CURRENTS IN BIOCHEMICAL RESEARCH Mi 3SB^^^^B3E3E3SQBBSE3E E3 IB I I I I 3 E3 a m I Marine Biological laboratory Library n H Woods Hole, Mass. ID D ' ra B ID i IB B 01 H Presented by uJ B 01 B Estate of Dr. ^tto LoeTfi C B August, 1962 CURRENTS IN BIOCHEMICAL RESEARCH J CURRENTS IN BIOCHEMICAL RESEARCH Ed^iitd^ h-j DAVID E. GREEN Thirty-one essays charting the present course of Biochemical Research and considering the intimate relationship of biochemistry to medicine, agriculture and social problems DR. OTTO LOEWF 133 ' -^ c-.-r /.- INTERSCIENCE PUBLISHERS, INC., NEW YORK 1946 Copyright, 1946, by Int'erscience Publishers, Inc. 215 Fourth Avenue, New York 3, N. Y. Printed in the United States of America by the Mack Printing Company, Easton, Pa. PREFAG E With the ever-increasing degree of specialization in scientific research and with the terrifying rate of growth of technical nomen- clature, men of science are literally compelled to know more and more about less and less. The scientific literature furthers this trend, since journals, textbooks (apart from those for students), and review articles are written primarily for the specialist. There is an acute need for stripping complex subjects and getting at the simple, essential concepts which are basic to their appreciation. After all, the same scientific principles are applicable to all fields of inquiry. The art of presentation consists in the elimination of the barriers of terminology which effectively conceal these fundamental principles. Currents in Biochemical Research represents an attempt by some thirty research workers to describe in as simple language as possible the important developments in their own fields and to speculate a little on the most likely paths of future progress. The aim of these essays has been to excite the imagination and to provide glimpses of some of the fas- cinating horizons of biochemical research. However, no populari- zations were intended. The various contributors were asked to write simply and provocatively but without sacrifice of scholarship. Dealing as they do on the one hand with pharmacology, chemotherapy, public health, genetics, photosynthesis, and agriculture and on the other with considerations of organic, analytical, and physical chemistry, they emphasize the focal position of biochemical research in biology, chemistry, and medicine. It is hoped that this survey from so many different points of view may assist biochemists, chemists, and medical doctors in seeing biochemistry in clearer perspective and in its proper relation to other fields of inquiry. David E. Green March, 1946 CONTENTS 1. The Gene and Biochemistry page by G. W. Beadle 1 '\^ 2. Viruses by W. M. Stanley 13 3. Photosynthesis and the Production of Organic Matter on Earth by H. Gaffron 25 4. The Bacterial Cell by Rene J. Dubos 49 ^ 5. The Nutrition and Biochemistry of Plants by D. R. Hoagland 61 6. Biological Significance of Vitamins by C. A. Elvehjem 79 7. Some Aspects of Vitamin Research by Karl Folkers 89 8. Quantitative Analysis in Biochemistry by Donald D. Van Slyke 109 9. Enzymic Hydrolysis and Synthesis of Peptide Bonds by Joseph S. Frufon 123 10. Metabolic Process Patterns by Fritz Lipmann 1 37 1 1 . Biochemistry from the Standpoint of Enzymes by David E. Green 149 12. Enzymic Mechanisms of Carbon Dioxide Assimilation by Severo Ochoa 165 13. Hormones hy B. A. Houssay 1 87 ^^ VII O ^ 1 1^ CONTENTS 14. Fundamentals of Oxidation and Reduction page by Leonor Michaelis 207 15. Mesomeric Concepts in the Biological Sciences by Herman M. Kalckar 229 16. Viscometry in Biochemical Investigations by Max A. Lauffer 241 17. Isotope Technique in the Study of Intermediary Metabolism by D. Rittenberg and David Shemin 261 18. Mucolytic Enzymes by Karl Meyer 277 19. Some Aspects of Intermediary Metabolism by Konrad Block 291 20. The Steroid Hormones by Gregory Pincus 305 21. Plant Hormones and the Analysis of Growth by Kenneth V. Thimann 321 22. Chemical Mechanism of Nervous Action by David Nachmansohn 335 23. Some Aspects of Biochemical Antagonism by D. W. Woolley 357 24. Chemotherapy: Applied Cytochemistry by Rollin D. Hotchkiss 379 25. Biochemical Aspects of Pharmacology by Arnold D. Welch and Ernest Bueding 399 26. Some Biochemical Problems Posed by a Disease of Muscle by Charles L. Hoagland 41 3 [y 27. Physiology and Biochemistry by Surgeon Captain C. H. Best 427 28. X-Ray Diffraction and the Study of Fibrous Proteins by /. Fankuchen and H. Mark 439 29. Immunochemistry by Michael Heidelberger 453 30. Social Aspects of Nutrition by W. H. Sebrell 461 31. Organization and Support of Science in the United States by L. C. Dunn 473 VIII THE GENE AND BIOCHEMISTRY G. W. BEADLE, professor of genetics, school of biological SCIENCES, STANFORD UNIVERSITY / "T IS both an accident of organic evolution and an indication of man's lack of foresight that the organisms studied in most detail by biochemists have not been those on which geneticists have concentrated. It is natural that man should have a prejudice in favor of himself, and it is therefore not remarkable that the urge of medicine on biochemistry has been in the direction of specialization on mammals, particularly on man himself. For obvious reasons, bacterial biochemistry has likewise been well nourished through medicine. Man has few inherent advantages for biochemical study while the bacteria abound in them. But both are most difficult for the geneticist — the one because of a long life cycle and social obstacles to controlled matings, the other because of the absence of a sexual cycle without which the geneticist cannot use his particular methods. The geneticist, on the other hand, has chosen to make the vinegar fly and Indian corn the classical organisms of his science. Both suffer disadvantages to the biochemist in not lending themselves readily to culture under precisely defined environmental conditions. Neither can be grown conveniently on a medium completely known from a chemical standpoint. In spite of this situation and additional impediments arising through divergence in outlook, such persons as Garrod, Onslow (n^e Wheldale), Troland, Goldschmidt, Wright, Haldane, and others have G. W. BEADLE urged that the two fields have much in common and that each stands to profit through contact with the other. Through the eff"orts of these individuals and others of like mind there are many instances known in which the relation of genetics to biochemistry is so clear that it can no longer be disregarded by intelligent investigators in either field. In fact, from this relation there tends to emerge a new interest, known as biochemical genetics, which promises to tell us what the genes do and how they do it, on the one hand, and to lead us to further knowledge in the ways of biosynthesis on the other. In both directions there obviously lie many opportunities. One of the earliest instances in which a Mendelian trait could be interpreted in terms of specific chemical reactions is that involving the human disease known as alcaptonuria. In individuals homo- zygous for the mutant gene responsible for this character, 2,5-di- hydroxyphenylacetic acid (homogentisic acid or alcapton) is excreted in the urine instead of being broken down to carbon dioxide and water, as it is in persons receiving the normal form of the alcaptonuric gene from one or both parents (15). Homogentisic acid is oxidized to a black pigment on exposure to air and it is this process that is re- sponsible for darkening of the urine, the most striking symptom of the disease. According to Gross (cited by Garrod), alcaptonurics lack a specific enzyme found in the blood of normal persons which catalyzes the degradation of homogentisic acid. Alcaptonuria therefore repre- sents the first recorded instance in which it could be said that a par- ticular chemical reaction is controlled by a known gene through the mediation of a specific enzyme. Within the past dozen years, additional examples have become known in which organisms unable to carry out specific reactions differ in a single gene from their chemically more successful relatives. In flower pigment synthesis, for example, the formation of carotenoids, anthocyanins, anthoxanthins, chalcones, and fiavocyanins is known to be genetically controlled in one plant or another (7,24). Specific oxidations of pelargonidin derivatives to cyanidin analogues and of cyanidin compounds to delphinidin counterparts are dependent on the activities of specific genes. The addition of sugars to anthocyani- dins through glycosidal linkages and the transformation of the an- thoxanthin quercetin-3-glucoside to the corresponding cyanidin-3- glucoside are likewise unable to proceed if specific genes are modified. THE GENE AND BIOCHEMISTRY Foiling (14) and Penrose (35) have shown that the genetically determined failure to oxidize phenylpyruvic acid in man is invariably associated with subnormal mentality. Here again there appears to be an intimate relation between a particular gene and a specific chemical reaction. Because of its obvious importance to an understanding of the mechanisms underlying mental processes, this case is of particular interest. It is of course related metabolically to alcaptonuria in so far as phenylalanine is concerned in both. Other abnormalities in phenylalanine-tyrosine metabolism are also known (15,19). Most remarkable progress has recently been made in under- standing the genetic and chemical mechanisms of sex determination and differentiation in the green alga, Chlamydomonas, by Moewus, Kuhn, and co-workers. From the carotenoid pigment, protocrocin, there is derived through cleavage the motility hormone, crocin, and a female-determining hormone known as gynotermone. This cleavage is known to be genetically controlled. In genetically male individuals gynotermone is hydrolyzed to a male-determining hormone known as androtermone. Under the direction of specific genes the cis and trans forms of the motility hormone, crocin, are converted into the corre- sponding cis and trans dimethyl esters of crocetin. In various specific mixtures, these serve as gamones, i. e., they render individuals of the specific genetic constitutions capable of conjugation. The relations between genes and chemical reactions disclosed by this work support the thesis that genes act in directing specific processes. The work on Chlamydomonas is so spectacular and its importance so great that inde- pendent confirmation is desirable (see 7,28,41). The splitting of specific di- and trisaccharides by yeasts is under genetic control, as shown by Winge and Laustsen (54) and Lindegren, Spiegelman, and Lindegren (25). It appears that the genes concerned determine whether or not specific enzymes are present in active form. Somewhat similar situations are known in the rabbit, where Sawin and Glick (37) have shown that the activity of the enzyme, atropine esterase, is dependent on the presence of the normal allele of a par- ticular gene, and in white clover, where an enzyme responsible for hydrolysis of specific cyanogenetic glucosides is known to show a similar dependence on a gene (2,9). In the bread mold, Neiirospora, Srb and Horowitz (44) have shown that there is present an ornithine cycle essentially similar to G. W. BEADLE that postulated by Krebs and Henseleit for the mammalian liver. In the bread mold it is known that mutation in any one of seven different genes will interi'upt the synthesis of ornithine or its conversion to arginine. So far as the data go, they are consistent with the assump- tion that each of the seven genes is normally concerned with a different chemical reaction in the system. It is an interesting point that it was possible to establish the presence of the ornithine cycle in Neurospora because of the existence of the mutant strains indicated. Tatum and Bonner (50) have shown that tryptophan is normally synthesized in Neurospora through the condensation of indole and serine. Evidently the indole is somehow derived from anthranilic acid, for there exist two mutant strains, one of which accumulates anthranilic acid when it is grown under suitable conditions, while the other is able to grow normally when supplied with anthranilic acid in place of indole or tryptophan (51). The gene by which the first strain differs from wild type is evidently concerned with the reac- tion by which anthranilic acid is converted into indole, whereas the mutant gene of the second strain appears to be concerned with failure of some reaction essential to the synthesis of anthranilic acid. It is obvious, in this case, that genetics has provided a tool of great useful- ness in investigating the biosynthesis of the important amino acid, tryptophan. Relations similar to those mentioned above are known for other biosyntheses in the bread mold and in other organisms. By following methods developed by Beadle and Tatum (8), it has been possible to obtain a series of mutant strains of Neurospora in each of which some particular reaction has been blocked. These are concerned with the synthesis of amino acids, vitamins, purines, pyrimidines, and other compounds of biological importance (6,21,48,49). We can be sure from such cases as those just cited that genes function in directing biochemical reactions. We know, further, that this direction may involve enzymes as intermediates between gene and reaction. All our information is consistent with the hypothesis that in all cases in which genes control specific reactions they do so indirectly through enzymes. In other words, genes direct enzyme specificities, and enzymes control reactions. This is not a new idea. Bateson (4), Moore (29), Troland (52), Goldschmidt (16), Muller (30), Alexander and Bridges (1), Haldane (18,19), Wright (55), and others have sug- THE GENE AND BIOCHEMISTRY gested it. We are only now beginning to do something definite about it from an experimental standpoint. Since the specificities of enzymes are referable to protein specificities, the hypothesis implies that genes direct protein specificities. In this case we might expect that the specificities of proteins other than those found in enzymes would show a direct relation to genes. This is indeed the situation as evidenced by the fact that in many organisms a general one-to-one relation be- tween genes and antigens has been shown (22,47). It is true that a few deviations from this correspondence are known, but they may well represent instances in which antigens have specificities made up of two components, each corresponding to one gene. If we knew the chemical nature of genes, we should be in a much better position than we are now to determine how they direct protein specificities. Direct chemical analyses of whole chromosomes show them to be largely nucleoprotein (27), which suggests that genes too are nucleoproteins. But since chromosomes probably contain much nongenic material, the deduction is not too satisfying. Ultra- violet radiation induces gene mutation; and its efficiency in this re- spect varies with wave length in the same way as does its absorption by nucleic acid (20,45), strongly indicating that the energy eff'ective in producing mutations in genes is absorbed by nucleic acid. The simplest assumption possible is that this is so because the nucleic acid is part of the gene. The similarity of genes and viru.ses constitutes a third line of evi- dence concerning the chemical nature of genes. Both have the property of self-duplication, which in both cases is dependent on the presence of a series of compounds such as those found in the living cell. Genes and viruses appear to be within the same size range (46). Both are capable of undergoing mutation to new forms which have altered biological activities but retain the power of self-duplication (46). Since viruses and genes have so many properties in common, it is probable that they are similar in chemical makeup. Following Stanley's isolation of crystalline tobacco mosaic virus, several other viruses have been prepared in pure form and all have been shown to be nucleoproteins (5,12,46). The circumstantial evidence that genes, too, are nucleoproteins, or at least contain nucleoproteins as essential parts, is therefore substantial. In duplicating themselves, genes have been assumed to act as G. W. BEADLE master molecules or models from which exact copies are made (11,18, 19,30,31,55). If this is so, their action may be visualized as one of directing the construction of specific protein types plus whatever other component parts genes may have. If the specificities of proteins generally are copied from genes, the observed relations between genes and enzymes and between genes and antigens should follow. For every specific protein there should exist a gene carrying this same protein. For every enzyme and antigen type there likewise should be a gene. Because of a general tendency of mutation pressure to eliminate genes that are of no advantage to the organisms, it might be expected that for every protein type there would be only one corre- sponding gene. The experimental evidence appears to support this general interpretation, although it must be recognized that, in dealing with genetic traits that can be described in terms of chemical reactions, there may be an unavoidable selection of those cases in which gene action is relatively simple. However proteins and other components of genes are synthesized — whether by an orthodox stepwise mechanism or by some as yet un- known mechanism by which many component parts are simultaneously directed into their proper places by the master molecule (11) — the pre- cursors of proteins, nucleic acids, and whatever other parts genes may have, must be synthesized. Their synthesis will involve many enzymes and a corresponding number of genes. Thus, before one gene can determine the specificity of a new protein molecule, many other genes must have acted. This amounts to saying that, in any multigenic organism, the genes constitute a highly organized system, just as the chemical reactions they direct are integrated in time and space in a manner characteristic of a particular species. Furthermore, while a particular gene will have only one primary action in determining specificity of an enzyme or an antigen, the final physiological conse- quences of a change in a single gene will be manifold. This can be appreciated when one considers the consequences of depriving an organism such as a rat of thiamin. The final consequence is of course death, but before death occurs a series of changes of increasing com- plexity take place. These can be brought about in the rat by remov- ing thiamin from the diet. In the bread mold, which normally syn- thesizes thiamin, the same end result can be effected as a result of an analogous series of changes initiated by replacing a normal gene 6 THE GENE AND BIOCHEMISTRY necessary for thiamin synthesis with a defective form of the same gene. In one particular case, the primary action of the gene is presumed to be in directing the specificity of the enzyme catalyzing the reaction by which thiazole and pyrimidine are combined (49). Viewed in this way, an understanding of gene action does not appear hopelessly difficult even though the final effects of a single gene change may involve alterations so complex as to defy complete description. Griine- berg (17) has pointed out that a similar type of interpretation in terms of one primary action of a given gene is tenable in the case of certain hereditary developmental defects in the mouse and rat that at first sight appear to involve several unrelated changes in the organism. That the gene has a functional as well as structural unity is therefore a hypothesis that has demonstrated its heuristic value. Until evidence with which it is inconsistent is presented, it will no doubt continue to play an important role in our concepts of what the gene is and how it acts. As Troland (52), Muller (30), Alexander and Bridges (1), Oparin (33), Plunkett (36), and others have pointed out, the similarity of viruses and genes suggests that the first living structures, i. e., those with the power of self-duplication, were probably somewhat similar to present- day viruses with the important difference that they were free-living. The evolution of systems of such units, each acquiring the property of directing the specificity of an enzyme or other protein, would be ex- pected to give rise to a series of forms of increasing complexity such as we see today in the larger and more complex viruses, the rickettsias, bacteria, and higher organisms. It is probable that the present viruses and rickettsias are not relics of these ancestral forms but are forms secondarily derived through specialization in connection with parasitism (10). The true ancestral types must have been capable of multiplying outside living cells in a kind of environment which, because of the presence of many organisms, is no longer likely to exist (33). In terms of genes directing chemical reactions through their control of enzyme specificities it is possible to imagine how, in principle, these simple forms evolved in the direction of the more highly specialized and complex forms of multicellular plants and animals (56), although it is of course not easy to visualize the way in which the process occurred in detail in particular instances. In the specialization of higher animals with respect to their nutri- G. W. BEADLE tion, it is possible to suggest a scheme of evolution that has some sup- port at least in analogy. It has become increasingly evident that, with respect to their need for and use of vitamins of the B group, purines, pyrimidines, choline, amino acids, and other compounds, all cellular organisms are fundamentally very similar (23,26,53). To consider a specific example, carboxylase is presumably present in all protoplasm and apparently always contains thiamin as thiamin pyrophosphate. Many organisms, e. g., most plants, are able to synthesize the thiamin they need, while others are dependent on an external supply of this essential compound. From an evolutionary standpoint, this difference is presumably determined by whether or not it is of advantage to a par- ticular organism to synthesize thiamin. Evidently for Neurospora there is selective advantage in being able to carry out this synthesis for we find in wild strains that all essential genes concerned with it are present in active form. In mammals, on the other hand, thiamin is presumably so frequently present in the diet that the genes originally concerned with its elaboration have been permitted by natural selection to become inactive so far as thiamin synthesis goes. It may well be that they have not disappeared entirely but have been modified so as to enable the mammal to carry out chemical reactions of which the bread mold is incapable. In a similar way, mammals have become specialized through loss of ability to synthesize other vitamins, the indispensable amino acids, and other compounds. With the develop- ment of parasitism it would be expected that still further loss in syn- thetic ability would be encountered. As Knight (23), Lwoff" (26), Schopfer (38), and others have pointed out, this is indeed the case. The work on Neurospora makes it most probable that the dropping out of specific chemical reactions no longer of selective advantage is the result of gene mutation. The limit of such parasitic specialization is probably represented in the molecular viruses that have lost all power of heterosynthesis and have retained only the one property essential for their continued existence in an environment in which all necessary compounds are available — the property of autosynthesis. One may quite properly raise the question as to the course of positive evolution in terms of chemical reactions — how are new syntheses developed in the course of organic evolution? Unfortunately, the experimental evidence bearing on this is meager, which is not surpris- ing, for obviously it should be much easier to destroy or inactivate a 8 THE GENE AND BIOCHEMISTRY complex self-duplicating unit than to modify it so as to give it a new and useful property without sacrificing its power of self-duplication. The first self-duplicating unit must have evolved from nonliving matter at some time, and more complex forms must have evolved from it — the alternative is some form of special creation. There would seem to be less difficulty in imagining a primitive "protogene" mutating to a true gene with a heterocatalytic property than its spontaneous origin in the first place, even if, as Oparin (33) supposes, it arose in a world containing preformed organic molecules of many kinds. Nor is there any apparent reason why such protogenes could not nmtate in many diflferent directions in order to give rise to many different organic catalysts. In present-day cellular organisms there exists a possible mechanism for acquiring totally new reactions. Occasionally, through accident, one or more genes become duplicated, i. e., a small segment of a chromosome occurs twice in every set. The duplicated genes will be unnecessary to the organism and will be expected to disappear through loss mutations, since such mutations are not disadvantageous. But such a duplicate gene may occasionally undergo mutation in such a way that it directs the formation of an entirely new enzyme. If this new enzyme should happen to catalyze a reaction that improved the organism in competition with its relatives, the new reaction would be retained. Such new reactions might add new compounds or they might bring about the reverse of the specialization process, which leads in the direction of parasitism. In this way, as Horowitz (20a) has pointed out, the first primitive organisms might gradually have built up systems of synthesis which freed them of their dependence on preformed organic molecules originally present in the environment. Through such advances as hav-e been indicated we appear to be moving rapidly in the direction of a better understanding of what genes are and what they do. We are no longer content with a knowl- edge of the laws by which they are transmitted from one generation to the next. We see that they are basic functional units of the organism and that, by taking advantage of their tendency to mutate, we can use them as powerful tools in determining the course of biosynthesis and in imderstanding other aspects of metabolism. Their relations to enzymes and antigens are becoming known. Precisely how they function in duplicating themselves and in directing the specificities of proteins, nucleic acids, and possibly other large molecules is a question G. W. BEADLE for the future. But there can be no doubt that the years that lie ahead will be exciting ones in this field. The work of Avery and his co-workers (3) on the transformation of types in Pneumococcus and that of Emerson (13) and of others suggests that we may one day learn to direct gene mutations in predetermined ways. Work on enzymes (32) and viruses (5,46) is so closely related to the general problem of gene structure and gene action that only short steps appear to be necessary to bridge the gaps that separate them. Nucleic acid certainly plays an important role in gene action and in protein synthesis (34,39,40), and it is not too much to hope that this role will be made clear in the near future. The relation of genes to cytoplasmic elements is not well understood, but after many years in which discouragingly little prog- ress has been made, important leads are being followed by Sonneborn (42), Spiegelman (43), and others. After half a century of growth, genetics seems to be assuming a position in the broad field of biology in which its close relations to evolution, development, physiology, and biochemistry are now more evident. References (1) Alexander, J., and Bridges, C. B., in Colloid Chemistry. Vol. II, Chemical Catalog Co., New York, 1928. (2) Atwood, S. S., and Sullivan, J. T., J. Heredity, 34, 311 (1943). (3) Avery, O. T. F., McLeod, C. M., and McCarty, M., J. Exptl. Med., 79, 137 (1944). (4) Bateson, W., MendePs Principles oj Heredity. Cambridge Univ. Press, London, 1909. (5) Bawden, F. C, Plant Viruses and Virus Diseases. Chronica Botanica, Waltham, 1943. (6) Beadle, G. W., Physiol. Revs., 25, 643 (1945). (7) Beadle, G. W., Chem. Revs., 37, 15 (1945). (8) Beadle, G. W., and Tatum, E. L., Proc. Natl. Acad. Sci. U. S., 27, 499 (1941). (9) Corkill, L., N. Zealand J. Sci. Tech., B23, 178 (1942). (10) Darlington, C. D., Nature, 154, 164 (1944). (11) Delbruck, M., Cold Spring Harbor Symposia Quant. Biol, 9, 122 (1941). (12) Delbruck, M., in Advances in Enzymology, Vol. II. Interscience, New York, 1942, p.l. (13) Emerson, S., Proc. Natl. Acad. Sci. U. S., 30, 179 (1944). (14) FoUing, A., Z- physiol. Chem., 227, 169 (1934). lO THE GENE AND BIOCHEMISTRY (15) Garrod, A. E., Inborn Errors of Metabolism. 2nd ed., Oxford Univ. Press, London, 1923. (16) Goldschmidt, R., Physiological Genetics. McGraw-Hill, New York, 1938. (17) Gruneberg, H., The Genetics oj the Mouse. Cambridge Univ. Press, London, 1943. (18) Haldane, J. B. S., "The biochemistry of the individual," in Perspec- tives in Biochemistry. Cambridge Univ. Press, London, 1938. (19) Haldane, J. B. S., New Paths in Genetics. Harper, New York, 1942. (20) Hollaender, A., and Emmons, C. W., Cold Spring Harbor Symposia Quant. Biol., 9, 179 (1941). (20a) Horowitz, N. H., Proc. Natl. Acad. Sci. U. S., 31, 153 (1945). (21) Horowitz, N. H., Bonner, D., Mitchell, H. K., Tatum, E. L., and Beadle, G. W., Am. Naturalist, 79, 304 (1945). (22) Irwin, M. R., and Gumley, R. W., Am. Naturalist, 77, 211 (1943). (23) Knight, B. C. J. G., Med. Research Council (Brit.), Special Rept. Series, No. 210 (1936). (24) Lawrence, W. J. C, and Price, J. R., Biol. Rev. Cambridge Phil. Soc, 15, 35 (1940). (25) Lindegren, C. C, Spiegelman, S., and Lindegren, G., Proc. Natl. Acad. Sci. U. S., 30, 346 (1944). (26) Lwofr, A., Ann. inst. Pasteur, 61, 580 (1938). (27) Mirsky, A. E., in Advances in Enzymology, Vol. III. Interscience, New York, 1943, p. 1. (28) Moewus, F., Ergeb. Biol, 18, 287 (1941). (29) Moore, A. R., Univ. Calif. Pub. Physiol., 4, 9 (1910). (30) MuUer, H. J., Am. Naturalist, 56, 32 (1922). (31) Muller, H. J., Cold Spring Harbor Symposia Quant. Biol., 9, 290 (1941). (32) Northrop, J. H., Crystalline Enzymes. Columbia Univ. Press, New York, 1939. (33) Oparin, A. I., The Origin of Life. Trans, by S. Morgulis. Macmillan, New York, 1938. (34) Painter, T. S., Texas Repts. Biol. Med., 2, 206 (1944). (35) Penrose, L. S., Lancet, 2, 192 (1935). (36) Plunkett, C. R., in Colloid Chemistry. Vol. V, Reinhold, New York, 1944. (37) Sawin, P. B., and Click, D., Proc. Natl. Acad. Sci. U. S., 29, 55 (1943). (38) Schopfer, W. H., Plants and Vitamins. Chronica Botanica, Waltham, 1943. (39) Schultz, J., Cold Spring Harbor Symposia Quant. Biol., 9, 55 (1 941). (40) Schultz, J., in Colloid Chemistry. Vol. V, Reinhold, New York, 1944. I I G. W. BEADLE (41) Sonneborn, T. M., Cold Spring Harbor Symposia Quant. Biol., 10, 111 (1942). (42) Sonneborn, T. M., Ann. Missouri Bolan. Garden, 32, 213 (1945). (43) Spiegelman, S., .inn. Missouri Botan. Garden, 32, 139 (1945). (44) Srb, A. M., and Horowitz, N. H., J. Biol. Chem., 154, 129 (1944). (45) Stadler, L. J., and Uber, F. M., Genetics, 27, 84 (1942). (46) Stanley, W. M., "Chemical structure and the mutation of viruses," in Virus Diseases. Cornell Univ. Press, Ithaca, 1943. (47) Stranskov, H. H., Physiol. Revs., 24, 445 (1944). (48) Tatum, E. L., Ann. Rev. Biochem., 13, 667 (1944). (49) Tatum, E. L., and Beadle, G. W., Ann. Missouri Botan. Garden, 32, 125 (1945). (50) Tatum, E. L., and Bonner, D., Proc. Natl. Acad. Sci. U. S., 30, 30 (1944). (51) Tatum, E. L., Bonner, D., and Beadle, G. W., Arch. Biochem., 3, 477 (1944). (52) Troland, L. T., Am. Naturalist, 51, 321 (1917). (53) Williams, R. R., Science, 94, 471 (1941). (54) Winge, O., and Laustsen, O., Compt. rend. trav. lab. Carlsberg, Ser. physiol., 22, 337 (1939). (55) Wright, S., Physiol. Revs., 21, 487 (1941). (56) Wright, S., Biol. Symposia, 6, 337 (1942). 12 VIRUSES W. M. STANLEY, member of the rockefeller institute for MEDICAL RESEARCH, PRINCETON; MEMBER OF THE NATIONAL ACADEMY OF SCIENCES D |URING the past ten years there has converged on a group of small, infectious, disease-producing entities, known as viruses, an array of scientific talent almost as diverse in nature as the Allied forces that were brought to bear on the Axis powers. The viruses are responsible for many diseases of man, animals, plants, and bacteria. There is no single criterion by means of which viruses can be differentiated from bacteria, yet the virus group has been segregated by means of certain general characteristics. Among the most important of these are small size, the ability to reproduce or multiply when within the living cells of a given host, the ability to change or mutate during multiplication, and the inability to reproduce or grow on artificial media. The sole means of recognizing the existence of a virus is provided by the manifestations of disease which result from the growth of the virus. Although pathologists have studied viruses for over fifty years and have added much to our knowledge, the new attack on viruses has been spearheaded by chemistry, chiefly biochemistry and physical chemistry, and their allied disciplines. The impact of these diverse disciplines on viruses has been accompanied by reverberations and repercussions, which, however, bode ill for viruses and good for scientific thought. Clcrtain dominant facts stand out in Ijold relief and these are understood and accepted by chemists and pathologists alike, but there is an undercurrent of indecision where neither feels 13 W. M. STANLEY quite sure of himself. This indecision becomes apparent when one considers the mode of reproduction and mutation of viruses, whether viruses are molecules or organisms, or whether some are molecules with others being organisms, or whether viruses represent a new type of structure, hitherto unrecognized and undefined, and whether one should speak of a solution or of a suspension of a given virus. For the moment, because of the lack of precise experimental data, discussion of these questions must remain more or less philosophical in nature. But since some of these questions pertain to the very nature of life itself, and others to fundamental physicochemical prob- lems, it is obvious that they are of great importance. During recent years the chemist, as well as the pathologist, has become aware of the limitations of his tools. It has become obvious that, if the tre- mendous problems posed by the viruses are to be solved, it will be necessary to forge new tools, both material and of the mind, and to carry forth the attack on a united front with a new perspective and with renewed courage and vigor. It is the purpose of this short essay to chart briefly the roads that the chemist and pathologist have already constructed into the field of viruses and to attempt to outline, in general terms, the manner in which these can be continued until they join and provide, through a common effort, a broad highway of fact and in- formation leading from the lowly electron to the lofty heights of man. Viruses range in size from about 10 m^i to about 300 m/x. Cer- tain small viruses, such as alfalfa mosaic virus, are smaller than certain accepted protein molecules, such as the Busycon hemocyanin molecules. On the other hand, certain large viruses, such as vaccine virus, are larger than certain accepted organisms, such as the minimal repro- ductive units of the microorganisms of the pleuropneumonia group. With respect to size, therefore, the viruses overlap with molecules at one extreme and with organisms at the other extreme. Since the discovery of viruses by Iwanowski, a plant pathologist, in 1892, and by Beijerinck, a chemist and botanist, in 1898, many investigations on the sizes of diflferent viruses have been carried out. Ultraviolet-light microscopes and ultrafiltration were used in most of these studies. Following the isolation of essentially pure virus preparations, methods involving ultracentrifugation, diffusion, x-ray diffraction, viscosity, osmotic pressure, and stream double refraction measurements were used with great success. Recently the chemist and pathologist have VIRUSES joined in the extensive use of the newly developed electron microscope. This instrument covers the entire range of sizes occupied by viruses and has proved, and will doubdess continue to prove, of the greatest value in the estimation of the sizes of viruses. In those cases in which more than one method has been employed, good agreement has usually been secured. The occasional discrepancies have been found to be due to errors or to a failure to appreciate the limitations of a given method, and generally have been resolved. At present the sizes of several viruses are well established and the values are accepted by both chemists and pathologists. Tobacco mosaic, the first virus to be discovered, was also the first virus to be prepared in essentially pure form and subjected to extensive chemical study. No difference was noted in virus samples prepared from a wide variety of hosts or from the same kind of host at different times of the year. The virus particles were found to consist of about 6% nucleic acid of the ribose type and about 94% protein. The exact nature of the linkage between protein and nucleic acid is unknown, but it appears to be considerably stronger than that which exists in the sperm nucleoproteins. The protein component contains definite and reproducible amounts of over thirteen amino acids. It is of interest that, in contrast to the sperm nucleoproteins, there does not appear to be an excess of basic amino acids. The single virus particles are about 280 m^u in length and 15 m^t in diameter. Tobacco mosaic virus activity has never been demonstrated to be associated with smaller particles. However, there is good evidence that a single virus particle is built up from similar subunits fitted together in a hexagonal lattice to yield the final structure which possesses virus activity. The nucleic acid of this final structure appears to exist in the form of eight threadlike units laid down along the length of the particle. Because of the repeat pattern within a single virus particle, it can be regarded as a sub- microscopic crystal. In addition, these single virus particles can aggregate with a two-dimensional regularity to form large needlelike crystals that are readily visible with a low-power hand lens. Of especial interest and significance is the fact that the virus particles appear to be utterly devoid of water and of any enzymic activity other than virus activity. The complete lack of water and the crystal- like inner structure of the individual virus particles would appear to preclude the existence of metabolism of the type usually associated with 15 W. M. STANLEY organisms. Yet, when introduced into the cells of susceptible h(jsts, these particles can direct or enter into the metabolic chain of events of the cell. The rod-shaped anhydrous tobacco mosaic virus particle is representative of a small group of viruses, but is certainly not repre- sentative of all viruses, for most viruses that have been studied ade- quately appear to be essentially spherical in shape and hydrated. Among the viruses that have been obtained in essentially pure form and studied in some detail are alfalfa mosaic virus with a diameter of about 17 m/i, tobacco ring spot virus with a diameter of about 19 m^u, tomato bushy stunt virus with a diameter of about 26 m;u, equine encephalo- myelitis and rabbit papilloma viruses with diameters near 40 m/i, influenza virus with a diameter of about 100 m/x and vaccine virus with a diameter of about 225 m/x. Of these, tomato bushy stunt virus is the only one that has been obtained in crystalline form. This virus, which contains about 17% nucleic acid, and about 83% protein, crystallizes in the form of large, beautiful, rhombic dodecahedra. The particles of bushy stunt virus appear to be strictly homogeneous with respect to size, shape, and density; hence the case for regarding these particles as molecules is as good as for any protein. The papilloma virus appears to be a nucleoprotein containing about 8% nucleic acid and little or no lipid. The particles of the equine encephalomyelitis virus appear to be a liponucleoprotein complex containing about 48% lipid, about 5% nucleic acid, and protein plus a small amount of excess carbohydrate. Data on influenza virus indicate that the 100 m/j, particle has a water content of about 60% by weight, with the solid portion being composed of about 65% pro- tein, about 25% lipid, about 7% carbohydrate, and a very small amount of nucleic acid. Vaccine virus is the largest and most complex virus that has been subjected to chemical investigation. The prepara- tions were found to contain protein, lipid, carbohydrate, and thymus nucleic acid in concentrations not materially different from those found in bacterial cells. The vaccine virus preparations were also found to contain phosphatase, catalase, lipase, biotin, riboflavin, flavin-adenine- dinucleotide, and apparently significant and reproducible amounts of copper. It is exceedingly difficult to prove that all of these repre- sent integral components of vaccine virus, but it must be regarded as significant that this large and complicated virus appears to retain i6 VIRUSES certain enzymic activities quite tenaciously, whereas tobacco mosaic and bushy stunt viruses appear lo possess no enzymic activities other than that of virus activity. It is also of interest that, in marked con- trast to bacteria and other ceils, tobacco mosaic and influenza viruses contain negligible amounts of the B vitamins. Electron micrographs of vaccine virus, as well as of certain bacteriophages, have revealed the presence within individual particles of an internal structure con- sisting of a pattern of granules. As a whole, the data now available on viruses indicate that, as one goes from the small to the large viruses, there is, with increase in mass, an increase in complexity of composition, structure, and function. The viruses appear to provide, in truth, a bridge between proteins and organisms. Indeed, if one wishes to regard the transformation agent of the pneumococcus as a virus, the bridge could be extended to nucleic acid. If one were starting out anew to construct a link between the molecules of the chemist and the organisms of the pathologist or bacteriologist, it is difficult to visualize how it would be possible to improve upon what Nature has already provided. The structural complexity encountered in tomato bushy stunt virus nucleoprotein is but little more complex than that of hemoglobin and no more complex than that of the hemocyanins. Physically and chemically these behave as molecules; and if it were not for the virus activity of the bushy stunt nucleoprotein it would not be given a second thought. Between bushy stunt virus and vaccine virus. Nature has provided a continuous series of structures of gradually increasing mass and complexity, all linked by a common biological property, virus activity. Vaccine virus is as large as some organisms that can be grown on artificial media; and if vaccine virus could be grown on artificial media it would be accepted generally as an organism. Even larger structures exist which cannot be grown on artificial media and which are just as fastidious as vaccine virus with respect to growth requirements; yet these are accepted as organisms. Nature has provided us with an accomplished fact, and it is time for all to brush away the barriers of the mind and to recognize the possibilities that are provided by the viruses. Although viruses are disease-producing agents and have caused untold suffering, the com- plete acceptance of Nature's dubious gift may not be without recom- pense. For its acceptance and exploitation may provide ihc key to broad and wonderful vistas. 17 W. M. STANLEY Pathologists and bacteriologists have labored long and arduously with viruses. They have found that a given virus will reproduce only within the cells of certain specific living hosts, and that, although some viruses will reproduce within the cells of several different hosts, other viruses will multiply only within the cells of one given host or sometimes only within certain specialized cells of that host. They have shown that the primary pathological changes produced in cells by viruses are either proliferative or degenerative in character. In some virus dis- eases, such as yellow fever, poliomyelitis, and tobacco necrosis, de- generative changes predominate; but in many, as in smallpox and fowl pox, both proliferation and necrosis occur. Still other virus diseases, such as Rous chicken sarcoma, Shope rabbit papilloma, and tobacco enation mosaic, are characterized by a rapid and unorganized cellular proliferation. Long before the discovery of viruses, a means was recognized of protecting man against the virus disease, smallpox. This was achieved by vaccination with active virus, presumably altered by passage in an unnatural host. However, it has only been in recent years that there has come a full realization of the great benefits, both with respect to methods of protection against virus diseases and to the study of the viruses themselves, that can be achieved through the use of new virus hosts. Yellow fever has been eliminated as a major health problem, and much has been learned of the virus because the virus was taken from man and grown in monkeys, in mice, and in chick embryos. The possibility of a recurrence of the 1918 influenza epidemic, which killed more people than have died from combat activities in World Wars I plus II (to date of writing), has been reduced and perhaps eliminated because of the production of a vaccine which was made possible by the growth of this virus in the chick embryo. Truly re- markable progress has already been made in the study of viruses and in the prevention of virus diseases, and it is to be expected that this progress will continue. Yet withal, the one fundamental and all- important problem posed by the viruses — that of the mode of virus reproduction — remains unsolved. For many years little hope for a solution of this problem was held, for viruses were generally regarded as living organisms and the nature of life was considered to be a hal- lowed, insoluble secret. However, chemists recognized in the virus activity of certain crystallizable nucleoproteins a type of biological i8 VIRUSES activity somewhat akin to that possessed by certain protein hormones and enzymes. To them, virus activity appeared no less wonderful and possibly no more complicated than the changes that can be induced within cells by enzymes and hormones. The mental barrier of the living state is being eliminated gradually and chemists are recognizing and accepting Nature's gift of the viruses. The true issue is only beclouded by the insistence in some quarters for a decision as to whether a given structure is living or inanimate. The fundamental meaning- lessness of such terms has been commented on before and is becoming ever more apparent. In the meantime, the requirements for the solution of the riddle of virus reproduction — perhaps the most impor- tant problem in all of biochemistry — are becoming clearer. Many studies on the nature and mode of virus reproduction are in progress, and it seems certain that such studies will increase in the future, not only in volume, but also in scope and in diversity. It is possible that the solution of this very important and fundamental problem will not be realized until after new weapons of the hand and mind have been brought into play. So far, the most spectacular ad- vances have been made along three lines, each of which merits con- siderable further attention: (a) studies on the very favorable bacterial cell-bacteriophage system; (b) studies on the changing of the chemical structure of a virus by means of known chemical reactions; and (c) studies on the nature of the differences in chemical structure that are responsible for the existence of virus strains. The bacterial cell- bacteriophage system provides an extraordinary opportunity to follow the interaction of a virus with its host cell. The host cells can be grown in vitro on artificial media in large quantities and under constant and reproducible conditions. The metabolism of the host can be con- trolled to a certain extent and analyses can be made on the system throughout the reaction. The bacterial and phage or virus materials can be differentiated up to the entrance of the virus into the cells and following the lysis of the cells. Studies on this system have permitted the conclusion that multiple infection of a bacterial cell with several virus particles of the same type has the same effect, both qualitatively and quantitatively, as infection with a single virus particle. It was also found that infection of a bacterial cell with two kinds of virus particles resulted in the reproduction of only one kind and the sup- pression of growth of the other kind. However, more than one virus 19 W. M. STANLEY can multiply in a single cell if the viruses are sufficiently distinct in their requirements. These and similar results have permitted many stimulating inferences regarding the mode of virus reproduction and have suggested innumerable approaches for future experimentation. Work on the changing of the chemical structure of a virus by means of known chemical reactions has been both encouraging and dis- couraging. It has, in fact, proved possible actually to change the chemical structure of a virus; but so far no change has been found to be perpetuated in the virus particles produced as a result of infection of a host with the altered structures. Thus, the abolishment of the sulfhydryl groups in tobacco mosaic virus or the introduction into the structure of this virus of several thousand acetyl, phenylureido, carbo- benzoxy, benzene sulfonyl, or malonyl groupings yields diverse altered virus structures. Although these are infectious, the disease which they produce is the ordinary tobacco mosaic disease, and is accompanied by the production of particles, not of the respective altered structures, but of ordinary tobacco mosaic virus. However, encouraging results were obtained in a study of the specific virus activity of these chemical derivatives on different hosts. It was found that a property of the virus, which perhaps can best be described as virulence, can remain constant for one host but be modified with respect to a different host, upon formation of a given chemical derivative of the virus. This result lends encouragement to the belief that, eventually, heritable structural changes in a virus will be achieved in the chemical labora- tory by means of known chemical reactions. Contemplation of the implications that would accompany the actual accomplishment of this feat tends to stagger the imagination. Spectacular progress has attended studies on the nature of the differences in chemical structure that are responsible for the existence of virus strains. It was indeed fortunate that Nature provided, and the pathologist recognized and separated, two or more strains of each of several different viruses. Strains of a virus appear to arise during the reproduction of a virus by a process which can be regarded as similar to that of gene mutation. It can be presumed that the strains of a virus have arisen from some parent strain, one by one, during the course of the years. Each strain thus probably bears a definite rela- tionship to the parent strain and to each of the similar strains. Each strain causes a more or less different disease; because of this fact it 20 VIRUSES has been possible foi' the pathologist to recognize, separate, and grow individual virus strains. Some of these have been found of girnt use- as vaccines for the prevention of certain virus diseases. In view of the large size and complexity of structure of \inises, ii may appear that chemists were somewhat optimistic in expecting to be able to detect differences in the chemical structure of different strains of a virus. Two similar large mountains may appear identical when viewed from a distance, but close inspection will reveal differences. So, too, with the viruses. The over-all structures of different strains of tobacco mosaic virus were found to be very similar; yet in several instances it has been possible to demonstrate definite differences in chemical structure. These consist of differences in the amount of one or more amino acids, in the presence of an entirely new amino acid, or in the complete elimination from the virus structure of a given amino acid. These changes represent deep seated and fundamental altera- tions in the virus structure, and it seems unlikely that they could have resulted from alterations of fully formed virus particles. It appears more likely that these changes occurred as a result of a dixersion of the synthetic process by means of which a virus reproduces. Since new strains tend to appear or to become dominant when a virus is grown in an unnatural host, it is possible that the altered environment of this host provides a somewhat different supply of amino acids and enzyme systems, and in the effort to adhere to some basic pattern it becomes necessary to build into the virus structure amino acids that would not be used normally. The drive to follow a basic pattern and the aberra- tions that result bear a certain kinship to the forces of heredity. As a matter of fact, there is a striking similarity between the properties of viruses and those that have been ascribed to genes. Both may be re- garded as large nucleoprotein structures that have the ability to perpetuate themselves within, and only within, certain specific living cells. Both can undergo sudden changes, apparently cither spon- taneously or as a result of external factors and those changes are then reproduced in subsequent generations. Within limits, the concen- tration of both in cells can be changed by proper treatment of the host. Some viruses appear to be concentrated in the cytoplasm of cells and others in the nuclei. The similarity between viruses and genes may not be without significance, for the abode of genes is the cell and no virus has been 21 W. M. STANLEY proved to arise de novo — they are always first found in cells. Viewed in this light, the diflferences in chemical structure that have been demon- strated to exist between different virus strains, and especially the chang- ing of the chemical structure of a virus in a definite manner by means of known chemical reactions, take on a new and perhaps startling significance. There are many virus diseases in which a symbiotic re- lationship is set up between virus and host. Although the virus enters into and alters the metabolic activity of the cell, the cell survives and continues to divide. The virus is carried continuously within the cells, and in some cases could almost be regarded as a normal com- ponent of the cells because of the lack of obvious damage. In fact, one virus has come to be known as the "healthy potato virus" because it is present in almost all potatoes grown in this country and yet the potato plants appear healthy. It is as if one had introduced; from without, a nucleoprotein which was accepted by the cell as a part of its own germ plasm. The fact that different strains of a virus, which can cause different manifestations of disease, are characterized by different nucleoproteins, and especially the fact that the structure of these nucleoproteins can be changed in the test tube by means of known chemical reactions, could be interpreted to mean that eventually the germ plasm of cells may prove to be susceptible to similar chemical manipulation. The viruses have assuredly provided a link between molecules and organisms, and there now exists a pathway leading from simple structures, such as the electron, to massive, highly complex structures, such as man. This pathway is broad and well established in some places and narrow and difficult to traverse in others. But as the latter are broadened and placed on a firm foundation, through the common effort of chemists and pathologists, it is possible that information will be acquired which could affect the future destiny of every living being in the world. Selected References Anson, M. L., and Stanley, W. M., J. Gen. Physiol., 24, 679 (1941). Bawden, F. C, Plant Viruses and Virus Diseases. Chronica Botanica, Wal- tham, 1943. Beard, J. W., Bryan, W. R., and Wyckoff, R. W. G., J. Infectious Diseases, 65, 43 (1939). Bernal, J. D., and Fankuchen, I., J. Gen. Physiol, 25, 111, 147 (1941). 22 VIRUSES Cohen, S. S., and Stanley, W. M., J. Biol. Chem., 144, 589 (1942). Doerr, R., and Hallauer, C, Handbuch der Virusjorschung. Springer, Berlin, 1938-39. Green, R. H., Anderson, T. F., and Smadel, J. E., J. Exptl. Med., 75, 651 (1942). Hoagland, C. L., Ann. Rev. Biochem., 12, 615 (1943). Hoagland, C. L., Ward, S. M., Smadel, J. E., and Rivers, T. M., /. Exptl. Med., 74,69,133(1941). Hoagland, C. L., Ward, S. M., Smadel, J. E., and Rivers, T. M., J. Exptl. Med., 76, 163 (1942). Knight, C. A., J. Biol. Chem., 145, 11 (1942). Knight, C. A., and Stanley, W. M., J. Biol. Chem., 141, 29 (1941). Lauffer, M. A., J. Biol. Chem., 143, 99 (1942). Lauffer, M. A., J. Am. Chem. Soc, 66, 1188 (1944). Lauffer, M. A., and Ross, A. F., J. Am. Chem. Soc, 62, 3296 (1940). Lauffer, M. A., and Stanley, W. M., J. Exptl. Med., 80, 531 (1944). Luria, S. E., and Anderson, T. F., Proc. Natl. Acad. Sci. U. S., 28, 127 (1942). Luria, S. E., and Delbriick, M., Arch. Biochem., 1, 207 (1942). MUler, G. L., and Stanley, W. M., J. Biol. Chem., 141, 905 (1941). Rivers, T. M., Viruses and Virus Diseases. Stanford Univ. Press, Stanford Uni- versity, 1939. Rivers, T. M., et al., Virus Diseases. Cornell Univ. Press, Ithaca, 1943. Stanley, W. M., Physiol. Revs., 19, 524 (1939). Stanley, W. M., Ann. Rev. Biochem., 9, 545 (1940). Stanley, W. M., J. Biol. Chem., 135, 437 (1940). Stanley, W. M., Sci. Monthly, 53, 197 (1941). Stanley, W. M., J. Exptl. Med., 79, 267 (1944). Stanley, W. M., and Anderson, T. F., J. Biol. Chem., 139, 325 (1941). Stanley, W. M., and Knight, C. A., Cold Spring Harbor Symposia Quant. Biol., 9, 255 (1941). Taylor, A. R., Sharp, D. G., Beard, D., and Beard, J. W., J. Infectious Diseases, 72, 31 (1943). Williams, R. J., Schlenk, F., and Eppright, M. A., J. Am. Chem. Sac, 66, 896 (1944). 23 PHOTOSYNTHESIS AND THE PRODUCTION OF ORGANIC MATTER ON EARTH H. GAFFRON, research associate, professor, departments of bio- chemistry AND chemistry (fELS FUNd), UNIVERSITY OF CHICAGO Practical Importance of the Assimilation of Carbon Dioxide We know, for example, that if we abuse the soil, it will lose its fertility, that if we massacre the forests, our children will lack timber and see their uplands eroded, their valleys swept by floods. Nevertheless, we continue to abuse the soil and massacre the forests ... in the simple affairs of nature, where we know quite well what is likely to happen we immolate the future to the present. " Those whom the gods would destroy they first make mad." K ALDOUS HUXLEY, Time Must Have a Stop, p. 298. TVEN the less informed layman is aware of the reaction which makes possible the abundance of living things on earth- — the conversion in daylight by plants of carbon dioxide and water, the waste products of plant and animal life, back into food. This reconversion is called the assimilation of carbon, or photosyn- thesis. Hardly ever does the layman spend another thought on this fundamental and quite spectacular achievement of living cells. If, however, he suddenly realizes that, without this reaction, life as we know it would perish quickly and completely (except for a few species of bacteria), he is likely to place false hopes on its artificial reproduction by the toiling scientist. He probably believes that mankind's in- 25 H. GAFFRON adequate supplies of food and dwindling stores of energy will turn into a surplus the moment photosynthesis has been duplicated in the test tube. Popular articles and radio talks — spreading the lamentable misconception that scientific research represents nothing but the first step in technology — foster this notion by stressing the practical im- portance of artificial food production from sunlight and carbon dioxide. The error is easily demonstrable. The solar energy flux per acre and year is a constant. Unless our devices are to be much more efficient than the green plants and unless we are willing to spoil with ugly ma- chinery an acreage equivalent to that now covered by beautiful forests and pastures, artificial photosynthesis is not only Utopian but im- practical. Even assuming we were to discover some sort of artificial photochemical reduction of carbon dioxide into a digestible carbo- hydrate with an over-all efficiency surpassing that of the plants which use on the average two per cent of the incident radiation, it would not help us much, since we need, not one product of plant metabolism, but a thousand (16). Let us mention only the proteins among foodstuffs, and wood and rubber among industrial raw materials. True, chemists have learned to make ammonia and nitrates from atmospheric nitrogen. In that respect man is no longer dependent on other natural sources. But up to now only the plant converts these simple nitrogen compounds into proteins,* and so we are forced in a second way, equally fundamental, to rely on the growth of plants for our continued existence. In order to stretch our limited supply of organic substances already formed, such as coal and petroleum, the production of most compounds which plants can synthesize and which we need in large quantities should be reserved for natural photosyn- thesis, regardless of whether the chemist can duplicate the synthesis in vitro. Oil, for instance, is too versatile a material to be converted into rubber as long as plants are capable of continuously producing essential raw materials like alcohol, or better still, rubber of excellent quality, t Denying that artificial photosynthesis will be the solution to a serious and fascinating problem is not equivalent to saying that sun- * Nonphotosynthetic plants like yeasts are a good source of protein, pro- vided they are fed with products of photosynthesis. t Guayule plants grown in California produce rubber far superior in automobile tires to the synthetic compounds now available. 26 PHOTOSYNTHESIS light will not be used at all in the technology of the future. In selected places, such as roof tops in sunny countries, bare rock, and deserts inac- cessible to irrigation, sunlight may soon be utilized to produce heat or electric power or even to drive some photochemical process like that of hydrolysis. These possibilities have exactly as much and as little re- lation to the problem under discussion as exists between any other common source of energy and a filled granary. There is every indica- tion that products of plant photosynthesis will be needed ever more urgently, not only as food but also as fuel and building material. Re- cent technological mventions indicate that wood, strengthened with wood-derived chemicals, will be in even greater demand than now. One and one-half billion of the two billion humans on earth are illnourished or permanently hungry, "There has never been enough food for the health of all people" (18). Since this globe offers only a certain area of habitable and tillable ground, it is obvious that populations will have to be adjusted to a certain optimum density determined by the general standard of living that man is capable of attaining or willing to endure. We have ample testimony of the improvident way in which the ancients exhausted their supply of wood. The mountains of Persia, Syria, Greece, Dalmatia, Italy, and Tripolitania are now to a large extent barren and infertile. The goats of the Arabs are said to have done away with the last traces of vegetation in North Africa, thus allowing the desert sands to advance to the shores of the Medi- terranean. The changes in climate brought about by the denuding and erosion of the sites of the most important early civilizations make it difficult or impossible to retrieve the lost fertility of the land. Are we wiser today? The rate of consumption of coal and pe- troleum in the world today (1.75 X 10^ tons carbon per year) is roughly one-tenth of the rate of the total carbon assimilation achieved by the land plants on earth (1.6 X lO^" tons carbon per year) (14). The coming industrialization of Asia will soon diminish the gap be- tween the demand for, and the supply of, organic material. The established reserves of oil in the United States, according to official figures, are about twenty billion barrels. They are being used up at a rate of one and one-half billion barrels per year. The industry insists upon a more optimistic outlook, based on the (ever-declining) rate of new discoveries. True or not, considering also the certainly 27 H. GAFFRON limited supply of coal, it is quite clear that we are spending mainly as fuel irreplaceable organic material. Fig. 1 . — Virgin forests in 1 850. Can these losses of stored products of photosynthesis be offset by the assimilation of carbon going on today? Within the cultivated areas of industrialized countries this is not the case, and foresters agree that virgin forests are more or less stationary. New growth balances natural decay. Only well-planned agriculture and expert forestry may perhaps furnish all that we need in the future provided the increasing demand for fuel and energy is eventually met by the general development of atomic power. At present, the products of agriculture are consumed within a few years after harvesting. There is no increase of our reserve of organic carbon due to this source. On the contrary, the improvident exploitation of the soil in many places leads to di- minishing returns. In the United States, fifty million acres now under cultivation are so badly eroded as to invite abandonment. These sad circumstances have received widespread attention, and effective measures are being taken to check further losses and to regain the lost fertility of the soil (13). Not so well known are the conditions regard- ing the forests in this country (1). The following figures speak for themselves: 28 PHOTOSYNTHESIS Billion board feet Original stand of forests 5,200 on 850 million acres Present (1927) stand of forests 2,500 on 550 million acres Rate of annual consumption 35 We are cutting down our forests three times as fast as they grow. Certain species of trees are in danger of extinction. War conditions have aggravated the situation. Figures referring to the years just before the war for the states of Washington, Oregon, Mon- tana, and Idaho, which now supply half of the timber cut in the United States, are as follows (3): Billion board feet Annual cut and losses in the Northwest 12.5 Current annual growth 4.2 Potential annual growth after universal adoption of best forest practices 23 Fig. 2. — Virgin forests in 1945. The last figure shows that we should not abandon all hope. There is a well-tested way to prevent the exhaustion of our wood reserves. But it is not enough for the government to administer the national forests i)rovidently. Industrialists and private owners hold 29 H. GAFFRON not only the larger area of commercial forest land but also the more valuable timber from the standpoint of accessibility and of quality. "The main objective of this group is to market their lumber as speedily as possible" (3). Management according to the principles of far- sighted modern forestry is encountered only on an estimated two per cent of the total area. The practices of commercial competition are such that in general only the State can afford to look ahead two or three generations. Meanwhile, the waste continues. According to N. G. Brown (1) less than half of the total amount of wood cut in this country is ulti- mately utilized. Hence it is imperative that the efforts of the Depart- ments of Agriculture and of the Interior to remedy the situation should have the conscious support of every citizen. Unless we plan for a permanent forestry everywhere, not only our supplies of oil but also those of wood will be gone in about seventy years and the country dependent upon foreign sources. No wonder that scientists begin to turn toward the oceans as a source of products of photosynthesis; it is estimated that the amount of carbon assimilated by marine algae surpasses by four or five times the yield of the land plants (14). But when can we expect to replace wood by plankton? The journals of chemical industry often display an advertisement showing a magnifi- cent mountain forest with a caption from which we quote: "Ever see a forest through a chemist's glasses? Do you see . . . plastics . . . plywood . . . laminated beams . . . silk . . . smokeless powder . . . rolls of newsprint . . . movie film . . . All these and many more items of beauty, strength, and utility the chemist makes from wood, holding out brighter hopes for a better future." If man continues to consume the products of photosynthesis in the way he does at the present time this kind of better future may not last very long.* The understanding of photosynthesis becomes essential if we are to solve part of these rapidly approaching difficulties, not in order to reproduce the process technically on a big scale but rather in order to * Two publications have recently appeared on the subject of our dwindling supply of wood: "Forestry and the public welfare," in Proc. Am. Phil. Soc, 89, 399 (1945); and "What's happening to the timber," by R. A. H. Thompson, in Harper'' s Magazine, issue of August, 1945, page 125. The instructive maps (Figs. 1 cmd 2) from the latter article are reproduced through the courtesy of Harper's Magazine. 30 PHOTOSYNTHESIS increase its natural efficiency. As everybody knows, different plants grow in different locations. The efficiency of the chloroplasts or of the photochemical mechanism proper may be equal in all plants, yet some plants may not have attained the optimum in utilizing their own photosynthetic products. Here the knowledge of internal factors influencing photosynthesis will show which plants to breed. Investi- gations on the photosynthetic behavior responsible for the over-all efficiency of plants in producing the coundess substances we need will increase in practical importance. Much has been done and will be done empirically by the gardener, agriculturist, and forester, but the shortest way to success lies in systematic investigations of the correlation between growth or fruition of plants and the rate of photosynthesis. Such work is in progress, for instance, at the California Institute of Technology (22). The intimate connection with studies concerning such factors as soil, climate, and inheritance are obvious. In contrast, it appears doubtful at this moment whether the analysis of the nature of the photosynthetic process can produce imme- diate practical results surpassing those obtainable by the kind of studies mentioned above. True, pure reseaixh remains the only source of sudden technical advances. At the present time, however, the ques- tion as to the nature of photosynthesis is a problem of science and not of so-called "applied science," that is, technology. The reward in joining the few who insist upon spending their time in tackling this problem will be only the pleasure of knowing a little more and seeing a little farther than those who worked on the same problem twenty years ago. Partial Reactions in Photosynthesis " The enormous amount of research upon the process of photosynthesis during the past half century has thrown little or no additional light upon the subject. The problem is evidently too complex for any specialists in any one field of scierue to solve. ..." E. C. MILLER, Plant Physiology, 2nd ed., 1938 During the past decade a new laboratory technique — that of enzyme chemistry — has been developed and is at present taught to every student in cell physiology. The isolation of an enzyme is an advance that requires no mental effort to appreciate. Equally obvious are the advantages of using traceable isotopes for elucidating the course 31 H. GAFFRON of obscure metabolic reactions. Among younger biochemists the attitude toward a field like photosynthesis is one of waiting until dis- agreeable obstacles have been cleared away by hand, so to speak, so that they may roll in with modern methods. During the past twenty years the moment when this will happen has quietly drawn nearer. How near we shall try to demonstrate in the rest of this essay. It is mostly forgotten that, after Buchner had demonstrated fermentation outside the living cell, some thirty years passed before enzymes were isolated and the mechanism of organic catalysis could be said to be clearly understood. Progress in the field of cellular break- down reactions, slow in the beginning of this century, nevertheless encouraged a handful of students to devote their lives to the investiga- tion of respiration and fermentation. In contrast, the return for the toil directed toward solving the problem of photosynthesis was so small that even the leading biochemists, after establishing a few important new facts such as the existence of a photochemical reaction distinguish- able from an enzymic one, saw no point in pursuing the analysis any further. In 1918, Willstatter wrote that it was evidently too early to try to elucidate the mechanism of carbon dioxide assimilation in living cells. We now know that he was right. At that time photo- synthesis appeared as an absolutely unique process showing no con- nection or analogy with other metabolic processes in the cell. It was looked upon as a direct photochemical decomposition of carbon dioxide and mysteriously connected with the phenomenon of life on earth. The absolute ignorance of the kind of reactions photosynthesis might involve and, consequently, the lack of a theoretical framework accurate enough to direct a reasonable approach barred further progress. And as to purely empirical experimental attempts, they all ended with the destruction of the intact living cell. Photosynthesis stops the moment the cell is hurt. While this is still true, of course, the great diflference is that today we know more or less why. Hence there is hope we shall be able to overcome this difficulty. Despite Miller's verdict quoted above, there has been very decisive progress between 1918 and 1938, not merely by the addition of several important observations to a hundred earlier ones but, mainly, by the change in our conception of what constitutes the problem of photosynthesis. This change was due, on the one hand, to the under- standing of the nature of other metabolic processes such as respiration 32 PHOTOSYNTFiESIS and fermentation and, on the other, to a clearer knowledge of the essence of photochemical reactions in complex molecules. Twenty- five years ago there was an overabvmdance of uncorrelated general observations and only a few \ery simple pictures of the mechanism of photosynthesis. The latter were hardly supported by any straight- forward experiments or permissible analogies. I need mention only the concepts of a chlorophyll carbonic acid complex, of the release of oxygen from this complex, and of the formation and polymerization of formaldehyde, which are still faithfully reproduced in most textbooks of botany. About 1930, van Niel (19) pointed out that photosynthesis should be considered as a coupled oxidoreduction comparable to other reactions of this kind. This approach to the problem proved very fruitful and was soon generally adopted. Today, in 1945, we can divide the process of photosynthesis into several partial reactions, each with its particular problems, none of which seems to present insurmountable conceptual difficulties. By drawing upon analogies with other metabolic processes and from the results of direct experi- ments it is possible to build up a theoretical picture which, though incomplete in many places, satisfies the essential requirement that the main observations can be correlated. The mystery of photo- synthesis is mainly gone, and this in a rather fundamental sense. We are now quite certain that photosynthesis promoted by chlorophyll in visible light has nothing to do with the origin of life. Instead, it must be regarded as a rather late achievement of the living cell. It is a unique combination of a few reactions found only in the green plant, with important devices characteristic of any kind of metabolism in living cells. With the acquisition of pigments, the early living cells became able to accelerate the reaction between hydrogen donors and acceptors by absorbing radiant energy. The biological currency, H and OH, the constituents of water, became available in larger quantities because of the interaction between the irradiated dye and water. Despite the photochemical reduction of carbon dioxide no true gain in free energy of the complete system was yet possible, since the resulting "hydroxyl- ated" counterparts {cf. reference 20) could only re-form water with some valuable hydrogen donor. Any increase in the amount of organic matter depended on the presence of inorganic hydrogen donors such as free hydrogen, hydrogen sulfide, sulfur, ferrous iron, 33 H. GAFFRON etc., and hence could not proceed very rapidly. The picture changed radically, however, the moment some cells happened to combine the photochemically accelerated utilization of water as hydrogen donor with a reaction allowing for the elimination of oxidation products by liberating free oxygen. As long as there was an abundant supply of carbon dioxide, of water, and of light energy, the accumulation of organic matter could continue unchecked. Oxygen appeared in quantities as free gas in the atmosphere. This was followed by the enormous multiplication of organisms capable of using the oxidation of organic compounds for synthetic reactions.* Carbon dioxide released by respiration was again utilized in the photosynthetic reaction. Some million years later, the cycle of carbon arrived at a steady state. The concentration of carbon dioxide in the air is now 0.03% and barely sufficient to support maximum photosynthesis in full sunlight. Most plants do better when carbon dioxide is added artificially. This picture of photosynthesis as a process that developed gradually from less complicated reactions is supported by the following observations. In the plant, the catalytic systems bringing about the assimilation of carbon dioxide can be distinguished by the differentiat- ing effects of metabolic poisons. The whole process is specific for carbon dioxide, just as other metabolic reactions are specific for their substrates. The various catalysts seem to consist of proteins combined with special reaction groups. Chlorophyll itself is bound to protein in the cell. The reduction of carbon dioxide is not an exclusive privilege of the green plants. Besides normal photosynthesis we dis- tinguish at present the following types of carbon dioxide reduction: (a) In the dark (10,23) coupled with (7) bacterial fermentations (methane bacteria, propionic acid bacteria), (2) bacterial and plant oxidations (sulfur bacteria, "Knallgas" bacteria, unicellular algae), (3) metabolic reactions in animal tissues. (b) In the light (6,7,19), with the simultaneous consumption * Van Niel has demonstrated a gradual adaptation to aerobic conditions with some strains of purple bacteria, which at first grew only anaerobically in the light (20). 34 PHOTOSYNTHESIS of inorganic or organic hydrogen donors (purple bacteria, unicellular algae). Unique to the green plants is the coupling of photochemical re- duction with an evolution of free oxygen. Such an evolution of oxygen can be obtained by illuminating isolated chloroplasts in the presence of oxidizing substances like ferric salts or jf?-benzoquinone (5,9,21). The existence of several more or less autonomous enzyme systems working together can best be demonstrated in unicellular algae be- longing to the Scenedesmaceae. In these algae it is possible to inter- change at will all three known types of carbon dioxide reduction; a chemoreduction promoted in the dark by the burning of hydrogen with oxygen to water; a photoreduction in which two hydrogen molecules disappear together with a molecule of carbon dioxide; and a complete photosynthesis with the liberation of an equivalent amount of oxygen (7). A rough summary of the way in which the problems of carbon dioxide reduction in an alga like Scenedesmus can be subdivided accord- ing to present knowledge is given in the scheme below. More complete and complicated schemes can be found in a recent review (7) and in Rabinowitch's new book (14). COj fixation of carbon dioxide / carboxyl group carbohydrate i (CHjO), Light absorbed by plant pigments (chlorophylls, xanthophylls, phycocyanins) V . . t . photochemical reaction dismutation to (chlorophylls, water, enzymes) oxygen and water ^ . ^.- ^ intermediate intermediate hydrogen a hydrogen donors > .2 < acceptors \ intermediate hydrogen donors \ u u \ hydrogen acceptors Oj oxidoreduction (different from respiration) intermediate hydrogen acceptors / ^ hydrogen donors H2 reduction to water i H,0 35 H. GAFFRON In addition to what has been said in the preceding paragraphs about the probable evolution of the photosynthetic system and about the three main types of carbon dioxide reduction, the scheme tries to coordinate the following fairly well-established facts (13): (7) Chlorophyll is necessary even in those cases in which the light is effectively absorbed by other plant pigments such as xantho- phyll and phycocyanin. (2) The oxygen liberated does not originate from carbon dioxide but from molecules of water entering somewhere as ultimate hydrogen donors into the reaction. (J) Carbon dioxide is fixed initially by way of a reaction which is reversible and nonphotochemical, probably in the form of a car- boxy 1 group. {4) There are several intermediate steps between the photo- chemical reaction proper and the appearance of free oxygen. The proof for this is found in experiments with specific inhibitors and in the fact that the reversible switching from oxygen evolution to an equivalent consumption of hydrogen ("photoreduction") does not change the quantum yield. (5) The chemoreduction in the dark has many traits in com- mon with photoreduction. Yet we should expect differences between some of the intermediates produced and utilized in the dark and those made under the impact of a light quantum with an energy content of 40,000 calories. Since neither the reactions of chlorophyll or of any other catalyst involved nor the nature of any one of the latter has been clearly estab- lished, the scheme will convince the reader that there is a host of problems to be solved. The interested investigator is likely to look upon that partial problem as the most exciting and, hence, important one to whose solution he believes he can contribute something sig- nificant. Since each of the partial reactions of which the whole of photosynthesis consists is indispensable, truly none can be more im- portant than the other. However, the utilization of radiant energy and the evolution of free oxygen sets photosynthesis apart from other better analyzed metabolic reactions. It is here, therefore, where we must expect the greatest deviations from the common types of cellular catalyses and where we may find the unprcdicted. Though the picture given in the scheme could be seen to PHOTOSYNTHESIS emerge gradually, since about 1930, from the discussions accompanying the newly gained experimental data (van Nicl, Stoll, Emerson, Wohl, Franck, and others), the interrelation between the different types of carbon dioxide assimilation has struck some investigators quite sud- denly. In their enthusiasm over this fact they are likely to overlook the peculiar problems which {photosynthesis offers in contrast to, and distinction from, the other metabolic processes. To them, now, all is very simple. The light absorbed by chlorophyll is used to mobilize hydrogen. Water is decomposed into H and OH. And once hy- drogen is available the reduction of carbon dioxide proceeds as a dark reaction exactly as in the cases mentioned above. This has led to some strange expressions and statements, such as "photosynthesis in the dark" or "light per se is not essential for photosynthesis." Lately, the analogy between thermal and photochemical re- actions has been pushed still further, and attempts have been made to bring the so-called energy-rich phosphate bond into the picture. Be- fore we analyze these attempts we must say a few words about inter- mediates. In Search of Intermediates In taking apart the mechanism of photosynthesis, we shall in all probability find two types of intermediates: on the one hand, the enzymes necessary for the transfer of hydrogen and the removal of oxygen; on the other, the final acceptors, the precursors of carbo- hydrates and of oxygen. The former we may compare to the pyridine nucleotides, flavoproteins, and cytochromes, the latter, to the deg- radation products of glucose, with no clear counterpart to what has been called the "photoperoxide" or "peroxide" or "moloxide" or "hydroxylated compounds" (van Niel) in photosynthesis, that is, the hypothetical substance which decomposes with the evolution of oxygen (we are pretty certain that it is not hydrogen peroxide and perhaps it is no peroxide at all) . Most of the intermediate catalytic steps in respiration and fermentation have been shown to be reversible — even the decarboxyla- tion of pyruvate, if it proceeds with the formation of acetyl phosphate (11). The only requirement is the coupling with another reaction, furnishing the energy to reverse the particular step. Without such a coupling the breakdown processes continue unchecked until the specific 37 H. G.\FFRON substraies are e>diausted. If such an easy reversibilirs^ existed e\er\- where in the reaction chain leading firom carbon dioxide to carbo- hNxlrate, an accumulation of photos^•nthetic products would be un- likely. In fact, it is quite essential for the efficiency of the s^Tithetic jMXXXSS that the carboh\tlrate finally formed is not broken dowTi again in the dark by the same catalytic s^^stems which helped to btiild it up. One of the difficulties we shall encounter in tr\-ing to reconstruct the phota5>-nthetic mechanism tn vitro with enz\-mes isolated fix)m the plant cell will probably be the reversibLlir>- of partial reactions, tinless we succeed in remoxing and thereby stabilizing the intermediate just fcHined. At least, the way "doxsTi" is the normal course of events as demonstrated in respiration and fermentation. Now, save for attack by the respiratory or glycoh-tic s\"5tems, the carbohydrate sxTithesized in the light is stable (a ver\- slow reversion is postulated for theoretical reasoDS not discussed here). The solution of the problem, therefore, may be hidden at either or both ends of the photosynthetic s%"stem. Tlie primar\- tinstable carbohNxirate may pohTnerize with the release of some free energ\-. One might even think of a reduction up to the alcc^ol ^OT h\-drocarbon) level and an oxidation by another path back to an aldeh\-de (or alcohol) . At the other end of the line, the ox\-gen- liberating s\"3tem should have prop>erties making it rather ineffective as an oxidase. At any rate, all attempts to cause an accimiulation of intomediates by mistreating the irradiated cell .with narcotics or specific poisons have failed. Hydrogen donors capable of reducing carbon dioxide afterwards in the dark do not survive a period of illxnnination. There is no easily detectable formation of partiy re- duced substances. Xe\"erthelesS; there must be at least three inter- mediate steps in the process of reduction, since four hydrogen atoms have to be transferred firom water to carbon dioxide. This particular problem ^ill probably be solved \sith the aid of carbon isotopes in continuation oi tbe \*T>rk oi Ruben, Kampn^ and Hassid {cf. 14). Several iod^jendent investigations have shov*"n that nine or ten l%hi qu2^:i l:z r.rcessary for this process. The TnaYi'mum ntmiber of initial 5 erre, could be ten. For reasons of stoichiometiy it is soKible uj assun^ that there are only eight photochemical reactions, a pair for each release and recovers- of a hydrogen atom by the mole- cule transforming electronic into chemical energy (4) . This molecule may be chlorophvll itself. The difference between the theoretical 38 PHOTOSYNTHESIS value of eight and the obser\-ed one of ten we attribute to losses due to inactive chlorophyll. The simplest explanation, then, for the fact that photosynthesis proceeds either all the way from carbon dioxide to carbohydrate \%-ith a permanent gain of 120 kcal. per mole, or not at all, is that the photochemically produced hydrogen donors and acceptors disappear by back reactions whenever the process becomes artificially inhibited. Despite the most drastic effects of certain poisons on the rate or the t\-pe of the reaction, there is no permanent shift of the assimilatory quotient.* Hence, all but the final reaction products, oxygen and starch or sucrose, must react back to form water. It is very unlikely that this disappearance of intermediates should proceed by an exact reversion of the photochemical process. A chemiluminescence with a 100% yield can certainly not occur. In other words, we may expect a special kind of oxidoreduction \N-ithin the assimilatory system (7) (compare the last part of this article, pages 43 et seq.). One way of avoiding immediate back reactions hes in the use of radiant instead of thermal energ^^ A Hght quantum heats up, as it were, a single molectile Ln an otherwise cool enwonment. Once the structure of the absorbing molecule allows for a sp>ontaneous con- version of electronic excitation energy into chemical energy (that is, a change of structure to a configuration with a higher free energ\- con- tent), the new tautomer has a good chance to sur\-ive until some catal^'tic action makes use of its potential energ\\ The new tautomer will live longer the higher the wall of activation energ\^ between the original and the photochemically changed structure. In other words, its lifetime depends on how much of the available light energ\' (in the case of chlorophyll, this is always ca. 40,000 calories and is independent of the size of the lieht quanttma absorbed) is expended in forming the new structure. Actually, the ver\' first photochemical product aj>- pears to be short-lived. It probably contains still much of the original 40,000 calories. J. Franck assimies, further, that a reduced inter- mediate hydrogen donor comparable to, let us say, the reduced p>Ti- dine nucleotides is not formed. Rather, the carbox\i group becomes reduced direcdy as a consequence of the photochemical reaction taking place in the chlorophyll-protein complex to which the essential car- * Assimilatory quotient = (oxxgen liberated") /(carbon dioxide consumed) or (hydrogen absorbed) /(carbon dio.xide consumed). 39 H. GAFFRON boxyl group seems to be attached, Franck considers the influence which the presence of carbon dioxide exerts on the intensity of chloro- phyll fluorescence in vivo as the most cogent evidence supporting this view. The scheme shown on page 35 does not do justice to these important considerations. In case, however, the fluorescence experiments could be ex- plained in another way, there would be no objection against assuming the existence of an intermediate hydrogen donor. Perhaps catalyst B of Franck and Herzfeld might play this part (4). Its concentration should be about two thousand times smaller than that of chlorophyll (the evidence can be found in the discussion about the photosynthetic unit). Summing up, we may state that the heart of the photosynthesis problem is the effective utilization of energy which has to be accepted in a few big lumps. Franck and Herzfeld (4) calculate an immediate and permanent gain of 21,000 calories for each quantum absorbed. Does comparison with chemosynthesis point to an orthodox solution of this problem? Possible Role of Energy- Rich Phosphale Recent research on the utilization of metabolic energy in living cells had led to the discovery that this energy is handled in parcels not surpassing 12,000 calories packed away in so-called energy-rich phosphate bonds. Such phosphate bonds occur, for instance, in substances like Lipmann's acetyl phosphate (11) which, for the purposes of our discussion, may be regarded as stable for an indefinite period. Recently, Emerson, Stauff"er, and Umbreit (2) suggested "that for each quantum of light absorbed one 'energy-rich' phosphate bond (of ca. 10,000 cal./mole) is formed." There is no experiment or observation new or old which requires such an assumption. The only merit of this proposal, therefore, would lie in the complete analogy of photosynthesis with synthesis by thermal reactions. Part of the energy contained in the unstable first product formed in the photochemical reaction, the tautomer mentioned above (4), would be stabilized by conversion into a phosphate bond. The loss involved would be two- thirds of the chlorophyll excitation energy. There is general agree- ment among those who have considered carefully the theoretical 40 PHOTOSYNTHESIS energy requirements of photosynthesis, that a minimum of 160,000 calories is needed per mole of carbon dioxide. Ten to twelve light quanta (let alone eight) transformed into phosphate bonds at 12,000 calories each would mean that the latter would have to be used with 120% efficiency. Actually, the efficiency with which the energy-rich phosphate bond can be used for synthetic reactions appears to be around 60%. If we postulate two phosphate bonds per quantum (that is, twenty bonds altogether), tliere would be enough energy. However, a complicated mechanism must be provided to divide the energy of the excited molecule between the two phosphate bonds. Finally, the resulting phosphorylated compounds should be stable enough to survive the end of an illumination period and cause the reduction of carbon dioxide for some time afterward in the dark. As said above, nothing of that kind has ever been observed, though many an investigator has looked for it. Hence, it is simpler to assume that Nature makes use of the particular advantages inherent in photo- chemical reactions and produces intermediate hydrogen donors which are capable of a one-step gain in free energy larger than those possible by way of energy-rich phosphate bonds. A question quite different from that discussed is whether carbon dioxide becomes reduced to carbohydrate in the form of a phos- phorylated compound. It is known that plant constituents combining with carbon dioxide in the dark are half saturated at a carbon dioxide partial pressure of less than 0.1 mm. mercury. The nature of the very first fixation of carbon dioxide preceding the reaction with excited chlorophyll is still under investigation. Suggestions like "reversal of a decarboxylation" are no more helpful than the dictum that the liberation of oxygen is "the reversal of respiration." What is needed is a working model fulfilling the energy requirements. A carboxyla- tion reaction as effective as that taking place in the first step of photo- synthesis needs the coupling with a reaction releasing some free energy {ca. 10,000 cal.). Ruben (15) suggests that the carboxylation occurs with the aid of an energy-rich phosphate bond present in the molecule which takes up carbon dioxide. Lipmann (11) recently has shown that, in the presence of certain enzyme preparations, acetyl phosphate, carbon dioxide, and molecular hydrogen can condense to pyruvate. By proposing that the pyruvate is removed quickly through further reduction and a new energy-rich phosphate furnished by an oxida- 41 H. GAFFRON tive side reaction, he has provided the first plausible model for a con- tinuous chenioreduction of carbon dioxide. As an experimental approach to tie up the phosphorylation reaction and photosynthesis, it is obviously insufficient to demonstrate that phosphorylations take part in the metabolism of plant cells and that the photosynthetic production of carbohydrates changes the dark equilibrium of various phosphate esters. It w^ould be strange if it were not so. In a respiring or fermenting cell, the model for the chemi- cal reduction of carbon dioxide as presented by Lipmann should work with any available metabolic hydrogen, particularly with free hydrogen activated by hydrogenase. Since, under the latter circumstances, the over-all energy requirements are very small, there is no reason why fermenting algae should not exhibit a clear-cut reduction of car- bon dioxide in the dark. This is not the case, at least not at a noticeable rate. Apparently some extra energy for activation is necessary. If so, we have still to explain why, as found by Rieke in unpublished experiments, the number of quanta for the reduction of one carbon dioxide molecule remains ten when the algae begin to consume hydrogen instead of liberating oxygen. Energetically, one or two quanta would now suffice. The simplest explanation would involve the assumption of the existence of a rigid coupling of the re- duction of carbon dioxide with the utilization of water as hydrogen donor. This would mean the exclusion of energetically "cheaper" hydrogen from a direct participation in the photochemical reaction during anaerobic photoreductions in algae and, by analogy, also in the photosynthetic purple bacteria. We cannot present in detail the points for and against this explanation. It is also not necessary, for in van Niel's comprehensive reviews (20) and Rabinowitch's book (14) the reader will find much stimulating discussion. Interrelation of Carbon Assimilation with Oxidation Reactions in Plants and Purple Bacteria We have seen above that there is a possibility that substrates, intermediates, and final products in photosynthesis may undergo re- duction and polymerization as phosphorylated compounds. Experi- ments supporting Ruben's or Lipmann's hypotheses (11,15) are still missing, and Ochoa (12) has just reported the formation of oxalo- 42 PHOTOSYNTHESIS succinic acid from a-ketoglutaric acid and carbon dioxide. This type of fixation reaction does not depend on energy-rich acyl phos- phate. Our knowledge of phosphorylation and the "energy-rich" phosphate bond is based entirely on the carbohydrate metabolism in- volved in respiration and fermentation. We know that both types of catabolic reactions make use of the same phosphorylated compounds. If phosphorylated substances participate in photosynthesis, are these intermediates related to, or identical with, those occurring in general cell metabolism? Some special reaction must provide for the energy- rich phosphate bond assumed to promote the initial fixation of carbon dioxide in the dark. Is this special reaction part of the normal re- spiratory system or something different? In this regard Ruben (15) mentioned as a possible source of energy the dismutation to oxygen or reduction to water of the intermediate hydrogen acceptors (hydrox- ylated substances). The reduction of carbon dioxide by a purely thermal oxidoreduction in green algae occurs under circumstances which exclude normal respiration (7). Here the energy-rich phosphate bond would present a welcome means of explaining not only the fixation or carboxylation but also the coupling between the oxidation of hydrogen and the reduction of carbon dioxide. Experimentally, one should try therefore to establish the existence of an independent phosphorylation cycle serving exclusively the oxyhydrogen or "Knall- gas" reaction. While the oxyhydrogen reaction may be considered to be a sort of respiration and consequently invites comparison with the normal cell respiration in air, conditions become more unfamiliar when we turn to photoreduction. We mentioned earlier in this article why special back reactions in photosynthesis must be assumed. The question arises whether such (hypothetical) back reactions contribute to the formation of energy-rich phosphate bonds. The autonomy of the mechanism of photochemical reduction becomes quite apparent in the following observation (8) . A concentration of 0.001 M phthiocol [also of methylnaphthoquinone (vitamin K) or of o-phenanthroline] inhibits strongly the respiration of algae like Scenedesmus, prevents completely normal aerobic photosynthesis, hinders the adaptation to and the reversion from photoreduction under anaerobic conditions, and severs the coupling between the oxyhydrogen reaction and the reduction of carbon dioxide in the dark. But if the inhibitor is added 43 H. GAFFRON to the plants after their adaptation to photoreduction it does not interfere with a reduction of carbon dioxide. The photochemical reaction proceeds with a normal quotient of two. Two volumes of hydrogen are absorbed together with one volume of carbon dioxide. The only change of importance concerns the quantum yield. It is exactly one-half of that found with the unpoisoned algae. Figure 3 shows how, with increasing concentrations of phthiocol at 560 lux, . 30 25 20 - 1 1 -1 1 r \ s.'"^" :^:: ^ 560 LUX A 400 LUX " 1 1 1 1 1 c LlI 2 o 15 :d o LU a: g 10 O X Q. o 5 Ixl < 2 4 6 8 10 12 CONCENTRATION OF PHTHIOCOL X 10"^ M Fig. 3. — The inhibition and stabilization of photo- reduction in Scenedesmus by increasing concentrations of phthiocol: • — •, rates at 560 lux; A — A, same a day later; X — X, rates at 400 lux; O — O, same a day later. the rate of the reaction falls from 26 to 13 mm. per 10 min. (or at 400 lux from 18 to 9 mm. per 10 min.) and then stays constant. What this means is quite obscure. A possible solution of the riddle may be found in the following considerations. The probable occurrence was mentioned above of back reactions between the reduced and oxidized products whenever photosynthesis is artificially inhibited. Now we postulate that, in the poisoned algae, all intermediates react back, and thereby activate a reduction of carbon dioxide with hydrogen in a manner similar to that brought about by the "Knallgas" reaction in the dark. Apart from any specific explanation we can state that, if 44 PHOTOSYNTHESIS phosphorylated compounds participate at all in the reduction of car- bon dioxide, they must be either drawn from a reservoir filled during preceding anaerobic periods or produced by a cycle belonging to the assimilatory system. No indication in favor of the first alternative has been obserxed. Special experiments to test it, however, have yet to be performed. Another approach to the question of whether phosphorylations occur within the assimilatory mechanism would be by way of investigat- ing further the metabolism of the photosynthetic purple bacteria (6,19,20). Studies on purple bacteria have proved extremely useful in the past for elucidating the similarities between photoreductions and the photosynthesis of green plants {cj. references 19 and 20). We now arrive at the point at which it would be of interest to analyze in detail the differences in their mechanisms. An important difference between the green plants and the purple bacteria from the point of view of the possible role of phos- phorylated compounds is the fact that the metabolism of the plant centers around carbohydrates, whereas most of the purple bacteria decline to utilize them in any way either in light or dark. They do not respire and they do not ferment glucose and consequently cannot grow in glucose media. The metabolism of purple bacteria revoh'es, if we consider only organic substrates, mainly around aliphatic acids and, in a few cases, simple alcohols. Acetate, propionate, butyrate, croto- nate, malate, etc., are favorite substrates. The elementary analysis of an entire green plant yields, in general, data indicating the pre- dominance of compounds of carbohydrate nature. The available elementary analyses of some purple bacteria have yielded figures indicating the presence of more hydrogen and of less oxygen than is found in carbohydrates and a composition very close to that of an extracted substance having the formula (C4H602)„. The latter proved to be a polymer of crotonic acid. The green plants store carbohydrates in more or less pure form whenever photosynthesis has lasted for a while in strong light, for respiration proceeds at a much slower pace. Some purple bacteria {Athiorhodaceae) do not seem to accumulate photoproducts in excess of what can be used for immediate •synthesis of integrated cell material, that is to say, for growth of more bacteria. Others {Thiorhodaceae) accumulate unknown photosynthetic l^roducts that break down anaerobically in the dark by a sort of back 45 H. GAFFRON reaction which yields again hydrogen donors capable of being utilized in the light (20). A second difference between plants and purple bacteria lies in the relation of cell multiplication to the assimilation of carbon. Green plants can, as a rule, grow normally in air on a heterotrophic diet without photosynthesis. Spoehr (17) succeeded in obtaining growth of hereditary chlorophyll-free albino corn plants by feeding them with only one organic compound, sucrose. On the other hand, we have not observed true growth of algae, during a photochemical or thermal reduction of carbon dioxide, under anaerobic conditions, that is, in the absence of respiration. By contrast, some strains of purple bacteria multiply exclusively under anaerobic conditions and only during periods of illumination, and never in the dark. They are not capable of linking a synthetic process to either oxidative reactions (which do occur in the presence of oxygen) or fermentations (which seem not to occur at all). They depend for synthesis and growth upon photoreduction with special hydrogen donors like fatty acids, hydrogen sulfide, or molecular hydrogen. A few species among the nonsulfur purple bacteria {Athiorhodaceae) appear to grow in ways similar to that of the green plants in that they do not depend upon the photochemical reaction alone. As van Niel has shown, they can also grow aerobically in the dark by oxidizing the same substrates which they use as hydrogen donors in the light. But the fact that no carbohydrates are attacked points to a deviation from the metabolism of the green plants. The third interesting difference between plants and purple bacteria concerns a direct interrelation between the respiratory and the assimilatory systems. In the ordinary green plant, respiration and photosynthesis can go on simultaneously. Metabolic measure- ments appear most consistent if we assume that both reactions run independently of one another and that any conspicuous increase of the rate of respiration in the light is caused in an indirect way. Photo- synthesis appears to provide only reserve material, while the synthetic reactions leading to cell multiplication are coupled exclusively with respiration (and perhaps with fermentation). In those purple bacteria capable of growing at the expense of oxidation reactions, the utiliza- tion of oxygen must compete with the utilization of carbon dioxide plus light for the same hydrogen donors. The reactions do not occur 46 PHOTOSYNTHESIS independently; they supplement or exclude each other. In a very interesting quantitative experiment, van Niel (20) found that the bacteria simply cease to take up oxygen when exposed to a sufficiently intense radiation. Here w^e have to search for a direct interrelation, an intermediate metabolic link. We may sum up as follows. Whether phosphorylated com- pounds participate in photosynthesis must be considered in the light of two sets of observations. First, the assimilatory mechanism in green plants and in purple bacteria must be "self-supporting" as far as phosphorylations are concerned, since it functions under conditions in which neither respiration nor fermentation of carbohydrates seem capable of providing enough ready-made phosphorylated compounds. Second, in purple bacteria a carbohydrate appears perhaps as the primary product of the photochemical reaction, but instead of being stored it is converted into cell material of different elementary com- position. These organisms are incapable of oxidizing or fermenting ordinary plant carbohydrates. From the evolutionary point of view it is interesting that the "respiration" of the purple bacteria corresponds to the oxyhydrogen reaction and related oxidations in anaerobically adapted algae and not to the "normal" respiration in plants. We may speculate that the liberation of oxygen from "hydroxylated" compounds could be com- bined effectively with the reduction of carbon dioxide only after the synthesis and the utilization of sugars had become separated. Under aerobic conditions, the back reactions with free oxygen in the assimila- tory system had to be prevented. We know that in adaptable algae this is brought about by the oxidative inactivation of one or two of the catalysts involved. Apparently we have here a parallel to the well- known case in which anaerobic fermentations are prevented from con- tinuing in air by a special oxidation, the so-called "Pasteur reaction." References (1) Brown, N. C, Forest Products. 2nd ed., Wiley, New York, 1927. (2) Emerson, R. L., Stauffer, T. F., and Umbreit, W. W., "Relation- ship between phosphorylation and photosynthesis in Chlorella" Am. J. Botany, 31, 107 (1944). (3) Forestry Depletion in Outline. Northwest Regional Council, Portland, 1940 (25^). 47 H. GAFFRON (4) Franck, J., and Herzfcld, K. F., "Contribution to a theory of photosynthesis," J. Phys. Chern., 45, 978 (1941). (5) French, C. S., and Rabideau, G. S., "The quantum yield of oxygen production by chloroplasts suspended in solutions containing ferric oxalate," J. Gen. Physiol., 28, 329 (1945). (6) GafTron, H., "tJber den Stoffwechsel der Purpurbakterien," Part I, Biochem. Z-, 260, 1 (1933); Part II, ibid., 275, 351 (1935). (7) Gaffron, H., "Photosynthesis, photoreduction and dark reduction of carbon dioxide in certain algae," Biol. Rev., 19, 1-20 (1944). (8) Gaffron, H., "o-Phenanthroline and derivatives of vitamin K as stabilizers of photoreduction in Scenedesmus," J. Gen. Physiol., 28, 259 (1945). (9) Hill, R., and Scarisbrick, R., Proc. Roy. Sac. London, B129, supple- ment, 39 (1940). (10) Krebs, H. A., "Carbon dioxide assimilation in heterotrophic or- ganisms," Ann. Rev. Biochem., 12, 529 (1943). (11) Lipmann, F., J. Biol. Chem., 158, 515 (1945). (12) Ochoa, S., "Isocitric dehydrogenase and carbon dioxide fixation," J. Biol. Chem., 159, 243 (1945). (13) Planning for a Permanent Agriculture. U. S. Dept. Agr., Misc. Pub. No. 351 (1939). (14) Rabinowitch, E., Photosynthesis, Vol. I. Interscience, New York, 1945. (15) Ruben, S., "Photosynthesis and phosphorylation," J. Am. Chem. Soc, 65, 279 (1943). (16) Sinnott, E. W., "Plants and the material basis of civilization," Am. Naturalist, 79, 28 (1945). (17) Spoehr, H. A., "The culture of albino maize," Plant Physiol., 17, 397 (1942). (18) United Nations Conference on Food and Agriculture, Report, Hot Springs, May, 1943. (19) van Niel, C. B., "On the morphology and physiology of the purple and green sulphur bacteria," Arch. MikrobioL, 3, 1 (1931). (20) van Niel, C. B., "The bacterial photosyntheses and their importance for the general problems of photosynthesis," in Advances in Enzymology, Vol. I. Interscience, New York, 1941, p. 263. "The culture, general physiology, morphology, and classification of the non-sulfur purple and brown bacteria," Bad. Revs., 8, 1 (1944). (21) Warburg, O., and Liittgens, W., "Weitere Experimente zur Kohlen- s'sLureassimilation," Naturwissenschaften, 32, 301 (1945). (22) Went, F. W., et al., "Plant growth under controlled conditions," several articles in Am. J. Botany, 31-32, 1944-1945. (23) Werkman, C. H., and Wood, H. G., "Heterotrophic assimilation of carbon dioxide," in Advances in Enzymology, Vol. 11, 1942, p. 135. 48 THE BACTERIAL CELL RENfi J. DUBOS, MEMBER OF THE ROCKEFELLER INSTITUTE FOR MEDICAL RESEARCH, NEW YORK; MEMBER OF THE NATIONAL ACADEMY OF SCIENCES; JOHN PHILLIPS MEMORIAL AWARD; MEAD JOHNSON AWARD Only such substances can be anchored at any particular part of the organism which fit into the molecule of the recipient combination as a piece of mosaic fits into a certain pattern. PAUL EHRLICH B BACTERIA appeared to the nineteenth century biologist as a type of protoplasmic material devoid of any organi- zation, almost as a link between the animate and the inanimate world. "They are," said Ferdinand Cohn, "the simplest and lowest of all living forms — beyond them, life does not exist." The compound microscope failed to reveal any structure within their cellular bound- aries, and the biochemist was inclined to consider the bacterial cell as a mere bag of enzymes which owed its enormous biochemical activity to its colloidal dimensions. The primitiveness of bacterial life appeared to be confirmed in chemical terms when Winogradsky demonstrated in 1887 that certain autotrophic species can grow in purely inorganic media and can synthesize their protoplasm from mineral salts and carbon dioxide, utilizing for the reduction of the latter the energy released by the oxidation of sulfur, iron, ammonia, nitrite, etc. (21). Was it not permissible to consider this production of organic matter from inorganic elements as the most primitive bio- chemical expression of life, as the beginning of life on earth? Advances on the diverse fronts of bacteriology were quick to dispel these early illusions concerning the biochemical primitiveness and the simplicity of organization of the bacterial cell. Analysis of the chemical activities of bacteria soon revealed that microbial life takes place through the agency of the same type of reactions, the same 49 R. J. DUBOS metabolic channels and products, and the same biocatalysts which constitute the mechanism of life in the highest and most evolved organisms. For example, the oxidation of sulfur by the autoti'ophic bacterium Thiobacillus thiooxidans depends upon an intimate linking between oxidation and phosphate turnover; the oxidative phase is accompanied by phosphate fixation and the reductive phase of carbon dioxide fixation is accompanied by a release of phosphate (22). More- over, Thiobacillus thiooxidans is fully equipped with the regular comple- ment of water-soluble vitamins found in other living organisms: thia- min, riboflavin, nicotinic acid, pantothenic acid, pyridoxine, biotin, etc. (17). In other words, the mechanisms of energy transfer and of intermediary metabolism are essentially as complex in the least exact- ing bacteria as they are in the most fastidious organism. The chemistry of Thiobacillus thiooxidans is not unique among bacteria; most of the known water-soluble vitamins — with the possible exception of ascorbic acid — have now been found to be either produced by, or required for the growth of, all the microbial species so far studied. Fat-soluble vitamins also probably play a part in microbial metabolism since at least one of them, vitamin K, is an essential growth factor for Johne's bacillus {Mycobacterium paratuberculosis) and the other myco- bacteria produce biologically active naphthoquinones during growth (24). Thus, bacteria utilize the multiple and complex biocatalysts which govern and integrate the metabolism of all living cells. More- over, many bacteria, the autotrophic species for example, possess in addition the ability to synthesize these same biocatalysts from inorganic elements in the course of their growth in synthetic media, a property which most plant and animal cells never possessed or have lost entirely. The high degree of cellular organization required for the performance and for the integration of these complex syntheses need not be em- phasized; at the biochemical level, at least, there is no ground to consider that bacteria represent primitive forms of life. The growth requirements of autotrophic bacteria are extremely simple indeed, but how complex their vital machinery, their performance, and their products ! If they are truly the first representatives of life on earth, they sprang, like Minerva, fully armed from the forehead of Jove. Simultaneously with the realization that the biochemical processes of bacteria are no simpler than those of other organisms cane 50 THE BACTERIAL CELL the recognition of the existence in the bacterial cell of a number of structures which, although often ill defined in nature and function, obviously express a morphological complexity parallel to the bio- chemical complexity of all known forms of life. Utilization of the classical methods of cytology soon revealed, for example, the existence in bacteria of flagella, spores, different kinds of membranes and capsules etc., which give to each bacterial type a fairly characteristic morpho- logical individuality (12). Little by little, bacteriological staining techniques are gaining the dignity of cytochemical reactions and give chemical definition to the cellular objects which they reveal; the more skillful utilization of Feulgen's reagent for instance, permits the identi- fication among the other basophilic constituents of the bacterial cell of discrete bodies rich in desoxyribonucleic acid which are almost cer- tainly the equivalent of the vesicular nucleus in larger cells (12,19). Photography in the ultraviolet and electron microscopy have per- mitted the optical resolution of cellular structures — intracellular granules, membranes, individual components of flagella, etc. — which are below the limit of resolution by ordinary microscopy. These classical cytological techniques aim at the direct visualization of the constituents of the cell. On the other hand, the analysis of the re- sponse of the cell to the effect of certain reagents and procedures pro- vides an indirect approach to cytological problems, by suggesting the existence and often the chemical nature of important cellular com- ponents which cannot be seen by any of the known methods of micros- copy. Interestingly enough, it is the study of pathogenic bacteria which has been the most fruitful from the point of view of this indirect approach to cytology. In order to analyze the host-parasite relationship, the student of infection must concern himself with those structures and products of bacteria — the cellular antigens and toxins — which affect the course of the infectious process and against which are directed the reactions of immunity. Similarly, the attempts to understand the mode of action of antiseptics on bacteria led to the study of the structures through which antiseptic agent and susceptible cell come in contact. Thus, many constituents of the bacterial cell have been recognized first by biological reactions; and the analysis of the phenomena of infection, immunity, and chemotherapy has provided important in- formation concerning the biochemical architecture of bacteria. Paul 51 R. J. DUBOS Ehrlich first stated clearly the possibility of describing these reactions in terms of cellular structure. He postulated that the living cell possesses a number of chemically reactive groups which he called "receptors" and with which dyes, bactericidal substances, and immune antibodies react selectively. Ehrlich regarded these "receptors" as definite chemical entities capable of entering into union with dyes, antiseptics, and antibodies. According to his theory, characteristic staining reactions, diflferential susceptibilities to toxic substances, and specific reactions with corresponding antibodies could all be explained by assuming the existence of a sufficient number of receptors in the bacterial cell. These phenomena can consequendy serve as tests to facilitate the recognition of the receptors and their isolation in pure form. During the past decades, immunochemists and students of the theory of chemotherapy have gone far toward identifying, and in several cases separating in a purified state, several of the cellular components with which antibodies and antibacterial agents react selectively. The specific chemical relationships involved in the chemo- therapeutic reactions are discussed in other essays of this volume and need not be considered here. There are certain aspects of the problem, however, which are too ill-defined to warrant discussion in terms of a chemical theory, but which deserve mention at this time since they bid fair to help in the elucidation of some interesting details of cellular structure. Empirical staining reactions led very early to the division of the bacterial world into three broad groups, the Gram-positive, the Gram-negative, and the acid-fast, which are defined by their behavior toward two staining techniques (the Gram and Ziehl-Neelsen methods) . These three bacterial groups not only differ in their staining properties, but also exhibit striking differential susceptibilities to the different types of antiseptics and antibacterial agents. By way of illustration, most Gram-negative bacilli grow readily in the presence of basic dyes, penicillin, or gramicidin, whereas these organisms are extremely susceptible to the bactericidal and lytic effect of immune serum. On the contrary, many Gram-positive species are completely resistant to lysis by immune serum, but are extremely susceptible to the bacterio- static and bactericidal effect of small concentrations of dyes, penicillin, and gramicidin. As for the tubercle bacilli (acid-fast), they are re- markably resistant to all the classical antisejitics and to a gi'eat variety 52 THE BACTERIAL CELL of other toxic agents, retaining their viabihty, for example, after ex- posure to 5% sodium hydroxide or sulfuric acid. In order to account for these strikingly different susceptibilities, many theories have been proposed which assume that the bacterial groups vary with reference to cellular permeability, presence of certain lipids in the cell mem- branes, acid-base properties of the cell body, etc. Thus, observations of electrokinetic behavior and of affinity for dyes at different pH levels suggest that the cell material in the colon-dysentery-typhoid group (Gram-negative) is less acidic than in the Gram-positive bacterial species (pneumococci, staphylococci, streptococci, anthrax bacilli, etc.) (1,20). The acid-fast bacilli (e. g., tubercle bacilli) produce astonishingly high concentrations of a variety of lipids which gives them marked hydrophobic properties (2,23). All the Gram-negative species readily yield in solution phospholipid-protein-polysaccharide complexes which constitute 5-10% of the total cell weight; similar complexes have not been obtained from the Gram-positive organisms (16,18). Recent observations have established a correlation between ability to retain the Gram stain (correlated with greater susceptibility to many antiseptics) and the presence around the cell of a magnesium ribonucleate complex (5,10). All these facts provide examples of the type of chemical information which results from the analysis of the biological behavior of the cell and which will undoubtedly reveal important differences of structure between the different bacterial groups. The specific antibodies produced as the result of the injection of bacterial antigens into the animal body have provided another set of reagents that have yielded important information of a cytochemical nature. The immune reaction to any one type of bacterial cell is not a simple phenomenon, since bacteria are made up of a multiplicity of chemical constituents many of which elicit the production of specific antibodies. In other words, the injection of one type of bacterial cell usually results in the production of several antibodies, each one of which is directed against one particular cellular component. These different cellular constituents obviously bear a definite spatial morphological relationship to each other in the intact cell. Some are masked by membranes and become exposed only as a result of cellular disintegra- tion; others are peripherally disposed and in direct contact with the environment. This stratification of cellular structures affects the im- 53 R. J. DUBOS mune response of the animal host to the whole bacterium and is re- flected in the type and amount of antibody produced. It also condi- tions the reaction of the bacterial cell with a given antibody, since the intact microorganism unites much more readily, if not solely, with the antibodies which are specifically directed against those of its con- stituents exposed at or near the surface. Analysis of antibody pro- duction and of antigen-antibody reaction can therefore help in formu- lating an approximate picture of the arrangement of the different antigens in the architecture of the cell. To summarize, the use of immunological procedures for the study of cellular structure involves a number of successive steps: (a) the preparation and separation of antibodies specific for each one of the chemical constituents of the cell; (b) the utilization of these antibodies as specific reagents for the detection and preparation in pure form of the cellular constituents; and finally, (c) the interpreta- tion of antigen-antibody reaction in an attempt to define the relative positions occupied by these chemical constituents in the living cell (7,9,11,14). The results obtained by immunochemical analysis have led to the recognition that the different bacterial groups (pneumococci, streptococci, sporulating aerobic bacilli, organisms of the colon- typhoid-dysentery group, etc.) are characterized by a general pattern of antigenic organization which is common to the different members of each group. On the other hand, the various species and immuno- chemical types within any general group differ from each other by virtue of the chemical specificity of the different components of their antigenic mosaic. Thus, all virulent pneumococci possess a capsule which is polysaccharide in nature, but the polysaccharide varies chemically and antigenically from type to type (3) . The virulent forms of human streptococci (group A) can also produce a capsule made up of hyaluronic acid; moreover, they all possess as surface constituents peculiar proteins (the M substances), and other substances of unknown chemical nature (the T antigens), which vary in immunochemical specificity from type to type (13). Several of the aerobic sporulating bacilli have been found to produce a capsule consisting essentially, if not solely, of polypeptides; the polypeptide, in the case of the anthrax bacillus, appears to be made up exclusively of ^/-glutamic acid (6). The virulent coliform bacilli (typhoid, dysentery, etc.) all produce the 54 THE BACTERIAL CELL lipid-protein-polysaccharide complexes mentioned above. The poly- saccharide components determine the immunochemical specificity of the different species. The protein component, on the contrary, ap- pears to be essentially the same in all the members of the group; in fact, it can be made to combine with the different specific polysac- charides to reconstitute complexes similar to those which normally occur in the cell (16). In general, the immunochemist has concerned himself primarily with the constituents which are present at the cell surface becaase these elements appear to play the most important part in the phe- nomena of immunity. However, other proteins, polysaccharides, etc., which certainly occupy less superficial positions in the cell have also been recognized by immunochemical analysis. A striking illustration of the potentialities of immunochemical methods in cytology is given by the case of the typhoid bacillus. Specific antibodies have been prepared for the following cellular components of this organism: tlie flagella; the O and Vi antigens of the cell surface; the R, p, and Q antigens, which are intracellular components. Living typhoid bacilli resuspended in solutions of these different antibodies exhibit a charac- teristic behavior which is determined by the relative position of the corresponding antigen in the cellular architecture. Loss of motility, different patterns of agglutination, bacteriolysis, etc., are phenomena which are characteristic for each antigen-antibody reaction and which can be interpreted in terms of cellular organization of the different antigens. Dyes, antiseptics, and antibodies are not the only reagents which can be used to recognize and identify the cellular receptors. If it be found, for example, that a given enzyme attacks the cells of a certain microbial species causing death or the alteration of a character- istic cellular property, it can be surmised that the chemical substrate which is susceptible to the enzyme is present in the cell under con- sideration and that it plays some part in the function altered by the enzyme. Thus, the fact that lysozyme causes the death and lysis of the cells of several bacterial species indicates that the mucopoly- saccharides hydrolyzed by this enzyme are essential components of the cellular structure of the susceptible species (15). It has been shown also that other polysaccharidases decompose the capsular polysac- charides of pneumococci and of streptococci, and that proteolytic 55 R. J. DUBOS enzymes inactivate the specific M protein antigens of group A strepto- cocci (8,13). Contrary to what is observed with lysozyme, however, the polysaccharidases and proteases do not aflfect in any way the viabiHty of the treated cells, even when they rid it entirely by hydrolysis of the specific polysaccharides or proteins. It is likely, therefore, that, instead of being considered as structural constituents of the bacterial bodies, the capsular polysaccharides and the M proteins should be regarded as excretion products which accumulate around the cell, since they can be removed or destroyed without interfering with the essential living processes. There are many other instances of biological reactions which, because of the specific relationships they bear to certain cellular com- ponents of bacteria, can be used as indirect methods for the analysis of cellular structures. Let us mention, for example, the remarkable selectivity of pure lines of bacteriophage with reference to the strains of bacteria which they can attack. The relationship between speci- ficity and cellular structure is illustrated by the fact that the bacterio- phage can be absorbed specifically by the cells, living or dead, of the susceptible bacterial cultures. It has also been found that, in certain cases, soluble fractions extracted from the susceptible bacterial cells inhibit specifically the lysis of the homologous organisms by the bac- teriophage. It is likely, therefore, that the phenomena of bacterio- phage lysis will also yield a number of new specific reactions which by revealing the existence and nature of new types of receptors could serve in the analysis of cellular structure. The biological phenomena which we have considered have in common the following characteristics permitting their utilization as indirect cytological methods. They are all the result of a reaction between a given reagent (antiseptic, enzyme, antibody, bacterio- phage) and a specific cellular receptor, this reaction manifesting itself by inhibition of growth, enzymic destruction of cellular component, agglutination or lysis by antibody, or lysis by bacteriophage. In many cases the reagent can be absorbed on the homologous cell sub- strate, and the reaction can be inhibited by the addition to the system of the specific substrate which constitutes the cellular receptor. In- hibition of growth, enzymic decomposition, agglutination, lysis, etc,, are only the secondary manifestations of primary reactions which de- pend upon the union between the cellular receptors, on the one hand, 56 THE BACTERIAL C:i;i,I, and the biological reagents, be they antiseptics, antibodies, enzymes, bacteriophages, on the other. One of the most intriguing applications of the indirect cyto- logical methods discussed in the preceding pages has been the analysis of the phenomena of bacteria variability. It has long been known that most bacterial cultures — even those arising from single cells — often undergo profound transmissible modifications of their morpho- logical, biochemical, and physiological properties. Immunochemical studies have revealed, in particular, a type of variation, now recognized in practically all bacterial species, which involves the loss of the specific surface components of the cell (the capsular polysaccharides of pncu- mococci, the M proteins of streptococci, the capsular polypeptide of anthrax, the lipid-protein-polysaccharide complexes of the dysentery and typhoid bacilli, etc.). These transmissible modifications of the surface of the bacterial cell have attracted particular attention because they are in many cases correlated with alteration of the virulence of the organism concerned. They constitute, however, only a very nar- row aspect of the total problem of bacterial variability. One can observe within one given bacterial culture transmissible modifications of many unrelated properties: ability to attack sugars or proteins, to synthesize amino acids or pigments, to resist antiseptics or other injurious procedures, to produce flagella, spores, capsules, and so on. All these variations occur independently of each other, thus giving to each bacterium the possibility of manifesting its existence under a great diversity of forms and properties. The production of these large numbers of variant forms deficient in one or another of the cellular components has greatly helped in the analysis of many immuno- chemical problems. From a more general point of view, it is of the greatest interest that a given organism can successfully continue to exist and to multiply as an independent living object after having lost a great variety of structures and functions which had appeared to constitute important components and attributes of the "normal" parent form. As already stated, these structures and functions can be lost and regained inde- pendently of each other, without altering the essential nature of the germ or the potentialities of the cell. Even more striking is the fact that it is possible to substitute experimentally one character for an- other. Thus, by adding to a strain of pneumococcus which has lost 57 R. J. DUBOS the ability to produce its specific capsular polysaccharide an extract of the cell of another type of encapsulated pneumococcus, one can convert the former organism into the type from which the extract was made. From then on, the cell can produce, and transfer to its progeny the ability to produce, a polysaccharide different from the one it had been known to synthesize heretofore. All available evidence indicates that the substance which is capable of inducing the transformation is a form of desoxyribonucleic acid- — specific for each pneumococcus type (4). Of equal interest is the unavoidable conclusion that the bacterial cell is not only an integrated complex of independent charac- ters, but that it is possible to substitute for one of these characters an- other one, homologous but different, without interfering essentially with cellular organization. Thus, a large body of knowledge concerning cytology is slowly emerging from the study of bacterial variability and of the behavior of the cell in the presence of a number of biological reagents. It is be- coming possible to recognize and to define in chemical terms a number of structures not yet detectable by any microscopic technique. Further- more, by bold, even though admittedly dangerous, extrapolation, one can guess at the approximate position of these cellular constituents in the architecture of bacteria. The history of science provides, of course, many examples of the fruitfulness of indirect methods, and in particular of the utilization of chemical and biological manifestations as indices and guides for the recognition and identification of morpho- logical structures. Claude Bernard stated as early as 1855 that "anatomical localization is often revealed first through the analysis of the physiological processes." Much of the morphology indirectly revealed by antibodies, enzymes, and cytotoxic substances lies beyond the microscopic range and in fact often reaches the molecular level. It concerns the organization of those molecular groupings which, because of their chemical reactivity, condition the behavior of the cell both as an independent functioning unit and in its relation to the environment. This knowledge is of obvious interest to the cytologist. It also forms the fundamental basis upon which is being erected the theory of immunity, since all the phenomena of host-parasite relation- ship are essentially a reflection of the biochemical architecture of the cell. 58 THE BACTERIAL CELL References* (i) Albert, A., "Chemistry and physics of antiseptics in relation to mode of action ," Lancet, 1942, II, 633-636. (2) Anderson, R. J., "The chemistry of the lipids of the tubercle bacilli," Harvey Lectures, 35, 271-313 (1939-1940). (3) Avery, O. T., "The role of specific carbohydrates in pncumococcus infection and immunity," Arm. Internal Med., 6, 1-9 (1932-1933). (4) Avery, O. T., MacLeod, C. M., and McCarLy, M., "Studies on the chemical nature of the substance inducing transformation of pneumococcal types," J. Exptl. Med., 79, 137-158 (1944). (5) Bartholomew, J. W., and Umbreit, W. W., "Ribonucleic acid and the Gram stain," J. Bact., 48, 567-578 (1944). (6) Bovarnick, M., "The formation of extracellular a'(-)glulamic acid polypeptide by Bacillus subtilis,'' J. Biol. Chem., 145, 415-424 (1942). (7) Boyd, Wm. C, Fundamentals oj Immunology. Interscience, New York 1943. (8) Dubos, R. J., "Enzymatic analysis of the antigenic structure of pneumo- cocci," Ergeb. Enzymjorsch., 8, 135-148 (1939). (9) Heidelberger, M., "Immunology as a tool in biological research," Am. Naturalist, 77, 193-198 (1943). (10) Henry, H., and Stacey, M., "Histochemistry of the Gram-staining reaction fcjr micro-organisms," Nature, 151, 671 (1943). (11) Kabat, E. A., "Immunochemistry of proteins," J. Immunol., 47,513-587 (1943). (12) Knaysi, G., Elements oj Bacterial Cytology. Comstock, Ithaca, 1944. (13) Lancefield, R. C., et al., "Studies on the antigenic composition of group A hemolytic streptococci," J. Exptl. Med., 78, 465-476 (1943); 79, 79-114 (1944). "Specific relationship of cell composition to biological activity of hemolytic streptococci," Harvey Lectures, 36, 251-290 (1940-1941). (14) Landsteiner, K., TheSpecificityoJ Serological Reactions. Rev. ed.. Harvard Univ. Press, Cambridge, 1944. (15) Meyer, K., Palmer, J. W., Thomf:)son, R., and Khorazo, D., "On the mechanism of lysozyme action," J. Biol. Chem., 113, 479-486 (1936). (16) Morgan, W. T. J., and Partridge, S. M., "An examination of the O antigenic complex of Bact. typhosum,'" Brit. J. Exptl. Path., 23, 151-165 (1942). "Studies in immunochemistry. 6. The use of phenol and of * The material discussed in this essay is presented in a more complete manner and with extensive documentation in a monograph, The Bacterial Cell, by R.J. Dubos and C. Robinow, Harvard University Press, Cambridge, 1945. 59 R. J. DUBOS alkali in the degradation of antigenic material isolated from Bad. dysenteriae (Shiga)," Biochem.J., 35, 1140-1163 (1941). (17) O'Kane, D. F., "The presence of growth factors in the cells of the auto- trophic sulphur bacteria," J. Bad., 43, 7 (1942). (18) Partridge, S. M., and Morgan, W. T. J., "Immunization experiments with artificial complexes formed from substances isolated from the antigen oiBad. Shigae," Brit. J. Exptl. Path., 21, 180-195 (1940). (19) Ribonow, C. F., "A study of the nuclear apparatus of bacteria," Proc. Roy. Soc. London, 130, 299-328 (1942). "Gytological observations on Bad. colt, Proteus vulgaris and various aerobic spore-forming bacteria with spe- cial reference to the nuclear structures," J. Hyg., 43, 413-423 (1942). (20) Stearn, A. E., and Stearn, E. W., "Metathetic staining reactions with special reference to bacterial systems," Protoplasma, 12, 435-464, 580-600 (1931). (21) van Niel, C. B., "Biochemical problems of the chemo-autotrophic bacteria," Physiol. Revs., 23, 338-354 (1943). (22) Vogler, K. G., and Umbreit, W. W., "Studies on the metabolism of the autotrophic bacteria," J. Gen. Physiol., 26, 157-167 (1942). (23) Wells, H. G., and Long, E. R., The Chemistry oj Tuberculosis, 2nd ed. rev., Williams & Wilkins, Baltimore, 1932. (24) Woollcy, D. W., and McCarter, J. R., "Antihcmorrhagic compounds as growth factors for the Johne's Bacillus," Proc. Soc. Exptl. Biol. Med., 45, 357-360 (1940). 6o THE NUTRITION AND BIO CHEMISTRY OF PLANTS D. R. HOAGLAND, professor of plant nutrition, college of agriculture; plant physiologist, agricultural experiment STATION, university OF CALIFORNIA OTHER writers for this volume will discuss the biochemistry of plants in relation to photosynthesis, plant hormones, and the activities of microorganisms. The present article is, there- fore, devoted primarily to an attempt to indicate the need and the opportunities for research on the biochemistry of higher plants, es- pecially plants of agricultural interest, as a foundation for the adequate understanding of many problems of plant nutrition and of plant physiology. This field of inquiry is relatively undeveloped in the modern period in comparison with the biochemistry of higher animals and of microorganisms, with its remarkable record of achievement during the past quarter of a century. The importance of the bio- chemistry of the higher plants for the cycles of living organisms in general, and for the basic occupation of agriculture is too obvious to require analysis. The disparity of achievement in fundamental biochemical re- search dealing with higher plants, on the one hand, and with the higher forms of aniinal organisms and some groups of microorganisms on the other, becomes apparent on examination of recent monographs dealing with advances in biochemistry. They are predominantly concerned with experiments on animal tissues, or on microorganisms, and the majority of contributors are associated with medical research in- 6i D. R. HOAGLAND stitutes. Recent texts on biochemistry give but little specific attention to the biochemistry of higher plants. Certain earlier treatises devoted largely to this latter subject have not appeared in new editions for a good many years. This general appraisal on a comparative basis appears to be justified even when the noteworthy contributions of individual workers or of certain groups of workers on the biochemistry of higher plants are kept in view. It receives support from some of the comments of Vickery, who is well known as an extensive con- tributor to several phases of the biochemistry of plants. The situation as described may seem surprising when one recalls the vast programs of research carried on by agricultural experiment stations. It is true that, in these stations, a great number of studies on plants have been made which are to some degree biochemical in nature. But it is rare to find groups of investigators assigned the definite objective of developing the knowledge of the fundamental biochemistry of crop plants. Generally, biochemical studies are encouraged in so far as they throw light on particular questions of agricultural importance as related to plant nutrition, horticulture, agronomy, or perhaps general plant physiology. In terms of crop production, the success of the coordinated attacks on plant problems through the application of the agricultural sciences and arts is well demonstrated by the enlarged production of crops under the difficulties of war conditions. This, however, does not meet the point we have under discussion. Further, there is reason to assume that, even from a utilitarian point of view, a more intensive development of research on the biochemistry of crop plants would in due course contribute to the basic knowledge essential to the control of plant production and supplement and guide the interpretation of results of practical experi- mentation. Ramifications of the biochemistry of plants into the applied field are manifold. The growth of crops needs to be appraised by the criteria not only of total yield but also of quality. This latter aspect is currently receiving much attention through consideration of plant composition in relation to the value of the plant product for animal nutrition, a point illustrated by the research program of the Federal Soil, Plant, and Nutrition Laboratory at Cornell. For example, the general problem of the synthesis of vitamins by plants might be cited, as well as the studies made to gain infoi-mation on the 62 BIOCHEMISTRY OF PLANTS relative influence of climate and mineral nutrition of the plant on its vitamin content. Also, the biochemical mechanisms in the plant that result in the synthesis of essential amino acids, carbohydrates, and higher fatty acids are of interest to students of both plant and animal metabolism. Knowledge of the biochemistry of the plant is one of the sources of information that constitutes a valuable asset to the plant pathologist, the plant geneticist, the horticulturist, the specialist in forestry, and indeed to all those who must of necessity come into contact with plant systems of biochemical reactions, whether or not this is consciously recognized. From some points of view, the plant offers, as compared with the animal, methods of study with marked advantages. On the other hand, there are disadvantages in the use of the higher plant for bio- chemical investigation. The nature of a plant's growth is such that most of its living cells perform the most diversified functions, a fact which renders plant cells less suitable for research on specific bio- chemical reactions. Tissues from animal organs often provide ma- terial with specialized and highly intense activities suitable for the investigation of enzyme reactions. It is doubtful that the plant in general is as favorable as the animal for similar investigations, but this view may be held only because fewer attempts have been made to exploit the possibilities of plant material. There is no opportunity to carry on with the plant the kind of experiments that are rendered possible by the presence of a blood stream and organs of elimination. The introduction of specific organic metabolites into the plant cell and the study of their transformations within the cell involve special compli- cations. In the examination of biochemical reactions taking place in excised plant tissues, avoidance of bacterial and fungal contamination often presents great difficulties. Further, the correlating eff"ects of various plant hormones and of normal translocation of metabolites make especially difficult the interpretation of the results of studies of plant tissues in terms of the intact growing plant. Plants of the kind under consideration may be regarded as normally complete synthetic systems, building up or breaking down compounds representing an extraordinary array of organic structures of biological interest, all derived from the simple substances required for plant growth, namely, carbon dioxide, water, and inorganic ele- ments to the number of fifteen or more. It might be postulated that 63 D. R. HOAGLAND certain species of plants growing in some types of soil high in organic matter may have lost the power of synthesizing at an adequate rate vitamins or other essential organic units, and that these species depend in part on the absorption of these units from an environment in which they have been synthesized by microorganisms. But most or all species of higher green plants so far intensively studied (these are mainly plants of economic importance) can go through their cycles of growth by virtue of their own synthetic powers, at least so far as can be ascer- tained by the use of purified inorganic media, although usually without complete exclusion of all microorganisms. It follows that the range and diversity of biochemical reactions that need to be investigated is enormous. Correspondingly great are the opportunities for the study of different synthetic processes in a living organism, to the extent that adequate methods can be devised for attacking such complex systems. The preoccupation of plant physiologists engaged in agricultural research with the inorganic elements absorbed by crop plants from the soil, and the frequent designation of these elements as "plant foods," tend to subordinate appreciation of the biochemical aspects of plant nutrition. The inorganic elements derived from the soil constitute only a small percentage of the dry weight of a plant, and one of the most important objectives of research in plant nutrition should be an understanding of the mechanisms by which these inorganic com- ponents become directly incorporated into the organic compounds synthesized by the plant, or activate the enzymes which catalyze the syntheses and breakdown of organic compounds. We have as a foundation the knowledge that plants of the kind in question absorb from an inorganic medium, and have an essential need for, the elements nitrogen, phosphorus, potassium, calcium, magnesium, sulfur, and iron. Also, as a result of research in plant nutrition during the past decade or two, other elements have been shown to be equally essential though required in only minute quan- tities. These additional elements include boron, manganese, copper, and zinc, with strong but still limited evidence that molybdenum is also essential. There may be, and probably are, still other chemical elements indispensable to plant growth, but conclusive proof of the general indispensability of other elements, over a wide range of plant species, has not yet been obtained. Limitations of technique are soon 64 BIOCHF.MISTRV OF PLANTS encountered as the effort is made to go further in the exckision of im- purities from the nutrient medium. It should be noted, however, that various chemical elements not indispensable for growth of the plant may modify its biochemical reactions either beneficially or adversely, under the conditions of a natural environment. Many of these facts have been established primarily through the use of artificial culture methods, among which the so-called water- culture method is especially useful for studying the effects of a deficiency of an element needed by the plant in minute quantity. Sometimes, however, special care in the selection and purification of a solid inert medium, to which a purified nutrient solution is applied, pro\ides an alternative technique, with certain advantages. Innumerable experiments, some of them with meticulous care, have been made on many species of plants with these artificial culture techniques. The control of the inorganic nutrient medium represents only a partial control of the environment; and frequently there remains the necessity or desirability of control of the atmospheric factors to which the plant is subjected: light, temperature, humidity, carbon di- oxide concentration, and air movement. The control of these factors obviously demands costly equipment and usually is not attempted in nutritional experiments with plants. But some laboratories have had the opportunity to grow plants under conditions of controlled air tem- perature and controlled artificial illumination. In recent years, fluorescent lamps have proved especially valuable for this purpose. Most artificial culture experiments have not been designed for the primary purpose of obtaining information about the metabolic mechanisms of the plant and the specific enzyme systems concerned. Observations on plants grown in the presence of selected combinations of inorganic nutrients are likely to be confined to measurements of rate of growth of the plants, total yields of the tissue produced on a fresh or dry weight basis, or on yields of some part of the plant of special interest from an agricultural point of view. Chemical studies are often limited to the determination at the end of a selected growth period of certain chemical elements absorbed by the plant, or of well- known organic compounds formed as a net result of innumerable bio- chemical processes that have proceeded perhaps for a considerable period of time during which the plant has increased in size and difleren- tiated its tissues. In other cases, the purpose may he to record the 65 D. R. HOAGLAND pathological symptoms resulting from a marked deficiency in the medium of some one of the essential inorganic elements. Valuable as this information is for the purpose in view, it does not advance our understanding of the biochemistry of plants in a manner at all com- parable with the advances made in the study of animal tissues or of some microorganisms, in which definite steps in a series of chemical reactions are identified or reasonably deduced from experimental data. The use of the artificial culture methods of plant nutrition makes feasible the growing of plants of many species with any desired combi- nation of inorganic nutrients, or with a given nutrient available in graduated quantities. Controlled modifications in the inorganic composition of the plant are thereby induced, although generally in no simple relation to the composition of the nutrient solution. As already stated, possibilities exist for control of illumination and tem- perature and, thus, to some degree for control of carbon assimilation and rates of metabolic reactions. It is tempting to propose that these methods of controlled culture afford techniques for endless rewarding studies on biochemical mecha- nisms in the plant. To what extent this is a realistic view is difficult to say. The extraordinary complexity of the growing plant and of the conditions of its nutrition may set narrow limits to what can be done of fundamental biochemical importance, but whatever opportunities do exist have yet to be adequately explored. There are, of course, other methods of experimentation on plants which can be adapted to biochemical research, such as embryo culture, culture of root tips, and experiments with excised leaves, roots, or other parts of the plant immersed in solutions of known composition. A technique has been recently described (13) for physiological and chemical studies on albino plants, whereby the transformations of a known carbo- hydrate supplied to the plant might be followed, without being com- plicated or obscured by reactions which are associated with photo- synthesis. It may be noted again that from the standpoint of plant nutri- tion the main biochemical problem is the fate of the chemical elements derived from the nutrient medium and the way in which they interact with the products of photosynthesis. Most of the elements essential to plants are also essential to all organisms; research on their metabolic functions has therefore a wide biochemical interest. For some cf 66 BIOCHEMISTRY OF PLANTS the essential inorganic elements, and especially certain "trace" ele- ments, we have little or no guidance from studies on other organisms. Boron is indispensable for all the higher plants so far properly investi- gated. It has not been shown to be indispensable to the animal, although little work has been done on this point. Boron is one of the elements plants require in minute amounts, yet deficiencies even under some soil conditions have assumed first-rate agricultural importance, a fact which accentuates interest in the biochemical functions of boron. Research by plant physiologists indicates that deficiency of boron often results in a pathological state in the plant nearly the same as that caused by a deficiency of calcium. One view is that an inadequate supply of boron may limit the maintenance of an effective level of calcium in a soluble or active form within the tissues of plants. Some workers think that formation of pectin compounds does not proceed normally when boron is deficient. Clearly these are questions which need the atten- tion of skilled biochemists. Possibly productive leads might come from comparative biochemistry. Certain groups of fungi, and perhaps some algae, appear not to require boron. The same fungi also can grow without calcium, or at least the amounts needed are too small to be removed from the media by present methods of purification. Furthei study of certain phases of organic metabolism in plant organisms with different boron or calcium requirements might conceivably point to biochemical reactions for which boron or calcium, or both, may be indispensable. Another among the chemical elements eff"ective in micro quantities to which an indispensable function in the growth of higher green plants must be assigned is zinc. This element, like boron, is not always adequately supplied by the soil, and the deficient plant becomes diseased ("little-leaf," "motde-leaf" of trees, and pathological ^conditions shown by other crop plants as a result of zinc deficiency). Physiological studies under the control of artificial culture disclose some of the effects of zinc deficiency. One such study in this laboratory yielded evidence that, without an adequate supply of zinc, plant growth substances of the auxin type are either not synthesized at a rate sufii- cient for normal growth or else are destroyed too rapidly (12). But it has also been learned that protein synthesis is retarded when the zinc concentration in the plant falls below a critical level. Tomato plants were grown with graduated supplies of zinc, so that some of the plants 67 D. R. HOAGLAND at the time of the experiment gave no visible symptomatic deficiency response, yet addition of zinc to the nutrient medium rapidly induced an increased rate of protein synthesis (1). Observations of this kind, however, only show that some unknown link in a chain of reactions is broken. The nature of the enzyme systems existing in the plant, of which zinc is an essential component, or activator, is an unanswered question. From investigations on blood cells comes evidence that zinc is a component of carbonic anhydrase; and there is a suggestion from studies on yeast that zinc is one of the activators of the enzyme aldolase. We have no positive evidence of this character derived from experiments performed directly on enzyme systems of higher plants. Zinc is often successfully applied in curing zinc deficiency disease by spraying the plant or treating the soil. This discovery, valuable as it is in agricultural practice, does not satisfy the curious investigator who seeks enlightenment on the function of zinc in plant metabolism. Knowledge of why zinc is needed by the plant might improve existing agricultural practice by providing a rational basis for the treatment of zinc deficiency disease; but clearly, as in other fields of research, sound progress in the study of plant biochemistry cannot be hoped for if the direction of research is to be governed by the degree of prob- ability that a given investigation will in itself have a practical outcome. These are only illustrations of the wide gaps in fundamental biochemical insight into the functions of the so-called plant foods. One could write of the lack of the kind of experimentation which might elucidate the role of potassium, one of the most important fertilizer elements in enzymic reactions in the plant. At the present time, it is possible to cite data from one source or another that could be in- terpreted in terms of an eff'ect of potassium on almost every general biochemical process of which the plant is capable. The net conclu-^ sion is that potassium is an essential element for plant growth and that its deficiency may impair growth in various ways or alter the compo- sition of the plant, depending upon the factors such as degree of potas- sium deficiency, the concentration in the media of calcium, sodium, or other ions, the species of plant studied, and the physiological age of the plant. The desirability of further basic information on the role of potassium in biochemical processes in the plant is evident. For example, a suggestion has been advanced that potassium has an im- 68 BIOCIII'MIS'FRY OI' I'LANIS portant effect on certain of the pliosphorylation processes in muscle, withi an antagonistic action by calcium. Studies of this ty])e, applied to the enzyme systems of higher plants, would obviously be of great significance for plant nutrition. If we turn to the essential element phosphorus, great encourage- ment is gained for the view that biochemical research on animal tissues and on microorganisms can serve as a guide for extension of research on the metabolism of the plant. In fact, current research on phosphorus metabolism of various organisms has far-reaching implica- tions for problems of plant nutrition from both a theoretical and practical standpoint. Following the work of Cori on the phosphorolytic system in animal tissues which brings about the synthesis of glycogen from glucose-1 -phosphate came the discovery by Hanes (5) "that an enzyme system can be prepared from plant tissues (potato, peas) which cata- lyzes reversible phosphorolytic reactions by which starch can he syn- thesized in vitro. Physical and chemical studies have been made of the synthesized starch, and its structure compjared with that of ii;itural plant starch (7,11). The conclusion is that the aitificially synthesized starch repre- sents only the amylose component of natural starch, that is, the one made up of long chains of glucose units, whereas the natural starch also includes a component characterized by a branched-chain structure. Recently, the failure to reproduce artificially the complete natural starch was apparently overcome. A preliminary report has been made of the isolation of another enzyme system from potato which can accomplish the synthesis in vitro of the amylopectin component of starch, with the branched-chain structure (8). The investigations as a whole on this question therefore represent clarification of the role of an essential inorganic element in synthesis by the plant of one of its most important carbohydrates. It is not difficult to appreciate the way in which this addition to plant biochemistry may aid in the guidance of researches in plant nutrition with reference to the utilization of phosphate. From the agricultural point of view, the information now available may well become of value in appraising the adequacy of the phosphate supply for high yields or starch content of a crop, especially as further research provides more data on concentiations of various forms of phosphate in the plant under diverse nutrient and atmospheric 69 D. R. HOAGLAND environments. To what extent, if at all, the relative proportions of amylose and amylopectin may be subject to modification by physio- logical conditions remains to be studied. The phosphorolytic mechanisms probably will supply the key to an understanding of the synthesis of another carbohydrate almost universally synthesized by higher plants, namely, sucrose. This is also the dominant sugar of commerce and, as one authority has pointed out, is manufactured commercially in far greater quantity than any other pure chemical product. It is natural, therefore, that many at- tempts have been made to analyze the mechanism of sucrose synthesis in the plant. Recently the enzymic synthesis of sucrose in vitro was accomplished (7); and, while the enzyme system responsible is derived from a bacterial organism (Pseudomonas saccarophila), the achievement has such great suggestive interest for the investigator of the nutrition of higher plants that it seems appropriate to mention it in this connection. The substrates utilized in the synthesis were glu- cose-! -phosphate and fructose. Sucrose is broken down into these components in a phosphorolytic reaction catalyzed by an enzyme system in the bacterial cell. By starting with glucose-1 -phosphate and fructose in the presence of the enzyme, pure crystalline sucrose was prepared which was identical with the natural product. The identity was established by all available physical and chemical criteria. This is the first well-authenticated synthesis of this sugar. It is true that a similar enzyme system has not yet been iso- lated from the tissues of higher plants, despite various attempts to do so in this laboratory. Nev^ertheless, biochemical studies on various species of plants strongly support the view that the synthesis of sucrose does proceed by chemical reactions in which glucose or fructose phosphate esters, or both, serve as substrates, although the mechanism is probably not identical with that of the bacterial enzyme system. Certain studies on sugar cane leaves suggest that, in the higher plants, fructose di- phosphate takes part in the synthetic reaction (6). It is of funda- mental importance that the experimental evidence now available shows that, for the synthesis of sucrose from glucose and fructose in the plant, aerobic metabolism is indispensable. Possibly aerobic oxida- tions are essential to the phosphorylation of one of the substrates in- volved in the synthesis of the sucrose. The question is complicated by the observation that various substrates other than glucose and fructose 70 BIOCHEMISTRY OF PLANTS may result in sucrose formation by plant tissues. For example, in experiments on barley shoots by infiltration procedures, galactose could be utilized for this purpose, as well as various other carbohydrates or related compounds. In the leaves of the sugar cane, as studied by Hartt, the oxidative system involved was not inhibited by cyanide, although in some experiments on other plant tissues in tliis laboratory the synthesis was cyanide sensitive. While the mechanisms of sucrose synthesis operating in the higher plant are by no means sufficiently elucidated as yet, there is reason for an optimistic view that further developments along the general lines of attack already pursued will eventuafiy lead to a satisfactory biochemical solution of this important problem of plant metabolism and plant nutrition. Closely related to the investigations just outlined is the long- standing question of the biochemical nature of the interconversion of starch and sucrose in the plant. As an illustration, the well-known sweetening of potatoes at low^ temperatures may be cited. In this process, starch is converted to sucrose. This conversion is also an aerobic process and is inhibited by cyanide and some other respiratory poisons. A mixture of hexose-6-phosphates has been i.solated from potato juice, and also phosphatases capable of hydrolyzing these com- pounds. A tentative scheme to explain the conversion of starch to sucrose has been advanced on the basis of reactions for which hexose phosphate esters are requisite; and, according to the explanation offered, both glucose-1 -phosphate ester and fructose diphosphate are essential (11). In the earlier experiments in Hawaii on sucrose synthesis by sugar cane leaves, fructose diphosphate was likewise re- garded as an essential substrate. On the other hand, in the bacterial enzyme synthesis referred to above, only glucose-1 -phosphate and fructose could be converted to sucrose. The great problem of cellulose synthesis remains without a biochemical explanation. Whether this synthesis can take place only from activities of the organized protoplasm, and whether phosphoro- lytic processes are involved, are in the realm of speculation at the present time. The point to be emphasized by the foregoing remarks is that biochemical research, by contributing to the basic knowledge of carbo- hydrate transformations, can influence profoundly the study of plant nutrition and physiology. There is, of course, open for further research 71 D. R. HOAGLAND the immense problem of the origin of polysaccharides other than cellu- lose and starch, which comprise a large fraction of many tissues of higher plants, such as hexosans, galactans, pentosans, and related com- pounds. There is need for more information about the synthesis of the important pectin compounds. As a brief digression from the main theme, it is of general bio- chemical interest to refer to the specificity of the enzyme system from Pseudomonas saccharophila which catalyzes the synthesis of sucrose. This enzyme system was not restricted in its catalytic potentiality to the synthesis of sucrose. It was also efTective in bringing about the synthesis of two new disaccharides. One was synthesized in vitro from glucose- 1 -phosphate and /-sorbose, the other from the glucose ester and a ketoxylose (3). The versatility of enzyme systems of this type was thereby demonstrated. To the student of practical plant nutrition interested in the application of fertilizers to soil, the utilization of simple nitrogen com- pounds by crop plants is always a topic of dominant interest. This interest is accentuated today because of the enormous expansion of industries for the fixation of atmospheric nitrogen. Greatly increased quantities of fi.xed nitrogen could be made available after the war for agriculture. Field and pot experiments on the effects of nitrogen fertilizers on crop growth are of course legion, but this type of investi- gation discloses little of the biochemical processes by which the nitrate or ammonia absorbed by the plant is elaborated into organic com- pounds such as the amino acids. Fortunately, this subject has attracted the efTorts of a number of able investigators whose primary concern has been that of biochem- istry; for reviews on the history of research on nitrogen metabolism of plants and recent trends, the monograph by Chibnall (2) and re- ports by Vickery ^< a/. (15-18) may be consulted. In the considera- tion of this aspect of plant biochemistry, it is apparent once more that guidance in the interpretation of data and in the design of experiments is greatly influenced by previous research concerned with the bio- chemistiy of muscle and other animal tissues. The importance of organic acids and their cycles of metabolism, including the tricarboxylic acid cycle, have been stressed by some investigators. Prominent among the organic constituents of plants are malic and citric acids. Oxalic acid also occurs very frequently, and 72 BIOCHEMISTRY OF PLANTS succinic acid has been established as a component of the tissues of several plant species whose content of organic acids has been examined. Frequently all the organic acid content of the plant is not accounted for, and unknown organic acids require identification and quantitative estimation. But it seems that all the organic acids postulated as com- ponents of organic acid cycles may be present in plant tissues. A general theory of protein metabolism based on experiments with seedlings and detached leaves has been evolved which assigns significant roles to the amides, asparagine and glutamine, and to various keto acids. As already noted, the mechanisms postulated draw heavily on the explanations advanced to account for transformations of nitrogen compounds in animal tissues; but caution is needed in applying mechanisms based on the study of animal tissues to plant processes, without adequate confirmatory evidence. Efforts have been made, however, to integrate available data on changes in the organic composition of excised plant leaves under experimental conditions, as well as data on the changes that occur in the intact growing plant, into schemes of protein synthesis and breakdown correlated with catalytic cycles. The possibilities of applying to this field of study in the plant the new tool of iso topic nitrogen have been opened by Vickery and his collaborators in a preliminary experiment with the tobacco plant, following the well-knowm research of the Schoenheimer group on animal metabolism. Another aspect of the problem ol nitrogen metabolism is con- cerned with the symbiotic fixation of nitrogen by leguminous plants. Much more will have to be learned about protein metabolism before the biochemical reactions of the nodule organism can be properly understood. Some progress has been made — compare the review by Wilson (19) — but the theories and evaluation of evidence now available are apparently subject to controversy. The extraordinary practical importance of nitrogen fixation and its scientific interest in- vites further efforts in research by biochemists. The brilliant study of oxidation systems in living organisms rests primarily on the experiments and insight of those who have been concerned with muscle and othci animal tissues, or with yeast. No comparable achievements by investigators of highrr pl.-mis come to mind. Some investigators of higher plants, howexcr, h;i\e sought to apply fundamental knowledge gained by studies on other organisms 73 D. R. HOAGLAND to the study of plant respiration. Goddard, for example, considered the role of the cytochrome system (4). It appears that this system is in fact active in some plant tissues, but not in all. Thus, cytochrome oxidase activity w^as demonstrated in the embryos and roots of certain species of plants, also in immature, but not in adult, leaves. A report is made of the successful isolation of cytochrome G from wheat germ. Wheat cytochrome G has the same absorption spectrum as heart cyto- chrome G. Its reduced form is oxidized by heart or wheat cytochrome oxidase. Succinic dehydrogenase was found to be present in wheat germ in small amounts. W. O. James and his associates (10) have undertaken a series of investigations on the enzyme systems extracted from the sap of the barley plant. Here, an ascorbic acid system appeared to be active in normal aerobic respiration. This oxidation system is thought to be characteristic of higher plants. The degradation of sugars, however, seems to proceed by way of well-known reactions of phosphorylation and hydrogen transfer, as described for other kinds of tissue. Various plant storage tissues have been selected by a considerable number of workers as suitable material for the study of respiratory processes. The effects of respiratory inhibitors on these and other plant tissues have also received much attention. A comprehensive survey of the literature of plant respiration from the point of view of modern concepts is greatly to be desired, but these examples may perhaps serve to illustrate the point that enzyme chemists can find broad opportunities in the higher plants. This ma- terial provides an important field of research in which the number of workers is inadequate to cope effectively with the many and formidable problems presented. It is reasonable to suppose that a sufficiently in- tensive effort directed at plant research might be calculated to advance the subject of respiratory systems in general, as well as supply much needed basic knowledge for plant nutrition and all its ramifications in agriculture. The growth of plants in soil and its dependence on the absorp- tion of inorganic salts or their ions by root cells, with its implications for soil and plant interrelations and for fertilizer practices, brings into the foreground the phenomenon of solute absorption and translocation by living cells — compare the review by Hoagland (9). While these phenomena are of general significance to the study of physiological 74 BIOCHEMISTRY OF PLANTS processes in all living organisms, they possess a peculiar importance in the consideration of the nutrition of higher plants for the reasons already suggested. At one time, the absorption of salts by plant cells would usually not have seemed to belong to the domain of biochemistry. The intake of solutes was regarded as a passive process, in which the per- meability of protoplasmic membranes was chiefly stressed. It is now well recognized that solutes may move into plant cells against concen- tration or activity gradients at the expense of metabolic energy. This kind of absorption of salts by living cells of the root, often referred to as salt accumulation, is dependent on aerobic respiration. The accumula- tion process is inliibited by many respiratory poisons, such as cyanide and iodoacetate, not only in the initial absorption of salts or ions by the root, but also in their polarized movements into the plant's upward conducting system, and their subsequent accumulation in the cells of the leaf or reproductive organs. There are various theories of the mechanisms by which solutes move through the living conducting system of the plant. It is agreed that simple diflTusion cannot explain the movement, and the conclusion cannot be escaped that at some point solutes move against gradients or are accelerated in their movement through coupling with some energy-yielding process. Concurrently with the accumulation of some ions by plant cells, various metabolic processes are stimulated, such as synthesis of organic acids, or proteins, and oxidation of sugars. In fact, the biochemistry of salt absorption by plant cells, in its various aspects, appears to offer a profitable branch of research in the plant field. Researches on storage tissues of plants provide eloquent testimony of the value of studying biochemical transformations as part of an investigation of salt accumulation — compare the review by Steward (14). In the development of studies on salt absorption or movement, the tool of radioactive isotopes may become of great value. Thus certain metabolic reactions may be related to the movement of a se- lected radioactive ion, which can be detected with extreme sensitivity of measurement. So far, the use of radioactive isotopes in research on higher plants has been limited in the main to simple tracer studies designed to obtain information on rates or direction of movement and to identify tissues through which translocation or accumulation takes place. A much wider field for the application of isotopes, stable and radioactive, awaits development, and with it may come knowledge ol 75 D. R. HOAGLAND the biochemical aspects of salt absorption and particularly of the energy-yielding reactions which are inextricably bound up with this process. While it is generally accepted now that the accumulation of salt by plant cells is in some way closely linked with metabolism, and particularly with aerobic respiration, the incompleteness of this knowl- edge should be clearly recognized. Even, if the steps in the particular respiratory cycles, and the active enzymes, coenzymes, and activators participating, should be identified to the extent that they sometimes have been in biochemical studies, the mechanisms by which metabolic reactions are coupled to the movement of solutes into the living plant cell, or to its polarized translocation from one tissue to another, would still be obscure. These are questions that have not been answered for any living cells, and the plant organisms may offer an especially favor- able, system for further research on solute movement. It is hoped that these remarks may serve to call attention to the abundant opportunities for fundamental biochemical research on higher plants, including the great groups of plants of economic impor- tance. There is a place in this field for more workers of the kind who have done so splendidly in advancing biochemistry, particularly in its relation to animal nutrition, medicine, and the metabolism of micro- organisms. References (1) Bean, R. S., Ph.D. thesis, University of California, 1943. (2) Chibnall, A. C, Protein Metabolism in the Plant. Yale Univ. Press, New Haven, 1939. (3) Doudoroff, M., Hassid, W. Z., and Barker, H. A., Science, 100, 315 (1944). (4) Goddard, D. R., Am. J. Botany, 31, 270 (1944). (5) Hanes, C. S., Proc. Roy. Soc. London B128, 1421 (1940); B129, 174 (1940). (6) Hartt, C. E., Hawaiian Planters' Record, 48, 31 (1944). (7) Hassid, W. Z., Doudoroff, M., and Barker, H. A., J. Am. Chem. Soc, 66, 1416 (1944). (8) Haworth, W. N., Peat, S., and Bourne, E. J., Mature, 154, 236 (1944). (9) Hoagland, D. R., Inorganic Nutrition of Plants. Chronica Botanica, Waltham, 1944. 76 BIOCHEMISTRY OF PLANTS (10) James, VV. O., Heard, C. R. C, and James, G. M., New Phytologist, 43, 62 (1944). (11) McCready, R. M., Ph.D. thesis. University of California, 1944. (12) Skoog, F., Am. J. Botany, 27, 939 (1940). (13) Spoehr, H., Plant Physiol., 17, 397 (1942). (14) Steward, F. C, Trans. Faraday Soc, 33, 1006 (1937). (15) Vickery, H. B., Leavenworth, C. S., and VVakeman, A. J., J. litol. Chem., 125, 527 (1938). (16) Vickery, H. B., Leavenworth, C. S., and Wakeman, A. J., Conn. Agr. Expt. Sta. Bull., No. 422 (1940). (17) Vickery, H. B., and Pucher, G. W., J. Biol. Chem., 128, 703 (1939). (18) Vickery, H. B., Pucher, G. W., Schoenheimer, R., and Rittenberg, D., J. Biol. Chem., 135, 531 (1940). (19) Wilson, P. W., The Biochemistry of Symbiotic Nitrogen Fixation. Univ. Wisconsin Press, Madison, 1940. 77 BIOLOGICAL SIGNIKIGANCE OF VITAMINS C. A. ELVEHJEM, professor of biochemistry, college of agri- culture, UNIVERSITY OF WISCONSIN; WILLARD GIBBS MEDALIST F PREVIOUS to the 20th century, thousands, and more likely millions, of people suffered and died because of a lack of scientific knowledge about vitamins or of an insufficient supply of foods rich in these essential nutrients. During the first three decades of this century, about a dozen vitamins not only have been identified, isolated, and synthesized, but manufacturing methods have been perfected to the point at which some vitamins can be supplied at a cost relatively lower than the cost of calories and proteins. Most nutri- tionists agree that there are more vitamins to be isolated and better methods of synthesis to be developed, but I wonder how many have given thought to the possibility that the uncontrolled production of synthetic nutrients may lead to sufficient economic disturbances in agricultural production to aflfect the health of the people of the world adversely. It is unwise, especially for a biochemist, to make any predictions of future developments. The field of vitamins, however, has now de- veloped to the point at which it is possible to look ahead in light of past experiences. The early workers on vitamins had a point of view or philoso- phy quite different from that held by present investigators. Many of the pioneers were motivated by the single purpose of alleviating human 79 C. A. ELVEHJEM suffering. When liver was given to relieve night blindness and fresh vegetables were used to cure scurvy, the practitioner knew nothing about vitamins — he was interested in healing the patient. Eijkman, I am sure, carried a mental picture of the severe cases of beriberi which he encountered in Java during all his attempts to relate this disease to a specific essential nutrient. R. R. Williams referred to his early con- tact with the disease in his Willard Gibbs award address as follows: "In short, beri-beri was a principal topic of conversation in scientific and medical circles in Manila during those early years of my enlistment with Vedder in the Philippines." When Goldberger was called upon in 1914 to undertake studies on the cause of pellagra, he knew very little about the disease, but on December 13, 1915, he wrote as follows to Dr. Milton Rosenau: "I can hardly describe the feeling that I experience as I go through our wards at the asylum and see the poor insane women who a year ago had pellagra but who this year are per- fectly well — so far as pellagra is concerned." Regardless of the satisfaction experienced by the individual workers, these phenomenal results did not captivate the interest of administrators of research funds. More support was given for studies on animal nutrition, since a premium was placed on production. Few recognized that the development of strong human bodies would also pay dividends. Perhaps we can now explain this difference in reaction. The animal husbandrymen took great interest in judging and selecting fine stock. If better nutrition produced better stock they were inter- ested. On the other hand, medical students have always been given sick people to study rather than the ultrahealthy. For example, Sir Robert McCarrison of England went to India to study disease but his most important contributions originated because he was impressed by the perfect physique of the Hunza race. At present many are interested in expanding our conception of the relation of nutrition to optimum health, but we still are not too certain about the procedure; some talk about extra quantities of vitamins, others advocate physical training. It was not surprising, therefore, that between 1910 and 1920 the following laboratory findings continued to attract widespread interest. (1) Calves maintained on diets balanced according to the recognized standards failed to grow on rations made entirely from products of the wheat plant but thrived on rations made from products of the corn plant. (2) Rats placed on purified diets developed nor- 8o BIOLOGICAL SIGNIFICANCE OF VITAMINS mally when the fats of the diet consisted largely of butterfat but failed when certain vegetable oils devoid of vitamin A were used; this ob- servation was especially significant at a time when Danish children given skim milk plus vegetable fats were developing the same symptoms as those observed in rats. (3) Chicks grew normally if allowed access to sunlight shortly after hatching, but developed severe leg weakness when hatched early in the spring; at that lime attempts were being made to start the chicks during the winter months so tliat broilers would be available when the demand was heavy. When the de- ficiency agent was found to be vitamin D, not only was the poultry industry saved but a program was initiated which led to the eradication of human rickets. By 1925, many laboratories were undertaking systematic nutritional studies, and further attempts were being made to produce specific beriberi in rats and chicks, vitamin A deficiency in rats and dogs, vitamin C deficiency in guinea pigs and monkeys, vitamin D deficiency in chicks, dogs and rats, and pellagra in dogs. Rough assay procedures were developed; foods richest in each of these vitamins were designated "protective" foods. Theoretically, this knowledge was all that was needed to prevent the deficiency disease resulting from the lack of each respective vitamin. Carlson, a few years ago, stated that we had sufficient knowledge to prevent all beriberi in the world long before vitamin Bi was synthesized, a statement which is true in a limited sense. Thus, Chamberlain and Vedder reduced the incidence of beri- beri in the Philippine Scouts by issuing unpolished rice, but the prob- lems of producing high-quality unpolished rice and of educating people to use this type of rice product as part of their diet are still with us. All these studies were most intriguing to ever-increasing numbers of research workers. The bars were now down to the progress of nutri- tion research and the studies gained momentum each year. Some wanted to know what happened during each deficiency and how the vitamin functioned in producing normal animals. One of the first attempts involved histological studies of tissues from vitamin- deficient animals; and it was soon recognized that the absence of a minute trace of a nutrient led to extensive structural changes in certain tissues. Even greater progress was made when the function of vita- mins was related to the dynamics of the living cell. In 1921, Seidell stated that, aside from a possible significant diff'crcnce in the degree of 8l C. A. ELVEHJEM dialyzability, there are no grounds for not classifying vitamins with enzymes. Prior to this, Harden and Young, in 1906, emphasized the importance of organic dialyzable substances in yeast fermentation and called these substances coenzymes. In 1918, Meyerhof found the co- enzymes of yeast to be present in a number of animal tissues; but animal workers were too busy compounding rations to pay any atten- tion to this finding. R. J. Williams, in 1919, concluded that the sub- stance or substances which stimulate the growth of yeast are identical with the substance or substances which in animal nutrition prevent beriberi or polyneuritis. Then, in 1932, Warburg and Christian found that the so-called yellow enzyme which they had shown to be active in a reconstructed oxidation system contained a derivative of riboflavin, the second mem- ber of the B complex, as the prosthetic group, A short time later, it was found that the coenzyme used in the same system and prepared from red blood cells contained nicotinic acid. In 1937, nicotinic acid was shown to be the antipellagra factor and thus became identified with the third member of the B complex. In 1932, Auhagen split carboxyl- ase, the enzyme necessary for the metabolism of pyruvic acid, into a protein component and a thermostable part called cocarboxylase. Soon cocarboxylase was identified as the pyrophosphoric acid ester of vitamin Bj. The fourth decade of the 20th century will undoubtedly be recognized as the period of greatest advance in our knowledge of the mechanism of action of the vitamins. We must recognize that much is still unknown and that some of the most difficult problems lie ahead. However, the results so far obtained have had a much broader in- fluence — they have given new impetus to the study of enzyme chemis- try. R. R. Williams states: "These enzyme molecules are too vast and complex for the chemist to decipher completely today, but we can now say that the prosthetic group or business ends of these molecules are in many instances what we earlier came to call vitamins. These vitamins are, therefore, the bits, the working ends, of the keys which unlock stores of vital energy from glucose and other foods." The enzyme approach has emphasized the close relationship of all types of life. Vitamins have turned out to be growth factors and metabolic regulators for plants, bacteria, protozoa and yeast, as well as for animals. See W. H, Peterson, Biol. Symposia, 5, 31 (1941). 82 BIOLOGICAL SIGNIFICANCE OF VITAMINS Others wanted to know what a vitamin looked hke. The first crystalHne material to be isolated from a natural concentrate having vitamin activity was probably nicotinic acid. It was obtained between 1912 and 1914 from rice bran and yeast; but unfortunately its bio- logical activity was tested for antineuritic activity rather than for anti- pellagra activity. The successful establishment of the chemical con- stitution of several of the vitamins depended upon an enormous amount of work and true chemical ability. The first of the vitamins to be given serious chemical consideration was undoubtedly vitamin D. In 1925, Steenbock and Black, and Hess and co-workers showed that crude cholesterol could be activated by ultraviolet light, and the following year ergosterol was recognized as the actual provitamin. These ob- servations stimulated the interest of organic chemists in the structure of sterols, a problem that had received only sporadic attention. A little later, the chemical basis for the relationship between carotene and vitamin A was established. It is interesting to note that, although the structure of vitamins A and D received early attention, these are the only well-known vitamins still not available in synthetic form. Vitamins C and Bi were the first to be made synthetically, but only about ten years ago. During the past decade, methods for the synthesis of ten different vitamins have been perfected. I have merely recorded the final results without paying tribute to the individual workers for their years of study of the details of chemical structure. It is true that some of the work was stimulated by commercial interest, but in many cases the individual workers were rewarded only by the satisfaction obtained from the successful proof of structure of "their" vitamin. That chemical industry did become interested in the production of vitamins was indeed fortunate. The availability of each new vitamin facilitated progress on the remaining vitamins. As far as I am aware, no one has formally expressed the grati- tude of laboratory workers for the large quantities of vitamins supplied gratis by industry for experimental purposes. Although it is true that many papers carry a footnote indicating indebtedness to a particular firm "for a generous supply of crystalline vitamins," such an acknowl- edgment is so common today that it is often taken for granted. I have no way of estimating the total expenditure involved, for the value of these gifts cannot be calculated merely by multiplying the number 83 C. A. ELVEIIJEM of pounds supplied by the current price. The crystalhne material was most valuable to the investigator \vhen the supply was still in limited production. For example, our work on the newer members of the B complex with the chick was directly dependent upon our ability to obtain adequate supplies of pure biotin. Currently, everyone is inter- ested in feeding his animals purified rations containing only the syn- thetic vitamins; and the vitamin requirements become rather large when dogs, monkeys, pigs, and even human subjects are used. Many of us today would be willing to pay a fancy price for even a few milli- grams of pure folic acid. If work on the chemistry of this and related compounds had not been limited by the war, sufficient quantities of it would undoubtedly now be available for experimental purposes. As the methods of synthesis improved and the demand for the compound increased, very substantial decreases in the wholesale prices of most of the vitamins were made. The cost of riboflavin has decreased from $17.50 per gram in April, 1938, to 30 cents per gram in gram lots in October, 1944. In January, 1934, vitamin G cost $213 per ounce; today, one ounce may be purchased for 95 cents. The reduced prices made these vitamins available for many purposes other than for the manufacture of elixirs, tablets, and cap- sules: synthetic ascorbic acid is added to the lemon powder used in army rations; B vitamins are added to flour, bread, and corn grits; and several of the vitamins were supplied to other countries through "Lend-Lease." In 1944, production of certain of the individual vita- mins ranged from 100,000 to 1,000,000 pounds. Vitamins are no longer limited to the laboratory and the doctor's office — they are now part of big business: more extensive use of vita- mins means greater dividends to the stockholders of many industries; therefore, large advertising campaigns have been instituted; and as profits increase, more funds become available for research. A few years ago some of us were highly pleased if we received $500 to sup- port a favorite project. Today, yearly grants as high as $50,000 are made for nutrition research — a magnificent start. We have the inter- est of the public and the support of industry, and we should have many well-trained and energetic investigators in the postwar period. But what of the future? First we must realize that many of the workers will be interested in research merely for the sake of research. Many will have had little contact with extensive deficiency diseases. 84 BIOLOGICAL SIGNIFICANCE OF VITAMINS This should not be disturbing if the findings are properly applied. The practical problems will involve economics as well as chemistry and physiology. In the field of medicine, the doctor will continue to prescribe vitamins for deficiencies which are clearly diagnosed, and in some cases he will try vitamins to determine if any beneficial effects can be obtained. If the patient recognizes some benefit, the use of vitamin supplements will be continued for some time; if no benefits are recognized, the box of capsules will probably remain on the shelf of the medicine cabinet. A certain group of people will buy vitamin preparations on their own initiative but few will take capsules con- tinually for any length of time. Extension of the use of vitamins will probably ha\c to come by way of the addition of vitamins to widely used foods. The supply of niacin did not become critical immediately after the discovery of its role in curing pellagra; but the supply was critical within a few weeks after the introduction of flour enrichment. If my calculations are correct, 1,000,000 pounds of niacin is almost sufficient to su[)ply the minimum requirement of all the people in the United States for one year, an amount which the annual production is now reaching. What will happen now that war is over and synthetic niacin will be in direct competition with niacin in meat, and synthetic ascorbic acid will be in competition with vitamin C in oranges, tomatoes, etc.? Will industry be willing to control the pioduction of synthetic vitamins in relation to the true demand for these products? I would be greatly disturbed by an extensive advertising campaign advocating greater use of syn- thetic vitamin C by the public at a time when the orange crop is rotting in orchards. On the other hand, if the synthetic vitamins are used to supplement rather than replace our food supply, we can plan for this country — yes, even the world — a continuous supply of nutrients which will be. little affected by crop failures. I hope such a plan can be made and executed before the problem becomes so acute that govern- ment may have to step in. At present, the use of vitamins is promoted largely among people who are rather adequately fed. What results we could expect from the proper use of vitamins in the contiol of famines in India ! It is true that vitamins cannot replace other food nutrients but certain vitamins at least increase the elliriency of utilization of the total nutrients and may also help in the synthesis of other vitanuns m the intestinal tract. 85 C. A. ELVEHJEM I believe there is another important angle which applies not only to the use of synthetic vitamins but also to synthetic amino acids, which undoubtedly will be produced in the postwar period. Since, in general, the synthetics have no taste appeal, and since mankind will continue to consume food for reasons other than that of mere nutrition, it may be more important to use a larger part of the synthetics in animal feeding. As we learn more about nutrition, we are finding that many of the more expensive animal feeds can be replaced by cheaper sub- stitutes. For example, riboflavin can be used in poultry feeding with- out relying upon more expensive milk products. Thus, the cost of animal production can be reduced to such an extent that animal products can be used more widely for human consumption. The necessity of fortifying certain human foods may continue for some time. Although new types of food fortifications may be intro- duced, we must recognize that any enrichment program is not neces- sarily permanent, and we should be willing to discontinue any one program when and if scientific evidence indicates that it is no longer necessary. It will be the duty of nutritionists to give careful considera- tion to these programs; proper decisions can be made only if we have extensive knowledge of the vitamin content of all foods. Food indus- tries have generously supported such programs but the work has cer- tainly not reached completion. Plans should be made to set aside funds which are now easily obtainable so that work of this kind can be carried out when personnel become available. There are two im- portant lines of approach: one deals with the production of food products high in vitamins and can be accomplished by improved breeding, cultivation, and fertilization; the other deals with improved methods of handling the food products after harvesting and slaughtering. Because fundamental research must continue in the field of vitamins, it will be fortunate to have young men interested in pure re- search. During the past few years, the practical problems have re- ceived greatest emphasis, but we have now reached the point at which fundamental research again becomes the limiting factor in further prog- ress. We must study cellular mechanisms within the body and the relation of bacterial cells to the vitamins within the intestinal tract. The relation of vitamins to enzymes has already been discussed. After all, sturdy bodies are largely dependent upon properly function- ing enzymes in all the cells of the body. Vitamins are only a small part 86 BIOLOGICAL SIGNIFICANCE OF VITAMINS of the enzyme molecule and we need to know more about the rest of the molecule. Perhaps proper exercise with limited amounts of vita- mins may be more conducive to rigorous enzyme systems than the consumption of vitamin cocktails while reclining in an easy chair. Studies on the enzyme systems will help to relate nutrition to such important problems as resistance to infection, prevention of cancer, and resistance to the process of aging. Biochemists have learned how to disorganize cells into parts, but greater integration of the parts into a whole must be attained. More attention also must be paid to the bacteria in the intes- tinal tract. Bacteriologists have studied the bacteria in soils, in milk, in foods, and in disease, but have largely disregarded the bacteria of our intestines. There are still nutritional disturbances which must be re- lated indirectly to the changes in the digestive tract.- Very recently, a most interesting report was presented on the nutritional status of about 800 individuals living in Newfoundland: certain symptoms observed were ascribed to riboflavin and niacin deficiencies, and yet a rough estimate indicated that the intake of these vitamins was not seriously inadequate. I believe some of the changes, at least, are due to a lack of as yet unknown vitamins which were not synthesized in suffi- cient amounts because of the type of dietary regime. Pellagra has always been associated with a large consumption of corn. Preliminary evidence in our laboratory indicates that an extensive corn intake may adversely affect the synthesis of vitamins in the intestinal tract, an effect which is overcome by high levels of nicotinic acid. High levels of protein also have a counteracting effect, which may explain why milk has been found to have antipellagra activity although it is known to contain little nicotinic acid. Intestinal synthesis is not limited to the production of the vita- mins which are the last to be discovered. Obviously, the degree of synthesis in the intestine must be less in the case of the older vitamins, or we would have had more difficulty in producing the deficiency state of these vitamins. Some time ago, we showed that the fat content of the diet had a marked effect on the riboflavin requirement of the rat. Diets which contained dextrin and low levels of ribofla\'in produced a much more severe riboflavin deficiency when a large jjortion of the dextrin was isocalorically replaced by fat. Later work has shown that this effect is direcdy dependent upon a decreased synthesis of riboflavin S7 C. A. ELVEHJEM in the presence of fat. There is much evidence to show that fat, carbo- hydrate, protein, and vitamins are all interrelated in their effect on the production of both known and unknown vitamins in the digestive tract of all animals; but the results for one animal cannot be predicted from the results obtained with another species. Just where the human fits into the picture is impossible to say. It is encouraging to find that several groups of workers are engaged in studying this problem on human subjects, a project which will undoubtedly clear up many of the difficulties now encountered in attempting to establish quantitative requirements for each of the vitamins in human subjects. The final answer can probably be made only when animals are rendered bac- teria-free and their requirements are studied under these conditions. Although this will answer the questions from an academic point of view, for practical purposes we must continue to recognize the inter- relationship of food and intestinal bacteria. No one need feel that further work in the field of vitamins will not be productive. Much human suffering has been alleviated through our knowledge of vita- mins and we can expect much success in the future if we learn more about these interesting compounds and if we apply what we learn in a sensible manner. Selected References Addinall, C. R., "Synthesis and production of vitamins," Chem. Eng. News, 22,2174 (1944). Black, J. D., ed., "Nutrition and food supply: The war and after," Ann. Am. Acad. Political Social Sci., 225 (1943). "Enrichment of flour and bread. A history of the movement," Bull. Natl. Research Council, No. 110 (1944). Evans, E. A., Jr., ed., Biological Action of the Vitamins. Univ. Chicago Press, Chicago, 1942. Major, R. T., "Industrial development of synthetic vitamins," Chem. Eng. News, 20, 517 (1942). "Medical survey of nutrition in Newfoundland by a group of investigators," Can. Med. Assoc. J., 52, 227 (1945). Rosenberg, H. R., Chemistry and Physiology of the Vitamins. Interscience, New York, 1945. Schultz, T. W., ed., Food for the World. Univ. Chicago Press, Chicago, 1945. Williams, R. R., and Williams, R. J., "Vitamins in the future," Science, 95,335-344 (1942). 88 SOME ASPECTS OF VITAMIN RESEARCH KARL FOLKERS, director of organic and biochemical research, MERCK & CO., inc.; AMERICAN CHEMICAL SOCIETY AWARD IN PURE chemistry; CORECIPIENT of the mead JOHNSON AWARD So little is known about the chemistry oj vitamins — not a single one has been isolated with absolute certainty — that I have hesitated to include this subject among the applications of organic chemistry. The very extensive contemporary literature on vitamins which takes up much space in journals devoted to biochemistry^ contains few chemical facts, and very few that are thoroughly well established. T\ ^HIS statement was made in 1928 at Cornell University by George Barger (1) during his lectures on some applications of organic chemistry to biology and medicine. These lectures were concerned with hormones, vitamins, chemical constitution and physio- logical action, chemotherapy, and blue adsorption compounds of iodine. The number of well-established chemical facts on the chemis- try of the vitamins developed so enormously during 1928 to 1945 that an entire university course could justifiably be devoted now to the organic chemical aspects of the vitamins. The companion develop- ments on the biochemistry of the vitamins and on the application of the vitamins in clinical medicine might also require a course each for adequate presentation to students. The industrial production of vitamins on a ton basis, a subject recently reviewed by Major (30), is no less amazing in the rapidity and magnitude of the develop- ment. Today there are so many excellent books and review articles on the chemistry of the vitamins available, that no effort will be made in the following sections to cover any topic completely. Instead, a 89 KARL FOLKERS few observations, facts, and results which may have unique interest have been selected for comment. On the Discovery of Vitamins The discovery of the "major vitamins"* has been based upon observations which related the syndrome of a human disease to con- stituents of natural materials used in nutrition. The discovery of the "lesser vitamins"* has been based upon observations which related biological reactions generally produced experimentally with animals or microorganisms to constituents of natural materials. Although the "lesser vitamins" are at present not known to correspond to any histori- cally recognized human disease, they probably are essential for the human being. They may be considered "lesser vitamins" today be- cause their absence in human diets is less frequent statistically or because the signs of their absence are not yet fully recognized, as was the case for riboflavin deficiency until 1938, when Sebrell and Butler (43) characterized ariboflavinosis. Oden, Oden, and Sebrell (37) con- cluded a little later that ariboflavinosis is "a common dietary-deficiency disease in the southern United States." It seems not unlikely that certain of the "lesser vitamins" will ascend to the class of "major vitamins" after further clinical research. Some of this future clinical research might lie in the borderline fields between biochemistry and psychology, according to R. J. Williams (63). His observations on "personality diff"erences" among animals in nutrition experiments, and the fact that hallucinations and mental symptoms of pellagra are known to be eliminated by administration of nicotinic acid, helped stimulate this interesting thought. Certainly, if other vitamins were found to benefit mental disease or psychological disturbances, the rank of importance of these vitamins would be elevated. R. R. Williams (69) believed it was probable that any * R. R. Williams, in his stimulating address on the occasion of the presenta- tion of the Chandler Medal in 1942, defined the "major vitamins" as thiamin, riboflavin, nicotinic acid, and vitamins A, D, and C. Five of these vitamins are related to ancient and widespread diseases. Ariboflavinosis, which is cured by riboflavin, had been confused with and masked by pellagra and was not recognized per se until recently. The "lesser vitamins" include choline, vitamin Be, pantothenic acid, biotin, inositol, etc. See reference (69). 90 VITAMIN RESEARCH vitamins yet to be discovered are destined to have lesser nutritional significance for human welfare and that the vitamins which are re- quired to check the nutritional plagues of mankind have already been discovered and produced. Nevertheless, he recognized that there may be exceptions, particularly in the case of obscure diseases. Un- doubtedly, chemists and nutritionists must cooperate scientifically for many years on the problems of the discovery of new vitamins. On the Isolation of Vitamins The isolation of a vitamin from the natural material in which it exists is essentially a chemical problem of the same nature as the older problems on the isolation of an alkaloid or a glycoside, but with at least two important dilTerences. One of these differences is that vita- mins generally occur in quantities amounting to a few parts per million of the natural material, whereas the common alkaloids and glycosides are found frequently in quantities amounting to a few parts per hundred. The greatest difficulties in vitamin isolation might be said to lie in the region of converting the natural materials with a few parts per million of the substance to a concentrate containing a few parts of the substance per hundred. New techniques and new procedures are frequently devised to surmount the difficulties in making such a puri- fication. The isolation of trace substances in milligram or gram quanti- ties requires the processing of hundreds of pounds of the natural ma- terial. A second important diff"erence in the isolation procedure is the necessity for countless biological assays throughout the whole isolation work to show the investigator the location (and loss !) of the vitamin in the fractionation. Pioneering researches on the isolation of a vitamin are very costly and time-consuming. It has been said (69) that the first gram of pure natural thiamin must have cost an aggregate of several hundred thousand dollars. Eight years transpired between the first success in isolating this vitamin m 1926 by Jansen and Donath (21) and the work of Williams, Waterman, and Keresztesy (70) m 1934, which resulted in greatly improved yields, so that a sufficient quantity of the pure vitamin could be made available for its structure determination. It took five years in Kogl's laboratory at the University of Utrecht in Holland to work out the pioneering methods which yielded seventy 91 KARL FOLKERS milligrams of crystalline biotin (26). The difficult and tedious scheme oi" hactionation involved a three million-fold purification. Kogl (26) estimated that to produce one gram of their biotin from ordinary yeast would have required 360 tons of the yeast as starting material. Although he found egg yolk to contain ten times as much biotin as yeast, the number of fresh eggs required for the production of one gram of biotin would have cost about $165,000 in 1937. Special chemical steps or new techniques often have to be de- vised or applied to the isolation process before the vitamin can be ob- tained in sufficient amounts for the complete elucidation of its structure. The improved yields (70) in the process for the isolation of thiamin depended upon the elution of the vitamin from fuller's earth with quinine acid sulfate instead of barium hydroxide and the introduction of a benzoylation step for purification. Probably one of the most im- portant factors contributing to the success of the isolation of additional quantities of crystalline biotin was the application of the chromato- graphic adsorption technique by du Vigneaud, Hofmann, Melville, and Gyorgy (59) to concentrates of the vitamin which had been ob- tained from beef liver (13). The concentrate contained 0.1% biotin, and, after esterification, the material was chromatographed twice over aluminum oxide. After the final sublimation and crystallization steps, pure crystalline biotin methyl ester was obtained in 38% yield based on the amount of the vitamin in the concentrate. On the Structure Determination of Vitamins Ordinarily, it is desirable to isolate any vmknown natural prod- uct in a state of complete purity before the carrying out of chemical reactions for establishment of molecular formula, identification of functional groups, and finally the determination of structure. Special attention to the question of purity is often justified, because natural products are frequently isolated which are extremely difficult to sepa- rate from final impurities of unknown but of allied properties. There are, however, exceptions to the purity requirement. The isolation of calcium pantothenate in pure form was found by Williams' group (67) to involve extraordinary difficulties, and it was necessary to conduct the structure studies with a highly purified con- centrate estimated to be about 90% pure (34,67). The establishment 92 VITAMIN RESEARCH of the structure of pantothenic acid under tliese conditions was accom- phshed by a rather unique series of developments. ^-Alanine was first reported by WiUiams, Weinstock, and Mitchell (61,68) to be formed from this calcium salt in acid or alkaline medium. The other hydrolytic product was found (34) to l)e an a-hydroxy acid capable of spontaneous transformation to a lactone. The crude lactone fraction was recombined (66,72) with /3-alanine to give material which possessed the physiological activity of pantothenic acid and, of course, actually was this acid. This condensation reaction constituted re- synthesis, and the results supported the earlier observation (34) that /3-alanine was combined as an amide through its /3-amino group, as in structure I. At this stage of the investigation it was clear that it R— CONHCH2CH.2CO..H (I) would be unnecessary to isolate pure pantothenic acid if the hydroxy acid fragment or its lactone could be isolated in pure form. The structure determination of the hydroxy acid would give also the struc- ture of pantothenic acid. Further research (34,53) showed how con- centrates containing only 3 to 40% of pantothenic acid could be purified and treated so as to yield the pure lactone. Once having obtained the pure lactone, Stiller, Keresztesy, and Finkelstein (53) applied degradation reactions which showed the lactone to be a- hydroxy-/3,/3-dimethyl-7-butyrolactone (II) : OH CH— C=0 OH (CH3)2G O (CH,)2C-CH-CONHCH2CH2C02H \ / 1 CH^ CH,OH (II) (III) Obviously, pantothenic acid was rt,7-dihydroxy-/3,/3-dimethylbutyryl- /3'-alanide (III). Because of the great cost and technical difficulties in the isolation of new vitamins, it has been necessary on occasion to open the program of structural research with a series of studies which are designed to cliaractcrize the functional groups of the substance and which can be 93 KARL FOLKERS carried out with micro quantities of the substance. For example, early evidence concerning the non-/3-alanine portion of pantothenic acid was obtained from the results of a series of new micro procedures (34). Some of these new micro procedures involved determination of active hydrogen atoms with deuterium oxide and of hydroxyl groups with hydriodic acid, selective oxidation with iodic acid, oxidation equivalent analysis, determination of a- and /3-hydroxy acids, and estimation of microorganisms in suspension. The use of such tech- niques represented an unconventional but fruitful approach to the study of pantothenic acid and often required only one or two milli- grams of the compound for each determination. Another type of study has been employed to open a program of structural research upon a costly vitamin available in only very limited quantities. This study, involving a series of "inactivation" experiments, requires only a milligram or less of the substance and yields information on functional groups and constitution. These experiments involve adding to the micro sample of the vitamin the chemical reagent (s) required for a given chemical reaction, such as nitrosation or hydrolysis, and following with a microbiological assay to test whether chemical reaction took place as judged by a change or lack of change of activity. These "inactivation" experiments yield valuable results, but they must nevertheless be interpreted with con- siderable caution. They do aid in guiding the exploratory efTorts in direct chemical studies. An example of such inactivation experiments may be found in certain biotin studies. Brown and du Vigneaud (2) described the effect of certain reagents on the activity of biotin. They obtained this preliminary information with experiments on 1- or 2-cc. aliquots of solution containing only 12.5 7 of biotin per cc, and the criterion of reaction was the effect upon yeast growth activity. Such reagents as 5% hydrogen peroxide solution, aqueous bromine, hydro- chloric acid, potassium hydroxide, formaldehyde, and nitrous acid caused "inactivation," indicating that a change in the structure had undoubtedly been brought about by the reagent. On the other hand, such reagents as acetic anhydride-sodium hydroxide, ketene, benzoyl chloride-pyridine, sodium ethoxide-methyl iodide, and ninhydrin caused no "inactivation," indicating that a change in structure had not been brought about by these reagents. The results of these and related experiments indicated to Brown and du Vigneaud that biotin 94 VITAMIN RESEARCH is inactivated by vigorous treatment with acid or alkali, is not an a- amino acid, is not destroyed by acylating or alkylating reagents, and is easily oxidized. Not all of the well-known vitamins have had their constitution elucidated by an application and development of micro methods. The correct structure of a-tocopherol of the vitamin E group was deter- mined largely by interpretation of the results of three chemical reac- tions, which were carried out on 2.1, 4.3, and 25 grams of the vitamin in each case. While repeating some of the isolation work of a-toco- pherol from cottonseed oil described by Emerson, Emerson, and Evans (7), the late Dr. E. Fernholz, working in the laboratory adjoin- ing that of the author, intimated that he desired to accumulate enough of the a-tocopherol to permit reactions on a gram-scale basis. It was feasible to isolate a-tocopherol on this scale. Subsequently, Fernholz described (8) the thermal decomposition of a-tocopherol which yielded durohydroquinone, and the details (9) of the experiment reveal that 2.1 g. of the vitamin had been heated for six hours at 355°. The crystalline sublimate yielded 257 mg., or 67%, of durohydroquinone and a hydrocarbon of the composition Ci8-i9H36- This characteriza- tion of durohydroquinone shattered the frequently discussed idea that a-tocopherol is related to the sterols. The absorption spectra and chemical properties of some synthetic monoethers of durohydro- quinone, when considered in conjunction with other suggestive evi- dence, led to the hypothesis that a-tocopherol was derived from chroman or coumaran. This hypothesis of a heterocyclic ring structure was justified by the results of an oxidation reaction with chromic acid (9). When 4.3 g. of a-tocopherol was dissolved in glacial acetic acid and oxidized with chromic acid, dimethylmaleic anhydride (IV) and /Cf /CH2 CHa— C \ / ^CHa II o o=c I CHa— G / \ C— CieH 33 (IV) (V) a lactone, C21H40O2, were isolated. Structure V was proposed for the lactone after a study of its properties and after a study of the oxidation of the acetate of a-tocophcrol obtained from 25 g. of a- 95 KARL FOLKERS tocopheryl allophanate. The chromic acid oxidation of this acetate yielded diacetyl, acetone, an acid C16H32O2 of probable structure VI, and a ketone CigHseO. CHg CHj CIH3 i I i HOOCCH2CH2CH(GH2)3CH(CH2)3GHCH3 (VI) The interpretation of these degradation products and related evidence led to the proposal by Fernholz (9) of structure VII for a- CH3 Cri2 HOr/N^ ^CHs CH3 CH3 CH, CHs^X Js.^ y-C(Cri2)3CH(Cri2)3Cri(Cri2)3CriCH3 CH3O I.H3 (VII) tocopherol. Although the related coumaran structure was con- sidered further by Karrer, Fritzsche, Ringier, and Salomon (22,23) all the results of the investigations — refer to review paper by Smith (44) — were soon interpreted in favor of the chroman structure VII, which was proposed after a study of a few relatively large-scale deg- radation reactions. An interesting aspect of the structure studies on thiamin and biotin was the application of new organic structural reactions. It had been observed during the isolation work that an attempt to use sulfurous acid as a preservative against the bacterial decay of extracts of vitamin Bi from rice polish caused a prompt and complete loss of vitamin activity. This observation led to the development of a pro- cedure by Williams, Waterman, Keresztesy, and Buchman (71) for the quantitative cleavage by sulfite of pure crystalline vitamin Bi at pa 5 and room temperature into two products of the composition C6H9N3SO3 and CeHgNSO. There are obvious advantages of this reaction over the usual oxidative and hydrolytic reactions which ap- peared to give a miscellany of substances. The reaction is expressed in terms of the structural formulas by the following equation: CH3 N=CNH3+C1- C CCH2CH2OH CH,C C— CH2— N'+ N— CH CI- CH- + Na2S03 96 VITAMIN RESEARCH c:h, N-=CNH2 c-^,,^cCH2CH.OFI / \ / CH3C CCH2SO3H + N N— CH CH — + 2 NaCl The new organic structural reaction used in the investigation of biotin was applied in the belief that the results would lead to a final selection between two alternative structures for biotin. It was believed that organic sulfides could be cleaved by the Raney nickel catalyst according to the equation: Ni(H) RSR' ^ J. RH + R'H This sulfur hydrogenolysis reaction was developed first on several "model" compounds. It was found by Mozingo, VVolf, Harris, and Folkers (35) that representative sulfides could be cleaved to their corresponding sulfur-free products in yields of 65 to 95% on both a macro and semimicro scale. As described by du Vigneaud, Melville, Folkers, Wolf, Mozingo, Keresztesy, and Harris (60), the application of this hydrogenolysis reaction to biotin methyl ester yielded dcsthio- biotin methyl ester, and the subsequent study of this hydrogenolysis product was the first of two independent methods which led to the final proof of the structvu-e of biotin. It was considered originally (35) that this reaction would be of general value in investigations on the structures of natural products containing sulfur. On the Use of Microorganisms in Vitamin Researcli It is beyond the scope of this article to present all of the interest- ing aspects of the application of microorganisms to vitamin research. Williams recently discussed the importance of microorganisms in \-ita- min research (62) and reviewed microbiological tests (64). Gyorg\' reviewed further developments in the use of microorganisms in vitamin research (12). There are, however, certain recent developments in the chemistry of vitamin Bg which were originally provoked l)\- the results of microbiological experiments, and recent results on new vita- mins which seem to merit comment. Snell, Guirard, and Williams (50) found that assays with Streptococcus lactis gave values for the pyridoxine content of certain 97 KARL FOLKERS natural materials which were several hundred to several thousand times the values obtained by different biological or chemical methods. Other experiments showed that the factor responsible for the increased activity was very similar in properties to pyridoxine, and they pro- visionally named the factor "pseudopyridoxine." Subsequent to these studies, Williams stated (65) in his review on the water-soluble vitamins for 1942: "Contrary to the common impression, the chemistry of vitamin Be is in a highly unsatisfactoiy state. There is no question, of course, regarding the fundamental chemistry of pyridoxine, that pyridoxine occurs naturally, or that pyridoxine has vitamin properties. The serious question is whether the vitamin Be activity of tissues is due solely to pyridoxine or whether there are other substances (probably closely related) which serve equally well or better." After about two years of further research at the University of Texas on the one or more substances in natural materials tentatively called "pseudopyridoxine," Snell (45) concluded that one of these substances was probably an aldehyde and another was probably an amine. These studies were concerned essentially with the effect of certain selected chemical treatments on the biological activity of pyri- doxine for Streptococcus faecalis R, Lactobacillus casei, and Saccharomyces cerevisiae. After consideration of the functional groups of pyridoxine which might be involved in the reactions, it was believed that the bio- logically active aldehyde would have one of three structures, VIII, IX, or X, and the biologically active amine would have the corre- sponding structure, XI, XII, or XIII. CH2OH HO/\cHO (VIII) CH20H H0/\CH2NH2 CH: \N^ (XI) CHO H0/\CH20H CH (IX) CH2NH2 H0/\CH20H CH (XII) CHO HO/\cHO CHsi^j^^ (X) CH2NH2 H0/\CH2NH2 CHi (XIII) Collaborative studies by Harris, Heyl, and Folkers (15) on the structure and synthesis of the active aldehyde and active amine resulted 98 VITAMIN RESEARCH in the synthesis of aldehydes VIII and IX and amines XI and XII. Biological tests on these synthetic compounds by Snell (46) showed that the biologically active aldehyde was 2-mcthyl-3-hydroxy-4-formyl-5- hydroxymethylpyridine (IX), and that the biologically active amine was 2-methyl-3-hydroxy-4-aminomethyl-5-hydroxymethylpyridine (see XII). The active aldehyde and amine were given the trivial names "pyridoxal" and "pyridoxaminc," respectively. The microbiological assays showed that pyridoxal was about 1400 times more active, and pyridoxamine about 10 times more active, than pyridoxine hydro- chloride for promoting the growth of L. casei. Pyridoxamine was about 8000 times more active, pyridoxal was about 5500 times more active, in promoting the growth of S. Jaecalis R, than was pyridoxine hydrochloride. The comparative activity of pyridoxal, pyridoxamine, and pyridoxine for Saccharomyces carlsbergensis was of the same order of magnitude. Evidence for the occurrence of pyridoxal and pyridoxamine in natural extracts was secured by Snell (47) by development of a differen- tial microbiological assay technique with the three organisms mentioned above and application of the assay to extracts of natural materials. Further evidence for the existence of pyridoxal and pyridoxamine in nature was secured by studying the effect of certain chemical treat- ments upon the "pyridoxine, pyridoxal, and pyridoxamine fractions" in comparison with the effect of these treatments upon the synthetic vitamins. Vitamin Be was originally considered to be a single pyridine derivative, pyridoxine. It may now be considered, as a result of these combined microbiological and organic chemical studies, as a name which designates a group of vitamins, i. e., the "vitamin Be group." Pyridoxal and pyridoxamine may occupy a place of equal or greater importance in this group as compared with that of pyri- doxine. In retrospect, it is interesting to note that, in the original isolation work, Keresztesy and Stevens (25) and Lcpkovsky (29) used rice bran as the source of their vitamin Be, while Kuhn and Wcndt (28) and Gyorgy (11) used yeast as their source of the crystalline vitamin Be (pyridoxine). Snell's microbiological differential assays showed (47) that a rice-bran concentrate contained far more pyri- doxine fraction than pyridoxal or pyridoxamine fractions, whereas 99 KARL FOLKERS yeast and liver extracts contained an excess of the pyridoxamine frac- tion, and animal assays showed no marked difference in the activity of the three substances. By assuming that the yeast supply used by Kuhn and Wendt and by Gyorgy contained an excess of pyridoxamine also, it is evident that at one or more of the steps in the isolation proc- ess the pyridoxamine was lost. Williams (65), in commenting on Gyorgy's communication (11) on isolation, noted that the yield of active substance in the first few steps of the concentration was only 10 to 30% of the original activity. If the isolation of vitamin Be from yeast had been guided by the results of microbiological assays with .S*. Jaecalis R instead of rats, one might predict today that it is quite probable that pyridoxamine would have been isolated instead of pyri- doxine and the pyridoxine present would have been lost at some step of the isolation procedure. In the field of new vitamins of unknown structure, several sub- stances are currently of great interest and papers concerning them are appearing frequently in the literature. Although studies of bio- logical activities in animals are being made, the use of microorganisms for tests of biological activities is resulting in the rapid accumulation of much valuable data on the differentiation of these substances. The following citations may exemplify the importance of the role of micro- organisms in the study of these new growth factors. Snell and Peterson (52) and Hutchings, Bohonos, and Peterson (19) have described the preparation and some properties of a con- centrate of a norite eluate factor from liver and yeast which resulted from a study of the nutrient requirements of L. casei and related lactic acid bacteria. Mitchell, Snell, and Williams (33) reported on the preparation of a highly purified nutrilite from spinach which they designated folic acid and defined as the material responsible for the growth stimulation of S. lactis R. Crystalline vitamin Bg from liver was highly active in growth activity for L. casei according to Pfiffner, Binkley, Bloom, Brown, Bird, Emmett, Hogan, and O'Dell (39). Stokstad (56) has described some properties of two crystalline prepara- tions, one from liver and one from yeast, which had somewhat different activities for promoting the growth of L. casei and S. lactis R. Keresz- tesy, Rickes, and Stokes (24) have reported the isolation of a different factor which was highly active for the growth of S. lactis R but rela- tively inactive for the growth of L. casei. Another new compound lOO VITAMIN RFSEARCH which is active lor the growth of L. casei was drscrihcfl hy llni. hings, Stokstad, Bohonos, and Slobodkin (20). The similarity of the biological activities of these several sub- stances suggests that they also may have a similarity in chemical structure. Concerning this chemical structure, Mitchell (32) pre- sented evidence which showed that the ultraviolet absorption spectrum of folic acid resembled that of xanthopterin (XIV). Addition of large amounts of thymine (XV) was found to substitute (biological pre- COH N CO /'\/\ /\ N G COH HN GCH, I II I I II H2NC C CH OG CH \/\/ \/ N N NH . (XIV) (XV) cursor?) for folic acid according to Snell and Mitchell (51). Thus, chemical relationships to the pterins and pyrimidincs are suggestive possibilities. Precise determination of the chemical structure of these several biological factors is probably necessary before the identity of any two or more of them, and before exact similarities in chemical structure, can be established with certainty. It is interesting that much of the micro- biological characterization of folic acid, vitamin B^, and related factors is being developed before the elucidation of their chemical structures. This situation is somewhat the reverse of that for the vitamin Be group since much of the chemical characterization of these factors was de- veloped before the clarification of the microbiological aspects. On the Synthesis of Vitamins The methods of the organic syntheses of the synthetic vitamins have been amply covered in appropriate review articles. These methods exemplify the adaptation and modification of classical organic laboratory reactions to the synthesis of the desired structure. The three asymmetric carbon atoms and the two fused five- membered saturated heterocyclic nuclei of biotin present interesting stereochemical features to the studies on the synthesis of this vitamin. There are two racemates which have cis forms of the rings and two lOI KARL FOLKERS racemates which have trans forms of the rings. The cis and trans relationship of the rings is shown by structures XVI and XVII, re- o 1 /\ ^fH NH G C 1 1 NH NH \ H H .• 1 CH2 CH(CH2)4C02H GH2 GH(CH2)4C02H (XVI) (XVII) 11 NH NH CH2 CH(CH2)4G02H (XVIII) spectively. Biotin is one of the eight stereoisomeric forms, dl- Biotin and related cis and trans forms were obtained by Harris, Mozingo, Wolf, Wilson, Arth, and Folkers (16). These other forms were des- ignated ^/-allobiotin and (//-epiallobiotin. Grussner, Bourquin, and Schnider (10) have described ,,>;„,;, K,,.;ii.,c NH CH diphtheria bacillus > CH CH CH2 CH(CH2)4C02H CH3 CH2(CH2)4C02H \s/ (XXI) (^^) 103 KARL FOLKERS of pimelic acid in promoting the biosynthesis of biotin. Cysteine or cystine, as sources of organic sulfur, enhanced the effect of pimelic acid. The studies by du Vigneaud, Dittmer, Hague, and Long (57) on the growth-stimulating effect of biotin for the diphtheria bacillus in the presence and absence of pimelic acid led to the interpretation that pimelic acid was being utilized as a precursor by the diphtheria bacillus for the biosynthesis of biotin. Experiments described by Dittmer, Melville, and du Vigneaud (3) on the activity of desthiobiotin (XXI) for stimulating the growth of S. cerevisiae showed that desthiobiotin disappeared from the incu- bating yeast cultures and was replaced by an equivalent amount of a substance possessing growth activity for L. casei. To these investi- gators, the most logical interpretation was that desthiobiotin (XXI) was transformed to biotin (XX) by the growing yeast cell. Further support for this interpretation was supplied by the experiments of Stokes and Gunness (54), which showed that, when extracts of yeast grown with desthiobiotin were treated with Raney nickel, the process (60) for converting biotin to desthiobiotin, the activity of the yeast- formed substance for L. casei was destroyed and activity for the yeast was retained. Avidin also neutralized the yeast-formed substance, as it does biotin. Another example of biosynthesis was found by Stokes, Keresz- tesy, and Foster (55), who reported that the S.L.R. factor was con- verted by S. lactis R into a substance which was active for the growth stimulation of L. casei. The possibility that alanine might be a biological precursor of vitamin Be was recognized two years ago by Snell and Guirard (49) when it was found that vitamin Be was not required for the growth of S. faecalis R if sufficient alanine was added to the medium. Subse- quent studies by Snell (48) showed that an enzymic digest of casein contained an unknown biological precursor which, together with dl- alanine, permitted the growth of L. casei in the absence of vitamin Be. Since fl?( — )alanine was active and /(4-)alanine was almost inactive, these data seem to be the first which indicate that the "unnatural" amino acids may be essential for normal metabolic processes. It is possible that future research on biological precursors and biosyntheses of the vitamins will make these substances available by new methods of startling simplicity. 104 VITAMIN RESEARCH References (1) Barger, G., Some Applications of Organic Chemistry to Biology and Medicine McGraw-Hill, New York, 1930, p. 62. (2) Brown, G. B., and du Vigneaud, V., J. Biol. Chem., 141, 85 (1941). (3) Dittmer, K., Melville, D. B., and du Vigneaud, V., Science, 99, 203 (1944). (4) Duschinsky, R., Dolan, L. A., Flower, D., and Rubin, S. A., Arch. Bio- chem., 6, 480 (1945). (5) Eakin, R. E., and Eakin, E. A., Science, 96, 187 (1942). (6) Emerson, G., J. Biol. Chem., 157, 127 (1945). (7) Emerson, O. H., Emerson, G. A., and Evans, H. M., Science, 83, 421 (1936). (8) Fernholz, E., J. Am. Chem. Soc, 59, 1154 (1937). (9) Fernholz, E., J. Am. Chem. Soc, 60, 700 (1938). (10) Grussner, A., Bourquin, J. P., and Schnider, O., Helv. Chim. Acta, 28, 517 (1945). (11) Gyorgy, P., J. Am. Chem. Soc, 60, 983 (1 938) . (12) Gyorgy, P., Ann. Rev. Biochem., 11, 310 (1942). (13) Gyorgy, P., Kuhn, R., and Lederer, E., J. Biol. Chem., 131, 745 (1939). (14) Hahn, G., and Schales, O., Ber., 68, 24 (1935). (15) Harris, S. A., Heyl, D., and Folkers, K., J. Biol. Chem., 154, 315 (1944); J. Am. Chem. Soc, 66, 2088 (1944). (16) Harris, S. A., Mozingo, R., Wolf, D. E., Wilson, A. N., Arth, G. E., and Folkers, K., J. Am. Chem. Soc, 66, 1800 (1944). (17) Hofmann, K., J. Am. Chem. Soc, 67 y 694 (1945). (18) Hofmann, K., J. Am. Chem. Soc, 67, 1459 (1945). (19) Hutchings, B. L., Bohonos, N., and Peterson, W. H., J. Biol. Chem., 141,521 (1941). (20) Hutchings, B. L., Stokstad, E. L. R., Bohonos, N., and Slobodkin, N. H., Science, 99, 2)1\ (1944). (21) Jansen, B. C. P., and Donath, W. F., Mededeel. Dienst. Volksgezondheid Nederland-Indie, 16, 186 (1926). (22) Karrer, P., Fritzsche, H., Ringier, B. H., and Salomon, H., Helv. Chim. Acta, 21, 820 (1938). (23) Karrer, P., Salomon, H., and Fritzsche, H., Ilelv. Chim. Acta, 21, 309 (1938). (24) Keresztesy, J. C., Rickes, E. R., and Stokes, J. L., Science, 97, 465 (1943). (25) Keresztesy, J. G., and Stevens, J. R., Pruc. Soc. Exptl. Biol. M,;l., 38, 64 (1938); Stiller, E. T., Keresztesy, J. C., and Stevens, J. R., J. Am. Chem. Soc, 61, 1237 (1939). 105 KARL FOLKERS (26) Kogl, F., J. Soc. Chem. Ind., 57, 49 (1938). (27) K5gl, F., and Borg, W. A. J., Z- physiol. Chem., 281, 65 (1944). (28) Kuhn, R., and Wendt, G., Ber., 71, 780, 1118 (1938). (29) Lepkovsky, S., Science, 87, 169 (1938); J. Biol. Chem., 124, 123 (1938). (30) Major, R. T., Chem. Eng. News, 20, 517 (1942). (31) Melville, D. B., in Vitamins and Hormones. Vol. II, Academic Press, New York, 1944, p. 29. (32) Mitchell, H. K., J. Am. Chem. Soc, 66, 274 (1944). (33) Mitchell, H. K., Snell, E. E., and Williams, R. J., J. Am. Chem. Soc, 63, 2284 (1941); 66, 267 (1944). (34) Mitchell, H. K., Weinstock, H. H., Jr., Snell, E. E., Stanbery, S. R., and Williams, R. J., J. Am. Chem. Soc, 62, 1776 (1940). (35) Mozingo, R., Wolf, D. E., Harris, S. A., and Folkers, K., J. Am. Chem. Soc, 65, 1013 (1943). (36) Mueller, J. H., J. Biol. Chem., 119, 121 (1937). (37) Oden, J. W., Oden, L. H., and Sebrell, W. H., U. S. Pub. Health Repts., 54, 790 (1939). (38) Ott, W. H., J. Biol. Chem., 157, 131 (1945). (39) Pfiflfner, J. J., Biiikley, S. B., Bloom, E. S., Brown, R. A., Bird, O. D., Emmett, A. D., Hogan, A. G., and O'Dell, B. L., Science, 97 y 404 (1943). (40) Pilgrim, F. J., Axelrod, A. E., Winnick, T., and Hofmann, K., Science, 102,35 (1945). (41) Robinson, R., J. Chem. Soc, 111, 876 (1917). (42) Schopf, C., Ann., 497, 1 (1932). (43) Sebrell, W. H., and Buder, R. E., U. S. Pub. Health Repts., 53, 2282 (1938). (44) Smith, L. I., Chem. Revs., 27, 287 (1940). (45) Snell, E. E., J. Am. Chem. Soc, 66, 2082 (1944). (46) Snell, E. E., J. Biol. Chem., 154, 313 (1944); 157, 475 (1945). (47) Snell, E. E., J. Biol. Chem., 157, 491 (1945). (48) Snell, E. E., J. Biol. Chem., 158, 497 (1945). (49) Snell, E. E., and Guirard, B. M., Proc Natl. Acad. Sci. U. S., 29, 66 (1943). (50) Snell, E. E., Guirard, B. M., and Williams, R. J., J. Biol. Chem., 143, 519 (1942). (51) Snell, E. E., and Mitchell, H. K., Proc Natl. Acad. Sci. U. S., 27, 1-7 (1941). (52) Snell, E. E., and Peterson, W. H., J. Bad., 39, 273 (1940). (53) Stiller, E. T., Keresztesy, J. C., and Finkelstein, J., J. Am. Chem. Soc, 62, 1779 (1940). (54) Stokes, J. L., and Gunness, M., J. Biol. Chem., 157, 121 (1945). io6 VITAMIN RESEARCH (55) Stokes, J. L., Keresztesy, J. C, and Foster, J. VV., Science, 100, 522 (1944). (56) Stokstad, E. L. R., J. Biol. Chem., 149, 573 (1943). (57) du Vigneaud, V., Dittmer, K., Hague, E., and Long, B., Science, 96, 186 (1942). (58) du Vigneaud, V., Hofmann, K., and Melville, D. B., J. Am. Cliem Soc 64, 188 (1942). (59) du Vigneaud, V., Hofmann, K., Melville, D. B., and Gyorgy, P., J. Biol. Chem., 140, 643 (1941). (60) du Vigneaud, V., MelvUle, D. B., Folkers, K., Wolf, D. E., Mozingo, R., Keresztesy, J. C, and Harris, S. A., J. Biol. Chem., 146, 475 (1942). (61) Weinstock, H. H., Jr., Mitchell, H. K., Pratt, E. F., and Williams, R. J., J. Am. Chem. Soc, 61, 1421 (1939). (62) Williams, R. J., Science, 93, 412 (1941). (63) WUliams, R. J., Science, 95, 340 (1942). (64) WUliams, R. J., Ann. Rev. Biochem., 12, 309 (1943). (65) Wiliams, R. J., Ann. Rev. Biochem., 12, 331 (1943). (66) Williams, R. J., Mitchell, H. K., Weinstock, H. H., Jr., and Snell, E. E., J. Am. Chem. Soc, 62, 1784 (1940). (67) Williams, R. J., Truesdail, J. H., Weinstock, H. H., Jr., Rohrmann, E., Lyman, C. M., and McBurney, C. H., J. Am. Chem. Soc, 60, 2719 (1938). (68) Williams, R. J., Weinstock, H. H., Jr., and Mitchell, H. K., Abstracts, 96th Meeting, Am. Chem. Soc, Sept., 1938, Division of Organic Chemistry, p. 34. (69) Williams, R. R., Science, 95, 335 (1942). (70) Williams, R. R., Waterman, R. E., and Keresztesy, J. C, J. Am. Chem. Soc, 56, 1187 (1934). (71) Williams, R. R., Waterman, R. E., Keresztesy, J. C, and Buchman, E. R., J. Am. Chem. Soc, 57, 536 (1935). (72) Woolley, D. W., Waisman, H. A., and Elvehjcm, C. A., J. Am. Chem. Soc, 61, 977 (1939); J. Biol. Chem., 129, 673 (1939). 107 8 QUANTITATIVE ANALYSTS IN BIOGHEIVIISTIIY DONALD D. VAN SLYKE, member of the rockefeller institute FOR medical research, NEW YORK; VVTLLARD GIBBS MEDALIST ^HE BIOCHEMISTRY of today is based to a large extent -*• on quantitative analyses by means of micro methods. It lias not only applied the methods developed in laboratories of organic and inorganic chemistry, but has also rapidly developed new pro- cedures to meet the demands of its expanding range of research . While Pregl's system of elementary organic microanalysis, published in 1911, was gaining application, Folin and Wu in 1919 and Bang in 1916 pub- lished quite different systems adapted to blood analyses. The ultra in microanalyses is represented by the methods of capillary colorimetry developed by A. N. Richards and his collaborators for the analysis of glomerular urine, and the extraordinary combination of physical and chemical procedures applied by Linderstr0m-Lang in his studies of the enzymes of cells. In a brief survey of the field it will be possible only to mention some of the different types of analytical procedure that have been applied in biochemistry, with examples of a few applications, and refer- ences to reviews in which descriptions and bibliographies can be found. Gravinielric A nalysis The appearance in 1911 of the Kuhlman micro balance, and of Pregl's system of micro elementary analyses decreased the size of 109 D. D. VAN SLYKE samples usually taken for analyses from 100-150 mg. to 4 or 5 mg., and opened a new epoch, not only in organic analysis, but also in organic chemistry. For it became possible to solve problems when amounts of material were available that would have been inadequate even for preliminary analyses by the old macro methods. It is ques- tionable, for example, that the rapid development of knowledge con- cerning the structure of the vitamins would have been possible without these micro methods. The techniques of micro analysis developed in Austria by Pregl and by Emich (1) were brought to this country, es- pecially by J. B. and V. Niederl (3), and are available in a recent volume by these authors. For ultra micro gravimetric work, Lowry (2) has recently described a simple quartz fiber balance which will weigh 200 gammas to 0.03 gamma. Volumetric Analysis Bang's work (4) introduced micro titrations into biochemistry, utilizing the principle of keeping the volume of titrated liquid small, to maintain the sharpness ^of the end point. Rehberg (9) followed shortly with his micro burette, in which a thread of standard solution in a calibrated capillary of about 1 mm. bore is expelled into the titrated solution by mercury moved by a plunger advanced by turning a steel screw. With this burette, the solution could be measured to within ±0.1 cu. mm., so that 50 cu. mm. was ample for a good titration. Linderstr0m-Lang (7) moved the error down to ±0.01 cu. mm. He used still narrower capillaries, and stirred the titrated solution in a minute glass thimble by a magnetic stirrer consisting of a bit of iron, enclosed in a glass droplet, which was lifted and lowered in the titrated solution by the automatic opening and closing of the circuit to a mag- net. Scholander (10,11) has applied the metal micrometer screw that is readily available because of its general use in accurate industrial mechanical work. The column of standard solution in a fine capillary is moved, as in the Rehberg burette, by pressure of a mercury column advanced by a screw and plunger; but in the Scholander burette the extent to which the screw is turned serves as a measure of the volume of solution delivered, and calibration of the capillary is not necessary. A limitation of the Rehberg and Scholander burettes is that they cannot be used to deliver solutions which react with mercury. For no QUANTITATIVE ANALYSIS such solutions, Longwell and Hill (8) have modified the Rehberg burette by introducing an elastic rubber diaphragm between the mercury and the solution. Clark, Levitan, Gleason, and Greenberg (5) have applied Scholander's micrometer principle, but employ the micrometer screw to push air (instead of mercury) from a hypodermic syringe into a capillary which delivers the solution. The general principles of volumetric microanalyses have been clearly elucidated by Conway (6). Gasometric Analysis One of the oldest quantitative analyses in biochemistry is the determination of urea by measurement of the nitrogen gas liberated by reaction with alkaline bromine solution. It exemplifies procedures in which a substance is measured by the amount of gas that it liberates when it reacts with properly chosen reagents. In such analyses, the measurement is based, as in gravimetric methods, on direct observation of the amount of substance obtained, independent of comparison with standard solutions, such as are required in titration and colorimetry. Combined with this independence are the advantages of a quick meas- urement and easy adaptation to micro quantities. Historically, micro gasometric procedures were introduced into biochemistry for deter- mination of the blood gases, and were then adapted to more general analyses. The first of such procedures was the blood gas analysis of Bar- croft and Haldane (12) in which the oxygen liberated from blood by ferricyanide, or the carbon dioxide liberated by acid, was measured by the gas displaced into a capillary tube. The procedure requires accurate temperature control and constant shaking until equilibrium between dissolved and supernatant gases is reached. The method was elaborated by Warburg (19), and has been used for a great variety of purposes by Warburg and others, particularly in following the course of enzymic reactions by measurement of the oxygen absorbed or the carbon dioxide evolved. The ease with which the course of a reaction can be followed by observing the increase in gas volume particularly adapts the procedure to the observation of comparative reaction veloci- ties (13). A recent refinement of the apparatus, and the principles of its use, are described by Summerson (15). Ill D. D. VAN SLYKE The manometric apparatus of Van Slyke and Neill (18), like the Barcroft-Haldane apparatus, was first developed for determination of the blood gases, and its use then spread to other micro analyses, including the determination of urea, reducing and fermentable sugars, the ammonia yielded by Kjeldahl digestions, amino nitrogen by measurement of the nitrogen yielded by reaction with nitrous acid (18), free alpha-amino acids by measurement of the carbon dioxide yielded by reaction with ninhydrin [RCH(NH2)COOH -> RCHO + CO2 + NH3] (16), organic carbon by measurement of the carbon dioxide evolved by a wet combustion completed in two minutes (17), and various other determinations. In these procedures, the gases are either evolved in, or transferred to, a 50-cc. chamber, provided at the top with small bulbs for measuring 0.5 and 2.0 cc. of gas, and connected at the bottom with a mercury manometer. The evolved gas is brought to 0.5 or 2.0 cc. volume, and its pressure is read on the manometer. In a mixture of gases, each gas can be measured separately by measur- ing the pressure before and after the absorption of each gas by intro- duction of a proper reagent. The carbon combustion method per- mits micro determination of any organic substance, such as the blood fats, that can be isolated by extraction with volatile solvents; the com- bustion also provides a micro measurement of any substance that can be isolated as a carbon-containing precipitate, e. g., sulfate as benzidine sulfate, magnesium as hydroxyquinolate, phosphorus as strychnine phosphomolybdate (14). While the Barcroft-Haldane-Warburg apparatus is adapted to following the course of time reactions, the Van Slyke-Neill apparatus is fitted for quick determination of the total amounts of gases evolved by rapid quantitative reactions. Hence the Haldane apparatus has found its chief application in following enzymic time reactions, while the Van Slyke-Neill apparatus is the one usually employed in quantita- tive micro determinations of specific substances. Photometric Analysis Under the chief initial stimulus of Folin and of S. R. Benedict, during the past forty years chromogenic reactions have been developed for estimating numerous biological substances by producing colored products from them (25). In some cases the colored products are I 12 QUANTITATIVF, ANALYSIS defined; in other cases neither ihe reactions nor the colored products are defined with certainty; hut emiiirical conditions ha\e heen fixed which relate the color quantitatively to the amount of the substance under analysis. Thus Folin (21) used the color produced by Nessler's reagent as the means for quantitative estimation of ammonia, and hence of nitrogen, through the ammonia obtained by Kjeldahl diges- tion, and of urea through the ammonia obtained by urea hydrolysis. The constitution of the colored compound of ammonia and potassium mercuri-iodide is still under dispute, but the colorimetric results are accurate. Sugars reducing Cu++ to Cu''" were determined by Folin and VVu (21) by letting the cuprous ion thus formed act on a molybdatc solution, with reduction of the colorless hexavalent molybdenum to a lower valence, which shows an intense blue color; the reactions do not appear to be stoichiometric, but quantitative relations can be obtained between colored molybdenum products and the initial sugar. For almost every substance of interest in quantitative biochemical analysis, chromogenic reactions have been devised which can be used for more or less accurate estimation. As a rule these procedures are rapid, and are adapted to minute amounts of material. During the first years of the colorimetric epoch, the instrument in general use was the familiar Duboscq colorimeter, in which the depths of colored solution layers in two parallel columns, one of the unknown solution and the other of a standard, are varied until the two fields viewed with the eye appear equal. The simplicity and versatility of the Duboscq colorimeter assisted greatly in the rapid adoption of colorimetric procedures. Photometers, in which the percentage transmission of light could be measured without standard solutions for direct comparison, were known long before this period, but were too complicated and ex- pensive for ordinary routine in biochemical laboratories. During the past two decades, however, photometers have been progressively made more adaptable to such routine, and have been gradually displacing colorimeters of the Duboscq type. In the photometer, the concentra- tion of light-absorbing solute is related to the optical density according to the simple linear formula of Beer's law (20,22,23): C = kl) (1) where C is the concentration and D is the optical density (or extinction). D. D. VAN SLYKE D is the logarithm of l/T, T being the fraction of Ught transmitted by the substance measured. The value of k for a given colored solute varies with the wave length of light. Hence Beer's law holds exactly only for monochromatic light; and the accuracy with which the formula applies to a given photometer depends partly on how narrow a spectral band can be given by the analyzing device interposed between the source of light and the solution, a limitation under which the Duboscq colorimeter does not suffer. Some solutes do not exactly follow Beer's law, even with the narrowest spectral bands. Most solutions, however, do follow the law over the concentration ranges used for analysis, and with the spectral bands provided by the instruments now available for routine analytical work. Solutions for which Beer's law does not hold can be analyzed by using empirical calculation curves of optical density vs. concentration. The photometer has several advantages over the Duboscq colorimeter. The validity of equation (1) makes it possible to measure the concentration of one colored solute in the presence of others, since the total optical density is additive: D = CiAx + C,/h. . . (2) Hence, if the medium in which the concentration of a solute is to be measured is itself colored or turbid, the increase in optical density due to the presence of the specific solute can be used as a measure of its concentration. Correction for nonspecific color is much less simple in a Duboscq colorimeter. Because the k value for each solute changes with the wave length, it is possible to determine two colored solutes in the same solution by measuring the optical densities at two different wave lengths. Two densities. Da and Z)j, are thus measured: Da = C,/k^ + C^/k^ (2a) D, = C,/k, + C2A4 (2b) If hi, kz, ks, and ki are known, Ci and C2 can be calculated by simul- taneous equations from the observed Da and i)j. The concentration of a turbid suspension can be estimated from its optical density, in the same way as the concentration of a colored solute. 1 14 QUANTITATIVE ANALYSIS The eye strain and subjective error accompanying the use of a visual instrument are obviated in most of the modern photometers by using photocells to measure the intensity of the transmitted light. The rapidity with which a series of observations can be carried out is also much greater when the eye is replaced by the photocell. A further advantage of the photometer is that, with photocell measuring devices, it can be used with light waves extending into the ranges of the ultraviolet and infrared, broadening the range of ac- cessible analyses. In the ultra micro colorimetric methods which Richards and his collaborators (24) developed for analyses of samples of 1 cu. mm. of glomerular filtrate, the sample is drawn into a small glass capillary in which it is mixed with chromogenic reagent, and the color is esti- mated by comparing the capillary under a microscope with a series of standards similarly prepared. The comparison with graded standards is a return to simplest first principles, but the technique of the applica- tion gave results to 1%. Fluorimetric Analysis The diffuse fluorescent light developed by passing light rays into solutions of fluorescent substances (28) can be measured by the same visual or electrical means employed in colorimeters and photom- eters, and serves in some cases, such as determination of riboflavin (27), quinine, atabrine (26), and related compounds, to measure substances in more dilute solutions than can be handled with a photom- eter. In fluorimetry, the concentration of fluorescent substance is directly proportional to the intensity of the light measured, instead of being inversely proportional to the log of the transmission, as in pho- tometry. Po larograph ic A nalys is This procedure, which has gained rapid utility during the past few years, was introduced by Heyrovsky in Prague in 1925, and has been applied to a multiplicity of analyses of substances, both organic and inorganic. It is based on measurement of the amperages obtained at observed voltages applied to solutions of electroreduciblc or electro- D. D. VAN SLYKE oxidizable substances in a cell in which one electrode consists of mercury falling from a fine capillary. When a current at gradually increasing voltage is passed to a small mercury electrode through a solution containing a solute capable of giving or receiving electrons ("redox solute") at a given potential, relatively little current is obtained until the decomposition potential of the redox solute is reached. Then the current rapidly rises, with further increase in voltage until a plateau is reached at which the redox solute is providing its maximal flow of electrons. This flow, and the resultant current, are proportional to the concentration of the redox solute, which determines the rate at which its molecules diff'use to the electrode and discharge or receive electrons. Since diff'usion is a proc- ess independent of the voltage, increase of voltage above that at which electrolytic decomposition of the redox solute equals its rate of diff'usion to the mercury electrode does not further increase the flow of electricity; the concentration of the active solute thus forms a bottleneck which limits the current. The curve of current vs. voltage then reaches a plateau, the height of which, in milliamperes, is proportional to the concentration of active redox solute (32-34). Maintenance of the proportionality requires a continually renewed surface of the mercury electrode; both renewal of the surface and setting its area at a small size are obtained by using mercury dropping from a fine capillary, of about 0.03 mm. diameter, as the electrode; a drop is delivered about once in two to four seconds. The current fluctuates somewhat as each drop of mercury expands and falls, but the average current, i, in microamperes is given by the "Ilkovic equation": i = 605 n D"'' C m'' t"'' (3) C is the millimoles of redox solute per liter, n is the number of electrons exchanged per molecule of redox solute in the electrolytic decomposi- tion, m is the weight of mercury flowing from the capillary per second, and t is the time required for formation of one drop of mercury. The procedure is adapted to analysis of highly dilute solutions, 0.001 molar and lower concentrations. Measurement of the current in such a system provides a measure of the concentration of the redox solute. Furthermore, if two different redox solutes are present with diff"erent decomposition potentials, they ii6 QUANTITATIVE ANALYSIS can be determined one after another by using voltages related (.. ilxir respective decomposition potentials. The setup also can be used lor electrometrir titrations, if .1 reagent which reacts with the redox solute is added while the current is measured at proper voltage, a drop in current will accompany the disappearance of the redox solute, and the end jjoint will be indicated when the current has fallen to a residual value, representing con- ductance by factors other than the redox solute. For literature, theoretical discussion, description of the different types of analysis to which the procedure has already been applied, and the precautions that must be observed, the reader is referred to the bibliography (35), in particular to KolthoflT (32,33) and Muller (34). The various inorganic cations and anions are determinable; also re- ducible organic compounds, such as aldehydes, ketones, and nitro compounds. Eisenbrand and Picher (31) found that the sex hormones with the 0=C — C=C — group are reducible at the mercury electrode and can be determined polarographically: these hormones include testosterone, progesterone, and desoxycorticosterone, but not andro- sterone nor dehydroandrosterone. Application to solve a hitherto difficult biochemical problem is illustrated by the work of Berggren (30) and of Beecher et al. (29) in determining the oxygen tension of arterial blood plasma and other body fluids. These authors also describe their apparatus in detail. Spectrograph ic A nalys is Measurement of the intensity of light of characteristic wave length emitted by incandescent elements has been long used in both qualitative and quantitative analysis, and in special problems for quantitative or semiquantitative estimation of minute amounts of specific elements. Lundegardh (39) in 1929 applied the principle in a manner which makes it applicable to micro determination of mineral bases in biological fluids. The solution is sprayed from an atomizer at a constant rate, and the stream of air with suspended fluid is mixed with acetylene, which is burned, heating the mineral bases in the suspension to such a temperature that they emit their charac- teristic light bands, the intensity of which is measured by an electro- photometer (37). The procedure is reviewed by Ells (38) and by D. D. VAN SLYKE Cholak and Hubbard (36). The latter describe the appHcation of the procedure to determination of minute amounts of cadmium in blood and urine, and compare it with polarographic and photometric pro- cedures for this purpose. An apparatus devised by the American Cyanamid Company, not yet in general production, determines potassium in serum in a few minutes with an error not over =*=5%. Micro Diffusion Analysis Conway (41) has devised a simple chamber for the determi- nation, primarily, of ammonia, but applicable also to estimation of other volatile substances that can be set free by quantitative reactions and transferred by diffusion at atmospheric pressure to absorbing solutions in which the diffused substance can be measured, by titration, photometry, or otherwise. The apparatus consists of a flat, cylin- drical dish, of 60-mm. diameter and 10 mm. high (inner measure- ments), from the inner bottom of which rises a ring of 33-mm. diameter and 5 mm. high. The top of the dish is ground accurately flat, so that when lubricated with vaseline or other proper material it can be closed gas tight by a flat glass cover. The chamber consists, therefore, of an outer ring, about 60 mm. wide, surrounding an inner low cylinder; when the chamber is covered a free space of 5 mm. is left open between the covering plate and the wall about the inner cylinder, permitting free diffusion of volatile substances from the outer compartment to the inner. To determine ammonia, the solution containing it is alka- linized in the outer compartment, and acid is placed in the inner compartment. In one or more hours, depending on temperature and the other conditions, the ammonia from the former diffuses through the air space of the chamber into the acid, where it can be measured in amounts of a few micrograms. Conway applied this procedure to determination of the minute amounts of ammonia in blood, to esti- mation of the ammonia formed by micro Kjeldahl nitrogen digestion, and of the ammonia formed by decomposition of urea with urease; also to chloride and bromide, which were oxidized to chlorine and bromine and diffused into potassium iodide solution for iodometric titration. Conway also applied the diffusion chamber to the determi- nation of carbon dioxide, which was caused to diffuse into a barium hydroxide solution, where the excess alkali was titrated. Borsook (40) ii8 QUANTITATIVE ANALYSIS developed applications of the ammonia procedures. VVinnick (42) applied the diffusion apparatus to the determination of: alcohol, with chromic acid in the inner chamber; lactic acid, which was oxi- dized in the outer chamber to acetaldehyde, the latter diffusing to a bisulfite solution in the inner chamber; acetone, which diffused to bisulfite; threonine, which was oxidized by periodate to acetaldehyde in the outer chamber, with diffusion of the aldehyde to bisulfite. References GRAVIMETRIC ANALYSIS (1) Emich, F., Microchemical Laboratory Manual. Trans, by F. Schneider. Wiley, New York, 1932. (2) Lowry, O. H., "A quartz fiber balance," J. Biol. C/iem., 140, 183 (1941). "A simple quartz torsion balance," ibid., 152, 293 (1944). (3) Niederl, J. B., and Niederl, V., Micromethods oj Quantitative Organic Analysis. Wiley, New York, 1942. VOLUMETRIC ANALYSIS (4) Bang, I., Melhoden zur Mikrobestimmung einiger Blutbestandteile. Berg- mann, Wiesbaden, 1916. (5) Clark, W. G., Levitan, N. I., Gleason, D. F., and Greenberg, G., "Ti- trimetric microdetermination of chloride, sodium, and potassium in a single tissue or blood sample," J. Biol. Chem., 145, 85 (1942). (6) Conway, E. S., Micro-diffusion Analysis and Volumetric Error. Van Nostrand, New York, 1 942. (7) Linderstr0m-Lang, K., "Distribution of enzymes in tissues and cells," Harvey Lectures, 34, 214 (1939). (8) Longwell, B., and Hill, R. M., "A modified Rehberg burette for use with titrating solutions which react with mercury," J. Biol. Chem., 112, 319 (1935). (9) Rehberg, P. B., "A method of microtitration," Biochem. J., 19, 270 (1925). (10) Scholander, P. F., "Microburette," Science, 95, 177 (1942). (11) Scholander, P. F., and Edwards, G. A., and Irving, L., "Improved microburette," J. Biol. Chem., 148, 495 (1943). GASOMETRIC ANALYSIS (12) Barcroft, J., and Haldane, J. S., "A method of estimating the oxygen and carbonic acid in small quantities of blood," J. Physiol., 28, 232 (1902). D. D. VAN SLYKE (13) Dixon, M., Manometric Methods. Cambridge Univ. Press, London, 1934. (14) Hoagland, C. L., "Microdetermination of sulfate and phosphate by manometric combustion of their organic precipitates," J. Biol. Chem., 136, 543 (1940). "Micro manometric determination of magnesium," ibid., 553. (15) Summerson, W. H., "A combination simple manometer and constant differential manometer for studies in metabolism," J. Biol. Chem., 131, 579 (1939). (16) Van Slyke, D. D., Dillon, R. T., MacFadyen, D. A., and Hamil- ton, P. B., "Gasometric determination of carboxyl groups in free amino acids," J. Biol. Chem., 141,627 (1941). Hamilton, P. B., and Van Slyke, D. D., "The gasometric determination of free amino acids in blood filtrates by the ninhydrin-carbon dioxide method," tZ»2a?., 150,231 (1943). (17) Van Slyke, D. D., and Folch, J., "Manometric carbon determina- tion," J. Biol. Chem., 136, 509 (1940). (18) Van Slyke, D. D., and Neill, J. M., "The determination of gases in blood and other solutions by vacuum extraction and manometric meas- urement," J. Biol. Chem., 61, 523 (1924); 83, 449 (1929). "Applications to other analyses," in Peters, J. P., and Van Slyke, D. D., Quantitative Clinical Chemistry. Methods. Williams & Wilkins, Baltimore, 1932; rev., 1943. (19) Warburg, O., "Verbesserte Methode zur Messung der Atmung und Glykolyse," Biochem. Z-, 152, 51 (1924). PHOTOMETRIC ANALYSIS (20) Ashley, S. E. Q., "Spectrophotometric methods in modern analytical chemistry," Ind. Eng. Chem., Anal. Ed., 11, 72 (1939). (21) Folin, O., and Wu, H., "A system of blood analysis," J. Biol. Chem., 38, 81 (1919). (22) Hamilton, R. H., "Photoelectric photometry. An analysis of errors at high and low absorption," Ind. Eng. Chem., Anal. Ed., 16, 123 (1944). (23) Miiller, Ralph H., "Photoelectric methods in analytical chemistry," Ind. Eng. Chem., Anal. Ed., 7, 223 (1935); 11, 1 (1939). (24) Richards, A. N., Bordley, J., 3rd, and Walker, A. M., J. Biol. Chem., 101, 179, 193, 223, 229 (1933). (25) Snell, F. D., and Snell, C. T., Colorimetric Methods oj Analysis, including Some Turbidimetric and Nephelometric Methods. 2nd ed., 2 vols., Van Nos- trand, New York, 1936-1937. FLUORIMETRIG ANALYSIS (26) Brodie, B. B., and Udenfriend, S., "The estimation of atabrine in bio- logical fluids," J. Biol. Chem., 151, 299 (1943). I20 QUANTITATIVE ANALYSIS (27) Hand, D. B., "Determination of riboflavine in milk by photoelectric fluorescence measurements," hid. Eng. Chem., Anal. Ed., II, 306 (1939). (28) Kavanagh, F., "New photoelectric fluorimeter and some applications," Ind. Eng. Chem., Anal. Ed., 13, 108 (1941). POLAROGRAPHIC ANALYSIS (29) Beecher, H. K., Follansbee, R., Murphy, A. J., and Craig, F. N., "Deter- mination of the o.xygen content of small quantities of body fluids by polaro- graphic analysis," J. Biol. Chem., 146, 197 (1942). (30) Berggren, S. M., "The oxygen deficit of arterial blood caused by non-ventilating parts of the lung," Acta Physiol. Scand. SuppL, 4, XI (1942). (31) Eisenbrand, J., and Picher, H., "t)bcr den polarographischen Nachweis von biologisch vvichtigen Ketonen der Steringruppe," <;. physiol. Chem , 260,83 (1939). (32) Kolthoff, I. M., "Factors to be considered in quantitative polarog- raphy," Ind. Eng. Chem., Anal. Ed., 14, 195 (1942). (33) Kolthoff", I. M., and Lingane, J. J., Polarography. Interscience, New York, 1941. (34) Miiller, O. H., The Polarographic Method oj Analysis. J. Chem. Educa- tion, Easton, 1941. (35) Sand, H. J. S., Bibliography of the Dropping-Mercury Electrode. Leeds & Northrup, Philadelphia, 1941. SPECTROGRAPHIC ANALYSIS (36) Cholak, J., and Hubbard, D. M., "Spectrochemical analysis with the air-acetylene flame," Ind. Eng. Chem., Anal. Ed., 16, 728 (1944). (37) Churchill, J. R., "Techniques of quantitative spectrographic analysis," Ind. Eng. Chem., Anal. Ed., 16, 653 (1944). (38) Ells, V. R., "The Lundegardh flame method of spectrographic analy- sis," J. Optical Soc. Am., 31, 534 (1941). (39) Lundegardh, H., Die Quantitative Spektralanalyse der Elemente. Fischer, Jena. Part I, 1929; Part II, 1934. MICRO DIFFUSION ANALYSIS (40) Borsook, H., "Micromethods for determination of ammonia, urea, and total nitrogen," J. Biol. Chem., 110, 481 (1935). (41) Conway, E. J., Micro-diffusion Analysis and Volumetric Error. Van Nos- trand, New York, 1940. (42) Winnick, T., "Micro-diffusion methods. Alcohol," Ind. Eng. Chem., Anal. Ed., 14, 523 (1942). "Acetone," J. Biol. Chem., 141,115 (1941). "Lactic acid," ibid., 142, 451 (1942). "Threonine," ibid., 142, 461 (l'M2). 121 ENZYMIC HYDROLYSIS AND SYNTHESIS OF PEPTIDE BONDS JOSEPH S. FRUTON, associate professor of physiological CHEMISTRY, YALE UNIVERSITY; LILLY AWARD IN BIOLOGICAL CHEMISTRY Specificity of Proteolytic Enzymes THE ACTION of proteolytic enzymes on peptide linkages involves a high degree of specificity. No proteolytic enzyme acts on peptide bonds indiscriminately, and each enzyme hydrolyzes only such peptide bonds as are present in the substrate in a certain structural setting. Thus, the nature of the requisite struc- tural attributes of the substrate is an expression of the specificity of the enzyme which hydrolyzes the substrate. In recent years, much attention has been devoted to the determination of those structural elements in the substrate molecule which are essential for the action of various proteolytic enzymes. These studies have permitted the formu- lation of several hypotheses concerning the specific action of the pro- teolytic enzymes. Modern theories concerning the specificity of proteolytic enzymes are based on the assumption, made by von Euler and Joseph- sohn (17) in their "dual-affinity" theory, that ereptic peptidase— it was not recognized at that time that "erepsin" represents a mixture of many peptidases — combines with two atomic groupings of the sub- strate molecule. Subsequent work of Balls and Kohler (2) presented 123 J. S. FRUTON further evidence for the "dual-affinity" hypothesis. More recently, the finding of synthetic substrates for the protein-spUtting enzymes (pepsin, trypsin, chymotrypsin, papain, etc.) has led to the extension of this hypothesis to all representative proteolytic enzymes (10). Of the two essential points in the substrate, one, of necessity, must be the sensitive CO — NH group or some part thereof. The other requisite point of contact lies in the "backbone"* of the substrate and varies with the nature of the enzyme. The nature of this second group and its position in the backbone relative to the sensitive peptide bond provide the basis for the classification of proteolytic enzymes into four groups as given in Table I (the requisite groups are italicized and the sensitive peptide linkage is indicated by means of a dotted line). Table I Classification of Proteolytic Enzymes Group No. Linkage attacked Classification R I NHiCUCO^NH- Aminojjeptidases R 1 Exopeptidases II ■ CO^NHCHCOOH Carboxypeptidases R III . CO—NH- CH • CO^NH- Proteinases R 1 V Endopeptidases IV • CO^NH- CH • CO—NH- Proteinases In groups I and II are included the enzymes restricted in their action to peptide bonds at the end of a peptide chain. The peptidases belonging to group I (aminopeptidases) selectively attack the chain at the peptide linkage adjacent to the amino end of the chain while the peptidases of group II (carboxypeptidases) attack the chain at the * In order to describe the structural setting of a peptide bond, it is desirable to speak of a "backbone" of the substrate, i. e., the sequence of — NH — CH — CO — groupings linked through peptide bonds, and the "side chains," :. e., the groups attached to the CH groups of the backbone. 124 HYDROLYSIS OF PEPTIDE BONDS peptide linkage adjacent to the carboxyl end of the chain. The amino- and carboxypeptidases cannot spUt Unkagcs that are centrally located in the peptide chain; for this reason they arc referred to as exopeptidases. All the protein-splitting enzymes whose backbone requirements have been determined belong to group III. These enzymes are capable of hydrolyzing central peptide bonds and, therefore, are re- ferred to as endopeptidases. They were found lo require, in their substrates, a peptide bond in close proximity to the carbonyl group of the peptide bond which is hydrolyzed by the enzyme. Although no known proteolytic enzyme has been identified as belonging to group IV, the suggestion was made recently that an enzyme which hydrolyzes leucylglycylglycine and which is found in intestinal mucosa may belong to this group (31). The presence of the indispensable groups in the backbone of a substrate in itself is insufficient to render the substrate susceptible to the action of an enzyme. It has been found that each of the proteolytic enzymes tested thus far also requires the presence, in the substrate, of a certain type of side chain (R) in a precisely defined location. In the second column of Table II, the recjuired location of side chain R is indicated for each of the enzymes mentioned; and the chemical nature of these R groups is given in the third column. The enzymes which have been listed require in their substrates one of the following side chains: isobutyl as in leucine, benzyl or /*-hydroxybenzyl as in phenyl- alanine or tyrosine, aminobutyl or guanidopropyl as in lysine or argi- nine. The side chains mentioned here represent only a few of those which jut out from the peptide chain of proteins. It will be a task of the future to determine precisely the specificity of proteolytic enzymes which require, in their substrates, the side chains of amino acids such as glycine, glutamic acid, histidine, tryptophane, etc. Homos pecific Proleolytic Enzymes In the course of the systematic study of the specificity of proteo- lytic enzymes, it was noted that several different enzymes exhibited the same backbone and side-chain requirements in their substrates. For example, as will be seen from Table II, an enzymic component of papain and an enzymic component of beef spleen cathepsin have the 125 J. S. FRUTON same type of specificity and split the same synthetic substrates as does crystalline pancreatic trypsin. Because of this relationship, the members of this group of enzymes have been designated "trypsinases." Similarly, evidence has been obtained for the existence of groups of enzymes related in specificity to pepsin, to leucine aminopeptidase, Table II Specificity of Proteolytic Enzymes Enzyme Requisite groups in substrate backbone Requisite groups in substrate side chain Peptidases (Exopepiidases) Leucine aminopeptidase from in- testinal mucosa, beef spleen, beef kidney, and swine kidney Chymotrypsin aminopeptidase Other aminopeptidases Carboxypeptidase from pancreas, beef spleen, beef kidney, and swine kidney Other carboxypeptidases R I NHiCHCO^NH... Same Same R I . . .CO^NH-CH-COOH Same CHi CHi \CH • CH2 . HO-C6H4CH2. or C8H6CH2... HGCsHi-CHj. or CeHsCHj... Proteinases (Endopeptidases) Pepsin Pepsinases from beef spleen, beef kidney, and swine kidney Trypsin Tripsinases from beef spleen, beef kidney, swine kidney, and papain Chymotrypsin R : I . . CO—NH- CHK-CO^J^H- CH . . CO—NH- CH- CO -^NH . . . R I . . . CO—NH- CH • CO^NH . . . HO.C6H4CH2... or CeHs-CHj... NH2-CH2(CH2)a.. or NHj. >C.NH.(CH2)3 HOC6H4CH2... or C6H6CH2... and to crystalline pancreatic carboxypeptidase. These are referred to as "pepsinases," "leucine aminopeptidases," and "carboxypep- tidases," respectively. Such groups of enzymes possessing identical backbone and side-chain requirements are termed homospecific enzymes (6) . On the other hand, two enzymes which diflfer from one another with respect to their backbone or side-chain requirements, or both, are designated heterospecific enzymes. 126 HYDROLYSIS OF PEPTIDE BONDS To our knowledge, the proteolytic enzymes represent the first class of enzymes for which liie property of homospecificity has been demonstrated. It is not unlikely, however, that similar relationships may exist in other classes of enzymes. Some years ago, the question was raised whether the hydrolysis of the various /3-glucosides by emulsin is to be attributed to a single iS-glucosidase or to several glucosidases of slightly different specificity, and also whether the emulsins of various plants contain identical or diff'erent /3-glucosidases. In particular, Weidenhagen (33) advocated the theory that the emulsins from various sources contain the same /3-D-glucosidase, and that this enzyme splits not only all /3-D-glucosides containing various aglucones, but also all oligosaccharides in which the sugar components are linked through /3-D-glucosidic linkage. More recently, Pigman (28), in his classi- fication of carbohydrases, suggested that the individual enzymes of Weidenhagen's system be considered as classes of enzymes acting on the same substrates but with different specificities. The finding of homo- specific proteolytic enzymes now raises the question whether similar groups of carbohydrases of identical specificity type exist. It may be added that Bisseger and Zeller (13) have applied the concept of homo- specificity to choline esterases obtained from various tissues. Mechanism of Enzymic Proteolysis It is inviting to speculate about the structural factors in the enzyme which give rise to the phenomenon of homospecificity. It was mentioned earlier that each proteolytic enzyme requires, for its action, certain atomic groupings in the backbone of its substrates. This backbone specificity may perhaps best be explained by the hypothesis that each enzyme molecule contains an essential center, composed of several distinct atomic groupings in a definite arrangement, and that the first step of enzymic action consists in a combination of several atomic groupings of the essential center of the enzyme with the indis- pensable backbone groups o' the substrate. As in the Michaelis concept of the enzyme-substrate compound, it is assumed that this combination would result in the activation, and subsequent hydrolysis, of an adjacent peptide bond of the substrate. To explain the homo- specificity phenomenon, it seems necessary to conclude that the enzymic action and its specificity originate from a rather restricted area 127 J. S. FRUTON of the enzyme-substrate complex — the reacting nucleus — while the rest of the complex may be expected to have some influence only on the rate of the enzymic action . Two homospecific enzymes thus may differ in many respects but are assumed to contain identical essential centers, and therefore yield, with the same substrate, two enzyme- substrate compounds containing identical reacting nuclei. ■ Support for this hypothesis comes from the finding that, if each of two homospecific enzymes is allowed to act on two substrates, the quotient of the rates of hydrolysis of each substrate by the two enzymes is independent of the nature of the substrate. For example, papain trypsinase was found to hydrolyze benzoylargininamide with a proteo- lytic coefficient* of 167, while beef spleen trypsinase splits the same substrate with a coefficient of 8.3. The quotient of these two values is 20.1. If the same enzymes are tested with benzoyllysinamide as the substrate, 78 and 3.8 are the coefficients found, giving a quotient of 20.5. Similar examples may be cited for other pairs of homospecific enzymes; for further data, cj. Bergmann (6). The fact that two homo- specific enzymes, Ei and Eo, give similar proteolytic quotients for the hydrolysis of the substrates Si and S2, may be explained by the assump- tion that the enzyme substrate compounds, EiSi, E2S1, E1S2, and E2S>, all contain identical reacting nuclei. Furthermore, if this concept is correct, it may be expected that parts of the enzyme molecule may be split off" without destruction of the enzymic activity, and without alteration of the enzymic specificity. Indeed, Kunitz (21) has shown that a-chymotrypsin may be trans- formed into 7-chymotrypsin, and that in the course of this transforma- tion about one-third of the a-chymotrypsin molecule is removed. However, neither the proteolytic activity toward proteins nor the specificity of action on synthetic substrates was altered. It may be added that both a- and 7-chymotrypsin have been found to exhibit two distinct proteolytic specificities, one of the amino- peptidase type and another of the proteinase (endopeptidase) type (18). Since all efforts to alter the ratio of the two specificities were unsuccessful, and also since a- and 7-chymotrypsin both conform to the * The proteolytic coefficient (C) is defined as the value of the reaction velocity constant (A") for the hydrolysis of a peptide bond in the presence of an amount of enzyme corresponding to one milligram of protein nitrogen per cubic centimeter of the test solution. 128 FrvDRor^vsis of pf.ptidf: bonds phase rule crilcria lor a pure protein (15), it wonUI .ippcar lli.K each of these two proteins exhibits more than one distinet enzymic speeifiriiy. The twofold specificity of the chymotrypsins may he ex|)lainecl on the basis of the hypothesis presented above by assuming that each of the protein molecules contains two distinct and different essential centers. The hypothesis of the predominant role of the essential center as against that of the remainder of the enzyme molecule cannot be considered the same as the well-known theory of Willstalter (34), who assumed that an enzyme molecule consists of a colloidal carrier and a prosthetic group, the latter being responsible for the enzymic activity and specificity. On the basis of this theory, Kraut distinguishes be- tween the pheron and the agon as the two components of enzymes (20). Numerous workers refer to the prosthetic group as the co- enzyme and the carrier as the apoenzyme (1), th6 dissociable flavo- proteins frequently being regarded as the prototype of this postulated dual structure. It should be recalled, however, that the flavin part of the flavoprotcins usually represents one of the partners in the chemi- cal reaction and that the essential catalytic activity resides in the protein moiety; cf. also Parnas (27). In the case of proteolytic enzymes that do not require activation by sulfhydryl compounds (pepsin, trypsin, etc.), no evidence of a dual structure is available. For the activatable proteolytic enzymes, our present knowledge indi- cates that the active enzyme represents a dissociable combination of a protein with one of several activators (19). However, in the case of these latter enzymes, it cannot be claimed that the protein part of the enzyme acts merely as a colloidal carrier for another active part of the enzyme. On the contrary, it is the protein part of the activated enzyme which contains the essential center and which determines the specificity. No proteolytic enzyme is known in which the nature of the activator determines the specificity type of the enzyme. Antipodal Specificily of Proteolylic Enzymes The majority of the known proteolytic enzymes of higher plants and animals has been found to be adapted to the hydrolysis of sub- strates in which the essential side chain belongs to an /-amino acid. For example, chymotrypsin endopeptidase rapidly hydrolyzcs the substrate benzoyl-/-tyrosylglycinamide, but does not hydrolyze bcnzoyl- 129 J. S. FRUTON ^-tyrosylglycinamide (9). The hypothesis of the essential center offers an explanation for such antipodal specificity. In order to act HO- -/ N— CH2 H H CH2— / V-OH ^ ^ \ / \ / ^ ^ C G /\ < E /\ < E HN CO HN CO i I II OC NH OC NH / \ / \ /-antipode (/-antipode Chymotrypsin I upon the substrate, the enzyme mast approach the substrate closely so that the groups in the essential center of the enzyme can combine with the indispensable groups in the backbone of the substrate. If we consider the tetrahedral arrangement of the groups about the asym- metric carbon atom, it becomes evident that the enzyme must approach the substrate from the right side in order to combine with the essential backbone groups of the substrate. In the case of benzoyl-af-tyrosyl- glycinamide, the side chain prevents the enzyme from approaching the backbone groups, and therefore the enzyme cannot split the substrate. When this theory of antipodal specificity was first proposed (5), it was suggested that the size of the side chain of a'-alanine should be sufficiently small so that the combination of the essential center of the enzyme with the backbone groups of the substrate would not be prevented completely, but would be made more difficult. This steric hindrance should result in a retardation of the enzymic action on peptides of (/-alanine as compared with its action on the /-form. In fact, it was found that crystalline pancreatic carboxypeptidase was able to hydrolyze carbobenzoxyglycyl-i^-alanine, albeit much more slowly than the /-form (8). It must be emphasized that the above conclusion can be correct only when it has been established that it is the same enzyme which hydrolyzes the two substrates containing d- and /-alanine. This res- ervation applies particularly to earlier experiments in which a study was made of the antipodal specificity of so-called dipeptidase and of aminopeptidase (12), since the existence of individual dipeptidases as well as the homogeneity of "aminopeptidase" have become doubtful. 130 HYDROLYSIS OF PEPTIDE BONDS Finally it should be mentioned that recent years have witnessed the discovery, in e\'er-increasing number, of proteolytic enzymes adapted to the hydrolysis of peptides of rf-amino acids (3,25,29). The existence of such enzymes presents us with an interesting problem: if the antipodal specificity of a peptidase has its basis in the nature of the essential center, then the essential center of a ^-peptidase should possess a configuration antipodal to that of the corresponding /-pepti- dase; cf. also Lettr6 (22). Role oj Proteolytic Enzymes in Peptide Synthesis In recent years, especial emphasis has been given to the dynamic character of protein metabolism. The studies of Whipple (24), Schoen- heimer (30), and others have given dramatic evidence for the view that, in the tissues of animals and plants, protein molecules are rapidly and continuously broken down and new protein molecules built up. The recognition of the "dynamic equilibrium" of proteins in vivo has brought to the fore the question of the nature of the enzymes that selectively catalyze the sequences of chemical reactions in protein synthesis and breakdown. More particularly, much attention has been given to the nature of the enzymes that micdiate the synthesis of peptide bonds between the individual amino acids. One view, which is held \videly, is that the biosynthesis of peptide bonds is catalyzed by the same enzymes that are responsible for the cleavage of peptide bonds; in other words, biological peptide synthesis is thought to repre- sent a reversal of the degradative action of the proteolytic enzymes. The tissue proteolytic enzymes which are presumed to perform in vivo synthesis are those frequently designated "cathepsins" (in the case of animal tissues) and "papainases" (in the case of plant tissues). The view that the hydrolytic action of proteolytic enzymes might be reversed was advanced by several workers at the start of the century and was later championed by Wasteneys and Borsook (32). More recently, it was shown unequivocally (7) that numerous proteo- lytic enzymes can catalyze, in model experiments, the synthesis of peptide bonds. One of many examples of such synthesis is the catalysis, by activated papain, of the following reaction: benzoyl-Z-leucine + /-leucinanilide -^ benzoyl-/-leucyl-/-leucinanilide. The fact that, in model experiments of this type, compoimds of known and relatively simple structure are involved, in contrast to J. S. FRUTON the heterogeneous character of the protein hydrolyzates employed by previous workers, has permitted a closer study of various factors which play a role in the synthesis of peptide bonds by proteolytic enzymes. The most important of these factors are: (a) the specificity of the enzyme action; (b) the role of activators in the enzyme action; and (c) the energy relationships involved in peptide synthesis. With respect to the specificity of peptide synthesis, it may be sufRcient, at this point, to recall that the action of a proteolytic enzyme is to catalyze the attainment of equilibrium between a peptide and its hydrolytic products. Consequently, one should expect the speci- ficity of synthesis to be the same as that of hydrolysis. Indeed, it has been found experimentally that, if the chemical nature of one of the groups near a peptide linkage is altered so that a given enzyme no longer is able to hydrolyze that linkage, then a similar structural change in the components for the enzymic synthesis will also prevent the formation of the peptide. As is well known, several of the intracellular proteolytic enzymes of animals and plants require, for their full catalytic activity, the addition of sulfhydryl compounds as activators. The view was ex- pressed some years ago (26) that, by oxidation of the sulfhydryl groups of a proteolytic enzyme into disulfide groups, the enzyme would cause peptide synthesis instead of hydrolysis. Experiments with model substrates soon showed, however, that the activation requirements were the same for the synthetic as for the hydrolytic reaction (7). Turning now to the energy relationships involved in peptide synthesis, we should recall that the hydrolysis of peptide bonds in proteins and peptides proceeds spontaneously in the presence of a suitable enzyme and that the equilibrium which is established is very far on the side of hydrolysis. Thus, in order to reverse the hydrolytic reaction, energy is required. Borsook (14) has calculated, from thermal data, that the energy needed for the synthesis of a peptide bond is approximately 3000 calories per mole. This energy may be obtained in a variety of ways. In the case of the synthesis of benzoyl- leucylleucinanilide, mentioned earlier, the driving force for synthesis comes from the removal, by crystallization, of the synthetic product from the solution. In order to restore the balance of the equilibrium reaction, synthesis occurs, which, in turn, causes more of the synthetic product to crystallize. 132 HYDROLYSIS OF PEPTIDE BONDS Another mechanism for favoring the formation of peptide bonds IS coupling the synthetic reaction with another cnergy-yiekHng chemical reaction. At the present writing, few experimentally demonstrated examples of such coupling can be cited (4). Surely the careful study of the coupling between peptide synthesis and energy-yielding systems represents one of the most interesting directions for future research; cj. Bergmann and Fruton (11). Before leaving the question of the role of the proteolytic en- zymes in peptide synthesis, it should be emphasized that the available experimental knowledge does not yet permit the conclusion that re- versal of proteolysis is actually the process employed in biological systems for the synthesis of peptide bonds. It is well to remember that in the metabolic transformation of the polysaccharides, for example, synthesis is not effected by the reversal of the hydrolytic action of the amylases, but rather through a different chemical pathway, namely, the synthetic action of the phosphorylases (16). The intervention of phosphate in the biosynthesis of peptides also has been suggested (23), but no experimental evidence for this view has as yet been brought forward. Other speculations concerning the biological mechanisms for peptide synthesis have been advanced, largely on the basis of in vitro reactions (11). Clearly a greater fund of data on coupled reactions in metabolic systems is required before it will be possible to decide which, if any, of these theories is correct. Perhaps the strongest reason for assuming, as a working hy- pothesis, the view that proteolytic enzymes do play an important role in protein synthesis is the fact they are the only known biocatalysts which, by virtue of their sharp specificity, could direct, precisely and reproducibly, the coupled sequence of successive peptide syntheses required for the formation of a protein. The considerations concerning specificity which have been discussed earlier in this article cannot fail to modify our picture of the possible role of the proteolytic enzymes in the biological synthesis of proteins. Until a few years ago, intra- cellular proteolytic enzymes, such as papain or cathepsin, were re- garded either as single enzymes or mixtures of very few enzymes. On this basis it was concluded that the specificity of a single enzyme can predetermine the molecular pattern of a protein. Thus it was assumed that the specificity range of an intracellular proteinase would be suffi- ciendy broad to comprise all the peptide bonds present in a protein J. S. FRUTON molecule. The demonstration of the extremely precise side-chain specificity of the proteolytic enzymes suggests that the synthesis of a protein from amino acids or small peptides could be accomplished only by the cooperative successive action of many enzymes of different specificity. In offering this suggestion, it must be emphasized again, however, that the view that the proteolytic enzymes mediate the bio- synthesis of proteins is not supported as yet by unequivocal experi- mental evidence. There can be no doubt that the efforts of numerous biochemists will be directed in the future to the elucidation of this fundamental problem. References (1) Albers, H., Angew. Chem., 49, 448 (1936). (2) Balls, A. K., and Kohler, F., Ber., 64, 34 (1931). (3) Bamann, E., and Schimke, O., Naturwissenschqften, 29, 558 (1941). (4) Behrens, O. K., and Bergmann, M., J. Biol. Chem., 129, 587 (1939), (5) Bergmann, M., Science, 79, 439 (1934). (6) Bergmann, M., in Advances in Enzyniology, Vol. II. Interscience, New York, 1942, p. 49. (7) Bergmann, M., and Fraenkel-Conrat, H., J. Biol. Chem., 119, 707 (1937). (8) Bergmann,M.,andFruton,J.S.,J.5to/.CAd'm., 117,189(1937). (9) Bergmann, M., and Fruton, J. S., J. Biol. Chem., 124, 321 (1938). (10) Bergmann, M., and Fruton, J. S., in Advances in Enzymology, Vol. I. Interscience, New York, 1941, p. 63. (11) Bergmann, M., and Fruton, J. S., Ann. N. T. Acad. Sci., 45, 357 (1944). (12) Bergmann, M., Zervas, L., Fruton, J. S., Schneider, F., and Schleich, H., J. Biol. Chem., 109, 325 (1935). (13) Bisseger, A., and Zeller, E. A., Helv. Physiol. Pharm. Acta, 1, C86 (1943). (14) Borsook, H., and Dubnoff, J. W., J. Biol. Chem., 132, 307 (1940). (15) Butler, J. A. V., J. Gen. Physiol., 24, 189 (1940). (16) Cori, C. F., Biol. Symposia, 5, 131 (1941). (17) Euler, H. v., and Josephsohn, K., Z- physiol. Chem., 162, 85 (1926). (18) Fruton, J. S., and Bergmann, M., J. Biol. Chem., 145, 253 (1942). (19) Irving, G. W., Fruton, J. S., and Bergmann, M., J. Biol. Chem., 139, 569 (1941). (20) Kraut, H., and Pantschenko-Jurewicz, W. v., Biochem. Z-> 275, 114 (1924). (21) Kunitz, M., J. Gen. Physiol., 22, 207 (1938). HYDROLYSIS OF PKFriDn BONDS (22) Lettr^, H., Angew. Chem., 50, 581 (1937). (23) Lipmann, F., in Advances in Enzymology, Vol. I. Intcrscicncc, New York, 1941, p. 154. (24) Madden, S. C, and Whipple, G. H., Physiol. Revs., 20, 194 (1940). (25) Maschmann, E., Biochem. Z-, 313, 129 (1942). (26) Maver, M. E., and Voegtlin, C, Enzymologa, 6, 219 (1939). (27) Parnas, J., Am. Rev. Soviet Med., 1, 485 (1944). (28) Pigman, W. W., J. Research. Natl. Bur. Standards, 30, 257 (1 943). (29) Schmitz, A., and Merten, R., Z- physiol. Chem., 278, 43 (1943). (30) Schoenheimer, R., Dynamic State of Body Constituents. Harvard Univ. Press, Cambridge, 1942. (31) Smith, E. L., and Bergmann, M., J. Biol. Chem., 153, 627 (1944). (32) Wasteneys, H., and Borsook, H., Physiol. Revs., 10, 110 (1930). (33) Weidenhagen, R., Ergeb. Enzymjorsch., 1, 205 (1932). (34) Willstatter, R., Ber., 59, 1 (1926). 135 70 METABOLIC PROCESS PATTERNS FRITZ LIPMANN, research chemist, Massachusetts general HOSPITAL, boston; RESEARCH FELLOW IN BIOCHEMISTRY AND SURGERY, HARVARD MEDICAL SCHOOL IVe have hitherto failed in our comprehension oj life mainly because we have been involved in tlie absolute method of dealing with things. E. NOBLE (14) THE CEASELESS occurrence of metabolic processes in a living cell has long been understood to imply at large a need of energy for maintenance of active life. There was, however, and still is, only a vague realization of the tasks for which uninterruptible flux of energy is needed. A first opening here appeared when, through fuller chemical resolution, recently, reaction chains unfolded which, rather unexpectedly, were found to involve a multitude of substances containing phosphate in peculiar linkages. When the way in which these phosphate intermediates are manipulated in the cell was inidcr- stood, it became possible to see clearly the connection between phos- phate cycles and transformation and transport of energy. \n all cells studied, a chemical network of energy distribution, the adenylic acid system, was found to be present and able to carry in the form of special energy-rich phosphate bonds standard portions of energy, amounting to about one-fiftieth of that liberated by total combustion of a mole of FRITZ LIPMANN carbohydrate. Therewith the view developed that catabolism con- sists to a considerable extent of a conversion of potential energy of food- stuffs into directly utilizable phosphate bond energy (7), and that, through alternate attachment and release of energy-rich phosphate bonds, catabolism and anabolism are knit together into a largely re- versible reaction continuum. This new appreciation of aspects of the metabolic apparatus which have hitherto been well concealed is beginning to affect our general attitude toward problems of metabolic chemistry. The more we recognize transformation of energy as a primary problem in meta- bolic processes, the more are we compelled to treat metabolic proce- dures for what they really are, namely, technical devices. In detail, the manner in which the living organism solves the problem of energy conversion is rather different from the technological methods employed by man. But whether in the case of the organism or man, the ultimate objective of energy conversion is the generation of energy in a utilizable form. A fundamental analogy appears, indeed, between the in- creasingly close dependence of our own daily life on electric current, gas pipes, and a variety of motors and that of our body cells on food and oxygen. In both instances the supply of energy is necessary to maintain an organization, although most of the energy is ultimately dissipated in the form of heat. In many respects a living cell is comparable to a chemical factory. The design of chemical factories, from the standpoint of a technologist, is based on a variety of technical principles (5). Only the process proper remains chemistry, but its technical execution is effected wholly by physicomechanical devices. This predominantly mechanical manipulation of unit processes represents a most significant difference between organismic chemistry and chemistry practiced by man, for, in living cells, both process design and process execution are based on chemical principles. Instead of the material being manipu- lated successively in spatially separated compartments, cellular chem- istry involves a harmonious series of consecutive reaction steps which are brought about on a molecular scale by a host of catalysts, all present together in the same reaction fluid. This difference in type of operation tends to obscure the basic analogy of both procedures. Most metabolic processes classify among what the chemical engineer calls "flow processes" (5), that is, procedures whereby streams 138 METABOLIC PROCESS PATTERNS of material enter and leave a reaction system uninterruptedly. The technical flow process is based on flow charts which map the route along which a compound is driven through a series of operations "An ideal flow process is characterized by steady states of flow, tem- perature and composition at any point of the process." This charac- terization holds likewise for almost any metabolic process. Process Characteristics of a Fermentation The physicomechanical environment in which we live has in- fluenced our thinking to the point at which we must overcome certain mental inhibitions in order to comprehend the almost exclusive reliance of the living organism on chemical operations. This is particularly the case with processes of power generation which we habitually associate with highly mechanical machinery, though most of the power ultimately derives, as in our bodies, from chemical combustion of carbon and hydrogen. There has been an additional, more inci- dental, obstacle to a ready understanding of biochemical energy transformation. In fermentation— we are just gathering the elemen- tary facts — human interest has centered long on manufacturing aspects, like the production of alcohol and other valued substances. From the biological point of view, a fermentation or a respiration is designed to produce power and the nature of the end product is more or less second- ary and accidental. In the simpler forms of anaerobic carbohydrate utilization, e. g., lactic and alcoholic fermentation, the mapping of the sequence of reactions is now completed. But it will take some time until their pattern and design are duly comprehended. However, life and multiplication of a large variety of organisms are maintained exclu- sively through fermentative transformation of energy, frequently involving simple organic and nitrogenous compounds as starting materials. Therefore principles derived from the chemical mechanics of fermentation allow a fair amount of generalization. In scheme I, the simplest fermentative process, the conversion of glucose into lactic acid and phosphate bond energy is represented. To emphasize process characteristics, the now rather well-known inter- mediaries are omitted in the scheme; the flow chart represents the reaction sequence: FRITZ LIPMANN hexose/2 *- ^^ph ph-glyceraildehydc + ph ph-glyceryl'^ph — 2 H —*■ '^ph ph-glycerate — H2O ph'^'enolpyruvate —*■ '^ph pyruvate -\r 1H lactate The terms '~ph, -ph, and ph characterize, respectively, the energy rich phosphate bond (12 kcal.), the ester phosphate bond (3 kcal.) and inorganic phosphate (7). Scheme i The Process Pattern of Lactic Acid Fermentation HEXOSE * ^ , )k — r^ 12 Kcal.(~ph) ad 12 Kcal.(~ph) LACTIC ACID-^ The flow line represents a projection into space of the catalytic pathway a hexose molecule travels to reach the inert end product, lactic acid. Initial fission in the middle of the six-carbon chain seems to be prompted by the introduction of phosphate groups at both ends. This phosphorylation is a rather costly investment absorbing just one- half of the gross yield of energy and thus reducing the net yield by about fifty per cent. An initial investment of part of the ultimate energy yield in the operation of the process is a notable feature. It is this need of induction energy which makes the fermentative process autocatalytic. The misleading statement is often made that, in fer- mentation, one half of the hexose molecule oxidizes the other half, suggesting a dismutative process. What happens, rather, is that a hydrogen donor, after unloading of phosphate bond energy, is trans- formed into a hydrogen acceptor. This manner of manipulation is expressed in the characteristic shape of the flow lines which, in all fermentations, fold back on themselves. The bending back, to accept a pair of hydrogens released in a previous stage on the molecular flow line, together with the initial expenditure of energy to start the process, may be considered as general and dominant characteristics of anaerobic metabolism. 140 METABOLIC PROCESS PATTERNS Technologically there are numerous disadvantages in operation of the anaerobic type of metabolism. Most prominent is low efficiency. The probable upper limit of the gross yield is about ten per cent, whereas ninety per cent of the potential energy of combustion remains unused in the waste products. A piling up of waste products, fre- quently of strongly acidic character, presents a further serious technical problem. In the more highly organized living systems, therefore, we find the aerobic type predominant. If in the higher organisms we meet, as we do sporadically, a well-developed system of anaerobic energy conversion, it is in places or stages of development at which, for structural or topographical reasons, the oxygen supply is poor or unsafe: in the embryo, in cancer tissue, in parts of the placenta, parts of the retina, in muscle, etc. An anaerobic energy supply grants greater independence (8). Process Characteristics of Respiration The introduction of oxygen as hydrogen acceptor increases considerably the complexity of the energy-yielding process. Our present insight in this case is spotty and far removed from the com- pleteness achieved in understanding simpler anaerobic fermentations. An attempt has been made here to coordinate the available data into a coherent process scheme, and at the same time to point out those stretches in the flow lines for which information is still missing. When, as in Figure 1 , the progressive catabolism of a substrate molecule in the manner of a flow chart is projected onto an energy- time coordinate system, some representative features emerge. The particulars of this scheme refer to degradation of half a glucose unit through the citric acid cycle. The gross energy available from this process, calculated by summation of the areas above the six consecutive steps of dehydrogenation is 6 X 57, = 342 kcal., a quantity practically identical with the theoretical yield for carbohydrate combustion. The dehydrogenation potentials of the intermediaries oscillate almost symmetrically around the hydrogen potential. Contrary to present convention, the hydrogen potential at pM 7 is made the reference potential, which coincides with the potential of the "average" respira- tory hydrogen donor. By plotting in a conventional manner oxidation-reduction potentials, E'o, as the difference between the normal potential of the system at pH 7 and the potential of atmospheric hydrogen gas at pH 0, a meaningless zero 141 FRITZ LIPMANN line cuts arbitrarUy through at the succinate/fumarate potential not far below the middle between hydrogen and oxygen potentials at pH 7. 0, +1.23 r + 0.42 + 0.2 "5 > -0.2 \ \ |2H |2I HC:0 + Spo HCOH HzCOpo" COa + COOH CH2 CH2 COOH C02 + COOH COOH COOH CH2 CH2 _^ CH H C:0 CH oPoi/^CHj COOH COOH _^COOH Fig, 1. — Citric Acid Cycle. The dotted lines mark off the constantly repeating process unit. Each turn — from condensation to oxalacetate regeneration — oxidizes a two-carbon unit of carbohydrate level. The two-carbon unit is fed into the system in the form of acetate radicals. Oxidation-reduction potential, volts> Absolute potential Hydrogen donator" Water system" Phosphate sy3tem<^ difference between oxygen and water system, volts* Phosphoglyccraldehyde Pyruvate Isocitrate Ketoglutarate Succinate Fumarate-malate -0.1 -0.35 0.13 -0,35 0.43 0.25 0.2 -0.05 -0.05 0.55 1.33 volts 1.58 1.10 1.58 0.80 0.98 Sum 7.37 volts Theory 7.46'* {Contintud on following page) METABOLIC PROCESS PATTERNS Unfortunately, all schemes following the well-established rule of assigning to oxidation-reduction potentials values increasingly positive toward oxygen depict respiratory processes ambiguously. In respiration, the chemical potential gradually diminishes, of course, toward oxygen, being eventually spent with oxygen reduction, when the listed potential attains the most positive value. As a characteristic of the process pattern, a separabihty of two main flow lines appears. The line representing catabolism of the substrate glides along on an approximately equipotential path; and at right angles to it the hydrogens which have been released by de- hydrogenation journey to oxygen. So far, the greatest progress has been made in the field of substrate catabolism in which workable schemes have emerged. Schemes like that of the citric acid cycle, however, do not supply information about the chemical pathways of respiratory transformations of energy beyond the stage of hydrogen donation. The pathway of the substrate supplies merely the level from which electrons are emitted, loaded with potential energy and ready to be used. We learn, thus, from our map that the potential difference between oxygen and each of the six dehydrogenation steps which sum up to complete oxidation of a triose averages very closely to the value of the oxyhydrogen potential. In other words, carbo- hydrate reacts grossly like a mixture of carbon dioxide and molecular hydrogen: (CHOH-HsO), = (C02-2H2), (1) At first approximation, it seems justified therefore to consider the respiratory process as a repeating series of rather uniform process " For isocitric and ketoglutaric acids, tentatively, the potentials of hydroxybutyric (4) and pyruvic acid (11), respectively, were used here. For other values of the oxidation-reduction potentials, cj . Green (4). ft Reference potential: hydrogen electrode at /)H 7; c} . page 141. c The terms, water system and. phosphate system, refer to the hydrated and the corresponding phosphorylated double bonds, as, for example, in phosphoglyceraldehyde hydrate and phosphoglycer- aldehyde phosphate (9). The difference of the oxidation-reduction potentials between the water series and the phosphate series is approximately constant and equal to the volt equivalent of the energy-rich phosphate. For acetyl phosphate, recently, a bond energy of approximately 15 kcal. was calculated (10, 121 which corresponds to roughly 0.3 v. This value is appropriate for calculations primarily con- cerned with bond generation. Th# average energy is somewhat lower, 12 kcal. or 0.25 v., which value is preferred for turnover calculations. In cases in which more or less arbitrarily the actually reacting dehydrogenation system is assumed to be of the phosphate type, the connecting line is drawn through the phosphate system. In these cases, a vertical line in the graph connects the potential points of water and phosphate system, indicating the energy transformation. Tht wide empty area above the line con- necting the potentials indicates the large part of the energy which remains here unaccounted /or {cJ. Fig. 2). <* Calculated from the combvistion heat of one-half mule of glucose, 343 kcal. The agreement between the rather roughly approximated voltage and the combustion heat is noteworthy. To be accurate, 0.25 v., the equivalent of the nonoxidative phosphate bond in phosphopyruvate, should be added to the sum of the oxidation-reduction potentials. FRITZ LIPMANN miits. Sucli a uiiil is essentially an oxygen hydrogen cell. This unit is further broken down in the scheme of Figure 2, which may be repre- sentative of the respiration not only of carbohydrate but also of other substrates. 1.2 -50 1.0- 0.8- 0.6 04 0.2 o > CYTOCHROMES -30 -FLAVOPROTEINS, 10 PYRIDINE NUCLEOTIDES o 2H ~POi- PO^ •po; ADENYLIC ACID METABOLITE Figure 2. — ^Transformation of Electron Potential into Phosphate Bond Energy. This graph accounts for the energy deficit which appeared in Figure 1 as empty space above the substrate flow line. A more detailed picture is ob- tained by fitting into the detours cycles of the type depicted in scheme II, page 146. Projection of the potential gradient onto a space scale enables us to fix approximately on the map those regions where transformation of electron potential into phosphate bond energy is to be expected. It is known from a gross comparison between oxygen consumption and the resulting phosphorylation (1,6,15) that, by transport of one pair of hy- drogen electrons from substrate to oxygen, 3+ energy-rich phosphate bonds may be generated. Energy-rich phosphate bonds average 12 kcal. per bond, which is equivalent to a span of 0.25 v. for a two- 144 METABOLIC PROCESS PATTERNS electron system. The available potential between oxygen and a pair of average substrate hydrogens is 1.2 v., a potential span which may accommodate theoretically 1.2 divided by 0.25, or 4+ energy-rich bonds. This calculation fixes the upper limit of yield and shows the experimental values to be well within this limit. In the particular scheme of Figure 2, a generation of the three phosphate bonds established by experiment is rationalized by merely cutting out, from the potential gradient, three 0.25-v. portions in succession. Such a mapping leads to the conclusion — unambiguously, it would appear — that at potential levels of around +0.1, +0.5, and +0.9 v., with reference to the hydrogen electrode of pH 7, the pair of hydrogen electrons is intercepted three times in succession by chemical devices which transform catalytically 0.25-v. portions into energy-rich phosphate bonds. This breaks the process of hydrogen transfer up into three smaller units; and here we probably meet the smallest units of the catalytic system designed for respiratory trans- formation of energy. The three, or perhaps four, transformers which are built into the pathway of the hydrogen electrons should be con- sidered as the actual power generators of the living organism. It is very significant that these transformer systems appear largely inde- pendent of the particular hydrogen donor. Such operation of sub- strate-independent catalysts for transformation may explain how phosphate bonds are generated in a constantly increasing variety of oxidations, such as sulfur oxidation in Thiobacillus (16), the oxyhy- drogen reaction (3), and fatty acid oxidation (13). To summarize, we may say that generation of phosphate bonds, regardless of the type of hydrogen donor, may be represented by the following equation: XH2) or > + 3 HO-POa" + 3 ad-H + O2/2 > X-HaO) or > + 3 ad ~ PO3— + 4 H2O; xo) average AFo, (3 X 12) - 57 = -21 kcal. (2) Equation (2) shows that a participation of phosphate and adenylic acid frequently is a reflection of general hydrogen transfer catalysis rather than a particular case of dehydrogenation. FRITZ LIPMANN The chemical nature of the transformer catalysts which operate on three 0.25-v, spans between the limits and -j-1.2 v. remains to be considered. In analogy to what is known about the carbonyl com- pounds, which in part could have the function of transformers on the lowest potential level (9), the following generalization may be illus- trative: The catalyst should offer opportunity to a phosphate molecule to add onto a double bond. From the addition product a pair of electrons is removed wherewith the energy-rich phosphate bond is generated. This step should involve little or no energy loss. Ex- change of adenylic acid hydrogen for the phosphate group now de- livers 12 kcal. into the cell, and the residual product may be rehydro- genated. This phase involves the refund of the delivered 12 kcal. (0.25 v.) and must ultimately lead to a regeneration of the double bond. Tentatively, the potential level around 0.1 v. may be assigned to a carbonyl double bond. For the subsequent level around 0.5 v. a C:C double bond would be suitable. Finally, a double bond of the ascorbic acid type appears as a possibility for the upper level. A trans- forming unit of the type outlined is charted arbitrarily for a C: C double bond in scheme II. Scheme ii A Catalytic Transformer of Electron Potential into Phosphate Bond Energy O/R-CATALYST HC-H I "2'-' ^ • V 3 -^H SYSTEM / V /^ Eo= 0.45 V. HCH HC H-ad I II + HCOH CO-POJ- WATER SYSTEM Eo=0.2V. ..Zfll u A ^ UTILIZA- METABOLITE HjC ad ~- P0i"-^-TION t C:0 = 0.25 V. With slight modification it is applicable to any double bond of a C:X type, X being, for example, :0 or :NH. The essential feature of the system is a shuttling between addition of water and of phosphate to the double bond. Starting from the bottom and moving clockwise, we distinguish four steps of the cyclic process. First, by hydrogenation of the water system the transformer is loaded with about 0.25 v. per mole, at least. Second, the reduced water 146 METABOLIC PROCESS PATTERNS system is then converted into a phosphate system by exchange reactions. Third, in the key reaction, a dehydrogenation of the phosphate system transforms with httle loss electron potential into phosphate bond energy. And fourth, the bond equivalent of 0.25 v. is unloaded to adenylic acid, thus returning the system to the original state. In scheme III, a condensed scheme of respiration is drawn. Scheme hi Summarizing Flow Scheme of Respiratory Energy Turnover 0; 'Ph -ph 5 o o a LU »-~ph- SUBSTRATE CATABOLISM In the space projection the two energy fields occupy two dimensions, the third being assigned to the equipotential flow of primary metabo- lites. In reality, these fields of electron and phosphorylation po- tential are not homogeneous, but are canalized through chemical specificity of hydrogen and phosphate transfer, with enzyme specificity acting in the manner of connection plugs. The underlying theme of this article is a reminder that, having pulled apart the chemical continuity of the living organism, we are challenged to reintegrate the scattered pieces into a whole. "There is more in a transition than a series of states or possible cuts, more in a movement than a sequence of positions or possible stops. We have 147 FRITZ LIPMANN to place ourselves along the transition and, from within, to cut across in thought, in order to appreciate the successive states (2)." References (1) Belitzer, V. A., and Tsibakova, E. T., Biokhimiya, 4, 518 (1939). (2) Bergson, H. L., Creative Evolution. Modern Library, New York, 1944. (3) Gaffron, H., J. Gen. Physiol., 26, 241 (1942). (4) Green, D. E., Mechanisms of Biological Oxidations. Cambridge, Univ. Press, London, 1940. (5) Hougen, O. A., and Watson, K. M., Chemical Process Principles. Wiley, New York, 1943. (6) Kalckar, H., Biochem. J., 33, 631 (1939). (7) Lipmann, F., in Advances in Enzymology, Vol. L Interscience, New York, 1941, p. 99. (8) Lipmann, F., "Pasteur effect" in A Symposium on Respiratory Enzymes. Univ. Wisconsin Press, Madison, 1942. (9) Lipmann, F., "Biological oxidations and reductions," Ann. Rev. Biochem., 7, 1 (1943). (10) Lipmann, F., J. Biol. Chem., 155, 55 (1944). (11) Lipmann, F., and Tuttle, L. C, J. Biol. Chem., 154, 725 (1944). (12) Meyerhof, O., Ann. N. T. Acad. Sci., 45, 357 (1944). (13) Munoz, J. M., and Leloir, L. F., J. Biol. Chem., 147, 355 (1942). (14) Noble, E., Purposive Evolution. Holt, New York, 1929. (15) Ochoa, S., J. Biol. Chem., 151, 493 (1943); 155, 87 (1944). (16) Vogler, K. G., and Umbreit, W. W., J. Gen. Physiol., 26, 157 (1942). 148 11 BIOCHEMISTRY FROM THE STANDPOINT OF ENZYMES DAVID E. GREEN, chief of the enzyme research laboratory DEPARTMENT OF MEDICINE, COLLEGE OF PHYSICIANS AND SURGEONS, COLUMBIA university; PAUL-LEWIS AWARD FOR ENZYME CHEMISTRY /: 'T WOULD be in the nature of a platitude to say that there is hardly a branch of biochemistry which cannot be analyzed or at least interpreted in terms of enzymes or enzymic phenomena. Yet few would maintain that enzymes represent more than an intel- lectual liqueur in the teaching of biochemistry in our graduate schools. Most modern textbooks of biochemistry often treat the subject of en- zymes much in the manner that feathers and scales are dealt with in textbooks of anatomy. If this were merely another instance of the lag of textbooks behind developments in research, there would be no cause for concern. But more appears to be involved than the traditional lag. The implications of enzyme chemistry have yet to be more generally understood; and until that time arrives the textbooks will continue to regard enzymes as chemical oddities, if not to ignore them altogether. During the past fifty years, biochemistry has developed from the ugly duckling of physiology to a science in its own right. During this period, interest has been focused largely on methodological and struc- tural problems: how to estimate and what the constitution is of the innumerable compounds which make up the living cell. In this phase in the development of biochemistry, the analytical chemist and struc- D. E. GREEN tural organic chemist played the leading roles, and indeed they laid the foundations for an exact science. Hence, it is hardly surprising that dynamic problems such as those of intermediary metabolism were approached more from the direction of what may be called chemical morphology than from the point of view of physiology. Perhaps the best way of illustrating the point is to recall the sensation which was produced by Schoenheimer's (5) early isotope experiments. His con- ception of an organism as a chemical system in a constant state of flux, in dynamic as opposed to static equilibrium, bore the same relation to the classical biochemical conception that the Schrodinger-Heisenberg conception of the atom bears to the rigid atom of the late 19th century. This is not to imply that the groundwork for Schoenheimer's concep- tion was not already laid in the literature. The students of enzymes have long been aware of reversible equilibrium systems; but biochem- ists generally were unable to project the implications of these reversible systems in terms of intermediary metabolism. The outstanding re- searches of Schoenheimer and Krebs have done much to orient research on intermediary metabolism along more peculiarly functional lines; but, nonetheless, not more than the fringe of biochemical thinking has been disturbed. Intermediary metabolism is still being taught and discussed without any reference to the catalysts responsible for each of the transformations. Discussing the behavior of a car without refer- ence to the motor represents an analogous situation. This state of af- fairs provides some reason for attempting an interpretation of biochem- istry in terms of enzymes and enzymic phenomena. What follows will be rather sketchy, not only because of the limitations of space, but also in some cases because of the inadequacy of available information. However, the purpose of this essay is more to show that enzymes pro- vide a logical and rational approach to many fields of biochemistry and medicine rather than to attempt a comprehensive survey of the enzyme field. Living systems carry on their activities by virtue of myriads of chemical reactions which collectively are referred to as intermediary metabolism. Physiological functions such as growth, reproduction, secretion, nerve conduction, muscular contraction, etc., are integra- tions of whole series of chemical events in intermediary metabolism. These chemical events with few exceptions are not spontaneous proc- esses. They require the presence of highly specialized protein cata- 150 ENZYMES lysts known as enzymes. Apparently, there is a different enzyme for practically every reaction, or at least every reaction type of intermediary metabolism. This can only mean that some several thousand en- zymes must exist. For a long time fears were expressed that the cramped quarters of the average small sized cell could not possibly accommodate so many different enzymes. But that bogey has been laid low by the isolation from one small cell such as the yeast cell of literally hundreds of different enzymes. Of course, not all types of cells have the same complement of enzymes. The number and amount of enzymes vary from one cell type to another, and in fact determine the individuality of each cell. There are many instantaneous ionic reactions which occur in the course of intermediary metabolism and which do not require enzymes, e. g., neutralization of acids or bases, and deposition of salts such as calcium phosphate. But even in the province of reactions which occur spontaneously, a surprise is occasionally in store. Thus, the decomposition of carbonic acid into carbon dioxide and water is catalyzed by a special enzyme known as carbonic anhydrase, which can speed up the rate of the reaction far beyond the spontaneous rate. The study of intermediary metabolism represents one of the oldest lines of biochemical investigation. It is not surprising, there- fore, that of the total number of reactions known to occur in inter- mediary metabolism only a very small proportion have been recon- structed with isolated enzyme systems. In fact, whole chapters of intermediary metabolism such as the metabolism of steroids, porphyrins, carotenoids, sulfur compounds, bile acids, fatty acids, etc., are prac- tically virgin territory as far as knowledge of enzymes is concerned. Enzyme chemistry has made its greatest strides in the field of fermenta- tion or glycolysis of sugar. Here, the entire process from glucose to glycogen or from glycogen to lactic acid has been reconstructed in vitro with some twenty odd enzymes, each prepared in pure or largely pure state. This in vitro reconstruction is not to be regarded as a stunt or merely as a triumph of biological engineering. In order to effect a successful reconstruction, it becomes necessary to understand precisely the way in which certain enzyme systems are linked together and the way in which different chemical reactions are synchronized. A successful reconstruction, therefore, implies mastery of most of the chemical details and a complete knowledge of the constituent enzyme D. E. GREEN systems. But the success in the field of glycolysis has been to some extent at the expense of progress in other fields. Nutrition in essence deals with the relative amounts and nature of the materials which have to be supplied ultimately to the enzyme systems, and with the resyntheses and replacement of enzymes. In other words, the study of nutrition and the study of enzymes represent two sides of the same coin. Since the nature, number, and amounts of enzymes vary from one living system to another, the nutritional problem varies in the same way. If we knew all the enzymes present in a par- ticular organism and the special components present in each of the enzymes, theoretically, we would have all the necessary data for determining the complete nutritional requirements. But since we are largely in the dark about the vast majority of enzymes, we use the data and knowledge of nutrition to ferret out information about enzymes. It has long been known, for example, that traces of certain metals such as manganese, iron, zinc, copper, and magnesium are essential in the diets of many animals. The indispensability of these metals in the diet has been correlated with the presence of these metals as structural elements of important enzyme systems. Thus, manganese has been shown to be an essential component of arginase; magnesium an essential component of carboxylase, chlorophyll, etc.; zinc of car- bonic anhydrase; copper of phenolases; and iron of catalase, per- oxidase, cytochromes, and lactic dehydrogenase. Traces of cobalt are known to be essential in the diets of sheep particularly. No doubt cobalt also will be identified as an essential component of some as yet unknown system. Among plants, boron and molybdenum are es- sential trace elements, and one must presume special enzyme systems in plants requiring these elements. The fact that, in most instances, small quantities of the metals are necessary correlates with the extraor- dinary activity of enzymes at high dilutions. In other words, only traces of metals are necessary for incorporation into the enzymes, since the enzymes also occur only in trace amounts. The identification of the P-P factor with nicotinamide, of vita- min Bi with thiamin, of vitamin B2 with riboflavin, and of vitamin Be with pyridoxal provides a moral for those who prefer to study one side of a coin without reference to the other. Cozymase, a dinucleotide of nicotinamide and adenine, has long been studied as the coenzyme of fermentation since its discovery by Harden and Young (4) in 1906. ENZYMES Yet it was not until 1937, some three years after Warburg had shown nicotinamide to be an essential component of the coenzyme, that Elvehjem and his colleagues (1) established a connection between the antipellagra vitamin and nicotinamide. Elvehjem's discovery was consistent with poetic justice because he had been trained both as a nutritionist and as an enzyme chemist. Vitamin B2 was identified in 1935 by Kuhn and Karrer with riboflavin, the prosthetic group of the so-called yellow enzyme which Warburg had isolated from yeast three years earlier. The time relations were reversed in the cases of vitamins Bi and Eg, since the identification of the vitamins preceded knowledge of their participation in enzymic reactions. The chemical identification of vitamin Bi with thiamin by Williams and Cline in 1936 preceded by one year the demonstration by Lohmann and Schuster that the prosthetic group of yeast carboxylase is a diphos- phoric ester of thiamine. Peters and his group at Oxford had estab- lished the role of vitamin Bj in the oxidation of pyruvic acid long before the chemical nature of the vitamin was established. Vitamin Be was identified with pyridoxine in 1938 by Folkers, Keresteszy et al., and Kuhn et al., but it was not until 1944 that a more active form of the vitamin, viz-, pyridoxal, was discovered by Snell and that the phos- phoric ester of pyridoxal was shown to be the coenzyme of tyrosine decarboxylase by Gunsalus. There are thus four authenticated identi- fications of vitamins with prosthetic groups. In the case of vitamin A, Wald has shown that it is an essential part of a photosensitive pigment in the eye known as visual purple. Many will not concede that visual purple has the properties of an enzyme but, whether or not that point is conceded, it is at any rate admissible that vitamin A fulfills the role of prosthetic group of a chromoprotein with an important physiological function. This relation between vitamins and prosthetic groups makes it easy to understand the basis for the body's continuous requirement of vitamins. Since enzymes have a limited life period in consequence of their destruction during activity, there is constant need for more of all enzymes. The minimum amount of any of the vitamin compatible with viability is therefore an approximate measure of the total amount of enzymes in the body whose prosthetic groups contain that vitamin. Here, again, if we knew precisely which enzymes, for example, require flavin, what the normal levels of these enzymes are in different organs. D. E. GREEN and the rates at which they are synthesized, we would have all the necessary data for determining the flavin requirements of that organ- ism. There are much easier methods of getting at that data at the mo- ment, but it is of considerable theoretical interest to arrive at the data from the enzyme side. There are some well-informed workers in the field of nutrition who are unwilling to concede that all vitamins must have a catalytic function. While they admit that the relationship has been established in at least four and possibly six instances, they do not regard these necessarily as precedents. In 1941 the enzyme-trace substance theory (3) was developed which predicted that any substance necessary in the diet in trace amounts must be an essential part of some enzyme system. The theory was not meant to imply that substances required in higher concentrations cannot be essential parts of enzymes. That may or may not be the case. However, amounts of the order of 10 fjLg. per kg. per day were regarded as conclusive evidence of a catalytic role for that substance. On the basis of the enzyme-trace substance theory we may confidently expect in the near future the identification of biotin, pantothenic acid, folic acid, vitamin A, vitamin K, and vita- min D with essential parts of new prosthetic groups. The daily re- quirements of vitamin C are about a thousand times greater than those for most of the other vitamins, and, certainly, are well beyond the trace level. The possibility is therefore open that vitamin C may have no catalytic function whatsoever. The recent investigations of Sealock on the role of vitamin G in the oxidation of tyrosine in the liver however, do not encourage the view that vitamin G is an exception to the vitamin-enzyme relation. In recent years the exact nutritional requirements of many bacteria and molds have been carefully investigated. A study of any of the synthetic diets proves very instructive from the standpoint of enzyme chemistry. The list of required substances can be easily divided into two categories: (7) substrates of enzyme systems, e. g., amino acids, glutamine, dextrose, fatty acids, purines, etc.; and (2) precursors of prosthetic groups, e. g., vitamins such as riboflavin, thiamin, etc. or the building stones thereof, hemin, and trace metals. The members of the first category can be classified not only on the basis of what we know about their intermediary metabolism but also on the basis of the amounts required, which are vastly in excess of the ENZYMES minimal amounts necessary of substances in the second category. The concentrations of the first category are in milHgrams per cubic centimeter, whereas those of the second are in fractions of a micro- gram per cubic centimeter. The enzyme-trace substance theory in the form stated above is appHcable only to naturally occurring substances in the diet or growth medium. There is an alternative form of the theory which applies to all substances regardless of their origin, and it may be stated in the following terms: Any substance which in trace amounts induces pro- found biological effects does so either by participating in or by spe- cifically afifecting some enzyme system. If valid, this theory must be regarded as one of the cornerstones of the science of pharmacology. An extensive list can now be compiled of substances, part if not all of whose pharmacological actions can be explained in terms of enzyme effects {cj. Table I). Table I Pharmacological agent Enzyme inhibited by the agent Fluoride Enolase One of the anterior pituitary Hexokinase hormones Cyanide Cytochrome oxidase Eserine Choline exterase Prostigmine Choline esterase Chlorinating agents Triosephosphoric dehydrogenase lodoacetic acid Triosephosphoric dehydrogenase Benzedrine Amine oxidase Phlorhizin Glucose phosphorylase Gramicidin One of the fermentation enzymes involved in phosphorylation Enzyme identical with the agent a-Toxin of Clostridium wdchii A lecithinase Lytic factor of cobra venom A lecithinase Spreading factor Hyaluronidase One of the toxins of CI. welchii CoUogenase One cannot but be impressed by the vindication of the enzyme hypothesis in at least fourteen instances, most of them reported within the last few years, in contrast to the absence of a single authenticated case in which any other principle of mechanism has been shown to operate. At present, the enzyme theory is really not much more than a good working hypothesis. But while it may be premature to regard all the biological eflfects of trace substances exclusively in terms of D. E. GREEN cnzymic effects at tiie jneseiit time, there is n(j alternative explanation that merits serious consideration. Alternative explanations usually amount to substituting an obscure phrase for an obscure phenomenon. Thus, some pharmacologists talk about trace substances upsetting an "active patch" of some important cellular membrane and thereby exerting their action. When interrogated about the properties of the "active patch," the pharmacologist usually admits that he has in mind some specific combining group and in effect admits a somewhat watered-down version of an enzyme reaction. Certainly there is no way of testing the "active patch" hypothesis as commonly stated nor is there any evidence that it has been productive either in explaining or predicting new phenomena. One cannot resist the conclusion that the "active patch" concept is a terminological device for cloaking ignorance. Another variant of the "active patch" concept is the so- called "active surface" which is sensitive to pharmacological agents and which controls certain key biological functions. The effect of agents on these surfaces is said to be exclusively a physical one, /. e., the "active surface" becomes covered by the pharmacological agent and consequently is inactivated. The extraordinary specificity of pharmacological effects and the high dilutions at which these effects occur render this interpretation in purely physical terms unlikely. The discovery by Woods (6) that the antibacterial action of the sulfonamides could be explained in terms of the resemblance of the sulfonamides to ^-aminobenzoic acid was an important milestone in our understanding of the mechanism of the action of drugs. It became at once clear that some cnzymic process was at the bottom of the chemo- therapeutic action of the sulfonamides. There is still no clue as to the nature of this enzymic process but there can be little doubt that the process is enzymic. /^-Aminobenzoic acid is a naturally occurring sub- stance in yeast and animal tissues. Many bacteria and molds are unable to grow unless it is present in the medium. The trace con- centrations in which it must be present for optimum growth exclude all but a catalytic role. In fact, /^-aminobenzoic acid has been shown to exist in yeast largely in the form of a polypeptide, which may well be its active catalytic form in the intact cell. One theory of sulfon- amide action assumes that the sulfa drugs displace /^-aminobenzoic acid from its combination with specific proteins and thereby inactivate enzymes important in the growth of certain microorganisms. In 156 ENZYMES other words, the natural substance, yz.c-,/'-aminobenzoic acid, competes with the drugs for the protein partner with which it forms the enzyme complex. The degree of inhibition is determined by the relative con- centrations of /^-aminobenzoic acid and drug and by their relative affinities for the protein partner. The phenomenon of competitive inhibition is well known in the enzyme literature and is perhaps the most characteristic hallmark of enzymic phenomena. The sulfonamides inhibit the growth of susceptible bacteria only when they are actively growing. Once active growth has taken place, the sulfonamides are without effect even on susceptible bacteria. This observation suggests that the sulfonamides interfere with the process of synthesizing the /^-aminobenzoic acid enzyme complex rather than with the activity or function of /i-aminobenzoic acid in its final catalytic form. Unquestionably, there are many alternative pathways in bacteria by which /^-aminobenzoic acid can be coupled with other substances to form the final catalytic complex. Only one pathway may be sulfonamide-sensitive, and only those bacteria which share that method of synthesizing the /)-aminobenzoic acid catalytic complex will be susceptible to the action of the sulfonamides. These considerations must be borne in mind in evaluating the enzymic theory of sulfonamide action. Ghemotherapeutic drugs may interfere either with the working of a key enzyme or with some stage in the process by which the key enzyme is synthesized. Since the synthesis of en- zymes is also enzymic in nature, the primary action of chemothera- peutic drugs must still be considered as one of interference with enzymes. The interpretation in terms of enzymes of the mode of action of sulfonamides is by no means in general currency. The so-called "essential metabolite" theory of Fildes (2) has gathered many adherents and is generally accepted among workers in the field of chemotherapy. This theory assumes that/^-aminobenzoic acid is an essential metabolite for the growth of certain microorganisms rather than a part of an essential enzyme system. A metabolite is usually defined as a sub- stance which undergoes chemical transformation; and the term is usually applied to substances like amino acids, fatty acids, etc. which are present in considerable concentration, and which are degraded or converted into more complex substances by enzyme systems. Since the amount of /^-aminobenzoic acid present in microorganisms is ^57 D. E. GREEN scarcely detectable by the most delicate chemical methods, p-am'ino- benzoic acid can hardly be classified as a typical metabolite. Quite clearly, the term metabolite as applied to /^-aminobenzoic acid must imply merely that it is involved in metabolism. Furthermore, since traces of essential substances are known to participate in metabolism only in the capacity of catalysts, the "essential metabolite" theory boils down to a disguised enzyme theory. In the present state of ignorance, there is some merit in talking of /)-aminobenzoic acid as an essential metabolite until such time as its precise enzymic role is clarified. But when the essential metabolite theory is seriously proposed as an alternative to the enzyme theory, it becomes important to recognize what the concept of essential metabolite really means. As applied to />-aminobenzoic acid, "essential metabolite" is just a term of caution to indicate by implication a catalytic role, without stating it in so many words. But as occasionally happens with terms of caution, their neu- trality defeats their purpose. Essential metabolite has become con- fused by many with metabolite, and its real significance has become lost among those who see only the letter and not the spirit of the term. The relation between the sulfonamides and j&-aminobenzoic acid provided a blueprint for designing other chemotherapeutic agents. Every vitamin or prosthetic group theoretically should find its nemesis in some antivitamin. So the hunt began, and not without success. The sulfonic acid analogue of pantothenic acid (pantoyl- taurine) was found to inhibit the growth of some bacteria which re- quired pantothenic acid. The ratio of pantothenic acid to the sulfonic acid analogue determined whether growth or inhibition would take place. Much of the same kind of results were obtained with pyri- thiamin, as antithiamin agent, with pyridine /3-suIfonic acid, as anti- nicotinic acid agent, etc. None of these antivitamins is comparable to the sulfonamides in their efficiency as chemotherapeutic agents; but in these antivitamins, at least, we have the hopeful beginnings of a rational program of chemotherapy. From the standpoint of enzymes, chemotherapy would appear to be the science of compounds which go for the enzymic Achilles' heel of an infectious organism without at the same time damaging the host unduly. In other words, the objective is first to find an enzyme • which is present or important in the infectious organism and not in the host, and second to find a drug which specifically inhibits this 158 ENZYMES enzyme. Clearly, little progress will be made along rational lines in chemotherapy unless progress in the enzyme chemistry of infectious organisms is stimulated. Many are not convinced that progress in chemotherapy can come from that direction. "Suppose," they say, "you find the enzymic Achilles' heel, how are you going to find the drug to inhibit that enzyme?" As an example of the direction from which the solution might come, the chemical nature of the prosthetic group or active groups of the enzyme which turns out to be the weak point might provide the necessary clue for the synthesis of specific inhibitors. None will maintain that the solution will be easy, but experience has taught us that the possibilities of solving a problem are greatly increased when the nature of the problem can be accurately defined. There is a school of thought well represented in our large phar- maceutical firms and also in the councils of our government scientific agencies which prefers to advance chemotherapy exclusively by the method of trial-and-error organic synthesis. In effect, the program followed is merely that of permutation and combination of the few effective chemotherapeutic agents we have as models. Instead of looking for new models, the old models are varied over and over again. The very limited success of this program of chemotherapy is hardly surprising. In the first place, not all pharmacologically active sub- stances tolerate any considerable structural change. Thus, no one has been able to prepare a more active form of the vitamin than thiamin. In fact, the slightest alteration of the molecule involves partial or complete loss of activity. The same considerations apply to the vast majority of biologically active substances such as acetyl- choline, histamine, flavin, ascorbic acid, etc. It is certainly true in other instances that a given effect is produced by large numbers of sub- stances sharing a common structure, e. g., adrenalin, and the hundreds of adrenalin-like bases which have been tested, sulfanilamide and the hordes of other sulfa drugs, etc. But one must keep the objective clearly in mind. In the case of adrenalin, the analogues merely imitate the adrenalin eff"ect. They accomplish nothing that adrenalin cannot do. Their virtue lies either in their greater stability or in their greater resistance to deterioration in the animal body. All the known sulfa drugs act in exactly the same way, though they are effective at different concentration levels. The same organisms can be inhibited both by the weakest and strongest sulfa drugs provided the weakest is D. E. GREEN sullicionlly s(jlul)k\ In othci- words, no new principle emerges. A more effective sulfa drug docs not extend the range of action of sulfa drugs in the sense of inhibiting organisms which are otherwise insensi- tive to the sulfa drugs. From the standpoint of enzyme chemistry, the direction which much of chemotherapy research has taken does not appear to be either profitable or rational. Chemotherapy cannot be attacked intelligently without a detailed knowledge of intermediary metabolism and enzyme chemistry. We may make allowances for the element of urgency in wartime, but after the war, there ought to be a better balance between the sums spent on sheer trial-and-error organic synthesis and the sums spent on fundamental investigations. Woolley (7) has pioneered in providing a framework for a rational pharmacology based on the antivitamin concept. He showed that certain antivitamins can produce a state of avitaminosis, in some cases in a matter of hours, merely by displacing the vitamins competi- tively from their catalytic complexes. Since profound pharmaco- logical effects attend the syndrome of avitaminosis, antivitamins have to be regarded as potential pharmacological agents. Thus, pyri- thiamin rapidly induces the disorders of the central nervous system which are characteristic of thiamin deficiency. The lesion is of course righted at once by addition of large enough amounts of thiamin. Since the quantitative importance of the catalytic reaction in which vitamins participate varies depending upon the organ or part of the organ, it does not follow that all antivitamins will exhibit similar pharmaco- logical effects. On the contrary, it would appear that each anti- vitamin would selectively poison only a particular portion of the nervous system as well as only particular organs. A complete series of anti- vitamins should provide a wide range of specific pharmacological agents, all of which are reversible by addition of the vitamins which they imitate. The beginnings in this new field of exploration are still modest but the horizons seem immense. The hormones represent a class of substance which, according to the enzyme-trace substance theory, ought unequivocally to qualify as enzymes or essential parts of enzymes. Yet no one has conclusively demonstrated that any one of this large class is either an enzyme or an essential part of an enzyme. Do we have in this class a notable excep- tion to the theory? There is no basis for answering this question defi- nitely one way or the other. There is a possibility that renin, the 1 60 ENZYMES hormone elaborated by I lie kidney, hydrolyzes hypertensinogen, one of the plasma proteins, with formation of a pressor substance. If this possibility is confirmed, we would have the first identification of a hormone with an enzyme function. Whatever the uncertainty about hormones as enzymes, the evidence leaves no doubt that hormones influence enzymic phenomena, e. g., the dramatic efTect of insulin* or adrenalin on carbohydrate metabolism. In practically every instance, hormone action has been boiled down to the regulation of some phase of intermediary metabolism. Do hormones regulate by being enzymes themselves, by influencing enzymes, or perhaps by controlling the synthesis of enzymes? There are many indications which point to some of the hormones controlling the synthesis of enzymes. But since the synthesis of proteins represents one of the most obscure corners of enzyme chemistry, there is little hope of any early clarification of the precise role which hormones might play in synthesis. Oddly enough, our most precise knowledge of the way in which the synthesis of enzymes is regulated has been acquired from the field of genetics. The experimentation which has led to this knowledge constitutes one of the most brilliant chapters of modern biology. Geneticists have succeeded in demonstrating that single genes control the syntheses of single enzymes. This implies that genes, like hor- mones, are regulators of intermediary metabolism. Genes accom- plish this regulation by controlling the synthesis of enzymes, whereas hormones operate in ways as yet not classified. Genes and hormones are distinguishable in another fundamental respect — hormones must be synthesized by the respective endocrine glands, while genes are autocatalytic and hence self-perpetuating. This autocatalytic property of the gene resolves the dilemma that, if enzymes are needed to synthe- size other enzymes, there must then be an infinite series of enzymes making enzymes. But recognition of autocatalysis as a phenomenon is a far cry from understanding the mechanism. As a matter of fact, guesswork constitutes the sum and total of our knowledge of the way in which certain protein molecules are able to reproduce themselves. * When this essay was in the proof stage, Cori and his group announced the epoch-making discovery that insuHn reverses the inhibition of hexokinase produced by one of the anterior pituitary hormones. In these two instances, at any rate, hormones must be regarded as regulators of enzymes by virtue of their inhibiting or releasing the inhibition of key enzymic processes. i6i D. E. GREEN The problem of the autocatalysis of genes is essentially similar to that of viruses. What is the starting material for the synthesis and how is the synthesis accomplished? No concrete answer is possible as yet, but there are indications that the solution to the problem will be in terms of enzyme chemistry. One intriguing development in recent years has been the successful application of the concept of com- petitive inhibition drawn from the field of enzymes to the virus problem. Two strains of the same virus are found to antagonize one another's growth in the same host, presumably by competing for the same pabulum. In chemotherapy, two substrates, one natural, the other "fraudulent," compete for the same enzyme. In the example above, two viruses are made to compete for the same substrate. This virus interference phenomenon occurs only when the two strains are closely related, just as competitive inhibition occurs only when the "fraudulent" substrate closely resembles the natural. The past decade has seen not only the extension of our knowl- edge of enzymes to other fields of biochemistry and medicine but also the extension in part of the enzyme concept to noncatalytic proteins. Let us compare, for example, two proteins, catalase and hemoglobin. Both contain iron protoporphyrin as prosthetic group and both are highly specific. Catalase catalyzes the decomposition of hydrogen peroxide into oxygen and water, whereas hemoglobin combines re- versibly with molecular oxygen. Catalase cannot function as hemo- globin and conversely hemoglobin for all practical purposes does not catalyze the decomposition of hydrogen peroxide. Catalase forms a compound with hydrogen peroxide and then the enzyme-substrate compound undergoes decomposition. This cyclical process repeats itself more than two million times per minute at 0°. In a similar way, hemoglobin forms a compound with molecular oxygen over a certain range of oxygen tension, and the complex is dissociable by lowering the oxygen tension. The speed of the combination is of about the same order of magnitude as for the combination of catalase with hydrogen peroxide. This comparison is not being made to infer that hemoglobin is an enzyme. There is little to be gained by a redefinition of the classical concept of an enzyme to include a proteinlike hemoglobin, because after all there is a real distinction between a catalyzed reaction and a noncatalyzed reaction. However, it is important to recognize the many properties in common which catalase and hemoglobin share. 162 ENZYMES While enzymes must be differentiated from noncatalytic proteins, nonetheless a broader classification of proteins is conceivable in which enzymes represent a special case of what we may call the functional type of protein. We may define the functional protein as one which performs a specific physiological function. Thus, catalase decom- poses hydrogen peroxide; hemoglobin combines reversibly with mo- lecular o.xygen; cytochrome C is reduced by the reduced forms of certain enzymes and in turn its reduced form is oxidized by cytochrome oxidase; prothrombin plays a specific role in blood clotting; visual purple acts as a photoreceptor, etc. Limitation of space precludes further development of the concept of functional proteins. Suffice to say that, in the author's opinion, processes like those of blood co- agulation, complement fixation, and antibody formation, are phe- nomena which have much in common with enzymic phenomena, and that the highly specific functional proteins responsible for these processes have more in common with enzymes than with purely structural proteins. The concept of functional proteins has the virtue of opening new horizons in the form of novel types of proteins. Just as myosin is a protein of muscle specialized to convert chemical energy to mechani- cal energy or as visual purple is a protein in the retina specialized to convert light energy, presumably, ultimately to the electrical energy of nerve conduction, so there may be analogous proteins in nerve, cellular membranes, etc. specialized to carry out the particular physio- logical functions of these organs. The trend in biochemistry would appear to be toward the inclusion of more and more proteins in the category of catalysts. In fact, it is conceivable that eventually all proteins apart from purely structural proteins will be found to perform in a highly specific way some physiological catalysis, and the currently prevalent idea of storage and inert proteins will soon be as outmoded as the so-called endogenous nitrogen metabolism. There is another aspect of myosin and visual purple that well merits consideration. Biochemists have long exercised themselves over the problem of the means by which the organism converts energy from one form to another. In one or two instances, the curtain sur- rounding these interconversions has been pierced. Myosin and visual purple may well be considered as examples of energy transformers. Thus, myosin in effect converts the chemical energy of hydrolysis of adenosine triphosphate into mechanical energy. The protein itself 163 D. E. GREEN acts as the transforming agent by combining two physiological func- tions. In the same way, visual purple becomes a vehicle for trans- forming light energy into chemical energy and, we must assume, eventually reacts in some way with nervous elements of the retina. Instances of enzymes with multiple functions and multiple active groups have been known in the literature, but the tendency hitherto has been to regard them as biological curiosities. Thus pyruvic oxidase of Lactobacillus delbrueckii contains two prosthetic groups, viz- flavin dinucleotide and diphosphothiamin. Milk flavoprotein con- tains two prosthetic groups (flavin dinucleotide and another as yet unidentified) and catalyzes the oxidation of purines, aldehydes, and dihydrocoenzyme I. The /-amino acid oxidase of rat kidney has two enzymic functions. These enzymes with multifunctions may be in- volved in the transfer of chemical energy from exergonic to endergonic processes. In other words, enzymes may ultimately turn out to be the energy transformers and converters of the cell. References (1) Elvehjem, C. A., Madden, R. J., Strong, F. M., and Woolley, D. W., J. Am. Chem. Soc, 59, 1767 (1937). (2) Fildes, P., Lancet, I, 955 (1940). (3) Green, D. E., Advances in Enzymology, Vol. I. Interscience, New York, 1941, p. 177. (4) Harden, A., and Young, W. J., Proc. Roy. Soc. London, B77, 405 (1906). (5) Schoenheimer, R., The Dynamic State oj Body Constituents. Harvard Univ. Press, Cambridge, 1942. (6) Woods, D. D., Brit. J. Exptl. Path., 21, 74 (1940). (7) Woolley, D. W., Science, 100, 579 (1945). 164 12 ENZYMIG MECHANISMS OF CARBON DIOXIDE ASSIMILATION SEVERO OCHOA, assistant professor of biochemistry, new york UNIVERSITY COLLEGE OF MEDICINE 7T IS well known that animal cells depend on a supply of ready-made organic materials (such as carbohydrates, fats, and proteins) or of their building stones (such as simple sugars, fatty acids, and amino acids) for either building up cell substance or replenishing their stores of energy-yielding foodstuffs. Green plants, however, are able to fix atmospheric carbon dioxide under the in- fluence of light and use it to synthesize the organic constituents they need; by means of this process — photosynthesis — they "assimilate" carbon dioxide. In its final over-all results, photosynthesis is essentially a reversal of respiration: In respiration, foodstuffs are oxidized to carbon dioxide and water with absorption of oxygen, the energy thereby released being utilized by the cell to carry out its activities; in photo- synthesis, the chlorophyll-containing chloroplasts utilize radiant energy to build up organic substance from carbon dioxide and water, and oxygen is liberated in the process. Some microorganisms, such as green algae and both green and purple bacteria, are photosynthetic. Certain bacteria do not possess the capacity to utilize radiant energy for carbon dioxide assimilation but are able to use, for the same purpose, the energy derived from oxidation of inorganic substances like hydrogen sulfide, thiosulfate, 165 SEVERO OCHOA sulfite, selenite, nitrite, elementary sulfur, ammonia, and molecular hydrogen. This process is known as chemosynthesis. Both photosynthetic and chemosynthetic organisms are referred to as autotrophic because they can grow in media composed of in- organic substances exclusively. However, a number of bacteria, in- cluding most of the pathogenic species, can live only in media that con- tain one or more organic components, and are known as heterotrophic (24,25,28). The importance of photosynthesis and, in general, of carbon dioxide assimilation can hardly be overemphasized, since animals depend for their subsistence on the materials formed through carbon dioxide assimilation by autotrophic organisms. Thus, carbon dioxide assimilation is one of the most fundamental of all life processes. Al- though it was known for some time that heterotrophic bacteria and some animal cells could utilize carbon dioxide to synthesize carbon-to- hydrogen bonds or carbon-to-nitrogen bonds (as is the case in the syn- thesis of formic acid from carbon dioxide and hydrogen by Escherichia coli, or in the synthesis of urea by the liver), the belief was current that such organisms lacked the capacity to utilize carbon dioxide for the synthesis of carbon-to-carbon bonds until the pioneer work of Wood and Werkman demonstrated that heterotrophic bacteria can fix carbon dioxide in this manner (28). The process is now known to occur also in animal cells. The synthesis of organic material from carbon dioxide is an endergonic reaction, which results in an increase of the free energy content of the system, and thus requires energy in order to proceed. In other words, such a synthesis must be coupled with exergonic* reac- tions involving a decrease in free energy. For this purpose, photo- and chemosynthetic organisms can use either radiant energy or the energy derived from oxidation of inorganic compounds. Heterotrophs, on the other hand, can assimilate carbon dioxide only at the expense of oxidizing organic foodstuffs, so that no net gain in organic cell constitu- ents can result from carbon dioxide assimilation under these conditions. It is, therefore, difficult to decide whether carbon dioxide fixation is * C. D. Coryell, in Science, 92, 380 (1940), introduced the terms "exergonic" and "endergonic" to characterize negative and positive changes in free energy (AF), respectively, and suggested that the use of the terms "exothermic" and "endothermic" be restricted to designate changes in heat (AH). 1 66 CARBON DIOXIDE essential or even important for heterotrophic organisms and animal cells. One could conceive, however, that fixation might be used for the synthesis of special cell constituents such as growth factors or the like. It is well known that most heterotrophic bacteria require the presence of carbon dioxide in the medium for optimal growth. The cellular mechanisms of carbon dioxide fixation have been obscure for a long time. It is only recently that light has been shed on the mechanism by which fixation occurs in heterotrophic bacteria and animal cells. As it now appears, the fundamental process in all types of carbon dioxide fixation is a reversal of the decarboxylation of some keto acids — a process catalyzed by enzymes. These reactions are reversible, but their equilibrium lies very far to the side of decar- boxylation, i. e., liberation of carbon dioxide. Thus, the problem faced by the cell is to shift the equilibrium as far as possible in the opposite or uphill direction, and this requires expenditure of energy. It has been established (15) that the free energy change of a re- versible chemical reaction is related to the equilibrium position in a manner expressed by the equation: ^F = -RT In K where AF represents the change in free energy (expressed in gram calories) and K is the equilibrium constant. If the equilibrium con- stant is expressed as: (decarboxylation product) (CO2) (carboxylated product) The equilibrium constant of some of the reversible decarboxylations is of the order of 10^, so that they proceed with a decrease in free energy of about —4000 to —5000 calories. Here are included the enzymic decarboxylation of oxalacetic acid to pyruvic acid and carbon dioxide, and that of oxalosuccinic acid to a-ketoglutaric acid and carbon di- oxide. In both cases the reaction involves a carboxyl in position /3 relative to the carbonyl group; this reaction type will be referred to here as j3-carboxylation. There is another group of enzymic decarboxylations involving simultaneous decarboxylation and dehydrogenation of a-keto acids that proceed with a much larger decrease in free energy than do /3- decarboxylations. Thus, the free energy change of reaction (1): pyruvate- + 2 H2O <^ acetate" + HCO3 " + 3 H+ + 2 e (1) 167 SEVERO OCHOA has been estimated to be —9400 cal., and that of reaction (2) —8000 cal. (8): a-ketoglutarate + 2 H2O ^ succinate-- + HCO3- + 3 H+ + 2 e (2) Reactions of this type are often referred to as oxidative decarboxylations and the reverse as reductive carboxylations. From the relationship between free energy change and equilib- rium constant discussed above, it is clear that shifts of equilibrium toward carboxylation, i. e., carbon dioxide fixation, can only be ac- complished by an input of energy into the system. We shall consider in some detail the enzymic mechanisms used by the cell for this purpose. The fundamental pattern of biological carbon dioxide fixation can be visualized in terms of the following steps: (7) Primary Fixation Reaction. Carboxylation. Since the equi- librium of the reversible reaction involved is very unfavorable, only small amounts of keto acid are formed at one time. (2) Reduction. Step 1 is followed by enzymic reduction of the keto acid to the corresponding hydroxy acid. This shifts the equilib- rium and more keto acid can be formed by step 7. (5) Reduction of the Pyridine Nucleotide Oxidized in Step 2. The second and third steps will be discussed below. (4) Dehydration, Hydration, Isomerization. Further equilibrium shifts can be brought about by secondary enzymic transformations of the hydroxy acid formed by step 2. Thus, the hydroxy acid may be dehydrated to the corresponding unsaturated fatty acid. The latter, in turn, may be hydrated in a different position of the molecule to form a new hydroxy acid isomeric with the one first formed. (5) Further Reduction. The unsaturated fatty acid formed in step 3 may undergo further reduction to the corresponding saturated acid. All the reactions concerned in the various steps just outlined are reversible. The reduction of the keto acid in step 2 is catalyzed by a specific pyridine nucleotide dehydrogenase. These dehydrogenases (27) con- sist of a protein and a prosthetic group or coenzyme that combine with 168 CARBON DIOXIDE one another, although the equiHbrium constant is preponderantly in favor of dissociation. The active group of the coenzyme is pyridine. Two such prosthetic groups are known at present, viz-, diphospho- pyridine nucleotide (abbreviated DPN, also known as cozymase and coenzyme I), and triphosphopyridine nucleotide (abbreviated TPN, also known as Warburg's coenzyme and coenzyme II). The coenzymes are dinucleotides, each containing two bases, adenine and nicotinamide, two molecules of ribose, and either two (DPN) or three (TPN) mole- cules of phosphoric acid. The structural formula of DPN is shown in scheme I. The point of attachment of the third phosphoric acid residue in TPN is as yet unknown. Most of the pyridine nucleotide dehydrogenases have DPN as the prosthetic group; only two are defi- Scheme I H CH 1 /\ N-=C— N HC C— CONH2 1 11 A i c— c— c 1 1 1 \+/ N N NH2 N CH HP HC 1 HCOH HCOH 1 1 HCOH HCOH 1 HC 1 HC H2C— O— P— O— P— O — CHj O- OH Structural formula of diphosphopyridine nucleotide CH HC C— CONH2 HC CH V R— O— P— OR' CH /\ HC C— CONH2 II I HC CHj V I H+O- 1 I R— O— P— OR' O O Reversible reduction of diphosphopyridine nucleotide 169 SEVERO OCHOA nitely known to function with TPN, and one at least can function with either coenzyme (Table I). Table I Pyridine Nucleotide Dehydrogenases Ox-redox System potential (£6; PH 7.0), volts Prosthetic group Occurrence of dehydrogenase Glutamic acid<=ia-ketoglutaric acid + NHi -0,03 DPN or TPN Yeast, bacteria, animal tissues Malic acid<=i oxalacetic acid -0.10 DPN Bacteria, plants, animal tissues Ethyl alcohol?^ acetaldehyde -0.16 DPN Yeast, bacteria Lactic acid ?=i pyruvic acid -0.18 DPN Bacteria, animal tissues 3-Phosphoglyceraldehyde + phosphate <=i 1,3-diphosphoglyceric acid -0.28 DPN Yeast, animal tissues j3-Hydroxybutyric acid <=^ acetoacetic acid -0.29 DPN Animal tissues Isocitric acid «=* oxalosuccinic acid -0.30 TPN Yeast, plants, animal tissues Glucose-6-phosphate ?^ 6-phosphoglu- conic acid -0.43 TPN Yeast, red blood cells Glucose <=^ gluconic acid -0.45 DPN Animal tissues The pyridine in the coenzymes acts by cyclic addition and re- moval of hydrogen (see scheme I). Through the reversible change pyridine ^ dihydropyridine, it functions as a hydrogen carrier. The protein component of the dehydrogenase, upon whose nature depends the specificity for its substrate, binds both substrate and coenzyme, and in this complex hydrogen from the substrate is transferred to the pyridine of the coenzyme; the substrate is thus oxidized and the pyri- dine reduced. This transfer of hydrogen is reversible, so that dihydro- pyridine and oxidized substrate when bound by the protein can react to form pyridine and reduced substrate. Thus, the action of a typical pyridine nucleotide dehydrogenase can be represented as follows: RHa + P-Go :^R + P-CoHa where RH2 and R represent reduced and oxidized substrate and P • Co and P-CoHa represent oxidized and reduced protein-coenzyme com- plex, respectively. The reduced coenzymes have a sharp absorption band at 340 m^, whereas there is no absorption at this wave length by the oxidized form. This distinction is important because the action of the pyridine nucleotide dehydrogenases can be conveniently followed by changes in the absorption of light of wave length 340 myu (27). 170 CARBON DIOXIDE It is thus dear that the biological reduction of the keto acid formed in step 7 of carbon dioxide assimilation requires the presence of the specific dehydrogenase and of the reduced form of its prosthetic group. Step 2 can then be represented as follows: keto acid + P • C0H2 t± hydroxy acid + P • Co (3) Since the complex formed by the coenzyme with the protein component of the dehydrogenases is dissociable to a relatively large degree, there are always small amounts of free coenzymes present in the cell. When equilibrium has been reached in a reaction of the type represented by reaction (3), it will stop unless provision is made for a reduction of the oxidized coenzyme formed so as to displace the equilibrium to the right. Step 3 occurs here. It is carried out through the action of another dehydrogenase which functions with the same prosthetic group as that acting in step 2. Such a reaction may be, for instance: hydroxy acid a + PrCo ?:± keto acid a + PrCoHa (4) and is possible because of the dissociable nature of the complex formed by the coenzyme with the protein components of the dehydrogenases, so that the coenzyme can alternatively be bound by either protein. Thus, some of the oxidized coenzyme dissociating from P • Co can be bound by the second protein, Pi, to form Pi -Co, as in reaction (4). Since the coenzyme oxidized in reaction (3) is reduced in reaction (4), the net result of these two reactions can be expressed by: keto acid + hydroxy acid a <=± hydroxy acid + keto acid a (5) This type of reaction is known as a coenzyme-linked dismutation, and is, in general, reversible. The extent to which it will proceed in a given direction depends on the equilibrium constants of the two dehydrogen- ase systems involved and on the concentration of reactants. Thus, in our case, a high concentration of hydroxy acid a will favor carbon dioxide fixation. We shall now consider the individual carbon dioxide fixation systems. ^-Cavhoxylaiion Carbon dioxide fixation by a-ketoglutaric acid. Although this type of carbon dioxide fixation is the most recently discovered 171 SEVERO OCHOA (20,21), it will be convenient to discuss it first because the methods used in the study of this system permit a clearer picture of the pat- tern into which the cellular mechanisms of carbon dioxide fixation can be fitted. The primary fixation reaction involves the reversal of the decarboxylation of oxalosuccinic acid to a-ketoglutaric acid and carbon dioxide (reaction la). This reaction is catalyzed by an OXALOSUCCINIC CARBOXYLASE COOH COOH CO CO CH2 + C02 — ^ HC COOH CH2 1 CH2 1 COOH COOH a-Ketoglutaric acid Oxalosuccinic acid (la) enzyme, oxalosuccinic carboxylase, present in heart muscle and prob- ably in other animal and plant tissues. The enzyme requires either magnesium or manganese ions for activity. The equilibrium of reaction la is so far to the left that the avail- able analytical methods would fail to show a formation of oxalosuccinic acid even when starting with very high concentrations of a-ketoglutaric acid and carbon dioxide in the presence of the enzyme. However, reversibility can easily be demonstrated by adding isocitric dehydro- genase (see Table I) and reduced triphosphopyridine nucleotide. When this is done, the oxalosuccinic acid formed by carboxylation is reduced to /-isocitric acid by TPNH2 which, in turn, is oxidized to TPN (reaction lb): COOH I CO ISOCITRIC DEHYDROGENASE COOH HC— COOH CH2 I COOH Oxalosuccinic acid + TPNH2 CHOH HC— COOH -f TPN I CH2 I COOH /-Isocitric acid (lb) Reaction lb is the second step of the series by which carbon dioxide is fixed in this system. The combined result of reactions la and lb is reaction Ic. 172 CARBON DIOXIDE OXALOSUCCINIC CARBOXYLASE AND ISOCITRIC DEHYDROGENASE COOH COOH i I CO CHOH CH2 + CO2 + TPNH2 , HC— COOH + TPN (Ic) CH2 CH2 I I COOH COOH a-Ketoglutaric acid /-Isocitric acid Since both reactions lb and Ic involve conversion of TPNH2 to TPN and vice versa, they can be followed spectrophotometrically in either direction by allowing the reaction to take place in a quartz cell and measuring the absorption of light at wave length 340 m^ by the test solution. As mentioned above, reduced pyridine nucleotides strongly absorb light of this wave length. The molar extinction coef- ficient, which is defined by the equation: log hi I a. — cl is 0.5644 X 10^ (cm. -/mole). For a transmittance of 95% and when / = 1 cm., the concentration of reduced pyridine nucleotide would be 0.04 X 10"'' moles per cc. In the case of TPN with a molecular weight of 743, 0.04 X 10"^ moles per cc. corresponds to 3 fxg. of TPN per cc, or 0.8 Mg- of isocitric acid per cc. This indicates the great sensitivity of the optical method and how suited it is for the study of reactions of this type. By using this method, it has been possible to determine the equilibrium constants of reactions la, lb, and Ic. The equilibrium constant of reaction lb, A"b = (/-isocitrate) (TPN)/(oxalosuccinate) (TPNH2), at/?H 7.0 and 22° C, is approximately 0.3. That of reac- tion Ic, K\ = (/-isocitrate) (TPN) /(a-ketogIutarate)(C02)(TPNH2), at the same pH and temperature is, on the average, 1.3 X lO"*. The equilibrium constant of reaction la can be calculated from these two values, since fi\ = (oxalosuccinate)/(a-ketoglutarate)(C02) = Kc/Kx, = 0.5 X 10^^ Thus, tlie equilibrium of reaction la is so un- favorable for carbon dioxide fixation that by this step alone only about 0.5% of the a-ketoglutarate would be carboxylated. SEVERO OCHOA The TPN formed in reaction Ic can be reduced through a coenzyme-linked dismutation as discussed above. This results in the shifting of the equilibrium toward the side of carbon dioxide fixation. Such a shifting has been accomplished with the glucose-6-phosphate dehydrogenase system (see Table I) which catalyzes reaction Id. The combined result of reactions Ic and Id, when a mixture of glucose-6- GLUCOSE-6-PHOSPHATE DEHYDROGENASE glucose-6-phosphate + TPN ^ 6-phosphogluconate + TPNH2 (Id) phosphate, a-ketoglutarate, and carbon dioxide is incubated with glucosephosphate dehydrogenase, oxalosuccinic carboxylase, isocitric dehydrogenase, manganese ions, and TPN, is reaction le: a-ketoglutarate + CO2 + glucose-6-phosphate <=^ /-isocitrate + 6- phosphogluconate (le) The equilibrium constant of reaction le has not yet been de- termined experimentally, but it can be calculated from free energy data. Thus, the free energy change of reaction Id can be estimated from the equation, —AF = nFAE, relating free energy change to the potential difference between two reacting oxidation-reduction systems (15). The reacting systems in reaction Id are the glucose-6-phosphate ^ phosphogluconate {Eq — —0.43 v. at pH 7.0) and the TPN ;=± TPNH2 system, the potential of which is unknown but can be con- sidered to be near that of the DPN «=± DPNH2 system (Eo = —0.28 v. at pH 7.0). For a potential difference of 0.43-0.28 = 0.15 v., the AF of reaction Id would be —6890 cal., corresponding to an equilib- rium constant K^ = (6-phosphogluconate) (TPNH2)/ (glucose-6-phos- phate)(TPN) of the order of 10^. Hence, the equilibrium constant of reaction le: K^. = (/-isocitrate) (6-phosphogluconate)/(glucose-6- phosphate) (a-ketoglutarate) (CO2) = K^X K^= \.'iX 10"" X 10^ = 13. A further shift of the equilibrium of reaction le toward carbon dioxide fixation occurs in the presence of aconitase. This enzyme is widely distributed in animal and plant cells and catalyzes the inter- conversion between /-isocitric, czV-aconitic, and citric acids according to reaction If, where the figures in parentheses give the percentage of the 174 COOH I CHOH I HC— COOH I CH2 I COOH ACONITASE COOH CH CARBON DIOXIDE COOH CH, ■HoO 11 +H2O I ^ C— COOH . HOC— COOH + H2O (If) -H2O CH2 I COOH /-Isocitric acid (7.7%) m-Aconitic acid (3.1%) CH2 COOH Citric acid (89.2%) individual components present at equilibrium at 37° C (18). Under these conditions, over 90% of the /-isocitric acid formed in step 3 is converted to <:2i--aconitic and citric acids. The free energy changes of the various steps of this system are given in Table II. Combination of the four steps gives an over-all balance of about —3000 cal., i. e., the complete system is exergonic by a fairly ample margin. Table II Carbon Dioxide Fixation by a-KETOGLUTARic Acid Step Reaction Enzyme AF, cal. Remarks (7) Carboxylation (2) Rfduction {3) Reduction of pyri- dine nucleotide (4) Over-all reaction Jor first three steps (5) Isomerization a-Ketoglutarate + CO2 <^ oxalosuccinate Oxalosuccinate + TP- NH2 <=i /-isocitrate + TPN Glucose-6-phosphate + TPN <=^ 6-phosphoglu- conate + TPNHo or-Ketoglutarate + CO2 + glucose-6-phosphate ^ 6-phosphogIuconate + /-isocitrate /-Isocitrate <=i citrate Oxalosuccinic carboxylase Isocitric dehydro- genase Glucosephosphate dehydrogenase + 4460 + 708 -6890 -1711 ca. -1500 Calc. for r = 295° from equil. const. Calc. as above Calc. from — AF = aFAE AF =■ S of AF of Aconitase partial reactions Calc. for T = 310° from equil. const. There is a possibility that the above reaction series might not stop with the formation of citric acid. Some microorganisms have been reported to split citrate to oxalacetate and acetate (2), a reaction that would favor carbon dioxide fixation via the a-ketoglutaric carboxyl- ation system by displacing the equilibrium still further. Since, as we shall see later, acetate can be converted to pyruvate by reductive carboxylation and pyruvate forms carbohydrate in cells, the biological formation of acetate from citrate would be of considerable importance. Carbon dioxide fixation by pyruvic acid. The primary re- 175 SEVERO OCHOA action of this system (5,9,1 1,12) involves the reversal of the decarboxy- lation of oxalacetic acid to pyruvic acid (reaction Ila). Reaction Ila OXALACETIC CARBOXYLASE CH, COOH i I CO + CO2 , GH2 I I (Ila) COOH CO I COOH Pyruvic acid Oxalacetic acid is catalyzed by oxalacetic carboxylase, an enzyme which is found in bacteria and liver and requires magnesium or manganese ions for activity. As in the case of oxalosuccinic carboxylase, the equilibrium of reaction Ila is very far to the left. Reversibility has been demon- strated by allowing the enzyme to act on oxalacetic acid in the presence of isotopic carbon dioxide. By stopping the enzyme action when about half of the oxalacetic acid was decarboxylated, the presence of isotopic carbon in the /S-carboxyl group was demonstrated. The equilibrium constant of reaction Ila has been calculated from its free energy change, in turn calculated from the free energies of formation (at 38 ° C.) of the substances involved (5) : oxalacetate + H2O > pyruvate" -\- HCOa" -184,210 cal. -56,200 cal. -106,460 cal. -139,200 cal. AF thus calculated is —5250 cal. for the decarboxylation, and A"a = (oxalacetate — )/ (pyruvate") (HCO3-) = 0.2 X 10"^ a value of the same order of magnitude as that of reaction la. Step 2 occurs when both malic dehydrogenase (see Table I) and reduced diphosphopyridine nucleotide are present, since the oxalacetic acid formed by reaction Ila is then reduced to /-malic acid (reaction lib). MALIC DEHYDROGENASE COOH COOH CH2 CH2 I + DPNH2 , I + DPN (Hb) CO CHOH COOH COOH Oxalacetic acid /-Malic acid 176 CARBON DIOXIDE The combined result of reactions Ila and lib is reaction IIc; OXALAGETIC CARBOXYLASE AND MALIC DEHYDROGENASE CH, COOH I I CO + CO2 + DPNH2 _1 CH2 + DPN (lie) COOH Pyruvic acid CHOH 1 COOH /-Malic acid In step 3, the DPN formed in reaction IIc can be reduced by lactic acid through a coenzyme-Hnked dismutation with the lactic de- hydrogenase system. Lactic dehydrogenase catalyzes reaction lid. CH, I CHOH + DPN ; LACTIC DEHYDROGENASE CHs COOH /( + )-Lactic acid CO + DPNHj I COOH Pyruvic acid (Hd) The combined result of reactions IIc and lid is reaction He. pyruvate + CO2 + /( + )-lactate ^ /-malate + pyruvate (He) The free energy change of this reaction (see Table III), is about +1500 Table III Carbon Dioxide Fixation by Pyruvic Acid Step Reaction Enzyme AF, cal. Remarks (7) Carboxylation Pyruvate -j- CO2 «=^ oxal- acetate Oxalacetic car- boxylase + 5250 Calc. from free energies of forma- tion Calc. from — AF = nFAE Calc. from — AF ■= nFAE AF = 2 of AF of (2) Reduction {3) Reduction oj pyri- dine nucleotide Over-all reaction (A) (4) Dehydration (5) Further reduction Over-all reaction (B) Oxalacetate + DPNH2?=i /-malate + DPN /(+)-Lactate + DPN ?ii pyruvate + DPNHj Pyruvate + CO2 -t- /(-i-)- lactate <=^ /-malate -|- pyruvate[C02-i-/(-f-)- actate?^ /-malate] /-Malate ?=i fumarate -)- H2O Fumarate -|- 2 H «=i suc- cinate Pyruvate + CO2 -f /(-!-)- lactate -{- 2 H ♦^i suc- cinate 4- pyruvate -|- H2O [/(-f Vlactate -f CO2 + 2H ^ suc- cinate + H2O] Malic dehydro- genase Lactic dehy- drogenase -8300 + 4600 +l}iO + 705 -20,450 -11,19} Fumarase Hydrogenasc; fumaric re- ductase (?) partial reactions Calc. from equil. const. Calc. from — AF ■• nFA£ AF - S of AF of partial reactions 177 SEVERO OCHOA cal. and the calculated equilibrium constant, K^ = (/-malate)/(/(+)- lactate)(C02), approximately 0.1. Step 3 could also involve dismutation with any other diphos- phopyridine nucleotide dehydrogenase system having an oxidation- reduction potential lower than that of the malic system (see Table I). We have considered the lactic dehydrogenase in this connection because there is evidence that it can participate in carbon dioxide fixation by pyruvic acid. In the presence of the enzyme fumarase, part of the /-malic acid formed in reaction He would be dehydrated to fumaric acid. Fumarase is widely distributed in plant and animal cells. Re- action lie would then be replaced by reaction Ilf: pyruvate + CO2 + /(+) -lactate ^ fumarate + H2O -f pyruvate (Ilf) Experimental support for the occurrence of the over-all reaction (Ilf) is gained from the observation that fumarate, when added to an enzyme preparation from liver in the presence of pyruvate, DPN, and manganese ions, is converted to lactate and carbon dioxide (5). This indicates that reaction Ilf can proceed from right to left. That it also proceeds from left to right is indicated by the presence of isotopic carbon in the carboxyl groups of the residual fumarate when the reaction is carried out in presence of isotopic carbon dioxide (29). The pyruvic oxalacetic system of carbon dioxide fixation has not yet been investi- gated by the methods used in the study of the ketoglutaric-oxalosuccinic system. A considerable shift of the equilibrium of reaction Ilf in the direction of carbon dioxide fixation can be brought about by reduction of the fumarate to succinate, a reaction that occurs in bacteria and liver tissue. In fermentation of glucose or glycerol by propionic acid bacteria, succinate is found to be one of the end products. Experi- ments with isotopic carbon dioxide have shown that the carboxyl groups of succinic acid become labeled. The carboxyl groups of mal- ate, fumarate, and succinate, formed by pyruvate fermentation with Escherichia coli in the presence of carbon dioxide containing isotopic carbon, also show excess of heavy carbon. This is also the case when pyruvate is incubated with liver preparations. Some bacteria, e. g., E. coli, can use molecular hydrogen for fumarate reduction and, in the 178 CARBON DIOXIDE presence of pyruvic acid, carbon dioxide, and hydrogen, all three are utilized to form succinic acid (9,10,28). The reduction of fumarate to succinate is strongly exergonic and provides ample energy to drive carbon dioxide fixation by this system to completion (see Table III). Reductive Carhoxylation This type of carbon dioxide fixation has been discussed recently by Lipmann (17) and will only be briefly considered here. The fixa- tion is a consequence of the reversal of the oxidative decarboxylation of a-keto acids catalyzed by specific enzymes. These reactions involve inorganic phosphate, and lead to thdf formation of an anhydride of the next lower fatty acid and phosphoric acid with liberation of either formic acid, or carbon dioxide and hydrogen; the hydrogen may either ap- pear as molecular hydrogen or reduce a hydrogen acceptor (see re- actions Ilia and 1 1 lb). The anhydride bond formed has a high energy content and is generally referred to as an energy-rich phosphate bond. There are other types of energy-rich phosphate bonds of biological im- portance, such as enol phosphate, guanidine phosphate, and pyro- phosphate bonds. The pyrophosphate type of bond has a special significance because, by the action of specific enzymes, it can give rise to any of the other phosphate bonds. Since the reactions involved in these conversions are reversible, it follows not only that any energy- rich phosphate bond can generate pyrophosphate bonds, but also that the various types of bonds can be converted into one another through the intermediate formation of pyrophosphate linkages. The biologically important pyrophosphate group is the one present in adenosine polyphosphates, that is, adenosine triphosphate (abbreviated ATP), and adenosine diphosphate (abbreviated ADP). N--CNH2 HG C— N I \^Tj OH OH OH OH OH I /^^ II III N— C— N CH— CH— CH— CH— CH2— O— P— O— P— O— P— OH I I II II II I o 1 000 Adenosine triphosphate By enzymic hydrolysis, ATP is dephosphorylated to ADP, and this, in turn, to adenosine monophosphate (adenine ribose 5-phosphate or SEVERO]]OCHOA muscle adenylic acid), with release of inorganic phosphate and of the energy of the bond. By enzymic transphosphorylation, ATP trans- fers phosphate to carboxyl, enol, and guanidine groups in a reversible manner. Since the energy content of the various types of energy-rich phosphate bonds is nearly the same, in the neighborhood of 12,000 cal., these transphosphorylations involve relatively small changes of free energy (16). The general type of oxidative decarboxylations can be repre- sented by reaction Ilia (17), where X stands for a hydrogen acceptor. GH3COOPO3- + CO2 + H2X (Ilia') cHaCO : cooH + H : opo,- Pyruvic acid ± CH3COOPO3— + CO2 + H2 (Ilia") CH3COOPO,-- + HCOOH Acetyl phosphate (Ilia ) The oxidative decarboxylation of a-ketoglutaric acid generates phos- phate bonds (19) and may be represented by reaction Illb. All these reactions occur in bacteria; reactions Ilia' and Illb also occur in ani- mal tissues. C00HCH2CH2C0 : COOH + H : OPO,— + x a-Ketoglutaric acid COOHGH2CH2COOPO,-- + CO2 + H2X (Illb) Succinyl phosphate Two reactions are known at present through whose reversibility reductive carboxylation can take place: (a) the splitting of formic acid to carbon dioxide and hydrogen; and (b) the splitting of pyruvic acid to acetyl phosphate and formic acid. Since acetyl phosphate can be formed enzymically by a reversible reaction between acetic acid and ATP, we have a biological system of reductive carboxylation of acetate to pyruvate. The pattern of carbon dioxide fixation established for /3-car- boxylations must be modified here to include two preliminary steps, viz., those of phosphorylation and of carbon dioxide reduction, respec- tively, and a final step for regenerating the ATP used at the beginning. 180 CARBON DIOXIDE A possible series of reactions, modified from Lipmann, is suggested in Table IV. The absolute value of the free energy change of step 6 Table IV Reductive Carboxylation Step Reaction Enzyme AF. calc. Remarks (7) Phosphorylation Acetate + ATP;(=i acetyl From Clostridium -1-3000 Calc. from equil. phosphate + ADP CO2 + H2<=i HCOOH butylicum const. (2) Carbon dioxide re- Formic hydro- - 200 Calc. from equil. duction genylase (£. coli) From E. coli const. (3) Carboxylation Acetyl phosphate + for- -F2800 Calc. from equil. mate «=i pyruvate -f- const. phosphate (4) Reduction oj car- Pyruvate + DPNH2 ^ Lactic dehydro- -4600 Calc. from — ^F — boxylation product I ( + ) -lactate + DPN genase nFAE (5) Reduction oj pyri- 3-Phosphoglyceraldehyde 3-Phosphoglycer- aldehyde de- ca. Calc. from equil. dine nucleotide + phosphate 4- DPN -t- 400 const. <=i 1,3-diphosphoglyc- hydrogenase erate + DPNHi (6) Regeneration of 1 ,3-Diphosphoglycerate From yeast -3000 See text ATP + ADP ?=i 3-phospho- glycerate -(- ATP Over-all reaction Acetate + COj -f H2 -f 3-phosphoglyceralde- ca. -1600 hyde<=i/( + )-lactate -f 3-phosphoglycerate has been considered to be the same as that of the step 7, on the assump- tion that the bond energy of the anhydride linkage in 1-phosphoglycer- ate is the same as that of the acetyl phosphate group. The reaction series postulated in Table IV would lead from acetate to lactate and the over-all reaction would be exergonic. A characteristic feature of reductive carboxylation is that it must be started by input of a large amount of energy from energy-rich phosphate bonds. Pyruvic acid formed by reductive carboxylation of acetate might also be converted to carbohydrate after reduction to triose in- stead of being reduced to lactic acid. Such a conversion would re- quire additional phosphate-bond energy which might be amply avail- able in chemosynthesis or photosynthesis. It is not unlikely that reactions Ilia' and Illb might also be reversible. The main difference between these reactions and reactions Ilia" and Ilia'" is one of mechanism, i. e., the hydrogen from the keto acid is first transferred to a hydrogen acceptor instead of being released either as molecular hydrogen or in combination with carbon dioxide as formic acid. If such were the case, all of the intermediate reactions involved in the oxidation of foodstuffs (i. e., in respiration) would be reversible. i8i SEVERO OCHOA Carbon Dioxide Fixation in Chemosynthesis and Photosynthesis It is now known that photosynthesis can be divided into two phases relatively independent of one another: (1) a so-called "dark" reaction which occurs in the absence of light and consists of a reversible fixation of carbon dioxide to form a carboxylic acid; and (2) the photo- lytic fission of water, yielding hydrogen with liberation of oxygen. Hydrogen produced by photolysis is used to reduce the products formed by carboxylation. In chemosynthesis, hydrogen is obtained by oxidation of inorganic compounds — a process that also supplies energy (6,7,23-25). It would appear that essentially the same mechanisms that func- tion in carbon dioxide fixation by heterotrophic organisms are operative in both photosynthesis and chemosynthesis, with the difference that, in the case of the heterotrophs, hydrogen and energy are derived by oxidation of organic materials. The widespread occurrence in plants of di- and tricarboxylic acids (malic, citric, isocitric) and of the enzymes that participate in their metabolism (fumarase, malic dehydrogenase, aconitase, isocitric dehydrogenase) lends support to such a view. Phos- phorylation processes which, as we have seen, are essential for reductive carboxylations, are connected with carbon dioxide fixation in the sulfur oxidizing autotroph Thiobacillus thiooxidans, and photosynthetic organ- isms may utilize radiant energy for the synthesis of energy-rich phos- phate bonds (3,22,26). We cannot yet formulate in any detail the course of events in photosynthesis and chemosynthesis, but, using the knowledge gained by the study of the mechanisms of carbon dioxide fixation in hetero- trophic organisms, we may attempt to draw a plausible picture of the chemical events. Reversal of the oxidative degradation of foodstufTs, i. e., of respiration, would now seem to be a definite possibility. We have seen that, on carboxylation and reduction, a-keto- glutaric acid can be converted to citric acid, and have indicated that the latter may be split to acetic and oxalacetic acids. Further, oxal- acetic acid can be reduced to succinic acid by way of malic and fumaric acids, and succinic acid could be converted to a-ketoglutaric acid by reductive carboxylation, i. e., by reversal of reaction Illb. In this way, a-ketoglutaric acid would be regenerated, while the acetic acid 182 CARBON DIOXIDE formed earlier can be converted to pyruvic acid by reductive carboxyl- ation, i. e., by reversal of reaction Ilia. We would thus have a cyclic mechanism whereby carbon dioxide and hydrogen entering at various points would emerge as pyruvic acid. The di- and tricarboxylic acids would only act catalytically as carriers of carbon dioxide and hydrogen. This is a reversal of the so-called tricarboxylic acid cycle, which is considered to be an important pathway for the oxidative breakdown of carbohydrate and fat in cells. Scheme II presents this metabolic cycle, incorporating recent findings concerning some of the intermediate reactions (1,4,13,28). For a more detailed discussion, see Lardy and Elvehjem (14). Scheme n Reversible Tricarboxylic Acid Cycle Carbohydrate fatty acids pyruvate ^ lactate malate -H2O czi'-aconitate. +H2O citrate +H2O -H2O -H:0 fumarate -HK5 +H2O isocitrate -2H + 2H + 2H ■2H succinate oxalosuccinate •2H -CO2 + 2H + CO2 + CO2 -CO2 a-ketoglutarate Another mechanism recently suggested for photo- and chemo- syntheses involves a sequence of carboxylations and reductions leading 183 SEVERO OCHOA from a carboxylic acid, through the next higher a-keto acid, to an a-hydroxy acid, which in turn would be carboxylated to the next higher Q;-keto-/3-hydroxy acid, and so on. Such a sequence would be es- sentially a reversal of a pathway of carbohydrate oxidation by way of phosphohexonic acid, a-ketophosphohexonic acid, phosphopentonic acid, etc., through alternating dehydrogenations and decarboxylations (17). Obviously, the above schemes are only gross approximations. The main point is that what we know about the mechanisms of carbon dioxide assimilation by heterotrophic organisms strongly suggests that all reactions involved in cellular respiration are essentially reversible. Thus, the processes of both photosynthesis and chemosynthesis may rep- resent reversals of the respiratory process not only from the standpoint of energy but also from the standpoint of the enzymic mechanisms. References (1) Buchanan, J. M., Sakami, W., Gurin, S., and Wilson, D. W., J. Biol Chem., 159,695 (1945). (2) Deffner, M., Ann., 536, 44 (1938). (3) Emerson, R. L., Stauffer, J. F., and Umbreit, W. W., Am. J. Botany, 31, 107 (1944). (4) Evans, E. A., Jr., and Slotin, L., J. Biol. Chem., 141, 439 (1941). (5) Evans, E. A., Jr., Vennesland, B., and Slotin, L., J. Biol. Chem., 147, 771 (1943). (6) Franck, J., and Gaffron, H., in Advances in Enzymology, Vol. I. Inter- science, New York, 1941, p. 199. (7) Gaffron, H., Biol. Rev. Cambridge Phil. Soc, 19, 1 (1944). (8) Kalckar, H. M., Chem. Revs., 28, 71 (1941). (9) Kalnitsky, G., and Werkman, C. H., Arch. Biochem., 4, 25 (1944). (10) Kalnitsky, G., Wood, H. G., and Werkman, C. H., Arch. Biochem., 2, 269 (1943). (11) Krampitz, L. O., and Werkman, C. H., Biochem. J., 35, 595 (1945). (12) Krampitz, L. O., Wood, H. G., and Werkman, C. H., J. Biol. Chem., 147,243(1943). (13) Krebs, H. A., in Advances in Enzymology, Vol. III. Interscience, New York, 1943, p. 191. (14) Lardy, H. A., and Elvehjem, C. A., Ann. Rev. Biochem., 14, 1 (1945). (15) Lewis, G. N., and Randall, M., Thermodynamics and the Free Energy of Chemical Substances. McGraw-Hill, New York, 1923. 184 CARBON DIOXIDE (16) Lipmann, F., in Advances in Enzymology, Vol. I. Interscience, New York. 1941, p. 99. (17) Lipmann, F., and Tuttle, L. C, J. Biol. Chem., 158, 505 (1945). (18) Martius, C, and Leonhardt, H., Z- physiol. Chem., 278, 208 (1943). (19) Ochoa, S., J. Biol. Chem., 155, 87 (1944). (20) Ochoa, S., J. Biol. Chem., 159, 243 (1945). (21) Ochoa, S., and Weisz-Tabori, E., J. Biol. Chem., 159, 245 (1945). (22) Ruben, S., J. Am. Chem. Soc, 65, 279 (1943). (23) Ruben, S., and Kamen, M. D., J. Am. Chem. Soc, 62, 3451 (1940). (24) Van Niel, C. B., in Advances in Enzymology, Vol. I. Interscience, New York, 1941, p. 263. (25) Van Niel, C. B., Physiol. Revs., 23, 338 (1943). (26) Vogler, K. G., and Umbreh, W. W., J. Geji. Physiol., 26, 157 (1942). (27) Warburg, O., Ergeb. Enzymforsch., 7, 210 (1938). (28) Werkman, C. H., and Wood, H. G., in Advances in Enzymology, Vol. II. Interscience, New York, 1942, p. 135. (29) Wood, H. G., Vennesland, B., and Evans, E. A., Jr., J. Biol. Chem., 159, 153 (1945). 185 13 HORMONES B. A. HOUSSAY, director of the institute of biology and EXPERIMENTAL MEDICINE, BUENOS AIRES Definition and Significance 'HE HORMONES are specific chemical substances pro- duced by an organ or tissue which, after being discharged into the circulating fluids (milieu inter ieur), may reach all parts of the organism and in small amounts markedly influence the functions of other organs or systems without themselves contributing important quantities of matter or energy. This definition diff"erentiates them from other important sub- stances which also reach the circulating fluids, such as: (a) nutritive substances which supply the tissues with materials and energy as, for example, glucose, amino acids, lipids, etc.; (b) vitamins, organic chemi- cal regulators contained in the food; (c) chemical mediators of nerve action, liberated by the nerve endings in the close vicinity of effector or other nerve cells and which exert a localized action; {d) the organizers, embryonic substances of regional origin which govern the differentia- tion of a determined organ, even if transplated to another zone or culti- vated in vitro; (e) the parhormones (Gley) which, although being excre- tory substances produced by the metabolic processes of all the tissues, nevertheless perform important regulatory functions, as is the case for carbon dioxide in the regulation of respiration. The action of what we today call glands of internal secretion is due to the hormones they produce; and the insufficiency of these glands, therefore, is only a matter of hormone deficiency. 187 B. A. HOUSSAY In a wider sense, which is not usual, internal secretion would mean any specific cellular elaboration discharged into the internal medium. It is in this sense that Claude Bernard considered glucose, which is produced in the liver and then passes into the blood, as an internal secretion. As we all know, the glands of internal secretion are referred to as such because they pour their elaborated products into the blood, in contrast to the glands of external secretion which pour their products out of the body or in cavities which communicate with the external medium. The first demonstration of a hormonal action was the induction of the comb of a capon by testicular graft (Berthold, 1849). The ex- pression "internal secretion" was used for the first time by Claude Bernard (1855) to point out that the liver "shed" sugar in the blood. The concept of internal secretions as we now understand it is due to Brown Sequard who, in 1891, stated in a paper with Arsonval, "Nous admettons que chaque tissue, et plus g^n^ralement, chaque cellule de I'organisme secrete pour son propre compte des produits ou des ferments sp^ciaux qui sont versus dans le sang et qui viennent influencer par I'interm^diaire de ce liquide toutes les autres cellules rendues ainsi solidaires les unes des autres par un m^chanisme autre que le tissue nerveux." The name "hormones" was proposed by Hardy to desig- nate the "chemical messengers" (Bayliss and Starling, 1904) which are secreted in the blood by one organ to stimulate the functions of another. Hormone etymologically means "I arouse the activity" or "I excite," but we are now aware of the existence of inhibitory hormonal actions. It is useless to classify the hormones as stimulating or inhibi- tory, inasmuch as the same hormone may often produce both effects, depending upon its concentration or the organ which it affects. It is understandable, therefore, why the designations proposed by Sharpey- Schafer, who gave them the general name of "autacoids" and sub- divided them into "hormones" (exciting) and "chalones" (inhibitory), did not gain currency. Chemical Nature According to their chemical nature the known hormones can be classified into three groups: (7) Phenolic Derivatives. — Both adrenalin, the hormone of the 1 88 HORMONES adrenal medulla, and thyroxine, the hormone of the thyroid, are phenolic substances. (2) Proteins. — To the proteins belong the hormones of the hypophysis (two gonadotropins, thyrotropin, adrenocorticotropin, lactogenic hormone and growth-promoting factor from the anterior lobe, and the vasopressor, oxytocic and melanic-expanding principles of the posterior lobe); the hormone of the pancreas (insulin), and the hormone of the parathyroid (parathormone). (J) Steroids. — To the steroid family belong the hormones isolated from the ovarium and corpus luteum, testicle, and adrenal cortex, twenty-eight steroids having been isolated so far from the adrenal cortex (Reichstein, 1943). They all have in common a cyclo- pentanoperhydrophenanthrene ring system. Attempts to prepare the nonprotein hormones synthetically have followed closely the determination of their chemical constitution. Besides, several synthetic substances of constitution similar to, but not identical with, the natural hormones have been prepared, and their pharmacologic action tested. Many vasoconstrictor and bronchodila- tor substances have thus been obtained which are more efficient than adrenalin for some therapeutic uses. Several synthetic estrogens (di- ethylstilbestiol, hexestrol, etc.) have also been obtained which are now largely used instead of natural estrogens to treat ovarian insufficiency. Desoxycorticosterone, a substance which is able to maintain adrenalec- tomized animals in normal condition, has been prepared by partial synthesis. Probably a wide field for pharmacological investigation of corticoadrenal hormones will soon be opened, since Reichstein (1943) announced a method of preparing steroids containing an atom of oxy- gen linked to carbon atom 1 1 . Steroid hormones are usually prepared by partial synthesis, starting from stigmasterol, a product of soybeans. The active substances extracted from an endocrine gland do not always represent the circulating natural hormone. Thus, while thy- roxine increases the oxygen consumption if given to the whole animal, it does not have such an action upon isolated tissues. Furthermore, whereas only from 25 to 50% of organically bound iodine is extract- able from the thyroid in the form of thyroxine, all the iodine compounds of the thyroid are able to increase the rate of metabolism. These and other reasons make it dubious that thyroxine is a constituent of the thyroid hormone or the hormone itself. 189 B. A. HOUSSAY Some of the protein hormones elicit such a small antibody re- sponse, as in the case of insulin, that their administration by repeated injections remains effective during scores of years. Other protein hor- mones become progressively less effective, which makes it necessary to increase the dose each time, as in the case of the parathyroid hormone. There are still other hormones (thyrotropin, gonadotropin, and in different degrees all the anteropituitary hormones) whose action de- creases quickly if they are administered daily and which induce the formation of antihormones. Thus, if an animal is treated with daily doses of thyrotropin, it shows, at the beginning, hypertrophy and hyper- function of the thyroid, but after a few weeks the action disappears and is followed by atrophy and hypofunction of the organ. The serum is then found to contain antithyrotropin, which not only inhibits the thyrotropin action but is also capable of inhibiting the action of the thyroid gland as shown by injecting such serum into another animal. Antihormones are only produced when protein hormones are administered parenterally. Gollip (1934) thought that they were sub- stances of physiological importance and that each hormone should have its corresponding antihormone to balance its effects. But it now seems that antihormones are antibodies or immunity mechanisms (Rowlands) reacting to injection of antigens from another species. It is to be noted that adrenalin, thyroxine, and the steroid hormones do not produce antihormones and are not proteins. For some of the actions of these hormones a certain habit may be produced without any demonstrable antihormones. The natural protein hormones do not seem to be completely equivalent to those which are extracted in the laboratories, for they do not induce the formation of antihormones. Thus, when a rat is castrated, great quantities of gonadotropins accumulate in its blood, as is readily demonstrated in parabiosis experiments — that is to say, experiments involving sewing two rats together side by side, after opening their bellies laterally, so that their peritoneal cavities, muscles, skin, and blood vessels are fused. In this way, the gonadotropin pres- ent in the blood of the castrated rat passes into the circulation of the normal rat, which, as a consequence, shows an intense stimulation of the gonads. This stimulation persists steadily for months, while para- biosis lasts, with no sign of antihormone production, in contrast to the 190 HORMONES rapid formation of antihormones and the inversion of effects caused by the injection of gonadotropins isolated from the gland. The preparation of active protein hormones that will not produce antihormones is one of the outstanding problems endocri- nology must solve. Meanwhile, the formation of antihormones imposes an important limitation on prolonged therapeutic application of certain hormones, particularly those of the pituitary and parathyroid. This is one of the causes of the discrepancy existing between, on the one hand, the great functional importance of the hypophysis as demon- strated by experiment and by the study of disease in man, and, on the other, the limited possibilities realized thus far from therapeutic ap- plications. The chemical mechanisms by which hormones act upon cells are not yet known. It has not yet been proved that they participate in enzyme systems, as is the case for vitamins such as thiamin, niacin, and riboflavin. The hormones are chemical regulators that probably modify some link in the chain of metabolic reactions, a field of study which has remained almost unexplored to date, in spite of its great importance. Role Some endocrine organs, phylogenetically, begin as external secretion glands that cast their secretions into the digestive system but afterward lose their excretory function and change into internal secre- tion glands. The pituitary, thyroid, and pancreatic islets of Langer- hans have evolved in some such way. The parathyroid, regulator of the metabolism of calcium and phosphorus, derives from the branchial arches. The ovarium, testicle, and adrenals, which produce the steroid hormones with sexual and metabolic actions, derive from the coelomic epithelium. The adrenal medulla that secretes adrenalin derives from the nervous sympathetic system, and the neurohypophysis from the diencephalon. The hormones of the vertebrates are better known than those of the invertebrates. In the latter, certain processes have been shown to be regulated by hormones, such as metamorphosis, color (in the case of Crustacea), and sexual dimorphism in some instances. The hormones regulate functions that exist before hormones appear and which often persist without them. Thus, all animal and plant cells consume glu- 191 B. A. HOUSSAY cose without the intervention of insulin, but in the great majority of the vertebrates insuhn is a new regulating mechanism of such impor- tance that, when missing, diabetes results, a disease which is fatal sooner or later depending upon the species. Sexuality exists in many inverte- brates without intervention of hormones, but in the vertebrates sexual characters do not reach their full development without the action of ovarian or testicular hormones, the production of which is governed by the pituitary gonadotropins. The functional unity of the organism is assured by nervous and humoral (chemical) mechanisms of correlation, which interrelate the different parts or tissues and regulate their recipro- cal activities. Sherrington has pointed out the unifying integrative action of the nervous system; a similar role is played by the humoral factors, among which are the hormones. Both types of mechanism maintain the stability of the milieu interieur (CI. Bernard) and of the organism as a whole (Cannon's homeostasis) in spite of varying condi- tions in the external environment and in the organism's activity. Modern studies have thus confirmed, extended, and given a more pre- cise meaning to the old and vague notions about the correlations be- tween the organs ("consensus partium" or "sympathies"). The roles played by the hormones can be classified somewhat conventionally as follows: (7) Metabolism. — Some of the hormones regulate the balance of metabolic processes. Their action may be general (stimulation of oxidation processes by thyroxine) or rather specialized (parathyroids and calcium, insulin and carbohydrate). One and the same hormone may modify several metabolic processes, e. g., adrenal steroids act both upon the metabolism of water and salt, and upon the metabolism of carbohydrates. (2) Morphogenesis. — The morphogenetic actions are the con- sequence of the selective role of the hormones in assimilation and growth phenomena. Some endocrine glands such as the pituitary, thyroid, parathyroid, and sexual glands play an important role in growth during a certain stage of development, principally because of their action upon the synthesis of proteins and on the development of bone. In other cases the growth-promoting action is exerted on special organs, as in the case of estrogens which promote the growth of the uterus and mammary gland. As the result of growth and differentiation, the proper morpho- 192 HORMONES logical constitution of each sex and of each individual is attained. The pars distalis of the pituitary gland stimulates the thyroid of tad- poles, and in turn the thyroid secretion promotes the metamorphosis of the larval form. The pars intermedia of the pituitai'y, and sometimes adrenalin, regulate the color of the skin of amphibians and fishes, by dispersing or concentrating the pigment gianulcs in the chromato- phore cells. (3) Endocrine Interrelation and Balance. — A close functional re- lation exists between the endocrine glands. Following Gley, we can, consider "correlation" as the relation of one organ with another; when it is a mutual correlation, we may call it "interrelation." Thus, while the secretion of pancreatic juice caused by secretin is an example of "correlation" between the duodenum and the pancreas, there exists an interrelation between the hypophysis and the gonads, for if the anterohypophysis is responsible for the final development and main- tenance of the functional activity of the gonads, at the same time the endocrine secretion of the latter regulates and moderates the gonad- stimulating function of the hypophysis. Analysis of each separate endocrine gland or of each hormone is clearly artificial. It was initially necessary because in order to study the behaviour of a gland, there were no procedures available other than extirpating the organ, injecting its extracts, and carrying out anatomical and functional studies of clinical cases showing visible altera- tions of some gland. Afterward the method persisted for didactic reasons; but since no endocrine gland can be considered as if its action were independent of other glands, the method is now being abandoned. A constellation of endocrine glands exists whose central organ is the hypophysis, the function of each gland being influenced more or less by the function of the others. Because the anterohypophysis con- tributes to the development and maintenance of the structure and function of various glands, extirpation of the pituitary produces a marked atrophy and hypofunction of the thyroid, gonads, and adre- nals, to such an extent that it has been said that the hypophysectomized rat is an endocrinological and metabolic ruin. But, at the same time, the structure and function of the anterohypophysis is governed by hormones secreted by the thyroid, gonads, and adrenal cortex. Each gland produces very specific hormones which play im- portant roles; nevertheless these hormones constitute simple parts of B. A. HOUSSAY complex functional mechanisms. Thus, the regulation of the metabo- lism of carbohydrates involves the liver, insuHn, hypophysial, cor- ticoadrenal, and thyroid hormones, as well as participation of the in- testine, kidney, and muscle. Sexual functions depend upon the com- bined action of the hormones of the hypophysis, ovary, corpus luteum, and placenta (during pregnancy), plus the action of still other glands, such as the adrenal cortex and the thyroid. As explained further on, in any single function of the organism more than one hormone plays its part, even when one of them exhibits a very specific and preponderant role. (4) Sexuality and Reproduction. — Among the endocrine actions, the sexual functions are especially important. These functions depend on the hormones of the gonads, governed in part by the pituitary go- nadotropins, in part by the hormones of the adrenal cortex and, to a cer- tain degree, by the thyroid. They regulate the production of the fe- male (ovule) and male (spermatozoid) germinal cells or the develop- ment of the organs which carry them, the development of the impulses and sexual acts that lead to fecundation, the progress of pregnancy and delivery, and the secretion of milk. The sexual hormones are mdis- pensable links for individual sexuality and for the maintenance of the species. (5) Mental and Nervous Functions. — The hormones influence several nervous and muscular activities. We only need to remember the differences shown by the sexes, the mental dullness due to hypo- thyroidism, the nervousness or mental instability of hyperthyroidism, the tetany of hypoparathyroidism, etc. These actions of the hormones are probably due to their influence on the metabolism of the nervous system, a point that has not received much study. (6) Vital Role. — The extirpation of some endocrine glands causes death. This was attributed to hypothetical poisons which ac- cumulated because they were not neutralized by the endocrine glands (antitoxic role), a theory that has been abandoned because no poison has been demonstrated and because it has been proved that death was due to metabolic disturbances provoked by the lack of the hormones normally produced by the organ in question. (7) Resistance. — The hormones are important factors in build- ing up the resistance of the organism to certain hazards such as high or low temperatures, low or excessive oxygen tension, overexercise, in- HORMONES fections, toxins, trauma, various poisons, etc. For example, pituitarec- tomized or adrenalectomized animals show a low resistance to all of these, and also are very sensitive to hypoglycemia and hypotensor agents or to the circumstances which provoke shock or hypothermy. Many of these responses depend upon metabolic phenomena. Some endocrine factors also influence the production of immunity antibodies or anaphylaxis and resistance to some infections. The harmful agents or circumstances (e. g., cold, fatigue, toxins, etc.) all produce similar reactions in a given organism, but the nature of these reac- tions varies in different species. The adrenals are largely responsible for these reactions, which pass through several phases: an initial "alarm reaction" followed by temporary compensation, and finally, decompensation. They have been thoroughly studied by Selye. Abnormal internal secretions. There are internal secretions produced under abnormal conditions. Thus, when the arterial hyper- tension is produced because the ischemic kidney produces renin, which acts enzymically upon the hypertensinogen of plasma to form hypertensin, a substance which increases the arterial blood pressure. During arterial hypotension, renin is secreted in the blood; but it has not yet been demonstrated with certainty that there is normally a small secretion of renin. Hormones and cancer. In certain cases, hormones enhance the development of tumors. Their role seems to consist more in pro- moting the growth of tumors than in initiating the malignant cellular transformation. The extirpation of glands (hypophysis, adrenals, gonads, etc.) or the injection of hormones may hasten or delay the growth of several types of tumors. Thus, testicular castration and injection of estrogenic hormones retard the development of prostatic cancer, whereas testicular hormone accelerates it. Some hormones promote the development of cancer. Thus, mammary cancer is less frequent in males or in castrated females of a strain of rat showing high incidence of this cancer in the adult female. The injection of estrogens (Lacassagne) stimulates growth of the mam- mary gland and produces cancer in many of the males of that strain. In these cases, it is debatable whether the estrogens initiate the cancer or merely promote its development. But recent work has shown that estrogenic induction of mammary cancer frequently is obtained in strains of rat in which the cancer occurs spontaneously only in rare B. A. HOUSSAY instances (Geschickter and Byrnes). Some estrogenic hormones have curious tumorigenic actions on the guinea pig, while other hormones can prevent this action (Lipschiitz). Benign and malignant tumors of the endocrine glands can pro- duce exaggerated quantities of hormones, as shown in cases of hyper- thyroidism, hyperparathyroidism, hyperinsulinism, acromegalia or giantism, and some adrenal, ovarian, and testicular tumors. Complex effects result from the action of adrenal tumors: They may exert either virilizing or feminizing actions, influence metabolic phenomena, alter the body shape, or modify arterial blood pressure. The ovarian tumors also produce varied effects such as feminizing or virilizing actions. Less familiar is the humoral mechanism by which the malignant tumors af- fect the metabolism of the whole organism. Specificity Endocrine glands and hormones have considerable specificity of action. The ovary stimulates the development of the feminine sexual characters, and the testicle that of the male sexual characters, as can be demonstrated by castration, which prevents their development or provokes their regression in both sexes. Conversely, the ovarian graft, or the administration of estrogenic hormones, develops the feminine characters in either castrated or entire animals. In the same manner, testicular grafts or androgenic hormones develop the masculine char- acters in either male or females, whether castrated or not. Each hormone produces its characteristic action upon diverse animal species. Thus, the insulin of a given animal produces hypo- glycemia in all mammals. It is also a rule that the hormone extracted from an animal is active in all other species of mammals. Thus, insulin extracted from a variety of mammals produces hypoglycemia in the rabbit, and insulin from bovine origin is active on all the verte- brates on which it has been tried. But some rare exceptions are known. Thus, gonadotropins from mammals have no effect on the gonads of the toad, Bufo arenarum, the cause of the anomaly being unknown. Some hormones are elaborated by more than one gland. For example, adrenalin is secreted by the chromaffin tissue but also exists in the cutaneous poison of the toads. The estrogenic hormones are found in the ovary, placenta, and adrenal cortex; and it is remarkable 196 HORMONES that one of the most abundant sources for industrial production is the urine of the staUion (in which it decreases after castration). The urine of men and women also contains estrogens and androgens. The andro- genic hormones, secreted normally by the testicle, can be secreted by an oVary grafted in the ear (Hill) if the ear is maintained at a low tempera- ture. They are also produced by some adrenal tumors and some ovarian tumors (arrhenoblastoma). The specificity of response of the reactive organs is not absolute. Androgens can produce some efTects on the endometrium, the vaginal epithelium, or the mammary gland. Estrogens can slightly influence the seminal vesicles. The injection of estrogens can provoke masculine erotization in some adult males, either normal or castrated. The relationship between the chemical constitution and the effect of the hormone is so close that a small modification of its mole- cule can profoundly change its actions. Thus, ethyl testosterone pre- pared from the male hormone is very active upon the endometrium. The recent book of Selye shows much of the multiplicity and complexity of the actions of the steroid hormones. Synergies and Antagonisms It would be impossible in a short essay to enumerate all the cases in which two hormones either strengthen (synergy) or oppose (antagonism) one another's actions. Estrogens in a certain adequate dose prepare the uterus and mammary glands, and sensitize them to progesterone; nevertheless, in other doses these substances can nullify each other's actions. Estrogenic and androgenic hoimones have, in certain cases, antagonistic actions; in others, their actions are inde- pendent and do not interfere with each other; and, finally, some- times they are mutually strengthened. Regulation of Secretion of Hormones Even though, for each function, several organs play a part, their participation is regulated so as to maintain a steady balance, as is demonstrated by the constancy of the blood sugar level, of the oxygen consumption, of the blood calcium, etc. These regulations are, there- fore, factors in the functional unity of the organism and in the equilib- rium of their functions. B. A. HOUSSAY REGULATION OF EACH GLAND Each endocrine gland has its own regulating mechanisms for the secretion of its hormones. As judged by experiments on extirpation and restitution, there must be a basal secretion, generally uninterrupted. In certain instances this has been well demonstrated (adrenalin, insulin). This secretion is submitted to regulating factors, the principal ones being humoral, and in some cases also nervous. Thus the basal amount of insulin secreted by the islets of Langerhans depends on the blood sugar level; and, reciprocally, the blood sugar level depends on the amount of insulin secreted. The basal secretion of insulin increases when glycemia increases; conversely, it decreases when the blood sugar level is lowered. The parathyroid secretion increases when calcemia de- creases. In both cases the secretion of either insulin or parathormone tends to restore the altered equilibrium of the internal medium. A similar regulation of the secretion of the thyroid hormone is probable, udging by the constancy of basal metabolism. The regulation of mechanisms of hormone secretion are very pre- cise and well designed to attain their objective. Thus, one-seventh of the pancreas is adequate to maintain the basal glycemia at the normal level; also, the basal glycemia continues at a normal level in an animal even if four pancreases are grafted by vascular anastomosis. These facts show that a normal secretion of insulin is maintained either with one-seventh of the pancreas or with five pancreases — this because gly- cemia governs insulin secretion. Only in abnormal cases (diabetes or hyperinsulinism) is regulation of insulin secretion deficient or excessive so that the gland works at a new level. In some cases of hyperplasia, adenomata, or cancers, the endocrine glands have been known to pro- duce hormonal hypersecretion. Although the reduction in mass of an endocrine organ may not alter its ability to function under basal conditions, it may lead to its insufficiency in cases of emergency. The pancreas of the dog reduced to one-fifth of its mass is enough to maintain normal glycemia, but if grafted to a diabetic dog it does not replace a normal pancreas in cor- recting the existing hyperglycemia. The resistance of the surgically reduced pancreas is diminished against the action of injurious agents such as extracts of the anterohypophysis and thyroid and, in conse- quence, diabetes develops readily. When the pancreas is normal, these 198 HORMONES injurious agents either do not, as in the case of sugar or thyroid extracts, produce diabetes or else do so, as in the case of anteropituitary extracts, only if much higher doses are used. In most instances the secretion of the hormones is governed by humoral factors, while the nervous factors play only an accessory, dispensable role. Thus, denervation of pancreas, thyroid, adrenal cortex, or gonads does not produce any insufficiency of their endocrine functions. Sometimes it is found that the nervous action makes the secretory regulation somewhat quicker and more precise, as in the case of insulin secretion. There are cases, however, in which the nervous regulation is important. Thus the splanchnic nerves exert a tonic and emergency action upon adrenalin secretion and their section reduces it to traces. The supraoptic nucleus governs secretion of the antidiuretic hormone of the neurohypophysis and section of the supraoptic neurohypophysial fibers is followed by insipid polyuria. Stimulation by light or desicca- tion inhibits the melanic-expanding secretion of the pars intermedia of the hypophysis through the medium of the nervous system. After cutting the pituitary stalk, the anteropituitary secretions are produced in sufhcient amount so that, generally, the thyroid, adrenal, and gonads are not modified. But there is no increase in the secretion in cases of emergency; and in animals with the pituitary stalk cut ovulation is no longer produced by mating {e. g., doe), embracing (e. g.,-toad), or vision (e. g., dove) as in normal animals. Intense light can induce the secre- tion of enough hypophysary gonadotropins to produce hypertrophy of the gonads when they are in the atrophic condition of winter rest. Cold no longer exerts its action on ovarian cycles or upon thyroid or adrenal cortex when the pituitary stalk has been cut in rats. Certain hypothalmic injuries may decrease the secretion of either all or some of the pituitary gonadotropins. REGUL.\TION OF COMPLEX EQUILIBRIA The individual regulation of each gland is at the same time submitted to more ample regulations which are often reciprocal. Thus, the blood sugar level is maintained constant in spite of the coexisting secretions which tend to produce either hypoglycemia, such as insulin, or hyperglycemia, such as those of the hypophysis and adrenal glands. There exists, therefore, a functional equilibrium of the secretion of each endocrine gland and, at the same time, a functional equilibrium of all B. A. HOUSSAY the secretions of a similar functional constellation (for example, the endo- crine secretions which regulate sex or those that govern carbohydrate metabolism, etc.)- These regulations are mainly humoral, depending to a great extent on endocrine factors, although the nervous system often has an important share. The tendency of the endocrine glands to reach a functional equilibrium and then to maintain it without overshooting seems to con- form to a sort of general law. Thus, when an ovary is extirpated, the remaining ovary produces the same number of ovules and cycles ("law of follicular constancy" of Lipschlitz). A testicular fragment either assures the total development of the sexual characters of a cock, or else it atrophies with no intermediate stable equilibrium being set up (the "all or none" law of Pezard and Gley). The tendency to all or none activity of the gland in situ, maintaining its secretion at a con- stant level, independent of its mass, is not inconsistent with the fact that the pharmacological effects of the hormones vary with the doses, the relationship following the typical S-shaped curve. The close associations existing among the glands of internal secretion explain why the disturbances affecting one of them are generally reflected in the others. It is rarely, if ever, that experimental or pathological disturbances of an endocrine organ are observed with- out modifications of the other glands. Thus, the extirpation of the hypophysis leads to atrophy and hypofunction of the adrenal cortex, thyroid, and gonads. The actions of each hormone can be direct or indirect. The testicle and androgenic hormones, for example, produce hypertrophy of the seminal vesicles and prostate directly. The pituitary gonadotropin (LH or ICSH) produces the same action, but since it induces the se- cretion of testicular hormones is inactive in the absence of the testicle. In certain cases the functional interactions of endocrine glands are simultaneous; in other cases they follow each other in sequence, as the proliferative (estrogenic) and secreting (progesteronic) phases of the endometrium during the menstrual cycle. HYPERSECRETION OF HORMONES That the secretion of the hormones is not normally at a maxi- mum is borne out by several facts: {a) castration very much increases the secretion of anterohypophysary gonadotropins: {b) adrenal 200 HORMONES ablation very much increases the secretion of pituitary corticotropin; (c) in clinical cases of endocrine hyperfunction supernormal effects are observed, similar in many cases to those produced by excessive administration of hormones. Hyperfunction of the endocrine glands is observed under several circumstances: (a) pathologic hyperplasias and tumors (be- nign, or adenomata, and malign) in which are produced exaggerated secretion of either normal or abnormal hormones; (b) injection of either glandular extracts or hormones; (c) excitation by a supernumary gland (for example, parabiosis of a castrated animal with a normal one); (d) rupture of the endocrine equilibrium. In some cases either hypo- or hyperfunction takes a certain time to develop. Thus if 95% of the pancreatic tissue of a rat is removed, the residual tissue is still able to keep a normal blood sugar level for two or three months, but later on a progressive diabetes de- velops. Following certain operations on the ovary (grafting, ligation, partial fragmentation), a hypersecretion of pituitary gonadotropin gradually develops, which, by promoting excessive secretion of ovaric estrogens, produces a marked hypertrophy of the uterus, hyperplasia of the endometrium, etc. In this case a new hyjDophyso-ovaric equi- librium is established at an abnormal level. Types of Endocrine Functional Associations The functional associations belong to various types. (/) A gland, such as the anterohypophysis, can develop and maintain the structure and function of one or several other glands, as the thyroid, adrenal cortex, ovary, or testicle. (2) One gland can moderate the function of another, e. g., the sexual hormones moderate the gonado- tropic pituitary function.* (3) Actions, sometimes antagonistic and sometimes synergetic, can be observed between two glands (or their hormones), as in the case of the ovary (or estrogens) and the corpus luteum (or progesterone). (4) Certain hormones increase the sensi- tivity to others, e. g., estrogens to the effect of progesterone upon the endometrium or upon the mammary gland, and thyroxine to the effect * In fact, the position is more complex, because estrogens, if given in in^li doses, may induce an increased secretion of luteinizing, adrenotropliic, and lacto- genic hormones of the anterior pituitary, but, in larger doses, suppress them all. 201 B. A. HOUSSAY of adrenalin. (5) Certain hormones can produce an insufficiency by damaging an organ; thus anteropituitary extracts and, in certain cases, those of the thyroid as well, by a repeated action bring about the disappearance of the j8-cells of the islets of Langerhans and produce a pancreatic diabetes, the cause of which could not be deduced by any one not familiar with the previous history. METHODS OF STUDY Many methods are required to study the endocrine glands and their hormones. Morphological. Important data can be obtained by the study of the weight and macro- and microscopic structure of the endocrine organs of different ages and under different conditions, created by extirpation of other organs and injections of hormones. It is also necessary to watch for the initiation and localization of the changes. Chemical. These include: {a) a search for and isolation and purification of the hormones by means of a combination of biological assays and chemical methods; {b) a study of the chemical changes and metabolic modifications produced in an animal by the suppression or administration of hormones; {c) action of the hormones on tissue slices or on chemical systems in vitro; and {d) a study of the origin, metabolism, and excretory products of the hormones. Physiological, These can be subdivided into experimental and clinical methods. The experimental method includes the study of: {a) glandular insufficiency and restitution through grafting, implants, or administration of either extracts or hormones; {b) hyperfunction induced by the methods already described; {c) measurement of the hormones in the organ, in the blood that comes away from it, in that of the general circulation, or even in the urine. It can also be indirectly measured by finding the amount necessary to substitute for the removed organ (substitution method). It is much safer to measure the hormone secreted by an organ than the amount the organ contains, because the amount secreted does not always vary parallel with the amount present in the organ. The clinical study is very valuable: (a) because it furnishes human data and {b) because the disease is a spontaneous experiment, the conditions being sometimes more delicate and varied than the experimental methods can secure. Many endocrine functions have 202 HORMONES been studied first clinically and afterward experimentally, e. g., Addi- son's disease, acromegaly, etc. Quantitative Determination of Hormones. The hormone or its derivatives can be determined in the organ, blood, or urine. The determination can be made by chemical, physical, or biological methods. The low concentration of the hormone rarely allows direct gravimetric estimations; in general, these must be based on colori- metric, spectrophotometric, or chromatographic methods. Biological assays are performed on entire animals, isolated organs, or tissue slices. Animals with insufficiency may be used. In some cases the organ is forced to function overloaded, e. g., a pancreas grafted in diabetic dogs by vascular grafting is tested for the time necessary to correct hyperglycemia. Hormones are assayed by comparing their effects with those produced by international standards, thus avoiding differences of sensitivity encountered in different races of animals. In order to confirm that a given function depends on an endocrine organ, the following consequences must be etablished: (7) the ablation or injury of the gland must produce an insufficiency of that function; (2) this deficiency must be compensated for by the graft or implantation of the gland, or by the injection of its extract or its hormone; (3) an excess of these substances must induce hyperfunction symptoms opposed to those of hypofunction; (4) the anatomoclinical facts must agree with the experimental findings; (5) both spontaneous and induced hyper- function must be improved by treatments which either eliminate the gland or decrease its action. In order to admit that a given action due to an endocrine gland is produced through the mediation of another gland, it is necessary to show: (7) that the removal of the second gland is followed by a condi- tion of hypofunction, just as, or even more, pronounced than that caused by the ablation of the first; (2) that the disturbances cannot be corrected either by implantation or injection of extracts or hormones of the first gland when the second one has been removed. For example, (a) ovariectomy leads to atrophy of the uterine and vaginal epithelium, just as much or more so than hypophysectomy, and (b) the pituitary gonadotropins produce hyperplasia of the uterine and vaginal epi- thelium only when the ovary is present. These facts lead to the conclusion that the effect of the hypophysis on the uterine and vaginal 203 B. A. HOUSSAY epithelium is due to the induecd hypersecretion of the ovarial hormones. To be active, certain hormones require a previous sensitization of the receptor organ by other hormones. Thus, estradiol does not produce hyperplasia of the adrenals in hypophysectomized rats, but this eflfect is produced if the atrophy of the adrenals is prevented by administration of two daily doses of anterohypophysis (Pinto). Chorionic gonadotropin does not induce ovulation in hypophysec- tomized rats, but does so if either serum gonadotropin or stilbestrol is previously injected. Metabolism of the Hormones At present we live in a period in which the metabolism of the hormones is being energetically studied. The studies include the investigations of origin, transformation, and elimination of the hor- mones. The study of origin includes that of their precursors within the organism or in the diet and that of the place and mechanism of elabora- tion within the endocrine gland. It is also interesting to know its absorption, its chemical trans- formations, and the site of these transformations. Thus it is known that the estrogens are destroyed principally in the liver and that thyrotropin is transformed by the thyroid into an inactive compound which can be reactivated at a certain temperature. In some cases the disappearance of the hormone can be quantitatively followed. The disappearance of the hormone from the blood has been followed in various cases. The unchanged hormone may be elimi- nated in the urine or, secondarily, in the milk or the bile, but in other cases only the transformation products of the hormone are eliminated through these routes. The urine has the advantage, for purposes of extraction, of being a concentrated ultrafiltrate of the plasma virtually free of protein. Hormones of protein nature may or may not filter through the kidney according to their molecular size. Thus chorionic gonadotropin and thyrotropin are found in the urine, but not so the gonadotropin of the pregnant mare serum. The form in which various steroids with estrogenic, androgenic, or corticoid action are eliminated, as well as the total elimination of the 17-ketosteroids, are being in- tensively studied. In some cases the metabolism of a particular hormone can be traced by the assay of some transformation product, 204 HORMONES e. g., that of progesterone through the urinary ehmination of preg- nanediol. Applications The study of hormones is of interest lo medicine and animal husbandry. For medicine it is important to investigate: {a) the physiological and nutritive conditions which will assure hormonal equilibrium and which will secure better physical and mental health during rest, exercise, or work; (b) the prevention and treatment of the endocrine disturbances and esjiecially the more common disturb- ances such as diabetes, endemic goiter, endocrine disturbances in women, etc., by genetic, dietetic, and pharmacological methods; (c) the prevention and treatment of disturbances due to abnormal internal secretions, e. g., nephrogenic hypertension. The distinction between specialists in diseases of metabolism and specialists in diseases of the endocrine organs is quite artificial. Diabetes, for example, is the most typical endocrine disease. The glands of internal secretion are the regulators of metabolism and it is impossible to study cither endo- crinology independently of metabolism or metabolism independently of endocrinology. The study of the hormones is germane to the problem of animal production because it gives a clearer understanding, and thereby wider possibilities, of controlling phenomena such as heat, ovulation, fer- tilization, pregnancy, number of offspring, breeding without de- pendence on factors such as lactation, castration, time of the year, etc. The study and production of hormones has been converted into a problem of national importance. There are enormous commercial interests involved and a great number of technicians devoted to the search, production, and commerce of hormones. This raises the danger of both excessive and inadequate use of the hormones and of exaggerated propaganda. At the present moment, the main problems of experimental endocrinology may be classified in the following groups: (7) isolating pure hormones and studying their actions, either separated or asso- ciated, simultaneous or successive; (2) establishing the mechanism of action of each hormone to determine whether they are direct or mediated by other organs; (J) studying each organ's secretion from the standpoint of the regulating factors and of its relations witli the 205 B. A. HOUSSAY regulatory mechanisms of other glands and with the wider systemic functions; (4) the metabolism of the hormones; (5) clinical applica- tions to prevent and cure endocrine diseases; (6) applications in animal industry. This exposition of the hormones for the sake of brevity has been necessarily general in character and, therefore, incomplete. The author had to avoid the double risk of being too elementary or too tedious. Briefness courts the danger of being dogmatic or of not pro- viding the factual basis for each statement. This essay has been written with a view to presenting to scientific laymen, but not to specialists in endocrinology, the current position of the problem of the hormones. Selected Bibliography Barker, L. F., Hoskins, R. G., and Mosenthal, H. O., Endocrinology and Metabo- lism. 5 vol., Appleton, New York, 1922. Bayliss, W. M., and Starling, E. H., "The chemical correlation of the secre- tory process," Proc. Roy. Soc. London, 67, 310-322 (1940). "Die che- mische Koordination der Funktionen des Korpers," Ergeb. Physiol., 5, 664 (1906). Berthold, "GeschlechtseigentiimHchkeiten," Wagners Handwort. Physiol., 1, 507 (1842). "Transplantation der Hoden," Arch. Physiol., 1849, 42. ^\cd\. A., Innere Sekretion. 2nd ed., 2 vol., 1913; 4th ed., 3 vol., 1919-1922. Urban & Schwarzenberg, Vienna. Del Castillo, E. B., Reforzo Membrives J., De La Baize, F., and Galli Mainini C., Endocrinologia Clinica. El Ateneo, Buenos Aires, 1944. Glandular Physiology and Therapy. Am. Med. Assoc., Chicago, 1935. 2nd ed., 1942. Gley, E., Les skrUions internes. Bailliere, Paris, 1914, 94 pp., 3rd ed., 1925. Hirsch, M., Handbuch der inneren Sekretion. 3 vol., Kabitzsch, Leipzig, 1928- 1933. Lucien, M., Parisot, J., and Richard, G., Traite d'endocrinologie. Doin, Paris, 1925 and 1934. Pende, N., Endocrinologia. 4th ed., 2 vol., Vallardi, Milan, 1934. Schaeflfer, E. A., "On internal secretions," Lancet, 2, 321, 324 (1895). The Endocrine Glands, Longmans, Green, London, 1916. Starling, E. A., "The chemical correlation of the functions of the body," Lancet, June (1905). Trendelenburg, P., Die Hormone. 2 vol., Springer, Berlin, 1929, 1934. Vincent, S., "Innere Sekretion and Driisen ohne Ausfiihrungsgang," Ergeb. Physiol., 11,218 (1911). 2o6 14 FUNDAMENTALS OF OXIDATION AND REDUCTION LEONOR MICHAELIS, member emeritus, the rockefeller INSTITUTE FOR MEDICAL RESEARCH / P"N ATTEMPTING to define the essential concepts involved in a problem, generally it is found that, because of the flexibility and, often, ambiguity of language, a definition cannot be formulated with perfect clarity. Nature is not so constructed that one can classify all its subject matter within a finite number of distinctly circumscribed terms. A great deal of confusion has arisen, and will henceforth arise again, from the fact that an author may use a given term according to one definition, and then during the dis- cussion consciously or unconsciously forget that definition. Further- more, it is a frequent fate of a definition that it be based on an assump- tion which later appears to be either erroneous or, at least, unsuitable. If the term, as commonly happens, is redefined according to the change in the underlying fundamental concepts, the new definition is likely to be in conflict with the older one. A typical instance of the flexibility of a concept is the term oxidation, and its reverse, reduction. Originally, oxidation meant combination with oxygen; the combination of hydrogen with oxygen to form water is the prototype of oxidation in this sense. According to this definition, the combination of hemoglobin with oxygen to form oxyhemoglobin should be the simplest, purest, and most unambiguous 207 LEONOR MICHAELIS case of oxidation in organic chcinisliy. However, according to our present usage of the term, this reaction is so dissimilar from what is now considered to be a typical oxidation that it is no longer classified among the oxidations — it has, in fact, been termed "oxygenation" by Conant. The real oxidation of hemoglobin, according to modern definition, is its conversion to methemoglobin, because, according to the present stage of knowledge and on the basis of our present model of atomic and molecular structures, the oxygen of oxyhemoglobin is attached to hemoglobin without aflfecting the electronic structure of the iron atom,* whereas in methemoglobin no oxygen is attached to hemoglobin at all — rather, the iron atom of hemoglobin contains one electron more than the iron atom of methemoglobin. The removal of that electron from hemoglobin is now considered as typical for its oxidation. When ferrous chloride reacts with chlorine, ferric chloride is formed. Since ferrous compounds can be converted to ferric com- pounds by oxygen also, ferric iron has always been considered an oxi- dation product of ferrous iron. Yet, if this conversion is brovight about by chlorine, oxygen plays no part in the "oxidation." What is common in most processes formerly designated as oxi- dation, disregarding such an exceptional case as the formation of oxy- hemoglobin, can only be stated in terms that would have been quite incomprehensible to the originators of the concept of oxidation. This common property can be defined, in terms of the present state of the atom model, by saying that, after its oxidation, a molecule has been deprived of one electron, or of two electrons; these two cases of oxida- tion are distinguished by the terms "univalent" and "bivalent." The oxidation of ferrous ion to ferric ion, or of ferrocyanide ion, [Fe(CN)6]^~ to ferricyanide ion, [Fe(CN)6]^~, are examples of univalent oxidation. The oxidation of stannous ion, Sn"'""'', to stannic ion, Sn^+, is a bivalent oxidation. In those cases in which the oxidation, or, in other words, the withdrawal of the electron, is brought about by oxygen, the oxygen is the acceptor of the electron. The fate of oxygen after acceptance of the electron will be discussed on page 220. Other reagents, such as * This statement involves another difficulty based on facts unknown until recently. Since hemoglobin is paramagnetic, but oxyhemoglobin diamagnetic, as Pauling and Coryell have shown, some change in electronic structure due to the attachment of oxygen must be postulated here too. But it is not of the type one would now call oxidation. 208 OXIDATION AND REDUCTION chlorine, can also accept :\n electron and thus "oxidize'' other sub- stances. Oxygen, therefore, is only one among many molecular species which can withdraw an electron from other molecules and so "oxidize" them. One must remember that the transition of the original definition of oxidation to the modern one has been gradual. Historically, there has been a transition stage inaugurated by Wieland, who developed his concept mainly with respect to oxidation of organic molecules. /H /H When alcohol, CH3C— H, is oxidized to aldehyde, CH3C , the over- \OH ^O all eflfect is the loss of two hydrogen atoms. According to the original definition of oxidation, one may say that the primary process is the addition of one oxygen atom to alcohol in order to form CH3C — OH \0H which, by splitting off one molecule of water, becomes CH3C ^O Wieland's suggestion is that the process does not pass through the stage of an addition of oxygen, but that what had been designated as oxida- tion is in fact a withdrawal of two hydrogen atoms. The oxidizing agent is the acceptor for the hydrogen atoms; for such cases, he re- places the term "oxidation" by "dehydrogenation." The reconciliation of this idea with the more modern one can be based on the fact that a hydrogen atom consists of a positively charged proton and a negatively charged electron. Then, one may say that the withdrawal of two electrons is essential for oxidation of alcohol. Since the two protons, which should remain, are no longer held by any noticeable force, they are detached also, and become bound to some proton acceptor, such as water, with which they form the hydrogen ion, OH3+, as it exists in the presence of water (called also the oxonium ion); or they may be bound to some anion that may be present, such as the acetate ion, GH3GOO~, with which they form an "acid," CH3COOH. This hypothesis should not involve the idea that the expulsion of the electron occurs first and the expulsion of the proton thereafter. The sequence is undecided. The principle is, rather, that the withdrawal of each hydrogen atom is the same as the 209 LEONOR MICHAELIS withdrawal of an electron together with a proton; this process is dehydrogenation. The reconciliation with the modern definition is the postulate that only the withdrawal of the electron is characteristic for oxidation, and that the simultaneous detachment of a proton does not belong to the process of oxidation proper. The attachment of a proton is termed a change in the level of "acidic ionization." This will become clear if one first considers an example showing the essential features in a very obvious way. Let us start with quinone (I), the reduction of which, according to modern concepts, consists essentially in the attachment of two electrons. The result is formula II, a doubly negatively charged ion of hydroquinone. O II /\ Y o ~ o- ~ OH 1 JA /N A V o- Y OI I V OH li III IV quino ne ion Hydroq uino ne anion Hyd ^oquinone Quinone If the reaction occurs in a strongly alkaline solution, the result of oxidation is molecule II. When the reaction occurs in a less alkaline solution, oipH about 9 to 10, one proton only is attached to II, and the univalent anion of hydroquinone is formed (III). When the reaction occurs in an acid solution, two protons are attached to II, and (union- ized) hydroquinone (IV) is formed. Depending upon the pH, either an electron or a full hydrogen atom may be attached to each oxygen. We now agree on the definition that II, III, and IV are on the same level of oxidation-reduction as, but on lower levels of oxidation than, quinone (I). The diff"erences between II, III, and IV are not in their level of oxidation, but in their level of acidic ionization. It is easy to transfer these ideas to the case of the oxidation of alcohol. On detaching two electrons only, one obtains structure V, where two protons exist in the molecule. The two protons, however, are held still less firmly than in hydroquinone ion IV, and under all CH3C^H + \OH + V CH3C' VI < 210 OXIDATION AND REDUCTION conditions possible, even in extremely acid solution, V changes its acidic ionization level by expelling the two protons and forming acet- aldehyde (VI). • Here again we do not claim that V is formed first and VI subsequently. It is, however, emphasized that the process of oxidation proper is the detachment of electrons, and that the simul- taneous loss of the two protons, although it may be a necessary con- sequence, does not belong to the process of oxidation itself. Dehy- drogenation is thus a special case of oxidation. The term dehydrogenation need not be discarded but may be used advantageously for those cases in which the release of an electron involves the simultaneous release of a proton. Whether or not this simultaneous release of the proton occurs depends on the pH of the solution. The term dehydrogenation can be reserved for the cases in which even in extremely alkaline solution the release of an electron is accompanied by a release of a proton. The conversion of succinic acid (VII) to fumaric acid (VIII) may be called a dehydrogenation because the intermediate stage (IX) is not capable of permanent existence. Molecular species VIII may be considered as an infinitely strong acid, which even in an acid solution is not capable of holding on to protons. Therefore the enzyme which catalyzes the reaction VII -^ VIII may COOH COOH COOH I I I CH2 CH CH-H + I II I CH2 CH CHH+ I I I COOH COOH COOH VII VIII IX be justly termed a dehydrogenase. When quinone (I) is reduced in an acid solution, we may speak also of dehydrogenation, whereas in the reduction of quinone in a more alkaline solution (IX and VIII) the synonymous use of the terms oxidation and dehydrogenation would involve a more generalized definition of dehydrogenation. Analogously, if we define oxidation as the loss of an electron, we are compelled to generalize the definition by the amendment that this definition holds whether or not a proton is also released; so whether one uses "oxidation" or "dehydrogenation" is a matter of nomenclature only. In this essay, we shall use "oxidation" and "reduction," and not "dehydrogenation" and "hydrogenation." For 21 I LEONOR MICHAELIS such reactions as the conversion of benzene to cyclohexane, however, it is recognized that one may prefer to use the term hydrogenation instead of reduction. As regards the chain of oxidation reactions as they occur in respiration, it is often said that hydrogen atoms are transferred from one molecular species to another. However, the oxidation of reduced cytochrome to cytochrome involves not the transfer of a whole hydrogen atom, but of one electron only. It is therefore not entirely true that the chain of respiratory processes con- sists in transferring hydrogen atoms from one molecular species to another, and finally to oxygen, but it is true that this chain consists in transferring electrons from one molecular species to another. Some- times the transfer of the electron is accompanied by a transfer of a proton and sometimes it is not. It may be added that the loss of a hydrogen atom in dehydro- genation is equivalent to the addition of a hydroxyl group, as far as the level of oxidation-reduction is concerned. When alcohol is oxi- dized to acetaldehyde, the process may be alternately described as follows. Alcohol detaches one hydrogen atom and accepts a hydroxyl group: /H /OH /O CHsC^OH > CH3C— OH > CHsC^OH > CHsC^^ + H2O \H \H \H \H Here, the first step is the detachment of the hydrogen atom, and the second, attachment of the hydroxyl group, while the process formerly was described in terms of loss of two hydrogen atoms. So, a'^hydro- genation is equivalent to "hydroxylation," and both terms can be avoided by describing the level of oxidation proper only in terms of electrons ejected or added, whether or not a proton or a hydroxyl ion is also involved in the process. Stepwise Oxidation and Reduction of Organic Compounds If one wants to formulate a simple example of a possible step- wise oxidation of an organic compound, one might propose the series: CH4 > CH3OH > CH2O > HCOOH > CO2 each step representing a bivalent oxidation. A univalent oxidation is not imaginable unless one abandons the a.ssumption that carbon is 212 OXIDATION AND REDUCTION tetravalent, oxygen bivalent, and iiydrogcn univalent: the uni\alenl oxidation of CH4 would give, as a step intermediary on the way to CH3OH, the free radical CH3 with "tervalent" carbon. To be sure, such free radicals have been recognized for a long time, but those known until recently always have one of two properties: either they are very unstable and have an extremely short lifetime (e. g., CH3); or they can be produced only from a very restricted number of com- pounds and only in a solution of a perfectly water-free organic solvent. The famous discovery of triphenylmethyl by Gomberg in 1890 pro- vided a prototype of such free radicals. It will now be shown that all oxidations of organic molecules, although they are bivalent, proceed in two successive univalent steps, the intermediate state being a free radical, and furthermore, that according to structural conditions these intermediate free radicals may be either just as unstable as CH3, somewhat more stable, or even perfectly stable compounds. First, however, the concept of stability must be discussed. The criterion of stability may be obtained in two ways. A substance may be said to be unstable if it rapidly undergoes a chemical change when exposed to such ubiquitous reagents as air; pyrogallol, for instance, is unstable in an alkaline solution because it is readily oxidized by oxygen, but is stable in the absence of oxygen. Or, a substance may be said to be unstable if it undergoes a change even in the absence of any foreign substance with which it might react; acetaldehyde, for instance, undergoes the Cannizzaro reaction in an alkaline solution, one molecule of aldehyde being oxidized to form acetic acid, while another is reduced to ethyl alcohol. It will now be shown that the alleged instability of free organic radicals is, in very many cases, essentially due to the latter, bimolecular interaction. It will be shown that the bivalent oxidation in fad occurs in two successive univalent steps as in the example of duroquinone (X). The methyl groups in the ring of X prevent the secondary, irreversible reactions which are of no interest in this discussion, and which would occur when working with the unmethylated, simple benzoquinone. When duroquinone in an alkaline solution is reduced, as by hydrogen plus palladium, or by any other suitable reducing agent, the faintly yellow solution turns brown first, then colorless. The colorless compound is the corresponding durohydroquinone, which 213 LEONOR MICHAELIS in extremely alkaline solution can be written as formula XI and which O -O OH i k CH3— /N— CHs CH3— / \ — CH3 CH3— Y X — CH3 CH3 \ -CHs — CH3 — CH3 -o XI in less alkaline solution would add first one, then another, proton to form XII. This is the customary bivalent reduction. The inter- mediate, brown substance is the result of a univalent reduction; and it can be shown by various methods, to be described later on (see pages 215 and 217), that it has the same molecular size as XI and differs from it only in so far as one oxygen atom is negatively charged and not the other. This brown substance is a free radical. One might say that one oxygen atom is bivalent and the other uni- valent, as in XIII. The same molecular species, in a more acid solu- tion, would have formula XIV. O O GHj CH, ^. -CH3 -CH3 CH3 CH3 ,-V^-CH3 -GHs o- XIII OH XIV Let us consider first the reaction in a strongly alkaline solution. When the solution of the quinone is mixed with increasing amounts of a reducing agent, there will always be, except for the very begin- ning and the end of the oxidation, a mixture of the quinone, the hydro- quinone, and the intermediate free radical which we shall call semi- quinone; and an equilibrium is established between these forms. It is important to emphasize that the equilibrium is established instan- taneously, and that there is no sluggishness in its formation, as is usually the case in the formation of equilibria with organic compounds (except acidic ionizations). The activation energy of this reaction leading to equilibrium is extremely small. So, in the solution, the intermediate free radical is never present without being in mixture with the quinone 214 OXIDATION AND REDUCTION and the hydroquinone. If we designate the hydroquinone (the re- duced form of this system) as R, the semioxidized form (the free radical) as S, and the totally oxidized form (the quinone) by T, the equilibrium is established according to the reversible reaction 2S , ' R + T (1) and the constant of equilibrium, which may be called the "semi- quinone formation constant," is: _[S?_ [R][T] Its reciprocal may be called the "dismutation" constant, because re- action (1) is a "dismutation" (or "disproportionation") of the free radical. If k is very small, very little of the radical exists in the state of equilibrium; if k is large, much of the radical can exist, and it can be distinguished by its particular color and by its paramagnetism, which will be discussed presently. Experience has shown for the case of duroquinone that k is very large in an alkaline solution, but very small in an acid solution. The transition of the behavior from alkaline to acid solution is continuous; and there can be no doubt that in acid solution a small amount of the radical is formed too, even if its concentration is too small to be notice- able by ordinary methods. Thus we are obliged to state that the ionized form of the radical (XIII) is a rather stable compound, and the unionized (XIV) rather unstable, "stability" being judged accord- ing to its capability of existing in equilibrium with its parent substances, the quinone and the hydroquinone. Why is XIII more stable than XIV? In formula XIII, the negative charge has been arbitrarily attributed to the upper oxygen atom, but it can just as well be attributed to the lower one. In fact, the location of the charge is undecided, for it oscillates between the two extreme positions through the chain of the atoms in the ring. Such a condition has been termed resonance. The two limiting struc- tures, one with the charge on top, the other with the charge at the bottom, are indistinguishable molecular species. One speaks about "equivalent" or "symmetrical" resonance, a condition which, accord- ing to quantum mechanics, contributes largely to the stability of a molecule. In form XIV, no such equivalent resonance prevails. The 215 LEONOR MICHAELIS proton attached holds the negative charge rather tightly in place. This molecule does exist but to a much smaller extent than XIII be- cause it is less stable. Free radical formation is a general reaction whether it takes place extensively, as in alkaline solution, or in traces, as in acid solution. It can be shown by numerous examples that the same behavior is true for organic compounds which form reversible oxidation-reduc- tion systems, such as many dyestuffs, e. g., methylene blue, or phenol- indophenol. No observable amount of any intermediate free radical can be demonstrated for irreversible oxidations, e. g., when alcohol is oxidized to aldehyde. So we are permitted to conclude that the formation of a free radical, in sufficient concentration, in the equi- librium involved, is a prerequisite for the reversibility of an oxidation- reduction system. Evidence for Free Radical Nature of the Intermediate Substance It is appropriate to discuss at this point the evidence for the assertion that the intermediate steps of oxidation are really free radicals of the same molecular size as the parent substance and not dimeric valence-saturated compounds made up in such a way that two radicals are combined to form a bond which abolishes the state of "unsatura- tion." Such a reaction may be imagined to occur as follows: 2 S > D (2) That is, two molecules of the semiquinone radical, S, combine with each other to form a dimeric compound, D, which no longer has the characteristics of a free radical, just as two "radicals" H (hydrogen atoms) combine as follows: 2.H > H2 (3) In fact, such reactions do occur, and an equilibrium is established, so that, instead of reaction (2) we should write: 2 S , D (4) Depending upon the conditions, which will not be discussed here, this equilibrium is sometimes greatly in favor of the free radical, sometimes 216 OXIDATION AND REDUCTION greatly in its disfavor. The main point is that reaction (2) does not, as a rule, proceed to completion, but that the free radical really does exist, and often even to such an extent that the dimerization is negli- gibly small. There are two powerful tools to decide whether the intermediate product is S, D, or an equilibrium mixture of the two, viz-, potentiometric titration and magnetic measurement. When a substance such as duroquinone, the example discussed above, is titrated with a reducing agent and the oxidation-reduction potential is plotted against per cent reduction, a curve is obtained the shape of which will depend markedly on whether the intermediate substance is R or D. Straightforward application of the law of mass action yields a complete theory as to the shape of the titration curve. The calculations involved, although not absolutely simple, are of such a nature that only a high-school student might consider them in the realm of "higher" mathematics. Application of the law of mass action has shown in many cases that the intermediate substance is a free radical, almost exclusively under certain conditions and, under other conditions, in equilibrium with its dimer. Furthermore, according to a theory first developed by G. N. Lewis and later amply confirmed by the quantum theory, a free radical, because it always contains an odd number of electrons, must always be paramagnetic, in contrast to the ordinary, valence-saturated organic compounds, which are diamagnetic (provided they do not contain metal atoms such as iron or cobalt). When a solution of duroquinone is slowly reduced in alkaline solution, by glucose, say, evidence can be produced for the gradual appearance of a paramagnetic molecular species and, on further reduction, for its disappearance. The paramagnetism is due to the spin of the odd electron, while in ordinary molecules the elec- trons always occur in pairs with opposite spin, whereby their para- magnetic effect is quenched. An elegant method of detecting such free radicals, even of low stability, has been established by G. N. Lewis. In the kind of experi- ments mentioned above, ejection of an electron is brought about by an oxidizing chemical, which serves as an acceptor of electrons. Lewis eflfects ejection of the electron by ultraviolet light, the substance being dissolved at the temperature of liquid air, in an organic solvent which has, at this temperature, a rigid, glasslike consistency without crystal- lizing. In such a rigid medium, no molecular collisions can occur 217 LEONOR MICHAELIS and free radicals, once they have been established, have no chance to undergo bimolecular reactions by which they may be rapidly elimi- nated. Radicals can be preserved, in this manner, to an extent w^hich by far exceeds their equilibrium concentration postulated by thermo- dynamics. Lewis showed that many organic compounds exposed in this way to ultraviolet radiation become colored; in suitable cases he identified, by spectrophotometrical comparison, the colored substance with the free radical obtained previously by chemical oxidation. The peculiar merit of this method is the fact that free radicals can be de- tected in the cases in which they would be unnoticeable in the state of equilibrium because of their low concentration. On the other hand, since the appearance of color is not necessarily evidence for a free radical, a scrutiny of the phenomenon is necessary for each individual case. In order to arrive, from these considerations, at the discussion of the final problem we have in mind, the varying degree of inclination of an organic substance toward oxidation (or reduction) must be con- sidered. Here we distinguish two essentially different properties. To state that a substance is easily, or difficultly, oxidized by an oxi- dizing agent may mean one of two things: that the speed of such an oxidation is great or small (a topic belonging to a discussion of chemical kinetics); or that the final state attainable by the interaction of the oxidizing agent and the oxidizable substrate is a complete, 100% oxidation, an incomplete oxidation, or almost no oxidation at all (a problem of thermodynamics). We shall start with the problem in thermodynamics, in par- ticular, the case in which both the oxidizing agent and the substance to be oxidized form reversible oxidation-reduction systems. If leuco- methylene blue is the substance to be oxidized and potassium ferri- cyanide the oxidizing agent, when the two are mixed in equivalent proportions practically all of the leuco dye is oxidized to the dye and all of the ferricyanide is reduced to ferrocyanide. But when leuco- methylene blue is mixed with an indophenol dye, both the oxidation of leucomethylene blue and the reduction of indophenol will be in- complete, and the final state is a mixture of four substances, the two leuco dyes and the two dyes. Furthermore, if leucomethylene blue is mixed with safranine, no change will occur, at least within practical measurable limits. Safranine does not oxidize leucomethylene blue 2l8 OXIDATION AND REDUCTION to any appreciable extent. On the other hand, leucosafranine readily reduces methylene blue. With this in mind, one can set up a sequence of all reversible oxidation-reduction systems according to their oxida- tive power. For the four systems mentioned, the sequence is: ferri- cyanide, indophenol, methylene blue, safranine. A quantitative expression of the oxidizing power is the "oxida- tion-reduction potential" of a dye in mixture with its leuco dye. Vari- ous substances capable of reversible oxidation and reduction can be arranged in a sequence according to their oxidation-reduction, or "redox," potentials. Each membei', when present in the reduced state, can be oxidized by any following member present in its oxidized state; and each member, when present in the oxidized state, can be reduced by any preceding member present in its reduced state. Let us now discuss the kinetic aspect of the problem. If we are dealing only with reversible redox systems, as above, establishment of the equilibrium is always so rapid that the reaction may be con- sidered almost instantaneous. But this is not the case, in general, if an irreversible reaction occurs. In the oxidation of alcohol to acetal- dehyde, for example, to oxidize alcohol, a powerful oxidizing agent such as chromic acid must be used; and to reduce acetaldehyde, a powerful reducing agent such as sodium amalgam must be used. Although an excess of oxidative (or reductive) power must be applied in order to make the reaction proceed, the reaction is sluggish. Addi- tional energy is required far beyond the quantity expected on a purely thermodynamic basis because, obviously, an obstacle has to be over- come. Despite the fact that the reaction between alcohol and chromic acid releases energy, energy must first be spent, which is of course eventually released again; and this extra energy is called the activa- tion energy. Although the path of energy is, as a whole, downward, it must first pass over a hill. Generally an activation energy is required for all bimolecular chemical reactions. In reversible reactions the energy of activation is very small, interaction occurring only when the two molecules "collide." The collision is impeded by the fact that molecules of any kind, on approaching each other in the course of thermal motion, will exhibit a mutual repulsion, and only those mole- cules which happen to have enough kinetic energy to overcome the repulsion will really collide and react with others. In irreversible o.xidations another, more serious, impediment 219 LEONOR MICHAELIS occurs. We have presented the postulate that all oxidations proceed in a sequence of univalent steps. The first step of oxidation of alcohol would lead to the free radical, in this case, an utterly unstable molecule, in the thermodynamic sense. To generate these radicals, an oxidizing agent of very high potential is required ; and even then their concen- tration remains small, so small that no direct evidence for their exist- ence is available. The free radicals, once generated, will then react by a dismutation: 2 radicals ^ ^ 1 alcohol -}- 1 aldehyde The velocity of the latter reaction depends, among other things, on the concentration of the molecules which are interacting with each other, and therefore on the square of the concentration of the free radicals. If this concentration is very small, it may be the limiting factor for the over-all process of oxidation of alcohol. We may say that the energy of activation for the oxidation of alcohol is essentially the energy necessary for the formation of the free radical. Unless the radical is relatively stable, as in reversible processes, the activation energy is very great. This high energy of activation is the reason why so many organic compounds are "stable." If all thermodynamically possible reactions could proceed unhampered there would be no such thing as organic chemistry. Inhibition due to high activation energy occurs only when the attainable concentration of the free radical is the limiting factor for the rate of the over-all reaction. It does not matter whether the attainable concentration of the free radical is 1 Af or, say, 10~* M. Factors other than the concentration of the free radical, such as the specific constants of reaction velocities, will determine the rate of the reaction. However, if the concentration of the free radical is so small as to be the limiting factor of the over-all process, sluggishness and irreversibility will arise. This consideration fulfills an important re- quirement for the understanding of reaction rates — it reduces a problem of kinetics to one of thermodynamics. Oxygen as an Oxidizing Agent These considerations also explain why oxygen is such a sluggish oxidizing agent despite its very large oxidative power in a thermo- 220 OXIDATION AND REDUCTION dynamic sense. If it can be reduced only in successive univalent steps, these steps must be: O2 > O2- »02-- >Oj" ^O*" Two of these steps are chemically identifiable. O2 is, after accept- ing two protons, O2H2, hydrogen peroxide. ©2" is, after accepting four protons, two molecules of H2O. However ©2" (or O2H), and 02~ (which may be written as O2H3, or OH + H2O) are intermediate, utterly unstable steps. Since the reaction must pass through these unstable steps, the activation energy involved in the reduction of O2 is very high. How is this activation energy overcome when oxygen does oxidize a substance? Overcoming the activation energy by working at high temperatures is a usual procedure in the laboratory but is not feasible under physiological conditions. Here the answer is that oxygen reacts with an oxidizable substance very often not only by means of a collision of the molecules but also by the establishment, after collision, of a relatively stable addition compound which can then undergo intramolecular redistribution of electrons. Certainly no claim is made that this mechanism is always the one involved in activation of oxygen. However, it is one of the possible mechanisms and very likely is correct for the particular case to be de- scribed in detail. In all probability it is the inechanism by which oxygen is activated in all those cases in which a heavy metal com- pound, especially of iron or copper, acts as activating catalyst. It has been shown that, at least in an acid solution, the oxida- tion of a leuco dye, or of cysteine, and many other substances, by means of free oxygen is accomplished, at least with any appreciable speed, only in the presence of a trace of an iron or copper salt. These metal atoms have two essential properties which render them useful for their catalytic action. First, they readily change their valence; iron may be bivalent or tervalent and copper, univalent or bivalent. Second, these metals are highly inclined to form complex compounds of the Werner type. The nature of such metal complex compounds may be demonstrated as follows. The doubly positively charged ferrous ion, Fe+"^, can combine, first of all, with two negatively charged univalent ions to form a saltlike compound, for instance, with two cyanide ions: 221 LEONOR MICHAELIS Fe++ + 2 CN- > Fe(CN)2 (a) However, the combining power of the ferrous ion is not exhausted by this reaction based on opposite charges. In fact, the molecular species Fe(CN)2 has never been shown to be capable of existence. More than two cyanide ions are attached to the iron due to the fact that each CN~ ion has one pair of electrons not used for chemical bonding. Each of such electron pairs can be shared with the iron to fill up its outermost incomplete electron shell to a complete shell, as in a noble gas. In addition to the two CN~ ions of equation (a), four more can be attached, which contribute four negative charges to the complex molecule, which is a ferrocyanide ion: Fe(CN)2 + 4 CN- > Fe(CN)J~ (b) The six CN groups are arranged around the central iron atom as the corners of an octahedron. Fe is said to possess "six coordination places" which may be occupied by atoms or atom groups. In analogy, when ferrous ion combines with cysteine, which we may write, briefly, as RSH, * we may imagine that, primarily, a saltlike compound between iron and two molecules of cysteine is formed in such a way that the two hydrogen atoms of the sulfhydryl groups are replaced by an iron atom: Fe++ + 2 RSH » Fe(RS)2 + 2 H++ (c) This scheme accounts, so far, for the saturation of two (of the six pos- sible) coordination places. Now atom group R contains another atom with an unused electron pair, viz-, the nitrogen atom of the amino group. Thus, the two molecules of RSH will occupy four coordination places. Probably because of the large size of the RSH molecule and steric hindrance involved in it, the two remaining coordination places cannot be occupied by a third molecule of RSH. In fact, no ferrous complex of cysteine can be prepared with more * SH is a sulfhydryl group, and R represents the rest of the molecule, which is, altogether: SH NHj 1 I H2C— C— COOH H 222 OXIDATION AND REDUCTION than two molecules of cysteine for one iron atom.* The two re- maining coordination places may be filled in by other atoms or atom groups of small size having an unused electron pair. One example of such an atom group is carbon monoxide, CO. In fact, the coordi- nation compound, Fe^^(RS)2(CO)2, can be readily prepared by the interaction of Fe++, cysteine, and carbon monoxide in the form of its well crystallizable alkali salt.f Just like CO, O2 also has (at least) one unused electron pair, and may combine instead of CO, as in the case of hemoglobin which combines either with O2 or with CO. In hemoglobin, four coordination places of iron are occupied by the four nitrogen atoms of the porphyrin ring, a fifth by protein, and the sixth can combine with O2 or with CO. So we arrive at hypothetical iron-cysteine-oxygen complexes, analogous to the well-known carbon monoxide complex, Fe"(RS)2(CO)2. One cannot tell whether one oxygen molecule is attached at one or at two coordination places, or whether even two oxygen molecules can be attached. Suffice it to imagine the complex Fe"(RS)202. This complex is said to be hypo- thetical because it cannot be prepared, and undergoes a redistribution of electrons, or, in other words, intramolecular oxidation-reduction, which may be symbolized in this way: The original, oxygen-con- taining complex may be imagined to consist of the following con- stituents: Fe++ + 2 RS~ + O2. The electron redistribution will occur thus: Fe3+ + 2 RS- + O2" (O2 withdraws one electron from Fe + +) (d) Fe3+ + RS- + RS + 02~ (O2- reduced to 02~, /. e., H2O2, by withdrawing one electron also from one RS ") (e) Fe++ + RS + RS + O2 — (Fe3+ withdraws one electron from RS") (f) An alternative scheme, probably of equal probability, is: * It is interesting to compare cysteine complexes of cobalt with those of iron. For the cobaltous state, no complex with more than two cysteine molecules can be obtained. For the cobaltic state, both a complex with two and another with three molecules of cysteine can be prepared. The cobaltous complex, then, is quite analogous to the ferrous complex. The cobaltic complex stands no com- parison, because the ferric-cysteine complex is too unstable, due to the rapid intramolecular rearrangements to be described presently. t The formula in the first footnote on page 222 shows that atom group R contains a carboxyl group. It is by means of the two carboxyl groups in the com- plex that alkali salts can be formed. 223 LEONOR MICHAELIS Fe3+ + 2RS- + O2- (g) Fe + + 4- RS- + RS + O2- (h) Fe++ + RS + RS + O2— (i) step (f) being identical in both cases. Now, the complex containing the constituents as in (f) is unstable and disintegrates to form three separate molecular species: Fe++; RSSR, cystine; and hydrogen peroxide. The latter may be used to oxidize more cysteine (stoichio- metrically, not catalytically), or to oxidize Fe++ to Fe^+j and this Fe3+ may oxidize (stoichiometrically) more cysteine. Finally, all the iron is again in the ferrous state and the whole cycle is repeated with iron thus acting as a catalyst. It is an essential prerequisite of this cycle that the change from the ferrous to the ferric state occurs readily and reversibly. In using cobalt instead of iron, the first stages are similar; but, once the cobaltic complex has been established, it shares with all cobalt complexes of the Werner type the property that the cobaltic state cannot be readily reduced to the cobaltous state, even by means of rather strong reducing agents. The final result is, therefore, the formation of the cobaltic complex, stoichiometrically, without starting a catalytic cycle. Cop- per, but not cobalt, can replace iron as a catalyst. What is furthermore essential in this process is the fact that each single step in this chain reaction consists of the transfer of a single electron. This assertion is more than a mere hypothesis. Since the change of ferrous to ferric state involves one electron only, the sub- division of the over-ail process into one-electron transfers is obvious. It is remarkable that, even for such a simple case of iron catalysis, the whole chain is of such an intricate nature, allowing for different path- ways leading to the same final result. Oxidation Catalysts and Enzymes , The physiologically occurring catalysts (or enzymes) for oxida- tion or reduction are characterized by their specificity. All oxidation enzymes have been recognized as compounds of reversible redox sys- tems and a specific protein. The same redox system, when attached to different specific proteins, may have a different specificity. The present state of our knowledge is on what may be called a descriptive 224 OXIDATION AND REDUCTION level: some of the enzymes can be prepared as pure crystalline entities, and their composition and activity can be examined. Since the first stages in their discovery by Warburg, a vast amount of knowledge has been accumulated which, on the one hand, demonstrates the highly complex nature of the problem and, on the other, shows that certain recognizable features are shared by these enzymes: they are proteins attached to prosthetic groups. The proteins are all specific and not identical with those otherwise occurring in the organism. But the chemical mechanism of the action of these enzymes is not yet worked out, at least to an extent comparable to that given in the simple example of iron catalysis in the oxidation of cysteine. It is quite natural that enzyme chemists have, thus far, been occupied with the discovery of many kinds of enzymes, the ingenious methods of preparing them, and the measurement of their activity. But at this point we must inquire into the chemical mechanism by which they work; and here only a few speculations can be brought up. One simple suggestion is this: if it is true that the sluggishness of an oxidation process is caused by the instability (in its thermo- dynamic sense) of the free radical through which the over-all oxidation has to pass, then the function of the enzyme may be that of increasing the stability of the radical, in other words, that of increasing the con- centration of the radical which can exist in equilibrium with the re- duced and the oxidized state of the substrate. The substrate com- bines, reversibly, with the enzyme, and the "semiquinone formation constant" of the enzyme-substrate compound may be greater than that of the uncombined compound. We may make another suggestion. Let us suppose that the enzyme can combine not only with the sub- strate to be oxidized but also with the oxidizing agent. For example, methylene blue can oxidize succinic acid to fumaric acid in the presence of the enzyme called succinodehydrogenase. Suppose this enzyme can combine with both succinic acid and methylene blue. The specific structure of the enzyme brings about a definite spatial orientation and juxtaposition of fumaric acid and methylene blue. When a mole- cule of one of these two substances collides with a molecule of the other in a solution, the chance of an electron transfer during the short time of collision is nil; but when these two molecules are held close together in appropriate juxtaposition and orientation with respect to each other, they remain in this spatial arrangement for a long time, during which 225 LEONOR MICHAELIS an electron transfer may occur once in a while. Now, the transfer of a single electron establishes the free radical, and from here on the second step of oxidation takes place readily and spontaneously. In order to account for the fact that the interaction of succinic acid and fumaric acid in the presence of the enzyme and methylene blue is reversible, one must postulate also that the enzyme can combine re- versibly not only with succinic acid, but also with fumaric acid; and not only with methylene blue, but also with leucomethylene blue. Such an assumption is not unreasonable and is supported by the ob- servation that, in very many cases, molecular species of a structure similar to that of the specific active substance, inhibit the function of that substance and so compete with it for the enzyme. Out of the numerous examples discovered in recent years, one may recall the antagonism of j&-aminobenzoic acid and sulfanilamide. At this stage of the argument, we have reached the realm of speculation. Any further advance depends on the clarification of the structure of the enzymes and especially of the steric structure of the spe- cific proteins. It will be the aim of future work to show that the specific structure of the enzyme forces the substrates attached to the en- zyme to stay in such a mutual orientation as to permit an electron transfer, which will not occur with a reasonable probability on a free collision. A similar principle may underlie all specific enzymic reactions, as well as those not concerned with oxidation-reduction. The astounding fact that all enzymes are or contain a protein of specific structure suggests that the attachment of the substrate to the enzyme with its specific protein structure increases the chance of a thermo- dynamically possible reaction by forcing a spatial orientation of the interacting molecules which has practically no opportunity to occur on spontaneous haphazard collision, and moreover, by holding the interacting molecules in this specific orientation for a length of time very much longer than on the occasion of a haphazard collision by thermal motion. This is the way in which we may imagine the activation energy is overcome, and in which out of a vast number of thermodynamically permissible reactions only a few reactions service- able for the metabolism are selected. The road for the exploration of the mechanisms of individual metabolic catalyses will be long. Although it is still far ahead, one is encouraged to believe that the correct road sign has been found. 226 OXIDATION AND REDUCTION Selected References Clark, W. M., et al., "Studies on oxidation-reduction," U. S. Pub. Health Service, Hyg. Lab. Bull. No. 151 (1928). Michaelis, L., "Occurrence and significance of semiquinone radicals," Ann. N. r. Acad. Sci., 40, 39 (1940). Michaelis, L., and Schubert, M. P., "The theory of two-step oxidations involv- ing free radicals," Chem. Revs., 22, 437 (1938). Pauling, L., and Coryell, D. C, Proc. Natl. Acad. Sci. U. S., 22, 210 (1936). Schubert, M. P., J. Am. Chem. Soc, 53, 3851 (1931); 54, 4077 (1932); 55,4563 (1933). Thunberg, T., "Zur Kenntniss des intermediaren StofTwechsels und der dabci wirksamen Enzyme," Skand. Arch. Physiol., 40, 1 (1920). Warburg, O., Uber die katalytischen Wirkungen der lebendigen Substanz- Springer, Berlin, 1928. Wieland, H., "Uber den Mechanismus der Oxydationsvorgange," Ergeb. Physiol., 20, All (1922). 227 15 MESOMERIG CONCEPTS IN THE BIOLOGICAL SCIENCES HERMAN M. KALCKAR, member of the research staff, division OF NUTRITION AND PHYSIOLOGY, THE PUBLIC HEALTH RESEARCH INSTITUTE OF THE CITY OF NEW YORK, INC. l\TO CHEMIST would question that the concept of meso- -^ » merism (or resonance), a concept which is actually based on quantum mechanics, has played an overwhelmingly important role in the development of modern physical and organic chemistry. This concept has made it possible to understand and explain the properties of numerous inorganic and organic compounds, and in a number of cases to predict chemical events. There are many indications that this concept will play an equally significant role in the biological sciences, and for that reason it merits the consideration of biologists. Mesomerism The terms resonance and mesomerism are synonomous. The former is based on quantum mechanical concepts, while the latter, more neutral term merely indicates that a substance exists as a hybrid of at least two more or less symmetric electronic states. The term "mesomerism" is more appropriate in a biological essay in which the symbols and formulas are not those of quantum mechanics, and will therefore be used here. 229 H. M. KALCKAR A group is said to display mesomerism if it can be described by two or more symmetric (or very nearly symmetric) electronic formulas, representing approximately the same potential energy. In that case, the chances that the electrons will occupy one or the other position are equally great. The electrons are therefore moving forth and back between these equivalent positions or, to use a term of physics, oscillat- ing between the positions. The mesomerisms of the carboxylate anion and the amidine cation represent some of the simplest examples of sym- metric mesomerism in a molecular group. O: • • • • R:C or :0 R:C • • :0:- • • O («) I (i) Carboxylate anion H:N:H • • :N: • • R:C or • • R:C H:N:H H:N:H + («) (b) II Amidine cation In these two sets of alternative structures, one of the oxygen or nitrogen atoms is surrounded by a complete set of eight electrons, the so-called octet, and the other by only six electrons, the last pair of elec- trons participating in the double bond. If this pair of electrons were moved up to complete the octet, the opposite oxygen or nitrogen would have to donate one pair of electrons from their octet in order to restore the double bond. Structures a and b are completely equivalent and indistinguishable . The well-known benzene mesomerism is usually illustrated by the two structures shown in formula III, /\ or f\ \/ V (a) (b) III Electronic mesomerism can also exist between two molecules. This so-called intermolecular mesomerism can be illustated by the two symmetric structures of formula IV. In structure a, the left A 230 MESOMERIC CONCEPTS IN BIOLOGY R:A- :A:R or R:A: -ArR IV possesses an unpaired electron (odd electron) and the right A, an elec- tron pair; in structure b, the situation is reversed. The shift can be represented as an oscillation of one electron between the two symmetrical molecules: R : A. : A : R. The existence of an unpaired electron gives rise to paramagnetism because the neutralized magnetic moments of paired electrons is abolished when the electron lacks its partner. The type of intermolecular mesomerism illustrated in formula IV will be discussed later in this essay. The frequency of the oscillations which the electrons in meso- meric structures undergo is very high; and it is therefore impossible to consider a mesomeric group as possessing for an appreciable interval of time either structure a or b. There is, however, a type of meso- merism in which it is possible to distinguish between the two structures. Tautomerism is a classical illustration of this type of mesomerism. In tautomerism, both an electron and a hydrogen atom parti- cipate in the oscillation. Since the hydrogen possesses a significant mass, the oscillation is considerably slower than in the purely electronic mesomerism; and it is therefore possible to distinguish between the two symmetrical states. Two examples of tautomerism are: (1) the enol-keto tautomerism of carbonyl compounds (V); and (2) the lac- tam-lactim shift of the hydroxy purines (VI). -H H3— c— c— N-=C— OH H— N— C— (0) II 1 1 1 1 Keto form HO- -C C— N. II II >c- N— C— N/ 0— C C— N. 1 II >c- H— N— C— N/ or -H (*) H2— C=C— 1 1 Enol form | H H OH («) (b) V VI The so-called hydrogen bond results from the attraction of a hydrogen atom attached to one electronegative atom (e. g., fluorine, oxygen, nitrogen) for an unshared electron pair of another electronega- tive atom (see formula VII): R— O— H . . . O— R VII 231 H. M. KALCKAR Biological Significance of Mesomerism What is the significance and what are the implications of meso- meric phenomena in biological reactions? The great significance of mesomerisin lies in the fact that it invariably endows the molecular group with a considerable amount of additional stability. The word stability in this connection is used in its broadest sense. It has been a custom to distinguish between thermodynamic stability and the kind of stability implied in terms such as "willingness" of a group to react Scheme i Sequence of Reaction "ACTIVATED" MOLECULE I ACTIVATION ENERGY OF REACTION: A:=:±:(B):;^C Scheme ii Sequence of Oxidation ■OF THE REACTION: A=;=i:C A^ r- Ia^aI a MOLECULAR ACTIVATION ENERGY 1) OF — NONCATALYTIC OXIDATION 2) OF CATALYTIC OXIDATION spontaneously. The distinction must be considered artificial. The work of Polanyi, Eyring and Stearn — cf. Eyring (3) — indicates strongly that the activation of the substrate, i. e., the problem of how reactive complexes are formed from stable molecular groups, is essentially a thermodynamic one. Scheme I illustrates some of these relationships. "A" signifies the starting product and "C," the end product of the reaction A — > C. AFis the change in free energy of the reaction, I. e., the amount of potential energy which is lost when A is converted to C. The free energy or potential energy of the system drops when A is converted to C. However, in order to start the reaction, A must be activated in some way, i. e., the potential barrier which represents the so-called activation energy must be overcome. Both the activation energy and the free energy change (AF) are influenced by mesomerism. Since the significance of mesomerism for our understanding of energy coupling in biological systems has been discussed elsewhere (6), this 232 MESOMERIC CONCEPTS IN BIOLOGY aspect will be treated rather briefly in the present essay. The influence of mesomerism on AFcan be summarized as follows: If C has an addi- tional amount of mesomeric stability which A does not possess, the drop in AF of the reaction A — > C is greater than it would have been if C did not possess that extra stability. We have already mentioned the carboxylate ion as a typical representative of a molecular group with extra mesomeric stability. If another molecule is introduced into such a mesomeric group, as by esterification, the symmetry of the group is disturbed and the mesomerism decreases or vanishes, which again implies that the potential energy of the complex is raised. Thus, acetic acid anhydride, CH3 — CO— O— CO — CH3, in which both carboxyl groups have lost their state of mesomerism, possesses a much higher potential energy than that of the two acetic acids formed by hydrolysis. In terms of scheme I, acetic acid anhydride would correspond to A and the two acetic acid molecules to C. The living cell contains at least three types of substances in which the mesomerism of two groups is mutually blocked. The first type includes the carboxyl phosphates (acyl phosphates), the second group, the amidine phosphates (phosphocreatine, phosphoarginine), and the third group, the pyrophosphates. The carboxyl phosphates are the primary oxidation products of a reaction in which a carbonyl phosphate complex undergoes enzymic oxidation. It is important to point out that the oxidation is catalyzed by an enzyme specific only for the phosphate complex. Thermodynamically speaking, the oxidation of a carbonyl-water complex to free carboxylate would be greatly favored, and the chances of forming a carboxyl phosphate would be vanishingly small, if the latter reaction were not specifically catalyzed by an enzyme. This brings up the question of the nature of enzyme catalysis. An unusually promising approach toward an understanding of oxidation-reduction catalysis on the basis of mesomeric concepts has been made by Michaelis and his group. Since this topic is discussed in Chapter 14, only certain aspects of the problem will be treated here (8). It is now generally recognized that oxidation of organic com- pounds, involving the removal of a pair of electrons, takes place step- wise. The removal of one electron prior to the other gives rise to the formation of a free radical displaying paramagnetism because of the H. M. KALCKAR presence of an unpaired electron, possessing unneutralized magnetic moment. This free radical has, generally, a very brief existence, since it either accepts an electron again or expels the remaining odd electron. Since the free radical has so little chance of existence, the removal of the first electron is barred, so to speak, by a high potential barrier. In scheme I, the group to be oxidized would be represented by A, the free radical by B, and the final oxidation products by C. The height of the potential barrier would represent the activation energy. The activation energy is the factor which particularly interests us in con- nection with the concept of catalysis. If the potential barrier is too high, the chance of forming the free radical is practically nil and the rate of the reaction is zero. When the temperature is raised sufficiently, the thermal movements of the molecule become so vigorous that a certain percentage of molecules will slip over the potential barrier. Although thermal movements, of course, are of importance for events in the living cell, physiological temperatures are usually too low to allow most reactions to proceed at measurable rates. In order to bring about a reasonable rate, the living cell has succeeded in lowering the potential barrier by a special device which we call catalysis. It is in this connection that mesomerism may turn out to be of paramount importance, as the model experiments of Michaelis and his associates have so strikingly demonstrated. The reduction of /)-benzoquinone to the corresponding hydro- quinone goes through a radical (semiquinone). The free radical has very little chance of existence because of its asymmetry (formula VIII). In basic solution, however, the semiquinone will exist as the symmetrical ion oscillating between two symmetrical electronic states. The two equivalent structures interchange the odd electron, similar to what Pauling (9) calls a three-electron bond, either through the benzene ring or by intermolecular bonds (formula IX). .6:H :6: < — ^ .6: I I I A fA A Y V Y :0: :0. < — > :0: VIII IX MESOMERIC CONCEPTS IN BIOLOGY The condition for stabilization of the semiquinone by meso- merism is satisfied only when the two structures are equivalent. This requirement is satisfied when the molecule is dissociated as an anion. The undissociated semiquinone in which the presence of the hydrogen atom eliminates the symmetry of the two structures does not fulfill the required condition. Correspondingly, Michaelis and his group showed that the semiquinone of phenanthrene-3-sulfonate is relatively very stable in alkaline solution in which it exists as the symmetrical anion. The semiquinone of /^-phenylenediamine, on the other hand, has a fair chance of existence in strongly acid solutions because only then does the symmetrical phenylenediaminium cation exist. In complete accordance with the ideas just developed is the observation that the free radical of paraquinones accumulates only in the alkaline pH range, whereas those of the paradiamine compounds accumulate in measur- able amounts only at strongly acid reactions. The "catalyzing" effect of hydroxyl ions or hydrogen ions on these two types of oxidation- reductions has actually been explained in terms of mesomerisms. The intriguing question is whether it is possible to explain en- zymic catalysis in terms of the same principles. There are observa- tions which may provide confirmation for such an explanation. Some years ago, Haas (4) found that riboflavin phosphate, when linked to a specific protein, forms a semiquinone when undergoing reduction. This semiquinone is not observed during the reduction of free riboflavin phosphate in neutral solution, but accumulates when the reaction is acid enough to insure complete ionization. In other words, the enzyme is able to stabilize a product at neutral reaction which other- wise would exist only at strongly acid reaction. This observation may be taken as a clear indication that oxidation-reduction enzymes in some way or other are concerned with the formation of mesomeric free radicals. Before going deeper into the discussion of the nature of enzyme catalysis, it is worth while to introduce a very ingenious theory proposed by Delbriick, dealing with the nature of reproduction phenomena. Delbriick advanced the idea that, in processes like gene reproduction, mes