I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I

s

iaS^^^^^^i3£3SS^B£

Marine Biological Laboratory Library

Woods Hole, Mass.

Presented by

Mrs, E. B. Harvey-
Sept. 1, 19S0

]]

II
11

E

D
D
D
D
II
U
E
B
II
D
D

3B3^^^^^^^^^^^^^E3E

The Chemistry
and Physiology of Growth

The Chemistry

and Physiology of

Growth

J. H. NORTHROP E. S. G. BARRON

F. O. SCHMITT P.WEISS

K. V. THIMANN J. S. NICHOLAS

KARL FOLKERS C. P. RHOADS

C. B. VAN NIEL C. N. H. LONG

EDITED XY ARTHUR %. TARPART

'i

Princeton University Press • 1949
Princeton, New Jersey

COPYRIGHT, 1949, PRINCETON UNIVERSITY PRESS
LONDON : GEOFFREY CUMBERLEGE, OXFORD UNIVERSITY PRESS

PRINTED IN THE UNITED STATES OF AMERICA
BY PRINCETON UNIVERSITY PRESS AT PRINCETON, NEW JERSEY

FOREWORD

AN architect must build upon the past. A study of the growth of
animal and plant organisms must start from a firm foundation
L in the mechanics of growth which finds its most eloquent expres-
sion in D'Arcy Thompson's Growth and Form.

The symmetrical groupings which cells take in forming parts of
organisms, the time required for the attainment of such groups, the
numerous internal and external forces that give form to the individual
cells, the symmetry of form of the whole organism and between organ-
isms, all these and numerous other precisely measured parameters are
the foundations upon which the students of growth must build.

The individual chapters of this book represent a number of sturdy
columns that are being erected upon these foundations. Examined
individually they make the reader want to take part in their construction.
Viewed as a whole they present a hope for a future understanding of
the phenomenon of growi:h.

One of the fundamental processes in growth of plant or animal con-
sists in the manufacture of an ever-increasing amount and variety of
proteins. These proteins form the major structural components as well
as the organic catalysts of all living cells. Hence an analysis of the
mechanisms involved in the "Synthesis of Proteins" (Chapter I) is an
important part of a study of growth.

"Morphology at the Molecular Level" (Chapter II) is rapidly becom-
ing an essential aspect of growth. It has already contributed greatly to
our understanding of the problem of the arrangement of protein and
other molecules within a cell, which determines the cells highly specific
form, symmetry and reactivity.

At this point a question arises. What determines the mass and direc-
tion which these structural units and their accompanying water and
salts acquire in the growing organism? The problems raised by this
question differ somewhat in the growth of plant from that of animal
organisms. For this reason Chapters III and IV deal with "Plant Growth
Hormones" and "Vitamins and Growth Factors."

Organisms, whether unicellular or multicellular, are composed of
cells which have arisen from preexisting cells. This is the fundamental
biological phenomenon of growth. Our present knowledge and future
prospects of understanding this phenomenon have been greatly acceler-
ated in recent years by studies on the "Kinetics of Growth of Micro-
organisms" (Chapter V). This leads naturally to a desire to know more

VI FOREWORD

about the energy sources for the growth processes. What starts and
maintains the chemical reactions that bring about the increase in mass,
the elaboration of new protoplasm? Up to this point we are in the posi-
tion of a man who has the plans for a building, all the many types of
material necessary to construct it, and many willing hands to do the
labor, but has failed to provide the foreman whose job it is to guide,
to direct, to speed up or slow down as occasion demands, the action of
those hands on the material available. In many respects the enzymes of
living organisms are analogous to these foremen. In the case of enzymes
the action is on the chemical reactions going on within cells. "Cellular
Metabolism and Growth" (Chapter VI) brings out many important
aspects of this problem.

The first six chapters of the book appraise many of the basic physio-
logical mechanisms and chemical reactions that confer on the individual
cell its form and function in growth. In the development of the multi-
cellular organism, however, these units of form and function appear as
part of the diverse types of structure that compose the adult organism.
The remaining chapters of the book suggest how to deal with these
problems.

Among the central problems of growth are those concerned with
differentiation of cells. How is it that in embryonic growth, and during
the various types of growth which occur in an adult organism, cells of
similar origin come to possess widely divergent characteristics? What
controls different modes and rates of growth? What are the factors
which govern the establishment of biochemical and morphological
differences among cells? These are the problems of "Differential
Growth" (Chapter VII).

The differentiation of the developing egg as it goes through embryonic
stages to the adult condition presents many "Problems of Organization"
(Chapter VIII). In some studies it has been shown that there are
localized regions of the egg, in some cases visibly different, which
determine the pattern of development. If these areas are disoriented,
visible alterations in the normal development of the embryo occur. In
other studies on the early stages of development of the embryo specific
groups of cells influence the activities of neighboring cells in such a way
that they determine the fate of their neighbors.

Growth does not always proceed in accordance with the functional
needs of the growing or adult organism. All too commonly the cells of
a particular tissue or organ will go on a growing spree of their own.
Such abnormal growths, neoplasms, are a part of the overall problem

FOREWORD vii

of growth. Examination of them should lead to a mutually stimulating
outlook in the future study of both normal and abnormal growth (Chap-
ter IX).

The last chapter of this book rounds out the work that has gone
before. Chemical syntheses, metabolic reactions, differentiation and
organization at the molecular and cellular levels must all be regulated
so that growth may be a coordinated series of events. Our knowledge
of these regulatory mechanisms stems largely from studies on the
adrenal gland. This pillar in the structure which the book attempts to
blueprint centers in the concluding chapter, "The Adrenal Gland, a
Regulatory Factor."

This book is an outgrowth of a conference on "The Chemistry and
Physiology of Growth" that was held in Princeton in September 1946
in celebration of the Bicentennial of Princeton University. Since the
conference, the authors have had the opportunity to change their papers
in the light of comments made then, or to comment on the work of
others. Thanks are here extended to the many persons whose work made
the conference and this book possible, and special thanks are due to the
authors of these contributions for their generous help throughout.

ARTHUR K. PARPART

Princeton University
January IQ48

CONTENTS

I. Enzymes and the Synthesis of Proteins

By John H. Northrop 3

II. Molecular Morphology and Growth

By Francis O. Schmitt 49

III. Plant Growth Hormones

By Kenneth V. Thimann 61

IV. Unidentified Vitamins and Growth Factors

By Karl Folkers ^2

V. The Kinetics of Growth of Microorganisms

By C. B. van Niel 91

VI. Cellular Metabolism and Growth

By E. S. Guzman Barron 106

VII. Differential Growth

By Paul Weiss 135

VIII. Problems of Organization

By J. S. Nicholas 187

IX. Neoplastic Abnormal Growth

ByC.P. Rhoads 217

X. The Adrenal Gland, a Regulatory Factor

By C. N. H. Long 266

Index 285

76922

The Chemistry
and Physiology of Growth

I. ENZYMES AND THE SYNTHESIS OF

PROTEINS

BY JOHN H. NORTHROP'

EVIDENCE is rapidly accumulating to show that all enzymes and at
least some viruses are proteins. As a result, three fundamental
problems, which previously appeared unrelated, may now be
considered as one general problem — the synthesis of proteins.

Solution of the problem requires the explanation of a large number
of very precise experimental facts on the one hand, and is confined by
strict theoretical limits on the other. Each tissue of each species of
plant or animal can form its own special proteins, and these proteins are
characteristic of the organ as well as of the species. It is probable, there-
fore, that millions of different proteins exist. In some cases, as in the
production of antibodies, or adaptive enzymes, the presence of a foreign
compound may cause the organism to develop a new protein having a
very definite and specific relation to the compound which caused its
production. In other cases the presence of a foreign protein (viruses)
results in the formation of more of the foreign protein. The formation
of the type-specific nucleic acid (Avery, MacLeod, and McCarty, 1944)
is an example of a similar reaction.

Whatever the mechanism of protein production, there is every reason
to believe it is controlled in one way or another by enzymes, and there
is good experimental evidence that many enzymes behave as theoretical
catalysts. The synthesis of proteins, therefore, must presumably obey
the laws of catalysis, and any hypothesis purporting to account for the
synthesis of proteins by means of enzyme reactions must conform to
the general theory of catalysis.

The correctness of this assumption is of fundamental importance
since, if true, it rules out many synthetic reactions which might other-
wise be possible. The experimental evidence for the correctness of the
statement is therefore reviewed in the following section.

I. Theory of Catalysis

The fundamental law of catalysis was formulated by Ostwald, who
defined catalysts as substances which affect the rate but do not affect the
equilibrium state of a reaction.^

1 The Rockefeller Institute for Medical Research, Princeton, New Jersey.
2Bredig (1902) modified the definition slightly in order to make it conform more

4 JOHN H. NORTHROP

Theoretical proof of this statement was furnished by van't Hoff, who
showed by thermodynamic reasoning that a substance which is not
changed during the course of a reaction cannot affect the equilibrium
condition (cf. Taylor, 1925, p. 318).^

This peculiarity of catalytic reactions also furnishes the most useful
method of determining experimentally whether a reaction is catalytic
or not. If the substance under investigation acts as a catalyst (enzyme)
the addition of increasing concentrations of the substance will not
change the final products formed, but only the time required for the
reaction to take place. If, on the other hand, the substance takes part in
the reaction, the nature or quantity of the products, as well as their rate
of formation, will change as the quantity of substance added is increased.

Since the catalyst does not change the equilibrium, it follows that if
any one catalyst causes the reaction to proceed from A to B then any
other catalyst which affects the reaction must also cause the reaction to
proceed in the same direction and to the same extent. The assumption of
"hydrolyzing" and "synthetic" catalysts for the same reaction under
the same conditions is therefore in contradiction to theory (cf. Borsook,

1935)-

It is true, however, that under some conditions the addition of a

catalyst can greatly increase the concentration of one or more reaction
products and it is for this reason that enzymes are so important in cel-
lular reactions and catalysts so useful in industry. Assume the presence
of a substance A, which decomposes "spontaneously," i.e. with the liber-
ation of energy, to form a series of substances, B, C , D, E — and that
at equilibrium all substances are present in equal concentration. Since
they are all in equilibrium with A they are also in equilibrium with each
other. \i A \s, then allowed to decompose spontaneously the system at
the end of the reaction will contain equal concentrations of A, B, C, D,
and E, and no catalyst can cause any further change to occur. Now sup-
pose that A alone is present and a catalyst is added which accelerates the

closely to the experimental facts. He defined catalysts as substances which do not appear
in the reaction products in simple stoicheiometric relationship.

3 This conclusion is strictly true only when the catalyst is present in low concentra-
tion and does not combine with any of the components of the reaction. If the catalyst
is present in high concentration it may affect the activity of the components in the
same way as would any indifferent substance and so change the equilibrium (Taylor,
1925). If the catalyst combines with one or more components the equilibrium will also
be changed (Euler, 1925, p. 305). Enzymes in general are present in such low con-
centration that effects on the activity are negligible. Many enzymes combine with the
products of reaction, however, and in such cases this combination may affect the
equilibrium point, although the principal result is to slow down the rate of reaction
so that experimentally it may be very difficult to reach final equilibrium.

THE SYNTHESIS OF PROTEINS 5

reaction A^ B but not the other reactions. Addition of this catalyst
will result in a rapid formation of B, until [^] = [-S] with the result
that a large amount of B and very little C, D, or E is formed. The re-
action is not at equilibrium and if allowed to proceed by itself will end
at the same state as though the catalyst had not been added. In the mean-
time, however, large amounts of B have been formed due to the catalyst
and could be removed. Similarly, if the concentration of B is greater
than that of ^ at the beginning, addition of the same catalyst will now
cause A to be formed until [^] = [5] and very little C, D, or E will be
formed. The reaction does not take place in the direction B —> A simply
because of the addition of the catalyst but because the energy changes
require the reaction to go in this direction, and the catalyst merely ac-
celerates the rate. In this case, also, the high concentration of A is only
temporary, since if left to itself the same final equilibrium conditions,
A=B=C=D=E, will be reached.

The condition may be roughly exemplified by the arrangement shown
in Fig. I . The five tubes are connected at the top by open tubing and at
the bottom are all connected to tube A by means of tubes which can be
closed. If tube A is now filled with water to position i, the water will

! /

4.

BCD

Figure i. Enzyme reaction model.

distribute itself until it is at the same level in all tubes, since this is the
equilibrium condition. If the connecting tubes at the bottom are closed
at the beginning the water will distribute itself slowly through the vapor
phase. This corresponds to the "spontaneous" reaction rate. If all the
bottom tubes are opened the same final condition will be reached much
more rapidly. This corresponds to what will happen if a catalyst is added

JOHN H. NORTHROP

which accelerates all the reactions. If, however, only the tube connecting
A and B .s opened the level of B will rise to position . and no measu able
change w.ll occur m the levels of C, D. and E during this time

The preceding example resembles more or less the hydrolysis of pro-
ems hydrolysts with acid leads to complete hydrops and esu tfht

also lead only to ammo aads at equdibrium under the same conditions
Pepsm or trypsnt, however, increase the rate of hydrolysis of some
protem hnkages much more than others, so that for all praciical purpose
he reacfon stops before any amino acids are formed, although h
still far from equilibrium. "

Enzymes as Theoretical Catalysts
The agreement between enzyme reaction and catalytic theory has been
much more fully confirmed in connection with carbohydrates and esters
than with proteins.

The first experimental synthesis by an enzyme was reported by Croft
Hill, who found that a disaccharide which he thought was maltose was
formed by the action of maltase on concentrated glucose solutions The
substance however, was eventually identified by Emmerling as iso-
maltose which is formed, under the same conditions, by acid catalysis
Iso-maltose is not hydrolyzed by maltase and this fact led Armstrong
to suggest that enzymes synthesize substances which they do not hydro-
lyze. This assumption is contrary to the law of catalysis since it predicts
that the direction of the reaction, and, hence, the equilibrium condition
vary with the catalyst. It was found later that the ''maltase" prepara'
tion used by Croft Hill contained emulsin which does hydrolyze iso-
maltose and, hence, also synthesizes it. This is one of many examples of
the confusion caused by the use of impure enzyme preparations

Bourquelot studied the simpler case of glucoside hydrolysis and syn-
thesis and found in several cases that enzymes caused the reaction to
proceed towards the equilibrium point from either side. This furnished
experimental proof of the statement that enzymes do not change the
equilibrium point.

Kastle and Loevenhart showed that the hydrolysis of esters by lipase
also agrees with theory.

Pottevin found that pancreatic lipase synthesizes esters from oleic
acid and methyl alcohol and that the equilibrium condition was inde-
pendent of the concentration of enzyme. (For review of the older litera-
ture see Euler, 1925, p. 295 ; and BayHss, 1925.)

THE SYNTHESIS OF PROTEINS 7

Borsook and Schott have found (1931) from a study of oxidation
potentials that the enzyme, fumarase, which catalyzes the reaction, suc-
cinate=fumarate, also acts as a theoretical catalyst.

The synthesis of glycogen described by Cori and Cori (1939) is
another example, since in this case the reaction proceeds in either direc-
tion, depending upon conditions. The synthetic reaction is remarkable
in that a trace of the end product, glycogen, is required to start it. The
enzyme, phosphorylase, however, is also necessary. The reaction as a
whole therefore appears to be an example of autocatalysis superimposed
on an enzyme reaction.

The experimental results with esters and glucosides in the presence
of enzymes are therefore in good agreement with theory. In the case of
the proteolytic enzymes, however, the results are more complicated, and
there is some doubt as to the occurrence of protein synthesis in the pres-
ence of proteinases.

Various Possible Synthetic Reactions

The various methods which have been suggested to account for the
synthesis of proteins may be roughly classified as follows :

1. Purely catalytic synthesis, no energy added.

a. Equilibrium point shifted in favor of protein by change
of concentration or other condition (plastein formation).

b. Removal of protein as soon as formed. Insoluble proteins
— plastein? Synthesis of anilides (Bergmann and Fruton,
1941). Protein surface films (Langmuir and Schaefer, 1938).

2. Energy added so as to shift equilibrium in favor of protein.

a. Coupled reaction (cf. Borsook and Dubnofif, 1940, hip-
pur ic acid).

b. Synthesis from building stones other than amino acids.

Purely Catalytic Synthesis

This mechanism (Wasteneys and Borsook, 1930) is the simplest and
the most thoroughly investigated.

Theory. Proteins are hydrolyzed by enzymes and therefore they must
also be synthesized by the enzyme which caused hydrolysis, since
catalysts must accelerate the reaction equally in both directions. In order
actually to obtain proteins in this way, therefore, it is only necessary to
find conditions under which significant quantities of protein will exist
at equilibrium. Theoretically, more protein will exist in concentrated
than in dilute solution. This follows from the law of mass action which

8 JOHN H. NORTHROP

States that the equiHbrium constant for the reaction A=nB is defined
by the equation K=[Br/[A] or [A] = [B]'^/K. [A] and [B] are
concentrations at equiHbrium, and n is the number of molecules of B
formed from A. If [A] is very small compared to [B] then the concen-
tration of A at equilibrium will increase in proportion to the «th power
of the total concentration. Thus if it be assumed that a protein is hydro-
lyzed to lOO amino acids, increasing the total concentration lo times
should increase the concentration of unhydrolyzed protein 10^°° times.
The fact that there are undoubtedly a number of intermediate steps in
the hydrolysis does not change the results, since according to thermo-
dynamic reasoning the condition at equilibrium must be independent of
the path by which the equilibrium is reached.

Effect of pH. It is possible to assume that the equilibrium is shifted
by changes in pH. of the solution, since the energy change is small in
any case and could be significantly affected by the energy of ionization
of the products of the reaction.

Experimental Results. Danilewsky found that a precipitate
(plastein) was formed when a preparation of pepsin was added to a
concentrated solution of the products of pepsin hydrolysis of protein.
The problem was carefully investigated by Wasteneys and Borsook
(1930). They confirmed the importance of the concentration, although
the experimental result is not exactly as expected by theory since it is
necessary also to change the pH. In dilute acid solution the plastein is
digested, whereas the synthetic reaction had an optimum rate of pti 4.0.
There is a decrease in amino and carboxyl groups.

The total amount of plastein formed is greater when greater amounts
of enzyme are added. This is apparently contrary to theory, but it is
very difficult to be sure that equilibrium is reached, especially with low
concentrations of enzyme in the presence of reaction products. The
products act as inhibitors and greatly decrease the reaction rate.

Wasteneys and Borsook also found that the reaction was accelerated
by the presence of some emulsifying agents (benzaldehyde) and took
place to some extent without the enzyme. This observation is difficult
to understand from the point of view of synthesis unless these substances
help remove the protein from solution. Wasteneys and Borsook con-
sider that the protein is synthesized as a soluble protein which soon
changes to an insoluble (denatured?) form and precipitates from the
solution. If this is correct then the emulsifying agents may act by
accelerating the denaturation reaction since many proteins are rapidly
denatured at interfaces.

THE SYNTHESIS OF PROTEINS 9

Plastein is also formed by the action of trypsin or papain on a solu-
tion of insulin which has been previously digested by pepsin (Haddock
and Thomas, 1942).

There seems little doubt that some synthetic reactions occur during
plastein formation, but whether proteins are produced is still uncertain.
According to FoUey (1932) no substance having a molecular weight
of over 1000 is formed. This result has been confirmed by Ecker ( 1947) .

Taylor (1907 and 1909) reported the synthesis of protamine sulfate
from a completely digested solution of the compound. The enzyme
preparation was a glycerine extract of liver. The substance formed was
identical, or very similar, to the original protamine.

The relation of the plastein to the original protein may be tested by
hydrolyzing enzyme proteins. Plastein is formed in concentrated solu-
tions of trypsin or pepsin which have been hydrolyzed by pepsin. The
plastein, however, has no enzymatic activity, nor does it have the general
properties of the enzyme protein from which it is derived. In these cases,
therefore, the plastein obtained is not related to the protein originally
hydrolyzed (Northrop, 1947).

There is some experimental evidence that proteins may be formed by
enzymatic synthesis under the proper conditions.

Removal of Protein From Reaction Mixture. The preceding
section summarizes attempts which have been made to synthesize pro-
teins by changing the concentration of the reaction mixture so as to
favor the presence of the protein.

Protein could also be accumulated if the quantity in equilibrium with
the hydrolysis products could be constantly removed. This operation
would require energy but would result in the formation of indefinite
quantities of protein.

The synthesis of anilides discovered by Bergmann and Fruton ( 1944)
is probably an example of this type of reaction.

Robertson (1926) suggested that protein might be synthesized at
liquid interfaces since the protein is concentrated there. The experi-
ments of Langmuir (1939) and Gorter (1937) have shown that native
proteins, although soluble in bulk, form an extremely insoluble surface
layer at the air-water interface. If this layer were constantly removed
a new one would form and in this way protein would be obtained. Lang-
muir has suggested (Langmuir and Schaefer, 1938) that the molecules
already present on the surface could act to regulate the formation of
more identical molecules. This mechanism also predicts that the syn-

lO JOHN H. NORTHROP

thesis of proteins would be associated with some type of structure such as
is present in cells which would provide the necessary surface.

This hypothesis solves both the energy and the specificity problem
and appears to be the most reasonable and simplest mechanism so far
suggested. It deserves the most careful consideration. The accumulation
of protein in the surface layer is an undoubted fact, so that only the
specific synthesis is in doubt. Unfortunately no experimental verification
has been obtained as yet for this step. It must be remembered, however,
that negative results are very unconvincing in experiments of this type,
since even after positive results have been reported it is sometimes
necessary for the problem to be studied for years before conditions are
sufficiently well established so that the results may be obtained at will.

Synthesis by the Addition of Energy

Coupled Reactions. Many cases are known in which a reaction
which requires energy and hence does not occur alone will take place if
another reaction which liberates energy takes place simultaneously. Such
reactions control carbohydrate metabolism (Meyerhof, 1944; and Kal-
ckar, 1944). In these reactions part of the energy liberated by hydrolysis
or oxidation is used for a synthetic reaction so that, although the total
change results in the liberation of energy, some steps occur which require
energy. Each step is controlled by a special enzyme. Phosphoric acid
esters are concerned in these energy exchanges and it is reasonable to
suppose that similar esters may take part in the synthesis of proteins, as
Bergmann and Fruton (1944) have suggested. Many other possible
sources of energy exist, but no direct experimental evidence has been
found to connect the synthetic and energy-providing reaction. The syn-
thesis of hippuric acid from glycine and benzoic acid in the presence of
liver slices is perhaps the nearest approach to such a reaction. Borsook
and Dubnoff (1940) found that rapid synthesis occurred in the presence
of intact liver cells. Extracts or even minced tissue failed to cause the
reaction.'' It seemed evident that the energy was obtained from respira-
tion and hence that the reaction would stop if the respiration were
stopped, and hydrolysis should then occur instead of synthesis. Addition
of HCN, which poisons the respiratory enzymes, did stop synthesis, but
it also stopped hydrolysis. This disconcerting result may be due to
poisoning of the enzyme which catalyzes the benzoic acid reaction, al-

4 It is perhaps significant that the three fundamental reactions, protein synthesis,
photosynthesis, and fixation of nitrogen all require energy and all are so far inseparable
from intact cells.

THE SYNTHESIS OF PROTEINS II

though such an effect would not be expected since only the respiratory
enzymes are known to be inactivated by HCN. The result might be due
to the fact that hydrolysis and synthesis are both caused by the same
enzyme.

Schoenheimer's beautiful experiments (1942) have shown that pro-
teins are continuously being synthesized and hydrolyzed so that it is
quite possible that part of the energy for synthesis of proteins is ob-
tained from the hydrolysis of other proteins or peptides.

The enzymes which hydrolyze proteins (pepsin, trypsin, papain,
cathepsin, etc.) do not carry digestion very far. The hydrolysis is com-
pleted by the various peptidases. Part of the energy liberated by the
hydrolysis of the peptides could, therefore, be returned to the protein
^peptide system and so result in the synthesis of proteins. Such a
system is highly organized and hence would be difficult to isolate in good
working order. It could not function indefinitely without an external
source of energy.

An interesting reaction which may be an example of this mechanism
has been studied by Behrens and Bergmann (1939). They found that
neither acetyl-J/-phenylalanyl glycine nor glycyl-/-leucine alone are
hydrolyzed by papain, but that glycine and leucine are formed when
both peptides are present. Similar results have been reported by Abder-
halden and Ehrenwall (1933). Behrens and Bergmann were also able
to show that an intermediate compound, acetyl phenyl-alanyl glycyl
glycyl leucine, is formed and then hydrolyzed. They consider that the
tripeptide acts as "co-substrate" for the hydrolysis of glycyl leucine
and that the energy required for the formation of the intermediate com-
pound is obtained from the hydrolysis of glycyl leucine. The tripeptide
appears to be a "co-enzyme," however, since its concentration is un-
changed at the end of the reaction, rather than a "co-substrate," which
is a substance changed by the reaction.

The reaction is an excellent example of a catalytic reaction in which
the acceleration is due to the formation of an intermediate compound.
The overall change is glycyl leucine ^ leucine and glycine and hence
the formation of the pentapeptide does not enter into the energy relation
at all, since thermodynamic equilibria are entirely independent of the
path by which they are reached. As Petrie (1943) has pointed out, the
reaction is actually an hydrolysis and does not by itself serve as a source
of the synthetic substance. Nevertheless, the reaction is of interest since
it shows that the enzyme does catalyze the synthetic reaction. If a mech-

12 JOHN H. NORTHROP

anism were present which would remove the pentapeptide, the reaction
could be used as a source of the substance. In this case the energy for
synthesis would be furnished by the mechanism which removed the
peptide.

Synthesis of Proteins from Compounds other than Amino
Acids. Several reactions are known (cf. Bergmann and Fruton, 1944)
which could give rise to proteins. Alcock (1936) notes that plants can
form proteins from sources other than amino acids. He claims with
some justification that two entirely different mechanisms for protein
synthesis would not be expected and suggests that animals first change
the amino acids to the hypothetical building stones which are common
to both plants and animals. An "ur-protein" molecule is then synthesized
and the other proteins derived from this "ur-protein." This hypothesis
is logically attractive but must be amplified somewhat to take into ac-
count the energy changes required (cf. page 10).

If it could be shown that the compounds taking part in these reactions
are formed from amino acids by the addition of energy and that these
compounds then reacted to form proteins with the liberation of energy,
the mechanics of synthesis would be greatly clarified. Synthesis of
proteins from such high energy compounds has been discussed by Del-
briick (1941) and Gulick (1944). Such a reaction would account for
the fact that N or C isotopes supplied in any amino acids are very soon
found in all amino acids (except lysin) (Schoenheimer, 1942). This is
difficult to understand if the proteins are synthesized from amino acids,
but would be expected if amino acids are first changed to other inter-
mediates before synthesis occurs.

At present, however, there does not appear to be any experimental
evidence for the existence of such compounds, or reactions, or the neces-
sary catalysts in biological material. The enzymatic synthesis of glu-
tamic acid from NH3 and the keto acid (Euler, Adler, Giinther, and
Das, 1938) is probably the nearest approach to such a reaction so far
discovered. Glutamic acid can be converted into several other amino
acids by transamination (Braunstein and Kritzmann, 1937; Braunstein,
1939). The transmigration of the methyl groups in certain amino acids,
studied by DuVigneaud and others (reviewed in Borsook and Dubnoff,
1943) furnishes another partial mechanism for equilibria between
amino acids. These reactions can account in part for transformation of
one amino acid into another but do not supply a complete synthetic
reaction for proteins.

THE SYNTHESIS OF PROTEINS I3

Formation of Native as Distinct from Denatured Proteins

It was stated earlier in this discussion that the reaction, protein and
water ^ amino acids is nearly complete and very little protein remains
at equilibrium. This is experimentally true in the case of denatured
proteins but merely an assumption as far as native proteins are con-
cerned. Most native proteins are digested very slowly by proteases and it
is possible to assume, as Linderstrom-Lang and Jacobsen (1941) have
suggested, that the first step in the hydrolysis is the formation of dena-
tured protein from the native protein. Linderstrom-Lang attempted to
answer the question by a study of the temperature coefficient of hy-
drolysis of native and denatured protein, but the results were inconclu-
sive. The hydrolysis of a protein like trypsin or chymotrypsin, which
exists in an equilibrium between native and denatured forms (Northrop,
1939), should furnish evidence in this connection.

Haurowitz and collaborators (1945) have reported that globular
proteins, such as egg albumin and serum globulin, are not hydrolyzed
by trypsin unless they are denatured, whereas "fibrous" proteins, such
as fibrin and myosin, are hydrolyzed at about the same rate in either
native or denatured forms.

If enzymatic hydrolysis does require preliminary denaturation, then
it could be assumed that native proteins can be synthesized without
adding energy, i.e. by a purely catalytic reaction. Energy is required to
denature the protein which then hydrolyzes with the liberation of energy.
The entire cycle would require three reactions :

enzyme

1. Amino acids ^ native protein + energy.

2. Native protein + energy ^ denatured protein.

enzyme

3. Denatured protein ^ amino acids + energy.

It is also possible to assume that denatured protein is formed first
from the amino acids and that the native proteins are then formed from
this denatured protein. This cycle would be as follows :

Amino acids + energy ^ denatured protein
Denatured ^ native protein + energy

The direction in which the reactions proceed is determined by the
energy changes and the rate of the reaction by the enzymes present.

14 JOHN H. NORTHROP

Specificity of Synthesis

The preceding outline shows that several mechanisms are known
which could result in the synthesis of proteins without coming in con-
flict with any of the accepted theories of chemical reactions. Even if
some of these overall mechanisms are correct, there still remains the
problem of the regulation of the reaction so that the desired protein only
is obtained and not simply a random assortment of proteins.

Nothing is to be gained, in the absence of experimental evidence, by
assuming a series of enzymes, each controlling the synthesis of a single
protein, since we are then faced with the mechanism of synthesis of the
enzymes themselves. These are presumably also proteins and we are
therefore back at the beginning again.

The known proteases hydrolyze practically all (denatured?) proteins
and hence must catalyze the synthesis of all proteins, so that these
enzymes can hardly account for the formation of any one protein. For
instance, horse hemoglobin and pig hemoglobin are both presumably
hydrolyzed completely to amino acids by the autolytic enzymes of either
the horse or pig. If, therefore, conditions were found in which the re-
action were reversed, addition of either of these enzymes would result
in the formation of at least two hemoglobins, horse and pig. In the
animal, however, this does not occur. Horse hemoglobin only is formed
in the horse and pig hemoglobin only is formed in the pig.

It is possible that the specificity of protein synthesis is not due to the
specific action of the enzyme but simply to the chemical nature of the
protein itself. Most chemical reactions are specific in that, as a rule, very
few and often only one of a large number of possible reaction products
actually appear. Thus if the reaction between NaOH and HCl is treated
purely statistically, there are a great many possible reaction products.
Actually, however, only two products are formed — NaCl and H2O. This
"specificity" is not due to any catalyst but simply to the fact that the
products, NaCl and HoO, are the only stable ones under the condition
of the reaction. It is therefore possible that the formation or synthesis
of specific proteins is regulated by the difference in stability of the many
structurally different proteins rather than by specific catalysts.

Autocatalysis

Autocatalysis is the only known chemical reaction by which indefinite
quantities of a required substance may be obtained without assuming
the existence of several other substances, the origin of which must also

THE SYNTHESIS OF PROTEINS I5

be accounted for. Autocatalysis, therefore, offers the only escape from
the dilemma of the preformationists and has been invoked by a number
of workers. Troland (1917) developed a general theory of cell forma-
tion on this basis. He concluded that complicated molecules reproduced
themselves by a process similar to crystallization. The molecular struc-
ture of the underlying layer determined the structure of the superim-
posed layer. Troland also reviewed earlier and less definite suggestions
of autocatalysis. Koltzoff (1928) developed a similar theory. He con-
cluded that antibody formation was a special example of such a reaction.
Koltzoff also realized the fact that an autocatalytic reaction cannot
start itself, but must have at least one molecule of the reaction product
present at the beginning (cf. Northrop, 1937). He assumed the "gen-
onema" contains at least one molecule of each protein of the species. It
is difficult to avoid the necessity for this assumption but it may be that,
if the total number of different kinds of protein molecule present in the
individual were known, genonema and even the sperm and ovum would
be too small to contain them. Gulick (1938) suggests that protein mole-
cules are biscuit-shaped discs and that this structure is impressed on
each succeeding molecule somewhat as a coin is stamped by a die. He
points out, however, that the contact must be "front to back" as other-
wise the impression of the disc and not the replica would be obtained.
Langmuir and Schaefer (1938) have also suggested the "template"
method of regulating synthesis. Pauling's (1940) theory of antibody
formation is somewhat similar. Darlington (1944) ascribes the forma-
tion of "plasma germ" and viruses to a similar autocatalytic reaction.

Stanley (1938a) suggests that a protein-nucleic acid complex is the
simplest structure capable of autocatalytic synthesis. He ascribes the
specific synthesis to a crystallization-like reaction similar to that sug-
gested by Troland.

A Working Hypothesis for the Synthesis of Proteins

The preceding section has summarized briefly theories and experi-
mental results relating to protein synthesis. These theories account more
or less satisfactorily for the specificity of the reaction but (with the
exception of the system of Langmuir and Schaefer) fail to provide for
the necessary energy. Since there is every reason to believe that energy
is required, the synthesis cannot be accounted for by a simple auto-
catalytic reaction, nor can it be said that any of the various types of
proteins or viruses "synthesize themselves," unless it be further assumed
that they are formed from unknown, high energy compounds. They

l6 JOHN H. NORTHROP

may be able to direct the synthesis so that the proper protein is formed,
but the energy must come from some other reaction, so that at least two
reactions, and not a single autocatalytic reaction, are necessary.

The working hypothesis outlined in the rest of this discussion is an
attempt to formulate the simplest assumption that is adequate to account
for the known facts. It is quite probable that future experiments will
force modification to be made and may render the entire mechanism un-
tenable. If the hypothesis leads to the discovery of new facts it will
have served its purpose even though the new facts destroy the assump-
tions which led to their discovery.

In general the hypothesis may be stated as follows :

The formation of proteins occurs in two steps. The first step consists
in the synthesis of one or more "type" proteins ("proteinogens") ("ur-
protein" of Alcock) which are specific for the species and perhaps for
the organ. This step requires energy which may be obtained from a
coupled reaction or from preliminary formation of high energy building
stones (not amino acids). It is "autocatalytic" in that the structure of
this proteinogen is determined by itself. No experimental evidence for
this step exists at present. It probably takes place in the cell, since respira-
tion and similar reactions that liberate energy occur usually in the cell.

Since even the existence of such a protein is purely hypothetical, dis-
cussion of its properties is hardly warranted. The only essential proper-
ties are that stable proteins may be formed from it without the addition
of energy and that its own formation is autocatalytic. There may be
several such proteinogen structures, one for each general group of
proteins. They probably contain nucleic acids. The virus proteins may
be special examples of these proteinogens.

Spiegelman and Kamen (1946) have suggested that nucleoproteins
are the controlling factor in protein synthesis, and present evidence
indicates that the energy may be supplied by the phosphoric acid of the
nucleoprotein. Muller (1945) has made a similar suggestion.*

The possibility that the proteinogen has the characteristics of a de-
natured, rather than a native protein, was suggested by M. Kunitz.
This assumption would account for the observed immunological reac-
tions as well as the energy requirements (cf . pp. 10, 27) , since denatured
proteins do not cross-react with the native proteins, nor do they show

♦Recent experiments by Reiner and Spiegelman (1948) and by Price (1948) bear
out this hypothesis and also indicate a relationship between virus and adaptive enzyme
formation. Spiegelman and Reiner found that a desoxyribonucleic acid fraction from
yeast cells stimulates the formation of adaptive enzyme in yeast. Price finds that the
same fraction greatly increases the yield of virus particles per cell, formed by S. muscac.

THE SYNTHESIS OF PROTEINS I7

such marked specificity. Therefore cross-reactions between the pro-
teinogen and the native proteins derived from it would not be expected.

It is probably unnecessary to assume that the molecular weight of the
proteinogen is equal to or greater than that of the largest molecules
which are to be formed from it since Svedberg (1937) and others have
found that very large protein molecules like the hemocyanins dissociate
very easily and reversibly. The formation of these very large molecules
from molecules the size of normal proteins, therefore, requires little or
no energy (Northrop, 1938, p. 363).

It is also unnecessary to assume that any appreciable quantity of this
protein exists at any one time. It could be decomposed as rapidly as
formed and still act as an intermediate between the energy-requiring
step and the specific step in protein synthesis.

In the second step the individual proteins are formed by a catalytic
or an autocatalytic reaction from this proteinogen. This reaction does
not require energy and may occur anywhere.

The formation of an enzyme from its precursor is an example of
such a reaction. It is the only mechanism so far discovered whereby
proteins may actually be produced in vitro. The reactions are specific
and are adequate to account for the formation of proteins in general,
provided the precursor is present. The exact chemical changes involved
in these reactions are not known.

The proposed mechanism accounts for the group specificity of pro-
teins from the same organ and species. Assumption of one (or a few)
synthetic reactions is simpler than assumption of a separate synthetic
reaction for each individual protein. It has the further advantage that
it is in keeping with the gradual development of the cell, since, once the
proteinogen is formed, other proteins can readily be derived from it
(cf. Troland, 191 7). If each protein must be synthesized individually,
each one requires an energy source and the evolution of a series of re-
lated proteins is more difficult.

Madden and Whipple (1940) have pointed out that partial synthesis
of all proteins in one organ (the liver?) would greatly simplify the
mechanism of tissue metabolism since it would then be unnecessary for
every cell of every tissue to contain the complete protein-synthesizing
system. According to the present mechanism, the proteinogen could well
be synthesized in one organ and distributed from there to the various
other organs and tissues where it would be modified into the specific
proteins needed.

It may be noted in this connection that many autocatalytic and many

l8 JOHN H. NORTHROP

coupled reactions are known, but no autocatalytic coupled reactions have
been described, so far as the writer is aware. If every cell in every tissue
had such coupled reactions occurring for each protein, the chances of
the discovery of the reaction would be far greater than if it occurred
only in a special organ or organs.

In the following sections an attempt is made to describe the formation
of normal proteins, viruses, adaptive enzymes, and antibodies in terms
of this working hypothesis.

II. Formation of Normal Proteins

It is becoming increasingly evident that some sort of equilibrium
exists between the various amino acids and the various proteins or both
(Madden and Whipple, 1940; Schoenheimer, 1942). Schoenheimer
demonstrated by means of isotopes that both the N and C atoms of an
ingested amino acid appear in nearly all the proteins and amino acids of
the body in the course of a few days. This result is totally unexpected
from the viewpoint of the classical theory of protein synthesis and
indicates the existence of a dynamic equilibrium (cf. Borsook and
Keighley, 1935), since such equilibria would result in a rapid distribu-
tion of the atoms. It is probable that, as Schoenheimer has suggested,
some unknown building stones take part in the equilibria, as otherwise
it is difficult to account for the transfer of the C atoms from one amino
acid to another.

Whipple and Madden's results are extremely interesting in this
connection. These workers find that dogs may be kept in health when
the only source of nitrogen is plasma injected into the vein. The animal
must, therefore, be able to form any required protein from the protein
present in plasma without preliminary hydrolysis by the enzymes of the
digestive tract. Whipple and Madden consider the results to show that
an equilibrium exists between plasma protein and the cell proteins. They
suggest that large polypeptides may be the intermediate compounds in
the equilibrium.

A similar equilibrium between the serum proteins themselves has
often been suggested, and there is considerable evidence to show that
these proteins can be transformed into each other.

Pederson has found by ultracentrifugal studies that a lipoid-contain-
ing protein of enormous molecular weight, which he calls X-protein,
exists in serum. He assumes that this protein is in equilibrium with the
albumin and globulin fractions. It is true that albumin and globulin

THE SYNTHESIS OF PROTEINS I9

are not in equilibrium with each other when isolated from serum, but this
fact by no means proves that they are not in equilibrium with each other
in the circulating blood.

Moore, Shen, and Alexander (1945) have found a special protein in
chicken embryos which is not found in adult chicken serum.

The various equilibria indicated by these experimental results may be
described in terms of the present hypothesis as follows :

Amino Acids

If (Energy added somewhere between amino acids and proteinogen.)
(Intermediate?)

\\ (Autocatalytic reaction.)
Proteinogen
± Enzyme Jf \\ (Autocatalytic or catalytic reaction.)
Protein A, Protein B, Protein C, etc.

Since all the proteins are in equilibrium with the proteinogen they
are in equilibrium with each other.

The quantity of proteinogen present at any time may be very small
and hence difficult to detect.

The proteinogen is probably synthesized only in cells. The formation
of the usual proteins from proteinogen may occur anywhere.

No experimental evidence for the formation of normal proteins in
this way is at hand. It is possible that such a reaction would be observed
if an organism could be completely freed from one of its normal pro-
teins. The re-addition of this protein to the organism should then result
in renewed production.

The formation of enzymes from their precursors is exactly analogous
to the hypothetical reaction assumed for the formation of all normal
proteins, and hence one of interest in this connection. The precursors of
the enzymes must themselves be formed from the proteinogen mole-
cules, and an indication of such a reaction has been noted by Allen, Ray,
and Bodine (1938).

Formation of Enzymes from Their Precursors

Trypsin from Trypsinogen. This reaction agrees quantitatively
with the theory for a simple autocatalytic reaction. The velocity is af-
fected by acidity and temperature in the same way as is the formation of
chymo-trypsin from chymotrypsinogen (Kunitz and Northrop, 1935)
or the hydrolysis of other proteins by trypsin. At pYi 5.0 and 8°C. the

20 JOHN H. NORTHROP

value of the constant is 14.6, i.e. with unit concentration of trypsinogen
the active trypsin would increase 14 times per hour. As in the case of
chymo-trypsin there is no evidence for the sphtting off of any part of
the molecule during this activation reaction, nor is there any marked
change in chemical composition so that the change from the inactive
protein to the enzyme is probably due to some internal rearrangement
in the structure of the molecule. However, the work on this question is
still in the preliminary stage. The autocatalytic nature of this reaction
was correctly described by Vernon but subsequently denied by Bayliss,
Starling, and others.

Kunitz (1938) has recently found that trypsin is also formed from
trypsinogen at pH 5.0 by a proteolytic enzyme secreted by a mold
(penicillium). Trypsin is only slightly active at this pH, so that the
activation curve is no longer autocatalytic but simply logarithmic as is
the activation of chymo-trypsinogen. Trypsin obtained in this way is
identical with that formed by autocatalytic activation. The mold enzyme,
therefore, must attack the trypsinogen molecule at the same place as does
trypsin.

The formation of trypsin from trypsinogen is also accelerated by
enterokinase. The reaction trypsinogen -^trypsin is therefore cata-
lyzed by three different catalysts — trypsin itself, mold kinase, and
enterokinase.

Activation of Trypsin by Enterokinase. Pancreatic juice as
secreted is inactive and becomes active in vivo only when mixed with
the contents of the small intestine. The subject has been controversial
for many years. Kunitz (1939b) has reinvestigated the problem with
purified materials and has found that the conflicting results are due to
the fact that an inert protein, as well as the active enzyme, are formed
simultaneously during the reaction. The relative amounts of the two
substances obtained depend on the acidity of the solution. In slightly
acid solution nearly all the precursor is transformed into the active
enzyme and very little inert protein is formed. In neutral or slightly
alkaline solution considerable amounts of inert protein are formed.

The entire reaction may be quantitatively predicted on the assump-
tion that the rate of formation of the active enzyme is proportional to
the concentration of precursor and of activator, while the rate of forma-
tion of the inert protein is proportional to the concentration of trypsin
and of the precursor. The relative rates of the two reactions are deter-
mined by the acidity.

Pepsin from Pepsinogen (Herriott, 1938). The reaction at /'H4.65

THE SYNTHESIS OF PROTEINS 21

is autocatalytic and hence is caused by pepsin itself. About 15 per cent
of the nitrogen is spHt off during this reaction. So far as is known,
pepsin attacks only peptide linkages, so there is reason to believe that
the rupture of one or more peptide links in the precursor leads to the
formation of the active enzyme. If swine pepsinogen is activated by
chicken pepsin, swine pepsin is formed (Herriott, Bartz, and Northrop,
1938) . The structure responsible for the species specificity of the enzyme
is therefore present in the precursor.

Relation of Protein to Precursor. The exact changes in chemical
structure which occur during these reactions is not yet known. It is
known, however, that the protein formed may be quite distinct in chemi-
cal, physical, and immunological properties from its precursor.

It is probable that all amino acids present in the enzyme must also be
present in the precursor although, in view of the remarkable reactions
discovered by Schoenheimer, even this conclusion may be questioned.

Immunological Relationship of Chymo-trypsin, Chymo-tryp-
siNOGEN, and Beef and Pig Trypsin (Ten Broeck, 1934). The tests
were carried out by the Dale technique for anaphylactic reactions.
Chymo-trypsinogen was found to be distinct from chymo-trypsin.
Guinea pigs sensitized with chymo-trypsinogen were negative in most
cases w^hen tested with chymo-trypsin, although four animals out of
twelve gave cross-reactions. No cross-reactions were observed when the
pigs were sensitized against chymo-trypsin and tested with chymo-tryp-
sinogen. There is a slow formation of chymo-trypsin from chymo-
trypsinogen even in the absence of trypsin (Kunitz and Northrop, 1935)
and it is quite possible that enough chymo-trypsin was formed in this
way during the experiment to sensitize some animals, and so produce
the cross-reaction.

Trypsin from beef pancreas was distinct from trypsin from pig
pancreas and both were distinct from chymo-trypsin and chymo-tryp-
sinogen.

Pepsinogen and Pepsin (Seastone and Herriott, 1937). These
tests were carried out by the precipitin reaction. They are complicated
by the fact that pepsin is inactivated rapidly at the pH of the blood so
that the antibody formed is presumably caused by the denatured rather
than the native protein. It is also possible that traces of pepsin are formed
from pepsinogen after injection into the blood stream.

Pepsinogen antisera reacted with pepsinogen but did not react with
pepsin nor with serum protein from the homologous species.

Antipepsin sera reacted with pepsin and also to a small extent with

22 JOHN H. NORTHROP

pepsinogen. They do not react with serum proteins of the homologous
species.

These resuhs show that there is at best only a faint cross-reaction
between the enzyme and its precursor, even when the experiments were
carried out with concentrated solutions of the two pure proteins. Had the
tests been carried out with crude tissue extracts of the precursors it is
probable that no cross-reaction between the enzyme and its precursor
would have been observed, owing to the small quantities present.

Specificity of the Reactions. Each enzyme, as a rule, has its own
precursor, but Bodine (1945) has reported that different tyrosinases
may be formed from protyrosinase by different methods. The enzyme
precursors isolated so far, therefore, do not represent the hypothetical
proteinogen, but must be considered a separate step between this and
the active enzyme. The formation of the precursor of tyrosinase has
been studied by Allen, Ray, and Bodine (1938) and found to be an
autocatalytic reaction. The enzymes whose precursors have been isolated
are secretory enzymes and are required in high concentration at special
times. It is possible that relatively large quantities of the precursors of
these enzymes are to be found in the organs for this reason, since they
serve as a convenient source of supply from which the active enzymes
can be obtained rapidly at any time. The cellular enzymes and the normal
proteins, on the other hand, are not required suddenly and there is no
necessity for a rapidly available reserve. This difference in the require-
ments may account for the fact that the precursors for the secretory
enzymes only have been found in quantity.

III. Formation of Viruses

There is good reason to believe that some viruses, at any rate, are pro-
teins (for a review of this work see Stanley, 1940).

No experimental evidence exists, so far as the writer is aware, to dis-
tinguish the formation of virus from the formation of normal proteins.
Both are found only in or on living cells and virus production is accom-
panied usually, if not always, by protein production. Experimentally the
only difference between the two processes is that "normal" proteins are
always produced by the cells but that the production of some virus pro-
tein may be started by introduction of the protein from outside the
organism (Northrop, 1937; Darlington, 1944). Phage produced by
lysogenic strains of bacteria and "indigenous virus" of plants and
animals are always produced just as are "normal" proteins. It can, of

THE SYNTHESIS OF PROTEINS 23

course, be assumed (cf. Gratia, 1938) that infection occurred sometime
in the past, but there is no direct evidence for this assumption. No means
is known to free lysogenic strains from their viruses.

It would be expected, then, that if an organism could be completely-
freed from a protein, it would no longer produce that protein until
reinoculated with it. The suggestion was made (Northrop, 1938) that
the substance producing type-specificity in pneumococci was an example
of such a reaction. The analogy between virus and the substance regu-
lating type-specificity in pneumococci was recognized by Gratia (1936).
Since then the substance has actually been isolated (Avery, MacLeod,
and McCarty, 1944). It turns out that it is a desoxyribonucleic acid
instead of a protein, but the mechanism of formation may well be the
same.

Virus proteins possess the following characteristics which must be
accounted for by any hypothesis for their formation :

1. They increase only in the presence of appropriate living cells.

2. Production of virus is usually associated with growth and
active metabolism of the host cells (Bordet and Jaumain, 1921 ;
Krueger and Northrop, 1930; Zinsser, 1937; Gratia, 1938;
Howe and Mellors, 1945, this paper contains other relevant
references).

This correlation between metabolism and virus production is strong
evidence that the synthetic part of the reaction is carried out by the cell
in conjunction with the synthesis of normal proteins, since the viruses
themselves do not contain the enzymes necessary to carry out these re-
actions.

If it should be found that virus production can take place only during
cell division, the analogy between virus and gene suggested by many
workers (cf. Beadle, 1945) would be greatly strengthened, since genes
also multiply only during cell division, whereas structural proteins are
formed continuously.

All hypotheses concerning the production of virus proteins (other
than those which consider the virus to be a cell) assume some sort of
autocatalytic reaction. Some writers (Troland, 191 7; Stanley, 1938b;
Caspersson, 1939; Jordan, 1944) assume that the virus can synthesize
itself from simple building stones. This mechanism requires energy and
entails a series of assumptions as to the source of energy and the manner
in which it is made available (cf. page 10).

Bordet (1931) pointed out that a simpler and equally adequate hy-

24 JOHN H, NORTHROP

pothesis is that the virus is formed from a precursor. This assumption
does not involve any energy mechanism in addition to that already pres-
ent in the host cells and has the further advantage that there is a close
experimental analogy in the formation of enzymes (Gratia, 1922 ; Nor-
throp, 1938). Darlington (1944) has recently suggested a similar
mechanism. Potter and Albaum (1943) consider viruses to be "mis-
placed enzymes."

Viruses have also been considered as obligate parasites. This point of
view is ably presented by Rivers (1939) and by Burnet (1945), who
consider that a virus is a fragment of a degenerate bacterial cell. This
fragment has maintained or developed the power of multiplying in the
presence of the host cell. Actually the original source of the virus is the
only significant difference between these points of view and that of
Bordet, who considered that the virus was originally part of a host cell,
rather than of a parasitic bacterial cell. It is quite possible that both
ideas are correct since the definition of virus is at present largely (and
quite arbitrarily) based on size. Vaccinia, for instance, is similar to bac-
teria in many ways, and would not be considered a protein even though
its peculiar physiological and pathological properties were not known.

On the other hand the well-defined crystalline viruses, tobacco mosaic
and especially bushy stunt of tomatoes, are typical proteins in many re-
spects and would never be considered as related to parasites were it not
for the pathological changes which they cause.

Final decision between the possibilities probably cannot be made
until the reaction can be produced in vitro and in the absence of host cells
(cf. Seifriz, 1939).

3. Viruses are formed as a rule only in the presence of cells from
related species, but may sometimes increase in the presence of
cells of widely separated species.

4. Viruses are immunologically distinct from the host proteins.

5. The virus protein produced in the presence of appropriate
living cells is usually identical with that used for inoculation
but may be different.

6. Inoculation with two viruses results in production of only one,
as a rule. In the case of bacteria (Delbriick and Luria) the virus
which is inoculated first is produced and not the second.

7. No evidence of metabolism has been found.

8. Viruses are of high molecular weight and contain nucleic acid.
Neither of these characteristics, however, suffices to distinguish
them from normal proteins.

THE SYNTHESIS OF PROTEINS 2$

9. Some viruses at least (bacteriophage, Northrop, 1939; tobacco
mosaic, Loring, 1940) can form saturated solutions. This result
shows that the virus can exist in solution or, more precisely, that
the liquid containing the virus is a single phase.

Such solubility experiments afford a decisive test of the question
whether the virus molecules are in "solution" or not. From the kinetic
point of view any particle is a molecule and there is no distinction
experimentally or theoretically between particles and molecules (cf.
Taylor, 1925, p. 1279). For this reason the word molecule seems appro-
priate until evidence is found contradicting its use.

From the point of view of thermodynamic equilibria as stated in
Gibb's phase rule, however, there is a sharp and definite distinction be-
tween suspensions and solutions. A suspension is a two-phase system and
a solution is a single-phase system. It is perfectly possible to decide,
therefore, whether a virus "solution" consists of one or two phases.^

Quantitative measurements have not been made on other viruses, but
any virus which crystallizes must exist as a saturated solution, in equi-
librium with the crystals.®

Such solubility experiments have not been carried out with larger
viruses, such as vaccinia, but it appears very unlikely that any indication
of a saturated solution would be obtained in these cases.

5 There is some uncertainty as to the application of the phase rule to colloidal solu-
tions, but the results obtained with proteins (Northrop, 1939) show clearly that the
rule may be applied to these compounds. The following experimental results are possible
in a study of the solubility of viruses :

/. The concentration of virus in the liquid is zero or variable and no equilibrium can
be shown to exist between the solid and liquid. Such systems are suspensions, but no
further conclusion can be drawn since the phase rule cannot be applied to systems
which are not in equilibrium.

2. The concentration of virus in the liquid is constant and in equilibrium with the
solid virus.

a. The concentration of virus in the solution is independent of the quantity of solid
virus present. The system consists of two phases and two components, and the virus
therefore is a single component. The virus is in solution in the liquid phase. No such
results have been reported.

b. The concentration of virus in solution varies with the amount of solid. There are
two phases and three or more components. The virus is in solution in the liquid phase.
The virus preparation consists of two or more components.

This is the result obtained with bacteriophage. (Northrop, 1938) and tobacco mosaic
preparations (Loring, 1940), although the bacteriophage curve is as good as that ob-
tained with many normal proteins.

^ The fact that the solutions may be concentrated in the centrifuge does not affect this
conclusion, since this simply amounts to increasing the force of gravity and any solu-
tion may theoretically be concentrated in this way. The gravitational field simply
represents another degree of freedom.

26 JOHN H. NORTHROP

The properties of viruses outlined above may be accounted for in
terms of the present hypothesis, as follows :

Specificity of Virus and Host. Viruses in general increase only in
related host species. It is known from immunological studies that the
proteins of related species are similar and some may be identical.

It would be expected, therefore, that the virus precursor (protein-
ogen) would be similar or identical in related species but different in
widely separated species and hence that a virus would usually be pro-
duced only in related species. This is usually the case.

The anatomical and physiological differences between species in-
creases as development proceeds, and embryos of different species are
much more alike than adults. It is quite possible that this similarity of
the embryos extends to the structure of the proteins and that embryo
proteins are more alike than adult proteins (cf. Pedersen, 1945). This
may account for the fact discovered by Murphy (1912) that heterol-
ogous transplants will grow in chick embryos. In general such trans-
plants do not grow when made in adults.

The fact that chemical differentiation of the proteins proceeds more
or less parallel to the morphological differentiation of the organs is
clearly shown by the results of Burke et al. (1944). These workers
followed the development of organ antigens in growing embryos and
found that organ specificity was closely related to the growth of the
organs themselves.

Woodruff and Goodpasture (1931) found that many viruses would
multiply in chick embryos, although they would not do so in the adult
chicken. These observations suggest that the proteinogens of embryos
are more closely related than are those of the adult animal and hence
are able to serve as precursors for heterologous proteins while the pro-
teinogens of the adult animal cannot.

Moore, Shen, and Alexander (1945) have found that plasma from
chicken embryos contains principally one high molecular weight pro-
tein which is absent, or present in low concentration, in adult plasma.
Pedersen (1945) has found a characteristic protein (fetuin) to be
present in foetal calf plasma, and he has suggested that this protein
may be a precursor of the normal serum proteins.

Burnet, Freeman, Jackson, and Lush (1941) and Dale have objected
to the assumption of a precursor on the grounds that the specific charac-
ter of the virus would then be determined by the host and not by the
virus, as is usually considered to be the case.

The character of a virus which has been modified chemically is deter-

THE SYNTHESIS OF PROTEINS 2/

mined by the host. Thus tobacco mosaic virus with oxidized SH groups
(Anson and Stanley, 1941) or the acetyl or phenyluriedo derivative of
tobacco mosaic (Miller and Stanley, 1941) all give rise to the original
tobacco mosaic protein and not to the modified form which was used
for inoculation.

In these cases the type of virus formed is determined by the host.
The changes in yellow fever virus after passage through mouse brain
(Lloyd, Theiler, and Ricci, 1936) and of equine encephalitis virus
grown in pigeons (Traub and Ten Broeck, 1935) may be other examples
in which the nature of the virus is determined by the host. This subject
has been reviewed in detail by Stanley (1943). Stanley considers these
changes in the virus to be mutations.

On the other hand, proof of the identity of the virus produced by two
different hosts rests in most cases only on immunological and pathologi-
cal properties. The same tests fail to differentiate between pepsin formed
from cattle or swine pepsinogen (Seastone and Herriott, 1937). The
differences between the two enzymes were established only by a mixed
solubility determination (Northrop, 1939, p. 32). Such tests have not
been made with virus proteins. In any case there is nothing to indicate
that the same precursor may not exist in many different species. For
example the crystalline lens proteins of many species are indistinguish-
able by immunological reactions and may be identical in various species.

The assumption of a precursor has also been criticized on the grounds
that its presence would be detected by immunological tests. This objec-
tion sounds very reasonable on a priori grounds, but as a matter of fact
only a questionable cross-reaction could be found between pepsin and
its precursor, pepsinogen^ (Seastone and Herriott, 1937), or chymo-
trypsin and its precursor, even with the classical Dale technique (Ten
Broeck, 1934; cf. page 16). These experiments were carried out with
concentrated solutions of both enzyme and precursor and it is probable
that no cross-reaction would have occurred with crude tissue extracts.
If the proteinogen is a denatured protein, such cross-reactions would
not be expected.

Recent work (Knight, 1946) has demonstrated that influenza virus
grown on chicken embryos reacts with normal chicken protein antiserum,

"^ These results, however, are open to the objection that the pepsin was denatured at
the /iH of the blood which the pepsinogen was not. Denaturation is known to modify
the specificity of proteins, and hence the failure to obtain a cross-reaction in this case
is not conclusive. The failure to differentiate bovine and swine pepsin is also complicated
by the fact that both proteins were denatured. This objection, however, does not apply
to Ten Broeck's results with chymo-trypsin.

28 JOHN H. NORTHROP

whereas influenza virus grown on mouse lung does not. The mouse-lung
virus, on the other hand, reacts with normal mouse antiserum but not
with normal chicken antiserum. After a single passage through chicken
embryos the mouse virus reacts with normal chicken protein antiserum.

The results of Gye and Purdy with Fujinama tumor virus from dif-
ferent hosts are very similar to those found by Knight for influenza
virus. In both cases the results agree with the predictions based on the
analogy between the formation of viruses and normal proteins.

It is not necessary to assume that any appreciable amount of pre-
cursor is present at any one time since it would serve its purpose as a
link between the synthetic and specific part of the reaction even though
it were only present in traces as an intermediate product.

Luria, Delbriick, and Anderson (1943) were unable to find any large
molecules, comparable to those of the virus, in uninfected cells. They
consider that this result proves that no precursor exists. There is no
necessity, however, for the precursor to be of the same molecular weight
as the virus since association of proteins to larger molecules is known to
occur readily. This step is quite different from the synthesis of the
primary bonds in the smaller units (Northrop, 1938).

Adaptation of Viruses.* In certain cases inoculation of a different
host results in the production of a different virus after a longer or
shorter time (equine encephalitis, yellow fever). This result may also
be accounted for on the basis of a precursor. Assume that Host i con-
tains Precursor Pi and that Host 2 contains a slightly different precur-
sor, P2. The original virus forms more of itself by reaction with Pre-
cursor Fi but forms more or less of a slightly different virus, V2, as well
as more of itself, by reaction with Precursor Pg- The reaction in Host
I is Pi + ^1 -* Vx. In Host 2 three reactions occur, P2 + ^1 -» ^1 ;
P2 + ^1 ~^ ^2 ; also P2 + f^2 -^ V2. If the production of Fi and V2 pro-
ceeds at exactly the same rate, both will be formed indefinitely. It is
extremely improbable that two such reactions have exactly the same
rates. If the rates differ, that reaction with the higher rate will eventu-

* Sudden changes in the properties of a virus are usually referred to as "mutations"
of the virus. As far as the writer is aware, there is no evidence to indicate whether the
mutation occurs in the virus or in the host cell. The experiments show that a different
virus appears as a result of some reaction in the host-virus system, but whether this is
due to a mutation in the host cell which then produces a different virus, or a mutation
in the virus itself, cannot be determined. If the mutation occurred first in the virus, it
would be expected that a chemically modified virus would reproduce itself, but as dis-
cussed above, it does not. Attempts to cause mutation by irradiating tobacco mosaic
virus were also unsuccessful, whereas mutations did arise if infected leaves were ir-
radiated (Kausche and Stube, 1939).

THE SYNTHESIS OF PROTEINS 29

ally be the only one which takes place. If Pg -^ V2, therefore, is more
rapid than P2 -^ Vi, the virus finally produced will be practically pure
V2, the length of time required depending on the relative reaction rates.

An example of such a transformation is furnished by Gratia's in-
teresting results (1936) with B. megatherium. This organism exists in
two forms, "lysogenic" and "sensitive." The lysogenic strain constantly
produces a virus T, which does not cause lysis of the parent lysogenic
strain but does cause lysis of the sensitive strain. Subculture of plaques
of this virus on the sensitive strain gives rise to two different viruses,
the original T and a new one, C, which causes lysis of both strains.
Similar results have been reported by Luria (1945).

It does not appear possible to make any generalization on theoretical
grounds as to whether or not the products of an "autocatalytic" reaction
are identical with the original catalyst. The hydrolysis of esters by acids
is an autocatalytic reaction and may give rise either to more of the same
acid or to different acids, depending on conditions. Thus the hydrolysis
of ethyl acetate, propyl acetate, and butyl acetate by means of acetic
acid will result in the formation of more acetic acid ; but the hydrolysis
of ethyl butyrate and ethyl propionate by acetic acid will result in the
production of butyric and propionic acids. Although hydrolysis is
caused primarily in this case by the hydrogen ion, which is the same in
all acids, the same relation holds presumably for proteins ; the ability
to produce certain reactions is primarily a property of certain groups
and these groups may be common to many different proteins.

Quantitative Data on the Formation of Bacterial Viruses
(Bacteriophages). If viruses are formed from precursors in the cell
it follows that the quantity of virus formed in any one cell should be
constant, provided no increase in the precursor occurs during the re-
action. Viruses in general, and bacterial viruses in particular, increase
most rapidly in the presence of growing host cells ; and under these
conditions the number of infected host cells is constantly changing and
more precursor is presumably being formed so that the amount of virus
formed per host cell cannot be determined. Krueger and his collaborators
(cf. Krueger and Scribner, 1939) found that bacteria viruses could be
produced under certain conditions without cell growth, and Krueger and
Baldwin (1937) obtained occasional virus increase in cell-free solution.
Krueger concluded that the virus was produced from a precursor in the
bacterial cell. Spizizen (1943) reports virus production in dilute solu-
tions of glycine anhydride without cell growth. Delbriick and Luria
separated virus production from cell growth by dilution of the culture

30 JOHN H. NORTHROP

after inoculation with virus. All of these workers agree that the bacterial
cell must be in good physiological condition and also that each cell forms
a certain definite amount of virus.

Delbriick and Luria (1942) have investigated the reaction in a series
of extremely careful and interesting papers and have obtained a detailed
picture of the steps in the increase of virus. Their results may be sum-
marized as follows :

1. Virus is rapidly absorbed by the host cell, the rate of absorption
being proportional to the concentration (of. Krueger, 1931).

2. No increase in infective centers, as indicated by plaque count, oc-
curs for a definite time interval which is characteristic of the virus.

3. The number of infective centers then increases rapidly for a few
minutes until a certain definite number have appeared for each host cell.

4. The greater the number of virus molecules originally present in the
bacterial cell, the more virus molecules are finally formed, but increasing
the original number causes much less than the proportional increase in
the number formed. Delbriick and Luria assume that the increase in
virus observed with increasing infection is due to secondary complica-
tions and that actually the number of virus molecules formed per host
cell is independent of the infecting number.

5. Two different viruses, if grown separately, produce the same num-
ber of molecules from the same bacterial cell.

6. If the bacteria are inoculated with two different viruses simul-
taneously, only one virus increases. If the bacteria are inoculated with
two viruses successively, only the one which is first inoculated increases.

Delbriick and Luria conclude that the "intracellular" increase in
virus is limited by the availability of some substrate, and that liberation
takes place when the reaction has run to completion. They assume that
a "key enzyme" exists which controls the reaction and which can react
with only one virus molecule. This peculiarity accounts for the fact that
either virus but not both may increase, and also for the fact that the time
for the growth cycle is nearly independent of the number of infective
units originally present. It does not account for the fact that the number
of molecules formed increases as the number of infective units increases.

These results may be explained equally well by the precursor mecha-
nism, without assuming the existince of a key enzyme. The following
assumptions are necessary:

1, Each host cell contains a definite quantity of precursor. The quan-
tity varies depending on the physiological state of the cell.

2. Virus is formed from this precursor in accordance with the usual

THE SYNTHESIS OF PROTEINS

31

autocatalytic equation dv/dt = KV{Ve — V) when Ve = total concen-
tration at end of reaction and V = concentration® of virus at time t.

The explanation of the next step depends on whether or not the varia-
tion in the number of virus molecules formed varies with the number
of molecules with which the bacteria is infected. The direct experimental
results show that it does, but Delbriick and Luria suggest that this is due
to secondary reactions.

If it be assumed that the number of molecules produced does vary
with the size of the infection, then it must be assumed that the time re-
quired for liberation of the virus depends to some extent on the bacteria
cell and not entirely on the number of virus molecules present. There is
some indication of this, since changing the temperature changes this
time interval in the same way as it does the growth rate of the bacteria.

We may assume that virus is liberated after a definite time has elapsed
and that this time is determined both by the state of the bacteria cell
and by the nature of the virus, but not by the number of molecules of
virus. The theory of autocatalytic reaction applied to these conditions
predicts that the number of molecules produced in the interval will in-
crease more slowly than does the original number. Fig. 2 shows a series

0 .4 .8 1.2 1.6 2.0 2.4 2.8 3.2 3.6 4.0 44 4.8

T

Figure 2. Effect of initial concentration of the end product on an autocatalytic reaction.

^ It may be recalled in this connection that most enzyme reactions deviate from simple
theory in that the velocity does not change in the expected way with either the enzyme
or substrate concentration. In the case of the autocatalytic formation of pepsin from
pepsinogen, for instance, increasing the concentration has much less effect on the
velocity than predicted by the equation given above (Herriott, 1938).

32 JOHN H. NORTHROP

of such autocatalytic curves calculated by the autocatalytic equation for
the conditions that i, 5, 10, or 20 per cent of the final maximum number
of molecules are present at the beginning. The curves show that after
the first part of the reaction the increase in the number of molecules
formed after a given time interval increases much more slowly than does
the original number. For instance (Table I), after 2.0 minutes, 50

TABLE I

Value of A corresponding to different values of Ao after different time

intervals

2.3 A{A—Ao)

Calculated from Autocatalytic Equation, T = Log

KAe Ao(Ae-A)

K — .023 ; Ae= 100 -{- Ao; At =^ Number of particles present at time

T (mins.)

Ao

^1.5

-^2.0

■^2.5

As.o

I

25

50

75

90

5

65

90

100

103

10

90

104

107

108

20

no

115

117

118

molecules would be present if one were present at the beginning, whereas
only 115 would be present if twenty were present at the beginning.
After 3.0 minutes the corresponding figures are 90 and 118. These
figures are similar to those actually found by Delbriick and Luria. They
were calculated on the assumption that the reaction followed the theo-
retical course. Actually the effect of changing the total concentration
on the rate of enzyme reactions is usually less than that predicted by
simple theory. This effect would tend to decrease the difference in the
figures given above.

3. If the number of molecules liberated per bacterial cell is actually
constant and independent of the infective quantity as Delbriick and
Luria suggest, then it is necessary to assume that the liberation occurs
as soon as the virus-forming reaction is completed. This assumption
predicts also that lysis occurs as soon as the ratio of virus to bacteria
reaches a definite value. This result agrees with the experiments of
Krueger and Northrop (1930). If liberation of virus occurs as soon as
this constant amount of virus is formed, the time required should be
shorter as the number of infective units increases, although the varia-

THE SYNTHESIS OF PROTEINS 33

tion is small. Interpolating from Fig. 2 again, it is found that if the
reaction is started with i per cent of the final quantity, 3.4 minutes are
required to produce 95 per cent, whereas if it is started with 5 per cent
about 2.5 minutes are required, while 10 per cent requires 1.8 minutes.
This variation is greater than that actually found by Delbriick and
Luria for the total elapsed time, but the variation in the time of forma-
tion itself may be greater than is indicated by the variation in ob-
served time. In other words, the observed time is the time required for
the virus to be formed plus the time required for it to be liberated
from all the infected cells; Tabs. = Tform. + Tub.. If an appreciable part
of this time is required for liberation to occur, then relatively large
changes in formation time would cause only slight variations in the
total observed time. Information in this connection could be obtained
by titrating the phage by the dynamic method (Krueger, 1930), which
determines the actual quantity present and not simply the number of
infective centers.

4. The results obtained with a mixture of viruses may be explained
in a similar way. The first virus added increases so that by the time the
second virus is added, more or less of the precursor is used up. The rate
of formation of the second strain will then be in proportion to its con-
centration (assuming equal rates of reactions), so that if 50 units of
virus A had been formed at the time one unit of B was added, ap-
proximately 50 times as much A as B would be formed subsequently.

This mechanism differs from that of Delbriick and Luria principally
in that it predicts that a cell can produce both strains simultaneously,
whereas Delbriick and Luria's hypothesis predicts an all or none result.

The data actually show that both viruses are produced by the culture,
but this is explained by Delbriick and Luria as due to production of the
different viruses by different cells. Anderson (1942) found that chick
embryo cells may contain inclusion bodies of two viruses. Sylverton and
Berry (1936) have reported similar results. It is quite certain that some
cells, therefore, may produce two viruses simultaneously.

5. In experiments with partially inactivated virus Luria and Delbriick
(1942) found that exposure to ultraviolet light changes the virus so
that it no longer produces lysis, but it still stops bacteria growth and
prevents formation of the active form of the same or of another virus.

The inactivated virus does not increase. This result may be due simply
to the fact that the bacteria do not grow in the presence of the inacti-
vated virus, since anything which stops bacterial growth generally stops
virus formation.

34 JOHN H, NORTHROP

It is possible to assume, however, that the partially inactivated virus
reacts with the precursor and forms more "inactive virus," but this
virus is not liberated from the cell since no lysis occurs. From this point
of view the interference caused by the "inactivated" virus is similar to
the interference between two viruses and is due to the destruction of the
precursor.

IV. Formation of Adaptive Enzymes

The formation of normal proteins and enzymes and of viruses out-
lined in the preceding section is purely autocatalytic and results in the
formation of more of the same molecules. The formation of adaptive
enzymes or antibodies, however, is not purely autocatalytic, since the
molecules formed are different from those originally present. The hy-
pothesis, therefore, must be modified to account for the formation of
these compounds.

Enzymes are sometimes formed as a specific response to the presence
of the corresponding substrate or its decomposition products. Enzymes
formed in this way are called "adaptive enzymes" (Karstrom, 1930).

The production of galactase by yeast is the best known example but
many similar cases are known (cf. Yudkin, 1938; Knight, 1936; Dubos,
1939). "Normal" yeast cannot ferment galactose, but Dienert (1900)
found that yeast acquires this power if grown in the presence of galac-
tose. It was thought at first that this result was due to selection of yeast
cells during growth of the culture. Later work established the fact that
the enzyme can be formed without any increase in the number of cells.
If this observation is correct the fermentation cannot be due to selection
of yeast cells, but must be due to production® of a new enzyme." The
formation of the enzyme is always associated with the formation of
protoplasm, even though no cell division occurs (Dubos, 1939).

In some cases, like the polysaccharide-splitting enzyme of Dubos
(1939), the enzyme displays extreme specificity, in fact the specificity
of this enzyme is more precise than the immunological tests.

Formation of the enzymes may be caused by the decomposition prod-

9 There does not appear to be any advantage in distinguishing between "production"
and "activation" of an enzyme. An inactive enzyme, strictly speaking, is not an enzyme
at all.

1° Sevag (1946) considers that no new enzymes are actually formed but that the new
substrate increases the activity of the normal enzyme in some unexplained way. Ac-
cording to this point of view the phenomenon is simply a change in the specificity of the
reaction similar to that which occurs when various amino acids are added to peptidase
systems (cf. page 11).

THE SYNTHESIS OF PROTEINS 35

ucts of the substrate. The production of yeast invertase is brought about
by glucose as well as by sucrose (Eulcr and Cramer, 1913).

The production of the enzyme is not inherited but stops when the
substrate is removed.

The potential ability to produce the enzyme, however, is inherited as
a mendelian character (Lindegren, Spiegelman, and Lindegren, 1944).

Yudkin suggests that the new enzyme is always present in minute
amounts in all cells and is in equilibrium with its precursor. Addition
of the substrate or products of the reaction or other compounds which
react with the enzyme will decrease the enzyme concentration. This
disturbs the equilibrium and hence more enzyme will be formed. This
explanation is adequate and also accounts for the fact that substances
which combine with the enzyme, other than the substrate, may bring
about its production.

It includes the assumption that minute amounts of all possible "adap-
tive" enzymes exist in normal cells and also that the equilibrium is such
that very little enzyme is present. It is further necessary to assume that
equilibrium is reached slowly, since the new enzyme does not appear at
once. The hypothesis predicts that the new enzyme should be obtainable
in zntro if the precursor were available.

Actually there is no experimental evidence of an equilibrium between
an enzyme and its precursor, since the known reactions of this type
run to completion {in vitro) as far as can be determined. It is possible
to assume equilibrium, but it must be very far in the direction of the
active enzyme.

If the reaction were autocatalytic the combination of the enzyme with
the products of reaction would decrease the rate of reaction and hence
delay the attainment of equilibrium conditions. The formation of trypsin
from trypsinogen, for instance, is stopped by the addition of trypsin
inhibitor (Kunitz and Northrop, 1936).

The assumption that the substrate acts by changing the course of
formation of the normal (closely related) enzyme appears to be equally
probable. This assumption has the advantage that closely analogous cases
are known and also that it may be applied to the formation of anti-
bodies, a similar reaction. It has the disadvantage that it does not explain
the specificity of the new enzyme quite so definitely.

According to this hypothesis the normal enzymes are continuously
being formed from their precursors in the cell by an autocatalytic reac-
tion. When a new substrate or other compound which combines with
the enzyme is added, it may act as a co-enzyme and change the course

2,6 JOHN H. NORTHROP

of the reaction so that a slightly different enzyme is formed as well as
the usual enzyme. When the new substrate is no longer present the re-
action returns to its normal course and the normal enzyme alone is
formed. Examples of similar reactions in which the course of an enzyme
reaction is affected by the presence of substances other than enzyme and
substrate are well known. Thus either trypsin or an inert protein may
be formed during the autocatalytic formation of trypsin from trypsino-
gen; the nature of the product depends on the />H of the solution
(Kunitz, 1939a). Different tyrosinases may be formed from protyrosi-
nase under different conditions (Bodine, 1945). Abderhalden and
Ehrenwall (1933) found that addition of certain peptides caused glycyl-
leucine to be hydrolyzed by trypsin, whereas the reaction did not occur
when trypsin and glycyl-leucine alone were present. The experiments
were extended by Bergmann and his collaborators (cf. page 11). The
added peptides appear to act as co-enzymes.

The specificity of the amino acid oxidases is completel}^ changed by
the addition of other proteins or amino acids to the reaction mixture
(Edlbacher, 1946).

The products of an enzyme reaction usually combine with the enzyme
so that the fact that a reaction occurs between the enzyme produced and
the substance which caused its production does not differ qualitatively
from many enzyme reactions. The distinguishing factor in the present
case is that one of the products of the reaction is itself an enzyme.

V. Formation of Antibodies

Antibodies are considered to be modified serum globulins which are
changed in such a way as to react with the homologous antigen (cf.
Marrack, 1938; Zinsser, Enders, and Fothergill, 1939; Burnet, Free-
man, Jackson, and Lush, 1941 ; Landsteiner, 1945; and Sevag, 1945).
There is no doubt that antibodies are proteins, but their relationship
to the serum globulins appears to be somewhat uncertain owing to the
fact that neither antibodies nor serum globulins have been obtained in
pure form (cf. Roche, Derrien, and Mandel, 1944a and 1944b). ^^ Any

11 Crystalline diphtheria antitoxin (Northrop, 1941) is probably a pure protein and
is not a globulin, although it cannot be separated from the globulin fraction if mixed
with it by any known procedure except precipitation with the homologous antigen.
This result would prove that antibodies are distinct chemically and immunologically
from the normal serum proteins, were it not for the fact that the toxin-antitoxin complex
used to prepare the crystalline antibody was digested with trypsin. It is possible to
assume, therefore, that the pure crystalline antibody does not represent the natural
antibody but is an hydrolysis product.

Antibodies are reported to be associated with different serum proteins in different

THE SYNTHESIS OF PROTEINS 37

hypothesis for the formation of antibodies must account for the follow-
ing facts (Burnet, Freeman, Jackson, and Lush, 1941 ; Landsteiner,

1945):

1. They are proteins which react specifically with their antigens.

2. They may be produced by so many different antigens that it is
extremely unlikely that even one molecule of all possible antibodies ex-
ists in the normal animal.

3. They are produced in great quantity as compared to the quantity
of antigen which causes their formation.

4. Antibody continues to be formed after the antigen has disappeared.
This statement unfortunately cannot be considered an experimental fact
since the evidence, although strong, is not conclusive (cf. Burnet, Free-
man, Jackson, and Lush, 1941 ; Landsteiner, 1945). There is no doubt
that antibody continues to be formed long after any antigen can be
detected experimentally in the circulation, but it is possible to assume
that minute quantities of antigen are still present in the cells. As Burnet
points out, this question is of fundamental importance to the theories
of antibody formation proposed by Ostromuislenskii (1916), Koltzofif
(1928), Mudd (1932), Pauling (1940), Pauling and Campbell (1942),
and Sevag (1945 ) , since these hypotheses all include the assumption that
antibody is formed as a result of some type of reaction between antigen
and globulin. Harris and Ehrich state that the decomposition products
of antigens occur autocatalytically.

5. The chemical nature of antibodies may be different in different
species or in animals of different ages.

This statement is open to question since no antibodies have been iso-
lated in pure form.

6. Antibodies are formed only in the animal receiving the antigen.
Injection of another animal with antisera (passive immunization) re-
sults in rapid disappearance of antibody and not in the formation of
more antibody. This fact precludes any purely autocatalytic mechanism
for the formation of antibodies, unless secondary hypotheses are added.

7. Antibodies are formed in cells, probably in the lymphocytes
(Dougherty, Chase, and White, 1944, 1945; Dougherty and White,
1945; Harris and Ehrich, 1946).

animals and have been found in all serum fractions (cf. Landsteiner, 1945, p. 133).
These results also indicate that the antibody itself may be a distinct protein ratlior than
a modified globulin. On the other hand, young animals which do not form antibodies
readily are said to possess very small amounts of serum globulin (Orcutt and Howe,
1922).

28 JOHN H. NORTHROP

A number of reports of the formation of antibodies in vitro have
been made (Ostromuislenskii, 1916; PauHng, 1940) but the identity
of the substance prepared in this way with the natural antibody has not
been estabhshed. It must be remembered in connection with these experi-
ments that owing to experimental difficulties whole blood cannot be
used in vitro. Failure of the experiment when carried out with serum,
plasma, defibrinated blood, or in the presence of anti-coagulants, there-
fore, does not prove that the reaction cannot occur in the circulation.

Burnet's Hypothesis. Burnet, Freeman, Jackson, and Lush (1941)
consider the formation of antibodies to be analogous to the formation
of adaptive enzymes. Indeed, there are many points of similarity. Anti-
bodies and enzymes may both be divided into two general classes-
normal or physiological, and acquired or adaptive. The "normal" en-
zymes or antibodies are present in all individuals of the species and are
inherited (cf. Landsteiner, 1945, p. 132). The ability to form adaptive
enzymes also appears to be inherited (Lindegren, Spiegelman, and
Lindegren, 1944). "Acquired" antibodies or adaptive enzymes are
formed in response to the presence of a foreign substance, and their
formation stops sooner or later after removal of this substance.

Both antibodies and adaptive enzymes have a specific relation to the
substance which caused their formation. In some cases adaptive enzyme
reactions are more specific than the antibody reactions (Dubos, 1939)-
The reaction of enzymes with substances formed during the reaction
is quantitatively similar to the reaction between antigen and antibody
(Northrop, 1922; and Zinsser, 1923). The antigen-antibody complex
is less dissociated than most enzyme-product compounds.

According to Burnet, normal globulin is synthesized by a special
proteinase. Injection of antigen results in a reaction between this pro-
teinase and the antigen, which destroys the antigen and modifies the
proteinase. This modified proteinase causes the formation of antibody
instead of normal globulins, and also causes replicas of itself to be
formed. The modified proteinase resulting from reaction with the anti-
gen decreases gradually when antigen is no longer present, as do adap-
tive enzymes.

The qualitative and quantitative differences in antibody formation are
therefore accounted for. The hypothesis also accounts for the fact that
small quantities of antigens can cause the production of very large
quantities of antibody and predicts that antibody formation may con-
tinue for an indefinite length of time after the antigen and even its
decomposition products have disappeared.

THE SYNTHESIS OF PROTEINS 39

It appears to the writer that this hypothesis predicts that injection
of an animal with antiserum should lead to the production of more anti-
bodies, since the production of the antibody-forming proteinase is as-
sumed to be autocatalytic. These proteinases are assumed to be present
in the blood and so should cause formation of antibody when blood con-
taining them is injected into a different animal. Actually the injection
of antiserum does not lead to the production of more antibodies, but on
the contrary the antibody injected disappears quite rapidly. It is possible
to avoid this discrepancy by means of secondary hypotheses, but it ap-
pears to be a serious objection to any purely autocatalytic mechanism.
There is no experimental evidence for the existence of the "proteinases,"
as far as the writer is aware. In the absence of such evidence the assump-
tion of their existence seems unnecessary and leads to the further com-
plication of accounting for their own formation.

Sevag's Hypothesis. Sevag (1945) has recently reviewed the
various hypotheses for the formation of antibodies and has carefully
and thoroughly discussed the relation of antigens to enzymes. There is
no doubt that many points of similarity exist between the action of
enzymes on their substrates and the formation of antibodies after injec-
tion of antigens. Sevag concludes that all proteins are enzymes and that
antibodies are formed as a result of the action of the antigens (enzymes)
on the normal globulins. According to this hypothesis antibodies should
be readily formed by the action of antigens on normal serum in vitro,
but this is not actually the case. This discrepancy, however, cannot be
considered conclusive owing to the difficulty of carrying out such an
experiment under the conditions which exist in the blood vessels.

Sevag's hypothesis also predicts that antibodies can be produced only
so long as antigen is present.

Antibody Formation from Proteinogen

The writer agrees with Burnet that antibodies and adaptive enzymes
are closely related and are probably formed by the same mechanism. The
hypothesis describing the formation of adaptive enzymes outlined above
is simpler than that of Burnet and agrees as well with the known facts
of antibody formation.

The reaction may be summarized as follows :

I. The normal serum proteins are formed from the "proteinogen"
by an autocatalytic reaction.

40 JOHN H. NORTHROP

2. Antigen (or its decomposition products) acts as a co-enzyme^" in
this reaction and causes the formation of a slightly different protein, in
the same way that the substrate causes the formation of the adaptive
enzymes (page 34). The resulting protein reacts with the co-enzyme
(antigen) which led to its formation since products of enzyme reactions
in general react with the enzyme which produced them.

3. The antibody is formed almost entirely in the cells where proteino-
gen is synthesized, since the quantity of proteinogen present in the blood
at any time is probably small.

These assumptions account for the following facts :

1. Antibodies are related to normal serum proteins and react with
the antigen which led to their formation.

2. Small quantities of antigen can produce indefinite amounts of
antibody.

3. Antibody production may continue after all antigen has disap-
peared from the circulation but not after all decomposition products of
the antigen have disappeared (Harris and Ehrich, 1946). In this respect
it differs from Burnet's and Sevag's hypotheses.

4. The antibody may be related to one or several normal proteins,
depending on which autocatalytic reaction is affected by the antigen.

5. More antibodies will not be formed when antiserum is injected
into an animal, since antibody formation itself is not assumed to be
autocatalytic. This prediction also differs from Burnet's hypothesis.

6. Antibodies are not formed in any appreciable amount in the absence
of cells. The proteinogen is formed in the cells, since it is only in the
cells that the necessary mechanism exists for the energy supply required
for synthesis. It will be transformed into the various special proteins of
the circulation as soon as it comes in contact with them. The formation
of antibodies will therefore occur principally in or near the synthesizing
cells. All the various special proteins, including antibodies, are assumed
to be in equilibrium with the proteinogen and hence are in equilibrium
with each other. Therefore it should be possible to prepare antibodies
in vitro by adding antigens to blood. The quantity of proteinogen ac-
tually present is assumed to be very small, so that the rate of reaction
may be slow. There is the further technical difficulty that whole blood
cannot be used (cf. page 39). Production of antibodies in this way has
been frequently reported (cf., for instance, Ostromuislenskii, 1916;
Pauling, 1940).

The hypothesis does not account in any simple way for the observa-

12 Sevag (1945) assumes that the antigens themselves act as enzymes.

THE SYNTHESIS OF PROTEINS

41

tion that antibody formation is sometimes accompanied by an increase
in normal serum globulins.

Rate of Appearance of Antibodies in the Circulation

Injection of antigen gives rise after a few days to a rapid increase
of antibody in the circulation. In some cases a second injection of anti-
gen gives rise to a second and larger increase in antibody, especially
if a small amount of antigen were injected the first time. The curves
are logarithmic in form and Burnet considers this to be evidence for
the existence of some process of biological multiplication. Actually the

E

o

jO

<0

o

4-

3-

00 1

5000-

0

-0
c
rvj

4000

1

-0

3000

0.

■0

E

2000

-0
0

1000

—

2nd injection

> —

•— —

1st injection

0

1

1

1

2

1
3

1
4

5

1 1 1
6 7 8

Days after injection

1st injection

2nd injection

lUU

0

1

0

a>

■ — \

flO

c

„_

vt

■•~

>^

60

CO
■C3

l_

a>

a.

T3

40

e

. V-

0

>»

?0

•Z3
0

J=>

,

C

<:

012345678

Days

Figure 3. Production of antibody after varying lengths of time.

42 JOHN H. NORTHROP

curves have long "lag periods" and are not strictly logarithmic except
for a small part of the reaction. They agree just as well or better with
the integral of the usual probability curves. (For the mathematical
relation between logarithmic and probability curves, see Yule, 1910.)
The case is very similar to the rate of death curves of bacterial cultures
or of insects (Loeb and Northrop, 191 7). The data may be made to fit
a log curve if the "lag period" is neglected. The appearance of antibody
in the blood, plotted as log curves (Burnet, Freeman, Jackson, and Lush,
1941), and as probability curves, is shown in Fig. 3. From the point
of view of probability curves, the results indicate that the cells in which
the antibody is produced die or are ruptured when antibody is released
into the blood stream. The rapid and large rise sometimes noted on a
second injection of antigen may thus be related to the phenomenon of
anaphylaxis. The cells are sensitized by the first small injection and
injured by the second. This mechanism agrees with the process of
antibody formation suggested by Sabin (1939).

REFERENCES

Abderhalden, E., and E. v. Ehrenwall. Continued studies on the provocation
of ereptic action in "erepsin-f ree" trypsin solutions prepared according
to Waldschmidt-Leitz and Harteneck. iv. Fennentforschung, 14, i,

1933-
Alcock, R. S. The synthesis of proteins in vivo. Physiol. Rev., 16, i, 1936.

Allen, T. H., O. M. Ray, and J. H. Bodine. Enzymes in ontogenesis of

Orthoptera. vi. Autocatalytic nature of formation in vivo of pro-

tyrosinase [in grasshopper eggs]. Proc. Soc. Exp. Biol. N.Y., jp, 549,

1938.

Anderson, K. Dual virus infection of single cells. Amer. J. Path., 18, 577,

1942.
Anson, M. L., and W. M. Stanley. Some effects of iodine and other reagents

on the structure and activity of tobacco mosaic virus. /. Gen. Physiol.,

24,679,1941.
Avery, O. T., C. M. MacLeod, and M. McCarty. Chemical nature of the

substance inducing transformation of pneumococcal types. Induction

of transformation by a desoxyribonucleic acid fraction isolated from

pneumococcus Type IIL /. Exp. Med., jg, 137, 1944.
Bayliss, W. M. The Nature of Enzymes. 5th edition, Longmans, Green and

Co., London, 1925.
Beadle, G. W. Biochemical genetics. Chem. Rev., 57, 15, 1945-
Behrens, O. K., and M. Bergmann. Cosubstrates in proteolysis. /. Biol.

Chem., 129, 587, 1939.

THE SYNTHESIS OF PROTEINS 43

Bergmann, M., and J. S. Fruton. The specificity of proteinases. Advances
in Enzymology, i, 63, 1941.

Bergmann, M., and J. S. Fruton. The significance of coupled reactions for
the enzymatic hydrolysis and synthesis of proteins. Ann. N.Y. Acad.
Sc, 45, 409, 1944.

Bodine, J. H. Tyrosinase and phenols. Action of diversely activated tyro-
sinase on monohydric and o-dihydric phenols. Proc. Soc. Exp. Biol.
N.Y., 58, 205, 1945.

Bordet, J. Theories of the bacteriophage. Proc. Roy. Soc. London, b., 10/,

398, 1931.
Bordet, J., A. Gratia, and D. Jaumain. Dualite du principe lytique du

colibacille et du staphylocoque. C. R. Soc. Biol. Paris, 83, 1095, 1921.
Borsook, H. Reversible and reversed enzymic reactions. Ergebn. Ensym-

forsch., 4, I, 1935.
Borsook, H., and J. W. Dubnoff. Biological synthesis of hippuric acid in

vitro. /. Biol. Cheni., 132, 307, 1940.
Borsook, H., and J. W. Dubnoff. The metabolism of proteins and amino

acids. Ann. Rev. Biochem., 12, 183, 1943.
Borsook, H., and G. L. Keighley. The "continuing" metabolism of nitrogen

in animals. Proc. Roy. Soc. London, b., 118, 488, 1935.
Borsook, H., and H. F. Schott. The role of the enzyme in the succinate-

enzyme-fumarate equilibrium. /. Biol. Chem., Q2, 535, 193 1.
Braunstein, A. E. Die enzymatische Umaminierung der Aminosauren und

ihre physiologische Bedeutung. Enzymologia, 7, 25, 1939.
Braunstein, A. E., and M. G. Kritzmann. tJber den Ab- und Aufbau von

Aminosauren durch Umaminierung. Enzymologia, 2, 129, 1937.
Bredig, G. Die Elemente der chemischen Kinetik, mit besonderer Beriick-

sichtigung der Katalyse und der Fermentwirkung. Ergebn. Physiol., i,

134, 1902.
Burke, V., N. P. Sullivan, H. Petersen, and R. Weed. Ontogenetic change

in antigenic specificity of the organs of the chick. /. Infect. Dis., 7^,

225, 1944.
Burnet, F. M. Virus as Organism. Harvard University Press, Cambridge,

1945-
Burnet, F. M., M. Freeman, A. V. Jackson, and D. Lush. Production of

Antibodies, Macmillan and Co., Ltd., Melbourne, 1941.
Caspersson, T. The role of nucleic acids in the cell. Proc. /th Intern. Ge-

netical Congr., Edinburgh, Aug. 23-30, 1941, 85, 1939.
Cori, G. T., and C. F. Cori. The activating effect of glycogen on the enzymic

synthesis of glycogen from glucose- 1 -phosphate. /. Biol. CJiem., iji,

397', 1939-
Darlington, C. D. Heredity, development and infection. Nature, 154, 164,

1944-
Delbriick, M. A theory of autocatalytic synthesis of polypeptides and its

application to the problem of chromosome reproduction. Cold Spring

Harbor Symposia, p, 122, 1941.
Delbriick, M., and S. E. Luria. Interference between bacterial viruses, i. In-

44 JOHN H. NORTHROP

terference between two bacterial viruses acting upon the same host,
and the mechanism of virus growth. Arch. Biochein., i, ill, 1942.

Dienert, F. Sur la fermentation du galactose et sur I'accoutumance des
levures a se sucre. Ann. Inst. Past., 14, 139, 1900.

Dougherty, T. F., J. H. Chase, and A. White. The demonstration of anti-
bodies in lymphocytes. Proc. Soc. Exp. Biol. N.Y., ^y, 295, 1944.

Dougherty, T. F., J. H. Chase, and A. White. Relationship of antibody con-
tent of normal and malignant lymphocytes. Proc. Soc. Exp. Biol. N.Y.,
59, 172, 1945.

Dougherty, T. F., and A. White. Functional alterations in lymphoid tissue
induced by adrenal cortical secretion. Am. J. Anat., 77, 81, 1945.

Dubos, R. J. Enzymic analysis of the antigenic structure of pneumococci.
Ergehn. Enzymforsch., 8, 135, 1939.

Ecker, P. G. Ultracentrifuge study of Plastein. /. Gen. Physiol., 50, 399,

1947.
Edlbacher, S. The problem of totality in biochemistry. Experientia, 2, 7,

1946.

Euler, H. v. Chemie der Enzyme, i. Verlag von J. F. Bergmann, 3rd edi-
tion, Miinchen, 1925.

Euler, H. v. Chemie der Enzyme, 11. Verlag von J. F. Bergmann, 2nd and
3rd editions, Munchen, 1927.

Euler, H. v., E. Adler, G. Giinther, and M. B. Das. Components of dehydro-
genase systems, xxi. Enzymic decomposition and synthesis of glutamic
acid in animal tissues. Z. physiol. Chem., 2^4, 61, 1938.

Euler, H. v., and H. Cramer. Untersuchungen iiber die chemische Zusam-
mensetzung und Bildung der Enzyme, ix. Zur Kenntnis der Invertase-
bildung. Z. phys. Chem., 88, 430, 191 3.

Folley, S. J. Nature of plastein. Biochem. J., 26, 99, 1932.

Gorter, E. Protein-Films. Tr. Faraday Soc, S3, 1125, 1937.

Gratia, A. On the so-called bacteriophage. Brit. Med. J., 2g6, 1922.

Gratia, A. Dissociation du bacteriophage du Bacillus megatherium lysogene
899 en deux varietes distinctes. C. R. Soc. Biol. Paris, 123, 1018, 1936.

Gratia, A. Nature des ultravirus, Levaditi et Lepine, Librairie Maloine,
Paris, 109, 1938.

Gulick, A. What are the genes? 11. The physico-chemical picture; conclu-
sions. Quart. Rev. Biol., 15, 140, 1938.

Gulick, A. The chemical formulation of gene structure and gene action.
Advances in Enzymology, 4, i, 1944.

Haddock, J. N., and L. E. Thomas. The synthesis of plasteins by the action
of trypsin and papain on digests of insulin. /. Biol. Chem., 144, 691,
1942.

Harris, T. N., and W. E. Ehrich. The fate of injected particulate antigens
in relation to the formation of antibodies. /. Exp. Med., 84, 157, 1946.

Haurowitz, F., M. Tunca, P. Schwerin, and V. Goksu. The action of trypsin
on native and denatured proteins. /. Biol. Chem., i^/, 621, 1945.

Herriott, R. M. Kinetics of the formation of pepsin from swine pepsinogen

THE SYNTHESIS OF PROTEINS 45

and identification of an intermediate compound. /. Gen. Physiol., 22,

65, 1938.
Herriott, R. M., Q. R. Bartz, and J. H. Northrop. Transformation of swine
pepsinogen into swine pepsin by chicken pepsin. /. Gen. Physiol., 21,

575. 1938. . , . , ,

Howe, H. A., and R. C. Mellors. Cytochrome oxidase m normal and re-
generating neurons. /. Exp. Med., 81, 489, 1945.
Jordan, P. Zum Problem der Eiweisz-Autokatalysen. Naturwissensch., 32,

20, 1944.
Kalckar, H. M. The function of phosphate in enzymatic syntheses. Ann.

N.Y. Acad. Sc., 45, 395, I944-
Karstrom, H. Enzyme production in bacteria. Lab. Butterexportges. Valio

m.b. H., Helsinki, 1930.
Kausche and Stube. Naturwis., 2J, 501, 1939.
Knight, B. C. J. G. Bacterial Nutrition. Sp. Rep. Ser. Med. Res. Coun.,

London, No. 210, 1936.
Knight, C. A. Sedimentable component of allantoic fluid and its relationship

to influenza viruses. /. Exp. Med., 80, 83, 1944-
Knight, C, A. Precipitin reactions of highly purified influenza viruses and

related materials. /. Exp. Med., 83, 281, 1946.
Koltzoff, N. K. Physikalisch-chemische Grundlage der Morphologic. Biol.

Zentr., 48, 345, 1928.
Krueger, A. P. A method for the quantitative determination of bacterio-
phage. /. Gen. Physiol., 13, 557, 1930.
Krueger, A. P. The sorption of bacteriophage by living and dead susceptible

bacteria, i. Equilibrium conditions. /. Gen. Physiol., 14, 493, 1931.
Krueger, A. P., and D. M. Baldwin. Production of phage in the absence of

bacterial cells. Proc. Soc. Exp. Biol. N.Y., 37, 393, 1937.
Krueger, A. P., and J. H. Northrop. The kinetics of the bacterium-bacte-

riophage reaction. /. Gen. Physiol., 14, 223, 1930.
Krueger, A. P., and E. J. Scribner. Serial production of phage from intra-
cellular phage precursor. Proc. Soc. Exp. Biol. N.Y., 40, 51, 1939.
Kunitz, M. Formation of trypsin from trypsinogen by an enzyme produced

by a mold of the genus penicillium. /. Gen. Physiol., 21, 601, 1938.
Kunitz, M. Effect of the formation of inert protein on the kinetics of the

autocatalytic formation of trypsin from trypsinogen. /. Gen. PJiysioL,

22, 293, 1939a.
Kunitz, M. Formation of trypsin from crystalline trypsinogen by means of

enterokinase. J. Gen. Physiol., 22, 429, 1939b.
Kunitz, M., and J. H. Northrop. Crystalline chymo-trypsin and chymo-

trypsinogen. i. Isolation, crystallization, and general properties of a

new proteolytic enzyme and its precursor. /. Gen. Physiol., 18, 433,

1935.
Kunitz, M., and J. H. Northrop. Isolation from beef pancreas of crystalline
trypsinogen, trypsin, a trypsin inhibitor, and an inhibitor-trypsin com-
pound. /. Gen. Physiol., ip, 991, 1936.

46 JOHN H. NORTHROP

Landsteiner, K, The Specificity of Serological Reactions. Harvard Uni-
versity Press, revised edition, Cambridge, Mass., 1945.
Langmuir, I. Molecular layers. Proc. Roy. Soc. London, a., //o, 940, i,

1939-
Langmuir, I., and V. J. Schaefer. Activities of urease and pepsin mono-
layers. /. Amer. Chem. Soc, 60, 1351, 1938.
Lindegren, C. C, S. Spiegelman, and G. Lindegren. Mendelian inheritance

of adaptive enzymes in yeast. Proc. Nat. Acad. Sc, 50, 346, 1944.
Linderstrom-Lang, K., and C. F. Jacobsen. The contraction accompanying

enzymic breakdown of proteins. C. R. Trav. Lab. Carlsherg, Sir.

Chim., 24, I, 1941.
Lloyd, W., M. Theiler, and N. L Ricci. Modification of the virulence of

yellow fever virus by cultivation in tissues in vitro. Tr. Roy. Soc. Trop.

Med. Hyg., 2Q, 481, 1936.
Loeb, J., and J. H. Northrop. On the influence of food and temperature

upon the duration of life. /. Biol. Chem., ^2, 103, 1917.
Loring, H. S. Solubility studies on purified tobacco mosaic virus. /. Gen.

Physiol., 2^, 719, 1940.
Luria, S. E., and M. Delbriick. Interference between inactivated bacterial

virus and active virus of the same strain and of a dififerent strain. Arch.

Biochem., i, 207, 1942.
Luria, S. E., M. Delbriick, and T. F. Anderson. Electron microscope

studies of bacterial viruses. /. Bad., 46, 57, 1943.
Luria, S. E. Mutations of bacterial viruses aflfecting their host range. Ge-
netics, so, 84, 1945.
Madden, S. C, and G. H. Whipple. Plasma proteins. Physiol. Rev., 20, 194,

1940.
Marrack, J. R. The Chemistry of Antigens and Antibodies. 2nd edition. Sp.

Rep. Ser. Med. Res. Coun., London, No. 230, 1938.
Meyerhof, O. Energy relationships in glycolysis and phosphorylation. Ann.

N.Y. Acad. Sc, 45, 2,77, 1944.
Miller, G. L., and W. M. Stanley, Derivatives of tobacco mosaic virus, i.

Acetyl and phenylureido virus. /. Biol. Chem., 141, 905, 1941.
Moore, D. H., S. C. Shen, and C. S. Alexander. The plasma of developing

chick and pig embryos. Proc Soc Exp. Biol. N.Y., 58, 307', 1945.
Mudd, S. A hypothetical mechanism of antibody production. /. Immunol.,

23, 423, 1932.
Muller, H. J. Pilgrim Lecture, London, November 1945 (in press).
Murphy, J. B. Transplantability of malignant tumors to the embryos of

foreign species. /. Amer. Med. Assn., 5q, 874, 191 2.
Northrop, J. H. The inactivation of trypsin, iii. Spontaneous inactivation.

/. Gen. Physiol., 4, 261, 1922.
Northrop, J. H. The formation of enzymes. Physiol. Rev., i/, 144, 1937.
Northrop, J. H. Concentration and purification of bacteriophage. /. Gen.

Physiol, 21, 335, 1938.
Northrop, J. H. Crystalline Enzymes. Columbia University Press, New

York, 1939.

THE SYNTHESIS OF PROTEINS 47

Northrop, J. H. Crystalline diphtheria antitoxin. Proc. Amer. Phil. Soc, 85,

13, 1941.
Northrop, J. H. Plastein from Pepsin and Trypsin. /. Gen. Physiol., jo,

Z77, 1947-
Orcutt, M. L., and P. E. Howe. Relation between the accumulation of

globulins and the appearance of agglutinins in the blood of new born

calves. /. Exp. Med., 36, 291, 1922.
Ostromuislenskii, I. I. Investigations of toxins and antitoxins. Chem.

Abstr., 10, 214, 1916.
Pauling, L. Theory of the structure and process of formation of antibodies.

/. Amer. Chem. Soc., 62, 2643, 1940.
Pauling, L., and D. H. Campbell. Manufacture of antibodies in vitro. /. Exp.

Med., 76, 211, 1942.
Pedersen, K. O. Ultracentrifugal Studies on Serum and Serum Fractions.

Almquist & Wiksells Boktryckeri AB, Uppsala, 1945.
Petrie, A. H. K. Protein synthesis in plants. Biol. Rev. Cambridge Phil.

Soc, 18, IDS, 1943.
Potter, V. R., and H. G. Albaum. Mechanism of hydrogen transport in

animal tissues, vii. Inhibition by ribonuclease. /. Gen. Physiol., 26, 443,

1943-
Price, W. H. Proc. Nat. Acad. Sci., 34, 317, 1948.
Reiner, J. M., and Spiegelman, S. Tech. Proc, 7, 98, 1948.
Rivers, T. M. Viruses and Virus Diseases. Stanford University Press,

Stanford University, California, 1943.
Robertson, T. B. The function of the lipoid in mitochondria. Australian J.

Exp. Biol, and Med. Sc, 3, 97, 1926.
Roche, J., Y. Derrien and S. Mandel. Heterogeneite des fractions eu- et

pseudoglobuliniques des proteines du serum de cheval et relargage

par les sels neutres. C. R. Soc. Biol. Paris, 138, 634, 1944a.
Roche, J., Y. Derrien, and S. Mandel. Separation de trois cristalbumines du

serum de cheval. C. R. Soc Biol. Paris, 138, 665, 1944b.
Sabin, F. R. Cellular reactions to a dye-protein with a concept of the

mechanism of antibody formation. /. Exp. Med., 70, 67, 1939.
Schoenheimer, R. The Dynamic State of Body Constituents. Harvard Uni-
versity Press, Cambridge, Mass., 1942.
Seastone, C. V., and R. M. Herriott. Immunological studies on pepsin and

pepsinogen. /. Gen. Physiol., 20, 797, 1937-
Seifriz, W. A materialistic interpretation of life. Phil. Sc, 6, 266, 1939.
Sevag, M. G. Immuno-Catalysis. Charles C. Thomas, Springfield, 111., and

Baltimore, Md., 1945.
Sevag, M. G. Enzyme problems in relation to chemotherapy, "adaptation,"

mutations, resistance, and immunity. Advances in En::ymology, 6, 33,

1946.
Spiegelman, S., and M. D. Kamen. Genes and nucleoproteins in the synthesis

of enzymes. Science, 104, 581, 1946.
Spizizen, J. Biochemical studies on the phenomenon of virus reproduction.

J. Infect. Dis., 72, 212, 1943.

48 JOHN H. NORTHROP

Stanley, W. M. Biochemistry and biophysics of viruses. In Doerr and

Hallauer, Handhuch der Virusforschung, Verlag von Julius Springer,

Wien, 1938a.
Stanley, W. M. The reproduction of virus proteins. Amer. Naturalist, 72,

no, 1938b.
Stanley, W. M. The biochemistry of viruses. Ann. Rev. Biochem., p, 545,

1940.
Stanley, W. M. Chemical structure and the mutation of viruses. From

Virus Diseases. Cornell University Press, Ithaca, N.Y., p. 35, 1943.
Svedberg, The. Protein molecules. Chem. Rev., 20, 81, 1937.
Sylverton, J. T., and G. P. Berry. Coexistent infections of individual cells

by more than one filterable virus. /. Bact., 32, 356, 1936.
Taylor, A. E. On the synthesis of protein through the action of trypsin.

/. Biol. Chem., 3, 87, 1907.
Taylor, A. E. On the synthesis of protamin through ferment action. /. Biol.

Chem., 5, 381, 1909.
Taylor, H. S. (editor). A Treatise on Physical Chemistry. Two volumes.

2nd printing. D. Van Nostrand Co., New York, 1925.
Ten Broeck, C. The differentiation of trypsins by means of the anaphylactic

test. /. Biol. Chem., 106, 729, 1934.
Traub, E., and C. Ten Broeck, Protective vaccination of horses with modi-
fied equine encephalomyelitis virus. Science, 81, 572, 1935-
Troland, L. T. Biological enigmas and the theory of enzyme action. Amer.

Naturalist, 57, 321, 191 7.
Wasteneys, H., and H. Borsook. The enzymatic synthesis of protein.

Physiol. Rev., 10, no, 1930.
Woodrufif, A. M., and E. W. Goodpasture. The susceptibility of the chorio-
allantoic membrane of chick embryos to infection with the fowl-pox

virus. Amer. J. Path., 7, 209, 1931.
Yudkin, J. Enzyme variation in microorganisms. Biog. Rev. Cambridge

Phil. Soc, 13, 93, 1938.
Yule, G. U. On the distribution of deaths with age when the causes of death

act cumulatively and similar frequency distributions. /. Roy. Statist.

Soc. London, 73, 26, 1910.
Zinsser, H. Infection and Resistance. 3rd edition. The Macmillan Co., New

York, p. 141, 1923.
Zinsser, H. On the nature of virus agents. Amer. J. Pub. Hlth., 2/, 1160,

1937.
Zinsser, H., J. F. Enders, and L. D. Fothergill. Immunity, principles and

application in medicine and public health. 5th edition of Resistance to

Infectious Diseases. The Macmillan Co., New York, 1939.

II. MOLECULAR MORPHOLOGY
AND GROWTH

BY FRANCIS O. SCHMITT^

MOLECULAR morphology and growth are both very large subjects.
Their combination encompasses a respectable portion of the
content of biology. If we are to bring into sharp focus a few of
the fundamental problems, to achieve a measure of synthesis and projec-
tion into profitable paths of future investigation, certain limitations of
scope must be imposed.

One of the uniquely characteristic properties of protoplasmic systems
is the structural organization by which the various energy-giving re-
actions are coupled and coordinated with the molecular effectors, the
"machinery," of the cell. While the molecular morphology of the com-
ponents which participate in cellular energetics, the enzymes, hormones,
metabolites and so on is also highly critical, we shall concern ourselves
chiefly with the macromolecular constituents of cell structure and in
particular with the fibrous proteins.

In discussing a subject as broad as that of growth, it would be im-
possible to exclude so important a subject as the globular proteins, since
many of the enzymes which preside over cellular energetics belong in
this category. As we shall see, a consideration of fibrous proteins in-
volves the globular proteins very closely. Unfortunately the space
allotted here is too limited to permit a full discussion of the detailed
structure of proteins generally.

I am not unmindful of the importance, for cell structure, of protein
complexes which are not perceptibly fibrous in nature. Included with
these are the various cytoplasmic particulates such as the microsomes,
secretion granules, Nissl substance and so on. However, polarization
and electron optical evidence indicates that the fibrous proteins form the
basis of the molecular fabrics which constitute the structural matrix of
the cell. Hence my own remarks will be limited to this aspect of molecu-
lar morphology.

With these apologies I shall proceed with a brief statement of the
present state of knowledge about the fibrous proteins and offer some
suggestions about fruitful fields for further study.

1 Department of Biology, Massachusetts Institute of Technology, Cambridge, Massa-
chusetts.

50 FRANCIS O. SCHMITT

The Technical Background

The determination of protein structure requires a complete and exact
knowledge of the amino acid composition and of the configuration of
the amino acid residues. To what extent are the physical and chemical
techniques so far developed adequate for this purpose and to what
extent have these techniques been applied?

Until recently amino acid data were not only relatively unreliable but
also incomplete with respect to the constituents present in small amounts.
If proteins are to be dealt with stoichiometrically, the analytical data
must be exact. Recent advances in amino acid analysis, particularly the
stable isotope and microbiological and adsorption techniques, should
help to fill many of the gaps in the cases where really pure preparations
can be obtained. Unfortunately few fibrous proteins have been prepared
in a state of purity approaching that which is required.

Fairly good technical and theoretical background exists for the
determination of particle size and shape. This includes the methods of
streaming double refraction, viscosity, diffusion and ultracentrifuga-
tion. But those methods yield appropriate information only when the
preparation is truly monodisperse. In the case of myosin, for example,
this criterion has not been fully met. The particles are fairly uniform
in cross section but have highly variable lengths. Hence the measure-
ments led to an average value for particle size, the larger sizes receiv-
ing greater weight, rather than to dimensions which could be ascribed
to the myosin molecule.

Electron staining and shadow casting techniques have greatly en-
hanced the value of the electron microscope in the investigation of the
structure of fibrous proteins. Resolution of the order of 25 A has
already been achieved with favorable materials and it is probable that
this may be further increased. Small-angle x-ray dififraction techniques
permit the recording of reflections corresponding to periodicities of the
order of hundreds of Angstrom units, thus widely overlapping the
resolving power of the electron microscope. The electron microscope
data assist greatly in the interpretation of the diffraction patterns of
fibrous proteins such as collagen and paramyosin which manifest a
large fiber-axis repeating period. In these cases fine structure within
the main period can be demonstrated in the electron micrographs ; this
fine structure may be correlated with the x-ray data to deduce the struc-
ture.

Only a few of the fibrous proteins have been studied with all of the

MOLECULAR MORPHOLOGY

51

physical and chemical techniques mentioned. An attempt to show the
state of our knowledge, or perhaps our ignorance, is presented in
Figure i. The individual rectangles represent eight methods of studying

KERATIM

HY03IN

riBRINOOEH-
FIBRIN

TOBACCO MOSAIC COLLAOEN
VIRUS PROTEIN

ELA3TIN

SILK

INSULIN
riBSRS

8TR0MAIIN

ACTIN

PARAMYOSIR

NEURONIN

NEURO-
KERATIN

l_a.

CHRONOaOMS
PROTEIN

RETINAL ROD
PROTEIN

PROTEIN FROM

TRICHOCXSt

SPERM TAIL

CILIARX

MITOTIC

DEVELOPINa

PROTEIN

PROTEIN

PROTEIN

PROTEIN

SMBFOrOS

Figure i. Rough appraisal of experimental data available on physical and chemical
properties of fibrous proteins. The eight rectangles, each characteristically shaded,
represent from left to right : amino acid analysis, x-ray diffraction, electron microscopy,
polarized light microscopy, streaming double refraction, viscosity, diffusion and ultra-
centrifuge analysis. The height of each rectangle represents the relative amount of
work published on the application of this method to the protein in question. Since this
chart was prepared important new results have been published, particularly in the case
of actin, myosin, and the actomyosin complex.

molecular structure. The height of each rectangle reflects the relative
amount of work which, to the author's knowledge, has been reported on
the subject. Of course, the mere fact that much work may have been
done on a subject does not necessarily mean that crucial conclusions have
been reached.

The figure is admittedly very rough, but the many gaps in our
knowledge are obvious. There must be concentration both on breadth
and depth. As more proteins are investigated the comparative aspect

^2 FRANCIS O. SCHMITT

will help to crystallize our concepts. At the same time there should be
heavy concentration on a few of the most favorable proteins which
can be isolated in relative purity and which manifest a high degree of
structural regularity. Lacking more direct and more sensitive methods,
only thus may we approach the desired goal.

Internal Architecture of Fibrous Proteins

Astbury distinguishes two classes of fibrous proteins: the keratin-
myosin-fibrin class, which gives the characteristic alpha and beta wide-
angle x-ray patterns, and the collagen class. Astbury's general point of
view is well summarized in his essay, 'The Form of Biological Mole-
cules," included in the set of essays presented to D'Arcy Thompson. The
correlations which he has drawn have greatly helped to weave a com-
mon thread through the maze of assorted information which existed
previously. The data do not yet warrant conclusions of detailed structure
which are necessarily unique, but, as Astbury says, it must be "some-
thing like that." Until specimens are prepared which have sufficient
regularity of structure to give a greater wealth of diffraction data, it is
unlikely that the details of the folded alpha configuration will be

deducible.

In certain of the fibrous proteins a large repeating fiber-axis spacing
has been demonstrated by x-ray diffraction and electron microscopy.
The value of this spacing and the intensities of the various diffraction
orders are characteristic of each protein. Yet the true significance of
this remarkable characteristic remains unknown. Astbury has attempted
to interpret long spacings in terms of the amino acid composition along
stoichiometric lines. The long spacing is regarded as the product of the
number of amino acid residues in the molecule and the length, along
the fiber axis, of each residue as indicated by the meridional diffrac-
tions of the wide-angle pattern. However, the data are not yet suf-
ficiently complete to warrant this procedure except as a rough guide to
thought. Moreover, Bear has demonstrated in the case of collagen that
physical and chemical alteration may cause the long and short fiber-axis
spacings to vary somewhat independently of each other.

While the wide-angle x-ray patterns reflect similarities of structure,
the small-angle pattern is characteristic of each particular proteim
Fairly well developed small-angle patterns have been obtamed for a and
/? keratin, collagen and paramyosin. Myosin patterns have also been
obtained (Bear), but as yet these lack detail. No x-ray long-spacings
have been reported for fibrin, though Wolpers has observed a periodic

MOLECULAR MORPHOLOGY 53

pattern in electron micrographs. The shafts of Paramecium trichocysts
show very regular periodic structure in the electron microscope, but
this object has not yet been studied with x-rays. In addition to the large
fiber-axis periodicities, the fibrous proteins may also show large lateral
separations, and these further assist the characterization of the protein.
The information on these large periods is summarized in Table I.

TABLE I

Large Repeating Periods in Fibrous Proteins

Period

Repeating

Period

Fiber-axis Period

Lateral Period

A

A

/? Keratin

95

34

a Keratin

198

82

Muscle (actomyosin?)

350-420

115

Paramyosin

720

200-300

Collagen

640

Trichocyst Protein

550
(2200)

Another approach to the problem is to make electron microscope and
x-ray studies of fibrous proteins in which reactive groups have been
altered chemically, as by deamination, deguanylation and so on. Such a
program is currently being undertaken at M.LT. on the protein collagen.

Finally, a word must be said about the nucleoproteins. In Astbury's
study of mixtures of thymonucleic acid with the protamine, clupein,
the x-ray patterns were very similar to those of nucleic acid alone. The
only aspect which is stressed by Astbury is a fiber-axis period of 3.34 A
which is characteristic of nucleic acid and is supposed to indicate the
separation between nucleotide residues. Since this spacing is close to
3.5 A, the distance between amino acid residues in fully extended beta
proteins, Astbury supposes a salt linkage between protein side chains
and phosphoric acid residues of the nucleotides.

Fluxional methods have shown that nucleic acid occurs as large
highly asymmetric particles of columnar shape. From polarization
optics it is known that in these columnar particles the nucleotide resi-
dues are oriented with long axes perpendicular to the axis of the col-
umns. The picture resembles stacks of coins. In each stack the coins are
bonded to each other near the periphery of the stack. Unfortunately, no
further detail can be furnished.

54 FRANCIS O. SCHMITT

The structure and composition of virus nucleoproteins, particularly
tobacco mosaic virus, has been closely studied. In this field the x-ray
work of Bernal and Fankuchen and the chemical studies of Stanley and
his associates are classical, but the published data do not yet permit a
detailed description of the structure of the protein and nucleic acid
constituents.

The Avery experiment, in which nucleic acid extracted from pneu-
mococci has been likened to a free gene, has further heightened the
interest in these substances. It is urgent, therefore, that the very meager
information about the structure of nucleic acids and nucleoproteins be
extended by further x-ray examination. Since, to the author's knowl-
edge, no electron microscope studies have thus far been reported, the
possibilities in this direction cannot be assessed.

The Continuous Polypeptide Chain Theory

versus the
Corpuscular Theory of Protein Structure

From the classical work of Emil Fischer it was natural to consider
the polypeptide chain as the unit of protein structure. This view was
adopted by those using x-ray methods, especially by Astbury. By stretch-
ing an alpha keratin fiber one was literally pulling the folded polypeptide
chains into the straight, beta configuration. In addition it was supposed
that these continuous polypeptide chains occur in well ordered semi-
crystalline regions and in poorly ordered amorphous regions; poly-
peptide chains coursing through both types of regions produce an anas-
tomotic system.

Recent experiments of Bear and his associates, in which x-ray patterns
are obtained from specimens which are tilted at various angles in the
x-ray beam, support the view that collagen is essentially a one-dimen-
sional structure. By this is meant that the linear elements have a thick-
ness of only a few polypeptide chains and have indefinite extension
longitudinally. Electron microscope evidence suggests that these units
associate with each other laterally so that their axial periodicities are
in phase, thus causing the entire fibril to appear cross-striated. It is even
possible to dissociate these units by treating rat tail collagen with acetic
acid, and to cause them to join up again, on elevation of the pH, so that
cross-striated fibrils are again formed. In the case of collagen, therefore,
the polypeptide chain theory seems most attractive at the present time.

Myosin, paramyosin and fibrin show wide-angle x-ray patterns

MOLECULAR MORPHOLOGY 55

similar to that of alpha keratin, from which it may be concluded that
some similarity of structure of the polypeptide units exists. However,
these proteins differ from keratin, at least superficially, in that they may
be composed of particulate units. In such cases the continuous polypep-
tide chain description requires considerable qualification. Some examples
of this class may be mentioned.

Fibrinogen molecules have been well characterized with respect to
size and shape. They are long, thin spindle-shaped particles. The chemi-
cal change which occurs when these units combine to form fibrin fibers
is very slight, making it improbable that the internal architecture suffers
much alteration in the process. In this case the aggregation of the
particles is not readily reversible.

Globular insulin molecules, which have relatively slight asymmetry,
may be converted into a fibrous modification without denaturation in
the strict sense of the term. The process is completely reversible, as
Waugh has shown. This is a particularly strategic case because much
is known about the composition and structure of the insulin molecule.
This should greatly aid in investigating the alterations which occur
when the globular molecules are recruited into fibrous arrays. Waugh
has pointed out that the phenomenon requires the assumption that
combination occurs at opposite poles of the molecule, otherwise only
random gel-like aggregation, rather than filament formation, would
result. It seems probable that similar considerations may apply to certain
other particle-fiber transformations.

Finally the remarkable properties of the muscle protein, actin, may
be mentioned. In distilled water these particles or molecules occur freely
dissociated. The solution has relatively low streaming double refrac-
tion and viscosity. Merely increasing the ionic strength greatly increases
the double refraction and viscosity due presumably to fibrous aggrega-
tion of the particles. Jakus and Hall have studied the phenomenon with
the electron microscope and find that the filaments have widths of the
order of lOO A and lengths up to several microns. Reduction of ionic
strength causes the filaments to break up. The recent work of Szent-
Gyorgyi, Cori, Price and others suggests that this type of reversible
aggregation may be involved in muscle contraction.

The Mechanism of Fibrogenesis

The points just discussed bear importantly on the chemical and
biological mechanisms of fibrogenesis.

56 FRANCIS O. SCHMITT

In the case of fibers constructed of globular units it is easy to picture
possible intracellular processes. Given a sufficient concentration of the
appropriate globular protein molecules within the cell, their aggregation
or disaggregation would depend on the local environment with respect
to salts or activating enzymes. The actin-myosin cycle seems particularly
attractive as a prototype. In this case the aggregation or disaggregation
would depend on the presence of free adenosinetriphosphate (ATP) or
similarly acting substance. The concentration of the latter depends, in
turn, on the presence of ATP-ase and on the coupled reactions of inter-
mediary metabolism. Thus would be exemplified the close relationship
between the energy yielding reactions and structure formation in cells.

Little vision is required to suggest such possible mechanisms, but
much experimentation is needed to isolate the proteins of cell structure
and to determine their physical and chemical properties. Especially
significant for the problem of growth is the fibrous system developed
during the mitotic cycle. Existing theories of the formation of the
spindle and astral fibers are well-nigh useless. It is necessary to tackle
this problem with the techniques of modern protein and enzyme chemis-
try.

Little can be said about the mechanism of the formation of fibers of
the collagen class, the components of which consist essentially of one-
dimensional arrays. The fibers are presumably formed extracellularly.
Do the cells produce a precursor, a procollagen, which is converted into
collagen under the influence of the intercellular environment? Or do the
fibroblasts produce the one-dimensional arrays of collagen which then
pass into the extracellular space and there form fibers in a manner
analogous to the reconstitution of fibers from an acetic acid solution of
collagen by the addition of salt or elevation of pH ?

To these questions there are as yet no satisfactory answers. Collagen
is a good case to investigate, for the fibrils can be identified in the elec-
tron microscope.

Comparative Macromolecular Morphology

Reference has been made to a classification of the fibrous proteins
according to two types, the collagen group and the group showing the
characteristic alpha x-ray pattern, including keratin, myosin, fibrin, and
paramyosin. Astbury has stressed the comparative aspect and has sug-
gested that nature employs only a few general patterns of structure with
major and minor variations on the main theme.

MOLECULAR MORPHOLOGY 57

Aside from the value of comparative studies in suggesting clues
about the structure of fibrous proteins, the matter has a biological
significance as well. Thus a detailed comparison of the structure and
composition of a particular protein, such as keratin, in the various
phyla and classes may reveal relationships quite as fundamental as those
of comparative anatomy and embryology,

Astbury and Rudall have begun such a study of the keratins.
"Feather" keratin, which is of the beta variety, is found only in reptiles
and birds ; in other vertebrates the keratin is of the alpha type. In reptiles
the mobile regions of the body are covered with the softer alpha keratin,
while the remainder of the body is enclosed in the harder beta variety.
Unfortunately the comparative studies on keratin have been made
chiefly on the basis of wide-angle x-ray patterns.

Bear and his associates have been making a careful comparative
x-ray study of the collagen type. All vertebrate collagens examined show
the same characteristic wide-angle pattern as do also many invertebrate
collagens, including those of representative mollusks, annelids, and
echinoderms. Recently these investigators have found a similar pattern
in gorgonin and spongin, thus furnishing evidence for the existence of
this class of proteins in the coelenterates and porifera. The meager data
that are available on the amino acid composition of these proteins in
the lower forms indicate significant differences from that of vertebrate
collagen. Bear's studies of the fiber-axis long spacings have added
greatly to the detailed information necessary for comparative studies.
It remains to be seen whether, with increasing phylogenetic complexity,
this ubiquitous protein type also increases in regularity of molecular
architecture or whether, even in the lowest forms, it exists already fully
established. The shaft membrane of the trichocysts of Paramecium
possesses a structure fully as regular as that of vertebrate collagen, to
which it bears some resemblance in chemical properties, as shown by
Jakus.

Will it turn out that the lowest organisms contain protein patterns as
complex as those of the higher organisms ; that such complexity of
molecular structure is required for life as we know it? If so, what are the
missing links between such structural proteins and the simpler organic
compounds ?

Macromolecular Ontogeny

The general problem of the chemical mechanism of the synthesis of
proteins has been considered previously in this symposium. The prob-

58 FRANCIS O. SCHMITT

lem may also be considered from the biological point of view. When in
the embryological sequence are the various structural proteins elabo-
rated? The embryologist can trace the steps in the differentiation of
the nervous system, muscles, connective tissue and so on, from the
earliest anlagen. But characteristic nerve fibers, myofibrils, and collagen
fibers are differentiated considerably after the tissue is determined.
When in these processes are the neuronin of nerve fibers, the myosin
and actin of muscle fibers, and the collagen of connective tissue elabo-
rated ? Are these fibrous proteins produced full blown or are they formed
from chemical "anlagen" or precursors? In other words, is there an
ontogeny of the structural proteins or are they formed under the in-
fluence of templates preexisting in the embryonic cells, possibly even
in the egg itself?

These are large questions which have been mostly academic in the
past because the means did not exist by which they might be answered,
but it is now possible to make a beginning in such investigations. For
example, collagen fibrils can now be readily recognized in the electron
microscope. It should be possible therefore to examine embryonic
material at various stages of development to determine when and where
collagen fibrils first make their appearance.

The technical aspects of the problem are by no means insurmountable
and will no doubt yield if a determined effort is made. The possibility
of laying the foundation for a new theoretical biology based on data
obtained at the molecular level is worthy of the best collaborative efforts
of biophysicists, biochemists and biologists. As the answers to these
problems are found, a clearer knowledge of the phenomena of growth,
normal and abnormal, must inevitably follow.

Molecular Ecology

The interaction of structural components with their environment is
at the basis of general and developmental physiology. Such interactions
depend not only on individual groups, such as particular amino acid side
chains, but on the molecular reacting system as a whole. Molecular
morphology should be studied not independent of or even consecutively
with the dynamic aspects, but simultaneously where possible. This
amounts to a study of molecular ecology, a term which I borrow from
Dr. Paul Weiss.

In a sense this is implied in what has been said before. Thus certain
fibrous proteins react spectacularly to changes in their environment.

MOLECULAR MORPHOLOGY 59

Fibrinogen molecules are recruited into fibrin fibers in the presence of
small amounts of thrombin. The actin-myosin system of muscle responds
to changes of potassium ion concentration or to ATP in a dramatic
manner. Such processes are readily appreciated in the light of the rich
physiological background in blood clotting and muscle contraction.

But the general concepts of the lability of macromolecular complexes,
their sensitivity to their chemical environment, is perhaps not fully
appreciated by students of cellular physiology. The heritage of classical
morphology conditions the investigator to static rather than to dynamic
concepts.

Many examples might be cited to illustrate the point. Investigators of
permeability phenomena in searching for the structure of the plasma
membrane are prone to regard this organelle as a fixed mosaic of
protein and lipid components. The small amount of evidence at hand
indicates that the protein component consists of a feltwork of very thin
protein filaments. Is it not possible that such proteins may be susceptible
to changes in ionic environment, to enzymes or to materials like ATP,
as is the case with the fibrous muscle proteins, and that such changes
may result in alterations of the protein which may profoundly affect
permeability properties? The protein of the erythrocyte envelope,
stromatin, may be isolated and studied for such properties. Perhaps this
would be a favorable system in which to test the suggested hypothesis.

The case of neurofibrils may also be mentioned. Clearly demonstrable
by suitable cytological techniques, these structures have been regarded
by physiologists as mere fixation artifacts. Polarization optics, however,
has demonstrated the existence of longitudinal submicroscopic strands
which form the framework of neurofibrils. This molecular lattice is
extremely sensitive to its chemical environment. In what ways is this
sensitivity involved in the propagation of the nerve impulse? In the
normal maintenance, growth, and repair of the neuron? These questions
cannot be answered until the neuronin complex is better characterized
biochemically and structurally.

The chromosome is of course the final and best illustration of the
importance of macromolecular ecology. On this structure is imposed
the task not only of maintaining through countless duplications a rigid
structural integrity — a process which physicists like Schrodinger find
difficult to understand — but also of presiding over the multitudinous
determinative processes of growth and development. The chromosome
is capable of a determinative role not only by virtue of its specific
intrinsic structure but also because of the very particular chemical

6o FRANCIS O. SCHMITT

environment provided by the orderly action of embryological organizers
upon the reacting system as a whole. Further progress requires a frontal
attack on the molecular structure of the chromosome with all the physical
and chemical tools at our disposal. But the investigator who ventures
in this difficult field must have a proper appreciation of the indetermi-
nacies involved. These stem largely from the fact that chromatin re-
moved from a nucleus for analytical study may be but a skeleton of the
biological entity as it exists in the living cell. This should not deter him
from making structural and chemical studies on chromosomal material
isolated in whatever form is most convenient for the purpose, but the
complexities imposed by the ecological factors should warn against
oversimplified generalizations which might be derived from such partial
systems.

III. PLANT GROWTH HORMONES

BY KENNETH V. THIMANN^

THE field of plant growth hormones, or auxins, has broadened
out so enormously in recent years that I shall be able here to
deal only with certain parts of it. The various applications to
agriculture and horticulture, for instance, I will scarcely touch on, not
because they are unimportant but because it seems preferable to center
attention on the basic phenomenon of growth itself and how it is
influenced by the hormones in plants.

One of the most striking features of the auxins is the multiplicity of
their actions. Although they were discovered through their growth
promoting effect on the Avena coleoptile, especially when applied to one
side, and have been assayed in this manner ever since, it was soon made
clear that they promote straight growth in many elongating stems and
organs, i.e. enlargement along the longitudinal axis. Also, when such
organs are slit and placed in a solution of an auxin they undergo dif-
ferential growth due to greater elongation of the outer than the inner
tissues. The auxins cause formation of roots on stems, they cause cell
division in the cambium and the formation of callus at cut surfaces;
on the other hand they inhibit the growth of buds, and they also inhibit
both the elongation of roots and the formation of the abscission layer
on leafstalks and fruitstalks. Further, they trigger in some way the
swelling of the ovary into a fruit, as pollination does, so that by applying
auxin in the absence of pollination a normal but seedless fruit is formed.
Finally, in pineapple but not (so far as known) in other plants, they
even cause flowering.

This multiple action is in strong contrast with that of some of the
animal hormones which typically cause specific effects on specifically
reactive tissues. The multiple action has led to the view, expressed
especially by Went (1938) and more recently by Gautheret (1944) - that
auxin acts as a mobilizing agent, bringing specific hormones to the
tissue at which it is applied. According to this view, root formation when
auxin is applied to the base of a cutting would be due to the mobiliza-
tion there of a specific root-forming hormone. Experiments of Went, of
Bouillenne, and of Cooper have given indirect support to this view,
but the support is not strong and there is evidence against it. For in-
stance, the rooting of small cubes of potato tubers when given high

1 Department of Biology, Harvard University.

62 KENNETHV.THIMANN

auxin concentrations, or the rooting of fragments of Helianthus hypo-
cotyls in tissue culture containing auxin, are difficult to explain on this
basis. The influence of leaves in promoting the rooting of cuttings after
auxin treatment, brought out especially clearly by Rappaport, was at
first thought to be due to the contribution of the specific hormone by
the leaves, but in the light of several more recent studies this effect is
almost certainly nutritional. In at least one plant which is very hard to
root with auxin treatment, the White Hibiscus studied by van Overbeek
and coworkers (1946), the entire effect of the leaves could be replaced
by nutrients, leaving no evidence for a hormonal factor.

Similarly the inhibition of lateral bud development by auxin applied
at the tip was explained by the mobilization of bud growth factors at
the point of auxin application, thus starving the buds of their essential
factors. But inhibition has been shown to occur when auxin is applied
to the lateral buds themselves ; and although its mechanism is certainly
not clear, it does seem to be essentially a direct effect of auxin on the
lateral buds rather than a diversion of other hormones away from these
buds. It is, however, possible in this case that the action is due to the
intermediate formation of an inhibiting substance by the auxin. Then
again, the inhibition of root growth, a phenomenon exhibited by all
auxins and in very low concentrations, is observed on isolated roots
and is definitely exerted on the root itself.

All these considerations very strongly support the idea which was
tentatively put forward more than ten years ago (Thimann, IQ35) that
auxin controls some fundamental master reaction in the cell. From this
reaction, growth in size dififerentation, or even inhibition may follow
according to the abilities of the cells concerned and their supply of
water and other factors. It remains true that during the occurrence of
the actual growth process there is some mobilization of materials, i.e.
of water and nutrients. But this is a part of the growth process itself,
rather than its primary cause. Its part will be referred to later.

Before dealing with this fundamental master reaction and the evi-
dence for it, it will be worthwhile to discuss one of the fields which has
been most active in the last few years, namely the relation between free
and hound auxin.

In many standard test objects growth is a simple function of the
auxin applied, otherwise bio-assay would not be possible. From this it is
but a short step to the concept that in the intact plant, or rather in a
part of an intact plant, growth is a simple function of the auxin made
available to that part. Of course nutrient materials have to be available

PLANT GROWTH HORMONES 63

too. This simple view had good success in explaining the facts so long as
the auxin was determined by diffusion out into agar. The two-factor
scheme given by Went and Thimann (1937) for many seedlings, and
founded on diffusion experiments, worked very well. Perhaps the best
case is furnished by Dolk's well-known study of geotropism in the
coleoptile, where the excess growth of the lower side, which causes the
geotropic curvature, was accounted for by the demonstration that 65%
of the auxin (instead of 50% when vertical) went to the lower side.
Many other organs have given essentially the same results.

The introduction of a new method, i.e. extraction of the auxin from
the plant tissue with an organic solvent, has gradually undermined all
this. The very first such experiments, on Avena, showed auxin all
through the plant, including the parts which are no longer growing.
This destroys the simple correlation between auxin and growth and

TABLE I

Auxin in Spinach Leaf Fractions

All samples extracted with ether 6 weeks at 2° after the treatment and
subsequent acidification. Figures are units of auxin per mg. dry weight.

FRACTION

TREATMENT

Incubated 24 hours at 37° :

Boiled at

None

at/)H8

at/jH 10.5

at/>H 10

with
chymo-
trypsin

pYi 10

Chloroplasts

0.21

3-54

0.07

9-4

0.16

Globulin I plus II

0.54

2.00

0.22

21.0

0.26

Albumin

2.43

0.80

I.I

0.50

introduces a distinction between free auxin, which diffuses readily into
agar, and bound auxin, which does not. As Went has shown (1942), it
is only the former which is redistributed under the influence of gravity.
Extension of the extraction method to other material has raised many
problems. Let us consider first the case of green leafy tissue. In Lemna,
for instance, auxin is set free into ether continuously but very slowly.
From dried material no liberation occurs, but on wetting it re-starts.
The auxin is apparently being set free by hydrolysis (Thimann, Skoog,
and Byer, 1942). Proteolytic enzymes, especially chymotrypsin, liberate
the auxin rapidly, so that we have here an auxin-protein. Using chymo-

64 KENNETH V. THIMANN

trypsin we have attempted to locate this auxin-protein in the leaves of
spinach, but there is more than one location. Some two-thirds of the
auxin is associated with a globulin in the cytoplasm, but there is also
much auxin in thoroughly washed intact chloroplasts. The other protein
preparations do not contain large amounts. The table summarizes one
such set of extractions. Bonner and Wildman (1946), using hydrolysis
by alkali, found all the auxin in a globulin, a very interesting compound
comprising three-quarters of the cytoplasm protein and having proper-
ties, including phosphatase activity, similar to those of myosin. Alkaline
hydrolysis does not liberate the chloroplast auxin, which explains our
inability to extract auxin from Lemna by alkali. Avery and coworkers
(1945) have reported the extraction of auxin from leaves of cabbage
by boiling in alkali ; but this cannot be a general phenomenon, for Miss
Mes, with tomato leaves (unpublished work), finds as we do with
spinach and Lemn-a that there is no liberation of auxin by alkali.

Very little auxin is liberated by simple short extraction of leaves
with ether, i.e. very little is free in the leaf. If this relatively large store
of auxin in protein form were available to the plant for growth, it
would be expected that leaves would continue growing much longer.
One might expect giant cells and other evidences of high auxin con-
centration ; but, further, one would expect to find the auxin very readily
set free under natural conditions. Nevertheless many extensive experi-
ments which we performed with different kinds of autolysis or action
of autogenous enzymes succeeded only in showing a very small libera-
tion ; no large fraction of the auxin was ever obtained even after many
days' autolysis.

A measure of the free auxin in the tissue is given by the results of
short-time extraction with ether, or in some cases, extraction of boiled
tissue with ether. Van Overbeek et al. (1945) have given evidence that
the free auxin is essentially extracted in two hours. From our own data,
based on twenty-four hours' extraction, the following figures may be
given for the percentage of the total auxin given up in this way, i.e. the
presumptive "free" auxin :

Fronds of Lemna

2%

The alga Ulva

2%

Avena roots

25%

Bean nodules

87%

Avena coleoptiles

Almost 100%

The contrast between the two extremes is very striking, for the auxin in

PLANT GROWTH HORMONES 65

nodules and coleoptiles is very largely in the free state while that in the
green tissue is overwhelmingly bound. It is difficult to resist the con-
clusion that this protein-bound auxin of leaves is not a reserve and is
(mainly at least) unavailable to the plant for growth.

Seeds, however, show a very different picture. When dry seeds of
Avena or of corn are dampened, auxin is very rapidly set free. Heating
with alkali liberates auxin very rapidly in this material, as shown by
Avery and coworkers (1941). It is now clear that this auxin, too, is
protein-bound, and Gordon (1946) has prepared pure proteins from
wheat which liberate auxin on hydrolysis with alkali or acid. We know
that seeds deliver to the seedling a precursor which the seedling tip con-
verts to auxin. While this mobile precursor, partially purified by Berger
and Avery (1944), does not itself seem to be a protein, it is clear that
the auxin-protein of seeds is a true reserve or auxin-source, available
to the plant for growth.

A word of warning is in place here. Tryptophane on heating with
alkali forms a small amount of indoleacetic acid; casein and other
proteins do the same. This phenomenon probably cannot account quan-
titatively for the auxin liberated by alkali from these seed proteins, but
it certainly enters in. The fact that some seedlings are able to convert
both tryptophane and tryptamine to free auxin should be remembered
also. The demonstration by Haagen Smit, and later by Avery et al., and
by Larsen and van Overbeek, that indoleacetic acid is no "heteroauxin"
but a true plant auxin is of importance here.

Besides the above instances, there is still another type of behavior.
When grass-stalks are laid horizontal, the nodes, which have long
ceased growing, begin to grow on the lower side and this produces geo-
tropic curvature of the stalks. Many years ago Schmitz showed by
diffusion that this is due to the new production of auxin in these nodes ;
here gravity, instead of merely redistributing the auxin, as in the coleop-
tile, causes renewed production of auxin. Now van Overbeek et al.
(1945) have shown in sugar cane that this new production is at the
expense of bound auxin in the node ; gravity in some way initiates the
conversion of the bound to the free form. This phenomenon of gravi-
tational release of auxin may be connected with another curious phenom-
enon. In pineapple (but not in other plants) application of auxin to
the growing tip causes formation of flowers. Van Overbeek and his
coworkers (unpublished) have now found that placing the pineapple
plant horizontal has the same effect. This remarkable result may be
due to liberation of bound auxin, but could of course be merely the result

66 KENNETH V. THIMANN

of geotropic accumulation of enough extra auxin on one side to start
the flowering process.

In summary we may conclude that growth depends not only on the
free auxin present but also on the ability of the tissue to convert bound
auxin into the free form. The bound auxin is itself of many types : ( i )
that of the seed and perhaps the ovary which is made available for
growth, (2) that of the leaf which is not, and (3) that of the main axis,
which may occupy an intermediate position, while (4) that in the
coleoptile appears much more loosely bound than the types discussed
above.

I come now to the mechanism of the action of auxin. One of the
greatest values of our knowledge of the auxins is that they provide a
tool for the analysis of the growth process itself. I will try to outline
briefly the progress made in this direction, but first I want to make the
point that all this evidence that auxin can combine with proteins, and in
many different ways, has another significance. From studies of the
structure of the many substances active as auxins, Koepfli, Went and I
have earlier deduced (1938) that there is a certain spatial relationship
between the various parts of the molecule which is essential for activity.
This suggests that the auxin has to combine with another molecule
whose spatial structure is fixed, such as a protein, for example. This
leads naturally to the idea that the auxin is a coenzyme for some enzyme
system which controls growth.

It has been known for a long time that growth of the coleoptile will
not take place anaerobically, and Bonner showed that growth is inhibited
by cyanide to the same degree as respiration is. This suggests that
growth is in some way linked to respiration, but it does not provide a
mechanism for studying it. However, Commoner and I (194O sub-
sequently found that various known inhibitors of dehydrogenase sys-
tems also inhibit growth. This led to a study of the whole relationship
between enzyme inhibitors and growth.

The technique was the straight growth of isolated coleoptile sections,
which grow little in water or sugar alone, more in auxin, and consider-
ably more in the combination of auxin and sugar. Growth was in
shallow layers of solution by placing the sections on the teeth of combs.

Of all the inhibitors studied, iodoacetate was one of the most effec-
tive, giving complete inhibition at about 3.10"^ molal, an exceedingly
low concentration for most enzyme inhibitions. Under these conditions
the respiration of the sections was very little reduced. We have here not
an effect on the general vitality or integrity of the cells, but specifically

PLANT GROWTH HORMONES 6/

on the growth process itself. Furthermore this inhibition of growth is
removed by malate and other four-carbon acids, and also by pyruvate
and (as we have found later) isocitrate. Also, the growth of the unin-
hibited sections in auxin and sugar is clearly promoted by four-carbon
acids after some hours. It is clear then that growth is not a function of
respiration as a whole but of some specific part of it, i.e. of a respiratory
system or at least a hydrogen-transferring system, involving those
organic acids which participate in the Krebs cycle.

Confirmation of this came from the study of protoplasmic streaming
in the coleoptile. Mrs. Sweeney and I (1938, 1942) studied this in the
epidermal cells, at first by direct observation with a stop watch and later
with a special device which provides an artificial comparison stream
whose rate can be adjusted to synchronize with that of the protoplasm
particles. The rate of streaming is accelerated by auxin. The reaction is
closely related to growth because :

(a) it takes place before visible growth occurs, being clear within
a few minutes,

(b) oxygen is required, as in growth,

(c) sugar is needed for the eflfect to last, just as in growth,

(d) the dependence on auxin concentration is parallel to that of
growth,

(e) the increased rate is prevented by iodoacetate,

(f) the auxin effect is promoted by malate, especially in old or
starved coleoptiles, whose sensitivity to auxin has been greatly
reduced.

The parallel is thus very complete ; the effect of auxin on streaming is
almost undoubtedly a part of its effect on growth.

If acceleration of growth and acceleration of streaming are due to a
dehydrogenation process, the effect of auxin on respiration should be
interesting.

All investigators are agreed that auxin does not raise the oxygen
consumption of coleoptile tissue in buffer solution ; but we found that in
sugar (sucrose and fructose) after a few hours' soaking, auxin does
produce a rise of some 10% in oxygen consumption. Furthermore, if
the tissue is soaked in malate or fumarate, the rise is larger (about
22%). This effect of auxin on respiration in the presence of malate
closely parallels the effect on growth.

The effect of auxin on respiration in presence of sugar has recently
been confirmed by Avery and Berger, though not however the influence

68 KENNETH V. THIMANN

of malate, which evidently needs further study to elucidate the con-
ditions under which it occurs.

The inhibition of growth by iodoacetate and its promotion by four-
carbon acids has been confirmed also for the growth of intact seedlings
by Albaum and coworkers (1941, 1943)- They found certain differ-
ences in the time relations, however, which may be due to the use of
whole seedlings with accompanying complications of transport, inactiva-
tion in the plant, and varying uptake by the roots, or may be due as they
suggest to changes in the enzyme systems with age of the seedlings.

Recently Bonner and Wildman have reported (1946) that fluoride
inhibits growth, and like iodoacetate the inhibition can be exerted with-
out any marked reduction in total respiration. They were led to this by
their finding (mentioned above) that the auxin-protein in the cytoplasm
of leaves contains a phosphatase and no other enzyme — a very sugges-
tive fact. Fluoride and iodoacetate both act on the phosphorylating break-
down of glucose, in fact on directly subsequent reactions, though it is
true that in the extracted systems so far studied they act only in higher
concentrations than are effective on plant growth. We know from the
work of James and his coworkers at Oxford that sugar breakdown in
leaves follows closely (though not identically) the chain of reactions
worked out in yeast and elsewhere. It is tempting therefore to postulate
that both interfere with growth through the sugar-phosphorylating
chain, and this would then be the system that delivers energy for growth.
The four-carbon acids might participate by allowing a phosphorylation
to occur at the expense of oxidizing malate, etc., as described for suc-
cinic, pyruvic, and ketoglutaric acids by Colowick et al. (1940, 1941)
and by Ochoa (1943, 1944).

In attempting to get some evidence on this we have carefully com-
pared the action of these two poisons on growth. Some time ago I
noticed that iodoacetate, although a growth inhibitor, increases the
auxin curvatures in the pea test, although it inhibits straight growth of
the same pea seedling internodes at the same cencentration. The curva-
ture may be increased by 50% while growth is decreased by some 50%.
This characteristic is shown only very slightly, if at all, by fluoride. The
highest nontoxic concentrations of both give opposite effects. Also, W.
Bonner and I have found that the inhibition by fluoride is not released
by the organic acids, though it can be released by certain hexose phos-
phates. Our tentative conclusion is that iodoacetate and phosphate
inhibit growth in different ways or through different systems. The one

PLANT GROWTH HORMONES 69

affected by iodoacetate is presumably of sulfhydryl nature, and there is
other evidence to support this.

At this point it is useful to digress for a moment to consider other
organisms. If auxins influence growth through such fundamental sys-
tems as the Krebs cycle and the phosphorylation of hexose, we should
not expect their action to be limited to higher plants. It is important in
this connection that the long disputed question whether they control the
growth of algae has recently been answered. Algeus (1946) has shown
that indoleacetic acid does in fact promote cell division in green unicellu-
lar forms and has given an explanation of why up to now the results
have been conflicting. One major factor is the strong influence of pK
on entry of auxin acids into the cell, and the other is that the light
necessary for continued growth of the algal cultures produces decom-
position products of the auxin which are toxic. If these points are
recognized and controlled, clear-cut growth promotion is found. Also
Dr. Dubos has informed me that there are in some bacteria very marked
effects of auxin on growth. This, together with some new evidence
(unpublished) about an influence of auxin on muscle tissue in animals,
justifies the view that the action is exerted via systems which are of
quite general occurrence.

Lastly we must ask how this purely respiratory process can bring
about growth. In the last analysis growth in plants means the intake of
water. One way in which this may be brought about is through an effect
on the accumulation of solutes. We know from the work of the Califor-
nia school that solute accumulation requires oxygen and consumes sugar
and that active protoplasmic streaming is characteristic of cells which
can accumulate. Following the demonstration by Reinders (1942) that
auxin promotes water uptake by potato slices. Commoner and coworkers
(1943) have shown that auxin even causes the uptake of water from a
solution which is hyper-tonic, i.e. a solution which without auxin would
cause wilting. This intake of water closely parallels other effects of
auxin on growth, and it is promoted by potassium ions in much the
same way as we earlier found the growth of coleoptiles in ordinary
auxin solutions to be increased by low concentrations of KCl.

Very tentatively, therefore, one may propose the following steps for
the simplest kind of growth of isolated plant parts immersed in solu-
tions :

I , Auxin enters the cells with great rapidity, accelerating the rate
of protoplasmic streaming.

70 KENNETH V. THIMANN

2. This acceleration causes increased intake of solutes.

3. The energy for the accumulation of solutes and for the stream-
ing is furnished from a sugar breakdown which involves
phosphorylation and in which the Krebs cycle participates ; one
of the key enzymes here is of the sulfhydryl type.

4. The intake of solutes is at once accompanied by intake of water ;
growing cells do not have any higher osmotic pressure than
non-growing ones.

5. The acceleration of streaming causes also a difference in the
rate of wall deposition, i.e. a change in plasticity.

6. The increased water intake coupled with the change in wall
properties means increased cell size — in other words, growth.

In this way one can begin to see how auxin can have the multiple
effects mentioned at the outset. For a cell may not be in a condition to
enlarge directly, due either to secondary wall formation or to its un-
favorable situation in other tissues, yet the increased streaming rate and
increased uptake of solutes may in some way activate numerous proc-
esses, causing cell division and leading to root initials and the like.
Such a tissue would accumulate organic solutes too, thus providing some
basis for the mobilization phenomena which, as was stated above, are
probably secondary and not causal.

This scheme is still largely hypothetical but there is enough circum-
stantial evidence to make one believe that it is at least somewhere near

the truth.

Note Added in Proof

The critical role of a sulfhydryl enzyme in growth has been confirmed (Thimann,
K. v., and W. D. Bonner, Jr., Amer. J. Bot., 35, 271-281, 1948; and in press). The con-
version of tryptophane to auxin has been clarified (Wildman, S. G., M. Ferri, and J.
Bonner, Arch. Biochem., 15, 131-144, 1947) and shown to occur in the living leaf. In
spite of these and many other recent results, the exact biochemical role of auxin in
growth remains elusive.

REFERENCES

Albaum, H. G., and B. Commoner. Biol. Bull., 80, 314-323. I94i-
Albaum, H. G., and B. Eichel. Amer. J. Bot., 30, 18-23, 1943.
Algeus, S. Bot. Notiser (Lund), 129-278, 1946.

Avery, G. S., J. Berger, and B. Shalucha. Amer. J. Bot., 28, 596-607, 1941.
Averyi G. S., J. Berger, and R. O. White. Amer. J. Bot., 32, 188-191, 1945-
Berger, J., and G. S. Avery, Jr. Amer. J. Bot., 31, 11-19, 199-203, I944-
Bonner, J., and S. Wildman. Grozvth, Supplement, Vol. 10, 51-68, 1946.
Colowick, S. P., M. S. Welch and C. Cori. /. Biol. Chem., 133, 359-373 J

PLANT GROWTH HORMONES 7I

641-642, 1940; Colowick, S. P., H. M. Kalckar, and C. Cori. Ibid., 137,

343-356, 1941.
Commoner, B., and K. V. Thimann. /. Gen. Physiol., 24, 279-296, 1941.
Commoner, B., S. Fogel, and W. H. Muller. Amer. J. Bot., 30, 23-28, 1943.
Gautheret, R. J. Rev. Cyt. Cytophysiol. Veg., 7, 45-185, 1944-
Gordon, S. A. Amer. J. Bot., 43, 160-169, 1946.
Koepfli, J. B., K. V. Thimann, and F. W. Went, /. Biol. Chem., 122, 763-

780, 1938.
Ochoa, S. /. Biol. Chem., 131, 493-501' ^943; i55> ^7-97, i944-
Overbeek, J. van. Ann. Rev. Biochem., 13, 631-666, 1944.
Overbeek, J. van, C. D. Olivo, and E. M. S. de Vazquez. Bot. Gaa., 106,

440-451, 1945.
Overbeek, J. van, S. A. Gordon, and L. E. Gregory. Amer. J. Bot., 33, 100-

107, 1946.
Reinders, D. E. Rec. trav. bot. Need., jp, 1-140, 1942.
Sweeney, B. M., and K. V. Thimann. /. Gen. Physiol., 21, 439-461, 1938;

25, 841-854, 1942.
Thimann, K. V., F. Skoog, and A. C. Byer. Amer. J. Bot., 29, 598-606,

1942.
Thimann, K. V. Proc. Kon. Akad. Wetenschap. (Amsterdam), 38, 896-912,

1935.
Went, F. W. Plant Physiol., 13, 55-80, 1938.
Went, F. W. Plant Physiol, 17, 236-249, 1942.
Went, F. W., and K. V. Thimann. Phytohormones. New York, Macmillan,

1937-

IV. UNIDENTIFIED VITAMINS
AND GROWTH FACTORS

BY KARL FOLKERS'

THERE are many investigations in progress which are concerned
with the elucidation of new vitamins and growth factors.
These investigations include nutritional experiments with ani-
mals, poultry, microorganisms, insects and other forms of life. Because
of the magnitude of the literature on unidentified vitamins and growth
factors, it seemed best for this review to select the results of certain
investigations for discussion.

Experience has shown that the initial phases of certain nutritional
investigations can be quite independent and diverse, and yet the final
phases may be the same.

Turning to the past, an excellent example of initially diverse nutri-
tional investigations was the independent researches on coenzyme R
which was described as a growth and respiration promoting factor for
Rhizohium trifolii, on biotin which is a yeast-growth factor, and on
vitamin H which was a factor for rat nutrition (Table I).

In the final phase of these investigations, it was concluded that these
three independent researches on the nutrition of Rhizohium trifolii,
yeast, and rats were concerned with a single entity — biotin.

Turning to the present, it is desirable to discuss first folic acid and
possible chemically related factors. These factors appear to deserve
mention first for several reasons. The chemistry of the first members of
the group to be elucidated is now becoming available. All the members
of this group are receiving much study at present in many laboratories.
These factors as a group typify again the development of independent,
but frequently converging lines of nutritional studies.

These factors were introduced into the literature by such designations
as vitamin M which is a factor for cytopenia in monkeys ; norite eluate
factor for the growth of L. casei; vitamin Be for a microcytic hyper-
chromic type of anemia in chicks ; factor R for chicks ; folic acid for the
growth of S. lactis and L. casei; liver L. casei factor; Streptococcus
lactis R factor ; vitamins Bio and Bn for chicks ; fermentation L. casei
factor; and the vitamin Be conjugate. The early papers on these factors
(Table II) appeared during 1938-1945.

1 Research Laboratories, Merck & Co., Inc., Rahway, N.J.

UNIDENTIFIED VITAMINS

71

TABLE I
Coenzyme R, Biotin and Vitamin H

Coenzyme R

(growth and respiration promoting effects
of extracts for Rhicobimn trifolii; Allison,
Hoover and Burk [1933])

Biotin

(yeast-growth factor;

K5gl [1935])

Vitamin H

(rat nutrition
Gyorgy [1939])

CH.. CH(CH04CO,H

Biotin

TABLE II
Folic Acid and Possible Chemically Related Factors

Vitamin M

Norite eluate factor
Vitamin Be

Factor R

Folic acid

Liver L. casei factor
Streptococcus lactis R factor

Vitamins Bio and Bn

Fermentation L. casei factor

Vitamin Be conjugate

for cytopenia in monkeys

for growth of L. casei

for anemia (macrocytic
hyperchromie type) in
chicks

for chicks

for growth of S. lactis
and L. casei

for chicks

for chicks

Day, Langstron, and Darby

(1938)
Snell and Peterson (1939)
Hogan and Parott (i939)

Schumacher, Heuser, and

Norris (1940)
Mitchell, Snell, and Williams

(1941)
Stokstad (1943)
Keresztesy, Rickes, and

Stokes (1943)
Briggs, Luckey, Elvehjem,

and Hart (i943)
Hutchings, Stokstad, Boho-

nos, and Slobodkin (i944)
Pfiffner, Calkins, O'Dell,

Bloom, Brown, Campbell,

and Bird (i945)

74

KARL FOLKERS

During these independent researches on the nutrition of monkeys,
chicks, and microorganisms, there were indications that some of the
studies were concerned with a single entity, but were of unproven rela-
tionship during the early work.

The last few months have marked the beginning of the clarification
of the chemical structural relationships between some of the factors
which appear to be chemically related to folic acid.

Recent papers presented at the New York Academy of Sciences
and published in Science showed that the liver "L. casei factor" is
A^ [4{( [2-amino-4-hydroxy-6-pteridyl] methyl) amino}-benzoyl] glu-
tamic acid. The name pteroylglutamic acid has been suggested as a more
convenient designation for this new vitamin. These elucidations of

TABLE III
Pteroylglutamic Acid

Liver "L. casei" factor
Stokstad (1943)

N N

/\./Xnh,

\

/ NHCH. ^;^^

N

N

OH

Pteroylglutamic acid

Vitamin Be

Hogan and Parrott (1940)

Isolation, Pfiffner and coworkers (1943)

UNIDENTIFIED VITAMINS 75

structure and synthesis studies were carried out in the Ledcrle Labora-
tories and the Calco Chemical Division of American Cyanamid.

A comparison of the synthetic pteroylglutamic acid with natural vita-
min Be in the Parke-Davis Laboratories showed that these two factors
were identical (Table III).

Structural studies showed that the fermentation L. casei factor con-
sists of three glutamic acid moieties in peptide linkage and attached
through the free amino group of the tripeptide to the carboxy group of
pteroic acid. The name pteroic acid was suggested for the acid :

N N

O
II

<:

HOC< \nHCH,

N I
OH

N

The vitamin Be conjugate consists of seven glutamic acid moieties as
a heptapeptide attached similarly to pteroic acid. The nature of the
peptide linkages in the glutamic acid peptides has not been described
as yet (Table IV).

The isolation of a polypeptide containing p-aminobenzoic acid from
yeast has been described. Analytical data showed that this peptide had
a free aromatic amino group, ten-eleven L-glutamic acid moieties and
an unknown amino acid moiety in a unit of the molecule.

A functional relationship of the folic acid conjugates to this peptide
of p-aminobenzoic acid is not established but is an obvious possibility.

The possible existence of other conjugates of pteroic acid containing
two, four, five, etc., moieties of glutamic acid appears even more pos-
sible after the announcement of this yeast peptide (Table V).

It is evident that several factors which appear to be related chemi-
cally to pteroylglutamic acid remain for further chemical characteriza-
tion. These factors are the norite eluate factor, folic acid, Streptococcus
lactis R factor, vitamins Bio and Bn, factor R, and vitamin M.

It does appear possible that some of these materials might have con-
tained one or more of the established pteroic acid conjugates of glutamic
acid or conjugates with a number of glutamic acid moieties other than
3 or 7 (Table VI).

76

KARL FOLKERS

TABLE IV
Fermentation "L. Casei" Factor and Vitamin Be Conjugate

Fermentation "L. casei" Factor
Isolation (1944)
Hutchings and coworkers

O

N N

N OH

N

R =: three glutamic acid moieties

Vitamin Be Conjugate
Isolation (1945)
PfifTner and coworkers

O
II

R'C<

N N

^NHCH,

N OH

NH,

N

R'zz: seven glutamic acid moieties

There are still other factors which may be related to pteroylglutamic
acid.

In their attempts to prepare a concentrate of folic acid from by-
products of commercial liver extract, Barton-Wright, Emery, and Rob-
inson found that there were chloroform-soluble fractions which were
active for the growth of L. helveticus and ^. lactis.

Daft and Sebrell have studied the anemia and granulocytopenia
which develop in rats on a pantothenic acid-deficient diet but with added
L. casei factor. Treatment with whole dried liver was more effective
than treatment with pantothenic acid. A factor or factors in the liver
were indicated.

The significance of these observations and a possible relationship to
pteroylglutamic acid is not yet clear (Table VII).

Because of abundant evidence on the antianemic properties of pteroyl-
glutamic acid, or vitamin Be, on experimental animals, it is not sur-
prising that pteroylglutamic acid showed antianemic properties in man.

UNIDENTIFIED VITAMINS

71

TABLE V

Isolation of a Peptide of p-Aniinobenzoic Acid from Yeast
(Ratner, Blanchard and Green [ 1946 J )

H.,N ^ >CO-polypeptide

OH N

N N

lo-ii /-glutamic acid residues
I unknown amino acid

-CH..NH /'

^ CO — glutamic acid
— tripeptide

— heptapeptide .

— dipeptide
— tetrapeptide
- — pentapeptide
— etc.

. known

TABLE VI

Factors Possibly Related Chemically to Pteroylglutamic Acid

Norite eluate factor

Folic acid

Streptococcus lactis R factor

Vitamins B^q and B^
Factor R
Vitamin M

Spies and his associates reported on nine cases of macrocytic anemia
in relapse which showed a positive reticulocyte response and an increase
in the red blood cell count after treatment with pteroylglutamic acid.

Darby and Jones reported satisfactory remissions when two patients
with sprue were similarly treated, and Spies reported satisfactory re-
missions of tropical sprue.

Satisfactory remissions for several cases of Addisonian pernicious

78

KARL FOLKERS

TABLE VII
Other Factors Related (?) to Pteroylglutamic Acid

Growth factors for
L. helveticus
S. lactis

Unidentified factor (s)
in Hver efifective in
blood dyscrasias in
rats

Chloroform soluble
fractions from liver
folic acid concentra-
tion process

Rats on pantothen-
ic acid deficient diet
plus L. casei factor

Barton-Wright.

Emery and Robinson,

(1945)

Daft and Sebrell
(1945)

anemia and pernicious anemia of pregnancy have been reported by
Moore and associates and by Amill and Wright.

Other such clinical observations have now been reported. However,
as Moore and his associates pointed out, there is no clinical information
as yet as to the effect of pteroylglutamic acid on the neurological mani-
festations of pernicious anemia (Table VIII).

The use of liver therapy for the treatment of pernicious anemia was
introduced by Minot and Murphy in 1926.

Since pteroylglutamic acid has been isolated from liver, it is desirable

TABLE VIII
Pteroylglutamic Acid as an Antipernicious Anemia Substance

Macrocytic anemia in
relapse (9 patients)

Sprue (2 patients)

Tropical Sprue

Addisonian pernicious
anemia (2 patients)

Pernicious anemia of
pregnancy ( i pa-
tient)

Addisonian pernicious
anemia (6 patients)

Positive reticulocyte
response increased red
blood cell count

satisfactory remissions

satisfactory remissions
satisfactory remissions

satisfactory remissions

Spies and associates
(1945)

Darby and Jones
(1946)

Spies (1945)

Moore and associates

(1945)
Moore and associates

(1945)

satisfactory remissions Amill and Wright

(1946)

UNIDENTIFIED VITAMINS 79

to know whether the effectiveness of commercial liver preparations is
due to its content of pteroylglutamic acid. Numerous microbiological
assays have been made with L. casei for the content of pteroylglutamic
acid in liver preparations. For example, Clark reported that commercial
parenteral liver extracts contained only 0.02-3.67JU, of pteroylglutamic
acid per unit of liver extract.

It was found that one U.S. P. unit of a certain liver extract when
administered daily for ten days was equivalent to only 3.8^0, of pteroyl-
glutamic acid. A dose of 20 mg. of pteroylglutamic acid daily for ten
days is 200,000^^. Alexander reported that only 0.035 ^S- ^^ ^ ^^"^^^
fraction was required to show antipernicious anemia activity.

Furthermore, Stokstad and Jukes have recently described the failure
of vitamin Be conjugase to liberate appreciable pteroylglutamic acid
from a parenteral liver extract.

Thus it appears that the antipernicious anemia factor in liver is not
pteroylglutamic acid, although the latter has antipernicious anemia
activity (Table IX).

TABLE IX
Treatment of Pernicious Anemia by Liver Therapy

Liver therapy, Minot and Murphy (1926)

Commercial parenteral liver extracts contain 0.02-3.67 /x/unit, Clark (1945)
I U.S. P. unit liver extract daily for ten days ^ 3.8 ix pteroylglutamic acid
20 mg. pteroylglutamic acid daily for ten days ^ 200,000 p., Moore and

associates (1945)
0.035 n^g- Liver fraction active, Alexander

"vitamin Be conjugase"

Parenteral liver extract > no appreciable libera-

dried chicken pancreas tion of pteroylglutamic

acid, Stokstad and
Jukes (1946)

Castle and his associates in 1944 studied the effect of the daily ad-
ministration of various substances together with neutralized normal
human gastric juice on several cases of pernicious anemia. The data on
two typical cases are summarized in Table X. In addition to an adequate
amount of "vitamin-free" casein and all the known B vitamins, xan-
thopterin and a folic acid concentrate were also administered. It was
concluded that folic acid had no effect upon the red blood cell count or
the reticulocyte peak. However, when beef muscle was used as a sup-

8o

KARL FOLKERS

TABLE X
Extrinsic Factor in Pernicious Anemia

Effect of daily administration of various substances together with 150
cc. of neutralized normal human gastric juice (Castle and associates, 1944)

Substances

Case 99

Case 100

First Period, 10 days

"Vitamin free" casein

50 gms. daily

50 gms. daily

all known B vitamins

adequate

adequate

xanthopterin

0.009

0.005

folic acid

0.0036

0.0023

Initial R.B.C. (loVcu. mm.)

1. 17

1.50

Reticulocyte peak

1.9

I.O

Interpretation

negative

negative

Second Period, 10 days

Beef muscle

200

—

Meat extract

35 cc.

Initial R.B.C.

1-54

1.79

Reticulocyte peak

13.6

8.9

Interpretation

positive

positive

plement for case 99 and a meat extract for case 100, a positive response
resulted as evidenced by a rise in the red blood cell count and the reticu-
locyte peak. It was concluded that the extrinsic factor for pernicious
anemia was not identical with folic acid, xanthopterin, or any of the
known B vitamins.

Much effort has been devoted to the isolation of the antipernicious
anemia factor in liver. The factor may be multiple. The properties of
the concentrates of this factor as studied by five groups of investigators
are given in Table XI. The properties of the concentrates differ con-
siderably. For example, the carbon content varied from 41 to 53 % , and
the nitrogen content varied from 10 to 15%. The molecular weight of
one preparation was believed to be in the range of 2000-5000. Two of
the materials were reported to have amino nitrogen, but none was found
in the material of another group. The fraction of Dakin and his asso-

UNIDENTIFIED VITAMINS

TABLE XI

8i

Properties of Certain Liver Fractions Containing Antipernicious Anemia

Factor

Cohn,

Dakin,

Laland,

Subbarow,

Karrer,

etal.

etal.

etal.

etal.

etal.

Composition

%C

—

46.8-48.1

53-64

41.56

45.68

%H

—

6.6-6.8

6.85

6.74

6.75

%N

10.8

15.2-16.8

^■33

1313

14.63

%S

none

—

0.74

1.2

present

Ash

—

—

2.05

—

—

Molecular weight

—

2000-5000

—

—

—

Presence of pentose

none

—

present

present

none

Amino nitrogen

before hydrolysis

none

0.5

—

5

—

after hydrolysis

none

10-10.4

—

75

—

Amino acids isolated

arginine
glycine
leucine
aspartic acid

hydroxyproline

proline

ciates yielded arginine, glycine, leucine, aspartic acid, hydroxyproline,
and proline on hydrolysis.

Now that the chemistry of pteroylglutamic acid and its related con-
jugates is known, it is likely that more rapid progress will be made in
the isolation of the liver antipernicious anemia factor (s). The unknown
factor or factors for pernicious anemia in liver and other sources are
possibly the outstanding unidentified factors of today, if importance
is judged by the amount of clinical evidence on therapeutic usefulness.

Just as pteroylglutamic acid has antipernicious anemia activity,
pteroyltriglutamic acid has been reported recently by Spies to have
activity also. It is effective (Table XII), but pteroylheptaglutamic acid
(vitamin Be conjugate) has been found by Welch and his associates
to be ineffective in pernicious anemia in relapse. A possible relationship
between the utilization of pteroylglutamic acid and the extrinsic factor
is a subject for interesting speculation and study.

About a year ago Cline, Berry, and Spies announced the isolation of
a new leukopoietic factor from a liver extract which was active in
leukopenia.

82

KARL FOLKERS

TABLE XII
Activity of Folic Acid Conjugates in Pernicious Anemia

O

li

RC

N N

<

Effective
(Spies, 1946)

■\ NHCH, ^^\^
N OH

N

R — three glutamic acid moieties

Pteroyltriglutamic acid

(Fermentation L. casei factor)

O
R'C^

N N

X\/Anh,

\ NHCH,

Inefifective
(Welch and associates, 1946)

N OH

N

R =z seven glutamic acid moieties

Pteroylheptaglutamic acid
(Vitamin Be conjugate)

The liver extract was fractionated with acetone and passed over
acid-washed permutit and fractionated again with water-acetone mix-
tures (Table XIII). The crystals which were obtained appeared to be
neutral in character and showed a positive test for primary aliphatic
amino groups. The material was active in doses of 20-60 mg. for
causing an increase in the white cell count and neutrophiles when given
intravenously to malnourished patients with leukopenia. These investi-
gators believed that this product might be related to the extrinsic factor
for pernicious anemia.

During 1 936-1 941 Wulzen and Bahrs studied the growth and disease
of planarian worms when the mashed tissues of guinea pigs which
were maintained on experimental diets were fed to these worms. The
growth of the worms was a measure of a change in the mammalian

UNIDENTIFIED VITAMINS 83

TABLE XIII

Isolation of a New Leukopoietic Factor from Liver (Cline, Berry, and

Spies, 1945)

Liver extract acetone acid-washed acetone

> . > > crystals

(reticulogen) fractionation permutit water fractionation

Properties of Crystals

1. positive test primary aliphatic amino group

2. contained "ash"

3. 20-60 mg. dose for malnourished patients with leukopenia caused in-
crease in white cell count and neutrophils

tissue which resulted from changing conditions in the animal as a
whole. These investigations resulted in the discovery that raw cream
contains a factor essential for the guinea pig which differs from those
factors previously described for the guinea pig. Deficiency of this
factor caused a stiffness of the wrists and elbows of the animals and
calcium triphosphate depositions throughout the whole body (Table
XIV).

TABLE XIV

A Dietary Factor Essential for Guinea Pigs, AntistiflFness Factor

Growth and disease of planarian worms fed tissues of guinea pigs (Wulzen
and Bahrs, 1936-1941).

Raw cream contains factor for guinea pigs.

Deficiency characterized by stiffness in wrists and elbows and calcium
triphosphate depositions.

Partial isolation of factor from raw cream (van Wagtendonk and Wulzen,
1943)-

A report on the partial isolation of this factor from raw cream was
made in 1943 by van Wagtendonk and Wulzen.

It was found that a deficiency of the antistiffness factor had no effect
upon the creatine excretion. In contrast with this result, vitamin E
deficiency produced muscular dystrophy accompanied by creatinurea.

The concentration of the easily hydrolyzable phosphorus fraction

84 KARL FOLKERS

(including adenosine triphosphate) decreased and the inorganic phos-
phorus increased in the liver and kidneys during the development of
the deficiency. Changes in the adenosine phosphates may involve the
metabolism of the purines.

Deficiency of the factor resulted in a lowering of the serum phos-
phatase level.

Administration of the factor resulted in a return to normal levels for
blood phosphorus and calcium, and prevention of the deposition of col-
loidal calcium phosphate in the muscle. Prolonged administration of the
factor resulted in removal of even large deposits of calcium phosphate
from the tissues.

Significant changes in the albumin-globulin ratio of the plasma pro-
teins also took place during the deficiency.

It seems apparent that this factor has a regulatory effect upon phos-
phorus metabolism (Table XV).

The work on the isolation of the antistiffness factor showed that

TABLE XV

Physiological Studies on the Antistiffness Factor (van Wagtendonk and

coworkers, 1943-1945)

No effect on creatine excretion — vitamin E deficiency produces severe
creatinuria.

Easily hydrolyzable P fraction decreases and inorganic P fraction increases.

Low serum phosphatase level.

Administration of factor resulted in normal levels of blood P and Ca.

Administration of factor resulted in removal of calcium phosphate deposits.

Changes in albumin-globulin ratios in plasma proteins take place.

Conclusion : Factor has regulatory effect on phosphorus metabolism.

raw cream was not a satisfactory source for the successful isolation of
the factor since such large amounts of raw cream would be needed.
Fifty-five gallons of raw cream gave only 3 mg. of a concentrate of the
factor which was curative at a dosage of o.iy.

Other raw materials were tested as a source of the factor. Table XVI
contains a summary of some of the results of tests on other materials.
Potato, beef muscle, liver, and alfalfa have about one-half as much ac-
tivity as raw cream. Yeast, broccoli, soybean, and beet molasses had no

UNIDENTIFIED VITAMINS 85

TABLE XVI
Sources of the Antistiffness Factor

55 gal. raw cream >3 mg. concentrate, curative at o.iy dosage

Activity of Source Materials (van Wagtendonk and Wulzen, 1946)

Material Activity, units/mg.

Raw cream ' i

Raw potato 0.5

Baker's yeast o

Beef muscle and liver 0.5

Alfalfa 0.5

Broccoli O

Soy bean o
Beet molasses o

Cane molasses 10
Cane juice 100- 1000

activity at the levels tested. Cane molasses and cane juice were discovered
to be excellent sources of the antistiffness factor and isolation work
was started with these raw materials.

A crystalline compound was eventually isolated from crude cane
juice. The process for the isolation of the crystalline compound is
shown schematically in Table XVII. The crude cane juice showed an
activity of 100 u./g. It was first extracted with ether. Distillation of
the ether left a residue which showed one million u./g. Distribution of

TABLE XVII

Isolation of the Antistiffness Factor from Cane Juice (van Wagtendonk

and Wulzen, 1946)

Crude cane juice ether f extract ~| distillation residue distribution

(55 gal.) > \ residual \ ^ (los u./g.) ^

(100 u./g.) extraction [ solution J pet. ether and

90% menthanol

activity in distillation residue MgO MgO adsorbate distillation

pet. ether ^ (1.5 x lo^ u./g.) > solution >

solution pet. ether

and benzene

residue molecular sublimations crystalline material

(10 X iQfi u./g.) > m.p. 81.5-82°

and recrystallizations activity, 500 x 10^ u./g.

yield, 100 mg.

dosage, 0.0027

86 KARL FOLKERS

this residue between petroleum ether and 90% methanol showed that
the activity remained in the petroleum ether solution. Distillation of the
petroleum ether phase left a residue showing one and one-half million
u./g. When this residue was dissolved in a petroleum ether-benzene
mixture and treated with magnesia, it was found that many impurities
were selectively adsorbed. Removal of the solvent left a residue show-
ing ten million u./g., and two molecular sublimations and many re-
crystallizations of this material gave a crystalline product which melted
constantly at 81.5-82°. It was effective in the biological test at a dosage
of o.oo2y and showed 500 million u./g. One hundred milligrams of
this crystalline compound was obtained from 55 gallons of crude cane
juice.

We may look forward to learning more about the chemistry of this
antistiffness factor,

Woolley demonstrated in 1941 the existence of a new growth factor
for the multiplication of certain organisms of Lancefield's group A
hemolytic streptococci (strains X40, S43, C203 and 594) on a chemi-
cally defined medium. It appeared that this factor was required for the
growth of other species of bacteria such as pneumococcus D39R. The
rate of growth of yeast on the Eastcott medium was also markedly
increased (Table XVIII).

TABLE XVIII
A New Growth Factor Required by Certain Hemolytic Streptococci

Growth factor (Woolley, 1941) for: a) Lancefield's group A hemolytic streptococci

strains X40, S43, C203 and 594

b) pneumococcus D39R

c) yeast on Eastcott medium

Sources : Liver "Fraction A," yeast and rice bran extracts

Dialyzability : Liver "Fraction A" purification steps concentrates

> (dialyzable form)

(non-dialyzable form)-

C^H^OH— H^O— HCl

material
-> (dialyzable form)

"Maximal unit" : Material/cc. to produce maximal growth and acidity

Purification: Concentrates showed i m.u./iOy

Properties : Not soluble in acetone, chloroform or ethyl acetate

Not readily adsorbed by norite (non-dialyzable form)

UNIDENTIFIED VITAMINS 87

That fraction of aqueous liver extract which was insoluble in 70%
alcohol ("Fraction A"), yeast extract (Difco), and rice bran extract
(Vitab) were good sources of the active factor. It appeared that this
factor was not a known amino acid, since the organisms did not grow
even when the amount of casein hydrolyzate in the medium was in-
creased to one per cent.

It was found that the factor in liver "Fraction A" was not readily
dialyzable, but some of the purified concentrates contained the factor
in dialyzable form. Treatment of the liver "fraction A" with alcoholic
hydrochloric acid converted the non-dialyzable form of the factor into
a dialyzable form.

A "maximal unit" for the assay of this factor was defined as the
amount of material per cubic centimeter of the medium which was re-
quired to produce the maximal growth and acidity.

The purification procedures yielded concentrates which showed
I m.u./ioy. The factor was not soluble in acetone, chloroform, or ethyl
acetate, and was not readily adsorbed on norite.

Pollack and Lindner showed that a substance was present in peptone
which was necessary for the growth of L. casei, and Smith showed that
certain strains of S. lactis grew better with a substance present in
yeast and peptone. It appeared to Sprince and WooUey that their factor,
strepogenin, was probably responsible for the growth phenomena ob-
served by Pollack and Lindner for L. casei and by Smith for S. lactis.
Sprince and Woolley prepared three concentrates of strepogenin by
dissimilar methods (dialysis, alcoholic butylamine extraction, and par-
tial hydrolysis of casein, etc.). Each of the three concentrates was as-
sayed for per cent recovery of activity by testing it for the growth of
the hemolytic Streptococcus X40, L. casei and S. lactis. It may be seen
from these per cent recoveries of activity that the active substance was
recovered about equally for these three organisms. Sprince and Woolley
concluded that the same substance, or very closely related substances,
were responsible for the growth activity in each case (Table XIX).

Since Wright and Skeggs had found strepogenin-like activity for
enzymic digests of casein, Sprince and Woolley treated purified proteins
with trypsin and examined the digests for strepogenin activity by assays
with L. casei (Table XX).

The relative strepogenin contents of several such proteins showed
that many of them are much richer sources of this growth factor than
is the liver fraction L. Crystalline trypsinogen and insulin were out-
standing in having 30-40 times as much strepogenin as liver fraction L.

88 KARL FOLKERS

TABLE XIX

Relationship of the Growth Factor (Strepogenin) Required by Certain
Hemolytic Streptococci to Growth Phenomena of Other Bacteria

Growth factor (Pollack and Lindner, 1943) for L. casei. Source: Peptone
Growth factor (Smith, 1943) for strains of 5. lactis. Source: Peptone, yeast
Growth factors for L. arabinosus, diphtheria bacillus and propionic acid
bacteria (1943).

Relative Potencies of Strepogenin Concentrates for Bacteria
(Sprince and Woolley, 1944)

Fraction prepared by Recovery activity (%)

strep. X40 L. casei S. lactis

Dialysis

54

54

50

Alcoholic butylamine extraction

25

27

21

Partial hydrolysis of casein, etc.

17

19

18

It appeared that the chemical nature of strepogenin is that of an
amphoteric peptide-like compound, or it is that of a proper combina-
tion of amino acids.

In an effort to learn whether strepogenin was of importance in animal
nutrition, Woolley tested the growth response of mice to certain diets.

TABLE XX

The Occurrence of Strepogenin in Purified Proteins

Wright and Skeggs (1944) : Casein > strepogenin-like activity

Chemical nature of strepogenin : amphoteric peptide-like compound or

proper combination of amino acids

Relative strepogenin contents of proteinaceous materials (after trypsin
digestion for 20 hours), Sprince and Woolley (1945)

Material
Liver fraction L
Crystalline insulin
Crystalline trypsinogen
Crystalline chymotrypsin
Crystalline tobacco mosaic virus, etc.

Relative strepogenin content
I
40

30
16

6

UNIDENTIFIED VITAMINS 89

The test animals grew at a suboptimal rate or gained only 1.7 grams/
week. The addition of 2% casein or 0.5% crystalHne trypsinogen to
the basal ration caused restoration of the normal growth or a gain of
3.1-3.6 grams/week. The addition of a strepogenin concentrate equiva-
lent to 2% casein also caused normal growth, but the addition of hydro-
lyzed strepogenin concentrate at twice the casein level (4%) had no
effect on the suboptimal growth rate (Table XXI).

These observations and others led to the conclusion that the growth

TABLE XXI

Growth Stimulant in Proteins for Animal Nutrition
Growth Response of Mice to Certain Proteinaceous Materials

(Woolley, 1945)

Average
Number change
of in weight

Addition to basal ration % Animals grams/week

12

1.7

Casein (vitamin-free)

2

4

3-1

Strepogenin concentrate

eq. 2%
casein

8

30

Denatured crystalline trypsinogen

0-5

4

3-6

Hydrolyzed strepogenin

concentrate

eq. 4%
casein

4

1-3

promoting powers of some proteins when added to diets based on amino
acids may be attributed to the presence of strepogenin.

Womack and Rose have described the results of their experiments
which indicate that proteins contain an unidentified growth stimulant
for rats, which is necessary for the maximum increases in weight and
which may be identical to strepogenin. For example, they found that
rats on a certain amino acid diet gained about 97 grams in weight in
28 days, but when a supplement of 5% casein was added, the average
gain was 126 grams. The addition of 5% acid-hydrolyzed casein was
considerably less effective in promoting growth (Table XXII). Such
experiments indicated that this factor was present in more than seven-
teen proteins and protein-rich foods.

90 KARL FOLKERS

TABLE XXII

Unidentified Growth Stimulant in Proteins for Rats
Growth response of rats to casein preparations (Womack and Rose, 1946)

Number of Average total gain

Supplement % animals in 28 days, grams

152

97 ± 0.7

Casein

5

10

126 ± 2.5

Acid-hydrolyzed casein

5

7

105 ± 2.9

There are numerous other investigations in progress on other factors.
A partial list of these other unidentified factors and vitamins is given
in Table XXIII. No doubt some of these nutritional studies are the
frontier investigations on new factors which will gain importance in
the future.

TABLE XXIII

Partial List of Other Unidentified Factors

Growth factor for Clostridmm sporogenes Knight and Fildes (1933)

Vitamin P (permeability vitamin) Rusznyak and Szent-Gyorgyi (1936)

Factor for reproduction in chickens McGinnis, Heuser, and Norris (1944)

Feeding factor for chicks Cravens, McGibbon, and Halpren (1944)

Anemia and weight factors for the pigeon Street (1944)

Chromotrichia factor (s) for the dog Frost and Dann (1944)

Unrecognized vitamins for the pig McRoberts and Hogan (1944)

Liver factor (s) for sulfapyridine anorexia
in rats and dogs Handler (1944)

Factor for growth of Tetrahymena geleii Kidder and Dewey (1944)

Growth factor for Clostridium perfringcns Ballentine and associates (1944)

Growth factor for guinea pigs Woolley and Sprince (1945)

Unidentified factor essential for rat growth. Hartman (1946)

Chick growth factor occurring in cow
manure Rubin and Bird (1946)

Growth factor for Lactobacillus gayoni
8289 Cheldelin and Riggs C1946)

V. THE KINETICS OF GROWTH
OF MICROORGANISMS

BY C. B. VAN NIEL^

THE consensus of opinion, expressed in the preceding chapters,
seems to have been that "growth is a very complex phenome-
non." I see no reason for amending this verdict when consider-
ing the growth of microorganisms. It is true that these creatures appear,
even when studied with the best optical or electronic equipment, as
rather simple structures; and, compared with higher plants and animals,
they certainly do not seem to be as highly differentiated. Nevertheless,
they display an astonishing diversity of functions which in many re-
spects they share with the trees, oysters, butterflies, and elephants.

Generally they are found swimming or floating about in an environ-
ment containing a more or less abundant variety of chemical compounds.
From these they pick up or absorb special ones which as a result of
metabolic activities are subsequently converted into cellular constituents.
When a large enough quantity and variety of the latter have been pro-
duced, and when the individual organism has sufficiently increased in
size, it divides into halves, each one of which then repeats the process.
This continues until the food supply, or specific ingredients thereof,
gives out, or until inhibitory metabolic products begin to interfere with
normal development.

The amazing feature of growth, in a microbial culture as in living be-
ings in general, lies in the fact that it involves the manufacture and or-
derly arrangement of the manifold structural and functional components
which characterize the organism in such a way as to duplicate accurately
an existing pattern. How this is accomplished is one of the major bio-
logical problems, fundamentally as complex in the lower as in the higher
forms of life. On the other hand, the comparative ease of handling
microorganisms, their rapid growth, and the opportunity for rigorously
controlling various environmental factors all combine to render them
favorable material for experimental studies of the basic mechanism of
the growth process.

In several instances the successful resolution of a complex phenome-
non into its constituent components, the latter more readily amenable to
scientific analysis than the overall event, has resulted from kinetic

1 Hopkins Marine Station of Stanford University, Pacific Grove, Calif.

92

C. B, VAN NIEL

Studies. By relating the rate of the process to a number of variables,
preferably one at a time, regularities and irregularities are often dis-
closed which serve as useful indicators of component steps. It therefore
seems appropriate to examine what kinetic studies on the growth of
microorganisms have contributed and may contribute to a deeper pene-
tration into the problem of growth in general.

Nearly all that is known about the kinetics of growth of micro-
organisms has been learned from studies of so-called growth curves.
By inoculating a specified culture medium with a suitable organism and
determining, at the start and after various time intervals, the number
of individuals present — frequently restricted to a determination of the
number of viable cells — the requisite data are collected for constructing
a graph which represents the relationship between the microbial popula-
tion (or density) and the time elapsed since the initiation of the culture.
The most convenient curves are those in which the number of individuals
per unit volume is plotted on a logarithmic scale against time. Figure i
demonstrates the commonly encountered shape of such growth curves.

Figure i. A representative growth curve.

Much attention has in the past been paid to an interpretation of the
various phases of these curves. It will be clear that the first phase [i]
denotes a period in which a gradual transition takes place from a condi-
tion where an increase in numbers does not occur to one where the in-
crease is exponential [phase 2]. This second phase is easily accounted
for ; it means that during this period all organisms divide regularly, with
the daughter cells behaving in an identical fashion. Towards the end of

KINETICS OF GROWTH 93

this phase the slope changes back to zero, evidence that when this point
is attained [phase 3] either no further divisions take place, or that the
increase in cell numbers is exactly balanced by the disappearance or
death of individuals. The transition period, often brief and rather sharp,
would imply that :

(a) either fewer and fewer cells are still in a position to divide, or

(b) that the division rate of the individuals becomes progressively
less.

During the final phase [4] the death of the cells is predominant ; from
the point of view of studies on growth this phase is of no interest.

A considerable amount of time has been devoted to the first and
originally quite enigmatic phase, referred to as the latency and lag
period. As a result of various investigations, mostly before 1930, the
attitude towards the problems it presents can be summed up as follows.

In many cases the cause for the existence of a latency and lag period
resides in the cells used as inoculum. It can be completely eliminated
by inoculating from a culture which itself is in the exponential phase.
This has been demonstrated by Chesney (i) for bacterial, and by
Phelps (2) for protozoan cultures. The careful observations of Henrici
(3) have shown that the morphology of a bacterial species dififers
markedly according to the phase of the culture. It is therefore clear that
the initial lag must be ascribed to a transformation of aged cells into
"embryonic forms" (3, p. 140). Depending upon the age of the culture
from which the inoculum is prepared, this change may require anywhere
from minutes to many hours.

Similarly, a period of preliminary adjustment generally follows upon
a change of environmental conditions, such as composition of the cul-
ture medium, change in temperature of incubation, etc. In that event a
lag phase can be observed even if the new cultures are inoculated from
a culture in the exponential phase, and the length of the lag depends upon
the extent and nature of the change in conditions.

Growth curves of microorganisms are as a rule accurately reproducible.
Chiefly responsible for this feature are ( i ) the rigorous control which
the experimenter can exercise over the environmental conditions, and
(2) the small size of the organisms. The small size eliminates many
complications connected with the accessibility of foodstuffs. In higher
plants and animals transport of nutrients, often over considerable dis-
tances, may be a controlling factor for growth ; in microorganisms this
is not the case. Furthermore, the small size insures that large numbers
of individuals are always involved. Hence a statistical treatment of

94 C.B.VANNIEL

results becomes superfluous ; every experiment automatically yields data
in which individual variations are represented. This will be obvious if it
be realized that with bacterial cultures populations of about lo® organ-
isms per ml. are the rule, and that even for pure cultures of protozoa
numbers up to lo*^ per ml. are far from rare.

That not all individuals in a culture behave in an identical manner,
and that the results obtained really express statistical averages, follow
from the fact that growth curves of microorganisms are always smooth
even if cultures are started with a single individual, where one might
expect a discontinuity such as was first observed by Ellis and Delbruck
(4) in their studies on bacteriophage multiplication. More direct evi-
dence for the existence of variation among individuals in bacterial cul-
tures can be found in the investigations of Vaas (5) and Doudoroff (6).

In spite of these advantages, the kinetic approach has not yet been
used very effectively or extensively. It is probable that this must chiefly
be ascribed to the unwieldy and laborious methodology that used to be
necessary for the determination of a growth curve, requiring either
many sets of culture plates, or the actual counting under the microscope
of thousands of cells for each point on the curve. The newer method-
ology of determining population densities by turbidity measurements
with the aid of photoelectric apparatus, though not always entirely free
from objections, may well bring about a renewed interest in this line of
work, as is foreshadowed by the important studies of Monod (7) during
the last few years.

Kinetic investigations on cultures of microorganisms are eminently
suited for establishing relations between growth and environmental
factors, especially the nature and amount of nutrients. Based upon the
concept developed by Liebig about a century ago, it could be anticipated
that in microbial cultures growth would be limited by those ingredi-
ents which are present in "minimum" quantity. Experiments such as
those of Meyer (8) on the growth of Aspergillus niger in media with
graded concentrations of phosphate and ammonium sulfate have gone
far towards establishing the validity of Liebig's "law of the minimum"
for microorganisms. But these were not, strictly speaking, kinetic
studies ; the measurements dealt with total crops as dependent upon
nutrient concentrations rather than with rates of growth. Furthermore,
these researches were carried out at a time when little was known con-
cerning the exact mineral requirements of the organism used, which
makes the interpretation of the results somewhat difficult.

Decidedly more conclusive in this respect have been the recent studies

KINETICS OF GROWTH 95

in connection with growth factor requirements of microorganisms. The
mere fact that microbiological assay methods have been developed for
many of the known vitamins and amino acids provides convincing
evidence for the thesis that a single nutrient factor may determine the
extent of growth. However, with a very few exceptions, these contribu-
tions too are based upon measuring final yields rather than rates. And
where the latter are employed, as for example in the thiamin assay by
means of yeast fermentation (9), it is the metabolic, not the growth
rates, that are the basis for the determination.

Before 1942 only a few investigations on the kinetics of microbial
growth in relation to nutrient concentration were reported. The results
have been well summarized by Rahn ( 10, p. 251 ) in the following state-
ment:

"With increasing amounts of food, the crop, i.e. the final number of
cells in a culture, usually increases.

"With a liberal supply of building material, increase of energy food
will cause no acceleration in growth, but a longer growth period, and,
therefore, a larger crop. This is limited by accumulation of fermenta-
tion products or of inhibiting cell secretions.

"With all food components increasing, the crop will still be limited by
fermentation products or cell excretions. The law of diminishing returns
becomes quite evident."

While these conclusions were based upon experiments with bacteria
and yeasts, they apply equally to the growth of protozoa in pure culture,
as was first established by Phelps (11). His studies also revealed a
direct proportionality between cell yield and food concentration within
very wide limits, a situation which, wrote Phelps, "has not yet been
demonstrated in the bacteria nor in most of the work on yeasts." ( 11, p.

494).

Two years later, however, Dagley and Hinshelwood (12) supplied

the experimental evidence that in bacterial cultures also the rate of
growth is independent of the concentration of nutrient materials over
a wide range, while the final crop of organisms is directly proportional
to this concentration. The latter part of these conclusions was corrobo-
rated by studies on the growth of Escherichia coli in beef broth (13).
All of the earlier studies on the kinetics of growth in relation to
nutrient concentration were carried out with complex media in which
not one but all constituents were present in varying amounts. It was
only recently that Monod (7) started his investigations in a manner
more likely to furnish basic information. By using bacteria which can

96 C.B.VANNIEL

develop in a strictly defined medium, it was possible to determine the
effect of variations in the concentration of a single nutrient constituent
upon both the rate of growth and the final crop. So far, this approach
has been restricted to experiments with sugars and related compounds,
applied over a wide range of concentrations. With three different bac-
terial species it was established that, for a given carbon source, the rate
of growth is independent of the substrate concentration, and the total
yield is strictly proportional to it, down to such low levels as 20 mg.
substrate per liter.

There are on record a number of observations which suggest a dif-
ferent response of microorganisms to the concentration of nutrients.
Thorne (14), for example, claims to have shown that the rate of yeast
growth in sugar media falls uniformly with increasing concentration of
the yeast, a behavior entirely at variance with the above-mentioned
results. And Porter (15) has recently resuscitated the concept of "bio-
logical space" as earlier developed by Bail (16). The argument here is
that, because the total crop of bacteria in liquid cultures reaches a cer-
tain maximum regardless of the concentration of foodstuffs, growth
must be limited by a space factor rather than by nutrients. The follow-
ing passage from Porter's book (15, p. 136) is worth quoting in this
connection :

"There is considerable evidence in favor of some of Bail's claims
concerning space theory and Af (maximum) -concentration, but many
of his points are not accepted at this time. Further work will have to be
done on this subject — why bacteria stop dividing at a maximum speed
when a certain population is reached — before any definite conclusions
can be reached."

Most, if not all, of these aberrant results can be readily ascribed to the
experimental conditions imposing a limiting factor which, in spite of
repeated exhortations, is still insufficiently taken into account. This
limiting factor is oxygen. The supply of this gas is governed by its
very low solubility coefficient and unless special measures are adopted
to insure a greatly increased rate of diffusion (e.g. by forced aeration,
continuous shaking, etc.), its availability in liquid media is apt to fall
below that required for an optimum growth rate of the organisms. The
studies of Phelps (11), Rottier (17), Rahn and Richardson (18), and
Anderson (19) attest to the correctness of this statement. It seems more
than likely that the "biological space" referred to above is simply a
measure of the oxygen supply. Particularly the close packing of bacteria
in colonies provides a convincing argument for the contention that

KINETICS OF GROWTH 97

space per se cannot be considered seriously as an important factor in
growth.

Thus the best accessible evidence indicates that the growth rate is
not affected by the concentration of nutrients between very wide limits,
and that the total yield, over an equally wide range, is proportional to
the amount of food. This situation carries important implications which
will next be considered.

Let me reiterate that growth consists in part of the transformation of
foodstuffs into cellular constituents. This, in turn, depends upon
metabolism. In Chapter III Thimann has presented a number of cases
in which growth responses are strictly paralleled by metabolic responses
following upon the exposure of plants to auxin solutions. It is therefore
of the greatest significance that the rate of metabolism, as has been
repeatedly ascertained, is generally unaffected by concentration dif-
ferences of the substrate down to extremely low levels." And it is not
surprising that growth would follow the same pattern.

In many respects the relation between metabolism and growth is
quite obvious. "No growth without metabolism" was a dictum as
familiar since the latter part of the nineteenth century as the phrase "No
thought without phosphorus" had been earlier. The reason for this has
become apparent as the concepts of thermodynamics were developed.
The growing cell must synthesize cellular constituents, and for these
syntheses energy is required.

As a first approximation the relation between growth and metabolism
could thus be understood as one of energy linkage ; the metabolic activi-
ties of the cell provide the energy necessary for the performance of
synthetic reactions. Based upon this concept, various investigations have
been conducted aimed at measuring the "energetic yield" of growth
processes. By determinations of the amount of cell material that could
be synthesized from a known quantity of substrate, of the caloric value
of the cell material, and of the energy liberated by oxidations accom-
panying the observed growth, data were assembled which permitted the
computation of such "energetic yields." In subsequent studies of the
same general nature more precise relations were calculated by distin-
guishing between a "metabolism for upkeep" and a "metabolism for
growth" (20, 2i).

However useful and important this sort of information may have
been, recent trends have brought about a considerable change in the

2 Some interesting cases have been found in which this is not true. These, however,
present problems which are outside the scope of this discussion.

98 C. B. VAN NIEL

primary approach to the problems of syntheses and growth. Biochemists
in particular have come to recognize that the thermodynamic formula-
tions beg the question which has become paramount in today's bio-
chemistry, viz. that of the mechanism whereby the result is achieved.
Applied to the problem under discussion, this means that the concept of
an energetic coupling between synthesis and breakdown, no matter how
sound fundamentally, is too vague to be permanently satisfactory, and
that ultimately our task is to discover the mechanisms whereby such
coupling is accomplished.

The most fruitful hypothesis that has been advanced in this connec-
tion assumes the existence of material links between breakdown and
synthesis. It may, somewhat primitively perhaps, be paraphrased in
this manner : instead of regarding biological syntheses as very special
processes which, because they require energy, are basically different
from the breakdown reactions in which energy is liberated, it would be
more appropriate to describe syntheses as series of enzyme-controlled
step reactions, each step comparable in every way but one with those
which collectively compose the breakdown. The differentiating feature
is that, whereas the breakdown reactions can always start with any one
of the utilizable substrates in the food, the synthetic processes require
particular substances which are not usually provided but which arise
from the initial substrates as intermediate products in the normal course
of their metabolic degradation. Such products are characterized by being
more unstable, i.e. more reactive, than their ultimate precursors, the
foodstuffs proper, thus implying that their energy level has been raised
with respect to the initial materials.

This concept, first enunciated by Kluyver (22, 23) some fifteen years
ago, does not conflict with the earlier one of energetic coupling. Its great
value lies in the penetrating insistence upon the functioning of an in-
telligible chemical mechanism which is capable of accounting for the
transfer of energy.

During the past few years tremendous strides have been made in this
direction. Ranking high among these is the discovery of the phosphory-
lated carbohydrate now known as "Cori-ester," a compound which can
serve as the immediate raw material for the spontaneous enzymatic syn-
thesis of a number of di- and polysaccharides (24-30). No less signifi-
cant are the studies of Lipmann (31, 32) on the labile intermediary
metabolic products characterized by a high-energy phosphate bond. It
has now been shown that these substances, too, can be used for per-

KINETICS OF GROWTH 99

forming a number of fundamental biochemical syntheses with isolated
enzyme systems.

A special corollary of Kluyver's concept is that the molecular struc-
ture of the substrate and of the intermediate products to which this can
lead during its breakdown, rather than the amount of energy liberated
during this process, becomes the all important determinant of its po-
tentialities as the starting point for the synthesis of cell materials. The
investigations by Clifton and Logan, Doudoroff, and Bernstein on the
assimilation of chemically related compounds at different energy levels
(33-36) provided suggestive results in this connection. Better yet can
this situation be illustrated by recent studies on the synthesis of some
characteristic cell constituents.

Having recognized a number of vitamins as parts of enzyme systems,
it has become increasingly clear that the biosynthesis of such entities
proceeds in a chemically intelligible manner, even though many of the
steps involved in their manufacture may still be obscure. The mere fact
that certain microorganisms are unable to develop in simple media
without being supplied with preformed parts of enzymes shows con-
clusively that the supply of energy is not the governing factor, and that
the availability of appropriate building blocks must be reckoned among
the requirements for growth. Such phenomena are now generally inter-
preted to mean that the cells lack the machinery for the synthesis of
essential compounds, which must consequently be supplied. It has fre-
quently been shown that the same organisms can also develop in an en-
vironment which contains, instead of the enzyme part itself, substances
that are chemically more or less closely related to it, and from which the
cellular constituent in question is elaborated.

Corresponding observations have been made pertaining to the syn-
thesis of specific amino acids, as the epoch-making investigations of
Beadle and Tatum and their collaborators with "biochemical mutants"
have so amply demonstrated. Through studies of this kind it has been
possible to elucidate various steps in the biosynthesis of a number of
cell materials and to develop the thesis that each step is controlled by
a specific gene. It is impossible to do this field justice here; reference is
made to some recent reviews (37, 38).

No student of the chemistry and physiology of growth is likely to
deny the importance of these advances in our knowledge of biosynthetic
mechanisms. Nevertheless the inclusion of this phase may appear as a
digression from my topic. What have such problems to do with the
kinetics of growth of microorganisms?

lOO C. B. VAN NIEL

It will be evident that even the growth of a microbe is subjected to
the "principle of limiting factors," If the synthesis of all cell con-
stituents can proceed faster in one medium than in another, growth will
obviously be faster in the former. Similarly, the rate at which a single
component is being synthesized may, under special circumstances, be
the determining factor for the rate of growth of the organism. In that
case our understanding of biosynthetic mechanisms might be advanced
by kinetic studies on growth, as well as by the more strictly chemical
ones which hitherto have been the sole method applied. And through a
kinetic approach it may be possible to learn something about the synthesis
of these cell constituents by organisms whose external environment does
not need to contain a supply of closely related substances because the
biosyntheses can proceed from a variety of simple ingredients of the
medium.

While no investigations have yet been conducted with this purpose in
mind, some data are available which support the idea here advanced. For
example, Monod includes in his monograph (7) figures on the rate of
growth of Bacillus suhtilis in a complex and in a synthetic medium, the
latter composed of inorganic substances and sucrose. Growth proceeds
faster in the former than in the latter medium ; the generation times are
36 and 50 minutes respectively. Such differences could well be due to
differences in the rate at which particular cell constituents become
available, especially in view of the fact that in the synthetic solution they
must all be built up from inorganic ingredients and sucrose. If it were
possible to increase the growth rate in the simple medium through the
addition of known chemical entities, it might also become clear what
special syntheses are responsible for the lower growth rate.

Another instance is provided by Whelton and Doudoroff's (39)
study on the assimilation of sugars and related compounds by Pseudo-
monas saccharophila. It includes a determination of the generation times
of this bacterium in a standard mineral medium with any one of a
number of carbon compounds. These were found to vary from 105-110
minutes in the presence of trehalose and sucrose, to 196 minutes in the
presence of acetate; cultures in glucose and lactate exhibited an inter-
mediate behavior with generation times of 178 and 136 minutes respec-
tively. Studies of this kind, if sufficiently extended, could yield informa-
tion concerning the nature of intermediate products which serve as
building blocks and the rate at which these are generated. Correlation of
such data with overall growth rates, supplemented with experiments on
the influence of special additions selected for the purpose of furnishing

KINETICS OF GROWTH Id

similar or different building blocks, might then reveal the nature of
limiting synthetic reactions.

I do not mean to convey the impression that the above-suggested
approach is an easy and certain way of solving the problem of growth.
They are intended to emphasize that kinetic studies of the growth of
microorganisms, hitherto rather neglected, should be seriously con-
sidered as potentially capable of supplying information which can con-
tribute toward a moie detailed analysis of the manifold and interlinking
events which, in their totality, constitute the growth process.

The foregoing reflections on the effects of external factors on the
growth rate of microorganisms, and on the interpretations and deduc-
tions that can be derived from the observed results, may have fostered
the impression that internal factors can be regarded as more or less
constant. But this is not the case; there exists an impressive body of
indisputable evidence to the contrary. Changes in the morphological,
physiological, and biochemical behavior of microbial cultures, attributed
to adaptations, modifications, mutations, etc., have long been recognized
as manifestations of the intrinsic variability of the organisms. Monod's
recent studies (7, 40-42) on the growth rate of bacteria in media con-
taining two different carbon sources instead of a single one offer an
important advance in this field.

Depending upon the nature of the organism and of the substrates,
either of two types of growth curves describes the observed response
of the bacteria. One represents a single period of exponential growth
extending over the entire time interval during which substrate is avail-
able, with the sharp break at the end marking the complete utilization of
the two carbon sources. The other kind consists of two distinct parts,
each in itself a characteristic growth curve, with its phases of acceler-
ated, maximal, and retarded growth rates, the two curves being
conspicuously separated by a horizontal stretch which indicates the
cessation of growth during the corresponding period. Monod has desig-
nated this latter behavior by the term "phenomene de diauxie" ("di-
phasic growth") and has shown that it involves the occurrence of an
adaptation process.

By chemical analyses of the medium at different times it was ascer-
tained that the first part of the growth curve corresponds to growth of
the organism at the expense of only one of the two substrates, to the
complete exclusion of the other. The second, separate growth curve
coincides with the decomposition of the second substrate. And the inter-
vening period during which the population remains constant represents

I02 . C. B. VAN NIEL

the interval needed by the bacteria to adapt themselves to the utilization
of this second substance.

Adaptations of this sort have been known for several years and have
been ascribed to an adaptive enzyme formation (43, 44). They are
strictly dependent upon the presence of a specific substrate, an observa-
tion which has led Yudkin (45) to develop an ingenious hypothesis
concerning the mechanism by which adaptive enzymes are generated,
based upon differences in stability of the free enzyme molecules and of
the enzyme-substrate complex. Monod's experiments (40) have shown
that the rate of such adaptations in bacterial cultures depends not only
upon the presence but also upon the concentration of the substrate. In
view of the well-established fact that metabolic rates are generally in-
dependent of this last-named factor, and also that the growth rate of
the adapted bacteria is constant over a wide concentration range,
Yudkin's hypothesis may have to be modified.

Later experiments by Monod (42) have made it probable that adap-
tive enzyme formation must be interpreted as a competitive phenomenon.
This follows from observations on diphasic growth in relation to
variations in the concentration ratio of the two substrates. The higher
this ratio in favor of the substrate inducing the adaptation, the shorter
is the period of suspended growth : the measure of the adaptation rate.
The lag can even be eliminated completely by sufficiently magnifying the
difference in concentrations of the substrates.

From these results it has been inferred that the two different sub-
strates compete for a common precursor of the specific enzymes. Dif-
ferences in affinity of the substrates for the precursor would then be
responsible for the quantitative variations in the enzyme systems of
organisms exposed to solutions which contain both substrates simul-
taneously but in different concentration ratios.

Growth is the expression par excellence of the dynamic nature of
living organisms. Among the general methods available for the scientific
investigation of dynamic phenomena, the most useful ones are those
which deal with the kinetic aspects. The relative paucity of such data
on the growth of microorganisms at the present time cannot be denied.

Nevertheless a few contributions of considerable importance have
been made in this respect. It can also be asserted with some confidence
that the recent advances in methodology for the determination of growth
curves will help in overcoming the justifiable hesitation on the part of
investigators, previously confronted with cumbersome techniques, to
adopt this approach for future studies on microbial growth. Only by

KINETICS OF GROWTH IO3

further experimentation can the potentialities of kinetic studies in this
field be duly appreciated and more clearly defined.

REFERENCES

1. Chesney, A. The latent period in the growth of bacteria. Jour. Exper.

Med., 24, 387-418, 1 91 6.

2. Phelps, A. Grov,^th of protozoa in pure culture. I. Efifect upon the

growth curve of the age of the inoculum and of the amount of the
inoculum. /. Exp. ZooL, 70, 109-130, 1935.

3. Henrici, A. T. Morphologic Variation and the Rate of Grozvth of

Bacteria. Charles C. Thomas, Springfield, III, 1928.

4. Ellis, E. L., and M, DelbriJck. The growth of bacteriophage. /. Gen.

Physiol., 22, 365-384, 1939.

5. Vaas, K. F. Studies on the Grozvth of Bacillus megatherium de Bary.

Diss. Leiden, 1938.

6. Doudoroff, M. Experiments on the adaptation of Escherichia coli to

sodium chloride. /. Gen. Physiol., 2j, 585-611, 1940.

7. Monod, J. Recherches sur la croissance des cultures bacteriennes. Her-

mann & Cie., Paris, 1942.

8. Meyer, R. Zum Ertragsgesetz bei Aspergillus niger. Arch. f. Mikro-

bioL, I, 277-303, 1930.

9. Schultz, A. S., L. Atkin, and C. N. Frey. Determination of vitamin B^

by yeast fermentation method. Ind. Eng. Chem., Anal. Ed., 14,
35-39, 1942.

10. Rahn, O. Physiology of Bacteria. Blakiston, Philadelphia, 1932.

11. Phelps. A. Growth of protozoa in pure culture. II. Effect upon the

growth curve of dififerent concentrations of nutrient materials.
/. Exp. ZooL, 72, 479-496, 1936.

12. Dagley, S., and C. N. Hinshelwood. Physicochemical aspects of bac-

terial growth. I. Dependence of growth of Bacterium lactis aero-
genes on concentration of medium. /, Chem. Soc, 1938, 1930-1936,
1938.

13. Korinek, J. Zum Problem der Bakterienpopulation. Centr. f. Bakt.,

11. Abt., 100, 16-25, 1939.

14. Thorne, R. S. W. The nitrogen nutrition of yeast. Wallerstein Labora-

tories Communic, p, 97-114, 1946.

15. Porter, J. R. Bacterial Chemistry and Physiology. Wiley, New York,

1946.

16. Bail, O. O. moznosti a zakladech pokusne nauky populacni. Cas.

Ces. lek. 1927; Immunitdts f., 60, 1-22, 1929.

17. Rottier, P. B. Recherches sus les courbes de croissance de Polytoma

uvella. L'influence de I'oxygenation. C. rend. Soc. Biol., 122, 65-67,
1936.

18. Rahn, O., and G. L. Richardson. Oxygen demand and oxygen supply.

Jour. Bact., 41, 225-249, 1940; ibid., 44, 321-332, 1942.

I04 C. B. VAN NIEL

19. Anderson, E. H. Nature of the growth factor for the colorless alga

Prototheca zopfii. /. Gen. Physiol., 28, 287-296, 1945.

20. Algera, L. Energiemessungen bei Aspergillus niger mit Hilfe eines

automatischen Mikro-Kompensations-Calorimeters. Diss. Groni-
gen; Rec. Trav. botan. neerland., 2g, 47-163, 1932.

21. Tamiya, H. Le Bilan materiel et I'energctique des syntheses biologiques.

Paris, Hermann & Cie., 1935.

22. Kluyver, A. J. Atmung, Garung und Synthese in ihrer gegenseitigen

Abhangigkeit. Arch. f. Mikrobiol., i, 181-196, 1930.

23. Kluyver, A. J. The Chemical Activities of Microorganisms. Univ. of

London Press, 1931.

24. Cori, G. T., and C. F. Cori. The kinetics of the enzymatic synthesis of

glycogen from glucose- 1 -phosphate. /. Biol. Chem., 155, 733-756,
1940.

25. Cori, C. F. Phosphorylation of glycogen and glucose. In Biological

Symposia, 5, 131-140. Lancaster, Jacques Cattell Press, 1941.

26. Hanes, C. S. The breakdown and synthesis of starch by an enzyme

system from pea seeds. Proc. Roy. Soc, B, 128, 421-450, 1940.

27. Hassid, W. Z. The molecular constitution of starch and the mechanism

of its formation. Quart. Rev. Biol., 18, 311-330, 1943.

28. Doudoroff, M., N. Kaplan, and W. Z. Hassid. Phosphorolysis and

synthesis of sucrose with a bacterial preparation. /. Biol. Chem.,

148, ^7-7 S, 1943-

29. Doudoroff, M., W. Z. Hassid, and H. A. Barker. Synthesis of two new

carbohydrates with bacterial phosphorylase. Science, 100, 315-316,
1944.

30. Hassid, W. Z., M. Doudoroff, and H. A. Barker. Enzymatically syn-

thesized crystalline sucrose. /. Amer. Chem. Soc, 66, 1416-1419,
1944.

31. Lipmann, F. Metabolic generation and utilization of phosphate bond

energy. Advances in Ensymology, /, 99-162, 1941.

32. Lipmann, F. Acetyl phosphate. Advances in Enzymology, 6, 216-268,

1946.

33. Clifton, C. E., and W. A. Logan. On the relation between assimilation

and respiration in suspensions and in cultures of Escherichia coli.
Jour. Bad., 37, 523-540, 1939.

34. Doudoroff, M. The oxidative assimilation of sugars and related sub-

stances by Pseudomonas saccharophila. Enzymologia, p, 59-72,

1940-

35. Bernstein, D. E. Studies on the assimilation of dicarboxylic acids by

Pseudomonas saccharophila. Arch. Biochem., j, 445-458, 1944.

36. Clifton, C. E. Microbial assimilations. Advances in Enzymology, 6,

269-308, 1946.

37. Knight, B. C. J. G. Growth factors in microbiology. Vitamins and

Hormones, j, 105-228, 1945.

38. Beadle, G. W. Biochemical genetics. Chem. Rev., 57, 15-96, 1945.

KINETICS OF GROWTH IO5

39. Whelton, R., and M. Doudoroff. Assimilation of glucose and related

compounds by growing cultures of Pseudomonas saccharophila.
Jour. Bact., 4p, 177-186, 1945.

40. Monod, J. Influence de la concentration des substrats sur la rapidite

d'adaptation chez le B. coli. Ann. Inst. Past., 6q, 179-181, 1943.

41. Monod, J. Remarques sur le probleme de la specificite des enzymes

bacteriens. Ibid., 70, 60-61, 1944.

42. Monod, J. Sur la nature du phenomene de diauxie. Ibid., Ji, 37-40,

1945-

43. Karstrom, H. t)ber die Enzymbildung in Bakterien und liber einige

physiologische Eigenschaften der untersuchten Bakterienarten.
Diss., Helsingfors, 1930.

44. Karstrom, H. Enzymatische Adaptation bei Mikroorganismen. Ergebn.

Enzymforsch., 7, 350-376, 1938.

45. Yudkin, J. Enzyme variation in microorganisms. Biol. Rev., ij, 93-106,

1939-

VI. CELLULAR METABOLISM
AND GROWTH

BY E. S. GUZMAN BARRON^

IT IS generally recognized that the energy necessary for cellular
activities, whether it be used to maintain the steady state of adult
life, the "dynamic equilibrium" of Schoenheimer, or the state of
growth and synthesis which predominates in dividing and growing cells,
is provided by enzymatic reactions largely of the oxidation-reduction
type. This has been amply demonstrated by the increased O2 utilization
during growth and by the necessity of aerobic processes for the syn-
thesis of carbohydrates, fats, and proteins which together make the bulk
of the cells.

It will be therefore necessary before discussing the role of enzyme
reactions in cellular growth to review some fundamental aspects of the
properties of the oxidation-reduction enzymes.

General Properties of Enzymes

Enzymatic oxidation-reduction reactions are distinguished from
chemical reactions in general by the facts that they are reversible, that
electron exchange occurs in steps, and that they have a great tendency
to form coupled reactions.

The reversibility of enzyme reactions can be accepted as a general
property, since more and more of those reactions hitherto considered
irreversible are being shown to be reversible. Witness the enzymatic
synthesis of phosphopyruvic acid from pyruvic acid (48), a reversal of
the dephosphorylation of phosphopyruvic acid, which had been reported
as an irreversible process (58). In many instances reversibility of oxida-
tion-reduction reactions is made easy by the entrance of P into the re-
actions and generation of energy-rich phosphate bonds, as was shown
by Warburg and Christian (83) in the compulsory coupling between
the oxidation of phosphoglyceraldehyde and the phosphorylation of
adenylic acid, a finding extensively generalized to other reactions by
Lipmann (53).

The first step in an enzymatic reaction is the formation of the enzyme-
substrate complex. Evidence in favor of such a concept, introduced by
Michaelis and Menten (61), is abundant. Knowledge of the nature of

1 The Chemical Division, Department of Medicine of the University of Chicago.

CELLULAR METABOLISM

107

this complex formation awaits clarification of the structure of proteins,
since it is probably the protein moiety which combines with the sub-
strate. The substrate to be oxidized must fit quite closely to the side
chains of the protein molecule where combination occurs. No lock, how-
ever, remains immune to the prying of false keys ; and substances with
chemical structure close to that of the substrate will compete for union
with the protein, thus affecting the activity of the enzyme. This is
strikingly demonstrated by the inhibition of acetate oxidation by fluoro-
acetate discovered in animal tissues (Bartlett, Ahrens, and Barron
[14]) and in yeast (Kalnitsky and Barron [42]). In yeast, acetate
oxidation, completely inhibited by monofluoroacetate, was not affected
at all by the other halogen acids, bromo-, iodo-, chloro- acetic acids, as
well as by trifluoroacetic acid (Table I). This inhibition is due to compe-

TABLE I

Effect of Halogen Acetates on the Oxidation of Acetate by Baker's Yeast.

pH, 6.2, Temp. 25°. Acetate concentration, o.oi M ; halogen acetate, o.ooi

M.

Inhibition

^ (%)

Acetate

Endogenous

Interatomic Distances

Halogen

Oxidation

Respiration

C— Halogen bond (A)

CH2FCOOH

95

24

1. 41

CHXICOOH

none

none

1.76

CH^BrCOOH

4.8

40

1. 91

CHJCOOH

1-5

none

2.10

CH3COOH

1.09

CF3COOH

none

none

tition of fluoroacetate with acetate in the protein of the enzyme, for
inhibition occurs in the first step of acetate metabolism, namely citric
acid formation (Table II). Of all these acids it is monofluoroacetic acid
which approaches most closely in size acetic acid, as the interatomic
distance between C and F (1.41 A) is only 0.32 A more than the inter-
atomic distance between C and H. An increase in the interatomic distance
of 0.35 A as in C and CI (1.76 A) was enough to avoid penetration of
the halogen acid into the space reserved for acetic acid in the protein

loS

E. S. GUZMAN BARRON

moiety. The protein-substrate complex formation resembles thus strik-
ingly the antibody-antigen complex formation in the specificity.

Some proteins require the presence of — SH groups for enzymatic
activity. Furthermore, in some cases only a portion of the — SH groups

TABLE II

Effect of Fluoroacetate on Citrate Formation by Baker's Yeast.

Yeast cells, 8.5 mgs. ; Mg acetate, 0.077 ^ I fluoroacetate, 0.005 ^- Duration
of experiments, 4 hours. Blank values for oxygen uptake and citric acid
have been subtracted.

Experiment

No.

Citric acid formation
(cmm.)

Inhibition
(%)

0, Uptake (cmm.)

Control

Inhibitor

Control

Inhibitor

I

II

III

4.99
3.86
5-0

0.045
0.038
1.4

91
90

72

lOIO

1005
1012

0
0

242

present in the protein are necessary for enzyme activity, as demonstrated
with urease by Hellerman and coworkers (35). A survey made by Bar-
ron and Singer (13, 78) showed that a large number of enzymes which
catalyze the metabolism of carbohydrates, fats and protein derivatives
need the electronegative — SH groups for activity (Table III). Doubt-
less there are still other enzymes, not yet studied, which belong to this
class. An indication in favor of this assumption is given by the inhibi-
tion of mitosis produced by small doses of x-rays and by nitrogen
mustards. Up to the time of Dale's experiments (26) little attention was
paid to enzyme inhibition by x-rays. In 1940 Dale demonstrated that
enzyme inhibition could be produced by small doses of x-rays provided
protein concentration was kept low. Barron, Dickman, and Singer (8)
have shown that x-rays, a, and fi particles and y radiations inhibit sulf-
hydryl enzymes at small doses by oxidation of the — SH groups of the
protein (Tables IV and V). Inhibitions produced by x-rays and a
particles could be released on addition of glutathione. This inhibition
is undoubtedly produced by the OH and OH2 radicals and H2O2 formed
by the ionization of water during irradiation. In fact, protection of the
— SH groups by mercaptide formation with Cl-Hg benzoic acid abol-

>
• ♦-<

-4-'

U

a
>^

N

c
W

o

h-l '^^

In} S
W o

< K
H CO

(U
en

C
<u
bO
O
u

i-i

(J
ca

o

c

<

I

rt

C
lU

bo
o
u
•o
>.

CI

u

E

CS

D

o

c

bo
O
u

x;

c

CJ

e

■§

_c

O

en

o.

U

n

,(-;

CJ

a

■*-*

>m

u,

rt

1^

H

U

3
u
u
o

CI

o
B

<u

c
a

E

<u

00

0\

o

a

'h

oi
-t-t

tl
o

<

en

oi

C
(U

be
o
u

>.

<u

w
■*-»

03

Uh

rt

CI

tn
nJ

C
(U

bo
O

tl
-o

<u

03
CI

C

bo

o

o
o

3
O

I

u

E
o

n

>

03
Ui

O
U

^

"o

CI

X

c«

C

U

bo

o

b

•o

>^

j::

o

CI

-a

«

w

(n

d

N

tn

rt

(U

rn

>.

c

CI
tn

CI

u

o

U
tn
>>

-1

W

<

U

kJ

1-1 M

In. OO

0\

C

_o

■*-»
of

'>>

Ui

O

tn
O

J=

o

Ui

CI

b

>

Nw/

^

«;

tn

<u

n1

tn

-^

nJ

rt

•4-*

0

-E

03
J3

03

tn

O.

O

tn

O

0

E

a

^

tn

0

O4

rt

t.J

0

0

n

c

lU

tn

>>

tn

OJ

o3
C

u

0

bo

0

tn

O

c

.s

o

c

0

tn

a

tn

CI

CJ

X

0

0

•o

•T3

CI

J3

JC

<

<

K

Oh

0^

o

t-H

N

ff)

Tf-

10

N

IN

M

N

N

w

3

6

N
C

w

tn
ns
C

CI

bo
O
u

o

'O

lU

'O

>.

j=
d

H

u

<u
tJ

>>

"bo
O

o.

tn

O

j::

HJ
tn

">.

X

o

u
03

u

tn

a

c

bo
o
u

>.

o
•O

lU

>
>,

pL,

c
o

-a

o!
>
3
I-i
>i
PL,

tn
03

CI

C

(L)

bo
o
I-i

•a

tu

'o

c
a

I-i

CI

u

CS

C

CI

bo
O

u.

•o

>.

-C
CI

•o

"o
>-i

CI

(J

>,

W O

0!

c

iJ

bo

0

t-i

c»

T)

>,

c

J2

M

"O

0

d

Ui

■4-»

Tl

0!

OS

d

3

T)

b/)

CI

n

4->

0!

<u

O!

t^

:^

ti

o)

^— .i^

c

X

d

CI

bO
0
u

i-i

>>

^"^

j::

CI

lU

t/i

-0

03

n

n

>.

'ci

E

CI

3

<

CO

«Q.

l-H (N

<^

10 \o

0\

no

E. S. GUZMAN BARRON

ished inhibition of urease by y radiation (Table VI). Destruction of
H2O2 by catalase, diminished the degree of inhibition of phosphoglyc-
eraldehyde dehydrogenase by a particles, and suppressed the inhibition
by /S particles. The combination of nitrogen mustards with — SH groups

TABLE IV

Effect of X-Rays on the Activity of Phosphoglyceraldehyde Dehydrogenase.

Buffer, phosphate, pH 7.0. Enzyme activity, K=z — X where t, time

t CoC

in minutes; Co (DPN) initial; C (DPN) at time t.

X-Ray
dose (r)

Enzyme
Activity
K X 10'

Inhibition (%)

Reactivation with
Glutathione (0.002 M)

none

4.80

25

4.80

none

50

4.80

none

100

3-81

21

complete

200

2.43

50

62

300

0.98

80

400

0.98

80

500

0.25

94

10

TABLE V

The Effect of a (Po), /? (Sr^"), and y (Ra) Radiation on the Activity of
Phosphoglyceraldehyde Dehydrogenase. Enzyme, 20y ; phosphoglyceralde-
hyde, 4X10-' M; DPN, 2.5X10-^ M; NaHASO^, 6X10-' M; pyro-
phosphate buffer, pU 8.5, 3X10"^ M. Temp. 26°.

a Radiation
Dose(r) Inhibition(%)

/3 Radiation
Dose(r) Inhibition (%)

y Radiation
Dose(r) Inhibition(%)

180

58

20

10

25

17

360

73

40

17

50

54

520

92

80

32

1080

complete

120

37

200

67

CELLULAR METABOLISM III

is known. However, as this combination is irreversible, it is difficult to
establish definitely whether enzyme inhibition is due to this reaction or
to combination with some other group of the protein. Very little is
known about the role of these — SH groups. Protection of the — SH
groups of the protein moiety of succinoxidase by succinate and by
malonate {2,7) would suggest that the substrate is anchored in close

TABLE VI

Inhibition of Crystalline Urease b}' Radium Radiation
Effect of Glutathione, Glycine, and p-Cl-Hg Benzoic Acid.

Dosage (r)

Additions

Inhibitions (%)

ICG

24

100

glutathione

30

200

36

100

glycine, iXio~® M

none

lOO

glycine, iXio~^ M

none

200

34

200

glutathione

39

200

p-Cl-

-Hg benzoic acid+glutathione

none

vicinity to the — SH groups. The proximity of the carboxyl electronega-
tive groups would decrease the activity of the — SH groups and thus
render them less susceptible to the action of sulfhydryl reagents. Some
other proteins require the presence of a metal (Mg, Mn, Zn, Co) for
enzyme activity; the metal does not participate in the reaction. Their
role is still more obscure than that of the — SH groups. It is known,
however, that — SH groups and metals contribute to increase the stability
of the native protein.

The combination of the substrate with the protein brings forth the
"activation" of the substrate. It is not yet clear what is meant by activa-
tion of the substrate. It is plausible, however, to postulate with Michaelis
that activation consists in the formation of a radical, the first step in the
process of oxidation of organic substances with a loss of two electrons.
This is implicit in Michaelis' theory of compulsory univalent oxidation
(6o).

112

E. S. GUZMAN BARRON

In condensation reactions where two substrates take part in the re-
action, as in the formation of acetylcholine, it seems that only one of
the substrates combines with the protein and becomes activated. Evi-
dence for this assumption is provided by inhibitions with competitive
inhibitors. For example, methyl bis (dichloroethyl) amine HCl inhibits
choline oxidase because of the similarity of its first transformation
product, ethylene imonium, to choline; it also inhibits the synthesis of
acetylcholine and its hydrolysis. It has no efifect on the hydrolysis of
tributyrin. Fluoroacetate inhibits the oxidation of acetate while it has
no effect on the synthesis of acetylcholine and acetylations in general
(Table VII). It must be concluded that in acetylations the protein
combines with the acetylated compound and not with the acetyl group.

TABLE VII

Effect of Methyl bis (Dichloroethyl) Amine (o.ooi M) and
Fluoroacetate (o.oi M) on Acetylations, Oxidations, and Hydrolysis.

Inhibition (%)

Reaction

N Mustard

Fluoroacetate

Acetylation of sulfanylamide

Acetylation of choline

Oxidation of choline

Oxidation of acetate

Choline esterase (hydrolysis of acetylcholine)

Esterase (hydrolysis of tributyrine)

none

75
complete

none

complete

none

none

none

none

complete

A number of proteins, components of enzyme systems, have been iso-
lated and prepared in crystalline form. Furthermore our knowledge of
the physical and chemical properties of proteins has considerably in-
creased during the last years. The time has come to use these techniques
for a serious study of the protein moiety of the enzyme. It is necessary
to know the architecture of the protein moiety of the enzyme, the spac-
ing of the side groups to which the substrates attach themselves, and
the nature of the activation of the substrate.

The Oxidation-Reduction Systems

In 1930 Barron and Hoffman (10) found that the catalytic power of
reversible oxidation-reduction dyes on cellular respiration was condi-
tioned by two factors: the oxidation-reduction potential of the dye
(Fig. i), and the permeability of the cell membrane (Table VIII).

CELLULAR METABOLISM

113

200

100

-300 -200

Figure i. Relation between the catalytic power of dyes and its Eo to the oxidation-
reduction potential of starfish eggs. (Dyes which penetrate through the cell wall.)
Abscissa, Eo (in millivolts) of dyes at /'H 7.0. Ordinate, per cent increase of O2 con-
sumption after dye addition, i. phenol-indophenol ; 2. toluylene blue chloride; 3. methy-
lene blue; 4. cresyl blue; 7. cresyl violet; 8. safranine; 9. janus green; 10. neutral red.
From Barron and Hoffman, /. Gen. Physiol., 13, 483, 1930.

TABLE VIII

The Effect of Reversible Dyes on the Oxygen Consumption of Starfish Eggs
(Mature, Unfertilized). The Efifect of the Membrane Permeability.

Eo'at/'H

7.00 in

volts

Oxygen consumption
per hour in cmm.

Permeability
of membrane

Name of dye

Before

dye
addition

After

dye

addition

Per cent
increase

Gallocyanine

+0.013

23.1

337

does not
penetrate

46

Methylene blue

-fo.oii

23-3

85-5

penetrates

267

Indigodisulphonate

—0.125

25.4

24.6

does not
penetrate

none

Cresyl violet

—0.165

25-3

65.8

penetrates

160

114

E. S. GUZMAN BARRON

The speed of this catalysis was correlated to (a) the speed at which the
dye was reduced by the cell, and (b) the speed at which the leucodye was
oxidized by atmospheric oxygen. Such a correlation between free energy
(as determined by the oxidation-reduction potential of the dye) and rate
of oxidation (as determined by the increase in the O2 uptake of the cell)
was naturally received with caution. At the suggestion of W. Mansfield
Clark and with his generous cooperation these experiments were re-
peated in homogenous systems by studying the rate of oxidation of
reduced dyes by atmospheric oxygen. The same relation between free
energies and rates of reaction was found (4) (Fig. 2). The theory of

Figure 2. The rate or autoxidation of reversible dyes. Abscissa, logarithm of time

(in minutes) required to oxidize the reduced dye; ordinate, Eo —

RT

In

2F I — o

expressed in millivolts. Curve i, phenolindophenol ; Curve 2, o-cresolindo-2, 6-dichloro-

phenol ; Curve 3, methylene blue ; Curve 4, i-naphthol-2-sulfonate-indophenol.

absolute reaction rates, developed mainly in Princeton University by
Eyring and his coworkers (29, 31, 89), which focuses its attention on
the thermodynamic probability of molecules entering the transition
state, has shown that the speed of reaction depends not on the heat of
activation alone but also on the entropy of activation :

KT

k =

-AF*/RT

h

where AF* is the free energy of formation of the activated state, and
k and h the Boltzman and Plank constants, respectively. This equation

CELLULAR METABOLISM II5

shows that the velocity constant (^i), is a function of the work
( — AF*) necessary to get the reactants into the transition state. If the
concept of the transition state is combined with quantum-mechanical
concepts of stabilization by resonance and the dependence of resonance
on the number of possible unperturbed forms, we arrive at a rather
satisfactory method of dealing with rate problems.

The activated substrate is oxidized by one or more of oxidation-
reduction systems, some sluggish, some electromotively active, which
transfer electrons to molecular oxygen. If more than one system is
interposed between substrate and oxygen, the electron transfer occurs
step by step from the system of more negative potential up to molecular
oxygen.

A study of the oxidation-reduction systems which take part in the
transfer of electrons from substrate to oxygen, shows that these systems
where the oxidation involves a total two-electron transfer (pyridine
nucleotides, flavins) do so in steps of one electron through radical forma-
tion as an intermediate step. The climbing over a large energy hill
involved in a two-electron oxidation is made easy by the interposition of
the intermediate step. Oxidation of foodstuflfs is thus performed by
one-electron steps.

Little is still known about the oxidation-reduction potentials of pyri-
dine nucleotides and flavins when in combination with the proteins.
Both combine reversibly with proteins, and the dissociation constants
of the oxidized and reduced states are sometimes different. However,
a model can be provided by a study of the oxidation-reduction potentials
of Fe-protoporphyrin and of Fe-protoporphyrin combined reversibly
to nitrogenous bases to form the hemochromogens. On addition of co-
ordinating bases to the Fe-protoporphyrin there is formation of a series
of oxidation-reduction systems which at pH 9.20 go from —0.292 volt
in Fe-protoporphyrin to +0.050 volt in nicotine hemochromogen (Fig.
3) (5). A similar change of potentials has been observed when flavin
is combined to a protein as in Warburg's old yellow ferment, where the
potential jumped from —0.208 volt at />H 7.0 for flavin to —0.061 volt
for the latter (47). W. M. Clark and his coworkers (22) have given a
quantitative treatment of the equilibrium reactions between a nitroge-
nous base and Fe-porphyrin and have shown the relation between the
change of potential and the dissociation constants of the metallo-porphy-
rin-nitrogenous base complex compounds. If association of the Fe"^"^"*"
and Fe"*"^ state with the nitrogenous base is equal in strength, there will
be no change in potential. If the potential becomes more positive it

+0.05 -

0.25 -

0.2 0.4 0.6 0.8

M/L of Nitrogenous Substance

Figure 3. The oxidation-reduction potentials of hemin and of some of its hemo-
chromogens. The influences of the concentration of the following nitrogenous sub-
stances: (i) cyanide-hemochromogen ; (2) pilocarpine-hemochromogen ; (3) histidine
hemochromogen ; (4) a-picoline-hemochromogen ; (5) pyridine-hemochromogen ; (6)
nicotine-hemochromogen. From Barron, /. Biol. Chem., 121, 285, 1937.

CELLULAR METABOLISM II7

means that the Fe++ porphyrin forms the more stable compound; if the
Fe"^"^"*" form associates more strongly than the Fe++ form the trend
of the potential will be in the negative direction. This is of great sig-
nificance for an understanding of the varied catalytic effects of the
different oxidation-reduction systems which form part of the enzyme
systems. In the same manner as the Fe-porphyrins, and its nitrogenous
complexes give a series of oxidation-reduction systems, the potentials
of the pyridine nucleotides and flavins in combination with different
proteins may change, at constant pK, according to the value of the dis-
sociation constants of the protein complex.

The breakdown of foodstuffs in stepwise manner is facilitated by the
interposition of reversible oxidation-reduction enzymatic systems, and
the stepwise transfer of electrons is facilitated by the property of the
three main electron-transfer systems (pyridine nucleotides, flavins, Fe-
porphyrins) of forming reversible complex compounds with proteins
with a wide variety of oxidation-reduction potentials. It must be re-
called, however, that electron transfer between two reversible enzymatic
systems does not occur but through the mediation of electroactive sys-
tems as shown by the experiments of Borsook (Table IX) (17).

The Pyruvate Oxidative Pathway

Of all the foodstuffs, carbohydrate provides most free useful energy
in its breakdown. It is mainly carbohydrate-provided energy which is
used during the process of embryonic growth and development. In gen-
eral, carbohydrate breakdown occurs in two phases : the anaerobic phase
which ends in pyruvic acid formation, and the oxidative phase which
involves the utilization of pyruvate via the oxidative pathway. In 1932
Barron and Miller (11) found that pyruvate was oxidized to acetate
and CO2 by gonococci. Although formation of acetate from pyruvate
by animal tissues has also been reported (46, 84, 55), there is a tendency
to ignore this oxidative pathway of pyruvate metabolism and to give as
the main oxidative pathway that of pyruvate carboxylation to oxaloace-
tate and condensation between pyruvate and oxaloacetate, as indicated
by Krebs in his "tricarboxylic acid cycle" (45). The powerful inhibit-
ing effect of fluoroacetate on the oxidation of acetate was used by Bart-
lett, Ahrens, and Barron ( 14) to test the oxidative pathway of pyruvate
metabolism. To discover this oxidative pathway fluoroacetate is indeed
the ideal agent, since if acetate is formed during the metabolism of
pyruvate, and it normally disappears in condensation reactions (acetyla-

-4->

>
;3

PLH

01

4J

X

.2

'■4-J
O

t/2

J >

1)

f^ .

II ^

II ^

. o

0 II

CQ a;

<

o
u

X

t

> ^ -^

I -t-i

I ^
+ 2

s °'

° II
fe II

■«i.

•n

■<->

W

c

H

«J

k^

rt

o

>

<u

3

m

<u

4-1

u

E

o

Q

o
E
o

u

-a c
(^ (/)

•"? 'S

PQ bo
\-,
o

c/i

C!

o ,/

>

^ ^

3

.y o

>>

E E

Ph

(U

E
u

o

micro
mols.

fe

3

d d d ►-■

o q »n 00 o vo

drii-i'cioooooocido

NO'-ii-iCMCMNfol

66006666

+ + I I I I I I

vo

1 ^

vo

1 ^

■o

N

1 CM

0)

1 0<

0)

I I

o

c

a>

J=
a
o

c

o
c
<u

Oh

CS

c

0

^

^

•4->

■M

+j 4->

•+-> 4-*

ID

"s

OJ

0

<u

«J (U

jj jj

^

tn

_o

_o

.2 .2

^ ."s

3

0
C
0

n

'>

w

•o

">

*> '>

•> •>

w

"o

<v

_c

<u

<u aj

W (U

G

E

■>

C

\IZ

c

3 3

3 3

1>

be

en

3

JJ

3

^M

OJ

^ ^

JJ JJ

>.

0

*>.

.3

1 ">»

[ ">. >.

">, ">>

r~!

be

j::

•*-*

1 J=

1 J= J=

1 J= JZ

-4—'

0
1-

U

-4-1

(/)

3

-*-»

-1-1 -i-j

-M *J

■U

•0

C

OJ

Q

<u

IL>

qj (U

(U <u

C
1— 1

03
1— »

s

Pi

J5

:^

:§ 1^

:§ s

IIIIIIIIIIII1I+++

+ + + + + + + + + I I+ + + +I +

000000000000

000

<N <N N

00000000000

o

000

M CS M

i-i N fO'^iOVO tvOO 0\0 "-i 0) rn -^ ^n \o t%

O Tt

i-i o

T

tjq

ijq

^-^

N

<u

+

0»

+

+

CM

x +

M

CO

+

0

U

1

ffi

-4-*

>

t

0

0. ffi

t +

I I

O
O
to

;-i
o

E

o

u

CELLULAR METABOLISM II9

tions, citrate formation, fatty acid synthesis), it will not be detected
unless some of these reactions are inhibited. Experiments where the
metabolism of pyruvate was studied in the presence of fluoroacetate in
kidney, liver, and muscle showed accumulation of acetate and inhibition
of pyruvate utilization due to accumulation of its oxidation product,
acetate. The other pathways of pyruvate metabolism (amination to
alanine, reduction to lactate, and dismutation to lactate and acetate)
were not affected. Furthermore none of the oxidative steps of the tri-
carboxylic acid cycle — oxidation of isocitric, a-ketogiutaric, succinic,
and malic acid — were affected by fluoroacetate. The experiments with
fatty acids showed an inhibition in the oxidation and utilization of these
substances with an inhibition of acetoacetate formation in liver slices,
but with acetate as substrate there was accumulation of acetoacetate.
This inhibition of acetoacetate formation in the presence of fluoroacetate
supports the view, now generally accepted, that the oxidation of fatty
acids occurs by steps until the formation of acetate. In the metabolism
of acetate there are then two pathways : oxidation inhibited by fluoro-
acetate, and condensation to acetoacetate not affected by it. Since acetate
accumulates in the presence of fluoroacetate, it is plausible to postulate
that acetate is the end product in the oxidation of fats and carbohydrates.
Furthermore the high toxicity of fluoroacetate is indication of the es-
sential role of acetate metabolism in higher animals and in some plants.

The oxidative pathway, however, is not the only avenue of utilization
of pyruvate. Pyruvate, the most highly reactive molecule among the
intermediates found during the breakdown of foodstuffs, is in fact
utilized in a variety of ways in the different cells, and even in the same
cell the utilization of pyruvic acid will depend on the oxygen tension,
thiamin concentration, glutathione concentration, etc. The multiple
pathway of pyruvate metabolism postulated in this laboratory for several
years may thus be summarized as follows, all these reactions being
reversible (see chart).

How much of the oxidative removal is performed via acetate and
how much via carboxylation to oxaloacetate has not been determined
in all tissues. In the experiments of Bartlett and Barron (14) with
kidney slices, at least 45 per cent of the pyruvate utilized could be ac-
counted for as due to direct oxidation.

It can be seen from this chart that the only difference between this
scheme and Krebs' tricarboxylic acid cycle is that here pyruvate is di-
rectly oxidized to acetate. The synthesis of citric acid from acetate was
demonstrated in yeast by Sonderhoff and Thomas (79). The oxidation

I20

E. S. GUZMAN BARRON

Lactate

The Multiple Pathway of Pyruvate Metabolism

+H, +NH3

^ Pyruvate '^ Alanine

C4

o

Acetoacetate

X

+CO,
+H,0

1

Oxaloacetate

Lactate + Acetate + COo

Acetaldehyde + CO^

+ Pyruvate

Acetylniethylcarbinol.+ 2 CO.
4-NH3

Fat ^5±^ Acetate <*— ;• "Active" acetate + Oxaloacetate -=:i Aspartate — i ^

Citrate ^= Isocitrate

O

+
Oxalosuccinate

O 4

4-

Malate
O/f

+
Fumarate

Protein
if

0

U
1

NT*

f

a-Ketoglutarate ; ^ Succmate

-C02

+NH3

i Glutamate

of isocitric acid by isocitric dehydrogenase and triphosphopyridine
nucleotide (TPN) to oxalosuccinic acid and its decarboxylation to
a-ketoglutaric acid by a specific carboxylase has been demonstrated by
the beautiful experiments of Ochoa (67, 69). The other systems have
been known for some time.

The interposition of enzymatic reversible oxidation-reduction systems
in the various steps during the burning of foodstuffs, already shown in
the scheme of pyruvate metabolism, allows the production of a series
of coupled reactions whose speed and direction will be determined not
only by the value of the equilibrium constants but also by the environ-
ment (temperature, O2 tension, CO2 tension, electrolytes, hormones,
etc.).

CELLULAR METABOLISM 121

Ochoa (68, 69) has thus shown a number of coupled reactions which
in in vitro experiments end in carboxylations :

(i) a-ketoglutarate + CO2 + glucose 6-phosphate ^ i-isocitrate +

phosphogluconate ;
(2) Malate + TPN ^ pyruvate + COo + TPNH,

These reactions do not occur in the presence of pure isolated enzyme
systems. Carboxylation of a-ketoglutarate and pyruvate, difficult of at-
tainment because of the values of the equilibrium constants, was ren-
dered possible by coupling carboxylation with an oxidation-reduction
reaction. Whether such reactions do occur in living cells is problematic.

Oxidations and Phosphorylations

Whenever an important discovery is made, there spring forth applica-
tions, similarities, extensions. Years have to pass by before harvesting
from the melee the useful crop. The coupling between phosphorylation
and oxidation observed by Meyerhof (59) in his studies on the mecha-
nism of glycolysis was explained in 1939 by the striking demonstrations
in Warburg's laboratory (83) which culminated in 1942 in the isolation
and crystallization of the specific transphosphorylase (18). Warburg
and his coworkers showed that the aldehyde group in phosphoglyceralde-
hyde is not oxidized as aldehyde hydrate but as aldehyde phosphate. By
a non-enzymatic reaction glyceraldehyde 3-phosphate takes up another
phosphate group to give 1,3-diphosphoglyceraldehyde, which is oxidized
by DPN to 1,3-diphosphoglyceric acid:

OH OH

HX-0-P=0 H^C-O-P^rO

I OH I \

HCOH HCOH OH

I +DPN ^:±: | ^O -f DPN Hj

HC-OH C^

'O-PO3H

a-^-^2

1 ^o« ^

0-P=:0

The phosphate transfer enzyme, a Mg-protein, transfers the unstable
phosphate group into adenosinediphosphate to give ATP. This reaction
has been taken as a model of phosphate transfer from aceylphosphates :

122 E. S. GUZMAN BARRON

O

//

-C^ OH O

\ / //

\^^ +ADP 7-^-r +ATP

OH \o_

O-

/

-c

\

The resonating structures impart stability to the group (39).

Lipmann (53) has appHed and extended these concepts to the oxida-
tion of pyruvate and a-keto acids in general, and indeed has found that
certain bacteria produce acetylphosphate on oxidation of pyruvic acid,
and that animal tissues hydrolyze acetylphosphate rather rapidly (54).
These important contributions of Lipmann were promptly accepted,
but little has been accomplished towards further clarification of the
mechanism of aerobic phosphorylations. The fact that they exist was
demonstrated by Belitzer and Tsibakowa (15), Ochoa (65) and Colo-
wick et al. (24). From Ochoa's work it is clear that on the oxidation
of pyruvate there are about 15 phosphorylations per mole of pyruvate
completely oxidized, i.e. 3 per atom of oxygen. Of these 15 phosphoryla-
tions, Ochoa (66) has reported that during the oxidation of a-keto-
glutaric acid to succinic acid in the presence of glucose and adenylic acid,
3 phosphate molecules are simultaneously esterified to the sugar forming
hexose diphosphate. The transfer of 2 electrons to oxygen was thus
accompanied by the release of energy in three successive steps in the
form of high-energy phosphate bonds. How they are formed remains
unknown. The postulated succinylphosphate formed as the first product
of oxidation has not been demonstrated. Ball's suggestion (2) that the
step by step transfer of electrons from oxidizable substrate to pyridine
nucleotide, from this to flavin, from fllavoprotein to cytochrome, and
from cytochrome to oxygen might release the energy (0.25 V.) neces-
sary to form a high-energy phosphate bond seems reasonable, but again
is still in the realm of speculation. To complicate the problem, Stumpf

CELLULAR METABOLISM I23

et al. (81) maintain that a-ketoglutarate is oxidized to succinate by
preparations from pigeon breast muscle without the necessity of in-
organic phosphate, adenosinetriphosphate, and magnesium, all com-
ponents considered by Ochoa essential for the oxidation process. The
phosphorylations in the first step of the cycle (see chart) — pyruvate -f-
y202—> acetate + CO2 — have not been established in animal tissues in
spite of the repeated efforts of Lipmann (54). Here again, Stumpf et al.
maintain that this oxidation occurs in the absence of phosphate and
with no formation of acetyl phosphate. The second oxidation step, iso-
citrate + TPN ^ oxalosuccinate + TPN H2, is postulated to give no
high-energy phosphate bonds; however in the experiments of Ochoa
(69) there was but one step in the transfer of electrons. If Ball's sug-
gestion is verified, here as well as in the oxidation of pyruvate, the step
by step release of energy might be utilized for the formation of high-
energy phosphate bonds. In the fourth step, succinate + ^Og ^ fuma-
rate + H2O, there has been found evidence of one phosphorylation
(66). This is probably due to the high potential of the system (o V. at
pH 7) and the fewer steps in electron transport to molecular oxygen
(0.8 v.). The final step, malic + ^02 ^oxaloacetic + H2O, was found
by Kalckar (40, 41) to produce at least two phosphorylations. It seems
as if the insistence on finding analogs to Warburg's reaction, which after
all is a fermentation reaction, has stopped progress and hindered the
search for other channels of energy transfer. Perhaps Ball's suggestion
ought to be studied experimentally. Aerobic phosphorylations may be
produced by the release of energy from the different oxidation-reduction
catalysts to adenylic acid, for it is accepted that adenylic acid enhances
completeness of oxidation. Lack of phosphorylation in isolated enzyme
systems where substrate is oxidized by only the first oxidation-reduc-
tion system and failure of phosphorylation in some tissue extracts when
cytochrome C is not present favor this opinion. The ATP thus produced
would represent the free useful energy ready to be used.

Energy transfer from electrons and its utilization by the cell through
adenosinetriphosphate occurs not only on the oxidation of carbohydrates
but also on that of fats, as the studies of Lehninger (50) and Munoz
and Leloir (64) have demonstrated.

The Oxidation and the Synthesis of Fats

Our knowledge of the mechanism of oxidation and synthesis of fatty
acids has advanced considerably in the last years through the efforts of

124 E. S. GUZMAN BARRON

a number of investigators, although no single enzyme has yet been
isolated. The classical concept of Knoop (43) and Dakin (25) with
the modifications suggested by MacKay (57) — oxidation at their /SC
atoms with subsequent removal of two C units which may condense
to acetoacetate or to citrate — seems now generally accepted. As in the
complete oxidation of carbohydrate, the series of oxidations which
take place during the breakdown of fatty acids require the presence of
ATP or of C4 dicarboxylic acids, which in their oxidation provide
energy-rich phosphate bonds (50, 64, 38). Whether there is a trans-
phosphorylation process with the formation of unstable acyl phosphates
is not known, in spite of some efforts to prove that such substances are
formed. The acyl phosphates (monopalmityl phosphoric acid and mono-
octanoylphosphoric acid) , although hydrolyzed much more slowly than
acetyl phosphate (51), were physiologically inactive. When octanoic
acid was oxidized by the liver in the absence of C* dicarboxylic acids,
acetoacetate was the end product (52, 85). In the presence of C4 dicar-
boxylic acid, the C2 fragments gave citric acid and its oxidation products,
a-ketoglutaric acid, and succinic acid. The interconversion of acetic
acid to acetoacetic acid has been demonstrated by direct interconversion
(49, 86) as well as by the synthesis of citric acid from acetoacetate pre-
sumably by its previous breakdown to "active" acetate (87). This acetic
acid, the end product of carbohydrate oxidation and of fat oxidation, is
the connecting link of carbohydrate and fat metabolism. The further
oxidation of acetic acid, we have already shown (see chart), occurs
through the tricarboxylic acid cycle. The synthesis of fat starts with
the condensation of acetate to acetoacetate. Indeed, Rittenberg and
Bloch (75) demonstrated with isotope experiments that acetic acid is
utilized for the synthesis of fats.

The Oxidation and Synthesis of Proteins

Proteins in the oxidation of their fundamental units, the amino acids,
contribute greatly to the formation of the carboxylic acids which by
their reversible coupled oxidation-reductions form the common pathway
for the final steps of complete oxidation of foodstuffs, whether it be
carbohydrate, fat or protein. The rapid and continuous breakdown and
synthesis of proteins so well demonstrated by Whipple and his co-
workers (56) and by Schoenheimer and his coworkers (77), as well
as the great specificity of proteinases and peptidases, brought forth the
suggestion that protein synthesis could be performed by coupled reac-
tions between the hydrolytic enzymes. Bergmann and Fruton (16), in

CELLULAR METABOLISM I25

fact, presented models of peptide bond synthesis by selecting suitable
amino acid derivatives and by producing the conditions necessary for
reversal of hydrolysis. Whether this mechanism actually takes place in
living cells is not known. The synthesis of proteins, like that of glycogen,
must be due to the cooperation of a large number of enzymes. In cyto-
chrome containing living cells it does not occur in the absence of oxygen,
and it is plausible to postulate that proteins are, like carbohydrates and
fats, synthesized after their breakdown to amino acids and that peptide
bond formation requires the utilization of energy produced during the
oxidation of carbohydrates and fats. Support for this assumption is
given by the experiments of Cohen and McGilvery (23) on the forma-
tion of /j-amino-hyppuric acid from /j-aminobenzoic acid and glycine
by rat tissue :

NH, ^^>COOH + NH,CH,COOH^ NH, €~>CO-NHCH,COOH + H,0

This may be taken as a model of peptide bond synthesis. This reaction
could not proceed anaerobically nor in the presence of oxidative inhibi-
tors (HCN, NaNs). Inhibition of this reaction by arsenite, iodoacetate
and malonate favors the view that the energy for the synthesis is pro-
vided by oxidations in the Krebs cycle where there are a number of
— SH enzymes. Whether this energy is furnished by the high-energy
phosphate bonds produced in these oxidations is not yet established. It
must be emphasized however that proteins possess a great specificity
and complexity. Besides enzymes forming peptide bonds there must be
others which add these bonds together to form the complicated and
varied structures of the different proteins.

Casper son (19, 20, 21) demonstrated that ribonucleic acid is present
in abundance in growing cells and gave some evidence in favor of the
view that it is essential for protein synthesis. The mechanism of action
of nucleic acids is not yet known. They must act in the last steps when
the architecture of the molecule acquires its specific characteristics.
Spiegelman and Kamen (80) believe that they are "the specific energy
donators which make possible reactions leading to protein and enzyme
synthesis." They found that new protein formation in yeast cells was
accompanied by a marked transfer of phosphate from the nucleoprotein.
However the phosphate groups in nucleoproteins are phosphate esters
which on hydrolysis provide small amount of energy. Northrop (64a)
has dealt extensively on this subject of the last stages of protein syn-
thesis.

126 E. S. GUZMAN BARRON

Regulatory Mechanisms

From this brief review of some enzyme reactions it can be seen that
what distinguishes biochemical reactions from chemical reactions in
general is the means provided for the rapid attainment of equilibrium,
so strikingly demonstrated by the enzymatic reversible oxidation-reduc-
tion systems. Furthermore, although on a thermodynamic basis a very
large number of reactions are theoretically possible in a given biologic
system, the presence of enzymes will result in the selective catalysis of
only a limited number of these reactions. But the isolation, purification,
and study of the catalytic action of individual enzymes is only the begin-
ning of the study of the biochemical reactions which occur in the cell,
whether it be in the state of "dynamic equilibrium" or in the state of
growth. Living matter is the result of the coordinated interplay of
substance and structure, as Pauling (71 ) has remarked, and the advance
of biology requires equal devotion to the work of isolation and identi-
fication of active chemical substances, to the study of the macromolecular
stromatic structures present in the living cell, and to the investigation
of the multiple regulating mechanisms which contribute to the cellular
dynamic equilibrium.

The rate of reaction of isolated enzyme systems is extremely high
when compared with the rate of reaction in the living cell. Enzyme
reactions are controlled in the living cell by a variety of regulating
mechanisms. To the usual physico-chemical factors which govern the
rate of reactions in solutions there must be added the effect of macro-
molecular structures which constitute the framework of the cell, such
as the genes, and which can orient or fix the position of the intracellular
enzymes by means of loose unions. As the cell increases in efficiency and
becomes more differentiated, the activity of the enzymes becomes more
and more controlled by specific regulating mechanisms, which in multi-
cellular organisms freely circulate through the intercellular fluid. In
vertebrates these regulatory mechanisms increase in number and speci-
ficity, and they control and orient the rate of enzymatic reactions by
interaction among each other. These regulatory mechanisms (hormones,
nucleoproteins, genes, electrolytes, certain oxidation-reduction systems)
do not act by forming part of enzyme components ; they act by control-
ling the rate of enzyme reactions or by changing the orientation of
these reactions. They act as the traffic policemen in the crowded streets
of a city. Thus the steady state of the adult cell is maintained, and the
synthetic processes required in growing cells are brought forth. An

CELLULAR METABOLISM I27

example of the presence of regulating mechanisms is seen in the effect
of dilution on the respiration of sea urchin sperm. The respiration of
sea urchin sperm is maintained at a low level under the regulating in-
fluence of some substances possessing — SH groups; on dilution or on
addition of iodoacetate or malonate (usual respiratory inhibitors), the
regulatory system becomes inactivated and the respiration is increased
(Fig. 4) (9). A striking demonstration of the role of hormones as
regulators of enzyme reactions, a theory postulated in 1943 (6) , is found
in the experiments of Price, Colowick, and Cori (72) of inhibition of
muscle hexokinase by a substance (hormone?) contained in the anterior
pituitary gland and the release of this inhibition by insulin. In the living
cell enzymes may be present which are so located as to be inactive in
the intact cell. Physiological alterations, rupture of the cell, destruction
of the normal architecture, may bring these enzymes into action. An
example of physiological alteration is given by the experiments of
Korr (44) who found that the respiration of the salivary glands, which
is insensitive to azide when the gland is in the resting state, becomes
azide sensitive when the gland is stimulated. An example of enzyme
liberation on damage to the cells is given by the marked and rapid
inhibition of the respiration of lung tissues after the tissue has been
ground (7). The QO2 value goes from 7.5 down to i.o. This inhibition
is due to destruction of the pyridine nucleotides by the nucleosidase
present in lung tissue in high concentration but kept inactive in the
intact cell. Evans (28), has dealt with the properties of the anterior
hypophyseal growth hormone isolated in his laboratory which plays an
important role in the synthesis of proteins, a role which in our opinion
is that of a regulating agent.

Cell Division

From the extensive investigations on the mechanism of cell division
it must be concluded that there is as yet no full understanding of the
physical and chemical mechanisms which bring forth this process. There
is agreement however that we are confronted with a series of events
which together culminate in the phenomena of membrane formation and
cell division. The initial process, formation of fertilization membrane
and entrance of the spermatozoon into the ovum with no nuclear
changes, may be called the atiaerobic phase, because it may take place
under strictly anaerobic conditions (3). If energy is required for this
process it must be obtained from glycolysis. The nuclear changes which

128

E. S. GUZMAN BARRON

Figure 4. Effect of iodoacetate on the O2 uptake of sea urchin sperm. Abscissa, time
in minutes. Ordinate, O- uptake in mm^. (i) Control (heavy suspension). (2) Heavy-
suspension with iodoacetate (o.ooiM). (3) Control (dilute suspension). (4) Dilute
suspension with iodoacetate (o.ooiM).

CELLULAR METABOLISM I29

follow membrane formation and end in cell division constitute the
aerobic phase and must require the energy provided by oxidative reac-
tions. The increased utilization of glycogen accompanied by the in-
creased utilization of pyruvate (Table X) may provide the energy nec-

TABLE X
Utilization of Pyruvate by Arbacia Eggs — Influence of Fertilization.

Date (1941)

Fertilization
(%)

Duration
(hrs.)

Pyruvate utilized

per gm. dry weight

per hr. (micrograms)

July 17

80

I

400

July 19

75

2

500

July 21

80

5

345

July 23

88

3

480

July 24

80

0.5

500

July 21

none

I

60

July 23

none

5

70

July 24

none

2

80

July 17

none

I

50

July 19

none

3

69

essary for the process of cell division (32). Furthermore, the intro-
duction during the cell division process of three of four oxidative steps
(the series of oxidations carried on by the cytochrome system) in the
series of reactions from substrate to molecular oxygen will increase the
efficiency of the reactions so that for the same amount of O2 utilized
there will be a more efficient utilization of energy. Whether carbo-
hydrate is metabolized entirely via glycolysis, or is oxidized through
the pathway of hexosemonophosphate oxidation as suggested by Or-
strom and Lindberg (70) is not known. Neither is the nature of the
acid formed on cell division known. Some of it may be acetic acid
formed on oxidation of pyruvic acid. Whether acetate is utilized or not
is not known. If it is, the pathway of its oxidation must be different
from that of most animal tissues because neither succinate nor a-keto-

130 E. S. GUZMAN BARRON

glutarate is oxidized by sea urchin eggs. Fluoroacetate had no effect
on the O2 uptake.

Cell Growth

Our knowledge of the enzymatic reactions which take part in cellular
growth is meager. In general cellular growth, new protein formation,
takes place at the expense of aerobic oxidations except in those cells
deprived of cytochromes.

Voegtlin (82), Hammett (33), and Rapkine (73) have for a long
time championed the importance of sulfhydryl groups in cell growth,
and evidence in support of this opinion is good. They have been found
increased in concentration in growing cells (62, 88) and growth was
inhibited on addition of substances which combine with — SH groups
(27, 34, 74, 76). It has already been pointed out that there are a large
number of enzymes which require the presence of — SH groups in their
protein moiety for activation. Destruction of these groups by oxidation
(x-rays, Oo, Cu), alkylations (CHJCOOH, CHJCOONH^, mustard,
nitrogen mustard), or mercaptide formation (Hg, Pb, Cd, Zn, Bi, As,
V, Sb) will produce enzyme inhibitions, an inhibition which affects
such high-energy yielding reactions as pyruvate oxidase, a-ketoglutarate
oxidase, succinoxidase, triosephosphate oxidase, adenosinetriphospha-
tase, malate oxidase. Sluggish oxidation-reduction systems, mostly in
their reduced state (glutathione, ascorbic acid), have also been found
abundant in cells during growth and development. These systems must
have for function that of keeping the — SH groups of the enzymes in
the reduced, physiologically active state.

It must be recalled that ionizing radiations (63), halogenated alkyla-
mines (30), and mustard gas (i) inhibit mitosis. Furthermore Hevesy
(36) has found that the synthesis of nucleoproteins (which take part
in mitosis) is inhibited by x-rays. Under certain conditions enzymes
like adenosinetriphosphatase from aged myosin are so susceptible to
ionizing radiations that amounts as small as ir produce inhibition
(Table XI). The inhibiting effect of x-rays may well indicate that
some enzyme necessary for the synthesis of nucleoproteins requires
— SH groups for its activity, groups which are oxidized by the OH and
O2H radicals, and H2O2 produced during the irradiation of water. The
inhibiting effect of mustard gas and of nitrogen mustard might also be
due to combination of these substances with the — SH groups. The fact
that lewisite produces no effect (i) is no serious objection, because

CELLULAR METABOLISM I3I

Barron et al. (12) have shown that rf-amino acid oxidase, a sulfhydryl
enzyme, was not inhibited by lewisite.

The close interrelation between the metabolism of carbohydrates, fats,
and proteins has shown that for the synthesis of protoplasm energy is
obtained by the aerobic oxidation of carbohydrate. The preponderant

TABLE XI

The Effect of Very Low Doses of X-Ray on Myosin (Aged).

Phosphorus liberated
in 10 min. at 38° (micrograms)

Dosage (r)

Treated

Control

Inhibition (%)

10

5
I

5-7
5.4
8.4

11.9
11.9
11.9

52
55
30

part of synthesis reactions over breakdown reactions in growing cells
must be due not only to the increased formation of enzymes but also
to the distribution of the regulating mechanisms which not only modify
the rate of reactions but are also able to change the direction of these
reactions. The study of these regulating mechanisms which channel the
multiple enzymatic reactions towards different pathways will clarify the
role of enzymes in growth processes and will help us in understanding
the change from normal controlled growth to the uncontrolled growth
of cancer.

REFERENCES

1. Auerbach, C, J. M. Robson, and J. G. Carr. Science, lofi, 247, 1947.

2. Ball, E. G. Ann. New York Acad. Sci, 45, 363, 1944.

3. Barron, E. S. G. Biol. Bull., 62, 42, 1932.

4. Barron, E. S. G. /. Biol. Chem., 97, 287, 1932.

5. Barron, E. S. G. /. Biol. Chem., 121, 285, 1937.

6. Barron, E. S. G. Adv. Enzymology, j, 149, 1943.

7. Barron, E. S. G., Z. B. Miller and G. R. Bardett. /. Biol. Chem., 171,

791. 1947-

8. Barron, E. S. G., S. R. Dickman, and T. P. Singer. Feder. Proc. Soc.

Biol. Chem., 6, i, 1947.

9. Barron, E. S. G., and J. M. Goldinger. Proc. Soc. Exp. Biol, and Med.,

48, 570, 1 941.

132 E. S. GUZMAN BARRON

10. Barron, E. S. G., and L. A. Hoffman. /. Gen. Physiol., 13, 483, 1930.

11. Barron, E. S. G., and C. P. Miller. /. Biol. Chem., p/, 691, 1932.

12. Barron, E. S. G., Z. B. Miller, G. R. Bartlett, J. Meyer, and T. P.

Singer. Biochem. J., 41, 69, 1947.

13. Barron, E. S. G., and T. P. Singer. /. Biol. Chem., 157, 221, 1945.

14. Bartlett, G. R., R. Ahrens, and E. S. G. Barron. /. Biol. Chem., 170,

67, 1947.

15. Belitzer, V. A., and E. T. Tsibakowa. Biokhtmtya, 4, 516, 1939.

16. Bergman, M., and J. S. Fruton. Advances in Ensymology, i, 63, 1941 ;

Ann. New York Acad. Sci., 45, 409, 1944.

17. Borsook, H. Ergilo. Ensymforsch., 4, I, 1935.

18. Bucher, H. Naturwiss., 30, 756, 1942.

19. Casperson, T. Chromosoma, i, 562, 1940; Naturwiss., 17, 33, 1941.

20. Casperson, T., and K. Brandt. Protoplasms, 35, 507, 1940.

21. Casperson, T., and L. Santerson. Acta Radiol. Supp., 46, 1942.

22. Clark. W. M., J. F. Taylor, T. H. Davies, and C. S. Vestling. /. Biol.

Chem., 135, 543, 1940.

23. Cohen, P. P., and R. W. McGilvery. /. Biol. Chem., 166, 261, 1946.

24. Colowick, S. P., M. S. Welch, and C. F. Cori. 7. Biol. Chem., 133, 359,

641, 1940.

25. Dakin, H. E. J. Biol. Chem., 6,221, 1909.

26. Dale, W. M. Biochem. J., 34, 1367, 1940.

27. Ephnisi, B. Compt. Rend., ip2, 1763, 1931.

28. Evans, H. M. Am-. Scientist, Vol. 35, p. 465, 1947.

29. Eyring, J. J. Chem. Physics, 3, 107, 1935.

30. Friedenwald, J. Personal communication, 1943.

31. Glasstone, Laidler, and J. Eyring. The Theory of Rate Processes.

McGraw-Hill Book Co., 1941.

32. Goldinger, J. M., and E. S. G. Barron. /. Gen. Physiol., 30, 73, 1946.

33. Hammett, F. S. Cold Spring Harbor Symp. Quant. Biol., 2, 78, 1934.

34. Hammett, F. S. Protopl., 7, 297, 1929.

35. Hellerman, L., F. P. Chinard, and V. R. Dietz. 7. Biol. Chem., 147,

443. 1943-

36. Hevesy, G. Rev. Med. Physics, 17, 162, 1945.

37. Hopkins, F. G., and E. J. Morgan. Biochem. J., 32, 611, 1938.

38. Hunter, F. E., and F. L. Leloir. 7. Biol. Chem., 159, 295, 1945.

39. Kalckar, H. Ann. New York Acad. Sci., 45, 395, 1944.

40. Kalckar, H. Ann. Rev. of Biochem., 14, 283, 1945.

41. Kalckar, H. Enzymologia, 2, 47, 1935; 5, 365, 1939.

42. Kalnitsky, G., and E. S. G. Barron. 7. Biol. Chem., 170, 23, 1947.

43. Knoop, F. Beit. Chem. Physiol. Path., 6, 150, 1904.

44. Korr, J. M. Am. J. Physiol., 133, 355, 1941.

45. Krebs, H. A. Advances in Ensymology, 3, 171, 1943.

46. Krebs, H. A., and W. A. Johnson. Biochem. J., 31, 645, 1937.

47. Kuhn, R., and P. Boulanger. Ber. Chem. Ges., 6g, 1557', 1936.

48. Lardy, H. A., and J. A. Ziegler. 7. Biol. Chem., 139, 343, 1945.

49. Lehninger, A. 7. Biol. Chem., 143, 147, 1942.

CELLULAR METABOLISM I33

50. Lehninger, A. /. Biol. Chem., 157, 363; 161, 437, 1945.

51. Lehninger, A. /. Biol. Chem., 162, 333, 1946.

52. Lehninger, A. /. Biol. Chem., 164, 291, 1946.

53. Lipmann, F. Advances in Enzymology, i, 99, 1941.

54. Lipmann, F. Advances in Enzymology, <5, 231, 1946.

55. Long, C. Biochem. J., 32, 171 1, 1938.

56. Madden, S. C, and G. H. Whipple. Physiol. Rev., 20, 194, 1940.

57. MacKay, E. M. /. Clin. Endoc, 3, loi, 1943.

58. Meyerhof, O., P. Ohlmeyer, W. Gentner, and H. Maier-Leibenitz.

Biochem. Z., 2p8, 376, 1938.

59. Meyerhof, O., P. Ohhiieyer, and W. Mohle. Biochem. Z., 2g/, 113,

1938.

60. MichaeUs, L. Transact. Electrochem. Soc, //, 107, 1940,

61. MichaeHs, L., and M. L. Menten. Biochem. Z., 4p, 333, 1913.

62. Miller, B. J., and S. P. Reimann. Arch. Path., 2p, 181, 1940.

63. Miiller, H. J. Verh. Vint. Kong. Verarbungsw., 234, \^2'j.

64. Muiioz, J. M., and L. F. Leloir. /. Biol. Chem., 147, 355, 1943.
64a. Northrop, J. This symposium.

65. Ochoa, S. /. Biol. Chem., 138, 751, 1941 ; 131, 493, 1943.

66. Ochoa, S. /. Biol. Chem., 155, 87, 1944.

67. Ochoa, S. /. Biol. Chem., 159, 243, 1945.

68. Ochoa, S., A. Mehler, and A. Kornberg. /, Biol. Chan., 16/, 871, 1947.

69. Ochoa, S., and E. Weisz-Tabori. /. Biol. Chem., 159, 245, 1945.

70. Orstrom, A., and O. Lindberg. Enzymologia, 8, 367, 1940.

71. Pauling, L. Chem. and Engin. News, 24, 1375, 1946.

72. Price, W., S. P. Colowick, and C. F. Cori. /. Biol. Chem., 160, 633,

1945-

73. Rapkine, L. Ann. Physiol., 7, 382, 1931 ; /. Chim. Phys., 33, 493,

1936; J-/, 416, 1937-

74. Rapkine, L., E. Chatton, and A. Lwoff. Compt. Rend. Soc. Biol., 106,

626, 1 93 1.

75. Rittenberg, D., and K. Bloch. /. Biol. Chem., 160, 41"/, 1945.

76. Ruffili, D. Boll. Soc. ital. Biol, sper., 77, 75, 1942.

yy. Schoenheimer, R. Dynamic State of Body Constituents. Harvard.
Univ. Press, Cambridge, 1942.

78. Singer, T. P., and E. S. G. Barron. /. Biol. Chem., 757, 241, 1945.

79. Saunderhoff, R., and H. Thomas. Ann. der C hemic, 530, 195, 1937.

80. Spiegelman, S., and M. D. Kamen. Science, 104, 581, 1946.

81. Stumpf, P. K., K. Zarudnaya, and D. E. Green. /. Biol. Chem., i6j,

817, 1947.

82. Voegtlin, C. Cold Spring Harbor Symp. Quant. Biol., 2, y^, 1934.

83. Warburg, O., and W. Christian. Biochem. Z., 303, 40, 1939.

84. Weil-Malherbe, H. Biochem. J., 31, 2202, 1937.

85. Weinhouse, S., G. Medes, and N. F. Floyd. /. Biol. Chem., 158, 411,

1945-

86. Weinhouse, S., G. Medes, and N. F. Floyd. /. Biol. Chem., 755, 143,

1944.

134

E. S. GUZMAN BARRON

87. Weinhouse, S., G. Medes, and N. F. Floyd. /. Biol. Chem., 166, 691,

1946.

88. Owen, S. E., H. A. Weiss, and L. H. Prince. Am. J. Cancer, S5, 424,

1939-

89. Wynne-Jones, and H. Eyring. /. Chem. Physics, j, 492, 1935.

VII. DIFFERENTIAL GROWTH

BY PAUL WEISS^

THE appearance of different properties in initially equivalent cells
is called "differentiation." Growth being only one cellular attri-
bute among many, "differential growth" is to be subsumed under
the larger topic of "differentiation," and our discussion therefore will
have to center on the latter. Since the accent of this book lies more on
prospects than on retrospects, I intend to concentrate on the gaps in our
concepts and knowledge of differentiation and point to possible ways of
filling them rather than dwell on past achievements, which, though im-
pressive in themselves, are dwarfed by the task that remains to be ac-
complished.

Biology has discovered the strength it can derive from a more rigorous
analysis of biological phenomena by the tools and standards of the
physical sciences, and if the study of development is to share the new
light with its more alert sister branches, such as physiology, genetics and
immunology, to name only a few, it will have to submit to the same
reorientation that has proved so beneficial with the others. This reorien-
tation is in the direction of greater precision, objectivity, and con-
sistency in the description and interpretation of phenomena, prefer-
ably in terms of physical and chemical order.

It implies the adoption not only of the tools of the physical sciences,
but above all of their disciplined methodology. Analytical embryology
is still full of problems which it can only settle within its own right
and by its own techniques. But in doing this it must apply standards,
criteria, and methods of procedure as rigorous as those of the physical
sciences. It must strive to replace the abstract formalism and verbalism
of its groping infancy by a mature realism, in which a term stands
as symbol for a known or knowable entity or relation, rather than as
a soothing device of convenient elasticity to pretend knowledge where
there is none. We must not allow its analytical principles to be con-
taminated by relapses into animistic mythology, as we are doing when
we charge the unresolved residue of meticulous determinations of
physico-chemical properties to "organizers," "inductors," and similar
ill-descript agents. Labels must no longer be allowed to pass for con-
tent, nor generalities for explanations. What we need most in entering

1 University of Chicago. Original work referred to in this article was aided by the
Dr. Wallace C. and Clara A. Abbott Memorial Fund of the University of Chicago.

136 PAUL WEISS

the era ahead is the will to drive the analysis of developmental phe-
nomena far beyond the mark where we used to stop to coin a term,
and the determination to become as exact, specific, objective, and realis-
tic in the formulation of our problems and in the description of our
observations and results as has become the rule in the physical sciences.
This is what we mean by reorientation. It is just as important as is
the gathering of more data. With this in mind, we turn to the question
which forms the core of my assignment: What is differentiation?

What Is Differentiation?

By the standards just advocated, this question can immediately be
recognized as far too indefinite to be answered by anything but another
generality. For "differentiation" refers not to a single circumscribed
natural phenomenon, but is a summary name for a heterogeneous col-
lection of notions about a variety of events associated with develop-
ment. Consequently, the various statements and generalizations which
have been made in the past concerning "differentiation" and "dedif-
ferentiation" are equally heterogeneous and full of contradictions.
Only a few authors have deemed it necessary to be explicit about terms
of such general currency. In order not to add to the confusion, I shall
begin by developing my own definition. Then whatever I shall have to
say about "differentiation" will refer to the phenomenon as here defined,
and not to anybody else's "differentiation."

A definition of "differentiation" is partly a matter of objective de-
scription, partly of plain logic. We shall deal with the logical portion
first, because it has been more persistently ignored.

Characters vs. Processes

One source of confusion is the tradition of confounding the process
of differentiation with its criteria. We usually tell differentiation by its
more conspicuous products; e.g. melanin granules in a pigment cell,
myofibrils in a muscle fiber, secretion granules in a gland cell. Our
criteria are predominantly morphological and we judge mostly by ap-
pearances. In the examples just mentioned, our judgment happens to
be well founded. It becomes dubious when based merely on superficial
differences of shape, arrangement, or size, and it misleads us completely
when we try to reverse the argument and take the absence of obvious
differences for evidence of lack of differentiation. To appraise a cell
correctly we must take into account all of its properties, not just the

DIFFERENTIAL GROWTH 1 37

ones that impress us optically, and to appreciate the development of
these properties we must bear in mind that we are dealing with material
systems in space and time, in which whatever is present at any one stage
is the result of antecedent processes. Thus the conventional "charac-
ters" by which we distinguish differentiation are in reality only inci-
dental manufacturing samples of the cellular production plant. And
just as one kind of industrial plant differs from another not only by its
different end products but by the whole machinery and processing
method turning out those products, so a differentiated cell of one kind
differs from that of another kind not only in its final equipment (pig-
ment, fibrils, secretions, etc.) but in the basic mechanisms through
which that equipment has been produced.

Tracing pigment cells, muscle cells, or blood cells back into earlier
phases of their development, we find, or postulate, that they contain
then, instead of the terminal products of melanin, myosin, or hemo-
globin, half-finished "precursors." But if we go back still further,
beyond the "precursors of precursors," the future products lose their
identity and we are faced with nothing but the plant capable of turning
them out, and perhaps some of the ingredients. Now, evidently a plant
designed to turn out chiefly melanin must be different from one de-
signed to make contractile myofibrils. We express this fact in the names
"melanoblast" and "myoblast" given to those early stages, and we
generally concede that they are intrinsically different, although the dif-
ferences are no longer demonstrable by ordinary microscopic or chem-
ical techniques. By further refinements of technique we can extend
the range of our discriminative powers ; structurally, by the ultramicro-
scope, polarization microscope, electron microscope and x-ray diffrac-
tion; chemically, by microtests (including cytological stains), spec-
troscopy, microincineration, etc. These techniques will undoubtedly
reveal criteria of differentiation much earlier in the embryonic history
than was possible with the cruder classical tools of observation. This
is a line of investigation worthy of vigorous prosecution. But we must
expect that it too will come to an end short of retracing the whole
ontogenetic course of a given cell strain, for we are still essentially
concerned with tracing "products" rather than the production processes
that lie behind. Once we can detect the "product," most of the story
of its production is over and we have certainly missed its essential be-
ginnings. It looks as if the analytical methods at hand are of no avail.

Preformistic theory would deny the validity of our dilemma. It
would assert that all distinctions of later stages have existed, though

138 PAUL WEISS

less manifestly, from the beginnings of ontogeny. But experimental
embryology has thoroughly discredited this concept by proving that
mature cells which differ profoundly in their physico-chemical consti-
tution may be derived from cells that are demonstrably equivalent and,
to all practical purposes, identical. When and by what means have their
developmental courses then become divergent? This is the crucial ques-
tion of cytodifferentiation. Fortunately, though analytical techniques
fail us, we can answer it by a test that might be called "behavioral."
This is where we must draw on logics rather than gadgets.

Suppose we are to compare two closed material systems, whether
they be molecules or cells or organisms, about whose character and
composition we know nothing. How are we to tell whether they are
the same or different? The only way of telling is by watching their
behavior. If both behave identically under identical conditions, we de-
cide that they are alike. However, if both, again under identical con-
ditions, behave differently, then we must conclude that they have been
intrinsically different from the very beginning of the test.

Against the cogency of this postulate even the most persistent fail-
ures to detect more tangible signs of difference would prevail nothing.
Two seeds giving rise to different plants in the same soil, or two eggs
giving rise to different animals in the same pond, would have to be
acknowledged as intrinsically different even if they were otherwise
indistinguishable. Similarly, if two cells under identical conditions
behave and respond differently in any respect (e.g. movement, growth
rate, chemical activity, sensitivity to radiation or drugs, etc.), then they
are different, which implies that if they have descended from a com-
mon source of identical cells, they must have become different, i.e. have
gone through a process of differentiation, no matter whether they show
it by morphological signs or not. The behavioral test is as compelling
as any morphological test, and far more pertinent; for morphological
criteria are but residues and convenient indicators of prior activities.

Early differences of this type, which are not immediately recogniza-
ble and can only be deduced from later formative activities, are com-
monly described as differences of "potency." If this were to imply
that they are virtual rather than real (in the sense of Driesch), then
it would mean reinstating the old "vital spirits" under another name.
Actually there can be no doubt that we are dealing with differences of
equipment and conditions which are real, yet too subtle to be detectable
by present techniques.

In conclusion, we propose to define "differentiation" as the unidirec-

DIFFERENTIAL GROWTH I39

tional changes in the character of cells and cell generations during their
life history, transformations by which they become increasingly dif-
ferent from their own former selves, their parent cells, and the cells of
other strains which have taken divergent courses. And, as we have
already intimated, "character" is to include the totality of reactions
of which the cell is capable at a given stage. In any given situation,
only a fraction of these possible reactions will be realized, and only a
small part of this fraction will become externally recognizable.

Modulation

We have said that cells that look alike may be intrinsically very
different. The opposite is equally true, namely, cells may look very dif-
ferent and yet be essentially alike in character. Let me cite a striking
example from my own experience, the vagaries of the adult sheath
cell of Schwann of adult mammalian nerve. In the embryonic stage
these cells are long spindles. They then envelop the nerve axons as thin
protoplasmic cylinders. When axons degenerate, as after nerve section,
the sheath cells regain mass, become spindly again, and glide out from
the nerve tubes. After nerve regeneration they resume their original
sheathing position and corresponding cylindrical shape. But they can
vary even much more than just shuttling back and forth between these
two states.

Embryonic Schwann cells, reared in tissue culture along the inter-
face of cover glass and fluid medium, tend to transform from the
spindly type into a flat round form of the appearance and behavior of
macrophages (49). A repetition of the experiments with adult nerve
proved that the fully mature Schwann cell is capable of the same trans-
formations and more (58). Plate I illustrates a series of cell forms
obtained in vitro from Schwann cells of peripheral nerve. ^ They include
tandem strands of filamentous sheath cells, ramified astrocyte-like
forms, giant macrophages with rufiled membrane in active phagocyto-
sis, monocyte-like round cells with clear ectoplasm and eccentric nucleus,
and foamy round cells whose vitality is proven by their mitotic activity.
Each one of these phases occurs under certain definable conditions,
among which the fine-structural constitution of the medium seems to
play a predominating role. Some of the transformations are easily
reversible, others have as yet been observed in one direction only. Since
they can be followed directly as they take place in the cultures, valuable

2 Save for a brief note, the results of these experiments have not yet been published.

I40 PAUL WEISS

insight into the processes underlying the morphological results has
been obtained. There was evidence of significant changes in the distri-
bution of the cell content, size of the ectoplasmic border, constitution,
elasticity, rigidity, mobility, etc., of the cell surface, average number
and shape of pseudopodia, mode and rate of locomotion, viscosity,
vacuolization, adhesiveness, orientation of fibrillar plasma constitu-
ents (as indicated by silver impregnation), fat production, distribution
of nuclear materials, shape and hydration of nucleus and nucleoli,
phagocytic power, and other associated properties. In short, these cells,
in response to different environments, have undergone rather profound
reorganizations of the kind unhesitatingly classified as "differentiation"
or "metaplasia" by a nomenclature based on sheer appearances.

But is this type of change directly comparable to the progressive
cytodifferentiation during ontogeny of which we have spoken before?
It evidently is not, and it is in this place that facts supplement logic.

As we mentioned before, many of the observed changes are reversi-
ble if the cell is returned to its prior environment. There are many more
cell types of the mature organism which behave similarly in that their
behavior and shape change reversibly with their environment. Changes
in the nutrient and hormone composition of the internal milieu, sub-
lethal toxic agents, inflammatory processes, seasonal fluctuations, de-
nervation, and in fact, on a more moderate scale, practically all ordinary
physiological stimulations, cause a more or less thorough reshufBing
of the cell content with consequent change in cell behavior, and often
in cell form. The important point to keep in mind is that in all these
instances the cell can revert to its original state when the original en-
vironmental conditions are restored. Evidently the cell passes through
these transmutations with no alteration, either gain or loss, of its basic
equipment. At the end of a cyclic change it turns out to be the same
cell that it was before, and it is quite beside the point that sometimes
the reversal may not occur in the same cell, but in one of its descendants
which then prove to be "chips off the old block." There is no reason
why the reversible changes observed in our Schwann cells should be
regarded in any difTferent light than, let us say, the contractions and
expansions of a melanophore in response to cyclic nervous or humoral
stimuli, or the change in the secretory state (and associated morphol-
ogy) of a hormone-sensitive cell in response to varying concentrations
of the respective hormone. They all are merely expressions of how a
given cell can react to a variety of conditions — its "reaction repertory,"
based on its material constitution such as it is at the time. The living

DIFFERENTIAL GROWTH I4I

cell is an unstable system which can have neither existence nor form
except in relation to its environment. Few animal cells (e.g. red blood
cells) freeze their shape by assuming a rigid envelope. For the rest,
shape is variable, merely an index of substance disposition attained in
given surroundings and valid only for these particular surroundings.

Obviously, then, the misleading mental picture of absoluteness and
static rigidity of cell shape, conjured by the microscopic picture of
fixed cells on slides or in textbooks, must be replaced by a dynamic
definition based not on the momentary expression of the cell but on
all the possible reactions of which it is capable. The variety of expres-
sions that thus can be assumed by the cell on any level of differentiation
I have called "modulations" (4, 47). If we may draw a chemical anal-
ogy, differentiation would be comparable to the synthesis of a new
compound by an irreversible reaction chain, while modulation would
correspond to the isomerism and allotropy of intermediate or terminal
products. The former involves a change of composition, the latter merely
of disposition in space, of the constituent elements. The implications of
this simile for the concept of the cell will be dealt with later.

It can be plainly recognized now how past concepts of differentia-
tion have been vitiated by undue reliance on visual criteria. Cells that
looked alike were thought to be essentially of like character, although
they often were radically different; and conversely, cells that looked
very different were considered as very disparate, although they actually
were often of the same kind. The literature of histology, embryology,
and pathology abounds with notable examples of victims of this "visual
illusion."

Differentiation and Dedifferentiation

If the test of differentiation, in contradistinction to modulation, is
the unidirectional, progressive, and irreversible trend of the former,
how safe is the evidence that differentiation in this strict sense ever
really occurs? Might not all differentiations be cases of modulation,
merely obscured by the difficulty of always finding the right conditions
to reverse the fate of the cell ?

As long as we confine ourselves to normal development, where
different cell behavior is generally associated with different local
conditions, the answer remains in doubt. For a crucial test of whether
different cells differ in character, we must remove them from their
different environments and observe them (a) in identical environ-

142 PAUL WEISS

ments, and (b) in each other's places. This can be done by the techniques
of explantation (tissue culture) and of transplantation, respectively.
The answer has been unequivocal.

Cells taken from different tissues of late embryonic or mature organ-
isms and transferred into standard extraneous media offering nutrition,
protection, and support lose many of their specialized aspects, i.e. criteria
by which we used to tell their "differentiation." Spectacular as their dif-
ferences may have been during residence in the body, they soon come to
look and behave much more alike, in adaptation to the common environ-
ment. In the early days of tissue culture this simplification of appearance
was actually often interpreted as a sign of "dedifferentiation," in the
sense of a return not only to a more primitive but to a veritably pri-
mordial state. This was again judgment by appearances with all its
pitfalls. Yet later work showed clearly that the explanted cells, in spite
of their similar disguise, retain distinguishing features of fundamental
nature which they pass on unaltered to their descendant cells for in-
definite numbers of generations. As a glance at Plate I will illustrate,
Schwann cells and endoneurial connective tissue cells, while assuming
both the aspect of common "mesenchym," i.e. spindle shape and amoe-
boid activity, can be well distinguished by their sizes, nuclear charac-
teristics, staining properties, affiliations, and other criteria revealed only
by subtler analysis.

In other words, the parallel modifications which the cell strains under-
go on transfer to tissue culture are merely modulations carried out by
variously "differentiated" cells that otherwise preserve their basic dis-
tinctions. Their distinctiveness persists even during such further modu-
lations in culture as the transformation to the macrophage stage. This
transformation has thus far been observed in cultures of various con-
nective tissues (25), Schwann cells (see above), and muscle (7). They
all can give rise under similar conditions to a highly phagocytic amoeboid
cell with hydrated nucleus, dense entoplasm, and a more hyaline ruffled
border, i.e. macrophage-like characters. But in doing so they also retain
the signs of their descendancy, so that a "macrophage" resulting from
the conversion of an endoneurial cell can usually be clearly distinguished
from one that has come from a Schwann cell.

Tissue culture thus does not really abolish character differences
established during ontogeny. This conclusion is fully borne out by the
critical literature on the subject (4, 40). It is strikingly confirmed by
experiments in which cells that had lost their more specialized aspects
during intensive proliferation in tissue culture were again given oppor-

DIFFERENTIAL GROWTH I43

tunity to produce their specialties, either by suppressing further pro-
liferation in vitro or by transplanting them back into an organism.
Under such conditions they were essentially true to the kind from which
they stemmed: those derived from pigment epithelium would resume
the production of pigment granules (8) ; those from thyroid, of colloid
(13) ; those from gut, of digestive enzymes (14). Cultured cells re-
turned to an organism, but into atypical locations (42), moreover,
showed no tendency to differentiate in conformity with their new en-
vironment, proving that the culture period had not restored to them the
ability of younger cells to switch into a new local course of differentia-
tion.

All this adds up to the realization that the measure of true differ-
entiation is the absence of true dedifferentiation, if dedifferentiation
connotes not the mere assumption of a semblance of primitivity but the
actual return to a state of wider potency, i.e. the recuperation of some-
thing that had ostensibly been lost. By the testimony of the reported and
similar experiments, vertebrate cells are incapable of such true dedif-
ferentiation ; hence, for them, true differentiation is the rule.

Tissue Differentiation

All our statements thus far refer to cytodifferentiation. Histodif-
ferentiation and organ differentiation require some additional comment.
Tissues and organs consist of cells and matrix in definite mutual rela-
tions and groupings. What happens during the differentiation of a
tissue is in large measure a direct reflection of the differentiation of its
constituent cells. Their order of proliferation, their movements, trans-
formations, affinities, secretions, etc., and the spatial arrangements and
restraining conditions resulting from group occupancy of a common
space determine, for the main part, the character and organization of the
tissue or organ. By their collective activity the cells set up conditions for
one another that would not develop if they remained single and inde-
pendent. For this reason, what is commonly called "histodifferentia-
tion" implies an added element of order and novelty over the mere
"cytodifferentiation" of the unit cells. This is particularly striking in the
formation of composite organs where the combination of cells from
diverse sources leads to novel results which neither of the contributing
components could have produced by itself. In addition to material contri-
butions, there are dynamic influences of mechanical, chemical, and
electrical nature acting from one tissue or organ upon another.

144 PAUL WEISS

But on close inspection there remains no property of a tissue that
could not be ultimately traced to the activity of either its own or other
cells. It is by virtue of properties acquired in cytodifferentiation that
cells can combine, interact, and arrange themselves in certain specific
ways, conditions permitting. "Histodifferentiation" implies merelv the
existence of conditions under which cytodififerentiated cells can realize
their individual and group faculties. This helps to clarify the meaning
of the term "tissue dedifferentiation." Evidently cells may lose their
tvpical associations and orderly arrangements just as we have seen be-
fore that they may lose internal differentiation products, without giving
up their essential character. Destruction of intercellular matrix, for
instance, can set cells free, and such cells mav resume motility, scatter,
and modulate in various ways. Their community has been dissolved, but
they themselves have retained their identity. To this extent, what is
called "tissue dedifferentiation" is merely tissue disintegration. Since
mobilization of cells is frequently attended by the dissolution of internal
differentiation products (e.g. myofibrils), the superficial picture becomes
one of general simplification and regression. There is no harm in con-
tinuing to speak of tissue dedifferentiation, so long as one bears in mind
that it need involve no true dedifferentiation of cells.

Which Cell Parts Differentiate*?

A cell is a highly complex and heterogeneous system. Then where
specifically do the changes occur that constitute differentiation? In the
cytoplasm, in the nucleus, or in both? Or rather, which ones of their
subdivisions undergo change, and which do not?

It is customary in speaking of cytodifferentiation to emphasize mainly
the concomitant profound changes in the content of the cytoplasm. But
the nucleus likewise undergoes patent modifications, which become
evident in the divergence of nuclear characters and behavior among
different cell types. Some of these nuclear distinctions can be deduced
from differential reactions to histological stains, which in essence con-
stitute microchemical tests. In special instances even more direct signs
of nuclear differentiation have been observed, as in the extrusion of
formed secretion bodies from the nuclei of certain nerve cells (30), or
in the transfer of the major mineral content from nucleus to cytoplasm
during the maturation of neuroblasts (36). On the other hand, it is
a basic tenet of genetics that the chromosomes, or at least those parts
of them that constitute the genes, remain unaltered in the process of

DIFFERENTIAL GROWTH I45

somatic differentiation. Accordingly each nucleus contains both "dif-
ferentiated" and "undifferentiated" portions.

The question arises whether the cytoplasm might not be similarly
subdivided into a differentiated fraction and an undifferentiated residue.
Such is generally conceded to be the case for the cytoplasm in the "germ
track," i.e. of those cells that lie in direct ascendancy of the primordial
germ cells. The same must be assumed for those "totipotent" somatic
cells of the lower invertebrates which can give rise to a new individual
by budding, regeneration, or other forms of asexual reproduction. The
specialized somatic cells of higher forms, on the other hand, have never
given any evidence of such totipotence. This could be interpreted to
indicate that all of their cytoplasm has been diverted into a specialized
course, has become liver-specific for the liver cell, kidney-specific for
the kidney cell, and so forth. Yet again, there is undeniably some stock
'of cytoplasmic equipment that practically all cells have in common (e.g.
respiratory enzymes, mitochondria, centrosomes, spindle fibers, etc.) and
which, thus, seems to have escaped the progressive changes of cytodif-
ferentiation.

In short, any attempts at making general statements for the nucleus
as a whole or the cytoplasm as a whole end up in contradictions. The
more concretely and specifically we try to state our problems, the more
clearly we realize the inadequacy and ambiguity of our traditional
vocabulary. So long as we address our questions not to a real living cell,
but to a vague and abstract symbol, we can expect no more than oracular
replies. Those general concepts of cell and cytoplasm with which we
conventionally operate, with their sundry textbook attributes of con-
tractility, excitability, adaptability, and so forth, have just about lived
out the term of their usefulness. We no longer view contractility and
excitability vaguely as "fundamental attributes of protoplasm," but
have recognized them as defined properties of circumscribed physico-
chemical systems, protein chains and surface membranes, respectively.
Differentiation deserves a similarly specific treatment. This will be
feasible only if we replace the abstract notion of the cell by a much more
specific picture. Such a picture is gradually emerging from the combined
efforts of cytology, general physiology, and biochemistry. According to
this picture, a cell is made up of known or knowable molecular elements,
and cellular activities are resolvable into elementary molecular processes.

Study of these elementary processes is well under way and has yielded
spectacular results. However, preoccupation with the elements has also
at times tended to obscure the fact that in order to constitute a living

146 PAUL WEISS

cell the constituent molecular processes must not just happen, but must
take place according to a highly ordered plan of interrelations based on
orderly spatial distributions, orderly chronological sequences, and or-
derly ratios of rates. No model of a cell can be pertinent unless it takes
into account both the elementary processes and their organizational
frame. It is imperative, therefore, that we match research on the former
by an equally vigorous search for the physical basis of the latter. To
facilitate this task I have tried to sketch a sort of crude molecular
blueprint of the cell, in which the molecular phenomena are placed in
relation to the organisation of the cell as a unit. This picture is of
necessity diagrammatic, oversimplified, tentative, and full of blank
spaces, but it does serve to crystallize the problems. Most of it has
already been presented on an earlier occasion (54).

Molecular Ecology

Science advances not only by discovering new truths but also by
eliminating old errors. Most older cell concepts held to the view that
cellular organization is based on rigid structural frameworks of micro-
scopic order, but this view has been ruled out by modern cell research.
Cytoplasm behaves as a viscous liquid. It can be thoroughly mixed up by
stirring or centrifugation, and yet its countless chemical workshops will
continue to operate in good order, which implies proper spatial segre-
gation. No orderly microscopic array could have survived in the com-
mingled content; what has survived is a fixed set of conditions — the
rule, as it were — according to which the original order can be restored.
And where do those conditions reside ? Evidently they must be looked for
in those parts of the system which can preserve organization in the
shuffie. There are two categories of cell constituents that qualify under
this title, and they are of quite different order of magnitude, one molecu-
lar and the other supramolecular. The former consists of the large
organic molecules or molecular aggregates with their specific structure
and faculty to perpetuate this structure in the presence of the necessary
ingredients and appropriate sources of metabolic energy. The latter
consists of the existing surfaces and interfaces of microscopic or sub-
microscopic order, whose molecular linings are anchored by adsorption
or chemical fixation and thus better protected against random disrup-
tion.

Organization within any sample space of a cell is then to be regarded
as a result of the counterplay between these two poles: the specific

DIFFERENTIAL GROWTH 1 47

faculties for selective combination and segregation of the molecular
constituents on the one hand; and the marginal reference frames of
surfaces on the other, which, by their power of adsorbing and orienting
the mobile elements, force them into a supraelemental collective order.
It seems impossible to account for cellular organization otherwise than
by such a dual concept. Its essence can be summarized as follows :
Manifest cell organization results from the response of organized ele-
ments to fields of organized (i.e. non-random) physical and chemical
conditions, here tentatively identified with conditions prevailing along
interfaces.

The nearest simile to this type of organization is found in the be-
havior of human or animal populations, which likewise consists of the
reaction of organized individuals to an organized environment. In both
cases, the order is one of ecological conditions. Thus, cell biology, in
molecular terms, becomes not merely molecular physics and chemistry
of the cellular constituents, but molecular ecology (54), a science not of
molecular individuals or pairs, but of molecular populations and their
existential conditions.

Some day we may hope to have a fairly complete list of the chemical
compounds present in any given cell, from simple elements to the highly
complex organic macromolecules, and thus develop a taxonomic cata-
logue of molecular species. But just as a museum collection of animals
is never lifelike unless it shows them in their natural environment and
mutual ecological relations, so this chemical catalogue would not be
representative of the cell unless it were supplemented by information on
the groupings, behavior, interdependence, and general operating con-
ditions that tie the various molecular species into a durable, and indeed
"viable," community.

Biochemistry is producing an ever-growing wealth of evidence for
the complexity of molecular interdependence in living systems. Even
very elementary physiological processes, as for instance the energy
transfer in respiration or photosynthesis, involve a great number of
coordinated steps, each of which requires the presence of a definite set
of properly dosed reactants, mediators (enzymes), and physical con-
ditions (/jH, temperature, pressure). Undisturbed operation of these
energy-yielding systems, in turn, is a prerequisite for the synthesis,
among others, of the complex protein molecules, some of which, as
enzymes, are again fed back as indispensable links into the energy-pro-
ducing systems. Thus neither system can maintain itself without the
other. Protein synthesis itself proceeds in series of steps, each of which

148 PAUL WEISS

can occur only under specified conditions which will vary for different
compounds. The various chemical systems of the cell thus form "sym-
biotic" interdependences, and the integrity of the cellular system as a
whole is contingent on their coexistence and cooperation. For each
particular chemical system we can thus postulate a definite set of "exis-
tential and operational prerequisites," that is, conditions indispensable
for the synthesis, preservation, and operation of specific organized
products. The character of a given animal population is determined by
the physical environment (climate, physiographic factors, etc.) and by
the organic environment (food, competition among species, biocoenosis,
etc.) in that particular area. Similarly, the character of the molecular
population occupying a given site in a cell will depend on the physical
conditions (electric potentials, surface tensions, etc.) and the chemical
interactions among the various molecular species in that locality.

The ecological simile is primarily a convenient device to describe the
behavior of the molecular populations of cells. It permits us, for instance,
to answer in principle the question of how the morphological segrega-
tion and orderly distribution of different chemical systems in the cell
is brought about, preserved, and restored after disturbance, despite the
absence of static internal frameworks. If each complex biochemical
system requires highly specific conditions for its formation and opera-
tion, then obviously the probability of finding that kind of system will
be high at sites where those conditions are satisfied, and low at others.
And after commingling, the content will again sort itself out according
to the frame of marginal conditions, and if the latter have remained
undisturbed, the re-sorting will likewise follow the old order. We will
thus find preferential sites for the synthesis of certain compounds,
namely, sites where the necessary ingredients, templates, energy sources,
and physical factors for that particular synthesis coexist. There must
also be preferential sites for the settlement of the finished products ; for
as synthesis continues, the accumulating products are driven further and
further from their production centers and thus brought under conditions
not necessarily compatible with their continued existence. I have sug-
gested on a previous occasion (54) that such preferential settlement
might be a matter of the configuration of the molecule, as a stable layer
of molecules with specifically shaped ends firmly adsorbed to an interface
could readily trap roaming molecules of complementarily fitting con-
figuration.'

3 This hypothesis leans on the immunochemical concepts of Pauling (32) .

DIFFERENTIAL GROWTH I49

The concept of molecular ecology gives a more truly representative
picture of cell life than the older morphological concepts. It gives expres-
sion to the fact that what is "determined" in a given cell is only the
general statistical norm of the proportions, arrangement, and distribu-
tion of its constituent elements, but not an absolutely fixed and stereo-
typed pattern exactly repeated in each individual cell specimen. No two
cells are ever precisely congruous. It is in the very nature of the cell
that we can predict no more than the degree of probability of the occur-
rence of a particular process in a given place, and this only with reference
to what is happening in the rest of the cell.

In animal ecology the sizes of the mobile units vary greatly. Some
animals move independently, others in flocks, and still others as hetero-
geneous groups. Similarly, in the molecular populations of the cell some
molecules move singly, others combine into larger bodies, submicro-
scopic or microscopic particulates, and still others form mixed ag-
gregates, with members of different species compounded in definite
proportions (e.g. lipoprotein complexes, coacervates, etc.). The forma-
tion of any such larger unit introduces a new element of complexity
(new specific surface, new conditions for adsorption, restraints of
mobility, etc.), and if one thinks the matter through to its logical con-
clusion, one realizes that cytodifferentiation effects its manifest order
through a series of steps of progressive complication, in which the basic
determinants are the properties and mutual affinities of the different
molecular species, and a frame of "conditions" prevailing in their com-
mon space, favoring the various species differentially. This frame or
ground plan, in relation to which the mobile elements will become or-
dered and sorted, must evidently possess a greater measure of physical
stability than the rest of the system. In a relatively fluid system, such as
a cell, the entities to answer this demand are the surfaces, in which
molecular mobility is restrained. Surfaces thus assume foremost rank as
ecological niches for the assembly and segregation of different segments
of the molecular population of the cell. Their mode of operation can be
briefly sketched as follows.

Surface Organization

Let us consider the interfaces which set off a cell from its surround-
ings, or a cytoplasmic inclusion from the matrix, or a chromosome from
the nuclear medium. Heterogeneous systems meeting along a common
border immediately affect each other across the border. This is due to

150 PAUL WEISS

the fact that the molecular population of one system begins to react
differentially to the conditions introduced by the new contact system
(electric charges, surface tensions, combinations and reactions between
border molecules). The resulting interactions lead to (i) the segrega-
tion of a selected fraction of the molecular population into a surface
layer, and (2) the immobilization, hence, solidification, of this portion.
The mechanisms involved will vary, but for our present purpose they
can simply be lumped under a common term as ''selective adsorption."
Their common denominator is the fact that a given interface of given
constitution will intercept and retain certain member species of a mixed
molecular population more easily and more strongly than others. Conse-
quently, the favored species will become concentrated at the surface and
the less favored ones correspondingly diluted. A change in the outer
conditions may force the replacement of the former border population
by new and more appropriate species from the interior. A complex sys-
tem of this kind will have a labile phase, during which the molecular
surface grouping responds to changes in the physical state and chemical
composition of the adjacent outer system, and a secondary stable phase,
which follows the consolidation of the surface due to chemical bonding,
gelation, precipitation, etc. If the outer conditions are not uniform over
the entire surface, but vary locally, the adsorbed molecular populations
will be correspondingly variegated. The same cell can thus develop dif-
ferent types of surface zones on different sides exposed to different
environments. Moreover, each of the manifold types of interfaces in-
side the cell will also acquire its own appropriate type of covering. The
segregative power of specific surfaces can thus be readily understood.
In becoming adsorbed to an interface, polar molecules assume a com-
mon orientation. Their exposed ends thus constitute a new free surface
to which further molecular species of fitting configuration may be built
on in a second layer. This, in turn, may serve for the deposition of a
third group, and so layer upon layer may be stacked up in a specific
sequence, the molecules of each interlocking with those of the next.*
Although factual information is scarce, it might be surmised that the
apposition of new layers in living systems is governed by highly specific
relations among the combining molecules, perhaps due to the steric
fitting, key-lock fashion, of the specifically shaped ends of the conjugat-

* Such stacking up of molecular films has actually been demonstrated by Langmnir
and Blodgett (23), and although their observations were made on rather simple models,
evidence of a similar molecular lamination at the surface has been found in at least
certain types of cells (10, 34).

2

/

., ^**

\

«

■*S'4^.

••

Z/^^*"

w

%■

\^

I'LAii: 1. Transformations of Schwann tells from
peripheral nerves of adult rats in tissue culture (niod-
ilietl from Weiss and Wang, 1945). All figures at iden-
tical magnification (X 710). Panel 1 shows the original
slender spindle shape of Schwann cells freshly emi-
grating from nerve. The picture also contains two
large endoneurial fibroblasts (large, oval, lightly-
stained nuclei ; dark, sutian-stained fat droplets in
cytoplasm). Panel 2 shows beginning ramification of
Schwann cells in plasma clot (note rod-shaped nuclei
and fat droplets). Panel S .shows giant multinucleate
"macrov)l>age" developed througli transformation and

•*v'*''

V

>» J*'

ir

N

^'.w'

i

6

coalescence of a group of Schwann cells at interphase
between cover glass and capillary exudate below
plasma clot. Panel 4 sliows extensively arborized
Schwann cells with polymorphic nuclei at fringe of
culture (note active extrusion of fat bodies). Panel 5
shows transformation stages from spindle type through
monopolar "flask" shape to spherical "monocytoid''
form with sickle-shaped eccentric nucleus and distinct
zone of hyaline exoplasm. Panel 6 shows small mac-
rophages formed from Schwann cells in liquid -ireas
of medium. (Note mitosis in telophase indicated by
arrow.)

f^^^

* **

Plate II. Induction of cartilage formation by dead cartilage graft in Amblystonia
tigrinum (X 185). A metatarsal bone, quick-frozen in isopentane at — i59°C, debydrated
in vacuo at — 40°C, reliydrated in Ringer's solution, and transplanted into the tarsus of
another premetamorphic specimen of the same species, has healed in; new cartilage and a
new perichondrium have formed along most of the surface of the epiphysis of the graft,
which has become detached from the bony shaft, seen in the right half of the picture. The
new cartilage can be clearly distinguished by its darkly stained nuclei from the dead graft
with unstained ghosts of dead nuclei in its capsules.

DIFFERENTIAL GROWTH I5I

ing molecules (54). As complex macromolecules will mostly have dif-
ferent configurations at opposite ends, successive layers will consist of
different molecular species. Here we have, therefore, a process that,
once it has started from a given surface, could lead to a progressive
segregation with stratification of molecular species which were originally
intermingled. In this process molecular orientation is of paramount
importance, since it exposes the maximum number of specific receptive
ends to the interior. This consideration gains added significance in the
case of those polar molecules that have enzymatic properties, as parallel
orientation and packing might increase their effectiveness by turning all
active groups toward the substrate. Then selective surface adsorption
would become a prime factor not only in the segregation of existing
molecular species, but also in the creation of new ones. Moreover, the
molecular fringe which settles along the surface acquires master control
over substance transfer across the cell boundary in that it can selectively
facilitate or prevent the exchange of materials between the areas it
divides.

The extent to which a surface layer grows inward by apposition, will
vary with the circumstances. For example, in some cells the outermost
layers form a thick and distinctive coat (myelin sheath of nerve fiber,
envelope of red blood cell, vitelline membrane of egg), while in others
they remain an inconspicuous "plasma membrane." The subjacent layers
(crust of egg, ectoplasms) likewise vary in thickness as well as in
sharpness of demarcation. It must be remembered, however, that strati-
fied molecular segregation need not be so massive as to become micro-
scopically discernible.

The picture of surface organization here presented is incomplete
and greatly oversimplified. For instance the specific key molecules, which
were given such prominence, obviously lie interspersed with many more,
smaller, simpler and trivial molecules of less specific behavior, a mosaic
feature which we have not duly taken into account. It has been suggested
on the basis of studies on drug action, immunological reactions, and
radiation effects, that target points of specific response are present in
isolated patches occupying only a small fraction of the cell surface. A
similar discontinuity characterizes the structure of the chromosomes.
Accordingly, many of the specific effects assigned above to surfaces in
general will actually be confined to certain portions of the total surface
only. An extension of this thought leads to the realization that complex
molecular aggregates (particulates) may be subject to specific adsorp-
tion and fixation, with one of their constituent key patches effecting the

152 PAUL WEISS

specific attachment, and the rest remaining uninvolved. But to discuss
these and other quaHfications in detail, would exceed the scope of this
paper.

Differentiation in Molecular Terms

We return from this excursion into the molecular realm better
equipped to phrase questions about differentiation and growth in tangible
form. We can now ask, for instance, whether dift'erentiation really
signifies a change in the composition of the molecular population of the
cell or merely a change in the distribution of preexisting molecular
species. The latter view has been implicit in those embryological and
pathological theories which hold that the various cells of a given organ-
ism are all composed of essentially the same basic and immutable "proto-
plasm," and that their diverse appearances and performances are merely
the overt responses to a variety of external situations or "stimuli." A
given response display (called "differentiation") is thought to last only
as long as the given stimulus situation lasts. Change the latter and the
display will change accordingly ("dedifferentiate" and "redifferentiate")
in a new direction.^ Yet, when we view the facts from the molecular
level, there seems to be no doubt that cytodifferentiation does not leave
the chemical composition of "protoplasm" basically unaltered.

The ubiquitous species of small and simple molecules going in and
out of cells remain about the same at all stages except for differences in
concentration. Some species of large and complex molecules, including
many of the proteins, are also present throughout development (5, 28).
All these comprise what we have called above the stock in common to
all cells. But in addition to these, differentiation brings with it, or virtu-
ally consists of, the continuous elaboration of molecular novelties which
have not been present from the beginning. Thus, a liver cell, a nerve cell,
a pigment cell, a mucus cell, a thyroid cell, etc., each produces its own
chemical specialties. Can this really be taken as evidence that the respec-
tive "protoplasms" are chemically different? Might it not be that all
mature cells have the synthetic faculty for the whole array of cell
products known to the body, and that a given cell is merely prevented
by local external conditions from materializing any but the appropriate
one ? This contention is contradicted by the experimental evidence quoted
above, showing that cells transferred into a different environment

5 A certain partiality toward such a view can be detected, for example, in the earlier
writings of Child, and there have appeared several more modern, if less thoughtful,
versions of the same theme.

DIFFERENTIAL GROWTH I 53

(explantation or transplantation), though they may cease to produce
their customary products, never reacquire the faculty to give rise to any
but the products of their original line, when given another chance.

Thus the conclusion is inescapable that the various cell types men-
tioned do not just temporarily exhibit different production processes,
but have themselves become chemically converted into different pro-
duction plants. That is, each cell type develops, in addition to the molecu-
lar species which it contains in common with all or several other cell
types, its own peculiar and distinctive array of molecular species. These
are the ones that account for the differential behavior of the various
protoplasms ; they are the true objects of differentiation. One is tempted
to identify them with specific proteins, but factual knowledge is still
far too inadequate for any generalization. We need much more precise
information on the time and mode of appearance of various cytochemical
constituents in the ontogeny of various cell strains. Such an inventory
is bound to contain the answers to numerous questions about which we
can as yet do no more than speculate.

In contrast to differentiation, cell modulation implies no essential
change in the composition of the molecular population, but merely a
regrouping of the existing species. According to our previous discussion,
any change in the environment of the cell that alters the conditions in
the surface could cause a reshuffling of the content and replacement of
the surface population with marked changes in the behavior and mor-
phology of the cell, but without necessarily affecting its basic composi-
tion. However, whether such a modulation is fully reversible, i.e.
whether or not the reshuffled population can return once more to its
initial distribution when the original environmental conditions are re-
stored, depends on the degree of consolidation the surface has undergone
in the interim.

These considerations suggest the possibility that all differentiations
start out as modulation, as mere molecular regrouping, in response to
changes in the cellular environment. Let us consider an embryonic cell
shifting from a site A to another site B, or, what amounts to the same,
a change in the environment of a stationary cell from a condition ^ to a
condition B. Exposure to the new contact conditions brings molecules
with affinities to B to the surface. Now, suppose this molecular coat
initiates a chain of chemical reactions, through which the composition
of the interior is materially changed. This implies the appearance of
novel molecular types as well as the disappearance of some of the exist-
ing species as they become converted into new ones. Such change cannot

154 PAUL WEISS

take place all at once throughout the cell. We must assume that it starts
from a localized area (in our concept, the surface) and spreads at a
measurable rate. Therefore if the cell were returned to its original
condition A during the early phases of this process, it would still con-
tain sufficient amounts of all the species it had before to resume its
original state. To the observer this would appear as reversible modula-
tion. But continued residence under condition B would permit con-
tinued progress of the changes indicated, until eventually the cell will
have lost a major segment of its earlier constituents. It would thus have
undergone an irreversible change of its constitution, i.e. true differentia-
tion. The lability of differentiation in its incipient stages is a matter of
record.

If a cell that has assumed character B then moves on to a site C (or if
its neighbor cells or the intercellular milieu changes to a condition C) ,
its surface population will again become reconditioned and then initiate
a new chain of processes, marking a further step of differentiation, the
earliest phase of which would again be labile. A cell transferred from
A to C directly may or may not respond in the same fashion as when
coming by way of B, depending on how decisively its molecular popula-
tion has been altered during the intermediary stay at B.

We are portraying "differentiation" as a chain of events consisting
of alternate physical regrouping and chemical alteration of the molecu-
lar populations, the latter phenomenon involving the emergence of novel
species of compounds. It can be seen that according to this concept the
potentialities for differentiation are strictly limited by the initial chemi-
cal endowment of the cell and that they become further restricted as the
cell passes through the various phases A, B, C. What we conventionally
call "loss of potency" is therefore merely the counterpart of the positive
increase in definite chemical specialization incurred by the "differentiat-
ing" segments of the molecular population. We need assume no separate
inhibitory agents.

This concept contains also, in principle, the answer to the problem of
divergent cytodifferentiation, that is, of how two initially identical cells
may become the source of two qualitatively different strains. Let us
assume that of two cells with identical molecular populations, including
key species a, ^, y, S, e, etc., one is exposed to an environment E, and the
other to a different environment F. Given some agitation and sufficient
mobility of the cell contents, we may further assume that in each cell a
selected fraction of key species will be concentrated at the surface,
namely, those species which best conform to the adjacent medium; let

DIFFERENTIAL GROWTH I55

US say, compounds e in the cell exposed to the E surface, and compounds
<t> in that at the F surface, where e and </> may be entirely unrelated.
According to the preceding section, such adsorbed and oriented border
species can exert key effects in turning the chemical reactions, segrega-
tions, and syntheses among the rest of the population into a definite
course. Consequently the chemical fates of the two cells, coated by
qualitatively different key species, will likewise become qualitatively
different. Thus a difference which started out as merely one of distribu-
tion gradually develops into one of composition and character. This
would be the fundamental step in the dichotomy of cell fates. This con-
cept remains, for the present, hypothetical. But being more concrete and
specific than the conventional verbal concepts, it has the advantage over
the latter of being amenable to experimental verification. Tentatively, it
fits the known facts of development very satisfactorily, as will be shown
in the following.

On an earlier occasion (47) I outlined three basic principles of
cytodifferentiation, which I called (i) discreteness, (2) exclusivity, and
(3) genetic limitation. Any valid concept of differentiation must be
consistent with these principles.

The principle of discreteness refers to the fact that "differentiation
produces a definite number of discrete, distinct, discontinuous and more
or less sharply delimited cell types which are not connected by inter-
gradations." The principle of exclusivity states that once one type of
differentiation has taken hold of a cell, no fraction of such a cell can
ever become engaged concurrently in any other type of differentiation ;
the cell in differentiation acts as an entity in all-or-none fashion. For
instance, there is a sharp dividing line at the rim of the optic cup sepa-
rating optic retina from tapetum ; one cell is a typical retinal cell and the
next cell is just as typical a pigmented tapetum cell. Although we know
from experimental evidence that originally both have the endowment to
transform into either type, they never form a half-way blend. Examples
of this kind could be multiplied at will. It is evident that these principles,
derived from empirical facts, can be logically deduced from the postu-
lated discreteness of the molecular key species «, P, y, 8, e, and their
mutual exclusiveness in the competition for surface positions. Thus, a
given cell must follow either the a-determined or the /3-determined or
the y-determined, etc., course, and follow it wholly and with no compro-
mises.

The third principle, genetic limitation, stating that each cytodif-
ferentiation occurs according to the special methods characteristic of

156 PAUL WEISS

the parent species, reflects the fact that the numbers and kinds of molecu-
lar key species available for differentiation are limited by the inherited,
hence, genetically controlled, initial endowment. It would seem prema-
ture, however, to speculate about the relation between our postulated
key species and genes.

Phenomena of dominance and recessivity of cellular activities can
likewise be explained in terms of our hypothesis. Any molecular species
assuming a monopolistic surface position thereby gains "dominance"
over competitively inferior species which, being barred from the sur-
face, are in no position to express themselves effectively.

Induction

After this consideration of cellular responses in differentiation, some
comment should be given to the factors calling them forth in the organ-
ized play of development. In view of the fact that the problems of
"organization" are discussed by Dr. Nicholas in the next chapter, we
shall forego any but the most cursory treatment of the subject here.

According to the testimony of experimental embryology, develop-
ment proceeds as follows. Cleavage subdivides the egg into blasto-
meres, each of which receives in addition to a nucleus a certain fraction
of the original cytoplasmic system. While the nuclei are essentially
equivalent, the cytoplasmic allotments vary, depending on the degree of
specific molecular segregation attained in the undivided egg. The role
of the most nearly stabilized part of the egg, namely its surface, must
be especially taken into account.^ Different blastomeres inherit different
parts of the original egg surface. This initial differential among egg
parts is then further elaborated and amplified by the subsequent reactive
transformations which the various cleaved-off cells undergo in response
to the organized conditions of the system which they constitute.

Descriptively, these conditions have been referred to as "fields"; but
we shall not concern ourselves here with their character and operation.
We need only keep in mind that in the early phases of development any
area of the cleaved egg is endowed with dual capacities. On the one hand,
it is composed of a large number of individual cells which in many cases,
according to experimental evidence, are practically equivalent ("equipo-
tential") and hence interchangeable as far as their response faculties
are concerned. On the other hand, the totality of these cells, as a collec-

^ The organizational role of the surface of the egg has been emphasized by several
embryologists (9, 21, 33).

DIFFERENTIAL GROWTH I57

tive unit, exercises specific functions not exhibited by the individual
elements as such. To mention a familiar though crude example, the
differential exposure to environmental factors of cells in the core of a
given group, as contrasted with those near the surface, is a property of
the collective not owned by its members.

Cell groups under the influence of such collective factors, but prior to
the appearance of a manifest response on their part, are usually referred
to as "determined." There follow material transformations of the kind
discussed in the preceding chapter, introducing novel relations among
parts. Stepwise, in assembly-line fashion, the diversity of activities in-
creases. As initial time differences become accentuated, some parts of
the germ soon outstrip others in their tempo of differentiation. From
then on, influences of the more advanced ("determined") upon the less
advanced ("undetermined") parts become increasingly prominent.
These influences find essentially two forms of expression : ( i ) influences
spreading within a primary cell continuum, and (2) influences exerted
from one cell continuum upon another contiguous one, usually after
secondary junction.

Influences of this type are customarily referred to as "inductions."
The term is purely descriptive and covers a considerable variety of het-
erogeneous phenomena which have often no more in common than the
label. There is not the least justification for the supposition that a single
operative mechanism underlies them all, much less so a single chemical
entity. Much more pertinent than the illusory search for a master solu-
tion to the "induction" problem is an objective analysis of each and
every one of the various concrete instances of induction that have thus
far been recognized. Therefore, instead of discussing induction in
general, let us concentrate on two specific examples, one for each of the
categories mentioned above.

The first example concerns what might be called "homoeo-induction,"
that is, the progressive recruiting of cells for a given type of differentia-
tion, spreading from a focal area like an infectious wave (e.g. the an-
tero-posterior progression of myotome formation). One gets the im-
pression that cells which have attained a certain differentiated character
can communicate their state to their neighbors, which then pass it on
further, and so on down the line. The experimental test lies in trans-
planting a differentiated fragment amidst less far advanced cells (37).
In many embryonic operations of this type, completion of the fragmen-
tary graft structures by recruitment from surrounding cell sources has
been observed, but in no case has the precise mechanism been revealed.

158 PAUL WEISS

Technically it might be easier to start with post-embryonic stages, par-
ticularly regeneration. For instance, the fact that transplants of mature
cartilage may induce new cartilage formation lends itself to a much more
detailed analysis than has as yet been undertaken. Nageotte {2'j^ re-
ported that the implantation of alcohol-fixed cartilage into the rabbit's
ear stimulates the formation of new cartilage in the adjacent connective
tissue. In preliminary experiments (50) in which I transplanted pieces
of salamander skeleton which had been devitalized by quick-freezing
and drying, I found that cells of the larval host built a new shell of
cartilage onto the dead model (Plate II) . Since new cartilage has formed
only in direct continuity with the graft, we must conclude that the influ-
ence in question is transmitted only by contact and not by diffusion. It
seems that objects of this kind would be very favorable for a detailed
physical and chemical analysis of communicable differentiation.

In terms of a molecular population concept, this type of inductive
effect could be described as follows. The exposed surface of the cartilage
contains certain cartilage-specific key compounds. Any potentially chon-
drogenic cell would contain similar key species in its interior. On making
contact with cartilage, these key species would become concentrated and
oriented along the surface by virtue of their affinity to the cartilage.
After saturating the surface, they would then direct the subsequent
chemical transformations in the cell interior into a cartilage-specific
course. Thereupon the cell can, in its turn, act on the next one, and so
forth. In the case of cartilage, the selective effect would be exerted by
the secreted ground substance rather than the cell body itself, but in other
instances it would pass directly from cell to cell. This concept has been
developed in greater detail in an earlier publication {$1). Its main thesis
is that adjacent cells influence each other's development by means of
selective attraction among specifically configurated molecules, key-lock
fashion. An alternative assumption would be an actual transfer of
specific master units from cell to cell, in the manner of a spreading
infection. Present evidence is insufficient for a final decision.

The second category to be illustrated here is "heteroinduction," in
which one tissue induces another tissue, with which it has secondarily
come in contact, to differentiate in a direction different from that of
the inducing part. Inductions of neural plate by chorda-mesoderm and
of lens by retina are familiar examples. According to our hypothesis the
mechanism might be as follows. The cells of the inductive layer have a
certain specific surface organization, that is, they possess already a
surface layer of specific key molecules in relatively stable configuration

DIFFERENTIAL GROWTH I 59

and orientation. Corresponding molecules of complementary configura-
tion would be present in the ectoderm cells ; but during the earlier labile
phase of these cells, these would still lie scattered in the interior, mixed
with an assortment of other species. As the organized substratum then
doubles under the ectoderm, its surface molecules would attract their
molecular counterparts in the ectoderm cells into the contact surface.
From their oriented surface positions, these molecular layers in the
basal face of the ectoderm cells would then control the further chemical
developments in their cell bodies, as outlined above. It can readily be
seen that the type of response to be obtained from such an inductive
action depends both on the character of the substratum and the types of
molecular species present in the reacting cells at the time of exposure.
The progressive transformation of the molecular equipment of the cell
explains why there is often a limited period of "competence" (43) dur-
ing which cells can appropriately conform to a given inductive stimulus.

In contrast to previous hypotheses, which for the most part have
ascribed the inductive effect to diffusible "evocator" substances, the
theory here advanced presumes a definite spatial organization of the
molecular population as essential for inductive success. Induction of
this type could be transmitted only by direct contact. This is in agree-
ment with experience. Lens induction requires intimate adhesion between
retina and epidermis. Neural plate arises in direct contact with the
chorda plate. Many more examples could be cited that would seem
to discount the diffusible nature of the inductive agent. One might
object that neural inductions have been obtained in explants floating in
body fluids, but it is conceivable that in these cases the active role is
played by a protein film adsorbed to existing surfaces to which the
explant has temporarily adhered, or a similar film settling along the
outer surface of the cells. Since there is a direct relation between the
presence of disintegrating cells and the occurrence of neural inductions
(18), it is plausible to assume that some neural key species, set free
from the destroyed cells, collect and become oriented along the outside
of the surviving cells and thereby become an inductive film in the sense
discussed under homoeoinduction.

These comments are conjectural. There are not enough facts known
on which to decide the issue of contact versus distance action conclu-
sively. It stands to reason that a precise and objective description of the
cytological changes attending induction, together with experimental tests
of their relevance for the process, could provide some of the wanting
clues. A suggestive instance is the following. In observing tissues in

l6o PAUL WEISS

inductive interaction I have often been struck by the remarkable corre-
spondence of their cell and nuclear orientation just during the crucial
phase. This phenomenon is very marked, for instance, in the formation
of the lens. When the eye cup makes contact with the overlying epi-
dermis, the irregular cells of the latter, within an area exactly coexten-
sive with the area of contact, turn into a prevalent radial orientation,
each elongate nucleus lining up in the direction of the axis of the
opposite retinal cell. Precisely these oriented cells will later constitute
the lens placode. The orientation effect is transitory, being obliterated
by the subsequent transformations of the placode. While experimental
tests remain to be applied, the cell-sharp coincidence between area of
contact, orientation, and lens determination makes it almost certain that
the observed cell changes are directly related to the primary inductive
effect. Evidently, what happens is that the content of the epidermal
cells undergoes a thorough reshuffling, with the underlying retinal cells
acting as attractive foci. In short, the elongation and orientation of cell
structure would be visible expressions of the oriented shifts of the
molecular population, which would have to be expected from our theory.
A systematic investigation of this orientation phenomenon is under way.

The foregoing comments on induction were presented mainly as
illustrations of how problems of development gain in sharpness when
resolved into terms of molecular behavior. The special solutions sug-
gested should be considered merely as a first crude and wholly tentative
effort to replace the vague, noncommittal, and unrealistic notion of
"inductor substances" by a specific, concrete, and testable concept. The
limitations of this concept are self-evident. Therefore our insistence lies
more on the merits of the particular manner of viewing the phenomena
than on the tentative views presented.

As for the limitations, one must not lose sight of the fact that the
concept pertains only to the instrumentalities of biological processes
and not to the overall order which these processes follow and which we
usually refer to as organization. It can explain the manifest differentia-
tion of a given molecular population, provided only that there is a regu-
lar frame of organized surface conditions to start with. The organiza-
tion of the cell content is thus referred back to a prior topographical
organization of the various surfaces of cell, nucleus, chromosomes, etc.,
and the principle, omnis organisatio ex organisatione (48), still holds.
The "field" character (47) of the organizing conditions in develop-
ment, which cannot be reduced to molecular terms, has not been touched
at all in our discussion.

DIFFERENTIAL GROWTH l6l

Another point is that our concept is perhaps too exclusive in assign-
ing the major ordering role to the surfaces. Some faculty of self-sorting
of molecular mixtures (as in the formation of tactoids [3] ) will perhaps
have to be brought into the picture to make it complete. Also our assump-
tion that selective transmission effects and affinities are essentially
explicable in terms of steric conformance of specifically shaped mole-
cules remains to be verified. It is still possible that ultimately we may
have to take recourse to some other sort of "resonance" mechanism,
"resonance-like" interrelations among tissues being clearly recognizable
on the biological level (46),

Growth

The problem of growth likewise resolves itself now into a number
of tangible issues. The definition of growth gains precision. Growth is
often defined as any increase in the mass of an organic system. Since
such an increase may occur by the mere intussusception of water — and
even a dead seed can swell — this popular definition is scientifically of no
value. I propose to define growth in a more restricted sense, as the
increase in that part of the molecular population of an organic system
which is synthesized within that system. The other part of the popula-
tion, consisting of inorganic substances like water and salts, passing in
and out of the system, is left out of consideration, and so are the smaller
organic compounds that enter as ready-made building blocks, e.g. sugars
and amino acids, until they have become assimilated and thereby lost
their identity. Since, as far as we know, the synthesis of the complex
high molecular systems in a cell occurs only in the presence and with the
intervention of other similarly complex and cell-generated systems, after
their own image, growth in our definition is essentially synonymous
with reproduction. In population terminology, we might say that im-
migration of molecules into the cell space as such does not constitute
growth, but their assimilation into complex systems after the model of
preexisting indigenous systems does.

We can thus divide the cell content roughly into two categories :
systems capable of turning out more of their own kind, and others
devoid of that faculty — reproductive and non-reproductive ones. The
non-reproductive group includes all the elementary ingredients for the
higher compounds, as well as the sterile terminal cell products — mem-
branes, fibers, secretions, pigments, etc. In thinking about these matters
it is rather distressing to note that the proteins about which we have the
most direct information, such as collagen, keratin, melanin, myosin,

l62 PAUL WEISS

pepsin, etc., belong in the latter class and hence can divulge nothing about
the reproductive systems, which are still very much of a mystery. At any
rate the distinction here proposed enucleates the core of the growth
problem, as it were. We shall say of a pigment cell that it grows not
when it synthesizes more melanin granules, but only when it synthesizes
more of that peculiar part of its protoplasm which, other conditions
favorable, can synthesize pigments. Similarly, growth of a myoblast is
not the synthesis of more myosin and actin, but of more of the basic
muscle protoplasm possessing, among others, the faculty to synthesize
myosin and actin ; and so forth. For the sake of simplicity let us ignore
for the moment the fact that "more" refers not to absolute production,
but to the net excess of production over destruction in opposite metabolic
phases. It would also be premature to regard the "basic" systems in
question as nothing but enzyme mixtures, although enzymatic activity
is one of their prime characteristics.

The fundamental feature of growth, therefore, is the capacity of a
reproductive system to procreate more systems with similar reproduc-
tive faculty. In comparison with this fundamental phenomenon of
multiplication of the molecular key population, the often superimposed
phenomenon of cell division is as incidental as the setting up of new
administrative districts in a growing country. The relation between
cell division and growth is by no means as close as the widespread habit
of treating them interchangeably would make it appear. During cleav-
age, an egg undergoes repeated cell divisions with practically no growth
at all, while conversely, certain cells, such as mature neurons or the
single-celled kidneys of some nematodes, grow to comparatively enor-
mous sizes without division of either cytoplasm or nucleus.

The growth process is being explored from many directions and with
increased momentum. Pertinent information comes now from genetics,
now from virus research, now from embryology. Views begin to con-
verge, and concepts, to become unified. Yet the more diversified the
contributions, the more we must guard against the obvious danger of
confusing a common label with a common object. If growth means cell
division to one, increase in weight to another, elongation to a third,
metabolism to a fourth, protein synthesis to a fifth, their mutual com-
munication on "growth" is apt to become deceptive unless they are fully
explicit. I want to make it clear, therefore, that I shall use the term
strictly as defined above, that is, as the multiplication of that part of
the molecular population capable of further continued reproduction,
irrespective of whether or not accompanied by cell division.

DIFFERENTIAL GROWTH 163

With that definition in mind we can now ask a number of specific
questions pertaining to the mode, the rate, and the site of growth. First,
do cells ever stop growing? Superficially, cells fall into two categories,
those that keep on proliferating, and those that do not. The ratio of the
former declines steadily during ontogeny. Yet even in the mature body
proliferation is still very active in many tissues, such as skin, blood-
forming organs, many glands, sex cords, and others. Other cell groups
become stationary. Since incremental growth can be observed only if
there is a net surplus of production over breakdown, it is quite con-
ceivable that the stationary cells are likewise engaged in growth, but at
a rate just sufficient to balance the metabolic losses. In this case, proto-
plasm would be continually regrown in all cells.

What evidence is there for this view ? Our habit of lumping all syn-
thetic processes in one class and all degradation processes in another
tends to make incremental growth appear simply as a shift in the
balance toward the former ; in other words, simply as more of the same
kind of activity that is going on in any metabolizing cell. But if we
distinguish, as we must, between different levels and types of synthesis,
the problem loses its aspect of simplicity. To crystallize the issue let us
focus on the most complex molecular systems, the proteins. We know
from isotope tracer work (35) that proteins are in a state of continuous
renovation. Now, this may mean either that the individual molecule
merely interchanges constituent parts with its environment while pre-
serving its identity and individuality, or that the molecule breaks down
completely while another whole one, freshly delivered from the repro-
ductive system, appears in its stead. The difference is as between keeping
an automobile in repair by replacing the worn-out parts one by one,
or scrapping it and replacing it by a new one. If the latter interpre-
tation is correct, protein replacement and reproduction, i.e. true growth,
would definitely be two different processes. To the best of my knowledge,
the matter has not been finally settled.

Let us then confine ourselves to the cells which are known to grow
perpetually (basal skin, lymph, intestine, etc.). Each of these cell types
is known to produce each more of its own kind. What does this imply
at the molecular level ? Again we are faced with alternatives that we can
only present, without deciding between them. The question is whether
the basic systems of reproduction are the same for all cell strains or
whether they change in character with progressive differentiation.
Perhaps all cell strains retain a certain germinal core population which
remains exempted from differentiation and constitutes the sole source

164 PAUL WEISS

of new protoplasmic compounds; in that case the synthesis of each
specific compound in skin, blood, gut, etc. would have to recapitulate
the whole course of the ontogenetic differentiation of that cell strain.
The new units would be turned out in a primordial mold common to all
cells and then gradually pass through a series of transformations after
the pattern of the already present skin-, blood-, gut-specific units, as the
case may be. The other possibility is that differentiation gradually
changes the reproductive apparatus of the various cell strains itself so
that the skin-, blood-, gut-specific systems would each be enabled to
reproduce themselves directly. If we were to postulate that the genes
represent the only self-reproducing systems, and assume, on the basis
of genetic evidence, that genes do not change in character during on-
togeny (60), the former alternative would become axiomatic. Some
new experimental contributions bearing on this problem will be reported
later.

Growth Rate

The fact that we do not know just what fraction of the molecular
population of a cell constitutes the actively reproductive growth appa-
ratus makes it impossible to determine "true" growth rates, i.e. the rate
at which the reproductive fraction procreates more of its kind^ This
places serious limitations on our interpretation of relative growth, for
in all our measurements of growth the reproductive and non-reproduc-
tive fractions of the cell are lumped. As long as their ratio remains
constant the error in estimating true growth from measurements of
total growth is constant and can be discounted for most purposes. For
example, if identical cells, kept in different conditions (temperature,
nutrients, etc.), grow at different rates, the difference can be properly
ascribed to effects on the true growth process. However, if cells of
different character grow unequally, even under identical external con-
ditions, this does not necessarily reflect differences of their true growth
rates; it may merely be an expression of the fact that different frac-
tions of their total mass are in reproductive activity.

If we designate the reproductive fraction of the population as R and
the nonreproductive, hence, constant fraction as D, the true growth

dR

rate would be Vr = while the measured growth rate would be

Rdt

"^ True growth rate could theoretically be expressed by the time required for a unit
quantity of the reproductive system to increase to two units.

DIFFERENTIAL GROWTH 165

d{R-\-D) R R

""-^V^T^J^^l^+U ' """ ^'"^ ^T77< ^' '''' "^^^^^^^^

growth rate will be smaller than the true growth rate by an amount vary-
ing inversely with the unknown constant D. Consequently, of two cells
with different proportions of R and D but identical fundamental growth
rates of the portions R, the one with the larger D will appear to grow
more slowly. The fraction D is itself heterogeneous, including water, in-
organic compounds, and all organic compounds that have either not yet
been incorporated in the reproductive systems or have become detached
from them as metabolites and differentiation products. Accordingly the
ratio between D and R will be significantly different in different cell
types, and hence different actual growth rates must be expected simply
as a corollary of differentiation. Whether there is also a concurrent
change in basic growth rate remains undecided.

Moreover, the ratio R/D does not remain constant for a given cell
strain over its whole ontogenetic history. As differentiation proceeds, the
proportion of "differentiated" terminal products, making up a large frac-
tion of D, increases progressively at the expense of R. This increasing
drain on the reproductive fraction of the cell is reflected in the well-
known decline of measured overall growth rate with age. After our
earlier discussion of differentiation it seems hardly necessary to reiterate
that the Z)- fraction in question includes not merely the formed cell prod-
ucts (pigments, secretions, fibrils, etc.) but also that part of their forma-
tive apparatus that is not self-reproductive. Even after the loss of the
formed products, as in tissue culture, cells of different strains of differen-
tiation exhibit different growth rates (31).

Comparative evaluation of differential growth rates among different
cell types is further complicated by the fact that in some cells the formed
differentiation products stay within the cell and are therefore included
in all measurements, while in other cell types they are extruded and lost.
For instance myofibrils, neurofibrils, most pigment granules, mem-
branes, and the like stay with the protoplasms from which they arose,
while formed secretions, intercellular ground substances, and the like do
not. Comparisons based on visible increments are therefore quite de-
ceptive.

Not until most of these matters have received much more precise
formulation and investigation will it be possible to develop a rational
theory of differential growth. Attempts have been made to account for
differential growth rates by different partition coefffcients in the appro-
priation of nutrients, by competition for "growth substrate," by differ-

l66 PAUL WEISS

ences in the efficiency of utilization of foodstuffs, and by differential
growth stimulators and growth inhibitors.^ Plainly, many of these
theories miss the point in that they try to explain differences which only
appear to be differences of basic growth but in reality are natural results
of differentiation. Further compelling evidence on this point will be
presented below, where we shall deal with tissue growth.

The Site of Growth

The distinction between a growing {R-) and non-growing {D-) frac-
tion in protoplasm raises the question of the sites of growth. Evidently,
the R- and D-fractions might be either segregated or interspersed. They
might be mingled as mosaics of molecular, submicroscopic, or micro-
scopic (particulate) dimensions; they might be uniformly dispersed or
each concentrated in certain localities, which might or might not corre-
spond to the nucleus, the chromosomes, nucleolus, cell surface, or other
topographical features.

Some unexpected light has been shed on these problems from recent
observations on nerve growth, which I shall briefly relate. A nerve fiber
is the long cytoplasmic extension of a nucleated cell body, surrounded
by sheath cells. The fiber grows enormously in length and width during
ontogeny, and even more dramatically during regeneration. When a fiber
is cut, the distal stump, severed from its nucleated base, perishes, while
the proximal stump, still connected with its central cell, grows out anew.
The rapid elongation of the regenerating fiber is a process of proto-
plasmic movement rather than of growth in the proper sense (51 )• But
true growth also supervenes and becomes most conspicuous in the in-
crease in width of the fiber (from less than i micron to often more
than 10 micra, i.e. more than a hundred times in mass). Since the fiber
is surrounded by accessory sheath cells and well supplied with blood, one
could well imagine that its growth is sustained from local sources. The
experimental evidence, however, is that growth takes place exclusively
in the nucleated central part of the cell and that all the specific cytoplasm
for the whole axon is synthesized there. The evidence comes from con-
striction experiments as follows (55).

A local constriction (by a ring of artery) reduces the diameters of
the neurilemmal tubes which contain the nerve fiber. Fibers which have
regenerated through such "bottlenecks" remain permanently undersized
from that point on distally. In contrast, these same fibers swell to ex-
cessive dimensions just proximal to the constriction. Plate III shows

8 For some recent examples, consult Needham (29) and Spiegelmann (38).

3^

t#^

i'l^

^

>\

•v ^-

't»

Plate III. Unequal growth of constricted nerve fibers. Cross sections through a constricted
nerve (right) and neighboring control nerve (left) at levels indicated in center diagram, all at
identical magnifications (X 425). (Rat. 35 weeks after operation; P, proximal; D, distal). The
same nerve fibers which are greatly shrunken distal to the constriction (lower right) have become
blown up to giant dimensions just proximal to tlie "bottleneck" (top right) ; some fibers have
become as wide as a small artery (compare the artery in the center of the distal control nerve
at the lower left ) .

.. . - » 1*

> •Ail.

L

Plate I\'. Oriented cell growth between three explants cultivated in a thin film of
coagulated blood plasma. Silver impregnation makes the condensations of oriented fibrin
between the centers (compare Figure 3) recognizable as darker streaks in the medium;
the cells can be seen to follow these dark bands.

DIFFERENTIAL GROWTH 167

cross sections through an experimental and adjacent control nerve at
the indicated levels, located one centrally and the other distally to the
bottleneck. A detailed analysis of this phenomenon in a large series of
varied experiments has led consistently to the same conclusion, namely,
that new axoplasm is constantly produced in the nucleated part of the
cell body and that the whole column of axoplasm is in steady slow mo-
tion in proximo-distal direction. If this translatory movement is par-
tially throttled by constriction, the supply to more peripheral levels is
correspondingly reduced, while the surplus piles up at the entrance to
the narrows. If the constriction is later removed, the effect is much as
that of opening flood gates; that is, the dammed up material moves
downward, its front advancing, according to preliminary determina-
tions, at the order of one millimeter per day.

There is evidence that this proximo-distal growth of the axoplasm
occurs not only during regeneration, when it adds directly to the width
of the fiber, but actually goes on continuously in all axons, even the
mature ones, which have reached stationary size. The central cell body
seems to turn out a steady supply of new axoplasm, most of which is
then conveyed distad. This raises the puzzling question as to where the
material goes, if it no longer adds to the size of the fiber. I submit that
it simply serves to replace the protein systems of the cell as they wear
off in a steady natural degradation process, and that the nucleated cell
body is, in fact, the sole source of such replacement. We know that any
isolated fragment of nerve can exhibit the synthetic phases of respira-
tory metabolism with the aid of the enzyme systems it contains.
But in the light of our experiments, it would seem that the enzyme sys-
tems themselves cannot be synthesized anywhere except in the central
cell body, whence they are then distributed over the whole neuron.

As supporting evidence for this view we can quote the intensive pro-
duction of nucleoproteins at the nucleus of the nerve cell (20), the iden-
tification of nucleoproteins in the axis cylinder (2), and finally what
might be taken to be the external sign of their peripheral breakdown,
namely the liberation of nitrogen in the form of ammonia from periph-
eral nerve (15). As a matter of fact, a preliminary estimate of the rate
of protein breakdown from the known values of ammonia production
leads to a figure which is satisfactorily close to the rate of proximo-
distal replacement determined from the damming-and-release experi-
ments reported above. While details of this concept remain to be worked
out, it leaves no doubt that the site of growth and replacement of the

l68 PAUL WEISS

neuron is strictly confined to the nucleated portion. It is probable, though
not proven, that the same would then hold for other cell types.

These facts have an immediate bearing on efforts, currently in the
foreground of interest, to bring concepts of genetics and growth in line.
Our experiments place the site of growth near the nucleus, not neces-
sarily in the nucleus. We mentioned before that if the reproductive
sources should prove to be identical in all types of cells (as is assumed
for the chromosomal genes), we would have to assume that their prod-
ucts are instantly recast into specific patterns corresponding to the sur-
rounding differentiated cytoplasm. In our present case the nucleus may
continuously reproduce basic protoplasmic systems of the general
character dictated by the genie constitution, which, however, under the
molding influence of the existing specific neural population to which
they are added, would assume specific nerve-character. The alternative
would be that the intranuclear extrachromosomal systems themselves
have become specifically transformed in the course of differentiation
(see p. 144), and in that case the proliferated new compounds (nucleo-
proteins, etc.) could acquire their full nerve-specific character at their
intranuclear site of origin in statu nascendi. Our data are noncommittal
on this point. They do prove, however, that growth is not a ubiquitous
process in cytoplasm and should not be treated at par with the type of
synthesis commonly met with in other types of metabolism.

Tissue Growth

Analogous considerations, as applied in the foregoing to cell growth,
apply to the growth of tissues and organs. Just as the molecular popu-
lation of the cell consists of a reproductive and a nonreproductive frac-
tion, so the cellular population of a tissue is composed of proliferative
and nonproliferating components. In most organs proliferation becomes
gradually confined to certain restricted areas, so-called germinal zones,
arranged in layers (e.g. central nervous system, skin), cords (e.g. sex
cords) or foci (e.g. buds). Of the newly formed cells, some remain in
the germinal zone, while the rest move away from it. The former con-
tinue to proliferate, while the latter cease to multiply and usually un-
dergo some terminal specialization.

This dichotomy has sometimes been interpreted as signifying an
antagonistic relation between proliferation and differentiation.^ Statis-
tically speaking, it is correct that as differentiation progresses, prolifera-
tion declines. But this does not imply that differentiation in the strict

9 For a brief summary, see Weiss (47). P- 85.

DIFFERENTIAL GROWTH 169

sense and proliferation are mutually exclusive (ii). From our earlier
discussion, it should be plain that the proliferating sources of epidermis,
bone, blood, muscle, glands, etc. are differentiated to the extent that
each can reproduce only its own kind and none of the others. Yet they
are still actively multiplying. Only the full transformation into terminal
stages seems to be held up in those cells that are kept in mitotic agita-
tion. Proliferation does not interfere with differentiation, but it does
impede the elaboration of certain manifest products of differentiation.
Conversely, the general impairment of proliferative capacity in cells
undergoing terminal specialization can be ascribed to the fact that most
of their substance, including mitotic prerequisites, is diverted into the
building of differentiation products (e.g. intracellular fiber systems).

In early developmental stages practically all cells proliferate. With
progressive differentiation, however, the outlined segregation into re-
productive and sterile groups takes place. Since the latter do not further
contribute to the increase in mass, the total growth of a tissue or organ
will vary directly with the ratio of reproductive to nonreproductive
cells. This ratio in turn depends on the geometric configuration of the
proliferating zone and the rate at which newly proliferated cells desert
it. Further variability is introduced by the fact that most organs receive
secondary additions from foreign sources (e.g. blood vessels, nerves,
pigment cells), and, conversely, lose some of their own cells by disper-
sion or destruction. Since each organ has its own rule, it is evident that
bulk measurements, which fail to separate the actual sources of the
measured materials from the "dead weight," can furnish no valid basis
for comparing "differential" growth of parts differing in constitution. If
percentage increase is faster in one organ than in another, this could be
due to either (a) a larger initial proportion of reproductive cells, ov (b)
a larger quota of retention of new cells in the germinal zone, or
(c) greater infiltration of foreign elements, or (d) less dissipation of
cells and cell products, or finally (£') an intrinsically higher rate of re-
production in the respective germinal cells. Only differences not ac-
counted for by (a), (b), (c) or (d) are to be ascribed to (e), that is, to
a higher "growth rate."

For an illustration let us choose the growing vertebrate eye. Due to
its simple form and accessibility it has been a favorite object for com-
parative studies on growth, with the increase in dimensions serving as
index. How heterogeneous a collection of events this index covers can
be readily seen from the following account. The original eye vesicle
consists of a certain initial allotment of cells from the embryonic brain

170 PAUL WEISS

wall. At first, all of these cells divide. The growth function at this stage
is therefore a volume function. In the cup stage the retina becomes
multilayered, with a sharp division into a germinal and a sterile zone.
Only the cell layer in contact with the outer surface, corresponding to
the ventricular (ependymal) layer of the brain, continues to proliferate,
while the cells released into deeper layers differentiate the various ret-
inal strata without further multiplication. The source of growth thus
has become reduced to a two-dimensional one, causing a marked decline
in the relative growth rate taken over the whole organ (e.g. from meas-
urements of diameter). Later, the cells of the germinal layer themselves
cease to proliferate and transform into sensory cells, a process which
starts from the center (macula) and spreads rapidly toward the pe-
riphery (ciliary zone) of the retina. Eventually, only the cells at the rim
retain residual capacity to multiply. Further growth is then essentially
by apposition from this rim ; that is, the growth source has shrunk
from planar to linear extension. Meanwhile some of the neuroblasts,
though no longer multiplying, grow in size as they sprout nerve proc-
esses, which, grouped into plexiform layers, add to the thickness of the
retina. During the later stages a gelatinous secretion, supposed to come
from cells of both retina and lens, fills the interior with vitreous humor,
thereby progressively distending the eyeball. In addition, blood vessels
and other mesenchym penetrate into the eye from the surroundings.

This diversity and complexity of the component processes contribut-
ing to eye size makes the search for a single "growth-controlling" prin-
ciple appear utterly unrealistic, and comparisons between different eyes
on the sole basis of size must be fallacious. To exemplify briefly the
direction that an analytical approach to "differential growth" would
have to take, I shall select a few of the concrete questions raised by this
picture of eye growth and point out their general significance.

First, what causes the confinement of proliferation to the single
cell layer at the convexity of the retina ? The answer is unknown. But a
clue lies in the fact that in this and most similar instances the site of pro-
liferation corresponds to some prominent geometrical feature of the
system, such as a surface or fold or tip or crest. For the cells concerned
this means unique physical conditions (e.g. tensions, pressures, unique
exposure to surrounding agencies) not shared by the rest of the tissue.
In neural tube and retina, for instance, the direct exposure to the fluid
of the central canal and ventricular spaces might be the crucial factor.
The fact that injury can activate cells of the interior to take part in pro-

DIFFERENTIAL GROWTH I7I

liferation^" could be explained by the penetration of the central fluid to
the deeper layers as a result of the break in the internal limiting mem-
brane. This is purely conjectural, but capable of an experimental test.

Next, what determines the further fate of the cells proliferated in
the germinal layer? Of the two daughter cells of a germinal cell, either
(a) both stay in the germinal layer, or (b) both move away, or (c) one
stays and the other moves off. The same alternatives recur at subsequent
divisions. Continuation of procedure (a) would lead to exponential
increase of the proliferative source with no cells left for terminal differ-
entiation, while procedure {b) would rapidly deplete the proliferative
source. The actual growth of the eye at this phase is therefore deter-
mined by the relative incidence of events (a), (b), and (c). But again,
factual information on this point is practically wholly lacking. In anuran
amphibians the number of dividing cells in the retina has been reported
to remain remarkably constant (i6), which would mean that all divi-
sions follow scheme (c). But this is definitely not true of the chick, for
instance, in which the number of proliferating cells increases with the
expansion of the retina. ^^ Here the rate of growth is therefore largely
dependent on the statistical apportionment of the newly formed cells
according to ((7), (b), or (c). This in turn seems to resolve itself into
a matter of the relative orientation assumed by the daughter cells,' either
during mitosis or immediately after, tangential orientation leading to
(a), radial orientation to (b) or (c). A plausible hypothesis is the fol-
lowing.

It has been demonstrated that dividing cells separate in the direction
of maximum cell elongation, which corresponds to the direction of
maximum tension. Hertwig's rules of cleavage are an illustration of
this principle. Assuming that retinal cells follow the same principle, we
may conclude that distribution of type (a) will result whenever the lo-
cally prevailing tension is tangential, while types (b) and (c) will re-
sult from radial tension or lateral pressure. According to this concept
the proliferating layer can expand only as long as the retina is under
tangential tension. The source of such force is found in the distension
of the eyeball by the growing lens and the vitreous body. Retina, lens,
and vitreous body would thereby form a self-regulating system operat-
ing as follows.

Accumulation of vitreous fluid would stretch the retina, producing a

1° See for instance the compensatory proliferation of cells throughout the remaining
half of an embryonic hindbrain after removal of one half (12).

11 Recently confirmed by unpublished investigations in this laboratory by Mrs.
Eleanor Gould.

172 PAUL WEISS

prevalence of tangential tensions, which, in turn, would cause a large
proportion of newly proliferated cells to stay in the germinal surface
according to (a). As these additional cells are intercalated, the lateral
stretch is reduced and may eventually even be reversed into pressure if
the cells begin to crowd laterally; in that case, radial movement (b) or
(c) would result. Meanwhile, however, further secretion of vitreous
humor has produced further distension so that more new cells can be
accommodated in the surface expanse, and the cycle repeats itself. Be-
cause of widely varying local conditions, these events will not cause a
regular alternation of phases of tangential and radial divisions as in
cleavage, but will merely statistically increase the probability of the oc-
currence of one or the other. There are some experimental results sug-
gesting such a mechanism of growth regulation. When the vitreous
humor of an embryonic chick eye is drained, the collapsed retina re-
mains greatly reduced in size, but increases in thickness (56). The loss
of a progressively distending core would properly account for the re-
sult. It has also been demonstrated that when incongruous eye-lens com-
binations are produced by heteroplastic transplantation, a mutual ad-
justment of size occurs (i, 17). This could be ascribed to the fact that
an undersized lens, both because of its smallness and because it will per-
mit vitreous body to flow out, will not be able to exercise the distensive
function which promotes retinal growth until it has caught up with the
size of the eye and sealed the borders. Reciprocally, an oversized lens
will produce greater distension, hence, expansion of the germinal layer,
resulting in intensified retinal growth. In a similar capacity, distension
of vesicles, glands, ducts, etc. by their own or foreign discharges may
be a major factor in regulating the extent of their proliferating layers,
and consequently of the amount of growth. Perhaps the excessive size
of the gut in amphibians fed on a bulky vegetable diet as compared with
meat-fed ones (22) is due to the same principle. One could even en-
visage that the air in the larval lungs of gill-breathing aquatic amphib-
ians is there for a growth-supporting rather than respiratory purpose.

These examples may suffice to illustrate the dependence of overall
"growth rate" on the special mode of growth of each organ, particularly
the geometric configuration of its germinal zone.

Orientation of Growth

Growth as such, i.e. the mere increase of the specific molecular popu-
lation of the cell, is a scalar process. It has no intrinsic direction. Its di-
rectiveness is given to it by vector components of the physical frame-

DIFFERENTIAL GROWTH IJT,

work in which it occurs. Yet growth itself plays a decisive role in creat-
ing those physical conditions which, in turn, will orient its further
course. A few examples may illustrate this principle of reverberation.

In the preceding section reference was made to the effect of tension
on the form of growth in epithelial tissues. Even more spectacular is its
effect on the architecture of tissues of mesenchymal origin. Connective
tissue, tendon, fascia, cartilage, bone, blood vessels, muscles, all respond
to mechanical tensions by orientation of the component cells and inter-
cellular fiber systems along the lines of stress. Experimental analysis
of this response in tissue culture showed that the observed cell orienta-
tion is a secondary product following a primary orientation of the col-
loidal matrix in which the cells lie embedded (44). The sequence is as
follows : ( I ) Tension produces alignment of aggregates of the aniso-
diametric molecules of the matrix into chains. (2) Such chains join to
build up submicroscopic fibrils, which eventually may merge into mi-
croscopic fibers. (3) Cytoplasmic filaments (filopodia) from cells in-
habiting the matrix extend along the interfaces between these solid
micellar threads and the interstitial liquid. (4) The rest of the cell body
follows the advancing filopodia. In this manner the cell becomes greatly
elongated and is made to trace the ultrastructural pattern of its sur-
roundings. Nerve processes advance in the same fashion, except that
their cell body remains anchored (45).

In the original tissue culture experiments the matrix consisted of
blood plasma, with fibrin constituting the orientable ultrastructural
reticulum. Later experiments (53) revealed that even in so-called cul-
tures in liquid medium, cell orientation is brought about by a similar
reticulum, namely, coagulable fibrous material exuding from the cells
themselves, spreading along the surfaces of the liquid and constituting
a trellis ("ground mat") on which the cells then advance. Tensional
fibrin organization in blood clots has likewise been disclosed as the
mechanism orienting cell growth in wound healing and nerve regenera-
tion in the body (57). In embryonic tissues, other fiber proteins can be
expected to operate in like manner. All pertinent observations and ex-
periments made thus far confirm the conclusion that tension, by way of
its organizing effect on the colloidal cellular milieu, can determine {a)
orientation of cell movement, (h) cell elongation, (r) cell shape, {d)
orientation of mitotic spindles, hence, direction of division, (r) rate of
movement, (/) rate of multiplication, {g) deposition of fibrous cell
products. Figure i illustrates diagrammatically how different degrees
of stretch applied to a fibrous colloid translate themselves into the mor-

174 PAUL WEISS

phology of inhabitant, as well as immigrating, cells. Potent as these
effects are, they could not be a major factor in development if only
tensions of extraneous origin were involved. But tensional stresses also
arise within the developing organism as a result of shifts, growth, and
chemical activity of its parts.

Figure i. Diagram illustrating the effect of graded stretching (in the direction of
the arrows) of a reticular matrix on the shape of mesenchymal cells.

Let us consider the two main intrinsic sources of morphogenetic ten-
sions— expansion and contraction. Their development is illustrated in
Figure 2. On the left side (Ao, Bo) are sectors of a tissue possessing a
continuous reticulated matrix and containing a circumscribed area
(dark circle) in which some localized developmental process goes on, as
follows. The lower half (Bo-^Be) represents a case in which the marked
area expands relative to its surroundings, either by rapid proliferation
or by deposition of large amounts of ground substance or, in the case of
hollow organs, by distension. This local expansion creates in the sur-
rounding continuum tensions that are oriented circumferentially, de-
clining in strength with increasing distance from the border. The fibrous
matrix assumes a corresponding pattern as depicted, and this in turn
imparts itself to the cells. Cell movements and cell divisions will follow
tangential courses. It is evident that the familiar concentric architecture
of the mesenchymal sheaths, coats, tunics, and membranes around both
solid organs and hollow ducts and vesicles finds a natural explanation on
this basis. Essentially the same effect is observed if the surrounding tis-
sue contracts against a rigid core, as in the formation of foreign-body
capsules.

A wholly different pattern results if a tissue contains a focal area
that contracts. In this case, illustrated in the top row of diagrams
(Ao-^Ac), the matrix adhering to the shrinking area is subject to radial
tensions, and the surrounding tissue will thus be gathered into a star-

DIFFERENTIAL GROWTH I75

shaped pattern. Cell movements and divisions will be directed at the
area, instead of around it as in the preceding example. Many embry-
onic invaginations (e.g. blastopore, neural tube, nasal pits) seem to
take their origin from such localized contractile centers in the surface
coat of the germ (24),

Figure 2. Diagrams illustrating the effect of local contraction (top) and local ex-
pansion (bottom) on the structure of a reticular matrix.

It was very illuminating to find that certain biochemical activities
attending growth and differentiation may lead to the same type of ex-
pansion and contraction patterns (45). The expansion type, for in-
stance, is observed around areas with proteolytic activity. When a local
process dissolves part of an erstwhile coherent protein network, the re-
maining portion, owing to its elastic tension, retracts from the hole and
assumes a pattern like that shown in Figure 2 Be, much like a pierced
cobweb. A contraction pattern such as in Figure 2 Ac, on the other hand,

176 PAUL WEISS

arises around actively proliferating centers through the following chain
of events. Cells in proliferation, and apparently only then, liberate agents
with strong dehydrating potency. The result is a progressive condensa-
tion of the surrounding fibrous phase of the medium. Such local con-
densation in a rigid tissue frame generates tensions, which in turn
orient the condensing system toward the center of shrinkage. The fa-
miliar ray-shaped growth of ordinary tissue cultures in blood plasma is
due to this efifect. Analogous configurations can be recognized in many
embryonic formations. By its radial orientation such a contraction pat-
tern can evidently guide peripheral cells toward the center. This is one
of the major mechanisms by which cells are drawn toward distant des-
tinations, as if "attracted." Its morphogenetic importance becomes par-
ticularly impressive when several such centers act on a common matrix,
where they create interconnections of great regularity among the other-
wise independent centers. As is readily understood from Figure 3, two

Figure 3. Diagram illustrating the orientation of a reticular matrix along the con-
necting line between two centers of contraction.

contractile centers produce the greatest amount of shrinkage, hence ten-
sion, and thereby orientation of the matrix along their connecting line;
this automatically establishes a bridge for cell traffic between the centers.
Three centers determine a triangle (Plate IV), and so forth. In prin-
ciple, these experiments have solved the problem of orientation in cell
movement and growth.

In acknowledging mechanical stress as a major morphogenetic factor,
two reservations must be borne in mind. Firstly, tension is not the only
agent capable of orienting tissue structure. Any other physical force that
is capable of afifecting the orientation and aggregation of polar molecules
(electrostatic fields, electrophoresis, streaming, etc.) may have com-

DIFFERENTIAL GROWTH 1 77

parable effects. Secondly, stress is decidedly not a determining factor in
cytodifferentiation. Though it seems to be prerequisite for some cell
types in order that they may be able to carry out terminal transforma-
tions for which they are already predisposed by previous differentiation,
it is not the cause of such differentiation. Without stretch, myoblasts
cannot transform into muscle fibers (39), but stretch itself neither
makes a cell into a myoblast nor ever transforms a non-myoblast into
a muscle fiber, as has once been claimed (6).

An interesting combination effect of both contractile tension and
proteolysis has recently been observed in the healing of nerve stumps
(57). Tension transmitted from stump to stump through an intervening
blood clot, intensified by the syncretic contraction of the clot itself, ori-
ents the fibrin network in the latter in a prevailingly longitudinal direc-
tion. The oriented strands tend to fuse and thus become heavier than
their irregularly arranged cross links. Then proteolytic enzymes appear
on the scene and rout out the cross links, while leaving the straight linear
bundles standing. Thereby preferentially linear order becomes strictly
linear.

The main value of these varied results is that they demonstrate in
principle how chemical activity (proteolysis, dehydration) can engender
spatial order of a high degree, although it has itself no space order.
Through its physical and chemical by-products, plain growth thus leads
to ordered growth. As the order, arrangement, and consequent exposure
of the growing tissue reciprocally affect the condition, manner, and
hence, rate of its further growth, we realize how complex and diverse
are the factors that contribute to the growth of a given tissue. Some of
the factual sequences are summarized in Figure 4. Many more are still
unexplored. Carrying this exploration of the "microtechnology" of de-
velopment, i.e. of "Entwicklungs-mechanik" in its original sense, for-

GROWTH

^ -^ \ ^ PROTEOLYSIS

PROLIFERATION \ secretion

7 SHIFTS
DEHYDRATION ^

\ <'

CONTRACTION STRETCH EXPANSION

TENSION

ELECTR.POT.f?) T HfORTLOWf?)

MICELLAR ORIENTATION

r

ORIENTED GROWTH
Figure 4.

178 PAUL WEISS

ward on an intensified scale will go far in rectifying the current rather
abstract and often patently unrealistic concepts of "differential growth."
But even the existing knowledge, of which we have presented some
samples, is sufficiently compelling to discredit the growing fashion of
reducing, on paper at least, problems of differential growth to terms of
simple chemical reaction rates, with "growth stimulators" and "growth
inhibitors" called upon to do the "regulating."

Elaboration of Growth Patterns

According to the preceding sections, differential growth is simply
a corollary of differentiation, and differences in growth rates result
from a great variety of causes. As such, we have singled out the unequal
depletion of reproductive protoplasm in the building of specialized dif-
ferentiation products in different cell types, the various inequalities aris-
ing from shifts, aggregations and dispersals of cell groups in accordance
with properties acquired during differentiations, the resultant differ-
ences of geometric configuration, and hence, exposure of different parts
of a tissue, with the ensuing confinement of proliferative activity to
circumscribed zones of varying sizes, the different fates of differentia-
tion products, which are retained in some tissues but extruded in others,
and other related disparities contingent upon the divergent courses of
differentiation. Such inequalities in the growth pattern can, in turn,
secondarily influence the setting for subsequent steps of differentiation,
and the resulting changes will further modify the growth pattern. The
actual elaboration of the mature body thus appears as a complex se-
quence of transforming patterns in which the contributions of differen-
tiation and differential growth are intimately interlocked. Basically,
however, differential growth is not an agent but merely a product and
index of differentiation.

This realization is of the utmost importance when it comes to inter-
preting the manifold general non-localized influences which modify the
growth process during later developmental stages, and which are im-
pressively illustrated by the diversity of phenotypic variants that can be
obtained from genetically identical germs: giants, dwarfs, duplications,
defects, disproportions, excesses, and suppressions. Since these can
often be directly related to aberrations of nutrition, hormones, oxygen
supply, physical factors (radiations, pressure), and even climatic
changes, all of these agents have at one time or other been claimed as the
"dominating" factors in the growth process. We realize now that any
factor that has an effect on the physico-chemical setting in which growth

DIFFERENTIAL GROWTH 1 79

occurs may thereby alter the manifest form which the growth process
will take. But it is evident that such actions will be in the nature of
merely modifying already established patterns rather than instituting
wholly new ones. Consequently an agent applied systemically (non-
topically) may accentuate or reduce inherent differentials but cannot
create differentials not yet in existence. A few illustrations may be in
order.

As the developmental realization of the primordial growth patterns
of the germ progresses, small initial differences among parts become
greatly amplified. Minor deviations from normal become exaggerated.
Therefore, if each part were to keep on growing independently of the
other parts, the harmony of development would soon be endangered.
The setting up of new systemic growth controls coordinating the sepa-
rate parts thus becomes vital for harmonious development. This is
achieved through the humoral, more specially the circulatory, systems
for effective and rapid diffusion of substances. But we understand now
that a humoral pool could exercise no discriminative control over the
various parts supplied by it unless these parts had already previously
acquired different response qualities. For instance, the fraction of a
common supply that a given part will be able to secure is a function of
the accessibility of the part to the circulatory channels; this relation,
however, is entirely a matter of previous differentiation and growth of
both the part and the vessels. Similarly, the capacity of a part electively
to remove, concentrate, utilize, or store certain components from a com-
mon pool only reflects affinities of its cells acquired during their prior
biochemical specialization; affinity for iodine in the thyroid cell, for
vitamin A in the sensory cell of the retina, etc. Or if one cell group re-
sponds to a circulating hormone in a manner or degree different from
that of another cell group, equally exposed, the two must have been
already different constitutionally at the time the hormone appeared on
the scene. The case is no different from that of selective drug response
of mature cells, which is likewise based on the specific sensitivity of the
reacting targets.

Since no nutrient or hormone or other chemical of ubiquitous distri-
bution can make a cell render any contribution for which that cell has
not been predisposed by its previous differentiation, none of those agents
can be rated as primary causes of differential cell behavior. But they do
affect the subsequent elaboration and final proportions of the growth
pattern in a variety of ways. A general deficiency of nutrients, oxygen,
or specific growth accessories, for instance, hits different growing parts

l8o PAUL WEISS

differentially, depending on their respective metabolic rates, margins of
tolerance, reserves, substitutive capacities, and the like. Faced with de-
ficiencies, they compete for the available requisites in what Roux has
described as the "struggle among parts." One part's loss may either be
another part's gain, or may doom other parts, depending on the mode
and degree of their mutual dependence. The production of differential
growth defects by different vitamin deficiencies (59) may be pointed
out as a most instructive example. But the lessons from such deficiencies
are not reversible. While increasing the supply of growth essentials
from deficient to adequate levels will benefit growth correspondingly, a
further increase to levels of overabundance does not keep on enhancing
growth indefinitely. A limit will be reached when the production plant
works at full capacity, the maximum rate of production being deter-
mined by the intrinsic rates of the component processes, which, because
of the assembly-line character of growth, are limited by the slowest links
in the chains. A fuller discussion of these problems has been ably pre-
sented in other chapters of this book.

Of course, since what we call the "normal" level of nutrients, vita-
mins, hormones, and other growth requisites still falls short of the at-
tainable optimum, the growth potentialities of the organism are "nor-
mally" not fully realized. Raising such growth factors to supranormal
dosages can therefore occasionally have the spectacular effect of bring-
ing out latent formations which under "normal" circumstances would
have remained unrealized. For example, overdosing a female opossum
with sex hormones causes the formation of prostate glands (26), a
capacity inherent in the urogenital tract but failing to come to expres-
sion at "normal" hormone levels.

To the various humoral principles thus far recognized as secondary
regulators of growth will perhaps have to be added an even more gen-
eral category, namely, systems of the antigen-antibody type consisting
of organ-specific discharges and their molecular counterparts. The case
for this assumption has been presented on an earlier occasion (54), to-
gether with supporting experimental evidence. In the most condensed
version, it is as follows. The specific key molecules which arise progres-
sively in the course of differentiation (see above, page 153) act as tem-
plates ("positives") in the intracellular reproduction of more of their
kind. At the same time they cause the formation of molecular species of
complementary shape ("negatives"), which may or may not be an in-
tegral step in the reproductive process itself (double reversal). Com-
plementary pairs could inactivate each other by conjugation. The

DIFFERENTIAL GROWTH l8l

amount of free active growth catalysts in a cell would thus vary with
the excess of one type over the other. Now, if we assume that the "posi-
tives" are retained in the cell (e.g. because of large dimensions), while
the "negatives" can escape, circulate, and enter other cells (e.g. because
of smaller size), we can see the general outlines of a highly specific and
sensitive mechanism of growth regulation, which would coordinate all
cells of common character into a single operational system, no matter
how much they are scattered. Local reduction (or augmentation) of
growth in any part of the system would imply a corresponding decline
(or increase) of both "positives" and "negatives" in that locality, and
because of their easier diffusibility, of the "negatives" only throughout
the system. Since systemic reduction of the "negatives" means an in-
creased ratio of active over conjugated "positives," a compensatory in-
crease of growth throughout the rest of the system will ensue; and con-
versely, enhancement of growth in one part will automatically entail
depression of growth in all other members of the system. Thus the total
growth of a given system would be held under joint control of all its
components.

That some such mechanisms exist is becoming increasingly evident
from a study of compensatory growth reactions. That these mechanisms
operate, at least in part, after the fashion of immune reactions is indi-
cated by experiments in which organ-specific antibodies were shown to
have selective effects on the growth of the homologous organs (54).
But the precise mode of operation here sketched by way of illustration
is sheer supposition. Whatever its nature may eventually turn out to be,
one must remember that it constitutes merely another contributing fac-
tor to the highly intricate system of dependencies and interactions which
control the growth process. It is no more the master key to differential
growth than are any of the other factors discussed before.

Conclusion

What our whole discussion adds up to is that there is no single master
clue to the problems of differential growth or of growth in general. The
measurable growth of different parts is so intimately dependent on their
peculiar configurations and cellular differentiations, which themselves
differ sharply and qualitatively, that a purely quantitative comparison
of different growth processes is a deliberate abstraction. There are in-
stances where such abstraction has led to the recognition of striking
regularities capable of mathematical formulation (41) (e.g. allotropic
growth (19)). Whether these regularities are due to the fact that the

l82 PAUL WEISS

compared systems resemble each other closely enough in their mode of
growth to be essentially commensurable, or to the existence of some un-
defined principle of organismic nature, exercising overall control over
the multiplicity of heterogeneous components involved in growth, can
only be decided after a detailed analysis of the underlying phenomena.
At any rate, a purely formal treatment of growth, as is often attempted
through the interpretation of growth curves, is only a valuable guide to
and supplement of, but never a substitute for, a precise analysis of the
different forms in which growth manifests itself.

There can be no research on growth as such. We can only study grow-
ing objects. And different growing objects follow different methods.
The growth of a differentiating metazoan introduces aspects not pre-
sented by the growth of a bacterial culture, and the growth of a higher
plant has still other peculiarities. As we have pointed out repeatedly,
each individual tissue and organ has its peculiar mode of growth. To
know growth we must therefore first break down each one of its mani-
festations into its constituent elementary processes and then study these
and describe them in objective terms. This is a long way to go, but
there is no short cut.

In our present primitive stage of knowledge impatient attempts to
formulate a general and universal theory of growth seem to have little
chance of success. Those general theories that are sporadically being ad-
vanced are usually "general" only in the sense that they are first derived
from some rather special segment of the growth problem and then
broadly generalized by proclamation or mere implication. Consequently
they may be sound and pertinent within the area of their derivation but
gratuitous and invalid in their illegitimate extensions. Oversimplification
of facts and overgeneralization in theory are the main causes of the
existing discordance among the several contending concepts of growth.
Unless we desist from those practices the confusion is bound to grow, in
spite of all the excellent experimental work being done in various areas.

Growth is not a simple and unitary phenomenon. Growth is a word,
a term, a notion, covering a variety of diverse and complex phenomena.
It is not even a scientific term with defined and constant meaning, but a
popular label that varies with the accidental traditions, predilections, and
purposes of the individual or school using it. It has come to connote all
and any of these: reproduction, increase in dimensions, linear increase,
gain in weight, gain in organic mass, cell multiplication, mitosis, cell
migration, protein synthesis, and perhaps more. It is gravely inconsist-
ent to apply the most exacting standards of precision to our research

DIFFERENTIAL GROWTH 183

data and then proceed to mix into their description and interpretation
such vague terminology as this. The mixture can be no more precise
than its vaguest ingredient. To deal with growth as an entity, which can
be "activated," "stimulated," "retarded," or "suppressed," is only part
science, and for the other part, fiction. The less we let our work and
thoughts be misled by the delusion that "growth" is basically but a
simple elementary process, like a "bimolecular reaction," the faster will
be our progress toward true insight into the real mechanisms of devel-
opment. To promote this more factual approach to the problem of
growth, I have tried in this sketchy survey to portray the problem in its
natural complexity. If the discussion has helped to make clear what the
problems are, it will have served its purpose, even if some of the special
interpretations and hypotheses presented may not stand the test of time.

REFERENCES

1. Ballard, William W. Mutual size regulation between eyeball and lens
in Amblystoma, studied by means of heteroplastic transplantation. /.
Exp. Zool., 81, 261, 1939.

2. Bear, Richard S., Francis O. Schmitt, and John Z. Young. Investiga-
tions on the protein constituents of nerve axoplasm. Proc. Roy. Soc.
London, Ser. B, 123, 520, 1937.

3. Bernal, J. D. Structural units in cellular physiology, in The Cell and
Protoplasm. Publ. Am. Assoc. Adv. Sci. No. 14, 199, 1940.

4. Bloom, William. Cellular differentiation and tissue culture. Physiol.
Rev., 17, 589, 1937.

5. Brachet, Jean. Embryologie Chimique. Masson & Cie., Paris, 1944.

6. Carey, Eben J. Direct observations on the transformation of the
mesenchyme in the thigh of the pig embryo (Sus scrofa), with especial
reference to the genesis of the thigh muscles, of the knee- and hip-
joints, and of the primary bone of the femur. /. Morph., 57, i, 1922.

7. Chevremont, M., and S. Chevremont-Comhaire. Recherches sur le
determinisme de la transformation histiocytaire. Acta Anat., i, 95,

1945-

8. Doljanski, L. Sur le rapport entre la proliferation et I'activite pig-

mentogene dans les cultures d'epithelium de I'iris. Compt. Rend. Soc.
Biol, 105, 343, 1930.

9. Dalcq, Albert. L'Oeuf et son Dynamisme Organisatciir. Albin Michel,
Paris, 1941.

10. Danielli, J. F. The cell surface and cell physiology, in Bourne's "Cy--
tology and Cell Physiology." Clarendon Press, Oxford, 1942.

11. Dawson, Alden B. Cell division in relation to differentiation. Growth
(Suppl., Second Symposium), 91, 1940.

184 PAUL WEISS

12. Detwiler, S. R. Restitution of the medulla following unilateral ex-
cision in the embryo. /. Exp. Zool., pd, 129, 1944.

13. Ebeling, A. H. A pure strain of thyroid cells and its characteristics.
J.Exp. Med., 41, 337, 1925.

14. Fischer, Albert. Tissue Culture. Levin and Munksgaard, Copenhagen,

1925.

15. Gerard, R. W. Nerve metabolism. Physiol. Rev., 12, 469, 1932.

16. Gliicksmann, A. Development and differentiation of the tadpole eye.
Brit. J. Ophthal., 24, 153, 1940.

17. Harrison, R. G. Correlation in the development and grovi^th of the eye
studied by means of heteroplastic transplantation. Arch. f. Entw.
-Mech., 120, I, 1929.

18. Holtfreter, Johannes. Neural differentiation of ectoderm through ex-
posure to saline solution. /. Exp. Zool., g^^, 307, 1944.

19. Huxley, Julian S. Problems of Relative Grozvth. Methuen & Co.,
London, 1932.

20. Hyden, Holger. Protein metabolism in the nerve cell during growth
and function. Acta Physiol. Scand., 6 (Suppl. xvii), i, 1943.

21. Just, E. E. The Biology of the Cell Surface. P. Blakiston's Son & Co.,
Philadelphia, 1939.

22. Klatt, B. Filtterungsversuche an Tritonen. L AUgemeines. Wachstums-
verhaltnisse. Darmlange. Arch. f. Enfzv. -Mech., 10/, 314, 1926.

23. Langmuir, L Molecular layers. Pilgrim Trust Lecture, Proc. Roy. Soc.
London, Ser. A, 770, i, 1939.

24. Lewis, W. H. The superficial gel layer and its role in development.
Biol. Bull., 8/, 154, 1944.

25. Mollendorff, M. von. Dber Histiozytenbildung aus Fibrocytenrein-
kulturen des erwachsenen Kaninchens nach leichten chronischen Reiz-
ungen. Zeit. Zellforsch., 12, 559, 1931.

26. Moore, Carl R. Prostate gland induction in the female opossum by
hormones, and the capacity of the gland for development. Am. J. Anat.,

7^, I. 1945-

27. Nageotte, J. L' Organisation de la Maticre dans ses Rapports avec la

Vie. Alcan, Paris, 1922.

28. Needham, Joseph. Chemical Embryology. Cambridge University Press,

1931-

29. Needham, Joseph. Chemical heterogony and the ground-plan of ani-
mal growth. Biol. Rev., p, 79, 1934.

30. Palay, Sanford L. Neurosecretion. V. The origin of neurosecretory
granules from the nuclei of nerve cells in fishes. /. Comp. Neur., jg,
247, 1943.

31. Parker, Raymond C. Physiologische Eigenschaften mesenchymaler
Zellen in vitro. Arch. f. Exp. Zellforsch., 8, 340, 1929.

32. Pauling, Linus. Molecular structure and intermolecular forces, in
Landsteiner's Specificity of Serological Reactions. Harvard Univer-
sity Press, 1946.

DIFFERENTIAL GROWTH 185

;^^. Runnstrom, John, and Ludwik Monne. On some properties of the
surface-layers of immature and- mature sea-urchin eggs, especially the
changes accompanying nuclear and cytoplasmic maturation. Arkiv fiir
Zoologi, jd, I, 1946.

34. Schmitt, Francis O., R. S. Bear, and K. J. Palmer. X-ray diffraction
studies on the structure of the nerve myelin sheath. /. Cell, and Coinp.
Physiol., 18, 31, 1 94 1.

35. Schoenheimer, R. The Dynamic State of Body Constituents. Harvard
University Press, 1942.

36. Scott, G. H. Mineral distribution in some nerve cells and fibres. Proc.
Soc. Exp. Biol. Med., 44, 397, 1940.

37. Spemann, Hans. Embryonic Development and Induction. Yale Uni-
versity Press, Nev^^ Haven, 1938.

38. Spiegelmann, S. Physiological competition as a regulatory mechanism ,
in morphogenesis. Quart. Rev. Biol., 20, 121, 1945.

39. Stillwell, E. Frances. Cytological study of chick heart muscle in tissue '
cultures. Arch. f. Exp. Zellforsch., 21, 446, 1938.

40. Thomas, J. Andre. Des epitheliums en culture et dans I'organisme.
Arch. f. Exp. Zellforsch., 22, 15, 1939.

41. Thompson, D'Arcy W. On Growth and Form. Cambridge University
Press, 1942.

42. Toro, E. t)ber Einpflanzung von Gewebekulturen. Arch. f. Zellforsch.,
75, 312, 1934.

43. Waddington, C. H. Organisers and Genes. Cambridge University
Press, 1940.

44. Weiss, Paul. Functional adaptation and the role of ground substances
in development. Am. Nat., 67, 322, 1933.

45. Weiss, Paul. In vitro experiments on the factors determining the
course of the outgrovi^ing nerve fiber. /. Exp. Zool., 68, 393, 1934.

46. Weiss, Paul. Selectivity controlling the central-peripheral relations
in the nervous system. Biol. Rev., 11, 494, 1936.

47. Weiss, Paul. Principles of Development. Henry Holt & Co., New
York, 1939.

48. Weiss, Paul. The problem of cell individuality in development. Am.
Nat., 74, 34, 1940.

49. Weiss, Paul. In vitro transformation of spindle cells of neural origin
into macrophages. Anat. Rec, 88, 205, 1944.

50. Weiss. Paul. The morphogenetic properties of frozen-dried tissues.
Anat. Rec, 88 (Suppl.), 48, 1944.

51. Weiss, Paul. The technology of nerve regeneration : A review. Suture-
less tubulation and related methods of nerve repair. /. Neurosurg.,
7. 400, 1944.

52. Weiss, Paul. Evidence of perpetual proximo-distal growth of nerve
fibers. Biol. Bull, 87, 160. 1044-

53. Weiss. Paul. Experiments on cell and axon orientation in vitro: the
role of colloidal exudates in tissue organization. /. Exp. Zool., 100,

353. 1945-

l86 PAUL WEISS

54. Weiss, Paul. The problem of specificity in growth and development.
Yale J. Biol. Med., 19, 235, 1947.

55. Weiss, Paul, and Helen B. Hiscoe. Experiments on the mechanism of
nerve growth. /, Exp. Zool., loj, 315, 1948.

56. Weiss, Paul, and R. Amprino. The effect of mechanical stress on the
differentiation of scleral cartilage in vitro and in the embryo. Growth,
4, 245, 1940.

57. Weiss, Paul, and A. Cecil Taylor. Histomechanical analysis of nerve
reunion in the rat after tubular splicing. Arch. Surg., 4/, 419, 1943.

58. Weiss, Paul, and Hsi Wang. Transformation of adult Schwann cells
into macrophages. Proc. Soc. Exp. Biol. Med., 38, 273, 1945.

59. Wolbach, S. B., and O. A. Bessey. Tissue changes in vitamin defi-
ciencies. Physiol. Rev., 22, 233, 1942.

60. Wright, Sewall. Physiological aspects of genetics. Ann. Rev. Physiol.,

7, 75, 1945.

VIII. PROBLEMS OF ORGANIZATION

BY J. S. NICHOLAS^

The mystical haze that often surrounds the word 'organization' has no
place in science. . . . Particulate units at any level are not wholly inde-
pendent of one another. The relations of the particles are part of the
system and it is their behavior in relation to one another that constitutes
'organisation.' These relations are amenable to experimental investiga-
tion and are usually expressed in some kind of configuration. . . . No
particle or unit can he clearly understood or its behavior predicted
unless its reactions with others are taken into consideration.

— HARRISON, 1940

IN the many years during which descriptive and experimental inves-
tigations have been carried out in the effort to elucidate the steps
by which the developing egg gives rise to the structures of the
adult there has been little change in the formulation of the main prob-
lems of organization. The attention of the embryologist is still largely
focused upon the questions arising from the investigation of the prob-
lems of localization, differentiation, polarity, movements of parts and
the interaction of tgg components.

His's expression of the dominant problem of his period, the localiza-
tion of parts in the ^gg, has developed into the newer aspect of the
localization of potencies for the development of these parts, and also
the changes that may be brought about by modifying the interaction of
such potencies. This may be on any level of organization whether it is
in the cellular layer or even at the tissue level. The movements of parts
and the mechanism responsible for cell and layer migrations forms an
important phase of the present approach to the fundamental problems
given above.

The problems of organization remain as expressed in the last century,
the expansion and methodological approach to those problems has
changed markedly and the concepts of the past have undergone radical
change in response to the present experimental approach. Some of the
changes in concept concern the ideas of the integrity of the cell as a
functional unit, while chemically the molecular level has evolved as
a biophysiological unit and physically the atom with its association of
electrons has come into being. Embryologists are still engaged with

1 Osborn Zoological Laboratory, Yale University.

l88 J. S. NICHOLAS

large units whose extrinsic relations are engaging attention, and in the
main they are paying little heed to those intrinsic cellular relations
which must be of such dominant import in the final extrinsic reactions
which are formative of the patterns through which the organizational
structural relationships arise.

The unitary influence of cytoplasm and nucleus in the initiation of
cell division was a moot question at the beginning of the century. Since
in some of the determinant forms the first division coincides with the
later axial relations the problem of nuclear or cytoplasmic dominance
in the activation of the primary division becomes a rather vital ques-
tion. It is clearly recognized today that normally there is an interacting
significance of each of these parts in which certain components are
actively engaged, but that experimentally there are multiple mechanisms
for attainment of the result.

The Problem of Nucleo-cytoplasmic Relationships

In the sea urchin Boveri ('07) has shown that so long as the chromo-
somal complex of the egg contained a normal complement of chro-
mosomal material development proceeded regularly. Double or half of
the normal series might be represented, the quantity apparently having
little significance so long as the quality of each of the constituent
chromosomes was present. Baltzer ('40) and Hadorn ('36) have been
able to eliminate the action of the female pronucleus and consider only
the reaction of the paternal pronucleus with reference to egg develop-
ment. Complete development has been attained, as would be expected
from some of the preceding results with invertebrate eggs. Conklin
('31) on the basis of much work has decided that the cytoplasm has a
very complex and interesting role in the process of development; his
results were indicative of the fact that the cytoplasm itself was prob-
ably of a greater independent activity than hitherto suspected. Fank-
hauser ('34) in his study of the polyploidy has likewise shown quanti-
tatively that the influence of the nucleus must be regarded as being
comparatively negligible upon the total process of differentiation.

E. B. Harvey ('36 and '46) has made an interesting series of studies
upon the Arbacia egg. She has employed centrifugation in separating
the uncleaved egg into four parts in each of which the constituents
separate regularly. After the process of separation is complete, the
fragments are activated parthenogenetically. Subsequent cell cleavage
takes place and may proceed to the formation of as many as 500 cells ;

PROBLEMS OF ORGANIZATION 1 89

all this may occur from a fragment which contained neither maternal
nor paternal chromosomes. There is a limit to development, and while
there is a preliminary organization, development ceases shortly after
the blastula is formed. In the ('46) contribution the clear quarter con-
taining the nucleus was fertilized after centrifugation and regular
plutei were obtained. In this series of experiments the action of the
cytoplasmic granules was analyzed. She concludes that the matrix or
ground substance is the material of importance and that the cytoplasmic
granules can be ruled out as requisite to the continued development here
attained. The granules are reconstituted after the blastular organiza-
tion is reached but not from previously localized granular material.

There is one important observation in Harvey's ('46) paper which
is not stressed in the interpretation but which may play an important
part in the organizational complex. When the clear quarter is fertilized,
soon after fertilization the fertilization membrane usually breaks with
extrusion of the oil cap. If, however, fertilization is performed, one to
three hours later development may be regular. It may be that time is a
requisite factor for the molecular reorientation necessary for the ade-
quate formation of the fertilization membrane. If this is caused to form
without the necessary cortical reorganization, the deficiency either
qualitative or quantitative in the surface molecular layer manifests
itself in a degree of fragility which causes the breaking of the layer
and the expulsion of the oil cap. The temporal factor permitting of
material organization seems here to be of tremendous import.

While the step by step analysis of the parts of Arbacia eggs has pro-
ceeded, it has shown clearly that the capacity of the embryo to undergo
its primary organization can be completed under rigorous experimental
ablation of many of its components. Certain combinations are necessary
for a complete development of the pattern but the pattern itself can be
broken down into parts which can be separated in time and action into
interrelated series, all combining in the interaction necessary for com-
plete embryo formation.

One clear example of the influence of nucleus in development is
shown in the inverse symmetry of snails (Boycott and Diver ('23)
and Diver, Boycott, and Garstang ('25)). It will be remembered that
they studied a group of left-spiraled or sinistral types with which they
conducted experiments breeding the dextral type to the sinistral type
and securing its results in the offspring which indicated that there was
a definite recessive present so that the genetic results appeared not in
the first generation of crossing but in the second generation. There is

190 J. S. NICHOLAS

a skip of one generation before the effect manifests itself. They have
interpreted this material as showing that the nucleus has a definite
effect upon the cytoplasmic relationship, and also that the cytoplasmic
forces cause a lag in the appearance of the nuclear constituents which
are responsible for the reversal of asymmetry in the second generation.
There is here an interesting link of reaction between cytoplasm and
nucleus, one which indicates that there is an interrelationship of rather
vital significance for the entire field of early organization. Lillie's ('29)
theory of the allocation of the various parts of the organism is well
known, a differentiation through segregation and differential dichot-
omy resulting in a definite localization of the different parts in a
rather quantitative way. Their qualities are quantitatively disseminated.
The analysis of the process of segregation shows that there is more
than a simple quality or quantity which has undergone localization in
the course of its movement to different regions. A tissue which may
become part of one definite organization in one region becomes part of
an entirely different set of organizations in another. The change in
positional setting is an important process in organization.

An interesting and entirely convincing experiment in which genie
control is dominant over all other factors is shown in Robertson's ('42)
paper. By transplanting the ovary of the homozygous yellow mouse into
black agouti hosts a normal environment was provided for develop-
ment. The lethal factor, however, appeared in exactly the same way as
if the embryos were developing in a heterozygous yellow female. The
genie mechanism is clearly responsible for the lethal effect in the homo-
zygous yellow. Just what occurs in the blastocyst at the critical stage
at which the lethal characters become evident is not yet known.

The Problem of Egg Constituents

Surface changes in the egg are but an indicator of internal changes
which are magnified in the peripheral surface movements. The egg is
not a simple sphere with an outer limiting surface on which one who
runs may read, but it is a nebulous system in which there are multiple
reaction spheres delimited by the relations of movement within an
inner and an outer layer.

If such could be held, then the primary arrangements within the egg
are due to differences in mobile capacity of the various substances
within a comparatively homogeneous system. The internal center is a
point of relative homogeneity from which forces emanate with a con-

PROBLEMS OF ORGANIZATION

191

stantly increasing reactivity towards the periphery. This concept is a
3 -dimensional translation of Dalcq's ('38) cortico-vegetal gradient and

Figure i.

according to it Dalcq's expression would be simply the resultant of the
activities of the multi-sphere relationships expressing their mobile
potentiality with reference to each other. Such a concept has the ad-
vantage of focusing attention upon the factors incident to the internal
configuration of the embryo which have received too little attention.

This thesis does not deny the importance of the surface in the read-
justments of polar diflferentiation subsequent to fertilization. It also
would explain the sorting out of processes which occur in the early
period as well as the differences in bioelectric potentials which could
easily be the resultants of movements of differently charged particles
and their reactions during passing phases in which electrical fields
might be formed.

192 J. S. NICHOLAS

Harrison ('45) bases the polarity of the egg upon three factors: the
dipole character of protein molecules and their resultant orientation,
regional gradational dififerences in the composition of dipole molecules,
and regional differences due to cytoplasmic inclusion. Schrodinger
('45) postulates an aperiodic crystal to account for some of the situa-
tions which occur in living systems. Needham ('42) invokes a para-
crystalline state or a liquid crystal, but this state would have to be a
relatively late one, probably occurring in the early neurula.

The question of gradients within the egg and the effects of different
things upon them has long been subject to much speculation. In a recent
article Child ('46) discusses the Spemann organizer concept in rela-
tion to the various gradient systems. The results are suggestive but not
conclusive. The experimental data concerning the Spemann organizer
is confined entirely to the amphibian embryo. Child points out that the
primary organizer is not primary, since it does not determine external
or other axiate patterns in natural development. He bases this idea on
Holtfreter's ('44a) and Earth's ('41) work on neuralization without
organizer action. He neglects to point out the fact that the amphibiolo-
gists themselves early deserted the idea of the Spemann organizer as
the only single factor responsible for all organization. While it is ex-
ceedingly important in nervous system formation, there are other fac-
tors which come into play, some of which have been set in motion far
earlier than the Spemann organizer, which cause the movements and
reactions of certain parts of the egg. Child has intentionally focused his
attention upon the later stages of development in the amphibian and
has not mentioned the work which deals with earlier organization
patterns.

This is particularly true as we look back and survey some of the
literature which has to do with the organization complex in fishes.
Morgan in 1893 reported that the yolk might be removed from the egg
of Fundulus and that the embryo would still continue to form. Oppen-
heimer ('36a) has shown that the blastomeres may be removed from
Fundulus eggs and that if not removed too early in development the
disc will gastrulate and will develop embryonic structures or even com-
plete embryos. Tung, Chang, and Tung ('45) have confirmed these
results upon the eggs of goldfish, Carassius auratus. Fragments con-
taining the blastodisc but less than half the yolk formed hyperblastulae.
Fragments of the same size derived from 4-cell stages formed embryos.
After the 8-cell stage the isolated blastodisc formed embryonic struc-
tures or embryos. Oppenheimer ('34, '36b) postulated that some sub-

PROBLEMS OF ORGANIZATION 193

Stance or material passed from the periblast during the course of early
cleavages. The postulate holds better for Fundulus where complete
embryos are not obtained until after the 32-cell stage as compared with
Carassius where similar formations are obtained at the 8-cell stage.*
Tung, Chang, and Tung consider that the passage of necessary materials
is accomplished earlier in Carassius because of the small proportion of
yolk present. Lewis and Roosen-Runge ('42) have studied cinemato-
graphically the movements in fish eggs (Danio) and find that there is
an early movement originating in the vegetal region of the tgg after
fertilization which up to now has not been considered a general reac-
tion of the differentiating organism. These movements are quite pro-
nounced, moving in a vortex from the vegetal pole towards the center
of the egg, resulting in rearrangements of materials at fertilization and
in migration of cortical parts definitely through the yolk substance to
form a layer, not uniform in its constituency, on top of the yolk between
the yolk and the blastoderm. This is a clear indication of the fact that
materials in early eggs still have certain visible effects which later are
to be correlated with the prospective parts of the developing organism.

The observations of Lewis and Roosen-Runge have their counter-
part in the study of rhythmical contractions observed by Yamamoto in
a series of papers extending from 1931 to 1938. All of his studies have
to do with the early rhythmical movements which are found during the
early stages of the egg before segmentation. The significance of these
is yet to be interpreted. They may be osmotic in nature but the resultant
end point of such a reaction will bring about changes within the egg
and cause material dispositions which have a bearing on later organic
localizations.

Morphogenetic movements have been plotted in various groups of
vertebrates, first in the amphibian by Vogt ('25, '29) in his classic
monographs on both the urodele and anuran, then on fishes by Pasteels
('33' '34) ^"d by Oppenheimer ('35, '36b) and later Luther ('36), by
Pasteels ('36) for the chick now supplemented by Spratt's ('46) studies
with carbon particles, and in mammals by Vanderbroek (unpublished).
The vital staining experiments analyze to a large degree the paths fol-
lowed by cells and by cellular aggregates during the course of their
subsequent assumption of position in the embryo.

It gives, however, no index of the effective agents which act upon the

* During the summer of 1948 Dr. T. C. Tung was able to secure complete embryos
in Fundulus after isolating the embryonic blastodisc at the 4-cell stage (personal com-
munication).

194 J. S. NICHOLAS

cells themselves or which cause them to do certain things. It gives a fate
map which, while telling the generalized movement of parts, tells noth-
ing about their specific mechanism of differentiation. For this phase of
study we are dependent upon other methods, all of which bring about
indirectly other effects which sometimes obscure the action of the ma-
terial itself, or the action of the system immediately surrounding it.
This is true of a great many of the transplantation or isolation studies
which, while they give interesting pictures of what a cell can do, tell
little concerning its reaction in its own specific pattern with regard to
its own specific organization. Furthermore, some materials can be
mapped less exactly than others. For instance, the teleost blastoderm
lacks pigment which is responsible for the maintenance of a persistent
color value. Stains likewise fade more rapidly in the teleost than in the
amphibian. The same is true in avian studies where the reduction of
dye sometimes leads to exceedingly erroneous interpretations. Extirpa-
tion or isolation of parts which are to be studied has given quite differ-
ent results from those secured when the same parts are integral com-
ponents included within a migrating tissue under normal circumstances
and surroundings. It is evident that some change occurs in the early
development of the fish egg which is responsible for later developments,
since an isolated blastodisc can continue development when isolated
after the 8- or i6-cell stage, but before that time it is incapable of
adjusting itself to the new environment in such a way as to give a com-
plete embryo formation. Parts may be formed but the total with its
organization is not consistently present. Tung, Chang, and Tung ('45)
have noted exactly similar conditions in the goldfish egg.

Another factor which seems to have received too little attention is the
egg-contained yolk. It constitutes a fundamental problem for each
embryo, for the developing structures are molded around the generally
spherical yolk content of the egg. The importance of this mass which
presents in many forms a surface around which the embryo must grow
was pointed out in a paper by Daniel and Yarwood ('39). They studied
the distribution of the yolk fragments in various stages of development
of the egg using coelomic, ovaducal, and fertilized eggs and compared
the stratification of the yolk platelets in each of these stages.

In the sectioned material there is a practical uniformity of distribu-
tion at the stages that were surveyed. Studies on later stages show that
there is practically no diminution in the size of the yolk platelets until
after it is formed in the particular cell in which it is destined to act as
a metabolite. Yolk platelets in different systems of the embryo are all

PROBLEMS OF ORGANIZATION 195

of the same magnitude, with the possible exception of the heavy aggre-
gates which are found in the lining of the alimentary tract. This is the
last material which is effective in differentiation and is really the begin-
ning of the circulating nutritive factor.

Further study of these stages with the application of the recently
developed histochemical tests is necessary for the interpretation of those
important agents that are deposited in the various substrates which are
present in yolk. Since this is the material which by chemical transference
becomes the tissue of the embryo, it is vital that the changes in it be
clearly analyzed in order that the relation between chemical changes and
the mechanics of development be adequately understood.

Organizational Dependencies

When we come to the study of relationships of organization in sepa-
rate systems, we begin to perceive how far organization enters into the
process and how much is intrinsic to cellular self-determination. Holt-
freter's ('39) early studies showed that tissues, when in combination,
produce far more than the individual elements by themselves even when
acting under an organization complex. When connective tissue is pres-
ent the cells tend to form more than when endoderm or ectoderm is
present by itself. Stableford ('39) found that it was difficult to isolate
the internal portion of the system without the necessary adjunct of
some of the mesoderm which is going to form connective tissues or
muscle layers. He indicates that the endoderm also is a source of meso-
derm formation and that isolated endoderm developing in Holtfreter's
standard solution can be shown to form mesoderm.*

Nieuwkoop ('46), however, has performed this experiment in re-
verse and has isolated the ectoderm-mesoderm complex from the endo-
derm. While there are definite deficiencies in the mesodermal structures,
there is an attempt by the organism to produce a complete somatopleure-
splanchnopleure combination, but the mesodermal structures are not
acted upon by the endoderm which has suffered early removal in the
early neurular stage. In these experiments Nieuwkoop has found that
two complete sets of limbs will evolve, one of them in the normal loca-
tion and from the normal structure of a somatopleure, the other formed
from the splanchnopleuric portion of the mesoderm at the same level
as the somatopleuric limbs. This is the strongest index that we have had

* Stableford in a subsequent study, as yet unpublished, concludes that marginal zone
is the sole source of mesoderm and therefore the endodermal capacity for mesodermal
origin must be subjected to further test.

196 J. S. NICHOLAS

SO far of what to expect when the action of the endoderm as a molding
structure is absent, but when the ectoderm and the nervous system are
serving in place of the endoderm. The process of regional localization
responsible for limb formation has been able to manifest itself because
of the absence of endodermal structures. The endoderm has not been in
place for the molding of the internal layer of the splanchnopleure and
in consequence the manifestation of limb formation has been expressed
in the splanchnopleuric component of the mesoderm. Is this an inhibi-
tion brought about by the endoderm, or is it an induction which has
moved regionally across in the mesoderm from what was formerly the
somatopleure to the splanchnopleure?

In Harrison's ('25) experiments in which the mediolateral axis of
the limb is reversed, we undoubtedly have the counterpart to this series
of experiments ; for Harrison, when he removed the mesoderm under-
neath the ectoderm comprising the limb bud, undoubtedly removed and
reversed the splanchnopleure, leaving the somatopleuric portion inter-
nally while the splanchnopleuric was on the outside. From this, perfect
limbs developed so that the splanchnopleuric part, already regionally
localized when placed in contact with reversed surroundings, cannot
effect its normal polarization of limb structures. Whether the somato-
pleuric layer of the mesoderm has influenced the splanchnopleuric or
not cannot be said, but the somatopleure and the splanchnopleure are, in
this experimental series, reversible.

That qualities do exist within groups of inducing cells which are
comparable to the factors responsible for regional determination in the
amphibian has been definitely demonstrated by Eakin ('39) for Salmo.
He has divided the invaginated archenteron roof or middle gastrulae
of Salmo into an anterior, a middle, and a posterior portion, wrapped
each in a tube consisting of extraembryonic half of a late gastrula, and
implanted the compound graft of the yolk sac epithelium of an older
Salmo embryo. In each case a control graft of the extraembryonic half
of a late gastrula was implanted in the same stage as the compounded
graft. The control pieces differentiated only into epidermis. When the
anterior portion of the archenteron roof was implanted together with
indifferent ectoderm, it exhibited little or no power of induction and
differentiated for the most part into digestive tube. When the middle
portion of the roof was transplanted, it differentiated notochord, so-
mites, gut, and pronephric ducts, inducing the accompanying ectoderm
to differentiate neural structures, primarily brain-like in character with
auditory vesicles. The posterior portion of the roof, when grafted, also

PROBLEMS OF ORGANIZATION 197

differentiated into chorda, somites, gut and mesonephric ducts ; in this
case, however, the neural structures induced from the accompanying
ectoderm were spinal rather than cranial in character. The three por-
tions of the archenteric roof are thus demonstrated to be essentially
similar to the prechordal plate, the head organizer, and the trunk
organizer respectively of the amphibian gastrula (cf. E. K. Hall ('37) ).

In fish organization before gastrulation the action of the different
components is more tentative than fixed, and while the tentative reac-
tions have certain definite sequelae which lead to the formation of
organized materials, they themselves are not irrevocably placed in posi-
tion where they cannot have different attributes from those which nor-
mally develop from them. During gastrulation there is a much more
specific organization of the blastoderm as a whole. The essential nature
of organizing forces operating in the integration of the developmental
processes remains to be elucidated by methods not yet at our command.
But in teleosts, where the formation of the embryo is limited to a rela-
tively small portion of blastoderm, there are demonstrable integrated
activities w^hose role is the coordination of the various parts of the
blastoderm. There is evidence both in Salmo and in Fundulus which
suggests the existence of such activity and which indicates that their
nature is not identical in the two forms.

The various parts of the blastoderm have different relationships with
one another. Luther ('36) divided or sectioned the blastoderm into six
sectors and found that not all parts of the blastoderm are equivalent in
embryo formation. At all stages investigated the potencies for differen-
tiation as indicated by the percentage of grafts differentiating were
highest in the embryonic sector and lowest in the lateral portions of the
blastoderm, diminishing gradually around the blastoderm during the
course of gastrulation. The differences between the median embryonic
and more outlying regions become progressively greater. No differentia-
tion was attained when a central portion of the blastoderm was isolated.
Luther interprets the results of these isolation experiments as demon-
strating the existence of w^hat he called a physiological field. He postu-
lated that the potency for differentiation in one particular sector was
related to the state of physiological activity of its constituent cells, which
in turn varies according to the position of the cells with respect to the
whole and according to their age. It is highest in the embryonic sector
and lowest in the extraembryonic. In the lateral embryonic and in the
extraembryonic regions it diminishes progressively during the course
of gastrulation. Luther ('37) considered his experiments both by

198 J. S. NICHOLAS

deletion and transplantation methods to strengthen the validity of his
interpretation. When a sector of more than 90 degrees was removed
from the embryonic area, no embryo formed. In the absence of embryo
formation, however, the blastoderm gradually expanded to cover the
yolk. When a 70-degree sector was removed from the embryonic area,
regulation occurred and embryo formation could still be accomplished by
what remained of the blastoderm. When an embryonic sector of 45
degrees was removed from the gastrula stage, the chorda and anterior
somites were lacking in some of the embryos which developed. No such
gradient has been observed in the Ggg of Fundulus (cf. Oppenheimer,
supra). However, the experiments are not exact counterparts of those
on Salmo. Small portions only of the germ ring were removed from
regions 90 degrees or 180 degrees away from the embryonic axis and
these were grafted into embryonic shields, or into the extraembryonic
membranes of an embryo of the same age as the donor. Such grafts
implanted on the extraembryonic membrane will differentiate any struc-
tures except epidermis, blood cells, and chromatophores. In contrast,
grafts from the 90-degree and the 180-degree germ ring of gastrulae
of various stages implanted into the shield differentiated without excep-
tion provided their cells were incorporated below the epidermis. The
nature of the structures differentiated bears no relationship to the sorts
of the graft ; head, trunk and tail structures were formed in graft from
both 90 degree and 180 degree regions of the germ ring. In some cases
grafts formed structures characteristic neither of the region of the host
to which they were transplanted nor of the regions of the embryo for
which the graft itself was originally destined.

Independent Movements

Movements of parts within the egg at early stages are largely deter-
mined by the physical reactions of the constituents. During later stages
the simpler physical determinants are obscured by specific and general-
ized processes of growth and differentiation. These processes are chem-
ical and distributional in nature. It has fallen to the descriptive embry-
ologist to relate the surface reactions which are immediately visible,
the analysis of the underlying orientations and their role in organization
as well as their later definitive action in aligning spatially localized parts
or organizers and relating these two components is a fundamental
problem of embryology which has so far gone unsolved.

There are many observations and partial attacks upon phases of this

PROBLEMSOFORGANIZATION 1 99

problem. Holtfreter ('43b, '44) has recently studied the problem in
the amphibian and has shown that many of the movements which we
associate with fertilization can occur in the unfertilized ^gg. The
surface movements of pigment and the initial cortical relationships of
the plasmogel simulate many of the relations of the fertilized egg
reactions. While some of these reactions were observed previously, the
various components have not been so completely traced and their sig-
nificance correlated as in the study referred to. In a second study he has
shown the contours of isolated endodermal cells taken from the am-
phibian gastrula. The blastoporal cells can be identified and spread in
the direction of their former proximal-distal axis. They exhibit a
definite precocious polarization whereas the non-polarized cells spread
in a constant radiating reaction. It must be remembered that these cells
are taken from the early gastrula, just as the endoderm is being invagi-
nated. There are two types of cells in the early gastrula, those which are
polarized and those which are nonpolarized. Nieuwkoop (see above)
assigns to the endoderm the role of generally organizing the mesoderm
which later is going to be found in the dorsal mesenteries and in a
primordial germ ridge with the general sensitivity for germ cell forma-
tion, acting therefore as a primary organizer of this tendency on the
part of certain of the mesoderm cells. If this be so, the polarization
which Holtfreter has observed has an added meaning, for it is the
polarized cells which exert the inductive influence.

In the study of the mechanics of gastrulation Holtfreter ('43a, '44b)
has demonstrated the influence of the surface coat which is found upon
the cells. Uncoated endoderm is incorporated into an endodermal sub-
strate, whereas a substrate of gill ectoderm will spread over a graft of
blastoporal endoderm. Endoderm will pentrate through a hole in the
epidermis and pass into an endodermal substrate. This and other factors
which are consistent with the formation of endoderm are treated by
Holtfreter and material which possesses a surface coat has a definite
affinity to invaginate into the substrate. However, when mesoderm is
placed upon a piece of glass and two pieces of ectoderm are placed in
contact with it, the ectoderm moves up along the outside of the meso-
derm and encloses it. In this case the mesoderm is uncoated, while the
ectoderm is coated. A reverse effect occurs when a coated piece of the
marginal zone is placed in contact with host ectoderm and invaginates
under the host ectoderm, which again is a covered type of tissue. In this
case we have two covered pieces moving together with the marginal
zone to form a type of archenteron. There seems to be no consistency

200 J. S. NICHOLAS

between the way in which the coat acts, but there is a definite specificity
in the way in which the tissues act, whether they are coated or uncoated.

Schechtman ('34) discovered the general process of ingression.
Schechtman placed stains upon the vegetal region of the egg and found
that they disappeared. When he came to examine his eggs further he
found that the stain had proceeded upward into the floor of the archen-
teron going into the blastocoelic floor and that there was a distinct
variation as to how far the material would progress according to the
stage at which it was placed upon the outside. If the stain was placed
upon the unfertilized egg it extended up into the floor of the blastocoele.
If it was so placed later, after thirty-two cells, it was just in the lower
portion of the floor of the blastocoele, extending to a very slight degree
outside the region of the cortex in which it was originally located. The
study of ingression seemed so important that it was deemed necessary
to carry the study through to the place where Vogt ('29) had stained
the early gastrula and followed the movements of particular parts. The
study has since been repeated by Nieuwkoop ('46) who confirms, in the
main, the results which were brought out by Schechtman, with one
minor variation. Since Nieuwkoop stains his materials in many cases
considerably later than Schechtman, he got a very distinctly different
picture of the final disposition of the stained material. When the stain
is applied early, it may proliferate according to extension action effects
by which it proceeds directly up the cleavage furrows, the so-called
centrifugal effect which Schechtman noticed when he repeated his
earlier results. This, however, is not a normal effect, but rather a diffu-
sion one, and the reaction of the egg must be measured in terms of its
diffusion constants as well as the changes to diffusion which are under-
gone according to its degree and state of development during the seg-
mentation stages.

Similar results are attained in my own study of this phenomenon
(Nicholas ('45) ). In my material, however, there is one step which can
be added. There are two distinct steps in the ingression system, the first
arising as described by Schechtman and the second occurring in those
specimens which are stained deeply and carry a second part of the stain
on the surface. It proceeds through the ventral lip and there is invagi-
nated, sometimes by apposition staining the dorsal lip. It is most promi-
nent in the floor of the archenteron and eventually merges with the
material which was formerly in the blastocoelic floor forming a con-
tinuum with it. The primary ingression is blastocoelic, the secondary
a gastrocoelic floor contribution. The first ingression forms the anterior

PROBLEMS OF ORGANIZATION 20I

portion of the gastrocoelic floor by an inversion of the blastocoelic floor
in the region in which the heart later forms. The gastrocoeHc invagina-
tion extends forward to the region of later liver formation. These
observations indicate mass movements of material which are responsible
for regional localization of determination reactions. This is corrobo-
rated by Nieuwkoop's results with a slightly different staining method
applied to different stages.

Mass Movements in Organization

Two examples of how mass movements are effective in organizational
phenomena will suffice. The first is the case of localization in the chick,
in which the differences between the older concept of the regional
differentiation of the parts of the blastoderm can be contrasted with the
results as secured from experimental data."

The His concept of the chick blastoderm was that of a central axial
localization in which the regional localizations are arranged in linear
order. The Rudnick diagram (even with the necessary modifications
imposed by Spratt's work) deserts the axial type of linear arrangement
in favor of a distribution of potencies for the formation of the parts
of the organism. Thus one finds heart, thyroid, liver, and chorda not
spatially localized in linear order but originally as potency areas which
later join to express the final types of specific organization.

The second example will be presented in a more complete form, for it
is the rather complete story of the localizations anticipatory to organiza-
tion in Drosophila.^

The egg nucleus lies close to the dorsal side of the egg; meiosis leads
to formation of three polar nuclei which fuse to form a residual body
under the dorsal surface near the female pronucleus. Following fertiliza-
tion the zygote nucleus divides a number of times and the nuclei are dis-
persed throughout the deutoplasm of the egg mass. The first migrations
or movements within the Drosophila egg are those towards the posterior
region of the egg where the pole cells later form. Two to five pole cells
are formed, each one of them a direct descendant of one of the nuclei
which lay originally within the deutoplasmic mass. These subsequently
divide to form a total of about thirty pole cells. After a latent period

2 Since the above comparison was made, Spratt ('46) has published his results with
carbon particles employed to study the movements of parts of the chick blastula. The
Rudnick diagram is therefore now subject to modification, but in the main shows the
cardinal localization of parts in the chick.

3 For this information I am deeply indebted to my colleague, Dr. Poulson.

202

J. S. NICHOLAS

A-UVER
t -HEART
® -CHORDA
A -THYROID
O-NEPHROS
X- INTESTINE
• -ERYTHROCYTES
T-MELANOPHORES

+ -skeletalmuscl£

Figure 2. The His concept of localization in the chick blastoderm compared with
Rudnick's summary of present findings in this form.

most of the remaining nuclei rather suddenly emerge from the central
deutoplasmic mass and are regularly arranged in a peripheral position,
where they remain, forming a single boundary layer completely enclos-
ing the remainder of egg materials. The pole cells remain separate from
the underlying blastoderm.

The events described can be regarded as the first generalized move-
ment of material within the egg of Drosophila. The deutoplasm still con-
tains a few nuclei, the yolk nuclei, which for some reason did not move
towards the periphery. During this process the fused polar bodies have
undergone a series of transformations during which the nuclear cycle is
synchronized directly with that of the blastoderm nuclei. Eventually the
polar body disappears and its chromatin disintegrates and is resorbed
into the general deutoplasmic mass.

After the nuclei have taken up their peripheral position and undergone
three mitoses, they incorporate some of the egg cytoplasm immediately
beneath them together with the cortical cytoplasm into a deep layer
which will become cellular material. Shortly afterward, cytoplasmic

PROBLEMS OF ORGANIZATION 203

cleavage begins as an infurrowing from the cortical layer. This separates
the blastoderm nuclei and incorporates each plus cortical material and
material which came from the underlying stratum of cytoplasm into in-
dividual cells. The peripheral cells undergo a growth by accretion or by
enlargement of their mass thrbugh the incorporation into them of ma-
terial which formerly lay in the cytoplasm of the central syncytium.

Upon completion of the blastoderm the first movements are those of
cells along the ventral mid-line and in the regions which later become the
cephalic furrow. They are slipping movements and appear to be conse-
quences of changes in cell shape and polarity. These movements lead to
the inturning of the mid-ventral cells to form a mesodermal tube and to
the establishment of the proctodaeal plate beneath the pole cells. These
movements are neither immediately preceded nor accompanied by mi-
totic activity. Anteriorly the mesodermal tube extends beyond the
cephalic furrow and terminates as a T-like structure just posterior to
the region which will produce the stomodaeal plate. At its posterior end
the mesodermal tube is for a short time confluent with the depression
which gradually deepens in the proctodaeal plate to receive the pole cells.
The pole cells subsequently migrate through the proctodaeum and sep-
arate into two groups which become the gonads. This period of special
movements is followed by one of general movement in which the newly
established germ band elongates and extends around the posterior pole
of the egg onto the dorsal side, while the dorsal and lateral regions of
the blastoderm fold and stretch in making way for the elongating germ
band. During this period the proctodaeal plate invaginates deeply and
mitoses are found among its cells. Simultaneously at the anterior end
certain cells at the level of the stomodaeal plate migrate beneath the
ectoderm and undergo mitosis to form a group of cells which become the
anterior rudiment of the mid-gut (endoderm). The origin of the pos-
terior rudiment of mid-gut is somewhat obscure, but probably of a simi-
lar nature, as there is no evidence of its formation from the mesodermal
tube. With the invagination of the stomodaeal plate the anterior rudi-
ment of mid-gut begins to extend by mitosis and cell movement pos-
teriorly and ventrally between the mesoderm (which has now become
a flattened layer of cells with no remnant of the original lumen) and the
yolk. A similar growth and extensions of the posterior rudiment lead
eventually to their fusion to complete the mid-gut.

A period of special cellular movement begins during the invagination
of the stomodaeal plate, when neuroblasts begin to be differentiated and
move from the ectoderm lateral to the ventral mid-line into the interior

204 J. S. NICHOLAS

against the mesoderm. The neuroblasts contain a large amount of cyto-
plasm and can be recognized as constituents which are distributed along
the entire length of the germ band. The neuroblasts immediately un-
dergo the characteristic unequal divisions which produce the ganglion
cells. The entire nervous system is originally bipartite, a condition re-
flected in its later bilateral structure. The differentiating ganglion cells
send out processes from the nerve cord into the mesoderm and these ex-
tend along with the mesoderm into the lateral areas, becoming the seg-
mental nerves.

Lateral ventral cells which remain at the surface after the neuroblasts
have moved to the interior become hypoderm cells and later produce the
larval integument, except in certain localized regions where they give
rise to salivary plates (later invaginated and pulled into the interior
through the stomodaeum) and the eleven pairs of tracheal pits which
give rise to the major part of the tracheal system.

By the middle period of development ( 10-12 hrs.) the principal larval
structures, or their rudiments, are all laid down. The shortening of the
germ band brings many cells back to somewhere near their original po-
sitions, but there are new shifts as the dorsal closure of the embryo is
completed and segmentation proceeds. Special movements are concerned
in the involution of the head region, which is characteristic of the larvae
of higher Diptera. The imaginal discs begin to make their appearance
only a short time before the hatching of the larva from the egg (at 20-
24 hrs. after fertilization, depending on the temperature).

The movements within the embryo of Drosophila are more complex
than in the amphibia where they are limited to two general types of
movement. The first of these is found in the egg just after fertilization,
continuing through the segmentation stages and blastulation, the second
comprises the later general movements, resulting in the formation and
distribution of the layers and the formation of the nervous system. Spe-
cific movements of the type performed in Drosophila seldom occur. In
Drosophila the specific movements apparently are those of aggregates
of cells which tend to adhere to each other and form functional masses
from which are derived the specialized tissues and organs of the embryo.
An amphibian would produce its nervous system if it were similar to a
Drosophila by the creation of more parts from which certain areas
would be deleted. Drosophila, however, has its parts fully formed, its
ganglion cells sending their processes out to the periphery, and then a
contraction or aggregation of the different parts of the system occurs,
with a degeneration of some of its parts to form a smaller unit. This

PROBLEMS OF ORGANIZATION 205

method is just as efficient in its complex potentiality as is that of the
amphibian.

There are four types of movement that one may designate as occur-
ring in the Drosophila egg. The first are general movements involving
nuclei and their peripheral distribution; the second, specific movements
of cells and of layers; the third, specific movements of cell aggregates
which later will form organs within the organism ; and finally the gen-
eral movements involved in the shortening of the embryo, which con-
tinue within the nervous system and result in the high degree of con-
densation characteristic of that system in the higher Diptera. The
condensation of the nervous system continues after the germinal band
has begun to proliferate upward along the sides to surround the intes-
tinal canal, and nerve fibers accompany the migrations of the mesoderm.
In this it is not unlike the vertebrate system. It differs from it, however,
in that (during the process) it sends proportionately very much longer
nerve fibers in connection with the contraction of the central system it-
self. The localization of the various parts and their relationships to
future organization are clearly shown in Poulson's ('45) rather re-
markable investigation of this form.

The movements described above may be reduced to a rather simplified
form in the following classification.

Classification of Movements in Drosophila

A. General movements (early egg)

1. Nuclear

a. early distribution

b. peripheral migration

2. Cytoplasmic

a. early (with nuclei)

b. disengagement from yolk

c. furrowing of peripheral cytoplasm

d. sync>1:ial layer to cellular layer

B. Special movements

1. Cellular migrations

a. endoderm in gastrulation

b. pole cells

c. neuroblasts

2. Aggregate movements

a. mesoderm in gastrulation

b. proctodaeum

206 J. S. NICHOLAS

c. stomodaeum

d. mid-gut

e. nervous system

C. General embryonic (germ band and later)

1. Surface movements

a. germ band extension

b. germ band shortening

c. dorsal closure

2. Invagination

a. involution of head

The story of the movements of parts during organization is not nearly
so complete in the amphibian (Gillette ('44)), Here a similar classifi-
cation scheme might be evolved, but there are tremendous gaps in the
analysis of the various types of movement which occur. There are many
observations which indicate that movements occur, but whether these
are molecular, particulate, cellular, or aggregate is in general not noted,
and accurate measurements of the amount of movement and the forces
necessary for the final position are not yet available. Gillette ('44) has
investigated the neurula and has given the most concrete example of
mass movements of cell aggregates in the formation of the neural plate.
There is a dorsal concentration of cells which formerly occupied a lateral
position. At the same time there is a smaller but definite relocation of
some of the ventral cells which move into the lateral position. There is
a loss of 1 1 % of the total ectoderm from the lateral lunes and 7% from
the ventral lunes in providing the ectodermal material which is taken
over in the dorsal region.

Chemical Organization

The work dealing with the chemistry of the neural organizing sub-
stance, or substances, beginning with Spemann, Fischer, and Wehmeier
('33) and continuing to the present (for reviews see Needham ('42)
and Boell ('42)) has shown how many different substances may cause
this effect. No particular one, however, can be said to be dominant in
the reaction. The analyses for glycogen in different regions by Woerde-
man ('33), Heatley ('35) and Heatley and Lindahl ('37) showed that
glycogen disappears in the dorsal lip during gastrulation. This fact has
been employed by Boell ('42) as showing the significance of the reac-
tions in association with the dorsal lip and carbohydrate metabolism.
Boell (loccit.) implies that the quantitative differences in metabolism

PROBLEMS OF ORGANIZATION

70
0'

207

I ANT ! MID • POST I
[THIRD THIRD" TH I R D

Z
oc

Ul

o
o

h

o

ai

_i
<

O

PORSALUJNES

u
o

a:

Q.

13 14 15 16

HARRISON'S STAGES

Figure 3. The left hand figure shows the (A) typical cross section of the neurula
with the ectodermal subdivisions marked, the (B) is a diagram of the left side showing
the subdivisions of (A) reconstructed as lunes on the lateral surface. The right hand
figure shows Gillette's graphs of the cellular redistribution of the ectodermal cells during
neural plate formation.

in the various parts of the gastrula to be of secondary importance in the
inductive action.

Brachet ('38) in cyto- and histochemical studies localized the SH
groups in various stages of development. With the exception of the two
cell stage, the SH groups are predominantly situated in the animal re-
gion of the embryo and during gastrulation are strongly centralized
around the dorsal lip, after which there is a heavy localization of these
substances in the nervous system. In his 1944 presentation of the vari-
ous lines of chemical investigation, Brachet shows that the content of
ribonucleic acid decreases in the dorsal lip during its invagination (cf.
Fig. 94 in Brachet ('44) ). There is a coordinate change in the distribu-
tion of the ribonucleoproteins and a marked shift in their localization
during neurulation. This is coincident with the pattern as shown in his
Figure 92.

208 J. S. NICHOLAS

On the basis of his own work Brachet ('44) concludes that the nu-
cleus is the center of protein synthesis, in contrast to Caspersson ('41),
who concludes that the nucleolus is the initiator of the reaction and that
the proteins are synthesized in the cytoplasm while the nucleolus acts as
the center for histones and the nucleus for the ribonucleic acid. Brachet's
studies on the roles of thymonucleic, ribonucleic, and desoxyribonucleic
acids in cell metabolism tend to strengthen his viewpoint concerning the
character and action of these specific substances.

While there have been numerous studies of oxidation rate in the de-
veloping embryo, there has been little correlation of the facts so derived
with the organizational picture. Boell's findings indicate that carbo-
hydrate metabolism rather than respiration is the major significant
factor. There have, however, been definite correlations between the de-
velopment of definite chemical entities with both morphological and
functional development, and a few examples of this phase of the work
will be given. Sawyer ('43) has shown the rate and amount of develop-
ment of cholinesterase and has correlated this with the onset of the em-
bryonic nerve-muscle reaction.

The enzyme remains constant in quantity during the early stages, de-
riving its activity in causing nerve-muscle reaction by the concentration
of practically the total amount of cholinesterase in the nervous and myo-
tomic structures. The increase in content begins with stage 36, when the
behavior pattern demonstrates the S-flexure, after which the enzyme
grows in quantity such that it can be represented as a sigmoid curve up
to the feeling stage. By placing Amblystoma in solutions of prostigmine
and eserine, Sawyer was able to change the rate and amount of the
cholinesterase development. In a further study Sawyer ('43) has studied
the cholinesterase concentration in the nervous system and in the total
animal, showing the increase in concentration during the period of func-
tional organization.

The cholinesterase activity reaches its peak in Amblystoma about fifty
days after fertilization in both nervous system and muscle and then de-
clines in both, but more drastically in muscle than in nerve. The adult
activity is represented by the same concentration of cholinesterase that
is found in some of the pre-feeding stages. When the spinal cord of the
embryo is removed so that the muscles develop in denervated condition,
the curve of the development of cholinesterase is much lower in the
muscle than in the innervated muscle. There is, however, a residuum
amounting to 31 % of the normal in late swimming stages, showing that

PROBLEMS OF ORGANIZATION 2O9

the nervous system is not entirely responsible for the organization of this
functional unit.

The results of Sawyer ('43) have been checked by Boell and Shen
('44) in their study of the cholinesterase content of induced neural
structures.

The results of measuring the cholinesterase content of various tissues
in developing the embryos of Amblystoma punctatum indicate a signifi-
cant localization of cholinesterase in the nervous system in comparison
with the ectoderm at the period of development marked by the closing
of the neural folds. As differentiation of the nervous system proceeds,
the concentration of cholinesterase in neural tissue becomes progres-
sively greater.

The cholinesterase activity of secondary or induced neural structures
produced by the action of implanted chorda mesoderm upon competent
ectoderm is of the same order of magnitude as that of the primary nerv-
ous tissues of the host and is much larger than that of ectoderm which
has not received the stimulus of the inductor. It appears then that the
phenomena of induction, in addition to causing the development of dis-
tinct morphological changes in the induced tissue, stimulates also the
latter to develop the biochemical machinery which is characteristic of
normal nervous tissue. They suggest that cholinesterase content of sec-
ondarily induced neural structures can be regarded as the measure of
their potential or incipient functional differentiation.

Boell ('45), in his study of the correlation of respiration and the
presence of cytochrome oxidase, finds that the respiratory rate increases
steadily throughout development and varies exponentially with time,
although it seems impossible to correlate specific morphogenetic effects
with abrupt changes in velocity of oxygen uptake. The rate of respira-
tory increase is not constant and shows a distinct break at stage 30. The
evidence supports the conclusion that the rise in respiratory rate noted
during development is the result of growth of the embryo, i.e. the trans-
formation of the inert yolk into metabolically active material. The res-
piration of Amblystoma punctatum embryos is markedly susceptible to
cyanide throughout development, but is especially so in the later stages.

As in the case of respiration, the cytochrome activity of Amblystoma
punctatum embryos increases steadily throughout the entire period of
development and apparently varies directly with increase in metabolically
active embryonic mass. Various lines of evidence suggest that respira-
tion is largely mediated by cytochrome oxidase throughout develop-
ment, but the absolute rate of oxygen uptake at any given stage is lim-

2IO J, S. NICHOLAS

ited by some factor other than the concentration of cytochrome oxidase.
By the application of the heterauxetic equations the relative growth rates
of cytochrome oxidase and cholinesterase have been determined and the
individual character of these processes of biochemical differentiation
shown.

The evidence in the work of Sawyer, Boell, and Shen shows clearly
the place which the development of the enzyme plays in the functional
organizational development. The fact that the enzymes have different
developmental rates indicates the significance which they have in the
general growth properties of the embryo and, what is still more im-
portant, how they are correlated with the specific parts of differentiation.
The interesting problems of the changes in the substrates from which
these materials are formed still await solution and will undoubtedly
form an important phase of the understanding of organization.

Regeneration

The problems of organization are not limited to those which can be
attacked with embryonic material, for in every regenerating structure
there is a new organization involving both growth and differentiation in
the attainment of the new structure. It is impossible here to cover the
wealth of vertebrate and invertebrate material which yields information
on various phases of the problem of morphogenesis during regeneration.
One example can be cited in order to show the changing attack in this
field, the relation of the nervous system to limb regeneration.

Schotte ('23, '26) on the basis of a great deal of work in the am-
phibian nerves came to the conclusion that some elements of the nervous
system were necessary in order that regeneration should be initiated and
completed. This work has been continued (Schotte and Butler ('41),
Butler and Schotte ('41), and Schotte and Butler ('44)). They have
studied the various phases of the formation of the blastema and the way
in which each is affected by the presence of nervous system. The nerves
seem to be necessary for the dedifferentiation phase leading to blastema
formation and also in the transformation of the blastema into a regen-
erate with morphogenetic determination. After morphogenetic deter-
mination is established, the presence of nerves is no longer necessary for
subsequent growth and differentiation.

The research referred to above seems to answer one of the problems
of regenerative organization in that it solves the particular role of nerv-
ous tissue in blastema organization. That there are other factors still

PROBLEMS OF ORGANIZATION 211

to be analyzed is shown in Rose's ('44) study of anuran forms, where
the cuticle forms a block to the regenerative process. By treating with
NaCl the stumps of amputated limbs in fully developed frogs or in frog
tadpoles after the normal time of regeneration, Rose was able to initiate
limb development which was sometimes complete, and at other times
with deficiencies in the hand.

Singer ('42, '43, '46a, '46b, '47a, and '47b) has analyzed the amount
of nerve tissue necessarily present in order to initiate regeneration. He
finds that the process is non-specific and that the number of nerve fibers,
whether autonomic, sensory, or motor, controls the reaction. In this way
the regenerative process of urodeles is reduced to a quantitative expres-
sion of nerve fiber relationship and our concept of the significance of
nerves to regeneration is clarified.

In the above pages a few of the problems of organization have been
presented. They constitute but a sample of the vast array which have at
one time or another been investigated and for which at present there is
no final answer. The work here given, however, shows the present attack
on problems many of which were formulated long ago. It must be ad-
mitted that their final solution has not been attained, but we have reason
for optimism, for science works slowly but surely toward the solution
of any major problem, such as embryonic organization has proved to be.

REFERENCES

Baltzer, F. Uber erbliche letale Entwicklung und Austauschbarkeit artver-
schiedener Kerne bei Bastarden. Naturiviss, 28, 177-187, 1940.

Barth, L. G. Neural differentiation without organizer. /. Exp. Zool., 8y,
371-383, 1941.

Boell, E. J. Biochemical and physiological analysis of organizer action.
Growth, Vol. 7 (Suppl.), 37-53, 1942.

Boell, E. J. Functional differentiation in embryonic development. II. Respi-
ration and cytochrome oxidase activity in Amblystoma punctatum.
/. Exp. Zool, IQO, 331-352, 1945.

Boell, E. J., and S. C. Shen. Functional differentiation in embryonic devel-
opment : I. Cholinesterase activity of induced neural structures in
Amblystoma punctatum. /. Exp. Zool., 97, 21-41, 1944.

Boveri, Th. Zellenstudien VI. Die Entwicklung dispermer Seeigeleier. Ein
Beitrag zur Befruchtungslehre und zur Theorie des Kerns. Jen. Zeit-
schr. f. Naturiv., 4^, 1-292, 1907'.

Boycott, A. E., and C. Diver. On the inheritance of sinistrality in Limnaea
peregra. Proc. Roy. Soc. London, (B) p^, 207-213, 1923.

212 J. S. NICHOLAS

Brachet, J. La localisation des proteines sulfhydrilees pendant le developpe-
ment des Amphibiens. Bull. Acad. Roy. Bclg., 1938:499-510, 1938.

Brachet, J. Emhryologie Chimique. Paris, Masson et Cie., 505 pp., 1944.

Butler, E. G., and O. E. Schotte. Histological alterations in denervated non-
regenerating limbs of urodele larvae. /. Exp. ZooL, 88, 307-341, 1944.

Caspersson, T., H. Landstrom-Hyden, and L. Aquilonius. Cytoplasma
nucleotide in eiweissproduzierenden Driisenzellen. Chromosoma, 2,
111-131, 1941.

Child, C. M. Organizers in development and the organizer concept. Physiol.
ZooL, ip, 89-148, 1946.

Conklin, E. G. The development of centrifuged eggs of ascidians. /. Exp.
ZooL, 60, 1-119, 1931.

Dalcq, A. M. Form and Causality in Early Development. Cambridge Uni-
versity Press, 1938.

Daniel, J. F., and E. A. Yarwood. The early embryology of Triturus toro-
sus. Univ. of Calif. Piihl. ZooL, 43, 321-356, 1939.

Diver, C, A. E. Boycott, and S. Garstang. Inheritance of Inverse Sym-
metry. /. Genetics, 75, 113-200, 1925.

Eakin, R. M. Regional determination in the development of the trout. Arcli.
f. Entw-mech., /jp, 274-281, 1939.

Fankhauser, G. Cytological studies on egg fragments of the salamander
Triton. IV. The cleavage of tgg fragments without the egg nucleus.
/. Exp. ZooL, 67, 349-394, 1934-

Gillette, R. Cell number and cell size in the ectoderm during neurulation
(Amblystoma maculatum). /. Exp. ZooL, pd, 201-222, 1944.

Hadorn, E. Ubertragung von Artmerkmalen durch das entkernte Eiplasma
beim merogonischen Triton-Bastard palmatus-Plasma X cristatus-
Kern. Verh. Dtsch. ZooL Ges., 1936, 97-104, 1936.

Hall, E. K. Regional differences in the action of the organization center.
Arch. f. Entw-mech., 135, 671-688, 1936.

Harrison, R. G. The effect of reversing the mediolateral or transverse axis
of the forelimb bud in the Salamander embryo (Amblystoma punc-
tatum Linn.) Arch. f. Entw-mech., 106, 469-502, 1925.

Harrison, R. G. Cellular differentiation and internal environment. A.A.A.S.,
piibl. no. 14, 77-97, 1940.

Harrison, R. G. Relations of symmetry in the developing embryo. Trans.
Conn. Acad. Arts and Sciences, 36, 277-330, 1945-

Harvey, E. B. Parthenogenetic merogony or cleavage without nuclei in
Arbacia punctulata. Biol. Bull., 77, 101-121, 1936.

Harvey, E. B. Structure and development of the clear quarter of the Ar-
bacia punctulata egg. J. Exp. ZooL, 102, 253-275, 1946.

Heatley, N. G. The distribution of glycogen in the regions of the Amphib-
ian gastrula ; with a method for the microdetermination of glycogen.
Bioch. Jour., 2g, 2568-2572, 1935.

Heatley, N. G., and P. E. Lindahl. Studies on the nature of the amphibian
organization centre. V. The distribution and nature of glycogen in the
amphibian embryo, Proc. Roy. Soc. London, (B) 122, 395-402, 1937.

PROBLEMS OF ORGANIZATION 213

His, W. Unsere K or perform nnd des physiologische Problem Hirer Enste-

hung. Leipzig, F. C. W. Vogel, 1874.
Holtfreter, J. Gewebeaffinitat, ein Mittel der embryonalen Formbildung.

Arch, ex per. Zellforschg., -?j, 169-209, 1939.
Holtfreter, J, A study of the mechanics of gastrulation, part I. /. Exp.

Zool., p4, 261-318, 1943a.
Holtfreter, J. Properties and functions of the surface coat in amphibian

embryos. /. Exp. Zool., 93, 251-323, 1943b.
Holtfreter, J. Neural differentiation of ectoderm through exposure to

saline solution. /. Exp. Zool., pj, 307-343, 1944a.
Holtfreter, J. A study of the mechanics of gastrulation, part H. /. Exp.

Zool., p5, 1 71 -212, 1944b.
Lewis, W. H., and E. C. Roosen-Runge. The formation of the blastodisc in

the egg of the zebra fish, Brachydanio rerio, illustrated with motion

pictures. Anat. Rec, 84, 13-14, 1942.
Lillie, F. R. Embryonic segregation and its role in the life history. Arch.

f. Entzv-mech., 118, 499-533, 1929.
Luther, W. Potenzpriifungen an isolierten Teilstiicken der Forellenkeim-

scheibe. Arch. f. Entw-mech., /J5, 359-383, 1936.
Luther, W. Transplantations und Defektversuche am Organisationszentrum

der Forellenkeimscheibe. Arch. f. Entzv-mech., /j/, 404-424, 1937.
Morgan, T. H. Experimental studies on the teleost eggs. Anat. Ans., 8,

803-814, 1893.
Needham, J. Biochemistry and Morphogenesis. Cambridge University

Press, 1942.
Nicholas, J. S. Blastulation, its role in pregastrular organization in Am-

blystoma punctatum. /. Exp. Zool., 100, 265-299, 1945.
Nieuwkoop, P. D. Experimental investigations on the origin and determi-
nation of the germ cells and on the development of the lateral plates

and germ ridges in Urodeles. Archives Neerlandaises de Zoologie., 8,

1-205, 1946.
Oppenheimer, J. Experiments on early developing stages of Fundulus.

Proc. Natl. Acad. Sci. U.S., 20, 536-538, 1934.
Oppenheimer, J. Processes of localization in developing Fundulus. Proc.

Natl. Acad. Sci. U.S., 21, 551-553, 1935.
Oppenheimer, J. The development of isolated blastoderms of Fundulus

heteroclitus. /. Exp. Zool., 72, 247-269, 1936a.
Oppenheimer, J. Processes of localization in developing Fundulus. /. Exp.

Zool, 73, 405-444. 1936b.
Pasteels, J. La gastrulation et la repartition des territoires dans la moitie

dorsale du blastodisque de Truite (Salmo iridaeus). C. R. Soc. Biol.,

113, 425-428, 1933.
Pasteels, J. Repartition des territoires et mouvements morphogenetiques

de la gastrulation de I'oeuf de truite (Salmo iridaeus). C. R. Ass. Anat.

Bruxelles, 1934, 451-458, 1934.
Pasteels, J. Etudes sur la gastrulation des vertebres meroblastiques. HL

Oiseaux. Arch, de Biol., 48, 382-463, 1936.

214 J- S. NICHOLAS

Poulson, D. Chromosomal control of embryogenesis in Drosophila. Am.

Nat., 6p, 340-363, 1945-
Robertson, G. G. An analysis of the development of homozygous yellow

mouse embryos. /. Exp. ZooL, 89, 197-231, 1942,
Rose, S. M. Methods of initiating limb regeneration in adult Anura. /. Exp.

Zool, 95, 149-170, 1944.
Rudnick, D. Early history and mechanics of the chick blastoderm. Quart.

Rev. Biol., ip, 187-212, 1944.
Sawyer, C. H. Cholinesterase and the behavior problem in Amblystoma.

I. The relationship between the development of the enzyme and early
motility. /. Exp. Zool., Q2, i-ii, 1943a.

Sawyer, C. H. Cholinesterase and the behavior problem in Amblystoma.

II. The effects of inhibiting cholinesterase. /. Exp. Zool., p2, 11-29,
1943b.

Sawyer, C. H. Cholinesterase and the behavior problem in Amblystoma.

III. The distribution of cholinesterase in nerve and muscle throughout
development. /. Exp. Zool., p4, 1-12, 1943c.

Sawyer, C. H. Cholinesterase and the behavior problem in Amblystoma.

IV. Cholinesterase in nerveless muscle. /. Exp. Zool., p4, 12-31, 1943d.
Schechtman, A. M. Unipolar ingression in Triturus torosus. Univ. of Calif.

Puhl. Zool, 39, 303-307. 1934-
Schotte, O. E. Influence de la section tardive des nerfs sur les pattes de

Tritons en regeneration. C. R. Soc. Phys. et Hist. fiat. Geneve., 40,

86-88, 1923.
Schotte, O. E. Systeme nerveux et regeneration chez le Triton. Rev. Suisse

de Zool., ss, 1-211, 1926.
Schotte, O. E., and E. G. Butler. Morphological effects of denervation and

amputation of limbs in urodele larvae. /. Exp. Zool., 8"/, 279-322, 1941.
Schotte, O. E., and E. G. Butler. Phases in regeneration of the urodele

limb and their dependence upon the nervous system. /. Exp. Zool., pj,

95-121, 1944.
Schr5dinger, E. What is Life? The physical aspect of the living cell. Cam-
bridge University Press, 1945.
Singer, M. The nervous system and regeneration of the forehmb of adult

Triturus. I. The role of the sympathetics. /. Exp. Zool., po, 377-399,

1942.
Singer, M. The nervous system and regeneration of the forelimb of adult

Triturus. II. The role of the sensory supply. /. Exp. Zool., p2, 297-

315. 1943-
Singer, M. The nervous system and regeneration of the forelimb of adult

Triturus. III. The role of the motor supply, including a note on the

anatomy of the brachial spinal nerve roots. /. Exp. Zool., 98, 1-21,

1945-
Singer, M. The nervous system and regeneration of the forelimb of adult

Triturus. IV. The stimulating action of a regenerated motor supply.

/. Exp. Zool., loi, 221-240, 1946a.

Singer, M. The nervous system and regeneration of the forelimb of adult

PROBLEMS OF ORGANIZATION 215

Triturus. V. The influence of number of nerve fibers, including a
quantitative study of limb innervation. /. Exp. ZooL, loi, 299-331,
1946b.

Singer, M. The nervous system and regeneration of the forelimb of adult
Triturus. VI. A further study of the importance of nerve number, in-
cluding quanitative measurements of limb innervation. /, Exp. ZooL,
104, 223-250, 1947a.

Singer, M. The nervous system and regeneration of the forelimb of adult
Triturus. VII. The relation between number of nerve fibers and sur-
face area of amputation. /, Exp. ZooL, 104, 251-265, 1947b.

Spemann, H., F. G. Fischer, and E. Wehmeier. Zur Kenntnis der Induk-
tionsmittel in der Embryonalentwicklung. Naturwiss., 21, 518, 1933.

Spratt, N. T., Jr. Formation of the primitive streak in the explanted chick
blastoderm marked with carbon particles. /. Exp. ZooL, 10^, 259-304,
1946.

Stableford, L. The potency of the isolated vegetal hemisphere (presump-
tive endoderm) of the blastula of Amblystoma punctatum. Anat. Rec,
75, abs. 35-36, 1939.

Tung, T. C, C. Y. Chang, and Y. F. Y. Tung. Experiments on the devel-
opmental potencies of blastoderms and fragments of teleostean eggs
separated latitudinally. Proc. ZooL Soc. Lond., 11^, 175-188, 1945.

Vogt, W. Gestaltungsanalyse am Amphibienkeim mit ortlicher Vitalfar-
bung. Vorwort iiber Wege und Ziele. I. Teil : Methodik und Wirkungs-
weise der ortlichen Vitalfarbung mit Agar als Farbtrager. Arch. f.
Entzv-mech., 106, 542-610, 1925.

Vogt, W. Gestaltungsanalyse am Amphibienkeim mit ortlicher Vitalfarb-
ung. II. Teil. Gastrulation und Mesodermbildung bei Urodelen und
Anuren. Arch. f. Entw-mech., 120, 384-706, 1929.

Woerdeman, M. W. tJber den Glykogenstofifwechsel des Organisations-
zentrums in der Amphibiengastrula. Proc. Acad. Sci. Amsterdam, 36,
189-193, 1933.

Yamamoto, T. Studies on the rhythmical movements of the early embryos
of Oryzias latipes. I. General description. /. Fac. Sci, Imp. Univ.
Tokyo, 2, 147-152, 1931a.

Yamamoto, T. Studies on the rhythmical movements of the early embryos
of Oryzias latipes. II. Relation between temperature and the frequency
of the rhythmical contractions. /. Fac. Sci, Imp. Univ. Tokyo, 2, 153-
162, 1931b.

Yamamoto, T. Studies on the rhythmical movements of the early embryos
of Oryzias latipes. III. Temperature and the amplitude of the contrac-
tion waves. /. Fac. Sci, Imp. Univ. Tokyo, 5, 105-110, 1933a,

Yamamoto, T. Studies on the rhythmical movements of the early embryos
of Oryzias latipes. IV. Temperature constants for the velocity of the
wave and for the pause. /. Fac. Sci, Imp. Univ. Tokyo, j, 111-117,
1933b.

Yamamoto, T. Studies on the rhythmical movements of the early embryos

2l6 J. S. NICHOLAS

of Oryzias latipes. V. The action of electrolytes and osmotic pressure.

/. Fac. Sci., Imp. Univ. Tokyo, j, 287-299, 1934.
Yamamoto, T. Studies on the rhythmical movements of the early embryos

of Oryzias latipes. VI. The action of hydrogen ion concentration. /.

Fac. Sci., Imp. Univ. Tokyo, 4, 221-232, 1936a.
Yamamoto, T. Studies on the rhythmical movements of the early embryos

of Oryzias latipes. VII. Anaerobic movements and oxidation-reduc-
tion potential of the egg limiting the rhythmical movements. /. Fac.

Sci., Imp. Univ. Tokyo, 4, 233-247, 1936b.
Yamamoto, T. Studies on the rhythmical movements of the early embryos

of Oryzias latipes. VIII. The effect of series of carbamates. /. Fac.

Sci., Imp. Univ. Tokyo, 5, 37-49, 1938a.
Yamamoto, T. Contractile movement of the egg of a bony fish Salanx

microdon. Proc. Imp. Acad. Tokyo, 14, 149-151, 1938b.
Yamamoto, T. On the distribution of temperature constants in Oryzias

latipes. Proc. Imp. Acad. Tokyo, 14, 393-395, 1938c.

IX. NEOPLASTIC ABNORMAL GROWTH

BY C. P. RHOADS^

^ BNORMAL growth IS SO broad a subject that the Hmits within
/\ which we are to use the term here must be precisely defined
X .A- before it is discussed. The purpose of the consideration here-
with presented is to explore the stockpile of possibly unrelated factual
material of a fundamental biological nature for ideas and principles
which may make more efifective the investigative attack on neoplastic
disease. This kind of abnormal growth is one which has no function
in, or integration with, any orderly cellular structure and, more than
this autonomy, it has a positive, malignant capacity to injure normal
tissue. It is analogous to the predatory animal in the domesticated herd.

Unique Characteristics of the Cancer Cell

Extensive studies in progress are designed to provide new informa-
tion about the nature of the cancer cell, a most important problem in the
field of abnormal growth. Though repetitious, it is relevant to state once
more that this cell is characterized by unique properties and is, sup-
posedly, the repository of secrets which, if known, would make simple
the solution of the whole great problem of neoplastic disease. Many in-
vestigators are dissatisfied with their limited knowledge of the mecha-
nism which provides such a destructive and uncontrolled structure with
its power, and some wish particularly to understand in greater detail
how the constituent cells dififer in their constitution and function from
their normal precursors. Not only is this question of fundamental inter-
est, but also, were adequate information concerning it at hand, a direct
therapeutic attack on neoplastic disease conceivably could be developed.

A histological preparation of cancer tissue in contrast to its back-
ground of normal cellular structure provides ample evidence to establish
the unusual capacities of the cancer cell. The very ability of the patholo-
gist to diagnose a neoplasm from the forms and staining reactions of its
components is proof enough ( i ) that they are in general strikingly ab-
normal. The cells are not only distinctive in form, but also show unique
competitive powers. The morphology of a gastric wall infiltrated with
cancer, or of a cervix uteri similarly involved, demonstrates adequately

1 Director, The Memorial Hospital for the Treatment of Cancer and Allied Diseases,
and the Sloan-Kettering Institute for Cancer Research. Professor of Pathology, Cornell
University Medical College, New York.

2l8 C. p. RHOADS

that the neoplastic cell possesses special privilege and a profound bio-
logical advantage over its normal fellows. It apparently can carry out
unusual reactions as shown by its abilities to invade, distort, and destroy
normal tissue, to survive in the midst of violent inflammatory reactions,
and to withstand invasion by microorganisms. It goes its own way — it
has the property of autonomy and in many instances something more.

A simple quantitative increase in the rate of growth of cells, associ-
ated with a decrease in their ability to differentiate, sometimes has been
advanced as a generalization sufficient to explain the unique properties
of neoplastic tissue. It is desirable to establish at the outset how sound
this view is, to decide whether it is sufficiently inclusive to comply with
all of the facts, and to ascertain to what extent it can be used as a basis
for planning new experiments, more conclusive than the ones now at
hand.

The fact that some cancer cells multiply rapidly is proved by measure-
ment of their metabolic processes. Rapid growth is not a distinctive
quality, however, since certain normal cells, devoid of autonomy, multi-
ply even more rapidly, as shown by Brues (2). The evidence for the
limited differentiation of cancer cells is adequate, but also, like rapid
growth, not uniformly present. In some instances, certainly, little evi-
dence for the existence of any normal function can be detected, and in
essentially all, the cellular activities seem to be less specialized than are
those of normal cells of similar types. In the vast majority of neoplasms,
however, some evidence of differentiated form and function is at hand
and it may be very impressive. Hyperthyroidism due to functioning
thyroid carcinoma in the absence of the thyroid gland (3) and the pro-
duction of hormones by tumors of other endocrine organs provide satis-
factory proof that highly specialized activities may continue even though
the cells responsible for them clearly possess wholly abnormal malignant
qualities of invasiveness and destructiveness. Clearly, something more
than a quantitative change in the rates of growth and differentiation is
required to convey the property of neoplastic growth.

The analogy is frequently drawn between cancer and embryonic cells
on the grounds that both grow rapidly and are little differentiated. The
embryonic cell certainly is not a neoplastic structure, however, nor is it
likely to become one without very considerable alteration. It has no au-
tonomy and lives dependent upon its associates. The cancer cell is far
more competent, aggressive, and malignant, and less susceptible to nor-
mal control and to orderly differentiation than is its normal embryonic

NEOPLASTIC ABNORMAL GROWTH 219

progenitor. These qualities must depend upon some profound deviation
from the normal structure or function, or both.

The distinctive characteristics of neoplastic grow^th are as striking as
those of the cells of the mutant bacterial colony described by Haddow
(4) : "The generation of a vigorously growing strain of bacterial cells
in a senescent colony presents a striking resemblance to the emergence
of a malignant variant in the cells of aging animal tissues." The com-
parison is supported by many examples of adaptive variations in bac-
teria. Here the cells of secondary colonies have been shown functionally
and structurally to be unlike the parents from which they arose, due to
the fact that they can metabolize constituents of the medium which the
progenitors are unable to attack at all. For lack of this competence the
parent cells, even though young and actively reproducing, are handi-
capped in their ability to carry on, with the substrate available, the meta-
bolic activities required for active proliferation. The variant, analogous
to the malignant cell, more versatile and perhaps better equipped bio-
chemically, can thrive under circumstances which are intolerable to its
less adaptable parent.

The Control of Cellular Growth
AND Differentiation

The role in cancer of a loss of the normal control of growth and dif-
ferentiation should be considered, whether or not we accept as correct
the conclusion that the constituent cell must involve something more
than a simple quantitative deviation of these factors from normal. There
are two possibilities : a wild growth, like that of cancer, might be due to
a local lack or a modification of some general control of development
and function; or, more possibly, the general force could remain un-
altered but be powerless to hold in restraint cells which have assumed
properties different from those of normal cells' properties through which
control is exerted under ordinary circumstances. Clearly, if a generally
applicable force has been modified to cause or to allow cancer to develop,
the modification must be one which afYects particularly a single specific
group of cells. It seems more probable that the cancer tissue has some
marked structural or functional difference from the normal, one which
should be detectable if searched for adequately. It is of the utmost im-
portance that any such difference be sought out and defined with the
greatest precision possible.

Certain controls of differentiation exist which, if disordered, may

220 C. P. RHOADS

concern the variation which is characteristic of cancer. One example is
found in the fact that when tissues composed of undifferentiated cells
come into contact, they unite. This is seen in wound healing, during em-
bryological development, and with neoplastic tissue. The temporary
union of undifferentiated cells and their subsequent permanent junction
is an important architectural process in embryology (5). The terminal
part of the floor of the penile urethra, for example, is formed by the
union of the lateral walls of the urethral groove, and in the normal de-
velopment of the embryo is replaced by ordinary firm connective tissue.
Gonadal hormones can cause the two uniting epithelial edges to become
keratinized so that union is impossible and hypospadias results. The
squamous cancer of the uterus induced in mice by Allen and Gardner
(6) provides a notable example of wholly abnormal differentiation
caused by an extracellular agent. The whole vast field of cancer produc-
tion by hormones is relevant to this question and makes available ample
evidence of the susceptibility of cells to general controls. It would be
desirable to have additional data concerning the effects of the hormones
and the carcinogenic chemicals on embryological development. We lack
precise comparative observations of their effects in causing differentia-
tion and possibly dedifferentiation of similar material susceptible to pre-
cise study. Reference to Needham (7) is desirable on this point.

The future may well see experiments designed to test more exactly the
flexibility of function of cells at various stages during differentiation. It
is not impossible that the assumption of adult differentiated form in-
volves a constant succession of losses of genes, or controlling elements,
with consequent loss of ability to carry out a variety of functions coinci-
dent with concentration on a comparatively small number. This em-
phasis on certain limited functions could well give the distinctive charac-
teristics and abilities possessed by highly specialized tissue. The constant
tendency toward decrease in the number of functions capable of being
executed would, of course, lead to a progressive handicap to the cell in
surviving in a standard environment and in time to the very impossibility
of doing so. This progressive change would eventually mean the death
of the cell, and indeed, this is precisely the fate of differentiated tissue.
Is it too fanciful to believe that the endocrine glands are accessory or-
gans created to m.aintain, in their terminal, highly differentiated phases,
the metabolic activity of cells which are necessary for only a part of
life's span? Certainly when these glands are extirpated, specific highly
differentiated cellular systems are put at a great disadvantage and die,
or at least become inactive.

NEOPLASTIC ABNORMAL GROWTH 221

The fact is remarkable that, to date, in spite of the enormously ex-
tensive studies which have been made of neoplastic tissue and the obvi-
ous morphological contrasts between normal and cancer cells, no quali-
tative chemical differences have been established conclusively. A
considerable series of studies culminating in the recent work of Friede-
wald (8), of Kidd (9, 10, 11), and of Green (12, 13) provide certain
immunologic evidence for the presence of abnormal constituent proteins
in the cancer tissue of animals. Very great extension of the work is re-
quired, since in few instances described did the tumor studied arise in
animals homozygous with the ones in which antibodies were induced.
The measurements of the enzymatic activities of neoplastic tissue insti-
tuted by the Warburg school (14) and so effectively continued by the
investigators of the National Cancer Institute (15), the McArdle labo-
ratory (16), by Salter (17) and others have not revealed completely
uniform qualitative differences between normal and neoplastic enzyme
function. Can it be that new enzymes have not been found because of
the tendency of investigators to examine cancer tissue for the activity
of normal systems against normal substrates, rather than to search em-
pirically for new metabolic pathways ? Differences of structure and func-
tion must exist, however, between normal and cancer cells, even though
they have not been satisfactorily defined. Perhaps a new approach can
be developed from a consideration of the possible origins of differences
between cells.

Cfxl Characteristics as a Function of Genes

The induction of changes in cells is frequently susceptible to analysis
in accord with the recognized principles of morphogenesis, under which
differentiation and growth are governed by genes. Whereas this con-
trol of the patterns of living organisms is well established, it is not en-
tirely clear how tissues and organs become differentiated into forms
which are completely distinct one from the other, when supposedly all
the component cells, being derived by mitotic divisions from a single
fertilized egg, have identical sets of genes. If genes are the sole regu-
lating agents, when a cell changes as it does in differentiating, the altera-
tion should follow and be the result of a mutation of the genes. It is
likely, however, that there exists in the cytoplasm of some cells a second
group of controlling factors which are distinct from the genes of the
chromosome thread, but which may depend on the genes for their origin
and to some extent for their continued growth. Conceivably both genie
and extragenic regulatory mechanisms operate coincidcntally.

222 C. P. RHOADS

The concept of governing factors in the cytoplasm, though almost as
old as cytology, has recently been supported by new and limited but con-
vincing evidence reviewed by Haddow ( i8). It is a subject of great im-
portance to the cancer problem because it provides a principle of mor-
phogenesis which has been involved to explain the transmission of some
unusual but actively studied examples of neoplastic disease by cell-free
agents (viruses) in part derived from avian tumors. Some have assumed
that these agents must be cytoplasmic entirely, and require little con-
sideration of a genie origin, though adequate justification or even need
for this assumption is hard to establish.

Satisfactory data attest to the view, however, that a very considerable
and important part of the form and function of sexual cells, as well as
of vegetative forms, is controlled by genes without the participation of
any non-genic or cytoplasmic factor. Hence, since a normal adult cell
differs profoundly from its primitive undifferentiated precursor, the
genes of the two forms should differ also. In support of this possibility
is the fact that differentiated cells contrast sharply with their primitive
relatives in their susceptibility to injury by one agent, x-rays, known to
have a profound effect on genes (19). Similarly, since the form and
function of the cancer cell differ distinctly from those of the normal cell,
the genes of these two cell types also should be different. Morphological
grounds to support this view are at hand in a long series of studies, of
which the most recent are those of Biesele (20). Although obvious ab-
normalities of chromosomes are not present in every cell believed to be
neoplastic, they frequently can be demonstrated. Furthermore, it is clear
from the results of experiments to be discussed later (21) that genie
changes adequate to cause profound cellular alterations may exist with-
out any morphological evidence of their presence.

As a basis for the concept that the progressive steps in cellular dif-
ferentiation, normal or abnormal, can be controlled by genes, proof must
be advanced that genes can mutate spontaneously in a regular fashion,
since only in this way could they control the changes in structure and
activity associated with orderly development. Demerec (22) has pre-
sented adequate evidence that genes vary greatly in their stability. Some
genes of Drosophila, for example, mutate rarely or not at all, whereas
others, and these are numerous, particularly in plants, mutate hundreds
of times in the development of a single individual. Plough (2T)) has
shown that the spontaneous lethal mutation rate in two of the three large
chromosomes of Drosophila is in direct proportion to temperature and
that the mutations so induced may subsequently disappear. Further-

NEOPLASTIC ABNORMAL GROWTH 223

more, the existence of genes which affect in turn the spontaneous muta-
bility of other genes has been proven by Rhoades (24).

Once the fundamental instability of genes has been established, if we
are to invoke them as regulators of differentiation, it is necessary to
specify what changes of cells are involved in that process and to estab-
lish that they are alterations of constituents which are dependent upon
specific genes for their existence. If this can be done, and if we are satis-
fied that those particular genes exist in mammalian cells, then, pending
the submission of strong evidence concerning the regulatory function of
cytoplasmic factors, it is not unreasonable to consider the possibility
that orderly, genie controlled variation is adequate to provide a basis for
development, either orderly or disorderly. Beadle states : "Cells of iden-
tical genotypes travel many divergent paths in the process of differentia-
tion of the individual organism" (25). The different paths involve the
acquisition of structural and functional characteristics which concern,
particularly, the protein components of the cell. Satisfactory proof is
available that organs of special function are constructed of different and
characteristic proteins ; the composition of the skin, for example, is
demonstrably quite different from that of other tissues. Functional dif-
ferences between cells depend upon kinetically active protein enzymes,
and examples of their specificities in differentiated tissues are almost
numberless. With the special proteins and enzymes established as charac-
teristic of cellular differentiation, proof is required that the production
of these constituents is controlled by genes.

Ample demonstration that genes control the constitution of tissue
proteins is found in the antigenic specificities of tissues as compliant
with the principles of heredity. From Landsteiner's (26) contributions
concerning the blood groups of man, through the more recent studies of
Wiener (27) on the Rh factor, a mass of evidence is available. Particu-
larly striking is the fact that a i :i gene-to-antigen relationship has been
shown to exist in experimental material by Tatum (28), a fact which
suggests that the antigens are direct products of the genes, or even fur-
ther, that the antigen is "the copy of the gene." The thesis of the con-
trolled production of antigens by genes is further supported by immuno-
logical studies referred to later, in which genes apparently have been
altered by use of antisera specifically directed against the antigen, or
gene product.

Striking evidence that enzyme production is also under the control of
genes is to be found in the experiments of Beadle and his associates, well
summarized in a recent publication (25). Elaborate studies of Neuro-

224 C. P. RHOADS

spora have proven beyond any question that the enzymatic activities of
those cells and, by inference, cells of any type, are formed by genes.
Beadle states as follows : "Genes evidently determine the final potentiali-
ties of differentiation in a given organism. Whereas the initiation of the
orderly process may be through the environment, whether or not a given
reaction . . . can take place is gene controlled." An example is at hand in
the MacDowell (29) dwarf mice, in which the normal gene concerned is
responsible for the development of pituitary eosinophils. Without this
gene the pituitaries of the dwarf animals lack the eosinophilic com-
ponent which, if supplied artificially, restores growth. Because of this
lack the growth extent and rate of all organs is enormously reduced.

If the control of normal cellular development and differentiation by
genes is indicated by the evidence for genie regulation of protein and
enzyme production as well as for the existence of spontaneous varia-
tions in genes, and if a similar mechanism is to be invoked in the con-
trol of abnormal differentiation, as of the cancer cell, it becomes neces-
sary to establish the possibility of ahnornml mutations. Mammalian cells
are vegetative and hence not susceptible to genetic study by standard
methods. For this reason, the last step possible in a correlation of ab-
normal variation with cancer is to show that the agents which produce
cancer cells also cause mutations in sexual forms. Mutations can be
produced artificially by certain physical and chemical agents, and appar-
ently by proteins with the properties of antisera. The physical agents
include x-rays, gamma rays, neutrons, and ultraviolet light, and a mass
of evidence proves that all of these are carcinogenic.

The Artificial Production of Gene Changes
AND Cancer by Physical Agents

The studies of mutations as caused by x-rays are classic. Chromosome
aberrations are said to result from x-rays more frequently than they
occur spontaneously (30), but true mutations of the spontaneous types
are also caused. Many have been proven to be due to the removal of
small segments of chromosomes followed by junction of the broken
ends. Some cannot be due to such a chromosome deficiency, however,
and rather must represent an actual change in the composition of the
gene, since back mutations occur. The mutations appear to start as
changes in individual atoms, in consequence of which, by a chain of re-
actions, the mutation results. The question of whether gene mutation

NEOPLASTIC ABNORMAL GROWTH 225

can be regarded as minute rearrangements of a size too small to be seen
cytologically has been reviewed by Muller (19).

Glucksmann reports that under certain conditions the process of dif-
ferentiation of mammalian embryonic cells is promoted by irradiation
(31), and that their sensitivity to injury decreases as differentiation
proceeds. Can it be that the differentiated cell is no longer sensitive to
certain dosage of x-rays because many of the points on the chromosome
which are particularly vulnerable to this agent have previously altered
spontaneously? Adequate data are at present not at hand to prove that
x-ray effects are not additive and that they cause variations which are
resistant to the agent. Work of this type is in progress and the results
are said to indicate that resistance does develop.

The experiments of Beadle (25) prove completely that defects in cer-
tain biosyntheses can be produced in Neurospora by irradiation with
x-rays with resultant mutations of specific genes. The growth factor re-
quirements of the mutated organism result from the induced inability
to manufacture essential substances from simpler constituents. These
requirements are characteristic of a given strain and are transmitted to
each descendant cell, unless back mutation occurs. Gray and Tatum
(32), furthermore, proved that x-ray treatment of two species of bac-
teria also produced strains characterized by their inability to carry out
specific biochemical reactions and suggested that biosyntheses in bac-
teria, like those of Neurospora, are controlled by specific genes.

The evidence indicates that a single molecular change affects a single
gene of sexual cells. A similar effect in vegetative cells is suggested by
the fact that in the case of both viruses and bacteria a single atom
ionized can kill. Lea (33) has shown that this atom cannot be one lo-
cated just anywhere; it must be in a sensitive component, just as in cells
which have demonstrable genes. Gowen (34) compared the mutation
rates induced by x-rays in Drosophila, tobacco mosaic virus and the
bacteria causing corn wilt disease. The frequency of variation was
increased at the same rate in all three, a fact which again suggests a
similar structure controlling mutation in both sexual and vegetative
cells. These observations provide examples of the similarity between the
variations induced by the same means in sexual (mold) and in vegetative
(bacterial) cells. Since genes are involved in the former, it is reasonable
to assume the participation of genes or gene-like substances in the latter.
This conclusion is, of course, not susceptible to absolute proof, a point
conceded by the authors. Furthermore, it is not possible to prove whether

226 C. P. RHOADS

the mutation has occurred in hypothetical nuclear genes or in hypothet-
ical cytoplasmic determiners or plasmagenes.

It is entertaining to speculate on cellular differentiation as a progres-
sive series of losses of function, like those undergone by mutated Neuro-
spora, by which the cell retains the ability to exercise a certain number
of special functions while, through mutation, the abilities are lost to
carry out certain other activities to completion. The formation of pig-
ment, for example, or the accumulation of any metabolite, could be
simply the result of loss of ability to carry out completely a series of
metabolic steps.

The simple gene-hit theory of x-ray effects has been brought into
question by the experiments reported by Dale (35) on the changes in-
duced by x-rays of two enzymes in solution. These suggest that the
molecules of the solvent are activated by the absorption of radiation
and transfer their energy to the dissolved protein. The fact that enzymes
can be so inactivated suggests the existence of an intracellular effect of
radiation on enzymes somewhat more complex than is explainable by the
simple gene-hit theory. The work on Neurospora proves that a direct
relationship exists between genes and enzymes. Hence the Dale mech-
anism may simply prove the existence of an indirect route, for the x-ray
effect, through enzymes to genes. Lea, Smith, Holmes, and Markham
(36) have provided further evidence on this point by showing that virus
inactivation by radiation may be either direct, independent of concentra-
tion of the solution, or indirect in dilute solution.

The ability of x-rays to cause cancer has been amply established by
tragic experience among radiologists and requires little documentation.
Particularly detailed confirmation is found in the recent experimental
studies of Henshaw (37) on the production of leukemia in mice by
irradiation. Furth (38) has reported that granulosa cell tumors of the
ovary can be similarly caused in mice. A most interesting observation
on the neoplastic effects of x-rays on man is that of March (39), who
finds that the incidence of leukemia in radiologists is more than ten times
that among physicians who are not radiologists.

Ultraviolet light, also a cause of cancer, seems to be somewhat less
effective than are x-rays in the induction of mutation. The effects of the
two agents on maize were carefully compared by Stadler (40), who
found certain differences. An explanation may be at hand in the length
of the time interval between the break and the separation of the chromo-
some fragments when the mutation is due to ultraviolet. The striking
effect of certain wave lengths of the ultraviolet is in sharp contrast to

NEOPLASTIC ABNORMAL GROWTH 22/

the apparent absence of wave length effect in the x-ray range. Beadle
(25) states: "The curve of effectiveness of ultraviolet in producing
mutations against the wave length effect is sufficiently similar to the
absorption spectrum of nucleic acid over the same wave length as to sup-
port strongly the view that this substance is closely associated with genes
if not a component of them." The communications of Hollaender (41)
should be consulted on this point. This investigator points out that on
living cells the 2600 A region of ultraviolet is the most productive of
genetic effects. With Greenstein (42), he irradiated sodium thymonu-
cleate in water and observed that a decrease in streaming double bi-
refringence and viscosity occurred. A depolymerization was postulated.
Hollaender concluded that there was a primary absorption of the ultra-
violet by nucleic acid followed by the breakdown of the protein and the
disruption of the relation between the nucleic acid and the protein.

It is important in this regard to consider the observations of Stanley
on variants of tobacco mosaic virus (43). Differences in the amount
of one or more amino acids, the presence of a new amino acid, and the
complete elimination of one amino acid were found. This indicates
again that, in variation, changes in protein composition may be quite as
important as effects confined solely to nucleic acid. Work of this type on
mammalian cells is urgently needed, and certain studies are in progress
under Brown in the Memorial Hospital laboratories.

A particularly interesting observation, and one in keeping with Stan-
ley's results, is that of MacKenzie and Muller (44), who induced muta-
tions of Drosophila by wave lengths from 2800 to 3650 A. The effect
was especially marked at 3100 A, with a second maximum at 2280 A,
presumably a result of the absorption of energy by the protein com-
ponent. It would be interesting to explore this point further with other
types of organisms.

Gray and Tatum (32) report failure in their efforts to cause varia-
tions analogous to mutations in sexual forms (Neurospora) in bacteria
by ultraviolet, but Beadle and Tatum (90) were successful in the case
of Neurospora, and rather more mutant strains were produced by ultra-
violet than by x-rays. Certain differences are reported between the effects
of the two agents. All but one of the mutations resulting in an inability
to synthesize thiamine were caused by x-rays and, in contrast, those in-
terfering with inositol and choline formation followed ultraviolet ex-
posure. This point is a particularly intriguing one since it suggests the
possibility that certain genes are more sensitive than others to one specific
physical agent.

228 C. P. RHOADS

Blum (46) has reviewed the effects of ultraviolet light in great
detail. It is clear from his work that the wave length range responsible
for mutations and the absorption spectrum of nucleic acid in the ultra-
violet correlate closely with the effects of the light in causing cancer
(as well as erythema) of the skin in experimental animals. Here again
is evidence that a very specific agent, a narrow wave length range of
ultraviolet, clearly causative of mutations (or back mutations) in genet-
ically controlled material, is also active in inducing the change from the
normal to the neoplastic cell.

The decreasing sensitivity to radiation shown by more differentiated
cells is striking and lends interest to any means by which the sensitivity
to mutation can be reduced. Swanson (47) has shown that if ultraviolet
is used followed by x-rays, the effects of the latter on chromosome
aberrations are partly suppressed. These findings suggest that the struc-
tures in the chromosome are not equally sensitive and that certain ones
are particularly susceptible to change induced by a specific physical
agent. This cannot be a rule, however, since a strain of Neurospora has
been reported with two different mutated genes, one the effect of x-rays
and the other of ultraviolet (48).

Mutations Caused by Chemical Agents

The induction of inherited changes, but not of cancer, by antisera
has been reported. Guyer and Smith (49) injected an antiserum to
rabbit lens protein into pregnant rabbits and reported that transmissible
ocular defects occurred in the offspring. Hyde (50) is stated to have
confirmed these observations. Lens protein is the product of a gene and
not a gene itself, but it may be that the modification of the product, an
antigen, modifies the gene in turn. It is also possible, as specified by
Sturtevant (50), that genes have chemical specificities corresponding
to the antigens which they produce, and so are affected directly by spe-
cific antibodies. The work of Emerson (51) tends to confirm this as-
sumption. He obtained from culture filtrates enzymes, described as
extracellular and adaptive in character, against which immune sera
were prepared and used to treat Neurospora. Resultant inherited
changes in the ability of the mold to produce the enzymes are reported.
It is possible that this principle provides an explanation of the results
of Kidd (52) and of Green (13) in injuring cancer cells by antisera
against them.

For a very considerable period chemical agents other than antisera

NEOPLASTIC ABNORMAL GROWTH 229

which would cause gene mutation were actively sought, but clear proof
of their existence was lacking. More recently, however, a group of com-
pounds known as the halogenated tertiary alkyl amines (nitrogen mus-
tards) has been studied. From the effects of these substances on Dro-
sophila (53), on Neurospora (54), and on tobacco mosaic virus (55),
it is clear that they are effective in causing mutations of classic types,
it is interesting, furthermore, that they affect tissues in vivo very much
as x-rays do. Inhibition of mitosis, injury to intestinal epithelium,
damage to hematopoietic and lymphoid tissue, and necrosis of skin are
all eft'ects caused by both agents (56). Furthermore, the studies of
Bodenstein (57) reveal that the nitrogen mustards, applied to growing
tissue of the proper type and stage of development, stop cellular pro-
liferation and cause extensive swelling, possibly an effort toward differ-
entiation. This has also been shown by Karnofsky (58) in the Memo-
rial Hospital laboratories to be true for neoplastic cells of one type,
mouse sarcoma 180, grown on the chick embryo.

Another set of genetically active chemical agents are the carcinogenic
substances of classic type containing a phenanthrene nucleus. When
applied to tissue they cause profound alterations of cells which lead to
self-reproducing alterations coincident with the assumption of neo-
plastic characteristics. Stowell (59) has shown that these changes are
associated with alterations of nucleic acid content and Biesele (20) has
demonstrated irregular alterations of chromosome structure. The
metabolic and chemical abnormalities of the cancer cells induced by
carcinogens have been so completely reviewed by Burk (60) and by the
students of the Cowdry school (61) as to leave little doubt of their
importance.

Unfortunately, few data are at hand on the use of carcinogenic
chemical agents on material susceptible to genetic study. Beadle (25)
states that tests with these substances have not resulted in mutations.
Tatum (62) is said, however, to have made certain new observations
which provide acceptable evidence that mutations can be induced in
Neurospora by methyl-cholanthrene. Mottram (63) and Spencer and
his associates (126) have reported striking changes in protozoa, and
Strong (64) has recently published impressive studies on genetic
changes in mice which have resulted from the administration of the
same compound to young animals. Further work appears to be indicated.

A possible explanation of the difficulty in causing mutations in
Neurospora by the chemical carcinogens is the fact that they are altered
by contact with body cells and that the active compounds are quite as

230 C. p. RHOADS

likely to be the intermediate metabolites as they are to be the original
materials. Dobriner and Rhoads (65), Boyland (66), Jones (67), and
Weigert (68) have all reported on studies of the metabolism of these
substances. The very great insolubility of the original materials and the
data which attest to the formation of highly labile intermediate prod-
ucts in the course of conversion support the view that the attempts to
cause mutations may not have subjected the cells studied to exactly the
compound which is active in the mammalian tissue.

The studies of Earle (125) are particularly important as they con-
cern the alterations undergone by cells in the course of becoming malig-
nant. He cultured mouse fibroblasts in the usual heterologous serum-
chick embryo extract mixture. Subcultures were exposed for different
periods to varying concentrations of methyl-cholanthrene, and the cells
were inoculated into mice of the strain from which they had been
derived. The cultures assumed characteristic growth characteristics and
certain of them, after a rather narrow range of brief exposure to the
carcinogen, were found to have neoplastic properties. This was also
true of the mouse fibroblasts cultured for longer periods without methyl-
cholanthrene.

The conclusion from this work is unavoidably that mouse fibroblasts
grown for long periods under very abnormal conditions, particularly
with a carcinogen, become sarcoma cells. Further details of the environ-
mental circumstances to which the cells are exposed and the composition
of the abnormal cultures will be anticipated.

As far as is known, no studies acceptable to the geneticist have been
made on cells of the mutation-producing eflFect of the azo-carcinogens
of the type of dimethylamino-azobenzene. In the case of these com-
pounds much information is at hand from the studies of Kensler (69)
and his associates, as well as from the Wisconsin school (70), con-
cerning the pathways followed in the breakdown of the parent sub-
stance and the biological activities of the various derivatives. For
example, dimethylamino-azobenzene is wholly inactive in inhibiting the
metabolism of the tissue cells which have been studied manometrically
in acute experiments. Certain derivatives, however, are exceedingly
toxic. Suggestive evidence is at hand, furthermore, that other deriva-
tives, not hitherto isolated, may also be biologically active. It is not
sufficient to conclude that because the juxtaposition of a carcinogen,
such as methyl-cholanthrene, and a cell susceptible to mutation, such as
Neurospora, fails to give mutation, the chemical is incapable of doing
so. It must be shown, before this conclusion is reached, that the cell

NEOPLASTIC ABNORMAL GROWTH 23 1

has been exposed to all the derivatives of the carcinogen to which
mammalian cells are exposed in the course of carcinogenesis.

Pure Speculation Concerning Normal
AND Abnormal Mutations

From the data it is clear that certain correlations exist between the
production of mutations and the production of cancer. In general these
correlations concern the ability of some agents to do both things, and
the fact that the agents afifect, in doing both, those constituents of cells
which are composed of a material, nucleic acid, which is known to be
involved in genes and in self -perpetuating alterations of vegetative
bacterial material. Opposed to this concept of cancer as due to a dis-
continuous variation (mutation) are two considerations: first, the so-
called "virus theory" by which the neoplasm is held to be due to an
infection of the cell by a foreign parasite; and second, the fact that the
cancer cell clearly is at a competitive advantage as compared to the
normal cell, apparently able to carry out a vast number of processes not
available to its more dififerentiated normal associate. Were the cancer
cell a mutation in the accepted sense in which the term is used in describ-
ing the experiments on Neurospora, it should be less rather than more
competent, since most mutations studied precisely have been shown to
involve a loss rather than a gain of function. This fact requires a fur-
ther consideration of development and of the possibility of back
mutation.

Perhaps normal differentiation of cells is, after all, simply the result
of mutations occurring normally at a predictable rate in an organized
fashion with the continued growth only of those forms which have a
superior adaptation to their environment. The word "mutation" as
applied commonly to cancer should by this concept be "abnormal muta-
tion," defined more precisely as an abnormal change occurring at an
unpredictable rate and in a specific fashion. This is productive in large
part of lethal effects or of pathological conditions including cells which
are out of environmental control and so can compete to great advantage
with their associates. Beadle (71) states, ". . . 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. But such a duplicate gene may undergo mutation in such a way that
it directs the formation of an entirely new enzyme. If this new enzyme

232 C. p. RHOADS

should happen to catalyze a reaction that improved the organism in
competition with its relatives the new reaction would be retained."

The most dramatic experiments in support of this statement are those
of Lindegren {^2'). He studied the enzymes which ferment melibiose
and lactose as they are produced by yeasts of three strains and hybrids
between them. The enzymes are adaptive in character and appear, in a
genically qualified yeast strain, upon exposure to their respective sub-
strates. All segregants from heterozygous diploids carrying a single
pair of genes were able to adapt. Furthermore, two of the four adapted
melibiose-fermenting cultures lost the function when dissimilated, ade-
quate proof that the enzyme was transferred from the cytoplasm of the
heterozygous hybrid maintained on melibiose to the cytoplasm of segre-
gants which did not contain the gene. The enzyme was maintained in
these segregants as long as the sugar was present. The evidence is good
that the enzymes are self-perpetuating cytoplasmic entities, initiated by
genes, with levels independent of the gene but dependent on interaction
between the enzyme and the sugar.

Unfortunately, all the genetic considerations of cell differentiation
as the result of mutations, either spontaneous or induced, are rendered
unsatisfactory by the fact that our knowledge of gene action is all de-
rived from organisms which reproduce sexually. Cancer cells are vege-
tative. Hence, if a consideration of variability and adaptation in other
forms has some pertinence to the cancer problem, we must examine the
facts regarding variations in other vegetative cells. Bacteria provide
such examples, and reveal again the fact that those agents which cause
mutations in sexual organisms susceptible to genetic study also cause
inherited alterations, usually toward losses in function, in vegetative
forms in which we have no way of demonstrating true genes, and are
active in the production of cancer.

Bacterial Variation as Applicable to Abnormal

Growth

Pure cultures of microorganisms, like the tissues of the body, even
though they are derived from isolated single forms, are known not to
consist entirely of identical individuals. Under standard conditions
bacteria vary at a constant rate. If the variation provides the cell with
an advantage in its environment, the variant continues to multiply and
overgrows the cells of the type from which it arose. If it has derived no
advantage from the variation, it is overgrown and dies. Modifications

NEOPLASTIC ABNORMAL GROWTH 233

may be transient, and if so are usually a direct effect of environment
and return to normal if the environmental alteration is removed; or
they may be permanent, of a type for which no environmental cause
can be found. It may be assumed that these permanent changes are
analogous to the spontaneous mutations which occur in sexual forms.
A third type of variation occurs during the growth cycle, and may be
analogous to the changes leading to differentiation in more complex
cells.

The changes which are transmitted by the altered cell to its progeny
are termed discontinuous variations and include certain of the consti-
tutive enzymes. The classical variation of this type is the unexplainable
occurrence of lactose-fermenting forms of E. coli mutabile, normally
a non-lactose-fermenting organism. It has been shown by Lewis (74)
that all colonies of this organism generate about the same number of
variant cells, but so few as to escape detection by conventional methods.
The variation may occur in the absence of lactose.

Dubos (73) states that in bacterial cultures non-beneficial variations
such as inadequate utilization of nutrients can be eliminated by over-
growth of the more vigorous original cells. This statement is an impor-
tant one, since it should also apply to the more complex cells of tissues.
However, normally it would not do so in the case of handicapping
variations of differentiated tissue cells because, in contrast to movable
bacteria, tissue components remain fixed in their relative positions. The
differentiated cell, which perhaps has normally mutated, becomes spe-
cialized in its function, loses its adaptability, and has to compete only
with similarly specialized cells adjacent to it. The most adult do not
come in direct contact with the young and particularly vigorous cells,
and so do not compete under the handicap imposed by the greater
adaptability of primitive forms. Thus, in tissue, cells limited to new
and specialized functions, such as seem to appear during differentiation,
are protected in their existence unless there develop in their midst more
competent and adaptable forms, such as neoplastic cells may be.

The development of abnormal cells in tissues can be caused experi-
mentally by the application of certain agents as specified previously.
All of these cause mutations, and all have a pronounced toxic or inhibi-
tory effect on the normal cells with which they come in contact, with
the result that the pathological forms, characteristic of neoplasm, ap-
pear. Perhaps analogous to this development of cancer following the
prolonged intoxication of normal cells are the discontinuous variations
of bacteria which yield abnormal cells characterized by their resistance

234 C. p. RHOADS

to injurious agents. Antibacterial substances such as gentian violet and
penicillin have been shown to cause the production of small resistant
colony variants of staphylococci (75). Many species of bacteria have
been exposed during their growth to immune sera with resultant
changes in their antigenic (chemical) structure (76). These experi-
ments must indicate, as discussed previously, that some factor which
controls the self -reproducing character of bacteria, analogous to genes
in the case of Neurospora, is changed as an effect of antisera, prepared
not against the genes themselves, but against the material which they
produce. It would be fascinating to know in what way, if any, the
further sensitivity of bacteria to agents such as ultraviolet light at
2650 A, a wave length known to injure specifically the genes of sexual
forms, is altered when a discontinuous variation due to the same agent
has occurred. Evidence is becoming available that resistance to ultra-
violet does follow exposure to it.

Whereas the evidence indicates that each bacterial organism has an
inherent potentiality for producing a variant form once in a given
number of divisions, it is difficult, as with tissue cells, to analyze these
phenomena in the terms of classical genetics. To do so would of course
require the ability to see and to describe alterations of structures in the
vegetative cell which could be interpreted as having a function like the
chromosomes and genes of sexual forms.

The studies of Lea (33) and of Gowen (34), mentioned previously,
bear on this point. From this work it is clear that the use of a physical
agent, x-rays, causes inherited alterations in viruses and bacteria similar
to the ones caused in the more complicated types of higher organisms.
In sexual forms the changes are known to be due to alterations of the
genes. In both sexual and vegetative cells the changes are believed to
follow hits on certain sensitive areas of the affected organism, areas
which appear to be functionally quite analogous to the genes.

The transmutation of pneumococcal types, although it has been dis-
cussed repeatedly, deserves mention here. It can be depicted by a simple
formula: Avirulent, unencapsulated Type II plus killed Type III equals
virulent, capsulated Type III.

The reaction requires for transformation a competent culture. This
culture may vary and result in non-encapsulated variants incapable of
reacting to the transforming substance. It is clear that there exist a
number of forms of unencapsulated variant pneumococci endowed with
varying potentialities and physiological abilities. The study is best
summarized by a direct quotation from Dubos (73a) :

NEOPLASTIC ABNORMAL GROWTH 235

"The data obtained by chemical, enzymatic, and serological analyses
together with the results of preliminary studies by electrophoresis,
ultracentrifugation, and ultraviolet spectroscopy indicate that, within
the limits of the methods, the active fraction contains no demonstrable
protein, unbound lipid, or serologically reactive polysaccharide, and con-
sists principally, if not solely, of a highly polymerized, viscous form
of desoxyribonucleic acid. On the other hand, the Type III capsular
substance, the synthesis of which is evoked by this transforming agent,
consists chiefly of a non-nitrogenous polysaccharide constituted of
glucose-glucuronic acid units linked in glycosidic union. The presence
of the newly formed capsule containing this type-specific polysaccharide
confers on the transformed cells all the distinguishing characteristics
of Pneumococcus Type III. Thus, it is evident that the inducing sub-
stance and the substance produced in turn are chemically distinct and
biologically specific in action and that both are requisite in determining
the type specificity of the cell of which they form a part."

Information which is perhaps pertinent to a better understanding
of the unique qualities of the cancer cell may be derived also from
studies of the factors which make a bacterium pathogenic. The loss of
the O type specific polysaccharide characteristic of the S— >R variation
of gram negative bacilli, like the loss of the specific polysaccharide of
the pneumococcus, is associated with loss of virulence. The M sub-
stance of group A streptococci gives type-specificity and seems to be
characteristic of all virulent strains. It is usually supposed that the dis-
appearance of characteristic substances in bacterial variants is due to
the fact that these substances are not produced. It may be, on the other
hand, that in the variant cells the special substances are metabolically
broken down by the acquisition of new functions. Although it is usually
stated that the non-specific or undififerentiated bacteria have lost the
ability to produce the O antigens or the capsular polysaccharide or
the M protein, the possibility has not been ruled out that in the trans-
formation from dififerentiated to undififerentiated cells they have be-
come able to metabolize those substances further and hence do not
accumulate them in recognizable form. Perhaps by this concept an
analogy can be drawn between bacterial variation and the differentia-
tion of mammalian cells. The assumption by tissue of special characters
and powers coincident with maturation may actually represent discon-
tinuous variations which involve the inability to carry certain reactions
to completion, with the consequent accumulation of the products of
these incomplete reactions.

236 C. p. RHOADS

Recent studies by Tatum {'jy) bear on this point. A mutant strain
of Neurospora unable to synthesize nicotinic acid accumulates a special
pigment, never formed by its progenitor, the wild type organism which
can synthesize from nitrogen and sugar all of its components except
biotin. The pigment is formed from an unused precursor of nicotinic
acid. It is not the precursor itself, but rather a product of conversion of
the precursor to pigment by an enzyme not demonstrable in the wild type
mold. By a mutant blocking of one system, that for synthesizing nico-
tinic acid from the precursor, a new enzyme previously unused because
unneeded, comes into play. The apparent appearance of a new function
is largely the disappearance of an old one.

Concerning the abnormal functions of cells, particularly those of
neoplastic tissue, a generally similar but abnormal process conceivably
could have taken place with the production of variants possessing wholly
unique abilities. For such a cell to survive requires that it not only be
successful in competition with the normal cells in the company of which
it finds itself, but also not susceptible to injury by whatever general
mechanisms are possessed by the body for the control of cell growth.

Genetics and the Cancer Cell

The material presented provides a basis for a consideration of cellu-
lar differentiation, normal and abnormal, in terms of the physiology of
simple forms. If differences between cells are genetic, they must be
due to somatic mutation, the alteration of genes. No final proof of gene
effects in the production of cancer is possible, since tissue cells are
vegetative and not sexual, and since chromosome changes are not regu-
larly demonstrable. Mutation is believed to be a possible cause because
cancer can and does arise in localized areas, perhaps from single cells,
and it results regularly from agents known to cause mutation. Since
genes control specific steps in metabolism, they must have altered in
the production of neoplastic cells of abnormal composition. Beyond
this point of speculation we can do little more than argue from analogy.
It may be useful, however, to set down what is known about how cancer
cells differ from their normal prototypes.

The earliest experimental information concerning the cancer cell
came from the demonstration that under certain circumstances it could
be transplanted into otherwise normal hosts. This naturally required
reference to the rules which apply to the transplantation of normal
tissue from one animal to another. In general, an animal will not accept

NEOPLASTIC ABNORMAL GROWTH 2^7

a graft containing a dominant allele which it itself does not have. The
reverse, however, is not true, and dominant alleles in the host are with-
out effect upon the acceptability of the graft. It is probable that these
dominant genes are concerned with antigen formation. Lacking a gene
competent to form an antigen, the foreign cell has no power to evoke
an antibody response and so to subject itself to destruction. If the
foreign cell or graft contains or can form antigens not shared by the
host, antibodies are induced which react with the graft antigens so that
the graft is cast off or resorbed. The publication of Medawar should
be consulted on this topic (78).

Ample evidence exists from the extensive work of the Little group
(79) that the ability of tumor cells to survive in the host is dependent
on both their constitution and that of the host. A tumor cell graft obeys,
in part, the rules which apply to a graft of normal tissue. If the tumor
carries one or more dominant genes which produce antigens and the
host has no, or inactive, alleles of these genes, the tumor graft induces
antibody formation against itself. Regression can result under these
circumstances. It is known, moreover, that transplantable tumors may
mutate so that fewer dominant genes are required in the host for suc-
cessful transplantation. This is, of course, the same thing as saying
that there are fewer dominant genes in the transplant capable of setting
up antibody formation. Some mouse tumors, for example sarcoma 180,
are almost without strain specificity, and further immunologic study
of this strain is desirable. Other mouse neoplasms are exceedingly sensi-
tive to genetic differences, particularly certain strains of mouse leu-
kemia (Furth (80)). MacDo well's (81) consideration of the immuno-
logic asp€cts of this disease has been particularly detailed. He advances
evidence that as soon as a leukemic cell is transplanted it becomes anti-
genically altered. Homozygous animals can be rendered immune to
leukemia of their own line, but this does not serve to prevent the devel-
opment of the spontaneous disease in the same animals. This concept,
however, does not take into consideration the possibility that when the
spontaneous disorder occurs the host may be altered in some way so
as to be no longer capable of immune response.

It is interesting to consider a correlation between decreasing numbers
of dominant genes and increased transplantability. It almost seems as
if the cell which has the least number of distinguishing characteristics
is the cell which is the most virulent, most capable of surviving under
circumstances set up by the body as adverse ones, and, perhaps the most
malignant. This is consistent with the thesis that the neoplastic cell is

238 C. p. RHOADS

a dedifferentiated one. Proof of this requires evidence, not now avail-
able, that normal differentiated cells are immunologically much more
complex than the primitive cell, and both more so than the neoplastic
cell.

The studies of Kidd (11), which are referred to later in detail, are
important in their bearing on the question of host resistance to trans-
planted tumors. They contain an admirable review of the literature on
the subject. Attention is called to the fact that some mechanism of pro-
tection against transplants exists other than the one characterized by
the presence of humoral antibodies. Kidd shows that whereas comple-
ment-fixing substances are frequently associated with regression of
the Brown-Pearce tumor, and injure the cells of that tissue, similarly
specific antibodies have no effect on the course of the V2 carcinoma.
Furthermore, examples are presented of solid immunity in the apparent
total absence of such antisera. Barrett (82) has published from Mur-
phy's laboratory beautiful experiments which prove the difficulty, if
not impossibility, of creating an immune state to cells homozygous
with the host.

In this respect an immunologic study of the differences between the
wild strain of Neurospora and the mutant derivatives would be of very
great interest. It is not impossible that the more biochemically com-
petent wild organism possesses fewer points of immunologic specificity
since there is less tendency to accumulate the products of incomplete or
abnormal metabolic processes. Were this concept, as presented by Led-
erberg (83), to be supported by experimental evidence, the cancer cell
could be considered as analogous to the wild strain Neurospora, a back
mutation to a completely competent form. It would then become neces-
sary to view the normal cell as the mutant, and again differentiation
would be seen as the result of a progressive series of mutations with
loss of function and thereby the discovery of special functions.

Abnormal Growth due to Hybridization

An important experiment in the induction of neoplasms is the work
on hybridization. Gordon (84) studied species of two genera of tropical
fish and found that hybrids between them resulted in the production of
melanotic neoplasms in greater numbers than would occur in the parent
species. Whitaker (85) described tumors which occurred in the Fi hy-
brids of crosses of Nicotiana glauca and Nicotiana Langsdorfi. He
ascribed this result to the introduction of chromosomes of the latter

NEOPLASTIC ABNORMAL GROWTH 23Q

into the cytoplasm of the former. Little (86) in a most significant ex-
periment crossed two species of mice, Mus bactrianus and C57 black
derived from Mus musculus. The two species differ markedly in size,
fertility, and rate of growth. The incidence of tumors in the Fi hybrids
was far greater than could be accounted for even by the addition of the
spontaneous tumor-forming tendencies of the parent strains.

It is curious that this type of experiment has not been made more
frequently, since it is wholly in keeping with other evidence which has
accumulated concerning the development of neoplasms. The work with
the Shope papilloma virus in the induction of tumors in domestic rab-
bits clearly involves the introduction of a cellular protein constituent
from one species into the cells of another and not too closely related
species, with resultant tumor production and disappearance of the virus.
The studies of Claude (87) concerning the enzymatic activities of cer-
tain cytoplasmic granules suggest the possibility that such protein par-
ticles might modify genes if introduced into the cytoplasm of a cell of
another species. Since the genes produce enzymes (25), and modifica-
tion of enzymes can alter the source genes (51), and since one group of
chemical carcinogens form in their metabolism substances toxic to en-
zymatic activity (69), the route of both carcinogen and filterable factor
may be to the gene through the enzyme. This type of thought must be
considered as the wildest speculation, however, until much more evi-
dence is available. One would like to have precise data on the ability of
enzymes from one group of cells to affect the activity of enzymes from
another group, clear proof of a change from normal of the enzyme
activities of the cancer cell, more facts bearing on the participation of
the filterable tumor-producing agents in the activity of kinetically active
chemical systems, and finally a satisfactory demonstration of the ability
to alter genes by the introduction of different enzymes into a cell.

The increased rate of tumor production by hybridization, like that by
carcinogenic chemicals in genetically suitable material (88), provides
another example of gene mutation caused by procedures also effective
in causing neoplastic disease. Sturtevant (21) proved that the offspring
of the back-crosses from hybrids between two races, A and B, of Dro-
sophila pseudoobscura show a very large increase in mutation fre-
quency. In this material 9% lethals and 0.5% of sex-linked viables
were produced without any evidence of increase in the number of major
chromosome aberrations like those resulting from x-radiation. This ex-
periment may be of great significance because it proves that mutations
can be caused without obvious morphological alteration of the geneti-

240 C. p. RHOADS

cally active elements. One argument against cancer as a mutation is the
"recognized irregularity of the chromosome equipment in cancer cells."
Whereas in given tumors no nuclear abnormality may be discernible,
other types present the appearance of extreme chromosome hetero-
geneity.

Abnormal Growth due to Cell Free Materials

The concept of cancer as a disease caused by a protein component of
cells, a "virus," has received increasing attention during the past few
years. Indeed, the point has been reached at which so much attention is
given to the filtrate or virus etiology as to obscure somewhat the exact
sense in which the term is used and its application to the possible pre-
vention or control of cancer.

The original proof of cancer induction by a cellular component came
with the demonstration by Rous (89) that a malignant mesoblastic
tumor of fowls, occurring spontaneously, could be transmitted to nor-
mal birds by the inoculation of cell-free antigenic material from the
original growth. Subsequent studies by Claude revealed that the active
material is a complex composed of lipo-protein and a ribose nucleic
acid, and that it is chemically and immunologically indistinguishable
from certain constituents of normal embryonic fowl cells. The con-
clusiveness of the evidence suggested that other, indeed perhaps all,
neoplasms are due to cell-free agents and that cancer in general is an
infectious contagious disease. This inclusive and far-reaching view was
supported by the fact that a number of other fowl tumors were proved
to be capable of being transmitted by cell-free material. Particularly
striking is the evidence of Ellermann (91) concerning fowl leukosis.
This neoplasm not only can be transmitted by filtrates but also appears
to be truly contagious, since contact transmission seems to have been
proven completely by the distinguished work of the Michigan investiga-
tors (92). Later studies revealed a number of additional examples of
active tumor-producing cell-free materials. The Shope papilloma of cot-
tontail rabbits, the oral papilloma of dogs, and the mammary tumor
inciter of mice are all agents which may be dissociated from whole cells
and which certainly cause benign neoplasms to appear.

The studies of Berry (93) are important in any consideration of the
virus etiology of neoplastic disease. This investigator used the principles
effective in demonstrating the transforming principle of pneumococci
in the transformation of the virus of rabbit fibromatosis to an agent

NEOPLASTIC ABNORMAL GROWTH 24I

capable of inducing in rabbits a wholly different cytological structure,
that of infectious myxomatosis. The material which causes the trans-
formation appears to reside in the myxoma virus nucleoprotein. This
mutability of a virus is a striking and frequently observed property.
The studies of Duran-Reynals (94) on the Rous agent prove that it has
wholly different pathological effects with the production of acute necro-
tizing lesions, when injected into very young chicks, than it does when
inoculated into adult fowls with the resultant tumor of avian mesen-
chyme. Indeed, this investigator has raised the question as to whether
tumor production by this agent may not be a special reaction given by a
partly immune host. The fact should be recalled, however, that the Rous
agent is quite unique among viruses. It is neutralized by rabbit anti-
serum to normal fowl tissues and the antibodies of such serum are ad-
sorbed out by extracts of normal chick embryo. Furthermore, chemical
studies on the agent reveal a close similarity with a substance obtained
from normal chick embryo. The contributions of Claude (95) should be
consulted on this subject.

The results reported by Rous and his associates (96) prove that the
cutaneous papillomas experimentally produced by the inoculation into
domestic rabbits of a cell-free material, the virus of Shope (97), ob-
tained from growths of western cottontail rabbits, frequently result in
the development of cancer in the involved tissue. An extensive series of
publications has appeared concerned with this experimental study and
with certain correlative data derived from immunologic studies with
the Brown-Pearce transplantable tumor of rabbits. The conclusion was
reached that both types of rabbit cancer investigated are caused by
"viruses" which cannot be recovered from the cancer tissue but actually
persist in a "masked" form. This is a most important conclusion, if
supported by the evidence, since it would prove that at least two and
possibly all forms of cancer in mammals are infectious, and, as such, to
be transmitted by contagion, even though no disease-producing material
can be demonstrated in them.

Morphological evidence has been advanced to prove that the cancers
associated with transmissible papillomas in both wild and domestic rab-
bits arise directly from the papilloma cells. This conclusion is based prin-
cipally on the finding that the virus causes the differentiating epidermal
cells to become much larger than normal and to have much larger vesic-
ulated nuclei with marginated chromatin. It is stated that "all these
changes may be encountered individually in tar cancers (for which no
virus etiology is claimed) yet when found together they are highly

242 C. P. RHOADS

characteristic of the action of the virus" (98). This differentiation by
cytological criteria could now conceivably be strengthened by evidence
derived from studies employing tracer elements.

Irrespective of whether virus can be recovered from the papillomas
or not, an antibody capable of fixing complement, in the presence of a
saline extract of the papilloma as an antigen, appears in the blood of
rabbits bearing the growths, and it increases in titer as the growths en-
large. Similarly, extracts of papillomas which yield no virus will, when
injected intra-peritoneally into normal rabbits, evoke specific antiviral
antibody, which fixes complement in the presence of the original antigen
and only that material, and, when mixed with virus, reduces or removes
(neutralizes) that agent. Not only are the papillomas antigenically ac-
tive, but the transplanted cancers are also, even though no virus can be
demonstrated in them. The virus-neutralizing antibody which circulates
in the blood of rabbits carrying the virus-induced papillomas has no
effect upon the course of the lesions however, or upon the growth of the
cancers, a fact which is ascribed to a protection of the virus by its host
cells from the action of the antibody.

The absence of virus from the papillomas and cancers of domestic
rabbits is notable, as is its lack in the cancers of the cottontails, although
it is abundant in the papillomas of that species. Kidd (99) sees two pos-
sible explanations of these findings : that the virus had disappeared from
the papilloma when the cancer developed and did not exist in the cancer,
or that in both the tissues virus was present but was masked, i.e. could
not be demonstrated, though actually present. To prove that the second
explanation was the correct one, a search was instituted for a virus-
neutralizing substance or inhibitor in the cancer. Table II of Kidd's
paper (99) presents the evidence that such an inhibitor was found. Ex-
tracts of glycerinated cancer tissue were mixed with papilloma virus
and inoculated into domestic rabbits. The controls were of virus mixed
with Tyrode solution. No controls with benign papilloma tissue were
used, since it was assumed that they could not contain an inhibitor be-
cause virus could be demonstrated in them, although the assumption
that the cancer did not contain a virus was not accepted. Of fourteen
tests with the virus cancer tissue mixtures, virus activity was found in
thirteen, but delay in papilloma growth as compared with the Tyrode
control was found in eleven. It should be noted that a second control re-
ported, with Tyrode and virus, showed no more activity than was seen
in eleven of the fourteen tests presented to prove the presence of an
inhibitor (100).

NEOPLASTIC ABNORMAL GROWTH 243

The presence of the inhibitor in tumor is invoked to explain the in-
ability to recover virus from mixed papillomas and cancers of cotton-
tails, as well as from the cancer tissue alone. Friedewald (loi) reported
that antibody to the papilloma virus is present in extracts of the virus-
induced growths of domestic rabbits, even though little or none may be
present in the serum. This fact is used to explain the failure to obtain
virus from the papillomas and also to explain, as due to extravasated
antibody, the presence of inhibitor in the cancer in the cottontail arising
in the virus-induced papilloma, a view in support of which morpholog-
ical evidence of hemorrhage and extravasation is advanced.

Since no way was thought feasible of separating extravasated anti-
body from the papilloma virus and so of demonstrating the masked
virus in the cancer derived from the papilloma of the cottontail, an ex-
periment was made to recover virus from a cancer not caused by it, and
so not productive of antibody (looa). Virus was mixed with the tissue
of tar-induced cancer of cottontails and the tissue transplanted into nor-
mal hosts. Antibodies to the virus were developed and the virus is stated
to have increased the speed of growth of the implanted tumors and
sometimes to have altered their morphology so that they looked, not
like tar cancers, but like virus cancers. From these transplanted virus-
infected tar cancers virus could sometimes be recovered if the anti-viral
activity of the host's blood had not risen to levels too high to allow a
satisfactory test. It is not clear how virus escapes from the cells bearing
it to act as an antigen, but the antibody fails to penetrate the cells and
kill the virus.

Whereas the papilloma caused by virus inoculation in the cottontail
yields virus regularly on test and is associated with virus-neutralizing
humoral antibodies, in domestic rabbits from flourishing papillomas
caused by the same material virus was obtained only irregularly. Fur-
thermore, the sera of the animals that carried growths yielding little or
no virus had little or no virus-neutralizing capacity. These findings are
summed up by the statement that the facts make it improbable that anti-
body is responsible for masking the virus in the growths produced by it
in domestic rabbits, although the presence of antibody as inhibitor in
tumor tissue is invoked to explain the inability to recover virus from the
inflamed papilloma cancers of cottontails.

It is remarkable that no antibodies can be shown to be associated with
the papilloma, which yields no virus but gives rise to a cancer also not
yielding virus but actively productive of antibody. Antibody cannot ex-
plain the failure to obtain virus from papilloma of domestic rabbits but

244 ^- ^- RHOADS

is invoked to explain the failure to do so from the cancer of cottontails.

The conclusions drawn from the immunological evidence presented
concerning the papilloma virus and the cancer supervening upon it in
rabbits is that "the cancers result from virus variation, this in many in-
stances being slight" (lOo). This conclusion is arrived at because of
evidence for the complete specificity for the virus of the antibodies as-
sociated with the cancer, and the complete failure to derive any infec-
tious material whatever from the cancers following virus-induced papil-
lomas which are supposedly due to the virus. If the serological evidence
for the virus etiology of the cancer is to be accepted, then the virus
should be the one reacting specifically with the antibodies, and not a
modified one. If the evidence is not to be accepted, it is hard to support
the view that the papilloma virus is the etiologic agent. If some modifi-
cation of the virus is to be invoked as etiologic, then evidence in support
of the existence of a modified agent is required. This is advanced in a
recent communication by Kidd ( 1 1 ) announcing the demonstration in
the V2 carcinoma of an antigenically specific protein wholly distinct im-
munologically from the papilloma virus. Further studies will be of inter-
est as they concern the place of the virus in this picture.

Support for the theme of a virus, or extraneous, infectious parasitic
etiology for the cancer in rabbits following the virus-induced papilloma
is further sought in immunologic studies of the Brown-Pearce trans-
plantable cancer in the same species (10). There was obtained from this
tumor a serologically active substance, demonstrable by its ability to fix
complement (no agglutination or precipitin reaction was noted) which
was insoluble in alcohol, was inactivated by heating to 65° for 30 min-
utes, by treating with acid to pH 4.5 or lower or alkali to />H 11.5 or
higher, with a uniform particle size over 348ja, and centrifuged down
by 20,000 r.p.m. for one hour. These general characteristics of a large
protein molecule with the ability to fix complement in the presence of a
specific antibody are advanced to favor the infectious etiology of the
tumor, though no infectious agent was obtained from it.

The matter of immunologic specificity of neoplastic cells is one of
great interest and importance. Beyond doubt new methods and new ex-
periments are desirable. The studies of Andrewes (102) and of Fouldes
(103) should be consulted in this regard. These investigators proved
that substances related antigenically to the chicken-tumor agents are
present in certain fowl tumors elicited with carcinogenic chemicals. The
latter have not been proved to be filterable, however, and hence the ob-
servation tells us only that the induction of cancer by a carcinogen

NEOPLASTIC ABNORMAL GROWTH 245

evokes alterations in avian tissue of molecular structures related in some
degree to those which occur spontaneously and transmit the disease.

Cytoplasmic Control of Inherited Cellular Traits

Since the early work in cytology of Roux and Weismann, evidence
has been available that constituents of cytoplasm play some role in the
development of the cell. The original experiments prove that removal
of particular regions of the ooplasm of the unsegmented egg without
injury to or disturbance of the nucleus is followed by corresponding
defects in the resultant embryos and larvae. This was taken to prove the
definite prelocalization in the cytoplasm of the unsegmented egg and
specification of the blastomeres due to the nature of the specific cyto-
plasmic materials which they receive during cleavage. This view was
further supported by the classic studies of the Spemann school on the
organizer substances. More recently, Sturtevant (104) examined the
evidence on the inheritance of dextral and sinistral coiling in the gas-
tropods and found that the eggs of a sinistral individual produce sinis-
tral offspring even if fertilized by sperm carrying dextral dominant
factor. J. A. Moore ( 105) showed that the cleavage rate of echinoderms
is a function of the cytoplasm, and E. B. Harvey (106) advanced evi-
dence that in the parthenogenetic merogony of enucleated egg frag-
ments chromatin plays only a minor role in early development.

The experiments of Lindegren, previously mentioned, on the meli-
biose-splitting enzyme in yeast formed under the control of a specific
gene, are important. Once formed, the enzyme is produced as long as
the substrate is present, even in the absence of the active allele of the
gene which produced it. Beadle (25) draws an interesting analogy be-
tween this observation and the formation of pepsin from pepsinogen.

Michaelis (107) studied the plant Epilobium and showed that recip-
rocal crosses between species and different races of single species show
differences between hybrids of first generation and behavior on back-
cross, indicating that cytoplasmic entities transmitted through the egg
are involved and depend upon the nuclear constitution of the plant in
which they are found. What is normal in plants of one genetic constitu-
tion is abnormal in those of another.

Kiihn and Plagge (108) showed that the eggs of Ephestia, homozy-
gous for the a allele of the Aa gene pair, carry a hormone which requires
the A allele for its production and which disappears in subsequent gen-
erations.

246 C. p. RHOADS

The studies of Imai (109) and of Rhoades (no) on cytoplasmic
control have attracted great attention in recent years. In the develop-
ment of chlorophyll by maize, pure genie control was proved in 100 in-
stances. In two cases, however, cytoplasmic inheritance was shown to
exist. In this plant, a specific gene has the power to cause a modification
of the plastid marked by smaller size and lack of chlorophyll. The modi-
fied plastids have genetic continuity not affected by the constitution of
the genes. To quote : ". . . although induced by a nuclear factor, the ij
gene, the mutated plastid, like a Frankenstein monster, is no longer un-
der the control of its maker."

The work of DuBuy and Woods ( 1 1 1 ) provides interesting new ma-
terial on cytoplasmic inheritance in plants and their reports should be
consulted directly. They discuss the evidence for regarding plastids as
analogous to mitochondria with particular reference to their content of
ribose nucleo-protein and to their tendency to break up unless their
handling is carried out with great precision. The modified plastids, as
the source of spontaneous variegations, are discussed and a virus
deemed not likely to be the cause. Of special interest are the data which
indicate that certain plant viruses compete with the host chromoprotein
for nitrogen and that both normal and abnormal (virus) enzyme sys-
tems can be inhibited by cyanide. Success in transplanting variegated
plastids with resultant self-propagating abnormalities in host plants
are described. The likeness between these studies and the disease in-
duced by the cytoplasmic material of tumor cells induced by the Rous
agent is dramatic.

Most interesting of the studies concerning the inheritance of cyto-
plasmic factors, and perhaps the most pertinent to the cancer problem,
are those recently reported by Sonneborn (112). This work is so im-
portant that it is discussed in extenso.

Sonneborn discovered that one race of Paramecium aurelia, when
placed in culture with organisms of another race, killed the latter by
means of a substance secreted into the culture medium. This ability to
kill was assigned to a gene called "killer gene" or K. The recessive allele
of K, which was present in the animals which were killed, was called
"sensitive" and designated by k. The characters K and k have been
shown to be inherited by simple Mendelian laws, and hence animals can
have the genotypes : KK, Kk, kk. It was found, however, that the
phenotypes, or the expression of these characters, did not necessarily
reflect the genotypic make-up. This was due to the elaboration into the
cytoplasm by X" of a substance "Kappa." Kappa is the material that

NEOPLASTIC ABNORMAL GROWTH 247

actually determines the phenotype of the organism as to whether it is a
killer or a sensitive. Paramecium aureli-a divides by simple fission, and
two sexual processes may intervene. The first of these is conjugation in
which two animals come together to form a cytoplasmic bond and ex-
change, reciprocally, one micro-nucleus. This is preceded by two reduc-
tion divisions in each animal. The micro-nuclei, which migrate, fuse
with corresponding micro-nuclei in each animal to form a zygote nu-
cleus. There may be, during conjugation, considerable exchange of
cytoplasm, a fact which is significant. The second process is autogamy,
in which the micro-nuclei in a single animal go through two reduction
divisions. Two daughter nuclei fuse to form a zygote nucleus and the
others degenerate. Sonneborn found that animals containing the gene
K and having Kappa present in their cytoplasm are invariably killers.
Animals of kk genotype which do not have Kappa in the cytoplasm are
invariably sensitive. However, the distribution of Kappa in the cyto-
plasm means that, due to the presence of the cytoplasmic bridge, it can
be distributed independently of the micro-nuclei on conjugation. Ani-
mals possessing the genotype Kk undergoing autogamy might change
in genotype from Kk to kk because of the two reduction divisions and
consequent independent genie segregation of these two alleles K and k.
It was possible, therefore, to obtain animals of the following make-up:

NUCLEUS CYTOPLASM

1. KK Kappa

2. Kk Kappa

3. kk Kappa

4. KK —

5. Kk —

6. kk —

Consequently, animals i, 2 and 3 were killers regardless of their geno-
types ; animals 4, 5 and 6 were sensitives even though 4 and 5 contain
K gene.

It was demonstrated that K can maintain and increase the amount of
Kappa present in the cytoplasm but cannot initiate its production. Once
an animal of composition KK or Kk is completely deprived of Kappa
it remains a sensitive animal regardless. Conversely, an animal of geno-
type kk containing Kappa in the cytoplasm will be a killer although it
will revert to sensitive within four to six generations due to dilution of
the substance Kappa below the point at which it can manifest the killer
character. These animals subsequently will be sensitive.

248 C. p. RHOADS

In examining a number of varieties of killer animals, Sonneborn
found, in variety 4, that if grown rapidly with fission occurring quickly,
after several generations some of the animals become sensitive. Subse-
quently, however, these sensitive animals would revert to killer when
grown slowly. If he continued to grow them rapidly more and more of
them were converted to permanently sensitive animals and fewer of
them would revert to killer under the effect of a slower rate of multipli-
cation. In other words, in this case the organisms grow more rapidly
than particles of Kappa are elaborated. Sensitive animals result when
the concentration of Kappa in the cytoplasm falls below a certain critical
level. In the animals which remain permanently sensitive, after the rapid
growing of many generations, Kappa becomes entirely diluted out, and
even though the genie make-up was K, the animals remained sensitive
due to the inability of K to initiate the production of Kappa.

From these data, Sonneborn was able to calculate the number of par-
ticles of Kappa in the normal killer animal, getting an estimate of be-
tween one and three hundred particles by two different methods of
calculation. He showed that a single particle of Kappa in the cytoplasm,
while not enough to manifest itself, was enough to enable the animal
ultimately to revert to killer provided the nucleus contained a K. His
data indicates that Kappa may have a limited power of independent
self -reproduction in the cytoplasm in the absence of i^ but that the major
factor controlling its elaboration in the cytoplasm is K. The number of
particles of Kappa necessary to cause the animal to be a killer ranges
perhaps from fifty to sixty.

These data indicate that a substance elaborated into the cytoplasm
controls the cell phenotype and may be to some extent independent of
the genotype. Thus an animal recessive {kk) for killer may, in effect, be
a killer phenotypically due to the presence of Kappa in the cytoplasm. It
indicates also that Kappa is elaborated at a rate independent of the gen-
eral rate of growth and multiplication. Thus sensitive animals with the
genotype KK or Kk and containing a sub-critical amount of Kappa in
the cytoplasm might become killers following autogamy because autog-
amy prevents fission for many hours and this allows the independent
production of Kappa in the cytoplasm to a point at which animals mani-
fest the killer characteristic. Killer animals held for twelve to thirty-
six hours at temperatures higher than normal (37°-38°) become sensi-
tive but revert subsequently to killers. This may be due to inhibition of
the production of Kappa at higher temperatures, or to its breakdown at
these temperatures.

NEOPLASTIC ABNORMAL GROWTH 249

Kappa particles are distributed to the cytoplasm normally in a ratio
of I particle/600: 1 000 cu. mm. of cytoplasm.

It should be noted that Kappa in many ways, however, fulfills the re-
quirements of a virus. Sonneborn believes it is not a virus since the fac-
tors determining mating types in Paramecia are likewise cytoplasmic,
and it is thought doubtful that mating types are determined by virus
infections. Clearly, these brilliant experiments deserve extension. It
would be most illuminating to have further data on the composition of
Kappa, its ability to survive in tissue culture or on chick embryo, and
the size of the particles.

The experiments specified concerning cytoplasmic inheritance are
discussed, of course, because of the interest in the "virus" theory of
cancer, which concerns to a considerable extent the activity of particles
supposed to be cytoplasmic in origin and location. The term "virus" is
generally employed to specify an agent entirely foreign to the normal
tissue it invades, a view derived from the concept of viruses as modi-
fied bacteria. It appears that ample evidence is at hand to prove that
autopropagating substances which reside in the cytoplasm of cells partic-
ipate, in certain instances, in the control of inherited characteristics. It
is not at all clear and seems unlikely that these cytoplasmic agents are
always of cytoplasmic origin and not genie, and subject, at least to some
extent, to genie control. The work of Sax (113) bears on this point.
He showed clearly that gene products essential for growth are able to
diffuse between the cytoplasm of attached cells but not between isolated
microspores.

It may be well in this connection to quote Wilson ( 1 14) : "The chro-
mosomes are as much concerned in the determination of the so-called
preformed or cytoplasmic characters as any others." This conservative
view is certainly supported by the most convincing evidence and serves
to strengthen our contention that somatic mutation, war-weary though
the term may be, is still adequate to explain neoplastic disease without
the addition of parasitic, foreign, or infectious agents.

Mammary Tumor Inciter (Milk Agent)

One of the most dramatic experimental studies of recent years is the
one which has to do with an extrachromosomal factor described by
Murray and Little (115) which governs the susceptibility to spontane-
ous mammary cancer in mice. This was shown by Bittner (116) to be
a factor transmitted to the young by the maternal milk. The data on this

250 C. p. RHOADS

factor are so numerous and the results so well defined that reference to
the original material is required for an adequate consideration of the
subject.

The reports of Green (117) have been particularly cogent in invok-
ing studies of the milk agent in support of a virus theory of cancer. The
theory is based clearly upon a concept of the disease as due to a "parasite
entirely foreign to normal mouse tissue. The origin of viruses from
microbes and their nature as highly adapted parasites are reasonable
concepts and correspond to the course of parasitism and the nature of
highly adapted parasites as observed throughout the biological world."

The milk agent is reported as having the following characteristics:
It is present in the milk secreted throughout the lactation period, and in
blood and normal tissue as well as in the spontaneous mammary cancer,
originally or after serial passages by nursing during the precancerous
stage. When passed serially by injection through highly susceptible but
not infected mice, the power to cause tumor declined and was lost by
the third passage (118). Dilute suspensions are more active than con-
centrated ones (119). The agent is as active by injection as when given
by mouth and resists drying, filtration, lyophilization, and treatment
with glycerine. It is heat labile and destroyed by extremes of pli. Most
of the activity is sedimented by centrifugation at 40,000 r.p.m. (iio,-
000 g. ) for one hour. These characteristics, like the ones shown for the
antigen in the Brown-Pearce tumor, are the ones of a large protein
molecule.

The studies of Andervont and Bryan (120) and of Green (12) show
conclusively that the milk agent is active as an antigen when centrifu-
gates of tumor tissue, collected between runs at 15,000 g. and 95,000 g.,
are injected into rats and rabbits. None of the animals given a mixture
of immune serum and agent developed tumors, whereas a large propor-
tion of the controls which received mixtures of the agent with normal
mouse tissue antiserum, normal rat serum, or saline did so. These re-
sults are interpreted as evidence that "the agent is a virus of exogenous

origin."

An equally striking experiment has been reported by Green (13) on
the toxic properties of the mouse milk agent antiserum for the mouse
mammary cancer cells. Samples of a suspension of mouse mammary
cancer cells were mixed with the agent antiserum, normal breast tissue
antiserum, normal rabbit serum and saline solution, and injected into
susceptible mice. None of the animals which received the mixtures of

NEOPLASTIC ABNORMAL GROWTH 25I

cancer cells and cancer agent antiserum developed tumor, although
nearly all the control animals did so.

The findings of the two studies just discussed indicate beyond doubt
that "mouse cancer cells are antigenically different from normal cells."
It is also stated that "the same antiserum will neutralize a virus and
inactivate the cancer cells stimulated by that virus."

The studies of antibody to the milk agent are similar to the prior
work of Kidd ( 121 ) on the Brown-Pearce tumor of rabbits, from which
no agent causative of cancer can be recovered. Kidd proved that an anti-
body which appears in the blood of some rabbits implanted with that
tumor or injected with cell-free extracts of it is capable of suppressing
the growth of the transplanted tumor cells. It was shown, however, that
the growth frequently regressed in animals which had failed to develop
antibody and, conversely, rare animals showed progressive growth of
the tumor in spite of a high titer of antibody. The sera of hosts in which
the growth had regressed did not suppress the subsequent growth of
the tumor cells by in vitro contact unless they contained specific anti-
body.

The admirable discussion of Kidd (52) should be consulted as a
reference review of the immunological factors concerned with neo-
plasms. He calls attention to the evidence that the fate of transplanted
tumor cells is controlled by multiple genetic factors which may decrease
during successive transplantation. The usual generalization, previously
mentioned, is that the transplanted tissues grow if the hosts have anti-
gens in common with the tumor cells and regress in hosts lacking the
antigens and so developing antibodies. The failure to transmit resistance
passively and the profound differences between the solid immunity
which follows the regression of transplanted tumor and the transient
resistance set up by inoculations of tumor material are striking.

It is unfortunate that in the studies of rabbit tumors concerned with
specific proteins (viruses) the material is of doubtful genetic homoge-
neity. The extensive studies of Bittner on the factors which concern
the susceptibility of mice to mammary tumors are revealing and should
be consulted (122). From these it is concluded that, whereas in the
strains of mice studied little or no cancer occurred in the absence of the
milk factor, cancer need not occur in its presence unless two other fac-
tors, the inherited hormonal influence and the inherited susceptibility,
are also present.

Bittner and Huseby (122) have contributed beautifully controlled

252 C. p. RHOADS

and most illuminating studies on the factors which bear on the develop-
ment of mammary cancer in mice. Six lines of animals were used:

1. Group A stock lacking hormone influence. These mice are of low
cancer incidence in virgins and high in breeding females. They carry
the milk factor as shown by the occurrence of tumors in breeders and
transmit it to their young as shown by a 92.5% incidence in nursed Fi
virgins. They clearly have the factor of susceptibility, or there would
be no tumors even with the combination of forced breeding and milk
influence as shown by the C57 Blacks. The lack of hormone influence
is proven by a 3.9% incidence in virgins without the hormone influence
provided by breeding. Multiple genetic factors appear to be required to
produce the inherited sensitivity to or power to develop the hormone in-
fluence,

2. C57 Black stock of low incidence, lacking inherited susceptibility.
These animals even when provided with milk factor by foster nursing
by A strain females and with hormonal stimulation through forced
breeding show only 10.3 % of cancer. The absence of inherited suscepti-
bility is apparent. Data obtained from crossing the A stock with the
C57 Black animals indicate that the inherited susceptibility is deter-
mined by multiple genetic factors, one of which is the gene for brown
coat color.

3. Zb and Ax animals of low incidence due to lack of milk factor.
These are mice of the high incidence A and Z strains fostered by CBA
mothers lacking the factor. Both the inherited susceptibility and hor-
mone influence are shown by a high incidence when the animals are
supplied milk factor and force bred.

4. Ax strain virgins of low incidence lacking both hormone stimula-
tion and milk influence. If given the latter only, as by nursing their own
mothers, they are, of course, the mice of Group i, still not liable to
cancer until provided with hormone influence by forced breeding.

5. Z strain virgins of high incidence due to the possession of all three
factors. AZFi hybrids also fall in this group.

6. Fi hybrids from crossing the A strain lacking hormone influence
and Z strain possessing it. A high incidence strain even in the virgin
females.

From the data presented in these experiments the fact is apparent
that three factors operate in causing breast cancer in mice and any one
of the three can be completely determining in its effects. Certainly two
of these factors are completely controlled by genes, the inherited sus-
ceptibility and the inherited hormone influence. Since from Green's data

NEOPLASTIC ABNORMAL GROWTH 253

the milk factor dies out by the third passage through animals not gen-
erating it, the likelihood is real that this factor also is gene controlled
and indeed has certain similarities to the Kappa factor described for
Paramecia by Sonneborn.

Huseby and Bittner (123) provide confirmation for the results just
described in their studies of the architecture of the mammary glands in
the same strains of animals. Comparisons were made between the three
groups each lacking one of the essential factors for the development of
breast cancer and with mice possessing all three factors. Precancerous
nodules of alveolar hyperplasia were found to occur only in animals
with a high incidence of cancer irrespective of which of the essential
factors was lacking. It is concluded that "the same three factors that
are etiologically important for the development of mammary cancer are
necessary for the development of precancerous alveolar hyperplasia."

If a virus is to be defined as an agent which is continually propagated
by living cells genetically qualified to manufacture it and which is regu-
larly associated with characteristic properties of those cells, the milk
factor and other substances found in tumors certainly qualify. If, how-
ever, a virus is to be defined as a parasite entirely foreign to its host
tissue, it must meet additional requirements. It should be capable of
causing disease in several species, it should be more independent of
genetic factors in its pathogenicity, it should be antigenic in the species
which is its natural host, it should be capable of being propagated in
tissue culture or on the chorio-allantoic membrane of the chick embryo,
and the disease caused by it should be contagious. It is clear that some
tumor cells contain substances which are in size and composition like
viruses and also like genes. One, and the only one clearly etiologic for
cancer in mammals, is dependent on genes for its continued production
and is equally dependent on genes for its ability to produce its charac-
teristic effects. It is also apparent from the studies of Claude (124) that
the protoplasm of mammalian cells, unaffected by viruses or any dis-
order, contains complexes of protein of similar size and composition.
Immunologic studies on these particles would be of great interest, and
particularly desirable would be the demonstration of their control by
genes or alteration of their composition by substances which are known
to change the structure and function of genes, or, particularly, the abil-
ity of these particles to alter genes.

There can be no doubt that cells contain substances which affect other
cells and can be propagated in turn by the recipients. Ample evidence
for the existence of substances of this type has been advanced from

254 C. p. RHOADS

work in botany and in embryology. The substances described as neo-
plastic viruses have not been shown to possess any single characteristic
which is not shared by substances which are not infectious parasitic
agents. Particle size, antigenicity, protein or even nucleo-protein com-
position, autocatalytic propagation, inactivation by heat, and changes
of pH are ubiquitous. Transmission by the maternal cytoplasm or by
nursing are all shared by other, non-viral, substances. With all the vast
amount of experimentation and highly illuminating, indeed brilliant,
results, perhaps transcending in importance those of any other field of
biology, the expression coined by Murphy for the Rous agent of "trans-
missible mutagen" seems to cover the facts more adequately than any
other.

Summary

A discussion has been presented of certain mechanisms known to
cause self-perpetuating alterations of cells in general. This has been
done because of the evidence that neoplastic abnormal growth is due to
unique, inherited properties of its constituent cells. It is hoped that from
a scrutiny of a variety of data, in part apparently irrelevant to the can-
cer problem, certain lines of investigation will stand out as potentially
profitable in the future.

Evidence has been advanced that the neoplastic process is a distinc-
tive, characteristic sort of abnormal growth, malignant as well as autono-
mous, something more than a quantitative deviation from the normal
rate and extent of differentiation.

Reference has been made to the factors which control normal growth
and changes in cells to ascertain to what extent principles normally
operative can be invoked to explain the acquisition of abnormal prop-
erties. It is clear that the constituents of cells which vary either normally
or pathologically are under the control of genes to a very great extent.
Gene mutation, either spontaneous or induced, is held to be able to effect
the types of changes of cell composition and function, normal and ab-
normal, which have been described. This principle of genie control is
invoked as applicable to self-perpetuating cytoplasmic factors and to
maternal influences, as well as to cell constituents known to be the direct
product of the gene. The evidence that the ability of substances formed
by the organs of internal secretion to affect the growth and differentia-
tion of cells depends upon the inherited properties of the target tissue
has been discussed.

Attention has been called to the correlation between the ability of cer-

NEOPLASTIC ABNORMAL GROWTH 255

tain agents to cause (a) gene mutations in sexual forms, (b) changes
with similar results in vegetative forms or bacteria, not susceptible to
study by the conventional methods of the geneticists, and (c) cancer in
avian and in mammalian tissue. Whereas this correlation may of course
be pure coincidence, it would be surprising if the basic mechanisms in
the three types of cells were not similar or perhaps identical.

The induction of certain forms of neoplastic abnormal growth by
filtrates of the affected cells, free of the whole cells as such, poses a par-
ticularly nice problem to the investigator. In some instances, such as that
of the Rous agent, the active material, proven conclusively to be related
closely to constituents of the normal cells which it affects, can be recov-
ered regularly from the resultant neoplasm. In other cases, as for ex-
ample with the cancer which arises on the Shope virus-induced papil-
loma, this is not true. Here the virus, though etiologically related to the
neoplasm, disappears, conceivably incorporated into the affected cells,
and a new protein constituent arises, immunologically quite distinct
from the inoculated material and from any substance present in the host
before inoculation.

Whereas self -perpetuating cytoplasmic entities are well-known in
lower forms, and some of these, as the Kappa factor of Paramecia and
perhaps certain plant constituents, can be passed from the cytoplasm of
one cell to that of another, they tend to be dependent to some extent
upon the genes of the host. It would be of great interest if a protein sub-
stance could be isolated which, when taken in by the cytoplasm of a cell,
could be proved to substitute either for a normal substrate or for a nor-
mal enzyme, and as a result would alter the gene to produce a wholly
new constituent. Such a compound might disappear and so not be re-
coverable, but it would have exerted its effect in causing a mutation.
There seems to be no fundamental reason why such a mechanism should
not exist, and indeed it is wholly possible that this may be one route
operative in the production of cancer by physical agents. Dale has shown
that x-rays need not affect the gene directly, but may do so indirectly
through their effects on enzyme systems. Kensler and Rhoads have
proven that the naturally occurring metabolites of one carcinogen af-
fect profoundly certain enzymes of the target tissue.

The studies on the Brown-Pearce tumor have been referred to as pro-
viding new evidence for the presence of antigenic constituents in a
transplantable neoplasm. The principle is extended by the preparation
of specific antibodies against the milk factor in mice and in the V2 carci-
noma of rabbits. There has been little doubt for many years that anti-

256 C. p. RHOADS

bodies can be induced in distantly related hosts by the inoculation of
tumor material. From the evidence, however, cancer of spontaneous
origin is not productive of antibodies capable of controlling the growth.
Some other factor appears to play a role in immunity, and a particularly
mysterious and important one. Intensive study is indicated of the anti-
genic constituents of the neoplasms induced by the conventional physical
and chemical agents as well as of those neoplasms known to follow the
inoculation of special proteins.

There is believed to be no fundamental incompatibility between the
thesis that some cancers are due to self -perpetuating, perhaps cytoplas-
mic, protein components, and the contention that most are the result of
gene change and represent somatic mutations. The dividing line between
these two mechanisms has become so indistinct, so much interrelation-
ship between nuclear genes and cytoplasmic constituents has been proved
to exist, and such a vast number of different kinds of self -perpetuating
entities have been described, that an agent can be found to suit almost
any taste.

Some cancers can be transmitted by filtrates and continue to produce
the active material, some result from filtrates but yield none of the eti-
ologic substance, some may conceivably result from altered enzymes,
either directly, or indirectly through their interaction with genes, and
some from direct gene hits with the production of standard mutations.
In spite of all these possibilities and the immense number of literary
and scientific contributions to the subject, there appears no conclusive
proof that any mammalian cancer is caused by an intracellular parasite
completely foreign to the host.

Reference has been made to the generalized effects upon cellular dif-
ferentiation, including the production of cancer, which follow the sys-
temic administration of certain compounds with hormone activity and
also of special chemical substances not known to have hormone effects.
The possibility has been advanced that perhaps one step in the produc-
tion of neoplastic abnormal growth may be the chronic exposure of cells
to an intolerable environment, with consequent death or defensive vari-
ation and the development of protective mechanisms. Such an environ-
ment might well allow the variants which occur spontaneously, if they
are equipped with protection, to persist and overgrow the normal cells.
Since the standard controls of growth are established to govern only
standard cells, certain variant cells might not be susceptible to them,
and, if so, might grow without restraint. This concept of neoplastic
growth will of course be immediately opposed by those who envision

NEOPLASTIC ABNORMAL GROWTH 257

the neoplasm only as composed of irreversibly altered constituents
which continue to reproduce themselves faithfully forever. Certainly
much evidence favors the view of permanent cellular change, but there
exist data which require the inclusion of an environmental factor to ex-
plain some neoplasms. Furthermore, all cancers need not arise and con-
tinue to grow because of exactly the same mechanism. Green (117)
has called attention to the fact that two neoplasms, histologically iden-
tical, may differ wholly in the degree to which they are autonomous. In
man, too many regressions of breast and prostate cancer occur to allow
the complete elimination of the general control of cellular environ-
mental factors as a possible participant in the neoplastic process.

From the vast collection of factual material existent on neoplastic
growth, one fact in particular stands out as worthy of special mention.
This is that further data are urgently required on the comparative com-
positions of normal and neoplastic cells. These data may concern genes,
enzymes, proteins, or any other component. They may be derived from
direct analytical methods, from tracer studies, or from the different
susceptibilities of the two cell types to toxic agents. Precise informa-
tion of this type is likely to advance significantly our knowledge of the
problem, somewhat more than discussions of whether the etiologic
agent is a component foreign to the host or of host origin. It is impor-
tant to prove only that it is different from any normal component and
to learn how to accomplish its destruction.

REFERENCES

1. Papanicolaou, G. N., and H. F. Traut. Diagnosis of Uterine Cancer by
the Vaginal Smear. Commonwealth Fund, New York, 1943.

2. Brues, A. M., M. M. Tracy, and W. E. Cohn. Nucleic Acids of Rat
Liver and Hepatoma: Their Metabolic Turnover in Relation to
Growth. /. Biol. Chem., 755, 619, 1944.

3. Seidlin, S. N., and L. D. MarineUi. Therapeutic Effect of Radio-
active Iodine on Adenocarcinoma of the Thyroid with Functioning
Metastases. Bulletin N.Y. Academy of Medicine, 21, 440, 1945.

4. Haddow, A. On Secondary Colony Development in Bacteria and an
Analogy with Tumour Production in Higher Forms. Acta de I'Union
Internationale contre le cancer, 2, 376, 1937.

5. Burrows, H. Biological Action of Sex Hormones. Cambridge Uni-
versity Press, Camljridge, 1945.

6. Allen, E., and W. U. Gardner. Cancer of Cervix of Uterus in Hybrid
Mice Following Long-continued Administration of Estrogen. Cancer
Research, i, 359, 1941.

258 C. p. RHOADS

7. Needham, J. Problems of Chemical Embryology. Fiziol. Zhur., ip,

241, 1935-

8. Friedewald, W. F., and J. G. Kidd. Induced Antibodies that React
in Vitro with Sedimentable Constituents of Normal and Neoplastic
Tissue Cells — Presence of the Antibodies in the Blood of Rabbits
Carrying Various Transplanted Cancers. /. Exp. Med., 82, 21, 1945.

9. Kidd, J. G. Serological Studies in Relation to the Problem of Tumor
Causation. /. Bacterial., jp, 349, 1940.

10. Kidd, J. G. A Distinctive Substance Associated with the Brown-
Pearce Rabbit Carcinoma. I. Presence and Specificity of the Sub-
stance as Determined by Serum Reactions. /. Exp. Med., Ji, 335,
1940.

11. Kidd, J. G. Distinctive Constituents of Tumor Cells and their Possible
Relations to the Phenomena of Autonomy, Anaplasia, and Cancer
Causation. Cold Spring Harbor Symposia on Quantitative Biology,
1947, 11^ 94. 1946.

12. Green, R. G., M. M. Moosey, and J. J. Bittner. Antigenic Character
of the Cancer Milk Agent in Mice. Proc. Soc. Exp. Biol. Med., 61,
115, 1946.

13. Green, R. G. Cytotoxic Property of Mouse Cancer Antiserum. Proc.
Soc. Exp. Biol. Med., 61, 113, 1046.

14. Warburg, O. The Metabolism of Tumors. (Trans, by F. Dickens.)
Constable and Co., Ltd., London, 1930.

15. Greenstein. J. P. Enzymes in Normal and Neoplastic Animal Tissues.
A.A.A.S. Research Conference on Cancer, The American Associa-
tion for the Advancement of Science, Washington, D.C., 1945, p. 192.

16. Potter, V. R. Advances in Ensymology ; Biological Energy Trans-
formations and the Cancer Problem. Interscience Publishers, New
York, 1944, pp. 201-250.

17. Roskelly, R. C, N. Mayer, B. N. Horwitt, and W. T. Salter. Studies
in Cancer; VIL Enzyme Deficiency in Human & Experimental Can-
cer. /. Clin. Inves., 22, 743, 1943.

18. Haddow, A. Transformation of Cells and Viruses. Nature, 154, 194,
1944.

19. Muller, H. J. Induced Mutations in Drosophila. Cold Spring Harbor
Symposia on Quantitative Biology, 1941, p, 151.

20. Biesele, J. J. Ribonucleic Acid and Heterochromatin in Epidermal
Carcinogenesis. Cancer Research, 4, 737, 1944.

21. Sturtevant, A. H. High Mutation Frequency Induced by Hybridiza-
tion. Proc. Nat. Acad. Sci., ^5, 308, 1939.

22. Demerec, M. Unstable Genes in Drosophila. Cold Spring Harbor
Symposia on Quantitative Biology, 1941, p, 145.

23. Plough, H. H. Spontaneous Mutability in Drosophila. Cold Spring
Harbor Sympo.Ha on Quantitative Biology, 1941, p, 127.

24. Rhoades, M. M. The Genetic Control of Mutability in Maize. Cold
Spring Harbor Symposia on Quantitative Biology, ip4i, p, 138.

NEOPLASTIC ABNORMAL GROWTH 259

25. Beadle, G. W. Biochemical Genetics. Cheni. Rev., 57, 15, 1945 (quote
from p. 23).

26. Landsteiner, K. The Specificity of Serological Reactions. Charles C.
Thomas, Springfield, Illinois, 1936.

27. Wiener, A. S. Blood Groups and Transfusion. Charles C. Thomas,
Springfield, Illinois, 1943.

28. Tatum, E. L., and G. W. Beadle. Biochemical Genetics of Neuro-
spora. Ann. Missouri Botanical Garden, j-?, 125, 1945.

29. MacDowell, E. C. What and Why is Leukemia? A Medico-Biological
Study on Mice. Carnegie Inst. Wash., Puh. 301, 353, 1938.

30. Sax, K. Types and Frequencies of Chromosomal Aberrations Induced
by X-rays. Cold Spring Harbor Symposia on Quantitative Biology,

1941, 9, 9Z-

31. Glucksmann, A. Quantitative Histological Analysis of Radiation-
Effects in Human Carcinomata. British Med. Bull., 4, 26, 1946.

32. Gray, C. H., and E. L. Tatum. X-ray Induced Growth Factor Re-
quirements in Bacteria. Proc. Nat. Acad. Sci., 30, 404, 1944.

33. Lea, D. E. The Action of Radiations on Viruses and Bacteria.
British Med. Bull., 4, 24, 1946.

34. Gowen, J. W. Mutation in Drosophila, Bacteria, and Viruses. Cold
Spring Harbor Symposia on Quantitative Biology, 194 1, p, 187.

35. Dale, W. M., W. J. Meredith, and M. C K. Tweedie. Mode of
Action of Ionizing Radiations on Aqueous Solutions. Nature, 131,
280, 1943.

36. Lea, D. E., K. M. Smith, B. Holmes, and R. Markham. Direct and
Indirect Actions of Radiation on Viruses and Enzymes. Parasitology,
36, 1 10- 1 18, 1944.

2)7 . Henshaw, P. S. Experimental Roentgen Injury ; IV. Effects of Re-
peated Small Doses of X-rays on Blood Picture, Tissue Morphology
and Life Span in Mice. /. Nat. Cancer Inst., 4, 513-522, 1944.

38. Furth, J., and M. C. Boon, Liver Changes Associated with a Trans-
plantable Granulosa Cell Carcinoma in Mice. Proc. Soc. Exp. Biol.
Med., 58, 112, 1945.

39. March, H. C. Leukemia in Radiologists. Radiology, 43, 275-278,
1944.

40. Stadler, L. J. The Comparison of Ultraviolet and X-ray Effects on
Mutation. Cold Spring Harbor Symposia on Quantitative Biology,
1941, p, 168.

41. HoUaender, A., and C. W. Emmons. Wavelength Dependence of
Mutation Production in the Ultraviolet with Special Emphasis on
Fungi. Cold Spring Harbor Symposia on Quantitative Biology, 1941,

P. 179-

42. Greenstein, J. P., and W. V. Jenrette. Physical Changes in Thymo-
nucleic Acid Induced by Proteins, Salts, Tissue Extracts, and Ultra-
violet Irradiation. Cold Spring Harbor Symposia on Quantitative
Biology, 1941, p, 236.

260 C. p. RHOADS

43. Stanley, W. M. Viruses: Currents in Biochemical Research. Inter-
science Publishers Inc., New York, 1946, p. 13.

44. MacKenzie, K., and H. J. Muller. Mutation Effects of Ultraviolet
Light in Drosophila, Proc. Roy. Soc, Ser. B., I2g, 491, 1940.

45. Beadle, G. W., and E. L. Tatum. Neurospora. II. Methods of Pro-
ducing and Detecting Mutations Concerned with Nutritional Re-
quirements. Am. J. Bot., 32, 678, 1945.

46. Blum, H. Physiological Effects of Sunlight on Man. Physiol. Rev.,

25, 483-530. 1945-

47. Swanson, C. P. X-ray and Ultraviolet Studies on Pollen Tube Chro-
mosomes. I. The Effect of Ultraviolet (2537 A) on X-ray Induced
Chromosomal Aberrations. Genetics, 2p, 61, 1944.

48. Tatum, E. L. Personal communication.

49. Guyer, M. F., and E. A. Smith. Further Studies on Inheritance of
Eye Defects Induced in Rabbits. /. Exp. Zool., 38, 449, 1924.

50. Sturtevant, A. H., and G. W. Beadle. An Introduction to Genetics.
W. B. Saunders Company, Philadelphia, 1939.

51. Emerson, S. Induction of Mutations by Antibodies. Proc. Nat. Acad.
Sci., 30, 179-183, 1944-

52. Kidd, J. G. Suppression of Growth of Brown-Pearce Tumor Cells
by a Specific Antibody; With a Consideration of the Nature of the
Reacting Cell Constitutent. /. Exp. Med., 83, 227-250, 1946.

53. Auerbach, C, M. Y. Ansari, and J. M. Robson. (Unpublished British
Work done under military security.)

54. Tatum, E. L. Personal communication.

55. Bawden, F. C, and N. W. Perie. The Effect of HN2 on Tobacco
Mosaic and Tomato Bushy Stunt Viruses. British Military Extra-
Mural Research, June 1944, Z4432.

56. Smith, H. Summary Report on the Systemic Pharmacology and
Pathology of the Sulfur and Nitrogen Mustards. OSRD Report 6325.

57. Bodenstein, D., and R. Gillette. Reports of the Medical Division
Chemical Warfare Service, U.S. Army, 1945.

58. Karnofsky, D. A., J. H. Burchenal, R. A. Ormsbee, I. Cornman, and
C. P. Rhoads. Experimental Observations on the Use of the Nitrogen
Mustards in the Treatment of Neoplastic Disease. Approaches to Tu-
mor Chemoflierapy. A.A.A.S. Washington. D.C., 1947.

59. Stowell, R. E., and Z. K. Cooper. Relative thymonucleic acid content
of human normal epidermis, hyperplastic epidermis and epidermoid
carcinomas. Cancer Research, 5, 295-301, 1945.

60. Burk, D., and R. J. Winzler. The Biochemistry of Malignant Tissue.
Annual Review of Biochemistry, 13, 487-532, 1944.

61. Suntzeff, v., and C. Carruthers. The Effect of Methylcholanthrene
upon Epidermal Sodium and Calcium. Cancer Research, 3, 431-433,

1943-

62. Tatum, E. L. Personal communication.

63. Mottram, J. C. Abnormal Paramecia Produced by Blastogenic Agents

NEOPLASTIC ABNORMAL GROWTH 261

and Their Bearing on the Cancer Problem. Cancer Research, i, 313-

323, 1941-

64. Strong, L. C. The Induction of Germinal Mutations by Methyl-

cholanthrene. Cancer Research, <5, 501, 1946.

65. Dobriner, K., C. P. Rhoads, and G. I. Lavin. The Spectroscopic
Study of Biological Extracts; II. Detection, Isolation and Biological
Effects of Metabolites of 1,2,5,6-Dibenzanthracene. Cancer Research,
2, 95-107, 1942.

66. Boyland, E., and A. A. Levi. Metabolism of Polycyclic Compounds ;
III. Anthryl-mercapturic acid. Biochem. Journal, 50, 1225-1227,
1936.

67. Jones, R. N., C. E. Dunlap, and C. J. Gogek. The Spectrographic
Analysis of Carcinogenic Hydrocarbons & Metabolites; IV. Elimina-
tion of 1,2,5,6-dibenzanthracene from the rat. Cancer Research, 4,
209-217, 1944.

68. Weigert, F. 22nd Annual Report, British Empire Campaign, 1945,
p. 38.

69. Kensler, C. J., and C. P. Rhoads. Biochemical Studies of Chemical
Carcinogenesis, with a Preliminary Note on the Susceptibility to
Chemicals of Certain Neoplasms, Both Animal and Human. A.A.A.S.
Research Conference on Cancer, p. 170. Washington, D.C., 1945.

70. Giese, J. E., J. A. Miller, and C. A. Baumann. The Carcinogenicity of
w'-Methyl-/'-Dimethylaminoazobenzene and /)-mono-Methylamino-
azobenzene. Cancer Research, 5, 337-340, 1945-

71. Beadle, G. W. Currents in Biochemical Research; the Gene in Bio-
chemistry. Interscience Publishers, New York, 1946.

72. Lindegren, C. C, S. Spiegelman, and G. Lindegren. Mendelian In-
heritance of Adaptive Enzymes in Yeast. Proc. Nat. Acad. Sci.,
30, 346, 1944.

JT)- Dubos, R. J. The Bacterial Cell in its Relation to Problems of Viru-
lence, Immunity, and Chemotherapy. Harvard University Press,
Cambridge, 1945.

74. Lewis, I. M. Bacterial Variation with Special Reference to Behavior
of Some Mutable Strains of Colon Bacteria in Synthetic Media. /.
Bad., 28, 619, 1934.

75. Schnitzer, R. J., L. J. Camagni, and M. Buck. Resistance of Small
Colony Variants (G-forms) of a Staphylococcus Towards the Bac-
teriostatic Activity of Penicillin. Proc. Soc. Exp. Biol. Med., 53, 75,

1943-

76. Griffith, F. The Influence of Immune Serum on the Biological Prop-
erties of Pneumococci. Repts. on Pub. Health Med. Subjects, Or.
Britain, No. 18, pp. 1-13, 1923.

yy. Tatum, E. L. Induced Biochemical Mutations in Bacteria. Cold
Spring Harbor Symposia on Quantitative Biology, 1946, 11, 278.

78. Medawar, P. B. A Second Study of the Behavior and Fate of Skin
Homografts in Rabbits. /. Anat., /p, 157, 1945.

79. Little, C. C, and P. A. Garer. The Genetics of Cancer in Mice, In

262 C. p. RHOADS

Genetics of the Mouse. Griineberg, H., Cambridge University Press,

1943-

80. Furth, J., M. C. Boon, and N. Kaliss. On the Genetic Character of

Neoplastic Cells as Determined in Transplantation Experiments.
Cancer Research, 4, i, 1944.

81. MacDowell, E. C, J. S. Potter, and M. J. Taylor. The Influence of
Transplantation upon Immunological Properties of Leukemic Cells.
Proc. Nat. Acad. Set., 25, 416, 1939.

82. Barrett, Morris K. The Influence of Genetic Constitution upon the
Induction of Resistance to Transplantable Mouse Tumors. Jour. Nat.
Cancer Inst., i, 387, 1940-1941.

83. Lederberg, J. Personal communication to Dr. Stock.

84. Gordon, M., and G. M. Smith. Production of Melanotic Neoplastic
Disease in Fishes by Selective Matings ; IV. Genetics of Geographical
Species Hybrids. Am. Jour. Cancer, S4, 543-565, 1938.

85. Whitaker, D. M. Physical Factors of Growth. Growth Supplement,
1940, p. 75.

86. Little, C. C. Hybridization and Tumor Formation in Mice. Proc.
Nat. Acad. ScL, 23, 452-455. i939-

87. Claude, A., and E. F. Fullam. Electron Microscope Study of Isolated
Mitochondria; Method and Preliminary Results. /. Exp. Med., 81,

51. 1945-

88. Snell, G. D. Biology of the Laboratory Mouse. The Blakiston Co.,

Philadelphia, 1941.

89. Rous, P. A Sarcoma of the Fowl Transmissible by an Agent Sepa-
rable from the Tumor Cells. 7. Exp. Med., 13, 397, 191 1.

90. Beadle, G. W., and E. L. Tatum. Neurospora. II. Methods of Pro-
ducing and Detecting Mutations Concerned with Nutritional Re-
quirements. Am. J. Botany, 32, 678, 1945.

91. EUermann, V. The Leucosis of Fowls and Leucemia Problems.
Gyldenal, London, 1921.

92. Waters, Nelson F. Personal communication.

93. Berry, G. P., and H. M. Dedrick, A Method for Changing the Virus
of Rabbit Fibroma (Shope) into that of Infectious Myxomatosis
(Sanorelli). /. Bact., 31, 50, 1936.

94. Milford, J. J., and Duran-Reynals. Growth of a Chicken Sarcoma
Virus in the Chick Embryo in the Absence of Neoplasia. Cancer Re-
search, 3, 578, 1943.

95. Claude, Albert, and Alexandre Rathen. Properties of the Causative
Agent of a Chicken Tumor. XIV. Relation between a Tumor Nu-
cleoprotein and the Active Principle. /. Exp. Med., /i, 619, 1940.

96. Rous, Peyton, and J. W. Beard. A Virus-Induced Mammalian
Growth with the Characters of a Tumor (Shope Rabbit Papilloma).
I. The Growth on Implantation within Favorable Hosts. /. Exp. Med.,
60, 701, 1934. II. Experimental Alterations of the Growth on the
Skin: Morphological Considerations: the Phenomena of Retrogres-

NEOPLASTIC ABNORMAL GROWTH 263

sion. /, Exp. Med., 60, 72^, 1934. III. Further Characters of the
Growth: General Discussion. /. Exp. Med., 60, 741, 1934. The Pro-
gression to Carcinoma of Virus-Induced Rabbit Papillomas (Shope).
/. Exp. Med., 62, 523, 1935.

97. Shope, R. E. A Transmissible Tumor-like Condition in Rabbits. /.
Exp. Med., 56, 793, 1932.

98. Kidd, J. G., and P. Rous. The Carcinogenic Effect of a Papilloma
Virus on the Tarred Skin of Rabbits. 11. Major Factors Determining
the Phenomenon: The Manifold Effects of Tarring. /. Exp. Med.,
68, 529, 1938.

99. Kidd, J. G. The Masking Effect of Extravasated Antibody on the
Rabbit Papilloma Virus (Shope). /. Exp. Med., 70, 583, 1939.

100. Kidd, J. E. and P. Rous. Cancers Deriving from the Virus Papil-
lomas of Wild Rabbits under Natural Conditions. /. Exp. Med., 71,
469, 1940.

loi. Friedewald, W. F. Identity of "Inhibitor" and Antibody in Extracts
of Virus-Induced Rabbit Papillomas. Proc. Soc. Exp. Biol. Med.,

42, 330, 1939-

102. Andrewes, C. H. Evidence for the Presence of Virus in a Non-filter-
able Tar Sarcoma of the Fowl. /. Path. Bad., 43, 23, 1936.

103. Fouldes, L. Observations on Non-filterable Fowl Tumors. The Pro-
duction of Neutralizing Sera against Filtrates of Rous Sarcoma.
I. By Non-infective Extracts of a Sarcoma Induced by i :2 :5 :6-
Dibenzanthracene. Atn. J. Cancer, 31, 404, 1937.

104. Sturtevant, A. H. Inheritance of Direction of Coiling in Limnaea.
Science, j8, 269, 1923.

105. Moore, J. A. Developmental Rate of Hybrid Frogs. /. Exp. ZooL,
86, 405, 1941.

106. Harvey, E. B. Parthenogenetic Merogony or Cleavage without Nu-
clei in Arbacia Punctulata. Biol. Bidl, 71, loi, 1936. Development of
Half-eggs of Chaetopterus Pergamentaceus with Special Reference
to Parthenogenetic Merogony. Ihid., 76, 384, 1939.

107. Michaelis, P. I. Uber Reziprok Verschiedene Sippen-Bastarde bei
Epilobium Hirsutum. Z. induktive Abstammungs -u. Vererhungs-
lehre, 78, 187, 1940. II. tJber die Konstanz des Plasmons der Sippe
Jena. Ihid., 78, 223, 1940. III. tJber die genischen Grundlagen der im
Jena Plasma auftretende Hemmungs-Reihe. Ihid., 78, 295, 1940. V.
tJber die Bedeutung der Genquantitat fur die Manifestation rezi-
proker Unterschiede. Ihid., 80, 429, 1942. VI. In welcher Weise sind
an der Manifestation der im Jena-Plasma auftretenden Entwicklungs-
tendenz die Gene dieser Sippe beteiligt? Ihid., 80, 454, 1942.

108. Kiihn, A. and E. Plagge. Pradetermination der Raupenaugenpigmen-
tierung bei Ephestia kuhniella Z. durch den Genotypus der Mutter
und durch arteigene und artfremde Implantate. Biol. Zentr., 57, 113,

1937-

109. Imai, Y. A Consideration of Variegation. Genetics, /j, 544-562,

1928.

264 C. p. RHOADS

no. Rhoades, M. M. Genie Induction of an Inherited Cytoplasmic Genie
Induction of an Inherited Cytoplasmic Difference. Proc. Nat. Acad.
ScL, 29, 327, 1943.

111. DuBuy, H. G., and Mark W. Woods. A Possible Common Mitochon-
drial Origin of the Variegational and Virus Diseases in Plants and
Cancers in Animals. A.A.A.S. Research Conference on Cancer,
Washington, D.C., 1945, p. 162.

112. Sonneborn, T. M. Gene and Cytoplasm. I. The Determination and
Inheritance of the Killer Character in Variety 4 of Paramecium
aurelia. Proc. Nat. Acad. ScL, 29, 329, 1943.

113. Sax, K. Diffusion of Gene Products. Proc. Nat. Acad. ScL, 28, 303,
1942.

114. Wilson, E. B. The Cell in Development and Heredity. 3rd edition,
Macmillan, New York, 1925.

115. Murray, W. S., and C. C. Little. Chromosomal and Extrachromo-
somal Influence in Relation to the Incidence of Mammary Tumors
in Mice. Am. J. Cancer, 57, 536, 1939.

116. Bittner, John J. Inciting Influences in the Etiology of Mammary Can-
cer in Mice. A.A.A.S. Research Conference on Cancer, Washington,
D.C., 1945. p. 63.

117. Green, Robert G. Virus Aspects of Carcinoma. Minnesota Medicine,
2Q, 277, 1946.

118. Green, R. G., M. M. Moosey, and J. J. Bittner. Serial Transmission
of the Milk Agent of Mouse Mammary Carcinoma. Proc. Soc. Exp.
Biol. Med., 61, 362, 1946.

119. Bittner, J. J. Characteristics of the Mammary Tumor Milk Agent
in Serial Dilution and Blood Studies. Proc. Soc. Exp. Biol. Med.,
59, 43, 1945.

120. Andervont, H. B., and W. R. Bryan. Properties of Mouse Mammary-
Tumor Agent. /. Nat. Cancer Inst., 5, 143-149, 1944.

121. Kidd, J. G. Suppression of Growth of the Brown-Pearce Tumor by
a Specific Antibody. Science, 99, 348, 1944.

122. Bittner, J. J., and R. A, Huseby. Relationship of the Inherited Sus-
ceptibility and the Inherited Hormonal Influence in the Development
of Mammary Cancer in Mice. Cancer Research, 6, 235, 1946.

123. Huseby, R. A., and J. J. Bittner. A Comparative Morphological Study
of the Mammary Glands with Reference to the Known Factors In-
fluencing the Development of Mammary Carcinoma in Mice. Cancer
Research, 6, 240, 1946.

124. Claude, Albert. Particulate Components of Cytoplasm. Cold Spring
Harbor Symposia on Quantitative Biology, 1941, 9, 263.

125. Earle, W. R. Production of Malignancy in vitro.

(A) I. Method of Cleaning Glassware. /. Nat. Cancer Inst., 4, 131,

1943-

(B) II. Photomicrographic Equipment. /. Nat. Cancer Inst., 4, 135,

1943-

NEOPLASTIC ABNORMAL GROWTH 265

(C) Earle, W. R., and L. R. Crisp. III. Microcinematographic
Equipment. /. Nat. Cancer Inst., 4, 14^, 1943.

(D) Earle, W. R., E. L. Schilling, T. H. Stark, N. P. Straus, M. F.
Brown, and E. Shelton. IV. The Mouse Fibroblast Cultures and
Changes Seen in the Living Cells. /. Nat. Cancer. Inst., 4, 165, 1943.

(E) Earle, W. R., A. Nettleship, E. L. Schilling, T. H. Stark. N. P.
Straus, M. F. Brown, and E. Shelton. V. Results of Injection of
Cultures into Mice. /. Nat. Cancer Inst., 4, 213, 1943.

(F) Nettleship, A., W. R. Earle, M. P. Clapp, and^E. Shelton. VI.
Pathology of Tumors Produced. /. Nat. Cancer Inst., 4, 229, 1943.

126. Daniel, G. E., R. R. Spencer, and Dorothy Calnan. Methylcholan-
thracene and the Environment of Paramecium. /. Nat. Cancer Inst.,
6, 157, 1945.

X. THE ADRENAL GLAND,
A REGULATORY FACTOR

BY C N. H. LONG'

Quel est I'usage des glandes surrenales?

— Academic des Sciences, Bordeaux, 17 16

SOME thirty years before the founding of Princeton University,
after due consideration of certain unexplored fields of anatomy
and physiology, the Academy of Sciences of Bordeaux offered a
prize for the best essay submitted on the function of the atrabiliary or
suprarenal glands. These organs had been described by Bartholomaeus
Eustachius in 1563; but in common with the easy practice of those
days, since nothing was known of their function, they were assigned
the lowly role of acting merely as supporting elements for their neigh-
boring structures. Even in this regard but little attention was paid then,
far less indeed than that lavished on their sister ductless gland, the
thyroid, of which it was said by Thomas Wharton in 1656 that its
function was "the filling up of the vacant spaces around the larynx,
thus contributing much to the beauty and rotundity of the neck, espe-
cially in females."

Since the Academy did not see fit to honor any of the contributors
with an award, the occasion might have passed unnoted and the record
lost to us were it not for the fact that the adjudicator for the Academy
was none other than Baron Montesquieu, the noted French jurist,
political writer and philosopher. Montesquieu, like many scholars of
his day, was equally well versed in the scientific as well as the human-
istic knowledge of the age. His skill as a writer and his knowledge of
contemporary anatomy and physiology were happily fused in the
learned and witty criticism which he later delivered to the Academy
( I ) . In this essay are set forth not only the reasons that prompted the
Academy to offer the prize but also those for the rejection of all the
contributions that had been submitted.

Let me quote a portion of his preamble in which the reasons for the
selection of this subject are set forth. Speaking of the functions of the
human body he says, "In this prodigious number of parts, of veins,
arteries, lymphatic vessels, cartilages, tendons, muscles, glands, one
cannot believe there could be any useless part. Everything contributes
to the well-being of the animated subject and if there is any organ

1 Department of Physiological Chemistry, Yale University Medical School.

THE ADRENAL GLAND 267

whose function is unknown to us we must be attacked by a noble un-
rest and seek to discover it. It is this which provoked the Academy
to select as a subject 'The Function of the Surrenal Glands or Atra-
biliary Capsules' and to encourage scientists to work on a problem
which, in spite of the researches of so many investigators, was still
completely new and seemed to have been so far more the object of their
desperation rather than of their understanding."

There is no need to quote all the devastating criticism heaped by
Montesquieu upon the unfortunates who had rashly accepted the offer
of the Academy to compete for this prize. Two quotations from his
criticism will suffice to illustrate what the hopeful contributors to the
literature of that day might anticipate from the critics of their papers.

"We have found an author who admits the existence of two kinds
of bile; one, the rougher kind, which is manufactured in the liver; the
other, more subtle, which is manufactured in the kidneys with the help
of the ferment which flows from the capsules through passages of
whose existence we are not aware and of whose existence we are even
threatened of never being aware. But as the Academy wishes to be
enlightened and not discouraged we do not stop at this system,

"Still another describes for us two little vessels which carry the
humours from the cavity of the gland into the vein which irrigates it.
This humour, which many experiments have led us to believe is alka-
line, is used according to this writer to give fluidity to the blood which
is coming out of the kidneys after having abandoned there the liquid
which composes urine. It must be said that this investigator has only
too good supporters for what he proposed : Sylvius, Manget and others
had expounded this theory before him. The Academy, which doesn't
like duplication and always desires new knowledge, which like an
avaricious person who, because of this cupidity wants always to acquire
new riches and seems to count as nothing what he already possesses,
has not awarded the palm to this system."

Yet Montesquieu could rise above mere destructive criticism. He
recognized the difficulty of the question raised by the Academy and the
inadequacy of the then existing knowledge of these organs. "Chance
[he writes] will perhaps some day achieve what this work could not."

And what has chance and scientific endeavor achieved in the past
230 years? Would Montesquieu now agree that the question posed by
the Academy has been answered or would we and our predecessors who
have worked on this problem again sit in trepidation as he delivered
judgment? While I would hesitate to give a categorical "Yes" to the

268 C. N, H, LONG

question asked so long ago by this French Academy, I believe that all,
including Montesquieu, would agree that our knowledge of the supra-
renal glands is now immeasurably greater, even though it is still far
from complete. The chance or chances to which Montesquieu looked for
future enlightenment were not soon seized by scientists, for it was not
until the period from 1855 to 1867 that another great Frenchman,
Claude Bernard, enunciated the two great generalizations that laid
not only the foundations of our present knowledge of the ductless
glands, but to use his own words established "the basis of general
physiology."

The first of these is nov/ known as the doctrine of "internal secre-
tion." In this it was pointed out that certain specialized groups of cells
or organs pour their secretions directly into the blood stream and that
such organs could thus be termed "ductless glands" to distinguish them
from that group whose products passed by ducts to an internal or ex-
ternal surface of the body. Curiously enough Bernard first used the
term "internal secretion" in connection with an organ not now generally
recognized as an endocrine gland. This was the liver, and he coined the
term to describe the capacity of this organ to "secrete" glucose directly
into the blood stream. However a few years later he included in this
category the adrenals and thyroid glands. To these and certain others
the term ductless or endocrine glands is now more generally applied.

Bernard's concept that the blood served as a transporting medium for
the products of secretion of certain organs and that these products
might be a foodstuff such as glucose or a more subtle agent such as a
hormone led in 1867 to the formulation of his general statement con-
cerning the relationship of the blood and body fluids to the preserva-
tion of the function or the organism. Bernard recognized that in the
higher animals all the cells and organs of the body exist in a fluid en-
vironment whose composition is rigidly maintained within limits that
are compatible with their existence. The importance of this concept
of the constancy of the internal environment to an understanding of
the physiology of the higher forms of life was not immediately appre-
ciated. It is only in our own time that it has been given its proper recog-
nition and this is largely due to the work and writings of Cannon,
Haldane, and Henderson. Briefly this doctrine states that not only is the
composition of the fluid environment of the cells kept within certain
narrow limits, but what is equally important, the organism strenuously
resists the operation of any factor or circumstance that acts to change
this composition.

THE ADRENAL GLAND 269

The physiological processes by which any significant change in the
composition of the body fluids is resisted or adjusted to the needs of
the moment are often termed compensatory mechanisms. These physio-
logical mechanisms are of many kinds : heat regulation by the skin and
lungs, the regulation of water, hydrogen ion and electrolyte concentra-
tions by the kidneys and lungs, and one which is the subject of the
present discussion. This is the activity of the endocrine glands which
regulate the composition of the fluid environment not only by the action
of their products on the excretory organs but also by the effect of these
secretions on the chemical processes within the cells themselves.

The factors that tend to distort the fluid composition of the body are
not only those imposed by disease, noxious substances and injuries.
The requirements of the organism for a continual replenishment of its
own material substance demands the ingestion of foodstuffs, minerals,
and water itself. It would be an unenviable kind of existence if these
had to be ingested at such slow rates that no significant changes
occurred in the composition of the fluids both within and without the
cells. It is common knowledge that we can ingest at one time consider-
able portions of our daily requirements of food, minerals, and water
without more than temporary dislocations in the composition of our
internal environment, so efficient are the bodily mechanisms that main-
tain this composition.

In addition to this capacity to assimilate at one time considerable
quantities of the necessities of life, the higher organisms also possess
means to support the level of certain essential constituents of the body
fluids at a time when none is being supplied from without. In the ab-
sence of such a function all the higher forms of life would have had
to remain in close proximity to an ever present food supply and this
would have greatly limited the colonization of the land and the possi-
bility of further evolutionary processes.

The vast amount of work carried out in the last fifty years on the
function of the endocrine glands has not only served to give an insight
into their individual mode of operation but has brought an increasing
consciousness of the fact that they act not as independent units but as
a correlated and integrated system that is closely associated with the
capacity of the organism to adapt itself to the varying conditions of
existence.

Thus we find that the endocrine glands not only regulate the metab-
olism of the organic and inorganic elements of the body but do so in

270 C. N. H. LONG

such a manner that their secretions are brought into operation in a way
best adapted to the needs of the moment and with consequences that
result in the preservation of the internal environment within certain
narrow hmits. In this regard the adrenal cortex is of particular perti-
nence, although it is not the only endocrine gland concerned in the
regulation of metabolism.

I have been asked to write particularly on the work of my colleagues
and myself on the function of the adrenal cortex.^ What we have done
represents only a small part of the many contributions that have been
made to this subject in recent years. Though it is quite impossible to
mention even briefly all these contributions, anything we have done has
been immeasurably helped by the work of our predecessors and con-
temporaries.

There are two further points I should mention in this general intro-
duction to the subject. The first is that, unknown to Montesquieu and
the French Academy, the adrenal glands really compose two endocrine
organs, the adrenal medulla and adrenal cortex. This paper is concerned
largely with the function of the latter, but it should be emphasized that
the methods used and the general hypotheses stated are common to a
study of any of the ductless glands. Furthermore the work done on the
adrenal cortex has revealed as clearly as in the case of any other gland
the integrative nature of the endocrine system.

The investigation of the function of any endocrine gland proceeds
along three main lines. These are (a) the chemical nature of its hormone
or hormones, (b) the factors that regulate the secretion of hormones
by the gland, and (c) the mode of action of the hormone or the manner
in which it produces the observed changes in the chemical processes
and consequently the degree of activity of the cells upon which it acts.

For my part I have nothing to report regarding the first of these
lines of investigation, but I cannot pass unnoticed the brilliant achieve-
ments of recent years that have led first to the isolation and more re-
cently to the synthesis of the hormones of the adrenal cortex. For it
must be remembered that the isolation in pure form of a hormone is
not only an achievement in itself, but makes immeasurably easier the
work of those who are investigating the function of the gland from
which it comes.

2 The experiments that are referred to in this paper were carried out in association
with Dr. A. White, Dr. and Mrs. G. Sayers, Miss E. Fry of the Department of Physio-
logical Chemistry, and Dr. T. Dougherty of the Department of Anatomy, Yale Univer-
sity.

THE ADRENAL GLAND 27I

Chemistry of Adrenal Cortical Hormones

The pioneer studies of Stewart and Rogoff and of Hartman had
indicated that various types of adrenal cortical extracts could effect a
moderate prolongation of the life of adrenalectomized animals. The
recognition that the activity resides primarily in lipid extracts of the
gland and that such extracts could indefinitely prolong the life of
adrenalectomized animals is to the credit of Professor Swingle of
Princeton University and his colleague Dr. Pfififner. Their methods of
extraction furnished the starting material for all subsequent isolations
of the active principles.

These active principles were soon recognized to belong to the great
class of chemical substances known as steroids. The isolation of these
and some twenty others not possessing biological activity is due to the
work of Pfiffner, Wintersteiner, and Kendall in this country and to
that of Reichstein in Switzerland. The wealth of steroid material pres-
ent in these glands, both in number of compounds and the diversity of
their action, exceeds that found in any other organ of the body, not
excluding the gonads, which in their capacity as endocrine glands also
secrete steroid hormones. Why the adrenal cortex should contain and
presumably manufacture not only the steroids with the characteristic
activity of this organ, but also those with gonadal activity, as well as a
large number with no known biological activity, is but one of the un-
solved problems of its function.

The minute concentrations in which the active steroids of the adre-
nal, the corticosterones, are present in the gland has made it difficult for
investigators to carry out the number of experiments with the pure
hormones that they have desired. The generosity of those who have
laboriously isolated and then donated these valuable compounds to their
colleagues for their work is beyond praise, and as one of the recipients
of their kindness I am happy to acknowledge how much it has meant
to our work.

It is also a matter of gratification to know that the partial synthesis
both of dehydro corticosterone and i i-dehydro-i /-hydroxy cortico-
sterone from desoxy cholic acid has now been accomplished. This im-
portant work is the result of the collaboration of a number of organic
chemists in different parts of the country. The combined efforts of
Wallis of Princeton, Kendall of the Mayo Clinic, Gallager at Chicago,
Riegel at Northwestern, Wintersteiner at the Squibb Research Insti-
tute and Sarett of the Research Laboratories of the Merck Company

272 C. N. H. LONG

have resulted in the preparation of quantities of the synthetic hormones
far exceeding those ever available from natural sources. The importance
of this to the further conduct of research in the field of adrenal physi-
ology is obvious, while the potential value of these new synthetic hor-
mones in the treatment of human disease can now, for the first time,
really be explored.

The Regulation of Adrenal Cortical Secretion

The manner in which the activity of the adrenal cortex is adjusted
to the needs of the organism and the circumstances which call forth an
increased secretion of its hormones is a problem engaging many investi-
gators at the present time. As I mentioned above, one of the tenets of
present day endocrinology is that the endocrine glands do not act in an
independent and haphazard manner but adjust their secretory rate very
closely to the actual needs of the moment.

The second question to be answered is : "What are the factors that
regulate the secretory activity of this gland?" This requires for its solu-
tion the development of methods that will tell us when the gland is
actually producing its hormone at an accelerated rate.

A good deal of information on this problem is now available. In 1930
P. E. Smith (2) showed that removal of the hypophysis in the rat was
followed by atrophy of the adrenal cortex and that this could be pre-
vented by the injection of crude pituitary extracts or by implants of the
anterior pituitary itself. A large amount of work has since shown that
all conditions that cause adrenal cortical enlargement, and these are too
numerous to mention, fail to do so in the absence of a factor from the
anterior lobe of the pituitary. This factor is a protein hormone now
known as the adrenotrophic or corticotrophic hormone. It has recently
been isolated from the anterior lobe in a highly purified if not com-
pletely pure form by workers in two laboratories. Such a preparation
has been obtained from sheep glands by Li, Simpson, and Evans (3),
and from hog glands by Sayers, White, and myself (4).

It is interesting to report that these two groups, working independ-
ently and with the pituitaries of different species, isolated a substance
whose biological, chemical, and physical properties agreed in a remark-
able manner.

When injected into animals this trophic hormone produces all the
effects given by the cortical hormones themselves. Thus it has been
found that it brings about an increased level of liver glycogen and of

THE ADRENAL GLAND 273

urinary nitrogen excretion. It causes glycosuria, sodium retention, and
increased potassium excretion in rats and has an effect identical with
that of the corticosterones in reducing the number of circulating lympho-
cytes. All these effects can be evoked in the hypophysectomized animal
in which no other known agent will cause an increased adrenal cortical
secretion. So far as is known at the present time the only effect of this
hormone is the stimulation of the secretory function of the adrenal
cortex.

But while the identification and isolation of this trophic hormone
answers the question of the agent responsible for an increased adrenal
cortical secretion, it merely transfers to the anterior lobe of the pituitary
the question of the mechanisms that cause this gland to secrete the
agent that is essential for an increased hormone production by the adre-
nal cortex. Are these mechanisms of nervous or humoral origin?

Before such a question can be tested experimentally it is necessary
to have available some method of determining the rate of secretion of
the adrenal cortical hormones. When we began our work, the available
methods for following adrenal cortical secretion were not very suitable
for studies in which small animals were to be exposed to a variety of
conditions. Those that had been used included (a) the determination
of "cortin" excretion in the urine by chemical or biological methods.
This is obviously impracticable in small animals, although in man it has
been successfully used by Browne and Venning and others to show that
trauma or infections increase the secretory rate. The second method
(b) used direct determinations of cortical hormone by biological means
in the blood of the adrenal vein. This method, applicable only to dogs
under anesthesia, had been used by Vogt (5) in England. It has yielded
important information, including the demonstration that epinephrine
injection is followed by an enhanced output of the hormone. The third
group (c) consisted of methods based on the actual content of hormone
in the gland. These are not practical at the present time.

Consideration of the question led to the suggestion that some lipid
constituent of the gland might vary with different degrees of its activity.
It was known that the stainable lipid of the gland decreased under con-
ditions associated with increased secretion, and there was some evidence
that the cholesterol content of the gland underwent similar changes.
Indeed there was a voluminous literature on this last subject, containing
many contradictions and expressions of opinion as to the relation of
adrenal cholesterol to the intermediary metabolism of this substance in
the rest of the organism. The general idea that cholesterol might be the

274 C. N. H. LONG

precursor of the steroid hormones has also been put forth by various
investigators but curiously enough no experiments to demonstrate that
it varied under conditions of increased secretion of the adrenal cortex
or gonads had been done.

We therefore injected the highly purified preparations of adreno-
trophic hormone into rats and determined the cholesterol content of the
glands after various intervals of time (6). It was found that single in-
jections of this trophic hormone decreased the cholesterol content of
the adrenal by some 50% in 3 to 6 hours, while its return to normal
levels required many hours. Similar results were observed in hypophy-
sectomized rats, provided too long a time had not elapsed since the oper-
ation. The cholesterol content of a variety of other tissues was not
affected by such injections. Finally, exposure of animals to such stresses
as cold, burns, trauma or hemorrhage brought about similar alterations
in the adrenal cholesterol of normal rats but did not alter the level in
hypophysectomized animals. Evidently the natural secretion of adreno-
trophic hormone following such procedures can also be detected by the
changes in the cholesterol content of the adrenal.

In reviewing these results it appeared entirely possible that the rather
long period, an hour or -more before the adrenal cholesterol fell signifi-
cantly, might be due to the fact that this substance acted as a reservoir
for the continual replenishment of the hormone itself and that other
substances in the gland might be more closely related to the actual secre-
tion.

The adrenal cortex is also unique in its high content of ascorbic acid
(Vitamin C). Furthermore this gland and the corpus luteum are the
only tissues in which such a high concentration of the vitamin is found
along with a very high concentration of cholesterol in the ester form.
It is unnecessary to remind you that ascorbic acid was originally iso-
lated from the adrenal, although the reasons for its presence in such
large amounts were not known. Further study of the literature indicated
that as in the case of cholesterol the level of this vitamin in the adrenal
was subject to wide variations, it being notable that under conditions of
stress low values were found.

Consequently, similar experiments were carried out with adreno-
trophic hormone and it was found that the ascorbic acid level of the
adrenal fell very promptly after the injection, reaching values of less
than fifty per cent of the original level within one hour (7). Indeed, a
significant decline could be detected within twenty minutes after injec-
tion. Exposure of normal animals to the types of stress mentioned

THE ADRENAL GLAND 2/5

above was followed by marked reductions in adrenal ascorbic acid, yet
similar treatment of hypophysectomized animals failed to affect the
level of the vitamin in the gland. The ascorbic acid content of other tis-
sues was not affected either by adrenotrophic hormone or by exposure
of animals to conditions that were followed by a decline in its content
in the adrenal.

While it might be argued that these changes in adrenal cholesterol
and ascorbic acid are only an indirect measure of adrenal cortical secre-
tion, our experience has convinced us that they furnish a reliable and
rapid indicator of this secretion. As I have pointed out above, this is ap-
parently mediated by the preliminary passage into the blood stream of
additional quantities of the pituitary adrenotrophic hormone. Even if
these changes in adrenal cholesterol and ascorbic acid were not directly
associated with the elaboration and secretion of the cortical hormones
but were merely expressions of a general heightening of adrenal metab-
olism, they would still be of value for following the secretory activ-
ity of the gland. However, there is other evidence that more directly
links these changes in adrenal chemistry with the actual formation and
release of the cortical steroids. Bloch and Rittenberg have shown by the
use of cholesterol tagged with deuterium that not only the bile acids but
also progesterone are formed from this steroid. The structure of proges-
terone is sufficiently akin to that of the corticosterones to make it ex-
tremely probable that the latter also are formed from cholesterol.

At least one role of ascorbic acid in the adrenal would appear to have
been clarified by the recent report of Lowenstein and Zwemer (8) that
it is possible to isolate from this gland a water soluble and biologically
active steroid in which the steroid nucleus is linked with ascorbic acid.
If this is correct then ascorbic acid is actually a part of the hormone in
the form it is secreted by the gland, cholesterol presumably furnishing
the material from which further quantities of the steroid portion may
be elaborated as it is required.

The association of a vitamin with a hormone in this manner is the
first known example of a common point of activity of members of these
two important groups of substances, although examples of vitamins
forming portions of co-enzymes are now fairly numerous. The recent
report of Price, Colowick and Cori (9) on the influence of insulin, an-
terior pituitary, and cortical hormones on the activity of the enzyme
hexokinase is an outstanding instance of such interplay, particularly
since nicotonic acid amide is a component of the hexokinase system. In

276 C. N. H. LONG

this system a hormone, vitamin, and protein enzyme are all part of a
cellular mechanism for carbohydrate metabolism.

Finally it may be pointed out that there is an excellent degree of cor-
relation between the decline in adrenal cholesterol and ascorbic acid and
such tangible and well-authenticated manifestations of adrenal cortical
activity as the increase in liver glycogen in fasting animals and the fall
in the number of circulating lymphocytes.

It is therefore possible not only to use these changes in the chemical
composition of the adrenal as indicators of increased secretory activity
but also under the right conditions to employ them as a method for the
assay of the adrenotrophic hormone itself. For this last purpose it is
desirable to use hypophysectomized animals as the test objects, since
their adrenal cholesterol and ascorbic acid will be affected only by the
trophic hormone and not by any nonspecific agent in the material in-
jected. By the use of such a method Dr. and Mrs. Sayers, now at the
University of Utah, have found that the blood level of adrenotrophic
hormone in normal rats is from 10-15 micrograms per 100 mis. and
that a rat pituitary gland contains about 20 micrograms of the hormone.
It should now be possible to apply this method to the determination of
the level of adrenotrophic hormone in the body fluids, a procedure that
should be of particular value in the diagnosis of certain endocrine dis-
orders in man.

We have already made a considerable number of observations on the
effect of various procedures on the secretory activity of the adrenal
cortex. In most of these experiments we have been satisfied to deter-
mine only ascorbic acid, since it reflects any change more rapidly than
cholesterol. We have found that exposure to cold (4°C for i hour),
trauma to muscles and particularly to the long bones, painful stimuli,
cutting or bruising the skin, burns, the intraperitoneal injection of cold
isotonic saline or other fluids, and the inhalation of volatile anesthetics,
all cause an immediate and prompt increased rate of secretion of the
cortical hormones (10). As might be expected, the degree of activation
of the gland is determined by the intensity and duration of the stimulus
applied.

We come now to an interesting and important point concerning the
manner in which these diverse types of stimuli bring about an increased
degree of cortical secretion. Three recent investigations bear on this
point.

(a) We have found that epinephrine, which, is certainly released after
the most varied types of stress, causes within an hour after subcutane-

THE ADRENAL GLAND 2/7

ous or intravenous injection a decrease in adrenal ascorbic acid in nor-
mal rats but not in rats 3 days after hypophysectomy. Tlie amounts
required are within the physiological range.

(b) Vogt (11) has also found that epinephrine causes a marked in-
crease in the quantity of cortical hormone in the adrenal vein blood of
dogs. In chronic experiments, epinephrine produced adrenal hypertrophy
only in normal but not in hypophysectomized rats.

In other words Vogt and ourselves are agreed that epinephrine is a
potent stimulus to adrenal cortical secretion, ahhough its exact mechan-
ism of action is still unknown.

It is possible to devise more decisive experiments to settle this point,
but whatever may be the answer it is of importance that, either directly
or indirectly through the adrenotrophic hormone, the stimulation of the
sympathetic nervous system and consequent liberation of epinephrine
is capable of increasing cortical hormone output. Such sympathetic ac-
tivity is always a component of those varied circumstances that are
known to call forth such secretion by the cortex. Heat, cold, hemor-
rhage, trauma, burns, toxic agents and painful stimuli of all kinds are
always associated with an increased degree of activity of the sympa-
thetic nervous system. Consequently it may be assumed that this is at
least a very important if not the only means by which an augmented
supply of cortical hormone is released to meet these states of emer-
gency.

I am also aware of the implications of the statement that epinephrine
can cause stimulation of the adrenotrophic function of the anterior
pituitary. For if it is an agent in this regard then it must act either by
the stimulation of adrenergic fibers in the anterior lobe, nerve fibers
whose existence is equivocal, or directly upon certain sensitive cells in
the gland itself, after the manner in which it accelerates glycogenolysis
in skeletal muscle.

(c) A third set of experiments expressing another point of view, or
rather another mechanism by which adrenal cortical secretion is regu-
lated, has recently been published by Dr. and Mrs. Sayers (12).

They have found that prior treatment of rats with cortical hormone
prevents the usual fall in ascorbic acid when the animals are exposed
to cold. We have confirmed this and shown that such treatment also
prevents the fall in adrenal ascorbic acid not only after painful stimuli,
trauma, etc. but also that produced by epinephrine. I have no doubt that
the effect of other types of stress on this component of the adrenal can
also be prevented by prior treatment with cortical hormone. Dr. and Mrs.

278 C. N. H. LONG

Sayers suggest that the blood level of cortical hormone determines the
secretion of the adrenotrophic hormone, A decline in the former, pre-
sumably caused by the stress applied to the animal, acts as a stimulus to
the secretion of the latter. If the fall in the blood level of cortical hor-
mone is prevented by injection of exogenous cortical hormone, then
the pituitary does not respond. In other words a differential and tem-
porary hypophysectomy is achieved by this means.

These experiments imply that stress produces conditions that cause
an increased rate of utilization of cortical hormone in the tissues. This
in turn brings about a reduction of the blood level of this hormone
which is the exciting factor to the anterior pituitary. Such a view is
perhaps half way between that of Vogt and ourselves. It regards the
adrenotrophic hormone as the essential factor for cortical secretion but
relegates epinephrine to a place with other nonspecific types of stress that
reduce the blood level of cortical hormones. Although our experiments
indicate that the stimulating effect of epinephrine on the pituitary (if
this exists) is abolished by raising the blood level of cortical hormones,
it must be remembered that epinephrine itself is a hormone with well-
defined points of action and is not in the usual sense "a nonspecific
agent." But it is also possible that the well-known metabolic effects of
epinephrine may cause an increased utilization of cortical hormone by
the tissues or change the composition of the blood passing through the
anterior lobe. Only further experimentation can settle these contro-
versial points, but I again repeat that the fact that epinephrine, by
whatever mechanism, increases the secretion of cortical hormone is in
itself a point of some importance in our understanding of the means
by which the organism adapts itself to stress. Indeed this observation
furnishes an interesting corollary to the emergency theory of adrenal
medullary function put forward by the late W. B. Cannon.

Even with these points unsettled, the picture of the control of adrenal
cortical secretion appears at the present time to be as follows :

(a) The changes in adrenal ascorbic acid and cholesterol may be used
as indicators of cortical activity. The former is probably directly related
to the secretion of the hormone, the latter acting as a reservoir of the
precursor of the hormone.

(b) The various circumstances that increase the cortical secretion do
so in all probability by first activating the adrenotrophic secretion of the
anterior pituitary.

(c) The manner in which anterior lobe activation is effected is either

THE ADRENAL GLAND 2/9

by a decrease in the blood level of cortical hormone or some other un-
known changes in the composition of the blood traversing the gland.

(d) The release of epinephrine, which occurs in so many circum-
stances, increases adrenal cortical secretion. It may do so either (i)
by a direct effect on the gland (which is Vogt's view), (2) by stimula-
tion of the anterior pituitary, or (3) as a result of its own effects on the
composition of the blood passing through the pituitary.

The series of events just outlined is a good example of the correla-
tion and interplay of three components of the endocrine system, anterior
pituitary, adrenal cortex, and medulla. Their combined operation facili-
tates the adaptation of the organism to external and internal stresses,
stresses which if unchecked would threaten its continued existence.
Their operation not only corrects changes in the fluid environment that
are detrimental to cellular function but also creates conditions within
the environment and the cells that are best adapted to meet the emer-
gency.

Mechanism of Action of Adrenal Cortical

Hormones

I would now like to turn from a consideration of the regulation of
adrenal cortical activity to the equally complicated question of the man-
ner in which the hormones of this gland produce the observed metabolic
changes in the cells and body fluids. For it must be in their capacity to
facilitate certain types of cellular activity that the protective properties
of these hormones reside.

There is not space here to review the large amount of work that has
been done on the function of the cortical hormones. I can but remind
you that these hormones are particularly active in the regulation of
many phases of metabolism. Thus they regulate the sodium, potassium,
and water balance by their influence on the rate of kidney tubular excre-
tion and reabsorption of these ions, and in this way they control the
shift of fluid and electrolytes between the various compartments of the
body. As is now well known, these hormones also have a marked effect
on several phases of the organic metabolism. It was recognized many
years ago that a subnormal blood glucose level is a frequent finding in
adrenal insufficiency in animals and man. Britton has shown that the
other carbohydrate constituents of the body, notably the liver and
muscle glycogen, are depleted and that the injection of the cortical hor-
mone, even without food, rapidly restores these levels.

28o C. N, H. LONG

In spite of this, it was soon surmised that the control of the electro-
lyte or carbohydrate levels did not fully explain the action of the hor-
mone. For while it is true that adrenalectomized animals or humans
with adrenal insufficiency might die from electrolyte loss or carbo-
hydrate depletion, the mere administration of sodium salts or carbohy-
drate was not sufficient to return them to a normal state, even though
life might b€ extended by their use.

In 1934 Lukens and I (13) found that adrenalectomy alleviated
many of the consequences of total pancreatectomy in the cat and dog.
The most notable effect, apart from a prolongation of life, was the
reduction in urinary glucose and nitrogen excretion, a fall from the
characteristically high to more normal blood glucose levels and a strik-
ing abatement of ketosis.

The general conclusion concerning these metabolic changes was that
adrenalectomy reduced the high rate of conversion of tissue proteins
to glucose (gluconeogenesis) while at the same time allowed a higher
rate of carbohydrate utilization. The converse conclusion would be that
the adrenal cortical hormones accelerated the conversion of tissue pro-
teins to glucose and/or inhibits carbohydrate utilization by the tissues.

Later experiments demonstrated that in other conditions accom-
panied by a high rate of gluconeogenesis from protein, adrenalectomy
was followed by a comparable reduction in the quantity of protein
undergoing catabolism. It has since been shown in several laboratories
besides our own that the injection of either the cortical steroids or
adrenotrophic hormone into normal fasting animals increases the rate
of protein breakdown to a marked degree, while at the same time large
quantities of glycogen are deposited in the liver. In fed animals, as Ingle
has shown, actual glycosuria follows prolonged injection of either of
these hormones, indicating that carbohydrate utilization was also sup-
pressed.

The adrenal cortical hormones, by their capacity to accelerate the
catabolic phases of protein metabolism, retard the growth of growing
animals. They therefore stand in opposition to the growth promoting
factor of the pituitary, which promotes the retention of protein in the
body. Since the activities of the adrenal cortex are also regulated by
the anterior pituitary, this gland exercises control over the rate of both
the catabolism and anabolism of protein.

Though we use such terms as "promotion of gluconeogenesis," "stim-
ulation of protein catabolism," and "inhibition of carbohydrate utiliza-
tion" to describe the function of the adrenal cortex, we have not yet

THE ADRENAL GLAND 281

succeeded in completely defining the action of these hormones. It is
true that a stimulation of glucose formation along with a suppression
of tissue utilization of carbohydrate is an important factor during
periods of fasting. Furthermore the present concept of a dynamic state
of protein catabolism indicates the necessity for the regulation of its
catabolism as well as anabolism. In spite of all these examples of the
teleological significance of the hormone, there is still one feature of
adrenalectomized animals that requires consideration before we can
accept any present hypothesis as a complete explanation of the function
of this gland. This is the remarkable inability of adrenalectomized or
hypophysectomized animals to withstand alterations in the external or
internal environment that are easily tolerated by intact animals. This
intolerance is too well known to need further description, but it should
be remembered that merely correcting the disturbances in the sodium
or carbohydrate metabolism by administration of these substances is not
sufficient in itself to ensure survival.

These measures, while they may bring about the extended survival
of adrenalectomized animals living under quiet undisturbed conditions,
fail conspicuously to protect such animals under conditions which are
easily tolerated by normal animals. Only the cortical hormones can con-
fer a normal resistance to stress to either adrenalectomized or hypophy-
sectomized animals.

This fact and many more that might be enumerated, for example the
effect of the hormones on the early fatigue of the skeletal muscles, so
well described by Ingle, indicate that there still remains to be found
some other basic mechanism of their action. The effects of adrenal cor-
tical insufficiency are extended to many organs, each of which expresses
this insufficiency in terms of its inherent function. Thus we observe the
rapid fatigue of working muscles, inadequate intestinal absorption, re-
duced ability of the kidney to separate sodium and potassium, and such
defects in liver function as a decreased rate of gluconeogenesis. Yet
even with such evidence of the necessity of cortical hormones for the
proper function of so many different organs, we must assume until it
is proved otherwise that in each organ the point of action of the hor-
mones is the same. At the moment we do not know, except in general
terms, what this effect is. All we can say is that without these hormones
the capacity of many types of cells to perform their particular function
is impaired, and that this becomes particularly evident in times of stress.

Recently some new observations on a well-known effect of cortical
hormones have been reported which indicate, if further proof were

282 C. N. H. LONG

needed, that we have still much to learn about their action in the body.

Those of you who are familiar with the beautifully illustrated mono-
graph in which Thomas Addison first clearly defined the relationship
of the adrenals to the disease in man which now bears his name, will
remember that one plate shows the remarkable enlargement of the ab-
dominal lymph nodes in the cases he described.

Since that time, numerous observations both in man and animals have
made us aware of the reciprocal relationship between the adrenal cortex
and the lymphoid elements of the body. Withdrawal of the hormone is
associated with lymphoid enlargement; an excess of hormone causes
rapid involution of these elements. This relationship has been particu-
larly emphasized by Selye in his many papers describing the phenomena
associated with the "alarm reaction." This investigator has shown that
a variety of insults to the organism cause, along with adrenal cortical
hypertrophy, a rapid involution of the thymus and lymph nodes.

Dr. White in my laboratory in association with Dr. Dougherty of
the Department of Anatomy at Yale have recently made a more detailed
analysis of this relationship (14). Their results may be summed up as
follows :

(a) The number of circulating lymphocytes is under pituitary-adre-
nal control. The cortical hormones or the adrenotrophic hormone when
injected into normal animals cause a rapid decrease in the blood lympho-
cytes, reaching a maximum a few hours after injection.

(b) The reduction in the blood lymphocytes is not due to an increased
rate of withdrawal by the tissues but is brought about by a rapid lym-
pholysis in the lymphoid tissues themselves. This is further emphasized
by the experiments of Reinhart, who found a fifty per cent reduction of
lymphocytes in the thoracic duct lymph within twenty minutes after the
injection of adrenotrophic hormone.

(c) This lympholysis liberates into the lymph and ultimately into the
blood serum the contents of the lymph cells. One of these constituents
has been identified as normal serum gamma globulin ; a second constitu-
ent is probably serum beta globulin. The rise in total serum globulin,
due almost entirely to these two globulins, has been shown by electro-
phoresis.

These experiments indicate an entirely new efifect of cortical hor-
mones on protein metabolism. For by this action, intact cellular protein
molecules are translocated through the body fluids to all parts of the
organism. As in all instances of hormone action this process is not
initiated by the hormone but merely altered in the rate at which it pro-

THE ADRENAL GLAND 283

ceeds. Once dispersed in this manner this protein must be metabolized
by the different tissues in a manner best suited to their requirements,
and these presumably vary according to their function.

(d) Since it is well known that the gamma globulins are the fraction
of the serum proteins that carry the specific immune bodies, White and
Dougherty immunized animals to various antigens, both with and with-
out the injection of cortical hormones. In all cases the titre of the spe-
cific antibodies in the serum rose faster and to higher levels in the group
receiving the hormone. But this was not all. For if the injection of both
hormone and antigen was stopped and sufficient time allowed for the
serum titre of antibody largely to disappear, the injection of a single
dose of cortical hormone caused antibodies to reappear in a few hours
in the serum to levels reached during the height of immunization. Ex-
posure of immunized animals to any stimulus that causes adrenal cor-
tical secretion was followed by similar rapid increases in serum anti-
bodies. In other words this "anamnestic reaction" is a manifestation of
the control of lympholysis and serum globulin release by the cortical
hormones.

It is apparent that the release of immune globulins in immunized ani-
mals is merely a special case of a general phenomenon, for similar quan-
tities of non-immune globulins are released along with other cellular
materials from the lymphocytes of non-immunized animals. Yet in the
immunized animals the actual release of cortical hormones confers a spe-
cific resistance to a particular infecting agent. Is there then a relation-
ship between the lympholysis occurring in non-immunized animals and
their resistance to such nonspecific stimuli as heat, cold, trauma, etc.? Is
such nonspecific resistance determined at least in part by the capacity of
the cortical hormones to liberate from the lymphoid elements into the
circulating fluids substances that are vital to the cells for a successful
defense against the distortion of their internal environment? This is an
unanswered but important question, but these experiments have opened
an entirely new line of thought, not only concerning the inability of
adrenalectomized or hypophysectomized animals to contend with the
factors that distort their internal environment, but also as to the mech-
anism of action of the cortical hormones. If nothing else, they lend em-
phasis to the point that there are still unexplored fields in our knowledge
of the function of the adrenal cortex.

Such studies are for the future; and since this is so, I too must stand
with my predecessors of some 230 years ago before Baron Montesquieu
and say that I can only answer in part the question posed by the Acad-

284 C. N. H. LONG

emy. I trust that my contemporaries will say as he did to them, "These
fruitless efforts are rather a proof of the obscurity of the problem than
the sterility of those who studied it."

REFERENCES

1. Biedl, A. Janus, 75, 193, 1910,

2. Smith, P. E. Amer. J. Anat., 45, 205, 1930. Anat. Rec, 52, 191, 1932.

3. Li, C. H., H. M. Evans, and M. E. Simpson. /. Biol. Chem., 14P, 413,

1943-

4. Sayers, G., A. White, and C. N. H. Long. 7. Biol. Chem., 14P, 425,

1943-

5. Vogt, M. /. Physiol., 103, 317, 1944.

6. Sayers, G., M. A. Sayers, E. G. Fry, A. White, and C. N. H. Long.
Yale J. Biol. Med., 16, 361, 1944.

7. Sayers, G., M. A. Sayers, T. Y. Liang, and C. N. H. Long. Endocrin-
ology, 3/, 96, 1945.

8. Lowenstein, B. E., and R. L. Zwemer. Endocrinology, 38, 6t^, 1946.

9. Price, W. H., S. P. Colowick, and C. F. Cori. /. Biol. Chem., 160,

633. 1945-

10. Long, C. N. H. Recent Progress in Hormone Research, I, 99, 1947.
Academic Press, New York.

11. Vogt, M. /. Physiol., 104,. 60, 1945.

12. Sayers, G., and M. A. Sayers. Endocrinology, 40, 265, 1947.

13. Long, C. N. H., and F. D. W. Lukens. /. Exp. Med., 63, 465, 1936.

14. White, A. Harvey Lectures 1947-48.

NAME INDEX

Abderhalden, E., ii, 36, 42

Adler, E., 12, 44

Ahrens, R., 107, 117, 132

Albaum, H. G., 24, 47, 68, 70

Alcock, R. S., 12, 42

Alexander, C. S., 19, 26, 46

Alexander, E. R., 79

Algera, L., 97, 104

Algeus, S., 69, 70

Allen, E., 220, 257

Allen, T. H., 19, 22, 42

Allison, F. E., y^

Amill, J. A., 78

Amprino, R., 172, 186

Anderson, E. H., 96, 104

Anderson, K., 28, 2Z, 42

Anderson, T. F., 46

Andervont, H. B., 250, 264

Andrewes, C. H., 244, 263

Ansari, M. Y., 229, 260

Anson, M. L., 27, 42

Astbury, W. T., 52, 53, 54, 56, 57

Atkin, L., 95, 103

Auerbach, C, 131, 229, 260

Avery, G., 54

Avery, G. S., 65, 67, 70

Avery, O. T., 3, 23, 42

Bahrs, A. M., 82, 83

Bail, O. O., 96, 103

Baldwin, D. M., 29, 45

Ball, E. G., 122, 123, 131

Ballard, William W., 172, 183

Ballentine, R., 90

Baltzer, F., 188, 211

Barker, H. A., 98, 104

Barrett, Morris K., 238, 262

Barron, E. S. G., 107, 108, 112, 113, 116,

117, 119, 131, 132, 133
Barth, L. G., 192, 211
Bartlett, G. R., 107, 117, 119, 131, 132
Barton-Wright, E. C, 76, 78
Bartz, Q. R., 21, 45
Baumann, C. A., 230, 261
Bawden, F. C, 229, 260
Bayliss, W. M., 6, 20, 42
Beadle, G. W., 23, 42, 99, 104, 223, 224,

225, 227, 229, 231, 239, 245, 259, 260,

261, 262
Bear, R. S., 52, 54, 57, 150, 167, 183, 185
Beard, J. W., 241, 262
Behrens, O. K., 11, 42
Belitzer, V. A., 122, 132
Berger, J., 65, 67, 70

Bergmann, M., 7, 9, 10, 11, 12, 36, 42, 43,

124, 132
Bernal, J. D., 54, 183
Bernstein, D. E., 99, 104
Berry, G. P., 33, 48, 240, 262
Berry, L. J., 81, 83
Bessey, O. A., 180, 186
Biedl, A., 266, 284
Biesele, J. J., 222, 229, 258
Bird, H. R., 73, 90
Bittner, J. J., 221, 249, 250, 251-253, 257,

258, 264
Blanchard, M., yy
Bloch, K., 124, 133
Blodgett, K. B., 150
Bloom, William, 141, 183
Blum, H., 228, 260
Bodenstein, D., 229, 260
Bodine, J. H., 19, 22, 36, 42, 43
Boell, E. J., 206, 208, 209, 210, 211
Bonner, J., 64, 66, 68, 70
Bonner, W. D., Jr., 68, 70
Boon, M. C, 226, 237, 259, 262
Bordet, J., 23, 24, 43
Borsook, H., 4, 7, 8, 10, 12, 18, 43, 48, 117,

118, 132
Boulanger, P., 132
Bourquelot, 6
Boveri, Th., 188, 211
Boycott, A. E., 189, 211, 212
Boyland, E., 230, 261
Brachet, Jean, 152, 183, 207, 208, 212
Brandt, K., 132
Braunstein, A. E., 12, 43
Bredig, G., 3, 43
Briggs, G. M., 73
Britton, S. W., 279
Brown, M. F., 265
Browne, J. S. L., 273
Brues, A. M., 218, 257
Bryan, W. R., 250, 264
Biicher, H., 132
Buck, M., 234, 261
Burchenal, J. H., 229, 260
Burk, D., y2, 229, 260
Burke, V., 26, 43
Burnet, F. M., 24, 26, 36, Z7, 38, 39, 40,

41, 42, 43
Burrows, H., 220, 257
Butler, E. G., 210, 212, 214
Byer, A. C, 63, yi

Calnan, Dorothy, 229
Camagni, L. J., 234, 261

286

INDEX

Campbell, D. H., 37, 47

Cannon, W. B., 268, 278

Carey, Eben J., 177, 183

Carr, J. G., 131

Carruthers, C, 229, 260

Caspersson, T., 23, 43, 125, 132, 208, 212

Castle, D. C, 79, 80

Chang, C. Y., 192, 193, 194, 215

Chase, J. H., yj, 44

Chatton, E., 133

Cheldelin, V. H., 90

Chesney, A., 93, 103

Chevremont, M., 142, 183

Chevremont-Comhaire, S., 142, 183

Child, C. M., 152, 192, 212

Chinard, F. P., 132

Christian, W., 106, 133

Clapp, M. P., 265

Clark, W. Mansfield, 114, 115, 132

Claude, A., 239, 240, 241, 253, 262, 264

Clifton, C. E., 99, 104

Cline, J. K, 81, 83

Cohen, P. P., 125, 132

Cohn, M., 81

Cohn, W. E., 218, 257

Colowick, S. P., 68, 70, 71, 122, 127, 132,

133, 27s
Commoner, B., dd, 69, 70, 71
Conklin, E. G., 188, 212
Cooper, Z. K., 229, 260
Cori, C. F., 7, 43, 55, 70, 71, 98, 104, 127,

132, 133. 275
Cori, G. T., 7, 43, 98, 104
Cornman, I., 229, 260
Cowdry, E. V., 229
Cramer, H., 35, 44
Cravens, W. W., 90
Crisp, L. R., 265

Daft, F. S., 76, 78

Dagley, S., 95, 103

Dakin, H. E., 124, 132

Dalcq, A. M., 156, 183, 191, 212

Dale, W. M., 21, 26, 27, 108, 132, 226, 255,

259
Daniel, G. E., 229, 265
Daniel, J. F., 194, 212
Danielli, J. F., 150, 183
Dann, W. J., 90
Darby, W. J., 73, -j-j, 78
Darlington, C. D., 15, 22, 24, 43
Das, M. B., 12, 44
Davies, T. H., 132
Dawson, Alden B., 183
Day, P. L., '7^
Dedrick, H. M., 240, 262

Delbriick, M., 12, 24, 28, 29, 30, 31, 32, 33,

43, 46, 94. 103
Demerec, M., 222, 258
Derrien, Y., 36, 47
Detwiler, S. R., 171, 184
Dewey, V., 90
Dickman, S. R., 108, 131
Dienert, F., 34, 44
Dietz, V. R., 132
Diver, C, 189, 211, 212
Dobriner, K., 230, 261
Doljanski, L., 143, 183
DoudorofT, M., 94, 98, 99, 100, 103, 104
Dougherty, T. F., yj, 44, 270, 282, 283
DubnofF, J. W., 7, 10, 12, 43
Dubos, R. J., 34, 38, 44, 69, 233, 234, 261
Du Buy, H. G., 246, 264
Dunlap, C. E., 230. 261
Duran-Reynals, F., 241, 262
Du Vigneaud, V., 12

Eakin, R. M., 196, 212
Earle, W. R., 230, 264, 265
Eastcott, R. v., 86
Ebeling, A. H., 143, 184
Ecker, P. G., 9, 44
Edibacher, S., 36, 44
Ehrenwall, E. v., 11, 36, 42
Ehrich, W. E., Z7, 4°, 44
Eichel, B., 70
Ellermann, V., 240, 262
Ellis, E. L., 94, 103
Elvehjem, C. A., "jz
Emerson, S., 228, 239, 260
Emery, W. B., "jd. 78
Emmons, C. W., 227, 259
Enders, J. F., z^^, 48
Ephrusi, B., 132
Euler, H. v., 4, 6, 12, 35, 44
Evans, H. M., 272, 284
Eyring, H., 114, 132, 134

Fankhauser, G., 212

Fankuchen, I., 54

Ferri, M., 70

Elides, P., 90

Fischer, Albert, 143, 184

Fischer, E., 54

Fischer, F. G., 206, 215

Floyd, N. F., 133, 134

Fogel, S., 71

Folley, S. J., 9, 44

Fothergill, L. D., 36, 48

Fouldes, L., 244, 263

Freeman, M., 26, 36, 37, 38, 42, 43

Frey, C. N., 95, 103

Friedenwald, J., 132

INDEX

287

Friedewald, W. F., 221, 243, 258, 263

Frost, D. v., 90

Fruton, J. S., 7, 9, 10, 12, 43, 124, 132

Fry, E. G., 270, 274, 284

Fullam, E. F., 239, 262

Furth, J., 226, 237, 259, 262

Gallager, T. F., 271

Gardner, W. U., 220, 257

Garer, P. A., 237, 261

Garstang, S., 189, 212

Gautheret, R. J., 61, 71

Gentner, W., 133

Gerard, R. W., 184

Gibb, 25

Giese, J. E., 230, 261

Gillette, R., 206, 207, 212, 229, 260

Glasstone, S., 132

Gliicksmann, A., 171, 184, 225, 259

Gogek, C. J., 230, 261

Goksu, v., 44

Goldinger, J. M., 131, 132

Goodpasture, E. W., 26, 48

Gordon, M, 238, 262

Gordon, S. A., 65, 71

Gorter, E., 9, 44

Gould, Eleanor, 171

Gowen, J. W., 225, 234, 259

Gratia, A., 23, 24, 29, 43, 44

Gray, C. H., 225, 227, 259

Green, D. E., ^^, 133

Green, R. G., 221, 228, 250, 252, 257,

264
Greenstein, J. P., 221, 227, 258, 259
Gregory, L. E., 71
Griffith, F., 234, 261
Griineberg, H., 262
Gulick, A., 12, 15, 44
Giinther, G., 12, 44
Guyer, M. F., 228, 260
Gye, 28
Gyorgy, P., jz

Haagen Smit, A. J., 65

Haddock, J. N., 9, 44

Haddow, A., 219, 222, 257, 258

Hadorn, E., 188, 212

Haldane, J. S., 268

Hall, C. E., 55

Hall, E. K., 197, 212

Hammett, F. S., 130, 132

Handler, P., 90

Hanes, C. S., 98, 104

Harris, T. N., 37, 40, 44

Harrison, R. G., 172, 184, 187, 192,

207, 212
Hart, E. B, n

Hartman, F. A., 271

Harvey, E. B., 188, 189, 212, 245, 263

Hassid, W. Z., 98, 104

Haurowitz, F., 13, 44

Heatley, N. G., 206, 212

Hellcrman, L., 108, 132

Henderson, L. J., 268

Henrici, A. T., 93, 103

Henshaw, P. S., 226, 259

Herriott, R. M., 20, 21, 27, 31, 44, 45, 47

Hertwig, O., 171

Heuser, G. F., ^z^ QO

Hevesy, G., 130, 132

Hinshelwood, C. N., 95, 103

His, W., 187, 201, 202, 213

Hiscoe, Helen B., 166, 186

Hoffman, L. A., 112, 113, 132

Hogan, A. G., yz, 74. 90

Hollaender, A., 227, 259

Holmes, B., 226, 259

Holtfreter, Johannes, 159, 184, 192, I95r

199, 213
Hoover, S. R., TZ
Hopkins, F. G., 132
Horwitt, B. N., 221, 258
Howe, H. A., 23, 37, 45
Howe, P. E., 47
Hunter, F. E., 132
Huseby, R. A^ 251-253, 264
Hutchings, B. L., TZ^ 7^
Huxley, Julian S., 181, 184
258. Hyde, 228

Hyden, Holger, 167, 184

Imai, Y., 246, 263
Ingle, D. J., 280, 281

Jackson, A. V., 26, 36, Z7, 38, 42, 43

Jacobsen, C. F., 13, 46

Jakus, M. A., 55, 57

James, 68

Jaumain, D., 23, 43

Jenrette, W. V., 227, 259

Johnson, W. A., 132

Jones, E., 77, 78

Jones, R. N., 230, 261

Jordan, P., 23, 45

Jukes, T. H., 79

Just, E. E., 156, 184

Kalckar, H. M., 10, 45. 7i, 123, 132
Kaliss, N., 237, 262
Kalnitsky, G., 107, 132
Kamen, M. D., 16, 47. 125, 133
196, Kaplan, N., 98, 104

Karnofsky, D. A., 229, 260
Karrer, P., 81

288

INDEX

Karstrom, H., 34, 45, 102, 105

Keighley, G. L., 18, 43

Kendall, E. C, 271

Kensler, C. J., 230, 239, 255, 261

Keresztesy, J. C, 7Z

Kidd, J. G., 221, 228, 238, 242, 244, 251,

258, 260, 263, 264
Kidder, G. W., 90
Klatt, B., 172, 184
Kluyver, A. J., 98, 99, 104
Knight, B. C. J. G., 34, 45, 90, 99, 104
Knight, C. A., 27, 45
Knoop, F., 124, 132
Koepfli, J. B., 66, 71
Kogl, F., 73

Koltzoff, N. K., 15, 2,T, 45
Korinek, J., 95, 103
Kornberg, A., 133
Korr, J. M., 127, 132
Krebs, H. A., 69, 117, 119, 125, 132
Kritzmann, M. G., 12, 43
Krueger, A. P., 23, 29, 30, Z2, 33, 45
Kiihn, A., 245, 263
Kuhn, R., 132
Kunitz, M., 16, 19, 20, 21, 35, 36, 45

Laidler, K. J., 132

Lancefield, R. C, 86

Landsteiner, K., 36, yj, 38, 46, 184, 223,

259
Landstrom-Hyden, H., 212
Langmuir, I., 7, 9, 15, 46, 150, 184
Langstron, W. C, "jz
Lardy, H. A., 132
Lavin, G. I., 230, 261
Lea, D. E., 225, 226, 234, 259
Lederberg, J., 238, 262
Lehninger, A., 123, 132, 133
Levi, A. A., 230, 261
Lewis, L M., 233, 261
Lewis, W. H., 175, 184, 193, 213
Li, C. H., 272, 284
Liang, T. Y., 274, 284
Lillie, F. R., 190, 213
Lindahl, P. E., 206, 212
Lindberg, O., 129, 133
Lindegren, C. C, 35, 38, 46, 232, 245, 261
Lindegren, G., 35, 38, 46, 232, 245 261
Lindner, M., 87, 88
Lindstrom-Lang, K., 13, 46
Lipmann, F., 98, 104, 106, 122, 133
Little, C. C, 237, 239, 249, 261, 262, 264
Lloyd, W., 27, 46
Loeb, J., 42, 46
Logan, W. A., 99, 104
Long, C, 133

Long, C. N. H., 272, 274, 276, 2'jT, 278,

280, 284
Loring, H. S., 25, 46
Lowenstein, B. E., 275, 284
Luckey, T. D, yz
Lukens, F. D. W., 280, 284
Luria, S. E., 24, 28, 29, 30, 31, 32, zz, 43.

46
Lush, D., 26, 36, 37, 38, 42, 43
Luther, W., 193, 197, 213
Lwoflf, A., 133

McCarty, M., 3, 23, 42

McGilvery, R. W., 125, 132

MacDowell, E. C, 224, 237, 259, 262

MacKay, E. M., 124, 133

MacKenzie, K., 227, 260

MacLeod, C. M., 3, 23, 42

Madden, S. C., 17, 18, 46, 133

Maier-Leibenitz, H., 133

Mandel, S., 36, 47

March, H. C, 226, 259

Marinelli, L. D., 218, 257

Markham, R., 226, 259

Marrack, J. R., 36, 46

Mayer, N., 221, 258

Medawar, P. B., 237, 261

Medes, G., 133, 134

Mehler, A., 133

Mellors, R. C., 23, 45

Menten, M. L., 106, 133

Meredith, W. J., 226, 259

Mes, M. G., 64

Meyer, J., 132

Meyer, R., 103

MeyerhofF, O., 10, 46, 121, 133

Michaelis, L., 106, iii, 133

Michaelis, P. L, 245, 263

Milford, J. J., 241, 262

Miller, B. J., 133

Miller, G. L., 27, 46

Miller, J. A., 230, 261

Miller, Z. B., 117, 131, 132

Minot, G. R., 78, 79

Mitchell, H. K., n

Mohle, W., 133

Mollendorff, M., 142, 184

Monne, Ludwik, 184

Monod, J., 94, 95, 100, Id, 102, 103, 103

Moore, C. V., 78

Moore, Carl R., 180, 184

Moore, D. H., 19, 26, 46

Moore, J. A., 245, 263

Moosey, M. M., 221, 250, 257, 258, 264

Morgan, E. J., 132

Morgan, T. H., 192, 213

Mottram, J. C., 229, 260

INDEX

289

Mudd, S., yj, 46

Muller, H. J., 16, 46, 133, 222, 225, 227,

258, 260
Muller, W. H., 71
Munoz, J. M., 123, 133
Murphy, J. B., 26, 46, 238, 254
Murray, W. S., 249, 264

Nageotte, J., 158, 184

Needham, Joseph, 152, 166, 184, 192, 206,

213, 220, 258
Nettleship, A., 265
Nicholas, J. S., 156, 200, 213
Nieuwkoop, P. D., 195, 199, 200, 201, 213
Norris, L. C, 73, 90
Northrop, J. H., 9, 13, 15, I7. I9. 21, 22,

23, 24, 25, 27, 28, 32, 35, 36, 38, 42, 45,

46, 47, 125, 133

Ochoa, S., 68, 71, 120, 121, 122, 123, 133

Ohlmeyer, P., 133

Olivo, C. D., 71

Oppenheimer, J., 192, 193, 198, 213

Orcutt, M. L., Z7, 47

Ormsbee, R. A., 229, 260

Orstrom, A., 129, 133

Ostromuislenskii, I. I., yj, 38, 40. 47

Overbeek, J. van, 62, 64, 65, 71

Owen, S. E., 134

Palay, Sandford L., 144, 184

Palmer, K. J., 150, 185

Papanicolaou, G. N., 257

Parker, Raymond C, 165, 184

Parott, E. M., -Ji, 74

Pasteels, J., 193, 213

Pauling, L., 15, ZJ, 38, 40, 47, 126, 133,

148, 184
Pedersen, K. O., 18, 26, 47
Perie, N. W., 229, 260
Petersen, H., 43
Peterson, W. H., -ji
Petrie, A. H. K., 11, 47
Pfiffner, J. J., -jz, 74, 76, 271
Phelps, A., 93, 95, 96, 103
Plagge, E., 245, 263
Plough, H. H., 222, 258
Pollack, M. A., 87, 88
Porter, J. R., 96, 103
Potter, J. S., 237, 262
Potter, V. R., 24, 47, 221, 258
Poulson, D., 201, 205, 214
Price, W. H., 16, 47, 55, 127, 133, 275
Prince, L. H., 134

Rahn, O., 95, 96, 103
Rapkine, L., 130, 133

Rappaport, J., 62

Rathen, Alexandre, 241, 262

Ratner, S., ^7

Ray, O. M., 19, 22, 42

Reimann, S. P., 133

Reinders, D. E., 69, 71

Reiner, J. M., 16, 47

Reinhart, H. L., 282

Rhoades, C. P., 229, 230, 239, 255, 260,

261
Rhoades, M. M., 223, 246, 258, 264
Ricci, N. I., 27, 46
Richardson, G. L., 96, 103
Riegel, E. B., 271
Rittenberg, D., 124, 133
Rivers, T. M., 24, 47
Robertson, G. G., 190, 214
Robertson, T. B., 9, 47
Robinson, F. A., 76, 78
Robson, J. M., 131, 229, 260
Roche, J., 36, 47
Rogoff, J. M., 271
Roosen-Runge, E. C, 193, 213
Rose, S. M., 211, 214
Rose, W. C, 89, 90
Roskelly, R. C, 221, 258
Rottier, P. B., 96, 103
Rous, P., 240, 241, 242, 244, 254, 262, 263
Rudall, K. M., 57
Rudnick, D., 201, 202, 214
Ruffili, D., 133
Runnstrom, John, 156, 185
Rusznyak, I., 90

Sabin, F. R., 42, 47

Salter, W. T., 221, 258

Santerson, L., 132

Sarett, L. H., 271

Sawyer, C. H., 208, 209, 210, 214

Sax, K., 224, 249, 259, 264

Sayers, G., 270, 274, 276, 277, 278, 284

Sayers, M. A., 270, 274, 276, 277, 278

Schaefer, V. J., 7. 9. I5, 46

Schechtman, A. M., 200, 214

Schilling, E. L., 265

Schmitt, Francis O., 150, 167, 183, 185

Schnitzer, R. J., 234, 261

Schoenheimer, R., 11, 12, 18, 47, 106, 124,

133. 163, 185
Schott, H. F., 7, 43
Schotte, O. E., 210, 212, 214
Schrodinger, E., 192, 214
Schultz, A. S., 95, 103
Schumacher, A. E., 73
Schwerin, P., 44
Scott, G. H., 144, 185
Scribner, E. J., 29, 45

290

INDEX

Seastone, C. V., 21, 27, 47

Sebrell, W. H, 76, 78

Seidlin, S. N., 218, 257

Scifriz, W. A., 24, 47

Selye, H, 282

Sevag, AI. G., 34, 36, zi, 39, 40, 47

Shalucha, B., 70

Shelton, E., 265

Shen, S. C, 19, 26, 46, 209, 210, 211

Shope, R. E., 241, 255, 263

Simpson, M. E., 272, 284

Singer, M., 211, 214, 215

Singer, T. P., 108, 131, 132, 133

Skeggs, H. R., 87, 88

Skoog, F., 63, 71

Smith, E. A., 226, 228, 260

Smith, G. M., 238, 262

Smith, H., 229, 260

Smith, K. M., 226, 259

Smith, P. E., 272, 284

Snell, E. E., yz

Snell, G. D., 239, 262

Sonderhoff, R., 119, 133

Sonneborn, T. M., 246-249, 264

Spemann, Hans, 157, 185, 192, 206, 215,

245
Spencer, R. R., 229, 265
Spiegclman, S., 16, 35, 38, 46, 47, 125, 133,

166, 185, 232, 261
Spies, T. D., -77, 78, 81, 82, 83
Spizizen, J., 29, 47
Spratt, N. T., 193, 201, 215
Sprince, H., 87, 88, 90
Stableford, L., 195, 215
Stadler, L. J., 226, 259
Stanley, W. M., 15, 22, 23, 27, 42, 46, 48,

227, 260
Stark, T. H., 265
Stewart, C. C., 271
StilKvell, E. Frances, 177, 185
Stokstad, E. L. R., 72,, 74, 79
Stowell, R. E., 229, 260
Straus, N. P., 265
Street, H. R., 90
Strong, L. C., 229, 261
Stumpf, P. K., 122, 123, 133
Sturtevant, A. H., 222, 228, 239, 245, 258,

260, 263
Subbarrow, Y., 81
Sullivan, N. P., 43
Suntzeff, v., 229, 260
Svedborg, 17, 48
Swanson, C. P., 228, 260
Sweeney, B. M., 67, 71
Swingle, W. W., 271
Sylverton, J. T., 33, 48
Szent-Gyorgi, A., 90

Tamiya, H., 97, 104

Tatum, E. L., 99, 223, 225, 227, 228, 229,.

236, 259, 260, 261, 262
Taylor, A. Cecil, 173, 177, 186
Taylor, A. E., 48
Taylor, H. S., 4. 25, 48
Taylor, J. F., 132
Taylor, M. J., 237, 262
Ten Broeck, C., 21, 27, 48
Theiler, M., 27, 46

Thimann, K. V., 62, 63, 66, 67, 70, 71, 97
Thomas, H., 119, 133
Thomas, J. Andre, 142, 185
Thomas, L. E., 9, 44
Thompson, D'Arcy W., v, 181, 185
Thorne, R. S. W., 96, 103
Toro, E., 143. 185
Tracy, M. M., 218, 257
Traub, E., 27, 48
Traut, H. F., 257
Troland, L. T., 15, 17, 23, 48
Tsibakowa, E. T., 122, 132
Tunca, M., 44

Tung, T. C., 192, 193, 194- 215
Tung, Y. F. Y., 192, 193, 194, 215
Tweedie, M. C. K., 226, 259

Vaas, K. F., 94, 103

Vanderbroek, 193

van Wagtendonk, W. J., 83, 84, 85

Vazquez, E. M. S. de, 71

Venning, E. M., 273

Vestling, C. S., 132

Voegtlin, C., 130, 133

Vogt, M, 273, 277, 278, 279, 284

Vogt, W., 193, 200, 215

Waddington, C. H., 159, 185

Wallis, E. S., 271

Wang, Hsi, 139, 186

Warburg, O., 106, 115, 121, 123, 133, 221,

258
Wasteneys, H., 7, 8, 48
Waters, Nelson F., 240, 262
Waugh, W. A., 55
Weed, R., 43
Wehmeier, E., 206, 215
Weigert, F., 230, 261
Wcil-Malherbe, H., 133
Weinhouse, S., 133, I34
Weiss, H. A., 134
Weiss, Paul, 58, 139. 141, 146-8, 151, 155,

158, 160, 166, 168, 172, 173, 175, 177, 180,

185, 186
Weisz-Tabori, E., 133
Welch, A. D., 81, 82
Welch, M. S., 70, 132

INDEX

291

Went, F. W., 61, 63, 66, 71
Whelton, R., 100, 105
Whipple, G. H., 17, 18, 46, 124, I33
Whitaker, D. M., 238, 262
White, A., 37. 44, 270, 282, 283, 284
White, R. O.. 70
Wiener, A. S., 223, 259
\Vildman. S. G., 64, 68, 70
Williams, R. J., 73
Wilson, E. B., 249, 264
Wintersteiner, O., 271
Winzler, R. J., 229, 260
Woerdeman, M. W., 206, 215
Wolbach, S. B., 180, 186
Wolpers, C, 52
Womack, M., 89, 90
■Woodruff, A. M., 26, 48

Woods, Mark W., 246, 264
W^oolev, D. W., 86, 87. 88, 89, 90
Wright, L. D., 78, 87, 88
Wright, Sewall, 164, 186
Wulzen, R., 82, 83, 85
Wynne-Jones, 134

Yamamoto, T., 193, 215, 216
Yarwood, E. A., 194, 212
Young, John Z., 167, 183
Yudkin, J., 34, 35, 48, 102, 105
Yule, G. U., 42, 48

Zarudnaya, K., 133
Ziegler, J. A., 132
Zinsser, H., 23, 36, 38, 48
Zwemer, R. L., 275, 284

SUBJECT INDEX

abnormal growth, 217-265; see also sub-
divisions under growth, differential
growth

adaptive enzymes, 18, 34-36, 102

adrenal gland, 266-284

adrenal gland and gluconeogenesis, 279-
284

amino acids, equilibria, 5, 18

antibiotics, 233, 234

antibodies, 14, 18, 36-41, 181, 221, 234, 235,
237, 242-245, 250, 251, 282-284

autocatalysis, 14-19, 31, 32, 254

auxin, 61-71, 97

auxin, mechanism of action, 66

bacteriophage, 29, 94

biological syntheses, 98, 100, 106-134

cancer cell, 217-265; sec also subdivisions
under growth, differential growth

carbohydrates, energy source for growth,
see growth, metabolism in ; glycogen,
synthesis

carcinogens, 220, 229-231, 239, 243-245

catalysis, theory of, 3-18

cell division, 127, 128, 156-164, 171-174,
182, 188

cell elongation, 69-70, 171-173

cellular metabolism, see growth, metabo-
lism in

cellular movements, see movements, or-
ganizational ; mass ; morphogenetic
movements

cholinesterase, 208-209

coupled reactions, 10, 49, 98

dedifferentiation, 141, 142, 237, 238
differentiation, 135-186, 187-216, 218-229,
231, 238

electron microscopy, 50, 52, 53, 91, 137

energetic coupling, 97, 98

enzymes, 3-49

enzyme formation, 9, 17, 223, 231, 232, 239

enzyme inhibition, 107-134

enzymes, pepsin, 11, 20, 21, 162, 245

enzymes, proteolytic, 177

enzymes, radiation effects, 108, no, in

enzymes, — SH groups in, 108, 109, 207

enzyme-substrate complexes, 106-134

enzymes — theoretical catalysts, 3, 4, 6, 7,

106-134
enzymes, trypsin, 11, 20, 21, 87

fats, oxidation and synthesis, 123, 124
fertilization, 189, 191, 199, 201

geotropism, 63

glycogen-syntheses, 7, 272-284

gradients, 191, 192

growth, 161-186, 187-216

growth, adaptations in, 101-103

growth, chromosomes in, 59, 149, 151, 166,

168, I SB, 222, 224, 226, 228, 239
growth, curves of, 91-105
growth, cytoplasm in, 188, 222, 223, 232,

239, 245-249
growth, environmental factors in, 93, 96,

97. 107, 120, 147, 148, 156, 157, 164, 170,

178, 230, 249, 250, 268, 269
growth, enzymotic activity in, 221, 223,

239, 245, 275
growth, eye, 169-173
growth factors, 72-90, 94, 95
growth, genes in, 221-229, 236-240, 245
growth, genetic factors in, loi, 155, 162,

178, 190, 221, 230, 236-240, 245-250
growth, hormones in, 61-71, 120, 126, 140,

179, 180, 218, 220, 224, 251-254, 266-283
growth, immunological reactions in, 221,

234, 235, 237, 240-245, 250, 251, 282-284
growth, inhibition of, 66, 68, 242-245
growth, kinetics of, 91-105, 138, 164-166,

183, 218, 219
growth, microorganisms, 74, 86-88, 91-

105, 182, 219, 225, 226, 231-236
growth, metabolism in, 84, 97, 98, 106-134,

161, 206-211, 218, 229, 238, 275, 276,

279-284
growth, nerve, 166-186, 192, 196, 197, 203,

204, 206, 208-211
growth, nucleus in, 201, 202, 203, 227, 246-

250
growth, osmosis in, 69, 70, 192
growth, oxidation-reduction in, 106-134
growth, plant, 61-71, 182, 246
growth, polarity in, 192
growth, proteins in, 57, 58, 135-186, 223,

240, 253
growth, radiation effects, 130, 131, 138,

178, 222, 224-228, 234, 239
growth, rate of, see growth, kinetics of
growth, regulatory mechanisms of, 126,

177, 178
growth, site of, 166-168
growth, substrates for, 66, 67, 93, 100, loi,

102, 106, 165, 279-284

INDEX

293

growth, vitamins in, 72-90, 99, 180, 274-

284
growth, symmetry in, 189, 190

hybridization and abnormal growth, 238-
240, 245

preformation, 137

protein mctaboHsm, 280-284

proteins, denatured, 13, 14, 16

proteins, fibrous, 13, 49, 51-56, 59, 161, 173

proteins, globular, 13, 54, 55

proteins, hydrolysis, li, 13

proteins, immunological reactions, 16, 21,

24
protein-metalo complexes, iii, 115, 116,

117
proteins, native, 9, 13, 14
proteins, non-amino acid sources of, 12
protein, specificity, 10, 14, 17, 153
protein structure, 54, 55, 112
protein substrate complexes, 107, 108, iii,

112

protein, synthesis, 3-48, 57, 119, 120, 124,

125, 147, 148, 162, 182
proteinogens, 16, 19, 39
protoplasmic streaming, 67, 70

radiation, see enzymes, radiation effects ;

growth, radiation effects
regeneration, 210, 211
ribonucleic acid, 207, 208

induction, 156-161, 221
ingression, 199-201
insulin, 55, 87
isotopes, 12, 50, 275

limb formation, 196, 197
localization, 187-190, 194-196, 199-206

metabolic cycles, see growth, metabolism
in

mitosis, 56, 173, 182, 202, 229

modulation, 139, 140, 141, 144, 153

molecular ecology, 58-60, 146-152, 189

molecular morphology, 49-59, 99, loi, 145,
146, 152-163, 187, 189

morphogenetic movements, 192, 193, 194,
204-206

movements, independent, 198-201

movements, organizational, 195-198

mutations and abnormal growth, 219-240 surface organization, 149-152, 160, 191,

194, 199

neoplastic growth, 217-265 ; see also sub-
divisions under growth, differential tissue culture, 142, 143, I73, 253
growth tissue differentiation, 143-146

neurofibrils, 59, 139, 140, 142, 165 tissue growth, 168-186

nucleo-cytoplasmic relationships, 188-190, transplantation, 142, 143, I53, 172, m,
202, 203 ^98, 237, 238, 242, 251

organization, 187-216
organization, chemical, 206-211
organization, mass movements in, 20i-2o5
organizers, 135-186, 192, 245
orientation of growth, 172-178
oxidation-reduction systems, 112-117

pituitary gland, 272-284

plastein, 8, 9

polarization optics, 53, 59, 137

ur-protein, 12

virus, 3, 15, 16, 222, 225-227, 231, 234,

240-254
virus, adaptations, 28
virus, formation, 22-34
virus, specificity, 26
vital staining, 193, 194, 200
vitamins, see growth, vitamins in

x-ray diffraction, 50, 52, 53, 137

1

K

5

yy\

r.^P-

,^%'s •

%

~\