ADVANCES IN CANCER RESEARCH Volume I
This Page Intentionally Left Blank
ADVANCES IN CANCER RESEARCH EDITED BY JESSE...
11 downloads
766 Views
36MB Size
Report
This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form
ADVANCES IN CANCER RESEARCH Volume I
This Page Intentionally Left Blank
ADVANCES IN CANCER RESEARCH EDITED BY JESSE P. GREENSTEIN National Cancer Institute, US.Public HealthService, Bethesdu, Maryland ALEXANDER HADDOW Chester Beatty Research Institute, Royal Cancer Hospital, London, England
Volume I
ACADEMIC PRESS INC., PUBLISHERS NEW YORK, N.Y. 1953
COPYRIGHT
1963, BY
ACADEMIC PRESS INC. 126 East 23rd Street, New York 10, N.Y.
All Rights Reaerved No part of this book may be reproduced in any form, by photostat, microfilm, or any other means, without written permission from the publishers. Library of Congress Catalog Card Number: 52-13360
PRINTED IN THE UNITED STATES OF AMERICA
CONTRIBUTORS TO VOLUME I C. A. COULSON, Wheatstone Physics Department, King’s College, London, England E. V. COWDRY, Wernse Cancer Research Laboratory and Department of Anatomy, Washington University, St. Louis, Missouri L. DMOCHOWSKI, Department of Experimental Pathology and Cancer Research, School of Medicine, University of Leeds, Leeds, England
W. U. GARDNER,Yale University School of Medicine, New Haven, Connecticut R. J. C. HARRIS,Chester Beatty Research Institute, Royal Cancer Hospital, London, England CHARLES HEIDELBERQER, The M cArdle Memorial Laboratory, for Cancer Research, The Medical School, University of Wisconsin, Madison, Wisconsin
ELIZABETH C. MILLER,The McArdEe Memorial Laboratory for Cancer Research, The Medical School, University of Wisconsin, Madison, Wisconsin
JAMESA. MILLER,The McArdle Memorial Laboratory for Cancer Research, The Medical School, University of Wisconsin, Madison, Wisconsin
W. C. J. Ross, Chester Beatty Research Institute, Royal Cancer Hospital, London, England HERBERT SILVERSTONE, Department of Cancer Research, Medical Research Institute, Michael Reese Hospital, Chicago, Illinois ALBERTTANNENBAUM, Department of Cancer Research, Medical Research Institute, Michael Reese Hospital, Chicago, Illinois RICHARD J. WINZLER,Department of Biological Chemistry, University of Illinois College of Medicine, Chicago, Illinois
V
This Page Intentionally Left Blank
PREFACE Cancer is a disease which has been recognized since ancient times, and which in every generation has claimed many victims of all ages and of all stations in life. Each generation through its medical practitioners has fought it with whatever ideas and tools were available at the time. It is a paradox that, as these ideas have become ever more clear and these tools ever more powerful, the proportion of individuals dying of cancer appears to have risen from year to year. Whatever the reason for these melancholy statistics may be, no matter whether they may be more apparent than real, it is not in the tradition of science to stand idly by in the face of this seeming failure. A society living in the midst of scientific miracles may rightly expect that this disease, like many others, should be comprehended and mastered. Although the comprehension and the successful therapy of a disease are not invariably related, it is the hope of rational men, confident in the scientific method, that if only a phenomenon were understood it could be controlled. Few more notable expressions of this faith in our time are evident than in the public and private support granted in many lands to the subject and field of cancer research. A host of scientific specialties has been marshaled to meet this challenge. As the search for the understanding of cancer proceeds, new scientific approaches develop and are emphasized, and older ones, for the time being perhaps, subside and diminish. The ebb and flow of ideas and experimental approaches in the field of cancer research, as in any of the creative areas of the arts and sciences, is a mysterious and inexplicable process. It is the purpose of the Editors that this and succeeding volumes of this series shall reflect this steady and inevitable march of the tides of our knowledge and increasing understanding. For this task, we must rely upon the generous cooperation of our colleagues in many lands, distinguished authorities in various branches of cancer research, t o review, synthesize, and interpret the advances made in their individual areas of investigation. It is our hope that these pages will reveal from year t o year the gallant and dedicated quest for comprehension and mastery of an ancient and elusive disease.
THE EDITORS
January, 1963 vii
This Page Intentionally Left Blank
CONTENTS CONTRIBUTORS TO VOLUME 1. . . . . . . . . . . . . . . . . . . . . . .
V
EDITORS' PREFACE . . . . . . . . . . . . . . . . . . . . . . . . . . .
vii
Electronic Configuration and Carcinogenesis
BY C . A. COULSON, Wheatstone Physics Department, King's College. London. England I . Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 I1. Historical Survey . . . . . . . . . . . . . . . . . . . . . . . . . 3 I11. Valence-Bond, or Resonance. Method . . . . . . . . . . . . . . . . 8 . . . . . . . . . . . . . . . . . . . . . 20 V . Electrical Index for the K-Region . . . . . . . . . . . . . . . . . . 30 VI . Possible Mechanisms . . . . . . . . . . . . . . . . . . . . . . . 46 V I I . Conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . 52 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
Epidermal Carcinogenesis
BY E. V. COWDRY, Wernse Cancer Research Laboratory and Department of Antomy, Washington University, St . Louis. Missouri I . Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . 58 I1. Experiments . . . . . . . . . . . . . . . . . . . . . . . . . . . 62 I11. Sequence in Experimental Epidermal Carcinogenesis . . . . . . . . . . 66 IV. Microscopic Properties . . . . . . . . . . . . . . . . . . . . . . . 69 V . Chemical Properties of Whole Epidermis . . . . . . . . . . . . . . . 74 VI Integration of Data . . . . . . . . . . . . . . . . . . . . . . . . 85 VII . Indications Concerning Human Epidermal Carcinogenesis . . . . . . . 93 VIII . Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
.
.
The Milk Agent in the Origin of Mammary Tumors in Mice
BY L DMOCHOWSKI. Department of Experimental Pathology and Cancer Research. School of Medicine. University of Leeds. Leeds. England I . Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . 104 I1 The Milk Agent and Genetic Factors . . . . . . . . . . . . . . . . 109 I11. The Milk Agent and Hormonal Factors . . . . . . . . . . . . . . . 119 IV The Milk Agent and Mammary Gland Structure . . . . . . . . . . . 127 V Inherited Hormonal Influence . . . . . . . . . . . . . . . . . . . . 129 VI . Properties of the Milk Agent . . . . . . . . . . . . . . . . . . . . 132 VII . Mammary Tumors in Hybrid Mice and the Milk Agent . . . . . . . . 148 VIII . The Nature of the Milk Agent . . . . . . . . . . . . . . . . . . . 156 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159
. . .
Hormonal Aspects of Experimental Tumorigenesis
BY W. U . GARDNER,Yale University School of Medicine, New Haven. Connecticut I Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . 173 I1. General Statements on Tumorigenesis . . . . . . . . . . . . . . . . 174 ix
.
X
.
CONTENTS
I11 Types of Experimental Hormonal Imbalances . . . . . . . . . . . . . IV . Influences of Differences in “Substrate” on Differences in Response . . V. OvarianTumors . . . . . . . . . . . . . . . . . . . . . . . . . VI Testicular Tumors . . . . . . . . . . . . . . . . . . . . . . . . . VII . Adrenal Tumors . . . . . . . . . . . . . . . . . . . . . . . . . . VIII . Pituitary Tumors . . . . . . . . . . . . . . . . . . . . . . . . . I X . Lymphoid Tumors . . . . . . . . . . . . . . . . . . . . . . . . . X . Uterine Tumors . . . . . . . . . . . . . . . . . . . . . . . . . . XI. Mammary Glands . . . . . . . . . . . . . . . . . . . . . . . . . XI1. Hormones in Relation to Tumors of the Secondary Sex Organs of Males . XI11 Other Tissues or Organs in Which Sex or Sex Hormones Modify the Appearance of Tumors . . . . . . . . . . . . . . . . . . . . . . XIV Urinary Tract . . . . . . . . . . . . . . . . . . . . . . . . . . XV . General Discussion . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . .
.
. .
.
178 180 184 194 198 200 204 207 211 219 220 221 221 223
Properties of the Agent of Rous No 1 Sarcoma
BY R . J . C. HARRIS,Chester Beatty Research Institute, Royal Cancer Hospital, London, England I . Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . 233 11 Agent and Host . . . . . . . . . . . . . . . . . . . . . . . . . . 235 I11 Agent and Malignant Cell . . . . . . . . . . . . . . . . . . . . . 243 I V . Isolation and Properties of Rous No . 1 Agent . . . . . . . . . . . . . 250 V. Relationship of Rous Agent to Fowl Tumors and Leucoses . . . . . . . 261 V I Origin of Rous Agent . . . . . . . . . . . . . . . . . . . . . . . 264 VII Conclusion. . . . . . . . . . . . . . . . . . . . . . . . . . . . 265 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . 265
. . . .
Applications of Radioisotopes to Studies of Carcinogenesis and Tumor Metabolism BY CHARLES HEIDELBERGER, The McArdle Memorial Laboratory for Cancer Research, The Medical School, University of Wisconsin, Madison, Wisconsin I . Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . 274 I1. Metabolism of Carcinogenic Hydrocarbons. . . . . . . . . . . . . . 279 I11. Other Carcinogenic Compounds . . . . . . . . . . . . . . . . . . . 290 IV Oxidative Metabolism of Tumors . . . . . . . . . . . . . . . . . . 293 V. Incorporation of Amino Acids into Tumor Proteins . . . . . . . . . . 301 V I Nucleic Acids . . . . . . . . . . . . . . . . . . . . . . . . . . . 309 VII . Miscellaneous Compounds . . . . . . . . . . . . . . . . . . . . . 326 VIII Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . 333 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . 334
. . .
The Carcinogenic Aminoazo Dyes
BY JAMES A . MILLERand ELIZABETH C . MILLER,The McArdle Memorial Laboratory for Cancer Research, The Medical School, University of Wisconsin, Madison, Wisconsin I . General Introduction . . . . . . . . . . . . . . . . . . . . . . . 340 I1. Early Observations . . . . . . . . . . . . . . . . . . . . . . . . 341 I11 4-Dimethylaminoaeohensene and Its Derivatives . . . . . . . . . . . 342
.
CONTENTS
xi
IV . Studies on the Hepato-Carcinogenicity of Other Azo Dyes . . . . . . . 379 V . On the Mechanism of Azo Dye Carcinogenesis . . . . . . . . . . . . 383 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . 390
.
The Chemistry of Cytotoxic Alkylating Agents
BY W C. J . Ross, Chester Beatty Research Institute, Royal Cancer Hospital. London. England I . Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . 397 I1. 2-Chloroethyl Sulfides (Sulfur Mustards) . . . . . . . . . . . . . . 399 I11 ZChloroethylamines . . . . . . . . . . . . . . . . . . . . . . . . 411 IV . 1.2-Epoxides . . . . . . . . . . . . . . . . . . . . . . . . . . . 429 V . Miscellaneous Agents . . . . . . . . . . . . . . . . . . . . . . . 436 V I . Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . 437 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . 446
.
Nutrition in Relation to Cancer BY ALBERT TANNENBAUM and HERBERT SILVERSTONE. Department of Cancer Research. Medical Research Institute. Michael Reese Hospital. Chicago. Illinois Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . 452 I . Some General Considerations . . . . . . . . . . . . . . . . . . . . 453 455 I1. Genesis of Tumors . . . . . . . . . . . . . . . . . . . . . . . . . . I11 Growth of Tumors . . . . . . . . . . . . . . . . . . . . . . . . 481 IV . Nutritional State and Cancer in Man . . . . . . . . . . . . . . . . 487 491 V . Conclusions and Commentary . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . 497
.
Plasma Proteins in Cancer
BY RICHARD J . WINZLER.Department of Biological Chemistry. University of Illinois College of Medicine. Chicago. Illinois I . Some Methods of Study of Plasma Proteins . . . . . . . . . . . . . 506 I1. Alterations of Plasma Proteins in Neoplastic Disease . . . . . . . . . . 513 I11. Plasma Enzymes and Inhibitors . . . . . . . . . . . . . . . . . . . 529 IV . Protein-Bound Carbohydrate . . . . . . . . . . . . . . . . . . . . 535 V. Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . 538 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . 539 AUTHOR INDEX . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
549
SUBJECT INDEX . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
573
This Page Intentionally Left Blank
Electronic Configuration and Carcinogenesis C. A. COULSON Whealstone Physics Department, King’s College, London*
CONTENTS
Page I. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 11. Historical Survey. . . . . . ............. 3 1. The K-Region.. . . . . ............. 3 2. Schmidt’s Box Mode ............................. 4 3. Svartholm’s Introdu ............. 6 7 4. The Work of Pullman, Daudel, and Others.. . . . . . . . . . . . . . . . . . . . . . . . . 111. Valence-Bond, or Resonance, Method.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 1. u and IT Electrons.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 2. Valence-Bond Structures. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 3. Some Complicating Features in the Valence-Bond Method.. . . . . . . . . . . 13 4. Derived Quantities.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 A. Bond Order.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 B. Charge Density and Distribution.. . . . . . . . . . . . C. Freevalence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16 5. Methyl Substitutions, Hyperconjugation, . . . . . . . . . 6. Aza Replacement. .. ...................... 18 7. Penney Bond Orders.. . .......... . . . . . . . . 19 IV. Molecular-Orbital Method. . . . . . . . . . . . . . . . . . . . . . . . . 1. Molecular Orbitals.. ... .... . . . . . . . . . . . . . . 20 2. The LCAO Representation.. . . . . . . . . . . . . . . . . . . . . 3. Fundamental Magni ...................... 24 4. Some Particular Results.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26 5. Polarizabilities. . . . . . . . . .................... 26 6. Hyperconjugation. .... ..................................... 27 7. Direct Tests of Theory ..................................... 28 V. Electrical Index for the .................. 1. Electrical Index. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2. Bond Orders in the 3. Free Valence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33 4. Charge on the Atoms.. . . . . . . . . . . . . . . . . . 33 5. Pullman’s Work on 6. Molecular-Orbital Indices. . . . . . ........................ 40 7. Resonance Energy, 8. Electronic Excitation. . . . . . . .................................. 43 VI. Possible Mechanisms. . 1. Interpretation of Pr
* Now at the Mathematical Institute, Oxford. 1
2
C. A. COULSON
2. Advantages of the K-Region.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3. Significance of the Total Charge on the K-Region.. . . . . . . . . . . . . . . . . . . 4. Some Speculations... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . VII. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Page
46 47 50 62 54
I. INTRODUCTION This account of the possible relationship between electronic configuration and carcinogenesis falls naturally into six sections. These are : (1) historical survey, (2) the quantum-mechanical resonance method of describing large unsaturated molecules, (3) the alternative molecularorbital method, (4) numerical applications of these two methods, ( 5 ) possible mechanisms of carcinogenic activity, and (6) conclusions. Of these (2) and (3) are not solely concerned with carcinogenic properties, but no reasonably simple and straightforward account of the two methods discussed seems to be available. This is particularly important if we bear in mind that both methods are approximations, whose reliability is not fully established. We shall see that although there is a considerable measure of mutual agreement, there are several places in which they disagree. As a result of this, there is still a certain amount of personal liberty possible in the interpretation of the calculations. Indeed, it may be admitted a t once that no final or complete account of the relation between electronic configuration and carcinogenesis has, or can, yet be given. The present account will therefore attempt to stress both the undoubted successes of the theory and at the same time some of its equally patent inadequacies. The whole field is a singularly interesting one, because it represents one of the very first serious attempts to relate what are obviously complex biological phenomena to quantum-mechanical principles. Its significance lies not only in the fact that carcinogenic potency (or otherwise) has been correctly predicted on purely theoretical grounds for quite a large number of molecules which had not, at that time, been investigated experimentally; but also in the fact that it opens up new fields of enquiry and discovery, and itself suggests new interpretations: this must inevitably lead to a better understanding of such phenomena as cocarcinogenesis, anticarcinogenesis, drug action, chemical mutations, and the mechanism of estrogenic and other hormone activity. There is little doubt but that in the next two or three decades we may expect enormous and far-reaching developments in these fields; and the developments are likely to be considerably facilitated if there is a fairly large body of experimenters who are familiar with the quantummechanical basis on which, quite evidently, the action of these chemicals must ultimately depend. It is for this reason that (2) and (3) have been
ELECTRONIC CONFIGURATION AND CARCINOOENESIS
3
written in their present form. Those who are already familiar with wave-mechanical methods can pass straight to the remaining sections of this review. 11. HISTORICAL SURVEY 1 . The K-Region
It was established about twenty years ago that certain polycyclic hydrocarbons have the property of inducing cancerous growths, either when painted on the skin or injected into the animal concerned. Nearly
Phenanthrene (1)
all these molecules could be regarded as derivatives of phenanthrene (I), though phenanthrene itself is not active. The staggered ring system of phenanthrene seems to be much more effective in promoting carcinogenic activity than does the straight type of annelation shown in anthracene (11),derivatives of which are hardly ever active. (An exception is 9,lOdimethylanthracene, which is the simplest known carcinogen among the hydrocarbon family. No satisfactory explanation of this has yet been suggested.) It is customary now t o distinguish between the anthracene -and phenanthrene-type skeletons by saying that in the latter the 9,lO-region has a character quite distinct from anything in the former. Following Mme. Pullman ( 1 9 4 6 ~1947c) ~ we shall call this the K-region. The K-region is easily recognized. Thus in the extremely important parent hydrocarbon 1,2-benzanthracene (111) there is one K-region, as shown by a thick line: in 3,4-benzphenanthrene (IV) and 1,2,5,6-dibenzanthracene (V) there are two K-regions. Our study of these molecules will largely consist of an enquiry concerning the electronic distribution in these regions and of the ways in which this distribution is affected by
1 ,l-Benxanthracene (111)
3,4-Benzphenanthrene (IV)
1,2,5,6-Dibenzanthracene (V)
4
C. A. COULSON
substitution (particularly methyl substitution), or by an aza replacement such as occurs when benzanthracene (111) is compared with benzacridine
(VI)-
8’
3,4-Benzacridine* (VI)
A t this stage reference must be made to the exceedingly valuable compilation of Hartwell (1941), who has listed all the available published (and unpublished) experimental results with molecules of the kind we are interested in. Practically all the experimental conclusions mentioned in this review are quoted from Hartwell’s quite invaluable tables. To this, and to the review article by Badger (1948) the writer is greatly indebted. 2. Schmidt’s Box Mode2 The first attempt to explain the significance of the K-region was due to Schmidt (1938, 1939a,b,c, 1941). In essence this amounted chiefly to an explanation, largely in classical terms (Schmidt’s wave-mechanical considerations were somewhat speculative, and have never gained general acceptance), to show why this region might be expected to behave differently from any other region in an aromatic system. Schmidt started with the hypothesis that there were certain regions, or groupings, of peculiar stability, which would try to preserve their character in any chemical reaction, so far as that was possible. Such groupings include benzene and naphthalene, but they do not include open-chain structures such as butadiene. The complete molecule can then be cut up into separate units as in Fig. 1, which suggests that phenanthrene and benzanthracene (a) should exhibit peculiarly strong reactivity precisely in their K-regions, but anthracene (a) should not. Rather is it that the reactivity of anthracene should reside in the meso, or 9,lO-positions. Now in the first and third of these molecules there is no doubt about the way in which the “boxes” should be drawn. Any attempts, such as those shown for anthracene (b), (c) to cut up the molecule in any other way, involves using an open-chain unit such as butadiene or ethylene. On the other hand, the rules do not prohibit the division shown in benz-
* Several distinct numbering systems are in use for this and other related molecules. For that reason we shall usually place our numbering scheme on the diagrams of the molecules in the text where they h t occur.
ELECTRONIC CONFIGURATION AND CARCINOGENESIS
5
anthracene (b), and we are left wondering whether there is, or is not, a special K-region in this molecule. It is certainly true that the units (or boxes) chosen by Schmidt are very stable ones, when they occur alone. Thus benzene has a resonance energy of 38 kcal./mole, and naphthalene nearly twice as much, compared with about 6 to 8 kcal./mole for butadiene. But there is no direct correlation between the behavior of the units when they occur as distinct molecules and when they occur as parts of a larger system. The evidence from ultraviolet absorption spectra suggests that molecules of this kind behave as single systems, whereas on Schmidt’s view we might have expected each component unit t o absorb separately from the rest. The
Phenanthrene
Benzanthracene (a)
Anthracene (a)
Anthraccne (b)
Benranthracene (b)
Anthracene (c)
Fro. 1. The “Box Model” of Schmidt.
theoretical basis for this theory is therefore destroyed, and, in view of the uncertainty about drawing the boxes and the almost complete lack of quantitative deduction that can be drawn from it, we can regard the theory merely as suggestive. It suggests (a) that the possession of a phenanthrene-type region is likely to make the behavior of a molecule different from that of molecules without this region, (b) that the K-region bond would be expected to be the seat of the enhanced reactivity, and (c) that in cases where no unique division into boxes is possible, there may be a double type of reactivity. This would mean that in benzanthracene, for example, both the K-region 3,4-and also the meso-positions 9,10 would have characteristic and important properties. We shall see later that all these suggestions correspond to the truth. In particular phenanthrene does differ from most aromatics in adding hydrogen at the 9,lO-positions (we shall discuss the so-called osmium tetroxide reaction later), and this addition is much easier than in benzene. Also the meso-positions in anthracene, and in benzanthracene too, are outstandingly important, both chemically (anthracene forms a photo-
6
C. A. COULSON
oxide in which -0-Olinks these two atoms) and carcinogenically (substituents at these positions tend to be more effective than at other places in the skeleton). 3. Svartholm’s Introduction of a Electrons
Schmidt’s work was unsatisfactory because it did not do justice to our knowledge of the electronic structure of molecules. The next stepand in some ways the most significant-was due to Svartholm (1941), though he does not appear to have fully appreciated the significance of his work. We shall not now describe in detail the analysis that he gave, because it has subsequently been developed much more fully by Mme. Pullman and R. Daudel, and will form the substance of Sec. 111. But, in essence, Svartholm accepted the description of an aromatic molecule in the form in which it had been given by Pauling and his co-workers. This treated the molecule as if it could be described as a simultaneous superposition of several so-called structures. These structures were nothing other than ways of drawing the necessary number of single and double bonds, as in the Kekul6 and Dewar structures of benzene. Wave mechanics showed that in the superposition some of these structures were more important than others, and thus that certain bonds were more nearly double bonds than were the others. Pauling, Brockway, and Beach (1935) had already used the concept of fractional bond order, which was implicit in Svartholm’s work. Svartholm showed that in phenanthrene and similar molecules the K-region did resemble a double bond much more closely than did any other bond. From this he concluded that this region could more easily add other groups or attach itself to the cell than could other regions. It was not unreasonable, therefore, to regard it as the seat of carcinogenic activity. A second conclusion followed, though it was less clearly stated. I n anthracene and other similar molecules there appeared to be a considerable unused bonding power a t the mesopositions 9,lO. These atoms would be expected also to be reactive for addition, so that, in cases such as benzanthracene where there was both a K-region and a pair of meso-atoms, we might expect one to be able to influence the other. The significant advance made by this work can be summarized: (a) it made use of a quantum-mechanical description of molecular structure ; (b) it showed that there really were certain special electrical properties associated with the K-region, and that these were capable of being calculated theoretically; finally (c) it threw emphasis on to the behavior of the electrons responsible for conferring double-bond character on the bonds of an aromatic molecule. These are the u electrons, which we shall describe more fully in Sec. 111. All this carried the problem into
ELECTRONIC CONFIGURATION A N D CARCINOGENESIS
7
the region of molecular structure, where, at this time, considerable advances both of a fundamental and a technical kind, were being made.
4. The Work of Pullman, Daudel and Others The next stage, which includes much the greater part of the work which we are reviewing, is due to the French school of theoretical chemists, Dr. and Mrs. Daudel, Dr. and Mrs. Pullman, and their colleagues. There has been a good deal of double publication, so that it is not easy always to be sure of the priorities, But the first full-scale report of the application to carcinogenic problems is given by Mme. Pullman (1945a,b, 1946a,c, 1947~). This includes the effects of aza substitution in the carbon skeleton and of methyl substitution around the periphery of the molecule. In order to develop this theory, however, it was necessary first to investigate the numbers of structures (in the sense used by Slater and Pauling-see Sec. 111)of given type; and in this work R. Daudel and Mme. Pullman largely published together (A. Pullman, 194613; B. Pullman, 1946; Daudel, 1946a; Daudel and Pullman, 1945b,c,d,e, 1946). More recently a review of some aspects of this problem has been given by Daudel and Daudel (1950), and a semihistorical account of much of the earlier work has been provided by R. Daudel (1946b). Very recently also a list of molecules for which carcinogenic potency has been predicted on theoretical grounds has been given by Daudel and Daudel and BuuHoi (1950). These workers effectively took Svartholm’s model, and made it quantitative. This involved, on the one hand, a very laborious solution of a large number of sets of simultaneous equations, and on the other hand, an estimate of the way in which the greater electronegativity of a nitrogen atom as compared with a carbon attracted electrons away from the K-region on to the nitrogen : and in which the electron-donating character of a methyl group replenished the supply of electrons in this region. Running through all this work was the conviction that some threshold existed for this region, and if what we may describe as the “electrical index” of this region exceeded the threshold value, the molecule would be carcinogenic: otherwise it would not be. The precise nature of what constitutes this electrical index is one of the major problems to be solved. Even now our judgment about it is continually changing as further experimental evidence accumulates. In Svartholm’s early work it was supposed that the bond order provided the effective index; in Pullman’s work on unsubstituted hydrocarbons it was the sum of bond order in the K-region and the two free valences at the carbons of this region; in substituted hydrocarbons and aza molecules it was the so-called total charge of this region (this is essentially the sum of bond order, free valences, and
8
C. A. COULSON
mesomeric charge migrations). More recently free valence alone has been suggested. But in no case is an unequivocal argument forthcoming that will deal with all possibilities. Some of the difficulties will be dealt with in Sec. V. They are of two kinds. I n the first place there is the definition of the electrical index in terms of the electronic distribution in and near the K-region. This is a matter of our own choice, and our object is clearly t o find that particular definition that seems t o fit the observed facts best. The second type of difficulty arises because, on account of the enormous complexity of a molecule with some twenty carbon atoms, the calculation of the quantities out of which the electrical index is built up (charges, bond orders, free valency, etc.) can itself be achieved only by making approximations. It is important, therefore, to distinguish between contributions to the first of these problems, which might be called the discovery of correlations, and contributions to the second, which may be called the improvement of technique. We shall see later that failure to recognize the inadequacies of some of the earlier techniques has not infrequently vitiated detailed numerical conclusions regarding the index. And the constant need to keep in mind the desirability of improving technic is shown by an entirely new discussion of charge migrations due to Nebbia (1950; Prato and Nebbia, 1950), and some much more detailed calculations of charge migration and bond order changes on substitution, recently completed by H. H. Greenwood (1951). We cannot, however, appreciate these arguments till we have explained the basis on which the bond orders, etc., are calculated. This we must now do. 111. VALENCE-BOND, OR RESONANCE, METHOD 1 . u and ?r Electrons*
The usual formulation of a chemical bond is to.describe it as the result of the pairing together of two electrons. In the isolated atoms from which the bond is formed, these electrons will occupy certain patterns, which may be calculated with reasonable accuracy. An approximate, though not quite rigorous, description of these atomic patterns (or atomic orbitals, as they are usually called) is obtained by imagining the electron spread out in the form of a cloud. The density of this charge cloud may be found by calculation. It is merely the square of the wave function, which is itself nothing more than the appropriate solution of the Schrodinger wave equation for the electron. The details of the whole pro-
* For further information Bee L. Pauling, Nature of the Chemical Bond, Cornell University Press, 1937, or C. A. Coulson, Valence, Oxford University Press, 1952.
ELECTRONIC CONFIGURATION AND CARCINOQENESIS
9
cedure do not matter; what does matter is that when the two atomic orbitals are “paired together,” there results a new charge cloud. This is the charge density for the bond, and with chemical single bonds it possesses two important properties: (a) It is almost completely localized in the region between the nuclei. (b) It is axially symmetrical around the line of the bond. These electrons are called u electrons. Normal single bonds (Hz,HC1, CHI) are of this type. They confer characteristic bond properties, such as length, dipole moment, ultraviolet absorption frequency, upon the bond in question, and the almost complete localization of the electrons makes each bond largely independent of the rest of the molecule.
(4
(b)
FIG.2. (a) An isolated atomic T orbital. (b) The formation of a R bond by pairing of two such orbitals.
But there is another type of electron cloud of much more interest to us. In the individual atom the density pattern consists (Fig. 2a) of two regions, or lobes, rather like the two halves of a complete dumbbell. In a normal double bond, such as occurs in ethylene, there is, first, a u bond of axially symmetric type, and then, superposed upon it, a ?r bond is formed (Fig. 2b) by the pairing together of two of these atomic patterns, in which the axes of the “dumbbells” are parallel and perpendicular to the plane of the CZH4 nuclei. In Fig. 2 the shaded regions denote those regions where the electronic charge cloud is chiefly concentrated. It can be seen from Fig. 2b that the *-bond pattern has zero density along the nuclear axis, which is just where the u-bond pattern is most concentrated. It is for this reason that we can attempt to deal with the two types of electron quite independently. The whole of our subsequent work will be concerned with the A electrons, because they confer aromatic character on a molecule, and, as Svartholm originally suggested, they give rise to the characteristic properties of the K-region. 2. Valence-Bond Structures
In ethylene (Fig. 2b) there was no difficulty in pairing the electrons in the formation of bonds. There were two u electrons to form a u bond,
10
C. A. COULSON
and two x electrons to form a ?r bond. But in an aromatic molecule like benzene this is no longer so straightforward. We can approach the matter most easily if we consider first the situation depicted in Fig. 3 where, on the left, we have three atomic dumbbell orbitals ABC on adjacent atoms. The three patterns are similar and parallel, pointing a t right angles to the line of centers. We are to make molecular patterns from these three atomic ones. Clearly one way is to link A and B, and so make a molecular pattern such as (3b), in which C is left as in an
(4
(4
FIQ. 3. (a) Three isolated unpaired atomic pairings to form molecular patterns.
(4 T
patterns. (b) (c) Alternative
@ ...
i
(4
(b)
...
..:
i
.
:
!
..
,
(4
FIG.4. Benzene pairings. (a) Separate unpaired r electron patterns. (b) (c) istinct alternative pairings to form molecular patterns.
isolated atom. But equally clearly there is another alternative (3c) in which we link B and C, and leave A isolated. Nobody can say that one of these pairing schemes is more favored than the other. In a word, the two possibilities (b) and (c), which Pauling calls valence-bond structures, or simply structures, are both equally possible. There is no unique pairing of the atomic electrons. And in particular, atom B must be supposed to pair with both of its neighbors A and C. This situation is enhanced in benzene and the aromatic systems generally. For in benzene, which is a planar molecule, we first allot electrons to the localized C-C and C-H bonds; after this we have six electrons left over, one on each carbon atom. In the isolated atoms they would occupy orbital patterns represented by the dumbbells of Fig. 4a. These are all parallel to each other and perpendicular to the molecular plane. Now there are
ELECTRONIC CONFIGURATION AND CARCINOGENESIS
11
obviously two distinct ways of joining these in pairs: just as we drew the patterns 3b and 3c, we can draw 4b and 4c. Neither of these is adequate by itself, and the mathematical reasoning of quantum theory tells us that the complete Schrodinger wave function for the molecule must be represented, at very least, by a sum of the functions representing the separate structures. These two structures are obviously the wavemechanical versions of KekulB’s bond patterns of 1865, and we should represent them by
WII)
respectively. The dynamic oscillation that he postulated is now replaced by the assertion that the complete wave function can be thought of as a simultaneous superposition of the two components. The result of this superposition (which we call resonance among the structures) is t o provide us with something quite different from either structure alone, though possessing certain characteristics of both of them, more or less in the way that a superposition of red and green gives a new color brown, distinct from either of its components. There are other ways of pairing these ?r electrons. For example we have three Dewar structures which we could represent as and
(VIII)
a
and
In these there is one long, or ineffective, bond. In such cases the pairing of these two orbits is purely formal since the internuclear distance is so large that no effective binding energy results. The significance to be attached to these structures is that, in them, the electrons a t the ends of a long bond may be treated as if they were almost free from constraint and could be used immediately for some other purpose such as starting a reaction. There are other structures that we could imagine, such as but it turns out (see later) that the 2 KekulB 3 Dewar structures
0,
+
provide a complete and sufficient collection and that any other structure could be expressed as some combination of these five. For that reason there is no point in introducing them, and we do best to adhere t o the chemically significant and relatively simple structures already described. It is convenient to have some terminology with which we may refer to these structures. We do this by the number of long bonds that they possess. Thus the Kekul6 structures of benzene are unexcited structures,
12
C. A. COULSON
the Dewar ones are singly excited, and so on. Figure 5 shows an assortment of structures for naphthalene, in which the total of forty-two independent structures divides into unexcited, singly, doubly, and triply excited members. The chief business at this stage is to determine the weights with which each of these structures enters the complete wave function. Each
Unexcited, or Kekule type
Fret excited
Total Number of This Type
Total Weight
3
0.55
a
@ J
0.39
Doubly excited
0.05
Triply excited
0.002
Total
PIO. 5.
1.oo
42
Structures in naphthalene.
structure is represented by a wave function. If these are the complete wave function is written
*
= a&+
+
*
'
$z
.- .,
*
(1) give the weights. They are usually normalized u2$2
The ratios uI2, u2), * - to unit sum. Thus benzene is written (in its ground state) 9 = 0.62 {$I
+ + 0.27 + +
(2) where $1 and +a are the Kekul6 structures, $8, $4, and $6 the Dewar structures. This means that the weights of a single Kekul6 and Dewar structure are (0.62)2 and (0.27)2, i.e., 0.39 and 0.07 respectively. The two Kekul6 structures amount in all t o 0.78, i.e., nearly 80% of the total, with the three Dewar structures contributing merely 21% in all. The corresponding values for naphthalene are shown in Fig. 5. In the process of determining the coefficients al, u2 - * in (l), the energy of the T electrons in the molecule is obtained. This is always less $2}
{$a
$4
$61,
-
ELECTRONIC CONFIGURATION AND CARCINOQENESIS
13
than that associated with the most stable structure (in benzene a Kekul6 structure). The amount by which the energy falls below that of the lowest component structure is the resonance energy of the molecule. With condensed aromatic hydrocarbons this amounts to about 36 kcal./mole per ring, being a little larger for the staggered rings like phenanthrene and chrysene than for straight ones like anthracene and napthacene. 3. Some Complicating Features in the Valence-Bond Method Several complicating features make this method much less easy in practice than might have been supposed. In the first place the choice of structures is not unique. Any set of indepenaent ones will do, though, if we are not careful later, they will give us quite different values for the desired quantities such as bond order and free valence. When there are odd-numbered rings, a8 in fulvene or azulene, no entirely satisfactory choice is possible. But when there are only even-numbered rings, as usually occurs, the difficulty is less severe. A systematic procedure has been developed by Moffitt (1949). The second difficulty lies in the enormous number of structures. For a molecule with 2n carbon atoms there are (2n)!/(n)!(n l ) !of these. Thus for benzene where 2n = 6, the number is 6!/3!4! = 5. For naphthalene we have 10!/5!6! = 42, and for anthracene 14!/7!8! = 429. This very soon introduces terrific complexity. Thus, to deal with the anthracene problem we require to solve a set of 429 simultaneous equations. This is a hopeless proposition. Some simplification must be introduced. Following a suggestion due to Pauling and Wheland (1933) all the Kekul6 structures are given the same weight. So are all the first excited, second excited, and so on. Without such simplification, Mme. Pullman (1947~)could never have dealt with molecules as large as beneanthracene. But it is important to realize that it is an approximation, which has been shown by Sherman (1934) for naphthalene to be correct only in its main conclusions; many details are incorrect. Certain other approximations turn out to be necessary even in setting up the simplified set of equations. Among these other approximations lies the question of which types of excited structures may be neglected. Figure 5 shows that in naphthalene the total weight of the triply excited structures is so small that these may evidently be neglected. But in naphthacene there are structures
+
Naphthacene
(1x1
C. A. COULSON
14
as much as sevenfold excited. It seems highly probable that in larger molecules the more highly excited structures become collectively more and more important. This makes the selection of structures a bit hazardous. Figure 6 (adapted from Pullman, 1947c) suggests that for molecules of carcinogenic interest, i.e., those containing four or five fused rings, the most important are the doubly and triply excited structures. 80r '
1
2
3
4
no. of condensed rings in molecule FIQ.6. Curve showing the total weight of all structures of a given class (degree of excitation) in terms of the number of rings in condensed polynuclear hydrocarbons. Curve 1 is for unexcited (KekulB) structures; curve 2 is for singly excited structures; curve 3 is for second-excited structures.
The number of these is prohibitively large (649 doubly excited structures in naphthacene), an unpleasant situation which adds greatly to the labor of computation. The third difficulty arises from the neglect of ionic structures. All the previous structures have been covalent, with exactly one n electron on each nucleus. But it was shown, a long time ago, by Sklar (1937),
0
that structures such as \ / play a part. These structures will occur in pairs, with reversed charges, and in the case of benzene will, by symmetry, not introduce any unevenness in the final charge distribution. But they are much more important in substituted molecules, or heteronuclear molecules like pyridine, where structures such as
ELECTRONIC CONFIGURATION AND CARCINOGENESIS
15
must play quite a large part. No satisfactory method yet exists for incorporating these ionic structures in systems bigger than benzene (Craig, 1950). The large but unavoidable amount of empirical character in this discussion of large molecules should now be apparent.
4. Derived Quantities Our chief interest in the calculations just described arises from the fact that they allow us t o derive certain magnitudes. It is with these magnitudes that any carcinogenic properties are to be correlated. The important magnitudes are bond order, charge and free valence. Let us describe them in turn. A. Bond Order. We have seen that our C-C bonds acquire a partial double-bond character as a result of resonance among the various allowed structures. Pauling, Brockway, and Beach (1935) suggested that a measure of this double-bond character could be obtained by adding together the weights of all the component structures in which the bond in question appeared as a double bond. An example will make this clear. A particular bond in benzene appears double in just one of the Kekul6 structures VII and just one of the Dewar structures VIII. The r-bond order is therefore 0.39 0.07 = 0.46. The total fractional bond order is 1.46; in other words, the percentage double-bond character is 46. Clearly all six C-C bonds in benzene are equivalent. But in naphthalene they are not. Using the weights referred to in Fig. 5 the total bond orders are calculated to be as in XI.
+
Bond orders in naphthalene (XI)
Here, and elsewhere, the bond order value is written along the relevant bond, and bonds not marked may be found by symmetry. In larger molecules where some of the approximations referred to on p. 13 have to be made, a good deal of analysis is needed t o determine in how many structures of each type a given bond appears double. But the requisite analysis has been provided by Wheland (1935) Pullman (1946c, 1947), and Daudel and Pullman (1945a,c,d,e, 1946). It should be mentioned here that in some of the early work of the French school, the term bond order was not used; instead values were given for the “charge de liaison.” This was simply twice the r-bond order. The justification for this term was that a complete pair of r electrons, as in ethylene, would give rise to a .rr bond of order unity.
10
C. A. COULBON
So, on a proportional basis, a fractional bond order p might be thought of as if it involved 2 X p electrons. The term bond order is now universally used. B. Charge Density and Distribution. In the case of hydrocarbons such as benzene, the ti charge must be evenly distributed by symmetry. This will always appear to be the case so long as we confine ourselves to structures such as VII, VIII, and those in Fig. 5. But as soon as we introduce structures such as XI1 or XI11 this is no longer true, and we
Naphthalene (ionic structures) (=I)
Azulene (ionic structures) (XIII)
do find certain charge migrations. Once again the total weight of structures of a particular kind will tell us how large these migrations are (Craig, 1950). Unless there are odd-numbered rings, however (Brown, 1948; B. Pullman et a2.,1950), these migrations are not very large, of the order of 1 or 2% of an electronic charge (f0.013e in butadiene). Except for some work by Nebbia (1950), which we shall discuss in Sec. IV, these shifts have generally been neglected. But with heteronuclear molecules and when methyl substitutents are present (see Sec. 11, 5, 6), this is obviously no longer the case, and structures analogous to X will appear, sometimes with a relatively large weight. C. Free Valence. Finally there is the free valence. On the old basis of Thiele and Werner, this could be regarded as the unused bonding of an atom. Thus it could be associated with the total weight of all the structures in which an ineffective, OF long, bond terminated on the atom. In benzene the only long bonds are in the Dewar structures, and each atom only has one such structure giving it a long bond. Thus the free valence at each carbon atom is 0.07. This was essentially the idea of Svartholm (1941), though it was considerably refined and systematized by Daudel and Pullman, (1945a) by Moffitt (1949), by R. Daudel and colleagues (1949), and by A. Pullman and B. Pullman (1949). In earlier work, on account of the fact that each long bond could be interpreted to mean an almost free electron a t each end, the term “charge du sommet” was used instead of free valence. But this latter term is now used exclusively. (An alternative definition of free valence will be mentioned in Secs. 111, 7 and IV, 3.) The combination of bond orders, charges, and free valences is often referred to as the “molecular diagram” of the molecule. The molecular diagrams of benzene and naphthalene* are shown in XIV and XV. On
* Calculated from the weights in Sherman (1934).
ELECTRONIC CONFIQURATION AND CARCINOQENESIS
17
account of the neglect of all ionic structures in the latter, all the 7r charges in both molecules are exactly equal to 1. When comparing molecular
(yJl 0.16
0.06
0"
1.46 0.07
Benzene (XW
?
1.59 0.10
Naphthalene
Molecular diagrams
(XV)
diagrams given by different authors it is most important to be sure of the particular approximations used in that diagram. Quite different values may be obtained by different approximations at various stages of the calculation. In general, however, the main chemical inferences (see Sec. VI) are unaffected by this unfortunate plurality of numerical values. 6. Methyl Substitutions, Hyperconjugation
It is known that methyl substitution of a hydrocarbon may profoundly alter its carcinogenic property. There are two mechanisms whereby a methyl group affects the electronic distribution in the rest of the molecule. Together they are described as hyperconjugation (Baker, 1951; Crawford, 1949; Mulliken et al., 1941). In the fist of those mechanisms, allowance is made for the well-known fact that a methyl group is an electron donor and is able to provide additional charge for the aromatic framework. Structures such as XVI(b) would be needed, in addition to normal structures XVI(a), to allow for this charge migration. In the second mechanism, no charge migration is allowed, but there is an increased conjugation (hence the name hyperconjugation). By analogy with NrC-, Mulliken argued that we might write H3=C-, in which the methyl group behaved as if it possessed a pseudo-triple bond. Thus there is a parallel between methyl cyanide Ha=C-C==N and cyanogen This parallel (which must not be pressed too closely) NrC-C=N. suggests that we could get conjugation of the pseudo-triple bond with adjacent aromatic groups. In the case of toluene we expect structures such as XVI(c) to express this additional conjugation. Structures (a)-(c) are, of course, only typical of many others thaf can be drawn. The only
(a) Normal
(b) Ionic (XVI) Structures in toluene
(c) Conjugation
difficulty now is to estimate the weights of these structures. Unfortunately this cannot be done by direct calculation, but must be inferred
18
C. A. COULSON
from evidence such as the C-CHs bond length and the dipole moment of the molecule. It will be noticed that all the new structures (b) and (c) show this bond as double. We should therefore expect that its length, which determines its double-bond character, should indicate the total weight of these additional structures. The results however can only be approximate since this length is not known sufficiently precisely and appears to vary somewhat erratically from molecule to molecule. Even this only tells us the sum of the weights of (b) and (c). By making certain assumptions about the distribution of net charge, A. Pullman (1947~)was able to estimate the separate weights of each type of structure. The resultant charge distribution is shown in XVII, where the figures shown at the various nuclei represent the net charge in electrons. Thus the methyl group in toluene would be supposed to carry a net positive charge O.lle. -0.037
6 -0.01
0
CHa + O N 2
-0.030
(XVII) (After Pullman)
Similar analysis may be used for methyl substitution in larger molecules: and the effects of two or more such substitutions may be treated as additive. There is no doubt but that this technique properly describes the phenomenon in a qualitative sense. But it can hardly be said to be satisfactory quantitatively. Thus as stated earlier, (1) the C-Me bond length is not known sufficiently accurately, (2) the argument which leads to the ortho positions carrying less charge than the para position is not completely convincing (see Sec IV), (3) the dipole moment calculated from XVII is about four times as large as the observed moment for toluene (0.4 D experimental, 1.4 D calculated). This suggests that the technique grossly exaggerates the charge migrations, which require to be scaled down by a factor of 3 or 4. But, on the other hand, the procedure is quite systematic, so that we may expect regular variations from one molecule to another to be exhibited by these calculations. Further applications of this analysis will be given in Sec. V. And in Sec. IV we shall show how the alternative molecular-orbital approximation attempts to deal with the difficulties which we have listed above. 6. Aza Replacement
A further effect of great importance is the replacement of a CH group in the aromatic framework by a N atom. In pyridine, for example, in addition to normal structures XVIII (a), (b) we have ionic structures
ELECTRONIC CONFIGURATION AND CARCINOGENESIS '
19
such as (c). Each structure (c) can be thought of as derived from an excited homopolar structure (b) by moving all the charge onto one end (i.e., the N atom) of a long bond.
(4
(b)
(4
(XVIII) Structures in pyridine
(4
(After Pullman)
Mme. Pullman made the hypothesis that the ratio of the weights of structures (c) and (b) was a constant, for all such pairs of corresponding structures, and for all molecules. If we accept this postulate, then it is not hard to estimate the charge distribution, since we need only work with homopolar, or covalent, structures, and these are evidently just the same as in the parent hydrocarbon (benzene). It is, unfortunately, impossible to test the validity of the postulate, because no one has yet succeeded in making any direct calculations of the weights in the different structures. Quite clearly this is a matter of considerable complexity. But it seems probable that here, as in Sec. 111, 5, systematic variations from one molecule, or series of molecules, to another, would be shown up by this analysis. It would, however, be foolish to claim anything like absolute accuracy for the final numerical values of the charges or bond orders. B. Pullman (1948) has endeavored to systematize the method of weighting a little more carefully. But several difficulties still remain. Thus his final charge distribution for the ?r electrons in pyridine is shown in XVIII (d). But the dipole moment associated with these charges is about 2.7 D, although the total molecular moment is known to be only 2.1 D, and the contribution from the u bonds is generally supposed to be in the region of 0.8 D. It looks as if Pullman's charge migrations are calculated t o be about twice as large as they should be (see also Daudel and Martin, 1948). 7. Penney Bond Orders One of the objections to the fractional bond orders described in Sec. 111, 4 arises from the fact that their precise values depend on which choice of structure we have made. We say that they are not true invariants. As a result they do not possess any very really fundamental theoretical significance, though in most cases, and particularly when no five-membered rings are present, the matter is not serious. It was shown by Penney (1937) some years ago that an alternative definition of bond
20
C. A. COULSON
order could be given, which did not suffer from this deficiency. Penney argued that in a molecule like benzene, if the spins of the two electrons associated with a bond were antiparallel (mutual angle of 180°), the bond would be pure double: and if they were quite randomly oriented relative t o each other (mutual angle SO”), it would be a pure single bond. Neither of these situations does in fact occur in large cyclic molecules. A situation arises in which we could think of a superposition of both possibilities. This could be described by saying that the mean value of angle between the spins had a value different from 90” or 180”. The magnitude of this mean value could be used as a measure of bond order. It turns out that this mean value can be calculated and hence the fractional bond order inferred. Such bond orders, of course, differ from those previously described. It is necessary, therefore, when comparing results by different authors, or by the same author at different times, to be sure which definition of bond order he has used. It may be said that the so-called Penney bond orders agree remarkably closely (Coulson et al., 1947) with those calculated according to the methods to be described in Sec. IV. The only unfortunate aspect of the matter is that a full Penney calculation is extremely lengthy. Attempts have been made with some success (Jean, 1948; R. Daudel, 1948c; Vroelant and Daudel, 1949a,c) under the title “method of spin states,’’ to simplify this procedure. And more recently, a much cruder method, taking account merely of the geometrical arrangement of the bonds in the vicinity of any selected bond, has been suggested by Vroelant and Daudel (1949b). All these distinct definitions give distinct bond orders, so that once again caution is necessary when quoting published values for comparison with other values. Further application of these ideas and techniques must be left till Sec. V. IV. MOLECULAR-ORBITAL METHOD 1. Molecdar Orbitals*
We must now describe the alternative method of representing the electrons in a conjugated compound. It is necessary to do so because this method has been used much more widely than the resonance method in recent years. Just as in Sec. 111, we consider merely the distribution of the ?r electrons. Since it is of the essence of these electrons that they are mobile, and resemble the conduction electrons of a metal, there is a certain advantage in recognizing this mobility right from the start. Now in a metal we do not attempt to localize individual electrons either in a bond or on an atom, but we think of each conduction electron as
* For further information, see C. A. Coulson, Valence, Oxford University Press. 1952.
ELECTRONIC CONFIGURATION AND CARCINOGENESIS
21
occupying an orbital which extends over the complete metal. I n the corresponding molecular problem we do precisely the same; that is to say, we suppose that each of the a electrons is assigned to a definite orbital, but these orbitals are polycentric and extend over all the resonating, or conjugated, part of the molecule. In this picture, there is no trace of the structures, or the resonance between them, which were the basis of the valence-bond method. In fact, the new method is fundamentally similar to the description of an isolated atom or of a block of metal; that is, we first calculate the possible orbitals (which we now call molecular orbitals, m.0.) and determine their energies. Next we “feed” the a electrons into these orbitals, two a t a time to allow for the Pauli exclusion principle which says that not more than two electrons may ever occupy the same space orbit, and then they must have opposed spins. We begin with the orbitals of lowest energy and continue until all the available a electrons have been disposed of in this way. The total energy is taken, rather approximately, to be the sum of the energies of the occupied orbitals. Excited levels occur by the promotion of one (or perhaps occasionally two) electrons from one of the occupied m.o.’s to one of the higher energy unoccupied m.o.’s. We can immediately see one important difference between the m.0. method and the resonance method. For now, each a electron may be found anywhere in the molecule. This means that each ?r electron contributes to the total a charge on each of the atoms and also to the bond order in each bond. In general any particular orbital will contribute different amounts to the charges on the various atoms and to the various bond orders. The greater part of the analysis referred to in this section is concerned with describing these orbitals and showing how a knowledge of the orbital enables us to compute the order of any bond and the total charge on any atom. The advantage of the m.0. method is that we can treat the electrons one a t a time, separately. Its chief disadvantage is that by not taking into account the simultaneous positions of any pair of electrons, we allow too many electrons to be in the same place simultaneously. The first statement of the m.0. method was due to Lennard-Jones; the greater part of the earlier theory was due to Bloch (for metals) and to Hund, Hilckel, and Mulliken (for molecules).
2. The LCAO Representation When one of the a electrons is in the close vicinity of a particular atom r, the forces acting on it are largely the same as the forces it would experience if it were attached to that atom. This means that the wave equation for this electron locally resembles the wave equation for this
22
C. A. COULSON
electron in a single atom-so also must the solution of this equation, i.e., the molecular orbital. Thus near atom T , $ resembles &, a known function; near atom 8, $ resembles 48, etc. Now +r and 4, are only of significant size in the neighborhood of their respective nuclei. A function, therefore, which satisfies the above conditions is $=
+
+
~141 ~ 2 4 2
*
. + Cn4n
(3)
where we suppose that there are n atoms in the resonating framework, and the cr are a set of constants yet to be determined. The m.0. method in its usual form asserts that by a proper choice of the c’s we can make (3) a good approximation to any of the real molecular orbitals. A definite technique (secular equations and secular determinant) exists for finding the appropriate c’s. It turns out that there are actually n distinct sets of c’s, and each set is to be associated with one m.0. Thus we have, in principle, a way of approximating to the required m.o.’s. In the process of determining the c’s we also determine the energy of this orbital, and we can therefore arrange the n orbitals in ascending sequence of energy. If there are 2m r electrons, then the lowest m of these m.o.’s willLbefully occupied in the ground state, and the total ?r energy can be found by a simple addition. The approximation used in (3) is the only one that has proved fruitful. It is called the LCAO (linear combination of atomic orbitals) approximation (Mulliken, 1932), since the molecular orbital $ is expressed as a linear combination of atomic orbitals 4,. There is a very simple interpretation of the coefficients Cr in (3). Since the square of $, rather than fi itself, has physical significance,and since $2
=
c12412
+ - + *
*
2ClC24142
+- *
we say that an electron in this m.0. is distributed among the atoms : cn2e If we normalize the orbital so that in the ratio c12:ca2 * cla cZ2 * * Cn2 = 1, we can say that an electron in $ contributes an amount c12 to the total r charge on atom 1, etc. Thus, by determining the sets of c’s, we can soon calculate the charge on each atom. There is a further point to be made here. In the density function 2. we have product terms such as ~ 1 ~ ~ 4 1 4These terms are quite insignificant unless atoms 1and 2 are neighbors, so let us restrict ourselves to such. We may say that the quantity clc2 ia a measure of the probability that the electron in question is associated with both atoms 1 and 2. But in such a case it is reasonable to suppose that it contributes to the bond between these atoms. We therefore define the contribution of an electron in this m.0. to the bond between atoms 1 and 2 as the product
+ +
+
--
ELECTRONIC CONFIGURATION A N D CARCINOGENESIB
23
clcz (Coulson, 1939). The total ?r-bond order is then the sum of products clcz for each of the a electrons. This definition can be shown to fit with the conventional single, double, and triple bonds; it is therefore a reason-
able definition to accept in other cases. It is obviously a matter of great importance to know what factors determine the coefficientscr and hence the other derived quantities (bond order, charge, energy, etc.) There are two sets of parameters on which everything else depends. As these will keep on coming up in later work, we must describe them carefully. They are the Coulomb terms a, of the atoms and the resonance integrals 0,. of the bonds. The Coulomb term arof atom T is a measure of the electronegativity (i.e., electron-attracting power) of this atom, or, more precisely, the electronegativity of this atom toward a electrons when in its proper position in the given molecule. It is defined as the energy of the a electron in the atomic orbital &, associated with nucleus T . It is nearly, but not quite, the same as the energy of the electron in the isolated atom. Although there are serious difficulties in determining exactly what a to use for any selected atom (e.g., nitrogen in pyridine) it is clear that a, which is negative, satisfies the inequality
I%(
< I a N I
10 -10
* The experimental data are recorded in J. A. Miller and Baumann (1946b),J. A. Miller and E.C. Miller (1948), J. A. Miller et 02. (1949), and J. A. Miller st 02. (1961b). t Figure corrected for poor absorption of dye from gastrointestinal tract. E. C. Miller, 1948). In conformity with the equal carcinogenic activities of DAB and MAB the 3-, 2‘-, 3’-, and 4I-methyl derivatives of MAB each had essentially the same activity as the corresponding dimethyl compounds (J. A. and E. C. Miller, 1948; Sugiura, 1948). Further studies showed that the nitro- and chloro-substituted derivatives of DAB also formed activity series of the order 3’ > 2’ > 4‘ (J. A. Miller and E. C. Miller, 1948). When methyl groups were substituted in the 3’ and 5’, 2’ and 5’, 2’ and 4’ (J. A. Miller and E. C. Miller, 1948), 2 and 4’ or 3 and 4’ (Nagao, 1941a) positions or when chloro groups were introduced into the 2’ and 5’ or 2’, 4’, and 6’ positions of DAB (J. A. Miller et al., 1949), the resulting compounds were inactive. The 3’-bromo (Kuhn and Quadbeck, 1949) and 2’,4’,6‘-tribromo (J. A. Miller et al., 1949) derivatives of DAB were also found to be noncarcinogenic; however, the tribromo derivative was only poorly absorbed from the gastrointestinal tract. Similarly, the 3‘-ethosy derivative of DAB had an activity less than one-sixth that of the parent dye (J. A. Miller and E. C. Miller, 1948). Introduction of a fluorine atom in the 2, 2‘, 3‘, or 4‘ positions of DAB resulted in compounds which were at least as active as the unsubstituted dye, and the activity series was very different from that obtained by the
358
JAMES A. MILLER AND ELIZABETH C. MILLER
introduction of a methyl, chloro or nitro group. Thus, the 2-, 3‘-, and 4’-fluoro dyes were about equally active and were approximately twice as carcinogenic as DAB, while the potency of the 2’-fluoro derivative was similar to that of the original compound (J. A. Miller et al., 1949, 1951b). 2’,4’-Difluoro- and 2’,4’,6’-trifluoro-DAB were also twice as active as the unsubstituted dye, and they are the only known carcinogenic polysubstituted derivatives of DAB (J. A. Miller et al., 1951b).2 The high carcinogenic activities of these compounds are particularly significant since they constitute evidence against the benzidine rearrangement theory of DAB carcinogenesis as suggested by Elson and Warren (1944) and Elson and Hoch-Ligeti (1946) (see Sec. V). The introduction of a trifluoromethyl group in the 2’, 3’, or 4’ positions (J. A. Miller et al., 1949) or of a hydroxy group in the 2, 2‘, 3’, or 4’ positions (J. A. Miller and E. C. Miller, 1948; J. A. Miller et al., 1949; Sugiura, 1948) has resulted in inactive compounds. The hydroxy compounds are of special interest since each could theoretically be formed in vivo from DAB. The formation of 4’-hydroxy-DAB following incubation of the dye with liver homogenates (Mueller and Miller, 1948) or liver slices (Kensler and Chu, 1950) has been demonstrated. The 4’-hydroxy derivatives of MAB and AB, both of which appear to be excreted in the urine of rats fed DAB, were also inactive (J. A. Miller and E. C. Miller, 1947, 1948). Several known and possible metabolites of DAB have been tested for carcinogenic activity, and, with the exception of MAB, all have failed to induce tumors with feeding periods of eight to twelve months. In addition to those compounds already discussed these include 4-hydroxyazobenzene, p-phenylene diamine, o-aminophenol, p-aminophenol, hydroquinone (J. A. Miller and E. C. Miller, 1948), and diacetyl-p-phenylene diamine (Sugiura et al., 1945). No tumors were obtained after feeding 2,4’-diamino-5-dimethylaminobiphenyl,the derivative obtained by the benzidine rearrangement of 4-dimethylaminohydrazobenzene in strong acid, for ten months at twice the molar level used for the azo dye (J. A. Miller and E. C. Miller, 1948; J. A. Miller et al., 1949). 3-Dimethylaminocarbazole was also found to be noncarcinogenic (J. A. Miller and E. C. Miller, 1948). This compound was tested since it has been postulated that 2,2’-aaonaphthalene, a hepatic carcinogen for mice, is converted in vivo t o 3,4,5,6-dibenzcarbazole,which is also carcinogenic for mouse liver (Boyland and Brues, 1937; Strong et al., 1938) and which has been a Since this manuscript was submitted 3’,5‘-difluoro-DAB and 2’,5’-difluoro-DAB have also been found to be stronger carcinogens than DAB (J. A. Miller and E. C. Miller, 1952a). Hence it now seems certain that none of the positions on the prime ring of DAB can be directly involved in the carcinogenic process.
THE CARCINOGENIC AMINOAZO D Y E S
359
indicted as the primary carcinogen in this case (Cook et al., 1940). 3-Dimethylaminocarbazole could bear a similar relationship to DAB. Finally, neither cirrhosis nor tumors were obtained when a mixture of the following known and possible metabolites was fed for eleven months: 0.04% of AB and 0.01% (calculated as the free base) each of 4’-hydroxyDAB, 4’-hydroxy-MAB1 4’-hydroxy-AB1 N,N-dimethyl-p-phenylene diamine dihydrochloride, N-monomethyl-p-phenylene diamine dihydrochloride, p-phenylene diamine, aniline hydrochloride, and p-aminophenol (J. A. Miller and E. C. Miller, 1948). 6. Metabolism b y the Rat
If detailed studies on the metabolism of chemical carcinogens are pursued with enough vigor, it is inevitable that at least the initial reactions in the carcinogenic process will be brought to light. Metabolic studies on such agents must be directed not only toward determining the action of the tissue on the carcinogen but also toward discovering the direct biochemical attack of the agent and its metabolites on the target tissue. As pointed out in the Introduction the importance of metabolic pathways in the reactions leading to carcinogenesis can only be assessed a t present by correlating their occurrence and intensity with the time and frequency of appearance of gross tumors in the tissue. As these correlative guideposts become established, the investigator should find it. easier to fill the gaps in our biochemical knowledge of the sequence of events which make up the carcinogenic process. A. Overall Metabolism. The metabolism of DAB in the rat is known in greater detail than that of any other carcinogen. This is largely due to the relative simplicity of the analytical problems involved, the relatively high carcinogenic dose required, and the ease with which large amounts of the susceptible tissue, the liver, can be obtained, The first conclusive study on the metabolism of this dye was that of Stevenson et al. (1942) who found that following the administration of the dye to rats isolable amounts of the N-acetyl derivatives of p-aminophenol and p-phenylene diamine were excreted in the urine. Approximately 50 to 60% of the dye can be accounted for in the urine in the form of acid-hydrolyzable conjugates of these two amines (J. A. Miller and E. C. Miller, 1947). The excretion of large quantities of these metabolites proves that extensive reduction, hydroxylation, and demethylation of the component structures of the dye occur in the rat. It is probable that all possible sequences of these three reactions occur t o some extent, but in general the demethylation reaction apparently precedes the other two (J. A. Miller et al., 1945). That most of the dye is subject t o reductive cleavage of the aao linkage at some stage in its metabolism is evident from the large
360
JAMES A. MILLER AND ELIZABETH C. MILLER
amounts of monophenyl amines excreted in the urine and the low levels of azo derivatives which can be detected in the tissues and excreted during the steady state. At least a part of the dye undergoes a fourth reaction in which it is chemically bound to the liver protein. It is this reaction which, though its exact nature is not known, appears t o be of importance in the carcinogenic reaction induced by the dye. Rats are usually fed approximately 5000 to BOO0 pg. of DAB per day to produce tumors but only 2 to 5 pg. of this dye, 1 to 5 pg. of MAB, and 6 to 10 pg. of AB are found free in the liver (J. A. Miller et al., 1945; Silverstone, 1948). About 25 t o 50 pg. of dye are found in combination with the liver protein. None of the other tissues contains the methylated or the protein-bound dyes. The rat also contains 200 to 300 pg. of AB, principally in the red blood cells, and approximately 25 pg. of this dye and the two methylated dyes are excreted in the urine and feces daily (J. A. Miller and Baumann, 1945a; J. A. Miller et al., 1945; E. C. Miller and J. A. Miller, 1947). In a study of the distribution of 3’-Me-CI4-DAB in the rat Salzberg et al. (1951) found that of the blood fractions tested, the formed elements had the greatest radioactivity; presumably this was due at least in part to the 3’-Me-AB known to be present. However, while they could not detect dye in the feces by colorimetric mcana, approximately 20% of the isotope was excreted by this route. Since it is known that AB is excreted in the bile by rats fed DAB (J. A. Miller et al., 1945), it seems likely that in the case of 3’-Me-C14-DAB bacterial destruction of the 3’-Me-AB in the tract leads to the excretion of nondye fragments containing CI4in the feces. The known and several possible metabolic pathways of DAB and its metabolites in the rat are shown in Fig. 2. B. Reduction of the Azo Linkage. The reductive cleavage of the azo linkage at some stage in the metabolism of DAB was first conclusively demonstrated by Stevenson et al. (1942), who isolated N-acetyl-paminophenol and N,N’-diacetyl-p-phenylene diamine from the urine of rats fed and injected with relatively large quantities of the dye. Earlier Kinosita (1940a) had reported that the carcinogen was reduced in vivo to N,N-dimethyl-p-phenylene diamine and aniline and that these compounds were excreted in the urine, but no experimental data were presented. Stevenson et aE. (1942) were unable to isolate these compounds unless sodium hydrosulfite was added to the urine, and under these conditions the amines may have been formed through the reduction of small amounts of dye in the urine. Using a sensitive analytical method (E. C. Miller et aZ., 194913) it was found that approximately 50% of the ingested DAB was excreted in the urine as acid-hydrolyzable conjugated forms of p-phenylene diamine and p-aminophenol (J. A. Miller and
-
-
Liver protein-Am dye compounds Liver proteins
=Reaction for which evidence exists --+ =Hypothetical reactioii
-t
Metabolicderivatives
I!
-
H
HCHO
demethylatedproducts
H H \
H H
H
- \N/0\N/ - \N/0\N/ 1 - 1
Monophenyl amines
f
H
H
------
1 - 1
H
--------I
FIG.2. The present knowledge of the metabolism of DAB and its derivatives in the rat.
4
CH,
362
JAMES A. MILLER AND ELIZABETH C. MILLER
E. C. Miller, 1947). Small amounts of conjugated forms of N-methylp-phenylene diamine, aniline, and o-aminophenol were also excreted, but only traces of N,N-dimethyl-p-phenylene diamine could be detected. Approximately the same quantities of these monophenyl amines were also excreted in the urine when MAB or AB was fed, except that no methylated diamines could be detected in the urine of rats fed the latter dye. Following the ingestion of any of these diamines the chief excretory product was a conjugate of p-phenylene diamine, but appreciable amounts of conjugates of the N-methyl- or N,N-dimethyl-p-phenylene diamine were also excreted when these compounds were fed. In each case the overall recovery was approximately 70 %. After the administration of aniline large amounts of conjugated p-aminophenol and smaller quantities of conjugated o-aminophenol and aniline were found in the urine. When either p-aminophenol or o-aminophenol was fed, the only metabolite identified was a conjugate of the isomer fed. The preponderance of AB rather than the methylated dyes in vivo and the much smaller amount of N,N-dimethyl-p-phenylene diamine excreted by rats fed DAB than by those fed the diamine itself suggest that in vivo the carcinogen may be largely demethylated to AB prior to reduction of the azo linkage. However, in vitro DAB is reduced by rat liver slices (Kensler, 1947,1948; Kensler and Chu, 1950) or, more rapidly, by rat liver homogenates (Mueller and Miller, 1948, 1949, 1950), and in the latter case the dye destroyed can be stoichiometrically accounted for as N,N-dimethyl-p-phenylene diamine and aniline. Maximum reduction of the dye by liver homogenates was obtained only when the system was anaerobic and fortified with triphosphopyridine nucleotide, diphosphopyridine nucleotide, magnesium ions, and an oxidizable substrate. A requirement for riboflavin-adenine dinucleotide could be demonstrated following carbon dioxide treatment of the homogenate ; riboflavin monophosphate was inactive in this system. A riboflavin coenzyme was also implicated in the liver slice system since the ability of the slices to destroy the dye could be correlated with their riboflavin content (Kensler, 1947, 1948, 1949). As the amines formed by reductive cleavage of the dye have so far exhibited little or no carcinogenic activity (Kinosita, 1940a; Sugiura et al., 1945; J. A. Miller and Baumann, 1945b; F. R. White et al., 1948; Druckrey, 1950a,b), at least a part of the protective action of dietary riboflavin against carcinogenesis by DAB probably results from its participation in the cleavage of the carcinogen to relatively inactive products. With either the slice or homogenate technics kidney was about one-third as active as liver and the other tissues studied, including liver tumors induced by the dye, were essentially inactive. Both slices and homogenates of liver similarly reduced MAB, AB, certain N-sub-
T H E CARCINOGENIC AMINOAZO DYE8
363
stituted derivatives of AB, and the ring-methyl derivatives of DAB.
In general DAB was reduced more rapidly than the other azo dyes studied, and there was no correlation between the carcinogenicity of a dye and its ease of destruction in vitro by liver preparations. Although reduction of the dyes implies that hydraao derivatives are formed as intermediates, there is no evidence that these compounds undergo a ((benzidine” rearrangement in vivo (see Sec. V). C . Hydroxylation of the “Aniline” Ring. The occurrence of large quantities of conjugated p-aminophenol and smaller quantities of conjugated o-aminophenol in the urine following the ingestion of DAB indicates that a high percentage of the ingested dye is hydroxylated at one or more stages in its metabolism. That hydroxylation can follow cleavage was shown by the finding of large quantities of these amines in the urine following the administration of aniline (J. A. Miller and E. C. Miller, 1947). However, the excretion of small quantities of the 2’- and 4’-hydroxy derivatives of MAB and AB in the urine of rats fed either DAB or MAB and the finding of small quantities of 4’-hydroxy derivatives following the incubation of DAB or certain of its ring-methyl derivatives with liver slices (Kensler and Chu, 1950) or homogenates (Mueller and Miller, 1948) demonstrate that some hydroxylation does occur prior to reduction. D. N-Demethylation. The isolation of N,N‘-diacetyl-p-phenylene diamine by Stevenson et al. (1942) also proved that the N-methyl groups were removed at some stage in the metabolism of DAB. This reaction can take place subsequent to reductive cleavage, since conjugated forms of N-methyl-p-phenylene diamine and p-phenylene diamine were found in the urine of rats fed N,N-dimethyl-p-phenylene diamine (J. A. Miller and E. C. Miller, 1947). However, the presence of small quantities of MAB and larger amounts of AB in the tissues (J. A. Miller et al., 1945) and of low levels of hydroxy derivatives of AB and its monomethyl derivative in the urine of rats fed DAB indicates that some demethylation must precede reductive cleavage. In fact, it appears likely that this may be the reaction sequence undergone by a major share of the dye, since the same levels of AB were found in the blood and tissues of rats fed either A% or its N-methyl derivatives (J. A. Miller et al., 1945). This logic requires the assumption that the amount of the primary aminoazo dye in the blood is an index of the amount of this substance being formed from the methylated dyes. This appears to be the case since the amount of AB in the blood is essentially proportional t o the amount of dye ingested (J. A. Miller el al., 1946). The demethylation of DAB to MAB is a reversible process since the same amounts of both dyes are found in the livers of rats fed either compound. The loss of the
JAMES A. MILLER AND ELIZABETH C. MILLER
364
second methyl group appears to be essentially irreversible; neither methylated dye has been detected in the livers of rats fed AB (J. A. Miller el al., 1945; E. C. Miller and Baumann, 1946).* N-Demethylation of the ring-methyl derivatives of DAB also occurs in a similar fashion; in each case primary, secondary, and tertiary aminoazo dyes were found in the liver while only the primary aminoazo dye could be detected in the blood (J. A. Miller and Baumann, 1945b). Dealkylation also occurs when azo dyes bearing other N-substituents are fed (Kensler et al., 1947; J. A. Miller and E. C. Miller, 1948). N-Ethyl groups are removed with relative ease, and the levels of AB in the blood and liver are essentially the same whether DAB, ethylmethyl-AB, or diethyl-AB is fed. On the other hand, 8-hydroxyethyl and benzyl groups are removed with difficulty, and only traces of AB can be detected in the blood of rats fed 8-hydroxyethylmethyl-AB, benzylmethyl-AB, or di-8-hydroxyethyl-AB. The fate in vivo of the N-methyl groups of DAB and certain of its derivatives has been demonstrated by the use of C14-labeled dyes. In the first study of this type Boissonnas et al. (1949) fed 4-dimethyl-C14aminoaaobenzene to an immature female rat and found about 50 % of the isotope in the respired air as C140a. They could not detect any significant quantity of C14 in the body choline or in the urinary creatinine obtained from this rat and concluded that transmethylation of the methyl groups of DAB did not occur under the conditions of their experiment. Somewhat different results were obtained by E. C. Miller et al. (1952b) and MacDonald el al. (1952b) employing similarly labeled dyes in its 3’-methyl adult rats of both sexes. 4-Dimethyl-C14-aminoazobenzene, and 4’-methyl derivatives, and 3-methyl-4-monomethyl-C14-aminoazobenzene were administered by stomach tube in oil solution in single or multiple doses for periods of five hours to twenty-eight days. In agreement with the data of Boissonnas et al. 50 to 70% of the label from each dye was respired as C1*OZwithin forty-eight hours. However, in each case 20 to 40% of the isotope remaining in the body (about 2 to 4% of the administered radioactivity) could be accounted for in the N-methyl groups of the body and liver choline and in the p-carbon of the serine in the body and liver proteins. This distribution of isotopt#is similar to that obtained after the administration of C14-labeled formaldehyde or formate (Sakami, 1948; Siekevitz and Greenberg, 1949; Plaut et al., 1950; Siege1 and LaFaye, 1950). Hence, although some transmethylation in these experiments cannot be excluded, it appears likely that the activity in the choline methyls is due to the formation of these one-carbon compounds from the N-methyl groups of the dye. In the case of the a
See footnote 1 on page 366.
T H E CARCINOQENIC AMINOAZO DYES
365
3'-methyl derivative it was noted that severe deficiencies of folic acid or vitamin Blz reduced but did not abolish the incorporation of the isotope from the dye into the choline and serine. In an earlier study Jacobi and Baumann (1942) reported that the feeding of DAB prevented or lessened the severity of kidney hemorrhages in rats fed a diet deficient in labile methyl groups. Neither AB nor AAT, which lack the N-methyl groups, protected the rats. That these results can be explained by the utilization of the methyl groups from DAB now appears unlikely in view of the relatively low order of incorporation of C14 from the N-methyl groups of the dye into the methyl groups of choline. Further information on the fate of the N-methyl groups of these dyes has been obtained from the metabolism of 3-methyl-MAB in liver homogenates (Mueller and Miller, 1951). This dye was chosen since in vitro demethylation of the secondary aminoazo dyes occurs much more rapidly than with the tertiary aminoazo dyes. Further, the 3-methyl group almost completely hinders the reductive cleavage of the azo group, so that the demethylation can be studied independently of other reactions. It was found that oxygen, triphosphopyridine nucleotide, diphosphopyridine nucleotide, adenosine triphosphate, and magnesium ions were required for optimum demethylation activity. When the reaction was carried out in the presence of semicarbazide, formaldehyde and 3-methyl-AB were recovered in amounts accounting stoichiometrically for the 3-Me-MAB metabolized. Hence it appears very likely that the reaction proceeds through a monomethylol derivative. This result together with the observed fate of these methyl groups in vivo indicates that a methylol derivative is probably an intermediate in the N-demethylation of DAB and its derivatives. Of course, it is entirely possible that in vivo some of the methylol derivative is further metabolized to the N-formyl or even to the N-carboxyl derivative and that these in turn yield formic acid and carbon dioxide, respectively. These interrelationships are shown in Fig. 3. In recent studies mouse liver homogenates have been found to contain two t o three times as much of the enzyme system responsible for the N-demethylation of these dyes as rat liver homogenates. Furthermore, the activity of mouse liver homogenates can be increased about 250% and of rat liver homogenates about 50% by feeding the animals diets containing certain meat products such as tryptone, liver powder, or oxidized cholesterol for at least one week before they are killed. The active factor is not identical with any of the known vitamins or accessory growth factors, and studies still in progress have shown that a number of organic peroxides (such as ascsridole or the photoxide of 9,lO-dimethyl-
366
JAMES A. MILLER AND ELIZABETH C. MILLER
1,2-benzanthracene) produce the same effect when incorporated in the diet at a level of about 0.05% (J. A. Miller et al., 1951a; Brown et al., 1952). E. The Formation of Protein-Bound Dyes. Several years ago we noticed that the liver protein from rats fed DAB turned pink when suspended in acid solution and light yellow when suspended in alkaline or neutral solvents. Like many other aminoazo dyes DAB is a sensitive acid-base indicator, and this observation showed the presence of dye associated with the liver protein. Subsequent studies have shown that the dye is bound through chemical linkages to certain of the liver proteins, and we have suggested that this reaction may play a role in car-
-N"H3
!
I
I
[
] = nd demonstrated
FIQ.3. The observed and possible interrelationships in the oxidative demethylation of the N-methylated aminoaro dyes by the rat.
cinogenesis by the dye (E. C. Miller and J. A. Miller, 1947; E. C. Miller et al., 1949b). Many of the initial observations have since been confirmed by Taki and Miyaji (1950). The protein-bound dyes appear in the liver shortly after feeding of DAB is begun and continue t o accumulate until a maximum level is reached at about one month. Thereafter, even though the dye-feeding is continued, the level of bound dye drops slowly and by four months is only about half-maximal. There are a number of reasons for believing that these dyes are chemically bound to the liver protein. Thus, they cannot be released by prolonged extraction with boiling organic solvents, by extraction with hot trichloroacetic acid, or by dialysis. The dyes are released upon destruction of the protein by t r y p i n or alkali; acid hydrolysis has not been used since aminoazo dyes are not stable in hot mid for prolonged periods. While the liver protein containing the bound dyes dissolves quickly in hot strong alcoholic alkali, the bound dyes are extractable only at a rate parallel with the progressive hydrolysis of the protein. This is also true of tryptic hydrolysis of these preparations. Following alkaline hydrolysis, about 10% of the liberated dye can be shown to be a mixture of MAB and AB. The major share of the dye, however, has
367
THE CARCINOGENIC AMINOAZO DYES
strongly polar properties since it can only be extracted from the alkaline hydrolyaate with a highly polar solvent mixture such as ethyl etherethanol. The bound dyes do not appear to be bound to the nucleic acids present in the liver protein preparations. Following the liberation of all the detectable nucleic acids by heating the protein preparations in 5% trichloroacetic acid for fifteen minutes (Schneider, 1945a), all the dye is still found combined with the residual protein precipitate. Whether or not the phosphorus remaining in such precipitates (as phosphoprotein or residual nucleic acid?) is involved in the binding is not known. In any case, of course, the finding that the dyes are not bound to the major share of the nucleic acids does not exclude them from being bound to the protein moieties of nucleoproteins. The released polar dye from the dye-protein compound has a spectrum in acid which is characteristic of N-disubstituted aminoazo dyes (J. A. Miller et al., 1948). While the nature of the polar group of this dye is obscure there are some clues as t o where and how the binding occurs on the dye molecule. On reduction the polar dye yields aniline and an unidentified polar amine (E. C. Miller et al., 1949b). This indicates that the dye is bound t o the protein either through some substituent on the ring bearing the -N(CHa)Z group or through a derivative of this group. It is felt that the latter possibility is the most likely one. Thus, it has been found that only N-methylated aminoazo dyes become bound to any large extent in vivo. An N-ethyl group permits only a slight amount of binding. This is also the case with the completely demethylated metabolite, 4-aminoa~obenzene.~Two dyes are known which bear N-methyl groups but which are not bound to an appreciable extent. These dyes also bear other groups (N-benayl or N-/3-hydroxyethyl) which have been found to hinder the demethylation of these dyes in the body (J. A. Miller and E. C. Miller, 1948, 1952a). It is the authors’ tentative conclusion that a N-hydroxymethyl derivative of the dye, which is presumably formed during the oxidative removal of the N-methyl groups, is the metabolite which becomes bound to the protein. N-Hydroxymethyl groups are known to be highly reactive and form stable bonds with compounds which possess reactive hydrogens through the elimination of the elements of water (Fraenkel-Conrat and Olcott, 1948). If such Mannich bases were formed with a protein grouping such as in tyrosine or histidine
I I
which has a reactive -CH,
the alkali stable
>
N-CH2-C-
would be formed. On the other hand reactive 4
See footnote on page 356.
I
I
grouping
NH groups in the
368
>
JAMES A, MILLER AND ELIZABETH C. MILLER
protein would yield the alkali labile
N-CHz-N/
\ grouping. The
small amounts of MAB and AB released in the alkaline hydrolysis of the protein may be derived from dyes bound through the latter linkage, i.e., the N-hydroxymethyl derivatives of DAB and MAB, respectively. However, the bound MAB and AB could also be attached through amide linkages. On this hypothesis the polar dyes would have been attached to the protein through the alkali-stable grouping,
>
N-CH2-C--,
I I
and,
following alkaline hydrolysis, would still be combined with an amino acid residue (Fig. 4).
0-y o e v r p i A + ~ < - 1c -+
' - 1 t e i n
KN
tyr-
' N o r dye'
protein
Wmbm wh
altUl In .mly*Is
w; ~ o a c " ~ N r343
"nm-pdor dye"
FIQ.4. A possible mechanism for the formation of protein-bound dye from DAB.
The protein-bound dyes may be conveniently estimated by hydrolyzAfter ing the liver proteins in alcoholic KOH for twenty hours at 80'. extraction with ethanol-ethyl ether the solvent is removed and the residue dissolved in alcoholic HC1. The optical density of the pink solution is obtained and the concentration is expressed as E per unit weight. Conversion factors based on the reduction of the released dyes to give known amines can be used to calculate the molar quantity of protein-bound dye present. Salzberg et al. (1951) found with 3'-MeC"-DAB that the amount of radioactivity in the extract of released bound dye agreed with the levels of dye as determined by the colorimetric methods. However, additional radioactivity was found in the residual alkaline solution after removal of the liberated dye. This component was not detectable by the colorimetric methods used for the dye, and it amounted to almost 50% of the C1*activity found in the extractable fraction. However, at least part of this component may consist of unreleased dye since the foregoing conditions of analysis only liberate 80% of the dye found after ninety-two hours of hydrolysis. After the latter time marked destruction of dye occurs. Hence the present
THE CARCINOGENIC AMINOAZO DYES
369
colorimetric method probably determines somewhat less than 80% ’ of the dye present in the protein. Thus in the case of Salzberg et al. some radioactive dye would be expected to be present in the alkaline phase. It is difficdlt to detect this dye by acidifying the alkaline phase since interfering pigments from the protein hydrolysis are present in addition t o the large volumes and quantities of salts produced upon acidification. The rate of disappearance of the protein-bound dyes in vivo has been determined following the removal of the dye from the diet. The half-life of the combination has been found to be about four days. This is approximately the same order of magnitude as the average half-life of three to seven days observed for the liver proteins of the normal rat (Tarver, 1951). It is of further interest that unpublished observations from this laboratory indicate that the protein-bound dyes are not attacked by the liver enzyme system which rapidly reduces the free dyes. Hence it appears likely that the destruction of protein-bound dye i n vivo consists largely of the breakdown of the protein moiety. The continual formation of protein-bound dye and its subsequent destruction could involve the turnover of a substantial quantity of protein in the liver during the several months required for the formation of tumors. Although quantitatively less important than other reactions involving the dye, the binding of derivatives of DAB to the liver proteins may be of major importance in the carcinogenic process induced by this dye since a number of correlations can be made between the levels of bound dye found under various conditions and the probability of tumor formation (E. C. Miller and J. A. Miller, 1947; E. C. Miller et al., 194913). Since the dye is highly specific for the liver of the rat, it is of interest that the protein-bound dyes have been found only in the liver and, to a small extent, in the blood plasma of rats fed DAB. Likewise the high species specificity of the dye permits another correlation to be made. No protein-bound dye could be detected in the livers of guinea pigs, hamsters, rabbits, cotton rats, chipmunks, or chickens fed the dye. No one has succeeded in producing liver tumors in these species by feeding DAB. Mice are much less susceptible than rats to the carcinogenic action of DAB, and only low levels of protein-bound dye are found in the livers of mice fed the dye. Two dietary conditions are known which lower the level of proteinbound dye in the rat liver and which also reduce the carcinogenicity of the dye. These are the addition of certain polycyclic hydrocarbons (Fig. 5 ) or extra riboflavin to the diet (E. C. Miller and J. A. Miller, 1947; E. C. Miller et aZ., 1952a). Since merely lowering the dye intake of the rat will also produce these results (E. C. Miller et al., 1949b), such effects must be interpreted with caution. In the case of riboflavin
370
JAMES A. MILLER AND ELIZABETH C. MILLER
it is known that this vitamin functions in an enzyme system which reduces the dye to inactive amines (Mueller and Miller, 1950), and it seems probable that the hydrocarbons also increase the rate of metabolism of the dye so that less dye is available to initiate the carcinogenic process. A further correlation is the requirement of a N-methyl group for the formation of high levels of protein-bound dye as well as for high carcinogenic activity toward the rat liver. Thus, as discussed earlier, those compounds which either lack a N-methyl group or which have a N-substituent hindering the metabolism of the N-methyl group are either noncarcinogenic or only weakly carcinogenic and are converted to protein,035,
T u r n at /\054%
'
3'-Me-DAB
10/[4 months
3 ,015
n .
P
l 3'-Me-DAB O + r .0033 MC
O
x
0/15
,005 5
10
15
20
25
30
35
DAYS
FIG.5. The effect of 20-methylcholanthrene (MC) in the diet on the formation of protein-bound dye and liver tumors by 3'-Me-DAB.
bound forms to only a small extent. Similarly the noncarcinogenic 4'-hydroxyl derivatives of DAB, MAB, and AB are not bound to the liver protein. Another correlation between carcinogenic activity and protein-bound dye formation was found on analysis of the livers from rats fed the ring-methyl derivatives of DAB. These isomers vary in carcinogenic activity from 0 to 12, and there is an approximate inverse correlation between the times of dye-feeding required before the maximum levels of bound dye are found in the liver and the carcinogenicities of these derivatives (E. C. Miller et al., 1949b). One of the most intriguing observations is that the protein-bound dyes are not detectable in the tumors induced by DAB and its derivatives even when the dyes are fed continuously (E. C. Miller and J. A. Miller, 1947; Price et al., 1949~). This difference between liver and tumor appears to be due to a difference in the intrinsic properties of these tissues, although its interpretation is complicated by the finding that
THE CARCINOGENIC AMINOAZO DYES
371
liver tumors are supplied only with arterial blood (Breedis and Young, 1949) while the liver receives 60 to 80% of its blood from the portal system. Thus, if the dye is absorbed into the portal system, the amount of dye reaching the liver would probably be much greater than that reaching the tumor. Actually, however, the main route by which ingested dye reaches the liver is not known since the arterial and portal blood in dye-fed rats do not contain detectable amounts of DAB (J. A. Miller et al., 1945) and estimations of this dye in the lymph have not been made. This circulatory problem has been circumvented to a large extent in an experiment (J. A. Miller and E. C. Miller, 1952a) in which rats bearing small liver tumors induced by DAB were fed a dye-free diet for two weeks so that no protein-bound dye remained in the liver. These rats were then given daily subcutaneous injections of an oil solution of DAB for a two-week period. The dye absorbed from the injection site would reach the heart via the vena cava (without passing through the liver) and then pass into the arterial circulation with branches to both the liver and the tumor. Under these conditions appreciable levels of protein-bound dye were found in the liver tissue but again none could be detected in the tumors. Further evidence has come from studies on the electrophoretic properties of the soluble proteins in the livers and liver tumors from rats fed DAB. Each of the morphological fractions of the liver from rats fed dye contain some protein-bound dye, but the major share is found in the soluble proteins (Price et al., 1948, 194913, 1950). On further fractionation of the soluble proteins Sorof et al. (1951) found that 70 to 90% of the bound dye was present in slow-moving proteins which comprised only 7 t o 15% of the total soluble proteins. Of considerable interest is the further finding of Sorof and Cohen (1951) that the liver tumors contained only greatly reduced amounts of this particular protein fraction and that other unrelated tumors exhibited a similar deficiency. The authors feel that the foregoing facts on the protein-bound dyes offer support for the hypothesis that the dyes induce neoplastic changes through the gradual deletion of key proteins essential for the control of growth (see Sec. V). 6 . Alterations in Chemical Composition Following Ingestion of the Dyes
As one approach to determining the changes associated with the induction of liver tumors a number of investigators have compared the composition of normal liver with that of liver from rats treated with a carcinogen and/or the resulting liver tumors. Regardless of the carcinogen employed such studies are complicated by the lack of a uniform histological picture in either the precancerous liver or in many of the
372
JAMES A. MILLER AND ELIZABETH C. MILLER
induced tumors. Some investigators have used cytochemical techniques in which the content of a given constituent in individual cells or small groups of cells can be determined, and these studies are of particular value in determining the distribution of the constituent among various classes of cells. From the standpoint of studying the early changes associated with the carcinogenic reaction, however, they have the disadvantage of depending on the investigator’s judgment as to which cells are undergoing changes associated with potential malignancy and which are undergoing alterations not related t o malignancy. Other investigators have analyzed either the whole tissue or the morphological fractions prepbred from it by differential centrifugation. These studies give better averages of the overall changes than cytochemical analyses, but give no information concerning the uniformity of composition of the various cells of the liver or tumor. Furthermore, significant alterations occurring in only a small percentage of the cells may be masked by large numbers of unaltered cells or by changes of the opposite type in another group of cells. The difficulties in the latter type of study can be partially offset by adequate histological study, but for the best histochemical picture of carcinogenesis both types of investigations appear desirable. Several Japanese workers (Nakahara et al., 1936; Fujiwara et al., 1937; Kishi et al., 1937; Masayama et al., 1938) have made biochemical studies on the gross composition of normal liver, liver undergoing carcinogenesis, and transplantable and primary liver tumors induced by AAT or DAB. Most of these differences and changes are too numerous and unrelated to list here. Unfortunately, only abstracts of these articles were generally available to the present authors. In another approach to the general problem outlined above Price et al. (1948, 1949b,c, 1950) compared the intracellulm composition of normal liver and of liver tumors induced by DAB with the composition of liver from rats fed DAB, its ring methyl derivatives, or 4’-F-DAB for approximately one month. The ingestion of the carcinogenic dyes resulted in large alterations from normal in the contents of nucleic acids, protein, and riboflavin in certain of the fractions, while little change was usually found when the noncarcinogenic dyes were fed, and, in some cases, the extent of change was roughly proportional to the carcinogenicity of the dye fed. These alterations in intracellular composition were even more exaggerated in the hepatic tumors induced by DAB. Figures 6 t o 9 compare the results obtained for the livers from rats fed the basal diet, AB, DAB, or one of the C-methyl dyes with the composition of the tumors induced by DAB. Thus, ingestion of the active carcinogen DAB produced a small increase in the desoxypentosenucleic acid content of the nuclear fraction
THE CARCINOQENIC AMINOAZO DYES
373
(Fig. S), while the 4’-fluoro and 3’-methyl derivatives, which are twice as active, caused 24 % and 100% increases in the amount of desoxypentosenucleic acid. The 3’-methyl derivative also caused a doubling of the amount of nuclear protein. The levels of both desoxypentosenucleic acid and protein in the nuclear fraction of the tumor were even higher than those found in the livers of rats fed 3‘-Me-DAB for just one month. These high levels of desoxypentosenucleicacid in tumors induced by DAB NUCLEAR FRACTION
5.0
9 a
OESOXYPLNTOSENUGLL I0 AG I D
40
4.0
W
2 c
30
3.0
a
$ I-
n t
W
W
a D
20
i
jL.0 1
Y
i
5
10
I .o
Bald
A0
2Ur DAB
Silo- * U o MA0 DA0
~ M o - DAB DAB
*-YO- Tumor DAB
FIG. 6. Comparison of the levels of desoxypentosenucleic acid and protein in the nuclear fractions from the livers of rats fed various azo dyes and in tumors induced by DAB. The number above each bar indicates the relative carcinogenic activity of the dye fed.
had been observed earlier by Masayama and Yokoyama (1940), Davidson and Waymouth (1944), and Schneider (1945b). However, as shown by Price et al. (1950), Cunningham et al. (1950b), and Mark and Ris (1949), the desoxypentosenucleic acid content per nucleus was the same for normal liver, precancerous liver, and induced liver tumors so that the increased level of this nucleic acid in these tissues resulted from greatly increased nuclear concentrations rather than from changes in the amount per cell. I n general, when the noncarcinogenic or weakly carcinogenic dyes were fed, the riboflavin and protein contents of the large granules were as high as those found in the livers of rats fed the basal diet (Fig. 7). Ingestion of the more carcinogenic of the C-methyl series of dyes appreciably reduced the amounts of protein and riboflavin in the large granule fraction, and in the case of 3’-Me-DAB the levels were almost as low as
374
JAMES A. MILLER AND ELIZABETH C. MILLER LARGE GRANULES Rl8OFLAVIN
1.0
0 PROTEIN
T .O
3 0.0 W
2
26.0 W
i
4.0
>
0
f 1.0 9 C.0 I .o
EmOl
A8
2-Yr OM
wk
MAE
DAB
-aDAB
DAB
*-Ma- Tumor DAB
FIG. 7. Comparison of the levels of riboflavin and protein in the large granule fractions from the livers of rats fed various azo dyes and in tumors induced by DAB.
= 0
LARGE GRANULES
AB
*MaDAB
mr MA0
*-Mc DAB
Bo
PENTOSENUCLEIC ACID PROTEIN
DA B
3-Me- Tumor DAB
FIG.8. Comparison of the levels of pentosenucleic acid and protein in the large granule fractions from the livers of rats fed various azo dyes and in tumors induced by DAB.
THE CARCINOGENIC AMINOAZO DYES
375
those found in the liver tumors. However, the strong carcinogen 4’-F-DAB decreased the protein and riboflavin contents of the large granules only slightly, and not in proportion to its carcinogenic activity when compared with the C-methyl dyes. All the aminoazo dyes that were studied caused some reduction in the pentosenucleic acid content of the large granules, but in general the largest reductions occurred when the most carcinogenic dyes were fed (Fig. 8); in the case of the livers from rats fed 3’-Me-DAB the content was very similar to that of tumor SUPERNATANT FLUID 0
1)WOI
AB
PENTOSENUCLEIC ACID PROTEIN
2-Y.DAB
*MIMA8
4-0-
DAB
@MaDAB
DAB
3’%c Tumor DAB
FIG.9. Comparison of the levels of pentosenucleic acid and protein in the supernatant fluid fractions from the livers of rats fed various azo dyes and in tumors induced by DAB.
tissue. From histological study it appeared most likely that the changes in the levels of protein, riboflavin, and pentosenucleic acid in the large granule fraction were largely the result of changes in the numbers of the granules per parenchymal liver cell rather than of gross changes in the qualitative composition of these particles. However, ingestion of the noncarcinogenic dye 2-Me-DAB apparently caused both an increased number of large granules in the parenchymal liver cells and an altered composition as reflected by the pentosenucleic acid to protein ratio. The protein contents of the supernatant fluid fraction from the livers of rats fed the basal or dye-containing diets or from liver tumors were very similar (Fig. 9). Similarly, the pentosenucleic acid content of this fraction was unchanged by the ingestion of any of the dyes except 3’-Me-DAB. The amount of this nucleic acid in the supernatant frac-
376
JAMES A. MILLBR AND ELIZABETH C. MILLER
tion from tumor was twice that found in normal liver, with a resultant increase in the ratio of nucleic acid to protein. A similar increase in the nucleic acid to protein ratio of the cytoplasm of tumor cells induced by DAB as compared to normal cells was found by Stowell (1949) who determined the light absorption of individual cells at 257 and 275 mp. No significant alteration from normal was found in the cytoplasm of nonneoplastic cells from tumor-bearing livers. He also observed that the nucleic acid to protein ratio in the nucleoli was higher in tumor cells than in normal liver cells. Opie (1946) and Opie and Lavin (1946) noted that ingestion of DAB caused degenerative changes in the liver which were accompanied by a chromatolysis of cytoplasmic structures and a loss of pentosenucleic acid. This stage was succeeded by a focal regeneration which was characterized by a reaccumulation of pentosenucleic acid and which appeared to be involved in the neoplastic changes. The composition of the livers from mice which had ingested DAB showed some of the same variations from normal as were observed in the livers from dye-fed rats, while the composition of hamster liver was unaffected by the ingestion of the dye (Price et al., 1951). As noted earlier mice develop tumors only slowly, while hamsters are resistant to the carcinogenic action of DAB. I n another study Price and Laird (1950) compared the intracellular composition of normal liver and DAB-induced liver tumors with regenerating liver one to twenty-three days after partial hepatectomy. Nuclear counts were made and all the results were calculated in terms of the amount per cell. Each type of liver cell had its own characteristic composition, and the intracellular compositions of tumor and regenerating liver were not similar enough to suggest a common pattern for growth. In a study of the sequence of changes following the feeding of 3'-MeDAB Griffin et al. (1948) and Cunningham et al. (1950a,b) found the hepatic level of desoxypentosenucleic acid to be 5001, above normal after only two weeks of dye-feeding and to be twice the normal level by eight weeks. On a fresh weight basis liver tumors induced by this dye contained three times as much desoxypentosenucleic acid as normal liver. Furthermore, radioactive phosphorus was incorporated to a much greater extent into the desoxypentose nucleic acid in the livers of dye-fed rats or in the induced tumors than into the desoxypentosenucleic acid of normal rat liver (Griffin et al., 1951). There was a significant increase in the globulin content of the liver following ingestion of the dye, and as seen earlier by other investigators (Kensler et at., 1941; E. C. Miller et al., 1948; Price et al., 1949b, 1950) there was a progressive decrease in total hepatic riboflavin and of the pentosenucleic acid content of the large
THE CARCINOGENIC AMINOAZO DYES
377
granule fraction following dye-feeding. This group of investigators (Cook et al., 1949) also found a threefold increase in the serum yglobulin level and a 15 to 20% decrease in the serum albumin level when rats were fed 3I-Me-DAB for two to eight weeks. Similar observations on the globulin level in rats fed DAB were made by Hoch-Ligeti et al. (1949) ; however, these workers observed no change in the albumin level and noted that the double peaks of albumin in normal rat serum changed to a single peak during gross tumor development. Homogenates of the livers from rats fed carcinogenic azo dyes also show a resistance to heat coagulation (Griffin and Baumann, 1948a). The livers from rats fed DAB or 3’-Me-DAB and the tumors induced by these dyes have essentially the same overall amino acid composition as normal liver (Schweigert et al., 1949; Sauberlich and Baumann, 1951). However, the proteins from each of the morphological fractions of the liver tumors contained 25 to 41% less methionine and 15 to 44% more cystine than the proteins from the same fractions of normal liver. In general tumors have been found to contain only low levels of most vitamins, and in the case of the tumors induced by DAB the levels of riboflavin, vitamin Be, biotin, pantothenic acid, and nicotinic acid are about one-fourth to one-tenth of those in normal liver (Kensler et al., 1941; Pollack et al., 1942a,b; Taylor et al., 1942a,b; West and Woglom, 1942; E. C. Miller et al., 1948; Price et al., 1949a,c; Higgins et al., 1950). Ingestion of DAB for periods short of tumor formation also results in marked decreases in the hepatic levels of riboflavin, vitamin Be, and biotin (Kensler et al., 1940; E. C. Miller et al., 1948). The level of riboflavin in the liver also decreases when other dyes are fed, and the extent of the loss is roughly proportional to the carcinogenicity of the dye fed (E. C. Miller et al., 1948; Griffin and Baumann, 1946, 194813; Griffin et al., 1948). Some studies have also been made on the enzymatic composition of liver which is undergoing carcinogenesis. Thus Roskelly et al. (1943) and Hoch-Ligeti (1947) showed that slices from the livers of rats fed DAB underwent a progressive loss of succinoxidase activity as the feeding time increased and that liver tumors induced by the dye contained even less of the enzyme system. Similar results were obtained by Potter et al. (1950) and by Viollier (1950b) with tissue homogenates. Potter et al. (1950) also analyzed the livers of rats fed other azo dyes and found that the succinoxidase activity depended on the dye fed and could be correlated with the amount of large granule protein in the liver. On the other hand the amount of oxalacetic acid oxidase, which is low in liver tumors, was not appreciably lower in preneoplastic than in control livers.
378
JAMES A. MILLER AND ELIZABETH C. MILLER
Gallico and Boretti (1948) found decreased levels of both succinic and malic dehydrogenases in the livers of rats fed AAT. The hepatic levels of glyoxalase (Cohen, 1945), lipase (Mark, 1950), xanthine oxidase (Westerfeld et al., 1950), arginase and histidase (Viollier, 1950a), choline oxidase (Viollier, 1950b), tributyrinase (Viollier and Waser, 1950), and the enzymes involved in the synthesis of citrulline and p-aminohippuric acid (Tung and Cohen, 1950) also decrease following the ingestion of DAB or 3’-Me-DAB. The levels of most of these enzyme systems are even lower in the induced liver tumors. On the other hand the acetylcholinesterase activity of liver and plasma (Viollier and Waser, 1950; Langemann and Kensler, 1951) and the alkaline phosphatase activity of liver (Woodard, 1943; Mellors and Sugiura, 1948; Pearson et al., 1950) increase following the ingestion of DAB or 3’-Me-DAB, and the levels of these enzymes are even higher in the tumors induced by these dyes. Zamecnik and his associates (1948, 1951) suggested that tumors induced by DAB had a faster rate of protein synthesis than normal rat liver, since the protein from tumor slices incubated with C14-labeled alanine contained approximately six times as much radioactivity as slices from normal liver incubated under the same conditions. The opposite conclusion might be inferred from studies on the in vivo incorporation of C14-labeled alanine or glycine into liver and tumor proteins (Zamecnik and Frantz, 1949; Griffin et al., 1950). Under these conditions the uptake of radioactivity into the tumor protein is slower than in the case of liver protein, but the maximum specific activities attained in each tissue are approximately equal. However, the latter studies were probably complicated by the differences in the blood supply of liver and liver tumors (Breedis and Young, 1949; Zamecnik et al., 1951), while the in vitro studies were complicated by differences in the amounts of endogenous substrates present in the slices from normal liver and liver tumor. In general the oxidative enzyme systems appear to be relatively deficient in the induced liver tumors as compared to normal liver (Potter, 1944, 1951), although the activity of lactic acid dehydrogenase is as high as in normal liver (Meister, 1950). On the other hand, the levels of certain phosphatases, nucleic acid depolymerases, cathepsins, desamidases, and nucleic acid desaminases are usually about as high in the liver tumors as in normal liver (Greenstein, 1947). For detailed discussions of theenzymatic composition and metabolism of the liver tumors induced by the aminoazo dyes the reader is referred to several recent reviews of this subject (Greenstein, 1947; Potter, 1944, 1951 ; Rusch and LePage, 1948; Zamecnik and Frantz, 1949; Zamecnik et al., 1951; Olson, 1951).
THE CARCINOGENIC AMINOAZO DYES
IV. STUDIESON
THE
370
HEPATO-CARCINOGENICITY OF OTHER Azo DYES
I. %',3-Dimethyl-.4-aminoazobenzene (o-Aminoazototuene, A A T ) and
Related Compounds Although the early observations on the carcinogenicity of AAT for rat liver (Sasaki and Yoshida, 1935; Shear, 1937) have been confirmed by other investigators (see Hartwell, 1941; Crabtree, 1949; J. A. Miller et al., 1949), its low carcinogenicity in this species as compared with DAB has discouraged extensive use. Considerably more study has been made on the carcinogenicity of AAT for mouse liver, since in this species it is more active and less toxic than DAB (see Shear, 1937; Hartwell, 1941; Law, 1941; Andervont and Edwards, 1943a; Andervont et al., 1944; Kirby, 1945a,b; Crabtree, 1949)., Early dietary studies showed that the activity of AAT was considerabIy diminished when either liver (Mori, 1941a), wheat or cod liver oil (Ando, 1941a,b) was included in the diet of rats, but the addition of cholesterol to the diet of mice was without effect (Baumann et al., 1940). Andervont and his associates (1942, 1943a, 1944) found that subcutaneous injection of oil solutions or glycerol suspensionsof AAT induced hepatomas in the mice of several strains and that the female mice of each strain were more susceptible than the male mice. A single subcutaneous injection of 2 mg. induced hepatomas in three of twenty-nine virgin female mice of the A strain and a single injection of 4 mg. induced liver tumors in seven of thirty-one mice. A single oral dose of 2 mg. was inactive in this experiment, but a 4-mg. oral dose elicited hepatomas in six of sixteen female mice (Andervont and Edwards, 194313; Andervont, 1947). In studying the effects of hormone stimulation Andervont and Dunn (1947) showed that intact females of the C strain were more susceptible to the induction of hepatomas than either castrated females or intact males. The susceptibility of male mice was considerably increased by castration, but castrate male or female mice bearing testosterone propionate pellets were almost as resistant as intact male mice. Andervont et al. (1942, 1943a,b, 1944, 1947) and Kirby (1945a) also found that the subcutaneous injection of AAT induced lung tumors, hemangioendotheliomas, and fibrosarcomas in many strains of mice. A number of compounds related t o AAT have been tested for carcinogenic activity in either rats or mice, but no extensive study has been made of the structural features required for activity. Crabtree (1949) prepared six isomeric aminoazotoluenes and fed them to both rats and mice. 2',3-Dimethyl-4-aminoazobenzene was the only isomer which
380
JAMES A. MILLER AND ELIZABETH C. MILLER
induced gross hepatic tumors in rats although microscopic areas which appeared to be hepatomas were found in the livers of rats fed 2’,5-dimethyl-2-aminoazobenzene. These two compounds and 2,4’-dimethyl4-aminoazobenzene induced gross hepatic tumors in mice while 4’,5-dimethyl-2-aminoazobenzene,2,3‘-dimethyl-4-aminoazobenzene, and 3,4’-dimethyl-4-aminoazobenzene were essentially inactive. Kirby (1945a) also failed to induce tumors in the livers of mice by subcutaneous injection of 4’,5-dimethyl-2-aminoazobenzene,while Yoshida (cited by Shear, 1937) obtained no tumors when the compound was fed t o rats for fifteen months. N-Methylation of AAT did not alter its activity toward either rat or mouse liver (J. A. Miller et al., 1948; J. A. Miller and E. C. Miller, 1952a), but acetylation reduced its potency (Kinosita, 1940a). Law (1941) found liver tumors in two of fifty-four and two of twenty mice which received subcutaneous injections of 4‘-hydroxy-2,3’azotoluene or 2,3’-azotoluene, respectively; Otsuka and Nagao (cited by Cook et al., 1937) had previously reported that 2,3’-azotoluene was inactive for rat liver. Nagao (1940) observed liver damage but no tumors following the ingestion of 2‘,3-dimethyl-4-aminoazoxybenzeneby rats, and Kirby (1947a, 194830) and Otsuka (see Cook et al., 1937) were unable to induce hepatomas in mice by oral, topical, or subcutaneous application of diazoaminobenzene. 2. Azonaphthalene Series Because of the similarity in the carcinogenic action of 2,2’-azonaphthalene and some of the arninoazo dyes, the studies on this dye are included here even though it lacks the amino substituent characteristic of the other hepato-carcinogenic azo dyes. Thus, high incidences of cholangiomas and hepatomas were reported by Cook and his associates (1940) in mice given oral, topical, or subcutaneous applications, although it should be noted that these terms were used to designate new growth of bile duct and liver cells and not necessarily to denote true tumors. The isomeric 1,l’-azonaphthalene produced liver damage in only a few mice, whereas 1,2’-azonaphthalene and 4-amino-l,2’-azonaphthalene were inactive. In strong acid solution 2,2’-azonaphthalene after partial reduction undergoes a benzidine rearrangement to 2,2’-diamino-l, 1‘dinaphthyl, and this compound readily loses ammonia to give 3,4,5,6dibenzcarbazole. Boyland and Brues (1937) and Strong et al. (1938) had reported that the latter compound was carcinogenic for mouse liver, while Andervont and Edwards (1941) have mentioned only hepatic lesions. Although this series of reactions has not been demonstrated in vivo, Cook et al. suggested that 2,2‘-azonaphthalene might be carcinogenic through in vivo conversion to the carbazole. This postulation was
THE CARCINOGENIC AMINOAZO DYES
381
strengthened by the reported inactivity of 1,2,7,8- and lJ215J6-dibenzcarbasole (Boyland and Brues, 1937) for the liver; these compounds would be formed if the inactive 1,l’-or lJ2’-aaonaphthalenes underwent a similar series of reactions in vivo, The carcinogenic activities of the and l12’-diamino-2,1’-dinaphcorresponding l,l’-diamino-2,2’-dinaphthyl thy1 were not determined. Neither the azonaphthalenes or 2,2’-diamino1,1’-dinaphthyl (Cook et al., 1940) nor 3,4,5,6-dibenzcarbazole (Boyland and Brues, 1937) induced hepatic tumors in rats. N-Methylation or N-ethylation of 3,4J5,6-dibenzcarbazole nearly abolished the hepatocarcinogenic activity of the compound, although these derivatives were both active toward the skin and subcutaneous tissue (Kirby and Peacock, 1946; Kirby, 1948a). 3. Trypan Blue Gillman and his associates (1949) have observed a marked hepatic response following the subcutaneous injection of trypan blue (sodium acid) into salt of o-tolidindisa~o-bis-l-amino-8-naphthol-3~6-disulfo~c rats a t weekly or biweekly intervals. In most of the animals the reaction was characterized by an extensive reticulosis which sometimes terminated in reticulum cell sarcoma of the liver. In an occasional rat the reticulosis was less extensive, and the connective tissue elements underwent malignant transformation to spindle-cell sarcoma. These observations are of particular importance since the induction of reticulum cell sarcomas by other means has not been reported. Cohen and Cohen (1951) have reported in a short note on the behavior of rats in which subcutaneous transplants of hepatomas induced by DAB grew and then regressed. When these rats were treated with total body irradiation by x-rays or with subcutaneous injections of trypan blue, all the tumors started growing again and grew progressively for the next four months until death. These rats, in contrast to the controls, had a high tendency toward bronchiectatic lesions, cysticercosis of the liver, and bartonella infection. Since cysticercosis is known to induce sarcoma of the liver (Bullock and Curtis, 1924) these investigators feel that the trypan blue may act by deranging some resistance mechanism and thus activating and disseminating latent spontaneous tumors as well as rendering the host susceptible to chronic infections such as bartonellosis and cysticercosis.
4. Commercial Food Dyes The synthetic food dyes are examples of the increasing number and variety of foreign compounds that humans are exposed to in modern society. Since the known carcinogens exhibit great differences in
382
JAMES A. MILLER AND ELIZABETH C. MILLER
activity in different species, tests on technologically useful compounds for carcinogenicity in experimental animals can never adequately ascertain their activity in man. Man has been shown to be susceptible to the carcinogenic action of several chemicals and occasional or frequent contact with a variety of foreign compounds in the air, food, drugs, and contact articles may summate with intrinsic factors to increase the incidence of cancer. The use of natural organic coloring matters is subject to the same objection except in the case of compounds of known physiological value such as the carotenes. Since the demand for the use of synthetic and natural coloring and flavoring materials of no food value is unlikely t o diminish it appears to the authors that these materials should a t least be subjected to extensive tests for carcinogenicity in a variety of experimental animals. Only a few such tests have been reported. The U. S. Government through the Food and Drug Administration permits the use of nineteen different coal-tar dyes in foods, drugs, and cosmetics after individual batches of these dyes have passed toxicity tests (Jablonski, 1951). Ten of these dyes are azo dyes, and these include four oil-soluble dyes. Two of the latter dyes are o-aminoazo derivatives and are used for coloring fats and oils; they are F D and C Yellow No. 3 (Yellow AB, phenylazo-8-naphthylamine)and FD and C Yellow No. 4 (Yellow OB, o-tolyaso-0-naphthylamine). So far these dyes have been found to be noncarcinogenic in rats (Sugiura, 1942, 1946) and in mice (Badger et al., 1942), although they have some toxic effects in these species. Cook et al. (1940) also failed to find liver tumors in mice after the administration of a number of water-soluble azo dyes used as food colorings in Great Britain and in this country. However, in Great Britain, Kirby and Peacock (1949) found hepatomas in six of eighteen male and one of eighteen female mice injected subcutaneously with up to 150 mg. of benzeneazo-p-naphthol (Sudan I or “Oil Orange E”). One of eighteen female and none of sixteen male mice developed hepatomas after similar injections of “Oil Yellow HA” (formula not published). It is of interest that the other two oil-soluble F D and C dyes are close relatives of benzeneazo-b-naphthol. They are F D and C Red No. 32 (Oil Red XO, m-xylylazo-p-naphthol) and F D and C Orange No. 2 (Orange SS, o-tolylazo-/%naphthol). Samples of another dye, watersoluble and not an azo dye, F D and C Green No. 2 (Light Green S F Yellowish, disodium salt of dibenzyldiethyldiaminotriphenylcarbinol trisulfonic acid anhydride) have produced sarcomas after subcutaneous injection in rats (Schiller, 1937; Harris, 1947a). The use of the trivial name “butter yellow” for DAB should be discontinued for it is no longer an accurate description of the use of this
THE CARCINOQENIC AMINOAZO DYES
383
compound and since several of the oil-soluble dyes now in use in foods are known by this term in commercial circles.
V. ON
THE
MECHANISM OF Azo DYE CARCINOGENESIS 1. General Considerations
The fate of any foreign compound in the body is obviously dependent on how it is attacked by the enzymes of the cells. The fate of the host, on the other hand, will depend on the extent to which the presence of the foreign agent causes quantitative or qualitative alterations in the constituents and potentialities of the cells. Both the metabolism of the compound and the alterations in the host cells may occur more or less simultaneously. Depending on the nature of the cellular changes, such an altered cell might recover and go on unchanged, or the cell and its descendants might even be able to meet future attacks more easily. Or if certain irreversible changes occur, the cell might die. However, still other irreversible changes might occur which, though permitting the cell to live, would place it in a new relation to the rest of the organism. If the cells merely ceased t o perform special functions and did not multiply a t an abnormally high rate, they might escape notice unless they were so numerous that the whole organism suffered from a lack of functional tissue. If, however, the affected cells suddenly, or gradually in future cell generations, escaped from the growth controls of the organism, they might reproduce themselves fast enough to form gross tumors possessing various degrees of autonomy. It is likely that each of the above responses is t o be found in any tissue susceptible to a carcinogen, and the fate of any one cell may be determined by the quantity of the primary carcinogen (i.e., the derivative directly initiating the process) which reaches it. Thus, as discussed in the section on Histology, following the administration of the aeo dyes some cells appear histologically undamaged, others die, and still others have altered characteristics; and it is in these areas of altered cells that the tumors appear t o arise. Likewise the magnitude of the liver changes and the survival of the animals are functions of the dose of carcinogen administered. The nature of the fundamental changes which make cancer cells relatively free of the factors which control the growth of normal tissues has not been defined, but they appear to be both irreversible and heritable. From the present knowledge of the composition of cells and of the maintenance of cell types through successive generations, it appears almost certain that this type of change must involve either c e w proteins or nucleic acids or both. With most carcinogenic procems there is little basis a t present for deciding in which of these constituents
384
JAMES A. MILLER AND ELIZABETH C. MILLER
the fundamental changes occur or in which part of the cell the crucial changes take place, but in the case of the a50 dyes the evidence points toward the importance of alterations in certain of the liver proteins. In studying the mechanism of chemical carcinogenesis there are a t least two major problems to be answered. The first is the identification of the compound or set of compounds which are directly involved in the initiation of the process. Thus, when a compound is administered, it may be metabolized to some derivative which is directly involved in the carcinogenic process and which could be properly considered as the primary carcinogen. Similarly, the identities of the initial sites of action of chemical carcinogens must be discovered before the whole process can be understood. Even a partial realization of these objectives should enable a more logical attack to be made on the mechanism of action of the carcinogen. These problems are analogous to those met in studies on vitamins, for the vitamins in the food must often be converted to coenzymes and the coenzymes must combine with specific proteins before the physiological activity of the vitamin is realized. The general property of chemical carcinogens that they do not need to be present in the tissue throughout the whole period of carcinogenesis should aid in the solution of these problems. 2. SpeciJic Hypotheses
A number of hypotheses have been proposed on the nature of the initial reactions involved in carcinogenesis with the azo dyes. These suggestions are outlined below. A. The Benzidine Rearrangement Hypothesis. Cook et al. (1940) have proposed that 2,2'-azonaphthalene is active in the mouse through conversion to 3,4,5,6-dibenzcarbazole. This conversion in vivo would, if analogous with the known in vitro reactions, involve a partial reduction to the hydrazo compound followed by a beneidine rearrangement to the corresponding diaminodinaphthyl and finally in a deaminative cyclization to the dibenzcarbazole (Fig. 10). However, the evidence that this actually occurs in vivo is incomplete (see above). A similar suggestion in the case of DAB has been made by Elson and his associates (1944, 1946), who have made use of the observations of Cook et al. on the azonaphthalenes as well as the work of Kensler et al. (1942a) on the toxicity of various diamines toward certain enzymes. Thus Elson and Warren (1944) found that the ingestion of azobenzene (a noncarcinogenic compound; Hartwell, 1941; Spitz et al., 1950) by rats led to the excretion in the urine of aniline and a substance which upon acid treatment was converted into 4,4'-diaminobiphenyl or benzidine. The authors assumed that the benzidine precursor was a derivative (ethereal sulfate?) of
T H E CARCINOGENIC AMINOAZO DYES
385
hydrasobenzene, and i t was suggested that a rearrangement of this precursor to benzidine might also take place in vivo. Elson and HochLigeti (1946) also considered it probable that DAB undergoes a similar partial reduction and rearrangement in vivo to yield 2,4’-diamino-5dimethylaminobiphenyl (Fig. 11) but they offered no direct supporting evidence. This compound, a substituted p-phenylene diamine, was found to be a potent inhibitor for certain enzymes and hence might be involved in the carcinogenic process induced by the parent dye. Direct and protracted tests of this rearrangement product for carcinogenic activity in the rat have yielded negative results (J. A. Miller and E. C. Miller, 1948; J. A. Miller et al., 1949). Positive evidence against the
FIQ. 10. The possible rearrangement products of 2,2’-azonaphthalene in the mouse (see Cook el al., 1940).
participation of a benzidine rearrangement in the carcinogenic process induced by DAB has been obtained through tests on the carcinogenicity of various mono- and polyfluoro derivatives of this dye (J. A. Miller et al., 1949, 1951b). These derivatives were chosen since the fluoro group is small and since the C-F bond is one of the strongest known. Thus 4’-F-DAB and 2-F-DAB proved t o be considerably more active than DAB and the 2‘-fluoro derivative had essentially the same activity as the parent dye. The 4’-flUOr0 substitution should have greatly reduced the carcinogenic activity and the 2’-fluoro and 2-fluoro compounds should have been less active if benaidine or semidine rearrangements occurred in the metabolism of the dye and were of decisive importance in the initiation of the carcinogenic process. More decisive positive evidence has been obtained in tests with 2’,4’-difluoro-DAB and 2’,4‘,6’-trifluoro-DAB. Each of these dyes has also proved to be more active than the parent dye. Neither of the two possible benzidine rearrangements and only one of the three possible semidine rearrange-
386
JAMES A. MILLER AND ELIZABETH C. MILLER
ments could occur in the case of the trifluoro derivative (Fig. 11). The remaining semidine rearrangement would presumably be inhibited in the case of 2-F-DAB where one of the two ortho positions is unavailable for reaction. Yet this dye is twice as carcinogenic as DAB. Thus it is felt that conclusive evidence now exists against the participation of bensidine or semidine rearrangements in the carcinogenic process induced by 4-dimethylaminoasobenzene. It is of interest that the 3'-flUOrO derivative of this dye is also considerably more active than the unsubQN=O(CH&
12H
Q-&yWk J \
Benzidine Rearronpemento
Semidine Rearrangements
0 ' positions blocked in 2',~,6'-trifluoro-DAB
FIQ. 11. The possible rearrangement products of Pdimethylaminohydrazobensene.
stituted dye. Indeed, the data on the high activities of the four monofluoro and the two polyfluoro derivatives can be used to argue that of the four positions involved, i.e., 4',3',2', and 2, probably none are directly concerned in the carcinogenic process.K While the 3 position remains to be tested in this regard, the above data and the essentiality of the azo group and of at least one N-methyl group for carcinogenicity indicate that the initial carcinogenic reaction probably involves one or both of these groups and not any of the ring positions. B. The Split Product Hypothesis. A different approach was taken by Kensler and his associates (1942a) in an attempt to identify the primary carcinogen and the initial sites of action in carcinogenesis by DAB. These workers advanced the suggestion that it is not the parent dye itself which initiatea the carcinogenic process but its reduction or split product, N,N-dimethyl-p-phenylene diamine, which was found to inhibit 6
See footnote on page 358.
THE CARCINOGENIC AMINOAZO DYES
387
certain important glycolytic and oxidative enzymes (Kensler et al., 1942a,b; Potter, 1942; Potter and DuBois, 1943). In support of this concept Kensler et al. correlated the toxicities of various p-diamines toward a yeast zymake system with the reported carcinogenicities of the aminoazo dyes from which they might arise in vivo. However, their correlation has been shown to rest on just two of the dyes whose carcinogenicities were quoted and extensive further tests of this hypothesis have not provided it with any support (J. A. Miller and Baumann, 1945b; Sugiura et al., 1945; J. A. Miller and E. C. Miller, 1948; J. A. Miller et al., 1949). In particular, N,N-dimethyl-p-phenylene diamine has not been found to be carcinogenic in the rat even when fed at high levels for long times (Kinosita, 1937; Sugiura et al., 1945; J. A. Miller and Baumann, 1945b; F. R. White et al., 1948; J. A. Miller and E. C. Miller, 1948). Furthermore, while it seems very likely from in vitro studies with liver homogenates (Mueller and Miller, 1949, 1950) that this diamine is formed, at least to some extent, from the dye in vivo it was also found that riboflavin-adenine-dinucleotideis the prosthetic group of the enzyme which reductively cleaves the azo dye. Since high levels of dietary riboflavin greatly delay the carcinogenic action of DAB (see Sec. III,3) it seems unlikely that the diamine could be the primary carcinogen when it should be present in vivo in the greatest amount under these conditions. Potter (1942) and Kuhn and Beinert (1943) presented evidence a t variance with the conclusions of Kensler et al. (1942a,b) that the diamines act as enzyme inhibitors through oxidation to SH-reactive free radicals. Kuhn and Beinert also felt that the carcinogenicity of the aminoazo dyes might depend on the oxidation of the dyes to various p-quinones. Although the present authors have published much of the data that do not support the hypothesis of Kensler et al., we feel that their basic concept still has definite heuristic value. Thus it is possible that reduction of the protein-bound derivatives of DAB occurs in vivo to form protein-bound diamines which could then function as enzyme inhibitors. Likewise a partial reduction-oxidation system such as azo+hydrazo may participate in the carcinogenic process. The hydrazo form of the dye is merely a substituted form of N,N-dimethyl-p-phenylene diamine and might act as an enzyme inhibitor in a reaction which would oxidize the hydrazo form back to the parent azo dye. Similarly, the reactivities of the N-methyl groups, which are required for this carcinogenic process, may differ greatly in these two forms. The report by Kinosita (1940b) that 4-dimethylaminohydrazobenzene is not carcinogenic cannot be taken as conclusive since no data were given on the preparation of this hitherto unknown (and, according to Kinosita, unstable) compound. C . The Methyl Deficiency Hypothesis. Although it is not known
388
JAMES A. MILLER AND ELIZABETH C. MILLER
whether there is any relationship between the induction of liver tumors in the rat by DAB and the choline deficiency-induced hepatomas described by the Alabama group (Copeland and Salmon, 1946; Engel et al., 1947), the properties of DAB and its metabolites make it possible t o consider a situation such as that depicted in Fig. 12. The removal of the methyl groups in an oxidized form and the presumed requirement for preformed methyl groups for remethylation could constitute a “leak” in the supply of available labile methyl groups in the liver. The inability of choline to protect against tumor formation and the failure of nicotinamide or glycocyamine to accelerate tumor formation by the dye do not suggest that this set of circumstances plays a significant role in the carcinogenic action of the dye (see Sec. 111,3), On the other hand, it is known that
’
- -DrI>diet
functional uses
[CH3-l
labile meth I POOf
MAB
I
C Hr;d: : I I I
I
end product
limited resynthesis
FIG.12. A possible mechanism for the production of a partial methyl deficiency in the rat liver by DAB and its metabolites.
DAB lowers the level of choline oxidase in the rat liver (Viollier, 1950b). This enzyme is necessary for transmethylation from choline, and as suggested by Kensler and Langemann (1951) a reduction in its level would be equivalent to a partial choline deficiency in the rat. Attempts to influence the activity of the dye by the products of this enzyme, such as betaine, would be of interest here. D. The Protein (or Enzyme) Deletion Hypothesis. The most recent suggestion concerning the mode of action of the aminoazo dyes has come from the authors’ laboratory (E. C. Miller and J. A. Miller, 1947; E. C. Miller et al., 1949a). This arose from the observation that the formation of liver tumors in rats fed DAB or any of its carcinogenic derivatives is preceded by the accumulation of protein-bound derivatives of the dye in the liver. As outlined in the section on Metabolism a number of correlations have been made between the presence or level of the bound dye and the probability of tumor formation under different conditions, and these correlations have indicated that the protein-bound dyes may bear a causal relationship to tumor induction by the dyes. Thus (1) under most conditions tumors are found only in the liver of the rat, and this is the only tissue in which the.protein-bound dyes can be detected. (2)
THE CARCINOGENIC AMINOAZO DYES
389
Of the eight species tested only two, the rat and the mouse, have proved susceptible to carcinogenesis by DAB. Similarly, these are the only two species in which the protein-bound dyes can be detected. More bound dye is found in the liver of the more susceptible species, the rat, than in the liver of the mouse. (3) MAB, the only known carcinogenic metabolite of DAB, gives rise to the same level of bound dye as DAB, while the essentially noncarcinogenic metabolite AB yields only a low level of bound dye. Little or no bound dye has been detected fdlowing the administration of the noncarcinogenic hydroxylated aminoazo dyes. (4) When the extents of protein-binding of several of the ring-monomethyl derivatives of DAB are compared, it is found that the rapidity with which maximum protein-binding occurs can be directly correlated with the carcinogenicity of the dye concerned. (5) The levels of bound dye in the liver are lower when the dyes are fed in diets (such as those containing high levels of riboflavin or low levels of certain polycyclic hydrocarbons) which inhibit tumor induction by the dye. (6) In differential centrifugation experiments some bound dye has been found in each morphological fraction from the liver cells. However, over 50% of the bound dye in the liver is associated with the soluble proteins, and of this at least SO% is attached to a slow-moving fraction which accounts for only 15% of the soluble proteins. The bound dye is not found in the tumors produced by continuous feeding of the dye, and in the soluble proteins, at least, the amount of the electrophoretic component bound to the dye in the liver is greatly reduced in the tumor. Furthermore, evidence for similar carcinogen-protein complexes has been found in the case of two of the polycyclic hydrocarbons (E. C. Miller, 1951; Heidelberger and Weiss, 1951, Wiest and Heidelberger, 1952). While these studies are less complete than those on the azo dyes, they indicate that a similar reaction mechanism (i.e., formation of carcinogen-protein complexes) might operate in many or all cases of chemical carcinogenesis (E. C. Miller and J. A. Miller, 1952). From these correlations it has seemed reasonable to suggest a tentative hypothesis of carcinogenesis by DAB and related dyes. According t o this proposal the administered dye is metabolized to a derivative (the primary carcinogen, possibly a N-hydroxymethyl aminoaso dye) capable of combining with certain liver proteins (the initial sites of action). These liver proteins are considered to play key roles in the response of the cell to its intrinsic growth controls (e.g., competitive reactions) and to the extrinsic growth controls (e.g., hormones) exercised by the rest of the organism. In any case the binding of these proteins with the dye metabolite is thought to reduce or prevent the further synthesis of these proteins so that eventually, in subsequent generations, cells would result with
390
JAMES A. MILLER AND ELIZABETH C. MILLER
less and finally none of the proteins originally attacked by the dye. Such cells could only respond to nutrition by continued growth and thus would be tumor cells, Although the centrifugation and electrophoresis data show that there is some specificity for the proteins attacked, not all the proteins thus affected need necessarily play a role in the carcinogenic reaction. The soluble proteins which bind the dye may not be the key proteins and may only serve as indicators of a similar reaction taking place with a protein in another part of the cell which is less amenable to study a t present. Since virus reproduction necessarily draws on the building blocks required by normal cells, it is clear that ultimately both a carcinogenic virus and a chemical carcinogen could produce the same result through altering the protein (enzyme) balance of the cell (see for example Potter, 1944, and Potter et al., 1950). Again, the destruction or partial inactivation of certain cell proteins by radiations could produce a similar state. Thus, the whole series of carcinogenic agents could be conceived as acting through the induction of abnormal protein patterns in a cell, The protein pattern leading to abnormal growth need not be identical in the induction of tumors by different carcinogens or even when different neoplastic cells are induced by a given carcinogen. But it must always leave the cell viable and free of many of the growth control systems. REFERENCES Andervont, H. B. 1947. J . Natl. Cancer Znst. 7 , 431-32. Andervont, H.B., and Dunn, T.B. 1947. J . Natl. Cancer Znst. 7 , 455-61. Andervont, H. B., and Edwards, J. E. 1941. J . Natl. Cancer Znst. 2, 139-49. Andervont, H. B., and Edwards, J. E. 1943a. J . Natl. Cancer Znst. 3,349-54. Andervont, H. B., and Edwards, J. E. 1943b. J . Natl. Cancer Inst. 3,355-58. Andervont, H. B., Grady, H. G., and Edwards, J. E. 1942. J . Natl. Cancer Znst. 3, 131-53. Andervont, H. B., White, J., and Edwards, J. E. 1944. J . Natl. Cancer Znst. 4, 583-86. Ando, T. 1941a. Gann 36,62-63. Ando, T. 1941b. Gann 86,20144. Antopol, W.,and Unna, K. 1942. Cancer Research 2, 694-96. Badger, G.M.,Cook, J. W., Hewett, C. L., Kennaway, E. L., Kennaway, N. M., and Martin, R. H. 1942. Proc. Roy. SOC.(London) Bl31, 170-82. Baumann, C. A. 1948. J . Am. Dietet. Assoc. 24,573-81. Baumann, C. A.,Rusch, H. P., Kline, B. E., and Jacobi, H. P. 1940. Am. J . Cancer 38, 76-80. Berman, C. 1951. Primary Carcinoma of the liver. Lewis, London. Boissonnae, R. A., Turner, R. A,, and du Vigneaud, V. 1949. J . Biol. Chem. 180, 1053-58. Boyland, E., and Brues, A. M. 1937. Proc. Roy. SOC.(London) Bl22, 429-41. Breedis, C., and Young, G. 1949. Federation Proc. 8, 351. Brock, N., Druckrey, H., and Hamperl, H. 1940. Z.Krebaforsch. 60,431-56.
THE CARCINOGENIC AMINOAZO DYES
391
Brown, R. R., Miller, J. A., and Miller, E. C. 1952. Federation Roc. 11, 192. Bullock, F. D., and Curtis, M. R. 1924. J . Cancer Research 8, 446-81. Clayton, C. C., and Baumann, C. A. 1949. Cancer Research 9,575-82. &hen, A., and Cohen, L. 1951. Nature 167, 1063. Cohen, P. P. 1945. Cancer Research 6, 626-30. Cook, H. A., Griffin, A. C., and Luck, J. M. 1949. J . Biol. Chem. 177,373-81. Cook, J. W., Haslewood, G. A. D., Hewett, C. L., Hieger, I., Kennaway, E. L., and Mayneord, W. V. 1937. Am. J . Cancer 29, 219-59. Cook, J. W., Hewett, C. L., Kennaway, E. L., and Kennaway, N. M. 1940. Am. J. Cancer 40, 62-77. Copeland, D. H., and Salmon, W. D. 1946. Am. J . Path. 22, 1059-67. Cortell, R. 1947. Cancer Research 7, 158-61. Crabtree, H. G. 1949. Brit. J . Cancer 3, 387-98. Cunningham, L.,Griffin, A. C., and Luck, J. M. 1950a. Cancer Research 10, 194-99. Cunningham, L., Griffin, A. C., and Luck, J. M. 1950b. Cancer Research 10,211. Dalton, A. J., and Edwards, J. E. 1942. J . Natl. Cancer Inst. 3, 319-29. Davidson, J. N., and Waymouth, C. 1944. Biochem. J . 38,379-85. Day, P. L., Payne, L. D., and Dinning, J. S. 1950. Proc. SOC.Ezptl. Biol. Med. 74, 854-55. Druckrey, H. 1950a. Arch. klin. Chir. 264, 45-55. Druckrey, H. 1950b. Arch. exptl. Path. Pharmakol. 210, 137-58. Druckrey, H.,and Kupfmtiller, K. 1948. 2.Naturforsch. 31, 254-66. Edwards, J. E.,and White, J. 1941. J. Natl. Cancer Inst. 2, 157-83. Elson, L. A., and Hoch-Ligeti, C. 1946. Biochem. J . 40, 380-91. Elson, L. A., and Warren, F. L. 1944. Biochem. J . 38,217-20. Engel, R. W., Copeland, D. H., and Salmon, W. D. 1947. Ann. N . Y. Acad. Sci. 49,Art. 1, 49-67. Fischer, B. 1906. Milnch. med. Wochschr. 63, 2041-47. Fischer-Wasels, B. 1936. Centr. allgem. Path. u . path. Anat. 66, 359-60. Fraenkel-Conrat, H., and Olcott, H. S. 1948. J . Biol. Chem. 174, 827-43. Fujiwara, T., Nakahara, W., and Kishi, S. 1937. Gann 31,51-62;abstract in 1938, Am. J. Cancer 33, 283. Gallico, E., and Boretti, G. 1948. Tumori 22, 130-138;abstract in 1949, Cuncer 8, 564. Giese, J. E., Clayton, C. C., Miller, E. C., and Baumann, C. A. 1946. Cancer Research 6, 679-84. Giese, J. E., Miller, J. A., and Baumann, C. A. 1945. Cancer Research 6, 337-40. Gillman, J., Gillman, T., and Gilbert, C. 1949. 8.African J . Med. Sci. 14,21-83. Greenstein, J. P. 1947. Biochemistry of Cancer. Academic Press, New York. Griffin, A. C., and Baumann, C. A. 1946. Arch. Biochem. 11, 467-76. Griffi, A. C., and Baumann, C. A. 1948a. Cancer Research 8, 135-38. Griffi, A. C., and Baumann, C. A. 1948b. Cuncer Research 8,279-84. Griffin, A. C., Bloom, S., Cunningham, L., Teresi, J. D., and Luck, J. M. 1950. Cancer 3, 316-20. Griffin, A. C., Clayton, C. C., and Baumann, C. A. 1949. Cancer Research 9,82-87. Griffin, A. C., Cunningham, L., Brandt, E. L., and Kupske, 0. W. 1951. Cancer 4, 410-15. Griffi, A. C., Nye, W. N., Noda, L., and Luck, J. M. 1948. J . Biol. Chem. 176, 1225-35. Gyorgy, P., Poling, E. C., and Goldblatt, H. 1941, Proc. SOC.Exptl. Biol. Med. 47, 41-44.
392
JAMES A. MIhLER AND ELIZABETH C. MILLER
Gyorgy, P., Tomarelli, R., Ostergard, R. P., and Brown, J. B. 1942. J . Exptl. Med. 76, 413-20. Haddow, A., Harris, R. J., Kon, G. A. R., and Roe, E. M. F. 1948. Trans. Roy. SOC.(London) A241, 147-95. Harris, P. N. 1947a. Cancer Research 7, 36-36. Harris, P. N. 1947b. Cancer Research 7 , 178-79. Harris, P. N. 1949. Cancer Reaearch 9, 602. Harris, P. N., Krahl, M. E., and Clowes, G. H. A. 1947% Cancer Research 7,162-75. Harris, P. N., Krahl, M. E., and Clowes, G. H. A. 194713. Cancer Research 7 , 176-77. Hartwell, J. L. 1941. Survey of Compounds Which Have Been Tested for Carcinogenic Activity. U.S.Public Health Service, Washington, D.C. Heep, W. 1936. Frankfurt. 2.Path. 60, 48-62. Heidelberger, C., and Weiss, S. M. 1951. Cancer Research 11, 885-91. Higgins, H., Miller, J. A., Price, J. M., and Strong, F. M. 1950. Proc. SOC.Exptl. Biol, Med. 76, 462-65. Hoch-Ligeti, C. 1947. Cancer Research 7 , 148-57. Hoch-Ligeti, C. 1949. Brit. J. Cancer 8, 285-88. Hoch-Ligeti, C., Hoch, H., and Goodall, K. 1949. Brit. J . Cancer 8, 140-47. Jablonski, C. F. 1951. Chap. IX. Coloring Matters in Foods. In Jacobs, M. B. (Editor), Food and Food Products, 2nd Ed. Interscience Publishers, New York. Jacobi, H. P., and Baumann, C. A. 1942. Cancer Research 2, 175-80. Kensler, C. J. 1947. Ann. N . Y . Acad. Sci. 49, Art. 1, 29-40. Kensler, C. J. 1948. Cancer 1, 483-88. Kensler, C. J. 1949. J . Biol. Chem. 179, 1079-84. Kensler, C. J., and Chu, W. C. 1950. Arch. Biochem. 26, 66-73. Kensler, C. J., Dexter, S. O., and Rhoads, C. P. 1942a. Cancer Research 2, 1-10, Kensler, C. J., and Langemann, H. 1951. Cancer Research 11, 264. Kensler, C. J., Magill, J. W., and Sugiura, K. 1947. Cancer Research 7, 95-98. Kensler, C. J., Sugiura, K., and Rhoads, C. P. 1940. Science 91, 623. Kensler, C. J., Sugiura, K., Young, N. F., Halter, C. R., and Rhoads, C. P. 1941. Science 98,308-10. Kensler, C. J., Young, N. F., and Rhoads, C. P. 1942b. J . B i d . Chern. 148,465-72. Kinosita, R. 1936. Gann SO, 423-26. (In Japanese.) Kinosita, R. 1937. Jap. Path. SOC.Trans. 27, 665-727. Kinosita, R. 1940a. Yale J . B i d . and Med. 12, 287-300. Kinosita, R. 1940b. Gann 84, 165-67. Kirby, A. H. M. 1945a. Cancer Research 6, 673-82. Kirby, A. H. M. 1945b. Cancer Research 6, 683-96. Kirby, A. H. M. 1947a. Cancer Research 7 , 263-67. Kirby, A. H. M. 194713. Cancer Research 7 , 333-41. Kirby, A. H. M. 1948a. Biochem. J. 43, lv. Kirby, A. H. M. 194813. Brit. J . Cancer& 290-94. Kirby, A. H. M., and Peacock, P. R. 1946. Brit. J . Exptl. Path. 27, 179-89. Kirby, A. H. M., and Peacock, P. R. 1947. J . Path. Bact. I S , 1-18. Kirby, A. H. M., and Peacock, P. R. 1949. Glasgour Med. J . 80,364-72. Kishi, S., Fujiwara, T., and Nakahara, W. 1937. Gann 81, 1-11; abstract in 1938. Am. J . Cancer 88, 283. Kline, B. E., Miller, J. A., and Rusch, H. P. 1946. Cancer Research 6 , 641-43. Kline, B. E., Miller, J. A., Rusch, H. P., and Baumann, C. A. 1946a. Cancer Research 0, 1-4.
THE CARCINOGENIC AMINOAZO DYES
393
Kline, B. E., Miller, J. A., Rusch, H. P., and Baumann, C. A. 1946b. C a n w Reaearch 6, 5-7. Kline, B. E., and Rusch, H. P. 1944. Cancer Research 4, 762-67. Kuhn, R., and Beinert, H. 1943. Ber. 76, 904-09. Kuhn, R., and Quadbeck, G. 1949. 2.Krebsfmsch. 66,242-45. Langemann, H., and Kensler, C. J. 1951. Cancer Research 11, 265. Langer, E. 1942. 2. Krebsforsch. 62, 443-54. Law, L. W. 1941. Cancer Research 1,397-401. Lowenhaupt, E. 1949. Cancer Research 9, 121-26. MacDonald, J. C., Miller, E. C., Miller, J. A., and Rusch, H. P. 1952a. Cancer Research 12, 50-54. MacDonald, J. C., Plescia, A. M., Miller, E. C., and Miller, J. A. 195213. Cancer Research 12, 280. Mark, D. D. 1950. Arch. Path. 49, 545-54. Mark, D. D., and Ris, H. 1949. Proc. SOC.Exptl. Biol. Med. 71, 727-29. Masayama, T., Iki, H., Yokoyama, T., and Hasimoto, H. 1938. Gann 32,303-06. Mwayama, T., and Yokoyama, T. 1940. Gann 34, 174-75. Meister, A. 1950. J. Natl. Cancer Inst. 10, 1263-71. Mellors, R. C., and Sugiura, K. 1948. Proc. SOC.Exptl. Biol. Med. 67, 242-46. Miller, E. C. 1951. Cancer Research 11, 100-08. Miller, E. C., and Baumann, C. A. 1946. Cancer Research 6, 289-95. Miller, E. C., Baumann, C. A., and Rusch, H. P. 1945. Cancer Research 6, 713-16. Miller, E. C., and Miller, J. A. 1947. Cancer Research 7 , 468-80. Miller, E. C., and Miller, J. A. 1952. Cancer Research 12, 547-56. Miller, E. C., Miller, J. A., and Brown, R. R. 1952a. Cancer Research 12,282-83. Miller, E. C., Miller, J. A., Kline, B. E., and Rusch, H. P. 1948. J. Exptl. Med. 88, 89-98. Miller, E. C., Miller, J. A., Sandin, R. B., and Brown, R. K. 1949a. Cancer Research 9, 504-09. Miller, E. C., Miller, J. A., Sapp, R. W., and Weber, G. M. l949b. Cancer Research 9, 336-43. Miller, E. C., Plescia, A. M., Miller, J. A,, and Heidelberger, C. 1952b. J. Biol. Chem. 196, 863-74. Miller, J. A. 1947. Ann. N . Y . Acad. Sci. 49, Art. 1, 19-28. Miller, J. A., and Baumann, C. A. 1945a. Cancer Research 6 , 157-61. Miller, J. A., and Baumann, C. A. 1945b. Cancer Research 6 , 227-34. Miller, J. A., Brown, R. R., Miller, E. C., and Mueller, G. C. 1951a. Cancer Research 11, 269. Miller, J. A., Kline, B. E., and Rusch, H. P. 1946. Cancer Research 6, 674-78. Miller, J. A., Kline, B. E., Rusch, H. P., and Baumann, C. A. 1944a. Cancer Research 4, 153-58. Miller, J. A., Kline, B. E., Rusch, H. P., and Baumann, C. A. 194413. Cancer Research 4, 756-61. Miller, J. A., and Miller, E. C. 1947. Cancer Research 7 , 39-41. Miller, J. A., and Miller, E. C. 1948. J. Ezptl. Med. 87, 139-56. Miller, J. A., and Miller, E. C. 1952a. Unpublished. Miller, J. A., and Miller, E. C. 1952b. Cancer Research 12, 283. Miller, J. A., Miller, E. C., and Baumann, C. A. 1945. Cancer Research 6, 162-68. Miller, J. A,, Miller, E. C., and Sapp, R. W. 1951b. Cancer Research 11, 269. Miller, J. A., Sapp, R. W., and Miller, E. C. 1948. J. Am. Chem. SOC.70,346&63. iller, J. A., Sapp, R. W., and Miller, E. C. 1949. Cancer Research 9, 652-60.
394
JAMES A. MILLER AND ELIZABETH C. MILLER
Miller, W. L., Jr., and Baumann, C. A. 1951. Cancer Research 11, 634-39. Miner, D. L., Miller, J. A., Baumann, C. A., and Rusch, H. P. 1943. Cancer Research 8, 296-302. Mori, K. 194la. Gann 36, 106-19. Mori, K. 1941b. Gann 36, 121-25. Mueller, G. C., and Miller, J. A. 1948. J. Biol. Chem. 176, 535-44. Mueller, G. C., and Miller, J. A. 1949. J. Biol. Chem. 180, 1125-36. Mueller, G. C., and Miller, J. A. 1950. J. Biol. Chem. 186, 145-54. Mueller, G. C., and Miller, J. A. 1951. Cancer Research 11,271. Nagao, N. 1940. Gann 34, 13-19. Nagao, N. 1941a. Gann 36,8-20. Nagao, N. 1941b. Gann 36,280-82. (In Japanese.) Nakahara, W., Kishi, S., and Fujiwara, T. 1936. Gann SO, 499-508; abstract in 1936. Am. J. Cancer 28, 790. Olson, R. E. 1951. Cancer Research 11, 571-84. Opie, E. L. 1944a. J. Exptl. Med. 80, 219-30. Opie, E. L. 1944b. J. Exptl. Med. 80, 231-46. Opie, E. L. 1946. J. Exptl. Med. 84, 91-106. Opie, E. L. 1947a. J. Exptl. Med. 86, 339-46. Opie, E. L. 1947b. J. Exptl. Med. 86, 45-54. Opie, E. L., and Lavin, G. I. 1946. J. Ezptl. Med. 84, 107-12. Orr, J. W. 1940. J. Path. Bact. 60, 393-408. Orr, J. W., and Price, D. E. 1948. J. Path. Bact. 60, 461-69. Pearson, B., Novikoff, A. B., and Morrione, T. G. 1950. Cancer Research 10, 557-64. Plaut, G. W. E., Betheil, J. J., and Lardy, H. A. 1950. J . B i d . Chem. 184,795-805. Pollack, M. A., Taylor, A., Taylor, J., and Williams, R. J. 1942a. Cancer Research a, 739-43. Pollack, M. A., Taylor, A., Woods, A., Thompson, R. C., and Williams, R. J. 1942b. Cancer Research 2, 748-51. Potter, V. R. 1942. Cancer Research 2, 688-93. Potter, V. R. 1944. Advances in Enzymol. 4, 201-56. Potter, V. R. 1951. Cancer Research 11, 565-70. Potter, V. R., and DuBois, K. P. 1943. J. Gen. Physiol. 26, 391-404. Potter, V. R., Price, J. M., Miller, E. C., and Miller, J. A. 1950. Cancer Research 10, 28-35. Price, J. M., Harman, J. W., Miller, E. C., and Miller, J. A. 1952. Cancer Research 12, 192-200. Price, J. M., and Laird, A. K. 1950. Cancer Research 10, 650-58. Price, J. M., Miller, E. C., and Miller, J. A. 1948. J. Bio2. Chem. 173, 345-53. Price, J. M., Miller, E. C., and Miller, J. A. 1949a. Proc. SOC.Exptl. Biol. Med. 71, 575-78. Price, J. M., Miller, E. C., Miller, J. A., and Weber, G. M. 1949b. Cancer Research 9, 398-402. Price, J. M., Miller, E. C., Miller, J. A., and Weber, G. M. 1950. Cancer Research 10, 18-27. Price, J. M., Miller, J. A., and Miller, E. C. 1951. Cancer Research 11, 523-28. Price, J. M., Miller, J. A., Miller, E. C., and Weber, G. M. 1949c. Cancer Research 9,96-102. Richardson, H. L., and Borsos-Naohtnebel, E. 1951. Cancer Research 11, 398-403. Richardson, H. L., Stier, A. R., and Borsos-Nachtnebel, E. 1952. Cancer Research 12,356-61.
THE CARCINOQENIC AMINOAZO DYES
395
Roskelley, R. C., Mayer, N., Horwitt, B. N., and Salter, W. T. 1943. J . Clin. Invest. 22, 743-51. Rumsfeld, H. W., Jr., Miller, W. L., Jr., and Baumann, C. A. 1951. Cancer Research 11, 814-19. Rusch, H. P., Baumann, C. A., Miller, J. A,, and Kline, B. E. 1945. Experimental Liver Tumors. In Moulton, F. R. (Editor). Research Conference on Cancer, American Association for the Advancement of Science, Washington, D. C. Rusch, H. P., and LePage, G. A. 1948. Ann. Rev. Biochem. 17,471-94. Rusch, H. P., and Miller, J. A. 1948. Proc. SOC.Exptl. Biol. Med. 08, 140-43. Sakami, W. 1948. J. Biol. Chem. 170,995-96. Salaberg, D. A., Hane, S., and Griffin, A. C. 1951. Cancer Research 11, 276. Sasaki, T., and Yoshida, T. 1935. Virchow’s Arch. path. Anat. 296, 175-200. Sauberlich, H. E., and Baumann, C. A. 1951. Cancer Research 11, 67-71. Schiller, W. 1937. Am. J. Cancer S1, 486-90. Schmidt, M. B. 1924. Virchow’s Arch. path. Anal. 26S, 432-51. Schneider, W.C. 1945a. J. Biol. Chem. 101, 293-303. Schneider, W. C. 194513. Cancer Research 6, 717-21. Schweigert, B. S.,Guthneck, B. T., Price, J. M., Miller, J. A., and Miller, E. C. 1949. Proc. SOC.Exptl. Biol. Med. 72, 495-501. Shear, M. J. 1937. Am. J . Cancer 29, 269-84. Siegel, I., and Lafaye, J. 1950. Proc. SOC.Exptl. Biol. Med. 74, 620-23. Siekevite, P., and Greenberg, D. M. 1949. J . Biol. Chem. 180,845-56. Silverstone, H. 1948. Cancer Research 8, 301-08. Sorof, S.,and Cohen, P. P. 1951. Cancer Research 11, 376-82. Sorof, S., Cohen, P. P., Miller, E. C., and Miller, J. A. 1951. Cancer Research 11, 383-87. Spite, S., Maguigan, W. H., and Dobriner, K. 1950. Cancer 3, 789-803. Stevenson, E. S., Dobriner, K., and Rhoads, C. P. 1942. Cancer Research 2,160-67. Stowell, R. E. 1949. Cancer 2, 121-31. Strong, L. C.,Smith, G. M., and Gardner, W. U. 1938. Yale J . Biol. and Med. 10, 335-46. Sugiura, K. 1942. Proc. SOC.Exp. Biol. Med. 60,214-15. Sugiura, K. 1946. Proc. SOC.Exp. Biol. Med. 61,301-02. Sugiura, K. 1948. Cancer Research 8, 141-44. Sugiura, K. 1951. J. Nutrition 44, 345-60. Sugiura, K.,Halter, C. R., Kensler, C. J., and Rhoads, C. P. 1945. Cancer Research 6,235-38. Sugiura, K., and Rhoads, C. P. 1941. Cancer Research 1, 3-16. Taki, I., and Miyaji, T. 1950. Gann 41, 194-95. Tarver, H. 1951. Chap. XIII. The Metabolism OF Amino Acids and Proteins. In Greenberg, D. M. (Editor), Amino Acids and Proteins. Charles C. Thomas, Springfield, Ill. Taylor, A., Pollack, M. A,, Hofer, M. J., and Williams, R. J. 1942a. Cancer Research 2, 744-47. Taylor, A., Pollack, M. A., Hofer, M. J., and Williams, R. J. 1942b. Cancer Research 2, 752-54. Tung, T. C., and Cohen, P. P. 1950. Cancer Research 10,793-96. du Vigneaud, V., Spangler, J. M., Burk, D., Kensler, C. J., Sugiura, K., and Rhoads, C. P. 1942. Science 96, 174-76. Viollier, G. 1950a. Helv. Physiol. Acta 8, C34-C36. Viollier, G. 1950b. Helv. Physiol. Acta 8, C37-C39. Viollier, G., and Waser, P. 1950. Helv. Physiol. Acta 8, C39-C41.
396
JAMES A. MILLER AND ELIZABETH C. MILLER
West, P. M., and Woglom, W. H. 1942. Cancer Research 2, 324-31. Westerfeld, W. W., Richert, D. A., and Hilfinger, M. F. 1950. Cancer Research 10, 486-94.
White, F. R., Eschenbrenner, A. B., and White, J. 1948. Acta union intern. conlre cancer 6, 75-78. White, F. R., and White, J. 1946. J . Natl. Cancer Znst. 7 , 99-101. White, J., and Edwards, J. E. 1942% J . Natl. Cancer Znst. 2, 535-38. White, J., and Edwards, J. E. 1942b. J . Natl. Cancer Inst. 8, 43-59. Wiest, W. G., and Heidelberger, C. 1952. Cancer Research 12, 308. Wilson, R. H., De Eds, F., and Cox, A. J. 1941. Cancer Research 1, 595-608. Woglom, W. H. 1913. Studies in Cancer and Allied Subjects, Vol. 1. Columbia University Press, New York. Woodard, H. Q. 1943. Cancer Research 9, 159-63. Yoshida, T. 1933. Trans. Jap. Path. SOC.as, 636-38. (Cited by Shear, M. J. AT J . Cancer 20, 269-84.) Zamecnlk, P. C., and Frantz, I. D. 1949. Cold Spring Harbor Symposia Quant. Biol. 14, 199-208. Zamecnik, P. C., Frantz, I. D., Loftfield, R. B., and Stephenson, M. L. 1948. J . Biol. Chem. 176, 299-314. Zamecnik, P. C., Loftfield, R.B., Stephenson, M. L., and Steele, J. M. 1951. Cancer Research 11, 592-602.
The Chemistry of Cytotoxic Alkylating Agents W. C. J. ROSS* Chester Beatty Research Institute, Royal Cancer Hospital, London, England
CONTENTS
I. Introduction 11. PChloroethyl Sulfides (Sulfur Mustards) 1. Reactions in Water 2. Reactions with Anions 3. Reactions with Bases 4. Reactions with Proteins 6. Reactions with Nucleic Acids 111. 2-Chloroethylamines Aliphatic Derivatives (Aliphatic Nitrogen Mustards) 1. Reactions in Water 2. Reactions with Anions 3. Reactions with Bases 4. Reactions with Proteins 5. Reactions with Nucleic Acids Aromatic Derivatives (Aromatic Nitrogen Mustards) 1. Reactions in Water 2. Reactions with Anions 3. Reactions with Bases 4. Reactions with Proteins IV. 1,2-Epoxides 1. Reactions in Water 2. Reactions with Anions 3. Reactions with Bases 4. Reactions with Proteins 5. Reactions with Nucleic Acids V. Miscellaneous Agents VI. Discussion References
Page 397 399 399 402 406 408 410 41 1 41 1 41 1 415 416 418 419 419 419 425 427 429 429 429 432 434 434 435 436 437 440
I. INTRODUCTION In recent years a considerable amount of work has been devoted to the study of the chemotherapy of cancer using agents related t o the vesicant war gas, di-2-chloroethyl sulfide (mustard gas). The first agents to be used extensively were two so-called nitrogen mustards, methyldi-2chloroethylamine (HNJ and tri-2-chloroethylamine (HNa); there is a * British Empire Cancer Campaign Research Fellow. 397
398
W. C. J. ROBS
very extensive literature on the use of these compounds; references to review articles are given by Philips (1950). These agents are administered intravenously as their water-soluble hydrochlorides. A compound of similar type, 2-naphthyldi-2‘-ch1oroethylamineJhas received clinical trials in England (Matthews, 1950); it has an effect essentially similar to that of HNa but is slower acting and more easily controlled.* It also has the advantage of being effective when administered orally. Some measure of success has attended the use of these mustards in thetreatment of neoplastic diseases of the hematopoietic organs. No significant effect on well-established malignant growths a t other sites has yet been observed. It is, however, hoped that these compounds will find an application as an adjunct to surgical and radiation treatment since they may well be able to control the growth of smaller metastases which are inaccessible t o these other technics. Boyland (1948) has drawn attention to the similarity between the biological actions of x-irradiation and of the mustards and used the term “radiomimetic,” which had previously been employed by Dustin (1947), to describe the effects of the chemical agents. These actions include vesication, the delayed lethal effect due to hemoconcentration and leucopenia, the inducement of chromosomal abnormalities, the mutagenic action, the characteristic damage caused to the bone marrow, the bleaching of the hair at the site of action, the inhibition of numerous enzyme systems, the retarding action on the growth of certain neoplasms, and the ability to suppress antibody formation. References to the work published on these effects may be found in Boyland’s review and in an article by Philips (1950). This latter author points out that while the term “radiomimetic” is an attractive one it should be used with due regard for the differences between the effects produced by the mustards and radiation. In particular the types of chromosomal aberration induced appear to depend upon the agent used (Ford, 1948; Auerbach, 1949) and even to vary from one chemical agent to another (Loveless and Revell, 1949, 1950). Although prior to 1939 di-2-chloroethylsulfide was not regarded as particularly reactive under physiological conditions, work done since that time has shown that both sulfur and nitrogen mustards are capable of direct reaction with many functional groups in proteins (Banks et at., 1946; Fruton et al., 1946a; Herriott et at., 1946) and in nucleic acids (Elmore et al., 1948; Fruton et al., 1946b). The close relationship between the biological activity and the chemical reactivity of the aromatic
* Since this review was written several Italian workers have published the results of clinical trials using the Znaphthylamine derivative, for example, P. Introzzi, and M. Ninni, 1950, Haematobgica 84, 925-65.
CYTOTOXIC ALKYLATINQ AGENTS
399
derivatives of nitrogen mustard gas strongly suggests that such biological activity is due t o a chemical reaction rather than to a purely physical effect on the system (Haddow et al., 1948). The importance of the chemical reactivity of these agents has been amply demonstrated by the recent finding of very similar biological activity in the case of epoxides (Rapoport, 1947b, 1948; Loveless and Revell, 1949; Ross, 1950c), ethyleneimines (Lewis and Crossley, 1950; Burchenal et al., 1950), p-tolueneand methanesulfonic acid esters (Timmis, 1949, 1950), and methyl sulfate (Loveless, 1950), all of which are capable of reacting in a similar manner under mild conditions. In this review a survey will be made of the reactions of a group of radiomimetic agents all of which may be regarded as alkylating agents under physiological conditions of temperature and pH. It is proposed to consider each class of compound with special reference to its reactions under mild conditions and with few exceptions no reaction which takes place under more drastic conditions will be discussed. In the main section of this article the reactions of each agent with water, anions, bases, proteins, and nucleic acids, where these have been studied, will be described, and in the final section it will be shown that each type of compound can be regarded as an electrophilic reagent capable of reacting with nucleophilic groups present in biological systems. The functional groups of such systems will be discussed in the light of their potential nucleophilic capacity.
11.
2-CHLOROETHYL SULFIDES (SULFUR
MUSTARDS)
The reactions of di-Zchloroethyl sulfide have recently been reviewed by Ogston (1948a), Boursnell (1948), and Needham (1948). 1. Reactions in Water
Unlike ordinary alkyl halides, compounds with a halogen atom in the 8-position to a nitrogen or sulfur atom (I) are very reactive in polar solvents, tliough they have no outstanding reactivity in nonpolar solvents. The activating effect of the hetero atom can be ascribed t o its inductive effect, electron repelling in character, which will facilitate the removal of the halogen atom leaving the positively charged carbonium ion (11)
400
W. C. J. ROSS
Price and Wakefield (1947) have suggested that this carbonium ion must be stabilized by,[passing into the cyclic ethylene sulfonium structure (111) since it is able t o discriminate between the molecular species with which it might react. It is also possible to regard the tendency to form a cyclic structure as the driving force for the ionization: CHz / \ R.8 ............ CHa.*.*.Cl+ R.5
+ c1-
the reaction being an internal bimolecular one (Lord, 1946; Hirst, 1950). However, Price and Wakefield (1947) consider that the formation of the ion precedes cyclization, and though Ogston’s work threw no light on the chemical nature of the activated form, he was able to rule out the existence in reaction mixtures of significant concentrations of an internal sulfonium compound. Stein and Fruton (1946) also found no evidence for the accumulation of a cyclic form in solutions containing di-2-chloroethyl sulfide. The formation of such a cyclic ion will depend on whether the tendency of the lone pair of electrons on the sulfur atom to form a bond with the 8-carbon atom is sufficiently powerful to distort the valency angles of carbon or whether a more stable structure can be produced if the &carbon atom coordinates with an electron pair from an external source such as a water molecule. That it is possible for an inductive effect to modify the reaction at the P-carbon atom without the necessity of postulating a cyclic intermediary is shown by the relative ease with which esters of hydroxyethyl sulfides are hydrolyzed as compared with ethyl esters (Davis and Ross, 1950). Whatever the merits of the sulfoniurn ion theory it is possible to interpret the reactions of di-2-chloroethyl sulfide on the basis of a primary ionization t o yield a carbonium ion (11). With a water molecule in dilute solution the ion will react thus: R.S*CH&Hz+
+
In more concentrated solutions the sulfur atom in a second molecule of sulfide becomes an effective competitor for reaction with the ion (11), the hydroxyethyl sulfide being more likely to react than the chloroethyl sulfide on account of the greater availability of the electrons in the sulfur atom of the former substance, and a sulfonium ion is formed:
401
CYTOTOXIC ALKYLATING AGENTS
R.SGHzCHz+
+ :S
+/
\
R.S.CH2CHzS CHaCHnOH
\
R
03) CHzCHzOH
Mustard gas hydrolyzes in dilute ( thiocyanate tion with epichlorohydrin was: iodide (2.3 X > bromide (1.4 X lo-*) > chloride (2.6 X > acetate (1.5 X (1.4 X > benzoate (1.2 X > formate (1.1 X sec.-l), the figures in brackets being the velocity constants for reaction with the anion at 20'. Ross (1950~)found the following order of reactivity for anions toward 1,2-3,4-diepoxybutane: thiosulphate (100) > iodide (97) > citrate (82.5) > chloride (80.3) > acetate (75.5) > benzoate (65.6) > tartrate (58.1) > oxalate (53.6) > formate (31.2) > nitrate (7.6); in this case the figures in brackets represent the percentage of the epoxide which is converted into ester under identical conditions of reaction. The order of reactivity of the anions is essentially the same in both cases and is comparable with the order of the competition factors of the same ions toward the mustard gas types of compound. The nitrate ion competes only feebly, and the formate ion is the least reactive of the organic ions. The high reactivity of the thiosulfate and iodide ions is common to both types of compound. The anions are clearly being placed in the order of their relative nucleophilic tendencies. The value of the velocity constant for reaction with glycidol is 5.7 X loFe set.-' for the chloride ion at 20' (Bronsted et al., 1929) and 2.3 X set.-' for the hydroxyl ion a t 37" (Ross, 1950~). Whilst these figures cannot be directly compared they do dispose of a statement by Bronsted et al. which has been accepted and discussed b y later workers
CYTOTOXIC ALKYLATING AGENTS
433
(Hammett, 1940; Bartlett et al., 194713; Grunwald and Winstein, 1948; Bartlett and Small, 1950) to the effect that the hydroxyl ion does not react with an epoxide a t a rate comparable with that of the chloride ion. The reaction of ethylene oxide and propylene oxide with the anionic forms of substituted phenols has been studied by Boyd and Marle (1914). Epoxides do not appear to react with aliphatic hydroxyl groups in neutral solution, but will do so if sufficient alkali is added. For example, sucrose, which does not react in purely aqueous solution, reacts readily in the presence of 0.25N-sodium hydroxide (LeMaistre and Seymour, 1948). The p H of such a solution is between 13 and 14 and since the pK of the hydroxyl groups in sucrose is 12.6, an appreciable proportion of these groups will be in the ionic form in the alkaline solution. Nevertheless, the reaction of epoxides with aliphatic hydroxyl groups is unlikely to be important under physiological conditions. Sulfur-containing anions are known to be powerfully nucleophilic and the reaction of an epoxide with the thiosulfate ion, described by Culvenor et al. (1949), has been utilized in a convenient rapid method for the estimation of oxirane oxygen (Ross, 1950~). The rate of reaction of various epoxides with the thiosulfate anion has also been shown to run parallel with their cytotoxic activities. Increasing substitution of the carbon atoms in the oxirane ring reduces this reaction rate and a t the same time reduces the toxicity of the compound and its capacity to inhibit the growth of a transplanted rat carcinoma (Ross, 1950~). Culvenor et al. (1949) have also shown that ethylene oxides react rapidly with sulfides, sulfites, and alkaline solutions of thiols. There are indications that in the same way that substitution on the @-carbonatom of a chloroethylamine reduces the ability of the compound to react with an anion (p. 426), substitution on the terminal carbon atom of an epoxide reduces its ability to react with an anion. Just as the reaction of an epoxide with water may be catalyzed by the presence of hydrogen ions so may the reaction with an anion. The intermediate conjugate acid (XLVII) reacts with the anion to give a hydroxy ester thus:
(XLVII)
However, the acid catalyzed reaction of most epoxides with anions is not important in the pH range 4-8.5 and so is of little significance in the reaction of these epoxides with biological material a t physiological pH.
W. C. J. ROBS
434
An exception may be the reaction of 1,2-5,6-diepoxyhexane which has a high value for the rate coefficient of the acid catalyzed reaction and so this reaction may become appreciable at pH 7.5 (see also below p. 438). 3. Reactions with Bases
As would be expected the nucleophilic alpino group reacts readily with the oxirane ring system. The stepwise formation of triethanolamine from ammonia demonstrates the types of product formed:
The reaction probably involves an attack by the amino group on the carbon atom of the epoxide followed by proton transfer from a water molecule to the oxygen atom (compare p. 430). This mechanism is in accord with the finding of Knorr (1899) that the reaction with amines proceeds much more readily in the presence of water and may not occur a t all under completely anhydrous conditions. Smith et al. (1946) have studied the reaction of different epoxides with ammonia and with primary, secondary, and tertiary amines. They find that the rates of reaction of amines with epichlorohydrin is greater than with ethylene oxide. As has been pointed out on p. 431 this is consistent with an initial nucleophilic attack on the terminal carbon atom. Smith et al. also find that the rates of reaction of an epoxide with substituted amines are in the following order, the most reactive being placed first: MesN
> MeZNH > MeNHz > NHs
this is the order of basicity of the amines and hence also the order of their nucleophilic capacities. Fraenkel-Conrat (1944) has observed that amine groups react preferentially in alkaline solution; this would be anticipated since the cationic form present in acid solutions, RNHS+, is not nucleophilic. Kiprianov (1926) has isolated a small proportion of the di(hydroxyethy1)amino acid (XLVIII) from the reaction of glycine ester with ethylene oxide but the lactone (XLIX) constitutes the major part of the product. (HOCHzCHz)~NCH,COgH
I /
HOCHzCHzNCHzCO
‘ 0
CHzCHz (XLVIII)
(XLIX)
4. Reactions with Proteins
Fraenkel-Conrat (1944) has shown that water-soluble epoxides such as ethylene oxide and propylene oxide readily react with crystalline egg
435
CYTOTOXIC ALKYLATINO AGENTS
albumin and with p-lactoglobulin under mild conditions. The isoelectric points of the protein derivatives were moved 1 to 3 pH units toward the alkaline side as compared with the starting material, and the new compounds were insoluble in the isoelectric region and more soluble on the acid side than on the alkaline side of the isoelectric point. These effects on the properties of the protein are consistent with the esterification of a high proportion of the carboxyl groups. Evidence was also obtained for the reaction of the epoxides, when present in large excess, with phenolic, primary amino, and sulfhydryl groups; the two former groups reacted more completely in alkaline than in acid solutions. No evidence for the further reaction of the newly introduced hydroxyalkyl groups with epoxide was obtained, and it was considered unlikely that aliphatic hydroxyl groups would react under the mild conditions used. The hydroxyalkyl group linkages were stated to be surprisingly resistant to both acid and alkaline hydrolysis though small proportions of the substituted carboxyl and amino groups were hydrolyzed. 1,2-5,6-Dianhydro-3,4-acetonemannitol (L, first prepared by Wiggins, 1946) reacts with the acidic side chains of the wool fiber giving a crosslinked structure (Speakman, 1948). Fibers treated with a 4% solution of this diepoxide a t pH 5 and at 50’ for twenty-four hours show an increased resistance to extension and a reduction in shrinkage in the milling process. CHz
‘o/
.
CH
.
CH
-
CH
A d
.
CH
*
CHz
‘d
‘C/ CH/, ‘CHa
&I It is of interest to note that the reaction of acidic groups in pectic and alginic acids with various epoxides has been demonstrated (Deuel, 1947) and one diepoxide, 1,2-3,4-diepoxybutane, has been shown to cross-link two alginic acid molecules by reaction with the carboxyl groups. 6. Reactions with Nucleic Acids Alexander (1951) has shown that propylene oxide and 1,2-4,5-diepoxypentane react with the phosphate groups in thymus nucleic acid. The reaction with the acid groups was assessed by measuring the uptake of a basic dye, methylene blue, before and after treatment with the agent. The amount of basic dye fixed by the nucleoprotein was practically restored to the initial value following the action of alkali due almost certainly to the hydrolysis of the phosphate ester linkages. The reaction of nucleic acid phosphate groups with the triaaine (LII), methyldi-2-
W. C. J. ROSS
436
chloroethylamine, dimethyl-2-chloroethylamine, and the bis-methanesulfonate (LV) has also been established by this method. An interesting observation was the finding that for equal amounts of esterification of the phosphate groups difunctional compounds were more effective in reducing the affinity of the nucleic acid for protamine than the corresponding monofunctional compounds. If the suggestion that these cytotoxic agents act by preventing the association of nucleic acid with a basic protein (p. 419) is valid then these results may well indicate a reason for the considerable difference in effectiveness between mono- and difunctional alkylating agents. Preliminary work by Butler (1950) has not revealed any significant viscosity changes following the treatment of solutions of thymus nucleic acid with 1,2-3,4-diepoxybutane or 1,2-5,6-diepoxyhexane.
V. MISCELLANEOUS AGENTS Ethyleneimine (LI) and a number of its derivatives, notably 2,4,6triethyleneimino-l,3,5-triazine (LII) and the diethylene derivative of hexamethylenurea (LIII), have recently been shown to have similar effects on cell division and upon the growth of animal tumors to those elicited by the mustard gas types (Buckley et al., 1950; Burchenal et aZ., 1950; Lewis and Crossley, 1950; Sugiure and Stock, 1950; Biesele et al., 1950; Rose et aZ., 1950; Haddow, 1950; Philips, 1950). CHz
CHz
CHz
CHI N.CO.NH(CH2)6NH.CO.N
\
\
N
N
CHz
(LIII)
Very little has been done as yet in studying the reactions of these imines under mild conditions in aqueous solution, but their reactions are probably similar to those of epoxides. For example, Braz and Skorodumov (1947) have shown that ethyleneimine reacts with diethylamine in the presence of water but not in its absence (compare p. 434). Ross (1950~) has established that the triazine (LII) reacts with anions and with water in a manner similar to that of the epoxides. The triazine was characterized by a very high value for the coefficient for the acid catalyzed reactions. This compound has been used as ti cross-linking agent in the
437
CYTOTOXIC ALKYLATINQ AQENTS
textile industry (Preston, 1949), since it will react with the wool fiber under mild conditions. Divinyl sulfone will produce typical “radiomimetic ” effects on the growing root tips of Vicia faba (Loveless, 1950). This sulfone can react with nucleophilic centers such as anions by a Lowry mechanism since the ethylenic double bond is polarized by the proximity of the electron attracting sulfoxide system: R O
HO-H
A
+ &&CH2
s-
+ R.SOzCHzCHzA
O
a+
+ OH-
+ A-
The reaction of the highly reactive thiosulfate ion with divinyl sulfone has been demonstrated by Stahman et al. (1946) and by Ross (1950~). Banks et al. (1946) consider that di-2-chloroethyl sulfone, which is converted into divinyl sulfone, and divinyl sulfone itself react mainly with amino and sulfhydryl groups in proteins, the reaction being of quite a different character from that of di-2-chloroethyl sulfide. Timmis (1949) has prepared p-toluenesulphonic esters of di-2hydroxyethylarylamines: two of these (LIV, R = H or C1) are active as tumor growth inhibitors. The reactions of these compounds will be exactly analogous to those of the halogenoethylarylamines (p. 419).
p-Toluenesulfonyl esters of glycols such as lJ3-propanediol and 2,4pentanediol are not biologically active, but activity has been demonstrated in the case of the dimethanesulfonyl ester of 1,4-butanediol (LV) (Timmis, 1950). This compound reacts with water and with the thiosulfate ion a t a rate comparable with that of the biologically active di(2,3-epoxypropyl)ether. The reactions of the dimethanesulfonyl esters are of the sN2 type; as in the case of the epoxides they involve nucleophilic displacements on carbon and the extent of the reaction is proportional to the concentration of the nucleophilic reagent (Ross, 1950d). VI. DISCUSSION The biologically active compounds discussed above fall into two groups in respect of their reactions in aqueous media. 2-Chloroethyl sulfides and amines react by an 8 N 1 mechanism while epoxides and sulfonyl esters of glycols react by an 812 mechanism. This difference in reaction mechanism has interesting consequences when the transforma-
438
W.
C. J. ROSS
tions of the agents in biological systems are considered. A feature of the S N 1 type of reaction is that the rate at which the reagent reacts is dependent upon the rate of ionization of the compound, which is determined by the medium and is largely independent of the nature and concentration of the groups with which reaction subsequently takes place. On the other hand in the s N 2 mechanism reaction only occurs on the approach of the reacting group. This difference may be represented quite simply thus: SN1 R X + R+ s N 2 ARX
+
+ A.***R*..*X XR+ + A-+ RA + RA 4-X-
The rate of combination of an agent which reacts by the second mechanism is entirely dependent upon the concentration of the group (A) which displaces group (X). This is probably of some importance in the reaction of the compounds with biological systems where the concentration of reacting centers will vary from site t o site. In particular, it will result in an increased amount of an agent which reacts by an s N 2 mechanism being involved in regions where there is a high concentration of groups capable of reacting. Another factor which will influence the amount of reaction at different sites is the hydrogen ion concentration of the environment. This is especially important in the case of those epoxides and ethyleneimines which have a high coefficient for the acid catalyzed reaction (see pp. 434,436). As has been stated, in the case of these compounds reaction is much more rapid under conditions of lower pH. This effect is obviously of importance when a compound is to be administered orally for clearly a compound with a high acid coefficient will be rapidly decomposed in the stomach where the pH is about 1. There is a possibility that this influence of pH upon reactivity may be exploited to obtain more selective action on tumor tissue. Stevens et al. (1950) have pointed out that cancerous tissue is often more acid than normal tissue on account of an increased production of lactic acid, and this acidity may be accentuated by injection of glucose, when the tumor fluids may reach pH 6.5 as compared with a pH of 7.5 in most normal tissues. They have demonstrated a selective deposition of sulfapyrazine, which is less soluble under acid conditions, in the tumor area in rats bearing the transplanted Walker carcinoma. It would seem worth while examining the possibility of obtaining an increased cytotoxic action in malignant growths by similar methods using appropriate epoxides or ethyleneimines. The definitions of the s N 1 and::&:! :mechanisms given above indicate that the group R in each type of compound has a tendency t o combine with a nucleophilic (electron-rich) center. In the case of the s N 1 mechanism this is due to the effective production of a carbonium ion and
CYTOTOXIC ALKYLATINO AGENTS
439
in the SN2prccess it is due to the displacement of a very weak nucleophilic group, e.g., the methanesulfonate ion, from the radical R by a more powerful nucleophilic group. All the compounds now under consideration can therefore be regarded as electrophilic reagents which will react with nucleophilic centers in biological systems. It now becomes of interest to decide which functional groups in such systems are nucleophilic under physiological conditions. The main types of nucleophilic groups likely to be encountered are ionized acid groups such as the anions of organic and inorganic acids (e.g., R C 0 2 - and
\
-P.O-),
/
ionized forms of hydroxy compounds (R-0-) and ionized
sulfhydryl groups (RaS-). Amines and thioethers are also nucleophilic on account of the unshared electron pair on the nitrogen or sulfur atom. Undissociated acid groups and the cationic form of amines (RNHd+) are not nucleophilic. When these facts are realized, it is possible to decide which functional groups are nucleophilic a t pH 7.5. Quite obviously the dissociation constant of a particular group will decide whether in the case of an acid it is present in the reactive anionic form or in the case of a base in the reactive undissociated form (e.g., as RNH2). It is convenient to illustrate this point by considering the groups present in two important biological materials, the proteins and nucleic acids. Table X shows the dissociation constants of the main groups likely to be involved, together with an estimate of the proportion of these groups which will be in the reactive form a t pH 7.5. A high percentage of the carboxyl groups in proteins and the phosphoryl groups in nucleic acids is reactive. The basic groups with which reaction is likely to occur are the terminal a-amino groups of peptide chains and the histidine amino group and also the aromatic type amino groups in the guanine, adenine, and cytosine components of nucleic acids. Relatively small proportions of phenolic, aliphatic hydroxyl, and thiol groups will be in the reactive anionic form a t pH 7.5. Some reaction can occur with undissociated water molecules and probably with undissociated alcoholic and phenolic groups but this will only be of importance if high concentrations of these groups are present in the system. Even in their reactive forms the various groups exhibit appreciable differences in their ability to react with an electrophilic center. This ability to react, which is a measure of the nucleophilic capacity, can be regarded as an “affinity” factor; it is this factor that is measured by Ogston (194813) in his determination of competition factors. It will be remembered (p. 403) that the competition factor of, say, an anion (FA)is defined as kA/lc,[HzO], this being a comparison of the rates a t which the
440
W. C. J. ROSS TABLE X pR. Values of Acidic and Basic Groups in Proteins and in Nucleic Acids
Group -
Proteins a-Carboxyl Carboxyl (aspartyl) Carboxyl (glutamyl) Phenolic hydroxyl (tyrosine) Sulfhydryl (cysteine) Imidazolium (histidine) a-Ammonium (terminal) +Ammonium (lysine) Guanidinium (arginine) Nucleic acids Primary phosphoryl Secondary phosphoryl Aromatic hydroxyl (uracil and thymine) (guanine) Sugar hydroxyl Aromatic amino (cytosine) (adenine) (guanine)
PKO
Fraction of the Groups in the Reactive Form at pH = 7.5 (f)
3.0-3.2 3 .O-4.7 4.4 10.4 10.8 5.6-7.O 7.6-8.4 9.4-10.6 11.6-12.6
0.9999 0.9999-0.999 0.999 0.001 0.0006 0.99-0.76 0.44-0.11 0.01-0.001 0.0001-0.00001
2.0 6.0 10.2 10.1 13 4.2 3.7 2.3
0.9999 0.96 0.002 0.0025 0.00001 0.999 0.999 0,9999
anion and a reference substance, water, react with the agent. Another factor which is important in deciding the extent to which a particular group will react is the concentration of that group in the system. This is especially important when the cytotoxic compound is administered to an animal in the relatively low dosage required t o inhibit tumor growth or to induce tumor formation, for under such conditions a small amount of reagent will be available to react with a large excess of reacting groups. The extent to which any particular type of group can compete for reaction with the electrophilic reagent a t pH 7.5 will be proportional to FA X f X e (where FA is the competition factor of the reactive form of A-this will be the anion in the case of an acid or a thiol and the free base in the case of an amine; f is the fraction of the groups which is in the reactive form a t pH 7.5; and c is the concentration of the particular group in the system). It may now be realized why the results obtained by allowing a large excess of radiomimetic agent to react with various compounds under conditions more acid or more alkaline than physiological pH can be misleading as an indication of the type of reaction likely to be encountered in vivo. In the first place, if a large excess of reagent is employed all
CYTOTOXIC ALKYLATING AGENTS
44 1
potential reacting groups will eventually combine for even in the case of groups with a low value off the combination of the reactive form will lead to the production of more of this form and ultimately the reaction will go to completion. Groups with high values of FA will react rapidly, but when these have combined the agent can then attack groups of lower nucleophilic capacity. Only if the reaction times are kept short can the information obtained by using large excesses of reagent be of value. Secondly, the pH of the solution has often been kept well on the alkaline side, particularly in reactions with the mustards where aciility would otherwise have developed. This procedure quite obviously leads to a far higher concentration of the reactive forms of amines and thiols; indeed, the higher reactivity of amines under alkaline conditions has often been observed (pp. 408, 417, 434). The relative inefficiency of the mustard gas type of compound as an inhibitor of the so-called -SH group of enzymes (Dixon and Needham, 1946) has surprised workers who have demonstrated a very high affinity of the thiol group for mustard gas used in large excess and/or in alkaline solution. Despite the very high value of the competition factor for the ionized thiol group, this high degree of combination would not be expected if limited amounts of reagent were available and the medium was kept at pH 7.5. The reaction of electrophilic reagents with mixed systems under acid conditions has not been studied to any extent, but as Olcott and Fraenkel-Conrat (1947) have pointed out under such conditions the thio ether linkage in proteins is one of the few groups likely to be involved. Consideration of the factors mentioned above suggests that the groups in proteins and nucleic acids most likely to combine with electrophilic reagents in viva are the acidic groups and certain amino groups of favorable dissociation constant. This conclusion appears to be supported by the work of Carpenter et al. (1948; pp. 409, 411) on the reaction of butyl 2-chloroethylsulfide with tobacco mosaic virus. Since compounds, such as diacid chlorides, diisocyanates, and diiodoacetyl compounds, which readily react in aqueous solutions with amino groups and thiols but not to any extent with acid groups, do not appear to be radiomimetic, obe is led t o the conclusion that the various cytotoxic agents owe their especial properties t o an ability to react with ionized acid groups under mild conditions. It is not, of course, suggested that these acids groups are necessarily those in nucleoproteins, the characteristic biological effects may be due to reaction with centers in essential enzymes, vitamins, hormones, or other growth factors. In view of the possible formation of esters by the reaction of the various cytotoxic agents with ionized acid groups, it was clearly of interest to study the relative stability of the ester linkages formed by
442
W. C. J. ROSS
different compounds. Davis and Ross (1950) prepared the acetate and benzoate corresponding t o several chloroethyl amines and sulfides and also certain epoxides and showed that these esters were more readily hydrolyzed under alkaline conditions than simple aliphatic esters. The rates of hydrolysis in alkaline solution do not necessarily bear any direct relationship to the rates at physiological pH; for example, the lability of hydroxy esters derived from epoxides is probably due to the assumption of the ionic form: CHaCHa0.CO.R
/’
-0
which will be present only in more strongly alkaline solution. Nevertheless it was thought that any acidic groups which reacted with the various agents would be only temporarily blocked. This view was expressed earlier by Peters (1947) who considered that regeneration of cell constituents attacked by mustard gas a t acidic groups was quite conceivable. Boursnell e2 al. (1946b) obtained further evidence for this statement when they showed that much of the mustard gas fixed in tissues had disappeared twelve hours after treatment. Until very recently it appeared necessary that there should be at least two alkylating groups in the molecule for a compound t o possess activity as a cytotoxic agent (compare Haddow et al., 1948). This led Goldacre et al. (1949) t o suggest that the two groups were required to permit the molecule to react at two distinct points, either on a single surface or upon two separate surfaces. This hypothesis was extended to suggest that a cross linkage might be formed between sister chromatids and that this would result in breakage of these chromatids during the process of cell division. The cytological picture of chromosome fragmentation and bridge formation was considered consistent with this theory of action. Rose et al. (1950) recognized that the mustards could function as cross-linking agents, but they did not consider that this property was a sufficient condition for biological activity. They suggest that only substances which can polymerize to give linear structures with reactive alkylating groups spaced along the axes are effective. Models showed that in such a polymer, e.g., (LVI) derived from a nitrogen mustard, the chloroethyl side chains appear on alternate sides of the chain a t distances apart which are multiples of 3.7 Angstrom units. Since this distance corresponds to the interpurine or interpyrimidine spacing in nucleic acids or to the interamino acid spacing in extended polypeptides, they considered that mitotic abnormalities might result from an attachment of the polymer at several points to the nucleic acid or protein chains of the chromatids.
443
CYTOTOXIC ALKYLATING AGENTS
Davis and Ross (1950) have pointed out that the formation of such polymers from aromatic nitrogen mustards is rather unlikely and that the polymer, if formed, would have unreactive side chains since the activating effect of the uncharged nitrogen atom on the halogen atom no longer operated. An alternative suggestion by Rose (1950) and his co-workers is more feasible, this is that one side chain of a mustard reacts as an alkylating agent with a center on the protein or nucleic acid chain and that subsequent polymerization occurs between molecules already in combination with this chain-they describe this as a type of “zipfastener ’’
n
(LVU
While the cross-linkage hypothesis was under consideration it became of interest t o decide to what extent the alkylating centers of a molecule could be separated while still retaining biological activity. It was known (see Kon and Roberts, 1950) that it was not essential for two chloroethyl groups to be attached to the same nitrogen atom, and these workers prepared a series of compounds of the general formula (LVII).
c:
Ar.N-(CH&-N.Ar ( H&,.Hal.
A
( H&,.Hal.
OlVII)
They concluded that biological activity is maintained when m = 2 or 3, but gradually falls off as the distance between t,he nitrogen atoms is further increased. Lengthening the halogenoalkyl side chain (e.g., to n = 3) leads to a complete loss of activity. Similar results have been obtained in a series of diepoxides of general formula (LVIII), prepared by Everett and Kon (1950).
(LVIII)
444
W. C. J. ROSS
Preliminary results indicate that in this series activity as an inhibitor of the growth of the transplanted Walker rat carcinoma falls off as n is increased (Haddow, 1950). It is of interest to note that in a series of chloroethyl sulfides of general formula (LIX) ClCHsCHsS(CH~)nSCHs*CHaCl (LIX)
Gasson et al. (1948) found that the vesicant power increased as n increased from 0-3 but then decreased as n increased from 4-10. Although it is diflicult t o draw any definite conclusions from these effects of chain lengthening since such ti modification will alter not only the distance between the alkylating centers but also certain physical properties of the molecule (e.g., lipoid solubility), the results do suggest that increasing this distance beyond a limiting value will reduce biological activity. Later work indicated (Ross, 1950c) that although two reactive centers in the molecule were required for activity these centers need not be of equal reactivity; for example, the effective 1,2-3,4diepoxy-2-methylbutane (LX) and N-2-chloroethyl-N-3-chloropropylaniline(LXII) (Kon and Roberts, 1950) both contained side chains of unequal reactivity.
CHiCHsCl CHsCHiCHiCl &XI11
CH&HsCH,Cl CHzCHnCHnCl (LXIII)
The compounds (LXI) and (LXIII) which contain two of the groups of lower reactivity in (LX) and (LXII) are not biologically active. These results suggested that only one group of high chemical reactivity was required-this would anchor the molecule to a receptor, facilitating the reaction of the second group with another center. The most recent studies have shown that certain monofunctional alkylating agents are radiomimetic in the sense that they will induce mutations, produce chromosome abnormalities, and inhibit the growth of transplanted tumors. Ethyleneimine and 2,4-dimethoxy-6-ethyleneimino-1,3,5-triazine (LXIV) cause cytological abnormalities in the dividing myelocytes in the bone marrow and elicit radiation-like damage in
CYTOTOXIC ALKYLATING AGENTS
445
the cells of the onion root tip. The triadne (LXIV) inhibits the growth of the mouse sarcoma 180. The monoepoxide, glycidol (LXV), evokes nuclear lesions in tissue cultures of mammalian cells (Philips, 1950; Biesele et al., 1950).
(LXIV)
(LXV)
Auerbach and Moser (1950) have shown that 2-chloroethyl-2-hydroxyethyl sulfide and butyl-2-chloroethyl sulfide are mutagenic, inducing sex-linked lethals in Drosophila. Stevens and Mylroie (1950)demonstrated the mutagenic action of butyl-2-chloroethyl sulfide, phenyl-2chloroethyl sulfide and diethyl-2-chloroethylamine on deficient mutant strains of Neurospora crassa, the effectiveness of the last-named compound on the same material has also been established by Jensen, Kirk, and Westergaard (1950). Loveless and Ross (1950) report that ethyl-2chloroethyl sulfide, dimethyl-2-chloroethylamine,ethyleneimine, ethylene oxide, and dimethyl sulfate produce chromosome breakage in plant material. Ethylene oxide has also been shown by Rapoport (1948)and by Bird (1950)to be mutagenic for Drosophila and the mutagenic activity of ethyleneimine was reported by Rapoport (1947a).
CLXVI)
N-(2,4-Dinitrophenyl)-ethyleneimine(LXVI) inhibits the growth of the transplanted Walker rat carcinoma (Rose, 1950; Haddow, 1950). It is of interest to recall that Butler and Smith (1950)found that ethyl-2chloroethyl sulfide was able to degrade nucleic acid in a manner similar to that of x-irradiation. It is thus now apparent that polyfunctional activity is not essential for the production of radiation-like effects in dividing cells. It is equally unnecessary t o postulate a cross linking or a polymerization mechanism for the production of chromosome aberrations. Nevertheless, it would still appear that difunctional compounds are more effective than the corresponding monofunctional agent and these mechanisms may account for the higher activity. Philips (1950)points out that a monoethyleneimine derivative has to be administered at 50 to 100 times the dose level
446
W. C. J. ROSS
of a polyethyleneimine in order to produce a comparable effect on hematopoietic organs or tumor tissue. Loveless (1950)has found that effects in plant material are elicited only by concentrations of monofunctional compounds, chloroethyl amines and sulfides and also epoxides, which are some fifty times that required of bifunctional analogues. The failure of earlier workers to detect cytotoxic effects of monofunctional compounds on tumor tissue in vivo may well have been that higher concentrations of such compounds would have been required than could be achieved in the body. Another point is that the monofunctional nitrogen mustard analogue, dimethyl-2-chloroethylamine,will not be particularly reactive toward nucleophilic centers at pH 7.5. This complication arises because the compound is a relatively strong base and at pH 7.5 about 90% of the amine will be present as an unreactive ammonium cation and also it will have a greater tendency to react to give a quaternary salt (p. 414) (Davis et al., 1950). It was the lack of biological activity of this amine that led earlier workers to believe that two reactive side chains were essential. Philips (1950) and Biesele et al. (1950) believe that cytotoxic action is dependent upon the presence in the compound, or its active intermediate, of an unstable three-membered heterocyclic ring system, but Loveless and Ross (1950)consider that the ability to react with nucleophilic centers effectively through a carbonium ion mechanism is a more general feature of this group of radiomimetic agents. The ability to react through a nucleophilic displacement on a carbon atom appears to cover the cases of the methane sulfonyl compound (p. 437) and dimethyl sulfate which cannot form three-membered heterocyclic ring systems. Although in the light of the later work it does not appear essential to have two or more functional groups in the molecule of a radiomimetic compound, one may conclude with Philips that duplication of reactive centers does increase the selective action of the compound toward proliferating cells. The outstanding feature of the group of cytotoxic agents discussed in this review is their ability to function as alkylating agents under the mild conditions arising in living tissues. REFERENCES Alexander, P. 1951. Personal communication. Alexander, P., and Fox, M. 1951. Personal communication. Auerbach, C. 1949. Heriditus, Suppl. Vol., 12847. Auerbach, C., and Moser, H. 1950. Nature 166, 1019-20. Baddeley, G., and Bennett, G. M. 1933. J. Chem. Sac. 261-68. Ball, E. G., Doering, W. E., and Linetead, R. P. 1942. OSRD Formal Report No. 1094, December 9. Banks, T. E., Boursnell, J. C., Francis, G . E., Hopwood, F. L., and Wormall, A. 1946. Biochem. J. 40, 745-56.
CYTOTOXIC ALKYLATING AGENTS
447
Barron, E. S. G., Bartlett, G. R., and Miller, Z. B. 1948. J . Ezptl. Med. 87, 489501. Bartlett, P. D., Davis, J. W., Ross, S. D., and Swain, C. G. 1947a. J . Am. Chem. SOC.69, 2977-83. Bartlett, P. D., Ross, S. D., and Swain, C. G. 1947b. J . Am. Chem. SOC.60, 2971-77. Bartlett, P. D., and Small, G. 1950. J . Am. Chem. SOC.72, 48674869. Bartlett, P. D., and Swain, C. G. 1949. J . Am. Chem. SOC.71, 1406-15. Berenblum, I., and Schoental, R. 1947. Nature 166, 727-29. Biesele, J. J., Philips, F. S., Thiersch, J. B., Burchenal, J. H., Buckley, S. M., and Stock, C. C. 1950. Nature 166, 1112-13. Bird, M. 1950. Paper read before the Genetical Society Symposia on Biochemical Genetics, London. Boursnell, J. C. 1948. Biochem. SOC.Symposia 2, 8-15. Boursnell, J. C., Cohen, J. A., Dixon, M., Francis, G. E., Greville, G. D., Needham, D. M., and Wormall, A. 1946b. Biochem. J . 40, 756-64. Boursnell, J. C., Francis, G. E., and Wormall, A. 1946a. Biochem. J . 40, 73742. Boyd, D. R., and Marle, E. R. 1914. J . Chem. SOC.106, 2117-39. Boyland, E. 1948. Biochem. SOC.Symposia 2, 61-70. Branch, G . E. K., and Calvin, M. 1941. The Theory of Organic Chemistry, Prentice Hall Inc., New York, p. 404. Braa, G. I., and Skorodumov, V. A. 1947. Compt. rend. acad. sci. U.R.S.S. 66, 315-17. Bronsted, J. N., Kilpatrick, Mary, and Kilpatrick, M. 1929. J . Am. Chem. SOC. 61,428-61. Buckley, S. M., Stock, C. C., Crossley, M. I,., and Rhoads, C. P. 1950. Cancer Research 10, 207-08. Burchenal, J. H., Crossley, M. L., Stock, C. C., and Rhoads, C. P. 1950. Arch. Biochem. 26, 321-23. Butler, J. A. V. 1950. Personal communication. Butler, J. A. V., and Conway, B. E. 1950. J. Chem. Soe. 3418-21. Butler, J. A. V., Gilbert, L. A., and Smith, K. A. 1950. Nature 166, 714-15. Butler, J. A. V., and Smith, K. A. 1950. J . Chem. SOC.3411-18. Carpenter, F. H., Wood, J. L., Stevens, C. M., and du Vigneaud, V. 1948. J . Am. Chem. SOC.70,2551-53. Cashmore, A. E., and McCombie, H. 1923. J . Chem. SOC.123, 2884-90. Chanutin, A., and Gjessing, E. C. 1946. Cancer Research 6, 599-601. Cohen, B., Van Artsdalen, E. R., and Harris, J. 1948. J . Am. Chem. SOC.70, 281-85. Conway, B. E., Gilbert, L., and Butler, J. A. V. 1950. J . Chem. SOC.3421-25. Culvenor, C. C. J., Davies, W., and Heath, N. S. 1949. J . Chem. SOC.278-82. Davies, W. 1920. J. Chem. SOC.117, 297-308. Davis, S. B., and Ross, W. F. 1947. J . Am. Chem. SOC.60, 1177-81. Davis, W. 1951. Ph.D. thesis, University of London. Davis, W., Everett, J. L., and Ross, W. C. J. 1950. J . Chem. SOC.1331-37. Davis, W., and Ross, W. C. J. 1949. J . Chem. SOC.2831-34. Davis, W., and Ross, W. C. J. 1950. J . Chem. SOC.3056-62. Deuel, H. 1947. Helu. Chim. Acta 30, 1523-34. Dixon, M., and Needham, D. M. 1946. Nature 168,432-38. Dustin, P., Jr. 1947. Nature 169, 794-97. Elmore, D. T., Gulland, J. M., Jordan, D. O., Taylor, H. F. W. 1948. Biochem. J . 42,308-16.
448
W. C. J. ROSS
Everett, J. L., and Kon, G. A. R. 1950. J. Chem. SOC.3131-35. Everett, J. L., and Ross, W. C. J. 1949. J. Chem. SOC.1972-83. Fleming, D. S.,Moore, A. M., and Butler, G. C. 1949. Biochem. J . 46, 646-52. Ford, C. E. 1948. Proc. 8th Intern. Congr. Genet. p. 570. Fraenkel-Conrat, H. 1944. J. Biol. Chem. 164,227-38. Fruton, J. A,, and Bergmann, M. 1946. J . Org. Chem. 11, 643-49. Fruton, J. S., Stein, W. H., and Bergmann, M. 1946a. J . Org. Chem. 11, 569-70. Fruton, J. S., Stein, W. H., Stahmann, M. A,, and Golumbia, C. 1946b. J. Org. Chem. 11, 571-80. Gasson, E. J., McCombie, H., Williams, A. H., Woodward, F. N. 1948. J. Chem. SOC.4446. Gjessing, E. C., and Chanutin, A. 1946. Cancer Research 593-98. Goldacre, R. J., Loveless, A., and Ross, W. C. J. 1949. Nature 168, 667-69. Golumbic, C., Fruton, J. S., and Bergmann, M. 1946a. J . Org. Chem. 11, 51831. Golumbic, C., Stahmann, M. A., and Bergmann, M. 1946b. J . O r g . Chem. 11, 550-58. Grunwald, E., and Winstein, 8. 1948. J. Am. Chem. SOC.70,84146. Gurin, S., Delluva, A. M., and Crandall, D. I. 1947. J. Org. Chem. 12, 608-11. Haddow, A. 1950. Personal communication. Haddow, A., Kon, G. A. R., and Ross, W. C. J. 1948. Nature 162,824-25. Hammett, L. P. 1940. Physical Organic Chemistry, McGraw Hill Book Co., New York, pp. 301-03. Hanby, W. E., Hartley, G. S., Powell, E. O., and Rydon, H. N. 1947. J. Chem. SOC. 519-27. Herriott, R. M. 1947. Advances in Protein Chem. 3, 170-225. Herriott, R. M., Anson, M. L., and Northrup, J. M. 1946. J . Gen. Physiol. 80, 185-210. Hirst, J. 1950. Ph.D. thesis, University of London. Hughes, E. D. 1941. ! h n 8 .Furaday SOC.37,603-32. Jenscn, K. A., Kirk, I., and Westergaard, M. 1950. Nature 166, 1020. Kadesch, R. G. 1946. J. Am. Chem. Soc. 68,41-46. Kerwin, J. F., Ullyot, G. E., Fumn, R. C., and Zirkle, C. L. 1947. J . Am. Chem. SOC.68, 2961-65. Kiprianov, A. 1926. Ukrain. Khem. Zhur. 2, 23649. Knorr, L. 1899. Ber. 82, 729-32. Kon, G. A. R., and Roberts, J. J. 1950. J . Chem. SOC.978-82. Kopac, M. J. 1947. Approaches to Tumor Chemotherapy. American Association for the Advancement of Science, Washington, D.C., pp. 27-57. Lawson, W. E., and Reid, E. E. 1925. J . Am. Chem. Boc. 47,2821-36. Le Maistre, J. W., and Seymour, R. B. 1948. J . Org. Chem. 18, 782-85. Lewis, M.R., and Crossley, M. L. 1950. Arch. Biochem. 26, 319-20. Lichtenstein, H. J., and Twigg, G. H. 1948. Trans. Furaday. SOC.44, 905-09. Lord, D. D. 1946. Ph.D. thesis, University of London. Loveless, A. 1950. Personal communication. Loveless, A., and Revell, 8. 1949. Nature 164, 938-44. Loveless, A., and Revell, 8. 1950. Personal communication. Loveless, A., and Ross, W. C. J. 1950. Nature 166, 1113-14. Matthews, W. B. 1950. Lancet 268,896-99. Moore, S.,Stein, W. H., and Fruton, J. A. 1946. J . Org. Chem. 11, 675-80. Needham, D. M. 1948. Biochem. SOC.Sgmposio 2, 16-27.
CYTOTOXIC ALKYLATING AGENTS
449
Neukom, H. 1949. Mitteilung aus dem Agrikulturchemischen Institut E. T. H., Zurich. Ogston, A. G. 1941. Report to the Ministry of Supply (London) by Peters, No. 34. Ogston, A. G. 1948a. Biochem. SOC.Symposia 2, 2-7. Ogston, A. G. 1948b. Trans. Faraday SOC.44, 45-52. Ok&E,A. 1934. Chem. Listy 28, 227. Olcott, H. A., and Fraenkel-Conrat, H. 1947. Chem. Reus. 41, 151-97. Peters, R. A. 1947. Nature 169, 149-51. Peters, R. A., and Wakelin, R. W. 1947. Biochem. J. 41, 550-55. Philips, F. S. 1950. J. Pharmacol. Exptl. Therap. 99, 281-323. Pirie, A. 1947. Biochem. J. 41, 185-90. Prelog, V., and Stephan, V. 1935. Coll. Czech. Chem. Comm. 7 , 93-102. Preston, J. M. 1949. Fibre Science, The Textile Institute, Manchester, p. 266. Price, C. C., and Wakefield, L. D. 1947. J . Org. Chem. 12, 232-37. Rapoport, I. A. 1947a. Bull. Exptl. Biol. Med. U.S.S.R. 23, 3. Rapoport, I. A. 1947b. J. Gen. Biol. (U.S.S.R.) 8, 359-79. Rapoport, I. A. 1948. Doklady Akad. Nauk. S.S.S.R. 60, 469-72. Rose, F. L. 1950. Personal communication. Rose, F. L., Hendry, J. A., Walpole, A. L. 1950. Nature 166, 993-96. Ross, S. D. 1947. J. Am. Chem. SOC.69, 2982-83. Ross, W. C. J. 1949a. J. Chem. SOC.183-91. Ross, W. C. J. 1949b. J. Chem. SOC.2589-96. Ross, W. C. J. 1949c. J. Chem. Soc. 2824-31. Ross, W. C. J. 1950a. J. Chem. SOC.815-18. Ross, W. C. J. 1950b. Nature 166, 808-09. Ross, W. C. J. 1950c. J . Chem. SOC.2257-72. Ross, W. C. J. 1950d. Unpublished work. Smith, L., Mattson, S., and Anderson, S. 1946. Kgl. Fysiograf. Sallskap. Lund, Handl. 42, No. 7, 1-18. Speakman, J. B. 1948. Proceedings of the Swedish Institute for Textile Research, No. 7. Stahmann, M. A., Golumbic, C., Stein, W. H., and Fruton, J. S. 1946. J. Org. C h m . 11, 719-35. Stein, W. H., and Fruton, J. S. 1946. J . Org. Chem. 11, 686-91. Stein, W. H., Fruton, J. S., and Bergmann, M. 1946b. J . Org. Chem. 11, 692-703. Stein, W. H., and Moore, S. 1946. J . Org. Chem. 11, 681-85. Stein, W. H., Moore, S., and Bergmann, M. 1946a. J. Org. Chem. 11, 664-74. Stevens, C. D., Quinlin, P. M., Meinken, M. A., and Kock, A. M. 1950. Science 112, 661. Stevens, C. M., McKennis, H., and du Vigneaud, V. 1948a. J. Am. Chem. SOC.70, 2556-59. Stevens, C. M., and Mylroie, A, 1950. Nature 166, 1019. Stevens, C. M., Wood, J. L., Rachele, J. R., and du Vigneaud, V. 194813. J, Am. C h m . SOC.70, 2554-56. Sugiura, K., and Stock, C. C. 1950. Cancer Research 10, 244. Timmis, G. M. 1949. Ann. Rept. Brit. Emp. Cancer Camp. No. 27, p. 43. Timmis, G. M. 1950. Personal communication. Twigg, G. H. 1950. Personal communication. Wiggins, L. F. 1946. J. Chem. SOC.384-88. Wiggins, L. F., and Wood, D. J. C. 1950. J. Chem. SOC.1566-75. Wood, D. J. C., and Wiggins, L. F. 1949 Nature 164, 402-03.
This Page Intentionally Left Blank
Nutrition in Relation to Cancer ALBERT TANNENBAUM AND HERBERT SILVERSTONE Department of Cancer Research, Medical Research Institute, Michael Reese Hospital, Chicago, Illinois CONTENTS Introduction I. Some General Considerations 11. Genesis of Tumors A. Influence of Caloric Intake 1. General Consequences of Calorie Restriction 2. Factors That Modify Extent of Inhibition a. Degree of Restriction b. Composition of Restricted Diet c. Potency of Carcinogenic Stimulus d. Kind of Tumor 3. Mode of Action of Caloric Restriction a. Duration of Restriction b. Stage of Carcinogenesis Where Action Occurs c. Significance of Body Weight d. Hormonal Factors e. Mitotic Activity of Tissue f. Generality of Caloric Influence B. Influence of Proportions of Dietary Components 1. Fat a. General Effects on Tumor Genesis b. Factors that Modify the Fat Effect c. Mode of Action of Dietary Fat 2. Protein a. Proportions of Protein Supporting Normal Body Weight b. Deficient Proportions of Protein 3. Vitamins a. Vitamin Deficiencies b. Vitamin Levels and Aeo Dye Liver Tumors c. Tumors Induced by Vitamin Deficiency d. Effects Independent of Calorie Intake and Body Weight 4. Minerals 111. Growth of Tumors 1. Caloric Intake 2. Fat 3. Protein a. Variations in Dietary Protein Only 45 1
Page 452 453 455 456 456 458 458 460 46 1 462 462 462 463 464 465 466 466 466 467 468 469 470 472 472 474 476 477 477 478 479 480 481 481 482 482 482
452
ALBERT TANNENBAUM AND HERBERT SILVERSTONE
b. Mode of Action of Protein Restriction.. . . . . . . . . . . . . . . . . . . . . . . . . . . c. Protein Restriction Combined with Other Treatment. . . . . . . . . . . . . . . .............................................. 4. Vitamins.. . . . . IV. Nutritional State and Cancer in Man.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1. Body Weight and Cancer Incidence.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2. Dietary Deficiencies and Cancer Production.. . . . . . . . . . . . . . . . . . . . . . . . a. Thyroid Cancer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . b. Pharyngeal Cancer.. . . . . . . . . ... ... ........ c. Liver Cancer.. . . . . . . . . . . . . . ....................... V. Conclusions and Commentary.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1. Summary of Present Knowledge., . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Page 484 484 485 487 488 489 489 490 490 491 491
b. Growth of Tumors in Animals.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . c. Cancer in Man.
493
4. Future Developments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
496 497
References.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
INTRODUCTION When living cells are subjected to carcinogenic influences they may undergo changes that finally result in a growing neoplasm. The energy and substance for the development of the first cancer cells are derived principally from the animal; the new cell type increases in number by assimilating nutrients from the host. It may be expected then that the diet and the nutritional state of the host influence the formation and growth of tumors. The existence, nature, and limits of the influence are the topic of this review. Despite gaps in our knowledge, and facts that do not easily fit into a simple viewpoint, sufficient data are available for a systematic approach. Inasmuch as this is the first volume of Advances i n Cancer Research, a relatively complete and possibly historical review of the relation of nutrition to cancer is desirable. As it happens, however, the main advances have been achieved during the past two decades, and we will center the discussion around the work of this period. It has been necessary to leave out many isolated findings; even the subjects taken up could not be treated exhaustively in discussion or bibliography. Perhaps these omissions will be corrected in reviews of smaller segments of the field in future volumes of Advances. The reader interested in other viewpoints and in a more extensive bibliography will find the following reviews and books helpful: Caspari, 1938; Waterman, 1938; Stern and Willheim, 1943; Morris, 1945; Rusch et al., 1945a; Greenstein, 1947; Tannenbaum, 1947; King et al., 1947; and Baumann, 1948. The subject matter is presented in five sections. The first is limited
NUTRITION IN RELATION TO CANCER
453
to some general and methodological considerations which may facilitate an understanding of the problems encountered. The second and third parts deal respectively with the influence of diet on the genesis and the growth of neoplasms; the significance of the level of caloric intake and the proportions of dietary fat, protein, vitamins, and minerals are discussed. Investigations on the action of crude materials and extracts are not included, since exact compositions and effective components are generally not known. The studies suggesting that nutrition plays a role in the formation of cancer in man are reviewed in the fourth section. Finally, an attempt is made to summarize the present knowledge. Liberties are taken in correlating it into a reasonable whole and in pointing out profitable areas for future work in this particular branch of cancer research.
I. SOMEGENERALCONSIDERATIONS The question as to whether and how a particular dietary change affects the origin or growth of a neoplasm immediately calls attention to experimental materials and procedures. These often determine the success or failure of an investigation and should be considered in evaluating the results. If neoplasms were not a large group of somewhat unrelated diseases, and if metabolic interrelationships were not so complex, experimental design and interpretation would be simpler. As it is, however, the complicating factors often weaken or deny the validity of the results of an experiment. Obvious features such as numbers of animals, duration of an experiment, and influence of deaths from other causes need not be discussed. On the other hand, there are a few general aspects of research and interpretation that might well be mentioned here. It is not intended t o supply an exhaustive list but only to take up some points that experience, sometimes unpleasant, has shown to be important. Expressedly or implicitly they provide criteria used in this review in evaluating the reported findings. They also illustrate the difficulties encountered and the possibilities for differences in technics and interpretations among those investigating the influence of diet on the cancer process. (1) Neoplasm. This term encompasses a group of pathologic entities varying in etiology, site, and kind of cell involved. Moreover, the process may be influenced by experimental procedures during any stage, from the initial host-carcinogen relationship to the large growing tumor. A dietary change might have diverse eff ects upon different carcinogens. Inasmuch as cells and tissues are not alike in metabolic pattern and in nutritive [requirements, it is not anticipated that neoplasms, arising from manifold cell types and growing in different sites, would
454
ALBERT TANNENBAUM AND HERBERT SILVERSTONE
be affected to the same extent or even in the same way by a particular alteration in diet. Even such characteristics as rate of growth and degree of malignancy are concerned in the response of the tumor to treatment. Much is gained by utilizing several kinds of tumors in a nutritional study. The dissimilarities as well as the uniform results provide a basis for more secure systematization and generalization. (9) Genesis and growth. The genesis of a neoplasm (also termed origin or formation) refers to the transformation of one or a few normal cells into cancer cells. It is distinct from the subsequent cellular proliferation or growth of the tumor. Not only in nutrition studies but in others as well, an experimental procedure may affect carcinogenesis in one way, tumor growth in another. Consequently, it is advantageous to consider separately the influence of diet on genesis and growth, rather than on the process as a whole. Not only is a more refined and correct interpretation possible, but also the information relevant to genesis may suggest preventive measures while that on growth might be applied to therapeutic procedures. (3) Potency of carcinogenic stimulus. The amount and potency of carcinogen, whether administered or of unknown or endogenous origin, determines the frequency of occurrence, the rate of formation, and possibly the malignancy of the resulting neoplasms. If the potency of the stimulus is of a high order, a moderately effective nutritional alteration may appear to be without influence. That is, a high “steam roller” dose of carcinogen may override the influence of an experimental procedure whose effect is obvious a t more “physiological” levels of carcinogen. Undoubtedly, negative findings have sometimes resulted because there was not a proper balance between carcinogenic potency and the other factors of the experiment. In this respect carcinogens are like other pharmacologic agents; doses must be within a range that permits recognition of modifying influences. (4) Characteristics of the diet. In planning nutritional studies, decisions must be made as to the nature and composition of the control dietand how best to insure that the experimental dietary change is actually and specifically attained. The rations may be made up of relatively purified, known components (synthetic diets) or of a mixture of natural foods (as in commercial diets). There is no special advantage of one over the other, provided the particular dietary modification being studied can be achieved with the same precision. It is generally easier to alter a diet composed of simple, known components; on the other hand, rations made up of natural foods are probably more “ complete.” At any rate, the composition of the rations should be specified, as well as the magnitude of the nutritional alteration and its relation to the diet as a whole.
NUTRITION IN RELATION TO CANCER
455
This permits adequacy to be judged and effects on tumors to be interpreted in terms of caloric content and proportions of components-and facilitates repetition and extension of the studies by other investigators. (5) Diet and the nutritional state, After choosing the experimental rations to be given the animals, there is still the question of what happens “within the body.’’ For it is the amount of food consumed and its effect upon the nutritional state of the host that determine the influence upon tumor formation and tumor growth. The caloric content of the ration as determined in the bomb calorimeter, and its composition as indicated by diet construction or analysis, may or may not represent what is available to the animal. The simplest example is a deficient diet which alters the appetite, resulting in a reduced intake. More subtle deviations occur through changes in digestibility and absorption, specific dynamic action, and synthesis or utilization of essential nutrients, particularly vitamins, by intestinal bacteria. All these factors can result in a nutritional state that in no way resembles that expected of a particular diet. And it is the nutritional state, not the diet, that influencea the neoplastic process! An alteration in diet may produce a distinct effect on a neoplasm. Can it be attributed specifically to the factor being studied or is it due to an indirect, secondary change-reproducible by a variety of measures having a common denominator? For example, some procedures that inhibit tumor formation produce voluntary restriction of food intake and resultant loss of body weight. And we now know that these themselves may strikingly hinder the formation of neoplasms. Perhaps the reduced caloric intake, not the particular experimental procedure, is responsible for the inhibition of tumor formation in these instances. Other situations could be cited where the findings are distinct but the interpretations questionable. It is the purpose here, however, t o emph* size that a dietary change can have many effects on the nutritional state of the host and t o stress the importance of trying to recognize which.of these might be responsible for the observed findings. Obviously, intimate metabolic reactions and mechanisms involved in these processes are not understood a t this time. However, we have advanced from the stage where only the experimental procedure and final result concern US. Between these lies the fertile field of understanding.
11. GENESISOF TUMORS The adequacy of a diet is conventionally described by the quantity and the relative amounts of essential foodstuffs-protein (or amino acids), lipids, carbohydrates, vitamins, and minerals. So far as metabolism is concerned, these are highly interdependent. The total amount of
456
ALBERT TANNENBAUM AND HERBERT SILVERSTONE
food, measured by its digestible caloric value, significantly influences the assimilation and interconversions of the components. Conversely, the proportion of one dietary constituent may modify the utilisation of the others and also the energy available to the animal. Nevertheless, mainly because of the regulation shown by the complex organism in the face of changes in its external environment, including its food supply, it is possible to investigate the influence of alterations in amounts of specific dietary factors. The interpretations, however, must always be qualified by the known and presumed interactions, particularly if large variations are being studied. The convenient dichotomy of quantity and quality suggests division of this section into (1) the influence of caloric intake and (2) the influence of the proportions of dietary components. A. INFLUENCE OF CALORIC INTAKE
There are no objective bases for definition of an optimal or normal caloric consumption. When food intake is unrestricted, mice, rats, domestic animals, and even man may become quite obese. Thus, under certain conditions, ad libitum food consumption may be supernormal. Only in this way have we a suggestion of the influence of high caloric intakes since there are no studies on carcinogenesis dealing with supervoluntary feeding uncomplicated by other treatment. In the main, the investigations on the effects of caloric intake have been concerned with restriction of the experimental animals to between one-half and twothirds of the rations eaten by their controls.
I. General Consequences of Calorie Restriction Caloric intake has been shown to influence the genesis of virtually all the mouse tumors that have been studied. Numerous and extensive investigations have established the fact that among mice chronically restricted in caloric intake the incidence of tumors is decreased and the time a t which the tumors appear is delayed. The impressive inhibitory action of long-term caloric restriction on the formation of a variety of tumors is summarixed in Table I. The neoplasms are of spontaneous origin or induced by known carcinogens, and occur in a variety of organs and tissues. In the mouse, eight tumor types of diverse nature are known to be affected by calorie restriction. Only a few studies have dealt with neoplasms of the rat, and these indicate that the formation of lymphosarcoma (Saxton, 1941) and induced mammary carcinoma (Dunning et al., 1949) is hindered. The reduced tumor incidence is not the consequence of any untoward
457
NUTRITION IN RELATION TO CANCER
TABLE I Results of Investigations on the Inhibitory Effect of Caloric Restriction on the ' Genesis of Tumors in Mice
Type of Tumor
DuraKind of tiont of Dietary Restric- Study tion* (weeks)
Mammary U carcinoma, C F spontaneous C F C C U Hepatoma, spontaneous C
+ +
Lung adenoma, primary
U U U U
U Leukemia, spontaneous C F Leukemia, induced by carcinogen U Skin tumors, induced by U carcinogen C C C Skin tumors, C induced by C UV light
+
Sarcoma, induced by carcinogen
*
- -
U C C C
80 78 96
Formation of Tumors: Control/Restricted InciStrain dencet Mean of (per Time! Mouse cent) (weeks)
References
64 58
dba CsH CsH dba CsH CJI
40/2 67/0 100/20 54/0 58/0 44/0
62
CaH
64/0
96 54 52 70
ABC Swiss Swiss A (males)
52/27 48/8 30/5 50/30
110
Ak
65/10
Tannenbaum, 1940a Visscher et al., 1942 White et al., 1944 Tannenbaum, 1942a Tannenbaum, 1945b - Tannenbaum and Silverstone, 1949c - Tannenbaum and Silverstone, 1949c - Tannenbaum, 1940a - Tannenbaum, 1942a - Tannenbaum, 1942a - Larsen and Heston, 1945 39/62 Saxton et al., 1944
55
dba
96/35
11/34 White et al., 1944
100
77 44 60 46 25 39 39 42 60 38 23
ABC 44/19 Swiss 62/40 dba 65/22 dba 100/71 Rockland 82/18 C 87/7 C 63/24 ABC CWBI dba C
47/15 74/44 16/2 85/70
63/71 41/35/64 74/47/-
28/55 20/27 28/38 18/39
Tannenbaum, 1940a Tannenbaum, 1942a Tannenbaum, 1942a Tannenbaum, 1945a - Boutwell et al., 1949a 33/37 Rusch et al., 1945c 33/33 Rusch et al., 1945c 26/32 24/28 30/15/17
+
Tannenbaum, 1942a Tannenbaum, 1942a Unpublished Rusch et al., 1945b
U underfeeding (all components of diet restricted); C F = only carbohydrate and fat restricted; C only carbohydrate restricted (caloric restriction per 80). t For spontaneous tumors this indicates age of mice at end of experiment; for induoed tumors it La the interval between 1st treatment with carcinogen and end of the experiment. 1Proportion of animals which developed tumors during course of experiment (numbers of miee were sufficient for statistical significance of observed differences). 8 Some of the figures in thia column were eatimated from summarizing tablea or curves in the indicated references.
458
ALBERT TANNENBAUM AND HERBERT SILVERSTONE
factors such as general debility or premature deaths among animals on the low-calorie diets. The restricted mice are active, sleek, and healthy; almost invariably their life span is increased, even when compared with the nontumor controls (Tannenbaum, 1942a;Ball et al., 1947). Indeed, since the classic work of MacCay el al. (1935,1939) it is an accepted fact that moderate degrees of caloric restriction prolong life and retard the onset of senescence. The influence of caloric restriction on the formation of tumors seems to have considerable consistency and generality. As more neoplasms are studied, however, exceptions may be found. Already there are two instances that require consideration. For the hepatic tumor induced in rats by feeding carcinogenic azo dyes, indirect evidence has been interpreted as suggesting that restriction of calories might not be inhibitory (Clayton and Baumann, 1949). An approach with fewer confounding factors is needed to clarify this point. King and associates (1949)have reported that the incidence of adrenal adenomas, arising subsequent to ovariectomy of CsH strain mice, was not reduced by caloric restriction. Since all the ovariectomized mice developed the adrenal tumors, the carcinogenic stimulus must have been of high intensity. It will be demonstrated later that the influence of caloric restriction can be masked by massive doses of carcinogen. If less potent tumorigenic conditions were designed, onset of adenomata in the restricted mice might be delayed, or even the incidence reduced. 2. Factors That Modify Extent of Inhibilion
The actual extent of the calorie effect is dependent on several factors, notably, the kind of tumor, the potency of the carcinogen, the degree of restriction imposed, and the composition of the restricted diet. These influences are additive, and the net influence is related to the experimental conditions. With one type of tumor, a low dose of carcinogen, and a 30 to 50% reduction in caloric intake, formation of neoplasms may be completely blocked. On the other hand, with another tumor, a high dose of carcinogen and only a small reduction in calories, inhibition may not be recognized. a. Degree of restriction. In the initial studies on the influence of caloric intake the restricted animals were given from one-half to twothirds the calories ingested by the full-fed controls. This degree of underfeeding is practicable and allows relatively good health and longevity. Many physiological functions are repressed at this level, however, particularly those concerned with reproduction. To ascertain whether or not less drastic treatment also inhibits the genesis of neoplasms, the influence of caloric intake a t graded levels was determined. These
NUTRITION IN RELATION TO CANCER
459
studies were performed with the induced skin tumor, the spontaneous mammary carcinoma, and the benign hepatoma (Tannenbaum, 1945a, 1945b; Tannenbaum and Silverstone, 1949c; Boutwell et al., 1949a). Carcinogenesis was found to be affected by even small degrees of caloric restriction, and the magnitude of the inhibition was dependent on the extent of the restriction. On an arithmetically scaled graph a complete relationship (covering a 0 to 100 per cent range of tumor incidence) may be pictured as an
i;/[ .-
/Moderate Dose
Low Dose
20
, / /
I
60
70
80
90
100
Mean Daily Caloric I n t a k e
(as percent of a d - l i b i t u r n i n t a k e ) FIG. 1. Idealized relation between degree of caloric restriction and tumor incidence: curves that can be obtained with low, moderate, or high carcinogenic doses.
J-shaped curve with the upper arm longer than the lower, and the s t e e p est slope at the 50 per cent level. Of course under actual experimental conditions either the upper or lower arm of the curve might be missing, depending on whether a small or large carcinogenic stimulus was employed (Fig. 1). The influence of caloric intake on the formation of neoplasms is evidenced not only in comparative incidence, but in rate of appearance as well. I n one experiment the dose of carcinogen was so large that nearly all mice developed tumors; nevertheless, the mean latent period of formation increased with decreasing caloric intake (Fig. 2). The relationship between tumor incidence and degree of calorie restriction seems to be of the type often found in pharmacological studies:
460
ALBERT TANNENBAUM AND HERBERT SILVERSTONE
the per cent of animals with tumors, expressed in probits,* is a straight line function of the logarithm of the caloric intake (Fig. 3). It offers no support to the possibility that there is a critical level or threshold of caloric supply essential for the genesis of tumors. Rather, it suggests that the energy value of the ration is only one of many factors in carcinogenesis.
8.1
I
8.9 9.6 10.3 11.0 11.7 12.8
I
14,3
Mean Daily Caloric Intake FIQ.2. Relationship between the degree of caloric restriction and the mean time of appearance of benzpyrene-inducedneoplasms of the skin. Shown for both carcinomas and total skin tumors (Tannenbaum, 1945a).
b. Composition of restricted diet. Experiments concerned with restriction of food intake have been performed in several ways: (1) The restricted diet was made identical in composition with the control diet, all components being restricted proportionately; (2) the restricted ration contained the same amounts and kinds of protein, vitamins, and minerals as the control, but the amounts of fat and carbohydrate were reduced; (3) the restricted diet was limited only in carbohydrate content. The first procedure has been described as “underfeeding,” the third as ‘ I caloric restriction per se.” Designating a decrease in carbohydrate as caloric restriction represents a simplification. Certainly, as the diet * Probits are the transformationof the percentage incidence to areas of the normal
distribution curve, these areas being measured in standard deviation units.
NUTRITION I N RlLATION TO CANCER
46 1
is diminished in amount by the removal of carbohydrate, there must occur increased diversion of fat and protein to meet energy requirements. Nevertheless, limiting the carbohydrate content offers the best approximation to “caloric restriction per se.” All three modes of limiting caloric intake affect the formation of tumors, but there are quantitative differences. Apparently the nature of the components of the restricted diet modifies the magnitude of the aJ -
g 95-
v, 90.n CI
2
.c1
c aJ
70-
2 50-
2
6 30u I= aJ
U .0
c 1050
a
I-
~~
60
70
80
90
100
Mean Daily Caloric I n t a k e (Log. Scale) ( a s p e r c e n t of a d - l i b i t u r n i n t a k e )
FIQ.3. Linear relationship between tumor incidence on a probit scale and degree of caloric restriction on a logarithmic scale. The straight lines satisfactorily fit the
data, taken from the authors’ experiments: A, spontaneous mammary carcinoma; B, spontaneous hepatoma; C, induced skin tumors. The statistical procedures are standard (Fisher and Yates, 1943).
inhibitory action. For instance, with restricted diets of equicaloric value, high fat rations may almost prevent the low-calorie effect observed with low fat rations (Tannenbaum, 1945b). The influence of specific dietary components is discussed later, but it can be stated now that the composition of the diet, as well as its energy value, plays a role in the carcinogenic process. c. Potency of carcinogenic stimulus. Moderately large doses of cutaneously applied carcinogen may evoke tumors in nearly all mice regardless of the level of caloric intake. Under such conditions caloric restriction does not evidence any restraining effect if only the incidences of animals with tumors are considered. However, the caloric influence
462
ALBERT TANNENBAUM AND HERBERT SILVERSTONE
is still mirrored by both the delayed appearance (increased latent period) and decreased number of lesions per animal (Tannenbaum and Silverstone, 1947). It seems likely that even these signs of inhibition might be obliterated if a sufficiently massive amount of carcinogen were used. The steam-roller action of large doses could be anticipated, as a feature common to pharmacologic experimentation. Along these lines, we regard the findings of White and White (1944a) with the spontaneous mammary carcinoma, and those of King and co-workers (1949) with the postcastrational adrenal adenoma, as instances where a high carcinogenic stimulus tended t o mask the low-calorie effect. d. Kind of tumor. Sufficient data are presented in Table I t o indicate that the inhibitory influence of caloric restriction is less striking for some tumors of the mouse than for others. Without exposition of the actual experimental conditions, it is seen that the formation of sarcomas and skin tumors induced by hydrocarbons is least affected; primary lung tumors and leukemias occupy an intermediate position; and spontaneous mammary carcinomas and benign hepatomas are most inhibited. These differences are the consequence of several factors, probably interrelated. Among them are: the potency of the carcinogenic stimulus and the resultant intrinsic vigor of the tumor (and its genesis) ; the time interval over which this stimulus operates; and the effect of the altered nutritional state on the carcinogen itself, and on the tissue in which the tumor arises. The authors have speculated as to how these factors produce the differential response of various tumor types. The arguments and conjectures are not being detailed, however, since more facts, not discussion, are needed. 3. Mode of Action of Caloric Restriction
The magnitude of carcinogenic inhibition brought about by limitation of caloric intake-and the fact that so many tumor types respond-has compelled interest in the question of mechanism. Some factors and influences that may alter the extent of the caloric effect have already been discussed. There remain, however, certain biological facts and a few hypotheses that bear on the way this simple nutritional alteration operates. Obviously, these are mere beginnings, but they may eventuate in a better understanding of the mechanism not only of caloric restriction, but of carcinogenesis itself. a. Duration of restriction. In most studies concerned with the influence of low-calorie diets, the experimental mice were underfed continuously throughout the period of observation, ranging from a few months in some instances to about two years in others. There are only a few experiments that give information as to the minimum period of restriction
463
NUTRITION I N RELATION TO CANCER
necessary to produce a caloric effect. Clearly, this is not a specified time applicable to every set of experimental conditions and mode of carcinogenesis. Even a few weeks or months may result in some delay in carcinogenesis, and it is probable that the extent of the inhibitory effect is correlated with duration of restriction. That it must be relatively continuous in order to be effective is suggested by an experiment in which twice-weekly fasts for twenty-four hours, with ad libitum feeding between fasts, had no inhibitory influence on the formation of mammary tumors (Tannenbaum and Silverstone, 1950). b. Stage of carcinogenesis where action occurs. In experiments with the spontaneous mammary carcinoma in strain dba mice, it was observed that institution of caloric restriction a t 2, 5, or 9 months of age, respectively, produced relatively the same degree of tumor inhibition (Tannenbaum, 1942a). This suggested that limiting the caloric intake would have an effect if begun at any time before tumors appeared. A study with strain C3H mice, which develop multiple mammary carcinomas, supports this inference. Each mouse was fed ad libitum until its first tumor appeared; subsequent caloric restriction resulted in a striking reduction in the incidence of second and third tumors (unpublished). The above experiments suggested that chronic restriction of calories probably acted, not during carcinogenic stimulation, but in some later part of the cancer process. This was tested in a study in which skin tumors induced by a carcinogenic hydrocarbon were utilized, because the period of treatment with the carcinogen could be better controlled (Tannenbaum, 1944a). Four equivalent groups of mice were employed, and all received the same amount of 3,4-benzpyrene to the skin; administration was discontinued before any tumors appeared. Thus the experiment was made up of two arbitrary intervals: the first, during which carcinogen was applied; and the second, during which tumors appeared. Four different sequences of a high- and a low-calorie ration were fed, as indicated in Table 11. The results were interpreted as showing that caloric restriction might have had a slight influence in the period of TABLE I1 Incidence of Induced Skin Tumors in Relation to Period of Caloric Restriction
Group H H HL L H L L
Diet during Period of Diet during Period Carcinogen Application of Tumor Appearance (10 weeks) (52 weeks) High calorie High calorie Low calorie Low calorie
High calorie Low calorie High calorie Low calorie
Tumor Incidence (per cent) 69 31 55 24
464
ALBERT TANNENBAUM AND HERBERT SILVERSTONE
carcinogen application, but that the major inhibitory action occurred during the period in which tumors were emerging. The concept that carcinogenesis proceeds through several sequential stages has arisen from the investigations of many workers, most prominently Rous and Berenblum. Although there are differences in points of view and nomenclature, it is generally agreed that at least two phases can be distinguished : (a) a stage variously designated as initiation, inception, or latency, in which the carcinogen acts upon normal cells to render them biased toward the cancerous transformation; these initial changes are limited but self-perpetuating; (b) a stage of development or promotion, in which the initiated or biased cells, under favoring conditions, develop into cancer cells. This ends carcinogenesis; the growth of the tumor begins with the active proliferation of the primary cancer cells. The following references are representative : Rous and Kidd, 1941; MacKenzie and ROUS,1941; Berenblum, 1941; Mottram, 1944; Tannenbaum, 194413; Kline and Rusch, 1944; Berenblum and Shubik, 1949; Friedewald and ROUS,1950. It is not within our province t o discuss in detail this important contribution to the understanding of carcinogenesis, but rather to emphasize that the work on nutrition is in agreement with these concepts and has expanded them. Specifically, caloric restriction appears to have its main influence on the developmental stage of carcinogenesis. c. Significance of body weight. In seeking information that might elucidate the mode of action of limited caloric intake, the concomitant phenomena of lower metabolic turnover and reduced growth or body weight were considered. The possible significance of low body weight, in the formation of the spontaneous mammary tumor, was examined in studies in which sodium fluoride or dinitrophenol was incorporated into the diet, or the mice were kept in a cold room a t about 50°F. Compared with the controls, the mice fed sodium fluoride consumed about 10 per cent less food; the animals of the other two groups ate about 10 per cent more than the controls-all that was offered them. All three procedures resulted in a significant retardation in body growth and a striking reduction in tumor incidence (Tannenbaum and Silverstone, 194913). Under similar conditions of food intake, the feeding of either sodium fluoride or dinitrophenol also resulted in reducing body weight and the incidence of spontaneous lung adenomas. It was postulated that it was not the level of caloric intake or total metabolic turnover, but rather the body weight level a t which a balance was struck between caloric intake and utili~ationthat might be a factor in the origin of some tumor types. This possibility was strengthened by the results of studies in which mice were fed thyroid extract (Silverstone
NUTRITION I N RELATION TO CANCER
465
and Tannenbaum, 1949). They consumed 40 to 50 per cent more food, yet weighed about 10 per cent less than the controls. This impressive augmentation of caloric intake and metabolic activity did not produce a large effect upon the formation of skin tumors or sarcomas induced by 3:4 benzpyrene. In these experiments the formation of tumors was more closely related to the average body weights of the groups. That the overall weight of an animal has importance is indicated by another experiment in which mice were intermittently fasted-twice a week for twenty-four hours with ad libitum feeding on the other days. They ate the same amount of food on a per week basis, grew as well, and had the same incidence of spontaneous mammary tumors as the controls (Tannenbaum and Silverstone, 1950). However, there are other observations that do not support the idea that the level of body weight is a major factor in the development of neoplasms. Two hundred and fifty strain CaH female mice were permitted to eat ad libitum of an adequate stock diet. The maximum weights attained by the mice ranged from 22 to 42 g. Yet there was no correlation between body weight and incidence or time of appearance of spontaneous mammary cancer (unpublished). I n other experiments, moreover, feeding of sodium fluoride or dinitrophenol failed to hinder the formation of induced skin tumors and sarcomas, although body weight was markedly reduced (Tannenbaum and Silverstone, 194913). Nevertheless, the weight of evidence available a t this time indicates that the inhibitory action of caloric restriction, and possibly some other experimental procedures, may be correlated with the accompanying limitation of body weight. At present, we prefer to look upon low caloric intake and low body weight as interconnected, without attempting to regard either as being more intimately related to the tumor-hindering action of calorie restriction. It is in this sense that the combined terms are used in the review. It might be well to mention here that there are special situations in which the experimental group weighs less yet develops more tumors than the controls. This can occur where the experimental diet differentially increases both the effective tissue dose and the toxicity of the carcinogen. d. Hormonal factors. In mice and rats given one-half to two-thirds the average ad libitum caloric intake there is a sequence of changes, notably in the ovaries, uteri, and mammae, that simulate those following hypophysectomy (Mulinos and Pomerantz, 1940). This so-called pseudohypophysectomy has been proposed as an explanation of the influence of caloric restriction on the formation of spontaneous mammary cancer. The action is presumed to be dual, involving a decrease in estrogen production and a diminished response of mammary tissue t o
466
ALBERT TANNENBAUM AND HERBERT SILVERSTONE
estrogen (Huseby et al., 1945). No attempt has been made to extend this hypothesis to other tumor types. I n contrast to this suggestion of pituitary insufficiency-arereports that caloric restriction may actually augment the output of adrenocorticotropic hormone (Boutwell et al., 1948; Sayers, 1950). This is indicated by a relative increase in adrenal size, a marked involution of lymphoid tissue (notably thymus), and an apparent increase in glyconeogenesis. Boutwell and associates have suggested that perhaps the changes in adrenal function may explain the tumor-inhibiting action of caloric restriction. e. Mitotic activity of tissue. Bullough (1950) has postulated that the developmental stage of carcinogenesis is influenced by the mean mitotic activity of the tissue; by inference this includes mitosis of the latent cancer cells. It was observed, for example, that mitotic activity was definitely inhibited by caloric restriction, in proportion to the degree of the limitation (Bullough and Eisa, 1950). These and other experiments led him to propound that a limiting factor in cell division is the amount of carbohydrate and carbohydrate intermediates available for the energy requirements of mitosis. The mitotic activity hypothesis is compatible with all that is known with regard to caloric restriction. f. Generality of caloric influence. Chronic caloric restriction results in many changes: in the absolute and relative weights of organs and tissues; in proportions of tissue and body fluid constituents; in mitotic activity; in hormone production; and in metabolism. Presumably there is a restricted supply of nutrients a t the loci of tumorigenesis. Which of these or other alterations are responsible for the influence on carcinogenesis? Whatever the answer, it must account for the generality of the action and its ultimate effectiveness in a wide variety of tissues and cell types. B. INFLUENCE O F PROPORTIONS OF DIETARY COMPONENTS
Each of the classes of dietary components, proteins, carbohydrates, fats, vitamins, and minerals, is constituted of a variety of substances. Proteins are similar on the basis of nitrogen present in peptide bonds, but differ widely in component amino acids and in physical and biological properties. Lipids consist of glycerides of various fatty acids and a host of other esters. Vitamins bear no chemical resemblance to one another, being classified together as organic substances needed in small amounts. Minerals, of course, are grouped only on the basis of being inorganic. It is a t once obvious that experiments on the influence of varying the components of diet can therefore yield only particular data. Rational generalization on the basis of collateral knowledge is necessary to give
NUTRITION IN RELATION TO CANCER
467
the results broader and useful significance. For example, suppose that varying the proportion of casein (as the sole source of protein) is found t o influence the incidence of tumors. It is not at all certain that exactly similar results would be obtained if lactalbumin, gelatin, or a mixed source, were employed as the dietary protein. Or suppose that experiments had been conducted utilizing partially hydrogenated cottonseed oil (easily available and commonly used) as the dietary fat. The results might differ if animal fats, completely hydrogenated oils, highly unsaturated oils, or pure synthetic triglycerides, were used instead. Such facts are tacitly understood in nutritional research and are of importance when interpreting the results obtained with a particular member of a class. The practical design of experiments involving different proportions of nutritional components poses another problem. Alterations in diet often lead t o changes in body weight relative to that of the control animals. This may be consequent to a voluntary increase or decrease in caloric intake, or determined by an influence on the metabolism of the animal. Inasmuch as the genesis of tumors can be modified by a change in caloric intake and body weight, it becomes necessary t o determine, either by logic or experimentation, whether the observed effects on tumor formation are due to the experimental modification or to the accompanying caloric and body weight changes. Investigators have tried to meet this problem by such technics as paired feeding or force-feeding, or by giving to all animals equicaloric quantities of diets a t a reduced level, below that consumed by those with the poorest appetite. Even under these conditions, however, differences in absorption, composition, interconversions, and excretion may effect notable differences in body weight. For this reason, and because the animal’s weight is an excellent index of usable calories, it has sometimes been found expedient to adjust calorie intake so that control and experimental animals have equal average body weights. Of course, even this does not necessarily result in comparable tissue weights. If deemed helpful-and whenever possible-more than one technic should be utilized. In this way the confounding factors may be resolved and a decision obtained as t o whether the dietary alteration itself, or the resulting caloric or weight changes, is responsible for the effect on carcinogenesis.
1. Fat Under this heading we intend to discuss principally the effects of dietary neutral fats. It is recognized that the investigations with f a t actually utilized either butter, lard, partially hydrogenated vegetable oils, etc., but in general the influences were dependent on their neutral
468
ALBERT TANNENBAUM AND HERBERT SILVERSTONE
fat content. The many reports on the significance of dietary phospholipids and of cholesterol are omitted since the results are not convincing. In early studies, fat enriched diets were fed without considering their greater energy value. When low- and high-fat rations were fed ad libitum, the latter were usually consumed in calorically larger amounts because of their compactness. Probably the observed effects on carcinogenesis were the resultant of both increased fat and increased calories. In more recent investigations the caloric intakes have been equalized, permitting better evaluation of the significance of dietary fat. The discussions that follow are based on equicaloric comparisons, although in fact this has not always been achieved with exactness. a. General egects on tumor genesis. When the proportion of dietary fat is increased from low levels (about 2 per cent) to moderate or high levels, the response is related to the tumor type. With some neoplasms, genesis is enhanced; with others there is either no effect or slight inhibition. Skin tumors induced in mice, by tarring, by carcinogenic hydrocarbons, or by ultraviolet light, have been most frequently used in investigating the influence of fat enriched diets. Many different kinds of fats have been employed, of animal or vegetable origin, of different fatty acid constitution and degrees of saturation. I n all these studies tumors appeared in greater incidence and at an earlier average time in the mice on the high-fat rations (Watson and Mellanby, 1930; Baumann et aZ., 1939; Baumann and Rusch, 1939; Jacobi and Baumann, 1940; Lavik and Baumann, 1941, 1943; Tannenbaum, 194213, 1944c, 1945b; Rusch et al., 1945c; Boutwell el al., 1949a). The genesis of the spontaneous mammary carcinoma also is significantly augmented in mice fed high-fat diets. Tumors appear earlier and in greater numbers (Tannenbaum, 1942b, 1945b; Silverstone and Tannenbaum, 1950). There is suggestive evidence that the same is true for mammary tumors induced in rats by implantation of stilbestrol (Dunning et al., 1949). The rate of formation of the spontaneous benign hepatoma of strain C8H mice is slightly accelerated by dietary fat enrichment (Silverstone and Tannenbaum, 1951a). The influence of this procedure on hepatic tumors induced in the rat by p-dimethylaminoazobenzene is not clear, although the weight of evidence suggests that here, too, enhancement occurs (Opie, 1944; Miller et al., 1944b; Kline et al., 1946; Silverstone, 1948). Special features related to this tumor will be discussed later. In contrast to the promoting influence of fat-enriched diets on the formation of spontaneous mammary carcinoma and induced skin tumors, and probably hepatic neoplasms, are the negative results with the sarcoma
NUTRITION IN RELATION TO CANCER
469
induced by subcutaneous injections of carcinogenic hydrocarbons (Baumann et al., 1939; Tannenbaum, 194213;Lavik and Baumann, 1943; Rusch et al., 1945b),the primary lung adenoma (Tannenbaum, 1942b) and the spontaneous and induced leukemia of mice (Lawrason and Kirschbaum, 1944). It has been reported that an increase in dietary fat decreased the incidence of ocular orbit tumors occurring in rats fed acetylaminofluorene (Engel, 1951). Where the genesis of a tumor is augmented by fat enrichment of the diet, the effect can be demonstrated a t many levels of calorie intake. With both the mammary carcinoma and induced skin tumor, augmentation was observed, for example, not only a t near ad libitum levels (about 12 Calories daily), but with restricted diets as well (about 7 Calories daily) (Tannenbaum, 1945b; Boutwell et al., 1949a). b. Factors that modify the fat efect. Some of the tumors that respond to changes in dietary fat content have been studied with the purpose of ascertaining factors that may modify the magnitude of the effect. These are the degree of fat enrichment, the potency of the carcinogenic stimulus, and the action on the carcinogen itself. (1) Degree of fat enrichment. I n experiments with the spontaneous mammary carcinoma the rate of tumor formation, as measured by both incidence and average time of appearance, increased with increasing levels of dietary fat (Silverstone and Tannenbaum, 1950). The response was not arithmetically proportional to the fat content of the diet; an increase from 2 to about 8 per cent resulted in as great an augmentation as that consequent to an increase from 8 to about 26 per cent. Furthermore, there appeared to be a definite plateauing or maximal effect with diets containing about 16 per cent fat. Consistent with these findings is the observation that a diet containing 27 per cent fat stimulated the formation of induced skin tumors to about the same degree as one containing 61 per cent fat (Boutwell et al., 1949a). The available data indicate that within limits the incidence of spontaneous mammary and induced skin tumors follows a doseresponse relationship to the proportion of dietary fat. ( d ) Potency of carcinogenic stimulus. When a moderate dose of carcinogen was employed to induce skin tumors, fat enrichment of the diet resulted in both increased incidence and accelerated onset. Practically all animals, whether on low- or high-fat diets, developed neoplasms when the dose of carcinogen was large. Nevertheless, the augmenting influence of fat enrichment was still evidenced by the significantly earlier appearance of the tumors (Tannenbaum and Silverstone, 1947). (3) E$ects o n carcinogen. It was briefly stated in the section concerned with general observations that a change in the diet might have
470
ALBERT TANNENBAUM AND HERBERT SILVERSTONE
profound influences on the carcinogen itself. For instance it might alter the effective tissue dose of the carcinogen. How such a situation might arise and affect the results and interpretation of an experiment is now discussed. I n studies on skin carcinogenesis, the high-fat ration might be constructed or fed in such manner that the skin of the animal became greasy or oily through external contact. There would follow a striking augmentation, considerably in excess of the enhancement that occurs in the absence of such external contamination (Watson and Mellanby, 1930). The mere application of oil to the skin during the period of carcinogen administration in itself results in an increased tumor incidence. This occurs not only with skin tumors induced by chemical carcinogens, in which case greater solubility and absorption into the skin might be evoked as an explanation, but also with skin tumors induced by ultraviolet light (Rusch et al., 1939). We believe it necessary to distinguish between this nondietary action of fat which influences the effective dosage of carcinogen in the initiatory stage of skin carcinogenesis, and the dietary action of fat which appears to operate principally during the developmental stage (see below), and is of relatively small magnitude (Lavik and Baumann, 1943; Tannenbaum, 1944~). c. Mode of action of dietary fat. ( 1 ) Stage of carcinogenesis where action occurs. It was previously pointed out that caloric restriction exerted its main effect in the developmental stage of carcinogenesis. Similarly, experiments with induced skin tumors have demonstrated that fat enrichment of the diet augments tumor formation during the developmental stage (after the limited period of carcinogen application). Little influence resulted from feeding high fat rations in the period of carcinogen application only-during the initiatory stage (Lavik and Baumann, 1941; Tannenbaum, 1944~). ( 2 ) Metabolic influence. The same qualitative effects on carcinogenesis are elicited by the various fats employed in altering the diet, except in the case of the azo dye induced liver tumor of the rat (discussed in next paragraph). The fats are mainly triglycerides of fatty acids along with a small percentage of other substances. Glycerol and the nonsaponifiable fraction from cottonseed oil had little influence on the induction of skin tumors. On the other hand, the reconstituted triglycerides (after removal of nonsaponifhble matter) were as stimulatory as the partially hydrogenated cottonseed oil from which they were obtained. Corn oil, coconut oil, lard, ethyl laurate (Lavik and Baumann, 1941, 1943) or butter (Watson and Mellanby, 1930), were also effective. Most likely the fatty acid moiety is responsible for the enhancing action of dietary fat on skin and mammary tumors. There are no conclusive
NUTRITION IN RELATION TO CANCER
471
data on the influence of metabolic products of fat or of short chain fatty acids; the only experiments in which these were added in amounts sufficient for exhibition of a response were performed with induced sarcoma, a tumor that is relatively unaffected by dietary fat. The relation of fat enrichment of the diet to hepatic tumors induced in the rat by feeding p-dimethylaminoazobenzene is of special interest. It has been reported that the genesis of these tumors is accelerated by rations with moderate or high fat content as compared with rations containing little fat, but this has not been found consistently. With this neoplasm, however, the nature of the fat is important (Miller et al., 1944a, 1944b; Kline et al., 1946). Tumors developed less rapidly in rats consuming rations containing olive oil than in those given corn oil, while hydrogenated coconut oil retarded tumor formation significantly. This latter result is not attributable to a deficiency of essential unsaturated fatty acids. Increasing the corn oil content from 5 per cent to 20 per cent of the diet strikingly enhanced tumor formation, but replacing the 5 per cent corn oil with 20 per cent partially hydrogenated cottonseed oil or 20 per cent lard produced no appreciable augmentation. This diverse action of different fats has not been shown for other tumors and despite excellent investigations has not yet been explained in terms of the chemical, physical, or biological properties of the several fats. I n an attempt to explain the augmenting action of fat enriched rations, attention has been called to the following phenomenon: Increasing the fat content of a diet results in an increased efficiency of energy utilization (Forbes et al., 1946a, 1946b). This metabolic characteristic has been advanced as the factor possibly responsible for augmenting the incidence of induced skin tumors inasmuch as this “saving” of net body energy might be regarded as equivalent to an increase in caloric intake (Boutwell e l al., 1949a). However, there are two points which negate this suggestion (Silverstone and Tannenbaum, 1950). First, the magnitude of the increase in net body energy can account for only a small part of the observed augmentation of tumor formation. Second, if the fat effect were mediated through the increased net body energy gain it would be reasonable to expect fat enriched diets to enhance the formation of all types of tumors influenced by the level of caloric intake; this is not the case. In fact, the converse idea, that the influence of caloric restriction is mediated through the depletion of cellular and tissue fat consequent to a restricted food intake, could better be supported. What are the net conclusions from all the above considerations as to how isocaloric dietary fat enrichment affects carcinogenesis? The tissue dosage of carcinogen and thus the initiatory stage may be modified. On the other hand, the alterations in fat content of the tissues in which the
472
ALBERT TANNENBAUM AND HERBERT SILVERSTONE
carcinogen is acting, the local nutritional state, and the changes in net body energy, may all influence the developmental stage. The variable response of different types of neoplasms may depend on the extent to which these separate factors are involved. 2. Protein
So far as the animal is concerned the usefulness of dietary protein depends not only on the amount but also on the nutritional quality. Briefly, adequate protein may be considered as that which supplies the kind and amount of amino acids essential to normal physiology and growth. Rations that fall short of this standard are inadequate. The point at which diets are designated as adequate or deficient is of course somewhat arbitrary. When otherwise adequate semipurified rations are given ad libitum to young adult mice, varying the proportion of dietary casein between 9 and 45 per cent has little influence on body weight. Actually, the mice ingesting the higher protein rations may tend to be somewhat smaller because of a voluntary decrease in caloric intake. At isocaloric levels, the mice on diets containing 18 per cent casein are about 10 per cent heavier, on the average, than those on the 9 per cent casein ration; further increases up to 45 per cent casein have no great influence on body weight. Thus mice can maintain normal body weight if allowed free access to a 9 per cent casein ration, even though the efficiency of utilization is aomewhat decreased. Under the same conditions, mice fed lower proportions of casein do not grow well. We have therefore accepted proportions of casein of 9 per cent or over as being adequate for adult mice. The plans, results, and interpretations of the investigations concerned with dietary protein direct the separation of the discussion into: the consequences of varying the proportions within limits adequate for growth; and the effects of protein-deficient diets. a. Proportions of protein supporting normal body weight. (1) Tumors not affected. The incidence and rate of appearance of 3,4-benzpyreneinduced skin tumors did not vary significantly among groups of mice which were fed ad libitum semipurified diets containing 9, 18, 27, 36, or 45 per cent casein (Tannenbaum and Silverstone, 1949a). With the same series of rations, the rate of formation of spontaneous mammary carcinoma in strain CsHfemales also was found to be unaffected by the level of dietary protein. In another experiment in which isocaloric diets were employed, mice on 9 per cent casein weighed somewhat less and formed mammary tumors a t a slightly slower rate as compared with those on 18 or 36 per cent casein. The induction of sarcomas by carcinogenic
NUTRITION I N RElLATION TO CANCER
473
hydrocarbons was not modified by an increase in the proportion of dietary casein from 18 to 32 per cent (Tannenbaum and Silverstone, 19490,) and from 13 t o 26 per cent in diets fed ad libitum, or from 20 to 40 per cent in calorie restricted rations (Rusch et al., 194513). Thus with three distinct types of tumors, carcinogenesis was uninfluenced by varying the proportion of casein between 9 and 45 per cent. Perhaps under other experimental conditions, particularly a less potent carcinogenic stimulus, small but significant effects would be exhibited. However, the results obtained might be expected since most of the tissues of the animal were not markedly affected (either in siae or protein content) by variations within this range. Notable exceptions are the kidney mass and protein content of the liver, which increase with increasing dietary protein, even within these limits. This “sensitivity” of the liver to changes in protein supply serves as an introduction to the following discussion. (2) Tumors of the liver. The rate of formation of the spontaneously occurring benign hepatoma in mice given 18,27,36, or 45 per cent dietary casein was of about the same order. However, on 9 per cent casein rations there was a striking inhibition of hepatoma formation. This result has been observed with two strains of mice and in females as well as males. Moreover, the reduced incidence on 9 per cent casein occurred with diets fed ad libitum or isocalorically, and also in experiments in which caloric intakes were controlled so as to maintain equivalent body weights among the several groups (Tannenbaum and Silverstone, 1949a; Silverstone and Tannenbaum, 1951b). Addition of 9 per cent of gelatin (an inadequate protein) to the 9 per cent casein ration, did not significantly enhance hepatoma formation. In contrast, supplementing the 9 per cent casein with methionine and cystine increased the incidence of hepatomas to that observed among mice on 18 per cent casein. This might be interpreted as indicating that the sulfur-containing amino acids play a specific role in the genesis of hepatomas. However, there is a more likely explanation. The proportion of sulfur-containing amino acids in casein is a limiting factor in its nutritive value, and its biological potentiality is raised by supplementing with methionine or cystine. I n fact, adding 0.2 per cent methionine to a ration containing 9 per cent casein produces a diet nutritionally equivalent to one with 12 to 14 per cent casein. It is likely that the augmenting effect on hepatoma formation was due to the increased “balanced ” protein, not the supplementary sulfur-containing amino acids per se. The induction of malignant liver neoplasms in rats fed carcinogenic azo dyes is dependent on a number of dietary features, one of these being the proportion of protein. Increasing the dietary protein or improving
474
ALBERT TANNENBAUM AND HERBERT SILVERSTONE
the biological value of casein by supplementation with sulfur-containing amino acids has been reported to be either without influence or more often inhibitory for genesis of these neoplasms (Kensler et al., 1941; Miller et al., 1941; Rusch et al., 1945a; Harris et al., 1947; Silverstone, 1948). The outcome, however, is dependent on other features of the ration. Most significant perhaps is the amount of dietary riboflavin and the resultant concentration of riboflavin in the liver, the latter being enhanced by high dietary protein (Sarett and Perlzweig, 1943; Czaczkes and Guggenheim, 1946) and depressed on ingestion of carcinogenic dyes (Kensler et al., 1940; Griffin and Baumann, 1946). “It is doubtful whether the beneficial effects of dietary protein in rats fed azo dyes involves the critical carcinogenic reaction. More probably they represent merely another example of the ability of dietary protein to increase the resistance of animals to such diverse toxic agents as benzene, chloroform, arsphenamine, selenium, and the carcinogenic hydrocarbons ” (Griffin et al., 1949). The divergent effects of protein on the two neoplasms described-of the same organ-emphasizes that the response to a particular dietary change may vary with the kind of tumor. Raising the casein from 9 to 18 per cent enhances the development of the spontaneous benign hepatoma in mice; a similar change either inhibits or does not modify the genesis of dye-induced malignant liver tumors in rats. b. Deficient proportions of protein. In certain sublines of strain dba mice, leukemia can be induced by cutaneous application of methylcholanthrene. The production of this type of leukemia was significantly inhibited when the mice were on diets containing only 4 or 5 per cent casein, which are deficient in cystine (J. White et al., 1947). If these diets were supplemented with cystine, the induction of leukemia proceeded at its maximal rate. In contrast, rations deficient in lysine or tryptophan had little influence, except for a slight retardation in time of appearance of the disease (Table 111). The divergence of the effects observed with the respective deficiencies prompted White and associates to conclude that “cystine played a role in the development of leukemia not associated with its properties as an essential amino acid for growth but with some other attribute not yet determined.” It was pointed out, however, that among the mice on the low casein diet (without the cystine supplement) sclerosis of the aorta occurred in about one-half the animals. Since these latter died (some without leukemia) a t about the average time of appearance of the leukemia,‘ the results are somewhat obscured. A diet containing 18 per cent gliadia (lysine-deficient) or one with 4 per cent casein (cystine-deficient) effected a considerable suppression of
475
NUTRITION IN RELATION TO CANCER
the formation of spontaneous mammary carcinoma in strain C,H mice (White and Andervont, 1943; White and White, 194413). On the gliadin ration, the incidence was only 25 per cent, on the low casein ration, zero; in contrast, tumors appeared in nearly 100 per cent of the controls. The mice fed the gliadin ration averaged 21 g. in body weight; those given the low casein ration, only 13 g. It seems likely that nearly identical effects on tumor formation would have been obtained if the mice had been maintained a t these body weight levels through caloric (carbohydrate) restriction alone. TABLE I11 Effect of Diet on the Induction of Leukemia by Methylcholanthrene
Diet * 1. Cystine deficient Cystine deficient 0.5 % bcystine 2. Lysine deficient Lysine deficient 0.62% L-lysine 3. Tryptophan-deficient Tryptophan-deficient 0.1% L-tryptophan
+ +
* The
+
Body Mice Developing Weight? Leukemia1 Change Incidence Mean Latent (grams) (per cent) Period (days) -0.2 +4.5
+0.4
+ 6 .2
0
+ 5 .3
55 92 90 90 85
88
113 97 124 110 136 91
diets were semisynthetic rations with the following protein sources: 1. cystine deficient,
5 % casein; 2. lysine deficient, 18 % gliadin; 3. tryptophan deficient, 3 % casein, 7 % peroxide-treated oasein, and 0.28 % methionine.
t Mean change in body weight of
the mice during first sixty days of experiment. Praotically all the animals died with leukemia except in the group fed the cystine-deficient diet. Here there were a number of deaths with aortic sclerosis, many without leukemia, a t an average latent period of 121 days. f There were 38 to 40 mice initially in each experimental group.
The influence of 5 per cent casein rations, with and without supplementary cystine, on the formation of hepatic tumors induced by azo dyes, has also been studied. The food intake and consequently the dose of carcinogen was adequately controlled by paired feeding. All the animals developed liver neoplasms, but they appeared much later in the rats given the cystine supplement (White and White, 1946). Formation of spontaneous lung adenoma was greater in mice on a 4 per cent casein ration supplemented with cystine than in those on the unsupplemented ration. Food intakes were ad libitum, and the latter ate less and weighed less. However, when the food intakes were maintained at isocaloric levels, there was no significant difference in tumor incidence (Larsen and Heston, 1945). Investigations utilizing amounts and proportions of protein that result in diminished metabolism and growth of the host present two
476
ALBERT TANNENBAUM AND HERBElRT SILVERSTONE
difficulties in the evaluation of effects on carcinogenesis. One has already been discussed in detail-the significance of differing caloric intakes. The other is related to the manner in which the deficient protein diet is altered to prepare the control diet. Sometimes this is done by supplementation with the deficient amino acid. It is obvious that this measure results in rations that differ not only in proportions of the particular amino acid, but also in the proportions of balanced, adequate protein. The question arises, therefore, whether an observed change in the formation of tumors is due t o the specific amino acid or to the resultant increase in adequate protein. In many instances the latter appears to be the correct explanation. Therefore, the conclusions of some of these investigations-that particular amino acids play a specific role in the genesis of particular neoplasms-require reexamination. It is concluded that modification of carcinogenesis through a change in the proportion of dietary protein, within limits that permit relatively good growth and body weight, has so far been unequivocally demonstrated for tumors of the liver only. The spontaneous mammary carcinoma, induced skin tumor, and induced sarcoma apparently are not responsive. As previously mentioned, this may be related to the fact that alterations in dietary protein are reflected in the composition of the liver, but much less so in that of most extrahepatic tissues. Probably the observed effects on hepatic tumors are related to amounts of protein in the liver that are marginal with respect to optimal health and function of the organ. Perhaps the establishment of similar critical conditions in other tissues, with carefully controlled caloric intake and body weight, might also reveal the influence of protein deficiency on the intimate changes leading to the formation of tumors in these sites. 3. Vitamins
Vitamins are so important to the nutritional state of the host, that they are brought to mind whenever growth processes are being considered. Cancer is no exception. Consequently, numerous attempts have been made to link carcinogenesis with both deficiencies and ahundances of these food components. It is therefore disconcerting that vitamin deprivations usually result in lower food intake and body weight and a shortened life span, events which in themselves either modify the onset of neoplasms or do not permit the completion of the process. These complications may interfere with the intent of a study and render the interpretation unconvincing. The early investigations have been extensively reviewed (Stern and Willheim, 1943; Burk and Winsler, 1944; Morris, 1947). We have
NUTRITION IN RELATION TO CANCER
477
chosen a few of these for illustration and will discuss the more recent contributions. In the main, however, the purpose has been to attempt an appraisal of the subject. Only the B vitamins are considered, since there are no conclusive data that the others have noteworthy effects. a. Vitamin de$ciencies. The influence of a partial riboflavin deprivation on the formation of spontaneous mammary carcinoma in CaH virgin female mice has been studied (Morris, 1947). The groups on the deficient diets developed significantly fewer tumors than those on the riboflavin-supplemented rations. However, the food consumption and body weights of the depleted mice were also lower, and to an extent that could account for much of the reduction in tumor incidence. That adequate nutrition is necessary for the genesis of tumors, as a general rule, is indicated by the following investigation. A semipurified ration was designed to contain levels of B vitamins just sufficient to maintain young mature rats a t their initial weight. Addition of a carcinogenic dose of p-dimethylaminoazobenzene (butter yellow) t o the diet caused the rats to lose weight; some died, and no liver tumors were found in survivors that had ingested the diet for as long as six months. Under these conditions of multiple vitamin deficiency, even a low level of riboflavin, known to favor the genesis of azo dye liver tumors, failed to evoke them (Miner et al., 1943). b. Vitamin levels and azo dye liver tumors. In contrast to the foregoing are the studies indicating that, with an otherwise adequate diet, low levels of riboflavin enhance the development of liver tumors in rats fed azo dyes (Kensler et al., 1940, 1941). The investigations of the University of Wisconsin group, including a survey of 34 diets, demonstrated that dietary supplements rich in both protein and B vitamins, particularly riboflavin, inhibited tumor formation (Miller et al., 1941). Protein is thought to have little direct effect, acting rather by altering the concentration of riboflavin in the liver. The significance of the concentration of riboflavin actually attained in the liver-not the amount in the diet-as a key factor in the genesis of liver tumors induced by means of azo dyes, is now established (Miller, 1947; Kensler, 1947; Griffin and Baumann, 1948). Low levels enhance, high levels inhibit tumor formation: There has even been progress as to the mechanism of action. Riboflavin appears to be a constituent of a coenzyme in the systems that break down dimethylaminoazobenzene and related compounds (Kensler, 1948; Mueller and Miller, 1950). This implies that riboflavin acts on the initiatory stage of carcinogenesis by modifying the effective dose of carcinogen. The consequence of varying the amount of riboflavin in $he diet is not of the same order for the various azo dyes (Giese et al., 1946); and an influence on the induction of
478
ALBERT TANNENBAUM AND HERBERT SILVERSTONE
hepatic tumors in rats fed a different type of carcinogen, acetylaminofluorene, seems to be absent or negligible (Harris, 1947). It should be mentioned here that dietary alterations might also affect the genesis of these neoplasms during the developmental stage. After a limited period of feeding p-dimethylaminoazobenzene in a brown rice diet, rats were continued on dye-free rations of brown rice, brown rice supplemented with yeast, or brown rice supplemented with milk powder. The incidences of liver neoplasms were significantly lower in the rats given the yeast or milk powder (Sugiura, 1944, 1951). Of the other B vitamins only biotin and pyridoxine appear to affect appreciably the rate of formation of azo-dye-induced hepatic tumors. Relative deficiencies of either of these vitamins hinder the genesis of the neoplasms (Burk et al., 1943; Miner et aE., 1943; Kline et al., 1945; E. C. Miller et al., 1945; Miller, 1947). The history and findings of the dietary studies, including vitamin alterations, on liver tumors induced in rats fed p-dimethylaminoazobenzene, have been more extensively reviewed elsewhere (Rusch et aE., 1945a; Opie, 1947). c. Tumors induced by vitamin deficiency. Up to this point in the review, experiments on diet and the nutritional state of the host as modifiers of carcinogenesis have been presented. There are two areas of investigation, however, suggesting that certain dietary deprivations may indeed be initiators of carcinogenesis-produce tumors without the administration of a known carcinogen. It has been recognized for some time that rats ingesting inadequate rations, deficient in vitamin A particularly, develop papillomatosis of the forestomach. The lesions may not be true neoplasia but rather an exaggerated example of the hyperplasia, hyperkeratinization, and metaplasia of epithelial tissues found in the vitamin A-deficient state (Passey et al., 1935; Fridericia et al., 1940). More recently it was reported that liver tumors develop in rats severely deficient in choline. The strain of rat used in these studies has a high requirement for choline, and the diets were so depleted that periodic administration of choline was necessary to prevent the death of the animals. The authors stated that “the choline deficient rats were in reasonably good condition for several months. Following this they gradually became unthrifty in appearance. There was marked loss of hair, muscular weakness, drowsiness, and lethargy ” (Copeland and Salmon, 1946; Engel et aE., 1947). A variety of pathological lesions was found, some obviously inflammatory. The state of the liver is of greatest ipterest. Severe degenerative and advanced cirrhotic changes were found in all animals which were on
NUTRITION IN RELATION TO CANCER
479
the choline-deficient diet for a t least eight months. Almost half the rats had hepatomas-these were not uniform in histologic appearance, and many were benign. This finding poses the important question as to whether the nodules are regenerating masses of liver cells in a severely damaged organ or true neoplasia. If it be the latter, do not these results support the often expressed view that tumors arise not by stimulation of normal cells, but as an adaptation to specific kinds of tissue injury? It is already known that cancer can be produced by physical, chemical, or parasitic agents. May severe dietary deficiencies, under special circumstances, also be carcinogenic? d. Effects independent of caloric intake and body weight. Several investigations have been carried out, in the past few years, on the influence of varying the level of B vitamins from what might be considered barely adequate for growth to a few times the optimal amounts. The studies utilized several tumor types, and control, or a t least appraisal, of caloric intake and body weight was accomplished. I n one such experiment the effect of multiple B vitamins, in low, control, and excess amounts, on the incidence of spontaneous lung adenoma in strain A mice was tested. Differences in tumor formation occurred, but these could easily be attributed to the differences in deaths from causes other than cancer (Taylor and Williams, 1945). Under similar dietary conditions the formation of skin tumors induced in mice with 3,4-benzpyrene was studied (Boutwell et al., 1949b). Ten of the B vitamins were varied as a group from levels estimated to be just adequate for maintenance to those considerably above the optimal. In other experiments, individual, pairs, and sets of B vitamins were given in minimal amounts, along with adequate quantities of the rest of the members. All diets were fed isocalorically. There were no significant differences in skin tumor formation, except for a somewhat lower incidence in the group fed low levels of all B vitamins. The mice of this group were reported to have shown some evidence of pyridoxine deficiency. Previously, the same laboratory reported an inhibition of skin tumor induction in mice fed a pyridoxine-deficient diet; caloric intakes and average body weights were controlled (Kline et al., 1943). We also have investigated the influence of B vitamins as a group, a t maintenance levels and in threefold or ninefold greater amounts, on the induced skin tumor (unpublished). Final incidences of neoplasms were of the same order in the three groups although the rate of formation was possibly slower in the mice of the low vitamin group. The same dietary factors were employed in experiments on the spontaneous mammary carcinoma of dba mice. The animals fed the low
480
ALBERT TANNENBAUM AND HERBERT SILVERSTONE
B vitamin diets weighed approximately 10 per cent less than those receiving three or nine times the minimal amounts. There were no noteworthy differences in the incidence of neoplasms, although here again they appeared in the groups on the low vitamin intake at a somewhat later average tihe (unpublished). With the spontaneous hepatoma of CsH mice, minimal levels of riboflavin or of B vitamins as a group had no important influence on rate of formation. Hepatomas developed in somewhat fewer of these animals, compared with controls, but the result is explicable on the basis of slightly lower average body weight (Silverstone and Tannenbaum, 1951a; unpublished). Analysis of the findings on the influence of varying the level of B vitamins suggests that wide changes in dietary content, in the range above minimal needs, have little effect on carcinogenesis. At least, this appears to be valid for the induced skin tumor and the spontaneous mammary, lung, and liver tumors of the mouse. As deficiency levels are approached there may be some inhibition of carcinogenesis, but in these instances caloric intake and body weight changes may be the major cause of the altered response. Except for the liver tumors induced by a50 dyes or choline deficiency, the levels of dietary vitamins have not been shown to affect significantly the genesis of tumors. Further work along these lines may prove fruitful, but only if the mistakes of the past are avoided, and the delicacy of the problem is recognized.
4. Minerals Several inorganic substances are implicated in the genesis of tumors. Clinical and experimental evidence reveal the carcinogenic action of arsenic, beryllium, chromates, and radium and other radioactive substances; however, none of these is a normal dietary constituent. There is no evidence that the formation of neoplasms is influenced by altering the proportions of inorganic components natural to the diet. The literature contains many contradictory reports; and the best studies, in the main, conclude that there are no important effects (Shear, 1933; Stern and Willheim, 1943). In recent investigations with the induced skin tumor, the spontaneous benign hepatoma, and the spontaneous mammary carcinoma, mice were fed semipurified diets containing a complete salt mixture at levels of 2, 4, or 8 per cent; a salt content of 4 per cent may be considered as normal. With all three tumor types there were no significant differences in carcinogenesis. The small effects observed could be attributed to the differences in caloric intake and body weight (unpublished).
NUTRITION I N RELATION TO CANCER
48 1
111. GROWTHOF TUMORS The neoplastic process is composed of a series of stages. Those comprising genesis have already been discussed; those grouped under the term “growth” are the subject of this section. For experimental consideration, the growth of a neoplasm can be divided into: 1) establishment in the host of the first few neoplastic cells, whether these developed in situ or were contained in fragments transplanted from another tumor; and 2) growth proper, that is, multiplication of the tumor cells to result in an increase in mass. Sufficiently distinct t o merit separate consideration is the spread of the tumor through invasion and release of metastastic emboli; this problem has been largely neglected in nutritional studies. The fact that this section on tumor growth is shorter than the foregoing discussion on genesis is no measure of their relative significance or of the number of investigations in these two areas. Actually, it seems a fair guess that more studies have dealt with growth. Disproportionate space has been allotted these subjects in the present review because changes in the nutritional state have definite and often striking influence on the genesis of tumors. In contrast, the studies concerned with the growth of tumors commonly reveal smaller effects or none at all; and for some diet components there is only a host of cloudy and contradictory reports. 1. Caloric Intake
Among the earliest studies on nutrition in relation to cancer are those indicating that underfeeding retards the growth of transplanted tumors (Moreschi, 1909; ROUS,1914; Sugiura and Benedict, 1926; Bischoff et al., 1935). Later this was shown to be a consequence of restriction in calories; i.e., it could be produced by reduction of only the carbohydrate (or fat) content of the diet (Bischoff and Long, 1938). I n some of these studies, there was a clear separation between establishment and growth of the tumors. When mice bearing mammary carcinoma were transferred to a low calorie ration, the growth rate of the tumors was significantly diminished; some even ceased growing and others became smaller, but complete regression was only rarely observed (Rous, 1914; Tannenbaum, 19404. The life span of the mice was increased to a small extent. In similar studies with sarcomas that had been previously induced in full-fed mice, caloric restriction hindered tumor growth but did not prolong life (unpublished). Mice inoculated with certain strains of leukemic cells lived longer, on the average, when underfed; however, with another strain of leukemia
482
ALBERT TANNENBAUM AND HERBERT SILVERSTONE
the average length of life was unaffected or possibly shortened (Flory et al., 1943). The underfeeding was begun the day after inoculation, and it is uncertain whether the restriction delayed the establishment of the leukemia or inhibited subsequent proliferation. Underfeeding or caloric restriction does not appear to be a successful means of controlling the growth of tumors. Not only is the neoplasm affected, but at the same time the normal tissues of the host diminish. It is questionable whether the life of the animal can be significantly or consistently lengthened. 2. Fat Variation in the proportion of dietary fat does not seem to influence the growth of tumors. Establiihment and growth of transplanted mouse and rat carcinomas and sarcomas were unaffected by fat enrichment of the diet (Sugiura and Benedict, 1930; Baumann et al., 1939). Sarcomas induced in mice on high fat diets grew a t the same rate as those induced in animals on a low fa t diet (Baumann et al., 1939; Tannenbaum, 1942b). The mean growth rate of spontaneous mammary carcinomas, arising in animals ingesting a semipurified diet containing 24 per cent fat, was the same as that in animals on a diet in which the fat was reduced to 2 per cent by equicaloric substitution of carbohydrate. Furthermore, the interval between recognition of the tumor and death of the mouse (survival time) was not correlated with the proportion of dietary fat-2, 4, 8, 16, or 24 per cent (Silverstone and Tannenbaum, 1950). 3. Protein
Numerous studies on dietary protein have succeeded in illustrating the vigorous ability of neoplasms to become established and grow, even in animals that are experiencing severe depletion. However, as with restriction of calories, protein deprivation decreases the rate at which these processes proceed. A more recent development has been the demonstration that variations in protein intake can modify the action of nondietary procedures. a, Variations in dietary protein only. The imposition of drastic dietary restriction of protein significantly hinders the establishment and growth of tumors. Spontaneous mammary carcinoma increased in size less rapidly in animals fed diets low in lysine or in cystine and methionine (Voegtlin and Thompson, 1936; Voegtlin and Maver, 1936; Voegtlin et al., 1936; Morris and Voegtlin, 1940). Kocher (1944) found that lysine-deficient diets produced only transient slowing of tumor growth; or, if the deficiency was instituted when the tumors were as large as 25 mm. in diameter, there was no effect. Although the results were
NUTRITION I N RELATION TO CANCER
483
discussed in terms of amino acid deficiencies, they can also be interpreted in terms of protein deprivation. In a well-designed study with a transplantable mammary adenocarcinoma, mice were fed a diet practically free of protein; the animals remained in continuous negative nitrogen balance and lost considerable weight during the experiment (White and Belkin, 1945). One week after institution of protein restriction the mice were inoculated with a mammary adenocarcinoma. The establishment of the tumors was not prevented by the protein deprivation. The rate of growth, however, was retarded but still was remarkably rapid-74 per cent of that in the controls-when one considers that the mice were consuming an almost protein-free diet. The final mean volume of the tumors in the depleted animals was nearly half that of the control tumors. A similar experiment with the Walker rat tumor 256 has been reported by Green and associates (1950). In this study the establishment of the implants was definitely delayed in rats protein-depleted by means of a diet containing only 1 per cent casein. Once the tumors became palpable, however, they increased in volume at about the same relative rate as those in control rats on a 22 per cent casein diet. Of course, since they became established later, the absolute rate of growth was less, and the tumors were not as large. Moderate dietary deficiency of protein has only a small effect on the growth of neoplasms. This has been shown in almost identical studies from two laboratories. Rats maintained on an otherwise adequate ration containing 5 per cent casein weighed only one-sixth less than those fed a 20 per cent casein diet. Correspondingly, the transplanted Walker 256 tumors in the protein-restricted animals attained weights about onesixth or one-twelfth less than those of the controls (Green and Lushbaugh, 1949; Devik et al., 1950). Essentially the same observations were made with another neoplasm, the rat hepatoma 31 (Voegtlin and Thompson, 1949). Inasmuch as diets containing as little as 5 per cent casein have only a small retarding effect upon the growth of tumors, one would expect that rations with higher proportions of casein would not be inhibitory. Actually, mice themselves grow well and at about the same rate when fed diets containing between 9 and 45 per cent casein. In mice consuming rations with casein contents within these ranges, spontaneous mammary carcinomas enlarged a t rates independent of protein intake; nor were there any effects upon the survival times and incidences of metastases (Tannenbaum and Silverstone, 1949a). Comparably, the growth rate of induced sarcomas in mice was not modified by varying the proportion of protein within the same limits (Rusch et al., 1945b; Tannenbaum and
484
ALBERT TANNENBAUM AND HERBERT SILVERSTONE
Silverstone, 1949a). Transplantable Walker tumor 256 grew as well in rats on a 10 per cent protein ration as in those on a 20 per cent protein intake (Elson and Haddow, 1947). I n summary, the response of tumor growth to alterations in dietary protein parallels the response of the host. Wide variations that have little effect upon the host, have no perceptible effect upon the neoplasm. Only with protein deprivation serious enough to cause striking loss of body weight is there significant retardation of the rate of establishment and rate of growth of tumors. b. Mode of action of protein restriction. Algire and Chalkley (1945) have demonstrated that growth of transplanted tumors does not become evident until completion of the primary vascularization process, on or about the fifth day. Complementary observations indicate that the initial inflammatory reaction to the implant is followed by development of a primitive connective tissue network into which the neoplastic cells migrate (Devik et al., 1950). In animals ingesting rations adequate in protein (for example, 20 per cent casein) the ingrowth of capillaries and organization of a connective tissue “capsule” is evident by the fourth to seventh day-establishment of the tumor is completed. A comparison of tumor implants in rats on 5 per cent casein with those in animals on 20 per cent casein revealed that the initial inflammatory reaction was greater on the low protein diet; on the other hand, the ingrowth of capillaries and the organiration of the connective tissue within and around the implant was delayed. These detailed findings indicate that one definite consequence of low dietary protein is delayed establishment of implants. There are no data as to whether these effects are specific or also follow other kinds of deficiency in dietary components or calories. Once established, even in animals markedly depleted in protein, a tumor increases in size and weight. It is remarkable that growth of the neoplasm proceeds while the host may be in continuous negative nitrogen balance. Apparently the protein of the tumor can be derived almost entirely from the host This fact should not lead to the oversimplification that the tumor has a specific affinity for protein only. Neoplasms grow, although at a reduced rate, in animals undergoing a variety of dietary deficiencies; such animals may be losing fat, vitamins, and minerals, as well as protein, yet these essential cellular constituents are accumulated in the growing neoplasm. Rather than regarding increase in protein as a special feature of tumors, it can be considered as only one aspect of their ability to grow under adverse circumstances. c . Protein restriction combined with other treatment. When 1,2,5,6dibennanthracene or other carcinogenic hydrocarbons were injected
NUTRITION I N RELATION TO CANCER
485
intraperitoneally in fairly large doses, the growth of tumors was inhibited (Haddow, 1935). More recently it has been shown that this action is highly dependent on the proportion of protein in the diet (Elson and Haddow, 1947). Rats were maintained on diets containing 20, 10, or 5 per cent protein for two weeks and then implanted with Walker tumor 256; the following day, half the rats on each protein level were injected with 50 mg. of dibenzanthracene. During the following thirteen days the tumors of the untreated rats grew to about the same average size; in other words, unrelated to the proportion of dietary protein. In contrast, the mean find tumor weights decreased with decreasing protein intake among the animals given dibenzanthracene ;the most pronounced inhibitory effect of the hydrocarbon was obtained with the rats on the 5 per cent protein diet. Likewise, Green and Lushbaugh (1949) found that intraperitoneal injections of 44imethylaminostilbene, beginning four days after the subcutaneous implantation of the Walker tumor 256, retarded tumor growth. This treatment was much more effective in the rats given a diet containing only 5 per cent casein than in those on a 22 per cent casein ration. The reports concerned with the influence of dietary protein on the response of tumors to x-radiation are of considerable interest (Elson and Lamerton, 1949; Devik el al., 1950). These studies were performed with Walker tumor 256 transplanted to rats on 5 per cent or 20 per cent protein diets. Irradiation of the tumor bed was begun six days after implantation. In the animals receiving the 5 per cent casein diet, the transplants first grew slowly and then more rapidly, ultimately causing the death of most of the animals; however, 15 per cent of the neoplasms completely regressed. Among the rats on the 20 per cent casein ration, on the other hand, the tumors grew well for the first few days, then began to shrink in size; in 90 per cent of the animals the neoplasms regressed completely. These three studies intimate synergistic action of protein restriction and the nondietary procedures. They also point out a potentially fruitful field for investigation. It should be noted, however, that in the described experiments the nondietary treatments were instituted at one, four, or six days, respectively, probably before the implants were fully established in the animals on the low protein diet. Thus, at least part of the differential effects might have occurred through action on tumor establishment.
4. Vitamins Critical examination of the numerous investigations on the influence of vitamins, in deficient and excess amounts, on the growth of neoplasms
486
ALBERT TANNENBAUM AND HERBERT SILVERSTONE
reveals relatively few instances of clearly demonstrated effects. On perusal of reviews on this subject, one is impressed by the number of studies, diversity of results, and the inherent complicating factors (Stern and Willheim, 1943; Burk and Winzler, 1944; Morris, 1947). Various multiple and individual vitamin deficiencies undoubtedly hinder the growth of tumors; but in only a few instances can any specificity be attached to the results. Recently, there has been improvement in the experimental approach; and the development of the antivitamin analogue technique has supplied a new and useful tool. The following few examples of current research are representative. Illustrating the problems encountered in studying the influence of dietary deficiencies on tumor growth are Morris’ (1947) experiments on thiamine deprivation in mice bearing spontaneous mammary carcinomas. When diets were fed ad libitum, the mice on the low-thiamine ration ate less food and lost weight in comparison with those on the thiaminesupplemented ration. As might be anticipated tumor growth was retarded. However, if the food intake of the control mice was restricted to that of the thiamine-deficient animals, the rates of tumor growth were essentially the same. Unexpected were the results of an experiment in which thiamine-deficient and thiamine-supplemented rations were force-fed in equal amounts, presumably at about the level that would be voluntarily ingested by the controls: The depleted mice showed no relative loss in body weight, and their tumors grew a t an apparently accelerated rate in comparison with those of the thiamine-supplemented animals. The findings of this last experiment are in agreement with those obtained by DobrovolskaIa-Zavadskaia (1945), who reported that daily subcutaneous injection of thiamine hindered the growth of spontaneous mammary carcinoma in mice. The influence of various vitamin deprivations on the growth of established mouse lymphosarcoma C3H-ED transplants was examined by Stoerck and Emerson (1949). The deficiencies were induced by the use of antagonistic analogues. Those of thiamine, niacin, and folk acid had no effect. However, production of deficiencies in either pyridoxine or riboflavin caused apparently complete regression of the tumors. Correction of the pyridoxine deficiency resulted in reappearance and growth of the tumors, whereas no recurrence was observed following relief of riboflavin deficiency. The establishment and possibly the growth of Rous chicken sarcoma was markedly inhibited by folic acid deficiency. In contrast, restriction of other vitamins did not influence tumor growth except when the chicks were depleted to dangerous levels (Little et al., 1948).
NUTRITION IN RELATION TO CANCER
487
Folic acid antagonists have a marked inhibitory effect on some lines of transmissible mouse leukemia, presumably through the induced folk acid deficiency. It has been pointed out that in most studies on mouse leukemia the antagonists were administered very shortly after inoculation of the leukemia cells. Possibly then, interference occurs in the tumorhost integration or establishment (Kirschbaum el al., 1950). In studies with four different transplanted leukemias neither aminopterin nor amethopterin prolonged the life span of the host if treatment was delayed until eight days after inoculation. Nevertheless, there must be some influence on the multiplication of leukemic cells; unquestionable support for this is the often dramatic effect.in children with acute leukemia. There is an interesting report on the consequences of ascorbic acid deficiency on the growth of a previously established sarcoma in the guinea pig. After institution of a scorbutogenic diet tumor growth slowed apprec'ably, even before any changes were noted in food consumption or body weight. The neoplasms attained a weight only about one-fourth those of the control pigs. A greatly altered stroma and decrease in collagen of the tumors in the scorbutic animals were regarded as significant (Robertson et al., 1949). The growth of several kinds of tumors is suppressed by administration of 8-azaguanine. In studies with the mammary adenocarcinoma 755 in strain C67 Black mice, it was found that the effect could be greatly enhanced by injection of other substances, begun four to fourteen days after tumor implantation (Shapiro and Gellhorn, 1951). Some of the observed synergisms might have been anticipated; for instance, desoxypyridoxine and 7-methylpteroylglutamic acid (in doses that had no consistent effect when administered alone) intensified the retardation. Against expectation was the observation that similar extension of the azaguanine inhibition generally followed administration of vitamin Bla or folic acid. The effects could not be attributed to obvious toxicity to the animal. In the main, our general remarks in the last paragraph under " Genesis of Tumors-Vitamins '' apply also to growth. Unfortunately, the increment to our knowledge through research on vitamins in relation to tumor growth is not commensurate with the time and funds expended. Perhaps, as more is learned about both the cancer process and the science of nutrition, studies in this area will become more definitive.
STATEAND CANCERIN MAN IV. NUTRITIONAL
mars
Our knowledge of the role of diet in the genesis and growth of has come mainly from investigations with animals. However, there are
488
ALBERT TANNENBAUM AND HERBERT SILVERSTONE
statistical and clinical studies on man relevant to the significance of the nutritional state, Some of these are in agreement with the findings with animals; others might well provide a basis for further animal experimentation. 1. Body Weight and Cancer Incidence
The genesis of many types of tumors is arrestingly curtailed in mice that are calorie-restricted or underweight. Does this phenomenon have its counterpart in human beings? This question cannot be answered readily by experimentation, but statistical and clinical studies throw light on the subject. For some years, insurance companies have been interested in the connection between various social and biological factors and the relative frequencies of the principal causes of death. Relevant to this review is the correlation between body weight and cancer mortality. The general methods of approach and the validity of these studies are acceptable (Tannenbaum, 1940b, 1947). Although the statistics relate to cancer mortality, the latter corresponds roughly to cancer incidence. In this regard, it is important to emphasirie that the insured were classified according to weight at issue of policy, many years before they were diagnosed as having cancer. Six of the insurance studies and the results of one extensive questionnaire on dietary habits have been summarized in a review (Tannenbaum, 1940b). A brief outline of the findings is presented in Table IV. They imply that individuals who overeat and are overweight when past middle age are more likely to die of cancer than persons of average weight or less. In one study the cancer mortality per 100,000 insured persons 25 per cent or more overweight, normal weight, and 15 per cent or more underweight, were 143, 111, and 95, respectively. A difference of 50 per cent in cancer mortality between overweight and underweight individuals! Inasmuch as this relationship between cancer incidence and body weight is strongly supported by controlled animal experimentation, it seems reasonable to expect that the avoidance of overweight would result in the prevention of a considerable number of cancers in man, or at least in the delay of their time of appearance. Furthermore, the dietary control need not be drastic since both the animal studies and the insurance statistics reveal a dose-response relationship. Even moderate continued caloric restriction or control of body weight deters the development of neoplasms. These remarks apply only to prevention of the genesis of tumors and not to the treatment of cancers once they have formed. Present evidence holds no hope that calorie restriction or lowering of body weight is a practical means of controlling the growth of an established cancer.
489
NUTRITION IN RELATION TO CANCER
TABLE IV Statistical Studies on Body Weight and Cancer Mortality Reference
Source of Material
Conclusions
Actuarial Society of America Contributing Insurance Cancer mortality increases and Association of Life Companies with increasing body weight Insurance Directors, 1913 Cancer mortality increases Union Central Life Dublin, 1929 with increasing body weight Insurance Company Metropolitan Life A high percentage of those Dublin, 1929 Insurance Company dying from cancer are overweight Actuarial Society of America Contributing Insurance No clear relationship between cancer mortality and and Association of Life Companies body weight Insurance Directors, 1932 Questionnaires of Overnutrition is unusually Hoffman, 1937 common in histories of hospitalized patients cancerous patients with cancer Cancer mortality increases Metropolitan Life Dublin and Marks, 1938, with increasing body weight Insurance Company 1939 (true for some sites but not for all) Hunter, 1939 New York Life Cancer group has a higher Insurance Company average weight than control group (true for some sites, not for all)
2. Dietary Dejiciencies and Cancer Production
The differences in the dietary of various races, coupled with clinical and pathologic data on the relative frequency of various types of neoplasms, have led to impressions that some tumor types in man might be related to dietary deficiencies. The evidence and reasoning with regard to the pathogenesis of these neoplasms are circumstantial. Before accepting these oft-repeated claims, it might be well to await further clinical study and experimental proof. a. Thyroid cancer. The presence of nodular goiter appears to influence the occurrence of cancer of the thyroid. BBrard and Dunet, cited by Wegelin (1928), found that in endemic goitrous regions from 2.5 to 4.0 per cent of all malignant tumors arose in the thyroid, whereas in relatively goiter-free areas the frequency was only 0.4 to 0.5 per cent. Wegelin himself commented upon the much greater incidence of thyroid cancer in necropsies in Berne, the center of a goitrous region, in comparison with that reported from Vienna, Prague, and Berlin. Statistics obtained from both operations and autopsies indicate that these tumors
490
ALBERT TANNENBAUM AND HERBERT SILVERSTONE
occur about ten times more frequently in Switzerland than in the United States. That nontoxic nodular goiter is a precursor of malignancy is also suggested by studies on individuals within this country (Ward, 1944; Cole et al., 1945). Inasmuch as goiter occurs with much greater frequency in regions where there is an iodine deficiency of the soil, drinking water, and foodstuffs, it is reasoned that a chronic iodine deficiency may be a factor in the genesis of thyroid cancer. Perhaps the iodine deficiency sensitizes the thyroid to goitrogens and even potential thyroid carcinogens. b. Pharyngeal cancer. Women of those areas of Sweden and Finland within the Arctic Circle are reputedly prone to develop carcinoma of the pharynx, oral cavity, and esophagus. Individuals with such tumors often have a history of Plummer-Vinson syndrome, characterized by anemia, achlorhydria, and atrophy of the mucous membranes; later, hyperkeratoses of the oral and pharyngeal mucosa develop (Ahlbom, 1936; Adair, 1947). Ahlbom found this syndrome in 80 of 123 women with cancer of the mouth, pharynx, or esophagus. Might it be related to an iron- and vitamindeficient dietary-reindeer meat and fish, with few green vegetablea-supposedly common to the inhabitants of the region? c. Liver cancer. The frequency of primary cancer of the liver, as a proportion of all neoplasia, is high in some regions of the world, low in others. Among natives of southeastern Asia-China, Japan, Java, and Sumatra-the relative incidences range between 7 and 41 per cent, and even higher among young male African Negroes (Bantus) employed as gold miners (Bonne, 1937; Berman, 1940; Gilbert and Gillman, 1944). In contrast, the disease accounts for only about 1 per cent of all cancers among Caucasians, and is also low among Negroes living in Europe and the United States (Pack and LeFevre, 1930; Berman, 1940; Kennaway, 1944). Specific racial susceptibility, endemic infections, and faulty nutrition have been suspected as probable etiologic factors of liver cancer among those groups generally afflicted. The dietary theory has attracted interest, particularly since cirrhosis of the liver is also common among the population prone to cancer of the liver (Berman, 1941; Gilbert and Gillman, 1944). Attention is thus focused on the rather general opinion that hepatic cirrhosis in man is frequently associated with primary cancer of that organ. There is little experimental evidence for a causal dependence. Liver cancers have been produced without cirrhosis either as a precursor or as an associated condition. On the other hand, severe liver damage, including cirrhosis, has been produced in rats by feeding cornmeal mush and sour milk, the principal food of the Bantus, and yet no tumors were found (Gillman, 1944; Gillman et al., 1945).
NUTRITION IN RELATION TO CANCER
49 1
More recently, it has been reported that primary cancer of the liver may be initiated by the ingestion of certain crude substances. Rats fed chilli peppers (Hoch-Ligeti, 1951) or alkaloids of Senecio jacobaea (Cook et al., 1950) developed tumors of the liver. These substances, while not considered normal dietary components, are nevertheless used by some of the populations revealing the abnormally high incidence of liver cancer. By analogy with the influence of diet on hepatocarcinogenesis through azo dyes, poor dietaries might favor the development of liver cancer among individuals ingesting such crude “ carcinogens.” The entire subject is illuminated in a recent monograph by Berman (1951). He suggests that liver cancer in man depends on a sequence of events, a combination of environmental circumstances. Malnutrition, by injuring the liver, may make it more vulnerable to a variety of infections and toxic substances that have carcinogenic potentialities. This reasoning also may apply to the neoplasms discussed above-of the pharynx and the thyroid.
V. CONCLUSIONS AND COMMENTARY I n the preceding sections we have attempted to organize the available knowledge on the significance of diet in the cancer process. Important contributions may have been omitted inadvertently. More often, however, reports have not been included because of their uncertain meaning due to one or more of the following: the omission of details revealing the composition of the diets, the very small numbers of animals utilized, the lack of data on food intake and body weights, the stormy course of the experiment including many nontumor deaths early in the study, or questionable results. Many interesting experiments concerned with the effects of animal tissues, organ extracts, and other relatively crude materials have not been taken up because of insufficient information as to their composition and the components responsible for the effects noted, and for lack of space. Nevertheless, a reasonably large number of facts have been accumulated, allowing a better degree of comprehension and systematization than possible a decade or two ago. It is intended in this section to summarize current knowledge, to discuss some implications of the findings, and the directions future work might take; and finally, to make some speculative and provocative comments as to the role that nutrition plays in the genesis of neoplasms. I . Summary of Present Knowledge a. Genesis of tumors in animals. It has been demonstrated that the origin of neoplasms depends in part on the nutritional state of the host. A noteworthy example is the inhibition of the formation of tumors that
492
ALBERT TANNENBAUM AND HERBERT SILVERSTONE
is brought about by dietary restriction of calories. Many diverse types of neoplasms respond to caloric deprivation by a reduction in tumor incidence and a delay in appearance: spontaneous mammary carcinoma, skin tumors induced by carcinogenic hydrocarbons or ultraviolet light, induced sarcoma, spontaneous lung adenoma, spontaneous hepatoma, and spontaneous and induced leukemia-all of the mouse; and lymphosarcoma and induced mammary carcinoma of rats. In fact, all tumor types tested extensively have been affected in this way. For completeness it should be mentioned that in two instances a low-calorie effect has not been found, but this may be due to masking by special factors of carcinogenesis. The magnitude of the inhibition of tumorigenesis by caloric restriction is related to the extent of the deprivation and the composition of the restricted diet. In addition, the repression occurs mainly during the developmental stage; the initiatory stage seems little affected. In contrast t o the arresting and relatively consistent influence of caloric restriction are the quite diverse effects of fat enrichment of the diet. The genesis of the spontaneous mammary carcinoma and induced skin tumor is somewhat enhanced by high fat diets. Hepatic neoplasms, spontaneous and induced, also tend to form more readily when the animals ingest these diets, Contrariwise, no noteworthy effect has been observed with the induced sarcoma, primary lung adenoma, and spontaneous and induced leukemia. Where high fat rations favor carcinogenesis, there is evidence that the influence occurs in the developmental stage; and its intensity is dependent on the degree of fat enrichment. Altering the proportion of dietary protein-within limits (9 to 45 per cent casein) that support relatively normal growth and weight of the mouse-influences the genesis of some tumor types but not that of others. The spontaneous mammary carcinoma and the induced skin tumor are not significantly affected, On the other hand, formation of the spontaneous hepatoma of the mouse is strikingly enhanced by changing the proportion of protein from 9 to 18 per cent, whereas a similar increase in dietary protein appears to hinder the formation of liver tumors induced by azo dyes in the rat. It has been reported that diets deJicient in total protein or essential amino acids suppressed the genesis of certain tumors. Increasing the proportion of B vitamins as a group, from levels considered minimal for good growth to those several times greater than optimal, exerts no significant influence on the formation of spontaneous mammary or liver tumors and of induced skin tumors of the mouse. Diets deJicient in vitamins generally retard the genesis of neoplasms, but the question arises as to whether the actions are specific or are mediated through the associated lower caloric intake and body weight. A clearly defined action of dietary vitamins on carcinogenesis is that
NUTRITION IN RELATION TO CANCER
493
of riboflavin on hepatic tumors induced in rats by certain azo dyes. Low proportions of the vitamin, or dietary alterations which result in a low level of hepatic riboflavin, enhance the formation of this neoplasm. Another interesting finding is the development of liver tumors in rats chronically deprived of choline, without the administration of a known carcinogen. Varying the level of a complete salt mixture, between reasonable limits of “normal” intake, has no recognizable effect on the genesis of spontaneous mammary carcinoma and benign hepatoma and induced skin tumors. b. Growth of tumors in animals. Dietary alterations and the nutritional state of the host influence the growth of neoplasms but less impressively than they affect the genesis. With transplanted tumors it should be kept in mind that the nutritional state may affect the establishment of the implant as well as its subsequent growth. Experiments designed to study these separately will yield the most definitive results. A tumor grows even while the animal is losing weight, incfact, at the expense of the host’s normal tissues. Consequently, although caloric restriction inhibits the growth of a neoplasm (it also causes the host to lose weight) the life span of the animal is not appreciably lengthened. Variations in the proportions of dietary fat, protein, vitamins, or minerals, within the limits that are necessary for relatively normal growth and body weight of the host, have not been found to modify tumor growth. On the other hand, protein or vitamin deficiencies may inhibit the rate of growth of neoplasms, but the retardation may be mediated in large part through voluntary restriction of food intake and consequent loss of body weight. From the presently available evidence, it seems that no specific tumor-controlling influence has been demonstrated for diets deficient in fats, proteins, and minerals. For the most part this is also true for vitamins, but there are a few instances in which specific vitamin deficiencies appear to inhibit particular tumor types. c. Cancer in man. Clinical and statistical findings suggest that the genesis of some kinds of tumors in man is partly dependent on the nutritional state, that is, whether the individual is overweight or underweight. Other aspects of the dietary, particularly deficiencies, conceivably play a role in the development of some neoplasms. However, there is no indication that controlling the nutritional state is a practical means of arresting the growth of tumors. 2. Implications
The increased interest and work of the past two decades have resulted
in a broadened recognition of the relationship of nutrition to cancer. A
494
ALBERT TANNENBAUM AND HERBERT SILVERSTONE
good many of the obvious experiments have been performed so that we now know, perhaps only generally and crudely, the significance of the level of caloric intake and of the proportions of dietary fat, protein, vitamins, and salts, in the genesis and growth of certain tumors. One cannot predict beforehand the importance of the isolated facts and correlations that have been unearthed. In the future they may or may not be integrated with information from other fields to furnish practical working tools. One significant implication is the following: the inhibitory influence of calorie restriction and lower body weight also has connotations for many nonnutritional investigations in cancer research. Restriction of calories or deprivation of particular essential components of the diet may be a planned part of a study, but can occur without the intention or knowledge of the investigator. For example, animals subjected to chemotherapeutic agents (hormones, antagonists, chemicals, etc.) or t o irradiation may reduce their food intake or develop an increased need for particular dietary essentials. Intercurrent infections can cause diarrhea or loss of appetite. These and other circumstances could bring on subnormal food consumption, altered metabolism, or lower body weight. Should any of the above conditions, recognized or not, become an unplanned factor of a cancer experiment, the genesis or growth of tumors might be retarded. It then becomes necessary to ascertain whether the results are directly related to the procedure under investigation, or are due to the accompanying side effects of undernutrition and decreased body weight. Of course, caloric restriction and low body weight are not the common denominators of all inhibitory effects, but for more exact interpretation the nutritional state of the animals should be known. 3. Speculations on Nutrition in Relation to Carcinogenesis
Considering the large variety of diseases that are classified together as neoplasms, it is not expected that they would all respond to a particular dietary change in the same manner. Actually, with some diet components there has been a striking conformity of effects. In the case of others, equally impressive diversity exists. At the present time it is difficult to determine how much of this disparity is dependent on real differences in the biology and biochemistry of the tumors themselves. Certainly there are suggestions, mainly untested, that part of the observed differences may be attributed to differences in factors of the experiments, such as the nature of the carcinogen, its potency, and how it reaches the site of action. Another determinant may be the varying effects of the particular dietary alteration on the respective tissues being acted on by the carcinogens.
NUTRITION IN RELATION TO CANCER
495
The uncommon and diverse responses of tumors of the liver to experimental nutritional procedures is an arresting problem for a reviewer attempting to systematize and generalize. The anatomy and manifold functions of this organ dispose it t o multiple types of reactions. Its special blood supply (portal vein) and property as a depot result in an intracellular composition that fluctuates widely with diet changes. Moreover, it is very susceptible to the action of toxic agents and necrotizing conditions. Certain carcinogens may be formed in the organ, while others, some deliberately administered, reach it indirectly. When more is known about the various sets of events that induce hepatic neoplasms and how a particular set is modified by a change in the nutritional state of the host (and liver), many of the seemingly divergent effects may fit into a more consistent pattern. The knowledge of agents that induce tumors is extensive, but our understanding of the way they act upon the cells and the cell environment to result in neoplasms is meager. A dietary change, and the consequent alteration in the nutritional state of the host, may influence carcinogenesis in the following ways: (1)by modifying the solubility, rate of metabolism, metabolic products, or amount of the carcinogen reaching the target tissues (the effective tissue dosage of actual carcinogen); (2) by modifying the susceptibility of the target cells to tumor-initiating action; and (3) by modifying the development of the initiated, biased cells. These latter two influences may involve not only cells but their environment: cell surface, ground substance, stroma, and blood supply. Stated otherwise, nutritional effects may be mediated through changes in the dose of effective carcinogen, the initiatory stage, or the developmental stage. For any particular nutritional modification none, one, or more than one of these possibilities exists. The influence may be determined systemically or locally, but the end result must be a t the actual site of tumor genesis. What is the relative significance of the three ways, stated above, in which modifications of the nutritive state can affect the genesis of tumors? There is no single answer. In the origin of some kinds of neoplasms striking changes may be produced because the nutritional alteration modifies the amount of actual carcinogen reaching the site of action. Such situations may not be uncommon. It is likely that for many tumor types, dietary changes that do not seriously affect the health of the host only slightly modify the susceptibility of the target cells to the carcinogen. The initiation of neoplasms is probably influenced to only a small extent under these conditions. However, in the origin of some tumors, dietary deficiencies may play an important role in modifying and sensitizing the target cells so that they become more vulnerable to
496
ALBERT TANNENBAUM AND HERBERT SILVERSTONE
carcinogens. From evidence now available, we are of the opinion that for most tumors nutritional alterations have their main effects during the developmental stage of carcinogenesis. Whether these involve energy only or are intimately bound up with essential cellular constituents is only conjecture at this time. This review has been concerned with the influence of known and relatively welldefined components of diets. However, natural foods contain a number of constituents which have been given little attention in nutrition and cancer research because they aoparently are not dietary essentials. In addition, there must be others yet undetected. Perhaps among these unregarded substances are some with carcinogenic activity; and others that potentiate or oppose the action of carcinogens. There is danger here that findings may be coupled with suggestions and guesses to build up concepts which by pyramided repetition become accepted. That is not the purpose of the reviewers, who have great respect for facts and recognize the pitfalls of speculation.
4. Future Developments We have no intention of making predictions, but believe it might be helpful to state our general views as to potentially fruitful fields of nutrition-cancer research. In fact, these are some of the problems of current interest to us. Although the subject has been probed and examined in a general way, there are many specific dietary components that have not been investigated. Hunch and logic suggest that most of such studies would not yield startling results, but the information would help to complete the overall picture. Another area for study, hardly begun, is the relationship of nutrition to the spread, establishment, and growth of metastases. In the section concerned with neoplastic growth, reference has been made to a few investigations in which dietary alterations were combined with a second experimental procedure. Two-factor approaches of this kind intimate that combined measures may have synergistic influences on the genesis and growth of tumors. Recently there has been a revival of interest in social and environmental factors as causes of cancer. Chronic malnutrition may be one of these, possibly not acting alone but preparing the soil for greater vulnerability to infectious and toxic carcinogenic agents. Other nutritive states, alone or in combination with special factors, may be implicated. As suggestive clinical and experimental evidence regarding carcinogenesis comes to our attention, it should be extended by vigorous clinical studies of the nutritive state and pathogenesis. Investigations in this area may result in revealing the nature of the essential conditions
NUTRITION I N RELATION TO CANCER
497
of some types of human cancer. Already there are indications that a search for carcinogenic compounds in human dietary regimens might be worthwhile. Early nutrition-cancer research had to be content with fact finding, and this has been accomplished to a satisfactory degree. More and more now, attention is being turned to studying the mode of action of those diet alterations which affect tumor genesis. This work may proceed to a better understanding of carcinogenesis itself. During the period, years or decades, that it takes to solve substantially a problem in medical research, perspective is both biased and clouded. It is often hampered by understandable high hopes and by unsubstantiated sensational claims. Nevertheless, an appraisal of the many solid facts unearthed by nutrition-cancer research and the present trend toward study of mechanisms leads one to conclude that the field is contributing t o the comprehension of the cancer process.
ACKNOWLEDGMENT The investigations of the authors, providing a background for this review, were supported by grants from the National Cancer Institute of the National Institutes of Health, Public Health Service; the American Cancer Society upon recommendation of the Committee on Growth of the National Research Council; the Foundation for Cancer Research, Chicago; the Michael Reese Research Foundation; and others. REFERENCES Actuarial Society of America and Association of Life Insurance Directors. 1913. Medico-Actuarial Mortality Investigation. Spectator Co., New York. Actuarial Society of America and Association of Life Insurance Directors. 1932. Supplement to Medical Impairment Study. New York. Adair, F. E. 1947. Bull. N . Y . Acad. Med. 23, 383-93. Ahlbom, H. E. 193G. Brit. Med. J . 2, 331-33. Algire, G. H.,and Chalkley, H. W. 1945. J . Natl. Cancer Inst. 6 , 73-85. Ball, Z.B., Barnes, R., and Visscher, M. B. 1947. Am. J . Physiol. 160,511-19. Baumann, C. A. 1948. J . Am. Dietet. Assoc. 24,673-81. Baumann, C. A., Jacobi, H. P., and Rusch, H. P. 1939. Am. J . Hyg. A. 30, 1-6. Baumann, C. A., and Rusch, H. P. 1939. Am. J . Cancer 36, 213-21. Berenblum, I. 1941. Cancer Research 1, 807-14. Berenblum, I., and Shubik, P. 1949. Brit. J. Cancer 3, 109-18. Berman, C. 1940. 8.African J . Med. Sci. 5, 54-72. Berman, C. 1941. S. Africun J . Med. Sci. 6, 11-26. Berman, C. 1951. Primary Carcinoma of the Liver. H. K. Lewis, London. Bischoff, F., and Long, M. L. 1938. Am. J . Cancer 32, 418-21. Bischoff, F., Long, M. L., and Maxwell, L. C. 1935. A m . J . Cancer 24,549-53. Bonne, C. 1937. Am. J . Cancer 30,435-54. Boutwell, R. K.,Brush, M. K., and Rusch, H. P. 1948. Am. J . Physiol. 154,517-24. Boutwell, R. K., Brush, M. K., and Rusch, H. P. 1949s. Cancer Research 9,741-46. Boutwell, R. K., Brush, M. K., and Rusch, H. P. 1949b. Cancer Research 9,747-52. Bullough, W. S. 1950. Brit. J . Cancer 4,329-36.
498
ALBERT TANNENBAUM AND HERBERT SILVERSTONE
Bullough, W. S., and Eisa, E. A. 1950. Brit. J . Cancer 4, 321-28. Burk, D., Spangler, J. M., du Vigneaud, V., Kensler, C., Sugiura, K., and Rhoads, C. P. 1943. Cancer Research 8, 130-31. Burk, D., and Whaler, R. J. 1944. Vitamins and Hormones 2,305-52. Academic Press, New York. Caspari, W. 1938. Nutrition and Cancer. Hermann & Cie, Paris. Clayton, C. C., and Baumann, C. A. 1949. Cancer Research 9,575-82. Cole, W. H., Slaughter, D. P., and Rossiter, L. J. 1945. J . Am. Med. Assoc. 127, 883-88.
Cook, J. W., Duffy, E., and Schoental, R. 1950. Brit. J . Cancer 4, 405-10. Copeland, D. H., and Salmon, W. D. 1946. Am. J . Path. 22, 1059-79. Czaczkes, J. W., and Guggenheim, K. 1946. J . Biol. Chem. 162, 267-74. Devik, F., Elson, L. A., Koller, P. C., and Lamerton, L. F. 1950. Brit. J . Cancer 4, 298-3 14.
Dobrovolska~a-Zavadskaia,N. 1945. Compt. rend. soc. biol. 189, 494-95. Dublin, L. I. 1929. Proc. Assoc. Life Insur. Med. Dir. America 16, 402-11. Dublin, L. I., and Marks, H. H. 1938. Proc. Assoc. Life Insur. Med. Dir. America 24, 47-76.
Dublin, L. I., and Marks, H. H. 26,203-25.
1939. Proc. Assoc. Life Insur. Med. Dir. America
Dunning, W. F., Curtis, M. R., and Maun, M. E. 1949. Cancer Research 9,354-61. Elson, L. A., and Haddow, A. 1947. Brit. J . Cancer 1,97-101. Elson, L. A., and Lamerton, L. F. 1949. Brit. J . Cancer 3, 414-26. Engel, R. W. 1951. Cancer Research 11, 246. Engel, R. W., Copeland, D. H., and Salmon, W. D. 1947. Ann. N. Y . Acad. Sci. 49, 49-67.
Fisher, R. A., and Yates, F. 1943. Statistical Tables for Biological, Agricultural, and Medical Research, 2nd Ed. Oliver and Boyd, Ltd., London. Flory, C. M., Furth, J., hxton, J. A., Jr., and Reiner, L. 1943. Cancer Research
a, 729-43.
Forbes, E. B., Swift, R. W., James, W. H., Bratzler, J. W., and Black, A. 1946a. J . Nutrition 82, 387-96. Forbes, E. B., Swift, R. W., Thacker, E. J., Swift, E. F., and French, C. E. 1946b. 3. Nutrition 82, 397-403. Fridericia, L. S., Gudjonsson, S., Vimtrup, B., Clemmesen, S., and Clemmesen, J. 1940. Am. J . Cancer 89, 01-69. Friedewald, W. F., and b u s , P. 1950. J . Exptl. Med. 91, 469-84. Giese, J. E., Clayton, C. C., Miller, E. C., and Baumann, C. A. 1946. Cancer Research 6, 679-84. Gilbert, C., and Gillman, J. 1944. Science 09, 398-99. Gillman, J. 1944. Brit. Med. J . 1, 149-50. Gillman, J., Gillman, T., Mandelstam, J., and Gilbert, C. 1945. Brit. J . Ezptl. Path. 26, 67-81. Green, J. W., Benditt, E. P., and Humphreys, E. M. 1950. Cancer Research 10, 769-74.
Green, J. W., and Lushbaugh, C. C. 1949. Cancer Research 9, 092-700. Greenstein, J. P. 1947. Biochemistry of Cancer. Academic Press, New York. Griffin, A. C., and Baumann, C. A. 1946. Arch. Biochem. 11, 467-76. Griffi, A. C., and Baumann, C. A. 1948. Cancer Research 8, 279-84. Griffin, A. C., Clayton, C. C., and Baumann, C. A. 1949. Cancer Research 9,82-87. Haddow, A. 1935. Nature 186,868.
NUTRITION IN RELATION TO CANCER
499
Harris, P. N. 1947. Cancer Research 7, 88-94. Harris, P. N., Krahl, M. E., and Clowes, G. H. A. 1947. Cancer Research 7 , 162-75. Hoch-Ligeti, C. 1951. Acta Unio Intern. contra Cancrum 7 , 606-11. Hoffman, F. L. 1937. Cancer and Diet. Williams & Wilkins Co., Baltimore, Md. Hunter, A. 1939. Trans. Actuarial SOC.America 40, 394-411. Huseby, R. A., Ball, Z. B., and Visscher, M. B. 1945. Cancer Research 6, 40-46. Jacobi, H. P., and Baumann, C . A. 1940. Am. J. Cancer 39,338-42. Kennaway, E. L. 1944. Cancer Research 4, 571-77. Kensler, C. J. 1947. Ann. N . Y. Acad. Sci. 49, Art. 1, 29-40. Kensler, C. J. 1948. Cancer 1, 483-88. Kensler, C. J., Sugiura, K., and Rhoads, C . P. 1940. Science 91, 623. Kensler, C. J., Sugiura, K., Young, N. F., Halter, C. R., and Rhoads, C. P. 1941. Science 93, 308-10. King, C. J., and 15 others. 1947. Ann. N . Y. Amd. Sci. 49, Art. 1, 1-140. King, J. T., Casas, C. B., and Visscher, M. B. 1949. Cancer Research 9, 436-37. Kirschbaum, A., Geisse, N. C., Judd, T., and Meyer, L. M. 1950. Cancer Research 10, 762-68. Kline, B. E., Miller, J. A., and Rusch, H. P. 1945. Cancer Research 6, 641-43. Kline, B. E., Miller, J. A., Rusch, H. P., and Baumann, C. A. 1946. Cancer Research 6, 5-7. Kline, B. E., and Rusch, H. P. 1944. Cancer Research 4, 762-67. Kline, B. E., Rusch, H. P., Baumann, C. A., and Lavik, P. S. 1943. Cancer Research 3, 825-29. Kocher, R. A. 1944. Cancer Research 4, 251-56. Larsen, C. D., and Heston, W. E. 1945. J . Natl. Cancer Znst., 6 , 31-40. Lavik, P. S., and Baumann, C. A. 1941. Cancer Research 1, 181-87. Lavik, P. S., and Baumann, C. A. 1943. Cancer Research 3, 749-56. Lawrason, F. D., and Kirschbaum, A. 1944. Proc. SOC.Exptl. Biol. Med. 66, 6-7. Little, P. A., Oleson, J. J., and SubbaRow, Y. 1948. J . Lab. Clin. Med. 33, 113943. MacKenzie, I., and Rous, P. 1941. J. Exptl. Med. 73, 391-416. McCay, C. M., Crowell, M. F., and Maynard, L. A. 1935. J. Nutrition 10, 63-79. McCay, C. M., Ellis, G. H., Barnes, L., Smith, C. A. H., and Sperling, G. 1939. J. Nutrition 18, 15-25. Miller, E. C., Baumann, C. A., and Rusch, H. P. 1945. Cancer Research 6,713-16. Miller, J. A. 1947. Ann. N . Y . Acad. Sci. 49, Art. 1, 19-28. Miller, J. A., Kline, B. E., Rusch, H. P., and Baumann, C. A. 1944a. Cancer Research 4, 153-58. Miller, J. A,, Kline, B. E., Rusch, H. P., and Baumann, C. A. 1944b. Cancer Research 4, 756-61. Miller, J. A., Miner, D. L., Rusch, H. P., and Beumann, C. A. 1941. Cancer Research 1,699-708. Miner, D. L., Miller, J. A., Baumann, C. A., and Rusch, H. P. 1943. Cancer Research 3, 296-302. Moreschi, C. 1909. 2. Zmmunitatsforsch. 2, 651-75. Morris, H. P. 1945. J. Natl. Cancer Znst. 6, 1-17. Morris, H. P. 1947. Ann. N . Y. Acad. Sci. 49, Art. I, 119-40. Morris, H. P., and Voegtlin, C. 1940. J . Biol. Chem. 133, lxix-lxx. Mottram, J. C. 1944. J. Path. Bact. 66, 181-87. Mueller, G. C., and Miller, J. A. 1950. J . Biol. Chem. 186, 145-54. Mulinos, M. G., and Pomerantz, L. 1940. J . Nutrition 19, 493-504.
500
ALBERT TANNENBAUM AND HERBERT SILVERSTONE
Opie, E. L. 1944. J . Exptl. Med. 80, 219-30. Opie, E. L. 1947. Approaches to Tumor Chemotherapy, pp. 128-38. Science Press, Lancaster, Pa. Pack, G. T., and LeFevre, R. G. 1930. J . Cancer Research 14, 167-294. Passey, R. D., Leese, A., and Knox, J. C. 1935. J . Path. Bact. 40, 198-99. Robertson, W. van B., Dalton, A. J., and Heston, W. E. 1949. 'J. Natl. Cancer Znst. 10, 53-60. ROUE,P. 1914. J. Expll. Med. 20, 433-51. &us, P., and Kidd, J. G. 1941. J . Exptl. Med. 78,365-90. Rusch, H. P., Baumann, C. A., and Kline, B. E. 1939. Proc. SOC.Expll. B i d . Med. 42, 508-12. Rusch, H. P., Baumann, C. A., Miller, J. A., and Kline, B. E. 1945a. Research Conference on Cancer, pp. 267-90. American Association for the Advancement of Science, Washington, D.C. Rusch, H. P., Johnson, R. O., and Kline, B. E. 1945b. Cancer Research 6, 705-12. Rusch, H. P., Kline, 13. E., and Baumann, C. A. 1945c. Cancer Research 6,431-35. Sarett, H. P., and Perlzweig, W. A. 1943. J. Nutrition 26, 173-83. Saxton, J. A., Jr. 1941. N.Y. Slate J . Med. Abatr. 41, 1095-96. Saxton, J. A., Jr., Boon, M. C., and Furth, J. 1944. Cancer Research 4, 401-09. Sayers, G. 1950. Physiol. R&s. 30, 241-320 (p. 293). Shapiro, D. M., and Gellhorn, A. 1951. Cancer Research 11, 35-41. Shear, M. J. 1933. Am. J . Cancer 18,924-1024. Silverstone, H. 1948. Cancer Research 8, 301-08. Silverstone, H., and Tannenbaum, A. 1949. Cancer Research 9, 684-88. Silverstone, H., and Tannenbaum, A. 1950. Cancer Research 10, 448-53. Silverstone, H., and Tannenbaum, A. 195111. Cancer Research 11, 200-03. Silverstone, H., and Tannenbaum, A. 1951b. Cancer Research 11, 442-46. Stern, K., and Willheim, R. 1943. The Biochemistry of Malignant Tumors. Reference Press, New York. Stoerck, H. C., and Emerson, G. A. 1949. Proc. SOC.Exptl. Biol. Med. 70, 703-04. Sugiura, K. 1944. Proc. SOC.Exptl. B i d . Med. 67, 231-34. Sugiura, K. 1951. J . Nutrition 44, 345-60. Sugiura, K., and Benedict, S. R. 1926. J. Cancer Research 10, 309-18. Hugiura, K., and Benedict, S. R. 1930. J . Cancer Research 14, 311-18. Tannenbaum, A. 1940a. Am. J . Cancer 38,335-50. Tannenbaum, A. 1940b. Arch. Path. 80,509-17. Tannenbaum, A. 1942a. Cancer Research 2,460-67. Tannenbaum, A. 1942b. Cancer Research 2, 468-75. Tannenbaum, A. 1944a. Cancer Research 4, 073-77. Tannenbaum, A. 1944b. Cancer Research 4, 678-82. Tannenbaum, A. 1944c. Cancer Research 4, 683-87. Tannenbaum, A. 1945a. Cancer Research 6,609-16. Tannenbaum, A. 1945b. Cancer Research 6, 616-25. Tannenbaum, A. 1947. Approaches to Tumor Chemotherapy, pp. 96-127. Science Press, Lancaster, Pa, Tannenbaum, A., and Silverstone, H. 1947. Cancer Research 7 , 567-74. Tannenbaum, A,, and Silverstone, H. l949a. Cancer Research 0, 162-73. Tannenbaum, A., and Silverstone, H. 194913. Cancer Research 9, 403-10. Tannenbaum, A., and Silverstone, H. 1949c. Cancer Research 9, 724-27. Tannenbaum, A., and Silverstone, H. 1950. Cancer Research 10, 577-79. Taylor, A., and Williams, R. J. 1945. Univ. Texas Pub. 4607, 119-22.
NUTRITION IN RELATION TO CANCER
501
Visscher, M. B., Ball, Z. B., Barnes, R. H., and Sivertsen, I. 1942. Surgery 11, 48-55. Voegtlin, C., Johnson, J. M., and Thompson, J. W. 1936. U.S. Pub. Health Repts. 61, 1689-97. Voegtlin, C., and Maver, M. E. 1936. U.S. Pub. Health Repts. 61, 1436-44. Voegtlin, C., and Thompson, J. W. 1936. U.S. Pub. Health Repts. 61, 1429-36. Voegtlin, C., and Thompson, J. W. 1949. J . Natl. Cancer Znst. 10, 29-52. Ward, R. 1944. Surgery 16, 783-803. Waterman, N. 1938. Diet and Cancer: An Experimental Study. D. B. Centen’s Uitgevers-Maatschappij, Amsterdam. Watson, A. F., and Mellanby, E. 1930. Brit. J . Exptl. Path. 11, 311-22. Wegelin, C. 1928. Cancer Rev. 3, 297-313. White, F. R., and Belkin, M. 1945. J . Natl. Cancer Znst. 6, 261-63. White, F. R., and White, J. 1944a. J . Natl. Cancer Znst. 4, 413-15. White, F. R., and White, J. 1944b. J . Natl. Cancer Znst. 6, 41-42. White, F. R., and White, J. 1946. J . Natl. Cancer Znst. 7, 99-101. White, F. R., White, J:, Mider, G. B., Kelly, M. G., and Heston, W. E. 1944. J . Natl. Cancer Znst. 6 , 43-48. White, J., and Andervont, H. B. 1943. J . Natl. Cancer Inst. 3, 449-51. White, J., White, F. R., and Mider, G. B. 1947. J . Natl. Cancer Znst. 7 , 199-202.
This Page Intentionally Left Blank
Plasma Proteins in Cancer RICHARD J. WINZLER Department of Biological Chemistry, University of Illinois College of Medicine, Chicago, 111. CONTENTS
I. Some Methods of Study of Plasma Proteins
Salt Fractionation Electrophoresis Fractionation with Alcohol at Low Temperature and Ionic Strength Other Methods 11. Alterations of Plasma Proteins in Neoplastic Disease 1. Albumin A. General Considerations B. Thermal Coagulation Procedures C. Polarographic Serum Test D. Binding of Anionic Dyes E. Methylene Blue Reduction 2. Alpha Globulins 3. Especially Soluble and Stable Proteins A. Polypeptidemia B. Polarographic Filtrate Test C. Mucoprotein D. Albumin A 4. Beta Globulins 5. Fibrinogen A. Sedimentation Rate B. Heat Turbidity Test 6. Gamma Globulins A. Amounts and Immunity B. Flocculation with Lipoidal “Antigens ” C. Multiple Myeloma 7. Changes in Protein Stability to Precipitating Agents 111. Plasma Enzymes and Inhibitors 1. Enzymes A. Acid and Alkaline Phosphatases B. Aldolase c. D-Peptidases D. Carcinolysis E. Plasmin F. Fuchs’ Test G. Abwehrferment 2. Enzyme Inhibitors A. Inhibition of Proteolytic Eneymes 1. 2. 3. 4.
SO3
Page 506 506 507 511 “13 513 514 514 516 518 518 519 520 521 522 522 523 524 524 524 525 525 525 525 526 527 528 529 529 529 530 530 530 531 53 1 532 532 532
RICHARD J. WINZLER
504
Page B. Inhibitions of Hyaluronidase . . . . . . . . . . . . . . . ..................... 534 534 C. Inhibitors of Oxidiiing Enzymes.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . IV. Protein-Bound Carbohydrate. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 535 1. Total Polysaccharide. ........................ 535 2. Tryptophan-Acid ........................................ 537 .. . . . . . . . . . . . . 538 3. Diphenylamine R V. Discussion.. . . . . . . . . . . . . . . . . . . . . . . . 538 References . . . . . . . . . . . . . . . . . . . . . . . . . . . 539
One of the enigmas of the cancer problem is the mechanism by which neoplastic disease may kill a host when no apparent interference with a vital structure is evident. That cancer produces systemic effects in the host is well known, but there is little information on the profound changes in physiology which must be a t the basis of such systemic effects. It has also been recognized that there are abnormalities in the protein composition of the blood. However, the nature of the changes in the plasma proteins and their physiological and pathological bases are still largely unanswered problems. There has been a growing knowledge of the numbers and properties of the plasma proteins. Yet we are only beginning to understand something of the source, the significance, and the metabolic activity of these agents. It is presumptive indeed, to attempt a discussion of the relation of plasma proteins to cancer when so little is yet known about fundamental aspects of plasma proteins in the normal individual. There may be, however, some value in reviewing what is known about the relationship of plasma proteins to malignant disease. It is in this humble vein that this discussion is written. In this respect this discussion may be considered a supplement to the reviews by Toennies (1947), Huggins (1949), and Gutman (1948). Some aspects of the problem of the serodiagnosis of cancer are related to changes in plasma proteins and will be briefly considered. In Table I are listed some of the changes in plasma proteins which TABLE I Some Changea in Plmma Proteins Associated with Neoplastic Disease in Humans Abnormality in Cancer Decreased albumin content Increased alpha-1 globulin content Increased alpha-2 globulin content Increased fibrinogen content
Page No. 507, 514, Table I1
Sample References
Mider et al. (1950) Seibert et al. (1947) 508, 520, Table I1 Mider el al. (1950) Seibert el al. (1947) 508, 520, Table I1 Mider et al. (1950) Seibert et al. (1947) 508, 524, Table I1 Mider et al. (1950)
PLASMA PROTEINS IN CANCER
505
TABLE I (Continued) Abnormality in Cancer Decreased amounts of polarographically determinable- SS i r -SH groups Decreased amounts of free SH groups determined amperometrically Decreased “reducing power” when heated with methylene blue Decreased tendency toward thermal coagulation Decreased amount of iodoacetate required to prevent thermal coagulation Increased tendency of plasma to become turbid on being heated Reduced ability to bind anionic dyes
Page No.
Sample References
518, Table IV BrdiEka (1937, 1939) 517, Table IV Schoenbach et al. (1950, 1951) 519, Table IV Savignac et al. (1945) Black (1947a,b) 516, Table IV Glass (1951); Huggins et al. (1949b) 516, Table I V Huggins et at. (1949a,b)
Black et al. (1948) 525 518, Table IV Huggins et al. (1949a) Westphal et al. (1951) 535, Table VI Shetlar et al. (1949, 1950) Seibert et al., (1947, 1948)
Increased total protein-bound carbohydrate content Increased total protein-bound hexosamine 523, 536, Table V Increased mucoprotein content Increased amounts of highly soluble 522, 523, Table V and stable proteins determined polarographically or chemically Decreased tendency to gel upon treatment with lactic acid Increased aldolase content
528, Table V 530
Increased acid or alkaline phospha tase content (metastatic cancer in bone)
529
Increased activity of inhibitors of hyaluronidase Increased activity of inhibitors of proteolytic enzymes Increased tendency toward flocculation by lipoidal “antigens”
534
Decreased amount of heat coagulable soluble 11 albumins ” Increased erythrocyte sedimentation rate
532 526
524 525
,
West and Clarke (1938) Winzler and Smyth (1948) Wolff (1921); Winzler and Burk (1944) Hahn (1921); BrdiEka (193913) Kopaczewski (1935, 1937) Warburg and Christian (1943) Sibley and Lehninger (1949) Gutman and Gutman (1938) Kay (1929, 1930) Huggins and Baker (1951) Hakanson and Glick (1948) Cliffton (1949, 1950) West el al. (1949a, 1951c) Shaw-Mac Kenzie (1922) Gruskin (1929) Penn et al., (1950) Kahn (1925) Gram (1922); Gilligan et al. (1950)
506
RICHARD J. WINZLER
have a t one time or another been studied in connection with neoplastic disease in humans. Some of these involve well-characterized changes in concentration or in activity of specific proteins and rest on considerable fundamental research effort. Others, regrettably, have been studied entirely from an applied point of view, and little effort has been made to discover the fundamental bases of the changes noted. Almost without exception the alterations listed have no specificity with regard to generalized cancer. Table I emphasizes the fact that advanced neoplastic disease is associated with a multitude of changes in the amounts and activities of a large number of plasma proteins, or in phenomena associated with their presence. Many of these abnormalities have been discussed in previous review articles on serodiagnosis of cancer (Bing and Marangos, 1934; Davidsohn, 1936; Woodhouse, 1940; Stern and Willheim, 1943; Maver, 1944; Homburger, 1950; and Sobotka, 1951).
I. SOMEMETHODSOF STUDYOF PLASMA PROTEINS A number of the proteins in plasma are present in rather large amounts and may be designated as the major components. The amounts of some of these proteins or of related groups have been determined by a number of methods, chief among which are fractional precipitation by neutral salts, fractional precipitation by ethanol a t low ionic strength and low temperature, and electrophoretic analysis. The application of these methods to the study of the plasma proteins in disease has been discussed in an excellent review by Gutman (1948). 1. Salt Fractionation One of the most extensively used methods for the study of plasma proteins has been by fractional salting out with ammonium sulfate or sodium sulfate. A considerable amount of the work up to 1948 on the procedure and its significance has been discussed by Gutman (1948) and by Edsall (1947). The bulk of the information that has been gathered on plasma protein fractions in cancer and in other pathological states has been obtained by the method of Howe (1921) using sodium sulfate fractionation. In this procedure, albumin is defined as the protein left in solution at 21.5% sodium sulfate, and the globulins are designated as the proteins insoluble a t this salt concentration. This separation is based on protein solubilitysalt concentration curves obtained with plasma in which a critical concentration zone between 21 and 22% sodium sulfate produced little increase in precipitated protein. Further fractionation of the globulins into euglobulin and pseudoglobulin was developed by Howe (1921) based
PLASMA PROTEINS IN CANCER
507
on inflections in the salt concentration-protein solubility curves a t 14 95 and 16.9% sodium sulfate. It is now recognized that the Howe procedure gives protein fractions which are electrophoretically inhomogeneous, with albumin values higher than those determined electrophoretically (Tiselius, 1937b ; Luetscher, 1940; Gutman et al., 1941; Svensson, 1941; Taylor and Keys, 1943; Dole, 1944; Majoor, 1946, 1947; Milne, 1947; Petermann et al., 1947; Martin and Morris, 1949; Jager et al., 1950). The albumin fraction contains considerable quantity of alpha and beta globulins, and may give albumin values that are 20 to 40% higher than values obtained electrophoretically. This may be particularly marked in pathological samples where albumin is low and the alpha globulins are raised. The distribution of the plasma proteins as measured by the Howe method shows a characteristic trend toward lowered albumin values and normal or even elevated globulin values in a large number of diseases including neoplasia. A number of other procedures for the separation of albumin and globulin by salt fractionation have been developed which are improvements over the Howe method, e.g., 26% sodium sulfate (Milne, 1947; Majoor, 1946, 1947), 26.9% sodium sulfite (Cohn and Wolfson, 1947, 1948), saturated magnesium sulfate (Popjak and McCarthy, 1946), 21.5% sodium sulfate in the presence of ether (Kingsley, 1940). 6. Electrophoresis
The development of moving boundary electrophoretic methods by Tiselius (19374 has resulted in considerable advance in knowledge of the plasma proteins. This procedure makes use of the rate of migration of proteins in an electrical field, the boundaries between components being visualized from changes in refractive index as a function of the distance from the starting point. The hope that plasma electrophoretic patterns might be characteristic of specific diseases, and thus be of great diagnostic significance, has not been generally supported by the large amount of work that has now been carried out. A large number of apparently unrelated pathological conditions result in very similar abnormalities in electrophoretic patterns of plasma. However, the electrophoretic method applied to plasma has given a great deal of accurate information on the changes in levels of albumin, of alpha, beta, and gamma globulins, of fibrinogen, and of the occasional presence of abnormal proteins in plasma in various diseases. The application of electrophoresis to the study of plasma proteins has been reviewed by Stern and Reiner (1946), by Luetscher (1947), and by Gutman (1948). Typical electrophoretic patterns obtained with normal and cancer
508
RICHARD J. WINZLER
serum in pH 8.5 veronal buffer a t ionic strength 0.1 are shown in Fig. 1. The lowered albumin and raised alpha globulin and fibrinogen levels in cancer plasma shown in Fig. 1 are in general agreement with the observations of numerous investigators who have made this comparison (Longsworth el al., 1939; Luetscher, 1941; Seibert el al., 1947; Petermann and Hogness, 1948a; Petermann el al., 1948; Dillard el al., 1949; Mider et al., 1950; Schoenbach et al., 1951). I n the most extensive of these studies Mider, Alling, and Morton (1950) have studied electrophoretically the distribution of plasma proteins in 258 patients with uncomplicated neoplastic disease under the most commonly employed electrophoretic conditions (pH 8.5 veronal buffer
FIQ.1. Electrophoresisof serum at pH 8.5 in veronal buffer at ionio strength 0.1. Ascending boundary. Left pattern normal, right pattern mammary carcinoma. (Courtesy of Dr. J. W. Mehl.)
a t ionic strength 0.10 (Longsworth, 1942)). Some of the data obtained by Mider e l al. (1950) is shown in Table 11. The data show a highly significant trend toward hypoalbuminemia, and equally significant increases in alpha-1 globulin, in alpha-2 globulin, and in fibrinogen as the severity of the disease progresses. Definite but less significant increases in beta and gamma globulins occurred. The increase in the total globulin was less than the decrease in albumin, resulting in a slight but definite tendency toward hypoproteinemia in the cancer patients. These changes, however, are not sufficiently characteristic of neoplastic disease to distinguish it from other ailments. One of the more important contributions of the electrophoretic method to the study of plasma proteins has been t o identify or characteril;e plasma components separated by other methods. The demonstration of the heterogeneity of the Howe fractions, for example, rests to a great extent on the presence of a number of electrophoretic components
cd I?
TABLE I1 Electrophoretic Components in Plasma of Normal Individuals and Cancer Patients (From Mider et al., 1950)
Normal adulta Cancer Advanced cancer 1
Standard deviation.
Total Protein g. %
Albumin g. %
6.83 f .07' 6.60 f .04 6.55 .14
4.04 k .05 2.94 k .04 2.38 .05
+
+
Alpha-1 Globulin g. %
Alpha-2 Globulin g. %
Beta Globulin g. %
0.38 f .01
0.66 f .02 0.90 .01 1.05 +_ .05-
0.76 k .02 0.89 f .01 0.99 rt .05
0.53 k .01 0.65 f .03
b-
m
Fibrinogen B- % 0.31
.02
0.58 k .01 0.82 +_ .09
Gamma Globulin g.
%
F
%
a
0.66 f .03 0.75 4 .02 0.82 f .04
ii
P
510
RICHARD J. WINZLER
in each of these fractions (Tiselius, 1937b; Gutman et al., 1941; Svensson, 1941; Dole, 1944; Majoor, 1946, 1947; Milne, 1947; Petermann et al., 1947). The designation, identity, and homogeneity of the various fractions obtained by the low temperature fractionation methods developed by Cohn and his associates (Cohn et al., 1946, 1950; Edsall, 1947), also rests to a large extent upon their electrophoretic behavior. It has become a fairly well-standardized procedure to carry out electrophoretic examination of plasma in veronal buffer a t pH 8.5 at ionic
Alb
I
t
-
t
FIQ.2. Electrophoresisof serum at pH 4.5 in acetate buffer at ionic strength 0.1. Descending boundary. Right pattern normal, left pattern mammary carcinoma. (Courtesy of Dr. J. W. Mehl.)
strength 0.1 following the demonstration of Longsworth (1942) that favorable resolution of the plasma proteins occurred under these conditions. The advantages of standardization of conditions are obvious, but the possibility that additional information may be obtained by carrying out electrophoresis at other pH values should be emphasized. Thus Petermann and Hogness (1948b) demonstrated an acid component with an isoelectric point lower than p H 4 which occurred in increased amounts in the plasma of patients with gastric carcinoma. Mehl et al. (1949a, 1950) using acetate buffer a t pH 4.5 showed that this acid component is identical with a plasma mucoprotein isolated by Winder et al. (1948) and by Weimer et al. (1950). An additional acidic component (designated M-2) was demonstrable in pathological sera examined electrophoretically a t pH 4.5 (Mehl et al., 194Ya, 1950). Figure 2 shows patterns from serum from a normal individual and a patient with cancer examined under these conditions. Miller et al. (1950) have made an extensive study of the electrophoretic patterns obtained with plasma over the pH range 3.0 to 11.4. These
PLASMA PROTEINS IN CANCER
511
studies may prove especially important when extended to pathological sera. The cost of the equipment and the time required for individual runs has prevented routine application of electrophoresis t o clinical problems. However, it now seems possible that electrophoresis on filter paper strips may provide a rapid and convenient method for determination of the electrophoretic components of serum. Proteins from 0.01 t o 0.1 ml. of serum separate into distinct electrophoretic components on filter paper wet with buffer across which an electrical potential is applied (e.g., Cremer and Tiselius, 1950; Durram, 1950; Turba and Enenkel, 1950; Grassman, 1951). The proteins may be fixed and stained at the end of the run or may be extracted and determined chemically. Awapara et al. (unpublished) have determined the relative amounts of albumin, alpha-1 globulin, alpha-2 globulin, beta globulin, and gamma globulin of sera from normal individuals and cancer patients by this method and have found the characteristic decrease in albumin and increase of alpha-1 and alpha-2 globulins already discussed.
3. Fractionation with Alcohol at Low Temperature and Ionic Strength E. J. Cohn and his associates have developed during the course of the last few years several methods for the separation of plasma proteins, using alcohol at low temperature and low ionic strength. This work has been almost entirely confined to the isolation and characterization of proteins present in normal human plasma, and little application has yet been made to the study of plasma from patients with various pathological conditions. This very extensive program has been reviewed in a number of papers (Cohn, 1941, 1948; Edsall, 1947; Oncley, 1950). About thirty protein components of plasma have been separated and partially characterized by the Harvard Group (Table 111). Recently, a modified procedure (method 10) involving fractional extraction, protein-protein and protein-heavy metal interactions has been developed (Cohn et al., 1950). This procedure is of special interest here since it lends itself to rapid analytical or preparative procedures on a small scale. Method 10 has been modified to the use of filtration instead of centrifugation (Lever et al., 1951), and the distribution of protein, of cholesterol, of phospholipid, and of protein-bound carbohydrate in several fractions has been carried out with normal plasma. There is little doubt that extension of this work to pathological plasmas will be of utmost value in future studies of the plasma proteins in cancer and other pathological states. Pillemer and Hutchinson (1945) have used methanol (42.5 per cent at 0°C. and pH 6.7 to 6.9) to separate albumin and globulin, the albumin
5 12
RICHARD J. WINZLER
TABLE I11 Protein Components of Human Plasma and Certain of Their Chemical Properties and Interactions'
Protein Component
Estimated % of Plasma Proteins
orl-Acid glycoprotein Caeruloplasmin Choline esterase a 1 - B i i b i n globulin Serum albumins Mercaptalbumin or TGlycoproteins c u r Mucoproteins Fibrinogen Cold insoluble globulin Antihemophilic globulin al-Lipoproteins &-Lipoproteins
Sedimen- Approx. Isotation Constant electric Sz0, W Point
0.5
-
0.005 0.05 52 (34) 1.2 0.5 4 0.15
-
3. 5.
&-Lipid-poor euglobulins Bl-Metal combining protein Isoagglutinins 82-Globulins
3 (0.03) 3
r-Globulins
11
Accelerator globulin Prothrombin Heparin complement Plasminogen Plasmin inhibitor Hypertensinogen Iodoproteins Complement component CI1 C'2 Amylase Alkaline phosphatase Peptidase &Glucuronidase aTProtein &Protein 1
-3 . 6 -
4.6
9 9 9
-
5 7 7 20 5.0
7
3.0 4.4 4.5 4.7 4.9
< 5.3
-
-
5 . 8 Iron & copper 6 . 3 Incompatible red cells 6.3 {7.3 6.3 Antigens
-
-
Prothrombin Calcium & thromboplastin Heparin Streptokinase Plasmin Tennin
-
Antigen-antibody complex
-
-
0.1 0.05
From Onoley (1960) and Cohn at d. (1960).
-
-
-
2.9
6.
-
5 . 2 Steroids & carotenoids 5 . 4 Steroids & carotenoids 5.5
-
-
Copper Choline esters Bilirubin Fatty acids-bile salts - Dyes, drugs & mercury 4 . 9 Carbohydrates & barium 4 . 9 Carbohydrates & barium < 5 . 3 Thrombin
-
4
Specific Chemical Reactions
-
-
-
-
Starch Phosphate esters c l e wylglycylglycine 8-Glucuronides Barium
-
PLABMA PROTEINS I N CANCER
513
globulin ratios being lower than those determined by the Howe method and corresponding closely to those obtained electrophoretically. Using this method Nitsche and Cohen (1947) have found subnormal albumin levels in patients with leukemia or Hodgkins disease.
4. Other Methods The ultracentrifugal study of plasma proteins has been of considerable value in evaluation of homogeneity and molecular weights of partially purified protein preparations. Three or four major ultracentrifugal components are discernible when normal or pathological human serum is examined (McFarlane, 1935; Pedersen, 1945; and Oncley et al., 1947). Too little work has been carried out to determine whether abnormal ultracentrifugal patterns occur with any frequency in the plasma of patients with neoplastic disease. The great specificity and sensitivity of immunological methods for the detection and identification of specific proteins can be expected to be of considerable value in future work on plasma proteins in health and disease. Especially useful in this connection has been the development of quantitative immunological procedures for the determination of specific proteins through the preparation of antibodies to these proteins in rabbits and subsequently determining the nitrogen content of antigenantibody precipitates (Heidelberger, 1939; Kabat, 1943; Treffers, 1944 ; and Jager et al., 1950). Studies of plasma proteins which have quantitatable biological activities (other than the antibodies mentioned above) may also be used to characterize differences in blood between normal and pathological states. The activity of specific plasma enzymes (Homburger, 1950) and the activity of inhibitors of enzyme action (Winzler, 1950) may be significantly altered in neoplastic disease. The presence of certain proteins which are characterized by chemical tags may also be altered in the plasma of cancer patients. Examples of this type are the polysaccharide-containing proteins of plasma, the levels of which rise markedly in cancer and in other pathological states (see later discussion). Many changes in the physicochemical characteristics of plasma in relation to malignant disease have been shown t o occur. Such changes as decreased tendency to coagulate by treatment with heat or acid and decreased ability to bind anionic dyes will be discussed in later sections.
11. ALTERATIONS OF PLASMA PROTEINS IN NEOPLASTIC DISEASE It has been brought out in the preceding discussion that there are abnormalities in the absolute and relative amounts of the major protein
514
RICHARD J. WINZLER
components of plasma of patients with neoplastic disease. There is, in addition, considerable evidence that even more pronounced alterations in amount or kind of minor plasma components may be associated with this disease. Many of these studies have been stimulated by the desire to develop serodiagnostic procedures for detection of cancer. While some success in this direction has frequently been reported, extensions of such studies have generally led to the conclusion that the changes are not characteristic of cancer, but are shared with other wasting diseases and infections. Nonetheless, the demonstration that such alterations and abnormalities occur regularly in neoplastic disease is sufficient reason for the pursuit of their fundamental physiological and biochemical bases. In the following discussion a few of the procedures which have been proposed for the serodiagnosis of cancer will be considered, as well as some of the observations which have been studied for their biological interest alone. It is the purpose of this discussion to consider the procedures from the point of view of their biochemical significance, rather than from the point of view of serodiagnosis of cancer. Some of the diagnostic procedures have been reviewed by Homburger (1950), Sobotka (1951), Maver (1944), Stern and Willheim (1943), Woodhouse (1940), Davidsohn (1938), Bing and Marangos (1934), and Kahn (1925).
1. Albumin A. General Considerations. The general decrease in plasma albumin in malignant disease is well established by both salt fractionation and electrophoretic methods. The possibility exists that this decrease is the result of a negative nitrogen balance which is especially reflected in albumin metabolism. However, it has been demonstrated that low serum albumin levels may persiet under conditions where an overall positive nitrogen balance is maintained (e.g. , Homburger and Young, 1948; Mider et al., 1948). This demonstration, however, would not dispose of the possibility that the growth of the tumor and the consequent withdrawal of nitrogen from the body pools may result in a net loss of protein to the patient, in spite of an overall nitrogen gain in tumor plus host. It seems likely, however, that the serum albumin level is reduced more severely in cancer than are the other protein stores of the body. The physiological basis of the reduced serum albumin in malignant disease is, thus, a, pressing problem. An hepatic dysfunction induced in some unknown manner by the presence of a neoplasm may be the basis for the lowered serum albumin levels in cancer patients. That the liver is the major site of albumin synthesis seems well established (Madden and Whipple, 1940). Abels et al. (1943), in reviewing the work of the Memorial Hospital group on
PLASMA PROTEINS IN CANCER
515
metabolic abnormalities in patients with cancer of the gastrointestinal tract, reported lowered serum albumin levels in a large proportion of their patients and showed that this lowered albumin level bore no relation to the degree of dietary deficiency or to loss of blood by internal bleeding. Using several criteria for adequacy of liver function in control individuals and in patients with gastrointestinal cancer, they showed that hepatic function was defective in a large proportion of the cancer patients in comparison with individuals with atrophic gastritis or oral leukoplakia. This suggested that defective protein fabrication in the liver of the cancer patients was a major cause of hypoalbuminemia. In support of this suggestion, Ariel et al. (1943) and Ariel (1949) found that the administration of glycine t o patients with gastrointestinal cancer resulted in a significantly higher and more persistent level of free amino acid in the blood than was the case with normal controls, an observation which suggested that the synthesis of protein from free amino acid might be reduced in the cancer patients. On the other hand Norberg and Greenberg (1951) found that the rate of uptake of C14-labeledglycine into the plasma proteins of mice was increased above normal rates in mice bearing transplanted tumors. If the suggestion is substantiated that hepatic dysfunction is associated with cancer of various types, the mechanism by which such an influence might be brought about would become a topic of considerable importance. It is reasonably well established that tumors may have effects on enzyme systems remote from the site of the tumor. Thus Greenstein et al. (1941) have shown that the liver catalase activity of rats and mice bearing transplanted tumors is much decreased from the control levels and that removal of the tumor results in the rapid return of liver catalase activity to normal. Recently an agent has been isolated from tumor tissue which caused significant decrease in liver catalase activity when injected into normal mice (Nakahara and Fukuoka, 1949; and Greenfield and Meister, 1951). Recently Madden (1950) has shown that the presence of turpentine abscesses in dogs increased the urinary nitrogen and sulfate output by 40 to 50% but had little influence on the incorporation of Ss6-labeled methionine into the tissue or plasma proteins. This might suggest that a negative nitrogen balance and hypoalbuminemia is more closely related to increased protein breakdown than to decreased protein synthesis. Another possibility that may be at the basis of the low serum albumin levels in malignant disease could be an increased lability of the serum albumin in this disease. Such a suggestion has been made, for example, to explain the abnormal alkali lability of serum from dogs with experimental pneumonia (Crossley et al., 1941).
RICHARD J. WINZLER
516
A number of procedures have been proposed for detection of cancer or for following the development of the disease which appear t o be based on quantitative or qualitative changes in the serum albumin of patients with neoplastic disease. Some of these procedures are briefly considered in the following paragraphs. B. Thermal Coagulation Procedures. Glass (1940, 1950a, 1950b, 1951) and Huggins and his associates (1949, 1949a, 1949b, 1950, 1951) have demonstrated that serum from cancer patients is frequently less coagulable upon exposure t o heat than is normal serum and have pointed out that the determination of thermal coagulability of serum may be a very useful procedure for evaluation of the clinical state of patients and that it may serve as a useful diagnostic aid for detection of neoplastic disease. Glass determined the coagulation temperature and used this as a measure of coagulability (whereas Huggins et al. determined the least coagulable concentration of serum as the end point (Table IV)). TABLE IV Summary of Some Differences between Normal and Pathological Serum Depending upon Albumin Procedure Lowest Clottable protein concentration' Iodoacetate index' Sulfhydryl content' Polarographic serum testa Reduction times Binding of phenol red9 1
Normal Serum
Cancer Serum
1.33 f ,169 1.72 f .2 10.8 f .76 6.43 k 2.36 63.9 f .7 36.4 f 1.9 36.6 26.0 8.6 f .1 12.3 f .7 67.07 5.79 44.61 f 11.91
*
Other
Referenoe
1.39 f ,178 Huggina et ol. (1949b) 10.38 f ,918 Huggins et 02. (l949b) 22.3 f 7.6' Scboenbach et al. (1961) 26.88 Rusch et Ol. (1940) 10.2 f .3 Blaok (1947) Huggina et al. (19498)
-
Grams protein/lM) ml.
* Standard deviation.
Miacellaneoua diseases. pM iodoacetate/g. protein. 6 pM SH/100 ml. serum. 0 Wave height in millimeters. 7 Miacellaneoua states with abnormal A/G ratios. * Minutes. 9 pg. PSP bound by 1 ml. serum g. of albumin per 100 ml. serum a
.
Both procedures appear to give comparable results. Huggins et al. (l949,1949a), and Jensen st al. (1950) have extended the thermal coagulability test to the determination of the amount of iodoacetic acid which will prevent serum from thermal coagulation (iodoacetate index = micromoles iodoacetate/g. protein). The effect of iodoacetate on thermal coagulation is presumably by virtue of its reaction with protein-bound sulfhydryl groups. A considerable decrease in the iodoacetate index was found in the serum of cancer patients in comparison to normal individuals
PLASMA PROTEINS IN CANCER
517
(Table IV). Ponder (1950) has developed a turbidometric measurement of the heat coagulability of small amounts of serum. A number of investigators, including the originators of the methods, have pointed out that decrease in thermal coagulability and the iodoacetate index of serum is not specific for neoplastic disease and is of limited value for the diagnosis of cancer (Cliffton, 1949; Bodansky and McInnes, 1950; Boyd, 1950; Dvoskin and La Due, 1950; Homburger et al., 1950; Kiefer et al., 1950; Pollack and Leonard, 1950; West and Keye, 1950; Gilligan et al., 1950; Finnegan et al., 1950; Ellerbrook, 1950; Henry et al., 1951; Ellerbrook et al., 1951b). The extent to which thermal coagulation is dependent upon specific protein fractions has as yet not been unequivocally established. It does seem clear, however, that the thermal coagulation of serum is intimately associated with its content of protein-bound sulfhydryl groups. A diminished sulfhydryl content of cancer serum was strongly indicated by the experiments of Waldschmidt-Leitz (1938), of Purr and Russel (1934), and of Meyer-Heck (1939) in which the reactivation of partially inactivated glyoxalase or papain by serum from cancer patients was less than with normal serum. Since approximately 80% of the sulfhydryl content of serum proteins is contained in the albumin (Weissman et al., 1950), this fraction would be implicated as the major factor in thermal coagulation of serum. Moreover, the thermal coagulation of serum was shown experimentally by Huggins et al. (1950) to be related primarily to the level of Berum albumin. These authors indicated however that a t any given level of albumin, the serum of cancer patients coagulated less readily than normal, and felt that a qualitative change in albumin was indicated. It is of special interest in this connection that purified albumin isolated from normal serum and from serum of cancer patients by the low temperature alcohol method (method 6 of Cohn et al., 194.6) showed no difference in iodoacetate index, although the serum from the same patients showed marked differences in this index (Huggins et al., 1949b), and it was suggested that substances associated with the serum albumin affecting its thermal coagulation were removed during the fractionation procedure. In this connection Schoenbach et al. (1950, 1951) showed by amperometric titration that the sulfhydryl content of serum of patients with cancer or with certain other diseases was markedly reduced from the normal levels (Table IV) and reported that this reduction could only partially be accounted for on the basis of reduced albumin content. This they attributed to some sort of qualitative change in the serum albumin, since the sulfhydryl content per gram of serum albumin was below the normal level.
518
RICHARD J. WINZLER
Ellerbrook (1950) also showed that the least coagulable serum concentration and the iodoacetate index, when corrected for the reduced albumin content, still showed values which were abnormal, again supporting some sort of qualitative change independent of albumin concentration. C. Polarographic Serum Test. When proteins are examined polarographically in ammoniacal buffer in the presence of cobalt ion, there appears a characteristic “double wave” on the current-voltage curves. This double wave is a result of the catalytic reduction of hydrogen ion by a cobalt-sulfhydryl complex and is given only by proteins which contain reactive or “unmasked” cystine or cysteine. BrdiEka (1937, 1939a,b) noted t,hat cancer serum gave a lower polarographic serum test than did normal serum (Table IV) an observation confirmed by a number of other investigators (Tropp, 1938; Meyer-Heck, 1939; WaldschmidtLeitz and Mayer, 1939; Rusch et al., 1940; Robinson, 1948). The subnormal albumin content of the serum of cancer patients appears to be the major basis for the polarographic serum test. Rusch et al. (1940) showed that the polarographic serum waves were closely proportional to the serum albumin concentration. The distinction between BrdiEka’s polarographic serum test and his polarographic filtrate test (BrdiEka, 193913) is sometimes overlooked. The polarographic seruw test depends upon reactive sulfhydryl and disulfide groups in the total serum proteins, whereas the filtrate test depends upon these groups in the filtrates of alkali-treated serum deproteinated with sulfosalicylic acid (see later discussion). Muller and Davis (1945,1947) have recently determined the ratio of the two determinations to give a ‘(protein index” which is independent of the temperature and of the properties of the particular dropping mercury electrode employed. D. Binding of Anionic Dyes. Certain proteins have the capacity to form reversible complexes with anionic dyes. Grollman (1925) and a number of observations have indicated that this capacity is reduced in the serum of cancer patients as compared to normal serum (Bennhold, 1932; Ehrstrom, 1937; Huggins et al., 1949a; Westphal et al., 1950, 1951 ; Huggins and Baker, 1951). The fundamental aspects of the binding of anions by proteins have recently been reviewed by Klota (1949), Scatchard (1949), Goldstein (1949), and Armstrong (1950). Different plasma proteins have very different capacities to bind anionic dyes, albumin being by far the most effective (Klotz, 1949; Huggins et al., 1949a; Westphal et al., 1950). The formation of protein-dye complex depends primarily upon the presence of free epsilon amino groups of lysine in accessible sites on the protein (Klotz, 1949). Methods employed for studying dye binding capacity of normal and pathological sera include the penetration of
PLASMA PROTEINS I N CANCER
519
unbound dye into gelatin (Bennhold, 1932), the adsorption of unbound dye with charcoal (Ehrstrom, 1937; Gottlieb and Ludwig, 1937), equilibrium dialysis (Huggins et al., 1949b); and chromatography (Westphal et al., 1950, 1951). In the most recent work on this subject a significantly subnormal ability of cancer serum to bind phenol red (Huggins et al., 1949a) or azorubin (Westphal et al., 1950, 1951) has been noted (Table IV). The reduction in dye binding capacity, however, was not limited to neoplastic disease. The amount of phenol red bound by serum was dependent primarily upon its albumin concentration. However, the amount of dye bound per milligram of albumin in serum from cancer patients was significantly less than in normal serum, suggesting an intrinsic difference between the albumins in the plasma of normal individuals and of cancer patients (see also Ehrstrom, 1937). However, when electrophoretically homogeneous albumin was isolated from normal serum and from cancerous serum by a low temperature ethanolic procedure (method 6 of Cohn et al., 1946), no difference in dye binding between the two were apparent (Huggins et al., 1949a). The explanation of subnormal dye binding by cancerous serum, thus, is not yet clear. E. Methylene Blue Reduction. Savignac et al. (1945) and Black (1947a,b, 1948) have reported that the time required for serum or plasma of cancer patients t o reduce methylene blue under defined conditions of dye concentration and alkalinity is increased over that obtained with normal plasma. This phenomenon, they found, was not limited to cancer, but was sufficiently well correlated with malignant disease to suggest that it might have diagnostic use. Measurements of methylene blue reduction time of serum carried out in a number of laboratories using Black’s method (Stadie, 1948; Henry et al., 1951; Eriksen et al., 1951a) have only roughly correlated with malignant disease. The increased methylene blue reduction time of serum is not specific for neoplastic disease and is therefore of limited use in cancer diagnosis. Black, Kleiner, and Bolker (1948) and Stettner et al. (1948) combined the methylene blue reduction test with the heat turbidity test (see later) and claimed that the accuracy of diagnosis of malignant disease was improved. The use of this combination of tests by other workers (Henry et al., 1951), however, did not significantly improve the specificity of the tests. Savignac et al. (1945) and Black (1947a,b) advanced the hypothesis that the methylene blue reduction time depended upon sulfhydryl groups in serum albumin and that the increased reduction time observed with plasma from cancer patients was a result of an abnormality in the sulfurcontaining amino acids of albumin. Stadie (1948) pointed out that heating proteins in alkaline solution converts a large part of their cystine and
520
RICHARD J. WINZLER
cysteine sulfur t o sulfide ion, which is capable of reducing methylene blue. Subsequently Black (1948) reported that the increase in the methylene blue reduction times obtained with abnormal sera did not reflect differences in total potential reducing groups but appeared to be associated with a slower liberation of these groups in the proteins of the abnormal serum upon exposure to heat. It is perhaps regrettable that more attention has not been paid to the control of pH in carrying out the Black test, since the rate of liberation of sulfide ion from the so-called labile sulfur of proteins is very sensitive t o pH changes. It cannot be considered established that the methylene blue reduction time depends primarily on serum albumin, nor is it certain that the difference between normal and abnormal serum is associated with differences in this serum component. However, the fact that albumin is especially rich in “labile sulfur” and that serum albumin levels fall in cancer may be construed as evidence in favor of the view that serum albumin is primarily concerned in the methylene blue reduction test. From the foregoing discussion it appears that differences in thermal coagulation, polarographic activity in the presence of cobalt, dye binding, and reducing power between normal and pathological plasma appear t o be closely related to the serum albumin concentration. In addition there is suggestive evidence that qualitative differences between the albumins may also be involved. Hughes (1947, 1949) has shown that serum albumin contains a component containing one free sulfhydryl group per mole of protein (mercoptalbumin) and also contains a component lacking free sulfhydryl groups. About two-thirds of the normal serum albumin is mercoptalbumin (Hughes, 1947, 1949; Weissman et al., 1950). The attractive possibility that the relative amount of sulfhydryl containing albumin might be reduced in the serum albumin of cancer patients was not substantiated in the work of Huggins et al. (1949a), who found about two-thirds of a mole of sulfhydryl per mole of serum albumin isolated both from normal individuals and from patients with malignant disease. Human serum albumin has also been shown to be separable into different electrophoretic components at pH 8 (Hoch-Ligetti and Hoch, 1948) and at pH 4 to 4.5 (Luetscher, 1939,1940; Sharp et al., 1942; Miller et al., 1950). There is yet no evidence indicating whether the relative amounts of the electrophoretically separable albumins in serum from patients with neoplastic disease are abnormal. 2. Alpha Globulins
The serum levels of alpha-1 and alpha-2 globulins, determined electrophoretically, may increase significantly in neoplastic disease as has already been pointed out, but little can yet be said about the physiological
PLASMA PROTEINS IN CANCER
521
significance of these observations. Each of these alpha globulin fractions contains a number of proteins which have the same mobility in an electric field at pH 8.5. Thus, from Table 111, there are a t least three components each in the alpha-1 and alpha-2 globulin fractions. Until detailed studies by such procedures as method 10 of the Harvard Group (Cohn et al., 1950; Lever et al., 1951) are applied to pathological plasma, the significance of the increase in the alpha globulins in neoplastic disease cannot be evaluated in terms of their individual components. An increase in the alpha globulins of plasma rather consistently accompanies acute febrile and inflammatory conditions, as well as cancer. Longsworth, Shedlovsky, and McInnes (1939) and Shedlovsky and Scudder (1942) pointed out the frequent association of raised alphaglobulin levels with inflammation and tissue destruction. The plasma alpha globulins levels may reflect the activity of generalized processes such as tissue destruction or replacement. However, much more information is needed on the sites of formation and the metabolic turnover rates of the individual protein components in health and disease before this thesis can be accepted. One component of the alpha-1 globulin fraction which rises markedly in neoplastic disease and which has been isolated and characterized as a mucoprotein (Winzler and Smyth, 1948; Winzler et al., 1948; Weimer et al., 1950; Smith et al., 1950) will be discussed in the section on especially soluble and stable proteins. Little is known about the specific components of the alpha-2 globulin fraction which increase in malignant disease. The plasma alpha-2 globulin contains at least two proteins which are rich in carbohydrate, and part of the increase in protein-bound polysaccharide noted in neoplastic disease may be associated with increases in one or both of these components. Seibert et al. (1947), for example, noted a close correlation between the serum polysaccharide levels and the amount of alpha-2 globulin in the serum of normal individuals and patients with tuberculosis or cancer (see section on serum polysaccharides). 3. Especially Soluble and Stable Proteins
A number of observations on sera of normal individuals and of cancer patients have indicated that the latter contain increased amounts of very soluble and stable proteins or protein derivatives. Some of these appear to be alpha globulins electrophoretically and are found with the albumin on fractionation by the Howe procedure. Some of the more thoroughly investigated of these soluble protein are briefly considered in the following paragraphs.
522
RICHARD J. WINZLER
A. Polypeptidemia. Plasma or serum contains proteins or their split products which are precipitated by tungstic or phosphotungstic acids but not by weaker precipitating conditions such as trichloroacetic acid (at high concentration), sulfosalicylic acid, perchloric acid, and heat coagulation. A number of investigators have assayed this fraction by determining the difference in the nitrogen or tyrosine content of phosphotungstic acid and of 20% trichloroacetic acid filtrates of serum or plasma (Hahn, 1921; Wolff, 1921; Goiffon and Spaey, 1934; Reding, 1936, 1938; Larizza, 1937; Godfried, 1939; Winzler and Burk, 1944). Trichloroacetic acid filtrates from normal serum or plasma contain about 5 mg. % more nitrogen and 2 mg. % more tyrosine than do tungstic acid filtrates, these differences being considerably increased in serum or plasma from patients with neoplastic or certain other disease (see Table V). The differences were considered to represent polypeptide nitrogen or tyrosine precipitated by phosphotungstic acid but left in solution by trichloroacetic acid. TABLE V Summary of Some Differences between Normal and Pathological Plasma Depending on Soluble and Stable Proteins Normal Polarographic filtrate test’ Mucoprotein Nitrogen Index of Polypeptidemia’
Cancer
Other
Reference
1 .O 2 . 4 rt .3* 1 . 5 f .2* BrdiEka et al. (1939) 2 . 7 f . 5 6 . 1 f 1 . 3 7 . 2 f . 8 Winzler and Smyth (1948) 3.7 f . 3 6.5 . 6 5 . 9 f .5 Winzler (unpublished)
mm. wave length for abnormal serum, ratio mm. wave length for normal serum Standard deviation. 8 Miscellaneous diseases. 4 Mg. mucoprotein tyrosine %. 6 Difference between nitrogen oontent of sulfaaalicylic acid and phosphotungstic acid filtrates of serum. Mg. %.
B. Polarographic Filtrate Test. BrdiEka has proposed a second nonspecific polarographic test for cancer based on the determination of high molecular weight, cystine-containing substances in filtrates of serum deproteinated with sulfosalicylic acid (BrdiEka et al., 1939; BrdiEka, 193913). The determination depends, as does the polarographic serum test, upon the catalytic reduction of hydrogen ion by a sulfhydryl-cobalt complex giving a characteristic “wave ” on a current-voltage curve. Sulfosalicylic acid filtrates of the sera of cancer patients gave significantly higher polarographic currents than did filtrates from normal sera (Table V). BrdiEka et al. (1939) attributed the increase in the polarographic filtrate test in pathological sera t o protein split products liberated as a
PLASMA PROTEINS IN CANCER
523
result of the activation of Abderhalden’s “Abwehrferment ’’ (see later). BrdiEka recognized that the polarographic filtrate test was not a specific test for malignant disease, but felt that it was a useful diagnostic aid. A number of investigators have confirmed BrdiEka’s observations (Albers, 1940; Schmidt, 1940; Winzler and Burk, 1944). One of the objections to the polarographic method is the lack of absolute units by which results can be compared without reference to arbitrary normal sera studied simultaneously. Recently, Muller and Davis (1945, 1947) have carried out BrdiEka’s serum and filtrate tests in parallel and expressed the results as a ratio of the filtrate wave to the serum wave, this ratio being designated the “protein index.” Since in abnormal serum the filtrate wave rises and the serum wave falls, differences between normal and abnormal sera are accentuated by the “protein index.” Of special value is the fact that the use of this ratio minimizes the influence of temperature and eliminates differences between different capillaries and instruments, permitting the expression of polarographic data in units reproducible in different laboratories. C. Mucoprotein. Winder and Burk (1944) showed that “polypeptidemia” and the polarographic filtrate test were given by substances in blood which had the solubility and stability properties of proteoses. The responsible fraction was isolated from sulfosalicylic acid filtrates of normal rat blood. Further work with normal human plasma (Winzler et al., 1948) showed that this material had the properties and composition of mucoprotein in agreement with previous observations made by Mayer (1942). The solubility of this fraction in 0.6 M perchloric acid and its precipitation by phosphotungstic acid provided a chemical method for its estimation in plasma or serum (Winzler et al., 1948). Patients with neoplastic disease, pneumonia, tuberculosis, rheumatic fever, and other conditions have significantly higher plasma mucoprotein levels than do normal individuals (Table V), (Simkin et al., 1949; Winzler and Smyth, 1948; Greenspan et al., 1951; Ellerbrook, 1950; Kelley et al., 1950, 1951; Henry et al., 1951; Jager et al., 1951). The major component responsible for the mucoprotein determination migrates with alpha-1 globulin a t pH 8.5, and has been shown (Weimer and Winzler unpublished) to be a major constituent of the seromucoid preparation of Bywaters (1909) and of Rimington (1940) as well as of the “mucoid-lihnliche substanz” of Mayer (1942). It seems very likely that it is also identical with Schmid’s (1950) small acid glycoprotein (Weimer and Winzler, unpublished ; Schmid, personal communication). Although this component migrates with the alpha-1 globulins at pH 8.5, its very acid isoelectric point permits its demonstration in untreated plasma at pH 4 t o 4.5 (Petermann and Hogness, 1948b; Mehl et al., 1949a; Mehl and Golden, 1950). The
524
RIcBARD J. WINZLER
significance of the acidic alpha-1 mucoprotein and of its increase in cancer is not yet known, nor is there any information on where this particular protein is synthesizied. D. Albumin A . A decrease in especially soluble “albumins” in the plasma of patients with malignant disease is the basis of the Kahn albumin A determinations (Kahn, 1925, 1930; Goldschmidt and Kahn, 1929; Hanke and Kahn, 1941; Hanke et al., 1944). Kahn’s procedure involves precipitation of the plasma proteins with 37.15 % ammonium sulfate, and testing the filtrate for soluble “albumins” by heat coagulation. This fraction is distinct from those responsible for the polypeptidemia, the polarographic filtrate or the mucoprotein determinations, in that it is heat coagulable, and its concentration is reduced in the plasma of cancer patients. A number of investigators (e.g., Tinozzi, 1927; Sievers, 1931, 1932; Lothammer and Pistofidis, 1937; Hinsberg et al., 1939) have confirmed the general aspects of Kahn’s work on albumin A, and have indicated that the decrease in this fraction is not at all specific for malignant disease. The nature of albumin A cannot be inferred from the meager information on its properties,
4. Beta Globulins At least four beta globulins have been demonstrated in plasma by low temperature-alcohol fractionation procedures (Table 111). The beta globulins are, in part a t least, associated with certain of the plasma lipids, and changes in the beta globulin levels are most frequently seen in association with accumulation of lipids and lipoproteins in the blood (e.g., Kunkel and Ahrens, 1949). A beta globulin is, for example, the serum component which gives the thymol turbidity test (Cohen and Thompson, 1947). Not enough work on the fractionation of the beta globulins has yet been done to assess the significance of the frequent rise in the serum beta globulin level in neoplastic disease. 5. Fibrinogen
The increase in plasma fibrinogen content which is often associated with cancer (e.g., Walton, 1933; Mider et al., 1950; Ellerbrook, 1950; Smith et al., 1951) is by no means characteristic of this disease, above normal levels being prominent in acute infections, nephrosis, pregnancy, after x-ray radiation and various injuries. Normal and pathological variations in plasma fibrinogen levels have been reviewed by Ham and Curtis (1938) and by Gram (1922). The factors affecting fibrinogen synthesis and breakdown are largely unexplored. There is good evidence that fibrinogen synthesis occurs in the liver. It is interesting to note that the argument that liver dysfunc-
PLASMA PROTEINS I N CANCER
525
tion may be the cause of the lowered plasma albumin levels in cancer patients cannot be used to explain the increased plasma fibrinogen noted in a good proportion of these patients. A. Sedimentation Rate. An increased erythrocyte sedimentation rate is frequently associated with neoplastic disease (e.g., Gram, 1922; Walton, 1933; Ham and Curtis, 1938; Moeschlin, 1944; Glass, 1950a; Finnegan et al., 1950; Gilligan et al., 1950). Since the plasma fibrinogen level has been recognized as an important factor influencing the sedimentation rate (e.g., Gray and Mitchell, 1942), this correlation is not surprising. Above-normal fibrinogen levels are not specific indications of neoplastic disease, and a similar nonspecificity is shown by the increased sedimentation rate associated with this disease. B. Heat Turbidity Test. Black, Kleiner, and Bolker (1948) showed that plasma from cancer patients tended to become more turbid on short exposure to heat than did normal plasma. The effect was dependent primarily upon fibrinogen, since serum did not show any increase in turbidity during the ten-second exposure t o boiling water involved in the test. The heat turbidity values paralleled the fibrinogen content of the plasma. The nonspecificity of increased fibrinogen levels limits the usefulness of the heat turbidity test for cancer diagnosis (Erickson et al., 1951b; Henry et al., 1951). By combining the heat turbidity test with the methylene blue reduction test already described, Black, Kleiner, and Bolker (1948) ; Black and Speer (1950); and Stettner et al. (1948) felt that the accuracy of diagnosis of malignant disease was significantly increased. However, Henry et al. (1951) did not find that the combination of tests materially improved the differentiation between neoplastic and other diseases. 6 . Gamma Globulins
A. Amounts and Immunity. There appears t o be no consistent change in the amounts of gamma globulins (electrophoretically determined) in neoplastic diseases (Mider et al., 1950). The gamma globulin may be reduced in late cancer, possibly due to decreased protein availability. Parfentjev and associates (Parfentjev and Duran-Reynals, 1951 ; Parfentjev et al., 1951) have, in fact, noted a pronounced reduction in the amount of euglobulin and of two naturally present immune bodies in tumor-bearing chickens and humans, and Wharton et al. (1951) noted a decreased antibody content of mice bearing transplanted tumors. Since the gamma globulin fraction contains a major proportion of the specific antibodies of the blood (e.g., Tiselius and Kabat, 1939; Enders, 1944) i t would appear that specific antibodie! are not formed in any quantity in response t o the presence of spontaneous neoplasia. This generalization
526
RICHARD J. WINZLER
probably does not apply to transplanted or virus induced tumors. For example, there is a sharp rise in the gamma globulin of chicken serum with the spontaneous regression of fowl lymphomas, the chickens being subsequently immune to implantation of the tumors (Krejci et al., 1948; Sanders et at., 1944). An immunity of rats or rabbits to transplanted tumors after tumor regressions has frequently been noted (e.g., Pearce and Brown, 1923; Besredka and Gross, 1938; Ostwald and Kuttelwascher, 1940; Cheever and Janeway, 1941; Gross, 1943; Harris, 1943). Friedewald and Kidd (1941) have found that an antibody against extracts of the Va carcinoma in rabbits may be demonstrated in the serum of rabbits bearing this transplanted tumor. The presence of Brown-Pearce carcinoma results in an increase in the complement fixation titer of rabbit serum titrated against extracts of the tumor (Cheever, 1940; Kidd, 1940; Appel et al., 1942; MacKenzie and Kidd, 1945; Thornton et al., 1950). Kidd (1944) also found that the transplantability of Brown-Pearce tumor was inhibited by incubating explants with serum containing the complement fixing antibody. The immunological aspects of neoplastic disease up to 1942 have been thoroughly considered by Stern and Willheim (1943). The hope that there might be changes in the antibody makeup of the plasma in neoplastic disease rests in good part upon the possible presence of abnormal substances in the tumor which would cause sensitization of the host. Evidence for such autogenous antigens in spontaneous cancer is largely lacking although Mann and Welker (1940, 1943, 1946) have advanced some evidence that specific proteins may be detected immunologically in human cancer tissue. B. Flocculation with Lipoidal Antigens.” Immunological reactions have been at the basis of a number of useful diagnostic (and therapeutic) procedures in a variety of diseases. Considerable effort has naturally been made to extend this approach to neoplastic disease. However, the fundamental basis upon which such immunological reactions rest i.e., the existence of antigens and antibodies specific for neoplastic disease, has not been established. A number of serological tests for cancer have been based on flocculation reactions of serum with various lipoidal “antigens,” greater flocculation generally being observed with cancer than with normal serum. Some examples of such flocculation with lipoidal “antigens” are found in papers by Ascoli and Izar (1910); Shaw-MacKenzie (1922); Kahn (1923, 1924) ; Fry (1925, 1926) ; Gruskin (1929) ; Weiss (1932) ;Lehmann-Facius (1932, 1936); Landau and German (1932). Repetition of these procedures by other investigators (see ,Stern and Willheim, 1943) has generally led to the conclusion that the distinction between serum from
PLASMA PROTEINS IN CANCER
527
normal individuals or patients with neoplastic or other disease lacks the specificity required for successful serodiagnosis of cancer. One of these test procedures (Gruskin, 1929) has recently been reexamined in considerable detail by Holmgren et al. (1951) and is worthy of mention since it illustrates the unsuspected difficulties that may be involved in such flocculation reactions. Gruskin’s procedure is based on the hypothesis that a foreign protein of embryonic character is present in malignant tissue and is also present in plasma and in fixed tissue cells. An alcohol-soluble “antigenic ” material was extracted from embryonic liver, and the serum t o be tested carefully layered over the “antigen.” Cancer serum formed characteristic floccules, whereas the control sera did not. In the reinvestigation by Holmgren et al. (1951), the principal agent bringing about the flocculation was shown t o be the alcohol used in the “antigen” preparations, and differences between normal and cancer sera were related to the rate of precipitation and the physical form 0: the protein precipitate at the alcohol-serum interface. There is considerable doubt as to whether immunological factors are involved in this or in other flocculation tests. The seroflocculation reaction for cancer recently described by Penn and his associates (Penn, 1950; Hall et al., 1950) is based on the hypothesis that endogenous nonsaponifiable lipids may become altered, carcinogenic, and essentially foreign substances with antigenic capacities. In the Penn procedure an “antigen ” prepared from the nonsaponifiable lipids of liver from patients with neoplastic disease is shaken with serum. The “antigen” with normal serum results in a fine persistent turbidity, whereas with cancer serum there is a rapid flocculation of the lipid and a clearing of the preparation. Preliminary reports (Hall et al., 1950) indicate a remarkable accuracy for this procedure, the number of erroneous diagnoses being relatively rare. Although this procedure differs from the older flocculation reactions in employing a nonsaponifiable “antigen,” it is not established that the phenomenon has an immunological basis. Fontaine et al. (1949) prepared a protein “antigen” from urine of cancer patients and, after heating the antigen with copper sulfate, found that more pronounced precipitation was produced with serum from cancer patients than with normal serum or serum from patients with various other diseases. As with the lipoidal antigens previously discussed, the antigenic nature of this urinary protein must be regarded with some skepticism until the reaction is shown to have an immunological basis. C. Multiple Myeloma. In multiple myeloma and in certain leukemoid diseases, the presence of exceedingly large amounts of abnormal beta or gamma globulin (more frequently the latter) may be demonstrated in serum by electrophoretic or salt fraction methods (Perlzweig
528
RICHARD J. WINZLER
et al., 1928; Gutman et al., 1941; Longsworth et al., 1939; Kekwick, 1940; Moore et al., 1943; Malmros and Blix, 194G; Adams et al., 1949; Pearsall
and Chanutin, 1949; Rundels et al., 1950; Schoenbach et al., 1951). High globulin values are frequently obtained by the Howe fractionation procedure. There has been, however, no consistency in the appearance of the abnormal globulin in any particular Howe fraction, the protein being in some cases associated with the pseudoglobulin and sometimes with the euglobulin fractions. In the low-temperature fraction method of Cohn et al. (1946) the abnormal globulin is usually found in fractions I1 and I11 but may also be found in fraction I (Pearsall and Chanutin, 1949). Similar inconsistencies have been observed in the electrophoretic studies where the abnormal protein may have mobilities varying mainly between those of the beta and the gamma globulins (Adams, Alling, and Lawrence, 1949). An interesting observation by Mehl and Golden (1950) is that the abnormal globulin may appear homogeneous at pH 8.5 but form two distinct components a t pH 4.5. It appears very probable that the hyperglobulinemia of multiple myeloma is due to the presence of large amounts of Bence-Jones or related abnormal proteins in the plasma. However, in some instances the abnormal globulins do not appear t o be typical Bence-Jones proteins. The present status of this work has been excellently reviewed by Gutman (1948) and by Adams, Alling, and Lawrence (1949). At the present time there is no definite evidence as to the site of formation of the abnormal globulins in multiple myeloma, although it seems likely that the protein is formed in the bone marrow lesions.
7. Changes in Protein Stability to Precipitating Agents A number of diagnostic tests for neoplastic disease have been proposed which are based on changes in behavior of the plasma proteins with various precipitating agents. Too little is known about the chemistry of most of these reactions to permit anything but speculation as to their fundamental basis. Some of these procedures may be based on such protective colloidal phenomena as have been studied by Munro (1944). A few examples of this sort of procedure are listed below, but are not discussed since their relation to specific proteins has not been established. In the lactogelification reaction of Kopaczewski (1935, 1936, 1937) serum from cancer patients forms a gel more rapidly upon the addition of lactic acid than does normal serum. The precipitation of serum proteins by vanadic and acetic acids follows a different dilution pattern in normal and cancer sera (Bendien test, Cronin-Lowe, 1933). Weltmann (1930) heated diluted serum in the presence of progressively decreasing calcium chloride concentration and found that the precipitation pattern was altered in cancer serum. Dilutions of normal serum treated with the cationic
PLASMA PROTEINS IN CANCER
529
detergent, Bradosol (beta-phenoxyethyldimethyldodecyl ammonium bromide) formed a characteristic precipitation pattern which was markedly altered when serum from patients with neoplastic or other disease was employed (Mayer and Eisman, 1951; Bronfin et al., 1951).
111. PLASMA ENZYMES AND INHIBITORS The enzymatic and enzyme-inhibiting components of plasma are present in amounts too small to be determined by electrophoretic or isolation methods, and consequently alterations in their amount can be determined only through measurement of their activities. Although the physiological significance of most of these agents in normal blood is not known, studies of their activities under pathological conditions may yield information pertaining t o the normal physiological significance. Of special interest is the possibility that specific enzymes may escape from tumors into the blood stream and yield significant increases in their circulating levels more or less directly. Such activity alterations may, perhaps, be more specific for cancer than changes which measure the systemic response of the host to various diseases. 1 . Enzymes
Although a number of enzymes have been demonstrated in plasma, only a few have been shown t o have any consistent relationship to neoplastic disease. Some of those that have appeared a t one time or another to be of special interest are briefly considered in the following paragraphs. A. Acid and Alkaline Phosphatases. The relation of acid phosphatase to prostatic cancer with metastases to the bone marrow, lymph nodes, or liver is perhaps one of the most prominent instances of an enzyme with important diagnostic value in malignant disease. Normally present in relative low amounts, the serum acid phosphatase may rise to very high levels in patients with metastatic prostatic cancer. The work of Gutman and Gutman (1938), Barringer and Woodward (1938), Huggins and Hodges (1941), Gutman (1942), Herbert (1946), Dillard et al. (1949), and Huggins and Baker (1951) are a few examples of the application of serum acid phosphatase determinations t o diagnosis, therapy, and prognosis of metastatic cancer of the prostate. Since this enzyme is especially concentrated in prostatic tissue, it seems very probable that its presence in blood in abnormally large amounts in patients with metastatic prostatic cancer represents a direct contribution by the malignant tissue. Increases in the alkaline phosphatase content of serum may occur in hyperplastic disturbances of bone, as was shown by Kay (1929, 1930). The increased level of this enzyme appears to be due to the increased
530
RICHARD J. WINZLER
activity of the osteoplastic cells either because of primary osteogenic tumors being the stimulation of the osteoblastic activity produced by the presence of metastatic tumors in the bone (Huggins and Hodges, 1941; Huggins and Baker, 1951). Roche et al. (1946, 1947) have reported that low concentrations of zinc ion usually had an inhibitory effect upon the serum alkaline phosphatase of cancer patients, whereas it usually increased the activity of this enzyme in the serum of normal individuals and patients with other diseases. This work could not be substantiated in recent carefully conducted work (Bodansky and Blumenfeld, 1949; Fishman et al., 1949; and Ellerbrook et al., 1951a). B. AZdolase. Warburg and Christian (1943) observed that the glycolytic enzyme aldolase (zymohexase) was significantly increased in the serum of tumor-bearing animals. Significant increases were not observed in the serum of cancer patients, presumably because the tumor mass was too small a proportion of the total body weight. Recently Sibley and Lehninger (1949) modified and simplified the method for the determination of this enzyme and confirmed the work of Warburg and Christian. In a large proportion of tumor-bearing animals the aldolase level was'significantly above normal, but in only 20 out of 104 cases of human cancer was there a significant increase in the serum level of this enzyme. Particularly interesting, however, was the observation that general cachexia, infection, or pregnancy did not lead to increased serum aldolase levels, perhaps indicating that aldolase escapes from tumor into the blood stream. An increase in circulating aldolase would thus represent a direct contribution of the tumor rather than a secondary reaction of the host. C. D-Peptidases. The presence of D-peptidases in abnormally high amounts in the blood of cancer patients was claimed by WaldschmidtLeitz and co-workers (1940a,b). These observations have been challenged by a number of workers (Bayerle and Podloucky, 1940a,b,c; Herken and Erxleben, 1940; Euler et al., 1940; Maver et al., 1941; Berger et al., 1941; Ahlstrom e l al., 1942). On the basis of present evidence it does not appear that the D-peptidase activity of serum has any particular relationship to neoplastic disease. D. Carcinolysis. The Freund-Kaminer test for malignancy (Freund and Kaminer, 1910a,b; Kaminer, 1933) was based on the relative abilities of normal and pathological wra to bring about the in uitro cytolysis of cancer cells, usually prepared from liver metastases. Normal serum was more effective than serum from cancer patients in lysing the cancer cells. In spite of the extensive literature on the carcinolytic reaction, it is not clear whether differences between normal and pathological sera are due
PLASMA PROTEINS IN CANCER
531
to alterations in enzyme content or to the presence of lytic inhibitors in the cancer serum. Stern and Willheim (1937) and Waterman (1931) have provided some evidence that the enzyme involved in the carcinolytic reaction is a Lipase. The literature on the carcinolytic reaction has been extensively reviewed in the monograph by Stern and Willheim (1943). E. Plasmin. The presence of proteolytic enzyme in normal plasma has been well established (e.g., Christensen and MacLeod, 1945; MacFarlane and Biggs, 1948), although considerable confusion has resulted from differences in terminology and methodology employed by various workers, the necessity of activation of the enzyme, and the presence of a t least one serum inhibitor of proteolytic enzymes. It seems very likely that the enzymes known as plasmin, fibrinolysin, serum protease, and serum tryptase are identical. This enzyme, if the identity is assumed, occurs in the plasma in an inactive form (plasminogen) and becomes activated when the plasma is treated with chloroform or with streptokinase. Quantitative measurement of this enzyme in serum is possible only after separation from the protease inhibitor which also occurs in serum. Dillard and Chanutin (1949) have studied the activities of plasmin (after separation from the inhibitor and activation with chloroform), of proteolysin (a spontaneously active proteolytic enzyme after separation from the inhibitor), and of trypsin inhibitor (see later) in the plasma of normal individuals, of cancer patients, and of patients with a miscellaneous group of diseases. Significant increases over the normal levels in all three principles were found both in the cancer and in the miscellaneous disease groups. Changes in the spontaneously active protease were the most marked and were also the most sensitive t o the clinical improvement following surgery. Dillard and Chanutin (1949) suggested that the proteolysin concentration may reflect the presence of a tissue product which stimulates the formation of the enzyme, or which induces the activation of plasminogen, and attributes its presence t o necrosis and inflammatory processes accompanied by elevated temperatures. An activator of fibrinolysin or plasmin (fibrinokinase) has indeed been shown to occur in animal tissues (e.g., Astrup et al., 1950; Tagnon and Pallade, 1950). F. Fuchs’ Test. A procedure for cancer diagnosis which has had wide discussion is the so-called Fuchs’ test (Fuchs, 1926, 1936). In this procedure chloroform-treated serum is incubated with fibrin or with dried trichloroacetic acid precipitated serum, the increase in nonprotein nitrogen being determined as a measure of proteolysis. The claim was made that fibrin from the blood of a patient with cancer was not split by proteases in the serum of other cancer patients, but was split by the serum
532
RICHARD J. WINZLER
proteases from normal individuals or from patients with other diseases. The converse was also claimed. Fuchs’ results have been supported by the work of some investigators (e.g., Bing and Marangos, 1934; Woodhouse, 1937; Wright and Wolf, 1939) and refuted by others (e.g., Hinsberg, 1940). Fuchs held the view that specific immune bodies were associated with the fibrin substrate and protected it from hydrolysis by proteolytic enzymes of the homologous type of serum. It seems more likely, however, that the differences observed by Fuchs were due t o the interrelations of plasmin, plasminogen, proteolysin, and the proteolytic inhibitor activities of serum, an interrelationship which requires considerable fundamental research before application to abnormal plasma is likely to be very fruitful. G . Abwehrferment. Abderhalden (1936, 1937, 1941, 1946) has studied the occurrence of a defensive enzyme designated “ abwehrferment ” in the plasma, this enzyme allegedly having the prbperty of breaking down specific substances which induced its formation. In the application to serodiagnosis of cancer, protein preparations from neoplastic tissues are employed, these presumably being more readily attacked by proteolytic enzymes in plasma of cancer patients than by those enzymes in normal plasma. A voluminous literature (reviewed by Stern and Willheim, 1943) has accumulated on this subject. The fundamental basis upon which this test rests, i.e., the production of enzymes with specific activity toward proteins not normally present, certainly needs more experimental foundation. There is also the inherent conclusion that there may be circulating in the blood stream an almost unlimited number of proteolytic enzymes, all different with respect to specificity toward their substrates. Again, any evaluation of the fundamental basis of the test must take into consideration its relation to plasmin, proteolysin, and the proteolytic enzyme inhibitors of serum. 2. E n z p e Inhibitors
Plasma has been shown to contain a number of factors which inhibit the action of certain enzymes. The activity of a number of these enzyme inhibitors may rise markedly in pathological states including neoplastic disease (Winzler, 1950). A. Inhibition of Proteolytic Enzymes. The most extensively studied of the serum enzyme inhibitors has been concerned with the inhibition of the action of trypsin. The early work of Brieger and Trebing (1908) showed that the plasma of cancer patients contained supernormal amounts of a very powerful inhibitor of trypsin-like enzymes. There
PLASMA PROTEINS IN CANCER
533
has been a continued interest in this type of inhibitor with a number of recent publications bearing on the subject (Grob, 1943, 1949; Kaplan, 1946; MacFarlane and Biggs, 1948; Clark et al., 1948; Guest et al., 1948; Cliffton, 1949, 1950; Dillard and Chanutin, 1949; Duthie and Lorenz, 1949a; Loomis et al., 1949; Wells et al., 1949; Waldvogel and Schmitt, 1950; Waldvogel et al., 1949; Cliffton and Young, 1950; Lewis and Ferguson, 1950; Ungar and Damgaard, 1951). This inhibitor is a remarkably potent one, 0.003 ml. of normal serum containing enough inhibitor to inhibit by 50% of the action of 0.02 mg. of crystalline trypsin (Clark et al., 1948). Serum from patients with neoplastic disease is even more effective. There is no evidence which would indicate that the agent inhibiting trypsin is different from the one inhibiting plasmin. The inhibitor appears in the albumin fraction of serum or plasma by both salt and electrophoretic separations, and its heat lability and nondialyzability indicate that it is a protein in nature. Beloff (1946) has shown that the proteolytic enzyme of skin was powerfully inhibited by plasma, the inhibitor being in the very soluble albumin fraction. Further work will be necessary t o establish Beloff’s contention that this inhibitor of skin protease is different from the plasma trypsin inhibitor. West et al. (1949a,b, 1951a,b,c) and Tauber (1950) have studied the inhibition of chymotrypsin by human serum and have shown that the activity of the chymotrypsin inhibitor rises markedly in cancer and in a number of other pathological conditions. The possible identity of serum chymotrypsin inhibitor with the serum trypsin and fibrinolysin inhibitors has not yet been investigated. If further work proves that these proteolytic enzymes are all inhibited by the same plasma agent, a remarkable degree of nonspecificity would be indicated. The great potency bf proteolytic enzyme inhibitors and their increase in activity in pathological states suggest that the factors may have important physiological roles. Cliffton (1950) has indicated that increases in this factor are not related to tissue necrosis. In addition to the above proteolytic enzyme inhibitor or inhibitors, West et al. (1949a,b, 1951a,b,c) have shown that serum contains an inhibitor of rennin activity, this inhibitor being less potent than the chymotrypsin inhibitor. The chymotrypsin and rennin inhibitor levels were followed in individual cancer and leukemia patients given various types of therapy. It was noted that these two inhibitors exhibit characteristic changes in activity during clinical improvement of disease or with its reactivation. Effectiveness of therapy in individual cases could be correlated with the balance between the rennin and chymotrypsin inhibitors (West et al., 1949a,b, 1951a,b,c; Ellis and West, 1951). The two inhibitors frequently responded in opposite directions to therapy indicating that they were separate agents.
534
RICHARD J. WINZLER
Still another proteolytic enByme inhibitor in serum has been studied by Duthie and Lorenz (1949b), who have shown that the inhibition of bacterial gelatinases by serum is brought about by an agent which has a different solubility in ammonium sulfate, a lesser stability, and a different rate of reaction with the enzyme than does the serum trypsin inhibitor. B. Inhibitions of Hyaluronidase. At least two hyaluronidase inhibitors have been demonstrated in serum, these differing in their heat lability, in their specificity toward hyaluronidase from different sources, and in their tendency toward abnormal levels in different pathological conditions. One of these appears to be a neutralizing antibody for streptococcal antigens and is associated with the gamma globulins (Moore and Harris, 1949). Another hyaluronidase inhibitor is a normal nonspecific constituent of plasma. The activity of this inhibitor rises significantly in cancer and in a number of pathological conditions (Dorfman et al., 1948; Fulton et at., 1948; Hakanson and Glick, 1948; Glick, 1950; Kiriluk et al., 1950). Glick (1950) has recently reviewed the behavior of this serum hyaluronidase inhibitor in health and disease. The nature of the nonspecific hyaluronidase inhibitor has been extensively studied by Glick and his associates, who have found that it is heat labile and presumably protein, that it occurs in the albumin fraction of plasma fractionated electrophoretically (Glick and Moore, 1948; Moore and Harris, 1949), that it bears no relationship t o plasma mucoprotein (Glick et al., 1949), and that it may be a heparin-lipoprotein complex originating in tissues rich in mast cells (Glick and Sylven, 1951). C . Inhibitors of Oxidizing Enzymes. Hirschfeld et al. (1946) reported that serum from cancer patients inhibited the oxidation of tyrosine by potato tyrosinase more effectively than normal serum and suggested that the change was a rather specific manifestation of neoplastic growth. Subsequently Marx (1949) and Stadie et al. (1947) found that such differences as were found were due to an increase in the induction period obtained in the presence of cancer sera. Shacter and Shimkin (1949) investigated the catecholase-inhibiting activity of normal serum and serum from cancer patients and observed that the induction time was significantly longer in the presence of normal than of pathological sera. This relationship, which is the reverse of the situation observed by Hirschfeld et al. (1946) was ascribed to the lowered sulfhydryl content of the sera of patients with cancer and other conditions. The experiments of Shacter and Shimkin would appear to be still another indication pointing toward decreased sulfhydryl content of the serum of patients with neoplastic or other types of disease.
PLASMA PROTEINS IN CANCER
535
IV. PROTEIN-BOUND CARBOHYDRATE 1. Total Polysaccharide
For a long time it has been recognized that a considerable amount of carbohydrate is associated with the serum proteins, the amount being somewhat in excess of the blood glucose levels. In a number of pathological conditions, including cancer, the amount of serum protein-bound carbohydrate may rise from normal values of about 100 mg. % ' to as high as 200 mg. % (Bierry et al., 1921; Lustig and Langer, 1931; Hinsberg and Merten, 1938; Lustig and Nassau, 1941; Novak and Lustig, 1947; Seibert TABLE VI Summary of Some Differences in Normal and Pathological Sera Depending on Protein-Bound Polysaccharide Normal Serum Nonglucosamine polysaccharide1 Glucosaminel
Tr yptophan-perchloric acid reaction' Diphenylamine reactions
Cancer Serum
Other Serum
Reference
111. f 9.32 171. f 32.5 149. f 29.18 Shetlar et al. (1949) 69 f 5.2 95 f 17.6 89 f 18.18 Shetlar et al. (1949) 60 106 1466 Seibert el al. (1948) 0.337 0.555 0.623' Niazi & State (1948)
Img. %. Standard deviation. 8 Non-neoplastic pathology. 4 Klett units. Advanced tuberculosis. 6 Optical denaity units. 7 Inflammations. 2
and Atno, 1946; Seibert et al., 1947, 1948; Shetlar, Erwin, and Everett, 1950; Shetlar et al., 1948, 1949, 1950). Similar increases in the proteinbound hexosamine are also noted in the plasma of cancer patients (Nilsson, 1937; West and Clark, 1938; Shetlar el al., 1949) some illustrative data is given in Table VI. The protein-bound carbohydrate is distributed among a number of proteins of plasma (Table 111),albumin being low and alpha globulin relatively high in carbohydrate content (Blix et al., 1941; Seibert et al., 1948). The protein-bound carbohydrate appears to consist of approximately equimolecular amounts of galactose, mannose, and glucosamine (Rimington, 1929, 1931, 1940; Friedmann, 1949; Waldron and Woodhouse, 1950).
536
RICHARD J. WINZLER
As has already been pointed out, one of the alpha-1 carbohydrate containing proteins has been isolated in a purified state (Weimer et al., 1950; Smith et al., 1950; Schmid, 1950). This mucoprotein has been shown (Weimer and Winzler, unpublished) to be a major constituent of the seromucoid preparations of Bywaters (1909) and Rimington (1940), Staub and Rimington (1948), as well as of the “mucoidlihnliche substanz” of Mayer (1942). It seems very likely that it is identical with the small acid glycoprotein described by Schmid (1950) (Schmid, personal communication; Weimer and Winzler, unpublished), and appears to be mucoprotein in nature (Meyer, 1945; Stacey, 1946). This component contains 16% hexose and 12% hexosamine and accounts for about 10% of the total serum polysaccharide (Winzler et al., 1948). Its concentration may rise several fold above normal in the plasma of cancer patients (Winzler and Smyth, 1948). This mucoprotein is isolated with the serum albumin fraction by most chemical separations, and therefore would account, in part at least, for the increase in the carbohydrate content of this fraction in pathological conditions (Novak and Lustig, 1947; Shetlar, Erwin, and Everett, 1950; Shetlar et al., 1950; Blix et al., 1941). Little is known about the behavior of other polysaccharide-containing proteins of plasm? in disease. It seems likely however, that the levels of a number of the carbohydrate-rich proteins in plasma rise in a variety of pathological states, including cancer. The levels of the different polysaccharidecontaining proteins may vary in a differential manner in different diseases. Seibert et al. (1947, 1948) showed that, in several wasting diseases, including tuberculosis, sarcoidosis and cancer, there was a parallel rise in the alpha-2 globulin and in serum polysaccharide, suggesting that the major serum polysaccharide increase was associated with the alpha-2 globulin fraction. Considerable interest and importance is concerned with the source of the protein-bound carbohydrate. One possibility is a local origin a t the site of the lesion which results in elevated levels. Seibert et al. (1947, 1948) suggested that the rise in alpha-2 globulin and serum polysaccharide was indicative of tissue destruction, a conclusion in accord with that of Longsworth el al. (1939) and Shedlovsky and Scudder (1942). Gersh and Catchpole (1949), Catchpole (1950), and Pirani and Catchpole (1951) have pointed out that the polysaccharide-containing proteins of serum may arise from the depolymerization of glycoprotein constituents in the ground substance of connective tissue in the region of certain lesions. Lustig and Nassau (1941) found that the carbohydrate content of the proteins of tubercular pleural effusions and cantharidin blisters frequently exceeded that of the serum proteins, suggesting that relatively carbohydrate-rich proteins entered the exudates from the tissues. Previously,
PLASMA PROTEINS IN CANCER
537
Lustig et al. (1937) had reported that the carbohydrate-content of the proteins of the venous blood frequently exceeded that of the arterial blood, again suggesting that carbohydrate-rich proteins were entering the blood from the tissues. Shetlar and his associates have studied the serum polysaccharides in normal individuals (1948) and in patients with cancer or other pathological conditions (Shetlar et al., 1949, 1950), a large proportion of whom showed abnormally high levels. Similar high serum polysaccharide levels were noted in tumor-bearing rats (Shetlar et al., 1950). These investigators suggested that an increase in serum polysaccharide level was associated with tissue proliferation rather than with tissue destruction per se (Shetlar et al., 1949). Werner (1949) showed that the serum glucosamine level rose significantly in rabbits bled repeatedly and that damage to the liver produced by administration of phosphorus or benzene prevented this increase. The data suggested that the carbohydrate-rich plasma proteins were produced in the liver under pathological conditions and that a subfraction of the alpha or beta globulins was especially concerned in this increase. Although conclusive evidence is thus still lacking with respect t o the origin and significance of the increased plasma mucoprotein and polysaccharide concentrations in cancer, the general increase of these components in a wide variety of diseases suggests that the increase represents a systemic effect of the tumor and that the components may not be formed in the lesion itself. It is clear from the many conditions in which increases in serum polysaccharides have been noted that such determinations are not likely to be of much use for the diagnosis of cancer. The possibility remains, however, that specific polysaccharide-containing proteins, present in relatively minor amounts, may be shown to be more directly associated with malignant disease than is the total serum polysaccharide level. 2. Tryptophan-Acid Reaction
An interesting observation was made by Seibert et al. (1948) in which it was noted that serum heated with tryptophan and perchloric acid developed a color with an absorption spectrum characteristic of fructose and not of other hexoses, although free or bound fructose could not be demonstrated in the serum. The amount of the color developed in this “ tryptophan-acid reaction’’ was significantly increased in the plasma of patients with tuberculosis or cancer. Riegel and Beatty (1950), Shetlar et al. (1949), Israel et al. (1949), and Weisbrod (1950), have also observed an increase in the tryptophan-acid reaction in the serum of patients with
538
RICHARD J. WINZLER
neoplastic disease. The nature of the substance, presumably carbohydrate, responsible for the reaction merits further study. 3. Diphenytamine Reaction
Niazi and State (1948) and Ayala, Moore, and Hess (1951) have determined the purple color produced when the diphenylamine reaction (Dische, 1929, 1930) is applied to hot trichloroacetic acid extracts of human serum from patients with cancer, rheumatic fever, or other conditions had significantly higher levels of the reactive substance than did normal serum. No information is available on the precise nature of the chromogenic substance, although it would appear t o be carbohydrate in nature. V. DISCUSSION
It is clear that many abnormalities occur in the plasma proteins in patients with neoplastic disease. The nature of most of these abnormalities and their physiological significance are largely unknown. Elucidation of these factors will await a great deal of fundamental work on the normal protein components. Studies of abnormalities may, however, be of great use in clarifying the normal physiology of particular proteins. The amount and kind of the plasma proteins reflect the activities of the various tissues and organs of the body, and the interplay between these tissues resulting in a steady state level of each constituent. Since neoplasms may have effects on tissues remote from the site of the tumor, the plasma changes associated with this disease clearly may have a highly complicated etiology. From the work that has thus far been carried out, it would appear that a large number of pathological conditions lead to very similar abnormalities in the plasma protein picture. This suggests that most of the abnormalities considered in previous pages are associated with systemic changes in the host elicited by the neoplasm, rather than a direct effect of the tumor. Most attention has thus far naturally been focused on those proteins present in the largest amounts. It would seem very possible, however, that the most pronounced and perhaps specific changes associated with neoplastic disease may reside in those proteins present in very small amounts or essentially absent in plasma of normal individuals. Again, any attempt to pursue this approach must be associated with an extensive study of the characterization of the minor plasma proteins. In addition to studies of the amounts of the various plasma proteins, considerable information may ultimately be obtained by investigation of the metabolic activities of the major and minor proteins using isotopic techniques. Studies on the turnover of individual proteins in the plasma
PLASMA PROTEINS IN CANCER
539
of normal individuals and of patients with neoplastic or other disease may be expected to yield very significant data on the fundamental causes of increased or decreased amounts of particular proteins. It is perhaps worthy of emphasis to note that it will scarcely be adequate to study the relative metabolic activity of the gross fractions obtainable by electrophoretic, ethanolic, or salt fractionation, since the most pronounced changes in a minor protein may be completely masked by contaminating major fractions. It is to be expected with some confidence that continuing improvements in fractionating techniques will permit the separation of plasma proteins into more and more homogeneous fractions. Perhaps one of the most pressing problems relating t o the plasma proteins is the elucidation of their source and their fundamental biological significance. Considerable fundamental work has already been carried out with respect to the source of a number of the major components, and with the significance of some of the proteins associated with blood coagulation. Much remains, however, in elucidating these points in connection with the minor plasma, proteins, and especially in disease states. The hope that specific abnormalities in plasma proteins might be of diagnostic significance in detection of cancer has not been substantiated by the work that has been reviewed in the previous discussion. Significant abnormalities are, indeed, associated with this disease, but thus far no specificity has been apparent. REFERENCES Abderhalden, E. 1936. Fermentforsehung 16, 245-50. Abderhalden, E. 1937. Ergeb. Enzymjorsch. 6, 189-200. Abderhalden, E. 1941. Abwehrfermente (Die Abderhaldenische Reaktion). Steinkopf, Dresden and Leipzig. Abderhalden, E. 1946. Schweiz. med. Wochnschr. 76, 47. Abels, J. C., Ariel, I., Rekers, P. E., Pack, G. T., and Rhoads, C. P. 1943. Arch. Surg. 46, 844-60. Adams, W. S., Alling, E. L., and Lawrence, J. S. 1949. Am. J. Med. 6, 141-61. Ahlatrom, L., Euler, H. V., and Hogberg, B. 1942. 2.physiol. Chem. 273, 129-57. Albers, D. 1940. Biochem. 2.306, 236-44. Appel, M., Saphir, O., Janota, M., and Straus, A. A. 1942. Cancer Research 2, 576-78. Ariel, I. M. 1949. Surg. Gynecol. Obstet. 88, 185-95. Ariel, I., Jones, F., Pack, G . T., and Rhoads, C. P. 1943. Ann. Surg. 117, 740-47. Armstrong, 5. H. 1950. Symposium on Nutrition 2. Plasma Proteins, C. C. Thomas and Co., pp. 22-61. Aacoli, M., and Izar, G. 1910. Miinch. med. Wochnschr. 67, 403-05; 954-56; 2 129-3 1. Astrup, T., Crockston, J., and MacIntyre, A. 1950. Acta Physiol. Scand. 21, 238-49. Ayala, W., Moore, L. V., and Hess, E. L. 1951. J . Clin. Invest. 90, 781-85.
540
RICHARD J. WINZLER
Barringer, B. S., and Woodward, H. Q. 1938. Trana. Am. Asaoc. Genito-Urin. Surgeuna 31,363-69. Bayerle, H., and Podloucky, F. H. 1940a. Biochem. 2. 304, 259-65. Bayerle, H., and Podloucky, F. H. 1940b. 2.phyajol. Chem. 264, 189-95. Bayerle, H., and Podloucky, F. H. 1940~. 2.Krebaforach. 60, 220-29. Beloff, A. 1946. Biochem. J. 40, 108-15. Bennhold, H. 1932. Ergeb. inn. Med. Kinderheilk. 42, 273-375. Berger, J., Johnson, M. J., and Baumann, C. A. 1941. J. Biol. Chem. 137, 389-95 Besredka, A,, and Gross, L. 1938. Ann. Inat. Paateur 60,5-12. Bierry, H., Rathery, F., and Levina, M. 1921. Compt. rend. 173, 56-57. Bing, M., and Marangos, G. 1934. Beitr. klin. Chir. 160, 417-44. Black, M. M. 1947a. Cancer Research 7, 321-25. Black, M. M. 194713. Cancer Reaearch 7, 592-94. Black, M. M. 1948. Science 108, 540-41. Black, M. M., Kleiner, I. S., and Bolker, H. 1948. Cancer Reaearch 8, 79-82. Black, M. M., and Speer, F. D. 1950. Am. J. Clin. Path. 20, 446-53. Blix, G., Tiselius, A,, and Svensson, H. 1941. J. Biol. Chem. 137, 485-94. Bodansky, O., and Blumenfeld, 0. 1949. Proc. SOC.Exptl. Biol. Med. 70,546-47. Bodansky, O., and McInnes, G. F. 1950. Cancer 3, 1-14. Boyd, L. J. 1950. N. Y. Slate J . Med. 60, 2167-72. BrdiEka, R. 1937. Nature 139, 1020-21. BrdiEka, R. 1939a. Acla radiol. et cancerol. Bohem. el Morav. 2, 7-16. BrdiEka, R. 1939b. Klin. Wochnachr. 18,305-08. BrdiEka, R., Novak, F. V., and Klumpar, J. 1939. Acla radiol. el cancerol. Bohem. el Morav. 2, 27-41. Brieger, L., and Trebing, J. 1908. Berlin. klin. Wochnachr. 46, 1041-44; 1349-51; 2260-6 1. Bronfin, G. J., Hart, R. W., Liebler, J. B., and Goldner, M. S. 1951. Proc. Soc. Exptl. Biol. Med. 77, 456-58. Bywaters, H. W. 1909. Biochem. 2. 16, 322-43. Catchpole, H. R. 1950. Proc. SOC.Exptl. Biol. Med. 76, 221-23. Cheever, F. S. 1940. Proc. SOC.Exptl. Biol. Med. 46, 517-22. Cheever, F. S., and Janeway, C. A. 1941. Cancer Research 1, 23-27. Christensen, L. R., and MacLeod, C. M. 1945. J. Gen. Phyaiol. 28, 559-83. Clark, D. G. C., Cliffton, E. E., and Newton, B. L. 1948. Proc. SOC.Exptl. Biol. Med. 69, 276-79. Cliffton, E. E. 1949. J. Nall. Cancer Znat. 10, 719-23. Cliffton, E. E. 1950. J. Natl. Cancer Znat. 11,33-50. Cliffton, E. E., and Young, L. E. 1950. Cancer 3, 488-92. Cohen, P. P., and Thompson, F. L. 1947. J. Lab. Clin. Med. 32, 475-80. Cohn, C., and Wolfson, W. Q . 1947. J. Lab. Clin. Med. 32, 1203-07. C o b , C., and Wolfson, W. Q . 1948. J . Lab. Clin. Med. 33, 367-70. Cohn, E. J. 1941. Chem. Revs. 28,395-417. Cohn, E. J. 1948. Blood 8,471-85. Cohn, E. J., Gurd, F. R. N., Surgenor, D. M., Barnes, B. A., Brown, R. K., Derouaux, G., Gillespie, J. M., Kahnt, F. W., Lever, W. F., Liu, C. H., Mittelman, D., Mouton, R. F., Schmid, K., and Uroma, E. 1950. J. Am. Chem. SOC.72, 465-74. Cohn, E. J., Oncley, J. L., Strong, L. E., Hughes, W. L., and Armstrong, S. H. 1946. J. Am. Chem.SOC.68, 459-75. Cremer, H. D., and Tiselius, A. 1950. Biochem. 2.380, 273-283.
PLASMA PROTEINS IN CANCER
54 1
Cronin-Lowe, E. 1933. Brit. Med. J . 1, 407-11. Crossley, M. L.,Kienle, R. H., Vassel, B., and Christopher, G. L. 1941. J . Lab. Clin. Med. 27, 213-22. Davidsohn, I. 1936. Am. J. Clin. Path. 6, 172-84. Dillard, G. H. L., and Chanutin, A. 1949. Cancer Research 9, 665-68. Dillard, G. H. L., and Pearsall, H. R., and Chanutin, A. 1949. Cancer. Research 9, 661-64. Dische, Z. 1929. Mikrocheniie 7, 33-68. Dische, 1;. 1930. Mikrochemie 8,4-32. Dole, V. P. 1944. J . Clin. Invest. 23, 708-13. Dorfman, A., Ott, M. L., and Whitney, R. 1948. J . Biol. Chem. 174,621-29. Durram, E. L. 1950. J. Am. Chern. SOC.72, 2943-48. Duthie, E. S.,and Lorenz, L. 1949a. Biochem. J. 44, 167-73. Duthie, E. S.,andLorenz, L. 194913. Biochem. J. 44, 173-78. Dvoskin, S.,and La Due, J. S. 1950. N . Y. State J . Med. 60, 1488. Edsall, J. T. 1947. Advances i n Protein Chem. 3, 383-479. Ehrstrom, M.C. 1937. Acta Med. Scand. 91, 191-96. Ellerbrook, L. D. 1950. First Conference on Cancer Diagnostic Tests of the National Cancer Inst., Chicago, Illinois. October 14. Ellerbrook, L. D., Lippincott, S. W., and Chipps, H. D. 1951a. J. Natl. Cancer Znst. 11, 739-56. Ellerbrook, L. D., Meek, E. C., and Lippincott, S. W. 1951b. J. Natl. Cancer Znst. la, 49-90. Ellis, F. W., and West, P. M. 1951. J . Clin. Invest. 30, 547-57. Enders, J. F. 1944. J. Clin. Invest. 23, 510-30. Eriksen, N., Ellerbrook, L. D., and Lippincott, S. W. 1951a. J . Natl. Cancer Znst. 11, 705-28. Eriksen, N., Ellerbrook, L. D., Meek, E. C., andLippincott, S. W. 1951b. J. Natl. Cancer Znst. 11,757-72. Euler, H. von, Ahlstrom, L., Skarzynski, B., and Hogberg, B. 1940. 2. Krebsforsch. 60,552-64. Finnegan, J. V., Brockland, I., Muether, R. O., Hawk, B. O., Inkley, J. J.,and Thoma, G. E. 1950. J. Lab. Clin. Med. 36, 708-12. Fishman, W. H., Wayne, A., and Homburger, F. 1949. Cancer Research 9, 681-83. Fontaine, R., Aron, M., and Buck, P. 1949. Schweiz. med. Wochnschr. 79, 227-30. Freund, E., and Kaminer, G. 191Oa. Wien. klin. Wochnschr. 28, 1221-23. Freund, E., and Kaminer, G. 1910b. Biochem. 2. 26, 312-24. Friedewald, W.F., and Kidd, J. G. 1941. Proc. Soc. Exptl. Biol. Med. 47, 130-32. Friedmann, R. 1949. Biochem. J. 44, 117-26. Fry, H. J. B. 1925. Brit. Med. J . 2 , 4-9. Fry, H. J. B. 1926. J . Path. Bad. 29, 353-64. Fuchs, H. J. 1926. Biochern. 2. 170, 76-101;176, 185-201; 178, 152-54. Fuchs, H. J. 1936. 2. ges. e z p t l . Med. 98, 70-132. Fulton, J. K.,Marcus, S., and Robinson, W. D. 1948. Proc. SOC.Exptl. Biol. Med. 69,258-62. Gersh, I., and Catchpole, H. R. 1949. Am. J . Anat. 86, 457-522. Gilligan, D. R., and Ernstene, A. C. 1934. Am. J . Med. Sci. 187,552-62. Gilligan, D. R., Rothwell, J. T., and Warren, 5. 1950. New En& J . Med. 242, 807-13. Glass, G. B. J. 1940. Presse mdd. 48,51-56. Glass, G.B. J. 1950a. Am, J . Med. 8, 745-54; 755-58; 759-66.
542
RICHARD J. WINZLER
Glass, G. B. J. 1950b. J . Mount Sinai Hosp. 17, 1-20. Glass, G. B. J., Boyd, L. J., and Dworecki, I. J. 1951. Proc. SOC.Exptl. Biol. Med. 76, 10-16. Glick, D. 1950. J. Mount Sinai Hosp. 17, 207-28. Glick, D., Good, R. A., Kelley, V. C., Winzler, R. J., and Mehl, J. W. 1949. Proc. SOC.Exptl. Biol. Med. 71, 412-15. Glick, D., and Moore, D. H. 1948. Arch. Biochem. 19, 173-75. Glick, D., and Sylven, B. 1951. Science 113, 388-89. Godfried, E. G. 1939. Biochem. J. 33, 955-59. Goiffon, R., and Spaey, J. 1934. Bull. SOC. chim. biol. 16, 1673-76. Goldschmidt, S., and Kahn, H. 1929. 2.physiol. Chem. 183, 19-31. Goldstein, A. 1949. Pharm. Rev. 1, 102-65. Gottlieb, H., and Ludwig, H. 1937. 2.klin. Med. 131, 358-82. Gram, H. C. 1922. Acta Med. Scand. 66, 107-61. Grassmann, W. 1951. Naturwissenschaffen 38, 200-06. Gray, 5.J., and Mitchell] E. B. 1942. Proc. SOC.Exptl. Biol. Med. 61, 403-04. Greenfield, R. E., and Meister, A. 1951. J . Natl. Cancer Znst. 11, 997-1005. Greenspan, E. M., Lehman, I., Graff, M. M., and Schoenbach, E. B. 1951. Cancer 4,972-83.
Greenstein, J. P. 1947. Biochemistry of Cancer. Academic Press, New York. Greenstein, J. P., Jenrette, W. V., and White, J. 1941. J . Natl. Cancer Znst. 2, 283-9 1.
Grob, D. 1943. J . Gen. Physiol. 26, 405-21. Grob, D. 1949. J. Gen. Physiol. 39, 103-24. Grollman, A. 1925. J. Biol. Chem. 64, 141-60. Grow, L. 1943. Quart. Bull. Polish. Znst. Arts and Sci. America 1, 418-30. Gruskin, B. 1929. Am. J. Med. Sci. 177, 476-87. Guest, M. M., Daly, B. M., Ware, A. G., and Seegers, W. H. 1948. J . Clin. Invest. 27, 785-92; 793-804. A. B. 1942. J . Am. Med. Assoc. 120, 1112-16. A. B. 1948. Advances in Protein Chem. 4, 155-250. A. B., and Gutman, E. B. 1938. J . Clin. Invest. 17, 473-78.
Gutman, Gutman, Gutman, Gutman,
A. B., Moore, D. H., Gutman, E. B., McClellan, V., and Kabat, E. A. 1941. J . Clin. Invest. 20, 765-83. Hahn, A. 1921. Biochem. 2. 121, 262-72. Hakanson, E. Y., and Glick, D. 1948. J . Natl. Cancer Znst. 9, 129-32. Hall, G. H., Penn, H. S., Dowdy, A. H., and Bellamy, A. W. 1950. Cancer Research 10, 223. Ham, T. H., and Curtis, F. C. 1938. Medicine 17, 413-445. Hanke, M. E., and Kahn, H. 1941. J . Biol. Chem. 114, 140. Hanke, M. E., Kahn, H., and Felsher, R. 1944. Federation Proc. 3, 58. Harris, M. 1943. J . Exptl. 2001.93, 131-45. Heidelberger, M. 1939. Bact. Revs. 3, 49-95. Henry, R. J., Berkman, S., Little, M. S., and Winder, R. J. 1951. J . Am. Med. A880C. 147, 37-39. Herbert, F. K. 1946. Quart. J . Med. 16, 221-41. Herken, H., and Erxleben, H. 1940. 2. physiol. Chem. 264,251-53. Hinsberg, K. 1940. Angew. Chem. 63, 356-62. Hinsberg, K., and Merten, R. 1938. 2. klin. Med. 136, 76-87. Hinsberg, K., Merten, R., and Mucke, K. 1939. 2. Krebsforsch. 40, 371-74. Hirshfeld, S., Duboff, G., and West, P. M. 1946. Cancer Research 6, 57-60, 224.
PLASMA PROTEINS IN CANCER
543
Hoch-Ligetti, C., and Hoch, H. 1948. Biochem. J. 43, 556-60. Holmgren, N., Denton, R. W., Levinson, S. A., Ivy, A. C., and Grubgeld, L. E. 1951. J . Natl. Cancer Znst. 11, 689-704. Homburger, F. 1950. Cancer 3, 143-172. Homburger, F., Pfeiffer, P. H., Page, O., Rizsone, G. P., and Benotti, J. 1950. Cancer 3, 15-25. Homburger, F., and Young, N. F. 1948. Blood 3, 1460-71. Howe, P. E. 1921. J . Biol. Chem. 49, 93-107; 109-113. Huggins, C. 1949. Cancer Research 9, 321-27. Huggins, C., and Baker, R. 1951. Cancer Research 11, 428-31. Huggins, C., Cleveland, A. S., and Jensen, E. V. 1950. J . Am. Med. Assoc. 143, 11-15.
Huggins, C., and Hodges, C. V. 1941. Cancer Research 1, 293-97. Huggins, C., and Jensen, E. V. 1949. J . Biol. Chem. 179, 645-54. Huggins, C., Jensen, E. V., Player, M. A., and Hospelhorn, V. 0. 1949a. Cancer Research 9, 753-57. Huggins, C., Miller, G. M., and Jensen, E. V. 1949b. Cancer Research 9, 177-82. Hughes, W. L. 1947. J . Am. Chem. Soc. 09, 1836-37. Hughes, W. L. 1949. Cold Spring Harbor Symposia Quant. Biol. 14, 79-84. Israel, H. L., Webster, M. B., and Maher, I. E. 1949. Am. J . Med. 0, 745-50. Jager, B. V., Brown, H., and Nickerson, M. 1951. J . Lab. Clin. Med. 37, 431-43. Jager, B. V., Schwartz, T. B., Smith, E. L., Nickerson, M., and Brown, D. M. 1950. J . Lab. Clin. Med. 36, 76-86. Jensen, E. V., Hospelhorn, V. D., Tapley, D. F., and Huggins, C. 1950. J . Biol. Chem. 186, 411-22. Kabat, E. A. 1943. J . Immunol. 47, 513-87. Kahn, H. 1923. Klin. Wochnschr. 2, 1364-65, 2241-42. Kahn, H. 1924. Klin. Wochnschr. 3, 920-22. Kahn, H. 1925. Ergebn. inn. Med. Kinderheilk. 27, 365-422. Kahn, H. 1930. Klin. Wochnschr. 9, 262-63. Kaminer, G. 1933. V i e n . klin. Wochnschr. 40, 1576-78. Kaplan, M. H. 1946. J. Clin. Invest. 26, 337-46. Kay, H. D. 1929. Brit. J . Exptl. Path. 10, 253-56. Kay, H. D. 1930. J . Biol. Chem. 89, 249-66. Kekwick, R. A. 1940. Biochem. J . 34, 1248-57. Kelley, V. C., Good, R. A., and Glick, D. 1951. J . Clin. Invest. 29, 1500-04. Kelley, V. C., Good, R. A., and McQuarrie, I. 1950. Pediatrics 6,824-35. Kidd, J. G. 1940. J . Exptl. Med. 71, 335-350; 351-371. Kidd, J. G. 1944. Science 99, 348-50. Kiefer, L., Sullivan, B. H., and Littlefield, W. 1950. Proc. Soe. Ezptl. Biol. Med. 73, 29-30. Kingsley, G. R. 1940. J . Biol. Chem. 133, 731-35. Kiriluk, L. B., Kremen, A. J., and Glick, D. 1950. J . Natl. Cancer Inst. 10, 9931000.
Klots, I. M. 1949. Cold Spring Harbor Symposia Quant. Biol. 14, 97-112. Kopacsewski, W. 1935. Bull. sci. pharmacol. 42, 135-145. Kopaczewski, W. 1936. Compt. rend. 202, 990-92. Kopaczewski, W. 1937. J. physiol. path. gen. 36, 50-70. Krejci, L. E., Ross, M. H., and Sweeney, L. 1948. Cancer Research 8, 575-80. Kunkel, H. G., and Ahrens, E. H. 1949. J . CZin. Invest. 28, 1575-79. Landau, J. L., and German, W,M. 1932. Am. J . Clin. Path. 2, 343-46.
544
RICHARD J. WINZLER
Larizza, P. 1937. Arch. exptl. Path. Pharmakol. 186, 232-46; 247-54; 255-61; 262-68. Lehmann-Facius, H. 1932. Klin. Wochnschr. 11, 775-76. Lehmann-Facius, H. 1936. Klin. Wochnschr. 16, 1591-92. Lever, W. F., Gurd, F. R. N., Uroma, E., Brown, R. K., Barnes, B. A., Schmid, K., and Schultz, E. L. 1951. J. Clin. Invest. SO, 90-111. Lewis, J. H., and Ferguson, J. H. 1950. J . Clin. Invest. 29, 486-90. Longsworth, L. G. 1942. Chem. Revs. SO, 323-40. Longsworth, L. G.,Shedlovsky, T., and MacInnes, D. A. 1939. J . Exptl. Med. 70, 399413. Loomis, E. C., Ryder, A., and George, C. 1949. Arch. Biochem. 20, 444-50. Lotthammer, R., and Rosenbohm, A. 1937. 2. Krebsforsch. 46, 28-35. Luetscher, J. A, 1939. J . Am. Chem. SOC.61, 2888-90. Luetscher, J. A. 1940. J . Clin. Invest. 19, 313-20. Luetscher, J. A. 1941. J . Clin. Invest. 20, 99-108. Luetscher, J. A. 1947. Physiol. Revs. 27, 621-42. Lustig, B.,and Langer, A. 1931. Biochem. 2.242, 320-37. Lustig, B.,Leiner, G., and Ernst, T. 1937. Ztsch. gee. exptl. Ned. 100, 492-501. Lustig, B., and Nassau, E. 1941. Am. Rev. Tuberc. 43, 817-24. McFarlane, A. S. 1935. Biochem. J . 29, 407-28; 860-93; 1175-1201; 1202-08; 1209-26. MacFarlane, R. G., and Biggs, R. 1948. Blood 3, 1167-87. MacKenzie, I., and Kidd, J. G. 1945. J . Exptl. Med. 82, 41-63. Madden, S. C. 1950. Symposia on Nutrition 2. Plasma Proteins, pp. 62-71. C. C. Thomas and Co. Madden, S. C., and Whipple, G. H. 1940. Physiol. Revs. 20, 194-217. Majoor, C. L. H. 1946. Yak J . Biol. and Med. 18, 419-41. Majoor, C.L. H. 1947. J. Biol. Chem. 169, 583-94. Malmros, H., and Blix, G. 1946. Acta Med. Scand. Suppl. 170, 280-306. Mann, L. B., and Welker, W. H. 1940. Am. J . Cancer 89, 360-64. Mann, L. S., and Welker, W. H. 1943. Cancer Research 3, 193-95; 196-97. Mann, L. S., and Welker, W. H. 1946. Cancer Research 6, 625-26. Martin, N. H., and Morris, R. 1949. J . Clin. Path. 2, 64-66. Marx, W. 1949. Cancer Research 9, 114-20. Maver, M.E.,Johnson, J. M., and Thompson, J. W. 1941. J . Natl. Cancer Znst. 1, 835-43. Maver, M. 1944. J. Natl. Cancer Inst. 4, 571-78. Mayer, K. 1942. Biochem. 2. 276, 16-24. Mayer, R.L., and Eisman, P. C. 1951. Proc. SOC.Exptl. Biol. Med. 77, 452-55. Mehl, J. W. 1950. Tex. Repts. Biol. Med. 8,169-82. Mehl, J. W., and Golden, F. 1950. J . Clin. Invest. 99, 1214-18. Mehl, J. W., Golden, F., and Winzler, R. J. 1949a. Proc. Soc. Exptl. Biol. Med. 72, 110-14. Mehl, J. W., Humphrey, J., and Winzler, R. J. 1949b. Proc. SOC.Exptl. Biol. Med. 72, 106-09. Meyer, K. 1945. Advances in Protein Chem. 2, 249-75. Meyer-Heck, P. 1939. 2. Krebsfmach. 49, 560-72. Mider, G. B.,Alling, E. L., and Morton, J. J. 1950. Cancer 8,56-65. Mider, G. B., Tesluk, H., and Morton, J. J. 1848. Acta union intern. cunlre cancer 6, 409-20. Miller, G. L., Miller, E. E., and Eitelman, E. 8. 1950. Arch. Biochern. 99, 413-19.
PLASMA
PROTEINS IN CANCER
645
Milne, J. 1947. J. Biol. Chem. 169, 595-99. Moeschlin, S. 1944. Helv. Med. Acta 11, 209-24. Moore, D. H., and Harris, T. N. 1949. J. Biol. Chem. 179, 377-81. Moore, D. H., Kabat, E. A., and Gutman, A. B. 1943. J. Clin. Invest. 22, 67-75. Miiller, 0. H., and Davis, J. S. 1945. J. Biol. Chem. 169, 667-79. Miiller, 0. H., and Davis, J. S. 1947. Arch. Biochem. 16, 39-54. Munro, L. A. 1944. J. Phys. Chem. 48, 187-95. Nakahara, W., and Fukuoka, F. 1949. Gann 40,45-69. Niaai, S., and State, D. 1948. Cancer Research 8, 653-56. Nilsson, I. 1937. Biochem. 2. 291, 254-258. Nitsche, G. A,, and Cohen, P. P. 1947. Blood 2, 363-70. Norberg, E., and Greenberg, D. M. 1951. Cancer 4, 383-386. Novak, J., and Lustig, B. 1947. J. Mount Sinai Hosp. 14, 534-41. Oncley, J. L. 1950. Symposium on Nutrition 2. Plasma Proteins, pp. 3-21. C. C. Thomas and Co. Oncley, J. L., Scatchard, G., and Brown, A. 1947. J . Phys. Chem. 61, 184-98. Oswald, W., and Kuttelwascher, H. 1940. 2.Zmmunitutsforsch. 98, 427-30. Parfentjev, I. A., Cliffton, E. E., and Duran-Reynals, F. 1951. Science 113,523-24. Parfentjev, I. A., and Duran-Reynals, F. 1951. Science 113, 690-91. Pearce, L., and Brown, W. H. 1923. J. Exptl. Med. 57, 811-12. Pearsall, H. R., and Chanutin, A. 1949. Am. J. Med. 7, 301-09. Pedersen, K. 0. 1945. Ultracentrifugal Studies on Serum and Serum Fractions. Almqvist and Wiksells, Uppsala. Penn, H. S. 1950. Cancer Research 10, 235. Perlsweig, W. A., Delrue, G., and Geschickter, C. 1928. J . Am. Med. Assoc. 90, 775-76.
Petermann, Petermann, Petermann, Petermann, 379-87.
M. L., and Hogness, K. R. 1948a. Cancer 1, 100-03. M. L., and Hogness, K. R. 1948b. Cancer 1, 104-08. M. L., Karnofsky, D. A., and Hogness, K. R. 1948. Cancer 1, 109-19. M. L., Young, N. F., and Hogness, K. R. 1947. J. Biol. Chem. 169,
Pillemer, L,, and Hutchinson, M. C. 1945. J. Biol. Chem. 168, 299-301. Pirani, C. L., and Catchpole, H. R. 1951. Arch. Path. 61, 597-601. Pollak, 0. J., and Leonard, A. 1950. J. Am. Med. Assoc. 124, 872-75. Ponder, E. 1950. Cancer Research 10, 139-40. Popjak, G., and McCarthy, E. F. 1946. Biochem J. 40, 789-803. Purr, A., and Russel, M. 1934. 2.physiol. Chem. 228, 198-206. Reding, R. 1936. Compt. rend. soc. biol. 123, 1238-40. Reding, R. 1938. Compt. rend. soc. biol. 129, 881-83. Riegel, C., and Beatty, P. R. 1950. Cancer Research 10, 495-97. Rimington, C. 1929. Biochem. J. 23, 430-43. Rimington, C. 1931. Biochem. J. 26, 1062-71. Rimington, C. 1940. Biochem. J. 34, 931-40. Robinson, A. M. 1948. Brit. J. Cancer 2, 360-68. Roche, J., van Thoai, N., Marcelet, J., and Desruisseaux, G. 1946. Compt. rend. soc. biol. 140, 632-34. Roche, J., Cornil, L., Desruisseaux, G., Bandoin, N., and Long, S. 1947. Compt. rend. soc. biol. 141, 1251-52. Rundles, R. W., Dillon, M. L., and Dillon, E. S. 1950. J . Clin. Invest. 29, 1243-60. Rusch, H. P., Klatt, T., Meloche, V. W., and Dirksen, A. J. 1940. Proc. SOC. Exptl. Biol. Med. 44, 362-65.
546
RICHARD J. WINZLER
Sanders, E., Huddleson, I. F., and Schaible, P. J. 1944. J . Biol. Chem. 166, 469-81. Savignac, R.J., Gant, J. C., and Sirer, I. W. 1945. Research Conference on Cancer. American Association for the Advancement of Science, Washington, D.C., pp. 241-52.
Scatchard, G. 1949. Ann. N . Y . Acad. Sci. 61, 660-72. Schmid, K. 1950. J. Am. Chem. Soc. 73, 2816. Schmidt, H. W. 1940. Ztschf. Krebsforsch. 60, 390-406. Schoenbach, E. B., Armistead, E. B., and Weissman, N. 1950. Proc. SOC.Erptl. Med. 78, 44-46. Schoenbach, E. B., Weissman, N., and Armistead, E. B. 1951. J. Clin. Invest. SO, 762-77.
Seibert, F. B., and Atno, J. 1946. J . Biol. Chem. 16S, 511-22. Seibert, F. B., Pfaff, M. L., and Seibert, M. V. 1948. Arch. Biochem. 18, 279-95. Seibert, F. B., Seibert, M. V., Atno, A. J., and Campbell, H. W. 1947. J . Clin. Invest. 26, 90-102. Shacter, B., and Shimkin, M. B. 1949. J . Natl. Cancer Znst. 10, 637-45. Sharp, D. G., Cooper, G. R., Erickson, J. O., and Neurath, H. 1942. J . Biol. Chem. 144, 139-47.
Shaw-MacKenzie, J. A. 1922. Lancet 2, 759-62. Shedlovsky, T., and Scudder, J. 1942. J . Exptl. Med. 76, 119-26. Shetlar, M. R., Erwin, C. P., and Everett, M. R. 1950. Cancer Research 10, 445-47.
Shetlar, M. R., Foster, J. V., Kelly, K. H., and Everett, M. R. 1948. Proc. SOC. Exptl. Biol. Med. 69, 507-11. Shetlar, M. R. Foster, J. V., Kelly, K. H., Shetlar, C. L., Byran, R. S., and Everett, M. R. 1949. Cancer Research 9, 515-19. Shetlar, M. R., Shetlar, C. L., Richmond, V., and Everett, M. R. 1950. Cancer Research 10, 681-83. Sibley, J. A., and Lehninger, A. L. 1949. J . Natl. Cancer Znst. 9, 303-09. Sievers, 0. 1931. Acta Path. Microbiol Scand. 8, 29-70. Sievers, 0. 1932. Acta Path. Microbiol Scand. 9, 11 1-38. Simkin, B., Bergman, H. C., and Prinzmetal, M. 1949. Am. J . Med. 6, 734-44. Smith, E. L., Brown, D. M., Weimer, H. E., and Winzler, R. J. 1950. J . Biol. Chem. 186, 569-75.
Smith, W. B., Rosenfeld, L., and Shinowara, G. Y. 501-07.
1951. Am. J . Clin. Path. 21,
Sobotka, H. 1951. J . Mount Sinai Hosp. 17, 1021-36. Stacey, M. 1946. Advances in Carbohydrate Chem. 2 , 161-201. Stadie, W. C. 1948. Science 108, 211. Stadie, W, C., Perlmutter, M., and Robinson, D. 1947. Am. J . Med. Sci. 2lS, 655-60.
Staube, A. M., and Rimington, C. 1948. Biochem J . 42, 5-13. Stern, K. G., and Reiner, M. 1946. Yale J . Biol. and Med. 19, 67-99. Stern, K., and Willheim, R. 1937. 2.Krebsforach. 46, 379-402. Stern, K., and Willheim, R. 1943. The Biochemistry of Malignant Tumors, Chemical Publishing Co., Brooklyn. Stettner, M. M., Baron, H., Lowenstein, B. E., Greenwald, A. E., Mershimer, W. L., and Kleiner, I. S. 1948. J . Lab. Clin. Med. 96, 136-40. Svensson, H. 1941. J. Biol. Chem. 139, 805-25. Tagnon, H. J., and Palade, G. E. 1950. J . Clin. Invest. 29, 317-24.
PLASMA PROTEINS IN CANCER
547
Tauber, H. 1950. Proc. Soc. Exptl. Biol. Med. 74,486-89. Taylor, H.L., and Keys, A. 1943. J . Biol. Chem. 148,379-81. Thornton, H., Ellerbrook, L. D., Rhees, M. C., Stowell, E. C., and Lippincott, S. W. 1950. Natl. Cancer Inst. 11, 97-124. Tinozzi, E. P. 1927. 2.Krebsforsch. 24, 477-87. Tiselius, A. 1937a. Trans. Faraday SOC.33, 524-37. Tiselius, A. 1937b. Biochem J . 31, 1464-77. Tiselius, A., and Kabat, E. A. 1939. J . Exptl. Med. 69, 119-31. Toennies, G. 1947. Cancer Research 7, 193-229. Treffers, H. P. 1944. Advances in Protein Chem. 1, 69-119. Tropp, C. 1938. Klin. Wochnschr. 17, 1141-48. Turba, F., and Enenkel, H. J. 1950. Naturwissenschaften 37, 93. Ungar, G., and Damgaard, E. 1951. J . Exptl. Med. D3, 89-97. Waldron, D. M., and Woodhouse, D. L. 1950. Nature 186, 186-87. Waldschmidt-Leitz, E. 1938. Angew. Chem. 61,324-27. Waldschmidt-Leitz, E., and Hatschek, R. 1940a. 2.physiol. Chem. 264, 196-97. Waldschmidt-Leitz, E., Hatschek, R., and Hausmann, R. 1940b. 2.physiol. Chem. 267, 79-90. Waldschmidt-Leitz, E., and Mayer, K. 1939. 2.physiol. Chem, 261, 1-19. Waldvogel, M. J., Marvin, H. N., and Wells, B. B. 1949. Proc. SOC.Exptl. Biol. Med. 72, 100-102. Waldvogel, M. J., and Schmitt, L. H. 1950. Cancer Research 10, 371-73. Walton, A. C. R. 1933. Quart. J . Med. 2, 79-115. Warburg, O., and Christian, W. 1943. Biochem. 2.314,399-408. Waterman, N. 1931. 2. Krebsforsch. 34,327-32. Weimer, H. E., Mehl, J. W., and Winzler, R. J. 1950. J . Biol. Chem. 186, 561-68. Weisbrod, F. G. 1950. J . Lab. Clin. Med. 36,408-10. Weiss, E. 1932. Arch. Path. 13, 106-11. Weissman, N., Schoenbach, E. B., and Armistead, E. B. 1950. J . Biol. Chem. 187, 153-165.
1949. Am. J . Clin. Path. lD, 448-53. Weltman, 0. 1930. Med. Klin. (Munich) 26, 240-45. Werner, I. 1949. Acta Physiol. Scand. 19,27-42. West, P. M , and Hilliard, J. 1949a. Proc. SOC.Exptl. Biol. Med. 71, 169-72. West, P. M., and Hilliard, J. 194913. Ann. West. Med. Surg. 3, 227-34. West, P. M., Hilliard, J., and Meitus, A. C. 1951a. Surg. Gynecol. Obstet. D2, 209-12. West, P. M., and Keye, W. R. 1950. J . Lab. Clin. Med. 36,675-80. West, P. M., Rapaport, S. I., and Tempereau, C. E. 1951b. Cancer 4, 177-83. West, P. M., and Smith, B. W. 1951c. J . Hematology 6, 330-36. West, R., and Clarke, D. H. 1938. J . Clin. Invest. 17, 173-78. Westphal, U., and Gedigk, P. 1951. Proc. SOC.Exptl. Biol. Med. 76, 838-43. Westphal, U., Gedigk, P., and Meyer, F. 1950. 2. physiol. Chem. 286, 36-58. m a r t o n , D. R. A., Miller, G. L., Wharton, M. L., Hankwitz, R. F., and Miller, E. E. 1951. Cancer Research 11, 127-33. Winzler, R. J. 1951. Proceedings of the First Conference on Cancer Diagnostic Tests of the National Cancer Inst., pages 53-57. Chicago, Illinois, October 14, 1950. Public Health Service Publication No. 96. U.S. Govt. Printing Office,
Wells, B. B., Marvin, H. N., and Waldvogel, M.
Washington, D.C.
548
RICHARD J. WINZLER
Winzler, R. J., and Burk, D. 1944. J . Natl. Cancer Znst. 4,417-28. Winzler, R. J., Devor, A. W., Mehl, J. W., and Smyth, I. M. 1948. J . Clin. Invest. 27, 609-16. Winzler, R. J., and Smyth, I. M. 1948. J . Clin. Invest. 27, 617-19. Wolff, E. 1921. Ann. mdd. 10, 185-97. Woodhouse, D.L. 1937. Lancet 1, 138-41. Woodhouse, D.L. 1940. Am. J. Cancer 40,359-74. Wright, W. M.,and Wolf, C. G. L. 1939. J. Cancer Research 14, 370-93.
Author Index Numbers in parentheses are references and are included to assist in locating references in which the authors”names are not mentioned in the text. Numbers in italics indicate the pages on which references are listed at the end of the article.
A
Andrew, V. W., 252, 271 Andrew, W., 66, 96, 99 Andrewes, C. H., 156,160, 246, 247, 255, 262, 264, 866, 267, 268 Anson, M. L., 398, 409, 448 Antopol, w., 347, 390 Apelgot, S., 328, 336 Apolant, H., 104, 117, 127, 160, 164 Appel, M., 526, 639 Argus, M. F., 293, 336 Ariel, I., 514, 515, 639 Armstead, E. B., 505, 508, 516, 517, 520, 528, 646, 647 Armstrong, E. C., 124, 160, 215(300), 218(320), 230, 831 Armstrong, M., 116, 121, 141, 160 Armstrong, S. H., 510, 517, 518, 519,
Abderhalden, E., 532, 639 Abels, J. C . , 514, 639 Adair, F. E., 490, 4.97 Adams, M. H., 256, 266 Adams, W. S., 528, 639 Ahlbom, H. E., 490, 497 AhlstrBm, L., 314, 315, 334,630,639, 641 Ahlstrom, C . G., 67, 98 Ahrens, E. H., 524, 643 Aiston, S.,252, 266 Aitken, H. A. A., 259, 266 Albers, D., 523, 639 Albert, S., 202 (158), 227 Alexander, P., 429, 435, 446 Algire, G. H., 484, 4.97 Allen, E., 119, 121, 135, 169, 164, 207 639,640 Arnesen, K., 146, 163 (205, 208), 210, 213(265), 288, 889 Alling, E. L., 504, 508,509, 528, 639, 644 Aron, M., 527, 641 Altman, S., 52, 64 Ascoli, M., 526, 639 Amies, C. R., 249,252, 256, 259, 260, 262, Astbury, W. T . , 136, 138, 139, 157, 169 Astrup, T., 531, 639 966,269 Athais, M., 195(93, 94), 212(93, 94), 826 Anderson, R. S., 256, 267 Atno, J., 504, 505, 508, 521, 535, 536, 646 Anderson, S., 434, 449 Au, M. H., 74, 100 Anderson, T. F., 253, 266 Auerbach, C., 398, 445, 446 Anderson, W., 50, 64 Andervont, H. B., 105, 106,107, 108,109, Austin, M. L., 157, 170 111, 112, 113, 114, 115, 116, 117, Ayala, W., 538, 639 121, 122, 124, 125, 126, 128, 129, 130, Ayengar, A. R. Gopal, 72, 98 132, 133, 134, 135, 136, 140, 142, 143, B 144, 148, 149, 150, 151, 152, 153, 154, 155, 156, 157, 159, 16.9, 160, 168, 163, 166, 166, 170, 182(22), 195, Back, A., 333, 334 198(107), 204 (87), 205, 213 (275), Bacon, R. L., 221(357, 358), 838 215(291, 301), 217(275, 291, 317), Baddeley, G.,420, 446 220, 883, 886, 230, 291, 344, 379, Badger, G. M., 4, 31, 35, 36, 48, 49, 64, 280, 334, 382, 390 380, 390, 475, 601 Bagg, H. J., 112, 120, 148, 160 Ando, T., 379, 390 549
550
AUTHOR INDEX
Baker, J. W., 17, 64 Beatty, P. R., 537, 646 Baker, R., 505, 516, 518, 529, 530, 645 Begg, A. M., 136, 166, 259, 266 Baker, S. L., 258, 259, 266 Beinert, H., 387, S9S Baldock, G. R., 32, 42, 64 Belding, T. C., 263, 266 Ball, Z. B., 106, 108, 110, 117, 128, 129, Belkin, M., 483, 601 132, 134, 135, 136, 137, 138, 139, Bellamy, A. W., 527, 64.2 160, 162, 166, 171, 214(287), 280, Beloff, A., 533, 640 401, 402, 403, 404, 405, 407, 446, Bendich, A., 319, ,934 457, 458, 466, 497, 499, 601 Benditt, E. P., 483, 498 Bandoin, N., 530, 646 Benedict, S. R., 250, 270, 481, 482, 600 Bang, F. B., 258, 266 Bennett, G. M., 420, 446 Banks, T. E., 398, 409,410, 437,446 Bennett, L. L., Jr., 323, 331, 332, SSd, Banyen, D., 64, 88, 99 SS6, SS7 Bardet, J., 139, 164 Bennhold, H., 518, 519, 640 Barlow, H., 117, 168 Bennison, B. E., 144, 146, 160, 168 Barnes, B. A., 510,511,512,521,640,6~~ Benotti, J., 645 Barnes, L., 458, 499 Berenblum, I., 91, 99, 410, 447, 464, 497 Barnes, R., 458, 497 Berg, N. O., 67, 98 Barnes, R. H., 136, 171, 457, 601 Berger, J., 530, 640 Barnes, W. A., 106, 135, 160 Berger, M., 328, 3.96 Barnum, C. P., 117, 132, 134, 135, 136, Bergman, H. C., 523, 646 137, 138, 139, 143, 160, 249, 269, Bergmann, M., 398, 401, 402, 405, 407, 314, 325, 334 412, 413, 414, 415, 416, 417, 418, Baron, H., 519, 525, 646 444449 Barrett, M. K,, 135, 160 Berkman, S., 517, 519, 523, 525, 642 Barringer, B. S., 529, 640 Berman, C., 351, 390, 490, 491, 497 Barron, E. 8. G., 249, 266, 268, 418, 447 Bernhard, W., 245, 255, 269 Bartlett, G. R., 401, 418, 447 Berry, G. P., 248, $370 Bartlett, P. D., 433, 447 Berthier, G., 16, 30, 31, 32, 33, 42, 43, 44, Barton, A. D., 304, $84 64, 66, 66 Bashford, E. F., 104, 117, 160 Besredka, A., 526, 640 Bauer, D. J., 250, 266 Baumann, C. A,, 63, 100, 341, 343, 346, Betheil, J. J., 364, 394 Biddulph, C., 189(48, 52), 224 347,348, 349,350,353,354,355,356, 357,359,360, 362,363,364,365,371, Bierry, H., 535, 640 377,379,387, S90,$91, S92, 39Sl S94, Biesele, J. J., 72, 74, 95, 99, 436, 445, 446, 447 896, 452, 457, 458,468, 469, 470, 471, 474, 477,478,479,482,497,4998,499, Biesele, M. M., 72, 74, 99 Biggs, R., 531, 533, 644 600, 530, 640 Billingham, R. E., 66, 99 Baumann, E. J., 256, 268 Bing, M., 506, 514, 532, 640 Baumberger, J. P., 74, 94, 99 Bird, M., 445, 447 Bawden, F. C., 247, 866 Bischoff, F., 120, 160, 204(177, 178), 205, Bayerle, H., 530, 640 827, 481, 497 Beach, J. Y., 6, 15, 66 Biskind, G. R., 185, 186(35), 192(35), Beadle, G. W., 109, 160 196, 197(101),224, 226 Beard, D., 251, 253, 254, 255, 256, 258, 263, 266, 869, 270, 271 Biskind, G. S., 185, 224 Beard, J. W., 137, 160, 233, 251, 253, 254, Biskind, M. S., 185, 186(35), 192(35), 255, 256, 257, 258, 263, 266, 267, ,969, 196, 197(101),224, 226 870, 871 Bissell, A. D., 213(272), 299
AUTHOR INDEX Bittner, J. J., 104, 105, 106, 107, 108, 109, 110, 111, 112, 113, 114, 115, 116, 117, 118, 120, 122, 123, 124, 125, 126, 128, 129, 130, 131, 132, 133, 134, 135, 136, 137, 138, 139, 141, 142, 143, 144, 145, 146, 148, 151, 152, 153, 154, 158, 160,161, 162, 164, 166, 166, 167, 169, 170, 171, 183(27), 200(125), 212, 213 (255, 276), 214(286), 215(292, 293), 217(316), 218(276, 318, 319, 322), 223, 226, 229, 230, 231 Black, A., 471, 498 Black, M. M., 505, 516, 519, 520, 525, 640 Blakemore, F., 150, 162 Bleyer, L., 235, 236, 267 Blix, G., 528, 535, 536, 640, 644 Bloch, H. S., 302, 334 Bloom, S., 302, 307, 336, 378, 391 Blumenfeld, O., 530, 640 Blumenthal, H. T., 121, 168, 212, 929 Bodansky, O., 517, 530,640 Boddaert, J., 188, 195, 196(91, 97), 205, 924, 226 Bohnel, E., 252, 266 Boissonnas, R. A., 290, 334, 364, 390 Bolker, H., 505, 519, 525, 640 Bonbme, R., 45, 66 Bonne, C., 490, 497 Bonser, G. M., 106, 122, 123, 124, 126, 155,160,162, 195,197(86), 213(271), 215(300), 226, 829, 230 Boon, M. C., 135, 164, 190(59), 191, 193(67), 194(67), 204 (185), 205 (185), 206(185), 224, 227, 457, 600 Boretti, G., 377, 391 Borgess, P. R. F., 107, 162 Borsook, H., 304, 307, 334 Borsos-Nachtnebel, E., 345,346,351,394 Boursnell, J. C., 398, 399, 405, 407, 408, 409, 410, 428, 437, 442, 446, 447 Boutwell, R. K., 108, 162, 457, 459, 466, 468, 469, 471, 479, 497 Boyd, D. R., 433,447 Boyd, L. J., 505, 516, 517, 640, 642 Boyland, E., 42, 46, 49,64, 207(210), 228, 280,288,289,334,336,358,380,381, 390, 398, 447 Brambel, F. W. R., 190, 924 Branch, G. E. K., 430, 447
551
Brandt, E. L., 317, 336, 376, 391 Brataler, J. W., 471, 498 Braa, G. I., 436, 447 Brdi6ka, R., 505, 518, 522, 640 Breedis, C., 346, 371, 378, 390 Brewer, J. I., 174(3), 223 Brewer, P., 280, 336 Bridgers, W. H., 257, 267 Brieger, L., 532, 640 Briggs, V. C., 198(107), 226 Brock, N., 345, 346, 390 Brockland, I., 517, 525, 641 Brockway, L. O., 6, 15, 66 Bronsted, J. N., 429, 430, 432, 447 Bronfin, G. J., 529, 640 Brown, A., 513, 646 Brown, D. M., 507, 513, 521, 536, 643, 646 Brown, E., 108, 164 Brown, G. B., 319, 320, 334, 336 Brown, H., 523, 643 Brown, J. B., 349, 391 Brown, M., 206(193), 207(203), 227, 228 Brown, R. D., 16, 48, 49, 64 Brown, R. K., 354, 388, 393, 510, 511, 512, 521, 640, 644 Brown, R. R., 351, 366, 369, 391, 393 Brown, W. G., 7, 27, 66 Brown, W. H., 526,646 Browning, H. C., 126, 162 Bruce, W. F., 45, 64 Brues, A. M., 275, 312, 319, 334, 358, 380, 381, 390 Brumfield, H. P., 252, 266 Brunschwig, A., 213(272), 229 Brush, M. K., 108, 162, 457, 459, 466, 468, 469, 471,479,497 Bryan, W. R., 106, 107, 110, 123, 132, 134, 135, 136, 137, 141, 144, 150,169, 160, 161, 166, 170, 201(139), 202 (130), 213(261), 226, 229, 240, 241, 242, 255, 257, 259, 266, 270, 284, 331, 332, 334, 337 Buchanan, J. M., 318, 334 Buck, P., 527, 641 Buckley, S. M., 436, 445, 446, 447 Bulliard, H., 73, 100 Bullock, F. D., 209(228), 228, 381, 391 Bullough, H. F., 63, 99 Bullough, W. S., 63, 99, 466, 497, 498
552
AUTHOR INDEX
Bunster, E., 189(50), 224 Burch, J. C., 208(215, 216), 228 Burchenal, J. H., 399, 436, 445, 446, 447 Burford, T. H., 219(332), ,93f Burk, D., 347, 396, 476, 478, 486, 498, 505, 522, 523, 648 Burkitt, F. C., 47, 64 Burmester, B. R., 263, 266 Burnet, F. M., 152, 168 Burnett, W. T., 203(160), 227 Burns, E. L., 210, 213(270, 273, 274), 220, 221(347), 229, 230, 231, 232 Burns, I. L., 121, 122, 168 BUITOWB, H., 104, 119, 162, 194(82), 195, 207(u)9, 210), 216(312), 219(330), 226, 227, 228, 231
Burrows, R., 174(2), 201(2), 219(2), 223 Busch, H., 302, 336 Butler, G. C., 409, 448 Butler, J. A. V., 410, 419, 436, 445, 447 Butterworth, J. S., 190((50),192(66), 224 Buu-Hoi, N. P., 7, 38, 47, 64, 66,. 293, 328,336 Buxton, L., 146, 163 Byran, R. S., 505, 535, 537, 646 Byrnes, E. W., 216, $30 Byrnes, W. W., 189(49),224 Bywaters, H. W., 523, 536, 6.40 C
Caldwell, A. L., 81, 100 Calnan, D., 240, 241, 266, 270 Calvin, M., 274, 276, 279, 334, 430, 447 Cambel, Perihan, 65, 67, 99 Campbell, H. W., 504,505,508,521,535, 536, 646 Carmichael, N., 117, 141, 171 Carpenter, F. H., 409, 411, 441, 447 Carpenter, G. E., 133, 166 Carr, C., 121, 166 Carr, J. G., 236, 237, 238, 240, 242, 246, 247,249,252,254, 255, 257, 258, 259, 260, 262, 266, ,966, 268 Carrel, A., 241, 243, 244, ,966 Carroll, D. M., 221(355), 232 Carruthers, C., 62, 63, 65, 74, 75, 76, 77, 81, 83, 84, 85, 88, 93, 95, 98, 99, 100, 101
Carter, C. E., 319, 337
Casas, C. B., 108, 121, 131, 16.9, 164, 166, 188(43), 214(288), 224, 230, 458, 462, @9 Cashmore, A. E., 407, 447 Caspari, W., 452, 498 Catchpole, H. R., 536, 640, 641, 646 Chaikoff, I. L., 203(161, 162), 227, 279, 313, 314, 334, 338 Chalkley, H. W., 484, 497 Chalvet, O., 47, 66 Chambers, L. A., 250, 253, 266 Chambers, R., 93, 99 Chamorro, A., 216, 230 Chang, C. H., 192(64), 224 Chanutin, A., 417, 419, 4.47, 448, 508, 528, 529, 531, 533, 641, 646 Chargaff, E., 310, 334 Cheever, F. S., 526, 640 Cheutin, A., 293, 336 Chipps, H. D., 530, 641 Chitre, R. G., 108, 166 Christensen, L. R., 531, 640 Christian, W., 505, 530, 647 Christopher, G. L., 515, 641 Chu, W. C., 358, 362, 363, 392 Clark, D. G. C., 533, 640 Clarke, D. H., 535, 647 Clarke, G. J., 204(177, 178), 205, 227' Claude, A., 137, 157, 162, 242, 245, 246, 249,250,251,252, 253,254, 255, 256, 257, 259, 267, 968, 269,270 Clayton, C. C., 343, 347, 348, 349, 350, 391, 458, 474, 477, 498 Clemmesen, J., 478, 498 Clemmesen, S.,478, 498 Cleveland, A. S., 516, 517, 643 Clifton, E. F., 135, 162 Cliffton,E. E., 193(72),194(72),224, 505, 517, 525, 533, 640, 646 Cloudman, A. M., 105, 115,135,162, 167 Clowea, G. H. A,, 81, 100, 347, 348, 392, 474,499 Cohen, A,, 381, 391 Cohen, B., 411, 412, 447 Cohen, J. A., 442, 4.47 Cohen, L., 381,392 Cohen, P. P., 371,378,381, 391, 396, 513, 524,640, 646 Cohn, C., 507, 640
AUTHOR INDEX
553
Cohn, E. J., 510, 511, 512, 517, 519, 521, Culvenor, C. C. J., 433, 447 Cunningham, L., 302, 307, 317, 336,373, 528, 640 Cohn, W. E., 312, 319, 334 376, 378, 391 Cunningham, R. S., 208(215, 216), 228 Cole, R. K., 106, 135, 160, 164 Curtis, F. C., 524, 642 Cole, W. H., 490, 498 Collins, V. J., 211(233), 228 Curtis, M. R., 201(147), 209(228), 216, Collip, J. B., 201(144), 202, 203(144, 220(335), 221(147), 226, 228, 230, 231, 381, 391, 456, 468, 498 163), 216, 226, 227 Coman, D. R., 93, 99 Czaczkes, J. W., 474, 498 Conway, B. E., 419, 447 Cook, H. A., 377, 391 D Cook, J. W., 280, S3g,359, 380, 381, 382, Dahl-Iversen, E., 208(219, 220), 228 384, 385, 390, 391, 491, 498 Dalton, A. J., 108, 166, 163, 168, 198 Coolen, M. L., 123, 168 Cooper, G. R., 253, 269, 520, 646 (107), 226, 345, 391, 487, 600 Cooper, 2. K., 64, 66, 68, 69, 70, 95, Daly, B. M., 533, 642 99, loo, 101 Damgaard, E., 533, 647 Dann, T. B., 200(127), 266 Copeland, D. H., 388, 391, 478, 498 Danysz, S., 123, 167 Cori, C. F., 120, 16'6, 212(250), 229 Darlington, C. D., 157, 162, 264, 266 Cornil, L., 530, 646 Dascomb, H. E., 248, 670 Cornman, I., 332, 536 Cortell, R., 345, 346, 391 Dauben, W. G., 280, 283, 284, 336 Costerousse, O., 328, 356 Daudel, P., 7, 25, 38, 47, 64, 66, 293, 328, Cottral, G. E., 263, 266 336 Daudel, R., 6, 7, 15, 16, 20, 25, 33, 38, Cottrell, T. L., 23, 66 47, 49, 50, 51, 64, 66,66, 328, 336 Coulson, C. A., 8, 20, 23, 24, 25, 26, 27, Davidsohn, I., 506, 514, 641 29, 30, 31, 33, 40, 46, 47, 64, 66 Cowdry, E. V., 62, 63, 65, 66, 70, 71, 72, Davidson, I., 135, 162 73, 74, 76, 81, 87, 92, 94, 95, 96, 98, Davidson, J. N., 309, 310, 3S6, 373, 391 99,100,101 Davies, M. C., 252, 669 Cox, A. J., 354, 396 Davies, W., 406, 433, 447 Davis, J. S., 518, 523, 646 Cox, E. C., 29, 66 Cox, H. R., 252, 266, 269 Davis, J. W., 433, 447 Crabtree, C. E., 221(353), ,232 Davis, S. B., 409, 447 Crabtree, H. G., 379, 391 Davis, W., 400, 409, 414, 419, 424, 428, 442, 443, 446, 447 Craig, D. P., 15, 16, 24, 66 Craigie, J., 136, 166 Day, P. L., 347, 348, 391 Cramer, W., 63, 66, 67, 73, 74, 90, 96, 99, Deansley, R., 202, 627 Deasy, C. L., 307, 334 100, 101, 120, 162, 201, 226 Crandall, D. I., 417, 448 de Bruyn, W. M., 244, 266,266 Crawford, V. A., 17, 27, 28,66 De Eds, F., 354, 396 Deihl, D. G., 259,266 Cremer, H. D., 511, 640 de Jongh, S. E., 122, 166, 189(54), 212, Crockston, J., 531, 639 Cronin-Lowe, E., 528, 641 224, 229 De Leon,H. P., 185(34), 186(34), 923 Crossen, R. J., 210, 229 Delluva, A. M., 318, 334, 417, 44.8 Crossett, A. D., Jr., 217(313), 231 Crossley, M. L., 399, 436, 447, 448, 515, Delrue, G., 527,628,646 Denton, R. W., 527, 64.9 641 De Ome, K. B., 106, 107, 110, 112,137, Crowell, M. F., 458, 499 Csaky, T. Z., 263, 266 166
554
AUTHOR INDEX
Deringer, M. K., 106, 107, 109, 111, 113,
121, 126, 130, 153, 155, 157, 159, 163, 166, 213(275), 217(275), 218 (323), 230,231 Derouaux, G., 510, 511, 512, 521, 640 Des Ligneris, M. J. A., 261, 262, 867 Desruisseaux, G., 530, 646 Deuel, H., 435, 447 Devik, F., 483, 484, 485, 498 Devor, A. W., 510, 521, 523, 536, 648 Dewar, M. J. S., 6, 11, 12, 15, 16, 24, 66 DeWitt Fox, J., 234, 267 Dexter, S . O., 384, 386, 387, 392 Dick, W., 23, 66 Dickie, M. M., 201(130), 226 Diddle, A. W., 213(265), 289 Dillard, G. H. L., 508, 529, 531, 533, 6-41 Dillon, E. S., 528, 646 Dillon, M. L., 528, 646 Dinning, J. S., 347, 348, 391 Dippel, R. V., 157, 1'70 Dirksen, A. J., 516, 518, 646 Dische, Z., 538, 641 Dixon, M., 441, 442, 447 Dixon, T. F., 157, 163 Dmochowski, L., 106, 107, 108, 110, 112, 113, 116, 117, 118, 120, 123, 124, 125, 126, 132, 133, 134, 135, 136, 138, 139, 140, 145, 146, 147, 148, 149, 150, 151, 152, 153, 154, 155, 157,163, 169, 204 (184), 205, 218 (325, 327) , 220 (339), 887, 231, 249, 250,252, 260, 262,267 Dobriner, K., 288,336,359,360,363,384, 396 Dobrovolska'La-ZavadskaSa,N., 486, 498 Doering, W. E., 401, 402, 403, 404, 405, 407, 446 Doerr, R., 235, 236, 267 Dole, V. P., 507, 510, 641 Doljanski, L., 239, 244, 262, 267, 269 Doniach, I., 64, 99 Dorfman, R. I., 120, 163, 199(121), 226, 534, 641 Dotti, L. B., 180(18), 223 Dougherty, T. F., 204(176, 179), 205,227 Dowdy, A. H., 527, 642 Drill, V. A., 194(79), 286 Druckrey, H., 343, 345, 346, 353, 362, 690, 691 Dry, F. W., 65, 99
Dublin, L. I., 489, 498 Dubnik, C. S., 108, 162, 163, 168 Duboff, G., 534, 642 Du Bois, K. P., 387, 394 DU B U ~ , H. G., 157,163, i r i , 278 Duffy, E., 491, 498 Dulaney, A. D., 146, 163 Dunlap, C. E., 281, 336 Dunn, J. B., 221 (356), 838 Dunn, T. B., 218(32!), 231, 379, 390 Dunn, Th. B., 104, 106, 124, 125, 126,
132, 133, 148, 149, 150, 151, 152, 153, 154, 155, 169,163, 164, 166 Dunn, W. J., 136, 168 Dunning, W. F., 201(147, 148), 202, 216, 220(335), 221(147), 286, 230, 231, 456, 468, 498 Duran-Reynals, F., 156, 164, 235, 236, 237, 238, 239, 243, 250, 252, 255, 259, 264, 267, 869, 870, 525, 646 Durram, E. L., 511, 641 Dustin, P., Jr., 398, 447 Duthie, E. S., 533, 534, 641 du Vigneaud, V., 290,334, 347, 364, 390, 396, 405, 409, 411, 441, 447, 449, 478, 498 Dvoskin, S., 517, 641 Dworecki, I. J., 505, 516, 648
E Eaves, G., 139, 157, 169 Ebeling, A. H., 243, 266 Eckert, E. A., 263, 866 Eddy, M. S., 134, 143, 150, 166 Edsall, J. T., 506, 510, 511, 641 Edwards, J . E., 198(107), 226, 343, 344,
346, 346,348,379, 380,390, 391, $96
Eggers, H. E., 260, 269 Ehrlich, P., 104, 117, 164 Ehrstrom, M. C., 518, 519, 641 Eidson, M., 337 Eisa, E. A., 466, 498 Eisen, M. J., 216, 230 Eisman, P. C., 529, 644 Eitelman, E. S., 510, 520, 524, 525, 6.64 Ekert, B., 293, 336 Elford, W. J., 255, 267 Ellerbrook, L. D., 517, 518, 519, 523, 524, 525, 526, 530, 641, 647
AUTHOR INDEX
555
Elliot, D. W., 135, 162 Ellis, F. W., 533, 641 Ellis, G. H., 458, 499 Elmore, D. T., 398, 410, 411, 447 Elson, L. A., 65, 100, 358, 384, 385, 391, 483, 484, 485, 498 Elwyn, D., 319, 323, 536 Emerson, G. A., 486, 600 Emmel, V. M., 221(354), 232 Enders, J. F., 525, 641 Enenkel, H. J., 511, 647 Engel, R. W., 388, 391, 469, 478, 4.98 Engelbreth-Holm, J., 123, 164, 241, 258, 263, 267, 270, 271 Engelman, M., 319, 336 Engle, E. T., 185, 207(207), 223, 228 Entenman, C., 279, 313, 338 Erf, L. A., 310, 311, 337 Erickson, J. O., 520, 646 Eriksen, N., 519, 525, 641 E r s t , T., 536, 644 Ernstene, A. C., 641 Erwin, C. P., 505, 535, 536, 537, 646 Erxleben, H., 530, 642 Eschenbrenner, A. B., 124,168,218(324), 231, 353, 355, 356, 362, 387, 306 Euler, H. von, 314, 315, 834, 336, 530,
Finnegan, J. V., 517, 525, 641 Firminger, H. I., 69, 99 Fischer, A., 243, 244, 267 Fischer, B., 341, 591 Fischer-Wasels, B., 341, 391 Fisher, J. C., 92, 93, 100 Fisher, R. A., 461, 498 Fishler, M. C., 279, 313, S38 Fishman, W. H., 530, 641 Fisk, A. A., 207(212), 208(212), 228 Flaks, J., 200(124), 226 Fleming, D. S., 409, 448 Flon, M., 293, 336 Flory, C. M., 482, 498 Folley, S. J., 212(248, 249), 229 Fontaine, R., 527, 641 Forbes, E. B., 471, 498 Forbes, T. R., 217(315), 231 Ford, C. E., 398, 448 Foster, G. L., 278, 536 Foster, J. V., 505, 535, 537, 646 Foulds, L., 148, 149, 151, 154, 164, 217 (314), 219(329), 231, 233, 246, 249, 260, 262, 263, 267 Fox, M., 429, 446 Fraenkel, E., 235, 236, 246, 255, 256, 267 Fraenkel-Conrat, H., 367, 391, 434, 441,
Evans, C. A., 116, 141, 161 Evans, H. M., 185,194(76,77), 198(105), 203(165, 166), 206(105, 198, 199), 223, 225, 227, 228 Evans, R., 66, 95, 99 Everett, J. L., 414, 419, 425, 443, 446,
Francis, G. E., 398, 405, 407, 409, 410, 428, 437, 442, 446, 447 Frankel, S., 77, 78, 82, 96, 100 Franklin, H. C., 64, 95, 99 Frants, I. D., 305, 338, 378, 396 Frantz, M., 121, 131, 164, 167, 188(43), 198(108, 109, 110), 199(110), 224, 226 Frederiksen, O., 258, 267 French, C. E., 471, 498 Freund, E., 530, 641 Fridericia, L. S., 478, 498 Friedberg, F., 301, 337 Friedman, 0. M., 330, 337 Friedmann, B., 284, 337 Friedmann, R., 535, 641 Friedwald, W. F., 247, 256, 267, 464, 498, 526, 641 Fruton, J. A., 398, 402, 407, 415, 416, 417,418, 448 Fruton, J. S., 400, 401, 402, 407, 412, 413, 437, 448, 449
639, 641
47,448
Everett, M. R., 505, 535, 536, 537, 646 Ewing, J., 117, 164
F Falk, H. L., 31, 66 Farber, E., 306, 336 Fekete, E., 105, 106, 112, 113, 115, 120, 121, 127,130, 131, 132, 133, 164, 172, 198(111, 112, 113, 114), 199, 214 (281), 226, 226, 250 Felsher, R., 524,646 Ferguson, J. H., 533, 6.44 Fevold, H. L., 207(206), 268 Findlay, G. M., 239, 263, 267 Finkelstein, H., 257, 267
448,449
556
AUTHOR INDEX
Fry, H. J. B.,526, 641 Fuchs, H.J., 531, 641 Fujiwara, T.,372,391,392,394 Fukuoka, F., 515,646 Fuller, R.H.,108, 164 Fulton, J. K.,534, 641 Furet, 8. S.,319,320,396 Furtado Dias, M. T., 195(94), 212(93, 94), 996 Furth, J., 106, 135, 147, 164, 166, 186, 190(58, 59, 60), 191, 193(67, 68, 69, 70), 194(67, 69),203(160), 204(185), !205(185), 206(58, 185), 884, 887, 260, 262, 263, 267, 268, 457, 482, 498,600
Furth, 0. B.,190(58), 191, 206(58), 224 Fuson, R.C.,413,448 G
Gaarenstroom, J. H., 189(54), 284 Gaines, J. A., 191, 192(65), 884 Gallico, E.,378, 391 Galloway, J., 259, 268 Gens, P.,189(53), 197, 294 Gant, J. C., 505, 519,646 Gard, S., 248, 249, 267 Gardner, R. F., 249,250,268 Gardner, W.U., 119, 120, 121, 123, 127, 128,131,135,148,155,169,163,164, 174(1, ll), 179(11), 180(11, 15, 16, 19), 185(33), 186, 187, 188(33), 190 (55), 191(62), 192(62, 63), 194(81), 195, 196(91, 97, 98),197(102a, 103), 198(104), 199, 201, 202(150, 1551, 204(87, 174, 175, 176), 205, 206(63, 192), 207(176), 208(214), 209(229, 230), 210, 211(229, 232, 233, 234), 213(260, 265, 266), 214(279, 280, 282), 215(15), 216(309), 218(149), 219(282), 221(348, 352, 354), 223, 224, 886, 226, 358,380,396 Gasson, E. J., 444,4.4.ff Gay, O., 185(34), 186(34), 223 Gedigk, P.,506, 518, 519,647 Geiser, R. C., 996 Geisse, N. C., 487,49G Geist, 8.H., 191, 192(65), 224 Gellhorn, A.,487,600 George, C.,533,644
Geren, B. B., 70, 99 German, W.M.,526,643 Gersh, I., 536, 641 Geschickter, C. F., 215(295), 216, 230, 527,528, 646 Gessler, A. E.,139, 158,164, 166 Giese, J. E.,343,347,349, 356, 391,477, 498 Gilbert, C.,381,391,490, 498 Gilbert, L. A., 419,447 Gillespie, H.B., 319,336 Gillespie, J. M., 510, 511, 512, 521, 640 Gilligan, D.R.,505, 517, 525, 641 Gillman, J., 381, 391, 490,498 Gillman, T.,381,991,490,498 Ginzton, L. C.,210,229 Ginaton, L.L.,123,169 Giroud, A.,73, 100 Gjessing, E. C., 417,419,447,448 Glass, G. B. J., 505,516, 525,641, 648 Glick, D., 132, 164, 249, 269, 505, 523, 534, 648, 643 Glucksmann, A,, 88, 100 Godfried, E.G., 522,648 Gogek, C.J., 281,336 Goiffon, R.,522,648 Goldacre, R. J., 442,448 Goldberg, R.C., 203(161, 162),887 Goldblatt, H.,348, 391 Golden, F., 510, 523,528, 644 Golden, J. R., 185, 283 Goldhaber, G., 262, 269 Goldner, M.S.,529,640 Goldschmidt, S.,524, 642 Goldsmith, Y.,146, 163 Goldstein, A.,518, 648 Goltz, H.L., 110, 171 Golub, 0.J., 252, 871 Golumbic, C., 398, 412, 413, 415, 418, 437,444 449 Good, R. A., 523,534,642,643 Good, 'f. A., 132, 164 Goodall, K.,377, 39W Gorbman, A.,202(159),887 Gorer, P.A., 115,135,145,164,166,167, 220(341), 231 Gottlieb, H.,519,648 Gottschalk, R.G., 246,260,262,263,264, 867
557
AUTHOR INDEX
Grady, H. G., 120, 128, 135, 170, 195, 204(87), 205, 226, 379, 390 Graff, A. M., 319, 336 Graff, M. M., 523, 642 Graff, S., 133, 136, 138, 140, 166,319,336 Graffi, A., 67, 100 Gram, H. C., 505, 524, 525, 642 Grassmann, W., 511, 642 Gray, S. J., 525, 642 Green, C. V., 120, 164 Green, R. G., 106, 116, 132, 136, 141, 142, 144, 145, 146, 156, 161, 166, 171 Green, J. W., 483, 485, 498 Greenberg, D. M., 298, 301, 302, 306, 307, 336, 336, 337, 364, 396, 515, 646 Greene, H. S. N., 201(133, 134, 135), 207(211, 212, 133, 134)) 208(211, 212), 211(211), 213(135, 264), 296, 228, 629, 239, 270 Greene, R. R., 174(3), 185(37), 186(37), 223, 924 Greenfield, R. E., 515, 642 Greenspan, E. M., 221(355), 232, 523, 642 Greenstein, J. P., 78, 87, 100, 108, 166, 170,275,336,340,341,378,391,452, 498, 515, 642 Greenwald, A. E., 519, 525, 646 Greenwood, H. H., 8, 27, 28, 30, 31, 35, 40, 41, 42, 48, 49, 64, 66 Greep, R. O., 189(51), 207(206), 224, 228 Gregory, J. E., 234, 867 Greig, J. R., 150, 156, 166 Greulich, W. W., 219(332), 231 Greville, G. D., 442, 447 Grey, C. E., 139, 164 1 Griffin, A. C., 290, 336, 302, 307, 317, 336,347, 348, 349, 360, 368, 369, 373, 376, 377, 378,391, 396, 474, 477, 498 Grob, D., 533, 642 Grollman, A., 518, 642 Gross, J., 214(289), 230 Gross, L., 135, 156, 158, 166, 526, 640, 642 Grubgeld, L. E., 527, 643 Gruenstein, M., 284, 337 Grunwald, E., 433, 448 Gruschow, J., 318, 336 Gruskin, B., 505, 526, 527, 642
Gudjonsson, S., 478, 498 Guerin, M., 201(137, 138), 226,245, 251, 255, 261, 262, 263, 269, Cuerin, P., 201(137, 138), 226, 251, 261, 269 Guest, M. M., 533, 642 Guggenheim, K., 474, 498 Gulland, J. M., 398, 410, 411, 447 Gumbreck, L. G., 189(48), 224 Curd, F. R. N., 510, 511, 512, 521, 640, 644 Gurin, S., 417, 448 Guthneck, B. T., 377, 396 Gutman, A. B., 504, 505, 506, 507, 510, 528, 529, 642, 646 Gutmann, H. R., 288, 336 Gye, W. E., 106, 123, 136, 147, 163, 166, 247, 256, 258, 267 Gyorgy, P., 348, 349, 391, 392
H Haagensen, C. D., 106,112,114,133,136, 138, 140, 142, 166, 207(207), 998 Haagen-Smit, A. J., 307, 334 Haaland, M., 117, 127, 166 Haddow, A., 50, 66, 65, 100, 157, 166, 243, 264, 268, 280, 336, 353, 354, 399, 399, 421, 436, 442, 444, 445, 448, 484, 485, 498 Hagopian, F., 120, 160 Hahn, A., 505, 522, 642 Hahn, L., 312, 336 Hahn, P. F., 275, 336 Hakanson, E. Y., 505, 534,64Z Hall, G. G., 52, 66 Hall, G. H., 527, 642 Halter, C. R., 346, 352, 353, 355, 358, 362, 377, 387, 392,396,474,'477, 499 Halvorson, H. O., 146, 166, 252, 266 Ham, A., 116, 141, 160 Ham, T. H., 524, 64.2 Hamburger, C., 208(219, n o ) , 228 Hamilton, J. B., 178(14), 923 Hammett, F. S., 158, 166 Hammett, L. P., 433, 44.8 Hamperl, H., 345, 346, 390 Hanby, W. E., 411, 412, 413, 448 Hane, S., 290, 336, 360, 368, 369, 396 Hanke, M. E., 524, 642
558
AUTHOR INDEX
Hankwitz, R. F., 525, 647 Harde, E., 213(268), 629 Harman, J. W., 345, 346, 394 Harris, J., 411, 412, 447 Harris, M., 526, 646 Harris, P. N., 347,348,392, 474,478,439 Harris, R. J. C., 238, 242, 249, 252, 254, 255, 257,259, 260, 866, 268 Harris, R. J., 349, 353, 354, 382, 392 Harris, T. N., 534, 646 Hart, R. W., 529, 640 Hartley, G. S., 411, 412, 413, 448 Hartwell, J. L., 4, 66, 341, 343, 352, 353, 379,384, 392 Hasirnoto, H., 372, 393 Haslewood, G. A. D., 380, 391 Hatschek, R., 530, 647 Hausmann, R., 530, 647 Havel, J., 245, 255, 269 Hawk, B. O., 517, 525,641 Hawkins, J. A., 256, 268 Heath, N. S., 433, 447 Heep, W., 341, 392 Heidelberger, C., 274, 276, 279, 280, 281, 283, 284, 285,280,287, 288, 289, 290, 291, 294,296, 300,302, 319,320, 333, 334, 336,336, 337, 364, 389, 392, 393, 396 Heidelberger, M., 513, 646 Heilman, F. R., 141, 161, 166, 207(200), 228
Heiman, J., 215(304), 210(311), 230, 231 Heinrich, M., 318, 336 Heller, C. G., 189(56), 624 Hellwig, C. A., 139, 166 Helmer, 0. M., 250, 259, 668, 269 Hendry, J. A., 436, 442, 4 9 Henle, G., 248, 871 Henle, W., 248, 250, 253, 266, 271 Henry, R. J., 517, 519, 523, 525, 64.9 Henshaw, P. S., 206(190), 627 Herbert, F. K., 529, 6-42? Herken, H., 530, 642 Herriott, R. M., 258, 666, 398, 408, 409,
w
Hertz, R., 174(10), 223 Hem, E. L., 538, 639 Heston, W. E., 105, 100, 107, 108, 109, 110, 111, 113, 118, 121, 124, 129, 130, 153, 155, 157, 158, 159,163, 166, 166,
168, 171, 200(127), 213(275), 215, (291), 217(275, 291), 218(323, 324), 626, 230, 231, 457, 475, 487, 499, 600, 601 Hevesy, G., 276, 312, 313, 314, 315, 320, 334, 336 Hewett, C. L., 359, 380, 381, 382, 384, 385,390,391 Hieger, I., 380, 391 Higgins, H., 377, 392 Hilfinger, M. F., 378, 396 Hilliard, J., 505, 533, 647 Hinsberg, K., 524, 535, 642 Hirshfeld, S., 534, 642 Hirst, J., 400, 448 Hisaw, F. L., 180(17), 223, 207(206), 228 Hitchcock, C. R., 302, 334, 336 Hoagland, C. L., 255, 268 Hoan, N., 293,336 Hoch, H., 377, 392, 520, 643 Hoch-Ligetti, C., 216(312), 231, 343, 358, 377,384,385,391,392,491,499,520, 643 Hodges, C. V., 529, 530, 643 Hofer, M. J., 377, 896 Hoffman, D. C., 259, 26'8 Hoffman, F. L., 489, 499 Hoffman, R. S., 244, 267 Hoffstadt, R. E., 260, 268 Hogberg, B., 530, 639, 641 Hogeboom, G. H., 249, 668, 270 Hognew, K. R., 507, 508, 510, 523, 646 Hoilomon, J. H., 92, 93, 100 Holmes, B. E., 258, 269, 315, 336 Holmgren, N., 527, 643 Homburger, F., 506, 513, 514, 530, 641, 643 Hooker, C. W., 194(74, 79), 195, 197(83), 264, 666, 203, 270 Hopwood, F. L., 398, 409, 410, 437, 446 Horne, H. W., 158, 166 Homing, E. S., 120, 162, 201, 204(184), 205,220(337,338, 339)) 226,227, 231 Horwitt, B. N., 377, 396 Hospelhorn, V. 0.) 505, 516, 518, 519, 520,643 Hoster, H. A., 139, 166 Hoster, M. S., 139, 166 Hotchkiss, R. D., 249, 268 Howe, P. E., 506, 643
559
AUTHOR INDEX
Hoyle, L., 163, 248, 868 Huddleson, I. F., 526, 646 Htickel, E., 21, 28, 66 Huff, J. W., 323, 337 Huggins, C., 219(333), 231, 504, 516,517, 518, 519, 520, 529, 530, 643 Hughes, E. D., 420, 448, 510, 517, 519, 520, 528, 640, 643 Hummel, K. P., 117, 133, 134, 143, 150, 151, 166 Humphrey, J., 644 Humphreys, E. M., 483, 498 Hungate, R. E., 141, 166, 171 Hunter, A., 489, 499 Hunter, S. W., 302, 336 Hurlbert, R. B., 302, 325, 336, 336 Huseby, R. A., 107, 108, 110, 111, 122, 125, 128, 129, 130, 131, 137, 138, 139, 143, 160, 162, 166, 200(125), 212, 213(255), 213(276), 214 (286, 287), 217(316), 218(276), 226, 229, 230, 231, 314, 326, 334, 466, 499 Hutchinson, M. C., 511, 646 Hutchison, 0. S., 331, 336 I Iki, H., 372, 393 Imagawa, D. T., 136, 145, 146, 162, 166 Inkley, J. J., 517, 525, 641 Introzai, P., 398 Israel, H. L., 537, 643 Ivy, A. C., 527, 643 Izar, G., 526, 639
J Jablonski, C. F., 382, 392 Jackson, J., 120, 148, 160 Jacobi, H. P., 365, 379, 390, 398, 468, 469, 482, 497, 499 Jacobs, J., 23, 24, 25, 33, 64, 66 Jacques, R., 16, 66 Jager, B. V., 507, 513, 523, 643 James, W. H., 471, 498 Janeway, C. A., 526, 640 Janota, M., 526, 639 Jean, M., 16, 20, 66 Jeener, R., 250, 268 Jefferies, M. E., 185(37), 186(37), 224 Jeffrey, G. A., 29, 65
Jenrette, W. V., 515, 642 Jensen, E. V., 505, 516, 517, 518, 519, 520, 643 Jensen, K. A,, 445, 448 Jobling, J. W., 235, 250, 256, 268, 270 Johnson, B. A., 235, 250, 270 Johnson, J. M., 482, 601, 530, 644 Johnson, M. J., 530, 640 Johnson, P., 136, 138, 139, 157, 169 Johnson, R. O., 457, 469, 473, 483, 500 Jones, E. E., 120, 166, 215(302), 230 Jones, F., 515, 639 Jones, H. B., 281, 284, 285, 301, 315, 336, SS6 Jones, R. N., 44, 66, 281, 336 Jordan, D. O., 398, 410, 411, 447 Jorgensen, H., 208(220), 228 Judd, T., 487, 499 Jungck, E. C., 189(56), 224
K Kabat, E. A., 147,166,260,263,668,507, 510,513,525, 528,642, 643,646, 647 Kadesch, R. G., 448 Kahler, H., 132, 136, 137, 141, 168, 166 Kahn, H., 505, 514, 524, 526, 642, 643 Kahnt, F. W., 510, 511, 512, 521, 640 Kamen, M. D., 75,100,157, 170,276,336 Kaminer, G., 530, 641, 543 Kaplan, H., 198(104), 226 Kaplan, H. S., 123, 167, 191(62, 62a), 192(62), 206(191, 193), 207(203), 984, 227, 228 Kaplan, M. H., 533, 643 Karnofsky, D. A., 508, 646 Kay, H. D., 505, 529, 643 Keighley, G., 307, 334 Kekwick, R. A., 528, 643 Keller, E. B., 307, 336 Kellcy, L. S., 315, 336 Kelley, V. C., 523, 534, 642, 643 Kelly, K. H., 505, 535, 537, 646 Kelly, M. G., 108, 171, 457, 601 Kelsch, J. J., 139, 164 Kendall, E. C., 207(200), 928 Kennaway, E. L., 359, 380,381, 382,384, 385, 390, 391, 490, 499 Kennaway, N. M., 358, 380, 381, 382, 384, 385, 390,3091
560
AUTHOR INDEX
Kensler, C. J., 341, 346, 347, 349, 352, 353, 355, 358, 362, 363, 364, 377,378, 386, 387, 388, 392, 393, 396, 474, 477, 478, 499 Kerwin, J. F., 413, 448 Keye, W. R., 617, 533,647 Keys, A., 507, 647 Khanolkar, V. R., 108, 166,214(285) Khanolkar, V. T., 111, 128, 166 Kidd, J. G., 91, 100, 247, 248, 867, 868, 270, 464, 600, 526, 641, 643,64.6 Kiefer, L., 517, 64.3 Kielley, M., 298, 336 Kienle, R. H., 515, 641 Kilpatrick, M., 429, 430, 434, 447 Kilpatrick, Mary, 429, 430, 432, 447 King, C. J., 452, 458, 462, @9 King, J. T., 108, 121, 166, 214(288), 830 King, J. W., 239, 867 King, T.T., 108, 162 Kingsley, G. R., 507, 643 Kinosita, R., 341, 343, 344, 345, 346,352, 353, 356, 360, 362, 380, 387, 392 Kiprianov, A., 434, 448 Kirby, A. H. M., 344, 355, 356, 379, 380, 381, 382, 392 Kiriluk, L. B., 534, 643 Kirk, I., 445, 448 Kirk, M. R., 283, 336 Kirkman, H., 221(357, 358), 232 Kirschbaum, A,, 121, 123, 124, 125, 126, 128, 131, 135, 155, 168, 164, 167, 188 (43), 191(61), 198(108, 109, 110), 199(110), 200(126), 204(172, 175), 205, 206(188), 218(322), 224, 226, 226, 227, 231, 263, 270, 469, 487,499 Kirtz, M. M., 121, 168, 212, 229 Kishi, S., 372, 391, 398, 394 Kit, S., 298, 306, 307, 336,336 Kits von Waveren, E., 244, 266 Klatt, T., 516, 518, 646 Klein, M., 187, 224 Kleiner, I. S., 505, 519, 525, 644 646 Klinck, G. H., 213(261, 263), 929 Kline, B. E., 341,346,348,349,350,362, 376, 377, 379,390, 392, 393,396, 452, 457, 468, 469, 470, 471, 473, 474, 478, 479, 483, 484, 499, 600 Klotz, I. M., 518, 643 Klumpar, J., 522, 640
Knight, C. A., 249,251,253,254.261.268 Knight, R. A., 147, 167 Knorr, L., 434, 448 Knowlton, N. P., Jr., 70, 100 Knox, J. C., 478, 600 Knox, R., 260, 262, 267, $68 Kocher, R. A,, 482, 499 Kock, A. M., 438, 449 K(lg1, F., 304, 336 Kohman, T. P., 311, 336 Koller, P. C., 483, 484, 485, 498 Kon, G. A. R., 353, 354, 392, 399, 442, 443, 444, 448 Koneff, A. A., 203(165, 166), 206(199), 297, 228 Koomen, J., 248, 270 Kopac, M. J., 419, 4.48 Kopacaewski, W., 505, 528, 643 Kortweg, R., 105, 106, 107, 112, 118, 121, 122, 127, 129,166,166, 167, 212, 214 (283, 284), 829, 930,244, 266 Kraemer, Dorothy Ziegler, 71, 100 Krahl, M. E., 347,348, 398, 474, 499 Krakower, C., 207(207), 228 Krejci, L. E., 526, 643 Kremen, A. J., 302, 334, 336, 534, 643 Krimsky, I., 250, 268, 269 Krueckel, B. J., 328, 337 Krysa, H. F., 66, 100 Kiipfmuller, K., 343, 391 Kuhn, R., 357, 387, 393 Kung, S. K., 65, 73, 74, 92, 100 Kunkel, H. G., 524,643 Kupke, D. W., 317, 336 Kupske, 0. W., 376, 391 Kurzrok, R., 209(2!26), 2.28 Kuttelwascher, H., 526, 646 Kynette, A., 141, l7f L
Lacassagne, A., 38, 64, 105, 122, 123, 167, 174(4, 5), 204(173), 205, 210, 213 (267, 269), 215, 216(305, 306, 307, 308), 223, 227, 830, 259, 868, 328, 336 La Due, J. S., 517, 641 Lafay, B., 240,269 Lafaye, J., 364, 396 Laird, A. K., 376, 394
AUTHOR INDEX
561
Lamerton, L. F., 483, 484, 485, 498 Levinson, S. A., 527, 643 Landau, J. L., 526, 643 Lewis,J. H., 533, 6.44 Landsteiner, K., 261, 269 Lewis, M. R., 249, 250, 257, 268, 399, 436, Lane, C. E., 196(95), 226 448 Langemann, H., 378, 388, 392, 393 Li, C. H., 194(76,77), 198(105),203(166), Langer, E., 345, 393, 535, 644 206(105, 198, 199), 826, 227, 228 Langham, W. H., 331, 337 Li, M. H., 186, 187, 191, 192, 193(71), Lansing, A. I., 74, 75, 100 197(102a), 198(104, 105), 224 Lardy, H. A., 364, 394 Liang, Hsu-mu, 64, 73, 95, 100 Larizza, P., 522, 644 Lichtenstein, H. J., 430, 448 Larsen, C. D., 457, 475, 499 Lick, L., 191, 224 Laser, H., 243, 244, 267 Liebler, J. B., 529, 640 Lasnitzki, I., 136, 169 Lindegren, C., 157, 170 Lathrop, A. E. C., 105, 119, 120, 167 Lindegren, G., 157, 170 Lauffer, M. A., 253, 266 Linsteed, R. P., 401, 402, 403, 404, 405, Lavin, G. I., 255, 268, 288, 336, 376, 394 407, 446 Lavik, P. S., 468, 469, 470, 479, ,499 Lippincott, S. W., 519, 525, 526, 530, Law, L. W., 104, 106, 120, 132, 135, 145, 641, 647 166, 167, 205(189), 206(189, 195), Lipschtitz, A., 174(6), 185(34), 186(34), 227, 344, 379, 380, 393 208(6, 217), 223, 228 Lawrason, F. D., 123, 167, 469, 499 Lischer, C. E., 65, 73, 96, 100 Lawrence, J. H., 310, 311, 337 Little, C. C., 104, 105, 106, 109, 111, 112, Lawrence, J. S., 528, 639 113, 115, 117, 118, 120, 130, 131, 132, Lawson, W. E., 406, 448 133, 134, 143, 148, 150, 151, 162, 164, 166, 167,169, 171, 172, 198(111, 112, Leblond, C. P., 68, 73, 100 Lecocq, J., 38, 47, 64 113, 114, 116, 117), 199(118, 119, Leese, A., 478, 600 122), 226, 226 Lefavre, H., 123, 167 Little, M. S., 517, 519, 523, 525, 642 Le Fevre, R. G., 490, 600 Little, P. A,, 240, 268, 269, 486, 499 Lehman, I., 523, 642 Littlefield, W., 517, 643 Lehmann-Facius, H., 526, 644 Liu, C. H., 510, 511, 512, 521, 640 Lehninger, A. L., 299, 337, 505, 530, 646 Llombart, A., 243, 268 Leidler, H. V., 194(75), 226 Loeb, L., 105, 117, 119, 120, 121, 122, Leiner, G., 536, 644 167, 168, 174(7), 210, 212(251), 213 (270, 273, 274), 214(251, 278), 223, Leiter, J., 326, 337 Le Maistre, J. W., 433, 448 229, 230 Lennard-Jones, J. E., 21, 24, 52, 66 Loeser, A. A., 215(303), 230 Lennette, E. H., 261, 268 Loewenthal, H., 258, 268 Leonard, A., 517, 646 Loftfield, R. B., 297, 305, 307, 338, 378, 396 Le Page, G. A., 280, 300, 302, 319, 320, 336,336, 337, 378, 396 Long, M. L., 204(177, 178), 205, 227, 481, Leseur, A., 157, 170 497 Long, S., 530, 646 Leslie, I., 309, 310, 336 Levaditi, C., 259, 268 Longsworth, L. G., 508, 510, 521, 528, Lever, W. F., 510, 511, 512, 521,640, 644 536, 644 Levi, A. A., 288, 336 Longuet-Higgins, H. C., 23, 24, 25, 26, Levillah, W. D., 106, 126, 153, 155, 166, 27, 40, 46, 47, 64, 66 218(323), 231 Loomis, E. C., 533, 644 Levina, M., 535, 640 Lord, D. D., 400, 448 Levine, M., 244, 256, 268 Lorens, E., 124, 168
562
AUTHOR INDEX
Lorenz, E., 218(324), 231, 255, 266 Lorenz, L., 533, 534, 541 Lotthammer, R.,524, 544 Loveless, A., 398, 399, 437, 442, 445, 446,
w
Lowenhaupt, E., 343, 393 Lowenstein, B. E., 519, 525, 646 Lowy, P. H., 307, 334 Luck, J. M., 302, 307, 336, 373, 376, 377, 378, 991 Luce-Clausen, E. M., 64, 100 Ludford, R. J., 117, 168, 243, 268 Ludwig, H., 519, 64.2 Luetscher, J. A., 507, 508, 520, 644 Luria, S. E., 246, 268 Lusbaugh, C. C., 483, 485, 498 Lustig, B., 535, 536, 644,64.5
M Ma, Chung K., 74, 100 Mabee, D., 283, 284, 936 McBee, B. J., 139, 166 McCarthy, E. F., 507, 646 McCarty, K., 139, 164 McCarty, K. S., 158, 166 McCawley, E. L., 137, 162 McCay, C. M., 458, 499 McClellan, V., 507, 510, 528, 642 McCombie, H., 407, 444, ,447, 4.48 McCord, W. M., 232 MacDonald, J. C., 343, 361, 364, 393 McEleney, W. J., 106, 107,142,169,160, 217(317), 220, 231 McEndy, D. P., 204(185), 205(185), 206(186), 2.27 McEuen, C. S., 201(144), 202, 203(144, 163), 206(197), 209(197), 216, 228 McFarlane, A. S., 513, 644 MacFarlane, R. G., 531, 533, 644 MacInnes, D. A., 608, 521, 528, 536, 644 McInnes, G. F., 517, 640 McIntosh, J., 240, 252, 255, 258, 261, 266, 268, 269, 270 MacIntyre, A., 531, 639 McKennis, H., 405, ,449 MacKenzie, I., 464, 499,526, 644 Mackenzie, R. D., 235, $69 McKewon, T., 219(334), 231 Macklin,LlM., 158, 168
McLean, I. W., 253, 256, 270 MacLeod, C. M., 531, 640 McQuarrie, I., 523, 64.3 Madden, 8. C., 514, 515, 644 Magat, M., 45, 66 Magill, J. W., 364, 392 Magnus, H. von, 156, 168 Magnus, P. von, 156, 168, 248, 249, 267 Maguigan, W. H., 384, 396 Maher, I. E., 537, 643 Mahoney, E. B., 64, 100 Maisin, J., 123, 168 Majoor, C. L. H., 507, 510, 644 Malmgren, R. A., 146, 168 Malmros, H., 528, 644 Malpress, F. H., 212(248, 249), 229 Mandelstam, J., 490, 498 Mann, I., 136, 166, 168 Mann, L. S., 526, 644 Mann, W., 318, 336 Marangos, G., 506, 514, 532, 640 Marcelet, J., 530, 645 March, N. H., 52, 66 Marcus, S.,534, 641 Marder, S. N., 207(203), 228 Mark, D. D., 373, 378, 393 Marks, H. H., 489, 498 Marle, E. R., 433, 447 Marshak, A., 311, 336 Martin, M., 38, 47, 64, 66 Martin, N. H., 507, 644 Martin, R. H., 382, 390 Marvin, H. N., 533, 647 Marx, W., 257, 267, 534, 644 Masayama, T., 372, 373, 39.9 Matthews, V. S., 221(357), 232 Matthews, W. B., 398, 448 Mattson, S., 434, 449 Maun, M. E., 216, 230,456, 468, 498 Maver, M. E., 257, 266, 482, 601, 506, 514, 530, 644 Mawson, C. A., 256, 267 Max, P. F., 65, 92, 100 Maxwell, L. C., 481, 497 Mayer, K., 518, 523, 536, 644, 647 Mayer, N., 377, 996 Mayer, R. L., 529, 644 Maynard, L. A., 458, 499 Mayneord, W. V., 380, 391 Mayot, M., 16, 66
AUTHOR INDEX
Meek, E. C., 517, 525, 641 Mehl, J. W., 508, 510, 521, 523, 528, 534, 5361 6&16&7 6471 648 Meinken, M. A., 438, 449 Meisel, D., 141, 171, 196, 197(102), 226, 329, SS7 Meister, A., 378, 393, 515, 642 Meitus, A. C., 533, 647 Melckionna, R. H., 220(336), 231 Mellanby, E., 236, 269, 468, 470, 501 Mellors, R. C., 378, S9S Meloche, V. W., 516, 518, 645 Mendelsohn, W., 250, 268 Mershimer, W. L., 519, 525, 646 Merten, R., 524, 535, 642 Meyer, F., 505, 518, 519, 647 Meyer, K., 536, 64.4 Meyer, L. M., 487, 499 Meyer, R. K., 189(48, 49, 50, 52), 224 Meyer-Heck, P., 517, 518, 644 Michaelis, L., 257, 268 Mider, G. B., 64, 90, 100, 108, 123, 124, 168, 171, 206(194), 227, 457, 474, 601,504, 508, 509,514,524, 525,644 Milford, J. J., 237, 238, 269 Millard, A., 136, 16S, 169 Miller, C . S., 323, SS7 Miller, E. C., 67, 100, 288, 290, 336, 343, 344,346,347,348,349,350, 351,353, 354, 355,356,357, 358,359,360,362, 363,364,365,366,367,369, 370, 371, 372, 373, 376, 377, 379, 380, 385,387, 388, 390, 391,393, S94, S96, 477, 478, 498,4.99, Miller, E. E., 510, 520, 524, 525, 644, 647 Miller, E. W., 106, 114, 120, 168, 210, 220, 221 (349, 350, 351), 329, 231 Miller, G. L., 510, 520, 525, 644, 647 Miller, G. M., 505, 516, 517, 519, 64.3 Miller, H., 74, 100 Miller, J. A., 290, 304, 3S6, 341, 343, 344, 345, 346, 347, 348, 349, 350, 351, 353, 354,355, 356,357,358, 359, 360,362, 363, 364, 365, 366, 367, 369, 370, 371, 372,373,376,377, 379,380,385,387, 388, 389, 390, 391, S92, S93, 394, 396, 452, 468, 471, 474, 477, @9, 600 Miller, J. K., 260, 869 Miller, 0. J., 188, 190(55), 224
563
Miller, W. L., Jr., 63, 100, 343, 350, S94, S96 Miller, Z. B., 418, 447 Millington, R. A., 295, 337 Mills, C. A., 108, 164, 168, 171 Milne, J., 507, 510, 646 Miner, D. L., 346, 347, 348, 394, 474, 477, 478, 499 Mitchell, E. B., 525, 642 Mitchell, J. H., Jr., 320, 331, 332, S34, 536, 336, 3S7 Mittelman, D., 510, 511, 512, 521, 640 Mixer, H., 191(61), 205(188), 206(188), 224, 227 Miyaji, T., 366, 996 Moen, J. K., 244, 269 Moeschlin, S., 525, 646 Mo5tt, W. E., 13, 16, 25, 66 Moloney, J. B., 255, 257, 266 Monod, J., 247, 269 Moon, H. D., 194(76, 77), 198(105), 203 (166), 206(105, 198, 199), 225, 227, 228
Moore, A. M., 409, 448 Moore, D. H., 136, 138, 140, 166, 507, 510, 528, 534, 642, 646 Moore, G. E., 302, 336 Moore, L. V., 538, 659 Moore, R. A., 220(336), 231 Moore, S., 402, 405, 407, 408, 448, 4.49 Moosey, M. M., 142, 144, 145, 166 Moreschi, C., 481, 499 Morgan, W. C., 135, 160 Mori, K., 348, 379, 394 Morrione, 378, 394 Morris, H. P., 108, 162, 163, 291, 336, 3S7, 452, 476, 477, 482, 486, 4.99 Morris, R., 507, 64.4 Morrow, A. G., 221(355), 232 Morse, A. H., 208(221), 228 Morton, D. E., 207(204), 228 Morton, J. J., 64, 90, 100, 123, 168, 206 (194), 227, 504, 508, 509, 514, 644 Moser, H., 445, 446 Moskop, M., 210, 213(270, 273, 274), 229, 2.90 Mottram, J. C., 64, 67, 99, 100, 243, 269, 464, 499 Moulder, P. V., 219(333), 231 Moulton, F. R., 174(8), 211(8), 22.3
564
AUTHOR INDEX
Mouton, R. F., 510, 511, 512, 521,640 Moyer, A. W., 252,269 Mucke, K.,524,642 Mtihlbock, O.,108, 120, 121, 133, 135, 150, 167, 168, 181(21), 182(23, 24), 218(21), 223 Mueller, G. C., 347, 358, 362, 363, 365, 366,370, 387, 393,394,477, 499 Mueller, J. H., 258, 269 Mtiller, 0.H., 518, 523, 646 Muether, R. O.,517, 525, 641 Mulinos, M. G.,465,499 Mulligan, R. M.,211(246), 929 Mulliken, R.S., 17,21, 22, 24, 27, 33, 66 Munro, L. A., 528,646 Murphy, J. B., 204(180), 205, 206(186), 207(201, 202), 927, 228, 235, 236, 237,243,250,260,261,967,269,270 Murray, J. A., 104, 117, 160,168 Murray, W.S., 105, 106, 109, 110, 111, 112,120,122,143,148,162, 167,168. 169,212, 229 Muxart, R.,293,336 Mylroie, A.,445,44.9
Noble, R. L., 216, 227 Noda, L.,376, 377, 391 Norberg, E.,302,336, 515, 646 Northrup, J. M.,398, 409,448 Norris, T.A., 141, 171 Novak, F. V., 522, 640 Novak, J., 535, 536,646 Novikoff, A. B.,378,394 Nye, W.,290, 336 Nye, W.N., 376,377,391 0
Oberling, C., 201(137, 138), 226, 245,251, 255, 261, 262, 263, 969 Ogston, A. G.,399, 401, 402, 403, 404, 405,408,415, 439, 449 Ok46, A.,406, 449 Olcott, H. A., 441,449 Olcott, H.S., 367,391 Oleson, J. J., 240, 868, 269, 486,@9 Olson, R.E.,294, 296, 336, 378, 394 Omachi, A.,249, 969 Oncley, J. L., 510, 511,512,513, 517,519, 528,640,646 N Opie, E.L.,344, 345, 348, 349, 376, 304, 468, 478,600 Nagao, N., 352, 356, 357, 380, 394 Orgel, L. E.,23, 66 Nakahara, W.,255, 271, 372, 391, 399, Orr, J. W.,70, 100, 123, 124, 125, 126, 394,515, 646 162, 163, 169, 218(325, 327), 221 Nassau, E.,535, 536,644 (349),231, 343,344, 345, 346, 394 Nathanson, I. T., 215(301), 230 Ostergard, R. P.,349,392 Nebbia, G.,8, 16, 27, 33, 34,66 Oswald, W.,526, 646 Needham, D.M., 399,408, 441,442,447, Ott, M. L.,534,641 Ottesen, J., 312, 313,336 4& Nelson, W. O., 189(56), 202, 209(224, Overholser, M. D.,207(205), 928 226), 215(152), 216, 2B4, 226,228 Nero, W. J., 135,162 P Neukom, H.,409, 449 Neurath, H.,253, 269,520, 646 Pack, G . T., 174(9), 211(9), 219(9),223, Newton, B. L., 207(211, 212), 208(211, 490, 600,514,515,639 212), 211(211), 213(264), 228, 229, Paesi, F. J. A., 189(54), 924 533,640 Page, O.,506, 513, 514,643 Newton, M. A., 323,332, 337 Palade, G. E.,531, 646 Niasi, S.,535, 538, 646 Paletta, F. X., 65, 71, 72, 73, 92, 95,96, Nickerson, M., 507, 513, 523,643 99, 100 Nielson, P. E.,66,95,99 Pan, S. C., 208(214), 210, 211(232, 234), Nibson, I., 535, 646 928,929 Ninni, M., 398 Pardee, A. B.,294, 296, 336 Nitsche, G . A., 513, 646 Parfentjev, I. A., 525, 646
565
AUTHOR INDEX
Parker, F., 259, 268 Parker, R. C., 244, 269 Parkes, H. S., 190, 224 Parkinson, M., 139, 164 Passey, R. D., 135, 136, 138, 139, 150, 157, 163, 169, 478, 600 Pading, L., 6, 7, 8, 10, 13, 15, 23, 29, 66 Payne, A. H., 316, 336 Payne, L. D., 347, 348, 391 Peacock, P. R., 246, 269, 355, 356, 381, 382, 392 Pearce, L., 520, 646 Pearsall, H. R., 508, 528, 529, 641, 646 Pearson, B., 378, 394 Pearsons, J., 120, 167 Peckham, B. M., 185, 186(37), 224 Pedersen, K. O., 513, 646 Penn, H. S., 505, 527, 642, 646 Penney, G. W., 20, 25, 66 Pennington, D., 141, 171 Pentimalli, F., 235, 236, 269 Perkins, M. S., 283, 336 Perlmutter, M., 534, 646 Perloff, W. H., 209(226), 228 Perlzweig, W. A., 474, 600, 527, 528, 646 Perry, I. H., 123, 269, 210, 229 Petermann, M. L., 507, 508,510,523,646 Peters, R. A., 409, 442, 449 Peyron, A., 240, 269 Pfaff, M. L., 505, 535, 536, 537, 646 Pfeiffer, C. A., 178(12, 13), 180(19), 188, 194(74, 79, 83), 195, 197(102a, 83), 204(87), 207(208), 208(222, 223, 12), 221(348, 354), 223, 224, 828 Pfeiffer, P. H., 643 Philips, F. S., 398, 408, 436, 445, 446, 447, 449 Pickels, E. G., 245, 255, 266 Pierson, H., 209(227), 228 Pikovski, M., 239, 262, 269 Pillemer, L., 511, 646 Pirani, c. L., 536, 646 Pirie, A., 250, 258, 259, 269, 409, 449 Pirie, N. W., 253, 269 Plant, G. W. E., 364, 394 Player, M. A., 505, 516,618,519,520,643 Plescia, A. M., 290, 336,364, 393 Podloucky, F. H., 530, 640 Poling, E. C., 348, 3991
Pollack, A. D., 191, 192(65), 224 Pollack, M. A., 377, 394, 395 Pollak, 0. J., 517, 646 Pollard, A., 252, 255, 256, 257, 258, 259, 260, 869 Pomerantz, L., 465, 499 Ponder, E., 517, 646 Popjak, G., 507, 646 Pople, J. A., 52, 65, 66 Porter, K. R., 140, 169, 245, 255, 266 Potter, V. R., 78, 100, 157, 169, 249, 270, 294, 296, 299, 302, 325, 333, 336, 336, 377, 378, 387, 390, 394 Poumeau-Delille, G., 240, 269 Powell, E. O., 411, 412, 413, 448 Prato, M., 8, 33, 34, 66 Preer, J. T., 157, 169, 170 Prelog, V., 416, 4 4 Pressman, D., 275, 336 Preston, J. M., 437, 449 Price, C. C., 400, 449 Price, D. E., 344, 394 Price, J. M., 345, 346, 370, 371, 372, 373, 376, 377, 390, 392, 394, 396 Prickett, C. O., 263, 266 Prhzmetal, M., 523, 646 Pullinger, B. D., 70, 71, 89, 100, 121, 128, 169 Pullinger, B. O., 219(328), 231 Pullman, A., 3, 6, 7, 13, 14, 15, 16, 18, 28, 30, 31, 32, 33, 34, 35, 36, 37, 41, 42, 43, 44, 47, 64, 66, 66, 279, 336 Pullman, B., 7, 16, 33, 42, 43, 44, 66, 279, 336 Purdy, W. J., 147,166, 256,258,262, 266, 267, 268 Purr, A., 517, 646 Pybus, F. C., 106,114,120,168,210,220, 221(349, 350, 351), 229, 231
Q Quadbeck, G., 357, 393 Quimby, E. H., 275, 336 Quinlin, P. M., 438, 443
R Rachele, J. R., 409, 449 Racker, E., 250, 268, 269
566
AUTHOR INDEX
Ragnotti, R., 236, 269 Ranadive, K. J., 111, 128,166, 214(285), 830 Randall, H. T.,106, 112, 114, 133, 136, 138, 140, 142, 166 Rapaport, I. A., 399, 445,&9 Rapaport, S. I., 633,647 Rask-Nielsen, R., 123, 164,206(196),228 Rathery, F.,535,640 Ray, F. E.,293,336 Raynaud, A., 216(307,308), 830 Reding, R., 522, 646 Reed, R., 136, 138, 139, 167,169 Reid, E.E.,406,448 Reid, J. C.,274,276,279, 301, 328, 334, S36 Reiner, L.,482,498 Rciner, M., 507,646 Reinhard, M.C.,110,171 Rekers, P. E.,514, 6S9 Reller, H.C.,68,69,70, 95,99, 100 Revell, S.,398,399,448 Rhees, M. C., 526,647 Rhoads, C.P.,288, 336, 342, 346, 347, 352,353,355,358,359,360, 362,363, 377,384,386,387,398,396, 399,436, 447,474, 477,478,499,514,639 Richardson, H. L., 346,346,351, 394 Richert, D.A., 378, 396 Richmond, V.,505, 536,536, 537, 646 Richter, M. N., 139,164 Riddle, O.,180(18),223 Riegel, C.,537,646 Rieke, C.A., 17,27,66 Rieke, H.S., 280, 291,336 Riley, V. T., 141,162,240, 241,242, 251, 259, 266, 269,270 Rimington, C.,523, 535, 536, 646,646 Ria, H., 373, 393 Ritchey, M. G.,76, 100, 101 Rittenberg, D.,278,301, 338,337 Rivers, T.M., 255,868 Rizzone, G. P.,643 Roberts, E., 76,77, 78, 81, 82, 96, 100 Roberta, J. J., 443,444,&8 Robertson, W.van B., 487, 600 Robinson, A. M., 518, 64.6 Robinson, D.,534, 646 Robinson, W.D.,534,641
Robson, J. M., 195, 197(86), 226 Roche, J., 530, 646 Roe, E. M. F.,65,100, 353, 354,392 Roll, P. M.,319,S34,966 Rolnick, H.A., 139,166 Rose, F. L.,436,442,443,445, 449 Rosenbohm, A.,644 Rosenfeld, L.,524, 646 Rosenthal, O., 297, SS6 Rosenthal, T.B.,74,75, 100 Roskelley, R. C., 377, 396 Ross, M.H.,526, 64.8 Ross, S.D.,433,447 Ross, W.C.J., 399, 400, 405, 409, 413, 414,419,421,422,425,426,427,428, 429,430,431,432,433,436,437,442, 444,445,446,447,448, 449 Rossiter, L.J., 490,498 Rothe-Meyer, A.,263,270, 971 Rothen, A.,242, 246, 252,257, 259, 266 Rothman, S.,66,76, 100, 101 Rothwell, J. T.,505, 517, 525, 641 Roue, P.,91,100,156,169,235,237,248, 260,268,870,464,481,498,499,600 Ruangsiri, C.,72 Rudali, G.,38,47,64 Rudall, K.M., 65,100 Rumsfeld, H.W.,Jr., 63,100, 343, 396 Rundles, R. W.,528,646 Rupp, J. J., 204(177, 1781, 205, 827 Rusch, H.P.,108,162,280,304, 311, 334, 336,341,343,346,347,348,349,350, 351,362,376,377,379, 390,393,394, 3g6,452,457,459,466,468,469,470, 471,473,474,477,478,479,482,483, 497,498, 499,600,516, 518, 646 Rush, B. F.,188(44), 224 Russ, S.,259, 270 Russel, B. R. G., 117,169 Russel, M.,517, 646 Rutenberg, A. M.,330,397 Ryder, A., 533,644 Rydon, H.N.,411,412,413, .bbs Rygaard, J., 206(192), 227 S
Sack, T., 326,337 Sakami, W.,364, 396 Salmon, W.D.,388, 391,478,498
AUTHOR INDEX
Salter, W. T., 377, 696 Salzberg, D. A., 290, 866, 360, 368, 369, S96 Sampath, A., 240, 368 Samuels, L. T., 131, 169, 218(319), 2Sl Sanders, E., 526, 646 Sandin, R. B., 354, 388, ,9995 Sandorfy, C., 16, 47, 66 Sannie, C., 201(138), 226 Saphir, O., 526, 659 Sapp, R. W., 344,347,355,357,358,360, 366, 367, 369, 370, 379, 380, 385, 387, s9s Sarett, H. P., 474, 600 Saaaki, T., 341, 379, 596 Sauberlich, H. E., 377, S96 Savignac, R. J., 505, 519, 646 Saxton, J. A., Jr., 201(134), 202(153, 1541, 207(134), 226, 456, 457, 482, 498, 600 Sayers, G., 466, 600 Scatchard, G., 513, 518, 646, 646 Schaible, P. J., 526, 646 Schenken, J. R., 220, 221(347), 2S1, $32 Schenken, L. R., 121, 122, 168 Schiff, A., 95, 99 Schiller, W., 382, 596 Schmid, K., 510, 511, 512, 521, 523, 536, 640, 6/14, 546 Schmidt, G. W., 235, 236, 267 Schmidt, H. W., 523, 646 Schmidt, I. G., 208(218), 228 Schmidt, M. B., 341, 396 Schmidt, O., 4, 5, 6, 49, 51, 66 Schmitt, L. H., 533, 647 Schneider, W. C., 78, 100, 249, 270, 296, 336, 367, 373, S96 Schoenbach, E. B., 505, 508, 516, 517, 520, 523, 528, 642,646, 647 Schoenewaldt, E. F., 329, 3S6, 357 Schoenheimer, R., 276, 336 Schoental, R., 280, 334, 410,447,491,498 Schultz, E. L., 511, 521, 644 Schuster, M. C., 139, 164 Schwarta, S., 214(289), 2SO Schwartz, T. B., 507, 513, 643 Schweigert, B. S., 377, 396 Scott, G. M., 259, 270 Scudder, J., 521, 536, 646 Seegers, W. H., 533, 643
567
Segaloff, A., 201(147, 148), 202, 216, 220 (335), 221(147), 226, 2Sl Seibert, F. B., 504, 505, 508, 521, 535, 536, 537, 646 Seibert, M. V., 504, 505, 508, 521, 535, 536, 537, 646 Selbie, F. R., 252, 255, 258, 261, 969, 370 Seligman, A. M., 326, 327, 330, 337 Sells, M. T., 257, 266 Selye, H., 201(144), 202, 203(144, 163), 2261 227 Sevringhaus, E. L., 185, 201(131), 22S, 226
Seymour, R. B., 433,448 Shacter, B., 534, 646 Shapiro, D. M., 487, 600 Shapiro, J. R., 200(126), 226 Sharp, D. G., 251,253,254,255,256,257, 258, 263, 266, 260, 370,271, 520,646 Sharpless, G. R., 252, 269 Shaw-MacKenzie, J. A., 505, 526, 646 Shay, H., 284, 3S7 Shear, M. J., 326, SSY, 341, 379, 380, 396, 480,600 Shedlovsky, T., 508, 521, 528, 536, 644, 646 Shemin, D., 250, 254, 270, 301, SS7 Sherman, J., 13, 16, 66 Shetlar, C. L., 505, 535, 536, 537, 648 Shetlar, M. R., 505, 535, 536, 537, 646 Shimkin, M. B., 105, 106, 107, 108, 110, 119, 120, 121, 122, 123, 128, 132, 134, 135, 136, 142, 148, 150,160,162, 169, 170, 182(22), 195, 204(87, 183),205, W S , 226, 227, 284, 3S4, 534, 646 Shinowara, G. Y., 524, 646 Shope, R. E., 247, 270 Shrigley, E. W., 236, 239, 267, 270 Shubik, P., 93, 100, 464, 497 Sibley, J. A., 505, 530, 646 Siedentopf, H. A., 106, 132, 136, 171 Siegel, I., 364, 396 Siekevita, P., 307, 337, 364, 396 Sievers, O., 524, 646 Silberberg, M., 122, 170, 183(26, 27), 194 (75), 204(181, 182), 205, 212, 215 (299), 223, 226, 227, 230 Silberberg, R., 122, 170, 183(26, 27), 194 (75), 204(181, 182), 205, 212, 215 (299), 223, 226,227, 2SO
568
AUTHOR INDEX
Silverstone, H., 108, 170, 171, 360, 396, 457,459,462,463,464,465,468,469, 471,472,473,474,480,482,483,484, 600 Silvertsen, I., 108, 171 Simkin, B.,523, 646 Simpson, L.,323, 332,337 Simpson, M. E., 185, 194(76, 77), 198 (105), 203(165, 166), 206(105, 198, 199), 223,626,227, 228 Simpson, W.L., 67,73, 74,96,99,100 Sittenfield, M.J., 235,250, 970 Sivertsen, I., 457,601 Sizer, I. W.,505,519,646 Skarzynski, B.,530,641 Skegga, H.R.,323,337 Skipper, H.E.,320, 331, 332, 334, 336, 336,337 Sklar, A. L., 14,66 Skorodumov, V. A.,436,447 Slaughter, D.P.,490,498 Sloviter, H.A.,327, 337 Slye, M.,104, 170 Smadel, J. E.,255, 268 Small, G.,433,447 Smiljanic, A. N.,66,76, 100, 101 Smith, B. W.,505,533,647 Smith, C.A. H., 458,499 Smith, E.L.,507,513,521,536,643,646 Smith, F., 110, 122, 129, 130, 162,166 Smith, F.W.,120,128,129,130,170,185 (33), 188(33), 198(104, 115), 199, 283,226,286 Smith, G. M., 123, 127, 164, 171, 201 (136), 210, 214(280), 218(326), 220 (340),926,229, 231, 358, 380, 396 Smith, K. A., 410,419,445,447 Smith, L.,434, 449 Smith, W.,66,95,99, 156, 170 Smith, W.B.,524, 646 Smith, W.E., 248,270 Smyth, I. M., 505,510,521,522,523,536, 648 Snell, G. D.,104, 170 Snider, H.,141, 166 Sobel, H.,186, 193(68, 70), 194(69), 8.34 Sobotka, H.,506, 514,646 Sonne, J. C.,318,334 Sonneborn, T.M.,157, 170 Sorof, S.,371,396
Spaey, J., 522,642 Spangler, J. M.,347, 396, 478,498 Speakman, J. B., 435,44.9 Speer, F.D., 525,640 Sperling, G.,458, 499 Spiegel, A., 198(106), B26 Spiegelman, S., 157, 170 Spirtes, M.A.,299, 337 Spitz, S.,384, 396 Sprinson, D.B.,319, 323, 336 Sproul, E.E., 250, 254, 256,268, 270 Stacey, M.,536, 646 Stadie, W.C.,519, 534, 646 Stahmann, M.A.,398,414,418,437,448, 449 Stanley, W.M.,136, 138, 140, 166,253, 270 Stare, F. J., 294,336 State, D.,535, 538,646 Staube, A. M., 536,646 Steele, J. M.,297, 307, 338,378, 396 Stein, W.H.,398, 400,401, 402,405,407, 408, 415,416,417,418,437,448,449 Steiner, P. E.,31,66 Stephan, V., 416, 449 Stephenson, M. L., 297, 305, 307, 338, 378,396 Stern, K., 135, 162, 452, 476, 480, 486, 600, 506,514, 526, 531, 532, 646 Stern, K.G., 250,252,255, 263,270,507,
646
Stettner, M.M., 519, 525, 646 Stevens, C. D.,405, 409, 411, 438, 441, 445, 447,449 Stevens, S., 256,268,270 Stevenson, E.S.,359,360, 363,396 Stickland, L. H.,136, 140, 163 Stier, A. R.,351, 394 Stock, C.C.,399, 436,445,446,447, 449 Stoerck, H.C.,486,600 Stoeaz, P.A., 300,337 Storey, W.F.,68, 100 Stout, A. P.,196, 197(102), 926 Stowell, E.C., 526, 647 Stowell, R.E.,63, 06,72,90,95,99, 100, 101, 376,396 Strait, L.A., 137, 161 Straw, A. A., 526,639 Strong, A.,123, 167 Strong, F. M.,377, 891
569
AUTHOR INDEX
Strong, L. C., 91, 101,104,115,118,123, Taylor, A. R., 251, 253, 254, 255, 256, 257, 258,269,270, 271 125,126,127, 135,148,164,167,170, 171, 196(98), 201(136), 204(175), Taylor, D.R.,141, 171 205, 210, 211(233), 213(265), 214 Taylor, H.C.,188(45), 224 (279,280), 218(321, 326), 220(340), Taylor, H.C.,Jr., 127,171 226,226,227,228,$29,231, 358,380, Taylor, H.F. W., 398,410,411,447 Taylor, H.L.,507, 646 396 Taylor, J., 377,394 Strong, L. E., 510, 517, 519, 528, 640 Taylor, J. C.,214(277), 230 Stulberg, C. R., 252, 266 Sturm, E.,204(180), 205, 207(201, 202) Tempereau, C.E., 533,647 227,228,235,250,259, 261, 269,270 Tenenbaum, E.,244, 267 Teresi, J. D.,302,307, 336,378, 391 SubbaRow, Y.,240, 268,486, 499 Sugiura, K.,237,241,250,258,270,319, Tesluk, H.,514, 6.44 323,334,341,342,343,346,347,350,Teutschlaender, O.,255, 271 352,353,355,357,358, 362,364,377, Thacker, J., 141, 171,471,498 378,382,387,392,393, 396, 436,443, Thiersch, J. B.,436,445,446, Thoma, G.E., 517,525, 641 474,477,478,481, 498,499,600 Sullivan, B. H.,517,6.43 Thomas, F.,121, 167,214(283), 230 Suntzeff, V., 62,63,66, 74,75,76,77,81, Thomas, M.A., 257, 266 83, 84, 92, 93, 94, 95, 96, 99, 1 , Thompson, C. R.,196(99, loo), 226 101,121,122,168,210,213(270, 273, Thompson, F.I,.,524,640 Thompson, H.C.,Jr., 70,99 274), 229, 230 Surgenor, D.M., 510, 511, 512, 521,640 Thompson, H.P.,140, 169 Sutton, L. E.,23, 66 Thompson, J. W.,482, 483, 601, 530, Svartholm, N. V.,6, 9, 16,66 644 Sveusson, H.,507, 510, 535, 536, 640, Thompson, R. C., 141, 166,377,394 Thornton, H.,526,647 643 Swain, C. G., 401,433,447 Timmis, G.M.,65, 100,399, 437,4.49 Sweeney, L.,526,643 Tinozsi, E. P.,524,647 Sweet, B.,326,337 Tiselius, A., 507,510,611,525, 535, 536, Swift, E.F., 471,498 640,647 Swift, R. W., 471, 498 Tishkoff, G. H.,82, 96,100 Sylven, B.,534, 642 Toennies, G.,504, 647 Symeonidis, A.,216(310), 230 Tolbert, B. M.,274, 276, 279, 328, 334, 337 Syverton, J. T., 145, 146, 166, 248,270 Tomarelli, R., 349,392 T Toosy, M.H.,66,97, 101 Totter, J. R.,319, 337 Tagnon, H.J., 531,646 Tracy, M.M.,312, 319,334 Taki, I., 366,396 Traub, E.,152, 171 Tannenbaum, A., 108,170, 171,452,457, Trebing, J., 532, 6.40 458,459,460,461,462,463, 464,465, Treffers, H.P., 513, 647 468,469,470,471,472,473,480,481,Trentin, J. J., 108, 121, 171, 181(20), 182(20), 183(25), 214(288a), 218(20), 482,483,484,488,600 Tapley, D. F., 516,643 223, 230 Tamer, H.,369, 396 Tripi, H.B., 260, 268 Tatum, E.L.,76, 100, 101 Troiser, J., 271 Tauber, H.,533, 647 Tropp, C., 518,647 Taylor, A,, 117, 141, 166, 171, 377, 394, Tung, T. C.,378, 396 396, 479,600 Turba, F.,511, 647
a?
570
AUTHOR INDEX
Turner, C. W., 108, 171, 212(247), 214, (288a), 2.99, 230 Turner, F. C., 326, 337 Turner, R. A,, 290, 334, 364, 390 Tuttle, L. W., 310, 311, 337 Twigg, G. H., 430, 431, 448, 44.9. Twombly, G. H., 123, 141, 171, 174(9), 188(45), 196, 197(102), 211(9), 223, 224, 926, 329, 336,337 Tyner, E. P., 302, 320, 937 Tytler, W. H., 235, 270 Tyzzer, E. E., 115, 167
W
Wagner, J. C., 252, d71 Wakefield, L. D., 400, 449 Wakelin, R. W., 409, 4.49 Walaszek, E., 333, 334 Waldron, D. M., 535, 647 Waldschmidt-Leitz, E., 517, 518, 530,647 Waldvogel, M. J., 533, 647 Walker, T. T., 259, 268 Wallace, E. W., 108, 171 Wallace, H., 108, 171 Walple, A. L., 436, 442, 449 Waltma D, C. A., 127, 171 U Wdtman, I#&"., 214(277), 230 Waitor, A. . R., 524, 647 Uhl, E., 263, 271 Waihrg, O., 293, 337, 505, 530, 647 Ullyot, G. E., 413, 448 Ward, R., 490, 601 Ungar, G., 533, 647 Ware, A. G., 533, 642 Unna, K., 347,390 Warner, S. G., 110, 111, 143, 169, 171 Uphoff, D., 218(324), 231 Warner, W., 104, 160 Uroma, E., 510, 511, 512, 521, 640, 644 Warren, F. L., 358, 384, 391 Warren, S., 505, 517, 525, 641 V Waser, P., 378, 396 N., 452, 601, 531, 647 Waterman, Van Artsdalen, E. R., 411,412,447 Watson, A. F., 468, 470, 601 van der Scheer, J., 252, 266 Watson, C. J., 132, 135, 16.9 Van Dyke, J. H., 70, 99 Waymouth, C., 244, $71, 373, 391 Van Eck, G., 192(64), 294 Wayne, A., 530, 641 van Gulik, P. J., 106, 107, 112, 121, 127, Weaver, J. C., 328, 336,337 166, 167, 214(284), 230 Weber, G. M., 344, 360, 366, 367, 369, van Thoai, N., 530, 646 370,371,372,373,376,377,393,394 van Wagenen, G., 194(78, 80), 208(221), Webster, M. B., 537, 643 219(331), 226, 228,231 Weed, L. L., 325, 337 van Wagtendonk, W. J., 157, 170 Wegelin, C., 489, 601 van Winkle, Q., 139, 166 Weigert, F., 49, 64, 67, 99, 100 Vasquea-Lopez, E., 203(164), 221(164), Weimer, H. E., 510,621, 536, 646, 647 227 Weinhouse, S., 284, 295, 298, 299, 387 Vassel, B., 515, 641 Weinstein, L., 209(230), 928 Vimtrup, B., 478, 498 Weisbrod, F. G., 537, 647 Viollier, G., 377, 378, 388, 396 Weisburger, E. K.,291, 336, 337 Visscher, M. B., 106, 108, 110, 128, 129, Weisburger, J. H., 291, 336, 337 132,136,160,162,166,167,171, 214 Weiss, E., 526, 647 (287, 288), 230, 457, 458, 462, 466, Weiss, 8.M., 281, 284, 286, 288,336, 389, 497, 499, 601 392 Voegtlin, C.,482, 483, 499,601 Weissman, N., 505, 508, 516, 517, 520, Volkin, E., 319, 337 528, 646, 647 Voorhees, V., 259, 266 Weitkamp, A. W., 75, 76, 101 Vroelant, C., 16, 20, 47, 66, 66 Welker, W. H., 526,644 Vyeki, E., 333,334 Wells, B. B., 633,647
57 1
AUTHOR INDEX Wells, E. B., 248, 270 Weltman, O., 528, 647 Werner, C. E., 295, 299,337 Werner, H. W., 196(99, loo), 226 Werner, I., 537, 647 West, P. M., 341, 377, 396, 505, 517, 533, 634, 641, 642, 647 West, R., 505, 535, 647 Westerfeld, W. W., 378, 396 Westergaard, M., 445, 448 Westman, A., 196(96), 226 Westphal, V., 505, 518, 519, 647 Wicks, L. F., 63, 75, 76, 96, 100,101 Wiener, M., 248, 871 Wiest, W. G., 285,287, 288, 289,55, 389, 396 Wiggins, L. F., 431, 435, 449 Wilheim, R., 452, 476, 480, 486, 600,506, 514, 526, 531, 532, 646 Williams, A. H., 444, 448 Williams-Aahman, H. G., 299, 337 Williams, R. J., 377, 394, 396, 479, 600 Williams, W. L., 123, 124, 125, 126, 128, 131, 167, i r i , i98(108), 204(176), 205, 208(215), 218(321, 322), 226, 227, 228, 231 Willie, R . A., 126, 171 Wilson, D. W., 318, 325, 336, 337 Wilson, J. G., 178(14), 223 Wilson, R. H., 354, 396 Winder, W. R., 70, 100 Winfield, K., 252, 269 Winnick, T., 301, 302, 306, 336, 337 Winstein, S., 433, 448 Winzler, R. J., 476, 486, 498, 505, 510, 513, 517, 519,521, 522,523, 525, 532, 534, 536, 642, 644, 646, 647, 648 Witschi, E., 185(32), 223 Wharton, D. R. A,, 525, 647 Wharton, M. L., 525, 647 Wheland, G. W., 6, 13, 15, 23, 24, 28, 66 Whipple, G. H., 514, 64.6 White, C. C., 257, 266 White, F. R., 108, 171, 348, 353,355, 356, 362,387, 396,457, 462, 474, 475, 483, 601 White, J., 108, 124, 171, 343, 345, 346, 348, 353, 355, 356, 362, 379, 387, 3.90, 391, 396, 457, 462, 474, 475,601, 515, 642
White, L., Jr., 331, 332, 334, 337 White, M. R., 316, 336 Whitney, R., 534, 641 Woglom, W. H., 208(213), 228,341, 377, 396 Wolbach, S. B., 65, 101 Wolf, C. G. L., 532, 648 Wolf, G., 287, 337 Wolfe, J. M., 201(1.32, 139, 140, 141, 142), 202, 208(216), 213(140, 261, 262, 263), 226, 828, 229 Wolff, E., 505, 522, 648 Wolfson, W. Q., 507, 640 WOU, E., 240, 2ri Wollman, E., 247, 269 Wolstenholme, J. T . , 193(72), 194(72), ,924 Wood, D. J. C., 431, 44.9 Wood, J. L., 288, 336, 409, 411, 441, 447,
449
Wood, M. T., 257, 266 Woodard, H. Q.; 378, 396, 529, 640 Woodhouse, D. L., 506, 514, 532, 535, 647, 648 Woods, A., 377, 394 Woods, M. W., 157, 163, 171, 172 Woodward, F. N., 444, 448 Woolley, G. W., 106, 115, 120, 130, 131, 132, 164, 172, 198(111, 112, 113, 114, 116, 117), 199(118, 119, 122), 200 (1231, 201(130), 226, 226 Wormall, A., 398, 405, 407, 409, 410, 428, 437, 442, 4461 447 Woywood, E., 185(34), 186(34), 223 Wright, A. W., 201(132, 139, 140, 142), 202, 213(140, 262, 263), 227 Wright, L. D., 323, 337 Wright, W. M., 532, 648 Wyckoff, R. W. G., 258, 266, 271 Wyman, R. S., 120, 122, 134, 170, 204 ( l a ) , 205, 227
Y Yankwich, P. E., 274,276,279, 334 Yaoi, H., 255, 271 Yates, F., 461, 498 Yokoyamrt, T., 372, 373, 393 Yoshida, T., 341, 379, 380, 396, 396 Young, G., 346, 371, 378, 390 Young, L. E., 533, 6-40
572
AUTHOR INDEX
Young, N. F., 346, 377, 392, 474, 477, 499, 507, 510, 514, 643, 646 Young, W. C., 178(14), 2?23 Yvan, P., 47, 66
z
Zamecnik, P. C., 297, 301, 305, 307, 309, 387, 888,378, 396
Zerahn, K., 315, 334 Ziegler, D. M., 71, 101 Zilversmit, D. B., 279, 313, 838 Zirkle, C. L., 413, 448 Zondek, B., 185, 201(143), 202, 223, 226, H7 Zuckerman, S., 219(334), 131 Zweifach, B. W., 93, 99
Subject Index A AAT, see under o-Aminoazotoluene AB, see under 4Aminoazobenzene Abwehrferment, 532 Acetate, labeled, metabolism, 294-296 2-Acetylaminofluorene, carcinogenic activity, 291 of 3-methyldiaminoazobenzene and, 351 mammary cancer in mice and, 124 metabolism, 291 Adrenals, effect on lymphomagenic action of estrogens, 207 of X-rays, 207 intermitotic time in, 70 ovary and, 188, 190 tumors, experimental, 198-200 caloric intake and, 458 gonadotropins and, 199 hormonal effects of, 199 strain differences in, 198, 199 types of, 198 Adrenocorticotropic hormone, caloric restriction and formation of, 466 gonadal hormones and, 199 tumorigenic activity of, 200 Adrenocorticotropin, see under Adrenocorticotropic hormone Alanine, incorporation into proteins of normal and tumor tissue, 305, 306,308 enzymes and, 307 reaction with aliphatic nitrogen mustards, 415,416 Albumin, egg, reaction of epoxides with crystalline, 434-435 plasma, effect of cancer on, 504, 509, 514,524
Albumin A, of plasma, cancer and, 524 Aldolase, serum, neoplastic disease and, 530 Alginic acids, reactions of epoxides with, 435 4-Alkyl-l-phenylpiperazines, formation, 416 Alkylating agents, see also under names of individual compounds cytotoxic, biological activity, 439ff. structure and, 442, 443, 444, 445, 446 chemistry, 397-449 reaction with nucleophilic groups in biological systems, 439ff. radiomimetic, 444 Amines, reaction of mustard gas with, 406 Amino acids, see also under names of individual compounds, carcinogenic aminoazo dyes and, 348, 377 epidermal, 82 effect of methylcholanthrene on, 81 of squamous cell carcinoma on, 82 incorporation into proteins, 309 labeled, incorporation into tumor proteins, 301-309 in vitro, 304-309 in vivo, 301-304, 307 reaction of alkyl-2-chloroethylamines with, 416,417 of mustards with, 407-408 D-Amino acids, in tumors, 304 4Aminoazobenzene, structure, 342 Aminoaao dyes, carcinogenic, see also under names of special dyes, 339-390 activity, 353, 354,355 amino acids and, 348 chemical structure and, 352-359,370
673
574
SUBJECT INDEX
dietary effects on, 346-351 mechanism, 371, 383-390 speaiea differences, 376 vitamins, 346-348 effecton chemical composition of liver, 371-378 liver tumors induced by, chemical composition of, 371-378 protein synthesis in, 378 o-Aminoazotoluene, carcinogenic activity, of, and its derivatives, 341, 379-380 sex differences in, 379 structure, 342 2-Aminofluorene, carcinogenic activity, 293 Amino groups, nucleophilic character, 427, 434 Ammonia, preformed, in epidermis, 78 Androgens, effect on carcinogenic action of estrogens, 215-216, 217, 219 on lymphagenic action of x-rays, 206 Aniline, carcinogenic activity, 353 Anthracene, carcinogenic activity, 37 K-region of, 37 Anticarcinogenesis, 2 Apyrase, epidermal, activity, effect of skin cancer on, 77 Arginase, epidermal, 78 effect of carcinogens on, 79, 81 of skin cancer on, 77, 81 role in urea formation in mammals, 78 Ar yl-2-chloroalk ylamines, reaction with anions, 425-427 Ascorbic acid, deficiency, effect on guinea pig sarcoma, 487 Atoms, meso, 5, 6, 30 %Azaguamine, tumor-inhibition by, 323, 487 uptake of labeled by nucleic acids, 323 Azo dyes, carcinogenic, see also under Amimoazo dyes and under names of individual dyes activity, diet and, 491 labeled, 290-293
2,2’-Azonaphthalene, carcinogenic activity of, and derivntives, 380-381 possible mechanism, 385 Azulene, ionic structures, 16
B Bases, reaction of I ,2-epoxides with, 434 of mustards with, 406-408, 416-418, 427-429 Bence-Jones protein, in multiple myeloma, 528 1 ,a-Benzacridine, K-region of, and its derivatives, 37, 39 carcinogenic activity and, 37, 39, 41 3,PBenzacridine, K-region of, and of derivatives, 4, 37 carcinogenic activity and, 37, 41 1,2-Benzanthracene, carcinogenic activity of, and its derivatives, 37, 42 K-region and, 37, 42, 43, 44 structure and, 40 K-region of, 3, 4, 5, 6, 32, 37 meso atoms of, 6, 30 molecular diagram, 30, 35 Benzene, I bond orders in, 15 molecular diagram, 17, 26 electron patterns, IOff. 3,4Benephenanthrene, carcinogenic activity of, and its derivatives, 31 K-region and, 36, 37, 43 K-region of, 3, 32 Benzpyrene, labeled, carcinogenic activity, 284 rate of elimination, 284-286 metabolism, 288 3,4-Benspyrene, carcinogenic activity, 67 caloric intake and, 463 labeled, 280 metabolism, 281, 282 Biological systems, nucleophilic groups in, 439-440 reaction of cytotoxic alkylatingagents with, 439ff.
SUBJECT INDEX
Biotin, azo dye-induced liver tumors and, 478 Bladder, tumors, fl-naphthalimine and, 175 Blood, adsorption of b u s No. 1 sarcoma agent on elements of, 236 Body weight, tumorigenesis and, 464 Boneb), neoplastic response to sex hormones, 22 1 tumors, alkaline serum phosphatase and, 530 Bradosol (fl-phenoxyethyl-dimethyldodecyl ammonium bromide), 529 Butadiene, molecular diagram, 26, 34 1,1Butanediol, dimethanesulfonyl ester, 437 Butter yellow, liver tumors produced by, 175 Butyl 2-chloroethylsulfide, reaction with tobacco mosaic virus, 441 C
Calcium, in normal and precancerous hyperplastic epidermis, 93 Cancer, see also under Tumors breast, see under Cancer, mammary effect on plasma globulins, 520, 524, 525-528 on serum protein-bound carbohydrate, 535-538 epidermal, 58 gastric cancer and, 58-61 gastric, epidermal cancer and, 58-61 hypoalbuminemia in, 514 mammary, in laboratory animals, 211, 212 hormonal imbalance and, 215 size of glands and, 215 in man, milk agent and, 158 in mice, 2-acetylaminofluorene and, 124 age and, 154 classification, 104 differentiation, 108 effect of pituitary on, 120
575
environmental factors and, 108 genetic factors and, 106, 108, 109ff., 119 hereditary factors and, 104, 105, 129 histology of, 125, 154 hormonal factors and, 106, 106, 108, 119, 129, 149 methylcholanthrene induced, 123 milk agent and, 103-172 effect of genetic factors on transmission of, 113 estrogens and, 121ff., 126 spontaneous, 118 in man, body weight and incidence of, 488489 dietary deficiencies and, 489-491 nutritional state and, 487-497 nitrogen mustards and chemotherapy of, 397 nutrition and, 451-497 plasma proteins and, 503-548 polarographic filtrate test for, 522-523 serodiagnosis of, 504, 514, 516-529, 530, 532 methods, 516-520, 525, 526-527, 530, 531, 532 skin, activity of epidermal enzymes and, 77, 78, 81 epidermal choline content and, 76 inositol content and, 76 squamous cell, experimental, 69, 90 chemical composition, 86, 87 effect on epidermal amino acids, 82 nitrogen metabolism, 80, 81 reducible substance in, 83 virus origin, 240 Carbohydrates, protein-bound of serum, distribution, 535 effect of cancer on, 535-538 nature of, 535-536 origin, 536-537 requirement for mitosis, 466 Carbonium ion, 412, 425, 426 affinity of iodide ion for, 427 formation, 411 reactions, 411, 415, 420, 422
576
SUBJECT INDEX
Carcinogenesis, see also under Tumorigenesis chemical, 280 mechanisms of, 384 effect on chemical composition of tissues, 87 electronic configuration and, 1-54 epidermal, 57-101 choline and, 86 hair follicle cycle and, 64-65 heredity and, 66 human, 66, 93-97 as compared with that in mice, 94-97
sunlight and, 94 latent period in, 89-90 in mice, 62-69 application of carcinogen, 63-64, 66 estrogen and, 65 factors influencing, 62-63, 64-65 nature of, 91, 97 stages in, 66-69, 91-93 mitotic activity of tissue and, 466 nutrition and, 495 radioisotopes in the study of, 273334
stages in, 66-09, 91-93, 464 vitamins and, 476-480 Carcinogens, see also under Hydrocarbons, carcinogenic and under names of individual compounds activity, 67 diet and, 461-462, 469-470, 485 factors influencing, 92 mechanism, 317 molecular size and, 2 structure and, 351 effect on chemical composition of epidermis, 72, 74-78 on cytoplasmic ribonucleic acid, 72 on dermis, 74 on epidermal alkaline phosphatase, 72
cells, 71, 72, 87-89, 90, 91, 92, 97 enzymes, 76-78 lipids, 76-76 minerals, 74-76 on hair follicles, 73 on sebaceous glands, 73
milk agent and, 123 responses to, 175 specificity, 175 transport through skin, 67, 68 Carcinolysis, 530-531 Carcinoma, see under Canwr Casein, tumorigenesis and dietary, 348, 472473
Catalase, cancer and activity of, in liver, 515 Catecholase, inhibition by serum, 534 Cell(s), divisibn, effect of ethyleneimine and its derivatives on, 436 epidermal, carcinogenic activity and, 92-93
number of, 70-71 spinous, effect of carcinogens on, 90, 91, 92, 97
tissue fluid environment of, 92 malignant, 244 conversion of normal into, 247 latent, 91 Rous No. 1 sarcoma agent and, 243-250
metabolism, effect of b u s No. 1 sarcoma agent on, 234, 246-250 virus-infected, metabolites of, 24 white, dendritic of human skin, pigment formation and, 66 Chick embryo agent, stability, 258 Chick embryo components, 253 chemical composition, 254 Chloroalkylamines, effect of structure on hydrolysis of, 424 2-Chloroethylamine (s), 411-429 biological activity of, and related compounds, 445 reaction with hexamethylenetetramine compounds, 418 in water, 411 N-2-Chloroethyl-N-3-chloropropylaniline, biological activity, 444 2-Chloroethyl sulfides, mutagenic action of, and related compounds, 445
SUBJECT INDNX
reaction mechanism in aqueous media, 4376. Choline, epidermal carcinogenesis and, 86 liver tumors and, 388, 478-479 Chromosomes, effect of mustard compounds on vege tal, 445 Chrysene, carcinogenic activity, 37 K-region of, 32, 37 Chymotrypsin inhibitor, of serum, neoplastic disease and, 533 Cocarcinogenesis, quantum-mechanical base of, 2 Colchicine, effect on incorporation of Pa* into DNA of tumors, 314 inhibition of Shope virus by, 240 labeled, metabolism in normal and tumor-bearing animals, 335 Configuration, carcinogenesis and electronic, 1-54 Cysteine, reaction of di-2-chloroethyl sulfide with, 405-406 Cystine, reaction with carcinogenic hydrocarbons, 288 Cytochrome oxidase, activity of epidermal, in cancer of skin and, 77 in precancerous hyperplastic epidermis, 77 Cytochromes, in tumors, 294 Cytoplaam, effect of carcinogenic hydrocarbons on, 289
D DAB, see under 4Dimethylaminoazobenzene DNA, see under Desoxyribose nucleic acid DN, see under Diphosphopyridine nucleotide Dehydrogenases, in transplantable tumors, 299 Dermis, effect of carcinogens on, 74
577
Desoxypentosenucleic acid, in liver following ingestion of carcinogenic aminoazo dyes, 372,373,374, 375-376 in viruses, 255 Desoxypyridoxine, 8-azaguanine and, 487 Desoxyribose nucleic acid, see also under Thymonucleic acid, 309 biosynthesis, 320 incorporation of Pa9 into, 312 colchicine and, 312,314,315, 316 radiation and, 31G315 tumors and, 312, 314, 315, 316 reaction of carcinogenic hydrocarbons with, 289 thymonucleic acid and, 312 1,2-5,6-Dianhydro-3, &acetone-mannitol, effect on wool fiber, 435 Dibenzanthracene, labeled, carcinogenic activity, 284, 288 metabolites, 283-284 rate of elimination, 284-286 lymphomagenic action, 206 1,2-Dibenzanthracene, K-region of, 286 metabolites, 286-288 1,2,5,6-Dibenzanthracene,33 K-region of, 3, 32 labeled, 280 metabolism, 281, 282, 283 1,2,7,8-Dibeneanthracenel33 K-region of, 32 1,2,3,4-Dibenaphenanthrenel carcinogenicity, 31 Di-2-chloroethylarylamines1 hydrolysis, 420ff. structure and rate of, 422, 423 tumor growth inhibitors, 421 NN-Di-2-chloroethyl-p-anisidine, cross linking in wool produced by, 429 Di-2-chloroethyl sulfide, 397, 399 reaction with amines, 406-407 with amino acids, 407 with anions, 402-406 with cysteine, 405-406 with proteins, 408 in water, 400, 401, 402 1,2,3,4Diepoxy-2-methylbutanebiological activity, 444
578
SUBJECT INDEX
Diet, liver cancer in humans and, 351 Diethyl-&iodoethylamine, labeled, metabolism in tumor-bearing mice, 330 Diethylstilbestrol, labeled, metabolism, 329 Di-2-halogenoalkylamines1 hydrolysis, structure and, 424 Di-2-hydroxyethylarylamines1 derivatives, tumor growth inhibition by, 437 4-Dimethylaminoarobeneene, administration, histological changes following, 343 pancreas and spleen tumors following, 344-346 carcinogenic activity, 341, 342-351 of derivatives, 343, 357-358 sex and species differences in, 343 of metabolites, 358-369 riboflavin and, 477, 478 thyroid and, 350-351 effect on chemical composition of normal and tumorous liver, 371378 on liver choline oxidase, 388 fluoro derivatives, carcinogenic activity, 385-386 labeled, metabolism, 290 metabolism by rats, 369-371 N-demethylation, 363-366 formation of protein-bound dyes, 366-371 hydroxylation of aniline ring, 363 pathway of, 361 reduction of azo linkage, 360, 362363 structure, 342 2,eDimet hoxy-6-et hyleneimino-l,3,6triarine, biological activity, 444 tumor growth inhibition by, 444 2',3-Dimethy1-~aminoaeobenrene1 see under o-Aminoarotoluene 9,I0-Dimethy1anthracenel3 3,l0-Dimethyl-6,6-benzacridinel 293 Dimethylbenzanthracene, auxinocarchogens, 46, 47 carcinogenophore, 46, 47
9,10-Dimethy1-1,2-dibenranthracene136 carcinogenic activity, 37 K-region of, 37 Dimethyl-2-chloroethylamine, reaction in water, 414 N-(2,4)-Dinitrophenyl-ethyleneimine tumor growth-inhibiting activity, 445 Diphenylamine reaction, 538 Diphosphopyridine nucleotide, effect on oxidative tumor metabolism, 299 Divinyl sulfone, radiomimetic effects, 437 Drugs, action, quantum-mechanical base of, 2 Dyes, see also under AZOdyes, Aminoazo dyes and under names of individual compounds protein-bound, 388, 389 formation of, 366-371 tumors induced by, 340ff. classification, 344-346 metastases, 345 origin, 345
E Electrons, methods of representing, in conjugated compounds, see under Valence bond method and under Molecular-orbital method II Electrons, 6-7, 8, 9, 20,21ff., 52ff. action of, 6 aromatic character of molecules and, 9-13 K-region and, 6, 9 nature of, 20 in pyridine, 19 Svartholm's model of, 6, 7 u Electrons, 8, 9, 52 Enryme inhibitors, in plasma, cancer and, 532-534 specificity of, 633 Enzymes, see also under names of individual enzymes effect of carcinogens on epidermal, 76-78 of mustards on sulfhydryl groups of, 441
SUBJECT INDEX
579
of tumors on, 515 with proteins, 434-435 of virus infection on cellular, 249-250 in water, 429-432 inactivation of Rous No. 1 sarcoma kinetics of, 430, 431 agent by, 258 mechanism of, 430-431, 438 in liver, effect of carcinogenic aminoazo Estradiol benzoate, effect on carcinodyes on, 377-378 genic activity, 92 oxidative, incorporation of labeled Estrogens, see also under names of indialanine into protein and, 307 vidual compounds inhibition by serum of cancer pacervical cancer in mice and, 176 tients, 534 effect on lymphomagenic action of location of, 298, 307 X-rays, 206 in tumors, 299, 306 on thymus, 204 plasma, neoplastic disease and, 529-532 on uterine cervix, 207 Epidermis, extrinsic, mammary tumorigenesis in calcium in normal and cancerous, 86, rodents and, 214ff., 218 87 inactivation by hepatic tissue, 185, 188 cancer, 58 incidence of liver tumors in mice and, gastric cancer and, 58-61 221 cells of, lymphoid tumors in rodents following growth, 70ff ., 58-59 treatment with, 204-205, 206 intermitotic time in, 70 adrenals and, 207 mitotic time in, 70 see differences in, 206 nuclei of, 71-72 strain differences in, 204 carcinogens and, 71, 72 mammary cancer in mice and, 121 number of, 70-71 milk agent and, 121, 123, 126 carcinogens and, 87-89 mammary response in mice to, 214ff., volume of, 71 218 chemical composition, 85-87, 97 pituitary tumors in rodents following effect of carcinogens on, 74-78 treatment with, 201, 202, 203 effect of carcinogens on nitrogen testicular response in mice to, 194, 195 metabolism, 78-82 strain differences in, 194, 196-197 experimental carcinogenesis in, 57-101 tumorigenic activity, 177, 207, 211, histochemistry of, 72-73 214ff. injuries to, 59, 60 androgens and, 215-216, 2 17, 219 carcinogenetic effect of, 59 progesterone and, 202, 204, 216-217 precancerous hyperplastic, 69 uterine response in rodents to, 208,209, biotin content, 76 210 calcium content, 93 vaginal response in mice to, 181-182 characteristics of, 89-90 strain differences in, 182, 183 cytochrome oxidase activity, 77 Estrone, mineral content of, 73 mammary response in mice to, 182 thymonucleic acid content, 72 strain differences in, 182-183 reducible substance in, 82, 83 Ethyl-di-2-chloroethylamine, replacement, 68 reactions with amino acids, 416, 417, urea content of, 78 418 1,2-Epoxides, with anions, 415, 416 cytotoxic action, 433, 436 with peptides, 416 reactions with anions, 432-434 Ethyleneimine, 436 with bases, 434 biological activity, 444, 445 with nucleic acid, 435-436 effect on cell division, 436
580
SUBJECT INDEX
on growth of animal tumors, 436 reaction in aqueous media, 438
F 4’-F-DAB, see under 4’-Fluoro-Pdimethylaminoazobenzene FSH, see under Follicle-stimulating hormone Fat, dietary, effect on tumorigenic activity of DAB and its derivatives, 469 growth of tumors and, 487 tumorigenesis and, 467-472 Fibrinogen, cancer and, 524-525 liver and synthesis of, 524 Fibroblasts, lymphoblasts and, 244 Roux No. 1 sarcoma agent and, 243, 244 transformation into macrophages, 244
4’-Fluoro-4dimethylaminoazobenzene, carcinogenic action, 343 effect on liver, 345 Folic acid, antagonists, effect on mouse leukemia, 48 Rous No. 1 sarcoma and, 240-241, 486 Follicle-stimulating hormone, ovarian tumors and, 189 testicular tumors and, 196, 197 Food dyes, commercial, carcinogenic activity, 380ff. Formate, labeled, incorporation into nucleic acid purines, 322, 323 Fowl leucoses, transmissible, 262-264 Rous No. 1 sarcoma and, 264 Fowl leucosis agent, isolation, 263 stability, 258 Fowl tumors, chemically-induced, 261-262 immunological properties, 262 Rous No. 1 sarcoma and, 261 Fuchs’ test, for cancer, 531-532 Fujinami tumor agent, enzymatic inactivation, 259
G
Gelatinases, bacterial, inhibitor of, in serum, 534 Glands, mammary, see under Mammary glands sebaceous, 73 effect of carcinogens on, 73 entry of carcinogens through, 68 a-2 Globulin, serum, effect of wasting diseases on, 536 a-Globulins, cancer and, 504, 508, 509,520-521 multiple myeloma and, 527-528 &Globulins, neoplastic disease and, 524, 527-528 y-Globulins, cancer and, 525-528 Glucose, labeled, metabolism, 295, 296, 297 D-Glutamic acid, labeled, metabolism in normal and tumorbearing animals, 304 Glycidol, biological activity, 445 Glycine, labeled, incorporation into proteins, 302-303,304,305,306 reaction of mustards with, 407 Glycol sulfonyl esters, reactions in aqueous media, 437 Glycolysis, in tumors, 294, 318 Gonadotropins, secretion of, age and, 188 ovary and, 185, 187, 188, 189 tumorigenic activity, 177, 194 Growth hormone, nucleic acids and, 326 tumorigenic action, 194, 203 Guanazolo, 323
H HN2, see under Methyldi-2-chloroethylamine HNs, see under Tri-2-chloroethylamine Hair follicles, cycle, 64-65 epidermal carcinogenesis and, 65 effect of carcinogens on, 73
58 1
SWJECT INDEX
Hexamethylenetetramine, reaction with chloroethylamines, 418 Hormonal imbalances, 178 cancer and, 178 contributing factors, 177 experimental, 178ff. lactation and, 212 reversibility of, 178, 179 types of, 178-180 ‘IHormonal influence, inherited,” mammary cancer in mice and, 129 Hormones, see also under names of individual hormones activity, quantum-mechanical base of, 2 caloric intake and production of, 465, 466 carcinogenic, 174, 176, 211, 222, 304 nature of, 176, 177 neoplastic response to, 175-176, 179ff. duration of stimulus and, 178 sex differences in, 183, 184 species differences in, 180 time factor and, 179 endogenous, hyperplasia induced by, 179 experimental tumorigenesia and, 173232 gonadal, adrenocorticotropin and, 199 gonadotropic, see under Gonadotropins growth, Bee under Growth hormone mammary, response to, 211, 212, 219 species differences in, 212 metabolites, carcinogenic activity, 176 sex, neoplastic response of bones to, 221 tumors of secondary male sex organs and, 219-220 Hyaluronidase, inhibitors of, in serum, 534 Hydrocarbons, carcinogenic, see also under names of individual compounds activity, 37 dietary cystine and, 288 fat and, 469 K-region and, 6, 30, 31-32, 33-40, 43-50, 52 of metabolites, 288 possible mechanisms, 50-5‘4
structure and, 3, 17, 36, 41, 46 cervical cancer in mice following application of, 211 effect on formation of protein-bound dyes in rat liver, 369, 370 on tumorigenic activity of 3‘methyldiaminoazobenzene, 351 interaction with tissue components, 288-289 K-region of, 3-8, 30-50, 279, 280 mammary cancer and, 123, 218 metabolism, 279-290 oxidation in vivo and in vitro, 49 synthesis of labeled, 280 Hyperconjugation, 17-18, 27-28 Hyperplasia, adrenal in mice, milk agent and, 130 endogenous hormones and, 179 Hypoalbuminemia, in cancer patients, 508, 514, 515
I Influenza virus, 251 composition, 248 host and, 263, 254 immunologicalproperties, 248,249,261 infective component, 253-254 metabolites, 248 Iodine deficiency, incidence of thyroid cancer and, 490 Irradiation, carcinogenic effects of, 275 by compounds localized in tumors, 327-328 inactivation of Rous No. 1 sarcoma agent by, 259 Isotopes, see also under names of individual compounds radioactive, methodology, 276-279 production and measurement, 276 in the study of carcinogenesia, 273334 of immunology of cancer, 276
K K-region, of carcinogenic hydrocarbons, bond orders in, 31-32 carcinogenic action and, 3-8,36,37,38, 39, 53, 279, 280
582
SUBJECT INDEX
charge at, 33-40 definition, 279 electrical properties, 6, 30-46 electronic distribution in and near, 8 free valences, 33 indexes, electrical, 30-31 molecular-orbital, 20-21 “Kappa ” factor, of paramecia, comparison with milk agent, 157 Ketosteroids, fecal excretion by mice with mammary tumor virus, 218 Kidneys, effect of sex on, of rodents, 221 Krebs cycle, conversion of carbohydrate to protein and, 307 localization of enzymes, 298 in tumors, 294-298, 307
L LH, see under Luteinieing hormone Lactation, hormonal imbalances and, 212 B-Lactoglobulin, reaction of epoxides with, 435 Lead-212 (thorium B), in the study of plant metabolism, 276 Leukemia, caloric intake and, 492 of mice, folic acid antagonists and, 487 milk agent and, 135 Lipids, effect of carcinogens on epidermal, 7576 Liver, cancer, see also under Liver, tumors, 490-491 diet and, in man, 351 etiologic factors, 490 liver cirrhosis and, 490 chemical composition, 85-87 effect of carcinogenic dyes on, 371378 effect of neoplastic disease on, 515 mitochondria in, 515 production of carbohydrate-rich plasma proteins by, 536 role in hypoalbuminemia of cancer patients, 515
tumors, see also under Liver, cancer butter yellow and, 175 caloric intake and experimental, 458, 492 choline and, 388, 478-479 dietary fat and, 468, 492 proteins and incidence of, 473-474, 475, 476, 477, 492 induced by aminoazo dyes, 340-343, 379-380 chemical composition of, 371-378 histology Of, 344-346 species differences in, 344 vitamin levels and, 477-478 in mice, hormones and, 221 sex differences in, 220-221 urethan and, 175 vitamins €3 and spontaneous, 480, 492 Luteinizing hormone, testicular tumors and, 196 Lymphoblasts, fibroblasts and, 244 Lymphoid tumors, experimental, 204-207 adrenals and, 207 following treatment with dibenzanthracene, 206 with estrogens, 204-205, 206 with methylcholanthrene, 206 sex differences in, 206 transplantability of, 204
M MAB, see under CMonomethylaminoasobeneene 3’-Me-DAB, see under 3’-Methyl-4dimethylaminoazobenzene Macrophages, fibroblasts and, 244 Malnutrition, tumorigenesis and, 491,496 Mammary glands, cancer, see also under Mammary glands, tumors caloric restriction and, 485, 492 dietary fat and, 468, 469, 492 proteins and, 472, 475, 476 hormones and, 212, 219 in mice, effect of labeled Nile blue 2B on, 327 vitamins B and incidence of, 477, 479-480, 486, 492
SUBJECT INDEX
hormonal influence on, 212 milk agent and structure of, 127 response to estrogens, inanition and, 214 tumors, experimental, 211-219 contributing factors, 218 ovarian function and, 214 role of hormones in, 211, 212, 217218 strain differences in, 213, 214, 217, 218 Metabolic pool, 276 Methionine, labeled, incorporation into prdteins of tumor-bearing animals, 302 reaction with di-2-chloroethyl sulfide, 408 5-Methylacridine, carcinogenic activity, 37 K-region of, 37 Methylbis (2-chloroethylamine), labeled, metabolism in normal and leukemic mice, 33 Methylcholanthrene, carcinogenic action, 63, 67, 69, 97, 175 effect on epidermal amino acids, 81 cells, 92, 93 epidermal hyperplasia and, 70, 71 labeled, 280 carcinogenic activity, 284 excretion into mother’s milk, 284 rate of elimination, 284-285 leukemia in mice induced by, 474 dietary proteins and, 474, 475 lymphomagenic action, 206 milk agent and, in mammary cancer in mice, 123 20-Methylcholanthrene, effect on carcinogenic activity of aminoazo dyes, 351, 370 labeled, distribution of radioactivity in tumors induced by, 283 2-Methyldiaminoazobenzene, carcinogenic activity, 356 4’-Methyldiaminoazobenzene, carcinogenic activity, 356 Methyldi-2-chloroethylamine, 392 reaction with amino acids, 416,417,418 with anions, 415, 416 with peptides, 416, 417
583
with proteins, 418 in water, 412413, 414 3’-Methyl-4-dimethylaminoazobenzene, carcinogenic action, 343 2-acetylaminofluorene and, 351 carcinogenic hydrocarbons and, 351 20-methylcholanthrene and, 351, 370 metabolism of labeled, 290 by rats, 360 7-Methylpteroylglutamic acid, synergism with desoxypyridoxine, 487 Milk agent, 152 adrenal hyperplasia and, 130 antigenic properties, 144 behavior in vivo, 140 chemical properties, 135 comparison with “kappa” factor of paramecia, 157 disappearance of, 142 distribution in body of mice, 107, 132, 133 effect of genetic factors on susceptibility to, 110, 114, 119 on transmiesion of, 113 on leukemia in mice, 135 hormonal factors and, 119, 152 inherited susceptibility to mammary cancer and, 129 introduction into body of mice, 107 isolation, 136 mammary cancer in mice and, 103-172 carcinogen-induced, 123ff ., 127 estrogen induced, 121ff., 126 in hybrid mice and, 148 in man and, 158 transplanted, 115 sarcomatous transformation in, 117 multiplication, 143 nature of, 147, 154, 156 neutralization, 144 occurrence, 132, 149, 150 physical properties, 135 purification, 140 serological behavior, 144 structure of mammary gland and, 115 transmission in mice, 107, 113, 140, 150 Minerals, carcinogenic action, 480 effect of carcinogens on epidermal, 74-75
584
BUBJECT INDEX
Mitochondria, in normal and tumorous liver, 298 Molecular-orbital method, 20-30 hyperconjugation, 27-28 LCAO representation, 21-24 magnitudes derived from, 24-26 molecular orbitals, 20-21 polarizabilities, 26-27 resulta, 25-26 tests of theory 28-30 Molecules, aromatic, r electrons and bond structure of, 9-13 carcinogenic potency, prediction of, 7 resonance energy and, 43 stability and, 42 electronic distribution in, effect of methyl groups on, 17-18, 27-28 structure, 6 “molecular diagrams,” 16 &Monomethylaminoazobenzene, carcinogenic activity, 356, 354 structure, 342 Mucoproteins, carbohydrate-containing of serum, 536 in normal and pathological plasma, 522,523 Muscle, chemical cornposition of epidermis and, 85-87 Mustards, see also under names of individual compounds antileukemic action, 329 biological effects, 398,410 relation between chemical activity and, 398-399 carcinogenic activity, 329 cross linking activity of, 442443 effect on sulfhydryl groups of enzymes,
with proteins, 398, 418,429 in water, 399402,411-415,437-439 sulfur, 399-411 Myeloma, multiple, serum globulins in, 527-528 therapeutic effect of stilbamidine in, 328
N
Naphthacene, carcinogenic activity, 37 K-region of, 32,37 Naphthhlene, T bond orders in, 15 carcinogenic activity, 37 ionic structures, 16 K-region, 37 molecular diagram, 17,26 valence bond structures, 12 &Naphthalimine, bladder tumors produced by, 175 2-Naphthyldi-2’-chloroethylamine, chemotherapy of cancer and, 398 Neoplastic disease, see also under Cancer and Tumors immunological aspects of, 526 plasma enzymes and, 529-532 protein stability in, 528-529 Nile blue 2B,labeled, 327 effect on mouse tumors, 327 properties, 327 Nitrogen, nucleophilic character, 412 Nitrogen metabolism, effect of carcinogens on epidermal, 78-82 Nitrogen mustards, see under Mustards, nitrogen Nucleic acids, see also under names of 441 individual compounds labeled, metabolism, 330-331 biochemistry, 310 nitrogen, aliphatic, 411-419 biosynthesis, 323 aromatic, 427-429 composition, 309 chemotherapy of cancer and, 397,398 cytoplasmic, DAB and, 344 radiomimetic, 398 growth and, 326 reactions with alanine, 415-416 incorporation of purines and pyrimiwith anions, 402-406,415-416 dines into, 308, 319,320,321 with bases, 406-408, 416-418, 427X-rays and, 320 429 metabolism, 310,312 with nucleic acids, 410,419 precursors, 314
SUBJECT lNDPlX
radioisotopes in the study of, 309-326 reactions with electrophilic groups in uico, 440, 441 of l12-epoxideswith, 435-436 of mustards with, 410411, 419 turnover rates, 312-314 uptake of Palby, 3106. Nutrition, cancer and, 451497 0
Orotic acid, labeled, 324-325 incorporation into nucleic acid pyrimidine nucleotides, 324 metabolism, 325 Ovary, adrenals and, 188, 189 effect on gonadotropin secretion, 185, 187 pituitary and, 188, 189 tumors, experimental, 184-194 age and, 187 gonadotropins and, 187-188, 189, 190, 194 histogenesis, 192-193 species differences in, 193 hormonal etiology, 184-187, 192 hormone production by, 194 induced by intrasplenic ovarian grafts, 185ff. by X-ray irradiation, 190-193 species differences in, 194 transplantability, 193
P PNA, see under Ribosenucleic acid Pancreas, tumors, following administration of, DAB, 343 Pectic acids, reactions of epoxides with, 435 Pentaphene, K-region of, 33 D-Peptidases, serum, neoplastic disease and, 530 Peptides, reaction of alkyl-2-chloroethylaminea with, 416, 417 Pharyngeal cancer, incidence in artic region, 490
585
Phenanthrene, carcinogenicderivatives, structure of, 3 double bond, see under Phenanthrene, K-region K-region of, 3, 4, 5, 6, 37 molecular diagram, 34 tumorigenic activity, 37 Pheochromocytomas, experimental, 198 Phosphatase, acid, metastatic prostatlic cancer and serum, 529 alkaline, effect of carcinogens on, 72 hyperplastic bone disease and, 529530 Phosphorus, radioactive, uptake by nucleic acids, 312 by nucleoproteins of normal and tumor-hearing mice, 310-312, 318 carcinogenic azo dyes and, 317 glucose and, 318 Phosphorylation, oxidative, in tumors, 299, 300 Pituitary, effect on mammary cancer in mice, 120 hormones, experimental testiculrtr tumors and, 194, 196, 197 ovary and, 188, 189 sex difference in function of, 178 in size of, 201 tumors, experimental, 200-204 following estrogen treatment, 201, 202, 203 nature of, 200, 202 possible etiology, 203 strain and species differences in, 201, 202, 203 thyroid and, 202-203 Plasma, electrophoretic components, 509 cancer and, 504-505, 509 enzymes, neoplastic disease and, 529532, 532-534 proteins, cancer and, 503-548 stability in neoplastic disease, 528529 Plasmin, 531 neoplastic disease and, 531 Polarographic filtrate test, for cancer, 522-523
586
SUBJECT INDEX
Polypeptides, in normal and pathological plasma, 522 Progesterone, carcinogenic action of estrogens and, 216-217
.
effect on lymphomagenic action of estrogens, 204 of X-rays, 204 tumorigenia action of estrogens and, 202
Prostate, cancer, hormones and, 220 serum acid phosphataae and metastatic, 629 Proteins, dietary resistance to toxic agents and, 474
tumor growth and, 482485 tumorigenesia and, 472-476 effect on tumorigenic action of DAB, 348
incorporation of amino acids into, 301309
in liver, effect of carcinogenic aminoazo dyes on, 373, 374, 375, 376, 377, 378
plasma, cancer and, 503-548 determination of, 506-513 stability in neoplastic disease, 528529
reaction with electrophilic agents in vivo, 440, 441 of 1,2-epoxides with, 434-435 with mustards, 398, 408-410, 418, 429
synthesis, circulation and rate of, 307
Roua No. 1 sarcoma agent and, 247 Provirusea, 264 Pseudohypophysectomy, 465 Purines, incorporation of labeled formate into, 318, 323
into nucleic acids, 318, 319, 320 grrene, K-region of, 32 Pyridine, molecular diagram, 26 stmctures, 19
Pyridoxine, azo dye-induced liver tumors and, 478 deficiency, effect on mouse lymphosarcoma, 486 Pyrimidines, incorporation into nucleic acids, 318 of nucleic acids, 323-326 Pyruvate, labeled, metabolism in tumors, 300
Q Quinoline, molecular diagram, 26
R RNA, see under Ribosenucleic acid Reducible substance, in liver, 85 in muscle, 85 in normal and cancerous epidermis, 82, 83, 84
ultraviolet characteristics, 83, 84 Rennin inhibitor, in serum, neoplastic disease and, 533 Resonance method, see under Valence bond method, Riboflavin, azo dye liver tumors and, 477478,492 deficiency, effect on mouse lymphosarcoma, 486 inhibitory effect on tumorigenic action of aminoazo dyes, 346-347, 369 mechanism of, 362, 370 mammary cancer in mice and, 477 Ribonucleic acid, cytoplasmic, effect of carcinogens on, 72
Ribosenucleic acids, 309 Roue No. 1sarcoma agent, adaptation, 237 adsorption on blood elements, 236 age and, 242 antigen of, 249 assay of, 241-242 cell metabolism and, 234, 245-250 distribution, 236-237 effect of host environment on antigenic structure, 239
587
BUBJECT INDEX
on type of lesions produced by, 238, 239 on protein-synthesizing systems, 247 folio acid and, 240-241 fowl leucoses agents and, 263-264 hemorrhagic lesions produced by, 237, 238 infectiveness of, 242, 246 inhibition in vivo, 240 isolation, 250-255 lipid content, 255, 256 localization of, 235, 240, 244 origin, 264-265 preparations, purity, 234,253 properties, 233-265 immunological, 260-261 physico-chemical, 250ff. relationship to host, 235-242 size of, 251, 255-256 stability, 257-260 purification and, 258 survival time, 237 variations, 238-240 vitamin Ble and, 240, 241 b u s sarcoma, cells of, 243-245 electron microscopy of, 245 chemically induced fowl tumora and, 261, enzymes in, 249 folk acid deficiency and, 486 histogenesis of, 243-245 S
Sarooma, caloric intake and, 492 dietary protein and incidence of experimental, 473,476 effect of iodinated polysaccharide of Serratia matcesnes on mouse, 326 Scarlet red, carcinogenic action, 341 structure, 342 Serodiagnosis, of cancer, 504, 514, 516529, 530, 532 methods, 516-520, 525, 526-527, 530, 531,532 Serratia marcesnes, iodinated polysaccharide of, 326-327
effect on mouse sarcomas, 326 properties, 326 Serum, protein-bound carbohydrate of, cancer and, 535-538 sulfhydryl content, 505, 519 neoplastic disease and, 534 Sex organs, secondary tumors of male, hormones and, 219-220 Skin, effect of carcinogenic hydrocarbons on metabolism of, 289 transport. of carcinogens through, 67, 68 tumors, experimental, caloric intake and, 492 dietary fat and 468, 469,471, 492 protein and, 476 vitamins B and, 479,492 Spleen tumors, following administration of DAB, 343 Stilbamidine, labeled, 328 nucleic acids and, 328 Stomach cancer, epidermal cancer and, 58-61 mortality, 58 vitamin A deficiency and, 478 Succinate, labeled, metabolism, 296-297 Sulfur, mustards, see under Mustards, sulfur radioactive, uptake by nucleoproteins, 315,316 Sunlight, carcinogenic action, 94 Switzerland, incidence of thyroid cancer in, 489,490
T Testis, atrophy in mice, following estrogen treatment, 194 strain differences in, 194, 196-197 tumors, experimental, 194-197 androgenic effects, 197 histology of, 197 hormonal etiology, 194, 195, 196 induced by estrogen treatment, 195
588
SWJE1CT INDEX
metastases, 197 pituitary gonadotropins and, 194, 196, 197 transplantability, 197 Texas, incidence of skin cancer in, 59 Thiamine, effect on mammary carcinoma in mice, 486 Thiazans, synthesis of aryl-substituted, 427 Thymonucleic acid, desoxyribosenucleic acid and, 312 in precancerous hyperplastic epidermis, 72 reaction of di-2-chloroethyl sulfide with, 410 of epoxides with, 435, 436 Thymus, effect of estrogens on, 204 experimental lymphoid tumors of, 204 Thyroid, cancer, 489-490 incidence in Switzerland, 489, 490 iodine deficiency and, 490 effect on tumorigenic action of 3'-MeDAB, 350-351 experimental pituitary tumors and, 202-203 Thyrotropic hormone, carcinogenicity, 177 Tissue, cancerous, acidity of, 438 connective, Rous virus and tumors of, 175 hepatic, inactivation of estrogens by, 185, 188 interaction of carcinogenic hydrocarbons with components of, 288-289 mitotic activity, caloric restriction and, 466 cancer and, 466 Tobacco mosaic virus, reaction with butyl2-chloroethyl sulfide, 411 Toluene, charge distribution in, 28 molecular structures, 17, 18 Z(p-Toluenesulfonyl) aminofluorene, labeled, 293
Tracers, radioactive, see under Isotopes, and under names of individual elements Tri-Zchloroethylamine, 392 reactions with amino acids, 416, 417, 418 with anions, 415, 416 with peptides, 416 2,4,6-Triethyleneimino-l,3,5-triazine, 436 cross linking action, 437 Triphenylbromoethylene, labeled, metabolism, 328-329 Triphenylene, carcinogenic activity, 37 K-region of, 37 Triphenylethylene, tumorigenic activity of, and its derivatives, 195 Tritium, 328 Trypan blue, tumorigenic activity, 381 Trypsin inhibitor, in plasma, 532-533 neoplastic disease and, 533 Tryptophan-acid reaction, 537-538 Tumorigenesis, see also under Carcinogenesis, 455480 caloric intake and, 492 effect of dietary fats on, 467472 minerals on, 480 proteins on, 472476 experimental, 174ff. hormonal aspects of, 173-232 ovarian, 184-194 inhibitory effect of caloric restriction, 456-462, 492 factors influencing, 458-465 mechanism, 462-466 malnutrition and, 491 mustard gases and, 410 Tumors, see also under Cancer adrenal, experimental, 198-200 types of, 198 D-amino acids in, 304 effect on enzyme systems, 515 on Pal uptake by DNA, 312, 314, 315, 316, 317 glycolysis in, 318 growth of, 454, 484, 481487 caloric intake and, 481-482 dietary fat and, 482 protein and, 482-485
589
BUBJECT INDEX
effect of ethyleneimine and its derivatives on, 436 inhibition by di-2-chloroethylanilines, 421 mechanism, 303, 304 stages, 481 vitamins and, 485-487 induced by dyes, 340ff. classification, 344, 345 metastases, 345 origin, 345 by vitamin deficiency, 478 inhibition by &azaguanine, 487 Krebs cycle in, 294298, 307 liver, induced by azo dyes, 340, 341, 342, 343 lymphoid, experimental, 204-207 mammary, experimental, 211-219 metabolism, oxidative, 293-300 radioisotopes in the study of, 273334 ovarian, experimental, 184-194 pituitary, experimental, 200-204 proteins of, incorporation of labeled amino acids into, 301-309 in &TO, 304309 i n uiuo, 301-304 testicular, experimental, 194-197 hormonal etiology, 195ff. transplantable, dehydrogenases in, 299 uterine, experimental, 207-21 1 Tyrosinase, inhibition by serum of cancer patients, 534 Tyrosine, labeled, incorporation into proteins of tumorbearing animals, 301-302
U Urea, formation in mammals, arginase and, 78 Urethan, activity, antileukemic, 331 carcinogenic, 333 mechanism of, 332 labeled, metabolism in tumor-bearing mice, 331-332 lung tumors produced by, 175
Uric acid, biosynthesis, 318 Uterus, tumors, experimental, following estrogen treatment, 208209 types of, 208, 209
v Valence bond method, 8-20 aza replacement, 18-19 complicating features, 13-15 derived magnitudes, 15-17 methyl substitution, 17-18 Valence bond orders, 15, 26, 29 Penney’s, 18-20 in K-region, 31-32 Valence bonds, structures of, 9-13 numbers of, 13 Virus(es), see also under ROUENo. 1 sarcoma agent animal, immunological properties, 261 avian tumor, effect on cell metabolism, 235 effect on cellular enzymes, 250 equine encephalomyelitis (Eastern strain) chemical composition, 254, 255 Rous No. 1 sarcoma virus and, 25 stability, 257, 258 fowl-tumor, antigenic relationship between, 262 host and chemical composition, 253 mammary tumor, 176, 182, 215, 218, 219 rabbit papilloma (Shope), 247-248,251 chemical composition, 254 immunological properties, 261 inhibition by colchicine, 240 purity of preparations, 253 ROUE,production of connective tissue tumors by, 175 Vitamin deficiency, tumors induced by, 478 Vitamins, effect on carcinogenic activity, 76,346348, 373, 374, 375, 377, 476-480 on tumor growth, 485-487
590
SUBJBCT INDEX
Vitamins B, see also under name8 of individual vitamins tumorigenesis and, 479, 492
W
X X-rays, effect on Pa* uptake by DNA, 315 lymphomagenic action, adrenals and, 207 progesterone and, 207 mutagenic activity, 97 ovarian tumors in mice produced by, 190-193
Water, reaction of l,%epoxides in, 429-432 of mustards in, 39-15, 419425 Wool, z effect of 1,2-5,6dianhydro-3,4-acetonemannitol on, 435 Zymohexase, see under Aldolase