CURRENTS

IN

BIOCHEMICAL

RESEARCH

Mi

3SB^^^^B3E3E3SQBBSE3E

E3

IB

I
I
I
I

3

E3

a m

I Marine Biological laboratory Library n

H Woods Hole, Mass. ID

D ' ra

B ID

i IB

B 01

H Presented by uJ

B 01

B Estate of Dr. ^tto LoeTfi C

B August, 1962

CURRENTS IN
BIOCHEMICAL RESEARCH

J

CURRENTS IN
BIOCHEMICAL RESEARCH

Ed^iitd^ h-j DAVID E. GREEN

Thirty-one essays charting the present course
of Biochemical Research and considering the
intimate relationship of biochemistry to
medicine, agriculture and social problems

DR. OTTO LOEWF

133 ' -^ c-.-r

/.-

INTERSCIENCE PUBLISHERS, INC., NEW YORK

1946

Copyright, 1946, by

Int'erscience Publishers, Inc.

215 Fourth Avenue, New York 3, N. Y.

Printed in the United States of America
by the Mack Printing Company, Easton, Pa.

PREFAG E

With the ever-increasing degree of specialization in scientific
research and with the terrifying rate of growth of technical nomen-
clature, men of science are literally compelled to know more and
more about less and less. The scientific literature furthers this trend,
since journals, textbooks (apart from those for students), and review
articles are written primarily for the specialist. There is an acute
need for stripping complex subjects and getting at the simple, essential
concepts which are basic to their appreciation. After all, the same
scientific principles are applicable to all fields of inquiry. The art
of presentation consists in the elimination of the barriers of terminology
which effectively conceal these fundamental principles. Currents in
Biochemical Research represents an attempt by some thirty research
workers to describe in as simple language as possible the important
developments in their own fields and to speculate a little on the most
likely paths of future progress. The aim of these essays has been to
excite the imagination and to provide glimpses of some of the fas-
cinating horizons of biochemical research. However, no populari-
zations were intended. The various contributors were asked to write
simply and provocatively but without sacrifice of scholarship. Dealing
as they do on the one hand with pharmacology, chemotherapy, public
health, genetics, photosynthesis, and agriculture and on the other
with considerations of organic, analytical, and physical chemistry,
they emphasize the focal position of biochemical research in biology,
chemistry, and medicine. It is hoped that this survey from so many
different points of view may assist biochemists, chemists, and medical
doctors in seeing biochemistry in clearer perspective and in its proper
relation to other fields of inquiry.

David E. Green

March, 1946

CONTENTS

1. The Gene and Biochemistry page

by G. W. Beadle 1 '\^

2. Viruses

by W. M. Stanley 13

3. Photosynthesis and the Production of Organic

Matter on Earth by H. Gaffron 25

4. The Bacterial Cell

by Rene J. Dubos 49 ^

5. The Nutrition and Biochemistry of Plants

by D. R. Hoagland 61

6. Biological Significance of Vitamins

by C. A. Elvehjem 79

7. Some Aspects of Vitamin Research

by Karl Folkers 89

8. Quantitative Analysis in Biochemistry

by Donald D. Van Slyke 109

9. Enzymic Hydrolysis and Synthesis of Peptide Bonds

by Joseph S. Frufon 123

10. Metabolic Process Patterns

by Fritz Lipmann 1 37

1 1 . Biochemistry from the Standpoint of Enzymes

by David E. Green 149

12. Enzymic Mechanisms of Carbon Dioxide Assimilation

by Severo Ochoa 165

13. Hormones

hy B. A. Houssay 1 87 ^^

VII

O ^ 1 1^

CONTENTS

14. Fundamentals of Oxidation and Reduction page
by Leonor Michaelis 207

15. Mesomeric Concepts in the Biological Sciences
by Herman M. Kalckar 229

16. Viscometry in Biochemical Investigations
by Max A. Lauffer 241

17. Isotope Technique in the Study of Intermediary
Metabolism by D. Rittenberg and David Shemin 261

18. Mucolytic Enzymes
by Karl Meyer 277

19. Some Aspects of Intermediary Metabolism
by Konrad Block 291

20. The Steroid Hormones
by Gregory Pincus 305

21. Plant Hormones and the Analysis of Growth
by Kenneth V. Thimann 321

22. Chemical Mechanism of Nervous Action
by David Nachmansohn 335

23. Some Aspects of Biochemical Antagonism
by D. W. Woolley 357

24. Chemotherapy: Applied Cytochemistry
by Rollin D. Hotchkiss 379

25. Biochemical Aspects of Pharmacology
by Arnold D. Welch and Ernest Bueding 399

26. Some Biochemical Problems Posed by a Disease of

Muscle by Charles L. Hoagland 41 3

[y 27. Physiology and Biochemistry

by Surgeon Captain C. H. Best 427

28. X-Ray Diffraction and the Study of Fibrous Proteins

by /. Fankuchen and H. Mark 439

29. Immunochemistry

by Michael Heidelberger 453

30. Social Aspects of Nutrition

by W. H. Sebrell 461

31. Organization and Support of Science in the United

States by L. C. Dunn 473

VIII

THE GENE
AND BIOCHEMISTRY

G. W. BEADLE, professor of genetics, school of biological

SCIENCES, STANFORD UNIVERSITY

/

"T IS both an accident of organic evolution and an indication
of man's lack of foresight that the organisms studied in most
detail by biochemists have not been those on which geneticists have
concentrated. It is natural that man should have a prejudice in
favor of himself, and it is therefore not remarkable that the urge of
medicine on biochemistry has been in the direction of specialization
on mammals, particularly on man himself. For obvious reasons,
bacterial biochemistry has likewise been well nourished through
medicine. Man has few inherent advantages for biochemical study
while the bacteria abound in them. But both are most difficult for
the geneticist — the one because of a long life cycle and social obstacles
to controlled matings, the other because of the absence of a sexual
cycle without which the geneticist cannot use his particular methods.
The geneticist, on the other hand, has chosen to make the vinegar fly
and Indian corn the classical organisms of his science. Both suffer
disadvantages to the biochemist in not lending themselves readily to
culture under precisely defined environmental conditions. Neither
can be grown conveniently on a medium completely known from a
chemical standpoint.

In spite of this situation and additional impediments arising
through divergence in outlook, such persons as Garrod, Onslow (n^e
Wheldale), Troland, Goldschmidt, Wright, Haldane, and others have

G. W. BEADLE

urged that the two fields have much in common and that each stands
to profit through contact with the other. Through the eff"orts of these
individuals and others of like mind there are many instances known
in which the relation of genetics to biochemistry is so clear that it can
no longer be disregarded by intelligent investigators in either field. In
fact, from this relation there tends to emerge a new interest, known as
biochemical genetics, which promises to tell us what the genes do and
how they do it, on the one hand, and to lead us to further knowledge
in the ways of biosynthesis on the other. In both directions there
obviously lie many opportunities.

One of the earliest instances in which a Mendelian trait could
be interpreted in terms of specific chemical reactions is that involving
the human disease known as alcaptonuria. In individuals homo-
zygous for the mutant gene responsible for this character, 2,5-di-
hydroxyphenylacetic acid (homogentisic acid or alcapton) is excreted
in the urine instead of being broken down to carbon dioxide and
water, as it is in persons receiving the normal form of the alcaptonuric
gene from one or both parents (15). Homogentisic acid is oxidized
to a black pigment on exposure to air and it is this process that is re-
sponsible for darkening of the urine, the most striking symptom of the
disease. According to Gross (cited by Garrod), alcaptonurics lack a
specific enzyme found in the blood of normal persons which catalyzes
the degradation of homogentisic acid. Alcaptonuria therefore repre-
sents the first recorded instance in which it could be said that a par-
ticular chemical reaction is controlled by a known gene through the
mediation of a specific enzyme.

Within the past dozen years, additional examples have become
known in which organisms unable to carry out specific reactions differ
in a single gene from their chemically more successful relatives. In
flower pigment synthesis, for example, the formation of carotenoids,
anthocyanins, anthoxanthins, chalcones, and fiavocyanins is known
to be genetically controlled in one plant or another (7,24). Specific
oxidations of pelargonidin derivatives to cyanidin analogues and of
cyanidin compounds to delphinidin counterparts are dependent on
the activities of specific genes. The addition of sugars to anthocyani-
dins through glycosidal linkages and the transformation of the an-
thoxanthin quercetin-3-glucoside to the corresponding cyanidin-3-
glucoside are likewise unable to proceed if specific genes are modified.

THE GENE AND BIOCHEMISTRY

Foiling (14) and Penrose (35) have shown that the genetically
determined failure to oxidize phenylpyruvic acid in man is invariably
associated with subnormal mentality. Here again there appears to be
an intimate relation between a particular gene and a specific chemical
reaction. Because of its obvious importance to an understanding of
the mechanisms underlying mental processes, this case is of particular
interest. It is of course related metabolically to alcaptonuria in so far
as phenylalanine is concerned in both. Other abnormalities in
phenylalanine-tyrosine metabolism are also known (15,19).

Most remarkable progress has recently been made in under-
standing the genetic and chemical mechanisms of sex determination
and differentiation in the green alga, Chlamydomonas, by Moewus,
Kuhn, and co-workers. From the carotenoid pigment, protocrocin,
there is derived through cleavage the motility hormone, crocin, and a
female-determining hormone known as gynotermone. This cleavage
is known to be genetically controlled. In genetically male individuals
gynotermone is hydrolyzed to a male-determining hormone known as
androtermone. Under the direction of specific genes the cis and trans
forms of the motility hormone, crocin, are converted into the corre-
sponding cis and trans dimethyl esters of crocetin. In various specific
mixtures, these serve as gamones, i. e., they render individuals of the
specific genetic constitutions capable of conjugation. The relations
between genes and chemical reactions disclosed by this work support
the thesis that genes act in directing specific processes. The work on
Chlamydomonas is so spectacular and its importance so great that inde-
pendent confirmation is desirable (see 7,28,41).

The splitting of specific di- and trisaccharides by yeasts is under
genetic control, as shown by Winge and Laustsen (54) and Lindegren,
Spiegelman, and Lindegren (25). It appears that the genes concerned
determine whether or not specific enzymes are present in active form.
Somewhat similar situations are known in the rabbit, where Sawin
and Glick (37) have shown that the activity of the enzyme, atropine
esterase, is dependent on the presence of the normal allele of a par-
ticular gene, and in white clover, where an enzyme responsible for
hydrolysis of specific cyanogenetic glucosides is known to show a similar
dependence on a gene (2,9).

In the bread mold, Neiirospora, Srb and Horowitz (44) have
shown that there is present an ornithine cycle essentially similar to

G. W. BEADLE

that postulated by Krebs and Henseleit for the mammalian liver. In
the bread mold it is known that mutation in any one of seven different
genes will interi'upt the synthesis of ornithine or its conversion to
arginine. So far as the data go, they are consistent with the assump-
tion that each of the seven genes is normally concerned with a different
chemical reaction in the system. It is an interesting point that it was
possible to establish the presence of the ornithine cycle in Neurospora
because of the existence of the mutant strains indicated.

Tatum and Bonner (50) have shown that tryptophan is normally
synthesized in Neurospora through the condensation of indole and
serine. Evidently the indole is somehow derived from anthranilic
acid, for there exist two mutant strains, one of which accumulates
anthranilic acid when it is grown under suitable conditions, while
the other is able to grow normally when supplied with anthranilic
acid in place of indole or tryptophan (51). The gene by which the
first strain differs from wild type is evidently concerned with the reac-
tion by which anthranilic acid is converted into indole, whereas the
mutant gene of the second strain appears to be concerned with failure
of some reaction essential to the synthesis of anthranilic acid. It is
obvious, in this case, that genetics has provided a tool of great useful-
ness in investigating the biosynthesis of the important amino acid,
tryptophan.

Relations similar to those mentioned above are known for other
biosyntheses in the bread mold and in other organisms. By following
methods developed by Beadle and Tatum (8), it has been possible to
obtain a series of mutant strains of Neurospora in each of which some
particular reaction has been blocked. These are concerned with the
synthesis of amino acids, vitamins, purines, pyrimidines, and other
compounds of biological importance (6,21,48,49).

We can be sure from such cases as those just cited that genes
function in directing biochemical reactions. We know, further, that
this direction may involve enzymes as intermediates between gene and
reaction. All our information is consistent with the hypothesis that
in all cases in which genes control specific reactions they do so indirectly
through enzymes. In other words, genes direct enzyme specificities,
and enzymes control reactions. This is not a new idea. Bateson (4),
Moore (29), Troland (52), Goldschmidt (16), Muller (30), Alexander
and Bridges (1), Haldane (18,19), Wright (55), and others have sug-

THE GENE AND BIOCHEMISTRY

gested it. We are only now beginning to do something definite about
it from an experimental standpoint. Since the specificities of enzymes
are referable to protein specificities, the hypothesis implies that genes
direct protein specificities. In this case we might expect that the
specificities of proteins other than those found in enzymes would show
a direct relation to genes. This is indeed the situation as evidenced
by the fact that in many organisms a general one-to-one relation be-
tween genes and antigens has been shown (22,47). It is true that a
few deviations from this correspondence are known, but they may well
represent instances in which antigens have specificities made up of two
components, each corresponding to one gene.

If we knew the chemical nature of genes, we should be in a
much better position than we are now to determine how they direct
protein specificities. Direct chemical analyses of whole chromosomes
show them to be largely nucleoprotein (27), which suggests that genes
too are nucleoproteins. But since chromosomes probably contain
much nongenic material, the deduction is not too satisfying. Ultra-
violet radiation induces gene mutation; and its efficiency in this re-
spect varies with wave length in the same way as does its absorption
by nucleic acid (20,45), strongly indicating that the energy eff'ective
in producing mutations in genes is absorbed by nucleic acid. The
simplest assumption possible is that this is so because the nucleic acid
is part of the gene.

The similarity of genes and viru.ses constitutes a third line of evi-
dence concerning the chemical nature of genes. Both have the property
of self-duplication, which in both cases is dependent on the presence of
a series of compounds such as those found in the living cell. Genes
and viruses appear to be within the same size range (46). Both are
capable of undergoing mutation to new forms which have altered
biological activities but retain the power of self-duplication (46).
Since viruses and genes have so many properties in common, it is
probable that they are similar in chemical makeup. Following
Stanley's isolation of crystalline tobacco mosaic virus, several other
viruses have been prepared in pure form and all have been shown to
be nucleoproteins (5,12,46). The circumstantial evidence that genes,
too, are nucleoproteins, or at least contain nucleoproteins as essential
parts, is therefore substantial.

In duplicating themselves, genes have been assumed to act as

G. W. BEADLE

master molecules or models from which exact copies are made (11,18,
19,30,31,55). If this is so, their action may be visualized as one of
directing the construction of specific protein types plus whatever other
component parts genes may have. If the specificities of proteins
generally are copied from genes, the observed relations between genes
and enzymes and between genes and antigens should follow. For
every specific protein there should exist a gene carrying this same
protein. For every enzyme and antigen type there likewise should
be a gene. Because of a general tendency of mutation pressure to
eliminate genes that are of no advantage to the organisms, it might
be expected that for every protein type there would be only one corre-
sponding gene. The experimental evidence appears to support this
general interpretation, although it must be recognized that, in dealing
with genetic traits that can be described in terms of chemical reactions,
there may be an unavoidable selection of those cases in which gene
action is relatively simple.

However proteins and other components of genes are synthesized
— whether by an orthodox stepwise mechanism or by some as yet un-
known mechanism by which many component parts are simultaneously
directed into their proper places by the master molecule (11) — the pre-
cursors of proteins, nucleic acids, and whatever other parts genes may
have, must be synthesized. Their synthesis will involve many enzymes
and a corresponding number of genes. Thus, before one gene can
determine the specificity of a new protein molecule, many other genes
must have acted. This amounts to saying that, in any multigenic
organism, the genes constitute a highly organized system, just as the
chemical reactions they direct are integrated in time and space in a
manner characteristic of a particular species. Furthermore, while a
particular gene will have only one primary action in determining
specificity of an enzyme or an antigen, the final physiological conse-
quences of a change in a single gene will be manifold. This can be
appreciated when one considers the consequences of depriving an
organism such as a rat of thiamin. The final consequence is of course
death, but before death occurs a series of changes of increasing com-
plexity take place. These can be brought about in the rat by remov-
ing thiamin from the diet. In the bread mold, which normally syn-
thesizes thiamin, the same end result can be effected as a result of an
analogous series of changes initiated by replacing a normal gene

6

THE GENE AND BIOCHEMISTRY

necessary for thiamin synthesis with a defective form of the same gene.
In one particular case, the primary action of the gene is presumed to
be in directing the specificity of the enzyme catalyzing the reaction by
which thiazole and pyrimidine are combined (49). Viewed in this
way, an understanding of gene action does not appear hopelessly
difficult even though the final effects of a single gene change may
involve alterations so complex as to defy complete description. Griine-
berg (17) has pointed out that a similar type of interpretation in terms
of one primary action of a given gene is tenable in the case of certain
hereditary developmental defects in the mouse and rat that at first
sight appear to involve several unrelated changes in the organism.
That the gene has a functional as well as structural unity is therefore
a hypothesis that has demonstrated its heuristic value. Until evidence
with which it is inconsistent is presented, it will no doubt continue to
play an important role in our concepts of what the gene is and how it
acts.

As Troland (52), Muller (30), Alexander and Bridges (1), Oparin
(33), Plunkett (36), and others have pointed out, the similarity of viruses
and genes suggests that the first living structures, i. e., those with the
power of self-duplication, were probably somewhat similar to present-
day viruses with the important difference that they were free-living.
The evolution of systems of such units, each acquiring the property of
directing the specificity of an enzyme or other protein, would be ex-
pected to give rise to a series of forms of increasing complexity such as
we see today in the larger and more complex viruses, the rickettsias,
bacteria, and higher organisms. It is probable that the present
viruses and rickettsias are not relics of these ancestral forms but are
forms secondarily derived through specialization in connection with
parasitism (10). The true ancestral types must have been capable of
multiplying outside living cells in a kind of environment which, because
of the presence of many organisms, is no longer likely to exist (33). In
terms of genes directing chemical reactions through their control of
enzyme specificities it is possible to imagine how, in principle, these
simple forms evolved in the direction of the more highly specialized
and complex forms of multicellular plants and animals (56), although
it is of course not easy to visualize the way in which the process occurred
in detail in particular instances.

In the specialization of higher animals with respect to their nutri-

G. W. BEADLE

tion, it is possible to suggest a scheme of evolution that has some sup-
port at least in analogy. It has become increasingly evident that, with
respect to their need for and use of vitamins of the B group, purines,
pyrimidines, choline, amino acids, and other compounds, all cellular
organisms are fundamentally very similar (23,26,53). To consider
a specific example, carboxylase is presumably present in all protoplasm
and apparently always contains thiamin as thiamin pyrophosphate.
Many organisms, e. g., most plants, are able to synthesize the thiamin
they need, while others are dependent on an external supply of this
essential compound. From an evolutionary standpoint, this difference
is presumably determined by whether or not it is of advantage to a par-
ticular organism to synthesize thiamin. Evidently for Neurospora
there is selective advantage in being able to carry out this synthesis
for we find in wild strains that all essential genes concerned with it are
present in active form. In mammals, on the other hand, thiamin is
presumably so frequently present in the diet that the genes originally
concerned with its elaboration have been permitted by natural selection
to become inactive so far as thiamin synthesis goes. It may well be
that they have not disappeared entirely but have been modified so
as to enable the mammal to carry out chemical reactions of which the
bread mold is incapable. In a similar way, mammals have become
specialized through loss of ability to synthesize other vitamins, the
indispensable amino acids, and other compounds. With the develop-
ment of parasitism it would be expected that still further loss in syn-
thetic ability would be encountered. As Knight (23), Lwoff" (26),
Schopfer (38), and others have pointed out, this is indeed the case.
The work on Neurospora makes it most probable that the dropping out
of specific chemical reactions no longer of selective advantage is the
result of gene mutation. The limit of such parasitic specialization is
probably represented in the molecular viruses that have lost all power
of heterosynthesis and have retained only the one property essential
for their continued existence in an environment in which all necessary
compounds are available — the property of autosynthesis.

One may quite properly raise the question as to the course of
positive evolution in terms of chemical reactions — how are new syntheses
developed in the course of organic evolution? Unfortunately, the
experimental evidence bearing on this is meager, which is not surpris-
ing, for obviously it should be much easier to destroy or inactivate a

8

THE GENE AND BIOCHEMISTRY

complex self-duplicating unit than to modify it so as to give it a new
and useful property without sacrificing its power of self-duplication.
The first self-duplicating unit must have evolved from nonliving matter
at some time, and more complex forms must have evolved from it —
the alternative is some form of special creation. There would seem
to be less difficulty in imagining a primitive "protogene" mutating to a
true gene with a heterocatalytic property than its spontaneous origin
in the first place, even if, as Oparin (33) supposes, it arose in a world
containing preformed organic molecules of many kinds. Nor is there
any apparent reason why such protogenes could not nmtate in many
diflferent directions in order to give rise to many different organic
catalysts. In present-day cellular organisms there exists a possible
mechanism for acquiring totally new reactions. Occasionally, through
accident, one or more genes become duplicated, i. e., a small segment
of a chromosome occurs twice in every set. The duplicated genes
will be unnecessary to the organism and will be expected to disappear
through loss mutations, since such mutations are not disadvantageous.
But such a duplicate gene may occasionally undergo mutation in such
a way that it directs the formation of an entirely new enzyme. If this
new enzyme should happen to catalyze a reaction that improved the
organism in competition with its relatives, the new reaction would be
retained. Such new reactions might add new compounds or they
might bring about the reverse of the specialization process, which leads
in the direction of parasitism. In this way, as Horowitz (20a) has
pointed out, the first primitive organisms might gradually have built
up systems of synthesis which freed them of their dependence on
preformed organic molecules originally present in the environment.

Through such advances as hav-e been indicated we appear to
be moving rapidly in the direction of a better understanding of what
genes are and what they do. We are no longer content with a knowl-
edge of the laws by which they are transmitted from one generation
to the next. We see that they are basic functional units of the organism
and that, by taking advantage of their tendency to mutate, we can use
them as powerful tools in determining the course of biosynthesis and
in imderstanding other aspects of metabolism. Their relations to
enzymes and antigens are becoming known. Precisely how they
function in duplicating themselves and in directing the specificities of
proteins, nucleic acids, and possibly other large molecules is a question

G. W. BEADLE

for the future. But there can be no doubt that the years that lie
ahead will be exciting ones in this field. The work of Avery and his
co-workers (3) on the transformation of types in Pneumococcus and that
of Emerson (13) and of others suggests that we may one day learn to
direct gene mutations in predetermined ways. Work on enzymes (32)
and viruses (5,46) is so closely related to the general problem of gene
structure and gene action that only short steps appear to be necessary
to bridge the gaps that separate them. Nucleic acid certainly plays
an important role in gene action and in protein synthesis (34,39,40),
and it is not too much to hope that this role will be made clear in the
near future. The relation of genes to cytoplasmic elements is not well
understood, but after many years in which discouragingly little prog-
ress has been made, important leads are being followed by Sonneborn
(42), Spiegelman (43), and others. After half a century of growth,
genetics seems to be assuming a position in the broad field of biology
in which its close relations to evolution, development, physiology, and
biochemistry are now more evident.

References

(1) Alexander, J., and Bridges, C. B., in Colloid Chemistry. Vol. II, Chemical
Catalog Co., New York, 1928.

(2) Atwood, S. S., and Sullivan, J. T., J. Heredity, 34, 311 (1943).

(3) Avery, O. T. F., McLeod, C. M., and McCarty, M., J. Exptl. Med.,
79, 137 (1944).

(4) Bateson, W., MendePs Principles oj Heredity. Cambridge Univ. Press,
London, 1909.

(5) Bawden, F. C, Plant Viruses and Virus Diseases. Chronica Botanica,
Waltham, 1943.

(6) Beadle, G. W., Physiol. Revs., 25, 643 (1945).

(7) Beadle, G. W., Chem. Revs., 37, 15 (1945).

(8) Beadle, G. W., and Tatum, E. L., Proc. Natl. Acad. Sci. U. S., 27, 499
(1941).

(9) Corkill, L., N. Zealand J. Sci. Tech., B23, 178 (1942).

(10) Darlington, C. D., Nature, 154, 164 (1944).

(11) Delbruck, M., Cold Spring Harbor Symposia Quant. Biol, 9, 122 (1941).

(12) Delbruck, M., in Advances in Enzymology, Vol. II. Interscience, New
York, 1942, p.l.

(13) Emerson, S., Proc. Natl. Acad. Sci. U. S., 30, 179 (1944).

(14) FoUing, A., Z- physiol. Chem., 227, 169 (1934).

lO

THE GENE AND BIOCHEMISTRY

(15) Garrod, A. E., Inborn Errors of Metabolism. 2nd ed., Oxford Univ.
Press, London, 1923.

(16) Goldschmidt, R., Physiological Genetics. McGraw-Hill, New York,
1938.

(17) Gruneberg, H., The Genetics oj the Mouse. Cambridge Univ. Press,
London, 1943.

(18) Haldane, J. B. S., "The biochemistry of the individual," in Perspec-
tives in Biochemistry. Cambridge Univ. Press, London, 1938.

(19) Haldane, J. B. S., New Paths in Genetics. Harper, New York, 1942.

(20) Hollaender, A., and Emmons, C. W., Cold Spring Harbor Symposia Quant.
Biol., 9, 179 (1941).

(20a) Horowitz, N. H., Proc. Natl. Acad. Sci. U. S., 31, 153 (1945).

(21) Horowitz, N. H., Bonner, D., Mitchell, H. K., Tatum, E. L., and
Beadle, G. W., Am. Naturalist, 79, 304 (1945).

(22) Irwin, M. R., and Gumley, R. W., Am. Naturalist, 77, 211 (1943).

(23) Knight, B. C. J. G., Med. Research Council (Brit.), Special Rept. Series,
No. 210 (1936).

(24) Lawrence, W. J. C, and Price, J. R., Biol. Rev. Cambridge Phil. Soc,
15, 35 (1940).

(25) Lindegren, C. C, Spiegelman, S., and Lindegren, G., Proc. Natl.
Acad. Sci. U. S., 30, 346 (1944).

(26) Lwofr, A., Ann. inst. Pasteur, 61, 580 (1938).

(27) Mirsky, A. E., in Advances in Enzymology, Vol. III. Interscience, New
York, 1943, p. 1.

(28) Moewus, F., Ergeb. Biol, 18, 287 (1941).

(29) Moore, A. R., Univ. Calif. Pub. Physiol., 4, 9 (1910).

(30) MuUer, H. J., Am. Naturalist, 56, 32 (1922).

(31) Muller, H. J., Cold Spring Harbor Symposia Quant. Biol., 9, 290 (1941).

(32) Northrop, J. H., Crystalline Enzymes. Columbia Univ. Press, New
York, 1939.

(33) Oparin, A. I., The Origin of Life. Trans, by S. Morgulis. Macmillan,

New York, 1938.

(34) Painter, T. S., Texas Repts. Biol. Med., 2, 206 (1944).

(35) Penrose, L. S., Lancet, 2, 192 (1935).

(36) Plunkett, C. R., in Colloid Chemistry. Vol. V, Reinhold, New York,
1944.

(37) Sawin, P. B., and Click, D., Proc. Natl. Acad. Sci. U. S., 29, 55 (1943).

(38) Schopfer, W. H., Plants and Vitamins. Chronica Botanica, Waltham,

1943.

(39) Schultz, J., Cold Spring Harbor Symposia Quant. Biol., 9, 55 (1 941).

(40) Schultz, J., in Colloid Chemistry. Vol. V, Reinhold, New York, 1944.

I I

G. W. BEADLE

(41) Sonneborn, T. M., Cold Spring Harbor Symposia Quant. Biol., 10, 111
(1942).

(42) Sonneborn, T. M., Ann. Missouri Bolan. Garden, 32, 213 (1945).

(43) Spiegelman, S., .inn. Missouri Botan. Garden, 32, 139 (1945).

(44) Srb, A. M., and Horowitz, N. H., J. Biol. Chem., 154, 129 (1944).

(45) Stadler, L. J., and Uber, F. M., Genetics, 27, 84 (1942).

(46) Stanley, W. M., "Chemical structure and the mutation of viruses," in
Virus Diseases. Cornell Univ. Press, Ithaca, 1943.

(47) Stranskov, H. H., Physiol. Revs., 24, 445 (1944).

(48) Tatum, E. L., Ann. Rev. Biochem., 13, 667 (1944).

(49) Tatum, E. L., and Beadle, G. W., Ann. Missouri Botan. Garden, 32, 125
(1945).

(50) Tatum, E. L., and Bonner, D., Proc. Natl. Acad. Sci. U. S., 30, 30 (1944).

(51) Tatum, E. L., Bonner, D., and Beadle, G. W., Arch. Biochem., 3, 477
(1944).

(52) Troland, L. T., Am. Naturalist, 51, 321 (1917).

(53) Williams, R. R., Science, 94, 471 (1941).

(54) Winge, O., and Laustsen, O., Compt. rend. trav. lab. Carlsberg, Ser.
physiol., 22, 337 (1939).

(55) Wright, S., Physiol. Revs., 21, 487 (1941).

(56) Wright, S., Biol. Symposia, 6, 337 (1942).

12

VIRUSES

W. M. STANLEY, member of the rockefeller institute for

MEDICAL RESEARCH, PRINCETON; MEMBER OF THE NATIONAL ACADEMY

OF SCIENCES

D

|URING the past ten years there has converged on a group
of small, infectious, disease-producing entities, known as
viruses, an array of scientific talent almost as diverse in nature as the
Allied forces that were brought to bear on the Axis powers. The
viruses are responsible for many diseases of man, animals, plants, and
bacteria. There is no single criterion by means of which viruses can
be differentiated from bacteria, yet the virus group has been segregated
by means of certain general characteristics. Among the most important
of these are small size, the ability to reproduce or multiply when within
the living cells of a given host, the ability to change or mutate during
multiplication, and the inability to reproduce or grow on artificial
media. The sole means of recognizing the existence of a virus is
provided by the manifestations of disease which result from the growth
of the virus. Although pathologists have studied viruses for over
fifty years and have added much to our knowledge, the new attack on
viruses has been spearheaded by chemistry, chiefly biochemistry and
physical chemistry, and their allied disciplines. The impact of these
diverse disciplines on viruses has been accompanied by reverberations
and repercussions, which, however, bode ill for viruses and good for
scientific thought. Clcrtain dominant facts stand out in Ijold relief
and these are understood and accepted by chemists and pathologists
alike, but there is an undercurrent of indecision where neither feels

13

W. M. STANLEY

quite sure of himself. This indecision becomes apparent when one
considers the mode of reproduction and mutation of viruses, whether
viruses are molecules or organisms, or whether some are molecules
with others being organisms, or whether viruses represent a new type
of structure, hitherto unrecognized and undefined, and whether
one should speak of a solution or of a suspension of a given virus.
For the moment, because of the lack of precise experimental data,
discussion of these questions must remain more or less philosophical
in nature. But since some of these questions pertain to the very
nature of life itself, and others to fundamental physicochemical prob-
lems, it is obvious that they are of great importance. During recent
years the chemist, as well as the pathologist, has become aware of
the limitations of his tools. It has become obvious that, if the tre-
mendous problems posed by the viruses are to be solved, it will be
necessary to forge new tools, both material and of the mind, and to
carry forth the attack on a united front with a new perspective and with
renewed courage and vigor. It is the purpose of this short essay to
chart briefly the roads that the chemist and pathologist have already
constructed into the field of viruses and to attempt to outline, in general
terms, the manner in which these can be continued until they join and
provide, through a common effort, a broad highway of fact and in-
formation leading from the lowly electron to the lofty heights of man.
Viruses range in size from about 10 m^i to about 300 m/x. Cer-
tain small viruses, such as alfalfa mosaic virus, are smaller than certain
accepted protein molecules, such as the Busycon hemocyanin molecules.
On the other hand, certain large viruses, such as vaccine virus, are
larger than certain accepted organisms, such as the minimal repro-
ductive units of the microorganisms of the pleuropneumonia group.
With respect to size, therefore, the viruses overlap with molecules at
one extreme and with organisms at the other extreme. Since the
discovery of viruses by Iwanowski, a plant pathologist, in 1892, and by
Beijerinck, a chemist and botanist, in 1898, many investigations on
the sizes of diflferent viruses have been carried out. Ultraviolet-light
microscopes and ultrafiltration were used in most of these studies.
Following the isolation of essentially pure virus preparations, methods
involving ultracentrifugation, diffusion, x-ray diffraction, viscosity,
osmotic pressure, and stream double refraction measurements were
used with great success. Recently the chemist and pathologist have

VIRUSES

joined in the extensive use of the newly developed electron microscope.
This instrument covers the entire range of sizes occupied by viruses
and has proved, and will doubdess continue to prove, of the greatest
value in the estimation of the sizes of viruses. In those cases in which
more than one method has been employed, good agreement has
usually been secured. The occasional discrepancies have been found
to be due to errors or to a failure to appreciate the limitations of a given
method, and generally have been resolved. At present the sizes of
several viruses are well established and the values are accepted by both
chemists and pathologists.

Tobacco mosaic, the first virus to be discovered, was also the
first virus to be prepared in essentially pure form and subjected to
extensive chemical study. No difference was noted in virus samples
prepared from a wide variety of hosts or from the same kind of host at
different times of the year. The virus particles were found to consist
of about 6% nucleic acid of the ribose type and about 94% protein.
The exact nature of the linkage between protein and nucleic acid is
unknown, but it appears to be considerably stronger than that which
exists in the sperm nucleoproteins. The protein component contains
definite and reproducible amounts of over thirteen amino acids. It is
of interest that, in contrast to the sperm nucleoproteins, there does not
appear to be an excess of basic amino acids. The single virus particles
are about 280 m^u in length and 15 m^t in diameter. Tobacco mosaic
virus activity has never been demonstrated to be associated with smaller
particles. However, there is good evidence that a single virus particle
is built up from similar subunits fitted together in a hexagonal lattice
to yield the final structure which possesses virus activity. The nucleic
acid of this final structure appears to exist in the form of eight threadlike
units laid down along the length of the particle. Because of the repeat
pattern within a single virus particle, it can be regarded as a sub-
microscopic crystal. In addition, these single virus particles can
aggregate with a two-dimensional regularity to form large needlelike
crystals that are readily visible with a low-power hand lens. Of
especial interest and significance is the fact that the virus particles
appear to be utterly devoid of water and of any enzymic activity
other than virus activity. The complete lack of water and the crystal-
like inner structure of the individual virus particles would appear to
preclude the existence of metabolism of the type usually associated with

15

W. M. STANLEY

organisms. Yet, when introduced into the cells of susceptible h(jsts,
these particles can direct or enter into the metabolic chain of events
of the cell.

The rod-shaped anhydrous tobacco mosaic virus particle is
representative of a small group of viruses, but is certainly not repre-
sentative of all viruses, for most viruses that have been studied ade-
quately appear to be essentially spherical in shape and hydrated.
Among the viruses that have been obtained in essentially pure form and
studied in some detail are alfalfa mosaic virus with a diameter of about
17 m/i, tobacco ring spot virus with a diameter of about 19 m^u, tomato
bushy stunt virus with a diameter of about 26 m;u, equine encephalo-
myelitis and rabbit papilloma viruses with diameters near 40 m/i,
influenza virus with a diameter of about 100 m/x and vaccine virus
with a diameter of about 225 m/x. Of these, tomato bushy stunt virus
is the only one that has been obtained in crystalline form. This virus,
which contains about 17% nucleic acid, and about 83% protein,
crystallizes in the form of large, beautiful, rhombic dodecahedra.
The particles of bushy stunt virus appear to be strictly homogeneous
with respect to size, shape, and density; hence the case for regarding
these particles as molecules is as good as for any protein.

The papilloma virus appears to be a nucleoprotein containing about
8% nucleic acid and little or no lipid. The particles of the equine
encephalomyelitis virus appear to be a liponucleoprotein complex
containing about 48% lipid, about 5% nucleic acid, and protein plus
a small amount of excess carbohydrate. Data on influenza virus
indicate that the 100 m/j, particle has a water content of about 60%
by weight, with the solid portion being composed of about 65% pro-
tein, about 25% lipid, about 7% carbohydrate, and a very small
amount of nucleic acid. Vaccine virus is the largest and most complex
virus that has been subjected to chemical investigation. The prepara-
tions were found to contain protein, lipid, carbohydrate, and thymus
nucleic acid in concentrations not materially different from those found
in bacterial cells. The vaccine virus preparations were also found to
contain phosphatase, catalase, lipase, biotin, riboflavin, flavin-adenine-
dinucleotide, and apparently significant and reproducible amounts
of copper. It is exceedingly difficult to prove that all of these repre-
sent integral components of vaccine virus, but it must be regarded as
significant that this large and complicated virus appears to retain

i6

VIRUSES

certain enzymic activities quite tenaciously, whereas tobacco mosaic
and bushy stunt viruses appear lo possess no enzymic activities other
than that of virus activity. It is also of interest that, in marked con-
trast to bacteria and other ceils, tobacco mosaic and influenza viruses
contain negligible amounts of the B vitamins. Electron micrographs
of vaccine virus, as well as of certain bacteriophages, have revealed
the presence within individual particles of an internal structure con-
sisting of a pattern of granules.

As a whole, the data now available on viruses indicate that, as
one goes from the small to the large viruses, there is, with increase in
mass, an increase in complexity of composition, structure, and function.
The viruses appear to provide, in truth, a bridge between proteins
and organisms. Indeed, if one wishes to regard the transformation
agent of the pneumococcus as a virus, the bridge could be extended to
nucleic acid. If one were starting out anew to construct a link between
the molecules of the chemist and the organisms of the pathologist or
bacteriologist, it is difficult to visualize how it would be possible to
improve upon what Nature has already provided. The structural
complexity encountered in tomato bushy stunt virus nucleoprotein is
but little more complex than that of hemoglobin and no more complex
than that of the hemocyanins. Physically and chemically these behave
as molecules; and if it were not for the virus activity of the bushy stunt
nucleoprotein it would not be given a second thought. Between bushy
stunt virus and vaccine virus. Nature has provided a continuous series
of structures of gradually increasing mass and complexity, all linked by
a common biological property, virus activity. Vaccine virus is as
large as some organisms that can be grown on artificial media; and if
vaccine virus could be grown on artificial media it would be accepted
generally as an organism. Even larger structures exist which cannot
be grown on artificial media and which are just as fastidious as vaccine
virus with respect to growth requirements; yet these are accepted as
organisms. Nature has provided us with an accomplished fact, and it
is time for all to brush away the barriers of the mind and to recognize
the possibilities that are provided by the viruses. Although viruses are
disease-producing agents and have caused untold suffering, the com-
plete acceptance of Nature's dubious gift may not be without recom-
pense. For its acceptance and exploitation may provide ihc key to
broad and wonderful vistas.

17

W. M. STANLEY

Pathologists and bacteriologists have labored long and arduously
with viruses. They have found that a given virus will reproduce only
within the cells of certain specific living hosts, and that, although some
viruses will reproduce within the cells of several different hosts, other
viruses will multiply only within the cells of one given host or sometimes
only within certain specialized cells of that host. They have shown
that the primary pathological changes produced in cells by viruses are
either proliferative or degenerative in character. In some virus dis-
eases, such as yellow fever, poliomyelitis, and tobacco necrosis, de-
generative changes predominate; but in many, as in smallpox and fowl
pox, both proliferation and necrosis occur. Still other virus diseases,
such as Rous chicken sarcoma, Shope rabbit papilloma, and tobacco
enation mosaic, are characterized by a rapid and unorganized cellular
proliferation.

Long before the discovery of viruses, a means was recognized of
protecting man against the virus disease, smallpox. This was achieved
by vaccination with active virus, presumably altered by passage in an
unnatural host. However, it has only been in recent years that there
has come a full realization of the great benefits, both with respect to
methods of protection against virus diseases and to the study of the
viruses themselves, that can be achieved through the use of new virus
hosts. Yellow fever has been eliminated as a major health problem,
and much has been learned of the virus because the virus was taken
from man and grown in monkeys, in mice, and in chick embryos.
The possibility of a recurrence of the 1918 influenza epidemic, which
killed more people than have died from combat activities in World
Wars I plus II (to date of writing), has been reduced and perhaps
eliminated because of the production of a vaccine which was made
possible by the growth of this virus in the chick embryo. Truly re-
markable progress has already been made in the study of viruses
and in the prevention of virus diseases, and it is to be expected that
this progress will continue. Yet withal, the one fundamental and all-
important problem posed by the viruses — that of the mode of virus
reproduction — remains unsolved. For many years little hope for a
solution of this problem was held, for viruses were generally regarded
as living organisms and the nature of life was considered to be a hal-
lowed, insoluble secret. However, chemists recognized in the virus
activity of certain crystallizable nucleoproteins a type of biological

i8

VIRUSES

activity somewhat akin to that possessed by certain protein hormones
and enzymes. To them, virus activity appeared no less wonderful
and possibly no more complicated than the changes that can be induced
within cells by enzymes and hormones. The mental barrier of the
living state is being eliminated gradually and chemists are recognizing
and accepting Nature's gift of the viruses. The true issue is only
beclouded by the insistence in some quarters for a decision as to whether
a given structure is living or inanimate. The fundamental meaning-
lessness of such terms has been commented on before and is becoming
ever more apparent. In the meantime, the requirements for the
solution of the riddle of virus reproduction — perhaps the most impor-
tant problem in all of biochemistry — are becoming clearer.

Many studies on the nature and mode of virus reproduction are in
progress, and it seems certain that such studies will increase in the
future, not only in volume, but also in scope and in diversity. It is
possible that the solution of this very important and fundamental
problem will not be realized until after new weapons of the hand and
mind have been brought into play. So far, the most spectacular ad-
vances have been made along three lines, each of which merits con-
siderable further attention: (a) studies on the very favorable bacterial
cell-bacteriophage system; (b) studies on the changing of the chemical
structure of a virus by means of known chemical reactions; and (c)
studies on the nature of the differences in chemical structure that are
responsible for the existence of virus strains. The bacterial cell-
bacteriophage system provides an extraordinary opportunity to follow
the interaction of a virus with its host cell. The host cells can be
grown in vitro on artificial media in large quantities and under constant
and reproducible conditions. The metabolism of the host can be con-
trolled to a certain extent and analyses can be made on the system
throughout the reaction. The bacterial and phage or virus materials
can be differentiated up to the entrance of the virus into the cells and
following the lysis of the cells. Studies on this system have permitted
the conclusion that multiple infection of a bacterial cell with several
virus particles of the same type has the same effect, both qualitatively
and quantitatively, as infection with a single virus particle. It was
also found that infection of a bacterial cell with two kinds of virus
particles resulted in the reproduction of only one kind and the sup-
pression of growth of the other kind. However, more than one virus

19

W. M. STANLEY

can multiply in a single cell if the viruses are sufficiently distinct in
their requirements. These and similar results have permitted many
stimulating inferences regarding the mode of virus reproduction and
have suggested innumerable approaches for future experimentation.

Work on the changing of the chemical structure of a virus by
means of known chemical reactions has been both encouraging and dis-
couraging. It has, in fact, proved possible actually to change the
chemical structure of a virus; but so far no change has been found to
be perpetuated in the virus particles produced as a result of infection
of a host with the altered structures. Thus, the abolishment of the
sulfhydryl groups in tobacco mosaic virus or the introduction into the
structure of this virus of several thousand acetyl, phenylureido, carbo-
benzoxy, benzene sulfonyl, or malonyl groupings yields diverse altered
virus structures. Although these are infectious, the disease which they
produce is the ordinary tobacco mosaic disease, and is accompanied
by the production of particles, not of the respective altered structures,
but of ordinary tobacco mosaic virus. However, encouraging results
were obtained in a study of the specific virus activity of these chemical
derivatives on different hosts. It was found that a property of the
virus, which perhaps can best be described as virulence, can remain
constant for one host but be modified with respect to a different host,
upon formation of a given chemical derivative of the virus. This
result lends encouragement to the belief that, eventually, heritable
structural changes in a virus will be achieved in the chemical labora-
tory by means of known chemical reactions. Contemplation of the
implications that would accompany the actual accomplishment of
this feat tends to stagger the imagination.

Spectacular progress has attended studies on the nature of the
differences in chemical structure that are responsible for the existence
of virus strains. It was indeed fortunate that Nature provided, and
the pathologist recognized and separated, two or more strains of each
of several different viruses. Strains of a virus appear to arise during
the reproduction of a virus by a process which can be regarded as
similar to that of gene mutation. It can be presumed that the strains
of a virus have arisen from some parent strain, one by one, during the
course of the years. Each strain thus probably bears a definite rela-
tionship to the parent strain and to each of the similar strains. Each
strain causes a more or less different disease; because of this fact it

20

VIRUSES

has been possible foi' the pathologist to recognize, separate, and grow
individual virus strains. Some of these have been found of girnt use-
as vaccines for the prevention of certain virus diseases.

In view of the large size and complexity of structure of \inises, ii
may appear that chemists were somewhat optimistic in expecting to
be able to detect differences in the chemical structure of different strains
of a virus. Two similar large mountains may appear identical when
viewed from a distance, but close inspection will reveal differences.
So, too, with the viruses. The over-all structures of different strains
of tobacco mosaic virus were found to be very similar; yet in several
instances it has been possible to demonstrate definite differences in
chemical structure. These consist of differences in the amount of one
or more amino acids, in the presence of an entirely new amino acid,
or in the complete elimination from the virus structure of a given amino
acid. These changes represent deep seated and fundamental altera-
tions in the virus structure, and it seems unlikely that they could have
resulted from alterations of fully formed virus particles. It appears
more likely that these changes occurred as a result of a dixersion of the
synthetic process by means of which a virus reproduces. Since new
strains tend to appear or to become dominant when a virus is grown
in an unnatural host, it is possible that the altered environment of this
host provides a somewhat different supply of amino acids and enzyme
systems, and in the effort to adhere to some basic pattern it becomes
necessary to build into the virus structure amino acids that would not
be used normally. The drive to follow a basic pattern and the aberra-
tions that result bear a certain kinship to the forces of heredity. As a
matter of fact, there is a striking similarity between the properties of
viruses and those that have been ascribed to genes. Both may be re-
garded as large nucleoprotein structures that have the ability to
perpetuate themselves within, and only within, certain specific living
cells. Both can undergo sudden changes, apparently cither spon-
taneously or as a result of external factors and those changes are then
reproduced in subsequent generations. Within limits, the concen-
tration of both in cells can be changed by proper treatment of the host.
Some viruses appear to be concentrated in the cytoplasm of cells and
others in the nuclei.

The similarity between viruses and genes may not be without
significance, for the abode of genes is the cell and no virus has been

21

W. M. STANLEY

proved to arise de novo — they are always first found in cells. Viewed in
this light, the diflferences in chemical structure that have been demon-
strated to exist between different virus strains, and especially the chang-
ing of the chemical structure of a virus in a definite manner by means
of known chemical reactions, take on a new and perhaps startling
significance. There are many virus diseases in which a symbiotic re-
lationship is set up between virus and host. Although the virus
enters into and alters the metabolic activity of the cell, the cell survives
and continues to divide. The virus is carried continuously within the
cells, and in some cases could almost be regarded as a normal com-
ponent of the cells because of the lack of obvious damage. In fact,
one virus has come to be known as the "healthy potato virus" because
it is present in almost all potatoes grown in this country and yet the
potato plants appear healthy. It is as if one had introduced; from
without, a nucleoprotein which was accepted by the cell as a part of
its own germ plasm. The fact that different strains of a virus, which
can cause different manifestations of disease, are characterized by
different nucleoproteins, and especially the fact that the structure of
these nucleoproteins can be changed in the test tube by means of known
chemical reactions, could be interpreted to mean that eventually the
germ plasm of cells may prove to be susceptible to similar chemical
manipulation. The viruses have assuredly provided a link between
molecules and organisms, and there now exists a pathway leading from
simple structures, such as the electron, to massive, highly complex
structures, such as man. This pathway is broad and well established
in some places and narrow and difficult to traverse in others. But as
the latter are broadened and placed on a firm foundation, through
the common effort of chemists and pathologists, it is possible that
information will be acquired which could affect the future destiny of
every living being in the world.

Selected References

Anson, M. L., and Stanley, W. M., J. Gen. Physiol., 24, 679 (1941).
Bawden, F. C, Plant Viruses and Virus Diseases. Chronica Botanica, Wal-

tham, 1943.
Beard, J. W., Bryan, W. R., and Wyckoff, R. W. G., J. Infectious Diseases, 65, 43

(1939).
Bernal, J. D., and Fankuchen, I., J. Gen. Physiol, 25, 111, 147 (1941).

22

VIRUSES

Cohen, S. S., and Stanley, W. M., J. Biol. Chem., 144, 589 (1942).
Doerr, R., and Hallauer, C, Handbuch der Virusjorschung. Springer, Berlin,

1938-39.
Green, R. H., Anderson, T. F., and Smadel, J. E., J. Exptl. Med., 75, 651

(1942).
Hoagland, C. L., Ann. Rev. Biochem., 12, 615 (1943).
Hoagland, C. L., Ward, S. M., Smadel, J. E., and Rivers, T. M., /. Exptl. Med.,

74,69,133(1941).
Hoagland, C. L., Ward, S. M., Smadel, J. E., and Rivers, T. M., J. Exptl. Med.,

76, 163 (1942).
Knight, C. A., J. Biol. Chem., 145, 11 (1942).

Knight, C. A., and Stanley, W. M., J. Biol. Chem., 141, 29 (1941).
Lauffer, M. A., J. Biol. Chem., 143, 99 (1942).
Lauffer, M. A., J. Am. Chem. Soc, 66, 1188 (1944).
Lauffer, M. A., and Ross, A. F., J. Am. Chem. Soc, 62, 3296 (1940).
Lauffer, M. A., and Stanley, W. M., J. Exptl. Med., 80, 531 (1944).
Luria, S. E., and Anderson, T. F., Proc. Natl. Acad. Sci. U. S., 28, 127 (1942).
Luria, S. E., and Delbriick, M., Arch. Biochem., 1, 207 (1942).
MUler, G. L., and Stanley, W. M., J. Biol. Chem., 141, 905 (1941).
Rivers, T. M., Viruses and Virus Diseases. Stanford Univ. Press, Stanford Uni-
versity, 1939.
Rivers, T. M., et al., Virus Diseases. Cornell Univ. Press, Ithaca, 1943.
Stanley, W. M., Physiol. Revs., 19, 524 (1939).
Stanley, W. M., Ann. Rev. Biochem., 9, 545 (1940).
Stanley, W. M., J. Biol. Chem., 135, 437 (1940).
Stanley, W. M., Sci. Monthly, 53, 197 (1941).
Stanley, W. M., J. Exptl. Med., 79, 267 (1944).

Stanley, W. M., and Anderson, T. F., J. Biol. Chem., 139, 325 (1941).
Stanley, W. M., and Knight, C. A., Cold Spring Harbor Symposia Quant. Biol.,

9, 255 (1941).
Taylor, A. R., Sharp, D. G., Beard, D., and Beard, J. W., J. Infectious

Diseases, 72, 31 (1943).
Williams, R. J., Schlenk, F., and Eppright, M. A., J. Am. Chem. Sac, 66,

896 (1944).

23

PHOTOSYNTHESIS

AND THE PRODUCTION OF

ORGANIC MATTER ON EARTH

H. GAFFRON, research associate, professor, departments of bio-
chemistry AND chemistry (fELS FUNd), UNIVERSITY OF CHICAGO

Practical Importance of the Assimilation of Carbon Dioxide

We know, for example, that if we abuse the soil, it will
lose its fertility, that if we massacre the forests, our
children will lack timber and see their uplands eroded,
their valleys swept by floods. Nevertheless, we continue
to abuse the soil and massacre the forests ... in the
simple affairs of nature, where we know quite well what
is likely to happen we immolate the future to the present.
" Those whom the gods would destroy they first make mad."

K

ALDOUS HUXLEY, Time Must Have a Stop, p. 298.

TVEN the less informed layman is aware of the reaction
which makes possible the abundance of living things on
earth- — the conversion in daylight by plants of carbon dioxide and
water, the waste products of plant and animal life, back into food.
This reconversion is called the assimilation of carbon, or photosyn-
thesis. Hardly ever does the layman spend another thought on this
fundamental and quite spectacular achievement of living cells. If,
however, he suddenly realizes that, without this reaction, life as we
know it would perish quickly and completely (except for a few species
of bacteria), he is likely to place false hopes on its artificial reproduction
by the toiling scientist. He probably believes that mankind's in-

25

H. GAFFRON

adequate supplies of food and dwindling stores of energy will turn into
a surplus the moment photosynthesis has been duplicated in the test
tube. Popular articles and radio talks — spreading the lamentable
misconception that scientific research represents nothing but the first
step in technology — foster this notion by stressing the practical im-
portance of artificial food production from sunlight and carbon dioxide.
The error is easily demonstrable. The solar energy flux per acre and
year is a constant. Unless our devices are to be much more efficient
than the green plants and unless we are willing to spoil with ugly ma-
chinery an acreage equivalent to that now covered by beautiful forests
and pastures, artificial photosynthesis is not only Utopian but im-
practical. Even assuming we were to discover some sort of artificial
photochemical reduction of carbon dioxide into a digestible carbo-
hydrate with an over-all efficiency surpassing that of the plants which
use on the average two per cent of the incident radiation, it would not
help us much, since we need, not one product of plant metabolism, but
a thousand (16). Let us mention only the proteins among foodstuffs,
and wood and rubber among industrial raw materials.

True, chemists have learned to make ammonia and nitrates
from atmospheric nitrogen. In that respect man is no longer dependent
on other natural sources. But up to now only the plant converts these
simple nitrogen compounds into proteins,* and so we are forced in a
second way, equally fundamental, to rely on the growth of plants for
our continued existence. In order to stretch our limited supply of
organic substances already formed, such as coal and petroleum, the
production of most compounds which plants can synthesize and which
we need in large quantities should be reserved for natural photosyn-
thesis, regardless of whether the chemist can duplicate the synthesis
in vitro. Oil, for instance, is too versatile a material to be converted
into rubber as long as plants are capable of continuously producing
essential raw materials like alcohol, or better still, rubber of excellent

quality, t

Denying that artificial photosynthesis will be the solution to a
serious and fascinating problem is not equivalent to saying that sun-

* Nonphotosynthetic plants like yeasts are a good source of protein, pro-
vided they are fed with products of photosynthesis.

t Guayule plants grown in California produce rubber far superior in
automobile tires to the synthetic compounds now available.

26

PHOTOSYNTHESIS

light will not be used at all in the technology of the future. In selected
places, such as roof tops in sunny countries, bare rock, and deserts inac-
cessible to irrigation, sunlight may soon be utilized to produce heat or
electric power or even to drive some photochemical process like that
of hydrolysis. These possibilities have exactly as much and as little re-
lation to the problem under discussion as exists between any other
common source of energy and a filled granary. There is every indica-
tion that products of plant photosynthesis will be needed ever more
urgently, not only as food but also as fuel and building material. Re-
cent technological mventions indicate that wood, strengthened with
wood-derived chemicals, will be in even greater demand than now.

One and one-half billion of the two billion humans on earth are
illnourished or permanently hungry, "There has never been enough
food for the health of all people" (18). Since this globe offers only a
certain area of habitable and tillable ground, it is obvious that
populations will have to be adjusted to a certain optimum density
determined by the general standard of living that man is capable of
attaining or willing to endure.

We have ample testimony of the improvident way in which
the ancients exhausted their supply of wood. The mountains of
Persia, Syria, Greece, Dalmatia, Italy, and Tripolitania are now to a
large extent barren and infertile. The goats of the Arabs are said to
have done away with the last traces of vegetation in North Africa,
thus allowing the desert sands to advance to the shores of the Medi-
terranean. The changes in climate brought about by the denuding
and erosion of the sites of the most important early civilizations make
it difficult or impossible to retrieve the lost fertility of the land.

Are we wiser today? The rate of consumption of coal and pe-
troleum in the world today (1.75 X 10^ tons carbon per year) is
roughly one-tenth of the rate of the total carbon assimilation achieved
by the land plants on earth (1.6 X lO^" tons carbon per year) (14).
The coming industrialization of Asia will soon diminish the gap be-
tween the demand for, and the supply of, organic material. The
established reserves of oil in the United States, according to official
figures, are about twenty billion barrels. They are being used up at
a rate of one and one-half billion barrels per year. The industry
insists upon a more optimistic outlook, based on the (ever-declining)
rate of new discoveries. True or not, considering also the certainly

27

H. GAFFRON

limited supply of coal, it is quite clear that we are spending mainly
as fuel irreplaceable organic material.

Fig. 1 . — Virgin forests in 1 850.

Can these losses of stored products of photosynthesis be offset
by the assimilation of carbon going on today? Within the cultivated
areas of industrialized countries this is not the case, and foresters agree
that virgin forests are more or less stationary. New growth balances
natural decay. Only well-planned agriculture and expert forestry
may perhaps furnish all that we need in the future provided the
increasing demand for fuel and energy is eventually met by the general
development of atomic power. At present, the products of agriculture
are consumed within a few years after harvesting. There is no increase
of our reserve of organic carbon due to this source. On the contrary,
the improvident exploitation of the soil in many places leads to di-
minishing returns. In the United States, fifty million acres now under
cultivation are so badly eroded as to invite abandonment. These sad
circumstances have received widespread attention, and effective
measures are being taken to check further losses and to regain the lost
fertility of the soil (13). Not so well known are the conditions regard-
ing the forests in this country (1). The following figures speak for
themselves:

28

PHOTOSYNTHESIS

Billion board feet

Original stand of forests 5,200 on 850 million acres

Present (1927) stand of forests 2,500 on 550 million acres

Rate of annual consumption 35

We are cutting down our forests three times as fast as they
grow. Certain species of trees are in danger of extinction. War
conditions have aggravated the situation. Figures referring to the
years just before the war for the states of Washington, Oregon, Mon-
tana, and Idaho, which now supply half of the timber cut in the
United States, are as follows (3):

Billion board feet

Annual cut and losses in the Northwest 12.5

Current annual growth 4.2

Potential annual growth after universal adoption of

best forest practices 23

Fig. 2. — Virgin forests in 1945.

The last figure shows that we should not abandon all hope.
There is a well-tested way to prevent the exhaustion of our wood
reserves. But it is not enough for the government to administer the
national forests i)rovidently. Industrialists and private owners hold

29

H. GAFFRON

not only the larger area of commercial forest land but also the more
valuable timber from the standpoint of accessibility and of quality.
"The main objective of this group is to market their lumber as speedily
as possible" (3). Management according to the principles of far-
sighted modern forestry is encountered only on an estimated two
per cent of the total area. The practices of commercial competition
are such that in general only the State can afford to look ahead two
or three generations.

Meanwhile, the waste continues. According to N. G. Brown (1)
less than half of the total amount of wood cut in this country is ulti-
mately utilized. Hence it is imperative that the efforts of the Depart-
ments of Agriculture and of the Interior to remedy the situation
should have the conscious support of every citizen. Unless we plan
for a permanent forestry everywhere, not only our supplies of oil but
also those of wood will be gone in about seventy years and the country
dependent upon foreign sources. No wonder that scientists begin to
turn toward the oceans as a source of products of photosynthesis; it
is estimated that the amount of carbon assimilated by marine algae
surpasses by four or five times the yield of the land plants (14). But
when can we expect to replace wood by plankton? The journals of
chemical industry often display an advertisement showing a magnifi-
cent mountain forest with a caption from which we quote: "Ever
see a forest through a chemist's glasses? Do you see . . . plastics . . .
plywood . . . laminated beams . . . silk . . . smokeless powder . . . rolls
of newsprint . . . movie film . . . All these and many more items of
beauty, strength, and utility the chemist makes from wood, holding
out brighter hopes for a better future." If man continues to consume
the products of photosynthesis in the way he does at the present time
this kind of better future may not last very long.*

The understanding of photosynthesis becomes essential if we are
to solve part of these rapidly approaching difficulties, not in order to
reproduce the process technically on a big scale but rather in order to

* Two publications have recently appeared on the subject of our dwindling
supply of wood: "Forestry and the public welfare," in Proc. Am. Phil. Soc, 89, 399
(1945); and "What's happening to the timber," by R. A. H. Thompson, in
Harper'' s Magazine, issue of August, 1945, page 125. The instructive maps (Figs.
1 cmd 2) from the latter article are reproduced through the courtesy of Harper's
Magazine.

30

PHOTOSYNTHESIS

increase its natural efficiency. As everybody knows, different plants
grow in different locations. The efficiency of the chloroplasts or of the
photochemical mechanism proper may be equal in all plants, yet
some plants may not have attained the optimum in utilizing their
own photosynthetic products. Here the knowledge of internal factors
influencing photosynthesis will show which plants to breed. Investi-
gations on the photosynthetic behavior responsible for the over-all
efficiency of plants in producing the coundess substances we need will
increase in practical importance. Much has been done and will be
done empirically by the gardener, agriculturist, and forester, but the
shortest way to success lies in systematic investigations of the correlation
between growth or fruition of plants and the rate of photosynthesis.
Such work is in progress, for instance, at the California Institute of
Technology (22). The intimate connection with studies concerning
such factors as soil, climate, and inheritance are obvious.

In contrast, it appears doubtful at this moment whether the
analysis of the nature of the photosynthetic process can produce imme-
diate practical results surpassing those obtainable by the kind of studies
mentioned above. True, pure reseaixh remains the only source of
sudden technical advances. At the present time, however, the ques-
tion as to the nature of photosynthesis is a problem of science and not
of so-called "applied science," that is, technology. The reward in
joining the few who insist upon spending their time in tackling this
problem will be only the pleasure of knowing a little more and seeing
a little farther than those who worked on the same problem twenty
years ago.

Partial Reactions in Photosynthesis

" The enormous amount of research upon the process of
photosynthesis during the past half century has thrown
little or no additional light upon the subject. The
problem is evidently too complex for any specialists in any
one field of scierue to solve. ..."

E. C. MILLER, Plant Physiology, 2nd ed., 1938

During the past decade a new laboratory technique — that of
enzyme chemistry — has been developed and is at present taught to
every student in cell physiology. The isolation of an enzyme is an
advance that requires no mental effort to appreciate. Equally obvious
are the advantages of using traceable isotopes for elucidating the course

31

H. GAFFRON

of obscure metabolic reactions. Among younger biochemists the
attitude toward a field like photosynthesis is one of waiting until dis-
agreeable obstacles have been cleared away by hand, so to speak, so
that they may roll in with modern methods. During the past twenty
years the moment when this will happen has quietly drawn nearer.
How near we shall try to demonstrate in the rest of this essay.

It is mostly forgotten that, after Buchner had demonstrated
fermentation outside the living cell, some thirty years passed before
enzymes were isolated and the mechanism of organic catalysis could
be said to be clearly understood. Progress in the field of cellular break-
down reactions, slow in the beginning of this century, nevertheless
encouraged a handful of students to devote their lives to the investiga-
tion of respiration and fermentation. In contrast, the return for the
toil directed toward solving the problem of photosynthesis was so small
that even the leading biochemists, after establishing a few important
new facts such as the existence of a photochemical reaction distinguish-
able from an enzymic one, saw no point in pursuing the analysis any
further. In 1918, Willstatter wrote that it was evidently too early
to try to elucidate the mechanism of carbon dioxide assimilation in
living cells. We now know that he was right. At that time photo-
synthesis appeared as an absolutely unique process showing no con-
nection or analogy with other metabolic processes in the cell. It was
looked upon as a direct photochemical decomposition of carbon dioxide
and mysteriously connected with the phenomenon of life on earth.

The absolute ignorance of the kind of reactions photosynthesis
might involve and, consequently, the lack of a theoretical framework
accurate enough to direct a reasonable approach barred further
progress. And as to purely empirical experimental attempts, they
all ended with the destruction of the intact living cell. Photosynthesis
stops the moment the cell is hurt. While this is still true, of course,
the great diflference is that today we know more or less why. Hence
there is hope we shall be able to overcome this difficulty.

Despite Miller's verdict quoted above, there has been very
decisive progress between 1918 and 1938, not merely by the addition
of several important observations to a hundred earlier ones but, mainly,
by the change in our conception of what constitutes the problem of
photosynthesis. This change was due, on the one hand, to the under-
standing of the nature of other metabolic processes such as respiration

32

PHOTOSYNTFiESIS

and fermentation and, on the other, to a clearer knowledge of the
essence of photochemical reactions in complex molecules. Twenty-
five years ago there was an overabvmdance of uncorrelated general
observations and only a few \ery simple pictures of the mechanism
of photosynthesis. The latter were hardly supported by any straight-
forward experiments or permissible analogies. I need mention only
the concepts of a chlorophyll carbonic acid complex, of the release
of oxygen from this complex, and of the formation and polymerization
of formaldehyde, which are still faithfully reproduced in most textbooks
of botany. About 1930, van Niel (19) pointed out that photosynthesis
should be considered as a coupled oxidoreduction comparable to
other reactions of this kind. This approach to the problem proved
very fruitful and was soon generally adopted. Today, in 1945, we
can divide the process of photosynthesis into several partial reactions,
each with its particular problems, none of which seems to present
insurmountable conceptual difficulties. By drawing upon analogies
with other metabolic processes and from the results of direct experi-
ments it is possible to build up a theoretical picture which, though
incomplete in many places, satisfies the essential requirement that
the main observations can be correlated. The mystery of photo-
synthesis is mainly gone, and this in a rather fundamental sense. We
are now quite certain that photosynthesis promoted by chlorophyll in
visible light has nothing to do with the origin of life. Instead, it must
be regarded as a rather late achievement of the living cell. It is a
unique combination of a few reactions found only in the green plant,
with important devices characteristic of any kind of metabolism in
living cells.

With the acquisition of pigments, the early living cells became
able to accelerate the reaction between hydrogen donors and acceptors
by absorbing radiant energy. The biological currency, H and OH,
the constituents of water, became available in larger quantities because
of the interaction between the irradiated dye and water. Despite the
photochemical reduction of carbon dioxide no true gain in free energy
of the complete system was yet possible, since the resulting "hydroxyl-
ated" counterparts {cf. reference 20) could only re-form water with
some valuable hydrogen donor. Any increase in the amount of
organic matter depended on the presence of inorganic hydrogen
donors such as free hydrogen, hydrogen sulfide, sulfur, ferrous iron,

33

H. GAFFRON

etc., and hence could not proceed very rapidly. The picture changed
radically, however, the moment some cells happened to combine the
photochemically accelerated utilization of water as hydrogen donor
with a reaction allowing for the elimination of oxidation products by
liberating free oxygen.

As long as there was an abundant supply of carbon dioxide,
of water, and of light energy, the accumulation of organic matter could
continue unchecked. Oxygen appeared in quantities as free gas in
the atmosphere. This was followed by the enormous multiplication
of organisms capable of using the oxidation of organic compounds for
synthetic reactions.* Carbon dioxide released by respiration was
again utilized in the photosynthetic reaction. Some million years
later, the cycle of carbon arrived at a steady state. The concentration
of carbon dioxide in the air is now 0.03% and barely sufficient to
support maximum photosynthesis in full sunlight. Most plants do
better when carbon dioxide is added artificially.

This picture of photosynthesis as a process that developed
gradually from less complicated reactions is supported by the following
observations. In the plant, the catalytic systems bringing about the
assimilation of carbon dioxide can be distinguished by the differentiat-
ing effects of metabolic poisons. The whole process is specific for
carbon dioxide, just as other metabolic reactions are specific for their
substrates. The various catalysts seem to consist of proteins combined
with special reaction groups. Chlorophyll itself is bound to protein
in the cell. The reduction of carbon dioxide is not an exclusive
privilege of the green plants. Besides normal photosynthesis we dis-
tinguish at present the following types of carbon dioxide reduction:

(a) In the dark (10,23) coupled with

(7) bacterial fermentations (methane bacteria, propionic
acid bacteria),

(2) bacterial and plant oxidations (sulfur bacteria,
"Knallgas" bacteria, unicellular algae),

(3) metabolic reactions in animal tissues.

(b) In the light (6,7,19), with the simultaneous consumption

* Van Niel has demonstrated a gradual adaptation to aerobic conditions
with some strains of purple bacteria, which at first grew only anaerobically in the
light (20).

34

PHOTOSYNTHESIS

of inorganic or organic hydrogen donors (purple bacteria,
unicellular algae).

Unique to the green plants is the coupling of photochemical re-
duction with an evolution of free oxygen. Such an evolution of oxygen
can be obtained by illuminating isolated chloroplasts in the presence
of oxidizing substances like ferric salts or jf?-benzoquinone (5,9,21).
The existence of several more or less autonomous enzyme systems
working together can best be demonstrated in unicellular algae be-
longing to the Scenedesmaceae. In these algae it is possible to inter-
change at will all three known types of carbon dioxide reduction; a
chemoreduction promoted in the dark by the burning of hydrogen
with oxygen to water; a photoreduction in which two hydrogen
molecules disappear together with a molecule of carbon dioxide; and
a complete photosynthesis with the liberation of an equivalent amount
of oxygen (7).

A rough summary of the way in which the problems of carbon
dioxide reduction in an alga like Scenedesmus can be subdivided accord-
ing to present knowledge is given in the scheme below. More complete
and complicated schemes can be found in a recent review (7) and in
Rabinowitch's new book (14).

COj

fixation of
carbon dioxide

/

carboxyl group

carbohydrate

i
(CHjO),

Light absorbed by plant pigments
(chlorophylls, xanthophylls, phycocyanins)

V . . t .

photochemical reaction dismutation to

(chlorophylls, water, enzymes) oxygen and water

^ . ^.- ^

intermediate intermediate

hydrogen a hydrogen

donors > .2 < acceptors

\

intermediate
hydrogen
donors
\

u

u

\

hydrogen

acceptors

Oj

oxidoreduction

(different from

respiration)

intermediate
hydrogen
acceptors

/ ^

hydrogen
donors

H2

reduction to water

i
H,0

35

H. GAFFRON

In addition to what has been said in the preceding paragraphs
about the probable evolution of the photosynthetic system and about
the three main types of carbon dioxide reduction, the scheme tries to
coordinate the following fairly well-established facts (13):

(7) Chlorophyll is necessary even in those cases in which the
light is effectively absorbed by other plant pigments such as xantho-
phyll and phycocyanin.

(2) The oxygen liberated does not originate from carbon
dioxide but from molecules of water entering somewhere as ultimate
hydrogen donors into the reaction.

(J) Carbon dioxide is fixed initially by way of a reaction which
is reversible and nonphotochemical, probably in the form of a car-
boxy 1 group.

{4) There are several intermediate steps between the photo-
chemical reaction proper and the appearance of free oxygen. The
proof for this is found in experiments with specific inhibitors and in
the fact that the reversible switching from oxygen evolution to an
equivalent consumption of hydrogen ("photoreduction") does not
change the quantum yield.

(5) The chemoreduction in the dark has many traits in com-
mon with photoreduction. Yet we should expect differences between
some of the intermediates produced and utilized in the dark and those
made under the impact of a light quantum with an energy content of
40,000 calories.

Since neither the reactions of chlorophyll or of any other catalyst
involved nor the nature of any one of the latter has been clearly estab-
lished, the scheme will convince the reader that there is a host of
problems to be solved. The interested investigator is likely to look
upon that partial problem as the most exciting and, hence, important
one to whose solution he believes he can contribute something sig-
nificant. Since each of the partial reactions of which the whole of
photosynthesis consists is indispensable, truly none can be more im-
portant than the other. However, the utilization of radiant energy
and the evolution of free oxygen sets photosynthesis apart from other
better analyzed metabolic reactions. It is here, therefore, where we
must expect the greatest deviations from the common types of cellular
catalyses and where we may find the unprcdicted.

Though the picture given in the scheme could be seen to

PHOTOSYNTHESIS

emerge gradually, since about 1930, from the discussions accompanying
the newly gained experimental data (van Nicl, Stoll, Emerson, Wohl,
Franck, and others), the interrelation between the different types of
carbon dioxide assimilation has struck some investigators quite sud-
denly. In their enthusiasm over this fact they are likely to overlook
the peculiar problems which {photosynthesis offers in contrast to, and
distinction from, the other metabolic processes. To them, now, all
is very simple. The light absorbed by chlorophyll is used to mobilize
hydrogen. Water is decomposed into H and OH. And once hy-
drogen is available the reduction of carbon dioxide proceeds as a dark
reaction exactly as in the cases mentioned above. This has led to
some strange expressions and statements, such as "photosynthesis in
the dark" or "light per se is not essential for photosynthesis."

Lately, the analogy between thermal and photochemical re-
actions has been pushed still further, and attempts have been made to
bring the so-called energy-rich phosphate bond into the picture. Be-
fore we analyze these attempts we must say a few words about inter-
mediates.

In Search of Intermediates

In taking apart the mechanism of photosynthesis, we shall in
all probability find two types of intermediates: on the one hand, the
enzymes necessary for the transfer of hydrogen and the removal of
oxygen; on the other, the final acceptors, the precursors of carbo-
hydrates and of oxygen. The former we may compare to the pyridine
nucleotides, flavoproteins, and cytochromes, the latter, to the deg-
radation products of glucose, with no clear counterpart to what has
been called the "photoperoxide" or "peroxide" or "moloxide" or
"hydroxylated compounds" (van Niel) in photosynthesis, that is, the
hypothetical substance which decomposes with the evolution of oxygen
(we are pretty certain that it is not hydrogen peroxide and perhaps it
is no peroxide at all) .

Most of the intermediate catalytic steps in respiration and
fermentation have been shown to be reversible — even the decarboxyla-
tion of pyruvate, if it proceeds with the formation of acetyl phosphate
(11). The only requirement is the coupling with another reaction,
furnishing the energy to reverse the particular step. Without such a
coupling the breakdown processes continue unchecked until the specific

37

H. G.\FFRON

substraies are e>diausted. If such an easy reversibilirs^ existed e\er\-
where in the reaction chain leading firom carbon dioxide to carbo-
hNxlrate, an accumulation of photos^•nthetic products would be un-
likely. In fact, it is quite essential for the efficiency of the s^Tithetic
jMXXXSS that the carboh\tlrate finally formed is not broken dowTi
again in the dark by the same catalytic s^^stems which helped to btiild
it up. One of the difficulties we shall encounter in tr\-ing to reconstruct
the phota5>-nthetic mechanism tn vitro with enz\-mes isolated fix)m the
plant cell will probably be the reversibLlir>- of partial reactions, tinless
we succeed in remoxing and thereby stabilizing the intermediate just
fcHined. At least, the way "doxsTi" is the normal course of events as
demonstrated in respiration and fermentation. Now, save for attack
by the respiratory or glycoh-tic s\"5tems, the carbohydrate sxTithesized
in the light is stable (a ver\- slow reversion is postulated for theoretical
reasoDS not discussed here). The solution of the problem, therefore,
may be hidden at either or both ends of the photosynthetic s%"stem.
Tlie primar\- tinstable carbohNxirate may pohTnerize with the release
of some free energ\-. One might even think of a reduction up to the
alcc^ol ^OT h\-drocarbon) level and an oxidation by another path back
to an aldeh\-de (or alcohol) . At the other end of the line, the ox\-gen-
liberating s\"3tem should have prop>erties making it rather ineffective
as an oxidase. At any rate, all attempts to cause an accimiulation of
intomediates by mistreating the irradiated cell .with narcotics or
specific poisons have failed. Hydrogen donors capable of reducing
carbon dioxide afterwards in the dark do not survive a period of
illxnnination. There is no easily detectable formation of partiy re-
duced substances. Xe\"erthelesS; there must be at least three inter-
mediate steps in the process of reduction, since four hydrogen atoms
have to be transferred firom water to carbon dioxide. This particular
problem ^ill probably be solved \sith the aid of carbon isotopes in
continuation oi tbe \*T>rk oi Ruben, Kampn^ and Hassid {cf. 14).

Several iod^jendent investigations have shov*"n that nine or ten
l%hi qu2^:i l:z r.rcessary for this process. The TnaYi'mum ntmiber
of initial 5 erre, could be ten. For reasons of stoichiometiy it

is soKible uj assun^ that there are only eight photochemical reactions,
a pair for each release and recovers- of a hydrogen atom by the mole-
cule transforming electronic into chemical energy (4) . This molecule
may be chlorophvll itself. The difference between the theoretical

38

PHOTOSYNTHESIS

value of eight and the obser\-ed one of ten we attribute to losses due to
inactive chlorophyll. The simplest explanation, then, for the fact
that photosynthesis proceeds either all the way from carbon dioxide
to carbohydrate \%-ith a permanent gain of 120 kcal. per mole, or not
at all, is that the photochemically produced hydrogen donors and
acceptors disappear by back reactions whenever the process becomes
artificially inhibited. Despite the most drastic effects of certain poisons
on the rate or the t\-pe of the reaction, there is no permanent shift of
the assimilatory quotient.* Hence, all but the final reaction products,
oxygen and starch or sucrose, must react back to form water.

It is very unlikely that this disappearance of intermediates
should proceed by an exact reversion of the photochemical process.
A chemiluminescence with a 100% yield can certainly not occur. In
other words, we may expect a special kind of oxidoreduction \N-ithin
the assimilatory system (7) (compare the last part of this article, pages
43 et seq.).

One way of avoiding immediate back reactions hes in the use of
radiant instead of thermal energ^^ A Hght quantum heats up, as it
were, a single molectile Ln an otherwise cool enwonment. Once the
structure of the absorbing molecule allows for a sp>ontaneous con-
version of electronic excitation energy into chemical energy (that is,
a change of structure to a configuration with a higher free energ\- con-
tent), the new tautomer has a good chance to sur\-ive until some
catal^'tic action makes use of its potential energ\\ The new tautomer
will live longer the higher the wall of activation energ\^ between the
original and the photochemically changed structure. In other words,
its lifetime depends on how much of the available light energ\' (in the
case of chlorophyll, this is always ca. 40,000 calories and is independent
of the size of the lieht quanttma absorbed) is expended in forming the
new structure. Actually, the ver\' first photochemical product aj>-
pears to be short-lived. It probably contains still much of the original
40,000 calories. J. Franck assimies, further, that a reduced inter-
mediate hydrogen donor comparable to, let us say, the reduced p>Ti-
dine nucleotides is not formed. Rather, the carbox\i group becomes
reduced direcdy as a consequence of the photochemical reaction taking
place in the chlorophyll-protein complex to which the essential car-

* Assimilatory quotient = (oxxgen liberated") /(carbon dioxide consumed)
or (hydrogen absorbed) /(carbon dio.xide consumed).

39

H. GAFFRON

boxyl group seems to be attached, Franck considers the influence
which the presence of carbon dioxide exerts on the intensity of chloro-
phyll fluorescence in vivo as the most cogent evidence supporting this
view. The scheme shown on page 35 does not do justice to these
important considerations.

In case, however, the fluorescence experiments could be ex-
plained in another way, there would be no objection against assuming
the existence of an intermediate hydrogen donor. Perhaps catalyst B
of Franck and Herzfeld might play this part (4). Its concentration
should be about two thousand times smaller than that of chlorophyll
(the evidence can be found in the discussion about the photosynthetic
unit).

Summing up, we may state that the heart of the photosynthesis
problem is the effective utilization of energy which has to be accepted
in a few big lumps. Franck and Herzfeld (4) calculate an immediate
and permanent gain of 21,000 calories for each quantum absorbed.
Does comparison with chemosynthesis point to an orthodox solution
of this problem?

Possible Role of Energy- Rich Phosphale

Recent research on the utilization of metabolic energy in living
cells had led to the discovery that this energy is handled in parcels
not surpassing 12,000 calories packed away in so-called energy-rich
phosphate bonds. Such phosphate bonds occur, for instance, in
substances like Lipmann's acetyl phosphate (11) which, for the purposes
of our discussion, may be regarded as stable for an indefinite period.

Recently, Emerson, Stauff"er, and Umbreit (2) suggested "that
for each quantum of light absorbed one 'energy-rich' phosphate bond
(of ca. 10,000 cal./mole) is formed." There is no experiment or
observation new or old which requires such an assumption. The only
merit of this proposal, therefore, would lie in the complete analogy of
photosynthesis with synthesis by thermal reactions. Part of the energy
contained in the unstable first product formed in the photochemical
reaction, the tautomer mentioned above (4), would be stabilized by
conversion into a phosphate bond. The loss involved would be two-
thirds of the chlorophyll excitation energy. There is general agree-
ment among those who have considered carefully the theoretical

40

PHOTOSYNTHESIS

energy requirements of photosynthesis, that a minimum of 160,000
calories is needed per mole of carbon dioxide. Ten to twelve light
quanta (let alone eight) transformed into phosphate bonds at 12,000
calories each would mean that the latter would have to be used with
120% efficiency. Actually, the efficiency with which the energy-rich
phosphate bond can be used for synthetic reactions appears to be around
60%. If we postulate two phosphate bonds per quantum (that is,
twenty bonds altogether), tliere would be enough energy. However,
a complicated mechanism must be provided to divide the energy of
the excited molecule between the two phosphate bonds. Finally, the
resulting phosphorylated compounds should be stable enough to
survive the end of an illumination period and cause the reduction
of carbon dioxide for some time afterward in the dark. As said
above, nothing of that kind has ever been observed, though many
an investigator has looked for it. Hence, it is simpler to assume that
Nature makes use of the particular advantages inherent in photo-
chemical reactions and produces intermediate hydrogen donors which
are capable of a one-step gain in free energy larger than those possible
by way of energy-rich phosphate bonds.

A question quite different from that discussed is whether carbon
dioxide becomes reduced to carbohydrate in the form of a phos-
phorylated compound. It is known that plant constituents combining
with carbon dioxide in the dark are half saturated at a carbon dioxide
partial pressure of less than 0.1 mm. mercury. The nature of the
very first fixation of carbon dioxide preceding the reaction with excited
chlorophyll is still under investigation. Suggestions like "reversal of
a decarboxylation" are no more helpful than the dictum that the
liberation of oxygen is "the reversal of respiration." What is needed
is a working model fulfilling the energy requirements. A carboxyla-
tion reaction as effective as that taking place in the first step of photo-
synthesis needs the coupling with a reaction releasing some free energy
{ca. 10,000 cal.). Ruben (15) suggests that the carboxylation occurs
with the aid of an energy-rich phosphate bond present in the molecule
which takes up carbon dioxide. Lipmann (11) recently has shown
that, in the presence of certain enzyme preparations, acetyl phosphate,
carbon dioxide, and molecular hydrogen can condense to pyruvate.
By proposing that the pyruvate is removed quickly through further
reduction and a new energy-rich phosphate furnished by an oxida-

41

H. GAFFRON

tive side reaction, he has provided the first plausible model for a con-
tinuous chenioreduction of carbon dioxide.

As an experimental approach to tie up the phosphorylation
reaction and photosynthesis, it is obviously insufficient to demonstrate
that phosphorylations take part in the metabolism of plant cells and
that the photosynthetic production of carbohydrates changes the dark
equilibrium of various phosphate esters. It w^ould be strange if it
were not so. In a respiring or fermenting cell, the model for the chemi-
cal reduction of carbon dioxide as presented by Lipmann should work
with any available metabolic hydrogen, particularly with free hydrogen
activated by hydrogenase. Since, under the latter circumstances,
the over-all energy requirements are very small, there is no reason
why fermenting algae should not exhibit a clear-cut reduction of car-
bon dioxide in the dark. This is not the case, at least not at a
noticeable rate. Apparently some extra energy for activation is
necessary. If so, we have still to explain why, as found by Rieke in
unpublished experiments, the number of quanta for the reduction of
one carbon dioxide molecule remains ten when the algae begin to
consume hydrogen instead of liberating oxygen. Energetically, one
or two quanta would now suffice. The simplest explanation would
involve the assumption of the existence of a rigid coupling of the re-
duction of carbon dioxide with the utilization of water as hydrogen
donor. This would mean the exclusion of energetically "cheaper"
hydrogen from a direct participation in the photochemical reaction
during anaerobic photoreductions in algae and, by analogy, also in
the photosynthetic purple bacteria. We cannot present in detail the
points for and against this explanation. It is also not necessary, for
in van Niel's comprehensive reviews (20) and Rabinowitch's book
(14) the reader will find much stimulating discussion.

Interrelation of Carbon Assimilation with Oxidation
Reactions in Plants and Purple Bacteria

We have seen above that there is a possibility that substrates,
intermediates, and final products in photosynthesis may undergo re-
duction and polymerization as phosphorylated compounds. Experi-
ments supporting Ruben's or Lipmann's hypotheses (11,15) are still
missing, and Ochoa (12) has just reported the formation of oxalo-

42

PHOTOSYNTHESIS

succinic acid from a-ketoglutaric acid and carbon dioxide. This
type of fixation reaction does not depend on energy-rich acyl phos-
phate. Our knowledge of phosphorylation and the "energy-rich"
phosphate bond is based entirely on the carbohydrate metabolism in-
volved in respiration and fermentation. We know that both types of
catabolic reactions make use of the same phosphorylated compounds.
If phosphorylated substances participate in photosynthesis, are these
intermediates related to, or identical with, those occurring in general
cell metabolism? Some special reaction must provide for the energy-
rich phosphate bond assumed to promote the initial fixation of carbon
dioxide in the dark. Is this special reaction part of the normal re-
spiratory system or something different? In this regard Ruben (15)
mentioned as a possible source of energy the dismutation to oxygen or
reduction to water of the intermediate hydrogen acceptors (hydrox-
ylated substances). The reduction of carbon dioxide by a purely
thermal oxidoreduction in green algae occurs under circumstances
which exclude normal respiration (7). Here the energy-rich phosphate
bond would present a welcome means of explaining not only the fixation
or carboxylation but also the coupling between the oxidation of
hydrogen and the reduction of carbon dioxide. Experimentally,
one should try therefore to establish the existence of an independent
phosphorylation cycle serving exclusively the oxyhydrogen or "Knall-
gas" reaction.

While the oxyhydrogen reaction may be considered to be a
sort of respiration and consequently invites comparison with the
normal cell respiration in air, conditions become more unfamiliar when
we turn to photoreduction. We mentioned earlier in this article why
special back reactions in photosynthesis must be assumed. The
question arises whether such (hypothetical) back reactions contribute
to the formation of energy-rich phosphate bonds. The autonomy of
the mechanism of photochemical reduction becomes quite apparent
in the following observation (8) . A concentration of 0.001 M phthiocol
[also of methylnaphthoquinone (vitamin K) or of o-phenanthroline]
inhibits strongly the respiration of algae like Scenedesmus, prevents
completely normal aerobic photosynthesis, hinders the adaptation to
and the reversion from photoreduction under anaerobic conditions,
and severs the coupling between the oxyhydrogen reaction and the
reduction of carbon dioxide in the dark. But if the inhibitor is added

43

H. GAFFRON

to the plants after their adaptation to photoreduction it does not
interfere with a reduction of carbon dioxide. The photochemical
reaction proceeds with a normal quotient of two. Two volumes of
hydrogen are absorbed together with one volume of carbon dioxide.
The only change of importance concerns the quantum yield. It is
exactly one-half of that found with the unpoisoned algae. Figure 3
shows how, with increasing concentrations of phthiocol at 560 lux,

. 30

25

20 -

1 1

-1 1 r

\

s.'"^"

:^::

^ 560 LUX

A

400 LUX "

1 1

1 1 1

c

LlI

2

o 15

:d
o

LU

a:

g 10
O

X
Q.

o 5

Ixl

<

0 2 4 6 8 10 12

CONCENTRATION OF PHTHIOCOL X 10"^ M
Fig. 3. — The inhibition and stabilization of photo-
reduction in Scenedesmus by increasing concentrations of
phthiocol: • — •, rates at 560 lux; A — A, same a day
later; X — X, rates at 400 lux; O — O, same a day later.

the rate of the reaction falls from 26 to 13 mm. per 10 min. (or at 400
lux from 18 to 9 mm. per 10 min.) and then stays constant. What this
means is quite obscure. A possible solution of the riddle may be
found in the following considerations. The probable occurrence was
mentioned above of back reactions between the reduced and oxidized
products whenever photosynthesis is artificially inhibited. Now we
postulate that, in the poisoned algae, all intermediates react back, and
thereby activate a reduction of carbon dioxide with hydrogen in a
manner similar to that brought about by the "Knallgas" reaction in
the dark. Apart from any specific explanation we can state that, if

44

PHOTOSYNTHESIS

phosphorylated compounds participate at all in the reduction of car-
bon dioxide, they must be either drawn from a reservoir filled during
preceding anaerobic periods or produced by a cycle belonging to the
assimilatory system. No indication in favor of the first alternative
has been obserxed. Special experiments to test it, however, have yet
to be performed.

Another approach to the question of whether phosphorylations
occur within the assimilatory mechanism would be by way of investigat-
ing further the metabolism of the photosynthetic purple bacteria
(6,19,20). Studies on purple bacteria have proved extremely useful
in the past for elucidating the similarities between photoreductions
and the photosynthesis of green plants {cj. references 19 and 20). We
now arrive at the point at which it would be of interest to analyze
in detail the differences in their mechanisms.

An important difference between the green plants and the
purple bacteria from the point of view of the possible role of phos-
phorylated compounds is the fact that the metabolism of the plant
centers around carbohydrates, whereas most of the purple bacteria
decline to utilize them in any way either in light or dark. They do not
respire and they do not ferment glucose and consequently cannot grow
in glucose media. The metabolism of purple bacteria revoh'es, if we
consider only organic substrates, mainly around aliphatic acids and,
in a few cases, simple alcohols. Acetate, propionate, butyrate, croto-
nate, malate, etc., are favorite substrates. The elementary analysis
of an entire green plant yields, in general, data indicating the pre-
dominance of compounds of carbohydrate nature. The available
elementary analyses of some purple bacteria have yielded figures
indicating the presence of more hydrogen and of less oxygen than is
found in carbohydrates and a composition very close to that of an
extracted substance having the formula (C4H602)„. The latter
proved to be a polymer of crotonic acid. The green plants store
carbohydrates in more or less pure form whenever photosynthesis
has lasted for a while in strong light, for respiration proceeds at a much
slower pace. Some purple bacteria {Athiorhodaceae) do not seem to
accumulate photoproducts in excess of what can be used for immediate
•synthesis of integrated cell material, that is to say, for growth of more
bacteria. Others {Thiorhodaceae) accumulate unknown photosynthetic
l^roducts that break down anaerobically in the dark by a sort of back

45

H. GAFFRON

reaction which yields again hydrogen donors capable of being utilized
in the light (20).

A second difference between plants and purple bacteria lies
in the relation of cell multiplication to the assimilation of carbon.
Green plants can, as a rule, grow normally in air on a heterotrophic
diet without photosynthesis. Spoehr (17) succeeded in obtaining
growth of hereditary chlorophyll-free albino corn plants by feeding
them with only one organic compound, sucrose. On the other hand,
we have not observed true growth of algae, during a photochemical or
thermal reduction of carbon dioxide, under anaerobic conditions,
that is, in the absence of respiration.

By contrast, some strains of purple bacteria multiply exclusively
under anaerobic conditions and only during periods of illumination,
and never in the dark. They are not capable of linking a synthetic
process to either oxidative reactions (which do occur in the presence
of oxygen) or fermentations (which seem not to occur at all). They
depend for synthesis and growth upon photoreduction with special
hydrogen donors like fatty acids, hydrogen sulfide, or molecular
hydrogen. A few species among the nonsulfur purple bacteria
{Athiorhodaceae) appear to grow in ways similar to that of the green
plants in that they do not depend upon the photochemical reaction
alone. As van Niel has shown, they can also grow aerobically in the
dark by oxidizing the same substrates which they use as hydrogen
donors in the light. But the fact that no carbohydrates are attacked
points to a deviation from the metabolism of the green plants.

The third interesting difference between plants and purple
bacteria concerns a direct interrelation between the respiratory and
the assimilatory systems. In the ordinary green plant, respiration
and photosynthesis can go on simultaneously. Metabolic measure-
ments appear most consistent if we assume that both reactions run
independently of one another and that any conspicuous increase of
the rate of respiration in the light is caused in an indirect way. Photo-
synthesis appears to provide only reserve material, while the synthetic
reactions leading to cell multiplication are coupled exclusively with
respiration (and perhaps with fermentation). In those purple bacteria
capable of growing at the expense of oxidation reactions, the utiliza-
tion of oxygen must compete with the utilization of carbon dioxide
plus light for the same hydrogen donors. The reactions do not occur

46

PHOTOSYNTHESIS

independently; they supplement or exclude each other. In a very
interesting quantitative experiment, van Niel (20) found that the
bacteria simply cease to take up oxygen when exposed to a sufficiently
intense radiation. Here w^e have to search for a direct interrelation,
an intermediate metabolic link.

We may sum up as follows. Whether phosphorylated com-
pounds participate in photosynthesis must be considered in the light
of two sets of observations. First, the assimilatory mechanism in
green plants and in purple bacteria must be "self-supporting" as far
as phosphorylations are concerned, since it functions under conditions
in which neither respiration nor fermentation of carbohydrates seem
capable of providing enough ready-made phosphorylated compounds.
Second, in purple bacteria a carbohydrate appears perhaps as the
primary product of the photochemical reaction, but instead of being
stored it is converted into cell material of different elementary com-
position. These organisms are incapable of oxidizing or fermenting
ordinary plant carbohydrates.

From the evolutionary point of view it is interesting that the
"respiration" of the purple bacteria corresponds to the oxyhydrogen
reaction and related oxidations in anaerobically adapted algae and not
to the "normal" respiration in plants. We may speculate that the
liberation of oxygen from "hydroxylated" compounds could be com-
bined effectively with the reduction of carbon dioxide only after the
synthesis and the utilization of sugars had become separated. Under
aerobic conditions, the back reactions with free oxygen in the assimila-
tory system had to be prevented. We know that in adaptable algae
this is brought about by the oxidative inactivation of one or two of the
catalysts involved. Apparently we have here a parallel to the well-
known case in which anaerobic fermentations are prevented from con-
tinuing in air by a special oxidation, the so-called "Pasteur reaction."

References

(1) Brown, N. C, Forest Products. 2nd ed., Wiley, New York, 1927.

(2) Emerson, R. L., Stauffer, T. F., and Umbreit, W. W., "Relation-
ship between phosphorylation and photosynthesis in Chlorella" Am. J.
Botany, 31, 107 (1944).

(3) Forestry Depletion in Outline. Northwest Regional Council, Portland,
1940 (25^).

47

H. GAFFRON

(4) Franck, J., and Herzfcld, K. F., "Contribution to a theory of
photosynthesis," J. Phys. Chern., 45, 978 (1941).

(5) French, C. S., and Rabideau, G. S., "The quantum yield of oxygen
production by chloroplasts suspended in solutions containing ferric
oxalate," J. Gen. Physiol., 28, 329 (1945).

(6) GafTron, H., "tJber den Stoffwechsel der Purpurbakterien," Part I,
Biochem. Z-, 260, 1 (1933); Part II, ibid., 275, 351 (1935).

(7) Gaffron, H., "Photosynthesis, photoreduction and dark reduction
of carbon dioxide in certain algae," Biol. Rev., 19, 1-20 (1944).

(8) Gaffron, H., "o-Phenanthroline and derivatives of vitamin K as
stabilizers of photoreduction in Scenedesmus," J. Gen. Physiol., 28, 259 (1945).

(9) Hill, R., and Scarisbrick, R., Proc. Roy. Sac. London, B129, supple-
ment, 39 (1940).

(10) Krebs, H. A., "Carbon dioxide assimilation in heterotrophic or-
ganisms," Ann. Rev. Biochem., 12, 529 (1943).

(11) Lipmann, F., J. Biol. Chem., 158, 515 (1945).

(12) Ochoa, S., "Isocitric dehydrogenase and carbon dioxide fixation,"
J. Biol. Chem., 159, 243 (1945).

(13) Planning for a Permanent Agriculture. U. S. Dept. Agr., Misc. Pub.
No. 351 (1939).

(14) Rabinowitch, E., Photosynthesis, Vol. I. Interscience, New York, 1945.

(15) Ruben, S., "Photosynthesis and phosphorylation," J. Am. Chem.
Soc, 65, 279 (1943).

(16) Sinnott, E. W., "Plants and the material basis of civilization," Am.
Naturalist, 79, 28 (1945).

(17) Spoehr, H. A., "The culture of albino maize," Plant Physiol., 17, 397
(1942).

(18) United Nations Conference on Food and Agriculture, Report, Hot
Springs, May, 1943.

(19) van Niel, C. B., "On the morphology and physiology of the purple
and green sulphur bacteria," Arch. MikrobioL, 3, 1 (1931).

(20) van Niel, C. B., "The bacterial photosyntheses and their importance
for the general problems of photosynthesis," in Advances in Enzymology, Vol. I.
Interscience, New York, 1941, p. 263.

"The culture, general physiology, morphology, and classification of the
non-sulfur purple and brown bacteria," Bad. Revs., 8, 1 (1944).

(21) Warburg, O., and Liittgens, W., "Weitere Experimente zur Kohlen-
s'sLureassimilation," Naturwissenschaften, 32, 301 (1945).

(22) Went, F. W., et al., "Plant growth under controlled conditions," several
articles in Am. J. Botany, 31-32, 1944-1945.

(23) Werkman, C. H., and Wood, H. G., "Heterotrophic assimilation of
carbon dioxide," in Advances in Enzymology, Vol. 11, 1942, p. 135.

48

THE BACTERIAL CELL

RENfi J. DUBOS, MEMBER OF THE ROCKEFELLER INSTITUTE FOR MEDICAL

RESEARCH, NEW YORK; MEMBER OF THE NATIONAL ACADEMY OF SCIENCES;

JOHN PHILLIPS MEMORIAL AWARD; MEAD JOHNSON AWARD

Only such substances can be anchored at any particular
part of the organism which fit into the molecule of the
recipient combination as a piece of mosaic fits into a
certain pattern.

PAUL EHRLICH

B

BACTERIA appeared to the nineteenth century biologist
as a type of protoplasmic material devoid of any organi-
zation, almost as a link between the animate and the inanimate world.
"They are," said Ferdinand Cohn, "the simplest and lowest of all
living forms — beyond them, life does not exist." The compound
microscope failed to reveal any structure within their cellular bound-
aries, and the biochemist was inclined to consider the bacterial cell
as a mere bag of enzymes which owed its enormous biochemical
activity to its colloidal dimensions. The primitiveness of bacterial
life appeared to be confirmed in chemical terms when Winogradsky
demonstrated in 1887 that certain autotrophic species can grow in
purely inorganic media and can synthesize their protoplasm from
mineral salts and carbon dioxide, utilizing for the reduction of the
latter the energy released by the oxidation of sulfur, iron, ammonia,
nitrite, etc. (21). Was it not permissible to consider this production
of organic matter from inorganic elements as the most primitive bio-
chemical expression of life, as the beginning of life on earth?

Advances on the diverse fronts of bacteriology were quick to
dispel these early illusions concerning the biochemical primitiveness
and the simplicity of organization of the bacterial cell. Analysis of
the chemical activities of bacteria soon revealed that microbial life takes
place through the agency of the same type of reactions, the same

49

R. J. DUBOS

metabolic channels and products, and the same biocatalysts which
constitute the mechanism of life in the highest and most evolved
organisms. For example, the oxidation of sulfur by the autoti'ophic
bacterium Thiobacillus thiooxidans depends upon an intimate linking
between oxidation and phosphate turnover; the oxidative phase is
accompanied by phosphate fixation and the reductive phase of carbon
dioxide fixation is accompanied by a release of phosphate (22). More-
over, Thiobacillus thiooxidans is fully equipped with the regular comple-
ment of water-soluble vitamins found in other living organisms: thia-
min, riboflavin, nicotinic acid, pantothenic acid, pyridoxine, biotin,
etc. (17). In other words, the mechanisms of energy transfer and of
intermediary metabolism are essentially as complex in the least exact-
ing bacteria as they are in the most fastidious organism.

The chemistry of Thiobacillus thiooxidans is not unique among
bacteria; most of the known water-soluble vitamins — with the possible
exception of ascorbic acid — have now been found to be either produced
by, or required for the growth of, all the microbial species so far studied.
Fat-soluble vitamins also probably play a part in microbial metabolism
since at least one of them, vitamin K, is an essential growth factor for
Johne's bacillus {Mycobacterium paratuberculosis) and the other myco-
bacteria produce biologically active naphthoquinones during growth
(24).

Thus, bacteria utilize the multiple and complex biocatalysts
which govern and integrate the metabolism of all living cells. More-
over, many bacteria, the autotrophic species for example, possess in
addition the ability to synthesize these same biocatalysts from inorganic
elements in the course of their growth in synthetic media, a property
which most plant and animal cells never possessed or have lost entirely.
The high degree of cellular organization required for the performance
and for the integration of these complex syntheses need not be em-
phasized; at the biochemical level, at least, there is no ground to
consider that bacteria represent primitive forms of life. The growth
requirements of autotrophic bacteria are extremely simple indeed,
but how complex their vital machinery, their performance, and their
products ! If they are truly the first representatives of life on earth,
they sprang, like Minerva, fully armed from the forehead of Jove.

Simultaneously with the realization that the biochemical
processes of bacteria are no simpler than those of other organisms cane

50

THE BACTERIAL CELL

the recognition of the existence in the bacterial cell of a number of
structures which, although often ill defined in nature and function,
obviously express a morphological complexity parallel to the bio-
chemical complexity of all known forms of life. Utilization of the
classical methods of cytology soon revealed, for example, the existence
in bacteria of flagella, spores, different kinds of membranes and capsules
etc., which give to each bacterial type a fairly characteristic morpho-
logical individuality (12). Little by little, bacteriological staining
techniques are gaining the dignity of cytochemical reactions and give
chemical definition to the cellular objects which they reveal; the more
skillful utilization of Feulgen's reagent for instance, permits the identi-
fication among the other basophilic constituents of the bacterial cell of
discrete bodies rich in desoxyribonucleic acid which are almost cer-
tainly the equivalent of the vesicular nucleus in larger cells (12,19).
Photography in the ultraviolet and electron microscopy have per-
mitted the optical resolution of cellular structures — intracellular
granules, membranes, individual components of flagella, etc. — which
are below the limit of resolution by ordinary microscopy. These
classical cytological techniques aim at the direct visualization of the
constituents of the cell. On the other hand, the analysis of the re-
sponse of the cell to the effect of certain reagents and procedures pro-
vides an indirect approach to cytological problems, by suggesting the
existence and often the chemical nature of important cellular com-
ponents which cannot be seen by any of the known methods of micros-
copy. Interestingly enough, it is the study of pathogenic bacteria
which has been the most fruitful from the point of view of this indirect
approach to cytology.

In order to analyze the host-parasite relationship, the student
of infection must concern himself with those structures and products
of bacteria — the cellular antigens and toxins — which affect the course
of the infectious process and against which are directed the reactions
of immunity. Similarly, the attempts to understand the mode of
action of antiseptics on bacteria led to the study of the structures
through which antiseptic agent and susceptible cell come in contact.
Thus, many constituents of the bacterial cell have been recognized
first by biological reactions; and the analysis of the phenomena of
infection, immunity, and chemotherapy has provided important in-
formation concerning the biochemical architecture of bacteria. Paul

51

R. J. DUBOS

Ehrlich first stated clearly the possibility of describing these reactions
in terms of cellular structure. He postulated that the living cell
possesses a number of chemically reactive groups which he called
"receptors" and with which dyes, bactericidal substances, and immune
antibodies react selectively. Ehrlich regarded these "receptors" as
definite chemical entities capable of entering into union with dyes,
antiseptics, and antibodies. According to his theory, characteristic
staining reactions, diflferential susceptibilities to toxic substances, and
specific reactions with corresponding antibodies could all be explained
by assuming the existence of a sufficient number of receptors in the
bacterial cell. These phenomena can consequendy serve as tests to
facilitate the recognition of the receptors and their isolation in pure
form. During the past decades, immunochemists and students of the
theory of chemotherapy have gone far toward identifying, and in
several cases separating in a purified state, several of the cellular
components with which antibodies and antibacterial agents react
selectively.

The specific chemical relationships involved in the chemo-
therapeutic reactions are discussed in other essays of this volume and
need not be considered here. There are certain aspects of the problem,
however, which are too ill-defined to warrant discussion in terms of a
chemical theory, but which deserve mention at this time since they bid
fair to help in the elucidation of some interesting details of cellular
structure. Empirical staining reactions led very early to the division
of the bacterial world into three broad groups, the Gram-positive, the
Gram-negative, and the acid-fast, which are defined by their behavior
toward two staining techniques (the Gram and Ziehl-Neelsen methods) .
These three bacterial groups not only differ in their staining properties,
but also exhibit striking differential susceptibilities to the different
types of antiseptics and antibacterial agents. By way of illustration,
most Gram-negative bacilli grow readily in the presence of basic dyes,
penicillin, or gramicidin, whereas these organisms are extremely
susceptible to the bactericidal and lytic effect of immune serum. On
the contrary, many Gram-positive species are completely resistant to
lysis by immune serum, but are extremely susceptible to the bacterio-
static and bactericidal effect of small concentrations of dyes, penicillin,
and gramicidin. As for the tubercle bacilli (acid-fast), they are re-
markably resistant to all the classical antisejitics and to a gi'eat variety

52

THE BACTERIAL CELL

of other toxic agents, retaining their viabihty, for example, after ex-
posure to 5% sodium hydroxide or sulfuric acid. In order to account
for these strikingly different susceptibilities, many theories have been
proposed which assume that the bacterial groups vary with reference
to cellular permeability, presence of certain lipids in the cell mem-
branes, acid-base properties of the cell body, etc. Thus, observations
of electrokinetic behavior and of affinity for dyes at different pH levels
suggest that the cell material in the colon-dysentery-typhoid group
(Gram-negative) is less acidic than in the Gram-positive bacterial
species (pneumococci, staphylococci, streptococci, anthrax bacilli,
etc.) (1,20). The acid-fast bacilli (e. g., tubercle bacilli) produce
astonishingly high concentrations of a variety of lipids which gives
them marked hydrophobic properties (2,23). All the Gram-negative
species readily yield in solution phospholipid-protein-polysaccharide
complexes which constitute 5-10% of the total cell weight; similar
complexes have not been obtained from the Gram-positive organisms
(16,18). Recent observations have established a correlation between
ability to retain the Gram stain (correlated with greater susceptibility
to many antiseptics) and the presence around the cell of a magnesium
ribonucleate complex (5,10). All these facts provide examples of the
type of chemical information which results from the analysis of the
biological behavior of the cell and which will undoubtedly reveal
important differences of structure between the different bacterial
groups.

The specific antibodies produced as the result of the injection
of bacterial antigens into the animal body have provided another set
of reagents that have yielded important information of a cytochemical
nature. The immune reaction to any one type of bacterial cell is not
a simple phenomenon, since bacteria are made up of a multiplicity of
chemical constituents many of which elicit the production of specific
antibodies. In other words, the injection of one type of bacterial cell
usually results in the production of several antibodies, each one of which
is directed against one particular cellular component. These different
cellular constituents obviously bear a definite spatial morphological
relationship to each other in the intact cell. Some are masked by
membranes and become exposed only as a result of cellular disintegra-
tion; others are peripherally disposed and in direct contact with the
environment. This stratification of cellular structures affects the im-

53

R. J. DUBOS

mune response of the animal host to the whole bacterium and is re-
flected in the type and amount of antibody produced. It also condi-
tions the reaction of the bacterial cell with a given antibody, since the
intact microorganism unites much more readily, if not solely, with the
antibodies which are specifically directed against those of its con-
stituents exposed at or near the surface. Analysis of antibody pro-
duction and of antigen-antibody reaction can therefore help in formu-
lating an approximate picture of the arrangement of the different
antigens in the architecture of the cell.

To summarize, the use of immunological procedures for the
study of cellular structure involves a number of successive steps:
(a) the preparation and separation of antibodies specific for each one
of the chemical constituents of the cell; (b) the utilization of these
antibodies as specific reagents for the detection and preparation in
pure form of the cellular constituents; and finally, (c) the interpreta-
tion of antigen-antibody reaction in an attempt to define the relative
positions occupied by these chemical constituents in the living cell
(7,9,11,14).

The results obtained by immunochemical analysis have led to
the recognition that the different bacterial groups (pneumococci,
streptococci, sporulating aerobic bacilli, organisms of the colon-
typhoid-dysentery group, etc.) are characterized by a general pattern
of antigenic organization which is common to the different members
of each group. On the other hand, the various species and immuno-
chemical types within any general group differ from each other by
virtue of the chemical specificity of the different components of their
antigenic mosaic. Thus, all virulent pneumococci possess a capsule
which is polysaccharide in nature, but the polysaccharide varies
chemically and antigenically from type to type (3) . The virulent forms
of human streptococci (group A) can also produce a capsule made up
of hyaluronic acid; moreover, they all possess as surface constituents
peculiar proteins (the M substances), and other substances of unknown
chemical nature (the T antigens), which vary in immunochemical
specificity from type to type (13). Several of the aerobic sporulating
bacilli have been found to produce a capsule consisting essentially, if
not solely, of polypeptides; the polypeptide, in the case of the anthrax
bacillus, appears to be made up exclusively of ^/-glutamic acid (6).
The virulent coliform bacilli (typhoid, dysentery, etc.) all produce the

54

THE BACTERIAL CELL

lipid-protein-polysaccharide complexes mentioned above. The poly-
saccharide components determine the immunochemical specificity of
the different species. The protein component, on the contrary, ap-
pears to be essentially the same in all the members of the group; in
fact, it can be made to combine with the different specific polysac-
charides to reconstitute complexes similar to those which normally
occur in the cell (16).

In general, the immunochemist has concerned himself primarily
with the constituents which are present at the cell surface becaase
these elements appear to play the most important part in the phe-
nomena of immunity. However, other proteins, polysaccharides, etc.,
which certainly occupy less superficial positions in the cell have also
been recognized by immunochemical analysis. A striking illustration
of the potentialities of immunochemical methods in cytology is given
by the case of the typhoid bacillus. Specific antibodies have been
prepared for the following cellular components of this organism: tlie
flagella; the O and Vi antigens of the cell surface; the R, p, and Q
antigens, which are intracellular components. Living typhoid bacilli
resuspended in solutions of these different antibodies exhibit a charac-
teristic behavior which is determined by the relative position of the
corresponding antigen in the cellular architecture. Loss of motility,
different patterns of agglutination, bacteriolysis, etc., are phenomena
which are characteristic for each antigen-antibody reaction and
which can be interpreted in terms of cellular organization of the
different antigens.

Dyes, antiseptics, and antibodies are not the only reagents
which can be used to recognize and identify the cellular receptors.
If it be found, for example, that a given enzyme attacks the cells of a
certain microbial species causing death or the alteration of a character-
istic cellular property, it can be surmised that the chemical substrate
which is susceptible to the enzyme is present in the cell under con-
sideration and that it plays some part in the function altered by the
enzyme. Thus, the fact that lysozyme causes the death and lysis of
the cells of several bacterial species indicates that the mucopoly-
saccharides hydrolyzed by this enzyme are essential components of
the cellular structure of the susceptible species (15). It has been shown
also that other polysaccharidases decompose the capsular polysac-
charides of pneumococci and of streptococci, and that proteolytic

55

R. J. DUBOS

enzymes inactivate the specific M protein antigens of group A strepto-
cocci (8,13). Contrary to what is observed with lysozyme, however,
the polysaccharidases and proteases do not aflfect in any way the
viabiHty of the treated cells, even when they rid it entirely by hydrolysis
of the specific polysaccharides or proteins. It is likely, therefore,
that, instead of being considered as structural constituents of the
bacterial bodies, the capsular polysaccharides and the M proteins
should be regarded as excretion products which accumulate around
the cell, since they can be removed or destroyed without interfering
with the essential living processes.

There are many other instances of biological reactions which,
because of the specific relationships they bear to certain cellular com-
ponents of bacteria, can be used as indirect methods for the analysis
of cellular structures. Let us mention, for example, the remarkable
selectivity of pure lines of bacteriophage with reference to the strains
of bacteria which they can attack. The relationship between speci-
ficity and cellular structure is illustrated by the fact that the bacterio-
phage can be absorbed specifically by the cells, living or dead, of the
susceptible bacterial cultures. It has also been found that, in certain
cases, soluble fractions extracted from the susceptible bacterial cells
inhibit specifically the lysis of the homologous organisms by the bac-
teriophage. It is likely, therefore, that the phenomena of bacterio-
phage lysis will also yield a number of new specific reactions which
by revealing the existence and nature of new types of receptors could
serve in the analysis of cellular structure.

The biological phenomena which we have considered have in
common the following characteristics permitting their utilization as
indirect cytological methods. They are all the result of a reaction
between a given reagent (antiseptic, enzyme, antibody, bacterio-
phage) and a specific cellular receptor, this reaction manifesting itself
by inhibition of growth, enzymic destruction of cellular component,
agglutination or lysis by antibody, or lysis by bacteriophage. In
many cases the reagent can be absorbed on the homologous cell sub-
strate, and the reaction can be inhibited by the addition to the system
of the specific substrate which constitutes the cellular receptor. In-
hibition of growth, enzymic decomposition, agglutination, lysis, etc,,
are only the secondary manifestations of primary reactions which de-
pend upon the union between the cellular receptors, on the one hand,

56

THE BACTERIAL C:i;i,I,

and the biological reagents, be they antiseptics, antibodies, enzymes,
bacteriophages, on the other.

One of the most intriguing applications of the indirect cyto-
logical methods discussed in the preceding pages has been the analysis
of the phenomena of bacteria variability. It has long been known
that most bacterial cultures — even those arising from single cells —
often undergo profound transmissible modifications of their morpho-
logical, biochemical, and physiological properties. Immunochemical
studies have revealed, in particular, a type of variation, now recognized
in practically all bacterial species, which involves the loss of the specific
surface components of the cell (the capsular polysaccharides of pncu-
mococci, the M proteins of streptococci, the capsular polypeptide of
anthrax, the lipid-protein-polysaccharide complexes of the dysentery
and typhoid bacilli, etc.). These transmissible modifications of the
surface of the bacterial cell have attracted particular attention because
they are in many cases correlated with alteration of the virulence of
the organism concerned. They constitute, however, only a very nar-
row aspect of the total problem of bacterial variability. One can
observe within one given bacterial culture transmissible modifications
of many unrelated properties: ability to attack sugars or proteins, to
synthesize amino acids or pigments, to resist antiseptics or other
injurious procedures, to produce flagella, spores, capsules, and so on.
All these variations occur independently of each other, thus giving to
each bacterium the possibility of manifesting its existence under a
great diversity of forms and properties. The production of these large
numbers of variant forms deficient in one or another of the cellular
components has greatly helped in the analysis of many immuno-
chemical problems.

From a more general point of view, it is of the greatest interest
that a given organism can successfully continue to exist and to multiply
as an independent living object after having lost a great variety of
structures and functions which had appeared to constitute important
components and attributes of the "normal" parent form. As already
stated, these structures and functions can be lost and regained inde-
pendently of each other, without altering the essential nature of the
germ or the potentialities of the cell. Even more striking is the fact
that it is possible to substitute experimentally one character for an-
other. Thus, by adding to a strain of pneumococcus which has lost

57

R. J. DUBOS

the ability to produce its specific capsular polysaccharide an extract
of the cell of another type of encapsulated pneumococcus, one can
convert the former organism into the type from which the extract was
made. From then on, the cell can produce, and transfer to its progeny
the ability to produce, a polysaccharide different from the one it had
been known to synthesize heretofore. All available evidence indicates
that the substance which is capable of inducing the transformation is
a form of desoxyribonucleic acid- — specific for each pneumococcus
type (4). Of equal interest is the unavoidable conclusion that the
bacterial cell is not only an integrated complex of independent charac-
ters, but that it is possible to substitute for one of these characters an-
other one, homologous but different, without interfering essentially
with cellular organization.

Thus, a large body of knowledge concerning cytology is slowly
emerging from the study of bacterial variability and of the behavior of
the cell in the presence of a number of biological reagents. It is be-
coming possible to recognize and to define in chemical terms a number
of structures not yet detectable by any microscopic technique. Further-
more, by bold, even though admittedly dangerous, extrapolation, one
can guess at the approximate position of these cellular constituents
in the architecture of bacteria. The history of science provides, of
course, many examples of the fruitfulness of indirect methods, and in
particular of the utilization of chemical and biological manifestations
as indices and guides for the recognition and identification of morpho-
logical structures. Claude Bernard stated as early as 1855 that
"anatomical localization is often revealed first through the analysis
of the physiological processes." Much of the morphology indirectly
revealed by antibodies, enzymes, and cytotoxic substances lies beyond
the microscopic range and in fact often reaches the molecular level.
It concerns the organization of those molecular groupings which,
because of their chemical reactivity, condition the behavior of the cell
both as an independent functioning unit and in its relation to the
environment. This knowledge is of obvious interest to the cytologist.
It also forms the fundamental basis upon which is being erected the
theory of immunity, since all the phenomena of host-parasite relation-
ship are essentially a reflection of the biochemical architecture of the
cell.

58

THE BACTERIAL CELL

References*

(i) Albert, A., "Chemistry and physics of antiseptics in relation to mode of
action ," Lancet, 1942, II, 633-636.

(2) Anderson, R. J., "The chemistry of the lipids of the tubercle bacilli,"
Harvey Lectures, 35, 271-313 (1939-1940).

(3) Avery, O. T., "The role of specific carbohydrates in pncumococcus
infection and immunity," Arm. Internal Med., 6, 1-9 (1932-1933).

(4) Avery, O. T., MacLeod, C. M., and McCarLy, M., "Studies on the
chemical nature of the substance inducing transformation of pneumococcal
types," J. Exptl. Med., 79, 137-158 (1944).

(5) Bartholomew, J. W., and Umbreit, W. W., "Ribonucleic acid and the
Gram stain," J. Bact., 48, 567-578 (1944).

(6) Bovarnick, M., "The formation of extracellular a'(-)glulamic acid
polypeptide by Bacillus subtilis,'' J. Biol. Chem., 145, 415-424 (1942).

(7) Boyd, Wm. C, Fundamentals oj Immunology. Interscience, New York
1943.

(8) Dubos, R. J., "Enzymatic analysis of the antigenic structure of pneumo-
cocci," Ergeb. Enzymjorsch., 8, 135-148 (1939).

(9) Heidelberger, M., "Immunology as a tool in biological research," Am.
Naturalist, 77, 193-198 (1943).

(10) Henry, H., and Stacey, M., "Histochemistry of the Gram-staining
reaction fcjr micro-organisms," Nature, 151, 671 (1943).

(11) Kabat, E. A., "Immunochemistry of proteins," J. Immunol., 47,513-587
(1943).

(12) Knaysi, G., Elements oj Bacterial Cytology. Comstock, Ithaca, 1944.

(13) Lancefield, R. C., et al., "Studies on the antigenic composition of group
A hemolytic streptococci," J. Exptl. Med., 78, 465-476 (1943); 79, 79-114
(1944). "Specific relationship of cell composition to biological activity of
hemolytic streptococci," Harvey Lectures, 36, 251-290 (1940-1941).

(14) Landsteiner, K., TheSpecificityoJ Serological Reactions. Rev. ed.. Harvard
Univ. Press, Cambridge, 1944.

(15) Meyer, K., Palmer, J. W., Thomf:)son, R., and Khorazo, D., "On
the mechanism of lysozyme action," J. Biol. Chem., 113, 479-486 (1936).

(16) Morgan, W. T. J., and Partridge, S. M., "An examination of the O
antigenic complex of Bact. typhosum,'" Brit. J. Exptl. Path., 23, 151-165
(1942). "Studies in immunochemistry. 6. The use of phenol and of

* The material discussed in this essay is presented in a more complete
manner and with extensive documentation in a monograph, The Bacterial Cell,
by R.J. Dubos and C. Robinow, Harvard University Press, Cambridge, 1945.

59

R. J. DUBOS

alkali in the degradation of antigenic material isolated from Bad. dysenteriae
(Shiga)," Biochem.J., 35, 1140-1163 (1941).

(17) O'Kane, D. F., "The presence of growth factors in the cells of the auto-
trophic sulphur bacteria," J. Bad., 43, 7 (1942).

(18) Partridge, S. M., and Morgan, W. T. J., "Immunization experiments
with artificial complexes formed from substances isolated from the antigen
oiBad. Shigae," Brit. J. Exptl. Path., 21, 180-195 (1940).

(19) Ribonow, C. F., "A study of the nuclear apparatus of bacteria," Proc.
Roy. Soc. London, 130, 299-328 (1942). "Gytological observations on Bad.
colt, Proteus vulgaris and various aerobic spore-forming bacteria with spe-
cial reference to the nuclear structures," J. Hyg., 43, 413-423 (1942).

(20) Stearn, A. E., and Stearn, E. W., "Metathetic staining reactions with
special reference to bacterial systems," Protoplasma, 12, 435-464, 580-600
(1931).

(21) van Niel, C. B., "Biochemical problems of the chemo-autotrophic
bacteria," Physiol. Revs., 23, 338-354 (1943).

(22) Vogler, K. G., and Umbreit, W. W., "Studies on the metabolism of the
autotrophic bacteria," J. Gen. Physiol., 26, 157-167 (1942).

(23) Wells, H. G., and Long, E. R., The Chemistry oj Tuberculosis, 2nd ed.
rev., Williams & Wilkins, Baltimore, 1932.

(24) Woollcy, D. W., and McCarter, J. R., "Antihcmorrhagic compounds
as growth factors for the Johne's Bacillus," Proc. Soc. Exptl. Biol. Med., 45,
357-360 (1940).

6o

THE NUTRITION AND BIO
CHEMISTRY OF PLANTS

D. R. HOAGLAND, professor of plant nutrition, college of
agriculture; plant physiologist, agricultural experiment

STATION, university OF CALIFORNIA

OTHER writers for this volume will discuss the biochemistry
of plants in relation to photosynthesis, plant hormones,
and the activities of microorganisms. The present article is, there-
fore, devoted primarily to an attempt to indicate the need and the
opportunities for research on the biochemistry of higher plants, es-
pecially plants of agricultural interest, as a foundation for the adequate
understanding of many problems of plant nutrition and of plant
physiology. This field of inquiry is relatively undeveloped in the
modern period in comparison with the biochemistry of higher animals
and of microorganisms, with its remarkable record of achievement
during the past quarter of a century. The importance of the bio-
chemistry of the higher plants for the cycles of living organisms in
general, and for the basic occupation of agriculture is too obvious to
require analysis.

The disparity of achievement in fundamental biochemical re-
search dealing with higher plants, on the one hand, and with the higher
forms of aniinal organisms and some groups of microorganisms on the
other, becomes apparent on examination of recent monographs dealing
with advances in biochemistry. They are predominantly concerned
with experiments on animal tissues, or on microorganisms, and the
majority of contributors are associated with medical research in-

6i

D. R. HOAGLAND

stitutes. Recent texts on biochemistry give but little specific attention
to the biochemistry of higher plants. Certain earlier treatises devoted
largely to this latter subject have not appeared in new editions for a
good many years. This general appraisal on a comparative basis
appears to be justified even when the noteworthy contributions of
individual workers or of certain groups of workers on the biochemistry
of higher plants are kept in view. It receives support from some of
the comments of Vickery, who is well known as an extensive con-
tributor to several phases of the biochemistry of plants.

The situation as described may seem surprising when one recalls
the vast programs of research carried on by agricultural experiment
stations. It is true that, in these stations, a great number of studies
on plants have been made which are to some degree biochemical in
nature. But it is rare to find groups of investigators assigned the
definite objective of developing the knowledge of the fundamental
biochemistry of crop plants. Generally, biochemical studies are
encouraged in so far as they throw light on particular questions of
agricultural importance as related to plant nutrition, horticulture,
agronomy, or perhaps general plant physiology. In terms of crop
production, the success of the coordinated attacks on plant problems
through the application of the agricultural sciences and arts is well
demonstrated by the enlarged production of crops under the difficulties
of war conditions. This, however, does not meet the point we have
under discussion. Further, there is reason to assume that, even from
a utilitarian point of view, a more intensive development of research
on the biochemistry of crop plants would in due course contribute to
the basic knowledge essential to the control of plant production and
supplement and guide the interpretation of results of practical experi-
mentation.

Ramifications of the biochemistry of plants into the applied
field are manifold. The growth of crops needs to be appraised by
the criteria not only of total yield but also of quality. This latter
aspect is currently receiving much attention through consideration
of plant composition in relation to the value of the plant product for
animal nutrition, a point illustrated by the research program of the
Federal Soil, Plant, and Nutrition Laboratory at Cornell. For
example, the general problem of the synthesis of vitamins by plants
might be cited, as well as the studies made to gain infoi-mation on the

62

BIOCHEMISTRY OF PLANTS

relative influence of climate and mineral nutrition of the plant on its
vitamin content. Also, the biochemical mechanisms in the plant that
result in the synthesis of essential amino acids, carbohydrates, and
higher fatty acids are of interest to students of both plant and animal
metabolism. Knowledge of the biochemistry of the plant is one of the
sources of information that constitutes a valuable asset to the plant
pathologist, the plant geneticist, the horticulturist, the specialist in
forestry, and indeed to all those who must of necessity come into contact
with plant systems of biochemical reactions, whether or not this is
consciously recognized.

From some points of view, the plant offers, as compared with
the animal, methods of study with marked advantages. On the other
hand, there are disadvantages in the use of the higher plant for bio-
chemical investigation. The nature of a plant's growth is such that
most of its living cells perform the most diversified functions, a fact
which renders plant cells less suitable for research on specific bio-
chemical reactions. Tissues from animal organs often provide ma-
terial with specialized and highly intense activities suitable for the
investigation of enzyme reactions. It is doubtful that the plant in
general is as favorable as the animal for similar investigations, but this
view may be held only because fewer attempts have been made to
exploit the possibilities of plant material. There is no opportunity to
carry on with the plant the kind of experiments that are rendered
possible by the presence of a blood stream and organs of elimination.
The introduction of specific organic metabolites into the plant cell and
the study of their transformations within the cell involve special compli-
cations. In the examination of biochemical reactions taking place in
excised plant tissues, avoidance of bacterial and fungal contamination
often presents great difficulties. Further, the correlating eff"ects of
various plant hormones and of normal translocation of metabolites
make especially difficult the interpretation of the results of studies of
plant tissues in terms of the intact growing plant.

Plants of the kind under consideration may be regarded as
normally complete synthetic systems, building up or breaking down
compounds representing an extraordinary array of organic structures
of biological interest, all derived from the simple substances required
for plant growth, namely, carbon dioxide, water, and inorganic ele-
ments to the number of fifteen or more. It might be postulated that

63

D. R. HOAGLAND

certain species of plants growing in some types of soil high in organic
matter may have lost the power of synthesizing at an adequate rate
vitamins or other essential organic units, and that these species depend
in part on the absorption of these units from an environment in which
they have been synthesized by microorganisms. But most or all species
of higher green plants so far intensively studied (these are mainly
plants of economic importance) can go through their cycles of growth
by virtue of their own synthetic powers, at least so far as can be ascer-
tained by the use of purified inorganic media, although usually without
complete exclusion of all microorganisms. It follows that the range
and diversity of biochemical reactions that need to be investigated
is enormous. Correspondingly great are the opportunities for the
study of different synthetic processes in a living organism, to the extent
that adequate methods can be devised for attacking such complex
systems.

The preoccupation of plant physiologists engaged in agricultural
research with the inorganic elements absorbed by crop plants from the
soil, and the frequent designation of these elements as "plant foods,"
tend to subordinate appreciation of the biochemical aspects of plant
nutrition. The inorganic elements derived from the soil constitute
only a small percentage of the dry weight of a plant, and one of the
most important objectives of research in plant nutrition should be an
understanding of the mechanisms by which these inorganic com-
ponents become directly incorporated into the organic compounds
synthesized by the plant, or activate the enzymes which catalyze the
syntheses and breakdown of organic compounds.

We have as a foundation the knowledge that plants of the kind
in question absorb from an inorganic medium, and have an essential
need for, the elements nitrogen, phosphorus, potassium, calcium,
magnesium, sulfur, and iron. Also, as a result of research in plant
nutrition during the past decade or two, other elements have been
shown to be equally essential though required in only minute quan-
tities. These additional elements include boron, manganese, copper,
and zinc, with strong but still limited evidence that molybdenum is
also essential. There may be, and probably are, still other chemical
elements indispensable to plant growth, but conclusive proof of the
general indispensability of other elements, over a wide range of plant
species, has not yet been obtained. Limitations of technique are soon

64

BIOCHF.MISTRV OF PLANTS

encountered as the effort is made to go further in the exckision of im-
purities from the nutrient medium. It should be noted, however, that
various chemical elements not indispensable for growth of the plant
may modify its biochemical reactions either beneficially or adversely,
under the conditions of a natural environment.

Many of these facts have been established primarily through the
use of artificial culture methods, among which the so-called water-
culture method is especially useful for studying the effects of a deficiency
of an element needed by the plant in minute quantity. Sometimes,
however, special care in the selection and purification of a solid inert
medium, to which a purified nutrient solution is applied, pro\ides an
alternative technique, with certain advantages.

Innumerable experiments, some of them with meticulous care,
have been made on many species of plants with these artificial culture
techniques. The control of the inorganic nutrient medium represents
only a partial control of the environment; and frequently there remains
the necessity or desirability of control of the atmospheric factors to
which the plant is subjected: light, temperature, humidity, carbon di-
oxide concentration, and air movement. The control of these factors
obviously demands costly equipment and usually is not attempted in
nutritional experiments with plants. But some laboratories have had
the opportunity to grow plants under conditions of controlled air tem-
perature and controlled artificial illumination. In recent years,
fluorescent lamps have proved especially valuable for this purpose.

Most artificial culture experiments have not been designed for
the primary purpose of obtaining information about the metabolic
mechanisms of the plant and the specific enzyme systems concerned.
Observations on plants grown in the presence of selected combinations
of inorganic nutrients are likely to be confined to measurements of rate
of growth of the plants, total yields of the tissue produced on a fresh
or dry weight basis, or on yields of some part of the plant of special
interest from an agricultural point of view. Chemical studies are
often limited to the determination at the end of a selected growth
period of certain chemical elements absorbed by the plant, or of well-
known organic compounds formed as a net result of innumerable bio-
chemical processes that have proceeded perhaps for a considerable
period of time during which the plant has increased in size and difleren-
tiated its tissues. In other cases, the purpose may he to record the

65

D. R. HOAGLAND

pathological symptoms resulting from a marked deficiency in the
medium of some one of the essential inorganic elements. Valuable as
this information is for the purpose in view, it does not advance our
understanding of the biochemistry of plants in a manner at all com-
parable with the advances made in the study of animal tissues or of some
microorganisms, in which definite steps in a series of chemical reactions
are identified or reasonably deduced from experimental data.

The use of the artificial culture methods of plant nutrition makes
feasible the growing of plants of many species with any desired combi-
nation of inorganic nutrients, or with a given nutrient available in
graduated quantities. Controlled modifications in the inorganic
composition of the plant are thereby induced, although generally in
no simple relation to the composition of the nutrient solution. As
already stated, possibilities exist for control of illumination and tem-
perature and, thus, to some degree for control of carbon assimilation
and rates of metabolic reactions.

It is tempting to propose that these methods of controlled culture
afford techniques for endless rewarding studies on biochemical mecha-
nisms in the plant. To what extent this is a realistic view is
difficult to say. The extraordinary complexity of the growing plant
and of the conditions of its nutrition may set narrow limits to what
can be done of fundamental biochemical importance, but whatever
opportunities do exist have yet to be adequately explored. There are,
of course, other methods of experimentation on plants which can be
adapted to biochemical research, such as embryo culture, culture of
root tips, and experiments with excised leaves, roots, or other parts of
the plant immersed in solutions of known composition. A technique
has been recently described (13) for physiological and chemical studies
on albino plants, whereby the transformations of a known carbo-
hydrate supplied to the plant might be followed, without being com-
plicated or obscured by reactions which are associated with photo-
synthesis.

It may be noted again that from the standpoint of plant nutri-
tion the main biochemical problem is the fate of the chemical elements
derived from the nutrient medium and the way in which they interact
with the products of photosynthesis. Most of the elements essential
to plants are also essential to all organisms; research on their metabolic
functions has therefore a wide biochemical interest. For some cf

66

BIOCHEMISTRY OF PLANTS

the essential inorganic elements, and especially certain "trace" ele-
ments, we have little or no guidance from studies on other organisms.
Boron is indispensable for all the higher plants so far properly investi-
gated. It has not been shown to be indispensable to the animal,
although little work has been done on this point. Boron is one of the
elements plants require in minute amounts, yet deficiencies even under
some soil conditions have assumed first-rate agricultural importance,
a fact which accentuates interest in the biochemical functions of boron.
Research by plant physiologists indicates that deficiency of boron often
results in a pathological state in the plant nearly the same as that caused
by a deficiency of calcium. One view is that an inadequate supply of
boron may limit the maintenance of an effective level of calcium in a
soluble or active form within the tissues of plants. Some workers think
that formation of pectin compounds does not proceed normally when
boron is deficient. Clearly these are questions which need the atten-
tion of skilled biochemists. Possibly productive leads might come from
comparative biochemistry. Certain groups of fungi, and perhaps some
algae, appear not to require boron. The same fungi also can grow
without calcium, or at least the amounts needed are too small to be
removed from the media by present methods of purification. Furthei
study of certain phases of organic metabolism in plant organisms with
different boron or calcium requirements might conceivably point to
biochemical reactions for which boron or calcium, or both, may be
indispensable.

Another among the chemical elements eff"ective in micro
quantities to which an indispensable function in the growth of higher
green plants must be assigned is zinc. This element, like boron, is
not always adequately supplied by the soil, and the deficient plant
becomes diseased ("little-leaf," "motde-leaf" of trees, and pathological
^conditions shown by other crop plants as a result of zinc deficiency).
Physiological studies under the control of artificial culture disclose
some of the effects of zinc deficiency. One such study in this laboratory
yielded evidence that, without an adequate supply of zinc, plant growth
substances of the auxin type are either not synthesized at a rate sufii-
cient for normal growth or else are destroyed too rapidly (12). But
it has also been learned that protein synthesis is retarded when the zinc
concentration in the plant falls below a critical level. Tomato plants
were grown with graduated supplies of zinc, so that some of the plants

67

D. R. HOAGLAND

at the time of the experiment gave no visible symptomatic deficiency
response, yet addition of zinc to the nutrient medium rapidly induced
an increased rate of protein synthesis (1). Observations of this kind,
however, only show that some unknown link in a chain of reactions
is broken. The nature of the enzyme systems existing in the plant, of
which zinc is an essential component, or activator, is an unanswered
question. From investigations on blood cells comes evidence that
zinc is a component of carbonic anhydrase; and there is a suggestion
from studies on yeast that zinc is one of the activators of the enzyme
aldolase. We have no positive evidence of this character derived
from experiments performed directly on enzyme systems of higher

plants.

Zinc is often successfully applied in curing zinc deficiency disease
by spraying the plant or treating the soil. This discovery, valuable as
it is in agricultural practice, does not satisfy the curious investigator
who seeks enlightenment on the function of zinc in plant metabolism.
Knowledge of why zinc is needed by the plant might improve existing
agricultural practice by providing a rational basis for the treatment of
zinc deficiency disease; but clearly, as in other fields of research,
sound progress in the study of plant biochemistry cannot be hoped for
if the direction of research is to be governed by the degree of prob-
ability that a given investigation will in itself have a practical outcome.

These are only illustrations of the wide gaps in fundamental
biochemical insight into the functions of the so-called plant foods.
One could write of the lack of the kind of experimentation which might
elucidate the role of potassium, one of the most important fertilizer
elements in enzymic reactions in the plant. At the present time, it
is possible to cite data from one source or another that could be in-
terpreted in terms of an eff'ect of potassium on almost every general
biochemical process of which the plant is capable. The net conclu-^
sion is that potassium is an essential element for plant growth and that
its deficiency may impair growth in various ways or alter the compo-
sition of the plant, depending upon the factors such as degree of potas-
sium deficiency, the concentration in the media of calcium, sodium,
or other ions, the species of plant studied, and the physiological age of
the plant. The desirability of further basic information on the role of
potassium in biochemical processes in the plant is evident. For
example, a suggestion has been advanced that potassium has an im-

68

BIOCIII'MIS'FRY OI' I'LANIS

portant effect on certain of the pliosphorylation processes in muscle,
withi an antagonistic action by calcium. Studies of this ty])e, applied
to the enzyme systems of higher plants, would obviously be of great
significance for plant nutrition.

If we turn to the essential element phosphorus, great encourage-
ment is gained for the view that biochemical research on animal
tissues and on microorganisms can serve as a guide for extension of
research on the metabolism of the plant. In fact, current research on
phosphorus metabolism of various organisms has far-reaching implica-
tions for problems of plant nutrition from both a theoretical and
practical standpoint.

Following the work of Cori on the phosphorolytic system in
animal tissues which brings about the synthesis of glycogen from
glucose-1 -phosphate came the discovery by Hanes (5) "that an enzyme
system can be prepared from plant tissues (potato, peas) which cata-
lyzes reversible phosphorolytic reactions by which starch can he syn-
thesized in vitro. Physical and chemical studies have been made of
the synthesized starch, and its structure compjared with that of ii;itural
plant starch (7,11).

The conclusion is that the aitificially synthesized starch repre-
sents only the amylose component of natural starch, that is, the one
made up of long chains of glucose units, whereas the natural starch
also includes a component characterized by a branched-chain structure.
Recently, the failure to reproduce artificially the complete natural
starch was apparently overcome. A preliminary report has been
made of the isolation of another enzyme system from potato which can
accomplish the synthesis in vitro of the amylopectin component of
starch, with the branched-chain structure (8). The investigations
as a whole on this question therefore represent clarification of the role
of an essential inorganic element in synthesis by the plant of one of its
most important carbohydrates. It is not difficult to appreciate the way
in which this addition to plant biochemistry may aid in the guidance
of researches in plant nutrition with reference to the utilization of
phosphate. From the agricultural point of view, the information now
available may well become of value in appraising the adequacy of the
phosphate supply for high yields or starch content of a crop, especially
as further research provides more data on concentiations of various
forms of phosphate in the plant under diverse nutrient and atmospheric

69

D. R. HOAGLAND

environments. To what extent, if at all, the relative proportions of
amylose and amylopectin may be subject to modification by physio-
logical conditions remains to be studied.

The phosphorolytic mechanisms probably will supply the key
to an understanding of the synthesis of another carbohydrate almost
universally synthesized by higher plants, namely, sucrose. This is also
the dominant sugar of commerce and, as one authority has pointed
out, is manufactured commercially in far greater quantity than any
other pure chemical product. It is natural, therefore, that many at-
tempts have been made to analyze the mechanism of sucrose synthesis
in the plant. Recently the enzymic synthesis of sucrose in vitro was
accomplished (7); and, while the enzyme system responsible is
derived from a bacterial organism (Pseudomonas saccarophila), the
achievement has such great suggestive interest for the investigator of
the nutrition of higher plants that it seems appropriate to mention it
in this connection. The substrates utilized in the synthesis were glu-
cose-!-phosphate and fructose. Sucrose is broken down into these
components in a phosphorolytic reaction catalyzed by an enzyme
system in the bacterial cell. By starting with glucose-1 -phosphate
and fructose in the presence of the enzyme, pure crystalline sucrose
was prepared which was identical with the natural product. The
identity was established by all available physical and chemical criteria.
This is the first well-authenticated synthesis of this sugar.

It is true that a similar enzyme system has not yet been iso-
lated from the tissues of higher plants, despite various attempts to do so
in this laboratory. Nev^ertheless, biochemical studies on various species
of plants strongly support the view that the synthesis of sucrose does
proceed by chemical reactions in which glucose or fructose phosphate
esters, or both, serve as substrates, although the mechanism is probably
not identical with that of the bacterial enzyme system. Certain studies
on sugar cane leaves suggest that, in the higher plants, fructose di-
phosphate takes part in the synthetic reaction (6). It is of funda-
mental importance that the experimental evidence now available
shows that, for the synthesis of sucrose from glucose and fructose in the
plant, aerobic metabolism is indispensable. Possibly aerobic oxida-
tions are essential to the phosphorylation of one of the substrates in-
volved in the synthesis of the sucrose. The question is complicated by
the observation that various substrates other than glucose and fructose

70

BIOCHEMISTRY OF PLANTS

may result in sucrose formation by plant tissues. For example, in
experiments on barley shoots by infiltration procedures, galactose could
be utilized for this purpose, as well as various other carbohydrates or
related compounds. In the leaves of the sugar cane, as studied by
Hartt, the oxidative system involved was not inhibited by cyanide,
although in some experiments on other plant tissues in tliis laboratory
the synthesis was cyanide sensitive. While the mechanisms of sucrose
synthesis operating in the higher plant are by no means sufficiently
elucidated as yet, there is reason for an optimistic view that further
developments along the general lines of attack already pursued will
eventuafiy lead to a satisfactory biochemical solution of this important
problem of plant metabolism and plant nutrition.

Closely related to the investigations just outlined is the long-
standing question of the biochemical nature of the interconversion of
starch and sucrose in the plant. As an illustration, the well-known
sweetening of potatoes at low^ temperatures may be cited. In this
process, starch is converted to sucrose. This conversion is also an
aerobic process and is inhibited by cyanide and some other respiratory
poisons. A mixture of hexose-6-phosphates has been i.solated from
potato juice, and also phosphatases capable of hydrolyzing these com-
pounds. A tentative scheme to explain the conversion of starch to
sucrose has been advanced on the basis of reactions for which hexose
phosphate esters are requisite; and, according to the explanation
offered, both glucose-1 -phosphate ester and fructose diphosphate are
essential (11). In the earlier experiments in Hawaii on sucrose
synthesis by sugar cane leaves, fructose diphosphate was likewise re-
garded as an essential substrate. On the other hand, in the bacterial
enzyme synthesis referred to above, only glucose-1 -phosphate and
fructose could be converted to sucrose.

The great problem of cellulose synthesis remains without a
biochemical explanation. Whether this synthesis can take place only
from activities of the organized protoplasm, and whether phosphoro-
lytic processes are involved, are in the realm of speculation at the
present time.

The point to be emphasized by the foregoing remarks is that
biochemical research, by contributing to the basic knowledge of carbo-
hydrate transformations, can influence profoundly the study of plant
nutrition and physiology. There is, of course, open for further research

71

D. R. HOAGLAND

the immense problem of the origin of polysaccharides other than cellu-
lose and starch, which comprise a large fraction of many tissues of
higher plants, such as hexosans, galactans, pentosans, and related com-
pounds. There is need for more information about the synthesis of the
important pectin compounds.

As a brief digression from the main theme, it is of general bio-
chemical interest to refer to the specificity of the enzyme system from
Pseudomonas saccharophila which catalyzes the synthesis of sucrose.
This enzyme system was not restricted in its catalytic potentiality to
the synthesis of sucrose. It was also efTective in bringing about the
synthesis of two new disaccharides. One was synthesized in vitro
from glucose- 1 -phosphate and /-sorbose, the other from the glucose
ester and a ketoxylose (3). The versatility of enzyme systems of this
type was thereby demonstrated.

To the student of practical plant nutrition interested in the
application of fertilizers to soil, the utilization of simple nitrogen com-
pounds by crop plants is always a topic of dominant interest. This
interest is accentuated today because of the enormous expansion of
industries for the fixation of atmospheric nitrogen. Greatly increased
quantities of fi.xed nitrogen could be made available after the war for
agriculture. Field and pot experiments on the effects of nitrogen
fertilizers on crop growth are of course legion, but this type of investi-
gation discloses little of the biochemical processes by which the nitrate
or ammonia absorbed by the plant is elaborated into organic com-
pounds such as the amino acids.

Fortunately, this subject has attracted the efTorts of a number
of able investigators whose primary concern has been that of biochem-
istry; for reviews on the history of research on nitrogen metabolism of
plants and recent trends, the monograph by Chibnall (2) and re-
ports by Vickery ^< a/. (15-18) may be consulted. In the considera-
tion of this aspect of plant biochemistry, it is apparent once more that
guidance in the interpretation of data and in the design of experiments
is greatly influenced by previous research concerned with the bio-
chemistiy of muscle and other animal tissues.

The importance of organic acids and their cycles of metabolism,
including the tricarboxylic acid cycle, have been stressed by some
investigators. Prominent among the organic constituents of plants are
malic and citric acids. Oxalic acid also occurs very frequently, and

72

BIOCHEMISTRY OF PLANTS

succinic acid has been established as a component of the tissues of
several plant species whose content of organic acids has been examined.
Frequently all the organic acid content of the plant is not accounted
for, and unknown organic acids require identification and quantitative
estimation. But it seems that all the organic acids postulated as com-
ponents of organic acid cycles may be present in plant tissues.

A general theory of protein metabolism based on experiments
with seedlings and detached leaves has been evolved which assigns
significant roles to the amides, asparagine and glutamine, and to
various keto acids. As already noted, the mechanisms postulated draw
heavily on the explanations advanced to account for transformations
of nitrogen compounds in animal tissues; but caution is needed in
applying mechanisms based on the study of animal tissues to plant
processes, without adequate confirmatory evidence. Efforts have been
made, however, to integrate available data on changes in the organic
composition of excised plant leaves under experimental conditions, as
well as data on the changes that occur in the intact growing plant, into
schemes of protein synthesis and breakdown correlated with catalytic
cycles. The possibilities of applying to this field of study in the plant
the new tool of iso topic nitrogen have been opened by Vickery and his
collaborators in a preliminary experiment with the tobacco plant,
following the well-knowm research of the Schoenheimer group on animal
metabolism.

Another aspect of the problem ol nitrogen metabolism is con-
cerned with the symbiotic fixation of nitrogen by leguminous plants.
Much more will have to be learned about protein metabolism before
the biochemical reactions of the nodule organism can be properly
understood. Some progress has been made — compare the review by
Wilson (19) — but the theories and evaluation of evidence now
available are apparently subject to controversy. The extraordinary
practical importance of nitrogen fixation and its scientific interest in-
vites further efforts in research by biochemists.

The brilliant study of oxidation systems in living organisms rests
primarily on the experiments and insight of those who have been
concerned with muscle and othci animal tissues, or with yeast. No
comparable achievements by investigators of highrr pl.-mis come to
mind. Some investigators of higher plants, howexcr, h;i\e sought to
apply fundamental knowledge gained by studies on other organisms

73

D. R. HOAGLAND

to the study of plant respiration. Goddard, for example, considered
the role of the cytochrome system (4). It appears that this system
is in fact active in some plant tissues, but not in all. Thus, cytochrome
oxidase activity w^as demonstrated in the embryos and roots of certain
species of plants, also in immature, but not in adult, leaves. A report
is made of the successful isolation of cytochrome G from wheat germ.
Wheat cytochrome G has the same absorption spectrum as heart cyto-
chrome G. Its reduced form is oxidized by heart or wheat cytochrome
oxidase. Succinic dehydrogenase was found to be present in wheat
germ in small amounts.

W. O. James and his associates (10) have undertaken a series
of investigations on the enzyme systems extracted from the sap of the
barley plant. Here, an ascorbic acid system appeared to be active in
normal aerobic respiration. This oxidation system is thought to be
characteristic of higher plants. The degradation of sugars, however,
seems to proceed by way of well-known reactions of phosphorylation
and hydrogen transfer, as described for other kinds of tissue. Various
plant storage tissues have been selected by a considerable number of
workers as suitable material for the study of respiratory processes.
The effects of respiratory inhibitors on these and other plant tissues
have also received much attention.

A comprehensive survey of the literature of plant respiration
from the point of view of modern concepts is greatly to be desired, but
these examples may perhaps serve to illustrate the point that enzyme
chemists can find broad opportunities in the higher plants. This ma-
terial provides an important field of research in which the number of
workers is inadequate to cope effectively with the many and formidable
problems presented. It is reasonable to suppose that a sufficiently in-
tensive effort directed at plant research might be calculated to advance
the subject of respiratory systems in general, as well as supply much
needed basic knowledge for plant nutrition and all its ramifications in
agriculture.

The growth of plants in soil and its dependence on the absorp-
tion of inorganic salts or their ions by root cells, with its implications for
soil and plant interrelations and for fertilizer practices, brings into the
foreground the phenomenon of solute absorption and translocation by
living cells — compare the review by Hoagland (9). While these
phenomena are of general significance to the study of physiological

74

BIOCHEMISTRY OF PLANTS

processes in all living organisms, they possess a peculiar importance in
the consideration of the nutrition of higher plants for the reasons already
suggested. At one time, the absorption of salts by plant cells would
usually not have seemed to belong to the domain of biochemistry. The
intake of solutes was regarded as a passive process, in which the per-
meability of protoplasmic membranes was chiefly stressed. It is now
well recognized that solutes may move into plant cells against concen-
tration or activity gradients at the expense of metabolic energy. This
kind of absorption of salts by living cells of the root, often referred to as
salt accumulation, is dependent on aerobic respiration. The accumula-
tion process is inliibited by many respiratory poisons, such as cyanide
and iodoacetate, not only in the initial absorption of salts or ions by the
root, but also in their polarized movements into the plant's upward
conducting system, and their subsequent accumulation in the cells of
the leaf or reproductive organs. There are various theories of the
mechanisms by which solutes move through the living conducting
system of the plant. It is agreed that simple diflTusion cannot explain
the movement, and the conclusion cannot be escaped that at some point
solutes move against gradients or are accelerated in their movement
through coupling with some energy-yielding process.

Concurrently with the accumulation of some ions by plant cells,
various metabolic processes are stimulated, such as synthesis of organic
acids, or proteins, and oxidation of sugars. In fact, the biochemistry of
salt absorption by plant cells, in its various aspects, appears to offer
a profitable branch of research in the plant field. Researches on
storage tissues of plants provide eloquent testimony of the value of
studying biochemical transformations as part of an investigation of salt
accumulation — compare the review by Steward (14).

In the development of studies on salt absorption or movement,
the tool of radioactive isotopes may become of great value. Thus
certain metabolic reactions may be related to the movement of a se-
lected radioactive ion, which can be detected with extreme sensitivity
of measurement. So far, the use of radioactive isotopes in research on
higher plants has been limited in the main to simple tracer studies
designed to obtain information on rates or direction of movement and
to identify tissues through which translocation or accumulation takes
place. A much wider field for the application of isotopes, stable and
radioactive, awaits development, and with it may come knowledge ol

75

D. R. HOAGLAND

the biochemical aspects of salt absorption and particularly of the
energy-yielding reactions which are inextricably bound up with this
process.

While it is generally accepted now that the accumulation of
salt by plant cells is in some way closely linked with metabolism, and
particularly with aerobic respiration, the incompleteness of this knowl-
edge should be clearly recognized. Even, if the steps in the particular
respiratory cycles, and the active enzymes, coenzymes, and activators
participating, should be identified to the extent that they sometimes
have been in biochemical studies, the mechanisms by which metabolic
reactions are coupled to the movement of solutes into the living plant
cell, or to its polarized translocation from one tissue to another, would
still be obscure. These are questions that have not been answered for
any living cells, and the plant organisms may offer an especially favor-
able, system for further research on solute movement.

It is hoped that these remarks may serve to call attention to
the abundant opportunities for fundamental biochemical research on
higher plants, including the great groups of plants of economic impor-
tance. There is a place in this field for more workers of the kind who
have done so splendidly in advancing biochemistry, particularly in its
relation to animal nutrition, medicine, and the metabolism of micro-
organisms.

References

(1) Bean, R. S., Ph.D. thesis, University of California, 1943.

(2) Chibnall, A. C, Protein Metabolism in the Plant. Yale Univ. Press,
New Haven, 1939.

(3) Doudoroff, M., Hassid, W. Z., and Barker, H. A., Science, 100, 315
(1944).

(4) Goddard, D. R., Am. J. Botany, 31, 270 (1944).

(5) Hanes, C. S., Proc. Roy. Soc. London B128, 1421 (1940); B129, 174
(1940).

(6) Hartt, C. E., Hawaiian Planters' Record, 48, 31 (1944).

(7) Hassid, W. Z., Doudoroff, M., and Barker, H. A., J. Am. Chem. Soc, 66,
1416 (1944).

(8) Haworth, W. N., Peat, S., and Bourne, E. J., Mature, 154, 236 (1944).

(9) Hoagland, D. R., Inorganic Nutrition of Plants. Chronica Botanica,
Waltham, 1944.

76

BIOCHEMISTRY OF PLANTS

(10) James, VV. O., Heard, C. R. C, and James, G. M., New Phytologist, 43,
62 (1944).

(11) McCready, R. M., Ph.D. thesis. University of California, 1944.

(12) Skoog, F., Am. J. Botany, 27, 939 (1940).

(13) Spoehr, H., Plant Physiol., 17, 397 (1942).

(14) Steward, F. C, Trans. Faraday Soc, 33, 1006 (1937).

(15) Vickery, H. B., Leavenworth, C. S., and VVakeman, A. J., J. litol.
Chem., 125, 527 (1938).

(16) Vickery, H. B., Leavenworth, C. S., and Wakeman, A. J., Conn.
Agr. Expt. Sta. Bull., No. 422 (1940).

(17) Vickery, H. B., and Pucher, G. W., J. Biol. Chem., 128, 703
(1939).

(18) Vickery, H. B., Pucher, G. W., Schoenheimer, R., and Rittenberg,
D., J. Biol. Chem., 135, 531 (1940).

(19) Wilson, P. W., The Biochemistry of Symbiotic Nitrogen Fixation. Univ.
Wisconsin Press, Madison, 1940.

77

BIOLOGICAL SIGNIKIGANCE
OF VITAMINS

C. A. ELVEHJEM, professor of biochemistry, college of agri-
culture, UNIVERSITY OF WISCONSIN; WILLARD GIBBS MEDALIST

F

PREVIOUS to the 20th century, thousands, and more likely
millions, of people suffered and died because of a lack of
scientific knowledge about vitamins or of an insufficient supply of
foods rich in these essential nutrients. During the first three decades
of this century, about a dozen vitamins not only have been identified,
isolated, and synthesized, but manufacturing methods have been
perfected to the point at which some vitamins can be supplied at a
cost relatively lower than the cost of calories and proteins. Most nutri-
tionists agree that there are more vitamins to be isolated and better
methods of synthesis to be developed, but I wonder how many have
given thought to the possibility that the uncontrolled production of
synthetic nutrients may lead to sufficient economic disturbances in
agricultural production to aflfect the health of the people of the world
adversely.

It is unwise, especially for a biochemist, to make any predictions
of future developments. The field of vitamins, however, has now de-
veloped to the point at which it is possible to look ahead in light of past
experiences.

The early workers on vitamins had a point of view or philoso-
phy quite different from that held by present investigators. Many of
the pioneers were motivated by the single purpose of alleviating human

79

C. A. ELVEHJEM

suffering. When liver was given to relieve night blindness and fresh
vegetables were used to cure scurvy, the practitioner knew nothing
about vitamins — he was interested in healing the patient. Eijkman,
I am sure, carried a mental picture of the severe cases of beriberi which
he encountered in Java during all his attempts to relate this disease to a
specific essential nutrient. R. R. Williams referred to his early con-
tact with the disease in his Willard Gibbs award address as follows:
"In short, beri-beri was a principal topic of conversation in scientific
and medical circles in Manila during those early years of my enlistment
with Vedder in the Philippines." When Goldberger was called upon
in 1914 to undertake studies on the cause of pellagra, he knew very
little about the disease, but on December 13, 1915, he wrote as follows
to Dr. Milton Rosenau: "I can hardly describe the feeling that I
experience as I go through our wards at the asylum and see the poor
insane women who a year ago had pellagra but who this year are per-
fectly well — so far as pellagra is concerned."

Regardless of the satisfaction experienced by the individual
workers, these phenomenal results did not captivate the interest of
administrators of research funds. More support was given for studies
on animal nutrition, since a premium was placed on production. Few
recognized that the development of strong human bodies would also
pay dividends. Perhaps we can now explain this difference in reaction.
The animal husbandrymen took great interest in judging and selecting
fine stock. If better nutrition produced better stock they were inter-
ested. On the other hand, medical students have always been given
sick people to study rather than the ultrahealthy. For example, Sir
Robert McCarrison of England went to India to study disease but his
most important contributions originated because he was impressed by
the perfect physique of the Hunza race. At present many are interested
in expanding our conception of the relation of nutrition to optimum
health, but we still are not too certain about the procedure; some talk
about extra quantities of vitamins, others advocate physical training.

It was not surprising, therefore, that between 1910 and 1920
the following laboratory findings continued to attract widespread
interest. (1) Calves maintained on diets balanced according to the
recognized standards failed to grow on rations made entirely from
products of the wheat plant but thrived on rations made from products
of the corn plant. (2) Rats placed on purified diets developed nor-

8o

BIOLOGICAL SIGNIFICANCE OF VITAMINS

mally when the fats of the diet consisted largely of butterfat but failed
when certain vegetable oils devoid of vitamin A were used; this ob-
servation was especially significant at a time when Danish children
given skim milk plus vegetable fats were developing the same symptoms
as those observed in rats. (3) Chicks grew normally if allowed access
to sunlight shortly after hatching, but developed severe leg weakness
when hatched early in the spring; at that lime attempts were being
made to start the chicks during the winter months so tliat broilers
would be available when the demand was heavy. When the de-
ficiency agent was found to be vitamin D, not only was the poultry
industry saved but a program was initiated which led to the eradication
of human rickets.

By 1925, many laboratories were undertaking systematic
nutritional studies, and further attempts were being made to produce
specific beriberi in rats and chicks, vitamin A deficiency in rats and
dogs, vitamin C deficiency in guinea pigs and monkeys, vitamin D
deficiency in chicks, dogs and rats, and pellagra in dogs. Rough assay
procedures were developed; foods richest in each of these vitamins
were designated "protective" foods. Theoretically, this knowledge was
all that was needed to prevent the deficiency disease resulting from the
lack of each respective vitamin. Carlson, a few years ago, stated that
we had sufficient knowledge to prevent all beriberi in the world long
before vitamin Bi was synthesized, a statement which is true in a limited
sense. Thus, Chamberlain and Vedder reduced the incidence of beri-
beri in the Philippine Scouts by issuing unpolished rice, but the prob-
lems of producing high-quality unpolished rice and of educating
people to use this type of rice product as part of their diet are still with us.

All these studies were most intriguing to ever-increasing numbers
of research workers. The bars were now down to the progress of nutri-
tion research and the studies gained momentum each year.

Some wanted to know what happened during each deficiency
and how the vitamin functioned in producing normal animals. One
of the first attempts involved histological studies of tissues from vitamin-
deficient animals; and it was soon recognized that the absence of a
minute trace of a nutrient led to extensive structural changes in certain
tissues. Even greater progress was made when the function of vita-
mins was related to the dynamics of the living cell. In 1921, Seidell
stated that, aside from a possible significant diff'crcnce in the degree of

8l

C. A. ELVEHJEM

dialyzability, there are no grounds for not classifying vitamins with
enzymes. Prior to this, Harden and Young, in 1906, emphasized the
importance of organic dialyzable substances in yeast fermentation and
called these substances coenzymes. In 1918, Meyerhof found the co-
enzymes of yeast to be present in a number of animal tissues; but
animal workers were too busy compounding rations to pay any atten-
tion to this finding. R. J. Williams, in 1919, concluded that the sub-
stance or substances which stimulate the growth of yeast are identical
with the substance or substances which in animal nutrition prevent
beriberi or polyneuritis.

Then, in 1932, Warburg and Christian found that the so-called
yellow enzyme which they had shown to be active in a reconstructed
oxidation system contained a derivative of riboflavin, the second mem-
ber of the B complex, as the prosthetic group, A short time later, it
was found that the coenzyme used in the same system and prepared
from red blood cells contained nicotinic acid. In 1937, nicotinic acid
was shown to be the antipellagra factor and thus became identified with
the third member of the B complex. In 1932, Auhagen split carboxyl-
ase, the enzyme necessary for the metabolism of pyruvic acid, into a
protein component and a thermostable part called cocarboxylase.
Soon cocarboxylase was identified as the pyrophosphoric acid ester of
vitamin Bj.

The fourth decade of the 20th century will undoubtedly be
recognized as the period of greatest advance in our knowledge of the
mechanism of action of the vitamins. We must recognize that much is
still unknown and that some of the most difficult problems lie ahead.
However, the results so far obtained have had a much broader in-
fluence— they have given new impetus to the study of enzyme chemis-
try. R. R. Williams states: "These enzyme molecules are too vast
and complex for the chemist to decipher completely today, but we can
now say that the prosthetic group or business ends of these molecules
are in many instances what we earlier came to call vitamins. These
vitamins are, therefore, the bits, the working ends, of the keys which
unlock stores of vital energy from glucose and other foods." The
enzyme approach has emphasized the close relationship of all types of
life. Vitamins have turned out to be growth factors and metabolic
regulators for plants, bacteria, protozoa and yeast, as well as for
animals. See W. H, Peterson, Biol. Symposia, 5, 31 (1941).

82

BIOLOGICAL SIGNIFICANCE OF VITAMINS

Others wanted to know what a vitamin looked hke. The first
crystalHne material to be isolated from a natural concentrate having
vitamin activity was probably nicotinic acid. It was obtained between
1912 and 1914 from rice bran and yeast; but unfortunately its bio-
logical activity was tested for antineuritic activity rather than for anti-
pellagra activity. The successful establishment of the chemical con-
stitution of several of the vitamins depended upon an enormous amount
of work and true chemical ability. The first of the vitamins to be given
serious chemical consideration was undoubtedly vitamin D. In 1925,
Steenbock and Black, and Hess and co-workers showed that crude
cholesterol could be activated by ultraviolet light, and the following
year ergosterol was recognized as the actual provitamin. These ob-
servations stimulated the interest of organic chemists in the structure of
sterols, a problem that had received only sporadic attention. A little
later, the chemical basis for the relationship between carotene and
vitamin A was established. It is interesting to note that, although the
structure of vitamins A and D received early attention, these are the
only well-known vitamins still not available in synthetic form. Vitamins
C and Bi were the first to be made synthetically, but only about ten
years ago. During the past decade, methods for the synthesis of ten
different vitamins have been perfected.

I have merely recorded the final results without paying tribute
to the individual workers for their years of study of the details of
chemical structure. It is true that some of the work was stimulated by
commercial interest, but in many cases the individual workers were
rewarded only by the satisfaction obtained from the successful proof
of structure of "their" vitamin. That chemical industry did become
interested in the production of vitamins was indeed fortunate. The
availability of each new vitamin facilitated progress on the remaining
vitamins.

As far as I am aware, no one has formally expressed the grati-
tude of laboratory workers for the large quantities of vitamins supplied
gratis by industry for experimental purposes. Although it is true that
many papers carry a footnote indicating indebtedness to a particular
firm "for a generous supply of crystalline vitamins," such an acknowl-
edgment is so common today that it is often taken for granted. I
have no way of estimating the total expenditure involved, for the value
of these gifts cannot be calculated merely by multiplying the number

83

C. A. ELVEIIJEM

of pounds supplied by the current price. The crystalhne material was
most valuable to the investigator \vhen the supply was still in limited
production. For example, our work on the newer members of the B
complex with the chick was directly dependent upon our ability to
obtain adequate supplies of pure biotin. Currently, everyone is inter-
ested in feeding his animals purified rations containing only the syn-
thetic vitamins; and the vitamin requirements become rather large
when dogs, monkeys, pigs, and even human subjects are used. Many
of us today would be willing to pay a fancy price for even a few milli-
grams of pure folic acid. If work on the chemistry of this and related
compounds had not been limited by the war, sufficient quantities of it
would undoubtedly now be available for experimental purposes.

As the methods of synthesis improved and the demand for the
compound increased, very substantial decreases in the wholesale prices
of most of the vitamins were made. The cost of riboflavin has decreased
from $17.50 per gram in April, 1938, to 30 cents per gram in gram lots
in October, 1944. In January, 1934, vitamin G cost $213 per ounce;
today, one ounce may be purchased for 95 cents.

The reduced prices made these vitamins available for many
purposes other than for the manufacture of elixirs, tablets, and cap-
sules: synthetic ascorbic acid is added to the lemon powder used in
army rations; B vitamins are added to flour, bread, and corn grits;
and several of the vitamins were supplied to other countries through
"Lend-Lease." In 1944, production of certain of the individual vita-
mins ranged from 100,000 to 1,000,000 pounds.

Vitamins are no longer limited to the laboratory and the doctor's
office — they are now part of big business: more extensive use of vita-
mins means greater dividends to the stockholders of many industries;
therefore, large advertising campaigns have been instituted; and as
profits increase, more funds become available for research. A few
years ago some of us were highly pleased if we received $500 to sup-
port a favorite project. Today, yearly grants as high as $50,000 are
made for nutrition research — a magnificent start. We have the inter-
est of the public and the support of industry, and we should have many
well-trained and energetic investigators in the postwar period.

But what of the future? First we must realize that many of the
workers will be interested in research merely for the sake of research.
Many will have had little contact with extensive deficiency diseases.

84

BIOLOGICAL SIGNIFICANCE OF VITAMINS

This should not be disturbing if the findings are properly applied.
The practical problems will involve economics as well as chemistry
and physiology. In the field of medicine, the doctor will continue to
prescribe vitamins for deficiencies which are clearly diagnosed, and in
some cases he will try vitamins to determine if any beneficial effects
can be obtained. If the patient recognizes some benefit, the use of
vitamin supplements will be continued for some time; if no benefits
are recognized, the box of capsules will probably remain on the shelf
of the medicine cabinet. A certain group of people will buy vitamin
preparations on their own initiative but few will take capsules con-
tinually for any length of time.

Extension of the use of vitamins will probably ha\c to come by
way of the addition of vitamins to widely used foods. The supply of
niacin did not become critical immediately after the discovery of its
role in curing pellagra; but the supply was critical within a few weeks
after the introduction of flour enrichment. If my calculations are
correct, 1,000,000 pounds of niacin is almost sufficient to su[)ply the
minimum requirement of all the people in the United States for one
year, an amount which the annual production is now reaching. What
will happen now that war is over and synthetic niacin will be in direct
competition with niacin in meat, and synthetic ascorbic acid will be in
competition with vitamin C in oranges, tomatoes, etc.? Will industry
be willing to control the pioduction of synthetic vitamins in relation
to the true demand for these products? I would be greatly disturbed
by an extensive advertising campaign advocating greater use of syn-
thetic vitamin C by the public at a time when the orange crop is rotting
in orchards. On the other hand, if the synthetic vitamins are used to
supplement rather than replace our food supply, we can plan for this
country — yes, even the world — a continuous supply of nutrients which
will be. little affected by crop failures. I hope such a plan can be
made and executed before the problem becomes so acute that govern-
ment may have to step in. At present, the use of vitamins is promoted
largely among people who are rather adequately fed. What results we
could expect from the proper use of vitamins in the contiol of famines
in India ! It is true that vitamins cannot replace other food nutrients
but certain vitamins at least increase the elliriency of utilization of the
total nutrients and may also help in the synthesis of other vitanuns m
the intestinal tract.

85

C. A. ELVEHJEM

I believe there is another important angle which applies not
only to the use of synthetic vitamins but also to synthetic amino acids,
which undoubtedly will be produced in the postwar period. Since, in
general, the synthetics have no taste appeal, and since mankind will
continue to consume food for reasons other than that of mere nutrition,
it may be more important to use a larger part of the synthetics in animal
feeding. As we learn more about nutrition, we are finding that many
of the more expensive animal feeds can be replaced by cheaper sub-
stitutes. For example, riboflavin can be used in poultry feeding with-
out relying upon more expensive milk products. Thus, the cost of
animal production can be reduced to such an extent that animal
products can be used more widely for human consumption.

The necessity of fortifying certain human foods may continue
for some time. Although new types of food fortifications may be intro-
duced, we must recognize that any enrichment program is not neces-
sarily permanent, and we should be willing to discontinue any one
program when and if scientific evidence indicates that it is no longer
necessary. It will be the duty of nutritionists to give careful considera-
tion to these programs; proper decisions can be made only if we have
extensive knowledge of the vitamin content of all foods. Food indus-
tries have generously supported such programs but the work has cer-
tainly not reached completion. Plans should be made to set aside
funds which are now easily obtainable so that work of this kind can be
carried out when personnel become available. There are two im-
portant lines of approach: one deals with the production of food
products high in vitamins and can be accomplished by improved
breeding, cultivation, and fertilization; the other deals with improved
methods of handling the food products after harvesting and slaughtering.

Because fundamental research must continue in the field of
vitamins, it will be fortunate to have young men interested in pure re-
search. During the past few years, the practical problems have re-
ceived greatest emphasis, but we have now reached the point at which
fundamental research again becomes the limiting factor in further prog-
ress. We must study cellular mechanisms within the body and the
relation of bacterial cells to the vitamins within the intestinal tract.
The relation of vitamins to enzymes has already been discussed.
After all, sturdy bodies are largely dependent upon properly function-
ing enzymes in all the cells of the body. Vitamins are only a small part

86

BIOLOGICAL SIGNIFICANCE OF VITAMINS

of the enzyme molecule and we need to know more about the rest of
the molecule. Perhaps proper exercise with limited amounts of vita-
mins may be more conducive to rigorous enzyme systems than the
consumption of vitamin cocktails while reclining in an easy chair.
Studies on the enzyme systems will help to relate nutrition to such
important problems as resistance to infection, prevention of cancer, and
resistance to the process of aging. Biochemists have learned how to
disorganize cells into parts, but greater integration of the parts into a
whole must be attained.

More attention also must be paid to the bacteria in the intes-
tinal tract. Bacteriologists have studied the bacteria in soils, in milk,
in foods, and in disease, but have largely disregarded the bacteria of our
intestines. There are still nutritional disturbances which must be re-
lated indirectly to the changes in the digestive tract.- Very recently,
a most interesting report was presented on the nutritional status of
about 800 individuals living in Newfoundland: certain symptoms
observed were ascribed to riboflavin and niacin deficiencies, and yet a
rough estimate indicated that the intake of these vitamins was not
seriously inadequate. I believe some of the changes, at least, are due to
a lack of as yet unknown vitamins which were not synthesized in suffi-
cient amounts because of the type of dietary regime. Pellagra has
always been associated with a large consumption of corn. Preliminary
evidence in our laboratory indicates that an extensive corn intake
may adversely affect the synthesis of vitamins in the intestinal tract, an
effect which is overcome by high levels of nicotinic acid. High levels
of protein also have a counteracting effect, which may explain why
milk has been found to have antipellagra activity although it is known
to contain little nicotinic acid.

Intestinal synthesis is not limited to the production of the vita-
mins which are the last to be discovered. Obviously, the degree of
synthesis in the intestine must be less in the case of the older vitamins,
or we would have had more difficulty in producing the deficiency state
of these vitamins. Some time ago, we showed that the fat content of
the diet had a marked effect on the riboflavin requirement of the rat.
Diets which contained dextrin and low levels of ribofla\'in produced a
much more severe riboflavin deficiency when a large jjortion of the
dextrin was isocalorically replaced by fat. Later work has shown that
this effect is direcdy dependent upon a decreased synthesis of riboflavin

S7

C. A. ELVEHJEM

in the presence of fat. There is much evidence to show that fat, carbo-
hydrate, protein, and vitamins are all interrelated in their effect on the
production of both known and unknown vitamins in the digestive tract
of all animals; but the results for one animal cannot be predicted from
the results obtained with another species. Just where the human fits
into the picture is impossible to say. It is encouraging to find that
several groups of workers are engaged in studying this problem on
human subjects, a project which will undoubtedly clear up many of the
difficulties now encountered in attempting to establish quantitative
requirements for each of the vitamins in human subjects. The final
answer can probably be made only when animals are rendered bac-
teria-free and their requirements are studied under these conditions.
Although this will answer the questions from an academic point of
view, for practical purposes we must continue to recognize the inter-
relationship of food and intestinal bacteria. No one need feel that
further work in the field of vitamins will not be productive. Much
human suffering has been alleviated through our knowledge of vita-
mins and we can expect much success in the future if we learn more
about these interesting compounds and if we apply what we learn in a
sensible manner.

Selected References

Addinall, C. R., "Synthesis and production of vitamins," Chem. Eng. News,

22,2174 (1944).
Black, J. D., ed., "Nutrition and food supply: The war and after," Ann. Am.

Acad. Political Social Sci., 225 (1943).
"Enrichment of flour and bread. A history of the movement," Bull. Natl.

Research Council, No. 110 (1944).
Evans, E. A., Jr., ed., Biological Action of the Vitamins. Univ. Chicago Press,

Chicago, 1942.
Major, R. T., "Industrial development of synthetic vitamins," Chem. Eng.

News, 20, 517 (1942).
"Medical survey of nutrition in Newfoundland by a group of investigators,"

Can. Med. Assoc. J., 52, 227 (1945).
Rosenberg, H. R., Chemistry and Physiology of the Vitamins. Interscience, New

York, 1945.
Schultz, T. W., ed., Food for the World. Univ. Chicago Press, Chicago, 1945.
Williams, R. R., and Williams, R. J., "Vitamins in the future," Science,

95,335-344 (1942).

88

SOME ASPECTS OF
VITAMIN RESEARCH

KARL FOLKERS, director of organic and biochemical research,

MERCK & CO., inc.; AMERICAN CHEMICAL SOCIETY AWARD IN PURE
chemistry; CORECIPIENT of the mead JOHNSON AWARD

So little is known about the chemistry oj vitamins — not
a single one has been isolated with absolute certainty —
that I have hesitated to include this subject among the
applications of organic chemistry. The very extensive
contemporary literature on vitamins which takes up
much space in journals devoted to biochemistry^ contains
few chemical facts, and very few that are thoroughly well
established.

T\

^HIS statement was made in 1928 at Cornell University by
George Barger (1) during his lectures on some applications
of organic chemistry to biology and medicine. These lectures were
concerned with hormones, vitamins, chemical constitution and physio-
logical action, chemotherapy, and blue adsorption compounds of
iodine. The number of well-established chemical facts on the chemis-
try of the vitamins developed so enormously during 1928 to 1945
that an entire university course could justifiably be devoted now to the
organic chemical aspects of the vitamins. The companion develop-
ments on the biochemistry of the vitamins and on the application of
the vitamins in clinical medicine might also require a course each for
adequate presentation to students. The industrial production of
vitamins on a ton basis, a subject recently reviewed by Major (30),
is no less amazing in the rapidity and magnitude of the develop-
ment.

Today there are so many excellent books and review articles
on the chemistry of the vitamins available, that no effort will be made
in the following sections to cover any topic completely. Instead, a

89

KARL FOLKERS

few observations, facts, and results which may have unique interest
have been selected for comment.

On the Discovery of Vitamins

The discovery of the "major vitamins"* has been based upon
observations which related the syndrome of a human disease to con-
stituents of natural materials used in nutrition. The discovery of the
"lesser vitamins"* has been based upon observations which related
biological reactions generally produced experimentally with animals
or microorganisms to constituents of natural materials. Although the
"lesser vitamins" are at present not known to correspond to any histori-
cally recognized human disease, they probably are essential for the
human being. They may be considered "lesser vitamins" today be-
cause their absence in human diets is less frequent statistically or because
the signs of their absence are not yet fully recognized, as was the case
for riboflavin deficiency until 1938, when Sebrell and Butler (43)
characterized ariboflavinosis. Oden, Oden, and Sebrell (37) con-
cluded a little later that ariboflavinosis is "a common dietary-deficiency
disease in the southern United States."

It seems not unlikely that certain of the "lesser vitamins" will
ascend to the class of "major vitamins" after further clinical research.
Some of this future clinical research might lie in the borderline fields
between biochemistry and psychology, according to R. J. Williams
(63). His observations on "personality diff"erences" among animals
in nutrition experiments, and the fact that hallucinations and mental
symptoms of pellagra are known to be eliminated by administration
of nicotinic acid, helped stimulate this interesting thought. Certainly,
if other vitamins were found to benefit mental disease or psychological
disturbances, the rank of importance of these vitamins would be
elevated. R. R. Williams (69) believed it was probable that any

* R. R. Williams, in his stimulating address on the occasion of the presenta-
tion of the Chandler Medal in 1942, defined the "major vitamins" as thiamin,
riboflavin, nicotinic acid, and vitamins A, D, and C. Five of these vitamins are
related to ancient and widespread diseases. Ariboflavinosis, which is cured by
riboflavin, had been confused with and masked by pellagra and was not recognized
per se until recently. The "lesser vitamins" include choline, vitamin Be, pantothenic
acid, biotin, inositol, etc. See reference (69).

90

VITAMIN RESEARCH

vitamins yet to be discovered are destined to have lesser nutritional
significance for human welfare and that the vitamins which are re-
quired to check the nutritional plagues of mankind have already been
discovered and produced. Nevertheless, he recognized that there
may be exceptions, particularly in the case of obscure diseases. Un-
doubtedly, chemists and nutritionists must cooperate scientifically
for many years on the problems of the discovery of new vitamins.

On the Isolation of Vitamins

The isolation of a vitamin from the natural material in which it
exists is essentially a chemical problem of the same nature as the older
problems on the isolation of an alkaloid or a glycoside, but with at
least two important dilTerences. One of these differences is that vita-
mins generally occur in quantities amounting to a few parts per million
of the natural material, whereas the common alkaloids and glycosides
are found frequently in quantities amounting to a few parts per
hundred. The greatest difficulties in vitamin isolation might be said to
lie in the region of converting the natural materials with a few parts per
million of the substance to a concentrate containing a few parts of the
substance per hundred. New techniques and new procedures are
frequently devised to surmount the difficulties in making such a puri-
fication. The isolation of trace substances in milligram or gram quanti-
ties requires the processing of hundreds of pounds of the natural ma-
terial. A second important diff"erence in the isolation procedure is
the necessity for countless biological assays throughout the whole
isolation work to show the investigator the location (and loss !) of the
vitamin in the fractionation.

Pioneering researches on the isolation of a vitamin are very costly
and time-consuming. It has been said (69) that the first gram of pure
natural thiamin must have cost an aggregate of several hundred
thousand dollars. Eight years transpired between the first success in
isolating this vitamin m 1926 by Jansen and Donath (21) and the work
of Williams, Waterman, and Keresztesy (70) m 1934, which resulted
in greatly improved yields, so that a sufficient quantity of the pure
vitamin could be made available for its structure determination. It
took five years in Kogl's laboratory at the University of Utrecht in
Holland to work out the pioneering methods which yielded seventy

91

KARL FOLKERS

milligrams of crystalline biotin (26). The difficult and tedious scheme
oi" hactionation involved a three million-fold purification. Kogl (26)
estimated that to produce one gram of their biotin from ordinary
yeast would have required 360 tons of the yeast as starting material.
Although he found egg yolk to contain ten times as much biotin as
yeast, the number of fresh eggs required for the production of one gram
of biotin would have cost about $165,000 in 1937.

Special chemical steps or new techniques often have to be de-
vised or applied to the isolation process before the vitamin can be ob-
tained in sufficient amounts for the complete elucidation of its structure.
The improved yields (70) in the process for the isolation of thiamin
depended upon the elution of the vitamin from fuller's earth with
quinine acid sulfate instead of barium hydroxide and the introduction
of a benzoylation step for purification. Probably one of the most im-
portant factors contributing to the success of the isolation of additional
quantities of crystalline biotin was the application of the chromato-
graphic adsorption technique by du Vigneaud, Hofmann, Melville,
and Gyorgy (59) to concentrates of the vitamin which had been ob-
tained from beef liver (13). The concentrate contained 0.1% biotin,
and, after esterification, the material was chromatographed twice over
aluminum oxide. After the final sublimation and crystallization steps,
pure crystalline biotin methyl ester was obtained in 38% yield based
on the amount of the vitamin in the concentrate.

On the Structure Determination of Vitamins

Ordinarily, it is desirable to isolate any vmknown natural prod-
uct in a state of complete purity before the carrying out of chemical
reactions for establishment of molecular formula, identification of
functional groups, and finally the determination of structure. Special
attention to the question of purity is often justified, because natural
products are frequently isolated which are extremely difficult to sepa-
rate from final impurities of unknown but of allied properties.

There are, however, exceptions to the purity requirement.
The isolation of calcium pantothenate in pure form was found by
Williams' group (67) to involve extraordinary difficulties, and it was
necessary to conduct the structure studies with a highly purified con-
centrate estimated to be about 90% pure (34,67). The establishment

92

VITAMIN RESEARCH

of the structure of pantothenic acid under tliese conditions was accom-
phshed by a rather unique series of developments. ^-Alanine was
first reported by WiUiams, Weinstock, and Mitchell (61,68) to be
formed from this calcium salt in acid or alkaline medium. The other
hydrolytic product was found (34) to l)e an a-hydroxy acid capable
of spontaneous transformation to a lactone. The crude lactone
fraction was recombined (66,72) with /3-alanine to give material which
possessed the physiological activity of pantothenic acid and, of course,
actually was this acid. This condensation reaction constituted re-
synthesis, and the results supported the earlier observation (34) that
/3-alanine was combined as an amide through its /3-amino group, as
in structure I. At this stage of the investigation it was clear that it

R— CONHCH2CH.2CO..H

(I)

would be unnecessary to isolate pure pantothenic acid if the hydroxy
acid fragment or its lactone could be isolated in pure form. The
structure determination of the hydroxy acid would give also the struc-
ture of pantothenic acid. Further research (34,53) showed how con-
centrates containing only 3 to 40% of pantothenic acid could be
purified and treated so as to yield the pure lactone. Once having
obtained the pure lactone, Stiller, Keresztesy, and Finkelstein (53)
applied degradation reactions which showed the lactone to be a-
hydroxy-/3,/3-dimethyl-7-butyrolactone (II) :

OH

CH— C=0 OH

(CH3)2G O (CH,)2C-CH-CONHCH2CH2C02H

\ / 1

CH^ CH,OH

(II) (III)

Obviously, pantothenic acid was rt,7-dihydroxy-/3,/3-dimethylbutyryl-

/3'-alanide (III).

Because of the great cost and technical difficulties in the isolation
of new vitamins, it has been necessary on occasion to open the program
of structural research with a series of studies which are designed to
cliaractcrize the functional groups of the substance and which can be

93

KARL FOLKERS

carried out with micro quantities of the substance. For example,
early evidence concerning the non-/3-alanine portion of pantothenic
acid was obtained from the results of a series of new micro procedures
(34). Some of these new micro procedures involved determination
of active hydrogen atoms with deuterium oxide and of hydroxyl groups
with hydriodic acid, selective oxidation with iodic acid, oxidation
equivalent analysis, determination of a- and /3-hydroxy acids, and
estimation of microorganisms in suspension. The use of such tech-
niques represented an unconventional but fruitful approach to the
study of pantothenic acid and often required only one or two milli-
grams of the compound for each determination.

Another type of study has been employed to open a program
of structural research upon a costly vitamin available in only very
limited quantities. This study, involving a series of "inactivation"
experiments, requires only a milligram or less of the substance and
yields information on functional groups and constitution. These
experiments involve adding to the micro sample of the vitamin the
chemical reagent (s) required for a given chemical reaction, such as
nitrosation or hydrolysis, and following with a microbiological assay
to test whether chemical reaction took place as judged by a change
or lack of change of activity. These "inactivation" experiments yield
valuable results, but they must nevertheless be interpreted with con-
siderable caution. They do aid in guiding the exploratory efTorts in
direct chemical studies. An example of such inactivation experiments
may be found in certain biotin studies. Brown and du Vigneaud (2)
described the effect of certain reagents on the activity of biotin. They
obtained this preliminary information with experiments on 1- or 2-cc.
aliquots of solution containing only 12.5 7 of biotin per cc, and the
criterion of reaction was the effect upon yeast growth activity. Such
reagents as 5% hydrogen peroxide solution, aqueous bromine, hydro-
chloric acid, potassium hydroxide, formaldehyde, and nitrous acid
caused "inactivation," indicating that a change in the structure had
undoubtedly been brought about by the reagent. On the other hand,
such reagents as acetic anhydride-sodium hydroxide, ketene, benzoyl
chloride-pyridine, sodium ethoxide-methyl iodide, and ninhydrin
caused no "inactivation," indicating that a change in structure had
not been brought about by these reagents. The results of these and
related experiments indicated to Brown and du Vigneaud that biotin

94

VITAMIN RESEARCH

is inactivated by vigorous treatment with acid or alkali, is not an a-
amino acid, is not destroyed by acylating or alkylating reagents, and
is easily oxidized.

Not all of the well-known vitamins have had their constitution
elucidated by an application and development of micro methods. The
correct structure of a-tocopherol of the vitamin E group was deter-
mined largely by interpretation of the results of three chemical reac-
tions, which were carried out on 2.1, 4.3, and 25 grams of the vitamin
in each case. While repeating some of the isolation work of a-toco-
pherol from cottonseed oil described by Emerson, Emerson, and
Evans (7), the late Dr. E. Fernholz, working in the laboratory adjoin-
ing that of the author, intimated that he desired to accumulate enough
of the a-tocopherol to permit reactions on a gram-scale basis. It was
feasible to isolate a-tocopherol on this scale. Subsequently, Fernholz
described (8) the thermal decomposition of a-tocopherol which yielded
durohydroquinone, and the details (9) of the experiment reveal that
2.1 g. of the vitamin had been heated for six hours at 355°. The
crystalline sublimate yielded 257 mg., or 67%, of durohydroquinone
and a hydrocarbon of the composition Ci8-i9H36- This characteriza-
tion of durohydroquinone shattered the frequently discussed idea that
a-tocopherol is related to the sterols. The absorption spectra and
chemical properties of some synthetic monoethers of durohydro-
quinone, when considered in conjunction with other suggestive evi-
dence, led to the hypothesis that a-tocopherol was derived from
chroman or coumaran. This hypothesis of a heterocyclic ring structure
was justified by the results of an oxidation reaction with chromic acid
(9). When 4.3 g. of a-tocopherol was dissolved in glacial acetic acid
and oxidized with chromic acid, dimethylmaleic anhydride (IV) and

/Cf /CH2

CHa— C \ / ^CHa

II o o=c I

CHa— G / \ C— CieH

33

(IV) (V)

a lactone, C21H40O2, were isolated. Structure V was proposed for
the lactone after a study of its properties and after a study of the
oxidation of the acetate of a-tocophcrol obtained from 25 g. of a-

95

KARL FOLKERS

tocopheryl allophanate. The chromic acid oxidation of this acetate
yielded diacetyl, acetone, an acid C16H32O2 of probable structure VI,
and a ketone CigHseO.

CHg CHj CIH3

i I i

HOOCCH2CH2CH(GH2)3CH(CH2)3GHCH3

(VI)

The interpretation of these degradation products and related
evidence led to the proposal by Fernholz (9) of structure VII for a-

CH3 Cri2
HOr/N^ ^CHs CH3 CH3 CH,

CHs^X Js.^ y-C(Cri2)3CH(Cri2)3Cri(Cri2)3CriCH3

CH3O I.H3

(VII)

tocopherol. Although the related coumaran structure was con-
sidered further by Karrer, Fritzsche, Ringier, and Salomon (22,23)
all the results of the investigations — refer to review paper by Smith
(44) — were soon interpreted in favor of the chroman structure VII,
which was proposed after a study of a few relatively large-scale deg-
radation reactions.

An interesting aspect of the structure studies on thiamin and
biotin was the application of new organic structural reactions. It
had been observed during the isolation work that an attempt to use
sulfurous acid as a preservative against the bacterial decay of extracts
of vitamin Bi from rice polish caused a prompt and complete loss of
vitamin activity. This observation led to the development of a pro-
cedure by Williams, Waterman, Keresztesy, and Buchman (71) for
the quantitative cleavage by sulfite of pure crystalline vitamin Bi at
pa 5 and room temperature into two products of the composition
C6H9N3SO3 and CeHgNSO. There are obvious advantages of this
reaction over the usual oxidative and hydrolytic reactions which ap-
peared to give a miscellany of substances. The reaction is expressed
in terms of the structural formulas by the following equation:

CH3

N=CNH3+C1- C CCH2CH2OH

CH,C C— CH2— N'+

N— CH CI- CH-

+ Na2S03

96

VITAMIN RESEARCH

c:h,

N-=CNH2 c-^,,^cCH2CH.OFI

/ \ /

CH3C CCH2SO3H + N

N— CH CH —

+ 2 NaCl

The new organic structural reaction used in the investigation
of biotin was applied in the belief that the results would lead to a final
selection between two alternative structures for biotin. It was believed
that organic sulfides could be cleaved by the Raney nickel catalyst
according to the equation:

Ni(H)
RSR' ^ J. RH + R'H

This sulfur hydrogenolysis reaction was developed first on several
"model" compounds. It was found by Mozingo, VVolf, Harris, and
Folkers (35) that representative sulfides could be cleaved to their
corresponding sulfur-free products in yields of 65 to 95% on both a
macro and semimicro scale. As described by du Vigneaud, Melville,
Folkers, Wolf, Mozingo, Keresztesy, and Harris (60), the application
of this hydrogenolysis reaction to biotin methyl ester yielded dcsthio-
biotin methyl ester, and the subsequent study of this hydrogenolysis
product was the first of two independent methods which led to the
final proof of the structvu-e of biotin. It was considered originally
(35) that this reaction would be of general value in investigations on
the structures of natural products containing sulfur.

On the Use of Microorganisms in Vitamin Researcli

It is beyond the scope of this article to present all of the interest-
ing aspects of the application of microorganisms to vitamin research.
Williams recently discussed the importance of microorganisms in \-ita-
min research (62) and reviewed microbiological tests (64). Gyorg\'
reviewed further developments in the use of microorganisms in vitamin
research (12). There are, however, certain recent developments in
the chemistry of vitamin Bg which were originally provoked l)\- the
results of microbiological experiments, and recent results on new vita-
mins which seem to merit comment.

Snell, Guirard, and Williams (50) found that assays with
Streptococcus lactis gave values for the pyridoxine content of certain

97

KARL FOLKERS

natural materials which were several hundred to several thousand times
the values obtained by different biological or chemical methods.
Other experiments showed that the factor responsible for the increased
activity was very similar in properties to pyridoxine, and they pro-
visionally named the factor "pseudopyridoxine." Subsequent to these
studies, Williams stated (65) in his review on the water-soluble vitamins
for 1942: "Contrary to the common impression, the chemistry of
vitamin Be is in a highly unsatisfactoiy state. There is no question,
of course, regarding the fundamental chemistry of pyridoxine, that
pyridoxine occurs naturally, or that pyridoxine has vitamin properties.
The serious question is whether the vitamin Be activity of tissues is due
solely to pyridoxine or whether there are other substances (probably
closely related) which serve equally well or better."

After about two years of further research at the University of
Texas on the one or more substances in natural materials tentatively
called "pseudopyridoxine," Snell (45) concluded that one of these
substances was probably an aldehyde and another was probably an
amine. These studies were concerned essentially with the effect of
certain selected chemical treatments on the biological activity of pyri-
doxine for Streptococcus faecalis R, Lactobacillus casei, and Saccharomyces
cerevisiae. After consideration of the functional groups of pyridoxine
which might be involved in the reactions, it was believed that the bio-
logically active aldehyde would have one of three structures, VIII,
IX, or X, and the biologically active amine would have the corre-
sponding structure, XI, XII, or XIII.

CH2OH
HO/\cHO

(VIII)

CH20H

H0/\CH2NH2

CH:

\N^

(XI)

CHO
H0/\CH20H

CH

(IX)

CH2NH2
H0/\CH20H

CH

(XII)

CHO

HO/\cHO

CHsi^j^^
(X)

CH2NH2
H0/\CH2NH2

CHi

(XIII)

Collaborative studies by Harris, Heyl, and Folkers (15) on the
structure and synthesis of the active aldehyde and active amine resulted

98

VITAMIN RESEARCH

in the synthesis of aldehydes VIII and IX and amines XI and XII.
Biological tests on these synthetic compounds by Snell (46) showed that
the biologically active aldehyde was 2-mcthyl-3-hydroxy-4-formyl-5-
hydroxymethylpyridine (IX), and that the biologically active amine
was 2-methyl-3-hydroxy-4-aminomethyl-5-hydroxymethylpyridine (see
XII). The active aldehyde and amine were given the trivial names
"pyridoxal" and "pyridoxaminc," respectively. The microbiological
assays showed that pyridoxal was about 1400 times more active, and
pyridoxamine about 10 times more active, than pyridoxine hydro-
chloride for promoting the growth of L. casei. Pyridoxamine was
about 8000 times more active, pyridoxal was about 5500 times more
active, in promoting the growth of S. Jaecalis R, than was pyridoxine
hydrochloride. The comparative activity of pyridoxal, pyridoxamine,
and pyridoxine for Saccharomyces carlsbergensis was of the same order of
magnitude.

Evidence for the occurrence of pyridoxal and pyridoxamine in
natural extracts was secured by Snell (47) by development of a differen-
tial microbiological assay technique with the three organisms mentioned
above and application of the assay to extracts of natural materials.
Further evidence for the existence of pyridoxal and pyridoxamine in
nature was secured by studying the effect of certain chemical treat-
ments upon the "pyridoxine, pyridoxal, and pyridoxamine fractions"
in comparison with the effect of these treatments upon the synthetic
vitamins.

Vitamin Be was originally considered to be a single pyridine
derivative, pyridoxine. It may now be considered, as a result of
these combined microbiological and organic chemical studies, as a
name which designates a group of vitamins, i. e., the "vitamin Be
group." Pyridoxal and pyridoxamine may occupy a place of equal
or greater importance in this group as compared with that of pyri-
doxine.

In retrospect, it is interesting to note that, in the original
isolation work, Keresztesy and Stevens (25) and Lcpkovsky (29) used
rice bran as the source of their vitamin Be, while Kuhn and Wcndt
(28) and Gyorgy (11) used yeast as their source of the crystalline
vitamin Be (pyridoxine). Snell's microbiological differential assays
showed (47) that a rice-bran concentrate contained far more pyri-
doxine fraction than pyridoxal or pyridoxamine fractions, whereas

99

KARL FOLKERS

yeast and liver extracts contained an excess of the pyridoxamine frac-
tion, and animal assays showed no marked difference in the activity
of the three substances. By assuming that the yeast supply used by
Kuhn and Wendt and by Gyorgy contained an excess of pyridoxamine
also, it is evident that at one or more of the steps in the isolation proc-
ess the pyridoxamine was lost. Williams (65), in commenting on
Gyorgy's communication (11) on isolation, noted that the yield of
active substance in the first few steps of the concentration was only
10 to 30% of the original activity. If the isolation of vitamin Be
from yeast had been guided by the results of microbiological assays with
.S*. Jaecalis R instead of rats, one might predict today that it is quite
probable that pyridoxamine would have been isolated instead of pyri-
doxine and the pyridoxine present would have been lost at some step
of the isolation procedure.

In the field of new vitamins of unknown structure, several sub-
stances are currently of great interest and papers concerning them
are appearing frequently in the literature. Although studies of bio-
logical activities in animals are being made, the use of microorganisms
for tests of biological activities is resulting in the rapid accumulation
of much valuable data on the differentiation of these substances. The
following citations may exemplify the importance of the role of micro-
organisms in the study of these new growth factors.

Snell and Peterson (52) and Hutchings, Bohonos, and Peterson
(19) have described the preparation and some properties of a con-
centrate of a norite eluate factor from liver and yeast which resulted
from a study of the nutrient requirements of L. casei and related lactic
acid bacteria. Mitchell, Snell, and Williams (33) reported on the
preparation of a highly purified nutrilite from spinach which they
designated folic acid and defined as the material responsible for the
growth stimulation of S. lactis R. Crystalline vitamin Bg from liver
was highly active in growth activity for L. casei according to Pfiffner,
Binkley, Bloom, Brown, Bird, Emmett, Hogan, and O'Dell (39).
Stokstad (56) has described some properties of two crystalline prepara-
tions, one from liver and one from yeast, which had somewhat different
activities for promoting the growth of L. casei and S. lactis R. Keresz-
tesy, Rickes, and Stokes (24) have reported the isolation of a different
factor which was highly active for the growth of S. lactis R but rela-
tively inactive for the growth of L. casei. Another new compound

lOO

VITAMIN RFSEARCH

which is active lor the growth of L. casei was drscrihcfl hy llni. hings,
Stokstad, Bohonos, and Slobodkin (20).

The similarity of the biological activities of these several sub-
stances suggests that they also may have a similarity in chemical
structure. Concerning this chemical structure, Mitchell (32) pre-
sented evidence which showed that the ultraviolet absorption spectrum
of folic acid resembled that of xanthopterin (XIV). Addition of large
amounts of thymine (XV) was found to substitute (biological pre-

COH N CO

/'\/\ /\

N G COH HN GCH,

I II I I II

H2NC C CH OG CH

\/\/ \/

N N NH .

(XIV) (XV)

cursor?) for folic acid according to Snell and Mitchell (51). Thus,
chemical relationships to the pterins and pyrimidincs are suggestive
possibilities.

Precise determination of the chemical structure of these several
biological factors is probably necessary before the identity of any two
or more of them, and before exact similarities in chemical structure, can
be established with certainty. It is interesting that much of the micro-
biological characterization of folic acid, vitamin B^, and related factors
is being developed before the elucidation of their chemical structures.
This situation is somewhat the reverse of that for the vitamin Be group
since much of the chemical characterization of these factors was de-
veloped before the clarification of the microbiological aspects.

On the Synthesis of Vitamins

The methods of the organic syntheses of the synthetic vitamins
have been amply covered in appropriate review articles. These
methods exemplify the adaptation and modification of classical organic
laboratory reactions to the synthesis of the desired structure.

The three asymmetric carbon atoms and the two fused five-
membered saturated heterocyclic nuclei of biotin present interesting
stereochemical features to the studies on the synthesis of this vitamin.
There are two racemates which have cis forms of the rings and two

lOI

KARL FOLKERS

racemates which have trans forms of the rings. The cis and trans
relationship of the rings is shown by structures XVI and XVII, re-

o

1

/\

^fH NH

G C

1

0

1

NH NH
\ H H .•

1

CH2 CH(CH2)4C02H

GH2 GH(CH2)4C02H

(XVI)

(XVII)

0

11

NH NH

CH2 CH(CH2)4G02H

(XVIII)

spectively. Biotin is one of the eight stereoisomeric forms, dl-
Biotin and related cis and trans forms were obtained by Harris, Mozingo,
Wolf, Wilson, Arth, and Folkers (16). These other forms were des-
ignated ^/-allobiotin and (//-epiallobiotin. Grussner, Bourquin, and
Schnider (10) have described <//-pseudo-jS-biotin and ^/-iso-^-biotin
in addition to ^/-/S-biotin. Each group may have certain racemates
in common. Further work on the comparison of the properties and
degradation products of a-biotin and j8-biotin is apparently in progress,
according to a i-ecent paper by Kogl and Borg (27).

Studies on synthesis provide interesting information concerning
the specificity of structure or the relationship between the constitution
and vitamin activity. The specificity of biotin has been discussed
in an excellent review article by Melville (31). Recent additions may
be made to the subject of specificity of biotin. The described cis and
trans forms related to ^/-biotin appear to be biologically inactive ac-
cording to Emerson (6), Ott (38), and Stokes and Gunness (54); see

102

VITAMIN RESEARCH

also Grussner et al. (10). An interesting oxygen analogue has been
described by Hofmann (17). It has significant biological activity
according to Pilgrim, Axelrod, Winnick, and Hofmann (40) ; an oxygen
analogue of about the same melting point was also reported as bio-
logically active by Duschinsky, Dolan, Flower, and Rubin (4). Sub-
sequent studies by Hofmann (18) suggest that rf/-oxybiotin (X\'III)
and biotin have identical spatial configurations and differ only in the
nature of one of the hetero atoms.

Biosynthesis of vitamins has received little attention. The
pioneering experiments on the synthesis of alkaloids under so-called
physiological conditions, as described by Robinson (41), Schopf (42),
and Hahn and Schales (14), involved the isolation of the alkaloid
synthesized as the criterion of the success of the experiments. Experi-
ments on biological precursors and on the biosynthesis of vitamins
have involved assays with microorganisms, and the production of
vitamin activity as the criterion of success. Once again, the utility
of microorganisms in vitamin research may be noted. The following
examples may serve to illustrate the trends of such vitamin synthesis
research.

Mueller's studies (36) on the nutritional requirements of certain
pathogenic bacteria led to the isolation of pimelic acid (XIX) from
cow's urine and the establishment of pimelic acid as a growth accessory
for the diphtheria bacillus. Eakin and Eakin (5) recognized that this
growth activity of pimelic acid, when considered in conjunction with
the tentative structural formulas for biotin, as published by du \'\-
gneaud, Hofmann, and Melville (58), might mean that pimelic acid
was being utilized as a biological precursor in the synthesis of biotin
(XX). They selected Aspergillus niger as a satisfactory mold to test
this biosynthesis and obtained data which demonstrated the activity

H02C(CH2)6C02H

(XIX)

NH NH

I I yeast

CH CH >

,,>;„,;, K,,.;ii.,c NH CH

diphtheria bacillus
>

CH CH

CH2 CH(CH2)4C02H

CH3 CH2(CH2)4C02H \s/

(XXI) (^^)

103

KARL FOLKERS

of pimelic acid in promoting the biosynthesis of biotin. Cysteine or
cystine, as sources of organic sulfur, enhanced the effect of pimelic
acid. The studies by du Vigneaud, Dittmer, Hague, and Long (57)
on the growth-stimulating effect of biotin for the diphtheria bacillus
in the presence and absence of pimelic acid led to the interpretation
that pimelic acid was being utilized as a precursor by the diphtheria
bacillus for the biosynthesis of biotin.

Experiments described by Dittmer, Melville, and du Vigneaud
(3) on the activity of desthiobiotin (XXI) for stimulating the growth
of S. cerevisiae showed that desthiobiotin disappeared from the incu-
bating yeast cultures and was replaced by an equivalent amount of
a substance possessing growth activity for L. casei. To these investi-
gators, the most logical interpretation was that desthiobiotin (XXI)
was transformed to biotin (XX) by the growing yeast cell. Further
support for this interpretation was supplied by the experiments of
Stokes and Gunness (54), which showed that, when extracts of yeast
grown with desthiobiotin were treated with Raney nickel, the process
(60) for converting biotin to desthiobiotin, the activity of the yeast-
formed substance for L. casei was destroyed and activity for the yeast
was retained. Avidin also neutralized the yeast-formed substance, as
it does biotin.

Another example of biosynthesis was found by Stokes, Keresz-
tesy, and Foster (55), who reported that the S.L.R. factor was con-
verted by S. lactis R into a substance which was active for the growth
stimulation of L. casei.

The possibility that alanine might be a biological precursor
of vitamin Be was recognized two years ago by Snell and Guirard (49)
when it was found that vitamin Be was not required for the growth of
S. faecalis R if sufficient alanine was added to the medium. Subse-
quent studies by Snell (48) showed that an enzymic digest of casein
contained an unknown biological precursor which, together with dl-
alanine, permitted the growth of L. casei in the absence of vitamin Be.
Since fl?( — )alanine was active and /(4-)alanine was almost inactive,
these data seem to be the first which indicate that the "unnatural"
amino acids may be essential for normal metabolic processes.

It is possible that future research on biological precursors and
biosyntheses of the vitamins will make these substances available by
new methods of startling simplicity.

104

VITAMIN RESEARCH

References

(1) Barger, G., Some Applications of Organic Chemistry to Biology and Medicine
McGraw-Hill, New York, 1930, p. 62.

(2) Brown, G. B., and du Vigneaud, V., J. Biol. Chem., 141, 85 (1941).

(3) Dittmer, K., Melville, D. B., and du Vigneaud, V., Science, 99, 203
(1944).

(4) Duschinsky, R., Dolan, L. A., Flower, D., and Rubin, S. A., Arch. Bio-
chem., 6, 480 (1945).

(5) Eakin, R. E., and Eakin, E. A., Science, 96, 187 (1942).

(6) Emerson, G., J. Biol. Chem., 157, 127 (1945).

(7) Emerson, O. H., Emerson, G. A., and Evans, H. M., Science, 83, 421
(1936).

(8) Fernholz, E., J. Am. Chem. Soc, 59, 1154 (1937).

(9) Fernholz, E., J. Am. Chem. Soc, 60, 700 (1938).

(10) Grussner, A., Bourquin, J. P., and Schnider, O., Helv. Chim. Acta, 28,
517 (1945).

(11) Gyorgy, P., J. Am. Chem. Soc, 60, 983 (1 938) .

(12) Gyorgy, P., Ann. Rev. Biochem., 11, 310 (1942).

(13) Gyorgy, P., Kuhn, R., and Lederer, E., J. Biol. Chem., 131, 745 (1939).

(14) Hahn, G., and Schales, O., Ber., 68, 24 (1935).

(15) Harris, S. A., Heyl, D., and Folkers, K., J. Biol. Chem., 154, 315 (1944);
J. Am. Chem. Soc, 66, 2088 (1944).

(16) Harris, S. A., Mozingo, R., Wolf, D. E., Wilson, A. N., Arth, G. E., and
Folkers, K., J. Am. Chem. Soc, 66, 1800 (1944).

(17) Hofmann, K., J. Am. Chem. Soc, 67 y 694 (1945).

(18) Hofmann, K., J. Am. Chem. Soc, 67, 1459 (1945).

(19) Hutchings, B. L., Bohonos, N., and Peterson, W. H., J. Biol. Chem.,
141,521 (1941).

(20) Hutchings, B. L., Stokstad, E. L. R., Bohonos, N., and Slobodkin, N. H.,
Science, 99, 2)1\ (1944).

(21) Jansen, B. C. P., and Donath, W. F., Mededeel. Dienst. Volksgezondheid
Nederland-Indie, 16, 186 (1926).

(22) Karrer, P., Fritzsche, H., Ringier, B. H., and Salomon, H., Helv. Chim.
Acta, 21, 820 (1938).

(23) Karrer, P., Salomon, H., and Fritzsche, H., Ilelv. Chim. Acta, 21, 309
(1938).

(24) Keresztesy, J. C., Rickes, E. R., and Stokes, J. L., Science, 97, 465
(1943).

(25) Keresztesy, J. G., and Stevens, J. R., Pruc. Soc. Exptl. Biol. M,;l., 38,
64 (1938); Stiller, E. T., Keresztesy, J. C., and Stevens, J. R., J. Am. Chem.
Soc, 61, 1237 (1939).

105

KARL FOLKERS

(26) Kogl, F., J. Soc. Chem. Ind., 57, 49 (1938).

(27) K5gl, F., and Borg, W. A. J., Z- physiol. Chem., 281, 65 (1944).

(28) Kuhn, R., and Wendt, G., Ber., 71, 780, 1118 (1938).

(29) Lepkovsky, S., Science, 87, 169 (1938); J. Biol. Chem., 124, 123 (1938).

(30) Major, R. T., Chem. Eng. News, 20, 517 (1942).

(31) Melville, D. B., in Vitamins and Hormones. Vol. II, Academic Press, New
York, 1944, p. 29.

(32) Mitchell, H. K., J. Am. Chem. Soc, 66, 274 (1944).

(33) Mitchell, H. K., Snell, E. E., and Williams, R. J., J. Am. Chem. Soc,
63, 2284 (1941); 66, 267 (1944).

(34) Mitchell, H. K., Weinstock, H. H., Jr., Snell, E. E., Stanbery, S. R.,
and Williams, R. J., J. Am. Chem. Soc, 62, 1776 (1940).

(35) Mozingo, R., Wolf, D. E., Harris, S. A., and Folkers, K., J. Am. Chem.
Soc, 65, 1013 (1943).

(36) Mueller, J. H., J. Biol. Chem., 119, 121 (1937).

(37) Oden, J. W., Oden, L. H., and Sebrell, W. H., U. S. Pub. Health Repts.,
54, 790 (1939).

(38) Ott, W. H., J. Biol. Chem., 157, 131 (1945).

(39) Pfiflfner, J. J., Biiikley, S. B., Bloom, E. S., Brown, R. A., Bird, O. D.,
Emmett, A. D., Hogan, A. G., and O'Dell, B. L., Science, 97 y 404 (1943).

(40) Pilgrim, F. J., Axelrod, A. E., Winnick, T., and Hofmann, K., Science,
102,35 (1945).

(41) Robinson, R., J. Chem. Soc, 111, 876 (1917).

(42) Schopf, C., Ann., 497, 1 (1932).

(43) Sebrell, W. H., and Buder, R. E., U. S. Pub. Health Repts., 53, 2282
(1938).

(44) Smith, L. I., Chem. Revs., 27, 287 (1940).

(45) Snell, E. E., J. Am. Chem. Soc, 66, 2082 (1944).

(46) Snell, E. E., J. Biol. Chem., 154, 313 (1944); 157, 475 (1945).

(47) Snell, E. E., J. Biol. Chem., 157, 491 (1945).

(48) Snell, E. E., J. Biol. Chem., 158, 497 (1945).

(49) Snell, E. E., and Guirard, B. M., Proc Natl. Acad. Sci. U. S., 29, 66
(1943).

(50) Snell, E. E., Guirard, B. M., and Williams, R. J., J. Biol. Chem., 143,
519 (1942).

(51) Snell, E. E., and Mitchell, H. K., Proc Natl. Acad. Sci. U. S., 27, 1-7
(1941).

(52) Snell, E. E., and Peterson, W. H., J. Bad., 39, 273 (1940).

(53) Stiller, E. T., Keresztesy, J. C., and Finkelstein, J., J. Am. Chem. Soc,
62, 1779 (1940).

(54) Stokes, J. L., and Gunness, M., J. Biol. Chem., 157, 121 (1945).

io6

VITAMIN RESEARCH

(55) Stokes, J. L., Keresztesy, J. C, and Foster, J. VV., Science, 100, 522
(1944).

(56) Stokstad, E. L. R., J. Biol. Chem., 149, 573 (1943).

(57) du Vigneaud, V., Dittmer, K., Hague, E., and Long, B., Science, 96,
186 (1942).

(58) du Vigneaud, V., Hofmann, K., and Melville, D. B., J. Am. Cliem Soc
64, 188 (1942).

(59) du Vigneaud, V., Hofmann, K., Melville, D. B., and Gyorgy, P., J.
Biol. Chem., 140, 643 (1941).

(60) du Vigneaud, V., MelvUle, D. B., Folkers, K., Wolf, D. E., Mozingo, R.,
Keresztesy, J. C, and Harris, S. A., J. Biol. Chem., 146, 475 (1942).

(61) Weinstock, H. H., Jr., Mitchell, H. K., Pratt, E. F., and Williams, R. J.,
J. Am. Chem. Soc, 61, 1421 (1939).

(62) Williams, R. J., Science, 93, 412 (1941).

(63) WUliams, R. J., Science, 95, 340 (1942).

(64) WUliams, R. J., Ann. Rev. Biochem., 12, 309 (1943).

(65) Wiliams, R. J., Ann. Rev. Biochem., 12, 331 (1943).

(66) Williams, R. J., Mitchell, H. K., Weinstock, H. H., Jr., and Snell,
E. E., J. Am. Chem. Soc, 62, 1784 (1940).

(67) Williams, R. J., Truesdail, J. H., Weinstock, H. H., Jr., Rohrmann, E.,
Lyman, C. M., and McBurney, C. H., J. Am. Chem. Soc, 60, 2719 (1938).

(68) Williams, R. J., Weinstock, H. H., Jr., and Mitchell, H. K., Abstracts,
96th Meeting, Am. Chem. Soc, Sept., 1938, Division of Organic Chemistry,
p. 34.

(69) Williams, R. R., Science, 95, 335 (1942).

(70) Williams, R. R., Waterman, R. E., and Keresztesy, J. C, J. Am. Chem.
Soc, 56, 1187 (1934).

(71) Williams, R. R., Waterman, R. E., Keresztesy, J. C, and Buchman,
E. R., J. Am. Chem. Soc, 57, 536 (1935).

(72) Woolley, D. W., Waisman, H. A., and Elvehjcm, C. A., J. Am. Chem.
Soc, 61, 977 (1939); J. Biol. Chem., 129, 673 (1939).

107

8

QUANTITATIVE ANALYSTS
IN BIOGHEIVIISTIIY

DONALD D. VAN SLYKE, member of the rockefeller institute

FOR medical research, NEW YORK; VVTLLARD GIBBS MEDALIST

^HE BIOCHEMISTRY of today is based to a large extent
-*• on quantitative analyses by means of micro methods. It
lias not only applied the methods developed in laboratories of organic
and inorganic chemistry, but has also rapidly developed new pro-
cedures to meet the demands of its expanding range of research . While
Pregl's system of elementary organic microanalysis, published in 1911,
was gaining application, Folin and Wu in 1919 and Bang in 1916 pub-
lished quite different systems adapted to blood analyses. The ultra
in microanalyses is represented by the methods of capillary colorimetry
developed by A. N. Richards and his collaborators for the analysis of
glomerular urine, and the extraordinary combination of physical and
chemical procedures applied by Linderstr0m-Lang in his studies of the
enzymes of cells.

In a brief survey of the field it will be possible only to mention
some of the different types of analytical procedure that have been
applied in biochemistry, with examples of a few applications, and refer-
ences to reviews in which descriptions and bibliographies can be found.

Gravinielric A nalysis

The appearance in 1911 of the Kuhlman micro balance, and
of Pregl's system of micro elementary analyses decreased the size of

109

D. D. VAN SLYKE

samples usually taken for analyses from 100-150 mg. to 4 or 5 mg.,
and opened a new epoch, not only in organic analysis, but also in
organic chemistry. For it became possible to solve problems when
amounts of material were available that would have been inadequate
even for preliminary analyses by the old macro methods. It is ques-
tionable, for example, that the rapid development of knowledge con-
cerning the structure of the vitamins would have been possible without
these micro methods. The techniques of micro analysis developed in
Austria by Pregl and by Emich (1) were brought to this country, es-
pecially by J. B. and V. Niederl (3), and are available in a recent
volume by these authors. For ultra micro gravimetric work, Lowry
(2) has recently described a simple quartz fiber balance which will
weigh 200 gammas to 0.03 gamma.

Volumetric Analysis

Bang's work (4) introduced micro titrations into biochemistry,
utilizing the principle of keeping the volume of titrated liquid small,
to maintain the sharpness ^of the end point. Rehberg (9) followed
shortly with his micro burette, in which a thread of standard solution in
a calibrated capillary of about 1 mm. bore is expelled into the titrated
solution by mercury moved by a plunger advanced by turning a steel
screw. With this burette, the solution could be measured to within
±0.1 cu. mm., so that 50 cu. mm. was ample for a good titration.
Linderstr0m-Lang (7) moved the error down to ±0.01 cu. mm. He
used still narrower capillaries, and stirred the titrated solution in a
minute glass thimble by a magnetic stirrer consisting of a bit of iron,
enclosed in a glass droplet, which was lifted and lowered in the titrated
solution by the automatic opening and closing of the circuit to a mag-
net. Scholander (10,11) has applied the metal micrometer screw
that is readily available because of its general use in accurate industrial
mechanical work. The column of standard solution in a fine capillary
is moved, as in the Rehberg burette, by pressure of a mercury column
advanced by a screw and plunger; but in the Scholander burette the
extent to which the screw is turned serves as a measure of the volume
of solution delivered, and calibration of the capillary is not necessary.

A limitation of the Rehberg and Scholander burettes is that they
cannot be used to deliver solutions which react with mercury. For

no

QUANTITATIVE ANALYSIS

such solutions, Longwell and Hill (8) have modified the Rehberg
burette by introducing an elastic rubber diaphragm between the
mercury and the solution. Clark, Levitan, Gleason, and Greenberg
(5) have applied Scholander's micrometer principle, but employ the
micrometer screw to push air (instead of mercury) from a hypodermic
syringe into a capillary which delivers the solution.

The general principles of volumetric microanalyses have been
clearly elucidated by Conway (6).

Gasometric Analysis

One of the oldest quantitative analyses in biochemistry is the
determination of urea by measurement of the nitrogen gas liberated by
reaction with alkaline bromine solution. It exemplifies procedures in
which a substance is measured by the amount of gas that it liberates
when it reacts with properly chosen reagents. In such analyses, the
measurement is based, as in gravimetric methods, on direct observation
of the amount of substance obtained, independent of comparison with
standard solutions, such as are required in titration and colorimetry.
Combined with this independence are the advantages of a quick meas-
urement and easy adaptation to micro quantities. Historically, micro
gasometric procedures were introduced into biochemistry for deter-
mination of the blood gases, and were then adapted to more general
analyses.

The first of such procedures was the blood gas analysis of Bar-
croft and Haldane (12) in which the oxygen liberated from blood by
ferricyanide, or the carbon dioxide liberated by acid, was measured
by the gas displaced into a capillary tube. The procedure requires
accurate temperature control and constant shaking until equilibrium
between dissolved and supernatant gases is reached. The method was
elaborated by Warburg (19), and has been used for a great variety of
purposes by Warburg and others, particularly in following the course
of enzymic reactions by measurement of the oxygen absorbed or the
carbon dioxide evolved. The ease with which the course of a reaction
can be followed by observing the increase in gas volume particularly
adapts the procedure to the observation of comparative reaction veloci-
ties (13). A recent refinement of the apparatus, and the principles
of its use, are described by Summerson (15).

Ill

D. D. VAN SLYKE

The manometric apparatus of Van Slyke and Neill (18), like
the Barcroft-Haldane apparatus, was first developed for determination
of the blood gases, and its use then spread to other micro analyses,
including the determination of urea, reducing and fermentable sugars,
the ammonia yielded by Kjeldahl digestions, amino nitrogen by
measurement of the nitrogen yielded by reaction with nitrous acid
(18), free alpha-amino acids by measurement of the carbon dioxide
yielded by reaction with ninhydrin [RCH(NH2)COOH -> RCHO +
CO2 + NH3] (16), organic carbon by measurement of the carbon
dioxide evolved by a wet combustion completed in two minutes (17),
and various other determinations. In these procedures, the gases are
either evolved in, or transferred to, a 50-cc. chamber, provided at the
top with small bulbs for measuring 0.5 and 2.0 cc. of gas, and connected
at the bottom with a mercury manometer. The evolved gas is brought
to 0.5 or 2.0 cc. volume, and its pressure is read on the manometer.
In a mixture of gases, each gas can be measured separately by measur-
ing the pressure before and after the absorption of each gas by intro-
duction of a proper reagent. The carbon combustion method per-
mits micro determination of any organic substance, such as the blood
fats, that can be isolated by extraction with volatile solvents; the com-
bustion also provides a micro measurement of any substance that can
be isolated as a carbon-containing precipitate, e. g., sulfate as benzidine
sulfate, magnesium as hydroxyquinolate, phosphorus as strychnine
phosphomolybdate (14).

While the Barcroft-Haldane-Warburg apparatus is adapted to
following the course of time reactions, the Van Slyke-Neill apparatus is
fitted for quick determination of the total amounts of gases evolved by
rapid quantitative reactions. Hence the Haldane apparatus has
found its chief application in following enzymic time reactions, while
the Van Slyke-Neill apparatus is the one usually employed in quantita-
tive micro determinations of specific substances.

Photometric Analysis

Under the chief initial stimulus of Folin and of S. R. Benedict,
during the past forty years chromogenic reactions have been developed
for estimating numerous biological substances by producing colored
products from them (25). In some cases the colored products are

I 12

QUANTITATIVF, ANALYSIS

defined; in other cases neither ihe reactions nor the colored products
are defined with certainty; hut emiiirical conditions ha\e heen fixed
which relate the color quantitatively to the amount of the substance
under analysis. Thus Folin (21) used the color produced by Nessler's
reagent as the means for quantitative estimation of ammonia, and
hence of nitrogen, through the ammonia obtained by Kjeldahl diges-
tion, and of urea through the ammonia obtained by urea hydrolysis.
The constitution of the colored compound of ammonia and potassium
mercuri-iodide is still under dispute, but the colorimetric results are
accurate. Sugars reducing Cu++ to Cu''" were determined by Folin
and VVu (21) by letting the cuprous ion thus formed act on a molybdatc
solution, with reduction of the colorless hexavalent molybdenum to a
lower valence, which shows an intense blue color; the reactions do not
appear to be stoichiometric, but quantitative relations can be obtained
between colored molybdenum products and the initial sugar. For
almost every substance of interest in quantitative biochemical analysis,
chromogenic reactions have been devised which can be used for more
or less accurate estimation. As a rule these procedures are rapid, and
are adapted to minute amounts of material.

During the first years of the colorimetric epoch, the instrument
in general use was the familiar Duboscq colorimeter, in which the
depths of colored solution layers in two parallel columns, one of the
unknown solution and the other of a standard, are varied until the two
fields viewed with the eye appear equal. The simplicity and versatility
of the Duboscq colorimeter assisted greatly in the rapid adoption of
colorimetric procedures.

Photometers, in which the percentage transmission of light could
be measured without standard solutions for direct comparison, were
known long before this period, but were too complicated and ex-
pensive for ordinary routine in biochemical laboratories. During the
past two decades, however, photometers have been progressively made
more adaptable to such routine, and have been gradually displacing
colorimeters of the Duboscq type. In the photometer, the concentra-
tion of light-absorbing solute is related to the optical density according
to the simple linear formula of Beer's law (20,22,23):

C = kl) (1)

where C is the concentration and D is the optical density (or extinction).

D. D. VAN SLYKE

D is the logarithm of l/T, T being the fraction of Ught transmitted by
the substance measured.

The value of k for a given colored solute varies with the wave
length of light. Hence Beer's law holds exactly only for monochromatic
light; and the accuracy with which the formula applies to a given
photometer depends partly on how narrow a spectral band can be
given by the analyzing device interposed between the source of light
and the solution, a limitation under which the Duboscq colorimeter
does not suffer. Some solutes do not exactly follow Beer's law, even
with the narrowest spectral bands. Most solutions, however, do follow
the law over the concentration ranges used for analysis, and with the
spectral bands provided by the instruments now available for routine
analytical work. Solutions for which Beer's law does not hold can be
analyzed by using empirical calculation curves of optical density vs.
concentration.

The photometer has several advantages over the Duboscq
colorimeter. The validity of equation (1) makes it possible to measure
the concentration of one colored solute in the presence of others, since
the total optical density is additive:

D = CiAx + C,/h. . . (2)

Hence, if the medium in which the concentration of a solute is to be
measured is itself colored or turbid, the increase in optical density due
to the presence of the specific solute can be used as a measure of its
concentration. Correction for nonspecific color is much less simple
in a Duboscq colorimeter.

Because the k value for each solute changes with the wave length,
it is possible to determine two colored solutes in the same solution by
measuring the optical densities at two different wave lengths. Two
densities. Da and Z)j, are thus measured:

Da = C,/k^ + C^/k^ (2a)

D, = C,/k, + C2A4 (2b)

If hi, kz, ks, and ki are known, Ci and C2 can be calculated by simul-
taneous equations from the observed Da and i)j.

The concentration of a turbid suspension can be estimated from
its optical density, in the same way as the concentration of a colored
solute.

1 14

QUANTITATIVE ANALYSIS

The eye strain and subjective error accompanying the use of a
visual instrument are obviated in most of the modern photometers by
using photocells to measure the intensity of the transmitted light. The
rapidity with which a series of observations can be carried out is also
much greater when the eye is replaced by the photocell.

A further advantage of the photometer is that, with photocell
measuring devices, it can be used with light waves extending into the
ranges of the ultraviolet and infrared, broadening the range of ac-
cessible analyses.

In the ultra micro colorimetric methods which Richards and
his collaborators (24) developed for analyses of samples of 1 cu. mm.
of glomerular filtrate, the sample is drawn into a small glass capillary
in which it is mixed with chromogenic reagent, and the color is esti-
mated by comparing the capillary under a microscope with a series of
standards similarly prepared. The comparison with graded standards
is a return to simplest first principles, but the technique of the applica-
tion gave results to 1%.

Fluorimetric Analysis

The diffuse fluorescent light developed by passing light rays
into solutions of fluorescent substances (28) can be measured by the
same visual or electrical means employed in colorimeters and photom-
eters, and serves in some cases, such as determination of riboflavin
(27), quinine, atabrine (26), and related compounds, to measure
substances in more dilute solutions than can be handled with a photom-
eter. In fluorimetry, the concentration of fluorescent substance is
directly proportional to the intensity of the light measured, instead of
being inversely proportional to the log of the transmission, as in pho-
tometry.

Po larograph ic A nalys is

This procedure, which has gained rapid utility during the past
few years, was introduced by Heyrovsky in Prague in 1925, and has
been applied to a multiplicity of analyses of substances, both organic
and inorganic. It is based on measurement of the amperages obtained
at observed voltages applied to solutions of electroreduciblc or electro-

D. D. VAN SLYKE

oxidizable substances in a cell in which one electrode consists of mercury
falling from a fine capillary.

When a current at gradually increasing voltage is passed to a
small mercury electrode through a solution containing a solute capable
of giving or receiving electrons ("redox solute") at a given potential,
relatively little current is obtained until the decomposition potential
of the redox solute is reached. Then the current rapidly rises, with
further increase in voltage until a plateau is reached at which the
redox solute is providing its maximal flow of electrons. This flow, and
the resultant current, are proportional to the concentration of the redox
solute, which determines the rate at which its molecules diff'use to the
electrode and discharge or receive electrons. Since diff'usion is a proc-
ess independent of the voltage, increase of voltage above that at which
electrolytic decomposition of the redox solute equals its rate of diff'usion
to the mercury electrode does not further increase the flow of electricity;
the concentration of the active solute thus forms a bottleneck which
limits the current. The curve of current vs. voltage then reaches a
plateau, the height of which, in milliamperes, is proportional to the
concentration of active redox solute (32-34).

Maintenance of the proportionality requires a continually
renewed surface of the mercury electrode; both renewal of the surface
and setting its area at a small size are obtained by using mercury
dropping from a fine capillary, of about 0.03 mm. diameter, as the
electrode; a drop is delivered about once in two to four seconds.

The current fluctuates somewhat as each drop of mercury
expands and falls, but the average current, i, in microamperes is given
by the "Ilkovic equation":

i = 605 n D"'' C m'' t"'' (3)

C is the millimoles of redox solute per liter, n is the number of electrons
exchanged per molecule of redox solute in the electrolytic decomposi-
tion, m is the weight of mercury flowing from the capillary per second,
and t is the time required for formation of one drop of mercury. The
procedure is adapted to analysis of highly dilute solutions, 0.001 molar
and lower concentrations.

Measurement of the current in such a system provides a measure
of the concentration of the redox solute. Furthermore, if two different
redox solutes are present with diff"erent decomposition potentials, they

ii6

QUANTITATIVE ANALYSIS

can be determined one after another by using voltages related (.. ilxir
respective decomposition potentials.

The setup also can be used lor electrometrir titrations, if .1
reagent which reacts with the redox solute is added while the current
is measured at proper voltage, a drop in current will accompany the
disappearance of the redox solute, and the end jjoint will be indicated
when the current has fallen to a residual value, representing con-
ductance by factors other than the redox solute.

For literature, theoretical discussion, description of the different
types of analysis to which the procedure has already been applied,
and the precautions that must be observed, the reader is referred to the
bibliography (35), in particular to KolthoflT (32,33) and Muller (34).
The various inorganic cations and anions are determinable; also re-
ducible organic compounds, such as aldehydes, ketones, and nitro
compounds. Eisenbrand and Picher (31) found that the sex hormones
with the 0=C — C=C — group are reducible at the mercury electrode
and can be determined polarographically: these hormones include
testosterone, progesterone, and desoxycorticosterone, but not andro-
sterone nor dehydroandrosterone. Application to solve a hitherto
difficult biochemical problem is illustrated by the work of Berggren
(30) and of Beecher et al. (29) in determining the oxygen tension of
arterial blood plasma and other body fluids. These authors also
describe their apparatus in detail.

Spectrograph ic A nalys is

Measurement of the intensity of light of characteristic wave
length emitted by incandescent elements has been long used in both
qualitative and quantitative analysis, and in special problems for
quantitative or semiquantitative estimation of minute amounts of
specific elements. Lundegardh (39) in 1929 applied the principle
in a manner which makes it applicable to micro determination of
mineral bases in biological fluids. The solution is sprayed from an
atomizer at a constant rate, and the stream of air with suspended fluid
is mixed with acetylene, which is burned, heating the mineral bases
in the suspension to such a temperature that they emit their charac-
teristic light bands, the intensity of which is measured by an electro-
photometer (37). The procedure is reviewed by Ells (38) and by

D. D. VAN SLYKE

Cholak and Hubbard (36). The latter describe the appHcation of the
procedure to determination of minute amounts of cadmium in blood
and urine, and compare it with polarographic and photometric pro-
cedures for this purpose. An apparatus devised by the American
Cyanamid Company, not yet in general production, determines
potassium in serum in a few minutes with an error not over =*=5%.

Micro Diffusion Analysis

Conway (41) has devised a simple chamber for the determi-
nation, primarily, of ammonia, but applicable also to estimation of
other volatile substances that can be set free by quantitative reactions
and transferred by diffusion at atmospheric pressure to absorbing
solutions in which the diffused substance can be measured, by titration,
photometry, or otherwise. The apparatus consists of a flat, cylin-
drical dish, of 60-mm. diameter and 10 mm. high (inner measure-
ments), from the inner bottom of which rises a ring of 33-mm. diameter
and 5 mm. high. The top of the dish is ground accurately flat, so that
when lubricated with vaseline or other proper material it can be closed
gas tight by a flat glass cover. The chamber consists, therefore, of an
outer ring, about 60 mm. wide, surrounding an inner low cylinder;
when the chamber is covered a free space of 5 mm. is left open between
the covering plate and the wall about the inner cylinder, permitting
free diffusion of volatile substances from the outer compartment to the
inner. To determine ammonia, the solution containing it is alka-
linized in the outer compartment, and acid is placed in the inner
compartment. In one or more hours, depending on temperature
and the other conditions, the ammonia from the former diffuses through
the air space of the chamber into the acid, where it can be measured
in amounts of a few micrograms. Conway applied this procedure to
determination of the minute amounts of ammonia in blood, to esti-
mation of the ammonia formed by micro Kjeldahl nitrogen digestion,
and of the ammonia formed by decomposition of urea with urease;
also to chloride and bromide, which were oxidized to chlorine and
bromine and diffused into potassium iodide solution for iodometric
titration. Conway also applied the diffusion chamber to the determi-
nation of carbon dioxide, which was caused to diffuse into a barium
hydroxide solution, where the excess alkali was titrated. Borsook (40)

ii8

QUANTITATIVE ANALYSIS

developed applications of the ammonia procedures. VVinnick (42)
applied the diffusion apparatus to the determination of: alcohol,
with chromic acid in the inner chamber; lactic acid, which was oxi-
dized in the outer chamber to acetaldehyde, the latter diffusing to a
bisulfite solution in the inner chamber; acetone, which diffused to
bisulfite; threonine, which was oxidized by periodate to acetaldehyde
in the outer chamber, with diffusion of the aldehyde to bisulfite.

References

GRAVIMETRIC ANALYSIS

(1) Emich, F., Microchemical Laboratory Manual. Trans, by F. Schneider.
Wiley, New York, 1932.

(2) Lowry, O. H., "A quartz fiber balance," J. Biol. C/iem., 140, 183
(1941). "A simple quartz torsion balance," ibid., 152, 293 (1944).

(3) Niederl, J. B., and Niederl, V., Micromethods oj Quantitative Organic
Analysis. Wiley, New York, 1942.

VOLUMETRIC ANALYSIS

(4) Bang, I., Melhoden zur Mikrobestimmung einiger Blutbestandteile. Berg-
mann, Wiesbaden, 1916.

(5) Clark, W. G., Levitan, N. I., Gleason, D. F., and Greenberg, G., "Ti-
trimetric microdetermination of chloride, sodium, and potassium in a single
tissue or blood sample," J. Biol. Chem., 145, 85 (1942).

(6) Conway, E. S., Micro-diffusion Analysis and Volumetric Error. Van
Nostrand, New York, 1 942.

(7) Linderstr0m-Lang, K., "Distribution of enzymes in tissues and cells,"
Harvey Lectures, 34, 214 (1939).

(8) Longwell, B., and Hill, R. M., "A modified Rehberg burette for use
with titrating solutions which react with mercury," J. Biol. Chem., 112, 319
(1935).

(9) Rehberg, P. B., "A method of microtitration," Biochem. J., 19, 270
(1925).

(10) Scholander, P. F., "Microburette," Science, 95, 177 (1942).

(11) Scholander, P. F., and Edwards, G. A., and Irving, L., "Improved
microburette," J. Biol. Chem., 148, 495 (1943).

GASOMETRIC ANALYSIS

(12) Barcroft, J., and Haldane, J. S., "A method of estimating the oxygen
and carbonic acid in small quantities of blood," J. Physiol., 28, 232 (1902).

D. D. VAN SLYKE

(13) Dixon, M., Manometric Methods. Cambridge Univ. Press, London,
1934.

(14) Hoagland, C. L., "Microdetermination of sulfate and phosphate by
manometric combustion of their organic precipitates," J. Biol. Chem., 136,
543 (1940). "Micro manometric determination of magnesium," ibid., 553.

(15) Summerson, W. H., "A combination simple manometer and constant
differential manometer for studies in metabolism," J. Biol. Chem., 131, 579
(1939).

(16) Van Slyke, D. D., Dillon, R. T., MacFadyen, D. A., and Hamil-
ton, P. B., "Gasometric determination of carboxyl groups in free amino
acids," J. Biol. Chem., 141,627 (1941). Hamilton, P. B., and Van Slyke,
D. D., "The gasometric determination of free amino acids in blood filtrates
by the ninhydrin-carbon dioxide method," tZ»2a?., 150,231 (1943).

(17) Van Slyke, D. D., and Folch, J., "Manometric carbon determina-
tion," J. Biol. Chem., 136, 509 (1940).

(18) Van Slyke, D. D., and Neill, J. M., "The determination of gases in
blood and other solutions by vacuum extraction and manometric meas-
urement," J. Biol. Chem., 61, 523 (1924); 83, 449 (1929). "Applications to
other analyses," in Peters, J. P., and Van Slyke, D. D., Quantitative Clinical
Chemistry. Methods. Williams & Wilkins, Baltimore, 1932; rev., 1943.

(19) Warburg, O., "Verbesserte Methode zur Messung der Atmung und
Glykolyse," Biochem. Z-, 152, 51 (1924).

PHOTOMETRIC ANALYSIS

(20) Ashley, S. E. Q., "Spectrophotometric methods in modern analytical
chemistry," Ind. Eng. Chem., Anal. Ed., 11, 72 (1939).

(21) Folin, O., and Wu, H., "A system of blood analysis," J. Biol. Chem.,
38, 81 (1919).

(22) Hamilton, R. H., "Photoelectric photometry. An analysis of errors
at high and low absorption," Ind. Eng. Chem., Anal. Ed., 16, 123 (1944).

(23) Miiller, Ralph H., "Photoelectric methods in analytical chemistry," Ind.
Eng. Chem., Anal. Ed., 7, 223 (1935); 11, 1 (1939).

(24) Richards, A. N., Bordley, J., 3rd, and Walker, A. M., J. Biol. Chem.,
101, 179, 193, 223, 229 (1933).

(25) Snell, F. D., and Snell, C. T., Colorimetric Methods oj Analysis, including
Some Turbidimetric and Nephelometric Methods. 2nd ed., 2 vols., Van Nos-
trand, New York, 1936-1937.

FLUORIMETRIG ANALYSIS

(26) Brodie, B. B., and Udenfriend, S., "The estimation of atabrine in bio-
logical fluids," J. Biol. Chem., 151, 299 (1943).

I20

QUANTITATIVE ANALYSIS

(27) Hand, D. B., "Determination of riboflavine in milk by photoelectric
fluorescence measurements," hid. Eng. Chem., Anal. Ed., II, 306 (1939).

(28) Kavanagh, F., "New photoelectric fluorimeter and some applications,"
Ind. Eng. Chem., Anal. Ed., 13, 108 (1941).

POLAROGRAPHIC ANALYSIS

(29) Beecher, H. K., Follansbee, R., Murphy, A. J., and Craig, F. N., "Deter-
mination of the o.xygen content of small quantities of body fluids by polaro-
graphic analysis," J. Biol. Chem., 146, 197 (1942).

(30) Berggren, S. M., "The oxygen deficit of arterial blood caused by
non-ventilating parts of the lung," Acta Physiol. Scand. SuppL, 4, XI (1942).

(31) Eisenbrand, J., and Picher, H., "t)bcr den polarographischen Nachweis
von biologisch vvichtigen Ketonen der Steringruppe," <;. physiol. Chem ,
260,83 (1939).

(32) Kolthoff, I. M., "Factors to be considered in quantitative polarog-
raphy," Ind. Eng. Chem., Anal. Ed., 14, 195 (1942).

(33) Kolthoff", I. M., and Lingane, J. J., Polarography. Interscience, New
York, 1941.

(34) Miiller, O. H., The Polarographic Method oj Analysis. J. Chem. Educa-
tion, Easton, 1941.

(35) Sand, H. J. S., Bibliography of the Dropping-Mercury Electrode. Leeds
& Northrup, Philadelphia, 1941.

SPECTROGRAPHIC ANALYSIS

(36) Cholak, J., and Hubbard, D. M., "Spectrochemical analysis with the
air-acetylene flame," Ind. Eng. Chem., Anal. Ed., 16, 728 (1944).

(37) Churchill, J. R., "Techniques of quantitative spectrographic analysis,"
Ind. Eng. Chem., Anal. Ed., 16, 653 (1944).

(38) Ells, V. R., "The Lundegardh flame method of spectrographic analy-
sis," J. Optical Soc. Am., 31, 534 (1941).

(39) Lundegardh, H., Die Quantitative Spektralanalyse der Elemente. Fischer,
Jena. Part I, 1929; Part II, 1934.

MICRO DIFFUSION ANALYSIS

(40) Borsook, H., "Micromethods for determination of ammonia, urea, and
total nitrogen," J. Biol. Chem., 110, 481 (1935).

(41) Conway, E. J., Micro-diffusion Analysis and Volumetric Error. Van Nos-
trand, New York, 1940.

(42) Winnick, T., "Micro-diffusion methods. Alcohol," Ind. Eng. Chem.,
Anal. Ed., 14, 523 (1942). "Acetone," J. Biol. Chem., 141,115 (1941).
"Lactic acid," ibid., 142, 451 (1942). "Threonine," ibid., 142, 461 (l'M2).

121

ENZYMIC HYDROLYSIS

AND SYNTHESIS OF

PEPTIDE BONDS

JOSEPH S. FRUTON, associate professor of physiological

CHEMISTRY, YALE UNIVERSITY; LILLY AWARD IN BIOLOGICAL CHEMISTRY

Specificity of Proteolytic Enzymes

THE ACTION of proteolytic enzymes on peptide linkages
involves a high degree of specificity. No proteolytic
enzyme acts on peptide bonds indiscriminately, and each enzyme
hydrolyzes only such peptide bonds as are present in the substrate in
a certain structural setting. Thus, the nature of the requisite struc-
tural attributes of the substrate is an expression of the specificity of the
enzyme which hydrolyzes the substrate. In recent years, much
attention has been devoted to the determination of those structural
elements in the substrate molecule which are essential for the action of
various proteolytic enzymes. These studies have permitted the formu-
lation of several hypotheses concerning the specific action of the pro-
teolytic enzymes.

Modern theories concerning the specificity of proteolytic
enzymes are based on the assumption, made by von Euler and Joseph-
sohn (17) in their "dual-affinity" theory, that ereptic peptidase— it
was not recognized at that time that "erepsin" represents a mixture of
many peptidases — combines with two atomic groupings of the sub-
strate molecule. Subsequent work of Balls and Kohler (2) presented

123

J. S. FRUTON

further evidence for the "dual-affinity" hypothesis. More recently,
the finding of synthetic substrates for the protein-spUtting enzymes
(pepsin, trypsin, chymotrypsin, papain, etc.) has led to the extension
of this hypothesis to all representative proteolytic enzymes (10).

Of the two essential points in the substrate, one, of necessity,
must be the sensitive CO — NH group or some part thereof. The other
requisite point of contact lies in the "backbone"* of the substrate and
varies with the nature of the enzyme. The nature of this second
group and its position in the backbone relative to the sensitive peptide
bond provide the basis for the classification of proteolytic enzymes into
four groups as given in Table I (the requisite groups are italicized and
the sensitive peptide linkage is indicated by means of a dotted line).

Table I
Classification of Proteolytic Enzymes

Group
No.

Linkage attacked

Classification

R

I

NHiCUCO^NH-

Aminojjeptidases

R

1

Exopeptidases

II

■ CO^NHCHCOOH

Carboxypeptidases

R

III

. CO—NH- CH • CO^NH-

Proteinases

R

1

V Endopeptidases

IV

• CO^NH- CH • CO—NH-

Proteinases

In groups I and II are included the enzymes restricted in their
action to peptide bonds at the end of a peptide chain. The peptidases
belonging to group I (aminopeptidases) selectively attack the chain
at the peptide linkage adjacent to the amino end of the chain while
the peptidases of group II (carboxypeptidases) attack the chain at the

* In order to describe the structural setting of a peptide bond, it is desirable

to speak of a "backbone" of the substrate, i. e., the sequence of — NH — CH — CO —
groupings linked through peptide bonds, and the "side chains," :. e., the groups
attached to the CH groups of the backbone.

124

HYDROLYSIS OF PEPTIDE BONDS

peptide linkage adjacent to the carboxyl end of the chain. The
amino- and carboxypeptidases cannot spUt Unkagcs that are centrally
located in the peptide chain; for this reason they arc referred to as
exopeptidases.

All the protein-splitting enzymes whose backbone requirements
have been determined belong to group III. These enzymes are
capable of hydrolyzing central peptide bonds and, therefore, are re-
ferred to as endopeptidases. They were found lo require, in their
substrates, a peptide bond in close proximity to the carbonyl group of
the peptide bond which is hydrolyzed by the enzyme. Although no
known proteolytic enzyme has been identified as belonging to group
IV, the suggestion was made recently that an enzyme which hydrolyzes
leucylglycylglycine and which is found in intestinal mucosa may belong
to this group (31).

The presence of the indispensable groups in the backbone of a
substrate in itself is insufficient to render the substrate susceptible to
the action of an enzyme. It has been found that each of the proteolytic
enzymes tested thus far also requires the presence, in the substrate, of
a certain type of side chain (R) in a precisely defined location. In the
second column of Table II, the recjuired location of side chain R is
indicated for each of the enzymes mentioned; and the chemical nature
of these R groups is given in the third column. The enzymes which
have been listed require in their substrates one of the following side
chains: isobutyl as in leucine, benzyl or /*-hydroxybenzyl as in phenyl-
alanine or tyrosine, aminobutyl or guanidopropyl as in lysine or argi-
nine. The side chains mentioned here represent only a few of those
which jut out from the peptide chain of proteins. It will be a task of
the future to determine precisely the specificity of proteolytic enzymes
which require, in their substrates, the side chains of amino acids such
as glycine, glutamic acid, histidine, tryptophane, etc.

Homos pecific Proleolytic Enzymes

In the course of the systematic study of the specificity of proteo-
lytic enzymes, it was noted that several different enzymes exhibited
the same backbone and side-chain requirements in their substrates.
For example, as will be seen from Table II, an enzymic component of
papain and an enzymic component of beef spleen cathepsin have the

125

J. S. FRUTON

same type of specificity and split the same synthetic substrates as does
crystalline pancreatic trypsin. Because of this relationship, the
members of this group of enzymes have been designated "trypsinases."
Similarly, evidence has been obtained for the existence of groups of
enzymes related in specificity to pepsin, to leucine aminopeptidase,

Table II
Specificity of Proteolytic Enzymes

Enzyme

Requisite groups in substrate
backbone

Requisite groups in
substrate side chain

Peptidases (Exopepiidases)

Leucine aminopeptidase from in-
testinal mucosa, beef spleen,
beef kidney, and swine kidney

Chymotrypsin aminopeptidase

Other aminopeptidases

Carboxypeptidase from pancreas,
beef spleen, beef kidney, and
swine kidney

Other carboxypeptidases

R

I
NHiCHCO^NH...

Same

Same

R

I
. . .CO^NH-CH-COOH

Same

CHi
CHi

\CH • CH2 .

HO-C6H4CH2.

or
C8H6CH2...

HGCsHi-CHj.

or
CeHsCHj...

Proteinases (Endopeptidases)

Pepsin

Pepsinases from beef spleen, beef
kidney, and swine kidney

Trypsin

Tripsinases from beef spleen, beef
kidney, swine kidney, and
papain

Chymotrypsin

R

: I

. . CO—NH- CHK-CO^J^H- CH .

. CO—NH- CH- CO -^NH . . .

R

I
. . . CO—NH- CH • CO^NH . . .

HO.C6H4CH2...

or
CeHs-CHj...

NH2-CH2(CH2)a..

or
NHj.

>C.NH.(CH2)3

HOC6H4CH2...

or
C6H6CH2...

and to crystalline pancreatic carboxypeptidase. These are referred
to as "pepsinases," "leucine aminopeptidases," and "carboxypep-
tidases," respectively. Such groups of enzymes possessing identical
backbone and side-chain requirements are termed homospecific
enzymes (6) . On the other hand, two enzymes which diflfer from one
another with respect to their backbone or side-chain requirements,
or both, are designated heterospecific enzymes.

126

HYDROLYSIS OF PEPTIDE BONDS

To our knowledge, the proteolytic enzymes represent the first
class of enzymes for which liie property of homospecificity has been
demonstrated. It is not unlikely, however, that similar relationships
may exist in other classes of enzymes. Some years ago, the question
was raised whether the hydrolysis of the various /3-glucosides by emulsin
is to be attributed to a single iS-glucosidase or to several glucosidases
of slightly different specificity, and also whether the emulsins of various
plants contain identical or diff'erent /3-glucosidases. In particular,
Weidenhagen (33) advocated the theory that the emulsins from various
sources contain the same /3-D-glucosidase, and that this enzyme splits
not only all /3-D-glucosides containing various aglucones, but also all
oligosaccharides in which the sugar components are linked through
/3-D-glucosidic linkage. More recently, Pigman (28), in his classi-
fication of carbohydrases, suggested that the individual enzymes of
Weidenhagen's system be considered as classes of enzymes acting on the
same substrates but with different specificities. The finding of homo-
specific proteolytic enzymes now raises the question whether similar
groups of carbohydrases of identical specificity type exist. It may be
added that Bisseger and Zeller (13) have applied the concept of homo-
specificity to choline esterases obtained from various tissues.

Mechanism of Enzymic Proteolysis

It is inviting to speculate about the structural factors in the
enzyme which give rise to the phenomenon of homospecificity. It was
mentioned earlier that each proteolytic enzyme requires, for its action,
certain atomic groupings in the backbone of its substrates. This
backbone specificity may perhaps best be explained by the hypothesis
that each enzyme molecule contains an essential center, composed of
several distinct atomic groupings in a definite arrangement, and that
the first step of enzymic action consists in a combination of several
atomic groupings of the essential center of the enzyme with the indis-
pensable backbone groups o' the substrate. As in the Michaelis
concept of the enzyme-substrate compound, it is assumed that this
combination would result in the activation, and subsequent hydrolysis,
of an adjacent peptide bond of the substrate. To explain the homo-
specificity phenomenon, it seems necessary to conclude that the
enzymic action and its specificity originate from a rather restricted area

127

J. S. FRUTON

of the enzyme-substrate complex — the reacting nucleus — while the
rest of the complex may be expected to have some influence only on
the rate of the enzymic action . Two homospecific enzymes thus may
differ in many respects but are assumed to contain identical essential
centers, and therefore yield, with the same substrate, two enzyme-
substrate compounds containing identical reacting nuclei.

■ Support for this hypothesis comes from the finding that, if each
of two homospecific enzymes is allowed to act on two substrates, the
quotient of the rates of hydrolysis of each substrate by the two enzymes
is independent of the nature of the substrate. For example, papain
trypsinase was found to hydrolyze benzoylargininamide with a proteo-
lytic coefficient* of 167, while beef spleen trypsinase splits the same
substrate with a coefficient of 8.3. The quotient of these two values is
20.1. If the same enzymes are tested with benzoyllysinamide as the
substrate, 78 and 3.8 are the coefficients found, giving a quotient of
20.5. Similar examples may be cited for other pairs of homospecific
enzymes; for further data, cj. Bergmann (6). The fact that two homo-
specific enzymes, Ei and Eo, give similar proteolytic quotients for the
hydrolysis of the substrates Si and S2, may be explained by the assump-
tion that the enzyme substrate compounds, EiSi, E2S1, E1S2, and
E2S>, all contain identical reacting nuclei.

Furthermore, if this concept is correct, it may be expected that
parts of the enzyme molecule may be split off" without destruction of
the enzymic activity, and without alteration of the enzymic specificity.
Indeed, Kunitz (21) has shown that a-chymotrypsin may be trans-
formed into 7-chymotrypsin, and that in the course of this transforma-
tion about one-third of the a-chymotrypsin molecule is removed.
However, neither the proteolytic activity toward proteins nor the
specificity of action on synthetic substrates was altered.

It may be added that both a- and 7-chymotrypsin have been
found to exhibit two distinct proteolytic specificities, one of the amino-
peptidase type and another of the proteinase (endopeptidase) type
(18). Since all efforts to alter the ratio of the two specificities were
unsuccessful, and also since a- and 7-chymotrypsin both conform to the

* The proteolytic coefficient (C) is defined as the value of the reaction
velocity constant (A") for the hydrolysis of a peptide bond in the presence of an
amount of enzyme corresponding to one milligram of protein nitrogen per cubic
centimeter of the test solution.

128

FrvDRor^vsis of pf.ptidf: bonds

phase rule crilcria lor a pure protein (15), it wonUI .ippcar lli.K each of
these two proteins exhibits more than one distinet enzymic speeifiriiy.
The twofold specificity of the chymotrypsins may he ex|)lainecl on the
basis of the hypothesis presented above by assuming that each of the
protein molecules contains two distinct and different essential centers.

The hypothesis of the predominant role of the essential center
as against that of the remainder of the enzyme molecule cannot be
considered the same as the well-known theory of Willstalter (34), who
assumed that an enzyme molecule consists of a colloidal carrier and a
prosthetic group, the latter being responsible for the enzymic activity
and specificity. On the basis of this theory, Kraut distinguishes be-
tween the pheron and the agon as the two components of enzymes
(20). Numerous workers refer to the prosthetic group as the co-
enzyme and the carrier as the apoenzyme (1), th6 dissociable flavo-
proteins frequently being regarded as the prototype of this postulated
dual structure. It should be recalled, however, that the flavin part
of the flavoprotcins usually represents one of the partners in the chemi-
cal reaction and that the essential catalytic activity resides in the
protein moiety; cf. also Parnas (27). In the case of proteolytic
enzymes that do not require activation by sulfhydryl compounds
(pepsin, trypsin, etc.), no evidence of a dual structure is available.
For the activatable proteolytic enzymes, our present knowledge indi-
cates that the active enzyme represents a dissociable combination of a
protein with one of several activators (19). However, in the case of
these latter enzymes, it cannot be claimed that the protein part of the
enzyme acts merely as a colloidal carrier for another active part of the
enzyme. On the contrary, it is the protein part of the activated
enzyme which contains the essential center and which determines the
specificity. No proteolytic enzyme is known in which the nature of
the activator determines the specificity type of the enzyme.

Antipodal Specificily of Proteolylic Enzymes

The majority of the known proteolytic enzymes of higher plants
and animals has been found to be adapted to the hydrolysis of sub-
strates in which the essential side chain belongs to an /-amino acid.
For example, chymotrypsin endopeptidase rapidly hydrolyzcs the
substrate benzoyl-/-tyrosylglycinamide, but does not hydrolyze bcnzoyl-

129

J. S. FRUTON

^-tyrosylglycinamide (9). The hypothesis of the essential center
offers an explanation for such antipodal specificity. In order to act

HO-

-/ N— CH2 H H CH2— / V-OH

^ ^ \ / \ / ^ ^

C G

/\ < E /\ < E

HN CO HN CO

i I II

OC NH OC NH

/ \ / \

/-antipode (/-antipode

Chymotrypsin

I

upon the substrate, the enzyme mast approach the substrate closely
so that the groups in the essential center of the enzyme can combine
with the indispensable groups in the backbone of the substrate. If we
consider the tetrahedral arrangement of the groups about the asym-
metric carbon atom, it becomes evident that the enzyme must approach
the substrate from the right side in order to combine with the essential
backbone groups of the substrate. In the case of benzoyl-af-tyrosyl-
glycinamide, the side chain prevents the enzyme from approaching the
backbone groups, and therefore the enzyme cannot split the substrate.

When this theory of antipodal specificity was first proposed
(5), it was suggested that the size of the side chain of a'-alanine should
be sufficiently small so that the combination of the essential center of
the enzyme with the backbone groups of the substrate would not be
prevented completely, but would be made more difficult. This steric
hindrance should result in a retardation of the enzymic action on
peptides of (/-alanine as compared with its action on the /-form. In
fact, it was found that crystalline pancreatic carboxypeptidase was
able to hydrolyze carbobenzoxyglycyl-i^-alanine, albeit much more
slowly than the /-form (8).

It must be emphasized that the above conclusion can be correct
only when it has been established that it is the same enzyme which
hydrolyzes the two substrates containing d- and /-alanine. This res-
ervation applies particularly to earlier experiments in which a study
was made of the antipodal specificity of so-called dipeptidase and of
aminopeptidase (12), since the existence of individual dipeptidases as
well as the homogeneity of "aminopeptidase" have become doubtful.

130

HYDROLYSIS OF PEPTIDE BONDS

Finally it should be mentioned that recent years have witnessed
the discovery, in e\'er-increasing number, of proteolytic enzymes
adapted to the hydrolysis of peptides of rf-amino acids (3,25,29). The
existence of such enzymes presents us with an interesting problem:
if the antipodal specificity of a peptidase has its basis in the nature of
the essential center, then the essential center of a ^-peptidase should
possess a configuration antipodal to that of the corresponding /-pepti-
dase; cf. also Lettr6 (22).

Role oj Proteolytic Enzymes in Peptide Synthesis

In recent years, especial emphasis has been given to the dynamic
character of protein metabolism. The studies of Whipple (24), Schoen-
heimer (30), and others have given dramatic evidence for the view
that, in the tissues of animals and plants, protein molecules are rapidly
and continuously broken down and new protein molecules built up.
The recognition of the "dynamic equilibrium" of proteins in vivo has
brought to the fore the question of the nature of the enzymes that
selectively catalyze the sequences of chemical reactions in protein
synthesis and breakdown. More particularly, much attention has
been given to the nature of the enzymes that micdiate the synthesis of
peptide bonds between the individual amino acids. One view, which
is held \videly, is that the biosynthesis of peptide bonds is catalyzed
by the same enzymes that are responsible for the cleavage of peptide
bonds; in other words, biological peptide synthesis is thought to repre-
sent a reversal of the degradative action of the proteolytic enzymes.
The tissue proteolytic enzymes which are presumed to perform in vivo
synthesis are those frequently designated "cathepsins" (in the case of
animal tissues) and "papainases" (in the case of plant tissues).

The view that the hydrolytic action of proteolytic enzymes
might be reversed was advanced by several workers at the start of the
century and was later championed by Wasteneys and Borsook (32).
More recently, it was shown unequivocally (7) that numerous proteo-
lytic enzymes can catalyze, in model experiments, the synthesis of
peptide bonds. One of many examples of such synthesis is the catalysis,
by activated papain, of the following reaction: benzoyl-Z-leucine +
/-leucinanilide -^ benzoyl-/-leucyl-/-leucinanilide.

The fact that, in model experiments of this type, compoimds
of known and relatively simple structure are involved, in contrast to

J. S. FRUTON

the heterogeneous character of the protein hydrolyzates employed by
previous workers, has permitted a closer study of various factors which
play a role in the synthesis of peptide bonds by proteolytic enzymes.
The most important of these factors are: (a) the specificity of the
enzyme action; (b) the role of activators in the enzyme action; and
(c) the energy relationships involved in peptide synthesis.

With respect to the specificity of peptide synthesis, it may be
sufRcient, at this point, to recall that the action of a proteolytic enzyme
is to catalyze the attainment of equilibrium between a peptide and
its hydrolytic products. Consequently, one should expect the speci-
ficity of synthesis to be the same as that of hydrolysis. Indeed, it has
been found experimentally that, if the chemical nature of one of the
groups near a peptide linkage is altered so that a given enzyme no
longer is able to hydrolyze that linkage, then a similar structural
change in the components for the enzymic synthesis will also prevent
the formation of the peptide.

As is well known, several of the intracellular proteolytic enzymes
of animals and plants require, for their full catalytic activity, the
addition of sulfhydryl compounds as activators. The view was ex-
pressed some years ago (26) that, by oxidation of the sulfhydryl groups
of a proteolytic enzyme into disulfide groups, the enzyme would cause
peptide synthesis instead of hydrolysis. Experiments with model
substrates soon showed, however, that the activation requirements
were the same for the synthetic as for the hydrolytic reaction (7).

Turning now to the energy relationships involved in peptide
synthesis, we should recall that the hydrolysis of peptide bonds in
proteins and peptides proceeds spontaneously in the presence of a
suitable enzyme and that the equilibrium which is established is very
far on the side of hydrolysis. Thus, in order to reverse the hydrolytic
reaction, energy is required. Borsook (14) has calculated, from
thermal data, that the energy needed for the synthesis of a peptide
bond is approximately 3000 calories per mole. This energy may be
obtained in a variety of ways. In the case of the synthesis of benzoyl-
leucylleucinanilide, mentioned earlier, the driving force for synthesis
comes from the removal, by crystallization, of the synthetic product
from the solution. In order to restore the balance of the equilibrium
reaction, synthesis occurs, which, in turn, causes more of the synthetic
product to crystallize.

132

HYDROLYSIS OF PEPTIDE BONDS

Another mechanism for favoring the formation of peptide bonds
IS coupling the synthetic reaction with another cnergy-yiekHng chemical
reaction. At the present writing, few experimentally demonstrated
examples of such coupling can be cited (4). Surely the careful study
of the coupling between peptide synthesis and energy-yielding systems
represents one of the most interesting directions for future research;
cj. Bergmann and Fruton (11).

Before leaving the question of the role of the proteolytic en-
zymes in peptide synthesis, it should be emphasized that the available
experimental knowledge does not yet permit the conclusion that re-
versal of proteolysis is actually the process employed in biological
systems for the synthesis of peptide bonds. It is well to remember that
in the metabolic transformation of the polysaccharides, for example,
synthesis is not effected by the reversal of the hydrolytic action of the
amylases, but rather through a different chemical pathway, namely,
the synthetic action of the phosphorylases (16). The intervention of
phosphate in the biosynthesis of peptides also has been suggested (23),
but no experimental evidence for this view has as yet been brought
forward. Other speculations concerning the biological mechanisms
for peptide synthesis have been advanced, largely on the basis of
in vitro reactions (11). Clearly a greater fund of data on coupled
reactions in metabolic systems is required before it will be possible to
decide which, if any, of these theories is correct.

Perhaps the strongest reason for assuming, as a working hy-
pothesis, the view that proteolytic enzymes do play an important
role in protein synthesis is the fact they are the only known biocatalysts
which, by virtue of their sharp specificity, could direct, precisely and
reproducibly, the coupled sequence of successive peptide syntheses
required for the formation of a protein. The considerations concerning
specificity which have been discussed earlier in this article cannot fail
to modify our picture of the possible role of the proteolytic enzymes
in the biological synthesis of proteins. Until a few years ago, intra-
cellular proteolytic enzymes, such as papain or cathepsin, were re-
garded either as single enzymes or mixtures of very few enzymes. On
this basis it was concluded that the specificity of a single enzyme can
predetermine the molecular pattern of a protein. Thus it was assumed
that the specificity range of an intracellular proteinase would be suffi-
ciendy broad to comprise all the peptide bonds present in a protein

J. S. FRUTON

molecule. The demonstration of the extremely precise side-chain
specificity of the proteolytic enzymes suggests that the synthesis of a
protein from amino acids or small peptides could be accomplished only
by the cooperative successive action of many enzymes of different
specificity. In offering this suggestion, it must be emphasized again,
however, that the view that the proteolytic enzymes mediate the bio-
synthesis of proteins is not supported as yet by unequivocal experi-
mental evidence. There can be no doubt that the efforts of numerous
biochemists will be directed in the future to the elucidation of this
fundamental problem.

References

(1) Albers, H., Angew. Chem., 49, 448 (1936).

(2) Balls, A. K., and Kohler, F., Ber., 64, 34 (1931).

(3) Bamann, E., and Schimke, O., Naturwissenschqften, 29, 558 (1941).

(4) Behrens, O. K., and Bergmann, M., J. Biol. Chem., 129, 587 (1939),

(5) Bergmann, M., Science, 79, 439 (1934).

(6) Bergmann, M., in Advances in Enzyniology, Vol. II. Interscience, New
York, 1942, p. 49.

(7) Bergmann, M., and Fraenkel-Conrat, H., J. Biol. Chem., 119, 707
(1937).

(8) Bergmann,M.,andFruton,J.S.,J.5to/.CAd'm., 117,189(1937).

(9) Bergmann, M., and Fruton, J. S., J. Biol. Chem., 124, 321 (1938).

(10) Bergmann, M., and Fruton, J. S., in Advances in Enzymology, Vol. I.
Interscience, New York, 1941, p. 63.

(11) Bergmann, M., and Fruton, J. S., Ann. N. T. Acad. Sci., 45, 357 (1944).

(12) Bergmann, M., Zervas, L., Fruton, J. S., Schneider, F., and Schleich,
H., J. Biol. Chem., 109, 325 (1935).

(13) Bisseger, A., and Zeller, E. A., Helv. Physiol. Pharm. Acta, 1, C86 (1943).

(14) Borsook, H., and Dubnoff, J. W., J. Biol. Chem., 132, 307 (1940).

(15) Butler, J. A. V., J. Gen. Physiol., 24, 189 (1940).

(16) Cori, C. F., Biol. Symposia, 5, 131 (1941).

(17) Euler, H. v., and Josephsohn, K., Z- physiol. Chem., 162, 85 (1926).

(18) Fruton, J. S., and Bergmann, M., J. Biol. Chem., 145, 253 (1942).

(19) Irving, G. W., Fruton, J. S., and Bergmann, M., J. Biol. Chem., 139,
569 (1941).

(20) Kraut, H., and Pantschenko-Jurewicz, W. v., Biochem. Z-> 275, 114
(1924).

(21) Kunitz, M., J. Gen. Physiol., 22, 207 (1938).

HYDROLYSIS OF PKFriDn BONDS

(22) Lettr^, H., Angew. Chem., 50, 581 (1937).

(23) Lipmann, F., in Advances in Enzymology, Vol. I. Intcrscicncc, New
York, 1941, p. 154.

(24) Madden, S. C, and Whipple, G. H., Physiol. Revs., 20, 194 (1940).

(25) Maschmann, E., Biochem. Z-, 313, 129 (1942).

(26) Maver, M. E., and Voegtlin, C, Enzymologa, 6, 219 (1939).

(27) Parnas, J., Am. Rev. Soviet Med., 1, 485 (1944).

(28) Pigman, W. W., J. Research. Natl. Bur. Standards, 30, 257 (1 943).

(29) Schmitz, A., and Merten, R., Z- physiol. Chem., 278, 43 (1943).

(30) Schoenheimer, R., Dynamic State of Body Constituents. Harvard Univ.
Press, Cambridge, 1942.

(31) Smith, E. L., and Bergmann, M., J. Biol. Chem., 153, 627 (1944).

(32) Wasteneys, H., and Borsook, H., Physiol. Revs., 10, 110 (1930).

(33) Weidenhagen, R., Ergeb. Enzymjorsch., 1, 205 (1932).

(34) Willstatter, R., Ber., 59, 1 (1926).

135

70

METABOLIC PROCESS
PATTERNS

FRITZ LIPMANN, research chemist, Massachusetts general

HOSPITAL, boston; RESEARCH FELLOW IN BIOCHEMISTRY AND SURGERY,

HARVARD MEDICAL SCHOOL

IVe have hitherto failed in our comprehension oj life
mainly because we have been involved in tlie absolute
method of dealing with things.

E. NOBLE (14)

THE CEASELESS occurrence of metabolic processes in a
living cell has long been understood to imply at large a need
of energy for maintenance of active life. There was, however, and still
is, only a vague realization of the tasks for which uninterruptible flux
of energy is needed. A first opening here appeared when, through
fuller chemical resolution, recently, reaction chains unfolded which,
rather unexpectedly, were found to involve a multitude of substances
containing phosphate in peculiar linkages. When the way in which
these phosphate intermediates are manipulated in the cell was inidcr-
stood, it became possible to see clearly the connection between phos-
phate cycles and transformation and transport of energy. \n all cells
studied, a chemical network of energy distribution, the adenylic acid
system, was found to be present and able to carry in the form of special
energy-rich phosphate bonds standard portions of energy, amounting
to about one-fiftieth of that liberated by total combustion of a mole of

FRITZ LIPMANN

carbohydrate. Therewith the view developed that catabolism con-
sists to a considerable extent of a conversion of potential energy of food-
stuffs into directly utilizable phosphate bond energy (7), and that,
through alternate attachment and release of energy-rich phosphate
bonds, catabolism and anabolism are knit together into a largely re-
versible reaction continuum.

This new appreciation of aspects of the metabolic apparatus
which have hitherto been well concealed is beginning to affect our
general attitude toward problems of metabolic chemistry. The more
we recognize transformation of energy as a primary problem in meta-
bolic processes, the more are we compelled to treat metabolic proce-
dures for what they really are, namely, technical devices. In detail,
the manner in which the living organism solves the problem of energy
conversion is rather different from the technological methods employed
by man. But whether in the case of the organism or man, the ultimate
objective of energy conversion is the generation of energy in a utilizable
form. A fundamental analogy appears, indeed, between the in-
creasingly close dependence of our own daily life on electric current,
gas pipes, and a variety of motors and that of our body cells on food
and oxygen. In both instances the supply of energy is necessary to
maintain an organization, although most of the energy is ultimately
dissipated in the form of heat.

In many respects a living cell is comparable to a chemical
factory. The design of chemical factories, from the standpoint of a
technologist, is based on a variety of technical principles (5). Only
the process proper remains chemistry, but its technical execution is
effected wholly by physicomechanical devices. This predominantly
mechanical manipulation of unit processes represents a most significant
difference between organismic chemistry and chemistry practiced by
man, for, in living cells, both process design and process execution are
based on chemical principles. Instead of the material being manipu-
lated successively in spatially separated compartments, cellular chem-
istry involves a harmonious series of consecutive reaction steps which
are brought about on a molecular scale by a host of catalysts, all present
together in the same reaction fluid. This difference in type of operation
tends to obscure the basic analogy of both procedures.

Most metabolic processes classify among what the chemical
engineer calls "flow processes" (5), that is, procedures whereby streams

138

METABOLIC PROCESS PATTERNS

of material enter and leave a reaction system uninterruptedly. The
technical flow process is based on flow charts which map the route
along which a compound is driven through a series of operations
"An ideal flow process is characterized by steady states of flow, tem-
perature and composition at any point of the process." This charac-
terization holds likewise for almost any metabolic process.

Process Characteristics of a Fermentation

The physicomechanical environment in which we live has in-
fluenced our thinking to the point at which we must overcome certain
mental inhibitions in order to comprehend the almost exclusive reliance
of the living organism on chemical operations. This is particularly
the case with processes of power generation which we habitually
associate with highly mechanical machinery, though most of the
power ultimately derives, as in our bodies, from chemical combustion
of carbon and hydrogen. There has been an additional, more inci-
dental, obstacle to a ready understanding of biochemical energy
transformation. In fermentation— we are just gathering the elemen-
tary facts — human interest has centered long on manufacturing aspects,
like the production of alcohol and other valued substances. From the
biological point of view, a fermentation or a respiration is designed to
produce power and the nature of the end product is more or less second-
ary and accidental.

In the simpler forms of anaerobic carbohydrate utilization,
e. g., lactic and alcoholic fermentation, the mapping of the sequence
of reactions is now completed. But it will take some time until their
pattern and design are duly comprehended. However, life and
multiplication of a large variety of organisms are maintained exclu-
sively through fermentative transformation of energy, frequently
involving simple organic and nitrogenous compounds as starting
materials. Therefore principles derived from the chemical mechanics
of fermentation allow a fair amount of generalization.

In scheme I, the simplest fermentative process, the conversion
of glucose into lactic acid and phosphate bond energy is represented.
To emphasize process characteristics, the now rather well-known inter-
mediaries are omitted in the scheme; the flow chart represents the
reaction sequence:

FRITZ LIPMANN

hexose/2 *- ^^ph
ph-glyceraildehydc + ph
ph-glyceryl'^ph — 2 H —*■ '^ph
ph-glycerate — H2O
ph'^'enolpyruvate —*■ '^ph

pyruvate -\r 1H

lactate

The terms '~ph, -ph, and ph characterize, respectively, the energy rich
phosphate bond (12 kcal.), the ester phosphate bond (3 kcal.) and inorganic
phosphate (7).

Scheme i
The Process Pattern of Lactic Acid Fermentation

HEXOSE

* ^ , )k — r^

12 Kcal.(~ph)
ad

12 Kcal.(~ph)
LACTIC ACID-^

The flow line represents a projection into space of the catalytic
pathway a hexose molecule travels to reach the inert end product,
lactic acid. Initial fission in the middle of the six-carbon chain seems
to be prompted by the introduction of phosphate groups at both ends.
This phosphorylation is a rather costly investment absorbing just one-
half of the gross yield of energy and thus reducing the net yield by
about fifty per cent. An initial investment of part of the ultimate
energy yield in the operation of the process is a notable feature. It is
this need of induction energy which makes the fermentative process
autocatalytic. The misleading statement is often made that, in fer-
mentation, one half of the hexose molecule oxidizes the other half,
suggesting a dismutative process. What happens, rather, is that a
hydrogen donor, after unloading of phosphate bond energy, is trans-
formed into a hydrogen acceptor. This manner of manipulation is
expressed in the characteristic shape of the flow lines which, in all
fermentations, fold back on themselves. The bending back, to accept
a pair of hydrogens released in a previous stage on the molecular flow
line, together with the initial expenditure of energy to start the process,
may be considered as general and dominant characteristics of anaerobic
metabolism.

140

METABOLIC PROCESS PATTERNS

Technologically there are numerous disadvantages in operation
of the anaerobic type of metabolism. Most prominent is low efficiency.
The probable upper limit of the gross yield is about ten per cent,
whereas ninety per cent of the potential energy of combustion remains
unused in the waste products. A piling up of waste products, fre-
quently of strongly acidic character, presents a further serious technical
problem. In the more highly organized living systems, therefore, we
find the aerobic type predominant. If in the higher organisms we
meet, as we do sporadically, a well-developed system of anaerobic
energy conversion, it is in places or stages of development at which,
for structural or topographical reasons, the oxygen supply is poor or
unsafe: in the embryo, in cancer tissue, in parts of the placenta, parts
of the retina, in muscle, etc. An anaerobic energy supply grants
greater independence (8).

Process Characteristics of Respiration

The introduction of oxygen as hydrogen acceptor increases
considerably the complexity of the energy-yielding process. Our
present insight in this case is spotty and far removed from the com-
pleteness achieved in understanding simpler anaerobic fermentations.
An attempt has been made here to coordinate the available data into
a coherent process scheme, and at the same time to point out those
stretches in the flow lines for which information is still missing.

When, as in Figure 1 , the progressive catabolism of a substrate
molecule in the manner of a flow chart is projected onto an energy-
time coordinate system, some representative features emerge. The
particulars of this scheme refer to degradation of half a glucose unit
through the citric acid cycle. The gross energy available from this
process, calculated by summation of the areas above the six consecutive
steps of dehydrogenation is 6 X 57, = 342 kcal., a quantity practically
identical with the theoretical yield for carbohydrate combustion. The
dehydrogenation potentials of the intermediaries oscillate almost
symmetrically around the hydrogen potential.

Contrary to present convention, the hydrogen potential at pM 7 is made
the reference potential, which coincides with the potential of the "average" respira-
tory hydrogen donor. By plotting in a conventional manner oxidation-reduction
potentials, E'o, as the difference between the normal potential of the system at
pH 7 and the potential of atmospheric hydrogen gas at pH 0, a meaningless zero

141

FRITZ LIPMANN

line cuts arbitrarUy through at the succinate/fumarate potential not far below the
middle between hydrogen and oxygen potentials at pH 7.

0,

+1.23 r

+ 0.42

+ 0.2

"5

>

0
-0.2

\ \

|2H |2I

HC:0 + Spo

HCOH

HzCOpo"

COa

+
COOH

CH2
CH2

COOH

C02

+

COOH

COOH COOH

CH2
CH2

_^ CH H C:0
CH oPoi/^CHj

COOH

COOH _^COOH

Fig, 1. — Citric Acid Cycle.

The dotted lines mark off the constantly repeating process unit. Each
turn — from condensation to oxalacetate regeneration — oxidizes a two-carbon unit
of carbohydrate level. The two-carbon unit is fed into the system in the form of
acetate radicals.

Oxidation-reduction potential, volts>

Absolute potential

Hydrogen donator"

Water
system"

Phosphate
sy3tem<^

difference between oxygen
and water system, volts*

Phosphoglyccraldehyde

Pyruvate

Isocitrate

Ketoglutarate

Succinate

Fumarate-malate

-0.1

-0.35

0.13

-0,35

0.43

0.25

0.2
-0.05

-0.05

0.55

1.33 volts

1.58

1.10

1.58

0.80

0.98

Sum 7.37 volts
Theory 7.46'*

{Contintud on following page)

METABOLIC PROCESS PATTERNS

Unfortunately, all schemes following the well-established rule of assigning
to oxidation-reduction potentials values increasingly positive toward oxygen depict
respiratory processes ambiguously. In respiration, the chemical potential gradually
diminishes, of course, toward oxygen, being eventually spent with oxygen reduction,
when the listed potential attains the most positive value.

As a characteristic of the process pattern, a separabihty of two
main flow lines appears. The line representing catabolism of the
substrate glides along on an approximately equipotential path; and
at right angles to it the hydrogens which have been released by de-
hydrogenation journey to oxygen. So far, the greatest progress has
been made in the field of substrate catabolism in which workable
schemes have emerged. Schemes like that of the citric acid cycle,
however, do not supply information about the chemical pathways of
respiratory transformations of energy beyond the stage of hydrogen
donation. The pathway of the substrate supplies merely the level
from which electrons are emitted, loaded with potential energy and
ready to be used. We learn, thus, from our map that the potential
difference between oxygen and each of the six dehydrogenation steps
which sum up to complete oxidation of a triose averages very closely
to the value of the oxyhydrogen potential. In other words, carbo-
hydrate reacts grossly like a mixture of carbon dioxide and molecular
hydrogen:

(CHOH-HsO), = (C02-2H2), (1)

At first approximation, it seems justified therefore to consider
the respiratory process as a repeating series of rather uniform process

" For isocitric and ketoglutaric acids, tentatively, the potentials of hydroxybutyric (4) and
pyruvic acid (11), respectively, were used here. For other values of the oxidation-reduction potentials,
cj . Green (4).

ft Reference potential: hydrogen electrode at /)H 7; c} . page 141.

c The terms, water system and. phosphate system, refer to the hydrated and the corresponding
phosphorylated double bonds, as, for example, in phosphoglyceraldehyde hydrate and phosphoglycer-
aldehyde phosphate (9). The difference of the oxidation-reduction potentials between the water series
and the phosphate series is approximately constant and equal to the volt equivalent of the energy-rich
phosphate. For acetyl phosphate, recently, a bond energy of approximately 15 kcal. was calculated
(10, 121 which corresponds to roughly 0.3 v. This value is appropriate for calculations primarily con-
cerned with bond generation. Th# average energy is somewhat lower, 12 kcal. or 0.25 v., which value
is preferred for turnover calculations. In cases in which more or less arbitrarily the actually reacting
dehydrogenation system is assumed to be of the phosphate type, the connecting line is drawn through
the phosphate system. In these cases, a vertical line in the graph connects the potential points of water
and phosphate system, indicating the energy transformation. Tht wide empty area above the line con-
necting the potentials indicates the large part of the energy which remains here unaccounted /or {cJ. Fig. 2).

<* Calculated from the combvistion heat of one-half mule of glucose, 343 kcal. The agreement
between the rather roughly approximated voltage and the combustion heat is noteworthy. To be
accurate, 0.25 v., the equivalent of the nonoxidative phosphate bond in phosphopyruvate, should be
added to the sum of the oxidation-reduction potentials.

FRITZ LIPMANN

miits. Sucli a uiiil is essentially an oxygen hydrogen cell. This unit
is further broken down in the scheme of Figure 2, which may be repre-
sentative of the respiration not only of carbohydrate but also of other
substrates.

1.2

-50

1.0-

0.8-

0.6

04

0.2

o

>

0

CYTOCHROMES

-30

-FLAVOPROTEINS,

10 PYRIDINE
NUCLEOTIDES

o

2H

~POi-

PO^

•po;

ADENYLIC
ACID

METABOLITE
Figure 2. — ^Transformation of Electron Potential into Phosphate Bond Energy.

This graph accounts for the energy deficit which appeared in Figure 1
as empty space above the substrate flow line. A more detailed picture is ob-
tained by fitting into the detours cycles of the type depicted in scheme II,
page 146.

Projection of the potential gradient onto a space scale enables us
to fix approximately on the map those regions where transformation of
electron potential into phosphate bond energy is to be expected. It
is known from a gross comparison between oxygen consumption and the
resulting phosphorylation (1,6,15) that, by transport of one pair of hy-
drogen electrons from substrate to oxygen, 3+ energy-rich phosphate
bonds may be generated. Energy-rich phosphate bonds average
12 kcal. per bond, which is equivalent to a span of 0.25 v. for a two-

144

METABOLIC PROCESS PATTERNS

electron system. The available potential between oxygen and a pair
of average substrate hydrogens is 1.2 v., a potential span which may
accommodate theoretically 1.2 divided by 0.25, or 4+ energy-rich
bonds. This calculation fixes the upper limit of yield and shows the
experimental values to be well within this limit.

In the particular scheme of Figure 2, a generation of the three
phosphate bonds established by experiment is rationalized by merely
cutting out, from the potential gradient, three 0.25-v. portions in
succession. Such a mapping leads to the conclusion — unambiguously,
it would appear — that at potential levels of around +0.1, +0.5, and
+0.9 v., with reference to the hydrogen electrode of pH 7, the pair
of hydrogen electrons is intercepted three times in succession by
chemical devices which transform catalytically 0.25-v. portions into
energy-rich phosphate bonds. This breaks the process of hydrogen
transfer up into three smaller units; and here we probably meet the
smallest units of the catalytic system designed for respiratory trans-
formation of energy. The three, or perhaps four, transformers which
are built into the pathway of the hydrogen electrons should be con-
sidered as the actual power generators of the living organism. It is
very significant that these transformer systems appear largely inde-
pendent of the particular hydrogen donor. Such operation of sub-
strate-independent catalysts for transformation may explain how
phosphate bonds are generated in a constantly increasing variety of
oxidations, such as sulfur oxidation in Thiobacillus (16), the oxyhy-
drogen reaction (3), and fatty acid oxidation (13). To summarize,
we may say that generation of phosphate bonds, regardless of the type
of hydrogen donor, may be represented by the following equation:

XH2)

or > + 3 HO-POa" + 3 ad-H + O2/2 >

X-HaO)

or > + 3 ad ~ PO3— + 4 H2O;

xo)

average AFo, (3 X 12) - 57 = -21 kcal. (2)

Equation (2) shows that a participation of phosphate and adenylic
acid frequently is a reflection of general hydrogen transfer catalysis
rather than a particular case of dehydrogenation.

FRITZ LIPMANN

The chemical nature of the transformer catalysts which operate
on three 0.25-v, spans between the limits 0 and -j-1.2 v. remains to be
considered. In analogy to what is known about the carbonyl com-
pounds, which in part could have the function of transformers on the
lowest potential level (9), the following generalization may be illus-
trative: The catalyst should offer opportunity to a phosphate molecule
to add onto a double bond. From the addition product a pair of
electrons is removed wherewith the energy-rich phosphate bond is
generated. This step should involve little or no energy loss. Ex-
change of adenylic acid hydrogen for the phosphate group now de-
livers 12 kcal. into the cell, and the residual product may be rehydro-
genated. This phase involves the refund of the delivered 12 kcal.
(0.25 v.) and must ultimately lead to a regeneration of the double
bond. Tentatively, the potential level around 0.1 v. may be assigned
to a carbonyl double bond. For the subsequent level around 0.5 v.
a C:C double bond would be suitable. Finally, a double bond of the
ascorbic acid type appears as a possibility for the upper level. A trans-
forming unit of the type outlined is charted arbitrarily for a C: C double
bond in scheme II.

Scheme ii
A Catalytic Transformer of Electron Potential into Phosphate Bond Energy

O/R-CATALYST
HC-H I

"2'-' ^ • V 3 -^H SYSTEM

/ V /^ Eo= 0.45 V.

HCH HC H-ad

I II +

HCOH CO-POJ-

WATER
SYSTEM

Eo=0.2V. ..Zfll u A ^ UTILIZA-

METABOLITE HjC ad ~- P0i"-^-TION

t

C:0 = 0.25 V.

With slight modification it is applicable to any double bond of a C:X
type, X being, for example, :0 or :NH. The essential feature of the system is
a shuttling between addition of water and of phosphate to the double bond.
Starting from the bottom and moving clockwise, we distinguish four steps of
the cyclic process. First, by hydrogenation of the water system the transformer is
loaded with about 0.25 v. per mole, at least. Second, the reduced water

146

METABOLIC PROCESS PATTERNS

system is then converted into a phosphate system by exchange reactions.
Third, in the key reaction, a dehydrogenation of the phosphate system transforms
with httle loss electron potential into phosphate bond energy. And fourth,
the bond equivalent of 0.25 v. is unloaded to adenylic acid, thus returning
the system to the original state.

In scheme III, a condensed scheme of respiration is drawn.

Scheme hi
Summarizing Flow Scheme of Respiratory Energy Turnover

0;

'Ph

-ph

5
o

o

a

LU

»-~ph-

SUBSTRATE CATABOLISM

In the space projection the two energy fields occupy two dimensions,
the third being assigned to the equipotential flow of primary metabo-
lites. In reality, these fields of electron and phosphorylation po-
tential are not homogeneous, but are canalized through chemical
specificity of hydrogen and phosphate transfer, with enzyme specificity
acting in the manner of connection plugs.

The underlying theme of this article is a reminder that, having
pulled apart the chemical continuity of the living organism, we are
challenged to reintegrate the scattered pieces into a whole. "There
is more in a transition than a series of states or possible cuts, more in
a movement than a sequence of positions or possible stops. We have

147

FRITZ LIPMANN

to place ourselves along the transition and, from within, to cut across
in thought, in order to appreciate the successive states (2)."

References

(1) Belitzer, V. A., and Tsibakova, E. T., Biokhimiya, 4, 518 (1939).

(2) Bergson, H. L., Creative Evolution. Modern Library, New York, 1944.

(3) Gaffron, H., J. Gen. Physiol., 26, 241 (1942).

(4) Green, D. E., Mechanisms of Biological Oxidations. Cambridge, Univ.
Press, London, 1940.

(5) Hougen, O. A., and Watson, K. M., Chemical Process Principles. Wiley,
New York, 1943.

(6) Kalckar, H., Biochem. J., 33, 631 (1939).

(7) Lipmann, F., in Advances in Enzymology, Vol. L Interscience, New
York, 1941, p. 99.

(8) Lipmann, F., "Pasteur effect" in A Symposium on Respiratory Enzymes.
Univ. Wisconsin Press, Madison, 1942.

(9) Lipmann, F., "Biological oxidations and reductions," Ann. Rev. Biochem.,
7, 1 (1943).

(10) Lipmann, F., J. Biol. Chem., 155, 55 (1944).

(11) Lipmann, F., and Tuttle, L. C, J. Biol. Chem., 154, 725 (1944).

(12) Meyerhof, O., Ann. N. T. Acad. Sci., 45, 357 (1944).

(13) Munoz, J. M., and Leloir, L. F., J. Biol. Chem., 147, 355 (1942).

(14) Noble, E., Purposive Evolution. Holt, New York, 1929.

(15) Ochoa, S., J. Biol. Chem., 151, 493 (1943); 155, 87 (1944).

(16) Vogler, K. G., and Umbreit, W. W., J. Gen. Physiol., 26, 157 (1942).

148

11

BIOCHEMISTRY
FROM THE STANDPOINT

OF ENZYMES

DAVID E. GREEN, chief of the enzyme research laboratory

DEPARTMENT OF MEDICINE, COLLEGE OF PHYSICIANS AND SURGEONS,
COLUMBIA university; PAUL-LEWIS AWARD FOR ENZYME CHEMISTRY

/:

'T WOULD be in the nature of a platitude to say that there
is hardly a branch of biochemistry which cannot be analyzed
or at least interpreted in terms of enzymes or enzymic phenomena.
Yet few would maintain that enzymes represent more than an intel-
lectual liqueur in the teaching of biochemistry in our graduate schools.
Most modern textbooks of biochemistry often treat the subject of en-
zymes much in the manner that feathers and scales are dealt with in
textbooks of anatomy. If this were merely another instance of the lag
of textbooks behind developments in research, there would be no cause
for concern. But more appears to be involved than the traditional
lag. The implications of enzyme chemistry have yet to be more
generally understood; and until that time arrives the textbooks will
continue to regard enzymes as chemical oddities, if not to ignore them
altogether.

During the past fifty years, biochemistry has developed from the
ugly duckling of physiology to a science in its own right. During this
period, interest has been focused largely on methodological and struc-
tural problems: how to estimate and what the constitution is of the
innumerable compounds which make up the living cell. In this phase
in the development of biochemistry, the analytical chemist and struc-

D. E. GREEN

tural organic chemist played the leading roles, and indeed they laid
the foundations for an exact science. Hence, it is hardly surprising
that dynamic problems such as those of intermediary metabolism were
approached more from the direction of what may be called chemical
morphology than from the point of view of physiology. Perhaps the
best way of illustrating the point is to recall the sensation which was
produced by Schoenheimer's (5) early isotope experiments. His con-
ception of an organism as a chemical system in a constant state of flux,
in dynamic as opposed to static equilibrium, bore the same relation to
the classical biochemical conception that the Schrodinger-Heisenberg
conception of the atom bears to the rigid atom of the late 19th century.
This is not to imply that the groundwork for Schoenheimer's concep-
tion was not already laid in the literature. The students of enzymes
have long been aware of reversible equilibrium systems; but biochem-
ists generally were unable to project the implications of these reversible
systems in terms of intermediary metabolism. The outstanding re-
searches of Schoenheimer and Krebs have done much to orient research
on intermediary metabolism along more peculiarly functional lines;
but, nonetheless, not more than the fringe of biochemical thinking has
been disturbed. Intermediary metabolism is still being taught and
discussed without any reference to the catalysts responsible for each of
the transformations. Discussing the behavior of a car without refer-
ence to the motor represents an analogous situation. This state of af-
fairs provides some reason for attempting an interpretation of biochem-
istry in terms of enzymes and enzymic phenomena. What follows will
be rather sketchy, not only because of the limitations of space, but also
in some cases because of the inadequacy of available information.
However, the purpose of this essay is more to show that enzymes pro-
vide a logical and rational approach to many fields of biochemistry
and medicine rather than to attempt a comprehensive survey of the
enzyme field.

Living systems carry on their activities by virtue of myriads of
chemical reactions which collectively are referred to as intermediary
metabolism. Physiological functions such as growth, reproduction,
secretion, nerve conduction, muscular contraction, etc., are integra-
tions of whole series of chemical events in intermediary metabolism.
These chemical events with few exceptions are not spontaneous proc-
esses. They require the presence of highly specialized protein cata-

150

ENZYMES

lysts known as enzymes. Apparently, there is a different enzyme for
practically every reaction, or at least every reaction type of intermediary
metabolism. This can only mean that some several thousand en-
zymes must exist. For a long time fears were expressed that the
cramped quarters of the average small sized cell could not possibly
accommodate so many different enzymes. But that bogey has been
laid low by the isolation from one small cell such as the yeast cell of
literally hundreds of different enzymes. Of course, not all types of
cells have the same complement of enzymes. The number and amount
of enzymes vary from one cell type to another, and in fact determine
the individuality of each cell.

There are many instantaneous ionic reactions which occur in
the course of intermediary metabolism and which do not require
enzymes, e. g., neutralization of acids or bases, and deposition of salts
such as calcium phosphate. But even in the province of reactions
which occur spontaneously, a surprise is occasionally in store. Thus,
the decomposition of carbonic acid into carbon dioxide and water is
catalyzed by a special enzyme known as carbonic anhydrase, which can
speed up the rate of the reaction far beyond the spontaneous rate.

The study of intermediary metabolism represents one of the
oldest lines of biochemical investigation. It is not surprising, there-
fore, that of the total number of reactions known to occur in inter-
mediary metabolism only a very small proportion have been recon-
structed with isolated enzyme systems. In fact, whole chapters of
intermediary metabolism such as the metabolism of steroids, porphyrins,
carotenoids, sulfur compounds, bile acids, fatty acids, etc., are prac-
tically virgin territory as far as knowledge of enzymes is concerned.
Enzyme chemistry has made its greatest strides in the field of fermenta-
tion or glycolysis of sugar. Here, the entire process from glucose to
glycogen or from glycogen to lactic acid has been reconstructed in
vitro with some twenty odd enzymes, each prepared in pure or largely
pure state. This in vitro reconstruction is not to be regarded as a
stunt or merely as a triumph of biological engineering. In order to
effect a successful reconstruction, it becomes necessary to understand
precisely the way in which certain enzyme systems are linked together
and the way in which different chemical reactions are synchronized.
A successful reconstruction, therefore, implies mastery of most of the
chemical details and a complete knowledge of the constituent enzyme

D. E. GREEN

systems. But the success in the field of glycolysis has been to some
extent at the expense of progress in other fields.

Nutrition in essence deals with the relative amounts and nature
of the materials which have to be supplied ultimately to the enzyme
systems, and with the resyntheses and replacement of enzymes. In
other words, the study of nutrition and the study of enzymes represent
two sides of the same coin. Since the nature, number, and amounts of
enzymes vary from one living system to another, the nutritional problem
varies in the same way. If we knew all the enzymes present in a par-
ticular organism and the special components present in each of the
enzymes, theoretically, we would have all the necessary data for
determining the complete nutritional requirements. But since we are
largely in the dark about the vast majority of enzymes, we use the data
and knowledge of nutrition to ferret out information about enzymes.
It has long been known, for example, that traces of certain metals
such as manganese, iron, zinc, copper, and magnesium are essential
in the diets of many animals. The indispensability of these metals
in the diet has been correlated with the presence of these metals as
structural elements of important enzyme systems. Thus, manganese
has been shown to be an essential component of arginase; magnesium
an essential component of carboxylase, chlorophyll, etc.; zinc of car-
bonic anhydrase; copper of phenolases; and iron of catalase, per-
oxidase, cytochromes, and lactic dehydrogenase. Traces of cobalt
are known to be essential in the diets of sheep particularly. No doubt
cobalt also will be identified as an essential component of some as yet
unknown system. Among plants, boron and molybdenum are es-
sential trace elements, and one must presume special enzyme systems
in plants requiring these elements. The fact that, in most instances,
small quantities of the metals are necessary correlates with the extraor-
dinary activity of enzymes at high dilutions. In other words, only
traces of metals are necessary for incorporation into the enzymes, since
the enzymes also occur only in trace amounts.

The identification of the P-P factor with nicotinamide, of vita-
min Bi with thiamin, of vitamin B2 with riboflavin, and of vitamin
Be with pyridoxal provides a moral for those who prefer to study one
side of a coin without reference to the other. Cozymase, a dinucleotide
of nicotinamide and adenine, has long been studied as the coenzyme
of fermentation since its discovery by Harden and Young (4) in 1906.

ENZYMES

Yet it was not until 1937, some three years after Warburg had shown
nicotinamide to be an essential component of the coenzyme, that
Elvehjem and his colleagues (1) established a connection between the
antipellagra vitamin and nicotinamide. Elvehjem's discovery was
consistent with poetic justice because he had been trained both as a
nutritionist and as an enzyme chemist. Vitamin B2 was identified
in 1935 by Kuhn and Karrer with riboflavin, the prosthetic group of
the so-called yellow enzyme which Warburg had isolated from yeast
three years earlier. The time relations were reversed in the cases of
vitamins Bi and Eg, since the identification of the vitamins preceded
knowledge of their participation in enzymic reactions. The chemical
identification of vitamin Bi with thiamin by Williams and Cline in
1936 preceded by one year the demonstration by Lohmann and
Schuster that the prosthetic group of yeast carboxylase is a diphos-
phoric ester of thiamine. Peters and his group at Oxford had estab-
lished the role of vitamin Bj in the oxidation of pyruvic acid long before
the chemical nature of the vitamin was established. Vitamin Be was
identified with pyridoxine in 1938 by Folkers, Keresteszy et al., and
Kuhn et al., but it was not until 1944 that a more active form of the
vitamin, viz-, pyridoxal, was discovered by Snell and that the phos-
phoric ester of pyridoxal was shown to be the coenzyme of tyrosine
decarboxylase by Gunsalus. There are thus four authenticated identi-
fications of vitamins with prosthetic groups. In the case of vitamin
A, Wald has shown that it is an essential part of a photosensitive
pigment in the eye known as visual purple. Many will not concede
that visual purple has the properties of an enzyme but, whether or not
that point is conceded, it is at any rate admissible that vitamin A
fulfills the role of prosthetic group of a chromoprotein with an important
physiological function.

This relation between vitamins and prosthetic groups makes
it easy to understand the basis for the body's continuous requirement of
vitamins. Since enzymes have a limited life period in consequence
of their destruction during activity, there is constant need for more of
all enzymes. The minimum amount of any of the vitamin compatible
with viability is therefore an approximate measure of the total amount
of enzymes in the body whose prosthetic groups contain that vitamin.
Here, again, if we knew precisely which enzymes, for example, require
flavin, what the normal levels of these enzymes are in different organs.

D. E. GREEN

and the rates at which they are synthesized, we would have all the
necessary data for determining the flavin requirements of that organ-
ism. There are much easier methods of getting at that data at the mo-
ment, but it is of considerable theoretical interest to arrive at the data
from the enzyme side.

There are some well-informed workers in the field of nutrition
who are unwilling to concede that all vitamins must have a catalytic
function. While they admit that the relationship has been established
in at least four and possibly six instances, they do not regard these
necessarily as precedents. In 1941 the enzyme-trace substance
theory (3) was developed which predicted that any substance necessary
in the diet in trace amounts must be an essential part of some enzyme
system. The theory was not meant to imply that substances required
in higher concentrations cannot be essential parts of enzymes. That
may or may not be the case. However, amounts of the order of 10
fjLg. per kg. per day were regarded as conclusive evidence of a catalytic
role for that substance. On the basis of the enzyme-trace substance
theory we may confidently expect in the near future the identification
of biotin, pantothenic acid, folic acid, vitamin A, vitamin K, and vita-
min D with essential parts of new prosthetic groups. The daily re-
quirements of vitamin C are about a thousand times greater than
those for most of the other vitamins, and, certainly, are well beyond
the trace level. The possibility is therefore open that vitamin C may
have no catalytic function whatsoever. The recent investigations of
Sealock on the role of vitamin G in the oxidation of tyrosine in the liver
however, do not encourage the view that vitamin G is an exception
to the vitamin-enzyme relation.

In recent years the exact nutritional requirements of many
bacteria and molds have been carefully investigated. A study of any
of the synthetic diets proves very instructive from the standpoint of
enzyme chemistry. The list of required substances can be easily
divided into two categories: (7) substrates of enzyme systems, e. g.,
amino acids, glutamine, dextrose, fatty acids, purines, etc.; and (2)
precursors of prosthetic groups, e. g., vitamins such as riboflavin,
thiamin, etc. or the building stones thereof, hemin, and trace metals.
The members of the first category can be classified not only on the basis
of what we know about their intermediary metabolism but also on
the basis of the amounts required, which are vastly in excess of the

ENZYMES

minimal amounts necessary of substances in the second category.
The concentrations of the first category are in milHgrams per cubic
centimeter, whereas those of the second are in fractions of a micro-
gram per cubic centimeter.

The enzyme-trace substance theory in the form stated above is
appHcable only to naturally occurring substances in the diet or growth
medium. There is an alternative form of the theory which applies
to all substances regardless of their origin, and it may be stated in the
following terms: Any substance which in trace amounts induces pro-
found biological effects does so either by participating in or by spe-
cifically afifecting some enzyme system. If valid, this theory must be
regarded as one of the cornerstones of the science of pharmacology. An
extensive list can now be compiled of substances, part if not all of whose
pharmacological actions can be explained in terms of enzyme effects
{cj. Table I).

Table I

Pharmacological agent

Enzyme inhibited by the agent

Fluoride

Enolase

One of the anterior pituitary

Hexokinase

hormones

Cyanide

Cytochrome oxidase

Eserine

Choline exterase

Prostigmine

Choline esterase

Chlorinating agents

Triosephosphoric dehydrogenase

lodoacetic acid

Triosephosphoric dehydrogenase

Benzedrine

Amine oxidase

Phlorhizin

Glucose phosphorylase

Gramicidin

One of the fermentation enzymes involved in

phosphorylation

Enzyme identical with the agent

a-Toxin of Clostridium wdchii

A lecithinase

Lytic factor of cobra venom

A lecithinase

Spreading factor

Hyaluronidase

One of the toxins of CI. welchii

CoUogenase

One cannot but be impressed by the vindication of the enzyme
hypothesis in at least fourteen instances, most of them reported within
the last few years, in contrast to the absence of a single authenticated
case in which any other principle of mechanism has been shown to
operate. At present, the enzyme theory is really not much more than
a good working hypothesis. But while it may be premature to regard
all the biological eflfects of trace substances exclusively in terms of

D. E. GREEN

cnzymic effects at tiie jneseiit time, there is n(j alternative explanation
that merits serious consideration. Alternative explanations usually
amount to substituting an obscure phrase for an obscure phenomenon.
Thus, some pharmacologists talk about trace substances upsetting an
"active patch" of some important cellular membrane and thereby
exerting their action. When interrogated about the properties of the
"active patch," the pharmacologist usually admits that he has in mind
some specific combining group and in effect admits a somewhat
watered-down version of an enzyme reaction. Certainly there is no
way of testing the "active patch" hypothesis as commonly stated nor
is there any evidence that it has been productive either in explaining
or predicting new phenomena. One cannot resist the conclusion that
the "active patch" concept is a terminological device for cloaking
ignorance. Another variant of the "active patch" concept is the so-
called "active surface" which is sensitive to pharmacological agents
and which controls certain key biological functions. The effect of
agents on these surfaces is said to be exclusively a physical one, /. e.,
the "active surface" becomes covered by the pharmacological agent
and consequently is inactivated. The extraordinary specificity of
pharmacological effects and the high dilutions at which these effects
occur render this interpretation in purely physical terms unlikely.

The discovery by Woods (6) that the antibacterial action of the
sulfonamides could be explained in terms of the resemblance of the
sulfonamides to ^-aminobenzoic acid was an important milestone in
our understanding of the mechanism of the action of drugs. It became
at once clear that some cnzymic process was at the bottom of the chemo-
therapeutic action of the sulfonamides. There is still no clue as to the
nature of this enzymic process but there can be little doubt that the
process is enzymic. /^-Aminobenzoic acid is a naturally occurring sub-
stance in yeast and animal tissues. Many bacteria and molds are
unable to grow unless it is present in the medium. The trace con-
centrations in which it must be present for optimum growth exclude
all but a catalytic role. In fact, /^-aminobenzoic acid has been shown
to exist in yeast largely in the form of a polypeptide, which may well
be its active catalytic form in the intact cell. One theory of sulfon-
amide action assumes that the sulfa drugs displace /^-aminobenzoic
acid from its combination with specific proteins and thereby inactivate
enzymes important in the growth of certain microorganisms. In

156

ENZYMES

other words, the natural substance, yz.c-,/'-aminobenzoic acid, competes
with the drugs for the protein partner with which it forms the enzyme
complex. The degree of inhibition is determined by the relative con-
centrations of /^-aminobenzoic acid and drug and by their relative
affinities for the protein partner. The phenomenon of competitive
inhibition is well known in the enzyme literature and is perhaps the
most characteristic hallmark of enzymic phenomena.

The sulfonamides inhibit the growth of susceptible bacteria
only when they are actively growing. Once active growth has taken
place, the sulfonamides are without effect even on susceptible bacteria.
This observation suggests that the sulfonamides interfere with the
process of synthesizing the /^-aminobenzoic acid enzyme complex
rather than with the activity or function of /i-aminobenzoic acid in its
final catalytic form. Unquestionably, there are many alternative
pathways in bacteria by which /^-aminobenzoic acid can be coupled
with other substances to form the final catalytic complex. Only one
pathway may be sulfonamide-sensitive, and only those bacteria which
share that method of synthesizing the /)-aminobenzoic acid catalytic
complex will be susceptible to the action of the sulfonamides. These
considerations must be borne in mind in evaluating the enzymic theory
of sulfonamide action. Ghemotherapeutic drugs may interfere either
with the working of a key enzyme or with some stage in the process
by which the key enzyme is synthesized. Since the synthesis of en-
zymes is also enzymic in nature, the primary action of chemothera-
peutic drugs must still be considered as one of interference with
enzymes.

The interpretation in terms of enzymes of the mode of action
of sulfonamides is by no means in general currency. The so-called
"essential metabolite" theory of Fildes (2) has gathered many adherents
and is generally accepted among workers in the field of chemotherapy.
This theory assumes that/^-aminobenzoic acid is an essential metabolite
for the growth of certain microorganisms rather than a part of an
essential enzyme system. A metabolite is usually defined as a sub-
stance which undergoes chemical transformation; and the term is
usually applied to substances like amino acids, fatty acids, etc. which
are present in considerable concentration, and which are degraded or
converted into more complex substances by enzyme systems. Since
the amount of /^-aminobenzoic acid present in microorganisms is

^57

D. E. GREEN

scarcely detectable by the most delicate chemical methods, p-am'ino-
benzoic acid can hardly be classified as a typical metabolite. Quite
clearly, the term metabolite as applied to /^-aminobenzoic acid must
imply merely that it is involved in metabolism. Furthermore, since
traces of essential substances are known to participate in metabolism
only in the capacity of catalysts, the "essential metabolite" theory boils
down to a disguised enzyme theory. In the present state of ignorance,
there is some merit in talking of /)-aminobenzoic acid as an essential
metabolite until such time as its precise enzymic role is clarified.
But when the essential metabolite theory is seriously proposed as an
alternative to the enzyme theory, it becomes important to recognize
what the concept of essential metabolite really means. As applied to
/>-aminobenzoic acid, "essential metabolite" is just a term of caution
to indicate by implication a catalytic role, without stating it in so many
words. But as occasionally happens with terms of caution, their neu-
trality defeats their purpose. Essential metabolite has become con-
fused by many with metabolite, and its real significance has become
lost among those who see only the letter and not the spirit of the term.

The relation between the sulfonamides and j&-aminobenzoic
acid provided a blueprint for designing other chemotherapeutic
agents. Every vitamin or prosthetic group theoretically should find
its nemesis in some antivitamin. So the hunt began, and not without
success. The sulfonic acid analogue of pantothenic acid (pantoyl-
taurine) was found to inhibit the growth of some bacteria which re-
quired pantothenic acid. The ratio of pantothenic acid to the sulfonic
acid analogue determined whether growth or inhibition would take
place. Much of the same kind of results were obtained with pyri-
thiamin, as antithiamin agent, with pyridine /3-suIfonic acid, as anti-
nicotinic acid agent, etc. None of these antivitamins is comparable
to the sulfonamides in their efficiency as chemotherapeutic agents;
but in these antivitamins, at least, we have the hopeful beginnings of
a rational program of chemotherapy.

From the standpoint of enzymes, chemotherapy would appear
to be the science of compounds which go for the enzymic Achilles'
heel of an infectious organism without at the same time damaging the
host unduly. In other words, the objective is first to find an enzyme •
which is present or important in the infectious organism and not in
the host, and second to find a drug which specifically inhibits this

158

ENZYMES

enzyme. Clearly, little progress will be made along rational lines in
chemotherapy unless progress in the enzyme chemistry of infectious
organisms is stimulated. Many are not convinced that progress in
chemotherapy can come from that direction. "Suppose," they say,
"you find the enzymic Achilles' heel, how are you going to find the drug
to inhibit that enzyme?" As an example of the direction from which
the solution might come, the chemical nature of the prosthetic group
or active groups of the enzyme which turns out to be the weak point
might provide the necessary clue for the synthesis of specific inhibitors.
None will maintain that the solution will be easy, but experience has
taught us that the possibilities of solving a problem are greatly increased
when the nature of the problem can be accurately defined.

There is a school of thought well represented in our large phar-
maceutical firms and also in the councils of our government scientific
agencies which prefers to advance chemotherapy exclusively by the
method of trial-and-error organic synthesis. In effect, the program
followed is merely that of permutation and combination of the few
effective chemotherapeutic agents we have as models. Instead of
looking for new models, the old models are varied over and over again.
The very limited success of this program of chemotherapy is hardly
surprising. In the first place, not all pharmacologically active sub-
stances tolerate any considerable structural change. Thus, no one
has been able to prepare a more active form of the vitamin than
thiamin. In fact, the slightest alteration of the molecule involves
partial or complete loss of activity. The same considerations apply
to the vast majority of biologically active substances such as acetyl-
choline, histamine, flavin, ascorbic acid, etc. It is certainly true in
other instances that a given effect is produced by large numbers of sub-
stances sharing a common structure, e. g., adrenalin, and the hundreds
of adrenalin-like bases which have been tested, sulfanilamide and the
hordes of other sulfa drugs, etc. But one must keep the objective
clearly in mind. In the case of adrenalin, the analogues merely
imitate the adrenalin eff"ect. They accomplish nothing that adrenalin
cannot do. Their virtue lies either in their greater stability or in their
greater resistance to deterioration in the animal body. All the known
sulfa drugs act in exactly the same way, though they are effective at
different concentration levels. The same organisms can be inhibited
both by the weakest and strongest sulfa drugs provided the weakest is

D. E. GREEN

sullicionlly s(jlul)k\ In othci- words, no new principle emerges. A
more effective sulfa drug docs not extend the range of action of sulfa
drugs in the sense of inhibiting organisms which are otherwise insensi-
tive to the sulfa drugs. From the standpoint of enzyme chemistry,
the direction which much of chemotherapy research has taken does
not appear to be either profitable or rational. Chemotherapy cannot
be attacked intelligently without a detailed knowledge of intermediary
metabolism and enzyme chemistry. We may make allowances for
the element of urgency in wartime, but after the war, there ought to
be a better balance between the sums spent on sheer trial-and-error
organic synthesis and the sums spent on fundamental investigations.

Woolley (7) has pioneered in providing a framework for a
rational pharmacology based on the antivitamin concept. He showed
that certain antivitamins can produce a state of avitaminosis, in some
cases in a matter of hours, merely by displacing the vitamins competi-
tively from their catalytic complexes. Since profound pharmaco-
logical effects attend the syndrome of avitaminosis, antivitamins have
to be regarded as potential pharmacological agents. Thus, pyri-
thiamin rapidly induces the disorders of the central nervous system
which are characteristic of thiamin deficiency. The lesion is of course
righted at once by addition of large enough amounts of thiamin. Since
the quantitative importance of the catalytic reaction in which vitamins
participate varies depending upon the organ or part of the organ,
it does not follow that all antivitamins will exhibit similar pharmaco-
logical effects. On the contrary, it would appear that each anti-
vitamin would selectively poison only a particular portion of the nervous
system as well as only particular organs. A complete series of anti-
vitamins should provide a wide range of specific pharmacological
agents, all of which are reversible by addition of the vitamins which
they imitate. The beginnings in this new field of exploration are still
modest but the horizons seem immense.

The hormones represent a class of substance which, according
to the enzyme-trace substance theory, ought unequivocally to qualify
as enzymes or essential parts of enzymes. Yet no one has conclusively
demonstrated that any one of this large class is either an enzyme or an
essential part of an enzyme. Do we have in this class a notable excep-
tion to the theory? There is no basis for answering this question defi-
nitely one way or the other. There is a possibility that renin, the

1 60

ENZYMES

hormone elaborated by I lie kidney, hydrolyzes hypertensinogen, one
of the plasma proteins, with formation of a pressor substance. If this
possibility is confirmed, we would have the first identification of a
hormone with an enzyme function. Whatever the uncertainty about
hormones as enzymes, the evidence leaves no doubt that hormones
influence enzymic phenomena, e. g., the dramatic efTect of insulin* or
adrenalin on carbohydrate metabolism. In practically every instance,
hormone action has been boiled down to the regulation of some phase
of intermediary metabolism. Do hormones regulate by being enzymes
themselves, by influencing enzymes, or perhaps by controlling the
synthesis of enzymes? There are many indications which point to
some of the hormones controlling the synthesis of enzymes. But since
the synthesis of proteins represents one of the most obscure corners of
enzyme chemistry, there is little hope of any early clarification of the
precise role which hormones might play in synthesis.

Oddly enough, our most precise knowledge of the way in which
the synthesis of enzymes is regulated has been acquired from the field
of genetics. The experimentation which has led to this knowledge
constitutes one of the most brilliant chapters of modern biology.
Geneticists have succeeded in demonstrating that single genes control
the syntheses of single enzymes. This implies that genes, like hor-
mones, are regulators of intermediary metabolism. Genes accom-
plish this regulation by controlling the synthesis of enzymes, whereas
hormones operate in ways as yet not classified. Genes and hormones
are distinguishable in another fundamental respect — hormones must
be synthesized by the respective endocrine glands, while genes are
autocatalytic and hence self-perpetuating. This autocatalytic property
of the gene resolves the dilemma that, if enzymes are needed to synthe-
size other enzymes, there must then be an infinite series of enzymes
making enzymes. But recognition of autocatalysis as a phenomenon
is a far cry from understanding the mechanism. As a matter of fact,
guesswork constitutes the sum and total of our knowledge of the way
in which certain protein molecules are able to reproduce themselves.

* When this essay was in the proof stage, Cori and his group announced the
epoch-making discovery that insuHn reverses the inhibition of hexokinase produced
by one of the anterior pituitary hormones. In these two instances, at any rate,
hormones must be regarded as regulators of enzymes by virtue of their inhibiting
or releasing the inhibition of key enzymic processes.

i6i

D. E. GREEN

The problem of the autocatalysis of genes is essentially similar
to that of viruses. What is the starting material for the synthesis and
how is the synthesis accomplished? No concrete answer is possible
as yet, but there are indications that the solution to the problem will
be in terms of enzyme chemistry. One intriguing development in
recent years has been the successful application of the concept of com-
petitive inhibition drawn from the field of enzymes to the virus problem.
Two strains of the same virus are found to antagonize one another's
growth in the same host, presumably by competing for the same
pabulum. In chemotherapy, two substrates, one natural, the other
"fraudulent," compete for the same enzyme. In the example above,
two viruses are made to compete for the same substrate. This virus
interference phenomenon occurs only when the two strains are closely
related, just as competitive inhibition occurs only when the
"fraudulent" substrate closely resembles the natural.

The past decade has seen not only the extension of our knowl-
edge of enzymes to other fields of biochemistry and medicine but also
the extension in part of the enzyme concept to noncatalytic proteins.
Let us compare, for example, two proteins, catalase and hemoglobin.
Both contain iron protoporphyrin as prosthetic group and both are
highly specific. Catalase catalyzes the decomposition of hydrogen
peroxide into oxygen and water, whereas hemoglobin combines re-
versibly with molecular oxygen. Catalase cannot function as hemo-
globin and conversely hemoglobin for all practical purposes does not
catalyze the decomposition of hydrogen peroxide. Catalase forms a
compound with hydrogen peroxide and then the enzyme-substrate
compound undergoes decomposition. This cyclical process repeats
itself more than two million times per minute at 0°. In a similar way,
hemoglobin forms a compound with molecular oxygen over a certain
range of oxygen tension, and the complex is dissociable by lowering
the oxygen tension. The speed of the combination is of about the same
order of magnitude as for the combination of catalase with hydrogen
peroxide. This comparison is not being made to infer that hemoglobin
is an enzyme. There is little to be gained by a redefinition of the
classical concept of an enzyme to include a proteinlike hemoglobin,
because after all there is a real distinction between a catalyzed reaction
and a noncatalyzed reaction. However, it is important to recognize
the many properties in common which catalase and hemoglobin share.

162

ENZYMES

While enzymes must be differentiated from noncatalytic proteins,
nonetheless a broader classification of proteins is conceivable in which
enzymes represent a special case of what we may call the functional
type of protein. We may define the functional protein as one which
performs a specific physiological function. Thus, catalase decom-
poses hydrogen peroxide; hemoglobin combines reversibly with mo-
lecular o.xygen; cytochrome C is reduced by the reduced forms of certain
enzymes and in turn its reduced form is oxidized by cytochrome
oxidase; prothrombin plays a specific role in blood clotting; visual
purple acts as a photoreceptor, etc. Limitation of space precludes
further development of the concept of functional proteins. Suffice
to say that, in the author's opinion, processes like those of blood co-
agulation, complement fixation, and antibody formation, are phe-
nomena which have much in common with enzymic phenomena, and
that the highly specific functional proteins responsible for these processes
have more in common with enzymes than with purely structural
proteins. The concept of functional proteins has the virtue of opening
new horizons in the form of novel types of proteins. Just as myosin is
a protein of muscle specialized to convert chemical energy to mechani-
cal energy or as visual purple is a protein in the retina specialized to
convert light energy, presumably, ultimately to the electrical energy
of nerve conduction, so there may be analogous proteins in nerve,
cellular membranes, etc. specialized to carry out the particular physio-
logical functions of these organs. The trend in biochemistry would
appear to be toward the inclusion of more and more proteins in the
category of catalysts. In fact, it is conceivable that eventually all
proteins apart from purely structural proteins will be found to perform
in a highly specific way some physiological catalysis, and the currently
prevalent idea of storage and inert proteins will soon be as outmoded
as the so-called endogenous nitrogen metabolism.

There is another aspect of myosin and visual purple that well
merits consideration. Biochemists have long exercised themselves
over the problem of the means by which the organism converts energy
from one form to another. In one or two instances, the curtain sur-
rounding these interconversions has been pierced. Myosin and visual
purple may well be considered as examples of energy transformers.
Thus, myosin in effect converts the chemical energy of hydrolysis of
adenosine triphosphate into mechanical energy. The protein itself

163

D. E. GREEN

acts as the transforming agent by combining two physiological func-
tions. In the same way, visual purple becomes a vehicle for trans-
forming light energy into chemical energy and, we must assume,
eventually reacts in some way with nervous elements of the retina.
Instances of enzymes with multiple functions and multiple active
groups have been known in the literature, but the tendency hitherto
has been to regard them as biological curiosities. Thus pyruvic
oxidase of Lactobacillus delbrueckii contains two prosthetic groups, viz-
flavin dinucleotide and diphosphothiamin. Milk flavoprotein con-
tains two prosthetic groups (flavin dinucleotide and another as yet
unidentified) and catalyzes the oxidation of purines, aldehydes, and
dihydrocoenzyme I. The /-amino acid oxidase of rat kidney has two
enzymic functions. These enzymes with multifunctions may be in-
volved in the transfer of chemical energy from exergonic to endergonic
processes. In other words, enzymes may ultimately turn out to be the
energy transformers and converters of the cell.

References

(1) Elvehjem, C. A., Madden, R. J., Strong, F. M., and Woolley, D. W.,
J. Am. Chem. Soc, 59, 1767 (1937).

(2) Fildes, P., Lancet, I, 955 (1940).

(3) Green, D. E., Advances in Enzymology, Vol. I. Interscience, New York,
1941, p. 177.

(4) Harden, A., and Young, W. J., Proc. Roy. Soc. London, B77, 405 (1906).

(5) Schoenheimer, R., The Dynamic State oj Body Constituents. Harvard Univ.
Press, Cambridge, 1942.

(6) Woods, D. D., Brit. J. Exptl. Path., 21, 74 (1940).

(7) Woolley, D. W., Science, 100, 579 (1945).

164

12

ENZYMIG MECHANISMS

OF CARBON DIOXIDE

ASSIMILATION

SEVERO OCHOA, assistant professor of biochemistry, new york

UNIVERSITY COLLEGE OF MEDICINE

7T IS well known that animal cells depend on a supply of
ready-made organic materials (such as carbohydrates,
fats, and proteins) or of their building stones (such as simple sugars,
fatty acids, and amino acids) for either building up cell substance or
replenishing their stores of energy-yielding foodstuffs. Green plants,
however, are able to fix atmospheric carbon dioxide under the in-
fluence of light and use it to synthesize the organic constituents
they need; by means of this process — photosynthesis — they "assimilate"
carbon dioxide. In its final over-all results, photosynthesis is essentially
a reversal of respiration: In respiration, foodstuffs are oxidized to
carbon dioxide and water with absorption of oxygen, the energy thereby
released being utilized by the cell to carry out its activities; in photo-
synthesis, the chlorophyll-containing chloroplasts utilize radiant energy
to build up organic substance from carbon dioxide and water, and
oxygen is liberated in the process.

Some microorganisms, such as green algae and both green and
purple bacteria, are photosynthetic. Certain bacteria do not possess
the capacity to utilize radiant energy for carbon dioxide assimilation
but are able to use, for the same purpose, the energy derived from
oxidation of inorganic substances like hydrogen sulfide, thiosulfate,

165

SEVERO OCHOA

sulfite, selenite, nitrite, elementary sulfur, ammonia, and molecular
hydrogen. This process is known as chemosynthesis.

Both photosynthetic and chemosynthetic organisms are referred
to as autotrophic because they can grow in media composed of in-
organic substances exclusively. However, a number of bacteria, in-
cluding most of the pathogenic species, can live only in media that con-
tain one or more organic components, and are known as heterotrophic
(24,25,28).

The importance of photosynthesis and, in general, of carbon
dioxide assimilation can hardly be overemphasized, since animals
depend for their subsistence on the materials formed through carbon
dioxide assimilation by autotrophic organisms. Thus, carbon dioxide
assimilation is one of the most fundamental of all life processes. Al-
though it was known for some time that heterotrophic bacteria and some
animal cells could utilize carbon dioxide to synthesize carbon-to-
hydrogen bonds or carbon-to-nitrogen bonds (as is the case in the syn-
thesis of formic acid from carbon dioxide and hydrogen by Escherichia
coli, or in the synthesis of urea by the liver), the belief was current that
such organisms lacked the capacity to utilize carbon dioxide for the
synthesis of carbon-to-carbon bonds until the pioneer work of Wood and
Werkman demonstrated that heterotrophic bacteria can fix carbon
dioxide in this manner (28). The process is now known to occur
also in animal cells.

The synthesis of organic material from carbon dioxide is an
endergonic reaction, which results in an increase of the free energy
content of the system, and thus requires energy in order to proceed.
In other words, such a synthesis must be coupled with exergonic* reac-
tions involving a decrease in free energy. For this purpose, photo- and
chemosynthetic organisms can use either radiant energy or the energy
derived from oxidation of inorganic compounds. Heterotrophs, on
the other hand, can assimilate carbon dioxide only at the expense of
oxidizing organic foodstuffs, so that no net gain in organic cell constitu-
ents can result from carbon dioxide assimilation under these conditions.
It is, therefore, difficult to decide whether carbon dioxide fixation is

* C. D. Coryell, in Science, 92, 380 (1940), introduced the terms "exergonic"
and "endergonic" to characterize negative and positive changes in free energy
(AF), respectively, and suggested that the use of the terms "exothermic" and
"endothermic" be restricted to designate changes in heat (AH).

1 66

CARBON DIOXIDE

essential or even important for heterotrophic organisms and animal
cells. One could conceive, however, that fixation might be used for
the synthesis of special cell constituents such as growth factors or the
like. It is well known that most heterotrophic bacteria require the
presence of carbon dioxide in the medium for optimal growth.

The cellular mechanisms of carbon dioxide fixation have been
obscure for a long time. It is only recently that light has been shed
on the mechanism by which fixation occurs in heterotrophic bacteria
and animal cells. As it now appears, the fundamental process in all
types of carbon dioxide fixation is a reversal of the decarboxylation of
some keto acids — a process catalyzed by enzymes. These reactions
are reversible, but their equilibrium lies very far to the side of decar-
boxylation, i. e., liberation of carbon dioxide. Thus, the problem faced
by the cell is to shift the equilibrium as far as possible in the opposite
or uphill direction, and this requires expenditure of energy.

It has been established (15) that the free energy change of a re-
versible chemical reaction is related to the equilibrium position in a
manner expressed by the equation:

^F = -RT In K
where AF represents the change in free energy (expressed in gram
calories) and K is the equilibrium constant. If the equilibrium con-
stant is expressed as:

(decarboxylation product) (CO2)
(carboxylated product)
The equilibrium constant of some of the reversible decarboxylations is
of the order of 10^, so that they proceed with a decrease in free energy
of about —4000 to —5000 calories. Here are included the enzymic
decarboxylation of oxalacetic acid to pyruvic acid and carbon dioxide,
and that of oxalosuccinic acid to a-ketoglutaric acid and carbon di-
oxide. In both cases the reaction involves a carboxyl in position
/3 relative to the carbonyl group; this reaction type will be referred to
here as j3-carboxylation.

There is another group of enzymic decarboxylations involving
simultaneous decarboxylation and dehydrogenation of a-keto acids
that proceed with a much larger decrease in free energy than do /3-
decarboxylations. Thus, the free energy change of reaction (1):
pyruvate- + 2 H2O <^ acetate" + HCO3 " + 3 H+ + 2 e (1)

167

SEVERO OCHOA

has been estimated to be —9400 cal., and that of reaction (2) —8000
cal. (8):

a-ketoglutarate + 2 H2O ^

succinate-- + HCO3- + 3 H+ + 2 e (2)

Reactions of this type are often referred to as oxidative decarboxylations
and the reverse as reductive carboxylations.

From the relationship between free energy change and equilib-
rium constant discussed above, it is clear that shifts of equilibrium
toward carboxylation, i. e., carbon dioxide fixation, can only be ac-
complished by an input of energy into the system. We shall consider
in some detail the enzymic mechanisms used by the cell for this
purpose.

The fundamental pattern of biological carbon dioxide fixation
can be visualized in terms of the following steps:

(7) Primary Fixation Reaction. Carboxylation. Since the equi-
librium of the reversible reaction involved is very unfavorable, only
small amounts of keto acid are formed at one time.

(2) Reduction. Step 1 is followed by enzymic reduction of the
keto acid to the corresponding hydroxy acid. This shifts the equilib-
rium and more keto acid can be formed by step 7.

(5) Reduction of the Pyridine Nucleotide Oxidized in Step 2. The
second and third steps will be discussed below.

(4) Dehydration, Hydration, Isomerization. Further equilibrium
shifts can be brought about by secondary enzymic transformations of
the hydroxy acid formed by step 2. Thus, the hydroxy acid may be
dehydrated to the corresponding unsaturated fatty acid. The latter,
in turn, may be hydrated in a different position of the molecule to form
a new hydroxy acid isomeric with the one first formed.

(5) Further Reduction. The unsaturated fatty acid formed in
step 3 may undergo further reduction to the corresponding saturated
acid.

All the reactions concerned in the various steps just outlined are
reversible.

The reduction of the keto acid in step 2 is catalyzed by a specific
pyridine nucleotide dehydrogenase. These dehydrogenases (27) con-
sist of a protein and a prosthetic group or coenzyme that combine with

168

CARBON DIOXIDE

one another, although the equiHbrium constant is preponderantly in
favor of dissociation. The active group of the coenzyme is pyridine.
Two such prosthetic groups are known at present, viz-, diphospho-
pyridine nucleotide (abbreviated DPN, also known as cozymase and
coenzyme I), and triphosphopyridine nucleotide (abbreviated TPN,
also known as Warburg's coenzyme and coenzyme II). The coenzymes
are dinucleotides, each containing two bases, adenine and nicotinamide,
two molecules of ribose, and either two (DPN) or three (TPN) mole-
cules of phosphoric acid. The structural formula of DPN is shown in
scheme I. The point of attachment of the third phosphoric acid
residue in TPN is as yet unknown. Most of the pyridine nucleotide
dehydrogenases have DPN as the prosthetic group; only two are defi-

Scheme

I

H

CH

1

/\

N-=C— N

HC C— CONH2

1 11

A i

c— c— c

1 1 1

\+/

N N NH2

N

CH

HP

HC

1

HCOH

HCOH

1 0

1 0

HCOH

HCOH

1

HC

0

0

1
HC

H2C— O— P— O— P— O — CHj

O- OH

Structural formula of diphosphopyridine nucleotide

CH
HC C— CONH2

HC CH

V

R— O— P— OR'

CH

/\
HC C— CONH2

II I

HC CHj

V

I H+O-

1 I

R— O— P— OR'

O O

Reversible reduction of diphosphopyridine nucleotide

169

SEVERO OCHOA

nitely known to function with TPN, and one at least can function with
either coenzyme (Table I).

Table I
Pyridine Nucleotide Dehydrogenases

Ox-redox

System

potential

(£6; PH
7.0), volts

Prosthetic
group

Occurrence of dehydrogenase

Glutamic acid<=ia-ketoglutaric acid +

NHi

-0,03

DPN or TPN

Yeast, bacteria, animal tissues

Malic acid<=i oxalacetic acid

-0.10

DPN

Bacteria, plants, animal tissues

Ethyl alcohol?^ acetaldehyde

-0.16

DPN

Yeast, bacteria

Lactic acid ?=i pyruvic acid

-0.18

DPN

Bacteria, animal tissues

3-Phosphoglyceraldehyde + phosphate

<=i 1,3-diphosphoglyceric acid

-0.28

DPN

Yeast, animal tissues

j3-Hydroxybutyric acid <=^ acetoacetic

acid

-0.29

DPN

Animal tissues

Isocitric acid «=* oxalosuccinic acid

-0.30

TPN

Yeast, plants, animal tissues

Glucose-6-phosphate ?^ 6-phosphoglu-

conic acid

-0.43

TPN

Yeast, red blood cells

Glucose <=^ gluconic acid

-0.45

DPN

Animal tissues

The pyridine in the coenzymes acts by cyclic addition and re-
moval of hydrogen (see scheme I). Through the reversible change
pyridine ^ dihydropyridine, it functions as a hydrogen carrier. The
protein component of the dehydrogenase, upon whose nature depends
the specificity for its substrate, binds both substrate and coenzyme, and
in this complex hydrogen from the substrate is transferred to the
pyridine of the coenzyme; the substrate is thus oxidized and the pyri-
dine reduced. This transfer of hydrogen is reversible, so that dihydro-
pyridine and oxidized substrate when bound by the protein can react
to form pyridine and reduced substrate. Thus, the action of a typical
pyridine nucleotide dehydrogenase can be represented as follows:

RHa + P-Go :^R + P-CoHa

where RH2 and R represent reduced and oxidized substrate and P • Co
and P-CoHa represent oxidized and reduced protein-coenzyme com-
plex, respectively.

The reduced coenzymes have a sharp absorption band at 340
m^, whereas there is no absorption at this wave length by the oxidized
form. This distinction is important because the action of the pyridine
nucleotide dehydrogenases can be conveniently followed by changes
in the absorption of light of wave length 340 myu (27).

170

CARBON DIOXIDE

It is thus dear that the biological reduction of the keto acid
formed in step 7 of carbon dioxide assimilation requires the presence
of the specific dehydrogenase and of the reduced form of its prosthetic
group. Step 2 can then be represented as follows:

keto acid + P • C0H2 t± hydroxy acid + P • Co (3)

Since the complex formed by the coenzyme with the protein
component of the dehydrogenases is dissociable to a relatively large
degree, there are always small amounts of free coenzymes present in the
cell. When equilibrium has been reached in a reaction of the type
represented by reaction (3), it will stop unless provision is made for a
reduction of the oxidized coenzyme formed so as to displace the
equilibrium to the right. Step 3 occurs here. It is carried out through
the action of another dehydrogenase which functions with the same
prosthetic group as that acting in step 2. Such a reaction may be, for
instance:

hydroxy acid a + PrCo ?:± keto acid a + PrCoHa (4)

and is possible because of the dissociable nature of the complex formed
by the coenzyme with the protein components of the dehydrogenases,
so that the coenzyme can alternatively be bound by either protein.
Thus, some of the oxidized coenzyme dissociating from P • Co can be
bound by the second protein, Pi, to form Pi -Co, as in reaction (4).
Since the coenzyme oxidized in reaction (3) is reduced in reaction (4),
the net result of these two reactions can be expressed by:

keto acid + hydroxy acid a <=± hydroxy acid + keto acid a (5)

This type of reaction is known as a coenzyme-linked dismutation, and
is, in general, reversible. The extent to which it will proceed in a given
direction depends on the equilibrium constants of the two dehydrogen-
ase systems involved and on the concentration of reactants. Thus, in
our case, a high concentration of hydroxy acid a will favor carbon
dioxide fixation.

We shall now consider the individual carbon dioxide fixation

systems.

^-Cavhoxylaiion

Carbon dioxide fixation by a-ketoglutaric acid. Although
this type of carbon dioxide fixation is the most recently discovered

171

SEVERO OCHOA

(20,21), it will be convenient to discuss it first because the methods
used in the study of this system permit a clearer picture of the pat-
tern into which the cellular mechanisms of carbon dioxide fixation
can be fitted. The primary fixation reaction involves the reversal of
the decarboxylation of oxalosuccinic acid to a-ketoglutaric acid and
carbon dioxide (reaction la). This reaction is catalyzed by an

OXALOSUCCINIC CARBOXYLASE

COOH COOH

CO

CO

CH2 + C02 —

^ HC COOH

CH2

1

CH2

1

COOH

COOH

a-Ketoglutaric acid

Oxalosuccinic acid

(la)

enzyme, oxalosuccinic carboxylase, present in heart muscle and prob-
ably in other animal and plant tissues. The enzyme requires either
magnesium or manganese ions for activity.

The equilibrium of reaction la is so far to the left that the avail-
able analytical methods would fail to show a formation of oxalosuccinic
acid even when starting with very high concentrations of a-ketoglutaric
acid and carbon dioxide in the presence of the enzyme. However,
reversibility can easily be demonstrated by adding isocitric dehydro-
genase (see Table I) and reduced triphosphopyridine nucleotide.
When this is done, the oxalosuccinic acid formed by carboxylation is
reduced to /-isocitric acid by TPNH2 which, in turn, is oxidized to
TPN (reaction lb):

COOH

I
CO

ISOCITRIC DEHYDROGENASE

COOH

HC— COOH

CH2

I
COOH

Oxalosuccinic acid

+ TPNH2

CHOH

HC— COOH -f TPN

I
CH2

I
COOH

/-Isocitric acid

(lb)

Reaction lb is the second step of the series by which carbon
dioxide is fixed in this system. The combined result of reactions la
and lb is reaction Ic.

172

CARBON DIOXIDE

OXALOSUCCINIC CARBOXYLASE AND ISOCITRIC DEHYDROGENASE

COOH COOH

i I

CO CHOH

CH2 + CO2 + TPNH2 , HC— COOH + TPN (Ic)

CH2 CH2

I I

COOH COOH

a-Ketoglutaric acid /-Isocitric acid

Since both reactions lb and Ic involve conversion of TPNH2
to TPN and vice versa, they can be followed spectrophotometrically in
either direction by allowing the reaction to take place in a quartz cell
and measuring the absorption of light at wave length 340 m^ by the
test solution. As mentioned above, reduced pyridine nucleotides
strongly absorb light of this wave length. The molar extinction coef-
ficient, which is defined by the equation:

log hi I
a. —

cl

is 0.5644 X 10^ (cm. -/mole). For a transmittance of 95% and when
/ = 1 cm., the concentration of reduced pyridine nucleotide would be
0.04 X 10"'' moles per cc. In the case of TPN with a molecular
weight of 743, 0.04 X 10"^ moles per cc. corresponds to 3 fxg. of TPN
per cc, or 0.8 Mg- of isocitric acid per cc. This indicates the great
sensitivity of the optical method and how suited it is for the study of
reactions of this type.

By using this method, it has been possible to determine the
equilibrium constants of reactions la, lb, and Ic. The equilibrium
constant of reaction lb, A"b = (/-isocitrate) (TPN)/(oxalosuccinate)
(TPNH2), at/?H 7.0 and 22° C, is approximately 0.3. That of reac-
tion Ic, K\ = (/-isocitrate) (TPN) /(a-ketogIutarate)(C02)(TPNH2),
at the same pH and temperature is, on the average, 1.3 X lO"*. The
equilibrium constant of reaction la can be calculated from these two
values, since fi\ = (oxalosuccinate)/(a-ketoglutarate)(C02) =
Kc/Kx, = 0.5 X 10^^ Thus, tlie equilibrium of reaction la is so un-
favorable for carbon dioxide fixation that by this step alone only about
0.5% of the a-ketoglutarate would be carboxylated.

SEVERO OCHOA

The TPN formed in reaction Ic can be reduced through a
coenzyme-linked dismutation as discussed above. This results in the
shifting of the equilibrium toward the side of carbon dioxide fixation.
Such a shifting has been accomplished with the glucose-6-phosphate
dehydrogenase system (see Table I) which catalyzes reaction Id. The
combined result of reactions Ic and Id, when a mixture of glucose-6-

GLUCOSE-6-PHOSPHATE DEHYDROGENASE

glucose-6-phosphate + TPN ^

6-phosphogluconate + TPNH2 (Id)

phosphate, a-ketoglutarate, and carbon dioxide is incubated with
glucosephosphate dehydrogenase, oxalosuccinic carboxylase, isocitric
dehydrogenase, manganese ions, and TPN, is reaction le:

a-ketoglutarate + CO2 + glucose-6-phosphate <=^

/-isocitrate + 6- phosphogluconate (le)

The equilibrium constant of reaction le has not yet been de-
termined experimentally, but it can be calculated from free energy data.
Thus, the free energy change of reaction Id can be estimated from the
equation, —AF = nFAE, relating free energy change to the potential
difference between two reacting oxidation-reduction systems (15).
The reacting systems in reaction Id are the glucose-6-phosphate ^
phosphogluconate {Eq — —0.43 v. at pH 7.0) and the TPN ;=±
TPNH2 system, the potential of which is unknown but can be con-
sidered to be near that of the DPN «=± DPNH2 system (Eo = —0.28 v.
at pH 7.0). For a potential difference of 0.43-0.28 = 0.15 v., the
AF of reaction Id would be —6890 cal., corresponding to an equilib-
rium constant K^ = (6-phosphogluconate) (TPNH2)/ (glucose-6-phos-
phate)(TPN) of the order of 10^. Hence, the equilibrium constant of
reaction le: K^. = (/-isocitrate) (6-phosphogluconate)/(glucose-6-
phosphate) (a-ketoglutarate) (CO2) = K^X K^= \.'iX 10"" X 10^ =
13.

A further shift of the equilibrium of reaction le toward carbon
dioxide fixation occurs in the presence of aconitase. This enzyme is
widely distributed in animal and plant cells and catalyzes the inter-
conversion between /-isocitric, czV-aconitic, and citric acids according to
reaction If, where the figures in parentheses give the percentage of the

174

COOH

I
CHOH

I
HC— COOH

I
CH2

I
COOH

ACONITASE
COOH

CH

CARBON DIOXIDE

COOH

CH,

■HoO 11 +H2O I

^ C— COOH . HOC— COOH

+ H2O

(If)

-H2O

CH2

I
COOH

/-Isocitric acid (7.7%) m-Aconitic acid (3.1%)

CH2

COOH

Citric acid (89.2%)

individual components present at equilibrium at 37° C (18). Under
these conditions, over 90% of the /-isocitric acid formed in step 3 is
converted to <:2i--aconitic and citric acids.

The free energy changes of the various steps of this system are
given in Table II. Combination of the four steps gives an over-all
balance of about —3000 cal., i. e., the complete system is exergonic
by a fairly ample margin.

Table II
Carbon Dioxide Fixation by a-KETOGLUTARic Acid

Step

Reaction

Enzyme

AF,
cal.

Remarks

(7) Carboxylation
(2) Rfduction

{3) Reduction of pyri-
dine nucleotide

(4) Over-all reaction Jor
first three steps

(5) Isomerization

a-Ketoglutarate + CO2
<^ oxalosuccinate

Oxalosuccinate + TP-
NH2 <=i /-isocitrate +
TPN

Glucose-6-phosphate +
TPN <=^ 6-phosphoglu-
conate + TPNHo

or-Ketoglutarate + CO2
+ glucose-6-phosphate
^ 6-phosphogIuconate
+ /-isocitrate

/-Isocitrate <=i citrate

Oxalosuccinic
carboxylase

Isocitric dehydro-
genase

Glucosephosphate
dehydrogenase

+ 4460
+ 708

-6890

-1711

ca.
-1500

Calc. for r = 295°
from equil. const.
Calc. as above

Calc. from — AF =
aFAE

AF =■ S of AF of

Aconitase

partial reactions

Calc. for T = 310°
from equil. const.

There is a possibility that the above reaction series might not
stop with the formation of citric acid. Some microorganisms have
been reported to split citrate to oxalacetate and acetate (2), a reaction
that would favor carbon dioxide fixation via the a-ketoglutaric carboxyl-
ation system by displacing the equilibrium still further. Since, as
we shall see later, acetate can be converted to pyruvate by reductive
carboxylation and pyruvate forms carbohydrate in cells, the biological
formation of acetate from citrate would be of considerable importance.

Carbon dioxide fixation by pyruvic acid. The primary re-

175

SEVERO OCHOA

action of this system (5,9,1 1,12) involves the reversal of the decarboxy-
lation of oxalacetic acid to pyruvic acid (reaction Ila). Reaction Ila

OXALACETIC CARBOXYLASE
CH, COOH

i I

CO + CO2 , GH2

I I (Ila)

COOH CO

I
COOH

Pyruvic acid Oxalacetic acid

is catalyzed by oxalacetic carboxylase, an enzyme which is found in
bacteria and liver and requires magnesium or manganese ions for
activity. As in the case of oxalosuccinic carboxylase, the equilibrium
of reaction Ila is very far to the left. Reversibility has been demon-
strated by allowing the enzyme to act on oxalacetic acid in the presence
of isotopic carbon dioxide. By stopping the enzyme action when
about half of the oxalacetic acid was decarboxylated, the presence of
isotopic carbon in the /S-carboxyl group was demonstrated.

The equilibrium constant of reaction Ila has been calculated
from its free energy change, in turn calculated from the free energies
of formation (at 38 ° C.) of the substances involved (5) :

oxalacetate + H2O > pyruvate" -\- HCOa"

-184,210 cal. -56,200 cal. -106,460 cal. -139,200 cal.

AF thus calculated is —5250 cal. for the decarboxylation, and A"a =
(oxalacetate — )/ (pyruvate") (HCO3-) = 0.2 X 10"^ a value of the
same order of magnitude as that of reaction la.

Step 2 occurs when both malic dehydrogenase (see Table I)
and reduced diphosphopyridine nucleotide are present, since the
oxalacetic acid formed by reaction Ila is then reduced to /-malic
acid (reaction lib).

MALIC DEHYDROGENASE
COOH COOH

CH2 CH2

I + DPNH2 , I + DPN (Hb)

CO CHOH

COOH COOH

Oxalacetic acid /-Malic acid

176

CARBON DIOXIDE

The combined result of reactions Ila and lib is reaction IIc;

OXALAGETIC CARBOXYLASE AND MALIC DEHYDROGENASE

CH, COOH

I I

CO + CO2 + DPNH2 _1 CH2 + DPN

(lie)

COOH

Pyruvic acid

CHOH

1
COOH

/-Malic acid

In step 3, the DPN formed in reaction IIc can be reduced by
lactic acid through a coenzyme-Hnked dismutation with the lactic de-
hydrogenase system. Lactic dehydrogenase catalyzes reaction lid.

CH,

I
CHOH + DPN ;

LACTIC DEHYDROGENASE
CHs

COOH

/( + )-Lactic acid

CO + DPNHj

I
COOH

Pyruvic acid

(Hd)

The combined result of reactions IIc and lid is reaction He.
pyruvate + CO2 + /( + )-lactate ^ /-malate + pyruvate (He)
The free energy change of this reaction (see Table III), is about +1500

Table III
Carbon Dioxide Fixation by Pyruvic Acid

Step

Reaction

Enzyme

AF,
cal.

Remarks

(7) Carboxylation

Pyruvate -j- CO2 «=^ oxal-
acetate

Oxalacetic car-
boxylase

+ 5250

Calc. from free
energies of forma-
tion

Calc. from — AF =
nFAE

Calc. from — AF ■=
nFAE

AF = 2 of AF of

(2) Reduction

{3) Reduction oj pyri-
dine nucleotide
Over-all reaction (A)

(4) Dehydration

(5) Further reduction

Over-all reaction (B)

Oxalacetate + DPNH2?=i
/-malate + DPN

/(+)-Lactate + DPN ?ii
pyruvate + DPNHj

Pyruvate + CO2 -t- /(-i-)-
lactate <=^ /-malate -|-
pyruvate[C02-i-/(-f-)-
actate?^ /-malate]

/-Malate ?=i fumarate -)-
H2O

Fumarate -|- 2 H «=i suc-
cinate

Pyruvate + CO2 -f /(-!-)-
lactate -{- 2 H ♦^i suc-
cinate 4- pyruvate -|-
H2O [/(-f Vlactate -f
CO2 + 2H ^ suc-
cinate + H2O]

Malic dehydro-
genase

Lactic dehy-
drogenase

-8300
+ 4600
+l}iO

+ 705
-20,450

-11,19}

Fumarase

Hydrogenasc;
fumaric re-
ductase (?)

partial reactions

Calc. from equil.

const.
Calc. from — AF ■•

nFA£

AF - S of AF of

partial reactions

177

SEVERO OCHOA

cal. and the calculated equilibrium constant, K^ = (/-malate)/(/(+)-
lactate)(C02), approximately 0.1.

Step 3 could also involve dismutation with any other diphos-
phopyridine nucleotide dehydrogenase system having an oxidation-
reduction potential lower than that of the malic system (see Table I).
We have considered the lactic dehydrogenase in this connection because
there is evidence that it can participate in carbon dioxide fixation by
pyruvic acid. In the presence of the enzyme fumarase, part of the
/-malic acid formed in reaction He would be dehydrated to fumaric
acid. Fumarase is widely distributed in plant and animal cells. Re-
action lie would then be replaced by reaction Ilf:

pyruvate + CO2 + /(+) -lactate ^

fumarate + H2O -f pyruvate (Ilf)

Experimental support for the occurrence of the over-all reaction
(Ilf) is gained from the observation that fumarate, when added to an
enzyme preparation from liver in the presence of pyruvate, DPN, and
manganese ions, is converted to lactate and carbon dioxide (5). This
indicates that reaction Ilf can proceed from right to left. That it also
proceeds from left to right is indicated by the presence of isotopic carbon
in the carboxyl groups of the residual fumarate when the reaction is
carried out in presence of isotopic carbon dioxide (29). The pyruvic
oxalacetic system of carbon dioxide fixation has not yet been investi-
gated by the methods used in the study of the ketoglutaric-oxalosuccinic
system.

A considerable shift of the equilibrium of reaction Ilf in the
direction of carbon dioxide fixation can be brought about by reduction
of the fumarate to succinate, a reaction that occurs in bacteria and
liver tissue. In fermentation of glucose or glycerol by propionic acid
bacteria, succinate is found to be one of the end products. Experi-
ments with isotopic carbon dioxide have shown that the carboxyl
groups of succinic acid become labeled. The carboxyl groups of mal-
ate, fumarate, and succinate, formed by pyruvate fermentation with
Escherichia coli in the presence of carbon dioxide containing isotopic
carbon, also show excess of heavy carbon. This is also the case when
pyruvate is incubated with liver preparations. Some bacteria, e. g.,
E. coli, can use molecular hydrogen for fumarate reduction and, in the

178

CARBON DIOXIDE

presence of pyruvic acid, carbon dioxide, and hydrogen, all three are
utilized to form succinic acid (9,10,28).

The reduction of fumarate to succinate is strongly exergonic
and provides ample energy to drive carbon dioxide fixation by this
system to completion (see Table III).

Reductive Carhoxylation

This type of carbon dioxide fixation has been discussed recently
by Lipmann (17) and will only be briefly considered here. The fixa-
tion is a consequence of the reversal of the oxidative decarboxylation of
a-keto acids catalyzed by specific enzymes. These reactions involve
inorganic phosphate, and lead to thdf formation of an anhydride of the
next lower fatty acid and phosphoric acid with liberation of either formic
acid, or carbon dioxide and hydrogen; the hydrogen may either ap-
pear as molecular hydrogen or reduce a hydrogen acceptor (see re-
actions Ilia and 1 1 lb). The anhydride bond formed has a high energy
content and is generally referred to as an energy-rich phosphate bond.
There are other types of energy-rich phosphate bonds of biological im-
portance, such as enol phosphate, guanidine phosphate, and pyro-
phosphate bonds. The pyrophosphate type of bond has a special
significance because, by the action of specific enzymes, it can give rise
to any of the other phosphate bonds. Since the reactions involved in
these conversions are reversible, it follows not only that any energy-
rich phosphate bond can generate pyrophosphate bonds, but also
that the various types of bonds can be converted into one another
through the intermediate formation of pyrophosphate linkages.

The biologically important pyrophosphate group is the one
present in adenosine polyphosphates, that is, adenosine triphosphate
(abbreviated ATP), and adenosine diphosphate (abbreviated ADP).

N--CNH2

HG C— N

I \^Tj OH OH OH OH OH

I /^^ II III

N— C— N CH— CH— CH— CH— CH2— O— P— O— P— O— P— OH

I I II II II

I o 1 000

Adenosine triphosphate

By enzymic hydrolysis, ATP is dephosphorylated to ADP, and this,
in turn, to adenosine monophosphate (adenine ribose 5-phosphate or

SEVERO]]OCHOA

muscle adenylic acid), with release of inorganic phosphate and of the
energy of the bond. By enzymic transphosphorylation, ATP trans-
fers phosphate to carboxyl, enol, and guanidine groups in a reversible
manner. Since the energy content of the various types of energy-rich
phosphate bonds is nearly the same, in the neighborhood of 12,000 cal.,
these transphosphorylations involve relatively small changes of free
energy (16).

The general type of oxidative decarboxylations can be repre-
sented by reaction Ilia (17), where X stands for a hydrogen acceptor.

GH3COOPO3-

+ CO2 + H2X

(Ilia')

cHaCO : cooH + H : opo,-

Pyruvic acid

± CH3COOPO3— + CO2 +

H2

(Ilia")

CH3COOPO,-- + HCOOH

Acetyl phosphate (Ilia )

The oxidative decarboxylation of a-ketoglutaric acid generates phos-
phate bonds (19) and may be represented by reaction Illb. All these
reactions occur in bacteria; reactions Ilia' and Illb also occur in ani-
mal tissues.

C00HCH2CH2C0 : COOH + H : OPO,— + x

a-Ketoglutaric acid

COOHGH2CH2COOPO,-- + CO2 + H2X (Illb)

Succinyl phosphate

Two reactions are known at present through whose reversibility
reductive carboxylation can take place: (a) the splitting of formic acid
to carbon dioxide and hydrogen; and (b) the splitting of pyruvic acid
to acetyl phosphate and formic acid. Since acetyl phosphate can be
formed enzymically by a reversible reaction between acetic acid and
ATP, we have a biological system of reductive carboxylation of acetate
to pyruvate.

The pattern of carbon dioxide fixation established for /3-car-
boxylations must be modified here to include two preliminary steps,
viz., those of phosphorylation and of carbon dioxide reduction, respec-
tively, and a final step for regenerating the ATP used at the beginning.

180

CARBON DIOXIDE

A possible series of reactions, modified from Lipmann, is suggested in
Table IV. The absolute value of the free energy change of step 6

Table IV
Reductive Carboxylation

Step

Reaction

Enzyme

AF.
calc.

Remarks

(7) Phosphorylation

Acetate + ATP;(=i acetyl

From Clostridium

-1-3000

Calc. from equil.

phosphate + ADP
CO2 + H2<=i HCOOH

butylicum

const.

(2) Carbon dioxide re-

Formic hydro-

- 200

Calc. from equil.

duction

genylase (£.
coli)
From E. coli

const.

(3) Carboxylation

Acetyl phosphate + for-

-F2800

Calc. from equil.

mate «=i pyruvate -f-

const.

phosphate

(4) Reduction oj car-

Pyruvate + DPNH2 ^

Lactic dehydro-

-4600

Calc. from — ^F —

boxylation product

I ( + ) -lactate + DPN

genase

nFAE

(5) Reduction oj pyri-

3-Phosphoglyceraldehyde

3-Phosphoglycer-
aldehyde de-

ca.

Calc. from equil.

dine nucleotide

+ phosphate 4- DPN

-t- 400

const.

<=i 1,3-diphosphoglyc-

hydrogenase

erate + DPNHi

(6) Regeneration of

1 ,3-Diphosphoglycerate

From yeast

-3000

See text

ATP

+ ADP ?=i 3-phospho-
glycerate -(- ATP

Over-all reaction

Acetate + COj -f H2 -f
3-phosphoglyceralde-

ca.
-1600

hyde<=i/( + )-lactate -f

3-phosphoglycerate

has been considered to be the same as that of the step 7, on the assump-
tion that the bond energy of the anhydride linkage in 1-phosphoglycer-
ate is the same as that of the acetyl phosphate group. The reaction
series postulated in Table IV would lead from acetate to lactate and
the over-all reaction would be exergonic. A characteristic feature of
reductive carboxylation is that it must be started by input of a large
amount of energy from energy-rich phosphate bonds.

Pyruvic acid formed by reductive carboxylation of acetate
might also be converted to carbohydrate after reduction to triose in-
stead of being reduced to lactic acid. Such a conversion would re-
quire additional phosphate-bond energy which might be amply avail-
able in chemosynthesis or photosynthesis.

It is not unlikely that reactions Ilia' and Illb might also be
reversible. The main difference between these reactions and reactions
Ilia" and Ilia'" is one of mechanism, i. e., the hydrogen from the
keto acid is first transferred to a hydrogen acceptor instead of being
released either as molecular hydrogen or in combination with carbon
dioxide as formic acid. If such were the case, all of the intermediate
reactions involved in the oxidation of foodstuffs (i. e., in respiration)
would be reversible.

i8i

SEVERO OCHOA

Carbon Dioxide Fixation in Chemosynthesis
and Photosynthesis

It is now known that photosynthesis can be divided into two
phases relatively independent of one another: (1) a so-called "dark"
reaction which occurs in the absence of light and consists of a reversible
fixation of carbon dioxide to form a carboxylic acid; and (2) the photo-
lytic fission of water, yielding hydrogen with liberation of oxygen.
Hydrogen produced by photolysis is used to reduce the products
formed by carboxylation. In chemosynthesis, hydrogen is obtained
by oxidation of inorganic compounds — a process that also supplies
energy (6,7,23-25).

It would appear that essentially the same mechanisms that func-
tion in carbon dioxide fixation by heterotrophic organisms are operative
in both photosynthesis and chemosynthesis, with the difference that,
in the case of the heterotrophs, hydrogen and energy are derived by
oxidation of organic materials. The widespread occurrence in plants
of di- and tricarboxylic acids (malic, citric, isocitric) and of the enzymes
that participate in their metabolism (fumarase, malic dehydrogenase,
aconitase, isocitric dehydrogenase) lends support to such a view. Phos-
phorylation processes which, as we have seen, are essential for reductive
carboxylations, are connected with carbon dioxide fixation in the sulfur
oxidizing autotroph Thiobacillus thiooxidans, and photosynthetic organ-
isms may utilize radiant energy for the synthesis of energy-rich phos-
phate bonds (3,22,26).

We cannot yet formulate in any detail the course of events in
photosynthesis and chemosynthesis, but, using the knowledge gained
by the study of the mechanisms of carbon dioxide fixation in hetero-
trophic organisms, we may attempt to draw a plausible picture of the
chemical events. Reversal of the oxidative degradation of foodstufTs,
i. e., of respiration, would now seem to be a definite possibility.

We have seen that, on carboxylation and reduction, a-keto-
glutaric acid can be converted to citric acid, and have indicated that
the latter may be split to acetic and oxalacetic acids. Further, oxal-
acetic acid can be reduced to succinic acid by way of malic and fumaric
acids, and succinic acid could be converted to a-ketoglutaric acid by
reductive carboxylation, i. e., by reversal of reaction Illb. In this
way, a-ketoglutaric acid would be regenerated, while the acetic acid

182

CARBON DIOXIDE

formed earlier can be converted to pyruvic acid by reductive carboxyl-
ation, i. e., by reversal of reaction Ilia. We would thus have a cyclic
mechanism whereby carbon dioxide and hydrogen entering at various
points would emerge as pyruvic acid. The di- and tricarboxylic acids
would only act catalytically as carriers of carbon dioxide and hydrogen.
This is a reversal of the so-called tricarboxylic acid cycle, which is
considered to be an important pathway for the oxidative breakdown of
carbohydrate and fat in cells. Scheme II presents this metabolic cycle,
incorporating recent findings concerning some of the intermediate
reactions (1,4,13,28). For a more detailed discussion, see Lardy and
Elvehjem (14).

Scheme n

Reversible Tricarboxylic Acid Cycle
Carbohydrate

fatty acids

pyruvate ^

lactate

malate

-H2O

czi'-aconitate.

+H2O

citrate

+H2O

-H2O

-H:0

fumarate

-HK5

+H2O

isocitrate

-2H

+ 2H

+ 2H

■2H

succinate

oxalosuccinate

•2H
-CO2

+ 2H
+ CO2

+ CO2

-CO2

a-ketoglutarate
Another mechanism recently suggested for photo- and chemo-
syntheses involves a sequence of carboxylations and reductions leading

183

SEVERO OCHOA

from a carboxylic acid, through the next higher a-keto acid, to an
a-hydroxy acid, which in turn would be carboxylated to the next higher
Q;-keto-/3-hydroxy acid, and so on. Such a sequence would be es-
sentially a reversal of a pathway of carbohydrate oxidation by way of
phosphohexonic acid, a-ketophosphohexonic acid, phosphopentonic
acid, etc., through alternating dehydrogenations and decarboxylations
(17).

Obviously, the above schemes are only gross approximations.
The main point is that what we know about the mechanisms of carbon
dioxide assimilation by heterotrophic organisms strongly suggests that
all reactions involved in cellular respiration are essentially reversible.
Thus, the processes of both photosynthesis and chemosynthesis may rep-
resent reversals of the respiratory process not only from the standpoint
of energy but also from the standpoint of the enzymic mechanisms.

References

(1) Buchanan, J. M., Sakami, W., Gurin, S., and Wilson, D. W., J. Biol
Chem., 159,695 (1945).

(2) Deffner, M., Ann., 536, 44 (1938).

(3) Emerson, R. L., Stauffer, J. F., and Umbreit, W. W., Am. J. Botany,
31, 107 (1944).

(4) Evans, E. A., Jr., and Slotin, L., J. Biol. Chem., 141, 439 (1941).

(5) Evans, E. A., Jr., Vennesland, B., and Slotin, L., J. Biol. Chem., 147,
771 (1943).

(6) Franck, J., and Gaffron, H., in Advances in Enzymology, Vol. I. Inter-
science, New York, 1941, p. 199.

(7) Gaffron, H., Biol. Rev. Cambridge Phil. Soc, 19, 1 (1944).

(8) Kalckar, H. M., Chem. Revs., 28, 71 (1941).

(9) Kalnitsky, G., and Werkman, C. H., Arch. Biochem., 4, 25 (1944).

(10) Kalnitsky, G., Wood, H. G., and Werkman, C. H., Arch. Biochem., 2,
269 (1943).

(11) Krampitz, L. O., and Werkman, C. H., Biochem. J., 35, 595 (1945).

(12) Krampitz, L. O., Wood, H. G., and Werkman, C. H., J. Biol. Chem.,
147,243(1943).

(13) Krebs, H. A., in Advances in Enzymology, Vol. III. Interscience, New
York, 1943, p. 191.

(14) Lardy, H. A., and Elvehjem, C. A., Ann. Rev. Biochem., 14, 1 (1945).

(15) Lewis, G. N., and Randall, M., Thermodynamics and the Free Energy of
Chemical Substances. McGraw-Hill, New York, 1923.

184

CARBON DIOXIDE

(16) Lipmann, F., in Advances in Enzymology, Vol. I. Interscience, New York.
1941, p. 99.

(17) Lipmann, F., and Tuttle, L. C, J. Biol. Chem., 158, 505 (1945).

(18) Martius, C, and Leonhardt, H., Z- physiol. Chem., 278, 208 (1943).

(19) Ochoa, S., J. Biol. Chem., 155, 87 (1944).

(20) Ochoa, S., J. Biol. Chem., 159, 243 (1945).

(21) Ochoa, S., and Weisz-Tabori, E., J. Biol. Chem., 159, 245 (1945).

(22) Ruben, S., J. Am. Chem. Soc, 65, 279 (1943).

(23) Ruben, S., and Kamen, M. D., J. Am. Chem. Soc, 62, 3451 (1940).

(24) Van Niel, C. B., in Advances in Enzymology, Vol. I. Interscience, New
York, 1941, p. 263.

(25) Van Niel, C. B., Physiol. Revs., 23, 338 (1943).

(26) Vogler, K. G., and Umbreh, W. W., J. Geji. Physiol., 26, 157 (1942).

(27) Warburg, O., Ergeb. Enzymforsch., 7, 210 (1938).

(28) Werkman, C. H., and Wood, H. G., in Advances in Enzymology, Vol. II.
Interscience, New York, 1942, p. 135.

(29) Wood, H. G., Vennesland, B., and Evans, E. A., Jr., J. Biol. Chem.,
159, 153 (1945).

185

13

HORMONES

B. A. HOUSSAY, director of the institute of biology and

EXPERIMENTAL MEDICINE, BUENOS AIRES

Definition and Significance

'HE HORMONES are specific chemical substances pro-
duced by an organ or tissue which, after being discharged
into the circulating fluids (milieu inter ieur), may reach all parts of the
organism and in small amounts markedly influence the functions of
other organs or systems without themselves contributing important
quantities of matter or energy.

This definition diff"erentiates them from other important sub-
stances which also reach the circulating fluids, such as: (a) nutritive
substances which supply the tissues with materials and energy as, for
example, glucose, amino acids, lipids, etc.; (b) vitamins, organic chemi-
cal regulators contained in the food; (c) chemical mediators of nerve
action, liberated by the nerve endings in the close vicinity of effector
or other nerve cells and which exert a localized action; {d) the organizers,
embryonic substances of regional origin which govern the differentia-
tion of a determined organ, even if transplated to another zone or culti-
vated in vitro; (e) the parhormones (Gley) which, although being excre-
tory substances produced by the metabolic processes of all the tissues,
nevertheless perform important regulatory functions, as is the case for
carbon dioxide in the regulation of respiration.

The action of what we today call glands of internal secretion
is due to the hormones they produce; and the insufficiency of these
glands, therefore, is only a matter of hormone deficiency.

187

B. A. HOUSSAY

In a wider sense, which is not usual, internal secretion would mean
any specific cellular elaboration discharged into the internal medium. It is
in this sense that Claude Bernard considered glucose, which is produced in
the liver and then passes into the blood, as an internal secretion.

As we all know, the glands of internal secretion are referred to
as such because they pour their elaborated products into the blood,
in contrast to the glands of external secretion which pour their products
out of the body or in cavities which communicate with the external
medium.

The first demonstration of a hormonal action was the induction
of the comb of a capon by testicular graft (Berthold, 1849). The ex-
pression "internal secretion" was used for the first time by Claude
Bernard (1855) to point out that the liver "shed" sugar in the blood.
The concept of internal secretions as we now understand it is due to
Brown Sequard who, in 1891, stated in a paper with Arsonval, "Nous
admettons que chaque tissue, et plus g^n^ralement, chaque cellule
de I'organisme secrete pour son propre compte des produits ou des
ferments sp^ciaux qui sont versus dans le sang et qui viennent influencer
par I'interm^diaire de ce liquide toutes les autres cellules rendues ainsi
solidaires les unes des autres par un m^chanisme autre que le tissue
nerveux." The name "hormones" was proposed by Hardy to desig-
nate the "chemical messengers" (Bayliss and Starling, 1904) which are
secreted in the blood by one organ to stimulate the functions of another.

Hormone etymologically means "I arouse the activity" or "I
excite," but we are now aware of the existence of inhibitory hormonal
actions. It is useless to classify the hormones as stimulating or inhibi-
tory, inasmuch as the same hormone may often produce both effects,
depending upon its concentration or the organ which it affects. It is
understandable, therefore, why the designations proposed by Sharpey-
Schafer, who gave them the general name of "autacoids" and sub-
divided them into "hormones" (exciting) and "chalones" (inhibitory),
did not gain currency.

Chemical Nature

According to their chemical nature the known hormones can be
classified into three groups:

(7) Phenolic Derivatives. — Both adrenalin, the hormone of the

1 88

HORMONES

adrenal medulla, and thyroxine, the hormone of the thyroid, are
phenolic substances.

(2) Proteins. — To the proteins belong the hormones of the
hypophysis (two gonadotropins, thyrotropin, adrenocorticotropin,
lactogenic hormone and growth-promoting factor from the anterior
lobe, and the vasopressor, oxytocic and melanic-expanding principles
of the posterior lobe); the hormone of the pancreas (insulin), and the
hormone of the parathyroid (parathormone).

(J) Steroids. — To the steroid family belong the hormones
isolated from the ovarium and corpus luteum, testicle, and adrenal
cortex, twenty-eight steroids having been isolated so far from the
adrenal cortex (Reichstein, 1943). They all have in common a cyclo-
pentanoperhydrophenanthrene ring system.

Attempts to prepare the nonprotein hormones synthetically
have followed closely the determination of their chemical constitution.
Besides, several synthetic substances of constitution similar to, but not
identical with, the natural hormones have been prepared, and their
pharmacologic action tested. Many vasoconstrictor and bronchodila-
tor substances have thus been obtained which are more efficient than
adrenalin for some therapeutic uses. Several synthetic estrogens (di-
ethylstilbestiol, hexestrol, etc.) have also been obtained which are now
largely used instead of natural estrogens to treat ovarian insufficiency.
Desoxycorticosterone, a substance which is able to maintain adrenalec-
tomized animals in normal condition, has been prepared by partial
synthesis. Probably a wide field for pharmacological investigation of
corticoadrenal hormones will soon be opened, since Reichstein (1943)
announced a method of preparing steroids containing an atom of oxy-
gen linked to carbon atom 1 1 . Steroid hormones are usually prepared
by partial synthesis, starting from stigmasterol, a product of soybeans.

The active substances extracted from an endocrine gland do not
always represent the circulating natural hormone. Thus, while thy-
roxine increases the oxygen consumption if given to the whole animal,
it does not have such an action upon isolated tissues. Furthermore,
whereas only from 25 to 50% of organically bound iodine is extract-
able from the thyroid in the form of thyroxine, all the iodine compounds
of the thyroid are able to increase the rate of metabolism. These and
other reasons make it dubious that thyroxine is a constituent of the
thyroid hormone or the hormone itself.

189

B. A. HOUSSAY

Some of the protein hormones elicit such a small antibody re-
sponse, as in the case of insulin, that their administration by repeated
injections remains effective during scores of years. Other protein hor-
mones become progressively less effective, which makes it necessary to
increase the dose each time, as in the case of the parathyroid hormone.
There are still other hormones (thyrotropin, gonadotropin, and in
different degrees all the anteropituitary hormones) whose action de-
creases quickly if they are administered daily and which induce the
formation of antihormones. Thus, if an animal is treated with daily
doses of thyrotropin, it shows, at the beginning, hypertrophy and hyper-
function of the thyroid, but after a few weeks the action disappears
and is followed by atrophy and hypofunction of the organ. The
serum is then found to contain antithyrotropin, which not only inhibits
the thyrotropin action but is also capable of inhibiting the action of
the thyroid gland as shown by injecting such serum into another
animal.

Antihormones are only produced when protein hormones are
administered parenterally. Gollip (1934) thought that they were sub-
stances of physiological importance and that each hormone should have
its corresponding antihormone to balance its effects. But it now seems
that antihormones are antibodies or immunity mechanisms (Rowlands)
reacting to injection of antigens from another species. It is to be noted
that adrenalin, thyroxine, and the steroid hormones do not produce
antihormones and are not proteins. For some of the actions of these
hormones a certain habit may be produced without any demonstrable
antihormones.

The natural protein hormones do not seem to be completely
equivalent to those which are extracted in the laboratories, for they
do not induce the formation of antihormones. Thus, when a rat is
castrated, great quantities of gonadotropins accumulate in its blood,
as is readily demonstrated in parabiosis experiments — that is to say,
experiments involving sewing two rats together side by side, after
opening their bellies laterally, so that their peritoneal cavities, muscles,
skin, and blood vessels are fused. In this way, the gonadotropin pres-
ent in the blood of the castrated rat passes into the circulation of the
normal rat, which, as a consequence, shows an intense stimulation of
the gonads. This stimulation persists steadily for months, while para-
biosis lasts, with no sign of antihormone production, in contrast to the

190

HORMONES

rapid formation of antihormones and the inversion of effects caused by
the injection of gonadotropins isolated from the gland.

The preparation of active protein hormones that will not
produce antihormones is one of the outstanding problems endocri-
nology must solve. Meanwhile, the formation of antihormones imposes
an important limitation on prolonged therapeutic application of
certain hormones, particularly those of the pituitary and parathyroid.
This is one of the causes of the discrepancy existing between, on the
one hand, the great functional importance of the hypophysis as demon-
strated by experiment and by the study of disease in man, and, on the
other, the limited possibilities realized thus far from therapeutic ap-
plications.

The chemical mechanisms by which hormones act upon cells
are not yet known. It has not yet been proved that they participate
in enzyme systems, as is the case for vitamins such as thiamin, niacin,
and riboflavin. The hormones are chemical regulators that probably
modify some link in the chain of metabolic reactions, a field of study
which has remained almost unexplored to date, in spite of its great
importance.

Role

Some endocrine organs, phylogenetically, begin as external
secretion glands that cast their secretions into the digestive system but
afterward lose their excretory function and change into internal secre-
tion glands. The pituitary, thyroid, and pancreatic islets of Langer-
hans have evolved in some such way. The parathyroid, regulator of the
metabolism of calcium and phosphorus, derives from the branchial
arches. The ovarium, testicle, and adrenals, which produce the
steroid hormones with sexual and metabolic actions, derive from the
coelomic epithelium. The adrenal medulla that secretes adrenalin
derives from the nervous sympathetic system, and the neurohypophysis
from the diencephalon.

The hormones of the vertebrates are better known than those of
the invertebrates. In the latter, certain processes have been shown to
be regulated by hormones, such as metamorphosis, color (in the case of
Crustacea), and sexual dimorphism in some instances. The hormones
regulate functions that exist before hormones appear and which often
persist without them. Thus, all animal and plant cells consume glu-

191

B. A. HOUSSAY

cose without the intervention of insulin, but in the great majority of
the vertebrates insuhn is a new regulating mechanism of such impor-
tance that, when missing, diabetes results, a disease which is fatal sooner
or later depending upon the species. Sexuality exists in many inverte-
brates without intervention of hormones, but in the vertebrates sexual
characters do not reach their full development without the action of
ovarian or testicular hormones, the production of which is governed by
the pituitary gonadotropins. The functional unity of the organism is
assured by nervous and humoral (chemical) mechanisms of correlation,
which interrelate the different parts or tissues and regulate their recipro-
cal activities. Sherrington has pointed out the unifying integrative
action of the nervous system; a similar role is played by the humoral
factors, among which are the hormones. Both types of mechanism
maintain the stability of the milieu interieur (CI. Bernard) and of the
organism as a whole (Cannon's homeostasis) in spite of varying condi-
tions in the external environment and in the organism's activity.
Modern studies have thus confirmed, extended, and given a more pre-
cise meaning to the old and vague notions about the correlations be-
tween the organs ("consensus partium" or "sympathies").

The roles played by the hormones can be classified somewhat
conventionally as follows:

(7) Metabolism. — Some of the hormones regulate the balance
of metabolic processes. Their action may be general (stimulation of
oxidation processes by thyroxine) or rather specialized (parathyroids
and calcium, insulin and carbohydrate). One and the same hormone
may modify several metabolic processes, e. g., adrenal steroids act both
upon the metabolism of water and salt, and upon the metabolism of
carbohydrates.

(2) Morphogenesis. — The morphogenetic actions are the con-
sequence of the selective role of the hormones in assimilation and
growth phenomena. Some endocrine glands such as the pituitary,
thyroid, parathyroid, and sexual glands play an important role in
growth during a certain stage of development, principally because of
their action upon the synthesis of proteins and on the development of
bone. In other cases the growth-promoting action is exerted on special
organs, as in the case of estrogens which promote the growth of the
uterus and mammary gland.

As the result of growth and differentiation, the proper morpho-

192

HORMONES

logical constitution of each sex and of each individual is attained.
The pars distalis of the pituitary gland stimulates the thyroid of tad-
poles, and in turn the thyroid secretion promotes the metamorphosis
of the larval form. The pars intermedia of the pituitai'y, and sometimes
adrenalin, regulate the color of the skin of amphibians and fishes, by
dispersing or concentrating the pigment gianulcs in the chromato-
phore cells.

(3) Endocrine Interrelation and Balance. — A close functional re-
lation exists between the endocrine glands. Following Gley, we can,
consider "correlation" as the relation of one organ with another;
when it is a mutual correlation, we may call it "interrelation." Thus,
while the secretion of pancreatic juice caused by secretin is an example
of "correlation" between the duodenum and the pancreas, there exists
an interrelation between the hypophysis and the gonads, for if the
anterohypophysis is responsible for the final development and main-
tenance of the functional activity of the gonads, at the same time the
endocrine secretion of the latter regulates and moderates the gonad-
stimulating function of the hypophysis.

Analysis of each separate endocrine gland or of each hormone is
clearly artificial. It was initially necessary because in order to study
the behaviour of a gland, there were no procedures available other
than extirpating the organ, injecting its extracts, and carrying out
anatomical and functional studies of clinical cases showing visible altera-
tions of some gland. Afterward the method persisted for didactic
reasons; but since no endocrine gland can be considered as if its action
were independent of other glands, the method is now being abandoned.

A constellation of endocrine glands exists whose central organ
is the hypophysis, the function of each gland being influenced more or
less by the function of the others. Because the anterohypophysis con-
tributes to the development and maintenance of the structure and
function of various glands, extirpation of the pituitary produces a
marked atrophy and hypofunction of the thyroid, gonads, and adre-
nals, to such an extent that it has been said that the hypophysectomized
rat is an endocrinological and metabolic ruin. But, at the same time,
the structure and function of the anterohypophysis is governed by
hormones secreted by the thyroid, gonads, and adrenal cortex.

Each gland produces very specific hormones which play im-
portant roles; nevertheless these hormones constitute simple parts of

B. A. HOUSSAY

complex functional mechanisms. Thus, the regulation of the metabo-
lism of carbohydrates involves the liver, insuHn, hypophysial, cor-
ticoadrenal, and thyroid hormones, as well as participation of the in-
testine, kidney, and muscle. Sexual functions depend upon the com-
bined action of the hormones of the hypophysis, ovary, corpus luteum,
and placenta (during pregnancy), plus the action of still other glands,
such as the adrenal cortex and the thyroid.

As explained further on, in any single function of the organism
more than one hormone plays its part, even when one of them exhibits
a very specific and preponderant role.

(4) Sexuality and Reproduction. — Among the endocrine actions,
the sexual functions are especially important. These functions depend
on the hormones of the gonads, governed in part by the pituitary go-
nadotropins, in part by the hormones of the adrenal cortex and, to a cer-
tain degree, by the thyroid. They regulate the production of the fe-
male (ovule) and male (spermatozoid) germinal cells or the develop-
ment of the organs which carry them, the development of the impulses
and sexual acts that lead to fecundation, the progress of pregnancy and
delivery, and the secretion of milk. The sexual hormones are mdis-
pensable links for individual sexuality and for the maintenance of the
species.

(5) Mental and Nervous Functions. — The hormones influence
several nervous and muscular activities. We only need to remember
the differences shown by the sexes, the mental dullness due to hypo-
thyroidism, the nervousness or mental instability of hyperthyroidism,
the tetany of hypoparathyroidism, etc. These actions of the hormones
are probably due to their influence on the metabolism of the nervous
system, a point that has not received much study.

(6) Vital Role. — The extirpation of some endocrine glands
causes death. This was attributed to hypothetical poisons which ac-
cumulated because they were not neutralized by the endocrine glands
(antitoxic role), a theory that has been abandoned because no poison
has been demonstrated and because it has been proved that death
was due to metabolic disturbances provoked by the lack of the hormones
normally produced by the organ in question.

(7) Resistance. — The hormones are important factors in build-
ing up the resistance of the organism to certain hazards such as high or
low temperatures, low or excessive oxygen tension, overexercise, in-

HORMONES

fections, toxins, trauma, various poisons, etc. For example, pituitarec-
tomized or adrenalectomized animals show a low resistance to all of
these, and also are very sensitive to hypoglycemia and hypotensor
agents or to the circumstances which provoke shock or hypothermy.
Many of these responses depend upon metabolic phenomena. Some
endocrine factors also influence the production of immunity antibodies
or anaphylaxis and resistance to some infections. The harmful
agents or circumstances (e. g., cold, fatigue, toxins, etc.) all produce
similar reactions in a given organism, but the nature of these reac-
tions varies in different species. The adrenals are largely responsible
for these reactions, which pass through several phases: an initial
"alarm reaction" followed by temporary compensation, and finally,
decompensation. They have been thoroughly studied by Selye.

Abnormal internal secretions. There are internal secretions
produced under abnormal conditions. Thus, when the arterial hyper-
tension is produced because the ischemic kidney produces renin,
which acts enzymically upon the hypertensinogen of plasma to
form hypertensin, a substance which increases the arterial blood
pressure. During arterial hypotension, renin is secreted in the
blood; but it has not yet been demonstrated with certainty that there
is normally a small secretion of renin.

Hormones and cancer. In certain cases, hormones enhance
the development of tumors. Their role seems to consist more in pro-
moting the growth of tumors than in initiating the malignant cellular
transformation. The extirpation of glands (hypophysis, adrenals,
gonads, etc.) or the injection of hormones may hasten or delay the
growth of several types of tumors. Thus, testicular castration and
injection of estrogenic hormones retard the development of prostatic
cancer, whereas testicular hormone accelerates it.

Some hormones promote the development of cancer. Thus,
mammary cancer is less frequent in males or in castrated females of a
strain of rat showing high incidence of this cancer in the adult female.
The injection of estrogens (Lacassagne) stimulates growth of the mam-
mary gland and produces cancer in many of the males of that strain.
In these cases, it is debatable whether the estrogens initiate the cancer
or merely promote its development. But recent work has shown that
estrogenic induction of mammary cancer frequently is obtained in
strains of rat in which the cancer occurs spontaneously only in rare

B. A. HOUSSAY

instances (Geschickter and Byrnes). Some estrogenic hormones have
curious tumorigenic actions on the guinea pig, while other hormones
can prevent this action (Lipschiitz).

Benign and malignant tumors of the endocrine glands can pro-
duce exaggerated quantities of hormones, as shown in cases of hyper-
thyroidism, hyperparathyroidism, hyperinsulinism, acromegalia or
giantism, and some adrenal, ovarian, and testicular tumors. Complex
effects result from the action of adrenal tumors: They may exert either
virilizing or feminizing actions, influence metabolic phenomena, alter
the body shape, or modify arterial blood pressure. The ovarian tumors
also produce varied effects such as feminizing or virilizing actions. Less
familiar is the humoral mechanism by which the malignant tumors af-
fect the metabolism of the whole organism.

Specificity

Endocrine glands and hormones have considerable specificity
of action. The ovary stimulates the development of the feminine sexual
characters, and the testicle that of the male sexual characters, as can be
demonstrated by castration, which prevents their development or
provokes their regression in both sexes. Conversely, the ovarian graft,
or the administration of estrogenic hormones, develops the feminine
characters in either castrated or entire animals. In the same manner,
testicular grafts or androgenic hormones develop the masculine char-
acters in either male or females, whether castrated or not.

Each hormone produces its characteristic action upon diverse
animal species. Thus, the insulin of a given animal produces hypo-
glycemia in all mammals. It is also a rule that the hormone extracted
from an animal is active in all other species of mammals. Thus,
insulin extracted from a variety of mammals produces hypoglycemia in
the rabbit, and insulin from bovine origin is active on all the verte-
brates on which it has been tried. But some rare exceptions are known.
Thus, gonadotropins from mammals have no effect on the gonads of the
toad, Bufo arenarum, the cause of the anomaly being unknown.

Some hormones are elaborated by more than one gland. For
example, adrenalin is secreted by the chromaffin tissue but also exists
in the cutaneous poison of the toads. The estrogenic hormones are
found in the ovary, placenta, and adrenal cortex; and it is remarkable

196

HORMONES

that one of the most abundant sources for industrial production is the
urine of the staUion (in which it decreases after castration). The urine
of men and women also contains estrogens and androgens. The andro-
genic hormones, secreted normally by the testicle, can be secreted by an
oVary grafted in the ear (Hill) if the ear is maintained at a low tempera-
ture. They are also produced by some adrenal tumors and some
ovarian tumors (arrhenoblastoma).

The specificity of response of the reactive organs is not absolute.
Androgens can produce some efTects on the endometrium, the vaginal
epithelium, or the mammary gland. Estrogens can slightly influence
the seminal vesicles. The injection of estrogens can provoke masculine
erotization in some adult males, either normal or castrated.

The relationship between the chemical constitution and the
effect of the hormone is so close that a small modification of its mole-
cule can profoundly change its actions. Thus, ethyl testosterone pre-
pared from the male hormone is very active upon the endometrium.
The recent book of Selye shows much of the multiplicity and complexity
of the actions of the steroid hormones.

Synergies and Antagonisms

It would be impossible in a short essay to enumerate all the
cases in which two hormones either strengthen (synergy) or oppose
(antagonism) one another's actions. Estrogens in a certain adequate
dose prepare the uterus and mammary glands, and sensitize them to
progesterone; nevertheless, in other doses these substances can nullify
each other's actions. Estrogenic and androgenic hoimones have, in
certain cases, antagonistic actions; in others, their actions are inde-
pendent and do not interfere with each other; and, finally, some-
times they are mutually strengthened.

Regulation of Secretion of Hormones

Even though, for each function, several organs play a part, their
participation is regulated so as to maintain a steady balance, as is
demonstrated by the constancy of the blood sugar level, of the oxygen
consumption, of the blood calcium, etc. These regulations are, there-
fore, factors in the functional unity of the organism and in the equilib-
rium of their functions.

B. A. HOUSSAY

REGULATION OF EACH GLAND

Each endocrine gland has its own regulating mechanisms for
the secretion of its hormones. As judged by experiments on extirpation
and restitution, there must be a basal secretion, generally uninterrupted.
In certain instances this has been well demonstrated (adrenalin, insulin).
This secretion is submitted to regulating factors, the principal ones being
humoral, and in some cases also nervous. Thus the basal amount of
insulin secreted by the islets of Langerhans depends on the blood sugar
level; and, reciprocally, the blood sugar level depends on the amount
of insulin secreted. The basal secretion of insulin increases when
glycemia increases; conversely, it decreases when the blood sugar level
is lowered. The parathyroid secretion increases when calcemia de-
creases. In both cases the secretion of either insulin or parathormone
tends to restore the altered equilibrium of the internal medium. A
similar regulation of the secretion of the thyroid hormone is probable,
udging by the constancy of basal metabolism.

The regulation of mechanisms of hormone secretion are very pre-
cise and well designed to attain their objective. Thus, one-seventh of
the pancreas is adequate to maintain the basal glycemia at the normal
level; also, the basal glycemia continues at a normal level in an animal
even if four pancreases are grafted by vascular anastomosis. These
facts show that a normal secretion of insulin is maintained either with
one-seventh of the pancreas or with five pancreases — this because gly-
cemia governs insulin secretion. Only in abnormal cases (diabetes or
hyperinsulinism) is regulation of insulin secretion deficient or excessive
so that the gland works at a new level. In some cases of hyperplasia,
adenomata, or cancers, the endocrine glands have been known to pro-
duce hormonal hypersecretion.

Although the reduction in mass of an endocrine organ may
not alter its ability to function under basal conditions, it may lead to its
insufficiency in cases of emergency. The pancreas of the dog reduced
to one-fifth of its mass is enough to maintain normal glycemia, but if
grafted to a diabetic dog it does not replace a normal pancreas in cor-
recting the existing hyperglycemia. The resistance of the surgically
reduced pancreas is diminished against the action of injurious agents
such as extracts of the anterohypophysis and thyroid and, in conse-
quence, diabetes develops readily. When the pancreas is normal, these

198

HORMONES

injurious agents either do not, as in the case of sugar or thyroid extracts,
produce diabetes or else do so, as in the case of anteropituitary extracts,
only if much higher doses are used. In most instances the secretion of
the hormones is governed by humoral factors, while the nervous factors
play only an accessory, dispensable role. Thus, denervation of pancreas,
thyroid, adrenal cortex, or gonads does not produce any insufficiency
of their endocrine functions. Sometimes it is found that the nervous
action makes the secretory regulation somewhat quicker and more
precise, as in the case of insulin secretion.

There are cases, however, in which the nervous regulation is
important. Thus the splanchnic nerves exert a tonic and emergency
action upon adrenalin secretion and their section reduces it to traces.
The supraoptic nucleus governs secretion of the antidiuretic hormone of
the neurohypophysis and section of the supraoptic neurohypophysial
fibers is followed by insipid polyuria. Stimulation by light or desicca-
tion inhibits the melanic-expanding secretion of the pars intermedia
of the hypophysis through the medium of the nervous system. After
cutting the pituitary stalk, the anteropituitary secretions are produced
in sufhcient amount so that, generally, the thyroid, adrenal, and gonads
are not modified. But there is no increase in the secretion in cases of
emergency; and in animals with the pituitary stalk cut ovulation is no
longer produced by mating {e. g., doe), embracing (e. g.,-toad), or vision
(e. g., dove) as in normal animals. Intense light can induce the secre-
tion of enough hypophysary gonadotropins to produce hypertrophy of
the gonads when they are in the atrophic condition of winter rest.
Cold no longer exerts its action on ovarian cycles or upon thyroid or
adrenal cortex when the pituitary stalk has been cut in rats. Certain
hypothalmic injuries may decrease the secretion of either all or some of
the pituitary gonadotropins.

REGUL.\TION OF COMPLEX EQUILIBRIA

The individual regulation of each gland is at the same time
submitted to more ample regulations which are often reciprocal. Thus,
the blood sugar level is maintained constant in spite of the coexisting
secretions which tend to produce either hypoglycemia, such as insulin,
or hyperglycemia, such as those of the hypophysis and adrenal glands.
There exists, therefore, a functional equilibrium of the secretion of each
endocrine gland and, at the same time, a functional equilibrium of all

B. A. HOUSSAY

the secretions of a similar functional constellation (for example, the endo-
crine secretions which regulate sex or those that govern carbohydrate
metabolism, etc.)- These regulations are mainly humoral, depending
to a great extent on endocrine factors, although the nervous system
often has an important share.

The tendency of the endocrine glands to reach a functional
equilibrium and then to maintain it without overshooting seems to con-
form to a sort of general law. Thus, when an ovary is extirpated, the
remaining ovary produces the same number of ovules and cycles
("law of follicular constancy" of Lipschlitz). A testicular fragment
either assures the total development of the sexual characters of a cock,
or else it atrophies with no intermediate stable equilibrium being set
up (the "all or none" law of Pezard and Gley). The tendency to all
or none activity of the gland in situ, maintaining its secretion at a con-
stant level, independent of its mass, is not inconsistent with the fact
that the pharmacological effects of the hormones vary with the doses,
the relationship following the typical S-shaped curve.

The close associations existing among the glands of internal
secretion explain why the disturbances affecting one of them are
generally reflected in the others. It is rarely, if ever, that experimental
or pathological disturbances of an endocrine organ are observed with-
out modifications of the other glands. Thus, the extirpation of the
hypophysis leads to atrophy and hypofunction of the adrenal cortex,
thyroid, and gonads.

The actions of each hormone can be direct or indirect. The
testicle and androgenic hormones, for example, produce hypertrophy of
the seminal vesicles and prostate directly. The pituitary gonadotropin
(LH or ICSH) produces the same action, but since it induces the se-
cretion of testicular hormones is inactive in the absence of the testicle.

In certain cases the functional interactions of endocrine glands
are simultaneous; in other cases they follow each other in sequence,
as the proliferative (estrogenic) and secreting (progesteronic) phases
of the endometrium during the menstrual cycle.

HYPERSECRETION OF HORMONES

That the secretion of the hormones is not normally at a maxi-
mum is borne out by several facts: {a) castration very much increases
the secretion of anterohypophysary gonadotropins: {b) adrenal

200

HORMONES

ablation very much increases the secretion of pituitary corticotropin;
(c) in clinical cases of endocrine hyperfunction supernormal effects
are observed, similar in many cases to those produced by excessive
administration of hormones.

Hyperfunction of the endocrine glands is observed under
several circumstances: (a) pathologic hyperplasias and tumors (be-
nign, or adenomata, and malign) in which are produced exaggerated
secretion of either normal or abnormal hormones; (b) injection of
either glandular extracts or hormones; (c) excitation by a supernumary
gland (for example, parabiosis of a castrated animal with a normal
one); (d) rupture of the endocrine equilibrium.

In some cases either hypo- or hyperfunction takes a certain
time to develop. Thus if 95% of the pancreatic tissue of a rat is
removed, the residual tissue is still able to keep a normal blood sugar
level for two or three months, but later on a progressive diabetes de-
velops. Following certain operations on the ovary (grafting, ligation,
partial fragmentation), a hypersecretion of pituitary gonadotropin
gradually develops, which, by promoting excessive secretion of ovaric
estrogens, produces a marked hypertrophy of the uterus, hyperplasia
of the endometrium, etc. In this case a new hyjDophyso-ovaric equi-
librium is established at an abnormal level.

Types of Endocrine Functional Associations

The functional associations belong to various types. (/) A
gland, such as the anterohypophysis, can develop and maintain the
structure and function of one or several other glands, as the thyroid,
adrenal cortex, ovary, or testicle. (2) One gland can moderate the
function of another, e. g., the sexual hormones moderate the gonado-
tropic pituitary function.* (3) Actions, sometimes antagonistic and
sometimes synergetic, can be observed between two glands (or their
hormones), as in the case of the ovary (or estrogens) and the corpus
luteum (or progesterone). (4) Certain hormones increase the sensi-
tivity to others, e. g., estrogens to the effect of progesterone upon the
endometrium or upon the mammary gland, and thyroxine to the effect

* In fact, the position is more complex, because estrogens, if given in in^li
doses, may induce an increased secretion of luteinizing, adrenotropliic, and lacto-
genic hormones of the anterior pituitary, but, in larger doses, suppress them all.

201

B. A. HOUSSAY

of adrenalin. (5) Certain hormones can produce an insufficiency by
damaging an organ; thus anteropituitary extracts and, in certain
cases, those of the thyroid as well, by a repeated action bring about the
disappearance of the j8-cells of the islets of Langerhans and produce a
pancreatic diabetes, the cause of which could not be deduced by any
one not familiar with the previous history.

METHODS OF STUDY

Many methods are required to study the endocrine glands and
their hormones.

Morphological. Important data can be obtained by the study
of the weight and macro- and microscopic structure of the endocrine
organs of different ages and under different conditions, created by
extirpation of other organs and injections of hormones. It is also
necessary to watch for the initiation and localization of the changes.

Chemical. These include: {a) a search for and isolation and
purification of the hormones by means of a combination of biological
assays and chemical methods; {b) a study of the chemical changes
and metabolic modifications produced in an animal by the suppression
or administration of hormones; {c) action of the hormones on tissue
slices or on chemical systems in vitro; and {d) a study of the origin,
metabolism, and excretory products of the hormones.

Physiological, These can be subdivided into experimental and
clinical methods. The experimental method includes the study of:
{a) glandular insufficiency and restitution through grafting, implants,
or administration of either extracts or hormones; {b) hyperfunction
induced by the methods already described; {c) measurement of the
hormones in the organ, in the blood that comes away from it, in that of
the general circulation, or even in the urine. It can also be indirectly
measured by finding the amount necessary to substitute for the removed
organ (substitution method). It is much safer to measure the hormone
secreted by an organ than the amount the organ contains, because the
amount secreted does not always vary parallel with the amount present
in the organ.

The clinical study is very valuable: (a) because it furnishes
human data and {b) because the disease is a spontaneous experiment,
the conditions being sometimes more delicate and varied than the
experimental methods can secure. Many endocrine functions have

202

HORMONES

been studied first clinically and afterward experimentally, e. g., Addi-
son's disease, acromegaly, etc.

Quantitative Determination of Hormones. The hormone or
its derivatives can be determined in the organ, blood, or urine. The
determination can be made by chemical, physical, or biological
methods. The low concentration of the hormone rarely allows direct
gravimetric estimations; in general, these must be based on colori-
metric, spectrophotometric, or chromatographic methods. Biological
assays are performed on entire animals, isolated organs, or tissue slices.
Animals with insufficiency may be used. In some cases the organ is
forced to function overloaded, e. g., a pancreas grafted in diabetic
dogs by vascular grafting is tested for the time necessary to correct
hyperglycemia.

Hormones are assayed by comparing their effects with those
produced by international standards, thus avoiding differences of
sensitivity encountered in different races of animals. In order to
confirm that a given function depends on an endocrine organ, the
following consequences must be etablished: (7) the ablation or injury
of the gland must produce an insufficiency of that function; (2) this
deficiency must be compensated for by the graft or implantation of the
gland, or by the injection of its extract or its hormone; (3) an excess
of these substances must induce hyperfunction symptoms opposed to
those of hypofunction; (4) the anatomoclinical facts must agree with
the experimental findings; (5) both spontaneous and induced hyper-
function must be improved by treatments which either eliminate the
gland or decrease its action.

In order to admit that a given action due to an endocrine gland
is produced through the mediation of another gland, it is necessary to
show: (7) that the removal of the second gland is followed by a condi-
tion of hypofunction, just as, or even more, pronounced than that
caused by the ablation of the first; (2) that the disturbances cannot
be corrected either by implantation or injection of extracts or hormones
of the first gland when the second one has been removed. For example,
(a) ovariectomy leads to atrophy of the uterine and vaginal epithelium,
just as much or more so than hypophysectomy, and (b) the pituitary
gonadotropins produce hyperplasia of the uterine and vaginal epi-
thelium only when the ovary is present. These facts lead to the
conclusion that the effect of the hypophysis on the uterine and vaginal

203

B. A. HOUSSAY

epithelium is due to the induecd hypersecretion of the ovarial hormones.
To be active, certain hormones require a previous sensitization
of the receptor organ by other hormones. Thus, estradiol does not
produce hyperplasia of the adrenals in hypophysectomized rats, but
this eflfect is produced if the atrophy of the adrenals is prevented by
administration of two daily doses of anterohypophysis (Pinto).
Chorionic gonadotropin does not induce ovulation in hypophysec-
tomized rats, but does so if either serum gonadotropin or stilbestrol
is previously injected.

Metabolism of the Hormones

At present we live in a period in which the metabolism of the
hormones is being energetically studied. The studies include the
investigations of origin, transformation, and elimination of the hor-
mones.

The study of origin includes that of their precursors within the
organism or in the diet and that of the place and mechanism of elabora-
tion within the endocrine gland.

It is also interesting to know its absorption, its chemical trans-
formations, and the site of these transformations. Thus it is known
that the estrogens are destroyed principally in the liver and that
thyrotropin is transformed by the thyroid into an inactive compound
which can be reactivated at a certain temperature. In some cases the
disappearance of the hormone can be quantitatively followed.

The disappearance of the hormone from the blood has been
followed in various cases. The unchanged hormone may be elimi-
nated in the urine or, secondarily, in the milk or the bile, but in other
cases only the transformation products of the hormone are eliminated
through these routes. The urine has the advantage, for purposes of
extraction, of being a concentrated ultrafiltrate of the plasma virtually
free of protein. Hormones of protein nature may or may not filter
through the kidney according to their molecular size. Thus chorionic
gonadotropin and thyrotropin are found in the urine, but not so the
gonadotropin of the pregnant mare serum. The form in which various
steroids with estrogenic, androgenic, or corticoid action are eliminated,
as well as the total elimination of the 17-ketosteroids, are being in-
tensively studied. In some cases the metabolism of a particular
hormone can be traced by the assay of some transformation product,

204

HORMONES

e. g., that of progesterone through the urinary ehmination of preg-
nanediol.

Applications

The study of hormones is of interest lo medicine and animal
husbandry. For medicine it is important to investigate: {a) the
physiological and nutritive conditions which will assure hormonal
equilibrium and which will secure better physical and mental health
during rest, exercise, or work; (b) the prevention and treatment of
the endocrine disturbances and esjiecially the more common disturb-
ances such as diabetes, endemic goiter, endocrine disturbances in
women, etc., by genetic, dietetic, and pharmacological methods; (c)
the prevention and treatment of disturbances due to abnormal internal
secretions, e. g., nephrogenic hypertension. The distinction between
specialists in diseases of metabolism and specialists in diseases of the
endocrine organs is quite artificial. Diabetes, for example, is the most
typical endocrine disease. The glands of internal secretion are the
regulators of metabolism and it is impossible to study cither endo-
crinology independently of metabolism or metabolism independently
of endocrinology.

The study of the hormones is germane to the problem of animal
production because it gives a clearer understanding, and thereby wider
possibilities, of controlling phenomena such as heat, ovulation, fer-
tilization, pregnancy, number of offspring, breeding without de-
pendence on factors such as lactation, castration, time of the year, etc.

The study and production of hormones has been converted into
a problem of national importance. There are enormous commercial
interests involved and a great number of technicians devoted to the
search, production, and commerce of hormones. This raises the
danger of both excessive and inadequate use of the hormones and of
exaggerated propaganda.

At the present moment, the main problems of experimental
endocrinology may be classified in the following groups: (7) isolating
pure hormones and studying their actions, either separated or asso-
ciated, simultaneous or successive; (2) establishing the mechanism
of action of each hormone to determine whether they are direct or
mediated by other organs; (J) studying each organ's secretion from
the standpoint of the regulating factors and of its relations witli the

205

B. A. HOUSSAY

regulatory mechanisms of other glands and with the wider systemic
functions; (4) the metabolism of the hormones; (5) clinical applica-
tions to prevent and cure endocrine diseases; (6) applications in
animal industry.

This exposition of the hormones for the sake of brevity has been
necessarily general in character and, therefore, incomplete. The
author had to avoid the double risk of being too elementary or too
tedious. Briefness courts the danger of being dogmatic or of not pro-
viding the factual basis for each statement. This essay has been written
with a view to presenting to scientific laymen, but not to specialists in
endocrinology, the current position of the problem of the hormones.

Selected Bibliography

Barker, L. F., Hoskins, R. G., and Mosenthal, H. O., Endocrinology and Metabo-
lism. 5 vol., Appleton, New York, 1922.

Bayliss, W. M., and Starling, E. H., "The chemical correlation of the secre-
tory process," Proc. Roy. Soc. London, 67, 310-322 (1940). "Die che-
mische Koordination der Funktionen des Korpers," Ergeb. Physiol., 5, 664
(1906).

Berthold, "GeschlechtseigentiimHchkeiten," Wagners Handwort. Physiol., 1,
507 (1842). "Transplantation der Hoden," Arch. Physiol., 1849, 42.

^\cd\. A., Innere Sekretion. 2nd ed., 2 vol., 1913; 4th ed., 3 vol., 1919-1922.
Urban & Schwarzenberg, Vienna.

Del Castillo, E. B., Reforzo Membrives J., De La Baize, F., and Galli Mainini
C., Endocrinologia Clinica. El Ateneo, Buenos Aires, 1944.

Glandular Physiology and Therapy. Am. Med. Assoc., Chicago, 1935. 2nd ed.,
1942.

Gley, E., Les skrUions internes. Bailliere, Paris, 1914, 94 pp., 3rd ed., 1925.

Hirsch, M., Handbuch der inneren Sekretion. 3 vol., Kabitzsch, Leipzig, 1928-
1933.

Lucien, M., Parisot, J., and Richard, G., Traite d'endocrinologie. Doin, Paris,
1925 and 1934.

Pende, N., Endocrinologia. 4th ed., 2 vol., Vallardi, Milan, 1934.

Schaeflfer, E. A., "On internal secretions," Lancet, 2, 321, 324 (1895). The
Endocrine Glands, Longmans, Green, London, 1916.

Starling, E. A., "The chemical correlation of the functions of the body,"
Lancet, June (1905).

Trendelenburg, P., Die Hormone. 2 vol., Springer, Berlin, 1929, 1934.

Vincent, S., "Innere Sekretion and Driisen ohne Ausfiihrungsgang," Ergeb.
Physiol., 11,218 (1911).

2o6

14

FUNDAMENTALS OF

OXIDATION AND

REDUCTION

LEONOR MICHAELIS, member emeritus, the rockefeller

INSTITUTE FOR MEDICAL RESEARCH

/

P"N ATTEMPTING to define the essential concepts involved
in a problem, generally it is found that, because of the
flexibility and, often, ambiguity of language, a definition cannot be
formulated with perfect clarity. Nature is not so constructed
that one can classify all its subject matter within a finite number of
distinctly circumscribed terms. A great deal of confusion has arisen,
and will henceforth arise again, from the fact that an author may use
a given term according to one definition, and then during the dis-
cussion consciously or unconsciously forget that definition. Further-
more, it is a frequent fate of a definition that it be based on an assump-
tion which later appears to be either erroneous or, at least, unsuitable.
If the term, as commonly happens, is redefined according to the change
in the underlying fundamental concepts, the new definition is likely
to be in conflict with the older one.

A typical instance of the flexibility of a concept is the term
oxidation, and its reverse, reduction. Originally, oxidation meant
combination with oxygen; the combination of hydrogen with oxygen
to form water is the prototype of oxidation in this sense. According
to this definition, the combination of hemoglobin with oxygen to form
oxyhemoglobin should be the simplest, purest, and most unambiguous

207

LEONOR MICHAELIS

case of oxidation in organic chcinisliy. However, according to our
present usage of the term, this reaction is so dissimilar from what is
now considered to be a typical oxidation that it is no longer classified
among the oxidations — it has, in fact, been termed "oxygenation" by
Conant. The real oxidation of hemoglobin, according to modern
definition, is its conversion to methemoglobin, because, according to
the present stage of knowledge and on the basis of our present model
of atomic and molecular structures, the oxygen of oxyhemoglobin is
attached to hemoglobin without aflfecting the electronic structure of
the iron atom,* whereas in methemoglobin no oxygen is attached to
hemoglobin at all — rather, the iron atom of hemoglobin contains one
electron more than the iron atom of methemoglobin. The removal of
that electron from hemoglobin is now considered as typical for its
oxidation. When ferrous chloride reacts with chlorine, ferric chloride
is formed. Since ferrous compounds can be converted to ferric com-
pounds by oxygen also, ferric iron has always been considered an oxi-
dation product of ferrous iron. Yet, if this conversion is brovight
about by chlorine, oxygen plays no part in the "oxidation."

What is common in most processes formerly designated as oxi-
dation, disregarding such an exceptional case as the formation of oxy-
hemoglobin, can only be stated in terms that would have been quite
incomprehensible to the originators of the concept of oxidation. This
common property can be defined, in terms of the present state of the
atom model, by saying that, after its oxidation, a molecule has been
deprived of one electron, or of two electrons; these two cases of oxida-
tion are distinguished by the terms "univalent" and "bivalent." The
oxidation of ferrous ion to ferric ion, or of ferrocyanide ion, [Fe(CN)6]^~
to ferricyanide ion, [Fe(CN)6]^~, are examples of univalent oxidation.
The oxidation of stannous ion, Sn"'""'', to stannic ion, Sn^+, is a bivalent
oxidation. In those cases in which the oxidation, or, in other words,
the withdrawal of the electron, is brought about by oxygen, the oxygen
is the acceptor of the electron. The fate of oxygen after acceptance
of the electron will be discussed on page 220. Other reagents, such as

* This statement involves another difficulty based on facts unknown until
recently. Since hemoglobin is paramagnetic, but oxyhemoglobin diamagnetic,
as Pauling and Coryell have shown, some change in electronic structure due to
the attachment of oxygen must be postulated here too. But it is not of the type
one would now call oxidation.

208

OXIDATION AND REDUCTION

chlorine, can also accept :\n electron and thus "oxidize'' other sub-
stances. Oxygen, therefore, is only one among many molecular species
which can withdraw an electron from other molecules and so "oxidize"
them.

One must remember that the transition of the original definition
of oxidation to the modern one has been gradual. Historically, there
has been a transition stage inaugurated by Wieland, who developed
his concept mainly with respect to oxidation of organic molecules.

/H /H

When alcohol, CH3C— H, is oxidized to aldehyde, CH3C , the over-

\OH ^O

all eflfect is the loss of two hydrogen atoms. According to the original
definition of oxidation, one may say that the primary process is the

addition of one oxygen atom to alcohol in order to form CH3C — OH

\0H

which, by splitting off one molecule of water, becomes CH3C

^O
Wieland's suggestion is that the process does not pass through the stage
of an addition of oxygen, but that what had been designated as oxida-
tion is in fact a withdrawal of two hydrogen atoms. The oxidizing
agent is the acceptor for the hydrogen atoms; for such cases, he re-
places the term "oxidation" by "dehydrogenation."

The reconciliation of this idea with the more modern one can
be based on the fact that a hydrogen atom consists of a positively
charged proton and a negatively charged electron. Then, one may
say that the withdrawal of two electrons is essential for oxidation of
alcohol. Since the two protons, which should remain, are no longer
held by any noticeable force, they are detached also, and become
bound to some proton acceptor, such as water, with which they form
the hydrogen ion, OH3+, as it exists in the presence of water (called
also the oxonium ion); or they may be bound to some anion that
may be present, such as the acetate ion, GH3GOO~, with which they
form an "acid," CH3COOH. This hypothesis should not involve the
idea that the expulsion of the electron occurs first and the expulsion of
the proton thereafter. The sequence is undecided. The principle is,
rather, that the withdrawal of each hydrogen atom is the same as the

209

LEONOR MICHAELIS

withdrawal of an electron together with a proton; this process is
dehydrogenation. The reconciliation with the modern definition is
the postulate that only the withdrawal of the electron is characteristic
for oxidation, and that the simultaneous detachment of a proton does
not belong to the process of oxidation proper.

The attachment of a proton is termed a change in the level of
"acidic ionization." This will become clear if one first considers an
example showing the essential features in a very obvious way. Let us
start with quinone (I), the reduction of which, according to modern
concepts, consists essentially in the attachment of two electrons. The
result is formula II, a doubly negatively charged ion of hydroquinone.

O

II
/\

Y

o

~

o-

~

OH

1

JA

/N

A

V

o-

Y

OI

I

V

OH

li III IV

quino

ne ion Hydroq

uino

ne anion Hyd

^oquinone

Quinone

If the reaction occurs in a strongly alkaline solution, the result of
oxidation is molecule II. When the reaction occurs in a less alkaline
solution, oipH about 9 to 10, one proton only is attached to II, and the
univalent anion of hydroquinone is formed (III). When the reaction
occurs in an acid solution, two protons are attached to II, and (union-
ized) hydroquinone (IV) is formed. Depending upon the pH, either
an electron or a full hydrogen atom may be attached to each oxygen.
We now agree on the definition that II, III, and IV are on the same
level of oxidation-reduction as, but on lower levels of oxidation than,
quinone (I). The diff"erences between II, III, and IV are not in their
level of oxidation, but in their level of acidic ionization.

It is easy to transfer these ideas to the case of the oxidation of
alcohol. On detaching two electrons only, one obtains structure V,
where two protons exist in the molecule. The two protons, however,
are held still less firmly than in hydroquinone ion IV, and under all

CH3C^H +
\OH +

V

CH3C'
VI

<

210

OXIDATION AND REDUCTION

conditions possible, even in extremely acid solution, V changes its
acidic ionization level by expelling the two protons and forming acet-
aldehyde (VI). • Here again we do not claim that V is formed first
and VI subsequently. It is, however, emphasized that the process
of oxidation proper is the detachment of electrons, and that the simul-
taneous loss of the two protons, although it may be a necessary con-
sequence, does not belong to the process of oxidation itself. Dehy-
drogenation is thus a special case of oxidation.

The term dehydrogenation need not be discarded but may be
used advantageously for those cases in which the release of an electron
involves the simultaneous release of a proton. Whether or not this
simultaneous release of the proton occurs depends on the pH of the
solution. The term dehydrogenation can be reserved for the cases in
which even in extremely alkaline solution the release of an electron is
accompanied by a release of a proton. The conversion of succinic acid
(VII) to fumaric acid (VIII) may be called a dehydrogenation because
the intermediate stage (IX) is not capable of permanent existence.
Molecular species VIII may be considered as an infinitely strong acid,
which even in an acid solution is not capable of holding on to protons.
Therefore the enzyme which catalyzes the reaction VII -^ VIII may

COOH COOH COOH

I I I

CH2 CH CH-H +

I II I

CH2 CH CHH+

I I I

COOH COOH COOH

VII VIII IX

be justly termed a dehydrogenase. When quinone (I) is reduced in
an acid solution, we may speak also of dehydrogenation, whereas in
the reduction of quinone in a more alkaline solution (IX and VIII)
the synonymous use of the terms oxidation and dehydrogenation would
involve a more generalized definition of dehydrogenation.

Analogously, if we define oxidation as the loss of an electron,
we are compelled to generalize the definition by the amendment that
this definition holds whether or not a proton is also released; so
whether one uses "oxidation" or "dehydrogenation" is a matter of
nomenclature only. In this essay, we shall use "oxidation" and
"reduction," and not "dehydrogenation" and "hydrogenation." For

21 I

LEONOR MICHAELIS

such reactions as the conversion of benzene to cyclohexane, however,
it is recognized that one may prefer to use the term hydrogenation
instead of reduction. As regards the chain of oxidation reactions as
they occur in respiration, it is often said that hydrogen atoms are
transferred from one molecular species to another. However, the
oxidation of reduced cytochrome to cytochrome involves not the
transfer of a whole hydrogen atom, but of one electron only. It is
therefore not entirely true that the chain of respiratory processes con-
sists in transferring hydrogen atoms from one molecular species to
another, and finally to oxygen, but it is true that this chain consists
in transferring electrons from one molecular species to another. Some-
times the transfer of the electron is accompanied by a transfer of a
proton and sometimes it is not.

It may be added that the loss of a hydrogen atom in dehydro-
genation is equivalent to the addition of a hydroxyl group, as far as
the level of oxidation-reduction is concerned. When alcohol is oxi-
dized to acetaldehyde, the process may be alternately described as
follows. Alcohol detaches one hydrogen atom and accepts a hydroxyl
group:

/H /OH /O

CHsC^OH > CH3C— OH > CHsC^OH > CHsC^^ + H2O

\H \H \H \H

Here, the first step is the detachment of the hydrogen atom, and the
second, attachment of the hydroxyl group, while the process formerly
was described in terms of loss of two hydrogen atoms. So, a'^hydro-
genation is equivalent to "hydroxylation," and both terms can be
avoided by describing the level of oxidation proper only in terms of
electrons ejected or added, whether or not a proton or a hydroxyl ion
is also involved in the process.

Stepwise Oxidation and Reduction of Organic Compounds

If one wants to formulate a simple example of a possible step-
wise oxidation of an organic compound, one might propose the series:

CH4 > CH3OH > CH2O > HCOOH > CO2

each step representing a bivalent oxidation. A univalent oxidation is
not imaginable unless one abandons the a.ssumption that carbon is

212

OXIDATION AND REDUCTION

tetravalent, oxygen bivalent, and iiydrogcn univalent: the uni\alenl
oxidation of CH4 would give, as a step intermediary on the way to
CH3OH, the free radical CH3 with "tervalent" carbon. To be sure,
such free radicals have been recognized for a long time, but those
known until recently always have one of two properties: either they
are very unstable and have an extremely short lifetime (e. g., CH3);
or they can be produced only from a very restricted number of com-
pounds and only in a solution of a perfectly water-free organic solvent.
The famous discovery of triphenylmethyl by Gomberg in 1890 pro-
vided a prototype of such free radicals.

It will now be shown that all oxidations of organic molecules,
although they are bivalent, proceed in two successive univalent steps,
the intermediate state being a free radical, and furthermore, that
according to structural conditions these intermediate free radicals may
be either just as unstable as CH3, somewhat more stable, or even
perfectly stable compounds.

First, however, the concept of stability must be discussed. The
criterion of stability may be obtained in two ways. A substance may
be said to be unstable if it rapidly undergoes a chemical change when
exposed to such ubiquitous reagents as air; pyrogallol, for instance, is
unstable in an alkaline solution because it is readily oxidized by
oxygen, but is stable in the absence of oxygen. Or, a substance may
be said to be unstable if it undergoes a change even in the absence of
any foreign substance with which it might react; acetaldehyde, for
instance, undergoes the Cannizzaro reaction in an alkaline solution,
one molecule of aldehyde being oxidized to form acetic acid, while
another is reduced to ethyl alcohol. It will now be shown that the
alleged instability of free organic radicals is, in very many cases,
essentially due to the latter, bimolecular interaction. It will be
shown that the bivalent oxidation in fad occurs in two successive
univalent steps as in the example of duroquinone (X).

The methyl groups in the ring of X prevent the secondary,
irreversible reactions which are of no interest in this discussion, and
which would occur when working with the unmethylated, simple
benzoquinone. When duroquinone in an alkaline solution is reduced,
as by hydrogen plus palladium, or by any other suitable reducing
agent, the faintly yellow solution turns brown first, then colorless.
The colorless compound is the corresponding durohydroquinone, which

213

LEONOR MICHAELIS

in extremely alkaline solution can be written as formula XI and which

O -O OH

i k

CH3— /N— CHs

CH3—

/ \

— CH3

CH3—

Y

0

X

— CH3

CH3

\

-CHs

— CH3

— CH3

-o

XI

in less alkaline solution would add first one, then another, proton to
form XII. This is the customary bivalent reduction. The inter-
mediate, brown substance is the result of a univalent reduction; and
it can be shown by various methods, to be described later on (see
pages 215 and 217), that it has the same molecular size as XI and
differs from it only in so far as one oxygen atom is negatively
charged and not the other. This brown substance is a free radical.
One might say that one oxygen atom is bivalent and the other uni-
valent, as in XIII. The same molecular species, in a more acid solu-
tion, would have formula XIV.

O O

GHj
CH,

^.

-CH3
-CH3

CH3
CH3

,-V^-CH3
-GHs

o-

XIII

OH

XIV

Let us consider first the reaction in a strongly alkaline solution.
When the solution of the quinone is mixed with increasing amounts
of a reducing agent, there will always be, except for the very begin-
ning and the end of the oxidation, a mixture of the quinone, the hydro-
quinone, and the intermediate free radical which we shall call semi-
quinone; and an equilibrium is established between these forms. It
is important to emphasize that the equilibrium is established instan-
taneously, and that there is no sluggishness in its formation, as is usually
the case in the formation of equilibria with organic compounds (except
acidic ionizations). The activation energy of this reaction leading to
equilibrium is extremely small. So, in the solution, the intermediate
free radical is never present without being in mixture with the quinone

214

OXIDATION AND REDUCTION

and the hydroquinone. If we designate the hydroquinone (the re-
duced form of this system) as R, the semioxidized form (the free radical)
as S, and the totally oxidized form (the quinone) by T, the equilibrium
is established according to the reversible reaction

2S , 'R + T (1)

and the constant of equilibrium, which may be called the "semi-
quinone formation constant," is:

_[S?_
[R][T]

Its reciprocal may be called the "dismutation" constant, because re-
action (1) is a "dismutation" (or "disproportionation") of the free
radical. If k is very small, very little of the radical exists in the state
of equilibrium; if k is large, much of the radical can exist, and it can
be distinguished by its particular color and by its paramagnetism,
which will be discussed presently.

Experience has shown for the case of duroquinone that k is very
large in an alkaline solution, but very small in an acid solution. The
transition of the behavior from alkaline to acid solution is continuous;
and there can be no doubt that in acid solution a small amount of the
radical is formed too, even if its concentration is too small to be notice-
able by ordinary methods. Thus we are obliged to state that the
ionized form of the radical (XIII) is a rather stable compound, and
the unionized (XIV) rather unstable, "stability" being judged accord-
ing to its capability of existing in equilibrium with its parent substances,
the quinone and the hydroquinone.

Why is XIII more stable than XIV? In formula XIII, the
negative charge has been arbitrarily attributed to the upper oxygen
atom, but it can just as well be attributed to the lower one. In fact,
the location of the charge is undecided, for it oscillates between the
two extreme positions through the chain of the atoms in the ring.
Such a condition has been termed resonance. The two limiting struc-
tures, one with the charge on top, the other with the charge at the
bottom, are indistinguishable molecular species. One speaks about
"equivalent" or "symmetrical" resonance, a condition which, accord-
ing to quantum mechanics, contributes largely to the stability of a
molecule. In form XIV, no such equivalent resonance prevails. The

215

LEONOR MICHAELIS

proton attached holds the negative charge rather tightly in place.
This molecule does exist but to a much smaller extent than XIII be-
cause it is less stable. Free radical formation is a general reaction
whether it takes place extensively, as in alkaline solution, or in traces,
as in acid solution.

It can be shown by numerous examples that the same behavior
is true for organic compounds which form reversible oxidation-reduc-
tion systems, such as many dyestuffs, e. g., methylene blue, or phenol-
indophenol. No observable amount of any intermediate free radical
can be demonstrated for irreversible oxidations, e. g., when alcohol is
oxidized to aldehyde. So we are permitted to conclude that the
formation of a free radical, in sufficient concentration, in the equi-
librium involved, is a prerequisite for the reversibility of an oxidation-
reduction system.

Evidence for Free Radical Nature
of the Intermediate Substance

It is appropriate to discuss at this point the evidence for the
assertion that the intermediate steps of oxidation are really free radicals
of the same molecular size as the parent substance and not dimeric
valence-saturated compounds made up in such a way that two radicals
are combined to form a bond which abolishes the state of "unsatura-
tion." Such a reaction may be imagined to occur as follows:

2 S > D (2)

That is, two molecules of the semiquinone radical, S, combine with
each other to form a dimeric compound, D, which no longer has the
characteristics of a free radical, just as two "radicals" H (hydrogen
atoms) combine as follows:

2.H > H2 (3)

In fact, such reactions do occur, and an equilibrium is established, so
that, instead of reaction (2) we should write:

2 S , D (4)

Depending upon the conditions, which will not be discussed here, this
equilibrium is sometimes greatly in favor of the free radical, sometimes

216

OXIDATION AND REDUCTION

greatly in its disfavor. The main point is that reaction (2) does not,
as a rule, proceed to completion, but that the free radical really does
exist, and often even to such an extent that the dimerization is negli-
gibly small. There are two powerful tools to decide whether the
intermediate product is S, D, or an equilibrium mixture of the two,
viz-, potentiometric titration and magnetic measurement.

When a substance such as duroquinone, the example discussed
above, is titrated with a reducing agent and the oxidation-reduction
potential is plotted against per cent reduction, a curve is obtained the
shape of which will depend markedly on whether the intermediate
substance is R or D. Straightforward application of the law of mass
action yields a complete theory as to the shape of the titration curve.
The calculations involved, although not absolutely simple, are of such
a nature that only a high-school student might consider them in the
realm of "higher" mathematics. Application of the law of mass action
has shown in many cases that the intermediate substance is a free
radical, almost exclusively under certain conditions and, under other
conditions, in equilibrium with its dimer. Furthermore, according
to a theory first developed by G. N. Lewis and later amply confirmed
by the quantum theory, a free radical, because it always contains an
odd number of electrons, must always be paramagnetic, in contrast
to the ordinary, valence-saturated organic compounds, which are
diamagnetic (provided they do not contain metal atoms such as iron
or cobalt). When a solution of duroquinone is slowly reduced in
alkaline solution, by glucose, say, evidence can be produced for the
gradual appearance of a paramagnetic molecular species and, on
further reduction, for its disappearance. The paramagnetism is due
to the spin of the odd electron, while in ordinary molecules the elec-
trons always occur in pairs with opposite spin, whereby their para-
magnetic effect is quenched.

An elegant method of detecting such free radicals, even of low
stability, has been established by G. N. Lewis. In the kind of experi-
ments mentioned above, ejection of an electron is brought about by
an oxidizing chemical, which serves as an acceptor of electrons. Lewis
eflfects ejection of the electron by ultraviolet light, the substance being
dissolved at the temperature of liquid air, in an organic solvent which
has, at this temperature, a rigid, glasslike consistency without crystal-
lizing. In such a rigid medium, no molecular collisions can occur

217

LEONOR MICHAELIS

and free radicals, once they have been established, have no chance to
undergo bimolecular reactions by which they may be rapidly elimi-
nated. Radicals can be preserved, in this manner, to an extent w^hich
by far exceeds their equilibrium concentration postulated by thermo-
dynamics. Lewis showed that many organic compounds exposed in
this way to ultraviolet radiation become colored; in suitable cases he
identified, by spectrophotometrical comparison, the colored substance
with the free radical obtained previously by chemical oxidation. The
peculiar merit of this method is the fact that free radicals can be de-
tected in the cases in which they would be unnoticeable in the state of
equilibrium because of their low concentration. On the other hand,
since the appearance of color is not necessarily evidence for a free
radical, a scrutiny of the phenomenon is necessary for each individual
case.

In order to arrive, from these considerations, at the discussion
of the final problem we have in mind, the varying degree of inclination
of an organic substance toward oxidation (or reduction) must be con-
sidered. Here we distinguish two essentially different properties.
To state that a substance is easily, or difficultly, oxidized by an oxi-
dizing agent may mean one of two things: that the speed of such an
oxidation is great or small (a topic belonging to a discussion of chemical
kinetics); or that the final state attainable by the interaction of the
oxidizing agent and the oxidizable substrate is a complete, 100%
oxidation, an incomplete oxidation, or almost no oxidation at all
(a problem of thermodynamics).

We shall start with the problem in thermodynamics, in par-
ticular, the case in which both the oxidizing agent and the substance
to be oxidized form reversible oxidation-reduction systems. If leuco-
methylene blue is the substance to be oxidized and potassium ferri-
cyanide the oxidizing agent, when the two are mixed in equivalent
proportions practically all of the leuco dye is oxidized to the dye and
all of the ferricyanide is reduced to ferrocyanide. But when leuco-
methylene blue is mixed with an indophenol dye, both the oxidation
of leucomethylene blue and the reduction of indophenol will be in-
complete, and the final state is a mixture of four substances, the two
leuco dyes and the two dyes. Furthermore, if leucomethylene blue is
mixed with safranine, no change will occur, at least within practical
measurable limits. Safranine does not oxidize leucomethylene blue

2l8

OXIDATION AND REDUCTION

to any appreciable extent. On the other hand, leucosafranine readily
reduces methylene blue. With this in mind, one can set up a sequence
of all reversible oxidation-reduction systems according to their oxida-
tive power. For the four systems mentioned, the sequence is: ferri-
cyanide, indophenol, methylene blue, safranine.

A quantitative expression of the oxidizing power is the "oxida-
tion-reduction potential" of a dye in mixture with its leuco dye. Vari-
ous substances capable of reversible oxidation and reduction can be
arranged in a sequence according to their oxidation-reduction, or
"redox," potentials. Each membei', when present in the reduced
state, can be oxidized by any following member present in its oxidized
state; and each member, when present in the oxidized state, can be
reduced by any preceding member present in its reduced state.

Let us now discuss the kinetic aspect of the problem. If we are
dealing only with reversible redox systems, as above, establishment
of the equilibrium is always so rapid that the reaction may be con-
sidered almost instantaneous. But this is not the case, in general, if
an irreversible reaction occurs. In the oxidation of alcohol to acetal-
dehyde, for example, to oxidize alcohol, a powerful oxidizing agent
such as chromic acid must be used; and to reduce acetaldehyde, a
powerful reducing agent such as sodium amalgam must be used.
Although an excess of oxidative (or reductive) power must be applied
in order to make the reaction proceed, the reaction is sluggish. Addi-
tional energy is required far beyond the quantity expected on a purely
thermodynamic basis because, obviously, an obstacle has to be over-
come. Despite the fact that the reaction between alcohol and chromic
acid releases energy, energy must first be spent, which is of course
eventually released again; and this extra energy is called the activa-
tion energy. Although the path of energy is, as a whole, downward,
it must first pass over a hill. Generally an activation energy is required
for all bimolecular chemical reactions. In reversible reactions the
energy of activation is very small, interaction occurring only when the
two molecules "collide." The collision is impeded by the fact that
molecules of any kind, on approaching each other in the course of
thermal motion, will exhibit a mutual repulsion, and only those mole-
cules which happen to have enough kinetic energy to overcome the
repulsion will really collide and react with others.

In irreversible o.xidations another, more serious, impediment

219

LEONOR MICHAELIS

occurs. We have presented the postulate that all oxidations proceed
in a sequence of univalent steps. The first step of oxidation of alcohol
would lead to the free radical, in this case, an utterly unstable molecule,
in the thermodynamic sense. To generate these radicals, an oxidizing
agent of very high potential is required ; and even then their concen-
tration remains small, so small that no direct evidence for their exist-
ence is available. The free radicals, once generated, will then react
by a dismutation:

2 radicals ^ ^ 1 alcohol -}- 1 aldehyde

The velocity of the latter reaction depends, among other things, on
the concentration of the molecules which are interacting with each
other, and therefore on the square of the concentration of the free
radicals. If this concentration is very small, it may be the limiting
factor for the over-all process of oxidation of alcohol. We may say
that the energy of activation for the oxidation of alcohol is essentially
the energy necessary for the formation of the free radical. Unless
the radical is relatively stable, as in reversible processes, the activation
energy is very great. This high energy of activation is the reason why
so many organic compounds are "stable." If all thermodynamically
possible reactions could proceed unhampered there would be no such
thing as organic chemistry.

Inhibition due to high activation energy occurs only when the
attainable concentration of the free radical is the limiting factor for
the rate of the over-all reaction. It does not matter whether the
attainable concentration of the free radical is 1 Af or, say, 10~* M.
Factors other than the concentration of the free radical, such as the
specific constants of reaction velocities, will determine the rate of the
reaction. However, if the concentration of the free radical is so small
as to be the limiting factor of the over-all process, sluggishness and
irreversibility will arise. This consideration fulfills an important re-
quirement for the understanding of reaction rates — it reduces a problem
of kinetics to one of thermodynamics.

Oxygen as an Oxidizing Agent

These considerations also explain why oxygen is such a sluggish
oxidizing agent despite its very large oxidative power in a thermo-

220

OXIDATION AND REDUCTION

dynamic sense. If it can be reduced only in successive univalent
steps, these steps must be:

O2 > O2- »02-- >Oj" ^O*"

Two of these steps are chemically identifiable. O2 is, after accept-
ing two protons, O2H2, hydrogen peroxide. ©2" is, after accepting
four protons, two molecules of H2O. However ©2" (or O2H), and
02~ (which may be written as O2H3, or OH + H2O) are intermediate,
utterly unstable steps. Since the reaction must pass through these
unstable steps, the activation energy involved in the reduction of O2
is very high.

How is this activation energy overcome when oxygen does
oxidize a substance? Overcoming the activation energy by working
at high temperatures is a usual procedure in the laboratory but is not
feasible under physiological conditions. Here the answer is that
oxygen reacts with an oxidizable substance very often not only by
means of a collision of the molecules but also by the establishment,
after collision, of a relatively stable addition compound which can
then undergo intramolecular redistribution of electrons.

Certainly no claim is made that this mechanism is always the
one involved in activation of oxygen. However, it is one of the possible
mechanisms and very likely is correct for the particular case to be de-
scribed in detail. In all probability it is the inechanism by which
oxygen is activated in all those cases in which a heavy metal com-
pound, especially of iron or copper, acts as activating catalyst.

It has been shown that, at least in an acid solution, the oxida-
tion of a leuco dye, or of cysteine, and many other substances, by
means of free oxygen is accomplished, at least with any appreciable
speed, only in the presence of a trace of an iron or copper salt. These
metal atoms have two essential properties which render them useful
for their catalytic action. First, they readily change their valence;
iron may be bivalent or tervalent and copper, univalent or bivalent.
Second, these metals are highly inclined to form complex compounds
of the Werner type. The nature of such metal complex compounds
may be demonstrated as follows. The doubly positively charged
ferrous ion, Fe+"^, can combine, first of all, with two negatively charged
univalent ions to form a saltlike compound, for instance, with two
cyanide ions:

221

LEONOR MICHAELIS

Fe++ + 2 CN- > Fe(CN)2 (a)

However, the combining power of the ferrous ion is not exhausted by
this reaction based on opposite charges. In fact, the molecular species
Fe(CN)2 has never been shown to be capable of existence. More
than two cyanide ions are attached to the iron due to the fact that
each CN~ ion has one pair of electrons not used for chemical bonding.
Each of such electron pairs can be shared with the iron to fill up its
outermost incomplete electron shell to a complete shell, as in a noble
gas. In addition to the two CN~ ions of equation (a), four more can
be attached, which contribute four negative charges to the complex
molecule, which is a ferrocyanide ion:

Fe(CN)2 + 4 CN- > Fe(CN)J~ (b)

The six CN groups are arranged around the central iron atom as the
corners of an octahedron. Fe is said to possess "six coordination
places" which may be occupied by atoms or atom groups. In analogy,
when ferrous ion combines with cysteine, which we may write, briefly,
as RSH, * we may imagine that, primarily, a saltlike compound
between iron and two molecules of cysteine is formed in such a way
that the two hydrogen atoms of the sulfhydryl groups are replaced
by an iron atom:

Fe++ + 2 RSH » Fe(RS)2 + 2 H++ (c)

This scheme accounts, so far, for the saturation of two (of the six pos-
sible) coordination places. Now atom group R contains another
atom with an unused electron pair, viz-, the nitrogen atom of the
amino group. Thus, the two molecules of RSH will occupy four
coordination places. Probably because of the large size of the RSH
molecule and steric hindrance involved in it, the two remaining
coordination places cannot be occupied by a third molecule of RSH.
In fact, no ferrous complex of cysteine can be prepared with more

* SH is a sulfhydryl group, and R represents the rest of the molecule,
which is, altogether:

SH NHj

1 I
H2C— C— COOH
H

222

OXIDATION AND REDUCTION

than two molecules of cysteine for one iron atom.* The two re-
maining coordination places may be filled in by other atoms or atom
groups of small size having an unused electron pair. One example
of such an atom group is carbon monoxide, CO. In fact, the coordi-
nation compound, Fe^^(RS)2(CO)2, can be readily prepared by the
interaction of Fe++, cysteine, and carbon monoxide in the form of its
well crystallizable alkali salt.f Just like CO, O2 also has (at least)
one unused electron pair, and may combine instead of CO, as in the
case of hemoglobin which combines either with O2 or with CO. In
hemoglobin, four coordination places of iron are occupied by the four
nitrogen atoms of the porphyrin ring, a fifth by protein, and the sixth
can combine with O2 or with CO. So we arrive at hypothetical
iron-cysteine-oxygen complexes, analogous to the well-known carbon
monoxide complex, Fe"(RS)2(CO)2. One cannot tell whether one
oxygen molecule is attached at one or at two coordination places, or
whether even two oxygen molecules can be attached. Suffice it to
imagine the complex Fe"(RS)202. This complex is said to be hypo-
thetical because it cannot be prepared, and undergoes a redistribution
of electrons, or, in other words, intramolecular oxidation-reduction,
which may be symbolized in this way: The original, oxygen-con-
taining complex may be imagined to consist of the following con-
stituents: Fe++ + 2 RS~ + O2. The electron redistribution will
occur thus:

Fe3+ + 2 RS- + O2" (O2 withdraws one electron from Fe + +) (d)

Fe3+ + RS- + RS + 02~ (O2- reduced to 02~, /. e., H2O2, by

withdrawing one electron also from one RS ") (e)

Fe++ + RS + RS + O2 — (Fe3+ withdraws one electron from RS") (f)
An alternative scheme, probably of equal probability, is:

* It is interesting to compare cysteine complexes of cobalt with those of
iron. For the cobaltous state, no complex with more than two cysteine molecules
can be obtained. For the cobaltic state, both a complex with two and another
with three molecules of cysteine can be prepared. The cobaltous complex, then,
is quite analogous to the ferrous complex. The cobaltic complex stands no com-
parison, because the ferric-cysteine complex is too unstable, due to the rapid
intramolecular rearrangements to be described presently.

t The formula in the first footnote on page 222 shows that atom group R
contains a carboxyl group. It is by means of the two carboxyl groups in the com-
plex that alkali salts can be formed.

223

LEONOR MICHAELIS

Fe3+ + 2RS- + O2- (g)

Fe + + 4- RS- + RS + O2- (h)

Fe++ + RS + RS + O2— (i)

step (f) being identical in both cases. Now, the complex containing
the constituents as in (f) is unstable and disintegrates to form three
separate molecular species: Fe++; RSSR, cystine; and hydrogen
peroxide. The latter may be used to oxidize more cysteine (stoichio-
metrically, not catalytically), or to oxidize Fe++ to Fe^+j and this
Fe3+ may oxidize (stoichiometrically) more cysteine. Finally, all the
iron is again in the ferrous state and the whole cycle is repeated with
iron thus acting as a catalyst.

It is an essential prerequisite of this cycle that the change from
the ferrous to the ferric state occurs readily and reversibly. In using
cobalt instead of iron, the first stages are similar; but, once the cobaltic
complex has been established, it shares with all cobalt complexes of
the Werner type the property that the cobaltic state cannot be readily
reduced to the cobaltous state, even by means of rather strong reducing
agents. The final result is, therefore, the formation of the cobaltic
complex, stoichiometrically, without starting a catalytic cycle. Cop-
per, but not cobalt, can replace iron as a catalyst.

What is furthermore essential in this process is the fact that
each single step in this chain reaction consists of the transfer of a single
electron. This assertion is more than a mere hypothesis. Since the
change of ferrous to ferric state involves one electron only, the sub-
division of the over-ail process into one-electron transfers is obvious.
It is remarkable that, even for such a simple case of iron catalysis, the
whole chain is of such an intricate nature, allowing for different path-
ways leading to the same final result.

Oxidation Catalysts and Enzymes

, The physiologically occurring catalysts (or enzymes) for oxida-
tion or reduction are characterized by their specificity. All oxidation
enzymes have been recognized as compounds of reversible redox sys-
tems and a specific protein. The same redox system, when attached
to different specific proteins, may have a different specificity. The
present state of our knowledge is on what may be called a descriptive

224

OXIDATION AND REDUCTION

level: some of the enzymes can be prepared as pure crystalline entities,
and their composition and activity can be examined. Since the first
stages in their discovery by Warburg, a vast amount of knowledge has
been accumulated which, on the one hand, demonstrates the highly
complex nature of the problem and, on the other, shows that certain
recognizable features are shared by these enzymes: they are proteins
attached to prosthetic groups. The proteins are all specific and not
identical with those otherwise occurring in the organism. But the
chemical mechanism of the action of these enzymes is not yet worked
out, at least to an extent comparable to that given in the simple example
of iron catalysis in the oxidation of cysteine.

It is quite natural that enzyme chemists have, thus far, been
occupied with the discovery of many kinds of enzymes, the ingenious
methods of preparing them, and the measurement of their activity.
But at this point we must inquire into the chemical mechanism by
which they work; and here only a few speculations can be brought
up. One simple suggestion is this: if it is true that the sluggishness
of an oxidation process is caused by the instability (in its thermo-
dynamic sense) of the free radical through which the over-all oxidation
has to pass, then the function of the enzyme may be that of increasing
the stability of the radical, in other words, that of increasing the con-
centration of the radical which can exist in equilibrium with the re-
duced and the oxidized state of the substrate. The substrate com-
bines, reversibly, with the enzyme, and the "semiquinone formation
constant" of the enzyme-substrate compound may be greater than that
of the uncombined compound. We may make another suggestion.
Let us suppose that the enzyme can combine not only with the sub-
strate to be oxidized but also with the oxidizing agent. For example,
methylene blue can oxidize succinic acid to fumaric acid in the presence
of the enzyme called succinodehydrogenase. Suppose this enzyme can
combine with both succinic acid and methylene blue. The specific
structure of the enzyme brings about a definite spatial orientation and
juxtaposition of fumaric acid and methylene blue. When a mole-
cule of one of these two substances collides with a molecule of the other
in a solution, the chance of an electron transfer during the short time
of collision is nil; but when these two molecules are held close together
in appropriate juxtaposition and orientation with respect to each other,
they remain in this spatial arrangement for a long time, during which

225

LEONOR MICHAELIS

an electron transfer may occur once in a while. Now, the transfer
of a single electron establishes the free radical, and from here on the
second step of oxidation takes place readily and spontaneously. In
order to account for the fact that the interaction of succinic acid and
fumaric acid in the presence of the enzyme and methylene blue is
reversible, one must postulate also that the enzyme can combine re-
versibly not only with succinic acid, but also with fumaric acid; and
not only with methylene blue, but also with leucomethylene blue.
Such an assumption is not unreasonable and is supported by the ob-
servation that, in very many cases, molecular species of a structure
similar to that of the specific active substance, inhibit the function of
that substance and so compete with it for the enzyme. Out of the
numerous examples discovered in recent years, one may recall the
antagonism of j&-aminobenzoic acid and sulfanilamide.

At this stage of the argument, we have reached the realm of
speculation. Any further advance depends on the clarification of the
structure of the enzymes and especially of the steric structure of the spe-
cific proteins. It will be the aim of future work to show that the
specific structure of the enzyme forces the substrates attached to the en-
zyme to stay in such a mutual orientation as to permit an electron
transfer, which will not occur with a reasonable probability on a free
collision. A similar principle may underlie all specific enzymic
reactions, as well as those not concerned with oxidation-reduction.
The astounding fact that all enzymes are or contain a protein of specific
structure suggests that the attachment of the substrate to the enzyme
with its specific protein structure increases the chance of a thermo-
dynamically possible reaction by forcing a spatial orientation of the
interacting molecules which has practically no opportunity to occur
on spontaneous haphazard collision, and moreover, by holding the
interacting molecules in this specific orientation for a length of time
very much longer than on the occasion of a haphazard collision by
thermal motion. This is the way in which we may imagine the
activation energy is overcome, and in which out of a vast number of
thermodynamically permissible reactions only a few reactions service-
able for the metabolism are selected.

The road for the exploration of the mechanisms of individual
metabolic catalyses will be long. Although it is still far ahead, one is
encouraged to believe that the correct road sign has been found.

226

OXIDATION AND REDUCTION

Selected References

Clark, W. M., et al., "Studies on oxidation-reduction," U. S. Pub. Health

Service, Hyg. Lab. Bull. No. 151 (1928).
Michaelis, L., "Occurrence and significance of semiquinone radicals," Ann.

N. r. Acad. Sci., 40, 39 (1940).
Michaelis, L., and Schubert, M. P., "The theory of two-step oxidations involv-
ing free radicals," Chem. Revs., 22, 437 (1938).
Pauling, L., and Coryell, D. C, Proc. Natl. Acad. Sci. U. S., 22, 210 (1936).
Schubert, M. P., J. Am. Chem. Soc, 53, 3851 (1931); 54, 4077 (1932);

55,4563 (1933).
Thunberg, T., "Zur Kenntniss des intermediaren StofTwechsels und der dabci

wirksamen Enzyme," Skand. Arch. Physiol., 40, 1 (1920).
Warburg, O., Uber die katalytischen Wirkungen der lebendigen Substanz- Springer,

Berlin, 1928.
Wieland, H., "Uber den Mechanismus der Oxydationsvorgange," Ergeb.

Physiol., 20, All (1922).

227

15

MESOMERIG CONCEPTS
IN THE BIOLOGICAL

SCIENCES

HERMAN M. KALCKAR, member of the research staff, division

OF NUTRITION AND PHYSIOLOGY, THE PUBLIC HEALTH RESEARCH INSTITUTE

OF THE CITY OF NEW YORK, INC.

l\TO CHEMIST would question that the concept of meso-
-^ » merism (or resonance), a concept which is actually based
on quantum mechanics, has played an overwhelmingly important role
in the development of modern physical and organic chemistry. This
concept has made it possible to understand and explain the properties
of numerous inorganic and organic compounds, and in a number of
cases to predict chemical events. There are many indications that this
concept will play an equally significant role in the biological sciences,
and for that reason it merits the consideration of biologists.

Mesomerism

The terms resonance and mesomerism are synonomous. The
former is based on quantum mechanical concepts, while the latter,
more neutral term merely indicates that a substance exists as a hybrid
of at least two more or less symmetric electronic states. The term
"mesomerism" is more appropriate in a biological essay in which the
symbols and formulas are not those of quantum mechanics, and will
therefore be used here.

229

H. M. KALCKAR

A group is said to display mesomerism if it can be described by
two or more symmetric (or very nearly symmetric) electronic formulas,
representing approximately the same potential energy. In that case,
the chances that the electrons will occupy one or the other position are
equally great. The electrons are therefore moving forth and back
between these equivalent positions or, to use a term of physics, oscillat-
ing between the positions. The mesomerisms of the carboxylate anion
and the amidine cation represent some of the simplest examples of sym-
metric mesomerism in a molecular group.

O:

• •

• •

R:C

or

:0
R:C

• •

:0:-

• •

O

(«)

I

(i)

Carboxylate

anion

H:N:H

• •

:N:

• •

R:C or

• •

R:C

H:N:H

H:N:H

+

(«)

(b)

II

Amidine cation

In these two sets of alternative structures, one of the oxygen or
nitrogen atoms is surrounded by a complete set of eight electrons, the
so-called octet, and the other by only six electrons, the last pair of elec-
trons participating in the double bond. If this pair of electrons were
moved up to complete the octet, the opposite oxygen or nitrogen would
have to donate one pair of electrons from their octet in order to restore
the double bond. Structures a and b are completely equivalent and
indistinguishable .

The well-known benzene mesomerism is usually illustrated by
the two structures shown in formula III,

/\

or

f\

\/

V

(a) (b)

III

Electronic mesomerism can also exist between two molecules.
This so-called intermolecular mesomerism can be illustated by the
two symmetric structures of formula IV. In structure a, the left A

230

MESOMERIC CONCEPTS IN BIOLOGY
R:A- :A:R or R:A: -ArR

IV

possesses an unpaired electron (odd electron) and the right A, an elec-
tron pair; in structure b, the situation is reversed. The shift can be
represented as an oscillation of one electron between the two symmetrical
molecules: R : A. : A : R. The existence of an unpaired electron gives
rise to paramagnetism because the neutralized magnetic moments of
paired electrons is abolished when the electron lacks its partner.
The type of intermolecular mesomerism illustrated in formula IV will
be discussed later in this essay.

The frequency of the oscillations which the electrons in meso-
meric structures undergo is very high; and it is therefore impossible
to consider a mesomeric group as possessing for an appreciable interval
of time either structure a or b. There is, however, a type of meso-
merism in which it is possible to distinguish between the two structures.
Tautomerism is a classical illustration of this type of mesomerism.

In tautomerism, both an electron and a hydrogen atom parti-
cipate in the oscillation. Since the hydrogen possesses a significant
mass, the oscillation is considerably slower than in the purely electronic
mesomerism; and it is therefore possible to distinguish between the
two symmetrical states. Two examples of tautomerism are: (1) the
enol-keto tautomerism of carbonyl compounds (V); and (2) the lac-
tam-lactim shift of the hydroxy purines (VI).

-H

H3— c— c—

N-=C— OH

H— N— C— 0

(0) II

1 1

1 1

Keto form 0

HO-

-C C— N.

II II >c-

N— C— N/

0— C C— N.

1 II >c-

H— N— C— N/

or

-H

(*) H2— C=C—

1

1

Enol form |

H

H

OH

(«)

(b)

V

VI

The so-called hydrogen bond results from the attraction of a
hydrogen atom attached to one electronegative atom (e. g., fluorine,
oxygen, nitrogen) for an unshared electron pair of another electronega-
tive atom (see formula VII):

R— O— H . . . O— R
VII

231

H. M. KALCKAR

Biological Significance of Mesomerism

What is the significance and what are the implications of meso-
meric phenomena in biological reactions? The great significance of
mesomerisin lies in the fact that it invariably endows the molecular
group with a considerable amount of additional stability. The word
stability in this connection is used in its broadest sense. It has been a
custom to distinguish between thermodynamic stability and the kind
of stability implied in terms such as "willingness" of a group to react

Scheme i
Sequence of Reaction

"ACTIVATED" MOLECULE
I

ACTIVATION ENERGY

OF REACTION:

A:=:±:(B):;^C

Scheme ii
Sequence of Oxidation

■OF THE REACTION:
A=;=i:C

A^

r-

Ia^aI

a

MOLECULAR
ACTIVATION ENERGY

1) OF
— NONCATALYTIC
OXIDATION

2) OF CATALYTIC
OXIDATION

spontaneously. The distinction must be considered artificial. The
work of Polanyi, Eyring and Stearn — cf. Eyring (3) — indicates strongly
that the activation of the substrate, i. e., the problem of how reactive
complexes are formed from stable molecular groups, is essentially a
thermodynamic one. Scheme I illustrates some of these relationships.
"A" signifies the starting product and "C," the end product
of the reaction A — > C. AFis the change in free energy of the reaction,
I. e., the amount of potential energy which is lost when A is converted to
C. The free energy or potential energy of the system drops when A is
converted to C. However, in order to start the reaction, A must be
activated in some way, i. e., the potential barrier which represents the
so-called activation energy must be overcome. Both the activation
energy and the free energy change (AF) are influenced by mesomerism.
Since the significance of mesomerism for our understanding of energy
coupling in biological systems has been discussed elsewhere (6), this

232

MESOMERIC CONCEPTS IN BIOLOGY

aspect will be treated rather briefly in the present essay. The influence
of mesomerism on AFcan be summarized as follows: If C has an addi-
tional amount of mesomeric stability which A does not possess, the drop
in AF of the reaction A — > C is greater than it would have been if C
did not possess that extra stability. We have already mentioned the
carboxylate ion as a typical representative of a molecular group with
extra mesomeric stability. If another molecule is introduced into such
a mesomeric group, as by esterification, the symmetry of the group is
disturbed and the mesomerism decreases or vanishes, which again
implies that the potential energy of the complex is raised. Thus, acetic
acid anhydride, CH3 — CO— O— CO — CH3, in which both carboxyl
groups have lost their state of mesomerism, possesses a much higher
potential energy than that of the two acetic acids formed by hydrolysis.
In terms of scheme I, acetic acid anhydride would correspond to A
and the two acetic acid molecules to C.

The living cell contains at least three types of substances in
which the mesomerism of two groups is mutually blocked. The first
type includes the carboxyl phosphates (acyl phosphates), the second
group, the amidine phosphates (phosphocreatine, phosphoarginine),
and the third group, the pyrophosphates. The carboxyl phosphates are
the primary oxidation products of a reaction in which a carbonyl
phosphate complex undergoes enzymic oxidation. It is important to
point out that the oxidation is catalyzed by an enzyme specific only
for the phosphate complex. Thermodynamically speaking, the
oxidation of a carbonyl-water complex to free carboxylate would be
greatly favored, and the chances of forming a carboxyl phosphate would
be vanishingly small, if the latter reaction were not specifically catalyzed
by an enzyme.

This brings up the question of the nature of enzyme catalysis.

An unusually promising approach toward an understanding of
oxidation-reduction catalysis on the basis of mesomeric concepts has
been made by Michaelis and his group. Since this topic is discussed in
Chapter 14, only certain aspects of the problem will be treated

here (8).

It is now generally recognized that oxidation of organic com-
pounds, involving the removal of a pair of electrons, takes place step-
wise. The removal of one electron prior to the other gives rise to the
formation of a free radical displaying paramagnetism because of the

H. M. KALCKAR

presence of an unpaired electron, possessing unneutralized magnetic
moment. This free radical has, generally, a very brief existence, since it
either accepts an electron again or expels the remaining odd electron.
Since the free radical has so little chance of existence, the removal of
the first electron is barred, so to speak, by a high potential barrier.
In scheme I, the group to be oxidized would be represented by A, the
free radical by B, and the final oxidation products by C. The height
of the potential barrier would represent the activation energy. The
activation energy is the factor which particularly interests us in con-
nection with the concept of catalysis. If the potential barrier is too
high, the chance of forming the free radical is practically nil and the
rate of the reaction is zero. When the temperature is raised sufficiently,
the thermal movements of the molecule become so vigorous that a
certain percentage of molecules will slip over the potential barrier.
Although thermal movements, of course, are of importance for events
in the living cell, physiological temperatures are usually too low to
allow most reactions to proceed at measurable rates. In order to bring
about a reasonable rate, the living cell has succeeded in lowering the
potential barrier by a special device which we call catalysis. It is in
this connection that mesomerism may turn out to be of paramount
importance, as the model experiments of Michaelis and his associates
have so strikingly demonstrated.

The reduction of /)-benzoquinone to the corresponding hydro-
quinone goes through a radical (semiquinone). The free radical has
very little chance of existence because of its asymmetry (formula VIII).
In basic solution, however, the semiquinone will exist as the symmetrical
ion oscillating between two symmetrical electronic states. The two
equivalent structures interchange the odd electron, similar to what
Pauling (9) calls a three-electron bond, either through the benzene
ring or by intermolecular bonds (formula IX).

.6:H :6: < — ^ .6:

I I I

A fA A

Y V Y

:0: :0. < — > :0:

VIII IX

MESOMERIC CONCEPTS IN BIOLOGY

The condition for stabilization of the semiquinone by meso-
merism is satisfied only when the two structures are equivalent. This
requirement is satisfied when the molecule is dissociated as an anion.
The undissociated semiquinone in which the presence of the hydrogen
atom eliminates the symmetry of the two structures does not fulfill the
required condition. Correspondingly, Michaelis and his group showed
that the semiquinone of phenanthrene-3-sulfonate is relatively very
stable in alkaline solution in which it exists as the symmetrical anion.
The semiquinone of /^-phenylenediamine, on the other hand, has a fair
chance of existence in strongly acid solutions because only then does
the symmetrical phenylenediaminium cation exist. In complete
accordance with the ideas just developed is the observation that the
free radical of paraquinones accumulates only in the alkaline pH range,
whereas those of the paradiamine compounds accumulate in measur-
able amounts only at strongly acid reactions. The "catalyzing" effect
of hydroxyl ions or hydrogen ions on these two types of oxidation-
reductions has actually been explained in terms of mesomerisms.

The intriguing question is whether it is possible to explain en-
zymic catalysis in terms of the same principles. There are observa-
tions which may provide confirmation for such an explanation. Some
years ago, Haas (4) found that riboflavin phosphate, when linked to a
specific protein, forms a semiquinone when undergoing reduction.
This semiquinone is not observed during the reduction of free riboflavin
phosphate in neutral solution, but accumulates when the reaction is
acid enough to insure complete ionization. In other words, the
enzyme is able to stabilize a product at neutral reaction which other-
wise would exist only at strongly acid reaction. This observation may
be taken as a clear indication that oxidation-reduction enzymes in
some way or other are concerned with the formation of mesomeric
free radicals.

Before going deeper into the discussion of the nature of enzyme
catalysis, it is worth while to introduce a very ingenious theory proposed
by Delbriick, dealing with the nature of reproduction phenomena.
Delbriick advanced the idea that, in processes like gene reproduction,
mesomeric phenomena may play an outstanding role. According to
classical concepts, the probability of bringing the same components
together in the same sequence as that of the original molecule is in-
finitely small, Delbriick is probably the first who has expressed the

H. M. KALCKAR

idea that the reproduction of a chain molecule may be accomplished
by the operation of mesomeric forces. According to this hypothesis,
the chances for reproduction are reasonable in case the new molecule
exists in the transitory state of a free radical (or as numerous free
radicals). Each unit of the daughter chain molecule existing as a free
radical will attempt to be picked up by the identical valence saturated
units of the original molecule with which they will form a mesomeric
three-electron bond of the type illustrated in formula IX. Delbriick
imagines the formation of such a chain of free radicals taking place as
the result of a one-step oxidation of an amino aldehyde "polypeptide"
chain, as in formula X (1). The peptide represents the fully oxidized
group, and the free radical, the partly reduced group.

-i- -i-

I I

N— H H— N

I .. .. I •

C::0 0:0:

I .. .. I

Peptide Free radical

X

A different quantum mechanical model of autocatalysis of pro-
teins has been discussed in a recent review by Jordan (5). It is hardly
necessary to stress that such hypotheses, which are based on quantum
mechanical concepts, must be taken only as the very first attempt to
explain the type of forces which operate in phenomena like reproduc-
tion. As Delbriick (2) expresses it: "The point I want to make is this:
Quantum mechanics offers a reason why a two-step oxidation may be
catalyzed by a structure which is closely similar to the oxidation
product."

The considerations just outlined ought to encourage us to look at
the problem of enzyme catalysis in terms of mesomerism. The idea
that a group of specific proteins called enzymes are capable of overcom-
ing potential barriers of 20,000 to 50,000 calories has always been
difficult to comprehend. It was mentioned above that the living cell,
in order to overcome the potential barrier which prevents the degrada-
tion of metabolites, had to evolve the device of catalysis. Suppose the
reaction A — ^ C illustrated in scheme I (page 232) were an oxidation
(scheme II) involving two electrons and that the intermediate B repre-

236

MESOMERIC CONCEPTS IN BIOLOGY

sented the first oxidation step, i. e., the product which lost only one
electron. In that case, the chance of existence of B, possessing as it
does an odd electron, would be infinitely small and the potential bar-
rier, therefore, very high. However, if B were able to form an inter-
molecular mesomeric bond with A of the type illustrated in formulas
IVfl and b, and in IX (that is, a three-electron bond), the chances of
getting the reaction started would be much better.

This could be illustrated in the following way: Let A be the

r

product to be oxidized, A the first oxidation product (^ signifies the un-
neutralized electromagnetic moment of an unpaired electron), and A°^
the final product. The broken line signifies the potential barrier
(or activation energy) of the noncatalyzed reaction, and the unbroken
line, that of the catalyzed reaction in which A forms a mesomeric com-

plex (signified by A f^^ A or by A: ?=^ .A) with A, both being in close
proximity to each other on the specific catalyst. In other words, the
state of activation is not confined to one molecule but is "spread"
over two or more molecules. Yet only one of the molecules, perhaps
that closest to the electron acceptor, donates the second electron and is
thus converted to A°^.

It might be useful, in order to make the idea more intelligible,
to select as a concrete example the enzymic oxidation of an alcohol to a
carbonyl group. The specific catalyst of this reaction could be classi-
fied as an alcohol dehydrogenase or as an acetaldehyde reductase
(hydrogenase) with equal right, inasmuch as the enzyme is equally
specific toward the alcohol and the aldehyde. The first step in the
oxidation is the formation of the free radical (see also Chapter 14)
which, however, has very little chance of formation if it cannot in some
way or other undergo mesomerism with another molecule. Inasmuch
as the protein-substrate combination brings the various oxidation
products of the substrate in close proximity, one is justified in assuming
that a good opportunity for the formation of mesomeric complexes is
at hand. The sequence of phenomena in such a dehydrogenation might
be described as follows. It is well known that, in oxidations of metab-
olites such as hydrocarbons, alcohols, and aldehydes, not only elec-
trons but hydrogen too is removed. Most metabolites like hydrocarbons
and alcohol are infinitely weak acids and have very little tendency to
form hydrogen ions except at extremely alkaline reactions.

H. M. KALCKAR

How, then, is the removal of hydrogen ions (which must pre-
cede the electron removal) accomplished in biological systems at neu-
tral reactions? In trying to answer this question one may refer to a
fact which has been particularly emphasized by Pauling (9), namely,
that the high dissociation constant of carboxyl hydroxyl is mainly attrib-
uted to the higher mesomeric stability of the symmetric carboxylate

ion, — C , as compared with the asymmetric carboxyl group,

\o-

— C . Suppose that an alcohol group bound to a specific enzyme

\0H
were able to undergo intermolecular mesomerism with two or more
other alcohol groups, provided that it existed in the ionized form. This
would imply that such an alcohol group showed greater tendency to
release hydrogen ions when forming a substrate-enzyme complex with
the specific dehydrogenase than when in free solution. If one imagines
two alcohol molecules brought together in close vicinity by the specific
enzyme protein, one sees immediately that the dissociation of a hydro-
gen ion from one of the alcohol groups would give rise to the formation
of a hydrogen bond (formula XI). Since a hydrogen bond is a struc-
tural feature with a certain degree of stability, it means that the
addition product would have a fair chance of existence. Now the way is
paved for the first oxidation step, i. e., the removal of one electron from
one of the alcohol groups. If the other alcohol group simultaneously
gives up its hydrogen ion, an ideal opportunity for forming a three-
electron bond has been created (formula XII). The free radical,
although possessing a certain degree of stability due to the existence of
the mesomeric three-electron bond, is nevertheless a labile configura-
tion, which tends to expel a second hydrogen ion and electron.
H ^ H H H

K:'6:'6::^=z ^ :0:'C:R ^r.n- ^ on-R

H H H H

: H+ : ; H+ : +e

XI XII

Waelsch (11) has suggested the possibility that the alcohol group
of the substrate may form a mesomeric complex with certain hydroxy

238

MESOMERIC CONCEPTS IN BIOLOGY

groups (serine, threonine) of the protein catalyst. This suggestion
may be in better accordance with kinetic data. Moreover, the sug-
gestion may illustrate the nature of the forces responsible not only for
the catalytic action of enzymes but also for their specificity.

The electron acceptor goes through most of the steps illustrated
here, only in the opposite order. The mesomerism between the alde-
hyde and the free radical is best illustrated as a complex between the
free radical and the aldehyde in the biradical (triplet) state in which
two electrons in the double bond are separated (formula XIII). In

H H

R:C:0: , .6:C:R

Free radical Biradical

aldehyde aldehyde

XIII

ihe biradical state, the group contains two unpaired electrons which
will give rise to the occurrence of paramagnetism. Most double bonds
are considered to exist only to a very small degree in the biradical state;
yet the occurrence of paramagnetism in molecules containing double
bonds is a reminder of the existence of unpaired electrons. In most
cases, the biradical state is considered a state of excitation. The energy
of this excitation has recently been made accessible to measurement,
thanks to the studies of phosphorescence spectra by Lewis and his
group. Readers interested in this new and interesting development are
referred to the papers of Lewis and his co-workers (7).

There are, however, a few facts concerning biradical double
bonds which might be of direct interest for biologists. One of them is
that molecules like oxygen and nitrate exist mainly in the biradical
state and that the double bond configuration represents the labile
state. Correspondingly, these molecules display a high degree of
paramagnetism. Pauling (9) describes the stable biradical (or triplet)
state of the oxygen molecule as a single bond with two three-electron
bonds (formula XIV). It would be interesting to know whether this

:0 : O;

• • -4 > •

XIV

is the form of oxygen which reacts with the coenzymes of the respira-
tion enzymes.

H. M. KALCKAR

Questioning the value of the speculations and considerations
put forward in this short essay might well be justified. The mesomeric
description of phenomena like reproduction and catalysis is, admittedly,
sheer speculation. Yet, it is a kind of speculation which may prove
to be highly constructive, inasmuch as phenomena are introduced
which are within the realm of experiments. The studies of Theorell,
Pauling, Coryell and Michaelis and his group, on the paramagnetic
susceptibility of biologically important substances point out the direc-
tion of future research. An adaptation of these methods on a micro
scale (10) might very likely pave the way for an experimental attack
on problems like catalysis and auto catalysis.

References

(1) Delbrlick, M., Cold Spring Harbor Symposia Quant. Biol., 9, 122 (1941).

(2) Delbriick, M., "Problems of Modern Biology in Relation to Atom
Physics." Lectures delivered at Vanderbilt University, 1944.

(3) Eyring, H., Chem. Revs., 17, 65 (1935).

(4) Haas, E., Biochem. Z, 290, 291 (1937).

(5) Jordan, P., Naturwissenschajten, 32, 20 (1944).

(6) Kalckar, H. M., Ann. N. T. Acad. Sci., 45, 395 (1944).

(7) Lewis, G. N., and Kasha, M., J. Am. Chem. Sac, 66, 2100 (1944).

(8) Michaelis, L., and Schubert, M. P., Chem. Revs., 22, 437 (1938).

(9) Pauling, L., The Nature oj the Chemical Bond. Cornell Univ. Press, Ithaca,
1939.

(10) Theorell, H., Arkiv Kemi Mineral. Geol., A16, Art. I, 1 (1942).

(11) Waelsch, H., private communication.

240

16

VISCOMETRY

IN BIOCHEMICAL

INVESTIGATIONS

MAX A. LAUFFER, associate research professor of physics,

UNIVERSITY OF PITTSBURGH; THE LILLY AWARD IN BIOCHEMISTRY

r.

ISCOSITY has been recognized for many years as an im-
portant characteristic of colloidal and biological materials.
In spite of this fact, however, in terms of the properties of molecules
and macromolecules, the meaning of viscosity remained almost com-
pletely obscure until recently. The cause of the sustained popu-
larity of this technique during the past years should probably be at-
tributed to the extreme ease of viscosity measurements, as contrasted
with other physical approaches to the study of biological and bio-
chemical systems, rather than to the theoretical significance of the
results obtained. The outlook for the future, however, is brighter,
for reasonably satisfactory theories have recently become available
which describe quantitatively the viscosity of liquids and of macro-
molecular solutions in terms of molecular size, shape, and thermo-
dynamic properties (22,24). Much progress has also been made in
late years from a purely empirical point of view in the correlation of
viscosity with the size and structure of molecules, particularly in the
field of high-polymer chemistiy (11).

In its most general sense, viscosity is a measure of the amount
of work which must be expended in order to maintain a certain rate
of flow. In the case of some liquids, the rate of flow at a fixed tem-

241

M. A. LAUFFER

perature is directly proportional to the force applied per unit area on
a plane parallel to the direction of flow. A fluid which flows in this
manner obeys Newton's law of flow and is said to be Newtonian. The
constant which relates the velocity gradient or the rate of shear in the
flowing liquid to the applied shearing force is called the viscosity
coefficient or, in less exact usage, simply the viscosity. In many liquids,
however, the rate of shear is not directly proportional to the shearing
force. Therefore, the ratio of the applied force to the resultant rate
of displacement is not a constant but a variable whose magnitude
depends upon the applied shearing stress or, viewed from a slightly
different but more usual position, upon the resultant rate of shear.
Thus, there is no such thing as a viscosity coefficient for such a fluid.
However, the ratio of the shearing stress to the resultant rate of shear
at a particular velocity gradient can be used as a partial description
of the flow characteristics of such fluids. This ratio can be called an
apparent viscosity coefficient. A complete description of the flow of a
liquid of this sort requires that the way in which the apparent viscosity
coeflficient varies with the rate of shear be specified. Fluids which
flow in this manner are often said to be non-Newtonian.

Until recently, it was fairly generally believed that Newtonian
and non-Newtonian fluids were qualitatively different, but newer
developments in the theory of viscous flow permit the interpretation
that the difference is only quantitative. Eyring and his associates (24)
have attempted to understand the flow characteristics of liquids on the
basis of concepts analogous to those employed in chemical kinetics.
The basic aspects of the Eyring point of view are that, in order for a
liquid to flow: (1) there must be regions or holes in the body of a liquid
into which molecules can jump; (2) there are potential energy barriers
which tend to prevent any particular molecule from jumping into a
vacant region near it; and (3) a shearing stress is a mechanical potential
which aids molecules jumping in the direction of the stress and hinders
those jumping in the reverse direction, thereby resulting in a net dis-
placement in the direction of the stress. Eyring has shown that,
based upon these concepts, the resultant rate of shear should be pro-
portional to the hyperbolic sine of the ratio of the work contributed
by the shearing force in moving a molecule over the energy barrier
to the product of the gas constant per molecule and the absolute
temperature. The proportionality constant is a function of the po-

242

VISCOMETRY

tential energy barrier. The hyperbolic sine of a variable has the
interesting property of being practically equal to the variable for values
less than about V2 and practically equal to the exponential of the
variable for all values greater than about 2. Therefore, in a liquid
in which the potential energy barrier is low, an observable rate of flow
can be obtained by the application of a force small enough to be in
the linear portion of the hyperbolic sine function. The rate of shear
will thus be pi'oportional to the shearing force. In other words, the
liquid will exhibit Newtonian flow. If, on the other hand, the force
required to cause an observable rate of shear is somewhat greater than
that just mentioned, the rate of shear will not be directly proportional
to the shearing force, but may be even an exponential function of the
force. Such a liquid obviously would not exhibit Newtonian flow,
but the diff'erence between it and a liquid which would exhibit New-
tonian flow would be a matter of degree and not of quality.

The application of viscosity to the problems of biology and bio-
chemistry usually involves solutions or dispersions in aqueous solvents.
In general, when a solute is dissolved in a solvent, the viscosity of the
solution, 77, is either greater than or less than that, 770, of the solvent.
The ratio of the viscosity of a solution to that of the solvent under the
same conditions, 77/770, is called the relative viscosity of that solution.
When rigid particles of colloidal or macromolecular dimensions are
dispersed or dissolved in an aqueous solvent, the viscosity of the re-
sultant solution is greater than that of the solvent; therefore the
relative viscosity is always greater than unity for such systems. The
relative viscosity of a solution is usually dependent upon the concen-
tration of the solute. Experience has shown that an equation of the
sort, 77/770 = 1 + AC + 5C" + . . . , usually can be written to describe
the relationship between relative viscosity and concentration, C. A
and B are arbitrary constants. For low values of concentration,
77/770 = 1 + AC or 77/770 — 1 = AC. The expression, 77/770 — 1, is
defined as the specific viscosity of the solution. In dilute solutions,
this specific viscosity is directly proportional to the concentration.
The proportionality constant, A, in the equation written above, was
defined by Kraemer (14) as the intrinsic viscosity and is often repre-
sented by the symbol, [77]. In strict mathematical language, the
intrinsic viscosity is the limit, as the concentration approaches zero,
of the ratio of specific viscosity to concentration. It can be repre-

243

M. A. LAUFFER

sented symbolically as follows: A = [r]]= [(v/vo - l)/C]c_o- Obviously,
the numerical value of the intrinsic viscosity is dependent upon the
manner in which the concentration is expressed. The latter must be
specified for the former to have meaning. The intrinsic viscosity is
an important constant describing a solute. As will be shown in the
paragraphs to follow, it is related to the molecular structure of the
solute particles and, in some cases, to the thermodynamic properties
of the solution. The principal concern of viscometry is the evaluation
of intrinsic viscosities.

All of the constants which have been defined thus far for solu-
tions apply only to those which exhibit Newtonian flow, that is, to
solutions for which a true viscosity coefficient can be assigned. Many,
if not most, of the solutions of interest to biologists and biochemists
do not exhibit Newtonian flow. For such systems, one obtains only
apparent viscosity coefficients, which depend upon the rate of shear.
Such systems must be described by a series of constants analogous to
the relative, specific, and intrinsic viscosities of Newtonian solutions.
These can be given the names: apparent relative viscosity, apparent
specific viscosity, and apparent intrinsic viscosity. The latter differ
from the former only in that each is dependent upon rate of shear.
Thus, in a study of a non-Newtonian solution, rate of shear must be
held essentially constant, or it must be specified, or the nature of the
variation with rate of shear must be described. Apparent intrinsic
viscosities are not necessarily any less useful than intrinsic viscosities.

It is the purpose of this discussion to consider the usefulness of
\iscosity in the study of biological and biochemical systems. In the
remaining paragraphs, a brief consideration of the methods of measur-
ing viscosity and a discussion of the meaning of the viscosity of solutions
or dispersions of rigid colloidal particles will be presented. Particular
attention will be devoted to the examination of direct experimental
evidence concerning the correctness of theories used to interpret the
various aspects of viscous behavior.

Intrinsic viscosities and apparent intrinsic viscosities are cal-
culated from measurements of the viscosity of the solvent, the viscosity
or the apparent viscosity of the solution, and the concentration of the
solute. It is entirely outside the scope of this article to discuss the
measurement of solute concentration. Suflftce it to say that any of
the methods of quantitative chemical and biochemical analysis might

244

VISCOMETRY

be used, depending upon the nature of the sokitc under study. There
are numerous methods of measuring viscosity or apparent viscosity.
Perhaps the simplest from a practical point of view is that involving
the rate of sedimentation of a sphere in the viscous liquid. According
to Stoke's law, the retarding force resisting the translational movement
of a large sphere in a viscous liquid is equal to dirr-qv, where r is the
radius and v the velocity of the sphere. When the sphere is falling
under the influence of gravity, the accelerating force, /a, acting upon
it is the mass of the sphere, Ws, minus the mass of the fluid it displaces,
Wl, all multiplied by the gravitational constant, G; fa = G{Ws — Wj).
Since the mass of the sphere is its volume, Va 7^r^ times its density,
fl'„ and the mass of the displaced liquid is equal to the volume of the
sphere times the density of the liquid, (11, this equation can be written
as /a = G Vs T^r^ids — dj). The sphere settles at a uniform rate when
the accelerating force and the retarding force are equal:

eirr-nv = Vs ^r^Gids - d^), or t? = [Vg r-G(^. - d:}]/v

Since it is possible to determine the radius and the density of a large
sphere, and the gravitational constant is known, the viscosity of a
liquid can be obtained by measuring its density and the velocity of
settling of the sphere. It is particularly easy to measure relative
viscosity* by this method, for:

r]/r]o = [{ds - dL)/(ds - dLo)](vQ/v).

The subscript, 0, indicates a quantity describing the solvent. To
obtain the relative viscosity of a solution, it is necessary to know only
the density of the sphere and to measure the densities of and the
velocities of fall of the sphere in both solvent and solution. The
velocity of the sphere can be obtained by observing the time required
for it to fall a fixed distance. The falling ball method has been used
quite satisfactorily, giving results which are in good agreement with
those obtained by other methods. This method is not particularly
satisfactory for the study of non-Newtonian solutions because the
velocity gradient in the liquid surrounding the falling sphere varies

* Practically all viscosity measurements are relative — even those of pure
liquids. Much effort has been expended to determine precisely the absolute vis-
cosity of water at 20° C. All other viscosity data are relative to the value assigned
to water and are subject to the same error as the absolute viscosity of water.

245

M. A. LAUFFER

with the position with respect to the sphere. Thus, the velocity
gradient cannot be specified. In addition, there is a tendency for the
flow around the sphere to be somewhat turbulent — an effect which
cannot be eliminated entirely.

The most commonly used method of determining viscosity is
that involving the flow of the liquid through a capillary tube.
Poiseuille (20) demonstrated empirically and Hagenbach (11) derived
theoretically from Newton's postulate that, when the rate of flow is
slow enough to exclude turbulence, the volume of liquid, V, flowing
through a tube in time, t, is directly proportional to the pressure
difference, P, across the capillary and to the fourth power of the capil-
lary radius, r, and inversely proportional to the length, /, of the capillary
and to the viscosity coefficient, rj:

V TrPr* irrHP

or 77 =

t 8r?/ SIV

If the dimensions of the capillary are known, absolute viscosity can be
measured; but, even if they are not known, relative viscosity can be
measured, for (VW = (i/h)(V(,/V){P/Po). The most usual procedure
is to observe the time for a definite volume of liquid to flow at a con-
stant pressure diff"erence, in which case v/vo = t/k, or to observe the
time for a definite volume to flow under the influence of gravity with
a constant average difference in liquid level. In this case, 77/^0 =
(t/to){d/do), where d and do are the densities of the solution and solvent,
respectively. These conditions are the ones encountered when using
the common Ostwald viscometer. Relative viscosities can be measured
with high precision by this technique. The capillary method has the
very considerable advantage of requiring easily constructed apparatus.
It has one important limitation, however. The rate of flow of a New-
tonian liquid at various distances from the wall of the capillary increases
parabolically from the wall to the center. The velocity gradient is
not constant throughout the cross section of the tube, but decreases
linearly from the wall to the center. With non-Newtonian liquids,
it is difficult to predict the exact nature of the variation of velocity
gradient with distance from the wall, but certainly the gradient varies
considerably. For that reason, this type of apparatus is not ideally
suited to the study of non-Newtonian systems.

The method most satisfactory from the theoretical point of

246

VISCOMETRY

view is that of Couette (6). Here, the liquid is placed between two
concentric cylinders, one of which is rotated while the other remains
stationary. The viscosity is usually determined by measuring the
moment due to viscous resistance, M, acting upon the inner cylinder,
when the outer one is rotated at a constant velocity. If R\ is the radius
of the inner cylinder, Ri the radius and fi the angular velocity of the outer
cylinder, and L the depth of liquid of viscosity t] between the cylinders:

R\ Rl \ Rl- R\ M

If the dimensions of the apparatus and the depth of the liquid layer
are known exactly, absolute viscosity can be evaluated from the ratios
of the moment to the angular velocity. Relative viscosity can be
obtained without knowing the dimensions exactly, for tj/tjo = (M/A/o)
(Oo/^)(Z-o/L). In practice, the best method is to measure M and Afo
at the same depth of liquid and the same angular velocity. Then
^/t^o = Af/A/o- There are numerous possible methods of measuring
the moment due to viscous resistance. The most common is to suspend
the inner cylinder by a wire of known restoring force, N. In this case,
the inner cylinder will turn through an angle 0, equal to M/N, and this
angle can be measured. Also, the inner cylinder could be mounted
on a mechanism similar to that of a galvanometer, and the moment
could thus be estimated in terms of the current through the galvan-
ometer which just prevents the cylinder from turning. Another possi-
bility would be to measure the rate of energy consumption with a
wattmeter, for this rate should be proportional to the moment due to
viscous resistance.

The velocity gradient at every point within the liquid enclosed
between the two cylinders is directly proportional to the angular
velocity of the outer cylinder and inversely proportional to the square
of the radius at the point in question. In a viscometer in which the
distance between the inner and outer cylinders is small compared with
the radius of the inner cylinder, the velocity gradient is essentially
constant everywhere throughout the enclosed liquid. Even in a vis-
cometer in which the distance between cylinders is one-tenth the radius
of the inner cylinder, the velocity gradient will not vary much more
than 10% from the mean. This essential constancy of velocity gradient
makes the rotating cylinder apparatus the most useful from a theoretical

247

M. A. LAUFFER

point of view, for it makes possible a study of the variation of apparent
viscosity with velocity gradient in non-Newtonian Hquids.

Certain precautions must be observed when any one of the
methods just discussed is used. These are adequately described in
readily available publications (12). The investigator who desires to
use viscometry must become familiar with them. However, since they
are not essential to the understanding of the meaning of viscosity data,
they will not be discussed here.

Usually the purpose of making viscosity measurements in bio-
logical and biochemical studies is to evaluate the intrinsic viscosity,
[iv/vo — l)/C]c_*o- If either specific viscosity, 17/770 — 1, or relative vis-
cosity, v/lo, were a linear function of concentration, this would be a
simple matter. As pointed out previously, however, specific viscosity
is approximately a linear function of concentration only for very low
concentrations. Many equations have been proposed to show the
relationship between specific viscosity or relative viscosity and con-
centration. Each one seems to work for some specific system or
systems. They have been reviewed by Huggins (13). All can be
expanded into series of the form t^/ijq — 1 = AC + BC^ + . . . . The only
difference between them is in the relative values of the constants B
and A.

The problem of determining intrinsic viscosity reduces itself to
the evaluation of the constant A, for it obviously represents the ratio of
specific viscosity to concentration as concentration approaches zero.
The simplest method of doing this is to plot the observed specific vis-
cosity against concentration and then draw a tangent to the curve at
the origin. The slope of the tangent will be A. This practice is
subject to the limitation that it uses only the data obtained at extreme
dilutions, where the experimental error is bound to be great. For that
reason, two other methods for evaluating A are sometimes used. The
two simplest expressions relating viscosity and concentration are those
of Arrhenius (1) and Bingham (2). The Arrhenius equation is
In T7/770 = AC, and the Bingham equation is 0/0o = Vo/v = 1 — AC,
where <j) is the fluidity or the reciprocal of the viscosity. Both equations
can be expanded into series, such as that shown above. Each can be
used to evaluate /I, or the intrinsic viscosity. From the Arrhenius
equation:

248

VISCOMETRY

In 77/770

-^ = ^ - hi

and from the Bingham equation:

(1 - «/>/0o)/C = A ^ [rj]

Thus, in the apphcation of the Arrhenius equation, one plots In 77/770
against C; and in the apphcation of the Bingham equation, one plots
(1 — 4>/(l>o) against C. In both cases, A or [77] is equal to the slope of
the straight line fitting the data in the region of low to moderate
concentrations. These methods employ a wider range of data than
the simpler procedure, and thus allow a more precise evaluation of the
intrinsic viscosity.

When rigid particles of colloidal dimensions are suspended in a
solvent, they increase the viscosity of the system. The first successful
theoretical treatment of this effect was made by Einstein (7). His
approach was from the point of view of hydrodynamics. When a
liquid undergoes plane laminar or simple viscous flow, infinitesimal
layers of the liquid glide over one another in the direction of flow, and
each layer moves slightly faster than the layer on one side of it and
slightly slower than the layer on the other side. In this process, energy
is dissipated, resulting in the viscosity of tlie fluid. If a rigid solid
object, large compared with the infinitesimal layers of liquid, is placed
in such a flowing system, some of the layers of liquid in which this
particle finds itself will move faster than the particle and some will
move more slowly. This will tend to cause the particle to rotate, and
it will also make it necessary for fluid to flow around the obstruction.
The resultant disturbance in the motion of the fluid results in an added
dissipation of energy by the system, in other words, in increased vis-
cosity. Einstein derived an equation for the relative viscosity of a
suspension of spherical particles as a function of the concentration of
the spheres: 77/770 = 1 + 2.5 C, where the concentration is expressed
as the volume fraction. On theoretical grounds, this equation should
be valid only for an infinitely dilute suspension of spherical particles
which are very large compared with the size of the solvent particles
and very small compared with the dimensions of the viscometer. More
recently, Guth (10) has shown that the viscosity of more concentrated
solutions of spheres should be given by the equation:

r,/rjo = 1 + 2.5 C + 14.1 C-

M. A. LAUFFER

The Einstein equation has been subjected to a rigorous test by
Eirich, Bunzl, and Margaretha (8). These investigators used rotating
cylinder, falling ball, and capillary viscometers to measure the vis-
cosity of glass spheres with a radius of 80 microns suspended in a
solution of mercuric nitrate in nitric acid, mushroom spores with a
radius of 4 microns dispersed in a mixture of olive oil and tetrachloro-
ethane, and yeast cells with an average radius of 2.5 microns dispersed
in water. The spherical natures of all of these particles were ascer-
tained by examination with the microscope. The intrinsic viscosities

Table I
Viscosity of Yeast Suspensions

Volume fraction of yeast

v/n<i

- 1

Einstein equation

Observed

0.0088

0.022

0.022
0.021

0.0175

0.044

0.042
0.042
0.043

0.0350

0.088

0.082
0.087
0.085
0.083

0.070

0.175

0.195
0.196
0.197

0.140

0.350

0.530
0.524
0.500
0.515

obtained in every case differed from that given by the Einstein equa-
tion, 2.5, by less than 10%. This means that the Einstein equation
is reliable when it is applied to systems which conform to the conditions
for which it was derived. Representative of the data of Eirich and
collaborators are the results presented in Table I, where the specific
viscosities obtained in a capillary viscometer of a yeast suspension at
various concentrations are compared with specific viscosities calculated
from the simple form of the Einstein equation. It can be seen that the
equation fits the data in very dilute suspensions.

250

VISCOMETRY

The Einstein equation relates the relative viscosity of a sus-
pension of spheres to the volume concentration. Thus, a viscosity
measurement on a dilute suspension of spheres would constitute a
means of determining the volume concentration of the suspended
matter. There are circumstances vmder which this information can
be of great value. For example, the water content of a pure prepa-
ration of particles of influenza virus can be determined by viscometry.
Electron micrographs show that the particles of this virus are spheres.
Therefore the total volume occupied by the particles (plus enclosed
or bound water) can be determined by measuring the viscosity. The
amount of solid matter can be determined by chemical analysis. Thus
the total volume in suspension associated with unit weight of dry matter
can be calculated; and, if the density of the dry matter is known, the
percentage by weight or by volume of water in the virus particle in
suspension can be calculated.

This problem can be viewed in a slightly different manner.
The specific viscosity of the suspension of virus particles would be equal
to 2.5 times the true volume fraction of the particles. If these particles
were composed of 80% by volume of water, the volume fraction of the
solid matter would be only one-fifth of the true volume fraction. Thus
the specific viscosity would be equal to 5 X 2.5 or 12.5 times the volume
fraction of solids. Therefore, for such a suspension, the intrinsic vis-
cosity would be equal to 12.5, when concentration is expressed as
volume fraction of solid matter. It is not common to find biological
materials composed of very much more than 80% by volume of water.
Hence, one should not expect intrinsic viscosities of spherical biological
materials to have values much greater than 12.5. Yet many biological
systems have much higher intrinsic viscosities. For example, tobacco
mosaic virus has a minimum intrinsic viscosity of about 39, and values
two or more times as great are often observed. Under some circum-
stances, solutions of tobacco mosaic virus nucleic acid have an intrinsic
viscosity of more than 60, and thymus nucleohistone has an intrinsic
viscosity of about 82 (16). The interpretation of such high values
must be sought in an extension of the theory of Einstein. Today, it
is generally believed that the high intrinsic viscosities are caused by
extreme departure of the colloidal particles from the spherical shape.

This question was first considered by Jeffery. He extended the
treatment of Einstein to inchide rodlike and platelike ellipsoids of

M. A. LAUFFER

rcxohition. Cicncially speaking, during laminar flow, the long axis
of a rigid rodlikc or platelike particle will be subjected to a tendency to
rotate aboui ilic direction in the (low plane perpendicular to the direc-
tion of flow. riiric will be ;i londency for (he particle to rotate most
rapidly wlicn I he loui; ;ixis is |)erpcndicular to the flow plane and most
slowly when the long axis is in the flow plane. This will result in an
average orientation in the direction of flow, the extent of which will
depend upon the degree of anisometry. With particles which are
small enough, Brownian motion will oppose the orientation and tend
to maintain a random state. The contribution that a particle makes
to the viscosity of the solution will depend upon its orientation. When
the long axis is perpendicular to the direction of flow the disturbance
of the motion of the fluid will be greatest, and when it is parallel the
disturbance will be a minimum. Thus, a suspension of randomly
oriented ellip.soids will have a higher viscosity than a suspension of
ellipsoids oriented more or less parallel to the direction of flow.

Simha (22) has solved the problem for the case of randomly
oriented rodlike and platelike ellipsoids of revolution. The intrinsic
viscosity depends upon the ratio of the major, b, to the minor, a, axis
of the ellipse whose rotation generates the ellipsoid. For elongated
ellipsoids, when b/a is considerably greater than 1 :

(b/ay 14

15

w

(b/ay

1

K

15

A a J

3'
2

1
5

1.

2.

The factor K is the ratio of the volume occupied by a hydrated particle
to its anhydrous volume. For disklike ellipsoids, when b/a is con-
siderably greater than 1 :

W _ 16 b/a

K 15 arc tan {b/a)

For the case of elongated ellipsoids oriented parallel to the direction
of flow, Eisenschitz (9) has derived the following equation:

f^ = 1.15 "'^

K In (lb/a)

By comparing this equation with Simha's, it will be obvious that the
contribution of a rodlike particle to the viscosity of its solution is much

252

viscometrv

less when the particle is oriented parallel to the direction of flow than
when it is randomly oriented. The situation with respect to flattened
ellipsoids is similar, for Peterlin (19) has shown that the intrinsic
viscosity of oriented plates is very little different from that of spheres.
The equations derived by Simha are potentially very useful,
for they make it possible to interpret the intrinsic viscosity of a suspen-
sion of rigid particles in terms of the ratio of length to thickness of the
particles. It is necessary, however, to have some independent means
of deciding whether the particles resemble rodlike or platelike ellipsoids
of revolution. The Simha equation, particularly the one for rods, is
rather cumbersome in the form presented. Mehl, Oncley, and Simha
(17) have calculated values of intrinsic viscosity which correspond to
various values of the axial ratio, h/a., for both rodlike and disklike
particles. These values are reproduced in Table II. Through the
use of this table, the determination of axial ratios from viscosity
measurements is greatly simplified.

Table II
Intrinsic Viscosity of Ellipsoids of Revolution (17)

bla

hl/A'

b/a

UVK

Elongated

Flattened

Elongated

Flattened

1.0

2.50

2.50

20.0

38.6

14.80

1.5

2.63

2.62

25.0

55.2

18.19

2.0

2.91

2.85

30.0

74.5

21.6

3.0

3.68

3.43

40.0

120.8

28.3

4.0

4.66

4.06

50.0

176.5

35.0

5.0

5.81

4.71

60,0

242.0

41.7

6.0

7.10

5.36

80.0

400.0

55.1

8.0

10.10

6.70

100.0

593.0

68.6

10.0

13.63

8.04

150.0

1222.0

102.0

12.0

17.76

9.39

200.0

2051.0

136.0

15.0

24.80

11.42

300.0

4278 . 0

204.0

The question of paramount importance is whether or not these
theoretical equations of Simha are actually valid. Mehl et al. at-
tempted to arrive at an answer to the question by comparing the
shapes calculated from viscosity data for numerous proteins with those
calculated from difl'usion and sedimentation data, on the assumption
that the protein particles are {a) elongated ellipsoids of revolution and
{b) flattened ellipsoids of revolution. The shapes were calculated
from sedimentation and difl'usion data by the following method.

M. A. LAUFFER

First, the molecular weights were calculated using the Svedberg (23)
equatioft, M = RTS/[D{\ — Vp)\ where R is the gas constant, T
the absolute temperature, S the sedimentation constant, D the diffusion
constant, V the partial specific volume, p the density of the medium,
and M the molecular weight. From the molecular weights and the
partial specific volumes, the particle volumes were computed, and
from these the radii, ro, the particles would have if they were spheres
were calculated. From these radii the friction factors, /o, which the
particles would have if they were spheres were calculated by Stoke's
law, /o = 67r?7ro. The actual friction factors of the various protein
particles,/, were calculated from the diffusion constants by using the
Einstein-Sutherland (7) equation: / = RT/ND, where jV is Avogadro's
constant. The friction ratios, ///o, were then obtained. Perrin (18)
and others have shown from hydrodynamic considerations that, for
elongated ellipsoids of revolution:

^.^ {b/aY^^ V\ - {a/by

J J^ ~ / =^

1 + Vl - {alby

In ;

alb

and for flattened ellipsoids of revolution:

^{hlaY - 1

///o =

(V«)'^'arctanA/(Vfl)2 - 1

Thus, with///o determined from experimental data, h/a values were
calculated on the assumption that the particles were (1) elongated
ellipsoids of revolution and (2) flattened ellipsoids of revolution. The
data considered by Mehl el al. are presented in Table III.

An obvious ambiguity exists in the interpretation of these data,
for it is possible to assume either a rodlike or a platelike ellipsoid as
the model for a protein molecule, unless independent evidence con-
cerning the shape is available. In general, however, it can be seen
that the viscosity data interpreted by the Simha equation are in better
agreement with the diffusion and sedimentation data as interpreted by
the Perrin equation when rodlike ellipsoids of revolution are assumed
to represent the shapes.

A far more serious source of ambiguity is derived from the fact
that Mehl and co-workers had to assume that the particles were un-

VISCOMETRY

Table III
The Shapes of Protein Molecules

///o

\r{\

b/a

Protein

Elongated

Flattened

Diff.+sed. Vise.

Diff. + sed. Vise.

Egg albumin

1.17

5.7

3.8

5.0

4.0

6.7

Serum albumin

1.25

6.5

5.0

5.6

5.4

7.7

Hemoglobin

1.16

5.3

3.7

4.6

3.9

6.0

Amandin

1.28

7.0

5.4

6.0

6.0

8.5

Octopus hemocyanin

1.38

9.0

7.2

7.3

8,2

11.4

Gliadin

1.60

14.6

10.9

10.5

13.6

21.0

Homarus hemocyanin

1.27

6.4

5.2

5.5

5.8

7.5

Helix pomatia hemocyanin

1.24

6.4

4.8

5.5

5.2

7.5

Serum globulin

1.41

9.0

7.6

7.3

8,9

11.4

Thyroglobulin

1.43

9.9

7.8

7.9

9,2

12.7

Lactoglobulin

1.26

6.0

5.2

5.4

5.7

6.9

Pepsin

1.08

5.2

2.5

4.5

2.6

5.8

Helix hemocyanin, /)H 8.6

1.89

18,0

16.6

12.0

23.9

26.0

hydrated in order to obtain the shapes from both intrinsic viscosity
data and friction ratios. Since both friction ratios and intrinsic vis-
cosities are increased by hydration, both sets of calculations could be
in error. Mehl et al. tried to assess the effect of hydration. Bull and
Cooper (5) discussed the issue in greater detail, and presented a
method of estimating the average degree of hydration for proteins from
a consideration of essentially the same data. Hov/ever, since their
method involved an assumption not consistent with the Simha and
Perrin equations, for the validity of both of which direct experimental
evidence has been presented (16), their treatment serves more to
emphasize the need for a reliable method of determining hydration than
to provide such a method. One of the serious obstacles to the under-
standing of the physical properties of proteins and other biological
materials remains the absence of adequate means for assessing quanti-
tatively the role of hydration.

The question of the validity of Simha's equation was recon-
sidered in an investigation of the physical properties of tobacco mosaic
virus preparations (16). Two preparations of the virus in different
states of aggregation were studied in the viscometer and also in the
electron microscope. The one preparation gave a relatively low in-
trinsic viscosity and the electron micrograph showed rod-shaped
particles with a unimodal distribution of lengths. The other prepara-

M. A. LAUFFER

lion held a much liiglicr intrinsic viscosity and a biniodal distribution of
lengths, with one maximum at the same particle length as in the case
of the first preparation and the other maximum at a greater value of
particle length. The ratios of the average length to thickness of the
particles of the two preparations were calculated from the intrinsic
viscosities by means of the Simha equation for elongated ellipsoids;
and the weight averages of the squared axial ratios were computed
for the two preparations from the lengths of the particles determined
by the electron microscope and from the thickness obtained by x-ray
diffraction measurements. The weight averages of the squared axial
ratios were calculated because, according to the Simha equation,
intrinsic viscosity is approximately a function of the amount of ma-
terial and of the square of the axial ratio. The results are shown in
Table IV.

Table IV
Axial Ratio of Tobacco Mosaic Virus Particles

Intrinsic viscosity

bfa

Virus sample

Simha

Electron microscope
plus x-ray

A
B

39
81

20
31

23
29

The determination of particle lengths with the electron micro-
scope is a direct measurement subject only to the error of the empirical
evaluation of the magnification factor of the microscope. There is
evidence to indicate that the tobacco mosaic virus particles in solution
are not appreciably hydrated. However, even if they are hydrated
to a considerable extent, the error in the intrinsic viscosity would be
small for such extremely elongated particles. Therefore, the data
shown in Table IV constitute a direct experimental verification of the
essential correctness of the Simha equation for the intrinsic viscosity
of rodlike ellipsoids of revolution.

The verification of the Simha equation is important because
it removes the most serious objection to the use of viscosity data in the
determination of particle shapes. It is quite important that one be
able to estimate the shape of a biological particle, for, from the shape,
the friction ratio can be computed by means of the Perrin equation.
The evaluation of the friction ratio is necessary for the interpretation of

256

VISCOMETRY

(he diffusion constant or the sedinirnla(if)n constant in terms of the
molecular weight and dimensions of nonsphcrical particles. Thus,
these quantities can he evaluated through viscosity and diffusion meas-
urements or through viscosity and sedimentation velocity measure-
ments. Viscometry, therefore, assumes dignity as a tool useful in
determining the dimensions of suspended particles.

The interpretation of viscosity data is rendered difficult by three
complications. Of these, the disturbing question of hydration is the
most serious. Until a reliable independent method for estimating the
degree of hydration is developed, there will be uncertainty in the
evaluation of the shapes of proteins or other biocolloids from viscosity
data and also from other physical data. The other two complications
are the effect of the charge of the particles upon their intrinsic viscosity
and the effect of the velocity gradient in the viscometer upon the
magnitude of the apparent viscosity coefffcient.

That there should be an effect upon viscosity due to the electric
charge of colloidal particles was first predicted from theoretical con-
siderations by von Smolochowski. The equation expressing this re-
lationship has been rederived by Krasny-Ergen (15) and can be pre-,
sented in the form:

bVK = 1 + ^

87r2 X77or2

The symbols f and r are the electrokinetic potential and the radius of
the colloidal particles, e and rjo are the dielectric constant and viscosity
of the solvent, respectively, and X is the specific conductance of the
solution. Kruyt and Bungenberg de Jong, Bull (4), and Briggs and
his associates (3) have all verified the existence of the electroviscous
effect. The intrinsic viscosity of colloidal particles with high electro-
kinetic potentials in solvents of low conductance can be many times
greater than the intrinsic viscosity of the same particles measured
under conditions under which the charge effects are suppressed. The
studies of Bull and of Briggs, however, show that the Krasny-Ergen
equation does not afford an adequate quantitative representation of
the effect of the electric charge. Thus, the electroviscous effect is
reduced to a pure hazard. The results of Bull and of Briggs show that
it can be suppressed by adjusting the pH to the isoelectric point of the
biocolloid or by adding neutral electrolytes such as sodium chloride.

M. A. LAUFFRR

Electrolyte concentrations as low as 0.02 M are sufficient to reduce
the effect to a relatively small value. An even better, but much more
laborious, method of eliminatins^ the effect is to measure the viscosity
and the other variables in the Krasny-Ergen equation at various
electrolyte concentrations, and then extrapolate the results to zero
electrokinetic potential. Briggs et ol. have shown that this procedure
is entirely satisfactory.

The error due to the variation of the apparent viscosity coeffi-
cient with the velocity gradient is more difficult to eliminate because
it can have more than one cau.se. One of the possible cau.ses of this
error, which will be referred to subsequently as anomalous viscosity,
is hydrodynamic. The Simha equations show that the intrinsic
viscosity of rodiike and platelike ellipsoids of revolution randomly
oriented in the flowing stream should be much greater than the
intrinsic viscosity for the same particles completely oriented parallel
to the direction of flow, as given by the treatments of Eisenschitz and
Peterlin. Thus, the intrinsic viscosity of rodlike or platelike particles
will vary with the velocity gradient for purely hydrodynamic reasons,
because the orientation of the particles depends upon the velocity
gradient. Were this the sole type of anomalous viscosity, the solution
of the problem would be simple, for all one would need to do is measure
intrinsic viscosity at various velocity gradients and extrapolate the
results to zero velocity gradient, where the particles are randomly
oriented and where, therefore, the Simha equations are valid. Robin-
son (21) has shown that the viscosity of tobacco mosaic virus actually
does vary with the degree of orientation of the particles as determined
by optical means, and that the data can apparently be extrapolated
to zero velocity gradient.

However, the anomalous viscosity could be due in part to inter-
particle forces, which confer upon the solution the properties of a
weak gel. This effect will be called structural anomalous viscosity.
Structural anomalous viscosity can be understood in terms of Eyring's
concepts (24). Eyring has postulated that a solution which shows
structural anomalous viscosity is composed of two mechanical systems.
The solvent and perhaps some of the dissolved particles constitute a
medium which obeys Newton's laws of flow. In addition to this,
however, relatively few bonds of high activation energy are assumed
to exist between some of the dispersed particles, forming a network

258

VISCOMETRY

with the properties of an elastic soHd. These bonds could be long-
range intermolecular forces, ordinary chemical bonds at points of
contact between dispersed particles, or forces of some unknown nature.
The only important factors are that they be relatively few and that
they have very high energies. This condition will result in a structure
which will undergo only elastic deformation under low-velocity gra-
dients, but which will flow when sufficiently high shearing forces are
applied. The whole solution, consisting of the Newtonian component
and the network, will exhibit anomalous viscosity. Eyring has shown
that flow curves resembling those actually obtained for many types of
systems exhibiting structural viscosity can be constructed on the basis
of such assumptions.

The difficulty presented by this situation is obvious. If the
anomaly is due solely to hydrodynamic efTects, the problem can be
solved by extrapolating the viscosity data to zero velocity gradient;
but, if the anomaly is due in part to structure, this procedure is de-
feated because the contribution of the structure — the unwanted factor
for our purposes — is at a maximum at zero velocity gradient.

The solution to the difficulty is very simple in theory. Hydro-
dynamic anomalous viscosity in dilute solutions should be independent
of concentration, that is, the ratio of the apparent specific viscosities
at any two values of velocity gradient should not vary with the con-
centration of solute. On the other hand, the structural effect should
tend to vanish as the concentration of solute approaches zero. Values
of the velocity gradient should exist for which the ratio of the apparent
specific viscosities changes with a change in the concentration. Thus,
it is theoretically possible to differentiate between structural and
hydrodynamic anomalous viscosities and to find experimental condi-
tions under which the latter vanishes. Practically, however, two
difficulties remain: first of all, these theoretical considerations have
not yet been investigated experimentally; and, second, it might be
necessary to work with extremely dilute solutions to realize the condi-
tions postulated. This might not be possible until after the develop-
ment of much greater precision than has heretofore been required in
viscometry.

It has been the purpose of this discussion to present some im-
pression of the simplicity of the viscometry technique, of the meaning
of viscosity data in terms of the shape of suspended or dissolved particles,

M. A. LAUFFER

and of the potential usefulness of viscometry in biological and biochemi-
cal investigations. The difficulties which remain as obstacles to the
interpretation of viscosity data have been discussed in terms of recent
theoretical developments. With the exception of the ambiguity due
to hydration, the theories discussed point the way to the resolution of
the uncertainties still encountered. It seems not too much to hope
that studies in the next few years will establish viscometry as a reliable
biochemical or biophysical tool, with dignity and meaning comparable
to diffusion and ultracentrifugation.

References

(1

(2

(3

(4
(5

1
(6
(7

(

(8
(9
10
11
12
13
14
15
16

(
17
18
19
20

21
22
23
24

Arrhenius, S., ^. physik. Chem., 1, 285 (1887).

Bingham, E. C, Fluidity and Plasticity. McGraw-Hill, New York, 1922.

Briggs, D. R., Hankinson, C. L., and Hanig, M., J. Phys. Chem., 45,
866, 943 (1941); 48, 1 (1944).

Bull, H. B., Trans. Faraday Soc, 36, 80 (1940).

Bull, H. B., and Cooper, J. A., Pub. Am. Assoc. Advancement Sci., No. 21,
50 (1943).

Couette, M., Ann. chim. phys., 21, 433 (1890).

Einstein, A., Ann. Physik, 17, 549 (1905); 19, 289 (1906); 34, 591
911).

Eirich, F., Bunzl, M., and Margaretha, H., Kollofd-Z-, 74, 276 (1936).

Eisenschitz, R., Z- physik. Chem., A158, 78 (1931).

Guth, E., Kolloid-Z; 74, 147 (1936).

Hagenbach, E., Ann. Physik Chem., 109, 385 (1860).

Hatscheck, E., Viscosity of Liquids. Van Nostrand, New York, 1928.

Huggins, M. L., J. Am. Chem. Soc, 64, 2716 (1942).

Kraemer, E. O., Ind. Eng. Chem., 30, 1200 (1938).

Krasny-Ergen, W., Kolloid-Z, 74, 172 (1936).

Lauffer, M. A., J. Am. Chem. Soc, 66, 1188 (1944); Chem. Revs., 31, 561
942).

Mehl, J. W., Oncley, J. L., and Simha, R., Science, 92, 132 (1940).

Perrin, F., J. phys. radium, 7, 1 (1936).

Petedin, A., Kolloid-Z-, 86, 230 (1939).

Poiseuille, J. L. M., AUm. Savants Etrangers, 9, 433 (1846).

Robinson, J. R., Proc Roy. Soc London, A170, 519 (1939).

Simha, R., J. Phys. Chem., 44, 25 (1940).

Svedberg, T., Chem. Revs., 14, 1 (1934); 20, 81 (1937).

Tobolsky, A., Powell, R. E., and Eyring, H., in The Chemistry of Large
Molecules. Intcrscience, New York, 1943, p. 125.

260

17

ISOTOPE TECHNIQUE IN
THE STUDY OF INTER-
MEDIARY METAROLISM

D. RITTENBERG, assistant professor in biochemistry, college

OF physicians and surgeons, COLUMBIA UNIVERSITY; THE LILLY

AWARD IN BIOCHEMISTRY

DAVID SHEMIN, associate in biochemistry, college of physicians

AND SURGEONS, COLUMBIA UNIVERSITY

THE STUDY of intermediary metabolism encompasses the
reactions undergone by cellular and dietary constituents as
well as the reactions involved in the formation of transient intermedi-
ates which occur during the degradation of various compounds to prod-
ucts of excretion. In studying these reactions, dietary components
were labeled in an attempt to trace their fate after ingestion by the
animal. These eflforts were but partially successful, since in all cases
the labels employed so changed the properties of the substance that
the cell treated it differently from its natural analogue. Indeed, in
most cases the success attained was actually due to the difference
between the metabolism of the foreign substance and that of its natu-
ral analogue. When fatty acids are administered to a normal dog
the only excretory products produced are carbon dioxide and water;
when the co-phenyl substituted fatty acids are fed (18) there are
excreted co-phenyl substituted acids having a carbon chain with
fewer carbon atoms than the original by one or more pairs. From

261

n. RITTENBERG AND D. SHEMIN

such data, however, Knoop deduced diat the oxidation of fatty acids
proceeds by the removal of two carbon atoms at a time; but the possi-
bihty still existed that the oxidation of the unnatural acids might
proceed by a mechanism entirely diflferent from that of the natural
fatty acids. Such techniques were of but restricted use. It was clear
that little further advance in the study of the intermediary metabolism
could be expected from this direction.

At about the time that Knoop was using the phenyl group to
label fatty acids and amino acids, a more powerful technique was in
the process of development as a result of the advances being made on
the structure of the atomic nucleus. During the 19th century, it
was universally believed that the chemical and physical properties of
an element were a function of its atomic weight. Mendelyeev, by
classifying atoms according to their masses, had been able to predict
the properties of elements as yet unknown. In 1906, however, Bolt-
wood (6) discovered a radioactive element, ionium, which was chemi-
cally and physically identical with thorium but which had a diflferent
atomic weight. Other such examples were found among the radio-
active elements and it soon became clear that the properties of an atom
were determined, not by its mass, but by its nuclear charge, its atomic
number. It was therefore possible to have two or more atoms having
the same nuclear charge but differing atomic weights. Such atoms
were called isotopes by Soddy (41). In 1910 he stated (40), "Chemical
homogeneity is no longer a guarantee that any supposed element is
not a mixture of several diflferent atomic weights, or that any atomic
weight is not merely a mean number."

With the development of the mass spectrometer by Dempster
(11) and Aston (1), it became possible to investigate the iso topic
composition of the stable elements; and it is now known that most
elements have at least two isotopes, though a few, such as phosphorus
and iodine, seem to exist in but one atomic variety. With the excep-
tion of the elements from hydrogen to carbon, every element of an even
atomic number has three or more stable isotopes, while every stable
element with an odd atomic number has two or less stable isotopes.

Our understanding of the laws governing the structure of atomic
nuclei is as yet in a primitive stage. We think that atomic nuclei are
built up from two elementary particles, the proton, a particle having
a mass of one atomic unit and a single positive charge, and the neutron,

262

ISOTOPE TECHNIQUE

an unchanged particle having a mass very nearly the same as that of
the proton. A particle containing 6 protons and 6 neutrons will be
a carbon nucleus, v^hich has a charge of 6. Using the notation of the
nuclear physicists, the particle will be denoted as Cg^ the superscript
indicating its mass and the subscript its nuclear charge. If we add
one more neutron to the nucleus, we obtain Cj^. This is an isotope
of carbon, for we have merely increased the mass and not the nuclear
charge. Introduction of a proton into the Cg" nucleus would yic^ld
Ny^, an isotope of nitrogen, for both the mass and the nuclear charge
have been increased by one unit. The particle Cg^ exists in nature,
for naturally occurring carbon is a mixture containing one atom of
C'l^ for every ninety-nine atoms of Cg". Introduction of one more
neutron into the nucleus of Cl^ would give rise to Cg^, and subtraction
of one neutron from Cf would form Cl^. Both these particles are
carbon atoms and both have been artificially prepared. Neither one
exists in nature, since each is radioactive, with a half-life of about 24,000
years and 21 minutes, respectively. Other carbon isotopes may be
visualized, such as Cl, Cl^, or Cg^, but presumably these are radioactive
and very unstable, i. e., their half-life times must be very short.

A study of the stable isotopes suggests that a nucleus is stable
only if it has approximately the same number of protons and neutrons
in its structure. The forces which bind the nucleus seem to be inter-
actions between pairs of protons and neutrons.

In the last ten years, the physicists ha\'e prepared artificially
radioactive isotopes of practically every element. During the same
period, chemists (16, 20, 28, 42-45), notably Urey, concentrated and
obtained in more or less pure form the heavy isotopes of hydrogen,
oxygen, nitrogen, carbon, and sulfur. At the present time, the bio-
chemist has available for his investigations isotopes, either stable or
radioactive, of every element of biological interest. Table I lists some
of these isotopes with their natural abundance, if stable, and their
half-life time, if radioactive. The elements chosen were those of bio-
logical interest, and it is fortunate that for every element there is avail-
able either a concentrate of a heavy stable isotope or a suitable radio-
active isotope.

The choice of the particular type of isotope to be used for any
specific problem depends on the isotopes available and the type of
problem. In the study of phosphorus metabolism by the isotope

263

D. RITTENBERG AND D. SHEMIN

technique, radioactive phosphorus, Pja, must be employed, since only
one stable atomic species exists, namely, Pis- The radioactive isotope
has a convenient half-life (14 days), and the radiation it emits is easy
to measure. Furthermore, the preparation of Pis is relatively simple.
The choice for the case of oxygen is also clear; the stable isotope O^
should be used since the radioactive oxygen isotope Og has a half-life
of only 126 seconds, a time too short to permit the carrying out of
almost any biological experiment. Similarly, the stable heavy hy-
drogen. Hi, is preferable to the radioactive isotope, Hi. The reason
in this case is not a restricted half-life but the fact that the radiation
emitted by Hi is so soft that it is extremely difficult to measure. Also,
in this particular case, the measurement of Hi concentrations is simpler
than that of the radioactivity.

By this reasoning we conclude that the stable isotopes should be
employed for studies of hydrogen, nitrogen, and oxygen, and the radio-
active isotopes for sodium, phosphorus, calcium, iron, and iodine.
The choice for carbon and sulfur is more difficult. While Gg cannot
compete with Cl^, isotope C" can. The actual choice here depends
on highly technical points involving the analytical procedures and
the relative cost of preparation of the isotopes. It seems likely that,
at least for the next five years, Gg^ rather than Cl^ will be employed
but that thereafter G" may supplant it. The choice in the case of
sulfur is also not easy, since neither form (Sjg or SH) has a definite
advantage over the other.

One of the problems which enters into the choice between the
stable and the radioactive isotope is the relative ease of the isotope
determinations. With the exception of hydrogen, all stable isotopes
require the use of a mass spectrometer for their measurement. The
apparent complexity and cost of the present type of mass spectrometers
have often acted as a deterrent to investigators who have contemplated
employing this technique. It is likely that, after the war, the mass
spectrometer will be greatly simplified and that its cost from some
commercial instrument maker will not exceed that of a good optical
spectrograph.

The analysis of organic compounds for deuterium may be car-
ried out by the falling drop procedure (17). The apparatus involved
and the procedures used are so simple that no competent investigator
need have any hesitation about setting up such equipment. However,

264

ISOTOPE TECHNIQUE

Table I
Some Natural and Radioactive Isotopes"

Relative

Relative

Element

Mass

abundance,

atom

per cent

Half-life

Element

Mass

abundance,

atom

per cent

Half-life

H

1

99.99

s

31

3 . 2 sec.

2

0.01

32

95.1

—

3

■ ■

30 yrs.

33

34

0.7
4.2

—

C

10

—

8 . 8 sec.

35

—

88 days

11

—

21 min.

36

0.02

—

12

98.9

—

13

1.1

—

Ca

39

— •

4 . 5 min.

14

10'-105 yrs.

40
41

96.97

8 . 5 days

N

13

—

9.9 min.

42

0.64

—

14

99.63

—

43

0.15

—

15

0.37

—

44

2.06

—

16

—

8 sec.

45
46

0.003

180 days

O

15

—

126 sec.

48

0.19

—

16

99.76

—

49

—

2.5 hrs.

17

0.04

—

18

0.20

—

Fe

53

—

8 . 9 min.

19

—

31 sec.

54
55

6.04

4 yrs.

Na

21

23 sec.

56

91.54

—

22

—

3 yrs.

57

2.11

—

23

100

—

58

0.28

— ■

24

—

14.8 hrs.

59

— •

47 days

Mg

23

11.6 sec.

124

— .

4 days

24

77.4

—

I

126

—

13 days

25

11.5

—

127

100

—

26

11.1

—

128

—

25 min.

27

—

10.2 min.

130
131

—

12.6 hrs.
8 days

P

29
30
31

32

100

4.6 sec.
2 . 6 min.

14.3 days

>131

2.4 hrs.

" Data from R. D. Evans in Medical Physics.
Chicago, 1944.

Year Book Publishers,

e\en in this case the ina.ss spectrometrir method is probably superior
to that of the falhng drop if only for the reason that smaller samples
can be analyzed. From 100 to 200 milligrams of organic compound
is necessary for analysis by the falling drop technique, but from 3 to
5 milligrams is sufficient for analysis by the mass spectrometer.

265

1). RriTENBERG AND D. SHEMIN

The essence of tiic isotope technique in biochemical research
consists in the preparation of a compound in which one or more of the
atomic comiDonents have an abnormal isotope concentration. In
t[iis manner, a substance is obtained which, by suitable isotope analysis,
can be detected and estimated quantitatively in the presence of either
its normal analogue or of other compounds, and which behaves like
its normal analogue in a biochemical system.

Since the electronic configurations of the isotopes of an element
are almost identical, isotopic compounds have similar chemical propei'-
ties. There are, however, some physical properties which are de-
pendent on the molecular mass. Diffusion is one such property, since
the rate of this process is inversely proportional to the square root of
the mass. Nevertheless, these differences in physical properties are
not of importance in cellular metabolism. Molecules of glycine,
heavier by one unit of mass because of the introduction of one atom
of a heavy isotope of either hydrogen, nitrogen, or carbon, will differ
from the normal analogue in their rate of diffusion by about 0.7%.
There is direct evidence available that the cell does not distinguish
between isotopic isomers. Compounds containing one or more heavy
atoms of hydrogen, nitrogen, etc. normally exist in nature. For
example, 0.37% of all glycine molecules contain N^^ atoms instead of
N*^. Such compounds are therefore not foreign to the cell. If living
cells accorded special treatment to isotopic isomers, we should expect
the isotope concentrations to vary as we pass from tissue to tissue and
from the living into the inorganic world. However, the natural
abundances of the heavy isotopes of all elements studied are uniformly
the same in organic and in inorganic substances. It is striking that
this normal abundance of isotopes even applies to meteorites originat-
ing from interstellar space (21).

The labeled compounds prepared must, of course, be of such a
nature that the isotope distribution will not be altered by mere ex-
change. Thus it is impossible to label ammonia in the water system
with either deuterium (Hj) or tritium (Hi), since the hydrogen atoms
of ammonia are readily replaced by those of the water. Several other
restrictions limit the scope of deuterium (33). When an organic
compound is labeled with deuterium, the carbon chain is not being
directly labeled, for deuterium may be lost from an organic compound
even though no reactions involving the carbon chain take place.

266

ISOTOPE TECHNIQUE

Tims, <Y.n:'-(li(leiiteriosiiccinic acid in llir li\ino' cell inav lose all its
deuteiium by repeated dehydrogenations and hydrogenation (12).
The appearance of deuterium in the cellular fluids cannot be regarded
as proof that the succinic acid has been oxidized to carbon dioxide and
water. If, on feeding a deuterio compound to an animal, one isolates
another compound containing a higher concentration of deuterium
than exists in the body fluids, it»may be taken as positive proof of the
utilization of the first compound in the synthesis of the second. The
power and elegance of this piocedurc is illustrated by the finding that
the feeding of deuterioacetate to rats leads to the formation of deuterio-
cholesterol (3). There is no other method at present available which
could have disclosed such a complex synthetic reaction. It further
illustrates the biological stability of the C — D bond.

While, at first sight, the fact that the C — D bond can be broken
in the course of chemical reactions (by dehydrogenations, enolizations,
etc.) seems to limit the field of applicability of deuterium, more careful
considerations suggest that this is, in actuality, ojie of the most useful
properties of deuterium. The same reactions which will remove
deuterium from an organic compound will also introduce it if the
medium in which the reaction occurs contains heavy water (D2O).
Injecting D2O into an animal makes possible the study of a host of
reactions and their rates, by determining the velocity of introduction
of deuterium into the organic compound. When the body fluids
of mice are enriched in D2O, the isotope concentration in the fatty
acids increases and finally reaches a value half that of the body fluids.
The isotope enters the fatty acid molecule in the course of the reactions
involved in the synthesis of the fatty acids. From such observations
(26) it was concluded that half the fatty acids of the mouse was de-
graded and resynthesized in about seven days.

On the whole, it appears that deuterium is at present the most
valuable single isotope available to the biochemist. In the use of
heavy nitrogen as a label for nitrogenous compounds,' it must also be
remembered that reactions exist which may detach the labeled atom
and transfer it to another molecule. This is of course especially true
in the case of an amino acid. The use of carbon for labeling organic
compounds is not .subject to such uncertainty becau.se no exchange
reactions occur. Further, the conversion of one organic compound
to another is, by definition, the utilization of the whole or part of the

267

D. RITTENBERC AND D. SMEMIN

carbon skeleton of one coiiipoinul loi' (he foiination of another. Tvahel-
int;- a compomul willi carbon i,sol()|)es is (liiis the most direct nielhod
for demonstratinej conversions.

The biochemical isoiopc (cchniqnc has been applied in four
fields: (1) in delectinL; die conxcrsion of one compound to another;
(2) in studying llie mechanism of biochemical reactions; (3) in measur-
ing the rates of reactions; and (4) in determining the amount of a
constituent in a mixture. It has the advantage over all other methods
that it may be carried out on the intact organism under normal physio-
logical conditions.
"^ The first biological experiment in which the isotope technique

( was used was worked out by Hevesy in 1923 (15). He traced the
transport of lead in plants whose roots were immersed in a medium
containing a radioactive isotope of lead. This type of experimentation
marked time until Urey, in 1933, discovered and concentrated heavy
water. At about the same time, artificial radioactivity was discovered.
As a result, there became available to the biochemist concentrates
either of heavy isotopes or of radioactive isotopes of all the elements of
importance in the study of intermediary metabolism.

Beside the numerous individual problems which in the past
ten years have been attacked by this method, the isotope technique
has revealed one phenomenon of fundamental significance in the living
cell, viz., the dynamic state of tissue constituents. These experiments
have shown that practically every component of the animal body is
constantly degraded and resynthesized at a rapid rate, and that the
former distinctions between structural and metabolic components of
the cell or of endogenous and exogenous metabolism are nonexistent.
The maintenance of the form and structure of the adult cell or tissue
is the result, not of a static state in which no reactions take place, but
rather of a dynamic equilibrium in which the synthetic and degradative
reactions proceed at equal rates. The existence of this stationary
state has been shown by experiments with several isotopes and with
every type of tissue, with the possible exception of the tissues of the adult
brain. Not only is the form and structure of the organ governed by
the rates of synthesis and degradation of its constituents but also its
composition. While the autolysis of tissues which becomes apparent
on the death of an animal was formerly thought to involve reactions
which began after death, we now believe that many of these degrada-

268

ISOTOPE TECHNIQUE

live reactions occur continuously. During life, the effects of these
degradative processes are nullified by synthetic reactions which re-form
as much tissue constituents per unit time as are destroyed by the auto-
lytic reactions; at death, the synthetic reactions, which in general are
coupled to oxidative reactions, cease and we observe the effects of
degradative processes alone. The mechanisms by which the cell
regulates these reactions are completely unknown. Since these reac-
tion rates can only be investigated in the intact animal, the only method
at present available which can be used is the isotope technique. It
can safely be assumed that the determination of these reaction rates
will occupy the attention of many investigatois in the coming years.

There are several published researches which illustrate the
method by which the isotope technique may be applied to the study
of the conversion of one compound to another in the intact animal.
It was known from feeding experiments with immature rats that,
while phenylalanine was an essential component of the diet, tyrosine
was not (29). From the chemical similarity of these amino acids, it
was suspected that phenylalanine could be oxidized to tyrosine. Fur-
thermore, in tyrosinosis, tlie feeding of either tyrosine or phenylalanine
results in excretion of /^-hydroxyphenylpyruvic acid (22). It should
be recognized, however, that this latter piece of evidence is not as
definitive as it appears. The fact that the feeding of a compound to
an animal results in the excretion of an excess of another related com-
pound cannot be taken as positive proof that the first compound has
been converted to the second, even if, as in some cases, a stoichiometric
relationship appears to exist between the amounts of these substances.
It is well known that the feeding of mineral oil to rats results in an
enhanced excretion of fecal steroids. This is certainly not the result
of the conversion of these hydrocarbons to cholesterol but is due to
interference by the mineral oil with the reabsorption of cholesterol by
the intestinal tract. Further, the addition of glycine to the diet
results, in the next 24 hours, in an added excretion of nitrogen in the
urine equivalent to that of the added glycine. The conclusion that
this extra urinary nitrogen is derived from the dietary glycine would
not be correct. We have fed u> luunans small quantities of glycine
labeled with N'^ (ab(jut 10 mg. per kilo weight) and have foimd that
only about 33% of the labeled glycine nitrogen appears in the urine
in the next 24 hours. E\'en after 3 days only al)Out 50% has been

269

D. RITTENBERG AND D. SHEMIN

excreted. The other nitrogen which is excreted must have arisen from
other sources, probably the tissue proteins (unpubhshed experiments).

The phenylalanine-tyrosine conversion was studied in rats by
feeding them a normal diet to which was added a small amount of
labeled phenylalanine (23) prepared by replacing the normal hydrogen
atoms of the benzene ring with deuterium atoms. From the tissues
of rats which had been fed this labeled phenylalanine, tyrosine was
isolated containing a high concentration of deuterium. Here is a
direct proof for the conversion. This transformation of phenylalanine
to tyrosine took place, not on a diet deficient in tyrosine and calculated
to force its synthesis, but on a diet containing an adequate amount.
The conversion of phenylalanine to tyrosine is independent of the
dietary composition. Such a reaction could have been detected by
no previously known experimental technique.

The method just described is of course of the broadest generality,
but there are some cases in which its experimental application is
difficult. Recourse must then be taken to more indirect methods.
The feeding to animals of benzoic acid results in the excretion of
hippuric acid, benzoylglycine. In experiments in which glycine
labeled with heavy nitrogen was fed to rats, it was found that one- third
of the glycine used for conjugation with the benzoic acid was of dietary
origin and two-thirds was supplied by the proteins of the tissues (27).
The feeding of benzoic acid thus results in the excretion of a sample
of glycine of the tissues. Conversions of other compounds to glycine
may easily be tested by feeding the labeled test substance, together
with benzoic acid. The feeding of these compounds which are con-
verted to glycine in vivo will result in the excretion of hippuric acid
containing a high concentration of N^l Serine gives such a positive
result, whereas leucine, alanine, and ethanolamine give negative
results (39). The fact that ethanolamine is not converted to glycine
indicates that the mechanism for the conversion of serine to glycine
does not involve decarboxylation to ethanolamine and subsequent
oxidation of the alcohol, but suggests that the /3-carbon atom of serine
is split off to yield glycine directly. The proof of the conversion of
serine to glycine is not as direct as that of phenylalanine to tyrosine,
for only the amino group of serine was labeled and not the carbon
chain, it is conceivable that only the amino group was transferred.
In control experiments it has been demonstrated that the feeding of

270

ISOTOPE TECHNIQUE

equivalent amounts of ammonia does not give rise to a high concentia-
tion of N^^ in the excreted hippuric acid. This eUminates the pos-
sibility that the serine was deaminated and that the ammonia was
used to form glycine. The conversion of serine to glycine was confirmed
and the mechanism established by labeling the carboxyl group of serine
with C^^. Serine is converted to glycine by the splitting oflf of the j8-
carbon atom (unpublished experiments).

A good example of the use of isotopes is to be found in the ex-
periments which led to establishment of the precursors of creatine.
The balance type of experimentation had completely failed to solve
this problem, and had not even been able to determine whether
creatine was biologically dehydrated to creatinine. Examination of
the structure of creatine suggested that glycine, arginine, and methio-
nine might be related to creatine. The feeding of labeled com-
pounds demonstrated that the COOH — CH2N — group of creatine

was derived from glycine (5), the Cf group trom the amidine

^ ■ \NHo

group of arginine (5) and the methyl group froni methionine (47).

In the last few years, a beginning has been made in the measure-
ment of the rates of the reactions involved in the synthesis of tissue
proteins. Feeding an animal with an amino acid labeled with N^^
results in the incorporation of this nitrogen into most of the amino
acids of the proteins (31), a process which is of course dependent on the
rate of the formation of peptide bonds and the rate of transfer of
nitrogen from one amino acid to another. Measurement of the rate
of increase or decrease of isotope concentration of tissue proteins gives
a measure of the rate of formation of the proteins. Such studies reveal
a remarkable rate for protein .synthesis in rat livers; half of the protein
is degraded and resynthesized in six days (36).

Not only are the "normal proteins" of animal tissues in a state
of dynamic equilibrium but also the antibodies, as shown with the aid
of isotopes in an actively immunized rabbit. The antibody protein
is broken down and newly synthesized even when the antibody titer ^
is declining. The rate of decline of the antibody titer is not deter-
mined by an uncompensated destruction of the antibody protein but
rather is the result of two opposing reactions of which the rate of break-
down is faster than the rate of synthesis. The feeding of isotopic

271

D. RITTENBERC AND D. SHEMIN

glycine to an actively immunized ral)l)il at llie stage at vvhicli the titer
is lalling results in the rapid incorporation of the isotopic nitrogen in
the antibody even in the phase of decline (32).

The rates of protein synthesis and destruction have been studied
in growing tissues, regenerating ii\er, and tumor tissues. Since it
has been shown by the isotope technique that tissues are in a state of
flux, it is reasonable to conclude that the synthetic or anabolic rate in
a tissue of constant weight must be equal to the degradative or cata-
bolic rate. In a growing tissue, however, the rate must be greater
than the degradative rate. This disparity in rates may be the result
of a relative increase of the synthetic rate as compared with that in
the nongr owing tissue, of a relative decrease of the rate of degradation,
or of a combination of the two possibilities. Which of the three
possibilities operates and accounts for growth in regenerating liver
tissue can be ascertained by the isotope technique. From unpublished
data obtained in our laboratory, it appears that, in the regenerating
tissue, the synthetic rate is similar to the rate found in the nongrowing
liver but the degradative rate is markedly decreased. Growth is the
result of an inhibition rather than the initiation or acceleration of re-
actions. It seems likely that the next few years will see an intensive
study of the rate of synthesis of various tissue components under diverse
physiological conditions in the intact animal.

An extension of the isotope technique is the isotope dilution
method of analysis (25), a technique which has been very useful in
various problems, particularly that of accurately determining the
amount of diflferent compounds in a mixture. The technique is based
on the fact that a compound which contains more than the normal
abundance of isotope is inseparable from its normal analogue by the
usual laboratory procedures. In this procedure, a small amount, for
example, of isotopic glutamic acid is added to a protein hydrolyzate
and a representative sample of glutamic acid is isolated from this
mixture. From the amount of glutamic acid added (A) and its N'^
concentration (Co) and the N^^ content of the isolated glutamic acid
(C), the amount of glutamic acid (B) originally present in the mixture
can be calculated from the formula: B = A[(Co/C) - 1]. This
technique applied to the determination of the amount of an amino
acid in a pure protein has provided the most reliable analytical values
at present available (14,38). One of the interesting developments of

272

ISOTOPE TECHNIQUE

the Lsolopc clilnlion iiuiliod li;is I)C(mi ils use in (Iclcniiiiiing whether
or not certain siihstances fire iinoKccI in aniinal inctabohsm. For
example, it has been rlemonslialed that large amounts of acetate are
produced daily in the rat thongh its (oneentratinn is never great enough
to he measured directly (4).

The role of r/-ainino acid oxidase in the animal organism is
obscure, though it appeared possible that its function was to deaminate
<f-amino acids that might arise from a partially or completely symmet-
rical synthesis of amino acids from a-keto acids (19). The demon-
stration that </-amino acids are not formed in the animal organism
would be very difficult with ordinary techniques. If, however, a
substance labeled with an isotope is administered to an animal and
subsequently isolated either from the tissues or the urine, then the
isotope concentration of the recovered compound will indicate whether
or not it has been diluted by some of the same nonisotopic substance
synthesized by the animal. f/jZ-Glutamic acid and f/,/-tyrosine labeled
with N'^ were administered to rats and the (/-amino acids isolated from
the urine. Since the isotopic content of the isolated (/-amino acid was
the same as the amino acid administered, it would appear that (/-amino
acids are not formed in the animal organism (35). By a similar pro-
cedure Bernhard (2) demonstrated that the long-chain dicarboxylic
acids are not intermediates in the metabolism of the fatty acids, as had
been suggested by Vei'kade (46). Bernhard used deuteriodicarboxylic
acids, and isolated from the urine of the test animal fatty acids con-
taining the same deuterium concentration as the sample that had been
fed.

The finding of N^^ in the a-amino group of amino acids of an
animal fed either labeled ammonia or an amino acid demonstrates
that the organism can synthesize these amino acids from the corre-
sponding keto acids. Such evidence throws no light on the question
as to whether an amino acid is essential or nonessential from the stand-
point of nutrition, since, while the cell may be able to convert a keto
acid to the required amino acid, it may not be able to synthesize the
appropriate keto acid. This seems to be true for most essential amino
acids. On the other hand, if, under these circumstances an amino
acid contains no labeled nitrogen, then we can conclude that the animal
cannot synthesize it, even though the carbon skeleton is available.
Examples of these two cases, in the rat, are leucine (24) and lysine

]->. RITTENBF.RG AND D. SHEMIN

(31). After the feeding of labeled amino acids, N^^ can be found in
the a-amino group of leucine, which suggests that a-ketoisovaleric
acid can be aminated. No N^^, however, is found in lysine, indicating
either that a-keto-e-aminohexanoic acid is not formed, or that it could
not be aminated if formed. In general, it is not possible by the use
of isotopic nitrogen to determine whether the carbon skeleton is essen-
tial. In certain special cases, such as that of histidine, it becomes
feasible. When the amino acid is isolated from rat tissues after the
administration of other labeled amino acids, it is found that all the
N^* is present in the a-amino group (34), none being found in the
imidazole ring. This, taken in conjunction with the nutritional re-
placeability of /-histidine by ^-histidine and by imidazole lactic acid
(9,10), forms direct evidence that the organism cannot synthesize the
histidine skeleton. It has been claimed that histidine is not essential
in man (30); by the feeding of a labeled amino acid to a human it
should be possible to settle this and other such problems.

A detailed study of the N^^ concentrations in the four nitrogen
atoms of arginine isolated from the tissue of rats which have received
glycine labeled with N'^ indicates that the a-amino group is not sub-
jected to the reversible deamination-reamination reaction. On deg-
radation of the arginine, it is found that the nitrogen atoms of the
a- and 5-carbon atoms contain the same N^^ concentration (37). The
nitrogen on the 5-carbon atom of arginine cannot be involved in a
reversible deamination-reamination reaction since it is blocked by the
amidine group. Were the a-amino group subject to reversible de-
amination-reamination, the likelihood of its isotope concentration being
the same as that of the nitrogen on the 5-carbon atom would be small.
The mechanism for the formation of arginine does not seem to involve
the amination or transamination of the corresponding keto acid. The
introduction ol equal concentrations of N^^ in the a- and 5-nitrogen
atoms of arginine is l^rought abovu through a cyclic process involving
the formation of ornithine from proline (37). Only in the cases of glu-
tamic acid (7,8,13), aspartic acid (7,8), and alanine (7,8), are the
mechanisms for their formation from keto acids known with certainty.
It is known, of course, that the keto acids of some essential amino acids
may substitute for the amino acid in growth experiments; but the
mechanism of the amination is unknown. Therefore, other mecha-
nisms may exist for the formation of amino acids, such as conversion of

274

ISOTOPE TECHNIQUE

one amino acid to another, examples being the conversion of serine to
glycine and the conversion of phenyl-alanine to tyrosine.

References

(1) Aston, F. W., Phil. Mag., 39, 449 (1920).

(2) Bernhard, K., Helv. Chim. Acta, 24, 1412 (1941).

(3) Bloch, K., and Rittenberg, D., J. Biol. Chem., 145, 625 (1942).

(4) Bloch, K., and Rittenberg, D., J. Biol. Chem. (in press).

(5) Bloch, K., and Schoenheimer, R., J. Biol. Chem., 138, 167 (1941).

(6) Boltvvood, B. B., Am. J. Sa., 24, 370 (1907); 25, 365 (1908).

(7) Braunshtein, A. E., and Kritsman, M. G., Enzymologia, 2, 129 (1937).

(8) Cohen, P. P., J. Biol. Chem., 136, 565 (1940).

(9) Cox, G. J., and Berg, C. P., J. Biol. Chem.. 107, 497 (1934).

10) Cox, G. J., and Rose, W. C, J. Biol. Chem., 68, 781 (1926).

11) Dempster, A. J., Phys. Rev., 11, 316 (1918).

12) Erlenmeyer, H., Schoenauer, VV., and Siillmann, H., Helv. Chim. Acta,
19, 1376 (1936).

13) Euler, H. v., Adler, E., GiJnther, G., and Das, N. B., Z- physiol. Chem.,
254, 61 (1938).

14) Foster, G. L., J. Biol. Chem., 159, 431 (1945).

15) Hcvesy, G., Biochem. J., 17, 439 (1923).

16) Huffmann, J. R., and Urey, H. C, bid. Eng. Chem., 29, 531 (1937).

17) Keston, A. S., Rittenberg, D., and Schoenheimer, R., J. Biol. Chem.,
122, 227 (1937).

18) Knoop, F., Beitr. chem. Physiol. Path., 6, 150 (1905).

19) Knoop, F., Z- physiol. Chem., 274, 291 (1942).

20) Lewis, G. N., and Cornish, R. E., J. Am. Chem. Soc, 55, 2616 (1933).

21) Manian, S. H., Urey, H. C, and Bleakney, W., J. Am. Chem. Soc, 5S,
2601 (1934).

22) Medes, G., Biochem. J., 26, 917 (1932).

23) Moss, A. R., and Schoenheimer, R., J. Biol. Chem., 135, 415 (1940).

24) Ratner, S., Schoenheimer, R., and Rittenberg, D., J. Biol. Chem.,
134, 653 (1940).

25) Rittenberg, D., and Foster, G. L., J. Biol. Chem., 133, 737 (1940).

26) Rittenberg, D., and Schoenheimer, R., J. Biol. Chem., 121, 235 (1937).

27) Rittenberg, D., and Schoenheimer, R., J. Biol. Chem., 127, 329 (1939).

28) Roberts, I., Thode, H. G., and Urey, H. C, J. Chem. Phys., 7, 137
(1939).

29) Rose, VV. C, Science, 86, 298 (1937).

30) Rose, VV. C. Haines, W. S., Johnson, J. E., and Warner, D. T., J. Biol.
Chem., 148,457 (1943).

D. RITTENBERG AND D. SHEMIN

(31) Schoenheimer, R., Ratner, S., and Rittenberg, D., J. Biol. Chem.,
130,703 (1939).

(32) Schoenheimer, R., Ratner, S., Rittenberg, D., and Heidelberger,
M., J. Biol. Chem., 144, 545 (1942).

(33) Schoenheimer, R., and Rittenberg, D., Physiol. Revs., 20, 218 (1940).

(34) Schoenheimer, R., Rittenberg, D., and Keston, A. S., J. Biol. Chem.,
127, 385 (1939).

(35) Shemin, D., and Rittenberg, D., J. Biol. Chem., 151, 507 (1943).

(36) Shemin, D., and Rittenberg, D., J. Biol. Chem., 153, 401 (1944).

(37) Shemin, D., and Rittenberg, D., J. Biol. Chem., 158, 71 (1945).

(38) Shemin, D., J. Biol. Chem., 159, 439 (1945).

(39) Shemin, D., Federation Proc, 4, 103 (1945).

(40) Soddy, F., Ann. Kept. Chem. Soc, 7, 285 (1910).

(41) Soddy, F., Ann. Kept. Chem. Soc, 10, 263 (1913).

(42) Thode, H. G., Gorham, J. E., and Urey, H. C, J. Chem. Phys., 6, 296
(1938).

(43) Thode, H. G., and Urey, H. C, J. Chem. Phys., 7, 34 (1939).

(44) Urey, H. C., Brickwedde, H. G., and Murphy, G. M., Phys. Rev., 39,
164 (1932); 40, 1 (1932).

(45) Urey, H. C, Fox, M., Huffmann, J. R., and Thode, H. G., J. Am.
Chem. Soc, 59, 1407 (1937).

(46) Verkade, P. E., and van der Lee, J., Z- physiol. Chem., 227, 213 (1934).

(47) du Vigneaud, V., Chandler, J. P., Cohn, M., and Brown, G. B., J.
Biol. Chem., 134,787 (1940).

276

IS

MUCOLYTIC ENZYMES

KARL MEYER, associate professor of biochemistry, department

OF OPHTHALMOLOGY, COLLEGE OF PHYSICIANS AND SURGEONS, COLUMBIA
university; chemist to the institute OF OPHTHALMOLOGY,

PRESBYTERIAN HOSPITAL

MUCOLYTIC ENZYMES may be defined as enzymes
which primarily catalyze the depolymerization of
highly polymerized mucopolysaccharides. * Mucopolysaccharides oc-
cur in many internal and external structures of animals and micro-
organisms. The biological importance of some of these substances
in a few instances is now well recognized, although at present our
knowledge of the functions and biological relationship of most of these
substances is still very meager. In some instances, either identical or
closely related mucopolysaccharides as, for example, hyaluronic acid,
occur both in the animal body and in microorganisms. In other
instances, mucolytic enzymes, such as lysozyme and hyaluronidase,
occur both in microorganisms and in the animal. Apart from their
biological interest, mucolytic enzymes are one of the most valuable
aids in the isolation and characterization of mucopolysaccharides.
Their high specificity is comparable to that of immunological reactions.
Their usefulness is somewhat limited by the lack of purity in the avail-
able enzyme preparations, a drawback, however, which can be over-
come by the judicious application of enzymes from different sources.

* Mucopolysaccharides are hexosamine-containing polysaccharides. For
their classification see reference (23).

277

KARL MEYER

This essay will deal with some of the problems connected with muco-
lytic enzymes and their substrates.

Lysozyme

Lysozyme is a bacteriolytic enzyme which effects the lysis of
some microorganisms, notably micrococci and sarcinae. Fleming
(13) recognized the wide distribution of this agent in nasal secretion,
saliva, tears, leucocytes, and egg white. It also occurs in some micro-
organisms in appreciable concentration, as in some molds (18), and
in a strain of Sarcina and in a white staphylococcus {S. muscae), both
of which are susceptible to lysis by lysozyme obtained from egg white
(40). The highest concentration of lysozyme is found in egg white,
although calculation on the basis of total protein shows that human
tears actually have a higher concentration of the enzyme than does
egg white.

Lysozyme is a basic protein of a molecular weight of about
18,000, quite stable toward heat and acid reaction, but relatively un-
stable toward oxidation and alkali (33). It was first obtained in
crystalline form by Abraham and Robinson (1). Recently an im-
prov^ed method for its preparation and crystallization was reported
(2). The biological activities of amorphous and crystalline lysozyme
are identical, as are the activities of lysozyme prepared by our method
and that of the California group. This activity is measured by ob-
serving the extent of clearing of a suspension of a susceptible test organ-
ism, either visually or photoelectrically. The test organism most
widely used is Micrococcus lysodeikticus, which was isolated from the air
by Fleming. This micrococcus is more susceptible than other organ-
isms, although a strain of Sarcina lutea has been found by Df . Rose
Feiner (unpublished) which is as susceptible as A/, lysodeikticus. Most
strains of sarcinae, however, have a susceptibility markedly less than
that of M. lysodeikticus.

Lysis by lysozyme does not take place at an acid pYi, although
the microorganisms have been shown to be killed by the enzyme at
such a/>H. When the /jH is raised to neutral, visible lysis takes place,
but this process is independent of the presence of lysozyme. In fact,
the optimum pW for the action of lysozyme is in the acid region {'^pH
4.5). Some susceptible microorganisms are not lysed at all by lyso-

278

MUCOLYTIC ENZYMES

zyme until, after incubation with the enzyme, llie/;H is raised to about
13, when visible lysis takes place, as in the case of S. muscae. There
are also instances of organisms which are killed without any visible
lysis even at high pH values, e. g., Bacillus mcgatherhwi (35).

Lysozyme, in contrast to most antibiotic agents, acts not only
on living" but also on killed microorganisms, but in the latter case
generally to a lesser degree. For example, M. lysodeiklicus killed by
chilled neutral acetone is lysed almost to the same degree as are living
organisms by lysozyme, while organisms killed by acidified acetone,
by alcohol, or especially by autoclaving are lysed to a much smaller
degree. Visible clearing apparently is a poor, but convenient, method
for the estimation of lysozyme.

The phenomenon of bacterial lysis, whether by lysozyme,
phage, or other lytic enzymes, obviously is complex. Under the
influence of lysozyme the susceptible organisms sw^ell to several times
their volume, the Gram stain becomes negative, uptake of oxygen
stops, and in the supernatant solution nonprotein. nitrogen, inorganic
phosphate, and reducing substances appear (31). U seems quite
ob\ ious that one enzyme cannot catalyze such dixcrse reactions.

The mechanisni of lysozyme action may be explained in terms of
hydrolysis of a substance of mucoid nature which is contained in the
bacterial membrane. That this type of hydrolysis takes place was
demonstrated by the appearance of reducing substances when ly.sozyme
was incubated with carbohydrate fractions extracted from susceptible
organisms following alkaline hydrolysis. The carbohydrate obviously
is not a capsular substance which is easily given off into the surrounding
media as is the specific soluble substances of pneumococci. The
carbohydrate, at least in highly susceptible organisms, occurs in a
quantity of about 3 to 5%.

The findings of the writer's laboratory have been confirmed by
Epstein and Chain (10) and by Pirie (35). The carbohydrate was
said by Epstein and Chain to contain both a ketohexose and an amino
sugar, evidence for the former being a positive SeliwanofT reaction in
the dialyzates after incubation with lysozyme. The SeliwanofT
reaction, however, is not given by the substrate of lysozyme, but by a
carbohydrate present in M. lysodeiktkits with which the preparations
of Epstein and Chain must have been contaminated.

As isolated in this laboratory, the substrate of lysozyme is a

KARL MEYER

mucopolysaccharide of high molecular weight, forming extremely
viscous aqueous solutions. It precipitates rabbit antisera against
M. lysodeikticus in a dilution of 1 : 1,000,000. Crystalline lysozyme in a
concentration of 1 to 5 mg. immediately precipitates 10 mg. of the
polysaccharide, as does protamine. In a concentration of a few
gamma, lysozyme abolishes within a few minutes the viscosity of
aqueous solutions of the polysaccharide. On incubation for two hours,
the hydrolysis of glucosidic linkages is almost complete. In the
presence of 1 mg. of the polysaccharide, the lytic action of lysozyme
on living M. lysodeikticus is decreased approximately 100-fold, indicating
a competition for the enzyme between the added substrate and the
substrate contained in the microorganism.

The action of lysozyme on this carbohydrate from M. lysodeik-
ticus is highly specific. Organ extracts and enzyme preparations which
contain no lysozyme have no depolymerizing effect on the carbo-
hydrate, while lysozyme from sources other than egg white depoly-
merizes the carbohydrate. From organisms pretreated with lysozyme,
no mucopolysaccharide fraction was obtained which Ulcerated re-
ducing groups when brought into contact with lysozyme.

The mucopolysaccharide appears to be firmly bound to the
bacterial membrane. The primary mechanism of lysozyme action
apparently consists of a depolymerization of this mucopolysaccharide,
leading to water imbibition by the organism and to disorganization of
the microbial cell. The breakdown of protein and of organic phos-
phates are effects due to autolytic enzymes, which are apparently re-
sponsible for some part of the visible clearing of the bacterial suspen-
sions.

It early became obvious to Fleming and his co-workers that
lysozyme was not identical with bacteriophage. But it appears from
available data that lysozyme or lysozyme-like enzymes are associated
with phage activity. As the result of the work of several investi-
gators—Gratia and Rhodes (14), Twort (41), WoUmann and Woll-
mann (42) — it seems established that, in some cases, as a result of phage
action, a lytic enzyme is released which is capable of dissolving"
heat-killed and phage-resistant living organisms of the same species
as the lysed strain. The.se lytic agents apparently originate in the
bacterial cells and in most instances can be expected to ha\'e speci-
ficities other than that of lysozyme.

280

MUCOLYTIC ENZYMES

Wollmann and Wollmann (42) found, however, that an enzyme
obtained from staphylococci which had been lysed by phage lysed
sarcinae, organisms sensitive to lysozyme. Pirie (35) found that
lysozyme from egg white released a specific phage from its union with
heat-killed cultures of B. megatherium, a release which occurred during
the same interval of time as hydrolysis of the bacterial polysaccharide
by the lysozyme. Heat-killed bacilli incubated with lysozyme could
no longer absorb phage.

Anderson, in a recent paper (4), reported the separation of a
lytic principle derived by ultraviolet irradiation from a phage of
Escherichia coli. The phage was purified by low- and high-speed
centrifugation. After irradiation with ultraviolet light, the appar-
ently homogeneous and pure phage was split into two components,
one of which was a low molecular weight protein found in the super-
natant fluid after ultracentrifugation. This protein of low molecular
weight caused the lysis of E. coli cells killed by ultraviolet irradiation.
The killed coli cells also could be lysed by crystalline lysozyme from
egg white. The lytic agent obtained from phage, however, could not
replace lysozyme in the lysis of M. lysodeikticus .

Northrop (34) found that suspensions of a highly purified phage
specific for iS". muscae caused the immediate lysis of suspensions of resting
living staphylococci but not of killed organisms. This organism is
killed (and lysed after alkalinization) by egg-white lysozyme (40).
From this staphylococcus, killed by acetone, a fraction was obtained
which not only lysed suspensions of the same species, but also a strain of
Sarcina lutea.

An explanation of these and other experiments may be at-
tempted along the following lines. Phage virus according to this
hypothesis has two components, one a highly specific substance of very
high molecular weight, the other a less specific component of lysozyme-
like nature which causes the actual lysis of the infected bacterial cells.
The lytic agent may be a normal component of the bacterial cell.

What are the functions of lysozyme in the microbial and animal
organism? From a teleological standpoint, lysozyme seems to act in
the animal organism as a protective enzyme against bacterial invasion.
But the most susceptible organisms are harmless saprophytes; patho-
genic organisms show little, if any, susceptibility, although some have
been reported to be susceptible to egg-white lysozyme.

281

KARL MEYER

The protective action of lysozyme, however, is strongly sug-
gested in vitamin A deficiency. Thus Findlay (12) reported a cure
of the xerophthalmic condition in vitamin A deficient rats by washing
the eyes with human tears. Andersen (3) found a subnormal amount
of lysozyme in the tears of human twins suffering from xerophthalmia;
on addition of vitamin A to the diet of the twins, the lysozyme increased
and the xerophthalmia concurrently improved. A renewed study of
nutritional factors in the production and activity of lysozyme seems
indicated.

The occurrence of lysozyme and lysozyme-like enzymes in many
microbes may be interpreted to mean that these enzymes are involved
in some metabolic process connected with the carbohydrate substrates
in the bacterial membranes. Since some of these membranes, as in
M. lysodeikticus, are extremely tough structures, they may serve in
facilitating the softening of the membrane in bacterial division. It
seems useful also to assume a metabolic role for lysozyme in the animal
body. Such a hypothesis would presuppose the occurrence of the
substrate of lysozyme in the animal organism and thus far the ocurrence
of such a substrate has been reported for egg white only (31). A study
of the chemical specificity of lysozyme undoubtedly will be useful in
establishing the presence or absence of such substrates in the animal
body.

The susceptible organisms become highly resistant to lysozyme
action when grown in the presence of sublethal doses of the enzyme.
It will be intei'esting to investigate this adaptation to determine whether
the adapted organism fails to produce the substrate of lysozyme at all,
or whether the adaptation is due to a chemical modification of the
membrane polysaccharide so that it no longer is attacked by lysozyme.

Recently, attention was called to the similarity in occurrence
and chemical properties of lysozyme and avidin, the protein which
neutralizes the vitamin biotin (22). Avidin and lysozyme are both
found in egg white, in the oviduct of birds, and in fish eggs; both are
basic proteins, very stable toward acid and unstable toward oxidation
and alkaline reaction. They undoubtedly are not identical, either
in respect to their chemical or biological properties, the main chemical
difference between the two being that of solubility, for lysozyme is a
very soluble, and avidin is a highly insoluble, protein, even at acid
pH values. They differ in their biological activity in that freshly

282

MUCOLYTIC ENZYMES

prepared, pure lysozyme has a negligible avidin activity. It was found
that biotin (10 7) increased the activity of our lysozyme preparation
(approximately 1 7) up to 500-fold. Lysis at such dilutions of lyso-
zyme had never been observed before. However, the activation was
shown by only a limited number of batches of M. lysodeikticus ; and the
activating effect of biotin was no longer shown by subcultures of the
original batch, while it was still reproducible with old samples. Mean-
while, a failure to activate lysozyme action with biotin was reported
(9). It seems probable that we were dealing with a nonpermanent
mutant in some of the colonies, which was the cause of an observed
autolysis in the affected strain. [Autolytic or "suicide" colonies have
been described as a variant of a Micrococcus tetragenus by Reimann
(38)]. It is hoped that an explanation for the activation will be found
in the near future.

The similarity between lysozyme and avidin seems so striking
that a relationship between the two deserves further investigation.
The formation of avidin, for example, by partial oxidation of lysozyme
is being investigated at present.

Other Mucolytic Enzymes

Many bacteriolytic agents have been encountered in animal
organs and in microorganisms, some of which may be classified as
lysozyme-like enzymes. With increasing knowledge of the muco-
polysaccharides, knowledge of specific mucolytic enzymes will un-
doubtedly increase as well.

Two enzymes have been encountered which may belong in
this group. One is an enzyme — thus far found in human saliva, sub-
maxillary glands of animals, and pneumococci — which hydrolyzes
one of the two mucopolysaccharides derived from the submaxillary
gland. The enzyme is identical neither with hyaluronidase nor with
amylase or lysozyme. A search for it in pathogens of the respiratory
tract might give some useful information about natural resistance to
respiratory infections. The substrate of this enzyme occurs in the
glandular extracts as a mucoprotein, composed of about equal parts
of a peptide chain and of an acid polysaccharide. The latter contains
acetylglucosamine and gluconic acid in equimolar portions (23).

Another mucolytic enzyme group, on which little work has

283

KARL MEYER

been clone, is that concerned with depolyTnerization and hydrolysis
of the neutral mucopolysaccharide fractions of gastric mucin. One
of these substances isolated from pig gastric mucosa is composed of
equimolar parts of acetylglucosamine and galactose (21). In these
mucopolysaccharide fractions some of the blood group substances
occur, substances responsible for the blood group specificities of man
and animal. Although the exact chemical composition of the blood
group substances is unknown at present, there seems little doubt that
the blood group A substance is closely related chemically to the
acetylglucosamine-galactose complexes of pig gastric mucosa.

Enzymes which destroy the blood group A activity have been
found in human tissues, feces, and a few species of bacteria, among
them Clostridium welchii (39). From one of these strains SchifT ob-
tained a filtrate which inactivated the A but not the B substance. In
the writer's laboratory extracts of one strain of C. welchii were prepared
which hydrolyzed the polysaccharide of gastric mucosa with the
liberation of reducing sugar, while from other strains no such enzymes
were obtained (29). The blood group A activity was unaffected by
extracts from the nonhydrolyzing strain, while the hydrolysis by the
active extracts paralleled the A-inactivating potency (unpublished
work of Schiff and Meyer). The wide occurrence of glucosamine-
galactose complexes in many bacterial polysaccharides would make a
study of these enzymes quite interesting.

Hyaliironidase

Hyaluronidase is the most extensively studied of all mucolytic
enzymes. Its main substrate, hyaluronic acid, has been isolated from
vitreous humor, umbilical cord (30), synovial fluid (32), skin (26), and
some tumors of mesodermal origin, such as fowl leucosis (15,36) and a
human mesothelioma (25). One of the most significant findings in
this field was its isolation from group A and G hemolytic streptococci
by Kendall, Heidelberger, and Dawson (17). Hyaluronidase, an
enzyme which depolymerizes and hydrolyzes hyaluronic acid, was
first obtained from a type II pneumococcus (28). Although it was at
first thought that the enzyme was identical with the autolytic enzyme
of pneumococci, this proved not to be the case. Hyaluronidase was
found, shortly afterwards, in a group A hemolytic streptococcus, in a

284

MUCOLYTIC ENZYMES

strain of C. welchii, and in extracts of spleen and ciliary body (29).
Chain and Duthie (6) made the important observation that purified
testis extracts, which contained in liigh concentrations the "spreading
factor" of Duran-Reynals (8) and of McClean (19), also contained
hyaluronidase; in fact, they proposed that hyaluronidase was identical
with the "spreading factor." Based on the findings of Chain and
Duthie, an intensive study was made of the hyaluronic acid-hyal-
uronidase system by a number of investigators.

Hyaluronidase and spreading factor have been found in extracts
of leach heads, in a number of snake and other venoms, in skin, in
staphylococci, and in many other microorganisms. Aside from some
unspecific spreading factors, which may act by oxidatively depolymer-
izing hyaluronic acid, the identity of hyaluronidase with spreading
factor seems well established. The only exception still is the occur-
rence of a strong spreading reaction in extracts or filtrates of some group
A hemolytic streptococci in which no hyaluronidase could be de-
tected (27).

In the spreading reaction, extracts are injected intradermally in
rabbits along with a suitable indicator such as certain dyes, hemo-
globin, bacterial toxins, or India ink. In the presence of spreading
factor, these indicators difTuse over an area of the skin proportional
to the concentration of agent, while in the controls the injected indicator
remains localized in a small bleb. The reaction apparently is caused
by the depolymerization of a hyaluronic acid gel present in the inter-
cellular substance of the dermis. Increased capillary permeability is
not caused by hyaluronidase. Obviously the capillary endothelial
cement does not contain hyaluronic acid. The spreading reaction can
also be demonstrated in the cornea (unpublished work), apparently
acting by the depolymerization of the monosulfuric acid ester of
hyaluronic acid, which occurs in the substantia propria (24).

Hyaluronic acid, a polymer of acetylglucosamine and glucuronic
acid, occurs in different tissues in polydisperse form. Its molecular
weight has been estimated from its double refraction of flow as between
200,000 and 400,000 (5), a figure which may be considerably higher
when measured on material derived from quite solid gels, as from some
malignant cysts or from the nucleus pulposus.

Three methods are available for the quantitative estimation of
hyaluronidase and they depend upon: (a) the hydrolysis of the gluco-

285

KARL MEYER

sidic linkages in hyaluronic acid, as measured by the increase in re-
ducing power; (b) the decrease in viscosity of hyaluronate-containing
solutions; and (c) the decrease of the protein-precipitating power of
hyaluronate after enzymic depolymerization. The last method in a
modification of that of Kass and Seastone (16) has proved most con-
venient and accurate.

Group A and C hemolytic streptococci are the only micro-
organisms in which hyaluronic acid has been demonstrated. It
apparently is formed only when these organisms possess a mucoid
capsule. Seastone has brought forward good evidence for the parallel-
ism of capsule formation, hyaluronic acid production, and the invasive-
ness of the organisms. Incubation with hyaluronidase causes the
disappearance of the capsule. Furthermore, intraperitoneal injection
of hyaluronidase was shown to protect mice against fatal doses of
intraperitoneally injected group A and G streptococci (16).

While some strains produce hyaluronic acid, other strains be-
longing to group A, C, and G produce hyaluronidase. Hyaluronic
acid and hyaluronidase production do not seem to occur simultane-
ously. In a recent study (7) involving 308 strains of group A strepto-
cocci, hyaluronidase production was found only in strains belonging
to type 4 and 22. The group A strain used mosdy in our studies also
belonged to type 4.

When hyaluronidase activity was determined chemically with
culture filtrates or fractions isolated from streptococci, a peculiar
behavior was noted, in that the activity measured viscometrically or
by reduction stopped after an initial reaction. In fact in many tests
with hyaluronidase-producing strains no enzyme could be demon-
strated in the majority of tests. The explanation for the abnormalities
may be found in an enzymic destruction of hyaluronidase by strepto-
coccus extracts.

For the concentration and purification of hyaluronidase, bull or
ram testis is the most convenient source. Thus, far, the enzyme has
never been obtained as a pure protein. Beside inert proteins, testicular
preparations contain an enzyme hydrolyzing chondroitin sulfate into
disaccharide units containing sulfate (27). The hyaluronidase activity
in most preparations runs parallel with that of the enzyme which
sphts chondroitin sulfate, including samples purified in the final step
by electrophoretic separation. The two enzymes are, however, sepa-

286

MUCOLYTIC ENZYMES

rate entities, since the enzyme sijlittincf chondrf)itin sulfate is absent
in pneumococcal hyahnonidase preparations. Furthermore, by
treatment with acetone, the former may be destroyed in testicular
preparations without affecting the hyaluronidase activity. The con-
centration of hyaluronidase in testis apparently is far smaller than that
found in the leech and probably than that found in some snake venoms.

Hyaluronidase in the testis or, more specifically, in spermatozoa
plays an important role in fertilization, as was shown recently (11,20).
The cumulus cells surrounding the ovum are embedded in a jelly
which apparently contains hyaluronic acid. This jelly is liquefied by
hyaluronidase furnished by a sufficiently large number of spermatozoa.
Fertilization then can proceed by a single spermatozoon.

In the skin — probably the largest store of hyaluronidase in the
body — the enzyme, although generally present in an inactive form
(27), may be supposed to regulate the velocity of water and metabolite
exchange by decreasing the viscosity of the intercellular matrix.

Clinically, the role of hyaluronic acid and hyaluronidase is yet
little explored. Hyaluronic acid concentration appears to be greatly
increased in some exudates of the joints. Injection of hyaluronidase
into pathological joints lowered the viscosity of the exudates, without
permanently improving the pathological condition (37). A dis-
appearance of ganglia of the tendon sheaths likewise has been noted on
injection of hyaluronidase. Intraperitoneal hyaluronidase injection
was used in a case of mesenthelioma of the pleura and peritoneum,
to facilitate the removal of a fluid of honeylike consistency: without
hyaluronidase injection, paracentesis was incomplete and very difficult;
after injection of purified testicular hyaluronidase, a fluid of low
viscosity could be completely removed in a short time. The con-
tinued injection of large quantities of hyaluronidase apparently had no
harmful effects. The continued growth of the malignant tumor
finally caused the death of the patient (Stewart and Meyer, unpub-
lished experiments).

Testicular preparations containing hyaluronidase in high
concentrations have been shown to lower the high erythrocyte sedi-
mentation rate when added to the blood of patients with rheumatic
fever and other diseases. This effect, which has all characteristics of
an enzymic reaction, does not seem to be due to a proteolytic action
on any of the plasma proteins. At present, it is doubtful, however,

287

KARL MEYER

whclhcr the action is cliic to the liyaluronidase contained in the prepa-
rations. The relationship between streptococcus infection and rheu-
matic fever clearly deserves an intensive study in which hyaluronic
acid-hyaluronidase production and liyaluronidase inhibition cannot
be overlooked.

In this essay, a few mucopolysaccharides and mucolytic enzymes
have been dealt with, and an attempt has been made to show some of
the interrelationship of these entities mainly with bacteriological, but
also with some physiological and medical, problems. Thus, in lyso-
zyme and its substrate, one is concerned not only with preparative
chemistry, but with various problems of bacteriological variation, of
immunology, of virus research, and of nutrition. A study of hyaluroni-
dase involves many bacteriological problems, the physiology of tissue
permeability and fertilization, and the fields of pathology and medi-
cine. The study of gastric mucin has led to the problem of blood
groups and is intimately connected with various tissue and bacterial
antigens. The finding in pneumococcal extracts of an enzyme which
depolymerizes the acid mucopolysaccharide of submaxillary gland
may suggest the presence of similar enzymes in other microorganisms,
especially of the upper respiratory tract.

Although other mucopolysaccharides occurring in the animal
body and their specific enzymes have not been discussed in this essay,
some of these systems are not less important, as, for example,
chondroitin sulfuric acid, heparin, and the carbohydrate matrix of
amyloid tissue. But because knowledge of the enzyme systems which
hydrolyze these substances is too meager, they may only be mentioned
at this point. It seems to the author that these and other systems de-
serve greater attention by the biochemist than they have hitherto
received.

Addendum

Recently a new and highly accurate method for the estimation of lysozyme
was developed in this laboratory. In this method the depolymerization of the
mucopolysaccharide isolated from M. lysodeikticus is measured viscometrically.
Furthermore, lysozymes, partly in very high concentration, were found in the fresh
and dried latex of some species of Ficus, in Euphorbia, and in papain. The Ficus
lysozyme is chemically distinct from that of egg white. Its potency is significantly
higher than the crystalline lysozyme of egg white.

288

MUCOLYTIC ENZYMES

J References

(1) Abraham, E. P., and Robinson, R., jVatuir, 140, 24 (1937).

(2) Alderton, G., Ward, W. H., and Fevold, H. L., J. Biol. Chem., 157, 43
(1945).

(3) Andersen, O., Acta Paediat., 14, 81 (1932).

(4) Anderson, T. F., J. Cellular Comp. Physiol., 25, 1 (1945).

(5) Blix, G., and Snellman, O., Nature, 153, 587 (1944).

(6) Chain, E., and Duthie, E. S., Nature, 144, 977 (1939); Brit. J. Exptl.
Path., 21, 324 (1940).

(7) Crowley, N., J. Path. Bad., 56, 27 (1944).

(8) Duran-Reynals, F., Bact. Revs., 6, 197 (1942).

(9) Elderton, G., Lewis, J. C, and Fevold, H. L., Science, 101, 151 (1945).

(10) Epstein, L. A., and Chain, E., Brit. J. Expf. Path., 21, 339 (1940).

(11) Fekete, E., and Duran-Reynals, F., Proc. Soc. Exptl. Biol. Med., 52, 119
(1943).

(12) Findlay, G. M., Brit. J. Exptl. Path., 6, 16 (1925).

(13) Fleming, A., Proc. Roy. Soc. London, B93, 306 (1922). Fleming, A., and
Allison, V. D., Lancet, 1, 1303 (1924); Brit. J. Exptl. Path., 8, 214 (1927).

(14) Gratia, A., and Rhodes, B., Compt. rend. soc. biol., 89, 1171 (1923);
90, 640 (1924).

(15) Kabat, E. A., J. Biol. Chem., 130, 143 (1939).

(16) Kass, E. H., and Seastone, G. V., J. Exptl. Med., 79, 319 (1944).

(17) Kendall, F. E., Heidelberger, M., and Dawson, M. H., J. Biol. Chem.,
118,61 (1937).

(18) Kriss, A. E., Microbiology (U.S.S.R.), 9, 32 (1940).

(19) McClean, D., J. Path. Bad., 33, 1045 (1930).

(20) McClean, D., and Rowlands, L W., Nature, 150, 627 (1942).

(21) Meyer, K., Cold Spring Harbor Symposia Quant. Biol., 6, 91 (1938).

(22) Meyer, K., Science, 99, 391 (1944).

(23) Meyer, K., "Mucoids and Glycoproteins," in Advances in Protein Chemis-
try. Vol. II, Academic Press, New York, 1945.

(24) Meyer, K., and Chaffee, E., Am. J. Ophthalmol., 23, 1320 (1940).

(25) Meyer, K., and Chaffee, E., J. Biol. Chem., 133, 83 (1940).

(26) Meyer, K., and Chaffee, E.,J. Biol. Chem., 138, 491 (1941).

(27) Meyer, K., Chaffee, E., Hobby, G. L., and Dawson, M. H., J. Exptl.
Med., 73,309 (1941).

(28) Meyer, K., Dubos, R., and Smyth, E. M., J. Biol. Chem., 118, 71 (1937).

(29) Meyer, K., Hobby, G. L., Chaffee, E., and Dawson, M. H., J. Exptl.
Med., 71, 137 (1940).

(30) Meyer, K., and Palmer, J. VV., J. Biol. Chem., 114, 689 (1936).

289

KARL MEYER

(31) Meyer, K., Palmer, J. W., Thompson, R., and Khorazo, D., J. Biol.
Chem., 113,479 (1936).

(32) Meyer, K., Smyth, E. M., and Dawson, M. H., J. Biol. Chem., 128, 319
(1939).

(33) Meyer, K., Thompson, R., Palmer, J. W., and Khorazo, D., J. Biol.
Chem., 113,303 (1936).

(34) Northrop, J. H., Crystalline Enzymes. Columbia Univ. Press, New York,
1939.

(35) Pirie, A., Brit. J. Exptl. Path., 21, 125 (1940).

(36) Pirie, A., Brit. J. Exptl. Path., 23, 277 (1942).

(37) Ragan, C, and De Lamater, A., Proc. Sac. Exptl. Biol. Med., 50, 349
(1942).

(38) Reimann, H. A., J. Bact., 31, 385 (1936).

(39) Schiff, F., J. Infectious Diseases, 65, 127 (1939).

(40) Thompson, R., Arch. Path., 30, 1096 (1940).

(41) Twort, F. W., Lancet, 2, 642 (1925).

(42) Wollman, E., and WoUman, E., Compt. rend. soc. biol., 110, 636 (1932);
112, 164 (1933).

290

19

SOME ASPECTS

OF INTERMEDIARY

METABOLISM

KONRAD BLOCH, associate in biochemistry, college of

PHYSICIANS AND SURGEONS, COLUMBIA UNIVERSITY

/:

'NTERMEDIARY metabolism is a branch of biochemistry
which attempts to trace the pathways of food components
in the animal body, with the ultimate aim of describing the sequence
of the chemical reactions involved, and of providing a quantitative
account of food utilization. Ideally, the study of normal intermediary
metabolism should be carried out under conditions which do not of
themselves aflfect the physiological state of the animal. In order
to maintain the "normal" state, it is necessary to employ a diet on
which the weight of the animal and the composition of the excretion
products remain substantially constant. Because the animal cell is
unable to distinguish between the various isotopes of an element,
substitution of a normal dietary constituent by one which contains an
excess of an isotopic element, but which is otherwise the same, does not
disturb its physiological state. In order to illustrate the type of in-
formation which can be gained with the aid of isotopically labeled
substances, it will be useful to consider briefly the scope and the limita-
tions of other methods which have been applied to the study of inter-
mediary metabolic reactions.

Chemical analysis of animal tissues can, of necessity, contribute
only limited information about intermediary reactions. It has been

291

KONRAD BLOCH

recognized that substances which are suspected of being intermediates
of either anabolic or catabolic reactions, but which are present in
amounts too small for chemical identification, may acquire quantita-
tive significance in the light of the rate at which they are regenerated.
Although their stationary concentration may be small, they arise and
are metabolized in large quantities over an extended period of time.
In the breakdown of the major food constituents, certain interme-
diates are known to be produced continuously, e. g., the deamination
products of amino acids, ornithine, the phosphorylated trioses, and
acetic acid. None of these intermediates has actually been isolated
as such from animal tissues. Hence, numerous attempts have been
made to arrest metabolism at an intermediate stage by inducing an
unphysiological state (diabetes, starvation, etc.) or by administration
of substances which contain groupings refractory to attack by animal
cells. The latter approach, of which the phenyl substituted amino
acids and fatty acids employed by Knoop are the most notable
examples, represents an early attempt to tag molecules in order to trace
them through the animal body.

A large part of the accepted knowledge of intermediary reac-
tions, particularly of carbohydrates and amino acids, has been gained
from studies with surviving tissues, tissue preparations, and isolated
enzymes. The in vitro techniques deal with reactions which the
enzyme systems of cells are able to carry out with an added substrate.
The result, although not the primary aim, of this type of experimenta-
tion, has been to provide knowledge of the organic chemistry of bio-
logical compounds in a "physiological" medium and under the in-
fluence of biocatalysts. The use of isolated tissues involves inherent
limitations as far as permitting a decision as to whether certain re-
actions occur in the living organism. The question as to whether
and to what extent an in vitro reaction is part of the normal metabolism
must be decided by an independent method. The fact that, for ex-
ample, surviving liver and kidney, or enzymes isolated from these
tissues, deaminate ^-amino acids at a rapid rate in no way signifies
that the cell normally has to deal with the unnatural amino acids.
Evidence that the amino acids of ^^-configuration are not normal con-
stituents of the animal cell is contained in the following experiment
(20). When fl',/-tyrosine or fl',/-glutamic acid labeled with N^^ was
administered to rats, the unnatural acids excreted in the urine had an

292

INTERMEDIARY METABOLISM

unchanged isotope concentration; if ^-tyrosine oi- (-/-glutamic acid
liad been formed in a normal synthetic process, it should have mixed
with the administered isotopic compound, and hence the excreted
amino acid should have had a lower isotope concentration.

As a great variety of foreign substances, e. ff., the odd-numbered
fatty acids, are not only modified but completely metabolized by the
animal cell, it must be inferred that the organism is al)le to handle
substances "to which it is ordinarily accustomed" as well as many
others which do or do not bear a structural relationship to natu-
rally occurring compounds. Conversely, compounds of established
biological activity, such as the lipids, are found to be unaffected by
isolated tissues. In vitro, the higher fatty acids, cholesterol, and other
steroids fail to show any evidence of synthetic or degradative reactions
involving carbon-carbon bonds. Limitations of solubility and cell
permeability must, to some extent, impair in vitro activity toward
lipids and may be responsible for their apparent metabolic inertia.

In view of our recent experiences with isotopes, another aspect
must be considered for the interpretation of these findings. The studies
of Schoenheimer and Rittenberg with isotopic test substances have
made it clear that the constancy of tissue composition in the normal
adult animal is maintained as a result of balanced synthesis and
degradation, and have provided an experimental basis for the concept
of dynamic equilibrium of the tissue constituents. It is conceivable
that a similar situation would obtain in isolated tissues so that a reaction
proceeds without over-all change of the substrate concentration. Evi-
dence that the dynamic equilibrium may obscure the recognition that
an in vitro reaction is proceeding may be seen in the following findings.
The utilization of acetic acid for cholesterol synthesis, first demon-
strated in intact animals with the aid of isotopic acetate (4), has re-
cently been observed to occur also in isolated tissue (unpublished re-
sults of Bloch, Borek, and Rittenberg). When liver slices were in-
cubated with deuterioacetic acid, deuteriocholesterol was formed at a
rate similar to that in the liver of intact animals. As the total amount
of cholesterol does not change noticeably under these conditions, it
is clear that a quantity equivalent to that synthesized had simultane-
ously been degraded. The sensitivity of the isotope method is such
that syntheses corresponding to as little as 0.1% of the cholesterol
present can be detected. Experiments in which liver slices were in-

293

KONRAD BLOCH

cubaled in the presence of DoO indicate that the deuteriocholesterol
formed amounted to at least 10% of the cholesterol present at the
start, an amount of change which, although close to the limits of
detection by available analytical methods, nevertheless should be
demonstrable.

The available data on the syntheses in which acetic acid is
employed indicate that, in the intact animal, fatty acids are regener-
ated at a faster rate than cholesterol (17). On the other hand, the
corresponding results obtained with liver slices would lead to the
erroneous conclusion that cholesterol synthesis is by far the more rapid
process. The lack of correlation between the rates in vitro and in vivo
must be taken to mean that, in the experiments in which acetic acid
was added to liver slices, the conditions chosen happened to be favor-
able for cholesterol synthesis, but that a component necessary for fatty
acid formation was lacking.

The use in tissue slice experiments of enzyme poisons, such as
malonate, fluoride, etc., makes possible an accumulation of inter-
mediates whose identification is thereby facilitated. If specific in-
hibitors for a particular reaction are not available, a suspected inter-
mediate may escape detection because it does not accumulate. The
isotope dilution method has recently been applied to establish the oc-
currence of a reaction which is not detectable by the balance method
(14). Incubation of kidney slices with acetic acid does not lead to a
demonstrable formation of acetoacetic acid. If acetic acid containing
heavy carbon is substituted for ordinary acetic acid and some non-
isotopic acetoacetic acid is added, subsequent analysis of the aceto-
acetate then isolated reveals the presence of heavy carbon in the keto
acid. Thus, clearly, acetoacetic acid is synthesized from acetic acid,
but its rapid removal by subsequent reactions prevents its accumulation
in quantities sufficient for analysis. In experiments of this type, only
the isotope concentration and not the absolute amount of acetoacetate
is relevant; the rate of acetoacetate synthesis can be calculated from
the amount of acetoacetate added initially and from its isotope concen-
tration at the end of the experiment. It is evident that many of the
difficulties pertaining to the use of tissue preparations can be circum-
vented by the use of pure enzymes.

In devising schemes for metabolic interconversions, the bio-
chemist often rests his case on little more than similarities of chemical

294

INTERMEDIARY METABOLISM

Structure between various substances. In spite of close chemical re-
semblances, the biological relationshi|)s of creatine to creatinine, of
phenylalanine to tyrosine, and of cholesterol to the Ijile acids and
steroid hormones have, until recently, remained hypothetical. The
balance type of experimentation with intact animals has, for various
reasons, failed to give conclusive answers to these problems. Even if
an oversupply in the diet of a suspected precursor leads to an increased
production of the correct reaction product, the conversion will remain
in doubt until the possibility of an indirect stimulating eflfect is excluded.
Balance experiments may be inconclusive because many biological
syntheses appear to proceed at rates which are, within limits, inde-
pendent of dietary variations. In such cases, the oversupply of a pre-
cursor will be without effect on the formation of the reaction product.
For instance, the immediate response to addition of excess creatine to
the diet is an increased excretion of creatine rather than of creatinine
(7). The conversion of creatine to creatinine becomes evident, how-
ever, by employing labeled creatine. When small amounts of creatine
containing N^'^ are fed to rats, the creatinine excreted contains isotopic
nitrogen (6). In this case, it is irrelevant whether or not creatinine is
increased in total amount; the presence of isotope is proof of the con-
version.

A biochemical conversion may proceed independently of the
exogenous supply of the conversion product or of both conversion
product and precursor. When an essential dietary constituent is the
precursor, its concentration in the diet may regulate the extent of the
conversion, but it may be immaterial whether or not the conversion
product is supplied. Conversion of phenylalanine into tyrosine and
the methyl transfer from methionine in the synthesis of choline proceed
normally even when these products of conversion are amply provided
in the diet (15,26).

If the precursor can be synthesized by the organism at an ade-
quate rate, its addition to the diet will likewise fail to increase produc-
tion of the conversion product. Interrelationships established with
the aid of labeled test substances (18,21,24), e. g., those existing between
the dispensable amino acids, proline, ornithine, and glutamic acid,
or between palmitic and stearic acids (19,22), could not be clarified
by quantitative methods because these reactions proceed independently
of the exogenous su{)ply of both precursor and reaction product.

295

KONRAD BLOCK

Recognition of these reactions which are independent of the nutritional
state of the animal illustrates the very limitations to which balance
experimentation in the intact animal is subject.

In considering the in vivo relationships of steroids, it is important
to bear in mind that the body possesses a store of cholesterol in every
cell. If cholesterol were the parent substance for all other steroids
(bile acids and steroid hormones), the precursor would be available in
amounts which are large in comparison with the quantities of bile acids
and hormones normally produced. Increase of the available choles-
terol supply by dietary addition can, therefore, be expected to affect
these conversions but little. With labeled cholesterol, the role of this
sterol as a precursor of bile acids and of the excretion form (pregnane-
diol) of at least one steroid hormone (progesterone) could be verified
(2,3). For the interpretation of the experiment, the quantities both
of precursor administered and of conversion product isolated are
immaterial. Appearance of the isotopic label in the reaction product
is sufficient proof of the occurrence of the postulated reaction, provided
that the uptake of isotope is not due to an unspecific reaction. When
deuterium is the isotope employed, the possibility that the reaction
product received its isotope by unspecific reactions such as exchange
with the body fluids must be excluded. The quantitative importance
of the conversion can be evaluated from the relative concentrations of
isotope in the circulating precursor and the product formed.

Degradation of Fatty Acids

The apparent inability of in vitro systems to synthesize or degrade
lipids may well be a consequence of the fact that the metabolism of
lipids requires the presence of intact cells. The enzyme systems
involved have remained entirely unknown, and the study of inter-
mediary fat metabolism has therefore been carried out primarily with
intact animals. Stable isotopes found their first biological application
in this field and piomise to yield further information particularly when
carbon C^^ becomes readily available. The recent observation that
lipid synthesis can be demonstrated to occur in surviving liver if marked
test substances are employed may facilitate a more detailed study of
reaction mechanisms which are not profitably investigated by experi-
mentation with intact animals.

296

INTERMEDIARY METABOLISM

The view that fatty acids arc metaboHzed by oxidation at their
j3-carbon atoms with consequent removal of two-carbon units has
been held ever since the classical experiments with phenyl substituted
fatty acids were carried out by Knoop. The experiments of Dakin
with the ammonium salts of fatty acids brought evidence for the chemi-
cal susceptibility of the beta positions to mild oxidizing agents in the
test tube. The facts that (1) the component fatty acids of animal
tissues contain, without exception, an even number of carbon atoms,
and (2) that acetoacetic acid is excreted under conditions of deficient
or impaired carbohydrate oxidation, have provided additional though
indirect support for the principle of beta oxidation. To this theory,
the objection has been raised that none of the postulated intermediary
keto acids nor any of the fatty acids below Cio, including the main
breakdown product, acetic acid, has been demonstrated in animal
tissues. In view of recent experience, the practical absence of such
intermediates is not surprising, since the failure to accumulate is
characteristic of the breakdown products of the major food constitu-
ents and must be attributed to their rapid rate of removal.

A finding apparently at variance with the beta oxidation theory
was the observation that, in isolated tissues, the uptake of oxygen and
the formation of ketone bodies are greater with the higher fatty acids
than with butyric acid (9), whereas Knoop's theory provided for only
one mole of acetoacetate per mole of fatty acid. An attempt to over-
come this inconsistency led to the concept of multiple alternate oxida-
tion (8) which visualizes simultaneous oxidation of the fatty acid chain
at alternate carbon atoms. In this manner palmitic acid could break
down to yield four molecules of acetoacetate. It is noteworthy that,
for years, little attention had been paid to the observation of Loeb (12)
that ketone bodies were formed on perfusion of liver with acetic acid.
If acetoacetate could arise synthetically from smaller units as well as
by primary breakdown, then the formation of more than one mole of
ketone bodies per molecule of fatty acid would no longer be incom-
patible with the theory of beta oxidation.

Another fact which could not be fitted into existing theories
was the finding that w-valeric acid is a ketogenic substance (9,13).
Neither Knoop's theory nor that of multiple alternate oxidation
could satisfactorily explain this observation unless it was supposed
that oxidation of odd-numbered fatty acids could be initiated at the

KONRAU BLOCII

a-carbon atom. McKay pcrspicaciously suggested that valeric acid
and other odd-numbered fatty acids yield, on oxidation, /3-keto acids
which then undergo hydrolysis to acetic acid. Condensation of two
acetic units would then give rise to acetoacetate.

By employing fatty acids labeled with deuterium and C^', it
has been possible to provide more direct support for the principle of
beta oxidation and to establish that acetic acid is the source of the
"extra" ketone bodies. The first experiment bearing on the question
of degradation by elimination of C2 units was carried out by Schoen-
heimer and Rittenberg (19) who demonstrated the biological conver-
sion of deuteriostearic acid to deuteriopalmitic acid. It was clear
from their data that this transformation was direct and not attributable
to utilization of smaller fragments derived from stearic acid. Ana-
lytical data for the myristic and lauric acid fractions indicated that the
process of C2 removal did not end at the Cie stage. It is noteworthy
that the fatty acids with carbon chains longer than Cio are constituents
of tissue fat but appear to be metabolically inert when added to tissue
slices or tissue extracts, whereas the lower acids from C2 to Cg are
oxidized in vitro but cannot be detected in tissue fat. Evidently the
rate of oxidative breakdown increases with decreasing chain length,
both in vitro and in vivo. Actually, the lower fatty acids may arise in
considerable quantities in the animal body over an extended period of
time, but their concentration in the stationary state is insignificant be-
cause they are rapidly oxidized further. Evidence supporting this
concept has been secured in the case of acetic acid. This acid should
be the principal intermediate which is formed by hydrolytic splitting
of the ^-keto acids in the course of beta oxidation. Although isolation
of this compound might be achieved by working up large quantities
of animal tissues or urine, its normal concentration in tissues and body
fluids is apparently too small to permit ready identification.

An investigation with the aid of deuterioacetic acid of the way
in which foreign amines are acetylated revealed that acetic acid is
utilized directly in the acetylation reaction (1,5). The excreted acetyl
compound contains only a fraction of the isotope present in the dietary
acetic acid; dilution of the isotope by acetic acid arising in inter-
mediary metabolism must have occurred in the tissues. If the de-
crease in isotope content occurring in the acetylation reaction could be
ascribed exclusively to the presence of endogenous acetic acid, then the

298

INTERMEDIARY METABOLISM

dilutions would be a measure of the amounts of acetic acid formed by
the intact animal. This assumj)tion proved to be valid for the acetyla-
tion of the aromatic amines, sulfanilamide and //-aminobenzoic acid
(unpublished results of Bloch and Rittenberg). The procedure em-
ployed represents an application of the isotope dilution technique to
in vivo systems. Dietary deuterioacetate and noni.sotopic endogenous
acetate merge, and a representative sample of the mixture becomes
fixed in a reaction product which resists attack by the animal cell and
is excreted in the urine. From the isotope concentration of the excreted
acetyl amino compound it can be estimated that, in 24 hours, the adult
rat produces a quantity of acetic acid equal to about one per cent of
its body weight. On the other hand, one can calculate from data on
fatty acid turnover (17) the amount of acetic acid which should arise
in the same period of time on the assumption that one molecule of
palmitic acid is broken down to eight molecules of acetate. The value
calculated on this basis is in close agreement with that found experi-
mentally by the isotope dilution method. These results suggest that,
during oxidation, most, if not all, carbon atoms of the fatty acid pass
through the acetic acid stage.

By taking advantage of the fact that, in the acetylation of
foreign amines a sample of endogenous acetate is trapped and elimi-
nated in the form of a metabolically inert derivative, it becomes pos-
sible to demonstrate acetic acid formation from individual fatty acids.
Appearance of deuterioacetyl groups after administration of a labeled
fatty acid will be indicative of intermediary acetic acid formation. In
accord with the beta oxidation theory, labeled myristic, ^-valeric,
isovaleric, and butyric acids were found to result in the excretion of
isotopic acetyl groups, while propionic and isobutyric acids failed to
do so (5). Extension of these experiments to other fatty acids which
are labeled at specific positions with deuterium as well as isotopic
carbon seems desirable.

As some of the intermediates in fatty acid catabolism contain
labile hydrogen which is subject to enolization and resultant exchange
with the hydrogen of the body fluids, deuterium may be partly lost.
Quantitative evaluation of the data on acetic acid formation from the
higher fatty acids may, therefore, be difficult and must await the
preparation of test substances containing isotopic carbon. On the
other hand, the lability of G — H bonds may often be of advantage by

299

KONRAD BLOCH

providins; information on reaction mechanisms. For instance, l)utyric
acid labeled with deuterium at either the a- or tlie 7-position will
yield deuterioacetyl groups, indicating fission of the molecule int(} two
C2 fragments. Since a-deuteriobutyrate yields an acetyl group of much
lower isotope concentration than does 7-deuteriobutyrate, it can be
inferred that the intermediate must be of a nature to permit loss of
deuterium at the a-carbon atom. In view of the known in vitro labil-
ity of the a-hydrogen atoms in /3-keto acids, acetoacetic acid appears
to be the most probable intermediate in the butyrate-acetyl con-
version.

Formation of Ketone Bodies

Theories of fat oxidation have been based, at least in part, on
the ketogenic action of certain substances and the ability of others to
suppress ketosis. The mechanism of ketone body formation long
remained controversial because of the "extra" ketone body formation
from even-numbered fatty acids which contain more than four carbon
atoms. On administering to fasting rats acetic acid which contained
a carboxyl group labeled with heavy carbon, Swendseid et al. (25)
observed the presence of C^^ in the excreted ketone bodies. Their
data provide unequivocal proof for the early contention of Loeb that
acetoacetic acid could be synthesized from smaller molecules. It
still remained to be decided whether this process involved a Claisen
type of condensation of acetic acid or a coupling with one molecule
of pyruvic acid to acetopyruvic acid with subsequent decarboxylation
(10). Conclusive evidence pertaining to the mechanism is supplied
by recent experiments of Weinhouse et al. (27). Octanoic acid labeled
with C^^ in the carboxyl group was incubated with liver slices. Isotope
analysis of the resulting acetoacetate revealed the presence of heavy
carbon in equal concentrations in the carboxyl and carbonyl groups,
a finding which must be ascribed to random condensation of the acetic
acid units arising from the oxidation of octanoic acid. The isotope
distribution observed can be reconciled neither with the classical view
that only the four terminal carbon atoms of a fatty acid are a source of
ketone bodies nor with the hypothesis of multiple oxidation at alternate
carbon atoms. Consequently, any metabolite capable of forming

300

INTERMEDIARY METABOLISM

acetic acid must be considered a potential source of ketone bodies. *
The ketogenic action of «-valcric acid, which has also been shown to
yield acetic acid in the normal animal (5), can be readily explained as
resulting from the condensation of two molecules of acetate.

In liver tissue in vitro, the oxidation of fatty acids appears to lead
to acetoacetic rather than to acetic acid as an end product. Since
such experiments are performed with slices from fasted animals, the
accumulation of ketone bodies may well be related to the lack of
carbohydrate or carbohydrate intermediates, conditions which, in the
intact animal, result in ketosis. It would be unwarranted to conclude
that in the normal, well-nourished animal the acetic acid formed by
fat breakdown necessarily condenses to acetoacetate prior to its further
oxidation. It is equally conceivable that normally fat oxidation takes
the course visualized by Knoop and that the acetate-acetoacetate
condensation goes into effect only when the catabolism of acetic acid
is interfered with.

Fatty Acid Synthesis

It is well recognized that fat can be synthesized from carbo-
hydrate, but information as to the chemical nature of the process is
totally lacking. Any proposed mechanism must be in accord with
the fact that the component fatty acids of tissue fat without exception
have an even number of carbon atoms and comprise all members of
the series up to C20 and higher. The simplest general scheme for the
biosynthesis of all fatty acids would be one involving building units
containing two carbon atoms. Proof of C2 condensation has been
established in at least one case by the finding that deuteriopalmitic
acid is converted to deuteriostearic acid by the rat (22). Since
deuteriostearic acid is also degraded biologically to deuteriopalmitic
acid, the reversible removal and addition of C2 units is clearly a normal
event. As to the nature of the fatty acid derivative undergoing the
chain elongation, the natural occurrence of fatty acid aldehydes

* In this connection, the question a.s to whether pyruvic acid can be con-
verted to acetic acid in vivo needs further investigation. From our findings witii
deuterioalanine, which we used as a source of deuteriopyruvic acid, it appears that
pyruvate-acetate conversion in the liver constitutes only a minor pathway for
pyruvate metabolism.

301

KONRAD BLOCK

and the fact that palmitic acid is in biological equilibrium with cetyl
alcohol (23), may be of significance.

In testing experimentally the hypothesis of C2 condensation, the
choice among compounds containing two carbon atoms is rather
limited. Except in the case of the amino acid glycine, the existence of
C2 compounds in animal tissues has never been substantiated by
isolation. Acetic acid suggests itself as a likely intermediate if only for
the reason that it arises as a product of fatty acid degradation. When
acetic acid containing deuterium in sufficient concentration to induce
detectable deuteriocholesterol formation was administered to rats, no
significant amounts of isotope appeared in the fatty acids (4). But
when the experiment was repeated with an acetate preparation contain-
ing very high concentrations of deuterium as well as a carboxyl group
labeled with C^^, the fatty acids of liver and depot were found to
contain both isotopic carbon and deuterium (16). The incorpora-
tion of both isotopes proved that the acetate molecule had been utilized
as such. However, the ratio of the concentrations of the two isotopes
in the fatty acids synthesized differed considerably from that in the fed
material, indicating that the conversion involved a loss of carbon-
bound hydrogen. Lability of the hydrogen atoms in the methyl group
of acetic acid cannot be responsible, since the ratio of the two isotopes
remains unchanged when acetic acid is employed in the acetylation of
foreign amines (unpublished results of Bloch and Rittenberg). It
follows that the intermediates in the fatty acid synthesis are of such a
nature as to permit loss of carbon-bound deuterium by exchanging it
with the hydrogen of the body fluids. An analogous conclusion was
necessary with respect to the intermediates in fatty acid oxidation.
The failure to detect fatty acid synthesis from acetate labeled by
deuterium alone and in relatively low concentrations, emphasizes the
caution which is necessary in the interpretation of negative data.
Use of a test compound containing the stable isotope of carbon as well
as deuterium not only eliminates the uncertainty arising from hydrogen
lability but also permits certain deductions from the change of the
D:Ci3 ratio.

The mechanism of condensation leading to the structure of
cholesterol must differ in at least some respects from that in\ olved in
fatty acid synthesis. Owing to the complexity of the chemical struc-
ture of cholesterol the nature of the condensation reactions has re-

302

INTERMEniARY METABOLISM

mained entirely olwcure, apart from the findinsr that the cholesterol
side chain, as well as the muleus, was derived from acetate. In the
case of the fatty acid, a wider variety of degradative procedures are
feasible, so that the isotope concentrations at specific positions of the
molecule can be ascertained. The acetic acid employed in these
experiments contained the carbon isotope only at the carboxyl carbon.
If the fatty acids were formed by successive condensation of C2 units,
only the odd-numbered carbon atoms would contain C'^. The car-
boxyl group obtained by thermal degradation of the fatty acid was
indeed about twice as rich in isotope as the total molecule. This
finding eliminates, as the principal synthetic reaction, the addition of
one C2 unit to either the a- or co-carbon atom of the preformed fatty
acid chain. The position of the isotope in the fatty acid chain can
be traced further by oxidative cleavage of the "oleic acid" fraction into
pelargonic acid, derived from carbon atoms 10 to 18, and azelaic acid
from carbon atoms 1 to 9, of the fatty acid. Both moieties contain
deuterium and C^^ to about the same extents. It therefore seems
probable that the C2 units are used imiformly for the various parts of
the fatty acid chain. If the conden.sation mechanism involved were
of the acetaldehyde-aldol type, in analogy to the in vitro synthesis of
stearic acid carried out by Kuhn et al. (11), a biological reduction of
acetic acid to acetaldehyde would be involved, and this is so far without
experimental basis. A Claisen condensation with acetic acid or an
acetic acid derivative is a more attractive possibility because the
simplest reaction of this type, the condensation to acetoacetate, has
definitely been shown to occur. The /3-keto acids which would have
to be postulated as intermediates, should contain readily enolizable
hydrogen at the alpha position. The fact that deuterium is lost in the
conversion of acetate to fatty acid is in accord with the postulated
scheme, which may be viewed as a reversal of the steps responsible for
fatty acid degradation. The fact that acetic acid is involved in the
synthesis as well as in the breakdown of fatty acids, and the revers-
ibility of two partial steps, viz., the acetate-acetoacetate reaction and
the palmitic acid-stearic acid conversion, are in favor of this scheme.
The hypothesis is not out of line with general biochemical experience,
if one considers the reversibility of individual reactions in the metabo-
lism of proteins and carbohydrates.

The role of acetic acid as a building unit for fat suggests that

KONRAD BLOCH

one of the [)atliways by which carbohydrates are converted to fatty
acids involves acetic acid. The sohition of this problem will necessitate
further clarification of the oxidative phases of carbohydrate metabo-
lism, and specifically that of pyruvic acid.

References

(1) Bernhard, K., <. physiol. Chem., 267, 91 (1940).

(2) Bloch, K., J. Biol. Chem., 157, 661 (1945).

(3) Bloch, K., Berg, B., and Rittenberg, D., J. Biol. Chem., 149, 511 (1943).

(4) Bloch, K., and Rittenberg, D., J. Biol. Chem., 143, 297 (1942).

(5) Bloch, K., and Rittenberg, D., J. Biol. Chem., 155, 243 (1944).

(6) Bloch, K., and Schoenheimer, R., J. Biol. Chem., 131,^111 (1939),

(7) Folin, O., in Festschrijt JiXr 0. Hammarsten, Uppsala Univ. Arsskr., 1906,
III, 1.

(8) Hurtley, W. H., Quart. J. Med., 9, 301 (1915).

(9) Jowett, M., and Quastel, J. H., Biochem. J., 29, 2159 (1935).

(10) Krebs, H. A., and Johnson, W. A., Biochem. J., 31, 772 (1937).

(11) Kuhn, R., Grundmann, C, and Trischmann, H., Z- physiol. Chem.,
248, 4 (1937).

(12) Loeb, A., Biochem. Z-, 47, 118 (1912).

(13) MacKay, E. M., Wick, A. N., and Barnum, C. P., J. Biol. Chem.,
136, 503 (1940).

(14) Medes, G., Weinhousc, S., and Floyd, N. F., J. Biol. Chem., 157, 752
(1945).

(15) Moss, A. R., and Schoenheimer, R., J. Biol. Chem., 135, 415 (1940).

(16) Rittenberg, D., and Bloch, K., J. Biol. Chem., 154, 311 (1944).

(17) Rittenberg, D., and Schoenheimer, R., J. Biol. Chem., 121, 235 (1937).

(18) RolofT, M., Ratner, S., and Schoenheimer, R., J. Biol. Chem., 136,
561 (1940).

(19) Schoenheimer, R., and Rittenberg, D., J. Biol. Chem., 120, 155 (1937).

(20) Shemin, D., and Rittenberg, D., J. Biol. Chem., 151, 507 (1943).

(21) Shemin, D., and Rittenberg, D., J. Biol. Chem., 158, 71 (1945).

(22) Stetten, D., Jr., and Schoenheimer, R., J. Biol. Chem., 133, 329 (1940).

(23) Stetten, D., Jr., and Schoenheimer, R., J. Biol. Chem., 133, 347 (1940).

(24) Stetten, M. R., and Schoenheimer, R., J. Biol. Chem., 153, 113 (1944).

(25) Swendseid, M. E., Barnes, R. H., HemingAvay, A., and Nier, A. O.,
J. Biol. Chem., 142, 47 (1942).

(26) du Vigneaud, V., Cohn, M., Chandler, J. P., Schenck, J. R., and
Simmonds, S., J. Biol. Chem., 140, 625 (1941).

(27) Weinhouse, S., Medes, G., and Floyd, N. F., J. Biol. Chem., 155, 143
(1944).

20

THE STEROID HORMONES

GREGORY PINCUS, visiting professor of experimental biology,

CLARK university; director of laboratories, the WORCESTER
foundation FOR EXPERIMENTAL BIOLOGY

THE MODERN biochemistry of the steroid hormones had
its inception in 1929 with the isolation of estrone from
human pregnancy urine by Doisy and by Butenandt. The previous
findings of Fellner, and of Doisy and Allen, that active ovarian mate-
rial was probably lipid in nature, and the discovery by Aschheim and
Zondek of larger amounts of folliculoid activity in lipid fractions of
human pregnancy urine paved the way for the simultaneous chemical
identifications of estrone in the United States and Germany. In the
fifteen years that followed, the organic chemistry of the steroid hor-
mones has developed explosively. Naturally occurring and synthetic
substances having testoid, folliculoid, luteoid, and corticoid effects have
been obtained in profusion. *

On the basis of the effects of these compounds, new techniques
in clinical medicine have developed. The nature of sex determina-
tion, pubertal growth and sexual involution have been delineated
experimentally with the aid of the steroid sex hormones. The avail-
ability of active corticoid steroid substances has led to a notable
elucidation of the hormonal control of electrolyte and water balances,

* The terminology suggested by Selye (10) will be used throughout this
chapter. Testoid is synonymous with androgenic, folliculoid with estrogenic or
gynecogenic, luteoid with progestational or luteogenic, and corticoid with adreno-
corticoid.

GREGORY PINCUS

as well as of certain phases of carbohydrate metabolism in vivo. The
remarkable effects of steroid hormones on pituitary activity have led
to new conceptions of hormonal balance. The combination of bril-
liant chemical investigation and the astute application of chemical
discoveries to physiological investigation has been so prolific that the
journals are crowded with papers in the field; and in fact new journals
have been founded specifically for the publication of data in endocri-
nology.

Nonetheless, on the biochemical side there have been two
notable gaps in the development of steroid hormone investigation:

(1) the nature of steroid hormone anabolism and catabolism and

(2) the role of the steroid hormones in cellular processes.

Biosynthesis of Steroid Hormones

Concerning the precursors and synthesis of the steroid hormones,
our factual basis is slight. Several authors have indicated possible
modes of degradation of cholesterol in vivo to active hormonal sub-
stance (3,6), and the synthesis of testosterone and progesterone is
indeed accomplished commercially in a series of reactions the first step
of which is the oxidation of the Cn side chain of cholesterol followed
by molecular rearrangement. Experimental verification of such
hypotheses was practically completely lacking until Bloch's recent
demonstration that deuterium-containing pregnanediol is found in
the urine of pregnancy after the administration of cholesterol contain-
ing heavy hydrogen. The recovery of marked pregnanediol was of
the order expected on the assumption that it arose from progesterone,
which in turn was synthesized from the cholesterol administered. The
nature of this cholesterol degradation and the probable intermediates
are not known, although Pearlman's isolation of a series of interesting
neutral steroids from bile suggests that a search in bile for the inter-
mediates may be profitable. Also suggestive is the evidence of Long
and his collaborators that cholesterol in the adrenal glands declines
markedly under conditions involving active secretion of corticosteroid.
It must still be demonstrated to what extent cholesterol is the precursor
of any or all steroid hormones. The evidence that cholesterol itself
is synthesized from simple precursors such as acetic acid renders un-
tenable the numerous theories concerning elementary steroid syn-

306

STEROID HORMONES

thesis, but provides a new approach to the investigation of steroid
synthesis.

It is customary to think of three principal glands as the endog-
enous steroid hormone factories; these are the testes, the ovaries,
and ihe adrenal cortex. There are numerous demonstrations of
folliculoid material in animal fluids, especially in liver and bile.
Estriol, estrone, and estradiol have been isolated from human placenta
and there have been indications of the presence of progesterone in
this organ; but the placenta as a synthesizing organ may be con-
sidered as a rather special case. In brief, the presence of notable
hormonal activity in certain tissues cannot be taken as prima Jacie
evidence that the hormones are produced in these tissues. For this
reason our concepts even of the sites of steroid hormone synthesis are
in a state of flux.

The female in full reproductive vigor leads a double life; she
is alternately mate and mother. According to current concepts, two
ovarian hormones are primarily responsible for this rhythm of func-
tion, estradiol and progesterone. Estradiol, the hormone of the

CH:,

OH

c=o

O

H3C

H3C

HsC

/\

/^

/\

H3C

/\

HO-

O-

/X

Estradiol

Progesterone

HO-

Estrone

ovarian follicle, is considered responsible for the estrous phenomena
that culminate in mating, and progesterone, the hormone of the corpus
luteum, is responsible for the maintenance of pregnancy. Doisy and
his collaborators, in 1936, isolated and chemically identified estradiol
from sow ovarian tissue; and in 1934 progesterone from luteal tissue
was chemically identified by four separate groups of investigators.
The uniqueness of progesterone as the ovarian luteoid is not open to
question, but the evidence is fairly clear that another folliculoid,
estrone, also arises from ovarian tissue. Westerfeld and collaborators
found it to occur in sow ovarian tissue in an amount nearly equal to
that of estradiol, and Dr. J. Schiller and the writer have recendy

307

GREGORY PINCUS

obtained evidence for its production by human ovaries perfused in vitro.
Estradiol is oxidized to estrone by various tissues and estrone is re-
duced to estradiol. Which of the two is formed first in the ovary
cannot therefore be determined a priori. We need to know the pre-
cursor of both to solve the problem.

The isolation of estrone (but not estradiol) from adrenal tissue
suggests that it may be supplied by the adrenal gland to the ovary for
conversion to estradiol. Production of estrone by the adrenal gland
must be in minor amount since the excision of the ovaries in adult
animals effectively abolishes the estrous cycle. Furthermore, adrenal-
ectomy in animals does not abolish estrous cycles, though irregularities
do occur. Finally, Samuels and collaborators have demonstrated that
women with Addison's disease (in which the adrenal glands are atro-
phied or destroyed) excrete normal amounts of folliculoids. It is clear,
therefore, that the adrenal glands are not essential for the produc-
tion of folliculoids. It is interesting to note, however, that very early
ovariectomy in mice leads to the eventual establishment of estrous
cycles and the excretion of notable amounts of folliculoid. Also, the
injection of pituitary adrenocorticotropic hormone into ovariectomized
rats stimulates folliculoid production by the adrenals. The most
likely conclusion is that the capacity for ovarian hormone synthesis is
present in adrenal tissue, but that optimal conditions are found in
normal ovarian tissue. This remarkable steroidogenic capacity of
the adrenal cortex is also confirmed by the isolation in small amounts
of progesterone from adrenal extracts.

CH3 CH3 CHg

c=o

HO— J

HgC

c=o

COH

H3C

VV
(I)

A'- Pregnenolone

/X

(II)

Progesterone

HO-

(III)

Pregnanediol

The immediate precursor of progesterone may be A^-pregneno-
lone, for we have recovered pregnanediol from the urine of men re-
ceiving pregnenolone by mouth or by injection. The transformation

308

STEROID HORMONES

would appear to be from I — > II -^ III. The conversion of I — > II
has been accomplished by bacteria. It is interesting that A^-prcg-
nenolone has been obtained as a naturally occurring substance thus
far only from testis tissue; yet progesterone is a female hormone par
excellence. Patient search for pregnenolone in female tissues appears
desirable.

Among the naturally occurring 19-carbon steroids, A^-dehydro-
isoandrosterone (formula IV) corresponds to A*-pregnenolone. We
would expect it, therefore, to be the immediate precursor of testos-
terone (V) (the primary testis hormone). That the transformation
of IV -^ V takes place appears probable on the basis of Danby's data
for the perfusion of testis tissue with IV. The resulting high testoid
activity in the perfusate might well be attributed to V which is the
most active of all testoid steroids. That VI cannot arise directly from
IV may be deduced from the fact that Dorfman and Hamilton (2)
were unable to recover VI from the urines of eunuchoid men after
the administration of dehydroisoandrosterone.

OH

O

II

/^

HO-

(IV)

Dehydroisoandrosterone

HO-

A/

Testosterone

(VI)

Androsterone

The light thrown on the nature of the immediate precursors of
the sex hormones is meager indeed, but in the field of the adreno-
corticosteroid hormones the darkness is almost total. We know only
that these hormones from adrenal tissue are accompanied by inactive
steroid substances which might be converted to active hormone by
appropriate oxidation and reduction. It would be most instructive
to perfuse such substances through adrenal glands to see if they are
convertible to active hormone.

While one may thus indicate possible chemical precursors of
the known steroid hormones by administering model substances either
iti vivo or in vitro, the normal course of metabolism will continue to be
a matter of speculation until a more direct attack is devised. The

309

GREGORY PINCUS

method of Bloch using deuterized cholesterol is limited by the need
lor relatively large amounts of the steroid products for mass spectros-
copy. Probably it is for this reason that only pregnanediol has been
identified as a transformation product of deuterized cholesterol. No
similar study has been carried out on other postulated intermediaries.
It is to be hoped that the development of micro methods will make
possible more extensive utilization of the isotope technique. Similarly,
in tracing the synthesis of steroids from postulated triose precursors,
the use of isotopes is clearly indicated.

Another approach to anabolic processes, particularly in mam-
mals, involves studies of the metabolism of precursors in vitro. The
perfusion of active steroidogenic organs should provide useful informa-
tion. On the basis of data derived from perfusion studies one might
attempt the isolation of the enzyme systems concerned with the ob-
served transformations. It is notable, however, that Danby, who
obtained increased testoid activity on perfusion of the bull testis with
dehydroisoandrosterone, was unable to observe any increase at all on
incubating dehydroisoandrosterone with testis mash.

One further note on the problem of steroid hormone synthesis
seems pertinent here. Emmens has made a useful distinction between
estrogens and proestrogens (folliculoids and profolliculoids in our
terminology). In dealing with substances particularly of the stilbene
series, he finds certain compounds which on systemic injection are
much less active per milligram than when applied intravaginally.
These are true estrogens because their action on the susceptible end-
organ requires much less material with direct application. Calling
the systemic median effective dose S and the median effective local
dose L, Emmens finds true estrogens (e. g., estrone, diethylstilbestrol)
give in spayed mice an S/L ratio of 50 or more. Proestrogens give an
S/L ratio approximating unity (e. g., a-phenylstilbestrol, triphenyl-
ethylene). SegalofT has shown that three proestrogens (triphenyl-
ethylene, triphenylchloroethylene, and 9,10-di-n-propyl-9,10-dihy-
droxy-l,2,5,6-dibenzanthracene) are definitely potentiated after in-
trasplenic injection (with the spleen in situ). Passage through the
hepatic portal system increased their activity some three to six times.
Since injection into the isolated spleen does not effect such potentia-
tion, it is probable that the liver converts these substances to more
active compounds. The possibility that the liver may play a role in

310

STEROID HORMONES

steroid hormone synthesis should not be oxci looked. Estrone per-
fused through the rat liver or iueuhaled wilh liver brei is at least in
part converted to a-estradiol, that is, into more active substance.
Estrone on this basis may be considered a precursor of a highly active
foUiculoid. Whatever the logical distinction may be, a special role in
steroid hormone synthesis may have to be assigned to the liver.

Destruction of Steroid Hormones

The nature of steroid hormone degradation in animals has been
more thoroughly investigated than hormone anabolic processes. The
available data has been reviewed in some detail (9). With certain
exceptions, most of the data have been arrived at by determination of
urinary steroid catabolites in normal human and animal urines or
similar urinary analyses after the administration of active hormone.
In the case of mammalian subjects, certain of the hormonally active
steroids probably undergo the series of conversions given in scheme I.

Adopting the convenient classification of the hormonally active
steroids on the basis of number of carbon atoms it can be seen tliat ihe
identified catabolic end product of C21 hormones is pregnanediol. It
is not established that this is the only end product. Furthermore,
pregnanediol is not found in the urine of certain species (c. g., cat and
monkey) nor is it found therein after progesterone administration.
Finally, quantitative studies indicate that only a portion of administered
progesterone and desoxycorticosterone is converted to pregnanediol.
It is probable, therefore, that other breakdown products of the C21
steroids should be sought. The pregnanolones, which have been
found in human pregnancy urine would seem to be likely metabolites,
but attempts to identify them in urine after the administration of C21
steroids to human and animal subjects have thus far been unsuccessful.

For the C19 steroids we have somewhat more detailed informa-
tion as to possible modes of catabolism. Dorfman and Hamilton
have attempted to set up a balance sheet for testosterone administered
to eunuchoid men. After making certain assumptions the data indi-
cate that androsterone may be taken as the principal catabolite,
accounting for a little over 60% of the administered hormone. In
somewhat better controlled experiments with guinea pigs, isoandros-
terone (the principal urinary catabolite to be identified in this species)

GREGORY PINCUS

CH,

I

c=o

I

/\

A/

Progesterone

Scheme i

CH3

I
COH

/X/X

/\

HO—

A/

Pregnanediol-3a, 20a

+ ?

/%

0=4

/X

A/

CH2OH

A/V

Dcsoxycorticosteronc

/X

/\

HO

CH,

I
COH

I

Pregnanediol-3a, 20a

+

OH

I

/^

0=

A/

Testosterone

HO.

0

1

0

II

/\

1

1

II

/X

/

V ^

/\

/

k

1 1
/

•\/

\

/

HO...

k/

\

/

H

Androsteronc

(isoandrostcrone

in guinea pig)

H

Etiocholan-3a-ol-17-one

+ ?

OH

I

/X/X

HO

Estradiol

O

II

/X/X

OH
/^/^-OH

HO

> /\

Estrone

HO—

+

Estriol

312

STEROID HORMONES

represents 30% of the administered hormone. From chimpanzees
receiving testosterone proprionate orally, Fish and Dorfman obtained
in the urine androsterone, androsten-17-one, and etiocholan-3-ol-17-
one.

Dorfman and Hamilton (2), on the basis of products isolated
from urine following administration of various precursors, have sug-
gested the mechanism of scheme II for the conversion of testosterone
to androsterone. It should be noted, however, that this scheme does
not exclude other degradation products which may be produced in

o=

OH

i

Scheme ii

O

II

/X!/\

- 2 H

Vv/

o=

+ 2H

V

. o=J

OH

O-d

+ 2H

/\

HO..

OH
/X

o

/X

+ 2 H

/\

HO.

H

H

Androsterone

313

GREGORY PINCUS

smaller yield. The acid test of the scheme for the interconversion of
testosterone and androsterone would be met by certain identification
of the intermediaries after administration of active hormone. Finally,
Dorfman has reported a conversion of exogenous testosterone to etio-
cholan-3-ol-17-one and to isoandrosterone in man. To account for
these reduction products several other mechanisms are indicated (7).

While testosterone has generally been recognized as the true
hormone of the testis, certain species may produce other, now un-
known, testoids. For example, Ruzicka was unable to obtain testos-
terone from hog testis tissue, and Hirano has reported the isolation
of testalolone and possible related steroids. If we want to ascertain
the fate of steroid hormone produced by the testis, a more complete
knowledge of just what the testis does produce is obviously required.

The course of conversion of the natural folliculoids (the Cis
steroid hormones) indicated above appears to be well established.
In the rabbit /3-estradiol is produced, whereas in man, the monkey,
and probably also in the rat and guinea pig a-estradiol is a conversion
product of estrone. Estriol is indicated as a metabolite in the various
animal species, but only in man has it been isolated as a urinary end
product following administration of estrone.

In all species investigated, the recovery of known catabolites
from the urine after the administration of estradiol or estrone to the
intact animal is low indeed, ranging from 1-2% to at most a possible
15% of the administered hormone. The fate of the balance has been
the subject of some investigation. It has been claimed that the
missing material undergoes an enterohepatic circulation much like
that of the bile acids. This claim requires more evidence, but there
is no doubt that a good amount of foUiculoid appears in the bile after
estradiol and estrone injection. Dr. J. Schiller and the writer have
recently shown that the removal of most of the liver of female rats
results in a large increase of urinary folliculoid. The injection of
estrone into such partially hepatectomized rats leads to the recovery
in the urine of folliculoids accounting for about 60% of the administered
hormone. From female rats with intact livers, 13%, at most, of ad-
ministered estrone can be recovered. Zondek has postulated a liver
estrinase, for inactivation can be oi«erved in vitro on incubating fol-
liculoid hormones with liver mash. Attempts to isolate a phenolase
that might be responsible for such inactivation have been unsuccessful.

3H

STF.ROin HORMONES

Since the inactivation of estradiol by rat liver slices appears to depend
on adequate supplies of thiamin and riboflavin, an approach to the
inactivating systems through enzymes dependent on these vitamins is
indicated.

It is fairly well established that the liver is largely responsible
for the conversions: estradiol ^ estrone — > estriol. Liver perfusion
not only results in a rapid interconversion of estradiol and estrone, but
seems also to hasten the production of estriol. In contrast, estriol is
produced in negligible amounts after perfusion of the kidney with
estrone. Marrian (8) has suggested that folliculoid breakdown pro-
ceeds through estriol to a diketone, thence to other inactive substances.
Since the liver is such an active producer of estriol, the postulated
diketone should be sought in liver extracts, particularly after the
administration of natural folliculoid or following the incubation of the
hormones with liver slices or mash.

OH

/X/\

-OH

O

/X/X

=o

HO-

Estriol

HO— l^

Estratriene-3-ol-16,17-dione

While the catabolic role of the liver has been most intensively
examined in studies with the folliculoids, its responsibility for the
breakdown of testosterone and progesterone is also established. It is
sufficient merely to implant pellets of these hormones into the hepatic
portal system (intrasplenic implantation is used principally) to observe
the remarkable absence of testoid or luteoid effects. Implantations
elsewhere are clearly effective. In contrast, the administration of
these steroids to partially hepatectomized animals results in remarkable
exaggeration of their typical effects.

The nature of the products which the liver produces from these
hormones is not too well established. Recent work indicates that
testosterone is converted to a mixture of 17-ketosteroids one of which,
A*-androstene-3,17-dione, has been identified chemically.

In contrast to the information available for the foUiculoid,
testoid, and luteoid hormones, practically nothing is known about the

315

GREGORY PINCUS

fate of the corticosteroids in the organism. We have mentioned that
desoxycorticosterone is degraded in part to pregnanediol. No parallel
information is available for the other active corticoids such as VII
and VIII, largely because these substances are available as pure

CH2OH CH2OH

0=

-^\/\

(VII)

Corticosterone

c=o

OH

(VIII)

ll-Dehydro-17-hydroxycorticosterone

chemicals in only very small amounts. (They have been obtainable
thus far only from adrenal gland extracts and the yields are very low.)
Although we know that they are very rapidly inactivated in vivo, their
breakdown products are unknown. It is suspected that certain of
them may be converted to 17-ketosteroids, since adrenalectomy or
atrophy of the adrenals leads to a marked reduction in urinary 17-keto-
steroid output; but since other possible precursors of 17-ketosteroids
have been isolated from adrenal tissue {e. g., androstenedione) experi-
mental verification is needed.

Eventually we shall have a detailed description of steroid
hormone transformation in animals and of the particular enzyme
systems involved. If we are to know the extent to which such trans-
formations are tied up with the function of the hormones, we will have
to know not only more about these metabolic changes but also the
precise role of the steroid hormones in cellular processes.

Role of Steroid Hormones in Cellular Processes

There is an extensive literature containing qualitative and
quantitative data on the effects of administering steroid hormones to
animals. It is not possible as yet to interpret these effects in terms of
the chemistry of the cell. Thus, the remarkable growth of vaginal
epithelium induced in spayed animals by folliculoids has been rather
intensively studied. That it can be induced by the direct application

316

STEROID HORMONES

of minute amounts of active hormone to the vaginal epithelium has
been demonstrated. That these catalytic amounts of hormone cause
a remarkable spurt of mitoses in the epithelial cells is also well known;
but the means by which the hormones induce this growth have not
been elucidated. In the case of the uterus, where parallel growth
transformations occur, there has been some indication that the follicu-
loids cause a vasodilation of the uterine vessels which is followed by
an influx of water into the uterine tissues. Similar large water reten-
tion occurs in the sexual skin of the pig-tailed monkey. Is this ab-
sorption of water accompanied by an intake of growth-promoting
substances from the blood? Or is the foUiculoid directly a growth
promoting factor? Our fragmentary knowledge of the biochemical
processes underlying end-organ responses to the steroid sex hormones
debars an answer to even such simple questions. In their recent
review on the influence of hormones on enzymic reactions, Jensen and
Tenenbaum (4) have a section entitled "Sex Hormones," in which
they state: "Estrogens and androgens no doubt aff'ect metabolism
either directly in the tissues or indirectly by stimulating or inhibiting
the rate of secretion of other endocrine organs. In vitro studies on the
eff"ect of this group of endocrine principles on a given enzyme system
have as yet not been reported."

The adrenal cortex steroids which on administration to animals
induce gluconeogenesis from protein and suppress the utilization of
glucose would seem to off"er material for elucidating the role of such
steroids in cellular processes. But the few attempts to define the role
of these steroids in carbohydrate metabolism have led to apparently
conflicting results. Evidence has been presented that adrenal cortex
extract inhibits glycogen breakdown and that it also speeds glycogen
synthesis in the rat liver. Certainly the adrenalectomized animal has
little or no liver glycogen. Glycogen synthesis from pyruvate and
^-lactate by liver slices is increased on addition of adrenal cortex
extract, but there is no increased synthesis from d',/-alanine or </-gluta-
mate. Kidney slices from adrenalectomized animals show decreased
formation of carbohydrate from a',/-alanine and no reduction in syn-
thesis from pyruvate. The implication is that adrenocortical hor-
mones affect carbohydrate synthesis in liver and kidney in diametri-
cally opposite ways. What is requisite is not only further data and
accurate checking, but a direct attack on the enzyme systems inxoKed.

GREGORY PINCUS

The tissue slice technique merely allows one to demonstrate an accelera-
tion or inhibition of a set of metabolic events. But unless the enzyme
systems involved are isolated and their activity measured directly, the
exact participation of the hormones cannot be accurately worked out.
They may be coenzymes at some step in the synthesis of glycogen or
they may accelerate the synthesis of an enzyme essential for glycogen
production.

It has indeed been suggested that the protein breakdown that
follows administration of adrenal cortex steroid is due to the increase
in liver arginase and not to gluconeogcnetic processes. But an increase
in kidney arginase occurs in mice receiving steroid substances that
promote protein synthesis. Can we explain this apparent discrepancy
by saying that liver arginase functions in protein breakdown and
kidney arginase in protein synthesis?

This promotion of protein synthesis by various steroid sub-
stances (e. g., testosterone, androstanediol, a-estradiol, and others) is
a biochemical phenomenon of great interest and importance. Its
elucidation in terms of cellular mechanisms is a preliminary to the
explanation of the remarkable series of growth phenomena affected
by the steroid hormones. Kenyon's review (5) of the data on man
clearly defines the effects of testoids and folliculoids on nitrogen balance.
The reader is also referred to Albright's engaging Harvey lecture (1)
on Cushing's syndrome for an astute and stimulating discussion of
the role of steroid hormones in protein metabolism and osteogenesis.

It is notable that the principal protein anabolic steroids are
testoids, and one is tempted to ascribe the muscularity of the virile
male to testis hormone. The proverbial flaccidity of the eunuch
tends to strengthen this notion. But when one examines more closely
the sex hormone complement of the male, curious anomalies appear.
The male animal produces not only testoid but also folliculoid. Cer-
tain stallions excrete more estrone into the urine than do pregnant
mares. The bull, according to Marker is a remarkable excretor of
pregnanediol, the urinary metabolite of progesterone, whereas little
or no pregnanediol is excreted by the steer. Conversely, the female
animal produces testoid hormone. Ovaries transplanted to the ear
of castrated male rats maintain the reproductive organs of the male,
but when they are transplanted to warmer sites (e. g., onto the kidney)
their testoid activity does not appear. Can we attribute the difference

318

STEROID HORMONES

between maleness and femaleness to the temperature of the gonads?
On the basis of urinary hormone measurements the difference between
men and women rests on a difference in the ratio of male to female
sex hormone, for the absolute amounts of each excreted varies widely
in each sex. It would seem to follow, therefore, that the development
of the typical secondary sexual characteristics depends on the balance
of certain steroid chemicals in the body. What cellular processes are
controlled by this balance? There is no direct neutralization of testoid
by folliculoid. Again we find a biochemical mystery.

Conclusion

In this presentation of certain problems in the biochemistry
of steroid hormones there has been no attempt at an inclusive review.
Essentially we have been engaged on a rather specialized gap-finding
mission. If there has seemed to be an emphasis on our ignorance, the
purpose thereof is ob\'ious. The frontiersman faces the unknown;
but behind him and at his disposal are innumerable known resources.

The known resources and the many frontiers of steroid hormone
endocrinology could occupy our attention for many pages. In the
decade that has followed the original identification of a few steroid
substances as active hormones, investigation has been made of their
role in a host of normal and pathological processes. There are data
on their involvement in menstruation, pregnancy, lactation, growth,
reproduction, embryogeny, puberty, senility, cancer, dermatology,
overt endocrinopathies, leukemia, inflammation, diabetes, melan-
cholia, schizophrenia, toxemia — most of the ills and physiological
goods of man and beast. These hormones are called upon to regulate
the other glands of internal secretion; they depress and stimulate
pituitary secretion and thyroid activity. The folliculoids affect not
only the organs of reproduction but also bone growth, sugar regula-
tion, skin growth, and carcinogenesis. Progesterone maintains em-
bryo life and pregnancy as well as affecting mammary growth, the
salt balance in blood and tissues, and smooth muscle contractility.
The testoids are not merely androgenic, they stimulate the synthesis
of protein and are often luteoid, acne-producing, and factors in bald-
ness. The adrenal cortex steroids function in sugar, fat, protein, and
salt metabolism; they affect emergency reactions to damage, processes
in convalescence, and lymphatic function.

GREGORY PINCUS

From conception to death the steroid hormones regulate, con-
trol, arbitrate, and defend. With dexterity and imagination chemists
have synthesized compounds that perform and facihtate each of the
various effects of the hormones. With the array of activities and
compounds available we are predestined to full biochemical explana-
tion of these activities. Unknov^^n only are the rate and manner of
the eventual revelation, but therein is the key to the charm and
excitement of experimentation.

Rejerences

(1) Albright, F., Harvey Lectures, 38, 123 (1943).

(2) Dorfman, R. I., and Hamilton, J. B., J. Biol. Chem., 133, 753 (1940).

(3) Fieser, L. F., The Chemistry of Natural Products Related to Phenanthrene.
2nd ed., Reinhold, New York, 1937.

(4) Jensen, H., and Tenenbaum, L. E., in Advances in Enzymology, Vol. IV.
Interscience, New York, 1944, p. 259.

(5) Kenyon, A. T., Biol. Symposia, 9, 11 (1942).

(6) Koch, F. C, in Sex and Internal Secretions. 2nd ed., Wood, Baltimore,
1939.

(7) Koch, F. C, Biol. Symposia, 9, 41 (1942).

(8) Marrian, G. F., Harvey Lectures, 34, 37 (1938).

(9) Pincus, G., and Pearlman, W. H., Vitamins and Hormones, 1, 293 (1943).
(10) Selye, H., Rev. can. bioL, 1, 577-632 (1942).

320

21

PLANT HORMONES

AND THE ANALYSIS
OF GROWTH

KENNETH V. THIMANN,* associate professor of plant

PHYSIOLOGY, THE BIOLOGICAL LABORATORIES, HARVARD UNIVERSITY

/T IS probably true that, in the field of biochemistry and
physiology, there are at least a dozen workers studying
animal material for every one studying plants. It would be expected,
therefore, that progress toward an understanding of the chemical
mechanisms operative in plants would be achieved more slowly than
in the corresponding field of animal biochemistry. Nevertheless,
when we survey the plant research of the last twenty years, the results
appear surprisingly good. The list of the elements needed for plant
nutrition has been more or less completed. The whole field of the
relationship between light and flowering has been opened wide, and
though the knowledge acquired has not yet been brought to the
chemical level, it is of great importance, both for physiology and for
agriculture. Somewhat parallel studies of the influence of tempera-
ture on flowering are developing. The chemistry of practically all
the known plant pigments has been thoroughly worked out; and,
while it is true that little is yet known about the formation and inter-
conversion of these pigments, this field is ripe for physiological ex-
ploitation. Classical problems such as the interconversion of starch

* At present on leave of absence for service in the Navy Department,

321

K. V. THIMANN

and sugar, and the accumulation of solutes by roots, have been greatly
clarified. Even the elusive field of photosynthesis has seen notable
progress through the investigation of the effect of intermittent light,
and the recent study of the "dark" reaction with radioactive isotopes.

In no field has more striking progress been made than in that
of the analysis of growth and its control by hormones. Darwin's work
on what was later to become a classical material for study — the oat
coleoptile — may be regarded as the first serious attempt to analyze
growth. The work was refined and advanced during the fifty years
that followed. In the first "shoot" of the dark-grown oat seedling, the
extreme tip — a region of about 0.5 mm. in length — was shown (by
the combined efforts of a number of workers) to produce a diffusible
substance which controls the responses to light and gravity of the whole
shoot. Later it became clear that the curvatures toward light, or
away from the earth, which constitute these responses, are special
cases of normal or symmetrical growth. The growth substance or
hormone, diffu.sing from the tip into the base of the coleoptile, is
asymmetrically distributed, with the result that one side grows faster
than the other, causing curvature. In the absence of external stimulus,
the hormone is symmetrically distributed, and it promotes and con-
trols the normal elongation of the organ.

Discovery of the growth-promoting activity in sources far
richer and more varied than the tips of seedlings, and the development
of quantitative methods for assay by Went and others, soon led to the
isolation of several "auxins" active in the test methods. Later it was
recognized that a large class of synthetic substances have auxin activity
(see below), an activity which is measured by methods that are now
standardized. The method most frequently used involves decapitating
the oat coleoptile, thus depriving it of its main source of auxin, and
applying the test material to one side. The curvature which results
is then measured. Alternatively, the auxin may be applied in such a
way as to induce straight growth, which is then measured under the
microscope. Small sections cut from the coleoptile, pea, or other
seedlings can be immersed in solutions of test substances and their
growth measured. Many other procedures have been used, all
involving the use of auxin-deficient material as the test object.

"Auxin a" (or auxentriolic acid) and "auxin b" (or auxenolonic
acid) were isolated by KogI, Haagen-Smit, and Erxleben from urine

322

PLANT HORMONES AND GROWTH

and from corn germ. Later, "heteroauxin" (or indole-3-acetic acid)
was isolated by the same workers (3) from urine and from yeast, and
by Thimann (10) from cultures of the fungus Rhizopus suinus. More
recently it has been isolated from wheat germ by Haagen-Smit. The
growth-promoting activity of these three substances is quantitatively
and qualitatively very similar.

CHs

1
CHC2H5

CjHbCH-

I

CH,

CzHsCH-

CH,

^— CHOHCH.CHOHCHOHCOOH

Auxentriolic acid ("auxin a")

/\

CH,
CHC2H6

V\ta/

-CH2COOH

Indoleacetic acid

3-

CHOHCH2COCH2COOH

Auxenolonic acid ("auxin b")

In addition to the three known naturally occurring substances,
a large number of unsaturated ring-containing acids have growth-
promoting activity. These include the propionic, butyric, and valeric
derivatives of indole, the acetic derivatives of indene, naphthalene,
anthracene, thiophene, and even (weakly) benzene. m-Cinnamic
acid is active, but the /rani--derivative is not. The simple esters
and amides, as well as methyl and chloro derivatives of some of these
acids, also have activity. Indeneacetic acid is of particular physio-
logical interest because its activity, though high, is localized, and unlike
the natural auxins it is not readily transported through plant tissue.
a-Naphthaleneacetic acid is also of importance because of its stability,
which has led to its wide adoption for certain practical uses in horti-
culture.

Since the extensive work on the relation between chemical
structure and growth-promoting activity has been fully reviewed
(7,10), there is no advantage in considering it here. The essential
groups are: (a) an unsaturated or an aromatic ring, and (b) a car-
boxyl or a group readily converted by the plant to a carboxyl. Ihe

323

K. V. THIMANN

carboxyl group must be separated from the ring by at least one carbon
(or oxygen) atom. The situation in which a number of chemically
different substances exert the same biological activity has occurred
also in the field of vitamins, though there, perhaps, the divergence
between the natural and synthetic active substances is not so great.
For synthetic materials with vitamin activity the term "vitamer" has
been proposed (1). On the same basis, synthetic substances with
growth-promoting activity in the standard auxin tests could be called
"auximers," but at present there seems little need for this term.

It is not proposed to discuss in this essay the chemistry of the
auxins but rather to consider the mechanism of their action, with par-
ticular reference to the nature of plant growth.

Historically, the first proof that the auxins influence more than
one type of growth came from the study of bud inhibition. It is typical
of shoot growth that, when one bud is vigorously developing, others
below it on the same stem are inhibited from doing so. The apical
bud thus "dominates" the lateral buds. Only when it is removed,
as in pruning, can the lateral buds develop. After evidence had been
brought forward, mainly by Snow, that this inhibition is exerted by
means of a substance diffusing out of the growing bud toward the
tissues below it, it was finally shown that pure auxin b, or also indole-
acetic acid, in quantities comparable to those of the auxin produced
by the growing bud, can duplicate the effect. Paradoxically, then, a
growth-promoting substance is here acting as a growth inhibitor (6) . It
should be noted that the movement of auxin in living tissue, as out of a
bud into the stem, is invariably in the direction from apex to base.
This movement is strictly polar, even taking place against an auxin
gradient. Transport occurs in the reverse direction only in dead
tissues, or when unphysiologically high auxin concentrations are
applied.

Very shortly afterward, two other important cases were brought
to light. It was demonstrated, by Went and Bouillenne, that a root-
forming hormone diffuses downward out of buds and leaves to stimulate
the formation of roots at the base of a cutting. In the purification of
this hormone. Went and Thimann found that root-forming activity was
concentrated parallel with growth-promoting activity (?. e., auxin),
and finally that numerous auxins — synthetic and otherwise — acted as
root-forming hormones. At about the saine time. Snow showed that,

PLANT HORMONES AND GROWTH

in young shoots, thr ramhium is stimulated to divide by a diffusible
substance and that pure indoleacetic acid liad the same effect. Like-
wise, the substance in orchid pollen, shown by Fitting to be responsible
for the swelling of the ovary on fertilization, was identified with auxin.
Finally the elongation of roots, unlike that of shoots, was shown to be
inhibited by auxins. Wherever synthetic substances have been tested,
their influence on these other growth processes has been to some
extent quantitatively comparable with their effectiveness in producing
simple growth by elongation (5,10).

Thus, unlike many animal hormones of specialized activity,
the functions of auxins are manifold. The stream of auxin coming
from the apical bud and the young leaves in it promotes the elongation
of the shoot, stimulates the cambium to produce thickening, inhibits
development of the lateral buds lower down, and promotes the forma-
tion of root initials toward the base. In the flower, the auxin con-
tributed by the pollen, or released from the ovary tissue by fertilization,
promotes the swelling of the ovary into a "fruit" and also prevents the
fruit stalk from becoming separated (inhibition of abscission) .

Morphologically, at least, these functions all appear to be very
different. Not only do they comprise effects of opposite sign (namely,
growth and inhibition) but even where the effect is positive the types
of growth controlled by auxin differ widely. The stimulation of
cambium or root initials to divide bears little resemblance to the
promotion of simple elongation without cell division, as in the oat
coleoptile. It is hard, therefore, to avoid the conclusion that the
primary effect of auxin is in all tissues the same — a fundamental
reaction whose morphological sequelae depend on the kind and age of
the tissue, the availability of other interacting substances, and the ex-
ternal and internal conditions. This concept and the experiments
bearing on it will be taken up in more detail below.

The general position described above has been established for
some time. During the last few years the field has developed mainly
along the following three lines.

(a) Practical Applications. Originally auxin was most gen-
erally used to induce the formation of roots on cuttings. Auxin
treatment both accelerates rooting and increases the number of roots
formed in cuttings of a wide variety of herbs, shrubs, and trees. It
does not, of course, solve all the problems of vegetative propagation.

K. V. THIMANN

In particular, the time of year and the age of the plant from which
cuttings are taken remain as important variables. A few species also
do not respond at all to auxin, their rooting being limited by other
factors which are not yet understood.

A second major application depends on the fact that auxin not
only inhibits lateral bud development, as described above, but also
inhibits the formation of the abscission layer, a special layer of cells at
the base of a petiole or fruit stalk, whose walls separate very readily,
resulting in the falling to the ground of the leaf or fruit. When the
fruit is sprayed with auxin solution, therefore, the formatipn of this
layer is delayed and the fruit remains on the tree for longer. Such
spraying has been very successful with apples, since the delay of a
week or two helps greatly to solve the long-standing problem of early
fruit fall. The most effective auxin for this purpose appears to be
a-naphthaleneacetic acid.

The effect of auxin in promoting the swelling of the ovary, men-
tioned above, has found interesting application in the production of
seedless fruit. The auxin is applied to the style of the mature flower,
either individually by hand, or by using a liquid spray or even vapor.
The ovary begins to swell in a normal-looking way and the process
continues, to produce seedless fruit. Tomatoes and holly are the out-
standing successes obtained by this procedure, but seedless pears and
even watermelons have been produced, and the method may have an
important future [see the review by Gustafson (2a)].

Lastly, the treatment of seeds with auxin stimulates root de-
velopment and may result in a promotion of growth which lasts for
many weeks or months. The conditions necessary for this stimulation
are, however, not fully understood and the results reported are so con-
flicting as to cast doubt on the legitimacy of this application. The
toxicity of some auxins in high concentrations has recently led to an
application as weed killer. An excellent summary of all of this ap-
plied work is given by van Overbeek (8).

All these applications have led to a scramble for trademarks,
and even patents, and the emergence of special methods of advertising
and marketing. Indoleacetic, indolebutyric and naphthaleneacetic
acids are now on the market in the form of solutions, pastes, powders,
and pills.

(b) Auxin Reserves. The auxin content of plant tissues is of

326

PLANT HORMONES AND GROWTH

great interesl. In the early work it was sufFicicnt to place Ijuds,
leaves or other auxui-produciny organs on such media as agar, and after
a given diffusion period to apply the agar to standard oat coletjptiles
for assay, a procedvu-e which measures the rate of production. How-
ever, since for some pur])oses it is necessary to know the amount of auxin
actually present, attention has reccndy turned to quaiUitative extrac-
tion of the tissues with organic solvents. This led to the discovery
that, while in most tissues a certain small amount of auxin can be
obtained by direct extraction, there is commonly from ten to fifty times
as much present in an inactive form. This inactive material yields
auxin slowly when wetted with water, but the process may be greatly
accelerated by hydrolysis with some enzymes, notably chymotrypsin,
or with alkali. Whether these inactive forms, or "auxin precursors,"
are the same in different plants and different types of tissues is not clear
as yet. At least one of them appears to be an auxin protein.

The biological significance of these large auxin reserves is worth
consideration. It seems probable that growth is commonly controlled,
not so much by the rate of transport or even the rate of true formation
of auxin, as by its rate of liberation from the stored form. The sudden
onset of growth in tree buds in the spring, and its cessation in early
summer just when conditions of light, temperatuie, and nutrition would
be expected to fa\ or both auxin production and the formation of
organic structural iriaterials, may find its explanation in this way.
Many types of infection, as by gall and nodule bacteria and certain
fungi, give rise to abnormal growth which corresponds well with the
assumption of excessive local auxin formation; in the case of legume
root nodules, large quantities of auxin are undoubtedly present. These
pathological phenomena may be due to liberation of auxin from the
inactive precursors just mentioned. It seems safe to predict that fur-
ther study of auxin precursors will not only bring to light some interest-
ing enzyme systems, but may lead to explanations of a number of
normal and abnormal growth reactions.

(c) Analysis of the Growth Process. The discovery of an active
substance, whether vitamin or hormone, is of importance for its own
sake and also because the experimental use of the substance can
elucidate the physiology of the processes which it controls. The study
of growth in plants has made marked progress through the use of
auxin as a tool. Before discussing this use of auxin, however, it

K. V. THIMANN

is necessary to consider the distinction between direct and indirect
effects on growth. Some of the observed effects of auxin are in part
at least indirect. When swelUngs are produced, or fruits induced to
grow, under the stimulus of auxin, there must be a continuous supply
of carbohydrates, amino acids, and especially water to the growing
zone. The plant, therefore, must have some mechanism whereby
materials for growth are accumulated at the place at which they are
being used. To take the simplest case, that of the carbohydrates, it
could be considered that, as the sugars are converted to polysaccharides
and deposited as such in the cell wall, the concentration of soluble
sugars decreases and hence more carbohydrates would flow there in
consequence of the concentration gradient. However, accumulation
in plant tissues does not always follow the gradient. Thus, salts are
concentrated in young roots by a factor many times the concentration
of the external nutrient solution, and sugars are sometimes accumulated
in quite high concentrations in storage organs. Auxin itself, as men-
tioned on page 324, is freely transported against its gradient. The
accumulation of materials in the growing organs, and the transport of
materials there, consequently do not necessarily constitute a "gradient"
process; and numerous suggestions have been made that auxin exerts
its growth-promoting activity through the accumulating or "mobiliz-
ing" of food materials. Auxin is envisaged as in some way controlling
transport, so that, wherever auxin is, other materials will be collected.
This view has been expanded to include mobilizing effects on hypo-
thetical special organ-forming substances, such as factors for root
growth, stem growth, and leaf growth (9). The evidence for such
organ-forming substances is indirect only; evidence for a mobilization
phenomenon as the cause of the growth-promoting activity of auxin is
at present by no means convincing.

It remains true, of course, that when auxin produces growth
there is an accompanying accumulation of materials. It has been
shown, for instance, that application of concentrated auxin paste to
decapitated bean plants produces localized swelling which is associated
with a large increase in soluble carbohydrates and total dry weight.
It has also been shown in ingenious experiments (9) on oat seedlings
that a zone of the plant which has been treated with auxin and later
cut off is subseqixently able to grow in solution better than a control
zone. Hence it is not the occurrence of such accumulations, but the

328

PLANT HORMONES AND GROWTH

problem of whether they arc causative of growth or not, tliat remains
unsettled.

To some extent, the need for postulating these indirect mecha-
nisms arises from the difficulty of designing experiments which bear
upon the growth process per se, and which introduce the minimum of
extraneous considerations. In the study of vitamin action, it is usual
to employ vitamin-deficient animals, but it is not .so easy to obtain with
certainty plants which are deficient in auxin. The discussion of auxin
reserves in the preceding section brings out the difficulties involved.
Green plants in the light, rich in stored auxin, vitamins, and carbo-
hydrates, would appear to provide the worst type of experimental
material. But even embryos and dark-grown seedlings, though more
nearly ideal, and certainly capable of yielding more quantitative data,
have their reservoirs of auxin and food factors. It is evident that
isolated parts of plants, grown in the dark under controlled conditions
and free from storage organs, approach nearest to the rigid require-
ments of experimental material for the study of auxin action. With
this in view many studies have been made of the growth of isolated
sections, 3 to 5 mm. long, cut from etiolated oat coleoptiles. Such
sections will elongate to a small extent in water, and to a much greater
extent in auxin solution, the amount of elongation being proportional
to auxin concentration up to a maximum at about 10 mg. per liter.
Higher concentrations cause reduction in growth and are toxic. This
elongation is achieved at the expense of the stored carbohydrates in
the cells. If sugar is added to the solution, the elongation is very
greatly increased both in amount and in duration. Nitrogenous
substances have not been found to increase growth any further, though
some salts, particularly potassium chloride, cause a further increase.
Optimal concentrations are: sucrose, 1%; potassium chloride. A// 100;
and indoleacetic acid, 1-5 mg. per liter. These isolated sections pro-
vide a simple growing system which is as free from indirect controls
as can readily be envisaged. Growth is a function of temperature, of
the osmotic gradient, and of the concentrations of auxin, potassium
ions, and sugar. It is difficult to see how auxin could cause growth
here by mobilizing anything.

Mainly with this material it has been shown that growth is
controlled by a specific process of respiration. It had been shown
many years earlier by Bonner that cyanide inhibits growth over the

K. V. THIMANN

«

same rans^e of concentrations and to the same degree as it inhibits
respiration. Recently, however, certain dehydrogenase inhibitors
have been found to inhibit growth at concentrations much lower than
those which are effective on respiration. lodoacetate is particularly
effective, inhibiting all growth in coleoptile sections at a concentration
of 5 X 10~^ M, a concentration which produces only a very slight
decrease in respiration rate (2). This growth inhibition is completely
removed by adding malate, fumarate, or to a lesser extent succinate
or pyruvate. The four-carbon acids, in fact, accelerate growth some-
what over that of controls, in presence of auxin and sugar. Thus a
typical respiratory system appears to exercise control over growth.
Also, in the presence of malate, and particularly if the sections are
soaked beforehand in solutions of malate, the addition of auxin causes
a definite small increase in respiration rate. It is of importance also
that the auxin concentrations which are effective in stimulating res-
piration closely parallel those concentrations at which growth is
affected. From the fact that growth can be completely inhibited with
only a small decrease in respiration it follows that growth is not con-
trolled by the respiration process as a whole, while the influence of
iodoacetate and the four-carbon acids suggests that the dehydrogena-
tion of the four-carbon acids is a controlling process in growth. Evi-
dently this system accounts for only a small part of the total respiration
of the coleoptile tissue.

Striking support for these conclusions comes from a study of the
effect of auxin on the rate of protoplasmic streaming. The long
epidermal cells of the coleoptile are particularly favorable for measure-
ments of streaming rate, which, under controlled external conditions,
is nearly constant and reproducible. When auxin solution is applied
to the coleoptile, there is an immediate rise in the streaming rate,
which, however, if no other materials are present, returns to normal
after ten to twenty minutes. The presence of sugar allows the in-
creased rate to be maintained for several hours. Oxygen is necessary
for the reaction; and if both oxygen and sugar are available there is
a good parallel between the concentrations of auxin which accelerate
streaming and those which accelerate growth. Since, also, the
increase in rate of streaming occurs before any increase in growth rate
can be detected, it is reasonable to conclude that the effect on stream-
ing is one of the first stages in the process, caused by auxin, which

PLANT HORMONES AND GROWTH

results in increased cj,ro\vth. Fnrther examination was therefore
made of the relationship between streaming and respiration.

lodoacetate, in the concentration which inhibits growth, com-
pletely inhibits the acceleration of streaming by auxin, and the in-
hibition, as with respiration and growth, is removed by the addition
of malate. Malate alone has no detectable effect on the streaming
rate; but in the presence of auxin it has the following very significant
effects. After coleoptiles reach a certain age (about five days at
25° C. in weak red light), they no longer show any acceleration, either
of growth or of streaming, when auxin is added. But when malate
is added, auxin is able to produce its normal acceleration of streaming
in this old material. Similarly, if young sections are given prolonged
soaking in water they behave like old coleoptiles, and here too the addi-
tion of malate restores the sensitivity of streaming to the influence of
auxin (4). Apparently, therefore, aging or prolonged soaking in
water depletes the coleoptile of its store of malate or other four-carbon
acids.

From these observations it can be concluded that there is a
special part of the respiratory system whose activation results in growth.
This growth is probably brought about in the first instance through
acceleration of the rate of protoplasmic streaming. Finally, the system,
involving streaming, a part of the respiration, and growth, is mediated
by the four-carbon acids, and auxin apparently acts as an essential
coenzyme. It should be added that some unpublished experiments
with other materials strongly indicate that this growth-controlling
system is present in similar form in materials quite other than the
coleoptile.

It will be seen that, while a good deal of information is now
available about the mechanism of auxin action, and therefore the
mechanism of growth, the interrelationships between the various
factors and processes are by no means clear. How, for instance, can
a process of respiration be linked to one of growth?

It is important to consider first what constitutes "growth" in
its simplest form. To a zoologist growth is commonly thought of in
terms of cell multiplication. In the oat coleoptile, however, the
number of cells reaches its final value at an early age and thereafter
growth is by cell elongation only. Such a situation, with modifica-
tions, is common in plant tissues. Furthermore, a coleoptile section

K. V. THIMANN

"growinj;'' in pure auxin solution decreases in dry weight as does,
for instance, a sea-urchin egg dividing in sea water. Thus the essential
minimum of growth, in plant tissues, comprises only an irreversible
increase in volume. This in turn means nothing more than the up-
take of water, or salt solutions. Now we know, principally from the
work of the California group under Hoagland, that the uptake of salts
from solution is dependent on respiratory processes, and the uptake of
water itself may well be a process of the same type. There is some
evidence to support this, for auxin has been shown to induce in potato
tissue an increased water uptake, which is dependent on the presence
of oxygen. The accumulation and concentration of water is thus seen
as a "vital" rather than a purely osmotic process. Respiration in some
way provides the free energy for the thermodynamic work involved.
The process in some ways parallels that of secretion in the kidney.
While a relation between a respiratory process and growth or water
uptake can thus be dimly envisaged, the role of the protoplasm, and
especially of its streaming, remains completely unassigned. Further
analysis may well lead to the correlation of many fundamental, and
superficially different, types of process.

Nothing as yet has been said about the part played by the cell
wall. It is only the primary wall which can stretch to allow visible
growth. Once secondary wall has been deposited, as in the older
coleoptile, extension ceases. Yet we have seen that at least one of the
reactions leading to growth, that of the acceleration of streaming, can
be initiated by auxin in presence of malate in this nongrowing tissue.
Part of the growth process may thus be completed, although visible
increase in size is mechanically prevented. It may become neces-
sary, then, to define growth in terms of its constituent biochemical
processes rather than in terms of its external appearance.

A suggestive parallel to this, though it may have little in common
with auxin effects, is furnished by the reaction caused by the "wound
hormone," traumatic acid, which is an unsaturated acid present in the
brei of many tissues, particularly in bean extract. It causes local
growth and swelling, being responsible for the so-called wound reaction.
This would appear, superficially, to be a very clearly defined response
both anatomically and physiologically, since wound-healing tissue
usually consists of masses of rapidly dividing cells. Yet in beans it
has been shown that, depending on the variety, the response to wound

PLANT HORMONES AND GROWTH

hormone may consist of typical rapidly dividing cells or may involve
principally enlarged (elongated) and undivided cells. No doubt the
underlying response is the same in both cases, but the external visible
effect may differ. It does not seem at present that there is any specific
hormonal stimulus for cell division per se. However, the control of
cell division is one of the problems which on further study may yield
information about the underlying mechanism of the wound reaction
and its relation to other forms of growth.

Lastly, there are a series of problems involving the interaction
of auxin with other special substances. Biotin and auxin act co-
operatively in the formation of roots by certain types of cuttings.
Thiamin and pyridoxine are necessary for the elongation of roots,
and auxin inhibits this elongation, but whether the availability of
these vitamins is inhibited is not known. In fact, the growth physi-
ology of the root is still largely unexplored, despite a great many in-
genious experiments (10). Thus it may be concluded that no more
than a beginning has been made in the study of plant growth
hormones.

References

(1) Burk, D., and Winsler, R. J., Vitamins and Hormones, 2, 306 (1944).

(2) Commoner, B., and Thimann, K. V., J. Gen. Physiol., 24, 279 (1941).
(2a) Gustafson, F. G., Botan. Rev., 8, 599 (1942).

(3) Kogl, F., Ber., A68, 16 (1935).

(4) Sweeney, B. M., and Thimann, K. V., J. Gen. Physiol., 25, 841 (1942).

(5) Thimann, K. V., Plant Physiol., 13, 437 (1938).

(6) Thimann, K. V., Biol. Rev. Cambridge Phil. Soc, 14, 314 (1939).

(7) Thimann, K. V., and Bonner, J., Physiol. Revs., 18, 524 (1938).

(8) van Overbeek, J., Ann. Rev. Biochem., 13, 631 (1944).

(9) Went, F. W., Am. J. Botany, 25, 44 (1938); Plant Physiol., 13, 55 (1938).
(10) Went, F. W., and Thimann, K. V., Phytohormones. Macmillan, New

York, 1937.

3'^:

o

22

CHEMICAL MECHANISM
OF NERVOUS ACTION

DAVID NACHMANSOHN, research associate, department of

NEUROLOGY, COLLEGE OF PHYSICL^NS AND SURGEONS,
COLUMBIA UNIVERSITY

r\ VIEW of the dominating role of the brain and of the
central nervous system in the body, the problem of the
mechanism of nervous action has always attracted general interest.
When, from Galvani's experiments it became clear that nerves generate
electricity, the news was received with great enthusiasm, not only
throughout the scientific world, but also among all educated people.
For more than a century the analysis of the electric changes constituted
the only means of studying nervous action. And yet, 150 years later,
Herbert S. Gasser, one of the leading electrophysiologists of our time,
compared the electric spikes to the ticks of the clock (5). Both are
only signs of activity in an underlying mechanism: "It follows then that
if spikes are but manifestations of activity in the inherent mechanism
of nerve fibers, the story of nerve is by no means told when the spikes
are described. We need to know something about the mechanism
which produces them — how it is maintained, its capacity for work,
and when and how the work is paid for." In spite of all the valuable
information obtained by the study of the physical aspect, knowledge
of the molecular changes, /. ^., of the chemical reactions involved, is
necessary for the understanding of the mechanism of nerve activity.

The special function of the nervous system, which is tliat of
carrying messages from one distant point of the body to another, may

335

DAVID NACHMANSOHN

be subdivided into three successive phases: a stimulus reaching a
neuron must initiate an impulse; the impulse, once initiated, must
be propagated along the axon; and, finally, it must be transmitted
either to a second neuron or to an effector cell. Early in this century,
the idea was evolved that a chemical compound may be connected with
the third phase, namely, the transmission of the nervous impulse from
the nerve ending to the effector cell. T. R. Elliot suggested, in 1905,
that adrenalin may be the transmitter of the impulse from the sympa-
thetic nerve ending to the effector cell. He based this idea on the
similarity between the action of adrenalin and the effect of stimulation
of sympathetic nerves. In 1921, Otto Loewi discovered that, follow-
ing vagus stimulation of the frog's heart, a compound appeared in the
perfusion fluid which, if transmitted to a second heart, produced an
effect similar to that of vagus stimulation. Accepting the basic idea
of Elliot, Loewi concluded that the compound (which was later identi-
fied with acetylcholine) is actually liberated from the nerve ending
and acts as a transmitter of the vagus impulse to the heart cell. Loewi's
concept of "neurohumoral" transmission was widely accepted among
physiologists.

In 1933, Dale tried to extend this idea of a "chemical media-
tor" of the nerve impulse to the neuromuscular junction and to the
ganglionic synapse. His theory was based essentially on the same
type of evidence as previously applied by Loewi in the case of auto-
nomic nerves. In this instance, however, the theory encountered
strong opposition. Besides many contradictions and difficulties there
were two main objections. The first was the time factor. This factor
was of lesser importance in the case of the slowly reacting cells in-
nervated by the autonomic nervous system. But the transmission of
nerve impulses across neuromuscular junctions and ganglionic synapses
occurs within milliseconds. No evidence was available that the
chemical process can occur at the high speed required, and Dale and
his associates admitted this difficulty. The second objection was still
more fundamental. According to leading neurophysiologists like
Sherrington, Fulton, Gasser, and Erlanger, the excitable properties of
axon and cell body are basically the same. The electric signs of
nervous action therefore did not support the assumption that the
transmission of the nerve impulse along the axon differs fundamentally
from that across the synapse (6).

NERVOUS ACTION

The concej)! of a cluMiiical mediator released ii die nerve ending
and acting directly on the seeond neuron thus appeared in many re-
spects to be unsatisfactory.

General Approach

Two features of nervous action are essential to an understanding
of the problems and difficulties involved, viz., the high speed of the
propagation of the impulse and the infinitely small energy required.
In meduUated mammalian nerve, the impulse travels at a rate of one
himdred meters per second and the energy required per impulse per
gram nerve is less than one-tenth of a millionth of a small calorie.
The recording of such an event offered many difficulties, even with the
use of physical methods. A really adequate electrical recording instru-
ment became available only with the introduction of the cathode-ray
oscillograph by Gasser and Erlanger. Still more difficult was the
detection of the energy involved. It is not surprising that Helmholtz,
who first demonstrated heat production in muscle, failed to demon-
strate it in nerves. Even A. V. Hill was unable to detect any heat
production in nerve for a long time, and only when he and his associates
developed thermoelectric methods of an amazingly high degree of
perfection did it become possible to measure amounts of heat of as
small an order of magnitude as are produced by nerve activity.

If even physical methods encountered so many obstacles, it is
obvious that the study of the chemical reactions connected with an
event of this kind must ofTer serious difficulties. No adequate methods
are available for determining directly chemical compounds appearing
in such infinitely small amounts and for such short periods. But the
development of biochemistry, especially during the last twenty years
has shown that, in such cases, much information may be obtained by
the study of biocatalysts. Nearly all chemical reactions in the living
cell are catalyzed by enzymes. Since Buchner's demonstration, in
1897, that fermentation may occur in cell-free extracts, a great number
of enzymes has been isolated, and the chemical reactions of living
cells have been studied in vitro. While many cell constituents and
intermediates, especially those closely connected with cell activities,
are extremely unstable and occur in concentrations too low for chemical
analysis, a great number of enzymes are relatively stable and, under

337

DAVID NACHMANSOHN

appropriate conditions, their activity is suitable for investigation.
Moreover, study of cellular constituents as such does not indicate which
changes they may undergo in the living cell and at what rate. It is
true that study of an isolated enzyme alone does not yet permit direct
correlation with its cell function, since there are so many simultaneous
reactions in the complex system of the living cell. But, by investigating
a great number of reactions and series of reactions, and by studying
their connection with events in the intact cell, valuable information
of the chemical mechanism of cell function may be obtained. One
of the most conspicuous examples of such an analysis is the development
of muscle physiology. Because of the pioneer work of A. V. Hill and
O. Meyerhof, many physical and chemical changes have been cor-
related; and our concept of the mechanism of muscular contraction
has gone through a real "revolution," according to an expression of
A. V. Hill, although we are still far from having a complete picture.

The most promising approach to the chemical reactions in-
volved in nerve activity appears, therefore, to be by way of a study of
the enzymes involved. Since Loewi's discovery suggested that the
release of acetylcholine may be connected with the transmission of
the nerve impulse, it could be expected that a study of the enzymes
involved in the formation and hydrolysis of the ester might lead to the
elucidation of the precise role of acetylcholine.

But even the best methods of enzyme chemistry available are
not adequate for the study of many problems. In many cases, in order
to find a satisfactory answer, it was necessary to combine enzyme
chemistry with the selection of special cases which offered particularly
favorable conditions for the different problems.

Role of Acetylcholine

Investigations based on this line of approach and carried out
over a period of nearly ten years have provided evidence for a new
concept of the role which acetylcholine may have in the mechanism
of nervous action (7,11). According to the new concept, the release
and the removal of the ester are considered as an intracellular process
occurring at points along the neuronal surface and directly connected
with the nerve action potential.

The action of acetylcholine may be pictured in the following
way: The nerve is, according to the generally accepted "membrane

338

NERVOUS ACTION

theory," surrounded by a polarized membrane. The polarized state
of the membrane is due to a selective permeability to potassium ions
which are many times more concentrated inside the axon than outside.
During the passage of the impulse the resistance of the membrane is
decreased. From experiments on the giant axon of the squid. Cole
and Curtis calculated that the resistance falls from 1000 ohms to about
25 ohms per sq. cm. (3). The permeability of the membrane to all
ions is increased and a depolarization occurs. This change in per-
meability may well be produced by the rapid appearance and re-
moval of acetylcholine (11). The depolarized point becomes negative
to the adjacent region and a flow of current results which stimulates
the next following point. There again acetylcholine is released and
the whole process repeated. The impulse is thus propagated along the
axon. At the nerve ending the surface is increased and the resistance
therefore decreased. This leads to a greater flow of current which
enables the impulse to cross the gap. Whereas, in earlier theories,
acetylcholine was considered as a "neurohumoral" or "synaptic"
transmitter, i. e., a substance released from the nerve ending and
acting directly on a second neuron, in the new concept the transmitting
agent is always the electric current, the action potential, but the cur-
rent is generated by acetylcholine.

The picture is consistent with the idea of the propagation of the
nerve impulse as developed by Keith Lucas and Adrian. It becomes
unnecessary to assume that the transmission along the axon diff"ers
fundamentally from that across the synapse. The assumption of a
special mechanism at the synapse diflferent from that m the axon was
the chief difficulty which had to be overcome for conciliating the
original theory with the conclusions of the electrophysiologists. This
appeared necessary for any satisfactory answer to the problem. For, if
it is true that physical methods alone are unable to explain the mecha-
nism in a living cell, it is equally true that conclusions based on
chemical methods should not be in contradiction to those obtained
with physical methods, in view of the much higher sensitivity of the
latter.

The facts on which the new concept is based have been recently
reviewed and discussed (11). Some essential features are based on
studies of choline esterase. A few examples may be given in order to
illustrate the new approach.

339

DAVID NACHMANSOHN

Choline Esterase

Time Factor. One of the essential results of these enzyme
studies is the evidence of the high rate of acetylcholine metabolism in
nerve tissue. Significant amounts may be split in milliseconds, that
is, a period of time sufficient for the passage of the impulse. Conse-
quently, the potential rate of acetylcholine metabolism is sufficiently
high to justify the assumption that it parallels the rate of the electric
changes and may therefore be directly connected with the nerve action
potential.

The frog sartorius muscle is a special case in which this problem
of the time factor has been studied and has received a satisfactory
answer. A small fraction of this muscle is free of nerve endings. By
determining the concentration of choline esterase in this part of the
muscle, in the part containing nerve endings, and in the nerve fibers,
it is possible to calculate the concentration of choline esterase at the
motor end plates. Since the number of end plates in a frog's sartorius
is known, the amount of acetylcholine which may be split during one
millisecond can be calculated. This turns out to be 1.6 X 10^ mole-
cules of the ester. About ^/s of the enzyme at the motor end plate
is localized inside the nerve ending. On the assumption that one

o

molecule of acetylcholine covers about 20 to 50 sq. A., the amount
which may be hydrolyzed during one millisecond at one end plate
would cover a surface of 100 to 250 sq. p..

Localization at the Neuronal Surface. A high concentration of
choline esterase, of similar order of magnitude to that at motor end
plates, exists at all synapses, whether central or peripheral. But the
difference between axon and synaptic region is only quantitative.
Fibers with a very thin myelin sheath, until recently considered as
nonmyelinated, like the sympathetic chain of mammals or the abdom-
inal chain of lobster, offer a favorable material for demonstrating this
distribution. Hence, the rate of acetylcholine metabolism may be
high everywhere in nerves.

Experiments on the superior cervical ganglion of cats indicate
that the enzyme might be concentrated at or near the neuronal surface.
The giant axon of squid, another special case, was used for testing such
an assumption. The axoplasm may be extruded and thus separated
from the sheath. Most of the sheath is connective tissue to which are

NERVOUS ACTION

attached two thin nicinl)ranes only a few microns thick. The whole
enzyme activity is in \hr sheath. The axoplasin is practically free of
choline esterase (1).

Bioelectric phenomena occur at the surface. The localization
of the enzyme at the neuronal surface and the high rate of acetylcholine
metabolism make possible the assumption that the ester is connected
with the electric manifestations of nerve activity. But for the interpre-
tation of the actual role of acetylcholine in the mechanism of nervous
action the activity of the enzyme had to be connected with an event in
the living cell. Such a correlation has been established in experiments
carried out on the third and perhaps most valuable special case, viz-,
the electric organ o fish.

Parallelism between Enzyme Activity and Voltage of the Nerve Action
Potential. The electric discharge in these organs is identical in nature
with the nerve action potential of ordinary nerves. The only dis-
tinction is the arrangement of the nervous elements, the electric plates,
in series. The potential difference developed by a single element is
about 0.1 v., which is the same order of magnitude as that found in
ordinary nerves. In the species with the most powerful electric organ
as yet known, Electrophorus electricus, the so-called electric eel, several
thousand elements are arranged in series from the head to the caudal
end of the organ. Thus, the voltage of a discharge amounts to
400-600 V. on the average; and, in some specimens, more than 800
V. have been observed. In Torpedo, another species with a powerful
electric organ, the elements are arranged in dorsoventral direction.
Since it is a fiat fish, the number of plates usually does not surpass
400 to 500; and, consequently, the discharge is only 30-60 v. on the
average. In the large Gymnotorpedo occidental is found on the North
American east coast, especially in the water surrounding Cape Cod,
the number of plates in scries and, consequently, the voltage may be
more than twice as high.

In 1937, the electric tissue was introduced by the writer as
material for the study of the role of acetylcholine in the transmission
of the nervous impulse. A high concentration of choline esterase was
found in the strong electric organs of Torpedo and E. eleclricus. These
organs hydrolyze in one hour amounts of acetylcholine equivalent to
one to five times their own weight. In the larger specimens, the
organs have a weight of several kilograms, so that the amount of acetyl-

DAVID NACHMANSOHN

choline which may be split in these organs may amount to several
kilograms per hour or several milligrams in one-thousandth of a second.
The concentration is of the same order of magnitude as was calculated
for the motor end plate by Marnay and Nachmansohn (10). It is
suiliciently high to make possible the assumption that acetylcholine
is directly connected with the discharge, which requires that the com-
pound must appear and disappear in milliseconds. If speculation
were to be excluded, the only means of removing this compound so
rapidly is enzymic action. The high concentration of a specific en-
zyme appeared particularly significant in view of the chemical con-
stitution of these organs. They contain 92% water and only 2% pro-
tein. The discharge is, in these organs, the main function; there is
no question here of a transmission of the nervous impulse to a second
unit, since there is none.

In the weak electric organ of the common ray, the concentration
of choline esterase is relatively low. If in the three species mentioned,
voltage and number of plates per centimeter are compared with the
concentration of choline esterase, a close relationship becomes obvious.

A more detailed analysis has been carried out of the electric
organ of E. electricus. This species is particularly favorable for such
studies because the number of plates per centimeter, and conse-
quently the voltage per centimeter, decreases from the head to the
caudal end of the organ. The choline esterase activity decreases in
the same proportion. If the electric changes are recorded and com-
pared with the chemical values at the same section, a close parallelism
is obtained between voltage and enzyme concentration. This is found
not only in regard to the variations which occur in the same specimen,
but even in absolute amounts for the variations between the individuals
(12). In measurements covering a range from 0.5-22 v. per cm.,
the line correlating the physical and chemical event was found to in-
dicate a direct proportionality.

The voltage developed in the discharge depends upon the elec-
tromotive force and the resistance. Two assumptions therefore appear
possible regarding the way in which acetylcholine may act. It may
produce an electromotive force directly by action on the surface, or
it may decrease the resistance by increasing the permeability of the
boundary. Resistance and electromotive force are closely related
properties. The drop in resistance during the passage of the impulse,

342

NERVOUS ACTION

found in the giant axon l:)y Cole and Curtis (3) and in the electric
tissue by Cox, Coates, and Brown (4), suggests that the paralleMsm
foinid between voltage and acetylcholine metabolism may be due
essentially to the effect of the ester on permeability in the way outlined
above.

Specificity oj the Enzyme. In all the experiments on the activity
of the enzyme, it was assumed that choline esterase is specific for
acetylcholine. In such a case, not only is the conclusion justified
that the substrate metabolized is acetylcholine, but the activity of a
specific enzyme determined in vitro may well be used as an indicator
for the rate of the substrate occurring in vivo. As pointed out by
Schoenheimer and Rittenberg (21), one of the most important general
results of the work with isotopes is the conclusion that "enzymes do
not lie dormant during life but are continuoiasly active." This, of
course, does not imply that all enzymes are working at an optimal rate
at every moment. In cells Uke nerve and muscle a considerable
difference must be expected between resting condition and a state of
activity. Lactic acid formation for instance, occurs in anaerobic
condition even in the resting muscle; but during tetanic stimulation
the rate increases several thousand times. Siinilaily, it cjumot be
expected that an en/.yme directly connected with the events during the
passage of the impulse is equally active in resting condition. It ap-
pears possible, moreover, and even probable, that enzymes are, to
some extent at least, present in excess above the optimum usually re-
quired. But all experience in enzyme chemistry appears to indicate
that a correlation exists between the concentration of an enzyme in a
cell and the rate at which the substrate is metabolized.

It thus appeared imperative to demonstrate the specificity of
the enzyme. The ester linkage in acetylcholine shows no peculiar
properties. It is therefore to be expected that this ester can be hy-
drolyzed by other esterases and, on the other hand, that choline
esterase can hydrolyze other esters. Specificity in this case would be
expected, on the basis of analogy, to be only relative and not absolute.
Choline esterase might be expected to split acetylcholine at a higher
rate than other esters, while other esterases might be expected to
behave diff'erently. By testing a number of substrates a pattern has
been established which makes it possible to distinguish choline esterase
from other esterases (17).

343

DAVID NACHMANSOHN

In the great variety of nerve tissue which has been used as basis
for establishing the new concept, the enzyme was found to be an esterase
specific for acetylcholine, viz-, mammalian brain, lobster nerve, squid
fiber containing the giant axon, and the electric tissue. All show a
similar pattern, typical for choline esterase. In contrast, the hy-
drolysis patterns of the esterase of other organs — liver, kidney and
pancreas — differ greatly from that of choline esterase. The esterase
in these tissues shows several variations; but this could be expected,
since the physiological substrate is unknown and probably varies in
the different organs. Only in muscle (free of nerve endings) was an
enzyme obtained whose properties corresponded with those of choline
esterase. It is possible that propagation of an impulse in the muscle
fiber has the same mechanism as in the nerve fiber. But the presence
of choline esterase alone is not sufficient to permit any conclusion.

Of special interest is the pattern obtained with purified choline
esterase. The enzyme extracted from the electric organ of E. electricus
has been purified to such a degree that one milligram of protein splits
three thousand milligrams of acetylcholine per hour. The rates of
hydrolysis of different svibstrates are exactly the same as those ob-
tained with freshly homogenized electric tissue. Thus, the enzyme
tested in fresh electric tissue is the same as that which is highly purified
and the parallelism established between voltage and enzyme activity
becomes particularly significant.

Energy Source

The electric organ also offers suitable material for investigating
the chemical reactions supplying the energy for the nerve action po-
tential. Both the electrical and the chemical energy released are in
the range of possible measurement, whereas, in ordinary nerves, the
methods available are not adequate for quantitative analysis. The
organ of Electrophorus electricus, for reasons discussed elsewhere, is again
particularly favorable for such a study.

Measurements carried out on these fish have revealed some facts
about the chemical source of energy for the action potential. The
electric energy released externally per gram and impulse was found to
be 8 X 10~® gcal. This is the maximum external energy which may
be obtained under the condition that the external resistance is approxi-

344

NERVOUS ACTION

mately equal to the internal. The total electric energy is, in this case,
about six times as high as the external, or about 48 X 10^® gcal. per
impulse per gram electric tissue (4). Under the same conditions and
tested simultaneously, the energy released by the breakdown of phos-
phocreatine was found to be about 32 X 10~^ gcal. per gram and im-
pulse (average of fifteen experiments). Lactic acid formation released
about 17 X 10^* gcal. per gram and impulse, averaging seven experi-
ments (13). The energy of lactic acid formation is probably used to
phosphorylate creatine just as in muscle, where phosphopyruvic acid
transfers its phosphate via adenosine triphosphate to creatine ("Parnas
reaction"). The figures are consistent with the conclusion that
energy-rich phosphate bonds are adequate to account for the energy
of the action potential. Hence, if the primary alterations of the
surface membrane during the passage of the impulse are due to the
release of acetylcholine, the figures suggest that phosphate bonds may
yield the energy for the synthesis of acetylcholine.

The amounts of acetylcholine actually released during a dis-
charge are not known. But the amount which may be split by one
gram of electric tissue during one discharge — about 5 X lO"*^ milli-
mole — may be used as an indication. The amount actually released
may be smaller, since the enzyme may be present in excess, but the
figures indicate the order of magnitude. The amount of phospho-
creatine actually split per gram and impulse is about 3 X 10~^ milli-
mole. Thus, the amounts of acetylcholine and phosphocreatine
metabolized seem to be of the same order of magnitude. Since,
however, one mole of phosphocreatine yields about 10,000 gcal., while
the acetylation of choline requires probably not more than 1500 to
2000 gcal., the fate of the remaining energy has yet to be explained
(11). There are, of course, several conceivable processes which could
account for this diff'erence. For instance, with the splitting of acetyl-
choline, a simultaneous change of a protein molecule could occur, due
to the acid formed, by which the protein is brought close to its iso-
electric point. In such a case, the energy required for bringing the
protein molecule back to its original condition would be of an order of
magnitude similar to that available. But, so far, there is no experi-
mental evidence for this or any other simultaneous reaction.

One of the essential facts supporting the new concept is, as
repeatedly emphasized, the extremely high concentration of choline

345

DAVID NACHMANSOHN

esterase at the neuronal surface, making possible a rate of acetylcholine
metabolism sufficiently high to parallel the electrical changes. In
electric tissue, the rate may be at least 100,000 times, but is probably
close to 1,000,000 times, as high as that of respiration. We must dis-
tinguish, however, between the possible rate and the absolute amounts
metabolized. Acetylcholine is released and hydrolyzed within a very
short period. The actual duration of one thousand discharges is about
three seconds. The recovery may require one to two hours during
which the rate of respiration may be increased. If the absolute
amounts of acetylcholine possibly metabolized are compared with those
of the phosphorylated compounds actually metabolized and the
observed rate of respiration, a satisfactory pictui-e is obtained. Since
this whole chain of reactions connected with the nerve action potential
is initiated by the release of acetylcholine, it has been called the
"acetylcholine cycle."

Choline Acetylase

It appeared essential to test whether or not phosphate bonds are
really the energy source of acetylcholine formation as these investiga-
tions suggest. Evidence for the correctness of this conclusion would
show that the energy of the primary recovery process is really used for
the resynthesis of the compound which, by its release, supposedly ini-
tiates the nerve impulse. It would, therefore, at the same time,
constitute a new support for the assumption that the "excitatory disturb-
ance" (Keith Lucas) which produces a propagated impulse may be
indeed the release of the ester. Also, we would have another example
of the "reconstruction of the chemical events in living cells," to use an
expression of Green (8).

In accordance with the assumption made, a new enzyme,
choline acetylase, could be extracted from brain which, in cell-free
solution, under strictly anaerobic conditions in the presence of adeno-
sine triphosphate, forms acetylcholine (16).

The enzyme may be extracted from homogenized brain. From
one gram of fresh rat or guinea pig brain an enzyme solution may be
prepared which forms 120 to 150 ng. of acetylcholine per hour. The
presence of eserine and fluoride is necessary to inhibit the action of
choline esterase and adenosine triphosphatase, respectively. The
enzyme may also be extracted from tlie powder of acetone-dried brain.

NERVOUS ACTION

In these extracts, the enzyme is about twice as pure as in those obtained
from fresh tissue — one gram of protein may form 3 to 4 mg. of acetyl-
choUne in one liour.

Since acetone inactivates choline esterase, this enzyme is largely
or sometimes completely inactivated in the extracts prepared from
powder of acetone-dried brain, so that addition of eserine may have
either a small effect or practically none on the formation of acetyl-
choline. It is thus demonstrated that the enzyme mechanism re-
sponsible for the formation of the ester is not identical with the hy-
drolyzing enzyme. But this is not surprising in view of the complexity
of the ester synthesis, which probably occurs in several steps involving
more than one enzyme. Adenosine triphosphatase is also removed
in extracts from acetone-dried brain. No addition of fluoride is there-
fore required.

The enzyme requires the presence of potassium in high con-
centration, close to that found in brain. It contains active sulfhydryl
groups which may be easily oxidized in air and which are readily in-
activated by monoiodoacetic acid or copper in low concentrations.
On dialysis, the enzyme rapidly loses its activity. Addition of glutamic
acid reactivates it partly. Only the naturally occurring /(+)-form
is effective (15). With potassium and glutamic acid, 50 to 80% of
the original activity may be restored. Further addition of cyanide or
replacement of glutamic acid by cysteine may reactivate the enzyme
nearly completely. /(+)-Alanine also has some effect; other amino
acids have either a weak effect or none. Citric acid has an effect nearly
as strong as glutamic acid, whereas dicarboxylic acids have practically
no effect.

The effect of glutamic acid, although weaker than that of cys-
teine, appears to be of special interest. Cysteine, like glutathione,
may enhance the activity of many enzymes containing sulfhydryl groups
which have been oxidized during preparation, whereas the effect of
glutamic acid cannot be explained by action on the sulfhydryl groups
of the enzyme.

Price, Waelsch, and Putnam (19) have observed a favorable
effect of glutamic acid on patients suffering from petit mal attacks.
The interest in the effect of glutamic acid on choline acetylase is, by
this clinical observation, further increased, since a relation between
the clinical effect and the enzyme reactivation is easily conceivable.

347

DAVID NACHMANSOHN

The oxidation products of amino acids, i. c, a-keto acids
(pyruvic, phenylpyru\'ic, oxyphenylpyruvic, and a-ketoglutaric acids
have been tested) have a strong inhibitory effect on the formation
of acetylcholine when present in concentrations of 10~' to 10~* molar.
These are close to the concentrations w^hich occur in living cells.
Pyruvic acid is known to be a "physiological anticonvulsant" (20).
The strong inhibitory effect of a-keto acids on the formation of acetyl-
choline is therefore obviously of physiological as well as of clinical
interest.

Action Potential and Inhibition of Choline Esterase

A new and significant relationship has been recently established
between enzyme activity and nerve action potential, this time using the
peripheral axon as material (Bullock, Nachmansohn, and Rothenberg) .
If acetylcholine is the depolarizing agent and if the function of choline
esterase is to remove the active ester so that polarization again becomes
possible after the passage of the impulse, then inhibition of the enzyme
should alter and, in sufficiently high concentration, should abolish the
nerve action potential. Experiments carried out on the giant axon
and on the fin nerve of squid have shown that eserine, known to be a
strong inhibitor of choline esterase, alters and finally abolishes the
nerve action potential. When the nerves are put back into sea water,
they quickly recover and conductivity reappears. The reversibility
of the effect is consistent with the fact that the inhibition of choline
esterase is easily reversible in vitro. Strychnine, another inhibitor of
choline esterase, was also found to alter and, in higher concentrations,
to abolish the nerve action potential reversibly.

Prostigmine has in vitro the same effect as eserine, but it has no
effect on the nerve action potential. Eserine is a tertiary amine and
may therefore, if undissociated, penetrate the lipoid membrane.
Prostigmine is a quaternary ammonium salt and therefore it cannot
penetrate the lipoid membrane. This has been demonstrated experi-
mentally: Eserine was found in the axoplasm of nerves kept in a solu-
tion containing this compound, while prostigmine was absent in the axo-
plasm tested under the same conditions. These observations explain
why prostigmine and acetylcholine, both quaternary ammonium salts,
act externally only on nerve endings which do not have a myelin

NERVOUS ACTION

sheath. But such compounds are inactive if apphed to the axon.
The pecuUarity of the synapse in reacting to injected acetylcholine can
no longer be referred to a difference in the fundamental physicochemical
process underlying the propagation of the nerve impulse, but to the
difference in histological structure. This may also be the explanation
of Claude Bernard's famous curare experiment, since the active prin-
ciple of this compound is, according to recent investigations, a quater-
nary ammonium salt.

Differences between Old and New Approaches

The facts described here may suffice to illustrate the new ap-
proach and may be used for analysis of a few general aspects of the
problem.

Loewi's discovery that a specific compound is released during
nerve activity was important and need not be minimized because of
the change in the original interpretation in the light of recent develop-
ments. Since the adoption of new methods led to.different conclusions,
the question appears of interest whether or not it is possible, with our
present knowledge, to find an explanation for the observations on which
the original conclusions and interpretations were based.

In the older theories, it was assumed that acetylcholine is
liberated at the nerve ending and that, having crossed the synapse or
motor end plate, it acts directly on the effector cell or the second
neuron. Evidence for the role of acetylcholine as such a "neuro-
humoral" or "synaptic" transmitter was considered as satisfactory if the
following three effects could be produced: (1) a stimulating effect by
injected acetylcholine; (2) appearance of acetylcholine in the per-
fusion fluid following nerve stimulation; and (3) the enhancing effect
of eserine on nerve stimulation.

The first effect is not necessarily physiological, but may well be
pharmacological. Other chemical compounds like nicotine or potas-
sium may have a similar effect. This would be, therefore, no conclu-
sive evidence for the assumption that acetylcholine is the mediator in
the original sense. Its appearance in the perfusion fluid following
nerve stimulation is a physiological event; this observation of Loewi
was significant because it connected the choline ester with nerve
activity. The mode of action, however, is not hereby explained.
Any compound forming j^art of an intracellular process may easily

349

DAVID NACHMANSOHN

appear outside the cell even if it is rapidly metabolized, because all
enzymic reactions follow a logarithmic curve. Therefore, a fraction
may persist long enough to escape to the outside. Even if the greatest
part of the acetylcholine released during the passage of the impulse
is split inside the cell, a small fraction may escape hydrolysis and
diffuse to the outside. The appearance of acetylcholine in the per-
fusion fluid does not, therefore, indicate that the compound acts
outside the cell.

Such an assumption would find support if evidence could be
provided that the amount leaving the cell is of an order of magnitude
similar to that which produces a response. This was assumed by
Dale and his associates when they attempted to obtain some quan-
titative estimates for comparing the two amounts. But in the two
cases in which such comparisons were made, a wide gap existed be-
tween the amounts appearing in the perfusion fluid and those necessary
to produce a single response: The acetylcholine found in the per-
fusion fluid is about 1/40,000 of the amount required for stimulation
in the case of the superior cervical ganglion, and in the case of muscle,
only 1/100,000. A difference of such an order of magnitude appears
rather puzzling. It becomes even more difficult to explain when one
considers that even these amounts were collected only if eserine was
present. Choline esterase is present in high concentration outside
as well as inside the cell; evidence for this distribution has been obtained
in experiments on denervated tissue. According to the new concept,
the main function of the choline esterase is to inactivate rapidly the es-
ter inside the cell after its function has been completed, and to restore
the membrane resistance thereby before passage of the next impulse.
The high concentration of the enzyme outside would have the function
of protecting the effector cell against those traces which may
escape and which, if allowed to accumulate, may interfere with the
normal functioning. The enzyme outside would thus act physiologi-
cally as a barrier for the acetylcholine escaping intracellular hydrolysis.
This barrier would be nullified, at least partly, by eserine. Even in
the presence of this drug, the amount collected is small. It is therefore
difficult to imagine that acetylcholine would cross the synapse or end
plate in a concentration sufficiently high to produce a stimulus if the
barrier were fully active. But the small amounts found arc entirely
consistent with the interpretation proposed here.

NERVOUS ACTION

In view of the hi<7h nfrmity of eserine for choline esterase, the
enhancing effect of tiiis compound on nerve stimulation, the third
kind of evidence which was considered as essential, may well be at-
tributed to the inhibition of choline esterase. But here again no
conclusion is possible as to the mode of action of acetylcholine. The
enzyme outside the cell will probably be inhibited; but a fraction of
the inhibitor may enter the cell. In both cases an enhancing effect of
nerve stimulation may result. Such an effect does, therefore, not per-
mit any conclusion as to whether acetylcholine acts physiologically
inside or outside the cell. There may be, also, no enhancing effect at
all, or it may be difficult to obtain, as in the case of the superior cervical
ganglion — a fact which appeared so puzzling to Dale.

We now arrive at a fundamental difference between the implica-
tions based on methods of enzyme chemistry and those based on
pharmacological tests. All facts observed on choline esterase point
unequivocally to the same role of acetylcholine in all nerves. All
contain the same specific enzyme, whether they belong to invertebrates
or vertebrates, to mammals or to fish. All contain a high concentration
of choline esterase whether sympathetic or parasympathetic, auto-
nomic or central nervous system, efferent or afferent. This con-
centration probably varies because the active surface per unit of tis-
sue varies. Other factors may also be important; but the order
of magnitude is always the same and sufficient to permit the assumption
that the substrate is metabolized at a rate parallel to that of the electric
changes, that is, the passage of the impulse. There are other indica-
tions. Wherever tested, a coincidence was found between the time of
appearance of the high enzyme concentration and the beginning of
nervous function during growth. The possibility of applying findings
obtained with the electric organ to the mechanism in brain, as illus-
trated by the discovery of choline acetylase, is another example. So far,
no biochemical fact has been found which is not consistent with the
conclusion that acetylcholine has the same function in all nerves. It is
difficult to conceive that such a specific and well-defined mechanism
either should have a variety of functions in different nerves or, as the
only other alternative, should be active in some, and inactive — although
present — in other nerves.

In striking contrast to the uniformity of the enzymic mecha-
nism, we find a great variety of effects if the pharmacological

DAVID NACHMANSOHN

method is applied. Only in a relatively few cases are the effects com-
parable. This variety is due to the fact that all pharmacological action
depends on a great number of unknown factors. The limitation of an
interpretation of drug effects has been well formulated by Clark (2) :
"Even in the most favorable cases, where quantitative relations have
been established for the action of drugs on cells, there probably remain
dozens of unknown variables, and there is usually a considerable range
of alternative explanations." The contrast between the effects of eser-
ine and prostigmine on the nerve action potential and, more generally
speaking, the great variety of effects obtained with drugs which are
strong inhibitors of choline esterase is an excellent illustration of the
truth of Clark's statement. The affinity for the enzyme is just one fac-
tor since the action depends on many others, like permeability, circu-
lation, concentration, interaction with other cell constituents and
enzymes, and so on. Changes in the chemical constitution of the
molecule are known to influence profoundly the effect of a drug.
Eserine acts mainly on the periphery, and strychnine on the central
nervous system. No explanation is as yet available for this difference
in preference, but it would be a mistake to conclude that there is a
difference in the basic mechanism of a cellular function because a drug
acts differently in two cases.

Most elaborate techniques were necessary to demonstrate the
three effects on which the evidence for the previous concept was based.
But if we keep in mind that the response of the cell to the injection of
acetylcholine is not distinguishable from other pharmacological effects,
that the enhancing effect of eserine is certainly a pharmacological effect,
and that the third observation-^the appearance of acetylcholine in
the perfusate requires the abolition of the physiological barrier by a
drug, then it is apparent why the three phenomena are so difficult to
demonstrate and so subject to variations under different conditions.
In spite of the fact that all biochemical data indicate an identical role
of acetylcholine in all nerves, there is still a great deal of discussion of
the question of whether or not this or that synapse is "cholinergic"
or whether or not acetylcholine is a "synaptic transmitter" in brain
because one or two of the necessary requirements cannot be demon-
strated. The answer does not appear difficult if Clark's statement and
the limitations of the methods used are kept in mind. If we see how
little deviation is found throughout the animal kingdom in the basic

NERVOUS ACTION

niechanisin of muscle function, il is dillicult to conceive that a great
variety of chemical mechanisms are required for the transmission of
the nervous impulse.

There is another aspect which requires some comment. Fol-
lowing the discovery of Loewi that an active compound, the Vagusstojjf,
is released if the vagus is stimulated, it was necessary to identify this
compound. Once its identity with acetylcholine had been established
and its release following nerve stimulation repeatedly demonstrated,
the question arose as to what information could be expected from the
determination of the ester itself and of its concentration in different
nerves or tissues. Acetylcholine is found in almost all tissues, but the
amounts are small, a few micrograms per gram or even a fraction of a
microgram per gram. As previously discussed, acetylcholine is, in
living tissue, a very unstable ester. We do not know what fraction
of the total amount present in the living tissue may be extracted with
our crude chemical methods, especially in nerve tissue, where acetyl-
choline may be so rapidly destroyed. Although the same values will
be obtained using the same conditions, the meaning appears obscure.
A concentration of 0.2 or 0.4 micrograms of acetylcholine per gram of
brain is difficult to interpret, if the same amounts are found in liver,
pancreas, or lung. In the spleen of ox and horse are found amounts
of 4 to 30 micrograms per gram, although there is no indication at
all that the ester has any function in this organ. No acetylcholine has
been found in the spleen of other animals. Chemical compounds may
occur in small amounts everywhere. Acetylcholine has been found
in the potato. The presence or absence of traces of it thus appears
less significant and not comparable to the finding of a mechanism such
as a powerful and specific enzymic system.

The enzyme systems specifically responsible for the formation,
and destruction of acetylcholine are, on the other hand, found mainly
in the nerve cell. Most other tissues do not contain them. But,
besides this localization, the data obtained on the concentration of
choline esterase do make possible the correlation with function. The
amounts of acetylcholine which may be split per gram of brain per
hour vary usually between 50 and 500 milligrams according to species
and center, although they are very constant for each species and each
center. This means that the amount of acetylcholine which may be
split per gram of brain in one millisecond is of the order of magnitude

353

DAVID NACHMANSOHN

of 10^^ to 10'^ molecules. On the assumption that one molecule of
acetylcholine may cover 20 to 50 sq. A., 10 to 100 millions of square
microns of nerve suiface may be covered by the amount which may be
metabolized in one gram of brain in one millisecond. Even if half
of the enzyme were localized outside the cell, the figures are still im-
pressive and suggestive of a relationship between function and quantity
found.

It is of paramount importance for the investigation of a problem
to know the inherent possibilities of methods, as well as their limita-
tions. Medicine depends on the use of drugs. Since it cannot wait
until the mechanism of their action has been explained, drug effects
must be tested on intact cells and intact animals. Such in vivo tests
are necessary because, owing to the great number of unknown factors
in in vivo studies, the effect of a drug is unpredictable merely on the
basis of its affinity for enzymes in vitro and of its possible relation to
cell mechanisms. But if we realize this, then, for the same reason, it
appears inadvisable to draw any conclusions about the physiological
mechanism from drug effects. In any case, it seems difficult to use
effects observed with drugs as an argument against conclusions based
on biochemical data.

It is equally important to know the limitations of enzyme
chemistry. As mentioned above, only if a number of facts or a series
of reactions and some relationship with events in the intact cell have
been established can conclusions as to the mechanism become possible.
In muscle, several physically recorded changes observed during con-
traction could be correlated with chemical reactions. In regard to
the mechanism of nerve activity, the observations on the electric fish
have permitted the correlation of biochemical data with occurrences
observed on the intact animal, and recent experiments on the giant
axon of squid have provided direct evidence for the dependence of the
nerve action potential on the normal function of choline esterase.
Many more are possible even with presently available methods, and it is
reasonable to hope that the continuous improvement of the available
methods and the development of new chemical and physical methods
will open more opportunities for approaching the great number of
unsolved problems. There is not yet evidence for the assumption that
all chemical reactions supplying energy have already been determined,
since the exact amount of heat produced has not yet been correlated

354

NERVOUS ACTION

with the chemical reactions measured. Although the heat production
observed in ordinary nerves is of the same order of magnitude as the
energy evolved during the discharge of the electric fish, direct correla-
tion would be desirable. Another important problem is the nature
of the physicochemical events occurring at the neuronal surface which
change the permeability. At present we do not even know in which
way acetylcholine is released.

It is needless to enumerate all the problems which still remain
unsolved. A great deal of information may still be obtained from
further studies which are being carried out on the enzyme systems
involved. But the use of more methods will be required until a satis-
factory picture is obtained.

Needham (18) stated that "science is the study of the quanti-
tative relationships in the world we live in." Pasteur called Lavoisier
the founder of modern chemistry on the ground that he has introduced
in chemistry the use of the balance, i. e., the notion of quantity. But
since living cells are amazingly complex systems, the interpretation
even of quantitative data requires caution. A great teacher, Sir
Frederick Gowland-Hopkins (9), has warned us: "All dogmatic teach-
ing about any aspect of the phenomena of life is apt to be checked
by the ultimate discovery that the living cell is before all things a here-
tic." For the ultimate goal, which is that of explaining the mechanism
of living cells in terms of physics and chemistry, research based on the
combination of a great number of devices will be necessary. Thus
far, enzyme chemistry has proved itself repeatedly to be one of the most
powerful tools in transforming our approach from pure description of
the phenomena of the living cell into scientific analysis. The history
of the problem of the mechanism of nervous action adds another illus-
tration of this development.

References

(1) Boell, E. F., and Nachinansohn, D., Science, 92, 513 (1940).

(2) Clark, A. J., The Mode of Action of Drugs nn Cells. Arnold, London,
1933.

(3) Cole, K. S., and Curtis, H. T., J. (rcn. PhyswL, 22, 619 (1939).

(4) Cox, R. T., Coalcs, C. W., and Brown, U. V., J. Gen. Physiol., 28, 187
(1945).

355

DAVID NACHMANSOHN

(5) Erlanger, J., and Gasser, H. S., Electric Signs of Nervous Activity. Univ.
Pennsylvania Press, Philadelphia, 1937.

(6) Fulton, J. F., Physiology of the Nervous System. Oxford Univ. Press,
New York, 1938. 2nd ed. rev., 1943.

(7) Fulton, J. F., and Nachmansohn, D., Science, 97, 569 (1943).

(8) Green, D. E., in Perspectives in Biochemistry. Cambridge Univ. Press,
London, 1937.

(9) Hopkins, F. G., Skand. Arch. Physiol., 49, 33 (1926).

(10) Marnay, A., and Nachmansohn, D., J. Physiol., 92, 37 (1938).

(11) Nachmansohn, D., in Harris, R. S., and Thimann, K. V., Vitamins
and Hormones, 3, 337 (1945).

(12) Nachmansohn, D., Cox, R. T., Coates, C. W., and Machado, A. L.,
J. NeurophysioL, 5, 499 (1942).

(13) Nachmansohn, D., Cox, R. T., Coates, C. W., and Machado, A. L.,
J. NeurophysioL, 6, 383 (1943).

(14) Nachmansohn, D., and John, H. G., Proc. Soc. Exptl. Biol. Med., 57, 361
(1944); J. Biol. Chem., 158, 157 (1945).

(15) Nachmansohn, D., John, H. M., and Waelsch, H. T., Biol. Chem.,
150, 485 (1943).

(16) Nachmansohn, D., and Machado, A. L., J. NeurophysioL, 6, 397 (1943).

(17) Nachmansohn, D., and Rothenberg, M. A., Science, 87, 158 (1944);
J. Biol. Chem., 158, 653 (1945).

(18) Needham,J., The Sceptical Biologist. Ghatto & Windus, London, 1929;
Norton, New York, 1930.

(19) Price, J. C, Waelsch, H., and Putnam, T. J., J. Am. Med. Assoc,
122, 1153 (1943).

(20) Putnam, T, J., and Merritt, H. H., Arch. Neurol. Psychiat., 45, 505
(1941).

(21) Schoenheimer, R., and Rittenberg, D., Physiol. Revs., 20, 218 (1940).

35^

23

SOME ASPECTS

OF BIOCHEMICAL

ANTAGONISM

D. VV. WOOLLEY, associate, the rockefeller institute for

MEDICAL RESEARCH, NEW YORK; THE LILLY AWARD IN BACTERIOLOGY;
MEAD JOHNSON AWARD IN NUTRITION

n

|URING the past five years, a series of observations has
been reported which show that compounds having spe-
cific antagonistic action to vitamins and to certain other metaboUtes
can be synthesized or obtained from nature. The uses to which these
materials may and have been put in the study of biochemistry, as well
as their promise in the field of pharmacology, have made it seem de-
sirable to winnow the data thus far obtained and to plant in a promi-
nent place whatever viable seeds may appear.

The agents antagonistic to metabolites may be divided, for
purposes of the present discussion, into two groups. The first contains
those inhibitory compounds which have structures analogous to the
metabolites in question. The second is composed of specific proteins
which react with certain metabolites in such a manner as to render
them biologically inactive in the systems in which they are studied.
The first of these two groups has been examined much more intensively,
and at the moment appears to bear more immediate promise than the
latter.

The work of Woods with //-aminobenzoic acid was the first to
cause widespread interest in compounds related structurally to vita-

357

D. W. WOOLLEY

mins but which were biochemically antagonistic to them. This work
was not the first* to demonstrate such antagonism, but it was the first
in which the relationship was seen in a light suflficiently bright to cast
a shadow ahead. Woods observed that the bacteriostatic action of
the sulfonamides was reversed competitively by />-aminobenzoic acid,
the structural analogue of sulfanilamide. The chemical relationship
here was that sulfanilamide was /)-aminobenzoic acid with a sulfon-
amide group instead of a carboxyl group. The hypothesis was ad-
vanced that the sulfonamides owed their action in inhibiting growth
of bacteria to their competition with /)-aminobenzoic acid in an essen-
tial metabolic reaction. This postulate gained some foundation in
fact when it was shown that /?-aminobenzoic acid occurred in yeast
and other living forms, and, more especially, that it was an essential
growth factor for several species of bacteria.

The discovery of the relationship between /?-aminobenzoic acid
and the sulfonamides prompted the application of similar types of
structural change to other vitamins in efforts to produce from these
vitamins bacteriostatic compounds. It was soon found that 3-pyridine-
sulfonic acid and its amide would inhibit the growth of certain bacteria
in a manner subject to reversal by nicotinic acid. Likewise, thio-
panic acid [pantoyltaurine, jV-(a,7-dihydroxy-/3,/3-dimethylbutyryl)-
taurine] acted competitively with pantothenic acid to produce bac-
teriostasis, and several a-aminosulfonic acids competed with a-amino-
carboxylic acids. (The structural formulas of some of these and suc-
ceeding compounds will be found on pages 371-373.) Thus it began
to appear that the principle of structural analogy was the basis of a
general means of producing bacteriostatic compounds.

The next advance was made when it was found that some types
of structural analogues of various vitamins would cause the appearance
of characteristic signs of vitamin deficiency diseases in animals, and
that these signs could be cured or prevented by adequate doses of the
vitamin involved. There is an excitement and appeal about a spec-
tacular experiment with animals which is never quite equaled by simi-

* The pioneer observation of Quastel and co-workers on the reversible
inhibition of the oxidation of succinate by malonate, and the finding of Woolley
et al. on the toxic effect of 3-pyridinesulfonic acid in nicotinic-acid-deficient dogs
antedated Woods' work; but these two investigations, especially the latter, lacked
sufficient appeal and interest to stimulate further searches,

358

BIOCHEMICAL ANTAGONISM

lar demonstrations in lower forms. Perhaps we feel closer kinship
with the higher phyla, but more probably we sense the wider ranj^e of
observation possible in tests with mammals. Be that as it may,
Woolley and White reported that the feeding of minute amounts of
pyrithiamin to mice caused the appearance of typical signs of thiamin
deficiency in these animals.

When mice are fed a thiamin-deficient diet, tliey do not exhibit most
of the characteristic signs which accompany thiamin deprivation in certain
other species. The fact that pyrithiamin evoked most of these signs may be
of use in the study of manifestations of deficiency in varied species. This idea
is strengthened by the observations that antagonistic analogues of some of the
other vitamins likewise produce typical signs of deficiency in species in which
such signs have not been seen previously.

The disease was prevented or cured by sufficient amounts of
thiamin. Pyrithiamin is the analogue of thiamin in which the thiazole
ring is replaced by a pyridine ring, or, more specifically, the sulfur atom
is replaced by — CH=CH — . It was then shown that glucoascorbic
acid, a structural analogue of ascorbic acid, produced a scurvylikc
disease of rats, mice, and guinea {ligs, and that, in guinea pigs, tlie
disea.se was prevented by adequate amounts of ascorbic acid. Fur-
thermore, signs of riboflavin deficiency were produced m rats by feeding
isoriboflavin, and in mice by feeding 2,4-dinitro-7,8-dimethyl-10-
ribityl-5,10-dihydrophenazine, the phcnazine analogue of riboflavin.
Moreover, manifestations of nicotinic acid deficiency were brought
about in mice by feeding 3-acetylpyridine, and prevented with nico-
tinic acid. Shortly before this, it was realized that the signs pre-
cipitated by 3,3'-methylenebis-[4-hydroxycoumarin] were those of
vitamin K deficiency and were preventable by the vitamin to which
the coumarin bears structural analogy. Finally, several results of
tocopherol and vitamin K deficiencies were produced in mice by
feeding tocopherol quinone.

Interspersed with these observations on animals there have been
a long series of findings of similar nature on the reversible inhibition
of microbial growth. It is not the purpose of this essay to catalogue
the rather impressive number of individual findings in this regard, but
rather to attempt to discover what general principles may lie beneath
them. Therefore, it will suffice to recount that 3-pyridincsuIfonic acid

359

D. W. WOOLLEY

inhibited growth competitively with nicotinic acid, thiopanic acid with
pantothenic acid, pyrithiamin with thiamin, benzimidazole with
adenine or guanine, 2,4-diamino-7,8-dimethyl-10-ribityl-5,10-dihydro-
phenazine and 6,7-dichloro-9-ribitylisoalIoxazine with riboflavin,
and desthiobiotin with biotin. Many of these examples, along with
others, will be brought into the discussion later to illustrate general
deductions. In every instance, the inhibitory action of the analogue
was negated by increasing the amount of the metabolite in the basal
medium. Therefore it could be said that the action of the analogues
was in some degree due to the production of deficiencies of the metabo-
lites. In this respect, however, the bacterial examples lacked some
of the force of the animal demonstrations since, in the latter cases, it
was possible to examine the specific anatomical and physiological
changes characteristic of various vitamin deficiencies which were
produced by the agents.

One feature of the action of antagonistic structural analogues
of metabolites on a wide variety of living things, both animal and
microbial, has been the marked species specificity shown by some of
these agents. With several of the compounds, inhibitory action was
manifest only against the species for which the metabolite concerned
was an essential growth factor. Against those species which syn-
thesized their own supply of the metabolite in question, the analogue
was ineffective. Thus, thiopanic acid was able to inhibit only those
bacteria that required pantothenic acid, and was ineffective in pre-
vention of growth of species not needing this vitamin. A somewhat
similar situation obtained with 3-pyridinesulfonic acid in relation to
propagation of organisms requiring nicotinic acid. With pyrithiamin,
the dependence of inhibition on the type of thiamin requirement was
highly developed. Species which demanded the intact vitamin for
growth were inhibited by minute amounts of pyrithiamin; those
which were content with the pyrimidine portion alone of the vitamin
could be prevented from multiplying only by ten times the amount
which was eff"ective with the former organisms; and those species
which used both pyrimidine and thiazole portions of thiamin needed
a further tenfold increase in the concentration of analogue. Those
species capable of good growth in the absence of thiamin or of its
component parts were quite resistant to any action of pyrithiamin,
and were able to thrive in media containing a million times the con-

360

BIOCHEMICAL ANTAGONISM

centration which was effective against thiamin-rcquiring microorgan-
isms. Another type of manifestation of the correlation of action with
requirement for the metaboUte preformed in the medium was seen
with phenyl pantothenone acting on bacteria and yeasts. This com-
pound is the one obtained by exchanging — CO — CeHs for the COOH
of pantothenic acid. It inhibited the growth of all microbial species
tested irrespective of their pantothenic acid requirements. However,
its effect was reversed by pantothenic acid only in the case of those
organisms which needed the vitamin. For the other species, panto-
thenic acid was without effect in reversing the action of the agent.
An analogous situation has been recorded in animals, where gluco-
ascorbic acid caused scurvylike conditions in rats and mice, species
not requiring dietary ascorbic acid, and in guinea pigs, animals that
do demand a dietary source of the vitamin. Ascorbic acid failed to
cure the condition in the former species, but was quite active in this
respect for the latter.

Thus far, only one instance of the correlation of effectiveness
with requirement for the related vitamin is known in the animal
world. 3-Pyridinesulfonic acid exhibited no action on mice, a species
not needing nicotinic acid in the diet, while it was quite effective in
killing nicotinic-acid-deficient dogs, and harmless in equal doses in
normal dogs.

In contrast to these correlations, several of the metabolite
analogues have been found to be effective against most types of micro-
bial growth regardless of whether or not the related metabolite was a
nutritive essential. This was true of sulfanilamide, the analogue of
/)-aminobenzoic acid, and of benzimidazole, the analogue of the

purines.

These correlations of action with requirement pose an interesting
problem. Why is it that an organism which obtains its vitamins
externally is susceptible to the antagonistic analogue while the one
which produces its metabolite internally is usually resistant? In
other words, why does the ability to synthesize the metabolite make
the organism resistant to the deficiency disease caused by the analogue?
In only a few instances can this be said to result from formation of the
metabolite in amounts sufficient to negate the inhibition. A very small
beginning has been made on this problem with pyrithiamin. It has
been shown that the types of bacteria which were resistant to pyri-

361

D. W. WOOLLEY

thiamin — and these were the ones which synthesized thiamin — con-
tained an enzyme system which actively spHt pyrithiamin to yield
the pyrimidine portion, and, presumably, the pyridine fragment of
the molecule.

Although it might seem that this pyrithiamin-splitting enzyme is
identical with that which synthesizes thiamin, this is probably not so. A
pyrithiamin-resistant strain of yeast developed in the laboratory by subculture
in the presence of the agent required thiamin as a growth factor. Therefore it
is possible to obtain an organism which can split pyrithiamin and yet cannot
synthesize thiamin.

In this case, then, resistance in the thiamin-synthesizing forms
was correlated with the presence of a mechanism for destroying the
antagonist. The possession or lack of such a system for destruction
of the inhibitory analogue cannot be the universal answer to the
problem posed, because it has been shown that in the case of 3-acetyl-
pyridine, an analogue of nicotinic acid, the resistant organisms were
not able to destroy the substance.

Another facet of the problem is the contrast between the action
of some metabolite analogues in microbial species and in animals.
Thus, thiopanic acid and the sulfonamides caused deficiencies of
pantothenic acid and of /?-aminobenzoic acid in bacteria, but were
innocuous in this respect in animals. On the other hand, 3-acetyl-
pyridine caused nicotinic acid deficiency in mice but not in bacteria.
Another nicotinic acid analogue, 3-pyridinesulfonic acid, was eff"ec-
tive in certain bacteria but not in mice. It may be that these phylal
dilTerences represent nothing more than variation in powers of excre-
tion, or it may be that they mirror various degrees of metabolism.

One metabolite may be capable of participation in several
reactions, some of which may function in a given type of organism but
be absent from the metabolic network of others. If an inhibitor inter-
fered with the action of the metabolite in just one process, it would be
expected that it would be effective in only those species which utilized
this particular function of the metabolite.

At about the same time that the facts discussed above were being
established, a second general type of biochemical antagonism was
being brought to light. This latter type dealt with proteins that were
antagonistic to vitamins by virtue of the ability of these proteins to

362

BIOCHEMICAL ANTAGONISM

inactivate the metabolites. The first such protein to be studied was
ascorbic acid oxidase, an enzyme which is present in several plant
tissues and which catalyzed the oxidation, and hence the inactivation,
of ascorbic acid. The next example was the curious protein of egg
white, variously known as avidin or antibiotin, which combined
stoichiometrically with biotin, and thus rendered it ineffective in many
biological systems. This case is noteworthy, since.it involves a non-
enzymic process in which a specific protein is one reactant. Finally,
there was the enzyme system in many aquatic animals which actively
destroyed thiamin by splitting from it the thiazole portion of the
molecule. The suspicion has existed in several laboratories that these
proteins have roles to play in the ordered progress of metabolism,
and that the avidity with which they attack such vital metabolites
may represent only a perversion of their true functions. However,
there are no data to indicate their biological significance, and until
more is learned about them, they cannot enter further into the dis-
cussion of biochemical antagonisms.

With this cursory scanning of the facts in the field of biochemical
antagonism, let us look about in an attempt to discover whatever light
may be shed on theoretical and on practical problems of biological
science. Particularly, let us consider some newly discernible aspects
of the study of metabolic mechanisms, and of the relationship of chemi-
cal structure to biological activity. It will be apparent as we progress
that there are very considerable gaps in our present scanty stock of
knowledge, and that in many cases these breaches are of such magni-
tude that, when they are eventually closed with solid knowledge, the
final picture which will emerge may differ rather considerably from the
one we are about to examine. With a field as new as this one is, there
are sure to be more questions than answers, and the answers available
are more in the nature of hypotheses than of facts. Nevertheless, let
us integrate present knowledge so that we may better be able to advance

in the future.

The first new aspect deals with the use of inhibitory structural
analogues of metabolites in the study of biochemical reactions. In
this aspect, particularly, hope exceeds realization at the present time,
for only local successes have been achieved. The use of specific
inhibition of a given reaction has been very serviceable in unraveling
the threads of many changes which occur. For example, fluoride

D. W. WOOLLEY

and cyanide have figured prominently in the discovery and differentia-
tion of certain enzymic processes. It has seemed probable that, since
the inhibitory structural analogues bring about specifically deficiencies
of various metabolites with which they act competitively, these
analogues might be highly selective inhibitors of an enzyme system
involving the structurally related metabolite. This has proved to be
so in the case of o-aniinobenzylmethylthiazolium chloride, a structural
analogue of thiamin. This compound was an effective inhibitor of
the enzyme found in certain marine animals which cleaves the thiazole
moiety from the vitamin. Likewise, glucoascorbic acid inhibited the
action of ascorbic acid in the oxidation of tyrosine by guinea pig liver.
The oxidation of tyrosine by liver was apparently a process which
involved ascorbic acid, and in this process glucoascorbic acid and its
related metabolite behaved competitively. Furthermore, the action
of cozymase, a derivative of nicotinic acid, was inhibited by 3-pyridine-
sulfonic acid. This effect of the analogue, however, was not specific,
since several other substances of dubious structural relationship to
nicotinic acid gave a similar result. The competitive inhibition by
malonate of oxidative systems involving the four-carbon dicarboxylic
acids was one of the first instances of biochemical antagonism to be
explored, and antedated by several years the observations on sulfanil-
amide and /?-aminobenzoic acid. Finally, utilization of pantothenic
acid by certain bacteria, a reaction which seemed to be enzymic, was
inhibited reversibly by thiopanic acid. Here, the antagonist was
of value in differentiating the reaction which involved pantothenic
acid from glycolysis. Several inhibitors of glycolysis were also in-
hibitors of the utilization of pantothenic acid. Thiopanic acid, how-
ever, prevented the latter without influencing the former, and thus
indicated that the two processes were separable.

The mechanism of action of various drugs is largely unknown.
It may be worth while in our gropings for the desired explanations
to consider some biological effects produced by benzimidazole, an
agent which causes in animals loss of muscular tone and of ability to
respond to stimuli. Also, it prevents the growth of many micro-
organisms. Its growth-inhibiting action is reversed competitively by
adenine, which, in the form of adenylic acid and of adenosine triphos-
phate, is believed to be of considerable importance in muscular con-
traction. Because of these facts, it would seem that the pharmaco-

364

BIOCHEMICAL ANTAGONISM

logical action of benzimidazolc mighl be an expression of its structural
relation to adenine. However, since adenine did not reverse the
effect of benzimidazolc on animals, more concrete evidence will be
necessary before the action of the drug can l)e explained. Eventually,
the modus operandi of some drugs may be viewed merely as the result of
the production of a deficiency of a structurally related metabolite at
the site of action of the drug. * Some drugs now under investigation
or in use can be pictured as structural analogues of certain metabolites.
Furthermore, the effects produced arc not unlike those which might
be expected to result from the removal of the analogous metabolite
from the field of metabolic action. Nevertheless, it must be re-
membered that these are speculations and that they demand experi-
mental proof before they can be accepted.

Further conjectures along this line may be applied in con-
sideration of antagonisms known to exist among various pairs of metab-
olites. For example, it has been shown that androgens and estrogens
may occur together in both sexes, and that pharmacologically they
exhibit certain antagonisms. Perhaps these are the result of the
structural relationship between these htjrmones.

Although we may think that antagonisms between metabolites
and their structural analogues pojiped into the world with the advent
of the nutritionist and the organic chemist, this is probably not so.
For example, Chromobacterium iodinum has been found to produce a
pigment which is bacteriostatic and is a structural relative of some
naturally occuring anthraquinones, and more remotely of vitamin K.
These anthraquinones, as well as vitamin K, reverse the action of
the antibiotic pigment. Similarly, 3,3'-methylenebis-[4-hydroxy-
coumarin], a structural analogue which elicits some of the signs of
vitamin K deficiency, was first discovered in spoiled sweet clover hay.

At this point it is well to contemplate some hypotheses re-
garding the mechanism of action of the sulfonamides in bacteriostasis,
and these same hypotheses as they are applied to inhibitory analogues
in general. After the discovery of the reversal of sulfonamide bac-
teriostasis by /;-aminobenzoic acid, the hypothesis was advanced that
the sulfonamides competed with />-aminobenzoic acid for a place on

* It is, of course, not expected that all types of drug action will find ex-
planation in biochemical antagonism. 1 1 is merely suggested that some pharma-
cological agenls may owe their biological potency to this phenomenon.

3^5

D. W. WOOLLEY

the receptor portion of an enzyme. The enzyme was supposed to
require />-aminobenzoic acid as a coenzyme, and the introduction of a
sulfonamide was supposed to result in a firm combination of the latter
with that portion of the enzyme to which /^-aminobenzoic acid normally
would be attached. The resultant foiling of the enzyme was said to
be overcome by an increase in the concentration of /?-aminobenzoic
acid which would shift the equilibrium in favor of the j&-aminobenzoic
acid-enzyme combination. Despite the fact that there has been no
direct experimental proof of an enzyme system involving p-amino-
benzoic acid, this attractive hypothesis has continued to flourish with
but slight opposition. Its chief buttresses have been the competitive
nature of the relationship between /^-aminobenzoic acid and the
sulfonamides, the success realized in application to other metabolites
of the same idea of structurally related inhibitors, and the lack of a
better hypothesis. A similar explanation has been given for the be-
havior of other inhibitory analogues. Since some of the metabolites
concerned with these antagonists are known to function as coenzymes
in well-defined enzyme systems it would seem advisable to put the
hypothesis in these instances to experimental test. There are some
shreds of evidence that are exceedingly difficult to fit into the original
postulate. For example, it has been noted with the sulfonamides,
with pyrithiamin, and with benzimidazole that subinhibitory quan-
tities of the compounds actually stimulated the growth of micro-
organisms. If the coenzyme replacement hypothesis is correct, it is
difficult to understand this stimulation. Furthermore, when resting
bacterial cells were treated with sulfonamides or with thiopanic acid,
no /?-aminobenzoic acid or pantothenic acid was liberated. This
observation, however, does not overthrow the hypothesis because the
amounts of metabolite liberated (either /j-aminobenzoic acid or panto-
thenic acid) may have been below the detectable amount, or may
have been retained inside the cells.

Much has been written about whether the antagonism with
/)-aminobenzoic acid explains the action of the sulfonamides in the
production of bacteriostasis and in the cure of certain infectious
diseases. Whatever the true explanation of the action may be, the
following must not be forgotten. First, /?-aminobenzoic acid reverses
competitively the effect of the sulfonamides, and second, j&-amino-
benzoic acid is a metabolite for microorganisms. In the light of these

366

BIOCHEMICAL ANTAGONISM

facts, it is reasonable to assume that the sulfonamides produce a
crippling deficiency of /)-aminobenzoic acid in the organisms. This
view is strengthened when one recalls that structural analogues of
certain other vitamins and hormones call forth in animals certain
signs associated with specific deficiency diseases. It cannot be said
that such antagonisms in any way lend proof to the hypotheses which
have been advanced to explain the action. Nevertheless, it is not
advisable to allow arguments about mechanism to obscure the well-
established facts in the case.

Finally, to end our consideration of the first aspect, let us note
examples of some biochemical reactions which have been uncovered
by means of microorganisms rendered resistant to the action of an
inhibitory analogue. By long-continued culture of a bacterial or
fungal species in the presence of gradually increasing concentrations of
an inhibitor it is possible to derive a strain which is resistant, or fast,
to the agent. By use of a strain of Endomyces vernalis made fast to
pyrithiamin, it was possible to show that the acquired resistance was
correlated with the appearance of a system which cleaved the inhibitor
into its component pyrimidine and pyridine portions. This demon-
stration prompted examination of naturally resistant species, that is,
those which synthesized thiamin. These forms were found to possess
the pyrithiamin-destroying system. Without the use of the "fast"
variant, the recognition of this biochemical reaction would have been
more difficult.

The second new aspect to open before us is the possibility of
producing new types of drugs by application of the knowledge that
pharmacological signs of a predictable nature may be produced by
metabolite analogues. The cataloguing of the various specific histo-
logical and biochemical manifestations of deficiency of each of the
vitamins and hormones is progressing rapidly. It may be desirable
to evoke one or more of these signs. If this should be so, then the
synthesis and trial of an inhibitory analogue of the metabolite in
question would be indicated. Already it has been seen that pharmaco-
logical manifestations predictable at least in part can be called forth
by the proper metabolite analogue. In this connection, a retrospec-
tive glance at the case of 3,3'-methylenebis-[4-hydroxycoumarin],
may be profitable. This agent produces signs similar to those seen
in vitamin K deficiency, and these manifestations form the basis of

D. W. WOOLLEY

Studies carried out with it. The compound was not discovered be-
cause of its structural similarity to vitamin K— only after its pharma-
cological properties were recognized was its structural similarity
to vitamin K seen, and its reversal by the vitamin demonstrated.
With this glance backward to fortify us, let us look further in this
direction.

The study of a-tocopherol quinone represents the next step along
the path. Tocopherol deficiency, at least in the first generation of
rats or mice, manifests itself in females, to a large degree, by resorption
of embryos during gestation. Other signs are generally not seen.
Therefore, it was thought that the development of a successful antago-
nistic analogue of tocopherol would serve as an example of what might
be done along lines of application to pharmacology of studies in bio-
chemical antagonisms. In due course it was demonstrated that a-
tocopherol quinone in sufficient doses caused hemorrhage and resorp-
tion of the embryos of pregnant mice.

Despite the fact that results siinilar to those of tocopherol deficiency
were achieved by administration of tocopherol quinone, tocopherol even in
large doses did not reverse these signs. It was of considerable interest to find
that vitamin K in very small quantities would prevent the vaginal hemor-
rhages which immediately preceded resorption. When these hemorrhages
were eliminated by vitamin K, the pharmacological effectiveness of the
quinone was markedly reduced. It was then realized that tocopherol quinone
is a structural analogue of vitamin K as well as of tocopherol. The action of
the quinone, however, differed from that of such an antivitamin K as 3,3'-
methylenebis-[4-hydroxycoumarin] in its selective action on pregnant animals
and in the fact that the coumarin did not produce similar signs.

When similar quantities of a-tocopherol quinone were given to
nonpregnant mice, no detectable signs of disease were observed. Al-
though these steps have been made along the path under considera-
tion, the final one has not been taken — the production of a thera-
peutically useful agent.

One of the features of the action of tocopherol quinone should be
discussed since it illustrates a general principle in the use of inhibitory
analogues related structurally to metabolites. Almost without excep-
tion the effective dose of an antagonist is very large compared with the
amount of metabolite involved. In other words, the ratio of inhibitor

368

BIOCHEMICAL ANTAGONISM

to metabolite is generally large.* Therefore, the required close of an
inhibitor depends on two things: tlie intrinsic activity of the metabo-
lite, and the ratio between metabolite and inhibitor which is neces-
sary for reversal. This latter ratio may vary widely among species
for the same inhibitor-metabolite pair. Now, tocopherol is a rela-
tively inactive vitamin as vitamins go. The effective dose in a rat
is about three milligrams as compared with a small per cent of this
amount for several other vitamins. Hence it was foreseen that very
large doses of tocopherol quinone would probably be required to
produce the effect; and this proved to be the case.

Antagonistic analogues of hormones may eventually find
practical application. It is now believed that certain diseases result
from overproduction, or diminished rate of destruction, of various
hormones. A possible method of treating such diseases may be by
administering an inhibitory analogue of the substance concerned.
The analogue would, in effect, remove the excessive amounts of the
hormone. A few such inhibitory analogues have already been pro-
duced, but the hormones to which they are related have not yet been
implicated as causative agents of disease as a result of overproduction
in the organism. It has therefore not been possible to test this in-
triguing postulate.

Because it had its beginnings in investigations of the action of
sulfonamides, work on inhibitory compounds related structurally to
metabolites has had a strong flavor of bacteriostasis. This has been
increased by the fact that a most fruitful method of determining whether
a given analogue behaves antagonistically with its related metabolite
has been to make observations on the effect on growth of microbial
species. Most of the early attention was directed toward the develop-
ment of chemotherapeutic agents which would be effective against
infectious diseases. Mcllwain and Hawking were able to show that
this method of approach held promise when they demonstrated that
thiopanic acid could prevent experimental infection of rats with
hemolytic streptococci. However, the amounts of the agent required

* Thus far, only in the cases of benzimidazole and of 3,3'-methylenebis-
[4-hydroxycoumarin] has the ratio been 1 : 1 or less. In these two exceptions the
quantity of inhibitor required for an effect was large, but once an adequate level
of the agent was established the ratio between inhibitor and metabolite was small.

D. W. WOOLLEY

were so large as to make its clinical application impractical. It is
possible that useful agents against infectious diseases may be produced
by investigation of inhibitory analogues, but at present this has not
been done. Meanwhile, it is well to remember that anti-infection
agents form only a part of the chemotherapeutic arsenal, and that
magic bullets against noninfectious diseases as well as against infections
may eventually be cast in the mold before us.

The third new aspect to be considered is an outgrowth of the
previous two general aspects. If it is possible to evoke the various
types of manifestations discussed, then it is well to examine the classes
of structural changes that must be made in a metabolite molecule in
order that it may exhibit inhibitory action. At the present time, two
general types of structural change which bring about this result can
be seen. In addition, there is a third group of structural alterations
about which it is not yet possible to make generalizations.

The first general type of structural change involves the replace-
ment of the carboxyl group of acidic metabolites by some other group.
The most frequently studied exchange has been that of sulfonic acid or
amide for the carboxyl. Examples of this class are: sulfanilamide and
its derivatives which are related to /^-aminobenzoic acid; S-pyridinc-
sulfonic acid, related to nicotinic acid; the a-aminosulfonic acids,
related to the a-aminocarboxylic acids; and thiopanic acid, related
to pantothenic acid. With the possible exception of 3-pyridine-
sulfonic acid, members of this class of inhibitors are not effective in
the production of deficiencies in animals. Furthermore, it is from this
class that many useful agents against infectious diseases, that is, the
sulfonamides and thiopanic acid, have come. Possibly the inactivity
toward animals contributes to their chemotherapeutic effectiveness,
for it may result in damage to the invading microorganisms without
causing undue violence to the host.

Inhibitory analogues have also been produced by replacement
of — COOH by — COR, but the class of compounds so derived has
not been studied as intensively as that just examined. Thus sub-
stitution of the — COOH of j&-aminobenzoic acid by — COCH3 (to
yield /'-aminoacetophenone) led to the formation of a bacteriostatic
substance whose action was reversible by /?-aminobenzoic acid. A
related ketone protected mice against experimental infections. Simi-
larly the exchange of — COOH in nicotinic acid for — COCH3 to

BIOCHEMICAL ANTAGONISM

Compounds of General Type I

A

CHjOH

NF

h

/\

— SO3H

CH,-

-C— CH,

CHOH

1

c— 0

1

V

S02NH2

Sulfanilamide

3-Pyridine3ulfonic
acid

NH
1
CH,

1
CH,

SO3H

Thiopanic acid

NH,

COOH

p-Aminobenzoic
acid

Class B

NH,

I

c=o

I
CH3

Aminoacetophenone

/\-

-COOH

Nicorinic acid

/\

o

— C— CH,

\N/

3-Acetylpyridine

CH2OH

I

CH3 — C — CH I
CHOH

c=o

I
NH

I
CH,

I
CH,

COOH

Pantodienic acid

CH2OH

I

CH 3 — C — CH ,

I
CHOH

1

c=o

I

NH

I
CH,

I
CH,

I

c=o

I

CjHj

Phenylpantothenone

371

0-

c

>

SB

V 3

C4

0 l-s

o

0 >ig

0-

<

— \ '^^
0 "^^

>

1

'>>

X a
^ 6S

X
0

X
0

«

K

1 hS' H'n

/\ II

o-

-o-

-u-
X

CO f-s 1

11 .2-0

X

X

a

C_>l

CL4

X
0-

^

!k

H

n

X /W'' ^

a

u

«1

O

1 " -5

XUXxt

o

X \ z<^

TJ

o— o— u

a

1 1

s

^,—u—z

a

1

a

O

u

O

X t

s

<

>o|

c

>

o

X

X

X ffi

/"

a— o— u— ^

>

o

J3

X
u

X'Z ^

I I

a

3

a
o

a

'e

<

M <M

I^ H! K
O— U— O— O-

-c/:!

U— O UE

'3|

o=u— o

I II

X

o-

I

o ^

% X

\x

c
o
c

'5

-<CJ>-

^ o

X

^ 8

o
H

H

u

C
w

o

'0

a

3

o
S

o

XX XX

O O E 0 O

111,1

\ X X

— 0 — -J

0:3
K 8

o
u

3

5

I X
O^ ^. /" J

O O

ffi OS

0 0

Vl— u=o— o— o— u I

X I

o

D. W. WOOLLEY

yield 3-acetylpyridine gave rise to a substance which caused nicotinic
acid deficiency in animals but not in microorganisms. Furthermore,
substitution of — CO — CeHs for the — COOH of pantothenic acid gave
rise to phenyl pantothenone which caused pantothenic acid deficiency.
The introduction of — COCH3 rather than — COCeHg did not yield a
compound that produced unequivocal pantothenic acid deficiency,
perhaps because a sufficiently negative ketone was not exchanged for
the carboxyl. In the cases of nicotinic acid and /?-aminobenzoic acid,
the carboxyl is attached to an aromatic type of nucleus, and therefore
the related ketone is more acidic than if this nucleus were aliphatic.
On the other hand, in pantothenic acid the carboxyl is aliphatic and
may require replacement by an aromatic ketone group.

The second general method for production of inhibitory com-
pounds consists of an exchange of one or more atoms in a ring system.
Since so many substances of biological importance are ring com-
pounds the method is rather inviting and has proved to be quite fruit-
ful. The first representative of the series to be studied was pyrithi-
amin, in which the group • — CH==CH — replaced the sulfur atom of
thiamin. Other examples are: benzimidazole, in which carbon
atoms in a benzene ring replace the nitrogen atoms of the structurally
related purines; 2,4-diamino-7,8-dimethyl-10-ribityl-5,10-dihydro-
phenazine, similarly related to riboflavin; triazolopyrimidines, in
which a nitrogen atom replaces a carbon atom of the imidazole ring
of the purines; benzoureidovaleric acid, in which a benzene ring
replaces the thiophane ring of biotin; o-aminobenzylmethylthia-
zolium chloride, in which a benzene nucleus is substituted for the
pyrimidine nucleus of thiamin; 3,3'-methylenebis-[4-hydroxy-
coumarin], derived from vitamin K by exchange of a carbon for an
oxygen atom along with changes in the side chains; and iodinin in
which two nitrogen atoms have replaced carbon atoms in the ring
system of the related anthra- and naphthoquinones. Here, as with the
first general type of structural modification, the method has worked
apparently in every case in which it has been tried. It will really be
remarkable if this record continues to stand in the face of more varied
application.

The third, or miscellaneous, group of types of structural change
is the scrap heap in which all atypical cases must be collected until a
general principle can be seen, and rescued from among them. To

374

BIOCHEMICAL ANTAGONISM

this group belong glucoascorbic acid, a-locophcrol quinone, and des-
thiobiotin, to mention only a few prominent members.

At the present very early stage of our knowledge, there arc
many questions demanding answers; and for many of these adequate
ones are not yet available. Let us examine two. From the examples
now at hand it appears that the exchange of carbon atoms for nitrogen
in six-membered rings leads to the production of inhibitory com-
pounds. Witness the effectiveness of conversion of pyrimidine de-
rivatives to benzene compounds. It would be well to know whether
the process is unilateral, that is, whether the exchange of nitrogen for
carbon would be equally as successful as the reverse process. From
the cases of iodinin, and of benzimidazole and triazolopyrimidine (see
next paragraph), and from the fact that 6-aminonicotinic acid behaves
competitively with /?-aminobenzoic acid, it would seem that the
process actually is not unilateral.

Take now the question of how great the change in structure
must be to attain maximal activity in the inhibitory analogue. Is it
preferable to make as small a change as possible, and if so, what is the
measure of the sizfe of the change which is made? If one examines a
case in which progressive change has been made it would seem that a
small alteration is better than a larger one, but this cannot be stated
with any assurance because of the lack of a standard of magnitude of
change. Benzimidazole is derived from the purines, adenine and
guanine, by elimination of the nuclear substituents, and by exchange
of the two pyrimidine nitrogen atoms for carbon. It is less active in
the inhibition of bacterial growth than are the triazolopyrimidines,
which differ from the purines only in that the imidazole carbon of the
latter has been replaced by a nitrogen atom. Thus, the exchange of
one carbon atom for a nitrogen gives a more potent compound than
was obtained by trading two nitrogen for two carbon atoms.

Of course, the two exchanges are not strictly comparable because in
one case the imidazole ring is altered and in the other the pyrimidine part is
transformed. It will be necessary to examine the activity of pyridinoimidazole
and its amino and hydroxy derivatives before unequivocal conclusions may be
made about the matter. A more unambiguous illustration deals with deriva-
tives of /?-aminobenzoic acid. Replacement of one carbon atom with nitrogen
yields 6-aminonicotinic acid which is antagonistic, while exchange of two
carbon for two nitrogen atoms to form 2-amino-5-carboxypyrimidine gives

375

D. W. WOOLLEY

rise to an analogue with no inhibitory powers but rather with weak /'-amino-
benzoic acid action.

Attendant on this increase in activity is an increase in speci-
ficity. Benzimidazole is reversed in its action by either adenine or
guanine. By contrast, hydroxyaminotriazolopyrimidine competes
only with guanine, and aminotriazolopyrimidine interferes only with
adenine. However, this increased specificity may not always be
desirable, for benzimidazole has pharmacological powers not possessed
by the triazolopyrimidines. Furthermore, the effects of the side chains
are of interest. In the case of benzimidazole it was thought that the
introduction of an amino group in the proper position would make the
compound more analogous to adenine, and hence more potent; but
the introduction of this side chain did not materially alter the potency
of the analogue. In the case of 3,3'-methylenebis-[4-hydroxycou-
marin] it would be of interest to learn the effect of the side chains on
biological action, that is, whether the introduction of a phytyl group
in position 3 (which would make the coumarin more analogous to
vitamin K) would increase its potency or modify its action.

It must not be assumed that it is only necessary to alter the
structure of a metabolite indiscriminately in order to achieve an
antagonistic agent. Much testing of compounds related structurally
to biologically active compounds has indicated that most of these
derivatives are inactive under the conditions of test.

On the other hand, there is no unique manner in which the
structure must be altered in order to produce antagonistic agents.
For example, 2,4-diamino-7,8-dimethyl-10-ribityl-5,10-dihydrophen-
azine, 6,7-dichloro-9-ribitylisoalloxazine, and 5,6-dimethyl-9-ribityl-
isoalloxazine, three analogues of riboflavin, all cause riboflavin defi-
ciency in various species even though the type of alteration in struc-
ture represented by the three compounds is fundamentally differ-
ent. Likewise, the structure of />-aminobenzoic acid may be altered
by changes either in the nucleus or in the nuclear substituents, with
resultant production of substances competing with />-aminobenzoic
acid in bacterial growth. It is of interest to note that only a few of these
derivatives are active against infections. Similarly, 3-pyTidinesulfonic
acid and 3-acetylpyridine represent two inhibitory compounds related
to nicotinic acid. Here too, as discussed previously, there are very

BIOCHEMICAL ANTAGONISM

marked differences qualitatively in the effects produced by the two
agents.

In conclusion it should be said that this essay is intended as a
guidepost set beside a new and dim trail to point the direction until
the road becomes a much traveled highway. Then a more preten-
tious marker may be raised and some of the errors of direction and
many of the convolutions of the trail may be obliterated. From the
work which has been done it seems that a study of antagonisms may be
a real current in biochemistry, and not merely an eddy.

Finally, since the present attempt was meant more as an essay
and less as a review, no literature has been cited. The reader is re-
ferred to references (2) and (7) for documentation of some of the points
discussed.

References

(1) Fildes, P., "The mechanism of the anti-bacterial action of mercury,"
Brit. J. Exptl. Path., 21, 67 (1940).

(2) McIIwain, H., "Theoretical aspects of bacterial chemotherapy," Biol.
Rev. Cambridge Phil. Soc, 19, 135 (1944).

(3) McIIwain, H., and Hughes, D. W., "Biochemical characterisation of
the actions of chemo therapeutic agents. II. A reaction of haemolytic
streptococci, involving pantothcnate-usage, inhibited by pantoyltaurine, and
associated with carbohydrate metabolism," Biochem. J., 38, 187 (1944).

(4) "Modes of drug action," a symposium in Trans. Faraday Soc, 39, 319
(1943).

(5) Quastel, J. H., and Woolridge, W. R., "Some properties of the dehy-
drogenating enzymes of bacteria," Biochem. J., 22, 689 (1928).

(6) Woods, D. D., "The relation of /)-aminobenzoic acid to the mecha-
nism of the action of sulphonamide," Brit. J. Exptl. Path., 21, 74 (1940).

(7) WooUey, D. W., "Some new aspects of the relationship of chemical
structure to biological activity," Science, 100, 579 (1944).

(8) Woolley, D. W., "Development of resistance to pyrithiamine in yeast
and some observations on its nature," Proc. Soc. Exptl. Biol. Med., 55, 179

(1944).

(9) Woolley, D. W., "Some biological effects produced by a-tocopherol
quinone," J. Biol. Chem., 159, 59 (1945).

377

24

CHEMOTHERAPY
APPLIED CYTO-
CHEMISTRY

ROLLIN D. HOTCHKISS,* associate, the rockefeller institute

FOR MEDICAL RESEARCH

THE FUNDAMENTAL procedure for determining the
properties of any object is to place that object in a series
of new environments and to look for changes produced thereby in
it or its surroundings. This procedure is a part of the innate be-
havior of primitive man — whether he be ancient aborigine or modern
infant — and is used, for example, in his examination of a new bauble.
The object is picked up, rubbed with the hand to feel its texture,
tested perhaps in the mouth, and withal scrutinized minutely in
various lights from various angles. A mentally competent primitive
soon builds up an enormous "library" of impressions of texture and
light-reflecting properties, etc., of the objects of his world, and uses
these to guide his activity. The steady accumulation — and, let us be
sure to add, the sharing — of these impressions has in time developed
large specialized bodies of scientific "knowledge." One of these,
more or less arbitrarily defined, may be called cytochemistry. Its
basic technique is "looking at" the cells of living matter with such

* On leave of absence, Lieutenant Commander in the Hospital Corps of
the U. S. Naval Reserve, Hospital of the Rockefeller Institute.

The views and opinions expressed here are those of the writer, and are not
to be construed as reflecting the official views or opinions of the Navy Department.

379

R. D. HOTCHKISS

agents as light or electron beams, electric currents, and especially
chemicals or enzymes, and noting changes in the cell or the environ-
ment caused by whatever interaction takes place.

Chemotherapy must have had its humble inception in self-
medication, the observation of the effect of natural herbs, plants,
juices, etc., upon the sick organism. During the course of time this
growing branch of medicine has come to make extensive use of the
knowledge accumulated by the pharmacologist, and of the materials
isolated from nature or synthesized by the chemist. In its present
phase, chemotherapeutic medicine is beginning to draw spasmodically
upon the pure science of biological chemistry. Despite the commer-
cial significance of chemotherapy, we are hardly justified in speaking
of a science of "theoretical chemotherapeutics." We shall entertain
here the somewhat limited view that chemotherapeutic medicine is an
applied cytochemistry. In so doing, we shall be able to look into
the future of what are believed to be some current trends, and to
consider some of the challenges that face the workers, and their teach-
ers, in this important field.

Perhaps the bright future of cytochemistry may be the better
envisioned if a comparison is made between this science and its more
established predecessor, organic chemistry. During the last century,
the organic chemist has been systematizing his experiences with the
effect of chemical reagents upon various complex molecules. He has
learned to detect atom groups which react together as units, to look
upon one of these groups as the unifying backbone or nucleus of the
molecule, and to locate the position of other groups relative to this
one. With the recent confirmation of many of his deductions by
physical methods, skepticism as to the general validity of his indirect
methods appears to have vanished. In a quite analogous way, the
cytochemist is beginning to study the way in which whole molecules
and layers of molecules are distributed in the living cell. It may be
said to be his task to be aware of the constituents of cells revealed by
chemical, physical, metabolic, immunological, pharmacological, or
nutritional analysis, and to learn to recognize these constituents, and
the evidences of their functioning, in the intact cell. Cytochemistry is
indeed the area of common interest which eventually will unite bio-
chemistry with biology in general.

Since most cells are large enough to reveal detail in light, or

380

CHEMOTHERAPY

electron microscopy, much has been done with stains and fixatives,
and we are able to assign generic names like protein, fat, carbohydrate,
and nucleoprotein to some cell structures. But enzymes, coenzymes,
metabolites, antigens, toxins, and such constituents of molecular size
are recognized by functional, rather than visual, attributes. Where
any means of identifying one of them in the anatomy of the cell is
available, it is likely to be an indirect method subject to the same
question and criticism as were the methods of early organic chemical
science. A few specific constituents of functional interest, like desoxy-
ribosenucleic acid, ribosenucleic acid, phosphatase, glycogen, cellulose,
heparin, etc., can now be recognized by staining methods when they
are localized in large amounts in some structure of the cell. Never-
theless, we seldom know both the function, in chemical terms, and the
site, of any cell constituent. And even when we do know the chemical
potentialities of some unseen constituent, we rarely are able to say
when in the life of the cell it is acting and when its activity is latent.

For information about the cell we can look to the stains, fixa-
tives, antiseptics, hypnotics, anesthetics, insecticides, hormones, meta-
bolic poisons, growth factors, etc. In short, we will eventually draw
upon every situation in which a substance of known constitution pro-
duces observable modifications in cells or cell activities. It is quite
natural that cytochemical reagents should all be, in varying degrees,
toxic substances. Before they can claim interest as therapeutic agents,
however, they must prove to be of less than general toxicity, and
capable of administration in a form not too easily hydrolyzed, de-
stroyed, diff'used away, precipitated out, excreted, or otherwise re-
moved from the site of activity in the body. The pharmacist and
pharmacologist have become so skilled in controlling and modifying
these last aspects of drug action that, in a sense, we shall be justified
in concentrating our attention upon the first-mentioned factor, cyto-
chemical specificity.

Ideally, chemotherapeutic agents should have maximal eff'ects
upon a chosen variety of cells (a parasitic foreign agent, or a diseased,
neoplastic, hyperactive, or otherwise abnormal host tissue) and a
minimal eff"ect upon other tissue cells of the organism. For this to
be possible it is necessary to capitalize upon structural difi'ercnces
between the cells at one of three levels of organization, the morpho-
logical, the colloidal, or the chemical level. This is now necessarily

381

R. D. HOTCHKISS

done blindly, and it is possible that it may always have to be done
empirically. For, on the one hand, although many visible sti'uctural
differences have been described between different cells, these can
hardly ever be described in terms of function, chemical reactivity or
vulnerability. On the other hand, it is safe to say that those few
chemical differences which have been detected probably reside in the
proteins, polysaccharides, or other macromolecules of the cells. Almost
the only agents with specificity great enough to show qualitatively
different behavior with different cells are the antibody and enzyme
proteins or organized agents like the viruses. Less complex chemical
agents sometimes behave differently in a quantitative sense in staining
or penetrating diverse cells, but even here we seldom find the desired
situation, viz-, that the reactivity of one type of cell far overshadows
that of any other type.

Chemical substances of low molecular weight are, however, not
devoid of specific affinities and are even capable of "biological" dis-
crimination. Enzymes are enzymes only if there are corresponding
substrates, and these latter are in many cases well-defined simple
substances. If the enzyme is credited with specificity in its action
upon substrates, it is perhaps only fair to ascribe specificity to the
substrate when it seeks out an active spot on the enzyme surface and
enters there into a transitory enzyme-substrate complex. Statistically
speaking, it is the rapidly moving substrate molecule and not the slug-
gish protein which does most of the hunting. However, we must
remember that there is virtually no evidence to indicate how many
"false complexes" the substrate enters into, at other sites on the enzyme
or other proteins, which do not result in chemical change, and there-
fore pass unnoticed. We cannot say that a key which fits many locks
is specific merely because most of the locks it fits will not work.

With antibody proteins, too, simple substances have a discrimi-
natory ability to inhibit the combination with, and precipitation of,
the specific antigens. Indeed large parts of the precipitating capacity
of the antibody proteins can be directed toward relatively simple
associations of chemical groups contained in the huge antigen particle.
Nevertheless, it is just when we turn to the smaller molecules and to
the phenomena of "specific inhibition" which they show that we begin
to find the broadest overlapping of reactivity shown by distantly
related structures. For example, an antibody can be sometimes more

382

CHEMOTHERAPY

interested in whether a group is on the ortho, meta, or para position in a
simple benzene derivative than whether that group is an amino, a
methyl, chloro, hydroxy, or nitro group.

As a third illustration of specificity shown by simple substances
let us consider crystallization from solution. In general, this process
involves a truly remarkably specific selection of particles of a single
molecular species, occurring as rapidly as diffusion processes will
permit. The molecule of appropriate configuration finds a place
v/here it precisely fits, and kinetic energy of motion becomes trans-
formed into potential energy of position. Symmetrical molecules
which can "sit down" in several ways {e. g., para-disubstituted benzene
compounds) find this easier to do and crystallize more readily than
unsymmetrical molecules; and their crystals are less soluble and
higher in melting point. Ions which have affinity for each other are
able to react if by reacting they may crystallize out as an insoluble
compound.

It may be worth some consideration that the demonstrations of
specificity in simple compounds just cited all involve as the specific act
the immobilization and orientation of the substance upon a surface,
of either a protein molecule or a crystal. We must recall that the
energy content of small molecules requires them to vibrate and move
about aimlessly in an activity far more rapid and extensive than that
of the protein molecules and crystal particles, which move with a
sluggish and restricted motion more like the Brownian movement.
We may picture the larger particle as furnishing a definite mold or
pattern into which the small particle in its bouncing, careening prog-
ress makes countless attempts to fit. When, by chance, there is close
approach at appropriate orientation to some matching area on the
surface, a specific adsorptive act occurs, the duration and outcome of
which depend upon intrinsic properties of the molecules. It may well
be promptly reversed and thus pass unnoticed in the confusion of
molecular traffic, or it may result in enzymic action, or inhibition
thereof, in precipitation or its inhibition, etc. We know biochemical
events to be replete with interactions of this general nature, and it
seems possible that specificity chiefly depends upon the existence of
anchoring areas on the surfaces of sluggishly moving large molecules
and cells. To learn the characteristics of their specificity we need to
know as much as possible about the way they are aflfected by all kinds

383

R. D. HOTCHKISS

of substances, whether or not these latter are cUnically useful drugs.
We must expect to find that, in intact cells, where diffusion rates, di-
electric constants, dissociation constants, and the like may be very
different from what they are in ordinary aqueous solutions, the speci-
ficities of biochemical systems also will not be the same as in the
test tube. The organizational factors which control the various
phases of chemical activity in the cell probably grade continuously
from slight "crowding" effects of molecule upon molecule which slightly
alter diffusion and reaction rates, through major crowding which in
effect makes the environment only semiaqueous, and finally all the
way to those special arrangements which amount to actual contiguity,
or separation, in space, of reactive elements.

We have attempted to portray as concretely as we are able a
general picture of the initial process of drug-protoplasm interaction.
It is, more or less, a simple expression in present-day physical terms
of views on the mode of drug action which go back at least to 1900.
Beginning about that time and throughout the years since, various
theories have been put forward, such as those relating narcosis and
surface action, associated with the names of Traube, Lillie, and War-
burg, among others. Many such hypotheses relating one or another
special kind of biological properties to surface behavior have also
appeared during this period. Like all old hypotheses, they are old
because they have not easily been disproved and discarded, and they
have remained hypotheses because they were propounded far in ad-
vance of experimental basis. The most inclusive theory is probably
that of Ehrlich, who accounted for the phenomena of immunology
and pharmacology in terms of combining side chains. Thus, drugs
had organotropic or parasitotropic "haptophore" groups able to
combine with "chemoreceptor" groups in tissues or parasites; and
when so anchored the drugs exerted their effect. To present-day
workers this theory has often appeared extravagantly endowed with
unsubstantial detail; it is, however, probable that Ehrlich's concepts
have remained implicit in the pharmacological thinking of the past
few decades, while his terminology has been almost entirely replaced
by specific chemical names for reactive groups.

By far the most significant special theory along these lines is
one which accounts for the adsorption of the agent, and also, in a some-
what explicit fashion, for the effect which this has upon the cell. In

384

CHEMOTHERAPY

1940, Fildes postulated ihat nmihactcrial substaiucs in q<-ncr;il would
be found to interfere with (\ss(MUi;ii (Mizyiiiir re, ui ions of llir f)acterial
cell, either by combining with important substrates ("essential metab-
olites"), or by competing with them for combination with the enzyme.
We shall see in what follows that a number of mechanisms other than
"competitive inhibition" can give rise to antibacterial action, but the
importance of Fildes' postulation is that it pointed the way clearly for
a large amount of experimental work, including the production of
many new substances of biological, and probably eventually thera-
peutic, activity. In Fildes' illustration — toxicity due to mercuric
ions^there was no actual demonstration of the identity or even of the
thiol nature of the natural metabolite. But an accompanying paper
of Woods contained a classic demonstration of competition between
the natural /^-aminobenzoic acid and the artificial drug p-a.m\no-
benzenesulfonamide and its derivatives. Since that time, others have
furnished many more examples of interference with mammalian as well
as bacterial physiological processes, caused by su Instances related
chemically to natural metabolites and \'itamins.

Since much has been written elsewhere and there are discussions
in this volume bearing upon competitive inhibition, let us content
ourselves with two or three points relating to our chosen theme. First,
let us contrast the rationale applied by Ehi'lich and others in synthe-
sizing new agents prior to 1940 with a rationale based upon the Fildes-
Woods hypothesis. It was customary previously to begin with a toxic
principle, e. g., arsenic, mercury, or a dye, and to build up around it
various organic structures and substituent groups in the hope of de-
veloping a selectivity toward a particular type of cell. The literature
of the last few decades is full of reported attempts to "reduce the
toxicity" of such agents by appropriate modifications in structure.
On the other hand, the warped metabolite molecules now being syn-
thesized are likely to contain a special arrangement of relatively
common chemical groups, which is at the same time the basis for a
rather subtle toxicity and for whatever specificity the substance shows.
For example, pantoyltaurine, an inhibitory analogue of pantothenic
acid, is an amide derivative formed by joining together two natural
products. Other synthetic inhibitors are homologues of natural metab-
olites, which presumably combine with the specific enzymes for the
very reason that their structure is not exotic. The spectacular recent

R. D. HOTCHKISS

successes in chemotherapy have come through such substances as
thiouracil, stilbestrol, sulfonamides, and penicillin, all more or less
specific agents which are by no means merely carriers of potent, widely
active groups. These particular substances were probably first hit
upon somewhat accidentally, but the approach through metabolite
inhibition would seem to provide a means of arriving at such com-
binations rationally.

Second, let us turn to some considerations which may suggest
directions to take in the synthesis of future chemotherapeutic agents.
One of the main desiderata, we have agreed, is selectivity for a given
type of cell. Comparative biochemistry seems to reveal that different
cells are very much alike in the way they break down and oxidize
substrates; so agents which inhibit catabolic enzymes are likely to
have broad activity. On the other hand, since different cells appear
to have developed somewhat different proteins and other macro-
molecules in their protoplasm, it is to be hoped that agents which can
inhibit synthetic processes will display greater cell specificity. Further-
more, in designing modifications in chemical structure which will
deceive the enzyme proteins, it should be helpful to look at some of
the data from immunochemistry. In his book, The Specificity of Ser-
ological Reactions, Landsteiner describes a large number of modifica-
tions in chemical structure which greatly alter the reactivity of an
antigen toward its antibody, and in addition, many examples of over-
lapping specificity, that is, of different chemical groups which to a
considerable extent look alike to an antibody protein. These relation-
ships, arrived at by trial and error after innumerable painstaking
syntheses, may to some degree decrease the number of errors per trial
in the synthesis of pharmaceuticals.

The isolation and characterization a few years ago of two
crystalline polypeptides derived from tyrothricin, an antibacterial
product extracted from cultures of Bacillus brevis, seemed to offer an
opportunity of studying some of the factors which are involved in
combating bacterial infections. One of these antibacterial poly-
peptides is basic, the other is neutral. Both appear to have molecular
weights of about 2500 and contain enough aromatic and higher ali-
phatic amino acids to be alcohol-soluble and rather insoluble in water.
Both substances markedly depress the surface tension of aqueous
solutions. A number of the amino acid residues in each substance

386

CHEMOTHERAPY

have the usual /-configuration but, surprisingly, a considerable frac-
tion of the residues have the -'unnatural" and rare ^-configuration.
Whatever their similarities, the basic peptide, tyrocidine, is essentially
inert against bacteria in vivo, while the neutral peptide, gramicidin, is
moderately successful in eliminating certain localized infections. This
contrast in activity between apparently similar substances suggested
that a study of their mode of action might reveal some significant
principles.

It was soon learned that, in the test tube, tyrocidine is in some
ways more potent than the other substance; small amounts kill bacteria
of many different species, while gramicidin affects only Gram-positive
bacteria, principally by inhibiting their growth rather than by killing
them. But tyi'ocidine, like many other potent antiseptics, is too un-
specific in its reactions to be effective in vivo. It can be demonstrated
to precipitate with certain protein anions, and in the presence of pro-
tein its action upon bacteria is diminished. It seems probable that
tyrocidine, by combining with tissue proteins, becomes at once both
harmful to the animal host and almost inert toward the infecting
parasite.

Does the milder, more selective, gramicidin also combine with
protein, but merely more specifically, so that it combines with, and
inhibits, say, some particular enzyme? Neither gramicidin nor tyro-
cidine has much effect upon ordinary hydrolytic enzymes, but when
they are tested upon the respiratory and fermentative systems of whole
bacteria the situation is somewhat different. Tyrocidine added in
lethal quantities to resting bacteria promptly depresses the respiration
to a small fraction of the original rate. Gramicidin seldom shows
direct inhibitory action; indeed, when environmental conditions are
favorable, this agent commonly stimulates susceptible bacteria to
respire at two or three times the normal rate. Yet, with staphylo-
cocci, the final end products, acetic acid and carbon dioxide, are pro-
duced in the same proportion and eventual yield whether or not
gramicidin is present. However, the bacteria are not interested in
preparing acetic acid (it may even accumulate sufficiently to hamper
their activity sometimes) or carbon dioxide as such: as far as we know
today their best reasons for oxidizing glucose are (a) to liberate its
energy into useful form and (b) to produce intermediates like pyruvic
acid, which instead of being further oxidized may serve as raw mate-

R. D. HOTCHKISS

rials for synthesis of protoplasmic constituents in growth. We have
some evidence that it is with one of these energy-utilizing functions
that gramicidin interferes.

The useful form of energy derived from such processes very
commonly appears to be the phosphate ester. An organic substance
phosphorylated at the proper place is thereby activated for oxidation
or other conversion in much the same way that a wooden match tipped
with a phosphorus composition is prepared for a function in which a
plain wooden splinter would be inert. Since bacteria, when growing,
are performing a great many complicated chemical reactions, it is very
important that we are sometimes able to study the first beginnings of
synthesis in the phosphorylations and accumulation of intermediates
that accompany respiration in washed, nonproliferating bacterial sus-
pensions. In such preparations of staphylococci, normal oxidation
of substrates results under certain conditions in the accumulation
inside the cells of: (a) phosphate, at the expense of external inorganic
phosphate; and (b) an unidentified phosphorylated carbohydrate ester.
In the absence of nutrients, these events appear to represent the most
that can be accomplished toward growth. But both of these events are
virtually abolished by gramicidin in the same small concentrations
that suffice to prevent bacterial growth in a more satisfactory environ-
ment. In short, gramicidin does not prevent staphylococci from con-
suming substrate, but appears to prevent them from deriving benefit
from the act. It is tempting to conclude that this effect accounts for
the growth-inhibitory properties of gramicidin. There is some evi-
dence in favor of this. The prevention by gramicidin of inorganic
phosphate uptake, at least, is remarkably clear-cut and requires but
small amounts of the agent. Under various conditions with different
susceptible and resistant strains, different species, in altered environ-
ments, and with and without gramicidin antagonists, this effect parallels
the ability to prevent growth.

As yet, it is not possible to suggest which specific enzyme is
being inhibited in this effect of gramicidin, although it appears quite
likely that the phenomena observed can result from the blocking of a
single enzyme. Naturally, therefore, it is not possible to judge whether
gramicidin inhibits because it resembles the natural substrate, but it
appears improbable that a polypeptide-like substrate would be in-
volved in these reactions. So we may tentatively look upon the action

388

CHEMOTHERAPY

of gramicidin as an example ol noncompetitive inhibition of some
enzyme involved in a phosphorylating system. Tlie effect can be
studied in cell-free systems, since muscle and kidney extracts are pre-
vented by this same agent from phosphorylating glucose, although the
oxidation which normally furnishes energy for the phosphorylation
continues unaffected. It is to be expected that future work with cell-
free preparations will reveal exactly which of the energy-utilizing
systems is being blocked, and at the same time provide a specific
inhibitor for use in the study of this important function in the bio-
chemical laboratory.

Tyrocidine is found to contrast sharply with gramicidin in
mechanism of action. Bacteria exposed to it respire at such a low
rate that syntheses and growth are not possible, if only for the reason
that energy is not made available rapidly enough. Yet it is note-
worthy that respiration, easily reduced to 5-10% of its normal rate,
appears to resist further reduction even when a tenfold excess of agent
is applied. This fact suggests that there is no single key enzyme being
inhibited by tyrocidine, but either that an organization of enzymes is
being rendered more inefficient or that an alternate, inefficient, meta-
bolic pathway is being utilized which is less susceptible. At this point
let us recall that tyrocidine, with its basic group and alcohol-soluble
side chains, has a pronounced surface activity. Like the natural
saponins and bile salts, and the synthetic detergents, wetting agents,
etc., it can bring about hemolysis of red blood cells. If it were to
produce an analogous injury to bacteria, we should have, as with the
hemoglobin of erythrocytes, various soluble cellular metabolites
liberated from the cells into the surrovmding medium. There is un-
mistakable evidence that exactly this sort of damage is produced by
tyrocidine; and, in exposures at different time, temperature, or con-
centration, it happens in just those cases in which the bacteria are
killed. Appropriate analyses show that five-minute exposure to about
one-hundredth their weight of tyrocidine at zero degrees centigrade
is sufficient to extract quantitatively from staphylococci the total tri-
chloroacetic-acid-soluble phosphorus and nitrogen compounds present
in the cell. Analyses have been made for such typical constituents as
inorganic phosphate, adenosine triphosphate, other phosphate esters,
total esters, total nitrogen, amino nitrogen, amino acids, and pentoses.
Analogous effects have been observed with other microorganisms. So

R. D. HOTCHKISS

we conclude that tyrocidine kills bacteria because it brings about a
cytolytic injury to their cell membranes.

Some thirty antiseptic substances, including mercury deriva-
tives, halogens, oxidizing agents, formaldehyde, dyes, etc., were able
to kill or completely repress growth of staphylococci without releasing
cell constituents. The increase in cellular permeability, which is often
supposed to occur generally with death, was not noted at all, and
certainly cannot, for these cells at least, be confused with the clear-cut,
almost instantaneous, physical change that accompanies killing by
tyrocidine.

However, the further analysis of the bactericidal effect is of
interest to us as cytochemists. It is not hard to explain a drop in the
respiratory rate in these dead cells. With the seepage of coenzymes,
activating ions, and intermediates out of the cells, there occurs an
enormous "dilution" of the multicomponent respiratory systems. It
is well known that dilution of a reacting system decreases its rate more
or less according to an exponential function of the degree of dilution,
in which the exponent is the number of reacting components. There-
fore dilution of, say, a three-component system, can be tremendously
"inhibitory" in its effect. It is actually found that the residual
metabolism is more dependent upon the volume into which this dilu-
tion takes place than upon the concentration of cytolyzing agent.
Accordingly, it is unnecessary to suppose that enzymes are inactivated
by the minute amounts of tyrocidine needed to kill bacteria.

Hydrolytic enzymes such as the peptidases, phosphatases, and
nucleosidases, however, are less affected by dilution than are respira-
tory enzymes, since the hydrolytic systems contain fewer components,
and the concentration of one of these, water, is scarcely changed by
dilution. In harmony with this, bacteria cytolyzed by tyrocidine
begin to undergo enzymic post-mortem changes which result in the
liberation of soluble degradation products, and with some species, but
not universally, may progress to actual bacteriolysis, i. e., partial or
even complete clearing of the suspension. These effects, if not avoided
by keeping the temperature low, or by adding enzyme poisons, may
obscure the observation of the initial cytolytic injury. It is apparent
that tyrocidine does not inactivate these hydrolytic enzymes. Does
it activate them, or why do they begin to work after the cytolytic
death? Only a few experiments with certain phosphatases have so

CHEMOTHERAPY

far been undertaken but they suggest an answer. Normal washed
staphylococci are able to hydrolyze added adenosine triphosphate, and
glycerophosphate. Tyrocidine-trealed cells, which show a burst of
phosphatase activity toward their own small complement of phosphate
esters, are not able to operate any faster upon an adequate concentra-
tion of added phosphate esters than the normal cells did before their
membranes were made permeable. With certain limitations, these
experiments indicate that phosphatase and adenosinepyrophosphatase
are situated on or near the surface of these bacteria, and cytolytic
injury allows substrates from the interior of the cell to leak out into
the medium, where they can be hydrolyzed. It will be evident that
tyrocidine can possibly afford us a novel means of studying the workings
of the cell.

We may stop to inquire whether there is not good reason why
phosphatases should be able to act at the periphery of the cell. Like
esterases, peptidases, amylases, etc., their function is probably always
to break dow^n complex substances since, like these other hydrolytic
enzymes, they appear unable to incorporate the energy needed for
the corresponding synthesis. Peripherally distributed phosphatases
may serve in part as agencies for converting phosphate esters, which
appear to be generally nondiffusible, into a form capable of entering the
cell. * Those cells able or obligated to use certain complex products
of other cells as nutrients must be prepared to degrade these extraneous
substrates to an acceptable form. In these cases we may expect other
hydrolytic enzymes such as those mentioned to be active, although
perhaps not entirely localized, at the cell boundary — more or less in
contrast with the enzymes of respiration, fermentation, and synthesis.
Many extracellular proteinases and polysaccharidases have been found
in the medium surrounding the cells which produce them. And re-
cently there have been interesting indications that the enzymes hy-
drolyzing trehalose (in yeasts) and lactose (in yeasts and colon bacilli)

* This is not in conflict with the possibility that, after hydrolysis, some
substrates may be phosphorylated by the cell itself during absorption, as suggested
by Lundsgaard, Verzar and others for the case of glucose absorption from the
lumen of the kidney tubules or the intestine. The whole selective mechanism
would ensure that foreign, and perhaps inimical, substances would not g£un ad-
mission.

R. D. HOTCHKISS

were active at the surface of cells in those strains capable of fermenting
these sugars.

We have waited until now to mention a large group of other
substances, which, so far as known, behave in essentially the same way
as tyrocidine. It is not surprising that fixatives such as hot water,
alcohol, acetone, and trichloroacetic acid also liberate cell constituents.
But, in the enormous class of synthetic detergents, wetting, dispersing,
and emulsifying agents, etc., there are many bactericidal surface-active
agents; and, so far as they have been investigated, these show the same
phenomena as tyrocidine when they kill bacteria. Cell solutes are
liberated, respiratory activity is depressed, autolytic changes are
initiated, whenever the substance and its concentration are adequate
to kill. Indeed, tyrocidine deserves no special consideration except
for the fact that its behavior has been investigated more fully than that
of any of the synthetic products. And the diversity of chemical struc-
tures which produce this cytolytic type of killing allows us to conclude
that the structural requirements for activity are chiefly physical and
are broad and ill-defined. Basic (cationic) agents are generally more
effective, especially toward the Gram-negative bacteria, but they may
be aliphatic straight-chain, aromatic or heterocyclic quarternary
ammonium bases, or simple fatty amines. Various alkyl and aryl
sulfates and sulfonates have similar activity, as do some fatty carboxylic
acids. When in concentration sufficient to kill, phenols and cresols
have the same eff"ect, but this is greatly increased in alkyl-substituted
phenols, which have more enhanced surface activity. The small
class of nonionizing surface-active agents has apparently not yet
produced a representative with notable bactericidal activity.

Somewhat systematic investigations have already been made
on the relation of structure to bactericidal activity in some of these
compounds, especially the phenols and the aliphatic acids. One of
the few conclusions that could be reached was that the most effective
substance in each series is usually one having approximately sixteen
to eighteen carbon atoms. Branched-carbon-chain compounds appear
to be commonly more toxic than those with straight chains. There
are almost no evidences of selectivity except that Gram-negative bac-
teria are mainly vulnerable only to basic substances unless the pH is
low enough that the cell proteins can combine with anionic detergents.
There are other small diff"erences of sensitivity among various species.

CHEMOTHERAPY

Classification of substances as inactive is undoubtedly largely a rela-
tive matter, since probably almost any surface-active agent worthy of
the name would kill bacteria if used in the concentration of around
1 % required for phenol and cresol, which are thought of as substances
toxic to bacteria.

The water insolubility of a large portion of the molecule is the
driving force that crowds even minute quantities of a surface-active
agent out to the boundaries of the aqueous phase. The affinities of
this nonpolar group for elements of the bacterial structure determine
where upon the cell it will become anchored and to some extent what
harm it will do there. The general toxicity of a wide variety of sub-
stances in this class suggests that the cellular elements having affinity
for hydrocarbon groups do not differ very much from species to species.
Evidently, whenever the nonpolar group carries basic or acidic groups
along with it the harm it will do may be extensive. If no strongly
ionized groups are present, either there is less of the substance adsorbed
or its effect is less drastic. Gramicidin, which is a rather insoluble
surface-active agent of the nonionizing class, is not notably bacteri-
cidal, but does display some specificity and delicacy of bacteriostatic
action, as indicated above. Perhaps systematic exploration or syn-
theses of further compounds in this general physical class of nonionic
agents may offer some promise for those anxious to find new medicinals
with some capacity for showing specific action upon cells.

The special cases of antibacterial agents just considered have
caused us to be concerned with cellular attributes of which we fre-
quently lose sight. Let us attempt to organize our rudimentary in-
formation about cell physiology so that we may see whether there are
other possible sites vulnerable to chemical action. We consider most
cells to be an organized system of enzymes, most of them at present
unkno\Nm, preserved together with vital coenzymes and metabolites
within a structural framework the outer boundary of which displays a
selective permeability. As indicated in the table on pages 394 and 395,
the chemical tasks of the living cell may be classified as several differ-
ent, obviously somewhat overlapping, types of function. This tabula-
tion wall, one hopes, look very naive indeed in a few years, but it may
help us a little now as we try to visualize the problem of exerting
chemical control over cells of different types.

Not only is the understanding of the phenomena under "mecha-

393

u

V

X!
<u
-0

o

■i I

cfl

(V.

a
gx: a

a o

V •-
bo >
O ?3

-C 3

•-^ tt-< '^

3

cr

o n..

•3 C

« -fj 4) 3

o .o •

«, 3

u (J
o „

o

c

u

5 9-'

o

3

O >s

U O

ti3

S w

o

O

■a ^
fix)

«^

3

cr

u
ho

n..

O
G
u
-C
(X
o

c
1)

'n

j2

re re
■3°

V -
w ^

«3

3

>

o

re

J2

3

U 3

o -a

a
re

3
3 cj

N X!

o

-a

a
re

— a. o u

X!

3 f.'

"3 .23 S
• -* re

V

CO

« x; -
> -S 3

>--3

u
bo

O

X3

re

3 ^^

" 3

V

o

OT 1/3 W ' '

< < o o

CO

o

^
w

o

V

8
o
w

3
u

3
li
■*-»
w
V
T3

a 3
"« o

tio re

re

3

^ ^ S

V

>

V

>

V

Oh
^•-

o o

".^

3 ^
O

1)

S
o
o

3
O
3

V
D.

3 cA

8 S

3 N c

^ O C O

.S JD .S (L> •-<

.■s 3.ti o.-s

J2 M j2 U JD

-3
C

x;

3

3
p

Sb

u

-a

X

3
O

u
re

w

3 -

N

3

o
o,

V

V V

> >

QQ

re
>

bD
u

V

a

3

_o

'to

V
>

re
6

V
+J

3

3
o

i2

>

Q Q

g^

■^ 3
-«; O

« ti
^ .>

o
re

8

3

o

X

4J

X

■♦-'

o

O 3
V

to o

.2 P u

ji; bn
>

I— I i-< w

u
3
X!

o

o
be

tv

3

U
1/3

c

J5

.a
u

o
a
a

3

X

"a

6
o
u

o o ii
.2 tS c

o

s

>- ;3

w

O
u

-a
x: ,
>.

u
re

3

O
XJ

re
re

o

H

o

re
■♦-»
re
u

>

o

ja
re
■*-»

V

6

Si

y re
X

bo o

^ X

'si

3 >.

■r o t« re
i3 ■!-> en ■!-•

322g

a

a.

3

3 u
3
T3
O

53S

.2 fi

u -o

T3

2 « «

Ci, re I,
"^ a. ?;

" 3 C

>> 3 bo
be. 13 o

U Si >,

3 rex^
V ^ bo

° re °

u 2 u

'•'^ bo

• re

be
re
ii
o

w in

Q

-v

X

o

Si

Si

re
u

en .3

X *.•

■*-»
3

(/5

N

3

V

V

X

3

C
V

3
3

-a

V
N

o

3

o

3
•T3

V

6

re
3

o

X

ex
o

X

0H

3 re nj

C ^ CO-

"013 "3
O 3

- re

3
SI

E/l

(U
Si

3

o

3

Si

re «J .

O J3

-3

V be

o

O 3

3 .t;x!

3 trt «

re

V

Pi

a
o
■a

a

.3

O

I
I

.§

■vs

s

r^

^

3

2

g;t^

1*5

a

c

£

a

[a,
'0

4J

.s

O

■«->

o
a.

to -w

03
X

3
'0

«J bo

6.S
c.S

be
oil

S a

N >

o

15

V

o

>^

>^

*-»

bo

<iS

w

O

o

«->

.. >.

n

y JV

t;

C bJC'

bJD

TJ iT

() c

w

-) o

>

i- 'M

O O

n

a j=

02

o

u

o

-1

W

<

a

12 nJ

3
bO
03
O
U

u

1/
bo

— , «^

-o

O

U

f>..

c

u
o

on-.
•■^ Vi

V
.^ C

o o -^ o

C V3 VJ

<<<

o

c

Oh

O

u

03

4 ^

s §

03
C
u

V

c
'c
o

V

>

o

3 J2 -'%• .5

crT3 y C 'H

- 3 ■- C o

y o P 03 j::
03 ^5d1c o-

" S 2 ^r ji

= .2 i^'12 ^
'aa.S aS
o C o

u y bo

♦i CO u

< a pj

c

V

bo

bn
c

13

u
o

-o

3
C/5

a

c
u
bO

u
03
I

u
o

3

3
>

J3

a

o

u

v 2

1- c

3 03

*-' u

t-> bo

^ s'e

— CO to

C3 03 TO

CO O

V

bo
03

S
O!

T3

V

n

03

y

o

V

V

hn

-^

n

>.

03

03

I-I

x;

Cl

u

n

^

01

c

y

V

"^

03
u
Si

e

E

u

V
CO

0!
V
u

o
n

V

bo

c

Ji -03
w C O

y 2 y

s

V

c
•2 c

> t;

c c

s

N

c

2°

CO S

—I ^

.a a

CO o

a"

c
o

a

03

O

o

E

c

O V

52

<

2 J=
"y y

N

c

bo

CO

y
be
n
«

■g y ^
J^ 2 ^

^ ?

c

03
O

y

y

"o

O

u

'bo

y
c
y
u

-o

<

03 N

■*-• C

03 y

O CO

x; c

D. «!

.- u

U *->

y y

C ■ ■

G CO

y o

o: 03

a

o

o
a

a

3

c
o

y y

.2-y

03

a

03

3

Vm

O

y

u

c

w-i

n

o

CO

01

u.

>

J3

y

ro

4J

U

CL,

.2

C flj

S - o

y S ..Q

S 'S «=
S

y

c,

« «

H i5

3
C ,

y

3
"O

^ 2

a

g y

u y
bo >

C§

.2 w

03 O

c S

CJ ^
U N „

c H y
o « ^

■^ ?■ o

o -Q M

c

•2 c
y.w

O
u

=2 O

c
o

c
>.

CO

y

03 03

o y

- 3

bo.S
2 y

03

y

y
bC
C
03

N

c

G

09
J3

y

03

C

o

c

0!

03

= i
J :2

y
_c

'ot

O

c
y

" b3

O y

o "
.- y
■t-» *j
OS 03

N j-

^ a ^

.n CO ^^

■- r\ 03
O 03

>- :2

03

O

V

o

G
03
C
y

c

§

u
CO

T3 3

c "

.2 -c
CO bo o

> c -a
^^i2

d

c2.2

— I U
U

u

a.

CO

o

OS

c
o
O

R. D. HOTCHKISS

nism" in the table all in the theoretical stage, but we have little more
than a crude terminology with which to describe them. Some detail
could be given under the mechanism of respiratory and fermentative
catabolism, but almost none under any of the anabolic, energy-
utilizing mechanisms. The examples of agents given in the last
column are, at best, somewhat questionable. Nevertheless, it is hoped
that the table will serve to demonstrate that it is almost inexcusable
that, when considering the mode of action of new or proposed drugs,
we have all almost universally limited ourselves to a consideration of
possible eflfects of the drug upon some aspect of the respiratory or fer-
mentative degradation of foodstuffs. As heating engineers trying to see
why the house is cold, we have been shortsightedly preoccupied with
the flicker of the furnace fire and have never bothered even to feel
the pipes and radiators upstairs.

Throughout the above, the viewpoint has been somewhat in-
sistently shifted back and forth from the planning of new chemo-
therapeutic agents to studying the action of the old ones. It is not
pretended that empirical ways of uncovering new agents have been
outmoded, but we do maintain that the most successful empiricism
will always use, consciously or unconsciously, what reasonable hy-
potheses are available. And the drugs are among our best cy to-
chemical reagents: they can give us information about the reactivities
of living cells, and this information will surely help us to foresee the
probable behavior of new drugs. Strictly speaking, there can be no
study of relation between pharmacological action and chemical struc-
ture and there can be no study of mechanisms of drug action — there
can only be the study of drug-protoplasm interaction.

Selected References

LOCALIZATION OF CELLULAR CONSTITUENTS

Caspersson, T., "Studies on the protein metabolism of the cell," Naturwissen-

schqften, 29, 33 (1941).
Claude, A., Hogeboom, G. H., and Hotclxkiss, R. D., "Distribution of enzymes

in cytoplasmic constituents," to be published.
Bounce, A. L., "Further studies on isolated cell nuclei of normal rat liver,"

J. Biol. Chem., 151, 221 (1943).
Gomori, G., "The distribution of phosphatase in normal organs and tissues,"

J. Cellular Comp. Physiol., 17, 71 (1941).

CHEMOTHERAPY

Feulgen, R., and Rossenbeck, H., Microchcmical test for nucleic acid of the
thymonucleic acid type and the selective staining of cell nuclei in micro-
scopic preparations, ^. phjisiol. Chem., 135, 203 (1924).

Lan, T. H., "The </-amino acid oxidase, uricase, and choline oxidase in nor-
mal rat liver and in nuclei of normal rat liver cells," J. Biol. Chem., 151,
171 (1943).

Linderstr0m-Lang, K., "Distribution of enzymes in tissues and cells," Harvey,
Lectures, 34, 214 (1938-1939).

Myrback, K., and Vasseur, E., Lactose fermentation and the localization of
enzymes in the yeast cell, Z- physiol. Chem., 277, 171 (1943).

Wolf, A., Kabat, E. A., and Newman, W., "A study of the distribution of acid
phosphatases with special reference to the nervous system," Am. J. Path., 19,
423 (1943).

ANTIGEN-ANTIBODY REACTIONS

Landsteiner, K., The Specificity oj Serological Reactions. Harvard Univ. Press,

Cambridge, 1945.
Marrack, J. R., The Chemistry of Antigens and Antibodies. H. M. Stationery Office,

London, 1934.

THEORIES OF DRUG ACTION

Ehrlich, P., E.xperimental Researches on Specific Therapeutics. Lewis, London, 1908.
Lillie, R. S., "Antagonism between salts and anaesthetics," Am. J. Physiol., 31,

255 (1913).
Traube, J., Theory of narcosis. Arch. ges. Physiol., 153, 276 (1913).
Warburg, O., Physical chemistry of cell respiration, Biochcm. Z-> H^) l^^*

(1921).

COMPETITIVE INHIBITION IN METABOLISM

Fildes, P., "The mechanism of the antibacterial action of mercury," Brit. J.

Exptl. Path., 21, 67 (1940).
Quastel, J. H., and Wooldridge, W. R., "Some properties of the dehydrogen-

ating enzymes of bacteria," Biochem. J., 22, 689 (1928).
Woods, D. D., "The relation of /^-aminobenzoic acid to the mechanism of the

action of sulfanilamide," Brit. J. Exptl. Path., 21, 74 (1940).

GRAMICIDIN AND TRVOCIDINE

Dubos, R. J., and Holchkiss, R. 1)., "I'he prodiu lion of bactericidal sub-
stances by aerobic sporulaung bacilli," J. Plxpll. Med., 73, 629 (1941).

Hotchkiss, R. D., "Gramicidin, tyrocidine, and tyrothricin," in Advances in
Enzymology, Vol. IV. Interscicncc, New York, 1944, p. 153.

397

R. D. HOTCHKISS

SURFACE-ACTIVE AGENTS

Baker, Z., Harrison, R. W., and Miller, B. F., "The bactericidal action of
synthetic detergents," J. Exptl. Med., 74, 611 (1941).

Hotchkiss, R. D., "The nature of the bactericidal action of surface-active
agents," Ann. N. T. Acad. Sci., 46, art. 6 (1946).

Kuhn, R., et al., Invert soaps, Ber., 73, 1080 et seq. (1940).

Stanley, W. M., Coleman, G. H., Greer, C. M., Sacks, J., and Adams, R.,
"Bacteriological action of certain synthetic organic acids toward Myco-
bacterium leprae and other acid-fast bacteria," J. Pharmacol., 45, 121 (1932).

Suter, C. M., "Relationships between the structure and the bactericidal prop-
erties of the phenols," Chem. Revs., 28, 269 (1941).

398

25

BIOCHEMICAL ASPECTS
OF PHARMACOLOGY

ARNOLD D. WELCH, professor of pharmacology, school of

MEDICINE, WESTERN RESERVE UNIVERSITY

ERNEST BUEDING, assistant professor of pharmacology, school
of medicine, western reserve university

f, TARL Y MAN, in his almost continuous search for food,
J-^ encountered a great variety of substances capable of
affecting the normal functions of the body. Some of these materials
produced effects which ultimately were correlated with their ingestion,
a circumstance which led to their use in attempts to alleviate suffer-
ing or to combat disease. Undoubtedly the experiences and ob-
servations of primitive medicine men added many other substances to
the early materia medica. Throughout the centuries, testing of the hit-
and-miss type has continued, and indeed few substances now exist
which have not at some time been used in an attempt to change the
course of bodily disorders. Of these, only a small number has proved
to have lasting value, although a much larger number of agents of
little merit remains to distend the dispensatories of the present day.
From the loss of most of these the world would suffer very little.

The search for new drugs continues, although they are usually
derived from a different source than formerly, for the synthetic activity
of nature has been replaced almost entirely by that of man. Also, in
the primary trials of new agents, human beings have now largely been
replaced by animals and microorganisms; yet the basic philosophy of
trial and error continues essentially unchanged.

399

A. D. WELCH AND E. BUEDING

|f Insight into the nicchanisiu of the aetioii of drugs on tissues

and organs of necessity awaited the attainment of considerable knowl-
edge in other fields, in particular that of physiology. As the prin-
ciples governing the function of organs and tissues began to be eluci-
dated and the science of physiology was developed, study of the mecha-
nism of the action of drugs began. Thus, not only has pharma-
cology developed logically from physiology, but the greater portion
of our present knowledge of drugs was acquired through the use of
methods devised primarily for physiological studies. The techniques
and concepts of physiology were gradually incorporated into the
matrix of the newer science, although the emphasis was shifted from a
primary interest in function to an exhaustive study of each drug, so
that all its multitudinous efTects might be uncovered, rather than to a
continued study of function and the manner in which it is influenced
by many chemical substances.

This approach to pharmacology has been to a considerable
degree self-sterilizing, since it led in the course of a comparatively few
years to relative exhaustion of the stock of drugs capable of exerting
sufficient physiological efTect to justify extensive study. As a conse-
quence, findings in the field of pharmacology became relatively prosaic,
interest in the science to some degree waned, and few young men were
attracted to or trained for a scientific career in the field. Where note-
worthy advances occurred, they came especially in those fields in which
) the closest cooperation between the pharmacologist and the synthetic
organic chemist was possible; but here too almost all advances, except
the most recent, resulted from approaches which were essentially
empirical.

In general, pharmacologists must face the indictment of having
viewed with too great complacency the limitations of the classical
physiological attack upon the basic mechanisms of drug action. The
continued emphasis upon the physiological approach led to many
improvements in our understanding of the action and proper usage of
important therapeutic agents, but, during recent years at least, such
methods have rarely opened new fields of research. Indeed, only
a few pharmacologists have given attention to the biochemical aspects
of the action of drugs or have turned to the investigation of more
fundamental systems.

The fact of the matter is that pharmacology, as a separate dis-

400

PHARMACOLOGY

ciplinc of the medical sciences, very nearly ceased to exist, a situation
which might not have been long delayed had it not been for the
chemotherapeutic advances oi the l&st decade. These, although
initiated largely by workers in cognate fields, gave pharmacology a
cogent reason for being, and placed renewed emphasis on the fact that
the "site" of drug action is not composed of static morphological units.
Rather is it to be found in the interrelated chains of chemical reactions
which constitute the dynamic equilibrium of the cell.

That the emphasis on the study of the action of drugs has been
placed primarily on the responses of organs and tissues, rather than on
those of cells, is well exemplified by the investigations of an important
pharmacological agent, viz., digitalis. The cardioactive glycosides of
the digitalis and related plants have long been known to affect bene-
ficially the function of the decompensated failing heart, an effect now
related largely to an increased contractility of the cardiac muscular
tissue (5). Yet, few studies have so far been made of the influence of
such drugs on the metabolic reactions supplying the energy required
for this function of myocardial cells. The validity of such an approach
can be examined by investigating the efTect of such drugs on various
metabolic reactions and isolated enzyme systems in cardiac and in
other forms of muscle.

The evidence available today, although in some cases circum-
stantial, strongly suggests the possibility that the action of many drugs
and poisons is mediated through a direct or indirect effect on enzyme
systems. This is not a new hypothesis. Myrback (23) made a
similar suggestion in 1926, and Clark (6) has reviewed the literature
of the subject up to 1937. Consideration of various classes of drugs
indicates that in only a few cases is it definitely unlikely that enzyme
systems participate in the primary action of drugs or poisons, for
instance, saline cathartics and diuretics, adsorbents, neutralizers of
alkali or of acid, and hemolytic compounds such as the saponins. It ' '
is equally true that there are only a few cases in which the action of
drugs or poisons has been proved beyond question to involve enzyme
systems primarily.

Of perhaps greatest importance, from the standpoint of un-
questionable enzymic participation in drug action, is physostigmine
(eserine), a potent drug and poison which was shown by Loewi and
Navratil (17) in 1926 to inhibit choline esterase specifically. Under

401

-fi

A. D. WELCH AND E. BUEDING

ordinary circumstances the neurohormone, acetylcholine, responsible
for the chemical transmission of nerve impulses at many sites, is de-
stroyed promptly, so that the effect of each nerve impulse is very
transitory. Partial inhibition of the enzyme which catalyzes the
hydrolytic cleavage of acetylcholine permits the accumulation of the
neurohormone in higher concentration, a circumstance equivalent in
its physiological effects to the injection of acetylcholine at the endings
of innumerable cholinergic nerves. Proof that the transitory action of
acetylcholine is due to its enzymic destruction and that this enzyme
is actually the "site of action" of physostigmine, constituted the first
clear-cut demonstration that the primary action of a drug can be based
solely on its effect on a single enzyme. It is of considerable practical
importance that the rate of hydrolysis of choline esters can be altered
by modifying the structure of the choline moiety or of the esterifying
acid. Thus, acetyl-j8-methylcholine and carbaminoylcholine have
the same pharmacological effect as acetylcholine, but they are slowly
or not at all hydrolyzed by choline esterase.

Other examples of drugs and poisons, the primary action of
which involves enzyme systems, are neostigmine, a compound related
to physostigmine, ionizable cyanides, and such vitamins as thiamin,
riboflavin, and niacin. However, there is strong presumptive evidence
for the involvement of enzyme systems in the action of various other
drugs and poisons.

Quastel et al. (26) have gtttempted to explain the mechanism
of action of narcotics in terms of their effects on the respiration of brain
tissue and on isolated enzyme systems. Various drugs which produce
anesthesia (barbiturates, chlorobutanol, and ether) were shown to
inhibit reversibly the oxygen consumption of the cortex in vitro. Oxi-
dation of glucose, lactate, or pyruvate is inhibited to a larger degree
than the oxidation of either succinate or glutamate, while the anerobic
utilization of glucose by the brain tissue is inhibited only slightly. In
continuation of this work, the effect of chlorobutanol on several
enzyme systems involved in the oxidation of glucose was examined by
Michaelis and Quastel (22). Since the activity of various dehydro-
genases, cozymase, Straub's flavoprotein ("diaphorase"), or cyto-
chrome oxidase is not decreased by chlorobutanol, in concentrations
which inhibit glucose oxidation in the brain, it was concluded that the
narcotic inhibits an as yet unidentified respiratory enzyme system.

402

PHARMACOLOGY

Because of disagreement concerning the concentration of various
narcotic agents in brain tissue, at varying depths of anesthesia, evalua-
tion of the observations of Quastel and his co-workers is not a simple
matter. Fuhrman and Field (8) concluded that, during anesthesia,
the concentration of certain narcotic agents in the brain is less than
one-fourth that required for the inhibition of glucose oxidation by
brain tissue in vitro; however, their conclusions do not appear to have
been based on adequate analytical data. Further studies will evi-
dently be required to establish whether those concentrations of narcotic
agents which produce anesthesia in vivo inhibit to an appropriate degree
any of the enzyme systems that are required for the normal functions
of the brain. The effect of narcotic agents on the functions of other
tissues must also be considered, since it has been shown (22) that
certain enzyme systems of skeletal muscle are inhibited by chlorobutanol
to the same degree as are comparable systems of brain. However, the
functional activity of muscle does not appear to be inhibited by those
concentrations of chlorobutanol attained during anesthesia.

Even if it were found that the concentration of narcotic agents
required to inhibit certain enzyme systems, or the over-all respiration,
in vitro, is significantly greater than the concentration attained during
anesthesia, a reaction more specifically concerned with the functional
integrity of the nervous system may be found that is inhibited by lower
concentrations. In favor of the theory that narcosis involves a more
general depression of energy-transferring mechanisms in nervous
tissue, however, is the wide variety of chemical structures, usually
lipid-soluble, which produce a reversible narcotic effect on many types
of cells. The importance of further study in this field of investigation
is clearly indicated.

" Seevers and Shideman (33) have made rather extensive studies

of morphine in an attempt to place its action on a biochemical basis.
In confirmation of the findings of Quastel and Wheatley (26) it was
shown that the drug in high concentration (0.12%) inhibits the oxygen
consumption, in vitro, of the cerebral tissues of rats, when certain
substrates, particularly lactate, are added. In the absence of added
substrates no effect was observed. Various mammalian dehydro-
genases (lactic, citric, glucose) were shown to be inhibited by similar
concentrations of morphine; levels much higher than any which could
be attained in the animal body. The results were interpreted very

A. D. WELCH AND E. BUEDING

cautiously by the investigators and serve as a basis for the further
studies which are needed.

More recently, Mayer and McCawley (21) have shown that
the N-allyl homologue of morphine is even more depressant of oxygen
consumption by slices of rat cerebrum than is morphine. Other
studies (13,18,35) have shown, however, that this derivative is antago-
nistic to many of the effects produced by morphine, particularly that
of respiratory depression. It would appear that depression of oxida-
tive reactions of brain tissue, in vitro, cannot serve to explain the
differences between the effects of morphine and the allyl homologue
and certainly cannot account for the narcotic and analgesic actions of
the parent alkaloid.

A striking observation made during the biochemical studies of
Shideman and Seevers (33) is worthy of note. They found that the
chronic administration of morphine to rats leads to a sustained accel-
eration in the oxidative metabolism of skeletal muscle, an effect which is
maintained when morphine has been withdrawn, and which is noted
when a major portion of the morphine would have been eliminated
from the body. The addition of morphine to muscle tissue, in vitro,
produces an increase in oxygen uptake, an effect observed whether the
tissue used is obtained from normal rats or from chronically mor-
phinized animals with an initial rate of oxygen consumption nearly
twice that of the muscle tissue of normal rats. Shideman and Seevers
do not believe that these observations are to be accounted for by the
oxidation of the drug. As was pointed out, the data available do not
clearly associate the effects produced in vitro with those which occur
uniformly in chronically morphinized muscle during withdrawal.
The fact that a concentration of azide which is without effect on the
respiration of normal muscle reduces to an approximately normal level
the oxygen utilization of chronically morphinized muscle should also
be mentioned. These interesting observations suggest the need for
additional studies of the mechanism of action of this very important
drug.

On the basis of studies in their own laboratory (9) and else-
where (3), Gaddum and Kwiatkowski advanced the hypothesis that
the action of ephedrine, a naturally occurring sympathomimetic com-
\ pound chemically related to epinephrine, can best be accounted for
by its inhibition of amine oxidase. This enzyme is one of several which

404

PHARMACOLOGY

participate in the inactivation of epinephrine and the closely related
or probably identical neurohormone, sympathin, responsible for the
chemical transmission of adrenergic impulses. Inhibition of the
enzyme leads to the accumulation of epinephrine in much the same
way that physostigmine protects acetylcholine from destruction. The
stimulant action of amphetamine (benzedrine) on certain functions
of the central nervous system was related by Mann and Quastel (20), ' i
in a rather complicated theory, to inhibition of the oxidative deami-
nation of compounds such as tyramine, which is deri\ ed from tyrosine.
It was suggested that the type of mental fatigue relieved by amphet-
amine is due to the formation, by amine oxidase, of toxic aldehydes
from compounds such as tyramine. Central excitation is thus con-
sidered to be secondary to an inhibition of amine oxidase. Needless
to say, the acceptance of such an hypothesis requires a great deal more
evidence than that so far made available.

Amine oxidase is undoubtedly responsible for the oxidative
deamination which many of the sympathomimetic amines undergo in
the body (2) and is thus responsible, in part at least, for their inactiva-
tion. Derivatives of phenethylamine which possess an alkyl group
alpha to the nitrogen, such as amphetamine and ephedrine, are not
only refractory to the action of amine oxidase, but are capable of
inhibiting markedly the action of the enzyme.

The enzymic approach to the study of the mode of action of
drugs is by no means limited to those compounds which aflfect mam-
malian tissues. The chemotherapy of infectious and parasitic diseases
is based on the use of compounds which in suitable dosage have less
efTect on chemical reactions essential to the well-being of the host ,
than on those reactions requisite to the life or reproduction of invasive
organisms. Thus, chemotherapeutic agents vary in their margins
of safety from those that are of the same order of toxicity for both host
and parasite to those that are essentially devoid of deleterious eflfects
on any mammalian host. Examples of such extremes are actually to
be found among drugs which may be used in the treatment of a single ; |
disease, like syphilis. Thus, metallic mercury, when used by inunction, /
is of such toxicity that it is possible only to arrest the progress of the \
infection, while with highly purified penicillin not only may apparent (
cures be produced but man appears to be essentially uninjured by any J
dose so far administered.

A. D. WELCH AND E. BUEDING

In a recent study of the effect of quinacrine (atabrine) on re-
spiratory enzymes, Haas (12) found that the drug inhibits the respira-
tion of Plasmodium knowlesi to a greater extent than that of mammahan
tissues. Furthermore, at relatively low concentrations the drug in-
hibits cytochrome reductase and glucose-6-phosphate dehydrogenase
{^wischenferment), while it has little or no effect on cytochrome oxidase,
cytochrome C, or triphosphopyndine nucleotide. The concentration
of atabrine required to inhibit the respiration of malaria parasites in
vitro is considerably higher than those concentrations of the drug in
plasma which exert an antimalarial effect in vivo; conceivably, how-
ever, the concentration of atabrine in the parasites may be considerably
higher than the concentration in the plasma. Quinine was less
active than atabrine in inhibiting both cytochrome reductase and the
respiration of malaria parasites, a difference which is in agreement with
the fact that the antimalarial activity of atabrine is now recognized as
definitely greater than that of quinine. In order to test the hypothesis
that the antimalarial activity of these two drugs is in part due to an
inhibition of the activity of cytochrome reductase, it might be helpful
to study the enzyme-inhibiting effect of a group of compounds closely
related structurally to atabrine, but among which varying degrees of
antimalarial activity are found. The work of Haas would take on
added significance if a close parallelism between the relative activities
in vitro and in vivo were found in such a series, and if the discrepancy
between the concentration causing enzymic inhibition and that which
is therapeutically active were satisfactorily explained. A closer corre-
lation between the effective concentration in vitro and in vivo was found
by Silverman et al. (34), who showed that quinine in concentrations
only slightly above those therapeutically active in vivo inhibits the
oxidation of glucose by Plasmodium gallinaceum, without affecting
glycolysis.

Earlier, but less striking, attempts have been made to correlate
the inhibition of enzyme systems with the action of drugs chemo-
therapeutically effective in vivo. In this connection there might be
mentioned the action of atoxyi and of quinine on certain lipases,
the antifumarase-activity of certain trypanocidal compounds (25),
and the inhibition with acriflavine of a cyanide-insensitive hydrogen-
transporting system in trypanosomes (30).

Thus, various investigations have been conducted in which the

406

PHARMACOLOGY

effect of a drug on isolated systems has been studied in an attempt to
explain the physiological action of the compound. In some of these
investigations attention has not been directed toward certain factors
which, in the opinion of the authors, are of importance. It is suggested
that, among others, the following criteria should be considered before
the physiological action of a drug is attributed to an effect on an isolated
enzyme system.

(1) The concentrations of the drug or poison which produce the
effect in vitro and which obtain at the anatomical site of action should be closely
similar.

(2) If a drug or poison exerts its effect primarily on a specific tissue
in vivo, either the effect on this tissue in vitro should be more pronounced than
the effects on other tissues, or the system inhibited must be shown to have more
functional significance in this than in other tissues.

(3) Among structurally related compounds there should be close
parallelism between the pharmacological activities in vitro and in vivo: all drugs
of the same chemical series which are active in vivo must also be active in vitro,
and those inactive in vivo must also be inactive in vitro, unless the discrepancy
can be accounted for by poor absorption, by inadequate distribution to or
penetration of the cells involved, by too rapid excretion, or by metabolic
alteration.

These criteria may be of value in avoiding premature explanations of
the physiological action of drugs on the basis of effects exerted on
isolated systems.

Until very recently, the search for new chemotherapeutic agents
has progressed along almost exclusively empirical lines. Such an
approach, though tedious, has, nevertheless, yielded many drugs of
importance, and will undoubtedly continue to do so in the future.
Recently, however, largely on the basis of the theory of the mode of
action of sulfonamides, concepts have been developed that promise
to be of value in the development of compounds of utility in chemo-
therapy and in the study of intermediary metabolism.

According to these concepts, which are reviewed in detail else-
where (19,37,38), interference with biological processes may result
from the use of compounds structurally related to, but not utilizable
in place of, substances essential for life or reproduction (essential
metabolites). Interference of this type can be explained most simply
on the basis of a direct competition between the essential metabolite

407

\\

A. D. WELCH AND E. BUEDING

and its structural analogue for some cellular component for which they
\ 1 / both have great affinity; in some cases, however, antagonism involves
other factors than simple competitive inhibition. In verification of
this concept, there have been developed various structural analogues
markedly antagonistic to the biological utilization of niacin, pyridoxine,
various amino acids, pantothenic acid, thiamin, riboflavin, ascorbic
acid, certain purines, biotin, and vitamin K. Against organisms
capable of synthesizing a given essential metabolite, the respective
analogues have in general been quite ineffective, the most notable of
the exceptions being of course, the sulfonamide group of drugs
(analogues of j&-aminobenzoic acid).

The effect of metabolite-analogues is probably exerted through
their affinity for enzymes or for the protein components of enzymes,
the substrates, or the prosthetic groups which they resemble struc-
turally. The competitive type of antagonism of the metabolite-
analogues is usually comparable to that exhibited by substrates and
their analogues competing for isolated enzyme systems, for instance,
succinate and malonate in the presence of succinic dehydrogenase.
In such cases, the normal and the abnormal compounds antagonize
one another competitively and the kinetics of the reaction usually can
be explained with mathematical precision. Although none of the
analogues of the various essential metabolites so far prepared, with the
exception of the sulfonamides, has proved to be a practical chemo-
therapeutic agent, there is reason to believe that this approach will
prove to be fruitful. It would be remarkable indeed if the groping
which led to the discovery of the sulfonamides should have yielded the
only group of compounds of practical utility in the antagonism of an
essential metabolite.

Another biochemical aspect of pharmacology to which more
attention should be given is concerned with the metabolic modification
of toxic or potentially toxic compounds. This process is generally
called "detoxication," a term which is misleading because it implies
that these compounds when metabolized are always converted into less
toxic substances. While this is generally true for compounds produced
in the body (for example, hormones), it is not always the case with
foreign substances (drugs, poisons, products of bacterial metabolism
in the intestine, such as indole and skatole). In some instances,
compounds of equal or even greater toxicity may be formed. It is

408

PHARMACOLOGY

suggested, therefore, that the term "detoxication" be reserved only for
those metabolic reactions which result in the formation of substances
less toxic than the compounds from which they are derived, and to
apply the term "metabolic alteration" to all changes which foreign
compounds undergo in the organism. Metabolic alteration will usually
be accomplished by means of oxidation, reduction, hydrolysis, or
conjugation.

It is not inconceivable that reactions resulting in detoxication
may be favorably influenced and, conversely, that reactions resulting
in increased toxicity can be inhibited. In order to investigate these
hypotheses, which could have considerable practical significance,
biological systems concerned with the metabolic alteration of foreign
compounds should be studied.

The transitory action of certain drugs in the body, as a result
of their rapid alteration, is often disadvantageous, and in some cases
the inhibition of such inactivation-reactions may be useful, as has been
shown for the inhibition of choline esterase by physostigmine or neo-
stigmine. With other transiently active compounds an inhibition of
inactivation may be difficult to accomplish, because of their suscepti-
bility to several enzyme systems which act on a variety of substrates.
Thus, epinephrine is oxidized by the lactic or malic dehydrogenase
systems, by cytochrome C (10), by amine oxidase (9), and probably
by other enzyme systems.

Richter (27) postulated that epinephrine is inactivated in the
body by conjugation with sulfate, a theory based partly on the observa-
tion that epinephrine increases the excretion of organic sulfate. How-
ever, the quantity of organic sulfate excreted after epinephrine ad-
ministration is much greater than can be accounted for by conjugation
with the drug administered (7). Since epinephrine decreases the
formation of glucuronic acid by liver slices (16), it is possible that the
hormone stimulates the formation of conjugated sulfates and decreases
the production of glucuronates without itself extensively participating
in the formation of the excreted organic sulfate.

While many foreign substances are enzymically oxidized in
the body, other compounds accept hydrogen as a result of enzymic
catalysis. In this manner trinitrotoluene (TNT) is partially reduced
in the body by flavoproteins (4). It is interesting to note that one
product of the reduction, hydroxylaminodinitrotoluene, is more toxic

409

<

''"^^^I^V

A. D. WELCH AND E. BUEDING

than the parent substance, a metabolic alteration which obviously
cannot be termed "detoxication."

As with the other mechanisms of metabolic alteration, conjuga-
tion does not necessarily result in a reduction in toxicity. Thus,
acetylation of sulfonamides renders these compounds not only thera-
peutically useless but also more toxic. Conjugations of other foreign
compounds with acetic, sulfuric, or glucuronic acids, or with glycine,
glutamine, or cysteine, may result, however, in true detoxications.
As yet, little is known about the mechanisms and enzymic reactions
involved in most conjugations. Conceivably, conjugation-reactions
may be enhanced by supplying the organism with the acids which are
found in combination with the foreign compounds, or with the pre-
cursors of the acids. Acetic acid (15), glycine (11,31), and sulfuric acid
(1) are used directly for conjugation, while three-carbon compounds,
rather than free glucuronic acid itself, are involved in the formation of
glucuronides (16). If a compound is conjugated with more than one
acid, one conjugate being more toxic than the other, a favorable result
might be attained by increasing the availability of the precursor of
the less toxic conjugate. Thus, it might be possible that the poten-
tially harmful acetylation of sulfonamides might be inhibited to some
degree if glucuronic acid conjugation (28,32,36) could be favored.

The striking decrease in the toxicity of certain pentavalent
arsenicals which results from the simultaneous administration of p-
aminobenzoic acid is at present an unexplained type of detoxication
(29). Although the decrease in the toxicity of such compounds is
accomplished without inhibition of trypanocidal action, it has been
shown (14,24) that the action of the arsonic acid derivative, atoxyl,
on Escherichia coli is antagonized by j&-aminobenzoic acid, in the same
manner as are the sulfonamides. Since the mammalian toxicity of
the sulfonamides is not antagonized by ^-aminobenzoic acid, it seems
unlikely that the pentavalent arsenicals influence an undetected
function of /?-aminobenzoic acid in animal tissues, unless, by virtue
of their different distribution, systems are affected on which the sulfon-
amides are inert. A study of the effect of /?-aminobenzoic acid on iso-
lated enzyme systems may throw light on the mechanism of this de-
toxication.

Discussion of the detoxication of the arsenicals would not be
complete without mention of the detoxifying action of sulfhydryl-

410

PHARMACOLOGY

containing compounds, particularly glutathione. In fact, extensive
studies by Voegtlin, Eagle, and their associates, indicate that the
therapeutic effect of the arsenicals in protozoan infections is dependent
on a reaction with cytoplasmic sulfhydryl groups.

As understanding of the biochemical mechanisms of energy
transfer and of its catalysis is advanced, the fundamental actions and
metabolic alterations of more and more of the known drugs and poisons
will be elucidated, and unprecedented opportunities will be afforded
for the rational development of new therapeutic agents.

References

(1) Bernheim, F., and Bernheim, M. L. C, J. Pharmacol. Exptl. Ther., 78,
394 (1943).

(2) Beyer, K. H., and Morrison, H., Ind. Eng. Chem., 37, 143 (1945).

(3) Blaschko, H., Richter, D., and Schlossmann, H., J. Physiol., 90, 1 (1937);
Biochem. J., 31, 21S7 il92>7).

(4) Bueding, E., unpublished research.

(5) Cattell, M., and Gold, H., J. Pharmacol. Exptl. Ther., 62, 116 (1938);
71, 114 (1941).

(6) Clark, A. J., in Heffter's Handbuch der experimentellen Pharmakologie.
Springer, Berlin, Suppl., Vol. 4, 1937.

(7) Deichmann, W. B., Proc. Sac. Exptl. Biol. Med., 54, 335 (1943).

(8) Fuhrman, F. A., and Field, J., J. Pharmacol. Exptl. Ther., 77, 392 (1943).

(9) Gaddum, J. H., and Kwiatkowski, H., J. Physiol., 94, 87 (1938); 96, 385
(1939).

(10) Green, D. E., and Richter, D., Biochem. J., 31, 590 (1937).

(11) Griffith, W. H., J. Biol. Chem., 69, 197 (1926); 82, 415 (1929).

(12) Haas, E., J. Biol. Chem., 155, 321 (1944).

(13) Hart, E. R., and McCawley, E. L., J. Pharmacol. Exptl. Ther., 82, 339
(1944).

(14) Hirsch, J., Science, 96, 139 (1942).

(15) Klein, J. R., and Harris, J. S., J. Biol. Chem., 124, 613 (1938).

(16) Lipschitz, W. L., and Bueding, E., J. Biol. Chem., 129, 333 (1939).

(17) Loewi, O., and Navradl, E., Arch. ges. Physiol. {PflUgers), 214, 678
(1926).

(18) McCawley, E. L., Federation Proc, 4, 129 (1945).

(19) Mcllwain, H., Bid. Rev. Cambridge Phil. Soc, 19, 135 (1944).

(20) Mann, P. J. G., and Quastel, J. H., Biochem. J., 34, 414 (1940).

(21) Mayer, N., and McCawley, E. L., Federation Proc, 4, 129 (1945).

41 I

r \

A. D. WELCH AND E. BUEDING

(22) Michaelis, M., and Quastel, J. H., Biochem. J., 35, 518 (1941).

(23) Myrback, K., Z- physiol. Chem., 158, 160 (1926).

(24) Peters, L., J. Pharmacol. Exptl. Ther., 79, 32 (1943).

(25) Quastel, J. H., Biochem. J., 25, 898 (1931).

(26) Quastel, J. H., and Wheatley, A. H. M., Proc. Roy. Soc. London, B112,
60 (1932); Biochem. J., 28, 1521 (1934); ibid., 32, 936 (1938).

(27) Richter, D., J. Physiol., 98, 361 (1940).

(28) Sammons, H. G., Shelswell, J., and Williams, R. T., Biochem. J., 35, 557
(1941).

(29) Sandground, J. H., J. Pharmacol., 78, 209 (1943); Proc. Soc. Exptl.
Biol. Med., 52, 188 (1943); Science, 97, 73 (1943).

(30) Scheff, G., and Hassko, A., Z^ntr. Bakt. Parasitenk. Orig., I, 136, 420
(1936).

(31) Schoenheimer, R., Rittenberg, D., Fox, M., Keston, A. S., and Ratner,
S., J. Am. Chem. Soc, 59, 1786 (1937).

(32) Scudi, J. v., Science, 91, 486 (1940).

(33) Seevers, M. H., and Shideman, F. E., J. Pharmacol. Exptl. Ther., 71, 373,
383 (1941); 74, 88 (1942).

(34) Silverman, M., Ceithaml, J., Taliaferro, L. G., and Evans, E. A.,
Jr., J. Infectious Diseases, 75, 212 (1944).

(35) Unna, K., J. Pharmacol. Exptl. Ther., 79, 27 (1943).

(36) Weber, C. J., Lalich, J. J., and Major, R. H., Proc. Soc. Exptl. Biol.
Med., 53, 190 (1943).

(37) Welch, A. D., Physiol. Revs., 25, 687 (1945).

(38) Woolley, D. W., cf. page 357.

412

26

SOME BIOCHEMICAL

PROBLEMS POSED BY A

DISEASE OF MUSCLE

CHARLES L. HOAGLAND, member of the rockefeller institute

FOR medical research, NEW YORK; PHYSICIAN TO THE ROCKEFELLER

HOSPITAL

THE MEDICAL biochemist hopes uhimately to provdde a
basis for the interpretation of disease phenomena which will
permit the description of a given syndrome in terms of the nature
of chemical and physiological alterations in the cellular mechanisms
of the affected organ. To achieve the synthesis of knowledge required
for this ambitious task he must rely on information supplied from
fields as dichotomous as that of the naturalist, whose efforts may fol-
low only the dictates of his curiosity, and that of the clinician charged
with the practical responsibility of alleviating disease. Situated in
this no man's land between the ill-defined borders of art and science,
the medical biochemist is in some danger of oversimplifying the com-
plex problems posed by his clinical colleagues, and of overextending
the concepts of his co-workers in biochemistry and physiology. It
follows naturally, therefore, that in cultivating the disputed stretch
between the fields of medicine and science he is frequently in jeopardy
of litigation from his neighbors on either side.

In no field of biology has intelligent effort been rewarded with
more striking or steady advances than in that devoted to a study of the
physiology and biochemistry of muscle. To a greater extent than for
any other organ, muscle has been the common meeting ground of the
biochemist, biophysicist, and physiologist. In the development of

C. L. HOAGLAND

our present knowledge concerning the underlying mechanisms of
muscular activity, the resources of practically every field of science have
been tapped. For the most part, the quality of the observations made
in the study of muscle has been high, due in a measure to the excellent
tools available for its study, and in part to the keen minds of the inves-
tigators for whom muscle biochemistry and physiology have held a
major interest.

Until recently, little interest was exhibited in the application of
information dealing with the chemical and physiological processes in
normal muscle to an elucidation of aberrant mechanisms at work in the
production of muscle disorders. As a consequence, these hetero-
geneous diseases continue to bear the confusing labels which were
given them by the classical pathologists. On occasions, serious efforts
have been made to apply the concepts and techniques of biochemistry
and physiology toward a solution of the problems presented by these
diseases, but the workers have been few, and significant results have
been scanty. Nevertheless, on the basis of what is known concerning
the normal physiology of muscle, we are perhaps in a better position
with respect to this organ than for any other to derive information of
an exact and fundamental character which would be helpful in form-
ing a common basis for clarifying the muscle syndromes. Moreover,
through a physiological consideration of the affections of muscle, a
unique opportunity is open to the biochemist by means of which he
may evaluate certain basic information with respect to its usefulness in
achieving the goal that he has set for himself, namely, that of defining
disease in terms of the nature of specific alterations in cellular
mechanisms.

Voluntary muscle constitutes nearly 43% of the total weight
of the body and is affected by a variety of diseases in which the primary
process appears to be located in the muscle fibers, and by yet another
group of diseases in which marked secondary changes in muscle are
apparent, but in which the primary process is located in the central or
peripheral nervous system. Few syndromes are more confusing than
the primary muscular disorders. The confusion is due in part to lack
of agreement on the clinical and pathological findings in these affec-
tions, and in part to the fact that, although these diseases have been
recognized as clinical entities for over fifty years, the pathogenesis has
remained obscure. It is becoming increasingly apparent that no

414

PROBLEMS OF A MUSCLE DISEASE

sharp lines can be drawn between many of the primary syndromes of
muscle, and that some unknown defect in metabolism is responsible
for the biological continuity and sequence of phenomena common to
these clinical disorders.

The problem of the secondai'y affections of muscle resolves
itself into a fundamental one having to do with the complex relation-
ship between the integrity of the innervation of muscle and the main-
tenance of an optimum state of nutrition of muscle cells. This so-
called "trophic efTect" which cells of the nervous system appear to
exert on the metabolism of the contractile cells of muscle is nearly as
complex as the problem of muscle metabolism itself, and cannot be
sharply divorced from a total consideration of other diseases afTecting
the muscular system in which no neurological component is recog-
nized.

Although the study of the muscle disorders has not kept pace
with advances in the field of biochemistry and physiology of muscle,
it has only recently received impetus from several new and important
sources of information. The discovery of a disease of dietary origin
in animals which resembles progressive muscular dystrophy, but
which, unlike the disease in man, is cured with vitamin E, is influencing
the study of human affections of muscle. The clinical observation
that prostigmine relieves to a variable extent th^ symptoms of myas-
thenia gravis has accelerated fundamental studies on the metabolism
of acetylcholine in this disease. The discovery of an aberration in
potassium metabolism in periodic muscular paralysis had led to some
success in a search for efTective therapy for this disorder, and to the
institution of basic studies on the relation of periodic paralysis to other
disturbances in muscle function associated with changes in the con-
centration of potassium. The presence of a disease in goats, indis-
tinguishable symptomatically from congenital myotonia in man, has
greatly facilitated studies on the group of myotonic maladies. In the
near future, therefore, increasing knowledge of the physiology of
muscle, particularly that pertaining to the physiology and chemistry
of contraction, transmission of excitation from nerve to muscle, and
the mode of action of specific pharmacological agents on the myoneural
junction, should provide many new and effective means for an attack
on the problem of the muscle diseases as a whole.

No one syndrome among the diseases of muscle has provided

C. L. HOAGLAND

more challenge to the clinician than progressive muscular dystrophy.
This disease, first clearly described by Gowers (6), is characterized by
primary degeneration and atrophy of voluntary muscles of the ex-
tremities, pelvis, and shoulder girdles. Atrophy is marked in all
forms, and hypertrophy or pseudohypertrophy is an early and promi-
nent symptom in one form. The disease has certain well-established
familial aspects, although the mode of genetic transmission is not
always clear. In the most common type, pseudohypertrophic mus-
cular dystrophy, the onset usually occurs before the fifth or sixth year.
The prognosis is grave, and death frequently results within several
years after the disease has become clearly manifest. No specific
therapy for the syndrome is known and little in the way of supportive
therapy has proved useful (16).

Morphological studies on progressive muscular dystrophy have
yielded little information, aside from the fact that there is hyaline
degeneration and fragmentation of muscle fibrils, and that the muscles
show marked increase in fat content. The recent development of a
simplified quartz microscope, with the 2537 A. line of mercury as the
light source, has made it possible to obtain ultraviolet photomicro-
graphs of muscle fixed and sectioned by current methods, and in cer-
tain instances on surviving specimens without previous fixation (8).
Photomicrographs of muscle made by this technique show selective
absorption of varying intensity, and reveal more detail than those
obtained with visible light on stained material. The technique has
particular advantages in the study of muscle when it is desired to
correlate morphological and physiological changes. In ordinary light
the appearance of unstained muscle is misleading, since the image
under these conditions is due to inhomogeneity of tissue and not to
the presence of material with selective absorption.

Description of the histological and pathological changes re-
vealed by ultraviolet light is difficult at the moment, because of the
inappropriateness of classical nomenclature for description of images
produced by specific absorption of light rather than by staining and
physical inhomogeneity of tissues. Certain analogies can be drawn,
however, between structures photographed in ultraviolet light and
those structures which are seen in visible light as a result of the appli-
cation of tissue stains, on the basis of size and location of the structures
within a given cell. It is evident from these studies that ultraviolet

416

PROBLEMS OF A MUSCLE DISEASE

photomicrography to be developed fully will require revision and re-
definition of the nomenclature of classical histology in order to include
those structures which are due to selective absorption as well as those
which are due to the effect of physical inhomogeneity and contrasting
stains.

Sections of human muscle photographed in ultraviolet light
reveal deeply absorbing transverse zones, spaced at regular intervals,
separated by alternate zones of low absorption occurring throughout
the length of the muscle fibers. The over-all appearance of the photo-
micrographs is the familiar one of cross striation which has long been
associated with voluntary muscle. Caspersson, using indirect methods,
recently concluded that the isotropic, or dark, striae of muscle revealed
by polarized light corresponds in position to the strongly absorbing
areas seen in ultraviolet light (3). It has been suggested, moreover,
that the absorption which is characteristic of these zones is due chiefly
to their content of adenylic acid or adenosine triphosphate. A direct
comparison of the position of the striae appearing in polarized light
with that occupied by striae appearing in ultraviolet light was made
recently, in our laboratory, by comparing half of a field of a given
muscle section photographed in polarized light with the remaining
half photographed in ultraviolet light. When identical magnifications
of the images are compared, inspection of the matched photomicro-
graphs leaves no doubt that isotropic zones in muscle revealed in
photomicrographs in polarized light correspond exactly to the deeply
absorbing striae revealed in the ultraviolet. In sections of muscle
taken from cases showing early signs of pseudohypertrophic muscular
dystrophy, the transverse zones were very faint, and in some areas
almost indiscernible, in marked contrast to the broad compound bands
observed in transverse sections of normal muscle. In cross sections,
areas of absorption which in normal muscle were punctate and equi-
distant in character, were revealed as coalescent areas in a state of
apparent disorganization.

That the method of simplified ultraviolet photomicrography
possesses marked advantages over the classical methods of histology
and pathology, which depend on staining and examination of speci-
mens in visible light, has been brought out convincingly in a study of
the histopathological changes associated with muscle disease. In
this respect muscle tissue has proved to be a happy choice. Because

417

C. L. HOAGLAND

of its specialized architecture, a fairly satisfactory comparison can be
made between structures revealed in ultraviolet light and those which
are observed after staining and photomicrography in the visible region
of the spectrum. Not only is greater resolution achieved with ultra-
violet light photography, but the image which is obtained may provide
in addition some idea of the chemical nature of the absorbing areas of
the tissue, since it results from the selective absorption of light by
proteins and by substances of high absorptive capacity, such as the
purines and pyrimidines.

A plethora of micro methods is available for the quantitative
determination of tissue constituents, and for almost every type of inter-
mediary compound known to arise in metabolism. However, inter-
pretation of quantitative data secured by analysis of diseased muscle
is enormously complicated by lack of information concerning the total
mass of muscle cells in these specimens due to their variable content
of fat, fibrous connective tissue, collagen, and water. It may be
highly desirable to determine quantitatively the various mineral con-
stituents of diseased muscle, and to analyze the material for its content
of creatine, adenylic acid, and various organic esters of phosphorus.
The results have little significance, however, unless some suitable basis
of reference can be found which will permit comparison of the mass
of muscle cells in one specimen with the mass of cells in another.
Two specimens of muscle removed at biopsy from approximately the
same area, in a subject showing incipient progressive muscular dys-
trophy, may show differences of 50% in fat content, 20% in concen-
tration of collagen and fibrous connective tissue, and 15% in the
content of free water and ash. Low values for creatine phosphate,
adenosine triphosphate, or other organic constituents in specimens of
diseased muscle have little significance unless some factor of correction
can be employed which will allow for the differential mass of muscle
cells present in the samples. Some success was achieved recently in
our laboratory in the use of myosin as a base of reference for certain
organic constituents of diseased muscle. Since myosin is the principal
protein component of the contractile cells of muscle, its concentration
must bear an important quantitative relationship to the total mass of
contractile cells in the specimen taken for analysis. Owing to its
property of critical solubility in potassium chloride, myosin can be
determined quantitatively, and with a fair degree of accuracy, on,

418

PROBLEMS OF A MUSCLE DISEASE

specimens of muscle weighing not more than one hundred milH-
grams.

The problem of adequate bases of reference for constituents of
diseased tissue in general is one of the principal barriers to the develop-
ment of a rational discipline of chemical pathology. It is becoming
increasingly evident that practical methods for determination of the
residual mass of parenchymal cells of affected organs will have to be
devised before full consideration can be given to possible alterations in
concentration of enzymes, substrates, metabolites, and metallic cata-
lysts occurring in tissue as a result of affection by disease. Since, in
the chemical analysis of diseased organs, we are going to be faced
constantly with the need of knowing the net quantities of various
tissue constituents, it is safe to assume that increasing attention will
be paid in the future to the problem of the relation of the concentration
of tissue constituents to differential cellular mass.

Some of the difficulties inherent in the selection of a proper base
of reference for constituents of diseased tissue can be avoided by ascer-
taining the ratio between the concentrations of two substances in a
given specimen of material and comparing it with the ratio which is
found for these constituents in similar tissues of normal organs. Not-
able among these is the familiar respiratory quotient, which, since it
measures the ratio of the production of carbon dioxide to oxygen con-
sumption, can be considered independently, within certain limits, of
the variable extraneous constituents of the specimen. Studies of the
respiratory quotient by means of the method of Dickens and Simers (4)
of specimens of muscle removed from patients exhibiting various stages
of progressive muscular dystrophy did not reveal values differing sig-
nificantly from those obtained for specimens of normal muscle removed
during surgical operation (17). The carbon dioxide and oxygen
quotients were extremely low in the diseased tissue, but the ratio of
carbon dioxide production to oxygen consumption was normal. It
should be pointed out that, with respect to the oxygen consumption,
specimens of surviving muscle from patients with progressive muscular
dystrophy differ markedly from specimens of muscle taken from animals
which have been deprived of vitamin E. Houchin and Matill have
shown that muscle tissue removed from animals on diets deficient in
vitamin E was characterized by marked increase in oxygen consump-
tion (10), and, moreover, that the oxygen uptake of the muscle speci-

C. L. HOAGLAND

mens was rendered normal by the in vitro addition of a-tocopherol
phosphate (9).

Notwithstanding the fact that progressive muscular dystrophy
has engaged the interest of clinicians since the original description of
the syndrome in the middle of the 19th century, almost no attempts
were made to study the general metabolism in this disease until the
pioneer work of Levene and Kristeller (11) in 1909. These workers
showed that the feeding of protein resulted in an excessive excretion
of creatine in patients with progressive muscular dystrophy. Subse-
quently, many studies have revealed that there is marked derangement
in metabolism of creatine in this disease, and that endogenous creatine,
formed from protein and amino acids, is not retained by the muscles
as effectively as in normal subjects. This observation has given rise
to the concept that there is, in progressive muscular dystrophy, a
diabetic-like state with respect to the ability of the patient to retain
either ingested creatine or creatine which is formed endogenously
from proteins and amino acids. Whether or not this is a true concept,
the recognition of a biochemical aberration in creatine metabolism is
perhaps the only truly significant contribution to have been made
within the last thirty years toward an understanding of the essential
nature of this disease.

Of perhaps even greater significance than an increase in excre-
tion of creatine in progressive muscular dystrophy is a diminished out-
put of creatinine. Because of the specific association of creatinine
formation with the integrity of muscle processes, the urinary concen-
tration of this material may, in certain cases, give a more reliable indi-
cation of the severity of the disease than the level of urinary creatine (7).
Moreover, the relatively great constancy in the excretion of creatinine,
even under widely diff'ering conditions of diet and health of the sub-
ject, permits the attachment of greater significance to small changes in
the urinary concentration of this material than is the case with crea-
tine, which may show wide fluctuations from day to day. The fact
that creatinine is not derived from all tissues, but that it arises as a
special process of tissue catabolism taking place "largely if not wholly
in the muscles," has given rise to the belief that the amount of creatinine
excreted in the urine "bears a direct relation to the potential elficiency
of the muscles and is a reliable index of the muscular development of
an individual" (15). The belief, originally stated by Foiin, that

420

PROBLEMS OF A MUSCLE DISEASE

creatinine formation represents a type of endogenous metabolism dis-
tinct from the exof^enous metabolism of food protein (5) is no longer
tenable, as a result of the studies performed by Bloch and Schoen-
heimer with the aid of isotopic nitrogen (14). The constant daily
excretion of creatinine, in contrast to the highly inconstant excretion
of other nitrogenous constituents of the urine, however, indicates that
the formation of creatinine is an orderly and well-regulated process,
the biological significance of which is not entirely apparent. That it
arises directly from creatine is abundantly clear. Moreover, there is
recent evidence to indicate that it arises in the process of dcphosphoryl-
ation of creatine phosphate, and that its formation is intimately con-
nected with the phenomenon of muscular activity (2,13).

The marked decrease in the output of creatinine observed con-
sistently in patients with progressive muscular dystrophy has been
attributed to a reduction in the total mass and cfTiciency of the mus-
cular system in this disease. In the normal subject there is a rather
remarkable relationship between the level of urinary creatinine and
the total mass of the musculature. By taking the muscle mass of the
normal subject as 43% of the body weight, a fairly quantitative idea
of the mass of functioning muscle in patients with progressive muscular
dystrophy can be obtained from a consideration of the value of the
urinary creatinine. This value has been found to vary from 35 to
40% in cases of incipient progressive muscular dystrophy to values as
low as 17% in patients in whom the disease was advanced. In our
laboratory, a marked correlation has been observed consistently be-
tween the muscle mass calculated as per cent of body weight and the
degree of physical performance of patients with dystrophy, in so far
as such performance could be appraised in a quantitative fashion.
Calculation of the muscle mass on the basis of creatinine excretion
appears to yield important information regarding the extent to which
the muscular system is involved in patients with this disease. More-
over, by performing this type of calculation at frequent intervals on a
given subject with progressive muscular dystrophy, a fairly quantita-
tive appraisal of the clinical condition of the patient and the rate of
progression of the disease at various intervals can be ascertained.

Following the elucidation of the role of methionine in the syn-
thesis of creatine and creatinine by Bloch and Schoenheimer (1) and
du Vigneaud (19), speculation arose in the laboratory over the possi-

421

C. L. HOAGLAND

bility that a deprivation of methyl groups may occur in certain diseases
characterized by excessive creatinuria. It has been shown that growth
in rats is inhibited by feeding of excess glycocyamine, and that growth
is resumed following the administration of choline or methionine (18).
Although alternative hypotheses must be entertained, the results of
these experiments indicate that, in the presence of an excess of sub-
stances which act as methyl acceptors, a loss of methyl groups may
occur, and that this deficiency may be prevented by an adequate intake
of methylating agents such as methionine and choline. Studies on
the total methyl output in patients with progressive muscular dystrophy
revealed that the loss in methyl groups, occurring as a result of exces-
sive creatine output, and greater than that seen in normal children,
was fully compensated for by a diminution in output of creatinine (7).
A surprising agreement was also found between the total quantity of
"methyl" excreted by the group of normal subjects and that excreted
by patients with progressive muscular dystrophy. These data suggest
that there is no absolute increase in the output of total creatine com-
pounds in progressive muscular dystrophy, and that the diseased
patient differs from the normal, with respect to creatine output, in the
differential partition of creatine and creatinine compounds in the
urine. In view of the fact that creatinine has been shown by isotope
studies to be derived from creatine, it would appear from a considera-
tion of these results that, in progressive muscular dystrophy, crea-
tinuria does not arise as a result of an increase in the synthesis of crea-
tine, but rather as a result of the incomplete metabolism of this material
in the muscle, with a consequent decrease in the amount converted to
creatinine.

In view of the intimate association existing between the exo-
thermic decomposition of adenosine triphosphate and the integrity of
the contractile processes of normal muscle, no consideration of muscle
disease would be complete without some discussion of the metabolism
of this compound. Moreover, since the discovery by Lyubimova and
Engelhardt of the adenosinetriphosphatase activity of myosin (12),
speculation on the nature of myosin in muscle disease would also
appear to be in order. The concentration of adenosine triphosphate
in the muscles in progressive muscular dystrophy was found to be
considerably lower than in specimens of muscle taken under similar
conditions and from corresponding areas in normal subjects; and only

422

PROBLEMS OF A MUSCLE DISEASE

traces of the compound were found in the muscles of patients in whom
the disease was advanced. However, absolute values for the concen-
tration of adenosine triphosphate in dystrophic muscle have not yet
been obtained, because of the factors which have been previously
discussed, namely, difficulties inherent in correction of such values for
the variable content of fat, connective tissue, water, etc., of the diseased
muscle. The difficulties are further enhanced in this instance by the
extraordinary lability of adenosine triphosphate. Determination of
the adenosinetriphosphatase activity of myosin prepared from speci-
mens of diseased muscle has been more feasible. Although the myosin
content of biopsy specimens of muscle from dystrophic patients appears
to be relatively low, it was probably not lower than would be expected
from a consideration of the diminished concentration of contractile
cells in the specimens. The enzymic activity of myosin prepared
from the diseased muscles was approximately of the same order of
magnitude as that prepared from normal muscle, when tested for its
ability to dephosphorylate adenosine triphosphate.

Although attention has been given thus far mainly to a con-
sideration of the chemical and physiological e\'ents occurring in the
muscles in progressive muscular dystrophy, there has been no adequate
evidence presented to show that the primary seat of the disease is
actually contained in the musculature. The possibility that the
muscular phenomena may be only secondary to an inherent defect
arising elsewhere must be kept constantly in mind, lest we become so
preoccupied with the organ in which the major symptoms arise that
we fail in our over-all search for the site of the initial disorder which
sets the muscular syndrome in operation.

Local and systemic changes in metabolism observed thus far in
progressive muscular dystrophy have been quantitative rather than
qualitative. Unless, and until, genuine qualitative changes in inter-
mediary metabolism in this disease can be demonstrated, the syndrome
will continue to escape classification with the rare diseases of inborn
metabolism such as alkaptonuria, cystinuria, and others in which
such changes are apparent.

Considered on the basis of incidence in population, progressive
muscular dystrophy cannot be regarded as a particularly important
syndrome, since happily for humanity it is a comparatively rare
disease. Viewed from the standpoint of what the successful solution

C. L. HOAGLAND

of this complex disorder would mean to the future development of
medical biochemistry, however, few diseases offer more exciting
potentialities. It will be recalled that earlier, in a consideration of
those tissues which have been of greatest use in a study of the chemical
and physical laws of living matter, scarcely a moment's reflection was
needed to place muscle first. Indeed, it was indicated that the physio-
logical basis of activity of any organ or organ structure as we regard
it today is but largely an extension of concepts first derived for muscle.
Is it too much to hope, therefore, that eventually the solution of the
clinical syndrome which we recognize as progressive muscular dys-
trophy may provide many of the new and fundamental concepts we
shall require in order ultimately to achieve the description of disease
in terms of the nature of altered physiological and chemical mecha-
nisms at work in the cells and tissues of the affected organism?

References

(1

(2
(3
(4

(5

(6
1

(7

(
(8

(

(9
(10
(11
(12
(13
(14
P
(15
(16

Bloch, K., and Schoenheimer, R., J. Biol. Chem., 138, 167 (1941).

Borsook, H., and DubnoflF, J. W., Ann. Rev. Biochem., 12, 183 (1943).

Caspersson, T., and Thorell, B., Acta Physiol. Scand., 4, 97 (1942).

Dixon, M., Manometric Methods. 2nd ed., Cambridge Univ. Press,
London, 1943, Chapter 7, p. 94.

Folin, O., J. Biol. Chem., 17, 469, 475 (1914).

Cowers, W. R., Pseudohypertrophic Muscular Paralysis. Churchill, London,
879.

Hoagland, C. L., Gilder, H., and Shank, R. E., J. Exptl. Med., 81, 423
945).

Hoagland, C. L., Shank, R. E., and Lavin, G. L, J. Exptl. Med., 80, 9
944).

Houchin, O. B., J. Biol. Chem., 146, 313 (1942).

Houchin, O. B., and Mattill, H. A., J. Biol. Chem., 146, 301 (1942).

Levene, P. A., and Kristeller, L., Am. J. Physiol, 24, 45 (1909).

Lyubimova, M. N., and Engelhardt, V. A., Biokhimiya, 4, 716 (1939).

Rosengart, V., Bull. Med. Coll. Dnepropetrovsk, U.S.S.R., 2, 87 (1940).

Schoenheimer, R., The Dynamic State of Body Constituents. Harvard Univ.
ess, Cambridge, 1942.

Shaflfer, P., Am. J. Physiol., 23, 1 (1908).

Shank, R. E., Gilder, H., and Hoagland, G. L., Arch. Neurol. Psychiatry,
52,431 (1944).

424

PROBLEMS OF A MUSCLE DISEASE

(17) Shank, R. E., and Hoagland, G. L., unpublished data.

(18) du Vigneaud, V., Chandler, J. P., Cohn, M., and Brown, G. B., J. Biol.
Chem., 134, 787 (1940).

(19) du Vigneaud, V., Gohn, M., Ghandler, J. P., Schenck, J. R., and Sim-
monds, S., J. Biol. Chem,, 140, 625 (1941).

425

27

PHYSIOLOGY
AND BIOCHEMISTRY

SURGEON CAPTAIN C. H. BEST, C.B.E., F.R.S., director of

MEDICAL RESEARCH, ROYAL CANADIAN NAVY

^HE RESPONSIBILITY of preparing an essay under this
-^ broad heading, even in normal times, would be very great.
Since, under present circumstances, a comprehensive treatment of the
subject would be still more difficult, with the consent of the editor I
am limiting my discussion to fields in which I have had personal ex-
perience.

Insulin and Diabetes

Four of the central problems in this field are: (a) the mecha-
nism of action of insulin; (b) the further study of the insulin molecule
and its synthesis; (c) the etiology of diabetes; and (d) the improve-
ment of insulin as a therapeutic agent.

Mechanism of Insulin Action. The broad picture seems bright
and clear. The administration of insulin to recently depancreatized
dogs or to the uncomplicated case of human diabetes completely restores
the organism. If the treatment is carefully continued and no com-
plications occur, the animal or patient may proceed with an essen-
tially normal existence and is, in fact, difficult to distinguish from the
nondiabetic. The deranged metabolism of carbohydrates, fats, pro-
teins, and phosphate compounds is corrected, except that the lack of a

427

C. H. BEST

physiological mechanism for the increased or decreased rate of libera-
tion of insulin in the severe diabetic permits wider fluctuations from
the normal range of values.

The detailed mechanism of the action of insulin is by no means
as clear as its specific effect on the diabetic state might suggest. It
was soon learned that the formation of the key polysaccharide —
glycogen — was stimulated by insulin and that the burning of sugar
was increased. The extremely wasteful and very dangerous break-
down of proteins to sugar and of fat to ketone bodies in the liver is
checked by insulin. The formation of fat from sugar is accelerated;
and there is some evidence that certain of the phosphate compounds
of paramount importance in the provision of energy for muscular
contraction may be regenerated at a greater rate when insulin is
present. Some workers believe that the amount of adenosine tri-
phosphate in the liver is increased by the action of insulin.

A very great deal has been learned in recent years which adds
to our understanding of the pathway of carbohydrate storage and
breakdown in the body. It would appear that most, if not all, of these
steps may be traversed without the action of insulin, but our knowledge
of these processes is not complete. A great step forward is represented
by the recent demonstration by Cori et al. that insulin releases the in-
hibition of the conversion of glucose to glucose-6-phosphate, which is
exerted by anterior pituitary extract.

While in a diabetic animal deposition of glycogen in both liver
and muscle is greatly increased by insulin, in the normal animal muscle
appears to have a "higher priority" and deposition of glycogen in this
tissue, after insulin administration, may be accompanied by a fall in
glycogen content of liver.

We have a great deal more to learn concerning the best estab-
lished effect of insulin, i. e., the promotion of glycogen deposition, and
we have even more to learn about the mechanism by which the break-
down of protein and fat is inhibited. It is possible that insulin in
some way controls the substrate which is presented to the tissue cells
for metabolism.

The intelligent use of such labeling agents as radioactive phos-
phorus and the stable isotopes of carbon, nitrogen, sulfur, and other
elements is certain to illuminate many of the dark passages through
which insulin passes in producing its effect on diabetes. We can

428

PHYSIOLOGY AND BIOCHEMISTRY

confidently expect further advances in this field, which has been
widened and cleared even during the present war.

We have no reliable method for the estimation of insulin in
small quantities of blood. Perfection of a technique for this purpose
would aid, not only in diagnosing the type of diabetes and perhaps in
facilitating the treatment, but in throwing a great deal of light on the
very real possibility that insulin concentration in the blood is a factor
in the control of insulin liberation from the pancreas. We need either
a sensitive chemical test or a microbiological procedure that is capable
of detecting the small amounts of the antidiabetic hormone which are
certainly present in varying quantities in blood. Thus far, very few
vigorous attacks have been made on this important problem.

Chemistry of Insulin and the Possibility of Synthesis. Ele\ en
amino acids have now (1945) been isolated from the crystalline insulin
protein. Little progress has been made in the identification of any
active chemical grouping which might exert an antidiabetic effect in
the absence of the whole insulin molecule. The synthesis of this
complex protein awaits further development in protein chemistry;
there are so many ways possible in which the various amino acids
might be linked that few investigators have thought it worth while to
give the matter serious consideration at this time. The problem is
attractive enough for its own sake, but when we consider that, if the
increase in insulin distribution persists at the present rate, all the
available pancreases in the world will be utilized for insulin produc-
tion in the not distant future, the situation acquires a new urgency.
For the last fifteen years, the total distribution of insulin in the United
States and Canada has doubled every five years — that is, the amount
used in 1935 was twice that in 1930, the amount in 1940 twice that
of 1935, and the same rate of increase has been maintained up to the
present time.

A vigorous search should be made for a source of insulin other
than mammalian pancreas. We know that considerable quantities
can be obtained from the principal islets of certain fish but this pro-
cedure has never been placed upon a commercial basis.

Etiology of Diabetes. There are no data available to tell us
what proportion of diabetic cases is produced by a primary lesion of
the pancreatic islet cells. There are many well-substantiated state-
ments in the literature which indicate that there is no obvious lesion

C. H. BEST

of the pancreas in many instances of human diabetes. On the other
hand, there is no convincing proof that the islet cells are really normal
in these cases; and many investigators feel that the burden of proof
that the pancreas is not primarily or secondarily involved rests with
those who make this assertion. Careful estimations of the insulin
content of pancreas, if it were possible to do this under reasonably
standard conditions in autopsies on diabetics, would yield valuable
results. It is obvious, however, that the opportunities to make this
study in uncomplicated cases will be rare. The administration of
insulin or fasting will lower the insulin content of the pancreas in
normal animals, a point which must be considered when insulin
estimations are made.

In experimental animals, permanent diabetes can be produced
by three procedures: by pancreatectomy, by destruction of the islet
tissue by the diabetogenic material of the anterior pituitary gland, and
by the administration of alloxan. All three methods effect removal
or practically complete destruction of the beta cells of the islands of
Langerhans. Substances are, however, present in the anterior
pituitary gland which aggravate diabetes in the completely depan-
creatized organism. A part of this effect appears to be due to a
deficient carbohydrate utilization in peripheral tissues.

Apart from the possibility of developmental defects in islet
structure which would presumably appear in very young children and
the well-established arteriosclerotic changes in the pancreas of older
individuals, there is little real evidence on which a reasonable theory
of the etiology of diabetes may be supported. The association of the
diabetic state with overactivity of the anterior pituitary lobe of the
pituitary, as in the acromegalic, encourages further study along a
path which has already been well outlined. We must develop accurate
procedures for the assay of the diabetogenic materials in blood. It
has recently been shown that diabetes may be most favorably affected
by removal of the exciting cause which, in a few clinical cases, has
been found in one of the adrenal glands. Extracts containing the
hormones of the cortex of the adrenal may produce diabetes in the
partially depancreatized animal, and it is well established that, under
certain conditions, removal of the adrenal cortices ameliorates the
diabetic state caused by pancreatectomy in much the same way as
does removal of the anterior lobe of the pituitary.

PHYSIOLOGY AND BIOCHEMISTRY

The fact that, in the experimental animal and presumably in
many diabetic patients, eight-tenths of the insulin-producing capacity
of the islet cells must disappear before it is possible to detect the insulin-
producing capacity by any procedure except direct examination, pre-
sents a challenge to clinicians and experimentalists alike. It is incon-
ceivable that a relatively clear-cut problem of this kind can long resist
the onslaught of the vigorous minds which we hope will attack it now that
peace is restored. Many physiological and chemical avenues of ap-
proach have not as yet been adequately explored.

Improvement of Insulin as a Therapeutic Agent. The prob-
lem of giving insulin in a more physiological way presents many diffi-
culties. It will not be easy to make a compound of insulin which is
broken down at an increasing rate as the sugar content of the tissues
rises. Improvements in protamine zinc insulin are quite possible and
should certainly engage the attention of those who have already made
great strides in this field. There are indications that a decrease in the
amount of protamine with respect to insulin, may serve a useful pur-
pose. Clear solutions of combined insulin which produce a prolonged
effect are desirable, but clinical opinion indicates that a completely
satisfactory one is not as yet available. Insulin by mouth for mild cases
of diabetes is by no means an impossibility, but it will probably always
remain a wasteful procedure and in cases in which dosage of insulin
must be very accurately regulated, a hazardous one.

It should be possible in the future to make a more definite
separation of the diabetic cases into those which have retained a definite
capacity for producing insulin and those in whom this function of the
islet cells is irretrievably lost. In the former type, it is obviously use-
less to plan diets with the idea of conserving islet function. In experi-
mental animals, diabetes resulting from partial removal of the pancreas
or from administration of the diabetogenic material of the anterior
pituitary gland, may now, under certain conditions, be prevented, or
in its early stages cured, by the appropriate use of insulin and diets
which do not tax the capacity of the remaining islands of Langerhans.
These findings have aroused the hope that a similar application may
be made in those cases of human diabetes in which a reasonable
number of functioning islet cells remains. This application will not
be easy until the potential human diabetic can be recognized much
earlier than is possible at present and until much more light is shed

C. H. BEST

on the etiology of the diabetic state in man. We have suggested that
it will be important to determine, by actual clinical trial, the extent
to which the results of these animal experiments are applicable to the
human subject. We have not suggested that insulin should be used
prophylactically in children with a diabetic family history until the
experimental results have been carefully tested by competent clinicians.
We have outlined many of the obvious difRculties which stand in the
way of such a clinical research. There may be many more which are
not apparent to the experimentalist.

The four problems discussed briefly are a few among the many
in this field which demand investigation. We can look forward with
confidence to the development of better diets for diabetics; and a
lead may have been supplied by the suggestion that certain fats may
be used as a source of energy without the production of excessive
amounts of the dangerous ketone bodies. The role of some of the
new accessory food factors in the intensity of diabetes has recently been
investigated and the use of small animals made diabetic by the injec-
tion of alloxan will surely be of great value.

It seems assured that there will be a better control of infections
which in many cases constitute the main problem in the treatment of
diabetics. Penicillin by its gentle and effective action may well intro-
duce a new era of progress in relation to the control of infection in
diabetic cases.

Histaminase and Histamine

Some sixteen years ago, we suggested the term "histaminase"
for the enzyme system, shown to be present in various tissues, which
inactivates histamine. This system has been partially purified as a
result of the efforts of various investigators, although a great deal
more remains to be done in this field as well as in the delineation of
its physiological significance. Histaminase has been useful, in con-
junction with other procedures, in the identification of histamine in
various tissue extracts and fluids. It is present in liver, lung, kidney,
and the intestinal mucosa, but is absent from the gastric mucosa.

A great deal has been learned about variations in histamine
content of tissues and about the liberation of histamine in anaphylactic
shock, whereas the physiological role of this substance is far from

432

PHYSIOLOGY AND BIOCHEMISTRY

completely established. A specific chemical procedure for the deter-
mination of histamine would be useful in any subsequent researches.
The present methods are not specific because they measure the amount
of the imidazole nucleus only.

It has been reported relatively recently that histamine may act
as a hapten when it is conjugated with certain proteins and that it is
possible to produce antibodies which are partially specific for the
histamine hapten.

While the histamine-histaminase system is a very useful one
for studies in enzyme chemistry, attempts at commercial exploitation
of the enzyme are without justification. There is no physiological
basis for the claim made by some clinicians that histaminase exerts
therapeutic properties.

Heparin

The discovery of heparin at Johns Hopkins University in 1918
stimulated a new series of researches on blood coagulation. The
anticoagulant, isolated in crystalline form in Toronto in 1935, prevents
the clotting of one hundred thousand times its own weight of blood.
The experimental and clinical demonstration that the pure substance
prevents thrombosis without the production of any harmful eff'ects
opened the way for its clinical use and aroused anew the interest of
physiologists and biochemists.

Heparin is one of the mucopolysaccharides. These compounds
are of considerable importance in physiology, as they are the chief
constituents of connective tissues. The variation in the properties
of this group of substances apparently permits the change in character-
istics of connected tissues. Heparin is the first substance of this group
to be crystallized. Its basic unit consists of two glycuronic acid resi-
dues, two glucosamine residues, and five sulfuric acid residues. Its
high content of sulfuric acid probably makes it the strongest organic
acid found in the body. Since no other mucopolysaccharide has any
significant anticoagulant activity, this property must depend on some
unique configuration of the molecule. Of considerable interest is the
fact that, while heparin in any species is identical from tissue to tissue,
on comparison of heparin from diff"erent species variations appear.
Thus, the relative anticoagulant potencies of the pure heparins from

433

C. H. BEST

sheep, pork, beef, and dog tissues are 1:2:5:10. While proteins dif-
fer with the species, there are few known examples of species variations
occurring with polysaccharides. Chemical studies of heparin would
therefore be of interest from several points of view.

The body contains enormous amounts of heparin. A dog
weighing ten kilograms contains approximately ten thousand units of
heparin, or sufficient to prevent clotting of ten kilograms of blood.
This is all contained in the tissues, since none can be detected in
normal blood. What is the function of such a large store of heparin?
It has been suggested that it provides an emergency mechanism which
can prevent intravascular thrombosis by the liberation of heparin into
the vessel threatened by thrombosis. Histological studies have demon-
strated that heparin is localized in the mast cells located in the walls
of blood vessels. Such a position would be suitable for a function of
this nature. Heparin is liberated from the mast cells and found in the
blood in peptone and anaphylactic shock. What is the initial stimulus
to heparin liberation? Is it the appearance of thrombin, of fibrin, or
the agglutination of platelets? How much heparin is liberated in
twenty-four hours? Is this increased when there is tissue damage,
as after operations? Does the heparin content of the mast cells and
tissues remain constant, or does it change with age, sex, operations,
etc.? Many of these questions could be quickly answered if we had
accurate methods for the determination of heparin in blood and tissues.
However, the amount in blood is too small for estimation by our present
methods, and the only procedure applicable to other tissues is to
attempt isolation by the method developed for one particular tissue —
beef lung.

On injection, heparin rapidly disappears from the blood. An
enzyme, heparinase, has been prepared from rabbit liver. The puri-
fication and study of the distribution of this enzyme has not yet been
attempted.

Heparin prevents the clotting of blood and the agglutination
of platelets, but we still do not know the mechanism of the action.
Heparin when added to oxalated plasma and thrombin prevents the
clotting of this system. However, if fibrinogen is substituted for the
plasma, heparin has no effect. This action of heparin on clotting
therefore depends on the presence of some unknown plasma factor.
The anticoagulant efTect of heparin on blood does not depend on this

434

PHYSIOLOGY AND BIOCHEMISTRY

antithrombic action of heparin because heparinized blood contains
the normal amount of prothrombin. Since, however, the nature of
the interaction of prothrombin, thromboplastin, calcium, and other
factors is at present unknown, it is not possible to indicate the particular
reaction in clotting prevented by heparin.

It can be seen that many questions remain unanswered in the
heparin field, the solution of the problems depending upon the develop-
ment of new methods and experimental approaches.

The Lipotropic Factors

The recognition of the significance of choline as a dietary factor,
and the rapid development of our knowledge of other lipotropic agents
such as betaine, methionine, and inositol, have opened a new chapter
in physiology and have stimulated almost as much experimental work
as did the isolation of insulin. The presence of choline or methionine
in the diet is essential for the survival of young animals of various
species. Death is apparently due to a failure of liver or kidney func-
tion. Three of the main problems demanding investigation are:
(a) the mechanism of action of lipotropic agents; (b) the physiological
significance of the lipotropic factors; and (c) the clinical significance
of the lipotropic factors. Choline apparently acts by virtue of its
incorporation as an intact unit into the molecule of phosphatidyl
choline (lecithin) ; and, as a result of this step, the rate of regeneration
of the choline-containing lipids of liver, kidney, blood, and perhaps
other tissue is increased. Betaine acts as a donator of one of its methyl
groups which, by methylating ethanolamine, forms choline. Me-
thionine is also a methyl donor; while it may also exert dramatic
efTects by providing a source of organic sulfur or perhaps by virtue of
its intact molecule, its lipotropic action is accomplished by donation
of its methyl group to ethanolamine, i. e., it acts as a precursor of
choline. Inositol, which has been well established as a lipotropic
factor under certain conditions, has been reported, without adequate
evidence, to aflfect particularly the cholesterol fatty liver produced by
various procedures. It may stimulate the regeneration of inositol-
containing phospholipids but this phospholipid has not, as yet, been
isolated from liver, the only tissue on which inositol has been shown to
exert a lipotropic action.

435

C. H. BEST

Physiological Significance. It remains to be decided whether
a supplement of choHne is necessary in a diet which contains an ade-
quate amount of choHne precursor, methionine. A small part of the
lipotropic action of protein may be due to the tyrosine it contains.
The lipotropic action of inositol is inhibited in young animals by some
component of corn oil, perhaps the essential fatty acids. Work which
is in progress should determine the relative importance of these various
agents under physiological conditions. The addition of choline to the
diets of puppies given an alipotropic ration, prevents failure of liver
function and death in this species. Presumably, methionine would
have the same effect. Choline or methionine prevents the develop-
ment of haemorrhagic kidneys which may lead to the death of young
rats through failure of kidney function. The extension of this work
may throw some light on the etiology of developmental defects in
kidneys in man.

Hepatic necrosis in rats is prevented by cystine and hepatic
cirrhosis by choline. Methionine, by providing a source of cystine or
of organic sulfur, prevents necrosis and, by enabling the body to make
choline from ethanolamine, prevents cirrhosis.

Clinical Significance. The provision of diets high in protein
may exert a favorable effect on hepatic cirrhosis in man. It is reason-
able, in the light of the experimental evidence, to consider that this
effect may be due to the methionine contained in protein which acts
as a source of choline. A little evidence has been provided that choline
is effective in human hepatic cirrhosis.

Evidence has been obtained that the high incidence of human
hepatic cirrhosis in certain districts of South Africa may be due to the
low "protein" content of the diet. This diet produces hepatic cirrhosis
in rats. We can no longer merely speak oi protein as a dietary con-
stituent without remembering that some of the amino acids, the
amounts of which vary in different proteins, have specific physiological
functions. At the moment, we are particularly interested in the
methionine contained in protein which, as stated above, provides a
source of choline.

Various claims have been made for the beneficial effect of
choline in preventing liver injuries due to a variety of infectious
processes. Methionine is now being studied extensively by clinicians
in the hope that its well-established physiological action can be utilized

PHYSIOLOGY AND BIOCHEMISTRY

in the solution of clinical problems. It will require a great deal of
time and effort to determine the relative importance of the various
possible actions of methionine, that is, as a source of organic sulfur,
as a source of a specific amino acid, or as a precursor of choline. It is
not reasonable to expect that the lipotropic factors will exert beneficial
effects when they are added to diets which already contain an abun-
dance of these agents unless clinical conditions exist in which their
absorption or utilization is abnormal.

The metabolism of fat which is, of course, intimately related to
that of protein and carbohydrate, can no longer be thought of as the
sluggish stream it was widely considered to be relatively few years ago.
From some viewpoints it is now easier to study the intermediate me-
tabolism of fats than that of carbohydrate or protein, and we can con-
fidently look forward to great progress in this field during the next
few years,

V/ar Medical Research

With the outbreak of war in September, 1939, many productive
peacetime medical studies were immediately discontinued. Our ene-
mies had apparently made this move many years previously and thus
had, for a time, the great advantage of the results of deliberate and
detailed planning for the medical aspects of warfare. Several busy
years were required to balance the scales anew.

Our experiences in war medical research in Canada are essen-
tially the same as those of our allies with whom and by whom problems
and solutions have been freely shared. The primitive state of our
organization for medical research in the Armed Services at the begin-
ning of the war may have introduced an element of novelty into some
of the arrangements subsequently made in Canada. In the Air Force
and Navy, medical research groups from university departments were
set up, in affiliation with the Medical Committees of the National
Research Council of Canada, as the nuclei of Service Medical Research
Units. The research workers, some placed on active service and some
as civilians supported by the National Research Council, continued to
use the university buildings and facilities for their Service researches.
In the Naval Service, this arrangement has continued with little
change up to the present time. In our particular unit, the first re-

437

C. H. BEST

searches were on shock, blood substitutes, nutrition, lighting, and the
development and application of new tests of vision, hearing, and fre-
quency discrimination. These were followed by attacks on the special
and \'ery numerous physiological problems associated with immersion
foot, cold and hot weather protective clothing, ventilation, diving,
underwater blast, etc.

In spite of the extremely satisfactory progress made under war-
time conditions in many countries, a host of Service medical problems
remain unsolved at the end of the war. No attempt will be made
to list them here. An important matter of concern is that, in our
eagerness to return to peacetime medical research, we shall not forget
the years when we narrowly missed becoming hopelessly enmeshed in
the "coils of slack" that had been allowed to accumulate.

438

28

X-RAY DIFFRACTION AND
THE STUDY OF FIRROUS

PROTEINS

I. FANKUGHEN, associate professor of crystal chemistry,

POLYTECHNIC INSTITUTE OF BROOKLYN
H. MARK, DIRECTOR, POLYMER RESEARCH INSTITUTE, POLYTECHNIC

INSTITUTE OF BROOKLYN

rHlS ESSAY will re\ iew some recent studies with x-rays of
the structure of proteins and will discuss a few questions
which are still unanswered, but which, presumably, may be success-
fully attacked with the aid of x-ray or electron diffraction within the
not too distant future.

The general purpose in applying x-rays for the elucidation of the
structure of matter is to obtain information about the exact position of
atoms or ions inside a crystal or molecule. Such information can be
obtained from diffraction patterns because the diameters of atoms like
carbon, nitrogen, and oxygen are of the same magnitude as the wave
length of x-rays as they are produced in commercial tubes. Both are
between one and two angstrom units, one angstrom unit being 10~^
centimeter. The x-rays diffracted from the three-dimensional atomic
lattice of a crystal or a large molecule show particularly high intensities
in certain preferred directions, from which one can figure out the dis-
tances between the various atoms in the sample.

This line of attack first led to a thorough understanding of the
fine structure of most simple inorganic and organic crystals such as

439

I. FANKUCHEN AND H. MARK

diamond, calcite, quartz, urea, benzene, etc. Later it was possible
also to elucidate the molecular structure of more complicated systems
such as silicates, dyestufFs, and hormones; and recently the x-ray dif-
fraction method has been successfully applied even to such complicated
systems as cellulose, starch, rubber, and protein. Molecular models
have been worked out for these high polymeric substances which reflect
in a fairly satisfactory way the various chemical and physical properties
of the materials.

Proteins of all kinds have for a long time been extensively
studied with x-rays. For recent comprehensive articles see particularly
references 1, 10, 13, 14, 15, 21 and 30. By 1920 diagrams of several
fiber proteins had been obtained by Herzog and his collaborators (8,
19); virus proteins were first successfully studied in 1936 by Bernal and
Fankuchen (7); and single protein crystals have been investigated
systematically by a number of authors since 1939 (11,12,27). Each of
these three types of proteins has challenged the x-ray method in a dif-
ferent way; and the investigations have both originated improvements
of the experimental technique and led to refinements of the theoretical
interpretation of the diagrams. In this essay we shall describe briefly
how fibrous proteins have been studied with x-rays, and will then pro-
ceed from the simple qualitative inspection of the diagrams to their
more quantitative evaluation.

Typical X-Ray Patterns of Fiber Proteins

Figure 1 shows the x-ray diagram of a bundle of native silk
fibers taken with a well-filtered and collimated beam of Cu K radiation.
Patterns of this kind can be obtained in about one hour or less with the
aid of modern commercial x-ray tubes now being produced in this
country by various companies. Since the taking of the diagram itself,
until a few years ago a difficult procedure, is now more or less a matter
of professional routine, all care and time can be spent on the prepara-
tion of the specimens and on the interpretation of the diffraction pat-
tern. It can be seen that the pattern of Figure 1 consists of a number of
diff'raction spots of moderate sharpness and varying intensities which
are arranged symmetrically about the axis of the incident beam.
The primary beam is very intense and would, therefore, produce a
large, black spot in the center of the diagram, which might ovei-shadow

440

X-RAY DIFFRACTION

part of the pattern. It is therefore customary to stop the primary beam
with a small lead plate, the shadow of which can be seen in the center
of Figure 1. In addition to these spots, one observes a fairly distinct
accumulation of scattered intensity in the immediate neighborhood of
the incident beam and a faint, but not negligible, diffuse background
upon which the diffraction spots are superimposed.

Theoretical considerations (see 9,24) and extensive experience
with x-ray diagrams of many inorganic and organic materials indicate

Fig. 1. — X-ray diagram of commercial silk fibroin
(silkworm). Beam normal to fiber axis (24).

that, under favorable conditions, conclusions of the following character
can be drawn from the analysis of such patterns.

Mere qualitative inspection permits the conclusion, on the basis of
the existence of distinct spots, that the sample investigated contains
geometrically organized areas inside of which long-chain molecules of
the silk protein are localized in a strictly periodic manner, like the indi-
vidual bristles in a paint brush or the single pencils in a package of
pencils. The presence of distinct spots instead of rings indicates that
these highly ordered domains all have one of their axes parallel (or
very nearly parallel) to the common axis of the investigated fibers,

441

I. FANKUCHEN AND H. MARK

comparable with many paint brushes the handles of which are all
parallel. Using a somewhat abbreviated and oversimplified expres-
sion, one can also say that the diagram indicates that the sample con-
tains a multitude of oriented crystallites of silk protein. The fact that
each of the spots is larger than the actual geometric projection of the
irradiated part of the fiber bundle and that their boundary is somewhat
diffuse indicates that the highly organized areas consist of only a limited
number of strictly periodically arranged elements, such as atoms,
groups of atoms, molecules, etc. In other words, the oriented "crystals"
of the fiber protein (or, in our analogy, the bundles of pencils) are very
small. The presence of the intensive halo around the incident beam
■ — compare Figure 1 — shows that there exist in this sample struc-
tural heterogeneities the dimensions of which are larger than the wave
length of the irradiated light. Cu Ka radiation has a wave length of
about 1.5 A., and the structural heterogeneities as indicated by the
central blackening have dimensions of about 100 A. Finally, the gen-
eral diffuse background of the pattern is caused by the presence in the
sample of a certain amount of disordered material which may, but
need not necessarily, be chemically different from the more highly
ordered constituents. This would be comparable to a situation in
which the bundles of pencils mentioned above are embedded in a mass
of randomly arranged pencils pointing in all possible directions.

Using the same simplicity and abbreviation as above, we may
summarize by saying that the qualitative inspection of a diagram such
as shown in Figure 1 leads to the conclusion that the investigated ma-
terial consists of very small, highly oriented crystallites, which are em-
bedded in an amorphous matrix in such a way that certain long-range,
quasiperiodic spacings are somewhat accentuated. Each of these
statements corresponds to a special feature of the diagram, as repre-
sented in Table I.

It may, perhaps, be appropriate to add a few brief remarks about
the justification of the use of "crystallites" and "amorphous areas"
in the qualitative interpretation of x-ray diagrams of this kind. With
reference to the crystallites, or micelles, we shall consider these to be
small volumes of somewhat indefinite size and shape, inside of which
the long-chain molecules are arranged according to a fairly regular
three-dimensional periodic pattern. There are also, however, in every
sample of a high polymer, certain portions which are not crystallized.

442

X-RAY DIFFRACTION

These are usually referred to as the amorphous, disordered, intermicellar,
or glassy fractions. Many investigators, particularly Baker and Fuller
(4), Gehman and Field (17), Goldfinger et al. (18), and Purves and his
collaborators (28) have used a variety of methods to obtain information
about these regions in a high-polymer material. It seems that the
regions contain the molecular chains in a less perfect ajrrangement, be-
cause, for instance, some irregularity may have prevented them from
reaching the proper equilibrium positions corresponding to the crystal-

Table I

Qualitative Inspection and Interpretation of a Typical Fiber

X-Ray Diagram

Observed phenomenon in the pattern

Conclusion about the structure

Comparatively distinct, intensive spots

Presence of highly organized areas (so-

of diffracted intensity

called crystallites)

The spots are not drawn out into long

The crystallites in the sample possess

segments or rings

a certain orientation

The spots have a certain radial width

The crystallites are very small (200 A.

and their boundary is somewhat dif-

or less)

fuse

There is a general diffuse background

The sample contains a certain amount

of disordered (amorphous) material

Comparatively intensive halo around

Certain long-range quasiperiodicities

the incident beam

exist in the samples

lized state. It is believed that one and the same chain can pass through
a crystalline area, enter an amorphous portion, go right through it,
and enter another crystalline area. This leads to the conclusion that
there is no sharp boundary betvteen the crystallized and disordered
domains but that the chains of a certain crystalline region somehow
became disordered, degenerate into fringes, and, finally, reach a com-
pletely disordered arrangement. Hence it may be appropriate, not to
make a sharp distinction between chains in a crystallized or amorphous
state, but rather to consider various degrees of disorder, just as one may
observe metallic inixed crystals which have the same composition but a
different arrangement of their constituents. Baker, Fuller, and Pape
(4) have suggested the existence of a me somorphous phase in such materials
as cellulose acetate or nylon; and Taylor (29) has repeatedly empha-

443

I. FANKUGHEN AND H. MARK

sized that chains with regularly distributed centers of attraction arc
unlikely to curl up in a completely random way but will lead to struc-
tures of an intermediate degree of order.

With these limitations and restrictions in mind, we shall now
proceed to a more quantitative discussion and interpretation of fiber
diagrams.

Small- and Large-Angle Diagrams

Inspection of Figure 1 has already shown that there are two dis-
tinct areas in which scattered intensity can be observed:

(a) In the immediate neighborhood of the incident beam, i. e.,
within very small angles of scattering.

(b) All over the photographic film, i. e., at comparatively
large angles of scattering.

Because of the reciprocal nature of all diffraction phenomena, *
small-angle scattering permits the study of large distances inside the sample,
while large-angle scattering permits the investigation of short-range order.
It has, therefore, become conventional to divide the quantitative study
of a fiber diagram into the evaluation of the "large-angle" and of the
"small-angle" pattern.

If one measures the position and intensity of all observable
points in the large-angle region, one can, first of all, compute from the
position of each spot the diffraction angle and, with the aid of Bragg's
law, the characteristic spacing which is responsible for each individual
spot. In this way one ends up with a table of spacings and relative
intensities of all observed diffraction points. In Table II, such a list is
given for the diagram of silk protein (Fig. 1, page 441). The "evalua-
tion" of the diagram is now based on the assumption that the spacings
of the lattice planes as they appear in Table II are the consequence of
the arrangements of all individual atoms in a three-dimensional lattice;
and the first goal is to find the fundamental distances and angles of this
lattice. Generally, this can be accomplished from x-ray data alone

* The famous Bragg diffraction law demands that the product of the scatter-
ing spacing, d, and the sine of the glancing angle, 6, must equal half the wave length
of the scattered radiation in order that an intensive diffracted beam be produced.
Small spacings produce, therefore, large diffraction angles, and large spacings,
small angles.

444

X-RAY DIFFRACTION

if one has a number of single crystal diagrams, taken nndcr very well-
defined geometric conditions, each diagram containing a large number
of individual spots. How^ever, if one is v^illing to combine x-ray data
with other evidence, coming from chemical analysis or from physico-
chemical investigations, it is possible to advance well-founded structural
proposals, provided that the investigated samples exhibit a so-called
twofold orientation, such as stretched or rolled thin films of cellulose,
rubber, or collagen (5,24). However, if one has only diagrams with

Table II
A Few Spacings of Lattice Planes in the Lattice of Natural Silk

Number of reflection in
Figure 1

Sine of Bragg's angle, 6 Lattice plane spacing, d, A.

Ai
A,
A4
A5

0.084
0.178
0.251
0.324

9.2
4.3
3.1
2.4

axial symmetry (so-called fiber diagrams), which contain spots of the
number and character as shown in Figure 1, it is, in general, not possible
to arrive at reliable quantitative conclusions regarding the exact size
and shape of the elementary cell of the lattice of the crystalline con-
stituents.

It seems that in some special cases, such as the protein in native
silk (8), the protein in native wool and hair called /3-keratin (3,26) and
a-keratin (2,3), and the protein in muscle called collagen (5,6) rather
reasonable proposals for the lattice parameters have been suggested by
supplementing the x-ray data with chemical and morphological data,
and with the aid of a certain amount of intuitive guessing. However,
even in these cases one cannot, at present, advance well-founded data
as to the exact location of the various substituents along the main poly-
peptide chains in the lattice, although one knows their approximate
arrangement. Summarizing, it may be said that the large-angle x-ray
study of fiber proteins to date has furnished enough data to arrive at
rather probable structures for the most important protein fibers if one
combines the x-ray data with all the other physical and chemical knowl-
edge of these materials.

445

I. FANKUCHEN AND H. MARK

If we now concentrate our attention on the immediate neiglibor-
hood of the incident beam in Figure 1, we find there is a distinct blacken-
ing of the film in this region. The intensity scattered at small angles
corresponds to large distances inside the investigated sample. If these
distances are arranged in some kind of periodic sequence, they produce,
according to Bragg's law, diffraction maxima the intensity and sharp-
ness of which depend upon the perfection of the periodic long-range
structure. If, therefore, the small-angle scattering consists of a number
of (more or less) well-defined lines, one concludes that something like a
"superlattice" exists in the crystalline domains of the sample. Figure 2

shows a typical case of such a long-
range order in a suspension of
I tobacco mosaic virus (7). The

very elongated rodlike particles
of this virus can be perfectly
oriented in their aqueous suspen-
sion even at such low concentra-
Fig. 2. — X-ray diaffram of tobacco ■ ^ ^rr-f ,

° . T, , , tion as 2 or 3%, and assume an

mosaic virus. Beam normal to molec-
ular direction. order similar to that in a bundle

of pencils: their axes are all
parallel and they maintain fixed average distances from each other.
If such a structure is irradiated with x-rays perpendicular to the
axis of orientation, patterns such as those in Figure 2 are obtained.
The sharp lines found on the equator of the diagram at small angles
indicate that there is a certain periodicity perpendicular to the axis of
orientation, which roughly corresponds, in the case of the dry virus,
to the average thickness of the rodlike parallelized virus particles. In
the case of tobacco mosaic virus, this thickness was found to be about
150 A.

Comparatively sharp reflections at small angles on the meridian of
protein fiber x-ray diagrams have also been observed, indicating the
existence of a certain long-range periodicity along the axis of the paral-
lelized and aligned molecules. Astbury (3) has observed that |S- keratin
exhibits meridional small-angle reflections, which correspond to a
spacing of about 66 A.; he suggests they are either the third order of a
period of 198 A. or the tenth order of 660 A. Porcupine quill and
feather rachis indicate long periods of 198 and 95 A., respectively
(23). Beef tendon (5) exhibits about thirty orders of a fundamental

446

X-RAY DIFFRACTION

period of 640 A. along the fiber axis; clam muscle fibrils (22) show
indications of a long hber period of 700 A. In the latter two cases,
independent observations with the electron microscope led to similar
periodicities in the investigated samples. Bear (6) reports a long fiber
period of about 600 A. for a large number of collagen specimens; and

-I -

o>-

^r-vMAj

i

'*^V>\v

^VV.^'^^

Q

m

Fig. 3. — Small and large angle scattering
of elongated, parallelized crystallites (16).

it is interesting to note that synthetic polyamides also show distinct
meridional small-angle reflections, which lead to long fiber periods of
about 90 A. (16,20).

Figure 3 shows diagrammatically the small- and large-angle
scattering of a highly stretched and subsequently relaxed nylon filament
and indicates the relations between the various reflections of the diagram
with the corresponding distances in the sample.

447

I. FANKUCHEN AND H. MARK

Sometimes the intensity scattered at small angles does not exhibit
distinct lines or spots but is of a more continuous character, decreasing
gradually with increasing diffraction angle. This points to the fact
that there are no repeating, well-defined long dimensions within the
sample but that these distances either scatter rather widely about a
certain average value or are randomly distributed in the sample. In
such cases, one must investigate quantitatively the whole intensity
distribution around the primary beam before one can draw conclusions
about the length of this pseudoperiodicity (16) or more generally say
something about the distribution of these long dimensioris within the
specimen.

Size of Crystallites, Degree of Orientation, and Crystallinity

It has already been mentioned that the sharpness and intensity
of the lines or spots of the small-angle diagram depend upon the num-
ber of repeating periodicities and upon the accuracy with which
they repeat. This is also true for the lines or spots of the large-angle
diagram. The more frequently a given lattice plane is repeated in one
of the crystalline regions, the more numerous will be the secondary
waves which cooperate with a favorable phase difference and, hence,
the sharper and more intense is the diffraction spot which they produce.
Sharp lines or spots indicate large crystals or crystalline regions; dif-
fuse lines indicate small crystals. In most fiber protein diagrams, the
large-angle diffraction spots are fairly broad and indicate that the aver-
age width of the bundle-shaped crystallites in a highly oriented protein
fiber, such as silk, muscle, or feather, is about 100 A. or less. Along
the axis of orientation the diffraction lines are sharper and, hence, the
domains of crystalline order are longer. Stretching and alignment
usually increase the size of the crystalline areas, whereas swelling and
relaxing decrease it. If the ordered domains become smaller than
20 or 30 A., the x-ray diagram of the system becomes more and more
diffuse and finally approaches that of a liquid.

Next to the approximate average size of the crystalline areas, one
is interested in the way in which they are oriented in a given sample.
Completely random arrangement of the scattering elements in a sample
produces an x-ray diagram which consists exclusively of rings. When
some or all of these rings degenerate into segments or spots, the presence

448

X-RAY DIFFRACTION

of a certain degree of orientation is indicated. Quantitatively one can
measure the length of each individual segment, parallel to the circum-
ference of the ring to which it belongs, and convert this length into the
angular opening of the segment. This angular opening describes the
degree of orientation in the investigated material: if it is large, the
crystallites scatter widely about certain directions and the orientation
is poor; if, however, the angles corresponding to all segments in the
diagram are small, i. e., the diagram consists of sharp spots rather than
smeared out segments, then the orientation of the crystalline domains
in the sample is good. In general, orientation of the crystallites is
closely connected with important mechanical properties such as
strength, rigidity, or toughness.

The last feature of a protein fiber x-ray pattern, which deserves
quantitative consideration is the general background of diffusely scat-
tered intensity. This indicates that, in addition to the well-ordered
fraction producing the comparatively sharp diffraction lines and spots,
a certain amount of material exists which is in a disordered state and which
cannot produce any characteristic diffraction phenomenon. This
amorphous or disordered matrix in which the crystalline areas are em-
bedded is mainly responsible for the resilience, flexibility, and swelling
properties of a protein fiber and hence may represent a very valuable
constituent of it. In principle, it would be possible to carry out a
measurement of the integrated intensity of this diffuse scattering and by
comparing it with the total intensity of the crystalline pattern to arrive
at a quantitative figure of the ratio of crystalline to amorphous material.
In practice, however, it is possible to obtain this ratio only as an esti-
mate rather than as a precise measurement. One can, for example,
find out whether a certain treatment of a sample has increased or de-
creased its degree of crystallinity. In general, procedures such as
stretching, rolling, drawing, deswelling, etc., increase the degree of
crystallinity, while all steps of relaxation and swelling tend to increase
the disorder in the material. Since the ratio between ordered and dis-
ordered constituents is of great importance for the mechanical and col-
loidal properties, even an approximate estimate of it can contribute to
an understanding of the behavior of certain systems.

Until recently, comparatively primitive x-ray diffraction tech-
niques were used to study proteins — primitive in the sense that they
were not the most advanced available at the time of the experiments in

449

I. FANKUCHEN AND H. MARK

question. More recent papers exhibit an awareness of this fact; and
it may be expected that much of the forthcoming work in this field will
utilize not only the modern developments of technique but also those
in interpretation.

In technique, the new approach starts, properly, with the speci-
men. While some protein fibers (wool, silk) can be handled with
ease, the study of others (like muscles) must be accompanied by extreme
precautions to prevent any appreciable changes from their natural
state. Because the use of normal specimens usually introduces many
fibers into the x-ray beam, the tendency is to reduce the size both of
the specimens and of the x-ray beam. This trend can be carried to
quite an extreme degree so that single fibers and beams of 0.05 mm.
diameter have been used (16); the results indicate that even much
finer beams could be economically employed. When the specimen is
a single intact fiber, the x-ray diagram is free from the possibility of
disorientation due to lack of parallelism between the individual fibers
making up a bundle.

More and more intense sources of x-rays are being made avail-
able to the investigator and these are shortening the time required to
obtain an x-ray diagram. A more exciting possibility arises from the
introduction of Geiger-Muller counters to record the scattered radiation.
This innovation permits the almost instantaneous determination of the
intensity of scattered radiation and will allow the planning of many
experiments which are impossible by photographic methods. Studies
of structure as a function of time (a contracting muscle, perhaps) could
be made now that a continuous recorder of scattered radiation is avail-
able. It will also be possible to study specimens which deteriorate
rapidly.

Efficient monochromatization of the x-ray beams will also be
useful in many cases. Biological specimens often give poor x-ray dia-
grams regardless of technique; and the additional background due to
incomplete monochromation may be enough to obscure important
details in the x-ray diagram.

The interpretation of the data is also being steadily improved.
Increased use of reciprocal-lattice concepts and the more widespread
realization of the shortcomings of ordinary fiber diagrams {i. e.,
diagrams from specimens which possess orientation with reference
to one direction only) lead the experimenter to attempt to prepare

X-RAY DIFFRACTION

specimens of higher orientation. As we have seen, much more certain
conclusions can be drawn from such studies.

Fourier methods have been widely used in the study of single
crystals (both protein and nonprotein). There seems to be no good
reason why these methods should not yield useful results when applied
to fibrous protein. Certainly, many recent experiments have yielded
data suitable for such studies.

When one considers in reti'ospect the considerable effort that
has been expended in the study of protein structures with x-rays, per-
haps the conclusion is inevitable that the return has been somewhat
meager; but at the same time there seems to be considerable justifica-
tion for an optimistic view of what is soon to come. The methods and
techniques of x-ray diff'raction which are now available should tell us
much that we want to know about the structure of proteins.

References

(1

(2
(3

(
(4

(
(5
(6
(7
(8
(9
(10

(II
(12
(13
(14
(15

(16
(17
(18

Astbury, W. T., Cold Spring Harbor Symposia Quant. Biol., 2, 15 (1934.)

Astbury, W. T., Ann. Rev. Biochem., 8, 113 (1939).

Astbury, W. T., and Sisson, W. A., Proc. Roy. Soc. London, A150, 533
935).

Baker, W. O., Fuller, G. S., and Pape, N. R., J. Am. Chem. Soc, 64, 776
942).

Bear, R. S., J. Am. Chem. Soc, 64, 727 (1942).

Bear, R. S., J. Am. Chem. Soc, 66, 1297 (1944).

Bernal, J. D., and Fankuchen, I., J. Gen. Physiol., 25, 111 (1941).

Brill, R., Ann., 434, 204 (1923).

Buerger, M. J., X-Ray Crystallography. Wiley, New York, 1942.

Bull, H. B., in Advances in Colloid Science, Vol. I. Interscience, New
York, 1941, p. 1.

Crowfoot, D., Chem. Revs., 28, 215 (1941).

Crowfoot, D., and Riley, D., J^ature, 144, 1011 (1939).

Fankuchen, I., Cold Spring Harbor Symposia Quant. Biol., 9, 198 (1941).

Fankuchen, I., Ann. j\. Y. Acad. Sci., 41, 157 (1941).

Fankuchen, I., in Advances in Protein Chemistry. Vol. II, Academic Press,
New York, 1945.

Fankuchen, I., and Mark, H., J. Applied Phys., 15, 364 (1944).

Gehman, S. D., and Field, J. E., J. Applied Phys., 10, 564 (1939).

Goldfinger, G., Siggia, S., and Mark, H., Ind. Eng. Chem., 35, 1083
943).

(

451

I. FANKUCHEN AND H. MARK

(19) Herzog, R. O., and Jancke, W., Ber., 53, 2162 (1920).

(20) Hess, K., and Kiessig, H., Naturwissenschqften, 31, 171 (1943).

(21) Huggins, M. L., Chem. Revs., 32, 195 (1943).

(22) Jakus, M. A., Hall, C. E., and Schmitt, F. O., J. Am. Chem. Soc, 66,
313 (1944).

(23) McArthur, I., Nature, 152, 38 (1943).

(24) Meyer, K. H., Natural and Synthetic High Polymers. Interscience, New
York, 1942.

(25) Nickerson, R. F., Ind. Eng. Chem., 32, 1454 (1940); 33, 85, 1022 (1941).

(26) Pauling, L., and Niemann, C, J. Am. Chem. Soc, 61, 1860 (1939).

(27) Perutz, M. F., Nature, 149, 491 (1942); 150, 324 (1942).

(28) Purves, C. B., Asaaf, A. G., and Haas, R. H., J. Am. Chem. Soc, 66, 59,
66 (1944).

(29) Taylor, H. S., at several lectures, particularly during the discussion at
the meeting of the American Chemical Society in Memphis, Tennessee,
April, 1942.

(30) Wrinch, D., Trans. Faraday Soc, 43, 1368 (1937).

452

29

IMMUNOGHEMISTRY

MICHAEL HEIDELBERGER, professor of biochemistry, college

OF PHYSICIANS AND SURGEONS, COLUMBIA UNIVERSITY; CHEMIST TO THE
PRESBYTERIAN HOSPITAL, NEW YORK

rHIS WAS the title, perhaps whimsically given, of a series
of lectures delivered in 1904 by Svante Arrhenius, a great
Swedish physical chemist, at the University of California, and pub-
lished in book form (la) three years later — so hectic was the pace in
immunology. True, Ehrlich (5) had insisted on the chemical nature
of immune processes and immune reactions, but his views were over-
shadowed by the widely accepted and facile concepts of Bordet (2a),
whose colossal contributions to immunology gave him vast influence.
He first classified immune reactions as essentially physical, then, when
this position became untenable, maintained that immune reactions
were "colloidal." The persuasive, amorphous terminology of the
early colloid chemistry, descriptive of everything but accounting for
little, found its way into books and papers on bacteriology and im-
munology, where, be it sadly whispered, it may often still be seen.
Doubtless, many of those who once found it useful and comfortable
still think of it with nostalgia.

But Arrhenius, with his Danish pupil, Madsen, proceeded to
show how, with certain assumptions, the laws of classical chemistry
could be applied to typical immune reactions, and so the term immiino-
chemistry came into serious use and was seriously taken, even though
the proposed analogy of antigen-antibody reactions to the union of
weak acids and weak bases was quickly shown to be im-alid.

453

MICHAEL HEIDELBERGER

Looking back from the present admittedly none too elevated
observation post, in which one must still duck occasional snipers' shots,
one cannot fail to be astonished that so many sound observations and
so many keen deductions were made in those early days. Next to
nothing was known of the chemical nature of antigens and antibodies,
the protagonists in the drama of immunity, while complement, the
powerful mediator of cell lysis, flitted in and out like the ghost in
Hamlet, between material existence and a mere "colloidal state" of
something else. Worse still, measurement of quantities of any of these
elusive unknowns could only be carried out in relative terms, such as
by the volume which just would kill or not kill an animal, or by the
dilution at which an inflamed skin area or a haze in a test tube just
faded out. Immunology, and the early immunochemistry as well,
staggered under the dictatorship of dilution and the tyranny of titer.

Relief, happily, was just around the corner. Discovery of the
specific polysaccharides (2,8) in Avery's laboratory gave concrete
direction to the relation between chemical constitution and bacterial
specificity. Extension of chemical concepts to protein antigens was
facilitated by the imaginative and painstaking studies under Land-
steiner's (29) direction, in which aromatic radicals, optically active
acids, and amino acids were coupled to proteins through the diazo
reaction and their eff"ects on specificity studied. As for antibodies,
prolongation of the discussion as to whether these were serum globulins
or unknown substances adsorbed on globulins seemed futile after
Felton's simple method for the concentration of pneumococcus anti-
bodies led him to the demonstration that zinc and aluminum salts of
these globulins were completely precipitable by the homologous type-
specific polysaccharide of pneumococcus (5a).

Though the chemistry of antigens and antibodies began to be
better understood, the fetters of the old, relative analytical methods re-
mained unbroken until Kendall and the writer (11), putting their faith
in the rigorous criteria of analytical chemistry, devised quantitative
micro methods for the accurate estimation of many antigens and anti-
bodies in absolute terms, that is, units of weight rather than of titer.
With Sia (19) they demonstrated the parallel between mouse protec-
tion and the amount of antibody pitrogen precipitated from anti-
pneumococcus type I serum by the specific polysaccharide of type I
pneumococcus. This was in 1930, but only in 1938 did the "brass

454

IMMUNOCHEMISTRY

hats" permit a substitution of antibody nitrogen values for mouse
protection, and then only in rabbit antisera to certain of the other
pneumococcus types. Millions of mice were finally saved, however,
from succumbing to protection tests by the advent of sulfa drugs and
penicillin, which made antipneumococcus sera practically obsolete.
Although large volumes of antipneumococcus horse sera, at least, have
enriched the efHuent from certain production centers of biologicals in
order to release storage space, infants and children are still being cured
ot influenzal meningitis by known amounts of antibody from rabbit
antisera to H. influenzae type b (1), for a drug to cure this dread disease
has not yet been found,* nor is the mouse protection test free from diffi-
culties in both its application and interpretation.

Combining proportions by weight in the precipitin reaction
were studied, and with their aid a quantitative theory of this reaction
was derived from the law of mass action, leading to a simple linear
relation applicable to many immune systems (12). The data obtained
were most easily explicable on the basis of the union oi' multivalent
antigen and multivalent antibody, giving concrete and quantitative
mathematical expression to the less definite "lattice" or "framework"
theory of Marrack (30).

The assumptions which were made admittedly involved over-
simplification, but this seemed pardonable in a pioneer theory which
explained much that had appeared mysterious and provided simple,
usable equations that could be readily understood. There was time
enough for taking care of the complications later, as several friends and
colleagues have done. Pauling and his group (33), brushing aside
the original assumptions as "arbitrary and unlikely," have made a
different application of the laws of chemical equilibrium but arrived at
the same approximate equations (were their faces red!). By a sta-
tistical approach Kendall (28) has paralleled and outstripped the
original theory, once more obtaining equations of the same form for
the precipitin reaction and succeeding in explaining quantitatively
even the limited range of toxin-antitoxin flocculation. One would
like to say more and know more about Hershey's (23) painstaking and
rigorous, but painfully complicated, analysis. Boyd (3) has been a
consistent Rightist in his attitude toward the theory of mutual multi-
valence, holding to the old notion that antigen-antibody combination
* Streptomycin shows promise (H. E. Alexander, Science, in press).

455

MICHAEL HEIDELBERGER

is a static affair and aggregation a nonspecific aftereffect. Insisting
that the evidence for the multivalence of antibody, at least, is inade-
quate, he has produced no satisfying alternatives and his publications
show evidence that underground movements are weakening his hold.

Study with Kabat (9) of a system of bacterial agglutination
showed this reaction to be subject to laws similar to those underlying
the precipitin reaction, and for the first time placed the important
function of salts in a perspective consistent with modern protein
chemistry. But a theory that does not lead to useful and exciting
predictions is like a sterile "Man O' War." In its explanation of salt
effects observed in a study with Teorell (14), the quantitative theory
developed with the new analytical methods led to a prediction as to
how analytically pure antibody globulin might be obtained (13,14);
and this was realized in two laboratories, (5b, 10). Since, with Pedersen
and Kabat (17,24) it was shown, with the help of Svedberg and his
ultracentrifuge and Tiselius (35) and his electrophoresis apparatus, that
purified antibodies have the properties of serum globulins, the protein
nature of antibodies can no longer be doubted.

The technically important toxin-antitoxin reaction has been
worked out analytically by Pappenheimer and Robinson (32), so
that it is now possible, aided by a single pair of nitrogen estimations,
to measure the unitage both of an unknown diphtheria toxin and an
unknown antitoxin. To Ehrlich this would have seemed a pro-
digious feat, but the "brass" of immunology has failed to legitimize it.

Quantitative immunochemical methods are, however, being
applied in many other directions. They were used, with Mayer, for
a study of reversibility and velocity in the precipitin reaction (20,31).
Through further modifications, with MacPherson, they have been
extended to the measurement of micrograms of antibody in human
sera (7,16). The new methods have also supplied an absolute measure
of complement, furnishing a weight unit independent of the lytic
function of this strange and little understood complex (6,15,18,22).
With this measure, the reacting proportions of complement, hemolysin,
and red cells could be calculated, as also those of antigen, antibody,
and complement involved in the process of complement fixation.
Oddly enough, complement could be fitted handily into the quanti-
tative theory of the precipitin reaction, which had been elaborated in
blissful neglect of what might have been an embarrassing apparition.

IMMUNOCHEMISTRY

For the first time, complement "belonged" and this Flying Dutchman
of immunology was at rest.

Stepping out with these quantitative methods, immunochemistry
has also contributed to the interpretation of isotope studies in protein
metabolism. Compounds of heavy nitrogen were found by Schoen-
heimer and his group to slip in and out of body proteins and their
constituent amino acids with unexpected speed and thoroughness.
In collaboration with these biochemists it was found possible to work
not only with labeled nitrogen but with labeled protein — antibody —
as well. In a rabbit immunized with type III pneumococci, anti-
body, quantitatively separable from the other serum globulins, was
found to take up dietary nitrogen at about the same rate as these
other proteins (21,34). All of this seemed eminently proper until
another test was made adopting TrefTers' idea of injecting a similarly
immunized rabbit with a difTerent antibody, type I pneumococcus
anticarbohydrate, from another rabbit. The type Ill-immunized
rabbit was fed heavy nitrogen. Here, then, were labeled dietary
nitrogen and two labeled indicator proteins, one being produced
(and destroyed) by the animal, the other merely being metabolized.
Each indicator protein in the rabbit's serum, drawn at intervals, was
quantitatively and independently precipitable by the corresponding
specific polysaccharide and so separable from the other serum proteins.
What happened was that, while the type III antibody quickly took
up heavy nitrogen as before, the passively injected type I antibody,
which was merely circulating in, but was not being elaborated by,
the test rabbit, remained free from heavy nitrogen. This was taken
to indicate that heavy nitrogen fails to enter serum protein molecules
which were complete at the start of heavy nitrogen feeding, but be-
comes part of the structure only of those proteins which are being
constantly synthesized. As a by-product, quantitative studies of the
rate of disappearance of heavy nitrogen from antibody indicated a
half-life (with respect to synthesis vs. destruction) of about two weeks,
much the same as that of other serum proteins in this animal.

Incidentally — and this was overlooked at the time — the absence
of N^^ in the passively injected type I antibody, after precipitation with
its homologous type-specific polysaccharide and washing under standard
conditions, affords as rigorous a test as has ever been made of the speci-
ficity of immune precipitation and the validity of the assumption under-

457

MICHAEL HEIDELBERGER

lying the estimation of antibody in absolute terms ; namely, that anti-
body, and antibody only, is quantitatively measured. Although the
serum of the type Ill-immunized rabbit contained up to 1 atom per
cent N^^ excess, the type I antibody isolated from this medium contain-
ing other antibodies, globulins, and albumins rich in heavy nitrogen re-
tained none of the N^^ within the error of the mass-spectrographic
method, about t. 0.03 atom per cent N^^ excess.

The immunochemist has learned quantitative relationships
with even so complicated a test tube as the guinea pig, for Kabat and
Landow (26,27) have shown that 0.2 mg. of antibody suffices to sensi-
tize the animal so that fatal anaphylactic shock ensues with 1 mg. of
antigen such as egg albumin or 0.1 mg. of a type-specific polysaccharide
of pneumococcus. Either of these shock doses is far in the region of
antigen excess in the in vitro precipitin reaction, suggesting at once the
first simple explanation of how much smaller quantities of antigen
may desensitize the animal. Because small quantities ot antigen com-
bine with relatively much larger amounts of antibody, after the in-
jection of "desensitizing" small doses of antigen relatively little anti-
body remains for the final destructive effect of an injection of excess
antigen.

All of which must not be taken to indicate that quantitative
immunochemistry, like a parachute flare, has had its episodic illu-
minating burst and is fading out. It has shown the way to objectives
of theoretical and practical interest, and many more applications are
apparent. One quantitative immunochemist has at last irritated our
native virologists into an appreciation of its potentialities in their vast
field of endeavor (25). Quantitative immunochemical methods are
also indispensable guides in the study of bacterial antigens and the
fractionation of many other natural carbohydrate and protein mixtures.
A recent application in the little studied field of lipoproteins was the
report of Cohen and Chargaff (4) on thromboplastin. Suspensions
of bovine thromboplastin removed measurable amounts of antibody
from specific antisera in rabbits, and, remarkably enough, the specific
precipitates showed greater thromboplastic activity than the antigen
contained in them. Displacement of the phosphatide in the antigen
by means of heparin did not destroy the power to precipitate with
antiserum.

Moreover, if proteins or proteinlike structures are ever synthe-

IMMUNOCHEMISTRY

sized, with or without the mediation of enzymes, immunochemistry
will afford incisive aids to their study. Protozoan and other parasitic
diseases are intruding upon us with a new insistence born of the war
and increased speed of travel, and it is confidently to be expected that
immunochemistry will become an essential factor in the understanding
and conquest of these scourges. The broadening horizons of immuno-
chemistry will steadily augment its opportunities for fundamental
contributions in the struggle against infectious disease.

Rejerences

(1) Alexander, H. E., Heidelberger, M., and Leidy, G., Tale J. Biol Med.,
16, 425 (1944), and earlier papers.

(la) Arrhenius, S., Immunochemistry. Macmillan, New York, 1907.

(2) Avery, O. T., and Heidelberger, M., J. Exptl. Med., 38, 81 (1923);
42, 367 (1925).

(2a) Bordet, J., Traite de I'lmmunite. Masson, Paris, 1920, 1939.

(3) Summarized in Boyd, W. C, Fundamentals of Immunology. Interscience,
New York, 1943.

(4) Cohen, S. S., and Chargaff, E., J. Biol. Chem., 136, 243 (1940).

(5) Summarized in Ehrlich, P., Studies on Immunity. Wiley, New York, 1906.
(5a) Felton, L. D., J. Immunol., 22, 453 (1932).

(5b) Goodner, K., and Horsfall, F. L., Jr., J. Exptl. Med., 66, 437 (1937).

(6) Heidelberger, M., Science, 92, 534 (1940); J. Exptl. Med., 73, 681 (1941).

(7) Heidelberger, M., and Anderson, D. G., J. Clin. Investigation, 23, 607
(1944).

(8) Heidelberger, M., and Avery, O. T., J. Exptl. Med., 38, 73 (1923);
40, 301 (1924).

(9) Heidelberger, M., and Kabat, E. A., J. Exptl. Med., 60, 643 (1934);
63, 737 (1936); 65, 885 (1937).

(10) Heidelberger, M., and Kabat, E. A., J. Exptl. Med., 67, 181 (1938).

(11) Heidelberger, M., and Kendall, F. E., J. Exptl. Med., 50, 809 (1929);

61, 559 (1935).

(12) Heidelberger, M., and Kendall, F. E., J. Exptl. Med., 61, 563 (1935);

62, 467 (1935); 65, 647 (1937).

(13) Heidelberger, M., and Kendall, F. E., J. Exptl. Med., 64, 161 (1936).

(14) Heidelberger, M., Kendall, F. E., and Teorell, T., J. Exptl. Med., 63,
819 (1936).

(15) Heidelberger, M., and Mayer, M., J. Exptl. Med., 75, 285 (1942).

(16) Heidelberger, M., and MacPherson, C. F. C., Science, 97, 405 (1943).

459

MICHAEL HEIDELBERGER

(17) Hcidelbergcr, M., and Pedersen, K. O., J. Exptl. Med., 65, 393 (1937).

(18) Hcidelbergcr, M., Rocha e Silva, M., and Mayer, M., J. Exptl. Med.
74, 359 (1941).

(19) Heidelberger, M., Sia, R. H. P., and Kendall, F. E., J. Exptl. Med., 52,
477 (1930).

(20) Heidelberger, M., Treffers, H. P., and Mayer, M., J. Exptl. Med., 71,
271 (1940).

(21) Heidelberger, M., Treffers, H. P., Schoenheimer, R., Ratner, S., and
Rittenberg, D., J. Biol. Chem., 144, 555 (1942).

(22) Heidelberger, M., Weil, A. J., and Treffers, H. P., J. Exptl. Med., 73,
695 (1941).

(23) Hershey, A. D., J. Immunol., 48, 381 (1944), and earlier publications.

(24) Kabat, E. A., J. Exptl. Med., 69, 103 (1939).

(25) Kabat, E. A., J. Immunol., 47, 513 (1943).

(26) Kabat, E. A., and Boldt, M. H., J. Immunol., 48, 181 (1944).

(27) Kabat, E. A., and Landow, H., J. Immunol., 44, 69 (1942).

(28) Kendall, F. E., Ann. N. T. Acad. Sci., 43, 85 (1942).

(29) Summarized in Landsteiner, K., The Specificity oj Serological Reactions.
Harvard Univ. Press, Cambridge, 1945.

(30) Summarized in Marrack, J. R., Chemistry oj Antigens and Antibodies.
H. M. Stationery Office, London, 1934; 2nd ed., 1938.

(31) Mayer, M., and Heidelberger, M., J. Biol. Chem., 143, 567 (1942).

(32) Pappenheimer, A. M., Jr., and Robinson, E. S., J. Immunol., 32, 291
(1937).

(33) Pauling, L., Campbell, D. H., and Pressman, D., Physiol. Revs., 23, 203
(1943).

(34) Schoenheimer, R., Ratner, S., Rittenberg, D., and Heidelberger, M.,
J. Biol. Chem., 144, 545 (1942).

(35) Tiselius, A., and Kabat, E. A., J. Exptl. Med., 69, 119 (1939).

460

■\

30

SOCIAL ASPECTS
OF NUTRITION

W. H. SEBRELL, medical director, u. s. public health service;

CHIEF, DIVISION OF PHYSIOLOGY, NATIONAL INSTITUTE OF HEALTH

'XTHE PHENOMENAL scientific progress in biochemistry
-^ in the past few decades has opened up new vistas in the
field of nutrition. The most important aspect of this progress has
been our greatly increased knowledge of the chemistry of the human
organism. Although we still have much to learn about the chemistry
of our vital processes, enough is now known to show that, by con-
trolling his nutritional environment, man can make far-reaching con-
tributions to his health and welfare.

Our enemies in this war have demonstrated how manipulation
of the food supply can be used to weaken nations by creating condi-
tions which cause malnutrition, ill health, and increasing mortality.
But our knowledge of nutrition also can be used to strengthen the
people of the world. If this knowledge is intelligently applied, progress
toward the ideal of adequate nutrition for everyone can become one
of the main roads leading to a greater degree of health and prosperity
than the world has ever known. This, if attained, can be a major
contribution to a permanent international peace. Since scientists
possess this vital knowledge, the question is: how can scientists best
play their parts in transforming the ideal into reality?

Most scientists feel they have completed their work when the
results of their experiments have been published in a technical journal

461

W. H. SEBRELL

and presumably made available to all. In fact, of course, this is not
true, for their results are actually made available only to a select few
who are capable of interpreting them correctly. The job of making
the results available to everyone in terms of their practical application
still remains to be done.

In our world of today, the biochemist has become so much a
part of medicine, and his findings are so important in the diagnosis
and treatment of disease, that his work can no longer be regarded as
finished when his new discoveries are published. It is a part of his
job and responsibility to assist in making his results of real benefit to
people. This does not mean that every biochemist should become a
crusader and public lecturer, but rather that some qualified investi-
gators should take part in cooperative local, State, and national eff'orts
to improve nutrition, adding their knowledge to that of the physician,
agriculturist, food distributor, social worker, and representatives of
government in the attempt to produce, distribute, and utilize our food
supply in the best possible manner.

Thus pellagra, a deficiency disease, has been known for about
twenty-five years to be due to a deficient diet. The foods which are
high in pellagra-preventive value are known and widely available in
the United States. Niacin was found to be the specific essential
vitamin, and has been available for seven years. Yet in the United
States in 1943 at least 1303 people died of pellagra, and a much larger
number suffered from its eflfects. Why did all this suffering and death
occur? The reasons were either that the necessary foods were not
available to the victims or that they were ignorant of the information
existing about this disease. Poverty and ignorance are the two great
factors involved in malnutrition of all kinds.

Would we not have been more successful in eradicating this
disease had all the scientists who knew the facts taken a more active
part in correcting the underlying causes? Although willing to argue
over technicalities, and the validity of observations, research workers
seem to be slow to see that it also should be one of their obligations to
devote at least some of their energy to devising practical measures for
the prevention of disease, and in imparting their knowledge to those
who can use it in the field. If we are determined to make democracy
work, scientists must make their special knowledge widely available.

In the less highly organized society which existed during the

462

SOCIAL ASPECTS OF NUTRITION

early development of this country, there was an opportunity to obtain
food by hunting, fishing, and farming. Today, it is worse
than useless; it is cruel to tell a malnourished man that he
needs better food when his lack of food is due to his inability to buy it,
or when the food is not available in his locality. A few years ago we
were destroying so-called food surpluses although many thousands of
people in this country had deficiency diseases. The trouble was not
overproduction in terms of adequate food for everyone but a failure
in distribution. Much food had to be destroyed because there was
no market for it, the producer was threatened with economic ruin,
and there was no mechanism by which the food could be made avail-
able to those who needed it for better nutrition.

This anomaly of want in the midst of plenty was partly due to
the fact that scientists were working independently on narrow aspects
of the problem and there was no organization for putting all the
knowledge together into a unified plan. The existing knowledge had
not been made available either to the public or to those responsible
for handling the food supply. This defect has been remedied to some
extent by action taken as a result of the National Nutrition Conference
in 1941, and by the establishment of the Food and Nutrition Board of
the National Research Council, which has played a useful part during
the war and which should continue to render valuable service to the
nation, by making scientific knowledge more readily available, by
enabling scientists to see nutritional problems as a whole, and by
giving them an opportunity to participate in the application of their
knowledge to human welfare. Had we, as a nation, recognized and
met our obligation to make adequate nutrition available to everyone
who needed it, the so-called surpluses never would have existed. No
nation has ever produced enough of the right kinds of food to meet
the needs of adequate nutrition for its entire population.

It is only within the past few years that research in nutrition and
biochemistry has brought our knowledge to the point at which we
could with some assurance formulate a dietary we know is adequate
for health. Such a dietary has been set forth in the recommended
dietary allowances of the National Research Council (4). With this
basic and essential information, the scientist is now ready to ask society
if it is willing to accept as a principle of democracy that everyone is
entitled to an opportunitv to secure a diet adequate for health. This

W. H. SEBRELL

responsibility was accepted, not only by the representatives of the
United States, but also by those of forty-three other nations at the
United Nations Conference on Food and Agriculture. This is a
great step forward, but it is still a long way from transforming the
ideal into actuality. It is important for this country and the world
that the objectives of the United Nations as regards nutrition should be
recognized and put into practice in every state and community in this
country. The implications are tremendous. A great expansion in
agriculture along selected food lines would be necessary to meet this
need. In 1940, Gavin, Stiebeling, and Farioletti (1) estimated that,
if the average consumption of protective foods could be raised to the
level of families whose food intake was rated as "good" from the
standpoint of nutrition, the increase in national consumption of milk
would be 20%; butter, 15%; eggs, 35%; tomatoes and citrus fruit, 70%;
and leafy green and yellow vegetables, 100%. That much of our
dietary inadequacy is due to lack of purchasing power is clearly indi-
cated by extensive data showing that it is the members of the lowest
income families who have the greatest deficiencies. It is also indicated
in the changes occurring in our food consumption during the war.
Stiebeling (5) estimates that, with the increased power to buy due to
high employment and better wages, at least one family out of every
seven which had a poor diet in 1936 was able to obtain a fair or good
diet in 1941.

What increased purchasing power will do is also shown by the
fact that rationing became necessary during the war although the amounts
of food available for civilians were greater than ever before. This has been
true of all foods except sugar and syrups, coffee, tea and cocoa, and
fats and oils. The increases in 1943 as compared with prewar levels
were about as follows: milk and milk products 17%; meat 6%;
eggs 16%; potatoes and sweet potatoes 9%; pulses 22%; tomatoes
and citrus fruit 17%; leafy green and yellow vegetables 9%; and
other vegetables 5%. It must be remembered that these figures
represent classes of foods and that there were decreases in certain
selected, popular food items such as beef and canned fruits. In terms
of nutrients, the increases are even more amazing: animal protein,
^; vegetable protein, 3%; calcium, 15%; iron, 14%; vitamin A,
^; ascorbic acid, 7%; thiamin, 37%; riboflavin, 18%; and niacin,
14% (2). Full employment and higher than usual wages made many

464

SOCIAL ASPECTS OF NUTRITION

good foods so scarce that they had to be rationed. That our at^riculture
can produce the foods we need if there is an effective detiiand was
shown by the fact that, in addition to the greater suppHes available
for civilians, huge amounts were allotted to the armed forces and
Lend-Lease.

Having accepted our responsibility as the nation which spon-
sored the Hot Springs Conference, the basic problem now is how the food
adequate for health can be made available to our low-income families.
That it can be done has been shown by war experience. We must
leave the method of doing it to our legislators. Once it is done, we
can expect the greatest era of national well-being and agricultural
prosperity we have ever known.

The attack on the problem is already under way. During the
days of the food "surpluses," one outlet for food was the school lunch
room. At that time, this use of the food did not conflict with the usual
trade channels and much food was disposed of in this manner. Now,
the school lunch program is just beginning to stand on its own feet in
the sense that the need for such a program is recognized: it is not being
carried out as a means of disposing of food surpluses. Many people
are at last seeing the fallacy of providing expensive buildings, equip-
ment, instruction, and books to a child with a mind and body retarded
by malnutrition. Yet, in many places, the idea still persists that it
is no concern of the educational system to make an adequate meal
available at the school. In many schools the child does not even have
the opportunity to buy an adequate meal, although the modern school
is often a long distance from the child's home; there are no feeding
facilities whatever. In other schools the lunch room is operated by
volunteers or with poorly paid and poorly trained personnel with
neither the necessary knowledge nor funds to supply the needed food.
Every medical appraisal of groups of school children reveals large
numbers with evidence of malnutrition. Children are found arriving
at school without breakfast, while many do not have the funds to
purchase an adequate lunch. Those who carry lunches from home
frequently bring inadequate food for lack of knowledge of nutrition in
the home and because of the difficulties of packing and carrying an
adequate, appetizing lunch. An obvious step would be a method of
supplying an opportunity to obtain an adequate lunch to all school
children throughout the country.

W. H. SEBRELL

In the past, the appHcation of our knowledge of nutrition to
industry has not received any more attention than we have given to
nutrition in our schools. The right of a workman to have an oppor-
tunity to obtain an adequate meal at his work received little recognition
until, under the stimulus of the necessity for the utmost production in
war industries, attention has finally been paid to malnutrition as a
factor affecting a man's ability to stay on the job and to do a good job
under pressure. In-plant eating facilities vary from none at all (and
little or no lunch interval) to the finest type of adequately supervised
feeding arrangements with ample time to eat. The benefits both to the
management and the worker have been so great that within the past
two years a large number of industrial establishments have made
great improvements in this respect even under wartime difficulties.
It has been estimated that before "Pearl Harbor" less than 20% of
our industrial workers had access to any type of in-plant food service.
By January, 1944, about 6,500,000 of the 22,000,000 people engaged
in war industries were receiving meals through industrial feeding
facilities (3). Here again we see a movement toward the goal of an
opportunity for an adequate diet for everyone, but with the goal still
a long way off. From the scientists' point of view, the mere supplying
of food is not the objective, although many think the goal has been
achieved when this is attained. The fundamental fact which always
should be in the forefront is that the objective is to supply nutritionally
adequate food. The mere installation of feeding facilities in schools
and industries does not furnish the opportunity the scientist has in
mind. The installation must also include supervision by adequately
trained dietitians and nutritionists so that the knowledge of nutrition
is applied in making every meal furnish a maximum of nutritive value
in appetizing form.

Although these two programs attack important parts of our
national problem, there are many other groups for which no wide
general provision has yet been made in this country, such as pregnant
and lactating women, the preschool child, and the housewife. Recent
research has demonstrated the very great importance of adequate
nutrition in the prenatal period to the welfare of both mother and
infant. The complications of pregnancy and delivery are less and the
infant is healthier, grows better, and is more likely to survive if pre-
natal nutrition has been adequate. With our future national welfare

466

SOCIAL ASPECTS OF NUTRITION

at stake, is not every infant entitled to the opportunity to obtain a diet
adequate for its growth and normal development? No national
program with such an aim exists in this country today. In some coun-
tries adequate infant feeding formulas are prepared, bottled, and
distributed under public supervision to infants in low-income families.
We are woefully behind in our application of the knowledge of nutri-
tion our scientists have supplied. We have hardly begun to create
the opportunity for our people to obtain adequate nutrition.

The other great cause of malnutrition in the United States, in
addition to the general dearth of opportunity to obtain nutritionally
adequate food, is the lack of knowledge of the need for, and what con-
stitutes, good nutrition, as reflected in bad food habits, poor food
preparation, and general ignorance and indifference about the types
of food necessary for good nutrition. The practical aspects of our
scientific knowledge of nutrition should be conveyed to everyone in
some manner other than through advertising or the educational ma-
terial supplied for the purpose of promoting the use of some particular
food or medicine. Nutrition education should be a part of the cur-
riculum of every school. Clear, understandable, correct, and un-
biased information should be available to every family. A start has
been made in this field also, but it is only a start; although it has been
going on for many years, it still has reached and affected only selected
groups of people. There is a wealth of literature available to those
who know how to obtain it and with enough education to read and
understand it. Agricultural agencies and health departments have
been especially active in this field. Under the war stimulus thousands
of women have attended Red Cross courses in nutrition; nutrition
committees and other organizations interested in nutrition have
reached many more thousands with educational material.

However, much more is needed; for literature is only part of
the answer. Demonstrations, cooking schools, exhibits, radio and
newspapers, moving pictures, and every other educational medium
should be used to the fullest extent. Lunch rooms in schools and
industrial establishments should be used, not only for feeding, but also
for practical education in good nutrition. Studies have demonstrated
that educational efforts here pay immediate dividends in influencing
food selection. The weakest point in nutrition education is that it
has not been given at the point of application, namely, the grocery

467

W. H. SEBRELL

Store, the restaurant, and the lunch room. The educational material
has not been tied in with the food seller and distributor. Until nutri-
tion education is applied at the point of sale of food and the available
food is related to the educational material, we cannot expect to obtain
maximum educational results. Furthermore, the group to whom
education would mean the most, namely, the children, has been largely
neglected. Nutrition education in our secondary and high schools
still consists very largely of teaching the way to prepare appetizing
food. Litde attention is given to nutritive value, to nutritive losses
by various methods of cooking, or to the necessity for an adequate diet.
The chance to establish good food habits at an early age in the ele-
mentary schools is ignored. No attempt is made to correct bad food
habits established at home, and a golden opportunity is lost to establish
good lifetime food habits. Class room lectures and cooking classes
should be closely correlated with the foods served in the school lunch
room. A start has been made in this field also — and an increasing
number of schools under the stimulus of the Office of Education are
making progress in this field, but again we still have a long way to go.
Recently, great advances were made in our knowledge of the
effects of different methods of food preparation and handling on the
nutritive value of food as it is eaten. It is now well known that some
methods of preparation result in practically complete loss of some
nutrients, while other methods of preparing the same food will con-
serve much of its nutritive value. The effect of this knowledge of food
preparation on the population at large has not been material because
it has not been applied: we still have mashed potatoes, overcooked
vegetables, and the use of too much water in cooking; cooked foods
frequendy are held on warming tables for hours before consumption.
These are all well known as methods of handling food which result in
a high loss of nutritive value; much educational effort has been spent
on attempts to correct these practices. In addition, most of us eat
certain foods only because we like them and not because they are good
for us. Too many of our cooks do not know how to prepare food
that is both appetizing and nutritious.

A new factor of great public interest has been introduced into
the nutrition picture by the discovery of the chemical structure of several
vitamins and the development of methods of synthesis on a commercial
basis. Here science has furnished us with a new, powerful, and valuable

468

SOCIAL ASPECTS OF NUTRITION

weapon against nialnutrition if we use it properly. Pure \itamins are
invaluable in the treatment of deficiency diseases, l)ul (licir widespread
use by the public requires technical guidance if we are to avoid eco-
nomic waste and a false sense of security. It is obvious that the use of
multivitamin and minercil pills can be no substitute for an adequate
diet. Since we still have not identified all of the dietary essentials, a
suitably varied diet covers these defects in our knowledge. Even if
we knew how to put all of the dietary essentials into one capsule, at-
tempts at mass use of such a preparation would fail because of lack of
public acceptance. Our aim should be to teach people how to meet
their nutritional requirements with food. The widespread use of vita-
min and mineral supplements seems clearly indicated in situations
in which, for any reason, an adequate food supply is not obtained. In
such cases, a survey of the diet available and clinical examination for
evidence of deficiency should be the basis for dietary supplementation
which then may be handled individually or collectively as the situation
indicates. One of our great nutritional advances has been the intro-
duction of enriched white bread and flour. Here, a cheap, widely used
food has been employed as a vehicle for additional supplies of vita-
mins and minerals which are not being obtained in sufficient amounts by
that part of the population using bread and flour in the greatest amount.

However, the addition of vitamins to foods is not something to
be done haphazardly. Carefully selected, widely used, suitable foods
supplemented by substances for which there has been shown to be a
widespread need are indicated. Other examples of the beneficial
and successful use of this principle are the addition of iodine to salt,
of vitamin D to milk, and of vitamin A to oleomargarine. The forti-
fication with vitamins and minerals of a large variety of foods of limited
use would be both wasteful and unnecessary. It might defeat the very
purpose of the practice by increasing the price of such foods, thus
tending to place them beyond the reach of those who need them most.

An important trend in nutrition research has been the forma-
tion of the Nutrition Foundation by the food industry in order to sup-
port fundamental research in the nutrition field. Grants for research
on nutrition are made to universities and scientific laboratories through
an independent scientific advisory committee. This enables the food
industries to contribute to the advance of our knowledge and to obtain
and make the best use of the most recent advances in nutrition research.

W. H. SEBRELL

The organization of the Nutrition Foundation represents the far-
sighted viewpoint of these industries, which reaHze that the closer we
approach dietary adequacy for everyone the sounder the basis for their
business, a viewpoint shown also in a change in much of the food
advertising, which is now frequently directed at the general field of
improved nutrition. The leaders in this field are not only promoting
the best interests of their business but also contributing to the welfare
of the country. The scientist can do much to direct this great public
educational force along sound lines. If the results of research in
nutrition are to be of public use and benefit, the scientist must come out
of his laboratory and assist in their application. He cannot, like a
famous mathematician, hope that his research will be useless so that
he can work without distraction.

In the field of medical science, the biochemist has given the
physician new and valuable methods in the diagnosis of nutritional
failure; but, because of difficulties in the interpretation of clinical
signs of early deficiencies, a controversy continues as to the extent and
importance of deficiency diseases in this country. We still need better
methods of diagnosis to enable us to recognize the earliest manifesta-
tions of nutritional deficiencies.

There has been a rapidly growing recognition that adequate
nutrition is of great importance in convalescence from any disease,
and that nutritional deficiencies frequently occur as complications of
many unrelated diseases, especially where there has been marked loss
of appetite, nausea and vomiting, diarrhea, disturbed metabolism, etc.
With the increased awareness of the importance of adequate nutrition
in medicine and the growing complexity of the subject, we now see
medical and public health schools teaching nutrition as a part of clinical
and preventive medicine, as well as of biochemistry and physiology.
There is reason to hope that, in a few years, our need for physicians
well grounded in the subject will be met.

The lack of statistics on the distribution and prevalence of the
deficiency diseases makes it exceedingly difficult for the public health
officer to attack the problem effectively from the viewpoint of pre-
ventive medicine, although here we find an active interest being taken
and a start being made. Many health officers now see that the pre-
vention of malnutrition is part of their obligation to provide an oppor-
tunity to attain the best degree of health. This aspect of nutrition

470

SOCIAL ASPECTS OF NUTRITION

could not develop until our knowledge had advanced to the point at
which the scientist could show the health officer that malnutrition is
a widespread health problem and that human dietary needs are suffi-
ciently well known for the health officer to apply adequate remedies.
In spite of the present lack of knowledge about the extent of the prob-
lem, the health officer is laying the foundation for more extensive work
in the field. A start has been made in almost every State, although
there is still no clear definition of the best method of approach or of the
way in which the health department can work to the best advantage.
The problem is made much more difficult for the health officer by the
fact that he recognizes that an adequate solution, involving as it does
education, agriculture, and food distribution, means that he must
cooperate with many agencies not concerned in most public health
problems. One of his most important duties is to determine what
deficiency diseases exist in his community, how many people are
affected, and where these people are located. He also can carry some
of the educational work, and his voice in pointing out the importance
of the problem can do much to secure action by other interested agen-
cies. In order to bring about the necessary coordination, nutrition
committees have been set up in every State, bringing together for the
first time the various agencies interested in different phases of nutrition.
These committees are wartime committees, but they should continue
in peacetime to form a unified approach to all aspects of nutrition
without which we cannot obtain a permanent correction of our diffi-
culties.

While the ultimate goal of an opportunity to obtain an adequate
diet for everyone is still far in the future, just the recognition of its im-
portance and the realization that it may be obtainable and that it
should be a part of any plan for permanent international peace is a
long step toward success. Unfortunately, only a few recognize these
factors today. The next great forward step must be a general recog-
nition that the future development of the nation to a maximum of
health and prosperity depends on adequate nutrition for everyone.
Scientists can do much to make their work of the greatest use to man-
kind by assisting in bringing this about.

This war has shown that great advances in nutrition and health
are within our reach. Gains during the war were made possible by
some factors on which we could not count in peacetime, but there

W. H. SEBRELL

is one factor which it will be our opportunity to continue to make
available now that the war is over — the participation of scientists in the
task of improving the nutrition of people. They must play a larger
role than in the past in applying the results of their work. Research
work in nutrition has, in recent years, achieved an enviable record.
It will earn the gratitude of mankind if a greater effort is made to give
human beings the benefits of its results. The nutrition scientist can
play an even more important part in the future than he has in the
past if he will, on occasion, come out of his laboratory and mingle with
the people in the market place. He will go back to his essential work
refreshed by a greater appreciation of the way in which his results
contribute to the improvement of human welfare and by his contact
with the every-day problems of life.

References

(1) Gavin, J. P., Stiebeling, H. K., and Farioletti, M., "Agricultural Sur-
pluses and Nutritional Deficits," in 1940 Yearbook oj Agriculture, Farmers in
a Changing World. U. S. Dept. Agr., Washington, 1940.

(2) Food Consumption Levels in the United States, Canada, and the United King-
dom. Report of Joint Committee of the Combined Food Board, U. S.
Dept. Agr., April, 1944.

(3) Goodhart, R., "Protecting the Health of the Industrial Worker: Nutri-
tion," in New Steps in Public Health. Milbank Memorial Fund, New York,
1945.

(4) "Recommended Dietary Allowances," Natl. Research Council Reprint Circ.
Series, No. 115 (1943).

(5) Stiebeling, H. K., "Adequacy of American Diets," in Handbook of Nutrition.
Am. Med. Assoc, Chicago, 1943.

472

31

ORGANIZATION AND

SUPPORT OF SCIENCE

IN THE UNITED STATES

L. C. DUNN, PROFESSOR OF ZOOLOGY, FACULTY OF PURE SCIENCE,

COLUMBIA UNIVERSITY

THE WAR and the sudden need to improvise means for
supporting and directing war research have brought into
high relief an important fact which has been dimly recognized for
many years: there has been in the United States no orderly means for
the continuous support of fundamental scientific research, and no
policy or method for the deliberate utilization of science by our society.
Science has been a hardy plant which grew where and how it could,
thriving in the comfortable greenhouse of a research institute, or
turning ample fertilizer into real fruit in an industrial laboratory, or
in the more usual case struggling for sustenance in the thin soil of
colleges and universities, occasionally enriched by temporary growth
stimulants from a foundation or private donor. Except in the case
of certain industrial developments and in a few government depart-
ments, the support of science in the United States has not been the
result of decision but of chance, operating in a milieu which contained
good scientists and a good deal of fluid wealth.

The most blunt and truthful statement we can make about the
reason for the lack of continuity and of public policy regarding science

473

L. C. DUNN

is that, as Americans, we did not want either continuous support or
direction or planned application of science. The detailed causes of
this attitude trace in part to reasoned premises and in part to prejudice;
and from these there has resulted a confusion of thought which the war
has now revealed.

The contradictions come out most clearly in the views of scien-
tists concerning the support of science after the war. Most of them
hope for release from the capricious and precarious methods by which
fundamental research was chiefly supported before the war, namely,
by periodic begging from donors, such as foundations who 'chose the
researches to be supported. Scientists generally hope for a more
orderly and stable means of support than this, yet most of them would
not turn to the Federal Government as the source of more continuous
support. They profess to fear infringements on their freedom more
when support comes from their government than when it comes from
private sources.

There is no sense in dodging or belittling the dilemma in which
this places science. On the one hand, the war agencies which have
guided and financed a large segment of scientific research propose to
withdraw from this function. If they do, the public investment in
scientific research will drop to a third or a quarter of its present level.
At the same time, the principal sums in the hands of the great founda-
tions are declining and science must adjust itself to diminishing support
from this and other private sources, and possibly to the extinction of
this sort of financial aid within another generation. There will even-
tually remain as sources of support chiefly industry and business,
through their reseai'ch laboratories and foundations, and the govern-
ment, through its own scientific agencies or through new channels
yet to be created.

Most scientists who do not like "domination of science by
government" like "domination of science by industry" even less; and
many have already objected to the influence which the foundations
wield because of their control of the fluid funds with which to supple-
ment the fixed investments of universities and research institutes in
men and permanent plant. It has often seemed that this small tail
of free funds has wagged the larger dog of solid investment.

Moreover, scientific research depends upon trained men and
women as much as upon material facilities, and we have as yet made

474

ORGANIZATION OF SCIENCE

no provision for assuring a steady flow of young scientists into research.
For advanced training we have rehed upon the existing scholarships
and fellowships of the universities, which are so meager that most young
scientists can devote only a portion of their time to learning, the rest
being needed for earning a living; and upon the advanced fellowships
supplied by foundations, private philanthropy, and industry. The
same considerations of approaching exhaustion of private funds apply
to the training of persons as to the provision of research funds.

The facts that must be faced are, then, that the present means
of support of science are running out and, whether we like it or not,
changes in the sources and form of support will occur; and that a
chief desideratum for scientists will be to keep science under the new
conditions as free as possible to develop according to its own inner
needs and according to its function in society.

In the following pages I propose to discuss, first, what the func-
tion of science is that entitles it to support; second, what determines
the attitudes of scientists toward forms of support; third, what general
public policy toward science would represent the best interests of
science and scientists; and, fourth, how this policy could be imple-
mented in practical ways.

At the bottom of every consideration of science in its public
aspects must lie the question: "What is science for?" When this
question is squai'ely and thoughtfully faced, scientists will agree that
science exists for man and not for itself alone. As a means of under-
standing the material world, it leads toward the improvement and
control of the environment in which human society must always
operate. Eventually, its results and the methods of thought which it
develops accrue to the public good, not merely by increasing the physi-
cal well-being of the people through technological applications, but
also by extending the domain of reason and by increasing our under-
standing and appreciation of nature. In discussing the material means
which have to be provided for scientific research, it is often forgotten
that the great and lasting changes wrought by science are in men's
minds, and that, in the end, science is to be supported for the same
reason that education is to be supported. The products of science are
primarily increase and diffusion of knowledge and increase in the
number of trained minds, and secondarily increase of technical facilities
and production of goods. Like other knowledge, scientific under-

475

L. C. DUNN

standing is one of the "rights" to which all citizens should have equal
access. Its support, like that of education generally, is thus to be
shared, as most essential activities are in our society, by the State and
by "public spirit" as it acts through foundations, private citizens,
and industry. At the material level, science in the modern world
has become a public necessity without which technical advances and
social developments determined by them cannot occur in an orderly
way. It has become so "affected with the public interest" that its
support must be a matter of public concern. The scientist has thus
become in some sense and in spite of himself a public servant.

The many scientists serving their country during the war as
scientists are less likely now than formerly to forget their public func-
tion; but in the past a failure to recognize this led scientists as a class
to have too little confidence in seeking support for scientific work.
They were not sure that science was worthy of public support, because
oftentimes science was not what the world needed, but only what they
enjoyed doing. They did not generally think about a public policy
for science because they were not clear about the public function of
science. Can we really expect (they would say) the public to support
this kind of work? Or as a small boy said to a scientist after a visit
to his research laboratory, "Uncle, do they really pay you for doing
this?"

When questions about the organization and support of science
were raised, however, other reasons were generally given for either
opposing the formulation of policy or avoiding the question altogether.
These reasons took different forms, but in general had their roots in
our tradition of individualism. Since scientists have usually been
strong individualists, the traditional public objections to schemes for
the support and direction of science have been strengthened and ra-
tionalized by the scientists themselves. They said: "Organization
kills initiative," "Planning interferes with free enterprise," or "Con-
tinuously assured support removes the need for periodic justification
of each research on its own merits." "Support implies direction,
and he who pays the fiddler will call the tune; and only scientists
can know what tunes can or should be played."

These are valid and weighty objections and they must be
squarely met by any general proposal for the maintenance or direction
of science. It is nevertheless true that these are not the primary or

476

ORGANIZATION OF SCIENCE

real reasons for opposing the formulation of a public policy or even
specifically for opposing the support of science from public funds,
since the same scientists who use them against government support
approve the use of organization, planning, continuous support, and
central direction when these are employed, as a matter nj policy, in the
great industrial laboratories. In fact, many scientists point with pride
to the splendid results which industrial laboratories have achieved
under the very conditions which they allege would impede and stifle
scientific research done at the expense of government. Moreover,
public support and direction appear to have been quite acceptable
in the great program of agricultural research which has been in opera-
tion since 1887 through the United States Department of Agriculture
and the State Agricultural Experiment Stations. These facts are
not cited to minimize the difficulties inv^olved in planned continuous
support and direction of research. They do show clearly, however,
that the objections are generally not to support and direction as such
but to these only when the authority which wields them is the Federal
Government. As the attitude toward agricultural research shows, the
objection does not apply with similar force to the State governments.
Many scientists have expressed the fear that central and especially
federal support of scientific research would put an end to "scientific
freedom" and lead to "regimentation." In most cases, it is the threat
to scientific individualism or "free enterprise in science" that is the
real cause of fear. Since such changes in modern society as the decline
of individualism are not due to deliberate acts of governments but re-
sult from the social and economic and technical developments of our
age, they call, not for fear, but for a greater effort to understand
them.

I believe that most scientists have come to realize the nature of
such objections to discussing general policies for the public support of
science. The central position that "pure science," especially physics,
came to occupy in war research revealed facts about science in the
modern world which simply could not be evaded or overlooked. Even
the need of "coordination," the blackest of the beasts which threaten
the research scientist, became evident as soon as the war imposed
pressing requirements which an unplanned, uncoordinated science
could not meet. The knowledge that our enemies had succeeded in
so organizing their researcli and development programs that they had

477

L. C. DUNN

"got the jump" on us in numerous ways persuaded even reluctant
individualists that coordination was absolutely necessary.

The war emergency also revealed the lack of balance which
obtains when science is directed by chance. Many fundamental prob-
lems, upon which other inquiries depended, had not been touched and
efforts had suddenly to be made to straighten the fiont. If this was
borne in upon those scientists who participated in war research, it
became even clearer to those who through lack of organization were
left out. There are now many biologists who would sacrifice their
cherished individualism for the sake of being identified with a great
national effort. They realize that the neglect, the omission almost, of
biology and biologists from the hastily improvised war agencies was
bad not only for biology and for other sciences, such as the medical
and agricultural sciences which depend upon biology, but for the
nation. Their state of mind is not improved by the reflection that, by
and large, the fault was their own.

Still other changes in the attitudes of scientists are due to the
growing realization that research workers need to recognize the connec-
tion between their own special work and the general scientific structure
in which it will find its place and its function. It is difficult for the re-
search worker to envisage this larger field without inquiring too about
the still wider frame of society in which science operates. Many more
scientists than formerly now believe not only that this social awareness
of the men who do the work of science is needed to make a social
being and a citizen of the scientist, but that this is essential in the na-
tional interest. Those who so believe will want to face the questions
involved in the public support of science.

By these paths we come to the problem itself: what public
policy toward science would encourage the best growth of science and
its use for the welfare of the people? The aims of policy must be to
reconcile two basic requirements, about which there is probably general
agreement.

(1) Science and scientists must be free to grow and change
in ways determined in part by the discoveries of science itself. This is
the way in which science has progressed in the past — and the autonomy
of small groups and the feeling of freedom of the individual to follow
the new idea wherever it may lead are goods which must be preserved.
This freedom must be accepted and guarded as a matter of principle;

478

ORGANIZATION OF SCIENCE

and provisions for freedom of publication and the prevention of arbi-
trary censorship must be a part of the basic poHcy.

(2) The forms of support and organization of science must be
determined by social needs and purposes and are therefore matters of
concern not only to scientists but to government and to the ultimate
beneficiaries of science, that is, the people, as consumers and workers.
Those who most directly need and use the results of scientific research
in education, industry, agriculture, medicine, and public health have a
special interest in the development of science, and means must be
provided by which this influence can be exercised. The two primary
conditions should therefore be: (a) a central organization by which the
conduct of science is made responsive to public requirements and needs;
and (b) the representative character of the directing agency or agencies,
insuring democratic methods in administration.

These two requirements of autonomy, on the one hand, and
subservience to social needs, on the other, have seemed antithetic to
some, but I do not believe this need be the case. There is much
evidence of the vitality and progressiveness of science in other countries
where it is largely under public control. The extreme example of public
control is in the Soviet Union, where the direction of scientific research
is centralized in the Academy of Sciences, through which the support
of the state flows to all of the research agencies. Other European
countries occupy positions intermediate between this maximum and
the minimum reached in the United States, where almost alone among
modern nations science has retained a predominantly private character.
Even here, the wartime activities of the Office of Scientific Research
and Development and the Committee for Medical Research show that
no essential incompatibility exists between research and public control;
while the long peacetime history of United States Government scien-
tific departments and especially of the Department of Agriculture illus-
trate the feasibility of accomplishing at once a scientific and a social
purpose.

Much experience in the United States and in other countries
indicates that, to obtain the maximum results from a given eflfort in
scientific research, the interests of the research workers themselves must
be consulted, but that these are not fundamentally diff'erent from those
of the community around them. Scientists traditionally are primarily
devoted to their work, often sacrificing other interests to it and exclud-

479

L. C. DUNN

ing other interests which tend to interfere with it. Yet, as the war
shows, they will voluntarily and gladly place this devotion and their
technical ability and intelligence at the service of an objective which is
clearly defined and compelling. On the other hand, directing agencies,
public or private, do not grudge to the scientist a greater measure
of freedom than to other workers, provided they are assured of his
adherence to the principles of service and to the general purpose which
they consider essential, and that this freedom actually produces the
results expected from it. Freedom within a general plan is a practical
ideal at which to aim, as the comparative freedom of local political
units within the general frame of Federal Union of the United States
shows.

Voluntary cooperation of scientists with public agencies in the
planning and execution of research would seem to provide the soundest
base. The greater tendency toward teamwork and pooling of ideas
by groups of scientists, the distribution of responsibility and credit for
scientific work among the whole staff of a laboratory, the greater dif-
fusion among younger scientists of the sense of social responsibility, and
the resulting tendency for social incentives to supplement more purely
personal motives — these facts all indicate that it is reasonable to expect
that scientists can and will participate in formulating the plans they
will execute. This leads to the kind of self-government to which demo-
cratic administration tends, and which industry has found valuable as
an incentive.

A further question that policy must meet is the ultimate dis-
position of the new knowledge which accrues from science. In the large
segment of scientific research under private control, it is generally
agreed that the ownership of valuable processes arising from research
is to be vested, not in the individual scientist, but in the laboratory or
the industry which has financed the research. Patents therefore gen-
erally become the property of the corporation by which the scientist is
employed.

The question of ownership has already arisen concerning values
accruing from war research, and it must enter inevitably into all plans
for the future support of science.

The clearest basis for policy in this regard is that research done
for a social or public purpose must be brought as quickly as possible
to serve this purpose. If it is carried out for the public and at public

480

ORGANIZATION OF SCIENCE

expense, it should belong to the public; and there is no more direct
way of making it public property than by publishing it as soon as the
facts are clear. Publication would preclude patenting and, with
certain precautions to be discussed below, would prevent the results of
public science from becoming private property. But, by the same
token, the results of private science would remain pri\'ate, subject to
patent or other ownership rights and restrictions.

A division of this sort already exists. Most agricultural research
in the United States is done at public expense and results are freely
published and can be consulted and used by anyone. The greatest
change in American agriculture in the present century, the introduction
of crossbred or hybrid corn, resulted chiefly from cooperative research
between the United States Department of Agriculture and the State
Agricultural Experiment Stations. The results were quickly utilized
by private seed companies, none of which was able to obtain a patent
or found a monopoly on it. Crossbred corn therefore came very
quickly into general use and its benefits were soon spread over all agri-
cultural communities.

Side by side with this development, it was possible for private
individuals and corporations to produce and patent new varieties of
other plants, such as roses, which could be propagated asexually. The
ownership of new rose varieties is thus (in general) private; but the new
method of corn breeding belongs to the public.

The question of property rights need then be faced only when
new values are created by publicly supported research; and the basic
policy stated above — that is, free publication of the results of public
research — need not interfere with existing arrangements under which
private research operates. As a matter of fact, the more fundamental
the research in the sense that the more general the truth that arises
from it, the less will property questions arise. It is hard to find a
patentable value in the general theory of relativity, or in the periodic
system of the elements, or in the theory of the gene. It is the fate and
the function of such ideas to become common property, and no man-
made rules should be allowed to interfere with their free circulation.
It is usually only the specific applications of general ideas which be-
come subject to property restriction; and public policy can only aim
at preventing such restriction from interfering with the advance of
science or with the spread of the benefits to the people.

481

L. C. DUNN

It is time now to deal briefly and in bare outline with the last
question: how can these ideas and hopes about the support of science
be brought into practical operation?

It seems evident that there must be an agency having as its
chief concern the preservation, advancement, and diffusion of scientific
knowledge. There are, in the United States, dozens of organizations
having this aim in limited spheres, but that not one of them fulfilled
the required functions in the national interest became evident when, in
the war emergency, a wholly new and temporary agency, the Office
of Scientific Research and Development, had to be created. The im-
portance of the work assigned to this office, and the power and facilities
which accompanied the responsibility, pointed not only to the need but
to the method of meeting the need for a central agency of government
concerned with science.

It is probable that nothing less than the creation of a cabinet
department of science under a Secretary of Science can permanently
meet the need. It ought to be connected directly with the central
executive body of the government, because only in such a position can
it be made aware of the basic problems which face the nation, and
only through the political power which attaches to cabinet rank can it
gain the means and facilities with which to support the study of both
immediate and long-term problems.

The structure of such a department may well be different from
that of other government departments because, in addition to policy
making and administrative functions, it would have to serve as a
coordinating agency for many existing scientific agencies, both public
and private. To name only two groups of interests, it would have to be
closely connected with the universities and research institutes, and with
industry, since in each of these institutions needs for new knowledge
are likely first to become apparent, and from each flows scientific and
technical information which can be put to use in national defense and
development.

At the heart of such a department could well be a board or
council of scientific research which could act at once as a granting
agency, allocating funds for specific researches, and as a board of strat-
egy, seeking out neglected areas, mobilizing disparate facts and dis-
tant persons, and shifting its forces from time to time to explore new
avenues of research. If it fulfilled its best purpose, it could not be

482

ORGANIZATION OF SCIENCE

content to sit and sift, but would itself have to search and ponder in a
more active way. Its basis of operation as a granting agency might
well be patterned upon the Office of Scientific Research and Develop-
ment in that it might receive applications for research funds from
universities, research laboratories, other government agencies, or even
individuals, and might enter into contracts with those it judged as
offering the best prospects for needed scientific advance. Like
O. S. R. D., it might find no need to become an operating agency with
plants and facilities of its own, although it should have some freedom
to use those methods best calculated to promote the best research.

Much would depend upon the composition of this board. It
should consist of working scientists who can judge the merits of various
research proposals and policies, and of representatives of those fo'-
whose benefit the research is done and who in the end pay the bills,
that is, the public as represented by labor, consumers, and industry,
small or large. Perhaps a proportion of eight scientists and four public
representatives would express both the purposes and responsibilities of
the board; and some of the scientists should be drawn from, or be pri-
marily interested in the scientific work of, the government departments.
Since there should be no disposition on the part of such a board
to displace any existing research agencies, but rather to supplement and
aid them, its most important function might well turn out to be,
especially in its initial operations, that of coordinating and facilitating
research generally. It would undoubtedly avoid competition with
industrial research, and direct its first attention to "unprofitable"
fields such as exploration looking toward new natural resources, hous-
ing, public health, etc. It would probably be concerned with such
public services as the provision of adequate means of publication, of
bibliographic and library services, of abstracts and translations of foreign
scientific literature, and similar functions.

Either this board or another one in the Department of Science
would of necessity concern itself with one of the basic questions in all
scientific research: how to insure an adequate supply of trained scien-
tists for research, for education, for industry, and for public service.
Its operation in this respect could well be patterned upon the fellow-
ship boards of the National Research Council, which at present ad-
ministers limited and temporary funds supplied from private sources.

Two main criticisms of the proposal outlined above may be

L. C. DUNN

anticipated. One is that research cannot be free under a central
direction, but will wither and die. Scientists, it is said, will not submit
to regimentation, nor can new ideas, the life blood of science, be cre-
ated by subsidy. The other criticism is that the needs are already met
by such existing agencies as the National Academy of Sciences and the
National Research Council.

The first criticism is certainly a cogent one when central con-
trol is proposed, but it applies with less force to a board which judges
applications initiated by working scientists as individuals or groups,
especially when many of the judges are themselves working scientists
who know how delicate a plant original research is and how necessary
is the atmosphere of freedom to its growth.

Much will depend upon the degree to which members of the
board realize that any organization of this sort exists primarily to
provide a material body for the mind of science. There are scientists
and others who know this and who apply to organizations proposed for
science two essential criteria; Does it provide the mind with adequate
and proper facilities? Does it leave the mind free to strike out in new
directions? Men who ask these questions are the ones whose sense of
public duty would bring them into the service of such a board, just as it
brought such men into the direction of war research.

In regard to the second criticism, it must be pointed out that
in the war emergency neither the National Research Council nor the
National Academy of Sciences proved to have the character needed
for an agency to guide and administer the organization and support of
science. Neither is an operating agency; and, as constituted at present,
neither could provide the initiative and the administrative services
which are required. The relative isolation in which they have func-
tioned has removed them from that close connection with problems of
public policy so essential for an agency to have which is to be responsive
to public needs. They have the confidence of scientists and close con-
nection with academic research and with the scientific societies and
organizations and are thus well prepared to serve an important ad-
visory function. The National Academy of Sciences, as a council of
elder statesmen, could well be called upon to pass upon the qualifica-
tions of scientists proposed for membership in the Board of Scientific
Research. The Academy would be less able to maintain sufficiently
close relations with consumers, with labor, and with industry, and it

484

ORGANIZATION OF SCIENCE

would be less competent to advise on questions bearing on the social
relations of science in these fields.

The Board might conduct its relations with the scientific socie-
ties through the National Research Council, which could then be
incorporated into the Department of Science and carry out other im-
portant functions, such as maintaining a permanent roster of scientific
personnel.

It is of course possible that the Academy and the present Na-
tional Research Council might be so changed as to assume the func-
tions it is proposed to assign to the Board. The changes would be so
fundamental as to constitute conversion of these older organizations
into a new department of the government; and it is probable that the
traditions of both institutions would make such conversion a slow and
difficult process, for, in spite of their "national" character, neither
has felt itself to be a truly public agency.

The foregoing brief sketch has given merely in outline form,
without any attempt to develop detailed procedures for giving them ef-
fect, some of the general ideas and principles which might underlie a
policy for the public support of fundamental research. Since it was
written (January, 1945) several events of cardinal importance have oc-
curred. One of these was the publication in July, 1945, of "Science,
the Endless Frontier," a report to the President of the United States
by Vannevar Bush, Director of the Office of Scientific Research and
Development. This is at once a report on the state of scientific re-
search in the United States and a set of recommendations concerning
its future support, in answer to specific questions posed by President
Roosevelt in 1944. Since it embodies the result of extensive study by
committees of competent and experienced scientists serving as advisers
to Dr. Bush and is documented with data on current scientific research
activity, it is destined to constitute one of the bases on which public
policy for science will rest. Another foundation had been provided
by the discussions and hearings centering around a bill submitted to the
Senate in 1943 by Senator Harley S. Kilgore of West Virginia. These
two major influences, one provided primarily by scientists, the other by
legislators, came together in the autumn of 1945 during the joint public
hearings on five bills for the public support of science. The message
to Congress of President Truman on September 6, 1945, definitely com-
mitted his administration to the support of fundamental research in

485

L. C. DUNN

both the natural and the social sciences, including the training of young
scientists and the dissemination of scientific information. This fusion
of the efforts of statesmen and scholars is a fact of the first importance
and transcends in its significance for the future the particular provi-
sions for support which are now under discussion. There is no doubt
that the atomic bomb greatly accelerated the speed with which scien-
tists and legislators were approaching each other and the swift realiza-
tion of the significance of science on the part of the public created the
atmosphere in which progress could be made.

As a result of these recent events I think it can be said that the pub-
lic, through its representatives, has now acknowledged that the support
of science is a public responsibility, to be assumed in a broad and perma-
nent form through a new agency of government to be known as a Na-
tional Science Foundation. This foundation will resemble the Board
of Scientific Research described in earlier paragraphs of this article,
but will have even broader powers since it will include a Division of
Social Sciences, and provisions for the training of young scientists and
for international collaboration in science which go beyond those pre-
viously suggested.

It appears that there is now substantial agreement among scien-
tists concerning the major purposes and functions of this Foundation, al-
though there is still some reluctance to grant that the administration of
the Foundation should be primarily responsible to the government
rather than to scientists.

However, it has now become so clear that the aggressive advance-
ment of fundamental scientific knowledge is a primary condition for the
maintenance of democratic government and for the attainment of a
good society that we may confidently expect methods to evolve in prac-
tice by which public control may be exercised without unduly limit-
ing the freedom which science needs. The attainment of this goal will
be hastened if scientists, acknowledging their public responsibility, will
build the means, through a guild or federation with a social purpose,
by which their influence on public policy can be brought to bear.

486

'■)';>''