ -Q(MCft(M.-r: ^

B I ^^° T ^ ^^ ^ ^ + + ^ 7

,_^^^ ,— vIMtO-^^ ^ — - CO r^ ^ --^ ^l^ ,^ ^C^--r-^ lO-— -

_|_o3 oj_|__l_cS I o3_|__l_o3 o3 OT _L._j_c3

S Si

O CO rj O) O ir- ,' Oi • "^cj '^T'-~-_

i;SS2--^Si5 00 (Mt-_^ op -^ J.J^■^'2

I II 1^' I cot^oo^ 1^ r^io^^co

eo^(MioJ2_Q 00 I ^E:; gooo oi ^^

73 00 »0

.2* c3 c3

00

-t- -^J -t-^

4J

o
(S3

a

"S

o

6

o

6

o,

=icjS ' SS So30c3^o3-r:o30

>-. 'tiC ""Gr' ' ' "C "C "^ _J' EC i3 d m

s '^^ >, o 't, ^^ 5 o ^_ ^~ '>, ^„ o -^^ O ^_ i ^„ O .S ^ O

_g co" -G .-2 co" co" r2 .- co" co" jg' co" :2 co' :S i^" ">i <^" :^ ^ S ^

a c-f (5 (m" csT iS is csT csT "S (N f5 (m" (5 of ■^ (>r Pi J2 ' Pi

jG

-ti

0)

aj

G

;^

0

1

0

0

03

»— <

N

04

G

0

03

-5

0

c3

>-,

i"^

*j

-G

1— >

<V

0

>>

u

>.

-1^

03

03

C

<u

^.

t-

QJ

0

y.

-*-i

J3

03

0

<B

a

9

-^

0

a

-3

o3

CO

0

Tfl

6

CI

X!

co"

r-T

si.

a a

^ Q Q W

74

W. G. OVEREND AND M. STACEY

O Tfi

T-l CO

Tti o i^ r^

■rtH iC iC lO
CO <M C^l (M

O

^

s s <^

Tj< rt rt

CD _Q

IC

iC

0 CO

1 CO lO CO

"-I CO ^ Ti

1-H t^

^

a

c

^^.^

^

^

^

OJ

~— '

o

-u

^-^

s

CO

>.

CO

sp

c3

<>)

r^

O

<M

CO

w

Q

+ +

o

o

^

■*

o

CD

—

1

,^^

-C

^

O

^

^^

X

00

o

C3

00

Ol

O

CO
1

00
1

^

o

o <u

r*l +^ .-.

O ^„ O
£ f^ .-5

c3

-^ 2

u <M

o ><■

tf tf

c3 a>

o3 o

ti CO o

p:;

° i^ ^ s

o) CO

:^ o

>) N 1

?? "P o

S -Q

J3

<i^ -S r^ ^ •:= "^

CHEMISTRY OF RIBOSE AND DEOXYRIBOSE

75

rt

CO (M r-H

lO lO 00
IM (M C^

O

o-^ T 3

2~

^

T ^ ^ ^

o

Oh

"^^ Ol -—^

-^^ 1 ^

00

^

b« ^-

t^

CO

o 1 C

1

1

-40

;(b)

(W);
.5 mi

1

2^

s;

T i^H s

^

'>>

r~> ^"^ ^-^ t-l .

^ S"

Ph

- b °^, -2 te

: O

^^

§ t 1

-=: "^ 2 SoO ^

O t-

t:^ I

iC

id

CO

3

1—1 ^

CO

^ +

J2

■* ^ CO
I I +

s
s

CO

. "5

lO

CO

23

s§

op ^ S

I

'3

a*

WW

+

.w

^^ +

00
CD

CO

CO

CO

• fO

-a CO ^

CO

00

o

CO

02

CO

o g-.

-p a

3

o
O

CO

c

Qi

O

♦J

CO

>>

c

>>

p CO

■ -> >>
7 -^

."2

It
o ^

o

T3
CO

">.

O

CO

O

o

id

'y.

o

-a
cp

■>>

c

o
-a

CO

OS

a;
'a

"a:
O
C

3

to
O

a
o

-^

^

co" co"

co"

Oh

-a
<

id

76

W. G. OVEREND AND M. STAGEY

00

CO ^ ^ O CO
CO ?0 CO CO C^

IM SS

+ +

00

+

-C

O

°§g

CO

i5

^_^ o

^ CO
V — CO

■- I

o ^

o 1?

CO

IM

+

0H 1

o 1

c3 ,-.

^ ^

^ X

cs

:^^

^o O

4S "5 03

o ^

O cci Z

- — o

^-- CO

o "^

0^ +^

CO ^

CO (M

lo -— ^

1 1

+ ^

o

-o

o

00

J3

-t^

s

■?

♦^

^_^

^

■73

J2

c^

c

t^

~ —

, — ^

cS

00

^

cc

00

c«o

o

o,

CO

o

o

0

00

'-*

iM

^

(M

■*

s

S

(M

^^

^ 1

3"

c3

2 ^

o

-d — -

cc

CO

a

D,

-►i

05

0)

£

'? o-

^

^

o

d

(M O

00

00 ^

^ 00
00

d

>.

a

o

a;

j2

"o

'E

N

o

03

X

1

o

C

T3
j3

5 -fl

21

OJ >,

^' -o

03 Q

6 >^

aw G

(D OJ

6

o >■■

CO O cS ^'

6 G, CO

"i <w

o

c

li

a, fc<

o o
V a; j^

i^h^

bC

a hH

^ S «

a o o3

2 fe -5^

t- -3 _ — -
Q -^ 'r 4) o3

X X

O O

« a

9 •- o

X ^^-o

a o 5S

.-, c;

Sh 4) oj

9 .S 6

X p -o
9 -^ .'

>.

J3

CHEMISTRY OF RIBOSE AND DEOXYRIBOSE i i

^

lO h^ t^ r^ t^ t^ r^ O <M ?^ r^ h- GO 00 00 Xi oc X X r^

e

+ +

T T

+

^ ' ' ' ' -°o a: f5 +

-H

t^

O

AT

,-H

00

''-^

>,

C2
+

+

^

,-,

S

o

c

lO

o

W

"■— ^

. «

r^

'5

1

>>

CI.

+

o

OC

■*

00

:3

c3

1

-4-

^^ % e?,i°2Sg^^^ r g ^ T

«5 "^ ^ d r=

CO (M fC O O

+ T+ 'f c!; ' ^ I + + i I + + r +

■a

C/3

G

S c3 O .-

>>

(B M «

7 .^ b 2 c ■=

•^i .7 -^ o o n

T3

(N CO

7" -7* T" "O -t; aj •«
Q C Q p S 2 J

QS'"'5 ^ "^ oj-^ 0/0)

xL'^S^ 999 9 9

o

s

H

H

--

^

§

p

p

CO

o

jO

^

d

T

lO

lO

o

^

o

04

1^

o

t^

o

'^

o

-f

C5

•^

,-H

T— "

T

c

'5

o

X

ut

1

O

1
O

^^^

o

T

cr

1

CO

X

o

T

>— 1

^

CO

t^

§

g

X

C<l

CO
C5

e

a

r"

X

o

lO

ca

-^

"H

t3

£ :h

£

-c

£ :h

a

a
a

3

'3

'5

3

cr

M

£

3

M

6

5

6

5

0)

o

'E

~

s

^

o

^

.;i

ts

03

?^

a,

"5

,

,

»j

Jj

.

^

O

^

o3

3

>-^

^j

<3

132.

^.

^,

^^

%

c3

£-1

a;

0)

x'

0)

o

0;

4=

OJ

J^

g

O

o

o

q

X'

0)

s

£

£

D

0/

Y

0)

c

o

3

>^

O

O

"?

T3

O

c^

-a

6

'0

n'

5

J3

-S

■5

(M

-r

o

^

C<1

G

'^

X

■^

lO

^'

■» J='

■« ■<*<

-c"

S

»c

">>

OJ

-C

'*'

'^

■^

CO

CO

'qj

o;

CO

0)

CO

^

S

<5

W

78

W. G. OVEREND AND M. STAGEY

Pi

o

o

«? fe

1 °^

2 (^

o3 "3

R +

I

o 2

+

+

c^i

c- I ^°o

^2 CO i- ""^ CO

'"' t^ 4> o (N

I (^i ii «= r

cS CO o

^

I

+

^

'3

S3
-►^ fc- .

.2 c

o
d

.-2 S
cr ^

>>

^

>>

c

<v

Xi

a.

',

',

J,

>i

>.

>>

^

^

rH

-fj

■*^

a;

a;

<f

s

s

0)

a

6

6

"2

6

.^

^

"n

,^

'?

-o

rt

^

't3

■^

-<*<

TS

lO

CHEMISTRY OF RIBOSE AND DEOXYRIBOSE

79

«

t^oooot^oOGOoooo
cocooocococococo

t^ t^ 00 00

C5 CO CO CC
CC M 00 CO

'*' -* 'rf l^

J3

o

S3

T3

+

«o JS

>— 1 ^

CD

>, 1

2^:^:^+ ^p-.

,. ^

<-H • •-

,-— ^

C^ ^

• • .-— s

ES

^§

g£S

O 0 O 0 >V ^

^^0 o

(M O <N t>. 0

o o

lO o

o lO t^

- . . . o lO o

CI 00

00 •

■^ t^ (M t^ (M • C2

O O

d ^

• t^ o

■^ rt •n' CD (M 00 CD

CD -H

CD (M ^

+

^_^ o

^ O

7. lo

>> ^

^^^

^0^^'

o (M

CI ^

0 • o

+ +

+ + I + I +

11+ II

CD CO rt rt

+

o

iM t^ <N CO CI

ca CO (M CO

I I I I

I I I I

r-HO'-COr-lC^Ob.liiCOcO -..■ "^ I-

XCO(MCO.-H^0^5'-^O^C200

o ^
— ■ o

^

—

>i

o

o*

N

«

't.

^

'^

^-1

CJ

0)

^i

>~

73

J2

o

o

lO

N

N

6

o

fl

c

CO

OJ

o

rt

•^

cJs

J

o

iM

<u

">>

«

6

6

o

+J

-f

.^

+

O

lO

>o

i:

-if

^

CO

rf

CO

+

lO

tf

c^

6

(M

CO

co"

a

—1

'-^

^<

<M

c^

"3 -S (N

o ^ b 1^

0)

7 ^ o 'n o ■'J- O
■§ J -5 <i *f 'i -^

, S5 -tJ *f -iJ 52

O _§ 75 rt 75 2

^ r ^ Qj ^ r

C C CO

7 3

c :2

n' CO N CO

4^ .:i cj

? 33

3 ^

4J

pq n

-•. u ^ I, 5 u

(N -2 J2 ^ M -2

.2 o

j2 _3

3 .5 J2 c3

^ p -n o

80

W. G. OVEREND AND M. STAGEY

rt

CO CO

00 ■*
CO CO

CO CO

to (N
I I

+

+

CO CO CO

+ + +

X!

3

o

t^ 1-H r-H

a

03

73

a

a>

a,

3

^_^

05

a

o

^ — ^

•i^

00

•

a

o

3

O

CO

00

T3

1-H

1-H

r-H

1

1

»o

o

s

t^

O

(M

(N

(N

'"'

■"^

1

^5

O

N

i

73

a

O

P,

a

o
O

03

'h<

-o

3

1

T3

a,

c

to

o

a;

'ra

c

£

O

03

c

ti

*>»

03

3

■d

^

t«

3

O

<L>

vm

o

.D

s

O

"cS

5

O- <5i

rt

>.

X

o

73

>i — H

o

>>
X

o

a

3

tn

s

benzo

Furano

>>

, to

d- o

73
>>

1"H QQ

'S

o3
73

S* 73 ^

^
>>

01

o3
O

C

<v

6

03

3
'jh
o3

1

C>3

CO

5-tri
-3,5
-D-r:

"^^ o

'T e

- a

to
O

J=l

a

m
O

Xi

a

o

c

o3
>>
O

'tH

a;

CO

o

-acetyl -2, 3,
di-0-acetyl
i-0-benzoyl

73

a

o

u

If

9 §

• -H aj

>. A

O <D

2 c

O 0)

9^

• •H -^^
P «

0

■>>

a

o3

73
O

CI

Oh Q

a

_Q

O JN P

x>

P -5

c3

-Q

a

C^

CO

'S,

5

ri -H kC

»C

U5

o

03 t?

CHAPTER 3

Chemistry of Purines and Pyrimidines
AARON BENDICH

I. Introduction 81

1 . Historical 81

2. Nomenclature 82

3. Occurrence and Distribution 84

a. Purines and Pyrimidines of Nucleic Acid Origin 84

h. Purines and Pyrimidines in Other Biological Sources 03

c. Purine and Pj-^rimidine Derivatives and Analogues of Biological im-
portance 101

II. General Properties of Purines and Pyrimidines 107

1. Physical Properties 107

a. Solubility; Distribution Studies; Chromatography 107

b. Criteria of Purity and Identity; the Value of Ultraviolet Absorption
Spectra; Ionization; Tautomerism; the Value of the Isost)estic Point . 110

2. Chemical Properties 110
a. Stability Towards Acid and Alkali ; Transformations; Nitrous Acid . 1 17
1). Action of Reducing Agents; Polarographic Behavior 119

c. Action of Oxidizing Agents and Ultiaviolot Light 120

d. Color Tests 123

e. Salt Formation 124

III. Synthetic Methods 125

1. General Methods for Pyrimidines and for the Introduction of Isotopes . 125

a. Transformations; the Value of Halogen and Mercai)to Derivatives . . 126

b. Newer Methods 129

2. General Methods for Purines and for the Introduction of Isotopes . . . . 130

a. From Purines 130

b. From I'yrimidines 131

c. From Imidazoles 134

Addendum 135

General References 136

I. Introduction

1. Historical

Pyrimidines and purines are intimately associated with all the living
systems which have thus far been studied, and have been found, as Rivers
put it,^ even "... in the twilight zone between the living and the non-

^ T. M. Rivers, "Viral and Rickettsial Infections of Man," 2nd ed., p. 5. Lippincott,
Philadelphia, 1952.

81

82 AARON BENDICH

living," the viruses. These compounds, in the free form as well as in a
wide variety of chemical combinations, have caught the attention and
imagination of biologists, enzymologists, chemists, oncologists, physicists,
and many students of chemotherapy.

Recognition of this group of compounds dates back to the discovery in
1776 of uric acid by Scheele and by Bergmann and of alloxan in 1818 by
Brugnatelli.'^ It includes the pioneering studies on uric acid of Liebig and
Wohler^'* begun in 1834 and those of Baeyer^ in 1863 which led to the
cycHc formulation (1875) of uric acid and guanine by Medicus^ and the
pyrimidine nucleus (1884-5) by Pinner.^ '^ The classic studies on purines
(1882-1907) by Fischer have been collected in a single volume^ which also
contains a description of the early work. There are other excellent treatises
which deal with the history and development of these substances and their
occurrence as components of the nucleic acids. ^"'^^

2. Nomenclature
The term pyrimidine was coined by Pinner^ from a combination of the
words pyridine and amidine. He was the first* to point out the structural
similarity of pyrimidines to benzene, pyridine, and s-triazine, and, ac-
cordingly, depicted the ring system in the form of a regular hexagon .''•^•i^
Although the hexagonal representation of pyrimidines (I) has been adopted
by Chemical Ahstracts^^ and The Ring Index,^'' many authorities have pre-
ferred the rectangular form (11)9. 13,14,18,19 numbered as shown (II), or as

2 G. Brugnatelli, Ann. Cheni. Phijs. [2] 8, 201 (1818).

3 J. Liebig, Ann. 10, 47 (1834).

< F. Wohler and J. Liebig, Ann. 26, 241 (1838).

^ A. Baeyer, Ann. 130, 129; 131, 291 (1864); See also Ann. 127, 199 (1863).
6 L. Medicus, Ann. 175, 230 (1875).
^ A. Pinner, Ber. 17, 2519 (1884).
8 A. Pinner, Ber. 18, 759, 2845 (1885).

^ E. Fischer, "Untersuchungen in der Puringruppe." Springer, Berlin, 1907.
'" C. Brahm and J. Schmid, in "Biochemisches Handle.xikon" (Abderhalden, ed.),

Vol. 4, pp. 1014, 1131. Springer, Berlin, 1911.
" W. Jones, "Nucleic Acids, Their Chemical Properties and Physiological Conduct."

Longmans. Green and Co., London, 1920.
^^ R. Feulgen, "Chemie und Physiologic der Nucleinstoffe nebst Einfiihrung in die

Chemie der Purinkorper." Borntraeger, Berlin, 1923.
^3 P. A. Levene and L. W. Bass, "Nucleic Acids." Chemical Catalog Co., New York,

1931.
1^ T. B. Johnson and D. A. Hahn, Chem. Revs. 13, 193 (1933).
'5 A. Pinner, Ber. 20, 2361 (1887); 22, 1600, 1612 (1889); 26, 2122 (1893).
i« Chem. Abstr. 39, 5867 (1945).
" A. M. Patterson and L. T. Capell, "The Ring Inde.x," p. 52. Reinhold, New York,

1940.
'* T. B. Johnson, m "Organic Chemistry, An Advanced Treatise" (Gilman, ed.).

Vol. 2, p. 948. Wiley, New York, 1938.
'^ F. Schlenk, Advances in Enzymol. 9, 455 (1949).

CHEMISTRY OF PURINES AND PYRIMIDINES

83

in III.-" Both systems, I and II, have been used in a single work.-' Con-
fusion also arises from nomenclature inconsistency. For example, in Chemi-

N

.N/4
3
I

IN-
2C

-C6

I
C5

3N— C4
II

3N-

-C4

2C

C5

IN-

-C6

III

cal Abstracts, IV is called 2,4,6(l//,3//,5//)-pyrimidinetrione, whereas V
is called 6-methyl-2,4-pyrimidinediol,'* and isobarbituric acid is listed as
5-hydroxy -2,4(1//, 3f/)-pyrimidinedione.^-

CH3

H

H

6

2 3 4

H

o I o

H
IV

N

An/

HO

OH

Since the pyrimidine nucleus is symmetrical about a plane including
C-2 and C-5, alternative numbering is possible (see II and III above).

Fischer^ -^ introduced the term purine which he derived from purum and
uricum, and employed the widely used construction (VI) based upon the
Medicus formula now ascribed to uric acid. This heterocycle (VI) consists

IN— C6

I I 7
2C 5C— N

C8

3N-

-N

VI

of a fused pyrimidine and imidazole nucleus in which carbon atoms 4 and
5 are shared by both rings.

The ring and numbering system employed by Chemical Abstracts^^ and

^^ V. Meyer and P. Jacobson, "Lehrbuch der organischen Chemie," Vol. II, p. 1172.

De Gruyter, Berlin and Leipzig, 1920.
*' A. A. Morton, "The Chemistry of Heterocyclic Compounds," p. 483. McGraw-Hill,

New York, 1946.
^^Chem. Ahstr. 44, 12451, 12845 (1950).
" E. Fischer, Ber. 31, 2550 (1898); 32, 435 (1899).

84 AARON BENDICH

The Ring Index^'' for purines is also used in this text (VII), and, for con-
sistency and to indicate more clearly the interrelationships between purines
and pyrimidines, the style (VIII) for pyrimidines is used here. The hex-

IN r

3

VII VIII

Purine Pyrimidine

agonal representations (VII and VIII) are preferred since they conform
closest to those revealed by X-ray studies.^*

The actual structures of pyrimidines and purines depend upon many
conditions (see below) and, even when the effect of a condition (such as
pH) is understood, it is not always possible to write a single structure for
a particular compound. For the sake of brevity it is necessary, therefore,
to be arbitrary in representing the structures. Except where otherwise
noted, amino, hydroxy, mercapto, or other derivatives of these hetero-
cycles will be written as if they had been formed by a replacement of a
hydrogen atom, leaving the K^kul^-type ring intact. Thus uric acid is
referred to as 2 , 6 , 8-trihydroxypurine and 5-methylcytosine as 2-hydroxy-
5-methyl-6-aminopyrimidine, without regard to the obvious operation of
lactim-lactam or amino-imino tautomerism. It is to be understood that
such representations do not necessarily reflect an actual state of a molecule.

3. Occurrence and Distribution

a. Purines and Pyrimidines of Nucleic Acid Origin

Thus far, only two purines are recognized as universal and normal con-
stituents of nucleic acids (PNA and DNA). Adenine^^ was discovered and
isolated from an acid hydrolysate of "nuclein" of beef pancreas in 1885 by
Kossel.2® '^^ He succeeded^^ in converting adenine into hypoxanthine (6-hy-
droxypurine) upon reaction with nitrous acid. Although Medicus^ proposed

24 D. O. Jordan, Progress in Biophysics 2, 51 (1951) ; Ann. Rev. Biochem. 21, 209 (1952).
" Beilstein's Handbuch der organischen Chemie.," 4th ed., 26, 420 (1937). (Herein-
after referred to as "Beilstein.")
2« A. Kossel, Ber. 18, 79 (1885).
" A. Kossel, Z. physiol. Chem. 10, 248 (1886).
28 A. Kossel, Ber. 18, 1928 (1885).

CHEMISTRY OF PURINES AND PYRIMIDINES

85

the correct cyclic structure for hypoxanthine, it was not until 1897 that
concrete chemical evidence for the stiiicture of adenine (IX) was made
available by Fischer in his partial synthesis of both adenine and hypoxan-
thine from 2,6,8-trichloropurine-^ obtained from uric acid:

POCh NH3

Uric acid -> 2,6,8-tnchloropuiine — > 6-amino-2,8-dichloropurine

HI

Adenine

HO NO

> 6-hydroxy-2,8-dichloropurine ^ Hypoxanthine

Total s>aitheses of adenine and hypoxanthine were accomplished in 1904
by Traube.^" (The Traube syntheses are discussed on page 131.) Certain
of the chemical and physical properties of adenine and other purines and
pyrimidines are listed in Table I and are discussed below.

NH

OH

HoN

Guanine (2-amino-6-hydroxypurine, X)^^ was found in 1844 in the ex-
creta (guano) of birds^- "^^ before it was recognized as a nucleic acid con-
stituent.^'* The accepted cyclic structure for guanine, postulated by Medi-
cus,® received considerable support from its conversion by nitrous acid to
xanthine (2,6-dih3droxypurine) and the chloric acid oxidation of guanine
to guanidine and parabanic acid by Strecker.^^ The relationship of guanine

" E. Fischer, Ber. 30, 2226 (1897) .
30 W. Traube, Ann. 331, 64 (1904).
'I Beilstein, 26, 449 (1937).

32 Magnus, Ann. 51, 395 (1844).

33 B. Unger, Ann. 58, 18 (1846); 59, 58 (1846).

34 A. Kossel, Z. physiol. Chem. 8, 404 (1883-1884); see also Z. physiol. Chem. 3, 284
(1879); 5, 152, 267 (1881).

3^ A. Strecker, Ann. 108, 129, 141 (1858); 118, 151 (1861).

86 AARON BENDICH

to the purine system (uric acid) was elucidated by Fischer. ^^ These reac-
tions are summarized below:

H

ii^l^'-^ NH2 I c=o

c \

1
c

H2N I / \ o ^

H2N NH I

xi

Guanidine Parabanic acid

2,6,8-trichloropurine > 6-hydroxy-2,8-dichloropurine ^ 2-amino-6-

CjHsONa

hydroxy-8-
chloropurine

HI

2,6-diethoxy-8-chloropurine > Xanthine < Guanine

Total syntheses of guanine and xanthine were first described by Traube.'^
Five pyrimidines have been isolated to date from nucleic acid sources.
The first to be discovered, thymine (2,6-dihydroxy-5-methylpyTimidine or
5-methyluracil, XI)," was isolated in 1893 by Kossel and Neumann^^'^^
from acid hydrolysates of the nucleic acids of calf thymus and beef spleen.
These investigators, as well as Jones'*" and Steudel and Kossel,^' found that
thymine was not identical with the isomeric 2 , 6-dihydroxy-4-methylpyrim-
idine which had been synthesized and named 4-methyluracil by Behrend''^
in 1884. The studies of Steudel,^^ in which urea was obtained following
permanganate oxidation of thymine, led him to conclude that XI expressed
the proper structure. This conclusion was soon confirmed by a number of

3« W. Traube, Ber. 33, 1371, 3035 (1900).

" Beilstein, 24, 353 (1936).

38 A. Kossel and A. Neumann, Ber. 26, 2753 (1893).

" A. Kossel and A. Neumann, Ber. 27, 2215 (1894).

" W. Jones, Z. physwl. Chem. 29, 20 (1899).

*' H. Steudel and A. Kossel, Z. physiol. Chem. 29, 303 (1899).

*^ R. Behrend, Ann. 229, 1 (1885).

" H. Steudel, Z. physiol. Chem. 30, 539 (1900).

CHEMISTRY OF PURINES AND PYRIMIDINES

87

syntheses of thymine,'*^ "^^ one of which'^^ is illustrated:

O

C2H5OC

CH3

NH2

C

+

KOH, H2O
room temp.

CH3S NH

iS-Methvlthiourea

C

/ \
HO H

Eth}'! formyl
propionate
(enol form)

OH

OH

CHj

//

N

CH3

cone. HCl

reflux 10 hr.

CH3S

2-Methylmercapto-

5-methyI-6-hydroxy-

pyrimidine (20% yield)

N

\n/

HO

Thymine
(99% yield)

Thymine, a ubiquitous component of the deoxy type of nucleic acid (see
Chapters 4 and 10), has not yet been found in the pentose nucleic acids or
in the free state in natural sources, although it has been isolated as an
A^'-xyloside (spongothymidine) from sponges (see below).

In 1894, Kossel and Neumann also^^ discovered as a cleavage product of
calf thymus nucleic acid a basic substance which they named cytosine
(2-hydroxy-6-aminopyrimidine, XII). ^^ Certain of its salts were described*^
^^■^> and this led to its correct empirical composition. Since uracil was
obtained from nitrous acid deamination, and biuret (rather than guanidine)
from permanganate oxidation, Kossel and SteudeP' proposed the accepted
structure of cytosine. The synthesis of cytosine^^ furnished further proof
of its structure. The synthesis is analogous to that for thymine previously
described above in that *S-ethylthiourea and ethyl formylacetate were

4^ H. Steudel, Z. physiol. Chem. 32, 241 (1901).

« H. L. Wheeler and H. F. Merriam, Am. Chem. J. 29, 478 (1903).

« O. Gerngross, Ber. 38, 3408 (1905).

" H. L. Wheeler and D. F. McFarland, Am. Chem. J. 43, 19 (1910).

« Beilstein, 24, 314 (1936).

'' A. Kossel and H. Steudel, Z. physiol Chem. 37, 177, 377 (1902-03).

5" P. A. Levene, Z. physio! . Chem. 38, 80 (1903).

" A. Kossel and H. Steudel, Z. physiol. Chem. 38, 49 (1903).

« W. L. Wheeler and T. B. Johnson, Am. Chem. J. 29, 492 (1903).

AARON BENDICH

condensed in alkaline solution at room temperature to afford 2-ethylmer-
capto-6-hydroxypyrimidine. The hydroxyl group was replaced (phosphorus
pentachloride) by chlorine, which in turn was replaced by an amino group
upon treatment with alcoholic ammonia. The ethylmercapto group was
replaced by hydroxyl upon hydrolysis (125°) with hydrobromic acid. It
was found ^' that the synthetic cytosine and that obtained from wheat
germ and spleen nucleic acids were identical. With the exception, recently
reported, of the DNA of T- even-numbered bacteriophages of E. coli,^^ all
specimens of PNA and DNA in which the pyrimidines have been charac-
terized have been found to contain cytosine.

OH

NH.,

OH

CH3

N

An/

N

An-

HO

N

/

NH.,

N
I ,

NHo

CH3

CH2OH

N

An/

HO HO HO HO

XI XII XIII XIV XV

Thymine Cytosine Uracil 5-Methylcytosine 5-Hydroxymethyl-

Pyrimidines occurring in the nucleic acids

cvtosine

Following a procedure for the isolation of thymine from herring sperm, ^^
Ascoli^® isolated from yeast nucleic acid (1900-1901) a new compound
(C4H4N2O2) which, as he pointed out, corresponded in empirical composi-
tion to the previously postulated (cf. Behrend^^) vracil (XIII)." This sur-
mise was confirmed by the following synthesis of uracil by Fischer and
Roeder,^* who considered their product to be identical wdth the natural
substance :

0

OH

NH2 II

I H

1 HO— C
c + \
/ \ C-H

0 NH2 II

210-220°

^^J H

Br2

cetic

100

-> IN

/\N/

A

acid

CH,

HO H

Urea Acrylic acid

Hydrouracil
OH

OH

1

A

N
I

A-Hr Pyridine
—HBr

100°

->

A

N
I

/

An/^x^

An/

HO

H

HO

Uracil

" W. L. Wheeler and T. B. Johnson, Am. Chem. J. 29, 505 (1903).

CHEMISTRY OF PURINES AND PYRIMIDINES 89

Uracil was also isolated from the nucleic acids of wheat germ ("Tritico-
nuclein"*^), calf thymus and herring sperm,*'' and beef spleen,*^ although
it was suspected*" '^^ that the uracil might have arisen from cytosine^''^2,62
as a result of hydrolysis (10% H2SO4 at 150°). I^racil is now recognized as
a universal constituent of the nucleic acids of only the pentose (or ribose)
type (Chapters 4 and 11). Uracil occurs naturally in the free form in ergot*'
and in more complex combinations (see below).

Until 1925, the composition of the nitrogenous constituents of the nucleic
acids was described only in terms of the five pyrimidines and purines dis-
cussed above; in that year, Johnson and Coghill reported*^ the occurrence
of a homologue of cytosine, o-methylcytosine (XI V),*^ among the hydro-
lytic products of a nucleic acid from tubercle bacilli, tuberculinic acid.
(Johnson had been seeking this substance in natural sources for some
twenty-one years.) It was concluded** that the picrate of the isolated
product was crystallographically identical with that of synthetic 5-methyl-
cytosine,** but no melting point was recorded for comparison. In 1949, an
examination of "5-methylcytosine" prepared by Johnson revealed*^ it to be
a mixture consisting mainly of cytosine. Although a careful chromatographic
study*'^ of the DNA from avian tubercle bacilli did not confirm the early
report,** a paper chromatogram of an hydrolysate of DNA from calf
thymus revealed the presence of a small component ("epicytosine"), the
properties of which led Hotchkiss*^ to the tentative view that it might be
5-methylc>i:osine. Wyatt*^''" has found 5-methylc>i;osine as a definite,
though minor, constituent of the DNA's of mammalian, fish, and insect
sources, and as a major pyrimidine component of wheat germ DNA. The
latter finding has been confirmed .'^^ It could not be detected in bacteriaP"

54 G. R. Wyatt and S. S. Cohen, Nature 170, 1072 (1952).

" W. Jones, Z. physiol. Chem. 29, 461 (1900).

^« A. Ascoli, Z. physiol. Chcm. 31, 161 '(1900-01).

»'Beilstein, 24, 312 (1936).

58 E. Fischer and G. Roeder, Bcr. 34, 3751 (1901).

•" T. B. Osborne and I. F. Harris, Z. physiol. Chem. 36, 85 (1902).

«" A. Kossel and H. Steudel, Z. physiol. Chem. 37, 245 (1902-03).

" P. A. Levene, Z. physiol. Chem. 38, 80 (1903).

" W. Jones and M. K. Perkins, J. Biol. Chem. 62, 557 (1924-25).

" R. p:ngeland and F. Kutscher, Zentr. Physiol. 24, 589 (1910).

«4 T. B. Johnson and R. D. Coghill, J. Am. Chem. Sac. 47, 2838 (1925).

85Beilstein,24, 355 (1936).

"« H. L. Wheeler and T. B. Johnson, Am. Chem. J. 31, 591 (1904).

6' E. Vischer, S. Zamenhof, and E. Chargaff, J. Biol. Chem. Ill, 429 (1949).

« R. D. Hotehkiss, J. Biol. Chem. 175, 315 (1948).

" G. R. Wyatt, Nature 166, 237 (1950).

'» G. R. Wyatt, Biochem. J. 48, 581, 584 (1951).

^' G. Brawerman and E. Chargaff, J . Am. Chem. Soc. 73, 4052 (1951).

90 AARON BENDICH

(including tubercle bacilli) or viral sources/^ and has not been found in the
pentosenucleic acids. A nucleotides^ and nucleosides^ of 5-methylcytosine
have been isolated from the DNA of thymus and wheat germ, respectively.
An excellent discussion concerning this pyrimidine is available .^^

Quite recently, the absence of cytosine in the DNA of coliphage T2 was
reported.ss Since hydrolysis of the DNA with 70% HCIO4 (100°), which
was employed," does not lead to a destruction of cytosine (see also Wyatt^^),
and since other analyses^^ ^9 indicated its apparent presence in T2 and Te ,
a reinvestigation was made by Wyatt and Cohen. ^^ Of great interest was
their discovery of the occurrence in the DNA's of T-even bacteriophages
(but not in the DNA's of thymus or of the host E. coli) of a new pyrimidine
base, 5-hydroxymethylcytosine (XV) in place of cytosine. The base XV,
which withstands the hydrolysis with 88% formic acid (175° for 30 min.)
necessary for its liberation from the DNA's, but which is largely destroyed
by the treatment with HCIO4, has spectral and chromatographic proper-
ties (Rf value in isopropanol-water-HCl)^^ very similar to those of cytosine
(and 5-methylcytosine). These considerations serve to explain the above-
mentioned apparent discrepancies.s6S8,79 a synthesis of the new base (un-
published), its nitrous acid deamination to a uracil-like derivative, and
the isolation of a nucleotide have been reported. ^^•*''

Because of the importance of this new base (XV) in DNA, and its struc-
tural similarity to vitamins Bi (thiamine, XVI) and Be (pyridoxine, XVII),
a more detailed discussion of the chemistry of hydroxymethylpyrimidines
and related compounds is given.

In an attempt to prepare A'^-methyluracil derivatives, Kircher*^ condensed 4-methyl-
uracil (XVIII)''^ with formaldehyde in either dilute acid or alkaline aqueous solution
and obtained 4-methyl-5-hydroxymethyluracil (XIX) in high yield. Proof of its struc-
ture was readily obtained upon its reduction to the known 4 , 5-dimethyluracil (XX) .^* • ^^
The compound XIX was unexpectedly labile to hydrolysis, undergoing a carbon-carbon
cleavage upon mere boiling with water, and the original 4-methyluracil (XVIII) and
formaldehyde were regenerated by this mild treatment. Other, ill-defined, products
(C11H12N4O4 and C12H14N4O5) resulted when XIX was subjected to hot mineral acid
(cf. Endicott and Johnson''-*^ for a description of similar substances made by treating

" G. R. Wyatt, J. Gen. Physiol. 36, 201 (1952).

''^W.E.Cohn, J. Am. Chem. Soc. 72, 2811 (1950); 73, 1539 (1951).

'* C. A. Dekker and D. T. Elmore, /. Chem. Soc. 1951, 2864.

" G. R. Wyatt, Exptl. Cell Research 3, Suppl. 2, 201 (1952).

" A. Marshak, Proc. Natl. Acad. Sci. U. S. 37, 299 (1951).

" A. Marshak and H. J. Vogel, /. Biol. Chem. 189, 597 (1951).

" J. D. Smith and G. R. Wyatt, Biochem. J. 49, 144 (1951).

" L. L. Weed and S. S. Cohen, /. Biol. Chem. 192, 693 (1951).

8" G. R. Wyatt and S. S. Cohen, Ann. inst. Pasteur 84, 143 (1953).

8' W. Kircher, Ann. 385, 293 (1911).

»2J. Schlenker, Ber. 34, 2812 (1901).

83 M. M. Endicott and T. B. Johnson, J. Am. Chem. Soc. 63, 1286 (1941).

" M. M. Endicott and T. B. Johnson, J. Am. Chem. Soc. 63, 2063 (1941).

CHEMISTRY OF PURINES AND PYRIMIDINES

91

NH,

CHo

CHa CH2CH2OH CH2OH

\ / HO i

c=c

CH3

N

\

c— s

CH3

/\N'

— CH2OH

XVI

Thiamine

XVII

Pyridoxine

HO

CH2NH2

-CHoOH

CH3

XVIIa
Pyridoxamine

XVIII with chloromethyl ether). This instability of a pyiimidine-carbinol linkage has
been found to be peculiar only to the hydroxymethji at position 5, since such a group
at position 4 (or 6) or 2, as in 4-h3-drox_ymethylthymine*^' ** and 4-hydroxymethyl-
uracil*' (originally considered as the simplest "nucleosides" of thymine and uracil,

OH

OH

OH

CH2OH

dil. acid
or alkali

hot H2O

N

Sn

CH3

HCl

N

HO CH3

XVIII

HO CHa

XIX

OH

CH2NH2

HO

/Vn/\

HO CH3

XX

OH

CH2OH

N

\^/

N

HO

/^N

XXI

XXII

respectively), 2-methyl-4-h3'droxy-6-hydroxymethylp3'rimidine*^ and 2,6-bis(hydroxy-
methyl)-4-hydroxy-5-methylpyrimidine,*' is not split off by hot water or hot mineral

85 T. B. Johnson and L. H. Chernoff, J. Biol. Chem. 14, 307 (1913).

86 T. B. Johnson and L. H. Chernoff, J. Am. Chem. Soc. 35, 585 (1913).
8^ T. B. Johnson and L. H. Chernoff, J. .4m. Chem. Soc. 36, 1742 (1914).

88 G. E. McCasland, D. S. Tarbell, R. B. Carlin, and N. Shakespeare, J. Ayn. Chem. Soc.
68,2390 (1946).

89 G. E. McCasland and D. S. Tarbell, J. A771. Chem. Soc. 68, 2393 (1946).

92 AARON BENDICH

acid. Since no direct method has been developed for the preparation of substitution
derivatives ( — OH, — NH2 , etc.) of the 5-methyl group of pyrimidines,'" recourse has
been made to indirect methods*^"'" such as the Curtius technique. When this reaction
was applied to ethyl-2-ethylmercapto-6-hydroxypyrimidine-5-acetate, the 5-carbeth-
oxymethyl group was converted to the ethyl ester of the methylurethan ( — CH2-
COOC2H5 -> — CH2NHCOOC2H5) and the urethan was hydrolyzed with liberation of
ethyl mercaptan to yield S-aminomethj'luracil (or "thyminylamine," XXI). ^'-^^ This
compound (XXI) was converted to 5-hydroxymethyluracil (or "thyminyl alcohol,"
XXII) by hydrolysis or nitrous acid deamination. ("Thyminyl alcohol," which should
prove to be identical with the deamination product of 5-hydroxymethylcytosine of
Wyatt and Cohen, ^* could not be prepared'' by the action of formaldehyde upon uracil.)
Both the 5-hydroxymethyl- and the 5-aminomethyluracils were unstable to treatment
with hot water, and in the case of the latter compound (XXI) a carbon-carbon cleavage
resulted with the formation of uracil, ammonia, and formaldehyde. The extreme in-
stability of these substances explains the difficulty '^ in preparing "absolutely pure"
specimens. A similar instability towards hydrolysis of substituted 5-aminomethylura-
cils is manifested by the cleavage of these compounds to uracil, formaldehyde, and a
substituted ammonia.'" The chloromethylation'^ of l,4-dimeth3'luracil with formalde-
hyde and HCl gives rise to l,4-dimethyl-5-chloromethyluracil which is also unstable
towards hot water and (presumably via a conversion to 5-hydroxymethyl) decomposes
into formaldehyde and an insoluble dimethyluracil condensation product.'^ A similar
behavior is described'^' *■* for 4-methyl-5-chloromethyluracil from which the unstable
5-hydroxymethylpyrimidine (XIX) has been prepared.*''

It will be pointed out in greater detail later that many analogies exist in the chemis-
try of pyrimidines and pyridines. For example, the 3 and 5 (or /3) positions of pyridine
and the chemistry of certain j8-substituted pyridines are often akin to those of carbon
5 of pyrimidines. However, the hydroxy or aminomethyl groups of pyridoxine (XVII)
and pyridoxamine (XVIIa) survive severe hydrolytic conditions with strong mineral
acid.9«-'8

As to the stability of thiamine (XVI), most of the available data generally deal
with the integrity of the intact vitamin, rather than the stability of the carbon-carbon
linkage between C-5 of the pyrimidine and its side chain, but it might be mentioned
that thiamine can be heated to 120°C at pH 3.5 without any decomposition.'' The
reader is referred to excellent discussions of the chemistry of the vitamin. i""-'"' As for

'« D. Riehl and T. B. Johnson, Rec. trav. chivi. 59, 87 (1940).

'1 T. B. Johnson and A. Litzinger, /. Am. Chem. Soc. 57, 1139 (1935).

'2 A. Litzinger and T. B. Johnson, /. Am. Chem. Soc. 58, 1936 (1936).

'3 T. B. Johnson and A. Litzinger, J. Am. Chem.. Soc. 58, 1940 (1936).

'^ R. C. Fuson and C. H. McKeever, in "Organic Reactions" (Adams, ed.). Vol. 1,
p. 63. Wiley, New York, 1942.

'^K. Schmedes, Ann. 441, 192 (1925).

'6 M. Hochberg, D. Melnick, and B. L. Oser, J. Biol. Chem. 155, 129 (1944).

'^ E. Cunningham and E. E. Snell, J. Biol. Chem. 158, 491 (1945).

'8 D. Melnick, M. Hochberg, H. W. Himes, and B. L. Oser, /. Biol. Chem. 160, 1 (1945).

"Ref. 101, p. 104.
lo" R. R. Williams and T. D. Spies, "Vitamin Bi (Thiamin), and Its Use in Medicine."

Macmillan, New York, 1938.
'"^ H. R. Rosenberg, "Chemistry and Physiology of the Vitamins." Interscience, New

York, 1942.
'"* F. A. Robinson, "The Vitamin B Complex." Wiley, New York, 1951.

CHEMISTRY OF PURINES AND PYRIMIDINES 93

compounds related to the pyrimidine moiety of thiamine, certain of the derivatives
possessing a substituent such as — CH3 , — CI, or — H at carbon 2 which is not capable
of protopic change (a change such as

c

c

C

/ \ /

^

/ \ /

^

/ \ .

0 N

•^

HO N

■^

-0 N

+ H^

H

etc.) do not show the instability associated with the corresponding uracil or cytosine
derivatives described above. For example, a cjuantitative yield of 2-methyl-5-amino-
methyl-6(4)-hydroxypyrimidine is obtained when the 5-urethanomethylpyrimidine is
heated at 100°C. for 2 hours with concentrated HCl;'"' the aminomethyl group is con-
verted to hydroxymethj-l upon nitrous acid treatment for 7 hours at 100°C.'^ The
hydroxymethj'l group of 4-methyl-5-hydroxymethyl-6-aminoi)yrimidine persists after
heating its 2-chloro derivative with zinc in water at 100° for 7 hours (no yields are cited,
however).'"'' Also, 2-methyl-5-bromomethyl-6-aminopyrimidine, in acid or neutral
solution, withstands several treatments in the autoclave'"^ (cf . Endicott and Johnson*^
and Schmedes'*).

The simple replacement of a hydrogen atom of the stably-bound methyl group of
5-methyl-uracil or -cytosine by hydroxyl or amino (or chloro) gives rise to compounds
possessing a carbon-carbon linkage which is unstable towards hot water or mineral
acid, but not (in the case of 5-hydroxymethylcytosine (XV)) towards hot formic acid
(which is a reducing agent). As yet, no theoretical explanation is available for this
phenomenon and any explanation that is put forward must take into account the
stalMlity of the other derivatives mentioned above.

h. Purines and Pyrimidines in Other Biological Sources

A detailed catalog listing the widespread natural distribution of purines
and pyrimidines has been compiled.'"^ For other listings, the general ref-
erences at the end of this chapter may be consulted. Because of this excel-
lent reference literature, greater emphasis will be given to more recent
developments.

Purine (C5H4N4), the parent member of the series, "^^ has recently been
isolated as a 9-D-riboside ("Nebularine") from a mushroom.'"*'"" Nebu-
larine, the detailed structure of which has been established by synthesis as

i«3 A. R. Todd, F. Bergel, H. 1.. Fraenkel-Conrat, and A. Jacob, J. Chem. Soc. 1936,

1601.
1" H. Andersag and K. Westphal, Ber. 70, 2035 (1937).

"OS W. J. Robbins and F. Kavanagh, Proc. Natl. Acad. Set. U. S. 24, 141 (1938).
los C. Wehmer and M. Hadders, in "Handbuch der Pflanzenanalyse" (G. Klein, ed.),

Vol. 4, p. 405. Springer, Vienna, 1933.
'" Beilstein, 26, 354 (1937).
'08 L. Ehrenberg, H. Hedstrom, N. Lofgren, and B. Takman, Svensk Kern. Tid. 58, 269

(1946).
'09 N. Lofgren, B. Takman and H. Hedstrom, Svensk Farm. Tidskr. 53, 321 (1949).
"" N. Lofgren and B. Liining, Acta Chem. Scand. 7, 15 (1953).

94 AARON BENDICH

9-/S-D-ribofuranosylpurine,"^ has an inhibitory effect on tubercle bacilli
and mitosis in Allium root cells."" It is of interest to record that Fischer
predicted (1907) the natural occurrence of purine when he wrote "... (ich)
halte . . . es nicht fiir unmoglich, das auch das Purin und die Methyl-
purine im tierischen oder pflanzhchen Organismus entstehen.""^ Syntheses
of purine from uric acid^' and from uracil'^' have been accomplished.

Other purines and simple purine derivatives that abound in nature are
mainly of the hydroxy and amino type. Perhaps the most ^^'idespread are
related to adenine. Adenine occurs in the free form, for example, in human
urine along ^^ith hypoxanthine and xanthine (and its methylated forms),"*
in human feces together with hypoxanthine, xanthine, and guanine,"^ and
with guanine in cow's milk."^ It is found as the free base in many
plants. "^""^ It is a constituent of adenosine triphosphate (ATP), and a
number of coenzymes such as DPN, coenzyme A, etc.'"°'^"^^ (see Chapter 4).
The thiomethjdpentoside of adenine isolated from yeast,^^®-^" which has
been sho\\'n by degradation and sjTithesis'-*''^^ to be 9-(5'-deoxy-5'-meth-
ylthio-/3-D-ribofuranosyl) adenine, remained a laboratory curiosity until
its participation (in the form of a sulfonium derivative wath homocysteine)
in transmethylation reactions"^ '^^^ (see also Smith and Schlenk'^*) was dis-

111 G. B. Brown and V. S. Weliky, /. Biol. Chem., 204, 1019 (1953).

112 Ref. 9, p. 68.

113 O. Isay, Ber. 39, 250 (1906).

"^ M. Kruger and G. Salomon, Z. physiol. Chem. 24, 364 (1898); 26, 350, 389 (1898-99).

115 M. Kruger and A. Schittenhelm. Z. physiol. Chem. 35, 153 (1902).

116 C. Voegtlin and C. P. Sherwin, /. Biol. Chem. 33, 145 (1918).

117 K. Yoshimura, Z. physiol. Chem. 88, 334 (1913).

11* A. Winterstein and F. Somlo, in "Handbuch der Pflanzenanalyse" (Klein, ed.),

Vol. 4, p. 362. Springer, Vienna, 1933.
11' H. Bredereck, in "Physiologishe Chemie" (Tlaschentrager and Lehnartz, eds.),

Vol. 1, p. 796. Springer, Berlin, 1951.
12" B. Lythgoe, Ann. Repts. on Progr. Chem. (Chem. Soc. London) 42, 175 (1945).

121 D. M. Needham, Advances in Enzymol. 13, 151 (1952).

122 Phosphorus Metabolism, 1 (1951).

123 Phosphorus Metabolism, 2 (1952).

124 G. W. Kenner, Fortschr. Chem. org. Naturstoffe 8, 96 (1951).

125 W. S. McNutt, Fortschr. Chem. org. Naturstoffe 9, 401 (1952).

126 J. A. Mandel and E. K. Dunham, /. Biol. Chem. 11, 85 (1912).

127 V. Suzuki, S. Odake, and T. Mori, Biochem. Z. 154, 278 (1924).

128 K. Satoh and K. Makino, Nature 165, 769 (1950).

129 F. Weygand, Angew. Chem. 62, 336 (1950).

130 J. Baddiley, O. Trauth, and F. Weygand, Nature 167, 359 (1951).

131 J. Baddiley, /. Chem. Soc. 1951, 1348.

132 G. L. Cantoni, ref. 123, p. 129.

133 G. L. Cantoni, /. Am. Chem. Soc. 74, 2942 (1952).

134 R. L. Smith and F. Schlenk, Arch. Biochem. and Biophys. 38, 159, 167 (1952).

CHEMISTRY OF PURINES AND PYRIMIDINES 95

covered by Cantoni. The antibacterial agent, cordycepin, isolated'*^ from
cultures of the mold Cordyceps militaris (Linn.) Link, is^^^ an adenine-9-
nucleoside of cordycepose, a branched 3-deoxypentose related structurally
to apiose. Adenine is present in pseudovitamin Bio , thus replacing 5,6-
dimethylimidazole in the nucleotide portion of B12 }'^' An antibiotic, Puro-
mycin, occurring in the mold Strepto7nyces alboniger, has been found to be
active against certain bacteria and Trypanosomes;'^* upon acid hydrolysis
it yields the dimethyladenine, 6-dimethylaminopurine, and D-3-amino-
ribose and O-methyl-L-tyrosine.^^^ In its proposed structure, Puromycin is
shown'*^ as a 9-[0-methyl-L-tyrosyl-A^-3'-aminoribosyl]-6-dimethylamino-
purine. G-Dimethylaminopurine has previously been synthesized by the
reaction of 6-methylmercaptopurine with dimethylamine.'^"

In addition to its presence in the free form in bird droppings and other sources al-
ready mentioned, guanine has an interesting distribution in nature. It accounts for the
iridescence of fish scales'^'"'" and as such in many teleosts is an inexpensive source of
"pearl essence."'"' ''*■' It was used in 1656 in France to impart a pearly appearance to
beads. "^ Guanine is found in the eyes of the dogfish''*® (cf. Hopkins'^'), in the excreta
of spiders, '''* in sugar and refuse molasses,'''^ and (with xanthine) in the shoots of a
food-herb (Aralia cordata) commonly used in Japan. '5° The white, shiny appearance
of the skin of many amphibia and reptiles is due to the presence of guanine (see Abder-
halden, "Biochemisches Handlexikon," in the General References section).

Guanine accumulates'*' in large crystalline masses in the bones and other tissues of
the pig in a metabolic disorder'^^ that bears a strong resemblance to human gout. The

"^ K. G. Cunningham, S. A. Hutchinson, W. Manson, and F. S. Spring, /. Cht-m. Sac.

1951, 2299.
i3« H. R. Bentley, K. G. Cunningham, and F. S. Spring, J. Chern. Soc. 1951, 2301.
1" H. W. Dion, D. G. Calkins, and J. J. Pfiffner, J. Am. Chem. Soc. 74, 1108 (1952).
"8 J. N. Porter, R. I. Hewitt, C. W. Hesseltine, G. Krupka, J. A. Lowery, W. S. Wallace,

N. Bohonos, and J. H. Williams, Antibiotics & Chemotherapy 2, 409 (1952).
"8 C. W. Waller, P. W. Fryth, B. L. Hutchings, and J. H. Williams, /. Am. Chem. Soc.

75, 2025 (1953).
"" G. B. Elion, E. Burgi, and G. H. Hitchings, J. Am. Chem. Soc. 74, 411 (1952).
1" Barreswill, Ann. 122, 128 (1862).
>« A. Bethe, Z. physiol. Chem. 20, 472 (1895).

'" G. H. Hitchings and E. A. Falco, Proc. Natl. Acad. Sci. U. S. 30, 294 (1944).
'■•^ D. K. Tressler and J. MacW. Lemon, "Marine Products of Commerce," 2nd ed.,

p. 117. Reinhold, New York, 1951.
'■•^ H. F. Taylor, in "Marine Products of Commerce" (D. K. Tressler, ed.), p. 161.

Chemical Catalog Co. New York, 1923.
"« A. Pirie and D. M. Simpson, Biochem. J. 40, 14 (1946).
1" F. G. Hopkins, Proc. Roy. Soc. (London) B130, 359 (1941-42).
"8 E. Gorup-Besanez and F. Will, Ann. 67, 117 (1849).
i^« E. C. Shorey, /. Am. Chem. Soc. 21, 609 (1899).
ISO K. Miyake, /. Biol. Chem. 21, 507 (1915).

"1 R. Virchow, Arch, pathol. Anat. u. Physiol. (Virchow's) 35, 358 (1866) ; 36 , 147 (1866).
1" W. Mendelson, Am. J. Med. Sci. 95, 109 (1888).

96

AARON BENDICH

absence of guanase in swine'" and of uricase in the human'" and the very low solu-
bility of guanine and uric acid (mono sodium salt) serve to explain, in part, the accu-
mulation of these purines often seen in gout (see Gutman'" Bauer and Kl'emperer'^^
for an extensive description of human gout and the participation of uric acid in this
disease). Uric acid gout also affects the bird.'" Originally mistaken for uric acid,'^*
the extremely insoluble 2,8-dihydroxyadenine (XXIII)^'. >" deposits in crystalline
condition in the kidney tubules of the rat'^o. 'S' following massive administration of
adenine.

Uric acid (XXIV) 1^2 has received a much more extensive treatment in
the Hterature than any other purine^"" "-^o. us. "s.ies (^f. also General Ref-
erences). It is the chief end-product of purine (and protein) metabolism in a
great many, but not all, animal species. Perhaps less well known is its pres-
ence in plants^^i^* and in the wings of certain butterfhes-^^'i^^ev in addi-
tion to its distribution in a variety of body fluids, uric acid is also found in
large amounts as a D-riboside in beef and human red corpusclesi**-"" and in
traces in those of other animal species. On the basis of ultraviolet spectral
properties, the ribose radical of a uric acid riboside from beef blood and
liver was suggested to be situated at position 9 of the purine.^" Another
preparation of uric acid riboside from beef erythrocytes possesses different
spectral properties. 1^2 Yot a synthesis of uric acid, see below.

In 1897, Fischer transformed uric acid into isoguanine (XXV), or 2-
hydroxyadenine) and considered it not unlikely^^ that the latter might be

'" W. Jones and C. R. Austrian, Z. physiol. Chem. 48, 110 (1906).

'" W. D. Geren, A. Bendich, O. Bodansky, and G. B. Brown, /. Biol. Chem 183, 21

(1950).
'" A. B. Gutman, Am. J. Med. 9, 799 (1950).
'^« W. Bauer and F. Klemperer, m "Diseases of Metabolism" (Duncan, ed.), 3rd ed.,

p. 683. Saunders, Philadelphia, 1952.
1" B. F. Kaupp, "Poultry Diseases," 4th ed., p. 228. Alexander Eger, Chicago, 1927.
'58 O. Minkowski, Arch, exptl. Pathol. Pharmakol. 41, 375 (1898).
159 A. Nicolaier, Z. klin. Med. 45, 359 (1902).
'«" A. Bendich, G. B. Brown, F. S. Philips, and J. B. Thiersch, /. Biol. Chem 183, 267

(1950).

'«' F. S. Philips, J. B. Thiersch, and A. Bendich, /. Pharmacol. Exptl. Therap. 104, 20

(1952).
'«2 Beilstein, 26, 513 (1937).
'" H. Biltz, "Die neuere Harnsaurechemie." J. A. Barth, Leipzig, 1936. (Reprinted from

/. prakt. Chem. [2] 145, 65 (1936).)
'" M. Sumi, Biochem. Z. 196, 161 (1928).

'" R. Fosse, P. DeGraeve, and P. Thomas, Compt. rend. 194, 1408 (1932).
'6« V. B. Wigglesworth, Proc. Roy. Soc. (London) B97, 149 (1925).
'" A. Tarttar, Z. physiol. Chem. 266, 130 (1940).

'«8 A. R. Davis, E. B. Newton, and S. R. Benedict, /. Biol. Chem. 54, 595 (1922).
i»9 E. B. Newton and A. R. Davis, /. Biol. Chem. 54, 601 (1922).
"0 E. B. Newton and A. R. Davis, /. Biol. Che7n. 54, 603 (1922).
'" R. Falconer and J. M. Gulland, J. Chem. Soc. 1939, 1369.
'" C. E. Carter and J. L. Potter, Federation Proc. 11, 195 (1952).

CHEMISTRY OF PURINES AND PYRIMIDINES

97

an animal product. The prediction appeared to have been borne out by the
reported isolation of isoguanine from pig blood, '^* although this finding could
not be confirmed'^'' (see also Schiitz"^). However, isoguanine occurs in the
wings of butterflies'^^ and was once'^^ thought to be a pterin ("guano-
pterin"). Isoguanine is also found as the aglycone of the riboside, crotono-
side, in the croton bean {Croton figlium L.)-"^''^^ The structure of crotono-
side as a 9-riboside of isoguanine was inferred from ultraviolet absorption
spectra, ^^° and established as 9-jS-D-ribofuranosylisoguanine by synthesis.^*'
A total synthesis of isoguanine has been accomplished. '^-

N

OH

NH,

OH

HO I

H

XXIII

2,8-Dihydroxyadenine

HO

/-^N/^N

H

H

XXVI

XXVII

anthine

Hypoxan thine

Reference to xanthine (XXVI) '^^ and hypoxanthine or "sarkin"
(XXVII)^*^ has been made a number of times above. These substances are
frequently present in animal sources as a result of the action of the enzymes
guanase and adenase on the parent purines or of nucleosidases on the cor-

" M. V. Buell and M. E. Perkins, J. Biol. Chem. 72, 745 (1927).

'^ S. Bergstrom, P. Edman, and O. Hail, Acta Chem. Scand. 3, 1128 (1949).

'^ F. A. Sciiiitz, Biochem. Z. 273, 52 (1934).

'«R. Punmann, Ann. 544, 182 (1940).

" C. 8ch6pf and E. Becker, Ann. 524, 49 (1936).

"* E. Ciierbuliez and K. Bernhard, Heh. Chim. Acta 15, 464, 978 (1932).

^9 J. R. Spies, ./. Am. Chem. Soc. 61, 350 (1939).

8» R. Falconer, J. M. Gulland, and L. F. Story, J. Chem. Soc. 1939, 1784.

s' J. Davoll, J. Am. Chem. Soc. 73, 3174 (1951).

"2 A. Bendich, J. F. Tinker, and G. B. Brown, ./. .4///. Chem. Soc. 70, 3109 (1948).

" Beilstein, 26, 447 (1937).

8^ Beilstein, 26, 416 (1937).

98 AARON BENDICH

responding nucleoside. Xanthine was discovered by Marcet (1817) in
bladder stones.'*^ Hypoxanthine was found in beef spleen (and so named
by Scherer^*^ in 1850), in beef and horse meat, and in creatine mother-
liquors by Strecker.^^^ The distribution of hypoxanthine and xanthine (and
its methylated derivatives such as caffeine, theophylline, and theobromine)
in tea,'^^ coffee, and cocoa and as urinary constituents is ably discussed
else where. ^^^•'^'' A purine nucleoside, spongosine, isolated from the sponge
Cryptoteihia crypta (Florida) , has been shown to be a pentosylmethylamino-
oxypurine, the aglycone of which has been deaminated with nitrous acid to
a methyldioxypurine.^^^

The free purines and their simple derivatives have been found much more
frequently in nature than the pyrimidines. Of the nucleic acid pyrimidines,
only uracil has been found in the free form,^^ and recently hydrouracil
(2 , 6-dihydroxy-4 , 5-dihydropyrimidine) has been obtained . from beef
spleen.' ^2 a. pentofuranoside of thymine, spongothymidine, thought to be a
xylofuranoside, was isolated^^^'^^ from Florida Cryptoteihia sponges. The
identity of the sugar portion as xylose was supported by paper chromato-
graphic analysis, ^^^ and the point of its attachment to N-3 of the thymine
follows from the resistance of the glycoside towards acid hydrolysis'^'
and from the character of the ultraviolet absorption spectra.'^' ''^^

The coenzyme of the system (galactowaldenase) which catalyzes the con-
version of galactose- 1 -phosphate into glucose-1-phosphate'^^ has been iso-
lated from bakers' yeast ;'"'^* the coenzjone contains uridine, two phos-
phate groups, and glucose and is named uridine diphosphate glucose
(UDPG). Upon gentle acid hydrolysis of TJDPG, glucose and a uridine
diphosphate (shown by synthesis to be uridine-5 '-pyrophosphate^ ^^) were
obtained and the formula XXVIII was assigned ;'^^-'^^ the formula

'** Marcet, "An Essay on the Chemical History and Medical Treatment of Calcul Dis-"

orders," London, 1817: cited in refs. 18, 119, and 183.
86 J. Scherer, Ann. 73, 328 (1850).
" A. Strecker, Ann. 102, 204 (1857); 108, 129 (1858).

88 A. Baginsky, Z. physiol. Chem. 8, 395 (1883-84).

89 W. C. Rose, Physiol. Revs. 3, 544 (1923).

^0 A. A. Christman, Physiol. Revs. 32, 303 (1952).

9' W. Bergmann and R. J. Feeney, /. Org. Chem. 16, 981 (1951).

«2 C. Funk, A. J. Merritt, and A. Ehrlich, Arch. Biochem. and Biophys. 35, 468 (1952).

93 W. Bergmann and R. J. Feeney, /. Am. Chem. Soc. 72, 2809 (1950).

9^ K. Makino and K. Satoh, Abstr. I2th Intern. Congr. Pure Appl. Chem. 1951, 317.

95 J. J. Fox and D. Shugar, Biochim. et Biophys. Acta 9, 369 (1952).

9«L. F. Leloir, ref. 122, p. 67.

97 C. E. Cardini, A. C. Paladini, R. Caputto, and L. F. Leloir, Nature 165, 191 (1950).

98 R. Caputto, L. F. Leloir, C. E. Cardini, and A. C. Paladini, /. Biol. Chem. 184, 333
(1950).

'99 N. Anand, V. M. Clark, R. H. Hall, and A. R. Todd, J. Chem. Soc. 1952, 3665.

CHEMISTRY OF PURINES AND PYRIMIDINES

99

(XXVIII) given here is arbitrary in that the configuration of the glucose
portion is not yet known. Three nucleotide derivatives, all of which in-
clude uridine-5'-pyrophosphate attached to an aminosugar by an acetal-
hke bond, accumulate in penicillin-treated cultures of S. Aureus i^"" '^^^
one of these substances contains an L-alanine residue and another contains
a peptide composed of L-lysine, D-glutamic acid, and three alanine residues.

CH.OH

OH OH OH

XXVIII

Uridine diphospate glucose

OH

H2N N

N

OH

OH

XXIX
Vicine

H,N N NH2

XXX

Divicine

Vicine (CioH]607N4) was isolated from legumes of the species vicia in
1870 by Ritthausen2''2-204 g^^^^j appears to be the first simple pyrimidine
derivative found in nature. It also occurs^'^ in beet juice-"^ and peas.^"®
Upon mild acid hydrolysis, it yields the aglucone divicine (C4H602N4)2''^'^''^
and D-glucose.2°^'2°^ Johnson'"* '^-"^'-'^ suggested a "pyrimidine-nucleoside"
structure for vicine. Levene found'^ ■-"•-'' that half of the nitrogen of vicine

^oo J. T. Park, ref. 122, p. 93.

201 J. T. Park, J. Biol. Chem. 194, 877, 885, 897 (1952).

2»2 H. Ritthausen and V. Kreusler, /. prakt. Chcm. [2] 2, 333 (1870).

203 H. Ritthausen, /. prakt. Chem. [2] 7, 374 (1873); 59, 480, 482 (1899).

^0* H. Ritthausen, Ber. 9, 301 (1876); 29, 2108 (1896).

205 E. O. von Lipmann, Ber. 29, 2653 (1896).

^os E. Schultz, Z. physiol. Chem. 15, 140 (1890); 17, 215 (1893).

20^ E. Fischer, Ber. 47, 2611 (1914).

208 H. H^rissey and J. Cheymol, Bull. soc. chim. hiol. 13, 29 (1931).

209 T. B. Johnson, /. Am. Chem. Soc. 36, 337 (1914).

2'o T. B. Johnson and C. O. Johns, /. Am. Chem. Soc. 36, 545 (1914 .

2'i P. A. Levene, /. Biol. Chem. 18, 305 (1914).

2'2 P. A. Levene and J. K. Senior, /. Biol. Chern. 25, 607 (1916).

100 AARON BENDICH

as well as of divicine was present as free amino groups, one of which was
part of a guanidino system, and he concluded that vicine was a 3-iV-gluco-
side of 2 ,5-diamino-4 ,6-dihydroxypyrimidine. Recent evidence^^^ has shown
this widely accepted formulation to be incorrect. Vicine was assigned the
structure XXIX (i.e., 2,4-diamino-6-hydroxy-5-(jS-D-glucopyranoside)) and
divicine that of 2,4-diamino-5,6-dihydroxypyrimidine (XXX) on the
basis of a study of ultraviolet absorption spectra and a variety of chemical
properties.2'3.214 Divicine belongs to a group of pyrimidines and other com-
pounds (such as ascorbic acid) that contain an enediol, aminoenol, or
amino-eneamine system conjugated to a carbonyl (or potential carbonyl)
group and therefore possess strong reducing properties ;^'^ vicine is non-
reducing. A similar substance, convicine (C]oH]608N3-H20), accompanies
vicine in viaa.^'^^ie Convicine is believed to be a hexoside of 4-amino-2 , 5 , 6-
trihydroxypyrimidine,2'7 and its chemical properties^'^-^'^ indicate that this
glycoside is very closely related to vicine in structure.

A compound of increasing interest and importance, orotic acid^'^ (XXXI,
or uracil-4-carboxylic acid) was discovered in the whey of cow's milk by
Biscaro and Belloni^^^ in 1905; the name is derived from the Greek: oros =
whey. It is found in the milk of sheep and goats, and in smaller amounts
in human, pig, and horse milk,22o-222 and has been isolated from "distillers
dried solubles. "^^s Orotic acid accumulates in large quantities during the
growth of mutants of the fungus Neurospora, which require uridine, cyti-
dine, or uracil,22'i and is required for the growth of L. hulgaricus 09.22'''22'
For a discussion of the participation of orotic acid in nucleic acid pyrimi-
dine biosynthesis see Chapters 23 and 25. A riboside of orotic acid (oroti-
dine) has also been isolated^^^ from Neurospora; unlike uridine and cytidine,

2'3 A. Bendich, Trans. N. Y. Acad. Sci., [2] 15, 58 (1952).
2'^ A. Bendich and G. C. Clements, Biochirn. et Biophys. Acta 12, 462 (1953).
215 H. Ritthausen, J. prakf. Chem. [2] 24, 202 (1881).
2'8 H. Ritthausen and Dr. Preuss, J. prakt. Chem. [2] 59, 489 (1899).
2" H. J. Fischer and T. B. Johnson, J. Am. Chem. Soc. 54, 2038 (1932).
218 Beilstein, 26, 253 (1936).

"9 G. Biscaro and E. Belloni, Ann. Soc. Chim. Milano 11, 18, 71 (1905) ; a fairly detailed
account of this work given in Chem. Centr. II, 76, 63 (1905).

220 O. p. Wieland, J. Avener, E. M. Boggiano, N. Bohonos, B. L. Hatchings, and J. H.
Williams, /. Biol. Chem. 186, 737 (1950).

221 J. W. Huff, D. K. Bosshardt, L. D. Wright, D. S. Spicer, K. A. Valentik, and H. R.
Skeggs, Proc. Soc. Exptl. Biol. Med. 75, 297 (1950); see also /. Arn. Chem. Soc. 72,
2312 (1950).

222 L. E. Hallanger, J. W. Laasko, and M. O. Schultze, /. Biol. Chem. 202, 83 (1953).

223 L. Manna and S. M. Hauge, /. Biol. Chem. 202, 91 (1953).

22" H. K. Mitchell, M. B. Houlahan, and J. F. Nye, J. Biol. Chem. 172, 525 (1948).
226 A. M. Michelson, W. Drell, and H. K. Mitchell, Proc. Nail. Acad. Sci. U. S. 37, 396
(1951).

CHEMISTRY OF PURINES AND PYRIMIDINES lOl

orotidiiie hydrolyzes in dilute mineral acid as readily as do the ribosides
of the purines.

Biscaro and Belloni degraded orotic acid to urea with permanganate and
proposed a 7-memliered cyclic ureido structure for their compound.^'"
Wheeler el a/.-"'*' pointed out that the empirical formula of orotic acid cor-
responded to that of a carboxyuracil, and they synthesized the 5-carboxylic
acid only to find it different from orotic acid. Wheeler also synthesized^^^
uracil-4-carboxylic acid, but its melting point (347°, dec.) was not the same
as that (260°, dec.) reported^^^ for orotic acid. In a reconsideration of this
problem twenty-two years later, Johnson and Caldwell described a syn-
thesis of 2,5-dihydroxypyrimidine-4-carboxylic acid; its melting point,
259°, led them to believe that it was identical wath orotic acid.^-^ It re-
mained for Bachstez"^^ to prove the identity of orotic acid when he com-
pared a specimen (prepared by Biscaro) which now melted at 345-6° (dec.)
with uracil-4-carboxylic acid previously synthesized by Wheeler-^ and by
Behrend and Struve.^^" Although Wheeler did in fact synthesize XXXI,--^
he did not realize that XXXI had resulted from the alkaline hydrolysis
and rearrangement of 5-(carbethoxymethylidene)hydantoin (XXXII, pre-
pared from oxalacetic ethyl ester and urea by MiillerV^^ method) rather than
(as he thought) from the saponification of the ethyl ester of XXXI. For the
clarification and a discussion of these points the reader is referred to the
interesting work of Mitchell and Nyc.^'"-"^ In the other early synthesis,^^"^
4-methyluracil (XVIII) was converted to orotic acid upon alkaline ferri-
cyanide oxidation. Thymine is not oxidized (see Johnson and Schroeder-^^)
to the isomeric 5-carboxylic acid.

An improved synthesis^^^ of orotic acid (which has been adapted-^^ for
the introduction of C'* in position 6) is shown below together with the re-
actions already discussed .

c. Purine and Pyrimidine Derivatives and Analogues oj Biological Importance

The early investigators of purines and pyrimidines could hardly have
suspected that what appeared as academic exercises in pure chemistry
would furnish a broad foundation for future biological thought and en-

"6 H. L. Wheeler, T. B. Johnson, and C. O. Johns, Am. Chem. J. 2,1, 392 (1907).

2" H. L. Wheeler, Am. Chem. J. 38, 358 (1907).

"8 T. B. Johnson and W. T. Caldwell, J. Am. Chem. Soc. 51, 873 (1929).

"9 M. Bachstez, Ber. 63, 1000 (1930).

"0 R. Behrend and K. Struve, Ann. 378, 153 (1911).

"1 R. Muller, J. prakt. Chem. [2] 55, 505; 56, 475 (1897).

"2 H. K. Mitchell and J. F. Nye, /. Am. Chem. Soc. 69, 674 (1947).

2" J. F. Nye and H. K. Mitchell, /. Am. Chem. Soc. 69, 1382 (1947).

23" T. B. Johnson and E. F. Schroeder, /. Am. Chem. Soc. 53, 1989 (1931).

236 C. Heidelberger and R. B. Hurlbert, /. Am. Chem. Soc. 72, 4704 (1950).

102

AARON BENDICH

NH,

C

/W
HS NH

Thiourea

EtO-C^

0
CH

C

/\

NaOEt, H*

HO CH(OEt)j

Ethyl 7,7-diethoxy-
acetoacetate

OH

HS^N'

XHO

2-Thiouracil-
4-aldehyde

OH

OH

N

HO'^N^CHa
XVIII

NaOH

K,Fe(CN),

N

A

HO N

COOH

XXXI

Orotic acid

H

I
N-C=0

^N-C=CHCOOH
I
H

KOH

.NH2 EtO-C=0
NH2 0 = C-CH,COOEt

HOAc,

HCl

0=C:

H

I

,N-

■C=0

Urea

Oxalacetic ethyl ester

^N-C=CHCOOEt
I
H

XXXII

deavor. Examples shall be cited here of the contributions made by these
workers which have strongly influenced modern medicine. Mention has
already been made, for instance, of uric acid and gout. Brugnatelli^ ob-
tained alloxan (2 , 6-dihydroxypryimidine-4 , 5-dione) in 1818 from the
nitric acid oxidation of uric acid. The Wohler and Liebig preparation of
alloxan"* is well known. In 1943, the important discovery was made^^^
that the administration of alloxan to rabbits leads to a necrosis of the
insulin-producing units of the pancreas (the beta-cells of the islets of
Langerhans), thereby providing a tool for the production and study of
experimental diabetes. Alloxan is diabetogenic in many animals including

"6 J. S. Dunn, H. L. Sheehen, and N. G. B. McLetchie, Lancet i, 484 (1943).

CHEMISTRY OF PURINES AND PYRIMIDINES 103

man (see Bailey-^'^ and Lazarow'^^). In 1864, Baeyer^ transformed alloxan
into barbituric acid^^^ (2,4,6-trihydroxypyrimidine), the total synthesis of
which (from urea and malonic acid) was carried out by Grimaux fourteen
years later.^^" The barbiturates are the best known of the hypnotics and
sedatives; barbital (5,5-diethylbarbituric acid) and phenobarbital (5-
ethyl-5-phenylbarbituric acid) have been in popular use since the first of
the two was introduced^"*^ into clinical practice in 1903 (see also Goodman
and Gilman242).

A single change in the structure of uracil, i.e., the substitution of its 2-
hydroxy by mercapto, gives rise to a compound, 2-thiouraciP'*^'^'*^ which
possesses several important biological properties. For example, 2-thio-
uracil and certain of its 4-aIkyl derivatives are effective in the treatment
of hyperthyroidism and thyrotoxicosis in man.^''^"^^^ Studies of antithyroid
activity and structure of substituted thiouracils have been made,^^^ -^^^ and
it has been suggested that these compounds may prevent iodination of
thyroxine precursors.^*^ The mode of action of these drugs is still a mystery.
Better understood, however, is the 2-thiouracil inhibition of E. coli, since
this bacteriostatic action is prevented by uracil.^^^ Uracil also causes a
reversal of the 2-thiouracil inhibition of tobacco mosaic virus biosyn thesis. ^^^
Thiouracil may therefore be classified as an antimetabolite and its antago-
nism by uracil may be a reflection either of the role of uracil as a component
of certain enzyme systems (see above) or of the nucleic acids. (For a dis-
cussion of antimetabolites and chemotherapy, see the literature^^^'"''.)

"' C. C. Bailey, Vitamins and Hormones 7, 365 (1949).

"8 A. Lazarow, Physiol. Revs. 29, 48 (1949).

2" Beilstein, 24, 467 (1936).

240^1. E. Grimaux, Bull. soc. chim. (Nouv. Sdr.) 31, 146 (1879); Compt. rend. 87, 752

(1878).
"1 E. Fischer and v. Mering, Chem. Cenlr. 74, I, 1155 (1903).
"^ L. Goodman and A. Gilman, "The Pharmacological Basis of Therapeutics," p. 126.

Macmillan, New York, 1941.
2« Beilstein, 24, 323 (1936).

2" H. L. Wheeler and H. S. Bristol, Am. Chem. J. 33, 448 (1905).
2^5 H. L. Wheeler and L. M. Liddle, Am. Chem. J. 40, 547 (1908).
"« E. B. Astwood, /. Am. Med. Assoc. 122, 78 (1943) ; Harvey Lectures 40, 195 (1945).
2" H. P. Himsvvorth, Lancet ii, 465 (1943) ; see also p. 483.
"8 W. W. van Winkle, Jr., S. M. Hardy, G. R. Hazel, D. C. Hines, H. S. Newcomer

E. A. Sharp, and W. N. Sisk, /. Am. Med. Assoc. 130, 343 (1946).
2" G. W. Anderson, I. F. Halverstadt, W. H. Miller, and R. O. Roblin, Jr., /. Am.

Chem. Soc. Q7, 2197 (1945).
"» R. H. Williams and G. A. Kay, A7n. J. Med. Sci. 213, 198 (1947).
"1 W. H. Miller, R. O. Roblin, Jr., and E. B. Astwood, /. Am. Chem. Soc. 67, 2201

(1945).
2" F. B. Strandskov and O. Wyss, /. Bacleriol. 50, 237 (1945).
2" B. Commoner and F. Mercer, Nature 168, 113 (1951); Arch. Biochem. and Biophys.

35,278 (1952).

104 AARON BENDICH

Other pyrimidines active as microbial inhibitors have also been known for many
years, and include 5-methylthiouracil,2^'' dithiothymine, and 5-hydroxy-, 5-amino-,
5-halo-, and S-nitrouracils.^**"^*" Certain 2,6(4)-diaminopyrimidines containing groups
such as benzyl, phenoxy, or phenyl at position 5 are powerful antagonists of folic or
folinic acids in a variety of biological systems^^' and inhibit the development of frog
gggg 262 (por a discussion of these acids and purine and pyrimidine biosynthesis, see
Shive.^^^") The structural resemblance between the antimalarial Paludrine {N'-p-
chlorophenyl-A'5-isopropylbiguanide) and certain of these 2,6(4)-diamino-5-aryloxy-
pyrimidines led to the suggestion that Paludrine might also possess anti-folic activity
and that these pyrimidines might be effective antimalarials. ^^^ These considerations
were confirmed upon experimentation, and many active derivatives were prepared. ^*^'^**
One of the most eflfective is 2,6(4)-diamino-5-(p-chlorophenyl)-6-ethylpyrimidine
(Daraprim),2^^'28^ which is about 1000 times more active than quinine and about 200
times more active than Paludrine. It is of interest that the open-chain Paludrine was
originally designed from a study of the active structural moiety of active 2-(phenyl-
guanidino)pyrimidines.^*' Other examples of pyrimidines used in chemotherapy are
the pyrimidine sulfonamides. ^^^

Just as in the cases of the pyrimidines cited above, alteration of the
structures of naturally occurring purines has yielded compounds of value
in biology and medicine. The changes that have furnished important ana-
logues to be discussed here have involved substituents at C-2 and C-6 as
well as the replacement of carbons 2 and 8 by nitrogen. The 2-amino-
substituted adenine, 2 , 6-diaminopurine or DAP^^-'^^'' has been found to

"4 R. O. Roblin, Jr., Chem. Revs. 38, 255 (1946).

2** L. D. Wright, Vitamins and Hormones 9, 131 (1951).

2*6 D. W. Woolley, "A Study of Antimetabolites." Wiley, New York, 1952.

2" A. Albert, "Selective Toxicity with Special Reference to Chemotherapy." Methuen,

London, 1951.
"8 G. H. Hitchings, G. B. Elion, and E. A. Falco, /. Biol. Chem. 185, 643 (1950).
2" G. H. Hitchings, G. B. Elion, E. A. Falco, P. B. Russell, and H. VanderWerfT, Ann.

N. Y. Acad. Set. 52, 1318 (1950).

260 G. H. Hitchings, G. B. Elion, E. A. Falco. P. B. Russell, M. B. Sherwood, and H.
VanderWerff, /. Biol. Chem. 183, 1 (1950).

261 G. H. Hitchings, E. A. Falco, H. VanderWerff, P. B. Russell, and G. B. Elion, J.
Biol. Chem. 199, 43 (1952).

262 S. Bieber, R. F. Nigrelli, and G. H. Hitchings, Proc. Soc. Exptl. Biol. Med. 79, 430
(1952).

262a \y Shive Vitamins and Hormones 9, 75 (1951).

263 E. A. Falco, G. H. Hitchings, P. B. Russell, and H. VanderWerff, Nature 164, 107
(1949).

264 L. G. Goodwin, Nature 164, 1133 (1949).

266 E. A. Falco, P. B. Russell, and G. H. Hitchings, /. Am. Chem. Soc. 73, 3753 (1951).

266 p. B. Russell and G. H. Hitchings, /. Am. Chem. Soc. 73, 3763 (1951).

267 L. H. Schmidt and C. S. Genther, /. Pharmacol. Exptl. Therap. 107, 61 (1953).

268 H. R. Ing, in "Organic Chemistry, An Advanced Treatise" (Oilman, ed.), Vol. 3,
p. 392. Wiley, New York, 1953.

269 E. H. Northey, "The Sulfonamides and Allied Compounds," p. 31. Reinhold, New
York, 1948.

270 W. Traube, Ber. 37, 4544 (1904).

CHEMISTRY OF PURINES AND PYRIMIDINES 105

serve as a precursor of the guanine of nucleic acids of the rat'-^' and the
purines of those of other species (see Chapters 23 and 25) . EarHer studies,
however, revealed DAP to be a potent inhibitor of L. casei; this inhibition
could be reversed by adenine.'-^" '"-^^ Many other effects of DAI^ have been
recorded (summarized in Wright-''^), and these include a prolongation of
the life of leukemic mice"'' '"^ and damage (which is prevented by adenine)
to sarcoma cells in tissue culture"^ and to the hematopoietic apparatus and
intestinal epithelium of small mammals.-" DAP affects a number of bio-
logical systems rich in nucleic acids: it inhibits the multiplication of vaccinia
virus in chick embryonic tissue,"^ of Russian Spring Summer Encephalitis
virus in vitro^''^ and in vivo,"^^^ and of kappa particles in killer Paramecium
aurelia}^^ '^^^ These effects are reversed by adenine. Other inhibitory ade-
nines are the 2-chloro,^^^ 2-thio, and 2-ethylthio derivatives.^^^

The ultimate replacement of the amino group of adenine by mercapto
has been accomplished by the treatment of hypoxanthine with phosphorus
pentasulfide.^'"' The resulting 6-mercaptopurine, which inhibits the growth
of L. casei,-^^ ■^^^'' possesses the unique and important property of rendering
a malignant tumor (mouse sarcoma 180) nonviable.^*^^ It has a marked
inhibitory effect on many mouse and rat tumors j^*^" it shows a toxicity to
mammals that is suggestive of an antagonism of polynucleotide biosyn-

"1 A. Bendich, S. S. Furst, and G. B. Brown, J. Biol. Chem. 185, 423 (1950).

"2 G. H. Hitchings, G. B. Elion, H. VanderWerff, and H A. Falco, J. Biol. Chan. 174,

765 (1948).
"3 G. B. Elion and G. H. Hitchings, /. Biol. Chem. 187, 511 (1950).
"" J. H. Burchenal, A. Bendich, G. B. Brown, G. B. Elion, G. H. Hitchings, C. P.

Rhoads, and C. C. Stock, Cancer 2, 119 (1949).
"6 H. E. Skipper, L. L. Bennett, Jr., P. C. Edwards, C. E. Bryan, O. S. Hutchison, J.

B. Chapman, and M. Bell, Cancer Research 10, 166 (1950).
"^ J. J. Biesele, R. E. Berger, A. Y. Wilson, G. H. Hitchings, and G. B. Elion, Cancer

4, 186 (1951).
2" F. S. Philips and J. B. Thiersch, Proc. Soc. Exptl. Biol. Med. 72, 401 (1949).
"8 R. L. Thompson, M. L. Price, S. A. Minton, Jr., G. B. Elion, and G. H. Hitchings,

J. Immunol. 65, 529 (1950).
"!• C. Friend, Proc. Soc. Exptl. Biol. Med. 78, 150 (1951).

280 A. E. Moore and C. Friend, Proc. Soc. Exptl. Biol. Med. 78, 153 (1951).

281 W. E. Jacobson, M. Williamson, and C. C. Stock, J. Exptl. Zool. 121, 505 (1952).

282 M. Williamson, W. Jacobson, and C. C. Stock, J. Biol. Chem. 197, 763 (1952).

283 J. J. Biesele, R. E. Berger, and M. Clarke, Cancer Research 12, 465 (1952).

284 G. B. Elion, G. H. Hitchings, and H. VanderWerff, ./. Biol. Chem. 192, 505 (1951).

285 Proc. Avier. Assoc. Cancer Research 1, 1953;
(«) G. B. Elion and G. H. Hitchings, p. 13.

(b) D. A. Clarke, F. S. Philips, S. S. Sternberg, C. C. Stock, and G. B. Elion, p. 9.
'<=> K. Sugiura, p. 55.

(■J) F. S. Philips, S. S. Sternberg, D. A. Clarke, and G. H. Hitchings, p. 42.
(<=> J. H. Burchenal, D. A. Karnofsky, L. Murphy, R. R. Ellison, and C. P. Rhoads,
p. 7.

106

AARON BENDICH

thesis ;^^^<^ and results of preliminary application to human neoplastic dis-
gg^gg285e have stimulated further clinical trials with this drug.

In 1901, Gabriel and Colman^^^ converted 4 , 5-diamino-6-methylpyrimi-
dine into 6-methyl-8-azapurine (or 7-methyl-l-i;-triazolo[c^]pyrimidine)
upon reaction with nitrous acid. The isosteric relationship of this compound
to 6-methylpurine2^ is obvious. With the expectation of preparing purine
antagonists, Roblin et al."^^ applied the above reaction to the appropriate
4,5-diaminopyrimidines and obtained the 8-aza analogues of adenine, of
guanine (XXXIII, or "guanazolo"), and of hypoxanthine. Those of adenine
and guanine exhibited an antibacterial activity (against E. coli and *S.

OH

I

N

NHo

HONG

H2N NH2

2,4,5-Triamino-6-
hydroxypyrimidine

H2N

OH

I

N

\

H
XXXIII

8-Azaguanine

NH

NHo

HN

4-Amino-5-imidazole-
carboxamide

XXXTV

2-Azaadenine

aureus), and this action was reversed by the respective parent purines.
Other 8-azapurines have been sjmthesized,^**'^^^ but 8-azaguanine has
provoked the most interest. It inhibits the growth of the guanine-requiring
protozoan T. geleii,'^^^ the mammary adenocarcinoma Eo 771 in C57 black
mice,2^^'2^2 and the development of several plant virus infections.^^' A

286 S. Gabriel and J. Colman, Ber. 34, 1234 (1901).

2" R. O. Roblin, Jr., J. O. Lampen, J. P. English, Q. P. Cole, and J. R. Vaughan, /.

Am. Chem. Soc. 67, 290 (1945).
288 L. F. Cavalieri, A. Bendich, J. F. Tinker, and G. B. Brown, /. Am. Chem. Soc. 70,

3875 (1948).
28» P. Bitterli and H. Erlenmejer, Helv. Chim. Acta, 34, 835 (1951).

290 G. W. Kidder and V. C. Dewey, /. Biol. Chem. 179, 181 (1949).

291 G. W. Kidder, V. C. Dewey, R. E. Parks, Jr., and G. L. Woodside, Science 109, 511
(1949).

CHEMISTRY OF PURINES AND PYRIMIDINES 107

clue to the mechanism of the inhibitory activity of this agent (and possibly
others) is furnished by the reports of its apparent incorporation into the
nucleic acids of T. geleii^^* and the normal and tumor tissue nucleic acids
of mice;-^^'^^^ its incorporation into tobacco mosaic virus nucleic acid is
supported by the isolation therefrom of a and b isomers of 8-azaguanylic
acid.^^^ Other instances of this type of phenomenon are the incorporation
of the inhibitors 5-bromouracil into the nucleic acids of S. faecalis-^^ and
2,6-diaminopurine into the nucleoside phosphates of the mouse.^^^

A new series of analogues in which carbon 2 of the purines is replaced by
nitrogen, i.e., the imidazo-1 ,2,3-triazines, has been developed.^"" A product
of the acid degradation of adenine, i-amino-S-imidazolecarboxamidine^"'
is converted upon reaction with nitrous acid into 2-azaadenine (XXXIV) i^""
in a similar fashion, 2-azahypoxanthine is obtained from the carboxamide.
The adenine analogue exerts a powerful inhibitory effect on a number of
microorganisms,^"" and mouse sarcoma 180 cells in tissue culture,^"^ and
these actions could be blocked by adenine. 2-Azaadenine also inhibits xan-
thine oxidase.^"^

II. General Properties of Purines and Pyrimidines

Many diverse techniques have been used to study the physical and chemi-
cal properties of the purines and pyrimidines and some of the&e are sum-
marized in this section. Detailed aspects of X-ray, infrared, and ultraviolet
studies are considered elsewhere (cf. Jordan, Chapter 13; Beaven, Holiday,
and Johnson, Chapter 14).

1. Physical Properties

a. Solubility; Distribution Studies; Chromatography

The pyrimidines and purines which are dealt with here should be classi-
fied as aqueous-soluble (rather than organic-soluble) despite the fact that

-'^ K. Sugiura, G. H. Ilitchings, L. F. Cavalieri, and C. C. Stock, ('(inccr Research 10,

178 (1950).
"5 R. E. F. Matthews, Nature 167, 892 (1951); J. Gen. Microbiol. 8, 277 (1953).
"* M. R. Heinrich, V. C. Dewej-, R. E. Parks, Jr., and G. W. Kidder, J. Biol. Chem.

197, 199 (1952).
"5 J. H. Mitchell, Jr., H. E. Skipper, and L. L. Bennett, Jr., Cancer Research 10, G47

(1950).
"« L. L. Bennett, Jr., H. E. Skipper, and L. W. Law, Federation Proc. 12, 300 (1953).
"^ R. E. F. Matthews, Nature 171, 1061 (1953).

298 F. Weygand, A. Wacker, and H. Dellweg, Z. Naturforsch. 7b, 19 (1952).
"9 G. P. Wheeler and H. E. Skipper, Federation Proc. 12, 289 (1953).
300 D. W. Woolley and E. Shaw, J. Biol. Chem. 189, 401 (1951).
30' L. F. Cavalieri, J. F. Tinker, and G. B. Brown, J. Am. Chem. Sac. 71, 3973 (1949),

302 J. J. Biesele, Cancer 5, 787 (1952).

303 E. Shaw and D. W. Woolley, J. Biol. Chem. 194, 641 (1952).

108

AARON BENDICH

t4-t

■a

-o -o -o

T3 -0 -0 -0

•~. ■-,

■-. -c

T3

■o

■o

<

•o -o

73

■0 -0 ■«

TS ■-

■"

n

O) O

CJl CO

CI

(M 00

o

'^O

t-. (N

00

IC CD

r— I

Cl

iM ^ CO

CO 05 «D iC

O

O

o

lO

OD

U5

t^ -<*<

lO

t^ t^ 00

in ^

en

on

X

05

^ « 1"^

00 „ „ O

r^

C3

(M

CO

Oi

^ -*

^

^ _ 05

CO C<)

05

CO

^^o

t^ 9 ^ q6

Oi

t^

00

CD

«o

t^

S "i

lO

^ «=? i^

lei CD

00

r^

s

(M ■*

O — 1

3

(N Ol

1—1 1— (

05

c^j (M <^' r^

-^ _ •■ CO
(M 1^ O (M

_- •- Ol 05 C3 CO o

>0 lO 00 t^ CD 00 CD

• • (N (N (N C^ (M

CI ■-*< CD lO
lO 00 r^ CD
<M (M (M C<l

^

Si JvJ to CI O (M O

"-^ "^ 00 00 CD CD r^

- - <M (N (M (M <M

(N

(N

(N

m

o

O

(M

O

^

C<)

o

■^

o

Tf

CO

o

^

■^

CO

l-H

^

o

CO

-^

(M

t^

CO

•^

?5

05

C5

a

o

ffi

1

tJ.

i-H

1

I— 1

1

I— 1

t^

^

C^

'-'

^

^

O

"3

m

ctj

tf

S P

^ i
^ ^

(M O

tf

O

S

T-H 00

I o

CD <N

3

o

a>

-4-3

a

a;

D,

o;

>>

a

s

fl

c

++

o ++

02

o

TS

-fj

>5 -iS

o

U

D

a

03

13

>.

o

o

QJ «

>,

>>

>,

a

S'S.

w

Ph

U

W5

lO

a

5 w

CHEMISTRY OF PURINES AND PYRIMIDINES

109

^ CO O GO

CO c^ ^ oo" Qo" o" 2 ^ — ^

00

-^

lO

;:^

N

CO

s

GO

Tf

1 lO CO O '^ l-~

COCOiO'^'^'-^iOiOCO'^C/DO

1 CO 00 t^ o

. CO
iM (M oq

(M

^ lO C2
CO CO ^
(M C^ <M

1^ c; CO oi C5 -^ 1—1
I • Tt* ■

i-HCO-— 'iCcOCOi— icOC<IC5>— lO

=^ ■ ^

a 6

O 0)
O -^^

O "5

CO 00
CO (M

•s s i

c

03 G

03

3.^

T)

U, 3

u,

O CO

<<

o

i-( (N

c

O
j3

>1

J3

G

3

.G

c^

Q

-r)

03

tn

03

ffi X

03 0<

i

s

CO

^

J

.^

s

g-

■o

03

03

C7i

c

^

'-s

o

^

-

1

a!

c
c

S t

.S

i c

ffl

•-^

2 2.

M-

c
2

— CO
IS ■«

5 s i

T3

c

— -t3
C

i

z

c

^_ c
S.I

S c

-J '^ c

C3 J=

i

o
W

IS

ao

c
c

0 .

■2 kJ c^

^ ^

0.

w

"S -c

s^

<

ftt.

fe < ^

=3 a; g
J? S

1-5 -' j:

M 4)

M

C

0.

to

05

1

■-2

M
C

o>

c
>

c3 o

^■^

o

.ic3

u^

is

J=

O

s

.2 0 -^

"5

i

h4

CO

W [i] <

^ hJ w

B S H

iL o

■3 -2

•^ ^ >>

" 2 "5 .sJ

S " •;:

'o "^ ,

X ^

S o

--. j:? io

c S

. 5 ^

c5

g5

. CJ .

s ^ a

J3 -S "o

o 05 H,

.A W

00 M ^~, . •

U5 03 oo *^

. ^ CI

• ~^ ?2 c?

■* o c i

>^ o o p

i£

m

d

§

m z;

Ol

c

c-

C

^-

o

-o

Fi

0!

>.

£

CO

^

tf

.^

c

ti

u

H

oo

fe
►^

°> p- 3 ,- "S •- "2

— I U <l;

^ C ^ .^ 0^

05 oq ^

pa ;;
> %
P-i iJ cc i-J

d w o s w <

11^
rt « E <

110 AARON BENDICH

many of the hydroxy and amino derivatives have a limited water solubihty
(Table I). The parent compounds are very soluble in water at ordinary
temperatures; the introduction of hydrogen-bonding groups ( — OH or
— NH2) into pyrimidine or purine results not only in a reduction in water
solubility, but in an increase in melting point as well. (An analogous situa-
tion is seen in the pteridine system.^"'') This effect is especially true of the
purines: the least soluble is 250,000 times less soluble than purine itself.
The replacement of the 2-hydroxyl of xanthine, or of the 6-hydroxyl of
uric acid, by an amino group is manifested by a further marked decrease
in water solubility (cf. Albert^"'*). However, all the compounds listed in
Table I are easily soluble in dilute aqueous alkali or mineral acid (or both) .
A common practice in the purification of the less-soluble compounds in-
volves their solution in dilute acid or alkali; crystallization often follows
neutralization. Whereas pyrimidine, purine, and the simple alkyl, aryl,
aryloxy, diarylamino, or halo derivatives are more or less soluble in com-
mon organic solvents, the other compounds listed in Table I are, in general,
extremely insoluble.

For these reasons, preference has been given to the use of aqueous sys-
tems for the purpose of separation, purification, and characterization of
pyrimidines and purines of biological interest. For example, the counter-
current distribution technique^^^ has been successfully applied to such
compounds^"^ in a system consisting of n-butanol and M phosphate, pH
6.5. The coefficients of distribution of pyrimidines and purines between
the two phases of this system (Table I) are very useful in identification and
characterization. The method is also valuable in the separation of mixtures
of these compounds and for the estimation of homogeneity. Another ex-
ample is the use of a variety of aqueous systems in column and paper
chromatography. These techniques are discussed in detail in Chapters 6
and 7. The Rf values of some pyrimidines and purines in an isopropanol-
water-HCl system are listed in Table I.

h. Criteria of Purity and Identity; the Value of Ultraviolet Absorption Spectra;
Ionization; Tautomerism; the Value of the Isoshestic Point

For discussions of the principles of the purity and identity of organic
compounds see Pirie^"^ and Eyring.^°^

Table I enumerates the melting and decomposition points of a few pyrimi-
dines and purines, and in some cases those of a salt. The substituted com-

304 A. Albert, Quart. Revs. (London) 6, 197 (1952).

306 L. C. Craig, Fortschr. chem. Forsch. 1, 292, 302, 312 (1949).

306 J. F. Tinker and G. B. Brown J. Biol. Chem. 173, 585 (1948).

307 N. W. Pirie, Biol. Revs. 15, 377 (1940).

308 H. Eyring, Anal. Chem. 20, 98 (1948).

CHEMISTRY OF PURINES AND PYRIMIDINES 111

pounds are characterized either by a very high point or by infusibility (the
latter is especially true of the hydroxylated purines). ^°^ This traditional
approach (including mixed melting points) can, therefore, only be of limited
value in the assessment of purity and identity, and in some instances (see
discussion on orotic acid, above) it may even be misleading. However, the
number of physical criteria of purity that are available (-when taken to-
gether) is sufficient to satisfy the most fastidious worker.

When the amount of a specimen to be tested is small, paper chromatog-
raphy (Chapter 7) is the method of choice. Microgram quantities are used,
and the sample, after elution and spectrographic evaluation, is subjected
to reanalysis on paper. The specimen, to be judged "pure," must show a
single "spot" in more than one solvent system and constant spectral
characteristics (at more than one pH) after each chromatographic run.
For radioactive materials, coincidence with radioactivity must be estab-
lished. The same considerations apply to column chromatography (Chapter
6) in that a single band must be demonstrated and the variables are the
adsorbents (anion- or cation-exchange resins, starch, etc.) as well as the
solvent systems. It is highly unlikely that a mixture of pyrimidines or
purines would behave as a homogeneous substance when subjected to all
these tests.^'"

When an adequate supply of a preparation is available, and a standard
specimen of unquestioned purity is desired, the above techniques as well
as the follo\ving are used. The substance is subjected to repeated fractiona-
tion by crystallization (salt as well as neutral forms), passage through a
column, countercurrent distribution, etc., until constancy is achieved with
respect to properties such as those hsted in Tables I and II, and a cor-
responding constancy with respect to isotope content if the compound is
so labeled. The determination of elementary composition presents no special
problem with pyrimidines and purines.

Additional criteria are the reproducibility of ultraviolet data, dissocia-

'•" The eutectic mixtures of a number of pyrimidine and purine derivatives with di-
cyandiamide exhibit well-defined melting points which are much lower than those
of the original derivatives (K. Dimroth and H.-G. Meyer-Brunot, Biochem. Z.-323,
343 (1952)). The determinations are easily carried out on tiny quantities.

"" It is possil)le, for example, that the growth response of a bacterial mutant may be
due to a small impuritj' in a purine or pyrimidine preparation. The importance of
using homogeneous compounds in biological studies, therefore, cannot be stressed
too strongly. Yet, despite the ease and simplicity of paper chromatography, the label
on a bottle has often been accepted too literally. As examples of extreme inhomo-
geneity, the author has examined by paper chromatography a number of commer-
cially available "hypo.xanthine" preparations only to find (with one exception)
these to contain, on the average, about 20% of adenine. One sample of "guanine"
was found to show some four to five "spots, " and only one small "spot" corresponded
in Rf value to guanine.

112 AARON BENDICH

tion constant values, and isosbestic points. These are of especial value since
all three may be determined on microgram quantities from the same set of
measurements without loss of the sample. These considerations depend
upon the f§;Ct (Chapter 14) that the nature and magnitude of absorption
of ultraviolet light by a variety of pyrimidines and purines vary wth the
pH of their aqueous solutions. This is attributable to ionization of one or
another group and is derived from the fact that the neutral and ionic species
of a given molecule exhibit different spectra. In order to. learn which of
these species (or mixture of species) is responsible for a spectrum at a
particular pH, the dissociation constant must be known. This has involved
the prior determination of (apparent) dissociation constant (s), and, from
this information,^"-^" spectra have been recorded at a pH so selected that
only the neutral, the anionic, or the cationic form (if any of these is pos-
sible) is present in solution.

A variation of this method^^^'*^' is based upon the observations that the
curves relating pH to absorption at particular wavelengths of ultraviolet
light are very similar to the dissociation curves. (Such curves, for example
for phenol,^'^ are superimposable.) In this technique, the spectra are de-
termined over a wide pH range to give a -more or less continuous picture
of the spectral changes. This often affords a convenient method for esti-
mating apparent dissociation constants and the proportions of absorbing
forms present in solution at specific pH values. Because of the difficulty
(or, at times, the impossibility) inherent in potentiometric measurements
at very low or high pH values, a very feeble dissociation may be missed,
and, accordingly, an erroneous conception of ionization behavior is ob-
tained. Such has been the case with the second dissociation of uraciP'*
(cf. Shugar and Fox^^O and xanthine^'^ (cf. Cavalieri et al}^^). The second
dissociations are demonstrable, however, by spectrophotometry. Apparent
dissociation constants should be determined by both methods, and a com-
parison for certain pyrimidines and purines is given in Table II. It is to be
noted that these values are not strictly valid in the thermodynamic sense

3'i D. J. Brown and L. N. Short, /. Chem. Soc. 1953, 331.

3'2 J. R. Marshall and J. Walker, J. Chem. Soc. 1951, 1004.

3'3 M. P. V. Boarland, and J. F. W. McOmie, /. Chem. Soc. 1952, 3716.

31" P. A. Levene, L. W. Bass, and H. S. Simms, /. Biol. Chem. 70, 229 (1926).

31B A. G. Ogston, /. Chem. Soc. 1935, 1376.

316 A. Albert, D. J. Brown, and G. Cheeseman, /. Chem. Soc. 1951, 474.

"T N. Whittaker, /. Chem.. Soc. 1951, 1565; 1953, 1646.

3>8 W. Stenstrom and N. Goldsmith, /. Phys. Chem. 30, 1683 (1926).

"' J. J. Fox and D. Shugar, Bull soc. chim. Beiges 61, 44 (1952).

320 D. Shugar and J. J. Fox, Bull soc. chim. Beiges 61, 293 (1952). -

321 D. Shugar and J. J. Fox, Biochim. et Biophys. Ada 9, 199 (1952).

322 L. F. Cavalieri, J. J. Fox, A. Stone, and N. Chang, /. Am. Chem. Soc. 76, 1119 (1954).

323 E. A. Johnson, Biochem. J. 51, 133 (1952).

TABLE II
Apparent Dissociation Constants* of Pyrimidines and Purines

Compound

Pj^rimidine

Cytosine

5-Meth}'lcytosine

Uracil

Thymine

l-MethyluraciH

3-Methyluracil*

1 ,3-Dimethylurat'il

Orotic acid

Barbituric acid

Barbital

Purine

Adenine

Guanine

Hypoxanthine

Xanthine

Uric acid

Method of measurement

Spectrophotometric

pA'a,

2.52

7.53
5.4

pA\

4.45'

12.2

4.6

12.4

9.5

>13

9.5

>13

9.95

—

9.75

—

none

~2.8

9.45/

3.9

12.5

7.85

12.7

8.90

11.63
10.6

Ref.

Titri

metric

Concen-

pA'ai

pAo2

tration

Uet.

1.30

.

.1//15

a

4.60

12.16

"^

9.45

.

c

9.94

—

<-•

9.99

—

A//25

<=

9.71

—

A//25

'

none

"

2.40

—

B

3.98

—

'

7.91

—

i

2.39

8.93

M/10, .1//100

I

4.15

9.80

m

3.3

9.2"

m

8.8

12.0

0

7 . 7

—

.17/1000

P

o.S

—

Q

5.78

(5.85)

.17/10,300

T

* Expre.ssed as pA'a , where \>Ka = 1/log Ka , and A'q = (A~)(H+)/(AH) for "acidic" and =
(B)(H + )/(BH'^) for "i)asic" dissociations. Thf designations pifai and pifo2 refer to the first and second dis-
sociations actually measured. Only for pyrimidine (unsubstituted) and purine do the listed p/Cai values
refer to the basicity of ring nitrogen.

" A. Albert, R. Goldacre, and J. Phillips, J. Chem. Soc. 1948, 2240.

'' D. Shugar and J. J. Fox, Btochim. el Biophys. Acta 9, 199 (1952).

<" P. A. Levene, L. W. Bass, and H. S. Simms, J. Biol. Chem. 70, 229 (1926)

'^ Referred to as "3-methyluracil" by an alternative nomenclature: D. Shugar and J. J. Fox, Biochim. et
Biophys. Acta 9, 199 (1952).

' Referred to as "1-methyluracil": see previous reference.

/ Orotic acid shows pKaz > 13. The carboxyl dissociation (pKai) appears to have a negligible effect on
the second and third dissociations since they are very similar to the corresponding values for uracil and thy-
mine.

» M. Bachstez, Ber.63, 1000 (1930).

'' J. J. Fox and D. Shugar, Bull. soc. chim. Beiges 61, 44 (1952).

' .1. K. Wood, J. Chem. Soc. 89, 1831 (1906); cone, probably .W/64.

' M. E. Krahl, J. Phys. Chem. 44, 449 (1940).

* A. Bendich, P. J. Russell, Jr., and J. J. Fox, J. Am. Chem. Soc. 76, (1954), in press.
' A. Albert and D. J. Brown, J. Chem. Soc. 1954, 2060.

■" H. F. W. Taylor, J. Chem. Sec. 1948, 765; cone, from 0.0012 to 0.007 M.

" pKas for guanine = 12.3.
H. F. W. Taylor, "Acid Base Properties of Nucleic Acids," Doctoral Thesis, London Univ., London,
1946; taken from D. O. .Jordan, Progr. Biophys. and Biophys. Chem. 2, 51 (1951); Ann. Rev. Biochem. 21, 209
(1952).

^ A. G. Ogston, J. Chem. Soc. 1935, 1376.

' W. His, Jr., and T. Paul, Z. physiol. Chem. 31, 1 (1900).

"^ A. L. Bernoulli and A. Loebeastein, Helv. Chim. Acta 23, 245 (1940); apparently these two .separate
dissociations (pKai = 5.78 and pKai = 5.85), involving a total of 2 equivalents of alkali per mole of uric acid,
escape detection in the spectrophotometric analysis and appear as one {pKai = 5.4).

' L. F. Cavalieri, J. J. Fox, A. Stone and N. Chang, J. Am. Chem. Soc. 76. 1119 (1954).

' E. A. Johnson, Biochem. J. 51, 133 (1952).

113

114

AARON BENDICH

(hence only "apparent") since in the spectrophotometric method the ac-
tivity coefficients (which probably do not differ greatly from unity, inas-
much as the concentrations are between 10"^ and 10~^ molar) are not
known and in the titrimetric technique (concentrations from lOr^ to
IQ-i molar) the liquid junction potentials are also not known. Nonethe-
less, the agreement (Table II) by the two methods is good.

The pKai values for the parent compounds pyrimidine and purine are a
measure of the basicity of ring nitrogen and are considerably lower than
those for pyridine (5.23) and imidazole (7.03). (For a discussion of the
basicity of nitrogen-containing heterocycles, see Albert et al.^^^) The pK„,
value of purine (8.9) concerns the removal of a proton from the imidazole
portion of the neutral molecule as does the pK^, (9.80) for adenine. While

the pKa, for adenine (4.15) and guanine (3.3) refer to the amino groups

+

( — NH3 ;:± — NH2 + H+), it is not yet established which group is involved
in the second and third dissociations for guanine.

As for the derivatives containing potentially tautomeric groups, the
question arises whether, for example, it is the carbonyl form (A)

^

N

H
A

O

/-

N

+ H^

A'

or the enol form (B)
C

\

\

O N "~ HO N "~ "O N

+ H+

H

B

which dissociates in aqueous solution; i.e., does enolization precede ioniza-
tion? Since the ionic forms A' and B' are resonance hybrids, they are not
capable of independent existence, and it is not to be expected that ultra-
violet spectroscopy can furnish a direct answer. Although the hypothetical
intermediate or uncharged enol (lactim) form may exist in solution, defini-
tive evidence for such existence has not been demonstrated by ultraviolet
spectroscopy. This might be due to its presence in too small a quantity, or
to the possibility that its spectrum might be the same as that for the anionic

32^ A. Albert, R. Goldacre, and J. Phillips, J. Chem. Soc. 1948, 2240.

CHEMISTRY OF PURINES AND PYRIMIDINES

115

\i

_

<\

— 1 —

r

— T

— 1 — 1

— 1 —

— ^ —

—

12
1 1

\i

f

"^

-

lO

~

\

,^d

-

\\\

^u

9
8

7

l''"\u

/

\

-

FV

\

/I
/ 4.4

-

-

\\ '■

-

V: -1

I ^

6
5

~

V*

/t

'-^'■h

k 1

-

-

\\

\¥

a

"f

\\\

-

.

y

t

\.

/b

\ \

4
3
2

\

v

\ \

' 1/

Ij

■• '

-

^

\

1

V^

..•'.1

13//

'

\ V

-

-

\-^

\'

>/

VW

■\ viA

-

\

\\,

/ ^^

; \

-

\

^

—^

V\ w

—

0

-

1

1

1

1

. ]_ _

1

1

1

1

- 1 "•■

k

210

230

250 270 290

310

W

AVEL

ENGT

H. nr

;j

Fig. 1. Ultraviolet absorption spectra of 5-»i€thylcytosine at pH values (1.0 to 14)
indicated. The curve for pH 2, not shown, is identical with that for pH 1.0; those (not
shown) for pH 7 to 10 are identical. Isoshestic point a is that for pA'ai (= 4.6); b, c,
and dare for pA'a2 (= 12.4). (Adapted from D. Shugar and J. J. Fox, Biochim. ctBtophys.
Acta 9, 199 (1952).)

form. Such a derivative would appear to exhibit the phenomenon of "meso-
meric tautomerism" which "results in a manifestation of dual character by
a compound essentially homogeneous."^-^ The existence of individual tauto-
mers (hence a slow tautomeric change) has been inferred^''* from titrimetric
data for the related 6-hydroxypteridine and xanthopterin (2-amino-4,6-
dihydroxpteridine), but thus far not for pyrimidines and purines.

The principles mentioned above are illustrated for 5-methylcytosine, the
spectral changes of which (cone. ca. 10~'* molar) are shown in Figure P^'
over the pH range 1 to 14. The curve for pH 1.0 is the same as that for
pH 2.0 and shows e = 9,790 at wavelength 283.5 m/x. As the pH is increased,
this absorption decreases (hypochromic effect) and the position of this
maximum shifts towards shorter wavelengths (hypsochromic effect). The
values remain constant from pH 7 through 10 (e = 6,230 at 273.5 mju) and
hyperchromic and bathochromic effects are observed with further increase
in pH until 14 is reached. (The basicity of ring nitrogens, which would be

3" L. Hunter, /. Chem. Soc. 1945, 806.

116 AARON BENDICH

manifested at values below pH 1, has not been measured.) The curves for
the cationic form, at pH 1.0, and for the neutral or nondissociated form,
at pH 7.2, cross at point a, indicating that at this wavelength (266 mju)
the two forms have the same extinction coefficient. This point is common
to the curves at all pH values from 1.0 to 7.2, and is called an isosbestic point
(point of equal extinction). ^^^'^27 Isosbestic point a delineates the equilibrium
(pK'ai) between the cationic and neutral forms; isosbestic points 6, c, and
d (pH 7 to 14) are concerned with the equilibrium between the neutral and
anionic forms. Since the neutral form is conmion to both equilibria, its
curve should (and does) pass through all the isosbestic points. The sharp-
ness of the isosbestic points is one measure of the purity of the absorbing
substance and of the precision in the recording of the spectra. "If curves at
a sufficient number of pH values are run, the plot of extinction coefficients
vs. pH at a given wavelength would give a titration curve. "^^^ From this
the apparent pKo can be calculated. For homogeneous substances, a pKa
value calculated from the extinctions at one wavelength should be the
same as that calculated from other wavelengths.

From these considerations, it follows that the nature and even the po-
sition (s) of potentially tautomeric groups of an unknown pyrimidine (or
purine) derivative may be diagnosed from a study of its spectral behavior
as a function of pH. The molar concentration need not be known since the
spectral behavior is essentially independent of the concentrations usually
employed in ultraviolet spectrophotometry.

Molecular extinction coefficients and absorption maxima are listed in
Table I for pyrimidines and purines often encountered in nucleic acid
studies. Whatever differences exist between these and other published
values are probably due to small variations in the instruments in use.

2. Chemical Properties

"As the simpler compounds are much less well known than the highly
hydroxylated or amino members, a rather distorted impression of pyrimi-
dine chemistry has grown up, much as if the behavior of benzene were
known only through the reactions of compounds like phloroglucinol."^^*
An analogous statement may be made in regard to purine. Pyrimidine and
purine are probably the most stable members of the series and can survive
treatment with strong oxidizing (but not reducing) agents, concentrated
sulfuric (100°) and nitric acids.^^'^^* The general stabihty of pyrimidine
may be attributed^^^ to the influence of the ring nitrogens which results in

'25 A. Thiel, A. Dassler, and F. Wiilfken, Forischr. Chem. Physik u. physik. Chem. 18,
(3), 79 (1924).

327 W. R. Erode, "Chemical Spectroscopy," 2nd ed. p. 249. Wiley, New York, 1943.

328 B. Lythgoe, Quart. Revs. (London) 3, 181 (1949J.

CHEMISTRY OF PURINES AND PYRIMIDINES 117

an electron deficiency at positions 2, 4, 6, and, to a smaller extent, at the
"meta" or 5-position. Hence, pyrimidine exhibits chemical properties that
are analogous^^^ to those of the more thoroughly studied pyridine system. ^^^
The stability is altered by the introduction of electron-donating groups
such as amino and hydroxyl, and the following sections deal primarily Avith
these derivatives. "Vestiges of the behavior of the parent compound, how-
ever, remain; the 2, 4 and 6 position retain their electrophilic character . . .;
and substitution by electrophilic reagents is still confined to position 5."^^^

a. Stability Towards Acid and Alkali; Transformations; Nitrous Acid

A practical problem concerning the stability of pyrimJdines and purines
is encountered in the acid hj^drolysis of nucleic acids. The liberation of
purines is achieved under much milder conditions than are needed for
pyrimidines (Chapters 5-7, 9-11). The usual procedure, A'' acid at 100°
for 1 hour, leads, essentially, to a quantitative recovery of the nucleic acid
purines as judged by chromatography^^^ and differential spectroscopy .^^^
However, the isotope dilution technique reveals^^^ that this treatment re-
sults in a 7 to 8 % deamination of adenine and guanine. A considerable
hydrolytic conversion of cytosine to uracil occurs during a 90-minute heat-
ing (175°) with 10 % HCl, but uracil and thymine escape destruction under
these circumstances.^^" 5-Methylcytosine is hydrolyzed to thymine by 20 %
sulfuric acid at 150-160°.^^ No detectable destruction of uracil, cytosine,
or thymine is observed upon heating with 98-100% formic acid at 175°
for 30-120 minutes f°'^^° the same is true for adenine and guanine during a
30-minutes treatment.'"^ A quantitative recovery of adenine, guanine, cjrto-
sine, uracil, and thymine is obtained''^ when these compounds are subjected
to the action of 12 A'' perchloric acid at 100° for 1 hour, but 5-hydroxy-
methylcytosine is thus destroyed^^ and a 15 % loss of thymine results when
the temperature is raised to 110°." (See addendum).

When guanine is refluxed for 32 hours vnih. 25% hydrochloric acid, a
50 % yield of xanthine is obtained ;^^^ a similar treatment of isoguanine for
47 hours also affords about a 50% conversion to xanthine.^^^'''^^ When
guanine is heated with 3.4 A^ HCl at 158° for 90 minutes, xanthine (52 %
yield), ammonia, glycine (53% of the theoretical yield), 4-(or 5-)guanido-
imidazole (14%), and a small quantity of glycocyamine are formed. It has
been demonstrated with isotopically labeled guanine that the carboxyl,

329 H. S. Mosher, in "Heterocyclic Compounds" (Elderfield, ed.), Vol. 1, p. 397. Wiley,

New York, 1950.
'30 E. Vischer and E. Chargaff, /. Biol. Chfim. 176, 715, 703 (1948).

331 H. S. Loring, J. L. Fairley, H. W. Bortner, and H. L. Seagran, /. Biol. Chem. 197,
809 (1952).

332 R. Abrams, Arch. Biochem. 30, 44 (1951).

333 E. Fischer, Ber. 43, 805 (1910).

118 AARON BENDICH

methylene, and amino groups of the glycine arise, respectively, from carbons
4 and 5 and N-7 of the guanine.^"^ Upon treatment of guanine with con-
centrated sulfuric acid at "high temperature," carbons 2 + 6 are liberated
as CO2 and carbon 8 as CO.'^^ Both hypoxanthine and adenine yield glycine,
ammonia, carbon dioxide, and carbon monoxide (the latter probably from
the decomposition of formic acid) upon hydrolysis with concentrated HCl
at 180-200°^^* The amino group of the liberated glycine arises from N-7
of the adenine i^"^ the formation of 4-amino-5-imidazolecarboxamidine from
such a hydrolysis was mentioned above. The formic acid resulting from the
hydrolysis of adenine with 30 % sulfuric acid'^^ may be collected by steam
distillation.

Data on the action of alkali on pyrimidines and purines are scanty.
Hypoxanthine is decomposed into ammonia and hydrocyanic acid upon
treatment with fused KOH at 200°;^" adenine and hypoxanthine are, how-
ever, unaffected by boiling in aqueous alkah.^^^-'" (At 200° in water alone,
hypoxanthine breaks down to CO 2 , NH3 , and formic acid.^") Pyrimidines
containing hydroxy or amino groups both at position 4 and 5 are unstable
in alkaline solution at room temperature; the instability is markedly in-
creased when, in addition, such a group occupies position 6.^^*

Reference has already been made to the nitrous acid deamination of
adenine, guanine, cytosine, and 5-hydrox3Tnethylcj4osine to the cor-
responding hydroxy derivatives. Nitrous acid deaminates both 5-methyl-
cytosine and 5-methylisocytosine to thymine.'^" 2,6-Diaminopurine is de-
aminated to isoguanine (cf. DavolP^O but the latter resists the action of
nitrous acid.^''^ 8-Aza-adenine and -guanine are converted by nitrous acid
to the corresponding hypoxanthine and xanthine analogues.^*^ The amino
group at C2, 4, and 6 of the pyrimidine moieties of these various compounds
therefore do not exhibit the properties expected of typical aromatic amino
derivatives, and accordingly are analogous in chemical behavior to the a-
and 7-aminopyridines.^^^ On the other hand, a behavior approximating that
of aromatic amino compounds is shown by 5-aminopyrimidines (cf. Whit-
taker,^^^ Lythgoe,^^^ and Rose^^^).

By the action of nitrous acid, a nitroso group can be introduced into the
5-position of certain pyrimidines. This reaction forms the basis of the
valuable Traube synthesis of purines (see below). Only those pyrimidines
can be nitrosated which contain amino or hydroxy groups in both positions

334 W. H. Marsh, /. Biol. Chem. 190, 633 (1951).

"5 M. Kriiger, Z. physiol. Chem. 16, 160 (1892); 18, 351, 423 (1894).

336 C. D. Stevens, /. Biol. Chem. 120, 751 (1937).

'" <''' A. Kossel, Z. physiol. Chem. 6, 422 (1882).

(b) 12, 241 (1888).
338 F. L. Rose, /. Chem. Soc. 1952, 3448.

CHEMISTRY OF PURINES AND PYRIMIDINES 119

4 and 6;^^^ this structural requirement is essentially independent of the
nature of the substituent at position 2.

When uric acid is heated for 30 hours with acetic anhydride and pyiidine,
8-methylxanthine is formed by a ring closure of the intermediate 4 , 5-diace-
tylaminouracil.^^" This result was confirmed by Biltz and Schmidt.^'*' Carbon
dioxide and acetic acid result from the ring closure. Bredereck, et al.^*'^
have recently studied this reaction in detail. Uric acid is converted, upon
reaction with formic acid at 220-230°, into xanthine and carbon dioxide.^^'^

h. Actian of Reducing Agents; Polarographic Behavior

Uracil is reduced to 4,5-dihydrouracil at 75° under 2 atmospheres of
hydrogen in the presence of a platinum catalyst.^^' Under identical condi-
tions, cytosine is converted to the same product, ^vith the liberation of
ammonia.^^^ The resulting dihydrouracil is much less stable to hot acid (as
well as alkah at room temperature^^^) and is thereby converted to /3-ala-
nine.^*^ These reactions have been exploited in a degradation procedure^^^
in which each carbon atom of isotopically labeled uracil can be directly
analyzed for its isotope content.

Whereas adenine and hypoxanthine are unaffected, aqueous solutions
of purine hydrochloride and 2-hydroxypurine each absorb one mole of
hydrogen at one atmosphere at room temperature under the influence of a
palladium-charcoal catalyst.^^^ The resulting dihydropurine derivatives are
quite unstable in the presence of dilute mineral acid giving rise to substances
containing a diazotizable amino group.

Agents such as sodium amalgam, sodium and ethanol, etc., reduce a
number of pyrimidines and effect a rupture of the ring system (cf. Johnson'*
for details). Electrolytic reduction (lead cathode, 7-9°) of 4-methyluracil
in 50% sulfuric acid gives 2-hydroxy-4-methyltetrahydropyrimidine and
1 ,3-diaminobutane.^''^ Purone (4,5-dihydro-6-deoxyuric acid) and tetra-
hydrouric acid result from a similar reduction of uric acid.^^*

Adenine, adenosine, and adenylic acid in 0.1 A'' perchloric acid are re-

339 B. Lythgoe, A. R. Todd, and A. Topham, /. Chern. Soc. 1944, 315.

3" H. Bredereck, I. Hennig, and W. Pfleiderer, Cheni. Ber. 86, 321, 333 (1953) : these in-
vestigators attribute the original work to C. F. Boehringer and sons, German Pats.
121,224 and 126,797 (1901).

3" H. Biltz and W. Schmidt, Ann. 431, 70 (1923).

3« H. Biltz and A. Beck, /. -prakt. Chem. [2] 118, 166 (1928).

3" E. B. Brown and T. B. Johnson, J. Am. Chem. Soc. 45, 2702 (1923).

3" E. B. Brown and T. B. Johnson, /. Am. Chem. Soc. 46, 702 (1924).

3^^V. Lagerkvist, Acta Chem. Scand. 7, 114 (1953).

3" A. Bendich, P. J. Russell, Jr., and J. J. Fox, /. Am. Chem. Soc. 76, in press (1954).

'" J. Tafel and A. Weinschenk, Ber. 33, 3378 (1900).

3« J. Tafel, Ber. 34, 258, 1181 (1901).

120 AARON BENDICH

ducible at the dropping mercury cathode, whereas guanine, guanosine,
guanyhc acid, cytidine, cytidylic acid, and uracil are not.^^^ This was pro-
posed as a method for the quantitative estimation of as httle as one micro-
gram of adenine (error ± 2 %) in a hydrolysate of PNA or DNA. (For
theoretical and practical aspects of polarography, see Kolthoff and Lin-
gane.^^") Pyrimidine, 2-amino-, 6-amino-, 6-hydroxy-, 2,6-diamino-, and
4,6-diaminopyrimidine are reduced polarographically whereas derivatives
such as thymine, isocytosine, and barbituric acid are not.^^^ From their
polarographic behavior, the reducible group of pyrimidines was found to

involve the • — C^C — C=N — system.^^^

KosseP^'"^ found that adenine and hypoxanthine (but not guanine or
caffeine) were decomposed upon heating with zinc and HCl, and that the
products (the structures of which were not elucidated) turned a ruby-red
when made alkahne. This reductive degradation of adenine (and hypoxan-
thine) leads to the formation of a diazotizable amine which couples with
A''-(l-naphthyl)ethylenediamine to form a red-colored dye^^^-^^^ and is em-
ployed for a colorimetric determination of adenine.^^^ Folic acid and ATP
are reduced by zinc and HCl and produce a red color, but guanine, cytosine,
isocytosine, thymine, and uracil do not.^^^ Hypoxanthine, isoguanine, and
xanthine, but not 2 , 6-diaminopurine or uric -acid, are also reduced by zinc
and HCl at 100°,^^* and it is quite hkely that the resulting diazotizable
amines are 5-substituted-4-aminoimidazoles.

c. Action of Oxidizing Agents and Ultraviolet Light

As pointed out above, thymine is oxidized by permanganate to urea,
and cytosine to biuret. The KMn04 oxidation to urea^^^-^^^ and to oxaluric
acid^^^ has been utilized for the analysis of the isotope content of specific
carbon atoms of labeled uracil (see Fairley et aZ.^"). Uracil is also oxidized
by ozone in glacial acetic acid, and the products which are obtained are
formylglyoxyiurea, oxaluric acid, urea, and oxalic and formic acids.^^^ Hy-
drogen peroxide in the presence of charcoal brings about the oxidation of

3" J. C. Heath, Nature 158, 23 (1946).

^5" I. M. Kolthoff and J. J. Lingane, "Polarography," 2nd ed., Vols. 1 and 2. Intersci-

ence, New York, 1952.
^51 L. F. Cavalieri and B. A. Lowy, Arch. Biochem. and Biophys. 35, 83 (1952).

352 A. J. Glazko and L. M. Wolf, Arch. Biochem. 21, 241 (1949).

353 D. L. Woodhouse, Arch. Biochem. 25, 347 (1950).

354 H. G. Koritz and F. Skoog, Arch. Biochem. and Biophys. 38, 15 (1952).

355 S. Friedman and J. S. Gets, Arch. Biochem. and Biophys. 39, 254 (1952).
355 M. R. Heinrich and D. W. Wilson, J. Biol. Chem. 186, 447 (1950).

357 J. L. Fairley, L. L. Daus, and B. Krueckel, /. Am. Chem. Soc. 75, 3842 (1953).

358 T. B. Johnson and R. B. Flint, /. Am. Chem. Soc. 53, 1077 (1931).

CHEMISTRY OF PURINES AND PYRIMIDINES 121

uracil to isobarbitiiric, isodialuric, and oxalic acids and urea.^^^ When ex-
posed to a solution of ferrous sulfate and sodium bicarbonate in the presence
of oxygen, thymine is oxidized and yields urea, acetol, and formic and
pyruvic acids?^" The acetol condenses with o-aminobenzaldehyde, yielding
the strongly fluorescent 3-oxyquinaldine which turns a deep red upon the
addition of ferric chloride, thus furnishing a sensitive color test for thymine.
The same oxidation mixture converts cytosine and uracil into dihydroiso-
barbituric acid, and this compound forms a red complex with ferrous sul-
fate.^" Thymine is also oxidized bj'- peroxide in the presence or absence of
ferrous sulfate and acetol is obtained upon heating the reaction mixture.^*^^

The mechanism of this oxidation was investigated by Baudisch and
Davidson^^^ with the possibility in mind that the intermediate oxidation
product might be thymine glycol (XXXV). Thymine was converted quan-
titatively ,^° upon treatment with bromine water, into 5-bromo-4-hydroxy-
hydrothymine (XXXVI), and this in turn was transformed^*^^ into thymine
glycol by shaking with moist silver oxide. Acetol (arbitrarily shown in the
keto form, XXXVII) and urea were obtained by boiling the glycol in either
sodium bicarbonate or barium hydroxide solution and it was concluded^^^
that thymine glycol was indeed the intermediate in the peroxide oxidation of
thymine. (Thymine can be regenerated from its glycol by hydriodic acid
reduction.) Advantage has been taken of this elegant degradation proce-
dure to locate the isotope (C^*) in the methyl group of biologically labeled
thymine ;'^^ the resulting acetol was converted by hypoiodite into iodoform.
Acetol is also obtained when 5-methylcytosine is subject ed^^^ to the Baudisch-
Davidson procedure.

Uracil, cytosine, and thjonine absorb iodine in aqueous bicarbonate solu-
tion; urea is formed when the reaction mixtures are heated, and, in the
case of thymine, acetol is also formed.^^^ Uracil, xanthine, and guanine
each absorb four equivalents of iodine from an alkaline solution,^" ■'®^
whereas uric acid absorbs only one mole^^'^ and adenine does not react;
although the chemistry of these reactions has not yet been clarified, they
serve as analytical procedures. [Cf. Dische, Chapter 9.]

2*^ C. R. Schwob and L. R. Cerecedo, Proc. Am. Soc. Biol. Chemists, J. Biol. Chem. 105,
Ixxvi (1934).

360 T. B. Johnson and O. Baudisch, J. Am. Chem. Soc. 43, 2670 (1921).

361 O. Baudisch, ./. Biol. Chem. 60, 155 (1924).

3«2 O. Baudisch and L. W. Bass, ./. Am.. Chem. Soc. 46, 184 (1024).

363 o. Baudisch and D. Davidson, /. Biol. Chem. 64, 233 (1925).

364 D. Elwyn and D. B. Sprinson, J. Am. Chem. Soc. 72, 3317 (1950).

365 H. H. Harkins and T. B. Johnson, J. Am. Chem. Soc. 51, 1237 (1929).
3.66 L. W. Bass and O. Baudisch, J. Am. Chem. Soc. 46, 181 (1924).

367 M. Z. Grynberg, Biochem. Z. 253, 143 (1932).

368 W. Klein, Z. physiol. Chem. 231, 125 (1935).

122

AARON BENDICH

III

OH

CH,

N

HO

/^N

H2O

Brz

OH

N

OH

CH3

/_Br Ag20

/

z

CH3

N

Thymine

^OH
HO H

XXXVl

5 -Bromo -4 -hydroxy -
hydrothymine

OH
OH

HO H

XXXV

Thymine glycol

H2O

NaHC03

or
Ba(OH)j

CH3

» NaOI 1

CHI3 ^ c=o *

!
CH2OH

XXXVII

Iodoform Acetol

CO2

COOH

4-

NH2

I
C

/- \
O NH2

Urea

The chloric acid oxidation of guanine'^ to guanidine (and parabanic acid),
described above, has been utilized in isotope studies to isolate and analyze
carbon ^2.^^' •^^^•^'''' Guanidine (from C-2) is also derived from an acid-
permanganate oxidation of guanine f^^ urea (mainly from C-8) and carbon
dioxide are formed as well (cf. Edmonds et aZ."^). The acid-permanganate
oxidation of triply labeled adenine produces urea which arises from the
1 ,7- and 3,9-nitrogen atoms.*"^

The oxidative degradations of uric acid to alloxan, murexide, allantoin,
etc., are well known and are ably discussed in detail elsewhere.'^ •^^•^*^ These
degradations have been applied to biological studies involving isotopes"^""®
and to isotopically labeled uric acid."^ Uric acid in alkaline solution is oxi-

3«9 A. A. Plentl and R. Schoenheimer, J. Biol. Chem. 153, 203 (1944).

370 G. B. Brown, P. M. Roll, A. A. Plentl, and L. F. Cavalieri, /. Biol. Chem. 172, 469
(1948).

371 M. Edmonds, A. M. Delluva, and D. W. Wilson, J. Biol. Chem. 197, 251 (1952).
3" J. M. Buchanan, J. C. Sonne, and A. M. Delluva, /. Biol. Chem. 173, 81 (1948).

373 J. M. Buchanan, /. Cellular Comp. Physiol. 38, Suppl. 1, 143 (1951).

374 D. Shemin and D. Rittenberg, J. Biol. Chem. 167, 875 (1947).

375 J. L. Karlsson and H. A. Barker, /. Biol. Chem. 177, 597 (1949).

376 D. Elwyn and D. B. Sprinson, J. Biol. Chem. 184, 465 (1950).

377 L. F. Cavalieri and G. B. Brown, J. Am. Chem. Soc. 70, 1242 (1948).

CHEMISTRY OF PURINES AND PYRIMIDINES 123

dized by oxygen in the presence of catalytic amounts of copper ;^^* the
spectral changes accompanying this oxidation are identicaP^^ with those
that occur when uric acid is oxidized by uricase (cf. Bentley and Neu-
berger."^).

In addition to the alterations brought about by the oxidizing agents
discussed above, certain pyrimidines and purines undergo profound, but
poorly understood, changes when their solutions are exposed to ultraviolet
light. For example, thymine,^^*^ uracil,^^^ isoguanine,^*^ and adenylic and
uric acids^*^ show extensive loss of their ultraviolet absorption spectra
after such exposure. In the case of thymine,^^" the pyrimidine ring is dis-
rupted with the formation of urea and pyruvic acid. When uracil and uri-
dine solutions are irradiated in the 230-280 m/x region, the loss in absorption
is largely restored by acidification of the irradiated solutions.^^^

d. Color Tests

Although they have been largely superseded by the modern and more
specific chromatographic, spectrophotometric, and polarographic tech-
niques, color reactions are often employed to follow the course of a chemical
reaction, the effectiveness of a fractionation or isolation procedure, or for
the identification and quantitative estimation of pyrimidines and purines.
Reference has already been made in the previous three sections to a number
of color tests. The Wheeler-Johnson test^^'' for uracil and cytosine depends
upon the conversion of these compounds to 5,5-dibromo-4-hydroxyhy-
drouracil, the bromine atoms of which are replaced by treatment with
barium hydroxide solution to give isodialuric and dialuric (2,4,5,6-tetra-
hydroxypyi'imidine) acids. The barium salts of these acids are purple. The
test is not given by thymine, 5-methylc5rtosine, or pyrimidine nucleosides.^^
In an adaptation of this reaction, ^^^ uracil, cytosine, isocytosine, 5-bromo-
cytosine, and thiouracil are also converted with bromine water to the di-
bromohydroxyhydrouracil ; the latter reduces the uric acid reagent, lithium
arsenotungstate, to produce a blue color which is measured quantitatively.
Uracil, cytosine, and thymine, but not N-3-substituted pyrimidines, couple
with the Pauly reagent, diazobenzenesulfonic acid, to give a red diazo
dye following alkalinization''*'^*^'^^^ (cf. Lythgoe et alP^). This reaction
serves as the basis of a microestimation technique for thymine.'^"^^"

378 M. Griffiths, /. Biol. Chem. 197, 399 (1952).

"9 R. Bentley and A. Neuberger, Biochem. J. 52, 694 (1952).

380 L. W. Bass, /. Am. Chem. Soc. 46, 190 (1924).

3" M. M. Stimson, J. Am. Chem. Soc. 64, 1604 (1942).

382 D. Rapport and A. Canzanelli, Science 112, 469 (1950).

383 R. L. Sinsheimer and R. Hastings, Science 110, 525 (1949).

384 H. L. wAeeler and T. B. Johnson, J. Biol. Chem. 3, 183 (1907).

385 M. Soodak, A. Pircio, and L. R. Cerecedo, /. Biol. Chem. 181, 713 (1949).

386 T. B. Johnson and S. H. Clapp, J. Biol. Chem. 5, 163 (1908).

124 AARON BENDICH

Guanine and xanthine, but not adenine or hypoxanthine, react with the
Folin phenol reagent to give a blue color in an analytical procedure.^^^

e. Salt Formation

The purines and pyrimidines form a variety of salts with many acids
and metal ions, a detailed hst of which has been compiled by Levene and
Bass.^^ Many investigators have, since 1858,^^ employed anmioniacal silver
nitrate to precipitate, isolate, and estimiate purines as sparingly soluble
silver salts.^^-*^^'*^^ Silver salts of the purines may also be formed quanti-
tatively in dilute sulfuric acid solution^^^ ■^'^^ and the purines are easily re-
generated upon treatment with hydrochloric acid. The pyrimidines of
nucleic acid origin do not precipitate as silver salts under acid conditions,
but do so in alkali.^' Poorly soluble cuprous-purine complexes are formed
when purines are boiled in the presence of cupric sulfate and sodium bi-
sulfite^^^ '^^^^ or cuprous oxide.^^^ The purines can be recovered following
reaction with hydrogen sulfide. The common purines and pyrimidines form
relatively insoluble complexes with mercury^^" when treated with mercuric
acetate solution at pH 6.2. Silver salts and chloromercuric derivatives of
the purines^*^ '^^^ condense with acetohalo-sugars to form purine nucleo-
sides.^^*-^^^

The picrates of many purines and pyrimidines are easily crystallizable,
well-defined salts and are valuable for purposes of isolation, characteriza-
tion, and quantitative estimation. Properties of certain picrates are given
in Table II. Picrate formation is the basis of a quantitative procedure for
adenine and guanine^^' ■^^^'^ The crystalline argentipicrate is used in a de-
termination of hypoxanthine.^"" The presence of a high -intensity band for
picric acid in the spectral region 350-400 mju and the absence of signficant
light absorption for many purines and pyrimidines in this region lend them-
selves to a simple micromethod for the determination of the molecular

3" T. B. Johnson and J. H. Derby, Am. Chem. /. 40, 444 (1908).

388 G. Hunter, Biochem. J. 30, 745 (1936).

389 D. L. Woodhouse, Biochem. J. 44, 185 (1949).

390 E. D. Day and W. A. Mosher, J. Biol. Chem. 197, 227 (1952).

391 G. H. Hitchings, J. Biol. Chem. 139, 843 (1941).

392 Z. Neubauer, Z. anal. Chem. 6, 33 (1867).

393 G. Bruhns, Ber. 23, 225 (1890); Z. -physiol. Chem. 14, 533 (1890).

394 E. Salkowski, Arch. ges. Physiol. 69, 268 (1897-98).

396 S. E. Kerr, K. Seraidarian, and M. Wargon, J. Biol. Chem. 181, 761 (1949).

396 G. H. Hitchings and C. H. Fiske, /. Biol. Chem. (a) 140, 491 (1941); (b) 141, 827
(1941).

397 S. Graff and A. Maculla, /. Biol. Chem. 110, 71 (1935).

398 E. Fischer and B. Helferich, Ber. 47, 210 (1914).

399 J. Davoll and B. A. Lowy. J. Am. Cheyn. Soc. 73, 1650 (1951).
"« G. H. Hitchings, /. Biol. Chem. 143, 43 (1942).

CHEMISTRY OF PURINES AND PYRIMIDINES 125

weights of picrates. This method'*°^ was used to establish the molecular
weight of the aglycone (adenine) of cordycepin.^^^'^^^ Purines and pyrimi-
dines can be recovered from their picrates by extracting hj^drochloric acid
solutions of these salts with benzene or ether,"" -^"^ by the use of anion-ex-
change resins (cf . Davoll and Lowy^^^) or by saturating dry ether or acetone
suspensions with HCl gas; in the last procedure, the purines and pyrimidines
are often obtained as the crystalline hydrochlorides.

m. Synthetic Methods

It is a tribute to the ingenuity of the earlier workers that many of the
recent syntheses of pyrimidines and purines employing isotopes have been
fashioned from methods (see General References) devised by the earlier
workers. There have been a few newer developments, and emphasis shall
be given to these in the following sections.

1. General Methods for Pyrimidines and for the Introduction of

Isotopes

Since they are cyclic amidines, pyrimidines have been synthesized by
condensing amidines, or substituted amidines such as ureas, thioureas, and
guanidines, with the appropriate compound containing at least three
carbon atoms in a chain. The latter include esters of malonic, acetoacetic,
cyanoacetic, and their substituted acids, or the free acids, /3-diketones,
a, /3-unsaturated esters, malononitrile, 1 ,3-dialdehydes, etc. Examples of
this type of synthesis were given for the preparation of thymine, uracil,
and orotic acid and have been adapted for the introduction of isotopes. ^^^ '^^g
Depending upon the choice of reactants, pyrimidines with varying degrees
of substitution may be obtained. For instance, guanidine and nitromalonal-
dehyde condense in mildly alkaline solution to give 2-amino-5-nitropyrimi-
dine which is hydrolyzed to 2-hydroxy-5-nitropyrimidine on boiling with
aqueous ammonia; the latter compound is prepared also from urea and
nitromalonaldehyde.^°^ 2-AminopyTimidine is formed by reacting guanidine
and iS-ethoxyacrolein acetal in ethanolic HCl.'*"^ Convenient sjTitheses of
uracil and isocytosine are afforded by the condensation, respectively, of
urea and guanidine with formylacetic acid (formed in situ from malic
acid) in fuming sulfuric acid.^"^-^'*® Similarly, thymine and 5-methyIisocy-
tosine are derived from ^S-methyhnalic acid.^°''

"1 K. G. Cunningham, W. Dawson, and F. S. Spring, J. Chem. Soc. 1951, 2305.

"2 J. K. Parnas, Biochem. Z. 206, 16 (1929).

«3 W. J. Hale and H. C. Brill, J. Am. Chem. Soc. 34, 82 (1912).

*o* R. W. Price and A. Moos,"/. Am. Chem. Soc. 67, 207 (1945).

"5 D. Davidson and O. Baudisch, J. Am. Chem. Soc. 48, 2379 (1926).

"8 W. T. Caldwell and H. B. Kime, /. Am. Chem. Soc. 62, 2365 (1940).

"7 H. W. Scherp, J. Am. Chem. Soc. 68, 912 (1946).

126 AARON BENDICH

An interesting synthesis of thymine, shown below, was developed by
Bergmann and Johnson /°* Urea and methylcyanoacetic acid are heated
with acetic anhydride to give methylcyanoacetylurea (XXXVIII), which,
as expected, rearranges under the influence of aqueous alkali to yield 4-
aminothymine. However, XXXVIII absorbs one mole of hydrogen in the
presence of a platinum catalyst to give anunonia and thymine:

O

OH II OH

CH3 H C CH3 I CH3

\ / \ /

N CH "= . N

N

NaOH

Pt

+ NH3

/\N/\ C CN /^N'

HO NH2 /' \ HO

O NH2

XXXVIII
4-Aminothymine Methylcyanoacetylurea Thymine

This reaction is based upon a similar conversion of cyanoacetylurea to
uracil and ammonia^'^^ under reducing conditions (H2 and a nickel catalyst) .
It was believed^"^ that the cyano group was first reduced to an imino group,
which in turn was hydrolyzed to ammonia and an aldehydo group prior to
cyclization. Although this explanation is plausible, it does not receive sup-
port from the catalytic reduction of 4-aminouracil to uracil.^^° Cyanoacetyl-
urea is prepared by the action of phosphorus oxychloride upon urea and
cyanoacetic acid'^ and undergoes a base-catalyzed ring closure to 4-amino-
uracil. The latter is also made by refluxing urea and cyanoacetic ester in
ethanol containing sodium ethoxide/"

Urea (N'^) and cyanoacetal [(C2H60)2CHCH2CN] condense in boihng
butanol in the presence of sodium butoxide to give an intermediate ureide
which cyclizes to 1,3-labeled cytosine upon acidification. *^2 j^ a similar
manner, guanidine (N'^) affords isotopically labeled 2,4(or 6)-diamino-
pyrimidine.'*^' Other examples of this type of synthesis are given below in
the section on purines.

a. Transformations; the Value of Halogen and Mercapto Derivatives

Because certain pyrimidines are difficult to prepare by more direct means,
a number of transformations of more easily accessible compounds have

"8 W. Bergmann and T. B. Johnson, J. Am. Chem. Soc. 55, 1733 (1933).

"9 H. Rupe, A. Metzger, and H. Vogler, Helv. Chim. Acta 8, 848 (1925).

*!" J. C. Ambelang and T. B. Johnson, /. Am. Chem. Soc. 63, 1934 (1941).

«i M. Conrad, Ann. 340, 310 (1905).

*i2 A. Bendich, H. Getler, and G. B. Brown, /. Biol. Chem. 177, 565 (1949).

^'3 A. Bendich, W. D. Geren, and G. B. Brown, /. Biol. Chem. 185, 435 (1950).

CHEMISTRY OF PURINES AND PYRIMIDINES 127

been developed, and some of these have been mentioned above. A wide
variety of 2,4- and 6-hydroxypyrimidines including uracil, 5-nitro-, and
4-methyluracil, barbituric acid, 6-amino-4-hydroxy-2-methylthiopyrimi-
dine, etc. are converted by means of POCI3 into the corresponding chloro
derivatives. (A considerable improvement in the chlorination procedure
involves the addition of dimethylaniline to the reaction mixtures.^^^'^^^'^''^-^'*)
Any or all of the chloro atoms of these compounds are often readily replaced
by hydrogen or by amino, mercapto, alkoxy groups, etc. (cf. Johnson and
Hahn'^). For example, 2,6(or 4)-dichloropyrimidine and benzenesulfon-
hydrazide condense smoothly and the resulting hydrazino derivative is
converted to pyrimidine upon alkaline hydrolysis.'*'^ Dechlorination may
also be effected by catalytic hydrogenation, or reduction with phosphonium
iodide or zinc and water, thus making available many new pyrimidines.'^"'-

286,312,317,415

Amination of chloropyrimidines usually proceeds smoothly, and, in the
cases of 5-nitro-2,6-dichloropyrimidine and its 4-methyl homologue, re-
placement of only the 6-chloro atom by an amino group may be
effected.i'^'286 However, amination of other 2 ,6(or 4)-dichloro- or -diethoxy-
pyrimidines leads to mixtures of monoamino isomers'^ and a method utiliz-
ing mercapto derivatives has been developed which obviates this difficulty.

The mercapto derivatives have been prepared by total synthesis involving
thiourea or substituted thioureas in the type of condensation discussed
above, and through the replacement of chloro atoms by alkali hydrosul-
fides.'^'*'''^^* A new method, bypassing the halogenated pyrimidines, de-
pends upon the direct thiation of uracils by heating with phosphorus penta-
sulfide in an inert medium such as xylene or tetrahn.^'^'^'* The resulting
2,6-pyrimidinedithiols usually lead to 6-amino-2-pyrimidinethiols exclu-
sively upon reaction with ammonia or amines.^" Thus, the dithiol com-
pounds derived from uracil and thjmine give excellent yields of 2-cytosine-
thiol and 5-methyl-2-cytosine-thiol, respectively.^'^ •'*'* These compounds,
in turn, are converted to cytosine and 5-methylcytosine^'^ '^'^ by an appli-
cation of the Wheeler-Liddle desulfurization technique. ^'^^ The reactions are
illustrated :

«* J. Baddiley and A. Topham, J. Chem. Soc. 1944, 678.

«5 M. P. V. Boarland, J. F. W. McOmie, and R. N. Timms. J. Chem. Soc. 1952, 4691.

«« G. B. Elion and G. H. Hitchings, J. Am. Chem. Soc. 69, 2138 (1947).

*" P. B. Russell, G. B. Elion, E. A. Falco, and G. H. Hitchings, /. Am. Chem. Soc. 71,

2279 (1949).
«8 D. J. Brown, J. Soc. Chem. Ind. (London) 69, 353 (1950).
"9 G. H. Hitchings, G. B. Elion, E. A. Falco, and P. B. Russell, /. Biol. Chem. 177,

357 (1949).

128

AARON BENDICH

OH

SH

NH,

P2S5

N

HO
Uracil

boiling
xylene

N

HS
Dithiouracil

NH3H2O
100°

> N

HS

2-Cytosinethiol

NH,

HOOCCHia
H2O 100°

NHo

N

Cone.

HCl

/

HOOCCH2S

2-Carbox3'methylthiocytosine

N

I

/\N/
HO

Cytosine

The carboxymethylthiopyrimidines easily form from chloroacetic acid and
mercaptopyrimidines and exhibit an extreme range of stabiUty towards
g^(.j^ 182 ,245 ,420 Coiwenient syntheses of 2-C'Mabeled uracil and thymine^'
exploit the ease with which the Wheeler-Liddle desulfurization proceeds
with the 2-thio derivatives.

An unwanted mercapto group may be replaced by a hydrogen atom by
treatment with Raney nickel. 4,5-Diamino-6-hydroxypyrimidine was ob-
tained from its 2-mercapto derivative in the first application of this tech-
nique to pyrimidines.-*^^ Other examples of compounds prepared by
reductive desulfurization with Raney nickel are 4 (or 6) -amino- and 4 (or
6)-hj^droxypyrimidine/22 '^'^ 4 ,5-diaminopyTimidine,'*^* 4-hydroxy-6-methyl-
pyrimidine,^^^ and pyrimidine.^'^

Another method for the sjmthesis of mercapto derivatives proceeds from
the interaction of a halo derivative with thiourea in boiling ethanol solution,
2-Mercapto-, 4-mercapto-, and dithiouracil are obtained directly from the
corresponding chloropyrimidines without isolation of the intermediate
thiouronium salts.^-^ The thiouronium salt from 2-chloro-4,6-dimethyl-
pyrimidine and thiourea can be isolated, and this yields the 2-mercapto
derivative upon alkaline hydrolysis. Thiourea transforms 2 , 5-dichloro- into
2-mercapto-5-chloro-pyrimidine.*^* This reaction is an example of the rela-
tive stability of halogen atoms at C-5 of pyrimidines."'^^*'^-^

«o G. H. Hitchings and P. B. Russell, /. Chem. Soc. 1949, 2454.

«i L. L. Bennett, J. Am. Chem. Soc. 74, 2432 (1952).

«2 L. F. Cavalieri and A. Bendich, /. Am. Chan. Soc. 72, 2587 (1950).

^" D. J. Brown, /. Appl. Chem. (London) 2, 239 (1952).

«4 J. F. W. McOmie and M. P. V. Boarland, Chemistry & Industry 1950, 602; /. Chem.

Soc. 1951, 1218; 1952,3722.
*" M. Yanai, /. Pharm. Soc. Japan 62, 95 (1942).

CHEMISTRY OF PURINES AND PYRIMIDINES 129

h. Newer Methods

A convenient method for the preparation of a variety of pyrimidines has
been developed by Whitehead.^^^ Urea, ethjd orthoformate, and ethyl
cyanoacetate condense when heated together to give ethyl ureidomethyl-
enecyanoacetate (XXXIX), in 69 % yield, and this in turn, by refluxing in
ethanol-sodium ethoxide, isomerizes quantitatively to 5-carbethoxycytosine
(XL). (Cf. Wheeler and Johns^-^ for a synthesis of XL by an indirect
method.) If ethyl malonate is used instead, diethyl ureidomethylenemalo-
nate (XLI) is formed; it can also be obtained by permitting urea and
ethoxymethylenemalonic ester to react at room temperature in ethanol-
sodium, ethoxide, luit the \deld is smaller. ^-^ XLI cyclizes to 5-carbethoxy-
uracil (XLII).

CN

H-CH-COOEt
+
NHj OEt CN

C + CH

NHo
NH."^CHCOOEt ''h?ft' N^^^°^'

XJ

//\ /\ (69%) 0 = r PH '^^%)

0 NHj EtO OEt ^ ^N^ "^ '"^

Urea Ethyl cyanoacetate + XXXIX XL

Ethyl orthoformate Ethyl ureido- S-Carbethoxycytosine

methylenecjanoacetate

COOEt
I
H-CH-COOEt
+
NHj OEt

I I

C + CH

// \ / \

0 NH2 EtO OEt

Ethyl malonate +
Ethyl orthoformate

OH

(40%) \^ COOEt

NH2 "^CHCOOEt "hTat" N

NaOEt, L ^COOEt

72%)

XJ

0-C^ ^CH ' ° HO N

N

XLI XLII

Diethyl ureido- 5-Carbethoxyuracil

methvlenemalonate

NaOEt /rooni temp.
(22%)

COOEt

NHj ^C-COOEt

I II

C + CH
//\ I

0 NH2 OEt

Ethoxymethylene-
malonic ester

130 AARON BENDICH

A^-substituted ureas and other active methylene compounds (malononitrile,
acetoacetate, etc.) may also be used in this method. The application of the
newer reducing agents, lithium aluminium and boron hydrides, to the 5-
carbethoxy or 5-carboxy compounds might prove useful in the preparation
of 5-hydroxymethylpyrimidines (cf. Wyatt and Cohen^^'^"). In connection
with the free acid forms of XL and XLII, they are both transformed into
uraciP^® •■*27 by heating with 20% sulfuric acid at 160-170°, yet uracil-4-
carboxylic acid (orotic acid) is not decarboxylated at 200°.^"

A new reaction of acetylene with nitriles leads to the synthesis of 2,4-
disubstituted pyrimidines.^^* For example, 2,4-dimethylpyrimidine results
when acetonitrile, potassium, and acetylene are autoclaved at 175-200°.

2. General Methods for Purines and for the Introduction of

Isotopes

a. From Purines

The syntheses and transformations in the purine series developed by
Fischer,^ some of which were discussed above, were based mainly upon
2,6,8-trichloropurine. The preparation of this valuable intermediate has
been improved considerably by using dimethylaniline in the reaction be-
tween uric acid and POCI3 .''^^ When a limited amount of triethylamine is
used in place of the dimethylaniline, 8-chloroxanthine is obtained from
monopotassium urate, and with an excess of triethylamine, 2,8-dichloro-6-
diethylaminopurine is formed.^^" The latter reaction may be of a general
nature since the hydroxyls of hypoxanthine and xanthine are all replaced
by diethylamino groups upon refluxing with triethylamine and POCI3 .
The resulting 6-diethylaminopurine is also prepared from 6-methylmer-
captopurine by heating with aqueous diethylamine in a sealed tube.'^°

Other 6-aminopurines are formed smoothly upon reacting 6-methylmer-
captopurine with the appropriate amines, but, unlike the mercaptopyrimi-
dines, 6-mercaptopurine does not react satisfactorily. As with pyrimi-
dines,^'^'^'* 6-mercaptopurine is prepared by heating hypoxanthine with
P2S5 in tetralin.^^" Another synthesis that has a counterpart in the pyrimi-
dine series'*'^ is the direct formation of 6-mercaptopurine from 6-chloropu-
rine by boiling with alcoholic thiourea.^^® 6-Chloropurine (formed by react-
ing hypoxanthine with POCI3 in the presence of dimethylaniline) and
6-mercaptopurine are reduced to purine, respectively, upon catalytic hy-

«6 C. W. Whitehead, /. Am. Chem. Soc. 74, 4267 (1952) ; 75, 671 (1953).

«7 H. L. Wheeler and C. O. Johns, A7n. Chem. J. 38, 594 (1907).

«8 T. L. Cairns, J. C. Sauer, and W. K. Wilkinson, /. Am. Chem. Soc. 74, 3989 (1952).

«9 J. Davoll and B. A. Lowy, J. Am. Chem. Soc. 73, 2936 (1951).

«« R. K. Robins and B. E. Christensen, /. Am. Chem. Soc. 74, 3624 (1952).

CHEMISTRY OF PURINES AND PYRIMIDINES 131

drogenation and Raney nickel desulfurization.^''^ The mercapto group of 2-
mercaptohypoxanthine was replaced by hydrogen via nitric acid oxidation
in an early synthesis of hypoxanthine.^" Syntheses of adenine may be ef-
fected by the nitric acid-hydrogen peroxide oxidation of 2-mercaptoade-
nine^° or by desulfurization with Raney nickel. ^^^

Other transformations have been utilized in partial syntheses of iso-
topically labeled purines. 1,3-Labeled xanthine^^^ '^^^ and hypoxanthine''^^
are obtained from the nitrous acid deamination, respectively, of N'*-
guanine and adenine. Stably-bound deuterium or tritium atoms are intro-
duced into the adenine and guanine molecules by isotope interchange from
heavy water in the presence of a platinum catalyst.'*^^ The isotopes un-
doubtedly enter at position 8 of guanine and positions 2 or 8 (or both)
of adenine, and the amount of exchange is a function of the isotope content
of the aqueous media used.

An amino group may be introduced into the 8-position of purines un-
substituted at C-8 by coupling \\'ith 2,4-dichlorobenzenediazonium chloride
and reducing the resulting diazo compound with sodium hydrosulfite.'*^^'^^-
For other transformations including methylations on ring nitrogen, see
Fischer,^ Johnson,'^ and Biltz.'®^

h. From Pyrimidines

The most versatile and widely used method for the synthesis of purines
was developed in 1900 by Traube.^^ This method may be considered to
consist of two parts: (a) the preparation of the appropriate 4,5-diamino-
pyrimidine, and (b) ring closure to the purine. Traube introduced the
amino group into the 5-position of 4-amino-6-hydroxy- and 4,6-diamino-
pyrimidines (bearing a mercapto, hydroxyl, or amino group at C-2) by
nitrosation (cf. Lythgoe et al}^^) followed by ammonium sulfide reduction.
Variations of this method include reduction of the nitroso group with
hydrosulfite,^^-''^^''^^ the use of 5-nitro derivatives and their subsequent
reduction to amines,"^ ^^^^^^■^-^•''^^ the use of 5-arylazopyrimidines which
are readily reduced to S-aminopyrimidines,^^^"*^* and the use of hydrolyzable

"• H. Getler, P. M. Roll, J. F. Tinker, and G. B. Brown, J. Biol. Chem. 178, 259 (1949).

«2 M. L. Eidinoff and J. E. Knoll, /. Am. Chem. Soc. 75, 1992 (1953).

«3 J. R. Spies and T. H. Harris, Jr., ./. Am. Chem. Soc. 61, 351 (1939).

«" M. F. Mallette, E. C. Taylor, and C. K. Cain, /. Atn. Chem. Soc. 69, 1814 (1947) ; see
also 68, 1996 (1946).

«5 R. K. Robins, K. J. Dille, C. H. Willits, and B. E. Christensen, J. Am. Chem. Soc.
75, 263 (1953); see correction, p. 6359.

«6 J. Baddiley, B. Lythgoe, and A. R. Todd, /. Chem. Soc. 1943, 386.

«7 B. Lythgoe, A. R. Todd, and A. Topham, /. Chem. Soc. 1944, 315.

«» L. F. Cavalieri, J. F. Tinker, and A. Bendich, /. Am. Chem. Soc. 71, 533 (1949); cor-
rection: 72, 5801 (1950).

132 AARON BENDICH

5-acylaminopyrimidines (cf. Cavalieri et al.*^^). A new variation employs
ethyl nitrosocyanoacetate (or ethyl hydroxyiminocyanoacetate) which,
upon condensation with the proper amidine or urea derivative, yields 5-
nitrosopyrimidines directly .^^°

Although the 4 , 5-diaminopyrimidines also serve routinely as intermedi-
ates in the synthesis of pteridines/"* -^^ their use in purine synthesis pre-
sents little ambiguity because of the reagents employed and the much
greater reactivity of the 5-amino group. Reaction of 4 , 5-diamino-2 , 6-di-
hydroxypyrimidine with ethyl chloroformate leads only to the 5-urethan,
and the urethan (sodium salt) cyclizes to uric acid upon heating.^^ With
formic acid, it appears that only the 5-formyl derivatives are formed from
di-, tri-, and tetra-aminopyrimidines.^'''^^'^'"''^^^ When 2,4,5-triamino-6-
hydroxypyrimidine is refluxed with formic acid-sodium formate, guanine is
formed directly in excellent yield without the necessity of isolating the in-
termediate 5-formyl compound .^^ Similarly, a nearly quantitative yield of
hypoxanthine is obtained merely by refluxing 4 , 5-diamino-6-hydroxy-
pyrimidine (sulfate) with formic acid."° However, other 4-amino-5-formyl-
aminopyrimidines show a reluctance to dehydrate to the corresponding
purines, and it is necessary to heat the dry compounds or their sodium or
potassium salts at elevated temperatures to effect ring closure. Thus,
xanthine, 2,6-diaminopurine, 2-mercaptoadenine, 2-mercaptohypoxan-
thine, 6-methylpurine, purine, and A^'-alkylxanthines are formed when
the dehydrations of the isolated intermediate 5-formyl derivatives are
carried out at i52-300°.3o, 36. 113,182,270 ,286 .442

For this reason, more suitable techniques of purine formation from
4 , 5-diaminopyrimidines were sought. In one method, 4 , 5 , 6-triamino-
pyrimidine is smoothly converted by treatment with sodium dithiofor-
mate into its 5-thioformyl derivative, and, by boiling in water, hydrogen
sulfide is eliminated to furnish adenine.'*^^ This procedure, which was em-
ployed for the preparation of 1,3-N^Mabeled adenine, yields a mixture
of products 65% of which is adenine and the remainder mainly 4,6-di-
amino-5-formylaminopyrimidine."' In another method, adenine is formed
directly by heating 4,5,6-triaminopyrimidine-2-sulfinic and formic acids
to 150-160°.^^^ The sulfinic acid is obtained by alkaline peroxide oxidation
of the 2-mercaptopyrimidine.

A more general method^^^ involves the heating (160°) of the sulfates of a

«' L. F. Cavalieri, V. E. Blair, and G. B. Brown, /. Am. Chem. Soc. 70, 1240 (1948).

"0 P. D. Landauer and H. N. Rydon, /. Chem. Soc. 1953, 3721.

^" M. Gates, Chem. Revs. 41, 63 (1947).

"« J. H. Speer and A. L. Raymond, /. Am. Chem. Soc. 75, 114 (1953).

"« M. Hoffer, Jubilee Vol. Dedicated to Emil Christoph Barell, 1946, 428.

CHEMISTRY OF PURINES AND PYRIMIDINES 133

number of 4 , 5-diaminopyrimidines with an equivalent amount of formic
acid in excess foiTnamide in a sealed tube. Excellent, and in some cases
quantitative, yields of isoguanine, 2,6-diaminopurine, 2-mercapto- and
2-carboxymethylmercapto-adenine, and adenine are obtained in this pro-
cedure which is useful in isotopic syntheses.^^^-^^^'^^^ With the proper sul-
fates, direct formation of adenine, xanthine, guanine, and 2-methylhypo-
xan thine occur in excellent yield when only formamide is used/^^'^*^*^^ It is
of interest that both formic acid and formamide are required to convert
2-hydroxy-4 , 5 , 6-triaminopyrimidine sulfate into isoguanine. '^^

By the judicious choice of labeled intennediates such as formic acid,
urea, thiourea, formamidine, guanidine, cyanoacetic ester, malononitrile,
phenylazomalononitrile, acetamidocyanoacetic ester, etc., isotopes have
been introduced in nearly every position of purine molecules when the
various methods detailed above have been employed. Since C-8-labeled
purines may be made suitably from C'^- or C^^-formic acid in the last step
of the synthesis, considerable attention has been drawn to this type of
synthesis. It is found that the labeled formyl group of 4,6-diamino-5-
formamidopyrimidine sulfate exchanges with nonisotopic formyl groups
when the ring closure is carried out in formamide, and a 75 % diluted C-8-
labeled adenine results.''^* The report*"*^ of only a 20% dilution when A^-
formylmorpholine was substituted for formamide could not be confirmed^^''
since a 50 % dilution was found upon reexamination. No dilution of the
resulting C-8-labeled adenine occurs when 4, 5, 6-triaminopyrimidine sul-
fate is heated with A^-formyl-C"-morpholine or when the dehydration of
4,6-diamino-5-C^*-formamidopyrimidine sulfate is carried out in diethanol-
amine at 210°.^^^ The above dilutions were tentatively explained on the
basis of a special type of exchange termed "reversible transformylation"
in which formylation of both the 4 (or 6)- and 5-amino groups was believed
to occur followed by a random deformylation during ring closure. Another
explanation lies in the possibility of a lability of carbon 8 in exchange re-
actions at elevated temperatures {ca. 200°) with carbon donors such as
formamide. An example of such an exchange is seen in the conversion of
uric acid into xanthine by means of hot formamide. ^^^

Some of the principles that have been discussed are illustrated below in
the synthesis of adenine labeled at N-1 and N-3 with N^^, and with C'^
or C^^ at C-4, C-6, and 0-8:438.444

4" L. F. Cavalieii and G. B. Brown, /. Am. Chem. Soc. 71, 2246 (1949),

446 H. Bredereck, H.-G. von Schuh, and A. Martini, Chem. Ber. 83, 201 (1950).
448 V. M. Clark and H. M. Kalckar, /. Chetn. Soc. 1950, 1029.

447 R. Abrams and L. Clark, /. Am. Chem. Soc. 73, 4609 (1951).

134

AARON BENDICH

NH

II
C

/ \.
H NH,

CN

+

CHN=N— 0

^CN

n-C4H90H

n-CiHgONa

(70-80%)

Formamidine Phenylazomalononitrile

NH2

NH,

N=N— (

*N
I .,

NH2
4,6-Diamino-5-phenyl-
azopyrimidine

Zn, H2O

NHo

(60-85%)

■^ *N

\N/*\

NH2
4,5,6-Triamino-
pyrimidine

HCOOH, HC0NH2

NH2

"N
I

(90%)

HCONH2

(75%)

(90%)

HCOOH

NH2

^N
I . ^

NHCHO

NH2

H

Adenine

4, 6-Diamino-5-f ormyl-
aminopyrimidine

c. From, Imidazoles

It is of interest that the rat transforms 4-amino-5-imidazolecarboxamide
into nucleic acid purines'"* (cf. Chapters 23 and 25) since the conversion of
this and other imidazoles into a variety of purines may be achieved chemi-
cally as well. In earlier syntheses of purines by this route (cf. Lythgoe^^^),
the intermediate imidazoles were relatively inaccessible. This has been
remedied to a great extent by newer methods of preparation. ^°' .448-452

The fusion of 4-amino-5-imidazolecarboxamide with urea leads to the
formation of xanthine in 75% yield^^^ (cf. Stetten and Fox^^^). When the

*« C. S. Miller, S. Gurin, and D. W. Wilson, Science 112, 654 (1950) ; /. Am. Chem. Soc.

74,2892 (1952).
"9 I. Heilbron, /. Chem. Soc. 1949, 2099.

«« A. H. Cook and I. Heilbron, Rec. trav. chim. 69, 351 (1950).
«' E. Shaw and D. W. Woolley, /. Biol. Chem. 181, 89 (1949).
«2 E. Shaw, /. Biol. Chem. 185, 439 (1950).
*" M. R. Stetten and C. L. Fox, Jr., J. Biol. Chem. 161, 333 (1945).

CHEMISTRY OF PURINES AND PYRIMIDINES 135

hydrochloride of this imidazole is heated with formamide at 185°, a good
yield of hypoxanthine is obtained; formylation yields 4-formamido-5-
imidazolecarboxamide, which in turn cyclizes to hypoxanthine by boiling
in dilute bicarbonate solution.^^-

The dihydrochloride of 4-amino-5-imidazolecarboxamidine, prepared by
the strong acid hydrolysis of adenine^*^^ or from malononitrile,''^^ gives
isoguanine upon fusion with urea or reaction with phosgene.^"^ ■'^^^ Adenine
is obtained in 80 % yield when the above carboxamidine is formylated and
the formyl derivative is cyclized in dilute bicarbonate solution.^^^ 'phis re-
action has been adapted for the preparation (in 61 % yield) of 2-C'^-ade-
nine:^^*

NH2 NH2 NH

C

HCOOH HN

OH

_ KHCO3

H— C*=N

H

4-Amino-5-imidazole- 4-Formamido-5-imidazole-

carboxamidine carboxamide (enol form)

Addendum

A new anti tubercular antibiotic, amicetin (C29H44N6O9), obtained from
Streptomyces, has been found to contain, in part, cytosine, p-aminobenzoic
acid, and dexiro-D-methylserine.'***

An inhibitor of S. faecalis, 8-aza-6-mercaptopurine has been prepared by
the direct thiation of 8-azahypoxan thine with P2S5 in boiling pyridine. ^^®

Further studies on the anti-tumor activity of 6-mercaptopurine have
been carried out.''" Of 45 children with acute leukemia, treatment with
this drug has produced^^* good remission in 15 and partial remission in 10.

The inhibitor 2-thiouracil-^^ is incorporated into the ribonucleic acid of

"^ A. R. P. Paterson and S. H. Zbarsky, /. Am. Chein. Soc. 75, 5753 (1953).

«5 E. H. Flynn, J. VV. Hinman, E. L. Caron, and D. O. Woolf , Jr., /. Am. Chem. Soc. 75,

5867 (1953).
«« C. T. Bahner, B. Stump, and M. E. Brown, /. Am. Chem. Soc. 75, 6301 (1953).
*" D. A. Clarke, F. S. Philips, S. S. Sternberg, C. C. Stock, G. B. Elion, and G. H.

Hitchings, Cancer Research 13, 593 (1953).
^58 J. H. Burchenal, M. L. Murphy, R. R. Ellison, M. P. Sykes, T. C. Tan, L. A. Leone,

D. A. Karnofsky, L. F. Graver, H. W. Dargeon, and C. P. Rhoads, Blood 8, 965

(1953).

136 AARON BENDICH

tobacco mosaic virus in amounts equal to about 20 % of the normal uracil
content/*^

Exposure of neutral aqueous solutions of uric acid to ultraviolet irradia-
tion gives rise to triuret (1 ,3-dicarba,mylurea).''^"

A detailed paper dealing with the occurrence of 5-hydroxymethylcytosine
in the DNA of bacteriophages T2 , T4 , and Te of E. coli has appeared.^"
The free synthetic base survives treatment with 72 % HCIO4 for one hour
at 100°, but such conditions lead to its destruction when in nucleic acid
linkage.

A thoughtful review of the chemistry of the simple p> rimidines has re-
cently been written. ^^^

The presence of 2-methyladenine in the vitamin Bi2-like factor "A" and
in crystalline pseudovitamin Bi2d has been announced .^*^''*^*

General References

E. Fischer, "Unterschungen in der Puringruppe." Springer, Berlin, 1907.

A. Fodor, in "Biochemisches Handlexikon" (Abderhalden, ed.), Vol. 9, p. 262. Springer,

Berlin, 1915.
V. Meyer and P. Jacobson, "Lehrbuch der organischen Chemie," Vol. II, pp. 1172,

1247. De Gruyter, Berlin and Leipzig, 1920.
S. J. Thannhauser, in "Biochemisches Handlexikon" (Abderhalden, ed.). Vol. 10, p.

113. Springer, Berlin, 1923.
P. A. Levene and L. W. Bass, "Nucleic Acids." Chemical Catalog Co., New York, 1931.
T. B. Johnson and D. A. Hahn, P3friniidines : their amino and aminooxy derivatives,

Chem.Revs. 13, 193 (1933).

A. Winterstein and F. SomI6, Purine, pyrimidine und verwandte verbindungen, in

"Handbuch der Pflanzenanalyse" (Klein, ed.), Vol. 4, p. 362. Springer, Vienna,
1933.
T. B. Johnson, The Chemistry of pyrimidines, purines, and nucleic acids, in "Organic
Chemistry, An Advanced Treatise" (Oilman, ed.). Vol. 2, p. 948. Wiley, New
York, 1938.

B. Lythgoe, Some aspects of pj^rimidine and purine chemistry, Quart Revs. {London)

3, 181 (1949).
H. Bredereck, Purin und pyrimidinverbindungen, in "Physiologische Chemie" (Flas-

chentrager and Lehnartz, eds.), Vol. 1, p. 796. Springer, Berlin, 1951.
D. O. Jordan, Nucleic acids, purines and pyrimidines, Ann. Rev. Biochem. 21, 209

(1952).

^** R. Jeener and J. Rosseels, Biochim. et Biophys. Acta 11, 438 (1953).

"« J. Fellig, Science 119, 129 (1953).

«i G. R. Wyatt and S. S. Cohen, Biochem. J. 55, 774 (1953).

«2 D. J. Brown, The Simple Pyrimidines, Rev. Pure Appl. Chem. 3, 115 (1953).

*^3 F. B. Brown and E. L. Smith, Biochem. J., in press.

«■» H. W. Dion, D. G. Calkins, and J. J. Pfiifner, /. Am. Chem. Sac. 76, 948 (1954),

CHAPTER 4

Chemistry of Nucleosides and Nucleotides*
J. BADDILEY

Vagc
1. Nucleosides ]3S

1. Introduction 13S

2. Structure of Nucleosides 140

a. The Nature of the Bases 140

b. Carbohj'drate Components 140

c. Nucleoside Conversions 142

d. Position of the Glj'cosidic Linkage 143

e. The Ring Structure of the Carbohj'drates 144

f. Configuration of the Glycosidic Linkage 148

3. Synthesis of Nucleosides 149

a. Purine Nucleosides 149

b. Pyrimidine Nucleosides . 155

4. General Properties of Nucleosides 157

5. Miscellaneous Nucleosides 157

a. Adenine Thiomethyl Peutoside . 157

b. Crotonoside 158

c. Uric Acid Riboside 159

d. Cordj'cepin 159

e. Spongothj'midine 159

f. Orotidine 159

g. Vicine 159

h. Puromycin 160

II. Nucleotides 160

1. Introduction 160

2. Structure of Nucleotides 162

a. Inosine-5'-phosphate 162

b. Adenosine-5'-phosphate . 163

c. Guanosine-5'-phosphate . 164

d. Cytidine-5'-phosphate 164

e. Uridine-5'-phosphate 165

f. Adenylic Acids a and h 165

g. Guanylic Acids a and h ■ 167

h. Cytidylic Acids a and b 169

i. Uridylic Acids a and h 170

j. Cj'clic Phosphates 170

* The author has preferred not to follow, in the graphical presentation of struc-
tures, the style used by Chemical Abstracts — [The Editoks].

137

138 J. BADDILEY

k. Deoxyribose Nucleotides 171

1. Deoxynucleoside Diphosphates 172

3. Synthesis of Nucleotides 172

a. The 5'-Phosphates 173

b. The a and b Isomers 174

c. The Cyclic Phosphates 174

d. Deoxyribose Nucleotides 175

e. Miscellaneous 176

4. Properties of Nucleotides 177

5. Nucleotide Coenzymes 177

a. Adenosine Di- and Triphosphates (ADP and ATP) 179

b. Di- and Triphosphopyridine Nucleotides (DPN and TPN) 181

c. Flavin Adenine Dinucleotide (FAD) 183

d. Coenzyme A (CoA) 184

e. Uridine Diphosphate Glucose (UDPG) 186

III. Addendum 188

I. Nucleosides
1. Introduction

In most living cells pyrimidines and purines are found in fairly large
amounts. [Cf. Bendich, Chapter 3.] Occasionally they occur in the free state,
e.g., theophylline, theobromine, caffeine, etc.; but those which predominate
in normal cells are usually present as glycosides. Purine and pyrimidine
glycosides are known as nucleosides. The nucleosides may be present as such
in cells but are found for the most part as their phosphoric esters, the nu-
cleotides, which play a vital role in all cells as coenzymes and, in a highly
polymerized condition, as nucleic acids. The term "nucleoside" has been
used in two senses. It has been considered as applying only to those pyrim-
idine and purine glycosides which are formed by hvdrolysis of nucleic acids,
but it is often used more generally for all naturally occurring pyrimidine
and purine glycosides. Nucleosides are, with very few exceptions, either |S-d-
ribofuranosides or D-2-deoxyribofuranosides. They are generally formed by
hydrolysis of nucleic acids with alkaline reagents and may be separated from
each other by methods described in Chapters 5 to 8.

The first nucleoside to be isolated was called vernine' and renamed
guanosine (I) later. It is found together with adenosine^ (II) in alkaline
hydrolysates of ribonucleic acid. These two substances are believed to rep-
resent the entire purine nucleoside components of ribonucleic acids from
different sources.

The pyrimidine nucleosides occurring in robonucleic acid are uridine (III)
and cytidine (IV). They are the 3-|S-D-ribofuranosides of uracil and cytosine,
respectively.

1 E. Schulze and E. Bosshard, Z. physiol. Chem. 9, 420 (1885); 10, 80 (1885).
« P. A. Levene and W. A. Jacobs, Ber. 42, 2703 (1909).

CHEMISTRY OF NUCLEOSIDES AND NUCLEOTIDES

139

o-

N.

OH OH
I I
■-CH-CH-CH-CH-CHrOH

OH

N

I

Guanosine

OH OH
I I
CH-CH-CH-CH-CHrOH

NH,

II

Adenosine

0

OH OH
I I
^CH-CH-CH-CH-CH^-OH

O-

OH OH
I [
-CH-CHCH-CH-CH,OH

N:

OH

NH,

III

Uridine

IV

Cytidine

The two purine nucleosides present in deoxyribonucleic acid are adenine
deoxyriboside (V) and guanine deoxy riboside (VI). These are both 9-d-
(2'-deoxy)ribofuranosides in which the configuration of the glycoside link-
age is not known.

OH
I
CH-CH:-CH-CH-CH,-OH

I

NH.

o

. N^

OH
I
L-CH-CHrCH-CH-CH. OH

H,N|^

OH

VI

Adenine deoxyriboside

Guanine deoxyriboside

The major pyrimidine nucleoside components of deoxyribonucleic acid
are cytosine deoxyriboside (VII) and thymidine (thymine deoxja'iboside)
(VIII). These are both 3-D-(2'-deoxy)ribofuranosides of undetermined con-
figuration at the glycosidic linkage. Some samples of deoxyribonucleic acid
contain small amounts of the pyrimidine nucleosides, 5-methylcytosine
deoxyriboside (IX) and uracil deoxyriboside (X), but it is possible that the
latter is an artifact arising by bacterial deamination of the cytosine nucleo-
side.^

3 C. A. Dekker and A. R. Todd, Nature 166, 557 (1950).

140

J. BADDILEY

OH

LCH-CHj-CH-CH-CHj-OH

OH

'-CH-CHrCH-CH-CH,-OH

NH,

N^Me
OH

VII
Cytosine deoxyriboside

VIII
Thymidine

OH

•^CH-CHj-CH-CH-CH^-OH

NH,

OH

l-CH-CHj-CH-CH-CHa-OH
I

N

OH

IX

5-Methylcytosine deoxyriboside

X

Uracil deoxyriboside

2. Structure of Nucleosides
a. The Nature of the Bases

Adenosine was discovered by Levene and Jacobs^ as a component of ribo-
nucleic acid. They showed that it yielded adenine (6-aminopurine) on acid
hydrolysis. The structure and synthesis of adenine and other purine and
pyrimidine bases from nucleic acids are discussed in Chapter 3. Guanine
(2-amino-6-hydroxypurine) was identified as the purine component of
guanosine by Schulze.^ The pyrimidine nucleosides uridine and cytidine
yield the pyrimidines uracil (2 , 4-dihydroxypyrimidine) and cytosine (4-
amino-2-hydroxypyrimidine), respectively, on vigorous acid hydrolysis.

The deoxyribosides of adenine, guanine, cytosine, and thymine yield the
respective purines and pyrimidines after acid hydrolysis.^- ^

6. Carbohydrate Components

The carbohydrate component of ribonucleic acid, and hence of the cor-
responding nucleosides, was described by Hanmiarsten^ as a pentose, and
its identification as D-ribose followed later. Levene and Jacobs^ "^ obtained it
crystalline and reported that it was a new sugar resembling arabinose in
many of its properties but differing from both d- and L-arabinose in melting

* P. A. Levene and E. S. London, J. Biol. Chem. 81, 711 (1929); 83, 793 (1929).

« W. Klein, Z. physiol. Chem. 224, 244 (1934); 255, 82 (1938).

« O. Hammarsten, Z. physiol. Chem. 19, 19 (1894).

' P. A. Levene and W. A. Jacobs, Ber. 41, 2703 (1908).

» P. A. Levene and W. A. Jacobs, Ber. 42, 1198 (1909).

9 P. A. Levene and W. A. Jacobs, Ber. 44, 746 (1911).

CHEMISTRY OF NUCLEOSIDES AND NUCLEOTIDES 141

point, rotation, etc. On oxidation it gave, first, D-ribonic acid and then an
optically inactive trihydroxyglutaric acid.

CH2-0H

1

CH2OH

1

CO2H

1
(CH.0H)3

1
-^ (CH-0H)3 -

> (CH-OH)
1

CHO

CO2H

1
CO2H

D-Ribonic acid

Full identification was obtained by the synthesis of D-ribose.^"

More diflficulty was experienced in the identification of the carbohydrate
component of the pyrimidine nucleosides. These substances are stable to
dilute acids and concentrated acids convert the liberated pentose to furfural.
However, simultaneous hydrolysis and oxidation of cytidine with hydro-
bromic acid and bromine gave 5-bromouracil and D-ribonic acid,^^ indicating
that this nucleoside contains a D-ribosyl residue. Furthermore, pyrimidine
glycosides may be reduced catalytically to 4 , 5-dihydropyrimidine glyco-
sides which are quite labile towards acids, giving the pyrimidines and free
sugar. Thus uridine is converted into dihydrouridine (XI) and this, after
acid hydrolysis, gives D-ribose."

-0-

OH OH
I I
^CH-CH-CH-CH-CH^-OH

O^N

OH

XI

4,5-dihydrouridine

The presence of D-ribose in the four nucleosides obtained from ribonucleic
acid has been confirmed by the conversion of their sugar components into
D-ribobenziminazole. The suggestion that this type of nucleic acid may con-
tain small amounts of L-lyxose derivatives is now believed to be incorrect.'^

The sugar component of the deoxyribonucleosides proved very difficult
to identify. Acid hydrolysis under conditions similar to those employed with
the purine ribonucleosides effects complete conversion of the deoxysugar
to levulinic acid. Howe\xr, short hydrolysis of the guanine deoxynucleoside
with very dilute acid gave the base and a crystalline deoxypentose.^' ^^

" W. A. van Ekenstein and J. J. Blanksma, Chem. Weekblad 10, 664 (1913).

" P. A. Levene and F. B. La Forge, Ber. 45, 608 (1912).

12 J. M. Gulland and G. R. Barker, /. Cheyn. Soc. 1943, 625; J. M. Gulland, ibid. 1944,

208; G. R. Barker, K. R. Cooke, and J. M. Gulland, ibid. 1944, 339; G. R. Barker,

K. R. Farrar, and J. M. Gulland, ibid. 1947, 21.
" P. A. Levene and T. Mori, J. Biol. Chem. 83, 803 (1929).

142 J. BADDILEY

It formed a benzylphenylhydrazone, but not an osazone, and gave color
tests typical of the group of 2-deoxysugars. It differed from D-2-deoxyxylose,
but was indistinguishable from synthetic L-2-deoxyribose in all respects
except the sign of rotation. '"^ It was then, D-2-deoxyribose. Recently it has
been shown that cautious acid hydrolysis of deoxyribonucleic acid in the
presence of a-toluenethiol gives the benzyl mercaptal of D-2-deoxyribose
which may be isolated in crystalline form.'^

c. Nucleoside Conversions

Some degree of generalization about structural features in nucleosides is
possible since, although all these features have not always been verified in
each nucleoside, the interconversions discussed below enable structural de-
ductions for one to be applied to others.

Adenosine (II) is converted into inosine (9-D-ribofuranosylhypoxanthine)
(XII) by the action of nitrous acid.^^'^^ Inosine was isolated from natural
sources by Haiser and Wenzel in 1908, before its relationship to adenosine
and the nucleic acids w^as established.^^ Similarly, nitrous acid converts
guanosine (I) into xanthosine (9-D-ribofuranosylxanthine) (XIII).^^' 2"' -^

■0

-0-

OH OH OH OH

II II

^CH-CH-CHCH-CHo-OH l-CH-CH-CH-CH -CHj-OH

OH OH ^^

xn XIII

Inosine Xanthosine

Adenine deoxyriboside is deaminated to hypoxanthine deoxyriboside by
an intestinal deaminase.^ This deaminase is usually present in the enzyme
concentrate used for the production of pyrimidine and purine deoxyribo-
sides from deoxyribonucleic acid ; consequently early methods for the isola-
tion of these nucleosides always yielded the hypoxanthine and not the
adenine derivative.^ The action of the deaminase may be inhibited by silver
ions.

" P. A. Levene, L. A. Mikeska, and T. Mori, J. Biol. Chem. 85, 785 (1930).

" P. W. Kent, Nature 166, 442 (1950).

18 P. A. Levene and W. A. Jacobs, Ber. 43, 3150 (1910) .

" P. A. Levene and R. S. Tipson, /. Biol Chem. Ill, 313 (1935).

18 J. M. Gulland and E. R. Holiday, J. Chem. Sac. 1936, 765.

19 F. Haiser and F. Wenzel, Monatsh. 29, 157 (1908).

"o J. M. Gulland and T. F. Macrae, J. Chem. Soc. 1933, 662.
21 P. A. Levene, J. Biol. Chem. 55, 437 (1923).

CHEMISTRY OF NUCLEOSIDES AND NUCLEOTIDES 143

The pyrimidine ribonucleosides have been related by the conversion of
cytidine into uridine with nitrous acid.^^

d. Position of the Glycosidic Linkage

The ease with which purine nucleosides may be liydrolyzed in acid
strongly suggests that the linkage between purine and sugar does not in-
volve a C — C bond, but is much more in keeping with the known properties
of A^'-glycosides. In view of the nucleoside conversions outlined above, the
primary amino groups in adenosine and guanosine cannot be involved. It
follows, then, that the N atoms at positions 1, 3, 7, or 9 may be involved.
Positions 1 and 3 were eliminated by methylation of xanthosine to give a
theophylline riboside (XI V).-^

OH OH

1 1

Me

MeN^
0

LcH-CH-CH-CH-CH,-OH

Me
NH. ^

XIV

Theophylline rib

oside

XV

9-Methyladenine

The early work established that the sugar was attached to one of the imi-
nazole N atoms in the purine nucleosides. The assumption that these were
7-glycosides was corrected later. The ultraviolet absorption spectra of
adenosine and inosine closely resemble those of 9-methyladenine (XV) and
9-methylhypoxanthine but differ markedly from those of the corresponding
7-methyl derivatives.^^ Similarly, guanosine-- and xanthosine-^ show ultra-
violet absorption spectra very similar to those of the corresponding 9-methyl
and not the 7-methyl derivatives. It was concluded that the natural purine
ribosides are 9- and not 7-glycosides.

This technique has also been applied to the purine deoxyribosides.
Adenine deoxyriboside-^ and guanine deo.xyriboside'-- show spectra very
similar to those of adenosine and guanosine, and it was concluded that these
must be 9-gl3^cosides.

Chemical evidence supporting the 9-glycosidic structure of the purine
ribosides has been forthcoming from the periodate oxidation studies dis-
cussed in the next section.

Chemical methods have been used for determining the position of the
glycosidic linkage in the pyrimidine ribosides. Since cytidine may be de-

22 J. M. Gulland and L. F. Story, /. Chem. Soc. 1938, 692.

23 J. M. Gulland, E. R. Holiday, and T. F. Macrae, J. Chem. Soc. 1934, 1639.
" J. M. Gulland and L. F. Story, J. Chem. Soc. 1938, 259.

144 J. BADDILEY

aminated to uridine without loss of the ribose residue it follows that the
glycosidic linkage does not involve the 6-substituent in these nucleosides.
Furthermore, position 5 may be dismissed from consideration since uridine
can be converted into 5-nitrouridine and 5-bromouridine without loss of the
carbohydrate residue. The formation of 4 , 5-diphenylhydrazinouridine
(XVI) by the action of bromine, then phenylhydrazine, on uridine indicates
the absence of substituents on positions 4 and 5.^^ This reaction is consid-
ered specific for 3-substituted uracil derivatives; consequently it seemed
likely that uridine and cytidine are 3-glycosides.

OH OH

1 1

0^

1

LcH-CH-CH-CH

1

f ^"l|NH-NHPh
Vj^NH-NHPh

CH,

OH

H

MeN JJ

OH

T
0

XVI
Diphenylhydrazi

nou

ridine

XVII

1-Methyluracil

The reaction between uridine and hydrazine to give pyrazolone also
indicates the absence of substituents at position 4.^^ Since uridine may
be converted into AT-methyluridine which, on hydrolysis, yields 1-methyl-
uracil (XVII), it follows that uridine and hence cytidine are 3-glycosides."

Thymidine is probably a 3-glycoside since methylation and hydrolysis
of deoxyribonucleic acid results in the production of 1-methylthymine.^*

e. The Ring Structure of the Carbohydrates

The size of the sugar ring in the purine nucleosides has been determined
by methylation and oxidation. Acetylation of adenosine and subsequent
methylation then deacetylation yields a trimethyl-A''-methyladenosine and
this, on hydrolysis with dilute acid, gives V^-methyladenine and a tri-
methylribose.^^ The same trimethylribose may be isolated after similar
transformations on guanosine.^" Its structure is established by oxidation to
2,3,5-trimethyl-7-D-ribonolactone and then to meso-dimethoxysuccinic
acid. The trimethylribose itself was identified subsequently by comparison
with synthetic 2,3,5-trimethyl-D-ribofuranose.^^

" P. A. Levene, /. Biol. Chem. 63, 653 (1925).

28 P. A. Levene and L. W. Bass, /. Biol. Chem. 71, 167 (1926).

" P. A. Levene and R. S. Tipson, /. Biol. Chem. 104, 385 (1935).

28 H. Bredereck, G. Muller, and E. Berger, Ber. 73, 1058 (1940).

29 P. A. Levene and R. S. Tipson, J. Biol. Chem. 94, 809 (1932).

30 P. A. Levene and R. S. Tipson, /. Biol. Chem. 97, 491 (1932).

31 P. A. Levene and E. T. Stiller, J. Biol Chem. 102, 187 (1933).

CHEMISTRY OF NUCLEOSIDES AND NUCLEOTIDES

145

o —

MeO OMe
I I
CH-CH-CH-CH-CH,-OMe

^N N

MeHN ^^

2',3',5',iV''-Tetramethyiadenosine

MeO OMe

I I

HO,C-CH-CH-UOoH

weso-Dimethoxysuccinic acid

O

MeO OMe
I I
HO-CH-CH-CHCH-CHj-OMe

2,3,5-Tnmethyl- D-ribofuranose

O —

MeO OMe
I I
CO-CH-CH-CH-CH,-OMe

2,3,5-Trimelliyl-D-ribonolactone

The ring structure of the ribose residue in the pyrnnidine nucleosides has
been determined in much the same way. Hydrogenation of triacetyl-
uridine gives the 4,5-dihydro derivative which may be methylated with
simulataneous deacetylation, then oxidized with bromine and hydrogen
bromide. The product is 2,3,5-trimethyl-7-D-ribonolactone, which may be
isolated and identified as before.'*'' Tt follows that uridine and cytidine are
furanosides.

The furanose configuration of ribonucleosides is supported by their be-
havior towards trityl chloride. It is known that this reagent reacts prefer-
entially with primary hydroxyl groups in sugars and their derivatives,
giving trityl ethers. Exceptions to the rule are known, Init inider con-
trolled conchtions reaction is indicative of the presence of a primary hy-
droxyl group in a sugar. Adenosine gives a mixture of mono- and ditrityl-
adenosine under these conditions.*^"^® Monotrityladenosine gives a tritosyl
derivative which is converted into tritosyladenosine by short acid hydroly-
sis. None of the tosyl groups in this compound is replaceable by iodine when
heated with sodium iodide. Since this replacement is usually confined to
tosyl esters of primary alcohols it follows that tritosyladenosine bears an
unsubstituted primary hydroxyl group, i.e., it must be A^®, 2', 3 '-tritosyl-
adenosine. The monotrityladenosine must be 5'-monotrityladenosine and
consequently adenosine must contain a furanose ring. The ditrityladenosine
is iY^,5'-ditrityladenosine, as shown by transformations on its diacetyl
derivative. Hydrolysis of the last-named compound with acetic acid yields
a diacetyladenosine (2',3'-diacetyladenosine), identical with that obtained
by hydrolysis of A*^,2' ,3'-triacetyl-5'-trityladenosine.

'2 p. A. Levene and R. S. Tipson, J. Biol. Chcm. 101, 529 (1933).

33 H. Bredereck, Ber. 66, 198 (1933).

"H. Bredereck, Z. physiol. Chem. 223, 61 (1934).

" H. Bredereck, Ber. 65, 1830 (1932).

36 P. A. Levene and R. S. Tipson, J. Biol. Chem. 121, 131 (1937).

146

J. BADDILEY

■0

OH OH
I I
•-CHCH-CHCH-CH.OTri

I

' T

5'-Trityladenosine

TsO OTs
I I
CHCH-CH-CH-CH^G

^ U>

TsHN ^

;V6,2',3'-Tritosyl-5'-trityladenosin

-0-

TriHN ^^

OH OH

•-CHCH-CH-CH-CHj OTri

(1) AciO

(2) HAc

N N
NH. ^^

AcO OAc
I I
CH-CH-CH-CH-CHj-Ol

iV^,5'-Ditrityladenosine

2',3'-Diacetyladenosine

The structure of 2',3'-diacetyladenosine is confirmed by reaction with
p-toluenesulfonyl chloride, giving the A^^,5'-ditosyl derivative. Only one
tosyloxy group in this compound may be replaced by iodine on treatment
with sodium iodide.*^

Similar transformations have established the furanose structure for uri-
dine. Uridine forms mono- and ditrityl derivatives."'^^" Methylation,
then hydrolysis, of monotrityluridine gives 2',3'-dimethyluridine. The lat-
ter substance forms a monotosyl derivative which reacts readily with so-
dium iodide to give crystalline 5'-iodo-2',3'-dimethyluridine. It follows
that the original trityl compound is 5'-trityluridine. Also, tosylation of this
compound gives 2',3'-ditosyl-5'-trityluridine, which after hydrolysis of the
trityl residue does not react with sodium iodide.

Tritylcytidine^^ • ^^ and tritylguanosine^^ are probably 5'-trityl deriva-
tives. In view of the insolubility of guanosine in pyridine the trityl deriva-
tive must be prepared by an indirect method.

Thymidine forms a monotrityl derivative.''" This is believed to be 5'-tri-
tylthymidine since it may be converted into a monotosyl compound (3'-
tosyl-5'-tritylthymidine) which does not react with sodium iodide. It fol-
lows that thymidine possesses a furanose ring.

Further support for the furanose configuration of thymidine is obtained
from the observation that, unlike guanosine and inosine, it does not increase

" P. A. Levene and R. S. Tipson, J. Biol. Chem. 105, 419 (1934).
38 H. Bredereck, E. Berger, and J. Ehrenberg, Ber. 73, 269 (1940).
" H. Bredereck and E. Berger, Ber. 73, 1124 (1940).
^» P. A. Levene and R. S. Tipson, J. Biol. Chem. 109, 623 (1935).

CHEMISTRY OF NUCLEOSIDES AND NUCLEOTIDES

147

•0-

OH OH
I 1
H-CH-CH-CH-CH,-OTri

■0-

N

N

MeO OMe
I I
•-CH-CH-CH-CH-CH^-OH

I

OH

5'-Trityluridine

-0-

TsO OTs
I I
"-CH-CH-CH-CH-CH.-OTri

OH

2',3'-Dimethyluridine

(1) Tosyl chloride

(2) Nal

0 —

MeO OMe
I I
i-CH-CH-CHCH-CH.-I

O^N.

OH
2',3'-Ditosyl-5'-trityIuridine

OH

5'-Iodo-2',3'-dimethyluridine

the acidity of boric acid in the Boeseken test'*' and consequently does not
possess a cis-l ,2-glycol grouping. Guanine and hypoxanthine deoxyribosides
resemble thymidine in this respect and so are probably furanosides/^

Strong evidence for the furanose structure in both rihonucleosides and
deoxyribonucleosides is provided by the action of periodate on these com-
pounds. The course of oxidation of nucleosides with periodate follows a
course similar to the corresponding reactions of the 0-glycosides. The na-
tural rihonucleosides consume 1 mol. of periodate, giving a dialdehj'de but
no formic acid. This behavior is consistent with a furanoside structure.^'

-O-

io«

-O-

-> RCHCHO OHCCHCH-OH

R-CH-CHOHCHOHCH-CHoOH

Under similar conditions the pyrimidine and purine deoxyribosides do not
consume periodate. The as- 1 ,2-glycol system must be absent from these
compounds which must be furanosides.^^

Adenosine, guanosine,^® inosine'^ and uridine^^ are converted into 2' ,3'-
isopropylidene derivatives by reaction with acetone in the presence of acidic
dehydrating agents, e.g., zinc chloride, copper sulfate, sulfuric acid. Both
2',3'-isopropylidene inosine and 2',3'-isopropylidene uridine form 5'-tosyl

^' P. A. Levene and R. S. Tipson, Z. physiol. Chem. 234, V (1935).

^2 K. Makino, Biochem. Z. 282, 263 (1935).

« B. Lythgoe and A. R. Todd, J. Chem. Soc. 1944, 592.

" D. M. Brown and B. Lythgoe, J. Chem. Soc. 1950, 1990.

^* P. A. Levene and R. S. Tipson, J. Biol. Chem. 106, 113 (1934).

148

J. BADDILEY

derivatives which yield the corresponding 5'-iodo derivatives by reaction
with sodium iodide. These conversions, like those already discussed for the
trityl derivatives, are consistent with furanose ring structures.

-0

OH ^^

MeCMe

/\
0 0

•-CH-CH-CH-CH-CH.-OH

N,

-0-

2',3'-Isopropylidene inosLne

MeCMe

/\
O 0

LCH-CH-CH-CH-CH,-OTs
2',3'-Isopropylidene-5'-tosylinosine

.N

/. Configuration of the Glycosidic Linkage

The behavior of ribonucleosides towards periodate has been used in a
chemical determination of the configuration of their glycosidic linkage and,
at the same time, provides confirmatory evidence for the location of the
sugar at position 9 in the purine nucleosides.

The dialdehyde obtained by the action of periodate on adenosine is
identical with the dialdehyde from 9-D-mannopyranosyladenine.''« This
mannosyl compound was prepared by the synthetic route outlined on page
153 and so must be a 9-glycoside. It follows that adenosine is also a 9-gly-
coside.

-0-

OH
-CH-CH-CH-CH-CH-CH,OH

,N. ^N, OH OH

C

— o-

"1

nQC >

NH, ^

CH-CHO 0HC-CH-CH2-0H

9-D-Mannopyranosyladenine

NH,

+ HCOoH

Periodate oxidation of a glucosyladenine also gives this dialdehyde. "^
This glucoside almost certainly has the jS-configuration since it was synthe-
sized from a-acetobromoglucose, and it is reasonable to expect inversion
during the formation of glucosides from this compound. The optical proper-
ties of the oxidation product are consistent with this view. It follows that
the natural nucleoside must also have the /3-configuration.

In the same way, the dialdehydes obtained by periodate oxidation of
uridine and cytidine are identical with those from synthetic samples of

" B. Lythgoe, H. Smith, and A. R. Todd, /. Chem. Soc. 1947, 355.
" J. Davoll, B. Lythgoe, and A. R. Todd, /. Chem. Soc. 1944, 833.

CHEMISTRY OF NUCLEOSIDES AND NUCLEOTIDES

149

3-/3-D-glucopyranosyluracil and 3-/3-D-glucopyranosylcytosine, respectively.
The glucosides were prepared by the synthetic route described on page 155
and must be 3-/3-derivatives.

N

0

OH OH

I I

L-CH-CH-CH-CH-CH-CHrOH
I I

N^ OH

■O-

N.

CH-CHO OHC-CH-CHrCH

I

OH

3-/3- D-Glucopyranosyluracil

OH

The /3-configuration of the natural nucleosides is also shown by the be-
havior of the 5'-tosyl derivatives of the basic members, adenosine and cyti-
dine. Both 2',3'-isopropylidene-5'-p-toluenesulfonyladenosine and the cor-
responding cytidine derivative isomerize with great ease to the respective
cyclonucleosides.^8 Steric requirements for these transformations are only
satisfied by /3-glycosides.

0 —

MeCMe

/\

O O

I I

LcH-CH-CH-CH-CH,-OTs

N
NH2

TsO-

X-ray analysis of several nucleosides has confirmed the structures al-
ready assigned on chemical evidence. In addition it appears that the gly-
cosidic linkage in adenosine lies in the same plane as the purine ring. The
purine ring is flat, in agreement with its aromatic properties, whereas the
sugar is slightly nonplanar. The sugar ring lies approximately perpendicular
to the purine ring in adenosine.*^ Certain other features have been noted.
Although the examination was incomplete, other nucleosides seem to con-
form to the same general pattern. [Cf. Jordan, Chapter 13.]

3. Synthesis of Nucleosides
a. Purine Nucleosides

A number of purine glycosides was synthesized by Emil Fischer and his
collaborators. These were mainly derivatives of theophylline, theobromine,

« V. M. Clark, A. R. Todd, and J. Zussman, J. Chem. Soc. 1951, 2952.
" S. Furberg, Acta Chem. Scand. 4, 751 (1950).

150

J. BADDILEY

and 2,6,8-trichloropurine and were prepared from the silver salts of the
appropriate purines and acetohalogeno-sugars. The products were deacetyl-
ated with ammonia in methanol.^" Theophylline and theobromine hexo-
sides,^"' ^1 pentosides,^!-^^ methyl pentosides,^^-^^ a glucodesoside,*» and a
rhamnofuranoside®" may be prepared in this way. A comparison of the
absorption spectra of these substances with known derivatives of theophyl-
line and theobromine indicates that the theophylline derivatives are 7-gly-
cosides and the theobromine derivatives are O-glycosides.^^ However, the
silver salts of 2,6,8-trichloropurine and 2 , 8-dichloroadenine also react
with acetohalogenosugars^o giving acetylated glycosides. The 2, 8-dichloro-
adenine glucoside is a 9-glucopyranoside as is shown by its absorption spec-
trum^" and by periodate oxidation. The deacetylated glucoside may be
converted into 9-D-glucopyranosyl-adenine or -guanine by appropriate
substitution of the chlorine atoms as shown.

OH

1

yj —

OH

1

-CH'CH-CH-CH-CH-CH,

•OH

R

1

^X^ ^ XT '

OH

—

H,

NH. N

ra.N

2Ha

(1) HNOj

(2) NH,

t

r

'

R

1

R

1

H.Nf-V\
OH

9-D-Glucopyr

anosyladen

ine

9-D-Glucopyranosylguanine

In order to apply this method of synthesis to the natural purine nucleo-
sides an acetohalogenoribofuranose is required. The bromo compound is
prepared from 5-trityl-D-ribose, through 5-trityl-l ,2,3-triacetylribose, from

60 E. Fischer and B. Helferich, Ber. 47, 210 (1914) .

" B. Helferich and M. von Kuhlewein, Ber. 53, 17 (1920).

52 E. Fischer, Ber. 47, 1377 (1914).

" J. Pryde and R. T. Williams, /. Chem. Soc. 1933, 640.

" G. A. Howard, B. Lythgoe, and A. R. Todd, J. Chem. Soc. 1947, 1052.

66 P. A. Levene and H. Sobotka, J. Biol. Chem. 65, 463 (1925).

" E. Fischer, B. Helferich, and P. Ostmann, Ber. 53, 873 (1920).

" E. Fischer and K. von Fodor, Ber. 47, 1058 (1914).

6» P. A. Levene and J. Compton, /. Biol. Chem. 117, 37 (1937).

" P. A. Levene and F. Cortese, /. Biol. Chem. 92, 53 (1931).

80 P. A. Levene and J. Compton, J. Biol. Chem. 114, 9 (1936).

CHEMISTRY OF NUCLEOSIDES AND NUCLEOTIDES

151

which the trityl group may he removed by hydrogenolysis. The resulting
1 ,2,3-triacetyl-D-ribofuranose is converted on acetylation into 1,2,3,5-
tetraacetyl-D-ribofuranose and hence into acetobromoribofuranose." jS-Tet-

CH.-OTn

HO-
HO-

CHa-OTri

CH,-OH

AcO-
AcO-

CHOH

■CH-OAc

AcO-
AcO-

- CH-OAc

CH,-OAc

CH,-OAc

AcO-
AcO-

CH-Br

Acelobrorno-D-nbofuranose

CH-OAc

raacetyl-D-ribofuranose is also stated to be formed by acetylation of D-ri-
bose at elevated temperatures.*'

Repetition of the nucleoside syntheses described above with 2,6-di-
chloroadenine and acetochloro-D-ribofuranose (prepared in an analogous
manner) gives adenosine*^ and guanosine.*^ Acetobromo-D-arabofuranose is
prepared similarly, but this gives a-glycosides with theophylline and aden-
ine.*^ Xylofuranosides may also be prepared by this route. This synthetic
method is not readily applicable to the silver salt of adenine itself since the
basicity of the purine is sufficient to effect dehydrohalogenation of the sugar
derivative. The difficulty may be overcome by acetylation or benzoylation
of the amino groups. From the chloromercury salt of 2 ,6-diacetamidopurine
and acetochlororibofuranose an acetylated 9-D-ribofuranoside is obtained.
Acetyl groups may be removed readil}^ from this compound giving 9-D-ribo-
furanosyl-2,6-diaminopurine. Partial removal of acetyl groups followed by
deamination with nitrous acid then complete deacetylation gives guanosine
in good yield.** Adenosine is prepared in an analogous manner and deamina-
tion of the diaminopurine riboside gives crotonoside (see p. 158).**

«' H. Zinner, Ber. 83, 153 (1950).

" J. Davoll, B. Lythgoe, and A. R. Todd, J. Chem. Soc. 1948, 967.

" J. Davoll, B. Lythgoe, and A. R. Todd, /. Chem. Soc. 1948, 1685.

" N. W. Bristow and B. Lythgoe, J. Chem. Soc. 1949, 2306; P. Chang and B. Lythgoe,

ibid. 1950, 1992.
«5 J. Davoll and B. A. Lovvy, J. Am. Chem. Soc. 73, 1650 (1951).
6« J. Davoll, J. Am. Chem. Soc. 73, 3174 (1951).

152

J. BADDILEY

■0

AcO OAc

AcHNf'^

N
AcHN

N

N.

CH-CH-CH-CH-CH^-OAc

^\ (1) NH, AcHNf^^"-

/ (2)HN0,* N:^

OH

R

f^\ NaOMe H.Nf^^V

[ ^ OH

R

R

\ HNOj

NHa^

6-diaminopurine

Crotonoside

NH,

The above syntheses are moderately convenient and were valuable for
demonstrating the jS-configuration of the glycosidic linkage in the natural
nucleosides. However, they do not in themselves establish the position of
attachment of sugar and purine. An alternative synthesis, although less
convenient, is unambiguous and may be regarded as a modification of a
general synthesis for purines and their 9-substituted derivatives.^^- ** In
the scheme described here the starting point is a substituted 4 , 6-diamino-
pyrimidine. For the synthesis of purine glycosides a 4-amino-6-glycosyl-
aminopyrimidine is employed. These may be prepared by acid-catalyzed
condensation of a diaminopyrimidine with a sugar in alcoholic solution. ^^
A 2', 4'- or 2',5'-dichlorophenylazo group is introduced at the 5-position in
the pyrimidine ring by coupling in neutral solution with the appropriate
diazonium salt.^" The hydroxyl groups in the sugar residue are usually
acetylated at this stage and the azo group reduced to amino with hydrogen
and a nickel catalyst. The resulting acetylated 4 , 5-diamino-6-glycosyl-
aminopyrimidine is converted into the corresponding 5-thioformamidopy-
rimidine by reaction with dithioformic acid or its sodium salt. Cyclization
of the thioformamido compound in pyridine may proceed in two directions,
giving the acetylated 9-glycosyladenine or 6-glycosylaminopurine, which
can be separated before or after deacetylation.'^^

The yield of purine glycoside obtained in the above synthesis is improved
by cyclization of the thioformamido compound with alcoholic alkoxide
solutions. Under these conditions unacetylated glycosylpyrimidines give
exclusively the 9-glycosylpurines whereas the acetylated derivatives yield

" J. Baddiley, B. Lythgoe, D. McNeil, and A. R. Todd, J. Chem. Soc, 1943, 383.

68 J. Baddiley, B. Lythgoe, and A. R. Todd, J. Chem. Soc. 1943, 386.

69 J. Baddiley, B. Lythgoe, and A. R. Todd, J. Chem. SOc. 1943, 571.
'» B. Lythgoe, A. R. Todd, and A. Topham, /. Chem. Soc. 1944, 315.
" J. Baddiley, B. Lythgoe, and A. R. Todd, /. Chem. Soc. 1944, 318.

CHEMISTRY OF NUCLEOSIDES AND NUCLEOTIDES

153

NH2

NH2

N^N^NCeHsCU'
NH,

N^NHj
NH,

N^
[f^NHRi

K^NH-CHS

NH,

Ri

. N. ^N

NH:

1

f

N

RiHN

H

N. ^N

R

R= glycosyl

Ri = acetylated glycosyl

NH,

mixtures of 9-glycosides and 7-glycosylaminopurinesJ^ The general method
has been used for the synthesis of 9-D-xylopyranosyl-/^ 9-D-ribopyranosyl-/*
9-D-mannopyranosyl-,'^® 9-D-glucop3"ranosyl-adenine/^ 9-D-xylopyranosyl-2-
methylthioadenine/^ and 9-D-glucopyranosyHsoguanine7^

Whereas the glycosides prepared by the general route are undoubtedly
9-/3-glycosides,^®' ^^' ''*• ^^ periodate titration shows that they are pyrano-
sides. In order to ensure a furanose structure in the glycosidic residue it is
necessary to start with a suitably protected aldehydo-sugar. For the syn-
thesis of adenosine^" along these lines 2,3,4-triacetyl-5-benzyl-D-ribose is
condensed with 4,6-diamino-2-methylthiopyrimidine to give a Schiff base.
A methylthio residue at position 2 is necessary to increase the reactivity of
the aminopja-imidine. After removal of the acetyl groups with methanolic
ammonia rearrangement occurs to the glycofiu-anoside. This is coupled
with diazotized 2,5-dichloroaniline, acetylated, and then reduced with zinc
and acetic acid to the amine, thioformylated, and cyclized. The 2-methyl-
thio and 5'-benzyl groups are removed by hydrogenolysis with nickel, then
acetyl groups removed with sodium methoxide, giving adenosine.

It is important in this synthesis to protect the 5-position in the initial
sugar derivative with a benzyl or similar group. When acetyl or benzojd
groups are used in this connection, mixtures of furanosides and pyranosides

" G. W. Kenner and A. R. Todd, J. Chem. Soc. 1946, 852.

" G. W. Kenner, B. Lythgoe, and A. R. Todd, J. Chem. Soc. 1944, 652.

■'* J. Baddiley, G. W. Kenner, B. Lythgoe, and A. R. Todd, /. Chem. Soc. 1944, 657.

" A. Holland, B. Lythgoe, and A. R. Todd, J. Chem. Soc. 1948, 965.

'« G. A. Howard, B. Lythgoe, and A. R. Todd, /. Chem. Soc. 1945, 556.

" K. J. M. Andrews, Nity Anand, A. R. Todd, and A. Topham, /. Chem. Soc. 1949,

2490.
'» G. A. Howard, G. W. Kenner, B. Lythgoe, and A. R. Todd, J. Chem. Soc. 1946, 855.
^^ G. A. Howard, G. W. Kenner, B. Lythgoe, and A. R. Todd, J. Chem.. Soc. 1946, 861.
«« G. W. Kenner, C. W. Taylor, and A. R. Todd, J. Chem. Soc. 1949, 1620.

154

J. BADDILEY

j^ CH.OCH.Ph

MeSf ^NH, I ^MeS(r^N = CH-(CHOAc),-CHo-OCHsPh

V I + (CH0Ac)3 »• XT

NHj CHO NH,

-N.

-o-

OH OH
I I
^MeSff^'^NH-CH-CH-CH-CH-CH.OCHjPh

NH,

-N:

-0-

AcO OAc

(f^'^NH-CH-CH-CH-CH-CHj-OCHjPh
N^NH-CHS
NH,

-0-

OH OH
I I
l-CH-CH-CH-CH,

•OCHoPh

MeSf^

N

N.^N

J'

(1) AcjO

NH,

(2) Ni/H

(3) NaOMe

N

-0

OH OH
I I
LCH-CH-CH-CH-CH,-OH

I

N.

NH,

5'-Benzyl-2-methylthioadenosine

Adenosine

are obtained.*' • *■ However, acetyl groups may be used for protection, if at
all stages deacetylation is avoided. In this way 9-D-galactofuranosyl-2-
methylthioadenine has been prepared from 2,3,5,6-tetraacetyl-D-galacto-
furanose.*^

Purine nucleosides may also be synthesized from iminazoles. This method
is more or less restricted to the xanthine glycosides,*'* • ** but with that
limitation is quite useful and has been applied to the synthesis of xanthosine
itself.*^ Starting from an acetohalogeno-sugar and the silver derivative of
4,5-dicarbomethoxyiminazole a glycoside is obtained. This is converted to
a diamide with simultaneous deacetylation by the action of ammonia, and
cyclization to the purine is effected by potassium hypobromite in a modi-
fied Hofmann reaction.

«i G. W. Kenner, B. Lythgoe, and A. R. Todd, /. Chem. Soc. 1948, 957.
"2 G. W. Kenner, H. J. Rodda, and A. R. Todd, J. Chem. Soc. 1949, 1613.
" K. J. M. Andrews, G. W. Kenner, and A. R. Todd, /. Chem. Soc. 1949, 2302.
8^ R. A. Baxter and F. S. Spring, /. Chem. Soc. 1947, 378.
" R. A. Baxter, A. C. McLean, and F. S. Spring, J. Chem. Soc. 1948, 523.
«6 G. A. Howard, A. C. McLean, G. T. Newbold, F. S. Spring, and A. R. Todd, /.
Chem. Soc. 1949, 232.

CHEMISTRY OF NUCLEOSIDES AND NUCLEOTIDES

155

1 > — " X^ ~

MeOjC N MeO^C N

H.N-OC n'
H,N-OC ^

KOBr

R.

YLQf

N:

OH

Xanthosine

„>

R = 2'3'5'-Triacetyl-D-ribofuranosyl
Ri = D-ribofuranosyl

h. Pyrimidine Nucleosides

Pyrimidine nucleosides have been synthesized from aeetohalogeno-sugars
and suitably substituted pyrimidines. When the pyrimidine employed bears
substituents, such as hydroxyl, which are capable of prototropic change, the
glycosyl residue becomes attached thereto and heterocyclic iV-glycosides
cannot be produced in this manner.^"' ^'-' " However, A^-glycosides are
obtained by the action of aeetohalogeno-sugars on 2 ,6-dialkoxypyrimidines,
in which tautomerization possibilities are excluded. This is an extension of
the reaction observed between alkyl halides and 2,6-dialkoxypyrimidines
whereby 3-alkyl-6-alkoxy-2-ketopyrimidines are formed.

>''''l

N:^

R-hal

»•

R

HCl

»•

Ri

OEt

OEt •

NH^\^

OH
Ri

NH

Thus, interaction of acetobromoglucose and 2,6-diethoxypyrimidine gives
an acetylglucoside which on hydrolysis with hydrogen chloride in methanol
yields 3-D-glucosyluracil. Alternatively, b}^ treatment of the acetylated
glucoside with ammonia, 3-D-glucosylcytosine is formed.*^' ** Several struc-
tural analogues have been prepared by this method.*'- ^° Starting from
acetobromoribofuranose and 2,6-diethoxypyrimidine the natural nucleo-

" G. E. Hilbert and T. B. Johnson, J. /l?n. Chern. Soc. 52, 4489 (1930).
«8 G. E. Hilbert and E. F. Jansen, J. Am. Chem. Soc. 58, 60 (1936).
«9 G. E. Hilbert, /. Biol. Chem. 117, 331 (1937).
90 G. E. Hilbert and C. E. Rist, J. Biol. Chem. 117, 371 (1937).

156

J. BADDILEY

TABLE I
Properties of Nucleosides

Nucleoside M.p., °C. [«]„ (H2O)

pK

Derivatives

Adenosine 229

Guanosine 237-240

Uridine

165

Cytidine 220-230

Inosine

Xanthosine

Adenine deoxy-
riboside

Hypoxanthine
deoxyriboside

Guanine deoxy-
riboside

Thymidine

Cytosine deox-
yriboside

Uracil deoxy-
riboside

5-Methylcyto-
sine deoxy-
riboside

218

189-19096

not sharp

not sharp

182-183
186

1633

60.0° "

amino 3.3"' ">*

picrate, m.p. 180-185°

sugar 12.5

(dec.)

60.52°

amino 1.6^^

2' ,3' ,5'-triacetylguan-

(alkali)

hydroxyl 9.16
sugar 12.5

osine, m.p. 226° «»

4.0° "

hydroxyl

5-bromo, m.p. 181-

-6.0°

9.17^^

184°;" 5-hydroxy,

alkali

sugar 12.5

m.p. 222-223°

29.63°

amino 4.22^3 ■ ^^

picrate, m.p. 185-187°;

sugar 12.5

sulfate, m.p. 233°; hy-
drochloride, m.p.
218°; nitrate, m.p.
197°

47.7° »

hydroxyl

2',3',5'-triacetylino-

8,7593. 94

sine, m.p. 236°

-51.2°
(alkali)

■21.0°

-37.5°

32.5° (in

alkali)

40.0°

sugar 12.5

hydrochloride, m.p.
161-164° (dec.)«6

picrate, m.p. 175-178°,
is converted into thy-
midine by nitrous
acid^fia

sides uridine and cytidine are obtained. ^^ Thymine nucleosides are synthe-
sized in a similar manner from 2,6-diethoxy-5-methylpyrimidine.^^

91 G. A. Howard, B. Lythgoe, and A. R. Todd, /. Chem. Soc. 1947, 1052.

92 D. W. Visser, I. Goodmann, and K. Dittmer, /. Am. Chem. Soc. 70, 1926 (1948).

93 P. A. Levene, /. Biol. Chem. 41, 483 (1920) ; cf . P. A. Levene, H. S. Simms, and L.
W. Bass, ibid. 70, 243 (1926).

9" P. A. Levene and H. S. Simms, /. Biol. Chem. 65, 519 (1925).

9* H. Steudel and R. Freise, Z. physiol. Chem. 120, 126 (1922).

96 W. Andersen, C. A. Dekker, and A. R. Todd, J. Chem. Soc. 1952, 2721.

96=' C. A. Dekker and D. T. Elmore, /. Chem. Soc. 1951, 2864.

chemistry of nucleosides and nucleotides 157

4. General Properties of Nucleosides

The nucleosides in general are colorless, crystalline substances with
rather high melting points. They vary somewhat in their solubility in water,
but most representatives are readily soluble in hot, considerably less soluble
in cold water. The pyrimidine nucleosides uridine and cytidine are more
soluble in water than are the purine nucleosides. Both pyrimidine and purine
nucleosides are insoluble in the more common organic solvents. Nucleosides
containing amino groups (adenosine, cytidine) are rather weak bases, while
those containing hydroxy 1 groups on the heterocyclic nucleus (xanthosine,
inosine, uridine) are weak acids.

Physical properties and useful derivatives of some of the better known
natural nucleosides are listed in Table I.

5. Miscellaneous Nucleosides

The nucleosides inosine, xanthosine, hypoxanthine deoxyriboside, and
uracil deoxyriboside which have been described in the previous sections
are thought to be absent from the native nucleic acids, but frequently arise
during isolation, or during hydrolysis of the isolated macromolecule to its
component nucleosides. In addition to these, some nucleosides which have
been isolated from natural sources are neither components nor degradation
products of nucleic acids. These are discussed below.

a. Adenine Thiomethyl Pentoside {5^ -deoxy-5' -methylthioadenosine)

This nucleoside was first isolated from yeast, ^^ but is probably present in
the tissues of many animals. On hydrolysis it gives adenine and a sulfur-
containing sugar, 5-deoxy-5-methylthioribose. The structure of the aldo-
pentose follows from its reduction to the pentitol, periodate titration, and
direct comparison with synthetic 5-deoxy-5-methylthiopentose deriva-
tives.^^'^"^ The ultraviolet absorption spectrum of the nucleoside supports a
9-glycoside structure."'- With nitrous acid it gives "hypoxanthine thio-
methyl pentoside, "^°^ the structure of which has been proved by synthe-
gjgio4,io5 Thig synthesis starts from 2',3'-isopropylidene-5'-p-toluene
sulfonylinosine,^^ from which is prepared 2',3'-isopropylidene-5'-methyl-
thio-5'-deoxyinosine. The isopropylidene residue may be removed by

" J. A. Mandel and E. K. Dunham, J. Biol. Chem. 11, 85 (1912).

98 P. A. Levene and H. Sobotka, J. Biol. Chem. 65, 551 (1925).

" G. Wendt, Z. physiol. Chem. 272, 152 (1942).
lo" K. Satoh and K. Makino, Nature, 165, 769 (1950).
>»' F. Weygand, O. Trauth, and R. Lowenfeld, Ber. 83, 563 (1950).
'»2 R. Falconer and J. M. Gulland, J. Chem. Soc. 1937, 1912.
i°3 R. Kuhn and K. Henkel, Z. physiol. Chem. 269, 41 (1941).
lo^ J. Baddiley, /. Chem. Soc. 1951, 1348.
•"^ K. Satoh and K. Makino, Nature 167, 238 (1951).

158

J. BADDILEY

cautious acid hydrolysis to give 5'-deoxy-5'-methylthioinosine, identical
with "hypoxanthine thiomethyl pentoside."

-0-

MeCMe
/\

0 0

1 I
'-CH-CH-CH-CH-CH,-OTs

N^^N

OH

KSMe

OH ^^

MeCMe

/\
O 0

CH • CH • CH • CH • CH, • SMe

O 1

OH OH I
I I I
L-CH-CH-CH CH-CH2-SMe

N

OH
Hypoxanthine thiomethyl pentoside

■0-

OH OH

L-CH • CH • CH • CH • CH2 • SMe

NH,

Adenine thiomethyl pentoside

This synthesis establishes the |8-configuration at the glycosidic center.
Adenine thiomethyl pentoside itself is synthesized in poor yield by a similar
series of reactions upon 2',3'-isopropylidene-5'-p-toluenesulfonyladeno-

104-106

sme/

It is now known^"^ that adenine thiomethyl pentoside arises in Nature
by decomposition of the transmethylation intermediate "active meth-
ionine." This intermediate has the structure^ "^^ ^^^ shown as XVIII.

-0-

OH OH

Me

NHj

I

•-CH-CH-CH-CH-CHj-S-CHs-CHj-CH-COr

NHa

XVIII
"Active methionme'

h. Crotonoside (9-j3-D-ribofuranosylisoguanine)

This was first isolated from the bean Croton tiglium. It is isomeric with
guanosine and gives on acid hydrolysis D-ribose and isoguanine.^"^ Its ab-

106 F. Weygand and O. Trauth, Ber. 84, 633 (1951).

»" J. Baddiley, G. L. Cantoni, and G. A. Jamieson, /. Chem. Soc, 1953, 2662.

'08 G. L. Cantoni, /. Am. Chem. Soc. 74, 2942 (1952).

109 E. Cherbuliez and K. Bernhard, Helv. Chim. Acta 15, 464 (1932).

CHEMISTRY OF NUCLEOSIDES AND NUCLEOTIDES 159

sorption spectrum indicates a 9-glycoside"° and with nitrous acid it is con-
verted into xanthosine. The formula shown on page 152 is confirmed by
synthesis.

c. Uric Acid Riboside

The nucleoside is present in beef blood. It is hydrolyzed in acid to D-ribose
and uric acid"^ and is probably a 9-glycoside.'^^

d. Cordycepin (9-cordyceposyladenine)

The mould Cordyceps militaris (Linn.) contains an antibiotic, cordy-
cepin,"^ which gives on hydrolysis adenine and a deoxyaldopentose, cordy-
cepose. Cordycepin is probably a 9-glycoside with the structure XIX."*

0 CHj

I I

CH-CHOH-CH

I I

NH

N

XIX

Cordycepin

e. Spongothymidine

This occurs in the sponge Cryptotethia. It yields thymine on hydrolysis
but the sugar component has not been fully identified."^' "^'^

/. Orotidine

A mutant of the mould Neurospora crassa contains a glycoside of orotic
acid. On hydrolysis it yields ribose and orotic acid."^

g. Vicine

Vetch meal contains a pyrimidine glycoside, vicine."^ It can be hy-
drolyzed by acids to D-glucose and the pja-imidine divicine."^ Evidence for
its structure is not entirely satisfactory. [Cf. Bendich, Chapter 3.]

"" R. Falconer and J. M. Gulland. J. Chem. Soc. 1939, 1784.

1" A. R. Davis, E. B. Newton, and S. R. Benedict, J. Biol. Chcm. 54, 595 (1922).

"^ R. Falconer and J. M. Gulland, J. Chem. Soc. 1939, 1369.

"^ K. G. Cunningham, S. A. Hutchinson, W. Manson, and F. S. Spring, J. Chem. Soc.

1951, 2299.
"< H. R. Bentley, K. G. Cunningham, and F. S. Spring, /. Chem. Soc. 1951, 2301.
"* W. Bergmann and R. J. Feeney, J. Am. Chem. Soc. 72, 2809 (1950).
"** According to a private communication from Drs. W. Bergmann and D. Burke,

Yale University, spongothymidine is an arabinoside [The Editors].
i'« A. M. Michelson, W. Drell, and H. K. Mitchell, Proc. Natl. Acad. Sci. U. S. 37,

396 (1951).
1'^ H. Ritthausen, Ber. 29, 894, 2108 (1896).
"8 P. A. Levene, J. Biol. Chem. 18, 305 (1914).

160 J. BADDILEY

h. Puromycin

An antibiotic, puromycin, is produced by Streptomyces alhoniger. On
alcoholysis it yields a 6-dimethylaminopurine, 0-methyl-L-tyrosine and
D-3-amino-3-deoxyribose. Its ultraviolet absorption spectrum indicates
that the carbohydrate residue is attached to the purine at position 9. Two
hydroxyls and an amino group are present and it does not consume period-
ate. These properties are consistent with the partial formula XX. "^* It is
not known whether the carbohydrate residue is in the furanose or pyranose
configuration.

NHs

CO-CH-CHjf 70Me

n<

OH NH 0 O
I I I I
CH-CH-CH-CH-CHj

r

a>

XX

Puromycin

II. Nucleotides
1. Introduction

Nucleotides are the phosphoric acid esters of nucleosides. They may be
prepared by chemical or enzymic hydrolysis of nucleic acids provided that
the conditions chosen do not effect hydrolysis of phosphomonoester groups.
Nucleotides are also formed during hydrolysis of certain coenzymes, but
nucleic acids are the most convenient source for these substances. The first
nucleotide to be isolated, inosinic acid,"^ is probably exceptional in this
respect. It was obtained from meat by fairly direct methods of extraction
and purification and is believed to be an artifact arising through deamina-
tion of "muscle adenylic acid" (adenosine-5'-phosphate), which occurs in
the free state in muscle tissues.^-"

Ribonucleotides may be obtained from ribonucleic acid by gentle chem-
ical or enzymic hydrolysis. Enzymic hydrolysis^'* is incomplete, however,
and chemical methods 'are generally employed for preparative purposes.
Hydrolysis of yeast ribonucleic acid with dilute ammonia gives the nucleo-

"8» C. W. Waller, P. W. Fryth, B. L. Hutchings, and J. H. Williams. /. Am. Chem. Soc.

75, 2025 (1953).
"9 J. von Liebig, Ann. 62, 257 (1847).

120 G. Embden andM. Zimmermann, Z. physiol. Chem. 167, 137 (1927).

121 H. S. Loring and F. H. Carpenter. J. Biol. Chem. 150, 381 (1943).

CHEMISTRY OF NUCLEOSIDES AND NUCLEOTIDES 161

tides known as yeast adenylic acid, guanylic, iiridylic, and cytidylic
acids.'--- '-^ Dilute alkali effects a nearly quantitative conversion to the nu-
cleotide mixture.'-^" Improved methods have been introduced from time to
time'^^ and cautious acid hydrolysis is convenient when only the more sta-
ble pyrimidine nucleotides are required.'-^-'" However, recent investigations
on these nucleotides using refined methods of ion-exchange and paper chro-
matography as well as selective fractional crystallization procedures, show
that they are not homogeneous and consequently the older isolation tech-
niques are no longer used. [Cf. Chapters 5-7, 11, 12.] By suitable combina-
tion of ion-exchange and paper chromatography it can be shown that the
nucleotides produced by hydrolysis of ribonucleic acid are mixtures of iso-
mers. Thus, yeast adenylic acid is a mixture of the two isomers adenylic
acid a and adenylic acid h. Similarly, guanylic, cytidylic, and uridylic acids
are all mixtures of a and b isomers.'-*"'^- Certain of the isomers may be
separated from each other by fractional crystallization of cyclohexylamine
and brucine salts. '^^- '^^ It should be explained that the terms a and 6 were
originally assigned arbitrarily and did not necessarily signify any structural
correlation between the different members. As will be seen later, it is now
known that cytidylic acid a is structurally related to uridylic acid a, sim-
ilarly cytidylic acid h is related to uridylic acid h. The isomerism in all
these nucleotides involves the position of the phosphate residue on C-2',
or C-3', in the respective nucleosides.

Nucleoside-5'-phosphates may also be obtained from ribonucleic acid.
For reasons discussed in Chapter 12, chemical hydrolysis cannot give these
nucleotides. However, ribonuclease hydrolysis under conditions which do
not permit 5'-phosphatases to operate, followed by ion-exchange separation
of the products, yields 5'-phosphates.'^^ [Cf. Cohn, Chapter 6; Schmidt,
Chapter 15.]

" P. A. Levene, J. Biol. Chem. 33, 425 (1918).
" P. A. Levene, J. Biol. Chem. 40, 415 (1919).
"» E. Chargaff, B. Magasanik, E. Vischer, C. Green, R. Doniger, and D. Elson, J.

Biol. Chem. 186, 51 (1950).
2^ H. S. Loring, P. M. Roll, and J. G. Pierce, J. Biol. Chem. 174, 729 (1948).
25 P. A. Levene and W. A. Jacobs, Ber. 44, 1027 (1911).
" S. J. Thannhauser and G. Dorfmiiller, Z. physiol. Chem. 104, 65 (1919).
" H. Bredereck and F. Richter, Ber. 71, 718 (1938).
28 W. E. Cohn, Science, 109, 377 (1949).
2» W. E. Cohn, J. Am. Chem. Soc. 71, 2275 (1949).
3" W. E. Cohn, J. Am. Chem. Soc. 72, 2811 (1950).
3' W. E. Cohn and C. E. Carter, /. Am. Chem. Soc. 72, 2606 (1950).
32 C. E. Carter, J. Am.. Ch^m. Soc. 72, 1466 (1950).
" H. S. Loring and N. G. Luthy, J. Am. Chem. Soc. 73, 4215 (1951).
'4 P. Reichard, Y. Takenaka, and H. S. Loring, J. Biol. Chem. 198, 599 (1952) .
" W. E. Cohn and E. Volkin, Nature 167, 483 (1951) .

162 J. BADDILEY

Ribonuclease digestion, or very mild alkaline hydrolysis of ribonucleic
acid, gives the four possible cyclic nucleoside-2', 3 '-hydrogen phosphates,
together with other products. ^^^ These are artifacts, cyclic phosphate struc-
tures not being present in nucleic acids, but their formation is significant in
connection with the structures proposed for ribonucleic acid (see Chapter
12). Ribonuclease digestion of ribonucleic acid also yields a number of di-
and trinucleotides, some of which contain cyclic phosphate structures.^"- ^^*

Pyrimidine deoxyribonucleotides are produced by mild acid hydrolysis
of deoxyribonucleic acid."^"''** Both 5'-phosphates and 3',5'-diphosphates
are obtained. The purine deoxyribonucleotides are very sensitive to acids
and cannot be prepared by chemical hydrolysis of the parent nucleic acid.
However, these substances are obtained from the nucleic acid by using an
intestinal deoxyribonuclease in the presence of sodium arsenate, thereby
inhibiting the action of an accompanying nucleotidase.^*^"''** Recent methods
for the isolation of deoxyribonucleotides employ ion-exchange resins."*-

150, 151

2. Structure of Nucleotides

As the nucleotides may differ from each other not only in the nature of
the pyrimidine and purine bases attached to either of the sugars D-ribose
and D-2-deoxyribose, but also in the location of the phosphoric ester link-
age, they are best considered individually.

a. InosineS' -phosphate (muscle inosinic acid) (XXI)

This nucleotide is a phosphoric ester containing the components hy-
poxanthine'^2 and D-ribose.^-* On neutral hydrolysis it gives phosphoric acid
and inosine; consequently it is a monophosphate of inosine in which the

i3« R. Markham and J. D. Smith, Biochem. J. 52, 552 (1952) .

'" R. Markham and J. D. Smith, Biochem. J. 52, 558 (1952) .

"8 R. Markham and J. D. Smith, Biochem. J. 52, 565 (1952).

"9 P. A. Levene and H. Mandel, Ber. 41, 1905 (1908).

"« P. A. Levene and W. A. Jacobs, /. Biol. Chem. 12, 411 (1912).

•" P. A. Levene, /. Biol. Chem. 48, 119 (1921).

•« S. J. Thannhauser and B. Ottenstein, Z. physiol. Chem. 114, 39 (1921).

1" S. J. Thannhauser and G. Blanco, Z. phijsiol. Chem. 161, 116 (1926).

1" P. A. Levene, /. Biol. Chem. 126, 63 (1938).

1" C. A. Dekker, A. M. Michelson, and A. R. Todd, J. Chem. Soc. 1953, 947.

•« W. Klein, Z. phijsiol. Chem. 218, 164 (1933).

1" W. Klein and S. J. Thannhauser, Z. physiol. Chem. 218, 173 (1933).

'48 W. Klein and S. J. Thannhauser, Z. physiol. Chem. 224, 252 (1934).

1" W. Klein and S. J. Thannhauser, Z. physiol. Chem. 231, 96 (1935).

160 E. Volkin, J. X. Khym, and W. E. Cohn, /. Am. Chem. Soc. 73, 1533 (1951).

1*1 R. L. Sinsheimer and J. F. Koerner, Science 114, 42 (1951).

1" F. Haiser, Monatsh. 16, 190 (1895).

CHEMISTRY OF NUCLEOSIDES AND NUCLEOTIDES

163

phosphate residue is located at the 2' ,3'- or 5'-position on the sugar chain.^
The ribose phosphate produced by acid hydrolysis must be ribose-5-phos-
phate since it is oxidized with nitric acid to a ribonic acid phosphate and
not a n6o-trihydroxyglutaric acid phosphate. »

OH OH

OH OH OH

H0CH-CH.CH-CH.CH.0P03H. -^ HO.C CH-CH- CH-CHrOPOaH,

D-Ribose-5-phosphate D-Ribonic acid-5-phosphate

The structure of ribonic acid-5-phosphate is confirmed by its slow con-
version into a 7-lactore.i^^' 155 Also, the sugar phosphate itself is converted
into a methylribofuranoside-5-phosphate rather than a mixture of furano-
side and pyranoside^^^ and is reduced to optically active ribitol-5-phosphate.
Full proof for the structure of ribose-5-phosphate is given by its synthesis
from methyl 2, 3-isopropylideneribofuranoside. Phosphorylation with phos-
phoryl chloride, then acid hydrolysis, gives the S-phosphate.^" Better yields
are obtained by phosphorylation with dibenzyl phosphorochloridate, then
hydrogenolysis and acid hydrolysis.'^*

-0-

OH OH
•-CH-CH-CH-CH-CH^-OPCH,

nJC>

OH ^^

XXI
Muscle inosinic acid

h. Adenosine-5' -phosphate (muscle adenylic acid) (XXII)

As its name implies, this nucleotide was first isolated from muscle.'^"
but it is now known to occur as a structural unit of ribonucleic acid and of
several coenzymes. Since it can be deaminated to inosine-5'-phosphate by
an enzyme present in muscle tissue,!^^ it must bear the same relationship
to that nucleotide as does adenosine to inosine, i.e., it must be adenosine-5'-
phosphate. Furthermore, the two nucleotides liberate phosphoric acid at

>53 p. A. Levene and W. A. Jacobs, Ber. 42, 335 (1909).

i5< P. A. Levene and H. S. Simms, J. Biol. Chem. 65, 31 (1925).

165 P A Levene and T. Mori, J. Biol. Chem. 81, 215 (1929).

156 p A. Levene, S. A. Harris, and E. T. Stiller, J. Biol. Chem. 105, 153 (1934).

157 P. A. Levene'and E. T. Stiller, J. Biol. Chem. 104, 299 (1934).

158 A. M. Michelson and A. R. Todd, J. Chem. Soc. 1949, 2476.

159 G. Schmidt, Z. physiol. Chem. 179, 243 (1928).

164

J. BADDILEY

the same rate during alkaline hydrolysisi«« and adenosine-5'-phosphate
gives adenosine when dephosphorylated by alkaline phosphatase. ^^i ^he
presence of an unsubstituted m-l,2-glycol structure in adenosine-5'-phos-
phate is indicated by the ready formation of its boric acid complex.'^^

f

N

— 0

OH OH
I I
•- CH • CH • CH • CH • CHs • OPO^Hj

NHj

. N

XXII

Muscle adenylic acid

c. Guanosine-5' -phosphate {XXIII)

It is now known that guanosine-5'-phosphate is produced during enzymic
hydrolysis of ribonucleic acid.'^* Although it is converted into guanosine
through the action of a 5'-phosphatase, it is different from the isomeric
guanylic acids a and h. The structure of this nucleotide was established by
direct comparison with synthetic guanosine-5'-phosphate.

-0-

•OPO3H,

N

OH OH
I I
L-CH-CH-CH-CH-CHs

OH

XXIII
G uanosine-5'-phosphate

d. Cytidine-S' -phosphate (XXIV)

Enzymic hydrolysates of ribonucleic acid may also contain cytidine-5'-
phosphate.125 Similar structural considerations apply to this nucleotide as

-O-

0<

N

OH OH
I I
•-CH • CH • CH • CH • CH2 • OPO3H2

N.

NH2

XXIV

Cytidine-5'-phosphate

IS" G. Embden and G. Schmidt, Z. physiol. Chem. 181, 130 (1929).
1" J. M. Gulland and E. R. Holiday, /. Chem. Soc. 1936, 765.
1" R. Klimek and J. K. Parnas, Biochem. Z. 292, 356 (1937).

CHEMISTRY OF NUCLEOSIDES AND NUCLEOTIDES 165

for giianosine-5'-phosphate, and its identity with synthetic cytidine-5'-
phosphate has been established.

e. Uridine-5' -phosphate (XXV)

Although uridine-5'-phosphate may be formed by enzymic hydrolysis of
ribonucleic acid'^^ it is also present as a structural unit in the uridme di-
phosphate coenzymes. It is hydrolyzed either enzymically or chemically to
uridine and consumes 1 mol. periodate, so it must be a 5'-phosphate.
It is identical with synthetic uridine-5'-phosphate.

-0-

OH OH

CH-CH-CH-CH-CH.-OPOjH^

I

N^
OH

XXV

Uridine-5'-phosphate

/ Adenylic Acids a and b

When ribonucleic acid is subjected to cautious hydrolysis, a material
known as "yeast adenylic acid" is obtained, together with other nucleotides
and nucleosides.!" • 1^^' ^"^ The early workers considered that this was a
single substance, but recent work has shown that two isomeric adenylic
acids (a and h) are formed by hydrolysis of ribonucleic acid, and it is not
quite certain which isomer was under investigation at each stage of the
initial structural work. In fact, it is probable that both isomers were present
on some occasions. For these reasons, and also because of the ready inter-
conversion of the isomers under mild acid conditions, the early work on the
location of the phosphate group in yeast adenylic acid is unreliable. How-
ever, it was clear that yeast adenylic acid is a monophosphate (or mixture
of monophosphates) of adenosine, since it yields adenosine and phosphoric
acid in equal amounts on hydrolysis in ammonia.i^^ Furthermore, the phos-
phate group must be situated on the sugar residue, since after treatment
with nitrous acid it gives an inosine phosphate which hydrolyzes readily to
hypoxanthine and a ribose phosphate (or mixture of ribose phosphates). i««
This "ribose phosphate" differs from ribose-5-phosphate. It would appear,
then, that yeast adenylic acid is adenosine-2'- or adenosine-3'-phosphate
(XXVI, XXVII). The ribose phosphate obtained from yeast adenylic acid

i«3 A. C. Paladini and L. F. Leioir, Biochem. J. 51, 426 (1952).
1" W. Jones and R. P. Kennedy, J. Pharmacol. 13, 45 (1919).
166 S. J. Thannhauser, Z. physiol. Chem. 107, 157 (1919).
166 p. A. Levene and S. A. Harris, J. Biol. Chem. 101, 419 (1933).

166

J. BADDILEY

PO3H2

-]-o-

P0,H2

0 OH

i I
■CH-CH'CH-CH-CH,-OH

XXVI

Adenosine-2'-phosphate

-0-

OH O
I I
"-CH-CH-CH-CH-CHs-OH

T N

NH2

XXVII

Adenosine-3'-phosphate

was believed to be identical with that from guanylic acid and gave on oxida-
tion a ribonic acid phosphate which differed in rate of lactonization and
hydrolysis from ribonic acid-S-phosphate.^®^ Furthermore, it was reduced
catalytically to an optically inactive ribitol phosphate. ^^^ If the yeast
adenylic acid in the first place was a single substance, and if no migration
of phosphate groups had occurred during subsequent hydrolysis and re-
duction, then the above transformations would indicate that the nucleotide
is adenosine-3'-phosphate yielding ribose-3-phosphate, and ribitol-3-phos-
phate.

PO3H,

-0-

OH O OH
I I I
HO-CH-CH-CH-CH-CHs

Ribose-3-phosphate

P0,H2
OH

OH

Ribonic acid-3-phosphate

OH O

I I I
H02C-CH-CH-CH-CH

POjHj

OH O OH

I I I

H0-CH2- CH-CH-CH-CH^-OH

Ribitol-3-phosphate

Since it is now known that both a and h forms of adenylic acid are produced
by hydrolysis of ribonucleic acid^^^ and also since it has been shown that
these acids are rapidly interconvertible under mild acid conditions, ^^"^ "'
the structural investigations on yeast adenylic acid are of little value.

Adenylic acids a and h are both dephosphorylated enzymically to adeno-
sine,^^^^ ^^^ are stable to periodate oxidation, '^^ and are relatively stable to
alkali. These properties agree with those expected for adenosine-2'- and

1" P. A. Levene and S. A. Harris, J. Biol. Chem. 95, 755 (1932).

i«» P. A. Levene and S. A. Harris, J. Biol. Chem. 98, 9 (1932).

»«» C. E, Carter and W. E. Cohn, Federation Proc. 8, 190 (1949).

"» D. M. Brown and A. R. Todd. J. Chem. Soc. 1952, 44.

"1 D. M. Brown, L. J. Haynes, and A. R. Todd, /. Chem. Soc. 1950, 2299.

CHEMISTRY OF NUCLEOSIDES AND NUCLEOTIDES 167

3'-phosphates^^^' "- and strong support for this view is furnished by the
synthesis of the two isomers by phosphorylation of 5'-trityladenosine with
dibenzyl phosphorochloridate, followed by hydrogenolysis of protecting
groups"" (see p. 174). Also, the ready cyclization of adenylic acids a and b
to the same adenosine-2',3'-hydrogen phosphate'^^ can only be explained in
this way. It has been shown recently'^"* that adenylic acid a may be hy-
drolyzed by short boiling in water with an ion-exchange resin (sulfonic
acid type) to ribose-2-phosphate. Under these conditions isomerization of
nucleotides and ribose phosphates does not occur to any significant ex-
tent. After similar treatment adenylic acid b gives ribose-3-phosphate. The
structure of the ribose phosphates follows from periodate oxidation of their
methyl glycosides and reduction to ribitol phosphates. Ribose-3-phosphate
is reduced to an optically inactive ribitol-3-phosphate, whereas the 2-phos-
phate gives an optically active substance, the activity of which may be
enhanced by formation of a boric acid complex. It follows that adenylic
acid a is adenosine-2'-phosphate and adenylic acid b is adenosine-3'-phos-
phate.

Adenosine-2',5'-diphosphate is obtained by enzymic hydrolysis of tri-
phosphopyridine nucleotide (p. 182) and the 3',5'-diphosphate is produced
in a similar manner from coenzyme A (p. 186).

PO3H2
-\ — 0

O OH OH
I I I
^^ „ MeO • CH • CH • CH • CH • CH2

I Q I (consumes 1 mol. lOr)

0 OH

-CH-CH-CH-CH-CHj-OH pQ^H^

/

N. ^N

^ r >

-o-

0 OH OH

NH2 ^ HO-CH-CH-CH-CH-CHj

Adenylic acid a Ribose-2-phosphate

\

P0,H2

0 OH OH

1 I I
H0-CH2- CH-CH-CH-CHrOH

Ribitol-2-phosphate

"2 D. M. Brown and A. R. Todd, /. Chem. Soc. 1952, 52.
1" D. M. Brown, D. I. Magrath, and A. R. Todd, J. Chem. Soc. 1952, 2708.
"* J. X. Khym, D. G. Doherty, E. Volkin, and W. E. Cohn, J. Am. Chem. Soc. 75,
1262 (1953).

168 J. BADDILEY

PO3H2

— o4— —

OH 0 OH
I I I
POH MeO-CH-CH-CH-CH-CHj

-0-

73x12

(stable to lOr)

OH 0
I I
^CH-CH-CH-CH-CHj-OH PO3H

/

31.1.2

N^^N

'S>

-0-

OH O OH

NH, ^ HO-CH-CH-CH-CH-CHj

Adenylic acid b Ribose-3-phosphate

\

PO3H2

OH O OH
I ! I
H0-CH2'CH-CH-CH-CHj-0H

Ribitol-3-phosphate

g. Guanylic Acids a and b

A substance known as guanylic acid was first isolated in 1898.^^^ Although
at first there was much controversy concerning its occurrence and composi-
tion, its existence was fully established eventually by Steudel.^^^ It can be
isolated in crystaUine form by hydrolysis of ribonucleic acid,^^' "^ and it
was believed that guanylic acid from different sources was a single sub-
stance.^^^ On treatment with phosphatase^" or by alkaline hydrolysis"^-'^^
guanylic acid is converted into guanosine, and with nitrous acid it is de-
aminated to xanthylic acid.^"- 1^2, iss j^ follows that the phosphate group is
situated on the sugar residue of guanosine. The ribose phosphate formed on
hydrolysis of xanthylic acid^^^' ^^^ -^y^s alleged to be identical with that
obtained from yeast adenylic acid, consequently guanylic acid was thought
to be guanosine-3 '-phosphate. However, it is now known that the guanylic
acid obtained by hydrolysis of ribonucleic acid is a mixture of a and h
isomers, ^^''' ^^^ and it is assumed, by analogy with the adenylic acids, that

"6 I. Bang, Z. phijsiol. Chern. 26, 133 (1898).

"» H. Steudel, Z. physiol. Chern. 53, 539 (1907).

1" P. A. Levene, J. Biol. Chern. 40, 171 (1919).

"« P. A. Levene and E. Jorpes, J. Biol. Chern. 81, 575 (1929).

"9 P. A. Levene and W. A. Jacobs, Ber. 42, 2469 (1909).

180 P. A. Levene and W. A. Jacobs, Ber. 42, 2474 (1909).

181 P. A. Levene and W. A. Jacobs, Biochem. Z. 28, 127 (1910).

»82 p. A. Levene and A. Dmochowski, /. Biol. Chern. 93, 563 (1931).
1" M. Knopf, Z. physiol. Chern. 92, 159 (1914).
18^ W. E. Cohn, J. Am. Chern. Soc. 72, 1471 (1950).

CHEMISTRY OF NUCLEOSIDES AND NUCLEOTIDES

169

these are 2'- and 3'-phosphates, but not necessarily respectively (XXVIII,

XXIX).

P03H,

+0-

POjH.

O OH
I I
■-CH-CH-CH-CH-CHj-OH

-CH-(

-0-

XXVIII

Guanosine-2'-phosphate

h. Cyiidylic Acids a and b

OH 0
I I
■-CH-CH-CH-CH-CH^-OH

I

H^Ni**

XXIX

Guanosine-3'-phosphate

As the pyrimidine nucleotides are more stable than, purine nucleotides
towards acids, they may be prepared by hydrolysis of ribonucleic acid with
dilute sulfuric acid.'*^' ^^^ By fractional crystallization of the barium salts
of the resulting pyrimidine nucleotide mixture, materials known as cytidylic
acid and uridylic acid are obtained. ^^' '*^' ^** Cytidylic acid yields cytidine
on neutral hydrolysis.'** In view of the stability of the pyrimidine nucleo-
tides it is not possible to hydrolyze them to free pyrimidines and ribose
phosphates. However, cytidylic acid may be reduced catalytically to dihy-
drocytidylic acid which, like the corresponding nucleoside dihydrocytosine,
is readily hydrolyzed by dilute acids to the dihydropyrimidine."^ Whereas
the phosphate residue in cytidylic acid is rather resistant towards hydroly-
sis, that in dihydrocytidylic acid has the same order of stability as the
phosphate residue in yeast adenylic acid.

Recently it has been shown that cytidylic acid from ribonucleic acid is a
mixture of a and h isomers. '^^''' '^'^ These may be separated by ion-exchange
chromatography or by careful fractionation of their salts, and, since they
are readily interconvertible in acid solution,'''" do not consume periodate,
and are both formed by alkaline hydrolysis of synthetic cytidine-2',3'-
hydrogen phosphate,'" they must be 2'- and 3'-phosphate (XXX, XXXI).
No decision has yet Vjeen made as to which is the 2'- and which is the
3'-phosphate on purely chemical grounds. However, careful measurements
of physical properties, e.g., spectra, dissociation constants, etc, suggest that
cytidytic acid a is the 2'-phosphate and cytidylic acid h the 3'-phos-
phate.'«9-""

185 p. A. Levene, Biochem. Z. 17, 120 (1909).
i8« P. A. Levene and W. A. Jacobs, Ber. 44, 1027 (1911).
»" P. A. Levene, Proc. Soc. Exptl. Biol. Med. 15, 21 (1917).
188 P. A. Levene, J. Biol. Chem. 41, 1 (1920).

170

J. BADDILEY

PO3H2

-J-0 —

P03H,

o<

O OH

CH-CH-CH-CH-CH2-0H

U

04-

OH 0
I I
i-CH-CH-CH-CH-CHj-OH

N,
NH2

XXXI

Cytidine-3'-phosphate

NH2

XXX

Cytidine-2'-phosphate

i. Uridylic Acids a and b

The substance known as uridylic acid was isolated, along with cytidylic
acid, from ribonucleic acid hydrolysates.^*^' 188-192 Ljj^g cytidylic acid, it can
be hydrolyzed to the nucleoside and phosphoric acid. Also, dihydrouridylic
acid yields uracil, ribose, and phosphoric acid on acid hydrolysis. Uridylic
acid may be separated into the two isomers, uridylic acid a and uridylic
acid 6^^° and it has been shown that cytidylic acid h is deaminated under
alkaline conditions to uridylic acid h}^^ Since the conditions of deamination
do not allow phosphate migration, it is concluded that uridylic acid a
corresponds to cytidylic acid a with respect to the position of the phosphate
group, and uridylic acid b corresponds to cytidylic acid b. Uridylic acids a
and b are probably the 2'- and 3'-phosphates of uridine, respectively
(XXXII, XXXIII).

PO3H2

-fo—

P03H2

0 OH

'-CH-CH-CH-CH-CH2-0H

o=r

OH

xxxn

Uridine-2'-phosphate

-0-

OH 0

LCH • CH • CH • CH • CH; • OH

i

o=r

OH

xxxni

Uridine-3'-phosphate

j. Cyclic Phosphates

Among the products of the action of ribonuclease on ribonucleic acid are
pyrimidine nucleotides which differ from any of those described so far.^^®' "^

'»5 L. F. Cavalieri, /. Am. Chem. Soc. 74, 5804 (1952).

ISO J. J. Fox and L. F. Cavalieri, Federation Proc. 12, 204 (1953).

1" R. J. C. Harris, S. F. D. Orr, E. M. F. Roe, and J. F. Thomas, J. Chem. Soc. 1953,

489.
192 S. J. Thannhauser, Z. physiol. Chem. 100, 121 (1917).
1" D. M. Brown, C. A. Dekker, and A. R. Todd, J. Chem. Soc. 1952, 2715.

CHEMISTRY OF NUCLEOSIDES AND NUCLEOTIDES

171

They show the properties of diesters of phosphoric acid (ion-exchange and
electrophoresis) and are converted by acids or further action of ribonuclease
into mixtures of the a and b isomers.^''^' '^^ These properties are consitent
with those expected for nucIeoside-2',3'-hydrogen phosphates, and this has
been estabhshed by direct comparison of the natural substances with syn-
thetic cytidine-2',3'-hydrogen phosphate (XXXIV) and uridine-2',3'-
hydrogen phosphate."^

■0-

0 OH

\/
P

1 I

-CH-CH-CH-CH-CHrOH

-0-

0 OH

^/
P

I I

L-CH-CH-CH-CH-CH,-OH

N.

N
NH2

XXXIV

Cytidine-2',3'-hydrogen phosphate

N. ^N

NH2

N

XXXV

Adenosme-2',3'-hjdrogen phosphate

CycUc phosphates of adenosine (XXXV) and guanosine are produced, in
addition to the cyclic pyrimidine nucleotides, by very mild alkaline hy-
drolysis of ribonucleic acid.^^^ The cyclic purine nucleotides, although read-
ily hydrolyzed by acids to the a and b isomers, are not attacked by ribo-
nuclease.'^^ The structure of the adenine nucleotide is fully established by
comparison with synthetic adenosine-2',3'-hydrogen phosphate.

k. Deoxyribosc Nucleotides

As mentioned before, enzymic or chemical hydrolysis of deoxyribonucleic
acid gives pyrimidine deoxyribonucleotides, whereas purine deoxyribonu-
cleotides can only be prepared enzymically. Four monophosphates of the
deoxynucleosides have been isolated, namely: thymidylic acid, cytosine
deoxyriboside phosphate, adenine deoxyriboside phosphate (XXXVI), and
guanine deoxyriboside phosphate. "''■'^^ In addition a fifth, 5-methylcytosine
deoxyriboside phosphate, has been demonstrated in deoxyribonucleic acid
hydrolysates by ion-exchange methods.''^ Until recently the phosphate
group in the deoxynucleotides was thought to be situated at position 3' in
the sugar chain. In the deoxynucleotides isomerism of the type shown by
the ribonucleoside-2'- and -3'-phosphates is not possible and by analogy
with these ribonucleotides the 3'-position was thought to be phosphorylated.
The isolation of 5'-phosphates from ribonuclease digests of ribonucleic acid
invalidates these earlier assumptions. Furthermore, the deoxyribonucleo-

>9^ W. E. Cohn, J. Am. Chem. Soc. 73, 1539 (1951).

172

J. BADDILEY

tides are dephosphorylated by a 5'-nucleotidase/^* and they behave in the
manner expected for 5'-phosphates on ion-exchange columns. '^^ [Cf. Cohn,
Chapter 6.]

The thymidine nucleotide has been shown to possess the structure thy-
midine-5'-phosphate (XXXVII) by synthesis^^^ (see p. 175).

OH
I
•-CH-CHi-CH-CH-CHi-OPOsHa

I

-0-

nJC >

NH,

XXXVI

Adenine deoxyriboside-5'-phosphate

OH
I
■-CH-CHs-CH-CH-CHs-OPOjHj

N^Me
OH

XXXVII

Thymidine-S'-phosphate

I. Deoxynucleoside Diphosphates

Acid hydrolysis of deoxyribonucleic acid gives, in addition to pyrimidine
deoxyriboside phosphates, varying amounts of diphosphates of thymidine
and cytosine deoxyriboside.^*""^^^ Chromatographic evidence also suggests
the presence in the hydrolysate of traces of the diphosphate of 5-methyl-
cytosine deoxyriboside."* Catalytic hydrogenation of thymidine diphos-
phate, followed by mild acid hydrolysis, gives dihydrothymidine and a
reducing sugar diphosphate. These diphosphates are the 3' , 5'-diphosphates
of thymidine (XXXVIII) and cytosine deoxyriboside (XXXIX) as is
shown by their synthesis from thymidine and cytosine deoxyriboside by
direct phosphorylation."*

-0

PO3H2

0

l-CH-CH.-CH-CH-CHiOPOaHj

N^Me
OH

XXXVIII

Thymidine-3', 5'- diphosphate

-0-

PO3H2

0

I I
L-CH-CHrCH-CH-CHj OPO3H2

0=< 1

NH2

XXXIX

Cytosine deoxynbosidehS' ,5'- diphosphate

3. Synthesis of Nucleotides

The formulas assigned to the nucleotides are supported, in some cases,
by synthesis. At the present time unambiguous syntheses exist for the

•" C. E. Carter, J. Am. Chem. Soc. 73, 1537 (1951).

136 E. Volkin, J. X. Khym, and W. E. Cohn, J. Am. Chem. Soc. 73, 1535 (1951).

1" A. M. Michelson and A. R. Todd, J. Chem. Soc. 1953, 951.

CHEMISTRY OF NUCLEOSIDES AND NUCLEOTIDES

173

nucleoside-5'-phosphates, the cyclic phosphates, and the pyrimidine deoxy-
ribonucleoside diphosphates. Successful methods employ unequivocal pro-
tection of appropriate hydroxyl groups on the sugar residue, then phos-
phorylation and removal of protecting groups.

a. The 5' -Phosphates

The most convenient derivatives for the synthesis of ribonucleoside-5'-
phosphates are the 2',3'-isopropylidene compounds. For example inosine-
5'-phosphate was first synthesized from 2',3'-isopropylidene inosine by
phosphorylation with phosphoryl chloride in pyridine, then removal of the
isopropylidene residue by cautious acid hydrolysis.^''

OH

— 0 —

MeCMe

/\

0 0

CH-CH-CH-CH-CHrOH

N

-0-

OH

MeCMe

/\
O O

L-CH-CH-CH-CH-CH.-OPOsH,

I

N

A synthesis of adenosine-5'-phosphate from 2', 3 '-isopropylidene adeno-
sine follows similar lines. ^^ The overall yield in this synthesis is low, how-
ever, and the final product difficult to purify. Similarly, phosphorylation
of 2',3'-diacetyladenosine with phosphoryl chloride, then removal of acetyl
groups with alkali, gives only low yields of adenosine-5'-phosphate.'**

The difficulties in these syntheses are probably associated with the nature
of the phosphorylating agent. The simultaneous formation of mono-, di-,
and triesters and subsequent difficulties arising in the hydrolysis of the
phosphorochloridates may be avoided by use of dibenzyl phosphorochlori-
date.^^* Thus, 2',3'-isopropylidene adenosine reacts readily with dibenzyl
phosphorochloridate in pyridine to give the 5'-dibenzyl phosphate. Benzyl

-O-

MeCMe

/\
0 0 OCH.Ph

I
'-CH-CH-CH-CH-CH.-OP = 0
I ■ I

N. ^N\ OCHaPh

X Hj/Pd

NH,

N

NH,

O —

MeCMe

CH-CH-CH-CH-CH^-OPOjH,

I

N

2',3'-Isopropylidene adenosine-5'-
dibenzyl phosphate

'^8 F. R. Atherton, H. T. Openshaw, and A. R. Todd, /. Chem. Soc. 1945, 382.

174 J. BADDILEY

groups are removed by catalytic hydrogenation and the isopropylidene
residue by acid hydrolysis as before. A good yield of adenosine-5'-phosphate
is obtained in this way.^'^

Phosphorylation of 2' ,3'-isopropylidene cytidine and 2' ,3'-isopropylidene
uridine with dibenzyl phosphorochloridate, followed by hydrogenation and
hydrolysis, gives cytidine-5'-phosphate and uridine-5'-phosphate.^^^ Uri-
dine-5'-phosphate has also been prepared by phosphorylation of 2',3'-iso-
propylidene uridine with phosphoryl chloride.^^ Guanosine-5'-phosphate
may be prepared from 2',3'-isopropylidene guanosine and phosphoryl
chloride but not from dibenzyl or diphenyl phosphorochloridate.'^^

b. The a and b Isomers

Syntheses of the 3'-phosphates of adenosine, guanosine, cytidine, and
uridine have been reported. These syntheses, however, are either ambiguous
or based on erroneous assumptions concerning the structure of intermedi-
ates. The ready interconversion of the a and b series and the unsuspected
difficulty in characterizing the products at the time invalidates much of this
work. Direct phosphorylation under various conditions with phosphoryl
chloride of adenosine,^"'' guanosine, and uridine^"' gives products which
were claimed to be identical with the nucleotides obtained from ribonu-
cleic acid. Similarly, phosphorylation of 5'-trityluridine and 5'-tritylcyti-
dine,^^' ~^- followed by removal of protecting groups, was thought to give
the 3'-phosphates. These syntheses are ambiguous, and it has been shown in
the case of adenosine that phosphorylation of the 5'-trityl compound leads
to mixtures of a and b nucleotides. '^°

Although unambiguous syntheses of the 2'- and 3 '-phosphates of purine
and pyrimidine nucleosides have been claimed, '^^' 203-205 j^ jg j^q^ known
that the compounds obtained were 5'-phosphates.'^'' 2"® These syntheses
were based on benzylidene nucleosides, which were believed to be 3',5'-
benzylidene compounds^^' ^"^ but which have since been shown to be
2',3'-compounds.

c. The Cyclic Phosphates

The four cyclic phosphates obtained by cautious alkaline hydrolysis of
ribonucleic acid are prepared from the appropriate a or 6 nucleotides by

199 J. Baddiley and A. R. Todd, J. Chem. Soc. 1947, 648.

^"0 G. R. Barker and J. M. Gulland, J. Chem. Soc. 1942, 231.

2o> J. M. Gulland and G. I. Hobday, J. Chem. Soc. 1940, 746.

202 H. Bredereck, Z. physiol. Chem. 224, 79 (1934).

^"^ J. M. Gulland and H. Smith, J. Chem. Soc. 1947, 338.

2»4 J. M. Gulland and H. Smith, /. Chem. Soc. 1948, 1527.

2" J. M. Gulland and H. Smith, J. Chem. Soc. 1948, 1532.

206 D. M. Brown, L. J. Haynes, and A. R. Todd, J. Chem. Soc. 1950, 408.

^o-' J. M. Gulland and W. G. Overend, /. Chem. Soc. 1948, 1380.

CHEMISTRY OF NUCLEOSIDES AND NUCLEOTIDES

175

reaction with trifluoroacetic anhydride.'" Both isomers give the same
product, and it is beheved that an intermediate mixed anhydride is formed
between the phosphate group of the starting nucleotide and a trifluoroacetyl
group. The anhydride then effects internal phosphorylation with elimina-
tion of the trifluoroacetyl group and production of a cyclic phosphate. A
cyclic phosphate of theophylline glucoside has been prepared by direct
phosphorylation with phosphoryl chloride in pyridine.-"* Its constitution
has not been established.

PO3H,

-O-

OH O
I I
•-CH-CH-CH-CH-CH.-OH

NH, ^

■0-

0 OH

W/
P

/\
O 0

^CH-CH-CH-CH-CH^-OH

I

NH2

Adenylic acid b

d. Deoxyribose Nucleotides

Thymidine-5'-phosphate and the luinatural 3'-phosphate are synthesized
from 5'-tritylthymidine.'^^ Direct phosphorylation of the trityl compound
with dibenzyl phosphorochloridate followed by treatment with hot acetic
acid gives thymidine-3'-benzyl phosphate, from which the benzyl group is
removed catalytically.

■0-

OH
I
•-CH-CHrCH-CH-CHrOTri

N^Me
OH

H0-P0(0CH2Phj

-o4-

N.

0
I
•-CH-CHj-CH-CH-CHs-OH

I

N

OH

PO:H,

-0-

O
I

l-CH-CH^-CH-CH-CHj-OH
I
N

OH

"8 E. Fischer, Ber. 47, 3193 (1914).

176

J. BADDILEY

Acetylation of 5'-tritylthymidine gives 3'-acetyl-5'-tritylthymidine, which
is readily converted into 3'-acetylthymidine by short heating with acetic
acid. Phosphorylation of this acetyl derivative with dibenzyl phosphoro-
chloridate, followed by removal of protecting groups, gives thymidine-5'-
phosphate.

-0-

AcO
I
CHCHj-CH-CH-CHrOTri

I

N^Me
OH

AcO
I
LCHCHi-CH-CH-CHj-OH

I

N,

■* ^=< II

N^Me

OH

OH
I
^CH-CH^-CH-CH-CHj-OPOsHj

OH

The 3',5'-diphosphates of thymidine and cytosine deoxyriboside are pre-
pared by direct phosphorylation of the nucleosides with an excess of diben-
zyl phosphorochloridate, then removal of benzyl groups by catalytic hy-
drogenation.^**

e. Miscellaneous

Several unnatural nucleotides have been synthesized. Thus adenosine-
5'-phenylphosphonate, adenosine-5'-ethylphosphonate, uridine-5'-phenyl-
phosphonate, and uridine-5'-ethylphosphonate are obtained from the
appropriate benzyl aryl (or alkyl) phosphonochloridate and the isopropyli-
dene derivatives of adenosine and uridine. ^"^

Two diribonucleoside phosphates have been prepared synthetically. The
first of these was given the formula diuridine-2',2''-phosphate, since it was
prepared from benzylidene uridine and phenyl phosphorodichloridate.^^^ It
is now known that the starting material was 2' , 3'-benzyhdene uridine and
not the supposed 3',5'-compound, consequently it seems probable that
the product was diuridine-5',5''-phosphate. The second member of this
group of substances, adenosine-5' uridine-5' phosphate has been prepared
by an unequivocal route.^^" 2',3'-Isopropylidene adenosine-5'-dibenzyl
phosphate (formula, p. 173) behave^ typically as a neutral benzyl ester of
phosphoric acid in that it loses one benzyl group on being heated with a

209 N. Anand and A. R. Todd, /. Chem. Soc. 1951, 1867.

210 D. T. Elmore and A. R. Todd, /. Chem. Soc. 1952, 3681.

CHEMISTRY OF NUCLEOSIDES AND NUCLEOTIDES

177

■0-

NH2 ^^

MeCMe

/\
O 0 OCH.Ph

I

l-CH-CH-CH-CH-CHrOP-OAg
II
O

-o-

MeCMe

LCH-CH-CH-CH-CHs-I
OH

(1) boiling toluene

(2) Huso,

NH,

k

t.y

OH

N^^N

N

o=k

■OH
-OH

■OH
•OH

CHrO-

OH

I
■P O-CH.

II
0

Adenosine-5' uridine-5' phosphate

tertiary base.^^' The resulting 2',3'-isopropylidene adenosine-5'-benzyl
phosphate is converted into its silver salt which in turn reacts with 2',3'-
isopropylidene-5'-iodo-5'-deoxyuridine. The isopropylidene and benzyl
groups were removed from the resulting neutral ester by gentle acid hy-
drolysis.

4. Properties of Nucleotides

The nucleotides resemble the nucleosides in many of their physical prop-
erties. They are colorless and for the most part crystalline solids with high
melting points, frequently decomposing before melting. They are strong
acids, soluble in water and insoluble in organic solvents. Their physical
properties (Table II) are often rather unreliable for characterization pur-
poses, and this fact was largely responsible for the failure of the early in-
vestigators to recognize the multiple nature of some of their materials. The
most satisfactory methods for the identication of nucleotides are those em-
ploying ion-exchange and paper-chromatographic analysis, often combined
with ultraviolet spectroscopy. [Cf. Chapters 6, 7, and 14.]

5. Nucleotide Coenzymes

Nucleotides of adenine and uridine occur as structural units not only of
the nucleic acids but also of certain coenzymes. The widespread occurrence

211 J. Baddiley, V. M. Clark, J. J. Michalski, and A. R. Todd, J. Chem. Soc, 1949, 815.

178

J. BADDILEY

TABLE II
Properties of Nucleotides

Nucleotide

M.p., °C.

Inosine-5' -phosphate
(muscle inosinic
acid)

Adenosine-5' -phos-
phate (muscle ad-
enylic acid)

Guanosine-5' -phos-
phate

Cytidine-5' -phos-
phate

Uridine-5' -phosphate

Adenylic acid a

Adenylic acid b

Cytidylic acid a

Cytidylic acid b
Uridylic acid a
Uridylic acid b

Thymidine -5'-phos-
phate

Cytosine deoxyribo-
side-5'-phosphate

Thymidine-3', 5' -di-
phosphate

Cytosine deoxyribo-
side-3', 5' -diphos-
phate

192

190-200
(dec.)
233

197 (dec.)

238-240

(dec.)

233-234

183-187

Derivatives

— 18.5° (2.5% forms crystalline ammonium,
HCl) barium, calcium, sodium, and

potassium salts
-46.4° (H2O) acridine salt, m.p. 208°
-47.5°
(NaOH)

dibrucine salt, m.p. 210-220°
(dec. with prelim, softening)
27.1° (H2O) dibrucine salt, softens 185°, m.p.
215° (dec.)
barium salt, hexagonal plates;
dibrucine salt, softens 185°
molten at 202°, needles [a] „
— 70.4° (pyridine)
187 (dec.) -84.3° (form- dibrucine salt, m.p. 165-175°
amide)
58.7° (form-
amide)
20.7° (H2O) free nucleotide polymorphous

49.0° (H2O) free nucleotide polymorphous

lo -58.9°

-4.4'

14.4° (Ba
salt)

dibrucine salt, [a]

(pyridine)
dibrucine salt, rosettes of

needles, softening at 140°,

m.p. about 175°

tetrabrucine salt, softens at
176°, m.p. 182-184°

tetrabrucine salt, needles sinter-
ing at 180°, m.p. 185°

and remarkable versatility of the nucleotide coenzymes warrants their
inclusion in this book. Although the chemical transformations which they
catalyze are of fundamental importance in all living cells, they are so numer-
ous and intimately interrelated that it is not possible to consider them here
and the following sections will be devoted entirely to their chemistry. It
will be seen that a common feature of the adenosine and uridine coenzymes
is the presence in the molecule of a pyrophosphate linkage.

CHEMISTRY OF NUCLEOSIDES AND NUCLEOTIDES

179

a. Adenosine Di- and Triphosphate (ADP and ATP)

These two substances participate in the reversible phosphorylation of a
very large number of important intermediates in metabolic processes. ATP
is usually isolated from muscle2i2-2i4 q^ after enzymic synthesis from adeno-
sine.2i* Two of its three phosphate groups are removed by acid hydrolysis,
the other products being adenine and ribose-5-phosphate. In alkali it gives
adenosine-5'-pho&phate and pyrophosphoric acid.^'^ Titration with alkali
indicates the presence of three primary and one secondary phosphate acidic
groups, suggesting a linear triphosphate. Positions 2' and 3' are unsubsti-
tuted since ATP increases the conductivity of boric acid and consumes 1
mol. periodate.2i^ These observations indicate the structure XL.

O —

OH OH OH OH

L-CH-CH-CH-CH-CHrO-P-O-P-OH

I II II

N. -N 0 0

NH.

XLI

Adenosine diphosphate

-O-

OH OH OH OH OH

l-CH-CH-CH-CH-CH^-O-P-O-P-O-P-OH

N. .N

0 0 0

NHj

N'

XL

Adenosine triphosphate

ADP (XLI) may be prepared by enzymic dephosphorylation of ATP^'s. 219
with the loss of one phosphate group. The presence of a pyrophosphate
residue is supported by titration, and the formulas of both these coenzymes
have been confirmed by synthesis.

212 K. Lohmann, Naturwissenschaften 17, 624 (1929).

213 C. H. Fiske and Y. Subbarow, Science 70, 381 (1929).
21* G. A. LePage, Biochem. Preparations 1, 5 (1949).

216 p. Ostern, T. Baranowski, and J. Terszakowec, Z. physiol. Chem. 251, 258 (1938).

216 K. Lohmann, Biochem. Z. 233, 460 (1931).

217 B. Lythgoe and A. R. Todd, Nature 155, 695 (1945).

218 K. Lohmann, Biochem. Z. 282, 104 (1935).

219 K. Lohmann, Biochem,. Z. 282, 120 (1935).

180

J. BADDILEY

In the synthesis of ADPi^^ adenosine-5'-benzyl phosphate is obtained by
simultaneous removal in acid of one benzyl and the isopropylidene residue
from 2',3'-isopropylidene adenosine-5'-dibenzyl phosphate (formula, p.
173). The silver salt of this monobenzyl ester, with dibenzyl phosphoro-
chloridate, gives adenosine-5'-tribenzyl pyrophosphate. Catalytic hydro-
genation of this gives ADP.

-0-

OH OH

OCH^Ph

f

N

N,

NH,

"-CH-CH-CH-CH-CH.-O-P-OH

I II

N 0

-0-

ADP

-, /Hs/Pd

OH OH PhCH,0 OCHjPh

II II

l-CH-CH-CH-CH-CH^-O-P-O-P-OCH^Ph

I II II

N. 0 0

NH,

T N

Adenosine-5'-tnbenzyl pyropnosphate

-0-

OH OH

f

N,

NH,

CH

I

N.

PhCH^O OCHjPh

CH-CH-CH-CHj-O-P-O-P-OHl

II II
0 0

-0

ATP

'Hj/Pd

OH OH

OCHjPh

I OCHjPh

f

N,

PhCH.O
I
^CH-CH-CH-CH-CHrO-P-O-P-O-P-OCHsPh

I II II II

N. 0 0 0

T N

NHs

When adenosine-5'-tribenzyl pyrophosphate is heated with a teritary
base-'^ one benzyl group is lost. The resulting dibenzyl ester, as its silver
salt, is phosphorylated wdth a further mol. of dibenzyl phosphorochloridate
giving a tetrabenzyl ester of ATP, from which benzyl groups may be re-
moved readily by catalytic hydrogenation.^^o -pj^g synthesis has been simpli-
fied by direct phosphorylation of the disilver salt of adenosine-5'-phosphate
with 2 mol. dibenzyl phosphorochloridate, and then removal of benzyl
groups as before. An unstable cyclic intermediate is thought to be formed in
this synthesis.^-^

"0 J. Baddiley, A. M. IVIichelson, and A. R. Todd, /. Chem. Soc. 1949, 582.
"1 A. M. Michelson and A. R. Todd, J. Chem. Soc. 1949, 2487.

CHEMISTRY OF NUCLEOSIDES AND NUCLEOTIDES 181

b. Di- and Triphosphopyridine Nucleotides {DPN and TPN)

The hydrogen-transporting coenzyme, DPN, was called cozymase when
first discovered'-" and is sometimes knoAMi as coenzyme I or codehydro-
genase I. The closely related TPN"-^ differs from it only in possessing an
extra phosphate group.-'' On acid hydrolysis DPN gives adenine (1 mol.),
nicotinamide (1 mol.), pentose (2 mol.), and phosphoric acid (2 mol.).--^"-^"
The amino group in the adenine residue of both coenzymes is unsubstituted,
since they are dcaminated with nitrous acid to the hydroxy analogues with-
out loss of other groups. They are easily reduced to dihydro derivatives
which, unlike the coenzymes themselves, are very labile towards acids but
relatively stable to alkali. These properties, together with the accompany-
ing changes in ultraviolet spectra, are typical for ciuaternary pyridinium
compounds, and it has been shown that synthetic A^-glycosides of nico-
tinamide are very similar in properties to DPN and TPN.^^'-^^ Furthermore,
enzymic hydrolysis of DPN gives A-(D-ribofuranosyl)nicotinamide-^''"-"
(XLII) which, in its dihydro form, is identical with the synthetic glycoside
prepared from dihydronicotinamide and acetobromoribofuranose.-^^ Only
one sugar phosphate (ribose-5-phosphate) is formed by hydrolysis of DPN,
and a pyrophosphate linkage is present as shown by titration and the isola-
tion from a hydrolysate of adenosine-5'-pyrophosphate.^^^ These facts are
consistent with formula XLIII for DPN.^""

The structures proposed for these coenzymes are strongly supported by
enzymic hydrolysis studies.-^' • -"- The pyrophosphate linkage of DPN is

=" A. Harden and W. J. Young, Proc. Royal Hoc. (London) B77, 405 (1906).

"3 O. Warburg and W. Christian, Riochem. Z. 254, 238 (1932).

"4 O. Warburg, W. Christian, and A. Griese, Biochem. Z. 282, 157 (1935).

"5 H. von Euler and K. Myrbiick, Z. physiol. 198, 236 (1931).

"6 H. von Euler and K. Myrback, Z. physiol. Chem. 203, 143 (1931).

2" H. von Euler and K. Myrback, Z. physiol. Chem. 212, 7 (1932).

"8 H. von Euler and K. Myrback, Z. physiol. Chem. 233, 95 (1935).

2" H. von Euler, H. Albers, and F. Schlenk, Z. physiol. Chem. 237, 1 (1935).

"0 H. von Euler, H. Albers, and F. Schlenk, Z. physiol. Chem. 240, 113 (1936).

"' P. Karrer, G. Schwarzenbach, F. Benz, and U. V. Solmssen, Helv. Chim. Acta 19,

811 (1936).
"2 H. von Euler, P. Karrer, and B. Becker, Helv. Chim. Acta 19, 1060 (1936).
"3 p. Karrer, B. H. Ringier, J. Biichi, H. Fritzsche, and U. Solmssen, Helv. Chim.

Acta 20, 55 (1937).
"< F. Schlenk, Svensk Vet. Akad. Arkiv Kemi 12B, 17 (1936).
"6 F. Schlenk, Svensk Vet Akad. Arkiv Kemi 14A, 13 (1941).
"« F. Schlenk, Naturwissenschaften 28, 46 (1940).
"' F. Schlenk, Arch. Biochem. 3, 93 (1943).
"» L. J. Haynes and A. R. Todd, J. Chem. Soc. 1950, 303.
"9 R. Vestin, F. Schlenk, and H. von Euler, Ber. 70, 1369 (1937).
"" H. von Euler and F. Schlenk, Z. physiol. Chem. 246, 64 (1937).
2" A. Kornberg, J. Biol. Chem. 182, 779 (1950).
"= A. Kornberg, J. Biol. Chem. 182, 805 (1950).

182

J. BADDILEY

CO-NHj

N'

-OH
-OH

CHj-OH

XLII

Nicotinamide riboside

Q

CO-NHo

-OH
-OH

0

N

N^N

NHj

N

-OH
-OH

0- OH
i I
CHj-O-P-O-P-O-CHs
II II
O 0
XLIII
DPN

OCO-NHj

1

N'

u

N-^N^

-OH
-OH

-PO3H,

■OH

0- OH

I I
CH2-0-P-0-P-0'CH2

II II
0 0

XLIV
TPN

POjHj

-1-0 —

0 OH

l-CH-CH-CH-CH-CHs-OPOaHj

j^N_N.

Yh.n

XLV

Adenosine-2',5'-diphosphate

hydrolyzed by a dinucleotide pyrophosphatase to give nicotinamide nucleo-
side-5'-phosphate. The coenzyme is resynthesized by this enzyme in the
presence of ATP. Also it has been shown that TPN may be synthesized by
enzymic phosphorylation of DPN. When TPN (XLIV) is hydrolyzed by
the pyrophosphatase a diphosphate of adenosine is obtained. This is adeno-
sine-2',5'-diphosphate (XLV), since it can be hydrolyzed by a specific
5'-nucleotidase to adenylic acid a (adenosine-2'-phosphate) .^''^

The coenzyme responsible for the oxidation of cysteinesulfinic acid to
cysteic acid is called coenzyme III.^*^ Although not yet isolated pure, ultra-
violet spectra of the reduced and oxidized forms of the coenzyme suggest
that it is a derivative of nicotinamide riboside. Furthermore, it is rapidly
destroyed by a pyrophosphatase; thus, it probably contains a pyrophos-
phate residue. The tentative formula, nicotinamide riboside-5'-pyrophos-
phate (XLVI) has been assigned to this substance.

2« A. Kornberg and W. E. Pricer, Jr., /. Biol. Chem. 186, 557 (1950).

2" T. P. Singer and E. B. Kearney, Biochim. et Biophys. Acta 8, 700 (1952).

CHEMISTRY OF NUCLEOSIDES AND NUCLEOTIDES
iCO-NH,

183

^N

-OH
-OH

OH 0-

I I
CHj-O-P-O-P-OH

II II
0 O

XLVI

Coenzyme III

c. Flavin Adenine Dinucleotide (FAD)

The coenzyme generally known as flavin adenine dinucleotide is not a
true dinucleotide, since one carbohydrate residue is not glycosidically
bound. It participates in oxidation-reduction reactions at a higher general
level than that for DPN. It is a derivative of riboflavin and gives adenine
and phosphoric acid on hydrolysis.^''* An unsubstituted amino group is
present since FAD can be deaminated with nitrous acid, and on careful
alkaline hydrolysis it gives adenosine-5'-phosphate. In acids it also gives
riboflavin-5'-phosphate.^''^ A formula for FAD in which the two phosphate
residue are joined together as a pyrophosphate (XLVII) is fully confirmed
by synthesis.

— O

OH OH

OH OH

-CH-CH-CH-CH-CHj-O-P-O-P-O-CHj-CH-CH-CH-CH^

-N,

NH,

0 0

XLVII
FAD

OH OH OH

Me
Me

Nv^ .N.

0

Riboflavin-5'-phosphate, which has coenzyme activity itself under certain
circumstances, has been synthesized by phosphorylation of riboflavin with
phosphoryl chloride-^^ • -^^ or from 2',3',4'-triacetylriboflavin.2^^ FAD has

2« O. Warburg and W. Christian, Biochem. Z. 298, 150 (1938).

"« E. P. Abraham, Biochem. J. 33, 543 (1939).

"" R. Kuhn and H. Rudy, Ber. 68, 383 (1935).

"« H. S. Forrest and A. R. Todd, J. Chem. Soc. 1950, 3295.

»" R. Kuhn, H. Rudy, and F. Weygand, Ber. 69, 1543 (1936).

184

J. BADDILEY

been synthesized in the following manner: A mixed anhydride of diphenyl
phosphate and benzyl hydrogen phosphite reacts with 2',3'-isopropylidene
adenosine to give the 5'-benzyl phosphite.-^" This is chlorinated with A''-
chlorosuccinimide to the phosphorochloridate,"^ which then reacts with
the monosilver salt of riboflavin-5'-phosphate. The benzyl group is lost
under the experimental conditions employed and the isopropylidene residue
removed by acid hydrolysis. From the resulting mixture pure FAD has
been isolated. "-

-0-

MeCMe

I I

l-CH • CH • CH • CH • CHj • OPH

OCH,Ph

-0-

N K O

N jC >

2',3'-Isopropylidene adenosine-5'-
benzyl phosphite

OH OH OH

MeCMe
/\
O 0 OCHjPh

I
•-CH-CH-CH-CH-CHrOPCl
I II

N. .N^ 0

NH, -^

Me
Me

FAD

CHj-CH- CH-CH-CHi-OPOsHj

0
Riboflavin-5'-phosphate

d. Coenzyme A (Co A)

The coenzyme responsible for the transfer of acetyl groups, Co A, is a
derivative of the vitamin pantothenic acid.^^^"^*^ Chemical or enzymic hy-
drolysis shows that Co A contains equimolar amounts of pantothenic acid,
adenine, pentose, and a sulfur compound^^^ subsequently identified as
2-mercaptoethylamine.^" • "^ Three phosphate groups are present.^^^ The

"0 N. S. Corby, G. W. Kenner, and A. R. Todd, J. Chem. Soc. 1952, 3669.

2" G. W. Kenner, A. R. Todd, and F. J. Weymouth, J. Chem. Soc. 1952, 3675.

"2 S. M. H. Christie, G. W. Kenner, and A. R. Todd, Nature 170, 924 (1952).

2" F. Lipmann, J. Biol. Chem. 160, 173 (1945).

2" F. Lipmann and N. O. Kaplan, J. Biol. Chem. 162, 743 (1946).

2" F. Lipmann, N. O. Kaplan, G. D. Novelli, L. C. Tuttle, and B. M. Guirard, /. Biol.

Chem. 167, Sm (1947).
"6 W. H. De Vries, W. M. Govier, J. S. Evans, J. D. Gregory, G. D. Novelli, M.

Soodak, and F. Lipmann, J. Am. Chem. Soc. 72, 4838 (1950).
"7 J. Baddiley and E. M. Thain, /. Chem. Soc. 1951, 2253.

"» J. D. Gregory, G. D. Novelli, and F. Lipmann, /. Am. Chem. Soc. 74, 854 (1952).
"9 G. D. Novelli, J. D. Gregory, R. M. Flynn, and F. J. Schmetz, Federation Proc.

10,229 (1951).

CHEMISTRY OF NUCLEOSIDES AND NUCLEOTIDES 185

close relationship between CoA and pantetheine (Lactobacillus bulgaricus
factor, LBF)-^"' -^^ is established by the conversion of the former into the
latter by enzymic digestion. Pantetheine (XLVIII) may also be converted
into CoA enzymically in the presence of ATP.-^-

OH

HO-CHj-CMes-CH-CO-NH-CHo.CHs-CO-NHCII.CH.-SH

XLVIII

Pantetheine (LBF)

The presence of a pyrophosphate linkage in CoA is indicated by its hy-
drolysis with a dinucleotide pyrophosphatase-*^^ • -^^ and supported by its
ready conversion into a cyclic phosphate of pantetheine with dilute alka-
jj 265, 266 q^j^g formation of cyclic phosphates has been observed under simi-
lar circumstances with UDPG (see below) and FAD.-^* The pyrophosphate
linkage joins the adenosine residue at position 5' with the pantetheine resi-
due at position 4'. This follows from the observation that acid hydrolysates
of CoA contain adenosine-5'-phosphate-°^ ■ "^ and pantothenic acid-4'-
phosphate (XIJX).-'''^^'' -^'' The formula for CoA is, then, represented
byL.

OH
I
H.O^PO-CHj-CMerCH-CO-NH-CHj-CHo-CO.H

XLIX

Pantothenic acid-4'-phosphate

POjHj

-o-

OH O OH OH OH

III II I

CH-CH-CH-CH-CH^-O-P-O-P-O-CHj-CMej-CH-CO-NH-CHj-CH^-CO-NH-CHrCHrSH

.N^/N. O O

NHj^

N L

Coenzyme A

The third phosphate group is a monophosphate, as shown by enzymic
hydrolysis with a monophosphatase, and periodate oxidation-^' indicates
that it is situated at positions 2' or 3' in the adenosine residue. Since this
phosphate is removed by an enzyme which hydrolyzes only 5 nucleotides,

2«« R. A. McRorie, P. M. Masley, and W. L. Williams, Arch.' Biochem. 27, 471, (1950).

"1 G. M. Brown, J. A. Craig, and E. E. Snell, Arch. Biochem. 27, 473 (1950).

"2 R. A. McRorie and W. L. Williams, J. Bacterial. 61, 737 (1951).

"3 G. D. Novelli, N. O. Kaplan, and F. Lipmann, Federation Proc. 9, 209 (1950) .

264 G.'D.Nove\U,Phosphoriis Metabolism 1, 4^(1951).

2" J. Baddiley and E. M. Thain, J. Chem. Soc. 1952, 3783.

"6 J. Baddiley and E. M. Thain, J. Chem. Soc. 1953, 903.

«" J. Baddiley and E. M. Thain, /. Chem. Soc. 1951, 3421.

186 J. BADDILEY

it must be at position 3' (h). This is confirmed by the demonstration that
one of the products of pyrophosphatase action on CoA is adenosine-3',5'-
diphosphate (LI), different from the diphosphate obtained from TPN.^^s

POjHa

-0-1-

OH 0
I I
'-CH-CH-CH-CH-CH2-OP03H2

o::>

LI

Adenosine-3' , 5'-diphosphate

OH

I
H203PO-CH2-CMe2-CH-CO-NH-CH2-CH2-CO-NH-CH2-CH2-SH

LII

Pantetheine- 4'-phosphate

Strong support for the formula assigned to CoA comes from the synthesis of
pantetheine-4'-phosphate (LII) by phosphorylation of pantetheine-0'',*S
dibenzyl ether with dibenzyl phosphorochloridate, then removal of benzyl
groups with sodium in liquid ammonia.^^^ This compound is identical with
the other product of pyrophosphatase action on the coenzyme, and is con-
verted into CoA by liver enzymes in the presence of ATP.""

The "trans-acetylation" function of CoA involves the thiol group which,
as its thiolacetate, is capable of donating an acetyl group to a variety of sub-
strates."^ CoA is also involved in the transfer of other important acyl resi-
dues in biological systems.

e. Uridine Diphosphate Glucose (UDPG)

The intercon version of glucose- 1 -phosphate and galactose- 1 -phosphate
requires an enzyme "galactowaldenase." The coenzyme for this transfor-
mation is known as uridine diphosphate glucose (UDPG)."^ It gives on
hydrolysis uridine, glucose, and 2 mol. of phosphoric acid. Titration indi-
cates the presence of two primary and no- secondary phosphate acidic groups
and cautious acid hydrolysis liberates one secondary acidic group, together
with glucose. Further hydrolysis liberates 1 mol. of inorganic phosphate.

268 T. p. Wang, L. Schuster, and N. O. Kaplan, J. Am. Chem. Soc. 74, 3204 (1952).
"9 J. Baddiley and E. M. Thain, /. Chein. Soc. 1953, 1610.

"0 J. Baddiley, E. M. Thain, G. D. Novelli, and F. Lipmann, Nature 171, 76 (1953).
"1 F. Lynen, E. Reichert, and L. Rueff, Ann. 574, 1 (1951).

"2 R. Caputto, L. F. Leloir, C. E. Cardini, and A. C. Paladini, J. Biol. Chcm. 184, 333
(1950).

CHEMISTRY OF NUCLEOSIDES AND NUCLEOTIDES

187

The uridine phosphate formed on hydrolysis consumes 1 mol. of periodate
and is identical with synthetic uridine-5'-phosphate.^^^ The product of
cautious acid hydrolysis is uridine-5'-pyrophosphate (LIII). The pyrophos-

N:

— o —

OH OH OH OH

II , ,

'-CH-CH-CH-CH-CHj-O-P-O-P-OH

I II II

N^ 0 0

-0-

N

MeCMe

/\
0 O OCH^Ph

I
LcH-CH-CH-CH-CH,-0-P-Cl
I II

N^ 0

OH

LHI

Uridine- 5'-pyrophosphate

OH

LIV

2',3'-Isopropj'lidene uridine-5'-
benzylphospliorochioridate

phate has been synthesized from 2',3'-isopropylidene-5'-iodo-5'-deoxyuri-
dine (formula, p. 177) by condensing the latter with silver tribenzyl pyro-
phosphate, then removing protecting groups.^''^ It may also be prepared
from 2',3'-isopropylidene uridine-5'-benzyl phosphorochloridate (LIV) and
dibenzyl phosphate.^^"

The formula LV is established for UDPG :

-O-

OH OH

OH OH

OH

OH

^CH-CH-CH-CH-CH>-0-P-0-P-0-CH-CH-CH-CH-CH-CH,-OH

I ' II II I

N 0 O OH

o=r ■

OH

LV
UDPG

The linkage between the pyrophosphate and glucose residues must be a
since cautious alkaline hydrolysis gives glucose-1 ,2-hydrogen phosphate,
which can only be formed from an a- 1 -phosphate.

Several other uridine diphosphate coenzymes have been encountered in
Nature. Staphylococcus aureus contains such substances bearing an amino
sugar and amino acids in the place of glucose.-^^ In addition to the galactose
analogue of UDPG there is some evidence for the presence in Nature of
UDP derivatives of other sugars.^®^- "^' "®

"3 N. Anand. V. M. Clark, R. H. Hall, and A. R. Todd, /. Chein. Soc. 1952, 3665.

"^ J. T. Park, /. Biol. Chem. 194, 877, 885, 897 (1952).

"^ J. G. Buchanan, J. A. Bassham, A. A. Benson, D. F. Bradley, M. Calvin, L. L.

Daus, M. Goodman, P. M. Hayes, V. H. Lynch, L. T. Norris, and A. T. Wilson,

Phosphorus Metabolism 2, 440 (1952).
"• G. J. Dutton and I. D. E. Storey, Biochem. J. 53, XXXVII (1953).

188 J. BADDILEY

III. Addendum

Introduction

Bacteriophage nucleic acid yields 5-hydroxymethylcytosine on hydroly-
sis.^^^ Cytosine is not present.

Miscellaneous Nucleosides

"Active methionine" has been synthesized from 5'-deoxy-5'-methyl-
thioadenosine and 2-amino-4-bromobutyric acid, BrCH2-CH2-CH-
(NH2) •C02H.2^* 5'-Deoxy-5'-ethylthioadenosine accumulates in yeast
which has been grown in the presence of ethionine."^

A synthesis of cordycepose has been reported .2*" Spongothymidine has
been shown to consume 1 mol. of periodate without yielding formic acid.
The pentose liberated upon hydrolysis is believed to be D-xylose ;2*^ see,
however, Ref. 115a.

Vicine has been shown to be a glucoside of 2,4-diamino-5,6-dihydroxy-
pyrimidine. The D-glucopyranosyl residue is probably involved in glycosidic
linkage with the hydroxyl group at position 5.-^^

A syntheis of 3-aminoribose, identical with that from puromycin, has
been reported.-*^

A nucleoside, nebularine, occurs in the mushroom Agaricus (Clitocyhe)
nebularis}^'' On hydrolysis it yields ribose and purine.^^^ Its identity with
9-/3-D-ribofuranosylpurine has been shown by two syntheses.^^®

Nucleotides

The structure of the adenylic acids has also been established through an
unambiguous synthesis of adenosine-2 '-phosphate. ^^^ Partial acetylation of
5'-acetyladenosine yielded only one diacetyladenosine. This was converted
into its monobenzyl phosphite and hence into a diacetyladenosine phos-
phate. Gentle alkaline hydrolysis removed both acetyl groups, yielding

2" G. R. Wyatt and S. S. Cohen, Biochem. J. 55, 774 (1953).

^'^^ J. Baddiley and G. A. Jamieson, Chemistry & Industry 1954, 375.

2" F. Schlenk and J. A. Tillotson, J. Biol. Chem. 206, 687 (1954).

280 R. A. Raphael and C. M. Roxburgh, Chemistry & Industry 1953, 1034.

2" K. Sato, J. Biochem. (Japan) 40, 273 (1953).

282 A. Bendich and G. C. Clements, Biochim.et Biophys. Acta 12, 462 (1953).

283 B. R. Baker and R. E. Schaub, /. Am. Chem. Soc. 75, 3864 (1953).

28'' L. Ehrenberg, H. Hedstrom, N. Lofgren, and B. Takman, Svensk Kem. Tidskr. 58,

269 (1946).
286 N. Lofgren and B. Luning, Ada Chem. Scand. 7, 225 (1953).

286 G. B. Brown and V. S. Weliky, /. Biol. Chem. 204, 1019 (1953).

287 D. M. Brown, G. D. Fasman, D. I. Magrath, A. R. Todd, W. Cochran, and M. M.
Woolfson, Nature 172, 1184 (1953).

CHEMISTRY OF NUCLEOSIDES AND NUCLEOTIDES 189

pure adenylic acid a, uncontaminated with any h isomer. Migration of a
phosphate group then cannot have occurred in this synthesis. The di-
acetyladenosine must have been 6,5'-diacetyladenosine. It yielded 6,5'-
diacetyl-a-tosyladenosine on treatment with p-toluenesulfonyl chloride
and methanolysis of this gave a methyl tosyl-D-ribofuranoside. As this
gave 3 , 5-dimethylribose after methylation and hydrolysis, it follows that
the original tosylated diacetyladenosine was 3',5'-diacetyl-2'-tosylad-
enosine, and hence that adenylic acid a is adenosine-2 '-phosphate.

X-ray crystallographic analysis indicates that adenylic acid b is the 3'-
phosphate.-^^

Comparison of the infrared spectra of cytidine and deoxycytidine phos-
phates strongly suggests that cytidylic acid h is cytidine-3 '-phosphate and
hence uridylic acid b is uridine-3 '-phosphate.-^''

Hydrolysis of ribonucleic acid with a crude snake venom yields, amongst
other products, 2', 5'- and 3 ',5 '-diphosphates of uridine and cytidine.-*^

A diphosphate of 5-hydroxymethylcytosine deoxyriboside has been
isolated from bacteriophage nucleic acid.^^"

The synthesis of deoxynucleoside phosphates has now been extended to
the preparation of the 3'- and 5 '-phosphates of deoxycytidine. The latter
compound is identical with the deoxycytidylic acid obtained by enzymic
hydrolysis of deoxyribonucleic acid.'^*^

Nucleotide Coenzymes

Uridine triphosphate^^^ and guanosine triphosphate^^^ have been isolated
from commercial ATP and ATP from rabbit muscle, respectively. Uridine
triphosphate has been synthesized enzymically from ATP and uridine-5'-
phosphate.^^* A tetraphosphate of adenosine is also present in some samples
of ATP.2^* All the mono-, di-, and triphosphates of adenosine, guanosine,
uridine, and cytidine have been detected in tumor extracts.^^^- -^^ Cytidine-
5 '-phosphate and related compounds have been isolated from Lactobacillus
arabinosusP'^

«88 A. M. Michelson and A. R. Todd, J. Chem. Soc. 1954, 34.

289 W. E. Cohn and E. Volkin, /. Biol. Chem. 203, 319 (1953).

290 L. L. Weed and T. A. Courtenay, J. Biol. Chem. 206, 735 (1954).

"1 S. H. Lipton, S. A. Morell, A. Frieden, and R. M. Bock, J. Am. Chem. Soc. 75,

5449 (1953).
"2 R. Bergkvist and A. Deutsch, Acta Chem. Scand. 7, 1307, (1953).
2" A. Munch-Petersen, H. M. Kalckar, E. Cutolo, and E. E. B. Smith, Nature 172,

1036 (1953).

294 D. H. Marrian, Biochim. ei Biophys. Acta 12, 492 (1953).

295 H. Schmitz, V. R. Potter, R. Hurlbert, and D. White, Cancer Research 14, 66,
(1954).

296 H. Schmitz, Biochim. et. Biophys. Ada 14, 160 (1954).

297 J. Baddiley and A. P. Mathias, Chemistry cfc Industry 1954, 277.

190 J. BADDILEY

Guanosine-diphosphate-mannose has been isolated from yeast.^^^ This
nucleotide yields guanosine-5 '-phosphate on hydrolysis, contains two phos-
phate groups, and gives mannose on gentle hydrolysis. From its titration
and other properties it is believed to be a pyrophosphate analogous to
UDPG.

An analogue of UDPG in which the glucose residue is substituted by
A^-acetylglucosamine has been isolated from yeast .^^^

"s E. Cabib and L. F. Leloir, J. Biol. Chem. 206, 779, (1954).

"9 E. Cabib, L. F. Leloir, and C. E. Cardini, J. Biol. Chem. 203, 1055, (1953).

CHAPTER 5

Hydrolysis of Nucleic Acids and Procedures for the Direct

Estimation of Purine and Pyrimidine Fractions by

Absorption Spectrophotometry

HUBERT S. LORING

Page

I. Hydrolysis of Nucleic Acids 191

a. General Considerations 191

1. Acid Hydrolysis of PNA 192

a. Liberation of Purine Bases and Ribose 192

b. Liberation of Inorganic Phosphate and Formation of Pyrimidine Nucleo-
tides 194

c. Liberation of Pyrimidine Bases 195

d. Formation of Di- and Oligonucleotides 196

2. Acid Hydrolysis of DNA 196

a. Liberation of Purine Bases 196

b. Liberation of Pyrimidine Nucleotides and Free Bases 197

3. Alkaline Hydrolysis of PNA 197

a. Formation of Nucleosides 197

b. Formation of Nucleotides 198

c. Liberation of Acid Groups 199

IL Estimation of Purine and Pyrimidine Components in PNA 199

1. Chemical Fractionation of Purine Bases and Pyrimidine Nucleotides and
Their Estimation by Absorption Spectrophotometry 199

2. Effect of Various Treatments on the Recovery of the Purine Bases and
the Pyrimidine Nucleotides 202

3. Experimental Procedure for the Analysis of Ribonucleic Acids by
Adsorption Spectrophotometry 202

4. Application to Purified Yeast Ribonucleic Acids 204

5. Application to Purified Nucleoproteins, Particulate Components of Cells,
and Tissues • 205

I. Hydrolysis of Nucleic Acids

a. General Considerations

The possible hydrolytic products of both pentose and deoxypentose
nucleic acids, PNA and DNA, respectively, include purine and pyrimidine
bases, pentose and pentose phosphate, nucleosides, nucleotides, and oli-
gonucleotides as well as various degradation products derived from the
various substances mentioned. It is characteristic of both types of nucleic

191

192 HUBERT S. LORING

acid, however, that the purine riboside or deoxy riboside linkage is unusu-
ally labile to acid hydrolysis, while the pyrimidine riboside or deoxyribo-
side linkage is relatively resistant. Similarly, phosphate ester groups
attached to purine nucleosides in the 2'- or 3'-positions are relatively acid-
labile in contrast to those attached to pyrimidine nucleosides. The free pur-
ine bases, adenine and guanine, are readily formed during acid hydroly-
sis of both types of nucleic acids, whereas the pyrimidine bases remain for
the most part as mononucleotides in the case of PNA or as nucleoside di-
phosphates in the case of DNA. As the purine bases are cleaved, reducing
groups from the A^- riboside or A^'-deoxy riboside linkages are liberated and,
depending oil the conditions used, free ribose or deoxy ribose may be
formed.

The two types of nucleic acids show, however, a characteristically dif-
ferent behavior in alkali. [Cf. Chapters 10 to 12.] Pentose nucleic acids are
split to mononucleotides by treatment in 1 A^ alkali at room temperature
in contrast to DNA which is little affected, as far as precipitability by tri-
chloroacetic acid is concerned, by such treatment. Such procedures have
served, therefore, for the fractionation and estimation of the relative
amounts of DNA and PNA in tissues.^- ^ [Cf. Leslie, Chapter 16.]

1. Acid Hydrolysis of PNA
a. Liberation of Purine Bases and Ribose

Acid hydrolysis has often been used in estimating the relative amounts of
adenine and guanine, reducing sugar, labile phosphate, and pyrimidine
nucleotides in PNA. It is of value, in using such procedures to estimate the
amounts of the respective components in different PNA's, to assess the de-
gree to which such procedures lead to a quantitative conversion to the
respective products.

The purine bases are liberated from PNA by treatment with alcoholic
HCP- 4 or by hydrolysis with 0.4 N,\N,2N,orQN HCl or H2SO4 at
100-120° for 1-2 hours.*"* Both types of procedures have been used for the
preparation of adenine and guanine from yeast PNA' • ^^"^ as well as in an-
alytical procedures for the estimation of purine and pyrimidine componelits.

» G. Schmidt and S. J. Thannhauser, J. Biol. Chem. 161, 83 (1945).

2 E. Hammarsten, Acta Med. Scand. Suppl. 196, 634 (1947).

3 P. A. Levene, J. Biol. Chem. 53, 441 (1922).

* E. Vischer and E. Chargaflf, J. Biol. Chem. 176, 715 (1948).

« R. D. Hotchkiss, /. Biol. Chem. 175, 315 (1948).

« S. E. Kerr, K. Seraidarian, and M. Wargon, /. Biol. Chem. 181, 761, 773 (1949).

7 J. D.' Smith and R. Markham, Biochem. J. 46, 509 (1950).

8 H. S. Loring, J. L. Fairley, and H. L. Seagran, J. Biol. Chem. 197, 823 (1952).

9 W. Jones, "Nucleic Acids," 2nd ed. Longmans Green & Co., New York, 1920: (a)
p. 107; (b) pp. 41,44.

HYDROLYSIS — ESTIMATION OF BASES IN PNA 193

There is some question about the complete hberation of the purine bases
in 0.5 N HCl at 120° for 2 hours,^- '« but, in 1 A^ acid at 100° for 1 hour,
adenine and guanine are completely liberated as judged by the recovery
of adenine from adenylic acid (95 %) and of guanine from sodium guanylate
(100%) by paper chromatography^ or by the recovery of adenine from
adenylic acid after hydrolysis and precipitation as the silver salt (97%).^
In 2 A^ H2SO4 at 100° purine nitrogen liberated from pancreas PNA reaches
a maximum in about 30 minutes, also indicating the complete liberation of
the purine bases from this type of PNA under these conditions.® The above-
mentioned experiments show that the purine bases are formed readily on
acid hydrolysis of PNA, without appreciable loss, either by paper chroma-
tography or by silver precipitation and without a significant amount of
deamination. Other experiments'^ in which an isotope dilution method was
used to estimate the purine concentration have indicated that as much as
7-8 % of both adenine and guanine may be destroyed by 1 A^ HCl at 100°.
The reasons for the apparent discrepancy between these results and those
mentioned above are not apparent. They probably depend in part on ex-
perimental technique and in part on experimental error. Direct evidence
that hydrolysis in 1 N HCl or H0SO4 at 100° for 1 hour liberates the purine
bases quantitatively is the recovery of from 98-99 % of the nitrogen of some
samples of yeast PNA as known purine and pyrimidine components.'^' ^

In agreement with the accepted theory that PNAs are polynucleotides
and that adenine and guanine are formed from adenylic and guanylic acids,
respectively, is the accompanying formation of reducing groups and inor-
ganic phosphate, as adenine and guanine are liberated during acid hydroly-
sis. The experiments of Levene and co-workers'^' " on the preparation of
a ribose phosphate from xanthylic acid and the inosinic acid derived from
yeast adenylic acid showed that the A^'-riboside linkage could be hydrolyzed
at pH 2 before appreciable amounts of inorganic phosphate were formed.
[Cf. Overend and Stacey, Chapter 2.] In 1 A^ or 2 A^ acid the liberation of
purine N likewise is slightly more rapid than the formation of inorganic
phosphate.® The methods which attempt to relate reducing sugar forma-
tion to adenine and guanine liberation involve the conversion of the ribose
formed to furfural and its distillation or estimation by colorimetry. [Cf.
Dische, Chapter 9.] The liberation of inorganic phosphate during acid hy-
drolysis of PNA has been used directly as a measure of the amount of purine
nucleotides present and indirectly as a measure of the acid-resistant pyrimi-
dine nucleotides.^'''' • ^*

i» M. M. Daly, V. G. Allfrey, and A. E. Minsky, /. Gen. Physiol. 33, 497 (1950).
" R. Abrams, Arch. Biochem. and Biophys. 30, 44 (1951).

12 P. A. Levene and A. Dmochowski, J. Biol. Chem. 93, 563 (1931).

13 P. A. Levene and S. A. Harris, J. Biol. Chem. 95, 755 (1932); 98, 9 (1932); 101, 419
(1933).

1^ W. Jones, J. Biol. Chem. 25, 87 (1916).

194 HUBERT S. LORING

6. Liberation of Inorganic Phosphate and Formation of Pyrimidine Nucleotides

The remarkable stability of the pyrimidine ribonucleotides to mild acid
hydrolysis, in contrast to the purine components, was first noted by Levene
and Jacobs in experiments in which the mixed nucleotides were prepared
as barium salts after a 2-hour hydrolysis of yeast PNA with 2 % H2SO4 .^^•
^®(*^ Subsequently, after the discovery of nucleotides as products of alkaline
PNA hydrolysis, the mixed pyrimidine nucleotides produced by acid hy-
drolysis were fractionated into crystalline barium and brucine uridylates
and into free cytidylic and uridylic acids. ^^- ^* That the phosphoric acid of
the pyrimidine components is slowly hydrolyzed by acid was shown by the
preparation of cytidine and uridine after acid hydrolysis of the nucleotides^^
and by experiments on the rate of liberation of inorganic phosphate from
yeast FNA^^^)- " in 5 % H2SO4 at 100°. Jones showed that an amount of in-
organic phosphate corresponding to 53.9 % of that present in the nucleic
acid was liberated during the first 2 hours, but that the amount formed sub-
sequently corresponded to a rate of only 10 mg. of magnesium ammonium
phosphate (1.26 mg. P) per gram of nucleic acid per hour. Assuming that
the relatively labile phosphate corresponded to that bound to purine nu-
cleosides and correcting the 2-hour period for inorganic phosphate formed
from the pyrimidine components, Jones calculated that approximately 50 %
(50.7 %) of the nucleic acid phosphate was bound to the purine nucleosides
and a similar quantity to the pyrimidine nucleosides in yeast PNA. This
occurrence of labile and stable phosphate in nearly equal proportions in
yeast PNA has been largely responsible for the tetranucleotide concept of
PNA structure.isC') In 2 N H2SO4 after 1 hour at 100° Kerr et aU have
found a slightly higher value for inorganic phosphate formed from yeast
PNA, namely about 57-58 %. The rates of phosphate liberation from pure
cytidylic and uridylic acids is appreciably higher than the value given
by Jones, but the agreement between different investigators is only fair,
e.g., approximately 9% for both nucleotides after 1 hour in 0.1 A^ H2SO4 at
]^00°i9a,b g^g compared with 13.8% for cytidylic acid under the same condi-
tions or with 15.5% in 1 AT acid.^" The recovery of purine and pyrimidine

16 P. A. Levene and W. A. Jacobs, Ber. 44, 1027 (1911).

" P. A. Levene and L. W. Bass, "Nucleic Acids," ACS Monograph Series. The

Chenucal Catalogue Co., New York, 1931 : (a) p. 221 ; (b) p. 274; (c) p. 57; (d) p. 193;

(e) p. 265.
" P. A. Levene, Proc. Soc. Exptl. Biol. Med. 15, 21 (1917).
18 S. J. Thannhauser and G. Dorfmuller, Z. physiol. Chem. 100, 121 (1917); 104, 65

(1919); Ber. 51,467 (1918).
19* A. M. Michelson and A. R. Todd, /. Chem. Soc. 1949, 2476.
i^b G. R. Barker, J. M. Gulland, H. Smith, and J. F. Thomas, /. Chem. Soc. 1949,

904, find an average value of 7% dephosphorylation for disodium uridylate, [a]^ =

19.8° (anhydrous) in 0.1 N H2SO4 at 100°.
2» P. A. Levene and E. Jorpes, /. Biol. Chem. 81, 575 (1929).

HYDROLYSIS — ESTIMATION OF BASES IN PNA 195

components in most of the recently published analyses lies within the range
expected from the above mentioned values, viz., 0.50-0.57 and 0.36-0.46
moles per mole of P for purines and pyrimidines, respectively. [Cf. Maga-
sanik, Chapter 11.]

In estimating the amounts of pyrimidine nucleotides in PNA after acid
hydrolysis, it is apparent from the above discussion that appreciable error
may result if hydrolysis to nucleosides is neglected .^^' ^- Of significance also
is the possible deamination or destruction of the pyrimidine compounds.
In 0.4 N sulfuric acid at 100° a 2 % deamination of cytidine per hour and
in 1 AT" HCl a 3-4 % deamination of cytidylic acid have been reported.^^- ^'

c. Liberation of Pyrimidine Bases

The hydrolysis of the pyrimidine components to the free pyrimidine
bases presents considerable difficulty. Procedures for the preparation of
cytosine and uracil from yeast PNA involve a 2-hour hydrolysis at 175° in
25 % H2SO4 .^^f=> Under these conditions there is a considerable deamination
of cytosine to uracil, a fact which in the earlier literature placed doubt on
the existence of uracil as a nucleic acid component. More recently the lib-
eration of the free pyrimidine bases from PNA by 0.4-6 N HCl at 120°,^ by
concentrated formic acid (98-100 %) and by 20 % HCl at 175°," and by 12
N perchloric acid-^ has been studied. The results with 0.4-6 A^ HCl at 120°
show an increasing liberation of pyrimidine bases as acid concentration is
increased, but the yields of cytosine relative to uracil are low, indicating a
considerable degree of deamination under these conditions. In concentrated
formic acid cytosine is apparently preserved from deamination to uracil,
but the recovery of pyrimidine bases in yeast PNA relative to P amounted
to only 0.327 mole per mole, in contrast to values as high as 0.46 found by
other methods after 1 N acid or alkaline hydrolysis.^' *• " Hydrolysis of
yeast PNA with 12 A^ perchloric acid^" for 1 hour at 100° causes no appreci-
able destruction of either adenine, guanine, cytosine, uracil, or thymine,
but the recovery of total pyrimidine base when this procedure was applied
to yeast PNA, 0.37 moles per mole P, again indicates an incomplete libera-
tion from PNA. At the present time, therefore, under the conditions studied
neither 20% HCl, concentrated formic acid, nor 12 N perchloric acid can
be said to lead to the quantitative liberation of the pyrimidine components
of PNA in the form of free bases.

21 21. R. Markham and J. D. Smith, Biochem. J. 49, 401 (1951).

" H. S. Loring, J. L. Fairley, H. W. Bortner, and H. L. Seagran, /. Biol. Chem. 197,

809 (1952).
" H. S. Loring, and J. Mc T. Ploeser, /. Biol. Chem. 178, 439 (1949).
" A. Marshak and H. J. Vogel, J. Biol. Chem. 189, 597 (1951).
26 E. Chargaff, B. Magasanik, E. Vischer, C. Green, R. Doniger, and D. Elson, J .

Biol. Chem. 186,51 (1950).

196 HUBERT S. LORINQ

d. Formation of Di- and Oligonucleotides

Considerable attention is given in the earlier literature'^^*^^ to the occurrence of di-
and oligonucleotides as hydrolytic products of nucleic acids. The existence of such
compounds, however, was largely disproved when the acid-resistant pyrimidine nu-
cleotide fraction, thought to be a cytidylic uridylic dinucleotide, was successfully
separated into cytidylic and uridylic acids. Accordingly the reported isolation bj'
Thannhauser^® of a trinucleotide after enzymic hydrolysis was never accepted by the
Levene school. Recent evidence of the occurrence of products larger than mononu-
cleotides in ribonuclease digests of PNA and deoxyribonuclease digests of DNA is
considered in subsequent sections of the book. [Cf. Chapters 6, 8, 10, 11, and 15.]

It has, in addition, been shown" that several dinucleotides and one trinucleotide
are formed when yeast PNA is treated at room temperature with 15 parts of 6 A^ HCl
for 3 minutes at room temperature. The products were successfully fractionated on
Dowex 1 (chloride) and were characterized as 5'-[(6 or a)-guanylyl]cytidylic acid b,
5'-[(b or a)-adenylyl]cytidylic acids a and b, 5'-[(b or a)-cytidylyl]cytidylic acid b,
uridylic acid-cytidylic acid dinucleotide, and adenylic acid diguanylic acid trinucleo-
tide.

2. Acid Hydrolysis of DNA
a. Liberation of Purine Bases

The purine bases of DNA are easily removed by mild acid treatment
(heating the free nucleic acid in 2 % solution at boiling water bath tempera-
ture for 10 minutes) apparently without complete degradation of the origi-
nal polynucleotide structure. The material remaining was early recognized
as a complex substance. '^^®^ It was believed free of cytosine as well as ade-
nine and guanine and was named thymic acid^^ because presumably the
only base remaining in the original polynucleotide structure was thymine.
Feulgen, in a highly important paper,^^ noted that a less degraded product
could be obtained under somewhat milder conditions than those mentioned
above, namely heating for 40 minutes in slightly acid solution below 80°.
This material gave N:P ratios corresponding to approximately equimolar
quantities of cytosine and thymine, which were both also isolated after
acid hydrolysis. Of special interest to cytology^" was the demonstration
that thymic acid gave a positive fuchsin test which later was correlated
with the occurrence of DNA in the cell nucleus. In more recent experiments
it was shown that the cytosine-thymine ratio of the original DNA was not
altered when DNA was dialyzed at 37° and pH 1.6 and the name "apurinic
acid" was proposed for the resulting product.^' [Cf. Chargaff, Chapter 10.]

26 S. J. Thannhauser, Z. physiol. Chem. 91, 329 (1914).

" R. B. Merrifield and D. W. Woolley, Federation Proc. 11, 258 (1950); J. Biol. Chem.

197,521 (1952).
28 A. Kossel and A. Neumann, Z. physiol. Chem. 22, 74 (1896-97).
" R. Feulgen, Z. Physiol. Chem. 101, 296 (1918).

^^ J. Brachet, "Chemical Embryology." Interscience Publishers, New York, 1950.
31 C. Tamm, M. E. Hodes, and E. Chargaff, /. Biol. Chem. 195, 49 (1952).

HYDROLYSIS — ESTIMATION OF BASES IN PNA 197

b. Liberation of Pyrimidine Nucleotides and Free Bases

The above-mentioned experiments show the relative ease with which the
purine bases are removed from DNA. In quantitative estimations alcoholic
HCl (methanol saturated with HCl for 3-5 hours at 50° ■* or 1.5 % metha-
nolic HCl for 20 hours at 37°) "^ has been used. If DNA is heated in 2 %
H2SO4 for 2 hours under reflux conditions, the remaining polynucleotide
structure after removal of the purine bases is further degraded to mixtures
of thymine deoxyribodiphosphoric acid and cytosine deoxyribodiphosphoric
acid, which may be fractionated and obtained as crystalline compounds.
[Cf. Baddiley, Chapter 4.] The extent to which other pyrimidine derivatives
also may be formed has not been determined.

The hydrolysis of DNA with concentrated formic acid at 175°,^ with 6
N HCl at 120° ^° or with 12 N perchloric acid^^ leades to a relatively quanti-
tative formation of the free pyrimidine bases. The merits of the different
hydrolysis procedures are discussed in Chapters 7 and 10 as regards the
recovery of individual bases.

3. Alkaline Hydrolysis of PNA

a. Formation of Nvcleosides

The acid instability of the purine ribosidic linkages, in contrast to their
alkali stability, was recognized when guanosine (then called vernine) and
inosine were discovered.^^' ^* This property of nucleosides has played an
important role in establishing the structures of the purine and pyrimidine
mononucleotides, since when alkaline hydrolysis is applied to yeast nucleic
acid (4-5 % NH3 for 3.5 hours at about 145°) ^^ or to mononucleotides, it is
the phosphate ester linkage which is hydrolyzed with the formation of in-
organic phosphate and purine and pyrimidine nucleosides. The conversion
of inosinic acid to inosine by alkaline hydrolysis or to hypoxanthine and
ribose phosphate by acid hydrolysis as carried out by Levene and Jacobs^^
thus laid the groundwork for the structure of all the nucleotides as phos-
phate esters of nucleosides linked through a hydroxyl group of the sugar as
shown below.

HO HO

^ H+ . OH-
purine + 0=P — O — sugar «- 0=P — O — sugar-punne >

/ /

HO HO

sugar-purine + HP04~

32 E. Schulze and E. Bosshard, Z. physiol. Chem. 10, 80 (1886).

33 F. Haiser and F. Wenzel, Monatsh. 29, 157 (1908).

34 P. A. Levene and W. A. Jacobs, Ber. 43, 3150 (1910).

35 P. A. Levene and W. A. Jacobs, Ber. 41, 2703 (1908) ; 42, 335, (1909) ; 42, 1198 (1909) ;
44,746 (1911).

198 HUBERT S. LORING

In the original procedure given by Levene and Jacobs^* for the prepara-
tion of guanosine, adenosine, cytidine, and uridine, yeast nucleic acid was
heated with concentrated ammonium hydroxide under pressure in an oil
bath at 175-180° for 3.5 hours. More recently aqueous pyridine^^ (under
reflux for 43^ days), a catalytic hydrolysis with lanthanum,^^ and phos-
phatase hydrolysis of nucleotides^^-^" have been used to prepare the purine
and pyrimidine ribonucleosides in moderate quantities. Their fractionation
by starch chromatography^' in milligram quantities has also been reported.

h. Formation of Nucleotides

As mentioned above, the pyrimidine nucleotides were recognized as acid-
resistant phosphate hydrolysis products of yeast PNA in 1911.'^ The dis-
covery that similar organic phosphate compounds could also be produced
during alkaline hydrolysis of yeast PNA was made independently by
Thannhauser and Dorfmuller'*^ an(j ^y Jones and Germann*^ when this sub-
stance was treated with ammonia under milder conditions (2.3 % NH3 for
1 hour at 115°) than those used by Levene and Jacobs^* for the formation of
nucleosides. The subsequent purification, isolation, and crystallization of
the compounds known as uridylic acid,'^ adenylic acid,'**- '^^ cytidylic acid,'^
and guanylic acid^^ resulted from work performed in the three laboratories
mentioned on either ammoniacal hydrolysates of yeast PNA or, for the
pyrimidine mononucleotides, on acid hydrolysates. The still milder hydroly-
sis with approximately 1 A^ NaOH at room temperature was discovered by
Steudel and Peiser^^ in experiments in which guanylic and adenylic acids
were isolated from yeast PNA after application of this procedure. More
recently 1 N alkali at 37°/ 0.6 N Ba(0H)2 ,^» and 0.1 N NaOH at 100° ''
have been used.

36 H. Bredereck, A. Martini, and F. Richter, Ber. 74, 694 (1941).

37 F. A. Allen and J. E. Bacher, /. Biol. Chem. 188, 59 (1951).

38 J. M. Gulland and T. F. Macrae, J. Chem. Soc. 1933, 662.

39 H. S. Loring, M. L. Hammell, L. W. Levy, and H. W. Bortner, J. Biol. Chem. 196,
821 (1952).

" P. Reichard, Y. Takenaka, and H. S. Loring, J. Biol. Chem. 198, 599 (1952).

« P. Reichard, Nature 162, 662 (1948) ; /. Biol. Chem. 176, 763 (1949).

« S. J. Thannhauser, Z. physiol. Chem. 91, 329 (1914); S. J. Thannhauser and G.
Dorfmiiller, 95, 259 (1915); 100, 121 (1917).

" W. Jones and H. C. Germann, J. Biol. Chem. 25, 93 (1916).

*^ W. Jones and R. P. Kennedy, /. Pharmacol. Exptl. Therap. 12, 253 (1918).

45 S. J. Thannhauser, Z. phyaiol. Chem. 107, 157 (1919).

*6 Guanylic acid was named by Bang in Olaf Hammarsten's laboratory [0. Bang, Z.
physiol. Chem. 26, 133 (1898-99)] as a hydrolysis product of pancreas. It was dis-
covered as a component of yeast RNA and crystallized by Levene [P. A. Levene,
J. Biol. Chem. 40, 171 (1919) ; 41, 483 (1920)].

" H. Steudel and E. Reiser, Z. physiol. Chem. 114, 201 (1921); 120, 292 (1922).

« H. S. Loring, P. M. Roll, and J. G. Pierce, J. Biol. Chem. 174, 729 (1948).

HYDROLYSIS — ESTIMATION OF BASES IN PNA 199

c. Liberation of Acid Groups

The earlier experiments mentioned above were concerned primarily with
the preparation of the nucleotides in pure form and with their chemistry.
It is likely that appreciable amounts of inorganic phosphate, nucleosides,
and deaminated products as well as nucleotides were formed under the
more severe conditions employed. When a 2 % solution of sodium nucleate
in 0.1 N NaOH is heated at 100°, there is at first a rapid liberation of acid
groups^^' ^'^ followed by a slower liberation of acid until a maximum value
is reached corresponding to about 1.08 equivalents per mole of nucleic acid
P. As no increase in inorganic phosphate was found under these conditions,
it is evident that the acid formed was due to the hydrolysis of phosphate
ester linkages in the original nucleic acid structure. Of interest is the fact
that the amount of acid formed, 1.08 equivalents per mole of P, is greater
by approximately 8 % than that expected from a nucleotide polymer con-
taining only diester linkages. The extent to which deamination of purine
nucleotides occurs under these conditions has not been determined and is
probably negligible. Cytidylic acid is deaminated to the extent of 2 and 12 %
in 0.01 A'' and 0.1 A^ NaOH, respectively, but is not significantly affected
at pH ll.-*' Appreciable deamination, from 10% to 33%, apparently occurs
in 1 N alkali at 37°.^!

II. Estimation of Purine and Pyrimidine Components in PNA"

1. Chemical Fractionation of Purine Bases and Pyrimidine
Nucleotides and Their Estimation by Absorption
Spectrophotometry

As discussed above, hydrolysis of PNA with 1 A'' acid at 100° for 1 hour
leads to the formation of a mixture of the purine bases and pyrimidine
nucleotides with small amounts of pyrimidine nucleosides and possibly oxy-
purines. As the purine bases form highly insoluble silver salts in acid solu-
tion^^' ^2' ^^ in contrast to either pyrimidine nucleotides or nucleosides, it is
possible to effect a relatively quantitative separation of the two types of
components by this procedure. If the purine bases are redissolved in dilute
HCl and^ the pyrimidine nucleotide fraction, after removal of silver ions,
treated with prostatic phosphatase, two relatively simple binary mixtures

*' H. S. Loring, H. W. Bortner, L. W. Levy, and M. L. Hammell, J. Biol. Chem. 196,

807 (1952).
*" H. W. Bortner, Dissertation, submitted to Stanford University for the Degree of

Doctor of Philosophy in Chemistry, 1952.
" D. H. Marrian, V. L. Spicer, M. E.Balis, and G. B. Brown, J. Biol. Cheyn. 189, 533

(1951).
" R. Feulgen, Z. phjsiol. Chem. 102, 244 (1918).
" S. E. Kerr and K. Seraidarian, J. Biol. Chem. 159, 211 (1945).

200

HUBERT S. LORING

Uj

9,4 00

240mp^

13,100^^-^

^

13,000

262 mv/^

~N. 278mji

ADENINE—-/

\

/■^ / 10,900

\

/ Y''252m>j

\cYTIDINE

/ A 9,900

\

/ / \260m)j

\

In

y \ 9,300 \
A r266 5m;j \

/ /7.650V/

/ /262m)rK

7,120 \
276m>i \

/ / L6.340 \ \

\_6,660 \
\ 280fnp \

/ / / 260m(j \ \

\ \

/^/s.sio /

\ \ \

/ 240m(j/

4.720 \

\ \ \

/ /

280 m)i \

\ \ \

/ /

4,500/

\ \

URIDINE /

278m(j

\ \

220 230

240 250 260 270 280

WAVE LENGTH (mp)

290 300

Fig. 1. The ultraviolet absorption spectra of adenine and guanine in 0.1 A' HCl and
of cytidine and uridine in 0.01 A' HCl.

are obtained of adenine, and guanine on the one hand and cytidine and uri-
dine on the other. ^^ The ultraviolet absorption spectra of each pair of com-
ponents in dilute acid as illustrated in Fig. 1«. 22, 49 are sufficiently different
to allow the estimation of the concentration of each component to an extent
limited only by the precision of the spectrophotometer and the degree to
which other ultraviolet-absorbing impurities may be present in either mix-
ture. [Cf. Beaven, Holiday, and Johnson, Chapter 14.]

The analysis of a binary mixture of components, A and B, having different absorp-
tion spectra, is made on the assumption that the total optical density, D, at each of

^* The relative amounts of the isomeric cytidylic and uridylic acids present in the
pyrimidine nucleotide fraction can be estimated directly^ by absorption spectro-
photometry without conversion to nucleosides. The uncertainty in such analyses
is larger, however, because the two isomeric cytidylic and uridylic acids present in
acid hydrolysates [H. S. Loring, N. G. Luthy, H. W. Bortner, and L. W. Levy, J.
Am. Chem. Soc. 72, 2811 (1950) ; W. E. Cohn, ibid. 72, 2811 (1950)] have somewhat
different absorption spectra.*'

HYDROLYSIS — ESTIMATION OF BASES IN PNA 201

two wavelengths, \i andX2, is the sum of the densities due to each component at each
wavelength.^' ^-^ ^^ Thus from the Beer-Lambert law two simultaneous equations
can be written in which the concentrations of A and B are expressed as a function of
their molecular extinction coefficients and the optical densities at the respective
wavelengths as follows :^^

CaEaXi + CbEbxi = jDxi
and

CaEa\2 + CbEb\2 = D\2

Ca and Cb are the respective concentrations of A and B expressed as moles per liter;
Ea\i, Eb\i, EA\2f s-rid Eb\2 are the respective molecular extinction coefficients of A and
B at Xi and X2; and Dx^ and Dx2 are the optical densities at the two wavelengths.
Solution of the equations for Ca and Cb gives the following:

^ £^sXiJDx2 — EbmDx,

EamEbXi — EaXiEbm

and

Cb =

EaXuDxi — EaXjDx^

EaXiEbXi — EaXxEbX".

The pairs of wavelengths selected for the estimation of adenine and guanine or of
cytidine and uridine are somewhat arbitrary, but, in order to achieve as high a sensi-
tivity as possible for the detection of either component, it is desirable to use wave-
lengths where relatively large differences in the molecular extinctions of the two
compounds occur, but at which both show appreciable absorption. If one of the
wavelength pairs is selected near either end of the ultraviolet spectrum as well, a
farther check is obtained as to whether or not ultraviolet-absorbing impurities are
present which absorb appreciably at the shorter or longer wavelengths. Similarly, by
using wavelengths at which intersection points of the respective absorption curves
occur, estimates of the total concentration of both components can be made, or, by
using a wavelength at which one component fails to absorb, the other can be esti-
mated independentl3^

In the analysis of PNA performed in the author's laboratory* • 2- t\vo
wavelength pairs, 262 and 280 m/x, and 262 and 240 m/i, were used for esti-
mation of adenine and gvianine and both intersection points at 252 m^z and
276 m/i for calculation of total purine bases. The respective equations using
molecular extinction values for highly purified samples of adenine and
guanine are as follows:
Adenine

^ 6.66D262 - 7.65Do8o ^ 9.40Dj62 - 7.65D.40

262-280 mti C = ■ 262-240 m^, C = —

5.U X 10^ 8.25 X 10^

'* If the optical density ratios of the pure substances at the wave lengths chosen are
appreciably-d-ifferent, the relative amounts of the two components present can be
determined most simply by extrapolation of the optical density ratio.

202 HUBERT S. LORING

Guanine

262-280 mil, C = ; 262-240 m/i, C =

Total purine

5.11 X 10^ ' 8.25 X 10^

252 mM, C = -^^ ; 276 m^, C = — -
10,950 ' ^' 7,120

The corresponding equations for cytidine and uridine at 260 and 278 m/x
and for total pyrimidine nucleosides at 266.5 m/u are as follows:

Cytidine, C =
Uridine, C =

9.902)278 - 4.50D260
1.0 X 10'

13.01)260 - 6.34D278
1.0 X 10«

■D266.5

Total pyrimidine nucleoside, C =

9,380

2. Effect of Various Treatments on the Recovery of the Purine

Bases and the Pyrimidine Nucleotides

Various recovery experiments have been performed to determine the extent to
which known crystalline samples of adenine, guanine, cytidylic acid, and uridylic
acid could be recovered after treatment by the various procedures outlined above.
The percentage recoveries of mixtures of adenine and guanine after precipitation as
silver salts, after heating for 1 hour at 100° in 1 A'' H2SO4 and after a combination of
the two procedures are shown in Table I. In the same table the recoveries of cytidylic
and uridylic acids as cytidine and uridine, respectively, are shown after dephos-
phorylation by prostatic phosphatase and after acid hydrolysis and dephosphoryl-
ation both without and with the addition and removal of Ag+. Because the re-
covery of cytidylic acid is low (96-97%) and the recovery of uridjdic acid high
(102-104%) after acid hydrolysis, it appears that, as with cytidine under similar con-
ditions,^^ a 3-4% deamination of cytidylic acid occurs after heating a mixture of these
substances in 1 A^ H2SO4 at 100° for 1 hour. Other recovery experiments on mixtures
of both purine bases and pyrimidine nucleotides gave similar recoveries ranging from
98-101% of the amounts used, provided a correction for 3.5% deamination of cytidylic
acid was made. It was furthermore shown that the pyrimidine nucleoside fraction
could be filtered through Dowex 1 bicarbonate without loss and that added aromatic
amino acids and the traces of purine bases remaining owing to the slight solubility
of the silver purines were effectively removed by this treatment.

3. Experimental Procedure for the Analysis of Ribonucleic Acids

BY Absorption Spectrophotometry

In analyzing PNA samples by the above-mentioned procedures, it is desirable to
relate the purine and pyrimidine composition to their nitrogen and phosphorus con-
tents. In the procedure described below, the Ma and Zuazaga method^^ for nitrogen
(micro-Kjeldahl) and a modified Fiske-Subbarow procedure^' were used.

" T. S. Ma and G. Zuazago, Ind. Eng. Chem., Anal. Ed. 14, 280 (1942).

HYDROLYSIS — ESTIMATION OF BASES IN PNA

203

TABLE I
Per Cent Recovery of Purine Bases and Pyrimidine Nucleotides After

Various Treatments

Treatment

Acid hydrolysis N H2SO4 , 1 hr.,

100°
Precipitation as silver salts
Acid hydrolysis and precipitation

as silver salts

Adenine

262-
280 mix

99

99
98

262-
240 mM

99

98
98

Guanine

262-
280 m/x

101.5

99
99

262-
240 mju

101.5

100
99

Total purine

252 mM

100.5

98.5
99

276 mM

100

98
98

Treatment

Dephosphorylation by acid phosphatase
Acid hydrolysis and dephosphorylation
Acid hydrolysis, Ag"*" added and removed,

and dephosphorylation
Dephosphorylation and Dowex 1 column

(amino acids present)

Cytidine

260-278
m/i

100
97
96

100

295 mM

102
96
96

98

Uridine

260-278
mM

102
104
102

99

Total
pyrimi-
dine

267 mM

100

100

99

99

A 150-mg. sample" in about 30 ml. of 1 N sulfuric acid in a test tube capped with a
small beaker is heated for 1 hour in a boiling water bath. The solution is filtered
through a sintered glass funnel to remove a few particles of material presumed to be
coagulated protein, and the filtrate and several washings are diluted to 50 ml. in a
volumetric flask. Portions are analyzed directly for total nitrogen and, after suitable
dilution, for total phosphorus.

For the determination of purine and pyrimidine content, 10-ml. portions of the
hydrolysate, in triplicate, are placed in 15-ml. centrifuge tubes. The following pro-
cedures are carried out in an identical manner on each of these aliquots as well as on
a 10-ml. sample of 1 A'' H2SO4, which provides a reagent blank for all subsequent op-
tical density measurements. The pH of the solution is brought to about 1.0 by the
addition of 11 A^ KOH, and, after warming to 90°, 1 ml. of a 20% AgNOs solution is
added to precipitate the adenine and guanine as silver salts. The suspension, after
standing overnight in the refrigerator, is centrifuged in the cold to pack the precipi-
tate of the silver-purines and a small amount of Ag2S()4 . The precipitate is washed
four times in the centrifuge tube with 3-ml. portions of ice-cold 0.1 A'^ H2SO4 . The
original supernatant solution and the washings are fiiltered through a sintered glass

" This amount of nucleic acid diluted as described is convenient when micro-Kjel-
dahl analyses for nitrogen are to be performed. If the sample is analyzed for pur-
ine, pyrimidine, and phosphorus only, the amount used can be decreased to about
2 mg. by decreasing the volumes of reagents used proportionally throughout.*

204 HUBERT S. LORING

funnel to remove stray particles of the flocculent silver-purines and set aside for anal-
ysis of the p3'rimidine components.

The silver-purine precipitate is subjected to four successive treatments with 10-ml.
portions of 0.1 N HCl to extract the free purine bases. In each case the suspension of
silver salts in the acid is heated in a boiling water bath for 5 minutes with occasional
stirring and then allowed to cool for 10 minutes. The solids are sedimented by cen-
trifugation, and the supernatant liquid is poured onto the funnel used in the separa-
tion of the purine and pyrimidine fractions. The acid solution is allowed to remain
in the funnel for several minutes in contact with a few particles of silver-purine re-
tained in the previous step before suction is applied. The combined filtrates from this
procedure, constituting the pure fraction, are diluted with 0.1 A'^ HCl to 50 ml. A
10-ml. aliquot diluted to 100 ml. with 0.1 N HCl is used for optical density measure-
ments at 240, 252, 262, 276, and 280 m^. The adenine, guanine, and total purine con-
centrations are calculated from the optical densities as previously described.

The pyrimidine fraction is warmed to about 70°, 3 ml. of 1 iV HCl added, and the
AgCl precipitate allowed to coagulate and removed by filtration. The precipitate is
washed three times by suspension in 5-ml. portions of 0.1 N HCl. The combined fil-
trates are diluted to about 45 ml., the pH of the solution adjusted to 4.7 with NaOH
(pH meter), and the volume brought to 50 ml. Two 5-ml. aliquots are treated with

I ml. of a filtered phosphatase solution containing 1 mg. of enzyme'' and the result-
ing solutions are incubated at 38° for 3 hours to bring about dephosphorylation of
the pyrimidine nucleotides. One aliquot, diluted to 100 ml. with 0.1 N HCl, is used
for optical density measurements at 260, 267, 278, and 295 m/x. Cytidine, uridine, and
total pyrimidine concentrations were evaluated as previouslj' described and corrected
for 3.5% deamination of cytidylic acid. The values obtained in this way may be
slightly high if small amounts of ultraviolet-absorbing amino acids were present
initially. The latter may be removed as follows: A second aliquot is adjusted to
pH 8.3 with NaOH, the solution filtered through a Dowex 1 (bicarbonate) column
(2 cm. X 3 sq.cm.), and the resin washed with about 70 ml. of 2% NaHCOs. The pH of
the combined effluent is carefully adjusted to 1.0 with H2SO4 and the volume to
100 ml. Optical density measurements are made and the concentrations of cytidine,
uridine, and total pyrimidine nucleoside calculated as previously described.

4. Application to Purified Yeast Ribonucleic Acids

The results of the application of the above mentioned analytical pro-
cedures to different purified samples of yeast PNA are summarized in Table

II in comparison with those of several other investigators. [Cf. Magasanik,
Chapter 11.] They show that from 97 to 99 % of the nitrogen of commercial
sodium ribonucleate or of a carefully prepared nucleic acid sample can be
accounted for in terms of known purine and pyrimidine components. Simi-
lar almost complete recoveries of the nitrogen of other yeast ribonucleic
acid samples are reported in several instances by other workers.^' 24. 25, ss. 59
In relation to phosphate content the results show that purine and pyrimi-
dine bases occur in very nearly equimolar quantities with phosphorus and
confirm the general opinion that ribonucleic acids are essentially poly-
ps A. Deutsch, R. Zuckerman, and M. S. Dunn, Ind. Eng. Chem., Anal Ed. 24, 1769

(1952).
" G. W. Crosbie, R. M. Smellie, and J. N. Davidson, Biochem. J. 54, 287 (1953).

HYDROLYSIS — ESTIMATION OF BASES IN PNA

205

Table II
Purine and Pyrimidine Components of Yeast Ribonucleic Acid Found by

Various Investigators

Sample"

Method
of hydro-
lysis

Mo

Ade-
nine

les pe

Gua-
nine

r mol

Cyti-
dine

e P

Uri-
dine

P ac-
counted
for

N ac-
counted
for

Re-
fer-
ence

Commercial N. A.

Acid

0.275

0.334

0.212

0.248

107

98

7

Commercial Preparation

Alkaline

0.280

0.290

0.178

0.203

95.1

89.4

25

1

Acid

0.288

0.258

0.165

0.195

90.6

86.0

25

Commercial Preparation

2
Preparation 3 bakers'

Alkaline

0.254

0.265

0.199

0.177

89.5

88.5

25

Alkaline

0.262

0.248

0.214

0.203

92.7

93.6

25

yeast

Alkaline

0.255

0.246

0.202

0.232

93.5

92.2

25

Acid

0.283

0.233

0.211

0.246

97.3

96.8

25

Preparation 4 bakers'

Alkaline

0.242

0.230

0.183

0.235

89.0

88.5

25

yeast

Alkaline

0.234

0.232

0.175

0.241

88.2

87.4

25

Commercial N. A.

Acid

0.25

0.29

0.18

0.19

91

95

24

Sodium nucleate

Acid

0.250

0.269

0.209

0.240

97

99

8

Bakers' yeast

Acid

0.249

0.272

0.212

0.236

97

97

8

Commercial N. A.

Acid

0.258

0.272

0.195

0.228

95

95

8

Commercial N. A.

Alkaline

0.278

0.273

0.206

0.265

102

102

58

Commercial N. A.

Alkaline

0.237

0.272

0.225

0.237

97

59

" The original publications .should be consulted for complete descriptions of the samples, the methods of
purification, and the methods of hydrolysis used. The cytidine values after acid hydrolysis, witli the excep-
tion of those of reference 8 have not been corrected for deamination. Commercial N. A. = commercial ribo-
nucleic acid. The preparation numbers given are the same as those of the original publication (Chargaff et
al.«).

nucleotides. A comparison of the various samples analyzed by different in-
vestigators reveals differences which appear to be significant. The results
indicate that considerable variation occurs in the composition of samples
of yeast ribonucleic acid, depending on the procedure employed during
their preparation and purification.^" Because of such variation it appears
that a better characterization of a particular ribonucleic acid lies in the
direct analysis of the purified nucleoprotein or of the particulate component
containing the nucleic acid.

5. Application to Purified Nucleoproteins, Particul.\te Components
OF Cells, and Tissues

The analysis of the ribonucleic acid occurring in a purified nucleoprotein,
in a separated particulate component of cells like mitochondria or micro-

^° For differences in yeast RNA composition depending on the conditions under
which yeast is grown see K. Dimroth and L. Jaenicke, Z. Naturforsch. 56, 185
(1950).

206 HUBERT S. LORING

somes, or in a particular tissue presents a number of difficulties depending
on the amounts of non-nucleic acid purine or pyrimidine or other similar
ultraviolet-absorbing compounds that may be present and the variability
of the respective materials. Of the many types of biological materials avail-
able for study, the plant viruses approach most closely substances that can
be described as relatively homogeneous, soluble nucleoproteins. Because of
their high molecular weight, it is relatively simple to separate them by dif-
ferential ultracentrifugation from the many low-molecular-weight com-
pounds which are associated with living cells and might interfere in the
analysis for the purine and pyrimidine components of the ribonucleic acid.
Analyses have been made on the nucleic acid prepared from purified tobacco
mosaic virus*^-^^ and many of its strains,^* by heat or alkali treatment as
well as on the trichloroacetic acid-extracted nucleoprotein (TCA nucleo-
protein)** and on the purified virus itself. [Cf. Magasanik, Chapter 11.] The
average results expressed as percentage molar proportions of the four bases
are summarized in Table III under the respective host species used for the
cultivation of the virus. While examination of the combined analyses re-
veals some differences in the composition of tobacco mosaic virus nucleic
acid as reported by the three laboratories, it is also clear that most of the
results are in good agreement and the differences within the experimental
error. Examination of Table III reveals that quite similar results were found
for both the TCA nucleoprotein and its isolated nucleic acid by the spectro-
photometric method described above. These results are also in good agree-
ment with those found by Knight for the isolated nucleic acid if the high
adenine value reported by him is corrected for a small amount of cytidine
likely to be present in the same area of the chromatogram and if a similar
correction is made for small losses of unrecovered uridine.^'- ^^ The results
from the two laboratories are thus in essential agreement for the composi-
tion of tobacco mosaic virus nucleic acid when the virus is produced in
Turkish tobacco plants. Knight in studies of a number of strains derived
from tobacco mosaic virus found a relatively constant composition for the
respective nucleic acids, the results suggesting, in fact, that the nucleic
acids of different strains of the same virus may have identical composi-
tions. ^^^ ^*

The fractionation of broken-cell preparations of tissues into large and
small granule fractions as begun by Bensley^^ and modified and improved

" R. Markham and J. D. Smith, Biochem. J. 46, 513 (1950).

«2 C. A. Knight, J. Biol. Chem. 197, 241 (1952).

*' W. D. Cooper and H. S. Loring, J. Biol. Chem., in press.

6* F. L. Black and C. A. Knight, J. Biol. Chem. 202, 51 (1953).

«6 R. Markham and J. D. Smith, Biochem. J. 49, 401 (1951).

«6 C. A. Knight, /. Biol. Chem. 171, 297 (1947).

" R. R. Bensley, Biol. Symposia 10, 323 (1943).

HYDROLYSIS — ESTIMATION OF BASES IN PNA

207

TABLE III
Percentage Molar Proportions of the Purine and Pyrimidine Bases in the
Nucleic Acid from Purified Tobacco Mosaic Virus and from Certain of
Its Strains, in the Purified Virus and in the Trichloroacetic Acid-
Insoluble Nucleoproetin (TCA Nucleoprotein) as Reported
BY Different Workers

Percentage molar

Strain"

Host

Material

proportions

Re-
fer-
ence

analyzed

Ade-

Gua-

Cyto-

Ura-

nine

nine

sine

cil

TMV

White Burley
tobacco

Nucleic acid

31.0

29.2

15.5

24.0

61

TMV

Turkish tobacco

Nucleic acid

29.8

25.2

18.5

26.2

62

TMV

Turkish tobacco

Purified virus

28.0

25.8

19.5

26.2

62

TMV

Turkish tobacco

Nucleic acid

28.0

24.0

20.0

28.0

63

TMV

Turkish tobacco

TCA-nucleo-
protein

27.8

24.2

19.8

28.6

63

M

Turkish tobacco

Nucleic acid

29.5

26.2

19.3

25.8

62

M

Turkish tobacco

Purified virus

27.5

25.0

20.5

27.5

62

M

Turkish tobacco

TCA-nucleo-
protein

27.5

24.5

19.8
18.5

28.3

63

Average for si:

< strains produced

Nucleic acid

29.5

25.5

26.2

62

in Turkish 1

obacco

±0.50

±0.65

±0.55

±0.63

Purified virus

28.0

24.7

20.5

26.8

62

±0.63

±0.65

±0.60

±0.62

Average for ^

PMV and M pro-

Nucleic acid

27.8

24.0

20.0

28.3

63

duced in Tu

rkish tobacco

TCA-nucleo-
protein

±0.55

±0.63

±0.23

±0.63

Aucuba

Potentate tomato

Nucleic acid

30.0

24.8

19.5

26.2

61

Tomato mo-

Potentate tomato

Nucleic acid

29.5

26.0

18.3

26.2

61

saic

Y.A.

Turkish tobacco

Nucleic acid

29.8

25.5

18.5

26.5

62

Rib grass

Kawala tobacco

Nucleic acid

29.2

27.0

17.2

26.2

61

Turkish tobacco

Nucleic acid

29.2

25.8

18.0

27.0

62

° The abbreviations are those used by Knight.*'
TMV = common strain of tobacco mosaic virus
M = Holmes masked strain
Y.A. = Yellow aucuba strain
The six strains include TMV, M, Y.A., JuDi (a lethal strain in young tobacco plants), GA (a green mottling
strain derived from yellow aucuba), and HR (the Holmes Rib-grass strain).

The standard errors of the mean as calculated from the data of Knight'^ and as found by Cooper and
Loring." The values are similar to those given by Black and Knight."

208 HUBERT S. LORING

TABLE IV

Adenine, Guanine, Cytidine, and Uridine Composition of Mitochondria,
Microsomes, and Nonsedimentable PNA Fraction from Normal Swiss

Mouse Liver"
(/iM per 100 mg. TCA and alcohol-ether extracted powder)

Total

Ade-

Gua-

Total

Cyti-

pyrimi-

Cytoplasmic fractions'*

nine

nine

purine

dine

Uridine

dine

Mitochondria pool II

2.6

4.9

7.5

3.7

3.4

7.1

Mitochondria pool IV

2.7

5.0

8.0

4.6

3.9

8.2

Mitochondria pool Via

2.7

4.9

7.7

4.3

2.8

7.0

Mitochondria pool VIb

3.4

5.1

8.7

4.0

2.8

7.0

Microsomes pool II

5.6

10.0

15.7

7.7

3.8

11.6

Microsomes pool III

5.0

9.4

14.4

8.4

5.6

13.9

Microsomes pool IV

4.3

8.2

12.7

6.3

4.7

11.1

Microsomes pool Via

4.2

8.1

12.4

7.8

5.2

12.6

Microsomes pool VIb

3.9

7.3

11.1

5.9

4.0

10.0

Microsomes pool VIII

4.7

9.5

14.0

7.6

5.3

12.8

Nonsedimentable pool

II

1.8

3.6

5.3

3.2

2.1

5.2

Nonsedimentable pool

IV

1.4

3.4

4.8

3.1

2.0

5.1

Nonsedimentable pool

Via

1.6

2.^

4.5

3.3

2.4

5.5

Nonsedimentable pool

VIb

1.5

3.2

4.6

Nonsedimentable pool VIII

1.6

3.1

4.6

3.2

2.6

5.6

" The means and standard deviations of the means for six separate analyses of a pool of normal
mouse liver were as follows: adenine, 2.77 ± 0.05; guanine, 4.98 ± 0.05; total purine, 7.74 ± 0.05; cyti-
dine, 4.44 ± 0.11; uridine, 2.70 ± 0.04; total pyrimidine, 7.13 ± 0 12.

* The fresh tissue from each pool was ground in 0.88 M buffered sucrose and fractionated by differential
centrifugation essentially as given by Hogeboom, Schneider, and Pallade.^' In VIb the tissue was stored
frozen for seven days before grinding.

particularly by Claude^* and by Hogeboom, Schneider, and Pallade^^ has
provided relatively characteristic ribonucleic acid-containing cell com-
ponents for analytical study. [Cf. Donace, Chapter 18, and Hogeboom and
Schneider, Chapter 21.1 Such materials from different animal species and
from several types of tissue after extraction with trichloroacetic acid and
alcohol-ether to remove acid-soluble and lipid components, respectively,
have been analyzed directly for their ribonucleic acid components'''-''^ or
have been used for the extraction of the ribonucleic acid, which subse-
quently has been analyzed.^^ Similarly ribonucleic acids prepared from
liver and other animal tissues have been extensively analyzed. 2^- 69. 72. 73

68 A. Claude, J. Exptl. Med. 80, 19 (1944); 84, 51 (1946).

89 G. H. Hogeboom, W. C. Schneider, and G. E. Pallade, J. Biol. Chem. 172, 619

(1948).
" A. Marshak, J. Biol. Chem. 189, 607 (1951).
" D. Elson and E. Chargaff, Federation Proc. 10, 180 (1951); Phosphorvs Metabolism

2,331 (1952).
" F. Leuthardt and B. Exer, Helv. Chim. Acta 36, 500 (1953).
" E. Volkin and C. E. Carter, J. Am. Chem. Soc. 73, 1516 (1951).

HYDROLYSIS — ESTIMATION OF BASES IN PNA 209

The results for the extracted ribonucleic acid preparations from several
types of tissues and animal species show a wide variation in composition
depending on the methods of extraction and analysis that are used.-^' ^^^ "• "
It is also clear that the composition of the ribonucleic acid fractions ex-
tracted from isolated mitochondria varies depending on the method used
for extraction.*'- " Thus, apparently more than one chemical species of
ribonucleic acid may be present in this tj^pe of cytoplasmic component.
The analytical results for the acid and alcohol-ether extracted mitochondria,
microsomes, and nonsedimentable ribonucleic acid fractions from mouse
liver found by the analytical procedure outlined above are summarized in
Table IV. (Unpublished experiments of the author with J. L. Fairley, H.
L. Seagran, R. S. Waritz, and M. D. Johnson.) The reproducibility of the
method is shown from the values of the standard deviations of the means
in a series of six separate analyses of a pool of normal mouse liver pow-
der. Because the variation between pools of mitochondria, microsomes,
or nonsedimentable ribonucleic acid fractions is larger than that in a single
pool, the results strongly suggest that variations occur in the ribonucleic
acid of the different cytoplasmic fractions even in animals under comparable
conditions. Under fasting conditions or in mice carrying the- Ehrlich ascites
tumor variations beyond those normally found also apparently occur.

Of considerable interest in the above-mentioned studies is the relative
amount of ribonucleic acid in mitochondria and microsomes as judged by
the molar ratios of total purine and purine and pyrimidine base to phos-
phorus present. If the assumption is made that the base to phosphorus
ratio in ribonucleic acid is one, then it can be concluded that significant
amounts of trichloroacetic acid-insoluble phosphoiTis compounds other than
nucleic acid are present in such cytoplasmic fractions and to a greater ex-
tent in mitochondria than in microsomes. In view of statements in the
literature" • '''^ that rat liver mitochondria may contain only small amounts
of ribonucleic acid, it may be pointed out that the amounts of purine
and pyrimidine bases found in trichloroacetic acid and alcohol-ether ex-
tracted mitochondria account for from 60 to 80% of the phosphorus
present.

In several instances isolated cell nuclei have also been analyzed for their
ribonucleic acid components. The results"' "• ''^ are apparently highly vari-
able depending on the method of preparation of the nuclei.

'< V. R. Potter, R. O. Rechnagel, and R. B. Hurlbert, Federation Proc. 10, 646 (1951) .
'* W. M. Mc Indoe and J. N. Davidson, Brit. J. Cancer 6, 200 (1952).

CHAPTER 6

The Separation of Nucleic Acid Derivatives by
Chromatography on Ion-Exchange Columns^

WALDO E. COHN

Page

I. Ion Exchange 212

1. Ion-Exchange Resins 212

2. Ion-Exchange Equilibria 213

a. Hydrophilic Character 213

b. The Distribution Coefficient 213

c. Nonpolar Affinity 214

d. Rate of Reaction 214

3. Ion-Exchange Chromatography 215

a. Sorption and Elution 215

b. Types of Chromatography 216

c. Scaling Up 216

II. The Separation of Bases and Nucleosides 217

1. Ionic Properties 217

2. Cation Exchange 218

3. Anion Exchange 219

a. General 219

b. Bases 219

c. Ribosides 221

III. Separation of Nucleotides 221

1. Mononucleotides 221

a. Ionic Properties 221

b. Distribution Coefficients 221

c. Elution with Dilute HCl 223

d. Use of Higher pH; the Separation of Inorganic Phosphate and Non-
nucleotide Phosphoric Acid Esters 224

e. Isomeric Nucleotides 225

(1) Choice of Anion 225

(2) pK's of Isomeric Nucleotides 226

(3) Cytidylic Acids 226

(4) Adenylic and Inosinic Acids 228

(5) Uridylic Acids 228

(6) Guanylic Acids 228

(7) Application to Digests 230

' Manuscript prepared under Contract No. W-7405-eng-26 for the Atomic Energy
Commission.

211

212 WALDO E. COHN

2. Polyphosphonucleosides 230

a. Diphosphates 230

b. Polyphosphates 232

3. Polynucleotides 233

a. Ionic Properties and Molecular Size 233

b. The Ribonuclease Digest 234

c. The Deoxyribonuclease Digest 235

IV. Separations Involving Sugar-Borate Complexing 235

1. Sugars (Borate Exchanger and Solution) 235

2. Sugar Phosphates (Borate in Solution Only) 236

a. Nucleosides 236

b. Sugar Phosphates and the Isomeric Ribose Phosphates 236

c. Nucleotides 241

V. Related Reviews 241

I. Ion Exchange
Ion-Exchange Resins^

The basic principles of ion exchange were first observed in clays and
minerals, in particular the zeolites (hydrated aluminum silicates), and led
to the use of synthetic zeolites to remove calcium ion from water in exchange
for sodium ("water softening"). These synthetic zeolites expanded the
limited range of usefulness of the natural materials, the first analytical
application being to remove armnonia from urine for colorimetric determi-
nation. It remained for the recent rapid developments of resin technology
to produce the synthetic ion-exchange materials which have made possible
the spectacular separations of the closely related members of such families
as the rare earths,' the amino acids,* and the isomeric nucleotides (Sect.
Ill.l.e). Significant among the properties of these newer materials in this
connection are: (1) strength of functional group (strong-acid cation ex-
changers or strong-base anion exchangers); (2) single functional species
(e.g., nuclear sulfonic acid devoid of phenoUc or other acid groups); (3)
use of the chemically stable polystyrene resins as the supporting matrix
(reducing side reactions of the matrix essentially to zero); (4) ability to
produce the exchangers in the form of spherical particles (with improve-
ment in hydrodynamic properties) ; and (5) control of the degree of cross-
linking (divinylbenzene) allowing the exchange of substances of high mo-
lecular weight.

Not all of these advances in technology occurred at once; hence, some
of the earlier separations of rare earths and amino acids were performed on

2 O. Samuelson, "Ion-Exchangers in Analytical Chemistry." John Wiley & Sons,
New York, 1953.

3 E. R. Tompkins, J. X. Khym, and W. E. Cohn, J. Am. Chem. Soc. 69, 2769 (1947);
B. Ketelle and G. E. Boyd, ibid. 2800.

" W. H. Stein and S. Moore, /. Biol. Chem. 192, 663 (1951); et ante.

ION-EXCHANGE CHROMATOGRAPHY 213

materials of different structure and with less satisfactory results than are
currently expected. The ion-exchange separations of the nucleic acid con-
stituents and related products had the advantage of the newer materials
from the start. Thus all that has so far been published in this field is
comparable, since all utihzed the same or very similar materials. For this
reason, the discussion of ion-exchange resins which follows will be limited
to those few which are currently in use.*

2. Ion-Exchange Equilibria-

a. Hydrophilic Character

We will consider only one type of matrix, the poljnmers of vinylbenzene
cross-linked with divinylbenzene to give three-dimensional polystyrene
beads of spherical shape. This matrix supports sulfonic acid (strong-acid
cation exchanger, such as Dowex-50 or Amberlite-120) or carboxylic acid
(weak-acid cation exchanger, such as Amberlite-IRC-50) or quaternary
ammonium (strong-base anion exchanger, such as Dowex-1, Dowex-2, and
Amberlite-IRA-400) residues attached to each one of the aromatic rings
("nuclear" substitution).* The hydrophilic character of these substituents
gives the beads, in water, the properties of gels. They swell in water up to
limits determined by the degree of cross-linking and show typical Donnan
equilibrium effects due to the nondifTusible nature of the highly ionized
substituents.

h. The Distribidion Coefficient- ^

The ionized substituents (e.g., resin SOs" H+) participate in ionic reac-
tions with diffusible ions (e.g., Na+) in the surrounding aqueous medium
according to the simple exchange reaction

1
(resin SO^) H+ + Na+^oin. <=^ (resin SOD Na+ + H+.oin.

2

With respect to the diffusible ion (Na+soin.), sorption consists in favoring
reaction 1 (e.g., by low H+soin. concentration), desorption or elution in
favoring reaction 2 (e.g., high acidity). Under any given set of conditions,
an equilibrium can be reached which may be expressed in terms of the
mass law but which is usually designated by a "distribution coefficient"

, _ moles C per g. exchanger

moles C per ml. solution

' A complete listing of commercially available ion exchangers, together with their
manufacturers, will be found in Samuelson," pp. 262-3. The same volume contains
a thorough discussion of all specifications as well as of the principles and applica-
tions of ion exchange. See also E. R. Tompkins.^

« E. R. Tompkins, Anal. Chem. 22, 1352 (1950).

214 WALDO E. COHN

where C is present in small amount compared to the bulk competing ion or
ions. The distribution coefficient is usually obtained by batch equilibra-
tion, but it is also related directly to the position of the C peak (50 % elu-
tion point) in column chromatography which, in turn, is equal to that
point in a breakthrough curve^ where the concentration of C in the effluent
is half of its concentration in the influent [(C) /(C)o = 0.5]. The ratio of the
distribution coefficients of two substances under the same set of conditions
is termed the "separation factor, "^ for it is also the ratio of the distances
from the origin to the peaks of the two substances when eluted under that
set of conditions. Thus the distribution coefficients define the peak positions.
The breadth or sharpness of each peak is related to column length and other
factors (size of particles, cross-linking, rate of flow) which are independent
of the distribution coefficient.

c. Nonpolar Affinity

The contrast in character between the polar substituents and the ben-
zenoid matrix underlies many of the anomalies in ion-exchange behavior.
Whereas the distribution coefficient of a particular solute will depend to a
large degree upon its charge, it will also depend upon any nonpolar affinity
of the solute for the polystyrene matrix and for the ions attached thereto.
The polar attractions are influenced by pH and by complex formation,
which affect the sign and degree of charge; the nonpolar attraction is rela-
tively independent of these factors. Nonpolar affinities exhibit a greater
temperature dependency than the polar. Reactions involving ions will
conform to the principles of stoichiometry ; those depending upon nonpolar
attractions will not conform so exactly and will deviate from equilibria
based upon stoichiometric considerations alone,

d. Rate of Reaction

Nonpolar interactions influence the rate of reaction. In general, the rates
of reaction between ions where at least one is "strong" (as is the case in
the sulfonic and quaternary amine resins) are more rapid and less tem-
perature-dependent than the nonpolar or "solubihty" reactions. The effect
of a slow rate of reaction upon the flow rates used in column chromtography
is adverse; a slow rate of reaction will require a slower flow rate to achieve
the symmetrical bell-shaped elution curve shown by Mayer and Tompkins*
to depend upon equilibrium conditions. Elevated temperatures can be used

^ A breakthrough curve is obtained when the saturation value of a given column for
a particular substance in a given solution is exceeded. When this occurs, the plot
of C (concentration in effluent) vs. volume is of a sigmoid nature^ • * and approaches
Co (influent concentration) as a maximum.

8 S. W. Mayer and E. R. Tompkins, /. Am. Chem. Soc. 69, 2866 (1947).

ION-EXCHANGE CHROMATOGRAPHY

215

to "sharpen" those eluticn peaks which are diffused due to noiipolar forces
as well as to shift their relative positions (see Stein and Moore'').

3. Ion-Exchange Chromatography

a. Sorption and Elution

Column chromatography in general requires two steps, the sorption of
the sample containing the components to be separated and an elution
sequence in which the various components are brought off the column sepa-
rately. The sorption step usually utilizes conditions of high affinity (high

LIQUID H£4D »ND FLOW
R4TE DETERMINED BY
LENGTH OF STEM

ION -EXCHANGE COLUMN

Fig. 1. lon-exchango column and receiver (after Tompkins).^' '

distribution coefficient) between solutes and exchanger (e.g., high charge
on the solute, low ionic strength in the solution) to bring about retention of
the sample in the topmost layers of the column (Fig. 1). Elution, on the
other hand, utilizes conditions in which a larger fraction of the constituent
in question is released from the resin (lower distribution coefficient), thus
setting up a distribution between solvent and exchanger which permits a
reasonable degree of movement of the solute do\vn the column with the
solvent flow. The conditions for elution are the opposite of those for sorp-
tion; hence, elution can be accomplished by a reduction in charge of the
sorbed constituent (by pH adjustment or complex formation or the reverse),
or by an increased concentration of competing ion (ionic strength adjust-
ment), or by increased temperature (decreased nonpolar attraction), or by
some combination of these.

9 E. R. Tompkins, J. Chem. PJduc. 26, 32, 92 (1949).

216 WALDO E. COHN

b. Types of Chromatography'^

Two methods of exploiting these factors are recognized. In "displace-
ment chromatography,"^" the chemical form of the exchanger is changed
during the elution sequence (e.g., from a hydrogen-form cation exchanger
to an ammonium-form, or from a hydroxide-form anion exchanger to a
chloride-form), and the various components trail the progressive change in
form by greater or lesser distances. This method has not had as wide-
spread practical success as the method of "development chromatography,"
also called "elution analysis," in which the salt form of the column remains
unchanged from sorption sequence to elution sequence. In the latter
method, the influent contains, as its competing, replacing, or bulk ion, the
diffusible ion of the exchanger itself. This method has been essentially the
only one applied to the separation of nucleic acid constituents. An innova-
tion, in which the eluting sequence is carried out by a gradually changing
concentration of eluting agent, instead of changing the solvent discontinu-
ously, has been termed "gradient development chromatography "^°' "

c. Scaling Up

It should be noted that a column separation is also a preparation. The
magnitude of the preparation is determined only by the cross-sectional area
of the column. Thus analytical separations can be scaled up by proportional
increase in the area of the column to prepare larger amounts of material
with no loss in resolution.^ Such large-scale ion-exchange chromatography
has been the main source of most of the new nucleotides discovered by ana-
lytical ion-exchange chromatography.

The factors and principles just discussed were known in 1948 when the
first serious efforts were made to develop precise ion-exchange chromato-
graphic separations of nucleic acid derivatives. Although the attempt was
made to find conditions giving good separations with elution curves con-
forming to those derived from equilibrium considerations, no systematic
examination of the many possible separation conditions has as yet been
made. Indeed, the stimulus for most of the developments arose from the
desire to recover degradation products as an approach to establishing struc-
tural relationships in the parent nucleic acid. The separations presented,
therefore, do not necessarily indicate the best possible apphcations of the
methods.

'" E. Glueckauf , Nature 170, 150 (1952) (as reported in Principles and Applications

of Ion Exchange).
11 H. Busch, R. B. Hurlbert, and V. R. Potter, /. Biol. Cheni. 196, 717 (1952).

ION-EXCHANGE CHROMATOGRAPHY

217

TABLE I

pK Values of Bases, Ribosides, and Nucleotides"

Bases

Nucleosides

Nucleotides'"

Anionic
(OH)

Phosphates

Cationic

Anionic (OH)

Cationic

Cationic

OH

First

Second

Purines

Adenine

4.l^.'i.^

4.22-''

9.8^'^

3.5'^-'', 3. a"

13''

3.7,4.4'

0.89

6.01,
6.4'

}

Guanine

3.2^

9.6/' 9.36*

1.6, 2.2*^

9.16, 9.5"

2.3

0.7

5.92,
6.4'

9.36,
9.7'

Hypoxanthine

(~2)M.98^

8.7", 8.9'-^

1.2S 1.5-^

8. 7-8.8^-^ ■»

J

1.54

6.04

8.88

Xanthine

(-0.8)'^

l.T-O 12'^

I, <2.5'^

5.7''A6.0i'

}

—

—

i

Pyrimidines

Uracil

(-0.5)^

9.45, 9.5'"

I

9.17,9.2^'"

—

1.02

5.88

9.43

Thymine

(~0)^

9.94, 9.9'"

I

9.8'"

—

1.6'

6.5'

10.0'

Cytosine

4.6, 4.5^'"

12.16, 12.2'"

4.22,4.1''.'"

12. 3^ ~13''

4.2^ 4.6'

0.80

5.97,
6.6'

13.2''

5-Methylcytosine

4.8/ 4.6'"

12.4'"

I

—

4.4"^

—

—

—

'^ From Levene and Bass'^ unless otherwise noted.
Ribosides and ribonucleotides except for thymidine and thymidine-5'-phcsphate. For isomeric nuc-
leotides, see Table H.

'' From .-ipectrophotometric observations. '^

■^ Alberty et al. (at 25°C.).»

" Taylor.i'

/ A. Albert, Biochem. J. 54, 646, 1953.

" Ogstnn."

Sugar group.
' 5' Deoxynucleotides: R. O. Hurst, A. M. Marko and G. C. Butler, J. Biol. Chem. 204, 847, 1953.
' Presumably same as riboside.
* Cavalieri et alA^

' Not measured but presumably similar to base.
'" Shugar and Fox."

II. The Separation of Bases and Nucleosides

1. Ionic Properties

All purine and pyrimidine bases have at least one group capable of form-
ing an ion. Most have more than one, and can ionize, under proper condi-
tions, to form either anions or cations (see Table I). The ribosides have
properties very similar to those of the bases, since the very weakly acid

'^ P. A. Levene and L. W. Bass, "The Nucleic Acids." Chemical Catalog Co., New

York, 193L
'' W. E. Cohn, unpublished observations.

'^ R. A. Alberty, R. M. Smith, and R. M. Bock, J. Biol. Chem. 193, 425 (1951).
'5 H. F. W. Taylor, J. Chem. Soc. 1948, 765.

'« L. Cavalieri, S. E. Kerr, and A. Angelos, /. .4m. Chem. Soc. 73, 2567 (1951).
" A. G. Ogston, J. Chem. Soc. 1935, 1376.
'» D. Shugar and J. J. Fox, Biochim. et Biophys. Acta 9, 199 (1952).

218

WALDO E. COHN

ribose residues do not interfere (except in the case of adenosine, where ribose
replaces the only easily dissociable hydrogen).

In acid solution, the amino groups of guanine, adenine, and cytosine are
cationic, with pK values increasing in that order (see Table I). Uracil and
thymine and their ribosides have no amino groups and are not sorbed to an
appreciable degree by cation exchangers in the hydrogen form. Hypoxan-
thine and xanthine and their ribosides, also not cationic, exhibit a small
degree of retention on cation exchangers." This retention may be ascribed
to a nonpolar attraction, seemingly general among the purines, giving them

fi

I U.^.J

300 400 500 600

VOLUME OF 2I1-HCI IN ml

Fig. 2. Separation of purine and pyrimidine bases by cation exchange in an acid
system.^'

Exchanger: Dowex-50-H+, ca. 300 mesh, 8.1 cm. X 0.74 cm. 2.

Solution: 2 N HCl, 0.6 ml./min.

Sorbed material: 0.5-1.0 mg. of each base in 7.5 ml. 2 A^ HCl. (Larger volumes of

more dilute acid may be used.)

a higher distribution coefficient than pyrimidines of equal pK values. Such
a family difference, noted in the nucleotides, ^^ may explain the greater re-
tention of adenosine over cytidine found on the carboxylic resin IRC-SO,^"
where ionic considerations alone would predict the reverse.

2. C.\TioN Exchange

In Fig. 2 is indicated the separation' of the four bases of ribonucleic acid
(such as could be derived by formic acid^ or perchloric acid hydrolysis^^)

19 W. E. Cohn, /. Am. Chem. Soc. 72, 1471 (1950).

2" W. Andersen, C. A. Dekker, and A. R. Todd, /. Chem. Soc. 1952, 2721.

=*! W. E. Cohn, Science 109, 377 (1949).

" E. Vischer and E. Chargaff, J. Biol. Chem. 176, 715 (1948).

23 A. Marshak and H. J. Vogel, J. Biol. Chem. 189, 597 (1951).

ION-EXCHANGE CHROMATOGRAPHY 219

by cation exchange, utilizing ionic-strength adjustment without any charge
adjustment (5-methylcytosine follows cytosine by a very slight margin,
whereas cytidine precedes it by a factor of 3.4'^). Uridine and cytidine have
been separated from pyridine or enzymic digests by sorption of the latter
on a sulfonic acid resin, followed by displacement with ammonia or pyri-
dine.2^'2* The possibilities of a separation of the bases by charge adjust-
ment (e.g., by increasing pH in a sodium or ammonium cycle, as Stein and
Moore^ have done for amino acids) has not been tried on the bases, but
Reichard and Estborn^'^ have utihzed this principle for the preparation of
four (N'^-labeled) deoxynucleosides in an ammonium cycle. Since adenosine
was not present in their mixture (the enzyme preparation having converted
it to deoxyinosine) , no information is at hand as to how well adenosine and
guanosme might separate in such a system. Extensive hydrolysis at the
acid-sensitive glycosidic linkage of the purines has accompanied other ex-
periments along this line.^" Such hydrolysis has also been observed with the
purine ribonucleotides in the hydrogen cycle. '^' ^'

3. Anion Exchange

a. General

The anionic properties of the bases reside chiefly in the keto groups,
which can enolize and thus develop ionizable hydrogens. While the hydroxyl
groups of ribose are also weak acids, it has seldom been necessary to resort
to the high pH required to ionize them. The attachment of ribose to the
purine and pyrimidine bases does not interfere with the acidic group (ex-
cept in the case of adenine), so that the behavior of bases and ribosides is
again very similar. The absence of glycosidic lability in alkaline regions
make the anion-exchange method preferable to cation exchange. In addi-
tion, it is possible to exploit the method of elation bj' charge adjustment,
thus removing the substances by very dilute (e.g., 0.01 M) solutions buf-
fered in the region of the pK's involved instead of the concentrated acid
solutions employed in cation-exchange separations.

h. Bases

The five bases normally considered as constituents of nucleic acid have
enolic pK's in the descending order cytosine, adenine, thymine, guanine,
and uracil. The order of elution by NH4OH-NH4CI buffers (Fig. 3) shows
two aberrations from this sequence: adenine follows guanine and uracil
precedes thymine. In the latter case, the additional methyl group may be
responsible for the higher distribution coefficient.

" D. T. Elmore, Nature 161, 931 (1948); J. Chem. Soc. 1950, 2084.

" R. J. C. Harris and J. F. Thomas, Nature 161, 931 (1948); /. Chem. Soc. 1948, 1936.

" P. Reichard and B. Estborn, Acta Chem. Scand. 4, 1047 (1950).

40r

•30

T^ZO

O 10

ADENINE

CYTOSINE

URACIL

THYMINE

GUANINE

100 200 300 400 500 600 700 800
VOLUME OF NH.OH-NH.GL BUFFER IN ml

Fig. 3. Separation of purine and pyrimidine bases by anion exchange in a chloride
system. ^^

Exchanger: Dowex-l-Cl~, ca. 300 mesh, 8.5 cm. X 0.74 cm. 2.

Solution: 0.2 M NH4OH + 0.025 M NH4CI, pH 10.6 (at a, pH -^ 10.0, CI" — 0.1

M); 0.25 ml./min.

Sorbed material: 1-2 mg. of each base in eluting buffer.

Influent

pH-

Rrmk; Acid + NH4OH buffers, formate ion« 0.01 M

lllo

it 25-

•fii

200 600

■^ N

ml . through column

Fig. 4. Separation of bases, ribosides, and nucleotides by anion exchange in a
formate system.''

Exchanger: Dowex-1-formate, 13 cm. X 0.74 cm.''.
Solutions: as shown, 0.5 ml./min.

Sorbed material : as shown. Substances in brackets were not included in this sepa-
ration but were examined separately. The dotted line indicates the method of
identification by spectrophotometric absorption ratios.

220

ION-EXCHANGE CHROMATOGRAPHY

221

.BEST CHROMATOGRAPHIC
RANGE

80 160 240 J20

RELATIVE DISTRIBUTION COEFFICIENT (ml per ml rRA-flOO TO C/C . 0 5 1

Fig. 5. Relative distribution coefficients (ml. effluent per ml. exchanger to C/Co =
0.5) of bases and ribosides as a function of pH in 0.01 M Cl~, as derived from break-
through experiments. '^

Exchanger: 1 ml. IRA-400.

(Circled letters T, C, A, U, and G are taken from Fig. 20, to show the influence of

borate ion on thymidine (no effect), cytidine, adenosine, uridine, and guanosine.

respectively.)

c. Ribosides

The similarity in the behavior of the ribosides is shown in Fig. 4, in
which a mixture of bases and ribosides has been chromatographed. This
method has been appUed to the large-scale separation of the deoxynucleo-
sides^" for preparative purposes.

A summary of breakthrough sorption experiments on bases and nucleo-
sides, for the purpose of determining relative distribution coefficients, is
given in Fig. 5.

III. Separation of Nucleotides
1. Mononucleotides

a. Ionic Properties

The nucleotides differ markedly in ion-exchange characteristics from the
bases and ribosides by virtue of the presence of the strongly acidic phos-
phoryl groups. Phosphate esters are stronger acids than inorganic ortho-
phosphate itself, with pK values of about 1 and 6 for the first and second
dissociation, respectively, as against 2 and 7 for orthosphophate. The dif-
ference in secondary ionization constants makes possible the easy separation
of inorganic phosphate from phosphoric acid esters in the region of pH 6
(see Sect. Ill.l.rf).

b. Distribution Coefficients

Among the nucleotides, anionic behavior is predominantly a function of
the phosphate group, but this is modified by the ionic and nonpolar proper-
ties of the nucleoside residues. An assessment of the total charge due to the

222

WALDO E. COHN

ionization of phosphate and of base groups as a function of pH is plotted in
Fig. 6. The order of elution predicted from these considerations is subject
to modification by the greater attraction of the exchanger for purines over
pyrimidines mentioned earlier. Thus, in order to superimpose the separa-
tions found (Fig. 7) on those predicted solely from the net-charge criterion
(Fig. 6), it must be assumed that the purine compounds are about three
times as strongly sorbed as the pyrimidines.^^ This factor remains fairly
constant over a wide range of pH (and hence of net charge), thus indicat-
ing that it probably has to do with intrinsic properties of the solutes — which
differ only in the base constituents — and not with ionic properties which
are pH dependent.

The variation of elution position with pH (Fig. 7) is of some interest,
particularly in view of the increasing use of volatile salts for elution rather
than mineral acids. At neutral pH, similar purine nucleotides appear to-
gether and similar pyrimidine nucleotides appear together, the latter sep-
arated (by the factor of about 3) from the former. It may be concluded that
the separation of each pair at more acid pH is due to the development of a
cationic group (the ammonium group) , causing adenylic to precede guanylic
acid (pK 3.7 vs. 2.3) and C3d:idylic to precede uridylic acid (pK 4.2 vs.

Fig. 6. Net charge per molecule of ribonucleotide as a function of pH, calculated"
from the pK values quoted in Levene and Bass."*

ION-EXCHANGE CHROMATOGRAPHY

223

none). Since uridylic acid has no cationic group, we express the data in
terms of iiridyhc acid ehition (the parallelism of ionosinic acid should be
noted). The relative positions of cytidylic and of adenylic acids are main-
tained at all pH values; those of uridylic and guanylic acids become re-
versed at low pH due to the development of a charge upon the latter. The
same is true of adenylic acid and its derivative, inosinic acid.

c. Elution with Dilute HCl

A separation of four purified ribose nucleotides by dilute HCl at pH 2.5
is shown in Fig. 8.^^ The mixture was sorbed from dilute ammoniacal solu-
tion, with total chloride concentration less than 0.01 J\I, followed by 0.01
M NH4CI until the effluent pH fell to 7. (This NH4CI wash removes any

-o..^ A-,^

<r^ o,,^__^ ^A^ 1

• \.

iV

A GU 3-

*\ ^5

■^

^°\^

I

\

1
1

A

• A 0

1 1 ' 1 ' ' 1 ' 1 • '1 ' 1 ' 1

0.04 0.07 0.1 0.2 0.4 0.7 1 2^4

RELATIVE DISTRIBUTION COEFFlClE

NT

Fig. 7. Observed relative distribution coefficients (positions of elution peaks) of
ribonucleotides as functions of pH, derived from column chromatography (see Fig. 8).

bicarbonate present and thus prevents CO2 formation when the acid is
started, as well as eliminating a long delay until all amphol^'tes on the
resin are neutralized; it is generally advisable when proceeding from alka-
line to acid solutions.) The order predicted from Fig. 6 is not observed with
respect to the uridylic-guanylic acid sequence, nor with respect to the dis-
tance between cytidylic and adenylic acids; the aberrations are accounted
for by the factor discussed and are set forth in Fig. 7. Fig. 8 also illustrates
the use of specific ultraviolet absorption characteristics to follow the course
of a separation and the extent to which the sj^stem follows the Mayer
and Tompkins eciuation,* derived upon the assumption that equilibrium
between exchanger and solute is achieved throughout the elution sequence.
The separation of the deoxynucleotides (Fig. 9-^"-^) follows a similar pat-
tern with thymidylic taking the place of uridylic acid. From the deoxy-

2' E. Volkin, J. X. Khym, and W. E. Cohn, J. Am. Chem. Soc. 73, 1533 (1951).

28 R. L. Sinsheimer and J. F. Koerner, Science 114, 42 (1951).

" R. O. Hurst, J. A. Little, and G. C. Butler, J. Biol. Chem. 188, 705 (1951).

224

WALDO E. COHN

cytidylic acid peak, deoxy-5-methylcytidylic acid can be separated [see
Sect. III.l.e(3)].

d. Use of Higher pH; the Separation of Inorganic Phosphate and Nonnucleo-
tide Phosphoric Acid Esters

Elution of nucleotides by dilute acids involves both principles of elution
mentioned earlier, charge adjustment (pH dependent) and ionic strength
adjustment (chloride-ion dependent). With increase in pH it is also neces-
sary to increase the chloride ion in order to maintain a reasonable rate of
elution; thus at pH 5.6, a chloride concentration of about 0.02 M is neces-
sary to maintain rates of elution comparable to those indicated for 0.003
N HCl in Fig. 8. Under these conditions, cytidylic and uridylic acids ap-
pear together, as do guanylic and adenylic acids (see Fig. 7). Inorganic
phosphate, which appears in the adenylic acid region at pH 2-3, precedes
the pyrimidine nucleotides by a wide margin at pH 6. Creatine phosphate
and sugar phosphates behave similarly in the acid region but are separable
from inorganic phosphate in the neutral region by virtue of the lower sec-

10 2.0 30

Uten HCl (from pH odiuttment )

Fig. 8. Separation of ribonucleotides by anion exchange in a chloride system
showing conformity with calculated theoretical curves (see text).''
Exchanger: Dowex-2-chloride, 2.5 cm. X 0.94 cm.*.
Solution: 0.003 A^ HCl, 0.8 ml./min.
Sorbed materials: purified nucleotides, in amounts shown.

ION-EXCHANGE CHROMATOGRAPHY

225

ondary pK of phosphoric acid esters. The data" on the separation of these
substances are summarized in Fig. 10.

e. Isomeric Nucleotides

(1) Choice of Anion. Other acids and their salts may be used in place of
chloride. In general, those anions \nth a greater affinity than chloride for
the exchanger (e.g., Br~, SO4 ) vaW effect a given rate of elution with
lower ionic strength or lower hydrogen-ion concentration or both; those
with a lesser affinity (e.g., formate, acetate) \\\\\ require a lower pH or a
higher anion concentration. The effectiveness of a given anion in replacing
a specific solute will depend upon its affinity for the exchanger, which is a
function of pK, and upon intrinsic properties. Formate has received much
attention,"' ^^ partly because of the ease of obtaining solutions of stable pH
in the region of 3-4 and partly because of the ease with which it and its

K

6X>-

eg

e

P 2

CYTIOYLIC

f /

~ X

/

Z a.

/

tt 1

y

5X3-

I 8
< 0

r

4.0-

I 1

1 h

3.0-

1 1

ADENYLIC

THYMIOYLIC

GUANYLIC

DEOKY- U

RIBOSIDES

■^

ZO-

BASES

n

i / \

/

Zl \

\ \

8

0 1 1

1

»- 1

1.0-

g 1

A

K

V L

,

/ V

v.^

— ^ v_

_• \

0 500 1000 1500 2000 Z500

m. FROM pH EQUILIBRATION

Fig. 9. Separation of deoxyribonucleotides by anion exchange."
Exchanger: Dowex-1-chloride, 200-400 mesh, 8 cm. X 0.72 cm.^.
Eluting solution: HCl as shown, 1 ml./min.

Sorbed material: mixed deoxynucleotides isolated by preliminary ion exchange
from 150 mg. DNA digested with deoxyribonuclease followed by intestinal phos-
phatase plus arsenate (method of Klein and Thannhauser used by Volkin ei alP
and Sinsheimer and Koerner^*).

226

WALDO E. COHN

INOBG P

ao-2-P

uB-J-P GU ACIDS

AD -5

-P i

A0-3-P

RiBP

X

\ "^

"--. ^

HEX P\^

CPEi

INE-P \

""^

"~~~^\

-.^

"~^ -^ \

^^ -^

\ ^C,

^'\^x \

S-METHYL- ■
DEOx'CrT aCiO

OEOXTCrT

°\ / \

\

RI9-2-P / -'

//

D

P^

\

//

, \

//

5 7 10

Fig. 10. Relative distribution coefficients of various nucleotides and organic phos-
phates relative to inorganic phosphate as a function of pH, derived from anion-ex-
change chromatography" (see Fig. 7).

ammonium salt can be volatilized. Acetate was used in the first ion-ex-
change separations of the 2' and 3' nucleotide isomers.'^- ^"'^^ On the other
hand, the dilute HCl solution has an advantage, if subsequent concentra-
tion by resorption and re-elution is considered,'^' ^' because of the low ionic
strength ; the more concentrated formate or acetate solutions must usually
be diluted before resorption, if such is indicated.

(2) pK's of Isomeric Nucleotides. The 5'-, 2'-, and 3'-phosphate isomers
of the various ribonucleosides appear in the order named upon elution un-
der the proper conditions for separation. It would appear, compositional
factors being the same, that the pK values of the phosphate and amino
groups, which differ slightly among the isomers, might be responsible for
the separations to be presented. These values are not known for all sets of
isomeric nucleotides; those so far reported are collected in Table II. It will
be seen that, in general, they vary in magnitude and in direction as one
might infer from the order of elution, and hence may properly be held to
be responsible for the separations.

(3) Cyiidylic Acids. The cytidylic acid isomers are eluted rapidly with
0.001 A^ HCl solution and with 0.01 A^ formic acid;^" hence, a pH of 3 or

'" J. X. Khym and W. E. Cohn, unpublished observations.

31 W. E. Cohn, J. Am. Chem. Soc. 72, 2811 (1950).

32 H. S. Loring, N. G. Luthy, H. W. Bortner, and L. W. Levy, /. Am. Chem. Soc.
72, 2811 (1950).

" W. E. Cohn, /. Am. Chem. Soc. 71, 2275 (1949).
3* W. E. Cohn, /. Cellular Comp. Physiol. 38, Suppl. 1, 21 (1951).
« L. F. Cavalieri, J. Am. Chem. Soc. 74, 5804 (1952); 75, 5268 (1953).
3« H. Wassermeyer, Z. physiol. Chem., 179, 238 (1928); see Levene and Bass,'^ p. 231.
37 P. A. Levene and H. S. Simms, J. Biol. Chem. 65, 519 (1925); 70, 327 (1926); see
Levene and Bass,!^ pp. 223, 228.

NH2 Group

Secondary Phosphate

4.12," 4.16''

—

3.63," 3.4"^

—

3.74,'' 3.89," 3.8'', 4.4*

6.05,"

6.48," 6.2^ 6.4*

3.80," 3.79"

6.15,"

6.21"

3.65,° 3.56,* 3.70'

5.88,"

6.06," 6.01'

—

6.86"

—

6.45"

—

6.14"

4.45,/ 4.6^

—

4.6,/ 4.8»

—

4.11,'' 4.15," 4.25»>

—

4.4"

—

4.44,M.45»'", 4.6*

6.6*

4.30, '■^•"4.36^

6.2'

4.3, M.28,'^-"-' 4.16''

6.0' ■•

ION-EXCHANGE CHROMATOGRAPHY 227

TABLE II
pK Values of Isomeric Nucleotides and Related Substances

Adenine
Adenosine

Adenosine-5 '-phosphate
Adenosine-2 '-phosphate
Adenosine-3 '-phosphate
H3PO4

/3-Glycerophosphate
Glucose-1 -phosphate
Cytosine
5-Methylcytosine
Cytidine

5-Methylcytidylic acid
Cytidine-5' -phosphate
Cytidine-2' -phosphate
Cytidine-3' -phosphate

" Alberty et alM (25°, 0.15 M NaCl).

" Kuna, M., U.S. Atomic Energy Comm. Rept. ORNL-318 (May 10, 1949); quoted by Cohn."

"^ Cavalieri.35

'' Wassermeyer'* (see Levene and Bass," p. 231).

* Levene and Simms" (see Levene and Bass,'^ pp. 223, 228).
/ Shugar and Fox" (spectrophotometric).

" Cohn'3 (spectrophotometric).

* Fox el a!.'* (spectrophotometric).
' Deoxy compound.

' Loring et al.'^

* For deoxy 5' nucleotides: R. O. Hurst, A. M. Marko, and G. C. Butler, J. Biol. Chem. 204, 842 (1953)

higher is necessary for their separation. Formate and acetate are more
manageable from the standpoint of pH control and buffer capacity in this
range and give good separations of the 5'-, 2'-, and 3'-phosphates of cytidine,
in that order. Attempts to prepare large amounts with 0.05-0.1 M formic
acid have given incomplete separation, apparently owing to an overloading
of the columns as well as to an excessively rapid elution with the higher acid
concentration.^^ Deoxycytidylic acid (also a 5' ester) is inseparable from
the ribose analogue under these conditions, but can be separated Avith the
aid of borate complexing (see Sect. IV.2.c). Deoxy-5-methylcytidylic acid
precedes the 5' nucleotide(s) by a small margin. Fig. lla^^ demonstrates
the separation of these substances with acetate buffers. This experiment is
similar to that in which were discovered the existence of 5-methylcytidylic
acid^^ in deoxycytidyhc acid and the existence of tw^o isomers (2' and 3')

38 J. J. Fox, L. F. Cavalieri, and N. Chang, J. Am. Chem. Soc. 75, 4315 (1953).

" H. S. Loring, H. W. Bortner, L. W. Levy, and M. L. Hammell, /. Biol. Chem.

196, 807 (1952).
" W. E. Cohn and E. Volkin, Nature 167, 483 (1951).
« W. E. Cohn, /. Am. Chem. Soc. 73, 1539 (1951).

228

WALDO E. COHN

URIDYLIC ACIDS

LITERS THROUGH COLUMNS

Fig. 11. Separation of (a) cytidylic and (b) uridylic acids by anion exchange.''
Exchanger: Dowex-1 -acetate (a), -formate (b), 200-400 mesh, (a) 11.3 cm. X
0.86 cm.^ (b) 9.5 cm. X 0.76 cm.^.

Solutions: (a) 0.025 M HAc + 0.025 M NaAc, (b) 0.04 M Na formate + 0.0004
M formic acid.

Sorbed materials: (a) crude deoxycytidylic acid fraction"' '" (see Fig. 9), con-
taining deoxy-5-methylcytidylic acid (shown as "X"), plus cytidylic acids a and
b (2' and 3') ; (b) uridylic acids a and b (2' and 3')-

of cytidylic acid. Similar separations are achieved in 0.01 A^ formic acid.
In Fig. lib is demonstrated the separation of the 2' and 3' isomers of
uridylic acid in a formate system. These are not well separated in acid
systems.

(4) Adenylic and Inosinic Acids. The 5'-, 2'-, and 3'-phosphate isomers
of adenosine are easily separable by 0.002 A'' HCP®' 2^' ^^ or by 0.1 M formic
acid.'^' *° (A separation at pH 5.5 in a chloride system has also been de-
scribed,^'* At this pH, the inosinic acid isomers, included in the experiment,
precede the corresponding adenylic acids just as do the uridylic acids;
see Fig. 12.) Fig. 13 shows the separation of a partially deaminated mixture
of the three adenylic acids by formic acid.

(5) UridyUc Acids. It will be noted that the inosinic acids not only follow
the adenylic acids at a considerable distance but are not well separated
by formic acid. The uridylic acids show the same behavior (see Fig. 5^').
Hence, in practice, resort has been made, for the separation of uridylic acid
isomers after the removal of the cytidylic and adenylic acids, to buffer
systems of higher pH and ionic strength (see Fig. 11). This also avoids
entangling the uridylic acids with the guanylic acids; it will be recalled
(see Fig. 7) that these two groups are not well separated at low pH.

(6) Guanylic Acids. The guanylic acid isomers separate well at all pH
values tested. Separations at pH 2.5 (HCl) and 5 (formate) have been re-
ported,^'' ^^ both as part of the analysis of an alkaline digest of PNA which
includes the four isomeric pairs of nucleotides.

ION-EXCHANGE CHROMATOGRAPHY

229

1.2

Adenylic
5

a.

425mq

o 10

ID

-

Inosinics
a b

< 0.8

.

I07mg. I0"5mg,

»

>

Inosinic

p.

1-

z 0.6

1 ^
1-., n

ng.

■

JT

< 0.4

1-

Q-

-

Adenylic

Adenylic

Ratio
O.Df5f

02

.

[

120 mg

b

\J

j^

— 1 1 —

T ■ I^

123 mq

-

0 1.0 2.0 3.0 4.0

LITERS THROUGH COLUMN

Fig. 12. Separation of three adenylic acids and the corresponding inosinic acids by
anion exchange at pH 5.5 in a chloride system.'"

Exchanger: Dowex-1-chloride, 250-500 mesh, 9.5 cm. X 0.9 cm.^.
Solution: 0.02 M NaCl in 0.01 M acetate buffer, pH 5.5, at 1.2ml./min.
Sorbed material: commercial adenylic acids (ca. 10 mg. each), partially deami-
nated with nitrous acid.

LITERS THROUGH COLUMN

Fig. 13. Separation of three adenylic acids and the corresponding inosinic acids
by anion exchange with formic acid.''

Exchanger: Dowex-1-formate, 250-500 mesh, 13 cm. X 0.74 cm.^
Solution: 0.125 M formic acid, 1.5 ml./min.
Sorbed material: same as in Fig. 12.

230

WALDO E. COHN

LITERS THROUGH COLUMN

Fig. 14. Separation of the isomeric 2', 3', and 5' nucleotides of ribonucleic acid by
anion exchange in a formate system. '^

Exchanger: Dowex-l(X 2%) -formate, 400 mesh, 5 cm. X 0.82 cm.^.
Solution: Formic acid, ammonium formate as shown.

Sorbed material : Combined alkaline digest and 5' nucleotides (diesterase digest)
prepared as follows: 25 mg. calf liver PNA + 1 ml. 0.5 A^ NaOH, 37°C., 18 hr.;
diluted to 50 ml.; added 2 mg. each of 5' nucleotides (prepared as described in
Fig. 16) + 0.5 meq. NH4 formate + NH4OH to 0.1 M.

(7) Application to Digests. A demonstration of the separation of the 5',
2', and 3' isomers of all four ribonucleotides is shown in Fig. 14. ^^ It has
been customary, in order to save time, to double the concentration of for-
mate and lower the pH with formic acid after removal of the uridylic acids
(see Fig. 15).

The use of this sytem in the analysis of unknown digests and of the usual
alkaline digest is shown in Fig. 15.'*'' This is the actual experiment in which
the 5' nucleotides were first isolated from an enzyme digest of PNA (See
also Cohn and Volkin^^ and Fig. 16).

2. POLYPHOSPHONUCLEOSIDES

a. Diphosphates

The mixed 2', 5'-, and 3',5'-diphosphates of cytidine and uridine have
been recovered by ion exchange from digests of PNA with rattlesnake
venom freed of 5' monoesterase.^^- ^^ These digests also contain the 5'
nucleotides of all four bases, and 3' pyrimidine nucleotides, and all four

« W. E. Cohn and E. Volkin, Arch. Biochem. and Biophys. 35, 465 (1952).
" W. E. Cohn and E. Volkin, /. Biol. Chem. 203, 319 (1953).

ION-EXCHANGE CHROMATOGRAPHY

231

GUANYLIC
5'o b

01 02 0 01 0.2

LITERS THROUGH COLUMN

Fig. 15. Analysis of (a) enzymic, (b) alkaline digest of calf liver PNA by anion
exchange.^"

Exchanger: Dowex-l -formate, 200-400 mesh, 7 cm. X 0.88 cm. 2.
Solutions: as shown.

Sorbed material: (a) 100 mg. (20 ml.) PNA + 7 mg. ribonuclease digested 5 hr.,
titrated with (112 /.leq.) NaOH to hold at pH 7.2; diluted to 25 ml.; 12 ml. + 0.8
ml. 0.1 M arsenate + 6 ml. (20 mg.) intestinal phosphatase; digested 90 min., ti-
trated with 1.5 ml. 0.05 M NaOH to hold at pH 8.5; diluted to 25 ml.; acidified to
pH 2.9 with formic acid ; centrif uged. Supernatant plus NH4OH sorbed on column,
(b) 4 ml. same PNA solution + 2 meq. NaOH; digested 15 hr. at 37°C.; diluted to
25 ml. with 2 meq. NH4 formate + 2 meq. NH4OH; sorbed on same column sub-
sequently.

nucleosides, the adenosine appearing as inosine.'*' The recovery of these
components and the mode of isolation and discovery of the diphosphates
is shown in Fig. 16.

The positions of the two diphosphates are of some interest. It might be
expected that the differentiation in ion-exchange behavior of these sub-
stances would depend on the bases to the same extent as in the monophos-
phates and thus bring about the same order of elution as is observed in the
monophosphates in pH ranges adjusted to offset the increased number of
acid groups. That such is the case is indicated by the order of the two
pyrimidine diphosphates in Fig. 16 and by the fact that adenosine-5'-
pyrophosphate (ADP) comes between them (see Fig. 17, in which the peak
positions of several substances examined independently are indicated by
arrows). The position of inorganic pyrophosphate between ADP and uridine
diphosphate is also similar to orthosphosphate in the monophosphate series.
It is also apparent that the diphosphates as a group are sufficiently strongly
bound to the anion-exchange resin to follow the last of the monophosphates,
overlapping only slightly with guanylic (and inosinic) acid. (The behavior

232

WALDO E. COHN

of riboflavin phosphate is ascribed to its added nonpolar affinity.) The
factors discussed in Section III. 1.6 may cause some shifting of the order of
elution in other pH ranges.

< 0.6-

■ 0 0

LITERS THROUGH COLUMN

Fig. 16. Analysis of snake venom diesterase digest of PNA by anion exchange."
Exchanger: Dowex-1 -formate, 400 mesh, 5.8 cm. X 0.9 cm.''.
Solutions: as shown, at 0.5 ml./min.

Sorbed material: 16 mg. calf liver PNA + 1 ml. 0.05 M MgCh + 1 ml. snake
venom diesterase preparation (rattlesnake venom freed of 5' monoesterase by
the m.ethod of Butler"), total volume 5 ml.; pH 8.6 maintained by addition of 0.02
A^ NaOH for 7 hr. at 25°C.; diluted to 100 ml. with 1 meq. NH4OH; sorbed.

h. Polyphosphates

Similarly, ATP follows uridine-3' , 5'-diphosphate (and also uridine-5'-
pyrophosphate^^) , and it has been demonstrated^'' that the entire group of
triphosphorylated nucleosides may be separated from the diphosphorylated
ones as well as the latter are from the monophosphates. The great differ-
ence in affinity has been exploited to effect rapid separations of adenosine-
5'-phosphate from ADP and ATP."

Finally, the position of hexametaphosphate may be noted in Fig. 17 as
further evidence that gross affinity increases with the degree of polyvalency
of the ion involved.

'' R. B. Hurlbert, and V. R. Potter, J. Biol. Chem. 209, 1 (1954).
« W. E. Cohn and C. E. Carter, J. Am. Chem. Soc. 72, 4273 (1950).

ION-EXCHANGE CHROMATOGRAPHY

233

3. Polynucleotides

a. Ionic Properties and Molecular Size

The polynucleotides derived from nucleic acid by the action of enzymes,
e.g., deoxyribonuclease on DNA^^ '*'' or ribonuclease on PNA,^^'*" or by
brief acid treatment^" can be separated by ion-exchange, if consideration
is given to their greater size and charge. These properties influence the rates

02 09
LITERS THROUGH COLUMN

Fig. 17. Separation of various phosphates by anion exchange in cliloiide system."
Exchanger: Dowex-l-chloride, 200-400 mesh, 10 cm. X 0.76 cm.''.
Solution: HCI + NaCl as shown, 1 ml./min.

Sorbed material: 2-5 mg. of each substance shown. (Substances with peak posi-
tions indicated by arrows onh' were not included in the separation shown, but
were examined on similar columns or the same column independently.)

of diffusion and the distribution coefficients. The latter must be compen-
sated for, as in the simpler polyphosphates, by increased acidity or ionic
strength. To accommodate the large size, it has been found advantage-
ous,^^' ''^ in ^vorking \vith polynucleotides, to utilize anion exchangers of a
lower degree of cross-linking (divinylbenzene content) than the so-called
"standard" resins although successful separations have been reported using

^« R. L. Sinsheimer and J. F. Koerner, J. Am. Chem. Soc. 74, 283 (1952).

" R. L. Sinsheimer. ./. Biol. Chem. 208, 445 (1954).

^» W. E. Cohn, D. G. Doherty, and E. Volkin, in "Phosphorus Metabolism" (Mc-

Elroy and Glass, eds.) Vol. II. Johns Hopkins Press, Baltimore, 1952.
^9 E. Volkin and W. E. Cohn, ,/. Biol. Chem.. 205, 767 (1953).
'» R. B. Merrifield and D. W. Woolley, J. Biol. Chem. 197, 521 (1952).

234

WALDO E. COHN

"standard" material.''^' *'^- ^^ Since the commercially available resins have
varied over quite a range of divinylbenzene content, it is not possible to
say, with respect to some reported separations, just what degree of cross-
linking existed in some of the exchangers used. It is known that the poly-
nucleotides of the ribonuclease digest, readily separable up to at least tetra-
nucleotides on 2 % divinylbenzene material in a chloride system,^^ were not
separated on 10% divinylbenzene resin by chloride;^^ very broad peaks,

Fig. 18. Separation of the products of ribonuclease digestion of PNA.*'
Exchanger: Dowex-l(2% DVB) -chloride, 400 mesh, 15 cm. X 3.7 cm. 2.
Solution: HCl + NaCl as follows: I, 0.005 N HCl; II, 0.01 N HCl; III-IX, 0.01
iVHCl + 0.0125, 0.025, 0.05, 0.1, 0.2, 0.3, and 1 A^ NaCl, respectively; X, 2 A^HCl.
Sorbed material: 700 mg. calf liver PNA -|- 10 mg. ribonuclease in 105 ml. H2O,
22 hr. at 37°C., + NaOH as required to keep at pH 7.0; pH lowered to 2.0; chilled;
centrifuged; supernatant made alkaline with NH4OH; sorbed.

indicating a poor degree of equilibration and leading to overlapping of
components, were obtained (qualitatively similar results have been reported
with polypeptides") .

h. The Ribonuclease Digest

An example of the separation of polynucleotides by anion exchange is
shown in Fig. 18.'*^ In this separation, the concentration of HCl was held
at 0.01 A'' (after a brief sequence at 0.005 A^) with stepwise increments in
NaCl to remove the more strongly sorbed polynucleotides. Merrifield and
Woolley^" used stronger HCl rather than NaCl to remove polynucleotides

" C. E. Carter and W. E. Cohn, J. Am. Chem. Soc. 72, 2604 (1950).
" S. Moore, personal communication.

ION-EXCHANGE CHROMATOGRAPHY 235

formed by brief acid hydrolysis. Either system (increasing NaCl at constant
pH or increasing HCl) can be adapted to gradient development and other
anions (e.g., formate) can be used, as has been demonstrated for the acid-
soluble polyphosphates.^'

It is of some interest to compare the relative positions of the polynu-
cleotides in the elution sequence at pH 2 in a chloride system with those
predicted from the positions of the mononucleotides at this pH. Polynu-
cleotides ending in C3d:idylic acid (e.g., AC and GC) precede those ending
in uridylic acid, just as cytidylic acid itself precedes uridylic acid. The
nature of the purine nucleotide attached to the end pyrimidine nucleotide
exerts an influence predictable from its own behavior, but secondary in
importance. Thus, AC precedes GC, AGC precedes GGC, and AAGC
precedes AGGC. From the observation that GAC precedes AGC, it would
appear that the third nucleotide exerts still less influence than the second
one. It should also be noted that the number of phosphate groups exerts
the expected influence; thus AC precedes AAC and GU precedes GGU.
Finally, it may be remarked that the elimination of the terminal (second-
ary) phosphoryl group from a dinucleotide makes the residual dinucleoside
monophosphate behave like a mononucleotide. Thus, dephosphorylated
GC appears in the cyiidylic acid region. The behavior of a dephosphorylated
trinucleotide has not been investigated.

c. The Deoxyrihonuclease Digest

The results of Sinsheimer,''^ obtained on the deoxyrihonuclease digest of
DNA at pH 4-5.5 in chloride and acetate systems, are in general agreement
with the above. The order of elution of dinucleotides of cytidylic acid is
CC, TC (plus CT), AC, and GC (plus CG). From his earlier studies,^^ it is
known that the order of deoxjonononucleotides in this pH range is C, T,
A, and G. In the trinucleotide region, the order is CCC, CCT, and CTT;
these precede GG. In every case where one cytidylic acid residue is replaced
by 5-methylcytidylic acid, the polynucleotide is advanced in position (it
will be recalled, as in Fig. 11, that 5-methylcytidylic acid precedes cyti-
dylic). Thus MCC precedes CCC, MCT precedes CCT, etc.

IV. Separations Involving Sugar-Borate Complexing

1. Sugars (Bor.\te Exchanger and Solution)

Although the sugars have acid ionization properties which could con-
ceivably be used in ion-exchange manipulations, these constants lie so far
on the alkaline side (ca. 13) as to be relatively unattractive or unmanage-
able. However, just as the use of complex-forming acids rendered the

" H. Schmitz, R. B. Hurlbert, and V. R. Potter, J. Biol. Chem. 209, 41 (1954).

236 WALDO E. COHN

strongly sorbed rare earth ions more easily handled, so has it been found
that the complexes which sugars form with borate ion possess acidic prop-
erties of sufficient strength and of sufficiently differing quahties to permit
practical ion-exchange chromatography.

Although the neutral sugars ribose and deoxyribose are seldom encoun-
tered in the usual nucleic acid digests, there exist other situations in nucleic
acid investigations (e.g., biosynthesis and precursor studies) in which it is
desirable to have at hand a method for the isolation of one or more sugar
components. The separation of a mixture of hexoses and pentoses on a
borate-anion exchanger with borate eluting solutions, as originally de-
veloped by Khym and Zill,"' " is shown in Fig. 19. Similar separations
have been demonstrated for specific groups of monosaccharides," for di-,
tri-, and tetrasaccharides,^^ and for sugar alcohols" (as well as the uronic
acids, which do not require borate for separation^).

The dependence of these separations (i.e., of the distribution coefficients)
upon the strength of the borate complex and the dependence of this, in turn,
upon pH and the structural details of the sugars (e.g., cts-glycol groups,
furanoid or pyranoid forms, etc.) is discussed by Khym and Zill.^^

2. Sugar Phosphates (Borate in Solution Only)

a. Nucleosides

This method has been applied to the nucleosides by Jaenicke and von
DahP' and by Khym and Cohn^" (see Fig. 20). It is possible to increase
markedly the sorption of the ribonucleosides, whereupon such weakly
sorbed nucleosides as cytidine and adenosine become more strongly sorbed
and more easily separable. The difference between deoxynucleosidesand ribo-
nucleosides with respect to borate complex formation can be exploited to
facilitate their separation (see Fig. 21 in comparison to Fig. 4).

h. Sugar Phosphates and the Isomeric Rihose Phosphates

In the usual acid eluting system, the sugar monophosphates behave in
an almost identical fashion and are not well separated from one another.
The borate complexes, however, have made it possible to separate various
hexose phosphates from each other and from ribose-5-phosphate, the latter
being strongly affected by borate. The separation of a mix-ture of sugar

6« J. X. Khym and L. P. Zill, J. Am. Chetn. Soc. 73, 2399 (1951).
--"J. X. Khym and L. P. Zill, /. Am. Chem. Soc. 74, 2090 (1952).
6« G. R. Noggle and L. P. Zill, Arch. Biochem. and Biophys. 41, 21 (1952).
" L. P. Zill, J. X. Khym, and G. M. Cheniae, J. Am. Chem. Soc. 75, 1339 (1953).
'8 J. X. Khym and D. G. Doherty, /. Am. Chem. Soc. 74, 3199 (1952).
" L. Jaenicke and K. von Dahl, N aturwissenschajten 39, 87 (1952).

ION-EXCHANGE CHROMATOGRAPHY

237

LITERS THROUGH COLUMN

EiG. 19. Separation of sugars by anion exchange in a borate system. ^^
Exchanger: Dowex-1 -borate, 11 cm. X 0.84 cm. 2.
Solution: tetraborate (K2B4O7) as shown at 1 ml./min.
Sorbed material: 2.5 mg. ribose, 5 mg. of other sugars in 10 ml. 0.01 M K2B4O7

0 4 0 6 08 16

LITERS THROUGH COLUMN

Fig. 20. Separation of ribonucleosides by anion exchange in a chloride system
with borate present'" (see Figs. 3, 4, 5, 21, and 22).

Exchanger: Dowex-1-chloride, 200-400 mesh, 11 cm. X 0.85 cm.'*.

Solution: 0.03 M KCl + 0.02 M K2B4O7 .

Sorbed material: ca. 2 mg. of each nucleoside in 10 ml. 0.01 M K2B4O7 .

238

WALDO E. COHN

0 6 08 0 0

LITERS THROUGH COLUMN

Fig. 21. Separation of uridine and thymidine by anion exchange in a borate sys-
tem^o (see Figs. 3, 4, 5, and 20).

Exchanger: same as in Fig. 20 but borate form.

Solutions: K2B4O7 as shown, 1 ml./min.

Sorbed material : ca. 3 mg. each of thymidine and uridine in 10 ml. 0.01 M K2B4O7

OOOI*NM/)M

"oom/iLeoT 0002s wNn,oM-|o,oo2s«Nfi,(x|"NM,ci "]

^ecgJco'se OOOliK^O, O000O,Vk^,0,

GtUCOSE-6-PO,
INCRGANIC-

"-i

4 M OsJoT

V

'4 OOiWMCi 1?°

1 Toe

i..

04 08 di 02 Q< 0 02 06 0 04 08 12 5 02 04 q Ci 06
ELUTINC solution ( LITERS!

Fig. 22. Separation of sugar acids and sugar phosphates by anion exchange in
chloride system with borate present*" (see Fig. 20).

Exchanger: Dowex-1-chloride, ca. 300 mesh, 12 cm. X 0.86 cm. 2.
Solutions: chloride and tetraborate as shown at 3.5 ml./min.
Sorbed material: 2-10 mg. of substance as shown.

JL — . 0 04 V NH^Cl — ]

ELUTING SOLUTIONS ILlTERSI

F^k;. 23. Separation of isomeric ribose phosphates by anion exchange in a chloride
system with borate present*' (see Figs. 20, 22, and 24).
Exchanger: same as in Fig. 22.

Solutions: chlorides and tetraborate as shown at 2 ml./min.
Sorbed material: 3-8 mg. of each substance shown in 10 ml. of 0.001 M NH4OH.

Rib-2-P04
27

Rib- 4 -PO4

^

0.005 M No,S04

Fig. 24. Separation of isomeric ribose phosphates by anion exchange in a sulfate
system with borate present*^ (ggg Figs. 20, 22, and 23).
Exchanger; same as Figs. 22 and 23, in sulfate form.

Solutions: 0.0018 M NaaSOi + 0.0018 M Na2B407 (for ribose-4-phosphate and
ribose-2-phosphate) , 0.005 M Na2S04 (for ribose -3-phosphate) .
Sorbed material: (a) 10 mg. commercial adenylic acid, 1 g. Dowex-50-H, 1 ml.
H2O; heated at 100°C. for 20 min., with stirring; filtered; NH4OH added to super-
natant; sorbed; (b) same as (a), but doubled quantities and heated for 2 hr.

239

240

WALDO E. COHN

phosphates and related substances, utihzing borate complexing, is shown
in Fig. 22.^° The separation of ribose-2- and-3-phosphates, which was an
essential part of the preparation and characterization of these two sub-
stances and of their identification as parts of the adenylic and guanylic
acids a and b, respectively, is showTi in Fig. 23.*'' ^^

In the course of these separations, it was found that an unexpected
isomer of ribose phosphate appeared®^ • *^ whenever ribose-2- or 3-phosphate

DESOXYCYT

-5

-P

d

1-
a.

0.8-

CYT-5-P

0.6-

—

0.4^

^

0.2-

^

=1-

X

02

04 06 08

LITERS THROUGH COLUMN

10

I 2

Fig. 25. Separation of deoxycytidine-5'-phosphate and cytidine-5'-phosphate by
anion exchange in a formate system with borate present."
Exchanger: same as Figs. 22, 23, and 24, in formate form.

Solutions: 0.1 M Na formate plus .0005 M tetraborate (peak 1) ; 0.1 M Na formate
(peak 2).
Sorbed material: ca. 5 mg. each in 10 ml. 0.01 M NH4OH.

was heated in acid for a period which produced a large degree of hydrolysis
to free ribose (which was identified as such by the method of Khym and
ZilP^). This new isomer, which was separable in a sulfate (also in a chloride)
system without borate*^ from the 2 and 3 isomers and partially separated
from the 5, has been identified as the 4 isomer. In Fig. 24^^ are showTi the
appearance of ribose-4-phosphate upon acid treatment and its separation

«» J. X. Khym and W. E. Cohn, /. Am. Chem. Soc. 75, 1153 (1953).

6' J. X. Khym, D. G. Doherty, E. Volkin, and W. E. Cohn, /. Am. Chem. Soc. 75,

1262 (1953).
« J. X. Khym and W. E. Cohn, J. Am Chem. Soc. 76, 1818 (1954).
«3 J. X. Khym, D. G. Doherty, and W. E. Cohn, for J. Am.. Chem. Soc, 77, in

press.
" J. X. Khym and W. E. Cohn, Biochim. el Biophys. Acta 15, 139 (1954).

ION-EXCHANGE CHROMATOGRAPHY 241

from ribose-2- and 3-phosphates in a sulfate-plus-borate system (the borate
influences the position of the 3-phosphate only, as neither the 4- nor the
2- derivatives form complexes appreciably with borate). ^^

Ribose-1 -phosphate, in sulfate-borate systems, precedes only very
slightly ribose-3-phosphate.^' In the absence of borate, it would be expected
to precede ribose-4-phosphate. Its acid-labihty precludes its presence in
the acid hydrolysates used to produce the other four isomers, but we may
list the order of elution, in the sulfate-borate systems, as 4, 2, 1, 3, 5.^^

c. Nucleotides

An extension of this method to the separation of cytidine-5'-phosphate
and deoxycytidine-5'-phosphate yields the expected result: the latter is not
complexed and thus precedes the ribose derivative (Fig. 25). This should
be applicable to all such pairs.

V. Related Reviews

The ion-exchange separation of the products of alkaline, diesterase and
nuclease digests of PNA and DNA, together with the methods of preparing
these digests, is described in a parallel review^* that includes the spectro-
photometric constants which are so useful in following the course of sepa-
ration. Detailed directions for the ion-exchange separation and concentra-
tion of the PNA and DNA mononucleotides are given elsewhere. ^^

^* W. E. Cohn, in Colowick and Kaplan, Methods in Enzymology, Vol. II, Academic

Press, New York, in press.
"W. E. Cohn and J. X. Khym, also W. E. Cohn, E. Volkin and J. X. Khym, in

W. W. Westerfeld, Biochemical Preparations, Vol. IV, John Wiley & Sons, New

York, in press.

CHAPTER 7

Separation of Nucleic Acid Components by
Chromatography on Filter Paper*

G. R. WYATT

Page

243
I. Introduction

II General Technique of Paper Chromatography ^**

III. Detection of Purine and Pyrimidine Derivatives on Filter Paper .... 246

1. Purine and Pyrimidine Bases

2. Nucleosides and Nucleotides

IV. Solvent Systems

1 General and Theoretical Considerations ^

2. Separation of Purine and Pyrimidine Bases and Nucleosides 250

3. Separation of Nucleotides ;\t ' , • ' » -j oi=;7

V. Quantitative Estimation of the Nitrogenous Components of Nucleic Acids 257

1. Hydrolysis of Deoxypentose Nucleic Acids 257

2. Hydrolysis of Pentose Nucleic Acids 25J

3. Quantitative Technique

VI. Chromatography of Nucleic Acid Sugars

VII. Addendum

I. Introduction

The feasibility of separating nucleic acid components by chromatography
on filter paper was first demonstrated in 1947-1948 by Vischer and Char-
gaffi and by Hotchkiss.^ Since then progress has been rapid and paper
chromatography as a quantitative technique has now attained more suc-
cess with this group of compounds than with any other. This is due largely
to the intense absorption of ultraviolet light by purine and pynmidme
derivatives, which facilitates their detection on paper and makes possible
their direct estimation, once separated, by spectrophotometry. The nitrog-
enous bases may be accurately estimated by ordinary methods of paper
chromatography from less than 0.5 mg. of nucleic acid, or, with special re-
finements of technique, from as little as a few micrograms. Application of
these methods has led to knowledge of the quantitative composition of
nucleic acids from a variety of sources, to recognition of two pyrimidine
bases not previously known to occur in nucleic acids, and, together with
* Contribution No. 116, Forest Biology Division, Science Service, Department of
Agriculture, Ottawa, Canada.

1 E. Vischer and E. Chargaff, /. Biol. Chem. 168, 781 (1947).

2 R. D. Hotchkiss, J. Biol. Chem. 175, 315 (1948).

243

244 G. R. WYATT

chromatography on ion-exchange columns and electrophoresis on filter
paper, is playing an important part in investigations into the specificity of
nucleases and the molecular structure of nucleic acids.

The present article aims to provide a practical guide to paper chromato-
graphic methods for quantitative analysis of nucleic acids and for separa-
tion and estimation of their components and related substances.

II. General Technique of Paper Chromatography

The general procedures used in chromatography on filter paper are by
now well known and are covered by several reviews and books.*-^* Both
the original descending technique (in which the upper edge of the paper
dips into a trough containing the solvent^), and the ascending technique
(in which the paper, rolled into a cylinder, stands in a dish of the solvent*)
have been widely used, and give similar results. The ascending method is
convenient for small two-dimensional chromatograms but becomes im-
practical when the solvent is required to flow more than about 25-30 cm. ;
the descending method has the advantages that the solvent may be allowed
to flow an indefinite distance and that the solvent in the trough may be re-
newed while retaining the vapor in the tank.

The troughs required for descending chromatography may be made of plastic or
stainless steel, but for chemical reasons glass is to be preferred, and it is perhaps
worth describing an especially simple method of making glass troughs.' Using a cut-
ting diamond mounted at the end of a steel rod,i" two longitudinal scratches are made
inside a length of glass tubing about 2.5 cm. in diameter, separated from one another
by about 90° around the circumference of the tube. The tube is then gently tapped on
the outside with a piece of metal following the lines of the scratches, and will crack
along them producing a trough whose ends can be sealed in a flame.

The grade of filter paper most frequently used has been Whatman No. 1, which is
satisfactory for quantitative and qualitative work with nucleic acid derivatives.
Whatman No. 4 is a faster running paper; in some solvent systems, however, this
may result in poorer resolution. Schleicher and Schvill.No. 597 paper has also been
used with similar results, although different grades of paper may give different mobil-
ities with the same solvent system. When larger quantities of material are to be sepa-

3 A. J. P. Martin, Ann. Rev. Biochem. 19, 517 (1950).

* R. J. Block, R. Le Strange, and G. Zweig, "Paper Chromatography." Academic
Press, New York, 1952

^ J. N. Balston and B. E. Talbot, "A Guide to Filter Paper and Cellulose Powder
Chromatography." Reeve Angel and Co., London, and W. and R. Balston Ltd.,
Maidstone, 1952.

* F. Cramer, "Papierchromatographie," 2nd ed. Verlag Chemie, Weinheim, 1953.
^^ E. and M. Lederer, "Chromatography." Elsevier Publishing Co., Amsterdam,

1953.
^ R. Consden, A. H. Gordon, and A. J. P. Martin, Biochem. J. 38, 224 (1944).
« R. J. Williams and H. Kirby, Science 107, 481 (1948).
' This process was first demonstrated to the writer by Dr. R. Markham.
'" Glass-cutting tool no. 6686, obtainable from A. Gallenkamp and Co. , Ltd. , London.

SEPARATION BY PAPER CHROMATOGRAPHY 245

rated, the thick papers Whatman No. 3 (grained surface) and No. 3MM (smooth
surface) are useful. No. 3 is approximately 2.2 times as thick as No. 1, and more than
10 mg. of a substance can be separated from a mixture applied as a band across a
single sheet (183^" x 22''^"). S-Hydroxymethjlcytosine was isolated from bacterio-
phage DNA on this type of chromatogram in amounts adequate for crj'stallization
and analysis."

Filter papers contain a certain amount of ultraviolet-absorbing material which
may be elated by a chromatographic solvent, especially if the latter is acid, and col-
lect in a band at or behind the solvent front. This can be removed by prior washing of
the paper, which is desirable in preparative work. In quantitative work, it may gen-
erally be allowed for by taking appropriate blanks. For successful chromatography
of phosphoric esters in many solvents, however, thorough washing of the paper is
essential, as otherwise the presence of metallic ions maj^ cause streaking or double
spots.'-' '^

With a given solvent system, precise Rf values^^ are influenced by (a) the
composition of the vapor phase in the chromatography vessel, which,
ideally, should be in equilibrium with the solvent mixture before a nni is
started, (b) the temperature, which affects partition coefficients, (c) the
direction (ascending or descending) and length of run, since the composition
of the solvent may change during its passage through the paper, ^^ and (d)
the paper, of which some variation is found even between batches of one
grade. If these conditions are adequately controlled Rp values can be ac-
curately reproduced. ^^ However, as has frequently been observed, a charac-
teristic pattern of the spots can be maintained despite considerable varia-
tion in absolute values of Rp , so that precise control is usually unnecessary.
The sensitivity to environmental conditions depends on the particular sol-
vent system.

III. Detection of Purine and Pyrimidine Derivatives on Filter Paper

1. Purine and Pyrimidine Bases

In the earlier experiments, a variety of means were tried for determining
the positions of nucleic acid derivatives on filter paper. Hotchkiss- cut the
paper into narrow bands, each of which was eluted for measurement of its
ultraviolet extinction in the spectrophotometer. Vischer and Chargaff ^ ■ '^
treated the paper with salts of mercury, and, after the excess had been
washed out, the mercury fixed by the purine and pyrimidine bases was
made visible by conversion to black mercuric sulfide. Another chemical

" G. R. Wyatt and S. S. Cohen, Biochem. J., 55, 774 (1953).
'2 C. S. Hanes and F. A. Isherwood, Nature 164, 1107 (1949).
13 K. C. Smith and F. W. Allen, Federation Proc. 12, 269 (1953).

movement of band

" Defined' as Rp = 1 — 7 — ^ '■ ^ : — ^ ,. . ,.

movement oi advancing front or liquid

16 L. Horner, W. Emrich, and A. Kirshner, Z. Elektrochem. 56, 987 (1952).

18 E. C. Bate-Smith, Biochem. Soc. Symposia {Cambridge, Engl.) No. 3, 62 (1949).

" E. Vischer and E. Chargaff, J. Biol. Chem. 176, 703 (1948).

246 G. R. WYATT

method of detecting purines'^ depends on their conversion to silver salts.
Purines may also be made visible by staining their mercury complexes with
eosin or bromphenol blue, or by their fluorescence after exposure to chlor-
ine.^^ A microbiological method for detecting purines and pyrimidine nucleo-
sides on paper chromatograms with the aid of deficient strains of Ophios-
toma has been used by Fries et al}'' In general, however, the most convenient
techniques are those taking advantage of the absorption of ultraviolet light
by the nucleic acid bases.

When a paper chromatogram is examined under an ultraviolet lamp hav-
ing a high emission in the range of maximal nucleic acid absorption, and
with visible light efficiently filtered out, spots of nucleic acid components
appear as dark regions against the background fluorescence of the filter
paper.^^"^^ A low-pressure mercury resonance lamp is suitable, and it is re-
ported that 0.2 ng. of adenine spread over a circle 1.5 cm. in diameter can
be detected.-'*

Another procedure of approximately equal sensitivity, which is more laborious
but provides a permanent record of each chromatogram, consists in making photo-
graphic contact prints in ultraviolet light. ^^^ ^^ A medium- or high-pressure mercury
lamp is used, with a filter system which isolates the 253.7-m/i and 265-myii emission
lines. 2' The dried paper chromatogram is pinned over a sheet of photographic paper
(a contact document or photostat paper is suitable) on a board and exposed to the
lamp for an appropriate time (usually less than a minute). In the developed print,

18 R. M. Reguera and I. Asimov, /. Am. Chem. Soc. 72, 5781 (1950).

19 H. Michl, Naturwissenschaflen 40, 390 (1953).

2" N. Fries and U. Bjorkman, Physiol. Plantarum 2, 212 (1949) ; N. Fries and B. Fors-
man, ibid. 4, 410 (1951).

21 E. R. Holiday and E. A. Johnson, Nature, 163, 216 (1949).

21a E. Chargaff, B. Magasanik, R. Doniger, and E. Vischer, /. Am. Chem. Soc. 71,
1513 (1949).

" C. E. Carter, J. Am. Chem. Soc. 72, 1466 (1950).

23 T. Wieland and L. Bauer, Angew. Chem. 63, 511 (1951).

2^ Marshak^'i reports this sensitivity using a General Electric Co. lamp No. G8T5
equipped with Corning filter No. 9863. "Mineralight" lamps have been widely
used, but are rather less sensitive. The effect has been photographed by J. P. Goel-
ler and S. Sherry, Proc. Soc. Exptl. Biol. Med. 74, 381 (1950).

26 R. Markham and J. D. Smith, Nature 163, 250 (1949).

26 R. Markham and J. D. Smith, Biochem. J. 45, 294 (1949).

2' Markham and Smith^" recommend the Mazda MB/V lamp with the glass bulb
removed, and a filter system made with two 25-ml. fused silica round-bottomed
flasks containing, respectively, a solution of cobalt and nickel sulfates (CoS04-
7H2O 10 g. and NiS04 35 g. per 100 ml.) and dry chlorine gas. The writer has found
satisfactory a General Electric AH-4 lamp equipped with this filter system. The
chlorine gas filter may be replaced by a 1-cm. layer of chlorine dissolved in carbon
tetrachloride. 39 The system transmits too much visible light to be satisfactory for
making spots visible by fluorescence quenching, but by adding a Corning filter
No. 9863 or by viewing the chromatograms through a blue filter, one may use it in
this way too.

SEPARATION BY PAPER CHROMATOGRAPHY 247

the positions of substances absorbing ultraviolet light of the wavelength used appear
as white areas on a dark background. These spots may then be traced on the chromat-
ogram itself. Slightly greater sensitivity is claimed for a technique identical in
principle but utilizing the 257-m/x and 275-miu emission lines from a cadmium arc.^^

It has been pointed out by Smith and Markham^^ that guanine and com-
pounds containing it fluoresce quite strongly in Hght of wavelengths 253.7
and 265 mju, and are thus easily differentiated from other nucleic acid de-
rivatives. Acid conditions are required, and may be created by exposing
the chromatogram to fumes of hydrochloric acid. Xanthine behaves simi-
larly .^^^ The effect may be recorded photographically by inserting between
the chromatogram and the photographic paper a sheet of cellulose nitrate,
which transmits only the fluorescent light. 8-Azaguanine and its compounds
fluoresce under both acid and basic conditions,-^'' and may be detected on
chromatograms by their fluorescence with greater sensitivity than by their
absorption of ultraviolet light.^"

With the device described by Paladini and Leloir,'' in conjunction with the Beck-
man spectrophotometer, a continuous record may be obtained of the ultraviolet ab-
sorption of strip chromatograms.

2. Nucleosides and Nucleotides

The techniques using ultraviolet light described above are of course also
applicable to nucleosides and nucleotides. Some color reactions of the sugar
and phosphate portions of these compounds are also useful on occasion, as
in identification of unknowns or where it is necessary to use a chroma-
tographic solvent which itself absorbs in the ultraviolet range.

Buchanan, Dekker, and Long'^ have developed means of detecting both ribo- and
deoxyribonucleosides on chromatograms by reactions of the sugars. The czs-glycol
structure present in ribosides may be oxidized either with periodate, the resulting
aldehydes being made visible with Schiff 's reagent, or with lead tetraacetate, in which
case white spots remain when the uncombined lead on the paper is converted to lead
dioxide. Other substances having this configuration, including adenosine-5'-phos-
phate, react, but riboside-2'- and -3'-phosphates and deoxyribosides do not. The
sensitivity of both methods is reported as about 20 /ug. of nucleoside. Deoxyribosides
on paper can be detected by adaptations of the Dische diphenylamine reaction or of
the Feulgen reaction, or of the reaction with cysteine.'' The last method is sensitive
to 10 iig. of deoxyriboside. [Cf. Chapters 9 and 17.]

The positions of deoxyribosides on paper chromatograms have also been deter-
mined by virtue of their ability to promote growth of Labctobacillus leichmannii.^*

28 J.-E. Edstrom, Nature 168, 876 (1951).

" J. D. Smith and R. Markham, Biochem. J. 46, 509 (1950).

"» J. Kream and E. Chargaff, J. A771. Chem. Soc. 74, 4274 (1952).

'« R. E. F. Matthews, Nature 171, 1065 (1953).

»i A. C. Paladini and L. F. Leloir, Anal. Chem. 24, 1024 (1952).

32 J. G. Buchanan, C. A. Dekker, and A. G. Long, J. Chem. Soc. 1950, 3162.

" J. G. Buchanan, Nature 168, 1091 (1951).

3* V. Kocher, R. Karrer, and H. R. Muller, Intern. Z. Vitaminforsch. 21, 403 (1950).

248 G. R. WYATT

Nucleotides and other phosphoric acid esters on paper chromatograms can be de-
tected by sprajang with an acid molybdate solution, partial hydrolysis of the ester,
and reduction of the resulting phosphomolybdate complex to a blue-colored com-
pound.^* The necessary hydrolysis may be effected by heating the papers after spray-
ing, by ultraviolet irradiation,'^ or by previous spraying with a solution of phospha-
tase."- '* It is claimed that 0.1 ng. P can be detected. If the water in the reagent is
partially replaced by acetone^' the papers may be dipped in it instead of sprayed.
Nucleotides can also be detected by fixation of uranium,^" and phosphates, by fixa-
tion of ferric iron;" these reactions avoid the need for hydrolysis.

IV. Solvent Systems
1. General and Theoretical Considerations

The solvent systems with which successful separations were first ob-
tained on paper chromatograms consisted of organic fluids saturated with
water. Their effect was satisfactorily interpreted as resulting from the par-
tition of solutes between a water-poor mobile phase and a water-rich phase
held by the strongly hydrophilic cellulose fibers ; and for a number of amino
acids and carboxylic acids, partition coefficients calculated from Rp values
on the basis of this theory agree well with the coefficients directly meas-
ured.''• ^2 A minor role may be played by adsorption and ion exchange, since
the cellulose fibers are electronegative in water, and carry a small number
of aldehyde and carboxyl groups.^ It was subsequently found that the sol-
vent need not be saturated with water, since the binding of water by the
cellulose results in a partition effect with miscible solvents just as with
water-saturated ones.^' In a further innovation, it was discovered that
separations may be obtained in the absence of any organic solvent, using
salt solutions, or even with water alone. Separations with water as the sol-
vent, apparently due to adsorption by the paper, may also be interpreted
as the result of partition between water and a water-cellulose complex.
[See below, Section IV.2.]

The influence of the composition of the solvent system on the movement

" Hanes and Isherwood'" spray the chromatograms at a rate of 1 ml. per 100 cm.^
with a solution containing: 5 ml. 60% HCIO4, 10 ml. N HCl, 25 ml. 4% (NH4)2Mo04 ,
and water to 100 ml.; then heat them to 85° for 7 min., and subsequently expose
them to H2S. Benson et al.*'' describe a similar reagent.

3« R. S. Bandurski and B. Axelrod, /. Biol. Chem. 193, 405 (1951).

" N. G. Doman and Z. S. Kagan, Biokhimiya 17, 719 (1952), seen only in abstract,
Chem. Abstracts 47, 4795 (1953).

38 E. Fletcher and F. H. Malpress, Nature 171, 838 (1953).

39 S. Burrows, F. S. M. Grylls, and J. S. Harrison, Nature 170, 800 (1952).

*° B. Magasanik, E. Vischer, R. Doniger, D. Elson, and E. Chargaff, /. Biol. Chem.

186,37 (1950).
" H. E. Wade and D. M. Morgan, Nature 171, 529 (1953)
<2 A. A. Benson, J. A. Bassham, M. Calvin, T. C. Goodale, V. A. Haas, and W.

Stepka, J. Am. Chem. Soc. 72, 1710 (1950).
« H. R. Bentley and J. K. Whitehead, Biochem. J. 46, 341 (1950).

SEPARATION BY PAPER CHROMATOGRAPHY 249

of substances has been discussed in terms of partition theory by Martin.^- ^^
The following factors are among the most important which may be utilized
in preparing solvents to effect desired separations and in using paper
chromatography in the identification of unknown substances.

a. Water content

By using a miscible organic solvent, the water content may be varied
over a wide range. As water is added to the moving phase, the rates of migra-
tion of solutes will increase in proportion to their polarity. For example, by
altering the water content of propanol-water mixtures, the relative posi-
tions of adenine and adenylic acid-^ or 5-methylcytosine and 5-hydroxy-
methylcytosine^^ may be reversed.

h.pH

Since ionization will alter the partition of a solute in favor of the aqueous
phase, one can regulate the relative rates of movement of ionizable sub-
stances by control of pH, having regard to their dissociation constants.
[Cf. Cohn, Chapter 6, and Jordan, Chapter 13.] Thus, in neutral aqueous
n-butanol uracil migrates more rapidly than cytosine, but if the solvent is
made basic by the presence of sufficient ammonia the movement of uracil
and thymine may be slowed until the former has an Rp less than that of
cytosine. This may be explained by ionization of enolic hydroxyls (pKi =
9.5 and 9.9 for uracil and thymine, respectively*^), the hydroxyl of cyto-
sine having too high a pK (12.2) to be more than slightly affected by am-
monia. The relative mobilities are of course sensitive to the precise con-
centration of ammonia (cf. Table I, solvents a, h, and c). Addition of strong
acid or base to a chromatographic solvent may have further effects on par-
tition coefficients; for example, hydrochloric acid added to aqueous alco-
hols decreases the Rp's of all the purine and pyrimidine bases, but the
effect is differential and at about 1 A^ HCl the relative positions of adenine
and cytosine are interchanged.'*^

c. Nature of the Organic Components

The importance of van der Waals' forces and of hydrogen bonding in
determining the partition of a system has been stressed by jNIartin.*^ When
solvent and solute are similar in structure, for example both aromatic or
both aliphatic, closer fit between their molecules, resulting in greater van
der Waals' forces and higher Rp values, may be expected. This principle is
of limited utility in separating nucleic acid bases, where we are concerned

** A. J. P. Martin, Biochem. Soc. Symposia iCa77ibndgc, Engl), No. 3, 4 (1949).

« G. R. Wyatt and S. S. Cohen, Nature 170, 1072 (1952).

« D. Shugar and J. J. Fox, Biochim. et Bwphys. Acta 9, 199 (1952).

" G. R. Wyatt, Biochem. J. 48, 584 (1951).

250 G. R. WYATT

exclusively with two similar ring systems, but does account for such facts as
the better separation of 5-methylcytosine from cytosine with increasing
chain length of the alcohol used.** The nucleic acid derivatives have num-
erous possibilities for hydrogen bonding, which undoubtedly contributes
to the different order of movement of the nucleotides given by solvents
active in forming hydrogen bonds, such as phenol and butyric acid, than
by aliphatic alcohols.

d. Salt Content

Salt will generally decrease the mutual solubility of water and an organic
solvent, and will thus alter the partition of a chromatographic solvent sys-
tem, and for this reason the local presence of salt may cause distorted spots.
Addition of salt to a system selectively slows the movement of solutes, and
relatively strong salt solutions containing little or no organic solvent may
be used to obtain chromatographic separations on filter paper. Hagdahl and
Tiselius*^ have separated amino acids using 3 M phosphate buffer as the
solvent; they term this "salting-out chromatography" and attribute the
effect to reductions in solubility due to the presence of salt, with consequent
increase in apparent adsorption. With similar "salting-out solvents" (5%
ammonium citrate or sodium or potassium phosphate overlayered with
isoamyl alcohol, which is slightly soluble in water and has the effect of pro-
ducing more compact spots), Carter^^ has obtained separations of nucleic
acid components. The principle has been further applied by Markham and
Smith, ^"^ who used 0.8 saturated ammonium sulfate containing 2 % isopropa-
nol. It is notable that these systems separate substances in a different order
from the more usual organic systems, and that they are capable of resolving
nucleoside-2'- and -3 '-phosphates, a separation not accomplished in any
system of the more usual type.

Two practical considerations to be borne in mind when preparing solvent
systems are: (a) it is preferable to select volatile substances (ammonium
sulfate might with advantage be replaced by ammonium carbonate) and
(b) when working with nucleic acid derivatives it is of course particularly
desirable to avoid substances absorbing in the ultraviolet range.

2. Separation of Purine and Pyrimidine Bases and Nucleosides

At the time of writing, over eighty solvent systems have been described
for separation of nucleic acid components on paper chromatograms. For-
tunately for the reviewer and for those using this technique, relatively few
of these show real advantages over others, and some of the simplest mix-

« G. R. Wyatt, Biochem. J. 48, 581 (1951).

" L. Hagdahl and A. Tiselius, Nature 170, 799 (1952).

6» R. Markham and J. D. Smith, Biochem. J. 49, 401 (1951).

SEPARATION BY PAPER CHROMATOGRAPHY 251

tures remain the most valuable. In Table I are listed the reported Rf values
of some purine and pyrimidine bases and nucleosides in a number of solvent
systems selected to include (a) those which, in the writer's estimation, have
proved the most broadly useful, and (b) representatives of different types
of mixture which have been tried with some degree of success. It is ob-
vious that for some purposes other solvents will be preferable to those
listed. For completeness, the Rp values reported by the original authors
have been supplemented with some determined by the writer.

Aqueous n-butanol, with and without added ammonia, was one of the
first solvents tested for separation of nucleic acid derivatives^ • ^^ and remains
among the most useful (Table I, solvents a, h, and c). Butanol saturated
with water has usually been used; however, the solution is most conveni-
ently made up at a fixed percentage composition slightly under-saturated,
thus making its composition independent of tempera ture.^i Since Rp values
are rather low, it is advantageous to use the descending method and to cut
the end of the paper to a number of teeth, from which the solvent is allowed
to drip.2^ Ammonia may be added either to the solvent irrigating the paper
or to that in the bottom of the tank; owing to the volatility of ammonia,
the latter practice affords the more constant conditions. It has the effect of
slowing the movement of substances with acidic substituents, and the
results given by two different concentrations are illustrated in Table I
(solvents h and c). With butanol-water-ammonia, all of the purine and pyrim-
idine bases known to occur in nucleic acids (except 5-hydroxymethylcy-
tosine, which runs close to guanine) may be resolved from one another. In
addition to the authors cited in the Table, Chargaff, el al.^'^- ^^^ Marshak
and Vogel,*' and others have used butanol-ammonia mixtures for quanti-
tative separation of nucleic acid components.

Addition of formic acid to aqueous butanol (solvent d) results in more rapid move-
ment of acidic substances such as uracil, thymine, xanthine, and hypoxanthine.

If butanol is saturated with a saturated solution of boric acid instead of with
water, ribosides, by virtue of their cj's-diol configuration, form borate complexes and
do not move. Complete separation of free bases from ribosides may thus be obtained.*^

Admixture of various other substances with butanol-water systems has been tried,
generally without much advantage. Butanol may be saturated with a 10% solution
of urea, instead of water, with similar results. ^^ Mixing morpholine or diethylene

^' n-Butanol satd. with water at 20° contains 84% by vol. of butanol, calcd. from

data given by A. Seidell, "Solubilities of Organic Compounds," Vol. 2, p. 266.

Van Nostrand Co., New York, 1941.
" E. Chargaff, R. Lipshitz, C. Green, and M. E. Hodes, /. Biol. Chem. 192, 223

(1951).
"» C. Tamm, H. S. Shapiro, R. Lipshitz, and E. Chargaff, J. Biol. Chcyn. 203, 673

(1953).
" A. Marshak and H. J. Vogel, J. Biol. Chem. 189, 597 (1951).
" I. A. Rose and B. S. Schweigert, J. Ain. Chem. Soc. 73, 5903 (1951).

TABLE I

Rp Values op Purine and Pyrimidine Bases and Nucleosides

Solvent

0

"o

c

a

3

0.38

s
0.28

"o

£3

s
0.40

0.33

a

c3
O "

CO

1— (

0.37

c

1

1— (

0.32

1 c

C O

o

0.34

3:2

O 03

CO

1— 1

0.83

1

03 .

^ 03

0.44

Adenine

0.37

Guanine

0.15

0.11

0.15

0.13

0.16

0.22

0.22

0.70

0.02

0.40

Hypoxanthine

0.26

0.12

0.19

0.30

0.16

0.29

0.44

0.69

0.57

0.63

Xanthine

0.18

0.05

0.01

0.24

0.11

0.21

0.62

0.60

0.49

0.62

Uracil

0.31

0.19

0.33

0.39

0.38

0.66

0.74

0.67

0.73

0.76

Thymine

0.52

0.35

0.50

0.56

0.52

0.76

0.84

0.78

0.73

0.74

Cytosine

0.22

0.24

0.28

0.26

0.32

0.44

0.21

0.80

0.73

0.70

5-Methylcytosine

0.29

0.27

0.36

—

0.37

0.52

—

—

—

0.73

5-Hydroxymethylcytosine

0.13

0.12

—

0.25

0.44

—

—

—

0.75

Adenosine

0.20

0.22

0.33

0.12

0.31

0.34

—

0.91

0.54

0.49

Guanosine

0.15

0.03

0.10

0.17

0.13

0.30

—

0.59

0.62

0.68

Inosine

—

0.03

0.08

—

0.14

0.30

—

—

—

0.81

Uridine

0.17

0.08

—

0.25

0.31

0.64

—

0.60

0.79

0.84

Cytidine

0.12

0.11

0.15

0.18

0.28

0.45

—

0.73

0.76

0.76

Adenine DR*

0.35

0.41

0.91

0.55

0.47

Guanine DR

0.21

—

0.18

—

—

—

—

0.67

0.62

—

Hypoxanthine DR

0.23

—

0.17

—

—

—

—

0.70

0.70

0.80

Uracil DR

0.38

—

0.34

—

—

—

—

0.67

0.79

0.83

Thymine DR

0.51

0.40

0.48

—

0.57

0.81

—

0.75

0.78

0.77

Cytosine DR

0.23

—

0.26

—

—

0.60

—

0.83

0.77

0.75

5-Methylcytosine DR

0.25

—

—

—

—

—

—

—

0.76

—

" 86% (vol. /vol.) aq. n-butanol; Whatman No. 1 paper, descending; Markham and Smith, Biochem. J'
45, 294 (1949). Values for deoxyribosides are from Buchanan, 33 using n-butanol satd. with water; values fo
methylcytosine and hydroxymethylcytosine detd. by the reviewer.

'' 86% (vol. /vol.) aq. n-butanol, with 5% by vol. of coned. NH3 soln. (sp. gr. 0.880) added to solvent in
bottom of tank; Whatman No. 1, descending; Markham and Smith, op. cit. Values for methylcytosine, hy-
droxymethylcytosine, inosine, and thymine deoxyriboside detd. by the reviewer.

" n-Butanol satd. with water at about 23° 100 ml., 15 N NH4OH 1 ml.; Whatman No. 4, ascending; Rlac-
Nutt, Biochem. J. 50, 384 (1952). Values for xanthine, adenosine, and guanosine are from Hotchkiss, J. Biol
Chem. 175, 315 (1948), with a similar solvent system.

** n-Butanol 77%, water 13%, formic acid 10% by vol.; Whatman No. 1, descending; Markham and Smith,
op. cit.

' Isopropanol 85 ml., water 15 ml., coned. (28%) NHj soln. 1;3 ml.; Whatman No. 1, descending; Hershey
et al., J. Gen. Physiol. 36, 777 (1953). 72;? values at 20-23° detd. by the reviewer.

/ Isopropanol 170 ml., coned. HCl (sp. gr. 1.19) 41 ml., water to make 250 ml.; Whatman No. 1, descending;
Wyatt, Biochem. J. 48, 584 (1951). Values redetd. at 20-23° by the reviewer.

" Collidine 1 vol., quinoline 2 vol., mixt. satd. with 1.5 vol. water; Schleicher and Schiill No. 597 paper,
descending, at about 22°; Vischer and Chargaff, /. Biol. Chem. 176, 703 (1948).

'' Isobutyric acid 400 ml., water 208 ml., 25% NHi soln. 0.4 ml.; Whatman No. 4, descending, at 22°; Lof-
gren, Acta Chem. Scand. 6, 1030 (1952); excepting deoxyribosides, for which solvent and conditions are as in
footnote e, Table II, and Rf values arecalcd. from therelative mobilities given by Tamm et al., J. Biol. Chem.
203, 6:3 (1953), taking the Rp oi thymidine as 0.75.

' 5% aq. Na2HP04 satd. with isoamyl ale, both aq. and nonaq. phases being present in the trough;
Whatman No. 1, descending; Carter, J. Am. Chem. Soc. 72, 1466 (1950). Values for deoxyribosides are from
Buchanan, Nature 168, 1091 (1951).

' Water adjusted to pH 10 with N NHiOH; Whatman No. 1, ascending, 22-23°; Levenbook, personal
communication, 1953. Values for methylcytosine and hydroxymethylcytosine detd. by the reviewer.

*DR = deoxyriboside.

252

SEPARATION BY PAPER CHROMATOGRAPHY 253

glycol with butanol permits addition of more water and gives higher Rp values, with-
out markedly altering the order of separation of the bases.'' Addition of ethanol gives
higher Rf values with some loss of resolution, ^^ Dioxane-* and 2-methoxyethanol
(methyl cellosolve)^^ have been added to butanol-water mixtures, and also lead to
increased Rf values. Water-saturated n-butanol has been mixed with acetic acid,
with ethyl acetate and morpholine, and with methyl glycol and morpholine with
some success in separating deoxj^ribosides.**

Amyl alcohol saturated with water gives inconveniently low Rf values for most
substances. 2^ Isopropanol-water-ammonia (Table I, solvent e) has been found a use-
ful mixture by Hershey, et al. :" it resolves 5-hydroxj'methylcytosine from the other
bases, but adenine and uracil run together, and the bases are less well separated from
their ribosides than in butanol. Mixtures of tetrahydrofurfuryl alcohol with propanol
and amyl alcohol buffered at different pH's have also been tried, ^* with some success
in separation of nucleotides, but for separation of the bases and nucleosides they
appear to be inferior to simple butanol solvents. Several other mixtures of alcohols
with NH3 and HCl have also been tested on a limited range of substances, ^^ and the
Rf values of orotic acid^" and of uric acid and its riboside*' in several solvents have
been recorded.

Solvent systems based on collidine and quinoline instead of alcohols have been
tested by Vischer and Chargaff.'' In these (e.g.. Table I, solvent g) the bases migrate
in a different order, xanthine and hypoxanthine running more rapidly and cytosine
much more slowly. However, the strong absorption of ultraviolet light by these sol-
vents is a serious drawback.

Isobutyric acid mi.xed with water and ammonia, as tested bj' Lofgren,*^ distrib-
utes the bases and nucleosides in a different order from other solvents, as shown in
Table I (solvent h), although Rf values are grouped undesirably close together.
(Compare also Tamm et al.^^'^) The effect is partially retained, with better spread of
Rf values, in a solvent containing u-butanol (75 ml.), isobutyric acid (37.5 ml.),
water (25 ml.), and ammonia (2.5 ml. of 25% soln.). A mixture containing piperidine,
tried by the same author, gave rather poor separations.

A limitation of all these neutral or weakly basic or acidic solvent systems
for quantitative analysis of nucleic acids is their low capacity for guanine.
Because of its insolubility, guanine in amounts of more than a few micro-
grams tends to form "tails" or double spots, or to remain partly at the ori-
gin. This difficulty may be avoided by use of solvents containing relatively
high concentrations of hydrochloric acid. Such a system was first used by
Smith and Markham^' for separation of the purine bases and pyrimidine
nucleotides obtained by mild acid hydrolysis of PNA (Table II, solvent a).
A mixture containing isopropanol and hydrochloric acid (Table I, solvent
/) was subsequently developed for separation of the bases from DNA.^^

" S. G. Laland, W. G. Overend, and M. Webb, J. Chem. Soc. 1952, 3224.

" W. S. MacNutt, Biochem. J. 50, 384 (1952).

" A. D. Hershey, J. Dixon, and M. Chase, J. Gen. Physwl. 36, 777 (1953).

" D. C. Carpenter, Anal. Chem. 24, 1203 (1952).

" B. Bheemeswar and M. Sreenivasaya, Current Sci. (India) 20, 61 (1951).

«" E. Leone and E. Scala, Boll. soc. ital. biol. sper. 26, 1223 (1950).

" E. Leone and D. Guerritore, Boll. soc. ital. biol. sper. 26, 609 (1950) .

62 N. Lofgren, Acta Chem. Scand. 6, 1030 (1952).

254

G. R. WYATT

fff in Isoproponol-HCI

Fig. 1. Diagram of the positions of nucleic acid derivatives on a two-dimensional
chromatogram run on Whatman No. 1 paper by the descending technique first in
solvent/ (Table I), then in solvent c (Table I; however, in the large chromatogram
tank used here, the effective NH3 concentration is reduced). A, adenine; AA, yeast
adenylic acid; AR, adenosine; C, cytosine; CA, cytidylic acid; CR, cytidine; G,
guanine; GA, guanylic acid; GR, guanosine; HMC, 5-hydroxymethylcytosine; HX,
hypoxanthine; HXR, inosine; MC, 5-methylcytosine; T, thymine; TDR, thymidine;
U, uracil; UA, uridylic acid; UR, uridine; X, xanthine.

This solvent resolves up to 75 ng. per spot of each of the DNA bases in 35
cm. movement of the front. Uracil is also resolved; 5-hydroxymethylcyto-
sine, however, runs together with cytosine. Ribosides run at similar rates
to their bases, and deoxyribosides rather faster; purine deoxyribosides are
decomposed by the acid.

When using solvents containing a high proportion of hydrochloric acid, the acid
must be thoroughly removed from the paper at the conclusion of the run by evapora-
tion at not too high a temperature (to avoid charring). Residual acid may damage
photographic paper used for printing the chromatograms and, according to Schramm
and Kerekjarto,^' on exposure to ultraviolet light may liberate chlorine which de-
stroys cytosine by oxidation.

For mixtures too complex to be resolved in one dimension, a useful two-
dimensional system is the isopropanol-HCl solvent in combination with
n-butanol-NHj or with isopropanol-NHs. Better movement of guanine is

" G. Schramm and B. von Kerekjdrt6, Z. Naturforsch. 7b, 589 (1952).

SEPARATION BY PAPER CHROMATOGRAPHY 255

obtained if the acid solvent is used first ; the rate of flow of the second solvent
is then more rapid than in untreated paper. The spacing of spots with such
a system is shown in Figure 1 .

Carter's dibasic sodium phosphate system is also included in Table I
(solvent i). Pyrimidines are well separated from purines, but within each
class there is little resolution. The use of water, or 0.01 M phosphate buffer,
to bring about rapid separation of pyrimidines from purines was first de-
scribed by Zamenhof et al.^^ Slightly greater dispersion oi Rp values is
given by water adjusted to pH 10, according to Levenbook,^^ whose data
on this system are included in Table I (solvent j). The chief virtue of water
as a solvent is its rapid flow, so that purine bases can be completely sepa-
rated from pyrimidines in 3 hours.

3. Separation of Nucleotides

Because of the strong polarity of their phosphoryl groups, nucleotides
do not move at appreciable rates in relatively nonpolar solvent systems
such as water-saturated n-butanol. Their movement may be accelerated by
increasing the content of water or other polar components in the system,
or by suppressing phosphoryl dissociation by addition of acid. These con-
ditions prevent full advantage being taken of the ionic differences in the
constituent purine and pyrimidine bases, and separation of the nucleotides
by paper chromatography has proven somewhat difficult. Effort has been
expended on the problem in a number of laboratories, and some of the more
satisfactory solvents which have been developed are shown in Table II.
None, however, is entirely satisfactory for separation of the nucleotides of
the four bases from PNA in a single run, although these may be resolved by
two-dimensional chromatography or by electrophoresis on filter paper.
[Cf. Smith, Chapter 8.]

The two solvent systems containing hydrochloric acid (Table II, solvents a and
b) both separate the pyrimidine ribonucleotides and the purine bases excellently from
one another. In solvent a cytosine and thymine deoxyriboside diphosphates are also
resolved from the pyrimidine ribonucleotides.*^ The purine nucleotides are not satis-
factorily resolved in either solvent. The isopropanol mixture is the faster running of
the two. A mixture of n-butanol, ethanol, and 5 N HCl (3:2:2 by vol.) resolves pur-
ines and pyrimidine nucleosides and nucleotides, with the following Rp values;" gua-
nine, 0.24; adenine, 0.35; cytidine, 0.43; cytidylic acid, 0.54; uridine, 0.63; uridylic
acid, 0.78.

Solvent c is one of several mixtures of acetone with carboxylic acids tested by
Burrows et al.^^ It affords good separation of guanylic, cytidylic, and uridylic acids;

^* S. Zamenhof, G. Brawerman, and E. Chargaff, Biochim. et Biophys. Acta 9, 402

(1952).
*^ L. Levenbook, personal communication, 1953.

66 L. L. Weed and D. W. Wilson, J. Biol. Chem. 202, 745 (1953).

67 J.-E. Edstrom, Biochim. et Biophys. Acta 9, 528 (1952).

256

G. R. WYATT

the Rf value of yeast adenylic acid is not recorded. Adenosine-5'-phosphate, adeno-
sine diphosphate, and adenosine triphosphate are also separable from one another in
this system, or, with rather better spacing, in a mixture of 65 vol. of acetone with 35
vol. of 15% trichloroacetic acid.

From the published Rp values, the buffered isoamyl alcohol-tetrahydrofurfuryl

TABLE II

Rf Values of Nucleotides

Solvent

c

ai

mo

0.50

'o

c

d

St.

Acetone-tri-
chloroacetic
acid''

Isoamyl alc.-te-
trahydrofur-
furyi ale. -buffer''

'E

TO

■g

u
, o

^ o

Ph

s

o
'T
6

Ph a

i

^ 03

O a

M o

>^ CO

Adenosine-2'-phosphate*l
Adenosine-3'-phosphate j

0.48

—

0.35

0.49

0.70

fO.74
\0.67

0.26
0.16

Guanosine-2'-phosphatel
Guanosine-3'-phosphatej

0.46

0.43

0.20

0.67

0.24

0.46

0.79

f0.50
\0.40

Uridylic acid

0.80

0.77

0.51

0.43

0.24

0.35

0.85

0.73

Cytidylic acid

0.56

0.58

0.34

0.26

0.37

0.57

0.85

0.73

Adenosine-5'-phosphate

—

0.43

0.37

0.28

0.43

0.69

Adenosine diphosphate

—

—

0.10

0.07

—

—

0.77

—

Adenosine triphosphate

—

—

0.04

0.08

—

—

0.83

—

Deoxycytidylic acid

—

0.64

—

—

—

—

—

—

Thymidylic acid

—

0.81

—

—

—

—

—

—

Orthophosphate

0.90

0.84

0.61

—

—

0.22

—

—

" tert-B\ita.no\ at 26° 700 ml., const.-boiling HCl 132 ml., water to make 1 liter; Whatman No. 1, descend-
ing; Smith and Markham, Biochem. J. '46, 509 (1950). Rf values (20-22°) are from Boulanger and Montreuil,
Bull. soc. chim. biol. 33, 784, 791 (1951).

'' See footnote /, Table I. The value for orthophosphate is from Markham and Smith, Biochem. J. 49,
401 (1951).

' Acetone 75 vol., 25% (wt./vol.) trichloroacetic acid 25 vol.; Whatman No. 1 paper washed in 2 A'' HCl
and water, ascending, at 4°; Burrows et al., Nature 170, 800 (1952).

"* Isoamyl ale. 1 vol., tetrahydrofurfuryl ale. 1 vol., 0.08 M potassium citrate buffer (pH 3.02) 1 vol.; What-
man No. 1, descending, 20-25°; Carpenter, Anal. Chem. 24, 1203 (1952).

* Isobutyric acid 10 vol., 0.5 A^ NH4OH 6 vol., pH 3.6-3.7; Schleicher and Schiill No. 597, descending,
21-25°; Magasanik et al., J. Biol. Chem. 186, 37 (1950). Rp values calcd. from the published relative mobilities
and figures.

^ 90% aq. phenol 84 vol., <er<-butanol 6 vol., formic acid 10 vol., water 100 vol.: after sepn. at a temp.
2-3° below that of the chromatography room, the nonaq. phase is used and the aq. phase is placed in the bot-
tom of the tank; Whatman No. 1 paper, descending, 20-22°; Boulanger and Montreuil, op. cit.

" See footnote i. Table I. Values for adenosine-5'-phosphate, diphosphate, and triphosphate are from
W.E. Cohn and C. E. Carter, J. Am. Chtm. Soc. 72, 4273 (1950).

'' Satd. (NHOiSOi in water 79 vol., 0.1 M buffer soln. (pH 6) 19 vol., isopropanol 2 vol.; Whatman No. 1
paper, descending; Markham and Smith, op. cit. Rp values are measured from the published figure.

* The "a" and "b" nucleotides of Carter, J. Am. Chem. Soc. 72, 1466 (1950), and subsequent authors are
now identified as the 2'-phosphates and 3'-phosphates, respectively, of the ribosides.

SEPARATION BY PAPER CHROMATOGRAPHY 257

alcohol system of Carpenter*' (solvent d) should give adequate separation of the
nucleotides from PNA. Use of this mixture in analytical work has not yet been re-
ported.

The isobutyric acid-ammonium isobutyrate solvent of Magasanik et al.*" (solvent
e) has been used for quantitative analysis of PNA's.** [Cf. Magasanik, Chapter 11.]
Since guanylic and uridylic acids occupy the same position on the chromatogram,
however, the amount of each must be calculated from extinction measurements taken
at two wavelengths. It is reported*^''' ^^ that the five deoxynucleotides from th.ymus
DNA are completely resolved from one another in this solvent, the respective dis-
tances travelled (relative to deoxyadenylic acid taken as 100) being: deoxyguanylic
acid, 53; thymidylic acid, 66; deoxycytidylic acid, 80; deoxyadenylic acid, 100; de-
oxy-5-methylcytidylic acid, 137.

The phenolic solvent system of Boulanger and MontreuiP" (solvent /) is capable
of separating the ribonucleotides with excellent spacing, and was used for quanti-
tative analysis of PNA's. Because of the ultraviolet absorption of the phenol, how-
ever, it is necessary to detect and estimate the nucleotides by reactions of the phos-
phoryl group, and anj' other nucleic acid derivatives would of course be missed.

For fractionation of the products of digestion of PNA by ribonuclease, Markham
and Smith''*' ''^ have used (1) 70% (bj- vol.) aqueous isopropanol and (2) 70% aqueous
isopropanol with concentrated ammonia added to the solvent in the bottom of the
tank at the rate of 0.35 ml. for each liter of gas space. The latter system proved espe-
cially useful for initial fractionation (prior to electrophoresis or chromatography in
other solvents) of the mixture of mono-, di-, and trinucleotides; in it the cyclic
nucleotides (riboside-2'3'-monohydrogen phosphates) move more rapidly than the
riboside-2'- and -3 '-phosphates. It also resolves nucleotides from their benzyl es-
ters.^'

With systems based on salt solutions instead of organic solvents (Table II, solvents
g and h) , the purine ribonucleotides, but not those of the pyrimidines, have been sepa-
rated. Carter's^^ sodium and potassium phosphate systems resolve adenosine-2'-phos-
phate and -3'-phosphate. The resolution of adenosine phosphates in dibasic sodium
phosphate is said to be improved by substituting for isoamyl alcohol 0.5% lauryl
amine in n-amjd alcohol.^'' The ammonium sulfate mixture of Markham and Smith*"
gives better spacing of the isomeric nucleotides and separates also those of guanylic
acid and 8-azaguanylic acid;'" in this solvent the cyclic nucleotides move more slowly
than the corresponding nucleoside-2'- and -3 '-phosphates. ''^

With water as the solvent, the Rp values of all the nucleotides are close to 0.9.^*

V. Quantitative Estimation of the Nitrogenous Components of

Nucleic Acids

1. Hydrolysis of Deoxypentose Nucleic Acids

The purine and pyrimidine components of deoxypentose nucleic acids

are satisfactorily estimated as free bases, since these may be obtained in

*' E. Chargaff, B. Magasanik, E. Vischer, C. Gi'een, R. Doniger, and D. Elson, J.

Biol. Chem. 186,51 (1950).
«' E. Chargaff, Federation Proc. 10, 654 (1951).

'" P. Boulanger and J. Montreuil, Bull. sac. chim. biol. 33, 784, 791 (1951).
" R. Markham and J. D. Smith, Biochem. J. 52, 552 (1952).
" R. Markham and J. D. Smith, Biochem. J. 52, 558, 565 (1952).
" D. M. Brown and A. R. Todd, /. Chem. Soc. 1953, 2040.
" O. Snellmann and B. Gelotte, Nature 168, 461 (1951).

258 G. R. WTATT

virtually quantitative yield by appropriate acid hydrolysis. Good yields of
nucleosides or nucleotides have not been obtained from DNA by any means
of hydrolysis with acid or alkali. Nucleotides may be obtained enzymically/*
but DNA containing 5-hydroxymethylcytosine resists this treatment''® and
chemical methods are preferable for routine work. [Cf. Chargaff, Chapter 10.]

a. Hydrolysis with hydrochloric acid

The purines are completely released from DNA by very mild acid treatment (pH
1 .6 at 37° for 24 hr . , or pH 2.8 at 100° for 1 hour") . More drastic hydrolysis with hydro-
chloric acid (10 mg. nucleic acid in 2 ml. 6 A' HCl at 120° for 2 hours''*- '') gives a good
yield of pyrimidines but causes some destruction of purines. Nearly quantitative
yields of all the bases (including 5-hydroxymethylcytosine but with evidence for
slight destruction of adenine) have been obtained by Hershey et al.^'' by subjecting
the partially hydrolyzed nucleic acid extracted from bacteriophage T2 with hot tri-
chloroacetic acid to further hydrolysis with 3 ml. redistilled 6 A' HCl under CO2 in
sealed tubes at 100° for 3 hours.

h. Hydrolysis with 'perchloric acid

The use of concentrated perchloric acid for liberation of bases from
nucleic acids was introduced by Marshak and Vogel.^^-^" With DNA,
either 7.5 N or 12 N (70 %) HCIO4 may be used, and maximal yields are
obtained in 1 hour at 100°. Loss of a small percentage of the thymine has
been noted during hydrolysis with 70 % HCIO4 , which may be minimized
by using not more than 15 yul. HCIO4 per mg. DNA, although if much pro-
tein is present the amount maybe increased.^' 5-Hydroxymethylcytosine is
extensively destroyed during hydrolysis of bacteriophage DNA with per-
chloric acid, although the isolated pyrimidine is apparently not attacked by
this acid."' " By hydrolyzing with HCIO4 it is possible to estimate the total
purine and pyrimidine bases from a mixture containing both DNA and
PNA, and to estimate the bases without isolation of nucleic acid from some
biological materials. Non-nucleic acid components are degraded to products
which interfere very little on the chromatograms. Perchloric acid hydroly-
sates may be diluted with water and applied directly to paper for chroma-
tography. If phosphorus is to be estimated in samples of the hydrolysate,
the insoluble residue from the nucleic acid carbohydrate, which tends to ad-
sorb phosphate, should first be brought into suspension. ^^

75 R. O. Hurst, J. A. Little, and G. C. Butler, /. Biol. Chem. 188, 705 (1951).

'« S. S. Cohen and G. R. Wyatt, unpublished results, 1953

" C. Tamm, M. E. Hodes, and E. Chargaff, /. Biol. Chem. 195, 49 (1952).

'8 M. M. Daly, V. G. Allfrey, and A. E. Mirsky, J. Gen. Physiol. 33, 497 (1950).

79 N. I. Gold and S. H. Sturgis, J. Biol. Chem. 196, 143 (1952).

«" A. Marshak and H. J. Vogel, Federation Proc. 9, 85 (1950).

" G. R. Wyatt, J. Gen. Physiol. 36, 201 (1952).

82 Levenbook** has obtained evidence for differential adsorption of purine and

pyrimidine bases also to the charcoal from nucleic acid samples contaminated with

carbohydrates.

SEPARATION BY PAPER CHROMATOGRAPHY 259

c. Hydrolysis with formic acid

The value of formic acid for liberating purine and pyrimidine bases from
nucleic acids without destruction was first shown by Vischer and Chargaff ,*^
and this acid has proved very satisfactory for the quantitative hydrolysis
of DNA. The conditions used in Chargaff 's laboratory have been 98-100 %
formic acid at 175° for 2 hours. Shorter periods of heating have also been
employed. ^^' ^^ Recently," good results were obtained with DNA from bac-
teriophages and other sources using 88 % formic acid at 175° for 30 minutes,
and it was found that whether 88 % or 98 % formic acid is used, recoveries
of guanine and hydroxymethylcytosine are substantially increased by using
a relatively large volume of it (500 jul. per mg. DNA) and by avoiding ex-
cessive oxidizing atmosphere while heating. The hydrolysis is carried out
in sealed Pyrex glass bomb tubes, and decomposition of the formic acid
produces considerable pressure, which it is advisable to release by melting
the tips of the tubes in a flame before opening them. The hydrolysate is
evaporated under reduced pressure to dryness, then redissolved in a small
volume of A^ HCl, and samples are taken for chromatography and for esti-
mation of phosphorus. Because of the high recoveries and the absence of
insoluble residue in the hydrolysate, this is probably the method of choice
for accurate microanalysis of purified preparations of DNA.

2. Hydrolysis of Pentose Nucleic Acids

The ribosides of both purines and pyrimidines are split with more dif-
ficulty than their deoxyribosides, and the resistance of the pyrimidine ribo-
sides is such that it is very difficult to get quantitative yields of the free
bases from PNA. However, the ease with which mononucleotides are ob-
tained from PNA makes it possible to estimate the pyrimidines, or all of
the bases, as nucleotides. [Cf. Loring, Chapter 5, and Magasanik, Chapter
11.]

a. Hydrolysis to purine and pyrimidine bases

Concentrated formic acid at 175° for 2 hours has been used for liberation of pyrimi-
dines from PNA;'3 however, the yield of uracil is very low and the method has been
found unsatisfactory for quantitative hydrolysis of uridylic acid.^* Nearly quantita-
tive yields of all the bases can be obtained with 70-72% perchloric acid at 100° for 1
hour/'- '3 and this method has been used for analysis of PNA's.*^

h. Hydrolysis to pnrine bases and pyrimidine nucleotides

The purine bases are liberated from PNA by relatively mild acid hydrolysis, e.g.,
N HsSOi^' or N HC186 at 100° for 1 hour. The latter treatment was selected by Smith

" E. Vischer and E. Chargaff, J. Biol. Chem. 176, 715 (1948).
«* A. Marshak, /. Biol. Chem. 189, 597 (1951).

85 Smith and Markham" include a critical discussion of previously used methods of
hydrolysis.

260 G. R. WYATT

and Markham for quantitative analysis of PNA's: pyrimidines are estimated as nu-
cleotides, a correction of 5% being applied to allow for degradation to nucleosides.^"
Abrams,*^ using isotope dilution to test comparable hydrolytic conditions, has demon-
strated about 7% destruction, probably deamination, of adenine and guanine.

c. Hydrolysis to nucleotides

As has long been known, PNA, unlike DNA, is readily hydrolyzed to
mononucleotides. [Cf. Chapter 11, and also Brown and Todd, Chapter 12.]
This may be achieved with N NaOH or A^" HCl at room temperature; the
latter is less convenient because the nucleic acid dissolves only slowly and
there is danger of splitting purine glycosidic linkages.^" Chargaff et al.^^
used pH 13 to 14 at 30° overnight to degrade PNA to nucleotides prior to
chromatographic separation. Boulanger and MontreuiP" used 0.5 N NaOH
at 20-22° for 18 hours, or concentrated ammonia solution (D° = 0.925)
at 45° for 8 days, with good results, although ammonia under other con-
ditions is liable to produce nucleosides. At 37°, N alkali causes partial de-
amination of cytidylic acid, but 0.3 N alkali does not." This was confirmed
by Davidson and Smellie,^^ who used 0.3 A^ KOH at 37° for 18 hours to
convert PNA to nucleotides, the potassium being removed as the perchlor-
ate prior to separation of nucleotides by electrophoresis on filter paper.
Crosbie et al}^ analyzed a number of samples of PNA by this method as well
as by hydrolysis in N HCl and in concentrated HCIO4 in conjunction with
paper chromatography: alkaline hydrolysis gave consistently higher results
for uracil, for cytosine, and for total recovery in terms of phosphorus than
either of the other procedures. Alkaline hydrolysis followed by^electro-
phoretic separation of nucleotides appears to be the most reliable method
yet devised for microanalysis of PNA. [Cf. Smith, Chapter 8.]

3. Quantitative Technique

Small volumes of hydrolysates or other solutions can be measured on to
filter paper for quantitative chromatography with a variety of micropipets
and burets. Especially convenient are self-filling capillary pipets,^° which
are not difficult to make; those available commercially often have exces-
sively thick tips and a bore so fine that they are liable to clogging and
drainage error. The volume which may be placed on the paper in one ap-
plication depends on the scale of working: in the range of 10 to 50 )ug. of
each purine or pyrimidine base, to be run in one dimension for about 18

8« R. Abrams, Arch. Biochem. 30, 44 (1951).

" D. H. Marrian, V. L. Spicer, M. E. Balis, and G. B. Brown, J. Biol. Chem. 189,
533 (1951).

88 J. N. Davidson and R. M. S. Smellie, Biochem. J. 52, 594 (1952).

89 G. W. Crosbie, R. M. S. Smellie, and J. N. Davidson, Biochein. J. 54, 287 (1953).
*" P. L. Kirk, "Quantitative Ultramicroanalysis," p. 22. John Wiley and Sons, New

York, 1950.

SEPARATION BY PAPER CHROMATOGRAPHY 261

hours, a suitable volume is about 10 fx\., and using less does not result in
significantly smaller final spots.

A substance may be eflficiently eluted in a small volume by cutting out
a pointed band of paper including the spot, and allowing water to diffuse
through it toward the point. The solute is carried with the water, which
may be collected from the point of the paper in a small vessel,^^ in a pipet^^
or directly on another sheet of filter paper for re -chromatography.^^ This
technique has been used to concentrate minor components such as 5-methyl-
cytosine from nucleic acids for estimation.*^

For quantitative estimation, spots are cut out and eluted by soaking in a
volume of liquid appropriate for the cells in which the extinctions are to be
read (4-5 ml. for the 1-cm. cells of the Beckman or similar spectrophotom-
eter). 0.1 iV HCl is an eluent'^ in which the nucleic acid bases are suffi-
ciently soluble, and in which some of them have higher extinction coeffi-
cients than in neutral or alkaline solutions. Standing overnight at room
temperature, with shaking before and after, effects quantitative elution.
To allow for ultraviolet-absorbing substances in the paper, blanks are cut
equal in area to the spots and at equal distances from the starting line, and
are eluted and read at the same wavelengths as the corresponding spots.
As a further precaution against error due to ultraviolet-absorbing contami-
nants, Vischer and Chargaff^^ read the extinctions of all eluates both at the
absorption peak of the substance being estimated and at another reference
wavelength where its absorption is low, basing their calculations on the
difference. The same procedure was adopted by Crosbie et al.^^ in analyzing
HCl and HCIO4 hydrolysates, and it is evident that it will generally reduce
the error from contaminants, although their absorption at the two wave-
lengths will rarely be precisely equal. Spectral data for use in estimating
nucleic acid derivatives by the absorption at their maxima and by the dif-
ference method are given in Tables III and IV, respectively. [Cf. Beaven,
Holiday and Johnson, Chapter 14.] Hotchkiss^ gives tables of the absorp-
tion of nucleic acid derivatives at 5-mM wavelength intervals, which may
be used to compute the composition of binary mixtures from density read-
ings at two or more wavelengths.

The methods which have been described are suitable for estimation of purine and
pyrimidine derivatives in the range of 5 to 100 Mg- per spot. Quantitative analyses in
duplicate have been conveniently carried out on samples of 0.3 to 1.0 mg. of nucleic
acid by making up hydrolysates to a volume of 25 m1-, from which two 8-^1. portions
are taken for chromatography and two 2-^1. portions for phosphorus estimation."' «'
In an ultra-micro adaptation of paper chromatography recently described by Ed-
strom," only one hundredth of this amount of material is required. The hydrolysate

" C. E. Dent, Biochem. J. 41, 240 (1947).

92 R. Consden, A. H. Gordon, and A. J. P. Martin, Biochem. J. 41, 590 (1947).

" A. M. Moore and J. B. Boylen, Science 118, 19 (1953).

262

G. R. WYATT

TABLE III
Ultraviolet Absorption Data on Purine and Pyrimidine Derivatives

Normal-

Millimolar

ity of

Wavelength,

extinction

pO.001%
-'^Icm.

Substance

HCl

m/x"

coefficient

Adenine''^

0.1

260

13.0

0.96

Guanine"

0.1

250

11.0

0.73

Uracil*'

0.1

260

7.9

0.705

Thymine*'

0.1

265

7.95

0.63

Cytosine'"

0.1

275

10.5

0.95

5-Methylcytosine"

0.1

283

9.8

0.785

5-Hydroxymethylcytosine'i

0.1

279

9.7

0.685

Adenylic acid'*

0.01

260

13.9

0.401

Guanylic acid '*

0.01

260

11.8

0.325

Uridylic acid**

0.01

262

9.89

0.305

Cytidylic acid"

0.01

278

12.72

0.393

" Not, in all cases, the precise absorption maximum.

TABLE IV
Ultraviolet Absorption Data on Purine and Pyrimidine Derivatives

Substance

Solvent

Absorption Reference

Maximum, wavelength,

mju m/i

Difference in

£i°« at the

two wavelengths

Adenine*'

0.1 N HCl

262.5

290

0.935

Guanine*'

1.6 A^ HCl

249

290

0.539

Cytosine*'

0.1 N HCl

275

290

0.546

Uracil*'

0.1 N HCl

259

280

0.609

Thymine''

Water

264.5

290

0.545

Cytidylic acid*'

0.01 A^ HCl

278

300

0.331

Uridylic acid*'

0.1 A^ HCl

260.5

280

0.814

of 1.5 to 5 Mg- of nucleic acid is applied in a volume of about 0.5 ^1. to a strip of filter
paper 0.5 to 1 mm. broad, which is subjected to chromatography in the ordinary way.
A contact print of the dried strip chromatogram is made on a process plate with light
of wavelengths 257 and 275 m^ from rotating cadmium electrodes, then the darkening
of the plate is measured with a recording microphotometer, the image of a rotating
sector serving for calibration. This refined technique will undoubtedly be of value in
certain biological problems; however, the equipment is not available in the average
laboratory, and one may note that by using narrow strips of paper as described by
Edstrom to prevent spreading of the spots, and eluting the bands for extinction
measurement in microcells'* on the spectrophotometer, comparable sensitivity could

«* E. Volkin and C. E. Carter, J. Am. Chem. Soc. 73, 1516 (1951).
'6 J. McT. Ploeser and H. S. Loring, J. Biol. Chevi. 178, 431 (1949).
96 O. H. Lowry and O. A. Bessey, J. Biol. Chem. 163, 633 (1946).

SEPARATION BY PAPER CHROMATOGRAPHY

263

TABLE V
Rf Values of Pentoses

Solvent

Methyl ethvl

Butanol-

Butanol-

Phenol"

ketone*

ethanol'^

pyridine''

0.54

0.075

0.12

0.44

0.090

0.15

—

—

0.125

0.19

—

0.59

0.165

0.21

0.49

0.59

0.180

0.30

0.56

0.73

—

0.44

0.60

Arabinose

Xylose

Lyxose

Ribose

Rhamnose

2-Deoxyribose

" Phenol satd. with water at 20°, with 1% NHj and a few crystals KCN in soln. in bottom of tank; What-
man No. 1 paper; Partridge, Biochem. J. 42, 238 (1948).

*■ Methyl ethyl ketone satd. with water at 20°, with 1% NHj in soln. in bottom of tank; Whatman No.
1; Partridge, op. cit. The value for lyxose is calcd. from the data of Crosbie et al., Biochem. J. 54, 287 (1953).

' n-Butanol 50 vol., ethanol 10 vol., water 40 vol., theupper layer beingused; Whatman No. 1. Rxa values
(movement relative to that of 2,3,4,6-tetramethylglucose taken as 1.00) are shown. E. L. Hirst and J. K. N
Jones, Discussions Faraday Soc. 7, 271 (1949).

** To the upper layer resulting from the mixture of 1 vol. pyridine, 1.5 vol. water, and 3 vol. n-butanol,
at about 28°, is added 1 vol. pyridine; Schleicher and Schilll No. 597 paper. Chargaff el al., J. Biol. Chem.
177,405 (1949).

be achieved. The bases from 5 ^g- of yeast PNA, each eluted in 0.1 ml. and examined
in a cell of 1 cm. path length, would give optical densities in the range 0.25 to 0.5, so
that it should be entirely practical to work on this scale without special equipment.

VI. Chromatography of Nucleic Acid Sugars

This account would be incomplete without reference to the identification
of nucleic acid carbohydrates by paper chromatography. However, since
the methods do not differ from those used for other carbohydrates, which
have been adequately reviewed,^"®- ''' and since the results with nucleic acid
components have been limited in scope, this description will be kept very
brief.

Numerous spray reagents have been described for detection of sugars on
paper chromatograms. Ammoniacal silver nitrate, used in the original in-
vestigations in this field, ^^ is sensitive but relatively nonspecific and sub-
ject to interference by impurities. w-Phenylenediamine^^ forms with a wide
range of aldoses and ketoses derivatives which fluoresce in ultraviolet light,
and has been used in Chargaff 's laboratory for recognition of nucleic acid
sugars. Aniline hydrogen phthalate^"" is a sensitive and convenient reagent

" S. M. Partridge, Biochem. Soc. Symposia {Cambridge, Engl.) No. 3, 52 (1949).
'8 S. M. Partridge, Biochem. J. 42, 238 (1948).

99 E. Chargaff, C. Levine, and C. Green, J. Biol. Chem. 175, 67 (1948).
""' Aniline, 0.93 g., and phthalic acid, 1.66 g., are dissolved in 100 ml. water-saturated

butanol. The papers are sprayed and then heated to 105° for 5 min. S. M. Partridge,

Nature 164,443 (1949).

264 G. R. WYATT

for aldoses, including ribose and deoxyribose, but not for ketoses; increased
range of sensitivity is reported for aniline hydrogen phosphate. ^"^ The re-
agents mentioned earlier (Section III. 2) for ribosides and deoxy ribosides
react also with the free sugars.

The Rf values of the pentoses, along with rhamnose and 2-deoxyribose,
in several chromatographic solvents are shown in Table V. Improved sepa-
rations of a number of sugars are reported by Jermyn and Isherwood^"^
using ethyl acetate-pyridine-water (2:1:2) and ethyl acetate-acetic acid-
water (3:1:3) as solvents ; however, Rp values of ribose and deoxyribose
were not determined. 2-Deoxyribose, 3-deoxyribose, and 4-deoxyribose are
separable with a butanol-ethanol mixture. ^"^ The d- and L-isomers of sugars
have not been separated on paper chromatograms.

The purine-bound sugar from pentose nucleic acids may be liberated by
hydrolysis in N H2SO4 at 100° for 1 hour,^^ and the hydrolysate can be
applied to paper for chromatography without removal of the acid.^"^ From
DNA, because of the lability of the sugar and the stability of the inter-
nucleotide linkages, enzymic hydrolysis is required: Chargaff et al}°^ used
deoxyribonuclease together with a phosphatase derived from Aspergillus
to give nucleosides, after which the purine nucleosides were split by heating
at pH 1.5 to 100° for 12 minutes.

The sugar components of all pentose nucleic acids yet examined by these
methods have proved to be chromatographically identical with ribose, and
those of deoxypentose nucleic acids, with 2-deoxyribose. [Cf. Chapters 2,
10, and 11.]

VII. Addendum

The following reports have appeared, or have come to the writer's atten-
tion, since preparation of this chapter.

A solvent composed of 7 vol. 95 % ethanol plus 3 vol. M ammonium ace-
tate buffer of pH 3.8, used in the study of uridine diphosphate glucose,
gives fair resolution of the nucleotides of PNA.^°^ Several systems suitable
for resolution of deoxyribosides have been described by Tamm et alJ''^"- Cald-
welP°^ has demonstrated resolution of adenosine-5 '-phosphate, diphosphate,
and triphosphate with Hanes and Isherwood's solvent containing 60 cc.

1"! N Aniline in butanol, 1 vol., mixed with 2 A'^ HjPO* in butanol, 2 vol.; J. L. Bryson

and T. J. Mitchell, Natxire 167, 864 (1951).
"2 M. A. Jermyn and F. A. Isherwood, Biochem. J. 44, 402 (1949).
103 p. Y. V. Ward and P. W. Kent, Nature 170, 936 (1952).
10^ B. D. E. Gaillard, Nature 171, 1160 (1953).
ifts E. Chargaff, E. Vischer, R. Doniger, C. Green, and F. Misani, J. Biol. Chem. 177,

405 (1949).
los A. C. Paladini and L. F. Leloir, Biochem. J. 51, 426 (1952).
1" P. C. Caldwell, Biochem. J. 55, 458 (1953).

SEPARATION BY PAPER CHROMATOGRAPHY 265

Ai-propanol, 30 cc. concentrated ammonia solution, and 10 cc. water. He has
also found useful for detection of ultraviolet absorbing spots a sheet of paper
impregnated with proflavine as a fluorescent screen.

The circular method of paper chromatography of Rutter^"^ and others,
for which improved resolution is claimed, has been applied to nucleic acid
derivatives.^"^

A comprehensive tabulation of ultraviolet absorption data, including
molar extinction coefficients and the r.atios of extinctions at different wave-
lengths, measured at three pH values, has been prepared by Volkin and
Cohn.ii"

'08 L. Rutter, Nature 161, 435 (1948).

'"^ K. V. Giri, P. R. Krishiiaswamy, G. D. Kalyankar, and P. L. N. Rao, Experientia

9, 296 (1953).
''" E. Volkin and W. E. Cohn, in "Methods of Biochemical Analysis," (D. Glick,ed.)

Vol. 1, Interscience Publishers, New York (1954), and m "Methods in Enzy-

mology" (S. P. Colowick and N. O. Kaplan, eds.), Vol. 2, Academic Press, New

York (in press).

CHAPTER 8

The Electrophoretic Separation of Nucleic Acid Components

J. D. SMITH

Page
I. Theory of the Electrophoretic Separation of Nucleic Acid Components. 267

1. General Considerations 267

2. Separation of Nucleotides and Polynucleotides 268

3. Purines, Pyrimidines, and Nucleosides 272

II. Apparatus and Techniques 274

III. Applications of the Method 276

1. Determination of Nucleic Acid Composition 276

2. Separation of Nucleotides from Other Phosphorus Compounds .... 277

3. Isolation of Nucleotides and Polynucleotides from Enzymic and Partial
Chemical Hydrolysates of Nucleic Acids 279

a. Cyclic Ribonucleotides 279

b. Polynucleotides from Ribonucleic Acid 280

c. Polynucleotides and Dinucleoside Monophosphates from Deoxyribo-
nucleic Acid 282

d. Nucleoside Diphosphates 282

4. Separation of Nucleosides and Nucleotides as Borate Complexes . . 283

Nucleic acids and their components bear a number and variety of ioniz-
able groups; and it is natural that electrophoretic techniques should play
an important part in their separation. The development of electrophoresis
on paper has greatly facilitated separations of nucleic acid derivatives which
would be impossible by paper chromatography and time-consuming by
ion-exchange methods. All the manipulations can be carried out on very
small amounts of material. An important feature of the method is that the
relative mobilities of nucleic acid components at any given pH may be
predicted quite accurately, thus aiding in the identification of unknown
components.

I. Theory of the Electrophoretic Separation of Nucleic Acid Components

1. Gener.\l Considerations

A molecule in a fluid subjected to a voltage gradient (E) is acted upon
by a force equal to (EQ), where Q is the net charge on the molecule and is
given by the algebraic sum of the products of the number of ionizing groups

267

268 J. D. SMITH

and their percentage dissociation. This force is opposed by that due to the
resistance of flow of the molecule through the liquid, which is proportional
to the velocity of motion of the molecule (v) and equal to (Kv) , where K is a
function of the size and shape of the molecule and the retarding effects of
other ions in the solution. The molecules thus migrate with a constant
velocity equal to (EQ/K) .

Using the dissociation constants of the ionizable groups in nucleic acid
components and making certain approximations as to the nature of the
resistance to flow of the charged molecules through the fluid, the relative
mobilities of these substances may be calculated. This is best illustrated in
the case of nucleotides and polynucleotides.

2. Separation of Nucleotides and Polynucleotides

The ionizing groups in nucleotides and polynucleotides are the primary
and secondary phosphate groups, the amino groups of adenine, guanine,
and cytosine and its derivatives, and the enol groups of guanine, cytosine,
uracil, and thymine. Table I gives the pK values of these groups for a num-
ber of nucleotides. [For a more complete discussion, compare Jordan, Chap-
ter 13.] The dissociation constants of the pairs of isomeric ribonucleoside-2'-
and -3 '-phosphates are very close to each other, so that all calculations
have been made with the values of Levene and Simms^'^ which are probably
based on mixtures of the two isomers. The dissociation constants of the
deoxy ribonucleotides have not been determined, but to a close approx-
imation they may be assumed to be identical with those of the correspond-
ing ribonucleotides. In Figure 1 are plotted the dissociation curves for the
ionizing groups of the four nucleotides, adenylic, guanylic, cytidylic, and
uridylic acids.

The enol groups of guanylic, cytidylic, and uridylic acids having pK
values of 9.5 or above are not dissociated at the pH values useful for most
separations, and so for the moment may be ignored, but the primary and
secondary phosphate groups and the amino groups are most important.
The pK values of the primary phosphate dissociations which lie between
0.7 and 1.0 are too close to each other to be used for the separation of the
nucleotides; and also some nucleotides and most polynucleotides are un-
stable in this pH range. The separation of adenylic, guanylic, cytidylic,
and uridylic acids is possible on the basis of the differences in the pK values
of the amino groups, which lie between 2 and 5. In this pH range the pri-
mary phosphate groups are completely dissociated and the secondary
phosphate groups uncharged. The most suitable pH for the separation is

» P. A. Levene and H. S. Simms, J. Biol. Chem. 65, 519 (1925).

* P. A. Levene, L. W. Bass, and H. S. Simms, J. Biol. Chem. 70, 229 (1926).
3 P. A. Levene, H. S. Simms, and L. W. Bass, J. Biol. Chem. 70, 243 (1926).

* P. A. Levene and H. S. Simms, /. Biol. Chem. 70, 327 (1926).

ELECTROPHORETIC SEPARATION OF NUCLEIC ACID COMPONENTS 269

TABLE I
The Dissociation Constants of Some Purines, Pyrimidines, Nucleosides, and

Nucleotides"

Primary

Secondary

Substance

(NH:)

First (OH)

phosphate

phosphate

Cytosine^

4.60

12.16

Uracil^

—

9.45

—

—

Thymine"

—

9.94

—

—

Adenosine*

3.3

—

—

—

Inosine'

—

8.75

—

—

Guanosine^

1.6

9.16

—

—

Cytidine^

4.22

12.3

—

—

Uridine 1

—

9.17

—

-r-

Adenylic acid^

3.70

—

0.89

6.01

Adenosine-5'-phosphate^

3.80

—

6.2

Inosine-5'-phosphate'

—

8.88

1.54

6.C4' " •

Guanylic acid^

2.3

9.7

0.70

5.92

Cytidylic acid^

4.24

13.2

0.80

5.97' i

Cytidylic acid^

4.36

.<-. '

(2'-phosphate)

Cytidylic acid*

4. 28

(3'-phosphate)

Uridylic acid^

—

9.43

1.02

5.88

" The values given under the headings adenylic, guanylic, cytidylic, and uridylic acids' were probably
determined on mixtures of the 2' and 3'-phosphates. The amino group pK given for adenosine-5'-phosphate»
does not bear the expected relation to that of the mixture of adenosine-2'- and -3'-phosphates and may be
incorrect.

3.5 where the degree of dissociation of the NH2 group of adenyUc acid is
0.54, giianyhc acid 0.05, cytidylic acid 0.84, while uridylic acid of course
has none. As each nucleotide carries a negative charge of 1 due to the pri-
mary phosphate group, the net negative charges are: adenylic acid 0.46,
guanylic acid 0.95, cytidylic acid 0.16, and uridylic acid 1.00. The four
mononucleotides have very nearly the same size and so approximately the
same resistance to motion; consequently these figures give their relative
mobilities at this pH. With the use of the dissociation curves the relative
mobilities of the nucleotides may be calculated for any other pH.

The pK values of the amino groups of the isomeric nucleoside-2'-, -3'-,
and -5 '-phosphates differ so little that they would only be expected to show
separation in a long electrophoretic mn and in practice generally run to-
gether, although Davidson et at? • * have observed incomplete separation
of guanosine-2'- and -3 '-phosphates at pH 3.5.

^ H. Wassermeyer, Z. physiol. Chern. 179, 238 (1928).

« L. F. Cavalieri, J. Am. Chem. Soc. 74, 5804 (1952).

■' J. N. Davidson and R. M. S. Smellie, Biochem. J. 52, 599 (1952).

« G. W. Crosbie, R. M. S. Smellie, and J. N. Davidson, Biochem. J. 54, 287 (1953).

270

J. D. SMITH

It might be thought that a mononucleotide and the corresponding di-
nucleotide, which below pH 5 has twice the charge and approximately
twice the size, would not differ in electrophoretic mobihty. This is not the

Adenylic acid

4 5

pH

Fig. 1. The dissociation curves of the ionizing groups of adenylic, guanylic, cytidy-
lic, and uridylic acids plotted from the data of Levene and Simms.*

case, for the separation of polynucleotides depends on the relative resist-
ances to motion of a mononucleotide, its dimer, trimer, etc. Before at-
tempting the electrophoretic separation of polynucleotides, Markham and
Smith^ ' ^o calculated the probable limits that could be set to the relative
resistance to motion of a mononucleotide and its dimer. From its diffusion

9 R. Markham and J. D. Smith, Nature 168, 406 (1951).
1" R. Markham and J. D. Smith, Biochem. J. 52, 558 (1952).

ELECTROPHORETIC SEPARATION OF NUCLEIC ACID COMPONENTS 271

constant"' ^^ adenylic acid would appear to have an effective axial ratio
of about 5:1. The axial ratio of the dinucleotide must thus lie between
2.5:1 and 10:1. Taking the shape to be an oblate ellipsoid, and assuming
the resistance to motion to be the same as in translational diffusion, the
ratio of the frictional resistance of the dimer to that of the monomer lies
between the limits 1.32 and 1.85. The observed relative mobilities on
paper of the mono-, di-, and trinucleotides of uridylic acid (after correction
for movement due to flow of the buffer solution through the paper) are
1:1.4:1.5 at pH 3.5. ^"^ At this pH the net charges on the mono-, di-, and
triuridylic acids are 1, 2, and 3 so that if the frictional resistance to motion
of a mononucleotide is taken as 1, that of a dinucleotide is 1.43 and of a
trinucleotide 2.0.

Using these figures and the dissociation cur\es, the mobilities of any
mono-, di-, or trinucleotide relative to uridylic acid may be calculated for
any pH. For example at pH 3.5 the relative mobilities of

AG = (0.46 + 0.95)/1.43 = 0.99

of AC = (0.46 + 0.16)/1.43 = 0.43

and of AAU = (0.46 + 0.46 + 1.0)/2 = 0.96

times that of uridylic acid. (Here A, G, C, and U represent adenylic, gua-
nylic, cytidylic, and uridylic acid residues, respectively.) It is again as-
sumed that the mononucleotides have approximately the same size and
shape, and that the dissociation constants of individual groups are the same,
whether in nucleotides or in polynucleotides.

Evidently the relative mobilities of the mono-, di-, and trimers of a sin-
gle nucleotide form a converging series so that separation of the tetra- and
higher polynucleotides by this method is unlikely.

Remarkably good agreement is found between the mobilities relative to
uridylic acid calculated on this basis and those observed in practice with
the technique of electrophoresis on paper immersed in carbon tetrachloride
(Section II). In Figure 2 are plotted the observed movements of 19 mono-
and polynucleotides in 0.05 m ammonium formate buffer pH 3.5 against
their calculated mobilities relative to that of uridylic acid. These data were
taken from preparative runs in which no particular care was taken to con-
trol the voltage gradient to better than ±10%. Under these conditions
about 25 % of the movement of uridylic acid is accounted for by flow of the
buffer solution due to endosmosis and incomplete saturation of the paper.
Table II gives the calculated relative mobilities and observed movement
of a number of nucleotides and polynucleotides at pH 3.5.

" A. H. Gordon and P. Reichard, Biochem. J. 48, 569 (1951).

»2 G. Schramm, W. Albrecht, and K. Munk, Z. Naturforsch. 7b, 10 (1952).

272

J. D. SMITH

T

0 10 20 30

Distance moved (cm.)

Fig. 2. Graph of mobility observed (in cm./2 hr./20 v. /cm.) against the mobility
(relative to uridylic acid) calculated from the charge and frictional resistance for
19 mono- and polynucleotides'" (two of the points are on top of others).

While 3.5 is one of the most useful pH values, the four nucleotides having
the greatest differences in net charge at pH 3.5, it is not suitable for all
separations. Another useful pH is 5 where all mono-, di-, and trinucleotides
have net negative charges of approximately 1,2, and 3, respectively, and
so separate with relative mobilities of 1:1.4:1.5. It is important to note
that at pH 7-8, where the secondary phosphate is fully dissociated, the
mono-, di-, and trinucleotides have relative charges of 2, 3, and 4 and their
relative mobilities are 1, 1.05, and 1 so that in this pH range their separation
is not possible.

3. Purines, Pyrimidines, and Nucleosides

The ionizing groups on the bases and nucleosides are the amino and enol
groups (the basic dissociations of the purine rings have pK values between
0 and V^ and may be neglected). The dissociation constants of these are
close to, although not identical with, those of the corresponding nucleotides
(Table I), so their electrophoretic separation is similar to that of the cor-
responding nucleotides. At pH 3.5, for example, uridine is uncharged,
while cytidine, adenosine, and guanosine migrate towards the cathode in

13 J. K Wood, J. Chem. Soc. 89, 1839 (1906).

ELECTROPHORETIC SEPARATION OF NUCLEIC ACID COMPONENTS 273

TABLE II

The Calculated Mobilities Relative to Uridylic Acid of Some Nucleotides

AND Polynucleotides at pH 3.5 and Their Observed Movement in 0.05 M

Ammonium Formate Buffer pH 3.5 at 20 v./cm.^" The Relative Positions

OF These Substances on Paper Chromatograms in Solvent 1 Are Also

Given (The Bands Are Numbered 1-6 in Order of Increasing Rf Value) .

Observed movement

Position on the

Calculated mobil-

in 0.05 M ammonium

chromatogram in

ity

relative to

formate buffe

r pH3.5

solvent 1 (bands

that of uridy-

(cm./2 hr.

at 20

numbered in order of

Substance

lie acid

v./cm.

)

increasing Rp value)

A

0.46

8

band 4

G

0.95

14

2

C

0.16

6.5

4

U

1.00

16

4

A!

0.46

8

6

G!

0.95

14

5a (between
5 and 6)

C!

0.16

7

6

U!

1.00

16

6

AC!

0.43

9

5

AU!

1.02

16

5

UU!

1.4

22

5

AC

0.43

8

2

AU

1.02

16

2

AG

0.99

15

1

GC!

0.78

15

3

GU!

1.36

20

3

GC

0.78

13.5

1

GU

1.36

19.5

1

UUU!

1.50

24

5

ACC!

0.39

6

2

AAC

0.54

13.5

1

AAU

0.96

17

1

AGU

1.21

19

1

" A, G, C, and U denote adenylic, guanylic, cytidylie, and uridylic acids, respectively. The symbol ! is
used to indicate a 2',3'-monohydrogen phosphate group.

the reverse order to that in which the mononucleotides move towards the
anode. At this pH the purines and pyrimidines move in the same order as
the nucleosides. These separations have been described by Dimroth et
al.,^* but the order they report for the migration of the bases at pH 3.2 is
not that expected theoretically, nor is it found in practice.

" K. Dimroth, L. Jaenicke, and I. Vollbrechtshausen, Z. physiol. Chem. 289, 71
(1952).

274 J. D. SMITH

n. Apparatus and Techniques

The first attempt at the separation of nucleic acid components by electro-
phoresis was made by Gordon and Reichard,^^ who ran deoxyribonuclease
digests of deoxyribonucleic acid on agar gel. The use of agar as a supporting
medium has obvious disadvantages. It is difficult to avoid overheating and
consequent "sweating" of the gel, and the recovery of the separated ma-
terial from the agar is cumbersome.

Paper is a much more suitable supporting material both for analytical
and preparative work. The chief problems in paper electrophoresis lie in
avoiding the siphoning of buffer down the paper from the electrode vessels,
and the evaporation of water from the paper due to heating during the
run. Both of these prevent the maintenance of a uniform concentration of
bufTer along the paper. Two types of apparatus originally designed for the
separation of amino acids have been used with nucleic acid derivatives.
One is the apparatus described by Durrum^* in which the paper strip passes
over a glass support and both ends dip into buffer solution, the whole being
enclosed in a glass vessel to minimize evaporation. In the second method
the paper is held between two glass plates which are cooled by circulating
liquid. ^^ Neither of these methods satisfactorily overcomes the problem of
evaporation from the paper, and both have to be used with comparatively
low voltage gradients.

A more convenient technique has been described by Markham and
Smith^' " where the paper is kept cool by immersion in carbon tetrachloride.
The apparatus (Figure 3) consists of three rectangular glass jars A, B,
and C (convenient sizes for most work are A, and B, 15 x 10 x 5 cm., C,
21 X 12 X 6.5 cm.). A and B contain the buffer solution and C contains car-
bon tetrachloride. A strip of Whatman No. 3 mm. paper, 56 x 8 cm., is
soaked in the buffer solution, blotted to remove surplus moisture, and
about 0.1-0.2 ml. of a solution containing the mixture to be separated ap-
plied in a line across the paper at a suitable place, usually about 12 cm.
from one end. The paper is placed in the apparatus so that the two ends
dip into vessels A and B and the center part of the paper is immersed in
the carbon tetrachloride in vessel C, being held in place by a piece of cel-
luloid or glass (celluloid often contains ultraviolet-absorbing substances
soluble in CCI4). The electrodes, two arc carbons, are placed in A and B
and the whole covered by a glass plate. The electrodes are connected to a
d.c. power supply, usually of 1000 v. thus giving a voltage gradient of
about 20 v./cm. (Suitable powder supplies are shown in Figure 4.) Under
these conditions a 2-hr. run is adequate for ost eparations. The sub-
's E. L. Durrum, J. Am. Chem. Soc. 72, 2943 (1950).
!» T. Wieland and E. Fischer, Naturwissenchaften 35, 29 (1948).
17 R. Markham and J. D. Smith, Biochem. J. 52, 552 (1952).

ELECTROPHORETIC SEPARATION OF NUCLEIC ACID COMPONENTS 275

5
(cm.)

10

Fig. 3. The electrophoresis apparatus."

stances are loeated by making a contact print in monochromatic ultra-
violet light of 260 m/i,^*^' *^ and may be observed during the run by the
technique of Holiday and Johnson.-"

The carbon tetrachloride keeps the paper cool and prevents evaporation.
It was chosen for three important reasons: (1) it is reasonably transparent
to ultraviolet light; (2) it is nonpolar and so has no tendency to dissolve
out the substances from the paper; (3) it is denser than water and so pre-
vents buffer from siphoning over from the electrode vessels and water-
logging the paper.

With Whatman No. 3 mm. paper about 2 mg. substances may be easily
run per centimeter width of paper; the only effect of increasing the amounts
put on is to make the separating bands broader. Placing the substances in a
line rather than in circular spots gives a much sharper separation.

After an electrophoretic run the substances often have to be submitted
to further chromatographic analyses. This is greatly facilitated if the buf-
fer components move as spots on the chromatograms, or if they are vol-
atile; for this reason buffers such as ammonium formate and ammonium
acetate are the most suitable. Using 0.05 M ammonium formate or acetate
buffer in the size of apparatus described above, at 20 v. /cm. the current
is about 8-10 ma. However, with the 0.05 M phosphate and borate buffers
used for the ranges pH 6-8 and 8-10, it is often necessary to reduce the
voltage gradient to about 15 v. /cm. to prevent overheating.

18 R. Markham and J. D. Smith, Biochem. J. 45, 294 (1949).
" R. Markham and J. D. Smith, Biochem. J. 49, 401 (1951).
=*» E. R. Holiday and E. A. Johnson, Nature 163, 216 (1949).

276

J. D. SMITH

ma.

/

-

+

Ir J

1000 V

V

Output

-

1000 V

?

Fig. 4. Two alternative 1000-v. power supplies for the electrophoresis apparatus.
In circuit B the transformer output must be at least 440 v. r.m.s. at 60 ma. The metal
rectifiers are in pairs of 350 v. 60 ma. The series resistance in the transformer primary
circuit is 1000-5000 ohms rated at 200 ma. and may be replaced "with advantage by a
variable autotransformer.

Under most conditions the movement due to endosmotic flow is neg-
ligible, but when the substances are placed near one end of the paper there
is an additional movement towards the opposite electrode which appears
to be largely due to the movement of buffer into the incompletely saturated
paper. In practice this is unimportant, but may be minimized by placing
the substances near the center of the paper.

III. Applications of the Method

1. Determination of Nucleic Acid Composition

The ease with which adenylic, guanylic, cytidylic, and uridylic acids
may be separated bj^ paper electrophoresis leads to a simple method for
the determination of the quantitative purine and pyrimidine composition
of pentosenucleic acids (PNA), which has been developed by Davidson
et al.'''^'^^ The PNA (about 5 mg./ml.) is hydrolyzed to nucleotides in

" J. N. Davidson and R. M. S. Smellie, Biochem. J. 52, 594 (1952).

ELECTROPHORETIC SEPARATION OF NUCLEIC ACID COMPONENTS 277

0.3 N KOH for 18 hr. at 37°, the solution brought to pH 1 with 60 % wt./wt.
HCIO4 , insoluble KCIO4 removed by centrifugation, and the combined
washings and supernatant fluid brought to pH 4. Aliquots of this solution
containing about 800 /xg. of ribonucleotides are mn at pH 3.5 in 0.02 M
citrate buffer on No. 3 mm. paper (using an apparatus similar to that
described by Durrum'^) for 18 hr. at 11 v./cm. The size of the paper strip
is 72 X 7 cm. (Actually by using the technique of electrophoresis on paper
immersed in CCU^ • ^'^ and with a shorter length of paper and correspond-
ingly higher voltage gradient this separation could be achieved in 30 min-
utes.) Whatman No. 1 paper may also be used for smaller quantities of
material (200 ng.).^ The nucleotides are located in ultraviolet light at
260 m/z, either visually^" or photographically/*'^' eluted, and estimated
spectrophotometrically and by phosphorus determinations. The phosphorus
recovery from runs of individual nucleotides is 96-99%. The PNA phos-
phorus recovered after alkaline hydrolysis and precipitation of KCIO4 is
about 97%, while that recovered as nucleotide P after electrophoresis is
94 % of that in the original PNA. Some of this 6 % loss is probably accounted
for by losses during the manipulations, but part has been shown to be due
to the presence of small amounts of non-nucleotide P in the PNA speci-
mens.

A similar method may be used for the determination of the composition
of deoxyribonucleic acids (DNA) by the electrophoretic separation of the
mononucleotides produced on successive hydrolysis with deoxyribonuclease
and purified snake venom diesterase. Deoxycytidylic and deoxy-5-methyl-
cytidylic acids, however, are not separated in the electrophoretic nm.

Matthews^^ has used paper electrophoresis in borate buffer pH 9 to sep-
arate 8-azaguanylic acid from alkaline hydrolysates of the PNA of tobacco
mosaic virus isolated from plants treated with 8-azaguanine. The pK of
the enol (OH) group of 8-azaguanine is about 3 pH units lower than that
of guanine, so that at pH 9 8-azaguanylic acid has a higher mobility than
any other mononucleotide.

2. Separation of Nucleotides from Other Phosphorus Compounds

A considerable part of the work on the incorporation of the radioactive
isotope P^^ into PNA has been based on the assumption that the ribonucleo-
tide fraction obtained by the Schmidt and Thannhauser procedure^^ con-
tains only PNA phosphorus.

By separating this ribonucleotide fraction through paper electrophoresis
Davidson and Smellie'' have shown that only about 75% of the P can be
accounted for as nucleotides, there being present at least six additional P

22 R. E. F. Matthews, Nature 171, 1065 (1953).

2» G. Schmidt and S. J. Thannhauser, J. Biol. Chem. 161, 83 (1945).

278

J. D. SMITH

T

F.

T
Cy.

T

Ad.

T
Gu.

T T T

E. Ur. D.

T
C.

T T
B.A.

Direction of run

Fig. 5. (a) Ultraviolet print at 260 mp and (b) autoradiograph of a short electro-
phoretic run of rat liver ribonucleotide fraction prepared by a modified Schmidt and
Thannhauser method. The substances showing in the autoradiograph contain P".
Cy, Ad, Gu, and Ur represent cytidylic, adenylic, guanylic, and uridylic acids, re-
spectively. (Reproduced from Davidson and Smellie.')

compounds. [Cf. Smellie, Chapter 26.] P^^ \vas administered subcutaneously
to rats and after suitable intervals the animals sacrificed and their liver
tissue treated with liquid solvents and cold trichloracetic acid to remove
lipid and acid-soluble P. The tissue was then hydrolyzed in 0.3 A'' KOH
for 18 hr. at 37°, 60 % wt./wt. HCIO4 added to give pH 1, thus precipitating
the DNA and KCIO4 , the combined supernatant and washings adjusted to
pH 4 and run on paper electrophoresis at pH 3.5 in 0.02 M citrate buffer.
The nucleotides were located by their ultraviolet absorption and P^^.^q^.
taining compounds shown up on autoradiographs. The autoradiographs
(Figure 5) show the presence of six P^--containing compounds other than
adenylic, guanylic, cytidylic, and uridylic acids, of which one (A) absorbs
in the ultraviolet at 260 mfx. Component B is inorganic phosphate; the
other substances are unidentified organic phosporus compounds. As P^^
(administered as inorganic phosphate) is initially incorporated into these
substances much more rapidly than into the ribonucleotides, contamination
of the latter by any of these substances leads to gross errors in the esti-
mation of P^2 incorporation into ribonucleic acid. By a long electrophoretic
run (18 hr., at 11 v. /cm.) cytidyhc, adenylic, and guanyhc acids may be
freed from other P^^.^ontaining substances and their true specific activ-
ities measured, but it is difficult to separate uridylic acid from components
D and E. These interfering substances can only be satisfactorily eliminated
by the prior isolation of PNA from the tissue.

ELECTROPHORETIC SEPARATION OF NUCLEIC ACID COMPONENTS 279

Deimel and Maiirer-* have described an electrophoretic method for the
separation of FN A and DNA fractions which is based on the Schmidt and
Thannhaiiser procedure.-^ After hydrolysis in KOH the solution containing
ribonucleotides and DNA is run on paper electrophoresis at pH 8.6, when
the ribonucleotides move together and more slowly than DNA. Inorganic
phosphate runs just ahead of the DNA. The extent to which these fractions
may be contaminated with other phosphorus compounds is not clear.

3. Isolation of Nucleotides and Polynucleotides from Enzymic or
Partial Chemical Hydrolysates of Nucleic Acids

The paper electrophoretic technique is perhaps shown to best advantage
in the isolation of the products of the partial breakdown of nucleic acids
by enzymic or chemical hydrolysis.

A mixture of polynucleotides such as that found in ribonuclease digests
contains 30 or more different substances whose direct separation by electro-
phoresis would be impossible. In practice complex mixtures of this type
are first fractionated by nnining chromatographically on Whatman No. 3
mm. paper in 70% isopropanol-water (vol. /vol.) with NH3 in the vapor
phase (solvent 1). This separates the substances produced in the incomplete
ribonuclease digestion of PNA into at least 6 bands (which for convenience
are numbered 1 to 6 in order of increasing Rp value). ^' ^°' " Each of these
bands contains a relatively small number of substances, many of which
may be isolated by a single electrophoretic run.

a. Cijclic Ribonucleotides

The cyclic ribonucleotides (nucleoside-2',3'-monohydrogen phosphates)
are intermediates in the acid and alkali breakdown of PNA [Cf. Brown and
Todd, Chapter 12, and Baddiley, Chapter 4] and may be isolated in small
yields from very weak alkaline hydrolysates of PNA''^ (e-g-, 1 hr. at 100°
in water over solid BaCOs). During the course of ribonuclease digestion
the cyclic pyrimidine nucleotides are formed as intermediates in the lib-
eration of the pjTimidine mononucleotides,^' " while complete ribonuclease
digests contain very small amounts of cyclic purine nucleotides, originating
from end groups of the nucleic acid."

As they lack a secondary phosphate group, the cyclic nucleotides move
faster than polynucleotides or other mononucleotides on chromatograms
run in solvent 1. G! moves just below the cyclic dinucleotides such as AC!,
etc. (band 5a), and A!, C !, and U! move together with a higher Rf value
(band 6). (The symbol ! is here used to denote a 2',3'-monohydrogen
phosphate group.) These bands to some extent overlap with those of the

2" M. Deimel and W. Maurer, Nalurwissenschaften 39, 489 (1952).
" R. Markham and J. D. Smith, Biochem. J. 52, 565 (1952).

280 J. D. SMITH

corresponding nucleosides produced during the BaCOs hydrolysis. After
elution from the bands the cyclic nucleotides are isolated by electrophoresis
at pH 3.5, when they run in the positions of the corresponding noncyclic
nucleotides.

As they bear no secondary phosphate group, at pH 7.4 the cyclic nucleo-
tides carry only one negative charge, while other nucleotides carry two
(Figure 1). This gives a simple electrophoretic method of separating cyclic
nucleotides from other mononucleotides.

b. Polynucleotides from Ribonucleic Acid

With the exception of the very small amounts of polynucleotides orig-
inating from end groups the polynucleotides in complete ribonuclease
digests are of one type.^' ^^'^^ They consist of chains of one or more purine
nucleoside residues joined through 3',5'-phosphodiester links, and ter-
minated by a pyrimidine nucleoside residue which is joined through a
phosphodiester link on its 5 '(OH)- to the 3'- position on the adjacent purine
nucleoside residue, and which bears a singly esterified phosphate in the
3'-position. These are thus of the type, AC, AU, AGC, etc. In partial di-
gests of PNA are also found the corresponding polynucleotides bearing
cyclic (2',3'-monohydrogen phosphate) groups on the terminal pyrimidine
nucleoside residue, AC!, AGC!, etc. The partial digests also contain small
amounts of pyrimidine polynucleotides with cyclic end groups such as
UU!, UUU!, CC!, etc.

The movement of these substances on a paper chromatogram in solvent
1 largely depends on three factors, namely, (1) mononucleotides move
faster than the corresponding dinucleotides, and dinucleotides move faster
than trinucleotides; (2) substances with cyclic terminal phosphate groups
move faster than the corresponding substances with singly esterified ter-
minal phosphates ; (3) guanylic acid and its derivatives move more slowly
than the corresponging mono- and polynucleotides. Chromatography in
this solvent thus fractionates ribonuclease digests into a number of bands
from which many individual polynucleotides may be separated by single
electrophoretic runs at pH 3.5. The chromatographic separation and electro-
phoretic mobilities of these are illustrated in Table II. Often more than
one electrophoretic run is necessary to separate the substances in a single
chromatographic band. For example band 1 of such a chromatogram eluted
and subjected to electrophoresis at pH 3.5 for 2 hr. at 20 v. /cm. gives a
band moving 13.5 cm. (towards the anode). This is a mixture of GC and
AAC which may be resolved by eluting and either (1) running at pH 5,
when the trinucleotide moves faster than the dinucleotide, or (2) running
at pH 2.5 for 3 hr. at 16 v. /cm., when GC moves 4.8 cm. and AAC 6.4 cm.

ELECTROPHORETIC SEPARATION OF NUCLEIC ACID COMPONENTS 281

Fig. 6. The dissociation curves of GU and GU! and the electrophoresis at pH 7.4
of GU! and GU (produced by the ribonuclease treatment of GU!).'" GU differs from
GU! in having an extra (secondary phosphoric OH) acidic group. The substances were
placed at the top of the paper, GU! on the left and GU on the right.

The dinucleotides with cycUc terminal phosphate groups are separated
from the corresponding nucleotides with singly esterified terminal phos-
phates on the initial chromatogram. These are also separated on electro-
phoresis at pH 7.4 when the noncyclic dinucleotide bears an additional
charge due to the dissociated secondary phosphate group (Figure 6). The
polynucleotides in incomplete acid hydrolysates of PNA may be separated
in the same way. Although polynucleotides bearing cyclic phosphate groups
are absent, these hydrolysates also contain polynucleotides with terminal
purine nucleotide residues; consequently, more than one electrophoretic
run is often necessary to obtain individual polynucleotides.

Isomers such as AGU, GAU, etc. are not separated on electrophoresis.

282 J. D. SMITH

TABLE III
The Observed Electrophoretic Mobilities of Some Dinucleoside
Monophosphate Diesters and Similar Substances at pH 3.5 in
0.05 M Ammonium Formate Buffer^^. 2^

Observed movement in 0.05 M
ammonium formate buffer pH 3.5
Substance" (cm./2 hr. at -20 v. /cm.)

A-p-A 4.2

A-p-G 6.0

A-p-C 2.0

A-p-T 6.2

G-p-C 3.0

G-p-MC 3.0

G-p-T 9.0

C-p-C 0

C-p-T 3.9

T-p-T 10.9

A-p-C-p-T 6.2

A-p-MC-p-T 6.2

" A, G, C, MC, and T represent deoxyadenosine, deoxyguanosine, deoxycytidine, deoxy-5-methylcytidine
and thymidine, respectively, and p represents a phosphate linking two nucleoside residues.

c. Polynucleotides and Dinucleoside Monophosphates from Deoxyribonucleic
Acid

Very similar techniques have been employed to isolate some of the poly-
nucleotides produced by the action of deoxyribonuclease on DNA.^^ A
combination of chromatography in solvent 1 and electrophoresis at pH
3.5 has been used to separate the dinucleoside monophosphates and analo-
gous compounds found in deoxyribonuclease digests which have been
treated with phosphomonoesterase.^^' 27 The principles of the electropho-
retic separation of these substances are similar to those for the separation
of polynucleotides. They contain one less primary and secondary dissoci-
ation per molecule than do the corresponding polynucleotides, and their
effective frictional resistances to motion are slightly different. The electro-
phoretic properties of some of these are given in Table III.

d. Nucleoside Diphosphates

The nucleoside diphosphates bear two primary and two secondary phos-
phate dissociating groups. At pH 7 and above they are the fastest-moving
nucleic acid components, while at pH 3.5 they are readily separated from
each other and from the corresponding mononucleotides as they bear an
additional negative charge due to the second primary phosphate group.

26 J. D. Smith and R. Markham, Nature 170, 120 (1952).

" J. D. Smith and R. Markham, Biochim. et Biophys. Acta 8, 350 (1952).

ELECTROPHORETIC SEPARATION OF NUCLEIC ACID COMPONENTS 283

4. Separation of Nucleosides and Nucleotides as Borate

Complexes

Under certain conditions sugars will form complexes with boric acid
believed to be of the types I, 11, and III.

— C— O

— C— O
I

I

B— OH

C— O OH

\ /
B

/ \
C— O OH

I

II

H+

— C— O

— C— O

O— C—

O— C—

H+

III

Types II and III are only formed when two adjacent cis-hydroxyls are
present in the sugar ring, and, in dilute solutions, when the relative pro-
portions of borate to sugar are high, type II predominates.-* It is thus
possible by borate complex formation to introduce a new dissociating acid
group into sugar residues possessing two m-hydroxyl groups.

The nucleic acid components bearing czV(OH) groups are ribose (which
in the pyranose form has four), and ribonucleosides, ribonucleoside-5'-
phosphates, and dinucleoside monophosphates from PNA (in the 2'- and
3 '-positions). Jaenicke and Vollbrechtshausen^^ have made use of borate
complex formation in the separation of some nucleosides and nucleotides
by electrophoresis in 0.1 ilf borate buffer pH 9.2. The pK of the first dissoci-
ation of boric acid is 9.0, so that at this pH the acidic group of the complex
is partially dissociated and the ribonucleoside-5 '-phosphates bear an ad-
ditional net charge and, therefore, move faster than other nucleotides.
Under these conditions the ribonucleosides move towards the anode in the
order (of increasing mobility): adenosine, guanosine and cytidine, uridine;
the rate of movement presumably being determined by the negative charges
due to the borate complex and those on the enol (OH) groups of the bases.
The deoxy ribonucleosides which only carry charges on the partially dis-
sociated enol (OH) groups move more slowly than the ribonucleosides.

" H. S. Isbell, J. F. Brewster, N. B. Holt, and H. L. Frush, J. Research Natl. Bur.

Standards iO, 129 (1948).
2' L. Jaenicke and I. Vollbrechtshausen, N aturwissenschaflen 39, 86 (1952).

284 J. D. SMITH

At pH 9.2 in borate buffer the bases, the ribonucleosides, and the nucleo-
tides may be separated into three groups in that order. ^"^

Addendum

Microtechnique of Electrophoresis

The electrophoretic techniques hitherto described are suitable for the
separation and analysis of the nucleotides or bases from 100 Mg- or more of
nucleic acid. Edstrom*" has developed a method which permits the analysis
of 100-1000 pg. of ribonucleic acid (1 pg. = 10~^- g.), which is of the same
order as the amount contained in single nerve cells.

The supporting medium is a piece of copper-silk (cellulose) previously
treated with alkali to swell the fibers. The fiber is soaked in a highly viscous
citrate buffer pH 3.6, containing glucose and glycerol, and stretched on a
quartz slide between dabs of paste moistened with the buffer. During the
electrophoresis the fiber is immersed in liquid paraffin. The substances
to be separated are placed on the fiber with a micropipet and a potential
applied through platinum electrodes between the dabs of paste. With a
voltage gradient of 12 v./mm. the adenine, guanine, and cytidyhc and
uridylic acids from a hydrolysate of PNA in A^ HCl are separated in 2 hr.
over a distance of 1 mm. The fiber is photographed together with a light
calibration system in monochromatic light of 257 mn and 275 m^, respec-
tively, and the densities of the calibration system and parts of the fiber are
measured photometrically on a photographic strip.

Detection of Terminal 5' -Phosphate Groups in Ribonucleic Acids

Markham, Matthews, and Smith^^ have shown that the ribonucleic acids
of tobacco mosaic virus and potato virus X contain a type of chain which is
terminated at one end by a nucleoside residue bearing a singly esterified
phosphate group on the 5'-position and at the other end by a nucleoside
residue with no singly esterified phosphate group. On alkaline hydrolysis
this type of chain gives in addition to mononucleotides, nucleoside-2' , 5'-
and nucleoside-3',5'-phosphates from the terminal residues bearing the
phosphate groups, and nucleosides from the other end of the chain. These
can be separated by paper electrophoresis at pH 7.4 when the nucleosides,
which are uncharged, remain at the starting point while the nucleoside
diphosphates which carry four negative charges move towards the anode
ahead of the nucleoside monophosphates.

30 J. E. Edstrom, Nature 172, 809 (1953).

" R. Markham, R. E. F. Matthews, and J. D. Smith, Nature 173, 537 (1954),

CHAPTER 9

Color Reactions of Nucleic Acid Components

ZACHARIAS DISCHE

Page
I. Color Reactions of Nucleic Acids and their Constituents Based on Reactions
of Their Sugars 286

1. Color Reactions of Deoxyribonucleic Acid and Its Nucleotides .... 286

a. Reaction with Diphenylamine 287

b. Reactions of 2-Deoxyribose, DNA and Its Nucleotides with Cj-steine
and Sulfuric Acid 290

(1) Cysteine Reaction of 2-Deoxyribose and DNA with 75 Vol. %
H2SO4 .' 291

(2) Reaction of DNA with Cysteine and Concentrated H2SO4. ... 294

c. Reaction of DNA with Tryptophan and Perchloric Acid 296

d. Reaction of DNA and 2-Deoxyribose with Indole and HCl 297

e. Reaction with Carbazole and Sulfuric Acid 298

f. Reaction of DNA with Schiff' s Reagent 299

g. Evaluation of Various Methods of Quantitative Determination of

DNA and Its Constituents 299

2. Reaction of Ribonucleic Acids 300

a. Orcinol Reaction • 300

b. Phloroglucinol Reaction of Euler and Hahn 302

c. Reaction of Pentose with Cysteine and H2SO4 302

II. Determination of Purine and Pyrimidine Bases of Nucleic Acids by Charac-
teristic Color Reactions 303

1. Determination of Adenine 304

2. Determination of Thymine 304

3. Determination of Cytosine and Uracil 305

The determination of nucleic acids can be based on the determination of
one of their three main components, namely phosphoric acid, the sugar, or
the nitrogenous bases. As regards the determination of phosphoric acid,
this procedure utilizes the differences in alkali lability of the two types of
nucleic acid and differences in their solubility in acids. [Of. Leslie, Chapter
16.] It is clear that these methods of determination cannot be regarded as
specific, and, while it was found that in certain tissues like liver,^ the meth-
ods give satisfactory results as far as agreement with other methods is
concerned, it was reported more recently^ that in some tissues, like certain

' W. C. Schneider, J. Biol. Chem. 161, 293 (1945); 164, 797 (1946).
' L. Drasher, Science 118, 181 (1953).

285

286 ZACHARIAS DISCHE

tumors, very large discrepancies are encountered between determinations
based on phosphorus and those based on other constituents. Recently,
several sensitive color reactions for the purine and pyrimidine bases were
developed. These reactions have not yet been sufficiently used and tested
as to their suitability for the determination of the bases in nucleic acids
and tissues. More widely used and better tested are methods using color
reactions of the sugars of nucleic acids. Because of the fundamental dif-
ference in the nature of the respective sugars in the ribo- and deoxyribo-
nucleic acids, the methods based on characteristic color reactions of these
sugars do not require any separation of the two types of nucleic acids, and
are therefore less tedious and simpler in operation.

I. Color Reactions of Nucleic Acids and their Constituents
Based on Reactions of Their Sugars

Of the two types of nucleic acid, deoxyribonucleic acid contains in all
its nucleotides 2-deoxyribose in the furanoid configuration, while the nucleo-
tides of the pentose nucleic acids were shown to be ribofuranosides. [Cf.
Chapters 2 and 4.] The color reactions of the two types of nucleic acid
based on the sugar components are therefore general reactions of the 2-
deoxypentoses or pentoses, respectively.

1. Color Reactions of Deoxyribonucleic Acid and Its Nucleotides

Among the color reactions of 2-deoxypentoses so far described three
groups can be distinguished. One represents a type of reaction character-
istic for aldehydes and is due to the fact that 2-deoxypentose, in contrast to
other saccharides, is present in solution largely in its aldehydic form.^ A
second group is represented by a type of reaction given with varying in-
tensity by various hydroxy and keto aldehydes; and a third by reactions of
furan derivatives which are formed from saccharides. It seems noteworthy
in this connection that the ring form of 2-deoxypentose can, by simple
dehydration, produce furfuryl alcohol, while the straight chain form of the
sugar by a simple dismutation between carbon 3 and 4 can be transformed
to a highly reactive compound, the w-hydroxylsevulic aldehyde. On the
other hand, the transformation of pentoses to the corresponding furan
derivatives is probably preceded by an inner rearrangement of the molecule
consisting of a dehydration and double bond formation between carbons 2
and 3,* and proceeds, therefore, less easily than the formation of furan
derivatives of 2-deoxypentose.

3 N. G. Overend, J. Chem. Soc. 1950, 2769.

*M. L. Wolfrom, R. D. Schuetz, and L. F. Cavalieri, /. Am. Chem. Soc. 70, 514
(1948).

COLOR REACTIONS OF NUCLEIC ACID COMPONENTS 287

a. Reaction with Diphenylamine

This reaction was first described in 1930^ and recommended for quanti-
tative determinations of DNA in animal tissues by colorimetry in white
light. In 1940 Seiberf^ adapted this method for quantitative determination
with a photoelectric colorimeter. The sensitivity of the reaction can be still
considerably increased by using a photoelectric spectrophotometer.

Procedure: — The procedure generally used for determinations of DNA is as follows.
One volume of the solution containing 50 to 500 ng. DNA per cc. is mixed with a
double volume of a reagent, which is prepared by dissolving 1 g. of diphenylamine
(twice recrystallized from 70% ethanol or petroleum ether) in 100 cc. of glacial acetic
acid of highest purity. 2.75 cc. of concentrated sulfuric acid (reagent grade) is added
to 100 cc. of the diphenylamine solution. The reaction mixture is heated at 100° for at
least 10 minutes. A blank containing water instead of the unknown is run simulta-
neously. The DXA solution gives a blue color which does not change significantly for
hours. The absorption curve determined with the Beckman spectrophotometer shows
a sharp maximum at 595 m^.

Influence of time of heating and concentration of the acid on the reaction: — Most
authors report that the color intensity reaches its maximum after 10 minutes' heating.
The color, however, develops very fast in the beginning, so that for qualitative pur-
poses a shorter heating of 3 minutes appears reasonable. Higher concentration of the
acid may increase the intensity of the color, but at the same time it decreases the
specificity.

Specificity of the reaction: — ^The blue color developed in the diphenylamine
reaction was shown to be produced not only by DNA and 2-deoxyribose,
but also by 2-deoxyxylose-7 it is therefore a reaction of 2-deoxypentoses
and not only of 2-deoxyribose. Of the constituent mononucleotides of DNA,
the two purine nucleotides give the reaction with an intensity twice that of
the equivalent amount of DNA.^ Preparations of apurinic acid, prepared
and analyzed in Dr. Chargaff's laboratory,^ showed a 4 to 10% lower ab-
sorption at 595 mju than equivalent amounts of DNA. [Cf. Chargaff, Chapter
10.] This difference appeared already after 3 minutes' heating and did not
increase on further heating. It can be assumed that it is due to a slight
destruction of the sugar of the purine nucleotide during the preparation
of the apurinic acid. Deriaz et al}^ report that a DNA preparation from
fish sperm showed a slightly lower color intensity measured with the Spek-
ker absorptiometer than one-half of the equivalent amount of 2-deoxyri-
bose. Of the pyrimidine nucleotides, thymidylic acid in higher concen-

6 Z. Dische, Mikrochemie 8, 4 (1930).

6 F. B. Seibert, J. Biol. Ctiem. 133, 593 (1940).

7 W. G. Overend, F. Shafizadeh, and M. Stacey, J. Chem. Soc. 1950, 1027.
* A. E. Mirsky, Advances in Enzymol. 3, 1 (1943).

9 C. Tamm, M. E. Hodes, and E. Chargaff, J. Biol. Chem. 195, 49 (1952) .
10 R. E. Deriaz, M. Stacey, E. G. Teece, and L. F. Wiggins, J. Chem. Soc. 1949, 1222

288 ZACHARIAS DISCHE

tration shows a faint blue color and an absorption maximum at 595 m/x.
The extinction coefficient of this nucleotide, however, is less than one-
sixtieth of that of DNA, while cy tidy lie acid shows no significant color at
all. The reaction of DNA, therefore, represents the reaction of its purine
nucleotides only. (Brady and McEvoy-Bowe^°'^ reported that it is possible
also to determine the deoxypentose of the pyrimidine nucleotides of DNA
with the diphenylamine reaction after preceding bromination which makes
the pyrimidine-sugar link more easily hydrolyzable.) It is obvious that the
splitting of the glycosidic linkage of the sugar is necessary for the reaction.
On the other hand, substitution in position 5 and 3 of the sugar by phos-
phoric acid apparently does not significantly affect the intensity of the
reaction. The blue color with the absorption maximum at 595 m/x is not
produced by any of the normal sugars or hexals, hexuronic acid, aliphatic
and aromatic aldehydes, glycolic aldehyde, trioses, aldo- and ketotetrose,
nor furfuraldehyde.^ Arabinal, on the other hand, produces the blue color
with the identical absorption spectrum,^ and the extinction coefficient,
as measured in the Spekker absorptiometer, is about one-third higher than
that of 2-deoxyribose. Finally, AUerton et al}^ recently repoited that 3-
deoxyxylose, as well as 2,3-deoxyxylose, produces a blue color with the
diphenylamine reagent. The extinction coefficient at the maximum of
absorption, however, is for both substances about one-thirtieth of that of
2-deoxyribose, and the color develops much more slowly. No data about
the absorption maximum and the form of the absorption curve were
reported for these two deoxysugars, and it is therefore not established
whether the blue color produced is identical with that derived from 2-
deoxypentoses.

Mechanism of the reaction: — The mechanism of the diphenylamine re-
action was extensively studied by Deriaz et al}'^ It was found that co-hy-
droxylaevulic aldehyde, but not laevulic acid, gives with the reagent a
blue color with an absorption spectrum identical or very similar to that
derived from 2-deoxyribose. They furthermore found that furfuryl alcohol
also gives the same reaction. Further studies showed that it is possible to
obtain co-hydroxylaevulic aldehyde by heating with methanolic HCl in lar-
ger quantities from furfural alcohol and arabinal than from 2-deoxyribose.
In agreement with that, the molar extinction coefficient of the first two
compounds in the diphenylamine reaction is considerably higher than of
the last one. The formation of co-hydroxylsevulic acid from 2-deoxyribose
and arabinal involves an inner dismutation between carbon 3 and 4 in which
the hydroxy 1 of carbon 3 is shifted to carbon 4, while in the case of furfuryl
alcohol this shift is brought about simply by hydrolysis of the oxygen ring.

i"* T. G. Brady and E. McEvoy-Bowe, Nature 168, 299 (1951).

" R. AUerton, E. G. Overend, and M. Stacey, J. Chem. Soc. 1952, 213.

COLOR REACTIONS OF NUCLEIC ACID COMPONENTS 289

The ease with which the hydroxyl carbon 3 is shifted is due to its beta po-
sition in relation to the carbonyl group which, as shown by Wolfrom ei
al.^ for other sugars, faciUtates this type of dismutation. The weak positive
reaction of 3-deoxyxylose could be explained in a similar way by assuming
that a small amount of the corresponding 3-deoxyketose is formed by isom-
erization between carbons 1 and 2. As can be easily seen in this sugar, the
shift of the hydroxyl from carbon 4 (in beta position to the carbonyl group)
to carbon 5, easily produces also w-hydroxylsevulic aldehyde. Deriaz et
al}'^ also isolated from the reaction mixture in the diphenylamine reaction
C^ different compounds by chromatography and showed that only one of
these compounds, representing apparently the product of a reaction be-
tween w-hydroxylsevulic aldehyde and diphenylamine, is responsible for
the blue color. This compound represents only 10% of the total reaction
products; it is yellow at neutral and alkaline reaction and turns blue at a
certain acidity.

Interference from other substances: — Most sugars and their analogues, as
well as aliphatic and hydroxy aldehydes, produce various colors with the
diphenylamine reagent, different from the color given by 2-deoxypentose.
When present in solution these compounds, therefore, can interfere with
the qualitative and quantitative determinations of deoxy nucleotides. The
green colors produced by most normal sugars appear with significant in-
tensity only at such high concentrations that in general they will not be of
practical importance. Certain carbohydrates, such as agar and carbohydrate
from carragheen moss, which contain anhydrosugars give an intensive
green color. ^- Of greater importance as far as interference with the reaction
of nucleic acids in concerned, are certain compounds of unknown nature
which are widely distributed in animal and plant tissues in combination
with proteins, ^2 and which give an intensive purple color which resembles
very much the color produced by hexals and 2-deoxyhexoses;^^ but, as was
found in our laboratory, while in the case of the latter two substances the
maximum absorption is at 510 m^ (Beckman spectrophotometer), the
substances associated with proteins have maxima around 530 m/i. A similar
color with absorption maximum at 560 m/x is produced by lipid-soluble sub-
stances found in plants.^'' When present in solution in significant amounts,
these substances can severely interfere with the reaction of the nucleic
acids. Overend et aU'" reported that histone and protamines as well as cer-
tain amino acids like glycine, isoleucine, and glutamic acid increase the

'2 N. W. Pirie, Brit. J. Exptl. Pathol. 17, 269 (1936).
'3 W. Ayala, L. V. Moore, and E. L. Hess, J. Clin. Invest. 30, 781 (1951).
'^ M. Ogur, R. O. Erickson, G. Rosen, K. B. Sax, and C. Holden, Exptl. Cell Re-
search 2, 73 (1951).
'!* W. G. Overend, /. Chem. Soc. 1951, 1484.

290 ZACHARIAS DISCHE

speed of the development of the blue color in the diphenylamine reaction,
even at concentrations corresponding to half of the concentration of DNA.
This effect, however, disappears almost completely when the heating is
continued until the maximum of the color is reached. Cohen^^ found pre-
viously that the nucleoprotein from calf thymus and the corresponding
nucleic acid, show identical color intensity when solutions of the two sub-
stances contained identical amounts of organic phosphorus.^®*

Detection of DNA and its constituents by dichromatic readings: — When substances
which produce colored products with the diphenylamine reagent are present in solu-
tion and obscure the reaction of 2-deoxyribose, it is often, accordingto our experi-
ments, possible to detect the presence of the latter by measuring the optical density
of the unknown at two different wavelengths. When substances which produce a green
color are present, the difference of the density at 595 and 650 m/x will in general be
either zero or negative for these substances, while it is positive for 2-deoxyribose. If,
therefore, the amount of the latter substance is not too high, a positive Dm — Duo
will be obtained even when the blue color is no more clearly recognizable. In the case
of acetaldehyde, which produces a brown color, Dm — Deso is almost zero^". In the
case of substances which yield purple or yellow colors, the difference of the densities
at 595 m/x and lower wavelengths may also be used to advantage for the detection of
2-deo.xypentose.i^

Quantitative determination of DNA and its constituents: — Dsgs , or its equivalent
with photoelectric instruments using filters, is proportional to the concentration of
DNA. In general, however, it can be recommended to use Dsgs — i^eso which, as was
found in our laboratory, is also proportional to the concentration of DNA as a meas-
ure of this concentration. Although this difference is about 40% smaller than Dm ,
the agreement between duplicate samples was found to be better than when Dggs is
used as a measure of the concentration. In mixtures of DNA and its breakdown prod-
ucts formed by the liberation of bases, the DNA can be determined by carrying out
first the reaction on the tissue extract and then bringing the latter to an alkalinity of
about O.l to 0.2 A'^, heating for 2 minutes at 100°, and repeating the determination.
The free sugar of the breakdown product, as was found in our laboratory, is com-
pletely destroyed by this procedure, while DNA is not affected. The difference be-
tween the two determinations gives then the amount of DNA. By the use of a spectro-
photometer as little as 50 ^g. of DNA can be determined fairly accurately. Steele
et al.^^ described a micromethod which permits the determination of as little as 1.6
tig. of DNA by the diphenylamine reaction.

6. Reactions of 2-deoxyribose, DNA and its nucleotides with cysteine and
sulfuric acid

2-Deoxyribose gives two completely different reactions when treated
with sulfuric acid in the presence of cysteine, depending upon the con-

1" S. S. Cohen, J. Biol. Chem. 167, 691 (1945).

'** H. V. Euler and L. Hahn reported {Arch, neerl. physiol. 28, 398, 1946) that certain
proteins increase the intensity of the diphenylamine reaction when present in
amounts corresponding to twice to three times that of DNA. G. Bergold (Z. Nat-
urforsch. 36, 406, 1948) could not confirm these results.

" E. Racker, Nature 167, 408 (1951).

" R. Steele, S. Theodore, and L. Ottolenghi, /. Biol. Chem. 177, 231 (1949).

COLOR REACTIONS OF NUCLEIC ACID COMPONENTS 291

centration of the acid and temperature of the reaction. In one of these
reactions concentrated acid is added to the solution and all carbohydrates
form reaction products with characteristic absorption spectra in the ultra-
violet and the blue part of the spectrum. The absorption curves, however,
differ from one class of saccharides to another. This reaction will be desig-
nated, therefore, as cysteine reaction of 2-deoxyribose with concentrated
H2SO4 . The second reaction is much more selective and can be regarded as a
characteristic color reaction of 2-deoxypentose. In it an H2SO4-H2O mixture
containing 75 vol. of H2SO4 in 100 cc. of the mixture is added to the sugar
solution.

(1) Cysteine Reaction of 2-Deoxyrihose and DNA with 75 Vol.7o H2S0^.
This reaction was first described in 1944'^ and recommended for quantita-
tive determination in white light. Later Stumpf^" modified slightly the
procedure and adapted it for spectrophotometric measurements. Every
one of these modifications offer certain advantages for special purposes, and
both precedures therefore are here described.

Original procedure of Dische: — 1 cc. of a solution containing 50 to 500 ng per cc. of
DNA is mixed with 0.05 cc. of cysteine hydrochloride and 5 cc. of 75 vol.% sulfuric
acid (70 vol. of H2SO4 cone, plus 30 vol. of H2O) under cooling in ice. The reaction
mixture is then transferred for 5 minutes into a water bath of 40°C. and then cooled
in tap water. A pink color develops which after 15 minutes reaches 90% of its maxi-
mum, and the maximum after about 1 hour. The absorption maximum is at 490 mfi.
The color is stable for hours.

Modification of Stumpf: — 0.05 cc. of cysteine hydrochloride and 5 cc. of 75 vol.%
sulfuric acid are added to 0.5 cc. of a solution containing 25 or more ^tg. per cc. of DNA
without cooling. The mixture is allowed to cool off at room temperature. The same
pink color with an absorption maximum at 490 m/i develops and reaches its maximum
after about 15 minutes.

Specificity of the reaction: — ^The molar extinction coefficients of 2-deoxyri-
bose, deoxyadenylic and -guanylic acids do not differ significantly in
Stumpf 's modification. 20 Preparations of apurinic acid, however, which
react 10 % weaker in the diphenylamine reaction than an equivalent amount
of DNA, react 18 % weaker in either modification of the cysteine reaction.
Thymidylic acid, which shows a negligible reaction with diphenylamine,
reacts strongly in the cysteine reaction. The molar extinction coefficient
for the free thymidylic acid was found^* to be 74 % of that for the purine
nucleotides in the Stumpf modification. In the original form of the reaction,
however, the thymidylic acid, according to our own observations, shows a
completely different behavior. When the optical density is determined
immediately after the heating period, it is less than 20 % of that of an equiv-
alent amount of DNA. The intensity of the color, however, increases rap-

>9 Z. Dische, Proc. Soc. Exptl. Biol. Med. 55, 217 (1944).

20 P. Stumpf, /. Biol. Chem. 169, 367 (1947).

21 L.A. Manson and J. O. Lampen, J. Biol. Chem. 191, 87 (1951).

292 ZACHARIAS DISCHE

idly and the maximum is reached only after about 6 hours. At this moment
the molar extinction coefficient corresponds to about 120% of that of DNA.
Deoxycy tidy lie acid, on the other hand, does not show any color in the
reaction. The reaction of thymidylic acid, however, is only so pronounced
when free thymidylic acid is present. Thymidylic acid bound in DNA does
not react at all in the cysteine reaction in its original form, as after the
destruction of the purine-bound sugar in the apurinic acid by 0.2 A^ alkali,
the remaining pyrimidine nucleotides do not show any appreciable color
even after hours. While the purine nucleosides react in the Stumpf mod-
ification \vith the same intensity as the corresponding nucleotides, this is
not the case with the pyrimidine nucleosides. While the molar extinction
coefficient of thymidylic acid is reported to be 26 % lower than that of the
purine nucleotides, these extinction coefficients are identical for thymidine
and the purine nucleosides. Cytidine, on the other hand, shows a much
lower color than thymidine and the maximum is reached only after 20
hours at room temperature.^^ It must be furthermore noted that the ab-
sorption spectrum of the pink color in both modifications of the cysteine
reaction is not completely identical in the case of thymidylic acid in DNA,
insofar as the absorption curve for the first one is somewhat steeper towards
the lower wavelengths than that for the second one.

According to our own experiments, the reason for this difference in the behavior of
the free and bound thymidylic acid is due to the fact that the latter in polynucleotide
linkage always reacts in the cysteine reaction like thymidine-3',5'-diphosphate. This
difference is particularly striking when the reaction is carried out according to the
original procedure with the more dilute acid. While under these circumstances the
purine nucleotides give the maximum color alreadj^ 60 minutes after the termination
of heating, it takes 24 hours for thymidylic acid to reach the maximum of color; and
thymidine diphosphate, isolated from DNA^'* or prepared synthetically, requires
more than 72 hours to reach the maximum. When the purine-bound sugar of DNA is
destroyed by alkali after being previously split from the purine by mild acid hydroly-
sis, the remaining pyrimidine nucleoside diphosphates react in the original form of
the cysteine reaction in such a way that the color developed during the first 48 hours
at room temperature corresponds to the amount of thymidine diphosphate present
in the DNA preparation. This shows that the cytidine diphosphate does not signifi-
cantly react, at least during the first 48 hours. With increasing concentrations of
thymidine diphosphate the intensity of the color increases disproportionately. Never-
theless, it is possible to determine quantitatively the concentration of thymidine
diphosphate in a DNA preparation with the cysteine reaction. To this end, the purine
sugar is destroyed by 2 minutes' heating at 100° with A' 112804 then adjusting the
neutralized hydrolysate to 0.2 A'' NaOH and again heating for 2 minutes. The cysteine
reaction is carried out on this preparation simultaneousl}' with a series of standards
prepared from a DNA of known pyrimidine composition, the concentration of the
standards to differ by no more than 10% from each other. The concentration of thymi-

'!» The author is greatly indebted to Professor A. R. Todd, University of Cambridge,
and to the Biochemical Research Foundation, California, for the preparations of
thymidine diphosphate used in these experiments.

COLOR REACTIONS OF NUCLEIC ACID COMPONENTS 293

dine diphosphate is then obtained by comparing the optical density of the unknown
with that of the standard closest to it. As the latter does not differ by more than 10%.
the deviation from proportionality can be neglected.

Neither aldopentoses or hexoses, nor short-chain sugars, glycolic al-
dehyde, trioses, erythrose, nor ahphatic aldehydes give any reaction with
the cysteine reagent, even in much higher concentrations than used for
DNA. The same is true of hexuronic acids, and amino, alpha-keto, and
alpha- and beta-hydroxy acids, which occur in significant quantities in
animal tissues. Fructose gives a yellow color; 2-deoxyhexoses and hexals a
yellow-red, and digitoxose a yellow color with completely different spectra
from that produced by 2-deoxyribose. The only compound so far investi-
gated related to saccharides which produces a cysteine reaction product
with a characteristic maximum at 490 m/u, is arabinal which, of course,
under conditions of the reaction should be partly hydrated to 2-deoxy-
pentose. Even in this case, however, the absorption curve differs con-
siderably from that of DNA insofar as it shows two more maxima— one
at 450 and one at 415 m^u. The molar extinction coefficient at 490 m/x is
only about one-third of that of DNA. Furthermore, arabinal produces a
yellow color with an absorption maximum at 470 m^t with sulfuric acid
alone in the absence of cysteine. This maximum at 470 mn disappears when
cysteine is added.

Mechanism of the reaction: — The cysteine reaction resembles the diphenyl-
amine reaction as far as its possible mechanism is concerned, insofar as,
according to our observation, furfuryl alcohol which gives the characteristic
blue color in the diphenylamine reaction, produces also a compound with
the characteristic maximum at 490 mju in the cysteine reaction. However,
while the absorption curve of the furfuryl alcohol in the first reaction was
found to be identical with that for 2-deoxyribose or DNA, this is not the
case in the cysteine reaction. Here the absorption curve shows again two
additional maxima which were found in the reaction of arabinal at 450 m/x
and 415 m^, and these two maxima are still more pronounced in the case of
furfuryl alcohol. This compound also gives the yellow color with the ab-
sorption maximum at 470 m/z with sulfuric acid alone. The molar extinction
coefficient at 490 m/i for furfuryl alcohol in both modifications of the cyste-
ine reaction, is considerably lower than that of DNA and arabinal. These
observations indicate that either furfuryl alcohol itself or a heterocyclic
derivative of it is an intermediate in the cysteine reaction of 2-deoxyribose.
It seems much less probable that w-hydroxylseviilic aldehyde or a similar
aliphatic aldehyde should be the intermediate in this reaction because
hydroxy aldehydes and keto aldehydes, in general, do not react with the
cysteine reagent while they give color reactions of varying intensities with
the .diphenylamine reaction. Furfural, on the other hand, gives a violet

294 ZACHARIAS DISCHE

color which differs from the absorption spectrum of the coloration obtained
with DNA or furfuryl alcohol. The fact that furfuryl alcohol and arabinal
produce less color than 2-deoxyribose does not necessarily contradict this
assumption about the mechanism of the reaction, as both these substances
enter side reactions which are responsible for the additional two absorption
maxima, while the intermediate in the case of 2-deoxyribose could be pro-
duced in small quantities during a certain time interval and immediately
removed by the action with cysteine without being able to enter into side
reactions. This assumption seems to be borne out by the fact that thy-
midylic acid in the original form of the cysteine reaction produces even a
higher color intensity than equivalent amounts of purine nucleotides, as
in this case the slow splitting of the glycosidic linkage should still more
slow down the formation of the intermediate. In agreement with that, the
steeper slope of the absorption curve towards lower wavelengths in the
case of thymidylic acid also indicates that in this case the side reactions
which produce the peaks at 450 and 415 mn are still more suppressed than
in the case of purine nucleotides.

Quantitative determination of DNA and its constituents by the cysteine reaction with
75 vol.% H2S0i : — The optical density at 490 m/i of the reaction mixture is proportional
to the concentration of DNA, purine nucleotides, and thymidylic acid in the range
between 25 and 500 iig. per cc. of DNA or the equivalent amount of the nucleotides in
both modifications of the cysteine reaction. It should be noted that, while the molar
extinction coefficient for DNA is lower in the original form of the reaction by about
one-third as compared with the Stumpf modification, it is possible to make determina-
tions at lower concentrations of the material as the amount of the unknown is twice
that used in the other form. As the pyrimidine nucleotides bound in the DNA do not
react at all in the reaction, any preparation of DNA can be used as standard solution
for its determination in tissues. In the determination of free thymidylic acid or the
pyrimidine nucleosides, however, standards of the substance which is to be deter-
mined are preferable. Small amounts of pure proteins and carbohydrates in concen-
trations which occur in animal tissues will not interfere with the reaction. How far
this is also true of plants and bacteria will have to be investigated.

(2) Reaction of DNA with Cysteine and Concentrated HiSd y^

Procedure: — To 1 cc. of a solution containing 50 to 500 ng. per cc. of DNA, 4 cc. of
concentrated sulfuric acid (reagent grade) is added under cooling in tap water. The
reaction mixture is left for 1 hour at room temperature to avoid air bubble formation,
and then 0.1 cc. of a 3% solution of cysteine hydrochloride is added and shaken. The
solution is left from 20 to 48 hours at room temperature, and the absorption is then
measured with the Beckman spectrophotometer.

Absorption curve and specificity of the reaction: — ^Pentoses, hexoses, hep-
toses, methylpentoses, and hexuronic acids form, under the conditions of
this reaction, furfural or the corresponding homologues which all show

" Z. Dische, /. Biol. Chem. 181, 379 (1949).

COLOR REACTIONS OF NUCLEIC ACID COMPONENTS 295

intensive absorption between 305 and 330 m/z,'^ or at 405 m^x.^^ This ab-
sorption appears immediately after the addition of the sulfuric acid to the
solution of the sugar. With DNA or its nucleotides, no such absorption
appears before addition of cysteine. Subsequently, a compound is slowly
formed which shows a sharp absorption maximum at 375 m^. The maximum
absorption is reached at room temperature only after 48 hours, but 80%
of the maximum is already obtained after 20 hours. Apurinic acid, thy-
midylic acid, arabinal, and furfuryl alcohol, according to our more recent
observation, give the same reaction and the slope of the absorption curve,
between 350 and 380 m;u, is identical in all these cases with that obtained
with DNA. The molar extinction coefficients at 375 m^u, however, are con-
siderably decreased, viz., for apurinic acid by 30%, for arabinal by 57%,
for furfuryl alcohol 53%, and by 40% for thymidylic acid. These values
indicate that the mechanism in this reaction is identical or similar to that
in the cysteine reaction of DNA with 75 vol. % H2SO4 , and that either
furfuryl alcohol itself or a furan derivative closely related to it is an inter-
mediate in this reaction. This conclusion is supported by the fact that
apurinic acid gives lower values than equivalent amounts of DNA, as it
can be assumed that in APA at least a part of the sugar is present as a
straight-chain compound. The lower values for furfuryl alcohol do not
exclude the possibility that this compound itself is an intermediate, as here
again side reactions may occur when the compound is present from the
beginning, and may be lessened or altogether voided when it is formed
slowly from another compound.

Interference from other siihstances: — As all carbohydrates give absorption
spectra in this cysteine reaction, with peaks around 400 m^u, they will all
interfere with the determination of DNA. As the reaction product of pen-
toses, however, is unstable and most of the absorption due to it disap-
pears after 24 hours' standing at room temperature,^- the interference from
small amounts of ribonucleic acid in general will not be significant. As far
as the interference from hexoses is concerned, it may be possible to elim-
inate it by dichromatic readings at 375 m/i, and a higher wave length chosen
in such a way that the absorption due to hexoses which has a maximum
at 414 m^, is equal to the absorption at 375 m/x. The difference in the op-
tical density of these two wavelengths will then be a measure of the amount
of DNA or one of its breakdown products. On the other hand, aliphatic
aldehydes and hydroxy aldehydes such as glycolic aldehyde, trioses, tet-
roses, and pyruvic aldehyde do not react at all in the general cysteine re-
action of carbohydrates and will not interfere with it. The same is true
of amino acids and pure proteins in considerable amounts, exceeding several

» Z. Dische, J. Biol. Chem. 204, 983 (1953).

296 ZACHARIAS DISCHE

times that of DNA, so that the reaction could be used to determine DNA
in nucleoproteins.

c. Reaction of DNA with Tryptophan and Perchloric Acid

This reaction, described by Cohen in 1944/^ represents a special case of
a more general reaction of carbohydrates with tryptophan in acid solution
described by Thomas.^*

Procedure: — To 1 cc. of a solution containing 100 to 500 fig. of DNA are added 0.2
cc. of 1% tryptophan solution and 1.2 cc. of 60% perchloric acid. The solution is
heated in a boiling water bath for 10 minutes, cooled in tap water, and read in the
spectrophotometer 5 minutes later. A purple color appears. The absorption spectrum
has so far only been determined with a Klett photoelectric colorimeter. The maximum
is obtained with a filter having a transmission range between 485 and 550 m/x- Accord-
ing to our own observation, an absorption maximum at 500 m// is obtained with the
Beckman spectrophotometer.

Specificity of the reaction: — Deoxyguanosine was found to react with a
molar extinction coefficient identical with that of DNA. This suggests that
the pyrimidine nucleotides of DNA react in this reaction to the same extent
as the purine nucleotides. This conclusion, however, should be further
checked by direct determinations on isolated pyrimidine nucleotides as
well as breakdown products in which the purine sugar is destroyed. Aldo-
pentoses, aldo- and ketohexoses, trioses, ascorbic acid, furfural, acetalde-
hyde, benzaldehyde, glyceraldehyde, and palmitaldehyde give color reac-
tions with the reagent which, however, differ in their absorption maximum
when determined with the Klett photoelectric colorimeter. Rhamnose
apparently showed the same maximum as DNA, while the intensity of
the color corresponded only to 4 % of that of an equal amount of DNA.

Mechanism of the reaction: — Cohen^® suggested that the colored product
is a Schiff base produced by the condensation of 2-deoxyribose with the
nitrogen of the pyrrol ring. This suggestion is based on the assumption that
2-deoxyribose, differing from other carbohydrates with a hydroxyl at carbon
2, would not be able to participate in the Amadori reaction and thus an
intermediate containing conjugated double bonds would accumulate.
Observations in our laboratory have shown that furfuryl alcohol reacts
with Cohen's reagent producing a pink color with an absorption maximum
at 500 van. The absorption curve is like that of arabinal and not identical
with that of DNA, and the molar extinction coefficient at 500 mn 40%
lower than that for DNA. This and the reactivity of hydroxy aldehydes
with the reagent suggest that here, as in the diphenylamine reaction, not
the aldehydic form of 2-deoxypentose, but an aldehydic intermediate which
can be formed from the sugar as well as from furfuryl alcohol, is responsible
for the color reaction.

" P. Thomas, Z. Physiol. Chem. 199, 10 (1931).

COLOR REACTIONS OF NUCLEIC ACID COMPONENTS 297

Quantitative determination of DN A in the presence of other substances: — The color
developed in the reaction is proportional to the concentration of DNA in the range of
0.01 to 0.05% when determined with a Klett photoelectric colorimeter. Aldehydes,
fructose and its derivatives can interfere, however, with the reaction. The protein
moiety in nucleoproteins interferes by creating a turbidity and producing unspecific
colors due to the protein constituents. The first impediment can be eliminated by
filtration; the second by extraction with isoamyl alcohol of a boiling point of 130-132°,
which does not extract the color produced by histone or protamine constituents but
extracts completely the color due to deoxyribose. The standard solution, of course,
must be treated in an identical way. Other proteins, however, may interfere and the
extraction with isoamyl alcohol may be of no avail.

d. Reaction of DNA and 2-deoxyrihose with Indole and HCl

This reaction was first described in 1929." As the original procedure
shows much lower sensitivity than other reactions, Ceriotti^® modified it by
increasing the acid concentration. This more sensitive modification, how-
ever, appears to be less specific.

Original procedure: — To 1 part of a solution containing 0.5 to 2.5 mg. per cc. of
DNA, are added an equal volume of 1% HCl and 0.1 cc. of a 1% solution of indole in
ethanol. The mixture is heated for 5 minutes in a boiling water bath. An intensive
reddish-brown color appears. The reaction mixture is thoroughly shaken with an
equal volume of chloroform, whereupon the water phase becomes jellow-brown while
reddish solid particles accumulate at the interphase.

Specificity of the reaction: — The reaction is produced only by the purine
nucleotides of DNA, as destruction of the sugar of purine nucleotides by
protracted heating in 2% sulfuric acid destroys the reactivity of DNA.
Pentoses and hexoses even at 2.5%, or ribonucleic acid at 0.25%, do not
produce any visible color in the water phase. On the other hand, glucosone
and the reaction products obtained by deamination of glucosamine with
nitrite (probably 2 , 5-anhydromannose) give the same yellowish-brown
color as DNA.

Procedure of Ceriotti for the quantitative determination of DNA: — To 2 cc. of a solu-
tion containing 2.5 to 15 ng. per cc. of DNA, are added 1 cc. of 0.04% indole C.P.
solution in distilled water and 1 cc. of concentrated HCl (sp. gr. 1.19). The test tube
is immersed for 10 minutes in a boiling water bath and cooled under running water,
the solution is extracted 3 times with 4 cc. of chloroform, and the water separated by
centrifugation from the chloroform. As the purity of the chloroform is of the utmost
importance, it should be purified bj' repeated extraction with concentrated H2SO4 ,
followed by water extraction, and then freed of water by keeping it for 48 hours over
CaCh . Finally, it is distilled to give a product with the boiling point of 61 °C. A
yellow color appears in the water phase which is read with the Beckman spectrophotom-
eter against a blank treated in an identical manner. A faint pink color appears in the
chloroform phase. The color given in the water phase is stable for several hours. On
long standing at room temperature, a pink color forms which can be again removed

" Z. Dische, Biochem. Z. 204, 431 (1929).
" A. Ceriotti, J. Biol. Chem. 198, 297 (1952).

298 ZACHARIAS DISCHE

by extraction with chloroform. The absorption curve has a sharp peak at 490 m/x and a
second small but constant peak at 460 m/i.

Specificity of the reaction according to Ceriotti.'—Apurimc as well as thy-
midylic acid according to our observations produce the same color as DNA;
the molar extinction coefficient of thymidylic acid, however, is about 20%
that of DNA. Free 2-deoxyribose gives a much lower color intensity than
DNA. The same is true of arabinal. Glucosamine, laevulic and uric acids,
as well as creatine and ascorbic acid, do not give any color. Hexoses, on the
other hand, give a pink color which is completely extracted by chloroform.
Arabinose, and probably also other pentoses as well as RNA, however,
produce a brown color in the water phase which corresponds to 8 % of the
color produced by an equivalent amount of DNA. Other pentoses and RNA
were not investigated but probably behave similarly. Galacturonic acid
also produces a considerable yellow-brown color in the water phase. These
two types of substances, therefore, may interfere with the determination
of DNA, if present in excess of DNA, and must be accounted for after
determination of their concentration in the unknown.

Mechanism of the reaction: — Not only tryptophan but other /3-substituted
derivatives of indole give color reactions when heated with HCIO4 and
DNA. This, and the fact that hydroxy and keto aldehydes produce more
or less intensive colors^® in both modifications of the indole reaction as well
as in the tryptophan reaction, suggests that the first reaction resembles
the second one in its mechanism. Furfuryl alcohol, however, does not pro-
duce any color in the water phase in the Ceriotti modification, and a very
faint one only in the original form of the indole reaction. This, like the weak
reaction of free deoxyribose, may be due to a rapid destruction of these
compounds by HCl at 100°.

e. Reaction with Carhazole and Sulfuric Acid.

This reaction, described in 1930^ is best carried out according to the
modification of Gurin and Hood.^^ DNA shows a purple color with an ab-
sorption maximum at 530 m/x. The maximum of the color is already reached
after 2 minutes at 100°.

Specificity of the reaction: — A characteristic feature of this reaction is
that after destruction of the sugar of the purine nucleotides by heating for
2 hours wuth 2 % H2SO4 , the remaining pyrimidine polynucleotides react
with about the same intensity as the whole DNA molecule, in spite of the
fact that the free purine nucleotides also react with great intensity.-^ This
indicates a considerable influence of the linkages between the nucleotides
on the intensity of the reaction. Free thymidylic acid reacts about 10%

" S. Gurin and D. B. Hood, J. Biol. Chem. 139, 775 (1941).

28 H. Angermann and F. Bielschowsky, Z. -phijsiol. Chem. 191, 123 (1930).

COLOR REACTIONS OF NUCLEIC ACID COMPONENTS 299

less than half the equivalent amount of DNA. The reaction is not specific
as practically all saccharides give purple colors with slightly different ab-
sorption maxima. Aliphatic aldehydes and hydroxy aldehydes also give
intensively colored products.

/. Reaction of DNA with Schiff's Reagent

The presence of considerable amounts of the aldehydic form of 2-deoxy-
ribose in its solutions manifests itself by its ability to produce a characteris-
tic red color with Schiff's reagent. Feulgen-^ found in 1924 that after a
brief hydrolysis with 1 N HCl, DNA reacts with Schiff's reagent. Tobie^"
introduced a modified reagent characterized by a much higher concentration
of SO2 which increases considerably the sensitivity of the reaction. Wid-
strom^' adapted the reaction with Schiff's reagent for quantitative pur-
poses. While the reaction proved of great importance for cyto- and histo-
chemical studies on nucleic acids [Cf. Swift, Chapter 17], it cannot compare
in sensitivity with the other methods here described, as far as quantitative
determinations are concerned.

g. Evaluation of Various Methods of Quantitative Determination of DNA
and Its Constituents

Of all the reactions here described, that with cysteine and 75 vol. %
sulfuric acid appears to be the most specific, as even a compound as closely
related to 2-deoxyribose as arabinal, produces a color which can be dis-
tinguished by its absorption spectrum from that of 2-deoxyribose. It will
be of particular interest to see whether 3-deoxypentose and 2,3-deoxy-
pentose which, according to Overend,^ show a weak blue color in the di-
phenylamine reagent, react at all with the cysteine reagent. For quanti-
tative determination, however, the difference in specificity between this
reaction and that with diphenylamine appears negligible and the sensitivity
of the latter reaction is about twice that obtained with the modification of
Stumpf of the cysteine reaction with 75 vol. % H2SO4 . The cysteine reaction
of DNA, furthermore, may prove valuable for the determination of free
thymidylic acid in the presence of other deoxynucleotides or DNA, par-
ticularly in its free form. However, more investigations appear necessary
to test the application of the cysteine-H2S()4 reaction for determinations
in living tissues. The tryptophan-perchloric acid reaction, when measured
with the Beckman spectrophotometer, appears to have about one-third of
the sensitivity of the diphenylamine reaction. Its use for quantitative pur-
poses in living cells has not sufficiently been studied, and the data so far

" R. Feulgen and H. Rosenbeck, Z. physiol. Chcrn. 135, 203 (1926).
30 W. C. Tobie, Ind. Eng. Chem., Anal. Ed. 15, 405 (1942).
3' G. Widstrom, Biochem Z. 199, 298 (1928).

300 ZACHARIAS DISCHE

published suggest that the chromogen is more labile than that of the di-
phenylamine reaction, and that the presence of even small amounts of
proteins makes the determination more laborious. As in this reaction the
pyrimidine nucleotides seem to react as well as purine nucleotides, it may
prove useful for the detection of breakdown products of DNA. The car-
bazole reaction could be used for the same purpose. Although the latter is
less specific than the tryptophan reaction and is strongly interfered with by
glucose, it is less affected by proteins. The reaction with indole and HCl in
Ceriotti's modification appears of interest, as it makes possible quantitative
determinations on about 10 times small amounts of material than all the
other methods. The use of this method in combination with special micro-
techniques like that of Steele et al}^ may make it possible to determine
DNA in quantities below 1.6 jug. per. cc. The lesser specificity, however, of
this reaction and the lack of sufficient data about its application for the
quantitative determination of nucleic acids in tissues and of the effect of
proteins on the reaction suggest caution in its use. The availability of sev-
eral color reactions of 2-deoxy pentose which differ in their mechanism,
specificity, and in the degree of interference from other substances is of
particular advantage, as far as detection and identification of DNA and
its derivatives in living cells is concerned. The uncertainties inherent in the
use of color reactions for this purpose are largely eliminated by checking,
on a quantitative basis, the results of one of these reactions by those ob-
tained from several other ones.'^*

2. Reaction of Ribonucleic Acids

All reactions of ribonucleic acids based on the sugar component, so far
published, are general reactions of pentoses. Three such reactions were
tested for quantitative determinations of PNA.

a. Orcinol Reaction

This reaction was recommended in two different forms — ^one, using FeCls
as catalyst, by Bial,^^ and one using CuCl2*^ which was employed for the
determinations of ribonucleic acids and its nucleotides by Massart and
Hoste.^* The Bial reaction, first used by Embden and Lenhartz'^ for white
light colorimetry of free pentose in 1924, was adapted for quantitative

"» The reaction of DNA with phloroglucinol described by H. v. Euler and L. Hahn
(Arch, neerl. physiol. 28, 398, 1946) appears to be 20 times less sensitive than the
diphenylamine reaction and does not seem to offer any advantage as compared
with other reactions of DNA.

'2 M. Bial, Deut. med. Wochschr. 29, 253; 29, 477 (1903).

" H. Barrenscheen and A. Peham, Z. -physiol. Chem. 272, 81 (1942).

^* L. Massart and J. Hoste, Biochim. et Biophys. Acta 1, 83 (1947).

36 G. Embden and E. Lenhartz, Z. physiol. Chem. 201, 149 (1931).

COLOR REACTIONS OF NUCLEIC ACID COMPONENTS 301

determinations and differentiation of various nucleotides by spectrophoto-
metric measurements in 1937.'^ Several modifications^'' of this reaction were
later developed which differ in the ratio between the solution and the re-
agent, and the composition of the latter, as regards the concentration of
HCl, FeCls , and orcinol. The original form of the reaction, as proposed in
1937, uses a higher HCl concentration than all later modifications and is,
therefore, the most sensitive of all. The most widely used modification
seems to be that proposed by Mejbaum^* in 1939 which uses 6 N HCl.

Quantitative determination of PNA according to Dische and Schwartz:^^ — To 1.5 cc.
of a solution of ribonucleic acid is added 3 cc. of the reagent which is prepared by dis-
solving 100 mg. FeClreHjG in 100 cc. of HCl, sp. gr. 1.19, and adding 3.5 cc. of a 6%
solution of orcinol (twice recrystallized from benzene) in ethanol. The reaction mix-
ture is heated in a water bath for 3 minutes and cooled in tap water. A standard of
PNA and a blank containing water instead of the unknown is run simultaneously.
The optical density is measured at 665 m^ with the Beckman spectrophotometer
against a blank containing water and the reagent. A solution of PNA containing 40
Atg. per cc. shows an optical density of 0.18 which does not differ significantly from the
density of a solution of 20 Mg- per cc. of adenosine-3-phosphate. The optical density is
proportional to the concentration of PNA in the range between 10 and 100 /ig. per cc.
Quantitative determination of PNA according to Mejbaum:^^ — To 1 part of the un-
known is added an equal volume of concentrated HCl, sp. gr. 1.19, containing 0.1%
FeCl3-6H20 and 0.1% of orcinol. The mixture is heated for 20 minutes, or, according
to Albaum and Umbreit,^' for 45 minutes. A standard solution of PNA and a water
blank are run simultaneously and the optical density of the solution is measured at
the absorption maximum, which with the Beckman spectrophotometer lies at 670 m/x

Specificity of the reaction: — In all the modifications of the Bial reaction
only the purine nucleotides and nucleosides react significantly. A more
involved procedure for the determination of pyrimidine nucleotides in
which bromination and prolonged heating with acid at 105° precedes the
performance of the Bial reaction, was developed by Massart and Hoste.^^
Rial's orcinol reaction has a rather low specificity as not only pentoses,
but also 2-deoxyribose and DNA, methylpentose, and hexuronic acids give
a green color with an absorption maximum around 670 mju. Certain aldo-
heptoses which can occur in bacterial polysaccharides also produce a green
color with an absorption maximum around 655 m^.'^^ All these substances
can occur in tissue extracts as they are prepared for determination of PNA.

Interference from other substances: — Aldo- and ketohexoses in free form
and in polysaccharides produce, in all modifications of Dial's orcinol re-
action, a reddish-brown color which can se^•erely interfere with the reaction

^* Z. Dische and K. Schwarz, Mikrochim. Acta 2, 13 (1937).

" G. L. Miller, R. H. Colder, and E. E. Miller, Anal. Chem. 23, 903 (1951).

38 W. Mejbaum, Z. physiol. Chem. 258, 117 (1939).

■39 H. G. Albaum and W. W. Umbreit, J. Biol. Chem. 167, 369 (1947).

" Z. Dische, J. Biol. Chem. 204, 983 (1953).

302 ZACHARIAS DISCHE

of PNA. Proteins and certain lipidic cell constituents can also interfere by
forming red products.

Quantitative determination of PNA in presence of aldohexose: — When the amount
of glucose or its polymers reaches a certain level, the absorption at 670 mju due to the
brown reaction products can increase the total reading. If the amount of glucose is
not too high, it is possible to eliminate this discrepancy by dichromatic readings.
With Mejbaum's reagent, the procedure recommended by Brown^^ and by Drury"
is used. This consists in the simultaneous determination of pentose and glucose from
two equations obtained from two measurements of optical densities at two different
wavelengths. Using the procedure of Dische and Schwartz, after the determination
of optical density at 665 niju, the optical density is determined at the wavelength
around 565 mju, at which the optical density of a glucose standard run simultaneously
with the unknown shows the same optical density at 665 m^." The difference of De^b —
Dses of the experimental sample divided by this difference in the internal standard is
then a measure of the concentration of PNA. This difference is only about 25% lower
than Dees so that this procedure does not involve a large decrease in sensitivity of the
reaction. It is not possible, however, to use such a procedure in the presence of keto-
hexoses or ketoheptoses.

h. Phloroglucinol reaction of Euler and Hahn**

Procedure: — To 1 cc. of the unknown containing 2 mg. PNA is added 8 cc. of a
0.1% solution of FeCls in a mixture of 1 part of concentrated HCl and 6 parts of glacial
acetic acid. After stirring, the tubes are immersed for 50 minutes in a boiling water
bath and cooled to room temperature. 1 cc. of a 25% phloroglucinol solution in a mix-
ture of 1 part concentrated HCl, 1 part H2O, and 2 parts of glacial acetic acid is then
added and kept for 20 minutes at room temperature. The tubes with the reaction mix-
ture are then immersed in a boiling water bath for exactly 4 minutes, cooled to room
temperature, and so kept from 2 to 24 hours. The maximum of the color intensity ap-
pears after 10 hours. The absorption maximum is at 680 m^i read with the Beckman
spectrophotometer.

Specificity of the reaction: — ^DNA does not give any color with this pro-
cedure; this, however, could be due to the prolonged heating before the
addition of phloroglucinol which, of course, destroys the sugar of the purine
nucleotides of DNA. No data were reported as to the reactivity of other
sugars and aldehydes.

c. Reaction of Pentose with Cysteine and HiSOi

PNA, its nucleotides and nucleosides all react in the general reaction of
carbohydrates with cysteine and sulfuric acid. The procedure was de-
scribed above under I.l b. The readings have to be carried out about 15

« A. H. Brown, Arch. Biochem. 11, 269 (1946).
« H. F. Drury, Arch. Biochem. 19, 455 (1948).
*^ Z. Dische, G. Ehrlich, C. Munoz, and L. von Sallmann, A7n. J. Opthalmol. 36, 54

(1953).
** H. V. Euler and L. Hahn, Svensk Kem. Tidskr. 58, 251 (1946); Chem. Abstr. 41,

2108 (1947).

COLOR REACTIONS OF NUCLEIC ACID COMPONENTS 303

minutes after addition of cysteine when the maximum absorption produced
by pentoses is observed. The peak of the absorption curve is at 390 mn.
Reaction of nucleotides and nucleosides of PNA : — ^Only the purine nucleo-
tides of PNA react significantly in the cysteine reaction. The molar ex-
tinction coefficient of adenosine-3-phosphate does not differ significantly
from that of pure ribose, and is twice as great as that of PNA.

QuaJititalive determination of PNA and its constituents in presence of DNA and
hexoses: — The optical density in the cysteine reaction is proportional to the concen-
tration of pentose in the range of concentration between 2 and 40 /xg- per cc. of ribose.
The density decreases slowly after maximum absorption is reached, and for this
reason the readings of the optical density of the unknown must be carried out simul-
taneously with the readings of the standard solution. The purest preparation of DNA
showed an optical density at 390 m^, which corresponded to no more than 1.5% of
that of equivalent amounts of PNA. As this ratio was obtained for several prepera-
tions of DNA, it is probably not due to contamination of the prepartion with
small amounts of PNA. Aldohexoses produce with the cysteine reaction a
yellow color with an absorption maximum at 412 m/i- This color interferes with the
determination of the pentose. This interference, however, can be eliminated by di-
chromatic readings. The latter are carried out by finding, around 424 m/x, a wavelength
at which a standard of glucose or another aldohexose shows the same optical density
as at 390 m/n. D390 — Dm, therefore, is zero for the hexose. It is only slightly lower
than Z)39o for the pentose. In this way adenosine -3-phosphate can be still determined
in presence of a 5-fold amount of glucose or a 10-fold amount of mannose or galactose.
£^390 — D424 , however, is not zero for ketohexoses. Hexuronic acids show an identi-
cal absorption curve as pentose, although the molar extinction coefficient is far
smaller. When present in large quantities, however, they will interfere with the re-
action, and the same is true of methylpentose. It is therefore necessary to correct for
the presence of these two classes of saccharides as in the case of the orcinol reaction.

Evaluation of the PNA reactions: — -Of the three reactions, that with Bial's
reagent appears most sensitive and the only one sufficiently investigated as
far as the use for quantitative determinations in tissues is concerned. The
procedure according to Dische and Schwarz is more sensitive than the other
modifications and therefore allows a shorter heating time. This in turn
decreases the interference from other sugars. As regards interference from
proteins, the orcinol reaction appears to be more influenced than the cyste-
ine-H2S04 reaction. The latter reaction is also more suitable for the de-
tection of small amounts of PNA in DNA. The reaction with phloroglucinol
of Euler and Hahn is much less sensitive, appears more laborious, and has
not been sufficiently investigated as far as interference from other sub-
stances is concerned.

II. Determination of Purine and Pyrimidine Bases of
Nucleic Acids by Characteristic Color Reactions

The discovery that the molar ratios of individual purines and pyrimidines
in nucleic acid varies from one preparation of nucleic acid to another de-

304 ZACHARIAS DISCHE

pending on the species from which it was obtained [Cf. Chapters 10, 11],
rendered the determination of the content of these bases by simple and
sensitive reactions a matter of considerable interest. In the last few years
such methods were developed for adenine, thymine, cytosine, and uracil,
which are fairly characteristic for these substances and could perhaps be
applied for the analysis of nucleic acids. A micromethod for the deter-
mination of guanine in urine has not yet been applied to the determination
of this purine in nucleic acids and will not be discussed here. Unfortunately,
the preparations used by various authors for testing the applicability of
these methods for the determination of the bases in nucleic acids were not
completely pure and it is difficult, therefore, at present to evaluate how
far these methods could give completely correct values for the content of
individual purines and pyrimidines in nucleic acids and how far they can be
applied for the determination of these substances in tissue extracts without
interference from other cell constituents.

1. Determination of Adenine

The procedure of Woodhouse*^ is based on a color reaction of adenine
after its reduction with zinc dust and diazotization with NaN02 with N-
1-naphthylethylenediamine hydrochloride. The red color is still visible at a
concentration of 5 ng. per cc. of adenine and. is proportional to its concen-
tration up to 40 fxg. per cc. The determination of adenine in adenosine and
adenosine-3-phosphate gave correct values. In two preparations of nucleic
acids, the P content of which was about 20% below the theoretical, the
values were 10% lower than corresponded to the content in P. The reaction
is not specific for adenine, as it was originally proposed for the determina-
tion of folic acid.

2. Determination of Thymine

Two procedures by Woodhouse and by Pircio and Cerecedo for the de-
termination of thymine are both based on the Hunter'*^ reaction for this
substance. Thymine is coupled first with sulfanilic acid in Na2C03 solution,
and the reaction product then treated with hydroxy lamine in NaOH. A
red color is produced which is still visible at a concentration of 10 ^g- per
cc. of thymine. Cytosine and uracil do not interfere, but purines must be
removed either as silver salt or by palladium chloride. Woodhouse^^ pre-
cipitates thymine as a silver salt before its determination. Pircio and Cere-
cedo^* extract the evaporated hydrolysates of nucleotides or nucleic acids

" D. L. Woodhouse, Arch. Biochem. 25, 347 (1950).

^« G. Hunter, Biochem. J. 30, 745 (1936).

" D. L. Woodhouse, Biochem. J. 44, 185 (1949).

« A. Pircio and L. R. Cerecedo, Arch. Biochem. 26, 209 (1950).

COLOR REACTIONS OF NUCLEIC ACID COMPONENTS 305

with ether to remove the laevuhc acid produced from the sugar. The re-
action can be obtained with solutions containing only 10 ng. per cc. of
thymine. The involved procedure for the purification of the base requires
much higher quantities when thymine is determined in DNA or PNA. The
values for thymine obtained on preparations of nucleic acids are in the range
found by other authors with these methods. The reaction of Hunter is by
no means characteristic for thymine, but is given by carbonyl compounds
which are capable of enolization in Na2C03 solutions (such as acetone).

3. Determination of Cytosine and Uracil

The method of Soodak et a/.** is based on the reduction of arsenotungstic
acid (uric acid reagent) by brominated cytosine and uracil. Solutions con-
taining no more than 5 jug. per cc. of cytosine can still give a clearly visible
color in mixtures of the two pyrimidines; both of them can be determined
after quantitative removal of cytosine by a zeolite, Decalgo. Thymine and
purines do not interfere.

^3 M. Soodak, A. Pircio, and L. R. Cerecedo, /. Biol. Chem. 181, 713 (1949).

CHAPTER 10

Isolation and Composition of the Deoxypentose Nucleic Acids
and of the Corresponding Nucleoproteins

ERWIN CHARGAFF

Page

I. Introductory Remarks 308

II. Deoxypentose Nucleoproteins 309

1. General 309

2. Classification 311

3. Isolation 312

a. Extraction with Solutions of Low Ionic Strength 313

(1) Preparation of Calf Thymus Nucleohistone 313

(2) Preparation of the Nucleoprotein of Avian Tubercle Bacilli . . 314

b. Extraction with Strong Salt Solution 315

(1) Preparation of the Nucleoprotamine of Trout Sperm 316

4. Properties 317

a. Some Chemical and Physical Characteristics 317

b. Cleavage and Degradation 319

c. Artifacts 319

III. Isolation of Deoxypentose Nucleic Acids 321

1. General 321

2. Preparative Procedures 323

a. Extraction with Strong Salt Solution, Deproteinization with Chloro-
form 323

(1) Sodium Deoxyribonucleate of Calf Thymus 323

(2) Sodium Deoxyribonucleate of Yeast 324

b. Extraction with Strong Salt Solution, Deproteinization by Satura-
tion with Sodium Chloride 325

(1) Sodium Deoxyribonucleate of Calf Thymus 325

c. Extraction with Water 326

(1) Sodium Deoxyribonucleate of Calf Thymus 326

d. Extraction with the Aid of Anionic Detergents 327

(1) Sodium Deoxyribonucleate of Calf Thymus 327

e. Comparison of Different Isolation Procedures 328

f. Miscellaneous Procedures and Applications 329

g. Nucleic Acids of Microorganisms and Viruses 330

h. Removal of Pentose Nucleic Acid and of Other Impurities .... 332

IV. Properties of Deoxj'pentose Nucleic Acids 333

1. Elementary Composition and Standards of Integrity 333

2. Denaturation and Degradation 337

V. Some Partial Degradation Products 340

1. General 340

2. Preferential Removal of Purines 341

307

308 ERWIN CHARGAFF

a. Thymic Acid 341

b. Apurinic Acid 341

(1) Preparation 341

(2) Properties 342

3. "Cores" (Limit Polynucleotides) 345

VI. Constituents of Deoxypentose Nucleic Acids 345

1. Sugar 346

2. Nitrogenous Constituents 346

3. Unidentified Constituents 347

VII. Composition of Deoxypentose Nucleic Acids 348

1. General 348

2. Procedures 350

3. Distribution of Purines and Pyrimidines 350

a. Grouping of Deoxypentose Nucleic Acids and Presentation of Results 350

b. Composition Differences and Similarities 351

VIII. Fractionation of Deoxypentose Nucleic Acids 358

1. General 358

2. Fractional Dissociation of Nucleohistone or Protein Nucleates ... 361
IX. Composition Studies and Structural Investigations 366

X. Correlations and Concluding Remarks 368

1. Simplifying Generalizations 368

2. Unifying Generalizations 369

3. A Concluding Remark 371

I. Introductory Remarks

The early history of the discovery of the deoxypentose nucleic acids
(DNA) has been mentioned in Chapter 1 of this book. A much lengthier,
and in part pleasantly autobiographical, treatment will be found in Levene's
admirable monograph;^ and the later stages were touched upon, though in
considerably less detail, by several authors.^"* The fascinating correspond-
ence which forms part of the collected papers of F. Miescher^ must, how-
ever, be specially mentioned in a review of this subject.

An account — not practicable in the present framework — of the evolu-
tionary sequence of our understanding of the nucleic acids would not be
uninstructive: it would show how often in the natural sciences, even behind
seemingly trivial experiments, there lies an entire vision of nature; it would
demonstrate that the answers which the investigator receives often are
contained in the questions he asks. As long as, in the field under considera-
tion here, feeble and faltering questions were asked, the answers were vague;

>P. A. Levene and L. W. Bass, "Nucleic Acids." The Chemical Catalog Company,

New York, 1931.
2 J. M. Gulland, G. R. Barker, and D. O. Jordan, Ann. Rev. Biochem. 14, 175 (1945).
' E. Chargaff and E. Vischer, Ann. Rev. Biochem. 17, 201 (1948).
^ E. Chargaff, Experientia 6, 201 (1950).
* "Die histochemischen und ph^'siologischen Arbeiten von Friedrich Miescher.

Gesammelt und herausgegeben von seinen Freunden," 2 Vols. F. C. W. Vogel,

Leipzig, 1897.

ISOLATION AND COMPOSITION OF DEOXYPENTOSE NUCLEIC ACIDS 309

but when the direction of the quest became reasonably well defined, the
results sharpened in definition. Though we still are very far from a solution
of the many problems that keep multiplying as our knowledge progresses, it
would be hazardous to erect a sign saying "Ignorabimus." But much is left
to future generations.

The deoxypentose nucleic acids have often been isolated by way of the
nucleoproteins. That the use of these conjugated proteins also has provided
an encouraging procedure for nucleic acid fractionation will be mentioned
later. In other cases direct isolation methods for the preparation of the
nucleic acids from a variety of tissues have been employed. These different
procedures will be described after a brief consideration of the nucleopro-
teins themselves. These sections will in turn be followed by a discussion of
the properties and of the composition, in both its qualitative and quantita-
tive aspects, of the deoxypentose nucleic acids. Their fractionation and
methods for their structural investigation will betaken up next; and, finally,
a provisional summation will be attempted.

II. Deoxypentose Nucleoproteins

1. General

Beneath the simple textbook definition of a nucleoprotein^ — a combina-
tion between a protein and a nucleic acid — there lies an ocean of uncertain-
ties. Studies on nucleoproteins lead right into one of the most neglected,
because most difficult, fields of present-day biochemistry, namely, the con-
jugated proteins. The deoxynucleoproteins are usually defined as conjugated
proteins in which the union between the deoxypentose nucleic acid, func-
tioning as a prosthetic group, and the protein is mediated by electrostatic
attraction or by secondary valence forces.®"* It is, however, as was pointed
out some years ago in a discussion of the cognate problem of the lipopro-
teins,^ extremely difficult to distinguish between these two types of combina-
tion when dealing with macromolecules. Moreover, a decision will have to
be made in every case whether what has been isolated really preexisted in
the cell as a conjugated protein or whether it was produced by the com-
bination between solutes fortuitously present in the same cell extract, thus
simulating a definite compound. One of the attributes of a conjugated pro-
tein is that it must differ in some of its properties from a mere mixture of
its components. A genuine nucleoprotein would, therefore, have to be re-
garded as a geometrically unique compound between two giant polyampho-

* J. P. Greenstein, Advances in Protein Chem. 1, 209 (1944).

^ E. Chargaff, Cold Spring Harbor Symposia Quant. Biol. 12, 28 (1947).

8 E. Chargaff, in "Some Conjugated Proteins," p. 36. Rutgers University Press,

• New Brunswick, New Jersey, 1953.

' E. Chargaff, Advances in Protein Chem. 1, 1 (1944).

310 ERWIN CHARGAFF

lytes; and a most extensive comparison of the properties of artifacts and
of the supposedly genuine nucleoprotein would have to precede a decision.
Such comparisons, extremely difficult in the absence of suitable biological
tests, have not yet been carried out to any extent, though a beginning has
recently been made by Crampton et al}°

If, as will be discussed later, the deoxypentose nucleic acid fraction of a
given nucleus is cofnposed of a large number of differently constituted indi-
viduals, a nucleoprotamine or a nucleohistone preparation must be assumed
to comprise many different conjugated proteins; a number that will be
even greater if not only different nucleic acids, but also different proteins
enter into partnership. The reassociation of the various nucleic acid and
protein moieties that may take place, if dissociation has occurred in the
course of isolation, will introduce additional complications. In this field,
simplification may relieve the mind, but will not ease the burden, of the
investigator. In any event, an inspection of the literature will lead to the
conclusion that very few, if any, entities that can be considered as genuine
deoxypentose nucleoproteins have been isolated. As is also true of the cor-
responding ribonucleoproteins (see Chapter 11) only those structures that
are characterized by a high degree of complexity, bordering on organization,
and by specific biological activity, the viruses and phages, can safely be so
designated.

Whether the major part, or all, of the deoxypentose nucleic acids occurs
in the cell in the form of nucleoproteins cannot yet be stated. But there is
little doubt that when nucleic acids and proteins are isolated together, i.e.,
in cell extracts, the deoxypentose nucleic acid can be freed of protein only
in the presence of a high salt concentration. ^^ Under these conditions a dis-
sociation of the conjugated protein may have taken place ;'^ and a nucleo-
protein recovered after the exposure of its solution to a high electrolyte con-
centration may represent an artifact, a protein nucleate.' A comparison of
several properties of nucleohistone preparations isolated in the absence of
electrolytes with those of specimens that had been in contact with salt re-
vealed a number of differences, but they were of a relatively minor nature.^"
The entire field of conjugated proteins suffers from our incomplete knowl-
edge of the conditions governing interactions between polyelectrolytes and
of the geometry of the resulting products.

Before discussing, in the following sections, several representative nucleo-
proteins, especially those that have served for the isolation of deoxypentose
nucleic acids, reference should be made to the reviews on nucleoproteins
written by Greenstein* and Markham and Smith. '^

'» C. F. Crampton, R. Lipshitz, and E. Chargaff, J. Biol. Chem. 206, 499 (1954).
" I. Bang, Beitr. chem. Physiol, u. Path. 4, 331 (1904) ; 5, 317 (1904).
'2 R. Markham and J. D.Smith, tn "The Proteins" (H. Neurath and K. Bailey, eds.),
Vol. II, p. 1. Academic Press, New York, 1954.

isolation and composition of deoxypentose nucleic acids 311

2. Classification

As in the case of the lipoproteins,^ the classification of conjugated pro-
teins in general must be based on the nature of both the prosthetic group
and the protein. The former division is implicit in the use of the terms
deoxypentose and pentose nucleoproteins. Beyond this it is not yet possible
to go, since in no instance, not even in that of the viruses, can the presence
of a single nucleic acid individual, homogeneous with respect to both struc-
ture and function, be affirmed.

As regards the protein moiety, a crude classification would distinguish
three principal groups: (a) The nucleoprotamines, discovered by Mie-
scher*'' in ripe sperm nuclei of the salmon and occurring in ripe sperma-
tozoa of many fish genera, (b) The nucleohistones, observed by KosseP'' in
bird erythrocytes and investigated in greater detail in the form of the pro-
totypal thymus nucleohistone by Lilienfeld,'* Huiskamp,^* and Bang,^^
though here again the first observations are due to Miescher.^ The extent
of occurrence of nucleohistones in the nuclei of mammalian cells is not yet
clearly defined, nor is it possible to draw an entirely satisfactory demarca-
tion line between these two types of nucleoprotein. The spermatozoa of sea
urchins and mollusks appear to contain basic proteins of a more complicated
composition than is usually encountered in the protamines;" rooster sperm
seems to contain a protamine.^* ^^ The most comprehensive monograph on
the basic proteins still is that of Kossel;^" a shorter modern treatment has
been given by Felix ;^^ the amino acid composition has been discussed by
Tristram.^^

The third, and in some respects perhaps the most interesting, group are
(c) the nucleoproteins in the proper sense of this term. In this type of com-
pound the deoxypentose nucleic acid is bound, most likely by secondary
valence forces, to a protein lacking the basic properties of the protamines
and histones. A nucleoprotein of this type has been isolated from avian
tubercle bacilli.^^ It is in contrast to the nucleoprotamines and nucleohis-

" F. Miescher, Hoppe-Seylers Med.-chtm. Untersuchungen 4, 441 (1871).

1* A. Kossel, Z. physiol. Chem. 8, 511 (1884).

'5 L. Lilienfeld, Z. physiol. Chem. 18, 473 (1894).

'« VV. Huiskamp, Z. physiol. Chem. 32, 145 (1901).

" T. Hultin and R. Heme, Arkiv Kemi, Mineral. Geol. 26A, No. 20 (1948).

'« M. M. Daly, A. E. Mirsky, and H. Ris, J. Gen. Physiol. 34, 439 (1951).

'9 H. Fischer and L. Kreuzer, Z. physiol. Chem. 293, 176 (1953).

^^ A. Kossel, "The Protamines and Histones." Longmans, Green and Co., London

and New York, 1928.
2' K. Feli.x, in "Physiologische Chemie" (B. Flaschentrager and E. Lehnartz, eds.),

Vol. I, p. 709. Springer, Berlin, Gottingen, Heidelberg, 1951.
" G. R. Tristram, in "The Proteins" (H. Neurath and K. Bailey, eds.), Vol. I, p.

181. Academic Press, New York, 1953.
" E. Chargaff and H. F. Saidel, /. Biol. Chem. 177, 417 (1949).

312 ERWIN CHARGAFF

tones soluble in isotonic salt solutions. Whether these nucleoproteins occur
only in -microorganisms cannot yet be stated. A histone has been claimed
to be present in the nucleoprotein fraction of Type III pneumococci.^^ It
is not impossible that the protein moiety of these nucleoproteins has a more
complicated amino acid composition than have the basic proteins of low
molecular weight. Moreover, the existence of similar nucleoproteins in the
nuclei of higher organisms cannot be excluded, although complications
arising from the contamination of the preparations with cytoplasmic ma-
terial may account for some of the results. While nucleoprotamine appears
to constitute almost the entire mass of defatted fish sperm nuclei,^^"" many
other varieties of nuclei include tryptophan-containing proteins.^'* -^^ ■" The
nucleoproteins of some of the animal and bacterial viruses possibly also
belong to this class; but, the chemical information available at present is
not sufficient for a decision.

It is quite obvious that we are very far from a meaningful classification
of the types of deoxypentose nucleoprotein occurring in cell nuclei, espe-
cially with regard to changes in the composition of the protein moiety
taking place during development and in the several phases of the mitotic
cycle.

3. Isolation

Two principal methods, both foreshadowed in the pioneering work of
Miescher,* have been used for the preparation of nucleoprotamines and of
nucleohistones, i.e., complexes in which the protein partner carries a posi-
tive charge. They are based on (a) extraction with solutions of low ionic
strength j^*'^*'^" (b) extraction with strong salt solutions.^* "'^ In recent
years, the studies of Hammarsten^" and of Mirsky and PoUister^^ were of
particular importance in providing the impulse for the isolation of a large
number of highly polymerized preparations of deoxypentose nucleic acid.
As already has been mentioned, certain bacterial nucleoproteins, which
appear to belong to an entirely different type of nucleoprotein, can be ex-
tracted with dilute buffers and fractionated by conventional means.^^

24 A. E. Mirsky and A. W. Pollister, J. Gen. Physiol. 30, 117 (1946).
26 A. W. Pollister and A. E. Mirsky, /. Gen. Physiol. 30, 101 (1946).

26 K. Felix, H. Fischer, A. Krekels, and R. Mohr, Z. physiol. Chem. 287, 224 (1951).

27 K. Felix, H. Fischer, A. Krekels, and R. Mohr, Z. physiol. Chem. 289, 10 (1951).

28 E. Stedman and E. Stedman, Cold Spring Harbor Symposia Quant. Biol. 12, 224
(1947).

29 A. E. Mirsky and H. Ris, J. Gen. Physiol. 31, 1, 7 (1947).

30 E. Hammarsten, Biochem. Z. 144, 383 (1924).

3' A. E. Mirsky and A. W. Pollister, Proc. Natl. Acad. Sci. U. S. 28, 344 (1942).

ISOLATION AND COMPOSITION OF DEOXYPENTOSE NUCLEIC ACIDS 313

a. Extraction with Solutions of Low Ionic Strength

This procedure, which has been used repeatedly for the isolation of
nucleoproteins,'"^-"^^ in particular of calf thymus nucleohistone, has the
advantage of avoiding the exposure of the conjugated protein to high salt
concentrations, and therefore to dissociating conditions, in the course of its
preparation. There exists, however, the danger of a partial enzymic degrada-
tion of the nucleic acid brought about by the release of nucleases. Attempts
are usually made, based on the behavior of the pancreatic deoxyribonuclease
(see Chapter 15), to suppress the enzymic attack by the use of arsenate,
citrate, or such chelating agents as sodium ethylenediamine tetraacetate.
There is, however, some doubt as to the role of these complex-forming
agents, owing to the existence of nucleases that differ in their requirements
from the pancreatic enzyme,""" and their efficacy may have to be explained
in a different manner.'*

A preparation, at a low electrolyte concentration, of calf thymus nucleo-
histone is given as the first example, and the nucleoprotein of tubercle
bacilli as the second.

(1) Prefaration of Calf Thymus Nucleohistone.^" Trimmed calf thymus was ob-
tained fresh at the slaughter-house, chilled immediately, and processed without de-
lay. All subsequent operations were performed at 4-6°. Fifty-gram portions of tissue
were triturated for 30 seconds in a high-speed mixer equipped with cutting blades
with 50 cc. of an ice-cold mixture of aqueous 0.1 M NaCl and 0.05 M sodium citrate
(previously adjusted to pH 7). The supernatant fluid resulting from centrifugation at
2000 X g for 30 minutes was discarded and the suspension of the sediment in 100 cc.
of saline-citrate once more centrifuged. The sediment was washed three times by
thorough resuspension and centrifugation, each time with 50 cc. of distilled water
(previously adjusted to pH 7 by being made about 0.0004 M with respect to NaHCOs) ,
in order to remove electrolytes. During the final washing the sediment swelled, but
yielded less than 0.5% of its total phosphorus to the supernatant fluid. The gelatinous
sediment then was blended (15 seconds in the high-speed mixer) with 250 cc. of dis-
tilled water (pH 7) and shaken overnight. The extremely viscous mixture was again
briefly stirred in the high-speed mixer and centrifuged for 30 minutes at 2000 X (j-
The P contents of the very viscous, opalescent supernatant fluids averaged 430 ^g.

'' K. G. Stern, G. Goldstein, J. Wagman, and J. Schryver, Federation P roc. 6, 296

(1947).
" D. C. Gajdusek, Biochim. et Biophys. Ada 5, 397 (1950).
^* K. G. Stern, G. Goldstein, and H. G. Albaum, J. Biol. Chem. 188, 273 (1951).
" M. H. Bernstein and D. Mazia, Biuchim. et. Biophys. Acta 10, 600 (1953).
•i« J. A, V. Butler, P. F. Davison, D. W. F. James, and K. V. Shooter, Biochim. et

Biophys. Acta 13, 224 (1954).
" S. Zamenhof and E. Chargaff, J. Biol. Chem. 180, 727 (1949).
38 M. E. Maver and A. E. Greco, J. Biol. Chem. 181, 861 (1949).
" M. Webb, Nature 169, 417 (1952).
« M. Webb, Exptl. Cell Research 5, 27 (1953).

314 ERWIN CHARGAFF

per cc. (about 90% of total P in the mixture) . The reextraction of the insoluble residue
with water or M NaCl yielded negligible amounts of P.

For the precipitation of the nucleohistone the aqueous solution was made 0.15 M
with respect to NaCl by the addition of 5.66 vol. of 0.177 M NaCl. The resulting
precipitate was collected 30 minutes later by centrifugation and washed on the centri-
fuge with 0.15 M NaCl and then, very briefly, with a very small amount of distilled
water. The weight ratios of protein to P in such preparations were around 12.

The extinction, expressed as eCP),"^ of nucleohistone thus prepared is
around 6600, regardless of the presence or absence of salt and close to the
figure found for most specimens of sodium deoxypentose nucleate. [See
below, Section IV. 1, and also Beaven, Holiday, and Johnson, Chapter 14.]
This is of interest, since it would have been conceivable that the extinction
of a nucleoprotein would be less than that of the free nucleic acid, if the
purines and pyrimidines were involved in the architecture of the intact
conjugated protein in such a manner as to bring about the suppression of
some of the chromophores.*^

Nucleohistone is soluble at a very low ionic strength. With increasing
salt concentration, the solubility first decreases steeply, reaching a mini-
mum in isotonic solutions, and then rises gradually. Though variations
occur, freshly prepared nucleohistone is, in general, soluble in NaCl solu-
tions stronger than 0.7 or 0.8 M. Small amounts of phosphorylated com-
pounds that can be extracted from nucleohistone by 0.02 to 0.15 M NaCl
solutions usually lack the typical nucleic acid spectrum and represent
impurities.

(2) Preparation of the Nucleoprotein of Avian Tubercle Bacilli.^^ A mixture of 25 g.
of ether-washed dry bacilli and 100 g. of washed, very fine, Pyrex glass powder (di-
ameter 3 n) was moistened with 0.1 M borate buffer (pH 8.3) and divided into eight
portions, each of which was ground for 30 minutes in a mortar. The ground cells were
united, shaken in a refrigerator with 500 cc. of the borate buffer for 2 days, and centri-
fuged for 30 minutes at 1900 X g. The strongly opalescent, slightly yellow supernatant
was decanted through a filter. The centrifugation residue was washed with 500 cc. of
borate buffer which then served for the extraction of a second 25-g. portion of disin-
tegrated bacilli. In this manner a total of 100 g. of organisms was processed. The ex-
tracts were dialyzed against running water for 48 hours, concentrated by pervapora-
tion to about one-third of the original volume, and again dialyzed against ice-cold
distilled water for 72 hours. Ethyl mercurithiosalicylate was added (0.01%) and the
bacterial glycogen" removed by sedimentation at 31,000 X g. The supernatants were
once more dialyzed, and the crude nucleoprotein fraction was recovered by evapora-
tion of the frozen solution in a vacuum (yield 2.7 g.). The yields varied for different
preparations between 2.4 and 3.4% of the starting material. The crude nucleoprotein
fractions, which could easily be dispersed in water or buffer solutions, gave positive
reactions for deoxypentose with diphenylamine and cysteine (see Chapter 9) and also

" E. ChargafT and S. Zamenhof, J. Biol. Chem. 173, 327 (1948).

« B. Magasanik and E. ChargafT, Biochim. etBiophys. Acta 7, 396 (1951).

« E. Chargaff and D. H. Moore, J. Biol. Chem. 156, 493 (1944).

ISOLATION AND COMPOSITION OF DEOXYPENTOSE NUCLEIC ACIDS

315

the Feulgen reaction (see Chapter 17). They were not precipitated when their solution
in M sodium chloride was diluted to a molarity of 0.15. Thej' were precipitated by tri-
chloroacetic acid and gave positive biuret, xanthoproteic, Millon, and Hopkins-Cole
reactions. Electrophoresis revealed the presence of three components.

For further purification, the dialyzed crude nucleoprotein solution (190 cc), con-
taining a total of 265.1 mg. of N and 16.1 mg. of P, was brought to pH 4.3 by the
addition of 2% acetic acid. The mixture in which a precipitate appeared immediately
was chilled for 3 hours and centrifuged. The sediment was washed with ice-cold 0.05 M
citrate buffer of pH 4.3, dissolved in borate buffer (pH 8.4), and dialyzed. To this
solution an equal volume of saturated ammonium sulfate solution was added and the
precipitate removed by centrifugation. The supernatant solution was, after prolonged
dialysis, evaporated in the frozen state in a vacuum. The purified nucleoprotein
formed a white fiber felt and weighed 0.29 g. It contained N 12.1%, P 3.2% (N:P per
cent ratio 3.8), accounting for about 60% of the phosphorus of the crude nucleopro-
tein fraction. The absorption spectra in the ultraviolet of both the crude and the
purified nucleoprotein fractions are reproduced in Fig. 1.

h. Extraction with Strong Salt Solution

This procedure, very widely used as the first step in the isolation of
deo.xy pentose nucleic acids (see below, Section 111.2), has also been much

9000

7000

• 0

e

/

«(P) 5000

■ 1

it

3000

■

-7

1000

■:

300 280

260

240 220

Fig. 1. Ultraviolet absorption spectra of deo-xypentose nucleoprotein of avian
tubercle bacilli in 0.03 M borate buffer at pH 7.9. Curve I, purified nucleoprotein;
Curve II, crude nucleoprotein. (Taken from Chargaff and Saidel.'')

316 ERWIN CHARGAFF

employed for the preparation of nucleoprotamine^^-®''^ and nucleohis-
tone.^^'^^'*^'^" There exists, however, considerable evidence that, as regards
the physical intactness of the nucleoproteins,*^ this method is inferior to that
described in the preceding section (II. 3. a) since, at any rate in molar or
stronger sodium chloride solutions {M glycine is considered less harmfuP^),
a far-reaching separation between nucleic acid and protein takes place.^°"'
24,25,30,62,63 fhls cau bcst bc shown under experimental conditions that lead
to the removal of one of the partners. ^°'^^ On the other hand, an advantage
of this procedure may be seen in the suppression of the activity of deoxy-
ribonuclease in strong salt solutions.^'*"

The preparation of a nucleoprotamine is given as an example.

(/) Preparation of the Nucleoprotamine of Trout Sperm.''' The operations were car-
ried out in a cold room. Before extraction the freshly collected trout sperm were
washed with a solution containing in 1000 cc. : 7.8 g. NaCl, 0.664 g. KCl, and 0.687 g.
K2SO4." After stirring, the suspension was centrifuged at 5000 r.p.m. for 15 minutes.
The washed sperm were extracted with M NaCl (final concentration after mixing).
On adding salt solution, the sperm mass immediately became sticky and gelatinous,
so that it appeared at first as if the cells were merely swelling. It was necessary to add
a large volume of solution and to stir vigorously in a Waring mixer before it became
apparent that the cells were breaking up as their contents passed into solution. For a
mass of sperm with a dry weight of 900 mg. the volume of the extraction mixture
should be about 500 cc. Even so, the mixture was quite viscous. After vigorous stir-
ring the mixture was centrifuged at 12,000 r.p.m. for 60 minutes. A perfectly clear,
viscous supernatant and a scanty residue were obtained. The material extracted from
the sperm was precipitated by pouring the supernatant into 6 volumes of water.** The

** I. Watanabe and K. Suzuki, J. Cheni. Soc. Japan, Pure Chem. Sect. 72, 578, 580,
604 (1951) ; Chem. Abstr. 46, 3666 (1952).

" H. V. Euler and L. Hahn, Arkiv. Kemi, Mineral. Geol. 22A, No. 17 (1946) ; 23A, No.
5 (1946).

" M. L. Petermann and C. M. Lamb, J. Biol. Chem. 176, 685 (1948).

« G. Frick, Biochim. et Biophys. Acta 3, 103 (1949).

" M. Fleming and D. O. Jordan, Discussions Faraday Soc. No. 13, 217 (1953).

" E. J. Ambrose and J. A. V. Butler, Discussions Faraday Soc. No. 13, 261 (1953).

«» J. M. Luck, D. W. Kupke, A. Rhein, and M. Hurd, J. Biol. Chem. 205, 235 (1953).

" K. G. Stern and S. Davis, Federation Proc. 5, 156 (1946).

" S. S. Cohen, J. Biol. Chem. 158, 255 (1945).

" M. H. Bernstein and D. Mazia, Biochim. et Biophys. Acta 11, 59 (1953).

" M. Laskowski, Arch. Biochem. 11, 41 (1946).

" M. Kunitz, J. Gen. Physiol. 33, 363 (1950).

** It may be of advantage to use a wash fluid containing sodium citrate. The solution
used by Chargaff et al." contained the following molarities: sodium chloride 0.123,
potassium chloride 0.009, potassium sulfate 0.004, sodium citraie 0.01. (Compare
also, Bernstein and Mazia. ^^)

" E. Chargaff, R. Lipshitz, C. Green, and M. E. Hodes, J. Biol. Chem. 192, 223 (1951) .

" This step is, in my experience, often accompanied by considerable losses, an ob-
servation also made by v. Euler and Hahn." If the nucleoprotein is to serve merely
as an intermediate in the preparation of deoxypentose nucleic acid, it is prefer-
able to employ precipitation with alcohol. (See below in Section III. 2.)

ISOLATION AND COMPOSITION OF DEOXYPENTOSE NUCLEIC ACIDS 317

precipitate was in the form of long fibrous strands which could easily be wound around
a rod. The fibrous precipitate redissolved in M NaCl. Vigorous stirring shortened the
time needed to dissolve the precipitate. Any suspended particles were removed by
centrifugation of the viscous solution, and the dissolved material was then reprecipi-
tated by pouring the solution into 6 vol. of water. The fibrous material was soluble in
M, and insoluble in 0.14 M, NaCl. Dissolved in M NaCl it kept well at 0° without
preservative. Dried preparations contained P 5.93%, 6.14%; N 18.1%, 18.4%.

4. Properties
a. Some Chemical and Physical Characteristics

In the absence of a proper biological test procedure no decision on the
native state or the attributes of intactness of a nucleoprotein can be made.
It is, perhaps, possible to distinguish less badly degraded nucleoproteins
from those suffering from excessive mistreatment; but before a systemati-
zation of nucleoproteins has been reached — and we are very far from it —
such discussions appear pointless. On the other hand, the failure to isolate
a nucleoprotein should not be taken, as is sometimes done,^^ as an indica-
tion of its absence. The nucleoproteins of fish sperm and calf thymus remain
the only easily accessible deoxynucleoproteins, and most of the work has
been done with them. Certain points of difference between nucleoproteins
and artificially prepared protein nucleates will be mentioned below (Sec-
tion II. 4. c).

The solubility properties of freshly prepared nucleohistone have been
mentioned in Section II.3.a.(l); those of nucleoprotamines seem to be
essentially similar.^^ Little is known about salts or complexes of nucleo-
proteins with heavy metals. The nucleoprotein of tubercle bacilli is com-
pletely precipitated by lanthanum.-^

The recognition of contaminating pentose nucleoproteins is usually based
on the available color reactions (Chapter 9) or on the differences in hy-
drolysis behavior of the respective nucleic acids (Chapters 5 and 16). There
exists, unfortunately, no procedure permitting the separation of deoxy-
pentose and pentose nucleoproteins, once they have been isolated together,
as for instance in some of the older preparations from liver. ^^ It is necessary
to undertake the isolation of the deoxypentose nucleoprotein from material
that has been freed of pentose nucleoprotein. This can in many, but not
all, cases be done by extensive preliminary washing of the cellular material
with 0.14 M NaCl or, better, with a mixture of 0.1 M NaCl and 0.05 M
sodium citrate. In certain instances a partial centrifugal separation of the
extracted nucleoproteins may be feasible. In a number of studies, con-
tamination was, at least in part, avoided by employing isolated nuclei as

6' V. L. Koenig, L. Larkins, and J. D. Perrings, Arch. Biochem. and Biophys. 39, 355

(1952).
«o J. P. Greenstein and W. V. Jenrette, J. Natl. Cancer Inst. 1, 91 (1940).

318 ERWIN CHARGAFF

the source material.'^'' •2®"'^°" The separation of the nucleic acids them-
selves will be discussed later.

The nucleic acid content of different nucleoproteins varies, with source
and preparation, from about 35 to about 60% of the dry weight. In most
nucleohistone preparations it is around 50%. In the nucleoprotamines of
trout and of herring Felix et al}^ found a phosphorus-to-arginine ratio of 1.
(Compare, also, Vendrely.*^) The protein content of nucleohistone may be
determined by a modified biuret reaction.^"

The absorption spectrum in the ultraviolet of a nucleoprotein is, in
general, identical with that of its nucleic acid moiety (maximum around
260 m/i).^^'"'^^'^" (Compare Fig. 1.) No depression of the e(P) of the nucleic
acid component is noticeable, nor does the extent of extinction change
materially under conditions of complete dissociation.^"

Our information on the physical characteristics of nucleoproteins (com-
pare also Chapter 13) is no less meager than that on their chemical proper-
ties. Much of what goes under this name in the literature really refers to
the nucleic acids themselves. Studies on the sedimentation behavior in the
ultracentrifuge of nucleohistone preparations were undertaken by Stern and
collaborators,^^ '^^ but have not yet been described in detail. Several other
investigations by means of the ultracentrifuge have also been published^^'*^'
47,63.64 a^g have numerous studies on electrophoretic^^'^®'^^'^^'*^** and vis-
cosity properties.^® ■^^■^^•^^•®*'" Despite several, often very discordant, esti-
mates no molecular weight can safely be stated. No more than passing
reference can be made here to studies of X-ray diffraction®^ and scattering,®^
of orientation phenomena in nucleoprotein films,^^ of dielectric properties,^"
and to observations in the electron microscope^' '^'^ and in polarized light."

" L. Ahlstrom, H. von Euler, and L. Hahn, Arkiv. Kemi, Mineral. Geol. 22A, No. 13

(1946).
«2 R. Vendrely and C. Vendrely, Nature 172, 30 (1953).
«' R. F. Steiner, Trans. Faraday Soc. 48, 1185 (1952).
" K. V. Shooter, P. F. Davison, and J. A. V. Butler, Biochim. et Biophys. Acta 13,

192 (1954).
" J. L. Hall, J. Am. Chem. Soc. 63, 794 (1941).

8« Q. Van Winkle and W. G. France, /. Phys. & Colloid Chem. 62, 207 (1948).
«' R. O. Carter and J. L. Hall, /. Am. Chem. Soc. 62, 1194 (1940).
68 M. H. F. Wilkins and J. T. Randall, Biochim. et Biophys. Acta 10, 192 (1953).
«9 D. P. Riley and U. W. Arndt, Nature 172, 294 (1953).
'0 L. G. Allg^n, Acta Physiol. Scand. 22, Suppl. 76 (1950).
" H. Fischer, O. Hug, and W. Lippert, Chromosoma 5, 69 (1952).
"« W. J. Frajola, M. H. Greider, and J. G. Rabotin, Biochim. et Biophys. Acta 14,

18 (1954).
" J. C. White and P. C. Elmes, Nature 169, 151 (1952).

ISOLATION AND COMPOSITION OF DEOXYPENTOSE NUCLEIC ACIDS 319

b. Cleavage and Degradation

It already has been mentioned (Section 11.3. b) that there exists con-
siderable evidence that deoxypentose nucleoproteins, or at any rate the
nucleoprotamines and nucleohistones, occur in strong salt solutions in a
largely dissociated state which has often been regarded as representing no
more than a mixture of protein and nucleic acid. This cleaving effect of a
high concentration of salt" has been employed in what, historically speak-
ing, is one of the most important procedures for the isolation of intact
deoxypentose nucleic acid, namely, that of Hammarsten.^" It is also demon-
strated by the observation that the nucleic acid can be freed entirely of
protamine by the dialysis of the nucleoprotamine of trout sperm in the
presence of M NaCl." The addition of alcohol to a salt solution of nucleo-
histone has a similar effect.^" '^'^ The fractional dissociation of nucleohistone
by increasing concentrations of salt has been studied in some detail by
Crampton et al.^" They discuss the possibility that even at a high ionic
strength there may remain some residual unbroken links between the
nucleic acid and the histone unless one of the partners is being removed
continually.

While agents known to break hydrogen bonds, e.g., urea, guanidine, etc.,
do not seem to have been much employed for preparatory purposes, the
detergents have found useful application. After the discovery of the effect
of sodium dodecyl sulfate on tobacco mosaic virus," this substance or
similar commercial detergent preparations often have served for the isola-
tion of deoxypentose nucleic acid. (See below. Section III.2.C?.) Sodium
deoxycholate has also been used occasionally.^''^*

The degradation of deoxypentose nucleoproteins has, in contrast to their
cleavage, been little studied, no doubt because of their lability and unfavor-
able solubility properties. Here again, though title or text may not indicate
it, much of the work probably deals with the nucleic acids rather than with
the nucleoproteins. Nemchinskaya^^ found depolymerization to be retarded
when pancreatic deoxyribonuclease acted on a nucleoprotein.

c. Artifacts

It is very easy to mix a nucleic acid with a protein, but very difficult
to describe the resulting product. The childish urge to pour all into one pot
has abated of late and the term nucleoprotein is no longer applied to every
concoction. This has been brought about by a realization of the enormous

" M. Sreenivasaya and N. W. Pirie, Biochem. J. 32, 1707 (1938).

'« O. T. Avery, C. M. MacLeod, and M. McCarty, J. Ezptl. Med. 79, 137 (1944).

"* V. L. Nemchinskaya, Biokhimiya 15, 478 (1950); Chem. Abstr. 45, 3439 (1951).

320 ERWIN CHARGAFF

range of variabilities that may prevail when two very large polyampholytes
combine with each otherJ^

When a polyacid, such as a deoxy pentose nucleic acid, combines with a
polybase, such as histone or protamine, many different salts or complexes
may form owing to the multiple possibilities of cross-linking taking place;
these may range from a close-knit fabric, the two partners being warp and
woof, to a treelike arrangement, in which protein molecules are attached
as branches to the nucleic acid trunk (or vice versa), or even to a stoichio-
metric sandwich, in which the minimum number of protein molecules that
can be accommodated is aligned lengthwise on a nucleic acid molecule. In
most cases, hybrids between all these forms probably will occur. This is a
field in which poor reproducibility is almost guaranteed, though, statis-
tically, there may be little difference between the results. The first event
that occurs when the two components are brought together may condition
all subsequent reactions; and the composition of the solvent, the relative
and absolute proportions of the partners, and even more the order and
rate of their mixing, will influence the quahty of the reaction products.
Moreover, the direction that electrostatic attraction takes first in a given
case will not be without influence on the type of secondary valence bonds
established subsequently. These considerations are of importance for a
decision whether a nucleoprotein isolated from a tissue may be considered
as intact or whether it has been converted to an artifact in the course of
its preparation owing to a random reassociation of its separated compo-
nents. (Compare also Section II. 1.)

The effect of a series of electrolyte concentrations on several properties
of calf thymus nucleohistone, prepared as described in Section II.3.a.(l),
has been studied by Crampton et aL^" (See also Section II.4.a.) "Native"
nucleohistone preparations, which had not been exposed to higher than
0.14 M salt concentrations, were much less soluble in 0.6 M NaCl than in
0.7 M; those that had previously been in contact with M NaCl showed no
such differences. The latter preparations also had a higher viscosity. Other
observations that are relevant to this problem will be mentioned later in
connection with the discussion of the fractionation of nucleic acids (Sec-
tion VIII).

The first observations on the formation of a precipitate when nucleic
acid and protamine are mixed are due to.Miescher.^ Since that time numer-
ous studies of the interaction of deoxypentose nucleic acids and proteins
have been published, of which only a few can be mentioned here. Much of
the early work is marred by the employment of degraded preparations.
Reactions with protamine" '^^ and histone,^''"'*^ egg albumin^"""^^ and

^8 A. Katchalsky, Progr. Biophys. and Biophys. Chem. 4, 1 (1954).
" P. Alexander, Nature 169, 226 (1952).

ISOLATION AND COMPOSITION OF DEOXYPENTOSE NUCLEIC ACIDS 321

serum albumin,*^"*'* and with many other proteins*"'*^'** were studied. One
of the most striking differences between a nucleoprotamine or a nucleo-
histone on the one hand and artifacts prepared by the mixing of a nucleic
acid and a basic protein on the other is the solubiUty of the first-mentioned
complexes in the absence of electrolytes. It now appears that it is possible
to prepare nucleic acid - protamine compounds of similar solubility in
water when very dilute aqueous solutions of the components are brought
together.'^ The combination of deoxypentose nucleic acid with enzymes has
been studied in several instances.*^ '^^ Of particular interest is a recent
investigation of Shapot" dealing with the formation of a complex between
deoxyribonucleic acid and pancreatic deoxyribonuclease which is stable in
the absence of Mg++ ions.^*'' Attention may also be drawn to studies on
the protection against heat coagulation that the presence of nucleic acid
confers on proteins. ^^ '^^

III. Isolation of Deoxypentose Nucleic Acids

1. General

The isolation of a cellular constituent of high molecular weight and
complex structure poses several problems of which the most important is
the decision whether the isolated preparation may still be regarded as
representative of the state in which it occurred in the living cell. Strictly
speaking, no compound, once it is isolated from the cell, can be considered
as native. When a pillar is hacked out of a building, neither pillar nor
building is left. If the nucleic acids occur in the cell in combination with

'* P. Alexander, Biochim. et Biophys Acta. 10, 595 (1953).

" K. B. Bjornesjo and T. Teorell, Arkiv. Kemi, Mineral. Geol. 19A, No. 34 (1945).

8" O. P. Chepinoga and R. Sh. Grosblat, Ukrain. Biokhim. Zhur. 21, 121 (1949) ; Chem.
^6s/r. 48, 4014 (1954).

8' J. P. Greenstein and M. L. Hoyer, J. Biol. Chem. 182, 457 (1950).

82 E. Stenhagen and T. Teorell, Trans. Faraday Soc. 35, 743 (1939).

w E. Goldwasser and F. W. Putnam, J. Phys. & Colloid Chem. 54, 79 (1950).

»* E. P. Geiduschek and P. Doty, Biochim. et Biophys. Ada 9, 609 (1952).

"5 A. N. Belozeiskii and G. D. Bazhilina, Biokhimiya 9, 134 (1944) ; Chem. Abstr. 39,
314 (1945).

»« P. Ohlmeyer, Biochim. el Biophys. Acta 4, 229 (1950).

" V. S. Shapot, Biokhimiya 17, 299 (1952) ; Chem. Abstr. 46, 10238 (1952).

88 V. L. Ryzhkov and G. I. Loidina, Doklady Akad. Nauk S.S.S.R. 86, 181 (1952);
Chem. Abstr. 47, 1243 (1953).

88a The spectral shift in the e.xtinction maximum from 260 to 254 m>t is, according to
unpublished experiments with Mr. H. S. Shapiro, not attributable to the formation
of the enzyme-substrate complex, as was thought by Shapot. 8' An identical spec-
trum can be reconstructed by the summation of the individual spectra of deoxy-
ribonucleic acid and deoxyribonuclease, when measured at the pH at which the
complex is examined.

8« C. E. Carter and J. P. Greenstein, J. Natl. Cancer Inst. 6, 219 (1946).

322 ERWIN CHARGAFF

proteins, and most likely as the prosthetic groups of conjugated proteins,
the .danger of secondary changes attending their liberation is particularly
great, since whatever forces had anchored them to the protein may now
have become free to interact. The series of degradative changes to which
a nucleic acid is exposed in the course of its isolation will, however, usually
be gradual; and, while it may not yet be possible to define the perfect
compound, the badly degraded one will, as a rule, be recognized.^" As in
all such cases, one must be satisfied with avoiding the avoidable.

Several physical criteria (viscosity, ultracentrifugal and electrophoretic
behavior, light scattering, ultraviolet absorption, etc.) that are in certain
cases useful for a decision on the integrity of a preparation are discussed
in Chapters 13 and 14. Unfortunately, from most of these tests we may
learn how bad a given specimen is, but only rarely how good it is. This
applies even more to the use of biological criteria of testing, the retention
or loss of transforming activity''^ (see Chapter 27) ; this expedient is, in its
present form, far from being a generally useful guide. ^^

While the early workers in this field, such as Miescher and Hoppe-Seyler
(see Chapter 1), conjectured the macromolecular and complex character of
the nucleic acids that they were the first to isolate, many of the precepts
and safeguards of biochemical etiquette were forsaken by the succeeding
generations. This may have been due in part to the powerful influence that
the great successes of the organic chemistry of small molecules had on the
development of biochemistry.'' A passage from Levene's book^ (p. 251) may
be instructive: "At that period of their work these investigators were still
under the influence of the traditional belief in the unusual lability of the
nucleic acids, and they accordingly avoided the use of heat for the libera-
tion of the nucleic acid from the protein." A revival of this, by no means
erroneous, traditional belief took place in the last thirty years.^"'®-'^^

Deoxypentose nucleic acids usually are isolated from tissues in the form
of their sodium salts. The requirements for a satisfactory preparation may
be listed as follows: (a) absence of proteins, polysaccharides, lipids, etc.;
(b) absence of pentose nucleic acids; (c) freedom from inorganic salts and
other easily dialyzable impurities; (d) a phosphorus content not far from
9.2%; (e) a maximum of absorption in the ultraviolet at pH 7 between
257 and 261 mn with an 6(P) around 6600 (see Chapter 14); (f) fibrous
character, a high and nonthixotropic viscosity,^" presence of streaming
birefringence (see Chapter 13). Several other requirements could be added
e.g., monodispersity in the ultracentrifuge, electrophoretic homogeneity,

90 S. Zamenhof and E. Chargaff, J. Biol. Chem. 186, 207 (1950).

9' S. Zamenhof, H. E. Alexander, and G. Leidy, J. Exptl. Med. 98, 373 (1953).

92 R. Feulgen, Z. physiol. Chem. 237, 261 (1935) ; 238, 105 (1936).

" R. Signer, T. Caspersson, and E. Hammarsten, Nature 141, 122 (1938).

ISOLATION AND COMPOSITION OF DEOXYPENTOSE NUCLEIC ACIDS 323

and, especially, those growing out of the analytical expectations and im-
pressive structural regularities to be discussed later in this article.

The extent to which these several requirements must be fulfilled in a
given case will, of course, depend upon the particular purpose. I shall give
only two examples. If a specimen of a deoxypentose nucleic acid is to serve
for a chemical and analytical study of its composition and of the quantita-
tive distribution of its constituents, a thorough dialysis and lyophilization,
both at 0°, will facilitate the sampling. Heating at high temperature should
be avoided, ^^ except for the determination of moisture. On the other hand,
preparations that are to be used for physical and structure studies, and
even more those to be tested for biological activity ,^^ should not be subjected
to dialysis in the absence of electrolytes or to lyophilization. More drastic
conditions, such as heat, acid, alkali, or even the excessive use of high-
speed mincers, ^''^ must be shunned.

2. Preparative Procedures

The preparative procedures for the isolation of highly polymerized
sodium deoxypentose nucleate, of which several representative examples
will be given, are based on the nucleohistone studies, discussed in detail
in Section II, of Bang," Hammarsten,^"' and Mirsky and Pollister.^^ These
methods are, in general, applicable to a wide variety of tissues, though
special precautions or modifications may occasionally become necessary.
Once the nucleoprotein or a mixture of nucleic acid and protein has been
extracted, the removal of protein is carried out by methods described by
Hammarsten,^^ Sevag,^^'^* and Pirie.^^-^^

a. Extraction with Strong Salt Solution, Deproteinization with Chloroform

(/) Sodixim Deoxyrihonudeate of Calf Thyrmis.^^ Fresh frozen calf thymus glands
(54.5 kg.) were minced and suspended in 0.9% sodium chloride (54 1.) and milled to
produce a fine suspension. This suspension was centrifuged (6300 r.p.m.) and the solid
material resuspended in 0.9% sodium chloride (45.5 1.) and milled and centrifuged as
before. The tissues, which were now free of material containing pentose, were sus-
pended in 10% sodium chloride (214 1.) with vigorous mechanical stirring at 0°. At

^* A. R. Peacocke, Biochim. et Biophys. Acta 14, 157 (1954).

'^* There exists no really good name for these useful, though sometimes almost too
effective, machines ("Waring Blendor," "Turmix," etc.). "Homogenizer" — a
horrible lucus a non lucendo — must be avoided; "disintegrator" or "macerator"
sound too much like science fiction; "blender" is culinary and not sufficiently
descriptive. "High-speed mincer" or "high-speed mixer" is perhaps least objec-
tionable.

" M. G. Sevag, Biochem. Z. 273, 419 (1934).

" M. G. Sevag, D. B. Lackman, and J. Smolens, /. Biol. Chem. 124, 425 (1938).

" F. C. Bawden and N. W. Pirie, Biochem. J. 34, 1278 (1940).

«8 J. M. Gulland, D. O. Jordan, and C. J. Threlfall, J. Chem. Sac. 1947, 1129.

324 ERWIN CHARGAFF

this stage the viscosity of the solution increased considerably. After extraction at 0°
for 48 hours, the insoluble material was removed by centrifuging (6300 r.p.m.) and
the deoxypentose nucleoprotein precipitated from the resultant solution (pH 6.5) by
the addition of an equal volume of industrial methanol. The precipitated solid was
washed with 70%, then 100% industrial methanol and dried in a vacuum at room
temperature. Yield, 1.69 kg. of a very slightly yellow fibrous solid.

The nucleoprotein (500 g.) was powdered to assist solution and dissolved in 10%
sodium chloride (45.5 1.) with vigorous mechanical stirring. The solution, which was
viscous, was clarified by centrifuging (6300 r.p.m.). To the clear solution was added
an equal volume of a mixture of chloroform (35 parts) and amyl alcohol (10 parts),
and the mixture was emulsified by rapid mechanical stirring. The emulsion was then
separated by centrifuging (DeLaval, model "500", disc -type bowl, 5500 r.p.m.) into
three parts: (a) the chloroform - amyl alcohol mixture; (b) a solution containing the
sodium salt of deoxypentose nucleic acid and nucleoprotein; (c) a gel of protein
hydrochloride and the chloroform - amyl alcohol mixture. The protein gel remained
in the bowl of the centrifuge whereas the chloroform - amyl alcohol mixture and the
solution of nucleic acid and nucleoprotein were discharged from separate outlets.
The last-mentioned solution was again emulsified with the chloroform - amyl alcohol
mixture and the process repeated until no gel was formed on emulsification; this re-
quired nine emulsifications. The sodium salt of the deoxypentose nucleic acid was
precipitated by the addition of an equal volume of methanol, washed free from chlo-
ride with 70% methanol, then 100% ethanol, and finally ether, and dried in a vacuum
at room temperature. Yields from two 500-g. quantities of nucleoprotein were 130 g.
and 150 g. of a white fibrous solid giving negative biuret and Sakaguchi tests.

(a) Small-scale preparation. A tj'pical preparation from our laboratory is briefly
described here. All operations were carried out at about 4°. Two hundred and twenty
grams of freshly obtained, well-trimmed calf thymus were cut into small pieces and
triturated for 3 minutes in a "Waring Blendor" in 400 cc. of a mixture of 0.1 M NaCl
and 0.05 M sodium citrate (adjusted to pH 7). The sediment resulting from centri-
fugation at 1900 X g for 45 minutes was washed twice more with 350-cc. portions of
the same fluid and stirred in a high-speed mixer with 1250 cc. of 10% NaCl. The mix-
ture was kept overnight and centrifuged at 20,000 X g for 45 minutes. The sediment
was reextracted for several hours with 800 cc. of 10% NaCl and the mixture centri-
fuged as above. The combined extracts (about 1600 cc.) were injected in a slow stream
into 2 vol. of 95% ethanol. The white strands of nucleohistone were lifted by spooling,
drained, and washed with 70% and 80% ethanol. The material was dissolved with the
aid of rapid stirring in 2500 cc. of 10% NaCl and the solution freed of protein by being
stirred eight times, each time for 150 seconds and being followed by centrifugation at
1900 X g, with one-third volume of chloroform - amyl alcohol (3:1). The supernatant
aqueous solution was finally injected into 2 vol. of 95% ethanol and the fibers were
spooled, lifted, drained, and washed with 70%, 80%, and 100% ethanol. They were
then dried in air or recovered by the lyophilization of their dialyzed aqueous solution.
The yield of sodium nucleate amounted to 1.4. to 1.9% of the fresh tissue.

(2) Sodium Deoxyribonucleate of Yeast^^ The isolation of this substance
may be regarded as an example of an extreme case, as yeast cells contain
almost 50 times as much pentose as deoxypentose nucleic acid.

One kilo of fresh bakers' yeast was washed with 1 1. of 0.1 M sodium citrate (pH
7.3). The 3'east cells, recovered by centrifugation, were suspended in 180 cc. of the
sodium citrate solution and the thick suspension was passed through an ice-cooled
wet crushing mill for bacteria. Each 50-cc. portion was ground for 30 minutes. Fol-

ISOLATION AND COMPOSITION OF DEOXYPENTOSE NUCLEIC ACIDS 325

lowing dilution with 590 cc. of sodium citrate, the crushed suspension was centrifuged
for 2 hours at 4000 r.p.m. Two solid layers sedimented underneath a very opalescent
supernatant (915 cc), a bottom layer of intact cells and an upper layer of cellular
fragments. A separate determination showed that about 55% of the cells had actually
been crushed.

The upper solid layer consisting of cellular fragments (485 cc.) was suspended in
850 cc. of ice-cold M sodium chloride solution of pH 6.3. The slimy mixture was kept
in the refrigerator for 72 hours and then centrifuged at 4000 r.p.m. for 2 hours. The
rapid addition of 2 vol. of chilled absolute ethanol to the very viscous supernatant
resulted in the precipitation of white threads that could easily be wound on a glass
rod and thereby separated from a granular precipitate suspended in the mother
liquor. The threads were washed thoroughly by successive immersion in three por-
tions of 73% ethanol, drained, and redissolved in 300 cc. of M sodium chloride with
the use of a high-speed mixer.

The turbid solution was freed of protein by being stirred in a high-speed mixer
with one-third of its volume of a 9:1 mixture of chloroform- octyl alcohol (for 5
minutes), followed by centrifugation at 4000 r.p.m. for 1 hour. After eight treatments
the solution was free of protein and gave no biuret test. At this stage, it was found to
contain 0.6 mg. of deoxypentose nucleic acid per cubic centimeter, corresponding to a
total of 180 mg. in the original 300 cc. of solution.

The addition of 2 vol. of ethanol to the clear protein-free solution again produced
white threads that were spooled on a rod, as described before. An additional amount
of fibrous nucleic acid could be recovered by reworking the granular precipitate re-
maining in the mother liquor.

The threads obtained by the above procedure were found to contain only 19% of
deoxypentose nucleic acid; the remainder consisted of ribonucleic acid (64%) and of
a polysaccharide. The well-drained threads were taken up in 20 cc. of neutral 10%
aqueous calcium chloride and the viscous milky solution was clarified by centrifuga-
tion at 20,000 r.p.m. for 2 hours. The sediment was washed twice, under the same
centrifugal conditions, with 6-cc. portions of 10% CaCh . The slow addition of 0.2
to 0.3 vol. of cold absolute ethanol to the combined clear supernat ants (32 cc.) brought
about the separation of white fibers that were lifted in the usual manner and washed
twice with 10% calcium chloride solution containing 0.3 vol. of ethanol.

This fraction was contaminated with about 20% of ribonucleic acid which could be
removed by enzymic digestion. To a solution of the precipitate in 45 cc. of 0.2 M so-
dium borate buffer of pH 7.8, 1.5 mg. of crystalline ribonuclease was added. The solu-
tion was subjected to dialysis at room temperature against two changes of 2-1. por-
tions of the borate buffer for 14 hours, against running tap water for 17 hours, and,
finally, against several changes of ice-cold distilled water for 26 hours. Then it was
again deproteinized, as described before, and evaporated in the frozen state in a
vacuum or precipitated with ethanol in the presence of sodium acetate. The sodium
salt of deoxypentose nucleic acid thus obtained weighed 105.5 mg. It formed a white
fluff which was readily soluble in water, giving a clear viscous solution. The biuret
reaction was negative.

6. Extraction with Strong Salt Solution, Dcproteinization by Saturation with
Sodium Chloride

(/) Sodium Deoxyribonucleate of Calf Thymus. ^^■^'"' All operations were carried
out at 0° with solutions that were 0.01 M with respect to sodium citrate. 450 g. of

95 R. Signer and H. Schwander, Helv. Chim. Acta 32, 853 (1949) .
!«« H. Schwander and R. Signer, Helv. Chim. Acta 33, 1521 (1950).

326 ERWIN CHARGAFF

freshly removed thymus were cut into small pieces, mixed with powdered solid CO2 ,
passed through a meat grinder, and treated in portions with a total of 4 1. of M NaCl
in a high-speed mixer for a short time. The mixture then was stirred slowly for several
days, until the gel was converted to a very viscous, reddish, turbid solution. The
dilution with water to a 6-fold volume (830-cc. portions were poured into 4.2-1. por-
tions of water) produced precipitation of the nucleoprotein which was stored in 0.01 M
citrate. The combined precipitate was washed by decantation with 8 1. of a 1% NaCl
solution and twice more dissolved in M NaCl and reprecipitated, as described above.
It was then dissolved in 5 1. of 10% NaCl and the solution, which was being stirred,
adjusted with saturated NaCl solution to a volume of 6 1. and saturated by the addi-
tion of solid NaCl. Stirring was continued for 4 days and the mixture then kept for
14 days. A suspension of 480 g. of Celite 545 in 1.5 1. saturated NaCl was added and
the mixture stirred vigorously for 24 hours. Following filtration by suction through
two filter papers and a layer of Celiie (and washing of the filter cake) the filtrate was
clarified after the admixture of "Hyflo Super-Cel" by filtration as above. It then was
poured into 1.5 vol. of alcohol and the fibers were washed with 70% alcohol, squeezed,
and dissolved in 4.5 1. of water while being stirred for several days. Precipitation was
repeated, this time with 2 vol. of alcohol, and the fibers were washed with 80%, 96%,
and 100% alcohol, and ether, and dried in vacuo over H2SO4 . The sodium nucleate
weighed 8 g. (1.8% of the tissue). Jt was stored in a desiccator over saturated NaCl
solution.

c. Extraction with Water

A rapid and simple preparation of deoxypentose nucleic acid, which has
the advantage of permitting the isolation of histone at the same time, is
based on the work of Crampton et al.^'^ It may start from purified nucleo-
histone, as described in Section II.3.a.(l), or be carried out directly. I
describe here a typical isolation often performed in our laboratory (based
on experiments of Drs. C. F. Crampton and M. E. Hodes and Miss R.
Lipshitz).

(/) Sodium Deoxyribonucleate of Calf Thymus. All operations were carried out at
4°. Two hundred grams of fresh, well-trimmed calf thymus were triturated for 30 to
45 seconds in 200 cc. of a mixture of 0.14 M NaCl and 0.01 M sodium citrate in a
"Waring Blendor." The mixture was centrifuged at 2000 X g for 20 minutes and
the sediment washed three more times with the same wash fluid. The supernatant
liquids were discarded. The sediment was distributed with rapid stirring in 850 cc.
of distilled water (pH 7) and the mixture kept for 15 to 20 hours. It was stirred briefly
to reduce its viscosity and passed quickly through a cooled Sharpies supercentrifuge
at 33,000 X g. The effluent was diluted to contain about 2 to 4 mg. of nucleic acid per
cubic centimeter and brought to a 2.6 M concentration by the addition of solid NaCl.
Immediately after the addition of 2 vol. of 95% ethanol the fibers were collected by
spooling, pressed free of mother liquor, washed with 66%, 80%, and 95% ethanol and
with acetone, and dried in air. The yield corresponded to 80% or better of the phos-
phorus present in the original aqueous extract. The sodium nucleate usually con-
tained 5% or slightly more of protein. The protein impurity could be removed com-
pletely by one treatment with sodium dodecyl sulfate (see the next section).

For a preparation of histone and of nucleic acid free of protein on a large scale
900 g. of fresh calf thymus were minced, washed with citrate - physiol. saline, ex-

ISOLATION AND COMPOSITION OF DEOXYPENTOSE NUCLEIC ACIDS 327

traded with water, and the nucleohistone was precipitated at 0.15 M NaCl. It was
collected in the Sharpies centrifuge, dissolved i water, brought to a 2.6 M NaCl
concentration, and the sodium nucleate precipitated with alcohol. The mother liquor
served for the isolation of histone. The sodium nucleate was dissolved in water, treated
once with sodium dodecyl sulfate (see below) and recovered as described in the next
section. The sodium deoxyribonucleate, dried in air, weighed 23.1 g. (2.6% of the
tissue) and contained only 0.65% of protein.

d. Extraction with the Aid of Anionic Detergents

In the recent past, deoxypentose nucleic acid preparations have often
been made with the aid of sodium dodec}^ sulfate'"^ or of detergents of the
Duponol type.'"' The use of sodium xylene sulfonate has also been pro-
posed.'"* As in all preparative methods, the type of source material will
not be without influence on the success of the isolation. In our own experi-
ence, a nucleic acid preparation from sea urchin sperm could not be freed
entirely of protein by treatment with a detergent. '"■*

(/) Sodium Deoxijribonucleale of Calf Thyrnus.^"^ Fifty grams of frozen calf thymus
were minced in a "Waring Blendor" for 3 minutes'"* in 200 cc. of ice-cold 0.9% NaCl
solution containing 0.01 M sodium citrate. The sediment resulting from centrifuga-
tion at 0° and 2500 r.p.m. for 30 minutes was three more times suspended in citrate -
saline, centrifuged as described before, and treated in the "Waring Blendor" for 3
minutes with 1 1. of ice-cold physiol. saline.'"^ The mixture was transferred to a large
beaker and 90 cc. of the detergent solution (5 g. of sodium dodecyl sulfate or purified
Duponol made up to lOOcc. with45%ethanol) were added. The resulting gel wasstirred
vigorously at room temperature for 3 hours during which time it turned gradually
into a very viscous solution. The NaCl concentration then was brought to 1 M by
the addition of 55 g. of salt,'" stirring was continued for 10 minutes, and the mixture
was centrifuged at 0° and 2500 r.p.m. (or at a higher speed) for 3 hours .'°* The crude
sodium nucleate was precipitated by the addition to the supernatant fluid of an equal
volume of 95% ethanol, the fibrous precipitate was lifted by spooling, pressed out,

><" A. M. Marko and G. C. Butler, J. Biol. Chr.m.. 190, 165 (1951).
'•"iE. R.M.Kay, N.S.Simmons, and A. L.Dounce, J. ^7n. Chcm.Soc. 74,1724 (1952).
"" N. S. Simmons, S. Chavos, and H. K. Orbach, Federation Proc. 11, 390 (1952).
10^ E. Chargaff, R. Lipshitz, and C. Green, /. Biol. Chem. 195, 155 (1952).

105 ^ grinding period of 30 to 45 seconds is sufficient in our experience. Excessively
long treatment in the "Waring Blendor" should, in general, be avoided.

106 Yor the preparation of large quantities we have found it advantageous to extract
the nucleohistone at this stage with 10% NaCl solution and to precipitate with
alcohol. The nucleohistone can be stored in the cold und^r aqueous alcohol and then
be processed as described in the following.

"" Care must be taken not to exceed this NaCl concentration, in order toprevent the
precipitation of the detergent.

'"* The mixture must not be stored in the cold at this stage, since it may jellify. We
centrifuge in the cold for 30 minutes at 13,000 r.p.m. in a rotor that has not been
precooled or, if large quantities are to be processed, in a Sharpies supercentrifuge
at 37,000 r.p.m. with a jet delivering 1 1 in 15 minutes. Filtration of the mixture""
is not recommended; it is accompanied by losses and reijuires a disproportionate
amount of patience.

328 ERWIN CHARGAFF

and washed 3 times with 95% ethanol and then with acetone until the washings were
no longer cloudy. It was dried in air (yield about 2 g.)-^"' The crude product was sus-
pended in 700 CO. of distilled water and brought into solution by being stirred rapidly
at room temperature; 63 cc. of the detergent solution was added and the mixture was
stirred for 1 hour. The mixture was made 1 Af by the addition of 45 g. of NaCl, centri-
fuged at 13,000 r.p.m. for 1 hour, and the nucleate precipitated with ethanol from the
supernatant and washed as before. It was dissolved once more in water and the solu-
tion adjusted to 0.14 M NaCl and centrifuged at 13,000 r.p.m. for 1 hour. The NaCl
concentration of the supernatant fluid was increased to 1 M and the sodium deoxy-
ribonucleate precipitated by the slow addition, with stirring, of an equal volume of
95% ethanol. The fibers, washed with ethanol and acetone and dried in air, weighed
about 1.3 to 1.4 g.

e. Comparison of Different Isolation Procedures

A systematic investigation of the merits of the several procedures still
remains to be carried out, except for some preliminary information on
relative yields which has recently been provided by Frick^^^ and a few
findings published by Schwander and Signer.^"'' Frick compared the methods
of Gulland et al.^^ (see Section III.2.a.(l)), of Hammarsten^" (compare the
modification by Schwander and Signer^"" described in Section 111.2.6.(1)),
and of Kay et al}°^ (see Section III.2.c?.(l)). His conclusion, only partially
borne out by the experience of this laboratory, was that the Gulland pro-
cedure gave yields amounting to onlv 5 to 10 % of the quantity of nucleic
acid present in the nucleohistone serving as the starting material; that the
Hammarsten procedure led to a product in good yield, but of a relatively
high protein content; and that the method of Kay et al. afforded the
greatest yield, but produced some degradation, as judged from ultraviolet
absorption.

It is quite clear that yield is only one of several criteria. The constancy
of physical and chemical characteristics of preparations isolated by differ-
ent procedures has been investigated only to a very limited extent; but the
rather extensive experience of our laboratory has failed to reveal divergen-
ces in composition or extinction that were due to the preparative methods,
if the precautions outlined in the preceding sections were observed. (Com-
pare, for instance, the study on deoxypentose nucleic acids of mammalian
origin."^) As regards biological properties it is worth mientioning that treat-
ment with neither sodium deoxycholate and chloroform^^ nor an anionic
detergent^^ appeared to destroy the transforming activity.

The content in deoxyribonucleic acid of fresh calf thymus may be esti-
mated as being in the neighborhood of 2.5%. [See Leslie, Chapter 16.]
Gulland et al.^^ report an average yield of 0.87% for a preparation made

i"' If the freshly precipitated and washed fibers are cut into very small pieces before
being suspended in distilled water, rapid stirring for 1 to 2 hours will be suflficient
to bring them into solution.

I'o G. Frick, Biochim. et Biophys. Acta 13, 41, 374 (1954).

1" E. Chargaff and R. Lipshitz, J. Am. Chem. Soc. 75, 3658 (1953).

ISOLATION AND COMPOSITION OF DEOXYPENTOSE NUCLEIC ACIDS 329

on an almost industrial scale. When the sodium nucleate is prepared by
this procedure in less gigantic dimensions and with efficient washing of the
chloroform gels to overcome entrapment, the yields range, in our experi-
ence, from 1.4 to 1.9%. Schwander and Signer^"" record 1.8%, Kay et a^.,'"^
2.6 to 2.8%. The latter method and the procedure described above under
Section III.2.c.'° (2.6% yield) appear to permit the greatest recovery; but
all the processes outlined in detail before are, on the whole, quite satisfac-
tory, at any rate for many mammalian organs and for fish sperm. Some
special cases will be mentioned below. Although the method outlined by
Schwander and Signer^"" consumes much more time than the others, it
should not be forgotten that one of their nucleate preparations has afforded
the best X-ray photographs."^ ''^ [See Jordan, Chapter 13.]

/. Miscellaneous Procedures and Applications

The isolation methods discussed above, which were developed principally
for calf thymus, may be applied not only to thymus tissue of other genera
(sheep, pig, man)'" but also to several other mammalian organs (e.g.,
spleen,"* kidney,"* thyroid"') without essential changes. For the prepara-
tion of satisfactory specimens of deoxypentose nucleic acid from other
organs, in particular liver,"' "^ exhaustive preliminaiy washing of the
minced tissue, to remove pentose-containing material, will be necessary;
and such specimens may have to be subjected to special purification pro-
cedures after their isolation, in order to free them entirely of pentose nucleic
acid, as will be discussed below. Nucleated erythrocytes have also served
as a source,"* as has hen's egg white. "^"^ Occasional observations on the
isolation of nucleic acids from malignant tissue are likewise recorded in
the literature.'" "^''^-'^^

While fish spermatozoa""*'^^ and testes,"* "^'-^ and also sea urchin

"2 M. H. F. Wilkins, A. R. Stokes, and H. R. Wilson, Nahire 171, 738 (1953).

113 R. E. Franklin and R. G. Gosling, Nature 171, 740 (1953).

ii< E. Chargaff, E. Vischer, R. Doniger, C. Green, and F. Misani, J. Biol. Chem. 177,

405 (1949).
116 M. M. Daly, V. G. Allfrey, and A. E. Mirsky, J. Gen. Physiol. 33, 497 (1950).
11* E. Chargaff, B. Magasanik, E. Vischer, C. Green, R. Doniger, and D. Elson, J .

Biol. Chem. 186, 51 (1950).
11' H. L. Fraenkel-Conrat, W. H. Ward, N. S. Snell, and E. D. Ducay, J. Am. Chem.

5oc. 72, 3826 (1950).

118 H. Fraenkel-Conrat and E. D. Ducay, Biochem. J. 49, xxxix (1951).

119 S. G. Laland, W. G. Overend, and M. Webb, J. Chem. Soc. 1952, 3224.

1*" P. C. Elmes, J. D. Smith, and J. C. White, He Congrhs International de Biochimie,

Resumes des covimun cations, Paris 1952, 7.
121 I. Asimov and R. R. Simon, Federation Proc. 12, 172 (1953).
1" D. L. Woodhouse, Biochem. J. 56, 349 (1954).
1" G. R. Wyatt, Biochem. J. 48, 584 (1951).
1" C. F. Emanuel and I. L. Chaikoff, J. Biol. Chem. 203, 167 (1953).

330 ERWIN CHARGAFF

spermatozoa,'"* "^'^^^ offer no particular difficulties of isolation, except for
the occasional need of 2 or 3 M NaCl solutions for the extraction of the
nucleoprotein,'"'*"^ the preparation of the intact deoxy pentose nucleic acids
of mammalian sperm often is no easy task. In what appears to be the first
isolation of deoxypentose nucleic acid from human sperm'" it was necessary
to treat the washed and defatted spermatozoa with crystalline trypsin
(free of deoxyribonuclease) before the extraction of nucleic acid could be
performed. In this manner, specimens of a high degree of polymerization
were obtained. In later experiments with ram and bull sperm a rather
unsatisfactory expedient, viz., extraction with KOH, seems to have been
employed. '2''^* It is known that even very strong salt solutions fail to
extract nucleoproteins from mammalian spermatozoa;'" but whether this
failure is due to unusual properties of the nucleoproteins themselves or to
obstruction by another protein that is removed by tryptic digestion'^^
cannot be decided.

Our information on deoxypentose nucleic acids from plant tissues is
regrettably meager. It is almost entirely limited to a few preparations from
several varieties of plant germ which since the pioneering experiments of
Kiesel and Belozersky'^^ and the later studies of Feulgen et a/.'-^ have been
known as good sources. Belozerskii and his colleagues''""'^^ have been
particularly interested in the isolation of nucleoproteins, mostly mixtures
of the deoxypentose and pentose varieties, from plant sources. The con-
tamination with pentose nucleic acid of the deoxypentose nucleic acid
specimens isolated in the recent past from wheat or rye germ"^"^'^'"'"*
is an obstacle to structural and analytical studies which appears to have
been overcome only rarely, either by degradation with alkali'-' or by
special purification"* (see Section III. 2. A. below).

g. Nucleic Acids of Microorganisms and Viruses

Unicellular organisms present a special problem as regards the isolation
of deoxypentose nucleic acids. No systematic treatment of preparative

'" S. Zamenhof, L. B. Shettles, and E. Chargaflf, Nature 165, 756 (1950).

126 x. Mann, in The biochemistry of fertilization and the gametes, Biochem. Soc.

Symposia (Cambridge, Engl.) No. 7, 11 (1951).
'" L. E. Thomas and D. T. Mayer, Science 110, 393 (1949).
128 A. Kiesel and A. N. Belozersky, Z. physiol.Chem. 229, 160 (1934).
1" R. Feulgen, M. Behrens, and S. Mahdihassan, Z. physiol. Chem. 246, 203 (1937).
"" A. N. Belozerskii and 1. 1. Dubrovskaya, Biokhimiya 1, 665 (1936) ; Chem. Ahstr. 31,

3100 (1937).
"' A. N. Belozerskii and L. A. Chernomordikova, Biokhimiya 5, 133 (1940); Chem.

Ahstr. 35, 1457 (1941).
"2 A. N. Belozerskii and M. S. Uspenskaya, Biokhimiya 7, 155 (1942); Chem. Abstr.

38, 131 (1944).
'" S. Laland, W. G. Overend, and M. Webb, Acta. Chem. Scand. 4, 885 (1950).
"< G. Brawerman and E. Chargaflf, J. Am. Chem. Soc. 73, 4052 (1951).

ISOLATION AND COMPOSITION OF DEOXYPENTOSE NUCLEIC ACIDS 331

methods can be offered though some progress has doubtless been made
during the time elapsed between the first and the second of two previous
reviews on this subject.'' ''^^ The uncertainties are in part due to the diffi-
culty of preparing sufficient starting material and of disintegrating it in a
suitable manner; but they are to an even larger extent inherent in our
lack of knowledge of microbial nucleoproteins and in the scarcity of informa-
tion on the topical separation of the various elements composing the internal
structure of the microbial cell. Furthermore, the frequent presence of
deoxypentose nucleases of different and largely unknown properties" often
renders the preparation of intact specimens very difficult.

Two examples have been quoted in detail above, namely, the nucleo-
protein of avian tubercle bacilli^' (Section II.3.a.(2)) and the deoxyribo-
nucleic acid of yeast^^'^'^ (Section III.2.a.(2)). From the former, crude
deoxypentose nucleic acid could be prepared by treatment with either
saturated NaCl solution or sodium deoxycholate.-^ The further purification
is discussed in the next section.

The first microbial nucleic acids in the course of whose isolation particular
attention was paid to purity and intactness were those from pneumococci
endowed with transforming activity.'''' •'^^•^^* [Compare Hotchkiss, Chapter
27.] They were isolated by extraction of the cells after lysis with sodium
deoxycholate in the presence of sodium citrate and purified by deproteiniza-
tion with chloroform (see Section 1 1 1. 2. a) and precipitation as the calcium
salt.*^* A similar procedure^^^ or treatment with anionic detergent^^ (Section
III.2.d) has served for the isolation of transforming nucleic acid specimens
from Hemophilus influenzae. While in these instances the disintegration of
the cell was achieved by mild lytic procedures, most microorganisms require
an efficient grinding for extraction to take place. Comminution with very
fine glass powder or in a wet crushing mill for bacteria was employed in
the examples quoted before,-^ •■*' and a similar procedure, shaking with glass
beads (ballotini), led to the extraction of nucleic acid from Hemophilus
periussis}'^^ From Escherichia coli, ground with glass powder, deoxypentose
nucleic acid could be extracted with salt solution;'*' for the extraction of
Serratia marcescens and other organisms 3.5 M aqueous NaCl was em-

'^* E. Chargaff, in Symposium sur le m^tabolisme microbien, He Congres Inter-
national de Biochimie, Paris 1952, 41.
'36 E. Chargaff and S. Zamenhof, J. Arn. Chem. Sac. 69, 975 (1947).
'" M. McCarty and O. T. Avery, J. Exptl. Med. 83, 89 (1946).

138 M. McCarty and O. T. Avery, J. Exptl. Med. 83, 97 (1946).

139 S. Zamenhof, G. Leidy, H. E. Alexander, P. L. FitzGerald, and E. Chargaff, Arch.
Biochem. and Biophys. 40, 50 (1952).

"0 W. G. Overend, M. Stacey, M. Webb, and J. Ungar, J. Gen. Microbiol. 5, 268 (1951).
'*' B. Gandelman, S. Zamenhof, and E. Chargaff, Biochim. et Biophys. Acta 9, 399
(1952).

332 ERWIN CHARGAFF

ployed, in order to suppress nuclease activity. ^''^■^^^ Nucleic acids from
acid-fast bacteria have been isolated by extraction with solutions of low
electrolyte concentration, ^^'^^ with salt and urea solutions,^" and with
alkali."^ The latter procedure, which has also been applied to E. coli,^*^
degrades, of course, the nucleic acids, though it is effective in removing
pentose nucleic acid. The deoxy pentose nucleic acid of Sarcina lutea has
also been described,'''*'^'' as have the nucleic acids from two rickettsiae:
R. prowazeki and R. hurneti}^^

Very little is known about the preparation of deoxy pentose nucleic acids
from viruses and even less about the physical state in which the often very
robust isolation procedures have left them. Following observations on the
behavior of plant viruses,"^ extraction of bacteriophages with urea solu-
tions has been used in several instances.^*" '^^^ A milder method, in which
the bacteriophage is broken by being passed through a colloid mill, thus
producing what has unfortunately been called a "phage grindate," has been
outlined recently. ^^^ In other cases, the polyhedral and capsule viruses of
insects, the composition of the deoxypentose nucleic acids was deter-
mined by the hydrolysis with HCIO4 '^' of the total virus preparations.^"

h. Removal of Pentose Nucleic Acid and of Other Impurities

Pentose nucleic acid is in many respects the most troublesome contami-
nant of deoxypentose nucleic acid preparations. Its presence vitiates spec-
troscopic measurements and makes impossible the analytical characteriza-
tion of the specimens. I am listing here some of the procedures that have
been employed for the removal of pentose nucleic acid. I. Separation by
electrophoresis: this mild, but laborious and wasteful, method has been
applied in several instances.^* '^^'^^ II. Fractionation by way of the calcium
salts :*^'^^^ the separation will, however, be far from complete. III. Purifica-
tion by treatment with crystalline ribonuclease:^^'^'^ this does not ensure
the complete removal of the pentose nucleic acid, which may leave behind

''" E. Chargaff, S. Zamenhof, G. Brawerman, and L. Kerin, J. Am. Chem. Soc. 72,

3825 (1950).
'■" S. Zamenhof, G. Brawerman, and E. Chargaff, Biochim. et Biophys. Acta 9, 402

(1952).
^** O. Snellman and G. Widstrom, Arkiv. Kemi, Mineral. Geol. 19A, No. 31 (1945).
1" A. S. Jones, Biochim. et Biophys. Acta 10, 607 (1953).
>« J. D. Smith and G. R. Wyatt, Biochem. J .iQ, 144 (1951).

'" S. K. Dutta, A. S. Jones, and M. Stacey, Biochim. et Biophys. Acta 10, 613 (1953).
'« G. R. Wyatt and S. S. Cohen, Nature 170, 846 (1952).
'^' F. C. Bawden and N. W. Pirie, Biochem. J. 34, 1258 (1940).
'SO S. S. Cohen, Cold Spring Harbor Symposia Quant. Biol. 12, 35 (1947).
'5' G. R. Wyatt and S. S. Cohen, Biochem. J. 56, 774 (1953).
1" A. Siegei and S. J. Singer, Biochim. et Biophys. Acta 10, 311 (1953).
•" A. Marshak and H. J. Vogel, J. Biol. Chem. 189, 597 (1951).
1" G. R. Wyatt, J. Gen. Physiol. 36, 201 (1952).

ISOLATION AND COMPOSITION OF DEOXYPENTOSE NUCLEIC ACIDS 333

enzyme-resistant portions. [Compare Chapters 11 and 15.] An example of
the application of Procedures II and III has been given above in Section
III.2.a.(2). IV. Dialysis against, or other treatment with, dilute alkah t^^i'*-
143,146 ^i^jg procedure injures, as it purifies, the deoxypentose nucleic acids,
though without noticeable change in composition;'^* the pentose nucleic
acids are converted quantitatively to mononucleotides."* [Compare Chap-
ters 5, 11, and 16.] Examples can, in fact, be found in the literature in which,
after the application of a modified Schmidt-Thannhauser procedure,'" the
composition of both deoxypentose and pentose nucleic acid was deter-
mined in the same cell or cellular fraction. '^*-'^^ V. Preferential adsorption
of pentose nucleic acid on activated charcoal:'*" this procedure is very
effective, if the preparation has been freed thoroughly of protein and if the
deoxypentose nucleic acid, but not the pentose nucleic acid, is highly
polymerized, as is usually the case. The adsorbed pentose nucleic acid can
be recovered by extraction of the adsorbent with aqueous phenol.'''^ VI.
Separation of deoxypentose and pentose nucleic acids by means of cetyl-
trimethylammonium bromide. ''•^^ — Procedures I, II, V, and VI permit the
recovery of pentose nucleic acid.

Impurities of low molecular weight are best removed by dialysis against
water or, preferably, physiol. saline. The removal of protein has been
described before (Sections III.2.a-c?). The elimination of high-molecular
contaminants, especially polysaccharides, often is not easy, since the op-
portunity of using a specific enzyme^^ will not frequently present itself.
Unless recourse is had to the mechanical separation of the fibrous sodium
or calcium deoxypentose nucleate from the usually granular polysaccharide
— a form of handicraft that often is surprisingly effective — a precipitating
agent for nucleic acids, specific within certain limits, may be employed,
e.g., lanthanum salts, "^■'*' from which the deoxypentose nucleic acid is
recovered as the potassium salt by treatment with potassium oxalate, or
such tervalent complex cations as hexammine cobaltic chloride.'*-

IV. Properties of Deoxypentose Nucleic Acids

1. Elementary Composition and Standards of Integrity

Deoxypentose nucleic acids are usually isolated as the sodium salts or,
occasionally, as the potassium salts. "^■'*' The sodium nucleate from calf

" G. Schmidt and S. J. Thannhauser, J. Biol. Chem. 161, 83 (1945).

^« A. Marshak, J. Biol. Chem. 189, 607 (1951).

" D. Elson and E. Chargaff, Experientia 8, 143 (1952).

" D. Elson and E. Chargaff, Phosphorus Metabolism 2, 329 (1952).

^« D. Elson, T. Gustafson, and E. Chargaff, J. Biol. Chem. 209, 285 (1954).

«» S. Zamenhof and E. Chargaff, Nature 168, 604 (1951).

«' E. Vischer, S. Zamenhof, and E. Chargaff, J. Biol. Chem. 177, 429 (1949).

«2 E. Chargaff and C. Green, J. Biol. Chem. 173, 263 (1948).

334 ERWIN CHARGAFF

thymus has been shown to include small amounts (0.01 to 0.1 %) of mag-
nesium.'*^ When cautiously prepared, these salts are obtained in the form
of white, tough strands of fibers, resembling asbestos, or as white fiber felts
after the evaporation of their frozen aqueous solution in a vacuum.'*'' In
addition to the cation, they contain the elements C, H, N, O, and P. The
determination of C and H usually does not contribute much to the charac-
terization of the substances, but N and P should be measured. Nitrogen is
best determined by the Dumas procedure, phosphorus by the Pregl-Lieb
method. For routine purposes, colorimetric estimation procedures for P
are convenient. We employ the method of King;'** compare also Jones et
a/.'** The sodium nucleates contain around 12 % of firmly bound water
which is best determined separately by drying the specimen at 60° for 3
hours and correcting the other analytical values for the moisture content.
The organic constituents should never be estimated in heated preparations.
(Compare also Peacocke.^*)

A selection of analytical data is given in Table I. It will be seen that the
N and P values and also the~ atomic N/P ratios are in many cases in good
agreement with the figures calculated from the distribution of purines and
pyrimidines actually found in the nucleic acids of the several genera. (See
below. Section VII.) Owing to the several structural regularities in all
deoxypentose nucleic acids that will be mentioned later, the N and P
contents are closely similar even for preparations showing wide divergences
in the proportions of individual bases; they cannot be utilized for the precise
definition of composition, except to indicate the degree of purity. The same
purpose is served by the determination of the deoxypentose and pentose
contents which is carried out by colorimetric comparison with suitable
nucleic acid standards."^"* [Compare Dische, Chapter 9.]

The physical tests usually applied are measurements of the absorption
spectrum in the ultraviolet and of viscosity. [See Chapters 13 and 14.]
While there may occur slight shifts in the position of the centers of absorp-
tion (between 257 and 261 m/x at pH 7), the extinction at the maximum
is, in my experience, almost constant. This has often been pointed out."^ •
116,167 When the extinction at the maximum and at pH 7 is expressed in the
most convenient way, namely, as the atomic extinction coefficient with
respect to phosphorus and designated c(P),^' preparations isolated cau-
tiously from a large variety of sources will show surprisingly little divergence

'" G. Jungner, Science 113, 378 (1951).

'*^ Lyophilized nucleic acid preparations often carry considerable static electricity,

and this makes difficult the accurate weighing of small samples. We use a static

eliminator containing a polonium source.
i«6 E. J. King, Biochem. J. 26, 292 (1932).

'»« A. S. Jones, W. A. Lee, and A. R. Peacocke, J. Chem. Soc, 1951, 623.
»" E. ChargafT, Federation Proc. 10, 654 (1951).

ISOLATION AND COMPOSITION OF DEOXYPENTOSE NUCLEIC ACIDS 335

< o :^

g^5

o —

fCCCCOCCMMCOfO

-a

3

cococcpocoeoeocc

^

a

CO
O

Oh

il

(M

CO

CO

gi

03 --^

Ol

Ol

Cl

Ol

C5

o

05

o

CI

s

CO
CO

'^

CO

CO

02

02

o

CI

o

Oi

Ol

05

GO

05

CO

00

00

fe5

Calcu-
lated*

Tt- "^ ■*

(M (M (N

Tf — • <r) CO i:^

(N iM '-I 04 — '

lO lO >o

lO iC iC lO lO

c
s
o

cocoiocO'sf'^ooc^

UJiOCO'O'O'O'^'C

q; ., C 1— I

i5 <u ^ o
rr r o u

a> ^

(1.

a.2 «

^-^ (33 ' — * o3 cd o3 c3 c3
e3rf'^^^o3c3c3

u

31

-a
_ c

C 4J «

i Z a.

£ Oh ^

D -o 7

'S Q -5

4) O

■^ * -5

5 a =

^ -H »

3 O- o

!-5 a

3
O
CO

oj J- Q e
" " o 'S

M cc »: i

3 3 3c

S £ S ■::

-^ -== -= " M

-.-i -M ->j D. .-. C3

.— icacO'^'OOt^cc

° c S

fl o o

^ a H

S Sf

OS »i?5

— OS

s:

o>

C<!

fc

^

o>

s

X

O

-«?

/.'

C)

Oi

o

^

o>

?

^^

CQ

m

at

-c

s

"^

^

"^

"^

5

"->

!«

-a

i

0)

a

B
3

0

D

a

i5

H

in

hJ

-e

>'

-»

*

Ci

-a

■^.

11

<

(J

c

33

fS

O

w

-0

-0

c

3!

C

0

-5

c

6

d

13

a

31

E

;=

o

3!

H

*

m"

„"

-^

V

.-

.ti

,^

-a

>— "

^

^

C

o

M

oi

i

=.

Q

a

Z

■a

-I

>J

•T3

a

3J

OS

a!

3)

n

;*:

^

to

to

n

p

s

3S

i\

rrt

J3

.c

s

H Oi o o c

-^

o a

H

H

a

336

ERWIN CHARGAFF

TABLE II

Ultraviolet Absorption of Several Preparations of Sodium Deoxypentose

Nucleate*

Prepa-
ration
No.

Source

Maximum

Minimum

Ref.

mn

6(P)

m^

231
232
231

231
230
232

e(P)

1
2
3
4
5
6
7
8
9

Calf thymus

Calf thymus

Ox spleen

Pig thymus

Pig spleen

Salmon sperm

Arbacia lixula sperm

Yeast

Avian tubercle bacilli

259
259
259
259
260
260
260
260
257

6600
6400
6500
6850
6800
6700
7300
6100
6400

2800
2700
2900

4600
3200
2500

b

b

d
f

a

* Preparations 1 to 8 were isolated bjr the procedures discussed in Section III.2.a; for Preparation 9 com-
pare Sections II.3.a.(2) and III.2.g. See also Chapter 14 and Section IV. 1 of this chapter.

References

° C. Tamm, M. E. Hodes, and E. Chargaff. J. Biol. Chem. 195, 49 (1952).

*" E. Chargaff, E. Vischer, R. Doniger, C. Green, and F. Misani, J. Biol. Chem. 177, 405 (1949).

" E. Chargafl and R. Lipshitz, J. Am. Chem. Soc. 75, 3658 (1953).

^ E. Chargaff, R. Lipshitz, C. Green, and M. E. Hodes, J. Biol. Chem. 192, 223 (1951).

^ E. Chargaff, R. Lipshitz, and C. Green, J. Biol. Chem. 195, 155 (1952).

■'■ E. Chargaff and S. Zamenhnf, J. Biol. Chem. 173, 327 (1948).

" E. Vischer, S. Zamenhof, and E. Chargaff, J. Biol. Chem. 177, 429 (1949).

from the value of €(P) = 6600. In eight preparations of sodium deoxy-
ribonucleate from ox tissues (seven from thymus, one from spleen) the
absorption maximum was at 259 mn and c(P) = 6650 with a standard
error of ±50."^ A selection of data is presented in Table II. In contrast
to a statement in the literature/"^ I am inclined to consider an e(P) higher
than about 7200 as a sign of denaturation. (Compare the discussion of the
general problem of hyperchromic effects by Magasanik and Chargaff"*- and
in Chapter 14, and also Thomas. ^^^•'^^)

While the absence of high viscosity indicates denaturation or degradation
of the sodium nucleate, no standard values can as yet be given for the
viscosity of carefully prepared specimens. [Compare the discussion by
Jordan, Chapter 13.]

Surprisingly little is known about the optical activity of intact nucleic
acids. To indicate the order of magnitude, some measurements by Tamm
et al}'"^ may be quoted. With a preparation of calf thymus sodium deoxy-

'«8 R. Thomas, Bull. soc. chirn. biol. 35, 609 (1953).

'69 R. Thomas, Biochim. et Biophys. Acta 14, 231 (1954).

i^» C. Tamm, M. E. Hodes, and E. Chargaff, J. Biol. Chem. 195, 49 (1952) .

ISOLATION AND COMPOSITION OF DEOXYPENTOSE NUCLEIC ACIDS 337

ribonucleate (similar to Preparation 2 in Table I) the following values were
found for approximately 0.1% solutions in 0.1 M phosphate buffer of pH
7.1: [a]f = +100° ± 10°; Hf(P) = +1350° ± 100°; [M]^?) = +420°
± 25°. [a]D(P) is the specific rotation with respect to phosphorus, equaling
100 ar,/bc(P), in which h is the layer thickness and c(P) the phosphorus
concentration in grams per 100 cc. of solution. [M]o(P) is the atomic rota-
tion with respect to phosphorus, equaling 0.3098[a]3(P). These terms, and
the corresponding viscosity expression Tjap^P) denoting the specific vis-
cosity divided by the molarity of the solution with respect to P,'" were
adopted for the same reasons that led to the introduction of the term for
extinction e(P).

The formation of insoluble deoxypentose nucleates by tervalent cations,
especially lanthanum, has often been investigated ;^*'^''"^''*'''^'"2 but there
is little information concerning other salts or metal complexes, nor do de-
tailed solubility studies of these compounds seem to have been carried out.
Complexes and precipitates with streptomycin have been described,'"''^* as
has also the interaction of deoxypentose nucleic acids with dyes.'^^'"^

The physical properties of the nucleic acids are outlined in Chapters 13
and 14. Estimates of the molecular weight of calf thymus deoxyribonucleic
acid vary, less with the specimen than with the method of determination,
from 820,0001^* to 7,700,000.1'^ Mention should be made here of investiga-
tions on the electron microscopy of sodium deoxyribonucleates.i*''-^^-

2. Denaturation and Degradation

A mild, but persistent, mistreatment of a protein leads to a state of
malaise known, vaguely, as denaturation. It is not astonishing that the
nucleic acids, especially the deoxypentose nucleic acids, which in the gradual
recognition of their complex properties have emulated the proteins in many
respects, have come also into this legacy. The decay of a macromolecule
of a specific and complicated structure will usually go through a number
of successive stages; the changes, almost imperceptible in the beginning,
multiply cumulatively, until the collapse makes itself known with almost

'" E. Hammarsten, G. Hammarsten, and T. Teorell, Acta Med. Scand. 68, 219 (1928).

•" K. G. Stern and M. A. Steinberg, Biochim. et Biophys. Ada 11, 553 (1953).

'" S. S. Cohen, J. Biol. Chem. 168, 511 (1947).

"* H. V. Euler and L. Heller, Arkiv. Kemi, Mineral. Geol. 26A, No. 14 (1948).

17' L. F. Cavalieri and A. Angelos, J. Am. Chem. Soc. 72, 4686 (1950).

"» L. F.^Cavalieri, A. Angelos, and M. E. Balis, J. Am. Chem. Soc. 73, 4902 (1951).

»" J. L. Irvin and E. M. Irvin, J. Biol. Chem. 206, 39 (1954):

"» R. Cecil and A. G. Ogston, J. Chem. Soc, 1948, 1382.

1" M. E. Reichmann, R. Varin, and P. Doty, J. Am. Chem. Soc. 74, 3203 (1952).

180 J. F. Scott, Biochim. et Biophys. Acta 2, 1 (1948).

'?' R. C. Williams, Biochim. et Biophys. Acta 9, 237 (1952).

'8« H. Kahler and B. J. Lloyd, Jr., Biochim. et Biophys. Acta 10, 355 (1953).

338 ERWIN CHARGAFF

explosive suddenness. The sequence leads probably from the rupture of
secondary valence bonds to the fission of covalent links; but the sense may
be opposite under circumstances, especially during enzymic attack: the
cleavage of covalent linkages could bring about the automatic snapping of,
for instance, hydrogen bonds. The line separating a denaturation product
from a degradation product is not clearly drawn; but one could define as
denaturation products those substances whose preparation caused inter-
ference with the physical properties, but not with the chemical composition,
of the parent nucleic acid, while the latter change will form part of the
description of a degradation product.^*'

Though proteins and nucleic acids share many features, there is one
essential distinction: the ideal "monomer" of a protein consists of one
molecular species, the amino acid; the ideal "monomer" of a nucleic acid,
the nucleotide, is composed of three, namely, base, sugar, phosphoric acid.
This brings about multiple possibilities of breakdown. The deoxypentose
nucleic acid chain can be degraded vertically, as it were, and horizontally,
i.e., perpendicularly to the long fiber axis and parallel to it. I shall return
to this point in the next section.

The denaturation of deoxypentose nucleic acids can be followed by two,
or perhaps three, different methods: (a) modifications in viscosity behavior;
(b) spectral changes; and finally, but only in a very limited number of
cases, (c) loss of transforming activity.^' The second procedure, viz., spec-
troscopy, appears, at least at present, the most fruitful.

As regards viscosity changes induced by acid or alkali, there exists an
extensive literature which cannot be reviewed here in detail [compare Jor-
dan, Chapter 13], though a few investigations should be cited. ^"•^"■^^•^^^-'^^
But viscosity is a treacherous guide. On the one hand it can be shown that
a solution of the sodium deoxyribonucleate of calf thymus in 0.05 M NaCl,
when adjusted to pH 3 by the careful addition of acid, shows a drop of
specific viscosity from 21.4 to 0.5; that the high viscosity of the solution
is regained upon neutralization within 30 minutes; but that under these
conditions an artifact is produced which, in contrast to the undenatured
preparation, has a highly thixotropic character.^" On the other hand, the
adjustment of similar solutions to pH 2.6 by dialysis did not affect the
molecular weight (7,700,000), as determined by light scattering.'*^ For a
discussion of the irreversible changes accompanying the titration of nucleic
acids Chapter 13 should be consulted.

'S3 C. Tamm, H. S. Shapiro, and E. Chargaff, J. Biol. Chem. 199, 313 (1952).

>" C. F. Vilbrandt and H. G. Tennent, J. Am. Chem. Soc. 65, 1806 (1943).

'86 J. M. Gulland, D. O. Jordan, and H. F. W. Taylor, J. Chem. Soc. 1947, 1131.

'«« J. M. Creeth, J. M. Gulland, and D. O. Jordan, J. Chem. Soc. 1947, 1141.

'*' H. V. Euler and A. Fono, Arkiv Kemi, Mineral. Geol. 25A, No. 3 (1947).

'88 ll.Schwa.ndeT,Helv.C him. Acta 32,2510(1949).

'89 M. E. Reichmann, B. H. Bunce, and P. Doty, J. Polymer Sci. 10, 109 (1953).

ISOLATION AND COMPOSITION OF DEOXYPENTOSE NUCLEIC ACIDS 339

The observations of the spectral changes attendmg the denaturation and
subsequent degradation of the nucleic acids start from the fact that the
extinction of intact preparations is lower than would correspond to the
sum of their constituent mononucleotides. The hyperchromic effect of
degradation has been discussed for pentose nucleic acids by Magasanik
and Chargaff.''^ [Compare Chapters 11 and 14.] As regards deoxypentose
nucleic acids, it was Kunitz-^* who first described the intensification of the
extinction brought about by deoxyribonuclease. Similar effects produced
by acid, alkali, heat, or the addition of salts have been studied fre-
quently,'^"-'^^ in greatest detail by Thomas.'*' Owing to the lability of the
pentose nucleic acids, the optical effects accompanying their denaturation
and their degradation can hardly be separated; and in the deoxypentose
nucleic acids, too, a further-reaching chemical degradation, namely, the
removal of the purines, has been shown to have a hyperchromic effect.'^"
But it may be concluded that for such effects to become noticeable in
deoxypentose nucleic acids a relatively mild trea.tment affecting only
secondary valence bonds is sufficient.'*'

The transforming activity of certain bacterial deoxypentose nucleic acids
is discussed in Chapter 27. Reference may be made here, however, to the
curious observation by McCarty that the transforming substance of Pneu-
mococcus Type III is inactivated reversibly by ascorbic acid.''''

The end results of chemical and enzymic degradation need not concern
us here; they are discussed, in different contexts, in Chapters 5, 12, and 15.
Certain aspects will also later be touched upon in this chapter as far as they
bear on questions of composition and structure. What should be mentioned
here — but it can only be done in the briefest form^ — is the existence of
numerous studies on the degradation of deoxypentose nucleic acids by
irradiation with ultraviolet light, "^-'^ by treatment with X-rays,^""-^"*

190 K. K. Tsuboi, Biochim. et Biophys. Acta 6, 202 (1950).

>" L. F. Cavalieri, J. Am. Chern. Soc. 74, 1242 (1952).

"2 G. Frick, Biochivi. et Biophys. Acta 8, 625 (1952).

'" E. R. Blout and A. Asadourian, Biochim. et Biophys. Acta 13, 161 (1954).

194 M. McCarty, J. Exptl. Med. 81, 501 (1945).

'9* A. Hollaender, J. P. Greenstein, and W. V. Jenrette, J. Natl. Cancer Inst. 2, 23

(1941).
i9« M. Errera, Biochim. et Biophys. Acta 8, 30, 115 (1952).

'" M. Seraydarian, A. Canzanelli, and D. Rapport, Am. J. Physiol. 172, 42 (1953).
'9« R. Setlow and B. Doyle, Biochim. et Biophys. Acta 12, 508 (1953).
'«' J. A. V. Butler and B. E. Conway, Proc. Roy. Soc. (London) B141, 562 (1953).

200 B. Taylor, J. P. Greenstein, and A. Hollaender, Arch. Biochem. 16, 19 (1948).

201 M. Errera, Bull. soc. chim. biol. 33, 555 (1951).

202 B. E. Conway and J. A. V. Butler, J. Chem. Soc. 1952, 834.

203 G. Scholes and J. Weiss, Biochem. J. 53, 567 (1953) ; 56, 65 (1954) ; Nature 171, 920
(1953).

204 M. Daniels, G. Scholes, and J. Weiss, iVa^ure 171, 1153 (1953).

20* V. L. Koenig and J. D. Perrings, Arch. Biochem. and Biophys. 44, 443 (1953).

340 ERWIN CHARGAFF

with ultrasound,^"^'^''^ or with radiomimetic agents, such as the sulfur and
nitrogen mustards." '^"^'^^^ Many of these studies have been considered in
detail by Errera.^'" The effect of phenol and urea has also been studied.^^^

V. Some Partial Degradation Products
1. General

Even if the assumption is made — and there is little justification for it,
as will be shown later — that a deoxypentose nucleic acid preparation from
a given cell is composed of only one molecular species, it will be readily
understood that the problem of desciibing its fine structure still is in-
solvable. It will affect this difficulty very little whether we are dealing
with a single polynucleotide chain or with two complementary chains
holding each other in a complicated embrace and exposing identical back-
sides to the outer world. A nucleic acid chain must, according to the molecu-
lar weight assigned to it, be composed of 2500 to 15,000 mononucleotides
of 4 or 5 or more varieties; and only through the most stultifying over-
simplification could a particular sequence be predicted. It will suffice to
point out that a chain consisting of 2500 nucleotides in the proportions
found for the total deoxypentose nucleic acid of ox tissues (see Section VII)
could exist in something like lO'*"" sequential isomers.'' Since the human
mind does not enjoy contemplating the impossible for a long time, it either
forgets, neglects, or reduces it. The latter operation results in the more
modest desire, not to write the entire sequence of nucleotides, but to
discern certain more general structural features, if any can be found.

A stepwise degradation of the nucleic acid chain appears to offer pos-
sibilities of distinction between different entities. That it may be carried
out in two essentially different directions has been pointed out in the
preceding section (IV.2). Owhig to the great difference in stability of the
glycosidic linkage in the purhie and in the pyrimidine nucleosides (compare
Chapters 5 and 9), it is possible to remove the purines preferentially from
a deoxypentose nucleic acid chain by a carefully controlled acid hydrolysis.
Products of this type for which the name apurinic acid has been proposed""
will be discussed in the next section. Less well-defined substances resulting
from a more vigorous breakdown liave long been known as thymic acid.

206 S. G. Laland, W. G. Overend, and M. Stacey, J. Chern. Soc. 1952, 303.
20' I. E. El'piner and A. V. Gerasimova, Doklady Akad. Nauk S.S.S.R. 86, 797 (1952) ;
Chem. Abstr. 47, 2225 (1953).

208 D. T. Elmore, J. M. Gulland, D. O. Jordan, and H. F. W. Taylor, Biochem. J. 42,
308 (1948) .

209 J. A. V. Butler, L. Gilbert, and D. W. F. James, J. Chem. Soc. 1952, 3268. Compare
this paper for references to other work.

2'o M. Errera, M^canismes de I'Action des Radiations sur le Noyau Cellulaire, Les

Editions "Acta Medica Belgica." Bruxelles, 1952.
211 B. E. Conway and J. A. V. Butler, /. Chem. Soc. 1952, 3075.

ISOLATION AND COMPOSITION OF DEOXYPENTOSE NUCLEIC ACIDS 341

The controlled action of deoxyribonucleases leads to breakdown products
of an entirely different character, namely, to a mixture of large oligonucleo-
tides, more resistant to enzymic action than the bulk of the nucleic acid
molecule and originally designated as the "core."-^- This name has given
rise to some criticism; but, since it seems to be more often used by its
critics than by its proponents, it may at least have the virtue of terseness.^^'
Other possible designations would be "limit polynucleotide" or "enzyme-
resistant residue."

2. Preferential Removal of Purines

a. Thymic Acid

The action of dilute mineral acid at 80° or 100° on deoxy pentose nucleic
acid yields a rather ill-defined degradation product which, as its original
discoverers believed it to contain only thj'mine, was designated "thymic
acid."-''' It was later, however, shown to contain also cytosine.^'* -'^ Prod-
ucts of this type, often of varying, always of incompletely known, composi-
tion have played an important part in discussions of the structure of
deoxypentose nucleic acid.''*^-^'^---" Their role in the nucleal reaction of
Feulgen also has often been considered. [Compare Swift, Chapter 17.] In
the course of their studies on partial degradation products and on the
formation of apurinic acid Tamm et a/.'" '^^-^ prepared and analyzed a
series of partial degradation products and compared their composition with
that of the parent calf thymus sodium deoxyribonucleate. Their studies
should be consulted for details. The rate of dialysis of the liberated purines
and the composition of the dialysate, when calf thymus nucleic acid is
exposed to the conditions leading to the formation of apurinic acid (pH 1.6,
37°), are shown in Fig. 2.

b. Apurinic Acid

A typical preparation of apurinic acid from the sodium deoxyribonucleate
of calf thymus will be described here.

(/) Preparation.^'"' To a solution of 105.0 mg. of the sodium nucleate (lyophilized,
moisture content 13%) in 42.5 cc. of water a total of 12.5 cc. of 0.1 A'^ aqueous HCl was

«'2 S. Zamenhof and E. ChargafT, J. Biol. Chem. 178, 531 (1949).

^^^ CORE: "A central part of different character from that which surrounds it."

(Oxford English Dictionary, Vol. II, p. 990.)
2'^ A. Kossel and A. Neumann, Ber. 26, 2753 (1893) ; Z. physiol. Chem. 22, 74 (1896-97).
2'^ H. Steudel and P. Brigl, Z. physiol. Chem. 70, 398 (1910-11).
"« R. Feulgen, Z. physiol. Chem. 101, 296 (1917-18).
2" H. Steudel and E. Reiser, Z. physiol. Chem. Ill, 297 (1920).
2'8 S. J. Thannhauser and B. Ottenstein, Z. phijsiol. Chem. 114, 39 (1921).
2" H. Bredereck and G. Miiller, Ber. 72, 115 (1939).

"0 J. M. Gulland, m Nucleic acid. Symposia Soc. Exptl. Biol. 1, 1 (1947).
«' C. Tamm, H. S. Shapiro, R. Lipshitz, and E. Chargaff, J. Biol. Chem. 203, 673

(1953).

342

ERWIN CHARGAFF

100

Hours
Fig. 2. Liberation of free purines from the sodium deoxyribonucleate of calf thy-
mus at 37° and pH 1.6. The duration of the reaction is plotted as the abscissa against
the concentration of adenine (A) and guanine (G) in the dialysate (left ordinate) and
against the quantities of liberated purines as per cent of total adenine and guanine
contained in the starting material (right ordinates) . The upper part of the graph indi-
cates the molar ratio of adenine to guanine in the dialysate. (Taken from Tamm
eia/.i")

added gradually, when a pH of 1.6 was reached. During the addition a heavy precipi-
tate formed in the viscous solution. The mixture was immediately transferred to a
cellophane bag and dialyzed at 37° against 440 cc. of dilute HCl of pH 1.6 for 26
hours. "2 No change in pH occurred during this time. The clear inside fluid was di-
alyzed against 750 cc. of 0.2 M borate buflfer (pH 7.3) for 22 hours at 4°, against run-
ning tap water (about 12°) for the same period, against frequent changes of distilled
water at 4° for 24 hours, and evaporated in the frozen state in a vacuum. The apurinic
acid formed a pure white fluff, weighing 76.7 mg. and containing 9.7% of moisture. It
was very hj^groscopic when completely dehydrated and was easily soluble in water
to give a solution of about pH 6.5. The yield corresponded to 94% of the starting ma-
terial, when allowance was made for the loss of purines amounting to about 20% of
the initial weight.

{2) Properties. Information about two typical preparations of apurinic
acid is provided in Table III. It will be seen that the recovery of the
pyrimidines and, even more significantly, of the phosphorus is almost
quantitative. The comparison with the data given below in Section VII will
show that no distortion of the inter pyrimidine ratios characteristic of the
parent nucleic acid has taken place. All this is not true of thymic acid
preparations. In apurinic acid the nucleotide sequence of the parent sub-
stance presumably is preserved, except that the positions of the purine

"'^ When larger quantities of apurinic acid are to be prepared, it is not necessary to
carry out the acid treatment under conditions of dialysis. The subsequent neutrali-
zation and dialysis are, of course, required.

ISOLATION AND COMPOSITION OF DEOXYPENTOSE NUCLEIC ACIDS

343

TABLE III
Apurinic Acid from Calf Thymus Sodium Deoxyribonucleate*

Preparation No.
1 2

Nitrogen, % 6.0 6.8

Phosphorus, % 10.7 10.6

Color yield with diphenylamine as % of standard DNAj 108 95

Moles per 100 g. -atoms of Pt

Adenine
Guanine

Total purines

Thymine

Cytosine

27
20

27
20

Total pyrimidines 47

47

Molar ratios

Thymine to cytosine 1.3
Phosphorus to total bases 2.1

1.3
2.1

Balances, % of DNA constituent recovered in product

Thymine
Cytosine
Phosphorus

89
89
89

100

101

96

* The figures refer to the dry preparations.

t The values are corrected for the loss in weight due to the removal of purines.

t When large amounts of hydrolysate are analyzed, between 1.2 and 1.3 mole % of 5-methylcytosine is
found and minute residues of adenine (0.5-1 mole %) and guanine (1 mole %).

nucleotides now are occupied by nonglycosidic deoxyribose phosphate units
which react as free aldehydes. When comparison is made on the basis of
P content, the color yield of the reaction of apurinic acid with diphenyl-
amine is almost identical with that given by deoxypentose nucleates.^-*
[Compare Dische, Chapter 9.]

The absorption spectra of the sodium salt of apurinic acid and of calf
thymus sodium deoxyribonucleate are compared in Fig. 3. The absorption
maximum is at 267 to 268 m/x with an e(P) of 4600 to 4800, the minimum

"' A previous statement'^" • '" that apurinic acid gives a greater color yield with di-
phenylamine than the corresponding quantity of deoxyribonucleic acid is incor-
rect. It was due to an error in the computation of a conversion factor.

344

ERWIN CHARGAFF

/"UUU

ma T-DNA

6000

/\

5000

1

f^APA \

f

V 4000

4

•^ \

1

7

3000

- A

1 1
li
1 1

1 1

V^ \

6

1

2000

^■<v^

(
<

\ ,

1 1 1 1

1

1

290

250

220

Fig. 3. Absorption spectra in the ultraviolet of sodium deoxyribonucleate of calf
thymus (T-DNA) and of the sodium salt of the corresponding apurinic acid (APA)
in phosphate buffer of pH 7.1. (Taken from Tamm et al.^'"')

at 238 m,x with an e(P) of 2900."" "i When the extinction coefficient is
computed with respect to nucleotide P rather than to total P, the cor-
responding e(NP) at the maximum is found at 9800 to 10,400: another
example of the hyperchromic effect of degradation mentioned above in
Section IV. The specific rotation [a]^ is +50°, the atomic rotation with
respect to phosphorus [M]f(P) is +140°. (Compare Section IV. 1.)

Other physical and chemical properties of apurinic acid and its deriva-
tives have been discussed by Tamm and Chargaff .^'^ In contrast to deoxy-
pentose nucleates, apurinic acid reduces Fehling's or Benedict's solution
on warming and reacts with ammoniacal silver solution. It gives a strong
Schiff test with reduced fuchsin at room temperature. It consumes one
mole of potassium bisulfite per mole of nonglycosidic sugar phosphate and
is stable toward periodic acid. It yields an oxime, a 2,4-dinitrophenyl-
hydrazone, and a benzyl mercaptal which is cleaved by alkali.

Apurinic acid preparations may be obtained from a large variety of
different deoxypentose nucleic acids. They have proved convenient inter-
mediates in the search for trace pyrimidines in nucleic acids.'"^'*^'"" In
addition, two points should be stressed, (a) Apurinic acid, while still a
polynucleotide, has an average molecular weight (estimated at 15,000^^^)
that is much smaller than that of the nucleic acid from which it is pre-
pared, (b) Apurinic acid is, in its composition, characteristic of the particu-
lar deoxypentose nucleic acid that served as the starting material. The use
of preparations of this type in structural studies will be considered later.

"« C. Tamm and E. Chargaff, /. Biol. Chem. 203, 689 (1953).

isolation and composition of deoxypentose nucleic acids 345

3. "Cores" (Limit Polynucleotides)

When a deoxyribonuclease acts on a deoxyribonucleic acid, we are dealing
with an enzyme of as yet unrecognized specificity attacking a polynucleo-
tide chain of as yet unknown sequence. It is a reaction whose study is
likely to spread more darkness than light. Moreover, the fact that deoxy-
ribonuclease is a phosphodiesterase of a closely circumscribed, as yet un-
definable, specificity"^ producing only a very small quantity of the ulti-
mate "monomers," the mononucleotides, in addition to a complex mixture
ranging from polynucleotides to dinucleotides,"® raises a problem of a
special kind. The enzymic attack on a substrate of the complexity of a
deoxypentose nucleic acid, which results in its partial cleavage, must go
through an intricate pattern: every break of the original molecule produces
substrates that are new and different; the enzyme must deal with kaleido-
scopic substrate changes. The attempt to solve this gigantic puzzle by
fitting the innumerable fragments. into a plausible sequence is doomed to
failure. Unless, however, deoxyribonuclease is specific only for the size,
and not the quality, of the oligonucleotide fragments which it is able to
cleave, in which case the first few random events would decide all subse-
quent ones, the order in which different pieces are detached, and the com-
position of those that are left behind, may serve as means of distinction
between different nucleic acids.

The study of the action of crystalline pancreatic deoxyribonuclease [com-
pare Schmidt, Chapter 15] on calf thymus nucleic acid has, in fact, shown
a characteristic trend in the composition of dialyzable digestion products
and the existence of polynucleotides, distinguished by greater resistance to
enzymic attack and characterized by greatly increased ratios of adenine
to guanine, thymine to cytosine, and purines to pyrimidines.^'--" Similar
studies have been carried out with nucleic acids from other sources."' •^^'*'
228-230 j^ j^g^y Q^^fy |-,g meutioned that the "cores" produced by the action
of two different deoxyribonucleases, those of pancreas and of germinating
barley, on the same nucleic acid differ markedly in their composition.-''

VI. Constituents of Deoxypentose Nucleic Acids

The chemistry of the ultimate breakdown products of the nucleic acids,
namely, the sugars and nitrogenous constituents, is discussed in Chapters
2 and 3; the nucleosides and nucleotides are considered in Chapter 4. In

"6 C. Tamm and E. Chargaff, Nahire 168. 916 (1951).

"6 R. L. Sinsheimer, /. Biol. Chem. 208, 445 (1954).

2" S. Zamenhof and E. Chargaff, J. Biol. Chem. 187, 1 (1950).

228 W. G. Overend and M. Webb, J. Chem. Soc. 1950, 2746.

«9 K. Moldave and C. Heidelberger, J. Am. Chem. Soc. 76, 679 (1954).

"" L. L. Uzman and C. Desoer, Arch. Biochem. and Biophtjs. 48, 63 (1954).

"' G. Brawerman and E. Chargaff, J. Biol. Chem. 210, 445 (1954).

346 ERWIN CHARGAFF

this brief section I shall limit myself to those aspects that are of importance
for the succeeding discussion of composition and structure.

1. Sugar

As regards the type of sugar occurring in deoxypentose nucleic acids,
very much has been taken for granted owing no doubt to the great difficulty
of isolating the free deoxysugars. Only for the nucleosides and nucleotides
obtainable from thymus nucleic acid, and strictly speaking, only for the
purine derivatives, is the chemical evidence decisive. There exist, however,
no indications that of the possible 2-deoxypentoses, 2-deoxyribose and
2-deoxyxylose, any but 2-deoxy-D-ribose occur in nucleic acids, though
proof by actual isolation has been scanty (see Chapter 2); nor has an
authentic case of the coexistence of both pentoses and deoxypentoses in the
same nucleic acid been described. The lack of information on the optical
activity of nucleosides and nucleotides isolated from other nucleic acids
than that of calf thymus is regrettable.

The chromatographic investigation of the sugars present in different
deoxypentose nucleic acids, or at any rate in their purine nucleotide moie-
ties, has, however, been pursued on a quite extensive scale in the past few
years. It is based on the enzymic degradation of the nucleic acid to the
nucleoside stage, on the release of the sugar by controlled heating of the
nucleosides at pH 1.5, and the chromatographic comparison, in at least
three different solvent systems, of the unknown sugar with 2-deoxyribose
liberated under identical conditions from calf thymus deoxyribonucleic
acid.'^'*'^^^ In all instances examined heretofore, the sugars were identical
and agreed in their chromatographic behavior with that of 2-deoxyribose.
This tentative identification has been performed with deoxypentose nucleic
acid preparations from the following sources: ox (thymus, spleen),"'' sheep
(thymus, liver), "^ pig (thymus, liver), "^ man (thymus, liver, carcinoma),"^
salmon sperm," sperm from four sea urchin genera,'"* yeast, and avian
tubercle bacilli.'^' 2-Deoxyribose has also been demonstrated in apurinic
acid.224

2. Nitrogenous Constituents

With very few exceptions, the bulk of the nitrogenous constituents (aboul
98 to 99%) of all deoxypentose nucleic acids is composed of two purines
adenine and guanine, and two pyrimidines, cytosine and thymine. The
principal exceptions are the deoxypentose nucleic acids of the bacterio-
phages T2, T4, and T6 of E. coli, in which 5-hydroxymethylcytosine takes
the place of cytosine,'*'-^' and the nucleic acid of wheat germ, in which

"2 E. Chargaff, C. Levine, and C. Green, J. Biol. Chem. 175, 67 (1948).
2" G. R. Wyatt and S. S. Cohen, Nature 170, 1072 (1952).

ISOLATION AND COMPOSITION OF DEOXYPENTOSE NUCLEIC ACIDS 347

about one-quarter of the cytosine is replaced by 5-methylcytosine.^^''-^'^^^
The latter has been recognized by Wyatt*-^ '^'^ '^^^ as occurring in traces in
several deoxypentose nucleic acids of animal origin (compare also reports
of other investigators'"* •"'•^^''), after an older claim of its being a constitu-
ent of the nucleic acid of tubercle bacilli^^* could not be confirmed.'"' The
chemistry of these purines and pyrimidines is discussed in Chapter 3 and
that of the corresponding nucleosides and nucleotides in Chapter 4. Certain
bacteria and bacteriophages appear to be able to incorporate, from the
medium, unusual pyrimidines, such as 5-bromouracil and 5-iodouracil, in
their deoxypentose nucleic acids in replacement of part of the thymine i^"'^^**
but these substances can hardly qualify as normal constituents.

3. Unidentified Constituents

There is little sense in attempting a catalogue of our ignorance. As will
appear later, most of the nitrogen and phosphorus of those deoxypentose
nucleic acids that have been investigated in detail has been accounted for;
and the remaining 1 to 3 % may well be attributed to analytical imperfec-
tions. But only very few nucleic acids have been studied at all, and even
they may contain as yet unrecognized nucleotide satellites. What must be
avoided at the present stage is the grand sweep of induction that extracts
general laws from single observations on ill-defined, and sometimes not
even specified, preparations (often known as "commercial sperm nucleic
acid").

It is not impossible that a careful search for novel nucleosides or nucleo-
tides by means of ion-exchange chromatography (see Chapter 6) would be
more rewarding than the separation of the free nitrogenous constituents by
chromatography on filter paper (see Chapter 7). Studies of this type have
been carried out on a few deoxypentose nucleic acids ;^^^-^''^ they have
yielded no unexpected results. There exists, however, a body of analytical
observations concerning differences in the hydrolysis behavior of different
nucleic acids that may provide some clues.'""'* '^'''"'■^^' The most in-
teresting ob.servation, perhaps, concerns the behavior of the adenine nucleo-
tides of the deoxypentose nucleic acid of wheat germ."'* As judged from the
extent of its liberation by various hydrolyzing agents, adenine seems to
occur in two types of linkage. The predominant type is broken with equal

"" G. R. Wyatt, Nature 166, 237 (1950).

"6 G. R. Wyatt, Biochem. J. 48, 581 (1951).

"6 T. B. Johnson and R. D. Coghill, J. Am. Chem. Soc. 47, 2838 (1925).

"7 F. Weygand, A. Wacker, and H. Dellweg, Z. Nalurforsch. 7b, 19 (1952).

"» D. B. Dunn and J. D. Smith, Biochem. J., 58, X (1954).

"9 E. Volkin, J. X. Khym, and W. E. Cohn, J. Am. Chem. Soc. 73, 1533 (1951).

"« R. L. Sinsheimer and J. F. Koerner, J. Biol. Chem. 198, 293 (1952).

"' R. O. Hurst, A. M. Marko, and G. C. Butler, J. Biol. Chem. 204, 847 (1953).

348

ERWIN CHARGAFF

TABLE IV

Liberation of Adenine from Sodium Deoxypentose Nucleate of
Wheat Germ and from Enzymically Produced "Cores"

Moles per 100 g. -atoms P

Hydrolysis conditions

Intact
DNA

19%
"core"

8%
"core"

N H2SO4 , 100°, 1 hr.
98% HCOOH, 175°, 2 hr.
7.5 N HCIO4 , 100°, 1 hr.

26.5
26.3

31.2
33.3
33.2

31.6
35.4

ease by A^ H2SO4, formic acid,^^^ and perchloric acid;^^* the other type, the
concentration of which is considerably increased in the nondiffusible "cores"
(see Section V.3), appears resistant to sulfuric acid, but not to the other
agents (see Table IV). In other words, a small proportion of the total
adenylic acid contained in this nucleic acid shows the resistance to cleavage
by mineral acid at 100° that is characteristic of the pyrimidine nucleotides.
Similar, though not equally striking, observations have been made in the
other instances cited above. It is not probable that they have to do with
the type of phosphate bridges linking the nucleosides in the deoxypentose
nucleic acids which presumably are 3': 5' in most cases, though exceptions
may occur. [See Chapter 12.] This stability is more likely connected with
the glycosidic bond (e.g., sugar at position 1 or 3 of the purine). But this
is, for the moment, only speculation.

VII. Composition of Deoxypentose Nucleic Acids

1. General

The study of the constitution and structure of a complicated cell con-
stituent of a high molecular weight, such as a protein, a polysaccharide or
a nucleic acid, goes through several well-defined stages. First, the substance
must be properly recognized and also distinguished from other compounds
with which it may share some, but not the decisive, properties. Next, all
its ultimate organic constituents — and it may comprise one variety or
many^ — ^have to be described in a qualitative fashion. As regards the nucleic
acids, no better account of these two stages will be found than that given
by Levene.^ Rut when quantity has to be translated into quality, when
discrimination has to be made not between substances that are palpably
different, but between those that are seemingly identical, the task becomes

2« E. Vischer and E. Chargaff, J. Biol. Chem. 176, 715 (1948).

ISOLATION AND COMPOSITION OF DEOXYPENTOSE NUCLEIC ACIDS 349

difficult, dull, controversial. The center begins to be surrounded by enor-
mous analytical slums: it is a world of decimals.

The problem of the recognition of identity and diversity in high-molecu-
lar cell constituents has been discussed before.'' Here, it will suffice to point
out that a chemical comparison with respect to identity or difference must
be based on the nature and the proportions of the constituents, on the
sequence in which these constituents are arranged in the molecule, and on
the type and the position of the linkages that hold them together. The
smaller the number of components of such a macromolecule is, the greater
is the difficulty of a decision. Deviations of the analytical results from
simple, integral proportions, of no importance for substances of small size,
become very significant when applying to compounds whose molecular
weights range in the millions; and variations in the proportions of their
several constituents often will provide the only proof of the occurrence of
different compounds.

The existence of significant chemical differences between deoxy pentose
nucleic acids of different cellular origin, or what has been called the chemical
specificity of nucleic acids,^ was discovered only a few years ago."^ Soon
after, it was possible to formulate most of the regularities that are recog-
nized today.'*'" -^' The steps that permitted this relatively ver^' rapid
progress were the following. After the development of partition chromatog-
raphy on filter paper and its qualitative application to amino acids'^^ it
became obvious that the high and specific absorption in the ultraviolet of
the purines and pyrimidines could form the basis of a quantitative ultra-
micromethod, if proper procedures for the hydrolysis of the nucleic acids
and for the complete separation of the hydrolysis products could be found.
Such procedures were indeed developed, 2"-''® and the elaboration of de-
tailed methods^^*"^^* was soon followed by their application to the analysis
of several nucleic acids.""* '"'^^^ The description of an arrangement per-
mitting the easy demonstration of the purine and pyrimidine spots^^^ and
the application of a similar, commercially available, ultraviolet lamp""
facilitated the performance of analyses. A photographic method useful for
the preservation of permanent records was also described.^*' "^

2« E. Chargaff, J. Cellular Camp. Physiol. 38, Suppl. 1, 41 (1951).

2« R. Consden, A. H. Gordon, and A. J. P. Martin, Biochem. J. 38, 224 (1944).

2" E. Vischer and E. Chargaff, J. Biol. Chem. 168, 781 (1947).

"» E. Vischer and E. Chargaff, Federation Proc. 7, 197 (1948).

2" R. D. Hotchkiss, J. Biol. Chem. 175, 315 (1948).

2« E. Vischer and E. Chargaff, J. Biol. Chem. 176, 703 (1948).

"9 E. R. Holiday and E. A. Johnson, Nature 163, 216 (1949).

2" E. Chargaff, B. Magasanilc, R. Doniger, and E. Vischer, J. Am. Chem. Sac. 71,

1513 (1949).
"' R. Markham and J. D. Smith, Nature 163, 250 (1949).
2" R. Markham and J. D. Smith, Biochem,. J. 45, 294 (1949).

350 erwin chargaff

2. Procedures

The principal analytical methods make use of chromatography of the
nitrogenous constituents on filter paper; they are discussed in Chapter 7.
For the subsequent spectrophotometric operations Chapter 14 should also
be consulted. In addition, very occasional use has been made of the chroma-
tography of the free bases on a starch column"* and of the separation of
enzymically liberated mononucleotides on ion exchangers.^'**'' -^^ [Compare
Chapter 6.] In contrast to their usefulness in other aspects, color reactions
have found little direct application in studies on composition. [Compare
Chapter 9.] The isotope dilution method does not appear to have been
much used either."^

Requirements with respect to the purity of the specimens have been
discussed in Sections III and IV. The most troublesome impurity is pentose
nucleic acid. The analytical results on preparations containing more than
3 % of this contaminant command little confidence; for careful comparisons
this impurity should be reduced to less than 1.5%. Its removal has been
discussed in Section lU.2.h. Proteins and polysaccharides interfere less,
but must be eliminated to an extent permitting the nucleic acid specimen
to have a P content of more than 8%.

The procedures used in the laboratory of this writer and some of the modi-
fications introduced in the course of time have been described in several
publications cited before." •"^■"*'2''8

3. Distribution of Purines and Pyrimidines

a. Grouping of Deoxypentose Nucleic Acids and Presentation of Results

When the composition of many specimens of deoxypentose nucleic acid
from different cellular sources is compared, a very striking feature emerges,
as was pointed out some years ago: two principal groups can be distin-
guished, namely the "AT type," in which adenine and thymine predominate,
and the "GC type," in which guanine and cytosine are the major constitu-
uents.'^--'" In addition, an intermediate group was discovered in E. coli
which is characterized by the presence of almost equimolar proportions of
the four components^^^'^^^'^" All total deoxypentose nucleic acid prepara-
tions from animal sources described up to this time belong to the AT type,
which seems to be much more frequent .than the others. The GC type has
been encountered in several microorganisms and in some insect viruses.
Consequently, the molar ratios of adenine to guanine, of thymine to cyto-
sine, and of adenine + thymine to guanine + cytosine will be above 1 in
the nucleic acids of the AT type and below 1 in those of the GC type. This
will appear with particular clarity when the composition of microbial de-

2" R. Abrams, Arch. Biochem. 30, 44 (1951).

ISOLATION AND COMPOSITION OF DEOXYPENTOSE NUCLEIC ACIDS 351

oxypentose nucleic acids is compared; in Table XII, which will be dis-
cussed later, Nos. 1-4 are AT, Nos. 5-9 belong to the intermediate group,
Nos. 10-16 to the GC type.

The subsequent account will demonstrate that different nucleic acid
specimens of the same origin show a remarkable identity in composition;
but this may be regarded, as was conjectured very early, "^ as "a statistical
expression of the unchanged state of the cell," for it is possible to obtain,
by the fractionation of nearly all deoxypentose nucleates, preparations of
a composition representative of all three types mentioned above. (See
Section VIII.)

The comparison of analytical results obtained by different workers using
procedures that are almost never identical and often radically different
suffers from a great deal of uncertainty. If, moreover, a large number of
analyses of several well-characterized and purified preparations are re-
ported in one instance, and, in another, single analyses of ill-described
specimens, a dilemma between encyclopedic comprehensiveness and over-
judicious exclusiveness will occur. For analytical results to be acceptable
the following minimum requirements should be fulfilled. At least two com-
pletely characterized preparations isolated separately from two independ-
ently collected specimens of the starting material should be analyzed in at
least two different hydrolysis experiments, and a full balance of recovered
components on a molar basis should be drawn. It is surprising how seldom
these modest wishes are met.

The compilation of data presented in Tables V to XIII comprises the
major part, if not all, of the analytical results with deoxypentose nucleic
acids. In general, only analyses performed on isolated nucleic acid speci-
mens have been included, and several results had to be omitted because of
lack of pertinent information. The total recovery has been indicated, when-
ever it was given in the literature; but, in order to facilitate comparison, all
molar proportions have been corrected for a 100% recovery in terms of
nucleic acid P.

b. Composition Differences and Similarities

After the existence of chemical differences between deoxypentose nucleic
acids of different origin was first suggested,*"^'"' -^'' agreement was not
general. This is not surprising, as in many cases very extensive work is re-
quired, in order to prove a divergent composition. But there is little doubt
at present that there exist many different deoxypentose nucleic acids and
that their number may be even much larger than would be revealed by
merely analytical differentiation. While, therefore, the question whether
there exists more than one deoxypentose nucleic acid can be answered by

2" E. Chargaff, S. Zamenhof, and C. Green, Nature 165, 756 (1950).

352

ERWIN CHARGAFF

TABLE V
Distribution of Purines and Pyrimidines in Deoxyribonucleic

Acids from Different Organs of Four Mammalian Genera*

Proportions in moles of nitrogenous constituent per 100 g. -atoms of

phosphorus in hydrolysate, corrected for a 100% recovery,

with their standard errors.

Genus Organ

Ox

Sheep

Pig

Man

Thymus
Liver

Thymus
Liver

Thymus
Liver
Spleen
Thyroid

Thymus
Liver

Ade-

Gua-

Cyt

o-

Thy-

No. of
prepa-
rations

nine

nine

sine

mine

29.0

1^
o

u

u

T3

0.2

21.2

o

u
<U

0.2

21.2

O

-6
■^
m

0.2

28.5

o

u

0.2

21

2

28.8

0.3

21.0

0.6

21.1

0.2

29.0

0.3

2

29.3

0.2

21.4

0.3

21.0

0.2

28.3

0.2

2

29.3

0.2

20.7

0.2

20.8

0.3

29.2

0.3

2

30.0

0.2

20.4

0.3

20.7

0.2

28.9

0.4

2

29.4

0.3

20.5

0.4

20.5

0.3

29.7

0.4

1

29.6

0.3

20.4

0.9

20.8

0.2

29.2

0.7

1

30.0

0.4

20.8

0.4

20.7

0.3

28.5

0.7

1

30.9

0.6

19.9

0.5

19.8

0.7

29.4

0.4

4

30.3

0.2

19.5

0.2

19.9

0.2

30.3

0.2

Average

total
recovery
Mole %

93.3
94.1

94.3
89.9

93.9
95.3
96.0
97.4

94.6
95.4

Ratio of

adenine -|-

thymine to

guanine +

cytosine

1.36
1.37

1.36
1.41

1.43
1.44
1.43
1.41

1.52
1.54

• Taken from E. Chargaff and R. Lipshitz, J. Am. Chem. Soc. 75, 3658 (1953).

pointing to the information collected in tabular "form in this section, a de-
cision whether different tissues of the same host yield identical or different
nucleic acid specimens is more difficult and will, perhaps, not be definite
before the analytical studies can be supplemented by other tests, for in-
stance, a completely reproducible fractionation process. For the moment,
it would appear that differences in the composition of preparations from
different organs, if they exist at all, are of very doubtful significance, al-
though some contrary findings have been reported following the administra-
tion of steroid sex hormones.-^*

The results of a comparative study of the composition of many prepara-
tions from different organs of ox, sheep, pig, and man'" are summarized in
Tables V and VI. In no case were significant differences in composition
found when the deoxyribonucleic acids isolated from the thymus and liver
of the same genus were compared statistically, nor were differences found
between human normal and carcinomatous liver. In Table VI the signifi-
cance of differences between different genera is considered. It will be seen

2" N. L Gold and S. H. Sturgis, J. Biol. Chem. 206, 51 (1954).

ISOLATION AND COMPOSITION OF DEOXYPENTOSE NUCLEIC ACIDS

353

^

J -o

S3

c3
en

U

>>

o

tn

fl)

(-1

>.

s

•^

M

o

c

r,

X

• —

<u

tr^ <i

w

f6

cS

(_

hJ

.

0)

CQ

iJ

bU

T1

7;

^

^^

•73

CB

^^

G

fc-

aj

a

CO

c

'S

u

■3

-C

«

-^

*^

-w

^

^

u
u

o

^

%

H

-1

u ^

o

So

-H I + +

Sheep
and
man

+ + + +
+ + + +

Sheep
and
pig

1 1 1 1

^ =3 S

+ + + +
+ + + +

o ^-a

+ -H -H 1

4S

IM CC (M <N

2-2-2-2-

Tf ;0 0> — I

Q Ol C5 O
CO ^ ^ CC

(M « (M CO

o o d o

GO •* t^ —

05 d d Si

f« (M C<l ?1

<M CO (N (M

o d d d

C<i a c<> c<\

2-2 2 2

O (M (M t^

o A

~" V

S a,

* 'i

T3 +

I 6§
V

a a

o -5 ~"
-3 -2 +
^^ +

lis

§■§ V

a S a,

>- C VS

a

a c >,

o

rr °

asi

« S c c

OJ

o S

E 0) £

c 'c g S

J2

11

OH-«

OJ c3 ^ >,

C

3

«< O U H

i^

^

a.'

354

ERWIN CHARGAFF

TABLE VII
Selection of Data on Purine and Pyrimidine Contents of Sodium

Deoxyribonucleate Preparations from Calf Thymus

Proportions in moles of nitrogenous constituent per 100 g. -atoms of P in

hydrolysate, corrected for a 100% recovery.*

No.

Adenine

Gua-
nine

Cytosinet

5-

Methyl-
cytosine

Thymine

A+T
G + C^
+MC

Actual

re-
covery

Re-
marks!

Ref.

1

28.0

22.0

19.9

30.1

1.35

88.4

?

a

2

27.3

22.7

21.6

28.4

1.26

93.7

a

3

27.9

23.8

19.9

28.3

1.29

98.8

b

4

27.4

22.4

20.7

29.5

1.32

97.0

?

b

5

28.0

23.5

20.4

28.1

1.28

100.2

b

6

29.2

20.8

20.8-

29.2

1.40

96

c

7

28.2

21.5

21.2

1.3

27.8

1.27

90

d

8

28.9

22.2

21.1

27.8

1.31

90

t

9

29.2

21.9

21.9

27.1

1.29

96

?

e

10

29.2

20.2

21.3

29.2

1.41

89

«

11§

20.0

1.3

28.7

92

«

12

30.1

21.5

20.4

28.0

1.39

93

?

/

13

28.4

21.1

22.1

28.4

1.31

95

/

14

27.6

21.5

21.2

1.9

27.9

1.24

89.7

9

15

28.0

20.9

21.4

1.9

27.8

1.26

91.0

B

16

28.7

21.7

20.7

1.7

27.2

1.27

?

h

17

27.1

23.2

21.8

27.9

1.22

96.5

'■

18

27.4

21.4

21.0

1.1

29.1

1.30

99.6

?

i

* When more than one value was reported for the same preparation, the average has been computed.
In Preparations 1 and 2, the first complete analyses reported in the literature, the purines and pyrimidines
were estimated in separate hydrolysates. For this reason, the total purines and the total pyrimidines were
adjusted individually to a 50% recovery.

t Where no methylcytosine figures are given, the values for cytosine include methylcytosine in most
cases. For the calculation of ratios cytosine and methylcytosine were taken together. The average of all ratios
reported here is 1.30. A = adenine; G = guanine; C = cytosine; MC = methylcytosine; T = thymine.

X A question mark indicates that the total recovery was below 85 mole %, or that the ratio of adenine to
thymine was below 0.95 or above 1.05. It has also been used when no recovery was reported or when other
factors rendered the results doubtful.

§ This preparation was analyzed as the apurinic acid from calf thymus sodium deoxyribonucleate.

References (Unless pointed out otherwise, the analytical methods were based on filter-paper chromatography.)

" E. Chargaff et al. J. Biol. Chem. m, 405 (1949).

* M. M. Daly et al., J. Gen. Physiol. 33, 497 (1950); chromatography of free bases on starch columns.
" A. Marshak and H. J. Vogel, J. Biol. Chem. 189, 597 (1951).

'' G. R. Wyatt, Biochem. J. 48, 584 (1951).
' C. Tamm et al., J. Biol. Chem. 195, 49 (1952).
^ C. Tamm et al., J. Biol. Chem. 199, 313 (1952).
0 S. G. Laland et al., J. Chem. Soc. 1952, 3224.

* R. L. Sinsheimer and J. F. Koerner, J. Biol. Chem. 198, 293 (1952); ion-exchange chromatography of
mononucleotides.

"• R. O. Hurst et al., J. Biol. Chem. 204, 847 (1963); ion-exchange chromatography of mononucleotides.

ISOLATION AND COMPOSITION OF DEOXYPENTOSE NUCLEIC ACIDS

355

that the adenine and thymine contents increase, and the guanine and
cytosine contents decrease, in the following order: ox, sheep, pig, man.
The differences between two adjoining columns are, in general, not signifi-
cant or of doubtful significance, when analyzed statistically ; but the prob-
ability of identity becomes very small, when the specimens from bovine
and human tissues are compared.

Table V includes information on the results of an analysis of 21 different
preparations of calf thymus sodium deoxyribonucleate. Many individual
analy.ses of calf thymus preparations are compiled in Table VII. It will be
seen that the agreement is, on the whole, better than would have been
anticipated. Table VIII collects data on other bovine tissues, Tables IX
and X on the nucleic acids from mammalian and other animal sources,
respectively.

The high content of 5-methylcytosine in the sodium deoxypentose nucle-

TABLE VIII
Selection of D.\ta on Purine and Pyrimidine Content.s of Sodiu.m

Deoxypentose Nucleate Preparations from Bovine Tissues

Proportions in moles of nitrogenous constituent per 100 g. -atoms of P in

hydrolysate, corrected for a 100% recovery.*

No.

Tissue

Ade-
nine

Gua-
nine

Cyto-
sine

5-

Methyl-
cytosine

Thy-
mine

A+T
0+0+"^
MC

Actual

re-
covery

Re-
marks

Ref.

1

Spleen

27.9

22.1

20.7

29.3

1.34

87.9

a

2

Spleen

28.2

21.2

21.0

1.3

28.2

1.30

90

6

3

Spleen

28.6

20.9

20.7

1.4

28.3

1.32

88.2

e

4

Spleen

27.9

22.7

20.8

1.3

27.3

1.23

100

d

5

Spleen

27.7

22.1

21.8

28.4

1.28

98.0

e

6

Liver

28.8

21.0

21.1

29.0

1.37

94.1

/

7

Liver

27.4

22.5

21.9

28.2

1.25

95.6

'

8

Pancreas

27.8

21.9

21.7

28.5

1.29

97.8

«

9

Kidney

28.3

22.6

20.9

28.2

1.30

96.9

0

10

Testes

27.0

22.9

22.3

27.8

1.21

96.0

«

11

Sperm

28.7

22.2

20.7

1.3

27.2

1.26

90

?

t>

• Compare footnotes in Table VII.

t The average of all ratios reported here is 1.29.

Referevces

" E. Chargaff et al., J. Biol. Chem. 177, 405 (1949).
* G. R. Wyatt, Biochem. J. 48, 584 (1951).
' S. G. Laland et at.. J. Chem. Svc. 1952, 3224.
■^ G. R. Wyatt and S. S. Cohen, Biochem. J. 55, 774 (1953).

' R. O. Hur.st etal., J. Biol. Chem. 204, 847 (1953); ion-exchange chromatography of mononucleotides.
/£. Chargaff and R. Lipshitz, J. Am. Chem. Soc. 75, 3658 (1953).
M. M. Daly et al., J. Gen. Physiol. 33, 497 (1950); chromatography of free bases on starch columns.

356

ERWIN CHARGAFF

TABLE IX

Selection of Data on Purine and Pyrimidine Contents of Sodium

Deoxypentose Nucleate Preparations from Different

Tissues of Several Mammalian Genera

Proportions in moles of nitrogenous constituent per 100 g. -atoms of P

in hydrolysate, corrected for a 100% recovery.*

d

Genus

Tissue

Ade-
nine

Gua-
nine

Cyto-
sine

Thy-
mine

A-fT

G+C

Actual

re-
covery

Re-
marks

Ref.

1

Mouse

Sarcoma

29.7

21.9

22.8

25.6

1.24

80

?

a

2

Rat

Bone
marrow

28.6

21.4

20. 4t

28.4

1.33

?

b

3

Sheep

Thymus

29.3

21.4

21.0

28.3

1.36

94.3

c

4

Shieep

Liver

29.3

20.7

20.8

29.2

1.41

89.9

e

5

Sheep

Spleen

28.0

22.3

21.1

28.6

1.30

92.9

d

6

Sheep

Sperm

28.8

22.0

21. OJ

27.2

1.27

88.5

?

b

7

Pig

Thymus

30.9

19.9

19.8

29.4

1.52

94.6

c

8

Pig

Liver

29.4

20.5

20.5

29.7

1.44

95.3

'

9

Pig

Spleen

29.6

20.4

20.8

29.2

1.43

96.0

e

10

Pig

Thyroid

30.0

20.8

20.7

28.5

1.41

97.4

?

e

11

Man

Thymus

29.8

20.2

18.2

31.8

1.60

91

?

e

12

Man

Thymus

30.9

19.9

19.8

29.4

1.52

94.6

c

13

Man

Liver

30.3

19.5

19.9

30.3

1.53

95.4

c

14

Man

Spleen

29.2

21.0

20.4

29.4

1.42

96.6

/

15

Man

Sperm

30.9

19.1

18.4

31.6

1.67

96

e

16

Man

Sperm

30.7

19.3

18.8

31.2

1.62

92

e

17

Horse

Spleen

29.6

22.9

20.1

27.5

1.33

95.7

?

d

* Compare footnotes in Table VII.

t In addition, 1.1 mole % of S-methylcytosine was found,
t In addition, 1.0 mole % of 5-methylcytosine was found.

References

" S. G. Laland et al., J. Chem. Soc. 1952, 3224.

* G. R. Wyatt, Bicchem. J. 48, 584 (1951).

'E. Chargaff and R. Lipshitz, J. Am. Chem. Soc. 75, 3658 (1953).

M. M. Daly et al., J. Gen. Phyaiol. 33, 497 (1950); chromatography of free bases on starch colymns.
' E. ChargafT el al., Nature 165, 756 (1950); correction as for Preparations 1 and 2 in Table VII.
' R. O. Hurst etal., J. Biol. Chem. 204, 847 (1953); ion-exchange chromatography of mononucleotides.

ate of wheat germ has been mentioned before. Data on the purine and py-
rimidine composition of different preparations of. this nucleic acid analyzed
in four laboratories are presented in Table XI ; the agreement is satisfactory.
As has been mentioned before, little information on other plant nucleic
acids is available.

What is known about the composition of microbial deoxypentose nucleic
acids and some related compounds from rickettsiae and viruses is collected
in Table XII. In view of the difficulty of securing sufficient material it is

ISOLATION AND COMPOSITION OF DEOXYPENTOSE NUCLEIC ACIDS

357

TABLE X

Purine and Pyrimidine Contents of Sodium Deoxypentose

Nucleates of Various Animals (from Birds Downward)

Proportions in moles of nitrogenous constituent per 100 g. -atoms of P

in hydrolysate, corrected for a 100% recovery.*

No.

Animal

Tissue

.£
c

T3
<

28.8

c

'a
a

3

O
20.5

c

O

21.5

1 <D

a

>.

H

A+T
G+C
+MC

<

93.7

03

s

t

1

Hen

Erythro-

29.2

1.38

a

2

Hen

cytes
Erythro-

28.0

22.0

21.6

28.4

1.29

97.4

b

3

Hen

cytes
Egg
white

29.7

21.5

21.3

27.5

1.34

?

^

4

Turtle

Erythro-

28.7

22.0

21.3

27.9

1.31

104.1

a

5

Shad

cytes
Testes

28.4

21.8

20.5

29.3

1.36

92.9

a

6

Herring

Testes

27.2

19.3

22.3

2.7

28.6

1.26

92.0

d

7

Herring

Testes

27.9

19.5

21.5

2.8

28.2

1.28

89.5

d

8

Herring

Sperm

27.8

22.2

20.7

1.9

27.5

1.23

91

'

9

Trout

Sperm

29.8

22.5

20.2

27.5

1.34

98.6

?

"

10

Salmon

Sperm

29.7

20.8

20.4

29.1

1.43

94.2

!

11

Locusta rnigratoria

Whole

29.3

20.5

20.7

0.2

29.3

1.41

?

'

12

Arbacia punctulata

Sperm

28.4

19.5

19.3

32.8

1.58

93.9

?

"

13

Arbacia lixula

Sperm

31.2

19.1

19.2

30.5

1.61

94.2

s

14

Echinus esculenius

Sperm

30.9

19.4

18.4

1.8

29.4

1.52

92

e

15

Echinocardium
cordatum

Sperm

32.9

17.0

17.9

32.2

1.86

96.4

9

16

Psammechinus
miliaris

Sperm

32.6

17.8

17.8

31.9

1.81

94.0

a

17

Paracentrotus
livid us

Sperm

32.8

17.7

17.3

(l.l)t

32.1

1.85

94.7

a

• Compare footnotes in Table VII.

t Separate analyses indicate a content in 5-metliylcytosine corresponding to 6.6% of the cytosine value.

References

" M. M. Daly et at.. J. Gen. I'hyswl. 33, 497 (1950); chromatography of free bases on starch columns.

* R. O. Hurst et al., J. Biol. Chem. 204, 847 (1953); ion-exchange chromatography of mononucleotides.
" H. Fraenkel-Conrat and E. D. Ducay, Biochem. J. 49, XXXIX (1951).

^ S. G. Laland et al.. J. Chem. Soc. 1952, 3224.
' G. R. Wyatt, Biochem. J. 48, 584 (1951).
■^ E. Chargaff et al.. J. Biol. Chem. 192, 223 (1951).
B E. Chargaff el al. , J. Biol. Chem. 195, 155 (1952).

not astonishing that some of the analyses are not of desirable quality, but
the trend is evident. It shows a very wide range of composition differences
going from the extreme AT type (yeast) to the extreme GC type (tubercle
bacilli). It is remarkable that all specimens from acid-fast bacteria (Nos.

358

ERWIN CHARGAFF

TABLE XI

Purine and Pyrimidine Contents of Sodium Deoxypentose Nucleate

Preparations from Wheat Germ

Proportions in moles of nitrogenous constituent per 100 g. -atoms of P

in hydrolysate, corrected for a 100% recovery.*

5-

A+T

Actual

Re-
marks

No.

Adenine

Guanine

Cytosine

Methyl-
cytosine

Thymine

G-fC

-fMC

re-
covery

Ref.

1

26.5

23.5

17.2

5.8

27.0

1.15

84.5

a

2

27.3

22.7

16.8

6.0

27.1

1.19

96.2

b

3

27.1

20.2

19.6

5.7

27.4

1.20

83.0

?

e

4

26.8

23.2

16.7

5.3

28.0

1.21

99.5

d

* Compare footnotes in Table VII.

References

°- G. R. Wyatt. Biochem. J. 48, 584 (1951).

^ G. Brawerman and E. Chargaff, J. Am. Chem. Soc. 73, 4052 (1951).
" S. G. Laland et al., J. Chem. Soc. 1952, 3224.
R. O. Hurst et al., J. Biol. Chem. 204, 847 (1953); ion-exchange chromatography of mononucleotides.

12-16 in Table XII) belong to this type. Even more remarkable, perhaps,
is the good agreement in the analytical results on 5 strains of E. coli (Nos.
5-9). The composition of the T2, T4, and T6 phages, based on the work of
Wyatt and Cohen, '^^ is shown in Table XIII. Attention should also be
drawn to a study of the composition of a series of insect viruses in which,
again, all three types of deoxypentose nucleic acid appear to have been
encountered.^*^

It should be mentioned that only in a few cases a statistical interpretation
of the significance of the analytical findings has been attempted."' ^'^^- "^' ^^'
The conclusions with respect to the mammahan nucleic acids have been
mentioned before. In a similar study of the nucleic acids of four different
sea urchin genera'"^ (compare also the work on the effect of these prepara-
tions on developing sea urchin eggs^*®) it was concluded that the differences
in the composition of the preparations from Echinocardium, Psammechinus,
and Paracentrotus (Nos. 15-17 in Table X) were not sufficient to permit a
distinction, but that the preparations from Arbacia lixula (No. 13) differed
significantly from the others.

VIII. Fractionation of Deoxypentose Nucleic Acids

1. Gener.\l

Some of the criteria on which a decision on the difference or identity of
macromolecules of an ostensibly identical composition must be based have

"6 S. Horstadius, I. J. Lorch, and E. ChargafT, Exptl. Cell Research 6, 440 (1954).

ISOLATION AND COMPOSITION OF DEOXYPENTOSE NUCLEIC ACIDS

359

TABLE XII

Purine and Pyrimidine Contents of Sodium Deoxypentose Nucleate

Preparations from Microorganisms, etc.

Proportions in moles of nitrogenous constituent per 100 g. -atoms of P

in hydrolysate, corrected for a 100% recovery*

No.

Organism

Ade-
nine

Gua-
nine

Cyto-
sine

Thy-
mine

A+T
G+C

Actual

re-
covery

Re-
marks

Ref.

1

Yeast

31.7

18.3

17.4

32.6

1.80

74.8

a

2

Yeast

31.3

18.7

17.1

32.9

1.79

92.0

b

3

Pneumococcus type
III

29.8

20.5

18.0

31.6

1.59

92.5

c

4

Hemophilus influenzae

31.9

18.2

19.6

30.2

1.64

92.7

d

5

type C
E. coli (mutant B/r)

22.5

24.5

25.8

27.2

0.99

e

6

E. coli (mutant B/r)

23.3

23.6

25.6

27.6

1.03

'

7

E. coli (K-12)

26.0

24.9

25.2

23.9

1.00

94.9

f

8

E. coli (UQ)

25.6

25.0

25.5

23.9

0.98

93.7

f

9

E. coli (thymineless)

25.4

24.1

25.7

24.8

1.01

89.2

f

10

Serratia marcescens

20.7

27.2

31.9

20.1

0.69

87.0

d

11

B. Schatz

19.9

29.1

32.3

18.6

0.63

86.9

d

12

Mb. tuberculosis
(human)

18.0

28.5

33.5

20.0

0.61

'

13

Mb. tuberculosis
(human)

19.3

28.2

34.9

17.5

0.58

67.6

a

14

Mb. tuberculosis
(bovine)

17.8

29.3

33.8

19.0

0.58

e

15

Mb. tuberculosis
(avian)

15.1

34.9

35.4

14.6

0.42

76.3

"

16

Mb. phlei

18.0

31.6

34.8

15.5

0.50

79.0

a

17

Rickettsia burneti

29.5

22.5

22.0

26.0

1.25

96

h

18

Rickettsia prowazeki

35.7

17.1

15.4

31.8

2.08

96

h

19

E. coli phage T5

30.3

19.5

19.5

30.8

1.57

93

i

20

Vaccinia virus

29.5

20.6

20.0

29.9

1.46

100

'■

• Compare footnotes in Table VII.

References

" E. Vischer et al.. J. Biol. Chem. 177, 429 (1949).

* S. Zamenhof and E. Chargaff, J. Biol. Chem. 187, 1 (1950).

' M. M. Daly el al., J. Gen. Physiol. 33, 497 (1950); chromatography of free bases on starch columns.

"* S. Zamenhof et al., Biochim. et Biophys. Acta 9, 402 (1952).

' J. D. Smith and G. R. Wyatt, Biochem. J. 49, 144 (1951).

•^ B. Gandelman et al., Biochim. et Biophys. Acta 9, 399 (1952).

" S. G. Laland et al., J. Chem. Soc. 1952, 3224.

^ G. R. Wyatt and S. S. Cohen, Nature 170, 846 (1952).

* G. R. Wyatt and S. S. Cohen. Biochem. J. 55, 774 (1953).

360

ERWIN CHARGAFF

TABLE XIII

Purine and Pyrimidine Contents of Sodium Deoxypentose Nucleate

Preparations from E. coli Phages T2, T4, and T6

Proportions in moles of nitrogenous constituent per 100 g. -atoms of P

in hydrolysate, corrected for a 100% recovery.*

No.

Material

Adenine

Guanine

5-

Hydroxy-

methyl-

cytosine

Thymine

A+T

Actual

re-
covery

Ref.

G+HMC

1

T2r+ DNA

32.5

18.2

16.7

32.6

1.86

97

a

2

T2r DNA

32.4

18.3

17.0

32.4

1.84

98

a

3

T6r+ DNA

32.5

18.3

16.7

32.5

1.86

99

a

4

T6r+ DNA

30.9

18.4

17.4

33.3

1.79

97

6

5

T2r+ virus

32.0

18.0

16.8

33.3

1.88

99

a

6

T2r virus

32.3

17.6

16.7

33.4

1.91

95

a

7

T4r+ virus

32.3

18.3

16.3

33.1

1.89

96

a

8

T4r virus

32.2

18.0

16.3

33.5

1.91

94

»

9

T6r+ virus

32.5

17.8

16.3

33.5

1.93

99

<»

10

T6r virus

32.3

17.7

16.6

33.4

1.91

88

• Compare footnotes in Table VII. HMC = 5-hydroxymethylcytosine.

References

" G. R. Wyatt and S. S. Cohen, Biochem. J. 55, 774 (1953).
*C. F. Crampton et al., J. Biol. Chem. 211, 125 (1954).

been pointed out in the preceding section. Strictly speaking, the vaUdity
of many of the methods of classical organic chemistry ends with the mixed
melting point. Beyond, it will often be a matter of preference or utility
what to call identical and what different. If a compound is assigned a
mainly mechanical function, its composition, let alone the sequence of its
constituents, will appear of little importance. The function of a bag is to
hold, of a trestle to support; what they are made of is of no consequence.
The biochemical literature, especially that dealing with the proteins and
polysaccharides, abounds in falsely generic terms which a profounder in-
sight will undoubtedly resolve into many different individuals. Our science,
so drunk with dynamics, is slowly learning to pay attention to the motions
of the immovable.

For this reason, attention began to be paid to the question of difference
only when specific biological functions were assigned to the nucleic acids.
In the light of current conceptions, a preparation of deoxypentose nucleic
acid, presumably an important component of the genetic material, could be
regarded as consisting of many chemically different, though closely related,
individuals, the constant composition of the whole being a statistical reflec-
tion of the unchanging condition of the cell. That there is, in fact, little

ISOLATION AND COMPOSITION OF DEOXYPENTOSE NUCLEIC ACIDS 361

fluctuation in composition in different nucleic acid preparations of the same
origin has been stressed before in connection with the results presented in
Tables V and VI. But, as has been pointed out (Section V.l), the number of
possible permutations within a nucleic acid chain, even without a change in
composition, is truly enormous; and many fine points cannot be considered
at the present state of our knowledge, as, for instance, whether two nucleic
acids of the same composition, but differing in some details of sequence or
in their terminal nucleotides, are to be regarded as different entities. One
must, for the time being, rely on relatively massive changes in composition.

Although the possibility that a deoxypentose nucleic acid preparation
from a given species represents a mixture of many different individuals has
been discussed occasionally, it is only in recent times that successful frac-
tionation experiments have been recorded. Attempts to separate highly
polymerized preparations by fractional centrifugation^^'* or adsorption on
charcoaP^" were of no avail. Indications of heterogeneity, with respect to
their metabolic origin, of deoxypentose nucleate preparations from rat
tissues have been presented by Bendich et al}^'' The evidence rests on dif-
ferences in the incorporation of isotope observed with two nucleic acid
preparations differing in their solubility in physiol. saline. In the absence
of analytical information it is not possible to say whether the fractions
differed in composition, nor, in fact, whether they were pure nucleic acids.
In the light of the experiments of Crampton el al}° on the dissociation and
reassociation of nucleohistone, the material insoluble in physiol. saline prob-
ably represented a, perhaps partially degraded, protein nucleate.

A process best described as the fractional dissociation of a nucleoprotein
has permitted the separation of many deoxypentose nucleate preparations
into a series of fractions of divergent purine and pyrimidine contents. The
procedure was first applied to calf thymus nucleohistone"^ and later ex-
tended to the fractionation of other nucleic acids through their complexes
with histone,"^ globin,"^ or polylysine.-^" Subsequently, a related procedure
was described in a preliminary form in which fractionation was achieved by
the gradual elution of the sodium nucleate from a histone-kieselguhr col-
umn.^^^

2. Fraction.\l Dissociation of Nucleohistone or Protein Nucleates

The experiments are based on the observation^" that when nucleohistone,
prepared as described in Section II.3.a.(l), is treated with chloroform in
the absence of electrolytes the entire nucleic acid P is found in the resulting

2" A. Bendich, P. J. Russell, Jr., and G. B. Brown, J. Biol. Cheni. 203, 305 (1953).

"8 E. Chargaff, C. F. Crampton, and R. Lipshitz, Nature 172, 289 (1953).

"9 C. F. Crampton, R. Lipshitz, and E. Chargaff, J. Biol. Chem., 211, 125 (1954).

"° P. Spitnik, R. Lipshitz, and E. Chargaff, in preparation.

"1 G. L. Brown and M. Watson, Nature 172, 339 (1953).

362

ERWIN CHARGAFF

TABLE XIV

Dissociation of Nucleohistone in Salt Solutions, after a Single
Treatment with Chloroform

Nucleic acid P in supernatant, as % of initial

NaCl molarity

Exp. 1

Exp. 2

0.60

39.5

42.2

0.65

52.4

45.9

0.70

58.4

58.9

0.80

78.6

79.0

1.00

94.6

gel, whereas a similar treatment in salt solutions of increasing strength
results in the proportionate detachment of increasing quantities of nucleic
acid. This is illustrated in Table XIV,* which is based on the experiments of
Ciampton et a/.*"

For details of the procedures the original papers^^*' ^*' should be con-
sulted. Full data on two abbreviated fractionation runs with calf thymus
nucleohistone, in which only three fractions were collected, are given in
Table XV P'

Nucleohistone preparation N-NH was isolated, and precipitated at 0.15 M NaCl
concentration, as described in Section II.3.a.(l); preparation R-NH was isolated
similarly, but from a solution that had first been exposed to M NaCl. Solutions of
these preparations in distilled water (previously adjusted to pH 7), containing about
90 Mg- P per cc, were mixed with an equal volume of 1.3 M NaCI solution and, after
an interval of 30 minutes, treated in a high-speed mixer with one-half volume of
chloroform- octanol (9:1) for 2 minutes. From the supernatant fluids, obtained by
centrifugation (15 minutes, 2000 X g), the sodium nucleate fractions l-I and 2-1
(Table XV) were isolated as described in Section III.2.C. The sedimented gels were
reextracted, as described above, first with portions of 0.9 M NaCl equal to one-half
of the volume of nucleohistone solution (fractions l-II and 2-II) and then with 2.6 M
NaCl (fractions l-lll and 2-III). The results are shown in Table XV. The deoxy-
ribose content of the fractions varied from 95 to 100% of that of a standard prepara-
tion; protein was absent from all preparations.

The results of seven similar fractionation experiments are shown in Fig.
4 in graphic form. This diagram contrasts the molar sums of adenine and
thymine found in each significant fraction with the corresponding sums of
guanine and cytosine. It will be noticed that the gradual extraction of the
nucleohistone with salt solutions of increasing strength yields a series of
nucleic acid fractions with diminishing concentrations of guanine and
cytosine and rising concentrations of adenine and thymine. But what should
be emphasized is that in all fractions the equimolarity of each pair of con-

ISOLATION AND COMPOSITION OF DEOXYPENTOSE NUCLEIC ACIDS

363

TABLE XV

Fractional Dissociation of Nucleohistone Preparations;

Composition of DNA Fractions

Experiment No.

1

2

Preparation

N-NH

R-NH

Protein/P, weight ratio

11.5

12.7

DNA fraction No.

I

II

III

I

II

III

NaCl molarity

0.65

0.90

2.6

0.65

0.90

2.6

% of nucleohistone P

30.1

53.5

14.6

32.4

48.0

11.0

Total P, %

8.9

9.2

8.9

9.0

9.0

8.8

Extinction, «(P)

6800

6700

6850

6400

6500

6650

Viscosity, Vsp.i^)

525

470

420

525

450

410

Moles per 100 g. -atoms P*

Total recovery

93.0

96.0

93.3

94.1

98.9

94.1

Adenine

26.0

30.3

31.4

26.4

30.7

30.1

Guanine

23.7

19.8

18.6

23.8

19.3

19.9

Cvtosine

24.7

20.9

19.7

24.5

20.4

21.2

Thymine

25.6

29.0

30.3

25.3

29.6

28.8

Molar ratios

Adenine + thymine to guanine +

1.07

H6

1.61

1.07

1.52

US

cytosine

Adenine to thymine

1.02

1.04

1.04

1.04

1.04

1.04

Guanine to cvtosine

0.96

0.95

0.94

0.97

0.95

0.94

Purines to pyrimidines

0.99

1.00

1.00

1.01

1.00

1.00

• The total average recovery of moles of nitrogenous constituents per 100 g.-atoms of P is given in the
first line. The mean proportions of each constituent have been corrected to a 100% recovery.

stituents and of total purines and pyrimidines — a characteristic feature of
all deoxy pentose nucleic acids, as will be discussed later in this chapter — is
fully maintained. (See also Table XV.)

In later studies these fractionation experiments were extended to many
nucleic acid preparations from the tissues of ox, pig, and man, from sea
urchin sperm, and from the r"^ strain of coliphage T6.'^^ In these experi-
ments, and in subsequent studies with pneumococcal transforming prepara-
tions, artificially prepared complexes with histone, or in a few cases with
globin, served as the starting material. A selection of such fractionation
experiments is shown in Fig. 5. The results are in agreement with those ob-
tained with calf thymus nucleohistone.

Of particular interest is the distribution of the pyrimidine satellite 5-
methylcytosine in the fractions thus obtained, as shown in Table XVI. It
will be noticed that significant divergences in the concentration of this

364

ERWIN CHARGAFF

lOOr

60

■ adenine + thymine
E3 guanine + cytosine

T 12 1234 23 23 455 2345 234567 1234

20

I II III IV V VI \1I

Fig. 4. Composition of sodium deoxyribonucleate fractions (in moles per 100 g.-
atoms of phosphorus) prepared by the fractional dissociation of calf thymus nucleo-
histone with salt solutions of increasing strength. Each block (I-VII) represents
a separate fractionation run; the consecutively numbered bars within each block
represent individual fractions, with the NaCl concentration rising from left to right.
The first column (T) indicates the average composition of the total deoxyribonucleic
acid. (Taken from Chargaff et al}^^)

J^ adenine -f- thymine
0 guanine -f- cylosine

Wi guanine ■>■ hydroxy met hy Icy losme
lOOrS ,l„„,? 3 4 T ,,l,^.,2,.,,,3„^'^,^

0 50 I GO

CALF THYMUS

0 50 100

PIG LIVER

0 50 100

HUMAN SPLEEN

0 50 100

C0LIPHA6E T6

Fig. 5. Composition of fractions (in moles per 100 g. -atoms of phosphorus) pre-
pared by the fractional dissociation of artificial histone complexes with the indicated
sodium deoxypentose nucleate preparations. The abscissa indicates the proportion of
original nucleic acid P recovered in each fraction. Compare Fig. 4 for other explana-
tions. (Based on the results of Crampton el alP-^^)

ISOLATION AND COMPOSITION OF DEOXYPENTOSE NUCLEIC ACIDS

365

TABLE XVI

S-Methylcytosine Distribution in Fractions of Calf Thymus
Sodium Deoxyribonucleate*

Fraction
1 2 3

NaCl molarity in extraction of nucleohistone gel

Corrected mean ■proportions
Thymine
Cytosine
5-Methylcytosine

Molar ratio

Thymine to 5-methylcytosine
Cytosine to 5-methylcytosine

0.40 0.45 1.7

2.6

24.2

24.0

29.0

28.7

23.9

23.4

19.8

21.3

1.9

2.7

1.2

0

12.7

8.9

24.2

12.6

8.7

16.5

• The mean proportions of each constituent have been corrected on the assumption that one-half of the
nucleic acid P is contributed by the pyrimidine nucleotide. In one hydrolysis experiment with Fraction 4 a
minute amount of 5-methylcytosine (about 0.6 mole) was found.

pyrimidine in the several fractions were found. This finding, it may be
pointed out, is difficult to reconcile with the structural hypothesis of Wat-
son and Crick-*^ (compare Chapter 13) which would have led to the expecta-
tion of a random replacement of cytosine by the methyl derivative. (See
also Section X.2.)

It is too early to attempt a detailed consideration of the reasons under-
lying the fractionation procedure. That nucleic acid chains relatively rich
in guanylic and cytidylic acids are detached more easily than those in which
adenylic and thymidylic acids predominate indicates that some property
inherent in their composition or sequence reduces the strength with which
they are bound to the histone. Brown and Watson^*^ have, on the basis of
Cavalieri's work,'^^ suggested that hydrogen-bonding between the 2-amino
group of guanine and a phosphate group of the sugar-phosphate backbone
of the nucleic acid could reduce the acidic properties of structures in which
this type of interaction occurs frequently and thereby weaken the link be-
tween protein and nucleic acid.

The fractionation experiments, which are discussed in detail in the pub-
lications cited before,^^*"-*' suggest that the deoxypentose nucleic acid of a
given cell is composed of a very large number of differently constituted
individuals and that it is possible to achieve the resolution of this entire
spectrum of structural gradations (descending contents of guanine and

"2 J. D. Watson and F. H. C. Crick, Nature 171, 737, 964 (1953) ; Cold Spring Harbor
Symposia Quant. Biol. 18, 123 (1953).

366 ERWIN CHARGAFF

cytosine, ascending contents of adenine and thymine) into several distinct
"bands," in which all the regularities characteristic of the entire nucleic
acid are maintained.

IX. Composition Studies and Structural Investigations

The problem of nucleotide sequence has frequently been mentioned in
the preceding pages. (See, in particular, Section V.l.) It is the most elusive
and vexing, and also the most important, part of what may be considered
as the structural investigation of the deoxypentose nucleic acids. There was
never much doubt that the principal connecting links between the mono-
nucleotides were 3 ',5 '-phosphate bridges (see Chapter 12), though the
proverbial exceptions that should prove or test this rule have, perhaps, not
yet been looked for in a sufficient number of cases. But even complete
certainty regarding the points of attachment of the phosphate bridges
would contribute no more to the problem of structure than the statement
that all proteins contain peptide bonds. It must be admitted that the task
of sequence analysis, beyond our present means if a single macromolecular
polynucleotide chain is to be unriddled, becomes so gigantic, if the con-
clusions from the fractionation experiments (Section VIII) are justified,
as to discourage the most sanguine of optimists.

At the present time, only the crudest form of mapping the order in which
the mononucleotides are aligned in a nucleic acid chain appears attainable.
The task is similar to that of an ancient geographer: no more than dim
contours, vague directions can be discerned.

As has been pointed out before (Section V), several partial degradation
products of deoxypentose nucleic acids appear to offer an opportunity of
searching for the existence of certain general structural features; these are
the apurinic acids,"" which can be prepared readily from many deoxy-
pentose nucleic acids, and the various large fragments formed by enzymic
attack.^'-' ^-^ If the production of pyrimidine deoxyribonucleoside diphos-
phates^^*' 263-266 (^ggg g^jgQ Chapters 4 and 12) by acid hydrolysis of nucleic
acids can be standardized and a procedure for the quantitative estimation
of individual diphosphates developed, this may also contribute to the char-
acterization of structural differences, provided the method is applied to
several purified deoxypentose nucleic acids of different origin rather than
to commercial material. If these diphosphates really are indicative of those
positions in the original nucleic acid chain in which a pyrimidine nucleotide
is flanked by purine nucleotides,-^® it is conceivable that different nucleic

2" P. A. Levene and W. A. Jacobs, J. Biol. Chem. 12, 411 (1912).

26* P. A. Levene, J. Biol. Chem. 48, 119 (1921); 126, 63 (1938).

2" S. J. Thannhauser and G. Blanco, Z. physiol. Chem. 161, 116 (1926).

"« C. A. Dekker, A. M. Michelson, and A. R. Todd, J. Chem. Soc. 1953, 947.

ISOLATION AND COMPOSITION OF DEOXYPENTOSE NUCLEIC ACIDS

367

(J3 O^^O (jJj— Q^^O (jJj O^^^O C3

I Ca' — OH /\ C4 OH /\

1^ HO 0^1^ HO O

C5.

Cs'

N.

HO ^0,

Cs-

j^3*— %^0 C3— 0^^^ C3— 0^ ^ G3.-

I Cv O^ \ Cv 0^ \ /\ —

05' Cs' 05' C5'

C^' 0. J) Cv 0. ,0 0»' 0. JD Gv-

1 Cv 0H/'\ +C4— OHy^ +

r HO OH 1^ ho oh

HO Q

p HO OH I
05' 05'— OH 05' — OH "05'

(a) (6) (c) (d)

Fig. 6. Cleavage of apurinic acid by alkali. (Taken from Tamm et a/."^)

acids will yield entirely different results according to the particular nucleo-
tide arrangement.

Up to this time, only the apurinic acid of calf thymus has been investi-
gated in some detail. While enzymic investigation methods were of little
avail — deoxyribonuclease apparently does not attack a chain that has been
deprived of its purines^*^' ^^^ — further degradation by alkali yielded some
results.^^' Following a mild treatment with alkali the apurinic acid is
cleaved in part with the formation of diffusible fragments and of a non-
diffusible residue. The nondialyzable fraction comprises about 85% of the
pyrimidine nucleotides, but only about 40 % of the nonglycosidic deoxyribo-
phosphate residues present in the starting material. If the sodium salt of
apurinic acid is assigned an average molecular weight of 15,000,-^* a "sta-
tistical polynucleotide" in accord with its composition and properties would
comprise 17 thymidylic acid residues, 12 cy tidy lie acid residues, and 29
deoxyribophosphate residues. About 10-* isomers are possible. As a plausible
explanation of the partial degradation of apurinic acid by alkali a mech-
anism similar to that proposed for the cleavage of ribonucleic acid-" (com-
pare, however, Lipkin et al}^^) was contemplated in which the formation,
and subsecjuent rupture, of a transitory cyclic triester of phosphoric acid
linking the 3'- and 4'-hydroxyls of the sugar phosphate residues occurring
in the aldehydo form was assumed (Fig. 6). A different explanation of this
reaction is given in Chapter 12. The experiments w^ere interpreted as per-
mitting the formulation of calf thymus deoxyribonucleic acid as

[(Pui7Thy2Cy2) (Thyi5CyioPui2)]. ,

2" D. M. Brown and A. R. Todd, J. Chem. Soc. 1952, 52.

"8 D. Lipkin, P. T. Talbert, and M. Cohn, J. Am. Chem. Soc. 76, 2871 (1954).

368 ERWIN CHARGAFF

with the first part of the expression representing the portion of apurinic
acid sensitive to alkaU and Pu, Thy, Cy denoting purine nucleotides and
thymidyhc and cytidyhc acids respectively. Whether the structure proposed
for calf thymus nucleic acid, namely, that of a chain in which tracts
consisting principally of pyrimidine nucleotides are followed by stretches
in which purine nucleotides predominate, will be borne out by further
work, remains to be seen. As regards the nondiffusible residue produced
by the action of alkali on apurinic acid, it may be of interest to note that
about one-half remains nondialyzable even against 2 M NaCl solution and
that an almost quantitative yield of the corresponding nucleosides is af-
forded by the action of phosphodiesterase and phosphomonoesterase.^^^

X. Correlations and Concluding Remarks
1. Simplifying Generalizations

Even that master of inductive reasoning, Macaulay's infant who "is led
by induction to expect milk from his mother or nurse, and none from his
father," would have rejected as spurious some of the generalizations about
nucleic acids that have been made in the past and are still found in some
textbooks. The effort to force nature into a strait-jacket of puerile approxi-
mations has yielded many short-term successes in the natural sciences. As
is true of soap sculpture, they were pretty, but easily washed away. Whether
the attempt to teach the "gesta Dei per mathematicos" in three easy lessons
was more harmful than useful, may remain a matter of controversy. Nor
is plausibility a criterion of durability; quite the contrary: the gospels of
the future often are the heresies of the present. But the writers of textbooks
thrive on premature explanations.

The "tetranucleotide theory" continues, for this reason, to lead a stub-
born existence. There never were any but psychological reasons for its
formulation, as has been pointed out before;^"'* but even in a very recent
and massive treatise there will be found the statement that thymus nucleic
acid is a large chain consisting of 500 to 1000 tetranucleotide units and that
each tetranucleotide is formed by the combination of four nucleotides con-
taining adenine, cytosine, guanine, and thymine, respectively."" Actually,
a glance at the tables in which the information on the composition of many
different deoxypentose nucleic acids is assembled (Tables V-XIII) will
show that out of almost 50 different species so far investigated only E. coli
(Nos. 5-9 in Table XII) has yielded preparations that could be called "sta-
tistical tetranucleotides," though not much good will come of it. One must

289 M. E. Hodes and E. Chargaff, in preparation.

"" B. Flaschentrager and E. Lehnartz (eds.), "Physiologische Chemie," Vol. I, p.
768. Springer, Berlin, Gottingen, Heidelberg, 1951.

ISOLATION AND COMPOSITION OF DEOXYPENTOSE NUCLEIC ACIDS

369

conclude that there is no subunit of recognizably recurrent structure larger-
than a mononucleotide.

2. Unifying Generalizations

The fear of premature generalizations, though justified on the whole, is
not without its own risk. By insisting only on the differences between the
various deoxypentose nucleic acids, of which there are many, one may not
see the unity for the decimals. The deoxypentose nucleic acid molecules
appear to possess no perceptible periodicity of their constituents: the nucleo-
tide sequence is probably arrhythmic. If some recurrent features exist, we
have no means of discerning them, as a bird's-eye view of this giant throw
of dice requires a distance which is denied us. It is, therefore, the more sur-
prising that the inspection of even the earliest analytical results compelled
the recognition of several regularities which, since the time when they were
first proposed,* have become well established: (a) The sum of the purine
nucleotides equals that of the pyrimidine nucleotides, (b) The molar ratio of

ADENINE

30 -

20

10

OSPM

THYMINE

GUANINE Cr TO SINE

OSPM

OSPM

OSPM

Fig. 7. Proportions of nitrogenous constituents (per 100 g. -atoms of phosphorus)
in the deoxyribonucleic acids of the ox (O), sheep (S), pig (P), and man (M). (Taken
from ChargafT and Lipshitz.'")

370

ERWIN CHARGAFF

TABLE XVII
MoLAK Relationships in All Deoxypentose Nucleic Acids*

Table No.

Number

of
entries

Adenine to

Guanine to

Purines to

6-Amino to

thymine

cytosine

pyrimidines

6-keto groups

V

10

1.020±0.007

0.997±0.004

1.010±0.005

1.013±0.004

VI

4

1.015±0.003

0.995±0.006

1.010±0.004

1.013±0.003

VII

15

1.007±0.011

1.015±0.022

1.008±0.010

0.999±0.011

VIII

10

0.998±0.009

1.016±0.014

1.002±0.007

0.995±0.008

IX

14

1.034±0.013

1.014db0.012

1.023±0.008

1.014±0.009

X

17

1.006±0.012

0.980±0.020

0.995db0.013

1.013±0.010

XI

4

0.985±0.010

0.968±0.057

0.975zfc0.025

1.010±0.032

XII

17

1.016±0.024

0.953±0.022

0.974±0.018

1.042zh0.017

XIII

10

0.974±0.007

1.083±0.007

1.006±0.008

0.955±0.004

All tables

101

1.009

1.001

1.000

1.008

(weighted

mean)

* The mean ratios and their standard errors are given for each table. In Table VII, Nos. 1, 2, and 11 were
not included; in VIII, No. 1; in IX, Nos. 11, 15, and 16; in XII, Nos. 1, 2, and 15. The values for cytosine
refer also to 5-methylcyto8ine or 6-hydroxymethylcytosine.

adenine to thymine equals 1. (c) The molar ratio of guanine to cytosine
equals 1. (d) The number of 6-amino groups is the same as that of 6-keto
groups."^ It is, in fact, because of these relationships that it is possible to
state that, for instance, in the deoxypentose nucleic acids of wheat germ
5-methylcytosine substitues for part of the cytosine,^" that in certain phage
nucleic acids 5-hydroxymethylcytosine replaces cytosine entirely,^*' or that,
as has been shown recently ,^^^ 5-bromo- or 5-iodouracil may under circum-
stances take the place of thymine. It is not unlikely that more instances of
such substitution will be discovered. What appears remarkable, however,
is that up to this time they have been found only among the pyrimidines
and that no purine satellite has been revealed.

Some of these relationships are, with respect to the composition of the
nucleic acids of different mammalian genera (Table VI), illustrated in Fig.
7. A complete survey of almost all data assembled in Tables V-XIII is
provided in Table XVII. The above-mentioned ratios were calculated
separately for each entry and their average values and standard errors com-
puted for each tabular division. (The assistance of Dr. D. Elson in these
calculations is gratefully acknowledged.) It will be seen that the agreement

"1 This means that the sum of guanine and thymine equals that of adenine and
cytosine (+ methylcytosine). Originally, the ratio of all amino groups to enolic
hydroxyls was introduced.'" The present expression is preferable, especially in
the light of recent findings on pentose nucleic acids.*"

"2 D. Elson and E. Chargaff, Nature 173, 1037 (1954).

ISOLATION AND COMPOSITION OF DEOXYPENTOSE NUCLEIC ACIDS 371

is most impressive, especially in view of the diversity of sources, prepara-
tions, and procedures.

An attempt has been made-^^ to derive a molecular structure of deoxy-
pentose nucleic acids from these regularities and from the available X-ray
evidence. [Compare the discussion in Chapter 13.] This hypothesis, which
postulates a helical dyad in which the two coiled strands are held together
by a specific pairing of the bases showing the unity relationships mentioned
above, has much to recommend itself on aesthetical grounds; it makes good
use of several experimentally established facts. Whether it does more than
to describe the structure of that portion of the processed preparation from
which the diffraction patterns are obtained, remains, however, to be estab-
lished. It is not improbable that the scheme is incomplete in some essential
features, at least insofar as substitution in position 5 of the pyrimidines is
concerned. If 5-methylcytosine or analogues could take the place of cyto-
sine and vice versa without restriction, the 6-amino pyrimidines should be
able to replace each other at random. This is obviously not the case, as
shown by the remarkable constancy of the 5-methylcytosine content of a
given species. (See, especially. Table XI.) The absence of uracil from de-
oxypentose nucleic acids also is not easy to understand on these grounds.
Even more disturbing, perhaps, is the fact, pointed out before (Section
VIII. 2 and Table XVI), that 5-methylcytosine is distributed unevenly in
the fractions obtained by the fractionation of calf thymus deoxyribonucleic
acid. One gains the impression that^ — ^just as certain phage nucleic acids
contain 5-hydroxymethylcytosine in total replacement of cytosine — there
exist nucleic acid molecules as part of the preparations from calf thymus or
from wheat germ in which cytosine is entirely replaced by 5-methylcyto-
sine, whereas other fractions are, in turn, completely devoid of the latter.

3. A Concluding Remark

It may be considered as intellectually quite unsatisfactory that a con-
siderable part of what is known about the composition and structure of
nucleic acids must, as has been shown here, rest on assiduous analytical
work. In our time, much stress is laid on the forces that govern the life and
the economy of the cell. The discovery that the cell runs while it rests has
effaced the other half of the truth: that it rests while it runs. The quiet
center is falling into oblivion ; and there is widespread contempt for what is
regarded as morphology or analysis. We should, however, comprehend that
it is by way of decimals that we penetrate into nature. We read in The
Wisdom of Solomon (11:21): "But thou hast ordered all things in measure
and number and weight." We can only hope that the span will not be too
wide, the count too high, the weight too heavy.

CHAPTER 11

Isolation and Composition of the Pentose Nucleic Acids
and of the Corresponding Nucleoproteins

B. MAGASANIK

Page

I. Introduction 373

II. Isolation of Pentose Nucleoproteins 374

1. General 374

2. Isolation of Nucleoprotein from Animal Tissues 375

a. Fractional Centrifugation 375

b. Extraction with 0.14 M Sodium Chloride Followed by Isoelectric
Precipitation 375

c. Salt Fractionation 375

(1) Nucleotropomyosin from Carp Muscle 376

3. Isolation of Nucleoprotein from the Tissues of Higher Plants 376

a. Fractional Centrifugation 376

(1) Preparation of Cucumber Virus 4 (CV4) 377

b. Salt Fractionation 378

(1) Preparation of Turnip Yellow Mosaic Virus 378

4. Isolation of Nucleoprotein from Microbial Tissues 379

III. The Nature of Pentose Nucleoprotein 379

IV. The Isolation of Pentose Nucleic Acids 384

1. General 384

2. Isolation of PNA from Animal Tissues 387

(1) Preparation of Mammalian Tissue Ribonucleic Acid 387

3. Isolation of PNA from the Tissues of Higher Plants 389

(1) Isolation of PNA from Strains of Tobacco Mosaic Virus and Cu-
cumber Virus 389

4. Isolation of PNA from Microbial Tissues 390

(1) Preparation of PNA from Yeast 391

V. The Nature of PNA 391

VI. The Nucleotide Composition of PNA 394

1. General Considerations 394

2. Analytical Procedures 395

3. PNA from Animal Tissues 396

4. PNA from Microbial Tissues 402

5. PNA from Plant Viruses 403

6. Conclusions 40o

I. Introduction

The early history of the discovery of the nucleic acids is described in
Chapter 1.

373

374 B. MAGASANIK

The development of modern methods of cytochemistry allowed the
demonstration that both types of nucleic acid are present in all cells, ani-
mal, plant, and microbial, deoxypentose nucleic acid in the nucleus, and
pentose nucleic acid in the cytoplasm as well as in the nucleus.^ While
early preparations of PNA usually were badly degraded, the lability of
PNA was later recognized, and consequently milder methods of isolation
came into use which resulted in PNA preparations of considerably higher
molecular weight.^'* Finally, the application of modern methods to the
analysis of nucleic acids, such as paper chromatography,^ ion-exchange
chromatography,^ and paper ionophoresis,^ demonstrated that nucleic
acid preparations isolated from different tissues differed in composition
and did not in general contain the four constituent nucleotides in equimolar
quantities.

The description of these newer methods of isolation and the results
obtained by the modern methods of analysis are the subject of this chap-
ter.

II. Isolation of Pentose Nucleoproteins
1. General

Pentose nucleic acids are found in tissue extracts in combination with
proteins as nucleoproteins, the "nucleins" of the early workers. Generally,
isolation of nucleoprotein is the first step in the isolation of PNA.

A short description of the distribution of PNA in the cell, a subject
discussed in detail in Chapters 16-21, is helpful for the understanding of
the experimental approach used for the isolation of nucleoproteins. PNA
is found in both nucleus and cytoplasm. The nuclear PNA accounts for
about one-tenth of the total PNA of the cell.^ The isolation of nucleoli and
the composition of their PNA constituent has been described.^ The major
portion of the PNA of the cytoplasm is contained in the particulate frac-
tions, and most of it is found in the microsomes. Chemical analysis has
shown microsomes to be complex macromolecular structures composed of
lipid, protein, and PNA.^ The fractionation of the components of the mi-
crosomes without the denaturation of the protein has not been accom-

1 T. Caspersson and B. Thorell, Chrovwsoma 2, 132 (1941).

« H. S. Loring, /. Biol. Chem. 130, 251 (1939).

3W. E. Fletcher, J. M. Gulland, D. O. Jordan, and H. E. Dribben, /. Chem. Soc.

1944, 30.
* E. Vischer and E. Chargaff, J. Biol. Chem. 176, 715 (1948).

6 W. E. Cohn, /. Am. Chem. Soc. 72, 1471 (1950).

« R. Markham and J. D. Smith, Biochem. J. 52, 552 (1952).

7 W. C. Schneider, J. Biol. Chem. 165, 585 (1946).

8 W. S. Vincent, Proc. Natl. Acad. Set. U. S. 38, 139 (1952).

9 A. Claude, Science 97, 451 (1943).

ISOLATION AND COMPOSITION OF PENTOSE NUCLEIC ACIDS 375

plished. The soluble PNA found in the supernatant fluid after sedimenta-
tion of the microsomes is also bound to protein.^"

The PNA-lipoprotein complexes which constitute the particulate frac-
tions of animal cells may be isolated by fractional centrifugation of tissue
homogenate.s. Analogous procedures may be used for the isolation of some
animal viruses and of all plant viruses which resemble the cytoplasmic
particles in size, composition, and distribution in the host cell. Alterna-
tively, the tissue may be extracted with isotonic saline and the cytoplasmic
PNA-proteins isolated by isoelectric precipitation of the extracts at a pH
of about 5. The soluble PNA-proteins as well as the microsomes and mito-
chondria are obtained together in this procedure. This method can be used
for the isolation of PNA-proteins from the organs of animals and from
yeast, and for the isolation of plant viruses. Other methods, such as the
fractionation of tissue extracts with ammonium sulfate or sodium sulfate,
have been useful in special cases.

2. Isolation of Nucleoprotein from Animal Tissues

a. Fractional Centrifugation

This method for the isolation of the cytoplasmic nucleoproteins is de-
scribed in Chapter 21. Similar procedures were used for the isolation of the
virus of eciuine encephalomyelitis from infected chick embryos by Taylor
et al}^ The virus was found to be composed of protein, phospholipid, and
PNA.

6. Extraction with 0.14 ^^ Sodium Chloride Followed by Isoelectric Precipi-
tation

This method takes advantage of the fact that PNA-protein but not
DNA-protein is readily soluble in 0.14 M sodium chloride. The PNA-
protein is subsequently precipitated by addition of acid to a pH of 4-5.
The preparation of calf thymus PNA-protein by this procedure has been
described by Mirsky and Pollister,^^ and that of pancreas PNA-protein by
Kerr and Seraidarian.'^ These preparations consisted of mixtures of particu-
late and soluble nucleoproteins, and were not further characterized but
used directly for the preparation of free PNA.

c. Salt Fractionation

This method has not been used extensively. It does, however, deserve
attention because of its recent successful application by Hamoir to the

'" D. Szafarz, Biochim. et Biophys. Acta 6, 562 (1951).

" A. R. Taylor, D. G. Sharp, D. Beard, and J. W. Beard, J. Infectious Diseases 72,
31 (1943).

12 A. E. Mirsky and A. W. Pollister, J. Gen. Physiol. 30, 101 (1946).

13 S. E. Kerr and K. Seraidarian, J. Biol. Chem. 180, 1203 (1949).

376 B. MAGASANIK

isolation of a crystalline PNA-protein from carp muscle.^* This compound,
nucleotropomyosin, is the first crystalline nucleoprotein to be prepared
from animal tissue and possesses considerable chemical and biological
interest. Together with nucleotropomyosin, another crystalline protein, the
well-known muscle component tropomyosin, was obtained. The two pro-
teins were indistinguishable by electrophoresis but behaved differently in
the ultracentrifuge. Tropomyosin was free from PNA and identical with
the protein portion of nucleotropomyosin. In contrast to the complex
cytoplasmic nucleoproteins isolated so far, nucleotropomyosin appears to
be free of phospholipids. The method of preparation is given in detail.

(1) Nucleotropomyosin from Carp Muscle.^^ The preparation is carried out through-
out in the cold and all separations are done by centrifugation. Carp muscles cut
with a freezing microtome into slices 40 fi thick are extracted for 20 minutes with
3 vol. of KCl-phosphate solution (0.50 M KCl and 0.1 M KH2PO4 brought to pH 5.).
This extract is diluted with 3.5 vol. of cold water; the precipitate is discarded and the
supernatant (I) kept.

The residue from the first extraction is reextracted for 10 minutes with 3 vol. of
0.5 M phosphate solution of pH 6.5 containing 0.3% sodium adenosinetriphosphate.
The residue is discarded and the extract (II) is diluted with 7 vol. of water. A pre-
cipitate of myosin and nucleotropomyosin forms, which is washed twice with water
and redissolved in 0.5 M KCl at neutral pH. The supernatant II and the supernatant
I are mixed and brought to 4.6. The precipitate containing tropomyosin is washed
twice with water and redissolved in 0.5 M KCl at neutral pH.

Both solutions are now centrifuged for 30 minutes at 14,000 r.p.m. to remove some
turbid material and are purified by a second precipitation by dilution with 8 vol. of
water at neutral pH (nucleotropomyosin) or at pH 4.6 (tropomyosin). Both precipi-
tates are washed twice with water and redissolved in 0.5 M KCl at neutral pH.

Both tropomyosins are isolated from these two solutions by (NH4)2S04 fractiona-
tion at neutral pH: the major part of the total protein content of the solutions pre-
cipitates between 30 and 50% saturation, while the tropomyosins precipitate between
50 and 66% saturation. The precipitate can be redissolved quickly by a slight dilu-
tion with water, giving a water-clear solution.

Although these methods of preparation are very reproducible, some variations are
observed in the yields obtained, which are usually 0.07% of the wet weight for nucleo-
tropomyosin and 0.03% for tropomyosin.

The undiluted salted-out precipitates are used for crystallization. This is carried
out by dialyzing an appro.ximate 1.5% solution against a solution containing 16 g.
(NH4)2S04 per liter and 0.01 M acetate buffer of pH 5.4. Nucleotropomyosin crystal-
lizes in elongated prisms; tropomyosin in the quadrangular plates previously de-
scribed by Bailey.'^

3. Isolation of Nucleoprotein from the Tissues of Higher Plants

a. Fractional Centrifugation

This method has been used extensively for the isolation of plant viruses
from leaves. The mildness of the procedure permits the isolation of some

" G. Hamoir, Biochem. J. 48, 146 (1951).
!•* K. Bailey, Biochem. J. 43, 271 (1948).

ISOLATION AND COMPOSITION OF PENTOSE NUCLEIC ACIDS 377

of the less-stable viruses which do not survive chemical fractionation pro-
cedures. The success of this method for the purification of plant viruses
depends on the absence of material of corresponding particle size and
stability in uninfected plants. A nucleoprotein of the size of plant viruses
has been isolated by Pirie'^ from the sap of young uninfected tobacco
leaves. However, the amount of this material in older leaves is very small,
and it is removed in the course of the procedures used to clarify the in-
fected sap prior to the sedimentation of the plant virus. Consequently,
plant viruses isolated from cell sap, clarified by precipitation with ethanol,
freezing, and exposure to inorganic phosphate, will not be contaminated
by the nucleoprotein of normal plant leaves.^®

Several strains of tobacco mosaic and related viruses have been isolated
by fractional centrifugation for a study of the amino acid and nucleotide
composition of the virus nucleoprotein. ^^-^^ The preparation of cucumber
virus 4 (CV4) is given in detail.

(1) Preparation of Cxicumber Virus 4 {CV^y-. Young cucumber plants were inocu-
lated with CV4 by rubbing one leaf on each plant with a gauze pad saturated with
infective juice. This juice was obtained from a cucumber plant showing the typical
yellow mottling produced by cucumber virus 4 in members of the Cucurbitaceae.
About 1 month after inoculation, the plants were harvested and placed in a room
kept at —12°. After a few days the frozen plants were ground. Three per cent by
weight of dipotassium phosphate in a 50% solution was mixed with the pulp, and
after about 2 hours the juice was expressed from the cold but completeh' thawed
pulp. The juice was passed through a celite filter to remove coarse particles of green
pigment and extraneous matter, and then the virus was sedimented in the form of
pellets by centrifugation at 20,000 to 30,000 r.p.m. for 30 minutes. The supernatant
liquid, which was practically free of virus, was discarded, and the pellets were dis-
solved in small amounts of distilled water. The combined solutions of virus pellets
were spun at about 3000 r.p.m. on an angle centrifuge for 30 minutes to remove green
pigment and insoluble colloidal matter. The supernatant liquid, which contained the
virus, was then returned to the high-speed centrifuge and the process was repeated
about three times, a longer period being allowed for the high-speed centrifugation as
the virus became more concentrated. In many cases it was found possible to effect a
better separation of green pigment from the virus by dissolving the pellets obtained
by the first two high-speed centrifugations in 0.1 M phosphate buffer at pH 7 and then
using distilled water as a solvent for the pellets of the last two centrifugations. All
of the preparations used for elementary analysis were further purified by dialysis
against flowing distilled water for 48 hours.

Pellets of purified virus ranged from 0.1 to 0.4 g. per liter of expressed juice as
compared with 2-2.5 g. per liter ordinarily obtained for tobacco mosaic virus from
the juice of diseased turkish tobacco plants.

'« N. W. Pirie, Biochem. J. 47, 614 (1950).

'7 C. A. Knight and W. M. Stanley, /. Biol. Chem. 141, 29 (1941).

18 C. A. Knight, J. Biol. Chem. 171, 297 (1947).

1" C. A. Knight, J. Biol. Chem. 197, 241 (1952).

378 B. MAGASANIK

h. Salt Fractionation

The original isolation of tobacco mosaic virus was achieved by salt
fractionation.^" The procedure recently described for the preparation of
turnip yellow mosaic virus by fractional precipitation with ammonium
sulfate is given in detail. ^^

(1) Preparation of Turnip Yellow Mosaic Virus.^^ The virus has been prepared
from a number of cruciferous plants, namely radish, Bronica carrinata, turnip, Chinese
cabbage, and from a specimen of broccoli found infected naturally. For general use
either turnip or Chinese cabbage is suitable and, generally speaking, old or "hard"
plants give better yields than rapidly growing "soft" plants. Yields up to 1 g. of virus
per liter of sap have been obtained from stunted turnip plants grown in the open, but
the usual jdelds from glasshouse plants vary from 200 to 500 mg. per liter.

Plants infected with the virus are minced, either fresh or after freezing overnight,
and the sap is expressed. If field turnips are used, it is necessary to use a hydraulic
press to remove the last half of the sap. The pH of the sap is usually about 5.7. The
sap is clarified by adding slowly, with stirring, 300 cc. of 90% ethanol to each liter.
This quantity of ethanol is fairly critical and should be measured carefully.

The copious precipitate which forms is centrifuged off at once at 3500 r.p.m. for
15 minutes. To the supernatant liquid, a volume of saturated ammonium sulfate equal
to half the volume of purified sap is added, and the solution is allowed to stand,
preferably overnight. After a few hours, large numbers of small octahedral crystals
are to be found in the liquid, and these increase in size with time. Many small, highly
refractive, birefringent crystals are also to be seen, especially in the sap from old
turnip plants, but these are inorganic and are removed later.

The crystalline precipitate is centrifuged off (30 minutes at 3500 r.p.m.), but the
supernatant liquid frequently deposits a second crop of crystals on standing for a
day or two, and so may be kept. In addition, it is very difficult to remove all the small
crystals by centrifuging as they sediment rather slowly because they are not very
dense.

The precipitate which contains the virus and much insoluble material is resus-
pended in a small volume of water (one-tenth to one-quarter of the original sap
volume) or buffer solution, and is centrifuged again to remove insoluble material.
Precipitation of the virus followed by resolution of the crystals and centrifuging to
remove debris is repeated several times. In favorable circumstances the virus prepara-
tion may be fairly clean by this stage.

Several methods may be used for further purification. The virus is not digested by
trypsin and treatment with commercial pancreatic extract (10 mg./cc), and incuba-
tion for a few hours with a few drops of chloroform as preservative will remove some
of the contaminants. The enzyme is removed by several recrystallizations of the
virus.

A treatment which is of use in removing brown pigment, if present, is crystalliza-
tion of the virus from dilute alcohol solutions. A solution containing some 5 mg./cc.
of virus and a trace of salt is cooled to 0° C. and 0.25-0.30 vol. of absolute alcohol is
added gradually with stirring. On adding drop by drop a solution containing 20 cc.
of absolute alcohol, 10 cc. of glacial acetic acid, and water to make 100 cc, the virus
solution becomes turbid, and this turbidity disappears on allowing the solution to

2« W. M. Stanley, Science 81, 644 (1935).

21 R. Markham and K. M. Smith, Parasitology 39, 330 (1949).

ISOLATION AND COMPOSITION OF PENTOSE NUCLEIC ACIDS 379

warm. If the solution is left in the refrigerator for a few hours, the virus crystallizes
out in the form of small birefringent needles; they may be left at 1°C. for a few days,
when the mass of crystals will be found to adhere to the bottom of the container and
the mother liquor may be decanted without difficulty. The crystals are then dissolved
in 0.1 M disodium phosphate or a neutral buffer, as they form cloudy suspensions in
plain water. Spinning at 5000 r.p.m. removes the pigmented material which by now is
insoluble.

4. Isolation of Nucleoprotein from Microbial Tissues

The protoplasm of microorganisms is especially rich in nucleoprotein.
Bacteria contain as much as 15% (dry weight) nucleic acid, of which
60-75 % is of the pentose type. In spite of the high concentration of PNA
in bacterial cells, it is difficult to obtain bacterial PNA-protein prepara-
tions free from DNA-protein and comparable in purity to the PNA-pro-
teins isolated from animal cells. The obstacles encountered are the difficulty
of breaking up bacterial cells without harm to their labile constituents, the
presence of potent nucleases in bacterial extracts, and finally, the presence
of DNA-protein throughout the cytoplasm rather than in a discrete remov-
able structure such as the nucleus.

Recently, Parsons has described the isolation of a PNA-protein free
from deoxyribose from Clostridium perfringcnsr^ The nucleic acid was
extracted from lyophilized cells with 0.14 M NaCl and purified by precipi-
tation with methanol at —10° C. and with ammonium sulfate.

The situation is more favorable in yeasts primarily because of their low
DNA-protein content. Thus the isolation of PNA-proteins from acetone-
dried ground yeast has been accomplished by Khouvine.^^ Several nucleo-
proteins which differed in nucleotide composition were obtained by iso-
electric precipitation at different pH's. The significance of this observation
will be discussed in the section on nucleotide composition.

III. The Nature of Pentose Nucleoprotein

The isolation of material consisting of nucleic acid and protein is in itself
no proof that an entity such as "nucleoprotein" really exists. The material
isolated may be a conglomerate consisting of a random collection of dis-
tinct protein and PNA units held together by the attraction of oppositely
charged groups located on their surfaces. Such salt-like combinations of
protein and PNA could be present in the cell, or the two components could
meet and become attached to one another in the course of the isolation
procedure. In either case, the proper approach would be to separate the
protein and nucleic acid moieties and to study their individual properties.

" C. H. Parsons, Jr., Arch. Biochem. and Biophys. 47, 76 (1953).
2' Y. Khouvine and H. De Robichon-Szulmajster, Bull. soc. chim. biol. 33, 1508
(1951).

380 B. MAGASANIK

On the other hand, nucleoproteins may be well-defined chemical struc-
tures in which nucleic acid and protein are arranged in a definite pattern,
and these entities may be the ultimate carriers of biological properties.
The detachment of the PNA from the protein would then present us with
fragments valuable for chemical and physical study but without biological
interest.

The question of the nature of PNA-protein is thus largely a question of
the nature of the bonds which hold PNA and protein together. Certain
relevant conclusions may be drawn by examining the conditions which
cause these bonds to break.

The early nucleic acid chemists used dilute alkali to separate PNA from
protein, but with the recognition of PNA as an alkali-labile polymer, other
methods came into use. The most generally applicable of these methods is
treatment with hot 10% sodium chloride.^^ Recently milder procedures,
such as treatment with guanidine hydrochloride^^ or with sodium dodecyl
sulfate^^ in the cold, have been found effective in separating PNA from
protein. In the special case of the plant viruses even milder treatment
suffices: the PNA of tobacco mosaic virus was liberated by very short
heating at neutral pH;-^ the PNA of turnip yellow mosaic virus could be
detached from the protein by treatment with 30 % ethanol in the cold.^^

It is quite evident from the ease with which nucleoprotein can be split
that the two moieties are not joined by covalent bonds. The agents used
for the separation of PNA from protein are all eff'ective protein dena-
turants, and, indeed, the removal of PNA is always accompanied by the
denaturation of the protein. It seems likely, therefore, that the forces
which are responsible for keeping protein molecules in their native folded
configuration are also responsible for the binding of PNA to protein. These
forces are the coulombic attractions of oppositely charged ions, the attrac-
tion of dipoles, and hydrogen bonds. The polar groups on protein and PNA
are well suited for such mutual attractions.

These forces of attraction may be quite nonspecific and cause the forma-
tion of nucleoprotein whenever PNA and protein are in close proximity.
Alternatively, nucleoproteins may be formed just like native proteins,
under the influence of specific directing forces active in the cell. Recent
work sheds some light on this problem, although sufficient evidence which
would enable us to choose between these alternatives has not been ob-
tained. Szafarz'" observed that the supernatant fluid of cytoplasmic ex-
tracts centrifuged at 60,000 g did not contain a component corresponding

2* J. N. Davidson and C. Waymouth, Biochem. J. 38, 375 (1944).
26 E. Volkin and C. E. Carter, J. Am. Chem. Soc. 73, 1516 (1951).

26 E. R. M. Kay and A. L. Bounce, J. Am. Chem. Soc. 75, 4041 (1953).

27 S. S. Cohen and W. M. Stanley, J. Biol. Chem. 144, 589 (1942).

ISOLATION AND COMPOSITION OF PENTOSE NUCLEIC ACIDS 381

in electrophoretic mobility to free PNA. The pH of the extracts could be
varied from 5.7 to 8.2 and the ionic strength from 0.08 to 0.31, without
resulting in the appearance of a band characteristic of free nucleic acid.
Szafarz therefore concluded that the nonsedimentable fraction of cyto-
plasmic PNA is bound to protein by bonds stable over a considerable
range of pH and ionic strength. In contrast, the bonds between simple
proteins such as albumin and PNA, which are formed when the two com-
ponents are mixed at a pH intermediate between their isoelectric points,
are not stable in solutions of an ionic strength equivalent to that used in
the experiments with the cytoplasmic nucleoproteins.-* The bond in the
cytoplasmic nucleoprotein may owe its strength to the particular proper-
ties of either the protein or the PNA, or to a particular steric relation
between the two components. An interesting observation by Szafarz'"
suggests that proteins of the cytoplasm possess special ability to form
stable complexes with PNA. Thus, small amounts of commercial yeast
PNA added to cytoplasmic extracts of the flagellate Polyf.omella coeca
could not be demonstrated by electrophoresis of the mixture over an
extended range of pH and ionic strength. The proteins of the cytoplasm
were capable of reacting in vitro with PNA isolated from a different species,
and the resulting complex possessed the stability of a cytoplasmic nucleo-
protein. The study of such artificial nucleoproteins may lead to the dis-
covery of the structural properties of cytoplasmic proteins which enable
them to combine with PNA. At present, it would seem most likely that
the charged groups on the surface of protein and PNA molecules are re-
sponsible for the ease of the interaction, and that the great number of such
coulombic bonds between each molecule of protein and PNA accounts for
the stability of the nucleoprotein complexes.

It is evident that free PNA could not be found in tissue extracts in the
presence of reactive cytoplasmic protein, even if it did exist in that state
in the intact cell. Without further evidence, the nucleoproteins isolated
from cytoplasmic extracts cannot be considered to be characteristic com-
ponents of the intact cell.

However, nucleoproteins which differ in properties from the complexes
formed by the direct interaction of protein and PNA seem to exist. An
interesting example is the crystalline nucleotropomyosin isolated by
Hamoir from carp muscle'^ by the method described in detail in the pre-
ceding section. The two crystalline proteins, tropomyosin and nucleo-
tropomyosin, showed identical behavior in an electric field, although the
former was free of nucleic acid while the latter contained 15% PNA. The
presence of nucleic acid thus did not affect the charges on the surface of
the protein, and it must be assumed that the PNA fits into the pattern of

28 L. G. Longsworth and D. A. Maclnnes, J. Gen. Physiol. 26, 507 (1942).

382 B. MAGASANIK

the protein in such a manner that its charged groups are not exposed and
the charged groups on the surface of the protein are not affected. Such an
orderly arrangement of protein and PNA would not be expected to result
from the random interaction of the polymers without the influence of a
directing force. Nucleotropomyosin does not have a fixed composition.^^
Exposure of this nucleoprotein to low pH results in the liberation of tropo-
myosin and in the accumulation of nucleotropomyosin of increased nucleic
acid content. Within large limits (10-20% of PNA) the physical properties
of these compounds which depend on surface charges are identical. When
the PNA content is increased to 30%, a nucleoprotein of lower isoelectric
point than ttie original nucleotropomyosin precipitates.

The properties of nucleotropomyosin are those of a compound in which
a core of PNA is completely covered by protein. Apparently a large por-
tion of the protein can be peeled off without exposing the charged groups
of the PNA core, and therefore without altering those physical properties
of the complex which depend on surface charges.

A similar relationship between protein and PNA has been discovered in
one of the crystalline plant viruses, turnip yellow mosaic. ^^ The sap of
infected plant leaves contains, in addition to the virus nucleoprotein,
another protein free of PNA but of similar size, shape, and crystal struc-
ture, and with identical electrophoretic behavior. The two proteins could
be separated by ultracentrifugation. They reacted quantitatively in the
same manner with antiserum directed against the nucleoprotein. However,
only the nucleoprotein could infect plants or cause the formation of anti-
bodies when injected into animals. Pancreatic ribonuclease did not attack
the intact nucleoprotein, but hydrolyzed the nucleic acid prepared from it
in the usual fashion.

Here again it appears that the PNA contained in the nucleoprotein is
completely covered by protein so that the surface properties of the com-
plex, such as electrophoretic mobility and reaction with antibodies, are
exclusively determined by the nature of the protein component. The
stability of the nucleoprotein to the action of pancreatic ribonuclease
confirms the assumption that the PNA moiety is not easily accessible. The
biological properties of the nucleoprotein, such as infectivity and anti-
genicity, depend on the presence of the PNA. However the PNA alone,
separated from the nucleoprotein by denaturation of the protein with
dilute ethanol, is neither infective nor antigenic. It would seem that the
nucleoprotein possesses activities which are not the sum of the activities
of its constituent parts; in this sense the nucleoprotein is an ultimate
biological unit. Still, it must be borne in mind that we do not know whether
the isolated PNA is in its native state; treatment sufficiently drastic to

" G. Hamoir, Biochem. J. 50, 140 (1952).

ISOLATION AND COMPOSITION OF PENTOSE NUCLEIC ACIDS 383

TABLE I
Properties of Crystalline Plant Virus Nucleoproteins

Molecul

ar

Nucleotides

Virus

weight X

10-6

PNA,

%

per mole

Ref.

Tobacco mosaic (TMV)

40

6

8000

30

Rib grass

40

6

8000

31

Cucumber (CV4)

40

6

8000

17

Tomato bushy stunt

10.6

17

5000

32,33

Southern bean mosaic

6.6

21

4000

34

Tobacco ring spot

3.4

40

4000

35

denature protein might well have a similar destructive effect on the integ-
rity of PXA. The possibility cannot be ignored that native PNA, if we
could obtain it, would be found to possess alone the biological activities
associated with the nucleoprotein.

The composition of the individual plant virus nucleoprotein is constant.
The material isolated from plants infected with turnip yellow mosaic
virus consists of only two components: the free protein, and the nucleo-
protein containing 28% PNA. Complexes of intermediate PNA content
are not found. The compositions of several crystalline plant viruses are
presented in Table I. The PNA content of these preparations is constant,
regardless of the host, the time of harvest, or the method of isolation. It is
further of interest that the different strains of tobacco mosaic virus all have
identical PNA content, and even, as will be discussed later, identical
nucleotide composition; however, the amino acid composition of their
proteins differs.^"' ^* The particle weight of different plant viruses and their
PNA content show great variation. Thus, tobacco mosaic virus has a
particle weight of 40 X 10^ with a PNA content of 6%, while tobacco ring
spot virus with a particle weight of only 3.4 X 10^ contains 40% PNA.
If their PNA content is expressed as the number of nucleotides per virus
particle, considerably less variation- — only from 8000 to 4000- — is observed.
When the assumption is made that each virus particle contains 1 molecule
of PNA, the molecular weight of tobacco mosaic PNA would be of the
order of 2,500,000; actually, the molecular weight of tobacco mosaic PNA
isolated by the mildest procedure^^ was estimated as 300,000. It appears

3" W. M. Stanley and H. S. Loring, Relazioni del IV Congresso Inlernazionale di
-patologia comparata 1, 45 (1939).

31 C. A. Knight, J. Biol. Chem. 45, 11 (1942).

32 W. M. Stanley, J. Biol. Chem. 135, 437 (1940).

33 H. Neurath and G. R. Cooper, J. Biol. Chem. 135, 455 (1940).
3" G. L. Miller and W. C. Price, Arch. Biochem. 10, 467 (1946).
36 W. M. Stanley, /. Biol. Chem. 129, 405 (1939).

36 C.A. Knight and W. M. Stanley, J. Biol. Chem. 141, 39 (1941).

384 B. MAGASANIK

that either extensive depolymerization occurred when the nucleoprotein
was spht, or that each particle of nucleoprotein contains as many as 10
molecules of PNA.

IV. The Isolation of Pentose Nucleic Acids
1. General

It is the aim of biochemical research to relate the chemical properties of
natural substances to their function in the living organism. An important
early step in work of this kind is, therefore, the isolation of pure compounds
which are characterized by possessing specific biological activity. In most
cases the activity is first observed in the intact organism. An organ or
group of cells rich in this activity is chosen as starting material for the
isolation of the compound responsible for the activity. The components of
the cells are fractionated by physical and chemical methods, using the
quantitative measurement of the biological activity as a guide. The mild-
ness or harshness of the isolation procedure depends on the nature of the
carrier of the activity; the disappearance of the activity at any step gives
immediate notice that this step must be avoided. Finally, the molecular
homogeneity of the material thought to be the smallest carrier of the
biological activity is investigated by physical methods.

The success of this approach in the isolation of pure proteins is, in large
measure, due to the striking and easily measurable biological activity so
many proteins such as enzymes, antibodies, and hormones possess. The
lessons learned concerning the conditions under which their biological
properties are lost, such as the action of heat, strong acids and bases, etc.,
could be applied to proteins with less striking biological activities. The
biological importance of proteins encouraged the development of physical
methods such as ultracentrifugation, electrophoresis, and the measurement
of diffusion by which their homogeneity could be investigated.

The situation is quite different with regard to nucleic acids, and particu-
larly pentose nucleic acids. At present, in spite of much work and greatly
more speculation, the function of PNA is not known and its biological
activity can, therefore, not be measured. One must be content to aim at
the recovery of material characterized in the cell only by the chemical
properties of its constituents, the purine and pyrimidine bases, pentose,
and phosphate; all that can be achieved is the separation of this material
from all compounds of different chemical composition. It is not known
whether a particular manipulation used in the course of the isolation
procedure will cause a change in the properties of the PNA which are
responsible for its function in the cell. Furthermore, the homogeneity of
the isolated material cannot as yet be deduced with any degree of certainty
from physical measurements. The difficulty of relating the results of diffu-

ISOLATION AND COMPOSITION OF PENTOSE NUCLEIC ACIDS 385

sion and sedimentation measurements on highly charged particles to their
size and shape is discussed in Chapter 13. In many instances physical
measurements clearly indicate the inhomogeneity of PNA preparations.
So far no method has been developed which permits the fractionation of
the molecular species contained in such heterogeneous preparations.

The methods used for the isolation of PNA aim at the complete recovery
of the PNA contained in the biological specimen under investigation. This
is usually an organ, such as liver or pancreas, a homogeneous collection of
cells, such as yeast paste, or a cell fraction, such as mitochondria or micro-
somes. The composition and properties of the isolated preparation will be
a cross-section of the properties of the various pentose nucleic acids present
in this material. A comparison of the nature of the PNA obtained from
different biological specimens will reflect the differences in the types of
PNA predominating in each specimen. It is important to keep in mind
that most methods of isolation do not, however, result in the complete
recovery of the PNA contained in the material. Fractionation may thus
take place in the course of isolation, and PNA samples isolated from the
same starting material by different isolation procedures may actually
differ in composition and properties.

As mentioned before, it cannot be decided whether a given preparation
of PNA has been degraded in the course of isolation as long as the biologi-
cal function of PNA is not known. Therefore, only the study of the action
of the chemical and physical agents employed in the course of the isolation
procedures on purified samples of PNA is helpful in choosing the least
destructive path. Obviously, any treatment that causes degradation of
purified PNA samples must be avoided in their preparation. In earlier
work PNA was extracted from tissues by dilute alkali; however, the ex-
treme ease with which PNA is hydrolyzed by alkali to a mixture of nucleo-
tides shows clearly that alkali should not be used in the preparation of
PNA. Dilute acid hydrolyzes PNA in a similar manner; in addition, the
purine-pentose bond is labile to acid and, although heat is usually neces-
sary to accomplish these transformations, prolonged contact with acid
even in the cold should be avoided. The most troublesome factor in the
isolation procedure is the presence of highly active and sturdy nucleolytic
enzymes in most tissues. The best known of these enzymes is pancreatic
ribonuclease;" it was found to accompany PNA through most of the stages
of isolation and was clearly responsible for the degraded state of the PNA
isolated.'^ This enzyme is still active at low temperatures and is not easily
inactivated by heat. The activity of the enzyme may be competitively

" M. Kunitz, J. Gen. Physiol. 24, 15 (1940).

38 J. E. Bacher and F. W. Allen, J. Biol. Cheni. 183, 641 (1950).

386 B. MAGASANIK

inhibited by heparin,^^- ^^ but so far this polysaccharide has not been used
to protect PNA during isolation. It is claimed that the detergent sodium
dodecyl sulfate^^ or guanidine hydrochloride^^ will denature the enzyme,
and that the use of these substances for the separation of the PNA from
protein will protect PNA from ribonuclease. The only nucleic acid speci-
mens whose degradation by ribonuclease during isolation can be excluded
with certainty are those derived from plant viruses, as the nucleoprotein,
which is the form in which the PNA is separated from host components,
is resistant to the action of nucleolytic enzymes.^^

The effect of the damaging agents referred to so far, viz., alkali, acid,
and ribonuclease, is manifested by a change in the nucleotide composition
of the PNA, Other agents seem to affect only the physical properties of
PNA. The PNA obtained from plant viruses, whose protein moiety was
denatured by short heating or by dilute ethanol, are viscous and possess
a molecular weight of about 300,000.^7 On standing, such PNA prepara-
tions break down spontaneously into smaller particles. The isolation of
PNA preparations of similar viscosity, particle size, and instability from
the liver of rat, rabbit, and calf, has recently been reported.*^ By analogy
with the observations on DNA where viscosity and high particle weight
are correlated with biological activity, it is thought that such PNA prep-
arations are more nearly representative of native PNA than others of
small particle weight.''^ The preparations of PNA of high particle weight
lose their viscosity and rapid rate of sedimentation in solutions of high
sodium chloride content or upon heating.*^ The use of strong electrolyte
solutions and heat in the process of isolation should probably be avoided.

In the succeeding sections the newer methods used for the isolation of
PNA are listed; several characteristic methods, which have served for the
preparation of the nucleic acids whose composition is discussed subse-
quently, are presented in detail. This chapter does not intend to give an
historical survey of the methods used for the isolation of PNA and will,
therefore, not describe the older methods in which the PNA was extracted
from tissue with dilute alkali. A description of these methods, which are
still used for the production of commercial yeast PNA, is found in Levene's
monograph. ^^

" J. S. Roth, Arch. Biochem. and Biophys. 44, 265 (1953).

" N. Zollner and J. Fellig, Am. J. Physiol. 173, 223 (1953).

« E. L. Grinnan and W. A. Mosher, J. Biol. Chem. 191, 719 (1951).

^ It should be emphasized that this is only an assumption; it may be that the vis-
cous preparations are artifacts resulting from the aggregation of PNA molecules
in the course of isolation.

" P. A. Levene and L. W. Bass, "Nucleic Acids." Chemical Catalog Company, New
York, 1931.

isolation and composition of pentose nucleic acids 387

2. Isolation of PNA from Animal Tissues

Chargaff ct al. isolated PNA from the livers of several species of animals.^*
Their procedure was based on a modification of the method of Davidson
and Waymouth.-* PNA was dissociated from protein by boiling in 10%
NaCl. An analogous procedure used by the same authors for the isolation
of PNA from baker's yeast, is described in detail below. Similar methods
were used by Tsuboi and Stowell^^ and by Davidson et al}^ for the isolation
of PNA from liver, and by Davidson and Smellie for the isolation of PNA
from cytoplasmic fractions of the liver/^

PNA was prepared from beef pancreas by Kerr and Seraidarian^^ by a
method in which the nucleoprotein, isolated by isoelectric precipitation,
was dissociated by exposure in a half-saturated solution of NaCl for a
period of 36 hours or more. The danger of degradation of PNA during
isolation by tissue ribonuclease is particularly great in the preparation of
PNA from pancreas. Bacher and Allen described the preparation of PNA
from pancreas after removal of ribonuclease by extraction with dilute acid
and acetone.^*

A general method for the preparation of PNA from different animal
tissues, including liver, spleen, thymus, and pancreas, was described by
Volkin and Carter.^^ It consists of the precipitation of PNA from a cold
2 M guanidine hydrochloride solution in which protein remains soluble.
Their procedure is described in detail below. A modification of this method
was used by Grinnan and Mosher for the preparation of highly polymer-
ized PNA from rat and rabbit liver.^^ The nucleotide composition of their
preparations was not determined. Recently, the preparation of PNA from
liver, pancreas, and tumor tissue by the use of sodium dodecyl sulfate has
been described by Kay and Dounce.^^

(1) Preparation of Mamvialian Tissue Ribonucleic Acid.-^ The method
of isolation of ribonucleic acid from tissue homogenates consisted of (a)
the removal of deoxyribonucleic acid as a nucleic acid - protein complex,
(b) the precipitation of the ribonucleic acid from a cold 2 M guanidine
hydrochloride solution in which the large bulk of protein remains soluble,
and (c) further purification of the ribonucleic acid by chloroform extrac-
tion and alcohol precipitations.

The possibility of the occurrence of nuclease action on ribonucleic acid
duringlthe preliminary steps of the preparation can be obviated by imme-

^* E. Chargaff, B. Magasanik, E. Vischer, C. Green, R. Doniger, and D. Elson,

/. Biol. Chem. 186, 51 (1950).
« K. K. Tsuboi and R. E. Stowell, Biochim. et Biophys. Acta 6, 192 (1950).
. ^6 J. N. Davidson, S. C. Frazer, and W. C. Hutchinson, Biochem. J. 49, 311 (1951).
■" J. N. Davidson and R. M. S. Smellie, Biochem. J. 52, 600 (1952).

388 B. MAGASANIK

diately homogenizing the tissue in concentrated guanidine hydrochloride.
The latter reagent is an effective protein denaturant. These procedures
were found to be applicable to a number of mammalian tissues; details of
the methods of preparation follow.

Fresh or frozen tissue was cut in small pieces and blended for 6 to 8 minutes
with 3 vol. per gram of tissue of a 0.15 M sodium chloride-0.02 M phosphate buffer,
pH 6.8. A few drops of octyl alcohol were added to reduce foaming. The homogen-
ate was then centrifuged at 3000 g for 30 minutes. Essentially all the deoxyribo-
nucleic acid was removed in the form of an insoluble nucleic acid-protein complex
as described by Mirsky and Pollister.i^ All operations were carried out between 2
and 5°.

To the supernatant solution enough solid guanidine hydrochloride was added,
with rapid stirring, to make the solution 2 M with respect to guanidine hydrochloride.
The solution was placed in a 38° bath and allowed to stand at this temperature for
30 minutes, then chilled at 0° for 1 hour. Under these conditions most of the protein
of the tissue extract remained soluble, while a gelatinous precipitate formed which
contained ribonucleic acid and a small amount of protein. The precipitate was washed
twice with a cold solution of 2 M guanidine hydrochloride (1 vol. per gram of original
tissue) and extracted with chloroform-octyl alcohol (5:1). The suspension of nucleic
acid in guanidine hydrochloride was added to an equal volume of the chloroform-
octyl alcohol mixture, warmed to 40°, then shaken mechanically for 30 minutes.
The mixture was centrifuged and the upper aqueous layer containing the nucleic acid
removed. The extraction of the aqueous solution at 40° was repeated twice with fresh
chloroform-octyl alcohol. Extractions in the cold, or in saline or water solutions,
resulted in incomplete separation of the nucleic acid from protein. Nucleic acid was
precipitated in the cold from the guanidine solution by adjusting the acidity to pH
4.2-4.5 with acetic acid and adding 2 vol. of cold ethanol. The white, flocculent ribo-
nucleic acid precipitate was centrifuged and washed twice with cold 70% alcohol. The
precipitate was then dissolved in water, carefully adjusted to pH 6.8 with dilute
sodium hydroxide, and any insoluble material (denatured protein) centrifuged off.
The ribonucleic acid was purified by adding enough 1 M sodium chloride to bring the
final concentration to 0.05 M sodium chloride and precipitating the sodium ribonu-
cleate with 2 vol. of cold ethanol. The product was washed twice with cold 70%
ethanol.

In the second method the tissue was immediately homogenized with 3 vol. per
gram of tissue of cold 2.5 M guanidine hydrochloride solution. The rest of the pro-
cedure followed that of the first method, except that the ribonucleic acid-protein
complex was washed at least three times with cold 2 M guanidine hydrochloride to
ensure complete removal of any contaminating DNA. Excess foaming, which occurred
during the blending in the presence of guanidine hydrochloride, was alleviated by
adding a* few drops of octyl alcohol after the solution had warmed a few minutes in
the 38° bath.

Dup icate liver ribonucleic acid preparations made bj^ the two methods had an
essentially identical analytical composition, indicating that in liver little or no en-
zymatic hydrolysis occurred in the first procedure.

The mammalian ribonucleic acids readily dissolved in water to give clear, color-
less solutions. Preparations to be stored were lyophilized from water solutions.
Concentrations as high as 20 mg. per cubic centimeter failed to give a reaction with
diphenylamine reagent, indicating that all the nucleic acid was of the ribose type.

ISOLATION AND COMPOSITION OF PENTOSE NUCLEIC ACIDS 389

Similar concentrations gave negative biuret tests. The yield of ribonucleic acids
varied from 20 to 30% of the total tissue ribonucleic acid.

3. Isolation of PNA from the Tissues of Higher Plants

The study of the nucleic acids of plant tissues has been rather neglected
in recent years. Kay and Dounce^^ reported the isolation of PNA from
wheat germ by a method analogous to that used for animal tissue, namely
by the use of sodium dodecyl sulfate. Their product is characterized by a
very low ratio of nitrogen to phosphorus and may be contaminated with
other phosphorus-containing compounds. So far, no analyses of the nucleo-
tide composition of preparations of plant PNA have been reported. On
the other hand, the nucleic acids of plant viruses have been isolated and
their composition and properties have been determined. The pentose
nucleic acids of several strains of tobacco mosaic virus and the related rib
grass viruses, were detached from protein by heating for 1 minute as
described by Cohen and Stanley," and modified by Knight. ^^ In the case
of turnip yellow mosaic virus, the protein was denatured and the PNA
liberated by treatment with 30% ethanol.^^

(1) Isolation of PNA from Strains of Tobacco Mosaic Virus and Cucumber Virus.^^
Six cubic centimeters of 0.13 M sodium chloride, 0.001 M with respect to Sorensen's
phosphate (K2HPO4-KH2PO4 , 11:5) and at a pH of about 7, was heated in a 15-cc.
conical glass centrifuge tube in a water bath at 100°. To this was added 2 cc. of virus
solution at a concentration of 30 to 80 mg. of virus per cubic centimeter. The mixture
was stirred by being drawn up and down in a dropping pipet for about 15 seconds.
By this time the mixture had reached a temperature of about 100° and heating was
continued for 1 minute; the tube was then withdrawn and placed in an ice bath. The
contents of several tubes were usually pooled and spun at 7000 r.p.m. in an angle
centrifuge in order to remove coagulated protein. The clear supernatant fluid con-
taining the sodium nucleate was dialyzed overnight at 4° against 18 1. of flowing
distilled water in a Kunitz-Simms rocking dialyzer. The ultraviolet absorption at
260 m^ of the nucleate before and after dialysis was essentially identical, which
indicates that no significant quantity of dialyzable nucleic acid material was pro-
duced by the cleavage method or by subsequent dialysis. The dialyzed nucleate solu-
tion was concentrated to 0.07 to 0.02 of its volume by pervaporation, and the small
amount of insoluble matter appearing during concentration, together with small
quantities of soluble virus fragments, removed by centrifugation at 40,000 r.p.m.
(102,000 X g, average) for 1 hour in the No. 40.2 rotor of the Spinco model L centri-
fuge. The clear supernatant fluid was lyophilized and the residue was dried to con-
stant weight in a drying oven at 110° or in vacuo over P2O6 at 78°.

Various modifications of the above procedure were tried with results which may
be summarized as follows. Lithium chloride can be substituted for sodium chloride.
If the final concentration of salt is about 0.1 M, coagulation of the denatured protein
seems to be greatly aided by the presence of a small amount of phosphate, but, if a
salt concentration in the neighborhood of 0.3 M is used, the phosphate is dispensable.
In tests with up to 1 M of sodium chloride, the yields of sodium nucleate were found

« R. Markham and J. D. Smith, Biochem. J. 49, 401 (1951).

390 B. MAGASANIK

to diminish above 0.3 M, owing probably to the rapid coagulation of virus before
cleavage of the nucleic acid. Presumably for the same reason, salts of polyvalent
metals, such as magnesium or aluminum chloride, gave low yields of nucleate. Heat-
ing times from 15 seconds up to 10 minutes at 100° were investigated and it was found
that 1 minute was optimum, the yield of nucleate being virtually quantitative at this
point. The yield of nucleate was greatly reduced when, owing to the size and shape of
the reaction vessel or to a diminution of the heating, the temperature within the
reaction mixture failed to rise above 99°. Hence, the optimum temperature seems to
be about 100°, although temperatures higher than 100° were not studied, and the
critical temperature between 96-100° was not ascertained. The concentration of virus
in the heated mixture affected the cleavage, as previously noted by Cohen and Stan-
ley." Moreover, when the concentration was high, there appeared to be more nucleo-
protein in the final preparation than when the final virus concentration did not exceed
20 mg. per cubic centimeter of salt-virus mixture.

4. Isolation of PNA from Microbial Tissues

PNA was first isolated from yeast, and yeast is still the preferred start-
ing material for PNA preparations. However, in most procedures for the
isolation of yeast PNA alkali is used, and the preparations are of low
molecular weight and have lost a portion of their pyrimidine nucleotides.
It is questionable whether any of the procedures described so far for the
isolation of PNA from yeast can result in the production of a preparation
that has escaped major degradation. (3ne reason for the failure to obtain
better preparations from a material as rich in PNA as yeast, is the diffi-
culty of breaking the wall of the yeast cell. Unless the cell is dried by
extraction with organic solvents such as ethanol or acetone, it is impossible
to isolate the major portion of the nucleoproteins. This rough treatment is
in all likelihood responsible for an extensive degradation of the PNA. The
nucleoproteins may be extracted with saline, purified, and fractionated by
isoelectric precipitation, and PNA prepared from the nucleoprotein frac-
tions by treatment with 10% NaCl.-^-^^ Another method consists of
grinding the defatted yeast in a bacterial mill, followed by extraction of
the PNA with 10% NaCl. This method is given in detail below.^^ Loring
and his collaborators used' short treatment with dilute alkali in the cold to
extract PNA from yeast. ^^

Little is known about bacterial PNA. Bernheimer^^ was able to show
that pentose nucleic acids from Streptococcus 'pyogenes, Clostridium welchii,
and Escherichia coli were inhibitors of group A streptococcal deoxyribo-
nuclease; PNA preparations from mammalian tissues, wheat germ, and
yeast failed to show inhibition, while tobacco mosaic virus PNA was
inhibitory, but only in relatively high concentrations.

*^ Y. Khouvine and H. De Robichon-Szulmajster, Bull. soc. chim. hiol. 34, 1056

(1952).
<>» H. S. Loring, J. L. Fairley, and H. L. Seagran, J. Biol. Chem. 197, 823 (1952).
" A. W. Bernheimer, Biochem. J. 53, 53 (1953).

ISOLATION AND COMPOSITION OF PENTOSE NUCLEIC ACIDS 391

The isolation of pure PNA is made difficult by the presence of large
amounts of DXA in bacterial cells. This difficulty seems to have been
overcome in a recently described procedure where a mixture of DNA and
PNA was extracted from disintegrated cells of Mycobacterium tuberculosis,
Mycobacterium phlei, and of Sarcina hitea, and precipitated with cetyltri-
methylammonium bromide.^'" PNA could be separated from DNA by
fractionation of the cetyltrimethylammonium salts with 0.5 M sodium
chloride at 0° C.^^^ In a second method PNA was preferentially adsorbed
on charcoal from 0.14 M sodium chloride^' and subsequently eluted by
means of 15% phenol at pH 7-7.5.^^^ Very recently the isolation of PNA
of Mycobacterium phlei by extraction w^ith 5 % NaCl and precipitation by
acid has been described in a short communication.^^

(1) Preparation of PNA from Yeast.'** The preparation was made from freshly
ground, defatted baker's yeast. The yeast cells (95 g.) were washed with 0.14 M
NaCl, and then with 50, 75, 95, and 100% ethanol. Their suspension in equal volumes
of 0.14 M NaCl and absolute alcohol was passed through an ice-cooled wet crushing
mill for bacteria and 2 vol. (240 cc.) of 70% ethanol was added to the mixture. The
precipitate of mostly crushed cells was washed repeatedly with 80 and 90% ethanol,
ethanol-ether (1:1), and ether, and dried in vacuo. It was twice extracted with 100-cc.
portions of 10% aqueous sodium chloride at 90° for }^ hour, and 2 vol. of ethanol was
added to the combined centrifuged extracts. The resulting precipitate, washed with
dilute and absolute alcohol and ether and dried, weighed 0.71 g. It was taken up in
35 cc. of water, the mixture was centrifuged, and 0.25 vol. of 20% barium acetate
solution (pH 7) and 1 vol. of ethanol were added to the supernatant. The precipitate
resulting from the centrifugation of the chilled mixture was washed with 5% barium
acetate and its aqueous suspension (17 cc.) stirred in a high-speed mixer in the pres-
ence of a small excess (150 mg.) of sodium sulfate. The solution, clarified by centrifu-
gation, was freed of protein by being stirred six times in a high-speed mixer with
chloroform-octanol (9:1) and then was poured into 2}4 vol. of ice-cold ethanol (50
cc.) that was made 0.05 A^ with respect to HCl. The mixture was chilled overnight and
the precipitate, after being washed with alcohol, was suspended in 20 cc. of water
and brought into solution by the cautious addition of dilute ammonia to pH 6. The
precipitation with acidified alcohol was repeated and the nucleic acid washed with
80 and 100% alcohol and ether and dried, when 0.17 g. of an almost white powder was
obtained.

V. The Nature of PNA

The final step in the preparation of PNA consists in the precipitation of
the material with ethanol at pH 7.0, 4.2, or 1. Consequently, PNA is ob-
tained as sodium nucleate, acid sodium nucleate, or free PNA. Recent
procediH'es have favored the isolation as sodium nucleate. ^^ In this way

"» A. S. Jones, Biochim. et Biophijs. Acta 10, 607 (1953).

"b S. K. Dutta, A. S. Jones, and M. Stacey, Biochim. et Biophys. Acta 10, 613 (1953).

82 S. Zamenhof and E. Chargaff, Nature 168, 604 (1951).

"Y. Khouvine, M. Barbier, and L. Wyssmann, Compt. rend. 236, 2118 (1953).

" B. Magasanik and E. Chargaff, Biochim. et Biophys. Acta 7, 396 (1951).

392 B. MAGASANIK

the possible degradation of the material by acid can be avoided. More-
over, sodium nucleate dissolves in water to give a clear, colorless solution,
whereas free PNA is almost insoluble.

The purity and homogeneity of PNA preparations can be determined
by physical and chemical methods. The former are discussed in Chapters
13 and 14; so far they have not been used to any great extent for the char-
acterization of PNA. The behavior of a number of PNA preparations in
the ultracentrifuge has been recorded. Cohen and Stanley observed that
PNA, freshly isolated from tobacco mosaic virus, sedimented from a 1 .0 %
solution in 0.2 M NaCl at pH 4.9 with a constant S2o,w of 5.9." The mate-
rial was found to be inhomogeneous. Similarly high sedimentation con-
stants were obtained by Grinnan and Mosher''^ with a preparation of rat
liver PNA, and by Kay and Bounce with rabbit liver PNA.^^ Comparison
of these results is difficult as different concentrations of PNA as well as
solvents of different ionic strength and pH were used. The importance of
the composition of the solvent is shown by the observations of Grinnan
and Mosher: their preparation was polydisperse in sodium chloride solution
but monodisperse in water. Volkin and Carter-* observed a doubling of the
sedimentation constant of a preparation of rabbit liver PNA when the pH
was lowered from 6.7 to 4.8. At the higher pH, several of their preparations
from different animal organs appeared to sediment as a single boundary
with Sao.w of about 2.3.

The behavior of liver PNA, yeast PNA, and pancreas PNA on dial-
ysis has been studied by Magasanik and Chargaff*'* and by Kerr and
Seraidarian." In all cases the isolated PNA preparations contained dialy-
zable fractions amounting to 10-25% of the preparation. The dialyzable
material consisted of polynucleotides. In the case of yeast PNA the compo-
sition of this fraction did not differ materially from that of the undialyzable
material, while in the case of the animal PNA a marked difference was
observed.^*

These few observations indicate that PNA preparations are far from
homogeneous. The lack of homogeneity of these preparations may be due
in part to their decomposition on storage. The stability of PNA has not
been carefully investigated, but isolated observations indicate that they
are indeed very labile substances. Some preparations cannot be stored
even in the cold. Thus, freshly isolated PNA of tobacco mosaic virus
(Nucleate A) whose molecular weight was estimated at 300,000, spon-
taneously decomposed at 4° and pH 7.0 in less than one week to give
material with a molecular weight of approximately 60,000." Similarly, the
viscosity of freshly prepared solutions of highly polymerized rat liver PNA
decreased with time, indicating spontaneous degradation. ^^ Beef pancreas
PNA isolated by Kerr and Seraidarian could not be dried without partial
decomposition." It has been mentioned before that this pancreas PNA

ISOLATION AND COMPOSITION OF PENTOSE NUCLEIC ACIDS 393

TABLE II
The Ultraviolet Extinction at 260 mp of PNA

e(P) after

Preparation

*(P)

alkaline hydrolysis

Increase, %

Ref.

Mouse liver

8,190

10,700

31

45

Pig liver

8,600

11,100

29

54

Calf pancreas

7,750

9,570

24

26

Yeast

10,000

12,400

24

54

Yeast

8,700

11,900

37

54

preparation contained nucleotide fragments which could be removed by
dialysis in the cold. Continuation of the dialysis at room temperature
resulted in the appearance of more polynucleotides in the dialysate, indi-
cating lack of stability at room temperature.

The extinction of ultraviolet light is a characteristic property of nucleic
acids. [Cf. Beaven, Holiday, and Johnson, Chapter 14.] The chromophores
are the conjugated double bond systems of the purines and pyrimidines.
The ultraviolet extinctions of nucleic acids and nucleotides are conveniently
expressed as e(P), the extinction of a solution containing one gram-atom
of nucleotide phosphorus per liter." The e(P) values for a number of
different PNA preparations are summarized in Table II. It can be seen
that the ultraviolet extinction of the PNA preparations is less than the
sum of the extinctions of the mononucleotides to which they can be hydro-
lyzed by alkali. It seems that the polymerization of mononucleotides to
form PNA is accompanied by the suppression of certain chromophores.
The level of polymerization at which this effect becomes noticeable cannot
be defined. It is not due to the simple union of a few mononucleotides,
since the dialyzable polynucleotides show the same extinction as the mix-
ture to which they can be hydrolyzed by alkali. On the other hand, the
highly polymerized PNA preparations obtained by the use of sodium
dodecyl sulfate do not show a greater degree of suppression of extinction
than the less highly polymerized preparations obtained by salt extraction.
Still, the measurements of the increase in ultraviolet absorption obtained
by exposure of PNA to alkali would appear to be of value in the charac-
terization of such preparations.

The chemical methods used to characterize PNA include the estimation
of the nucleotide composition, which will be discussed separately, the
determination of nitrogen, of total and acid-hydrolyzable phosphorus, and
of pentose. The nitrogen and total phosphorus content differs with the
nature of the sample. Free PNA contains about 15-16% nitrogen and
8.5-9% phosphorus, while the values for sodium nucleates range around

" E. Chargaff and S. Zamenhof, J. Biol. Chem. 173, 327 (1948).

394 B. MAGASANIK

12% and 8%, respectively. More important than the actual values is the
atomic N/P ratio. A nucleic acid composed of equimolar amounts of the
four nucleotides would have an N/P ratio of 3.75. Preponderance of purines
over pyrimidines and of cytosine over uracil results in an increased N/P
ratio. PNA preparations from animal sources, particularly from pancreas,
have N/P ratios as. high as 4.3. N/P ratios higher than this are indicative
of contamination of the preparation with other nitrogenous substances,
generally with protein.

The amount of inorganic phosphorus obtained after hydrolysis of PNA
with dilute acid is a measure of the purine content of the preparation, for
only purine-nucleoside-bound phosphate is hydrolyzed under these condi-
tions. Similarly, the estimation of pentose by the orcinol reaction is a
measure of purine-bound pentose. Pyrimidine nucleotides are not hydro-
lyzed under the conditions of this test, and consequently, the pyrimidine-
bound pentose does not react with orcinol. [Cf. Dische, Chapter 9.]

It is usual to measure the amount of protein and DNA in the PNA
preparations by specific methods, such as the biuret test, and the diphenyl-
amine reaction, respectively. The PNA preparations which are obtained
by the isolation methods presented above are usually free from protein,
and do not contain more than 5% DNA. In general, contamination with
DNA is more effectively excluded when the nuclei are removed prior to
the isolation of PNA. The presence of DNA does not interfere with the
estimation of pentose nucleotides by paper chromatography or ionophore-
sis. [Cf. Chapters 7 and 8.]

The colorimetric estimation of pentoses by the orcinol method does not
differentiate between ribose and other pentoses. However, the nature of
the sugar component may be determined after hydrolysis by isolation and
conversion into derivatives, or by paper chromatography in various sol-
vents. By these methods the pentose components of PNA preparations
isolated from mammalian organs,^ '^^'^^ from tobacco mosaic virus, ^^^ from
cucumber virus CV3," and from C. perfringens,^^ have been unequivocally
identified as D-ribose. At present, pentose nucleic acids isolated from these
sources may safely be called ribonucleic acids.

VI. The Nucleotide Composition of PNA

1. General Considerations

The analysis of purified DNA preparations isolated from a great variety
of cells has led to the conclusion that (a) the deoxypentose nucleic acids
of different species of organisms have different nucleotide compositions,
(b) those of different organs of the same species have identical composi-
tions, and (c) the ratios of adenine to thymine and guanine to cytosine

" R. Markham and J. D. Smith, Biochem. J. 46, 513 (1950).

" D. L. MacDonald and C. A. Knight, /. Biol. Chem. 202, 45 (1953).

ISOLATION AND COMPOSITION OF PENTOSE NUCLEIC ACIDS 395

(including methylcytosine) are always unity. ^* The implications of these
findings are discussed in Chapter 10 and elsewhere in this book. Here the
results obtained by the analysis of PNA preparations will be examined
with the view to finding out whether similar or different generalizations
can be made with regard to the composition of PNA.

2. Analytical Procedures

The methods used for the estimation of nucleic acid components have
been discussed in Chapters 5 to 9. Here only a brief description will be
given of those methods which served to obtain most of the results pre-
sented in the succeeding sections.

Method 1.^^ Hydrolysis of PNA by N HCl to a mixture of purine bases and pyrimi-
dine mononucleotides followed by separation of the components of the mixture by
paper chromatography, elution, and spectrophotometric estimation.

Method 2^°- " Hydrolysis of PNA by dilute alkali to a mixture of mononucleotides,
separation of the mixture by paper chromatography, followed by elution and spec-
trophotometric estimation. In this method buffered isobutyric acid is used as the
organic phase in chromatography. Guanylic acid and uridylic acid, which occupy the
same position on the chromatogram, are eluted together and their respective concen-
trations determined from the extinction values of the mixed eluate at two different
wavelengths by means of simultaneous eciuations. The value for the molecular extinc-
tion of guanylic acid in M phosphate buffer of pH 7.0 which was used by the authors
to calculate the concentration of guanylic acid and of uridylic acid was obtained
from measurements carried out on an impure sample of sodium guanylate and cor-
rected according to the nitrogen content of this sample. Recent determinations of
the molecular extinction of guanylic acid purified bj' ion-exchange chromatography
have shown that the value used was too high. [Cf. Chapter 14.] Consequently the
results of the analysis of pentose nucleic acids by Chargaff et «/."< and by Magasanik
and Chargaff" report values for guanylic acid which are too low. The compositions
of these preparations have been recalculated using the following extinction coeffi-
cients taken from a recent paper of Elson et al.^^ (Ae is the difference between the
molecular extinction at the wavelengths indicated and that at 290 fi.)

Nucleotide \,mn At

Adenylic acid 260 15.12

Guanylic acid 265 7.18

252.5 10.74

245 8.87

Cytidylic acid 270 6.87

Uridylic acid 265 9.49

261 9.80

245 5.47

58 E. Chargaff, Experientia 6, 201 (1950): Federation Proc. 10, 654 (1951).

" J. D. Smith and R. Markham, Biochem. J. 46, 509 (1950).

60 B. Magasanik, E. Vischer, R. Doniger, D. Elson, and E. Chargaff, J . Biol. Chem.

186, 37 (1950).
«' D. Elson, T. Gustafson, and E. Chargaff, J. Biol. Chem. 209, 285 (1954).

396 B. MAGASANIK

The results reported in the subsequent sections were calculated using these values.

Method S.2* Hydrolysis of PNA by dilute alkali to a mixture of mononucleotides,
separation of the mixture by chromatography on a basic ion-exchange resin (Dowex
1) [compare Cohn, Chapter 6], elution with acid, and spectrophotometric estimation.
In this method the 2'- and 3'-phosphates of the purine nucleosides, which are formed
by the action of alkali, are separated. The results recorded here give the sum of the
2'- and 3'-phosphates of adenine or guanine.

Method 4-^'^- *' Hydrolysis of PNA by dilute alkali, followed by separation of the
mixture of mononucleotides by electrophoresis on filter paper, elution, and spectro-
photometric estimation or phosphorus analysis. [Cf. Smith, Chapter 8.]

Method 5.**' ®^ Hydrolysis of PNA by N H2SO4 to a mixture of purines and pyrimi-
dine nucleotides. The purines are precipitated as insoluble silver salts; these are de-
composed by hydrochloric acid, and the concentration of adenine and of guanine in
the supernatant solution determined by measurement of its ultraviolet extinction at
two wavelengths. The concentrations of the pyrimidine nucleotides in the super-
natant solution of the purine silver salts are similarly determined. Alternatively the
pyrimidine nucleotides may be converted to nucleosides by treatment with prostatic
phosphatase prior to spectrophotometric estimation. [Compare also Loring, Chapter
5.]

Methods 1, 3, and 4 permit the separation of the four nucleic acid com-
ponents and are therefore presumably superior in accuracy to methods 2
and 5. The recovery of pyrimidine nucleotides in method 1 may be low,
even when the correction described by the authors is applied. ^^

In general the reliability of a method is judged by the completeness with
which the sum of the products obtained will account for the phosphorus
and nitrogen content of the PNA preparation. The methods described
allow 90-100 % of the nitrogen and the phosphorus content to be accounted
for as bases or nucleotides. Usually the composition of the PNA sample is
presented as the fraction of PNA phosphorus accounted for in each nucleo-
tide. However, in order to facilitate the comparison of the composition of
different PNA preparations it is advantageous to express the results as
the molar ratios of the nucleotides relative to adenine as 10.** It is in this
manner that the composition of PNA preparations is presented in the
succeeding sections.

3. PNA FROM Animal Tissues

A sufficiently large number of animal PNA preparations has been isolated
and analyzed to permit consideration of the cjuestion whether animal
pentose nucleic acids are species-specific or organ -specific or fall into no
clearly discernible pattern. The composition of PNA preparations isolated

«2 J. N. Davidson and R. M. S. Smellie, Biochem. J. 52, 594 (1952).
«3 G. W. Crosbie, R. M. S. Smellie, and J. N. Davidson, Biochem. J. 54, 287 (1953).
6^ S. E. Kerr, K. Seraidarian, and M. Wargon, J. Biol. Chem. 181, 761 (1949).
" H. S. Loring, J. L. Fairley, H. W. Bortner, and H. L. Seagran, J. Biol. Chem. 197,
809 (1952).

ISOLATION AND COMPOSITION OF PENTOSE NUCLEIC ACIDS 397

TABLE III
Nucleotide Composition of PNA Isolated from Liver

Prepa- Adenylic Guany- Cytidylic Uridylic

ration Animal acid lie acid acid acid Pu/Py Method" Ref .

1

Rabbit

10

16.9

14,6

10.3

1.08

4

63

2

Rabbit

10

19.7

4

63

3

Rabbit

10

17.0

4

63

4

Rabbit

10

16.5

4

63

5

Rabbit

10

15.4

4

63

G

Rabbit (preg-
nant)

10

15.8

15.2

11 1

0.98

4

63

7

Rabbit (fetal)

10

15.4

15.6

10.0

0.99

4

63

8

Rabbit

10

20.2

16.8

9.9

1.13

3

25

9

Rabbit

10

19.4

16.1

9.5

1.12

3

25

10

Rat

10

17.5

13.9

10.9

1.10

4

63

11

Rat

10

17.6

14.3

10.8

1.10

4

63

12

Rat (regenerat-
ing)

10

16.6

14.1

10.2

1.08

4

63

13

Rat

10

18.3

18.9

8.5

1.04

3

25

14

Rat (regenerat-
ing)

10

19.0

18.3

9.3

1.05

3

25

15

Beef

10

17.0

10.4

7.8

1.49

2

44

16

Beef (calf)

10

18.8

10.8

6.7

1.65

2

44

17

Beef (calf)

10

17.9

14.9

8.4

1.20

3

25

18

Beef (calf)

10

18.7

16.0

8.7

1.17

3

25

19 Chicken

10

17.1

13.6

10.6

1.12

4

63

20 Mouse

10

16.2

12.9

8.6

1.22

3

25

21 Sheep

10

19.4

12.7

7.1

1.49

2

44

22 Pig

10

18.8

15.3

9.1

1.23

2

44

23 Man

10

38.6

27.5

11.0

1.26

2

44

° See page 395.

from the liver of a variety of species of animals is presented in Table III
and will be considered first. It can be seen that the liver PNA preparations
of all the animals studied are rich in guanylic acid and cytidylic acid, and
poor in adenylic acid and uridj'lic acid. The ratio of purines to pyrimidines
is generallj' not far from unity. The preparations exhibit considerable
variation in nucleotide ratios, but no striking differences in the composi-
tion of liver PNA of different animals indicative of species-specificity can

398 B. MAGASANIK

TABLE IV
Nucleotide Composition of PNA Isolated from Different Organs of Calf

Prepa- Adenylic Guanylic Cytidylic Uridylic

ration Organ acid acid acid acid Pu/Py Method" Ref.

1 Liver 10 17.9 14.9 8.4 1.20 3 25

2 Pancreas 10 34.5 16.8 9.5 1.69 3 25

3 Spleen 10 19.7 17.7 8.6 1.13 3 25

4 Thymus 10 23.8 13.9 6.5 1.65 3 25

° See page 395.

be discovered. Some of the observed differences are undoubtedly due to
differences in the method of isolation. Thus different values are reported
for the cytidylic acid content of calf liver PNA isolated in different labora-
tories (preparations 16 and 17). Of greater interest are the differences in
the composition of PNA preparations isolated by the same method. A
large number of preparations were isolated from the livers of different
rabbits and rats by Davidson and his collaborators.®^ The individual
variations in the composition of liver PNA of different rabbits (prepara-
tions 1-7) were found to be greater than the differences between the average
composition of rabbit liver PNA and rat liver PNA. The only PNA prepa-
ration which differs significantly from all the others in composition is that
isolated from a human liver (preparation 23). It seems, however, unwar-
ranted to attach significance to an isolated observation.

In Table IV the composition of PNA preparations obtained from differ-
ent organs of the same animal are compared. Calf spleen PNA can be seen
to differ little from liver PNA, whereas thymus and particularly pancreas
PNA are considerably richer in guanylic acid. In consequence, the purine-
to-pyrimidine ratios of calf thymus PNA and of pancreas PNA approach
2.0. The high guanylic acid content of pancreas PNA has also been ob-
served in other laboratories. However, the preparations isolated by different
procedures vary greatly in composition (Table V). This variation may
partly be ascribed to the extent of degradation by pancreatic ribonuclease
which PNA undergoes during the course of the isolation procedure. The
influence of the action of the enzyme on the composition of the final pro-
duct is clearly demonstrated by comparison of preparations 3 and 4 in
Table V. The former was prepared in the usual manner, while the latter
was isolated from the tissue after removal of ribonuclease by extraction
with dilute acid and acetone ;^^ preparation 4 is considerably richer in
pyrimidines and seems to be essentially identical in composition with
liver PNA.

The action of pancreatic ribonuclease on PNA preparations isolated

ISOLATION AND COMPOSITION OF PENTOSE NUCLEIC ACIDS

399

TABLE V
Nucleotide Composition of PNA Isolated from Pancreas

Aden-

Guan-

Cytid-

Urid-

Prepa-

ylic

ylic

ylic

ylic

ration

Animal

acid

acid

acid

acid

Pu/Py

Method"

Ref.

1

Beef

10

42

32

10

1.2

5

65a

2

Beef (calf)

10

34.5

16.8

9.5

1.7

3

25

3»

Beef

10

21

15

2.0

d

38

4.

Beef

10

18

30

1.0

d

38

5

Pig

10

26.7

9.3

4.9

2.6

2

44

" See page 395.
'' Prepared in usual way.
'^ Ribonuclease removed prior to isolation.
Indirect colorimetric methods.

TABLE VI
Nucleotide Composition of PNA "Cores" Resistant to Pancreatic

Ribonuclease

Prepa-
ration

Tissue

Aden- Guan- Cytid-
ylic ylic ylic

acid acid acid

Uridylic
acid Pu/Py Method" Ref.

1

Beef pan-
creas

10

23

15

2.2

b

38

2

Pig liver

10

49.1

8.2

6.7

4.0

2

54

3

4

Yeast
Yeast

10
10

43
14

4

10

7

4.8
2.4

2

6

54
38

" See page 395.

' Indirect colorimetric methods.

from liver or from yeast results in the formation of pyrimidine mononucleo-
tides and of polynucleotides of different sizes consisting mostly of purines
[see Chapter 15]. The largest of these polynucleotides, the so-called "core,"
is very rich in guanylic acid and poor in pyrimidine nucleotides (Table VI).
Liver PNA is thus transformed by the action of pancreatic ribonuclease
into polynucleotides similar in composition to pancreas PNA. It is there-
fore quite possible that "native" pancreas PNA is of the same composition
as liver PNA, but that the action of the ribonuclease converts this native
PNA during the isolation procedure into a degraded product of high gua-
nylic acid and low pyrimidine nucleotide content.

65" S. E. Kerr, K. Seraidarian, and M. Wargon, J. Biol. Chem. 181, 773 (1949).

400 B. MAGASANIK

TABLE VII
Nucleotide Composition of PNA from Miscellaneous Animal Sources

Aden-

Guan-

Cytid-

Urid-

Prepa-

ylic

ylic

ylic

ylic

ration Tissue

acid

acid

acid

acid

Pu/Py

Method"

Ref.

1 Cat brain

10

14.7

12.0

9.5

1.15

2

66

2 Carp muscle (Nu-

10

21

19

11

1.0

1

29

cleotropomyosin)

3 Sea urchin eggs

10

13.3

12.3

9.3

1.07

2

61

4 Starfish eggs

10

15

14

11

1.0

1

8

" See page 395.

Table VII presents the composition of PNA preparations isolated from
miscellaneous animal tissues. It is of interest that the PNA of cat brain*^
and of carp muscle tropomyosin-^ have essentially the same composition
as liver PNA. The PNA of starfish^ and of sea urchin^^-^^ eggs similarly
possess the "high guanylic acid, high cytidylic acid" pattern found in all
other PNA preparations isolated from the tissues of animals. The composi-
tion of the sea urchin egg PNA did not change after fertilization during
the course of 48 hours of embryonic development.*^-"

The problem of the heterogeneity of PNA in different cell fractions has
been studied in several laboratories.*^""*^ The cell nuclei were isolated in
citric acid or by centrifugation of a tissue mince at 700 X g. The elements
of the cytoplasm were fractionated by centrifugation (see Chapter 21).
PNA was not isolated, but the total polyribonucleotide composition of the
fractions determined directly. Some of the results obtained in these studies
are presented in Table VIII. There are no significant differences in nucleo-
tide composition between the different cytoplasmic fractions, which,
however, appear to differ from the nuclear fractions. The familiar "high
guanylic acid, high cytidylic acid" pattern of animal PNA was found in
every case. However, the nuclear PNA composition of different animals of
the same species seemed to vary appreciably. For instance, the composi-
tion of preparation 1 , which is the average of the nuclei obtained from the
livers of four rats, contains more guanylic acid than most of the other
preparations from nuclei. Even higher values for guanylic acid, which
would make the composition of nuclear PNA more closely similar to that
of cytoplasmic PNA, were found by Mclndoe in many batches of nuclei
(unpublished results quoted by Crosbie elal.)}^ The composition of nuclear

"H. A. Deluca, R. J. Rossiter, and K. P. Strickland, Biochem. J. 55, 193 (1953).

"D. Elson and E. Chargaff, Phosphorus Metabolism 2, 329 (1952).

«8 A. Marshak, J. Biol. Chem. 189, 607 (1951).

" W. M. Mclndoe and J. N. Davidson, Brit. J. Cancer 6, 200 (1952).

ISOLATION AND COMPOSITION OF PENTOSE NUCLEIC ACIDS

401

li

P-,

05 03 CO t~~ t-- l~-

CO CO CO CO CO CO

-^ -^ Tf< (M iM <N
-H 00 -^ ■<** t^ CO

Ci 03 00 00 GO CO

05 t^^ (M lOOO

coo

O -—I

oeo

CO coco CO
O O O -H

CO ■* 05 CO

coco t- t^
CO CO coo

CO"0 00(N

05
CO

oo
d

O

3 1.00 4 63

4 0.98 4 63

coco
coco

CO CO
coco

(MOO

(M .-I rt C^ -H Tt<

CO .— 1 lO CO CJ CO

-H O

0 002 05
l~-. ^ 1— 1 r^

^0 05 05

CO t^»0 05

■<ti

oo
CO >c

OO

OOi

coco

•* c^ 1— 1 -^ -^ 00

GO-* o t-o o

■^ CO O 05 CO <M

oooooo

Q^ Q^ QJ Q^ Q^ Qj

'« "« "o '« "« "o
3 3 3 3 3 3

^^ >:> ^^ -*^ ^_5 -^
oj c3 oj c^ 02 Gj

.a

1— ( C^ CO -^ lO ^

C30

CO t^

oo
.2 2

-OT3

c c
O o

o o
t>.oo

9 Rat Microsomes 10 16.9 14
10-^ Rat Microsomes 10 16.8 15

11 Rat Microsomes 10 17.9 17

12 Rat Microsomes 10 18.2 16

■*■<*< CO "5

oi ojco o
CO t^ t^ 00

OOOO

a, a.aa

e3 03 ci3 e3
00 tn oo (c

"3 "^"^ "^
OOOO

-^ -^ +J -*^

^ cj oj oj
CO -^ lO CO

CO

CO

CO

o

3

13

c3

IC"0

CT> •*

oo

2 2

-OTD
G C

o o
o o

SI

03 03

00 Oi

CO -^
»CcO

OO

72 tC

a s

o o

OQ CO

O O

03 03
(Mc5

iC"5

OO

aa

03 03
OQ 09

OO

33

03 03
?3?5

0, oj C

« > «
M < rt

402 B. MAGASANIK

PNA was also studied by Marshak^^ by a method in which PNA was
hydrolyzed to a mixture of bases with hot perchloric acid/" This procedure
was later found to lead to an incomplete hydrolysis of the pyrimidine
nucleotides and to poor recoveries of cytosine and uracil.^* The use of this
method may account in part for the exceptionally low cytosine and uracil
(;ontent of nuclear PNA reported by Marshak.^^

4. PNA FROM Microbial Tissues

The composition of yeast PNA has been determined in many labora-
tories. Most investigators who were intent upon developing methods for
the estimation of the nucleotide composition of PNA used commercial
yeast nucleic acid, purified to various degrees, as an object for testing
their method. However, it has been mentioned earlier that all commercial
yeast PNA preparations have been isolated by extraction with alkali, and
consequently are badly degraded. The composition of such preparations
cannot be considered to represent yeast PNA. For this reason only the
composition of yeast nucleic acid specimens isolated in the laboratory by
mild procedures will be considered here.'*^^^'^° The composition of five
preparations of PNA from baker's yeast are shown in Table IX. In all
cases the yeast was dried with ethanol before extraction of PNA. It can
be seen that these five preparations are quite similar in composition, al-
though the method of isolation was different, except for preparations 4
and 5. Preparation 6, which was isolated from brewer's yeast seems to be
of slightly different composition."*^ The composition of yeast PNA ap-
proaches a "statistical tetranucleotide" more closely than that of animal
PNA. However, baker's yeast PNA was in all cases found to be somewhat
richer in guanine than in adenine, and somewhat richer in uracil than in
cytosine. It must be kept in mind that all purified preparations of yeast
PNA account for only a portion of the PNA originally present in yeast.
Therefore the good agreement in the results obtained may simply indicate
that the same portion of yeast PNA is isolated in the procedures used. A
different procedure of drying, such as substituting acetone for ethanol, led
to a PNA preparation with a guanine content of 14.3, almost one-third
higher than that of the ethanol-dried preparations.^^ An attempt was made
by Khouvine and her collaborators to investigate the heterogeneity of
baker's yeast PNA by separating the PNA-proteins isolated from acetone-
ground yeast into several fractions by isoelectric precipitations.-^ The
nucleotide composition of the fractions precipitated at pH 5 and 4.3 closely
resembled that of the preparations presented in Table IX, whereas the
nucleotide composition of the fraction precipitated at a pH of 2.3 was
richer in giianylic acid.

" A. Marshak and H. J. Vogel, J. Biol. Chem. 189, 597 (1951).

ISOLATION AND COMPOSITION OF PENTOSE NUCLEIC ACIDS 403

TABLE IX
Nucleotide Composition of PNA from Microorganisms

Aden-

Guan-

Cvtid-

Urid-

Prepa-

ylic

ylic

ylic

ylic

Me-

ration

Organism

acid

acid

acid

acid

Pu/Py

thod"

Ref.

1

Baker's yeast

10

12.0

8.0

9.8

1.23

2

44

2

Baker's yeast

10

11.9

7.2

11.4

1.17

2

44

3

Baker's yeast

10

10.9

8.5

9.5

1.16

5

50

4

Baker's yeast
(Springer Sp5)

10

11.4

7.5

9.0

1.30

1

49

5

Baker's yeast
(Koenig-Gist)

10

11.1

7.1

8.9

1.32

1

49

6

Brewer's yeast
(Karcher)

10

10.4

9.2

9.6

1.08

1

49

7

Mycobacterium
phlei

10

15.5

8.5

3.5

2.13

1

53

8

Serratia viarces-
cens

10

10.2

8.5

8.3

1.2

2

71

9

Escherichia coli

10

10.2

8.5

8.3

1.2

2

71

10

Clostridium per-
fringens

10

11.6

9.4

7.4

1.29

2

22

" See page 395.

The only PNA isolated from a bacterial species whose nucleotide com-
position has so far been determined is that of M ycobacterium phlei.^^ The
composition of PNA of Serratia marcescens'^ and of Escherichia coli,^^ and
of a nucleoprotein of C. perfringens^^ were determined without prior isola-
tion of the nucleic acids. The results of the analyses are shown in Table IX.

5. PNA FROM Plant Viruses

The composition of a large number of plant virus PNA preparations has
recently been determined by Knight and his collaborators'^"''^ and by
Markham and Smith.'*^ The results are of particular interest since in the
case of plant viruses we can accept them as describing the composition of
native virus PNA without reservations regarding degradation or fraction-
ation during isolation. The purity of the virus preparations can be ascer-
tained by physical methods, and the nucleotide composition of the isolated
PNA was found to agree closely with that of the virus, determined without
prior isolation of PNA. The results of these determinations are presented

" D. Elson and E. Chargaff, Nature 173, 1037 (1954).

" R. W. Dorner and C. A. Knight, J. Biol. Chem. 205, 959 (1953).

" F. L. Black and C. A. Knight, J. Biol. Chem. 202, 51 (1953).

404

B. MAGASANIK

X
m

t3

Z
O

o
O

ci

Cm

Oh

o

CI X

<

a. 2

Ct^

O O Oi Ol o o

00 00 05 05 05 C^
QO QO 00 00 00 05

(M "3 C<J O IM <N

CO cO O ^ ^ ^

iC O '*' 05 «D 00

00 00 00 00 00 00

o o o o o o

^ « ;2: -< -< p^
H S ^ o ^ a

o o o « «

CO tc

o p

^ o3 c3 ^

tB 02 CO oc

O O O O

s a s s

O O O O O Q

U « O tJ CJ o

Si -O

H H

^ ^ c3 c3

^ JD X5 -Q

o o o o

H H H H

^ (M CO -^ O ^

0) 0)

s s

3 3

3 3

O O

m

m

■J3

3
O

ISOLATION AND COMPOSITION OF PENTOSE NUCLEIC ACIDS 405

in Table X. It can be seen that different viruses have completely different
nucleotide composition. For instance, cytidylic acid varies from 6.6 in
potato virus X to 16.8 in turnip yellow mosaic virus. There does not appear
to be a regularity comparable to the constant purine-to-pyrimidine ratio
of 1, observed in all DNA preparations. [Compare Chapter 10.] The most
striking observation made by Knight and his collaborators is that all
sixteen strains of tobacco mosaic virus, are identical in composition.^'"
Similarlj", the nucleotide composition of two strains of cucumber virus is
identical, but different from that of the closely related tobacco mosaic
virus.*' These results present impressive evidence that the PNA of plant
viruses is species-specific, and that strains which have developed by muta-
tion of the parent strain do not differ from it in nucleotide composition.
A similar relation was found in the DNA of bacteria: different species
possess nucleic acids of different composition, but different strains of the
same species yield specimens of the same composition.^^ It has recently
been shown that the DNA preparations are composed of a large number
of molecules of different composition.''^ The DNA composition of the DNA
preparation isolated from an organ or a group of cells represents the aver-
age composition of the individuals present in this material. It is quite
possible that the viral PNA is similarly composed of a population of mole-
cules of different composition. It would appear then that all strains of the
same virus contain these nucleic acids in the same proportions.

The species-specificity of PNA in plant viruses agrees well with the
postulate of a genetic function for this PNA. The plant viruses are able
to bring about their own production in susceptible cells. Consequently,
they must possess a genetic determinant. The lack of DNA, which is con-
sidered to be the carrier of heredity in most organisms, suggests that this
function is carried out by the PNA present in the virus.

6. Conclusions

The results of the determinations of the nucleotide composition of PNA
preparations which have been presented in the preceding sections are too
scattered, and even contradictory, to permit any definite statements re-
garding regularities in PNA composition to be made. However, a few
tentative conclusions can be drawn from the results obtained so far, and
these may point out the directions for future work.

(1) PNA may he species-specific. PNA is clearly species-specific in plant
viruses: different viral species possess PNA of characteristically different
composition, while different strains of the same species possess PNA of

''* B. Gandelman, S. Zamenhof, and E. Chargaff, Biochim. et Biophys. Acta 9, 399

(1952).
" E. Chargaff, C. F. Crampton, and R. Lipshitz, Nature 172, 289 (1953).

406 B. MAGASANIK

identical composition. Yeast PNA and animal PNA differ in composition
from one another and from viral PNA. However the variations encountered
in the composition of PNA preparations isolated from different animals of
the same species are as great as the variations encountered in animals of
different species. These individual variations in PNA composition remain
unexplained. It is unlikely that they are due to genetic differences, for then
greater differences between species would be expected. They do not seem
to depend on the metabolic state of the organ from which the PNA is
isolated, as preparations isolated from normal and from regenerating rat
liver did not differ appreciably in composition. It will be necessary to
determine critically the range of variation of PNA composition within
members of the same species before fruitful comparisons of PNA prepara-
tions from different species can be made.

(2) PNA is presumably not organ-specific. So far no convincing differ-
ences in the composition of PNA preparations isolated from different
organs of the same animal have been observed. The high guanylic acid and
low pyrimidine nucleotide content of pancreas PNA seems to be due to the
extensive degradation by pancreatic ribonuclease during isolation, and not
to reflect a difTerence in composition between native PNA from pancreas
and from other organs.

(3) The nuclear and cytoplasmic fractions of cells differ in PNA composi-
tion. PNA isolated from whole organs consists almost exclusively of cyto-
plasmic PNA, as the nuclei are largely removed during isolation, and as
nuclear PNA accounts for only 10 % of the total PNA of the cell. [Compare
Chapters 18, 19, and 21.] Cytoplasmic PNA does, therefore, not differ
appreciably from "whole organ PNA" in composition. The PNA of rat
liver nuclei appears to be richer in uridylic acid and poorer in guanylic
acid than cytoplasmic PNA. Nuclei isolated from the livers of different
rats show greater variation in composition than the corresponding cyto-
plasmic fractions. Nuclear PNA is known to be more active metabolically
than cytoplasmic PNA . The correlation of metabolic activity with changes
in nucleotide composition might prove to be of interest.

(4) The composition of animal, yeast, and bacterial PNA shows certain
regularities. The nucleotide ratios of viral PNA preparations do not fall
into an easily discernible pattern. On the other hand, both yeast PNA and
PNA of the cytoplasm of animal cells contain purine and pyrimidine
nucleotides in nearly equimolar quantities. In animal PNA, guanylic and
cytidylic acids predominate, while in yeast PNA the four nucleotides are
present in nearly equimolar concentrations. Recently Elson and Chargaff^^
have pointed out that the regularities in the composition of PNA can be
demonstrated in a more striking fashion w^hen whole cells or centrifugally
prepared cell fractions are subjected to hydrolysis and analysis ensuring

ISOLATION AND COMPOSITION OF PENTOSE NUCLEIC ACIDS 407

the quantitative recovery of PNA as mononucleotides. In this way changes
in nucleotide composition due to the great susceptibility of nucleic acid to
enzymic and other degradations during isolation are avoided. These authors
were able to show that all pentose nucleic acids analyzed in their labora-
tory, including specimens from bacteria, yeast, and sea urchin and starfish
eggs, as well as from the cytoplasm and the nuclei of vertebrate cells,
possessed an ecjual number of 6-keto (guanine and uracil) and 6-amino
(adenine and cytosine) groups. The preparations from animal cytoplasm,
marine eggs, and yeast showed the additional regularity of a purine-to-
pyrimidine ratio of 1. These relations are well illustrated in the examples
taken from the work of these authors as well as from that of others pre-
sented in Tables VII, VIII, and IX (preparations 7-10). It is of interest to
compare the composition of PNA and DNA of animal cells. PNA is com-
posed of equal amounts of guanylie acid and cytidylic acid, and of adenylic
acid and uridylic acid, with the first pair predominating.^^ DNA is com-
posed of ecjual amounts of guanylie acid and cytidylic acid, and of adenylic
acid and thymidylic acid, but the second pair predominates.^** The com-
position of DNA is compatible with a double-stranded helical structure,
proposed by Watson and Crick on the basis of X-ray diffraction measure-
ments.''^ A similar correlation of physical and chemical properties of PNA
has not yet been made.

These are the conclusions which can be drawn from the work of the last
seven years, during which modern, accurate methods of nucleotide analysis
were used. Further progress will not depend on the development of more
accurate and efficient methods of analysis, but rather on new approaches
to the problem of the isolation and physical characterization of PNA.
Until homogeneous PNA preparations can be isolated, the knowledge of
the nucleotide composition of individual PNA specimens will not greatly
further the study of the structure and biological role of PNA. The most
fruitful approach to the problem of PNA structure has been the study of
the degradation products produced enzymically from PNA. The results of
such studies, which are discussed in Chapter 15, have shown that animal,
yeast, and viral PNA preparations are broken down by ribonuclease into
qualitatively very similar mixtures of mono-, di-, and polynucleotides.
For example, the largest nucleotide fragments are in all cases extremely
rich in guanylie acid. It would thus appear that these PNA preparations,
although differing in composition, are constructed according to very similar
patterns. The definition of these patterns remains a task for future research.

'" J. D. Watson and F. H. C. Crick, Nature 171, 737 (1953).

CHAPTER 12

Evidence on the Nature of the Chemical Bonds
in Nucleic Acids

D. M. BROWN AND A. R. TODD

Page

I. Introduction 409

II. Chemistry of the Ribonucleic Acids 413

1. Mononucleotides 413

2. Esters of the Mononucleotides 415

3. General Structure of Polyribonucleotides based on Chemical Degradation 418

4. Evidence for Nucleoside-S'-linkages • . . , . 422

5. Chemistry of Ribonuclease Action 423

6. Structure of Oligonucleotides derived from Ribonucleic Acids 428

7. Evidence for 3'(6) -linkages in Ribonucleic Acids 430

8. Chain-branching in Ribonucleic Acids 432

9. Nucleotide Sequence in Polyribonucleotides 436

III. Structure of the Deoxyribonucleic Acids 439

1. Nucleotide Sequence in Deoxyribonucleic Acids 442

I. Introduction

Since the emergence of reasonably clear ideas on the nature of the simple
nucleotides, and the recognition that the nucleic acids can be regarded as
polynucleotides, many attempts have been made to formulate general struc-
tures for them. In all such attempts the nature of the internucleotidic link-
age has occupied a central position and at one time or another most of the
possible types have been considered. The earlier developments in this field
have been adequately treated in a number of reviews,'"® and since they are
primarily of historical interest they will not be discussed here. Fortunately,
work in recent years has led to precise and generally accepted views on at
least the major portion of the internucleotidic linkages present in both
ribonucleic and deoxyribonucleic acids; it is therefore only necessary to

' P. A. Levene and L. W. Bass, "The Nucleic Acids." The Chemical Catalog Co.,

New York, 1931.
« R. S. Tipson, Advances in Carbohydrate Chem. 1, 193 (1945).
'J. M. Gulland, J. Chem. Soc. 1938. 1722.

* J. M. Gulland, J. Chem. Soc. 1944, 208.

' J. M. Gulland, Nucleic acids, Symposia Soc. Exptl. Biol. 1, 1 (1947).

* F. Schlenk, Advances in Enzymol. 9, 455 (1949).

409

410 D. M. BROWN AND A. R. TODD

consider the development of these modern views and to mention briefly
suggestions which have been made regarding other types of Hnkage which
might possibly be present to a minor extent.

Electrometric titration of monoribonucleotides and of ribo- and de-
oxyribonucleic acids by Levene and Simms^ led directly to the general con-
cept I for nucleic acid structure, in which the nucleoside residues were bound
together by phosphodiester linkages, and eliminated from consideration
ether or pyrophosphate linkages. This simple formulation of the internucleo-
tidic linkage has now become generally accepted, and recent work has
served to define the position of the linkage points in the nucleoside residues.*
These developments will form the main theme of this chapter, but before
turning to them mention may be made of the recorded evidence for other
types of linkage.

I
Base — sugar — PO (OH)

I
Base— sugar— PO(OH)

!

Base— sugar— PO (OH)

Electrometric titration of ribonucleic acids using refined techniques^"^^
and dye-binding studies^^'^'' led various workers to the conclusion that the
amount of secondary phosphoryl dissociation they exhibit is greater than
would be expected of a high-molecular- weight linear polydiester of type I.
This led Gulland et aZ.'"''^'^^ to propose structures (e.g., la) incorporating
a number of phosphotriester linka.ges, with a consequent increase in the
number of phosphomonoester end groups. It is perhaps significant that
recent studies using nucleic acids isolated by very mild procedures have
tended to minimize the proportion of secondary phosphoryl dissociation. ^^-^^
The general lack of agreement on this point probably stems from the use by

■> P. A. Levene and H. S. Simms, J. Biol. Chevi. 65, 519 (1925); 70, 327 (1926).

8 A. R. Todd, Angew. Chem. 65, 12 (1953).

9 F. W. Allen and J. J. Eiler, J. Biol. Chem. 137, 757 (1941).

i" W. E. Fletcher, J. M. Gulland, and D. O. Jordan, J. Chem. Soc. 1944, 33.

" Y. Khouvine and J. Gr^goire, Bull. soc. chim. biol. 26, 424 (1944).

12 H. Chantrenne, Bull. soc. chim. Beiges 55, 5 (1946).

'3 L. F. Cavalieri, S. E. Kerr, and A. Angelos, J. Am. Chem. Soc. 73, 2567 (1951).

'" L. F. Cavalieri, A. Angelos, and M. E. Balis, J. Am. Chem. Soc. 73, 4902 (1951).

15 J. M. Gulland and D. O. Jordan, Nature 161, 561 (1948).

16 J. M. Gulland, Cold Spring Harbor Symposia Quanl. Biol. 12, 95 (1947).
1' C. A. Zittle, J. Biol. Chem. 166, 491 (1946).

18 S. Weiner, E. L. Duggan, and F. W. Allen, J. Biol. Chem. 185, 163 (1950).

19 L. Vandendriessche, Compt. rend. trav. lab. Carlsberg 27, 341 (1951).

CHEMICAL BONDS IN NUCLEIC ACIDS 411

different workers of a variety of rather ill-characterized nucleic acid prep-
arations.

I
Base— sugar— PO(OH) P0(0H)2

I 1

Base — sugar — PO — sugar — Base

Base— sugai — PO (OH) 2
la

No direct chemical evidence has been adduced in support of this type of
branched structure; it would be expected to be relatively unstable both to
acid and alkali and might, therefore, have been missed in chemical degrada-
tions. More recent evidence for this type of linkage derived from enzyme
experiments will be considered later in connection with the general problem
of chain-branching in ril)onucleic acids.

Electrometric titration of deoxyribonucleic acids of high molecular weight
shows only small amounts of secondary phosphoryl dissociation. -° These
acids are thus generally considered to be straight-chain polydiesters of
type I, a structure which also accords with their other physical properties.^'
[Cf. Jordan, Chapter 13.] Nevertheless, Lee and Peacocke-^ have inter-
preted their titration data on the basis of a branched structure including
phosphotriester linkages, a conclusion also reached from dye adsorption
studies.'^

Euler and Fono'^ have observed the liberation of base-binding groups
when deoxyribonucleic acid is treated with alkaH (pH 11.5). They suggest
that this represents the fission of linkages between phosphate and the
enolic hydroxyl groups of purine or pyrimidine residues. A similar explana-
tion has been suggested by Little and Butler-^ to account for their observa-
tion that during the action of deoxyribonuclease on deoxyribonucleic acids
groups of pK 9-10 are liberated, in addition to secondary phosphoryl
groups. It seems more likely that these represent enolic hydroxyl groups
on the purine and pyrimidine residues, which are masked by hydrogen-
bonding and are set free during degradation of the molecule, than that they
originate in covalent internucleotidic linkages. Since the specificity and
mode of action of deoxyribonuclease has not been clearly defined, it is pos-
sible, as Zamenhof and Chargaff-^ point out, that two effects may be super-

" J. M. Gulland, D. O. Jordan, and H. F. W. Taylor, J. Chem. Soc. 1947, 1131.
^' D. O. Jordan, Progr. Biophys. and Biophys. Chem. 2, 51 (1951).
" W. A. Lee and A. R. Peacocke, J. Chem. Soc. 1951, 3361.

23 L. F. Cavalieri and A. Angelos, J. Am. Chem. Soc. 72, 4686 (1950).

24 H. von Euler and A. Fono, Arkiv. Kemi Mineral. Geol. 25A, No. 3 (1947).
"J. A. Little and G. C. Butler, J. Biol. Chem. 188, 695 (1951).

2« S. Zamenhof and E. Chargaff, J. Biol. Chem. 187, 1 (1950).

412 D. M. BROWN AND A. R. TODD

imposed — ^an enzymic cleavage of covalent links followed by a spontaneous
rupture of secondary valence bonds. Similar observations have been made
by Vandendriessche^^ in a study of the degradation of yeast ribonucleic acid
by ribonuclease. He, too, points out that the liberation of such titratable
groups does not necessarily indicate that they were originally present in
covalent linkages. Cavalieri, Kerr, and Angelos'^ report discrepancies be-
tween base content and titratable groups in the so-called "core" of ribo-
nucleic acid, and suggest that in this material (see later), which constituted
some 3-10% of the original nucleic acid, the enolic hydroxyl of uracil and
guanine may be involved in internucleotidic linkages. Corresponding
changes in ultraviolet absorption when this material'^ '^^ and intact ribo-
nucleic acid^^ are hydrolyzed by alkali, or when deoxyribonucleic acids are
degraded by deoxyribonuclease^^ or snake venom,'" suggest, however, that
the changes are more likely to originate in cleavage of secondary valence
bonds, which interfere with the resonating system of purine and pyrimi-
dine rings,^^''^ than in rupture of covalent linkages. [Cf. Beaven, Holiday,
and Johnson, Chapter 14.]

Based on the reaction of deoxyribonucleic acids with Feulgen's reagent
after very mild treatment with acid and alkali, or even simple dialysis,
Stacey et al.^^'^^ have proposed that a small number of labile phosphoryl
linkages attached to Ci of deoxyribose residues are present. This type of
structure would, of course, be at variance with the generally accepted poly-
nucleotide structure of nucleic acids, although it could be argued that a
very small proportion of such links in a nucleic acid of high molecular weight
might make little difference to the nitrogen-phosphorus ratio as determined
by analysis.

Consideration of the various types of linkage which have been proposed
in recent years leads to the conclusion that the phosphodiester linkage
between the sugar residues of the individual nucleosides remains at least
the major one. The others, apart possibly from the phosphotriester link,
may have been advanced on reasonable grounds, but the absence of cor-
roborative chemical evidence and, in general, their apparently very small
occurrence in proportion to the others, suggests that they are of minor

" B. Magasanik and E. Chargaff, Biochim. et Biophrjs. Acta 7, 396 (1951).

" M. Kunitz, /. Biol. Chem. 164, 563 (1946).

" M. Kunitz, J. Gen. Physiol. 33, 349, 363 (1950).

3" R. L. Sinsheimer and J. F. Koerner, J. Biol. Chem. 198, 293 (1952).

" B. Commoner, Science 110, 31 (1949).

« C. F. Li, W. G. Overend, and M. Stacey, Nature 163, 538 (1949).

" W. G. Overend, M. Stacey, and M. Webb, J. Chem. Soc. 1951, 2450.

»* W. G. Overend, A. R. Peacocke, and M. Stacey, J. Sci. Food Agr. 3, 105 (1952).

« W. A. Lee and A. R. Peacocke, J. Chem. Soc. 1952, 130.

CHEMICAL BONDS IN NUCLEIC ACIDS 413

significance, and may even represent false interpretations of experimental
findings.

II. Chemistry of the Ribonucleic Acids

1. Mononucleotides

Until 1949 only four mononucleotides had been isolated from hydroly-
sates of ribonucleic acids. Levene and his co-workers^®" concluded from
degradative studies and arguments based on analogy- that these were the
3 '-phosphates of the four nucleosides adenosine, guanosine, cytidine, and
uridine, i.e., that they could be represented by structure lib in which R rep-
resents a purine (adenine or guanine) or pyrimidine (cytosine or uracil)
residue. The fact that only these four nucleotides were obtained made it
very difficult to define the position of the internucleotidic linkage. Clearly
it could not be 3', 3'-, but if it were not then it seemed necessary to assume
that a phosphoryl linkage at a position other than 3'- would be preferen-
tially hydrolyzed. The 5 '-position seemed unlikely on this basis since
adenosine-5 '-phosphate (muscle adenylic acid) had a stability towards hy-
drolysis comparable with that of yeast adenylic acid and Levene and Tip-
son^* therefore formulated the ribonucleic acids as 2 ',3 '-linked polynucleo-
tides, assuming, in the absence of any experimental evidence, that the
Cz' — O^ — P linkage would be less stable than the C3- — O- — P linkage, and
would therefore always be ruptured on hydrolysis, yielding only the nucleo-
side-3 '-phosphates. Gulland,^'^* on the basis of enzymic studies, did at one
stage propose a 3 ',5 '-linked structure for ribonucleic acid, but as he found
it necessary to invoke a rather improbable phosphoryl migration from C^-
to C3- to account for the products of alkaline hydrolysis, he later retracted
this proposal.^"

-O-

I OPO3H2 OH
R-CH \ —

OH OPO3H2
■CH^-OH R-CH-

-CHa-OH

H H H H H

Ila lib

The isolation by Carter and Cohn^^ in 1949 of two isomeric adenylic
acids, termed a and b, from alkaline hydrolysates of yeast ribonucleic acid,
and the subsequent demonstration that similar pairs of isomeric nucleotides

" P. A. Levene and S. A. Harris, J. Biol. Chem. 98, 9 (1932).
" P. A. Levene and S. A. Harris, J. Biol. Chem. 101, 419 (1933).
38 P. A. Levene and R. S. Tipson, J. Biol. Chem. 109, 623 (1935).
" J. M. Gulland and E. M. Jackson, J. Chem. Soc. 1938, 1492.
^» J. M. Gulland and E. O. Walsh, J. Chem. Soc. 1945, 172.

*^C. E. Carter and W. E. Cohn, Federation Proc. 8, 190 (1949), and subsequent
papers.

414 D. M. BROWN AND A. R. TODD

derived from each of the other three nucleosides were also produced^^-'is
put a new complexion on the polynucletide problem. The a and h nucleotides
were shown^® to be the 2'- and 3 '-phosphates of the corresponding nucleo-
sides (i.e., Ila and b), although not necessarily respectively, and much sub-
sequent work has confirmed this conclusion. The acid-catalyzed intercon-
version of the isomeric a and b nucleotides,^^'^^ which will be discussed later,
rendered invalid the conclusion of Levene and Harris that their isolated
nucleotides were 3 '-phosphates, since their degradative method could not
really distinguish between the 2'- and 3 '-isomers. Very recent studies*^"^"
have provided strong evidence for the view that the a nucleotides are the
2 '-phosphates (Ila), and the h nucleotides the 3 '-phosphates (lib), of the
respective nucleosides. While additional confirmation"^ of these findings
would no doubt be desirable, it is reasonable to accept them for the pur-
poses of our discussion. On their validity rests any final conclusion regard-
ing the absolute orientation of the internucleotidic linkages in intact ribo-
nucleic acids.

The salient points in the chemistry of the nucleoside-2'- and -3 '-phos-
phates [cf. Baddiley, Chapter 4] may be briefly restated. Their intercon-
version in acid solution to an equilibrium mixture^^''^^ has been shown*«
by analogy with the glycerol-a- and -|3-phosphates^i-^^ to depend on migra-
tion of the phosphoryl group via an intermediate cyclic 2 ',3 '-phosphate
(III). The intermediate cychc 2 ',3 '-phosphates have been synthesized ^^
and shown to possess the predicted properties, i.e., they are unstable and

« W. E. Cohn, J. Am. Chem. Soc. 72, 1471 (1950).

« W. E. Cohn, J. Am. Chem. Soc. 72, 2811 (1950).

** H. S. Loring, N. G. Luthy, H. W. Bortner, and L. W. Levy, J. Am. Chem. Soc. 72,

2811 (1950).
« W. E. Cohn, J. Cellular Comp. Physiol, 38, Suppl. 1, 21 (1951).
^6 D. M. Brown and A. R. Todd, J. Chem. Soc. 1952, 44.
« H. S. Loring, M. L. Hammell, L. W. Levy, and H. W. Bortner, J. Biol. Chem. 196,

821 (1952).
« L. F. Cavalieri, J. Am. Chem. Soc. 74, 5804 (1952); 75, 5268 (1953).
" J. J. Fox, L. F. Cavalieri, and N. Chang, J. Am. Chem. Soc. 75, 4315 (1953).
''0 J. X. Khym, D. G. Doherty, E. Volkin, and W. E. Cohn, J. Am. Chem. Soc. 75,

1262 (1953).
''»» This has recently been given in the case of the isomeric adenylic acids by X-ray

analysis (D. M. Brown, G. D. Fasman, D.I. Magrath, A. R. Todd, W. Cochran,

and M. M. Woolfson, Nature 172, 1184, 1953), by chemical synthesis {loc. cit.; D.

M. Brown, G. D. Fasman, D. I. Magrath, and A. R. Todd, J. Chem. Soc. 1954,

1448), and by degradation (J. X. Khym and W. E. Cohn, J. Am. Chem. Soc. 76,

1818, 1954).
" M. C. Bailly, Compt. rend. 206, 1902 (1938); 208, 443 (1939).
62 P. E. Verkade, J. C. Stoppelenburg, and W. D. Cohen, Rec. trav. chim. 59, 886

(1940).
" E. Chargaff, J. Biol. Chem. 145, 455 (1942).
" D. M. Brown, D. I. Magrath, and A. R. Todd, J. Chem. Soc. 1952, 2708.

CHEMICAL BONDS IN NUCLEIC ACIDS

415

are readily hydrolyzed to a mixture of the nucleoside-2'- and nucleoside-3'-
phosphates. The free isomeric mononucleotides are, like the glycerol mono-
phosphates, stable without interconversion in alkaline solution.

Ila^

R-CH

CH,-OH ^

lib

2. Esters of the Mononucleotides

Bailly and Gaume^^ showed that whereas glycerol-a-phosphate is stable
to alkali, glycerol-a methyl hydrogen phosphate (IV; R = Me) is readily
hydrolyzed under alkaline conditions to methanol and a mixture of glycerol-
a- and -iS-phosphate ; no methyl phosphate is produced. Similar observations
by Baer and Kates^^ show that the choline phosphate (IV; R = choline
residue) is hydrolyzed to choline and glycerol-a- and -/3-phosphate, while
alkaline degradation of the lecithins also results in phosphoryl migration."
2-Hydroxyethyl dimethyl phosphate (V) under the action of dilute alkali
yields 2-hydroxyethyl phosphate (VI) with loss of methanol. ^^ Mild acid
treatment also effects these degradations. It is noteworthy, however, that
2-methoxyethyl dimethyl phosphate (VII) forms 2-methoxy ethyl methyl
hydrogen phosphate (VIII) with alkali, and this is now stable to further
hydrolysis, a result which is in accord with the generally recognized stability
of dialkyl esters of phosphoric acid towards alkaline reagents.^*"®'

OH

I
H0-CH2-CH-CH,-0P0(0H)-OR

IV

HO-CH2-CH2-OPO(OMe)2
V

0 OR

W/
P

I I

HO-CH^-CH-CHz
IX

HO-CHj-CH^-OPOaH^
VI

0 OH

W/

p

-*H0-CH,-CH-CH,
X

Me0-CH2-CH.2-0P0(0Me)
VII

Me0-CH2-CH2-0P0(0H)-0Me
VIII

" O. Bailly and J. Gaurn^, Bull. soc. chim. 2, 354 (1935).
^6 E. Baer and M. Kates, J. Biol. Chem. 175, 79 (1948).
" E. Baer and M. Kates, J. Biol. Chem. 185, 615 (1950).
^8 O. Bailly and J. Gaum^, Bull. soc. chim. 3, 1396 (1936).
"J. Cavalier, Compt. rend. 127, 114 (1898).

416 D. M. BROWN AND A. R. TODD

The prerequisite for alkali-lability in dialkyl phosphates is thus the pres-
ence of hydroxyl function in proximity to the phosphoryl group.

It is well known that triesters of phosphoric acid are sensitive both to
alkali and to acid.*'^®^ Brown and Todd®^ envisaged a mechanism for the
hydrolysis of the hydroxylated dialkyl phosphates in which a cyclic inter-
mediate is involved. Thus, in the alkaline hydrolysis of glycerol-a methyl
phosphate (IV; R = Me) an intermediate e.g. IX (R = Me) is produced
and immediately cleaved to methanol and the cyclic phosphate X which
then gives rise to glycerol-a- and -/3-phosphate. No other major products
are to be expected, since it is evident that although the three ester linkages
are of comparable reactivity, yet only by fission of the bond retaining
the singly-linked substituent can degradation of the molecule occur. Other
structures for the intermediate IX have been proposed.^® ^^ Brown and
Todd®* point out that the classical neutral triester intermediate of type IX
was originally advanced to simplify discussion, and suggest that the true
mechanism of processes of type IV —^ X depends probably on an acid- or
base-catalyzed attack by the vicinal hydroxyl group on the — P=0
bond with simultaneous elimination of the R group as an alkoxy anion,
possibly with the intervention of a pentacovalent transition complex.***

Fon6,®^ as early as 1947, noted the importance of the neighboring hy-
droxyl groups in causing lability of diesters of phosphoric acid containing a
glycerol or ethylene glycol residue, and postulated the existence of a cyclic
triester in their hydrolysis. He put forward the view that the alkali-labihty
of ribonucleic acid, in contrast to the alkali -stability of deoxyribonucleic
acid, may depend on the extra hydroxyl group present at C 2 in the sugar
residues of the former, i.e., that ribonucleic acids are analogous to glycerol
alkyl phosphates in their hydrolytic behavior. Although evidence was not
available at the time to permit detailed conclusions to be drawn, subsequent
work has clearly established the validity of Fono's basic idea.

Brown and Todd,^*'®^ during their synthesis of adenylic acids a and 6,
isolated two substances shown to be adenosine-2' benzyl phosphate and
adenosine-3' benzyl phosphate (XI and XII; R' = CH2C6H6 ; R = adenine
residue). These substances were found to be readily hydrolyzed in both

•0 R. H. A. Plimmer and W. J. N. Burch, J. Chem. Soc. 1929, 279.

*i G. M. Kosolapoff, "Organophosphorus Compounds," p. 232. John Wiley & Sons,
New York, 1950.

^•^ D. M. Brown and A. R. Todd, J. Chem. Soc. 1952, 52.

" D. M. Brown and A. R. Todd, J. Chem. Soc. 1953, 2040.

**" This view of the mechanism finds confirmation in the experiments of D. Lipkin,
P. T. Talbert and M. Cohn (/. Am. Chem. Soc. 76, 2871, 1954) who showed that
alkaline hydrolysis of yeast ribonucleic acid in H2O'* yields mononucleotides con-
taining only one atom of O'* per atom of phosphorus

" A. Fono, Arkiv. Kemi Mineral. Geol. 24A, No. 33, 14, 15 (1947).

CHEMICAL BONDS IN NUCLEIC ACIDS

417

acidic and alkaline media with simultaneous phosphoryl migration yielding
benzyl alcohol and a mixture of adenosine-2'- and -3 '-phosphate; no other
products were observed. Clearly these degradations are entirely analogous
to those recorded in the glycerol phosphate series, and proceed by way of a
cyclic intermediate (e.g., XIII); this then yields the cyclic phosphate XIV
(R = adenine residue), which subsequently hydrolyzes to adenosine-2 '-

■0

R-CH

OPO(OH)-OR' OH

CH,-OH

H

H H

■CH,-OH

XI

-0-

R-CH

OH OPO(OH)-OR'

H H

H

XII

•CH,-OH

OH

HO-CH

and -3 '-phosphate. The properties of the synthetic nucleoside-2', 3 '-phos-
phates^"* (XIV) are in every way consistent with the hypothesis that they
participate in the hydrolytic breakdown of nucleotide esters. Comparable
observations have since been made on benzyl, methyl, and ethyl esters of
the cy tidy lie and uridyUc acids. "■''^

It was also noted that adenosine-5' benzyl phosphate*^ was stable under
conditions which led to the hydrolysis of the 2'- and 3 '-nucleotide esters.
Nor was adenosine-o '-phosphate converted under acidic conditions into the
2'- or 3'-nucleotide.

Consideration of the stereochemistry of the natural ribonucleosides (XV ;
R = purine or pyrimidine residue) shows that the hydroxyl groups at Co
and Cs of the sugar residue bear a fi'.s'-relationship to each other, thus per-

" J. Baddiley and A. R. Todd, J. Chern. Soc. 1947, 648.

418 D. M. BROWN AND A. R. TODD

mitting ready formation of cyclic 2',3'-phosphoryl intermediates in the
intercon versions of the 2'- and 3 '-nucleotides, whereas the ^rans-relation-
ship between the hydroxy] groups at Cs and Cb prevents the formation of
cyclic phosphoryl and other^^ derivatives at these positions. As a corollary,
it is evident from the above that the course of hydrolysis of analogous de-
rivatives would be expected to be different if sugar residues (e.g., xylofura-
nose), in which the stereochemical relationships of the hydroxyl functions
were altered, were substituted for ribofuranose in the nucleosides. Ribose
and deoxyribose in their furanose forms are uniquely suited for incorpora-
tion in the nucleic acids.

The above observations on the reactions of simple esters of the mono-
nucleotides can be applied directly to the elucidation of ribonucleic acid
structure.

3. General Structure of Polyribonucleotides Based on Chemical

Degradation

It was early shown that when ribonucleic acids are treated with mild
alkaline reagents under a variety of conditions they are rapidly converted
to a mixture of their component mononucleotides. "''^^ [Cf. Chapters 5 and
11.] Claims to the isolation of larger f ragments^^ '^^ have generally been
relinquished or refuted by other workers on the general grounds that the
products described were separable mixtures of mononucleotides^^''''® (com-
pare, however, Smith and Allen"). In the same way, mild acid hydrolysis
also yields mononucleotides, although further degradation of the purine
nucleotides complicates the picture. The early observation that the final
products of alkaline hydrolysis are mononucleotides has been confirmed by
recent studies using chromatographic*^'^*'*" and electrophoretic*^ separation,

66 D. M. Brown, L. J. Haynes, and A. R. Todd, J. Chem. Soc. 1950, 3299.
" H. Steudel and E. Peiser, Z. phijsiol. Chem. 120, 292 (1922).

68 P. A. Levene, /. Biol. Chem. 40, 415 (1919); 55, 9 (1923).

69 W. Jones and M. E. Perkins, J. Biol. Chem. 62, 557 (1925).
'6 H. 0. Calvery, J. Biol. Chem. 72, 27 (1927).

" G. Schmidt and S. J. Thannhauser, J. Biol. Chem. 161, 83 (1945).

" H. S. Loring, P. M. Roll, and J. G. Pierce, /. Biol. Chem. 174, 729 (1948).

" W. Jones and H. C. Germann, J. Biol. Chem. 25, 93 (1916); W. Jones and B. E.

Read, ihid. 29, Ul; 31, 39 (1917).
7*S. J. Thannhauser and G. Dorfmuller, Z. physiol. Chem. 95, 259 (1915).
75 P. A. Levene, J. Biol. Chem. 33, 229 (1919).

'6 W. Jones, "Nucleic Acids," p. 36. Longmans, Green and Co., London, 1920.
" K. C. Smith and F. W. Allen, J. Am. Chem. Soc. 75, 2131 (1953).
'8E. Chargaff, B. Magasanik, E. Vischer, C. Green, R. Doniger, and D. F. Elson,

J. Biol. Chem. 186,51 (1950).
'9 R. Markham and J. D. Smith, Biochem. J. 49, 401 (1951).
8» J. Montreuil and P. Boulanger, Bull. soc. chim. biol. 33, 784, 791 (1951).
81 J. N. Davidson and R. M. S. Smellie, Biochem. J. 52, 594 (1952).

CHEMICAL BONDS IN NUCLEIC ACIDS 419

and physical methods for identifying the products. Recovery of the mono-
nucleotides as mixtures of the 2'- and 3 '-isomers (generally in the ratio
40:60) from alkaline hydrolysates of ribonucleic acids, in amounts ap-
proaching the theoretical value, have been reported. ^^■^-

Brown and Todd^- have discussed these observations on the basis of
their findings with the esters of the mononucleotides (see above), and have
developed from them general structures for the ribonucleic acids. If the
linear polynucleotide sequence XVI is considered, in which the expression
C2' — Cs'^ — C5- is used as an abbreviation for individual nucleoside residues,
alkaline degradation will proceed through a cyclic intermediate (formulated
for simplicity as XVII) yielding cyclic nucleoside-2', 3 '-phosphates by
exclusive fission of the Cs- — 0 — P bond, in strict conformity with the be-
havior of the fundamentally analogous simple nucleotide esters. The cyclic
phosphates then yield, by further hydrolysis, the mixture of nucleoside-2 '-
phosphates and nucleoside-3 '-phosphates which is normally isolated. This
postulate that cyclic nucleoside-2 ',3 '-phosphates should be produced dur-
ing alkaline hydrolysis of ribonucleic acids has since been substantiated by
their demonstration^-^ in barium carbonate (ca. pH 9 at 100°) and dilute
ammonia hydrolysates. Kinetic studies, too, give results consistent with
the proposed hydrolytic mechanism.*^ The earlier failure to recognize cyclic
phosphates among hydrolytic products was doubtless due to their labile
character.

C2.-OH C2.-OH C2.-OH C2.-OH

I _ /^v' _ -/^^

C5. c.

0 — C5. 0-^ 0—

■c, o- ■<

XVI

0 c,-o^^^o

r°\,y

0 c.-o-^Xq 0,-0-^ X,

XVII

It should be noted that no such mechanism of hydrolysis can occur in
the case of deoxyribonucleic acids where the absence of a hydroxyl at C2'
in the deoxyribofuranose residues prevents the essential cyclization which

" E. Volkin and C. E. Carter, J. Am. Chem. Soc. 73, 1516 (1951).

83 R. Markham and J. D. Smith, Biochem. J. 52, 552 (1952).

8^ J. E, Bacher and W. Kauzmann, J. Am. Chem. Soc. 74, 3779 (1952).

420

D. M. BROWN AND A. R. TODD

is a prerequisite for alkali-lability. It is for this reason that they are not
degraded to small molecules by alkali,*^ '^^ and their stability in this respect
is in accord with the generally accepted resistance of dialkyl phosphates to
alkaline hydrolysis. ^^•*'- Acid hydrolysis of deoxyribonucleic acids clearly
depends on other factors and will be discussed later.

Arguments based on complete hydrolysis to mononucleotides do not, of
themselves, permit rigid conclusions to be drawn regarding the precise
position of the internucleotidic linkages in ribonucleic acids, but they cer-
tainly limit the number of possible structures. In addition to structure
XVI, in which the linkage is shoASTi joining C3- in one nucleoside residue
to C5' in the next, a comparable structure in which C2' is linked to Cs- can
be drawn, which would be indistinguishable from XVI by alkaline hydroly-
sis. Such structures as XVIII, which involve C^r — Cs- linkages can, how-
ever, be dismissed, since they should yield dinucleotides stable to further
hydrolysis. This conclusion that Cb- — C5. linkages are incompatible with
hydrolysis to mononucleotides is implicit in the stability of adenosine-5'
benzyl phosphate towards alkali and acid, and is confirmed by the stability
of synthetic dinucleoside-5', 5 '-phosphates. Diuridine-5', 5 '-diphosphate*^
(originally synthesized by Gulland and Smith, and believed erroneously to
be the 2 ',2 '-isomer), and adenosine-5' uridine-5' phosphate^^ (XIX) are
both stable to alkali under conditions which bring about complete hydroly-
sis of ribonucleic acid to mononucleotides; this stability is due to the ab-
sence of the vicinal hydroxyl group necessary for cyclization and hence
easy fission of the phosphoryl group in these substances.

-c.

C

Co.

Cs- P Co-

C3-P-

C,.-P — c,

C,,-P-C,.
XVIIl

NH,

0-

i OH OH
CH I I
I H H H

N'

O

II
■CH.-O-P-O-CH-r
I
OH

XIX

— 0 —
OH OH

■CH

H H H

0=f

N

OH

" J. M. Gulland and H. Smith, J. Chem. Soc. 1948, 1532.
8« D. T. Elmore and A. R. Todd, J. Chem. Soc. 1952, 3681.

CHEMICAL BONDS IN NUCLEIC ACIDS 421

c.-P-

-c,.

c.-

-p-

'?■

Cx

C3.-

-p— C3.

C3-P-

C5.

c.

XX

Ca-

C3.

4

■-I:

p>.

4?-

Cfi.

Cs-

XXI

c,-

Two other structural types (XX and XXI, or structures containing the
features of both) would also yield mononucleotides by the hydrolytic mech-
anism under discussion. In each case, however, hydrolysis would necessarily
have to proceed stepwise by removal of mononucleotide units from that
end of the polynucleotide chain bearing a free hydroxyl group at C2' or Cs- .
Structures of type XVI, embodying a 3 ',5 '-linkage, could in theory be
degraded by simultaneous attack at m.any points in the chain. Relevant
chemical evidence, although scanty, is in favor of simultaneous rather than
stepwise attack.*^ (Compare, however, Magasanik and Chargaff.^') Mer-
rifield and WooUey*^ showed that controlled acid hydrolysis of yeast ribo-
nucleic acid yielded a variety of small oligonucleotides, some of which
they separated by ion-exchange chromatography and were able to charac-
terize. Production of molecules of this type shows that structural features
in type XVI, in which 3',5'(or 2',5')-liiikages are present, must exist in
ribonucleic acids as distinct from such structures as XX and XXI, where
Cs' is not involved in the internucleotidic linkage and where hydrolysis,
which requires a free hydroxyl at C2' or Cs- , can only occur at a terminal
linkage; in other words C5. must be involved as one of the internucleotidic
linkage points. The isolation of nucleoside-5 '-phosphates after enzymic
hydrolysis of ribonucleic acids to be discussed later also indicates that Cs-
is one of the major linkage points.

The above considerations led Brown and Todd^^ to postulate the general
structure XVI, involving a recurring 3',5'-phosphodiester linkage for the
polynucleotide sequence in ribonucleic acids, as the one which accounts
most satisfactorily for the chemical evidence. They pointed out, however,
that a 2 ',5 '-linkage would be equally admissible on the known facts of
hydrolysis. Preference for the 3 ',5 '-structure was expressed mainly on the
ground of analogy with deoxyribonucleic acids where a 2',5'-linkage is
structurally impossible; justification for this preference has since come from

" R. B. Merrifield and D. W. Woolley, J. Biol. Chem. 197, 521 (1952).

422 D. M. BROWN AND A. R. TODD

further studies using enzymes. Evidence from chemical hydrolysis to mono-
nucleotides does not permit any answer to the question whether ribonucleic
acids are straight-chain or branched-chain polynucleotides. It does, how-
ever, clearly define the types of branching which can be considered. Given
that the polynucleotide chain in each branch follows the standard 3',5'-
linkage pattern, then two types of branching can be envisaged: (a) branch-
ing on phosphorus, i.e., through alkali-labile phosphotriester groupings,
and (b) branching at C2' in one of the nucleoside residues in the main chain,
attachment being through the usual phosphodiester linkage to C3. (or C2')
in the first nucleoside residue of the branch; linkage to Cs- would give an
alkali-stable, and therefore inadmissible, structure. The question of chain-
branching in ribonucleic acids will be discussed later; it is mentioned at
this point because the evidence of chemical hydrolysis clearly defines the
only types of branching which need be considered.

4. Evidence for Nucleoside-5 '-Linkages

Although evidence has gradually accumulated, mainly as a result of en-
zymic studies, that 5'-phosphoester linkages occur in nucleic acids, the
complete failure to isolate nucleoside-5 '-phosphates from chemical hydroly-
sates for long prevented the general acceptance of their presence. Follow-
ing early studies by Takahashi** and Klein and Rossi,^^ Gulland and
Jackson^^ examined the action of Russell's viper venom. This venom had,
in addition to 5 '-nucleotidase and phosphodiesterase, a weak nonspecific
monoesterase activity; however, they showed that when it acted on yeast
ribonucleic acid it liberated 25% of the bound phosphorus as inorganic
phosphate, and this was increased to 75% by the addition of bone phos-
phomonoesterase. These early indications of 5'-phosphoester linkages have
been confirmed in recent years by the work of Cohn and his co-workers.
Thus Cohn and Volkin^" treated ribonucleic acid with ribonuclease, followed
by intestinal phosphatase in the presence of arsenate, to inhibit phospho-
monoesterase.^^ Analysis of the hydrolysate by ion-exchange chromato-
graphy yielded, in addition to unidentified products, the 5 '-phosphates of
adenosine, guanosine, uridine, and cytidine. These were characterized by
comparison with the synthetic nucleoside-5 '-phosphates prepared by
Michelson and Todd.^'^ Using venom diesterase virtually free from mono-
esterase (Hurst, Little, and Butler^^), Cohn andVolkin^* obtained a yield of

88 H. Takahashi, /. Biochem. Japan 16, 463 (1932).

89 W. Klein and A. Rossi, Z. physiol. Chem. 231, 104 (1935).

90 W. E. Cohn and E. Volkin, Nature 167, 483 (1951).

91 cf. W. Klein, Z. physiol. Chem. 218, 164 (1933).

92 A. M. Michelson and A. R. Todd, J. Chem. Soc. 1949, 2476.

93 R. O. Hurst, J. A. Little, and G. C. Butler, J. Biol. Chem. 188, 705 (1951).

94 W. E. Cohn and E. Volkin, Arch. Biochem. and Biophys. 35, 465 (1952).

CHEMICAL BONDS IN NUCLEIC ACIDS 423

more than 60% of niicleoside-5 '-phosphates from the ribonucleic acids of
yeast and calf liver, in addition to some free nucleosides and nucleoside
diphosphates. These latter substances, which are discussed later in connec-
tion with chain-branching, have one of their phosphate groups at the 5'-
position of the nucleoside residue. Thus there is good evidence that the
C 5 '-position is of major importance as an internucleotidic linkage point.
Further strong confirmatory evidence is to be found in the chemistry of
the oligonucleotides found in ribonuclease digests of ribonucleic acids. The
presence of 5'-phosphoester linkages in ribonucleic acids and the absence
of nucleoside-5 '-phosphates in chemical hydrolysates find an adequate
explanation in the theory of hydrolysis discussed above. ^^

5. Chemistry of Ribonuclease Action

As clearly indicated, alkaline hydrolysis degrades ribonucleic acid to
mononucleotides without yielding larger fragments (oligonucleotides) con-
taining more than one nucleotide unit. Such larger fragments are of impor-
tance in the study both of nucleotide sequence and of the detail of the inter-
nucleotidic linkage. They can be obtained by the action of certain enzymes,
among which the most widely used is pancreatic ribonuclease. This enzyme
was obtained in crystalline form by Kunitz^* in 1940, and its use in recent
years has shed light on a number of important details of ribonucleic acid
structure.

Allen and Eiler^ found that 0.25 equivalents of secondary phosphoryl
groups per atom of phosphorus were liberated during ribonuclease action,
and higher values (0.4-0.5) have since been reported ;*2'^^ no inorganic
phosphate was liberated. The enzyme, indeed, seemed to be a specific
type of diesterase and as such it gave both dialyzable and nondialyzable
fission products when allowed to act on ribonucleic acids. Weiner, Duggan,
and AUen^* found by titration that the ratios of monoesterified to diesterified
phosphate in the intact nucleic acid, and in the dialyzable and the non-
dialyzable fractions of the digests, were 1:10, 1:2, and 1:3, indicating that
considerable breakdown had occurred even in the nondialyzable fragments.
Analysis of the fractions obtained" showed that the smaller (dialyzable)
fragments were rich in pyrimidine nucleotide derivatives, while the larger
(nondialyzable) showed a high purine-pyrimidine ratio. These observa-
tions led to the suggestion"-^* that the enzyme acted at pyrimidine nucleo-
tide sites in the molecule. Evidence pointing in the same direction came

S6M. Kunitz, J. Gen. Physiol. 24, 15 (1940).

" E. Volkin and W. E. Cohn, J. Biol. Chem. 205, 767 (1953).

" J. E. Bacher and F. W. Allen, J. Biol. Chem. 183, 633 (1950).

98 See also G. Schmidt, R. Cubiles, B. H. Swartz, and S. J. Thannhauser, J. Biol.

Chem. 170, 759 (1947) ; G. Schmidt, R. Cubiles, and S. J. Thannhauser, Cold Spring

Harbor Symposia Quant. Biol. 12, 161 (1947).

424 D. M. BROWN AND A. R. TODD

from the demonstration" •^^•^'"' that in addition to larger fragments, sub-
substantial amounts of pyrimidine mononucleotides were formed by ribo-
nuclease digestion without the concomitant production of purine mono-
nucleotides. A single claim to have isolated purine mononucleotides from
ribonuclease digests^"^ has not been confirmed by later work. Further evi-
dence for a specificity of ribonuclease towards pyrimidine nucleotide resi-
dues came from the work of Schmidt, Thannhauser, and their co-workers.
It was found^'''''^*'^ that after exhaustive digestion of ribonucleic acid with
ribonuclease and subsequent treatment with prostatic phosphomonoester-
ase, inorganic phosphate was produced in an amount corresponding to at
least 93 % of the pyrimidine nucleotide phosphorus content of the original
nucleic acid. The remaining organically bound phosphorus was present
entirely, or almost entirely, as purine nucleotide phosphorus.'*'^ The con-
clusion was drawn that in ribonuclease digests the pyrimidine nucleotide
residues were present either as mononucleotides or as terminal residues
carrying a monoesterified phosphoryl group in oligonucleotides consisting
otherwise solely of purine nucleotide residues. The same workers added
materially to this conclusion by periodate oxidation studies. Ribonucleic
acid and the products of its digestion with ribonuclease were stable to
periodic acid, but after treating the digest with phosphomonesterase, a
large periodate upta.ke -was observed which was equivalent to the amount
of inorganic phosphate liberated, and hence corresponded approximately
to the pyrimidine nucleotide content of the original nucleic acid. They also
showed that the larger, phosphorus-containing, fragments in the phos-
phomonoesterase-treated digests, i.e., oligonucleotides containing a termi-
nal pyrimidine residue, were also susceptible to periodate oxidation. Since
oxidation of nucleotides by periodate depends on the presence of unsubsti-
tuted hydroxyl groups at both C2' and C3. ,^''* they concluded that the
terminal (pyrimidine) residue in these oligonucleotides was linked to the
rest of the molecule through a position other than C2' or C3' . Claims that
ribonucleic acids themselves, and ribonuclease digests before phosphomono-
esterase treatment, reduce periodate'^- '"^ and lead tetraacetate'^^ have been
denied in more recent publications.^®'"- Periodate uptake after phospho-
rs C. E. Carter and W. E. Cohn, /. Am. Chern. Soc. 72, 2604 (1950).
lo" G. Schmidt, R. Cubiles, and S. J. Thannhauser, J. Cellular Comp. Physiol. 38,

Suppl. 1,61 (1951).
i"' H. S. Loring and F. H. Carpenter, J. Biol. Chem. 150, 381 (1943).
i"" G. Schmidt, R. Cubiles, N. Zollner, L. Hecht, N. Strickler, K. Seraidarian, M.

Seraidarian, and S. J. Thannhauser, J . Biol. Chem. 192, 715 (1951).
'" cf. W. Jones, J. Biol. Chem. 24, iii (1916).
i»* B. Lythgoe and A. R. Todd, J. Chem. Soc. 1944, 592.
los F. W. Allen, Federation Proc. 10, 155 (1951).
i"* R. A. Becher and F. W. Allen, J. Biol. Chem. 195, 429 (1952).

CHEMICAL BONDS IN NUCLEIC ACIDS 425

monoesterase treatment has, on the other hand, been confirmed in several
laboratories."'^^ Brown and Todd®- pointed out that an uptake of periodate
after ribonuclease action but without phosphomonoesterase treatment
would only be possible if Cz- — 0 — P linkages were split. The picture of
ribonuclease action which emerges from all these observations is that the
enzyme cleaves Cs- — 0 — P linkages in the postulated structure XVI giving,
in addition to pyrimidine-2'- or -3 '-mononucleotides, fragments containing
terminal pyrimidine nucleotide residues, bearing monoesterified phosphoryl
groups at C2' or C3.; enzymic removal of these phosphomonoester groups
then yields periodate-oxidizable pyrimidine nucleoside residues linked
through Cs' to the rest of the (purine) polynucleotide chain.

When ribonucleic acid is treated for short periods with ribonuclease,
two substances are produced^"^ which are converted, respectively, to uri-
dylic and cy tidy lie acids by further action of the same enzyme. These two
substances have been shown to be identical®^ '^"^ with the cyclic phosphates,
uridine-2', 3 '-phosphate and cytidine-2', 3 '-phosphate, synthesized by
Brown, Magrath, and Todd.^^ It is clear that these cyclic phosphates are
the precursors of the pyrimidine mononucleotides found in ribonuclease
digests and that they must have arisen by a mechanism akin to that postu-
lated for the chemical hydrolysis of the nucleic acids.^'- This was confirmed
by Brown, Dekker, and Todd^^* in a further investigation of the action
of ribonuclease on the cyclic phosphates. Volkin and Cohn"° had shown
that the pyrimidine nucleotide fraction in ribonuclease digests consists
solely of uridylic acid b and cytidylic acid b (i.e., the 3 '-phosphates), and
Brown, Dekker, and Todd provided an adequate explanation of this fact
by showing that ribonuclease effects a unidirectional cleavage of the cyclic
2 ',3 '-phosphates of uridine and cytidine yielding exclusively the 6(3')-
isomers; ribonuclease, however, had no action on the cyclic 2 ',3 '-phosphates
of adenosine and guanosine, a fact which supports the specificity of ribo-
nuclease for pyrimidine nucleotide derivatives. In the course of this inves-
tigation it was shown that the phosphoryl group in cytidjdic acid b oc-
cupied the same position (probably 3') as in uridylic acid b by converting
the former to the latter by deamination under alkaline conditions which pre-
cluded phosphoryl migration. In both the cyclic pyrimidine nucleotides,
then, it was clear that ribonuclease cleaved the C2' — ^O — P linkage. It
is of some interest to note in connection with the demonstration that cyclic
phosphates are intermediates in ribonuclease digestion that in 1946
Schramm, Bergold, and Flammersfeld''^ had suggested the presence of

1" R. Markham and J. D. Smith, Research 4, 344 (1951).

108 R. Markham and J. D. Smith, Nature 168, 406 (1951).

">» D. M. Brown, C. A. Dekker, and A. R. Todd, J. Chem. Soc. 1952, 2715.

i^« E. Volkin and W. E. Cohn, Federation Proc. 11, 303 (1952).

I'l G. Schramm, G. Bergold, and H. Flammersfeld, Z. Naturforsch. 1, 328 (1946).

426

D. M. BROWN AND A. R. TODD

"inner esterified" nucleotides in ribonuclease digests. The evidence for this
suggestion, however, rested on the disparity between the number of acidic
functions Uberated by the enzyme in relation to the conclusion, now known
to be erroneous, "2 that the products were all mononucleotidic.

•0

OH OPO3H2

CH2-0H

RNase 0=^

*' N

CH-H — h
I H H

N.

H

■CH,-OH

NHo

NH,

alkali

■0-

CHj-OH

RNase

N,

CH-

I

N.

OH OPO3H2

■CH,-OH

H H

H

OH

Markham and Smith"^ have studied the oligonucleotide fragments in
ribonuclease digests by paper chromatographic and paper electrophoretic
techniques, thereby separating a variety of di- and trinucleotides. They
found that during the rapid initial action of the enzyme substances were
produced which were then transformed more slowly into the oligonucleo-
tides without change in the gross analytical composition (base-phosphorus
ratio). They suggested, on the basis of the hydrolytic mechanism advanced
by Brown and Todd,®^ that these initially produced substances carried a
cyclic 2',3'-phosphoryl group on the terminal (pyrimidine) nucleoside
residue; further action of the enzyme on these cyclic groups then yielded
the true oligonucleotides. Brown, Dekker, and Todd^°^ pointed out that
this latter process, being analogous to that studied in their work on the ac-
tion of ribonuclease on cyclic phosphates of uridine and cytidine, should
yield oligonucleotides bearing a terminal 6 (3') -nucleotide residue, and
hence that on alkaline hydrolysis pyrimidine h nucleotides would be pro-
duced from them. In agreement with this view, Volkin and Cohn^^ found

"2 Inter al., G. Schramm, W. Albrecht, and K. Munk, Z. Naturforsch. 7b, 10 (1952).
"3 R. Markham, and J. D. Smith, Biochem. J. 52, 558 (1952).

CHEMICAL BONDS IN NUCLEIC ACIDS

427

that digestion of ribonucleic acid with ribonuclease, followed by treat-
ment with alkali, yielded only the b isomers of the pyrimidine nucleotides,
together with a mixture of the a and b isomers of the purine nucleotides.
The processes involved are shown in the annexed formulas in which Py =
uracil or cytosine residue and R = remainder of polynucleotide chain;
where R = H the final products of ribonuclease action are the pyrimidine
b mononucleotides which are unaffected by treatment with alkali.

Ribonucleic acid

Pyrimidine 6 nucleotides and
purine a+ b nucleotides

CHj-OR

The evidence so far presented in this section regarding ribonuclease ac-
tion may now be summarized. The initial reaction of the enzyme evidently
occurs specifically at pyrimidine nucleotide sites in the ribonucleic acid
molecule. This specificity is discussed in more detail later, but it is supported
by the observed faihire of ribonuclease to attack the cyclic 2 ',3 '-phos-
phates of adenosine and guanosine,*^'^"^ although it readily attacks their
uridine and cytidine analogues. Subject to this specificity, and to the pro-
duction of only the b isomers of pyrimidine mononucleotides, ribonuclease
digestion follows a course akin to that of alkaline hydrolysis. The course
of ribonuclease hydrolysis and the various degradations described in this
section can be conveniently represented in the appended scheme using the
earlier postulated 3',5'-linked polynucleotide to represent ribonucleic acid
(in this scheme Py = uracil or cytosine residue, Pu = adenine or guanine
residue, and C2' — C3. — €5- is again used as an abbreviated form of the sugar
residue). In this representation of ribonuclease action, the pyrimidine
mononucleotide fraction arises from sites in the nucleic acid where two or
more pyrimidine nucleotide residues are adjacent, a dinucleotide from sites
where a purine nucleotide is flanked on either side by a pyrimidine nucleo-
tide, and larger fragments from positions where several purine nucleotides
occur consecutively."'' '"'* The production of another type of fragment by
ribonuclease action, viz., the so-called "core,"^''" is not substantiated by
recent work.^^"^ It should be pointed out that although the reactions are

1'^ R. Markham and J. D. Smith, Biochem. J. 52, 565 (1952).

428

D. M. BROWN AND A. R. TODD

represented in the scheme as applying to a straight-chain polynucleotide,
they are equally readily accommodated on branched-chain structures
which fulfil the conditions laid down on the basis of the results of chemical
hydrolysis (p. 422).

Py-Cj. Pu-Ca- Py-Cr Py-Cz- Pu-Cj-

Ca- Ca- Cs- Cs-

I ^P I ^P ^P I ^P
C^. C5. C5. C5.

.^

0 Pu-C2- Py-Cz-.

OH

Pu-Ca- Py-Ci

RNase (rapid)
^0 Py-C2;

RNase (slow)

I

Py-Cr

>^

0 PU — Cy

OH

r

\i

Co-OPOsHs

C3-OP03Hj

C,.

prostate phospliatase

Pu-C2- Py-C2.-0H Py-Cj-OH

alkali

1

Pyrimidine b nucleotides and
purine a+b nucleotides

a.

C,.-OH

C,.-OH

Cfi.

C5.-OH

J

periodate

Oxidation products

Purine nucleotides and
pyrimidine nucleosides

6. Structure of Oligonucleotides Derived from Ribonucleic Acids

That the general stucture of the oligonucleotides present in ribonuclease
digests is correct seems clear from the degradations described above. Their
chemistry appears to be wholly consistent with the requirements set out by
Brown and Todd^^ f^om their discussion of chemical hydrolytic mechanisms.
It is unfortunate that up to the present, no oligonucleotides have been
isolated in substance, apart from the compounds obtained by Merrifield

CHEMICAL BONDS IN NUCLEIC ACIDS

429

and WooUey" from acid hydroly sates of yeast ribonucleic acid. Neverthe-
less, fractionations of oligonucleotide mixtures in ribonuclease digests have
been achieved by paper chromatography and electrophoresis,"^ and by
ion-exchange chromatography,^^ which have given solutions of apparently
homogeneous substances. Using ion-exchange chromatography, as many
as 30-40 distinct elution peaks have been observed, and many of these
have been shown to represent solutions of virtually pure oligonucleotides.
Using such solutions, the structure of a number of oligonucleotides has been
determined. The following example,^^ set out schematically below, will
suffice as an example of such a determination carried out on a trinucleotide
containing the bases uracil, adenine, and guanine in the ratio 1:1:1 ob-
tained from a ribonuclease digest (in the formulas U, A, and G represent
the uracil, adenine, and guanine residues).

OPOjHj
U-C2.-C3.-C5.

P

/
A C2-^C3: -Cj.

P

/
G— C2— C3.— Cj.

U-C2.-C3.--C5

barley 6 (3')- " J y 0

/

venom
diesterase

Uridylic acid b and

idenylic and guanylic

acids a and b

G_C2.-C3.-C5.
XXII

alkali

Uridine and adenylic

and guanylic acids

a and 6

Uridine-5'-phosphate
and

Adenosine-5'-phosphate
and

Guanosine

Alkaline hydrolysis yielded uridylic acid h, together with the mixed a and
b isomers of adenylic and guanylic acid. By removal of the terminal phos-
phoryl group with phosphomonesterase or barley 6(3')-niit-leotidase,"^ a
product XXII was produced whose molecular weight could be deduced
from the ratio of inorganic to total phosphorus. With snake venom diester-
ase XXII underwent fission at the Cs- — ^O — ^P linkages, giving guanosine
and the 5 '-phosphates of uridine and adenosine, while with alkali it gave
uridine and the mixed purine a and h nucleotides. Treatment of other oli-
gonucleotides first with phosphomonoesterase, and then with snake venom
diesterase, always yielded only nucleoside-5 '-phosphates and a nucleoside
representing the terminal residue, indicating that all were linear structures
since chain-branching would have been expected to lead to other types of
breakdown products by the action of snake venom diesterase.

The earlier work of Merrifield and Woolley," in which they isolated and

"» L. Shuster and N. O. Kaplan, Federation Proc. 11, 286 (1952).

430 D. M. BROWN AND A. R. TODD

characterized several di nucleotides from acid hydrolysates of yeast ribonu-
cleic acid, is of considerable interest. For structural determination they
used methods similar to those just described except that intestinal phos-
phatase, inhibited with arsenate, was used instead of snake venom diesterase
to demonstrate the C 5 /-linkages. Since the formation of their dinucleotides
depended on a random fission of the polynucleotide chain, they obtained
products containing purine and pyrimidine residues isomeric with those
found in ribonuclease digests, but in which the terminal phosphoryl group
was attached to the purine nucleoside residue and hence could be further
degraded by ribonuclease. Moreover, they were able to separate dinucleo-
tides which were isomeric with one another by virtue of the position (a or
h) of the terminal phosphoryl group, in accordance with prediction from
the postulated mechanism of chemical hydrolysis, ^^ in which phosphoryl
migration must always accompany fission of the diester linkages between
individual nucleoside residues in the ribonucleic acid molecule.

7. Evidence for 3'(6)-Linkages in Ribonucleic Acids

In discussing the evidence so far presented on ribonuclease action, the
3',5'-linked polynucleotide structure has been assumed. However, since
degradation with ribonuclease proceeds via intermediates with cyclic 2',3'-
phosphate groupings it is clear that, as in the case of chemical hydrolytic
studies, no differentiation can be made between C2' and Ca- as a linkage
point in the original ribonucleic acids. These could be either 2', 5'- or 3',5'-
linked structures, or might contain both types of linkage. Clearly, finality
regarding the linkage requires a decision between C2' and Cs-.

A solution to this problem was found in the study of the action of
enzymes on simple esters of the mononucleotides. Brown and Todd®' ex-
amined the action of ribonuclease on the benzyl esters of the a and b iso-
mers of the pyrimidine nucleotides. Cytidine benzyl phosphate b and uri-
dine benzyl phosphate b were converted by ribonuclease into cytidylic acid
b (cytidine-3 '-phosphate) and uridylic acid b (uridine-3 '-phosphate), hy-
drolysis proceeding in each case by way of the intermediate cyclic 2',3'-
phosphates. The benzyl esters employed in these experiments were oriented
by catalytic hydrogenation to the parent nucleotides, a process unlikely to
involve phosphoryl migration. The methyl and ethyl esters were similarly
hydrolyzed by ribonuclease. The isomeric esters of the pyrimidine a nu-
cleotides were completely unaffected, nor had ribonuclease any action on
esters of either a or 6 isomers of the purine nucleotides. Not only does this
further justify the belief in specificity of ribonuclease for pyrimidine nu-
cleotide ester linkages in ribonucleic acids, but it leads to important con-
clusions about the linkage position. Since, as shown above, all the inter-
nucleotidic linkages involving pyrimidine nucleotide residues are apparently

CHEMICAL BONDS IN NUCLEIC ACIDS

431

hydrolyzed by ribonuclease, it is clear that the 6(3') -position of the pyrimi-
dine nucleoside residues is involved exclusively in the internucleotidic
linkage. Only in the case of branched-chain structures could the a (2') -posi-
tion be involved and then only as a branching point; this exceptional case
will be discussed later.

•0-

OCH,Ph
I
OHOPO-OH

R-CH-

H H

H

CH,-OH

CHo-OH

R-CH

CHo-OH

H H

In discussing the intimate, action of ribonuclease, Brown and Todd*^
suggest that the fundamental catalytic action is the formation and subse-
quent rupture of Ca{2-) — 0 — P bond as indicated schematically below.
The elimination of the residue R as the alcohol ROH is a necessary conse-
c^uence of this action; the close similarity between ribonuclease and chemi-
cal hydrolytic agents is apparent.

•OH .0-
P^O -
O OR

R Nase

(fast)

.^'

— o ^0-

+ R-OH

(slow)

-OH
OPO(OH)j

This interpretation gives a simple explanation for the apparently two-
fold action of ribonuclease on ribonucleic acids. The initial "depolymeriza-
tion" without liberation of acid functions, followed by the slow liberation
of acidic secondary phosphoryl groups observed b}^ Chantrenne, Linder-
str0m-Lang, and Vandendriessche^^^-^^ using a dilatometric method is
undoubtedly related to the above formation and cleavage of cyclic phos-
phoryl groups. Vandendriessche"' has indeed shown that the volume
changes observed during the action of ribonuclease on ribonucleic acids are
closely parallel to those observed in the ribonuclease hydrolysis of the
mononucleotide esters described above.

Although the study of ribonuclease action on mononucleotide esters es-

"^ H. Chantrenne, K. Linderstr0m-Lang, and L. Vandendriessche, Nature 159, 877,

(1947).
'•' L. Vandendriessche, Acta Chem. Scand. 7, 699 (1953).

432 D. M. BROWN AND A. R. TODD

tablishes 6(3') as the linkage point for pyrimidine nucleoside residues,
studies with this enzyme can give no information about the corresponding
purine nucleoside residues. Evidence on this point comes from studies
using other nuclease preparations with different specificities, several of
which appear to exist. ^'"' Volkin and Cohn"** showed that spleen nuclease
prepared according to Maver and Greco"^ yields the h isomers of both
pyrimidine and purine mononucleotides when it acts upon ribonucleic acid.
Further purification of this enzyme preparation by Heppel and Hilmoe"'
gave a product which degrades ribonucleic acids and oligonucleotides,
giving high yields of mononucleotides which, in the case of adenylic and
guanylic acids, were shown to be the 6 (3') -isomers -j'^" since no evidence for
intermediate cyclic phosphates was observed, it was concluded that the
purine nucleotide residues were linked at the 6 (3') -position in the intact nu-
cleic acid. More definite evidence has been provided by Brown, Heppel,
and Hilmoe,*-^ who have shown that the same enzyme preparation, as well
as others from intestine, potato, and rye-grass, hydrolyze cytidine benzyl
phosphate b and adenosine benzyl phosphate b to the corresponding b
nucleotides while they have no action on esters of the a isomers.

Regardless of the mechanism of spleen nuclease action it is clear that
these results, together with the above studies of ribonuclease hydrolysis
of nucleotide esters, establish with a high degree of certainty that the inter-
nucleotidic linkage in ribonucleic acids involves the 6(3') -position and not
the a(2')-position in both purine and pyrimidine nucleoside residues. As-
suming the validity of the constitutions proposed for the mononucleotides,
i.e., that the b isomers are the 3 '-phosphates, the ribonucleic acids must, on
the evidence presented, be considered to be polynucleotides in which the
individual nucleoside residues are joined one to the other by phosphodi-
ester linkages between the 3'- and 5 '-positions as indicated in structure
XVI.

Although the oligonucleotides produced by ribonuclease action appear to
be unbranched, this fact cannot of itself be accepted as evidence that intact
ribonucleic acids are linear polynucleotides. To complete our discussion it
is necessary to consider the question of chain-branching.

8. Chain-Branching in Ribonucleic Acids

There has been much discussion in the past about branched-chain as
distinct from linear structures for the ribonucleic acids. This involved, in

"8 M. E. Maver and A. E. Greco, J. Biol. Chem. 181, 861 (1949).

"9 L. A. Heppel and R. J. Hilmoe, Federation Proc. 12, 217 (1953).

'2" L. A. Heppel, R. Markham, and R. J. Hilmoe, Nature 171, 1152 (1953).

'2' D. M. Brown, L. A. Heppel, and R. J. Hilmoe, J. Chem. Soc, 1954, 40.

CHEMICAL BONDS IN NUCLEIC ACIDS

433

the main, discussion of branching from the main cliain by incorporation of
phosphotriester linkages.'" However, with the advent of an adequate ex-
planation of the hydrolytic behavior of ribonucleic acids it became possible
to define with some certainty the possible types of branching which might
be considered. As has already been briefly mentioned, Brown and Todd,®^
starting from the requirement that ribonucleic acids are hydrolyzed
to mononucleotides under mild alkaline conditions, pointed out that in
addition to branching on phosphorus, branching by attachment of polynu-
cleotide chains to C2' in one of the nucleoside residues in a main polynu-
cleotide chain of type XVI could be considered, provided that the attach-
ment was through a normal phosphodiester group attached to C3. (or C2')
in the first nucleoside residue of the branching chain; only in this way
could alkali-lability be maintained. C5. could not be considered as the point
of linkage in the first residue of the branch as the result would be an alkali-
stable system, cyclization either in that residue or in the residue forming
the branch-point in the main chain being impossible. These authors there-
fore proposed a general structure of the type XXIII for branched ribo-
nucleic acids. In this structure the branches can be extended following the
normal C3. — C5. .sequence as hi the main chain. The incorporation of a
certain number of branches involving phosphotriester linkages into such a
structure would present no problem provided, again, that the Cs- — Cs-
sequence was followed in the branches.

C^. C3. C5.

X

C. /C,.— 03,-05.

\

■ Co. O c

Oo. Oo. Oi

"p

\

XXIII

Some support for the branched-chain structure XXIII was provided by
methylation studies.'- '^-^ Methylation of yeast ribonucleic acid, followed
by hydrolysis and separation of the mixture of sugar derivatives obtained,
yielded ribose as well as its mono- and dimethyl derivatives. The presence
of ribose was taken to indicate triply-substituted ribofuranose residues in
the nucleic acid.''^^ It is, however, doubtful if the method of methylation
can be safely applied in this field on account of the lability of the internu-

'" A. S. Anderson, G. R. Barker, and K. R. Farrar, Nature 163, 445 (1949).
•" A. S. Anderson, G. R. Barker, J. M. Gulland, and M. V. Lock, ./. Chevi. Soc. 1952,
369.

434 D. M. BROWN AND A. R. TODD

cleotidic linkages in the presence of alkaline reagents and the difficulty of
assessing the completeness of methylation.^^sa

In discussing structure XXIII, Brown and Todd*^- suggested that the rapid
production of pyrimidine mononucleotides during ribonuclease digestion
might be explained by postulating that branches were short and frequent —
perhaps consisting of only one nucleoside unit in some cases — and consisted
of pyrimidine nucleotide residues. Fission at the C2' — O — P bond at the
branch-point would yield pyrimidine nucleotides (via cyclic phosphates)
readily. Rather similar suggestions based, however, on branched structures
of undefined type, were also advanced by Carter and Cohn^^ and Magasanik
and Chargaff.27 ^pj^g more recent developments in our knowledge of ribo-
nuclease action make it clear that it is not necessary to postulate branches
as the source of the pyrimidine mononucleotides, but it has made it evi-
dent that in any branching of the type shown in XXIII the first residue in
the branch must be a pyrimidine nucleotide i^^ if it were not, branched oligo-
nucleotides would be formed.

Cohn and Volkin^^^ have brought forward further evidence bearing on the
problem of branching from studies using snake venom diesterase, and have
discussed their results in terms of structure XXIII. Following the work of
Gulland and Jackson^^ with snake venoms which contain diesterases and
5 '-nucleotidases, they were able to confirm that, acting on ribonucleic acids,
large amounts of inorganic phosphate are liberated. In addition, however,
they showed that cytidine diphosphate and uridine diphosphate were pro-
duced simultaneously in an amount corresponding to about 30% of the
pyrimidine content of the nucleic acid. That the inorganic phosphate origi-
nated mainly in nucleoside-5 '-phosphates was shown by experiments using
venom which had been freed of 5 '-nucleotidase by the method of Hurst,
Little, and Butler. ^^ This purified preparation acting on ribonucleic acids
from calf-liver, thymus, and yeast, liberated very little inorganic phos-
phate but yielded large amounts of all four nucleoside-5 '-phosphates, the
pyrimidine nucleoside diphosphates, some nucleosides (mainly purine),
and about 10 % of pyrimidine h nucleotides. The diphosphates were shown
by enzymic degradation to be mixtures of cytidine-2' ,5'- and cytidine 3',5'-
diphosphate (XXI Va and b; R = cytosine residue) and the corresponding
uridine diphosphates (XXI Va and b; R = uracil residue), structurally

123a More recent work (D. M. Brown, D. I. Magrath, and A. R. Todd, J. Chem. Soc.
1954, 1442) has shown that during the methylation of uridylic acid b, phosphoryl
migration occurs; the method is therefore unlikely to afford reliable evidence
when applied to polynucleotides since internucleotidic bond fission must on this
evidence be expected to accompany methylation.

12^ E. Volkin and W. E. Cohn, /. Biol. Chem., 203, 319 (1953).

CHEMICAL BONDS IN NUCLEIC ACIDS 435

analogous to the adenosine diphosphates obtained by degradation of tri-
phosphopyridine nucleotide^-* and coenzyme A.^^^

I OPO3H., OH
R-CH

-CHo-OPOsH,

I OH OPO3H2

•CHo-OPOsHj R-CH \ \

H H H H H H

XXIVa XXIVb

The pyrimidine nucleoside diphosphates do not appear to be produced
from the end-groups of the ribonucleic acids bearing monoesterified phos-
phate since they are obtained in undiminished yield from nucleic acids pre-
viously treated with bone phosphomonoesterase. This enzyme removed
about 10 % of bound phosphate, considered on other grounds to represent
end-group monoesterified phosphate. Oligonucleotides produced by ribo-
nuclease action from the nucleic acids were attacked very slowly, if at all,
by the purified venom.

Volkin and Cohn^^* explain these observations on the basis of a branched-
chain structure based on XXIII which can be represented by XXIIIa. They

C,.— C3.— C,.

C,. C3.— C,. a

e c ^.

€2- Cj. ^Cj. C3. Cj, d

\ \P "^P
/ \
C3.1 C3. Cg. C3. Cj. 6

xxnia

suggest that the diesterase causes fission of the internucleotidic linkages at
the Czr — 0- — P bonds (i.e., at the broken lines in XXIIIa) liberating nu-
cleoside-5 '-phosphates (from residues a, 6, and c) as in the case of the poly-
deoxy ribonucleotides.^^ The diphosphates would arise from branch points
(residue d) and the (purine) nucleosides from the terminal residue of the
branch (residue e). The purine nucleoside produced in their experiments was
in fact equivalent in amount to the nucleoside diphosphates. For this ex-
planation to hold good, they had to postulate that the residue at the branch-
ing point {d) is a pyrimidine nucleoside residue. The first residue (c) in the
branch must also be a pyrimidine nucleoside residue to accomodate the

'25 A. Romberg and W. E. Pricer, Jr., J. Biol. Chem. 186, 557 (1950).

'26 T. P. Wang, L. Shuster, and N. O. Kaplan, J. Am. Chem. Soc. 74, 3204 (1952).

436 D. M. BROWN AND A. R. TODD

results of ribonuclease studies, as already pointed out. The small amount of
pyrimidine h mononucleotide present in the venom digests was tentatively
regarded^^^'^" as originating in side chains consisting of only one nucleotide
residue attached through triply-esterified phosphorus, as in the early struc-
tural proposals of GuUand and his colleagues^". Their own and other'^** obser-
vations indicate that some 10% of the phosphoryl groups in ribonucleic
acids are singly esterified.

While these experiments favor a branched-chain structure for ribonu-
cleic acids, it would be unwise to rely entirely on evidence from only one
type of degradation and further confirmation would be welcome. In any
case, it is not impossible that among the natural ribonucleic acids both
linear and branched-chain polynucleotides may occur; such a possibility
has, in fact, been suggested^^^ and certainly there appears to be no evidence
against it at the present time.

9. Nucleotide Sequence in Polyribonucleotides

In previous pages the problem of the nature and position of the internu-
cleotidic linkage has been discussed in the light of recent work. A clear pic-
ture has emerged in which the nucleoside residues in the ribonucleic acids
are joined at their 3'- and 5 '-positions through phosphodiester linkages;
the possible types of branching have been indicated and evidence for
their occurrence has been reviewed. In brief, it appears that for a straight-
chain ribonucleic acid the general structure XVI will apply, and for
branched-chain acids structure XXIII, with the possible addition of some
branching through phosphotriester linkages. The degree to which branch-
ing may occur in ribonucleic acids generally is not yet established, but this
does not affect the general structures as far as the internucleotidic linkage is
concerned.

In discussing these structures little attention has been paid to the order
in which the four different nucleotide residues occur in the polynucleotide
chain. It has, however, been pointed out that in a branched structure
XXIII the first residue in the branch must be a pyrimidine nucleotide resi-
due, and that in the more specific structure XXIIIa the branching point
in the main chain must also be a pyrimidine nucleotide residue. Only in
the case of the small oligonucleotides has residue sequence been deter-
mined and the results of such determinations do not shed a great deal of
light on the situation in the intact nucleic acids. Nucleotide sequence in
ribonucleic acids cannot be simply elucidated by partial degradation to

1" W. E. Cohn, D. G. Doherty, and E. Volkin, Phosphorus Metabolism 2, 339 (1952).
1" G. Schmidt, M. Seraidarian, K. Seraidarian, and S. J. Thannhauser, Federation

Proc. 11,283 (1952).
1" G. Schramm and B. von Ker6kjdrt6, Z. Natyrforsch. 7b, 589 (1952).

CHEMICAL BONDS IN NUCLEIC ACIDS 437

polynucleotides, followed by structural determinations on these and re-
construction to yield a unique solution. The fact that only four mononu-
cleotides are involved makes this type of approach in some respects less
likely to succeed than in the case of the proteins, as is evident when one
considers the results obtained by analyzing the oligonucleotides isolated
from ribonuclease digests of ribonucleic acids. ^^•"•^•^29a /^\\ ^\^q possible
residue arrangments are found in the di- and trinucleotides (with terminal
pyrimidine residues) obtained in this way, and there is little reason to be-
lieve that the same will not also hold for the mixture of larger nondialyzable
fragments which are obtained. The problem would be even more compli-
cated if, as Markham and Smith'" suggest, ribonucleic acids are to be re-
garded as mixtures of large numbers of relatively small molecules rather
than single species of very high molecular weight. These authors arrived
at this suggestion on the basis of a method of end-group determination de-
pending on removal of terminal monoesterified phosphate with prostatic
phosphatase; using it, they found all four nucleotides as end-groups in
addition to nucleotides bearing a cyclic phosphoryl group. Their conclu-
sions seem open to criticism, however, firstly because chain-branching,
which they do not consider, could account for increased numbers of end-
groups, and secondly because the presence of traces of diesterases in their
enzyme preparations would lead to high apparent figures for end-groups
and correspondingly low estimates of molecular weight.

At the present time, available evidence does not indicate anything more
than a rather random sequence of residues, and it seems clear that progress
must depend (a) on the development of methods of stepwise degradation
and (b) some certainty as to the individuality of a given ribonucleic acid —
clearly if it were a mixture of different molecular species"" efforts to de-
termine sequence would be useless. A method of stepwise degradation is
in any case of considerable value, since quite apart from its use on intact
ribonucleic acids, such a method is necessary if the structure of the larger
oligonucleotides is to be determined. The essential requirement for such a
method is that the terminal internucleotidic linkage in a chain should be
rendered more labile than any other similar linkage in the molecule so as to
permit removal (and identification) of that residue, leaving the remainder
of the polynucleotide intact and ready for a repetition of the same process.
The lability of the internucleotidic linkage severely limits the choice of
possible methods. Brown, Fried, and Todd'^' have proposed a method

129a 'YYiQ recent demonstration (L. A. Heppel and P. R. Whitfeld, Proc. Biochem. Soc.
56, ii, 1954; L. A. Heppel, P. R. Whitfeld, and R. Markham, ibid. 56, iii, 1954)
that ribonuclease acts reversibly must render ineffectual any attempt to derive
the structure of a polynucleotide from a consideration of the products of its action.

'■'n V. Desreux and J. M. Ghuysen, Bull. soc. chim. Beiges 60, 410 (1951).

"1 D. M. Brown, M. Fried, and A. R. Todd, Chemistry & Industry 1953, 352.

438 D- M. BROWN AND A. R. TODD

which meets the above requirements and which depends on the known
lability towards alkali of esters, and in particular, phosphates of /3-keto and
/3-aldehydo alcohols. Such phosphates readily undergo an elimination reac-
tion under these conditions ;^^2 the reaction is illustrated by the conversion
of glyceraldehyde-3-phosphate to lactic acid under very mild alkaline and
pyruvaldehyde under acid conditions. ^^'-^^^ In model experiments,^^^
adenosine-5 '-phosphate and adenosine-5' benzyl phosphate were oxidized
with periodic acid to give dialdehydes of formula XXV (Ad = adenine
residue, R = H or CH2C6H5) in which the phosphate residue is attached
in the /3-position to one of the aldehyde groups. In accordance with expecta-
tion, these products were extremely labile to alkali, the phosphoryl (or
benzylphosphoryl) group being rapidly removed at room temperature
even at pH 10.5; under these conditions the unoxidized nucleotides, as well
as ribonucleic acids,^^^^ are completely stable. Since adenosine-5' benzyl
phosphate is structurally analogous to a polynucleotide (in the latter the
benzyl group is replaced by a polynucleotide chain) the potentialities of
this observation are evident. If formula XXVI represents a polynucleotide
made up of n nucleotides, removal of the terminal phosphoryl groups by
means of a phosphomonoesterase would yield XXVII, which, containing
a free a-glycol system, would be oxidized by periodic acid to XXVIII. By
analogy with oxidized adenosine-5 '-phosphate, mild alkaline treatment at
pH 10.5 would be expected to remove the oxidized terminal nucleoside
residue (which could be identified by various methods) yielding XXIX,
which is a polynucleotide containing n — 1 residues but otherwise exactly
Uke the original XXVI ; on this product the whole process could be repeated.

Although the method has not yet been applied to a large polynucleotide,
there is no reason to doubt its validity in such a case, especially as it has
been found applicable to dinucleotides where n = 2}^^ At the present time
this method represents the only reasoned approach to determination of
nucleotide sequence; if successful in its application it should also provide
definite information on the nature and extent of any branching which exists
in a given polynucleotide.

It is of interest to note that the polynucleotide XXVI can be regarded as
a polymer in which the monomeric units are nucleoside-3 '-phosphates. It
would be equally reasonable (and it would not affect the previous discus-
sion) if it were represented as XXVIa, in which the monomeric units are
nucleoside-5 '-phosphates. To which of these two types the natural nucleic
acids belong is at present unknown. That apparently successful end-group

132 R. p. Linstead, L. N. Owen, and R. F. Webb, /. Chem. Soc 1953, 1211.

133 O. Meyerhof and K. Lohmann, Biochem. Z. 271, 89 (1934).
13* E. Baer and H. O. L. Fischer, J. Biol. Chem. 150, 223 (1943).
134a Cf. also C. A. Zittle, J. Franklin Inst. 242, 221 (1946).

136 P. R. Whitfeld and R. Markham, Nature 171, 1151 (1953).

CHEMICAL BONDS IN NUCLEIC ACIDS

439

-O

OH OH

Ad-CH-

H H H

OPO3H2

Base— C2-— Cg— C5.

P
Base — C2' — Cg.— C^,

P
Base — C2- — C3. — Cg.
XXVI

OPO3H2

Base — C 2- — C3. — Cj,

P'
Base-Co.— cC-a.

Base — Cj.— Cg.— Cs-
XXIX

O

II
ch^-op:

,0R
^OH

•0-

0

Ad-CHCHO OHC-CH-CH,-OP:

XXV

Base — C 2'""G3'— Cj.

Base-C2--C3.-C5._j^_,

Base-C2.-C3.-Cj.
XXVII

.OR
^OH

-0

Base-CHCHO OHC-CH^CHj
P

Base — Cj. C3. Cg,

Jn-2

Base — Co.-

— CgT-

XXVIII

■'C.,

Base-C

2-

n-2

"^3- /Cg.
P

Base — C J. ~ Cg. — C5.

P^
Base— Cj. — Cg. "Cg.

H2O3PO
XXVIa

determinations have been made using phosphomonoesterase to remove a
terminal Cs'-phosphoryl group is not valid evidence of XXVI unless it can
be shown that the nucleic acid is completely intact; any partial degradation
of either type of polynucleotide would inevitably lead to structures of type
XXVI, on the basis of the accepted mechanism of hydrolytic breakdown. ^^^^
A similar uncertainty exists in the case of the deoxyribonucleic acids.

III. Structure of the Deoxyribonucleic Acids

The deoxyribonucleic acids have generally been considered on titrimetric
data to be essentially high-molecular-weight polynucleotides which are

"6a Since this review was written, Markham, Matthews, and Smith {Nature 173, 537,
1954) have shown that the nucleic acids of tobacco mosaic virus and potato virus
X belong in large proportion to the class represented by XXVIa; alkaline hydroly-
sis yields nucleoside-2'(and -3'),5'-diphosphates and nucleosides originating from
terminal residues. Other nucleic acids, e.g., from turnip yellow mosaic virus, ap-
pear to conform to structures of type XXVI, although, as indicated above, this
could be a reflection of their degraded state.

440 D. M. BROWN AND A. R. TODD

substantially iinbranched, and in which the individual nucleotides are
joined by phosphodiester linkages.-" Certainly most of the available evi-
denced^ ^ is in agreement with this although there has been an isolated sug-
gestion,^" on titrimetric results, that chain-branching through phosphotri-
ester linkages (about once in every 10-20 linkages) occurs and some slight
degree of branching has been also suggested from work using light-scatter-
ing techniques.^" These suggestions are at present unsubstantiated by other
physical studies, and no evidence for branching has been found in degrada-
tive studies. The observations on which they rest may be susceptible of
other interpretations in the light of recent structural models proposed by
other workers^^**'^^^ in which hydrogen-bonding between associated poly-
nucleotide chains plays a major role. For the purposes of the present discus-
sion, therefore, the deoxyribonucleic acids will be considered as essentially
unbranched polynucleotides.

In their discussion of the hydrolytic behavior of the nucleic acids. Brown
and Todd''- adopted for deoxyribonucleic acids a general structure of type
XXX (Base — C3-— Cs- represents a nucleoside unit), in which the deoxy-
nucleoside units are linked by 3 ',5 '-phosphodiester groupings. This struc-
ture, they pointed out, was in accord with the alkali-stability of deoxyribo-
nucleic acids, and it is also borne out by the results of chemical and enzymic
degradation.

Base Base Base Base
C3. C3. C3. C3.

Cs' Cg- Cs' Cg-

XXX

It was claimed many years ago by Levene and Jacobs^*" that acid hydroly-
sis of thymus deoxyribonucleic acid under fairly vigorous conditions yielded,
among other products, two substances which were diphosphates of the two
pyrimidine deoxyribonucleosides, thymidine and deoxycytidine ; this claim
was subsequently disputed by other workers. ^^^ Dekker, Michelson, and
Todd^^2 have reinvestigated this matter using deoxyribonucleic acid from
herring sperm and have vindicated the claim of Levene and Jacobs."" "^'^^^

138 Inter al., R. Signer and H. Schwander, Trans. Faraday Soc. 46, 790 (1950).

1" M. E. Reichmann, R. Varin, and P. Doty, J. Am. Chem. Soc. 74, 3203 (1952); P.

Doty and B. H. Bunce, ibid. 74, 5029.
138 L. Pauling and R. B. Corey, Nature 171, 346 (1953).
"9 J. D. Watson and F. H. C. Crick, Nature 171, 737 (1953).
1" P. A. Levene and W. A. Jacobs, /. Biol. Chem. 12, 411 (1912).
1^1 H. Bredereck and G. Caro, Z. physiol. Chem. 253, 170 (1938).
"2 C. A Dekker, A. M. Michelson, and A. R. Todd, J. Chem. Soc. 1953, 947.
1" See also P. A. Levene, J. Biol. Chem. 48, 119 (1921); 126, 63 (1938).

CHEMICAL BONDS IN NUCLEIC ACIDS 441

They were able to isolate from acid hydrolysates thymidine-3', 5 '-diphos-
phate (XXXI; R = thymine residue) and deoxycytidine-3', 5 '-diphosphate
(XXXI ; R = cytosine residue) and to establish their structure by synthesis;
they also obtained some evidence for the presence of small amounts of the
corresponding 3 ',5 '-diphosphate of 5-methyldeoxycytidine in their hydroly-
sates. The isolation of these substances is a clear indication that both the
3'- and 5 '-positions in the deoxyribonucleosides are involved in the inter-
nucleotidic linkage and emphasizes the difference between the hydrolytic
mechanism which operates in ribonucleic and deoxyribonucleic acids.

O-

I OPO3H2

R-CH-CHj

I OH

CHj-OPOsH. R-CH-CH2 — \-

CHz-OPOjHa

H H H H

XXXI XXXII

Chemical hydrolysis of deoxyribonucleic acids is not a satisfactory
method for obtaining simple mononucleotides mainly on account of the
lability of the glycosidic linkage in the purine deoxyribonucleotides to acid,
and the resistance of the normal phosphodiester group to alkaline hy-
drolysis. Klein and Thannhauser/"** however, using an arsenate-inhibited
intestinal phosphatase, succeeded by enzymic hydrolysis in isolating four
nucleotides (deoxyadenylic, deoxyguanylic, deoxycy tidy lie, and thymidylic
acid) ; Volkin, Khym, and Cohn'"*^ also described their separation from simi-
lar hydrolysates. Similarity to the ribonucleoside-5 '-phosphates in their
ion-exchange characteristics and their dephosphorylation^'*^ by a specific
5 '-nucleotidase""* indicated that, contrary to earlier assumptions, they were
the 5'-phosphates of the respective nucleosides (XXXII; R = purine or
pyrimidine residue). Conclusive proof for this formulation has been pro-
vided at least in the case of the pyrimidine nucleotides by identification of
the natural substances through direct comparison with thy midine-5 '-phos-
phate (XXXII; R = thymine residue) and deoxycy tidine-5 '-phosphate
(XXXII; R = cytosine residue) prepared by unambiguous methods by
Michelson and Todd."' The almost quantitative recovery (ca. 92%) of the
deoxyribonucleoside-o '-phosphates from snake venom diesterase hydroly-
sates of deoxyribonuclease-treated deoxyribonucleic acids^"''^ shows that

'••^ See also S. J. Thannhauser and B. Ottenstein, Z. physiol. ('hem. 114, 39 (1921);

S. J. Thannhauser and G. Blanco, ibid. 161, 116 (1926).
'" W. Klein and S. J. Thannhauser, Z. -phisiol. Chem. 218, 173 (1933); 224, 252 (1934);

231,96 (1935).
'" E. Volkin, J. X. Khym, and W. E. Cohn, J. Am. Chem. Soc. 73, 1533 (1951).
1" C. E. Carter, J. Ain. Chem. Soc. 73, 1537 (1951).
'« L. A. Heppel and R. J. Hilmoe, J. Biol. Chem. 188, 665 (1951).
'<' A. M. Michelson and A. R. Todd, J. Chem.. Soc. 1953, 951; 1954, 34.

442 D. M. BROWN AND A. R. TODD

all, or almost all, the internucleotidic linkages must involve the 5 '-position
of a nucleoside residue. These facts all point clearly to an essential structure
of type XXX.

Evidence from other enzymic studies supports this conclusion. The mode
of action of the crystalline deoxyribonuclease^^ is not yet understood, but
several workers^^''"^^^ have shown that a number of oligonucleotides of
varying size are produced when it acts on deoxyribonucleic acids. The use
of electrophoretic,^*^'^^^ paper chromatographic,^*^ and ion-exchange^*^
methods have permitted the separation and identification of some of these
oligonucleotides.. Sinsheimer and Koerner^*' have described two dinucleo-
tides, one containing two cytosine residues and the other an adenine and a
cytosine residue, and have determined their structures by enzymic hydroly-
sis. The method used is indicated below, using the deoxycytidine dinucleo-
tide XXXIII as an example. Phosphomonoesterase treatment yielded the

OPO3H2

Cyt-C3-^C5. ^,^^^^^^^^

P *■ 2 mol. deoxycytidine- 5'-phosphate

Cyt-Cg.— C5.
XXXIII

monoesterase
T

Cyt-Co.— C5.

\ diesterase 1 mol. deoxycytidine and

^\ 1 mol. deoxycytidine-5'-phosphate

Cyt-Cg.— Cs-

XXXIV

dinucleoside phosphate XXXI V^** which in turn yielded, with snake venom
diesterase, both deoxycytidine-5 '-phosphate and the free nucleoside, deoxy-
cytidine. Direct enzymic hydrolysis of the original dinucleotide (XXXIII),
on the other hand, yielded only deoxycytidine-5 '-phosphate, indicating
that a phosphoryl group is attached at C5- in both residues.

1. Nucleotide Sequence in Deoxyribonucleic Acids

Although, as the above discussion shows, the nature and position of the
main internucleotidic linkage in deoxyribonucleic acids seems clear, the
problem of nucleotide sequence remains. [Cf. Chargaff, Chapter 10.] On
this no definite statement can be made. As yet no method of stepwise degra-
dation comparable to that proposed for ribonucleic acids"^ has been evolved.

15" W. G. Overend and M. Webb, J. Chem. Soc. 1960, 2746.

1" A. H. Gordon and P. Reichard, Biochem. J. 48, 569 (1951).

162 J. D. Smith and R. Markham, Nature 170, 120 (1952).

1" R. L. Sinsheimer and J. F. Koerner, J. Am. Chem. Soc. 74, 283 (1952).

1" Cf. also J. D. Smith and R. Markham, Biochim. et Biophys. Acta 8, 350 (1952).

CHEMICAL BONDS IN NUCLEIC ACIDS 443

No attempts have been made to piece together information based on the
structure of oligonucleotides found in deoxyribonuclease digests; for rea-
sons similar to those discussed in connection with the ribonucleic acids (p.
437) it is, in any case, unlikely that such attempts would lead to a unique
solution. Chargaff and his co-workers, however, have sought to ascertain
whether there is any recognizable order in the nucleotide sequence of typi-
cal deoxyribonucleic acids. Thus Zamenhof and Chargaff'^ ''^^'^^^ have
studied the distribution of purine and pyrimidine bases in the dialyzable
and nondialyzable fractions of deoxyribonuclease digests. Their results
suggest a very complex pattern; it should be remembered, of course, that
the as yet unknown specificity of deoxyribonuclease and the effect of differ-
ent dialysis rates of the fragments must be taken into account in drawing
any definite conclusions from such studies. The separation of six dinucleo-
tides from such digests^ ^- seems to indicate that no simple regularity is to
be expected in the nucleotide distribution.

When deoxyribonucleic acids are subjected to mild acid hydrolysis,
materials are formed which were formerly described by the generic term
"thymic acid."^" Although most of the earlier preparations were highly
degraded, Chargaff and his co-workers have reinvestigated their prepara-
tion and obtained an interesting group of substances of molecular weight
ca. 15,000 which they term apurinic acids. ^^^ As their name suggests, they
are produced by removal of all purine residues from the polynucleotide,
through fission of the labile purine A'^-glycosidic linkages, and thus have
deoxyribose phosphate residues in place of the purine nucleotide residues
originally present in the nucleic acid. It is obvious that such products are
of great structural interest and their hydrolysis would merit close study.

Relatively vigorous acid hydrolysis of deoxyribonucleic acids yields, as
already indicated, the 3 ',5 '-diphosphates of thymidine and deoxycytidine.
Brown and Todd*'- originally suggested that a certain amount of these di-
phosphates might be expected as a result of random fission of internucleo-
tidic linkages. Dekker, Michelson, and Todd,'^^ however, have commented
on the unexpectedly large amounts of the diphosphates in acid hydroly-
sates and suggest that they owe their origin to some other mechanism
operating at those positions in the polynucleotide chain where pyrimidine
and purine residues are adjacent to one another in view of the above-men-
tioned evidence^" that the initial action of acid is to remove purine residues
from deoxyribonucleic acids. A probable mechanism is that discussed by
Brown, Fried, and Todd^^' as a basis for stepwise degradation of polyribo-

'" S. Zamenhof and E. Chargaff, /. Biol. Chem. 178, 531 (1949).

if^^S. Zamenhof, Phosphorus Metabolism 2, 301 (1952).

'" See footnote 158 for bibliography.

1*8 C. Tamm, M. E. Hodes, and E. Chargaff, J. Biol. Chem. 195, 49 (1952).

444 D. M. BROWN AND A. R. TODD

nucleotides, i.e., the tendency of phosphates of /3-aldehydo alcohols to un-
dergo ehmination reactions (cf. p. 437). The initial action of acid must lead
to products of the apurinic acid type. If we consider a section (XXXV;
R = next residue in the chain) of such a molecule, ready elimination should
occur at Cs- of the deoxyribose residue (here written in the aldehyde form)
as indicated by the broken line a a.

CHO Py-CH-

T o

CH r°\ OH CH-

"^0^ CHOH " p1 / CH— I

^CH2 0 ,' 0 CH2

0

XXXV

The alternative fission i.e., at the point indicated by the broken line h
h in XXXV, which would result from intervention of a cyclization be-
tween C3- and C4' in the deoxyribose residue, analogous to the mechanism
involved in ribonucleic acid hydrolysis, ^^ is less likely; if it occurred, one
would expect nucleosides or nucleotidic materials deficient in phosphorus
to be produced. If the next residue (R) ni the formula (XXXV) represents
a pyrimidine nucleoside residue, a product larger than a mononucleotide
will be first formed, but in the event of R being another deoxyribose residue
(i.e., the original site of a purine nucleoside residue) production of a pyrimi-
dine nucleoside diphosphate might be expected. It is clear that further de-
tailed investigation of the products of acid hydrolysis is desirable.

Tamm et a/.'^^ have found that apurinic acids are rapidly degraded in
alkaline solution to give a mixture of dialyzable and nondialyzable prod-
ucts. The nondialyzable fraction contains ca. 85 % of the pyrimidine nucleo-
tide residues but only some 40% of the deoxyribose phosphate residues
present in the starting material. They interpret this degradation by invok-
ing the cyclization mechanism mentioned above in discussing the break-
down of XXXV, and from their findings they have deduced that calf thy-
mus deoxyribonucleic acid consists of a chain in which sections of the chain
containing principally pyrimidine nucleotides are followed by stretches in
which purine nucleotides predominate. There appears to be some discrep-
ancy between this conclusion and the isolation of large amounts of pyrimi-
dine nucleoside diphosphates from acid hydrolysates if these latter sub-
stances arise by the mechanism discussed above. The appearance of the
diphosphates suggests a fairly even or random distribution of purine and
pyrimidine nucleotide residues. A likelier mechanism for the alkaline degra-

1^3 C. Tamm, H. S. Shapiro, R. Lipshitz, and E. Chargaff, J. Biol. Chem. 203, 673
(1953).

CHEMICAL BONDS IN NUCLEIC ACIDS 445

dation of apurinic acid is that discussed above for structure XXXV, in
which fission at the broken Hne a o by an ehmination reaction or by hydroly-
sis is a major factor;- this mechanism would lead to products in which the
pyrimidine nucleoside residues retain phosphorus in excess of the mono-
nucleotide ratio (1:1), e.g., nucleoside diphosphates, dinucleoside triphos-
phates, etc. The low rate of dialysis of the pyrimidine-containing fragments
of Tamm et al}^^ against distilled water might be explained by the associa-
tion of low molecular weight with high ionic charge."* At present it is not
possible to resolve these matters, and further evidence bearing on nucleotide
sequence is required.

CHAPTER 13

The Physical Properties of Nucleic Acids
D. O. JORDAN

Page
I. Pyrimidines, Purines, Nucleosides, and Nucleotides 447

1. Structure Determination by X-ray Diffraction 447

a. Pyrimidines 447

b. Purines 451

c. Nucleosides 452

2. Dissociation Constants 455

a. Pyrimidines and Purines 455

b. Nucleosides and Nucleotides 458

II. Nucleic Acids 461

1. Structure Determinations by X-ray Diffraction 461

2. Determination of JVIolecular Weight 470

a. Deoxypentose Nucleic Acids 470

b. Pentose Nucleic Acids 474

3. Acid-Base Properties 475

a. Deoxypentose Nucleic Acids 475

b. Pentose Nucleic Acids 480

4. The Solution Properties of Sodium Deoxypentose Nucleate 483

a. The Influence of Ionic Strength in Neutral Solution 483

b. Conclusions as to the Size and Shape of the Deoxypentose Nucleate Ion

in Neutral Solution 488

c. The Influence of Changes of pH on the Size and Shape of the Deoxy-
pentose Nucleate Ion in Solution 490

I. Pyrimidines, Purines, Nucleosides, and Nucleotides
1. Structure Determinations by X-Ray Diffraction
a. Pyrimidines

The full details of the structures of uracil (I), thymine (5-methyluracil,
II), cytosine (III), and 5-methylcytosine (IV), the pyrimidines known to
occur in nucleic acids [cf. Bendich, Chapter 3], have not been established
by X-ray diffraction. Some conclusions may be drawn concerning their
structures, however, from those of some related pyrimidine derivatives

447

448 D. O. JORDAN

which have been studied. Clews and Cochran' -^ have determined the struc-
O O NH2 NH2

C C CH3 C C CHs

/\ / \ / ^ \ ^ \ /

HN CH HN C N CH N C

I II I II I II I II

C CH C CH C CH C CH

^ \ / ^ \ / /\/ /\/

ON ON ON ON

H H H H

I II III IV

ture of 2-amino-4-methyl-6-chloropyrimidine, 2-amino-4 , 6-dichloropyrimi-
dine, 4-amino-2,6-dichloropyrimidine, and 5-bromo-4,6-diaminopyrimi-
dine and their results are shown in Fig. 1. Clews and Cochran conclude
that the pyrimidine ring is planar, in agreement with the result of Schneider^
obtained from dipole moment measurements. The C^ — N and C — C bond
distances in the pyrimidine ring of these derivatives correspond to approxi-
mately 50 % double bond character, as would be expected for a six-mem-
bered ring in which resonance of the benzene type is possible. Whether
resonance of the benzene type can occur in the amino- and hydroxypyrimi-
dines which are found in nucleic acids is dependent on whether these groups
exist in the amino or imino and the enol or keto forms, respectively.

An amino group in position 6(4) of the pyrimidine nucleus may show
greatly reduced amino behavior owing to the possibility of a tautomeric
change converting the 6-aminopyrimidine (V) into the iminodihydropyrimi-
dine form (VI). The X-ray evidence shows that 4-amino-2,6-dichloro-
pyrimidine and 5-bromo-4 , 6-diaminopyrimidine^ are in the amino form

NH2 NH

I II

C C

/ \ / \

N CH HN CH

II I I II

HC CH HC CH

\ Z' \ /

N N

V VI

in the crystal and it is therefore probable that the structures of cytosine
and of 5-methylcytosine are analogous.

' C. J. B. Clews and W. Cochran, Acta Cryst. 1, 4 (1948).

2 C. J. B. Clews and W. Cochran, Ada Cryst. 2, 46 (1949).

3 W. C. Schneider, J. Am. Chem. Soc. 70, 627 (1948).

PHYSICAL PROPERTIES OF NUCLEIC ACIDS

449

CI

NH.

CH3 NHj

Fig. 1. Structure of (a) 2-amino-4-methyl-6-chloropyrimidine, (b) 2-amino-4,6-
dichloropyrimidine, (c) 4-amino-2,6-dichloropyrimidine, and (d) 5-bromo-4,6-di-
aminopjTimidine (Clews and Cochran''^).

The position of the hydrogen atom in the — NH^ — CO — • groups, i.e.,
whether the group exists in the keto or enol form, may be determined from
the C — 0 bond length in 2-hydroxy-4,6-dimethylpyrimidine obtained by
Pitt,^ the structure of which is shown in Fig. 2. The length of the C^^^OH
bond is 1.25 ± 0.04 A. and is clearly different from the phenolic hydroxyl
group in resorcinol (C — ^OH distance 1.36-1.37 A.) and is to be compared
to that in oxalic acid dihydrate (1.24-1.30 A.). The bond between the
carbon and oxygen atoms thus possesses considerable double bond charac-
ter, but the hydrogen atom is nevertheless covalently bound to the oxygen
and takes part in a hydrogen bond between the oxygen atom and a water
molecule. In aqueous solution, the position of the hydrogen atom is not
necessarily the same as in the crystal, and evidence from the ultraviolet
absorption spectra of various pyrimidines shows that the keto form is
predominant in solution. At room temperatures, in neutral aqueous solution,

^ G. J. Pitt, Ada Cryst. 1, 168 (1948).

450

D. O. JORDAN

OH

Fig. 2. Structure of 2-hydroxy-4,6-dimethylpyrimidine (Pitt^).

the characteristic absorption spectrum of pyrimidines, purines, and of
nucleic acids is a broad band around 2400-2800 A. [Cf , Beaven, Holiday and
Johnson, Chapter 14.]. On increasing the pH of the sokition, a shift in Xmax.
to longer wavelengths is observed which has been interpreted by Loof-
bourow et al.^ and by Stimson and Reuter^-^ as an enolization of the pyrimi-
done. However, as Marshall and Walker^" point out, this shift is more prob-
ably due to ionization than to enolization. The problem has been consider-
ably clarified by the work of Marshall and Walker,'" who have examined
the ultraviolet absorption spectra of a number of 2- and 4 (6) -substituted
pyrimidines and the corresponding Ni- and Na-methylated derivatives in
aqueous solution at various pH values. The ^K'a value of each pyrimidine
was determined and the pH of the solution to be measured chosen so as to
be pK'o ± 2. In this way only neutral molecules or ions were present in the
solution and the spectra were not confused by being those of mixtures of
ions and neutral molecules. Marshall and Walker conclude that the shift in
Xmax. is due to ionization and that potential 2- and 4(6)-hydroxypyrimidines
should be represented in the keto form, i.e., as pyrimidones. A similar con-

« J . R. Loofbourow, M. M. Stimson, and M. J. Hart, J. Am. Chem. Soc. 65, 148 (1943) .
« M. M. Stimson and M. A. Reuter, J. Am. Chem. Soc. 65, 151 (1943).
^ M. M. Stimson and M. A. Reuter, /. Am. Chem. Soc. 67, 847 (1945).

8 M. M. Stimson and M. A. Reuter, J. Am. Chem. Soc. 67, 2191 (1945).

9 M. M. Stimson, /. Ajn. Chem. Soc. 71, 1470 (1949).

'« J. R. Marshall and J. Walker, J. Chem. Soc. 1951, 1004.

PHYSICAL PROPERTIES OF NUCLEIC ACIDS

451

NH

a b

Fig. 3. Structure of (a) adenine, (b) guanine (Broomhead''''^).

elusion is reached for uracil, which should therefore be represented as
2 , 4(6)-pyrimidinedione.

b. Purines

X-ray studies of adenine hydrochloride were first made by Bernal and
Crowfoot," but no attempt was made to determine the positions of the
atoms in the unit cell. These have, however, been determined by Broom-
head'^"^^ for both adenine (VII) and guanine (VIII) hydrochlorides. More
precise data on adenine hydrochloride have been obtained by Cochran."
The bond distances and interbond angles are given in Fig. 3. The structures
of adenine and guanine are very similar, the main differences being that

NH,

O

C N

C N

/-^x /^

/ \ / '

Ni 6C X

HN C

1 11 ^CH

1 II

HC2 4C /

c c

\3/ \./

/ \ / \

N N

H2N N N

H

H

VII

VIII

\

X

CH

the amino nitrogen atom (Nio) of guanine appears to be displaced by 0.11 A.
from the plane containing the other atoms and that there is a difference of

" J. D. Bernal and D. Crowfoot, Nature 131, 911 (1934).
12 J. M. Broomhead, Acta Cryst. 1, 324 (1948).
»3 J. M. Broomhead, Acta Cryst. 4, 92 (1951).
" W. Cochran, Acta Cryst. 4, 81 (1951).

452 D. O. JORDAN

0.10 A. in the C4— Cs bond distance, that in adenine being the longer. This
difference may be due to experimental error," but nevertheless is in agree-
ment with the different acid-base properties of adenine and guanine (see
p. 457). The Ce — 0 bond in guanine, which has a length of 1.20 A. appears
to have predominantly double bond character, thus confirming the con-
ventional keto formula (VIII) ascribed to guanine. Broomhead" points out,
however, that an error of only —0.05 A. in this value (i.e., giving a bond
length of 1.25 A.) would make the bond of comparable length to that of
the C — 0 bond in glycine and diketopiperazine, where the bond possesses
only 50% double bond character.

The position of the hydrogen atoms has been considered by both Broom-
head" and Cochran,^* who conclude that in the adenine cation, hydrogen
atoms are at Ni , C2 , Cs , Ng , and two hydrogen atoms are at the amino
nitrogen Nio . The positive charge maybe located at Ni , N9 , or Nio and the
short length of the Ce — ^Nio bond (1.34 A. compared with the single C — N
bond length of 1.47 A.) may be attributed to the contribution of resonance
forms in which the Nio bears a positive charge and in which the Ce — Nio
bond is double. The location of the hydrogen atoms in the guanine cation
is not known so precisely. Four hydrogen atoms are located at Ni , Ng , Nio ,
and Cs , and hydrogen bonds exist between N3 and Nio and between O and
N7 . There are thus four possible tautomeric forms, and a decision as to
which is the correct structure must await the results of a more precise
study. In aqueous solution, it will be difficult to distinguish between the
four "mesohydric tautomers" and indeed such distinction maybe meaning-
less.

c. Nucleosides

The structure of only one nucleoside, viz., cytidine (IX), has been de-
termined in detail although some preliminary studies have been made on
the other ribonucleosides. The deoxypentose nucleotides have yet to be ex-
amined. The structure of cytidine has been determined by Furberg^*''^
and in addition to confirming the furanose structure of the D-ribose, the
point of attachment of the sugar radical as being atNs , and the /3-configura-
tion of the Na-glycosidic link, other information of fundamental impor-
tance to the structure of nucleic acids has emerged. The molecular projec-
tion is given in Fig. 4. The six atoms of the pyrimidine ring lie in the same
plane in agreement with the observations of Clews and Cochran''- and of
Pitf on some substituted pyrimidines. The C2 — 0 distance (see IX) is
1.25 A., identical with the value found by Pitf* for the corresponding bond
in 2-hydroxy-4,6-dimethylpyrimidine, and the bond therefore possesses
some double bond character. The bond from the ring to the amino group,

16 S. Furberg, Nature 164, 22 (1949).
'« S. Furberg, Ada Cryst. 3, 325 (1950) .

PHYSICAL PROPERTIES OF NUCLEIC ACIDS 453

NHo O

c

/ \

N CH

I II
C CH

/ \ /

0 N

C

/ \
HN CH

1 II
C CH

/- \ /

0 N

HC— CHOH

/ \
0 CHOH

\ /
CH

1

HC— CHOH

/ \
0 CHOH

\ /
CH

1
CH2OH

CH2OH

IX

X

NH2

0

II

C N
HN C X

C N

/ \ / V

H C X

1 II CH

1 II CH

HC C /

\ / \ /

N N
1

c c /

/ \ / \ /

H2N N N

1

1
HC— CHOH

/ \
0 CHOH

1
HC— CHOH

/ \
0 CHOH

CH CH

I I

CH2OH CH2OH

XI XII

Ce — N, is short (1.31 A.), suggesting that it participates in the resonance
of the pyrimidine ring. Four of the atoms of the D-ribose ring lie nearly in
one plane, viz., Ci , Oi , C2 , and C4 , but the fifth atom of the ring, C3 , is out
of the plane l)y nearly 0.5 A. The central bond, N3 — Ci , joining the pyrimi-
dine ring to the sugar radical has a length of 1.47 A. and is clearly a single
bond. This bond lies in the plane of the pyrimidine ring and forms angles
of 109° and 115° with the adjacent ring bonds in D-ribose. Contrary to the
previous assumption that the two rings were parallel [see Astbury"], it is
evident that they are approximately perpendicular. The bearing of this

" W. T. Astbury, Symposia Soc. Exptl. Biol. 1, 66 (1947).

454

D. O. JORDAN

©

Q

Fig. 4. Dimensions of cytidine (Furberg'^^*)

important result on the structure of nucleic acid will be discussed below (see
p. 462).

Preliminary X-ray investigations have also been carried out by Furberg
on uridine (X), adenosine (XI), and guanosine (XII). ^^ From the resem-
blance of the cell dimensions of uridine to those of cytidine, it is concluded
that the stereochemistry of the two molecules are very similar. The data
on adenosine confirm that the glycosidic linkage is at N9 and is of the jS-
type. The direction of the N9 — Ci bond connecting the two rings was con-
sidered by Hendricks^^ and Astbury" to form an angle with the plane of
the purine ring so as to make the purine and D-ribose rings parallel. As

18 S. Furberg, Acta Chem. Scand. 4, 751 (1950).

19 S. B. Hendricks, /. Phys. Chem. 45, 65 (1941).

PHYSICAL PROPERTIES OF NUCLEIC ACIDS 455

Furberg has shown in cytidine, the corresponding bond Hes in, or very
nearly in, the plane of the pyrimidine ring, and the considerable resonance
in the adenine molecule would suggest that this is true for adenosine. If so,
the two rings will be nearly perpendicular to each other, as in cytidine.
Although the existing data do not definitely exclude other configurations,
this structure appears to give the most satisfactory explanation of the
available information. The data of Hendricks'^ can be reinterpreted on the
basis of this model. Gulland et al.-^ have pointed out that, in the structure
proposed by Hendricks, the sugar was a-lyxose and not D-ribose; and, in
the structure proposed by Furberg, two parallel planes approximately 1.5
A. apart can be recognized, one containing the purine ring and Ci , C4 , and
Cs , the other passing through C2 , C3 , and O2 .^^ Furberg^* considers that
guanosine has a structure similar to adenosine.

2, Dissociation Constants
a. Pyrimidines and Purines

The ^K'a values of some pyrimidines and purines are given in Table I.
The acid-base properties of uracil were first investigated by Levene et
aL,2i'22 who concluded from electrometric titration data that it possessed
two acid dissociations with p/C'a values 9.28 and 13.56, which were attrib-
uted to the two — CO — -XH — groups. Later the same authors-^ showed thai
an error had been made in applying the water correction for the titration in
strongly alkaline solution and a recalculation of the data showed the pres-
ence of only a single dissociation of p/^'a 9.45. This conclusion has been
confirmed by Taylor.^^ The reason for the extreme weakness of the second
dissociation constant of uracil is still obscure. Levene et alP considered that
in the neutral molecule the two — CO — NH — groups were of comparable
strength, a view that is supported by the ^K'a values of 1- and 3-methyl-
uracil which are similar to that of uracil (Table I), and that the ionization
of the second group was inhibited by the ionization of the first to such an
extent that its pK'a was too high to be detected by titration in aqueous
solution. However, calculation of the difference between pA'^^ and p/v'ao ,"
using the formulas of Kirkwood and Westheimer-^-^ and assuming that

" J. M. Gulland, G. R. Barker, and D. O. Jordan, Nature 151, 109 (1943).

^^ P. A. Levene and H. S. Simms, J. Biol. Chem. 65, 519 (1925).

« P. A. Levene, H. S. Simms, and L. W. Bass, /. Biol. Chem. 70, 243 (1925).

" P. A. Levene, L. W. Bass, and H. S. Simms, /. Biol. Chem. 70, 229 (1926).

" H. F. W. Taylor, Acid Base Properties of Nucleic Acids, Doctoral Thesis, London

Univ., London, England, 1946.
^* D. O. Jordan, Progr. Biophys. and Biophys. Chem. 2, 51 (1951).
" J. G. Kirkwood and F. H. Westheimer, J. Chem. Phys. 6, 506 (1938).
" F. H. Westheimer and J. G. Kirkwood, J. Chem. Phys. 6, 513 (1938).

456 D. O. JORDAN

the dimensions of the uracil molecule are similar to those of 2-hydroxy-4,6-
dimethylpyrimidine/ gives the low value of 1.57. It would therefore appear
that the weakness of the second dissociation cannot be due to the electro-
static field effect and must be caused by a real difference in the groups.
This conclusion is not entirely unexpected, since NiH lies between two
carbonyl groups, whereas NjH is attached to a carbonyl group and a CH
group. This view is also supported by the extreme weakness of the second
dissociation of cytosine, pK'a^ 12.1, as compared with that of isocytosine
(XIII), pK'a^ 9.42, which suggests that the 2-carbonyl group in uracil pro-
motes stronger basic properties in the — NH — CO — grouping than does the
6-carbonyl group. This conclusion is sinilar to that reached by Ogston^^ in
the somewhat analogous case of the xanthines.

The introduction of the methyl group in the 5-position of uracil in
thymine has little effect on the pK'a value. The basic properties of cytosine
and isocytosine can be ascribed to the amino group, the pK'a values being

O

II
C

/ \

N CH

II II

C CH

/ \ /

H2N N

H

XIII

comparable with those of the aromatic amines; this view has been con-
firmed by titration in the presence of formaldehyde.^"*

The dissociation constants of adenine and guanine are given in Table I.
There is considerable evidence that the weakest dissociation of adenine and

guanine represents the dissociation — NH — to — N — in the imidazole
ring. Benzimidazole, which is analogous to purine, has been found by

Taylor^^ to have a basic association of pK'a^ 5.30 and an acidic dissociation

+

of pK'a2 12.3. These values must represent the dissociation of the — NH=
and — NH— groups, respectively. The value of pK'ai 5.30 may be com-
pared with that of 7.1 given by Dedichen^" for imidazole, whereas the
value of 12.3 is of the same order as those of the weakest dissociations of
guanine, 1- and 3-methylxanthine, and of hypoxan thine, all of which are
unsubstituted in the imidazole ring. As would be expected, 7- and 9-methyl-

28 A. G. Ogston, /. Chem. Soc. 1935, 1376.
" H. F. W. Taylor, J. Chem. Soc. 1948, 765.
»» G. Dedichen, Ber. 39, 1831 (1906).

PHYSICAL PROPERTIES OF NUCLEIC ACIDS

457

TABLE I
pK'o Values of Pyrimidines and Purines

P^'a.

pK'a

pK'a

Reference

Uracil

9.45

l-Methyluracil

9.99

3-Methyluracil

9.71

Thymine

9.82

Cytosine

4.60

12.16

Isocytosine

4.01

9.42

Adenine

4.15

9.80

Guanine

3.3

9.20

Xanthine

7.7

1-Methylxanthine

7.7

12.05

3-Methylxanthine

8.5,"

8.10''

11.3

7-Methylxanthine

S.S,"

8.30'-

9-Methylxanthine

6.3,<'

6.25*

Hypoxanthine

8.8

12.0

" Results of Ogston.2s

* Results of Taylor .29

12.3

21-24

23

23

23

23,24

23

29

29

28

28,29

28,29

28,29

28,29

24

xanthine show no dissociation in the pH range 11.0-12.5. In adenine, the
pK'a2 value, 9.80, is surprisingly low when compared with the correspond-
ing dissociation in guanine and has been explained by Taylor'^ on the basis
of the crystallographic data of Broomhead.'-'" Taylor explains the unusual
acid strength of this group in adenine as being due to the considerable con-
tribution to the resonance of the structure, not only of the uncharged
structures, but also of those in which there is a negative charge on the
nitrogen atoms in the pyrimidine ring and a positive charge on those in
the imidazole ring. This distribution of charge would lower the p/C'aj value

from that characteristic of an uncharged — NH — group towards one

+
characteristic of a charged — NH= group. The respective pK'a values for
these groups in benzimidazole are 12.3 and 5.3 and the observed value of
9.8 for adenine may thus be explained. In guanine, the dissociation of the
imidazole — NH — group takes place in a molecule which already bears a
negative charge in the region of the pyrimidine nitrogen atoms as a result
of the dissociation of the — NH — CO — group. It is therefore to be expected
that the resonance structures bearing a negative charge on the negative
atoms of the pyrimidine ring will make a smaller contribution in the gua-
nine than in the adenine anion, and, since resonance structures of this type
necessarily contain a single bond in the C4 — C5 position, Taylor^^ concludes
that the greater length of this bond in adenine is explained. This argument
has been criticized by Cochran'"* on the grounds that the resonance struc-

81 H. F. W. Taylor, Nature 164, 750 (1949)

458 D. O. JORDAN

tures of the type considered by Taylor consistently make the bonds N9 — Cs
and N: — Cs single bonds and the bond N7 — Cs a double bond, whereas the
X-ray data indicate that these bonds in adenine all possess 20-50 % double
bond character. However, the correlation of bond lengths obtained for the
crystal with other properties obtained from measurements made on solu-
tions must be carried out with care as the different environments may
produce small changes in the molecule. In the present instance, the influence
of the chloride ion, which exists close to the adenine ion in the crystal lat-
tice, may produce significant effects.

The group responsible for the acid dissociations of pK'a 6-9 which appear in the
oxypurines is not known with certainty. Existing evidence indicates clearly that it
is to be associated with the presence of the 6-oxy group in guanine, hypoxanthine,
and xanthine; the second oxy group in the 2-position of xanthine does not give rise to
a further dissociation, the case being somewhat analogous to that of uracil. Ogston^*
considers that the acidic properties of xanthine do not depend on the — NH — CO —
group in the 1- and 2-positions since substitution by a methyl group in the 1 -position
does not greatly alter the pi^'oi value of xanthine, nor that of the 3-, 7-, and 9-methyl-
xanthines. The actual form of the dissociating groups has been studied by Ogston^*
by electrometric titration in water and 90% ethyl alcohol. In xanthine, the group is
considered to be in the enolic form and the structure is — Ni^CeOH. In the 3-, 7-,
and 9-methylxanthines, however, the zwitterionic form is considered to predominate,

+
the acid dissociation being represented as that of an — NH^group in the imidazole
ring. If this interpretation is correct, the second dissociation in the imidazole ring
must be preceded, according to Taylor,*' by a tautomeric change, since Ogston''^
has shown that the — NH — group in the 1 -position does not show acid properties in
these molecules.

h. Nucleosides and Nucleotides

The pK'a values of the ribonucleosides and ribonucleotides are given in
Table II. The values for the nucleosides correspond closely to those of the
parent purines and pyrimidines although there is a general tendency for the
pK'a values to be lowered, i.e., for the acid dissociations to be strengthened.
Except in the case of xanthosine, which has not been studied to a suffi-
ciently high pH value, an additional dissociation is observed at pK'a 12.3-
12.6, which is of the correct order for the first acid dissociation of a sugar.
No data are available for D-ribose, but glucose has a pK'a value of 12. 1.^--^^

The pK'a values of the ribonucleotides obtained from yeast nucleic acid
are given in Table II. These values will all refer to mixtures of the a and b
nucleotides, i.e., of the nucleoside-2'- and -3'-phosphates, in unknown
ratios. This isomerism, however, will have only a small effect on the dis-
sociation constants (see below). Comparison of the values for the nucleo-

32 P. Hirsch and R. Schlags, Z. physik. Chem. 141, 387 (1929).

33 F. Urban and P. A. Shaffer, J. Biol. Chem. 94, 697 (1931).

3* F. Urban and R. D. Williams, J. Biol. Chem. 100, 237 (1933).

PHYSICAL PROPERTIES OF NUCLEIC ACIDS 459

TABLE II

pK'o Values of Nucleosides and Nucleotides

piv-'o,

VK'.,

pK'.,

pK'a,

Reference

Adenosine

3.5

12.5

21,22

Guanosine

1.6

9.2

12.3

21,22

Cytidine

4.2

12.3

21,22

Uridine

9.2

12.5

21,22,35

Inosine

8.8

12.3

21,22

Xanthosine

6.0

28

Adenylic acid

0.9

3.7

6.0

21,22,35

Guanylic acid

0.7

2.4

6.0

9.3

21,22,35

Cytidylic acid

0.8

4.2

6.0

21,22,35

Uridylic acid

1.0

5.9

9.4

21

tides with those of the respective nucleosides, shows that the dissociating
groups of the latter are supplemented by two additional dissociations hav-
ing pK'a values 0.7-1.0 and 5.9-6.0. These are of the correct order for the
first and second dissociations of a sugar phosphate; ribose phosphates have
not been studied, the analogous glucose-3-phosphate has pK'a values of
0.84 and 5.67,^® and similar values have been obtained for other sugar
phosphates.^^

The assignment of the pK'a values to the various groups is best carried
out by reference to the data for the corresponding nucleosides and sugar
phosphates. The following are the relevant data for adenylic acid:

Adenosine pK'a 3.5

Sugar phosphates p^'oi 0.8-1.1 pK'a2 6.0-6.5

Adenylic acid pK'ai 0.9 pK'a2 3.7 pK'a^ 6.0

The pK'ai value of adenylic acid is clearly that of a primary phosphoric
acid dissociation, pK'a2 an amino dissociation, and pK'as a secondary phos-
phoric acid dissociation. In the isoelectric region, therefore, it is evident
that adenylic acid, and hence also guanylic and cytidylic acids, will exist
largely in the zwitterionic form. It is therefore to be expected that the
nucleic acids will behave similarly.

The values of K^ for the nucleotides, where

[H3NROPO(OH)o]

K, =

[H.NR0P0(0H)2]

3^ W. E. Fletcher, On the Structure of Nucleic Acids, Doctoral thesis, London Univ.,

London, England, 1948.
" O. Meyerhof and K. Lohmann, Biochem. Z. 185, 113 (1927).

1.6

7

3.5

3.6 X 102

4.2

2.6 X 103

460 D. O. JORDAN

TABLE III
Values of Kj for the Ampholytic Nucleotides

(nucleotide) (nucleoside) Ki

Guanylic acid 0.7

Adenylic acid 0.9

Cytidylic acid 0.8

have been calculated^^ assuming (1) that only the primary phosphoric acid group
and the amino group are important in determining the position of the isoelectric
region and the value of Ki , (2) that the constant for the dissociation of the nucleotide

H3NROPO(OH)2 + H2O ^ H2NROPO(OH)2 + H3O+
is identical with that for the nucleoside

H3NR + H2O ^ H2NR + H3O+
Then^s

A'l (nucleotide)

K.t = ~ ' ~ r — 1
Ai(nucleoside)

This assumption is analogous to that generally made for calculating K^ for the amino
acids, when it is assumed that the dissociation of the amino group in the positively
charged acid is identical with that of the corresponding ester.'* It is justified here
by the fact that the undissociated phosphoric acid group is unlikely to have any
inductive effect on the amino group, from which it is separated by six atoms, and
the furanose ring is a nonresonating system. The values of K^ so calculated are given
in Table III. In guanylic acid, because of the low value of the pK'a of the amino group
in guanosine, the concentration of the zwitterionic form is only seven times that of
the uncharged form. In adenylic and cytidylic acids, however, the ratio is very much
greater and the concentration of the uncharged form in the isoelectric region is
negligible.

It is evident from Table II that the introduction of the negatively charged
phosphate ion into the nucleoside produces a weakening of the acid strength
of the amino and the — NH — CO — dissociations. This observation is in
agreement with the field effect, first recognized by Bjerrum^^ and treated
c^uantitatively by Kirkwood and Westheimer.-^" The degree of the shift
of the pK'a will be a function of the distance l)etween the charged groups,
the smaller this distance the greater the p/v'a shift. This relationship has
been used by Cavalieri'"' in an attempt to establish the configuration of the

37 W. D. Kumler and J. J. Eiler, J. Am. Chem. Soc. 65, 2355 (1943).

38 E. J. Cohn and J. T. Edsall, "Proteins, Amino Acids and Peptides." Chemical
Catalog Co., New York, 1943.

39 N. Bjerrum, Z. phtjsik. Chem. 106, 219 (1923).

'0 L. F. Cavalieri, /. Am. Chem.. Soc. 74, 5804 (1952).

PHYSICAL PROPERTIES OF NUCLEIC ACIDS 461

a and b isomers of cy tidy lie acid. From electrometric titration (glass elec-
trode) the following pK'a values were obtained:

a pK'aj 4.36 (4.3) pK'aj 6.17 (6.2)

6 pK'a2 4.28 (4.3) pK'a^ 6.0 (6.0)

The values in parenthesis were those obtained by Loring et al.,^^ and Cava-
lieri considers that since the amino and phosphate groups will be closer in
the 2'- than in the 3'-phosphate the a isomer is probably the 2'-phosphate.
A similar conclusion was reached by Loring et al.^"^ on the grounds that the
isomer with the smaller distance between the charged groups would show
the greater tendency towards zwitterion formation and hence have the
lower solubility, acidity, and ultraviolet absorption.

II. Nucleic Acids
1. Structure Determinations by X-Ray Diffraction

Although the general chemical picture of the nucleotides and of the nu-
cleic acids is one of some complexity, the basic macromolecular configura-
tion which emerges from X-ray diffraction studies is one of comparative
simplicity. The studies, in the main, have been concerned with the sodium
salt of deoxypentose nucleic acid since this substance can be comparatively
readily isolated in a state of purity and of high molecular weight. Ribo-
nucleic acids have only been examined when a high-molecular-weight
sample, such as that obtained from tobacco mosaic virus, has been available.

The main features of the molecular structure of deoxypentose nucleic
acid have been known since 1938, but it is only recently that the accumula-
tion of knowledge has permitted the suggestion of a structure concerning
which there appears to be a large measure of agreement. Astbury and
BelP^ showed first that the X-ray fiber photograph gave a prominent re-
flection which corresponded to a spacing of 3.34 A. along the fiber axis.
This was interpreted as due to a succession of fiat nucleotides standing out
perpendicularly to the fiber axis to form a relatively rigid structure. This
rather small distance between the nucleotides is in agreement with the
density of the dried nucleic acid, Avhich is 1.62-1.63 g./cc. In addition to
this spacing of the nucleotides, Astbury and Bell and later Astbury^^ con-
cluded that the structure pattern along the axis of the molecule repeated
at a distance of about 27 A. Since the Astbury structure assumed a pile of
nucleotides, one on top of the other, this distance corresponds to approxi-

■" H. S. Loring, H. W. Bortner, L. W. Levy, and M. L. Hammell, /. Biol. Chem. 196,

807 (1952).
« H. S. Loring, M. L. Hammell, L. W. Levy, and H. W. Bortner, J. Biol. Chem. 196,

821 (1952).
« W. T. Astbury and F. O. Bell, Nature 141, 747 (1938).

462 D. O. JORDAN

mately eight nucleotides. In view of the more recent developments it should
be pointed out that Astbury^^ did not completely reject the possibility that
the nucleotides might be disposed spiralwise around the long axis of the
molecule, but he concluded that if the nucleotides did not lie closely on top
of one another, then the state of packing must be dimensionally equivalent
to such an arrangement so that the neighboring nucleotide molecules were
closely interleaved in a surprisingly regular pattern. This, Astbury con-
sidered, was unlikely.

The Astbury structure rested, in part, on the assumption that the nucleo-
tides were flat or approximately flat molecules, i.e., the purine or pyrimi-
dine ring systems were in the same plane as that of the sugar. This was
shown to be wrong by Furberg,'*-'^ who showed that in the nucleoside
cytidine the two ring systems instead of being parallel were almost per-
pendicular. This observation is of fundamental importance in the de-
velopment of the structure of nucleic acids and has been confirmed for
2' ,3'-isopropylidene-3 ,5'-cycloadenosine iodide by Clark et aZ.** and by Zuss-
man.'*'*^ As has been pointed out above, the bond between sugar and purine
or pyrimidine (Ng or 3 — Ci) is a single bond and so rotation about this
bond is possible. However, Furberg** finds that not all orientations of the
sugar and purine or pyrimidine are equally feasible and the most favorable
position is considered that shown in Fig. 5 which is that found in the crystal
structure of cytidine.'*''^ In view of these observations on the structure of
the nucleosides and nucleotides, Furberg'*^ revised the Astbury structures
and proposed two possible models which are given in Fig. 6. In contrast to
the Astbury model, these two modifications have the planes of the sugar
rings, as well as the P — O3 bonds (Fig. 5), approximately parallel to the
long axis of the molecule. Most atoms, including the phosphorus atoms,
lie in planes 3.4 A. apart, thus explaining the strong 3.4-A. reflection.

In model 1 (Fig. 6) the pyrimidine and purine rings are piled almost on top of each
other; they cannot be piled directly on top of each other without bringing some of
the atoms in successive nucleotides too near together. Van der Waals attraction be-
tween the rings will stabilize the structure. The ribose rings and the phosphate groups
form a spiral enclosing the column, the spiral repeating itself after eight nucleotides
as required by the Astbury model. In model 2, the ribose rings and phosphate groups
form a central column from which the purines and pyrimidines stand out perpendicu-
larly. This model has the disadvantage that there are no intramolecular Van der
Waals forces between the purines and pyrimidines, but the rigidity of the molecule
will depend on such forces between the sugar molecules.

Following the successful formulation of the structure of some of the pro-

** V. M. Clark, A. R. Todd, and J. Zussman, J. Chern. Soc. 1951, 2952.

^^^ J. Zussman, Acta Cryst. 6, 504 (1953).

" S. Furberg, Acta Chem. Scand. 6, 634 (1952).

PHYSICAL PROPERTIES OF NUCLEIC ACIDS

463

D C

3.4 A

Fig. 5. Structure of a pyrimidine nucleoside (Furberg"**).

teins as helical polypeptide chains, Pauling and Corey'**'''' suggested a
structure for nucleic acid which they considered as compatible with the
main features of the X-ray diagram. The structure involved three inter-
twined helical polynucleotide chains, each chain having approximately
twenty -four nucleotides in seven turns of the helix. This structure, however,
has been severely criticized on several grounds. First, the Pauling and Corey
structure has the phosphate groups closely packed about the axis of the
molecule, with the pentose residues surrounding them and the purine and
pyrimidine groups projecting radially; the general behavior of the nucleate
ion in solution, however, would suggest the reverse configuration, i.e., with
the phosphate groups on the outside of the molecule and the purine and
pyrimidine rings on the inside. There is considerable experimental evidence
for this, the most important probably being that of Gulland et al.,^^ who
showed by titration that the phosphate groups are available for acid-base
equilibria in the normal way and that the amino and the — NH — CO —
groups on the purines and pyrimidines were inaccessible, unless the nucleic
acid was first subject to extremes of pH. Dye and protein absorption con-
firm that the phosphate groups must be accessible to large ions. Secondly,

« L. Pauling and R. B. Corey, Nature 171, 346 (1953).

« L. Pauling and R. B. Corey, Proc. Natl. Acad. Set. U. S. 39, 84 (1953).

« J. M. Gulland, D. O. Jordan, and H. F. W. Taylor, J. Chem. Soc. 1947, 1131.

464

D. O. JORDAN

Fig. 6. Two models of deoxypentose nucleic acid based on nucleotides of the
standard configuration (see Fig. 5) (Furberg").

Watson and Crick*^ point out that many of the Van der Waals distances in
the Pauhng and Corey structure are too small and also that while the X-ray
diagrams are for the sodium salt of deoxypentose nucleic acid, the Pauling
and Corey structure refers to the free acid. The great importance of the
Pauling and Corey contribution probably lies in the suggestion that the
structures of nucleic acids were of the helical type.

In view of these criticisms a new structure was proposed by Watson and
Crick^^ which has many novel features and appears to be in harmony with
a large amount of different experimental data, and it would appear, apart
from alterations of detail, that this structure is correct for the deoxyribo-
nucleate ion as obtained from calf thymus. This structure has two helical
chains each coiled round the same axis (see Fig. 7). Both chains follow
right-handed helices, but the sequences of the atoms in the two chains run
in opposite directions. Each chain resembles the Furberg^^ formula model
1 (Fig. 6) and the configuration of the nucleotides are approximate to that
given in Fig. 5. The pyrimidines and purines are thus on the inside of the
helix and the phosphates on the outside, and a nucleotide occurs in the
direction of the long axis every 3.4 A. The structure repeats everj'- ten
nucleotides or 34 A., the angle between adjacent nucleotides in the same
chain being assumed to be 36°.

The novel feature of the structure is the manner in which the two chains
are held together by the purine and pyrimidine bases. The planes of these
bases are perpendicular to the fiber axis and are joined together in pairs by

^' J. D. Watson and F. H. C. Crick, Nature 171, 737 (1953).

PHYSICAL PROPERTIES OF NUCLEIC ACIDS

465

34 A

T
3.4 A

Fig. 7. The helical structure of sodium deoxypentose nucleate proposed by Wat-
son and Crick.""

hydrogen bonds. A single base from one chain is hydrogen-bonded to a
single base in the other chain so that the two bases lie side by side with
identical ^-coordinates. The formation of these hydrogen bonds between
different bases is found, on the basis of molecular models, to be highly
specific and only certain pairs of bases will fit into the structure. One mem-
ber of a pair must be a purine and the other a pyrimidine in order to bridge
between the two chains. A bridge of two pyrimidines is not large enough
to form the link and there is not room for two purines. If the most probable
tautomeric forms of the purines and pyrimidines are assumed it is then
found that the only pairs of bases that are possible are:

adenine with thymine

guanine with cytosine

The ways in which these can be joined are shown in Fig. 8. A given pair can

466

D. O. JORDAN

Adcnint

Th/mint

Cyxot'int

Fig. 8. The permitted hydrogen bonds in the Watson and Crick formula.

be either way around, thus adenine can occur in either chain, but when it
does, its partner on the other chain must always be thymine.

The pairing of the bases in this manner is strongly supported by the
recent analytical results of Chargaff^'^ and Wyatt.*''*^ The molar ratios of
adenine-thymine and guanine-cytosine for a variety of deoxypentose
nucleic acids are listed in Chapter 10. These ratios approximate very closely
to unity in accordance with the Watson and Crick hypothesis. The ratios
of adenine to guanine on the other hand [see Chapter 10] are generally
much greater than unity and are variable in different nucleic acids. The
formation of hydrogen bonds between the paired bases, as shown in Fig.
8, is also in agreement with the titration results of Gulland et al.,*^ who first
suggested the formation of hydrogen bonds between the amino and — ^NH —
CO — dissociations in order to explain the nonavailabihty of this group

" E. Chargaff, Experientia 6, 201 (1950); Federation Proc. 10, 654 (1951); for other

references see Chapter 10 of this book.
" G. R. Wyatt, Biochem. J. 48, 584 (1951).
" G. R. Wyatt, J. Gen. Physiol. 36, 201 (1952).

PHYSICAL PROPERTIES OF NUCLEIC ACIDS 467

Fig. 9. X-ray diagram of the crystalline form of sodium deoxypentose nucleate
(Franklin and Gosling'^).

except -when the nucleic acid had been treated in aqueous solution with
acid or alkali. The titration results further show^^ that the amino group of
guanine does not take part in the hydrogen-bonding in agreement with the
arrangement shown in Fig. 8b.

Confirmation of the general features of the Watson and Crick formula-
tion, which was apparently derived mainly from geometrical considera-
tions, and more precise details concerning the dimensions of the helix and
of the unit cell have been given by the X-ray data of Wilkins et alr'^ and
Franklin and Gosling.""^* Fibers of the sodium salt of deoxypentose nucleic
acid may be prepared" with a high degree of orientation by withdrawing a
needle point slowly from a stiff gel of the nucleate. By suitably varying the
speed of withdrawal and the water content of the gel, fibers of any diameter
between 1 and 100 m may be obtained. Such fibers give two di.stinct types

" D. J. Cosgrove, R. H. Garner, D. O. Jordan, and S. M. Matty, in press.
" M. F. H. Wilkins, A. R. Stokes, and H. R. Wilson, Nature 171, 738 (1953).
" R. E. Franklin and R. G. Gosling, Nature 171, 740 (1953).
*« R. E. Franklin and R. G. Gosling, Nature 172, 156 (1953).
F R. E. Franklin and R. G. Gosling, Ada Cri/st., 6, 673 (1953).
M R. E. Franklin and R. G. Gosling, Acta Cryst., 6, 678 (1953).

468 D. O. JORDAN

o

Fig. 10. X-ray diagram of the paracrystalline form of sodium deoxj^pentose
nucleate (Franklin and Gosling^*).

of X-ray diagram. The first (Fig. 9), which corresponds to a crystalhne
form (termed structure A by Frankhn and Goshng*^), is obtained in an at-
mosphere of about 75% relative humidity. This corresponds to a water
content of the sodium nucleate of 40-45% of the dry weight. At higher
humidities a different structure (B) showing a paracrystalline form with a
lower degree of order appears and persists over a wide range of humidity
(Fig. 10). The change from A to B is normally reversible.*'' In view of the
high water content of the form B it seems reasonable to suppose that in
this form the structural units of the sodium nucleate are relatively free
from the influence of neighboring molecules, each unit being shielded by a
sheath of water. Analysis of the X-ray diagrams of the B form of the sodium
deoxypentose nucleate from E. toli^^ and of the sodium deoxyribonucleate
of calf thymus** leads to a general confirmation of the helical structure
although the evidence is somewhat circumstantial. Furthermore it is
shown that the phosphate groups must lie on the outside of the structural
unit on a helix of diameter approximately 20 A.

One of the important features of the Watson and Crick structure is that
it consists of two coaxial helical chains related by a dyad axis, the third
coaxial chain being absent in the nucleic acid but presumably occupied by
the protein in the nucleoprotein. Direct evidence of this two-chain helix
has been obtained by Franklin and Gosling,***^ who for structure A, which

PHYSICAL PROPERTIES OF NUCLEIC ACIDS

469

gives an X-ray diagram with sixty-six independent reflections distributed
on nine well-defined layer-lines, have calculated the cylindrically averaged
Patterson function (Fig. 11). The theoretical curves of the Patterson
function for a smooth two-chain helix of radius 9.0 A. and in which the two
chains are separated by one-half the length of the c axis of the unit cell are
found to pass through a large proportion of the Patterson peaks. Using the
cylindrical Patterson function, Franklin and Gosling determined that the
unit cell is face-centered monoclinic having a = 22.0 A., b = 39.8 A., c =
28.1 A. and j8 = 96.5°. In order to calculate the number of nucleotides in
the unit cell, it is necessary to know both the density and the water con-
tent. These are interdependent and can only be measured on a polycrystal-
line mass, and some doubt must rest on the values. For a density of 1.47
g./cc. at 75% relative humidity and a water content of about 40% of the
dry weight, a value of twenty-three nucleotides per lattice point results, so
that the unit cell will contain eleven nucleotides on each of two chains.

Origin __

of curvc/ii)

Orig in

o,curve(i),. ■ '" [-- "■■[' ^ p^pT

5 lO 15 20 25

Fig. 11. Cylindrical Patterson function of crystalline sodium deoxypentose
nucleate.

470 D. O. JORDAN

The complete three-dimensional Patterson function suggests that only
a part of the structure repeats itself in the plane at c = 3^ in the unit cell.
This is what would be expected for two coaxial chains related by a dyad
axis, as suggested by Watson and Crick. The phosphate groups repeat at
c = M, but since the two chains run in opposite directions, this will not be
true of the rest of the molecule.

These important conclusions refer only to the structure of deoxypentose
nucleic acids and not of ribonucleic acids. The latter acids have yet to be
prepared in a form that will yield satisfactory X-ray diagrams and the
structural evidence based on chemical and enzymatic studies (see Chap-
ters 11 and 12) suggests that these acids have a branched-chain structure
in marked contrast to the unbranched chain of deoxypentose nucleic acids.

2. Determination of Molecular Weight

a. Deoxypentose Nucleic Acids

The apparent molecular weight (particle weight) of the deoxypentose
nucleate ion in solution has been measured by a variety of methods, the
majority leading to a value of the order of 1 X 10^. The determination of
the molecular weight of the nucleate ion in the ultracentrifuge is compli-
cated by a marked variation of the sedimentation coefficient with concen-
tration of the nucleate ion. This difficulty, which is typical for poly electro-
lytes, may be reduced by increasing the ionic strength of the solution by
the addition of neutral salt, but even so, the results shown in Fig. 12 indi-
cate a marked variation of the sedimentation coefficient in a buffer of ionic
strength 3. This dependence upon concentration necessitates the extrapola-
tion of the sedimentation results to obtain the value of the sedimentation
coefficient at infinite dilution. Tennent and Vilbrandt*^ observed a par-
ticularly rapid increase of sedimentation constant at low concentrations
which did not permit extrapolation. This observation has not been con-
firmed by Cecil and Ogston,*^ Atlas and Stern, ^^ or by Kahler,®^ whose data
shown in Fig. 12 are in close agreement. Kahler^^ ^nd Ogston^^ extrapolated
to infinite dilution by making use of the fact that the reciprocal of the
sedimentation coefficient is a linear function of the concentration of the
nucleate ion at low concentrations (Fig. 13). The values obtained were
12.5 X 10-1^ c.g.s. (Kahler«2), 13.2 X 10-" c.g.s. (Ogston^^). These values
have been confirmed by the less detailed investigations of Conway et al.^*
and Krejci et al.^^

" H. G. Tennent and C. F. Vilbrandt, J. Am. Chem. Soc. 65, 424 (1943).

•0 R. Cecil and A. G. Ogston, /. Chem. Soc. 1948, 1382.

•1 S. M. Atlas and K. G. Stern, unpublished data.

^^ H. Kahler, /. Phys. & Colloid Chem. 52, 676 (1948).

"A. G. Ogston, Trans. Faraday Soc. 46, 791 (1950).

'* B. E. Conway, L. Gilbert, and J. A. V. Butler, J. Chem. Soc, 1950, 3421.

«5 L. E. Krejci, L. Sweeny, and J. Hambleton, J. Franklin Inst. 248, 177 (1949).

PHYSICAL PROPERTIES OF NUCLEIC ACIDS

471

t?

12 —

\

lO —

*v

\

8 —

\

\

•

2 6 -

X

\

ors^

--..Jo

•

4 —

^

-«-

2 ^

1

1

1

0-4

06

O O-l 0-2 0-3

Concentration per cent

Fig. 12. The variation of sedimentation constant with concentration of sodium
deoxypentose nucleate.

3 data of Cecil and Ogston^"
O data of Atlas and Stern"
O data of Tennent and Vilbrandt*'
• data of Kahler.6?

The diffusion coefficient also varies with the concentration and although
the values obtained by extrapolation to infinite dilution by Kahler^- and
Cecil and Ogston^"-^^ are in fairly close agreement, the diffusion coefficient
was found to vary with concentration in opposite senses by these two
groups of observers. Kahler's results show a marked increase of the diffusion
coefficient with concentration, while those of Cecil and Ogston show an
equally marked decrease. Although the experimental technique was differ-
ent in the two cases — ^Cecil and Ogston used the Gouy method and Kahler
the more common Neurath cell with the Lamm scale or Longsw orth scan-
ning optical methods — it is clearly evide;it that further experimental work
is desirable to determine the true concentration dependence. The determi-
nation of the diffusion coefficient is, furthermore, made more difficult,
since owing to the concentration dependence of the diffusion coefficient, a
skewed boundary is obtained (Fig. 14) and there may also be an apparent
movement of the boundary.^* The value of the diffusion coefficient ob-

" J. A. V. Butler and D. W. F. James, Nature 167, 844 (1951).

472

D. O. JORDAN

1 — 1 — \ — r

Oi 0-2 0-3 0-4 OS 0-6

Conc«ntrotlon per cent

Fig. 13. Variation of 1/S with concentration (Kahler*^).

/^

^c

f-.v

/ ^

— f^

^

1

- \

1

r^T°^ —

Fig. 14. Diffusion of 0.9% sodium deoxypentose nucleate in buffer at pH 7.4
and 20° in normal coordinates (O). Ideal normal curve (•) (Kahler^^).

tained by Ogston*^ was between 0.55 and 0.70 X 10~^ and that found by
Kahler,^2 0.45 x 10~^ Combination of the sedimentation and diffusion re-
sults yield the values 1.3 X lO^.^o^^ 1.5 X 10«,«2 and 1.3 X 10^ ^^ for differ-
ent preparations of the sodium deoxyribonucleate of calf thymus.

In view of the difficulties attendant upon the determination of the
of the molecular weights of nucleic acids by sedimentation and diffusion
methods, other methods have been used, and probably the most accurate
information on the molecular weight and size and shape of the deoxypentose
nucleate ion in neutral solution has been obtained by Doty and Bunce^^

PHYSICAL PROPERTIES OF NUCLEIC ACIDS 473

TABLE IV
Molecular weight and Dimensions of Deoxypentose Nucleate Ion in

0.2 M NaCl

Gulland^o

Signer'2

Doty

Molecular weight

4 X 10'=

6.7 X 10«

4 X 10«

7.7 X 10"

Radius of gj^ration (A.)

1170

2200

1630

Radius for coil (A.)

2850

5400

4000

pH of solution

4.4

6.5

6.6

using the light-scattering method. Earher studies by Oster"^ and Smith and
Sheffer^^ had shown that nucleic acid solutions were suitable for such
studies, and a value of the molecular weight of 4.4 X 10^ was obtained by
Smith and Sheffer. The values obtained by Doty and Bunce^^ for three
different samples of sodium deoxypentose nucleate are given in Table IV.
The sample of Gulland et al?^ was identical with that examined by Cecil and
Ogston^" and shows that the light-scattering method yields a value approxi-
mately four times as great as that obtained by sedimentation. This dis-
crepancy may be due in part to the possibility that the scattering envelope
is not identical with the sedimenting unit, but is also very probably due to
the value of the sedimentation coefficient obtained by existing extrapola-
tion methods being too low. More recently, Reichmann et al.''^ have ob-
tained the value of 7.7 X 10^ for the molecular weight of a new sample,
prepared by the method of Signer,^'- and Katz^^ has recorded a value of 8.0
X 10^ for a similar preparation.

Lower estimates of the molecular weight have been obtained by Jungner et alJ^-''^
from dielectric dispersion measurements (1.2-8.4 X 10^). For the Gulland prepara-
tion'" the value 6.1 X 10* was obtained (compared with 1.3 X 10* by sedimentation
and difTusion*"'" and 4 X 10* by light-scattering*' methods). The dielectric molecu-
lar weight represents the weight of individual units orienting independently in an
electric field. Whether these units represent a single nucleic acid molecule or sections
of a larger molecule is difficult to decide at present. Jungner et al. assume the former
possibility to be correct and consider that the sedimentation and light-scattering
methods give the molecular weight of aggregates of molecules. This point was, how-
ever, carefully investigated by Doty and Bunce," who from the linearity of the zero-
angle curve conclude that no aggregation occurs. It would therefore appear more
probable that the dielectric molecular weight is not the true molecular weight, but
that of an orienting unit.

6' P. Doty and B. Bunce, J. Am. Chem. Soc. 74, 5029 (1952).

*» G. Oster, Trans. Faraday Soc. 46, 794 (1950).

69 D. B. Smith and H. Sheffer, Can. J. Research B28, 96 (1950).

'» J. M. Gulland, D. O. Jordan, and C. J. Threlfall, J. Chem. Soc. 1947, 1129.

■"■ M. E. Reichmann, R. Varin, and P. Doty, J. Am. Chem. Soc. 74, 3203 (1952).

'2 R. Signer and H. Schwander, Helv. Chim. Acta 32, 853 (1949).

" S. Katz, J. Am. Chem. Soc. 74, 2238 (1952).

474

D. O. JORDAN

TABLE V

Some Examples of Molecular Weights op Various Pentose Nucleic Acids

Source of nucleic acid

Molecular weight

Method Reference

Yeast

1.7 X 10*

Diffusion

79

Yeast

1.03 X 10*

Diffusion

80

Yeast

1.03-2.33 X 10*

Diffusion

81

Yeast

1.7-3.5 X 10*

Diffusion

82,83

Yeast

6.0-7.0 X 10*

Dielectric

84

Liver

0.5-2.39 X 10*'

Escherichia coli

1.75 X 10*

Diffusion and sedimenta-

85

Pancreas

1.15 X 10*

tion

Malt

3.45-6.97 X 10*

Tobacco mosaic

virus

3.7 X 10*

\ Diffusion and sedimenta-

[79

Tobacco mosaic

virus

5.9 X 10*-2.9 X 10^/ tion

\86

Rat Liver

2.64 X 105

Sedimentation

85b.

h. Pentose Nucleic Acids

The molecular weights which have been recorded for pentose nucleic
acids vary from 1.31 X 10* to 2.9 X 10^ according to the source, method
of preparation, and method of measurement. The low value of 1.31 X 10*
was obtained by Myrback and Jorpes^^ and for many years was the main
evidence in support of the tetranucleotide hypothesis, now long discarded.
The experimental method employed by Myrback and Jorpes has been
criticized by Fletcher,^* who considered, on a re-evaluation of their data,
that the molecular weight of the sample was near 6 X 10*, which although

7* G. Jungner, I. Jungner, and L-G Allg^n, Nature 163, 849 (1949).

76 I. Jungner, Acta Physiol. Scand. 20, Suppl. 69, 1 (1950).

7' G. Jungner, Trans. Faraday Soc. 46, 792 (1950) ; G. Jungner and I. Jungner, Acta

Chem. Scand. 6, 1391 (1952); G. Junger, Acta Chem. Scand. 6, 1405 (1952).
" K. Myrback and E. Jorpes, Z. phijsiol. Chem. 237, 159 (1935).
7* W. E. Fletcher, On the Structure of Nucleic Acids, Doctoral thesis, London Univ.,

London, England, 1948.
79 H. S. Loring, J. Biol. Chem. 128, Sci. Proc. 33, 61 (1939).
*" F. G. Fischer, I. Bottger, and H. Lehmann-Echternacht, Z. physiol. Chem. 271,

246 (1941).
81 W. E. Fletcher, J. M. Gulland, D. O. Jordan, and H. E. Dibben, /. Chem. Soc.

1944, 30.
" I. Watanabe and K. Iso, J. Chem. Soc. Japan 71, 280 (1950).
" I. Watanabe and K. Iso, J. Am. Chem. Soc. 72, 4836 (1950).
8* G. Jungner and L.-G. Allg^n, Acta Chem. Scand. 4, 1300 (1950).
8' L. Delcambe and V. Desreux, Bull. soc. chim. Beiges 59, 521 (1950).
86a E. Volkin and C. E. Carter, J. Am. Chem. Soc. 73, 1516 (1951).
86b E. L. Grinnan and W. A. Mosher, J. Biol. Chem. 191, 719 (1951).
860 E. R. M. Kay and A. L. Dounce, J. Am. Chem. Soc. 75, 4041 (1953).
8« S. S. Cohen and W. M. Stanley, J. Biol. Chem. 144, 589 (1942).

PHYSICAL PROPERTIES OF NUCLEIC ACIDS 475

still low did not correspond to the theoretical value for a tetranucleotide
(1300). In view of the degradation which must generally occur during the
isolation of pentose nucleic acids, it is doubtful whether the values recorded,
of which examples are given in Table Y, have any direct connection with
the molecular weight of the parent nucleic acid in the cell and can only
refer to a particular sample isolated in a certain way. Furthermore, in view
of the degradation, the determinations will, in general, have been carried
out on polydisperse solutions, and it is evident that an analysis for hetero-
geneity and a fractionation should be made prior to molecular weight
determinations. Delcambe*^ and Ghuysen^^ have carried out such analyses
by means of solubility measurements, and Bacher and Allen^^ have com-
bined solubility and sedimentation studies in order to characterize pentose
nucleic acids.

High-molecular-weight (greater than 10^) nucleic acids have been ob-
tained from virus and, more recently, from a variety of different tissues
(85a, 85b, 85c). The highly polymerized ribonucleic acid obtained from
tobacco mosaic virus has been studied by Cohen and Stanley.*® The virus
nucleic acid was prepared by heat denaturation of the virus and had a
molecular weight of between 1.5 and 2.9 X 10\ This preparation was
heterogeneous, however, and decomposed spontaneously to give a nucleic
acid of molecular weight 5.9-7.0 X 10^ which possessed a higher degree of
homogeneity. From the original highly polymeric material, by treatment
with 5 % sodium hydroxide, a fairly homogeneous nucleic acid sample hav-
ing a molecular weight of 1.5 X 10^ was obtained.

3. Acid-Base Properties

The nucleotides, as has been shown, all possess characteristic acidic dis-
sociations owing to the presence of the primary and secondary phosphoric
acid groups, the amino group, or the — NH — CO^ — group in the molecule.
These groups again appear in the nucleic acids, and the analysis of the
electrometric titration curves yields information both as to the nature of
the internucleotide linkage and the macromolecular structure.

a. Deoxypentose Nucleic Acids

The early titration data on the deoxyribonucleic acid of thymus^"''' were conflict-
ing, the acid, in titration up to pH 8.0, being classified as pentabasic or tetrabasic by

"L. Delcambe, Bull. soc. chim. Beiges 59, 508 (1950).

88 J. M. Ghuysen, Bull. soc. chim. Beiges 59, 490 (1950).

89 J. E. Bacher and F. W. Allen, J. Biol. Chem. 184, 511 (1950).
'"H. Steudel, Z. physiol. Chem. 77, 497 (1912).

" R. Feulgen, Z. physiol. Chem. 104, 189 (1919).

92 P. A. Levene and H. S. Simms, J. Biol. Chem. 65, 519 (1925).

»» P. A. Levene and H. S. Simms, J. Biol. Chem. 70, 327 (1926).

476 D. O. JORDAN

different observers and the results interpreted in terms of an open-chain and cyclic
structure, respectively. The samples studied, however, were considerably degraded
owing to the methods of extraction that were employed, and the conflicting results have
been ascribed to the different degrees of degradation of the samples studied.^'- ss-ioo
The first study of an acid isolated by a mild procedure was that of Hammarsten,'"!
who concluded from conductivity titrations on the free acid that the latter showed
four acid dissociations for every four atoms of phosphorus, having the very approxi-
mate pK'a values of 2.4, 3.7, 4.3, and 5.2. Incomplete electrometric titration results
were obtained on similar preparations by Jorpes'"'' and by Stenhagen and Teorell,^"'
which indicated the absence of any appreciable secondary phosphoric acid dissocia-
tion since the solutions were found to be almost entirely unbuffered in the region
pH 6.0-9.0.

More extensive studies on a carefully prepared high-molecular-weight
sample of the sodium salt of the deoxyribonucleic acid of calf thymus have
been made by Gulland et al.,^^ whose results have been confirmed by Signer
and Schwander,^^ Cosgrove and Jordan,^"* and Lee and Peacocke'"^ for the
deoxypentose nucleic acids from lamb thymus, herring sperm, wheat germ,
and mouse sarcoma. The titration curve of the sodium salt of the deoxy-
ribonucleic acid of calf thymus is shown in Fig. 15. The addition of acid
or alkali to the solution in water does not at first bring about the ionization
of groups between pH 5.0 and 11.0, but outside these limits there occurs a
rapid liberation of groups titrating in the ranges pH 2.0-6.0 and pH 9.0-
12.0. On back-titration, either with acid from pH 12.0 or with alkali from
pH 2.5, a curve is obtained which is different from that representing the
initial (forward) titration and which exhibits a well-defined point of inflec-
tion in the neutral region and shows incipient points of inflection in the
regions of pH 12.0 and 2.0 corresponding, respectively, to approximately
2.0 equivalents of alkali and 3.0 equivalents of acid for each four atoms of
phosphorus. Gulland et al}^ found that the same back-titration curve was
obtained irrespective of whether the titration was commenced at pH 12.0
or 2.5. Lee and Peacocke,^^^ however, found that this was not so and that
slightly different curves were obtained on back-titrating from the different

s^K. Makino, Z. physiol. Chem. 232, 229 (1935).

95 H. Bredereck, M. Kothnig, and G. Lehmann, Ber. 71, 613 (1938).

98 H. Bredereck, M. Kothnig, and G. Lehmann, Ber., 72, 121 (1939).

" L. Ahlstrom, H. von Euler, I. Fischer, L. Hahn, and B. Hogberg, Arkiv. Kenii.

Mineral. Geol. A20, No. 15 (1945).
"» G. Schmidt, E. G. Pickels, and P. A. Levene, J. Biol. Chem. 127, 251 (1939).

99 S. S. Cohen, J. Biol. Chem. 146, 471 (1942).

"» J. M. Gulland, G. R. Barker, and D. O. Jordan, Ann. Rev. Biochem. 17, 175 (1945).

'»! E. Hammarsten, Biochem. Z. 144, 383 (1924).

"2 E. Jorpes, Biochem. J. 28, 2102 (1934).

'"3 E. Stenhagen and T. Teorell, Trans. Faraday Soc. 35, 743 (1939).

1"^ D. J. Cosgrove and D. O. Jordan, /. Chem. Soc. 1949, 1413.

'9* W. A. Lee and A. R. Peacocke, /. Chem. Soc. 1951, 3361.

PHYSICAL PROPERTIES OF NUCLEIC ACIDS

477

I I I I I I

2 3 4 5 6 7 8

PH

Fig. 15. The titration curve of sodium deoxypentose nucleate (Gulland, Jordan,
and Taylor«).

I Forward titration curve
II Back-titration curve

extremes of pH (Fig. 16). Whether this discrepancy is due to a difference
in drying procedure is not clear, but recent work by Garner et al}°^ using
undried nucleic acid has confirmed the results of Lee and Peacocke.

As the acid-base characteristics of the deoxypentose nucleotides have not
yet been studied, interpretation of the back-titration curve has to be made
with reference to the p/v'a values of the ribonucleotides. This procedure is to
some extent justified by the similarity of the acid-base properties of 9-
methylxanthine and xanthosine, which suggests that the acid-base prop-
erties of the pyrimidines and purines are not dependent on the nature of the
substituent radical in the 9(or 3)-position so long as it is a nonresonating
system and that the glycosidic, C — N, link remains a single bond. The
titration curve (Figs. 15 and 16) shows that two main types of titratable
group exist in the molecule: that titrating in the range pH 2.0-6.0 and that
titrating in the alkahne range pH 8.0-12.0. From the discussion of the dis-
sociating groups in the nucleotides it is evident that these groups are the

108 R. H. Garner, D. O. Jordan, and S. M. Matty, unpublished results.

478

D. O. JORDAN

Fig. 16. The titration curve of sodium deoxypentose nucleate (Lee and Pea-
cocked"*).

I Titration with alkali from pH 6.4 to 12.0
II Back-titration with acid from pH 12.0

III Titration with acid from pH 6.4 to 2.4

IV Back-titration with alkali from pH 2.4

purine-pyrimidine amino and the — NH — CO- — dissociations of guanine
and thymine, respectively. The number of these dissociating groups is
known with some accuracy from the analytical results of Chargaff^° and
Wyatt*'-^"^ [cf. Chargaff, Chapter 10], and in Table VI the results of Lee and
Peacocke^"^ are summarized for three deoxypentose nucleic acids. The
analysis'** of the sodium salt of the deoxyribonucleic acid from calf thymus
for sodium shows that there is one sodium ion for every phosphorus atom,
and, in view of the fact that the amount of the secondary phosphoric acid
dissociation is small (Table VI), these must be combined largely or entirely
with the primary phosphoric acid dissociations. The presence of the theo-
retical number of amino, primary phosphoric acid, and — NH- — CO — dis-
sociations is in agreement with the view that the deoxypentose nucleic
acids have a long-chain structure in which the internucleotide bond is
through a phosphoester linkage. The difficulty in accepting this type of
linkage in view of the very different behavior of deoxypentose nucleic acids,
compared with ribonucleic acids, towards alkali has now been overcome by
the conclusion of Brown and Todd^^^ that the lability of the latter acid is
due to the ease of formation of a cyclic intermediate involving the adjacent

1" D. M. Brown and A. R. Todd, J. Chem. Soc. 1952, 52.

PHYSICAL PROPERTIES OF NUCLEIC ACIDS

479

TABLE VI
The Titratable Groups of the Sodium Salts of Some Desoxypentose Nucleic

Acids

Groups

Calf Herring Wheat

thymus sperm germ

1. Amino (guanine)

2. Amino (adenine)

\ Equivalents/4P

(pK'a

\Equivalents/4P

3. Amino (cytosine and 5- jpK'a
methylcytosine") \Equivalents/4P

4. Secondary phosphoric acid ipK'a
dissociation \Equivalents/4P

5. — NH — CO— dissociation of jpK'a
guanine and thymine \Equivalents/4P

2.5
0.75

3.7
1.02

4.75
0.84

6.5
0.33

2.35
0.71

3.7
1.0

4.85
0.92

6.5
0.18

10.4,11.4 10.4,11.4
1.76 1.76

2.9
0.66

3.85
0.9

4.5

0.84

5.5
0.42

1.57

" pK'o of 5-methylcytosine assumed the same as that of cytosine.

sugar hydroxyl group to that bearing the phosphate group [cf . Brown and
Todd, Chapter 12.]. The presence of a small amount of secondary phosphoric
acid dissociation can only indicate that there is some chain-branching,
which will occur every ten or twenty nucleotides. This view is confirmed by
the dye-adsorption measurements of Cavalieri and Angelos,^"^ who, in order
to explain the experimental results, found it necessary to assume the exist-
ence of two binding sites. The ratio of these sites is lower than that found
from titration, being one in thirty.

The initial (forward) titration curve of the high polymeric deoxypentose
nucleic acids is anomalous in that on the addition of acid or alkali to the
solution in water, no groups are titrated at first between pH 5.0 and 11.0,
but outside these limits there occurs a rapid ionization of the amino groups
and the — NH — CO — dissociations. This general behavior was first ob-
served by Gulland et al^^ for the acid from calf thymus and has been con-
firmed by electrometric titration of other acids and for different methods
of preparation by Signer and Schwander,''^ Cosgrove and Jordan, ^"^ and Lee
and Peacocke.^"^ It has also been confirmed by spectrophotometric titration
by Shack and Thompsett.^"^ The ionization of groups at pH 11.5 and in the
range pH 3.5-4.5 is accompanied by a marked fall in the viscosity'^" and

io» L. F. Cavalieri and A. Angelos, /. A»i. Chem. Soc. 72, 4686 (1950).

i«9 J. Shack and J. M. Thompsett, /. Biol. Chem. 197, 17 (1952).

"0 J. M. Creeth, J. M. Gulland, and D. O. Jordan, J. Chem. Soc. 1947, 1141.

480 D. O. JORDAN

the disappearance of streaming birefringence.^"^ ■""•^^ This decrease in the
viscosity on the addition of acid and alkaU was considered by Vilbrandt
and Tennent"^ to be caused by depolymerization, which slowly reversed
when the solution was returned to neutrality. Gulland et al.,'^^ however,
have shown that this depolymerization cannot involve the rupture of the
internucleotide phosphoester linkages since there is no increase in the
amount of secondary phosphoric acid dissociation on back-titration, and
these authors suggested that some of the amino groups of the pyrimidines
and purines are linked by hydrogen bonds to the — NH — CO — dissocia-
tions; the ionization of either group would then bring about the liberation
of both as shown by the position of the back-titration curve (Figs. 15 and
16). The liberation of the groups at pH values more alkaline or acid than
would be expected from the normal pK'a values of the — NH^ — ^CO — and
amino groups, respectively, will be due, most probably, to the stability of
the nucleic acid hydrogen-bonded structure and the need to break several
of the hydrogen bonds simultaneously. The ionization would thus be
shifted in the direction of higher hydrogen and hydroxyl ion concentrations
as is shown by the titration curve.

From the titration data alone it is not possible to decide whether the
bonds are intra- or intermolecular. The marked decrease in viscosity which
has been observed by Creeth et al,^^'^ at almost the same pH values as the
ionization of the groups would serve to indicate that the bonds join smaller
molecular units, which in solution show lower viscosity than the hydrogen-
bonded macromolecules. The sedimentation data,^" although not conclusive
on this point, tend to confirm this conclusion.

These conclusions, drawn from titration results, are in remarkable agree-
ment with the structure proposed by Watson and Crick. ^^ The hydrogen
bonds between the single nucleic acid chains involve the amino groups of
adenine and cytosine and the — NH — CO— groups of guanine and thymine.
Furthermore, more careful examination of the titration curve^^ shows that
the anomalous behavior ceases at approximately pH 3.5 after the titration
of 1.8-2.0 equivalents of amino group, i.e., the back- and forvvard-titration
curves are coincident at pH values more acid than 3.5, which indicates that
the most acid amino group, viz., that of guanine, does not partake in the
formation of hydrogen bonds as suggested by Watson and Crick.

b. Pentose Nucleic Acids

The titration of pentose nucleic acids has yielded important information
concerning the nature of the ionizable groups and of the internucleotide
linkage. No information concerning the macromolecular structure has been
obtained, however, since the samples studied have all been isolated by use

111 C. F. Vilbrandt and H. G. Tennent, J. Am. Chem. Soc. 63, 1806 (1943).

PHYSICAL PROPERTIES OF NUCLEIC ACIDS 481

of fairly strong reagents and the nucleic acid has consequently suffered deg-
radation. It is not known, therefore, whether the pentose nucleic acids
exist in a similar hydrogen-bonded structure to that of the deoxypentose
nucleic acids.

Early work by Levene and Simms,^^ Jorpes,^°- Makino,^^ and Bredereck
et a/."^'''^ had shown that the ribonucleic acid of yeast contained three to
four ionizable groups per four atoms of phosphorus on titration to pH 8.0.
The first systematic study was that of Allen and Eiler,i'^ who were the first
to observe that the titration curve indicated the presence of three amino
groups for every four atoms of phosphorus, together with one group titrat-
ing in the range pH 5.0-7.5 and which they regarded as a primary phos-
phoric acid group that had been weakened by virtue of the ionization of
the remaining phosphoric acid dissociations. Fletcher et al.,^^^ who obtained
titration curves (Fig. 17) very similar to those of Allen and Filer, inter-
preted their results as indicating the presence, in the ribonucleic acid of
yeast, of three amino groups, one secondary phosphoric acid, two — NH —
CO — -, and three primary phosphoric acid dissociations for every four
atoms of phosphorus. The pK'a values of the nucleotides were assumed to
be unchanged in the nucleic acid. This conclusion was confirmed by the
titration curve of the deaminated acid (Fig. 17); the removal of the amino
groups permits the direct titration of the primary phosphoric acid dissocia-
tions, and the curve indicates the presence of three primary and one sec-
ondary phosphoric acid dissociations, the — NH — -CO — dissociation of
xanthosine which titrates in the range pH 5.0-7.0 (pK'a 6.0), and three
— NH— CO — dissociations titrating in the range pH 8.0-12.0. These re-
sults suggest that, on average, three secondary and one primary phosphoric
acid groups are utilized in yeast ribonucleic acid for every four atoms of
phosphorus, in forming the phosphoester internucleotide bond. This con-
clusion indicates that one in every four atoms of phosphorus (on average) is
triply esterified. The shape of the electrometric titration curve has been
confirmed by several workers, '^^"'", who have generally interpreted their
results in the same way although there is no general agreement about the

"2 H. Bredereck and M. Kothnig, Ber. 72, 121 (1939).
"3 H. Bredereck and I. Jochman, Ber. 75, 395 (1942).
"4 F. W. Allen and J. J. Eiler, J. Biol. Chem. 137, 757 (1941).
"5 W. E. Fletcher, J. M. Gulland, and D. O. Jordan, J. Chem. Soc. 1944, 34.
^'^ H. Chantrenne, Bull. soc. chim. Beiges 55, 5 (1946).

1" H. Chantrenne, K. Linderstr0m-Lang, and L. Vandendriessche, Nature 159, 877
(1947).

118 L. Vandendriessche, Com-pt. rend. trav. lab. Carlsberg, Ser. chim. 27, 341 (1951).

119 Y. Khouvine and J. Gregoire, Bull. soc. chim. biol. 28, 424 (1944).

120 C. A. Zittle, J. Biol. Chem. 166, 491 (1946).

121 G. Wiener, E. L. Duggan, and F. W. Allen, /. Biol. Chem. 185, 163 (1950).

1" L. F. Cavalieri, S. E. Kerr, and A. Angelos, /. Am. Chem. Soc. 73, 2567 (1951).

482

D. O. JORDAN

pH

Fig. 17. Titration curve of yeast ribonucleic acid, I, and of deaminated j'east
ribonucleic acid, II (Fletcher, Gulland, and Jordan"^).

relative amount of the secondary phosphoric acid dissociation and this
may well vary both with th^ method of preparation and the source of the
nucleic acid. The presence of a triply esterified phosphorus in the poly-
nucleotide necessarily involves the formation of a branched-chain structure.
It is not possible at present to say to which nucleotide the triply esterified
atom belongs; Cavalieri et al.,^'^^ in ascribing it to uridylic acid, have made
the same mistake as Fletcher et o/.,"^ later corrected by Gulland et al.,^^ in
that they have regarded the nucleotides containing amino groups as being
in the nonzwitterionic form.

The conclusions derived from electrometric titration data have been confirmed by
the dye-adsorption studies of Cavalieri et aZ.'"' '" In their studies of the adsorption
of rosaniline by pentose nucleic acid using the method of equilibrium dialysis, they
found that their results did not follow the simple equation: r/c = kn — kr (where r
is the amount of dye bound per mole of nucleic acid, c is the equilibrium concentra-
tion of the dye, n — rmax. , and k is the binding constant), but it was found necessary to
assume two types of binding sites in the nucleic acid and the equation r/c = niA:i/(l +
kic) + 712^2/(1 -f- k^c) was found to describe the experimental results. Since rosaniline

'" L. F. Cavalieri, A. Angelos, and M. E. Balis, /. Am. Chem. Soc. 73, 4902 (1951).

PHYSICAL PROPERTIES OF NUCLEIC ACIDS 483

is a cationic dye, it will be adsorbed only at negatively charged sites, and at neutral
pll these will be the phosphoric acid dissociations.

A further possibility of testing these conclusions lies in the different stability of
the tertiary and secondary esters of phosphoric acid towards alkaline reagents. '^^
The alkaline hydrolysis constants have been given by Cavalieri'^^ as being ca. 3.5 X
10"^ min.~i for the reaction: triester phosphate -^ diester phosphate, and ca. 2.9 X
10~^ min.~' for the reaction: diester phosphate — > monoester phosphate. Cavalieri's
results of a kinetic study''^^ of the alkaline hydrolysis of yeast ribonucleic acid again
suggests that there are at least two types of phosphate linkage, one more labile than the
other. The analysis of the kinetic data is, however, difficult in view of possible dif-
ferences in the rate of hydrolysis of bonds between different nucleotides, and Cavalieri
did not consider it justifiable to relate the two types of linkage definitely to the tri-
and diester phosphates.

4. The Solution Properties of Sodium Deoxypentose
Nucleate

Although studies of the properties and structure of nucleic acids and
their sodium salts in the solid state have yielded results of fundamental
importance, it is, nevertheless, the size, shape, and rigidity of the nucleate
ion in solution which is probably of greatest interest from the biological
viewpoint. In view of the very different nature of the environments, con-
clusions drawn from studies of the solid state may require some modifica-
tion when the properties in solution are considered. It has been known for
some time that the properties of the nucleate ion in solution are very de-
pendent on the pH and ionic strength of the solvent as well as upon the
previous treatment of the nucleic acid, and the nature of these changes
will now be considered.

Owing to the difficulty in obtaining high-molecular-weight preparations
of pentose nucleic acids, most studies have been made on the sodium salt
of a deoxypentose nucleic acid and this discussion will be confined to the
properties of this group of acids.

a. The Influence of Ionic Strength in Neutral Solution

Measurements of viscosity and streaming birefringence made on solu-
tions of sodium deoxypentose nucleate containing no added electrolyte
have not been found to be reproducible (see for example Sadron'-®), par-
ticularly at concentrations above about 0.005 %, the actual value varying
with the nucleate sample, being lower for high-molecular-weight prepara-
tions. This lack of reproducibility will be due to the slow degradation of the
nucleate which occurs in aqueous solution and which is accelerated by in-

'" G. M. Kosolapoff, "Organophosphorus Compounds," p. 232. John Wiley and Sons,

Inc. New York, 1950.
"6 L. F. Cavalieri, J. Am. Chem. Soc. 73, 4899 (1951).
'28 C. Sadron, Prog. Biophys. and Biophys. Chem. 3, 237 (1953).

484 D. O. JORDAN

crease of temperature and retarded by the presence of electrolyte.^^^ It
could also be caused by aggregation of the ions to form ionic micelles in the
more concentrated solutions. A further cause of the nonreproducibility is
that the properties vary considerably with the method of preparation of the
solution, and the most satisfactory method is to permit the fibrous nucleic
acid to dissolve slowly without stirring at ca. 0°, when solution is normally
complete within twelve hours. Solutions of sodium deoxypentose nucleate
in pure water as solvent and prepared in this way show a very high vis-
cosity, which is dependent to a marked degree on the rate of shear, and also
exhibit strong streaming birefringence. On the addition of electrolyte all
three effects are considerably reduced and the reproducibility of the meas-
urements greatly increased.

The decrease of viscosity on the addition of electrolyte''^ ■"''•^2^'^2' was
first explained by Greenstein and Jenrette^^* as a reversible depolymeriza-
tion of the deoxypentose nucleate. However, comparison of the acid-base
properties in 1 .0 M potassium chloride solution with those in pure water as
solvent^^ shows that there is no increase in the titratable acidic and basic
groups when the deoxypentose nucleate is in the former solvent and further-
more indicates that the hydrogen-bonded structure is not affected by the
increase in ionic strength. More definite evidence that the process is not one
of depolymerization is afforded by the light-scattering measurements of
Reichmann et alJ^ which show that an increase of ionic strength does not
change the molecular weight, but only produces relatively small changes in
the shape of the ion. The similarity of the behavior of the deoxypentose
nucleate ion and of synthetic polyelectrolytes such as poly-A''-n-butyl-4-
vinylpyridonium bromide to changes of ionic strength, led Jordan^^" to sug-
gest that in water solution the nucleate ion was fully stretched by virtue of
the repulsion between the charged (-^P0~) groups. The fall in viscosity
on the addition of electrolyte would then be produced by a coiling of the
molecule permitted by a neutralization of these charged groups. This mech-
anism is identical with that suggested by Fuoss and Strauss^^' for the syn-
thetic polyelectrolytes. The early work of Bungenberg de Jong and Kwan'^-
had indicated that there was a similarity between the viscosimetric behavior
of nucleic acids and polyelectrolytes, and this has been confirmed by Basu.^^^
In a more recent study, however, Pouyet^^^ has made measurements of the

" T. Miyaji and V. E. Price, Proc. Soc. Exptl. Biol. Med. 75, 311 (1950).

28 J. P. Greenstein and W. V. Jenrette, /. Natl. Cancer Inst. 1, 77 (1940).

29 G. Vallet and H. Schwander, Helv. Chim. Acta 32, 2508 (1949).
'» D. O. Jordan, Trans. Faraday Soc. 46, 792 (1950).

" R. M. Fuoss and U. P. Strauss, /. Polymer Sci. 3, 246; 602 (1948).

32 H. G. Bungenberg de Jong and U. S. Kwan, Kolloidchem. Beih. 31, 89 (1930).

" S. Basu, Nature 168, 341 (1951).

3< J. Pouyet, Compt. rend. 234, 152 (1952).

PHYSICAL PROPERTIES OF NUCLEIC ACIDS

485

1 — \ — \ — r

23456789 lO

Concentration gx IO/ti'-

Fig. 18. Variation of Vsp./c with concentration of sodium deoxypentose nucleate
at different concentrations of sodium chloride (Conway and Butler''^).
• 10-» N O 10-3 A^

3 10-" N € No added salt

viscosity of deoxypentose nucleate solutions at low rates of shear in a
Couette viscometer and has interpreted his results as indicating that at
infinite dilution of nucleate there is no influence of ionic strength on the
viscosity of these solutions. This view has been supported by the more ex-
tensive results of Conway and Butler/^^ A\ho determined \-iscosities at
sufficiently low rates of shear to permit extrapolation to zero shear (Fig.
18). These conclusions would indicate that the deoxj^pentose nucleate ion
does not show a contraction, as was suggested by Jordan/^° on the addition
of electrolyte, and Conway and Butler'^^ consider that the variation of
viscosity is due entirely to electrostatic interaction l^etween the nucleate
ions, which would be reduced by the addition of sodium ions.

The problem of the fall in viscosity on the increase of ionic strength is
clearly intimately related to the rigidity of the deoxypentose nucleate ion.
Further evidence on this problem comes from streaming birefringence
studies. '^^"'^^ Schwander and Cerf" find that the streaming birefringence

1" B. E. Conway and J. A. V. Butler, /. Polymer Sci., 12, 199 (1954).

'3^ O. Snellman and G. Widstrom, Arkiv Kemi Mineral. Geol. A19, No. 31 (1945).

>" H. Schwander and R. Cerf, Helv. Chim. Acta 32, 2356 (1949).

138 H. Schwander, J. chim. -phys. 47, 718 (1950).

139 H. Schwander and R. Cerf, Helv. Chim. Acta 34, 436 (1951).
i^» H. Schwander and R. Cerf, Experientia 7, 95 (1951).

"1 IL Schwander and R. Signer, Helv. Chim. Acta 34, 1344 (1951).

486

D. O. JORDAN

?

^o-'.

o°,>j:.'-^'* "

»9^< *

.r^'

t> O.OI9l«>/o
O O.OI4 7'>/o
• O.OI03»/o

400
G

Fig. 19. Streaming birefringence of solutions of sodium deoxypentose nucleate
in 10% sodium chloride (Schwander and Cerf'^^).

decreases with ionic strength, but the rotational diffusion constant is ap-
parently independent of both ionic strength and nucleate concentration
(Fig. 19) and has a value of 36.6 sec.~'. From these results and the magni-
tude of the birefringence, Sadron^^^ has concluded that the particles are
very long and not capable of contraction. The experiments of Schwander
and Cerf'^^ in which the birefringence was measured using solutions of
sodium deoxypentose nucleate in which the viscosity was varied by adding
glycerol, are, however, probably of greater significance. The birefringence
may arise either from the orientation of a rigid particle in the stream lines
or from the deformation of a random coil to give increased asymmetry
followed by orientation; the contribution of these two effects may be
analyzed by varying the viscosity of the solvent (see Cerf'^^ for theoretical
details). On increasing the viscosity of the solvent, tan a (where a is the
initial slope of the curve relating extinction angle and velocity gradient)
should be a linear function of the solvent viscosity if the particles are rigid.
Schwander and Cerf'^^ and Sadron^-^ first interpreted these results (Fig. 20)
as indicating that the nucleate ion was a rigid particle, the discrepancy at
higher values of the viscosity being attributed to heating effects. On the
basis of a more detailed theoretical treatment of the problem, Cerf"^ con-
siders that the change of slope at higher viscosities (Fig. 20) is due to a de-

i« R. Cerf, /. Polymer Sci., 12, 15 (1954).

PHYSICAL PROPERTIES OF NUCLEIC ACIDS

487

20-

15 —

10—

10 15

centipoises

Fig. 20. Values of tan ao for a solution of sodium deoxypentose nucleate with
sodium chloride when viscosity is varied by addition of glycerine (Schwander and
Cerf'^s).

formation of the nucleate ion, and the more detailed experimental study of
Horn et al}^^ confirms this view although the neutral ion in neutral solution
is much less deformable than the acid-treated material.

The light-scattering results of Reichmann et alJ^ and of llo\ven et al}^^
confirm the conclusion that the deoxypentose nucleate ion is deformable,
but the contraction on the addition of electrolyte to a water solution is much
less than for a typical polyelectrolyte. Thus Rowen et al}^'^ found that the
maximum dimension decreased from 6800 A. to 4500 A. when the ionic
strength was increased from 0 to 2 X 10"^. A similar change of solvent
with sodium polymethacrylate would produce about a tenfold decrease in
the maximum dimension."^ Furthermore, Reichmann et al?^ conclude that
the deoxypentose nucleate ion in 0.2 M sodium chloride solution is only
slightly more asymmetric than that of a random coil; if this be so, the
streaming birefringence can only arise by a deformation of the molecule.

Direct evidence that sodium ions are bound to the nucleate ion has been obtained
from determinations of the charge on the nucleate ion at various ionic strengths.
The charge has so far only been determined from measurements of the membrane
potential, and the measurements of Creeth and Jordan'^* have recently been extended
by Shack ei oL'" Their results are very similar in that the charge on the nucleate ion

»" P. Horn, J. Leray, J. Pouyet, and C. Sadron, J. Polymer Sci. 9, 531 (1952).

"^ J. W. Rowen, M. Eden, and H. Kahler, Biochim. et Biophys. Acta 10, 89 (1953).

i« A. 0th and P. Doty, J. Phys. & Collid Chern. 56, 43 (1951).

"« J. M. Creeth and D. O. Jordan, /. Chem. Soc. 1949, 1409.

488

D. O. JORDAN

at the ionic strengths studied is less than the theoretical value for the fully charged
ion, thus indicating a considerable degree of ion-pair formation. Shack et al.^" ob-
serve only a small variation of the charge with ionic strength, the average value in
the range 0.005 and 0.1 being 1.89 negative charges for every four atoms of phosphorus.
The variation of charge observed by both Creeth and Jordan and Shack et al. was an
increase of negative charge with increase of ionic strength. The reason for this varia-
tion is not clear, and, in view of the limitations and inaccuracies of the membrane-
potential method, it is clearly desirable that the charge should be investigated by
other methods.

Changes in the ultraviolet absorption spectrum on the addition of electro-
lyte which have been studied systematically by Thomas/^s Shack et al.,^^^
and Lawleyi^o are also consistent with the view that the nucleate ion is
capable of some deformation. These authors find that the value of the ex-
tinction coefficient at 260 m^ is lowered by the addition of electrolytes, the
cations having a specific effect which is dependent upon their charge. The
adsorption of Na+, K+, Mg^, Ba++, and Zn^ probably occurs at the phos-
phate groups and is nonspecific since the overall shape of the adsorption
band is not affected. H3O+ and Ag+ ions, however, appear to have a more
specific action since the shape of the absorption curve changes and Xmax.
shifts to longer wavelengths. The lowering of the absorption is unlikely to
be due to a shadowing effect and is more probably a true difference in the
absorption of the ring systems due to changes in the configuration of the
rings in the macro-ion. In the Watson and Crick structure^^ the extent of
stretching of the polynucleotide chain is determined by an equilibrium be-
tween the hydrogen-bonding and the Van der Waals forces on the one hand
and the repulsion between the charged groups on the other. A decrease of
the latter due to ion-pair formation will permit a small decrease in the
length of the ion and, owing to the new relative configurations of the purine
and pyrimidine rings, a change in the absorption spectrum. [Cf. Beaven,
Holiday, and Johnson, Chapter 14.]

6. Conclusions as to the Size and Shape of the Deoxypentose Nucleate Ion in
Neutral Solution

The problem of determining the dimensions of a polyelectrolyte of un-
known macro-structure is a most exacting one, as the small number of
successful studies with synthetic polyelectrolytes of known structure
testifies. However, from the experimental evidence discussed above it is
possible to draw some conclusions concerning the size and shape of the

1" J. Shack, R. J. Jenkins, and J. M. Thompsett, J. Biol. Chem. 198, 85 (1952).
"* R. Thomas, Bull. soc. chim. biol., in press.

1" J. Shack, R. J. Jenkins, and J. M. Thompsett, J. Biol. Chem., 203, 373 (1953).
'^» P. D. Lawley, Studies in the Behaviour of Polyelectrolytes Doctoral thesis, Not-
tingham Univ., Nottingham, England, 1953.

PHYSICAL PROPERTIES OF NUCLEIC ACIDS 489

deoxypentose nucleate ion. Such conclusions must, as yet, be tentative
owing to the marked disagreement which exists between some of the ex-
perimental results and more particularly in the conclusions draw^n there-
from. The maximum dimension of the particle has been given as 8000 A.'^^
from streaming birefringence measurements and as 6500 A.,^^ 6800 A.''''*
and 5100-7200 A.^*' from light-scattering measurements on solutions of
ionic strength 0.2. There is little information on the shorter dimension, but
Schwander'^^ gives the value of 10 A. as determined from streaming bire-
fringence. This value is somewhat less than the value of 15-24 A. given by
Rowen ct al.^'^'^ from high-resolution electron microscopy, which is in agree-
ment with the previous value of 15 A. given by Williams'^- and also ob-
tained from measurements with the electron microscope. If these values
are of the correct order then the axial ratio will be about 350, which is to
be compared with the approximate value of 120 obtained by Cecil and
Ogston^" from sedimentation measurements.

The relatively small deformability possessed by the nucleate ion when
compared to a typical polyelectrolyte can be attributed to the hydrogen-
bonded structure suggested by Gulland ct al.'^^ and to the presence of strong
intramolecular Van der Waals forces as postulated by Schwander and Sig-
ner.'^^ There is also, as has been pointed out by Conway and Butler,'^^
an important difference between the nucleate ion and a typical polyelec-
trolyte, since in the former, the charges (— >P0~) are carried on the phos-
phoester "backbone" and in the latter are generally carried on short side
chains, thereby permitting greater relative movement of the charged
groups.

Comparison of these views on the shape and structure of the molecule
with the Crick and Watson structure^^ is interesting. This structure indi-
cates a comparatively rigid molecule in view of the strong hydrogen-bond-
ing and Van der Waals forces between the purine and pyrimidine ring sys-
tems of the two chains. The length and rigidity of the molecule will be
determined by the equilibrium between the repulsion between the charged
groups tending to extend the molecule on the one hand and the hydrogen-
bonding and Van der Waals forces tending to contract the molecule on the
other. Absorption of sodium ions, with the resulting neutralization of some
of the charged groups through ion-pair formation, Avill disturb this equilib-
rium and lead to a more compact molecule. The evidence for a rigid non-
deformable molecule rests on the independence of the viscosity of infinite
dilution on ionic strength as determined from viscosity measurements^^'* -^^^
and on the variation of the streaming birefringence in solutions of different
viscosity."" The latter evidence is capable of different interpretation, ^^^•"^

"1 R. F. Steiner, Trans. Faraday Soc. 48, 1185 (1952).
1" R. C. Williams, Biochim. et Biophys. Acta 9, 237 (1952).

490

D. O. JORDAN

whereas in the case of the former measurements it is doubtful whether vis-
cosities in very dilute solutions can be determined with sufficient accuracy
to decide whether the rjap./c — vs. — c curves obtained with and without
added electrolyte extrapolate to the same point. It is to be concluded there-
fore that the deoxypentose nucleate ion in solution is deformable, but to a
much lesser extent than a typical polyelectrolyte or uncharged polymer.

One further fundamental problem should be mentioned. For a deoxy-
pentose nucleate ion having the structure suggested by Crick and Watson''^
and a molecular weight of 8 X 10«, the dimensions would be 40,000 A. by
20 A. This length is too great by a factor of five or six and can clearly only
be reduced by increasing the smaller dimension either by coiling or by chain-
branching. It would therefore appear impossible for the experimental values
of 8 X 10« for the molecular weight and 6800 A. by 20 A. for the dimen-
sions to be correct for a particular set of conditions. The value for the
molecular weight and length were obtained from the same measurements
(light-scattering) but the width is much less certain and depends largely
on measurements on the dried material. It is unlikely that a particle of size
40,000 A. by 20 A. should remain as a rigid rod in solution, but would coil
to give a much less asymmetric particle as the light-scattering evidence
indicates, ^^-^i The suggestion ^^ that the symmetry is due entirely to chain-
branching would appear not to be in agreement with the streaming bire-
fringence results since deformation would not then be possible. It is clearly
evident that more reliable experimental data is required before a complete
picture of the size and shape of the nucleate ion in solution is obtained.

c. The Influence of Changes of pH on the Size and Shape of the Deoxypentose
Nucleate Ion in Solution

The comparison of the forward- with the back-titration curve of sodium
deoxypentose nucleate (p. 477, Fig. 15) shows that the action of both acid
and alkali produces a marked change in the properties of the nucleic acid.
This behavior has been ascribed by Gulland et al}^ to an irreversible break-
ing of the hydrogen bonds existing between the amino and — NH— CO —
groups in the nucleate ion. The breaking of the hydrogen bonds occurs at
pH values of 5.0 and 11.0, between these values no groups are titrated on
the addition of acid or alkali and no hydrogen bonds are broken. In a re-
lated study of the viscosity changes produced in a 0.24 % solution of sodium
deoxypentose nucleate, Creeth et al}^^ observed that the viscosity dropped
sharply in the region of the same critical pH values of 5.0 and 11.0 at which
the titration of hydrogen-bonded groups occurred. This behavior was in-
terpreted as indicating that the hydrogen bonds unite molecular units which
are more symmetrical and of lower molecular weight than the original
nucleate. These units become the disperse species in acid and alkaline solu-

PHYSICAL PROPERTIES OF NUCLEIC ACIDS 491

tion. The earlier suggestion of Vilbrandt and Tennent"' that the fall in
viscosity produced by the action of acid or alkali was due to a depolymeriza-
tion of the nucleate appears to be untenable in \dew of the titration evi-
denced^ that phosphoester bonds are not broken by these reagents.

The solutions studied by Creeth et al}^^ were too concentrated to permit
any real analysis in terms of molecular dimensions to be made, and more
recently Schwander^^^ has measured the viscosities of more dilute solutions
in 1 % sodium chloride solution at pH 3.70 and 6.60 and finds a much
smaller decrease of viscosity than that observed by Creeth et al. Sedimenta-
tion studies by Vilbrandt and Tennent"' and Cecil and Ogston*° showed
that the action of both acid and alkali causes a decrease in the sedimenta-
tion coefficient. Cecil and Ogston found that two separate components
appeared after the addition of acid {\ M HCl); one was homogeneous and
resembled the original nucleate in neutral solution and the other was poly-
disperse and had apparently been formed by the disaggregation of the
original material. The addition of alkali produced a larger lowering of the
sedimentation coefficient than did the action of acid and the disaggregation
was more complete.

Creeth ct al}^'^ observed that on neutralizing to pH 7.0 a solution of
deoxypentose nucleate which had been treated with alkali at pH 12.0, a
slow increase of viscosity occurred and after 90 hours the viscosity resem-
bled that of the original solution. This behavioui' has been shown by
Zamenhof and Chargaff^^^^ to be due to an artifact having highly thixotropic
behaviour. The reversibility of the disaggregation was not observed on
treatment with acid.

A very much clearer picture of the changes that occur in sodium deoxy-
pentose nucleate on treatment with dilute acid has been given in a very
careful study by Reichmann et al}^^ A great criticism of all the previous
work is that the pH of solutions was changed by adding small amounts of
relatively strong acid. This procedure must inevitably produce transient
regions of much lower pH (or higher pH if alkali is added) than that ulti-
mately attained at equilibrium. This treatment will thus produce, in some
particles, greater degradation than in the remainder. In order to prevent
the deoxypentose nucleate particles from coming into contact with con-
centrated acid, Reichmann ct al}^^ adjusted the pH by dialysis. Using the
light-scattering method, these authors found that at pH 6.5, 3.0, and 2.6
in 0.2 M sodium chloride solution, the deoxypentose nucleate ion has the
same molecular weight (7.7 X 10®). There is, however, a marked contrac-
tion of the ion which is quite pronounced at pH 3.0, but very much greater

»" H. Schwander, Helv. Chim. Acta 32, 2510 (1949).

15^ S. Zamenhof and E. Chargaff, J. Biol. Chem. 186, 207 (1950).

1" M. E. Reichmann, B. H. Bunce, and P. Doty, J. Polymer Set. 10, 109 (1953).

492 D. O. JORDAN

at pH 2.6. This is shown by the values of the root-mean-square end-to-end
distance, which assuming a monodisperse species are: at pH 6.5, 5030 A.;
pH 3.0, 4340 A.; pH 2.6, 2150 A. Furthermore this contraction is com-
pletely reversible, the value of the end-to-end distance being 5200 A. on
neutralizing the solution to pH 6.5 from pH 2.6. It is thus evident that, at
pH 2.6 in 0.2 M sodium chloride solution, degradation of the particle does
not occur, but only a marked contraction. It is also evident that the de-
formability of the molecule at pH 2.6 is greater than at pH 6. This will be
due to the removal of many of the hydrogen bonds, thus reducing the
rigidity of the molecule so that when the charged groups are neutralized by
combination with a proton, the molecule will assume the shape approaching
that of a random coil. The reversibility of the contraction (which is also
reflected in the viscosity measurements of Reichmann et al.^^^) is due to the
fact that, at pH 2.6 in 0.2 M sodium chloride, only about one-half of the
amino groups have been titrated and therefore a number of the hydrogen
bonds will remain unbroken and these are apparently sufficient to hold the
macro-ion together. At lower pH values, at this ionic strength, all the
hydrogen bonds will be broken and the molecular sub-units, capable of
independent existence, produced. Such a change will, in all probability, be
irreversible.

The results of Reichmann et al}^^ are not in complete agreement with a
similar investigation carried out by Horn et al.^^^ The latter authors found
that an irreversible change in viscosity occurred when the pH of the solu-
tion was changed from pH 7.0 to 3.8 in 1 M sodium chloride. However, the
pH was changed by the addition of 0.01 M hydrochloric acid to the solution
and not by dialysis, which may account for the irreversible nature of the
change. The higher ionic strength used compared with that employed by
Reichmann et at. will make the degree of ionization at the different pH
values more comparable in view of the influence of ionic strength on the
dissociation constants of the amino groups. It would appear important that
much more precise information concerning the degree of ionization is neces-
sary for a true analysis of these studies.

The properties of the products of acid treatment obtained by Horn et al}*^
are interesting. Since the change was irreversible, it would appear that the
hydrogen-bonded structure had been destroyed and the sub-units, two or
more per molecule, liberated. The streaming birefringence results in solu-
tions of different viscosity (adjusted by the addition of glycerol), show that
molecular sub-units are much more deformable than the original nucleic
acid and appear to possess a much more typical polyelectrolyte behavior
as would be expected for a single, non-hydrogen-bonded polynucleotide
chain.

CHAPTER 14

Optical Properties of Nucleic Acids and Their Components
Cx. H. BEAVEN, E. R. HOLIDAY, AND E. A. JOHNSON

1. Introduction 493

II. Bases, Nucleosides, and Mononucleotides 495

1. Bases 498

2. Nucleosides 504

3. Nucleotides 511

III. Nucleic Acids and Polynucleotides 514

1. General 514

2. The Absorption Spectra of Nucleic Acids 516

3. Effect of pH on the e (P) of DNA 522

4. Effect of Salts on the Absorptivity of DNA 525

5. Effect of Other Agents on the e(P) of DNA 529

6. Spectrophotometric Estimation of Nucleic Acids 529

7. Nucleoproteins 530

IV. Ultraviolet Dichroism 532

1. General Theory 533

2. Form Dichroism 534

3. Definitions and Equations 535

4. Dichroism of Simple Compounds 537

5. Dichroism of Nucleic Acid 538

6. Influence of Dichroism on the Microspectrographic Estimation of Nucleic
Acid in Intact Cells 542

7. Dichroism of Stained Nucleic Acids 544

V. Infrared Absorption Spectra 545

I. Introduction

Although ultraviolet absorption spectroscopy has been employed for the
characterization of nucleic acids and their derivatives for more than twenty
years, technical improvements in that period, in particular the introduction
of the photoelectric spectrophotometer, have reduced considerably the
value of most of the earlier work, which is in consequence not usually re-
ferred to here. Similarly, recent advances in instrumentation have stimu-
lated the application of infrared absorption methods to the study of nu-
cleic acids, although the full value of such methods for both analysis and
molecular structure determination has not yet been exploited.

In the present field absorption spectroscopy fulfils two principal func-
tions, as an analytical tool and as a basis for deductions concerning struc-
ture. To date, the first of these functions has been by far the more important
since the hydrolysis products of nucleic acids lend themselves readily to

493

494 G. H. BEAVEN, E. R. HOLIDAY, AND E. A. JOHNSON

such methods. A large proportion of spectroscopic work in the ultraviolet
region has however been carried out with a view to serving an author's own
particular analytical requirements, and both the scope of the work and the
manner of publication have sometimes reduced its general utility, resulting
in much unnecessary duplication. A critical survey of work in this field also
demonstrates forcibly that the use of compounds of a very high order of
purity does not of itself ensure the production of spectroscopic data of
comparable accuracy. The measurement of molecular extinction coefficients
to an absolute accuracy of better than 1%, or even 5% (Caster^), neces-
sitates careful attention to possible instrumental and operational errors
and to inter-instrumental discrepancies. These causes of error have re-
cently been discussed in some detaiP (Goldring et al}).

The data given here are reproduced in accordance with the suggested
nomenclature of Brode^ and Hughes.^ The following symbols are used:-

A = absorbance (optical density) = ecd = log y

where e = molar absorptivity;

c = molar concentration;

d = internal cell length in centimeters;

/o = intensity of incident radiation;

7 = intensity of transmitted radiation.
Wavelengths (X) are given in millimicrons (mju). From the point of view of
theoretical interpretation of absorption spectra it would be preferable to
tabulate or plot frequencies rather than wavelengths, and if a wide spectral
band is to be covered such a plot also gives a more compact spectrum,
avoiding excessive spread at longer wavelengths. Since, however, the cali-
bration of all current ultraviolet instruments is given in terms of wave-
lengths, these are retained here. The manner of presentation of absorption
curves has in the past depended on the format favored by the journal in
question, which is, with a very few noteworthy exceptions, invariably
unsatisfactory from the point of view of the working spectroscopist. All
such curves should be given on an adequate scale, have coordinate grids
fine enough to permit accurate interpolation in both directions, and if
possible be printed in such a way that their scales have a metric basis to
permit direct measurement with an ordinary ruler.* It is rarely necessary

* See pocket on inside back cover for scale drawings of Figs. 3 through 20.

1 W. O. Caster, Anal. Chem. 23, 1229 (1951).

^Photoelectric Spectrometry Group Bull., No. 3 (Oct. 1950).

3 L. S. Goldring, R. C. Hawes, G. H. Hare, A. O. Beckman, and M. E. Stickney,

Anal. Chem. 25, 869 (1953).
^ W. R. Erode, /. Opt. Soc. Amer. 39, 1022 (1949).
6 H. K. Hughes, Anal. Chem. 24, 1349 (1952).

OPTICAL PROPERTIES OF NUCLEIC ACIDS 495

to plot absorptivities on a logarithmic basis for relatively simple spectra
such as those of purines and pyrimidines. The details of shape are thereby
blunted and the tedious process of obtaining quantitative estimates from
small-scale curves is made even more tiresome and inaccurate. Even if the
above conditions are fulfilled it is always desirable to provide in addition a
table giving molar absorptivities and wavelengths of principal features,
including minima as well as maxima.

II. Bases, Nucleosides, and Mononucleotides

The structural significance of the ultraviolet absorption spectra of purines
and pyrimidines has been well reviewed by Jordan,^ who, following Mar-
shall and Walker,' has emphasized the role played by ionization in the
spectral changes observed on varying the pH of solutions. Marshall and
Walker have pointed out that the more accessible pyrimidines, including
those concerned here, have a number of functional groups as substituents
which make detailed interpretation of their spectra very difficult. These
workers, and also Boarland and McOmie^ and Brown and Short, ^ give
spectra for many simple pyrimidines, including pyrimidine itself. The
ultraviolet absorption spectra of the simple diazines have been discussed
by Halverson and Hirt,'" who, however, employ the older spectrum of
pyrimidine determined by Heyroth and Loofbourow." Less attention has
been devoted to a general examination of the purines. Here again, however,
Stimson and Reuther^- and Cavalieri et al.^^ have attributed the spectral
changes observed with change in pH to keto-enol tautomerism, although
the close coincidence of pH values determined by titration with the pH
regions where such changes occur, points unmistakably to their ionic charac-
ter. The importance of pH control when measuring purine absorption
spectra was emphasised as early as 1930 by one of us (E. R. H.^^), but
measurements are still frequently made at an arbitrary pH value which
lies so close to a pK that a mixture of ionic species is present. Spectroscopic
measurements have in fact proved a useful method for the determination of
many additional pK values, some of which lie outside the range of con-
venient measurement by conventional means. The work of Shugar and

* D. O. Jordan, Progr. Biophys. and Biophys. Chem. 2, 51 (1951) ; Ann. Rev. Biochem.
21, 209 (1952).

7 J. R. Marshall and J. Walker, J. Chem. Soc. 1951, 1004.

8 M. P. V. Boarland and J. F. W. McOmie, J. Chem. Soc. 1952, 3716, 3722, 4942.

9 D. J. Brown and L. N. Short, J. Chem. Soc. 1953, 331 .

1" F. Halverson and R. C. Hirt, J. Chem. Phys. 19, 711 (1951).

'1 F. F. Heyroth and J. R. Loofbourow, J. Am. Chem.. Soc. 56, 1728- (19.34).

12 M. M. Stimson and M. A. Reuther, J. Am. Chem. Soc. 65, 153 (1943).

" L. F. Cavalieri, A. Bendich, J. F. Tinker, and G. B. Brown, J. Am.. Chem. Soc. 70,

3875 (1948).
" E. R. Holiday, Biochem. J. 24, 619 (1930).

496

G. H. BEAVEN, E. R. HOLIDAY, AND E. A. JOHNSON

^ 6

220

240

300

320

260 280

Wavelength, idm

Fig. 1. Isomeric cytidylic acids (Cohn, unpublished)

5'- and deoxy-

3'-

Pqx1*i6 is a good illustration of the application of such methods, and is
noteworthy for the excellent style of publication of the spectral data.

The absorption characteristics of a given compound in aqueous solution
can therefore be given as a series of spectra of the individual ionic species,
together with a list of the corresponding pi^ values preferably as deter-
mined spectroscopically.

The pK value for a given ionization can be determined from the follow-
ing expression (see Hammett and co-workers'^ and Edwards'^) :

pKa = pH — log

1^ D. Shugar and J. J. Fox, Biochim. et Biophys. Acta 9, 199 (1952).

i« J. J. Fox and D. Shugar, Biochim. et Biophys. Acta 9, 369 (1952).

" L. A. Flexser, L. P. Hammett, and A. Dingwall, J. Am. Chem. Soc. 57, 2103 (1935).

" L. J. Edwards, Trans. Faraday Soc. 46, 723 (1950).

OPTICAL PROPERTIES OF NUCLEIC ACIDS

497

Fig. 2. Relation between absorbance ratios and pH for the isomeric cytidylic
acids (Cohn, unpublished).

5'- and deoxy-

3'-

2'-

where cha and €a are the absorptivities of the two ionic species, and « is
the absorptivity determined at the given pH. The thermodynamic vakie
of pKa may be obtained by appropriate correction for the ionic strengths.
The tables given here also include absorbance ratios based largely on
the very comprehensive work of Cohn^^ (with subsequent additions and
revisions). These have considerable utility for the identification of com-
pounds in solutions of unknown concentration, but their precise values
will be critically dependent on the wavelength calibration of the spectro-
photometer, and variations must be expected when different instruments
are used. The relationships of these ratios to pH values and to the absorp-
tion spectra are illustrated for the isomeric cytidylic acids by Figs. 1 and 2
(Cohn, unpublished). It may be noted that the phosphate ionizations ap-
pear to be without significant effect on the spectra, but this is not neces-
sarily always true (see Table III).

'« E. Volkin and W. E. Cohn, Methods of Biochemical Analysis, Vol. 1, p. 287, Inter-
science, New York, 1954.

498

G. H. BEAVEN, E. R. HOLIDAY, AND E. A. JOHNSON

260

Wavelength, m^

Fig. 3. Adenine (Johnson, unpublished). The absorption spectra illustarted here
and in Figs. 4-20 will be found drawn to scale in the pocket on the inside back cover
of the volume.

From such data it is possible to select wavelengths and obtain coefficients
for the identification and quantitative estimation of bases, nucleosides,
and nucleotides obtained from nucleic acids by any of the methods des-
cribed in other chapters.

1. Bases

In spite of the fact that ultraviolet absorption data for adenine have
been reported in at least thirteen separate papers, no set of data has yet
been published which can be considered in any way comprehensive. This
is equally true of the other purines found in or derived from nucleic acids.
It has therefore seemed necessary to undertake a more thorough examin-
ation of them. Spectra for adenine, hypoxanthine, guanine, and xanthine
(measured with a calibrated Unicam SP 500 spectrophotometer, using
specimens all purified by various methods in this laboratory) are given in
Figs. 3 to 6, respectively, and corresponding numerical data in Table I.

Spectral data for adenine and guanine in dilute acid and in some cases
at other pH values have been determined recently by Hotchkiss,^'' Cav-
alieri et al.,^^ Vischer and Chargaff,^! Kerr et al.p Tsuboi and Stowell,^^

2» R. D. Hotchkiss, J. Biol. Chern. 175, 315 (1948).

" E. Vischer and E. Chargaff, J. Biol. Chem. 176, 703 (1948;.

OPTICAL PROPERTIES OF NUCLEIC ACIDS

499

200

220

240

280

300

320

260
Wavelength, rrtfi

Fig. 4. Hypoxanthine (Johnson, unpublished).

Wyatt,-^ Markham and Smith," and Loring et al.f^ and the results pre-
sented here agree quite well with the more recent of these. It does not seem
to have been observed previously, however, that both compounds have an
additional pK in the region of pH 0, and show marked spectral changes
presumably associated with the ionization of one of the ring nitrogen atoms.
Some methods of spectroscopic analysis have failed to take this into ac-
count.

The first two references of this group also contain data for hypoxanthine
and xanthine, absorption curves for which have in addition been published
by Stimson and Reuther.^'^

Absorption data for cytosine, uracil, and thymine have been published
in recent years by Hotchkiss,^^ ^'ischer and Chargaff,^! Stimson," Wyatt,^''
and Shugar and Fox.^^ Excellent data for the first two of these are given by

" S. E. Kerr, K. Seraidarian, and M. Wargon, J. Biol. Chem. 181, 761 (1949).

" K. K. Tsuboi and R. E. Stowell, Biochim. et Biophys. Acta 6, 192 (1950).

24 G. R. Wyatt, Biochem. J. 48, 584 (1951).

" R. Markham and J. D. Smith, Biochem. J. 49, 401 (1951).

" H. S. Loring, J. L. Fairley, H. W. Bortner, and H. L. Seagran, /. Biol. Chem. 197,

809 (1952).
" M. M. Stimson, /. Am. Chem. Soc. 71, 1470 (1949).

500

G. H. BEAVEN, E. R. HOLIDAY, AND E. A. JOHNSON

in6NHCI

inO.lA'HCI

in pH 7 buffer

inNaOH, pH 10.9

inA'NaOH

260
Wavelengtn, m/j

320

Fig. 5. Guanine (Johnson, unpublished).

Ploeser and Loring,^^ and Andrisano and Modena^^ discuss the spectra of
numerous pyrimidines including uracil, at various pH values, without
mention of a pK. Wyatt'*^ and Cohn^^ first gave data for 5-methylcytosine,
and Wyatt and Cohen^^ have reported 5-hydroxymethylcytosine isolated
from bacteriophage nucleic acids, with maxima at 279 m/x, e = 9700, in
0.1 N hydrochloric acid, at 269.5 mn in pH 7.4 buffer, and at 283.5 m^ in
0.1 A^ sodium hydroxide. For all except this last compound very complete
data have been given by Shu gar and Fox,^^ and the curves published here
(Figs. 7 to 10) have been re-drawn from their paper. It has been pointed
out by Cohn (private communication) that uracil has a pK in the region of
pH 0.5; the spectrum in 6 A" HCl was therefore determined in this labora-
tory and added to those of the other ionic species.

Good general agreement was found between the data for the pyrimidines
by Shugar and Fox, those given here for the purines, and an extensive series

" J. M. Ploeser and H. S. Loring, J. Biol. Chem. 178, 431 (1949).

29 R. Andrisano and G. Modena, Gazz. chim. ital. 81, 405 (1951).

30 G. R. Wyatt, Biocheyn. J. 48, 581 (1951).

31 W. E. Cohn, /. Am. Chem. Soc. 73, 1539 (1951).

32 G. R. Wyatt and S. S. Cohen, Nature 170, 1072 (1952).

OPTICAL PROPERTIES OF NUCLEIC ACIDS

501

200

220

240 260

Wavelength, m/i

Fig. 6. Xanthine (Johnson, unpublished).

320

of absorbance ratios and values for e at 260 m/x determined by Cohn'^
(with subsequent revisions and additions), whose complete spectra were
unfortunately not available. In view of the limitations to the accuracy of
absorbance ratios mentioned above, in Table I the values of Cohn are
quoted for the sake of consistency whenever they are available, supple-
mented from other sources when necessary. Some variations from values
obtained by inspection of the absorption curves will therefore be found.
Table I is intended to give data for individual ionic species, and the pK
values in the second column are placed between the two species to whose
intercon version they apply. Wherever possible the determinations for each
species were carried out at a pH removed at least 1 .5 and preferably 2 units
from the nearest pK, and in solutions of low ionic strength.

The determination of absorption spectra of purines and pyrimidines at
low temperatures has been carried out on sublimed films at 90° K. by Brown
and Randall,^^ and on sublimed films and also solutions in mixed organic
solvents which form glasses at 77° K. and 23° K. by Sinsheimer ct al.^*
The vibrational fine-structure is enhanced, particularly in pyrimidines such

" G. L. Brown and J. T. Randall, Nature 163, 209 (1949).

" R. L. Sinsheimer, J. F. Scott, and J. R. Loofbourow, J. Biol. Chem. 187, 299, 313
(1950); J. Am. Chem. Soc. 74, 275 (1952).

502

G. H. BEAVEN, E. R. HOLIDAY, AND E. A. JOHNSON

o

1

c

o3
J3
(i
O
M

<1

o o

IC lO »c

CO O IM

1 O O O

(600

0 i^ t^

d d d d

10 10 IC

1 05 --*< 00 0

1 -^ m lO CO

Q t^ CM 1^

0 0 CI CM

00 00 a
t^ 0 CO

0000

0 0 0 CM

0 0 CM

1 t^ (M 0

d d d

Tfi Cl (N o>

d d d d

0 «

-* -^ CO ^

1 00 0 <-! CM

iO ^ ^^ Gi
■-< CD t^ CO

CO 00 00

lO iCi CM

0 >— 1 I-H >— 1

0 0 ^ CM

r-l 0 CO

o o
to to

1 CO CD t^

1 t^ I>- w

di d <6

10

0 IM Tt< 0

•* CO 00 t^
,-i ,-t d, d

10 >?5
1 1^ CM 00 0

1 CO ■* 0 00

r-i 7-^ d d

10 "O «
CO CD C5 »— 1

t^ 10 CI I-H

d d .-< .-I

00 00 C5
-^ t^ "3

d d d

o

ic 0 CO ■*

CO Oi oi CO

00 0 CM ■* CO

^ 00 CM t^

0 10 CO

^ CO CO 0
IV

t- r^ 0 ^

^ ,-1

00 00 r^ CD t^
lA

ci 00 10 CO

CD iC CM

03
S

"S

c3
B

a
1

1 1 1 1

1 1 1 1

1 "5 CO 0 1

CO -*' d

^. 1 =? 1
10 in

1 1 1

a

i

1 1 1 1

1 1 1 1

1 T}< Tt Tt< 1

' CM CM CO '
CM CM CM

CM CM

1 1 1

OS

a

1 1 1 1

1 1 1 1

1 Tf t^ CO 1
^ ^ d d ^

>0

CO 1 05 1
d ' 00 '

S 1 1

h-l

M 02

< PQ

i

a

1 1 1 1

1 1 1 1

10

1 06 CD CO 1
CM CM CM

10 10
d 1 0 1

CO ' >* '
CM CM

2 1 1

CM

a
1

1 10 lO 10
1 ic 10 CO

C^ (N CO

t^ lO t^ lO

r-H <N CO CO

1 kC lO 10 lO

1 rH 0 0 .-1

CM t^ 0 CO

CM CO -^

r^ t^ CO o

kc C-? 10 CO

^ ->*< ,-(

a

1 0 CD t^
<N C^ CO
<M C<) CM

10 .-i iM CO
-H (M CO CO
(M (M (M (M

1

1 t^ CM iC OS
CO CD 10 CO
CM CM CM CM

iC lO

CM d l^ t^
•^ CO iC IC
CM CM CM CM

06 t- d

CO "* 10
CM CM CM

1

. 10 0

i-H CO CO

CO CO (M

10.8
10.7

11.05
11.45

10 10

CO — 1 0 05

10 10

^ CM CO •*

o2S

r- 00 00 oi

C5 0 C5 C5

0 CO t^

a

1 (N d CI

1 CD CD CD
(M (M (M

248
249.5
258.5
262.5

lO lO uO

1 10 ic CO ■*
t^ t^ 1^ t^

CM CM CM CM

0 t^ t^ "^

S S CM CM

CD t^ CM
t^ CO 00
CM CM CM

d --I 00
V ^ d

(M 05 (^
0 CJ /\

10

CM

d CM CO •

V CO 05 U

00
==00

10 CM
■>*< CM

a

o

a
S
o

•0

a
"S

•0

a;
a

'.a

c

X

a,

■0

0)

.2
'S

03

3

0

4)

a

c

c3
><1

a
■§
>>

OPTICAL PROPERTIES OF NUCLEIC ACIDS

503

c^ »c >o

ti lO t^

(m' d -^

0.09
1.41

0.05«

0.01

1.27

2.66
1.20
3.75"

0.53
1.31

0.30'
0.175
1.40

0.41
0.81
0.85<'

0.67
0.65

0.795'

0.84

0.71

CO Tfi r-

CO '^' '-H

■^ t^ CD
1 t^ CO CO

' IV

(X) (M .-H ,-(
(^ 00 -^ ■*

1 1 I

MM

1 1 1 1

1 1 1

1 M 1

1 1 1 1

12.0

14.2

9.5

1 1 1 1

210.5
210.5

207

I 1 1 1

05 (O CO
O CO t-^

1 05 c^ ]

--J c<i

1.52'

1.8

2.15

242

251.5

253.5

233.5
244

228.5'
227.5
241.5

OJ CO iC

r^ c^ o

05 d 00

7.89
5.44

t^ 00 CO

»C »0 Id

CO CO o^

00 (^ 00
iM C-^ (M

264.5
291

» ^- 1

a> Oi ^ \
CO lo 00

C<) Ol iM

00 Tt

ca. 0

9.9

>13

ca. 0.5

9.5

>13

a;
c

■§

1

c

1

o

„ E

P - = E

. £

a: C

.3 - ■»

^ =5 "•

3 .S O

c S —

3 S e

S > .o S .o

:: > s -

T3 » s., »

504

G. H. BEAVEN, E. R. HOLIDAY, AND E. A. JOHNSON

210

230

250

270

290

Wavelength, rriM
Fig. 7. Cytosine (Shugar and Fox'*).

as thymine, and to a much greater extent in subhmed films than in glasses.
Changes with aging of sublimed films are observed which suggest that these
are initially largely amorphous but slowly become crystalline, a process
accelerated by the presence of water vapor.

2. Nucleosides

Since ribosides may be obtained by hydrolysis of PNA with much greater
ease than the deoxy- derivatives from DNA, the greater part of the avail-
able spectroscopic data refers to the former. A good series of absorption
curves for deoxyribosides has been published by MacNutt,^^ who examined
these derivatives of guanine, hypoxanthine, thymine, cytosine, and uracil.
Manson and Lampen^^ give absorption curves for the deoxyribosides of

36 W. S. MacNutt, Biochem. J. 50, 384 (1952).

3« L. A. Manson and J. O. Lampen, J. Biol. Chem. 191, 87 (1951).

OPTICAL PROPERTIES OF NUCLEIC ACIDS

505

T f\

/

\

/

\

'

•--at pH 1

at pH 7

at pH 1

o

1 "^

\

-

4

/

\\

10

\
\

\\

V\

i

0

\\

1

\

1

1
1

\
\

1

1

1
1
1

i

1

6

1

1

\

1

1
1

'^^.J

1

1

\

\

1

\

4

1

\

\

/

//

/ /

/

1

1

\

1
1

1
\

\

N

1

1

1

1

1

\

o

I

t

V

\

1

1
1
1

1
/

\

\

\
\

\

/

(_^y

/

\

\

\

V

\

210

230

290

250 270

Wavelength, rriM
Fig. 8. 5-MethyIcytosine (Shugar and Fox'").

310

hypoxaiithine, thymine, and cytosine at neutral pH, and GuUand and
Story^^ compared guanosine and guanine deoxyriboside. For purine ribosides
Hotchkiss'" gives data for adenosine and guanosine, although in a form
quite unsuitable for general use without extensive arithmetical conversion,
and Kalckar^^ includes data for these two and for inosine. The differences
between ribo- and the corresponding deoxyribo- derivatives may be expected
to be very small (compare curves given by MacNutt^^ for guanine and
hypoxaiithine deoxyribosides with Figs. 12 and 13), but Fox and Shugar^®
have shown that, at least for pyrimidine derivatives, real distinctions can
be detected. The realization of the precision obtainable by the use of photo-
electric spectrophotometers has come very recently in some quarters, since

" J. M. Gulland and L. F. Story, J. Chem. Soc. 1938, 692.
38 H. M. Kalckar, /. Biol. Chem. 167, 429, 445 (1947).

506

G. H. BEAVEN, E. R. HOLIDAY, AND E. A. JOHNSON

\

1

at pH 7.2

~N

\

-

at pH 13

at pH 14

\

\

\

b

X

\

\

/

r\

\

UJ

>

\

X,

/

\

Q. 5

o

I \

/

\

pK^x

O

\

1

/

n

1

w

\

4

\

/

/

/

\

W

>

J

/

\

\\

\

V

\\

210 230 250 270 290

Wavelength, rriM
Fig. 9. Thymine (Shugar and Fox'*).

310

Kalckar^^ was apparently unable to distinguish adenosine from adenine
even though using such an instrument. Schlenk and co-workers^' give an
absorption curve for xanthosine at pH 7.5, and curves for synthetic xantho-
sine are given by Todd and co-workers.*" Even allowing for unsuitable pH
values the latter are in poor agreement with the results of Schlenk and
those given here, and also with the much older values of Gulland et al.,*^
all of which are reasonably consistent.

The incompleteness of the data available for purine ribosides seemed
again to require a more systematic investigation, the results of which are
given in Figs. 11 to 14, and in Table II. For the pyrimidine nucleosides
recourse was had to Fox and Shugar'^ (Figs. 15 to 18), again by far the most
complete set of data, and for 5-methylcytosine deoxyriboside to Dekker
and Elmore.*- Absorbance ratios are from Cohn wherever available. Other

" M. L. Schaedel, M. J. Waldvogel, and F. Schlenk, J. Biol. Chem. 171, 135 (1947).
" G. A. Howard, A. C. McLean, G. T. Newbold, F. S. Spring, and A. R. Todd, J.

Chem. Soc. 1949, 232.
*' J. M. Gulland, E. R. Holiday, and T. F. Macrae, J. Chem. Soc. 1934, 1639.
« C. A. Dekker and D. T. Elmore, J. Chem. Soc. 1951, 2864.

OPTICAL PROPERTIES OF NUCLEIC ACIDS

507

210

230 250 270

Wavelength, rriM
Fig. 10. Uracil (Shugar and Fox'*).

290

data on p^'rimidine nucleosides have been published by Hotchkiss,-" Loring
et al.,''^ and Lampen and co-workers."

The ionization in the pH 12-13 region observed in the free pyrimidine
bases is probably attributable to the acidic dissociation of the 2-hydroxyl
group, and in the purines to the acidic dissociation of the iminazole NH
group. In the nucleosides both of these are blocked, and the spectroscopic
changes in the same pH region can in this case be attributed to a sugar
hydroxyl dissociation. (See Jordan^ and Shugar and Fox^^'^^.)

The first demonstration that the sugar group in the naturally occurriiig nucleosides
was attached to the purine nucleus in the 9-position rather than in the 7- was carried

« H. S. Loring. H. W. Bortner, L. W. Levy, and M. L. Hammell, J. Biol. Chem. 196,

807 (1952).
" T. P. Wang, H. Z. Sable, and J. O. Lampen, J. Biol. Chem. 184, 17 (1950).

508

G. H. BEAVEN, E. R. HOLIDAY, AND E. A. JOHNSON

<

& a

O O

(N

lO 00

» 10

<a

a »0 10

y

o» «0

§8 1

c^ o

0 t^ CO

1 CO -H 05

lO 00 CO

r-i 0 0

O

^ '^'

o o

IC C^J 1-H

1 0 CO 10

10 <M lO

CD CO -(<

-*^

d d

odd

d d d

d d d

^' d d

^ d d

e«

(h

»

a

«

o o

lO ^

V

10

Cb

a iC a

O

s M

rt M< 1

rH IC 00

C5 r- ^

1 00 CO CO

0 CO I^

lO CO CI

a

o3

^ ^

(M r-H 1

^ (M 1-H

CD 0 0

1 C<l ^ rt

i-H C3 ^

r-H 0 0

O

d d

0(60

0 0 (6

d .-i r-n'

(N d i-i

CO 0 "H

ai

Cb

Xi

o o

V

«

10

ce <:» C2

<

s s

-t< 00 1

1-1 ao tJZi

■^ IC 05

1 »o 0 C^

10 CD CO

CO CO —1

'^ ^

00 t-- 1

(M CD 0

oi --I 00

1 r^ CO ^

'^f 00 00

tT 00 GO

d d

^ ^ ^

<6 ^ <6

0 i-H 1— I

d> d> d>

d d d>

10

kO 10

0

in u5 vc

o

CO oi

0 1-1 r-

t^ (^ CO

t-^ CO CD

^ 10 0

.-H CO 0

C)

-*' TjH 1

Oi t^ ^

1— i .— i >-H

1 00 i^ t^

CO t^ t^

CD t~- t^

1 1 i-H i-H

c^

IC

a

1 1 1

1 1 1

1 1 1

1 05 t^ 0

1 CO 1

1 1 1

J

1 1 1

CO (m' '*'

06

■^

0

.5

1 1 1

1 1 1

1 1 1

1 t^ (m' 0

1 ^ 1

1 1 1

a

1 1 1

1 1 1

1 1 1

' .-H (M CO

' C<1 '

/<

C^l (M C<1

iM

:;

1 1 1

1 ^ C^ CD

r-l CO 1

CO 1 1

J

1 1 1

1 1 1

1 1 1

' 00 d 00

d 06 '

d ' '

T— 1

Q

M
C/2

c3

"a

^

1 1 1

1 1 1

1 1

1 0 00 T-H

10 iC
d d 1

10

CO 1 1

O

►J

G

a

1 i 1

I 1 I

1 1 1

' CO -^ IC

^ (M '

1-H

'^

/<

(M (M <M

<M (M

CO

"0

a

p

w

-^

to

10 >o

iCi

10

CO

CO 0

^

c4

a

»0 (N 1

-^ -^ lO

^ cc 1-1

Tt< 0 10

t^ 10 0

10 --H t^

a

1

CO ci

<M CO (M

<n' c<i Tt<'

CO l> t^

,-H d CO

T-i d "5

-

lO

»o

10

IC »0 lO

lO

a

O t^ 1

.— 1 CO •*

00 CO ^H

1 00 -* (M

^ d — <

r-H 0 06

CO (M

<M (M !M

C<» (M CO

-r CD CD

-* lO 10

•^ in ^r

^

(M (M

(M C<) (M

(M C^ C^

(M <M (M

(M C^ (N

CO C^l CO

^

lO

iO

lO

00

x

CO 05 1

Oi IM ^

(M CO CO

1 Oi 0 CO

Tf .-1 (M

CO 0 CO

OS

a

-*' -^ '

d <N CO

(m' CO ^

' 06 00 Ci

CO 05 0

CO 0 Oi

<u

l-H

1— 1

1

CD

'^

»c

lO

lO IC CD
CO <M 1

»o

r- ci 1

.— 1 00 CO

1 CO 00 CD

0 —1 CO

0 r— 1 CO

a

iO lO '

10 •* 10

10 10 00

' CO t^ r-

00 t-- t^

00 (^ 1^

><

(N (M

iM IM (M

iM (M lO

(M IM (M

C<) CO IM

CO CO CO

CO

CO

CO

^

0 '-^

T—H

1— f

•*

(N t^

(M lO

. 10 .

^H

CO

5a.

• CO

53 ■ C

■ ts

■ e

CO T-H

^ 06

C^ 05

cj 0 CJ

rf 0

■^ u

-o

>>

X

c

0

3
O

s

o

o

■IS

a

■§

a

•a

■0

.2

0

c

13

.S "S

0 J2

OJ

1— 1

c3

c

V'

-*^ -r

T3

3

OS

>>

>> ^

<

0

X

0

0

OPTICAL PROPERTIES OF NUCLEIC ACIDS

509

3.55
0.93

0.235
0.16

0.03
0.02

Ml

3.4
1.45

0.72
0.67

0.35
0.29

CO CO 1

d d

0.45
0.95

0.65
0.75

0,74
0.83

0.72"
0.81"

(M O

CO CO 1

8.75
6.65
<6.5

9.95
7.35
<7.1

10.1
7.55
>7.65

M,

M,

Ml

1 1 1

1 11

111

1 1 1

1 11

9.78
13.38

9.55

Ml

1 1 1

05 CO 1

<M c5

207.5

11,

1 1 1

1.05
5.43

2.25
4.58

2.05
4.48

2.2
5.35

(N (M

235
240.5

230.5
236.5

^ (M 1
CO ^
(M (N

11.61

8.81

9.65
7.38

10.1

8,5

10.2

7.63

CO t-
CO t^ '

CO CO '

(M c^

CO CO '

C^ (M
CO CO '
(N iM

lO
■ CO

CO
00 —1

o A

9.25
>13

CO
CO >-H

d A

a oj

'to "^

o .-a

■5 ^-

U5

a
1

a

o

."2

X

o
o

-a

S ti

O

■n

•a

o

H

B

a

w

o

as

«

T3

03

h

h

C3

X

*z

_

a

T3

^

F

_!;'

S
s

J=

S

vt

1)

c

^

t-

E

U

Q

a

X

u.

X

£

a

cs

^

Z2

B

V

o

a

o

-a
a

5

Oj

s

c

n

M

-0

it

i£

p

'-^

p

fi;

<

510

G. H. BEAVEN, E. R. HOLIDAY, AND E, A. JOHNSON

16

>i 10

200

/^

-^

m(

DOeNHCI

4

1
1

L

\

in V

vater, pH 6 4

\
1
1

\
\
1

^

1

1

ll
1

— H

1

ll
ll
ll
i

if
ll

//

\ ^

■' /

\

\'\^

\^ ^

-~~~-—

220

240 260

Wavelength, m/i

280

300

Fig. 11. Adenosine (Johnson, unpublished).

Fig. 12. Inosine (Johnson, unpublished).

OPTICAL PROPERTIES OF NUCLEIC ACIDS

511

16

14

12

>; 10

200

— -in0.2NHCI

In water, pH 6

/

^

In NaOH, pH 11.3

1

\

/

V""

v^

\

\

\

/

V

\ 1

\

^ /

\\

\

\ \
\ \
\ \

\ \

\

/
/

"

'"--^

\ \

\ \

\ \
\ \

\
\

\

\

\ \

\

/-^'

L.//

\\

\

^

^.

1

\ \
\ ^

\

\

\

\

220

240

260

Wavelength, m/i

280

300

320

Fig. 13. Guanosine (Johnson, unpublished).

out b\- comparing their absorption spectra with those of the corresponding methyl-
substituted compounds (Gulland et oZ." •'"■■'*).

It may perhaps be pointed out here that the deduction of Press and Butler/*
on spectroscopic grounds, that the ring system of guanosine remains intact after
treatment with di(2-chloroethyl)methylamine seems ill-founded. From the data
given, there seems no reason to assume that the iminazole ring is unbroken, since
pyrimidine derivatives frequently possess absorption maxima at longer wavelengths
than purines.

3. Nucleotides

The separation of the phosphate group in a nucleotide from the effective
ultraviolet chromophore by the saturated carbon chain of the sugar residue
may be expected to ensure that its effect on the absorption characteristics
will be small. This is in fact the case, and many nucleotides may be regarded
as barely distinguishable from their parent nucleosides in the present state
of ultraviolet spectroscopic technique.

The phosphate group may be substituted in the 2'-, 3'-, or 5'-position in
ribonucleotides, and Markham and Smith" have recently shown that cyclic
2',3'-monohydrogen phosphate esters may also be isolated from enzymic or

^s J. M. Gulland and E. R. Holiday, /. Chem. Soc. 1936, 765.
" E. M. Press and J. A. V. Butler, /. Chem. Soc. 1952, 626.
" R. Markham and J. D. Smith, Biochem. J. 52, 552 (1952).

512

G. H. BEAVEN, E. R. HOLIDAY, AND E. A. JOHNSON

200

220

240

280

260
Wavelength, m/i

Fig. 14. Xanthosine (Johnson, unpublished).

300

alkaline digests of PNA. These have absorption spectra very similar to
those of the other isomers. In addition to the hydroxyl and amino pK's of
the bases and the sugar dissociation, further pX's are contributed in the
nucleotides by the phosphate dissociations. These would again not be ex-
pected to influence the spectrum greatly, but inspection of the absorbance
ratios in Table III for the uridylic and thymidylic acids shows that changes
occur between pH 2 and pH 7 which are not attributable to any other cause.
Separation techniques for the isomeric nucleotides have been worked out
only quite recently, and much of the spectroscopic work has therefore been
done on mixtures of unknown composition (Kerr et al.,-^ Ploeser and Lor-
ing,^' Chargaff and co-workers,^* Dunn and co-workers^'). Kalckar^* gives
data for adenylic acid-5', inosinic acid-5', and also ATP, and Kaplan et oZ.^"
give a curve for ITP. Data for cytidylic acids a(2') and 6(3') have been
published by Loring et al.*^ who found .that at pH 1 the 3'-isomer gave
results very close to those of cytidine. Fig. 1 shows the curves for the iso-
meric cytidylic acids (Cohn, private communication), and a comparison

*^ B. Magasanik, E. Vischer, R. Doniger, D. Elson, and E. Chargaff, /. Biol. Chem.

186, 37 (1950).
<' A. Deutsch, R. Zuckermann, and M. S. Dunn, Anal. Chem. 24, 1769 (1952).
" N. O. Kaplan, S. P. Colowick, and M. M. Ciotti, J. Biol. Chem. 194, 579 (1952).

OPTICAL PROPERTIES OF NUCLEIC ACIDS

513

<M

(M

_^

'"'

o

^111

'^ 1

t^ 1

1

■— '

(M

a

lO

I-- '

CO '

1

ffi

^

&

'"'

ft

^^

t^

(^

^

o

«5 -t* 1 1

c 1

^r 1

1

■^

o

ffi

lO

r^ t^ ' '

d '

^ i

1

w

^

o.

ft

""^

c^<

•o

"M

OO (M 1 1

o 1

-f 1

1

GO

■^

O O ' '

d '

00 '

ft

05

<M

I— 1

CO O CO o

-M

t^

iM

1—4

1

05 CO CO CO

o 1

—1 1

O

ft

-^

o

d> d> d> d

d

d

'-'

d

•^

t-

C5

O O CO o o-i

CO 1

~V 1

OJ

t^

00

o

W

o

(M CO CO CO CO

o 1

est 1

o

X

iM

s

a

d

d d d d d

d

d

,-H

d

'^

(M

00
CO

(M CO >0 lO 00

CO ■*

CO o

^

00 05 ^

w

o

CSi -^ lO lO -^

o o

(M C<l

Tt"

^^

■^ f lO

a,

d

d d

d o

CO

d d d

M

IM

T— t

lO

iC CO O Ci

ic 00 CO

t^

(M

o

OQ

a

^^

00 CI oi c» 1

C5 C^ CO 1

CO 1

lO

ft

CO

O

d

d d d d

d d d

d

-^

d

o

9

'^

1^

iC

iC CO Ci O CO

O lO o

CO

(M

t^

00

ffi

'^

00 C2 o CI Ci

CO CO -f

1- 1

lO

ffi

CO

"

ft

d

d d d d d

d d d

d

■—1

ft

d

o

'tl

,

.

CO

<:

(N

CO 0-1 iM

O O O M Q

00 (M 00 'f

(M O

iO

1—1

OO C5 o

ft

(M (M iM

00 O "— ' .— 1 o

(M CO CO CO

t^ 1^

CO

ft

CO CO t^

d d d

d> d d

^ oq c^i (N c<i

d d d d

d d

(M

(M

^-1

O O C5

O O ^ iM

lO CO c^

-f

CO

05

ft

GO 00 1-

C> 00 OO 00 1

00 00 GO 1

t^ 1

o

OO

d> di <6

d d d d

d d d

d

d

l-M
ft

d

'^'

t^

o o c»

O CO -t* O? 00

OO CO CO

»o

CO

t^

lO

ft

CO 00 t^

o 00 00 00 00

'- t^ f- 1

CO 1

Cl

ft

o

d d d

d d d d d

o o o

d

d

,-i

'^

•

C^

lo lo -r

OO lO CO CO CO

O CO -t' o

~r lo

CO

^

O CO Ol

ft

00 00 CO

■^ ■* -rf -r 'f

00 t^ t^ OO

CO CO

CO

d

ft

Oi O^ Oi

d d d

d d d

d d d d d

d d d d

d d

<u

CJ

u

cu

T^,

o

. . . ^^

,^

-o

S

Ph

-a

'-D

3

c^ CO lb

tidylic 2
tidylic 3
tidylic 5
oxycytid
tidine di

c<j CO lo ;g

.^ .2 .i CD

-o -a

oi CO lb

." ." ."

sec
a> « lU

73, 72 72 72

c c c

se c3 c3

-o -a -o

>>>>>, OJ >,

'C "u 'C 'C

J3 ^

3 3 3

<1 <1 -<

O O O P O

\:d^^^

H H

lO

COO

514 G. H. BEAVEN, E. R. HOLIDAY, AND E. A. JOHNSON

of Table II with Table III, also compiled from data kindly provided by
Cohn, shows that in general the nucleotide in which the phosphate is near-
est the base, i.e., the 2', shows the greatest differences from the correspond-
ing nucleoside, which might perhaps be expected. The curves for guanyhc
acid (Fig. 19) were determined in this laboratory. Although separation of
the isomers was not specifically attempted, the compound was purified by
ion-exchange chromatography and appears from the absorbance ratios to
consist very largely of the 3'-isomer. The very close resemblance to guano-
sine (Fig. 13) may be noted.

The absorption curves given for 5-methylcytidylic acid (Fig. 20, from
Cohn^^) appear to be the only ones so far available for a deoxynucleotide.
These are however relative, although a very close estimate of absolute
values may be obtained from the corresponding nucleoside.

III. Nucleic Acids and Polynucleotides
1. General

As is to be expected from their content of purines and pyrimidines, intact
nucleic acids show selective absorption in the X 260-mM region.

Spectrophotometry has been widely used for estimating the purine and
pyrimidine bases obtained from hydrolytic degradation of nucleic acids.
[Cf. Chapters 5-7, 10, and 11.] Reviewing in 1940 the small amount of prior
spectrophotometric work, Loofbourow" concluded that the absorptivity of
a nucleic acid was, to a close approximation, the sum of the contributions
of the constituent bases, assuming the tetranucleotide theory of their com-
position to be correct. The spectrophotometric analysis of an intact nucleic
acid would thus constitute a difficult four- or perhaps five-component analy-
sis. This does not prove possible, howeyer, since nucleic acids, both DNA^^*
and PNA, show lower absorptivities, by up to 35%, than the value cal-
culated from summing the molar absorptivities of the contained nucleotides
(as determined by separation and analysis) . Possible causes for this anomaly
are of interest in relation to theories of the structure of nucleic acids. The
reduction in total absorptivity may be regarded as arising from modifi-
cations of the chromophoric groups of one or more of the individual nucleo-
tide units in the macromolecule. Further, it appears that the absorptivity
value of a nucleic acid is very sensitive to its previous treatment and in
some respects may be more easily affected than other physical properties.
Although the results of investigations of these phenomena are beginning to
appear, it should be emphasized that complete explanation of the anomalous

" J. R. Loofbourow, Revs. Mod. Phys. 12, 320 (1940).

*'* The abbreviation DNA maj^ be understood to refer to the sodium salt where ap-
propriate.

OPTICAL PROPERTIES OF NUCLEIC ACIDS

515

1

1
at pH

1

/'

\

at pH /.Z -

atpH 14

1

/
/
/

\
\
\
\

12

\

/
/

\
\
\

X

1
1
1

\

\

10

\
\

V

1
1
1

1 J.

*C«v

\

\
\

^

■>

I\

8

\
\

\

\,

\

\

\

\
\

V

s

//\

1

w

\

6

\
\

\

1

1
1
1

\\

\

\

1
1
1

1

V\

1
\

4

\
\

1
1

1

\

\
\
\

\

/

1
1

\

\

2

V

^/

K\

\

y

\

\ ^

220

240

280

260

Wavelength, rriM
Fig. 15. Cytidine (Fox and Shugar'*).

300

ultraviolet absorption characteristics of nucleic acids is not possible at the
moment. The recent spectrophotometric studies by Cavalieri,^^ Shack
et a?.,^^'^^ Thomas, ^^ and Lawley (working in Jordan's laboratory),*^ indicate
that hitherto unsuspected factors may cause irreversible changes in the
absorption spectrum of DNA. These changes in the macromolecular struc-
ture we may term, from protein analogies, denaturation. Consequently,
many preparations and physical studies of DNA have been carried out
under conditions which have since been found to cause denaturation. In
the light of these findings the ultraviolet absorption characteristics of a

" L. F. Cavalieri, J. Am. Chem. Soc. 74, 1244 (1952).

" J. Shack, R. J. Jenkins, and J. M. Thompsett, /. Biol. Chem. 198, 85 (1952).

" J. Shack, R. J. Jenkins, and J. M. Thompsett, J. Biol. Chem. 203, 373 (1953).

" R. Thomas, Bull. soc. chim. biol. 35, 609 (1953).

" P. D. Lawley, Ph.D. Thesis, Nottingham University, 1953.

516

G, H. BEAVEN, E. R. HOLIDAY, AND E. A. JOHNSON

210

230

250

270

290

310

Wavelength, rriM

Fig. 16. Cytosine deoxyriboside (Fox and Shugari*).

nucleic acid acquire an importance comparable with the viscosity and the
sedimentation constant in the assessment of the macromolecular status of
the nucleic acid.

2. The Absorption Spectra of Nucleic Acids

Because of their varying content of water the molar absorptivities of
nucleic acids cannot reliably be based on weight. A more practicable con-
vention has been suggested by Chargaff and Zamenhof." Since nucleic
acids contain one phosphorus atom per nitrogenous base, these workers
relate the absorbance of a solution of nucleic acid to phosphorus concen-
tration and derive a molar absorptivity, e(P), for nucleic acid, based on
one gram-atom of phosphorus per liter.

" E. Chargaff and S. Zamenhof, J. Biol. Chem. 173, 327 (1948).

OPTICAL PROPERTIES OF NUCLEIC ACIDS

517

10

\

-at pH

12

V

\
\

-at pH

13

r

8

\

\
\
\

/

f

\

\
\

/

'■ — ,

\ \

6

\

\
\

/ /

1 /

\ \
\ ^
\

\
\
\

;

/ /
' /

/

\
\
\

4

^-1

r

1

\ \

1

\ \
\ \

2

w

\ \
\\

r-

\
\

210

230

270

290

250

Wavelength, rriM
Fig. 17. Thymidine (Fox and Shugar'^).

As defined by Chargaff and Zamenhof

6(P) = AlC-d

A and d have been defined above; C is measured in gram-atoms of phos-
phorus per Uter.

A review of pubUshed values of e(P) shows very wide variation, from
G,000 to 8,000 for DXA, and from 7,000 to 10,000 for PNA. When results
have not been expressed in accordance with this convention, comparison
with other work is often impossible. The €(P) value of a nucleic acid meas-
ured under standard conditions may eventually serve as an index of the
degree of denaturation.

Two general conclusions may be stated:

(a) All recorded values of €(P) for intact DNA and PNA are lower than

518

G. H. BEAVEN, E. R. HOLIDAY, AND E. A. JOHNSON

210

230

270

290

250

Wavelength, rriA
Fig. 18. Uridine (Fox and Shugar'«).

the values calculated from the summation of the absorptivities of the con-
tained purine and pyrimidine nucleotides.

(b) Values of e(P) show poor reproducibility even from the same prepara-
tion of nucleic acid.^'*"'^^ It is of interest therefore to find out at what level
of hydrolytic degradation the anomalously low e(P) of nucleic acid is abol-
ished.

Kunitz^^ had observed slight increases in the absorption maxima of
yeast PNA samples when depolymerized with ribonuclease, although here
the primary effect was a shift of the entire curve to shorter wavelengths.
Oster and Grimsson®" and Ogur and Rosen^^ found that other PNA Sam-
s' A. M. Marko and G. C. Butler, /. Biol. Chem. 190, 165 (1951).
69 M. Kunitz, /. Biol. Chem. 164, 563 (1946).
60 G. Oster and H. Grimsson, Arch. Biochem. 24, 119 (1949).
" M. Ogur and G. Rosen, Arch. Biochem. 25, 262 (1950).

OPTICAL PROPERTIES OF NUCLEIC ACIDS

519

260
Wavelength, m/i

Fig. 19. Guanylic acid (largely 3'-isonier) (Johnson, unpublished).

320

pies showed similar increases in absorption of from 5 to 15% on enzymic
degradation at neutral pH or on partial alkaline or acid hydrolysis. Ku-
nitz^- also first observed that depolymerization of DXA at pH 5 by de-
oxyribonuclease was accompanied by an ultimate increase in absorption at
2G0 m/x of nearly 30%, and increases of a similar order were reported for
both DNA and PXA by Tsuboi.^^

The relation between the absorptivity of the intact nucleic acids and the
value obtained by summing the absorptivities of the constituent nucleotides
was first pointed out, for a number of PXA samples, by Magasanik and
Chargaff." They found that the absorbance of the total alkaline hydrolysis
products was invariably greater by some 24-37 % than that of the original
sample, and that if, after partial enzymic hydrolysis of the PXA, the prod-
ucts were separated into dialyzable and nondialyzable fractions, the in-
crease in absorbance or "hyperchromic effect" on total alkaline hydrolysis
was almost negligible for the former and very marked for the latter. These
latter polynucleotide "cores" were especially rich in guanylic acid, and it
appeared that this hyperchromic effect might be especially associated with

«2M. Kunitz, J. Gen. Physiol. 33, 349 (1950).

63 K. K. Tsuboi, Biochim. et Bwphrjs. Acta 6, 202 (1950).

64 B. Magasanik and E. Chargaff, Biochim. et Biophxjs. Acta 7, 396 (1951).

)20

G. H. BEAVEN, E. R. HOLIDAY, AND E. A. JOHNSON

220

240

300

260 280

Wavelength, rriM
Fig. 20. Deoxy-5-methylcytidylic acid (Cohn'')-

320

the presence of this nucleotide, since the nondialyzable PNA cores obtained
by partial alkaline hydrolysis, which were rich rather in adenylic than in
guanylic acid, showed a much smaller hyperchromic effect than either the
parent PNA or the enzymic core.

In the case of still smaller fragments there seems to be no evidence of anomalous
absorption. Merrifield and Woolley** have confirmed that a number of dinucleotides
and a trinucleotide derived from DNA show absorptivities which are within experi-
mental error the sums of those of their consistent mononucleotides (see also Elmore
and Todd^^), and Sinsheimer and Koerner*' make the same claim for di-deoxyribo-

" R. B. Merrifield and D. W. Woolley, /. Biol. Chem. 197, 521 (1952).

" D. T. Elmore and A. R. Todd, /. Chem. Soc. 1952, 3681.

" R. L. Sinsheimer and J. F. Koerner, /. Am. Chem. Soc. 74, 283 (1952).

OPTICAL PROPERTIES OF NUCLEIC ACIDS

521

f(p)

200

\
\

--- pH 16
pH 7.0

\—

\
\

^

^.

— pH

20

\

K^/^

/x

y' \

uuu
000
,000
,000

\\

A

\
\

// /

/

/

\
\

s

/
/
1

N

\ \

\\

\

\

/~^-~ -

220

240

280

300

320

260
Wavelength, m/i

Fig. 21. Calculated absorption curves for a mixture of nucleosides containing
adenosine, guanosine, cytidine, and thymidine in the molar proportions 4:3:3:4. At
pH 1.5 values for adenine and guanine are employed.

nucleotides. More recentlj' Smith and Allen^* state that the absorptivities of poly-
ribonucleotides containing up to 20 nucleotide residues appear to be normal, and the
same may be concluded of the "apurinic acid" of Chargaff and co-workers, ^^ which,
however, can scarcely be regarded as a normal polynucleotide.

This absorbance anomaly therefore appears to be essentially a feature of
polynucleotides of relatively high molecular weight and probably denotes
some considerable degree of intramolecular organization. It is not due to
the binding of the nucleotides by glycosido-phosphate ester linkages. The
causes for it may therefore be of direct concern when considering theories
of the structure of nucleic acids.

The work reviewed below has been mainly concerned with calf thj^mus
DNA, which is perhaps the most readily accessible highly polymeric nucleic
acid. In general, the ribonucleic acids are far less well-characterized ma-
terials, as a class, than the deoxyribonucleic acids. For this reason, although
the causes of the absorptivity anomalies may be similar in both types of
nucleic acid, the results of studies of the effects of pH, cations, heat, and

«8 K. C. Smith and F. W. Allen, /. Am. Chem. Soc. 75, 2131 (1953).

69 C. Tamm, M. E. Hodes, and E. Chargaff, J. Biol. Chem. 195, 49 (1952).

522

G. H. BEAVEN, E. R. HOLIDAY, AND E. A. JOHNSON

10.000

8,000

m

6,000

4,000

2,000

200

---pH2.0
pH 7 0

—

— pH 12.3

.<-

cn

^f

/'^

N

\ \

\ \
\

— '

\;n\

:i^

\

\

^

V

220

240 260 280
Wavelength, m^

300

Fig. 22. Calf thymus DNA (Shack and Thompsett^") .

320

dehydration on the €(P) of PNA will be difficult to interpret until an un-
denatured form (Cf. Kay and Dounce^^) can be unequivocally recognized.

3. Effect of pH on the €(P) of DNA

In Fig. 21 the absorption curve of a mixture of nucleotides corresponding
in nitrogen base content to a typical DNA is shown at three pH values.
In Fig. 22 is shown the absorption curve of a calf thymus DNA preparation
under similar conditions. These curves clearly indicate that the marked
changes in absorptivity of DNA with change of pH cannot be attributed
simply to ionization of the purine and pyrimidine bases. The effects of
change in pH when summed for the four principal nucleotides in a nucleic
acid (Fig. 21) result in a small change in absorptivity and a slight shift in
wavelength, whereas there are gross changes for DNA itself under similar
conditions. Further it can be seen that on exposing DNA to acid or alkali
its absorption curve approaches more closely to the calculated one. It is
noteworthy, however, to find that at least part of the well-known anomalous
electrometric titration curve of DNA"" may also be obtained spectro-

" J. Shack and J. M. Thompsett, J. Biol. Chem. 197, 17 (1952).

'1 E. R. M. Kay and A. L. Bounce, J. Am. Chem. Soc. 75, 4041 (1953).

" J. M. Gulland, D. O. Jordan, and H. F. W. Taylor, /. Chem. Soc. 1947, 1131.

" W. A. Lee and A. R. Peacocke, /. Chem. Soc. 1951, 3361.

OPTICAL PROPERTIES OF NUCLEIC ACIDS

523

. Forward titration from pH 7 U hotometric

* Backtitration I

* Forward titration from pH 7 U|,c,rometric
» Backtitration '

8 9 10 11 12

pH

Fig. 23. Spectrophotometric and electrometric titration data for calf thymus
DNA (Shack and Thompsett'") . (a represents the proportion of the high-e(P) form
of the nucleate, or the proportion of nucleate anion.)

photometricalh^^^•^* Fig. 23, taken from a paper by Shack and Thomp-
sett,^" compares the electrometric and spectrophotometric titration curves
of DNA when titrated with alkaU. It can be seen from these results that
over the pH range 9-12 the increase in e(P) corresponds with ionization,
presumably of the hydroxyl groups of guanosine (pKa 9.5) and thymidine
(pKa 9.8). It is evident that the pK^ values of these groups are significantly
higher in the DXA macromolecule.

On the acid side of neutrality comparative electrometric and spectro-
photometric titration data at pH values lower than 3.5 are lacking and such
a correspondence does not emerge.

Fig. 24 (excluding curve 1) shows spectrophotometric titration curves
obtained in different laboratories ^^ 74.75 qj^ samples of calf thymus DNA in
absence of salts. These show that on titration of DNA with acid at the
lowest possible salt concentration, the e(P) of the nucleic acid begins to
rise at a pH as high as 8.0 and has almost attained its maximum value at
pH 5.0. A typical electrometric titration ".72,73 qj^ i\^q other hand shows that
in this pH region no ionic groups are being titrated. On the acid side of
pH 8.0, therefore, it appears that an increase in €(P) occurs, in the absence
of salts, at a higher pH than ionization. The effect of the addition of salts

''* G. Frick, Biochim. et Biophys. Acta 8, 625 (1952).

^* E. R. Blout and A. Asadourian, Biochim. et Biophys. Acta 13, 161 (1954).

524

11.000
10,000

G. H. BEAVEN, E. R. HOLIDAY, AND E. A. JOHNSON

Fig. 24. Variation of 6(P) of calf thymus DNA with pH.

(1) Calculated for a mixture of nucleosides containing adenosine, guanosine, cyti-
dine, and thymidine in the molar proportions 4:3:3.4. Below pH 2 values for adenine
and guanine are employed.

(2) Titration of calf thymus DNA in absence of salt (Blout and Asadourian''^).

(3) Titration in 1 M NaCl solution (Frick^^).

(4) Titration in absence of added salts.
(5a) in IQ-^ M MgSOi .

(5b) In 3.4 X IQ-^ M MgSOi .

(5c) In 0.1 M NaCl.

(5d) In M NaCl (Lawley^s).

is to decrease the pH value at which the increase in e(P) starts to occur
(see below). These observations indicate that there are two independent
processes, ionization on the one hand and a change of e(P) on the other.
On the alkaline side the process responsible for the changes in ^(P) and
ionization is concurrent with change in pH, whereas on the acid side, at
any rate in the absence of salts, they are independent.

OPTICAL PROPERTIES OF NUCLEIC ACIDS 525

The lowered absorptivity of the bases, resulting in the low €(P) of DNA
is most plausibly interpreted as due to a disturbance of the tautomeric
equilibrium of one or more of the bases. It is difficult to envisage any other
type of explanation: no exception to Lambert's law has been recorded so
that the lowering of €(P) cannot be accounted for on the basis of "optical"
interference between chromophoric groups (i.e., screening affects^^).

Purine and pyrimidine bases exhibit keto-enol tautomerism, and it has
been shown^-^ that free pyrimidines occur in a predominantly keto form,
although ionization to the anion may be presumed to involve an increase
in enolic character. [Cf. Jordan, Chapter 13.] The observation that the
electrometric and spectrophotometric titrations proceed concurrently in
the alkaline pH region suggests that hydrogen ions are being released from
the ionizable hydroxyl groups (with modified ^Ka values) of bases in a
predominantly enolic form. Such a degree of enolic character for the bases
in intact DNA would correspond, at least qualititively, to certain structures
for DNA proposed by Stern^^ and by Watson and Crick." In the structure
proposed by the latter, specific pairs of adjacent bases are hydrogen-bonded
through the 1,1- and 6,6-positions. [Compare Fig. 8 in Chapter 13.] The
effect of ionization is always to raise the value of €(P), but the salt effects
at constant pH indicate that the factors operating in the native macro-
molecule which are responsible for the absorption anomaly can also be
counteracted (in part reversibly) in other ways.

4. Effect of Salts on the Absorptivity of DNA

It has been known for some time that exposure of nucleic acids to ex-
tremes of pH, temperature, dehydration, etc. causes marked irreversible
changes in physical properties, including €(P). More recently the effects of
salts in particular on the €(P) value of DNA have been studied in several
laboratories.

Thomas^^-''^ has shown that exposure of DNA to low sodium chloride con-
centration (<10~' M) results in irreversible changes in its structure. The
evidence for this may be summarized as follows. A DNA preparation, which
has never been exposed to low salt concentration, extremes of pH, or other
denaturing influences, when dissolved in sodium chloride solution (con-
centration > 10~^ M) has an e(P) value in the region of 6000. Such a solution
is quite stable for several days as judged by the constancy of e(P) and by
the fact that increasing the concentration of sodium chloride does not lower
the e(P) value. However, a solution of DNA in which the concentration
of sodium chloride is less than 10~^ M has an €(P) value considerably higher
than 6000, reaching a value of 8000 at a concentration of ca. 10~^ M . When

76 K. G. Stern, Yale J. Biol. Med. 19, 937 (1947).

" J. D. Watson and F. H. C. Crick, Nature 171, 737, 964 (1953).

" R. Thomas, Biochim. et Biophys. Acta 14, 231 (1954).

526 G. H. BEAVEN, E. R. HOLIDAY, AND E. A. JOHNSON

the concentration of sodium chloride in such a solution is subsequently
raised, the «(?) value is reduced, but not to the value that would have been
found had the DNA not been previously exposed to the lower concentration.
It may be concluded that the irreversible increase in e(P) which occurs at
low salt concentration must be due to irreversible structural changes in the
DNA molecule. The possibility that such a change is due to contamination
by depolymerizing enzyme has been suggested by Gilbert et al?^ to explain
the spontaneous changes in the viscosity of DNA dissolved in distilled water.
In the light of these results some doubt exists about the interpretation
that can be placed on the careful work of Shack et al}^ on the effects of
varying metal-ion concentrations on the absorptivity of DNA. Since their
solutions were initially made up in distilled water, the DNA may have
been irreversibly changed to an unknown extent, though evidently not
sufficiently to obscure the effects under study. Fig. 25 contains information
derived from their paper,^^ showing the lowering of e(P) of DNA, initially
dissolved in distilled water, following the addition of monovalent cations.
It also demonstrates the much greater activity in this respect of divalent
cations. In this connection Lawley*^ has shown that cations compete for
the DNA anion in the order H+, Mg++, Na+, K+. Cavalieri^^ has shown that
adding a large excess of magnesium sulfate to a solution of sodium DNA
lowers the pH. Fig. 24 shows that in the presence of metal cations the pH
value at which the increase in «(?) commences is lowered. This figure also
includes Frick's data^'* in molar sodium chloride solution. On the alkaline
side of neutrality metal cations have very little effect. In Fig. 26 are shown
titration curves illustrating the combined effects of variations in salt con-
centration and pH on the e(P) value of DNA. The effect of exposure to low
salt concentration is to increase the c(P) value at pH 6, although the final
value at low pH is independent of this prior treatment. Denaturation in
this manner thus reduces the eventual increase in e(P) on acidification.
This behavior suggests itself as a useful test for the state of denaturation
of a given sample of DNA. It is dissolved in 10"^ M sodium chloride and its
absorbance determined at pH 3.0 and pH 7.0. The ratio {Ra) of these two
values may be taken as a measure of the extent of denaturation of the
sample. From Fig. 26, Ra is seen to be 1.30 for Thomas' undenatured
material and 1.12 for the material used to obtain the other two curves.
Both Frick'^ and Shack et al.^* have suggested that the c(P) value of a
DNA preparation is a good indication of its state of denaturation. The test
given here has the added advantage that phosphorus determinations are
unnecessary. It is evident from the results of this work that metal cations
play an important role in the stabilization of the structure of native DNA.

" L. Gilbert, W. G. Overend, and M. Webb, Exptl. Cell Research 2, 349 (1951).

OPTICAL PROPERTIES OF NUCLEIC ACIDS

527

10

AlAr

09

08

Mg-

as chloride

» Ca

• Na*^ )

o Na* as sulfate

• 8

°m

-4 -3 -2

Log equivalents cations per liter

Fig. 25. EfYect of metal cations on e(P) of calf thymus DNA. The ordinate scale
is the ratio of A, the absorbance in the presence of a given concentration of salt, to
.4o , the absorbance in the absence of salt (Shack, Jenkins, and Thompsett").

The nature of this role is still far from clear, but it is at least certain that
the "ionic" history of a DNA preparation is of great importance.

The fact that divalent cations such as Ca++ and Mg++ stabilize DNA at much
lower concentrations than do monovalent ions introduces an additional practical
complication. Many preparations of DNA may have been dialyzed against tap water
containing sufficient Ca++ ions to ensure stabilization. It may be expected that such
preparations will be more stable at low salt concentrations than DNA that has not
been so treated, and that, on subsequent dialysis against distilled water, they will
tend to lose any remaining Na"*" ions before Ca*"'" ions.

Certain observations which were difficult to explain are now readily accounted for
on the basis of the above findings. For instance it has been reported that the t(P) of
a preparation of DNA varies with the concentration, i.e., that Beer's law is not
obeyed, but that in the presence of salt «(?) is virtually constant over a wide range
of concentration.*^ •'*•*"

It has also been observed that the e(P) values of a given sample of DNA in dis-
tilled water solution are not reproducible."" Thomas'* has shown that the extent of

80 J. Pouyet, G. Scheibling, and H. Schwander, /. chhn. phys. 47, 417 (1950).

528

G. H. BEAVEN, E. R. HOLIDAY, AND E. A. JOHNSON

1.30

1.20

110

1.00

"Undenatured" preparation in lO""^ M NaCI
Preparation partially denatured by dilution to 10""'' M

NaCI before titration
Preparation partially denatured by dilution as above.

brought back to IQ-^M NaCI before titration

PH

Fig. 26. The variation of e(P) with pH of "undenatured" and partially denatured
preparations of calf thymus DNA (Thomas''*).

The ordinate scale is the ratio, Ra , of the observed e(P) value to the minimum
value (measured at pH 6 in 10"^ M NaCI solution) . The titration in each case is carried
out from the higher pH value.

denaturation in such solutions depends on the time lapse between the preparation of
the solution and examination.

A further protective effect of salts against heat denaturation of DNA solutions
has been observed by Thomas.'* -^^ For calf thjrmus DNA in sodium chloride solutions
of various concentrations, some denaturation occurs at room temperature at <10~^ M
and below 70° in 10"^ M. Denaturation is strongly inhibited even at 100° C. in 10"^ M
and M solutions. It is claimed that heat denaturation occurs stepwise and that the
critical temperature for each step varies for DNA from different sources. Thus DNA
from starfish testis in 10"^ M sodium chloride undergoes the first step in denaturation
at 55° C. compared with 70° C. for that from calf thymus.

The bearing of these results on theories of the structure of DNA may
become apparent when they are correlated with the study of other physical
properties. This may not be a simple matter because of the very different
nucleic acid concentrations at which spectrophotometric and other physical

81 R. Thomas, Trans. Faraday Soc. 50, 290 (1954).

OPTICAL PROPERTIES OF NUCLEIC ACIDS 529

measurements are made. The concentrations of DNA recjuired for elec-
trometric, conductimetric, and viscosimetric measurements may be one
hundred times greater than those required for spectrophotometric mea-
surements.^^ ■*-

5. Effect of Other Agents on the e(P) of DNA

Thomas^^ has shown that urea (6 M at 20° C.) is without denaturing
effect on DNA. Blout and Asadourian/^ working with a DNA stock solu-
tion in distilled water, have found that urea lowers e(P) slightly, whereas
plasma albumin and lysozyme have a marked effect in the same direction.
They also find that non-ionic compounds such as polyvinyl alcohol, sucrose,
and glycerol have no effect even in high concentration. It appears, however,
from Lawley's results^^ that ethanol increases «(?) : it is maximally effective
in 50% concentration while at higher or lower concentrations e(P) ap-
proaches its distilled water value. He has also shown that Ag+ has a specific
action on DNA, unlike that of other metal cations. Addition of silver nitrate
to a solution of DNA first lowers the e(P) value, this lowering being maxi-
mal for one equivalent of Ag^ per 4 atoms of DNA phosphorus. Further
addition causes a restoration of e(P) to the value for the DNA in distilled
water. The explanation for this is still obscure, but Lawley suggests that
the bases themselves are involved. It would be of interest to examine the
action of other heavy metals on the e(P) value of a nucleic acid. Such mea-
surements will be limited in many cases by the insolubility of the prod-
ucts.83.84

6. Spectrophotometric Estimation of Nucleic Acids

In general the spectrophotometric analysis of nucleic acids is carried out
by hydrolysis to free nucleotides, nucleosides, or bases, followed by their
separation and individual estimation, from which an estimate of the total
nucleic acid may be obtained. [Cf. Chapters 5-7, 9-11.]

However, methods for direct estimation of nucleic acids by spectro-
photometric means have been developed. Ogur and Rosen^^ by differential
extraction with perchloric acid have estimated DNA and PNA in corn
root tips and rabbit liver in this way. Tsuboi®^ and Logan et al.^^ have used
trichloracetic acid in a similar method. These procedures cause considerable
degradation of the nucleic acids which, however, does not result in any
net loss of absorbing material. It is, in fact, an essential feature of these
methods that the nucleic acids should be degraded so that the €(P) anomaly

"2 J. A. V. Butler and B. E. Conway, Nature 172, 153 (1953) .

S3 E. Hammarsten, Biochem. Z. 144, 383 (1924).

" K. G. Stern and M. A. Steinberg, Biochim. el Biophys. Acta 11, 553 (1953).

85 J. E. Logan, W. A. Mannell, and R. J. Rossiter, Biochem. J. 61, 480 (1952).

530 G. H. BEAVEN, E. R. HOLIDAY, AND E. A. JOHNSON

is completely abolished, thus enabling a direct correlation between ab-
sorptivity and concentration to be made. It is evident from Ogur and
Rosen's analyses that the c(P) of yeast PNA from different sources, after
such degradation, is remarkably reproducible, having a value of 10.800
(in A^" perchloric acid) with a standard deviation of 1.3%. Their correspond-
ing figure for thymus DNA is 8,780, with a standard deviation of 2.3%.
Such values of c(P) will vary with the purine and pyrimidine contents of
the particular nucleic acid, so that in order to apply the method generally
this e(P) value must be established in each case. [Cf. Chapter 10; Chapter
11.]

7. NUCLEOPROTEINS

The ultraviolet absorption spectrum of a nucleoprotein is composed of
contributions from both constituents and therefore differs from that of the
nucleic acid alone. The protein absorption,^^ which depends mainly on its
content of aromatic amino acid residues, is maximal at ca. X 280 m/x, and
rises again at shorter wavelengths from a minimum at ca. X 250 mju. On a
weight basis the peak absorption of protein will be only about one-tenth or
less that of nucleic acid, so that for a nucleoprotein containing, say, 40%
of nucleic acid, the absorption contribution of the latter therefore dominates
the collective absorption curve, and the contribution of the protein usually
appears as an inflection at ca. X 280 m^ while the nucleic acid minimum at
ca. X 230 m^ is shifted to longer wavelength. Since the nucleic acid content
of nucleoproteins varies over a wide range (e.g., 5-40% for some plant
viruses,^^ while the absorption of some of the proteins associated with
nucleic acids tends to be very low, the spectra of nucleoproteins will ob-
viously vary between rather wide limits; many examples can be found in
the literature.^^ '^^

Various suggestions have been made regarding the possible mode of combination
of the nucleic acid and protein moieties of nucleoprotein. Astbury'" pointed out that
the ca. 3.4-A. X-ray diffraction spacing of nucleic acids was comparable with the
side-chain spacing of a fully extended /3-type polypeptide chain and suggested that
this correspondence might allow salt-like electrostatic bonds to be formed between
the phosphate groups of the Astbury polynucleotide chain model and the guanidino
groups of arginine side chains, which form a large proportion of the amino acid
residues in basic proteins of the histone and protamine types. Support for this view

86 G. H. Beaven and E. R. Holiday, Advances in Protein Chem. 7, 319-386 (1952).
8^ J. P. Greenstein, Advances in Protein Chem. 1, 209 (1944).

88 R. Markham, R. E. F Matthews, and K. M. Smith, Nature 162, 88 (1948).

89 T. Caspersson, "Cell Growth and Cell Function." Norton, New York, 1951;
Symposia Sac. Exptl. Biol. 1, 127-151 (1951).

9» W. T. Astbury and F. O. Bell, Nature 141, 747 (1938); see also W. T. Astbury,
Symposia Soc. Exptl. Biol. 1, 66-76 (1951).

OPTICAL PROPERTIES OF NUCLEIC ACIDS 531

comes from the analytical data of Vendrely and Vendrely," who find a constant DNA-
arginine ratio in the erythrocyte nuclei of many fish species which show large varia-
tions in DNA content. Some interesting observations on the composition and solu-
bilitj' of synthetic DNA-protamine complexes in relation to natural nuceloproteins
are discussed by Alexander." Stern'' has found that model polynucleotide and poly-
peptide chains can interlace without straining of bond angles to allow the formation
of salt linkages, as envisaged by Astbury, in such a waj- as to lock the polynucleotide
chain in a specific configuration. In this connection it should perhaps be noted that
free DNA preparations from different species show the same structure independent
of the base constitution.'*

Riley and Arndt'* have concluded from quantitative X-ray scattering studies of
herring sperm and calf thymus nucleoproteins and of the constituent nucleic acids
and proteins (clupeine and histone, respectively) that a nucleoprotein is best regarded
as a fairly gross addition complex, presumably held together by salt linkages, rather
than an intimate and specific association of the two components, for example, a fully
extended /3-type polypeptide chain wound around the double-helix polynucleotide
chain, as envisaged by Watson and Crick."

It is not yet possible, on the basis of these conflicting views on the nature of the
protein-nucleic acid association, to make any prediction about the additivity or
otherwise of the absorptivity contributions of the two components of a nucleoprotein.
Some experimental data are available. Shack and co-workers'"'^ find that the anom-
alous spectrophotometric titration of thymus DNA is also shown by the related
nucleohistone and by one isolated from a transplantable mouse lymphoma. They
conclude that the same alkali-labile hydrogen-bonded structure is present in a nucleo-
protein and the derived nucleic acid.

Blout and Asadourian" have measured the absorbance of calf thymus DNA solu-
tions containing added protein differentially against control solutions of protein
and find that both lysozyme and bovine plasma albumin lower the DNA absorbance;
in the presence of sodium chloride, the lowering is additional to the effect of the salt
alone. These workers recognize three factors which determine the absorptivity of a
given solution of DNA, (1) pH, (2) a "small ion" effect, and (3) a "macromolecular
ion" effect. The first two factors have been discussed above; the third is supposed by
Blout and Asadourian to operate in the same was as (2), viz., by decreasing the
interaction (of unspecified character) between neighboring units in the polynucleotide
chain. Non-ionic macromolecular substances such as polyvinyl alcohol are without
effect.

Brachet is reported " to have evidence that the absorption of mixtures of thymus
DNA and lysozyme (but not histone) is not additive.

Seiberf has found deviations from additivity for the total absorbance of mixtures

" R. Vendrely and C. Vendrely, Nature 172, 30 (1953).

92 P. Alexander, Biochim. et Biophys. Acta 10, 595 (1953).

" K. G. Stern, The chemistry and phj'siology of the nucleus, Exptl. Cell Research,

Suppl. 2 (1952).
" M. H. F. Wilkins, A. R. Stokes, and H. R. Wilson, Nature 171, 738 (1953).
'* D. P. Riley and U. W. Arndt, Nature 172, 294 (1953).
'6 J. Shack and J. M. Thompsett, J. Natl. Cancer Inst. 13, 1425 (1953).
" H. G. Davies and M. P. B. Walker, Microspectrometry of living and fixed cells,

Progr. Biophys. and Biophys. Chem. 3, 195-236 (1953).
" F. B. Seibert, The physical chemistry of proteins. Discussions Faraday Sac. 13,

251 (1953).

532 G. H. BEAVEN, E, R. HOLIDAY, AND E. A. JOHNSON

of tuberculin proteins and thymus DNA over a wide range of pH. At low pH values
the measurements are complicated by opalescence, but above pH 5 the differences
between the calculated and measured mixture absorbances are much greater than
could reasonably be ascribed to spectrophotometric errors, and indicate that the
total absorbance is reduced in the mixture, deficits of 10% or more being observed.

The limited experimental work on this subject does not reveal which
component of a nucleic acid-protein system gives rise to the absorbance
deficit, or if both components are affected. In particular it is not yet known
if combination with a protein has the same effect as Mg++ or Na+ ions (in
appropriate concentrations) in restoring the low e(P) value characteristic
of the native state to a DNA preparation that has previously been partly
denatured (with consequent increase in €(P)) by exposure to low salt con-
centration, etc. Clearly there is scope for accurate spectrophotometric
studies on nucleoproteins in relation to the absorption properties of their
constituent proteins and nucleic acids and the effects of denaturation.

IV. Ultraviolet Dichroism

The dichroism of the ultraviolet absorption of oriented fibers and sheets
of high-molecular-weight nucleic acid may be used to obtain information
concerning the orientation of the purine and pyrimidine rings with respect
to the long axis of the macromolecule. This axis may be taken as parallel to
the fiber axis or direction of shear in specimens which have been oriented
by stretching or shearing, respectively. Dichroism measurements may also
be used to study the state of nucleoprotein as it exists in intact cells. The
possibility that the dichroism of nucleic acid in biological material may
give rise to serious errors in the microspectrometric estimation of nucleic
acid has been the subject of much recent comment. For these reasons it
has seemed advisable to review. briefly the present status of dichroism
studies, although it must be emphasized that a full treatment would require
an extensive development of the underlying principles of crystal optics.
Here it is only possible to give references to some standard texts on the
subject and to state the appropriate concepts and equations without de-
tailed exposition. ^''^"^

89 H. Ambronn and A. Frey, "Das Polarisationsmikroskop." Akad. Verlagsgesell-

schaft, Leipzig, 1926.
1"" C. W. Bunn, "Chemical Crystallography," 1st ed. Oxford University Press,

London, 1945.
i"! N. H. Hartshorne and A. Stuart, "Crystals and the Polarising Microscope,"

2nd ed. Edward Arnold, London, 1950.
102 E E. Jelley, Microscopy, in "Physical Methods of Organic Chemistry," Vol. 1,

Part 1 of "Technique of Organic Chemistry," (A. Weissberger, ed.), 2nd ed. In-

terscience Publishers, New York, 1949.

optical properties of nucleic acids 533

1. General Theory

The nature of an electronic transition in a molecule caused by the ab-
sorption of radiation of appropriate frequency requires that for maximum
probability the electric vector of the radiation should have a definite di-
rection with respect to the chromophoric group of the molecule. For a
linear conjugated system, such as a polyene, the electric vector must be
parallel to the direction of the conjugated multiple bonds, while for a planar
conjugated system, of which benzene is the simplest example, the electric
vector must be in the plane of the ring. The absorption is greatest when
these conditions are satisfied. If the electric vector is perpendicular to the
plane of the ring the transition probability, and therefore the intensity of
absorption, is reduced, though not necessarily to zero, because of inter-
action between the two modes of excitation. These polarization properties
apply both to allowed transitions and also to the much weaker transitions,
e.g., that giving rise to the 260-m)u benzene band, which acquire allowed
character because of the disturbing effect of simultaneous changes in the
vibrational states. If the ring chromophore is unsymmetrical, the strength
of the transition may also depend on the direction of the electric vector
in the plane of the ring.

The intensity of absorption of polarized radiation by a crystal or oriented
specimen will therefore depend on the orientation of the electric vector of
the radiation with respect to the absorbing groups of the constituent mol-
ecules, and hence in the case of a crystal, with respect to the crystallographic
axes. For an oriented fiber or sheet, the axis of reference will be the fiber
axis or shear direction, as noted above. The crystal or specimen will there-
fore require more than one absorptivity to express its absorption properties,
and may be described as dichroic (or trichroic if three such coefficients are
required). These are known as the principal absorptivities and can be re-
ferred to specific directions with the crystal or system. Since the anisotropy
of refractive index (birefringence) for unabsorbed light in a region of trans-
parency is also a conseciuence of the molecular orientation of the specimen,
the directions of the principal refractive indices, and hence of the extinction
directions, will be related to, and for uniaxial orientation, will coincide
with, the directions of the principal absorptivities. The birefringence and
dichroism are not necessarily parallel indications of the orientation of the
same groups in a molecule since in a region of transparency the birefringence
is the sum of contributions by all the bonds in the molecule whereas in an
absorbing region the dichroism is a characteristic of the absorbing groups
only.

Absorption measurements in solution refer to randomly oriented mol-
ecules (except for the special case of solutions of long-chain macromolecules

534 G. H. BEAVEN, E. R. HOLIDAY, AND E. A. JOHNSON

oriented by flow) in which the principal absorption directions are in all
possible directions with respect to the electric vector of the (nominally)
unpolarized incident radiation. In order to measure the principal absorptiv-
ities it is necessary to use crystals or oriented specimens and plane-polarized
radiation. Ideally, the arrangement of the molecules in a crystal with re-
spect to the crystallographic axes should be known from an X-ray structure
determination, but even if this information is lacking the dichroism alone
can be used to obtain information about the probable orientation of the
absorbing groups within the crystal.

In practice absorption measurements on crystals or oriented material
usually entail the use of very small specimens. This limitation arises partly
because of the difficulty in preparing large crystals or specimens (or the
inherent size of specimens of biological origin), and partly because of the
high absorbances that would be encountered with thick samples. For this
reason, it is necessary to use some form of fully achromatic reflecting micro-
scope in conjunction with the spectrophotometric equipment. With regard
to crystals, the number of compounds in which the arrangement of the
molecules within the crystal lattice is known from X-ray structure data to
be favorable for the study of pleochroism is very small. The total volume
of work in the field is therefore limited, though the introduction of various
designs of reflecting microscope has led to increased activity.

Seeds^"' has reviewed the polarized ultraviolet microspectrography of
crystals and oriented systems of biological interest, and this work should
be consulted for experimental details. General microspectrophotometry
has been dealt with in considerable detail by Seeds, Wilkins, Barer, Davies,
Mellors, Walker, Commoner, and other contributors to a recent sym-
posium^"* which contains much information on reflecting optics and micro-
spectrophotometric systems, together with typical results on a wide variety
of biological material.

2. Form Dichroism

Form anisotropy, both birefringence and dichroism, is encountered in
composite systems containing oriented assemblages of macromolecules,
even if the latter are themselves isotropic, when the systems contain amor-
phous regions between the ordered ones. The fonn anisotropy is additional
to any intrinsic anisotropy that the oriented macromolecules may exhibit
and can therefore make an important contribution to the total observed
anisotropy of a system in which the conditions for form anisotropy are
satisfied. The two models which have been treated theoretically by Wie-

'<" W. E. Seeds, Polarized ultraviolet microspectrography and molecular structure,

Progr. Biophys. and Biophys. Chem. 3, 27-46 (1953).
1"* Spectroscopy and molecular structure and optical methods of investigating cell

structure, Discussions Faraday Soc. No. 9 (1950).

OPTICAL PROPERTIES OF NUCLEIC ACIDS 535

ner^"* are isotropic rods arranged in parallel bundles and flat plates in paral-
lel stacks, separated in both cases by an amorphous component of different
refractive index. Wiener's equations apply strictly to systems in which
the volume concentration of the ordered phase is small and is therefore
not accurate for, e.g., a slightly swollen fiber, in which the ordered phase is
a large proportion of the total volume. It has been stated^°^ that a more
accurate theory for hexagonally packed rods is being prepared. Examples
of the use of the Wiener formulas have been given by many authors'"^"^"^
in connection with the optical properties of natural cellulosic fibrous struc-
tures.

Form anisotropy can be detected by changing the refractive index of
the amorphous component (by immersion in liquids of various refractive
indices) and observing the effect on the birefringence. At a particular re-
fractive index of the immersion li(iuid the form birefringence (positive for
rod model, negative for lamellar model) falls to zero, and any residual
birefringence is then intrinsic, due to anisotropy of the macromolecules in
the ordered phase.

A good example of this behavior was given by Weber"" (quoted by Doty
and Geiduschek'") for artificial gelatin and myosin fibers; both showed
positive l)irefringence, but, at the critical refractive index, the residual
birefringence of the gelatin fiber was zero, while for the myosin fiber a large
intrinsic birefringence persisted. Some observations on nucleoproteins and
nucleic acids in which form anisotropy may have been involved will be
noted later in this section.

3. Definitions and Equations'"'

A linear absorbing system is characterized by two absorptivities, ay and
a± for the electric vector parallel and perpendicular, respectively, to the
direction of the absorbing system. Then

Dichroism = an — a±

Dichroic Ratio (D) = a\\/ax

In general a biaxial crystal will have three absorbances (Ax , A„ , A^)
corresponding to the principal orthogonal axes of the triaxial absorption

'"« O. Wiener, Abhandl. math.-phys. Kl. sacks. Akad. Wiss. (Leipzig) 32, 507 (1912)-,
KoUoidchem. Beih. 23, 189, 198 (1926).

»»» M. H. F. Wilkins, A. R. Stokes, W. E. Seeds, and G. Oster, Nature 166, 127 (1950).

I" F. O. Schmidt, Advances in Protein Chem. 1, 25-68 (1944).

'"* A. Frey-Wyssling, "Submicroscopic Morphology of Protoplasm," Elsevier, Ams-
terdam, London, New York, Houston, 2nd English Ed., 1953.

»»» A. Frey, Jahrb. wiss. Botan. 67, 597 (1927).

"« H. H. "Weber, Pfliigers Arch. ges. Physiol. 235, 205 (1934).

'» P. Doty and E. P. Geiduschek, in "The Proteins" (H. Neurath and K. Bailey,
eds:). Vol. 1, part A. Academic Press, New York, 1953.

536 G. H. BEAVEN, E. R. HOLIDAY, AND E. A. JOHNSON

ellipsoid. These are related to the non-oriented value (A') for the same
compound by the equation

A' = {Ax + A^ + At)/S (biaxial case)

or

A' = i2Ax + A||)/3 (uniaxial case)

The calculated value of A' can be used to estimate the thickness of the
crystal specimen by applying the Beer-Lambert absorption laws to solution
data for the same compound, assuming that the specific absorptivity (a or
e) is identical in the crystal and in solution (with identical concentration
units) .

Except in certain favorable cases (e.g., hexamethylbenzene, see below)
the absorbing groups in crystals and oriented materials are not all parallel
to each other, but may be in arrangements in which the dichroism of one
set of parallel groups is neutralized, to a greater or lesser extent, by the
opposing dichroism of other sets of parallel groups. Consequently the ob-
served dichroic ratio is often much smaller (2 or less) than would be ex-
pected from simple theory. As emphasized by Seeds^"' the interpretation
that can be given to such low dichroic ratios depends on making a choice
between two assumptions. The system under study can be assumed to be
partly disoriented, so that the observed low dichroism is taken as some
measure of the degree of orientation. Such an assumption is obviously not
valid for crystals proper, for which, in any case. X-ray data might be avail-
able to indicate that the crystal lattice is unfavorable. For artificially ori-
ented materials of natural or synthetic origin, imperfect orientation is
much more likely, although errors arising from it can be minimized in
microspectrographic procedures by confining the measurements to the
specimen areas with the highest birefringence. It may then be assumed
that the specimen is fully oriented and that the observed dichroism is a
measure of the average angle of tilt of the absorbing groups in the molecule
to the principal axis of the specimen. If it is further assumed that the di-
chroic ratio for a planar ring is very large and that dichroism in the plane
of the ring is small, the variation of D with angle of tilt for some simple
model structures can be calculated.

For the important practical case of a uniaxial fiber with rotational sym-
metry, in which the normals to the planes of the absorbing groups lie on a
cone of semi-angle 6 generated about the fiber axis, Fraser"^ has shown that

Dfiber = Sin2 e/l - i Sin« 0

"2 R. D. B. Fraser, Ph. D. Thesis, London University, 1951, quoted by Seeds. '"'

OPTICAL PROPERTIES OF NUCLEIC ACIDS 537

For oriented sheets

I>sheet I = Sin^ ^/COS^ e
I>sl,eet II = Sin2 e

Sheet model type I is for the case in which the normal to the planes of the
absorbing groups makes an angle 6 with the direction of shear and lies in a
plane normal to the direction of the incident radiation. In the type II model
the normal to the absorbing planes lies at an angle 6 to the direction of
shear and in the plane of incidence of the radiation.

By using these equations some indication of the mean orientation of the
absorbing groups can be obtained. Since observations on crystals with
favorably disposed absorbing groups indicate that absorption is not van-
ishingly small when the electric vector is perpendicular to the ring plane,
these simple relationships have to be modified if it is required to estimate
d from very large or very small dichroic ratios.

4. DicHROisM OF Simple Compounds

A much-quoted demonstration of the dichroism of the benzene chromophore is due
to Scheibe, Hartwig, and Miiller using hexamethylbenzene.ii'This compound crystal-
lizes with all the planar molecules parallel to one another and can be obtained as
crystalline sheets or sublimed crystalline films in which the planes of the benzene
groups are roughly normal to the specimen plane. Such a preparation has a dichroic
ratio of ca. 10 over most of the absorption band with X^ax. ca. 270 niju (corresponding
to the well-known "forbidden" benzene band at ca. 255 van). Scheibe et al. also find
that the wavelengths of the vibrational fine-structure bands are slightly different
for the electric vector parallel and perpendicular to the ring planes.

Craig and Lyons"^ have extended these observations to the 230-mM band, which
is also polarized in the molecular plane, the dichroic ratio for this band being of the
same order. This result, which is applicable by analogy to the benzene "allowed"
system of ca. 200 m^, is of great importance for current theoretical arguments about
the assignment of the benzene 200-mM system and the nature of the molecular orbitals
involved.

Dichroism studies on the particular pyrimidines and purines found in nucleic acids
have not yet been reported, probably because of difficulty in obtaining suitable
crystal specimens and the lack of X-ray structure data. Seeds'o^.ne.ne has studied
2-hydroxy-4,6-dimethylpyrimidine (I) for which favorable X-ray data are available.

Me Me Me

y\ ./\ .^

N
ill

N N

HO^N/Me HoN'^j^t/C1 C^n/^®

I II III

"3 G. Scheibe, St. Hartwig, R. Muller, Z. Elektrochem. 49, 372 (1943).
"" D. P. Craig and L. E. Lyons, Nahire 169, 1102 (1952).
"6 W. E. Seeds, Ph.D. Thesis, London University, 1951.
"6 W. E. Seeds, Discussions Faraday Soc. No. 9, 394 (1950).

538 G. H. BEAVEN, E. K. HOLIDAY, AND E. A. JOHNSON

Absorption measurements with the electric vector in and perpendicular to the ring
plane show very marked dichroism; D approaches 100 at 300 mn where the perpendic-
ular absorption is very small, while the in-plane absorption extends to a much longer
wavelength. The dichroism decreases with decreasing wavelength and at 250 m^ D
has fallen to ca. 2. By a special crystallization method Seeds has also been able to
make observations with the electric vector in the ring plane and parallel or perpendic-
ular to the pseudo-dyad axis through C2-C6 and finds a small but definite dichroism
below 300 m/i.This work appears to be the first in which the three absorption spec-
tra of a crystal have been measured and dichroism in the ring plane demonstrated;
Craig and Lyons''^ have since demonstrated that in naphthalene the transition
dipole moment for the weak transition at ca. 310 m/x lies predominantly in the mo-
lecular plane and along the shorter molecular axis.

Lyons"* has examined 2-amino-4-chloro-6-meth3'lpyrimidine (II), for which X-ray
structure data are available, and 2-chloro-4,6-dimethylpyrimidine (III). Of the three
band systems generally shown by pyrimidines^' the weak longwave band (a) is ob-
scured (though not in III) by system (b) located at 250-290 m/i; the third sj'stem (c)
lies at 200-250 m/i. Band (a) disappears in acidic solution and is a transition of the
n— n- type'^' involving non-bonding electrons on the nitrogen atoms while bands (b)
and (c) are due to tt-tt transitions corresponding to the 260-mM and 200-m/i benzene
systems. For the type (b) band of II the electric vector is in the molecular plane for
strong absorption in agreement with the results for I and there is a favored direction
(x) in the plane parallel to a line through the two nitrogen atoms. For III, with un-
known crystal structure, the polarization of bands (a) and (c) are similar and different
from that of band (b). Since by analogy with hexamethylbenzene band (c) of II and
III will be polarized in the molecular plane, their polarization direction in the plane
must be mainly y, at right angles to that of the (b) band. From these findings Lyons
is able to give assignments to the type (b) and (c) systems and to draw some important
conclusions about the nature of the upper state of the type (a) n—n- transition.

5. Dichroism of Nucleic Acid

The optical anisotropy of biological materials is the subject of an ex-
tensive literature (see review^-") mainly concerned with birefringence stud-
ies. Sperm heads have long been known^-^ to show a large optical aniso-
tropy, indicating parallelism of the nucleoprotein molecules, and it has
very recently been shown by X-ray diffraction studies that this anisotropy
arises from a true three-dimensional crystallinity.^^- The birefringence of
cytoplasm has been the subject of much work, but it has recently been
pointed out^-" that form and intrinsic birefringence have not always been
distinguished, while the sign of the birefringence is certainly not sufficient

"7 D. P. Craig and L. E. Lyons, /. Chem. Phys. 20, 1499 (1952).

"s L. E. Lyons, /. Chem. Phys. 20, 1814 (1952).

113 M. Kasha, Discussions Faraday Soc. No. 9, 14 (1950).

12" M. M. Swann and J. M. Mitchison, Birefringence of cytoplasm and cell membranes,

Progr. Biophys. and Biophys. Chem. 2, 1-16 (1951).
121 W. J. Schmidt, "Die Doppelbrechung von Karyoplasma, Zytoplasma und Meta-

plasma." Borntraeger, Berlin, 1937.
1" M. H. F. Wilkins and J. T. Randall, Biochim. et Biophys. Acta 10, 192 (1953).

OPTICAL PROPERTIES OF NUCLEIC ACIDS 539

to distinguish oriented protein from oriented nucleic acid. The high*"*
negative birefringence of extracted high-molecular-weight thymus DNA
solution, oriented by flow, was immediately interpreted by Signer, Caspers-
son, and Hammarsten*-' as indicating that the absorbing groups were ori-
ented with their ring planes perpendicular to the long axis of the macro-
molecule, and this conclusion was utilized by Astbury and Bell'" in their
model of the DNA molecule in which the nucleotide residues were stacked
parallel to each other along one side of the long molecular axis. [Cf. Jordan,
Chapter 13.]

The negative ultraviolet dichroism of a thymus DNA film oriented by
stretching was first demonstrated by Caspersson,*^* whose results show a
dichroic ratio at the 260-mM absorption maximum of ca. 1.6 and also good
agreement between the observed absorption curve for unpolarized light
and that calculated from the values obtained with the electric vector paral-
lel and perpendicular to the direction of stretch. No dichroism was observed
with yeast PNA films, in agreement with indications from other techniques
that this material was unoriented. It has usually been considered that
because of its lower molecular weight and possibly because of structural
differences, PXA will not show molecular orientation to the striking extent
found with DNA. However, PNA from calf and rat liver, prepared with
precautions against depolymerization,*" shows flow birefringence (about
one-tenth of the value for highly polymerized DNA) and forms birefringent
fibers which are elastic when undried (cf. sodium DNA fibers, see below).
It seems likely that, when the problems of preparing undenatured PNA
have been solved, this class of nucleic acids will also lend itself to studies of
optical anisotropy.

Seeds*"* "^'^^^ has obtained dichroic ratios of 1.7-2 for air-dried {ca. 30%
water) films of thymus DNA oriented by shearing a viscous gel, the di-
chroism being roughly constant over the entire absorption band. Obser-
vations over a range of controlled humidity show that the dichroism in-
creases with increasing humidity up to 90%, above which the specimen
becomes unstable to irradiation and changes eventually to an isotropic
form. The highest dichroic ratio recorded by Seeds is 4.7 at the 265-mM
maximum for a film in air at 93 % humidity ; the average value for air-

122a Pq^ oriented fibers with rotational symmetry about the long axis and for their
films oriented by shearing, etc.. the sign convention for birefringence and dichro-
ism is that a positive fiber has its greater refractive index (or absorbance) along
the fiber length. When the greater refractive index (or absorbance) is perpendicu-
lar to the fiber axis the sign is negative, as in nucleic acid. (Cf. ref. 101, pp. 148, 441).

^" R. Signer, T. Caspersson, and E. Hammarsten, Nahire 141, 122 (1938).

'" T. Caspersson, Chromosoma 1, 605 (1940); cf. J. P. Greenstein, ref. 87.

1" E. L. Grinnan and W. A. Mosher, J. Biol. Chern. 191, 719 (1951).

'26 W. E. Seeds and M. H. F. Wilkins, Discussions Faraday Sac. No. 9, 417 (1950).

540

G. H. BEAVEN, E. R. HOLIDAY, AND E. A, JOHNSON

1 1 1 \ 1

— Electric vector perpendicular to direction of shear
• — Electric vector parallel to direction of shear

240

250

290

260 270 280

Wavelength, rriM
Fig. 27. Ultraviolet absorption spectra of oriented film of thymus DNA sodium
salt: (+) at 90% humidity, (O) at 60% humidity (Seeds 'o^, ii6).

dried oriented DNA is ca. 2, increasing to nearly 4 at 90 % humidity, (Fig.
27.) The change in dichroic ratio with humidity can be interpreted, using
the equations for model systems as a rotation of the planes of the absorbing
groups, the angle obtained depending on whether the uniaxial fiber or type
II sheet model is selected. Seeds concludes that the bases are so arranged
that, on the average, they lie at a small angle to the normal to the mo-
lecular axis and that this angle decreases with increasing humidity; the
actual structure is probably intermediate between the two model structures
considered.

The influence of humidity on the anisotropy of oriented DNA fibers has
been studied in greater detail by Wilkins, Gosling, and Seeds,^" who rec-

1" M. H. F. Wilkins, R. G. Gosling, and W. E. Seeds, Nature 167, 759 (1951).

OPTICAL PROPERTIES OF NUCLEIC ACIDS 541

ognize two distinct forms with strikingly different properties:
Type I (stable at high humidity) with :

Negative birefringence

Negative ultraviolet dichroism

Mainly negative infrared dichroism

X-ray diffraction pattern of well-ordered crystallites
Type II (stable at 50% humidity) with:

Positive birefringence

No ultraviolet dichroism

Mainly positive infrared dichroism

Diffuse-X-raj'- diffraction pattern
Type II can be obtained from a type I fiber by stretching up to nearly
100 % elongation in air of 50 % relative humidity. Reconversion to type I,
with appreciable contraction, occurs on exposure to higher humidity.
Wilkins ct al. consider that the dimensional change associated with hy-
dration is due to conversion from amorphous to the crystalline state and not
to water molecules packing between the crystallites. On dehydration it is
suggested that the molecular backbone of phosphate ester linkages crumples
and the bases tilt with consequent destruction of the ultraviolet dichroism
(and reversal of birefringence), while in the hydrated I-form the extended
molecules pack together regularly with the planes of the bases approx-
imately at right angles to the long axis; in the Il-form the rings are estimated
to lie at an average angle of ca. 45° to the fiber length. Additional infor-
mation on the difference between the I- and II- forms of DNA is given by
Franklin and Gosling^-^ in a discussion of X-ray diffraction evidence sup-
porting the Watson and Crick" double coaxial helical structure for sodium-
DNA. They consider that in the highly ordered I-form at 75% humidity
the planes of the base rings are at an angle of ca. 25° to the fiber axis; this
value, and that estimated by Wilkins ct al.™ for the Il-form, appear to be
consistent with the changes in dichroism observed for the I :^ II intercon-
version process.

It is evident that in these structural studies the ultraviolet dichroism
has provided valuable supplementary information to the X-ray diffraction
and other optical investigations. [Cf. Jordan, Chapter 13.]

Work has been reported on the ultraviolet dichroism of the nucleoprotein
(containing PNA) associated with tobacco mosaic virus. Butenandt and
co-workers^-' measured the dichroism of solutions oriented by streaming,
and their essential results have been confirmed by Seeds and Wilkins^^^"^ '^-^
working with both flow-oriented solutions and shear-oriented gels, who find,

128 R. E. Franklin and R. G. Gosling, Nature 171, 740 (1953); 172, 156 (1953).

129 A. Butenandt, H. Friedrich-Freksa, St. Hartwig, and G. Scheibe, Z. -physiol.
Chem. 274, 276 (1942).

542 G. H. BEAVEN, E. R. HOLIDAY, AND E. A. JOHNSON

however, the dichroism to be of opposite sign from that reported by Bute-
nandt et al., i.e., absorption is greatest with the electric vector parallel to
the flow (or shear) direction. This alteration in sign is confirmed by Perutz,
Jope, and Barer,^'° using the microspectrographic technique of Barer et al.,^^^
adapted for work with polarized radiation. These three groups of workers
are mainly concerned with the significance to be attached to the parallel
dichroism of the tryptophan fine-structure band at ca. 290 m/i of the protein
moiety. Seeds and Wilkins reject the conclusion of Butenandt et al. that
the dichroism of similar sign and magnitude in the 260-mM region is an
indication of orientation of the nucleic acid, and suggest that it is due to
form dichroism. The optical properties of the hexagonal crystals present
in the leaf hairs and other cells of plants infected with tobacco mosaic
virus have also been studied, '°^ but the results obtained are too complex
to be reviewed here. Calf thymus nucleoprotein (containing DNA) gives
oriented films and fibers (though less readily than the derived DNA) which
are negatively birefringent. As with nucleic acid, the ultraviolet dichroism
is negative, with maximum absorption when the electric vector is perpen-
dicular to the direction of shear, and dichroic ratios of 1.5-1.7 have been
observed for air-dried specimens of birefringence ca. —0.01 4. ^"^-'-^

6. Influence of Dichroism on the Microspectrographic Estimation
OF Nucleic Acid in Intact Cells

The use of a refracting microscope for the ultraviolet microspectrometry
of intact cells, was initiated by Caspersson^^^ ^^ 1936 and has since been
intensively developed^^* and exploited*' by him and his co-workers as a
valuable cytochemical tool. [Cf. Swift, Chapter 17.] More recently, fully
achromatic reflecting microscopes have been devised for such work; the
results obtained have been reviewed by Davies and Walker,'^ and instrumen-
tal problems elsewhere ^°*

Commoner and Lipkin^^'* suggest that orientation, which is known from
birefringence observations to be often present to a greater or lesser extent in
biological material, may be expected to invalidate absorption measure-
ments with unpolarized light because of Beer-Lambert absorption law
deviations which are, according to their arguments, inherent in such sys-
tems. Their views have been the subject of much comment.

Thorell and Ruch'" have measured the absorption at 260 m^i of sodium-
's" M. F. Perutz, E. M. Jope, and R. Barer, Discussions Faraday Soc. No. 9, 423 (1950).
I'l R. Barer, E. R. Holiday, and E. M. Jope, Biochim. et Biophys. Acta 6, 123 (1950).
'32 T. Caspersson, Skand. Arch. Physiol. 73, Suppl. 8 (1936).
'" T. Caspersson, Exptl. Cell. Research 1, 595 (1950); T. Caspersson, F. Jacobsson,

and G. Lomakka, ibid. 2, 301 (1952).
'3< B. Commoner, Science 110, 31 (1949); B. Commoner and D. Lipkin, ibid. 110, 41

(1949).
'" B. Thorell and F. Ruch, Nature 167, 815 (1951).

OPTICAL PROPERTIES OF NUCLEIC ACIDS 543

DNA solutions oriented by streaming and of gels oriented by stretching,
both with unpolarized radiation and with radiation polarized parallel and
perpendicular, respectively, to the orientation direction. Their theoretical
analysis indicates that the error in absorption measurements on such ori-
ented systems with unpolarized radiation will be dependent on the dichroic
ratio, and they suggest that the magnitude of the error must be investigated
experimentally for each particular case. Their results show small errors,
ca. 5%, for specimens with quite high dichroic ratios (up to 3) and high
transmission, but the error is very much larger for low transmissions. They
conclude, however, that the dichroism of the nucleic acid in cytological
material is so low, except for material such as sperm heads, that the effect
of molecular orientation on the absorption measured with unpolarized light
is unlikely to introduce the serious errors adduced by Commoner and
Lipkin.

As noted above, Caspersson's early ultraviolet absorption measure-
ments^'* on an oriented film of DNA show good agreement between the
curve for unpolarized radiation and the average of the two curves for radi-
ation polarized parallel and perpendicular to the orientation axis. In this
example the average absorbance is ca. 0.6 (approx. 25 % transmission) and
the dichroic ratio about 1 .6; the experimental value for the peak absorbance
with unpolarized radiation is not more than 3 % lower than the calculated
figure.

In a critical evaluation of quantitative cytochemical techniques Click,
Engstrom, and Malmstrom"® accept the conclusions of Thorell and Ruch,
and cite other examples of biological material in which orientation might
be expected but where the observed ultraviolet dichroism is low, so that
absorption measurements with unpolarized radiation should not be subject
to appreciable error. They also draw attention to the possibility of error due
to inhomogeneous distribution of absorbing material over the total area of
measurement.^" It can easily be shown that error due to such inhomogeneity
is most serious at high absorbance values. Since the proportional effect
of a constant fractional error in transmission on absorbance (and hence on
concentration) measurement also depends on the absorbance. Click et al.
emphasize the importance of working at fairly low absorbance levels. The
concept of an optimum absorbance range, approximately 0.2-0.7, for precise
spectrophotometry has long been accepted for measurements in solution,
where problems due to heterogeneity do not arise. "^■^''

In a discussion of errors in microspectrography Wilkins"" comments on

"6 D. Glick, A. Engstrom, and B. G. Malmstrom, Science 114, 253 (1951).
'" R. N. Jones, J. Am. Chem. Soc. 74, 2681 (1952).

"» G. F. Lothian, "Absorption Spectrophotometry," p. 52. Hilger and Watts, Lon-
don, 1949.
"9 N. T. Gridgeman, Anal. Chem. 24, 445 (1952).
i<° M. H. F. Wilkins, Discussions Faraday Soc. No. 9, 363 (1950).

544 G. H. BEAVEN, E. R. HOLIDAY, AND E. A. JOHNSON

the views of Commoner and Lipkin and concludes with Thorell and Ruch
that errors due to orientation would only be serious in biological material
of exceptional character, in which the degree of orientation is very high. In
spite of considerable debate^^^ it does not appear that Commoner and
Lipkin 's arguments have been completely refuted, however, and it seems
likely that more experimental work on carefully chosen specimens will be
necessary to resolve this issue.

7. DicHROiSM OF Stained Nucleic Acids

The combination of a dye with an oriented macromolecule may lead to
orientation of the dye molecules, causing the visible light absorption of the
dyed macromolecule to show dichroism. There is an extensive literature on
the visible dichroism of substantively dyed cellulose fibers, and White and
Elmes^"*^ have made some observations on dyed birefringent fibers of DNA
and the corresponding nucleoproteins isolated from various human tissues.
With pyronine and methyl green, the positive dye dichroism indicates that
the absorbing groups of the dyes lie parallel to the fiber axis, although
imperfectly oriented. For toluidine blue-stained fibers the visible dichroism
is negative, and the absorbing groups of the dye appear to lie roughly at
right angles to the fiber length. It is of interest to note that the oriented
DNA fibers used in this work were laid down within crystals of sodium
chloride depositing from solutions of the nucleic acid or nucleoprotein in
sodium chloride solutions. The salt could be dissolved away with methanol
leaving arrays of birefringent fibers which show the type A-type B inter-
conversions described by Wilkins et alP'' The dye dichroism of stained,
oriented nucleic acid is of great interest in relation to the use of methyl
green as a specific dye for high-molecular- weight DNA; conversely pyro-
nine stains depolymerized DNA and undegraded PNA. The differential
staining of tissues by methyl green-pyronine (Unna-Pappenheim) is thus
intimately related to the nature and state of the nucleic acids present. ^^^
The subject has been discussed in detail by Kurnick"* and Thomas."^ The
phosphate groups are certainly implicated in the binding of basic dyes
by nucleic acids,^*^ but an explanation for the binding of specific dyes in
terms of dye molecular structure has not yet been attempted. It is perhaps

1" B. Commoner, Discussions Faraday Soc. No. 9, 393 (1950); H. R. Catchpole and

I. Gersh, ibid., p. 471.
"2 J. C. White and P. C. Elmes, Nature 169, 151 (1952).
1" J. Brachet, Compt. rend. soc. biol. 133, 88 (1940).
1" N. B. Kurnick, Exptl. Cell Research 1, 151 (1950) ; ibid. 3, 649 (1952) ; Stain Technol.

27, 233 (1952); Arch. Biochem. 29, 41 (1950).
1" R. Thomas, Arch, intern, physiol. 61, 270 (1953).
1" L. F. Cavalieri and A. Angelos, /. Am. Chem. Soc. 72, 4686 (1950) ; L. F. Cavalieri,

S. E. Kerr, and A. Angelos, ibid. 73, 2567 (1951).

OPTICAL PROPERTIES OF NUCLEIC ACIDS 545

noteworthy that unhke other biological macromolecules with acidic func-
tional groups, PNA is a unique substrate for some basic dyes in that it re-
presses dimerization and hence the change in dye color known as metachro-
masy. The subject has been discussed by Michaelis."^ [Cf. Swijt, Chapter
17.]

V. Infrared Absorption Spectra

The application of infrared absorption techniques to the study of nucleic
acids is relatively new as in Loofbourow's extensive review article published
in 1940," on the investigation of materials of biological interest by physical
methods, there is not a single reference to this topic. Even Schlenk's account
of nucleic acid chemistry which appeared in 1949^^^ cites only one paper,
that of Blout and Fields^*' published in 1948. Subsequent investigations on
nucleic acids proper have been reviewed very recently by Fraser.^^" Several
extensive studies, however, of simple pyrimidines and related compounds
have been made and are briefly noted here.

Broadly speaking, infrared absorption methods have been used for two
different purposes in the nucleic acid field. On the one hand, the wealth of
detail in an infrared spectrum may be used as a "finger-print" to identify a
particular compound, and to detect it in mixtures. The infrared absorption
spectrum may also be used as a proof of identity between synthetic and
natural specimens of the same compound, especially when the existence of
isomers with very similar properties results in other characterization meth-
ods being inadequate for such a purpose. Many examples of this application
can be found in the recent literature dealing with the synthesis of nucleo-
sides and nucleotides. For example, Brown and Todd^" were able to dis-
tinguish between natural and synthetic adenylic acids a and h and muscle
adenylic acid (adenosine-5'-phosphate) by their infrared absorption spectra,
obtained on mulled samples, and the isomeric cytidylic acids can also be
identified in spite of the additional complication of polymorphism.^^' A
strong characteristic band free from overlapping can be used for the esti-
mation of a compound in mixtures if suitable solvents are available, though
even mull spectra are of analytical value. ^^^ Work of this sort entails the use

"' L. Michaelis, Nucleic acids and nucleoproteins, Cold Spring Harbor Symposia

Quant. Biol. 12, 131-142 (1947); see also G. Oster and H. Grimsson, ref. 60.
1" F. Schlenk, Chemistry and enzymology of nucleic acids, Advances in Emymol. 9,

455-535 (1949).
1" E. R. Blout and M. Fields, Science 107, 252 (1948).
IS" R. D. B. Fraser, The infra-red spectra of biologically important molecules, Progr.

Biophys. and Biophys. Chem. 3, 47-60 (1953).
1" D. M. Brown and A. R. Todd, J. Chem. Soc. 1951, 44.
162 R. J. C. Harris, S. F. D. Orr, E. M. F. Roe, and J. F. Thomas, J. Chem. Soc. 1953,

489.

546 G. H. BEAVEN, E. R. HOLIDAY, AND E. A. JOHNSON

of infrared absorption spectra on an essentially empirical basis and no
fundamental interpretation of the spectrum is really required. The other
application is for structural studies, and in this case very detailed analysis
of the spectrum is necessary. This may take the form of the correlation of
band frequencies with particular structural elements of the sample, i.e.,
the use of "group correlations," for which there is an extensive empirical
and theoretical literature (see Fraser^*" for references). For more funda-
mental indications of structure, e.g., the detection of hydrogen-bonding
in polynucleotide chains and other detailed aspects of macromolecular
structure, reliable assignments of bands to known modes of vibration are
required. The progress which has been made in this direction has also been
critically discussed by Fraser^*" for biological macromolecules, and by
Thompson et aZ.^" and Sutherland^ ^'^ for large molecules in general.

Problems of sample preparation and of solvent selection are prominent
in the infrared spectroscopy of nucleic acids and the related simple com-
pounds. Because of its heavy absorption water is virtually useless as a
solvent, quite apart from its solvent power for the rock-salt windows nor-
mally used in absorption cells. The available solvents with useful trans-
mission windows in the infrared^*^ are of little value; thus Lacher and co-
workers'^^ went to the trouble of using antimony trichloride at 100° C. to
obtain absorption data in the 1-2. 5-m overtone region on some pyrimidines.
Pyrimidines and purines have therefore mainly been examined as mulls,
in which a finely powdered solid is mixed with a suitable nonvolatile liquid.
The spectrum of a mulled sample will be partly obscured by any absorption
bands of the mulling liquid, and the observed band intensities are subject
to uncertainties arising from scattering losses and the difficulties in con-
trolling sample concentration and film thickness. The fact that the spec-
trum refers to the solid state must also be borne in mind when theoretical
interpretation is attempted. Many pyrimidines and purines are sufficiently
stable to be sublimed in vacuo, however, and thin-film samples prepared
in this way have been used by many workers for both infrared''*^ •'"■'** and
ultraviolet^^ ■^'* absorption studies. A new method'^* for preparing solid
samples for infrared absorption studies makes use of the fact that powdered
potassium bromide can be converted under moderate pressure into clear
plates with good infrared transmission. The finely ground sample is uni-

'" H. W. Thompson, D. Nicholson, L. N. Short, Discn^sions Faraday Soc. No. 9, 222

(1950).
'" G. B. B. M. Sutherland, Discussions Faraday Soc. No. 9, 274 (1950).
1" P. Torkington and H. W. Thompson, Trans. Faraday Soc. 41, 184 (1945).
i6« J. R. Lacher, D. E. Campion, and J. D. Park, Science 110, 300 (1949).
1" E. R. Blout and M. Fields, J. Biol. Chem. 178, 335 (1949).

168 E. R. Blout and M. Fields, /. Am. Chem. Soc. 72, 479 (1950).

169 M. M. Stimson, J. Am. Chem. Soc. 74, 1805 (1952); U. Schiedt, Z. Naturforsch 76,
270 (1952) ; Appl. Spectroscopy 7, 75 (1953).

OPTICAL PROPERTIES OF NUCLEIC ACIDS 547

formly mixed with the haUde salt before pressing, and its vokime concentra-
tion in the resulting "sample plate" can be readily controlled. The spectra
obtained from samples prepared in this way are free from solvent absorption
and scattering errors and are of quantitative value, so that this technique
promises to be a very real advance in infrared analytical technique.

Nucleic acid samples can be examined as cast films, prepared by evap-
orating aqueous solutions on silver chloride plates.'''^ Oriented samples can
be obtained by shearing viscous gels or by stretching fibers ;^°' such samples
are usually too small to be examined directly with conventional spectrom-
eters and require some form of reflecting micro.scope to form an enlarged
image of the specimen on the entrance slit.^®*^

The infrared absorption spectra of the pyrimidines and purines found in
nucleic acids and of xanthine, hypoxanthine, and of some methylated
xanthines have been measured by Blout and Fields,"' '^"'^^^ mainly on
evaporated films. Several features in the spectra that are of value for identi-
fication and analysis are noted and some bond assignments proposed.
Brownlie^" has examined twenty-five pyrimidines, containing two or more
substituents, as mulls over the 2-1 5-m range and has attempted a large
number of possible bond assignments. Some of his conclusions are debated
jby Short and Thompson'^- in a study of no less than eighty pyrimidines
ncluding many important mono-substituted derivatives, over the range
2-25 M- These workers used lithium fluoride, sodium chloride, and potas-
sium bromide prisms to cover this wide range of wavelengths, and treated
some of their compounds with deuterium oxide to assist the recognition of
vibrations involving hydrogen atoms. Their discussion of this large volume
of data is an excellent example of the partly empirical, partly theoretical
approach that has to be employed in the interpretation of such complex
spectra. All the spectra are highly characteristic and contain many strong
sharp bands suitable for identification and analysis, even when precise
structural correlations are difficult. Interpretations based on previously
established group correlations favor a ketonic structure for the 2-hydroxy-
and 4-hydroxypyrimidines and probably a diketonic structure for the
2 , 4-dihydroxy derivatives. Amino groups appear to exist in the non-tauto-
merized form. Both the tautomerism and the hydrogen-bonding in the
solid state are influenced by the various substituents present in the com-
pounds studied. Short and Thompson conclude that frequency assignments
in pyrimidine spectra must be made with considerable reserve, in view of
the meager X-ray structure data that can be utilized to provide confirma-
tion. Even so, their tentative findings are obviously of great interest in

»«» R. D. B. Fraser, Discussions Faradaxj Sor. No. 9, 378 (1950).

i6» I. A. Brownlie, J. Chem. Soc. 1950, 3062.

'»2 L. N. Short and H. W. Thompson, J. Chem. Soc. 1952, 168.

548

G. H. BEAVEN, E. K. HOLIDAY, AND E. A, JOHNSON

4000 3000

2000

Frequency, cm.'i
1500 1200 1000 900

800

700

2 4 6 8 10 12 14

Wavelength, m
Fig. 28. Infrared absorption spectrum of cast film of DNA sodium salt (Rowen^^^).

relation to the specific interactions between base substituent groups en-
visaged in the Watson and Crick structure for DNA."

With regard to infrared absorption studies on nucleic acids proper there
is as yet very little to add to Fraser's excellent account/^" which deals also
with proteins and polysaccharides. The following frequency ranges are
listed by him for the fundamental vibrations that may be readily recognized
in the spectra of nucleic acids (e.g., Fig. 28).

(1) Vibrations involving hydrogen atoms:

Bond stretching O— H 3000-3700 cm.-i

N— H 3000-3500

C— H 2800-3100
Bond bending O— H ca. 1100

N— H 1500-1600

C— H 1300-1500

(2) Multiple-bond stretching:

C=0 1600-1800 cm.-i
C=N ca. 1650
C=C ca. 1650
P=0 1250-1300 "

(3) Skeletal frequencies involving many linked atoms:

\ /

— C— O— C— ca. 1100 cm.-i

163 J. W. Rowen, Biochim. et Biophys. Acta 10, 391 (1953).

OPTICAL PROPERTIES OF NUCLEIC ACIDS 549

(4) Ionic group frequencies:

PO4 ca. 1080 and 980 cm -1

The large number of different atomic groupings in a biological macro-
molecule such as nucleic acid gives rise to a very complex set of characteris-
tic and skeletal frequencies so that only a few of the more important can be
assigned with any certainty. The situation is complicated by overlapping of
bands and by severe broadening of those bands arising from groups which
may be hydrogen-bonded. The most detailed analyses have been made on
the spectra of proteins and related compounds/^^ and the interpretation of
nucleic acid spectra has followed along essentially similar lines, extended to
include the structural features peculiar to these compounds.

Blout and Fields"^ ■^" find that the sodium salts of PNA and DNA can be
distinguished by their infrared absorption spectra, especially at frequencies
lower than ca. llOOcm."^ Fraserand Chayenhave used microspectrographic
methods^®' to extend this finding to intact tissue sections. The specific
frequencies recognized by these workers are 860, 916, 969, and 997 cm.~'
for PNA, and 895, 930, and 967 cm.~' for DNA, as confirmed by selective
extraction procedures. Since the technique can be made at least semiquan-
titative, it offers the possibility that the two types of nucleic acid can be
differentiated and estimated separately in situ in biological material. A
further advantage is that any mononucleotide absorption contributions
can be distinguished and evaluated separately. Quantitative infrared micro-
spectrographic procedures may therefore prove to be a valuable supplement
to the more refined ultraviolet methods which have been given so much
more attention in the past.

Macroscopic tissue sections have been studied by Blout and Mellors,^^^ the sam-
ples being evenly wetted with high-molecular-weight fluorocarbon or hydrocarbon
(mineral) oil to reduce scattering. These authors were particularly interested in the
possibility of clinical diagnostic applications, and observed that a band of 9.3 m
(1075 cm."'), which they associated with nucleic acid, was more prominent in can-
cerous than in healthy mammary tissue. They also found definite differences in the
spectra of fixed and unfixed samples of the same tissue.

Cavalieri, Kerr, and Angelos,'^^ in a study of the enzyme-resistant residues of
various PNA samples, concluded that the infrared absorption spectra of mulls were
not sufficiently characteristic to reveal differences in macromolecular structure, al-
though intact nucleic acids could be differentiated from mixtures of mononucleotides.
More recently, however, Rowen'" has examined the effect of treatment with deoxy-
ribonuclease on the infrared absorption spectra of cast films of DNA (Fig. 28). The

'" G. B. B. M. Sutherland, Infrared analysis of the structure of amino acids, poly-
peptides and proteins. Advances in Protein Chem. 7, 291-318 (1952).
'«5 R. D. B. Fraser and J. Chayen, Exptl. Cell Research 3, 492 (1952).
i«6 E. R. Blout and R. C. Mellors, Science 110, 137 (1949).

550 G. H. BEAVEN, E. R. HOLIDAY, AND E. A. JOHNSON

band frequencies and assignments proposed by this author are in substantial agree-
ment with those given by Fraser;'*" the 960-cm.~' band is greatly reduced in intensity
by enzymic degradation and is therefore correlated with the internucleotide ester
linkage (P — O — Cs'), the 1015-cmr' band then being regarded as arising in the nucle-
oside-phosphate bond (P — O — Cs')- The assignments of vibrations involving the
phosphate group are very important for the interpretation of the infrared absorption
and dichroism of nucleic acids and are briefly discussed by Fraser.^'"

The dichroism of infrared absorption bands has been used in recent structural
studies on nucleic acids. The theoretical basis of the method is the assumption that
the transition moment associated with a vibration which is active in absorption is
either along or perpendicular to the band involved, depending on whether the vibra-
tion is bond-stretching or bond-bending, respectively. The dichroism of the C=0
stretching and N — H bending bands has been widely used to investigate the a- and
/3-configurations of oriented structural proteins and synthetic polypeptides (for
literature references see Fraser'^" and Short and Thompson'®^) Fraser and Price'*'
have shown, however, that in the important case of the peptide linkage this simple
assumption may not be justifiable. A consideration of the mechanical interactions
involving atoms other than the bonded pair in the C=0 group, and of the effect of
the vibration on the resonance structures of the peptide group, indicate that the di-
rection of the transition moment for the 0=^0 stretching vibration will not be exactly
along the bond. The interpretation of infrared dichroism may therefore have to be
made with considerable caution.

Fraser and Fraser'** have studied the infrared absorption and dichroism of shear-
oriented films of thymus DNA and assigned the principal bands in accordance with
the general classification given above. The N — H stretching vibrations of the base
amino and imino-groups and the double-bond (C==0, C=N, C=C) stretching vibra-
tions of the same ring systems show perpendicular dichroism, i.e., absorption is
greatest when the electric vector is perpendicular to the direction of shear, which is
taken as the direction of the polynucleotide chain. In agreement with the negative
birefringence and ultraviolet dichroism, the perpendicular infrared dichroism indi-
cates that the planar bases are approximately perpendicular to the chain axis. The
dichroism of other bands which are associated with the sugar residue and the phos-
phate group also appears to be consistent with the modifications proposed by Fur-
berg'" to the Astbury polynucleotide model. [Cf. Jordan, Chapter 13.] Fraser and
Fraser have also found that the type A — > type B transformation of DNA which
occurs on stretching at a suitable humidity (cf . Section IV) is accompanied by a re-
versal from perpendicular to parallel of the dichroism of manj^ important bands; the
effect is most marked with the 967-cm.~' bond, which has not yet been assigned with
certainty. The bond at 1235 cm."' with a dichroic ratio of unity, is not affected by
this structural transformation. The assignment of this bond to a P==() stretching
vibration'*' has been reconsidered.'**

Fraser"* has studied the infrared dichroism of shear-oriented tobacco mosaic virus
gels and finds the frequencies and dichroism of theC=Ostretchingand N — H bending
vibration bonds to be consistent with a structure in which the protein is in an a-con-

'" R. D. B. Fraser and W. C. Price, Nature 170, 490 (1952); see also A. Elliott, ihiil.

172, 359 (1953).
'** M. J. Fraser and R. D. B. Fraser, Nature 167, 761 (1951).

'*^ S. Furberg, Acta Chem. Scand. 4, 751 (1950) ; see D. O. Jordan, ref . 6, for summary.
"* R. D. B. Fraser, Nature 170, 490 (1952).

OPTICAL PROPERTIES OF NUCLEIC ACIDS 551

figuration with the polypeptide chains perpendicular to the axis of the rod-like virus
particle, unlike some fibrous proteins which have been studied by the same method.
Ambrose and Butler'^' have determined the dichroism in the 1500-1800-cm.~'
region of herring sperm nucleoprotein oriented by being cast on a rubber film base
which may be stretched to produce about 100% extension. The dichroism of the
1695-cm.~' stretching vibrations of the base C=0 and C==N bonds are consistent
with the dichroism found bj' Fraser and Fraser'^* for the moderately extended and
optically negative type A nucleic acid fiber. The protein, however, is in the |3-con-
figuration in this nucleoprotein. There seem to be interactions between the two com-
ponents of a nucleoprotein system which determine, on the one hand, the configura-
tion assured by the protein and, on the other hand, the internal order of the poly-
nucleotide chain.

It can be seen from this brief account of previous work that there are
many practical and theoretical difficulties associated with the infrared
absorption spectroscopy of nucleic acids. Further progress in the assign-
ment of frequencies and in resolving uncertainties in the directions of transi-
tion moments should lead to significant contributions to knowledge of the
detailed structure of nucleic acids.

Addendum

Bases, Nucleosides, and Nucleotides

Fox et al."^'' report that the ultraviolet absorption spectra of cytidylic
acids a and b behave in a significantly different manner in the alkaline pH
range, and attribute this to the presence or absence of an ionizable 2'-
hydroxyl group. The spectrum of cytidylic acid a, like that of 2'-deoxy-
cytidine, remains constant in the pH 12 to 14 region, and hence may be
presumed to be the 2'-phosphate ester, in which no 2'-hydroxyl group
ionization can occur. This agrees with other evidence. The specific effect of
2'-hydroxyl ionization in the b or 3'-isomer may be due to hydrogen bonding
between this group and the 2-keto group of the pyrimidine nucleus, since
in the 2'- and 3'-adenylic acids, where such hydrogen-bonding cannot occur,
no changes in the alkaline pH region are observed with either isomer.
Spectroscopic constants for the isomeric cytidylic acids have also been re-
ported by Harris ef al.^^^ and by Cavalieri"^''.

E. L. Bennett^" and L. L. Bennett"^ have synthesized a number of C^^-
containing purines and pyrimidines, respectively, and report brief spectro-
scopic data which are in good agreement with values for the normal com-

'" E. J. Ambrose and J. A. V. Butler, The physical chemistry of proteins, Disc^issions

Faraday Soc. No. 13, 261 (1953).
"1" J. J. Fox, L. F. Cavalieri, and N. Chang, J. Am. Chem. Soc, 75, 4315 (1953).
''!'' L. F. Cavalieri, J. Am. Chem. Soc, 75, 5268 (1953).
"2 E. L. Bennett, J. Ain. Chem. Soc. 74, 2420 (1952).
1" L. L. Bennett, J. Am.. Chem. Soc. 74, 2432 (1952).

552 G. H. BEAVEN, E. R. HOLIDAY, AND E. A. JOHNSON

pounds. Cavalieri et al.^'^* have discussed the effects of methyl substitution
on the ultraviolet absorption spectra of xanthines and give absorption
curves for xanthine and xanthosine which agree well with those published
here. Hamer et al."^ have reported the effects on the ultraviolet absorption
spectra of purines and pyrimidines when these compounds are reduced
under acid and alkaline conditions.

Nucleic Acids and Polynucleotides.

Laland et al."^^ have determined 8(P) values for DNA preparations from
animal, plant and microbial sources, before and after degradation by acid,
alkali, heat, ultrasonic irradiation or desoxyribonuclease. From their re-
sults they conclude that any treatment of DNA which only leads to a de-
crease in intermolecular bonding does not alter the 8(P) value, whereas the
breaking of intramolecular hydrogen bonds results in an increase in e(P).
These workers therefore prefer to regard any alteration in the state of DNA
in solution which is not accompanied by an increase in 8(P) as a disaggrega-
tion, to distinguish it from the breakdown of the intermolecularly-bonded
structure which gives rise to the absorption anomaly.

Ultraviolet Dichroism

Franklin and Gosling's X-ray diffraction studies of oriented DNA fibers
have now been reported in greater detail."^ These authors recognize a
highly crystalline A-form, stable at 75% relative humidity, and a para-
crystalHne B-f orm which occurs at humidities of 92 % and higher. Riley and
Oster^" had previously studied the DNA-water system over a very wide
range of composition, and the micelle state in which DNA shows a partly
ordered liquid-crystalline structure, may correspond to the paracrystalline
B-f orm of Franklin and Gosling. The relation between the various forms of
hydrated DNA recognized by X-ray diffraction and those recognized mainly
by dichroism by Wilkins, Gosling, and Seeds^-^ (Section IV.5) is not yet
clear, but it seems certain that the macromolecular order, and presumably
the optical properties of the oriented DNA-water system, are critically
dependent on water content, as explicitly stated by Franklin and Gosling.

"^ L. F. Cavalieri, J. J. Fox, A. Stone, and N. Chang, J. Am. Chem. Soc. 76, 1119

(1954).
"* D. Hamer, Deirdre M. Waldron, and D. L. Woodhouse, Arch. Biochem. and Bio-

phys. 47, 272 (1953).
"^* S. G. Laland, W. A. Lee, W. G. Overend and A. R. Peacocke, Biochirn. et Biophys.

Acta, 14, 356 (1954).
"« R. E. Franklin and R. G. Gosling, Acta Cryst. 6, 673, 678 (1953).
1" D. P. Riley and G. Oster, Biochirn. et Biophys. Acta 7, 526 (1951).

OPTICAL PROPERTIES OF NUCLEIC ACIDS 553

Infrared Absorption Spectra

Blout and Lenormant'''^ have obtained infrared absorption spectra of
concentrated solutions of RNA, DNA, and DNP in both water and deu-
terium oxide, using very thin cells to reduce the solvent absorption. Under
these conditions the absorption bands are sharper than those given by
dried films. The characteristic 1020-cm.~^ band of DNA is retained in solu-
tion with a slight shift in wavelength, and DNA and RNA can be clearly
differentiated.

Frick and Rosenberg"^ have obtained some infrared absorption evidence
for the existence of hydrogen bonds in native highly polymeric DNA. Work-
ing with pressed KBr sample plates^^* and a calcium fluoride prism, they
find that a small maximum at ca. 3.1 n, ascribed to hydrogen-bonded NH
stretching, is reduced to an inflection by prior exposure of the sample to pH
12, the remainder of the spectrum being unaltered.

"8 E. R. Blout and H. Lenormant, /. Opt. Soc. Amcr. 43, 1093 (1953).
"' G. Frick and A. Rosenberg, Biochim. et Bio-phys. Ada 13, 455 (1954).

CHAPTER 15

Nucleases and Enzymes Attacking Nucleic Acid Components

G. SCHMIDT

Page

I. Introduction 556

II. Enzymes Ciitahziiig the Cleavage of Bonds Between Nucleotides 558

1. Ribonucleases 558

a. Pancreas Ribonuclease (Ribonuclease I) 558

b. Other Ribonucleases • 575

2. Deoxyribonucleases 576

a. Pancreas Deoxj'ribonuclease (Deoxyribonuclease I) 576

b. Other Deoxyribonucleases 584

3. "Unspecific" Phosphodiesterases 585

a. Snake Venom Phosphodiesterases 585

b. "Phosphodiesterase" of Intestinal Mucosa 588

III. Enzymes Catalj-zing the Hydrolytic Cleavage of the Phosphoryl Groujjs of

Mononucleotides 590

1. Phosphatases of Low Degrees of Specificitj- 590

2. Specific Nucleotide Phosphatases (Nucleotidases) 591

a. 5'-Nucleotidases 591

b. 3'-Nucleotidase 592

3. Phosphatases as Phosphotransferases Between Phosphoric Acid Esters
and Nucleosides 593

IV. Nucleoside Kinases 594

V. Enzj'mes Acting on the Amino Groups of Purine and Pyrimidir.e Com-
pounds 595

1. Enzymic Deamination of the Adenine Group 595

a. Behavior of Free Adenine 595

b. Adenyl Deaminase of Aspergillvs oryzae 596

c. Adenosine Deaminase 596

d. 5' -Adenylic Acid Deaminase 597

2. Enzymic Deamination of the Guanine Group: Guanase, Guanosine
Deaminase, Guanylic Acid Deaminase 598

3. Enzymes Deaminating the Cytosine Group 599

4. Evidence for Transamination in the Purine and the Pyrimidine Field . . 599
VI. Enz.ymes Acting on the Linkages Between the l^asic and the Carbohydrate

Groups of Nucleic Acid Derivatives 600

] . Nucleoside Phosphorylases 602

a. Purine Nucleoside Phosphorylases 602

(1) Animal Enzymes 602

(2) Enzymes of Microorganisms 603

b. Pj'rimidine Nucleoside Phosphorylases 604

555

556 G. SCHMIDT

Page

(1) Animal Thymidine Phosphorylase 605

(2) Bacterial Uridine Phosphorylase 605

2. Nucleoside Hydrolases 606

a. Purine Nucleoside Hj'drolase of Yeast 606

b. Uridine Hydrolase of Yeast 606

c. Nucleoside Hydrolases of Lactobacillus pentosus 607

3. Nucleotide-l'-phosphorylase 607

4. Nucleotide-l'-pyrophosphorylase 607

5. DPN Hydrolases 608

VII. Xanthine Oxidase 609

VIII. Enzymes Involving the Opening of the Purine Ring 614

1. Inosinic Acid Transformylase 614

2. Uricase 614

IX. Enzymes Involving the Opening of the Pyrimidine Ring 619

1. Reversible Degradation of Orotic Acid by Bacterial Enzymes .... 620

a. Dihydroorotic Acid Dehydrogenase 620

b. Dihydroorotase 621

c. 5-Carboxymethylh}^dantoinase 621

2. Conversion of Dihydrouracil and Dihydrothymine to /3-Amino Acids

by Tissue Slices 621

3. Degradation of Uracil and Thymine by Bacterial Enzymes 622

a. Bacterial Oxidase of Uracil and Thymine 622

b. Barbiturase 622

X. Some Data Concerning the Intracellular Distribution of Enzymes of

Nucleic Acid Metabolism 623

1. Deoxyribonuclease 624

2. Ribonuclease 624

3. Intracellular Localization of Deoxyribonuclease I and Ribonuclease I

in Pancreas 625

4. Other Enzymes of Nucleic Acid Metabolism 625

XI. Addendum (Concerning the Enzymic Formation of Nucleoside Di- and Tri-
phosphates) 625

I. Introduction

Owing to the large number of the several types of bonds between the
component groups in each nucleic acid molecule, various sequences of
reactions are theoretically possible for the catabolism and for the biosyn-
thesis of the nucleic acids. As in other metabolic fields, several different
specific enzymes exist in many instances for the cleavage or formation of a
given bond in nucleic acids or their derivatives. The pathway of nucleic
acid metabolism is thus not uniform. The biological significance of the
many alternative pathways of nucleic acid metabolism is not known at
present, but the highly specialized distribution and the high specificity of
some of these enzymes strongly suggests the idea that, in these instances,
the enzymic degradation products have specific biological functions.

One generalization regarding the intermediary metabolism of nucleic
acids can be made, at least as a summary statement reflecting the present

ENZYMES ATTACKING NUCLEIC ACIDS 557

state of knowledge: the only bonds known to be susceptible to the action
of enzymes in nucleic acids or polynucleotides are the phosphoric ester
linkages. All known deaminases or enzymes involving the glycosidic bonds
act only on mononucleotides or their cleavage products. The degradation
of nucleic acid in the laboratory to substances such as GuUand's deamino-
ribonucleic acid or Chargaff's apurinic acid has so far no analogy in the
living organism. Obviously, confirmation or refutation of this statement
will be an important topic of future research. The validity of this generali-
zation would mean that the polymerization of mononucleotides would be
the exclusive mechanism of the biosynthesis of the nucleic acids and the
only determining factor for the sequence of the various nucleotide groups.
It would mean that actions such as the enzymic exchange of purine groups
would take place exclusively at the mononucleotide or nucleoside level.

The degrees of specificity of the enzymes of nucleic acid metabolism
show widely different ranges. Many of these enzymes catalyze reactions
not only of natural substrates, but also of analogous substances. Such
cases are of particular interest when biosynthetic enzyme reactions are
inv^olved. The observations concerning the incorporation of antimetabo-
lites into important cell constituents of the intact organism are paralleled
by the behavior of isolated enzyme fractions. The concept of such incor-
poration is becoming increasingly important in the explanation of physio-
logical effects of certain antimetabolites.

The number of different enzymic mechanisms in the field of nucleic acid
metabolism is too large to fit into the original system of nomenclature
proposed by Levene and Medigreceanu.^ The present terminology is some-
what arbitrary because some original designations are now applied with a
modified meaning. Levene and Medigreceanu introduced the collective
term "nucleases" for all enzymes involved in the metabolism of nucleic
acids or their degradation products or precursors. In contrast to this broad
application of the term "nucleases,"- the terms "ribonuclease" and "de-
oxyribonuclease"- — introduced by Dubos and Thompson^ and by Kunitz,^^
respectively — are now usually limited to the designation of enzymes which
catalyze the cleavage of the phosphoric ester bonds interlinking the nucleo-
tide groups of nucleic acids or polynucleotides.*

' P. A. Levene and F. Medigreceanu, J . Biol. Cheni. 9, 389 (1911).

2 O. Hoffmann-Ostenhof, Advances in Enzymol. 14, 219, 225 (1953).

» R. Dubos and R. H. S. Thompson, J. Biol. Chem., 124, 501 (1938).

3» M. Kunitz, J. Gen. Physiol, 33, 349 (1950).

* H. S. Loring and F. H. Carpenter suggested replacement of the term "ribonu-
clease" by the term "ribonucleinase" [J. Biol. Chem. 150, 381 (1943)], but the
customary use of the shorter term is continued in the literature. The reviewer
feels that, owing to the overlapping specificities of manj' nucleases, it will be next
to impossible to devise a framework of nomenclature which will not be riddled
with inconsistencies bj' future observations.

558 G. SCHMIDT

Only a few of the known enzymes of nucleic acid metabolism have been
thoroughly investigated, and, as yet, only two have been obtained in
crystalline form. The properties of these nucleases will be discussed in
some detail in the following sections. Some of these enzymes have become
important tools for the study of the structure of nucleic acids. The wealth
of information accumulated during the intense recent study of ribonuclease
I and deoxyribonuclease I is in striking contrast to the scarcity of descrip-
tive data available for the majority of the enzymes of nucleic acid metab-
olism. A review of the present knowledge of this field is, therefore, of
necessity very unbalanced. It should be pointed out that the relative space
devoted to some individual enzymes reflects the intensity of work devoted
to their study, rather than the degree of their importance in nucleic acid
metabolism.

II. Enzymes Catalyzing the Cleavage of Bonds Between Nucleotides

1. RiBONUCLEASES

a. Pancreas Ribonuclease (Ribonuclease I)

History. In 1920, Walter Jones observed that boiled extracts of pancreas
were capable of transforming yeast ribonucleic acid to acid-soluble products
without the liberation of inorganic phosphate or of purine and pyrimidine
bases. ^ This observation was a decisive advance in our knowledge of nucleic
acid metabolism because it demonstrated for the first time the existence of
a specific enzyme the activity of which is limited to the cleavage of inter-
nucleotide bonds without the formation of ammonia, inorganic phosphate,
or free purines or pyrimidines. The separation of this specific activity from
those of the rather ubiquitous deaminases, phosphatases, and nucleosidases
was facilitated by the exceptional heat stability of Jones' enzyme, since at
the time of its discovery the technique of enzyme fractionation was very
primitive. The enzymic degradation of nucleic acids in pancreas had been
observed much earlier by Jones himself® as well as by others; but, in these
experiments, the hydrolysis always proceeded beyond the stage of nucleo-
tides. It is, therefore, justified to consider the date of the discovery of
ribonuclease as coincident with that of the detection of its heat stability.
Interestingly enough, in his first paper, Jones himself expressed doubt as
to whether the activity of his boiled pancreas extracts could be attributed
to the presence of an enzyme. This doubt was shared by his contemporaries
and was during a considerable period a discouraging influence on the
further investigation of Jones' observation. Eighteen years had elapsed
after the publication of Jones' paper, when interest in the heat-stable

6 W. Jones and M. E. Perkins, Am. J. Physiol. 55, 557 (1923).
6 W. Jones, Z. physiol. Chem. 41, 101 (1904).

ENZYMES ATTACKING NUCLEIC ACIDS 559

nuclease of pancreas was reawakened by the observation of Dubos and
his associates,^ '^ who found that the Gram-positive staining properties of
killed pneumococci disappeared after incubation of the cells with boiled
pancreas extracts. They succeeded in a considerable purification of Jones'
enzyme for which they proposed the term "ribonuclease,"'and they found
evidence supporting the assumption that the ribonucleic acid fraction of
pneumococci was related to their Gram-positive properties. The work of
these authors stimulated P. A. Levene** to undertake a study of pancreas
ribonuclease in collaboration with Schmidt. These authors confirmed
Jones' concept that the action of ribonuclease was limited to the cleavage
of the interlinkages between the nucleotide groups of ribonucleic acid, but
they emphasized that only part of these interlinkages could be split by
the enzyme.

A new phase in the investigation of ribonucleases began in 1940 when
Kunitz^ obtained the enzyme in crystalline form (Fig. 1) and found that
its action was accompanied by the liberation of acidic groups. It will be
seen in the following sections and in other chapters of this book that this
achievement not only opened the way for a clearer understanding of the
properties of Jones' enzyme, but that it also created a valuable tool for
the study of the structure of its substrates, the ribonucleic acids.

Specificity. Ribonuclease I is a highly specific phosphodiesterase which
hydrolyzes all known ribonucleic acids, certain ribonucleotides, ^° and syn-
thetic ribonucleotide-P-esters" which will be defined in the following
sections (see also Chapter 12), and a natural polymer of ribosephosphoric
acid which was discovered by Zamenhof et al}- (Fig. 2). It also hydrolyzes
polynucleotides obtained by deamination of PNA uith nitrous acid.^*'^''
On the other hand, ribonuclease is entirely inactive toward deoxyribo-
nucleic acid and its hydrolysis products,^ except apurinic acid ("thymic
acid"),^^ and toward the P-esters of a-glycerophosphoric acid such as
L-a-glycerylphosphorylcholine and L-a-glycerylphosphorylethanolamine.'^
Diphenyl phosphate and dinitrophenyl phosphate, which are used as

' R. H. S. Thompson and R. Dubos, J. Biol. Chem. 125, 65 (1938).

« G. Schmidt and P. A. Levene, J. Biol. Chem. 126, 423 (1938).

9 M. Kunitz, /. Gen. Physiol. 24, 15 (1940).

10 D. M. Brown, D. I. McGrath, and A. R. Todd, J. Chem. Soc. 1952, 2708.
" D. M. Brown, E. A. Dekker, and A. R. Todd, J. Chem. Soc. 1952, 2715.
'2 S. Zamenhof, G. Leidy, P. L. FitzGerald, H. E. Alexander, and E. ChargafT, J.

Biol. Chem. 203, 695 (1953).
'3 W. E. Fletcher, J. M. Gulland, D. O. Jordan, and H. E. Dibben, J. Chem. Soc.

1944, 30.
'■* L. Vandendriessche, Compt. rend. trav. lab. Carlsberg, Ser. chim. 27, 341 (1951).
'6 M. C. Durand and R. Thomas, Biochim. et Biophys. Acta 12, 416 (1953).

560

G. SCHMIDT

Acid extract
of beef pancreas

Trypsin

Precipitate
Trypsinogen

Compound

*^ >!^i^ trypsin*

'. a t€ n inhibitor

4

^^.t?,

Trypsin
inhibitor

Ribonuclease

01

0.2

-0,3

04-

-05

-0.6

Desoxy-
nbonuclecse

Chymo - cc- Chyjno-

trypsinogen trypsin

-0.7

0.8

09

Saturation of
ammonium sulfate

0-Chymo-
tpypsin

if"- Chymo -
trypsin

Fig. 1. Fractionation of pancreas enzj^mes. [From. M. Kunitz.'*]

model substrates for other phosphodiesterases, are resistant against ribo-
nuclease.^®

Hydrolysis Products Resulting from the Action of Rihonuclease I on Ribo-
nucleic Acids. Ribonuclease I does not hydrolyze all bonds between the
component nucleotides of ribonucleic acid, but it catalyzes the cleavage
only of certain strictly defined internucleotide bonds, namely those be-
tween the 3'-pyrimidine nucleoside phosphoryl groups and the 5 '-hydroxy
groups of the adjacent purine or pyrimidine nucleotide groups.^* '^^^
Thus, neither the bonds between adjacent purine nucleotide groups nor

16 G. Schmidt, R. Cubiles, N. Zollner, L. Hecht, N. Strickler, K. Seraidarian, M.

Seraidarian, and S. J. Thannhauser, /. Biol. Cheni. 192, 715 (1951).
!«• H. S. Loring, F. H. Carpenter, and P. M. Roll, /. Biol. Chem. 169, 601 (1947).

ENZYMES ATTACKING NUCLEIC ACIDS

561

Fig. 2. Probable structure of ribose phosphate poljsaccharide of Hemophilus
influenzae. (From Zamenhof et al.^^]

H-C-

H-C-OH

1
H-C-0

H-C

Hj-C-OH

0-

I

in
o

Py (or Pu)

H-C

I
H-C-OH

H-C-0

H-C-

•O-C-H2

0-

0

Step 1

Transphosphorylation

Py

H2-C-0H

Py (or Pu)
H-C

H,-C-OH

Py (or Pu)
H-C

H-C-OH

+ H2O Step II

Hydrolysis

H-C-0

P=0 H-C

0- H2-C-OH 0

Fig. 3. Mechanism of action of ribonuclease I according to Brown and Todd.

562 G. SCHMIDT

the bonds between a 5'-pyrimidine phosphoryl group and a 3'-hydroxj^
group of the adjacent purine or pyrimidine nucleotide group are hydro-
lyzed. Consequently, after exhaustive incubation with ribonuclease, all
3 '-phosphoryl groups attached to pyrimidine nucleoside groups are present
in the form of secondarj^ phosphoryls. This observation^ — ^in addition to
kinetic observations^ — ^supports the conclusion that the incomplete hydrolj'-
sis of the internucleotide bonds by ribonuclease is not due to inhibitory
influences on the enzyme, but to its sharply defined specificity toward the
internucleotide bonds as defined in this chapter. The products obtained
after exhaustive digestion of ribonucleic acids with ribonuclease are 3'-
uridylic acid, 3'-cytidylic acid, and a considerable number of oligonucleo-
tides of various degrees of polymerization. These oligonucleotides were
designated as limit polynucleotides by Schmidt ei al}^ in analogy to the
terminology used for the end-products of amylase action on polysaccha-
rides. The simpler components of the complex mixture of the limit poly-
nucleotides of low molecular weights were successfully separated by Volkin
and Cohn^^ on ion-exchange columns and by Markham and Smith^^ by
ionophoresis. They are unbranched di-, tri-, and tetranucleotides of differ-
ent compositions, but all have in common one structural property: each
of these oligonucleotides contains one pyrimidine nucleotide group per
molecule. The pyrimidine nucleotide group is always terminal and carries
the secondary phosphoryl group of the chain on its 3 '-carbon atom. Be-
tween 30 and 49 % of the purines of yeast PN A and between 45 and 50 %
of those of liver PNA are found as oligonucleotides containing four or
fewer mononucleotide groups after exhaustive digestion with ribonuclease.
The remainder are present in the form of polj^nucleotides of higher order
which have not been separated as yet.

A considerable proportion of these oligonucleotides, corresponding to 15
to 25 % of the total phosphorus of the substrate, is not dialyzable against
water. Until recently this fraction was considered to consist of polynucleo-
tides of relatively high molecular weight which have been termed "cores"
or "limit nucleic acids" by Zamenhof and Chargaff.^^ Markham and
Smith^* showed recently, however, that the diffusibility of oligonucleotides
through dialysis membranes was greatly enhanced by the presence of
sufficient concentrations of sodium chloride in the solution.

The description of the internucleotide bonds hydrolyzed by ribonuclease
implies that the degree of polymerization is not important for the action
of this enzyme.-" This conclusion is borne out by the observations of Merri-

" E. Volkin and W. E. Cohn, J. Biol. Chem. 205, 767 (1953).

'8 R. Markham and J. D. Smith, Biochem. J. 52, 565 (1952).

»' S. Zamenhof and E. Chargaff, /. Bwl. Chem. 178, 531 (1949).

" B. Magasanik and E. Chargaff, Biochim. et Biophys. Acta 7, 396 (1951).

ENZYMES ATTACKING NUCLEIC ACIDS 563

field and Woollej',-^ who succeeded in isolating by ion-exchange chromatog-
raphy a considerable number of di- and trinucleotides from hydrolysates
of yeast ribonucleic acid obtained by short treatment of PNA with 6 N
hydrochloric acid at room temperature. All those oligonucleotides in which
the 3'-phosphoryl groups of pyrimidine nucleotides were esterified with
other nucleotides were cleaved by ribonuclease, whereas all oligonucleo-
tides in which such a structure was absent were resistant toward the
enzyme.

The presence of pyrimidine residues in those nucleotide groups which
react with ribonuclease is, however, not a specific recjuirement for the
catalytic action of the enzyme. Zamenhof et al}- isolated from Hemophilus
influenzae, type b, a polymer of ribosyl-3-phosphoric acid which was hydro-
lyzed by ribonuclease (Fig. 2). It would appear, then, that the contrast in
the behavior of pyrimidine and purine nucleotide groups against ribonu-
clease must be interpreted as a specific resistance of the purine nucleotide
groups against the action of ribonuclease rather than by a specific impor-
tance of the pyrimidine group for the catalytic activity of this enzyme.

Mechanism of Action of Ribonuclease (see also Chapter 12). The enzymic
cleavage of internucleotide bonds by ribonuclease is not a simple hydrolysis
but an intramolecular transphosphorylation followed by hydrolysis. The
presence of transphosphorylating and hydrolyzing activities in the same
enzyme is not a unique property of ribonuclease, but has recently been
observed in phosphomonoesterases. However, transphosphorylation as a
necessary intermediary step of hydrolysis on the substrate level has so
far been established only for the action of ribonuclease on polynucleotides.-

The concept of the biphasic nature of the cleavage of ribonucleic acids
by ribonuclease is based on the following observations. Chantrenne, Linder-
str0m-Lang, and Vandendriessche'^"-^ made the interesting discovery that
the volume changes of a ribonucleate solution during the action of ribo-
nuclease were biphasic: A short and transitory dilatation of the solution
was followed by a slow contraction. Since the simple hydrolytic clea\age
of an ester bond would be expected to result exclusively in a small con-
traction, these authors concluded that the cleavage of the internucleotide
bonds of ribonucleic acid by ribonuclease was not a simple hydrolysis, but
that it followed a more complex mechanism.

2' R. B. Merrifield and D. W. Woolley, J. Biol. Chem. 197, 521 (1952).

"^ As contrasted with hypothetical transphosphorylation between an "active" en-
zyme phosphate compound and the substrate.

*' H. Chantrenne, K. Linderstr0m-Lang, and L. Vandendriessche, Nature 159, 877
(1947).

^* L. Vandendriessche, Conipt. rend. trav. lab. Carlsberg, Ser. chim. 27, 341 (1951).

-^ L. Vandendriessche, Acta Chem. Scand. 7, 699 (1953).

564

G. SCHMIDT

(RNA-ase)

(RNA-ase)

-^ (RNA-ase)

Fig. 4. Interlinkages in oligo- and polynucleotides and in ribonucleic acids cleaved
by ribonuclease I. [Modified from the scheme of W. E. Cohn et al. in "Phosphorus
Metabolism" (McElroy and Glass, eds.), Vol. 2, p. 344. Johns Hopkins Press, Balti-
more, 1952.]

These physicochemical changes are explained by observations of Brown,
Dekker, and Todd^^ and of Markham and Smith" that cychc 2',3'-nucleo-
side phosphoric acid diesters are- intermediaries in the cleavage of the
internucleotide bonds by ribonuclease.^^ In a second step the cyclic pyrim-
idine nucleotides are hydrolyzed by ribonuclease to ordinary 3 '-nucleotides
(or nucleotide groups). The diphasic enzymic cleavage of the internucleo-
tide bonds is a peculiarity of the action of ribonuclease I on ribonucleic acid
and ribopolynucleotides. These substrates have the structural requirements
for such a reaction mechanism because they have 2 '-hydroxy groups in
the vicinal position to the corresponding 3'-phosphoryl groups of the
internucleotide bonds (Fig. 4). The action of deoxyribonuclease on deoxy-
ribonucleate causes only the usual volume contraction of 2 cm. per equiva-
lent and is not accompanied by an intermediary phase of dilatation.

"D. M. Brown, C. A. Dekker, and A. R. Todd, /. Chem. Soc. 1952, 512.

" R. Markham and J. D. Smith, Biochem. J. 52, 552 (1952).

" L. A. Heppel, P. R. Whitfeld, and R. Markham, Biochem. J. 56, Proc. iii (1954).

ENZYMES ATTACKING NUCLEIC ACIDS 565

The intermediary formation of 2 ',3 '-cyclic pyrimidine nucleotides was
suggested by the observation that they could be detected by paper iono-
phoresis of ribonuclease digests obtained by very short incubation of
ribonucleate with very small quantities of the enzyme. They could be
accumulated in considerable ciuantities when the digestion was carried out
in dialysis bags so that the cyclic nucleotides were separated from the
enzj'me soon after they had been formed. The absence of cyclic nucleotides
in later stages of the incubation or in digests obtained with large amounts
of ribonuclease is explained by the observation that ribonuclease catalyzes
the hydrolysis of cyclic pyrimidine nucleotides to the corresponding 3'-
nucleotides. As shown hi Fig. 3, it was concluded that the cleavage of
the phosphoryl diester bonds by ribonuclease is not a simple hydroly-
sis, but a transphosphorylation (step I) followed by hydrolysis (step
II). If A is a pyrimidine nucleotide with a 3 '-phosphoryl group, and B
the adjacent nucleotide esterified with this phosphoryl group at the 5 '-pos-
ition, the first rapid step of ribonuclease action consists in the shift of the
5'-phosphoryl ester bond of B to the 2'-position of A with the formation
of a cyclic diester of phosphoric acid (cj^clic nucleotide group or cyclic
3 '-nucleotide). The second slow step in ribonuclease action is the spe-
cific enzymic hydrolysis of the cyclic diester at the 2 '-phosphoryl ester
linkage with the formation of a 3 '-nucleotide group (or nucleotide).

Nucleotide-P -esters as Substrates of Ribonuclease. Brown and Todd
succeeded in synthesizing model esters of nucleotides with benzyl, methyl
and ethyl alcohols. It was found that the esters of 3'-cytidine and 3'-
uridine phosphates were split by ribonuclease, but that those of the 5'-
nucleotides and all esters of adenine nucleotides were resistant to the
enzyme. The intermediary formation of cyclic nucleotides was demon-
strated during the cleavage of all synthetic nucleotide esters susceptible
to the action of the enzyme.^^

The behavior of synthetic nucleotide-P-esters of known structure is thus
in complete analogy to that of polynucleotides and shows that only one of
the two ester bonds of phosphoric acid must be attached to a pyrimidine
nucleoside to make the compound susceptible to the action of ribonuclease,
and that this bond must be attached to the 3'-hydro.\y group of the ribose
moiety.

Some Synthetic Reactions Catalyzed by Ribonuclease. The conversion of
internucleotide bonds to cyclic phosphodiester bonds by ribonuclease raises
the c}uestion as to whether or not the reverse reaction might occur bet^^■een
cyclic nucleotides and other hydroxy compounds under the influence of
this enzyme. Whereas the former reaction results in degradation, the
latter would lead to the formation of larger molecules. Model reactions
of such a synthetic action of ribonuclease have been demonstrated by

566 G. SCHMIDT

Heppel and Whitfeld,^^ ■^*' who found that ribonuclease catalyzes the en-
zymic synthesis of benzyl- and of methylcytidylic acid from cyclic 2',3'-
cytidylic acid and the corresponding alcohols (which must be present in
excess). Similarly, the incubation of a mixture of cyclic 2',3'-cytidylic
acid with cytidine results in the formation of the nucleotide-nucleoside-P-
diester cytidylyl-cytidine,^" that of cyclic uridylic acid and cytidine in the
formation of uridylyl-cytidine. The observed transesterification reactions
catalyzed by ribonuclease are not limited to the cyclic nucleotides as
substrates, but occur between acyclic P diesters of pyrimidine-3 '-nucleo-
tides and other alcohols: for example, the incubation of a mixture of P-
methyl-3'-cytidylic acid with cytidine results in the formation of cytidylyl-
cytidine and methyl alcohol. Finally, the formation of oligonucleotides by
enzymic exchange reactions of cyclic nucleotides with pyrimidine-3'-
nucleotides were reported by Heppel et al.^^^

The possible occurrence of such exchange reactions must of course be
carefully considered in the interpretation of experiments carried out with
ribonuclease as a hydrolyzing agent in investigations concerning the
structure of PNA.

Principles of Assay of Ribonuclease. (a) Formation of acid-soluble degrada-
tion products of PNA.^^-^"- The mononucleotides and many oligonucleotides
are soluble at pH values of 2, whereas PNA is precipitated under these con-
ditions. Hydrochloric (0.5 A^), sulfuric, perchloric acids or glacial acetic acid
have been used for such partitions. The precipitation of undigested nucleic
acid in easily filtrable form is facilitated by the use of alcohol-containing
acid solutions.^^ The fine dispersion of nucleic acid precipitates obtained in
very dilute substrate solutions with acjueous acids has been used as a
principle for a turbidimetric assay method for ribonuclease.^* The acid-
soluble hydrolysis products can be determined in the filtrates by phos-
phorus determination or by ultraviolet spectrophotometry. Roth and
Milstein^' used P'--labeled PNA as substrate and determined the radio-
activity of the filtrates obtained after incubation with ribonuclease-con-
taining tissue extracts.

A very suitable reagent for the assay of ribonuclease by determination
of its acid-soluble degradation products is 1.5% uranyl chloride in 10% of
trichloroacetic acid.^^ According to MacFadyen,'^ this reagent precipitates

29 L. A. Heppel and P. R. Whitfeld, Biochem. J., 56, Proc. ii (1954).

30 L. A. Heppel, R. Markham, and R. J. Hilmoe, Nature 171, 1151 (1953).

'"* L. A. Heppel, P. R. Whitfeld, and R. Markham, Abstr. 126th Meeting Am. Chem.

Soc, New York p. 52c (1954).
3' C. E. Carter and J. P. Greenstein, J. Natl. Cancer Inst. 1, 29 (1946).
^'^ A. Cantero, R. Daoust, and G. de Lamirande, Science, 112, 221 (1950).
" J. S. Roth and S. W. Milstein, J. Biol. Chem. 196, 489 (1952).
" M. McCarty, J. Expll. Med. 88, 181 (1948).
36 D. A. MacFadyen, /. Biol. Chem. 107, 297 (1934).

ENZYMES ATTACKING NUCLEIC ACIDS 567

PNA quantitatively in flocculent form even from very dilute solutions,
but does not precipitate mononucleotides and oligonucleotides of low
molecular weight. Care must be taken to have the uranyl chloride present
in excess when concentrated solutions of PNA are used as substrate. The
acid-soluble fraction contains the mononucleotides and probably the
dinucleotides quantitatively; some oligonucleotides of higher molecular
weight, however, are difficultly soluble in acids and are partially precipi-
tated together with PNA.

(6) Formation of titratable acidic groups. The cleavage of each nucleotide
interlinkage involving a diesterified phosphoryl group results in the appear-
ance of a titratable secondary phosphoryl group. The appearance of acidic
groups during ribonuclease action was established by Kunitz^ and by
Allen and Eiler.^^ When ribonucleates are used as substrates, the accuracy
of titration is limited by the possibility of pK shifts during their enzymic
degradation. In addition, the interpretation of the titration curves is
impeded by the overlapping of the dissociation ranges of the amino groups
with the beginning, and of those of the phenolic groups with the end, of
the dissociations of the secondary phosphoryls. According to Schmidt,
Seraidarian, and Thannhauser," this difficulty can be overcome by calcu-
lating the amounts of secondary phosphoryl groups from the slope of the
titration curve between two close pH values in the neighborhood of pH 6
(e.g., between pH 5.9 and pH 6.2). In this range, which is the region of
the p/^2 values of the nucleotides, the dissociation of the amino groups is
negligible.

A simple transformation of the Henderson-Hasselbach eciuation shows
that the amounts of secondary phosphoryl groups (T) in a nucleotide
mixture can be calculated from the e(]uivalents of alkali consumed between
(pH)i and pH 8 (Ui), and those consumed between (pH)2 and pH 8 (U2)
provided (pH)i and (pH)2 are in the range between pH 5.8 and pH 6.4.

The equation is:

(C/i) X

(Hi+)

- 1

r(H,+)

(H,)

- 1

T =

The validity of this equation can easily be checked with pure nucleotides
for which the T values obtained for arbitrary pairs of (pH)i and (pH)2
within the range from pH 5.8 and 6.4 are constant. For nucleic acids or
nucleotide mixtures which contain different secondary phosphoryl groups

" F. W. Allen and J. J. Eiler, J. Biol. Chem. 137, 757 (1941).

" M. Seraidarian, Thesis, Science Faculty, Tufts College, 1952.

568 ^- SCHMIDT

with slightly varying pK values, constancy of the T values prevails only
over a slightly narrower range of pH values, and the T values tend to de-
crease with the shift of the selected pair of (pH) values toward the alkaline
range. It is easy, however, to find the range in which the decrease of T is
minimal. The T value obtained in this range represents a close approxima-
tion to the correct value for the amount of secondary phosphoryl groups
in the titrated sample.

These remarks will be sufficient to show that the increment of acidic
groups during the hydrolysis of ribonucleic acid by ribonuclease can only
be calculated from complete titration curves. This procedure is obviously
too slow for measurements in the initial phases of the hydrolysis, but it is
suitable for the determination of the final extent of hydrolysis. If only
approximate measurements are required, the titrimetric technique can be
adapted to determinations of hydrolysis by adjusting the pH of the sub-
strate solution to a value around 8. The action of added ribonuclease
results in a gradual decrease of the pH of the digest. At certain time inter-
vals, the initial pH is reestablished by the addition of measured amounts
of 0.1 A^ alkali. These amounts correspond to the newly formed acidic
groups if pK shifts during the hydrolysis remain negligible, and if, in the
pH range covered by the measurements, no titratable groups other than
phosphoryl groups are liberated. The occurrence of pK shifts cannot be
excluded, and the selection of pH 8 as end-point is arbitrary, particularly
in view of the fact that the range of the inflection point of nucleotides in
the region is usually very narrow and that the slopes of the titration curves
of most oligonucleotides around these inflection points are rather steep.

Vandendriessche^^ as well as Cavalieri ei aU^ reported titrimetric observa-
tions from which they concluded that the action of ribonuclease results in
the liberation of phenolic hydroxy groups the titration of which overlaps
with that of secondary phosphoryl groups. The existence of internucleotide
linkages involving phenolic hydroxy groups is by no means excluded by
the wealth of recent evidence in favor of the concept that the majority of
the internucleotide linkages are phosphoric acid ester bonds with 3'- and
5'-hydroxy groups, respectively.

(c) Manometric determination of ribonuclease according to Bain and Rusch.^^
The manometric determination of the liberated acidic groups ofi'ers the
advantage that the initial stages of the hydrolysis can be quantitatively
studied. Bain and Rusch reported a straight-line time-activity curve
during the first 30 minutes of hydrolysis when sufficiently dilute enzyme
solutions were used. Under such conditions, proportionality between the
amounts of carbon dioxide developed and between the concentrations of

38 L. F. Cavalieri, S. E. Kerr, and A. Angelos, J. Am. Chevi. Soc. 73, 2567 (1951).

39 J. A. Bain and H. P. Rusch, /. Biol. Chem. 153, 659 (1944).

ENZYMES ATTACKING NUCLEIC ACIDS 569

ribonuclease were obtained. The procedure has been used for the deter-
mination of the concentration of ribonuclease in tissues.

The application of this method to kinetic studies, mainly by Zittle,'^" is
complicated by the corrections which must be applied for the retention of
carbon dioxide. The necessity for this correction interferes particularly
with studies concerning Michaelis-Menten constants or concerning the
effect of phosphate esters on the action of ribonuclease.

(d) The secondary phosphoryl groups of ribonuclease digests can also be de-
termined enzymically with acid prostatic phosphatase as a specific hydrolyz-
ing agent for secondary hydroxy groups. Schmidt et a/.'^ found that mono-
esterified phosphoryl groups of nucleotides are rapidly and completely
hydrolyzed by prostatic phosphatase, whereas diesterified phosphoryl
groups are resistant toward this enzyme. These observations are in agree-
ment with results obtained with chromatographic analyses of ribonuclease
digests." '^^-^ The transformation of phosphodiester groups into phos-
phomonoester groups by ribonuclease explains the fact that exhaustive
digestion with prostate phosphatase releases a much larger amount of
inorganic phosphate from ribonuclease digests than that formed under
similar conditions from ribonucleic acid.^^'^°'^'

So far, the phosphatase method has mainly been used for analyzing
ribonuclease digests in the final phase of hydrolysis, whereas the others
are better suited for enzyme assays or kinetic studies.

(e) Light absorption in the ultraviolet region. The action of ribonuclease on
ribonucleic acids is accompanied by changes in the absorption of ultraviolet
light for which no theoretical explanation can be given as yet. Kunitz'*-
found a decrease of the absorption at 290 mju which can be used for the
assay of purified preparations of ribonuclease.

In the region around 260 m/x, the characteristic range of the ultraviolet
absorption of the bases, the action of ribonuclease causes no appreciable
optical changes.^" '^^ This is of interest since the quantitative degradation
of PNA to mononucleotides by alkali is accompanied by an increase of
approximately 20% in the absorption at 260 m^i (hyperchromic effect).
The smaller absorption of PNA samples in comparison with the sum of
the absorption effects of their mononucleotides is ascribed to an alteration
of the resonance behavior of the bases when they are bound in polynucleo-
tides of relatively high molecular weight. Magasanik and Chargaff'-" ob-
served hj^perchromic effects during the alkali hydrolysis of the high-
molecular, but not of the low-molecular limit polynucleotides obtained by
digestion of PNA with ribonuclease. They concluded from these observa-

"> C. A. Zittle, J. Biol. Chem. 163, 119 (1946).

^' R. A. Bolomey and F. W. Allen, J. Biol. Chem. 144, 113 (1942).

« M. Kunitz, /. Biol. Chem. 164, 563 (1946).

570 G. SCHMIDT

tions that the structures responsible for the comparatively low ultraviolet
absorption are mainly the purine nucleotide groups of the cores.

Observations on the Kinetics of Ribonuclease. Substrates of well-defined
and simple structure, i.e., oligonucleotides containing per molecule only
one susceptible hnkage or a well-defined number of such linkages would
obviously be the most suitable materials for studies of the kinetics of
ribonuclease as well as for its assay. Such substances, e.g., certain dinucleo-
tides^i or cyclic nucleotides,^"'" became available only very recently, and
no detailed kinetic studies on these highly interesting substrates have been
published as yet.

For these reasons, all available kinetic data are based on the action of
ribonuclease on ribonucleates. It is obvious that such data are of very
limited value for the understanding of the action of ribonuclease. Many of
the kinetic studies were carried out on commercial preparations of PNA.
Recent experiences have shown that commercial preparations of yeast
PNA must be considered as more or less degraded products which differ
in many properties such as the number of terminal groups per molecule
from PNA samples prepared by the mildest available procedures in the
laboratory. But even genuine nucleic acid samples offer very complex
conditions for kinetic studies since any ribonucleic acid preparation from
a given tissue probably consists of a mixture of different nucleic acids and
each individual PNA molecule contains a large number of linkages which
are hydrolyzed by ribonuclease. Although some important features are
common to all these linkages, they differ amongst each other in regard to
their structure. It is very possible that the various linkages hydrolyzed by
ribonuclease are cleaved at different rates so that the rate of ribonuclease
action as measured, e.g., by the rate of formation of acidic groups from
PNA, represents an overall rate resulting from a large number of individual
enzyme reactions each of which might have its own characteristic kinetics.

According to the manometric measurements of J. A. Bain and H. P.
Rusch,^^ the effect of the concentration of ribonucleate on the rate of
ribonuclease action is very considerable; the maximal initial velocity is
only approached at a substrate concentration of 6.5% (Fig. 5).

In addition, the curves of Figure 5 (according to Bain and Rusch^^)
demonstrate that relatively high substrate concentrations (at least 5%)
are required in order to obtain constant hydrolysis rates during the initial
phases of the enzyme reaction. At low substrate concentrations, the rates
are falling continuously even during the first seconds of incubation.

This might suggest a strong competitive effect of the hydrolysis products
of ribonuclease action on the enzyme. Actual data concerning such effects,
however, are scanty and controversial. So far, only the effect of added

ENZYMES ATTACKING NUCLEIC ACIDS

571

mononucleotides has been studied and these experiments were carried out
with mixtures of the 2'- and 3 '-isomers.

Zittle"*^ found approximately 50 % inhibition in the presence of guanylic
acid in 0.04 M concentration. Adenylic acid inhibited less than guanylic
acid or than the mixture of the four nucleotides obtained by alkali hy-
drolysis from PNA. He defined the inhibition as noncompetitive, although
the range of substrate concentrations studied appears as too narrow to
permit this conclusion. He also found that commercial nucleic acid prep-
arations (particularly free ribonucleic acid) contained acid-soluble con-

250

250

150

100

50

7.0

8,0

pH

5 10 15

Time, min.

Fig. 5. Fig. 6.

Fig. 5. Effect of variation of substrate level on ribonuclease activity. [From Bain
and Rusch."] Beginning with the highest, the curves represent 200, 150, 75, 50, and
25 mg. of substrate per flask, respectively.

Fig. 6. Effect of pH on the activity of ribonuclease I. [From Bain and Rusch.^^]
3.0-ml. reaction mixture containing 200 mg. of sodium ribonucleate, and varying
concentrations of sodium bicarbonate, together with 20 fig. crystalline ribonuclease.

taminants which considerably inhibited ribonuclease action. This suggests
inhibitory effects of acid soluble ribooligonucleotides. The available data
must be considered as preliminary for reasons explained above. It should
also be mentioned that the manometric determination of ribonuclease
action, particularly in its initial phases, is not very accurate owing to the
necessity to apply considerable and complicated corrections for the reten-
tion of carbon dio.xide.

The kinetic observations just discussed were made before the discovery
of cyclic nucleotides as intermediaries of the hydrolysis of PNA by ribonu-
clease. It is obvious that the understanding of the kinetics of ribonuclease
action requires supplementary data which will permit the interpretation of

" C. A. Zittle, /. Biol. Chem. 160, 527 (1945).

572 G. SCHMIDT

the earlier results in the light of the recent information on the reaction
mechanism of the enzyme.

Influence of pH. The pH optimum of crystallized pancreas ribonuclease
was found by Kunitz^ to be in the region of pH 7.7 (Fig. 6). Appreciable
spontaneous degradation of ribonucleic acids occur already at this pH in
experiments of longer duration. Many studies on ribonuclease action are
therefore carried out at slightly acid pH values in the range between 5 and
6 in which the enzyme has still considerable activity.

Influence of Temperature. At pH 5, the optimal temperature of ribonu-
clease action is 65°.^ In the range between pH 2 and pH 5 ribonuclease is
rather heat-stable. Only about 20% of the activity is lost when ribonuclease
is kept under these conditions at a temperature of 100° for thirty minutes.

Activators, Inhibitors. Ribonuclease I requires for its optimal activity an
ionic strength of about 0.1. The activating effects of various univalent cat-
ions are quantitatively similar and larger than those observed with mag-
nesium ions.^^ '"^^ Zittle observed that contamination of ribonucleates with
copper ions impaired their susceptibility to ribonuclease action.^^' ■** Sodium
fluoride even at 0.1 M concentration is without appreciable effect on
ribonuclease. Zollner and Fellig^^ showed good evidence for a specific
competitive inhibition of ribonuclease activity by heparin. Some other
acidic polysaccharides were without effect. Massart et aU^ reported an
inhibitory influence of penicillin on ribonuclease on the basis of a histochem-
ical assay of enzyme activity.

Ledoux"^ studied the influence of the sulfhydryl reagent p-chloromercuri-
benzoate. He concluded that the complex influence of this substance
resulted on the one hand from the inhibitory efi'ect of its combination with
the sulfur groups of the enzyme, on the other hand, from the activating
influence of its combination with polar groups of the substrate. (See also *^^)

Ribonuclease as a Protein. The crystaflization of ribonuclease was
achieved by Kunitz^ in 1940 by fractionation with ammonium sulfate.
The enzyme prepared according to this procedure retains traces of con-
taminating proteolytic enzymes even after several recrystallizations.^^
Since crystalline ribonuclease is frequently applied as a histochemical
reagent, an important advance was made when McDonald succeeded in
removing these contaminations by introducing a heating step into the pro-

«=» S. R. Dickman, R. B. Knopf, and J. B. Aboskar, Abstr. 126th Meeting Am. Chem.

Soc, New York p. 71C (1954).
■*« The effects of copper ions can be abolished by versene."^
" N. Zollner and J. Fellig, Am. J. Physiol. 173, 223 (1953).
46 L. Massart, G. Peters, and A. Vanhauke, Experientia 3, 494 (1947).
4' L. Ledoux, Biochim. et Biophys. Acta 11, 517 (1953).
« M. R. McDonald, J. Gen. Physiol. 32, 33 (1948).

ENZYMES ATTACKING NUCLEIC ACIDS 573

cedure of purification.'*^ Labeled ribonuclease was prepared by Anfinsen^"
by incubating slices of beef pancreas in the presence of C^'*02 .

Despite the homogeneous behavior of ribonuclease during sedimentation
and electrophoresis (Rothen^^), recent studies of Martin and Porter^^ and
of Plirs, Moore, and Stein^^ with ion-exchange chromatography demon-
strated that crystalline ribonuclease, as well as fresh pancreas extracts,
contains at least two ribonucleases (Ribonuclease lA and IB). Ledoux's^^^
suggestion that the chromatographic inhomogeneity of ribonuclease might
be attributed to different stages of oxidoreduction of its sulfur groups is
not in agreement with observations of other authors.'*^''

Chromatographically purified ribonuclease lA (the slower-moving band)
was obtained in crystalline form, whereas the isolation of ribonuclease IB
has so far not been achieved on the preparative scale. The assay methods
used for the activity determinations on both enzymes were based on the
formation of acid-soluble phosphorus compounds from yeast PNA. At
present, no information regarding the finer enzymic specificities of the two
pancreas ribonucleases is available.

All data regarding the physicochemical properties of ribonuclease are
still based on measurements carried out in 1940 by Rothen^^ on ribonuclease
samples prepared without chromatographic resolution into ribonucleases
lA and IB. According to these measurements, the sedimentation rate of
crystalline ribonuclease I in 0.5 M ammonium sulfate is *S-^ = 1.85 X 10~'^.
The diffusion coefficient in 0.5 M ammonium sulfate is D = 1.36 X 10~^.
These data correspond to a molecular weight of 12,700, which is in good
agreement with the mean value of 13,000 obtained from sedimentation
ecjuilibria by Rothen. Kunitz^ arrived at the value of 15,000 (±1000)
from osmotic pressure measurements. Most of the calculations in the
current literature are based on the assumed value of 13,500. From the most
recent amino acid analyses of Hirs, the value of 14,100 for the molecular
weight of ribonuclease I is obtained. ^'•^^''•^^"=

The isoelectric point of ribonuclease I was found at pH 7.8. The specific

volume has the rather low value of 0.709. Ribonuclease I is a globular

/
protein with a dissymmetry factor r = 1.04.

Jo

*^ M. R. McDonald, J. Gen. Physiol. 32, 39 (1948).
5« C. B. Anfinsen, J. Biol. Chem. 186, 827 (1950).
" A. Rothen, J. Gen. Physiol. 24, 203 (1940).
« A. J. P. Martin and R. R. Porter, Biochem. J. 49, 215 (1951).
" C. H. W. Hirs, S. Moore, and W. H. Stein, J. Biol. Chem. 200, 493 (1953).
"» L. Ledoux, Biochim. et Biophrjs. Acta 14, 267 (1954).
"b C. H. W. Hirs, Federation Proc. 13, 230 (1954).

"<= C. H. W. Hirs, W. H. Stein, and S. Moore, Abstr. 126th Meeting Am. Chem. Sac,
New York p. 89C (1954).

574 G. SCHMIDT

Observations of Anfinsen et al}^-^^ suggest that ribonuclease I contains
only one C terminal group (valine) and one N terminal group (lysine).
The analysis of ribonuclease I by means of X-ray diffraction suggests the
presence of five crystallographic chains per molecule.^® Elliott" studied
the infrared spectrum of single crystals of ribonuclease I and observed
dichroism indicating the presence of folded polypeptide chains. According
to Anfinsen, the available information suggests the working hypothesis
that ribonuclease I consists of a single peptide chain which is folded to the
five crystallographic chains postulated on the basis of the X-ray diffraction
patterns of ribonuclease I crystals. ^^ Possibly the four cystine groups" '^^-ei
present in each molecule of ribonuclease are involved in maintaining the
folded structure of the protein (Fig. 7). According to Anfinsen,^^ ^^g g^-

[NHJ

LYS-GLU-THR-ALA

r

[PRO]

S ^

i [PRO]

r

[PRO]

[PRO]

■ [MET,TYR,ALA,LEU,PHE]- VAL

Fig. 7. Scheme of the shape of the molecule of ribonuclease I. [From Anfinsen ei
al.^^\ (The abbreviations are symbols for amino acids.)

zymic activity of ribonuclease I is destroyed by digestion with crystallized
pepsin. During the first phase of this digestion approximately 65% of the
activity is lost without a measurable appearance of new end-groups and
without a significant decrease of the sedimentation rate.^^ This slow phase

'^ C. B. Anfinsen, M. Flavin, and J. Farnsworth, Biochim. et Biophys. Acta 9, 468

(1952).
" C. B. Anfinsen, R. R. Redfield, W. L. Choate, J. Page, and W. R. Carroll, /. Biol.

Chem. 207, 201 (1954).
" H. Carlisle and H. Scouloudi, Proc. Roy. Soc. (London) A207, 496 (1951).
"A. Elliott, Proc. Roy. Soc. (London) A211, 490 (1952).
68 E. Brand, Ann. N. Y. Acad. Sci. 47, 187 (1946).
6' G. R. Tristram, Advances in Protein Chem. 5, 145 (1949).
*" J. H. Northrop, M. Kunitz, and R. M. Herriott, "Crystalline Enzymes," p. 26,

New York, 1948.
*' "Symposium on Mechanism of Enzyme Action" (W. D. McElroy and B. Glass,

eds.). Johns Hopkins Press, Baltimore, 1953.
«2C. B. Anfinsen, /. Biol. Chem. 196, 201 (1952).
83 D. Shugar, Biochem. J. 52, 142 (1952).

ENZYMES ATTACKING NUCLEIC ACIDS 575

is followed by a rapid deeper cleavage of the ribonuclease I accompanied
by a practically total loss of activity. Hirs^^'^ and Hirs, Stein, and Moore*^''
applied oxidation of ribonuclease with performic acid at — 10° and subse-
cjuent digestion of the oxidized enzyme with trypsin to the investigation of
the amino acid sequence in the ribonuclease molecule. The treatment of
the protein with performic acid oxidized exclusively the sulfur groups of
the cystine and methionine components. The number of polypeptides ob-
tained by chromatography of the tryptic digest closely agreed with that
predicted from the experimentally determined values of 10 lysine and 4
arginine rests for each molecule of ribonuclease.

b. Other Ribomiclcases

The existence of ribonucleases other than the enzyme obtained in crystal-
lized form from pancreas can now be considered as certain. They differ
from pancreas ribonuclease in regard to their specificity toward the inter-
nucleotide bonds of the substrate, in regard to their heat lability and to
their pH optima. Schmidt ct a/.^^ found in 1950 that incubation of PNA
with crude pancreas extracts resulted in the cleavage of internucleotide
bonds which were resistant to the action of crystallized pancreas ribonu-
clease I. These bonds were the ester linkages of 3 '-purine nucleotide groups
with the adjacent nucleotide groups. Maver and Greco®^ reported the
presence in spleen extracts of a heat-labile ribonuclease which differed
from ribonuclease I also by the different pH range (optimum in the region
of pH 5.2, in the presence of Mg; at pH 6.6 without addition of Mg salts^^)
of its optimal activity.

Clear evidence for the existence of different ribonucleases and for the
nature of their catalytic activity was obtained by Hilmoe and Heppel,"
who fractionated the nucleolytic enzymes of spleen extracts. They studied
the action of these enzymes not only on PNA, but on the limit polynucleo-
tides obtained by exhaustive hydrolysis of PNA with ribonuclease I as
well as on some simple synthetic nucleotide-P-esters of well-defined struc-
ture. Hilmoe and Heppel prepared from spleen a ribonuclease which was
free from phosphomonoesterases and which hydrolyzed the limit poly-
nucleotide fraction of PNA approximately four times faster than PNA
itself. The pH optimum of this enzyme was at pH 6.6. The enzyme was
rapidly inactivated at 60°.

Mechanism of Action. The spleen ribonuclease does not hydrolyze cyclic
nucleotides. It is capable of catalyzing exchange reactions between nucleo-

"^ G. Schmidt, R. Cubiles, and S. J. Thannhauser, /. Cellular Coywp. Physiol 38,

Suppl. 1,61 (1950).
«* M. E. Maver and A. E. Greco, /. Biol. Chem. 181, 861 (1949).
6« M. E. Maver and A. E. Greco, Federation Proc. 13, 261 (1954).
" R.. J. Hilmoe and L. A. Heppel, Federation Proc. 12, 217 (1953).

576 G. SCHMIDT

side-containing diesters of phosphoric acid, such as:

2 cytidine-3'-benzylphosphate

= cytidylyl-3',5'-cytidylyl-3'-benzylphosphate + benzyl alcohol

Ribopolynucleotidases of Plants. Bredereck et al.^^-^^ found that aqueous
extracts of sweet almonds, lucern seeds, or sprouted peas contain an en-
zyme system capable of converting ribonucleate practically completely to
the nucleosides and phosphoric acid, at a pH range between 5 and 6. A
fractionation of this system has so far not been attempted.

Bacterial Rihonucleases. Muggleton and Webb^" found in culture filtrates
of soil actinomyces (strain A) a very heat-labile ribonuclease, which was
capable of hydrolyzing PNA as well as the polynucleotides resistant toward
ribonuclease I. The presence of this ribonuclease is responsible for the
power of the culture filtrates to render suspensions of heat-killed pneumo-
cocci gram-negative.

2. Deoxyribonucleases

a. Pancreas Deoxyribonuclease {Deoxy ribonuclease I)

History. Araki^"^ observed in 1903 that extracts of several tissues such
as liver, spleen, and thymus had the power to liquefy gels of deoxyri-
bonucleic acid. Abderhalden and Schittenhelm" (1906) as well as de la
Blanchardiere^i'^ (1913) found that pancreatic juice of dogs effected this
liquefaction without liberation of purine bases or inorganic phosphate.
Feulgen72.73 added the important observation that the degradation of
deoxyribonucleic acid by pancreas preparations (pancreatin) did not
yield mononucleotides, but stopped at the formation of oligonucleotides.
Schmidt, Pickels, and Levene^" found in 1938 that the degradation of DNA
to oligonucleotides by the pancreas enzyme or by alkali was an essential
intermediary step for the enzymic cleavage of DNA to mononucleotides
(and subsequently to nucleosides) by the phosphatase of intestinal mucosa.
Intestinal phosphatase does not act on highly polymerized DNA, but it
hydrolyzes all interlinkages of the mononucleotide groups in the oligo-
nucleotide mixture obtained by the action of deoxyribonuclease. Some-

«8 H. Bredereck and G. Rothe, Ber. 71B, 408 (1938).

«» H. Bredereck, G. Caro, and F. Richter, Ber. 71B, 2389 (1938).

70 P. W. Muggleton and M. Webb, Biochim. et Biophys. Acta 9, 343 (1952).

70a T. Araki, Z. physiol. Chem. 38, 84 (1903).

" E. Abderhalden and A. Schittenhelm, Z. physiol. Chem. 47, 452 (1906).

"a P. de la Blanchardiere, Z. physiol. Chem. 87, 291 (1913).

" R. Feulgen, "Chemie und Physiologie der Nucleinstoffe." Berlin, 1923.

73 R. Feulgen, Z. physiol. Chem. 237, 261 (1935).

7" G. Schmidt, E. G. Pickels, and P. A. Levene, /. Biol. Chem. 127, 251 (1939).

ENZYMES ATTACKING NUCLEIC ACIDS 577

what earlier, Schmidt^^ had ah-eady found that fresh extracts of nucleo-
histone from thymus glands were not hydrolyzed by intestinal phosphatase,
but that they were dephosphorylated by this enzyme after a preceding
incubation with crude trypsin. This effect was undoubtedly due to the
enzymic depolymerization of the nucleic acid component of the nucleo-
histone by the deoxyribonuclease present in the trypsin preparations, but
not to proteolysis. Laskowski^^ (1946) and McCarty" achieved a consider-
able purification of the enzyme obtained from pancreas. The most essential
contribution of this work was the separation of deoxyribonuclease from
the powerful ribonuclease of pancreas and the evidence for the strict
specificity of deoxyribonuclease for DNA. The crystallization of the enzyme
which precipitates at much lower ammonium sulfate concentration (0.4
sat.) than does ribonuclease I (0.8 sat.) was achieved by Kunitz^'' in 1950.

An interesting historical fact is the influence which the now abandoned
working hypothesis of the tetranucleotide structure of nucleic acids had on
the interpretation of the action of deoxyribonuclease. All investigators
agreed that, after exhaustive incubation of DNA with deoxyribonuclease,
about one out of four phosphoryl groups is present as a terminal secondary
phosphoryl group. For a long time this observation was considered as
evidence suggesting that the main products of the enzymic degradation
were tetranucleotides and that, therefore, natural DNA should be con-
sidered as polymers of tetranucleotide units." This view was predominant
until, in 1951, chromatography was developed by Cohn and Volkin, and
Smith and Markham, as a tool for the fractionation of oligonucleotide
mixtures. We know now that there is no correlation between the relative
amount of terminal phosphoryl groups and the chain length of the oligo-
nucleotide units, and that deoxyribonuclease digests of DNA contain
dinucleotides as well as hexanucleotides and oligonucleotides of other
degrees of polymerization.

Specificity. The number of substrates tested for their behavior toward
deoxyribonuclease is far less extensive than that studied with ribonuclease I.
Nevertheless, the available observations justify the conclusion that deoxy-
ribonuclease is a highly specific phosphodiesterase; in particular, it is
without any effect on ribonucleic acids and ribopolynucleotides. In analogy
to the action of ribonucleases, that of deoxyribonuclease I never results in
the formation of inorganic phosphate. According to Tamm, Shapiro, and
Chargaff,^'" the degree of polymerization of DNA is not of essential influence
on the rate and on the extent of the action of deoxyribonuclease; partial

7*G. Schmidt, Enzymologia 1, 135 (1936).

'« M. Laskowski, Arch. Biochem. H, 41 (1946).

" M. McCarty, J. Gen. Physiol. 32, 39 (1948).

" C. Tamm, H. S. Shapiro, and E. Chargaff, J. Biol. Chem. 199, 313 (1952).

578 G. SCHMIDT

or total removal of the purine groups, however, greatly diminishes the
catalytic effects of the enzyme. It should be mentioned that Tamm et al.
observed in the absence of deoxyribonuclease I degrading effects of the
bivalent cations (Mg++, Mn++) required as activators of deoxyribonuclease
I on the deoxypolynucleotides of lower molecular weights. These effects
obviously render it difficult to evaluate accurately the action of deoxyribo-
nuclease on DNA fragments of lower molecular weights. Little and Butler^^
isolated the nucleotide fraction formed by exhaustive enzymic degradation
of highly polymerized (acid-insoluble) DNA to acid-soluble nucleotides.
The titration of this fraction demonstrated the formation of secondary
phosphoryl groups amounting to 25 % of the total phosphoryl groups. This
result is in agreement with earlier observations of Fischer, Bottger, and
Lehmann-Echternacht.^" Gordon and Reichard,^^ Sinsheimer and Koer-
ner,^2 ,83 ,83a as y^Qii as Smith and Markham^^ have succeeded recently in
identifying by chromatography on paper and on ion-exchange columns
some of the oligonucleotides as dinucleotides and trinucleotides. So far,
deoxyadenyl-cytidylic, adenyl-thymidylic, guanyl-thymidyhc, thymidylyl-
thymidylic, cytidylyl-cytidylic, methylcytidylyl-cytidylic, methylcytidy-
lyl-guanylic, adenyl-adenylic, adenyl-guanylic, guanyl-guanylic acids and
a trinucleotide consisting of one thymidylic and two cytidylic acid groups
were identified amongst the degradation products present in deoxyribonu-
clease digests of calf thymus DNA. Digests of wheat germ DNA contained
methylcytidylyl-cytidylic and methylcytidylyl-adenylic acids in addition
to the dinucleotides just mentioned. It is interesting that not all dinucleo-
tides which have been analyzed up to now contain a pyrimidine component
in contrast to the oligonucleotides formed from PNA by the action of
ribonuclease. The amounts of mononucleotides formed by the action of
deoxyribonuclease on DNA are much smaller than those formed by ribo-
nuclease from PNA. The first- evidence for the formation of mononucleo-
tides by the hydrolysis of DNA with deoxyribonuclease I was obtained
by Potter, Brown, and Laskowski,^^ who detected the appearance of
appreciable quantities of deoxycytidylic acid in such digests. They repre-
sent only approximately 1 % of the total phosphorus of the substrate.*^^

79 J. A. Little and G. C. Butler, J. Biol. Chem. 188, 695 (1951).

ST. G. Fischer, I. Bottger, and H. Lehmann-Echternacht, Z. physwl. Chem. 271,
246 (1941).

81 A. H. Gordon and P. Reichard, Biochem. J. 48, 569 (1951).

82 R. L. Sinsheimer and J. F. Koerner, Science 114, 42 (1951).

83 R. L. Sinsheimer and J. F. Koerner, /. Am. Chem. Soc. 74, 283 (1952).
83» R. L. Sinsheimer, J. Biol. Chem. 208, 445 (1954).

84 J. D. Smith and R. Markham, Biochim. et Biophys. Ada 8, 350 (1952).

85 J. L. Potter, K. D. Brown, and M. Laskowski, Biochim. et Biophys. Acta 9, 150
(1952).

ENZYMES ATTACKING NUCLEIC ACIDS 579

In deoxyribonuclease digests of calf thymus DNA, thymidylic, cytidylic,
adenylic, and guanylic acids were detected; in those of wheat germ DNA,
methylcytidylic acid was found in addition to the other mononucleotides.
The amounts of thymidylic acid accounted for 50 to 60% of the total
mononucleotides formed.

Dinucleotides accounted for 15 to 18% of the substrate phosphorus.
Although most of the dinucleotides contained a pyrimidine group, appre-
ciable amounts of purine dinucleotides were found in deoxyribonuclease
digests of calf thymus and wheat germ DNA.

From the fact that exclusively 5 '-mononucleotides were obtained by
the action of phosphodiesterase^^ on oligonucleotides obtained with deoxy-
ribonuclease I, one can imply that the terminal secondary phosphoryl
groups of these oligonucleotides are attached to the 5 '-positions of the
corresponding deoxyribose moieties.

An appreciable portion of the higher oligonucleotides of deoxyribonu-
clease I digests is not dialyzable against water. This undialyzable fraction
was designated by Zamenhof and Chargaff*® as "cores." These cores have
much higher adenine: guanine, thymine :cytosine, and purine: pyrimidine
ratios than are found in the original substrate. The possibility of a correla-
tion between the sequence of the bases in DNA and the points of attack
of deoxyribonuclease is not yet clear, but it appears that some seeming
resemblance of deoxyribonuclease and ribonuclease action in this respect
does not justify the assumption of similar underlying specificities.

The effect of deoxyribonuclease on the viscosity of DNA has recently
been used for some therapeutic purposes. ^""^

Assay Methods. It is obvious that some of the tests for deoxyribonuclease
activity resemble in principle those for ribonuclease activity and need not
be discussed in detail. Formation of acidic groups: The transformation of
primary into secondary phosphoryl groups can be measured by titration
or manometric methods. The conditions for such measurements with DNA
as substrate are more favorable than similar determinations of ribonuclease
activity because of the negligible amounts of preformed secondary phos-
phoryl groups in highly polymerized DNA. Formation of acid-soluble
phosphorus or deoxyribose compounds: The merits of these assay methods
have been discussed by Laskowski,''^ Kurnick,*^ and Allfrey and Mirsky.^"^
The determination of acid-soluble P compounds in deoxyribonuclease

8« S. Zamenhof and E. Chargaff, J. Biol. Chem. 187, 1 (1950).
86" J. N. Davidson, Brit. Med. Bull. 9, 154 (1953).
" N. B. Kurnick, Arch. Biochem. 29, 41 (1950).
s"* V. G. Allfrey and A. E. Mirsky, J. Gen. Physiol. 36, 227 (1952).
"■^ S. G. Laland, W. A. Lee, W. G. Overend, and A. R. Peacocke, Biochim. el Biophys.
^c<a 14, 356 (1954).

580 G. SCHMIDT

digests is likewise more accurate than that in ribonuclease digests owing
to the practically complete insolubility of DNA even in very dilute solu-
tions. The colorimetric determination of acid-soluble deoxyribose com-
pounds by means of Dische's reagent is particularly convenient according
to Allfrey and Mirsky since the time-consuming ashing procedure required
for determination of the acid-soluble phosphorus is eliminated. Optical
changes: Kunitz^^ found that the action of deoxyribonuclease on DNA is
accompanied by an increase of absorption at 260 m/x amounting maximally
to 30%. Under standard conditions, the initial increase per minute of the
optical density is proportional to the amounts of deoxyribonuclease, and
its determination is a useful assay method particularly for purified prepara-
tions of deoxyribonuclease. Similar optical changes were observed by
Kurnick^^ during the nonenzymic depolymerization of DNA by heat.
Viscosity changes: The hydrolysis of DNA is accompanied by a rapid
decrease of the viscosity of the enzyme-substrate mixtures. This pheno-
menon is a useful basis for enzyme determinations since the initial rates of
decrease are proportional to the amounts of deoxyribonuclease." '^^-^^
Since the viscosity of DNA solutions is influenced by various other factors,
such assays must be carried out under strictly comparable conditions,
particularly in regard to electrolyte concentrations. Binding of methyl
green {Kurnick's method) •?'^ '^"^ Kurnick found that methyl green (ethylated
hexamethylpararosaniline) combines only with highly polymerized DNA,
and that the color of this compound is stable at pH 7.5 in contrast to free
methyl green which fades at this pH. When DNA containing methyl
green is used as substrate solution, the depolymerization of DNA by
deoxyribonuclease (or by heat) is accompanied by the liberation of a
corresponding amount of methyl green. The rate of decrease of absorption
at 640 m/x at pH 7.5 can be used as a measure of deoxyribonuclease activity.
Since the fading of the color of the "liberated" methyl green is not instan-
taneous, the enzyme action is stopped by the addition of citrate, and the
colorimetric readings are carried out after four hours' standing.

The behavior of another basic dye, methylene blue, during the action of
deoxyribonuclease on DNA was studied by Vercauteren.^^-^^ The binding
of this dye, unlike that of methyl green, does not require a very high degree

88 N. B. Kurnick, /. Gen. Physiol. 33, 243 (1950).

89 J. P. Greenstein and W. V. Jenrette, Cold Spring Harbor Symposia Quant. Biol. 9,
236 (1941).

s" J. P. Greenstein, J. Natl. Cancer Inst. 2, 357 (1942).

" M. McCarty, J. Gen. Physiol. 29, 123 (1946).

« N. B. Kurnick, Arch. Biochem. and Biophys. 43, 97 (1953).

93 R. Vercauteren, Nature 165, 603 (1950).

" R. Vercauteren, Enzymologia 14, 134 (1950).

" R. Vercauteren, Arch, intern, physiol. 57, 214 (1949).

ENZYMES ATTACKING NUCI EIC ACIDS 581

of polymerization of DXA; its liberation (as measured by the appearance
of methylene blue in the supernatant solution after the precipitation of
the dye-polynucleotide compounds with acid) begins only in the later
phases of deoxyribonuclease action. The characteristic initial drop of
viscosity is therefore not accompanied by the liberation of methylene
blue. This interpretation is in analogy with the observations of Jungner
et al.,^^'^'' who found that the dielectric constants of deoxyribonuclease
digests remained unchanged during phases of rapid decreases of viscosity.

Most methods of assay yield satisfactory results with purified deoxyribo-
nuclease preparations but are not easily adaptable to enzyme assays in
crude tissue extracts or serum. Kurnick's procedure, however, is applicable
to serum. It should be mentioned that Alfert reported evidence suggesting
that changes of methyl green binding by DXA can be caused by factors
other than depolymerization.^*

Kinetics. The kinetics of the degradation of DXA by deoxyribonuclease
I is obviously an overall picture of many different individual enzyme
processes. Xo simple substrates for deoxyribonuclease are available as yet.

At low concentrations of DXA (160 )ug. per ml.) the hydrolysis curve
follows the course of a pseudounimolecular reaction, and the velocity
constants are proportional to the concentrations of deoxyribonuclease in
the digestion mixture.^^ At high concentrations of the substrate, the hydrol-
ysis curve is an asymmetric S-shaped curve owing to a short initial lag
period. ^^

Activators. Unlike ribonuclease I, deoxyribonuclease recjuires for its
action the presence of bivalent cations.^'' •^^•^^•^^•^°°-'''^ Magnesium, man-
ganese, ferrous, and cobaltous ions are the most effective activators and
resemble each other in regard to the quantitative intensity of activation.
The observations regarding the effects of some other cations, such as zinc
and calcium, are contradictory.^"^ The concentration rerjuired for optimal
effects of the activating cations is practically independent of the enzyme
concentration but increases with increasing substrate concentrations.

9« G. Jungner, I. Jungner, and L. G. Allg^n, Kahire 163, 849 (1949).
"G. Jungner, Ada Physiol. Scand. 10, Suppl. 32, 1 (1945).
98 M. Alfert, Biol. Bull. 103, 145 (1952).
"9 J. Gr^goire, Compt. rend. 231, 384 (1950).

100 F. G. Fischer, H. Lehmann-Echternacht, and I. Bottger, J. prakt. Chem. 159, 59
(1941).

101 M. Laskowski and M. K. Seidel, Arch. Biochem. 7, 465 (1945).

io» C. E. Carter and J. P. Greenstein, J. Natl. Cancer Inst. 7, 29 (1946).

1" J. P. Greenstein, C. E. Carter and H. W. Chalkley, Cold Spring Harbor Symposia

Quant. Biol. 12, 64 (1947).
10^ W. G. Overend and M. Webb, Research 2, 99 (1949).

lo^ T. Miyaji and J. P. Greenstein, Arch. Biochem. and Biophys. 32, 414 (1951).
»<" L. M. Gilbert, W. G. Overend, and M. Webb, Exptl. Cell Research 2, 349 (1951).

582 G. SCHMIDT

Optimal activation by magnesium ions was observed at a molar proportion
of 3 between the magnesium ions and the total phosphorus of the sub-
strate. Increases of the total salt concentration of the incubation mixture
beyond 0.02 M are progressively inhibitory.

The observations just reported suggest that the salt activation of deoxy-
ribonuclease is produced by the combination of the ions with the substrate
and not by the formation of an enzyme-activator compound.'"'^ '^"^

General Stability and Influence of Heat. Solutions of deoxyribonuclease
of relatively high concentration (0.1 mg. of enzyme protein per ml. of
buffer solutions) retain their activities for several days at 5°. Highly dilute
solutions, however, require the presence of stabilizing colloids (gelatin,
peptones) for preservation.

When solutions of deoxyribonuclease I are kept at 90° for 5 minutes at
pH 2.8, the activity disappears completely, but most of the original activity
is restored by standing at 20° for 30 hours.

Inhibitors. (1) Substances removing activating bivalent cations. The re-
quirement of certain bivalent cations for the activity of deoxyribonuclease
I explains the inhibitory effect of certain anions such as citrate and fluoride
on this enzyme.

(2) Other inhibitors. Copper, selenite, and arsenate ions are strongly
inhibitory at 0.001 M, borate at 0.02 M. Hydroxylamine, cysteine, and
iodoacetic acid have no appreciable effects. Mitotic poisons, such as col-
chicine and aminopterine, and SH-reactants, such as iodoacetamide and
chloroacetophenone, caused no inhibition of the action of deoxyribonuclease
I on deoxyribonucleohistone. Incubation of nucleohistone with nitrogen
mustards resulted in a diminished susceptibility of the nucleoprotein
toward deoxyribonuclease.

(3) Specific inhibitors of deoxyribonuclease in tissues. Tissue inhibitors of
deoxyribonucleases are of great biological interest because of their possible
regulatory function in the DNA metabolism of growing cells. ^"^ Recently,
protein inhibitors of deoxyribonuclease I were found in several tissues.
Dabrowska, Cooper, and Laskowski purified such a protein from growing
pigeon crop glands."" On the basis of kinetic studies, Laskowski and his
associates concluded that this effect was caused by a reversible association
of a specific protein with deoxyribonuclease I. The inhibitor protein is easily
extractable by water from the tissue, and is precipitated by ammonium
sulfate between 0.3 and 0.6 saturation. It is destroyed by heat and by the
action of trypsin. Protein fractions with similar inhibitory effects on

107 N. Weissman and J. Fisher, /. Biol. Chem. 178, 1007 (1949).

108 V. S. Shapot, Biokhimiya 17, 290 (1952).

109 S. Zamenhof and E. Chargaff, J. Biol. Chem. 180, 727 (1949).

110 W. Dabrowska, E. J. Cooper, and M. Laskowski, /. Biol. Chem. VII, 991 (1949).

ENZYMES ATTACKING NUCLEIC ACIDS 583

deoxyribonuclease I were also found in other animal tissues and in certain
malignant tumors. In the pigeon crop glands, the concentration of the
inhibitor protein is strongly increased during the hypertrophy of the
epithelium occurring in glands of brooding birds; in other animal tissues,
however, no definite correlation between inhibitor concentrations and cell
growth could be established so far. Cooper, Trautman, and Laskowski
demonstrated the presence of inhibition of deoxyribonuclease I in various
mammalian tissues.^" Henstell and Freedman"' reported considerable
inhibitory influence of extracts from mature white blood cells on pancreas
deoxyribonuclease I. The concentration of this inhibiting activity was
greatly diminished in white blood cells of leukemia patients"^ provided
that immature cells were predominant in the blood. Kurnick ct ai."'* corre-
lated changes of the inhibiting activity of the white blood cells with the
characteristic appearance of the L. E. cells in the blood and bone marrow
of patients with lupus erythematosus."* McCarty demonstrated specific
antigenic effects of purified preparations of pancreas deoxyribonuclease."®
Bernheimer and Ruffier"^ found that extracts of streptococci contained an
inhibitor of pancreas deoxyribonuclease I, the activity of which was de-
stroyed by the action of ribonuclease."^

pH Optimum. Deoxyribonuclease I has maximal activity between pH 6
and pH 7. The precise values of the pH optimum differ somewhat for
different kinds of activating bivalent cations. ^°*

Deoxynbonudease I as a Protein. Owing to the small yields (Kunitz
obtained from 10 kg. of pancreas approximately 20 mg. of the enzyme),
information regarding the chemical properties of deoxyribonuclease I is
much less extensive than the corresponding data available for crystallized
ribonuclease.

The molecular weight of deoxyribonuclease I (based on its diffusion in
the Northrop-Anson cell) was found by Kunitz^'' to be approximately
60,000. It contains 8 % of tyrosine and 2 % of tryptophan. The ultraviolet
absorption of deoxyribonuclease I is a consequence of the presence of these
two amino acids in the molecule.

The pH of minimal electrophoretic mobility is approximately 5.

1" E. J. Cooper, M. L. Trautmann, and M. Laskowski, Proc. Soc. Exptl. Biol. Med.

73, 219 (1950).
"2 H. H. Henstell and R. L. Freedman, Cancer Research 12, 341 (1952).
"3 H. H. Henstell, R. L. Freedman, and B. Ginsburg, Cancer Research 12, 346 (1952).
"4 N. B. Kurnick, L. I. Schwartz, S. Pariser, and S. L. Lee, /. Clin. Invest. 52, 193

(1953).
"6 S. L. Lee, S. R. Michael, and L L. Vural, Aju. J. Med. 10, 446 (1951).
116 M. McCarty, J. Gen. Physiol. 29, 123 (1946).
1" A. W. Bernheimer and N. K. Ruffier, J. Exptl. Med. 93, 399 (1951).

584 G. SCHMIDT

h. Other Deoxyribonucleases

Recent observations from several laboratories demonstrated the exist-
ence of several different enzymes catalyzing the cleavage of the internu-
cleotide bonds of DNA."^

Maver and Greco®^-^® and Brown, Jacobs, and Laskowski"^ found that
the considerable deoxyribonuclease activity of thymus glands is largely
due to the presence of an enzyme which can be differentiated from deoxy-
ribonuclease I. Laskowski and his collaborators^^^-^-^ as well as Webb'-'^^^
succeeded in purifying the extractable fraction of thymus deoxyribonu-
clease several hundred fold. The enzyme was inhibited by 0.025 M mag-
nesium sulfate, but it was not affected by 0.005 M magnesium sulfate. Its
activity was not influenced by the specific protein inhibitor of deoxyribo-
nuclease I. Its pH optimum was at pH 5.

Thymus deoxyribonuclease II hydrolyzes DNA without the formation of
undialyzable cores.

The association of the tissue deoxyribonucleases with particulate frac-
tions^^*'^-^ is discussed in Section X of this Chapter.

Deoxyribonucleases of Plants. The occurrence of deoxyribonucleases in
plants has been reported,^--* '^-^ but very little detailed information concern-
ing the properties of these enzymes is as yet available. The high concen-
trations of deoxyribonuclease activity in various germinating seeds^"^'^'*
are of particular interest.

Deoxyribonucleases of Microorganisms. Various microorganisms^"^ 'i-^-^^"
contain highly active deoxyribonucleases. Only a few of these enzymes
have so far been studied in some detail. Zamenhof and Chargaff^''^ obtained
from extracts of bakers' yeast a complex of a deoxyribonuclease with a
specific inhibitor protein by precipitation in 0.6 saturated ammonium
sulfate solution. On prolonged autolysis at low temperature, the inhibitor
protein is destroyed by proteolytic enzymes, and the deoxyribonuclease

118 G. Herbert, K. Lang, and A. Corbet, Biochem. Z. 320, 418 (1950).

n9 K. D. Brown, G. Jacobs, and M. Laskowski, J. Biol. Chem. 194, 445 (1952).

120 M. Laskowski, E. A. Steberl, R. Akka, and P. Watson, Biochirn. et Biophys. Ada
(1954), in press.

121 M. Privat de Garilhe and M. Laskowski, Biochirn. et Biophys Acta (1954), in press.
1" M. Webb, Exptl. Cell Research 5, 27 (1953).

123 M. Webb, Exptl. Cell Research 5, 16 (1953).

124 J. P. Greenstein, Federation Proc. 1, 113 (1942).

125 G. Brawermann and E. Chargaff, J. Biol. Chein. 210, 445 (1954).

126 H. Plenge, Z. physiol. Chem. 39, 190 (1903).

127 M. McCarty and O. T. Avery, J. Exptl. Med. 83, 97 (1946).

128 S. S. Cohen, J. Biol. Chem. 168, 511 (1947).

129 W. S. T. Tillett, S. Sherry, and L. R. Christensen, Proc. Soc. Exptl. Biol. Med. 68,
184 (1948).

130 W. C. Schneider and C. H. Hogeboom, J. Biol. Chem. 198, 155 (1952).

ENZYMES ATTACKING NUCLEIC ACIDS 585

activity increases correspondingly, e.g., during 90 days' autolysis, the
deoxyribonuclease activity of a yeast extract increased fifty fold according
to the results of a viscosimetric assay procedure. The deoxyribonuclease
differs from deoxyribonuclease I of pancreas by its pH optimum which is
in the region of pH 6. It is insoluble in water but soluble in dilute salt
solution. It requires magnesium ions for its activity. The protein inhibitor
which is destroyed by various crystallized proteolytic enzymes is specific
for the yeast enzyme and has no activity toward deoxyribonuclease I, and
toward deoxyribonucleases of thymus and of neurospora.

Muggleton and Webb^" found in culture filtrates of a soil actinomyces
(strain A) a very heat-labile deoxyribonuclease which hydrolyzed DNA
and the cores obtained from digests of DNA with deoxyribonuclease I.

3. "Unspecific" Phosphodiesterases
a. Snake Venom Phosphodiesterases

The internucleotide bonds of PNA and of all deoxyribo- and ribonucleo-
tides which do not contain terminal secondary phosphoryl groups in the
3 '-position are hydrolyzed by purified extracts of the venoms of a variety
of poisonous snakes, e.g., those of Crotalus, of Bothrops, of the Japanese
snake "Habu." Since none of these enzymes has been obtained in pure,
crystalline form, it is not possible at present to decide as to whether the
phosphodiesterases of various snake venoms are identical or not. Histor-
ically, the snake venom phosphatases, whose action on nucleic acids was
discovered in 1919 by Delezenne and Morel, '^^ are of interest because
Akamatsu's concept'^- (1931) of the existence of specific enzymes for the
cleavage of phosphomonoesters and of phosphodiesters originated from
observations with this group of enzymes. Crude snake venoms with few
exceptions contain mixtures of phosphodiesterases and phosphomono-
esterases. Hurst and Butler^'^^ succeeded, however, in separating both
types of phosphatases chromatographically; Sinsheimcr and Koerner'^^
described a procedure involving alcohol fractionation for the preparation
of phosphodiesterase containing only negligible quantities of phospho-
monoesterases.

Specificity. Snake venom phosphodiesterase hydrolyzes not only oligo-
nucleotides of the structures defined, but also diphenyl phosphate and
6zs-p-nitrophenyl phosphate to the corresponding monoesters and phenols
without the formation of inorganic phosphate. It can be assayed by phenol

131 C. Delezenne and H. Morel, Compt. rend. 168, 241 (1919).

132 T. Uzawa, J. Biochem. (Japan) 15, 19 (1932).

133 R. O. Hurst and G. C. Butler, J. Biol. Chem. 193, 91 (1951).

"" R. L. Sinsheimer and T. J. Koerner, J. Biol. Chem. 198, 293 (1952).

586 G. SCHMIDT

determination or, in the case of the phosphodiesters of nitrophenols which
are practically colorless, by direct photometric determination of the yellow
color of the liberated nitrophenol.

Monophenyl phosphate, p-nitrophenyl phosphate, the glycerophosphates
and all mononucleotides are resistant to the enzyme.

Nature of Degradation Products. Incubation of PNA as well as of deoxy-
ribonuclease I digests of DNA results in the cleavage of practically all
internucleotide bonds. (The appearance of small amounts of inorganic
phosphate in many digests originates most likely from small contamina-
tions of the phosphodiesterase preparations with phosphomonoesterases.)
The action of snake venom phosphodiesterase on PNA was first studied by
GuUand and Jackson,^'* who found that such digests contained phosphoryl
groups which were hydrolyzed by a specific 5 '-nucleotidase (see below)
and concluded that a considerable amount of 5 '-phosphoryl groups must
be present in PNA itself. The action of this enzyme on PNA was con-
vincingly elucidated in the investigations of Cohn and Volkin.^^^ In the
case of deoxyribooligonucleotides, the final hydrolysis products are exclu-
sively the four 5'-mononucleotides. The exhaustive digestion of PNA with
snake venom phosphodiesterase results in a mixture of the four 5 '-nucleo-
tides, of pyrimidine-2',5'- and -3',5'-diphosphomononucleotides, and of
an amount of purine nucleosides approximately equivalent to that of the
diphosphonucleotides.

Thus, in contrast to the action of ribonuclease, that of snake venom
diesterase causes specifically the cleavage of the 3 '-ester linkages between
the phosphoryl groups and the corresponding nucleoside residues. The
structural implications of these observations have been discussed in Chap-
ter 12.

The practically complete cleavage of PNA and of all deoxyribooligo-
nucleotides to mononucleotides or — in the case of PNA — to mixtures of
mononucleotides and mononucleotide phosphate esters leads to the con-
clusion that all internucleotide bonds in these polynucleotides regardless
of their structural differences are cleaved by snake venom phosphodieste-
rase. However, this general statement requires one qualification: Only
those compounds are hydrolyzed by the enzyme which do not have a sec-
ondary phosphoryl group in the 3'-position. Thus, dinucleotides of the type
shown in the lower part of Figure 8 are resistant to the enzyme, but nu-
cleotide-nucleoside esters originating from dinucleotides by the removal
of the terminal phosphoryl group by a phosphomonoesterase are hydrolyzed
by snake venom diesterase.

Consequently, all oligonucleotides of deoxyribonuclease I-digests are

1" J. M. Gulland and E. M. Jackson, Biochem. J. 32, 597 (1938).
i'" W. E. Cohn and E. Volkin, /. Biol. Chem. 203, 319 (1953).

ENZYMES ATTACKING NUCLEIC ACIDS 587

1' 2' 3' 4' 5'

(Diesterase)

/

(Diesterase)

(Bone P-ase) P

Fig. 8. Action of the phosphodiesterases of intestinal mucosa and of snake venom,
and of bone phosphomonoesterase on PNA. [From Volkin and Cohn."]

hydrolyzed by snake venom phosphodiesterase because their terminal phos-
phomonoester groups are in the 5'-position; on the other hand, the ohgo-
nucleotides of digests of PXA with ribonuclease I which contain the termi-
nal secondary phosphoryl groups in the 3'-positions are hydrolyzed by snake
venom phosphodiesterase only after the removal of these groups by a pre-
liminary digestion with a phosphomonoesterase (bone phosphomonoesterase
or prostatic phosphatase).

The question remains why undegraded PNA (prepared by mild pro-
cedures in the laboratory) is hydrolyzed by the enzyme without a preceding
monoesterase treatment^^^ despite its content of terminal 3'-phosphoryls
which amounts to about 6 to 10% of its total phosphorus. Possibly the in-
hibiting influence of these terminal groups decreases with increasing degrees
of polymerization of the substrates.

The resistance of highly polymerized deoxyribonucleic acids toward
snake venom diesterase could be attributed either to their molecular size
or to structural reasons. No decision between these alternatives is possible
at present.

It may well be that the action of the snake venom enzyme is a simple
hydrolysis; at least it is obvious that its action on DNA cannot involve an
intermediary formation of 2,3-cyclic nucleotide groups.

Properties of Snake Venom Phosphodiesterases

Correlation with toxicity. According to Taborda et al.,^^'' the average
phosphodiesterase concentration in venoms of various species shows a

1" A. R. Taborda, L. C. Taborda, J. N. Williams, Jr., and C. A. Elvehjem, J. Biol.
Chem. 195, 207 (1952).

588 G. SCHMIDT

significant parallelism with their toxicity. The activity of these enzymes is
strongly inhibited by some factors known to depress the toxicity, such as
antivenom serum, cysteine, and formaldehyde. The activity determinations
of Taborda's group were carried out on crude extracts; since a manometric
method was used, the values obtained represent figures of the phosphodi-
esterase activities.

-pH Optimum. The optimum for the hydrolysis of PNA and of deoxy-
oligonucleotides is at pH 9.2. ^^^'^^^ GuUand and Jackson reported optimum
hydrolysis of diphenyl phosphate at pH 8.6.^^*

Activators and inhibitors. Snake venom phosphodiesterase is considerably
activated by 0.003 M magnesium ions, and to a lesser extent by manganese
ions. Its activity seems to be inhibited by the characteristic inhibitors of
alkaline phosphomonoesterases such as cyanide and cysteine, but the evalu-
ation of the available data is difficult because of the high concentrations
at which the inhibitors were studied. Adrenocorticotropic hormone and
cortisone, both of which are known to alleviate the clinical toxicity of the
venoms, has no effect on their phosphodiesterase activity,

h. ^^Phosphodiesterase" of Intestinal Mucosa

The studies of the enzymic cleavage of nucleic acids into lower poly-
nucleotides and mononucleotides by duodenal juice or by extracts of in-
testinal mucosa were of great historical importance in the development of
the chemistry of the nucleic acids. The use of this enzyme for the partial
hydrolysis of nucleic acids in the laboratory of P. A. Levene"^ and in that of
S. J. Thannhauser^'^ was essential for the successful isolation of the nu-
cleosides and nucleotides of deoxyribonucleic acids and for the isolation of
D-deoxyribose. After the advent of ion-exchange chromatography, Cohn
and Carter^^" succeeded for the first time in isolating the four 5'-ribonucleo-
tides as products of the enzymic hydrolysis of PNA by intestinal phospho-
diesterase.

Specificity. Unlike snake venom phosphodiesterase, intestinal phospho-
diesterase has not been obtained free from phosphomonoesterase activity
by fractionation methods."' "' The phosphodiesterase activity of highly
purified preparations of intestinal phosphatase is specific toward the bonds
between the nucleotide groups of PNA and those of DNA polynucleotides
of relatively low molecular weights. Cyclic 2',3'-mononucleotides are not
hydrolyzed by intestinal phosphodiesterase.^^ Diphenyl phosphate, glyc-

''* P. A. Levene and L. W. Bass, "Nucleic Acids." Chemical Catalog Co., New York,

1931.
139 W. Klein and S. J. Thannhauser, Z. physiol. Chem. 231, 96 (1935).
1^0 W. E. Cohn and C. E. Carter, Nahire 167, 483 (1951).
1" G. Schmidt and S. J. Thannhauser, J. Biol. Chem. 149, 369 (1943).

ENZYMES ATTACKING NUCLEIC ACIDS 589

erylphosphorylcholine, and glycerylphosphorylethanolamine are hydro-
lyzed by homogenates and crude autolysates of intestinal mucosa of calves,
but not by highly purified intestinal phosphatase.^^

Highly polymerized DNA is likewise resistant toward purified intestinal
phosphatase/^' ^^ but digests of DNA obtained by incubation with deoxy-
ribonuclease I are rapidly and completely hydrolyzed to nucleosides and
inorganic phosphate.

Yeast and pancreas PNA are completely cleaved to nucleosides and ortho-
phosphate by highly purified intestinal phosphodiesterase provided the
incubation is carried out with relatively large amounts of the enzyme. The
rates of hydrolysis of DNA polynucleotides by intestinal phosphatase are
considerably higher than those of PNA.^*-

The hydrolysis of phosphomonoesters by intestinal phosphatase proceeds
much more rapidly than that of polynucleotides; since the latter is thus the
limiting factor, mononucleotides are usually not detectable in the digests.
Klein and Thannhauser^^^ found, however, that the phosphomonoesterase
activity of intestinal mucosa is inhibited by arsenate to a higher degree than
is its phosphodiesterase action. In the presence of arsenate, appreciable
amounts of mononucleotides accumulate in the digests. The amounts of
deoxymononucleotides in arsenate-containing digests of depolymerized
DNA are sufficient to permit their isolation by fractionation procedures.
The presence of mononucleotides in PNA hydrolysates obtained with
intestinal phosphatase in the presence of arsenate was demonstrated by
Cohn and Volkin by means of ion-exchange chromatography. "''■^^^ In con-
trast to pancreas ribonuclease I, which forms only 3'-pyrimidine nucleo-
tides, the nucleotides originating from the action of intestinal phosphatase
are 5'-purine and pyrimidine nucleotides.

The strong nucleophosphodiesterase activity of intestinal phosphatase
raises the ciuestion as to whether alkaline phosphatases of other tissues
are likewise capable of hydrolyzing certain phosphodiester linkages. So far,
the alkaline phosphatases in various tissues have been considered as identi-
cal although GuUand"^ considered alkaline bone phosphatase as a strict
phosphomonoesterase. On the strength of Gulland's interpretation, this
enzyme has recently been used as a reagent for terminal secondary phos-
phoryl groups of nucleic acids. '^^ The author of this review, however, is not
convinced of the evidence in favor of the exclusive monophosphoesterase
nature of bone phosphatase. Gulland's view was based on experiments with

1" W. Klein, Z. physwl. Chem. 207, 164 (1933).

>" W. Klein and S. J. Thannhauser, Z. physiol. Chem. 218, 164 (1933).

1^^ W. E. Cohn, D. G. Doherty, and E. Volkin in "Phosphorus Metabolism" (McElroy

and Glass, eds.), Vol. 2, p. 339. Johns Hopkins Press, Baltimore, 1952.
'« J. M. Gulland and E. M. Jackson, /. Chem. Soc. 1938, 1492.

590 G. SCHMIDT

relatively weakly active phosphatase preparations, whereas the action of
intestinal phosphatase was studied on highly active phosphatase samples.
Owing to the relatively slow rate of the enzymic depolymerization of PNA
and of deoxypolyniicleotides, the diesterase action of weakly active phos-
phatase preparation might have been overlooked. It seems that the prob-
lem of nucleophosphodiesterase action of alkaline phosphatases requires
clarification by further investigations.

yH Optimum. The pH optimum of intestinal nucleophosphodiesterase
has been reported to be in the alkaline range ; since all available observations
are based on the liberation of inorganic phosphate from the substrate, it is
possible that these data describe the behavior of the phosphomono- rather
than that of the phosphodiesterase activity.

Assay. Approximate figures for the activities of intestinal phosphodi-
esterases are usually obtained by the determination of inorganic phosphate
or of phenols which originate from the intermediary phosphomonoesters by
the action of the phosphomonoesterase which is always present. In the case
of other nucleophosphodiesterases, it might be preferable to determine the
organic acid-soluble phosphorus compounds formed, particularly for studies
of the effects of enzyme inhibitors.

III. Enzymes Catalyzing the Hydrolytic Cleavage of the Phosphoryl
Groups of Mononucleotides

1, Phosphatases of Low Degrees of Specificity

Mononucleotides are rapidly dephosphorylated by many phosphatases
such as intestinal phosphatase, bone phosphomonoesterase, acid prostatic
phosphatase, almond phosphatase. Since these phosphatases are active
toward many other phosphoric acid monoesters, their detailed description
is beyond the scope of this book. It should be mentioned, however, that in
some cases the hydrolysis of nucleotides shows "kinetic differences in com-
parison with that of other substrates. The rates of hydrolysis of 3 '-nucleo-
tides by prostate phosphatase follow the course of a unimolecular reaction,
at least up to degrees of 70 % hydrolysis, whereas those of the hydrolysis of
many other substrates (in particular, of the glycerophosphates) fall much
more rapidly even in the initial stages."

The phosphatases just mentioned hydrolyze all mononucleotides regard-
less of the position of the phosphoryl groups. In some cases, however, the
rates of hydrolysis catalyzed by one enzyme are very markedly influenced
by the position of the phosphoryl groups in the nucleoside moiety. For
example, the hydrolysis of 5'-adenylic acid by prostatic phosphatase is
approximately one-third of that of 2'- or of 3'-adenylic acid."

enzymes attacking nucleic acids 591

2. Specific Nucleotide Phosphatases (Nucleotidases)
a. 5' -N iicleotidases

The existence of specific phosphatases for the hydrolysis of 5'-nucleotides
was implicitly suggested by the old observation that mammalian skeletal
muscle and heart tissue which do not hydrolyze glycerophosphate, 2'- and
3'-nucleotides, and many other monoesters of phosphoric ac^d, are capable
of hydrolyzing 5'-adenylic and 5'-inosinic acids. Reis^'*^"^''* established in
systematic studies the widespread occurrence in mammalian tissues of a
specific 5'-nucleotidase which is particularly abundant in muscle and nerve
tissue and in bull testicle. ^^^ In these tissues, the largest part of the enzyme
is closely associated with the insoluble particles.

In the cytoplasm and the nuclei of the cells of smooth muscle (such as
uterus and aorta) and in fibroblasts, the presence of a specific 5'-nucleo-
tidase was demonstrated histochemically by Wachstein and Meisel.^^^ Par-
ticularly high concentrations of o'-nucleotidase were found histochemically
in the posterior lobe of the human pituitary gland. ^^^

Soluble 5'-nucleotidases of strong activity were found in snake venoms.
In the plant kingdom, potatoes are a convenient source for the preparation
of purified extracts with highl}^ specific 5'-nucleotidase activity. ^^- The
(juestion as to whether the 5'-imcleotidases obtained from the various
sources are closely related in regard to their chemical properties or not, can-
not be decided at present because the purification procedures of these en-
zymes were designed for the purpose of obtaining preparations with high
enzymic specificity rather than homogeneous enzyme proteins.

Properties of 5' -Nucleotidase of Bull Semen. Some of the properties of
5'-nucleotidase from bull semen, which was studied by Heppel and Hil-
moe,^^^ are as follows.

pH Optimum. Seminal 5'-nucleotidase has its optimal activity at pH
8.5; 5'-nucleotidases from other sources have likewise a pH optimum in this
region.

Potency. The activity per milligram of the most highly purified prepara-
tions of seminal 5'-nucleotidase is about twice that of purified intestinal
phosphatase. The concentration of the enzymic activity in seminal plasma
is much higher than that of intestinal phosphatase in the intestinal mucosa.

1" J. Reis, Bull. sue. chim. biol. 22, 36 (1934).

1" J. Reis, Enzymologia 2, 183 (1937).

"8 J. Reis, Enzymologia 5, 251 (1938).

1" T. Mann, Biochem. J. 39, 451 (1945).

150 M Wachstein and E. Meisel, Science 115, 652 (1952).

^" A. G. E. Pearse and J. L. Reis, Biochem. ./. 50, 534 (1952).

162 A. Kornberg and W. E. Pricer, Jr., J. Biol. Chem. 186, 557 (1950).

>" L. A. Heppel and R. J. Hilmoe, J. Biol. Chem. 188, 665 (1951).

592 G. SCHMIDT

Speci^city. The specificity of bull semen phosphatase was thoroughly-
investigated by Heppel and Hilmoe, who found that 5'-adenylic acid, 5'-
uridylic acid, and nicotinamide-5-nucleotide, as well as ribose-5-phosphate,
were rapidly hydrolyzed by the enzyme, whereas all other phosphoric acid
esters tested, in particular, the 2'- and 3'-ribonucleotides and the adenyl
pyrophosphates, were resistant toward the enzyme. Of particular interest
is the fact that ribose-5-phosphate is hydrolyzed at a very considerable rate
whereas glucose-6-phosphate, phosphogluconate, and 1,6-fructose diphos-
phate are practically resistant against the enzyme.

Michaelis-Menten constants. The affinity of seminal 5'-nucleotidase for
the purine and pyrimidine-5'-nucleotides (iv^^deoyUc <l^~\ A'jfuridyiic
<10~'*, i^Afcytidyiic = 2 X 10~*) Is much higher than those for ribose-5'-

phOSphate (/Vjifrlbose-S-phosphate — 2.6 X 10~ ; ivAfnicotinamide-S-nucIeotide ~ 1-6

X 10-3).

Activators and inhibitors. Magnesium ions in approximately 0.01 M con-
centration have considerable activating effect; they cannot be replaced by
manganese ions. The extent of the magnesium activation depends on the
nature of the buffer : Heppel and Hilmoe found that in tris(hydroxymethyl)
aminomethane buffer the enzyme has 75 % of its full activity without the
addition of magnesium ions, whereas, in glycine buffer the activity without
magnesium ions amounts only to 30% of the optimal activity. ^^

Fluorides in 0.01 M concentration inhibit the enzyme activity by 70%,
in 0.1 M concentration completely. Borate in 0.08 M concentration inhibi-
ted the activity 85 % in comparison with glycine buffer at the same pH.

Stability. Crude as well as purified preparations of seminal 5'-nucleotidase
can be preserved in active form for a considerable time in the frozen state,
before or after lyophilization.

b. 3' -Nucleotidase (formerly b -nucleotidase)

Shuster and Kaplan^ ^^ discovered the presence in germinating rye grass
and germinating barley of a phosphatase which specifically catalyzes the
hydrolysis of 3 '-nucleotides to nucleosides and inorganic phosphate. En-
zymes of similar specificity occur in other vegetable materials, such as wheat
and corn leaves, soy bean leaves, and lawn grass leaves. Takadiastase like-
wise dephosphorylates 3'-nucleotides much faster than 2'- or 5'-nucleotides.
So far, the most powerful 3 '-nucleotidase activity was found in germinating
rye grass. 2'-Nucleotides, 5'-nucleotides, and pyrophosphates were not
hydrolyzed by purified preparations of the enzyme. The fact that one of
the phosphoryl groups of coenzyme A is hydrolyzed by 3'-nucleotidase is
interpreted by assigning to this phosphoryl group the 3'-position of the

1" L. Shuster and N. O. Kaplan, J. Biol. Chem. 201, 535 (1953).

ENZYMES ATTACKING NUCLEIC ACIDS 593

adenylic acid group. The phosphomonoester group of TPN, on the other
hand, is resistant toward 3'-nucleotidase, and it is assumed that this group
is esterified at the 2'-position of the adenyhc acid moiety. The glycerophos-
phates and the hexose and pentose phosphates are not hydrolyzed by the
enzyme.

The phosphoric acid diester, diphenyl phosphate, is resistant to 3'-nucleo-
tidase, but the phosphodiester linkages of ribopolynucleotides are hydro-
lyzed by the enzyme. During this hydrolysis, only up to G% of the total
phosphorus is liberated in the form of inorganic phosphate. This would
suggest that the interlinkages of the ribopolynucleotides are cleaved at the
3'-phosphoric acid ester bonds. It cannot yet be decided whether the ribo-
nuclease activity of 3 '-nucleotidase preparations is a property of the 3'-nu-
cleotidase itself, or whether it must be attributed to the presence of a
second enzyme in the preparations.

pH Optimum. 3'-Nucleotidase has a rather narrow range of optimal ac-
tivity in the neighborhood of p?! 7.5.

Kinetics. Different 3'-nucleotides are hydrolyzed by 3'-nucleotidase at
very different rates. Shuster and Kaplan found the following velocities:
3'-adenylic acid, 1; 3'-inosinic acid, 0.47; 3'-guanyHc acid, 0.27; 3'-uridylic
acid, 0.19; 3'-cytidylic acid, 0.11. The Km values for some nucleotides are
as follows: 3'-adenylic acid, 0.3 X 10"^ M; 3'-cytidylic acid, 2.5 X 10"' M\
3'-uridylic acid, 2.7 X 10-^ M.

No additive effect on the hydrolysis rates were observed in mixtures of
various 3'-nucleotides; this observation suggests the assumption that the
hydrolysis of the various 3'-nucleotides is catalyzed by one enzyme.

The action of 3'-nucleotidase on 3'-nucleotides is competitively inhibited
by 2'- and by 5'-nucleotides.

Inhibition of 50% of the optimal activity is effected by the following
compounds: 5.9 X 10-^ M phosphate; 17 X lO"' M arsenate; 0.9 X 10-^ M
cysteine; 5.2 X 10~^ M potassium cyanide; 5 X 10~^ M glutathione. The
effect of cysteine can be reversed by dialysis.

Heat Lahility. Most of the enzyme activity is destroyed within 10 min-
utes at 80°, and within one or 2 minutes at 100°.

3. Phosphatases as Phosphotransferases between Phosphoric Acid
Esters and Nucleosides

Brawerman and Chargaff^^'''^-^ obtained evidence for the power of various
phosphatases to transfer phosphate groups from "low energy" donors, such
as 5'-nucleotides or phenyl phosphates, to nucleosides. This transferase

i"aG. Brawerman and E. Chargaff, /. Am. Chem. Soc. 75, 2020 (1953).
i"b G. Brawerman and E. Chargaff, J. Am. Chem. Soc. 75, 4113 (1953).
'^^^ G. Brawerman and E. Chargaff, Federation Proc. 13, 186 (1954); Biochim. el
Biophys. Acta (in press).

594 G. SCHMIDT

activity is of interest as a mechanism for the biosynthesis of nucleotides.
The donor requirements are different for different phosphatases. 5'-Nucleo-
tides are particularly efficient donors with malt phosphatase as the transfer
enzyme, but ineffective with prostate phosphatase. ^^''^ On the other hand,
phenyl phosphate is a good phosphoryl donor with prostate phosphatase.
No essential differences between the efficiency of various donors were found
for liver phosphatase. ^^^=

The structure of the nucleotides formed was different according to the
nature of the transferring phosphatases. With the barley phosphatase, only
5'-nucleotides were obtained (e.g., 5'-cytidylic, 5'-thymidylic acids),
whereas with prostate phosphatase, all three isomers of each nucleotide
were obtained. Between 5 and 10% of the added nucleosides (in 40 milf
concentrations) were phosphorylated at donor concentrations of approxi-
mately 200 mM.

IV. Nucleoside Kinases

Adenosine kinase was discovered in yeast by Ostern, Baranowski, and
Terszakowec,^^^*^ in 1938 and partially purified by Caputto^^'** and by Rom-
berg and Pricer^^^f in 1951, who demonstrated also the presence of a different
kinase in some animal tissues. It catalyzes the reaction:

adenosine -|- ATP = AMP + ADP

pH Optima. The pH optimum of the yeast enzyme is between 6.9 and 7 ;
that of the liver enzyme lies between 4.9 and 5.0.

Specificity. The yeast enzyme, as well as the liver enzyme, is specific for
adenosine and for 2 , 6-diaminopurine riboside. ^^*^ No phosphorylation of
guanosine, inosine, uridine, or cytidine could be detected.

Activators. The yeast enzyme is optimally activated by 0.02 M magne-
sium ions or by 0.002 M manganese ions. The enzyme can be stabilized by
reduced glutathione or by an unknown stabilizing factor present in boiled
yeast extracts.

Assay. According to Ostern et al.^^*'^ acetone-dried yeast is sufficiently
active to permit the isolation of 5'-adenylic acid in good yield in large-scale
experiments with adenosine as substrate.

In maceration juice or extracts of animal tissues, the enzyme can only be
detected after a preceding purification designed to remove contaminating
enzymes. Caputto used in his assay procedures manometric determinations
of 5'-adenylic deaminase, phosphate balances, or chromatography, and he
succeeded in the quantitative analytical separation of adenosine and the

'"d P. Ostern, T. Baranowski, and J. Terszakowec, Z. physiol. Chem. 251, 258 (1938).

i"«R. Caputto, J. Biol. Chem. 189, 801 (1951).

'"f A. Kornberg and W. E. Pricer, /. Biol. Chem. 193, 481 (1951).

ENZYMES ATTACKING NUCLEIC ACIDS 595

nucleotides by precipitation of the latter with zinc suKate and barium
hydroxide.

The presence in yeast autolysates and pigeon liver extracts of a phospho-
kinase converting the 4-amino-5-imidazolecarboxamide riboside to its ribo-
tide was reported by Greenberg.'''^'^

V. Enzymes Acting on the Amino Groups of Purine and Pyrimidine

Compounds

1, Enzymic Deamination of the Adenine Group

a. Behavior of Free Adenine

The tissues of many higher animals do not seem to contain enzymes ca-
pable of deaminating free adenine. Some statements to the contrary in the
literature are based on indirect evidence such as the decoloration of meth-
ylene blue by adenine in the presence of xanthine dehydrogenase. It is now
known that the enzymic dehydrogenation of adenine results in the forma-
tion of 2 ,8-dihydroxyadenine, and that it is not preceded by its deamination
to hypoxanthine.^^"'^^^ The metabolic resistance of the amino group
of adenine in rats and dogs is demonstrated by Nicolaier's'^^ observation
that ingested adenine is transformed into 2,8-dihydroxyadenine, which
forms crystalline deposits in the kidney. The mechanism of this transforma-
tion has recently been studied b.y Bendich, Brown, Philips, and Thiersch'^*
who found that ingested 2-hydroxyadenine (isoguanine) and 8-hydroxy-
adenine can be converted to 2,8-dihydroxyadenine (see also Chapter 25).
It should be pointed out, however, that the basis of comparative biochemi-
cal studies of this question is not broad enough to permit the conclusion that
the resistance of the amino group of free adenine is representative of the
behavior of this base in higher animals.

The desirability of extensive comparatiA-e studies is emphasized by the
observations of Duchateau-Bosson and his collaborators, ^"'^^^ who found
that extracts of certain lower animals such as anodonta, crustaceans, and
insects contain adenine deaminases, but lack enzymes capable of deaminat-
ing adenosine or adenylic acid. The authors consider the different distribu-
tion of the deaminases of the adenine group in lower and higher animals as

>"8G. R. Greenberg, Federation Proc. 12, 211 (1953).

1" A. Nicolaier, Z. klin. Med. 45, 359 (1902).

i5« A. Bendich, G. B. Brown, F. S. Philips, and J. B. Thiersch, J. Biol. Chew. 183,

267 (1950).
1*' G. Duchateau-Bosson, M. Florkin, and G. Frappez, Compt. rend. soc. biol. 133,

433 (1940).
'" G. Duchateau-Bosson, M. Florkin, and G. Frappez, Compt. rend. soc. biol. 133,

274 (1940).
'6' G. Duchateau-Bosson, M. Florkin, andG. Frappez, Acad. roy. Belg. 27, 169 (1941).

596 G. SCHMIDT

an example of phylogenetic changes in the pattern of certain metaboHc
pathways.

b. Adenyl Deaminase of Aspergillus Oryzae

Mitchell and McElroy found in 1946 that takadiastase contains an en-
zyme capable of deaminating adenosine.^®" They succeeded in separating
taka-adenosine deaminase from some contaminating enzymes on a chro-
matopile in quantities suitable for isolation experiments.^®^ Kaplan, Colo-
wick, and Ciotti^®' achieved a considerable further purification of the en-
zyme by acetone-alcohol and ammonium sulfate fractionations with the
practically complete removal of phosphatases. According to Kaplan et al}^-
adenosine, adenosine-5'-phosphate, adenosine-3'-phosphate, ATP, ADP,
oxidized and reduced DPN, and adenosine-diphosphate-ribose are deami-
nated by the purified enzyme at rates decreasing in this order. Adenine,
adenosine-2'-phosphate and TPN are not deaminated.

The lack of specificity of the Aspergillus oryzae adenyl deaminase is in
remarkable contrast to the highly selective action of animal adenyl de-
aminases. The deamination of DPN with preservation of the pyrophosphate
bond represents the first example of the enzymic deamination of an adenyl
compound of a pyrophosphoryl dinucleotide.

pH Optimum. Aspergillus oryzae deaminase has a broad zone of optimal
activity between pH 5 and pH 8.

Michaelis-Menten Constants. The following K values were calculated for
the various substrates by Kaplan et al.:^^'- adenosine, 0.6 X 10"^ M; ADP,
0.7 X 10-3 M; 5'-AMP 0.8 X IQ-^ M; ATP, 1.2 X 10-^ M; adenosine-
diphosphate-ribose 1.5 X IQr' M; 3'-AMP, 1.7 X 10"^ M.

c. Adenosine Deaminase

Most tissues (intestines, liver, kidney, spleen, brain, striated muscle,
heart) of higher animals contain adenosine deaminase in very active con-
centrations. The enzyme was discovered by Gyorgy and Rothler;^®^ the
high degree of its specificity was recognized by Schmidt^®^ in an investiga-
tion on muscle deaminases, and its distribution in animal tissues was de-
scribed by Conway and Cooke. ^^^ Procedures for its partial purification

160 H. K. Mitchell and W. D. McElroy, Arch. Biochem. 10, 351 (1946).

i«i H. K. Mitchell, M. Gordon, and F. A. Haskins, J. Biol. Chem. 180, 1071 (1949).

'62 N. O. Kaplan, S. P. Colowick, and M. M. Ciotti, J. Biol. Chem. 194, 579 (1952).

1" P. Gyorgy and H. Rothler, Biochem. Z. 187, 194 (1927).

i«^ G. Schmidt, Z. physiol. Chem. 179, 243 (1928).

1" E. J. Conway and R. Cooke, Biochem. J. 33, 479 (1939).

ENZYMES ATTACKING NUCLEIC ACIDS 597

from aqueous extracts of intestinal mucosa have been described by Brady^^®
and by Kalckar.^"

Specificity. Adenosine deaminase is strictly specific toward adenine ribo-
and deoxyribonucleosides/^^ and it is probable that both nucleosides are
deaminated by the same enzyme. The amino groups of free adenine and of
other purines, pyrimidines, and their derivatives, as well as those of the
adenine nucleotides, are resistant to the enzyme. Byrne'^^-' reported, how-
ever, that highly purified samples of adenosine deaminase from beef spleen
were capable of hydrolysing the amino groups of 2'-aden3dic acid and 2',
3',-cyclic adenylic acid as rates corresponding approximately to }^q of the
rate of the deamination of adenosine. The pH-optimum for the nucleotides
was at pH 5.2, that for the nucleoside between pH 7.5 and 9.3. 3'- and
5'-Adenylic acid, DPN, and TPX were not appreciably deaminated.

pH Optimum. According to Kalckar, the pH optimum of adenosine de-
aminase is near the neutral point, but its activity is still considerable at
pH 9 and pH 6.

Stability. Adenosine deaminase of intestinal mucosa is rapidly inactivated
by standing at pH 3. On dialysis against water, the activity disappears
rapidly and is not restored by combination with the dialysate.

Inhibitors. According to Brady ,'®^ the activity of intestinal adenosine
deaminase is not influenced by low concentrations of fluoride, phosphate,
and cyanide ions.

d. 5'-Adenylic Acid Deaminase

5'-Adenylic acid deaminase occurs in relatively high concentration in
striated muscle^^^ but is practically absent in heart. ^^^'^^

5'-Adenylic acid deaminase is a highly specific enzyme which converts
5'-riboadenylic acid to inosinic acid. According to Carter,^^^ 5'-deoxyadenylic
acid is slowly deaminated by the enzyme. All other amino compounds are
resistant, in particular, the o'-adenosine pyrophosphates and 2'- and 3'-
adenosine phosphates. Owing to its high specificity, the enzyme is used as a
convenient analytical tool for the quantitativ^e determination of 5'-adenylic
acid.

The enzyme can be easily prepared from muscle in enzymically homo-
geneous form. Its action can be followed either by ammonia determinations
or more conveniently by measuring the changes of the absorption at 265
mM caused by the conversion of the adenine to the hypoxanthine group.^"-""

'««T. Brady, Biochem. J. 36, 478 (1942).

'" H. M. Kalckar, /. Biol. Chem. 167, 461 (1947).

'«^* W. L. Byrne, Abstr. 126th Meeting Am. Chem. Soc, New York p. 73C (1954).

'88 W. Kutscher and W. Sarreither, Klin. Wcchschr. 26, e05 (1948).

'«» C. E. Carter, J. Am. Chem. Soc. 73, 1537 (1951).

"» H. M. Kalckar, J Biol. Chem. 167, 429 (1947).

598 G. SCHMIDT

yH Optimum. 5'-Adenylic acid deaminase has a sharp optimum at pH
5.9.

Extractability. 5'-Adenyhc deaminase is easily extracted from minced
muscle by 2 % solutions of sodium bicarbonate or, to a lesser extent, by
water. Very little of the enzyme is extracted by 0.85% sodium chloride
solution. Removal of soluble proteins by exhaustive extraction of minced
rabbit muscle with the latter solution, and subsequent extraction of the
residue with 2% sodium bicarbonate solution, yields a highly active de-
aminase extract, free of myoglobin, which represents a suitable starting
material for further purification.

These solubility properties are probably responsible for the fact that
purified myosin preparations frequently exhibit deaminating activity
toward 5'-adenylic acid.^^ The association of this deaminase with myosin,
however, is much looser than that of the Ca-activated adenosinetriphos-
phatase. At present, it is hardly justifiable to consider adenylic deaminase
activity as an inherent property of myosin,!^^ particularly in view of the
absence of adenylic acid deaminase activity in heart muscle.

2. Enzymic Deamination of the Guanine Group: Guanase, Guanosine
Deaminase, Guanylic Acid Deaminase

Unlike the adenine group, the guanine group can be deaminated enzymi-
cally by many tissues of higher animals in the form of the free base as well
as in that of its riboside.i^^a Aqueous extracts of fresh rabbit liver or of acetone
powders prepared from this tissue are capable of deaminating guanylic acid,
guanosine, and guanine. The crude, as well as the purified, enzyme prep-
arations used in these investigations contained, in addition, phosphatase,
nucleoside phosphorylase, and possibly enzymes cleaving the purine-ribose
linkage of nucleotides. The available evidence^^^ "^ does not permit a deci-
sion of the question as to whether or not guanosine and guanylic acid are
deaminated before the cleavage of the guanine-ribose linkage.

Specificity of Guanase. Guanase is highly, but not absolutely, specific for
guanine. According to Hitchings and Falco,"^ the enzyme also deaminates
1-methylguanine. Roush and Norris"^ found that highly purified guanase
preparations prepared from rat liver according to Kalckar^^^^ deaminate
azaguanine. The deamination of azaguanine is competitively inhibited by
guanine. Measurable inhibitory effects of some pteridine derivatives (xan-

I'l V. Sz. Hermann and G. Josepovits, Nature 164, 865 (1949).

"2 B. A. Askonas, Biocheyn. J. 48, 42 (1951).

1"* For older references, see W. Jones, "Nucleic Acids," p. 70. Longmans Green &

Co., London, 1920.
1" G. Schmidt, Z. physiol. Chem. 208, 185 (1932).
"4 Y. Wakabayashi, J. Biocheyn. (Japan) 28, 185 (1938).

"6 G. H. Hitchings and E. A. Falco, Proc. Natl. Acad. Sci. U. S. 30, 294 (1944).
"6 A. Roush and E. R. Norris, Arch. Biochem. 29, 124 (1950).

ENZYMES ATTACKING NUCLEIC ACIDS 599

thopterine, 6-hydroxymethyl pteridine, and 6-formyl pteridine) on the
enzjnnatic deamination of azaguanine were reported by Dietrich and
Shapiro. 1^^"

pH Optima of Guanase. According to Roush and Norris, guanase de-
aminates guanine over a pH range between 5 and 9 with a rather flat opti-
mum at pH 8 ; the deamination of azaguanine has a much sharper optimum
at pH 6.5, and the activity decreases steeply at both sides of the optimum.
At pH 8 the activity toward azaguanine is only slight.

Michaelis Constants. For Michaelis constants, Roush and Norris calcu-
lated for guanine as substrate Kjugu^nine = 5 X 10~^ M; for azaguanine,
^i»/a^aguanine = 7 X lO"* M iu phosphate buffers of pH 6.5. The high dilution
of the substrates necessitated by their small solubilities, must be considered
in the evaluation of these constants.

Assay Methods. Guanase may be assayed spectrophotometrically by de-
termining the changes of the extinction at 245 mju, or by determination of
the liberated ammonia. The latter procedure requires substrate solutions of
concentrations exceeding the solubility of guanine at the pH range of op-
timal activity. Sufficiently stable colloidal guanine suspensions can be
prepared by neutralizing guanine solutions in dilute sodium hydroxide in
the presence of 0.5 % gelatin. ^^^

3. Enzymes Deaminating the Cytosine Group

Very little information is available regarding deaminases of the cytosine
group of nucleic acid derivatives. Highly active extracts of cytosine de-
aminase were obtained by Kream and Chargaff"^ from ground cells of yeast
and E. coli. The extracts had no activity toward adenine and guanine.
Cytidine and cytidylic acid (possibly after enzymic dephosphorylation) are
deaminated by extracts of mouse kidney."*

4. Evidence for Transamination in the Purine and Pyrimidine Field

One of the least explored questions in nucleic acid metabolism is that of
the formation of the amino groups of the nucleic acid purines and pyrimi-
dines. Rapid turnover rates for the amino group of the adenine compounds
of muscle were demonstrated by Kalckar and Rittenberg."^ Only a few
pertinent enzymic observations are so far available.

Weil-Malherbe'*" reported that the formation of ammonia during the

»^6» L. S. Dietrich and D. M. Shapiro, J. Biol. Chem. 203, 89 (1953).

"7 E. Chargaff and J. Kream, /. Biol. Chem. 175, 993 (1948) ; J. Kream and E. Chargaff,

J. Am. Chem. Soc. 74, 4274, 5157 (1952).
'^8 J. P. Greenstein, C. E. Carter, H. W. Chalkley, and F. M. Leuthardt, J. Nail.

Cancer Inst. 7, 9 (1946).
'" H. M. Kalckar and D. Rittenberg, /. Biol. Chem. 170, 455 (1947).
'80 H. Weil-Malherbe, Biochem. J. 54, vi (1953).

600 G. SCHMIDT

incubation of dialyzed brain homogenates increased considerably over the
sum of the amounts of ammonia in all controls when glutamine and inosine
triphosphate were added to the system. The most plausible explanation of
this observation is the assumption that the amide group of glutamine was
enzymically transferred to the hypoxanthine group of inosine triphosphate,
which was in turn deaminated by the enzymes of the homogenate.

Stephenson and Trim^^"^ found that the rate of deamination of adenine
by E. coli suspensions was strongly accelerated by catalytic amounts of
adenosine. The authors discussed— very cautiously— the possibility of the
deamination of the added adenosine to inosine and of the subsequent trans-
amination of the amino group of adenine to inosine. It is just as reasonable,
however, to assume a phosphorolytic or nonphosphorolytic exchange of the
whole adenine molecule with the hypoxanthine group of inosine.

According to Gunsalus and Tonzetich,^^! adenine, guanine, or cytosine
are amino donors for glutamate formation from ketoglutaric acid by cell-
free extracts of E. coli. Pyridoxal is required for the reaction; ammonium
ions are without effect.

VI. Enzymes Acting on the Linkages Between the Basic and the
Carbohydrate Groups of Nucleic Acid Derivatives

The enzymic cleavage of the linkages between the basic and the carbo-
hydrate groups of polynucleotide derivatives can take place either in the
free nucleosides or in pyrophosphoryl dinucleotides (e.g., DPN). So far, no
such cleavage has been observed on nucleic acids or polynucleotides.

The enzymic cleavage of iV-glycoside bonds of nucleic acid derivatives
was first observed in nucleosides, and the term "nucleosidases" introduced
by Levene and Medigreceanu^ is still used.^*^ i^ the current literature this
name tends to be replaced by the designations "nucleoside phosphorylases"
and "nucleoside hydrolases." Those terms are more satisfactory because of
their adaptability to similar enzymes acting on nucleotides as substrates.
In the case of the latter enzymes, it might be advisable to change the term
"nucleotide phosphorylases" (proposed by Saffran and Scarano^^^) to "nu-
cleotide-l'-phosphorylases." Although the term "phosphorylases" still im-
plied (in analogy to Cori's original designation) enzyme reaction of glyco-
sidic groups with phosphate, this cannot be said for pyrophosphorylases.
In fact, most of the known nucleotide pyrophosphorylases act on groups
other than glycosidic groups.

The reversible phosphorolysis of nucleosides offers a plausible pathway

180a M. Stephenson and A. R. Trim, Biochem. J. 32, 1740 (1938).
1" C. F. Gunsalus and T. Tonzetich, Nature 170, 162 (1952).

182 M. Dixon and R. Lemberg, Biochem. J. 28, 2065 (1934).

183 M. Saffran and E. Scarano, Nature 152, 949 (1953).

ENZYMES ATTACKING NUCLEIC ACIDS 601

for the biosynthesis of nucleotides involving as its first step the biosynthesis
of a nucleoside and as the second step its esterification with phosphate. The
first experimental evidence for the occurrence of this second step in cells
was reported by Ostern and Terszacowec,'*^ who found that acetone-dried
yeast or toluene-poisoned yeast was capable of converting added adenosine
to AMP and ATP in the presence of fructose diphosphate or phosphogly-
ceric acid.

Esterification of nucleosides with phosphate, however, is not the only,
and perhaps not the most important, mechanism for the biosynthesis of
nucleotides. Observations suggesting the cleavage of the bond between
purine bases and ribose at the nucleotide stage had been made by Schmidt^"
in 1932 and by Wajzer and Baron'*^ in 1949. Since 1952, several enzyme
reactions have been found in which certain purines, pyrimidines, or some of
their nitrogenous precursors are condensed with the 1 '-carbon atoms of
5'-phosphorylated ribose derivatives to nucleotides. Hydrolases acting on
the A^-glucoside linkage of certain nucleotides have also been described. The
study of these enzyme reactions which was initiated mainly in the labora-
tories of G. R. Greenberg,^«« of J. M. Buchanan,^" of H. M. Kalckar,i^« and
of A. Kornberg^^^ is rapidly becoming one of the central problems of cur-
rent research on nucleotide metabolism.

History. Purine nucleosidases were discovered by Levene and Medi-
greceanu^ soon after the elucidation of the general structure of nucleotides.
The specificity of these enzymes toward purines and their inactivity toward
nucleosides of pyrimidines and dihydropyrimidines was recognized in early
investigationsby Levene. '^''•^^^ Much later, Deutsch and Laser'^'- discovered
in Thannhauser's laboratory the presence of specific pyrimidine nucleo-
sidases in bone marrow. An important advance was made when Klein' ^^
found in 1935 that purine nucleosidase prepared from beef spleen lost its
acti\aty by dialysis but was reactivated by phosphate or arsenate. The
explanation for this behavior was given in 19-47 by Kalckar,'^^ who estab-
lished the phosphorolytic nature of phosphate-dependent nucleosidases.
Carter, '^^ however, furnished the first evidence for the conclusion that not

>8^ P. Ostern and J. Terszacowec, Z. physiol. Chem. 250, 155 (1937).
'85 J. Wajzer and F. Baron, Bull. soc. chim. hiol. 31, 750 (1949).

186 G. R. Greenberg, Federation Proc. 13, 745 (1954).

187 W. J. Williams and J. M. Buchanan, J. Biol. Chem. 203, 5S3 (1953).

188 H. M. Kalckar, Biochim. et Biophys. Acta 12, 250 (1953).

189 A. Romberg, I. Lieberman, and E. S. Simms, J. Am. Chem. Soc. 76, 2027 (1954).

190 P. A. Levene and I. Weber, /. Biol. Chem. 60, 707 (1924).

191 P. A. Levene, *Yamagawa, and I. Weber, J. Biol. Chem. 60, 693 (1924).

192 W. Deutsch and R. Laser, Z. physiol. Chem. 186, 1 (1927).

193 W. Klein, Z. physiol. Chem. 231, 125 (1935).

194 H. M. Kalckar, J. Biol. Chem.. 167, 461 (1947).
1" C. E. Carter, J. Am. Chem. Soc. 73, 1508 (1951).

602 G. SCHMIDT

all nucleosidases are phosphorylases, but that hydrolytic enzymes for the
cleavage of the bonds between purine or pyrimidine bases and ribose do
exist. The cleavage of nucleosides is, therefore, another example of the ex-
istence of dual mechanisms for the enzymic cleavage of glycosidic bonds.

1. Nucleoside Phosphorylases
a. Purine Nucleoside Phosphorylases

(1) Animal Enzymes. The first thorough studies of these enzymes were
carried out on purified rat and calf liver extracts by Kalckar,!^* who ob-
tained active preparations from the supernatant solutions obtained by
high-speed centrifugation (16,000 r.p.m.). The extracts were purified by
ammonium sulfate fractionation. The fraction which precipitated between
0.4 and 0.6 saturation was further purified by isoelectric precipitation at
pH 6. The enzyme activity was recovered in the supernatant.

An enzyme of very similar specificity was purified by Heppel and Hil-
moe^^^ from yeast autolysates by ammonium sulfate fractionation and sub-
sequent adsorption on aged calcium phosphate gel. A 19-fold purification
was achieved but considerable losses occurred during the procedure.

Specificity: liver enzyme. Originally Kalckar found that only ribo- and
deoxyriboguanosine and -hypoxanthine were split by purine nucleoside
phosphorylase of rat liver, and that adenosine, xanthosine, and the pyrimi-
dine nucleosides were resistant. Cardini and his associates^^^ as well as
Friedkin^^^'i^^ found later, however, that uridine and xanthosine were also
split by highly concentrated liver enzyme. The phosphorolysis of uridine
by highly concentrated solutions of purine nucleoside phosphorylase is
most likely caused by contamination with thymidine phosphorylase (see
below). Isoguanosine and adenine thiomethylriboside are not split by the
enzyme.^""

Highly purified purine nucleoside phosphorylase was recently obtained
by Korn and Buchanan^oi .202 from aqueous extracts of acetone-dried beef
liver, by alcohol and ammonium sulfate fractionation and by adsorption on
silica gel. The final preparation, which was approximately 200 times as
active as the crude aqueous extract, was capable of converting adenine to
inosine in the presence of rib ose-1 -phosphate. 202 It is as yet undecided

i9« L. A. Heppel and R. J. Hilmoe, J. Biol. Chem. 198, 683 (1952).

1" C. E. Cardini, A. C. Paladini, R. Caputto, and L. F. Leloir, Acta Physiol. Lati-

noamer. 1, 57 (1950).
188 M. Friedkin, J. Am. Chem.. Soc. 74, 112 (1952).
199 M. Friedkin, Federation Proc. 11, 216 (1952).

20° M. L. Schaedel, M. J. Waldvogel, and F. Schlenk, /. Biol. Chem. 171, 135 (1947).
2«i E. D. Korn and J. M. Buchanan, Federation Proc. 12, 233 (1953).
202 E. D. Korn, F. C. Charalampous, and J. M. Buchanan, J. Am. Chem. Soc. 75,

3610 (1953).

ENZYMES ATTACKING NUCLEIC ACIDS 603

whether the resistance of adenine toward the cruder enzyme preparation
from rat Hver is a peculiarity of this species or whether it is caused by the
presence of an inhibitor in these enzyme preparations. Rowen and Korn-
berg-°' found that hver contains a nicotinamide nucleoside phosphorylase
and reported evidence in favor of its identity with the purine nucleoside
phosphorylase. Reduced nicotinamide riboside is not cleaved by the en-
zyme.^°*

(2) Enzymes of Microorganisms. According to Heppel and Hilmoe,^^^ the
yeast enzyme splits only guanosine, hypoxanthosine, and nicotinamide
nucleoside, but is without action toward adenosine, xanthosine, and some
synthetic purine nucleosides. No information on the behavior of the deoxy-
ribonucleosides toward the yeast enzyme is available as yet. E. coli (strain
15, American Type Culture Collection Xo. 9723) contains potent purine
and pyrimidine deoxynucleoside phosphorylases according to Manson and
Lampen.-°^

Mechanism of the Reaction (see also Chapter 24). With hypoxanthosine
as substrate, the following reaction is catalyzed by the purine nucleoside
phosphorylase :^^^

hypoxanthosine + phosphate ;^ hypoxanthine + ribose-1-phosphate

The phosphate in this reaction can be replaced by arsenate; Heppel and
Hilmoe state that the rate of arsenolysis of hypoxanthosine is approxi-
mately two-thirds of that of its phosphorolysis. 1-Riboarsenate undergoes
spontaneous hydrolysis according to Manson and Lampen f^^ this observa-
tion probably accounts for Klein's opinion (based on the results of a reducto-
metric assaj^ procedure) that arsenate would be a more efTective activator
of nucleosidase than phosphate. It follows from the eciuation that, for ex-
ample, a mixture of guanine riboside, hypoxanthine, and phosphate is
transformed by the enzyme into a mixture of guanine and hypoxanthine
ribosides and ribose-1 -phosphate. This has been verified by Friedkin and
Kalckar-"^ for the deoxyribosides.

According to Friedkin and Kalckar, a mixture of ribo- and deoxyribo-
nucleosides is not phosphorylated at a higher rate than ecjuimolar amounts
of the pure nucleosides. This suggests that the phosphorolysis of purine
ribo- and deoxyribonucleosides is catalyzed by the same enzyme.

Influence of pH. The optimum for the action of the yeast enzyme was
found near pH 7 by Heppel and Hilmoe ;^^^ the experiments with the liver

"3 J. W. Rowen and A. Romberg, J. Biol. Chem. 193, 497 (1951).
2"^ H. M. Kalckar, Biochim. et Biophijs. Acta 12, 250 (1953).
^oe L. A. Manson and J. O. Lampen, J. Biol. Chem. 193, 539 (1951).
"« L. A. Manson and J. O. Lampen, J. Biol. Chem. 191, 95 (1951).
2" M. Friedkin and H. M. Kalckar, /. Biol. Chem. 184, 437 (1950).

604 G. SCHMIDT

enzyme were carried out at pH 7.5, but no systematic pH activity curves
have been reported.

Influence of Substrate Concentration. The Michaelis-Menten constants of
the calf Hver enzyme are 1.7 X 10'^ M for ribonucleosides, 1.8 X 10"^ M
for deoxyribonucleosides, and 1.1 X 10"^ M for nicotinamide nucleoside.
According to Heppel and Hilmoe/^® the yeast enzyme has the higher Km
values of 1.1 X 10"^ M for guanosine, 6.5 X lO"" M for nicotinamide ribo-
side, and 8 X lO"" M for phosphate.

Equilibrium Constant. The equilibrium of the phosphorolysis of purine
nucleosides is in favor of nucleoside formation. The equilibrium constant
for hypoxanthosine phosphorolysis at pH 7.4 is 0.03 according to deter-
minations with the yeast enzyme. The value for nicotinamide nucleoside
phosphorolysis in which H+ enters is 1 X lO-^^"' ^os

Stability. Highly purified liver enzyme preparations can be stored at
-20° with very little loss, whereas the yeast nucleoside phosphorylase is
less stable, at least in highly purified conditions.

Assaij Methods. In the earlier investigations of the period preceding the
work of Kalckar.i^" ^hg activity of nucleosidases was assayed reducto-
metrically. The liberation of reducing groups after the action of nucleo-
side phosphorylases is, however, a secondary reaction caused by acid
hydrolysis of the highly acid-labile ribose-1 -phosphate.

In recent investigations, the action of the enzyme was measured either by
determination of inorganic phosphate or of the liberated purines according
to the sensitive and convenient procedures developed by Kalckar.^^* The
formation of deoxyribosides can also be followed with the microbiological
method of Hoff-j0rgensen2O9.2io which is based on the observation that
Thermobacterium acidophilus R26 (Orla Jensen Collection) requires deoxy-
ribosides for growth.

b. Pyrimidine Nucleoside Phosphorylases

The first observations suggesting the existence of specific phosphorylases
for pyrimidine nucleosides were made by Deutsch and Laser,!^^ ^y^Q dis-
covered the power of bone marrow extracts to cleave deoxyribonucleosides
of pyrimidines, and by Klein, ^^^ who demonstrated the phosphate— or
arsenate — requirement of these enzymes. He also found that kidney ex-
tracts were much more active toward uridine than toward deoxyriboguanine,
whereas the opposite behavior was found for spleen extracts. In Klein's ex-
periments guanine "inhibited" the cleavage of purine nucleosides, but not
that of uridine. The cleavage of cytidine was much slower than that of uri-
dine and thymidine.

208 L. J. Zatman, N. O. Kaplan, and S. P. Colowick, J. Biol. Chem. 200, 197 (1953).

203 E. Hoff-J0rgensen, J. Biol. Chem. 178, 525 (1949).

210 E. Hoff-J0rgensen, M. Friedkin, and H. M. Kalckar, J. Biol. Chem. 184, 461 (1950).

ENZYMES ATTACKING NUCLEIC ACIDS 605

Paege and Schlenk^^^ demonstrated the presence of potent uracil ribo-
nucleoside phosphorylases in aciueous cell-free extracts of E. cold, Aerohacter
aerogenes, and Micrococcus hjsodcicticus. Cytidine was inert toward these
enzymes.

Direct evidence for the existence of specific pyrimidine nucleoside phos-
phorylases Avas subsequently furnished by the purification of two enzyme
fractions whose properties will be briefly described below.

(1) Animal Thymidine Phosphonjlase. The reversible phosphorolytic
cleavage of deoxythymidine was first observed in kidney extracts by Man-
son and Lampen.'-^''"''' Friedkin-"<='<i obtained from horse liver a highly pu-
rified thymidine nucleoside phosphorylase.

Specificity. Friedkin's enzyme catalyzes the reversible exchange of the
phosphoryl group of deoxyribose-1 -phosphate and- — at slower rates — of
ribose-1-phosphate with thymine or uracil. 2-Thiouracil, 5-aminouracil,
5-iodouracil, and 5-bromouracil are likewise substrates of the enzyme.
2-Deoxy thymidine was isolated by Friedkin and Roberts ;^^^'='i thiouracil
riboside and deoxyriboside were isolated by Strominger and Friedkin^^'^
from digests of the respective substrate mixtures with thymidine phos-
phorylase.

Cytosine and orotic acid, as well as adenosine, guanosine, and hypo-
xanthosine, are inert in the enzyme system.

Equilibrium. Since thymine inhibits the reaction by an unknown mech-
anism, a well-defined efjuilibrium constant has not been obtained. At suf-
ficiently high concentrations of deoxyribose-1-phosphate 80 to 90% thymine
formation was observed.

yH optimum. The pH optimum for the enzymic arsenolysis catalyzed by
thymidine phosphorylase is between 5.7 and G. The activity decreases rap-
idly at the acid, slowly at the alkaline, side of the optimum and has a
plateau at 03% of the optimal activity between pH 8.1 and 8.8.

Heat inactivation. The activity decreases by 9% during 8 miiuites' heat-
ing at 50°, by 44 % at 00°, and by 97 % at 70°.

(2) Bacterial Uridine Phosphorylase. A specific uridine phosphorylase was
purified from E. coli E-2G (Iowa State College Laboratory culture) by Paege
and Schlenk'-"' by ammonium sulfate fractionation, adsorption on alumi-
num hydroxide, and elution with sodium phosphate.

Specificity. The enzyme catalyzes only the reversible phosphorolysis of

2" L..M. Paege and F. Schlenk, Arch. Biochem. 28, 348 (1950).

2'>» L. A. Manson and J. O. Lampen, Federation Proc. 8, 224 (1949).

2'"^ L. A. Manson, Thesis, Washington University, St. Louis, 1949.

2'"= M. Friedkin and D. Roberts, J. Biol. Chern. 207, 245 (1954).

*"<iM. Friedkin and D. Roberts, J. Biol. Chern. 207, 257 (1954).

2"«D. B. Strominger and M. Friedkin, J. Biol. Chern. 208, 663 (1954).

"" L. M. Paege and F. Schlenk, Arch. Biochem. and Biophys. 40, 57 (1952).

606 ^- SCHMIDT

uracil riboside. Adenosine, guanosine, hypoxanthine, ribocytidine, and de-
oxy thymidine are inert toward the enzyme.

Equilibrium constant. The equilibrium constant of 0.62 is less in favor of
the synthetic direction than that of purine nucleoside phosphorylases.

pH Optimum. The optimal pH is near pH 7.2.

Assaij. The activity of the enzyme can be measured either by colori-
metric pentose determination (pyrimidine ribosides yield only very little
pigment in comparison with equimolar amounts of ribose-1 -phosphate or
free ribose) or by ultraviolet spectrophotometry.

2. Nucleoside Hydrolases
a. Purine Nucleoside Hydrolase of Yeast

Heppel and Hiknoe^^^ found in 1952 that autolysates of bakers' yeast
were capable of splitting nucleosides in absence of phosphate or arsenate.
They succeeded in separating the nucleoside hydrolase from the nucleoside
phosphorylase by ammonium sulfate fractionation and in achieving further
purification of the hydrolase by subsequent adsorption on calcium phos-
phate gel. The presence of a purine nucleoside hydrolase in potatoes was
likewise demonstrated by Heppel and Hilmoe.

Specificitij. The purified nucleoside hydrolase splits adenosine, guanosine,
hypoxanthine, xanthosine, nicotinamide nucleoside, and some synthetic
nucleosides such as 2,6-diaminopurine-9-ribofuranoside. Amongst the
latter are o-ribopyranosyl adenine, D-glucopyranosyl adenine, D-arabo-
furanosyl adenine, D-ribofuranosyl adenine, which were substituted at the
2-carbon, and D-ribofuranosyl nucleosides of some synthetic purines.

Mechanism. The enzyme catalyzes the reaction:

hypoxanthosine + water = D-ribose -|- hypoxanthine

No exchange reaction was detected in the presence of isotopically labelled
free purine in the reaction mixture.

The pH optimum of the purine nucleoside hydrolase is somewhat higher
than that of the phosphorylase, namely, at about pH 8.

b. Uridine Hydrolase of Yeast

Carter^^^ obtained from autolysates of bakers' yeast in 1950 a highly spe-
cific uridine nucleosidase which did not require phosphate or arsenate for
its activity. No hydrolytic uridine nucleosidase has been found so far in
mammalian tissues. The enzyme was partially purified 10 to 15 times by
ammonium sulfate fractionation. Adenine, inosine, guanosine, cytidine, and
thymidine were not degraded by the enzyme. The reaction followed a uni-
molecular course up to 85% of hydrolysis. The pH optimum was found to
be at pH 7.

ENZYMES ATTACKING NUCLEIC ACIDS 607

Activity Determination. The activity of the hydrolytic uridine nucleosidase
can be determined spectrophotometrically by utilizing the facts that, at
pH 7, the molar extinction of uridine at 280 van is 3.5 X 10^ that of uracil
1.4 X 10^ or that in 0.01 A^ sodium hydroxide the molar extinctions of uri-
dine and uracil are 30 and 5.4 X lO'', respectively.

c. Nucleoside Hydrolases of Lactobacillus pentosus

Lampen and Wang'^'^ demonstrated that extracts of Lactobacillus pento-
sus contain nucleoside hydrolases catalyzing the cleavage of purine ribo-
sides (including uric acid riboside) and pyrimidine ribosides. The hy-
drolysis of the pyrimidine nucleosides was enhanced, that of the purine
nucleosides depressed, by phosphate and arsenate ions. The authors
demonstrated that the enhancing effects of phosphate and arsenate on the
cleavage of the pyrimidine nucleosides were caused by the stabilizing effect
of bivalent anions on the enzyme proteins, but not by the participation
of these ions in the enzyme reaction. The hydrolysis of the purine ribosides
was strongly inhibited by phosphate and arsenate ions.

3. NUCLEOTIDE-I'-PHOSPHORYLASE

The requirement of ribose-5-phosphate and of ATP for nucleotide syn-
thesis in pigeon liver was first recognized by Williams and Buchanan.^*^

The role of ribose compounds esterified with phosphoryl groups in the 1
as well as in the 5 position as intermediaries in the enzymic formation of
nucleotides was suggested by SafTran and Scarano'*' who found that a
phosphorylated ribose believed to be 1,5-ribose diphosphate was trans-
formed into 5'-adenylic acid by pigeon liver homogenates in presence of
C'*-labeled adenine. It is very possible, however, that the actual inter-
mediary in this and in similar reactions is 1-pyrophosphoryl, 5-phosphoryl-
ribose, which was isolated by Kornberg et al.^^^ (see next paragraph).
According to J. M. Buchanan (private communication to the author) this
substance is the intermediary involved in the enzymic synthesis of ino-
sinic acid in pigeon liver extracts in presence of ATP and ribose-5-phosphate.

4. NUCLEOTIDE-I'-PYROPHOSPHORYLASE

According to Kornberg et al.,^^^ a partially purified enzyme fraction from
yeast catalyzes the reversible condensation between adenine or orotic acid
and a pyrophosphoryl ribose compound of the probable strucure of 1-pyro-
phosphoryl-ribose-5-phosphate to 5'-adenylic acid or 5'-orotidyl nucleotide,
respectively. Detailed data regarding the equilibria and the kinetics of
these important enzyme reactions are not as yet available. The phosphoryl-
ated ribose compound was obtained by the action of a purified enzyme

2"8 J. O. Lampen and T. P. Wang, J. Biol. Chem. 198, 385 (1952).

608 G. SCHMIDT

preparation from pigeon liver acetone powder on a mixture of ATP and
ribose-5-phosphate.

Orotidyl nucleotide is decarboxylated by another enzyme fraction from
yeast to uridylic acid. It appears that the enzymic decarboxylation of orotic
acid requires its intermediate conversion to 5'-orotidylic acid.

The complete enzyme system is therefore capable of converting adenine
or orotic acid in the presence of ribose-5-phosphate and ATP to 5'-ade-
nylic or 5'-uridylic acid, respectively. The presence of the complete enzyme
system in rat and pigeon liver has been demonstrated by Hurlbert and
Reichard.^i'^'

Karger and Carter-'- obtained evidence for the enzymic conversion of
orotic acid to uracil by liver extracts and for the essential role of uridylic
acid as an intermediary step in this conversion.

5. DPN Hydrolases

A more detailed description of the enzymes which hydrolyze DPN to
nicotinamide and adenosine diphosphorylribose is beyond the scope of this
book since they are not enzymes of nucleic acid metabolism in the strict
sense. Two aspects, however, are of interest in correlation with the proper-
ties of the enzymes discussed in this chapter: (a) The group of DPN hydro-
lases is so far the only group of hydrolases known to cleave an A^'-glycoside-
like linkage in a compound whose structure resembles that of dinucleotides.
(b) Some DPN hydrolase preparations catalyze phosphate-independent
exchange reactions of the bound nicotinamide group of DPN. With regard
to these exchange reactions, the occurrence of three different types of DPN
hydrolases has so far been established: (1) DPN hydrolase of Neurospora
crassa was obtained as a soluble, highly purified preparation by Kaplan,
Colowick, and Ciotti.^^' It is competitively inhibited by nicotinamide in
relatively high concentration (1.45 X 10~^ M is required for 50 % inhibition)
and does not catalyze the exchange of the nicotinamide group of the sub-
strate with free nicotinamide. (2) DPN hydrolases of some animal organs
such as beef spleen, lung, and brain. The enzyme of beef spleen was obtained
as a particulate centrifugal fraction, the specific activity of which was ap-
proximately 1 % that of the purified Neurospora enzyme. It is inhibited to
a degree of 50% by a 1.5 X 10~^ M nicotinamide concentration. This inhi-
bition is noncompetitive with respect to DPN, and must be attributed to
the enzymically catalyzed exchange of the bound nicotinamide group with
free nicotinamide. The exchange was demonstrated with C'Mabelled nico-
tinamide.^'* (3) Whereas the action of some of the animal DPN hydrolases

"ih R. B. Hurlbert and P. Reichard, Acta Chem. Scand. 8, 701 (1954).

^"2 B. Karger and C. E. Carter, personal communication (1954).

2" N. O. Kaplan, S. P. Colowick, and M. M. Ciotti, J. Biol. Chem. 194, 579 (1952).

2'* L. J. Zatman, N. O. Kaplan, and S. P. Colowick, J. Biol. Chem. 200, 197 (1953).

ENZYMES ATTACKING NUCLEIC ACIDS 609

is strongly inhibited by the antituberculous drug isonicotinic acid hydrazide,
that of some other animal DPX hydrolase preparations (enzyme prepara-
tions from pigs' brain, human spleen and prostate) is not appreciably in-
fluenced by this substance.-^* Zatman ct al. demonstrated that the latter
enzyme preparations are capable of catalyzing the exchange of the nico-
tinamide group of DPX with isonicotinic acid hydrazide.

They isolated from incubation mixtures of the enzyme from pigs' brain
with DPX and isonicotinic hydrazide a compound which appears to be an
analogue of DPX in which the nicotinamide group is replaced by isonico-
tinic acid hydrazide. They therefore postulate as a working hypothesis the
formation of a complex between the enzyme and the adenosine diphospho-
ribosyl group. This complex is believed to contain the free energy of the
glycosidic bond. Water and nicotinamide or isonicotinic hydrazide are be-
lieved to compete for the reactions with this complex.

VII. Xanthine Oxidase

The discovery of xanthine oxidase dates back to Horbaczewski,^^® who
observed in 1882 that extracts of various tissues were capable of oxidizing
xanthine to uric acid. The early history of the enzymic formation and de-
struction of uric acid is connected with the work of many outstanding in-
vestigators at the end of the 19th century and cannot be re\'iewed in detail
in this article. Much pertinent information can be found in the monograph
on uric acid by McCrudden.-'^ It is worth mentioning that as early as 1905,
Schittenhelm-^* successfully applied fractionation with ammonium sulfate
and dialysis to the partial purification of xanthine oxidase from spleen ex-
tracts. His preparation was a mixture of xanthine oxidase and guanase,
but was free of uricase activity. The model for Schittenhelm's method
was the earlier use of ammonium sulfate for enzyme fractionation by
Jacoby,^'^ who developed his procedure in 1899 in Hofmeister's laboratory.
Morgan, Stewart, and Hopkins--^ found that milk contained a dehydro-
genase which catalyzed the oxidation of hypoxanthine and xanthine to uric
acid, and raised the question of its identity with the aldehyde dehydro-
genase discovered earlier in milk by Schardinger.^-^

2'5 L. T. Zatman, N. O. Kaplan, S. P. Colovvick, and M. M. Ciotti, J. Am. Chem. Soc.

75, 3293 (1953).
"« J. Horbaczevvski, Monatsh. 12, 221 (1882).
2'7 F. H. McCrudden, "Uric Acid." Boston, 1905.
2'« A. Schittenhelm, Z. physiol. Chem. 43, 228 (1904).
215 M. Jacoby, Z. physiol. Chem. 30, 135 (1900).

220 E. J. Morgan, C. P. Stewart, and F. G. Hopkins, Proc. Roy. Soc. (London) B94,
109 (1922-1923).

221 F. Schardinger, Z. Untersuch. Nahr. u. Gemissm. 5, 22 (1902).

610 G. SCHMIDT

Ball'-^- and later Corran, Dewan, Gordon, and Green--^ succeeded in a
considerable purification of the enzyme obtained from milk and established
its nature as that of a conjugated protein containing a flavin adenine dinu-
cleotide as a prosthetic group which can be separated from the protein by
heat or treatment with acid or alcohol. The prosthetic group of xanthine
oxidase can replace the coenzyme of amino acid oxidase in Warburg and
Christian's^-^ test, and both flavin adenine dinucleotides have closely similar
absorption spectra.

Evidence suggesting the participation of the flavin group of xanthine
oxidase in its catalytic activity was obtained by Ball, and his conclusions
were later supported by Horecker and Heppel'"^ and by MorelP^^ on much
purer enzyme preparations. Morell established the proportionality of the
catalytic activity of his preparations with their contents of a specific flavin
compound. This was inferred from the fact that only a part of the total
flavine of xanthine oxidase is reduced by the substrate. According to ob-
servations by Mackler et al.,"^^^ this behavior must be attributed to the
oxido-reduction equilibria between the enzyme- and the substrate-systems.
The standard potential of xanthine oxidase is more negative than those of
other flavoproteins.

The essential role of the flavin adenine dinucleotide for the catalytic
action of xanthine dehydrogenase is further supported by the striking
parallelism between the riboflavin content of the diet and the xanthine
oxidase concentration in the tissues of growing rats--^ —a parallelism which
resembles the behavior of other respiratory enzymes.

Methylene blue,^-" oxygen,^^- or cytochrome c-^^ can act as hydrogen ac-
ceptors.

In addition to flavine adenine dinucleotide, xanthine oxidase has an-
other, possibly iron containing chromophoric group.--^^'^"^ The presence of
this group is the reason for the differences between the absorption spectrum
of xanthine oxidase and that of a flavoprotein.

The complex nature of xanthine oxidase is impressively demonstrated
by the surprising identification of sodium molybdate as an additional
dietary factor required for maintaining normal xanthine oxidase levels in

"2 E. G. Ball, /. Biol. Chem. 128, 51 (1939).

2" H. S. Corran, J. G. Dewan, A. H. Gordon, and D. E. Green, Biochem. J. 33, 1694

(1939).
"4 O. Warburg and W. Christian, Biochem. Z. 298, 150 (1938).

225 B. L. Horecker and L. A. Heppel, J. Biol. Chem. 178, 683 (1949).

226 D. B. Morell, Biochem. J., 51, 657 (1952).

226" B. Mackler, H. B. Mahler, and D. E. Green, J. Biol. Chem. 210, 149 (1954).

227 E. C. de Renzo, E. Kaleita, P. Heyther, J. J. Cleson, B. L. Hutchings, and J. H.
Williams, /. Am. Chem. Soc. 75, 753 (1953).

22'" D. A. Richert and W. W. Westerfeld, J. Biol. Chem. 209, 179 (1954).

ENZYMES ATTACKING NUCLEIC ACIDS 611

the livers and particularly in the intestines of weanling rats and some other
mammals.227 228 According to Westerfeld and his collaborators,--^ xanthine
dehydrogenase preparations purified by Ball's method contained 0.03%
molybdenum. It is possible that the molybdenum-containing group repre-
sents the "second active group" of xanthine oxidase already suspected by
Ball.222

The role of molybdenum as a functionally important component of
xanthine oxidase was demonstrated by Totter,--*=^ by Green and Beinert,-^''
and by Mackler et alr^^'^ Totter et al. found that labeled molybdate in-
jected into a cow appeared in the xanthine oxidase isolated from the milk,
and that the proportion between the molybdenum and the flavine remained
constant at the value of 0.5. Green and Beinert arrived at the same pro-
portion in their analyses of highly purified preparations of the enzyme.
According to Mackler et al, molybdenum can be removed from the enzyme
by dialysis in alkaline solution. The molybdenum-free enzyme catalyzes the
oxidation of substrate by indophenol or oxygen, but is incapable of reacting
with one-electron acceptors such as cytochrome c or ferric cyanide. On
addition of molybdic oxide, this catalytic function is restored. This specific
role of the metal in facilitating the reaction of the enzyme with one-electron
acceptors applies to other enzymes belonging to the group of metalloflavo-
proteins.

The metal-free xanthine oxidase can be partially reactivated by uranyl
ions, but not by tungstate or by any representative of a large number of
other metals. The metal-catalyzed oxidation of the substrate by one-elec-
tron acceptors has an absolute requirement for phosphate ions.

The role of molybdenum is not an exceptional property of xanthine
oxidase since the nitrate reductase of Neurospora crassa was shown by
Nason and Evans-''''= to contain flavine adenine dinucleotide and molyb-
denum as essential constituents.

According to Corran et al.,'^-^ the xanthine oxidase of milk is most likely
identical with xanthine oxidase of liver, but some observations suggest that
the enzymes in milk and in tissues might be associated with different
carrier proteins.

Assay Methods. Xanthine oxidase may be assayed by the reduction of
methylene blue,-'-^ •-^'' by the manometric determination of oxygen up-

"8 D. A. Richeit and W. W. Westerfeld, J. Biol. Chevi. 203, 915 (1953).

229 W. W. Westerfeld, J. M. McKibbiiis, J. C. Roemel, and M. F. Hilfinger, Avi. J.
Physiol. 157, 18-1 (1949).

229a J. R. Totter, W. T. Burnett, Jr., R. A. Monroe, I. B. Witney, and C. L. Comer,

Science 118, 555 (1953).
229b D. E. Green and H. Beinert, Biochim. et Biophys. Acta 11, 599 (1953).
2290 A. Nason and H. J. Evans, J. Biol. Chem. 202, 655 (1953).

230 V. H. Booth, Biochem. J. 32, 494 (1938).

612 G. SCHMIDT

take,^-2'2'''2*'^ by the spectrophotometric determination of reduced cyto-
chrome c, or by determination of the uric acid formed.

Specificity P^ Xanthine oxidase catalyzes the oxidation of hypoxanthine,
xanthine, adenine (to dihydroxyadenine),^^^ 8-hydroxypurine, 6-amino-8-
hydroxypurine, 6-amino-2-hydroxypurine, 2-azaadenine,2^-^ and xanthop-
terin.-^^ It is noteworthy that the oxidation of the aminopurines is not
accompanied by the formation of ammonia.

Even highly purified preparations of xanthine oxidase are active toward
a considerable number of aliphatic and aromatic aldehydes and toward
DPNH.

Certain substances related to aldehydes such as chloral hydrate, butyryl-
chloraldehyde, paraldehyde, glucosone, glucose, mannose, and gluconate
are not oxidized by the enzyme.

Inhibitors. Xanthine oxidase is completely and irreversibly inhibited by
low sodium cyanide concentrations (e.g., 0.0025 M^). 222 ,235.235a ^he reaction
of the cyanide with the enzyme protein takes an appreciable time and the
cyanide effect can only be demonstrated if the solutions of the enzyme and
of the cyanide are mixed before the addition of the substrate. Hypoxanthine
and other purines, e.g., uric acid, protect the enzyme against the effect of
cyanide, presumably because of their higher affinities to the enzyme. The
fact that not only the oxidation of hypoxanthine and xanthine, but also that
of aldehydes, is protected by purines against the effect of cyanide strongly
suggests the identity of xanthine oxidase and the Schardinger enzyme.
Xanthine oxidase is inhibited by borate ions, most likely owing to the
formation of borate complexes with the ribityl group of the riboflavin
moiety of the enzyme. In experiments of Roush and Norris,-^^ the inhibition
which is competitive was 50% at concentrations of 0.03 M borate and
0.034 mM xanthine.

Effect of some substances of quinoid structure. Xanthine oxidase of liver is
practically completely inhibited by oxidized p-aminophenol at 1 X lO""*
jl^ 236a 'pi^g inhibition can be fully reversed by reductive removal of the
inhibitor from the enzymic digest. With the milk enzyme, the effects of

"' D. A. Richert, S. Edwards, and W. W. Westerfeld, J. Biol. Chern. 181, 254 (1949).

"la S. B. Dhungat and A. Sreenivasan, /. Biol. Chem. 208, 845 (1954).

"2 H. Klenow, Biochem. J. 50, 404 (1952).

2"a E. Shaw and D. W. Woolley, J. Biol. Chem. 194, 641 (1952).

2" H. Wieland and R. Liebig, Ann. 555, 146 (1944).

"^ H. M. Kalckar, N. O. Kjeldgaard, and H. Klenow, Biochim. el Biophys. Acta 5,

575 (1950).
"5 A. Szent-Gyorgyi, Biochem. Z. 173, 275 (1926).

"sa M. Dixon and D. Keilin, Proc. Roy. Soc. (London) B119, 159 (1936).
"6 A. Roush and E. R. Norris, Arch. Biochem. 29, 344 (1950).
23«a F. Bernheim and M. C. L. Bernheim, /. Biol. Chem.. 123, 307 (1938).

ENZYMES ATTACKING NUCLEIC ACIDS 613

quinone imine are variable, presumably because of the presence of a sub-
stance which reacts with quinone imine. Xanthine oxidase of milk is in-
hibited to an extent of 50% by pyrogallol at 1 X lO""* M concentration.-^^''
Other known oxidases of liver are not appreciably inhibited by quinone
imine. Many other phenols such as p-nitrophenol, p-bromophenol, phloro-
glucinol, anthraquinone, and a-naphthol do not significantly influence the
action of xanthine oxidase.

Effect of pterin derivatives. Kalckar et al.-^^'^ found that 2-amino-4;-hydrox3^-
6-formylpterin inhibits xanthine oxidase practically completely at a
concentration of 10""* M. 2-Amino-4-hydroxy-6-carboxypterine"^^'* and 2-
amino-4-hydroxy-6-hydroxymethylpterin'^*'^ inhibit the enzyme at higher
concentrations. Other respiratory enzymes tested were not inhibited by
these pterins.

It is of interest that these specific inhibitors of xanthine oxidase also have
strong depressing effects on the endogenous oxygen consumption of liver
homogenates. This raises the question of the role of xanthine oxidase in the
endogenous respiration of isolated tissue preparations.

L-Ascorbic acid-^^^ at 1 X 10~^ M inhibits irreversibly most of the activ-
ity of xanthine oxidase purified from cream.

Antabuse-''' (tetraethylthiuram disulfide, (C2H5)2N(CS) -S-S- (SO)X-
(C2H5)2 , has a considerable inhibitory effect on crude and purified prepara-
tions of xanthine oxidase from rat liver, but not on preparations from milk.
The inhibition of the liver enzyme can be overcome by methylene blue.

The differences in the behavior of xanthine oxidase preparations from
milk and from tissues toward inhibitors are ascribed to the association of
the enzyme with different carrier substances depending on the source of the
preparations. An interesting example of such an association is furnished by
the observation of Philpot,-^^^ who found that the xanthine oxidase activity
of milk is small until it was exposed to temperatures below 15°. The solidifi-
cation of the fats releases the enzyme which was associated with the fat
globules of the milk.

Copper ions inactivate xanthine oxidase in verj' small concentrations.-^^*'
The use of water distilled from copper \-essels must therefore be avoided
during experiments with this enzyme.

""b s. J. Gray and R. Z. Felsher, Proc. Soc. Exptl. Biol. Med. 59, 287 (1947).

2"c H. M. Kalckar, N. O. Kjeldgaard, and H. Klenow, J. Biol. Chem. 174, 771 (1940).

"6d H. G. Petering and J. A. Schmitt, /. Am. Chem. Soc. 72, 2995 (1950).

236ep Feigelson, /. Biol. Chem. 197, 843 (1952).

"6fD. A. Richert, R. Vanderlinde, and W. W. Westerfeld, /. Biol. Chem. 186, 261

(1950).
"«8F. J. Philpot, Bioc.hem. J. 32, 2013 (1938).
"6h M. R. Stetten and C. L. Fox, Jr., J. Biol. Chem. 161, 333 (1945).

614 G. SCHMIDT

VIII. Enzymes Involving the Opening of the Purine Ring

Up to the discovery of 4-amino-5-imidazolecarboxamide"^^^'' and its
role in the formation of the purine ring, the only enzyme known to catalyze
the biological degradation of the purine ring was uricase. Whereas the latter
enzyme reaction is most likely irreversible in the sense that no evidence
exists for the participation of allantoin and its intermediary precursors in
purine biosynthesis, the formation of 4-amino-5-imidazolecarboxamide com-
pounds has been shown to be an important step in the biological formation
of the purine ring. The study of the enzymes involved in the metabolism of
4-amino-5-imidazolecarboxamide is still in the early stages and the role of
this compound in purine biosynthesis is discussed in Chapter 23, but, in
anticipation of the increasing importance of this field, a brief review of the
available observations is included in this chapter.

In contrast to the action of uricase, the enzyme reactions involving the
4-amino-5-imidazolecarboxamide group occur exclusively on nucleotide
compounds.

1. Inosinic Acid Transformylase

An enzyme fraction obtained from pigeon liver extracts by fractional
precipitation with alcohol catalyzes the reversible-^^J ••= transformation of
inosinic acid to 4-amino-5-carboxamide ribotide in the presence of certain
formyl acceptors. The yield of the reaction product is strongly dependent
on the concentration of inosinic acid. The reaction requires ATP and mag-
nesium ions.^'®'-™ It is enhanced by glycine (but not by other amino acids
such as L-methionine, L-alanine, L-serine, L-leucine, and sarcosine) and by
leucovorin. The reaction is strongly stimulated by copper ions at 1 X 10"'' M.

The enzyme is completely inhibited by cyanide or versene.

Assay. The action of inosinic transformylase can be followed by the de-
termination of the quantities of the arylamine formed by means of the
diazo reaction of Bratton and Marshall.^^^"

2. Uricase

History. The earliest observations on the power of mammalian tissues to
destroy uric acid date back to Ascoli-" and Wiener.^^^ Schittenhelm^^^ came

"6i W. Shive, W. W. Ackermann, M. Gordon, M. E. Getzendaner, and R. E. Eakin,

J. Am. Chem. Soc. 69, 725 (1947).
236i J. M. Buchanan and M. P. Schulman, J. Biol. Chem. 202, 241 (1953).
236k J. G. Flaks and J. M. Buchanan, J. Am. Chem. Soc. 76, 2275 (1954).
2361 G. R. Greenberg, Federation Proc. 13, 221 (1954).
236m Q R Greenberg, Federation Proc. 13, (1954), in press.
2'6n A. C. Bratton and E. K. Marshall, Jr., /. Biol. Cherti. 128, 537 (1939).

237 G. Ascoli, Pflilgers Arch. ges. Physiol. 72, 340 (1898).

238 W. Wiener, Z. physiol. Chem. 43, 532 (1904).

239 J. A. Schittenhelm, Z. physiol. Chem. 45, 121 (1905).

ENZYMES ATTACKING NUCLEIC ACIDS 615

to the conclusion that the oxidation of xanthine and that of uric acid were
catalyzed by different enzymes and introduced the name uricase for the
enzyme system responsible for the latter process. In 1007, Wiechowski-^"
identified allantoin as the main end-product of enzymic uricolysis. This sub-
stance had been known much earlier as the main reaction product of oxi-
dation of uric acid with permanganate. The first thorough study of the
mechanism of uricase action was carried out by Batelh and Stern .-^^

Definitioti. Uricase is a uric acid oxidase which^ — under conditions resem-
bling those in hving cells — transforms uric acid in yields of more than 95 %
to racemic-^- allantoin. Under these conditions, the overall reaction closely
approaches the equation given by Keilin and Hartree.-''^ '-"

CO + CO2 + H2O1

0 H

H

HN X'^ \

0 ,N

1 II CO -F O2 -1- 2H2O

= 1 1

oc^ ^c.^ /

OC. n ^

H H

H H

Presen t Concepts Regarding the Mechanism of Enzymic Uricolysis. A compar-
ison of the structures of uric acid and of allantoin shows that enzymic uri-
colysis is a complex process involving oxidation and a decarboxylation
hivolving the opening of the pyrimidine ring of uric acid. This suggests
the assumption that the transformation of uric acid to allantoin is a se-
quence of at least two reactions, one oxidative and one decarboxylating
step. Since it is most unlikely that these two reactions are catalyzed by
the same enzyme, one must either assume, according to Felix et alr'^^ that
uricase consists of a mixture of at least two enzymes, or that uric acid is
oxidized enzymically to a labile intermediary product which subsequently
yields allantoin by a nonenzymic degradation. At present, the latter alterna-
tive is favored as a working hypothesis. In the following, some current con-
cepts regarding the nature of the labile intermediary compound will be
briefly discussed.

Mechanism of the oxidation reaction. Bentley and Neuberger-^^ demon-
strated in experiments with 0^^ and H20^^ that the oxygen of the hydrogen
peroxide formed during uricolysis originated exclusively from gaseous oxy-
gen. The oxidation of uric acid consists therefore in the transfer of two elec-
trons from each uric acid molecule to molecular oxygen.

240 w. Wiechowski, Bcitr. chem. Path. Physiol. 9, 295 (1907).

2^' F. Batelli and L. Stern, Biochem. Z. 19, 219 (1909).

2« L. B. Mendel and H. D. Dakin, J. Biol. Chem. 7, 153 (1916).

2« D. Keilin and E. F. Hartree, Proc. Roy. Soc. (London) B119, 114 (1936).

2" C. G. Holmberg, Biochem. J. 33, 1901 (1939).

2« J. N. Davidson, Biochem. J. 36, 252 (1942).

2^6 K. Feli.x, F. Scheel, and W. Schuler, Z. ph^jsiol. Chem. 215, 258 (1929).

2" R. Bentley and A. Neuberger, Biochem. J. 52, 694 (1952).

616 G. SCHMIDT

According to Keilin and Hartree the quotient C02formed-O2con8umed has the
value of 1 with purified uricase preparations which are free of catalase.
Crude uricase preparations are always contaminated with catalase, and
their action results in an increased respiratory quotient.

Mechanism of the decarboxylation reaction. Progress in the understanding
of uricase action was initiated by three approaches: (1) The direct experi-
mental proof for the origin of the carbon dioxide from the 6-C atom of uric
acid. This evidence was furnished by Bentley and Xeuberger^^'^ by the en-
zymic uricolysis of 6-C^*-labelled uric acid. (2) The experimental separation
of uricase action into its oxidation and decarboxylation phases. The feasi-
bility of such a separation was first demonstrated in 1929 by Felix, Scheel,
and Schuler,^*® who found that, at pH 8.9, the oxygen consumption was
much more rapid than the carbon dioxide formation. A further important
advance was Klemperer's observation^'*^ (1945) that borate ions have a spe-
cific inhibitory effect on the decarboxylation phase and allantoin formation
of uricase action. The study of enzymic uricolysis in borate buffers has
become an indispensable method for the exploration of the mechanism of
uricase action. (3) The oxidation of uric acid with alkaline permanganate
which also results in the formation of allantoin is a useful model reaction in
the search for the unstable intermediary of the biological formation of
allantoin.249-251

Nature of the unstable intermediary product of the enzymic oxidation of uric
acid (see also Chapter 26). Cavalieri and Brown"- found that 1,3-N'^-
labelled uric acid yielded an allantoin which was equally labelled in both
ureido groups. The same holds true for allantoin isolated from the urine of
rats fed with 1 ,3-N^^-labelled uric acid. On the basis of these observations
Brown et al}^^ suggested an oxidation product of symmetrical structure as
the immediate precursor of allantoin. A similar postulate for the permanga-
nate oxidation of uric acid had already been made by Behrend,^^^ who ex-
plained by this assumption earlier observations by Fischer and Ach."^ These
authors found that permanganate oxidation of 1-methyluric acid and 7-
methyluric acid yielded 3-methylallantoin, whereas the oxidation of both
3-methyl uric acid and 9-methyluric acid yielded 1-methylallantoin.

The postulated symmetrical intermediary (Fig. 9) formed during per-
manganate oxidation was isolated by Schuler and ReindeF" as the trisilver

2« F. W. Klemperer, J. Biol. Chem. 160, 111 (1945).

2" R. Behrend, Ann. 333, 141 (1904).

"»R. Behrend and R. Schultz, Ann. 366, 21 (1909).

2" W. Schuler and W. Reindel, Z. physiol. Chem. 208, 248 (1932).

262 L. F. Cavalieri and G. B. Brown, J. Am. Chem. Soc. 70, 1242 (1948).

2" G. B. Brown, P. M. Roll, A. A. Plentl, and L. F. Cavalieri, J. Biol. Chem. 172,

469 (1948).
254 E. Fischer and F. Ach, Ber. 32, 2745 (1899).

ENZYMES ATTACKING NUCLEIC ACIDS

617

Oh Oh

HN^ ^C-" \ _ H^ HN^^C^ \ - 2c

I II co^= = ^=^ I I CO

OC-^ /C. /

H H

+ H2O + JO,

'

'

COOH

H 1 H

OC, CO

\ ^c. /

H H

(

3H

OC.

HN
I
OC.

^N'

^Sc^

^C\

\,

CO

^N

Uric acid isomer with
ionized CH group

III

HN-
OC.

H

.CO

IV

End products

+ HjO'

End products

COOH

I H

^C^ \

OC,

CO

\ ^c^ /

H
II

Fig. 9. Intermediary steps in uricolysis.

Schuler and Reindel's concept

Bentley and Neuberger's concept

salt of hydroxyacetylenediureidocarboxylic acid (I). On heating on a boil-
ing water bath for 30 minutes, in 50% acetic acid, the silver salt yields al-
lantoin in 50% yield; by heating in alkali under similar conditions, the sym-
metric ring compound I is oxidized to uroxanic acid (diureidomalonic acid,
II) with opening of both rings.

Schuler and Reindel reported-^" the isolation of a silver salt identical with
that of substance I from digests obtained by enzymic uricolysis. The initial
increase of the acidity of such digests also favors the assumption of the in-
termediary formation of a carboxyl compound prior to the formation of
carbon dioxide and allantoin.

Bentley and Xeuberger raised two objections to the assumption that hy-
droxyacetylenediureidocarboxylic acid (I) would be the intermediary prod-
uct of enzymic uricolysis: (1) the transient increases of the light absorptions
at 260 and 325 ran during the enzymic uricolysis in phosphate buffers
(Praetorius-^^) ; (2) the relatively drastic conditions required for the non-
enzymic decarboxylation of I to allantoin.

255 W Schuler and W. Reindel, Z. physiol. Chem. 215, 258 (1933).
"« E. Praetorius, Biochim. et Biophys. Acta 2, 602 (1949).

618 G. SCHMIDT

These authors proposed instead for the intermediary a hypothetical struc-
ture which represents a dehydration product of Schuler and Reindel's car-
boxylic acid, and which contains a tautomeric double bond in one of the
cyclic ureide moieties (II). According to these authors, uric acid is dehy-
drogenated to a carbonium compound (III) which isomerizes to an unstable
cyclic quaternary ammonium compound (IV). This is spontaneously
hydrolyzed to II.

Evidence in favor of the nonenzymic nature of the decarboxylation reaction
ofuricolysis. The recent hypotheses regarding the nature of the intermediary
substance or substances^^^ of enzymic uricolysis depend largely on the as-
sumption of a nonenzymic terminal step of decarboxylation. Three reasons
are usually given in favor of the nonenzymic nature of this step and of the
enzymic homogeneity of uricase: (1) the preservation of the complete
enzyme system through numerous fractionation steps in all purification
procedures; (2) the chemical lability of the intermediary products; (3) the
observations initiated by Klempererj^** who found that, in addition to
allantoin, other end products are obtained during the enzymic uricolysis
in borate buffer. In such digests, urea, allantoin, alloxanic acid, and oxal-
uric acid were recently identified by chromatographic procedures.-"

It is clear that none of these three reasons excludes the possibility of the
presence of a specific decarboxylase in the uricase preparations, although the
heterogeneity of the end products obtained in borate buffers renders their
direct enzymic formation unlikely. The spontaneous degradation of the
labile intermediary in the presence of borate could be explained by the pres-
ence of a borate-sensitive decarboxylase in the uricase system.

Specificity. Uricase is a highly specific enzyme, uric acid being the only
purine derivative which is oxidized under its influence. Some uric acid de-
rivatives, however, are competitive inhibitors of uricase, such as 7-methyl-
uric acid and 1,3,7-trimethyluric acid.-*^ Owing to this high specificity,
uricase is used as a specific reagent for the determination of uric acid. Many
of the recent procedures for the purification^^^'^^^ of uricase were devised
for this purpose.

pH Optimum. Keilin and Hartree^^^ found the pH optimum for the action

257 E. L. Canellakis and P. P. Cohen, Federation Proc. 13, 189 (1954).

258 C. G. Holmberg, Biochem. J. 33, 1901 (1939).

259 K. I. Altman, K. Sunell, and G. E. L. Barron, Arch. Biochem. 21, 158 (1949).

260 M. B. Blauch and F. C. Koch, J. Biol. Chem. 130, 443 (1939).

261 H. A. Bulger and H. E. Johns, /. Biol. Chem. 140, 427 (1941).

262 N. K. Schaffer, J. Biol. Chem. 153, 163 (1944).

263 W. D. Block and N. C. Geib, /. Biol. Chem. 168, 747 (1947).
26" H. M. Kalckar, /. Biol. Chem. 167, 429 (1947).

265 H. Silverman and I. Gubernick, J. Biol. Chem. 167, 363 (1947).

266 E. Leone, Biochem. J. 54, 393 (1953).

ENZYMES ATTACKING NUCLEIC ACIDS C19

of purified uricase in the region of 9.2. Since uricase action is usually assayed
by determination of the oxygen uptake or by that of uric acid disappear-
ance, these observations refer to the oxidative phase of uricase action. Felix
ct al.;-^^ who worked with much cruder enzyme preparations, found the opti-
mum pH of the oxidative phase at pH 9.2 and that of the decarboxylation
phase at pH 9.9. Owing to the still problematic nature of the "decarboxyl-
ation reaction," the question of the pH optimum of the decarboxylation
requires further clarification.

Inhibitors. Uricase action is inhibited to an extent of 94 % in presence of
1 X 10-^ M potassium cyanide, and to an extent of 83 % in the presence of
3 X IQ-^ilf potassium cyanide.-^^'-'^^-^'^^ The cyanide inhibition is com-
pletely reversed by dialysis.-"*^

Carbon monoxide is without influence on uricase according to David-
son.2*s

Many heavy metals (Cu, Fe, Mn, Zn, Co, Ni) have considerable inhibi-
tory effects at concentrations in the range of 0.001 il/.'-"

Pyrophosphates, fluorides, and urethan have no inhibitory effect. In the
presence of sodium azide the activity decreases by 28 % at pH 8 and by 40 %
at pH 6.8.

Role of Heavy Metals. The possible significance of the zinc (0.1 %) and iron
(0.2 %) contents of purified uricase preparations is not yet clear.-**^ Prae-
torius concluded from the failure of British Anti-Lewisite (BAL) to inter-
fere with uricase action that the zinc content of uricase preparations has no
connection with their activity .^^^

IX. Enzymes Involving the Opening of the Pyrimidine Ring

Intact animals metabolize administered uracil and thymine with the for-
mation of urea'^°-'^^ or of /3-amino acids"^'' (see also Chapter 26). The catab-
olism of pyrimidines in cell-free systems has so far been studied mainly on
bacterial enzymes. No observations on cell-free systems from animal sources
are as yet available, but the results obtained with tissue slices offer interest-
ing correlations with the mechanisms demonstrated for the action of bac-
terial enzymes.

2" H. R. Fosse, Compt. rend. 191, 1153 (1930).

268 R. Truszkowski, Biochevi. J. 24, 1340 (1930).

269 E. Praetorius, Biochim. et Biophys. Acta 2, 590 (1948).
27» H. Steudel, Z. phrjsiol. Chevi. 32, 285 (1901).

271 L. R. Cerecedo, J. Biol. Chem. 93, 269 (1931).

272 L. B. Mendel and V. C. Myers, Am. J. Physiol. 26, 77 (1910).

273 D. W. Wilson, J. Biol. Chem. 56, 215 (1923).

274 H. Deuel, /. Biol. Chem. 60, 749 (1924).

274a K. Fink, R. B. Henderson, and R. M. Fink, J. Biol. Chem. 197, 441 (1952).

620

G. SCHMIDT

O

HN CH

I II + DPNH + H+

^N I
H COOH

Orotic acid

HN XH2

oa ^CH

^N I
H COOH

L-Dihydroorotic acid

0

HN CHj

I I + H2O

OC. .CH
^N I
H COOH

L-Dihydroorotic acid

H2N COOH

I I
OC CH2

I I

HN-CH
COOH
L-Ureidosuccinic acid

HN-CH-CHj

I
OC

-COOH

+ H2O

III

HN-CH-CHj-COOH

I
OC

HN-CO H2N COOH

L-5-Carboxymethylhydantom L-Ureidosuccinic acid

Fig. 10. Action of dihydroorotic acid dehydrogenase (I), dihydroorotase (II),
and 5-carboxymethylhydantoinase (III).

1. Reversible Degradation Of Orotic Acid by Bacterial Enzymes
The enzymes involved in metabolism of orotic acid are of great interest
since this naturally occurring pyrimidine derivative is utilized for the biosyn-
thesis of nucleic acid pyrimidine groups of various organisms (see Chapter 23).
Lieberman and Romberg"^'' ■<= found in the soil organism Zymobacterium
oroticum^''*^ an enzyme system capable of catalyzing the reversible transfor-
mation of orotic acid to L-ureidosuccinic acid (Fig. 10). The first step of this
overall reaction is the hydrogenation of orotic acid to dihydroorotic acid
under the influence of a specific enzyme, dihydroorotic acid dehydrogenase.
Observations with isotopes leave no doubt that formation of orotic acid
in the mammalian organism follows the same pathway. The enzymic syn-
thesis of ureidosuccinic acid from aspartic acid, carbon dioxide and ammonia
was recently demonstrated by Reichard"^^ with cell-free rat liver prepara-
tions. The reaction requires adenosine triphosphate and magnesium ions.

a. Dihydroorotic Acid Dehydrogenase'^''*^

Dihydroorotic acid dehydrogenase was separated from cell-free extracts
of Zymobacterium oroticum^''*^ by precipitation with protamine and fractiona-
tion of the solution of the protamine precipitate with ammonium sulfate.

^^^'^ I. Lieberman and A. Romberg, Biochim. et Biophys. Acta 12, 223 (1953).
"4c I. Lieberman and A. Romberg, J. Biol. Chem. 207, 911 (1954).
"^d J. T Wachsman and H. A. Barker, /. Bacteriol. 68, 400 (1954).
27<« P. Reichard, Acta Chem. Scand. 8, 795 (1954).

ENZYMES ATTACKING NUCLEIC ACIDS 621

The enzyme reaction requires either the presence of stoichiometric amounts
of reduced DPN or of catalytic amounts of DPX in the presence of a hy-
drogen-donor system.

Specificity. Dihydroorotic acid dehydrogenase is strictly specific for orotic
acid. Uracil, cytosine, thymine, and 5-methylcytosine are inert and have no
influence on the rate of hydrogenation of orotic acid.

pH Optimum. The optimal pH range is between pH 6.4 and 6.5.

The Michaelis-Menten constant has the value of 1.1 X 10~^ M concentra-
tion of orotic acid.

The presence of the enzyme in other microorganisms is suggested by the
fact that dihydroorotic acid is a growth factor for Lactobacillus bulgaricus.

b. Dihydroorotase^'^'^'^

The reversible hydrolysis of dihydroorotic acid to ureidosuccinic acid is
catalyzed by dihydroorotase Avhich is present in cell-free extracts of Zymo-
bacterium oroticum.

Equilibrium. Lieberman and Kornberg measured the equilibrium of the
reaction by using C"-labelled substrates and by determining the total radio-
activity of the components of the incubation mixture after chromatographic
separation. At pH 6.1, the ratio of ureidosuccinate to dihydroorotate at
equilibriam is approximately 1.9.

c. 5-Carboxymcthylhydantoinase-''^''

It is interesting that the formation of dihydroorotic acid is not the only
type of enzymic ring-closure which ureidosuccinic acid undergoes in cell-
free extracts of Zymobacterium oroticum. These extracts contain a specific
enzyme, 5-carboxymethylhydantoinase, which catalyzes the reversible ring-
closure between the ureido group and the vicinal carboxy group to l-5-
carboxymethylhydantoin.

Equilibrium. The eciuilibrium constant is similar to that of dihydrooro-
tase, amounting to the value of 1.9 for the proportion between 5-carboxy-
methylhydantoin and ureidosuccinate.

The specificity of 5-carboxymethylhydantoinase has not yet been investi-
gated. Liver homogenates which hydrolyze hydantoin-^^f were found to be
without activity toward L-5-carboxymethylhydantoin.

2. Conversion of Dihydrouracil and Dihydrothymine to j8-Amino
Acids by Tissue Slices

The reductive pathway of the reversible degradation of orotic acid under
the influence of the enzyme system of Zymobacterium oroticum has an
interesting parallel in the behavior of liver and kidney slices toward di-
hydrothymine and dihydrouracil. Fink et a/."^^ demonstrated by paper

""F. Bernheim and M. Bernheim, J. Biol. Chem. 163, 683 (1946).

2'^8 R. M. Fink, K. Fink, and R. B. Henderson, J. Biol. Chem. 201, 349 (1953).

622 G. SCHMIDT

chromatography the conversion of dihydrouracil to /3-alanine and that of
dihydrothymine to j8-aminoisobutyric acid by slices of hver and — to a
lesser extent — of kidney. Slices obtained from rat, rabbit, guinea pig, cow,
and monkey gave similar results.

Specificity. Only slight formation of /3-aminoisobutyric acid occurred when
thymine was used as substrate. No production of j8-amino acids was de-
tected with uracil, uridine, uridylic acid, cytosine, 5-methylcytosine, and
orotic acid.

The excretion of jS-aminoisobutyric acid by rats after administration of
thymine, of deoxyribonucleic acid, or — to a greater extent- — of dihydrothy-
midine suggests the assumption that the dihydropyrimidines are physiologi-
cal intermediaries in this pathway of mammalian pyrimidine catabolism.
Furthermore, dihydrouracil has been isolated from beef spleen by Funk et

3. Degradation of Uracil and Thymine by Bacterial Enzymes
a. Bacterial Oxidase of Uracil and Thymine (Fig. 11)

Hayaishi and Kornberg,^^^ as well as Wang and Lampen,"^ demonstrated
in cell-free extracts of three different species of soil bacteria the presence of
an enzyme which catalyzed the oxidation of uracil and thymine. The oxi-
dation product of uracil is barbituric acid, that of thymine, 5-methylbar-
tituric acid. Since the determinations of the Michaelis constants gave prac-
tically identical results for the affinities to uracil, whether determined with
uracil as substrate, or with uracil as competitive inhibitor of thymine oxi-
dation, it appears that both pyrimidines are oxidized by the same enzyme.

Specificity. The oxidase is very specific for uracil, thymine, and 5-amino
uracil. Barbituric acid, isobarbituric acid, 5-methylbarbituric acid, 6-
methyluracil, dihydrothymine, dihydrouracil, 2-thiouracil, 2-thiothymine,
cytosine, and methylcytosine are inert.

pH Optimum. The optimal pH is approximately at 8.5; at neutral reac-
tion only one-tenth of the optimal activity was observed.

Mechanism. The presence of methylene blue is required for the action of
the enzyme; so far it could not be replaced by the physiological hydrogen
acceptors or mediators of electron transfer.

Michaelis Constants. The enzyme-substrate dissociation constants are
0.35 X 10-4 M for thymine and 1.3 X 10-" M for uracil.
h. Barhiturase (Fig. 12)

The end-products of the oxidation of uracil by the intact bacteria men-
tioned in the previous section are carbon dioxide, ammonia, and water.

2'^^ C. Funk, A. J. Merritt, and A. Ehrlich, Arch. Biochem. and Bio-phys . 35, 468 (1952).
"5 O. Hayaishi and A. Kornberg, J. Biol. Chem. 197, 717 (1952).
"« T. P. Wang and J. O. Lampen, /. Biol. Chem. 194, 785 (1952).

ENZYMES ATTACKING NUCLEIC ACIDS 623

HN^ ^CH +^0, HN "^CH.

OC CH ^ OC^^/CO

H H

Uracil Barbituric acid

HN^ ^C-CH3 +10, HN HC-CH,

I II * I T

OC^ /CH OC CO

H H

Thymine 5-Methylbarbituric acid

Fig. 11. Bacterial oxidation of uracil and thymine.

0

HN CH^

1 1 +2H,0
OC. XO
^N^

NHs COOH

1 1
CO + CH2

1 1
NH2 COOH

H

Fig. 12. Action of barbiturase.

Hayaishi and Romberg succeeded in obtaining from cell-free extracts of
mycobacteria (obtained at pH 6.65) besides urease an enzyme which cata-
lyzes the cleavage of barbituric acid into urea and malonate. The enzyme
could be separated from most of the urease by purification with protamine
sulfate. In crude extracts, the urease acti\'ity could be selectively suppressed
by the addition of silver ions in 0.5 X 10~^ M concentration.

Specificity. Barbiturase is a highly specific enzyme. 5-Methylbarbituric
acid, orotic acid, barbital, pentobarbital, 2-thiobarbituric acid, and iso-
barbituric acid are inert.

pH Optimum. The optimal range of the activity is between pH 8 and pH
9.

Michaelis Constant. The approximate Km value is 3.4 X 10^* M.

The enzymic mechanism of the final degradation of 5'-methylbarbiturate
which is rapidly metabolized by the intact bacteria under aerobic conditions
is as yet unknown.

X. Some Data Concerning the Intracellular Distribution of Enzymes of
Nucleic Acid Metabolism

Owing to the significance of nucleic acids as constituents of vital struc-
tural elements of cells (chromosomes, nucleoli, mitochondria, Nissl bodies),
the intracellular distribution of the enzymes of nucleic acid metabolism is
of great interest and is discussed also in Chapters 18 and 21. The interpre-

624 G. SCHMIDT

tation of the available data, however, requires caution because various
interfering factors must be considered. Some of those factors are: (1) The
presence of abundant amounts of substrates in some cell fractions, particu-
larly in studies concerning deoxyribonuclease and ribonuclease. (2) The pos-
sibility of absorption of enzymes by particles during homogenization. This
source of error is serious when citric acid procedures are employed for the
isolation of nuclei. (3) The influence of the ionic environment and other ac-
tivators and inhibitors on the action of the enzymes.

1 . Deoxyribonuclease*^'' -"^ -^-^ ,277-279

According to Allfrey and Mirsky,*^^ deoxyribonuclease I appears to be a
specific digestive enzyme of pancreas. It is not found in significant concen-
trations in other tissues, but its activity in pancreas (as differentiated by
means of its pH optimum) exceeds that of other deoxyribonucleases more
than 200-fold. It was found in high concentration in the pancreas fistula
juice of a dog, whereas no appreciable amounts of other deoxyribonucleases
could be demonstrated.

Deoxyribonucleases of acid pH optima were found in many tissues in
which they are mainly associated with the cytoplasmic fraction, particularly
with the mitochondria (Chapter 18). In contrast to some earlier reports, it
appears now that the nuclei of liver, heart, and spleen have little if any de-
oxyribonuclease activity (see also Chapter 21).

Kowlessar et al}^^-"^^^ found that considerable amounts of deoxyribonu-
cleases of acid as well as alkaline pH optima are excreted in the urine by
mice which had been irradiated with lethal doses of X-rays. This suggests
the liberation of intracellular nucleases under conditions producing exten-
sive cell damage.

2. Ribonuclease

According to Roth,^*^ ribonucleases of acid as well as alkaline pH optima
occur in many tissues although the concentration of ribonuclease I in pan-
creas exceeds by far the ribonuclease activity of any other animal tissue.
In liver, the mitochondrial fraction contains a larger part of "acid" and
"alkaline" ribonuclease activity than do other fractions of the homogenates.
Liver nuclei contain small, but measurable, amounts of ribonuclease activity

2" W. C. Schneider and G. H. Hogeboom, J. Biol. Chem. 198, 155 (1952).

"8 R. J. Neff, Thesis, Univ. of Missouri, 1951.

"9 K. Land, G. Siebert, I. Baldus, and A. Corbet, Ex-perientia 6, 59 (1950).

28" O. D. Kowlessar, K. I. Altman, and I. H. Hempelman, Arch. Biochevi. and Biophys.

43, 233 (1953).
"1 O. D. Kowlessar, K. I. Altman, and I. H. Hempelman, Nature 172, 867 (1953).
"2 J. S. Roth, /. Biol. Chem. 208, 181 (1954).

ENZYMES ATTACKING NUCLEIC ACIDS 625

with a single optimum at pH 7.8. Miller and Kozloff-^^ demonstrated the
presence of ribonuclease activity in the nuclear fraction of chicken erythro-
cytes obtained by saponin hemolysis. These authors believ^e that the ribo-
nuclease activity of the nuclei is localized in the nuclear membrane since
it is not increased by freezing and thawing of the enzyme preparation.

3. Intracellular Localization of Deoxyribonuclease I and
Ribonuclease I in Pancreas

The crystallized ribo- and deoxyribonucleases of the pancreas offer the
experimental advantage that specific antibiodies for these two enzymes can
be prepared. Such antibodies were used by Marshall for studies on the intra-
cellular distribution of ribonuclease I and deoxyribonuclease I by histo-
chemical application of the specific antibodies labelled with fluores-
cein.-^'*-*^ Ribonuclease I and deoxyribonuclease I were found in the
zymogen granules and in the cytoplasm, but not in mitochondria and nu-
clei. Schneider and Hogeboom-" found ribo- and deoxyribonuclease activity
in liver mitochondria. Probably the mitochondrial nucleases are not identi-
cal with the crystallized enzymes of pancreas.

4. Other Enzymes of Nucleic Acid Metabolism

According to Stern et al.,^^^ adenosine deaminase, guanase, and nucleoside
phosphorylase are present in high concentrations in nuclei of many tissues,
whereas nucleotide phosphatases and uricase are present only in relatively
small amounts if at all. This question is discussed further in Chapters 18
and 21.

XI. Addendum

{Concerning the Enzymic Formation of Nucleoside Di- and Triphosphates)

When the organization of Chapter 15 was undertaken it was decided to
exclude enzyme reactions involving the pyrophosphoryl groups of nucle-
oside polyphosphates since their discussion seemed to be pertinent to the
field of energy metabolism rather than to that of nucleic acids. During the
last year, however, the presence of polyphosphates of various ribonucleo-
sides in animal tissues was demonstrated by Schmitz, Hurlbert and Potter.^"
All these compounds are phosphorylated in their 5'-positions. Conceiv-

283 Z. B. Miller and L. M. Kozloff, J. Biol. Chem. 170, 105, (1947).
28" J. M. Marshall, Exptl. Cell Research 6, 240 (1954).

285 A. H. Coons, H. J. Creech, R. N. Jones, and E. Berliner, J. Immunol. 45, 159
(1942).

286 H. Stern, V. G. Allfrey, A. E. Mirsky, and H. Laetren, J. Gen. Phijsiol. 35, 559
(1952).

287 H. Schmitz, R. B. Hurlbert, and V. R. Potter, /. Biol. Chem. 209, 41 (1954).

626 G. SCHMIDT

ably, these substances might be intermediaries in the biosynthesis of nu-
cleic acids although no direct evidence in favor of such a hypothesis exists
as yet.

Several enzymes catalyzing the formation of nucleoside polyphosphates
are known. The pathways involved include transphosphorylation between
nucleoside polyphosphates as well as reactions in which inorganic phosphate
participates in the enzymatic synthesis.

(1) Diphosphokinase reactions: Berg and Joklik found that enzyme frac-
tions of muscle and of bakers' and brewers' yeast mediate the reversible
transphosphorylation between adenosine triphosphate and hypoxanthosine
diphosphate with formation of adenosine diphosphate and hypoxanthosine
triphosphate.^*^

(2) Reaction of the myokinase'^^^ type (Transphosphorylations between
nucleoside triphosphates and nucleoside-5'-monophosphates) . According to
Lieberman, Romberg, and SimmSj^^" cell free yeast preparations catalyze
a transphosphorylation between adenosine triphosphate or uridine triphos-
phate, on the one hand, and 5'-uridylate on the other hand, to the diphos-
phates (uridine diphosphate is inert in this system).

(3) Formation of guanosine triphosphate or hypoxanthosine triphosphate
with participation of inorganic phosphate. Sanadi et al.^^^ found that the
enzymic synthesis of adenosine triphosphate from adenosine diphosphate,
inorganic phosphate and succinyl coenzyme A requires the intermediary
formation of either guanosine- or inosine triphosphate from the respective
diphosphates. Guanosine-, as well as hypoxanthosine triphosphate, in turn
can act as phosphate donors in reversible enzymic phosphorylations of
other mono- or dinucleotides.

In some cases^*^'^^" it has been demonstrated that the enzymes cata-
lyzing transphosphorylations between nucleoside triphosphates and nu-
cleoside diphosphates are different from those catalyzing transphos-
phorylations between nucleoside tri- and nucleoside monophosphates. The
discussion of such reactions in the three groups of this chapter, however,
does not imply a general validity of this classification with regard to enzyme
specificit> .

*88 p. Berg and W. K. Joklik, Nature 172, 1008 (1953).

289 S. P. Colowick and H. M. Kalckar, J. Biol. Chern. 148, 117 (1943).

«9o I. Lieberman, A. Kornberg, and E. S. Simms, J. Am. Chern. Soc. 76, 3608 (1954).

"1 R. Sanadi, D. M. Gibson, and P. Ayengar, Biochim. et Biophys. Acta 14, 434 (1954).

Author Index

Figures in parentheses are reference numbers. They are included to assist the reader in locating
a reference when a work is cited, but the author's name is not mentioned.

Abderhalden, E., 95, 576

Aboskar, J. B., 572, 573 (43a)

Abraham, E. P., 183

Abrams, R., 117, 133, 193, 260, 350

Ach, F., 616

Ackermann, W. W., 614

Ahlstrom, L., 318, 475(97)

Aichner, F. X., 17, 18(74), 20(74), 49(74),

59(74), 67(74), 74(74), 78(74)
Akka, R., 584

Albaum, H. G., 34, 35, 301, 313
Albers, H., 13, 181
Albert, A., 103(257), 108(k), 109, 110, 112,

113,114, 115(304), 132(304)
Alberty, R. A., 217, 227
Albrecht, W., 271, 426
Alexander, Hattie, E., 60, 322, 323(91),

328(91), 331(91), 332(139), 338(91),

559, 561(12), 563(12)
Alexander, P., 320(78), 321, 340(77), 531
Alfert, M., 581
Allen, F. W., 198, 245, 385, 387(38), 398

(38), 399(38), 410, 418, 423, 424, 475,

481, 521, 567, 569
Allerton, R., 47, 51(252), 52(252), 59,

60(252, 311), 75(252), 288
Allfrey, V. G., 193, 197(10), 258, 329, 330

(115), 350(115), 354, 355, 356, 357,

579, 580, 584(87a), 624(87a)
Allg^n, L. G., 318, 473(74), 474, 581
Altman, K. I., 618, 624
Altmann, R., 2
Ambelang, J. C., 126
Ambronn, H., 532
Ambrose, E. J., 316, 318(49), 551
Amirad, G., 54
Anand, N., 98, 153, 176, 187
Andersag, H., 93
Andersen, W., 49, 156, 218, 219(20), 221

(20)
Anderson, A., S., 433

Anderson, G. W., 103

Andrews, K. J. M., 153, 154

Andrisano, R., 500, 538(29)

Anfinsen, C. B., 573, 574

Angelos, A., 217, 337, 410, 411, 412, 424

(13), 425(14), 426(13), 479, 481, 482

(122),544, 549, 568
Angermann, H., 298
Aquillonius, L., 5
Araki, T., 576
Arndt, U. W., 318, 531
Asadourian, A., 339, 523, 524, 527(75),

529, 531
Ascoli, A., 3, 88
Ascoli, G., 614
Asimov, I., 246, 329
Askonas, B. A., 598
Astbury, W. T., 453, 454, 461, 462, 530,

539, 550
Astwood, E. B., 103
Atherton, F. R., 173
Atlas, S. M., 470, 471(61)
Austin, W. C., 16, 17, 20(76), 73(76)
Austrian, C. R., 96
Avener, J., 100
Avery, O. T., 319, 322(74), 331(74), 332

(138), 333(74), 584
Avison, A. W. D.,33
Axelrod, B., 35, 248
Ayala, W., 289
Ayengar, P., 626

B

Babson, R. D., 45, 47(240), 72(240), 73

(240)
Bacher, J. E., 198, 385, 387(38), 398(38),

399(38), 419, 423, 475
Bachstez, M., 101, 108(i), 113
Baddiley, J., 6, 8(67), 12, 13, 14, 15, 94,

127, 131, 132(436), 152, 153, 157, 158

(104), 174, 176, 180(199, 211), 184,

627

628

AUTHOR INDEX

185(257), 186, 188, 189, 197, 279, 414,
417

Baer, E., 415, 416(56), 438

Baeyer, A., 82, 103

Baginsky, A., 98

Bahner, C. T., 135

Bailey, C. C, 103

Bailey, K., 310, 311, 376, 535

Bailly, M. C, 414

Bailly, O., 415

Bain, J. A., 568, 570, 571

Baker, B. R., 15(60), 188

Baldus, I., 624

Balis, M. E., 199, 260, 337, 410, 411, 425
(14), 482

Ball, E. G.,610, 611,612(222)

Balston, J. N., 244, 248(5), 263(5)

Bandurski, R. S., 35, 248

Bang, I., 3, 168, 198, 310, 311, 316(11), 319
(11), 323

Baranowski, T., 179, 594

Barbier, M., 391, 403(53)

Barclay, J. L., 29, 30(156)

Barer, R., 534, 542

Barker, G. R., 12, 21(19, 20, 21), 22, 25,
31(19, 20, 21), 43(134), 46, 67(134),
68(134), 69(19, 20, 21), 70(134), 73
(19, 134), 141, 174, 194, 308, 433, 455,
476

Barker, H. A., 122, 620 (274d)

Baron, Frangoise, 37, 41(204), 601

Barrenscheen, H., 300

Barreswill, 95

Barron, G. E. L., 618

Bass, L. W., 3, 7, 82, 96(13), 99(13), 112,
113, 119, 121, 122(13), 123, 124, 136,
144, 156, 168(93), 169(93), 194, 195
(16c), 196(16d, 16e), 198(16), 217,
222, 226, 227, 268, 269(2, 3), 308,
322(1), 341(1), 348(1), 386, 409, 455,
457(22, 23), 459(22), 588

Bassham, J. A., 35, 187, 248

Basu, S., 484

Bate-Smith, E. C., 245

Bates, F. J., 19

Batelli, F., 615

Baudisch, O., 121, 125

Bauer, L., 246

Bauer, W., QQ

Bawden, F. C., 323, 332

Baxter, R. A., 47, 69(254), 154

Bazhilina, G. D., 321

Beard, D., 375

Beard, J. W., 375

Beaven, G. H., 107, 200, 261, 314, 393,
412, 450, 488, 530

Beaven, G. R., 13, 31(37)

Becher, R. A., 424

Beck, A., 109, 119

Becker, B., 17, 20(79), 65(79), 67(79), 181

Becker, E., 97

Beckman, A. O., 494

Beebe, S. P., 5

Behrend, R., 86, 88, 90(42), 101, 616

Behrens, M., 4

Beinert, H., 610, 611

Bell, F. O., 461, 530, 539

Bell, M., 105

Belloni, E., 100, 101

Belozerski, (kii, ky), A. N., 4, 321, 330

Bendich, A., 96, 97, 100, 104(182), 105,
106, 109, 113, 118(214), 119, 122(271),
126, 128(214), 130(346), 131(182, 346,
422), 132(182, 422), 133(182, 438), 138,
159, 188, 361, 447, 495, 498(13), 595

Benedict, S. R., 13, 96, 159

Bennett, E. L., 551

Bennett, L. L., 128, 133(421), 551

Bennett, L. L., Jr., 105, 107

Bensley, R. R., 206

Benson, A. A., 35, 187, 248

Bentley, H. R., 95, 125(136), 159, 248

Bentley, R., 123, 615, 617

Benz, F., 17, 20(79), 65(79), 67(79), 181

Berariu, C., 4

Berg, P., 626

Bergel, F., 93

Berger, L., 19, 29, 30(89), 45(88, 90), 65
(86, 88, 90), 70(86, 87, 89, 157, 158),
71(87, 157, 158)

Berger, R. E., 105

Bergkvist, R., 189

Bergmann, M., 11, 17

Bergmann, W., 98, 126, 159

Bergold, G., 290, 425

Bergstrom, S., 97

Berlin, H., 35

Berliner, E., 625

Bernal, J. D., 451

Bernha^ ' K., 13, 67(31, 32), 68(31), 97,

Bernl , 612, 621

AUTHOR INDEX

629

Bernheim, M. C. L., 612, 621

Beriiheimer, A. W., 390, 583

Bernoulli, A. L., 113

Bernstein, 1. A., 41

Bernstein, M. H., 313, 316, 318(53)

Bessey, O. A., 262

Bethe, A., 95

Bheemeswar, B., 253

Bial, M., 300

Bieber, S., 104

Bielschowsky, F., 67, 298

Biesele, J. J., 105, 107

Biltz, H., 96, 109, 119, 122(163), 131(163)

Birkofer, L., 33, 71(170)

Biscaro, G., 100, 101

Bitterli, P., 106, 127(289), 131(289)

Bjerrum, N., 460

Bjorkman, U., 246

Bjornesjo, K. B., 320(79)

Black, F. L., 206, 207, 403, 405(73)

Blair, V. E., 132

Blanchard, K. C, 4

Blanco, G., 162, 172(143), 366, 440(144),

441
Blanksma, J. J., 12, 15, 17, 18(16), 20(15),
64, 65(16, 100), 67(16, 100), 73(15, 69,
100, 343), 141
Blass, Judith, 21
Blauch, M. B., 618
Block, R. J., 244, 263(4)
Block, W. D., 618

Blout, E. R., 339, 523, 524, 527(75), 529,

531, 545, 546(149), 547(149), 549, 553

Boarland, M. P. V., 108(b), 112, 127, 128

(415), 130(424), 495
Bock, R. M., 189, 217, 227(14)
Bodansky, O., 96
Boehringer, C. F., 119
Boggiano, E. M., 100
Bohonos, N., 95, 100
Bolomey, R. A., 569
Booth, V. H., 595(230), 611, 612(230)
Bortner, H. W., 62, 109, 117, 124(331),
195, 198, 199(22), 200(22, 49), 201(22),
202(49), 204(39), 226, 227, 396, 414,
461, 499, 507, 512(43)
Bosshard, E., 138, 140(1), 197
Bosshardt, D. K., 100
Bottger, I., 474, 578, 581
Boulanger, P., 7, 8(77), 256, 25 , 418

Boyd, G. E., 212, 215(3) — >

Boylen, J. B., 261
Brachet, J., 4, 6, 8, 196, 544
Brackken, H., 25
Bradley, D. F., 187
Brady, T., 64, 75(347), 288, 597
Brahm, C, 82, 96(10)
Brand, E., 574
Bratton, A. C, 614

Brawerman, G., 89, 255, 330, 332, 333(142,
143), 345(134), 347(134, 231), 350
(142), 358, 359, 584, 593, 594(154b,
154c)
Bredereck, H., 20, 22(96), 43, 44, 46, 65
(96), 66(236), 67(96, 234), 94, 96(119),
109, 119, 133, 136, 144, 145, 146(33,
35), 161, 173(38), 174(39), 198, 341,
440, 475(95, 96), 481, 576
Brewster, J. F., 283
Brigl, P., 341
Brill, H. C., 125

Brink, N. G., 13, 31(34, 35), 70(34)
Bristol, H. S., 103
Bristow, N. W., 151
Britton, H. T. S., 42
Brode, W. R., 116, 494
Broomhead, J. M., 451, 452, 457
Brown, A. H., 302

Brown, D. J., 108(k), 109, 112, 113(1), 127,

128(418), 130(418), 131(423), 136, 495,

525(9)

Brown, D. M., 14, 38, 62, 147, 166, 167

(170), 169(173), 170(172), 171(173,

193), 174(170, 171), 175(173), 188,

189^287), 257, 260, 279, 367, 414, 416

(46), 417(54, 63), 418, 419(62), 420

(62), 421, 423(62), 425, 426, 427(109),

428, 430(62), 431, 432, 433, 434, 437,

438(131), 440, 443, 444(62), 479, 545,

559, 564, 565(26), 570(10, 11), 588(11)

Brown, E. B., 119

Brown, E. V., 33, 44(169), 66(169), 72

(169), 73(169)
Brown, F. B., 136

Brown, G. B., 6, 8(69), 44, 60, 61(317),
64(237), 80(317), 94, 96, 97, 104(182),
105, 106, 107, 109, 110, 118(301), 122
(271, 301), 125(370), 126, 131(182), 132
(182, 370), 133(182), 134(301), 135
(301),188, 199, 260,361, 495, 498(13),
595, 616
Brown, G. L , 361, 365, 501, 546(33)

630

AUTHOR INDEX

Brown, G. M., 185

Brown, K. D., 578, 584, 624(119)

Brown, M. E., 135

Brownlie, I. A., 547

Brugnatelli, G., 82, 102

Bruhns, G., 124

Bryan, C. E., 105

Buchanan, J. G., 13, 31(38), 187, 247, 252

Buchanan, J. M., 37, 122, 601, 602, 607,

614
Bulger, H. A., 618
Bunce, B. H., 338, 440, 472(67), 473, 489

(67), 490(67), 491, 492(155)
Bungenberg de Jong, H. G., 484
Buchi, J., 181
Bunn, C. W., 532
Burch, W. J. N., 415(60), 416
Burchenal, J. H., 105, 106(285e), 135
Burgi, E., 95, 105(140), 130(140), 132(140)
Burke, D., 159

Burnett, W. T., Jr., 610, 611
Burrows, S., 246(39), 248, 255, 256
Busch, H., 216, 225(11)
Butenandt, A., 541, 542
Butler, G. C., 223, 227, 232, 258, 327, 336

(101), 347, 350(241), 354, 355, 356,

357, 411, 422, 434, 435(93), 441(93),

518, 527(58), 578, 585
Butler, J. A. V., 313, 316, 318(36, 49), 339,

340, 470, 471, 472(64), 485, 489, 511,

529, 551
Butler, K., 30, 57(160), 70(160), 78(160)
Bryson, J. L., 264
Byrne, W. L., 596, 597 (165a)

Cabib, E., 60, 190

Cahill, J. J., 13, 31(34), 32, 33, 70(34)

Cain, C. K., 131

Cairns, T. L., 130

Caldwell, P. C, 264

Caldwell, W. T., 101, 125

Calkins, D. G., 95, 136

Calvery, H. O., 4, 418

Calvin, M., 35, 187, 248

Campion, D. E., 546

Canellakis, E. L., 618

Cantero, A., 566

Cantoni, G. L., 94, 158

Cantor, S. M., 22

Canzanelli, A., 123, 339

Capell, L. T., 8, 824(17)
Caputto, R., 13, 98, 186, 594, 602
Cardini, C. E., 13, 98, 186, 190, 602
Carlin, R. B., 91, 92(88)
Carlisle, H., 574
Caro, G., 440, 576
Caron, E. L., 135
Carpenter, D. C, 253, 256, 257
Carpenter, F. H., 160, 424, 557
Carroll, W. R., 574

Carter, C. E., 49, 96, 161, 166(132), 167
(132), 172, 208, 209(73), 232, 234, 246,
250, 251(22), 252, 256, 257, 262, 321,
380, 386(25), 387, 392, 396, 397(25),
398(25), 399(25), 413, 419, 423(82),
424, 434, 441, 474, 475(85a), 566, 581,
588, 589(140), 597, 599, 601, 606, 608
Carter, R. O., 318
Caspersson, T., 4, 5, 6, 322, 374, 530, 539,

542(89), 543
Caster, W. O., 494
Catchpole, H. R., 544
Cavalier, J., 415

Cavalieri, L. F., 62, 106(292), 107, 109,
112, 113, 118(301), 120, 122(301), 125
(370), 128, 131(422), 132(370, 422),
133(438), 134(301), 135(301), 169(189,
190), 170, 217, 227, 269, 286, 289(4),
337, 339, 365, 410, 411, 412, 414, 424
(13), 425(14), 426(13), 460, 461, 479,
481, 482, 483, 495, 498, 515, 526, 544,
549,551,552,568, 616
Cecil, R., 337, 470, 471, 472(60), 473, 480

(60), 489, 491
Cerecedo, L. R., 121, 123, 304, 305, 619
Cerf, R., 485, 486, 487. 489(140, 142)
Ceriotti, A., 297
ChaikoflF, I. L., 329
Chalkley, H. W., 581, 599
Chang, N., 109, 112, 227, 414, 551, 552
Chang, P., 151

Chantrenne, H., 410, 431, 481, 563
Chapman, B., 105
Charalampous, F. C, 37, 602
Chargaff, E., 6, 7, 8(64), 14, 60, 89, 117,
124(330), 161, 192, 193(4), 195, 196,
197(4), 204(25), 205, 208(25), 209(25,
71), 218, 243, 245, 246, 247, 248, 251,
252, 253 (52a), 255, 256, 257(40, 52a),
258, 259(68), 260, 261, 262(17), 263,

AUTHOR INDEX

631

264(52a, 83), 287, 308, 309, 310(3),
311(9), 313(10), 314, 315, 316(10),
317(23), 318(10, 23), 319(10, 23), 320
(10), 322(4), 324(41), 326(10), 327, 328,
329(10, 57, 111), 330(104), 331 (7,23,37,
41), 332(23, 41, 139), 333(41, 114, 116,
142, 143), 334(41, 114, 116), 335, 336
(42, 111), 337(10, 23, 114, 161), 338
(90), 339(170), 340(4, 170), 341(170),
342(170), 343(170, 183), 344(104, 170),
345(134, 212), 346(57, 111, 114, 161,
224), 347(104, 111, 114, 134, 161, 170,
231), 348, 349(4, 114, 161, 167), 350(57,
111, 114, 141, 142, 167, 248), 351(4,
114, 161), 352(111), 354, 355, 356, 357,
358(104, 57, 111), 359, 360, 361(10,
160), 362(8, 10, 258, 259), 363(259),
364(259), 365 (258, 259, 260), 366 (170,
212, 227), 367(183, 221, 224, 225), 368
(4, 114), 369(4), 370(167), 374, 387,
390(44), 391(44), 392, 393(54), 394(4,
44), 395, 396(44), 397(44), 399(44, 54),
400(61), 401(67), 402(44), 403(44), 405,
406, 407(58, 71), 411, 412, 414, 418,
421, 424(27), 427(27), 434, 441, 443
(26), 444, 445(159), 466, 478, 491, 498,
499, 512, 516, 518(57), 519, 521, 557,
559, 561(12), 562, 563(12), 569, 577,
578(78), 579, 582, 584, 593, 594 (154b,
154c), 599
Charney, J., 20
Chase, Edith C, 61, 62(326), 79(326),

80(326)
Chase, M., 252, 253, 258(57)
Chavos, S., 327
Chayen, J., 549
Cheeseman, G., 112
Cheniae, G. M., 236, 237(55), 240(57)
Chepinoga, O. P., 320(80), 321
Cherbuliez, E., 13, 67(31, 32), 68(31), 97

158
Chernoff, L. H., 91, 92(85, 86, 87)
Chernomordikova, L. A., 330
Cheymol, J., 99
Choate, W. L., 574

Christensen, B. E., 130, 131, 133(435)
Christensen, L. R., 584
Christian, W., 13, 181, 183, 610
Christie, S. M. H., 184
Christman, A. A., 98
Christman, C. C, 34, 42

Ciotti, M. M., 512, 596, 608, 609

Clapp, S. H., 123

Clark, E. P., 19

Clark, L., 133

Clark, V. M., 98, 133, 149, 177, 180(211),

187, 462
Clarke, D. A., 105, 106(285d), 135
Clarke, M., 105
Claude, A,, 208, 374
Clements, G. C, 100, 118(214), 128(214),

188
Cleson, J. J., 610
Clews, C. J. B., 448, 449, 452
Cochran, W., 188, 189(287), 414, 448, 449,

451, 452, 457
Coghill, R. D., 89, 347
Cohen, P. P., 618

Cohen, S. S., 35, 41(185), 53, 61, 88(54),
89, 90, 92(54), 108(f), 109, 117(54),
130, 136, 188, 245, 249, 258(11), 259
(11), 261(11), 262(11), 290, 296, 298
(16), 316, 319(52), 332, 337, 344(151),
346(151), 355, 358, 359, 360, 370(151),
380, 383(27), 386(27), 389, 390, 392,
474, 475, 476, 500, 584
Cohn, E. J., 460
Cohn, Mildred, 37, 367, 416
Cohn, W. E., 20, 35, 36, 37(99), 38(99,
189, 190), 49, 57, 62, 63, 75(270), 76
(270), 79(334), 80(334), 90, 161, 162,
164(135), 165(135), 166, 167(130),
169(130), 170(130), 171, 172, 189, 200,
212, 215(3), 217, 218(13), 219(13, 19,
21), 220(19, 21), 221(13), 222(19), 223
(19), 224(19), 225(13, 19), 226(13, 19),
227(31), 228(19, 27, 31, 33, 34, 40, 41),
229(13, 34), 230(13, 40), 231(40, 43),
232(43), 233(13), 234(49), 236, 237(30),
238(60), 239(30, 61), 240(30), 241(63),
249, 256, 265, 347, 374, 413, 414. 418
(45), 419(45), 422, 423, 424(96), 425
(96), 426, 427(96), 429(96), 432,
434, 435, 436(124), 437(96), 441, 496,
497, 500, 501, 502(c), 506, 508(c),
512, 514, 520, 562, 564, 569(17), 577,
586, 587(136), 588(136), 589(136, 140)
Cole, Q. P., 106, 118(287), 128(287)
Colman, J., 106, 127(286), 131(286), 132

(286)
Colowick, S. P., 42, 241, 265, 512, 596, 604,
608, 609, 626

632

AUTHOR INDEX

Comer, C. L., 610, 611

Commoner, B., 103, 135(253), 412, 534,

542, 544
Compton, J., 42, 67(231), 150
Conrad, M., 126

Consden, R., 244, 245(7), 248(7), 261, 349
Conway, B. E., 339, 340, 470, 472(64),

485, 489, 529
Conway, E. J., 596, 597(165)
Cook, A. H., 134
Cooke, Kathleen, R., 12, 21(20), 31(20),

69(20), 141
Cooke, R., 596, 597(165)
Cooley, G., 32, 57(166), 70(166), 76(166)
Coons, A. H., 625
Cooper, E. J., 582, 583
Cooper, G. R., 383
Cooper, W. D., 206, 207
Corbet, A., 584, 624
Corby, N. S., 184, 187(250)
Corey, R. B., 440, 463, 464
Cori, C. F., 42, 600
Cori, Gerty T., 42, 600
Corran, H. S., 610, 611
Cortese, F., 150

Cosgrove, D. J., 467, 476, 479, 480(53)
Courtenay, T. A., 61, 189
Craig, D. P., 537, 538
Craig, J. A., 185
Craig, L. C, 110
Cramer, F., 244, 263(6)
Crampton, C. F., 310, 313(10), 316(10),
318(10), 319, 320, 326, 329(10), 337
(10), 360, 361, 362(258, 259), 363(259),
364(258), 365(258, 259), 405
Graver, L. F., 135
Creech, H. J., 625
Creeth, J. M., 338, 479, 480(110), 484

(110), 487, 488, 490(110), 491(110)
Crick, F. H. C, 365, 371 (262), 407, 440, 464,
465, 466, 467, 468, 470, 480, 488, 489,
525, 531, 541, 548
Crosbie, G. W., 204, 205(59), 208(59),
209(59), 260, 261, 262(89), 263, 269,
276(8), 277(8), 396, 397(63), 398(63),
400(63), 401(63), 402(63)
Crowfoot, D., 450

Cubiles, R., 423, 424, 432(100), 560, 562
(16), 569(16), 575, 589(16)

Cunningham, E., 92, 95, 125(135, 136),

159
Cutolo, E., 189

D

Dabrowska, W., 582
Dakin, H. D., 615

Daly, M. M., 193, 197(10), 258, 311, 329,
330(115), 350(115), 354, 355, 356, 357,
359
Daniels, M., 339
Daoust, R., 566
Dargeon, H. W., 135
Dassler, A., 116
Daus, L. L., 120, 187
Davidson, D., 121, 125
Davidson, J. N., 5, 6, 7(53), 8(65, 70), 204,
205(59), 208(59), 209(59), 260, 261
(89), 262(89), 263, 269, 276(7, 8), 277
(8), 278, 380, 387, 394(24), 396, 397
(63), 398, 400(63), 401(63, 69), 402
(63),418, 579, 615, 619(245)
Davies,H.G., 531,542
Davies, M. T., 33
Davis, Alice, R., 13, 96, 159
Davis, S., 316, 318(51)
Davison, P. F., 313, 318(36)
Davoll, J., 26, 37, 44, 51, 55, 57(282, 302),
61, 64, 75(302), 97, 118, 124, 125(399),
130, 148, 151, 153(47)
Dawson, F. S., 125
Day, E. D., 123(390)
Dedichen, G., 456
de Garilhe, Privat M., 584
DeGraeve, P., 96
Deimel, M., 279

Dekker, C. A., 14, 49, 62, 77(52), 90, 139,
156(3), 162, 170, 171(193), 172(145),
176(145), 218, 219(20), 221(20), 247,
366, 425, 426, 427(109), 440, 443, 506,
509(h), 559, 564, 565(62), 570(11),
588(11)
de la Blanchardiere, P., 576
de Lamirande, G., 566
Delcambe, L., 474, 475
Dd^zenne, C, 585
Delluva, A. M., 122
Dellweg, H., 107, 347
Deluca, H. A., 400
Dent, C. E., 261
Derby, J. H., 123(387)

AUTHOR INDEX

633

de Renzo, E. C, 610

Deriaz, R. E., 49, 50(6, 277), 53, 55(281),

56(277, 281), 59(291), 60(277, 281),

75(277, 281), 77(277, 281), 287, 288

(10), 289, 290(10)

De Robichon-Szulmajster, H., 379, 389,

390(23), 402(23, 49), 403(49)
Desclin, L., 4
Desoer, C, 345
Desreux, V., 437, 474
Deuel, H., 619

Deutsch, A., 189, 204, 205(58), 512
Deutsch, W., 601, 604
De Vries, W. H., 184
Dewan, J. G., 610, 611(223)
DeWayne, Roberts, 63
Dewey, V. C, 106, 107
Dhungat, S. B., 612
Dibben, H. E., 474, 559
Dickens, F., 41
Dickman, S. R., 572, 573(43a)
Diehl, H. W., 52, 63, 79(338), 80(338)
Dietrich, L. S., 599
Dille, K. J., 131, 133(435)
Dimler, R. J., 21, 47, 69(107), 70(253)
Dimroth, K., Ill, 205, 273, 284(14)
Dingwall, A., 496
Dion, H. W., 95, 136
Dische, Z., 53, 54, 121, 193, 287, 291, 294,
295(22), 297, 301, 302, 303, 334, 343,
394
Dittmer, K., 156
Dixon, J., 252, 253, 258(57)
Dixon, M., 600, 612
Dmochowski, A., 39, 168(182), 193
Doherty, D. G., 36, 38(190), 62, 63(334),
167, 233, 236, 240, 241(63), 414, 436,
589
Doman, N. G., 248
Donace, 208

Doniger, R., 14, 161, 195, 204(25), 205(25),
208(25), 209(25), 246, 248, 256, 257
(40), 259(68), 260(68), 263, 264, 329,
333(114, 116), 334(114, 116), 336, 337
(114), 346(114), 347(114), 349(114),
350(114), 351(114), 354, 368(114), 387,
390(44), 391(44), 394(44), 395(44),
396(44), 397(44), 399(44), 402(44),
403(44), 418, 512
Doree, C., 22
Dorfmiiller, G., 161, 194, 198, 418

Dorner, R. W., 403

Doty, P., 321, 337, 338, 440, 472, 473, 484

(71), 487(71), 489(67), 490(67, 71),

491, 492(155), 535
Dounce, A. L., 327, 328(102), 329(102),

335, 380, 386(26), 387, 389, 392, 393

(26), 474, 475 (85c), 522
Doyle, B., 339
Drake, N. L., 13, 20(33), 21(33), 43(33),

65(33), 66(33), 67(33), 68(33)
Drascher, L., 285
Drell, W., 100, 159
Drey wood, R., 54
Dribben, H. E., 374
Drury, H. F., 302
Dubos, R., 557, 559
Dubrovskaya, I. I., 330
Ducay, E. D., 329, 357
Duchateau-Bosson, G., 595
Duggan, E. L., 410, 423, 481
Dunham, E. K., 15(61), 94, 157
Dunn, D. B., 347, 370(238)
Dunn, J. S., 102
Dunn, M. S., 204, 205(58), 512
Dunstan, Sonia, 22
Durand, M. C., 559
Durrum, E. L., 274, 277
Dutta, S. K., 332, 391
Dutton, G. J., 187

Eakin, R. E., 614

Easterby, D. G., 18, 20(85), 22(85), 6{

(85)
Eden, M., 487, 489(144)
Edman, P., 97
Edmonds, M., 122
Edsall, J. T., 460
PJdstrom, J. E., 247, 253, 255, 261(67),

262, 284
Edward, J. T., 54
Edwards, L. J., 496
Edwards, P. C., 105
Edwards, S., 612
Egami, F., 8

Ehrenberg, J., 146, 173(38)
Ehrenberg, L., 60, 93, 188
Ehrlich, .A., 98, 622
Ehrlich, G., 302
Eidinoff, M. L., 131
Eiler, J. J., 410, 423, 459(37), 481, 567

634

AUTHOR INDEX

El Heweihi.Z., 21, 71(112), 72(112)
Elion, G. B., 95, 104, 105(140, 260), 127,

130(140,416,417), 132(140). 135
EUinghaus, J., 22
Elliott, A., 550, 574
Ellis, B., 13, 31(37), 32, 57(166), 70(166),

76(166)
Ellis, G. P., 30, 70(161), 71(161)
Ellison, R. R., 105, 106(285e), 135
Elmes, P. C, 318, 329, 544
Elmore, D. T., 90, 156, 176, 219, 340, 420,

506, 509(h), 520
El'piner, I. E., 340

Elson, D., 161, 195, 204(25), 205(25), 208
(25), 209(25, 71), 248, 256, 257(40),
259(68), 260(68), 329, 333(116), 334
(116), 370, 387, 390(44), 391(44),
394(44), 395(44), 396(44), 397(44),
399(44), 400(61), 401(67), 402(44),
403(44), 406, 407(71), 418, 512
Elvehjem, C. A., 587
Elwyn, D., 121, 122
Emanuel, C. F., 329
Embden, G., 160, 163(120), 164, 300
Emerson, Gladys A., 31, 61, 62(326), 79

(326), 80(326)
Emrich, W., 245

Endicott, M. M., 90, 92(83, 84), 83
Engeland, R., 89, 98(63)
English, J. P., 106, 118(287), 128(287)
Engstrom, A., 543
Erickson, R. O., 289
Erienmeyer, H., 106, 127(289), 131(289)
Errera, M., 339, 340
Estborn, B., 49, 219
Evans, E. F., 48, 72(255), 73(255)
Evans, H. J., 610, 611
Evans, J. S., 184
Evans, W. L., 20, 27, 65(102), 66(102),

67(102), 69(102), 73(102), 74(102)
Eyring, H., 110
Exer, B., 208, 209(72)

Fairley, J. L., 109, 117, 120, 124(331), 192,
193(8), 195(8), 199(22), 200(22), 201
(8, 22), 203(8), 205(8), 209, 390, 396,
402(50), 403(50), 499

Falco, E. A., 95, 104, 105(260), 127, 130
(417), 598

Falconer, R., 96, 97, 157, 159

Farnsworth, J., 574

Farrar, Kathleen R., 12, 21(21), 31(21),

44, 64, 69(21), 433
Fasman, G. D., 188, 189(287), 414
Feeney, R. J., 98, 159
Feigelson, P., 613
Felix, K., 311, 312, 316(26), 318(26, 27),

615, 619
Fellig, J., 136, 386, 572
Fells, E., 19, 30(89), 70(89)
Felsher, R. Z., 613
Felton, G. E., 49, 75(275)
Feulgen, R., 3, 4, 6(33), 7, 82, 196, 199,

299, 322, 341, 475, 576, 577(73)
Fields, M., 545, 546(149), 547(149), 549
Fink, K., 619, 621
Fink, R. M., 619, 621
Fischer, E., 12, 15, 18(14), 20(14), 48, 73
(14), 74(14), 82, 83, 85, 86, 87, 94(23),
96(9, 29), 99, 103, 108(j), 109, 116(23),
117, 124, 130, 131, 136, 149, 150, 155
(50, 52,57), 175, 274, 616
Fischer, F. G., 474, 578, 581
Fischer, H., 311, 312, 316(26), 318(26, 27)
Fischer, H. J., 100
Fischer, H. 0. L., 18, 48, 438
Fischer, I., 475(97), 476
Fisher, J., 582
Fiske, C. H., 124, 179, 202
FitzGerald, P. L., 60, 331, 332(139), 559,

561(12), 563(12)
Flaks, J. G., 614
Flammersfeld, H., 425
Flaschentrager, B., 311, 368
Flavin, M., 574
Fleming, M., 316, 318(48)
Fletcher, E., 248

Fletcher, H. G., Jr., 10, 15(62), 16, 21, 26
(111), 27(146), 45(111, 145, 146), 47(1,
111), 52, 63, 64, 65(111, 244), 66(111,
146, 244), 68(145, 146, 244), 69 (111,
146), 79(337, 338), 80(338, 340)
Fletcher, W. E., 374, 410, 433(10), 459,

474, 481, 482, 559
Flexser, L. A., 48, 74(256), 496
Flint, R. B., 120
Florkin, M., 595
Flynn, E. H., 135
Flynn, R. M., 184, 185(259)
Fodor, A., 136

AUTHOR INDEX

635

Folkers, K., 13, 31(34, 35), 32, 33, 61, 62,

70(34), 79(326), 80(326)
Fono, A., 338, 411, 416
Forrest, H. S., 183, 185(248)
Forsman, B., 246
Fosse, R., 96, 619
Foster, A. B., 29, 30(156), 33 .
Fox, C. L., Jr., 134, 613, 614(236h)
Fox, H. H., 48, 74(257)
Fox, J. J., 98, 108(d), 109, 112, 113, 115
(321), 116(319), 130(346), 131(346),
169(190), 170, 217, 227, 249, 414,
495(15), 496, 499, 500, 502(f, g), 503
(f, g), 504, 505, 506, 507, 508(f, g),
509(f, g, h), 515, 516, 517, 518, 551,
552
Fraenkel-Conrat, H. L., 93, 329, 357
Frajola, W. J., 318
France, W. G., 318
Franklin, R. E., 329, 467, 468(55, 57),

469, 541, 552
Frappez, G., 595
Fraser, M. J., 550, 551
Fraser, R. D. B., 536, 545, 546, 547, 548,

549, 550, 551
Frazer, S. C., 387
Freedman, R. L., 583
Frei, P., 17, 20(79), 65(79), 67(79)
Freise, R., 156

Freudenberg, W., 49, 75(275)
Frey, A., 532, 535
Frey-Wyssling, A., 535
Frick, G., 316, 318(47), 328, 339, 523, 524,

526, 553
Fried, M., 437, 438(131), 443
Frieden, A., 189
Friedkin, M., 36, 58, 63, 77(308), 602, 603

604, 605
Friedman, S., 120
Friedrich-Freska, H., 541, 542(129)
Friend, C., 105
Fries, N., 246
Fritzsche, H., 181
Frush, Harriet L., 16, 61, 283
Fryth, P. W., 15(59), 95, 160
Funk, C., 98, 622
Fuoss, R. M., 484
Furberg, S., 12, 14, 149, 452, 454, 455, 461,

462, 463, 464, 550
Furst, S. S., 105, 122(271)
Fuson, R. C, 92

Gabriel, S., 106, 127(286), 131(286), 132

(286)
Gaillard, B. D. E., 264

Gajdusek, D. C., 313, 318(33)

Gakhokidze, A. M., 49, 59(276), 60(276),
64, 75(276), 77(276), 78(276, 348)

Gandelman, B., 331, 350(141), 359, 405

Garner, R. H., 477, 480(53)

Gates, M., 132

Gaum^, J., 415

Gehrke, M., 17, 18(74), 20(74), 49(74),
59(74), 67(74), 74(74), 78(74)

Geib, N. C., 618

Geiduschek, E. P., 321, 535

Gelotte, G., 257

Genther, C. S., 104

Gerasimova, A. V., 339

Geren, W. D., 96, 126

Germann, H. C., 198, 418

Gerngross, O., 87

Gersh, I., 544

Getler, H., 126, 131

Getzendaner, M. E., 614

Ghuysen, J. M., 437, 475

Gibson, D. M., 626

Gilbert, L., 340, 470, 472(64), 526, 581

Gilbert, R., 27

Gillam, A. E., 22

Gilman, A., 103

Ginsburg, B., 583

Giri, K. V., 265
Glass,B.,564, 574, 589
Glazko, A. J., 120
Glick, D., 265, 543

Glock, Gertrude E., 33, 37, 41, 58, 59
Glueckauf, E., 216
Gold, N. I., 258, 352
Goldacre, R., 113, 114
Golder, R. H., 301
Goldring, L. S., 494
Goldsmith, N., 112
Goldstein, G., 313, 318(32)
Goldwasser, E., 321
Goodale, T. C., 35, 248
Goodman, I., 156
Goodman, L., 103
Goodman, M., 187
Goodwin, L. G., 104

Gordon, A. H., 244, 245(7), 248(7), 261,
271, 274, 349, 442, 578, 610, 611(223)

636

AUTHOR INDEX

Gordon, M., 596, 614

Gorin, P. A. J., 61

Gorup-Besanez, E., 95

Gosling, R. G., 329, 467, 468(55, 57), 469,

540, 541(127), 544(127), 552
Gots, J. S., 120
Govier, W. M., 184
Graff, S., 124
Gray, S. J., 613
Greco, A. E., 313, 432, 575, 584
Green, Charlotte, 14, 161, 195, 204(25),
205(25), 208(25), 209(25), 251, 257,
259(68), 260(68), 263, 264, 316, 327,
329(57), 330(104), 333(114, 116), 334
(114, 116), 335, 336, 337(114), 344
(104), 346(57, 114), 347 (104, 114),
349(114), 350(57, 114), 351(114), 354,
356, 357, 358(57, 104), 368(114), 387,
390(44), 391(44), 394(44), 395(44),
396(44), 397(44), 399(44), 402(44),
403(44), 418
Green, D. E., 610, 611 (222c, 223)
Greenberg, G. R., 595, 601, 614
Greenstein, J. P., 309, 310, 317, 320(81),
321, 339, 484, 530, 566, 580, 581, 583
(105), 584, 599
Gr^goire, J., 410, 581
Gregory, J. D., 184, 185(259), 481
Greider, M. H., 318
Greiner, W., 43, 67(234)
Gridgeman, N. T., 543
Griese, A., 13, 181
Griffiths, M., 123
Grimaux, M. E., 103
Grimsson, H., 518, 545
Grinnan, E. L., 386, 387(41), 392, 474,

475 (85b), 539
Grosblat,R.Sh., 320(80), 321
Groshens, Barbara P., 64
Groth, D. P., 40
Grylls, F. S. M., 246(39), 248, 255(39),

256
Grynberg, M. Z., 121
Gubernick, I., 618
Guerritore, D., 253
Guirard, B. M., 184

Gulland, J. M., 12, 14, 21(19, 20, 21), 31
(19, 20, 21), 69(19, 20, 21), 73(19), 96,
97, 141, 142, 143, 150(23, 2A), 157, 159,
164, 167(20), 174, 176(205), 194, 198,
308, 323, 328, 335, 338, 340, 341(185),

374, 409, 410, 411, 413(3), 420, 422,
433(10), 434, 436, 440(20), 455, 463,
466, 473, 474, 476, 477, 478(48), 479,
480(110), 481, 482(115), 484(48, 110),
489, 490, 491, 505, 506, 511, 522,
523(72), 557, 559, 586, 588(135), 589

Gunsalus, C. F., 600

Gurin, S., 54, 134, 298

Gustafson, T., 333, 395, 400(61)

Gugman, A. B., 96

Guttag, A., 48

Gyorgy, P., 596

H

Haas, V. A., 35, 248

Hadders, M., 93

Hagdahl, L., 250

Hahn, D. A., 82, 99(14), 123(14), 127,

128(14), 136
Hahn, L., 290, 300, 302, 303, 316, 318,

475(97), 476
Haiser, F., 142, 162, 197
Hale, W. J., 125
Hall, J. L., 318
Hall, O., 97
Hall, R. H., 98, 187
Hallanger, L. E., 100
Halverson, F., 495
Halverstadt, I. F., 103
Hambleton, J., 470
Hamer, D., 552

Hammarsten, E., 4, 14, 192, 312, 316(30),
319, 320(30), 322(30), 323, 328, 337
(30), 338(30), 476, 480(101), 529, 539

Hammarsten, G., 337

Hammarsten, O., 3, 11, 140, 198

Hammell, Myrtle L., 62, 198, 199, 200(49),
202(49), 204(39), 227, 414, 461, 507,
512(43)

Hammet, L. P., 496

Hamoir, G., 376, 381(14), 382. 400(29)

Hanes, C. S., 35, 245, 248

Hardegger, E., 21, 71(112), 72(112)

Harden, A., 181

Hardy, S. M., 103

Hare, G. H., 494

Harkins, H. H., 121

Hart, M. J., 450

Harris, I. F., 3, 89

Harris, R. J. C., 62, 169(191), 170, 219,
545, 551

AUTHOR INDEX

637

Harris, S. A., 34, 39(179), 40, 62(179, 213,

214), 63, 67(214), 68(213, 214), 74

(219), 163, 165, 166, 167, 168(167),

193, 413, 414
Harris, T. H., Jr., 131
Harrison, J. S., 246(39), 248, 255(39), 256
Hartree, E. F., 615, 616, 618, 619(243)
Hartshorne, N. H., 532, 539(101)
Hartwig, St., 537, 541, 542(129)
Haskins, F. A., 596
Hassid, W. Z., 34, 42(177)
Hastings, R., 123
Hauge, S. M., 100
Hawes, R. C, 494
Hawkins, J. D., 33
Haworth, W. N., 27
Hayaishi, 0., 622, 623
Hayes, P. M., 187

Haynes, L. J., 38, 166, 174(170), 181, 418
Hazel, G. R., 103
Heath, J. C, 120
Hecht, L., 424, 560, 562(16), 569(16),

589(16)
Hedstrom, H., 60, 93, 188
Heidelberger, C, 101, 125(235), 345
Heilbron, I., 134

Heinrich, M. R., 107, 120, 122(356)
Helferich, B., 20, 68(104), 124, 150, 155

(50)
Heller, L., 337
Hempelman, I. H., 624
Henderson, R. B., 619, 621
Hendricks, S.B., 454,455
Henkel, K., 157
Hennig, I., 119
Henstell, H. H., 583
Heppel, L. A., 432, 437, 441, 564, 566, 575,

588(67), 581, 592, 602, 603, 604, 606,

610
Herbert, G., 584
H^rissey, H.,99
Hermann, V. Sz., 598
Heme, R., 311
Herriott. R. M., 574
Hershey, A. D., 252, 253, 258
Hess, E. L., 289
Hesseltine, C. W., 95
Hewitt, R. I., 95
Heyl, Dorothea, 61, 62(326), 79(326),

80(326)
Heyroth, F. F., 495

Heyther, P., 610

Hilbert, G. E., 155

Hilfinger, M. F., 611

Hilmoe, R. J., 432, 441, 566, 575, 588(67),
591, 592, 602, 603, 604, 606

Himes, H. W., 92

Himsworth, H. P., 103

Hines, D. C., 103

Hinman, J. W., 135

Hirs, C. H. W., 573, 575(53b, 53c)

Hirsch, P., 458

Hirst, E. L., 15(64), 22, 27, 263

Hirt, R. C., 495

His, W., Jr., 2, 109, 113

Hitchings, G. H., 95, 104, 105(140, 260),
106(285d, 292), 124, 127, 128, 130(140,
416, 417), 132(140), 135, 598

Hobday, G. I., 174

Hochberg, M., 92

Hochreuter, R., 64, 80(342)

Hockett, R. C., 18,29,48,52

Hodes, M. E., 196, 251, 258, 287, 316, 326,
329(57), 335, 336, 339(170), 340(170),
341(170), 342(170), 343(170), 344(170),
347(170), 350(57), 354, 357, 358(57),
366(170), 368, 443, 521

Hoepfner, Eva, 44, 66(236)

Hoff-J0rgensen, E., 604

Hoffer, M., 132

Hoffmann-Ostenhof, O., 557

Hogberg, B., 475(97)

Hogeboom, G. H., 208, 584, 624, 625

Holden, C., 289

Holiday, E. R., 13, 31(36, 37), 107, 142,
143, 150(23), 164, 200, 246, 261, 275,
277(20), 314, 349, 393, 412, 450, 488,
495, 506, 511(41), 530, 542

Hollaender, A., 339

Holland, A., 153

Holly, F. W., 13, 31(34), 32, 33, 70(34)

Holmberg, C. G., 615, 618

Holt, N. B., 283

Honeyman, J., 29, 30, 47, 70(161), 71 (161)

Hood, Dorothy, B., 54, 298

Hopkins, F. G., 95, 96(147), 609

Hoppe-Seyler, F., 1,2,322

Horbaczewski, J., 609

Horecker, B. L., 35, 41 (188), 67 (188b), 610

Horn, P., 487, 492

Horner, L., 245

Horrocks, R. H., 21

638

AUTHOR INDEX

Horstadius, S., 358

Hoste, J., 300

Hotchkiss, R. D., 6, 89, 192, 193(5), 243,

245, 252, 261, 331, 349, 498, 499, 504,

507
Hough, L., 18, 20(85), 21, 22(85), 24, 52,

59, 68(85)
Houlahan, M. B., 100
Howard, G. A., 26, 30, 37(144), 43, 44, 45

(159), 66(144, 159), 70(159), 150, 151

(54), 153, 156, 506
Hoyer, M. L., 320(81), 321
Hudson, C. S., 17, 18, 20(80), 21, 24, 25,

26(111), 27(146), 29, 45(111, 145, 146),

47(111), 48, 52, 56, 64, 65(111, 138),

66(111, 146), 68(138, 145, 146), 69(106,

111, 146), 74(345)
Huebner, C. F., 47, 70(253)
Huff, J. W., 100
Hug, 0., 318
Hughes, H. K., 494
Huiskamp, W., 311, 312(16)
Hultin, T., 311

Humoller, F. C, 17, 20(76), 73(76)
Hunter, G., 123(388), 124, 304, 305
Hunter, L., 115
Hurd, M., 316, 318(50)
Hurlbert, R. B., 101, 125(235), 189, 209,

216, 225(11), 232, 235,608, 625
Hurst, R. O., 223, 227, 232(29), 258, 347,

350(241), 354, 355, 356, 357, 358, 422,

434, 435(93), 441(93), 585
Hutchings, B. L., 15(59), 95, 100, 160, 610
Hutchinson, S. A., 95, 125(135), 159
Hutchinson, W. C., 387
Hutchison, O. S., 105

I

Imanaga, Y., 53, 75(289)

Ing, H. R., 104

Irvin, E. M., 337

Irvin, J. L., 337

Isay, O., 94 127(113), 131(113), 132(113)

Isbell, H. S., 16, 21, 22, 23(129), 61, 65

(129), 73(129), 283
Isherwood, F. A., 35, 245, 248, 264
Iso, K., 474

J

Jackson, E. L., 25, 65(138), 68(138), 422
Jackson, E. M., 413, 434, 586, 588(136),
589

Jacob, A., 93

Jacobs, G., 584, 624(119)

Jacobs, W. A., 3, 7(14), 11, 12, 20(12),
39(11, 12), 65(12), 67(12, 103), 68(12),
138, 140, 142(16), 156(8), 161, 162(7,
8, 9), 163(9), 168, 169, 170(186),
172(140), 194, 197, 198, 366, 440

Jacobson, P., 83, 96(20), 136

Jacobson, W. E., 105

Jacobsson, F., 542

Jacoby, M., 609

Jaenicke, L., 205, 236, 273, 283, 284(14)

James, D. W. F., 313, 318(36), 340, 471

Jamieson, G. A., 158, 188

Jansen, E. F., 155

Jeanloz, R. W., 10, 15(62), 21, 26, 45(111,
145, 146), 46, 47(1, 111), 65(111), 66
(111), 68(145), 69(111)

Jeener, R., 136

Jelley, E. E., 532

Jenkins, R. J., 488, 515, 518(54), 523(53,
54), 526(54), 527, 529(53)

Jenrette, W. V., 317, 339, 484, 580

Jermyn, M. A., 264

Jochman, I., 481

Johns, C. O., 99, 101, 129, 130(226)

Johns, H. E., 618

Johnson, A. W., 13, 31(38), 61, 79(327),
80(327)

Johnson, E. A., 13, 31(37), 107, 109, 112,
113, 200, 246, 261, 275, 277(20), 314,
349, 393, 412, 450, 488, 498, 499, 500,
501, 502(d, e), 503(e), 508(d, e), 509,
510, 511, 512, 513(b), 519

Johnson, M. D., 209

Johnson, T. B., 82, 87, 88, 89(52), 90, 91,
92(83, 84, 85, 86, 87), 93(92), 99, 100,
101, 108(c, e), 109, 117(66), 119, 120,
121,122(18), 123(14, 387), 124, 126,
127, 128(14), 130(226), 131, 136, 155,
347

Joklik, W. K., 626

Jones, A. S., 332, 334, 391

Jones, J. K. N., 18, 20(85), 21, 22(85), 24,
59, 61, 68(85), 263

Jones, R. N., 543, 625

Jones, W., 3, 4, 7, 82, 86, 88(55), 89,
96(11), 121(40), 165, 192, 193(9b),
194(9b, 14), 198, 418, 424, 558, 559

Jones, W. H., 62

Jope, E. M., 542

AUTHOR INDEX

639

Jordan, D. O., 6, 8(68), 12, 84, 107, 113,
136, 149, 249, 308, 316, 318(48), 323,
328(98), 329, 335, 336, 338, 340, 341
(185), 374, 410, 411, 433(10), 436, 440
(20), 455, 460(25), 463, 466(48), 467,
473, 474, 476(48), 477, 478(48), 479
(48), 480(48, 110), 481, 482(48, 115),
484(48, 110), 485, 487, 488, 489(48),
490(48, 110), 491(110), 495, 507, 514,
522, 523(72), 525, 539, 541, 550, 559

Jorpes, E., 4, 38, 167, 169(178), 194, 474,
476, 481

Josepovits, G., 598

Jung, H., 49, 77(280)

Jungner, G., 334, 473(74, 76), 474, 581

Jungner, I., 473(74, 75, 76), 581

Kaczka, E. A., 62
Kagan, Z. S., 248
Kahler, H., 337, 470, 471, 472, 487, 489

(144)
Kalckar, H. M., 33, 34(174), 36(174), 37

(194), 58, 133, 189, 505, 506, 512, 597,

598(167), 599, 601, 602, 603(194), 604,

612, 613, 618, 626
Kaleita, E., 610
Kalyankar, G. D., 265
Kaplan, N. O., 184, 185, 186, 241, 265, 429,

435, 512, 592, 593, 596, 604, 608, 609
Karabinos, J. V., 44, 66(239), 73(239)
Karger, B., 608

Karnofsky, D. A., 105, 106(285e), 135
Karlsson, J. L., 122
Karrer, P., 13, 20(79), 40(41), 64, 65(79),

67(79), 80(342), 181, 247
Kasha, M., 538
Katchalsky, A., 320
Kates, M., 415, 416(56)
Katz, S., 73
Kaupp, B. F., 96
Kauzmann, W., 419
Kavanagh, F., 93
Kay, E. R. M., 327, 328, 329, 335, 380, 386

(26), 387, 389, 392, 393(26), 474, 475

(85c), 522
Kay, G. A., 103
Kearny, E. B., 182
Keenan, G. T., 22
Keilin, D., 612, 615, 616, 618, 619(243)
Kennedy, R. P., 165, 198

Kenner, G. W., 10, 12(3), 30, 43, 44(232),
45(159), 46, 65(232, 245), 66(159, 232,
245), 68(232), 70(159). 71(245), 72
(232, 245), 94, 153, 154, 184, 187(250)

Kent, P. W., 22, 47, 49, 55(251), 75(251),
142, 264

Kerin, L., 332, 333(142), 350(142)
Kerr, S. E., 124, 192, 193(6), 194, 199, 200
(6), 201(6), 217, 375, 387, 392, 396,
399, 410, 411, 412, 424(13), 426(13),
481, 482(122), 498, 512, 544, 549

Ketelle, B., 212, 215(3)

Khouvine, Y., 379, 389(23), 391, 402(23,
49), 403(49, 53), 410, 481

Khym, J.X.,36,38(189, 190), 49, 57(270),
62, 63(334), 75(270), 76(270), 79(334),
80(334), 162, 167, 172, 212, 215(3),
223, 226, 228(27), 236, 237(30, 55),
238(60), 239(30, 61), 240(30, 57),
241(63), 347, 414, 441

Kidder, G. W., 106, 107

Kiesel, A., 4, 330

Kime, H. B., 125

King, E. J., 334

Kirby, H., 244

Kircher, W., 90

Kirk, P. L., 260

Kirkwood, J. G., 455, 460

Kirshner, A., 245

Kjeldgaard, N. O., 612, 613

Klein, J. R., 15(65), 16

Klein, W., 49, 75(267), 121, 140, 142(5),
162, 171(147, 148, 149), 225, 422, 441,
588, 589, 601, 604

Klemperer, F., 96

Klemperer, F. W., 616, 618

Klenovv, Hans, 37, 41(201), 612, 613

Klimek, R., 35, 164

Klingensmith, C. W., 20, 27, 65(102), 66
(102), 67(102), 69(102), 73(102), 74
(102)

Knight, C. A., 206, 207, 377, 383(17, 18),
389, 390(19), 394, 403(19), 404(19, 72),
405

Knoll, J. E., 131

Knopf, M., 168

Knopf, R. B., 572, 573(43a)

Koch, F. C., 618

Kocher, V., 247

Koenig, V. L., 317, 339

Koerner, J. F., 49, 162, 223, 225, 233, 235

640

AUTHOR INDEX

(28), 347, 350(240), 354, 412, 441(30),

442, 520, 578, 585, 588(134)
Kohn, H. I., 15(65)
Kolthoff, I. M., 120
Komatsy, S., 64, 74(345)
Konkova, V. A., 61
Koritz, H. G., 120
Korn, E. D., 37, 602
Romberg, A., 13, 36, 181, 182, 435, 591,

594,601, 603, 604(203), 607(189), 620,

621(274c),622, 623, 626
Kosolapoff, G. M., 415(61), 416, 420(61),

483
Kossel, A., 2, 3, 11, 84, 85, 86, 87, 89(51),

118, 120, 124(27), 196,311, 341
Kothnig, M., 20, 22(96), 43(96), 46(96),

65(96), 67(96), 475(95, 96), 481
Kowlessar, O. D., 624
Kozloflf, L. M., 625
Krahl, M. E., 113
Kream, J., 247, 599
Krejci, L. E., 470

Krekels, A., 312, 316(26), 318(26, 27)
Kreusler, V., 99
Kreuzer, L., 311
Krishnaswamy, P. R., 265
Krueckel, B., 120
Kruger, M., 94, 118, 124(335)
Krupka, G., 95
Kuhn, L. P., 24
Kuhn, R., 17, 29, 33, 65(78), 68(78), 71

(154, 170), 157, 183
Kumler, W. D., 459(37)
Kunitz, M., 316, 339, 385, 412, 423, 441
(29), 518, 519, 557, 559, 560, 567, 569,

572, 573, 574, 577, 580, 581(3a), 583
Kupke, D. W., 316, 318(50)
Kurnick, N. B., 544, 579, 580, 581(88),

583
Kutscher, F., 89, 98(63)
Kutscher, W., 597
Kwan, U. S., 484

Laasko, J. W., 100

Lacher, J. R., 546

Lackman, D. B., 323

Ladenburg, K., 45, 47(240), 72(240), 73

(240)
Laetren, H., 625
La Forge, F. B., 12, 29, 141, 156(11)

Lagerkvist, V., 119, 120(345)
Laland, S. G., 30, 57(160), 70(160), 78
(160), 253, 259(55), 329, 330(119),
340, 345(119), 347(119), 354, 355, 356
357, 358, 359, 552, 579
Lamanna, C, 572
Lamb, CM., 316, 318(46)
Lampen, J. O., 14, 58, 63, 76(306), 106,
118(287), 128(287), 291, 292(21), 504,
507, 603, 605, 607, 622
Land, K., 624
Landauer, P. D., 132
Landstrom-Hyden, H., 5
Lang, K., 584
Larkins, L., 317
Larsen, B., 37, 41(201)
Laser, R., 601, 604
Laskowski, M., 316, 577, 578, 579, 581

(76), 582, 583, 584, 624(119)
Laufer, L., 20
Lavigne, J. B., 32
Law, L. W., 107
Lawley, P. D., 488, 515, 523(56) 524, 525

(56), 526, 527(56), 529
Lazarow, A., 103
Lechinsky, W., 11
Lederer, E., 244
Lederer, M., 244
Ledoux, L., 572, 573

Lee, J., 19, 29, 30(89), 45(88, 90), 65(86,
88, 90), 70(86, 87, 89 157, 158), 71
(87, 157, 158)
Lee, S. L., 583
Lee, W. A., 334, 411, 412, 440(22), 476,

477, 478, 479, 522, 523(73) 552, 579
Lehmann, G., 475(95, 96)
Lehmann-Echternacht, H., 474, 578, 581
Lehnartz, E., 311, 368
Leidy, Grace, 60, 322, 323(91), 328(91),
331(91), 332(139), 338(91), 559, 561
(12), 563(12)
Leloir, L. F., 10, 13, 33, 60, 98, 165, 186,

187(163), 190, 247, 264, 602
Lemberg, R., 600
Lenhartz, E., 300
Lenormant, H., 553
Leonard, F., 19, 45(88), 65(88)
Leone, E., 253, 618
Leone, L. A., 135
LePage, G. A.,39, 40, 179
Leray, J., 487, 492(143)

AUTHOR INDEX

641

Leslie, I., 192, 285, 328

Le Strange, R., 244, 263(4)

Leuthardt, F., 208, 209(72)

Leuthardt, F. M., 599

Levenbook. L., 252, 255, 257(65), 258

Levene, P. A., 3, 4, 7(14, 16), 11, 12, 14,
19, 20(12), 21, 24, 25, 26, 27, 29, 34,
38, 39(11, 12, 179, 218), 40(175), 42
(136, 178, 181, 210, 220), 43(110), 46
(175, 220), 47(101), 48, 49(264, 266),
57(264), 60(266), 62(179, 213, 214),
63, 65(12, 110, 220), 66(101, 110, 220),
67(12, 101, 103, 136, 175, 210, 214,
220, 331), 68(12, 110, 136, 175, 213, 214,
220, 247), 69(110, 220), 73(136, 175,
220), 74(219), 75(265, 266), 76(264),
77(266), 78(266), 87, 89, 96(13), 99,
112, 113, 119(314), 122(13), 124, 136,
138, 140, 141(4), 142(4), 143(16, 21),
144, 145, 146(27, 36), 147(17, 36), 156
(8, 11), 157(17), 159, 161, 162(7, 8, 9),
163(9), 165(122), 166, 168(93, 167),
168(167, 182), 169(93, 178), 170(186,
188), 172(140, 141, 144), 173(17, 36),
174(45), 192, 193, 194, 195(16c), 196
(16d, 16e), 197, 198(16, 17), 217, 222,
226, 227, 268, 269(1, 2, 3), 270, 308,
322, 341(1), 348, 366, 386, 409, 410,
413, 414, 418, 440, 455, 457(21, 22, 23),
459(21, 22), 475, 476, 481, 557, 559,
576, 588, 589(74), 600, 601

Levine, C, 263, 346

Levy, L. W., 62, 198, 199, 200(49), 202(49),
204(39), 226, 227, 414, 461, 507, 512(43)

Lewis, W. L., 17

Li, C. F., 412

Liddle, L. M., 103, 127(245), 128(245)

Lieberman, L, 601, 607(189), 620, 621
(274c), 626

Liebig, J., 82, 102

Liebig, R., 612

Lilienfeld, L., 311, 312(15)

Linderstr0m-Lang, K., 431, 481, 563

Lingane, J. J., 120

Link, K. P., 21, 47, 69(107), 70(253)

Linstead, R. P., 438

Lipkin, D., 367, 416, 542, 544

Lipmann, F., 13, 184, 185, 186

Lippert, W., 318

Lipshitz, Rakoma, 14, 251, 252, 253(52a)
257 (52a), 264(52a), 310, 313(10), 316

(10), 318(10), 319(10), 320(10), 326(10),
327, 328, 329(10, 57, 111), 330(104),
335, 336(111), 337(10), 341, 344(104),
346(57, 111), 347(104, 111), 350(57, 111),
352, 355, 356, 357, 358(57, 104, 111),
360, 361 (10), 362(10, 258, 259), 363(259),
364(258, 259), 365(258, 259, 260), 367
(221), 369, 405, 444, 445(159)

Lipton, S. H., 189

Little, J. A., 223, 233(29), 258, 411, 422,
434, 435(93), 441(93), 578

Litzinger, A., 92, 93(92)

Lloyd, B. J., Jr., 337

Lock, M. v., 43, 46, 433

Loebenstein, A., 113

Lofgren, N., 60, 93, 188, 252, 253

Logan, J. E., 529

Lohmann, K., 179, 438, 459

Lohmar, R., 47, 70(253)

Loldina, G. L, 321

Lomakka, G., 542

London, E. S., 3, 48, 49(264), 57(264), 76
(264), 140, 141(4), 142(4)

Long, A. G., 247

Longsworth, L. G., 381

Loofbourow, J. R., 450, 495, 501, 514, 545,
546(34)

Lorch, I. J., 358

Loring, H. S., 62, 109, 117, 124(331), 160,
161, 169(133), 192, 193(8), 195(8),
198, 199(22), 200(22, 49), 201(8, 22,
23), 202(23, 49), 203(8), 204(39), 205
(8), 206, 207, 226, 227, 259, 374, 383,
390, 396, 402(50), 403(50), 414, 418,
424, 461, 474, 499, 500, 507, 512, 557

Lothian, G. F., 543

Lowenfeld, R., 157

Lowrey, J. A., 95

Lowry, O. H., 262

Lowy, B. A., 61, 120, 124, 125(399), 130,
151

Liibavin, N., 2

Luck, J. M., 316,318(50)

Luning, B., 60, 93, 188

Luthy, N. G., 161, 169(133), 200, 226, 414

Lynch, V. H., 187

Lynen, F., 186

Lyons, L. E., 537, 538

Lythgoe, B., 14, 26, 30, 37(144), 43(144),
44(144), 45(159), 51, 55, 57(282, 302),
66(144, 159), 70(159), 75(302), 94, 116,

642

AUTHOR INDEX

117(328), 118, 119, 123(339), 131(339),
132(436), 134, 136, 147, 148, 150, 151
(54), 152, 153, 154(46, 47), 156, 157,
179, 424

M

Ma, T. S., 202

McCarthy, M., 566, 577, 580, 581(91), 583
584

McCarty, M., 319, 322(74), 331(74), 332
(138), 333(74), 339

McCasland, G. E., 91, 92(88, 89)

McCrudden, F. H., 609

MacDonald, D. L., 394

McDonald, M. R., 572, 573

McElroy, W. D., 564, 574, 589, 596

McEvoy-Bowe, E., 288

MacFadyen, D. A., 566

McFarland, D. F., 87, 127

McGrath, D. I., 559, 570(10)

Macheboeuf, M., 21

Mclndoe, W. M., 209, 400, 401(69)

Maclnnes, D. A., 381

McKeever, C. H., 92

McKibbins, J. M., 611

Mackler, B., 610, 611

McLean, A. C, 26, 47, 69(254), 154, 506

MacW. Lemon, J., 95

MacLeod, C. M., 319, 322(74), 331(74),
333(74)

McLetchie, N. G. B., 102

McNair Scott, D. B., 35, 41(185)

McNeil, D., 152

McOmie, J. F. W., 108(b), 112, 127, 128
(415), 130(424), 495

MacNutt, W. S., 94, 252, 253, 504, 505

McT. Ploeser, J., 195, 201(23), 202(23),
262

Macrae, T. F., 142, 150(23), 167(20), 198,
506, 511(41)

McRorie, R. A., 185

Maculla, A., 124

Magasanik, B., 161, 195, 204(25), 205(25),
206, 208(25), 209(25), 246, 248, 256,
257, 259(68), 260(68), 314, 329, 333
(116), 334(116), 336(42), 339, 349, 387,
390(44), 391(44), 392(54), 394(44),
395(44), 396(44), 397(44), 399(44, 54),
402(44), 403(44), 412, 418, 421, 424
(27), 427(27), 434, 512, 519, 562, 569

Magnus, 85

Magrath, D. L, 38, 167, 169(173), 171
(173), 175(197), 188, 189(287), 414, 417
(54), 425, 434

Mahler, H. B., 610, 611 (222c)

Makino, K., 49, 60(279), 75(279), 94, 98,
147, 157, 158(105), 475(94), 481

Mallette, M. F., 131,572

Malmstrom, B. G., 543

Malpress, F. H., 248

Mamalis, P., 13, 31(37), 32, 33, 57(166),
70(166), 76(166)

Mamur, J., 40

Mandel, J. A., 4, 15(61), 94, 157, 162

Mann, T., 330, 591

Manna, L., 100

Mannell, W. A., 529

Manson, L. A., 58, 76(306), 291, 292(21),
504, 603, 605

Manson, W., 95, 125(135), 159

Marcet, 98

Markham, R., 162, 170(137), 171(137),
192, 193(7), 195(7), 204(7), 205(7),
206, 207(61), 244, 246, 247, 249(26),
250, 251(26), 252, 253(26), 256, 259,

260, 270, 271(10), 272(10), 273(10),
274, 275(17), 277(9, 17, 18, 19), 279
(9, 10, 17), 280(9, 10, 25), 281(10),
282, 284, 310, 349, 374, 378, 380(21),
382(21), 386(21), 389, 394, 395, 403,
404(48), 418, 419, 425, 426, 427(83,
113), 429(113), 432, 437(113), 438, 439,
442, 443(152), 445(114), 499, 511, 530,
562, 564, 565, 566, 569(18), 577, 578

Marko, A. M., 227, 327, 336(101), 347,
350(241), 354, 355, 356, 357, 518, 527
(58)

Marrian, D. H., 61, 189, 199, 260

Marsh, W. H., 118

Marshak, A., 90, 117(77), 195, 197(24),
204(24), 205(24), 208, 209(70), 218,
246, 251, 258, 259(53), 332, 333, 348
(153), 354, 400, 402

Marshall, E. K., Jr., 614

Marshall, J. M., 625

Marshall, J. R., 112, 127(312), 128(312),
450, 495, 525(7)

Martin, A. J. P., 244, 245(7), 248(7), 249,

261, 349, 573
Martini, A., 133, 198
Masley, P. M., 185
Massart, L., 300, 301, 572

AUTHOR INDEX

G43

Mathias, A. P., 189

Matthews, R. E. F., 106(293), 107, 247,

257(30), 277, 284, 439, 530
Matsushima, Y., 53, 75(289)
Matty, S. M., 467, 477, 480(53)
Maurer, W., 279
Maver, M. E., 313, 432, 575, 584
Mayer, D. T., 330
Mayer, S. W., 214, 223(8)
Mazia, D., 313, 316, 318(53)
Meagher, W. R., 34, 42(177)
Medicus, L., 82, 84, 85
Medigreceanu, F., 557, 600, 601
Meisel, E., 591

Meisenheimer, J., 49, 77(280), 109
Mejbaum, W., 301
Mellors, R. C, 534, 549
Melnick, D., 92
Mendel, L. B., 615, 619
Mendelson, W., 95
Mercer, F., 103, 135(253)
Merck, E., 15(63)
Merriam, H. F., 87, 90(45), 108(g)
Merriefield, R. B., 196, 233, 234(50), 421,

428, 429, 520, 562(21), 563, 569(21),

570(21)
Merritt, A. J., 98, 622
Metzger, A., 126
Meyer, V., 83, 96(20), 136
Meyer-Brunot, H. G., Ill
Meyerhof, O., 42, 438, 459
Michael, S. R., 583
Michaelis, L., 545
Michalski, J. J., 177, 180(211)
Michelson, A. M., 12, 34, 40, 62, 73(180),

100, 159, 162, 163, 172(145), 174(158),

175(197), 176(145), 180, 189, 194, 366,

422, 440, 441, 443
Michl, H., 246

Miescher, F., 1, 2, 3, 308, 311, 312, 320, 322
Mikeska, L. A., 3, 7(16), 48, 49(266), 60

(266), 75(266), 77(266), 78(266), 142
Miller, C. S. 134
Miller, E. E., 301
Miller, G. L., 301, 383
Miller, G. W., 61, 79(327), 80(327)
Miller, W. H., 103
Miller, Z. B., 625
Mills, J. A., 13, 31(38), 61, 79(327), 80

(327)
Milstein, S. W., 566, 572(33)

Minkowski, A., 96

Minsaas, J., 25, 68(137)

Minton, S. A., Jr., 105

Mirsky, A. E., 6, 7, 8(78), 193, 197(10),

258, 287, 311, 312, 316(24, 25, 31), 317

(25), 318(24, 25), 319(25), 323, 329,

330(115), 350(115), 354, 355, 356, 357,

367(24), 375, 388, 579, 580, 624(87a),

625
Misani, F., 14, 263, 264, 329, 333(114), 334

(114), 336, 337(114), 346(114), 347

(114), 349(114), 350(114), 351(114),

354, 368(114)
Mitchell, H. K., 100, 101, 159, 596
Mitchell, J. H., Jr., 107
Mitchell, T. J., 264
Mitchison, J. M., 538
Miyaji, T., 484, 581, 583(105)
Miyake, K., 95
Modena, G., 500, 538(29)
Mohr, R., 312, 316(26), 318(26, 27)
Moldave, K., 345
Monroe, R. A., 610, 611
Montgomery, Edma, M., 24
Montreuil, J., 7, 8(77), 256, 257, 260, 418
Moore, A. E., 105
Moore, A. M., 261
Moore, D. H., 314
Moore, L. V., 289
Moore, Marjorie U., 61, 62(326), 79(326),

80(326)
Moore, S., 47, 70(253), 212, 215, 219, 234,

573, 575 (53c)
Moos, A., 125
Mora, T. P., 22
Morel, H., 585
Morell, D. B., 610
Morell, S. A., 189
Morgan, D. M., 248
Morgan, E. J., 609
Mori, T., 3, 7(16), 48, 49(266), 60(266),

75(265, 266), 77(266), 78(266), 94, 141,

142, 163
Morton, A. A., 83
Mosher, H. S., 117, 118(329)
Mosher, W. A., 123(390), 124, 386, 392,

474, 475(85b), 539
Mueller, G. C., 40
Muggleton, P. W., 576, 585
Mukherjee, S., 47, 51(250)
Muller, G., 144, 341

644

AUTHOR INDEX

Mliller, H. R., 247

Miiller, R., 101, 537

Munch-Petersen, A., 189

Munk, K., 271, 426

Munoz, C, 302

Murphy, (M.) L., 105,.106(285e), 135

Myers, V. C, 619

Myrback, K., 181, 474

N

Najjar, V. A., 37

Nason, A., 610, 611

Needham, D. M., 94

Nef, J. U., 52

Neff, R. J., 624

Neker, H. T., 17

Nemchinskaya, V. L., 319

Ness, R. K., 26, 27(146), 45, 63, 64, 65

(244), 66(146, 244), 68(244), 69(146),

79(337, 338), 80(338, 340)
Neubauer, Z., 124
Neuberger, A., 123, 615, 617
Neumann, A., 2, 3, 86, 87, 196, 341
Neurath, H., 310, 311, 383, 535
Newbold, G. T., 26, 154, 506
Newcomer, H. S., 103
Newton, Eleanor, B., 13, 96, 159
Nicolaier, A., 96, 595
Nicholson, D., 546
Nigrelli, R. F., 104
Noggle, G. R., 236
Norris, E. R., 598, 599, 612
Norris, L. T., 187
Northey, E. H., 104
Northrop, J. H., 574
Novelli, G. D., 13, 184, 185(259), 186
Nunez, G., 21
Nye, J. F., 100, 101
Nystrom, C., 5

O

Ochoa, S., 15(66)

Odake, S., 94

Odenius, R., 4

Ogston, A. G., 112, 113, 217, 337, 456, 457,

458, 459(28), 470, 471, 472(60), 473

(63), 480(60), 489, 491
Ogur,M., 289, 518, 529,530
Ohlmeyer, P., 321

Ohta, K., 49, 60(278, 279), 75(278, 279)
Ollendorf, G., 12

Openshaw, H. T., 173

Orbach, H. K., 327

Orr, S. F. D., 62, 169(191), 170, 545, 551
(152)

Osborne, T. B., 3, 89

Oser, B. L., 92

Oster, G., 473, 518, 535, 542(106), 545,
552

Ostern, P., 179, 594, 601

Ostmann, P., 150

0th, A., 487

Ottenstein, B., 48, 162, 172(142), 341, 366
(218), 440(144), 441

Ottolenghi, L., 290, 300(18)

Overend, W. G., 10, 29, 30(156), 33, 47, 49,
50(6, 277), 51(252), 53(4), 54, 55(252,
277, 281), 56(277, 281), 57(160), 59,
60(252, 277, 281, 311), 70(160), 75(6,
252, 277, 281), 77(277, 281), 78(160,
277, 281), 174, 193, 253, 259(55), 287,
288(7), 289, 299, 329, 330(119), 331,
340, 345(119), 347(119), 354, 355, 356,
357, 412, 442, 526, 552, 579, 581

Overland, R. N., 40
Owen, L. N., 438

Paege, L. M., 605

Page, J., 574

Painter, T. S., 4

Paladin, A. C., 13, 98, 165, 186, 187(163),

247, 264, 602
Pallade, G. E., 208
Pariser, S., 583
Park, J. T., 99, 187, 546
Parks, R. E., Jr., 106, 107
Parnas, J. K., 35, 125, 164
Parsons, C. H., Jr., 379, 394(22), 403(22)
Partdrige, S. M., 21, 24, 54, 263
Pascu, E., 22, 27
Pasternack, R., 33, 44(169), 66(169), 72

(169), 73(169)
Paterson, A. R. P., 135
Patterson, A. M., 82, 84(17)
Patterson, A. L., 64
Paul, T., 109, 113
Pauling, L., 440, 463, 464
Peacocke, A. R., 323, 334, 411, 412 440

(22), 476, 477, 478, 479, 522, 523(73),

552, 579

AUTHOR INDEX

645

Pearse, A. G. E., 591

Peel, Elizabeth, W., 13, 31(34), 32, 33,

70(34)
Peham, A., 300

Peiser, E., 198, 341, 418, 420(67)
Peniston, Q. P., 22
Perkins, M. E., 4, 89, 97, 418, 558
Perrings, J. D., 317, 339
Perutz, M. F., 542
Pesez, M., 54
Petering, H. G., 613
Petermann, M. L., 316, 318(46)
Peters, G., 572
Petrow, v., 13, 31(36, 37), 32, 33, 57(166),

70(166), 76(166)
Pfiffner, J. J., 95, 136
Pfleiderer, W., 119
Phelps, F. P., 19, 22, 23(129), 65(129), 73

(129)
Philips, F. S., 96, 105, 106(285d), 109, 135,

595
Phillips, J., 113, 114
Philpot, F. J., 613
Piccard, J., 2

Pickels, E. G., 476, 576, 589(74)
Pierce, J. G., 161, 198, 418
Pigman, W., 22, 23(129), 65(129), 73(129)
Piloty, O., 12, 15, 18(14), 20(14), 73(14),

74(14)
Pinner, A., 82
Pircio, A., 123, 304, 305
Pirie, A., 95

Pirie, N. W., 110, 289, 319, 323, 332, 377
Pitt, G. J., 449, 456(4)
Plenge, H., 584
Plentl, A. A., 122, 125(369, 370), 131(369),

132(370), 616
Plimmer, R. H. A., 415(60), 416
Ploeser, J. M., 500, 512
Plosz, A., 2

Podwissotzky, W. V., 15(62)
PoUister, A. W., 6, 312, 316(24, 25, 31),

317(25), 318(24, 25), 319(25), 323, 367

(24), 375, 388
Porter, J. N., 95
Porter, R. R., 573
Potter, J. L., 96, 578
Potter, V. R., 189, 209, 216, 225(11), 232,

235, 625
Pouyet, J., 484, 487, 489(134), 492(143),

527

Praetorius, E., 617, 619

Press, E. M., 511

Price, M. L., 105

Price, R. W., 125

Price, V. E., 484

Price, W. C., 383, 550

Pricer, W. E., Jr., 13, 182, 435, 591, 594

Preuss, Dr., 100

Privat de Garilhe, M., 584(121)

Pryde, J., 150

Purrmann, R., 97, 117(176)

Putnam, F. W., 321

R

Rabotin, J. G., 318

Racker, E., 59, 290

Randall, J. T., 318, 501, 538, 546(33)

Rao, P. L. N., 265

Raphael, R. A., 188

Rapport, D., 123, 339

Raymond, A. L., 34, 42(178), 132

Rechnagel, R. O., 209

Redfield, R. R., 574

Reguera, R. M., 246

Rehorst, K., 64, 74(344)

Reichard, P., 14, 49, 161, 198, 219, 271,

274, 442, 578, 608, 620
Reichert, E., 186
Reichmann, M. E., 337, 338, 440, 473,

484(71), 487, 490(71), 491, 492(155)
Reichstein, T., 56
Reindel, W., 616, 617, 618
Reinemund, K., 17, 65(78), 68(78)
Reis, J., 591
Reis, T. L., 591
Remick, A. E., 27
Reuter, M. A., 450, 495, 499
Rhein, A., 316, 318(50)
Rhoads, C. P., 105, 106(285e), 135
Richards, G. N., 52, 75(284)
Richberg, D. A., 612
Richert, D. A., 610, 611, 613
Richter, F., 161, 198,576
Richtmyer, N. K., 17, 18, 20(80), 21, 48,

69(106)
Riehl, D., 92

Riley, D. P., 318, 531, 552
Ringier, B. H., 181
Ris, H., 311, 312
Rist, C. E., 155
Rittenberg, D., 122, 599

646

AUTHOR INDEX

Ritthausen, H., 99, 100, 159

Rivers, T. M., 81

Robbins, G. B., 45

Robbins, W. J., 93

Roberts, D., 605

Robins, R. K., 130, 131, 133(435)

Robinson, F. A., 92

Robinson, R. A., 42

Roblin, R. O., Jr., 103(254), 106, 118(287),

128(287)
Rodda, H. J., 46, 65(245), 66(245), 71

(245), 72(245), 154
Roe, E. M. F., 62, 169(191), 170, 545, 551

(152)
Roeder, G., 89
Roemel, J. C, 611
Roll, P. M., 122, 125(370), 131, 132(370),

161, 198, 418, 616
Rose, F. L., 118, 123(338)
Rose, I. A., 251
Rose, W. C, 98
Rosen, G., 289, 518, 529, 530
Rosenbeck, H., 299
Rosenberg, A., 553
Rosenberg, H. R., 92
Rosseels, J., 136
Rossenbeck, H., 4, 6(33)
Rossi, A., 422

Rossiter, R. J., 15(66), 400, 529
Roth, J. S., 386, 566, 572(33), 624
Rothe, G., 20, 576
Rothen, A., 573
Rothler, H., 596
Roush, A., 598, 599, 612
Rowen, J. W., 36, 487, 489, 548(166), 549,

603, 604(203)
Roxburgh, C. M., 188
Ruch, F., 542, 544
Rudy, H., 183
Rueff, L., 186
Ruff, O., 12, 18
Ruffier, N. K., 583
Rupe, H., 126
Rush, H. P., 568, 570, 571
Russell, P. B., 104, 105(260), 127, 128, 130

(417)
Russell, P. J., Jr., 109, 113, 119, 130(346),

131(346), 361
Rutter, L., 265
Rydon, H. N., 132
Ryzhkov, V. L., 321

Sable, H. Z., 14, 507

Sadron, C, 483, 486, 487, 489(126), 492

(143)
Saffran,M., 600, 607,608
Saidel, H. F., 311, 315, 317(23), 318(23),

319(23), 331(23), 332(23), 337(23)
Salkowski, E., 124

Salomen, H., 17, 20(79), 65(79), 67(79)
Salomon, G., 94
Saluste, E., 14

Samuelson, O., 212, 213, 214(2), 216(2)
Sanadi, R., 626
Sankstone, M. I., 16
Santesson, L., 5
Sarreither, W., 597
Satoh, K., 94, 98, 157, 158(105), 188
Sattler, L., 54
Sauer, J. S., 130
Sax, K. B., 289
Scala, E., 253
Scarano, E., 600, 607, 608
Schaedel, M. L., 506, 602
Schaffer, N. K., 618
Schardinger, F., 609
Schaub, R. E., 15(60), 188
Scheel, F., 615, 619(246)
Scheele, 82

Scheibe, G., 537, 541, 542(129)
Scheibling, G., 527
Scherer, J., 98
Scherp, H. W., 125
Schiedt,U., 546, 553(159)
SchifT, H., 54
Schindler, O., 49, 76(268)
Schirp, H., 20, 68(104)
Schittenhelm, A., 94, 576, 609, 614
Schlags, R., 458

Schlenk, F., 7, 8(72), 13, 33, 35(45), 36,
40(45), 58, 73(45), 82, 94, 181, 188,
409, 506, 545, 602, 605
Schlenker, J., 90
Schlubach, H. H., 27
Schmedes, K., 92
Schmetz, F. J., 184, 185(259)
Schmid, J., 82, 96(10)
Schmidt, F. O., 535

Schmidt, G., 6, 8(66), 161, 163, 164, 192,
277, 279, 333, 345, 418, 420(71), 423,
424, 432(100), 436, 476, 559, 560, 562,

AUTHOR INDEX

647

569, 575, 576, 577, 580(75), 588, 589
(16, 74, 75), 596, 598, 601(173)
Schmidt, L. H., 104
Schmidt, W., 119
Schmidt, W. J., 538
Schmitt, J. A., 613
Schmitz, H., 189, 235, 625
Schneider, W. C, 208, 285, 374, 448, 584,

624, 625
Schoenheimer, R., 122, 125(369), 131(369)
Scholes, G., 339
Schopf, C, 97

Schopp, K., 17, 20(79), 65(79), 67(79)
Schorger, A. W., 22
Schotte, H., 11, 17
Schramm, G., 254, 271, 425, 426, 436
Schreier, E., 21, 71(112), 72(112)
Schroeder, E. F., 101
Schryver, J., 313, 318(32)
Schuetz, R. D., 286, 289(4)
Schuler, W., 615, 616, 617, 618, 619(246)
Schulman, M. P., 614
Schultz, E., 99
Schultz, J., 5
Schultz, R., 616
Schultze, M. O., 100

Schuize, E., 138, 140, 197

Schuster, L., 186

Schiitz, F. A., 97

Schwander, H., 325, 328, 329, 338(99),
440, 473, 476, 479, 484(72), 485, 486,
487, 489(138, 140), 527

Schwartz, L. I., 583

Schwarz, K., 301, 302, 303

Schwarzenbach, G., 181

Schweigert, B. S., 251

Schwob, C. R., 121

Scott, J. F., 337, 501, 546(34)

Scouloudi, H., 574

Seagran, H. L., 109, 117, 124(331), 192,
193(8), 195(8), 199(22), 200(22), 201
(8, 22), 203(8), 205(8), 209, 390, 396,
402(50), 403(50), 499

Seeds, W. E., 534, 535(103), 536, 537, 538,
539, 540, 541(127), 542(103, 106, 126),
544(127), 547, 552

Seegmiller, J. E., 35, 41(188), 67 (188b)

Seibert, F. B., 287, 531

Seidel, M. K., 581

Seidell, A., 251

Senior, J. K., 99

Seraidarian, K., 124, 192, 193(6), 194(6),

199, 200(6), 201(6,) 375, 387, 392,

396, 399, 424, 436, 498(22), 512, 560,

562(16), 569(16), 589(16)

Seraidarian, M., 424, 436, 560, 562(16),

567, 569(16), 589(16), 590(37)
Seraydarian, M., 339
Set low, R., 339
Sevag, M. G., 323
Shack, J., 479, 488, 515, 518(54), 522, 523

(53, 54), 526, 527, 529, 531
Shaffer, B., 5
Shaffer, P. A., 458
Shafizadeh, F., 53, 287, 288(7), 299(7)
Shakespeare, N., 91, 92(88)
Shapiro, D. M., 599

Shapiro, H. S., 251, 252, 253(52a), 264
(52a), 321, 338, 341, 343(183), 354,
367(183, 221), 444, 445(159), 577, 578
(78)
Shapot, V. S., 321, 582
Sharp, D. G., 375
Sharp, E. A., 103

Shaw, E., 107, 134, 135(451, 452), 612
Sheehen, H. L., 102
Sheffer, H., 473
Shemin, D., 122
Sherry, S., 246,584
Sherwin, C. P., 94
Sherwood, M. B., 104, 105(260)
Shettles, L. B., 330
Shive, W., 104, 614
Shooter, K. V., 313, 318(36)
Shorey, E. C., 95

Short, L. N., 112, 495, 525(9), 546, 547, 550
Shugar, D., 98, 108(d), 112, 113, 115(321),
116(321), 217, 227, 249, 495(15), 496,
499, 500, 502(g, f), 503 (f, g), 504, 505,
506, 507, 508(f, g), 509(f, g, h), 515,
516, 517, 518, 574
Shunk, C. H., 13, 31(34), 32, 61, 62(326),

70(34), 79(326), 80(326)
Shuster, L., 429, 435, 592, 593
Siebert, G., 624

Siefken-Angermann, Marianne, 57
Siegel, A., 332
Sigmund, W., 46, 65(246)
Signer, R., 322, 325, 328, 329, 338(99), 440,

473, 476, 479, 484(72), 485, 489, 539
Silverman, H., 618

648

AUTHOR INDEX

Simmons, N. S., 327, 328(102), 329(102),

335
Simms, E. S., 601, 607(189), 626
Simms, H. S., 39, 112, 113, 119(314), 156,
163, 168(93), 169(93), 227, 268,269(1,
2, 3), 270, 410, 455, 457(21, 22, 23),
459(21, 22), 475, 481
Simon, R. R., 329
Simpson, D. M., 95
Singer, S. J., 332
Singer, T. P., 182

Sinsheimer, R. L., 49, 123, 162, 223, 225,
233, 234(46, 47), 235(28), 345, 347, 350
(240), 354, 412, 441(30), 442, 501, 520,
546(34), 578, 585, 588(134)
Sisk, W. N., 103
Skeggs, H. R., 100
Skipper, H. E., 105, 107
Skoog, F., 120
Smaller, C. J., 16

Smellie, R. M., 204, 205(59), 208(59), 209
(59), 260, 261(89), 262(89), 26?, 269,
276(7, 8), 277(8), 278, 387, 396, 397
(63), 398(63), 400(63), 401(63), 402
(63), 418
Smith, D. B., 473
Smith, E. E. B., 189
Smith, E. L., 136
Smith, H., 148, 153(46), 174, 176(205), 194,

420
Smith, J. D., 90, 162, 170(136), 171(136),
192, 193(7), 195(7), 204(7), 205(7),
206, 207(61), 246, 247, 249(26), 250,
251(26), 252, 253(26), 255, 256, 259,
260, 270, 271(10), 272(10), 273(10),
274, 275(17), 277(9, 17, 18, 19), 279(9,
10, 17), 280(9, 10, 25), 281(10), 282,
284, 310, 329, 332, 333(146), 347, 349,
350(146), 359, 370(238), 374, 389, 394,
395, 396, 403, 404(48), 418, 419, 425,
426, 427(83, 113), 429(113), 437(113),
439, 442, 443(152), 445(114), 499, 511,
562, 564, 565, 569(18), 577, 578
Smith, K. C, 245, 418, 521
Smith, K. M., 378, 380(21), 382(21), 386

(21), 530
Smith, M. E., 48
Smith, R. L., 94
Smith, R. M., 217, 227(14)
Smolens, J., 323
Smyrniotis, P. Z., 35, 41(188), 67(188b)

Snell, E. E., 92, 185

Snell, N. S., 329

Snellmann, O., 257, 332, 485

Sobotka, H., 150, 157

Solmssen, U. V., 19, 29, 30, 45(88), 65(86,

88), 70(86, 158), 71(158), 181
Somlo, F., 94, 99(118), 136
Sonne, J. C, 122
Soodak,M.,123,184,305
Sowden, J. C, 10, 11, 18, 48, 50, 67(5),

74(5)
Speer, J. H., 132
Spicer, D. S., 100
Spicer, V. L., 199, 260
Spies, J. R., 13, 20(33), 21(33), 43(33),
65(33), 66(33), 67(33), 68(33), 97, 117
(179), 118(179), 131
Spies, T. D., 92
Spitnik, P., 361, 365(260)
Spring, F. S., 26, 47, 69(254), 95, 125(135,

136), 154, 159, 506
Sprinson, D. B., 121, 122
Sreenivasan, A., 612
Sreenivasaya, M., 253, 319, 323(73)
Stacey, M., 10, 27, 30, 33, 47, 49, 50(6,
277), 53(4), 55(251, 277, 281), 56(277),
57(160), 59(291), 60(277, 281, 311), 70
(160), 75(6, 251, 277, 281), 77(277,
281), 78(160, 277, 281), 193, 287, 288(7,
10), 289(10), 290(10), 299(7), 331, 332,
340, 354, 355, 356, 357, 391, 412
Stanley, W. M., 377, 378, 380, 383(17, 27),

386(27), 389, 390, 392, 474, 475
Steberl,E. A.,584
Stedman, E., 312
Stedman, Ellen, 312
Steele, R., 290, 300
Steiger, Marguerite, 16, 18(70), 20(70),

67(70), 73(70)
Stein, W. H., 212, 215, 219, 573, 575(53c)
Steinberg, M. A., 337, 529
Steiner, R. F., 318, 489
Stenhagen, E., 321, 476
Stenstrom, W., 112
Stephenson, M., 600
Stepka, W., 35, 248
Stern, H., 625
Stern, K. G., 313, 316, 318(32, 51), 337,

470, 471(61), 525, 529, 531
Stern, L., 615
Sternberg, S. S., 105, 106(285d), 135

AUTHOR INDEX

649

Stetten, M. R., 134, 613, 614(236h)
Steudel, H., 3, 86, 87, 89(51), 156, 168,

198, 341,418, 420(67), 475, 619
Stevens, C. D., 118
Stewart, C. P., 609
Stickney, M. E., 494
Stiller, E. T., 34, 40(175), 42(220), 46(175,

220), 65(220), 66(220), 67(175, 220),

68(175, 220, 247), 69(220), 73(175,

220), 74(219), 144, 163
Stimson, M. M., 123, 450, 495, 499, 546,

553(159)
Stock, C. C, 105, 106(292), 135
Stokes, A. R., 329, 531, 535, 542(106)
Stone, A., 109, 112, 552
Stoppelenburg, J. C, 414
Storey, I. D. E., 187
Story, L. F., 14, 97, 143, 150(24), 505, 511

(37)
Stowell, R. E., 387, 393(45), 498
Stump, B., 135
Stuart, A., 532, 539(101)
Strandskov, F. B., 103, 104(252)
Strauss, U. P., 484
Strecker, A., 85, 98, 122(35), 124(35)
Strickland, K. P., 400
Strickler, N., 424, 560, 562(16), 569(16),

589(16)
Strobele, R., 17, 29, 65(78), 71(154)
Strominger, D. B., 605
Struve, K., 101
Stumpf, P., 291
Sturgeon, B., 13, 31(37), 32, 33, 57(166),

70(166), 76(166)
Sturgis, S. H., 258, 352
SubbaRow, Y., 179, 202
Sugiura, K., 105, 106(292)
Sumi, I\I., 96
Sunell, K., 618
Suranyi, J., 42
Sutherland, E. W., 37
Sutherland, G. B. B. M., 546, 549
Suzuki, v., 94, 316
Swann, M. M., 538
Swartz, B. H., 423
Sweeny, L., 470
Swift, H., 299, 542, 545
Sykes, M. P., 135
Szarfarz, D., 375, 380, 381
Szent-Gyorgyi, A., 612

Taborda, A. R., 587

Taborda, L. C, 587

Tafel, J., 119

Takahashi, H., 422

Takenaka, Y., 161, 198

Takman, B., 60, 93, 188

Talbert, P. T., 367, 416

Talbot, B. E., 244, 248(5), 263(5)

Tamm, C, 196, 251, 252, 253, 257 (52a),
258, 264, 287, 335, 336, 338, 339(170),
340(170), 341(170), 342, 343(170, 183),
344, 345, 346(224), 347(170), 354, 366
(170), 367(183, 224, 225), 443, 444, 445,
521, 577, 578

Tan, T. C, 135

Tarbell, D. S., 91, 92(88, 89)

Tarr, H. L. A., 14

Tarttar, A., 96

Tatchel, A. R., 29

Taylor, A. N., 4

Taylor, A. R., 375

Taylor, B., 339

Taylor, C. W., 43, 44(232), 65(232), 66
(232), 68(232), 72(232), 153

Taylor, E. C, 131

Taylor, H. F., 95

Taylor, H. F. W., 113, 217, 338, 340, 341
(185), 411, 440(20), 455, 456(24), 457
(24), 458, 463, 466(48), 476(48), 477,
478(48), 479(48), 480(48), 482(48),
484(48), 489(48), 490(48), 522, 523(72)

Teece, Ethel G., 49, 50(277), 53, 55(277),
56(277), 59(291), 60(277), 75(277), 77
(277), 78(277), 287, 288(10), 289(10),
290(10)

Tennent, H. G., 338, 470, 471, 476(59),
480, 491

Teorell, T., 320(79), 337, 476

Terszakowec, J., 179, 594, 601

Thain, E. M., 13, 184, 185(257)

Thannhauser, S. J., 48, 136, 161, 162, 165,
170, 172(142, 143), 192, 194, 196, 198
(18), 225, 277, 279, 333, 341, 366(218),
418, 420(71), 423, 424, 432(100), 436,
440(144), 441, 560, 562(16), 569(16),
575, 588, 589(16)

Theodore, S., 290, 300(18)

Thiel, A., 116

Thiersch, J. B., 96, 105, 109, 595

650

AUTHOR INDEX

Thomas, J. F., 62, 169(191), 170, 194, 219,
545, 551(152)

Thomas, L. E., 330

Thomas, P., 4, 96, 296

Thomas, R., 336, 339, 488, 515, 525(55, 79) ,
527, 528, 529, 544, 559

Thompsett, J. M., 479, 488, 515, 518(54),
522, 523(53, 54), 526(54), 527, 529, 531
(70)

Thompson, A., 48, 72(255), 73(255)

Thompson, H. W., 546, 547, 550

Thompson, R. H. S., 557, 559(3)

Thompson, R. L., 105

Thorell, B., 5, 374, 542, 544

Threlfall, C. J., 323, 328(98), 335, 473

Tillett, W. S. T., 584

Tillotson, J. A., 188

Timms, R. N., 127, 128(415)

Tinker, J. F., 97, 104(182), 106, 107, 109,
110, 118(301), 122(301), 131(182), 132
(182), 133(182, 438), 134(301), 135
(301), 495, 498(13)

Tipson, R. S., 12, 14, 20, 21, 24, 25, 26,
27, 38, 42(136, 210), 43(110), 47(101),
65(110), 66(101, 110), 67(101, 136,
210), 68(110, 136), 69(110), 73(136),
142, 144, 145, 146(27, 36), 147(17, 36),
157(17), 173(17, 36), 174(45), 409, 413
(2)

Tiselius, A., 250

Tishler, M., 33, 45, 47(240), 72(240), 73
(240)

Todd, A. R., 12, 13, 14, 26, 30, 31(38), 34,
37(144), 38, 40, 43(144), 44(144, 232)
45(159), 46, 47, 49, 51(250), 61, 62, 65
(232, 245), 66(144, 159, 232, 245), 68
(232), 70(159), 71(245), 72(232, 245),
73(180), 77(52), 79(327), 80(327), 93,
98, 119, 123(339), 131(339), 132(436),
139, 147, 148, 149, 150, 151(54), 152,
153, 154(46, 47), 156(3), 157, 162, 163,
166, 167(170), 169(173), 170(172), 171
(173, 193), 172(145), 173, 174(158, 170,
171), 175(173, 197), 176(145), 177, 179,
180(199, 211), 181, 183, 184, 185(248),
187(250), 188(287), 189, 194, 218, 219
(20), 221(20), 257, 260, 279, 292, 366,
367, 410, 414, 416(46), 417(54, 63),
418, 419(62), 420(62), 421, 422, 423
(62), 424, 425, 426, 427(109), 428, 430
(62), 431, 433, 434, 437, 438(131), 440,

441, 443, 444(62), 462, 479, 506, 520,
545, 559, 564, 565(62), 570(10, 11), 588

(11)
Tobie, W. C, 299
Tompkins, E. R., 212, 213, 214(6), 215,

216(9), 223(8)
Tonzetich, T., 600
Topham, A., 119, 123(339), 127, 131(339),

152, 153
Torkington, P., 546
Totter, J. R., 610, 611
Traube, W., 85, 86, 104, 122(30), 126(36),

131(30), 132(30, 36, 270)
Trauth, O., 94, 157, 158
Trautmann, M. L., 583
Tressler, D. K., 95
Trim, A. R., 600
Trippett, S., 51, 57(282)
Tristram, G.R., 311, 574
Trufanov, A. V., 15(67, 68)
Truszkowski, R., 619
Tsuboi, K. K., 339, 387, 393(45), 498, 519,

529
Tuttle, L. C, 184

U

Umbreit, W. W., 34, 35, 39, 301
Ungar, J., 331
Urban, F., 458
Upson, F. W., 45
Uspenskaya, M. S., 330
Usteri, E.', 13, 40(41)
Uzawa, T., 585
Uzman, L. L., 345

Valentik, K. A., 100

Vallet, G., 484

Vandendriessche, L., 410, 412, 431, 481,

559, 563, 568
Vanderlinde, R., 613
VanderWerff, H., 104, 105(260)
van Ekenstein, Alberda W., 12, 15, 17,

18(16), 20(15), 64, 65(16, 100), 67(16,

100), 73(15,69, 100,343), 141
Vanhauke, A., 572
Van Slyke, D. D., 42
Van Winkle, Q., 318
van Winkle, W. W., Jr., 103
Varin, R., 337, 440, 473, 484(71), 487(71),

490(71)

AUTHOR INDEX

651

Vaughan, J. R., 106, 118(287), 128(287)

Vellu, L., 54

Vendrely, C, 318, 531

Vendrely, R., 7, 8(71), 318, 531

Vercauteren, R., 580

Verkade, P. E., 414

Vestin, R., 181

Vilbrandt, C. F., 338, 470, 471, 476(59),

480, 491
Vincent, W. S., 374, 400(8)
Virchow, R., 95
Vischer, E., 6, 8(64), 14, 89, 117, 124(330),

161, 192, 193(4), 195, 197(4), 204(25),
205(25), 208(25), 209(25), 218, 243,
245, 246, 248, 252, 253, 256, 257(40),
259(68), 260(68), 261, 262(17), 263, 264
(83), 308, 310(3), 329, 333(114, 116),
334(114, 116), 336, 337(114, 161), 346
(114, 161), 347(114, 161), 348, 349(114,
161), 350(114, 248), 351(114, 161), 354,
359, 368(114), 374, 387, 390(44), 391
(44), 394(4, 44), 395(44), 396(44), 397
(44), 399(44), 402(44), 403(44), 418,
498, 499, 512

Viscontini, M., 64, 80(342)

Visser, D. W., 44. 64(237), 156

Voegtlin, C, 94

Vogel, H. J., 90, 117(77), 195, 197(24),
204(24), 205(24), 218, 251, 258, 259
(53), 332, 348(153), 354, 402

Vogler, H., 126

Volkin, E., 36, 38(190), 49, 57(270), 75
(270), 76(270), 79(334), 80(334), 161,

162, 164(135), 165(135), 167, 172, 189,
208, 209(73), 223, 225, 226(40), 227,
228(27, 40), 230(40), 231(40, 43), 232
(43), 233, 234(49), 240, 241, 262, 265,
347, 380, 386(25), 387, 392, 396, 397
(25), 398(25), 399(25), 414, 419, 422,
423(82), 424(96), 425(96), 426, 427
(96), 429(96), 432(110), 434, 435, 436
(124), 437(96), 441, 474, 475(85a), 497,
501(19), 562, 569(17), 577, 586, 587
(136), 588(136), 589

Vollbrechthausen, I., 273, 284, 283(14)

von Braun, J., 21, 68(105)

von Dahl, K., 236

von Euler, H., 13, 40(41), 181, 290, 300,

302, 303, 316, 318, 337, 338, 411, 475

(97)
von Fodor, K., 150, 155(57)

von Ker^kjdrto, B., 254, 436
von Kiihlevvein, M., 150
von Liebig, J., 160
von Lipmann, E. O., 99
von Mering, 103
von Sallmann, L., 302
von Schuh, H.-G., 133
Vural, I. L., 583

W

Wachsman, J. T., 620(274d)

Wachstein, M., 591

Wacker, A., 31, 107, 347

Wade, H. E., 248

Wadman, W. H., 21, 24

Wagman, J., 313, 318(32)

Wajzer, J.,37,41(204),601

Wakabayashi, Y., 598, 599(174)

Waldron, D. M, 54, 552

Wald vogel, M. J., 506, 602

Walker, J., 112, 127(312), 128(312), 450,

495, 525(7), 534
Walker, M. P. B., 531, 542
Wallace, W. S., 95
Waller, C. W., 15(59), 95, 160
Walsh, E. O., 413

Wang, T. P., 14, 186, 435, 507, 607, 622
Warburg, O., 13, 181, 183, 610
Ward, P. F. V., 264
Ward, W. H., 329
Wargon, M., 124, 192, 193(6), 194(6), 200

(6), 201(6), 396, 399, 498(22), 512
Waritz, R. S., 209
Wassermeyer, H., 227, 269
Watanabe, I., 316, 474
Watson, J. D., 365, 371(262), 407, 440,464,

465, 466, 467, 468, 470, 480, 488, 489,

490, 525, 531, 541, 548
Watson, M., 361, 365
Watson, P., 584
Waymouth, C, 5, 6, 7(53), 380, 387, 394

(24)
Webb, M., 253, 259(55), 313, 329, 330(119) ,

331, 345(119), 347(119), 412, 438, 442,

526, 576, 581, 584, 585
Weber, H. H., 535
Weber, I., 601

Weed, L. L., 61, 90, 189, 255
Weerman, R. A., 64, 74(346)
Wehmer, C, 93
Weil-Malherbe, H., 599

652

AUTHOR INDEX

Weiner, S., 410, 423

Weinschenk, A., 119

Weiss, J., 339

Weissberger, A., 532

Weissman, N., 582

Weliky, Virginia S., 60, 61(317), 80(317),

94, 188
Wellman, J. W., 33, 45, 47(240), 72(240),

73(240)
Wendt, G., 157
Wenis, Ed., 19, 45(88), 65(88)
Wenzel, F., 142, 197
Westerberg, J., 35

Westerfeld, W. W., 241, 610, 611, 612, 613
Westheimer, F. H., 455, 460
Westphal, K., 93
Weygand, F., 17, 31, 46, 64, 65(78, 246),

94, 107, 157, 158, 183, 347
Weymouth, F. J., 184
Wheeler, H. L., 87, 88, 8^(52) , 90 (45) , 101,
103, 107, 108(c, e, g), 117(66), 123, 127
(245), 128(245), 129, 130(226, 227)
White, D., 189
White, J. C, 318, 329, 544
Whitehead, C. W., 129, 130
Whitehead, J. K., 248
Whitfeld, P. R., 437, 438, 564, 566
Whittaker, N., 108(a), 112, 118, 127(317)
Widstrom, G., 299, 332, 485
Wiechowski, W., 615
Wieland, H., 612
Wieland, O. P., 100
Wieland, T., 246, 274
Wiener, G., 481
Wiener, O., 534(105), 535
Wiener, W., 614

Wiggins, L. F., 10, 47, 49, 50(6, 277), 53,
55(251, 277, 281), 56(277, 281), 59
(291), 60(277, 281), 75(251, 277, 281),
77(277, 281), 78(277, 281), 287, 288
(10), 289(10), 290(10)
Wigglesworth, V. B., 96
Wilkins, M. H. F., 318, 329, 467, 468(54),
531, 534, 535, 538, 539, 540, 541, 542
(106, 126), 543, 544, 552
Wilkinson, W. K., 130
Will, F., 95
Williams, J. H., 15(59), 16, 95, 100, 160,

610
Williams, J. N., Jr., 587
Williams, R. C, 337, 489

Williams, R. D., 458

Williams, R. H., 103

Williams, R. J., 244

Williams, R. R., 92

Williams, R. T., 150

Williams, W. J., 601, 607

Williams, W. L., 185

Williamson, M., 105

Willits, C. H., 131, 133(435)

Wilson, A. T., 187

Wilson, A. Y., 105

Wilson, D. W., 120, 122(356), 134, 255, 619

Wilson, PI. R., 329, 467, 468(54), 531

Winter, L. B., 13, 24

Winterstein, A., 94, 99(118), 136

Wirth, F., 64

Witney, I. B., 610, 611

Wohler, F., 82, 102

Wolf, L. M., 120

Wolfrom, M. L., 40, 44, 48, 66(239), 72
(255), 73(239, 255), 286, 289

Wood, J. K., 109, 113, 272

Woodhouse, D. L., 120, 123(389), 304, 329,
552

Woodside, G. L., 106

Woolf, D. O., Jr., 135

Woolfson, M. M., 188, 189(287, 414)

Woolley, D. W., 103(256), 107, 134, 135
(451), 196, 233, 234(50), 421, 428, 429,
520, 562(21), 563, 569(21), 570(21)
612

Wiilfken, F., 116

Wright, L. D., 100, 103(255), 105

Wyatt, G. R., 61, 88(54), 89, 90(70), 92
(54), 108(f, h), 109, 117(54, 70, 72),
118(70), 130, 136, 188, 245, 249, 250,
252, 253(47), 258(11), 259(11, 47), 261
(11, 48, 81), 262(11, 47), 329, 330(123),
332, 333(146), 344(151), 346(151), 347
(123), 350(146), 354, 355, 356, 357, 358
(123, 154), 359, 360, 361(234), 370
(151), 466, 478, 499, 500
Wyss, O., 103, 104(252)
Wyssmann, L., 391, 403(53)

Yamagawa, 601
Yanai, M., 128
Yoshimura, K., 94
Young, W. J., 181

AUTHOR INDEX

653

Zamenhof, S., 60, 89, 255, 313, 314, 322,
323(91), 324(41), 328(91), 330, 331(37,
41, 91), 332(41, 139), 333(41, 142, 143),
334(41), 336, 337(161), 338(90, 91),
341, 345(212), 346(161), 347(161), 349
(161), 350(141, 142), 351, 356(161),
359, 361(160), 366(212, 227), 391, 393,
405, 411, 443(26), 491, 516, 518(57),
559, 561, 562, 563, 579, 582, 584

Zatman, L. J., 604, 608, 609

Zbarsky, S. H., 135

Zerban, F. W., 54

Zill, L. P., 236, 237(55), 240(57)

Zimmerman, M., 160

Zinner, H., 21, 26, 43, 44, 45, 47(114), 65
(141, 235), 66(141, 235, 238), 67(113),
68(113, 114), 69(141), 71(113, 114a,
114b), 72(113, 114a, 114b, 238), 151

Zittle, C. A., 410, 438, 481, 569, 571, 572

Zollner, N., 386, 424, 560, 562(16), 569(16),
572, 589(16)

Zuazago, G., 202

Zuckerman, R., 204, 205(58), 512

Zussman, J., 149, 462

Zweig, G., 244, 263(4)

Subject Index

Absorptivities,

biaxial crystals and, 535-536
linear systems and, 535
principle, measurement of, 533-534
Acetic acid,

paper chromatography and, 253
salts, nucleotide separation by, 226
Acetoacetic acid, pyrimidine synthesis

and, 125, 130
Acetaldehyde, deoxyribose biosynthesis

and, 59
Acetamidocyanoacetate, purine syn-
thesis and, 133
Acetobromo-D-arabofuranose, nucleoside

synthesis and, 151
Acetobromoglucose, nucleoside synthesis

and, 155
Acetobromoribofuranose, nucleoside

synthesis from, 155-156
Acetochloro-D-ribofuranose, nucleoside

synthesis and, 151
Acetol, formation from thymine, 121
Acetone,

paper chromatography and, 255-256
thymine estimation and, 305
Acetonitrile, pyrimidine synthesis and,

130
Acetyl phosphate, hydrolysis of, 37
Acetylene, pyrimidine synthesis and, 130
3'-Acetylthymidine, nucleotide synthesis

and, 176
3'-Acetyl-5'-tritylthymidine, 176
5-Acylaminopyrimidines, purine syn-
thesis and, 132
Adenase, 97, see also Adenine deaminase
Adenine, 2, 3, 14, 40, 124, 346
adenosine deaminase and, 596
amino group turnover, 599
bond distances and angles, 451-452
cationic exchange of, 218
color reaction for, 304
deamination, 118, 131, 260
adenosine and, 600

2,6-diaminopurine and, 105
discovery of, 84
dissociation of, 114, 268
dissociation constants of, 456, 457

resonance and, 457-458
glutamate and, 600
hydrogen atoms of, 452
hydrolysis of, 135
inosine formation from, 602-603
iodine and, 121

nucleohistone fractionation and, 362
nucleotide pyrophosphorylase and,

607-608
occurrence of free base, 94
oxidation of, 122

paper chromatography of, 249, 253, 255
reaction with acetobromo-D-arabo-

furanose, 151
reaction with ribose-1 -phosphate, 37
reducing agents and, 119, 120
spectrophotometric estimation of, 200-

202
stability,
acid, 117-118
alkali, 118
synthesis of, 85, 131, 133, 135
thymine ratio, 352-355
ultraviolet absorption of, 498-499, 502
wheat germ nucleic acid and, 347-348
xanthine oxidase and, 612
Adenine deaminase, distribution of,

595-596
Adenine deoxyriboside, 139, 140
deamination of, 142, 597
glycosidic linkage of, 143
hydrolysis of, 57
Adenine deoxyriboside phosphate, 171,

see also Deoxyadenylic acid
Adenine nucleotide esters, ribonuclease

and, 565
Adenine picrate, 124
Adenine riboside, see Adenosine
Adenine thiomethylpentoside, isolation
of, 94-95, 157

655

656

INDEX

Adenine thiomethylriboside, nucleoside

phosphorylase and, 602
Adenosine, 15, 19, 36, 138, 157

adenine deamination and, 600

adenine estimation in, 304

borate complex, electrophoresis of, 283

conversion to inosine, 142

deamination of, 595

discovery of, 140

dissociation constants of, 459

glycosidic linkage of, 143

hydrolysis of, 57, 606

ionic properties of, 218, 219, 272

methylation of, 42, 144

phosphate linkages, 413-415

purine nucleoside phosphorylase and,
602, 603

reduction of, 119-120

structure, x-ray diffraction and, 454-
455

synthesis of, 151

thymidine phosphorylase and, 605

trityl derivatives of, 145

ultraviolet absorption of, 505, 508, 510

uridine hydrolase and, 606

uridine phosphorylase and, 606
Adenosine-2' benzyl phosphate, hydroly-
sis of, 416-417
Adenosine-3' benzyl phosphate, hydroly-
sis of, 416-417
Adenosine-5'-benzyl phosphate, perio-
date oxidation of, 438

stability of, 417
Adenosine benzyl phosphate b, nucleases

and, 432
Adenosine deaminase,

Aspergillus, specificity of, 596

distribution of, 596

inhibitors and, 597

intracellular localization of, 625

specificity of, 597

stability of, 597
Adenosine diphosphate, 13, 167, see also
Adenosine-5'-pyrophosphate

coenzyme A and, 186

deamination of, 596

paper chromatography of, 256, 264-265

synthesis of, 180
Adenosine-5'-ethylphosphonate, 176
Adenosine kinase,

activators, 594

assay of, 594

specificity of, 594
Adenosine-5'-phenylphosphonate, 176
Adenosine-2'-phosphate, 596, 597

paper chromatography of, 257

prostatic phosphatase and, 590

synthesis of, 188-189

ultraviolet absorption of, 513
Adenosine-2', 3'-phosphate, 597, see also
Nucleotides, cyclic

ribonuclease and, 425
Adenosine-3'-phosphate,

adenine estimation in, 304

cleavage of, 38-39

cysteine-sulfuric acid reaction and, 303

deamination of, 39, 596

paper chromatography of, 257

prostatic phosphatase and, 590

ultraviolet absorption of, 513
Adenosine-5'-phosphate,

biosynthesis of, 607

coenzymes and, 179, 183, 185

deamination of, 596

hydrolysis of, 63

ion-exchange of, 232

occurrence of, 160

paper chromatography of, 256, 264-265

periodate oxidation of, 438

preparation of, 594

prostatic phosphatase and, 590

structure of, 163-164

synthesis of, 173

ultraviolet absorption of, 512, 513
Adenosine-D-phosphate-ribose, deamina-
tion of, 596
Adenosine-5'-pyrophosphate, see also
Adenosine diphosphate

coenzymes and, 181

ion-exchange of, 231, 232
Adenosine-5'-tetraphosphate, 189

occurrence of, 60
Adenosine triphosphate, 13

deamination of, 596

hydrolysis of, 39, 40-41, 179

ion-exchange of, 232

paper chromatography of, 256, 264-265

reduction of, 120

ribose-5-phosphate preparation and, 63

synthesis of, 180

transphosphorylation and, 626

ultraviolet absorption of, 512

INDEX

657

Adenosine-5'-uridine-5'-phosphate,
stability of, 420
synthesis of, 176-177
Adenj'lic acid, 4, see also Adenosine

phosphates
biosynthesis of, 607-608
chromatography,

ion exchange, 222, 223, 224

paper, 249
deamination of, 595
dissociation constants of, 459-460
hj'perchromic effect and, 520
ionization of, 268-269
isomers, 37-38

cyclization of, 167

formation of, 165

hydrolj'sis of, 165

identification of, 167

infrared absorption and, 545

ion -exchange of, 228

interconversion of, 413-414

properties of, 166-167

ribose phosphate in, 165-166

spleen nuclease and, 432

ultraviolet absorption of, 551
5'-nucleotidase and, 592
pentose nucleic acid,

liver, 397-398

ratios, 407

yeast, 402
reduction of, 119-120
ribonuclease and, 571
ultraviolet light and, 123
5'-Adenylic acid deaminase,
adenosine kinase and, 594
assay of, 597
extraction of, 598
occurrence, 597
specificity of, 597
Adenylthiomethylpentose, 15, see also

Adenine thiomethylpentoside
Adonitol, 48, see also Ribitol
ADP, see Adenosine diphosphate
Adrenocorticotropic hormone,
phosphodiesterase and, 588
Adsorption, nucleic acid separation

by, 333
Agar,

diphenylamine and, 289
electrophoresis and, 274

L-Alanine, 614

nucleotide derivatives and, 99
/3-Alanine, formation,

dihydrouracil and, 621-622

uracil and, 119
Albumin,

deoxypentose nucleic acid denatura-
tion and, 529, 531

pentose nucleic acid and, 381
Alcohols, paper chromatography and, 250
Aldehyde dehydrogenase, 609
Aldehydes,

carbazole-sulfuric acid reaction and,
299

cysteine reaction and, 293

cysteine-sulfuric acid reaction of, 295

diphenylamine and, 288

trj'ptophan -perchloric acid reaction
and, 296

xanthine oxidase and, 612
Aldoheptoses, orcinol reaction and, 301
Aldohexoses,

cysteine reaction and, 293

cysteine-sulfuric acid reaction and, 303

orcinol reaction and, 301-302

tryptophan-perchloric acid reaction
and, 296
Aldopentoses,

cysteine reaction and, 293

tryptophan-perchloric acid reaction
and, 296
Alkali,

nucleic acid separation by, 333

pentose nucleoprotein separation and,
380
Allantoin, 122, 614

formation from uric acid, 615, 618
Alloxan, 82, 122

biological activity of, 102-103

sj'nthesis of, 102
Alloxanic acid, formation from uric

acid, 618
D-Altronic acid, ribose synthesis and,

17-18
Amicetin, components of, 135
Amidines,

purine synthesis and, 132

pyrimidine synthesis and, 125
Amino acid oxidase, coenzyme of, 610
Amino acids,

cysteine reaction and, 293

658

INDEX

cysteine-sulfuric acid reaction and,
295-296

;3-Amino acids, formation from pyrimi-
dines, 619

o-Aminobenzaldehyde, thymine estima-
tion and, 121

p-Aminobenzoic acid, amicetin and, 135

2 - Amino -4 - chloro - 6 - methylpyrimidine,
dichroism of, 538

D-3-Amino-3-deoxyribose, occurrence, 160

2-Amino-4,6-dichloropyrimidine, x-ray
diffraction and, 448-449

4-Amino-2,6-dichloropyrimidine, x-ray
diffraction and, 448-449

4-Amino-6-glycosylaminopyrimidine, pu-
rine glycoside synthesis and, 152

2-Amino - 4 - hydroxy - 6 - carboxypterin,
xanthine oxidase and, 613

2-Amino-4-hydroxy-6-formylpterin, xan-
thine oxidase and, 613

2-Amino -4 - hydroxy - 6 - hydroxymethyl-
pterin, xanthine oxidase and, 613

6-Amino-4-hydroxy-2-methylthiopyrimi-
dine, conversion to chloro derivative,
127

2-Amino-6-hydroxypurine, see Guanine

6-Amino-2-hydroxypurine, xanthine oxi-
dase and, 612

6-Amino-8-hydroxypurine, xanthine oxi-
dase and, 612

4-Amino-6-hydroxy pyrimidine, 2-substi-
tuted, purine synthesis and, 131

4-Amino-5-imidazolecarboxamide
phosphorylation of, 595
purine formation and, 134-135, 614

4-Amino-5-imidazolecarboxamide ribo-
side, 36-37

4-Amino-5-imidazolecarboxamide ri bo-
tide, formation of, 614

4-Amino-5-imidazolecarboxamidine, 107,
118
purines and, 135

/3-Aminoisobutyric acid, dihydrothymine
and, 621-622

2-Amino-4-methyl-6-chloropyrimidine, x-
ray diffraction and, 448-449

5-Aminomethyluracil, preparation of, 92

2-Amino-5-nitropyrimidine, hydrolysis
of, 125

p-Aminophenol, xanthine oxidase and,
612-613

Aminopterin, deoxyribonuclease and, 582
6-Aminopurine, see Adenine
8-Aminopurines, synthesis of, 131
Aminopyridines, 118
2-Aminopyrimidine,
formation of, 125
reduction of, 120
4-Aminopyrimidine, synthesis of, 128
5-Aminopyrimidine, purine synthesis

and, 131
6-Aminopyrimidine,
reduction of, 120
tautomerism of, 448
6-Amino-2-pyrimidinethiols, synthesis

of, 127
3-Amino-D-ribose, 188
occurrence of, 15, 95
Aminosugars, nucleotide derivatives

and, 99
4-Aminothymine, 126
4-Amino-2,5,6-trihydroxypyrimidine, oc-
currence of, 100
4-Aminouracil, reduction of, 126
5-Aminouracil, 104
oxidation of, 622

thymidine phosphorylase and, 605
Ammonia, paper chromatography and,

249, 251, 253, 254, 257, 264-265
Ammonium sulfate, paper chromatogra-
phy and, 250, 257
Amyl alcohol, paper chromatography

and, 253
Amylase, 562

1,5-Anhydro-D-arabitol, 45
2,5-Anhydromannose, indole-hydrochlo-

ric acid reaction and, 297
Anhydroribitol, 45, 47
1,5-Anhydro-D-xylitol, 45
Animal tissue,
nucleoprotein isolation from, 375-376
pentose nucleic acid,
isolation, 387-389
nucleotides, 396-402
Anisotropy, 533
form, 534-535
Antabuse, see Tetraethylthiuram di-
sulfide
Anthrone reaction,

deoxyribose and, 54
Antimetabolites, incorporation of, 557

INDEX

659

Antimony trichloride, infrared spectros-
copy and, 546
Apiose, 95
Apurinic acid,

c,ysteine reaction and, 291

cysteine-sulfuric acid reaction and, 295

degradation, 367-368
alkaline, 444-445

deoxyribonuclease and, 367

diphenylamine and, 287

formation of, 196

indole-hydrochloric acid reaction and.
298

molar absorptivities and, 521

molecular weight of, 443

preparation of, 341-342

properties, 342-344

pyrimidines and, 344

ribonuclease and, 559

structural studies on, 366-368
Arabinal, 17

cysteine reaction and, 293-294, 299

cysteine-sulfuric acid reaction and, 295

deoxyribose synthesis and. 49-50

diphenylamine and, 288

indole-hydrochloric acid reaction and,
298

tryptophan-perchloric acid reaction
and, 296
Arabinose, 10, 11-12, 55, 140

deoxyriljose synthesis from, 50, 53

indole-hydrochloric acid reaction and,
298

ribose synthesis and, 16-17
Arabinose-5-phosphate, 35, 42

metabolism of, 41
D-Arabofuranosyl adenine, nucleoside

hydrolase and, 606
Arabonic acid, 21

ribose synthesis and, 16
Arginine, 575

nucleoprotein formation and, 530-531
Arsenate,

intestinal phosphatase and, 589

nucleoprotein isolation and, 313

nucleoside phosphorylase and, 603

3'-nucleotidase and, 593
Arsenite, deoxyribonuclease and, 582
Arsenotungstic acid, pyrimidines and,
305

5-Arylazopyrimidines, purine synthesis

and, 131
Ascorbic acid,

indole-hydrochloric acid reaction and,
298

transforming activity and, 339

tryptophan-perchloric acid reaction
and, 296

xanthine oxidase and, 613
ATP, see Adenosine triphosphate
2-Azaadenine,

biological activity of, 107

synthesis of, 107

xanthine oxidase and, 612
8-Azaadenine, 106

deamination of, 118
8-Azaguanine,

biological activity of, 106-107

deamination of, 118, 598-599

fluorescence of, 247

incorporation of, 107
8-Azaguanine riboside, 36
8-Azaguanylic acid,

electrophoresis of, 277

isolation of, 107

paper chromatography and, 257
2-Azahypoxanthine, 107
8-Azahypoxanthine, 106

thiation of, 135
8-Aza-6-mercaptopurine, synthesis of,
135

B

Bacteriophage, 346

deoxyribonucleic acids of, 61

pyrimidines of. 88. 89, 90, 136, 188
Barbital, 103

barbiturase and, 623
Barbiturase, properties of, 622-623
Barbituric acid,

chloro derivative, 127

formation, 622

polarography and, 120

synthesis of, 103

uracil oxidase and, 622
Barium, deoxypentose nucleic acid and,

527
Benzimidazole, dissociation constants of,

456
Benzimidazole deoxyriboside, prepara-
tion of, 57

660

INDEX

Benzylcytidylic acid, synthesis, enzy-
matic, 566
Birefringence,

anisotropy and, 533, 534-535

cytoplasm and, 538-539

deoxypentose nucleic acid, 483-484,
485, 487, 489, 490, 492

dichroism and, 533
Biuret,

formation from cytosine, 87, 120

pentose nucleic acid and, 394
Borate,

ion-exchange and, 235-241

nucleoprotein extraction and, 314-315

5'-nucleotidase and, 592

nucleotide separation and, 227

paper chromatography and, 251

uricase action and, 616, 618

xanthine oxidase and, 612
British Anti-Lewisite, uricase and, 619
Bromine, thymine and, 121
5-Bromocytosine, color tests for, 123
5-Bromo-4,6-diaminopyrimidine, x-ray

diffraction and, 448-449
5-Bromo-5-hydroxyhydrothymine, for-
mation from thymine, 121
5-Bromouracil, 12, 370

incorporation of, 107, 347

production from cytidine, 141

thymidine phosphorylase and, 605
5-Bromouridine, 144
n-Butanol, paper chromatography and,

249, 251-253, 254, 255
Butyric acid, paper chromatography
and, 250

Caffeine, 120, 138
Calcium,
deoxypentose nucleic acid and, 526-527
deoxyribonuclease and, 581
nucleic acid separation by, 332
Calcium D-gluconate, ribose synthesis

and, 17
Carbazole reaction, deoxyribose and, 54
Carbazole-sulfuric acid,

reaction with deoxyribonucleic acid,
comparative value of, 300
specificity of, 298-299
5-Carbethoxycytosine, synthesis of, 129

5-(Carbethoxymethylidene)hydantoin,
orotic acid and, 101

5-Carbethoxyuracil, pyrimidine syn-
thesis and, 129

Carbon tetrachloride, electrophoresis
and, 274, 275

5-Carboxymethylhydantoinase, proper-
ties of, 621

2-Carboxymethylmercapto-adenine, syn-
thesis of, 133

Carp muscle, nucleotropomyosin from,
376

Carragheen moss, carbohydrate, di-
phenylamine and, 289

Catalase, uricase and, 616

Centrifugation, fractional, nucleoprotein
isolation and, 376-377

Cetyltrimethylammonium bromide, nu-
cleic acid separation by, 333, 391

Chelating agents, 614

nucleoprotein isolation and, 313

Chloroacetic acid, reaction with mer-
captopyrimidines, 128

Chloroacetophenone, deoxyribonuclease
and, 582

2-Chloroadenine, 105

2-Chloro-4 , 6-dimethylpyrimidine,
dichroism of, 538
thiouronium salt of, 128

Chloroform, 328

Chloroform-amyl alcohol, deproteiniza-
tion by, 324

Chloroform-octyl alcohol, deoxyribo-
nucleate and, 325

Chloroformate, purine synthesis and, 132

p-Chloromercurybenzoate, ribonuclease
and, 572

6-Chloropurine,

reaction with thiourea, 130
synthesis of, 130

Chloropyrimidines, reactions of, 127

8-Chloroxanthine, synthesis of, 130

Chromatography,
deoxypentose nucleic acid composition

and, 349, 350
ion-exchange, 6-7
"development", 216
"displacement", 216
"gradient development", 216
nucleotides, 396
sorption and elution, 215-216

INDEX

661

paper, 6

choice of filter paper, 244-245
general technique, 244-245
nucleotides, 395
phosphate esters, 245
solvent systems, 248-257
troughs for, 244
"Chromonucleic acid," see Deoxypentose

nucleic acid
Citrate,
deoxyribonuclease and, 582
nucleoprotein isolation and, 313
Clupeine, nucleic acid combination, 531
Cobalt,

deoxyribonuclease and, 581
Coenzyme A, 167, 435
hydrolysis of, 184-185
3 '-nucleotidase and, 592-593
phosphate position in, 185-186
ribose in, 13

succinyl, adenosine triphosphate and,
626
Coenzyme I, see Diphosphopyridine

nucleotide
Coenzyme II, see Triphosphopyridine

nucleotide
Coenzyme III, probable structure of, 182
Coenzymes, nucleotides in, 177-187
Colchicine, deoxyribonuclease and, 582
Collidine, paper chromatography and,

253
Convicine, properties of, 100
Copper, 614

deoxyribonuclease and, 582
ribonuclease and, 572
xanthine oxidase and, 613
Cordycepin, 125
adenine in, 95
structure of, 159
Cordecepose, 188

structural relations of, 95
"Cores", see Polynucleotides, limit
Cortisone, phosphodiesterase and, 588
Cozymase 27, see also Diphosphopyridine

nucleotide
Creatine, indole-hydrochloric acid reac-
tion and, 298
Creatine phosphate, separation from

nucleotides, 224
Crotonoside, see also Isoguanosine

isolation and structure, 97, 158-159
synthesis of, 151
Cucumber virus,
isolation of, 377
pentose nucleic acid,
isolation, 389-390
nucleotides, 405
Cuprous salts, purine, 124
Cyanide, 614
3'-nucleotidase, 593
phosphodiesterase and, 588
uricase and, 619
xanthine oxidase and, 612
Cyanoacetal, cytosine synthesis and, 126
Cyanoacetate, purine synthesis and, 133
Cyanoacetic acid, pyrimidine synthesis

and, 125, 126, 129
Cyanoacetylurea, uracil synthesis and,

126
Cysteic acid, 182
Cysteine,

deoxyribonuclease and, 582
3'-nucleotidase and, 593
phosphodiesterase and, 588
Cysteinesulfinic acid, oxidation of, 182
Cysteine-sulfuric acid reaction,
comparative value of, 299
deoxypentose nucleic acid and, 299
interfering substances, 295-296
mechanism of, 293-294
procedures, 291, 294
specificity of, 291-293, 294-295
deoxyribose and, 54
ribonucleic acid and, 302-303
Cystine, 574, 575
Cytidine, 100, 120, 157, 169, 594
bond lengths of, 452-454
borate complex, electrophoresis of, 283
carbohydrate in, 141
cationic exchangers and, 218, 219
deamination of, 143-144, 599
dissociation constants of, 459
estimation, spectrophotometric, 200-

202
glycosidic linkage of, 143-144
ionization of, 272
oxidation of, 12
paper chromatography of, 255
phosphate linkages, 413-415
phosphorolysis of, 604
structure,

662

INDEX

crystal, 462-463
x-ray diffraction, 452, 454
synthesis of, 155-156
trityl derivative of, 146
ultraviolet absorption of, 508, 515
uridine hydrolase and, 606
uridine phosphorylase and, 606
Cytidine benzyl phosphate h,
nucleases and, 432
ribonuclease and, 430
Cytidine diphosphate,
cysteine reaction and, 292
diesterase and, 434
ion-exchange of, 230-231
production of, 434-435
ultraviolet absorption of, 513
Cytidine methyl phosphate a, ribo-
nuclease and, 430
Cytidine-2'-phosphate, ultraviolet ab-
sorption of, 512-514
Cytidine-2',3'-phosphate, see also Nu-
cleotides, cyclic
ribonuclease and, 425
synthesis, enzymatic, 566
Cytidine-3'-phosphate,
deamination of, 425
esters, ribonuclease and, 565
identification of, 62
ribonuclease and, 425, 562
ultraviolet absorption of, 512-514
Cytidine-5'-phosphate, 62
esters, ribonuclease and, 565
ion-exchange of, 241
proof of structure, 164-165
synthesis of, 174
ultraviolet absorption of, 513
Cytidylic acid, 4, 120, see also Cytidine
phosphates
apurinic acid and, 367-368
deamination of, 260, 599
dissociation constants, 459-460
esters, hydrolysis of, 417
ion-exchange of, 222, 223, 224
ionization of, 268-269
isolation of, 169
isomers, 38

dissociation constants of, 460-461
hydrolysis of, 169
identification of, 169
infrared absorption and, 545
ion-exchange of, 226-228

phosphate position in, 189
separation of, 169
ultraviolet absorption of, 497, 551
paper chromatography of, 255-256, 257
pentose nucleic acid,
cytoplasmic, 400
liver, 397-398
ratios, 407
virus, 405
yeast, 402
Cytidylyl-cytidine, synthesis, enzymatic,

566
Cytochrome, 15
Cytoplasm,
birefringence of, 538-539
fractions, pentose nucleic acid, 400-
401, 406
Cytosine, 3, 14, 90, 138, 140, 346, 622
amino group of, 448
antibiotics and, 135
cationic exchange of, 218
color tests for, 123, 305
deamination of, 118, 195
dihydroorotic acid dehydrogenase and,

621
discovery of, 87

dissociation constants of, 456, 457
glutamate and, 600
ionization of, 268

nucleohistone fractionation and, 362
nucleoside phosphorylases and, 605
oxidation of, 120, 121
paper chromatography, 249, 250, 253,

254
reduction of, 119, 120
stability to acids, 117
synthesis of, 87-88, 126, 127-128
thymidine phosphorylase and, 605
thymine estimation and, 304
thymic acid and, 341
ultraviolet absorption of, 499, 502, 504
uracil oxidase and, 622
Cytosine deaminase, specificity of 599
Cytosine deoxyribodiphosphoric acid,

formation of, 197
Cytosine deoxyriboside, 48, 57, 139, 140,
see also Deoxycytidine
ultraviolet absorption of, 504, 508, 516
Cytosine deoxyriboside deaminase, 14
Cytosine deoxyribose diphosphate, 172
paper chromatography of, 255

INDEX

G63

Cytosine deoxyriboside phosphate, 171
2-Cytosine-thiol, synthesis of, 127

Daraprim, 104
Deaminase, 142
Deaminoribonucleic acid, 557

ribonuclease and, 559
Deoxyadenosine, .sfe Adenine deoxyri-
boside
Deoxyadenylic acid, 441, 578
cysteine reaction of, 291
deamination of, 597
paper chromatography of, 257
2-Deoxy-D-arabinose, 10
Deoxycytidine, see also Cytosi de-
oxyriboside
cysteine reaction of, 292
infrared spectrum of, 62
Deoxycytidine dinucleotide, hj-drolysis,

enzymatic, 442
Deoxycytidine-2'-phosphate, ultraviolet

absorption of, 551
Deoxycytidine -3'-phosphate, 62
Deoxycytidine-3',5'-diphosphate,
deoxyribonucleic acid and, 440-441
production, mechanism of, 443
Deoxycytidine phosphate, see alsu De-
oxycytidylic acid,
synthesis, 189
Deoxycytidine-5'-phosphate,
deoxyribonucleic acid digests and, 441
ion-exchange of, 241
Deoxycytidylic acid, 441, 578, 579, see
also Deoxycytidine phosphates
C3'steine reaction and, 292
diphenylamine and, 288
electrophoresis of, 277
ion-exchange of, 223-224, 227
paper chromatography of, 257
ultraviolet absorption of, 513
5'-Deoxy-5'-ethylthioadenosine, accu-
mulation of, 188
3-Deoxy-D-gluconate,

deoxyribose synthesis and, 52
3-Deoxy-D-glucose, 61

deoxyribose synthesis and, 52
Deoxyguanosine, see also Guanine de-
oxyriboside
phosphorolysis of, 602, 604

tryptophan -perchloric acid reaction
and, 296
Deoxyguanylic acid, 171, 441, 579
cysteine reaction and, 291
paper chromatography of, 257
2-Deoxyhexoses, cysteine reaction and,

293"
Deoxyinosine, see also Hypoxanthiue
deoxyriboside
ion-exchange of, 219
phosphorol3sis of, 602
Deoxy-5-methylcytidylic acid, 171, 579
electrophoresis of, 277
ion-exchange of, 223-224, 227
paper chromatography of, 257
ultraviolet absorption of, 520
5'-Deoxy-5'-methylthioadenine, see Ade-
nine thiomethylpentoside
5-Deoxy-5-methylthioribose , proof of

structure, 157
Deoxymononucleotides, intestinal phos-
phatase and, 589
Deoxynucleoproteins, definition, 309
Deoxynucleoside diphosphates,
formation of, 172
position of phosphate in, 172
Deoxypentose nucleic acids, see also
Deoxyribonucleic acids
acid-base properties of, 475-480
AT type, 350
birefringence of, 539
chain-branching of, 479, 490
coiling of, 484, 490
composition,
elementary, 333-334
general, 348-349
procedures, 350
purine distribution, 350-358
pyrimidine distribution, 350-358
constituents, 345-348
carbohydrate, 346
nitrogenous, 346-347
unidentified, 347-348
deformation, 487, 489-490, 491-492
hydrogen-bonding and, 489
phosphate charges and, 489
viscosity and, 489-490
degradation, 339-340, 340-345
degraded, molar absorptivity of, 552
denaturation, 336, 337-339

664

INDEX

heat and, 528
measurement of, 526-527
deproteinization,
chloroform, 323-324
sodium chloride, 325-326
dichroism,
humidity and, 539--541
infrared, 550
stained, 544-545
digests,
intestinal phosphatase and, 589
phosphodiesterase and, 586
dimensions of, 489-490
diversity of, 351-358
enzyme combination with, 321
fractionation,
general, 358-361
of protein nucleates, 361-366
theories regarding, 365-366
GC type, 350

hydrogen-bonding in, 480, 489, 490
hydrogen-ion concentration and, 490-

492
hydrolysis of, 196-197
formic acid, 259
hydrochloric acid, 258
perchloric acid, 258
products, 191-192
hyperchromic effect, 339
infrared absorption of, 548-549, 553
integrity, 321-322

standards of, 333-337
inter-nucleotide distance, 461
intestinal phosphatase and, 589
isolation of, 321-333

anionic detergents, 327-328
comparison of procedures, 328-329
microorganisms, 330-332
miscellaneous, 329-330
removal of impurities, 332-333
strong salt solution, 323-326
water, 326-327
light scattering and, 484, 487, 489, 490,

491-492
localization of, 6

mammalian, composition of, 352-355
microbial, composition of, 356-358
models,
Astbury, 461-462, 539
Furberg, 462, 464
Pauling and Corey, 463-464

Watson and Crick, 371, 464-470, 480,
541,548
moisture content, 334, 552
molecular weight determination, 337
dielectric dispersion, 473
diffusion coefficient, 471-472
light-scattering, 472^73
ultracentrifugation, 470-471
nucleotide sequence, 366
optical activity, 336-337
paracrystalline form of, 468, 552
phosphate groups, position of, 463
phosphodiesterase and, 587
protein combination of, 311-312, 320-

321
purity, criteria of, 322-323
sedimentation of, 491
solution properties,
ionic strength, 483-488
pH effects, 490^92
size and shape, 488-490
stabilization, cations and, 527-528
structure, 366-368
helical, 463-464, 468^70
pattern, 461-462
similarities, 369-371
titration, 490

anomalies, 476^77, 522-525
transforming activity, 328, 331
ultraviolet absorption,

deoxyribonuclease and, 519
electrolytes and, 488
viscosity, 489-492
pH and, 479^80
x-ray diffraction of, 461, 467, 469-470,

552
yeast, purification of, 325
Deoxypentose nucleoprotein,
infrared absorption of, 553
D-2-Deoxyribofuranosides, 138
2-Deoxyribonic acid, formation of, 59-60
Deoxyribonuclease, 162, 233, 277, 321,
367, 441, 552
action,
products of, 345, 442
volume changes and, 564
antibodies against, 625
assay,

acid group formation, 579
acid-soluble phosphorus and, 579-580
dye-binding and, 580-581

INDEX

665

optical changes, 580

viscosity changes, 580
definition of, 557
digests,

electrophoresis of, 274

ion -exchange of, 235
enolic groups liberated by, 411-412
excretion of, 624
infrared absorption and, 549-550
inhibition of, 316, 390, 582-583, 584-585
intracellular localization of, 624, 625
limit polynucleotide and, 341
microorganisms, 584-585
molar absorptivities and, 519
nucleoprotein and, 319
pancreas,

activators, 581-582

assay, 579-581

historj', 576-577

hydrogen ion concentration and, 583

inhibitors, 582-583

kinetics, 581

protein properties, 583

specificity, 577-579

stability, 582
plant, 584

thymus, properties of, 584
ultraviolet absorption and, 339
Deoxyribonuclease I, sec Deoxyribo-

nuclease, pancreas
Deoxyribonucleic acids, 3, see also De-

oxypentose nucleic acids
acid hydrolysis, mechanism of, 444
alkali,

hydrolysis, 419-420

stability, 416
bacterial, removal of, 391
chain-branching of, 439-440
color reactions for, 294

carbazole-sulfuric acid, 298-299

cysteine-sulfuric acid, 290-296, 303

diphenylamine, 287-290

evaluation of, 299-300

indole-hydrochloric acid, 297-298

Schiff's reagent, 299

tryptophan-perchloric acid, 296-297
composition, electrophoresis and, 276
digests, dialysis of, 443
electrometric titration of, 410, 411
estimation of, 290, 294, 297-298, 303
molar absorptivity,

miscellaneous effects, 529

nucleic acid estimation and, 529-530

nucleoprotein and, 530-532

pH and, 522-525

salts and, 525-529
nucleotide sequence in, 442—145
phloroglucinol reaction and, 302
protein complex, nature of, 309-310
removal of, 387
ribonuclease and, 559
separation from pentose nucleic acid,

279
structure of, 439-445
Deoxyribonucleosides, 604, see also De-

oxyribosides
anion exchange of, 221
biosynthesis of, 58
borate complexes, electrophoresis of,

283
chemical properties of, 57
chromatography of, 49
formation of, 56-57
ion-exchange of, 219
phosphorolysis of, 603
Deoxyribonucleotides, ion-exchange of,

223-224
2-Deoxy-D-ribose, 3
acid and, 59
biosynthesis of, 59
chemistry of, 48-60
derivatives, properties of, 75-78
dismutation of, 288-289
esters of, 60
N-glycosides of, 56-57
0-glycosides of, 55-56
hydrazines and, 60
identification of, 53-54, 141-142
isolation of, 48-49, 142
mutarotation of, 55
nomenclature, 10-11
nucleic acid and, 346
occurrence of, 14, 346
orcinol reaction and, 301
paper chromatography of, 263-264
phosphates of, 58-59
synthesis of, 47, 53, 61
3-Deoxyribose,
derivatives, 51

paper chromatography of, 264
4-Deoxyribose, paper chromatography

of, 264

666

INDEX

Deoxyribose nucleotides, see also Deoxy-
ribotides
formation of, 171
position of phosphate in, 171-172
2-Deoxyribose-l -phosphate,
formation of, 58, 63
thymidine formation and, 63
thymidine phosphorylase and, 605
2-Deoxyribose-5-phosphate, 49

formation of, 58
Deoxyribosides, see also Deoxyribo-
nucleosides
configuration of, 14-15
paper chromatography of, 247-248, 253,

254, 264
purine, cysteine reaction and, 292
ultraviolet absorption of, 504
Deoxyribotides, see also Deoxyribo-
nucleotides
purine,

carbazole-sulfuric acid reaction and,

298
diphenylamine and, 287
indole-hydrochloric acid reaction
and, 297
pyrimidine,

carbazole-sulfuric acid reaction and,

298
diphenylamine and, 288
tryptophan-perchloric acid reaction
and, 296
Deoxysugars, chromatography of, 54
2-Deoxyxylose, 142

deoxypentose nucleic acid and, 346
diphenylamine and, 287
formation of, 53
3-Deoxyxylose, diphenylamine and, 288,

289
2, 3-Deoxyxylose, diphenylamine, 288
Desoxyribose, see Deoxyribose
Detergents, anionic, deoxypentose nu-
cleic acid and, 327-328
2,6-Diacetamidopurine, chloromercury

salt in nucleoside synthesis, 151
2',3'-Diacetyladenosine,

nucleotide synthesis and, 173
structure of, 145-146
4,5-Diacetylaminouracil, 119
1,3-Dialdehydes, pyrimidine synthesis

and, 125
Dialuric acid, 123

Dialysis, nucleic acid purification by, 333
1,3-Diaminobutane, 119
2,4-Diamino-5,6-dihydroxypyrimidine,
188
properties of, 100
4 , 5-Diamino-2 , 6-dihydroxypyrimidine,

uric acid synthesis and, 132
4,6-Diamino-5-formamidopyrimidine,
dehydration of, 133
isotopic, reaction with formamide, 133
4,6-Diamino-5-formylaminopyrimidine,

adenine synthesis and, 132
4,5-Diamino-6-glycosylaminopyrimi-

dine, cyclization of, 152
4 , 5-Diamino-6-hydroxypyrimidine,
hypoxanthine synthesis and, 132
synthesis of, 128
4,5-Diamino-6-methylpyrimidine, 106
4,6-Diamino-2-methylthiopyrimidine,

adenosine synthesis from, 153
2 , 6-Diaminopurine,

biological activity of, 104-105
deamination of, 118
incorporation of, 107
reducing agents and, 120
synthesis of, 132, 133
2, 6-Diaminopurine-9-ribofuranoside, nu-
cleoside hydrolase and, 606
2, 6-Diaminopurine riboside, phosphory-
lation of, 594
2 , 6 (4) -Diaminopyrimidine ,
derivatives, as antimetabolites, 104
reduction of, 120
synthesis of, 126
4 , 5-Diaminopyrimidines ,
pteridines and, 132
purine synthesis and, 132-133
synthesis of, 128, 131
4,6-Diaminopyrimidine,
reduction of, 120

2-substituted, in purine synthesis, 131
Diazobenzenesulfonic acid, color tests

and, 123
5,5-Dibromo-4-hydroxyhydrouracil, py-
rimidine color tests and, 123
1,3-Dicarbamylurea, 136
4,5-Dicarbomethoxyiminazole, xantho-

sine synthesis from, 154
2,6-Dichloroadenine, adenosine syn-
thesis and, 151

INDEX

667

2,8-Dichloroadenine nucleosides, syn-
thesis of, 150
2,4-Dichlorobenzenediazonium chloride,

131
2,8-Dichloro-6-diethylaminopurine, syn-
thesis of, 130
2,5-Dichloropyrimidine, reaction with

thiourea, 128
2,6-Dichloropyrimidine, amination of,

127
Dichroism,

orientation and, 536-537
ultraviolet, 532-545, 552
Diesterase, 277, 442

action, mechanism of, 435
deoxyribonucleic acid digests and,

441-442
ribonucleic acid branching and, 434
trinucleotide structure and, 429
Diethanolamine, purine synthesis and,

133
2,6-Diethoxy-5-methylpyrimidine, nu-
cleoside synthesis and, 156
2,6-Diethoxypyrimidine, nucleoside sjn-

thesis and, 155-156
6-Diethylaminopurine, synthesis of, 130
5,5-Diethylbarbituric acid, see Barbital
Diethjdene glycol, paper chromatogra-
phy, 251-253
Digitoxose, cysteine reaction and, 293
Dihydrocytidylic acid, hydrolysis of, 169
4,5-Dihydro-6-deoxyuric acid, see Purone
Dihydroisobarbituric acid, formation

from pyrimidines, 121
Dihydroorotase, properties of, 621
Dihydroorotic acid,

bacterial growth and, 621
hj'drolysis of, 621
Dihydroorotic acid dehydrogenase, prop-
erties of, 620-621
4,5-Dihydropyrimidine glycosides, hj^-

drolj'sis of, 141
Dihydrothymine, 172
degradation of, 621-622
uracil oxidase and, 622
Dihydrouracil,

degradation of, 621-622
formation of, 119
occurrence of, 622
uracil oxidase and, 622
Dihydrouridine, hydrolysis of, 141

Dihydrouridylic acid, hydrolysis of, 170
2,8-Dihj'droxyadenine,
deposition of, 96
formation of, 595
2,6-Dihydroxy-4,5-dihydropyrimidine,

see Hydrouracil
2,6-Dihydrox3'-4-methylpyrimidine, 86
2,6-Dihydroxy-5-methylpyrimidine, see

Thymine
2,4-Dihydroxypyrimidines, infrared ab-
sorption of, 547
2,6-Dihydroxypyrimidine-4,5-dione, see

Alloxan
/3-Diketones, pyrimidine synthesis and,

125
Diketopiperazine, bond lengths in, 452
Dimethoxysuccinic acid, 42, 144
6-Dimethylaminopurine, occurrence of,

95, 160
Dimethylaniline, 127, 130
5,6-DimethyIbenzimidazole, 95
5,6-Dimethylbenzimidazole deoxyribo-
side,
hydrolysis of, 57
preparation of, 57, 62
5,6-Dimethylbenzimidazole riboside,
hj'drolysis of, 57
isolation and sj-nthesis of, 31-33
1 ,4-Dimethyl-5-chloromethyluracil, prep-
aration of, 92
2,4-Dimethylpyrimidine, synthesis of,

130
1,4-Dimethyluracil, 92
4,5-Dimethyluracil, 90
2',3'-DimethyIuridine, 146
Dinitrophenyl phosphate, ribonuclease

and, 559-560
Dinucleotide pyrophosphatase, coen-
zymes and, 182, 185
Dinucleotides,

deoxyribonuclease and, 579
electrophoretic mobilities of, 270-271
molar absorptivities and, 520-521
nucleic acid hydrolysis and, 196
paper chromatography of, 257
phosphodiesterase and, 586
ribonuclease and, 427
Dioxane, paper chromatography and, 253
Diphenylamine reaction,
apurinic acid and, 343
comparative value of, 299

668

INDEX

deoxypentose nucleic acid,

interfering substances, 289-290
mechanism of, 288-289
procedure, 287
specificity, 287-288

deoxypentoses and, 53

pentose nucleic acid and, 394
Diphenyl phosphate,

3'-nucleotidase and, 593

phosphodiesterase and, 585-586, 588

ribonuclease and, 559-560
Diphosphates, nucleoside, ion-exchange

of, 230-232
3',5'-Diphosphates, synthesis of, 176
Diphosphokinase, transphosphorylation

and, 626
Diphosphopyridine nucleotide, 597

analog, isonicotinic hydrazide and, 609

deamination of, 596

dihydroorotic acid dehydrogenase and,
620

hydrolysis of, 40, 181-182

ribose in, 13
Diphosphopyridine nucleotide hydro-
lase,

animal, exchange reaction of, 608-609

Neurospora, inhibition of, 608
Diribosylamine, 29
Dissociation constants,

nucleosides, 458-461

nucleotides, 458-461

purines, 455-458

pyrimidines, 455-458
Distribution coefficient,

ion-exchange, 213-214

mononucleotides, 221-223
Dithioformate, purine synthesis and, 132
Dithioformic acid, purine nucleoside

synthesis and, 152
Dithiothymine, 104
Diureidomalonic acid, 617
Diuridine-5',5'-diphosphate,

stability of, 420

synthesis of, 176
Divicine, 159

structure of, 100
DNA, see Deoxypentose nucleic acid
DPN, see Diphosphopyridine nucleotide
Dyes, deoxypentose nucleic acid and, 337

Egg albumin, nucleic acid and, 320-321
Electrolytes, deoxypentose nucleic acids

and, 484-488
Electrophoresis,
buffers and, 275
endosmosis and, 276
general considerations, 267-268
microtechnique, 284
nucleic acid separation by, 332
paper, nucleotides and, 396
Endosmosis, electrophoresis and, 276
D-£/r?/</iro-2-Deoxypentose, 11
Erythrose, 10
cysteine reaction and, 293
deoxyribose synthesis and, 50
Ethanol,

deoxypentose nucleic acid absorp-
tivity and, 529
paper chromatography and, 253, 255,

264
pentose nucleoprotein separation and,
380
Ethionine, 188
/3-Ethoxyacrolein acetal, 125
Ethyl acetate, paper chromatography

and, 253
Ethylenediamine tetraacetate, nucleo-
protein isolation and, 313
Ethylmercurithiosalicylate, nucleopro-
tein extraction and, 314
5-Ethyl-5-phenylbarbituric acid, see

Phenobarbital
Ethyl riboside, 13-14,24
2-Ethylthioadenine, 105

FAD, see Flavin adenine dinucleotide
Feulgen reaction, 6
Fiber, axis, absorptivity and, 533
Flavin adenine dinucleotide,

hydrolysis of, 183

ribose in, 15

synthesis of, 184

xanthine oxidase and, 610
Flavin mononucleotide, 15
Fluoride,

deoxyribonuclease and, 582

INDEX

669

S'-nucleotidase and, 592
ribonuclease and, 572
FMN, see Flavin mononucleotide
Folic acid,
adenine estimation and, 304
antagonists, 104
reduction of, 120
Folinic acid, antagonists, 104
Formaldehyde, phosphodiesterase and,

588
Formamide, purine synthesis and, 133,

135
Formamidine, purine synthesis and, 133
Formate, nucleotide separation by, 225-

226, 227-228
Formic acid,
formation,

adenine and, 118
pyrimidines and, 120, 121
nucleic acid hj'drolysis by, 195, 197
paper chromatography and, 251
purine synthesis and, 132, 133
Formylacetic acid, pyrimidine synthesis

and, 125
Formylglyoxylurea, formation from ura-
cil, 120
N-Formylmorpholine, purine synthesis

and, 133
6-Formyl pteridine, guanase and, 598-599
Fructose, cysteine reaction and, 293
Fructose-l ,6-diphosphate, 5'-nucleo-

tidase and, 592
Furfural, 141
cysteine reaction and, 293-294
diphenylamine and, 288
tryptophan -perchloric acid reaction
and, 296
Furfuryl alcohol,

cysteine reaction and, 293-294
cysteine-sulfuric acid reaction and, 295
diphenylamine and, 288
indole-hydrochloric acid reaction and,

298
trj^ptophan-perchloric acid reaction
and, 296

9-D-Galactofuranosyl-2-methj'lthioade-
nine, synthesis of, 154

Galactose, cysteine-sulfuric acid reac-
tion and, 303
Galactose-1 -phosphate, 98
Galactowaldenase, 13

coenzyme of, 98, 186
Galacturonic acid, indole-hj'drochloric

acid reaction and, 298
Gelatin, birefringence of, 535
Globin, nucleic acid fractionation and,

363
D-Gluconic acid 6-phosphate, ribose for-
mation from, 41
D-Glucopyranosyladenine,
nucleoside hydrolase and, 606
synthesis of, 150, 152-153
3-/3-D-Glucopyranosylcytosine, reaction

with periodate, 148-149
9-D-Glucopyranosylguanine, synthesis

of, 150
9-D-Glucopyranosylisoguanine, sj-nthesis

of, 152-153
3-/3-D-Glucopyranosyluracil, reaction

with periodate, 148-149
Glucosamine, indole-hydrochloric acid

reaction and, 297, 298
Glucose,
cysteine-sulfuric acid reaction and, 303
dissociation constant of, 458
orcinol reaction and, 302
ribose synthesis and, 18
vicine and, 99
Glucose-l,6-diphosphate, 37
Glucose-1 -phosphate, 98
Glucose-3-phosphate, dissociation con-
stants of, 459
Glucose-6-phosphate, 5'-nucleotidase

and, 592
Glucosone, indole-hydrochloric acid re-
action and, 297
Glucosyladenine, 152

configuration of, 148
3-D-Glucosylcytosine, synthesis of, 155
3-D-Glucosyluracil, synthesis of, 155
Glutamic acid

diphenylamine and, 289-290
formation, purines and, 600
D-Glutamic acid, nucleotide derivatives

and, 99
Glutamine, purine amination and, 600

670

INDEX

Glutathione,

adenosine kinase and, 594

3 '-nucleotidase and, 593
Glyceraldehyde, deoxyribose biosynthe-
sis and, 59
Glyceraldehyde-3-phosphate,

conversion to lactic acid, 438

pyruvaldehyde and, 438
Glycerol, deoxypentose nucleic acid and,

529
Glycerol-a-methyl hydrogen phosphate

hydrolysis of, 415
Glycerol-a-methyl phosphate, hydroly-
sis, mechanism of, 416
Glycerol phosphates,

alkali and, 415

interconversion of, 414

3'-nucleotidase and, 593
L-a-Glycerylphosphorylcholine, ribonu-

clease and, 559
L-a-Glycerylphosphorylethanolamine, ri-

bonuclease and, 559
Glycine,

bond lengths in, 452

diphenylamine and, 289-290

formation from purines, 117-118

inosinic acid transformylase and, 614

nucleoprotamine isolation and, 316
Glycocyamine, formation from guanine,

117
Glycolic aldehyde,

cysteine reaction and, 293

cysteine-sulfuric acid reaction and, 295

diphenylamine and, 288
6-Glycosylaminopurine, 152
Gout, guanine and, 95-96
Guanase, 96, 97, 609

assay of, 599

inhibition of, 598-599

intracellular localization of, 625

properties of, 599

specificity, 598-599
Guanazolo, see 8-Azaguanine
Guanidine,

formation from guanine, 85, 122

nucleoprotein and, 319, 380

pentose nucleic acid isolation and,
387-388

purine synthesis and, 133

pyrimidine synthesis and, 125-126

ribonuclease and, 386

Guanidoimidazole, formation from gua-
nine, 117
Guanine, 2, 3, 14, 82, 120, 346

bond distances and angles of, 451-452

cationic exchange of, 218

color reaction of, 124

conversion to xanthine, 117

cytosine ratio, 352-355

deamination of, 118, 131, 260

2,6-diaminopurine and, 105

discovery of, 85

dissociation constants of, 456, 457
resonance and, 457-458

dissociation of, 114, 268

fluorescence of, 247

formic acid and, 259

glutamate and, 600

gout and, 95-96

hydrogen atoms of, 452

internucleotide linkages and, 412

keto group of, 452

nucleohistone fractionation and, 362

occurrence of free base, 94, 95-96

oxidation of, 121-122

paper chromatography of, 251, 253,
254-255

pyrimidine nucleoside phosphorylases
and, 604

spectrophotometric estimation of, 200-
202

stable suspensions of, 599

stability to acid, 117-118

synthesis of, 132, 133

tautomers of, 452

ultraviolet absorption of, 498-499, 500.
502
Guanine deoxyriboside, 139, 140, see also
Deoxyguanosine

glycosidic linkage of, 143, 147

hydrolysis of, 48-49, 141

phosphorolysis of, 58

ultraviolet absorption of, 504
Guanine picrate, 124
Guanopterin, see Isoguanine
Guanosine, 19, 36, 120, 158, 167, 594, 604

borate complex, electrophoresis of, 283

conversion to xanthosine, 142, 598

discovery of, 140

dissociation constants of, 459

glycosidic linkage of, 143

hydrolysis of, 606

INDEX

671

ionization of, 272

methylation of, 42

phosphate linkages, 413-415

phosphorolysis of, 602, 603

structure, x-ray diffraction and, 455

synthesis of, 151

thymidine phosphorylase and, 605

titration anomalies and, 523

trit}'! derivative of, 146

ultraviolet absorption of, 505, 508, 511

uridine hydrolase and, 606

uridine phosphorj-lase and, 606
Guanosine diphosphate mannose,

isolation of, 190

occurrence of, 60
Guanosine-2'-phosphate,

electrophoresis of, 269

ultraviolet absorption of, 513
Guanosine-2',3'-phosphate, see also Nu-
cleotides, C3'clic,

ribonuclease and, 425
Guanosine-3'-phosphate,

electrophoresis of, 269

ultraviolet absorption of, 513
Guanosine-5'-phosphate,

proof of structure, 164

synthesis of, 174

ultraviolet absorption of, 513
Guanosine phosphates, see also Guanylic
acid

tumors and, 189
Guanosine triphosphate,

formation of, 626

isolation of, 189
Guanylic acid, 4, 120

analysis for, 395

cytoplasmic pentose nucleic acid and,
400

deamination of, 598

discovery of, 167

dissociation constants, 459-460

hyperchromic effect and, 519-520

ion-exchange of, 222, 223, 224

ionization of, 268-269

isomers of, 38, 168-169
ion-exchange of, 228
spleen nuclease and, 432

paper chromatography of, 255-256, 257

pentose nucleic acid
liver, 397-398
pancreas, 398

ratios, 407

thymus, 398

yeast, 402
ribonuclease digests and, 407, 571
ultraviolet absorption of, 514, 519

H

5-Halouracils, 104

Heparin, ribonuclease and, 385-386, 572

Heptoses, cysteine-sulfuric acid reaction

and, 294-295
Hexals, cj-steine reaction and, 293
Hexamethylbenzene, dichroic ratios of,

537
Hexammine cobaltic chloride, nucleic

acid purification and, 333
Hexoses,

cysteine-sulfuric acid reaction and,

294-295
diphenylamine and, 288
indole-hydrochloric acid reaction and,
297, 298
Hexuronic acids,
cysteine-sulfuric acid reaction and,

294-295, 303
cysteine reaction and, 293
diphenylamine and, 288
orcinol reaction and, 301
Histone,
diphenylamine and, 289-290
isolation of, 326-327
nucleic acid combination, 320-321, 531
nucleic acid fractionation and, 363
occurrence of, 311-312
Humidity, deoxypentose nucleic acid

and, 539-541
Hydantoin, 621

Hj'drazine, reaction with uridine, 144
Hydrochloric acid, paper chromatog-
raphy and, 249, 253-254, 255
Hydrogen peroxide, uracil and, 120-121
Hydrouracil, 98

Hydroxyacetylenediureidocarboxylic
acid, formation from uric acid, 616-
617
a-Hydroxyacids, cj'steine reaction and,

293
/3-Hydroxyacids, cysteine reaction and,

293
2-Hydroxyadenine, see Isoguanine

672

INDEX

S-Hydroxyadenine, 2,8-dihydroxyade-

nine and, 595
Hydroxy aldehydes, cysteine-sulfuric

acid reaction and, 295
2-Hydroxy-6-aminopurine, see Isogua-

nine
2-Hydroxy-6-aminopyrimidine, see Cyto-

sine
2-Hydroxy-4,6-dimethylpyrimidine,
dichroism of, 537-538
dissociation constants of, 456
hydroxyl bond length in, 449
2-Hydroxyethyl dimethyl phosphate,

hydrolysis of, 415
Hydroxyiminocyanoacetate, purine syn-
thesis and, 132
Hydroxylamine,

deoxyribonuclease and, 582
reaction with thymine, 304
co-Hydroxylevulinaldehyde, 53
o;-Hydroxylevulinic acid,
cysteine reaction and, 293
diphenylamine and, 288
formation of, 59
2-Hydroxy-5-methyl-6-aminopyrimidine,

84
5-Hydroxymethylcytosine, 258, 346, 370,
371
deamination of, 118
discovery of, 88, 89
formic acid and, 259
isolation of, 188, 245
occurrence of, 61, 90, 136, 188, 346
paper chromatography of, 249, 251,

253, 254
perchloric acid and, 258
properties of, 90
stability to acids, 117
ultraviolet absorption of, 500
5-Hydroxymethylcytosine deoxyribo-

side, 189
2,6- bis (Hydroxymethyl) - 4 - hydroxy - 5 -

methylpyrimidine, 91
6-Hydroxymethyl pteridine, guanase

and, 598-599
4-Hydroxy-6-methylpyrimidine, synthe-
sis of, 128
5-Hydroxymethylpyrimidines, synthesis

of, 130
2 - Hydroxy - 4 - methyltetrahydropyrimi -
dine, 119

4-Hydroxymethylthymine, 91

4-Hydroxymethyluracil, 91

5-Hydroxymethyluracil,
preparation of, 92
stability of, 92

2-Hydroxy-5-nitropyrimidine, formation
of, 125

6-Hydroxypteridine, tautomerism of , 115

2-Hydroxypurine, reduction of, 119

6-Hydroxypurine, see Hypoxanthine

8-Hydroxypurine, xanthine oxidase and,
612

2-Hydroxypyrimidines, infrared absorp-
tion of, 547

4-Hydroxypyrimidines, infrared absorp-
tion of, 547
synthesis of, 128

6-Hydroxypyrimidine, reduction of, 120

2 -Hydroxy -4,5,6 - triaminopyrimidine,
isoguanine synthesis and, 133

5-Hydroxyuracil, 104

Hyperchromic effect,

ribonucleic acid digests and, 569-570

Hypoxanthine, 2, 124, 165
adenine oxidation and, 595
cationic exchangers and, 218
discovery of, 97-98

dissociation constants of, 456-457, 458
formation of, 84
infrared absorption of, 547
occurrence of free base, 94
oxidation of, 609, 612
paper chromatography of, 251, 253
reaction with triethylamine, 130
reducing agents and, 119, 120
stability to alkali, 118
synthesis of, 85, 131, 132, 135
ultraviolet absorption of, 498-499, 502
uridine phosphorylase and, 606
xanthine oxidase inhibition and, 612

Hypoxanthine argentipicrate, 124

Hypoxanthine deoxyriboside, 48-49, 157,
see also Deoxyinosine
arsenolysis of, 58-59
formation of, 142
glycosidic structure of, 147
ultraviolet absorption of, 504

Hypoxanthine thiomethylpentoside,

synthesis of, 157-158

Hypoxanthosine, see Inosine

INDEX

673

I

Imidazole, dissociation constant, 456, 458
6-Iminodih3'dropyrimidine, 448
Indole-hj'drochloric acid reaction,

comparative value of, 300

deoxypentose nucleic acid,
mechanism of, 298
procedures, 297-298
specificity of, 297, 298
Infrared absorption,

dichroism and, 550-551

mulled samples, 546

nucleic acids and, 545-551

solvents and, 546

sublimed samples, 546
Inosine, 36, 157, 165, 594

adenine deamination and, 600

dissociation constants of, 459

formation,

adenine and, 602-603
adenosine and, 142

glycosidic linkage of, 143

hydrolysis of, 606

phosphorolysis of, 602, 603, 604

thymidine phosphorylase and, 605

uridine hydrolase and, 606
Inosine diphosphate, transphosphoryla-

tion of, 626
Inosine-5'-phosphate, see also Inosinic
acid

structure of, 162-163

ultraviolet absorption of, 505, 508, 510

synthesis of, 173

ultraviolet absorption of, 512
Inosine triphosphate,

ami nation of, 600

ultraviolet absorption of, 512
Inosinic acid, 39, 40, 597, see also Inosine
phosphate

biosynthesis of, 607

conversion to hypoxanthine, 197

conversion to inosine, 197

isomers, ion-exchange of, 228

origin of, 160
Inosinic acid transformj'lase,

assay of, 614

cofactors of, 614

purine synthesis and, 614
Iodine, reaction with pyrimidines and
purines, 121

lodoacetamide, deoxyribonuclease and,

582
lodoacetic acid, deoxyribonuclease and,

582
5-Iodouracil, 370
incorporation of, 347
thymidine phosphorylase and, 605
Ion-exchange,
anionic, nucleic acid derivatives and,

219-220
cationic, nucleic acid derivatives and,

218-219
equilibria,
distribution coefficient, 213-214
hj'drophilic character, 213
non-polar affinity, 214
rate of reaction, 214-215
Iron,
deoxyribonuclease and, 581
uricase and, 619
Isoamyl alcohol, paper chromatography

and, 256-257
Isobarbituric acid, 83
barbiturase and, 623
formation from uracil, 120-121
uracil oxidase and, 622
Isobutyric acid, paper chromatography

and, 253, 257
Isocytosine,
color tests for, 123
dissociation constants of, 456, 457
polarography and, 120
synthesis of, 125
reducing agents and, 120
Isodialuric acid, 123

formation from uracil, 120-121
Isoguanine,

conversion to xanthine, 117
2,8-dihydroxyadenine and, 595
formation from uric acid, 96
nitrous acid and, 118
occurrence of, 97
reducing agents and, 120
synthesis of, 133, 135
ultraviolet light and, 123
Isoguanosine, see also Crotonoside
isolation of, 13

nucleoside phosphorylase and, 602
Isoleucine, diphenylamine and, 289-290
Isonicotinic acid hj'drazide, diphospho-
pyridine nucleotide and, 609

674

INDEX

Isopropanol, paper chromatography and,
253-254, 255, 257

2',3'-Isopropylidene adenosine, nucleo-
tide synthesis and, 173-174

2',3'-Isopropylidene-3 , 5'-cycloadenosine
iodide, crystal structure of, 462

2,3-Isopropylidene-D-glyceraldehyde, de-
oxyribose synthesis from, 52-53

2',3'-Isopropylidene inosine, nucleotide
synthesis and, 173

Isosbestic points, identification of com-
pounds and, 116

a-Ketoacids, cysteine reaction and, 293
a-Ketoglutaric acid, amination, purines

and, 600
Ketohexoses,

orcinol reaction and, 301-302
tryptophane-perchloric acid reaction
and, 296

Lactic acid, gl3'ceraldehyde-3-phosphate

and, 438
Laevulic acid, see Levulinic acid
Lanthanum,

deoxypentose nucleic acid and, 337

nucleic acid hydrolysis and, 198

nucleic acid purification and, 333

nucleoprotein and, 317
Lauryl amine, paper chromatography

and, 257
Lecithins, 1

hydrolysis of, 415
Leucine, 614
Levulinic acid, 48, 49, 54

diphenylamine and, 288

formation of, 59, 141

indole-hydrochloric acid reaction and,
298

thymine estimation and, 304-305
Lithium arsenotungstate, color tests

and, 123
Lithium chloride, pentose nucleic acid

isolation and, 389
Liver, pentose nucleic acid of, 397-398
L-Lysine, 574, 575

nucleotide derivatives and, 99

Lysozyme, deoxypentose nucleic acid

denaturation and, 529, 531
Lyxose, 12, 141

M

Magnesium,

adenosine kinase and, 594

deoxypentose nucleic acid and, 526-527

desoxyribonuclease and, 578, 581, 582,
584, 585

5'-nucleotidase and, 592

phosphodiesterase and, 588

ribonuclease and, 572
Malic acid, pj^rimidine synthesis and, 125
Malonic acid,

formation from uracil, 623

pyrimidine synthesis and, 125, 129
Malononitriles,

purine synthesis and, 133, 135

pyrimidine synthesis and, 125, 130
Manganese,

adenosine kinase and, 594

deoxyribonuclease and, 581

5'-nucleotidase and, 592

phosphodiesterase and, 588
9-D-Mannopyranosyladenine,

reactions of, 148

synthesis of, 152-153
Mannose,

cysteine-sulfuric acid reaction and, 303

nucleotides and, 60
2-Mercaptoadenine,

adenine synthesis and, 131

synthesis of, 132, 133
2-Mercapto-5-chloropyrimidine, synthe-
sis of, 128
2-Mercaptoethylamine, 184
2-Mercaptohypoxanthine,

conversion to hypoxanthine, 131

synthesis of, 132
6-Mercaptopurine,

biological activity of, 105-106

reaction with amines, 130

synthesis of, 130

tumors and, 135
Mercaptopurines, reactions of, 130-131
Mercaptopyrimidines,

purine sj'nthesis and, 132

reactions of, 127-128

synthesis of, 127, 128

INDEX

675

Mercury salts, purines and pyrimidines,

124
Methionine, 575, 614
"active", 158
synthesis of, 188
dexiro-Metasaccharinate, deoxyribose

synthesis from, 52
2-Methox3'ethanol, see Methyl cellosolve
2-Methoxyethyl dimethyl phosphate,

hydrolysis of, 415
2-Methoxyethyl methyl hydrogen phos-
phate, hydrolysis of, 415
2-Methyladenine, occurrence of, 136
6-N6-Methyladenine, 144
7-Methyladenine, 143
9-Methyladenine, absorption spectra of,

143
1-Methylallantoin, formation of, 616
3-MethylaIlantoin, formation of, 616
2-Meth3'l-5-aminomethyl-6(4)-hydroxy-

pyrimidine, preparation of, 93
Methyl 2,3-anhydroribose, deoxyribose

and, 50-51
6-Methyl-8-azapurine, synthesis of, 106
5-Methylbarbituric acid,
barbiturase and, 623
degradation of, 623
formation from thymine, 622
uracil oxidase and, 622
2-Methyl-5-bromomethyl-6-aminopyrim-

idine, stability of, 93
Methyl cellosolve, paper chromatography

and, 253
4-Methyl-5-chloromethyluracil, stability

of, 92
Methylcyanoacetic acid, thymine syn-
thesis and, 126
5-Methylcytidylic acid,
polynucleotide sequence and, 235
synthesis, enzymatic, 566
ultraviolet absorption of, 513
5-Methylcytosine, 84, 90, 123, 347, 370, 622
amino group of, 448
cation exchange behavior, 219
deamination of, 118
deoxypentose nucleic acid structure

and, 371
dihydroorotic acid dehydrogenase and,

621
discovery of, 89-90
nucleic acid fractionation and, 363-365

occurrence of, 14, 346-347, 355-356

oxidation of, 121

paper chromatography of, 249, 250

synthesis of, 127-128

ultraviolet absorption of, 115-116, 500.

503, 505
uracil oxidase and, 622
5-Methylc3'tosine deoxyriboside, 139
ultraviolet absorption of, 506, 509
5-Methylcytosine deoxyriboside diphos-
phate, 172
deoxyribonucleic acid and, 441
5-Methyl-2-cytosine-thiol, synthesis of,

127
Methyl-2-deoxyribosides, properties of,

55-56
Methylene blue,
deoxyribonuclease assay and, 580-581
uracil oxidase and, 622
Methyl glycol, paper chromatography'^

and, 253
Methyl green,

deoxyribonuclease assay and, 580-581
specificity of, 544
1-Methylguanine, deamination of, 598
7-Methylguanine, 143
9-Methylguanine, 143
2 -Methyl -4 -hydroxy -6- hydroxymethyl-

pyrimidine, 91
4 -Methyl -5-hydroxymethyl -6 - ami nopy -

rimidine, stability of, 93
4-Methyl-5-hydroxymethyluracil, prepa-
ration and properties, 90
2-Methylhypoxanthine, synthesis of, 133
7-Methylhypoxanthine, 143
9-Methylhypoxanthine, 143
5-Methylisocytosine,
deamination of, 118
synthesis of, 125
/3-Methylmalic acid, pyrimidine synthe-
sis and, 125
6-Methylmercaptopurine, reaction with

amines, 130
Methylpentoses,

cysteine-sulfuric acid reaction and,

294-295, 303
orcinol reaction and, 301
6-Methylpurine, synthesis of, 132
Methyl riboside, preparation of, 24-26
D-Methylserine, amicetin and, 135
5-Methylthiouracil, 104

676

INDEX

1-Methylthymine, formation of, 144
0-Methyl-L-tyrosine, occurrence of, 95,

160
1-Methyluracil,

dissociation constants of, 455, 457

formation from uridine, 144
3-Methyluracil, dissociation constants

of, 455, 457
4-Methyluracil, 86

conversion to chloro derivative, 127

orotic acid and, 101

reduction of, 119
5-Methyluracil, see Thymine
6-Methyluracil, uracil oxidase and, 622
1-Methyluric acid, oxidation of, 616
3-Methyluric acid, oxidation of, 616
7-Methyluric acid,

oxidation of, 616

uricase and, 618
9-Methyluric acid, oxidation of, 616
1-Methylxanthine, dissociation constants

of, 456-457, 458
3-Methj'lxanthine, dissociation constants

of, 456-457, 458
7-Methylxanthine, 143

dissociation constant, 456-457, 458
8-Methylxanthine, formation of, 119
9-Methylxanthine, 143

acid-base properties of, 477

dissociation constant, 456-457, 458
Methylxanthines,

infrared absorption of, 547

ultraviolet absorption of, 552
Microorganisms,

deoxypentose nucleic acid,
composition of, 356-358
extraction of, 330-332

deoxyribonucleases of, 584-585

pentose nucleic acid,
isolation, 390-391
nucleotides of, 402-403

pentose nucleoprotein isolation, 379

ribonucleases of, 576
Microsomes,

composition of, 374-375

nucleic acid, analysis of, 207-209
Mitochondria,

deoxyribonuclease and, 624, 625

nucleic acid, analysis of, 207-209

ribonuclease and, 624, 625

Molecular weight,

deoxypentose nucleic acids, 470-474
pentose nucleic acids, 474-475
Molybdenum, xanthine oxidase and, 611
Mononucleotides,

deoxyribonuclease and, 578
electrometric titration of, 410
esters of, 415-418
infrared absorption and, 549-550
intestinal phosphatase and, 589
ion exchange,
distribution coefficients, 221-223
elution, 223-224
ionic properties, 221
isomer separation, 225-230
separation from other phosphates,
224-225
phosphodiesterase and, 585
polymerization of, 557
ribonuclease and, 427
Morpholine, paper chromatography and,

251-253
Murexide, 122
Muscle adenylic acid, see Adenosine-5'-

phosphate
Muscle inosinic acid, see Inosine-5'-

phosphate
Myokinase, transphosphorylation and,

626
Myosin, birefringence of, 535

N

Naphthalene, dichroism of, 538
N-(1-Naphthyl)ethylenediamine, ade-
nine determination and, 120, 304
Nebularine,

biological activity of, 94
isolation of, 60
structure of, 93-94, 188
synthesis of, 61
Nicotinamide, 40
diphosphopyridine nucleotide hydro-
lase and, 608
N-glycosides of, 181
Nicotinamide nucleoside phosphorjdase,

603
Nicotinamide-S'-nucleotide, 5'-nucleo-

tidase and, 592
Nicotinamide riboside, 36, 604

hydrolysis of, 606
Nitrate reductase, molybdenum and, 611

INDEX

677

5-Nitro-2,6-dichloropyrimidine, amina-

tion of, 127
Nitrogen,

content of pentose nucleic acid, 393-394
deoxypentose nucleic acid and, 334
Nitrogen mustards, deoxyribonuclease

and, 582
Nitromalonaldehj'de, 125
bts-p-Nitrophenyl phosphate, phospho-
diesterase assay and, 585-586
5-Nitropyrimidines, purine synthesis

and, 131
Nitrosocyanoacetate, purine synthesis

and, 132
5-Nitrouracil, 104

conversion to chloro derivative, 127
5-Nitrouridine, 143
Nitrous acid, effect on purines and

pjrimidines, 118-119
Nucleases, 331, 332
definition, 557

nucleoprotein isolation and, 313
spleen, specificity of, 432
Nuclei,
deoxyribonuclease and, 624, 625
enzymes in, 625
nucleic acid, analysis of, 209
pentose nucleic acid of, 400-402, 406
ribonuclease and, 624-625
Nucleic acids,
cellular,

dichroism and, 542-544
infrared absorption and, 549
denaturation, ultraviolet absorption

and, 514-516
derivatives, estimation of, 260-263
early history, 1-3
estimation,

phosphorus and, 285-286
spectrophotometric, 529-530
hydrolysis of, 19-20
infrared absorption spectra of, 545-551
metabolism, 556-558

enzymes of, 623-625
molar absorptivities of, 516-518
protein combinations, 530-532
ribose isolation from, 19-20
separation of, 192
structure,
infrared absorption and, 546
x-ray diffraction and, 461-470

ultraviolet absorption, tetranucleotide

theory and, 514
ultraviolet dichroism of, 532, 538-542
definition and equations, 535-537
form dichroism, 534-535
general theory of, 533-534
simple compounds, 537-538
stained nucleic acid, 544-545
Xucleohistone,
extinction of, 314
fractionation of, 361-366
isolation of, 310, 312-317
occurrence of, 311
phosphatase and, 577
solubility of, 314
Nucleoprotamines,
isolation of, 310, 312-317
occurrence of, 311-312
Nucleoproteins,
artifacts, 319-321
bacterial, isolation of, 314-315
cleavage of, 319
cysteine-sulfuric acid reaction and,

295-296
dichroism, 541-542

infrared, 551
diphenylamine and, 290
nature of, 311-312
nucleic acid content of, 318
properties, 317-321
purification of, 315
separation of, 317-318
spectrophotometric titration anomalies

of, 531
ultraviolet absorption of, 318, 530-532
x-ray diffraction of, 531
/5-Nucleoprotein, 3
Nucleosidases, terminology, 600
Nucleoside diphosphates,
diesterase and, 422-423
electrophoresis, 282
Nucleoside-5'-phosphates, production of,

161
Nucleoside phosphorylase,

intracellular localization of, 625
riboside synthesis and, 36-37
Nucleoside polyphosphates, formation

of, 625-626
Nucleosides,
borate complexes,

678

INDEX

electrophoresis of, 283-284
ion-exchange of, 236
carbohydrate ring structure of, 144-

148
conversions of, 142-143
detection on paper, 247-248
dissociation constants of, 458-461
electrophoresis of, 272-273
formation from pentose nucleic acid,

197-198
glycosidic linkage of, 143-144, 148-149
ionic properties of, 217-218, 219
occurrence of, 138

paper chromatography of, 251, 253, 254
phosphorylation of, 593-594
properties of, 157
purine, orcinol reaction and, 301
reaction with acetone, 147-148
reaction with periodate, 147, 148-149
structure, x-ray diffraction and, 452-

455
synthesis,

purines, 149-155
pyrimidine, 155-156
terminal, ribonuclease and, 426
ultraviolet absorption, 504-511

carbohydrate and, 507-511
x-ray analysis of, 149
Nucleotidase, 172

inhibition of, 162
3'-Nucleotidase,

heat stability of, 593
inhibition of, 593
kinetics of, 593
occurrence of, 592
specificity of, 592-593
trinucleotide structure and, 429
5'-Nucleotidase,
buffers and, 592
bull semen,

activators and inhibitors, 592
potency, 591
specificity, 592
stability, 592
deoxyribonucleotides and, 441
occurrence, 591
ribonucleic acid and, 422, 434
Nucleotide-l'-phosphorylase, ribose

phosphates and, 607
Nucleotide-1 '-pyrophosphorylase, nucle-
otide synthesis and, 607-608

Nucleotides,

analysis of, 395-396
biosynthesis of, 600-601
borate complexes,

electrophoresis of, 283-284

ion-exchange of, 241
coenzymes and, 177-187
cyclic, 162, 571-572

formation, 170, 171

hydrolysis of, 171, 414-415, 417

paper chromatography of, 257

phosphodiesterase and, 588

ribonucleases and, 564-565, 575

ultraviolet absorption of, 511-512
detection on paper, 247-248
dissociation constants of, 458-461
electrophoresis of, 268-272
ion-exchange chromatography of,

mononucleotides, 221-230

polynucleotides, 233-235

polyphosphonucleosides, 230-232
isolation of, 161
isomeric,

elution order, 226

ion-exchange of, 225-230
liberation from pentose nucleic acid,

198
occurrence of, 138
paper chromatography of, 250, 253,

255-257, 264
properties of, 177
purine,

cysteine-sulfuric acid reaction and,
303

orcinol reaction and, 301
pyrimidine,

cysteine-sulfuric acid reaction and,
303

orcinol reaction and, 301
sequence,

deoxyribonucleic acid, 442-445

polyribonucleotides, 436-439
sources of, 160-162
spacing of, 461
structure of, 162-172
synthesis,
cylic, 174-175

deoxyribose nucleotides, 175-176
isomeric, 174
phosphatases and, 594

INDEX

679

5'-phosphates, 173-174

ultraviolet absorption of, 511-514
2'-Nucleotides, 3'-nucleotidase and, 593
5'-Nucleotides, 3'-nucleotidase and, 593
Nucleotropomyosin ,

isolation of, 376

pentose nucleic acid content of, 382

surface charge of, 381-382

O

Oligonucleotides,

chromatographic separation of, 429
deoxypentose, [jhosphodiesterase and,

587
deoxyribonuclease and, 442, 578
hydrolysis,

alkaline, 426-427

enzymatic, 562-563
nucleic acid hydrolj-sis and, 196
pentose, phosphodiesterase and, 587
periodate and, 424
phosphate, position of, 421
production of, 423
ribonuclease and, 426-427, 562
ribonucleic acid, structure of, 428-430
synthesis, enzymatic, 566
uranyl chloride and, 567
Orcinol, reaction, 35
deoxyribose and, 54
ribonucleic acid,

interfering substances, 301-302

purines and, 394

quantitative, 301, 302

specificity, 301
Orotic acid, 622
dihj'droorotic acid dehydrogenase and,

621
barbiturase and, 623
conversion to uracil, 608
decarboxylation of, 130
degradation, enzymatic, 620-621
nucleotide pyrophosphorylase and,

607-608
occurrence of, 100
paper chromatography of, 253
properties of, 101
thymidine phosphorylase and, 605
Orotidine, 159

isolation of, 100-101
Orotidylic acid.

biosynthesis of, 607-608
decarboxylation of, 608
Orthoformate, pyrimidine synthesis and,

129
Oxalic acid, formation from uracil, 120,

121
Oxalic acid dihydrate, hydroxyl bond

length in, 449
Oxaluric acid,

formation from uracil, 120
formation from uric acid, 618
3-Ox3'quinaldine, thymine estimation
and, 121

Paludrine, resemblance to pyrimidines,

104
Pancreas,

pentose nucleic acid of, 398
ribonucleases of, 558-575
Pantetheine, 185
Pantetheine-4'-phosphate, synthesis of,

186
Pantothenic acid, 184
Paper Chromatography, see Chromato-
graphy, paper
Paper, electrophoresis and, 274, 275
Parabanic acid, formation from guanine,

85, 122
Penicillin, ribonuclease and, 572
Pentobarbital, barbiturase and, 623
Pentose nucleic acids, see also Ribo-
nucleic acids
acid-base properties of, 480-483
adsorption of, 391
bacterial, inhibition by, 390
birefringence of, 539
cellular,

analysis of, 205-209
distribution, 374
chain-branching of, 482
composition,
electrophoresis and, 276-277
elementary, 393-394
contamination, 350
detection of, 317
deaminated, titratable groups of, 481
decomposition of, 392-393
dialyzable fraction of, 391
degradation,

680

INDEX

criteria of, 384-385

ribonuclease and, 407
dichroism and, 538
distribution of, 4-6
dye-binding and, 482-483
homogeneity, 392

preservation of, 384-386
hydrolysis, 259-260

acid, 192-196

alkaline, 197-199, 483

products, 191-192
infrared absorption of, 549, 553
intestinal phosphatase and, 589
isolation,

animal tissues, 387-389

general, 384-386

microbial, 390-391

plant tissues, 389-390
liberation of acid groups from, 199
metachromasy and, 545
molecular weight determination, 386,

474-475
nature of, 391-394
nucleotide composition,

analytical procedures, 395-396

animal, 396-402

general considerations, 394-395

microbial, 402-403

plant, 403-405
organ-specificity of, 406
phosphodiesterase and, 586, 587
polymerization of, 393
position in nucleotropomyosin, 382
purified, analysis of, 204-205
reaction with albumin, 381
reaction with cytoplasm, 381
regularity, inter specific, 406-407
removal from deoxypentose nucleic

acid, 332-333
removal from tissues, 329
ribonuclease kinetics and, 570-572
species-specificity of, 405-406
titratable groups of, 481-482
ultracentrifugation of, 391
ultraviolet absorption, 393

digests, 569-570

ribonuclease and, 518-519
virus, function of, 405
Pentose nucleoprotein,
centrifugation of, 380-381
components, separation of, 380-382

isolation,

animal, 375-376

general considerations, 374-375
microbial, 379
plant, 376-379
procedures, 376, 377, 378
nature of bonds, 379-383
Pentose phosphate isomerase, ribose-5-

phosphate preparation and, 63
Pentoses,
cysteine-sulfuric acid reaction and,

294-295
indole-hydrochloric acid reaction and,
297, 298
Pepsin, ribonuclease and, 574-575
Peptides, infrared dichroism and, 550
Perchloric acid, nucleic acid hydrolysis

by, 195, 197
Periodate,
reaction of nucleosides, 147, 148-149
ribonuclease digests and, 424-425
ribonucleic acid end-group analysis
and, 438
Permanganate, pyrimidines and, 120
Phenobarbital, 103
Phenol,
paper chromatography and, 250, 257
xanthine oxidase and, 613
N-Phenyl-2-deoxyribosylamines, 56
o-Phenylene diamine, 21
Phenylhydrazine, reaction with uridine,

144
Phloroglucinol, reaction with ribonucleic
acid,
procedure, 302
specificity, 302
Phosgene, purine synthesis and, 135
Phosphatase, 167, 199
action on nucleotides, 164
bone, nature of, 589-590
deoxyribonucleic acid and, 441
deoxyribonucleic acid digests and,

576-577
liver, phosphate transfer by, 594
low specificity, 590
malt, phosphate transfer by, 594
nucleic acid hydrolysis and, 198
prostatic,
end-group analysis and, 437
nucleotides and, 590
phosphate transfer and, 594

INDEX

681

ribonuclease digests and, 569
specific, 591-593

transphosphorylation by, 593-594
5'-Phosphatase, 164
Phosphate,
donors, phosphatases and, 594
inorganic, in nucleotide separation,

224
liberation from pentose nucleic acid,

193, 194-195
3'-nucleotidase and, 593
position, ultraviolet absorption and,

511-514
sugar, separation from nucleotides,
224-225
Phosphocreatine, hjdrolysis of, 37
Phosphodeoxyriboaldolase, deoxjribose

synthesis and, 59
Phosphodiesterase,

antivenoni serum and, 588
deoxyribonuclease digests and, 579
inhibition of, 588
intestinal,
assay, 590
specificity, 588-590
ribonucleic acid and, 422-423
snake venom,

activators and inhibitors, 588
degradation products, 586-587
toxicity and, 587-588
specificity, 585-586
Phosphodiesters,
deoxyribonucleic acids and, 440
nucleic acids and, 410, 412
nucleic acid branching and, 422, 433
ribonuclease and, 423, 569
Phosphoglucomutase, 37
Phosphogluconate, 35

S'-nucleotidase and, 592
Phosphomonoesterase, 442, 587, 589-590
action, mechanism of, 563
periodate and, 424-425
ribonucleic acid and, 435
digests, 422, 424
end-group analysis, 438
snake venom, 585
trinucleotide structure and, 429
Phosphomonoesters,
nucleic acids and, 410-411
ribonuclease and, 423

Phosphoriboaldolase, deoxyribose syn-
thesis and, 59
Phosphoribomutase, 37
Phosphorus,
compounds, separation from nucleo-
tides, 277-279
deoxypentose nucleic acid and, 334
estimation, 258

nucleic acids and, 285-286
pentose nucleic acid, 393-394
Phosphoryl groups, secondar}^ ribo-
nuclease and, 567-568, 569
Phosphotriesters,

nucleic acids and, 410-411, 412, 440,
482, 483
branching of, 422, 433, 436
stability of, 416
Phytonucleic acid, 3

Picrates, purine and p3-rimidine, 124-125
Piperidine, paper chromatography and,

253
Plants,

deoxypentose nucleic acid extraction,

330
deoxyribonuclease of, 584
nucleoprotein isolation from, 376-379
pentose nucleic acid isolation from,

389-390
ribopolj'nucleotidases of, 576
viruses, nucleotides of, 403-405
"Plasmonucleic acid", see Pentose

nucleic acids
PNA, see Pentose nucleic acids
Polarography, purines and pyrimidines,

119-120
Poly-N-n-butj'l-4-vinylpyridonium bro-
mide, electrolytes and, 484
Polynucleotides,
electrophoresis of, 268-272, 280-281,

282
elation sequence, 235
ion-exchange of, 233-235
limit,
common structures of, 562
composition, 579

deoxyribonuclease, 345, 579, 584-585
dialysis of, 562

hyperchromia effect and, 569-570
infrared absorption and, 549-550
molar absorptivities of, 519-520

682

INDEX

ribonuclease and, 427, 562, 575
separation of, 562
molar absorptivities and, 521
3'-nucleotidase and, 593
nucleotide sequence in, 436^39
phosphate, position of, 420-421
Polyphosphates, nucleoside, ion-ex-
change of, 230-232
Polyribophosphate, 60
Polysaccharides, 562
contamination by, 350
nucleic acid separation from, 333
ribonuclease and, 572
Polytomella caeca, cytoplasm, reaction

with pentose nucleic acid, 381
Polyvinyl alcohol, deoxypentose nucleic

acid and, 529
Potassium bromide, infrared spectros-
copy and, 546-547
Potato virus X, pentose nucleic acid,

nucleotides of, 405
Proflavine, paper chromatography and,

265
Propanol, paper chromatography and,

249, 264-265
Protamines, 2

diphenylamine and, 289-290
occurrence of, 311
reaction with nucleic acid, 320-321
Proteins,
conjugated,
classification, 311
definition, 309
contamination by, 350
cysteine-sulfuric acid reaction and,

295-296
deoxypentose nucleic acid absorptivity

and, 531-532
infrared dichroism and, 550
nucleic acid combination, 530-532
pentose nucleic acid separation from,

380
ultraviolet absorption of, 530
Protein nucleate, 310
Pseudovitamin B12, adenine and, 95
Pseudovitamin Bi2d, purine in, 136
keto -T>-psicose, 47-48
Pterins, xanthine oxidase and, 613
Purine,
isolation of, 93
reduction of, 119

stability of, 116-117
synthesis of, 130-131, 132
Purine deoxyribonucleotides, formation

of, 162
Purine nucleosidases, history of, 601-602
Purine nucleoside hydrolase, bacterial,

607
Purine nucleoside phosphorylase,
animal, specificity, 602-603
microorganismal,

assay, 604

hydrogen ion concentration and, 603-
604

kinetics, 604

mechanism of action, 603

stability, 604
purification of, 602
Purine nucleosides, stability of, 192
Purine nucleotides,
hyperchromia effect and, 569-570
linkage in ribonucleic acid, 432
paper chromatography of, 255, 257
ribonuclease and, 423-424, 563
Purines,

anionic exchange of, 219-220
arsenotungstic acid reaction and, 305
cellular, 138

color reactions for, 303-304
counter-current distribution of, 110
deoxy ribonuclease and, 577-578
detection on paper, 245-247
discovery of, 2

dissociation of, 112-116, 217-218, 219
electrophoresis of, 272-273
enolic hydroxyls of, 411
estimation, 260-263

pentose nucleic acid, 199-209, 394

procedure, 202-204
hydrogen bonding of, 464-467
identification of, 110-116
infrared absorption of, 547
isotopic,

synthesis of, 131, 133

ultraviolet absorption of, 551-552
liberation from nucleic acids, 193, 196,

258
naturally occurring, 84-86
nomenclature of, 83-84
paper chromatography of, 250, 255
physical properties of, 108-109
ratio in pentose nucleic acid. 397

INDEX

G83

recovery from hydrolj^sates, 202
reduced, ultraviolet absorption of, 552
salts of, 124-125

separation from pyrimidines, 199-200
solubility, effect of substituents on, 110
spectrophotometry of, 112-116
structure, x-ray diffraction and, 451-

452
sublimed, ultraviolet absorption of,

501-504
synthesis of, 130-135

from imidazoles, 134-135
from purines, 130-131
from pyrimidines, 131-134
thymine estimation and, 304
transamination and, 599-600
ultraviolet absorption of, 450-451
Puromycin, 15, 188

structure of, 95, 160
Purone, formation from uric acid, 119
Pus, nucleic acids and, 1
Pyrazolone, formation of, 144
Pyridine, nucleic acid hydrolysis, and,

198
Pyridine nucleotides, 15
Pyridoxal, purines and, 600
Pyridoxine, 92
Pyrimidine,
reduction of, 120
stability of, 116-117
synthesis of, 128
Pyrimidine deoxyribonucleoside diphos-
phates, structural significance of,
366-367
Pyrimidine deoxyribonucleotides,
dialysis of, 445
formation of, 162
Pyrimidine -2', 5'-diphosphomononucleo-

tides, phosphodiesterase and, 586
Pyrimidine -3', 5'-diphosphomononucleo-

tides, phosphodiesterase and, 586
Pyrimidine nucleosidases, history of,

601-602
Pyrimidine nucleoside hydrolases,
bacterial, 607
yeast,
isolation, 606
mechanism of action, 606
specificity of, 606
Pyrimidine nucleoside phosphorylases,
specificity of, 604-605

Pyrimidine nucleosides, 36, 123
stability of, 192
terminal, ribonuclease and, 426
Pyrimidine nucleotides,
liberation from deoxypentose nucleic

acid, 197
paper chromatography of, 255
pentose nucleic acid isolation and, 390
recovery from hydrolysates, 202
ribonuclease and, 423-424
ribonucleic acid branching and, 434-

436
separation from purines, 199-200
stability of, 194-195
Pyrimidines,
amino groups of, 547
anionic exchange of, 219-220
antimetabolites and, 104
bond distances in, 448-451
catabolism of, 619-623
cellular, 138

color reactions for, 303-305
counter-current distribution of, 110
deoxyribonuclease and, 578
detection on paper, 245-247
dissociation of, 112-116, 217-218, 219
electrophoresis of, 272-273
enolic hydroxyls of, 411
estimation, 260-263
pentose nucleic acid, 199-209
procedure, 202-204
hydrogen bonding of, 464-467
identification of, 110-116
infrared absorption of, 547-548
isotopic, ultraviolet absorption of, 551-

552
keto form of, 449-451
liberation from nucleic acids, 195, 197,

258
naturally occurring, 86-93
nomenclature of, 82-83
paper chromatography of, 250, 255
physical properties of, 108-109
ratio in pentose nucleic acid, 397
reduced, ultraviolet absorption of, 552
resonance in, 448
salts of, 124-125

solubility, effect of substituents on, 110
spectrophotometry of, 112-116
structure, x-ray diffraction and, 447-

451

684

INDEX

sublimed, ultraviolet absorption of,
501-504

substituted, stability to alkali, 118

synthesis of, 125-130

transamination and, 599-600

ultraviolet absorption of, 450-451
2,6-Pyrimidinethiols, sj^nthesis of, 127
Pyrimidones, 450

Pyrogallol, xanthine oxidase and, 613
Pyronine-methyl green,

dichroism and, 544

specificity of, 544
Pyrophosphatase, 186
l-Pyrophosphoryl-5-phosphoryl ribose,

nucleotide synthesis and, 607
Pyruvic acid, formation from thymine,

121, 123
Pyruvic aldehyde,

cysteine-sulfuric acid reaction and, 295

glyceraldehyde-3-phosphate and, 438

Quinoline, paper chromatography and,

253
Quinone imine, xanthine oxidase and,

612-613

Resins,

cross-linkage, effect on nucleotide
separation, 233-234

ion-exchange, properties of, 212-213
Resorcinol, hydroxyl bond length in, 449
Rhamnose,

paper chromatographj' of, 264

tryptophan-perchloric acid reaction
and, 296
D-Ribal, 17

Ribazoles, see also 5,6-Dimethylben-
zimidazole riboside

activity of, 31

synthesis of, 61

synthetic, activity of, 62
Ribazole phosphate, synthesis of, 61-62
Ribitol, 16, 48

occurrence of, 15
Ribitol-5-phosphate, 163
1-Riboarsenate, nucleoside phosphoryl-

ase and, 603
Ribobenzimidazole, 47, 141

importance of, 31
preparation of, 21
Ribodesose, see Deoxyribose
Riboflavin, 15, 29

Riboflavin phosphate, see also Flavin
mononucleotide
ion-exchange of, 232
synthesis of, 183
D-Ribofuranosyl adenine, nucleoside

hydrolase and, 606
9-D -Ribofuranosyl -2,6-diaminopurine,

synthesis of, 151
l-(Q:-D-Ribofuranosyl)-5,6-dimethylben-

zimidazole, 13
D-Ribonic acid, 12, 21, 141
reactions of, 47
ribose synthesis and, 16
Ribonolactone, 16, 17
Ribonuclease I, see Ribonuclease, pan-
creas
Ribonuclease, 430
action,
chemistry of, 423-428
mechanism of, 431
products of, 161-162
volume changes and, 563
amino acids of, 574-575
antibodies against, 625
assay,
manometric, 568-569
phosphorus and, 566
spectrophotometric, 566-567
turbidimetric, 566
bacterial, 576

cyclic phosphates and, 425-426
definition of, 557
digests,

dialysis of, 423
electrophoresis of, 279-282
ion -exchange of, 234-235
nucleotide isomers in, 425
periodate and, 424
enolic group liberation by, 412
inhibition of, 385-386, 571
intracellular localization of, 624-625
isolation of, 572-573
molar absorptivities and, 518-519
mononucleotide esters and, 430
nucleic acid separation by, 332-333
oligonucleotides and, 426-427

INDEX

685

pancreas,

assay of, 566-567, 568-569

history, 558-559

homogeneity of, 573

hydrogen ion concentration and, 572

hydrolytic products of action, 560-
563

inhibitors, 572

kinetics of, 570-572

mechanism of action, 563-565

nucleotide esters and, 565

physicochemical properties of, 573-
575

protein nature of, 572-575

specificity of, 559-560

synthetic reactions of, 565-566

temperature and, 572

titrable acidic groups and, 567-568,
569

ultraviolet absorption and, 569-570
pentose nucleic acid composition and,

398-399
pepsin and, 574-575
performic oxidation of, 575
5'-phosphates and, 422
plant, 576

pyrimidine nucleotides and, 423-424
removal of, 387, 398
secondarj' phosphoryl and, 423
specificity of, 427-428
spleen,

mechanism of action, 575-576

specificity of, 575
test for pentose nucleic acid, 4, 6
transesterification and, 566
transphosphorylation and, 563
turnip yellow mosaic virus and, 382
Ribonucleic acid, 3, see also Pentose

nucleic acid
adenine estimation in, 304
alkali lability of, 416
branching of, 422, 432-436
color reactions for,

cysteine-sulfuric acid, 302-303

evaluation of, 303

orcinol, 300-302

phloroglucinol, 302
contamination, removal of, 325
cores, 412

depolymerization of, 431
diesterase and, 434-435

electrometric titration of, 410
end-group analysis of, 437-438
hydrolysis, 429-430

alkaline, 418-419

cyclic intermediates, 419

enzymatic, 421
indole-hydrochloric acid reaction and,

297
internucleotidic linkages, 420-422
lability of, 478-479
methylation of, 433-434
monomers of, 438-439
phosphate linkages in, 422-423, 430-432
spleen nuclease and, 432
x-ray diffraction of, 461, 470
Ribonucleosides, 604

stereochemistry of, 417-418
Ribonucleotides,

acid-base properties of, 477-479
cation exchange behavior, 219
cyclic, electrophoresis of, 279-280
separation from other phosphorus com-
pounds, 277-279
D-Ribopyranosyl adenine,

nucleoside hydrolase and, 606
synthesis of, 152-153
D-Ribose, 3, 10
acetals of, 46

acj'lderivatives, purification of, 19
anhydrides of, 46-47
browning of, 14

chromatography of, 21-22, 24, 263-264
compounds containing, 13
derivatives, 20-21

occurrence of, 15

properties of, 65-74, 79-80
dissociation constants and, 458
esters of, 43-46, 63-64
ethers of, 42-43
form in nucleic acid, 12
N -glycosides of, 27-33
0-glycosides of, 24-27
identification and estimation, 11-12,

20-22, 140-141
ionic behavior of, 219
isotopic, 61
liberation from pentose nucleic acid,

193
mutarotation of, 22-23
natural occurrence of, 60
pentose nucleic acids and, 394

686

INDEX

phosphates, 33-43
chromatography of, 35-36
differentiation of, 35
hydrolysis of, 33-35

physical properties of, 22-24

position, in nucleosides, 453-455

preparation of, 16-20

purification of, 18-19

reactions of, 27-31, 47-48
Ribose-l,5-diphosphate, 37

as coenzyme, 13

nucleotide phosphorylase and, 607
Ribose-1 -phosphate, 13, 604

conversion to 5-phosphate, 37

formation of, 36

hydrolysis of, 37

inosine formation and, 602-603

nucleoside phosphorylase and, 603

riboside synthesis and, 36-37

thymidine phosphorylase and, 605
Ribose-2-phosphate,

identification of, 38

isolation and properties of, 62-63
Ribose-3-phosphate,

identification of, 38

isolation and properties of, 38-39, 62-63

polymerized, ribonuclease and, 559, 563
Ribose-4-phosphate,

separation of, 240-241

preparation of, 63
Ribose-5-phosphate, 13

chromatography of, 40-41

deoxyribose biosynthesis and, 59

formation from ribose-1 -phosphate, 41

identification of, 39-40

isolation of, 39, 40

metabolism of, 41

5'-nucleotidase and, 592

nucleotide synthesis and, 607

preparation of, 63

proof of structure, 163

properties of, 63

pyrophosphoryl ribose phosphate and,
607

synthesis of, 40
Ribose phosphates, borate complex, ion-
exchange of, 236-241
Ribosides, see also Nucleosides,

alkyl, synthesis of, 26-27

anionic exchange of, 221
Ribosimine, 29

Ribolose, 53

Rosaniline, pentose nucleic acid and,
482-483

Salt, paper chromatography and, 250

Sarcosine, 614

Sarkin, see Hypoxanthine

Schiff's reagent,

deoxyribonucleic acid and, 299
deoxyribose and, 54
Selenite,

deoxyribonuclease and, 582
Serine, 614

Serum albumin, nucleic acid and, 320-321
Silver, deoxypentose nucleic acid ab-
sorptivity and, 529
Silver salts, purine and pyrimidine, 124
Snake venom, phosphodiesterase of, 585-

588
Sodium, deoxypentose nucleic acid

rigidity and, 489
Sodium chloride,
deoxypentose nucleic acid absorptivity

and, 525-526
nucleohistone extraction and, 313-314
nucleoprotamine extraction and, 316-

317
pentose nucleic acid and, 386, 391, 393
pentose nucleic acid isolation and,

375, 387
pentose nucleoprotein separation and,
380
Sodium deoxycholate, 328, 331

nucleoprotein and, 319
Sodium dodecyl sulfate,
deoxypentose nucleic acid and, 327-328
nucleoprotein and, 319
pentose nucleic acid isolation and, 387,

389, 393
pentose nucleoprotein separation and,

380
ribonuclease and, 386
Sodium phosphate, paper chromatogra-
phy and, 255, 257
Sodium xylene sulfonate, 327
Solvents, chromatographic,
hydrogen ion concentration, 249
organic components, 249-250
nucleoside separation, 250-255
nucleotide separation, 255-257

INDEX

687

purine and pyrimidine separation, 250-

255
salt content, 250

theoretical considerations, 248-249
water content, 249
Spectrophotometry,

estimation of purines and pyrimidines

and, 199-202
nucleic acids and, 5
Sperm,

anisotropy of, 538

human, deoxypentose nucleic acid ex-
traction, 330
nucleic acid and, 2

nucleoprotamine isolation from, 316-
317
Spleen,
pentose nucleic acid of, 398
ribonuclease, properties of, 575
Spongosine, purine in, 98
Spongothymidine, 98, 159, 188

source of, 87
Streptomycin, deoxj'pentose nucleic acid

and, 337
Sucrose, deoxypentose nucleic acid and,

529
Sugars, borate complexes, ion-exchange

of, 235-236
Sulfanilic acid, reaction with thymine,

304
Sulfonamides, 104

Takadiastase,

adenosine deaminase of, 596
nucleotidase activity of, 592
2,3,5,6- Tetraacetyl - d - galactof uranose,

154
1,2,3,5-Tetraacetyl-D-ribofuranose, for-
mation of, 151
Tetraethylthiuram disulfide, xanthine

oxidase and, 613
Tetrahydrofurfuryl alcohol, paper chro-
matography and, 253, 256-257
Tetrahydrouric acid, 119
2,4,5 , 6-Tetrahydroxypyrimidine, see

Dialuric acid
Tetranucleotide,

hypothetical, 474-475, 577

nucleic acid structure and, 368-369

Tetroses,

cj'steine-sulfuric acid and, 295

diphenylamine and, 288
Theobromine, 138
Theobromine nucleosides, synthesis of,

149-150
Theophylline, 138

deoxyribonucleoside of, 56-57
Theophylline glucoside, cyclic phosphate

of, 175
Theophylline nucleoside,

formation of, 143

synthesis of, 149-150, 151
Thiamine, stability of, 92-93
2-Thioadenine, 105
2-Thiobarbituric acid, barbiturase and,

623
5-Thiomethyl-D-ribose, occurrence of, 15
2-Thiothymine, uracil oxidase and, 622
Thiouracil,

color tests for, 123

derivatives, biological activity of, 103

incorporation of, 135-136

thymidine phosphorylase and, 605

uracil oxidase and, 622
Thiouracil deoxyriboside, 605
Thiouracil riboside, 605
Thiourea,

purine synthesis and, 133

pyrimidine synthesis and, 125, 127, 128

reaction with chloropurines, 130
Thymic acid, 342, 443

composition of, 341

formation of, 196
Thymidine, 139, 140

cysteine reaction and, 292

glycosidic linkage of, 144, 146-147

phosphorolysis of, 58, 63, 604

titration anomalies and, 523

trityl derivative of, 146

ultraviolet absorption of, 504, 509, 517

uridine hydrolase and, 606

uridine phosphorylase and, 606
Thymidine diphosphate, 172

cysteine reaction and, 292-293

deoxyribonucleic acid and, 440-441

paper chromatography of, 255

ultraviolet absorption of, 513

production, mechanism of, 443-444
Thymidine-5'-phosphate, deoxyribo-
nucleic acid digests and, 441

688

INDEX

Thymidine phosphates, synthesis of, 175
Thymidine phosphorylase, 63, 602

animal,

equilibrium, 605
properties, 605
specificity, 605

inhibition of, 605
Thymidylic acid, 171, 172, 441, 579

apurinic acid and, 367-368

carbazole-sulfuric acid reaction and,
298-299

cysteine reaction and, 291-292, 294

cysteine-sulfuric acid reaction and, 295

diphenylamine and, 287-288

estimation of, 299

indole-hydrochloric acid reaction and,
298

ion -exchange of, 223

paper chromatography of, 257

ultraviolet absorption of, 512, 513
Thymine, 3, 14, 48, 118, 123, 341, 346, 347

arsenotungstic acid reaction and, 305

cationic exchangers and, 218

color reaction for, 304-305

degradation of, 622

dihydroorotic acid dehydrogenase and,
621

discovery of, 86

dissociation constants of, 456, 457

ionization of, 268

isotopic, degradation of, 121

nucleohistone fractionation and, 362

oxidation of, 120, 121, 622-623

paper chromatography of, 249, 251

perchloric acid and, 258

polarography and, 120

reducing agents and, 120

stability to acids, 117

synthesis of, 86-87, 125, 126

thymidine formation from, 63

thymidine phosphorylase and, 605

ultraviolet absorption of, 123, 499, 503,
506
Thymine deoxyribodiphosphoric acid,

formation of, 197
Thymine deoxyriboside, see Thymidine
Thymine glycol, formation of, 121
Thymine nucleosides, synthesis of, 155-

156
Thymine xyloside, 87
Thyminose, 11

Thyminyl alcohol, 92
Thyminylamine, 92
Thymus,

deoxyribonucleases of, 584
deoxyribonucleate isolation from,

323-324, 325-326, 326-327, 327-328
nucleohistone isolation from, 313-314
pentose nucleic acid of, 398
Titration, electrometric, nucleic acids

and, 410
Tobacco mosaic virus, 277, 319
dichroism of, 541-542

infrared, 550
hydrolysates, electrophoresis, 284
molecular weight of, 383-384, 475
nucleotides of, 405
particle size, 383-384
pentose nucleic acid isolation from, 389
protein,

composition, 383
orientation, 550-551
x-ray diffraction of, 461
Tobacco ring spot virus, composition of,

383
2-0-p-Toluenesulfonyl-^-L-arabinoside,

deoxyribose synthesis and, 51-52
Toluidine blue, dichroism and, 544
Transforming activity, 338

nucleic acid denaturation and, 339
2,3, 4-Triacetyl-5-benzyl-D-ribose, ade-
nosine synthesis from, 153
2,4, 5-Triamino-6-hydroxypyrimidine,

guanine synthesis and, 132
4 , 5 , 6-Triaminopyrimidine,
adenine synthesis and, 132
reaction with formylmorpholine, 133
4,5,6- Triaminopyrimidine - 2 - sulfinic

acid, adenine synthesis and, 132
Trichloroacetic acid, paper chromatog-
raphy and, 256
2,6,8-Trichloropurine, 85

sj^nthesis of, 130
2,6,8-Trichloropurine nucleosides, syn-
thesis of, 149-150
Triethylamine, purine synthesis and, 130
Trihydroxyglutaric acid, 12, 141
2,6,8-Trihydroxypurine, see Uric acid
2,4,6-Trihydroxypyrimidine, see Barbi-
turic acid
Trimethoxyglutaric acid, 42

INDEX

689

Trimethyl-N-methyladenosine, hydroly-
sis of, 144
2,3,5-Trimethyl-7-D-ribonolactone, 144,

145
Trimethylribose, identification of, 144
1,3,7-Trimethyluric acid, uricase and,

618
Trinucleotides,

electrophoretic mobilities of, 270-271
molar absorptivities and, 520
paper chromatography of, 257
structure of, 429
Triose phosphate, deoxyribose biosyn-
thesis and, 59
Trioses,

cysteine reaction and, 293
cysteine-sulfuric acid reaction and, 295
diphenylamine and, 288
tryptophan-perchloric acid reaction
and, 296
Triphosphopyridine nucleotide, 167, 435,
596, 597
hydrolysis of, 182
3'-nucIeotidase and, 593
ribose in, 13
Triticonucleic acid, 3
Trityl chloride, reaction with adenosine,

145
5'-Tritylcytosine, nucleotide synthesis

and, 174
5-Trityl-D-ribose, nucleoside synthesis

and, 150-151
5'-Tritylth3-midine, nucleotide synthesis

and, 175-176
5'-Trityluridine, nucleotide synthesis and

174
Triuret, formation of, 136
Tropomyosin,
isolation of, 376
surface charge of, 381-382
Trypsin, turnip virus and, 378
Tryptophan, 583

Tryptophan-perchloric acid reaction,
comparative value of, 299-300
deoxypentose nucleic acid,
mechanism of, 296
procedure, 296
specificity of, 296
deoxyribose and, 53-54

Tubercle bacilli, nucleoprotein isolation

from, 314-315
Tuberculinic acid, hydrolytic products

of, 89
Tumors,

8-azaguanine and, 106-107

deoxyribonuclease inhibitors and, 583

2,6-diaminopurine and, 105

6-mercaptopurine and, 105

nucleic acid in, 285-286

ribosides and, 33
Turnip yellow mosaic virus,

antisera and, 382

isolation of, 378-379

nucleoprotein, composition of, 383

nucleotides of, 405

protein of, 382

ribonuclease and, 382
TjTosine, 583

U

UDPG, see Uridine diphosphate glucose
Ultraviolet, absorption, 265

apurinic acid and, 343-344

deoxypentose nucleic acid and, 334-
335, 338-339

nomenclature, 494

pentose nucleic acid and, 393

pH and, 495-498

ribonuclease and, 569-570
Ultraviolet light,

effect on pyrimidines and purines, 123

paper chromatography and, 246-247
Q:,(3-Unsaturated esters, pyrimidine syn-
thesis and, 125
Uracil, 3, 94, 112, 120, 622

cationic exchangers and, 218

color tests for, 123, 305

dihydoorotic acid dehydrogenase and,
621

dissociation constants of, 455-456, 457

formation from cytosine, 87

formic acid and, 259

ionization of, 268

internucleotidic linkages and, 412

isolation and synthesis of, 88-89

oxidation of, 120, 121, 622-623

paper chromatography of, 249, 251,
253, 254

reduction of, 119, 120

690

INDEX

representation of, 451

stability to acids, 117

synthesis of, 125, 126, 130

thiouracil and, 103

thymidine phosphorylase and, 605

thymine estimation and, 304

ultraviolet absorption of 123, 499, 503,
507
Uracil-4-carboxylic acid, see Orotic acid
Uracil-5-carboxylic acid, 101
Uracil deoxyriboside, 139, 157

occurrence of, 14

ultraviolet absorption of, 504, 509
Uranyl chloride, ribonuclease assay and,

566-567
Urea,

deoxypentose nucleic acid and, 332,
529

formation from guanine, 122

formation from pyrimidines, 86, 120,
121, 123, 619, 623

formation from uric acid, 618

nucleoprotein and, 319

paper chromatography and, 251
Ureas,

purine synthesis and, 132, 133, 134

pyrimidine synthesis and, 125, 126,
129-130
Urease, barbiturase and, 623
Ureidomethylenecyanoacetate, pyrimi-
dine synthesis and, 129
Ureidomethylenemalonate, pyrimidine

synthesis and, 129
Ureidosuccinic acid,

formation of, 620, 621
Ureidosuccinic acid, ring-closure of, 621
Uric acid, 13, 82, 84, 94

dehydrogenation of, 618

formation, 609

gout and, 96

indole-hydrochloric acid reaction and,
298

8-methylxanthine formation and, 119

occurrence of, 96

oxidation of, 121-123

paper chromatography of, 253

permanganate oxidation of, 616

purine synthesis and, 130

reduction of, 119, 120

ultraviolet light and, 123, 136

xanthine oxidase inhibition and, 612

Uric acid riboside, 159

occurrence of, 96

nucleoside hydrolase and, 607

paper chromatography of, 253
Uricase, 96, 123, 609, 614

history, 614-615

inhibitors, 619

intracellular localization of, 625

mechanism of action,

decarboxylation reaction, 616, 618
intermediary product, 616-618
oxidation reaction, 615-616

optimal pH of, 618-619

specificity of, 618
Uridine, 12, 100, 138, 140, 157, 594, 622

borate complex, electrophoresis of, 283

cationic exchangers and, 219

dissociation constants of, 459

formation from cytidine, 143, 144

ionization of, 272

methylation of, 145

paper chromatography of, 255

phosphate linkages, 413-415

phosphorolysis of, 602, 604, 605

reduction of, 141

reaction with hydrazines, 144

spectrophotometric estimation of, 200-
202

structure, x-ray diffraction and, 454

synthesis of, 155-156

trityl derivatives of, 146

ultraviolet absorption of, 123, 509, 518
Uridine benzyl phosphate 6, ribonuclease

and, 430
Uridine diphosphate,

coenzymes and, 187

diesterase and, 434

ion-exchange of, 230-231, 232

production of, 434-435

transphosphorylation and, 626

ultraviolet absorption of, 513
Uridine diphosphate glucose,

hydrolysis of, 186-187

paper chromatography and, 264

ribose in, 13

structure of, 98-99
Uridine ethyl phosphate a, ribonuclease

and, 430
Uridine-5'-ethyl phosphate, 176
Uridine hydrolase.

INDEX

G91

assay of, 607

properties of, 606-607
Uridine-5'-phenyl phosphonate, 176
ljridine-2'-phosphate, see also Uridylic
acid

ultraviolet absorption of, 513
Uridine-2',3'-phosphate, ribonuclease

and, 425
Uridine-3'-phosphate,

esters, ribonuclease and, 565

preparation of, 62

ribonuclease and, 425, 562

ultraviolet absorption of, 513
Uridine-5'-phosphate,

coenzymes and, 187

esters, ribonuclease and, 565

structure of, 165

sj'nthesis of, 174

transphosphorylation of, 626

ultraviolet absorption of, 513
Uridine phosphates, tumors and, 189
Uridine phosphorylase, bacterial,

assay, 606

equilibrium, 606

specificity, 605-606
Uridine-5'-pyrophosphate,

derivatives of, 99

ion-exchange of, 232
Uridine triphosphate,

isolation of, 189

transphosphorylation and, 626
Uridylic acid, 169, 622, see also Uridine
phosphate

analysis for, 395

dissociation constants, 459

esters, hydrolysis of, 417

estimation of, 259

hyrolysis of, 170

ion-exchange of, 222-223, 224

ionization of, 268-269

isomers, 38
identification of, 170
ion-exchange of, 228
phosphate position in, 189

liver pentose nucleic acid and, 397-398

5'-nucleotidase and, 592

paper chromatography of, 255-256

pentose nucleic acid ratios, 407

phosphotriesters and, 482

ultraviolet absorption of, 512

uracil formation and, 608
yeast pentose nucleic acid and, 402
Uridylyl-cytidine, synthesis, enzymatic,

566
Uroxanic acid, formation from uric acid,
617

Valine, 574
Vector, electric, 533
Vernine, 138, see also Guanosine
Vicine, 159

isolation and occurrence of, 99

structure of, 188
Virus, 81-82

composition, 375

deoxypentose nucleic acid extraction,
332

isolation from plants, 376-379

nucleic acid, analysis of, 206

nucleoproteins of, 383, 530
Viscosity,

deoxypentose nucleic acid and, 336,
338, 483-487, 489-490

deoxyribonuclease and, 580
Vitamin B,2, 27

ribose and, 13

Xanthine, 2, 112

cationic exchangers and, 218

color test for, 124

discovery of, 97-98

dissociation constant, 457, 458

fluorescence of, 247

formation of, 85, 117, 131

infrared absorption of, 547

occurrence of free base, 94

oxidation of, 121

paper chromatography of, 251, 253

reaction with triethylamine, 130

reducing agents and, 120

synthesis of, 132, 133, 134

ultraviolet absorption of, 498-499, 501,
502, 552
Xanthine dehydrogenase, adenine and,

595
Xanthine oxidase,

assay of, 611-612

carrier protein of, 613

692

INDEX

history, 609-611
inhibition of, 107, 612-613
iron and, 610
molybdenum and, 611
oxygen acceptors of, 610
specificity of, 612
Xanthopterin,
guanase and, 598-599
tautomerism of, 115
xanthine oxidase and, 612
Xanthosine, 36, 143, 157, 159
acid-base properties of, 477, 481
dissociation constant, 459
formation from guanosine, 142
hydrolysis of, 606
methylation of, 143
phosphorolysis of, 602, 603
synthesis of, 154

ultraviolet absorption of, 506, 508,
512, 552
Xanthylic acid, hydrolysis of, 39, 167-168
X-Ray diffraction,

deoxyribonuclease and, 541, 552, 624
dichroism and, 534

nucleic acid structure and, 461-470
structure determination and,
nucleosides, 452-455
purines, 451-452
pyrimidines, 447-451

Xylofuranose, 418
Xylofuranosides, synthesis of, 151
9-D-Xylopyranosyladenine, synthesis of
152-153

9-D-Xylopyranosyl-2-methylthio-adenine,

synthesis of, 152-153
Xylose, 12, 188

nucleosides of, 98

phosphates of, 42

ribose-5-phosphate preparation and.
63
Xylose-5-phosphate, 42
metabolism of, 41

Yeast,

deoxyribonuclease, 584-585
deoxyribonucleate isolation from 324-

325
pentose nucleic acid isolation from, 391
pentose nucleic acid, nucleotides of
402

Zinc,

deoxyribonuclease and, 581

uricase and, 619
Zoonucleic acid, 3
Zymogen granules, nucleases and, 625

\

t