Copyright 0 1991 by the Genetics Society of America Perspectives Anecdotal, Historical and Critical Commentaries on Genetics Edited by James F. Crow and William F. Dove The Regulationof Arginine Biosynthesis: Its Contributionto Understanding the Controlof Gene Expression Werner K. Maas Department of Microbiology, New York University Medical Center, 550 First Avenue, New York, New York 10016 N the 1950s there occurred a fundamental change in the outlook of biologists on the regulation of biochemical processes. Certain studies on the regulation of arginine synthesis in Escherichia coli played a role in bringing about this change and the present article is an account of these studies. Because I started to work on the regulation of arginine biosynthesis in the mid-I950s, the story told here will be of a somewhat personal nature. What was the nature of the fundamental change that occurred in the 195Os?To answer this question I shall first describe the general picture of the regulation of biochemical pathways and of gene expression that was in vogue before the change. Regulation of biochemical pathways belonged at that time in the domain of biochemists, whereas regulation of gene expression belonged in the domain of geneticists, and there was little interchange of ideas between the two. Inregardto biochemical pathways, regulation was thought to be somehow built into the chemical structures of enzymes and their substrates and products. Rates of reactions were thought to be controlled by affinities of substrates for their enzymes and by inhibition by products formed during a reaction. These conceptsseemed to be implicit in the thinking of biochemists and very little was said explicitly about regulation in textbooks published at that time, such as in the excellent book by FRUTONand SIMMONDS (1 953). In regard to the control of gene expression, the thinking of geneticists was very much influenced by the one gene, oneenzyme hypothesis of BEADLE and TATUM, which in a general form stated that a gene determinesa specific protein molecule. Little was known about how a gene determines a protein molecule and even less about how this gene activity was I This article is dedicated to the memory of LEOSZILARD. Genetics 128: 489-494 (July, 1991) controlled. Again, this lackof concrete ideas was reflected in then current textbooks of genetics. An exception to the lack of specific information was the work on adaptive enzymes, which at that time was carried out mainly by JACQUES MONOD and his coworkers at the Pasteur Institute in Paris. It was shown that for these enzymes, such as &galactosidase, the substrate or an analog of the substrate was required to bring about synthesis of the enzyme and that this “induction” was controlled by a specific gene. Constitutive mutants were isolated in which the “inducer” was no longer required for enzyme synthesis. MONOD’sinterpretation of the findings was that the inducer somehow molded the informationintoan enzymeforming system to produce a specific enzyme. In a constitutive mutant, a specific gene gave rise to an internal inducer, whereas in the wild type, this gene was inactive and the inducer had to be supplied externally. This picture of gene control, like that of the control of enzymatic reactions, involves chemical information contained in the substrate, which brings about regulation of, in one case, enzymatic reactions and, in the other case, the synthesis of enzymes. The thinking behind these concepts was based largely on then current concepts of interactions between proteins and low molecular substrates and also on template mechanisms, such as that invoked for the synthesis of antibodies. The latter was developed largely by HAUROWITZ (1 952) and has been referred toas the or “directtemplate”theory HAUROWITZ-PAULING (BURNET 1959). holds It that antibody molecules have their specificity determined by being synthesized against atemplate of theantigen molecules themselves. The big change in thinking about regulation was the realization that there is no direct chemical relationshipbetween the regulating substance and the 490 W. K. Maas protein whose production is regulated by this substance. Instead, the response to the regulating substance is mediated by the products of genes whose only function is the regulation of gene expression. This view was expressed by LEDERBERG (1 956) during the discussion of a lecture by MONOD(1 956)in which the latter had stated that“the ‘S’ substance (a common precursor) might be capable of giving rise to several different enzymes, depending on the nature of the inducer with which it reacted.” LEDERBERG’S response phosphate was, “Dr. MONOD asked whether the inducer carries the information needed for the specification of the enzyme. One permissible view holds that the enzyme, or its critical surface, is directly molded on the inducing substrate. The alternative, which I prefer, is that all the specifications are already inherent in the genetic constitution of the cell. The inducer signals a regulatory system to accelerate the synthesis of the corresponding enzyme protein.” This notion of specific regulatory gene products (presumably proteins) was novel to biochemists and had a profound effect on subsequent research in biochemistry. It brought biochemists together with geneticists and it did much to createthe field of molecular biology, inwhich guidance was provided mainly bythe results of genetic experiments. I have tried to formulate this change in an article written for a Lipmann Symposium (MAAS 1974) by saying, “It has become clearthat during evolution, mechanisms previously considered unlikely, can become established in living cells, as long as they give the organism a selective advantage over its competitors. The cell is no longer considered as a chemical machine, but as a cybernetic chemical machine.” The source of these novel ideas was the discovery of end-product repression in biosynthetic pathways, including that of arginine. These end productscould control thesynthesis of enzymes whose substrates bore little structural similarity to the end product. Discovery of the repression of arginine biosynthesis: T o make it easier for the reader to follow the text, the pathway of arginine biosynthesis in E . coli is shown in Figure 1. T h e first studies were carried out by HENRY VOGEL(1953). He noticed that in an arginine-requiring mutant blocked before N-acetylornithine (Figure l), the level of acetylornithinase (Figure 1, step 5) was higher in cells grown with acetylornithinethan incells grown with arginine. He subsequently extended these studies by growing the cells on mixed supplements of arginineand acetylornithine, in order to distinguish between induction by acetylornithine or inhibition by arginine as an explanation for his observations. In 1957, in a paper given at the McCollum Pratt Symposium (VOGEL1957), he presented data showing that his observed effects were due to inhibition by arginine. He coined theterm 1 Glutamate -N-acetylglutamate 2 4N-acetyl lutamyl phoshate \“I N-acetylglutamlc semialdehyde 1. Arginine 7 Arglnlno +Citrulline succinate 0 A A + Ornithine + Carbamyl t9 ATP + Glutamine + CO2 FIGURE 1.-The biosynthesis of arginine. T h e reactionsteps shown arecontrolled by theargininerepressor. T h e enzymes catalyzing these reactions are,in order of the steps: 1, N-acetylglutamate synthase; 2, N-acetylglutamate kinase; 3, N-acetylglutamylphosphate reductase; 4,N-acetylornithine transaminase; 5 , N-acetyl ornithinase; 6, ornithine carbamoyl transferase; 7, argininosuccinate synthetase;8, argininosuccinase; 9, carbamoylphosphate synthetase. “enzymerepression” to describe this phenomenon. The substance inhibitingenzyme synthesis was named a “represser.”(The alternative spelling “repressor”has since been accepted, whereas “promotor” has not.) My own involvement with the regulation of arginine biosynthesis began in 1955 and was unrelated to VOGEL’S work. At that time I was interested in how genes control theproduction of enzymes. In 1952 I had shown that in a temperature-sensitive mutant of E . coli blocked in the synthesis of pantothenic acid, the enzyme catalyzing the conditionally blocked reaction was more heat-labile than the corresponding enzyme in the wild type (MAASand DAVIS1952). The results showed that an altered protein was produced as a result of mutation and that a gene could therefore control the structure of its protein product. After a stay in LIPMANN’S laboratory in 1953 and 1954, I returned to New York University to continue the study of gene-enzyme relationships. I looked for mutants of a known gene in which the rate of enzyme synthesis was altered. T o d othis, I screened revertants from completely blocked mutantsthat could grow without the required growth factor at 37” but could not at 25 (cold-sensitive mutants). One of the strains from which I obtained such mutants was an arginine auxotroph blocked between ornithine and citrulline (Figure 1,step 6) and unable to make the enzyme ornithine transcarbamoyltransferase (OTCase). From this strain the desiredrevertants were obtained. OTCase was produced when the cells were grown at 37” but not at 25”. At 25” the mutant had to be grown with arginine, but at 37” it grew in a minimal medium without arginine. The surprise came when the cells were grown at 37 ” with arginine: no OTCase was produced.Subsequently, inhibition of OTCase formation by arginine was found in the wild-type O Perspectives strain as well, at both 37” and 25”. Itwas shown that the inhibition was specific for arginine. Growth with citrulline did not inhibit OTCase synthesis, whereas growth with arginine did (MAAS 1956). Presumable citrulline is converted to arginine too slowly to establish thearginineconcentrationrequiredfor full repression. In this roundabout way I had serendipitously discovered thatend-productrepression of OTCase was a general phenomenon and had nothing to dowith the cold-sensitive mutations. By 1957 it was thus established that arginine could repress the formation of at least two enzymes in its biosynthetic pathway. Both enzymes catalyzed reactions before the last step in the pathway. It seemed most probablethat this repression was not due to arginine alone and that other components must be involved. The most obvious way to look for such components was to isolate mutants in which OTCase formation was not repressed by arginine. I shall describe such mutants after first considering theimpact of the argininestudies on the then current ideas about the regulation of enzyme formation. A unified hypothesis for the regulationof enzyme synthesis: In 1957, enzyme induction and enzyme repression were two different ways in which the formation of an enzyme could be controlled. At that time the notion was prevalent that theprimary mechanism for the two types of control should be thesame. This view was especially propounded by MONOD.It is curious today to see this insistence on a unitary mechanism. The reason may be that, at thetime, our thinking was muchinfluenced by thethen fashionable concept of the “Unity of Biochemistry.” I stated this view during the discussion of a paper I presented in 1957 (MAAS and GORINI1958): “I think that most people feel the distinction between adaptive (ie., inducible) and constitutive (it?.,biosynthetic) enzymes is an artificial one. In the case of constitutive enzymes the inducer is always present, whereas in the case of adaptive enzymes it is ordinarily not present. Alternatively, it is possible that adaptive enzyme formation is due to the inhibition of a feed-back inhibitor normally present, and thatinducers actby preventing the action of such feed-back inhibitors.” On the basis of his studies on the induction of pgalactosidase, MONOD at first believed strongly in induction as the primary mechanism (positive and “instructive” control). The story of how he changed his mind and how the work on arginine regulation contributed to his conversion to a hypothesis involving negative control has been told several times UIJDSON 1970; SCHAFFNER 1974; BROCK1990). Nevertheless, I shall give my personal account of this history and of the seminal role played byLEO SZILARDin the exchange of ideas. I had already demonstrated (MAAS 1956) that the 49 1 substrate ornithine was not required for the synthesis of OTCase and concluded, “here the substrate may not function as an inducer forenzyme formation.” At that time LUIGIGORINI,who was a visiting investigator in our pharmacology department at New York University, joined me in investigating the kinetics of OTCase formation. Following a discussion with NOVICK and SZILARD, we decided to study OTCase synthesis in a chemostat, a device for constant growth devised by NOVICKand SZILARD (1 950). our For studies we used a mutant blocked before ornithine andwe grew it in achemostat with arginine limiting the growthrate.Under theseconditionsOTCase was produced at a high “derepressed” rate, yet its substrate ornithine was not present in the bacteria (GORINIand MAAS 1957). In April, 1957 I attended the Federation Meetings in Chicago and stayed at the Quadrangle Club where SZILARD was also staying. I met him during breakfast and told him about the synthesis of OTCase in the absence of a substrate. Yet the kinetics of enzyme synthesis were like those observed during the induction of p-galactosidase. Onthe basisof aunitary hypothesis, one had to conclude that induction was a removal of repression andthattheinducer acted secondarily in counteracting a repressor. SZILARD was impressed with this idea and said that it could explain some findings on P-galactosidase formation that MILTON WEINERhad made in their laboratory. We had more discussions during breakfast on subsequent days and at the endSZILARD was ready to publish a paper. I felt reluctant because we had no experimental evidence in an inducible system but SZILARD did publish a theoretical paper (SZILARD 1960). Early in 1958, SZILARD presented the unitary hypothesis we had formulated in Chicago at a seminar at thePasteur Institutein Paris. At that time, ARTHUR PARDEE in MONOD’Slaboratory hadjust begun to carry out experiments tostudy the expression of the inducibility genefor @-galactosidaseformation following transfer of this gene during matings (later known as the PARDEE-JACOB-MONOD or “Pajamo”experiment or, in a lighter vein, the “Pajama” experiment). The thinking of MONOD at that time was still in favor of an “internal inducer” hypothesis but it appears that SZILARD’S seminar and the subsequent discussion had a strong influence in convincing him of the validity of the “removal of repressor” hypothesis to explain pgalactosidase induction. T o quote from MONOD’SNobel lecture, “Why not suppose, then, since the existence of repressible systems and their extreme generality were now proven, that induction could be effected by an anti-repressor, rather than repression by an anti-inducer? This is precisely the thesis that LEO SZILARD,while passing through Paris, happened to propose to us during a seminar. We had only recently 492 W. K. Maas obtained the first results of the injection (‘Pajama’) experiment, andwe were still not sure aboutits interpretation. I saw thatour preliminaryobservations confirmed SZILARD’S penetrating intuition, and when he had finished his presentation, my doubts about the theory of the ‘double bluff‘ (inductionequals removal of repression) had been removed. . .” (MONOD1966). T h e results of the completed experimentsdid indeed establish the validity of the “removal of repressor” model as originally postulated from results with the repression of OTCase. It is interesting that the belief in a unitary mechanism had such a strong influence on the willingness of MONODand others toaccept the interpretation of theinducer as an anti-repressor. Nowadays, and for goodreasons, our attitude has become muchmore flexible because bothprimary negative actions and primary positive actionshave been found in the control of enzyme synthesis. Isolation and properties of derepressed (argR-) mutants: In order to study components other than arginine required for repression,I started to look for derepressed mutants in 1957. Canavanine, an analog of arginine,was found toinhibit growth by competing for arginine in protein synthesis (SCHWARTZ and MAAS 1960). It seemed reasonable to suppose that mutants resistant to canavanine might be derepressedbecause they would produce more arginine than the repressible wild type and thus overcome inhibition by canavanine. At that time our wild-type strain was E. coli W. We isolated many canavanine-resistant mutants but they all turned out to be defective in the uptake of canavanine, arginine and other basic amino acids MAAS and SIMON 1959). These mutants (SCHWARTZ, started us on further investigations of permeases for basic amino acids. During the summer of 1959 1 worked in JACOB’S laboratory at the Pasteur Institute in Paris and again I looked for canavanine-resistant mutants, this time starting with E. coli K12. The first mutants I isolated were derepressed for OTCase.Later they were shown to be derepressed also for other enzymes of the arginine pathway. I named the gene that had mutated argR, I supposed that argR controlled the production of an “aporepressor,” arginine being the “corepressor.” After my return to New York University in the fall, 1 mapped the argR gene as well as several genes of the arginine biosynthetic pathway. The results of these studies were presented at the 1961Cold Spring Harbor Symposium (MAAS 1961). At the same meeting, GORINIpresented his work on argR mutants which he had isolated by another, more complicated method (GORINI, GUNDERSEN and BURGER196 1).The Operon Model was also described at this meeting by JACOB and MONOD(196 1). One of the unusual features of the arginine system was that, in contrast to the genes of other biosynthetic pathways (tryptophan, histidine), its genes were not next toeach other butwere scatteredoverthechromosome, yet were all controlled by the same regulator gene. Itwas shown that the tryptophan genes and the histidine genes were united in single operons, but the argininegenes were present in several operons. A. J. CLARK and I, in a paper demonstrating the dominance of repressibility over nonrepressibility in diploids (MAASand CLARK 1964), introduced the term “regulon” to describe the situation of the argininegenes. This paperestablished clearly that the control of the arginine pathway is negative via a repressor, the product of the geneargR. The nature of the arginine repressor and its targets: In 1961 the general outline of how the synthesis of enzymes is regulated had been established, thanks largely to genetic experiments, and now it became necessary to fill in details in different systems. For example, the chemical nature of repressors was not known, and it was also not known whether they worked at the level of transcription or translation or even post-translationally (MAAS and MCFALL1964). These questions were answeredduring the 1960s and 1970s for many systems, including the arginine system. It was shown that repressors are proteins and that they act at the level of transcription. Their targets, the operator sites, are short, specific DNA sequences located close to the promoters of the regulated genes. In the case of arginine, the question was raised whether arginine or a derivative of arginine is the corepressor.Afterconsiderableefforts it was shown thatarginine itself is the corepressor. The developments that occurred in the study of arginine regulation in prokaryotes up to 1985 have been reviewed by CUNINet al. (1986). Since then, the argR gene of E. coli K12 has been cloned and sequenced and the arginine repressor protein has been isolated and its interactions with its operator sites analyzed (LIMet al. 1987). Several unusual features distinguish the arginine repressor from other repressors. It appears tobe a hexamer,whereas other repressors are either dimers or tetramers. Its target site consists of two 18-bp palindromicsequences (ARG boxes) located near each other within the promoter region of each gene of the arginine regulon. Recently, co-crystals of the repressorprotein with synthetic ARG boxes have been obtained by JOACHIMIAK in SIGLER’S laboratory at Yale University and are being subjected to X-ray diffraction analysis. Studies on the interaction of the arginine repressor with its operator sites have thus reached thelevel ofatomic resolution. Another feature of the argininesystem is of general interest. It had already been notedin 1961 by GORINI that theregulation of arginine synthesis is different in E. coli B and E. coli K12 (ENNISand GORINI 1961). 493 Perspectives During growth in the absence of added arginine, endogenously produced arginine brings aboutthe same degree of repression of the arginine enzymes (about 90%) in both strains. With added arginine, the arginine enzymes of E. coli K 12 are totally repressed, whereas the arginine enzymes of E. coli B are only about 80% repressed. Later, this difference was shown by J A c o B Y and GORINI (1967) and by us (KADNERand MAAS1971) to be due to different argR alleles in the two strains. Recently, DONCBIN LIM inour laboratory has cloned and sequenced the argR gene of E. coli B and has isolatedthe arginine repressor from this strain (LIM et al. 1988). The B repressor differs by one amino acid from the K12 repressor but some of its properties, such as solubility,are quite different. The question is, how could the difference between the two repressors account for the observed difference in the regulation of enzyme synthesis by arginine? We have now found a partial answer to this question. It should be recalled that the active repressor consists of the ArgR protein (aporepressor) and arginine (corepressor). The degree of repression is therefore determined by the concentrations of repressor protein and arginine. Limiting either canlead to derepression. The formation of the K12 repressor is autoregulated (LIMet al. 1987) and there is presumptive evidencethat this is also true for the B repressor. In fact, it appears that autoregulation by the B repressor is so strong that the repressor concentration can become limiting for repression, so that derepressionof OTCase andother arginine enzymesmay ensue. When the operator and promoter sites of the B repressor are removed and replaced by the potent tac promoter, which does not have the argR operator sites, the addition of arginine brings about complete repression, as it does with the K12 argR gene. After induction with the inducer IPTG, presumably more repressor protein is produced with the tac promoter than with the normal argR promoter. These findings raise an interesting question. Why should a wild-type strain of E. coli increase the level of arginine-forming enzymes when it is supplied with external arginine? So far I have no clear answer for this paradoxical situation, but the persistence ofE. coli B in nature suggests that there must be a selective advantage for B-type regulation of arginine biosynthesis in the ecological niche occupied by E. coli B. If we do find an answer, we may uncover new principles that apply to the regulation of gene expression in general. Thus, even after 35 years of investigation, the regulation of arginine biosynthesis raises challenging questions for futureresearch. The author is grateful for continuous supported since 1958 by I’ublic Health Service Grant GM-06048 from the National Institute for General Medical Sciences for studies on the regulation of arginine biosynthesis. LITERATURE CITED BROCK,T. D., 1990 The Emergence of BacterialGenetics, Chapter 10. Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y. BURNET, M., 1959 The Clonal Selection Theory of Acquired Immunity, pp. 50-51. Vanderbilt University Press, Nashville, Tenn. CUNIN, R., N. GLANSDORFF, A. PIERARD and V. STALON, 1986 Biosynthesis and metabolism of arginine in bacteria. Microbiol. Rev. 50: 314-352. ENNIS,H. L., and L. GORINI,1961 Control of arginine biosynthesis in strains of Escherichia coli not repressible by arginine. J. Mol. Biol. 3: 439-446. FRUTON, J. S., and S. SIMMONDS, 1953 General Biochemistry. John Wiley It Sons, New York. GORINI,L., W. GUNDERSEN and M. BURGER,1961 Genetics of regulation of enzyme synthesisin thearginine biosynthetic pathway of Escherichia coli. Cold Spring Harbor Symp. Quant. Biol. 2 6 173-182. GORINI,L., and W. K. MAAS,1957 The potential for the formation of a biosynthetic enzyme in Escherichia coli. Biochim. Biophys. Acta 25: 208-209. HAUROWITZ, F., 1952 The mechanism of the immunological response. Biol. Rev. 27: 247-280. JACOB,F., and J. MONOD,1961 On the regulation of gene activity. Cold Spring Harbor Symp. Quant. Biol. 2 6 193-21 1. JACOBY, G. A., and L. GORINI,1967 Genetics of control of the arginine pathway in Escherichia coli B and K. J. Mol. Biol. 2 4 4 1-50. JUDSON, H. F., 1970 The Eighth Day of Creation, Chapter 7. Simon It Schuster, New York. KADNER, R. J., and W. K. MAAS,1971 Regulatory gene mutations affecting arginine biosynthesisin Escherichia coli. Mol. Gen. Genet. 111: 1-14. LEDERBERG,J.,1956 Comments on gene-enzyme relationships, pp. 161-162 in Enzymes: Units of Biologzcal Structure and FuncAcademic Press, New York. tion, edited by 0. H. GAEBLER. LIM, D., J. D. OPPENHEIM, T. ECKHARDT and W. K. MAAS, 1987 Nucleotide sequence of the argR gene of Escherichia coli K-12 and isolation of its product, the arginine repressor. Proc. Natl. Acad. Sci. USA 84: 6697-6701. LIM,D., J. OPPENHEIM, T. ECKHARDTand W. K. MAAS, 1988 The unitary hypothesis for the repression mechanism of arginine biosynthesis in E. coli B and E. coli K12 revisited after 18 years, pp. 55-63 in Gene Expression and Regulation: The Legacy of Luigi Gorini, edited by M. BISSELL,G. DEHO,G . SIRONIand A. TORRIANI. Excerpta Medica, New York. MAAS, W.K., 1956 Regulation of arginine biosynthesis in Escherichia coli. Biol. Bull. 111: 319. MAAS,W. K., 1961 Studies on repression of arginine biosynthesis in Escherichia cola. Cold Spring Harbor Symp. Quant. Biol. 26: 183-191. MAAS,W. K., 1974 Some thoughts on the regulation of arginine biosynthesis and its relation to biochemistry, pp 399-403 in Lippmann Symposium: Energy, Biosynthesis andRegulationin Molecular Biology, edited by D. RICHTER.Walter de Gruyter, New York. MAAS,W. K., and A. J. CLARK,1964 Studies on the mechanism of repression of arginine biosynthesisin Escherichia coli. 11. Dominance of repressibility in diploids. J. Mol. Biol. 8: 365370. MAAS,W. K., and B. D. DAVIS,1952 Production of analtered pantothenate-synthesizing enzyme by a temperature-sensitive mutant of Escherichia coli. Proc. Natl. Acad. Sci. USA 38: 785797. MAAS,W. K . , and L. GORINI,1958 Negative feed-back control of the formation of biosynthetic enzymes, pp. 15 1- 158 in Physiological Adaptation, edited by C. L. PROSSER.American Physiological Society, Washington, D.C. 494 W. K. Maas MAAS,W. K., and E. MCFALL,1964 Genetic aspects of metabolic control. Annu. Rev. Microbiol. 18: 95-110. MONOD, J., 1956 Remarks on the mechanismof enzyme induction, p. 26 in Enzymes: Units of Biological Structure and Function, edited by 0. H. GAEBLER. Academic Press, New York. MONOD, J., 1966 From enzymatic adaptation to allosteric transitions. Science 154: 475-483. A., and L. SZILARD, 1950 Description of the chemostat. NOVICK, Science 112: 715-716. K. F., 1974 Logic of discovery and justification in SCHAFFNER, regulatory genetics. Stud. Hist. Philos. Sci. 4: 349-385. J. H., andW. K. MAAS,1960 Analysisof the inhibition SCHWARTZ, of growth produced by canavanine in Escherichia coli. J. Bacterial. 7 9 794-799. SCHWARTZ, J. H., W. K. MAAS and E. J. SIMON,1959 An impaired concentrating mechanism for amino acids in mutants of Escherichia coli resistant to L-canavanine and Dserine. Biochim. Biophys. Acta 32: 582-583. SZILARD, L., 1960 The control of the formation of specific proteins in bacteria and in animal cells. Proc. Natl. Acad. Sci. USA 4 6 277-29 1. VOGEL,H. J., 1953 On growth-limiting utilization of a-N-acetylL-ornithine. Proc. 6th Int. Congr. Microbiol. (Rome) 1: 269271. VOGEL,H. J., 1957 Repression and induction as control mechanisms of enzyme biogenesis: the “adaptive” formation of acetylornithinase, pp 276-289 in The Chemical Basis of Heredity, edited by W. D. MCELROY and B. GLASS.The Johns Hopkins Press, Baltimore.
© Copyright 2026 Paperzz