THE NITROGEN

METABOLISM OF

MICRO-ORGANISMS

B. A. FRY

! METHUEN'S MONOGRAPHS ON i^&.iJIOCHEMICAL SUBJECTS

]\ Marine Biological Laboratory Library

I Woods Hole, Mass.

I

I I I

[0

•/'^"iw^v.

Presented by

J John Wiley and Sons, Inc,

J] New lork City

I I

IB

B3^{

CD

^~ m

^_ CD

METHUEN'S

MONOGRAPHS ON

BIOCHEMICAL SUBJECTS

General Editors : SIR RUDOLPH PETERS, f.r.s. and F. G. YOUNG, f.r.s.

THE NITROGEN METABOLISM OF MICRO-ORGANISMS

The Nitrogen Metabolism of Micro-organisms

B. A. FRY,

B.A., Ph.D.

Lecturer in Microbiology

in the University of Sheffield

WITH 3 PLATES & 14 DIAGRAMS

LONDON: METHUEN & CO. LTD NEW YORK: JOHN WILEY & SONS, INC.

First published in ig55

I.I

CATALOGUE NO. 4141/u (mETHUEN)

PRINTED AND BOUND IN GREAT BRITAIN BY BUTLER AND TANNER LTD., FROME AND LONDON

PREFACE

Little reflection is required to realize that nitrogen is a constituent of numerous compounds of biological interest, and all acquainted with present-day biochemistry are aware that during the last ten years the former emphasis on the study of the degradation of complex substances has been largely replaced by an active interest in mechanisms of syn- thesis, and in particular, in the synthesis of proteins and the metabolic role of the nucleic acids. Micro-organisms are proving to be of great value in the unravelling of the routes whereby amino-acids, nucleotides and other compounds are synthesized in mvo, and for a long time they have been used with great success in experiments designed to elucidate the functions of the many water-soluble substances now in- cluded in the B group of vitamins.

In this monograph an attempt has been made to survey as comprehensively as possible the nitrogen metabolism of micro-organisms and to treat the subject in such a manner as to reflect current trends in modern microbiology. The monograph is based on a series of lectures given in a one- year post-graduate course of microbiology held in the Uni- versity of Sheffield, and it is hoped that advanced students at other universities and research workers in allied fields will find it a convenient and concise introduction to one impor- tant section of microbial biochemistry. If it be thought that some topics receive more attention than they warrant, then the author accepts full responsibility for his choice and de- fends it on the grounds that these topics either encompass ideas of wider significance or serve to focus attention on how little has really been established. Though the title of the monograph is all-embracing and in the text examples are drawn from experiments with bacteria, fungi, algae and pro- tozoa, the m.ain emphasis is naturally on the first two of these four groups, since most work has been done with species of bacteria and yeasts. There is not space to mention every

vi NITROGEN METABOLISM

organism which has been studied, but the reader's search for additional information should be aided by the books re- commended for general reading and the detailed biblio- graphy appended to each chapter.

It is with very great pleasure that I record my thanks to Dr. S. R. Elsden for his interest in the preparation of this monograph and to him and Dr. J. L. Peel for reading the drafts of the various chapters and making many helpful sug- gestions. I am also grateful to Dr. E. F. Gale, F.R.S., for reading the completed manuscript.

Sheffield

1953

ACKNOWLEDGMENTS

I AM indebted to the following authors, editors and pub- lishers for their permission to reproduce figures which have appeared in the literature: Fig. 3.1, Prof. A. L. Audus and Fig. 10.2, Dr. P. Mitchell, and the Editors of Nature; Fig. 3.2, Prof. J. H. Quastel and Dr. P. G. Scholefield, and Messrs. Williams and Wilkins Co., U.S.A.; Fig. 4.1, Prof. P. W\ Wilson and the Editors of the Biochemical Journal', Figs. 6.1 and 6.2, Dr. E. F. Gale, F.R.S., the Academic Press Inc., U.S.A., and the Editors oi t\\t J ounml of General Microbiology; Fig. 8.1, Prof. L. Gorini and Prof. CI. Fro- mageot, and the Elsevier Publishing Co., Inc.; Plate I, Dr. B. Davis and the Editors of Experientia; Plate III, Prof. R. Tulasne and Dr. R. Vendrely, and the Long Island Bio- logical Association: Plate II is a photograph of a chromato- gram kindly prepared by Dr. R. Markham.

CONTENTS

CHAP. PAGE

PREFACE V

I INTRODUCTION I

II AMINO- ACID CATABOLISM lO

III NITRIFICATION AND DENITRIFICATION 32

IV THE FIXATION OF NITROGEN 45

V SYNTHESIS OF AMINO-ACIDS 6o

VI ABSORPTION OF AMINO-ACIDS BY MICRO-ORGANISMS 8o

VII PEPTIDES AND PROTEINS 95

VIII PROTEOLYTIC ENZYMES 112

IX NUCLEOTIDES AND NUCLEIC ACIDS 1 26

X MODE OF ACTION OF CHEMOTHERAPEUTIC AGENTS I45 INDEX 159

PLATES

FACING PAGE

I SYNTROPHISM AMONG ARGININE REQUIRING

MUTANTS OF ESCH. COLI 69

II ULTRAVIOLET PHOTOGRAPH OF CHROMATOGRAM

SHOWING SEPARATION OF PURINES AND PYRIMI- DINES IN YEAST NUCLEIC ACID I31

III PHOTOMICROGRAPHS OF A COLON BACILLUS BEFORE

AND AFTER TREATMENT WITH NUCLEASES 1 33

ao9o6

CHAPTER I

INTRODUCTION

Energetics of biological systems

In recent years the attempts to analyse the energetics of biological systems in terms of established thermodynamic principles have naturally focused much attention on the reactions in such systems which yield energy and those which utilize energy [6]. When energy is supplied to or liberated in a system, there are limitations regarding the conversion of one form of energy into another (Second Law of Thermodynamics). In other words, only part of the energy content of any system is available for doing further work, and this useful energy is termed /r^^ energy. Chemical reactions in which there is an output of free energy are described as exergonic and those in which there is an uptake of free energy as endergonic. Reproduction, growth and the maintenance of life are all endergonic processes and are therefore intimately associated with mechanisms able to supply them with energy.

It is generally believed that energy becomes available in biological systems as the direct or ultimate result of oxida- tion reactions [8, lo]. The oxidation of one substance must necessarily be accompanied by the reduction of another and a biological oxido-reduction reaction involves the trans- fer of hydrogen atoms or electrons [14]. Consequently the substance which is oxidized is sometimes described as being a hydrogen donor [H-donor], whilst the one being reduced is termed the hydrogen acceptor [H-acceptor]. The com- plete oxidation of any one substance proceeds by one or more simple steps, each catalysed by the appropriate enzyme, and in all known reactions the transfer of hydrogen atoms or electrons to the ultimate H-acceptor is effected by one or more intermediate carriers. In aerobic organisms, molecular oxygen serv^es as the H-acceptor and, according to the enzyme concerned, the end-product is water or

2 NITROGEN METABOLISM

hydrogen peroxide. Organisms which by chance or by necessity are Hving in an anaerobic environment must use a substance other than oxygen for this purpose. Such a sub- stance may be derived from the environment (e.g. CO 2 , nitrate or acetate) or may be a product of the organism's cataboHsm (e.g. in the lactic acid bacteria, pyruvate is reduced to lactate).

The esterification of inorganic orthophosphate is an integral part of the mechanism whereby endergonic reac- tions are able to utilize the energy made available by oxido- reduction reactions. Our conception of this mechanism is mainly due to Lipmann [10], who pointed out that phos- phorylated compounds can be divided into two groups according to the amount of energy released by their hydro- lysis: some yield about 3,000 cal. per mole whilst others liberate 10,000 to 16,000 cal. per mole. Lipmann proposed that the latter should be known as high-energy (or energy- rich) phosphate compounds and that they contain what he termed high-energy (or energy-rich) phosphate bonds, the hydrolysis of which yields 10,000 or more calories of free energy per mole of inorganic orthophosphate liberated. The significance of certain biological oxido-reduction reactions lies in the fact that they are associated with the formation of energy-rich phosphate bonds: these arise either during the actual oxidation of the organic substrate or else during the transfer of hydrogen (or electrons) to a H-acceptor. In the former case oxidation of the organic substrate is accom- panied by its esterification with inorganic orthophosphate and in consequence most of the energy made available by the oxidation reaction is not liberated as heat but is retained in the oxidized substrate in association with the newly in- corporated phosphate group. The only known example of an energy-rich phosphate group arising by a non-oxidative re- action is found in e«o/-2-phosphopyruvic acid, a substance formed by the dehydration of 2-phosphoglyceric acid under the influence of enolase. The phosphate groups and their associated energy can be transferred, in the presence of the appropriate enzyme (a phosphokinase), to adenosine diphos- phate (ADP), or sometimes to adenosine monophosphate

INTRODUCTION 3

(AMP), thus forming adenosine triphosphate (ATP) or ADP respectively.

In a biological system, the only known way in which the energy released by an exergonic reaction can be made available to an endergonic reaction, is for the two reactions to be coupled together by means of a substance which participates in both. This is the function of ATP, which by virtue of its high-energy phosphate groups acts as an energy carrier between reactions yielding energy and those utilizing energy. Adenosine triphosphate participates in the latter by reacting with, and thus activating, one of the reactants, and by this means the total free-energy content of the reactants is raised to a value at least approximately equal to, and often far greater than, that of the products. From the standpoint of energy relationships, the conditions are now such as to favour formation of the products, and the utilization of ATP in this manner is accompanied by the appearance of inorganic orthophosphate.

Although it is generally accepted that the energy meta- bolism of all organisms is associated with energy-rich phosphate bonds, little is known about how they are formed except during the anaerobic catabolism of glucose and pyruvate. The results of contemporary research indicate that co-factors containing thiol groups probably play an important role both in the production of energy-rich phos- phate groups and in their utilization, and that the synthesis of thiol esters may be an essential intermediate stage in these reactions (cf. the role of glutathione in the triosephosphate dehydrogenase system [16], and coenzyme A (Co. A) in the synthesis of citric acid and other compounds [2]). A sub- stance having the properties of ATP is believed to be present in all organisms and ATP has in fact been isolated from yeast [cf. 4], green plants [i] and animals, but its occurrence in bacteria is based more on inference rather than its isola- tion in a pure state [3, 7, 9, 11, 12, 13, 15].

Nutrition: general aspects

Irrespective of the organism, the continuance of life and the synthesis of cytoplasm are dependent on the availability

4 NITROGEN METABOLISM

of the same basic materials, namely, mineral salts, water and sources of carbon and nitrogen together with a mechanism providing energy in a form that can be utilized in biological systems. Autotrophs are organisms whose carbon require- ments are entirely satisfied by CO 2 (perhaps in some by CO). On the other hand, heterotrophs require a more com- plex carbon source, i.e. an organic compound, as well as CO 2 . Moreover, heterotrophs usually derive their energy by catabolism of the organic carbon source and are therefore to be contrasted with autotrophs which obtain their energy either from light (photosynthetic autotrophs) or by the oxidation of inorganic substances, e.g. H2S, S, NagSaOa , NHt, NOi", H2 or Fe"*""^ (chemosynthetic autotrophs). Each chemosynthetic autotroph oxidizes one specific com- pound, or in certain cases, a limited number of chemically related compounds, and presumably part of the energy released during these oxidations becomes available in the form of energy-rich phosphate groups. How the light energy absorbed by the chlorophyll of photosynthetic organisms becomes converted into a form that can be utilized in enzymic reactions is not yet known, though recent experi- ments have provided some indications of a possible mechanism [17].

All autotrophs derive their nitrogen from an inorganic source and, depending on the organism, use molecular Ng , NHt J nitrate or nitrite. Although one or more of the latter may serve as a complete source of nitrogen for certain heterotrophs, the nutritional requirements of many of these organisms are not so simple. It appears that such hetero- trophs are unable to synthesize one or more of the organic constituents of cytoplasm and they are therefore only able to grow if these substances are present in their environment, i.e. they are exacting towards these substances. The ability to synthesize complex organic nitrogenous compounds is especially variable, and whilst some organisms are exacting towards only one compound, e.g. Salmonella typhosa to tryptophan and Proteus vulgaris to nicotinic acid, the nutri- tion of other heterotrophs is far more complex, e.g. Leuco- nostoc mesenteroides P-60 requires eighteen amino-acids and

INTRODUCTION 5

at least eleven growth factors. (The term growth factor is used here in the same sense as vitamin in animal nutrition.) With regard to the amount of carbon used for the synthesis of cellular material, the contribution of the organic com- pound serving as a source of carbon and energy varies in- versely with the number of cytoplasmic constituents which the heterotroph derives preformed from the environment: in a rich medium this compound may function primarily as a source of energy. At one time, autotrophs were diiferen- tiated from heterotrophs on two counts: firstly that hetero- trophs were unable to incorporate the carbon of CO 2 into organic molecules, and secondly that autotrophs live entirely and exclusively at the expense of inorganic substances. There is now adequate information to show that both of these statements require modification [18, 19, 5]. The growth of heterotrophs is in fact dependent on the presence of CO 2 and they are known to possess enzyme systems accomplishing its fixation: but, although essential, CO2 is neither a complete nor a major source of carbon for hetero- trophs. Furthermore, it has been established that several organisms regarded as autotrophs can exist heterotrophic- ally. For example, in the presence of a suitable H-donor, the purple sulphur bacteria (Thiorhodaceae) obtain their energy from light, whilst CO2 and NH3 (or N2) serve as complete sources of C and N. The H-donor may be an inorganic form of sulphur or an organic substance such as a fatty acid, and the Thiorhodaceae can therefore be regarded as facultative autotrophs. On the other hand, the green sulphur bacteria use only an inorganic H-donor and appear to be obligate autotrophs. The Athiorhodaceae (non-sulphur purple bac- teria) require certain growth factors and usually an organic H-donor, i.e. they are heterotrophs, although they too- derive their energy from light.

Synopsis of monograph

Many organisms can derive their energy either directly or indirectly from nitrogenous compounds, and examples of this feature of their metabolism are given in separate:

6 NITROGEN METABOLISM

chapters devoted firstly to the fermentation and oxidation of amino-acids by heterotrophs, and secondly to the auto- trophic nitrifying bacteria. It will be seen that in nature the ammonia produced during the decomposition of amino- acids may suffer one of three fates: (i) oxidation by the nitrifying bacteria to nitrate (Chap. Ill), (2) after oxida- tion to nitrate, conversion to molecular Ng ^^^ nitrous oxide (Chap. Ill), (3) incorporation into organic molecules (Chap. V). The anabolic aspects of nitrogen metabolism culminate in the formation of two major groups of complex substances, proteins and nucleic acids. The latter are con- sidered in a separate chapter whilst protein synthesis is traced step by step, beginning with the mode of incorpora- tion of nitrogen from molecular Ng and NH3 into organic molecules. After dealing with the synthesis of amino-acids and with the mechanisms operative in the absorption of these compounds from the environment, attention is next directed to the significance of peptides in intermediary meta- bolism, the problems of protein synthesis and how amino- acids become joined together by peptide bonds. This part of the monograph concludes with a chapter devoted to the enzymes responsible for proteolysis, a process which ulti- mately yields free amino-acids. The catabolism of the latter is discussed at the beginning of the monograph, conse- quently it will be appreciated that the metabolism of amino- acids and proteins has been studied at various stages in a cycle. The underlying theme of the monograph is none other than that known to all biologists as the nitrogen cycle, and an attempt has here been made to analyse some of the component steps of the cycle from the standpoint of the biochemistry of the various reactions and the micro- organisms concerned (Fig. i.i).

Purely for convenience, and in order to avoid possible confusion, the microbial metabolism of nucleotides, nucleo- sides, purines and pyrimidines is discussed in a separate chapter. This field is now being studied intensively and there has been little time to correlate many of the experimental facts rapidly being placed at our disposal. Partly for this reason, and partly because of limitations in the amount of

CATABOLISM

ANABOLISM

denitrification

NO"

NO3 (ui)

nitrogen \

(Tv) NHjOH

fixation ^

PEPTIDES'*

FIG. I.I. — The Nitrogen Cycle. The roman numbers enclosed within circles denote the numbers of the chapters dealing with the various topics shown in the schenje

8 NITROGEN METABOLISM

Space available, the subject-matter of the chapter is con- fined to a few selected topics.

In the concluding chapter the mode of action of chemo- therapeutic agents is considered in terms of their observed effects on the metabolism of compounds containing nitrogen.

BIBLIOGRAPHY

The following books are recommended for general reading and as reference books for detailed information concerning specific topics:

FOSTER, J. w. (1949), Chemical Activities of Fungi, Academic Press, U.S.A.

FRY, B. A. and PEEL, J. L. (editors), (1954), Autotrophic Micro- organisms, Soc. gen. Microbiol. Symp., 4, Cambridge Univer- sity Press, G.B.

GALE, E. F. (1949), Chemical Activities of Bacteria, University Tutorial Press, G.B.

LWOFF, A. ( 1 95 1 ), Biochemistry and Physiology of Protozoa, Academic Press, U.S.A.

STEPHENSON, M. (1949), Bacterial Metabolism, Longmans Green, G.B.

SUMNER, J. B. and myrback, k. (editors), (1950), The Enzymes, Academic Press, U.S.A.

WERKMAn, c. H. and wilson, p. w. (editors), (195 1), Bacterial Physiology, Academic Press, U.S.A.

REFERENCES

1. Albaum, H. G., Ogur, M. and Hirshfeld, A. (1950), Arch.

Biochem., 27, 130

2. Barker, H. A. (1950), in Phosphorus Metabolism, i, 204 (Ed.

McElroy, W. D. and Glass, B., Johns Hopkins Press, U.S.A.)

3. and Lipmann, F. (1949), J. biol. Che?n., 179, 247

4. Dounce, A. L., Rothstein, A., Beyer, G. T., Meier, R. and

Freer, R. M. (1948), J. biol. Chem., 174, 361

5. Gest, H. (195 1), Bact. Rev., 15, 183

6. Hearon, J. Z. (195 1), Fed. Proc, 10, 602

7. Hersey, D. F. and Ajl, S. J. (1951), J. biol. Chem., 191, 113

8. Kaplan, N. O. in The Enzymes, 2 (i), Chap. 45

9. LePage, G. A. and Umbreit, W. W. (1943),^. biol. Chem., 147,

263; 148, 255

10. Lipmann, F. (1941), Advances in Enzymology, i, 99; (1946), 6,

231; (1949), Fed. Proc, 8, 597

11. Lohmann, K. (1928), Biochem. Z., 203, 164

INTRODUCTION

12. 13-

14- 15-

i6.

17- i8. 19.

Lutwak-Mann, C. (1936), Biochem. J., 30, 1405 Mesrobeaunu, L. (1936), Thesis: Paris, Contribution a Vetude

des corps puriques de la cellule bacterienne Michaelis, L. in The Enzymes, 2 (i), Chap. 44 O'Kane, D. J. and Umbreit, W. W. (1942). J. biol. Chem.,

142, 25 Racker, E. and Krimsky, I. (1952), Nature, 169, 1043 Vishniac, W. and Ochoa, S. (1952), J^. biol. Chem., 195, 75 Umbreit, W. W. (1947), Bact. Rev., 11, 157 Bacterial Physiology, Chaps. 11 and 19

CHAPTER II

AMINO-ACID CATABOLISM

Many heterotrophs can utilize organic nitrogen compounds, in addition to carbohydrates, as primary sources of carbon and energy. In general, the nitrogen is first removed from the compound and the product is then fermented or oxidized by the same terminal pathways that are operative in the catabolism of carbohydrates and fatty acids. Certain hetero- trophs, apparently lacking the ability to metabolize exogen- ous sugars, are entirely dependent on organic nitrogen compounds, such as amino-acids, purines or pyrimidines, as sources of carbon and energy. Although the end-products of the catabolism of these organisms have been studied, little is yet known about the routes of their formation or the enzymes responsible for the individual steps.

The catabolism of amino-acids commences either with an oxidative deamination or with the removal of a specific group by a non-oxidative process. It is unlikely that the latter is directly responsible for making energy available to the organism, but in either case examples are known in which the further metabolism of the products proceeds by routes which result in the formation of energy-rich phos- phate groups. Thus pyruvate may arise by the non-oxida- tive deamination of serine (p. 23) or the oxidative deamina- tion of alanine (p. 11), and its oxidation by the pyruvic oxidase system is accompanied by the formation of energy- rich phosphate groups [39]. The first part of this chapter is concerned with mechanisms and enzymes which accom- plish the oxidative catabolism of amino-acids, whilst the second part is devoted to enzyme systems whose primary mode of attack is non-oxidative.

Amino-acid oxidases

The amino-acid oxidases oxidize amino-acids to the corresponding keto acids and are specific for either the L or

10

AMINO-ACID CATABOLISM U

the D stereo-isomers of their substrates,

RCH(NH,)COOH+H20=RCOCOOH+NH3+2H

The transfer of hydrogen from the amino-acid to a suitable acceptor, typically Og , appears to be mediated by one or more carrier substances, and usually the enzyme has a prosthetic group capable of functioning in this manner. Enzymes of this type are the L-amino-acid oxidases of Neurospora crassa and N. sitophila [7], Proteus vulgaris [58], Penicillium notatum and Aspergillus niger [37]. Each of these oxidases attacks a wide variety of amino-acids, although the possibility that the observed activity is due to several very similar, but specific, enzymes has not been ruled out. Oxygen can be replaced in vitro by reducible dyes, such as methylene blue, or by ferricyanide. There is evidence that the enzyme from N. crassa possesses a prosthetic group, adenine flavindinucleotide, which enables hydrogen to be transferred directly to Og , resulting in the formation of H2O2 [10]. In the presence of catalase (present in Neuro- spora), the oxidation of one gram mole of amino-acid involves the overall uptake of one gram atom of oxygen. The mycelium of N. crassa also contains a similar oxidase specific for D-amino-acids [7].

Whether the oxidase from Pr. vulgaris also has a flavin prosthetic group has not yet been established, and although one atom of oxygen is taken up per molecule of amino-acid oxidized, there is no evidence that HgOg is first formed and subsequently decomposed by catalase. There must be more than one oxidase in Pr. vulgaris since washed suspensions oxidize more amino-acids than the cell-free enzyme pre- paration [58]. Oxygen is required for the deamination of glycine, alanine and glutamic acid by washed cell sus- pensions of Escherichia coli, Pseudomonas fluorescens and Bacillus mycoides [cf. 25]: using cells treated with toluene to prevent the further metabolism of pyruvate, it can be shown that the deamination of alanine by Esch. coli proceeds quantitatively according to the following equation:

CH3CH(NH2)COOH+i02=CH3COCOOH+NH3

12 NITROGEN METABOLISM

but nothing is known about the properties of the enzyme concerned.

Glutamic acid dehydrogenase

The deamination of glutamic acid by Esch. coli is due to the enzyme L-glutamic acid dehydrogenase with the co- enzyme triphosphopyridine nucleotide (TPN) [2]: the end- product is a-ketoglutaric acid and it is believed that the reduction takes place in two stages:

COOH.(CH2)2CH(NH2)COOH+TPN+ ^

COOH.(CH2)2C(:NH)COOH+TPN.H+H+

COOH.(CH2)2C(:NH)COOH+H20 ^

COOH.(CH2)2COCOOH+NH3

The system is reversible, with the equilibrium in favour of the synthesis of glutamic acid. A similar specific glutamic dehydrogenase occurs in Saccharomyces cerevisiae [i], Clostridium sporogenes [44], N. crassa [23], and probably in Haemophilus pertussis [34] and H. parainfluenzae [36]. Haemophilus influenzae will not grow except in the presence of a porphyrin (the X-factor) and diphosphopyridine nucleotide (DPN), TPN or nicotinamide riboside (the V- factor). The oxidative activity of cells harvested from a medium deficient in the V-factor was considerably increased by the addition of DPN or TPN, and in this way Klein has shown that the latter are involved in the oxidation of aspartic and glutamic acids to CO 2 , NH3 and acetic acid [36]. Unlike H. parainfluenzae, no volatile fatty acid was formed during the oxidation of amino-acids by H. pertussis, an organism which requires neither X nor V fac- tors, and in the experimental conditions employed by Jebb and Tomlinson, only carbon from glutamic acid was incor- porated into cell substance [34].

Oxidation of tryptophan by Pseudomonas spp.

A characteristic feature of species of Pseudomonas is that they possess or quickly acquire the ability to utilize any one of a wide variety of oxidizable organic substances as sources

AMINO-ACID CATABOLISM I3

of carbon and energy, a property which has been widely exploited by Stanier in the elucidation of metabolic path- ways by the technique of 'simultaneous adaptation' [54]. If an organism exhibits little or no detectable activity against a certain substance, and if the inclusion of this substance in its environment evokes, in the absence of cell division, a marked increase in the organism's ability to metabolize that substance, then adaptation is said to have taken place. If an organism can metabolize a particular compound, the hypo- thesis of simultaneous adaptation postulates that it can also metabolize immediately, and at a comparable rate, any sub- stance which is an intermediate in the metabolism of that compound (assuming that the intermediate can pass into the cells). If there is a lag period prior to the rates of utilization becoming comparable, then it may be concluded that the substance cannot be attacked by the existing metabolic systems, in other words, it is not an intermediate, and is only metabolized after adaptation has taken place. From such data it may be possible to deduce the probable route by which a substance is catabolized, but unequivocal proof requires not only direct evidence of formation of the inter- mediates but also isolation of the appropriate enzymes. The aerobic nature of the Pseudomonas spp. means that the overall catabolism of whole cells can be studied mano- metrically in terms of an uptake of O2 , and an example of this technique is provided by the investigations concerned with the degradation of tryptophan [55]. After being grown on, or otherwise adapted to tryptophan, some strains of Pseudomonas are simultaneously adapted to formylkynure- nine, kynurenine, anthranilic acid and catechol; whilst others are adapted to kynurenine and kynurenic acid, but not to anthranilic acid or catechol. Work with cell-free extracts [30] revealed that the pyrrole ring of tryptophan (Fig. 2.1) is first ruptured by a peroxidase-oxidase system in which both H2O2 and Og are involved, and the product, formylkynurenine, is then hydrolysed by formylase into formic acid and kynurenine. In some strains, the pyrrole ring is now reformed, thus producing kynurenic acid, but the route by which this substance is metabolized remains

14

NITROGEN METABOLISM

unknown. In other strains, kynurenine is split by kynuren- inase into alanine and anthranilic acid. The oxidation of the latter to COg proceeds via catechol, m-m-muconic acid, and ^-ketoadipic acid. Whilst the three enzyme systems, tryptophan peroxidase, kynureninase and pyrocatechase, were all highly active in cell-free extracts from organisms

tryptophan /^\ j|CH2Ch(nH2)cOOH

formyl kynurenine

COOH

CH.COOH

I CH.CH

II HOOC.CH

muconic acid

HOOC.CH2CO(CH2)2COOH

/g-ketoadipic acid

FIG. 2.1. — Pathways of tryptophan degradation in Pseudomonas spp,

grown in the presence of tryptophan, extracts from cells grown on asparagine exhibited negligible activity. It is notable that in none of the oxidative reactions could dyes like methylene blue replace O2 as the H-acceptor. The degradation of tryptophan via catechol has been termed the 'aromatic pathway' whilst that by way of kynurenic acid is known as the 'quinoline pathway'. Animal tissues degrade tryptophan by the former route. The routes of tryptophan

AMINO-ACID CATABOLISM 15

catabolism by Bacillus subtilis, another aerobe, appear to be similar to those found in the Pseudomonas spp. [40].

Tryptophanase

The appearance of indole in the culture medium is a valuable diagnostic test in bacteriology and it is readily detected by the formation of a pink compound in the presence of ^-dimethylaminobenzaldehyde and acid. Hop- kins and Cole were the first to isolate tryptophan and imme- diately suspected and proved that it was the natural precursor of this indole. The enzyme system concerned has been termed tryptophanase, and is found in some species of Escherichia, Proteus and Vibrio, though there are strain differences in any one species. Oxygen is probably not directly involved in the initial step of the tryptophanase reaction, but indole tends to accumulate only in aerobic conditions; e.g. in the absence of Og , Esch. coli formed indolepropionic acid and little, if any, indole. When washed cells of Esch. coli were incubated with tryptophan and the system adequately aerated, the uptake of oxygen corres- ponded to that required for the complete oxidation of the alanine side chain [64]. Most of the experiments concerned with the mode of action of tryptophanase have been per- formed with preparations of Esch. coli, and prior to attempt- ing to prepare the enzyme system in a cell-free state atten- tion was directed to the factors affecting its activity in whole cells [29]. Tryptophanase was found to be adaptive, and extremely active cells were obtained from vigorously aerated media containing tryptophan. Whether Esch. coli grown in the absence of the substrate exhibits detectable trypto- phanase activity appears to depend on the strain of the organism concerned. The inclusion of glucose in the medium may result in the suppression of indole formation, and if it does, the cells from such cultures do not exhibit tryptophanase activity. This effect is not due to growth in an acidic medium since other sugars are metabolized with the formation of acidic end-products and yet they do not suppress the adaptive formation of tryptophanase. How- ever, several workers have reported activity in cells derived

l6 NITROGEN METABOLISM

from cultures grown in the presence of glucose and a probable explanation of their results is that tryptophanase is only developed after all or most of the glucose has been decomposed. There is evidence that this enzyme system is not developed in cultures grown in an amino-acid rich medium containing glucose because phenylalanine and tyrosine exert an inhibitory effect. The specificity of trypto- phanase is high and indole is only formed from compounds related to L-tryptophan provided that the a-carboxyl and a-amino groups, the /^-position in the side chain, and the N of the indole ring are unsubstituted.

Wood, Gunsalus and Umbreit [63] have obtained from Esch. coll a cell-free preparation which attacked tryptophan with the formation of equimolecular amounts of indole, NH3 and pyruvic acid. Their tryptophanase preparation was activated by pyridoxal phosphate and would not deaminate either serine or alanine. They therefore concluded that neither of these amino-acids is an intermediate in the degradation of tryptophan by this route. Dawes and Hap- pold have performed similar experiments and reached the same conclusion. Although their system produced equi- molecular amounts of indole and pyruvic acid there was an excess of NH3 . Whilst no correlation with tryptophan dis- appearance was attempted, these observations may indicate that the initial step is one of deamination which is perhaps catalysed by a type of L-amino-acid oxidase. If this were so, the formation of ^-indolepyruvic acid would simultaneously make hydrogen available for the reductive rupture of the bond linking indole to the beta carbon of pyruvic acid:

i/^. nCH2CH(NH2)COOH

+ H0O, -NH3, -2H

NH

nCH,COCOOH

/\ . CH3

T^ Ml + CO

\A/ I

NH NH COOH

There is little direct evidence to support this hypothesis,

AMINO-ACID CATABOLISM I7

although it was noted that in addition to pyridoxal phos- phate, the enzyme preparations contained riboflavin, a pos- sible carrier of hydrogen. The bond joining indole to the side chain appears to be susceptible to reduction since indole is readily formed in vitro when tryptophan is either refluxed with Raney nickel and absolute alcohol containing a little HCl, or boiled with aqueous NaOH and catalytic amounts of Cu"^"^ or Co"^"*". A summary of the other mechanisms which have been proposed to explain the mode of action of trypto- phanase will be found in the review by Happold [29].

With the possible exception of tryptophanase, all the enzyme systems discussed above accomplish the oxidative catabolism of amino-acids in association with molecular O2 as the ultimate and natural H-acceptor. In anaerobic organ- isms either amino-acids themselves or compounds derived from them may fulfil this function.

Stickland reaction

Stickland was the first to demonstrate that amino-acids take part in anaerobic oxido-reduction reactions, certain acids acting as H-donors whilst others function as Pi- acceptors. The original experiments were done with the strict anaerobe CI. sporogenes which is capable of growing in an amino-acid medium in the absence of carbohydrates, and he suggested that the organism derived its energy from reactions of this type [56]. Nutritional studies later revealed that the organism only grew well in a medium containing adequate amounts of the amino-acids shown by Stickland to be H-acceptors and H-donors [22]. In Stickland's experi- ments, washed cell suspensions were incubated anaerobi- cally with the appropriate substrates in Thunberg tubes. Hydrogen-donor amino-acids were detected by their ability to reduce methylene blue or cresyl blue to the colourless leuco form. Alanine, valine, leucine and pyruvate were all active H-donors whilst phenylalanine, aspartic and glutamic acids showed some activity, reduction of the dye being accompanied by deamination of the amino-acid. Hydrogen acceptors were detected by their ability to accept hydrogen from the leuco form of a dye of suitable redox potential

l8 NITROGEN METABOLISM

(e.g. phenosafranine and benzyl viologen, but not methy- lene blue) and thus restore the original dye colour. Proline, hydroxyproline and glycine were reduced by such a system, though ammonia was only formed from glycine. Incubation of alanine with proline resulted in the production of i mole of NH3 per mole of alanine: none was produced from the separate amino-acids. Stickland thus demonstrated that CI. sporogenes catalysed oxido-reduction reactions between pairs of amino-acids, one acid acting as a H-donor, the other as a H-acceptor. The reduction of proline resulted in opening of the ring and the formation of (5-aminovaleric acid, whilst the products derived from alanine were NH3 , CO 2 and acetic acid. When cresyl blue accepted hydrogen from alanine, 2 moles of dye were decolorized for each mole of NH3 released, indicating that the overall oxidation of alanine involved the donation of four hydrogen atoms. Although there was no direct evidence, it seemed highly probable that pyruvate was an intermediate in the decompo- sition of alanine and Stickland therefore proposed that the overall reaction {d) represented the sum of three separate reactions (a, b, c):

CH3CH(NH2)COOH+H.30=NH3+CH3COCOOH+2H (a) CH3COCOOH+H20=CH3COOH+C02+2H (b)

4H+2NH(CH2)3CHCOOH=2NH2(CH2)4COOH (c)

CH3CH(NH2)COOH+2NH(CH2)3CHCOOH+2H20

=2NH,(CHo)4COOH+NH3+C02+CH3COOH (d)

similarly:

CH3CH(NH2)COOH+2NHoCH,COOH+2H20

=3NH3+C02+3CH3COOH

Clostridium botiilinum [15] and all the proteolytic Clos- tridia examined by Nisman, Raynaud and Cohen [45] were capable of performing the Stickland reaction; amongst the organisms which could not were CI. tetani, CI. tetano- morphum, CI. welchii and CI. saccharobutyricum. Apart from the substances already mentioned, histidine, serine, iso-

AMINO-ACID CATABOLISM I9

leucine, tyrosine, methionine, ornithine, tryptophan, phenylalanine, cysteine and ethanol also act as H-donors, whilst tryptophan, tyrosine, ornithine and arginine function as H-acceptors [33, 65]. It will be noted that some amino- acids serve both as an acceptor and as a donor, and whilst the reaction is specific for the L-isomer of H-donors [cf. 56], there is no stereochemical specificity w^ith respect to H-acceptors [65].

If CI. sporogenes is grown in the presence of glucose, it develops an active hydrogenase which enables the reducing component of the Stickland system to be replaced by mole- cular H2 [33]. The reaction can then be followed in terms of an uptake of Hg and the end-products are only those derived from the amino-acid added as the H-acceptor. In an analogous manner, and perhaps unexpectedly since the organisms are strict anaerobes, the H-acceptor part of the system can be replaced by Og [46]. A number of amino- acids, all H-donors in the Stickland reaction, were oxida- tively deaminated by washed suspensions of CI. sporogenes to the corresponding a-keto acid. The uptake of Og was appreciably reduced by the presence of a H-acceptor amino- acid and restored to its former value by the addition of arsenite. The latter can readily be explained since although arsenite completely inhibited H-transfer to a H-acceptor amino-acid (e.g. from leucophenosafranine to proline) it had no eflFect on the H-donor part of the system (e.g. the reduction of phenosafranine by alanine) [47]. Hence the first step in the Stickland reaction is probably catalysed by a type of L-amino-acid oxidase which, perhaps with the aid of one or more carriers, can transfer hydrogen to Og, a dye or another amino-acid. In cell-free extracts, DPN but not TPN was readily reduced by H-donor acids [44], but whether reduced DPN is the natural carrier and reacts directly with H-acceptor amino-acids is not known. Pyru- vate and other a-keto acids are oxidized by CI. sporogenes to CO 2 and a fatty acid containing one less carbon atom by a mechanism which can lead to the fornjation of energy- rich phosphate groups (cf. reaction h):

CH3COCOOH+H3PO4— > CHgCO^POsHa+zH+CO., [47]

20 NITROGEN METABOLISM

Nisman and his colleagues therefore support Stickland's concept of the reaction mechanism and they believe that reactions «, h and c are catalysed by an L-amino-acid oxidase [51], a keto acid oxidase [cf. 39] and an amino-acid reductase respectively. There is a complete lack of know- ledge concerning the mechanism by which an amino-acid is reduced to a fatty acid (reaction c). Reaction a is inhibited by KCN and secondary octyl alcohol, b by iodoacetate and c by arsenite [47]. The growth of CI. sporogenes at the ex- pense of energy derived from amino-acids is characterized by the production of acetic acid together with isobutyric, isovaleric and optically active valeric acids, derived respec- tively from valine, leucine and isoleucine by the Stickland reaction. On the other hand, owing to its inhibitory effect on the Stickland reaction, the utilization of glucose as the energy source yields only acetic and butyric acids [16].

Fermentation of amino-acids by other organisms [20]

Certain organisms live anaerobically by the fermentation of one particular organic nitrogen compound whilst others, although not so specific, are restricted to the utilization of a small number of chemically related compounds. These organisms have usually been isolated by the enrichment culture technique and the anaerobic incubation of a sample of mud or soil in a medium containing an organic nitrogen compound as the major source of carbon. In the event of the organisms requiring certain growth factors, a small amount of yeast extract is usually included in the medium after the first transfer.

The anaerobic cocci Diplococcus glycinophilus [11], Micro- coccus anaerobius and M. variabilis [19] are specific for glycine, and no other substance is readily metabolized unless glycine is also present. If cultures of D. glycinophilus are not shaken, the overall fermentation is expressed by:

4NH2CH2COOH+2H20=4NH3+3CH3COOH+2COo

The fermentation is not a simple dismutation involving the oxidation of one molecule of glycine to CO 2 and NH3 and the reduction of three molecules to acetic acid and NH3.

AMINO-ACID CATABOLISM 21

Experiments with isotopically labelled glycine have shown that most of the methylene carbon appeared as acetic acid whilst the carboxyl carbon appeared as CO 2 [4].

Alanine, serine or threonine serve as sole sources of carbon and energy for CI. propionicum and are fermented to CO 2 , NH3 and fatty acids [11, 12]: 3CH3CH(NH2)COOH+2HoO =

2CH3CH,COOH+CH3COOH+CO,+3NH3 3CHo(OH)CH(NH.,)COOH+H.,0 =

CH3CH2C6OH+2CH3COOH+2CO0+3NH3 The mechanism of alanine fermentation may be comparable with that proposed for the Stickland reaction, alanine acting both as the H-acceptor and as the H-donor. Lactate and pyruvate are fermented in a similar manner and these fer- mentations are at least superficially comparable to the fer- mentation of lactate by the propionibacteria. However, whilst in the latter propionic acid arises by the decarboxyla- tion of succinate, in CI. propionicum it is probably formed from acrylate [35]. Barker and Wiken have concluded that acetate is not an intermediate in the fermentation of threo- nine to butyric and propionic acids, and that butyric acid probably arises directly from a C4-compound (a-ketobu- tyrate?)[5].

Unlike D. glycinophilus and CI. propionicum, CI. tetano-

morphum ferments glucose as well as certain amino-acids.

The end-products of both glutamic acid and histidine

fermentations include Hg , CO2 , NH3 , acetic and butyric

acids [66], and by analogy with Edlbacher's work with liver,

Woods and Clifton were the first to suggest that glutamic

acid was an intermediate in the fermentation of histidine.

Confirmation of this hypothesis has been recently obtained

and the first step in the conversion of histidine (I) to

glutamic acid, HCOOH and NH3 involves deamination to

urocanic acid (II) [60].

CH=CH.CH2CH(NH2)COOH

h ^i >

N3 iNH -NH3

CH

(I)

22 NITROGEN METABOLISM

CH=CH.CH:CHCOOH (CH2)2COOH

II I 1

N NH — >CH(NH2) +HCOOH+NH3

\ / I

CH COOH

(11) Urocanic acid and glutamic acid are also intermediates in the oxidation of histidine by Ps. fluorescens, and isotopes have been used to show that the amino-nitrogen of glutamic acid and the carbon of HCOOH are derived from the N and C in positions i and 2 of the imidazole ring [59]. The fer- mentations of C/. tetani [48Z>]and CI. cochlearum [3] resemble those of CI. tetanomorphum. The former ferments a number of amino-acids, only one of which, histidine, is among those essential for growth; aspartic acid and serine give rise to alcohols as well as to fatty acids. It is well known that yeasts fermenting carbohydrates in the presence of amino-acids produce a number of the higher aliphatic alcohols (fusel oil), and Ehrlich showed that the latter contain one less carbon atom than the amino-acids from which they were derived, the overall reaction being:

Such a process may be the means whereby the nitrogen of the amino-acid is made available in the form of NH3 , but the details of the mechanism are unknown. Most of the alcohols are formed after the amino-acids have disappeared from the medium, indicating that there are intermedi- ate stages between deamination and production of the alcohol [13].

Anaerobic ac- deaminases

A characteristic feature of the anaerobic a- deaminases is that the reaction product is unsaturated and therefore such enzymes are sometimes described as desaturases. One typical example has already been mentioned, namely, the enzyme forming urocanic acid from histidine, and there is evidence that it also occurs in certain strains of Esch. colt, Salmonella paratyphi and Shigella paradysenteriae [50]. A similar type of reaction is catalysed by aspartase, an enzyme

AMINO-ACID CATABOLISM 23

found in Esch. colt, Ps. fluorescens, Serratia marcescens, Pr. vulgaris, Lb. casei, and perhaps in yeasts [67, 17, cf. 21]:

COOH.CH2CH(NH2)COOH # COOH.CHrCH.COOH+NHs

The system is reversible and by using cells treated with cyclohexanol to prevent conversion of the fumarate to suc- cinate or malate, it can be shown that the equilibrium favours the synthesis of L-aspartic acid [67]. Serine and threonine are deaminated anaerobically to pyruvic and a-ketobutyric acid respectively, though whether one de- aminase catalyses both reactions is not yet known. Unlike aspartase, the reaction does not appear to be reversible and the first step is thought to be the removal of the elements of water, followed by the spontaneous hydrolysis of the resulting imino-compound [i4<2]:

CH2(OH)CH(NH2)COOH > (CH2:C(NH2)COOH) -^

CH3C(:NH)COOH

CH3C(:NH)COOH ^ CH3COCOOH+NH3

The dehydration step is analogous to that catalysed by enolase, and cysteine desulphurase may likewise be regarded as first removing the elements of HgS:

CH2(SH)CH(NH2)COOH+H20=CH3COCOOH+H2S+NH3

Serine and threonine deaminase activity is found in Esch. colt, CI. welchii, Ps. pyocyanea, Proteus OX-ig and staphylo- cocci: cysteine desulphurase occurs in Sac. cerevisiae, Esch. colt, Pr. vulgaris, B. subtilis and Propionibacterium pentosea- ceum [24].

Aspartase, cysteine desulphurase, and the serine and threonine deaminases are all alike in that their activity is dependent on the presence of certain co-factors, the identity of which has not yet been completely established. For example, there was a marked reduction in the aspartase activity of washed cell suspensions of Esch. coli after they had been kept standing in water or buffer. This decay in activity could be prevented by the addition of a small amount of either adenylic acid (AMP) or orthophosphate, together 3

24 NITROGEN METABOLISM

with a reducing agent such as cysteine, lactate or for- mate [28, 25]. Similarly, a loss of serine deaminase activity was prevented by glutathione, formate or AMP. During investigations of the nutrition of lactobacilli and Strep, faecalisy organisms exacting towards aspartic acid, Stokes, Larsen and Gunness observed that they grew well in the absence of this amino-acid provided the medium contained adequate amounts of biotin. They therefore suggested that biotin plays an importarit role in aspartic acid metabolism, and perhaps especially in the aspartase system, although they were unable to demonstrate the presence of the latter in these organisms [57].

Meanwhile Lichstein and his colleagues were studying the catabolism of aspartic acid and serine by Esch. coli, Pr. vulgaris and Bacterium cadaveris. They found that of the known B vitamins only biotin was capable of reactivating washed cell suspensions whose deaminase activity had been reduced by standing for 30 minutes in molar phosphate buffer, pH 4 [42]. This loss of deaminase activity and subse- quent reactivation by biotin or by much larger concentra- tions of AMP was only exhibited by cells which had been grown in the presence of a yeast extract. A concentrate of compounds containing biotin was prepared from yeast and was found to be a hundred times more effective in enhancing the aspartase activity of 'aged' suspensions than was to be expected on the basis of its biotin content. Moreover, when fresh, a partially resolved cell-free preparation of aspartase was activated either by yeast extract or by biotin plus adenylic acid, but after standing at 0° C, only the former was effective [41]. A bound form of biotin, termed biocytin, has been isolated recently from yeast, crystallized and shown to be £-N-biotinyl-L-lysine [69]. Biocytin itself is probably not the natural coenzyme, since, unlike the impure concen- trates, its activity is only comparable to that produced by an equivalent amount of free biotin [68]. Wood and Gunsalus have prepared from Esch. coli a purified cell-free system possessing serine and threonine deaminase activity, and although both adenylic acid and glutathione were required, biotin did not appear to be necessary [62\

AMINO-ACID CATABOLISM 25

Furthermore, Binkley's dialysed cell-free preparation of serine deaminase was reactivated by Zn [8]. In contrast with this work with bacteria, it is now clear that Neurospora possesses two deaminases — one specific for the L isomers of threonine and serine [71], the other for the D isomers [70] — and that both of these enzymes are activated by pyridoxal phosphate. Biotin, AMP and GSH, either separately or together, did not affect the activity of cell-free prepara- tions. Though the D-serine deaminase of Esch. coli is like- wise activated by pyridoxal phosphate, attempts to show that the corresponding L-serine deaminase is activated by this substance have been unsuccessful [43^]. Pyridoxal phosphate is also reported to be the activator of the cysteine desulphurase of Proteus vulgaris [48^]. In view of these diverse observations, it is not yet possible to define the co- factors naturally associated with the anaerobic deaminases. The confusion in this field has recently been increased since a substance produced by heating glucose with acid under pressure was found to be as active as yeast extract in the activation of the resolved aspartase system of Bact. cadaveris and yet did not contain biotin [146]. Winzler, Burk and du Vigneaud [61] have observed that unless biotin was added to the system, washed cells of biotin-deficient Sac. cerevisiae were incapable of assimilating exogenous NH|. Their ob- servation can be readily explained, if, as Lichstein's work indicates, this growth factor is an essential component of aspartase, one of the systems by which inorganic nitrogen is incorporated into organic molecules (p. 60).

In general, the anaerobic deaminases are most active in cells harvested from cultures at the cessation of active cell-division. Moreover, the presence of Og favours the development of the aerobic amino-acid oxidases whilst anaerobic conditions favour the anaerobic deaminases. The metabolic quotients of the former (QnHs about 30) are very much lower than those of the latter (Qnh3=2oo-i,ooo). After growth in the presence of glucose, organisms usually possess poor deaminase activity, and although in general the optimum pH for deaminase activity is in the range 8-10, this effect cannot be explained in terms of growth in an

26 NITROGEN METABOLISM

acidic environment since activity is not increased by buffer- ing the medium. Nor must it be assumed that glucose is used preferentially as a source of carbon and energy and that in consequence the organisms tend not to synthesize the deaminases. On the contrary, the reduced activity may be a reflection of the lack of essential co-factors since Boyd and Lichstein [9] found that the low serine deaminase activity of a washed suspension of Esch. coli (grown in the presence of glucose) was almost immediately increased by the addition of biotin, adenylic acid, yeast extract or liver extract.

Having considered the enzymes which deaminate amino- acids, it is now convenient to deal with those which attack X)ther groups in the molecule.

Arginine dihydrolase

Washed suspensions of Strep, faecalis, Staph, aureus^ 'CI. septicum and CI. sporogenes decompose arginine into ornithine, NH3 and CO 2 :

C.NH(CH2)3CH(NH2)COOH+2H20

NH

=2NH3+C02+NH2(CH2)3CH(NH2)COOH

Since urea is not attacked by Strep, faecalis, the overall reaction cannot be explained in terms of the splitting of arginine into ornithine and urea by arginase and the subse- quent decomposition of urea by urease. Hills [32] therefore proposed that the enzyme system in Strep, faecalis and Staph, aureus should be known as arginine dihydrolase. Recent work has shown that more than one enzyme is in- volved. The first step in the reaction is the formation of am- monia and citrulline, NH2CONH(CH2)3CH(NH2)COOH, which is then degraded to ornithine by an enzyme system activated by inorganic phosphate, Mg"*""^ or Mn"^"*", and ATP or AMP [38, 52]. Strep, faecalis is exacting towards arginine, and arginine can be replaced by ornithine only if the medium also contains adequate amounts of CO 2 . Carbon

AMINO-ACID CATABOLISM 27

dioxide is not produced during the fermentation of glucose by Strep, faecalis, and since growth is not possible in its absence Gale has suggested that arginine dihydrolase is a mechanism whereby this metabolite is made available [26]. He found that strains with high arginine dihydrolase activity grew better if the amount of arginine initially present in the medium was such that it was not all decomposed by the time the pH became unfavourable for further growth. Alternatively, it is feasible that arginine serves as a source of energy and that arginine dihydrolase activity is connected with the organism's energy metabolism, since Knivett [38] has shown that the conversion of citrulline to ornithine is accompanied by the phosphorylation of ADP to ATP.

Amino-acid decarboxylases

Although for several years bacteria have been known to form amines from amino-acids, no study of the enzymes concerned was made until the work of Gale [27]. The amines arise by decarboxylation of a-amino-acids:

RCH(NH2)COOH=RCH2NH2+C02

and the initial experiments demonstrated the existence of six amino-acid decarboxylases, specific respectively for the L-isomers of tyrosine, lysine, ornithine, arginine, histidine and glutamic acid. They have been found in the genera Escherichia, Streptococcus, Clostridium, Proteus and Lacto- bacillus. Whilst an organism may possess more than one decarboxylase, the distribution of the enzymes amongst the strains of any one species is variable. These enzymes are active in the pH range 2-6 and have a sharp optimum. In general the substrates possess an a-amino and an a-carboxyl group together with another polar group at the opposite end of the molecule. Substitution in any of these groups yields substances which are not attacked and the introduction of a hydroxyl group in another part of the molecule is the only known structural modification which does not affect suscep- tibility to decarboxylation, e.g. ^-hydroxyglutamic acid, /5-hydroxylysine and 3:4-dihydroxyphenylalanine are decar- boxylated by the glutamic, lysine and tyrosine decarboxylases

28 NITROGEN METABOLISM

respectively. Because of their specificity and the fact that decarboxylation proceeds to completion, these enzymes are invaluable for the quantitative determination of the corresponding amino-acids. The analytical procedure origi- nated by Gale is based on the use of a washed suspension or an acetone powder of the appropriate organism and the manometric determination of the CO 2 released.

The growth conditions are especially important in deter- mining whether the cells develop highly active decar- boxylases, and owing to their adaptive nature, the medium must contain the specific substrate. The enzymes are not formed unless growth occurs in an acid environment, and this is usually attained by allowing the organisms to meta- bolize glucose. Temperature is also important: for example, the decarboxylase activity of Esch. colt is better when cul- tures are grown at 20-26° C. rather than at 37° C, but the opposite is true of Strep, faecalis and CI. welchii. Decar- boxylase activity becomes maximal at the cessation of active cell division. Except for the histidine enzyme, there is evidence that all the amino-acid decarboxylases possess a prosthetic group, codecarboxylase, the existence of which was first discovered by Gale during the purification of the lysine decarboxylase. Fractionation with (NH4)2S04 in alkaline conditions resulted in the precipitation of an in- active apoenzyme to which decarboxylase activity could be restored by the addition of extracts of bacteria, yeast or animal tissues. Gunsalus and his co-workers later identified codecarboxylase as being pyridoxal phosphate. It is there- fore essential that the growth medium should contain ade- quate amounts of pyridoxin, since even when the organism is not exacting towards this factor, the rate of synthesis may be insufficient to saturate the apoenzyme. The apoenzyme of tyrosine decarboxylase can be prepared [6] by growing Strep, faecalis in a pyridoxin-deficient medium containing D-alanine, a substance which, according to the strain, replaces or reduces the organism's pyridoxin requirements. Washed cells of bacteria grown in this way exhibit no activity unless the experimental system contains either synthetic pyridoxal phosphate, pyridoxal or natural code-

AMINO-ACID CATABOLISM 29

carboxylase, and they can therefore be used for assaying codecarboxylase .

More recent work has shown that the tyrosine de- carboxylase will also attack phenylalanine [43^2] and m- tyrosine [53], and that Esch. colt possesses a specific enzyme which decarboxylates a,e-diaminopimelic acid to lysine. Unlike the other decarboxylases, the activity of the latter is high even in cells grown in the absence of the substrate, and furthermore there is evidence that the reaction is reversible: like the other decarboxylases the prosthetic group appears to be pyridoxal phosphate [18]. Cultures of Proteus spp. grown in an amino-acid medium have been found to contain amines derived from the branched chain amino-acids valine, leucine and isoleucine, but their mode of formation is not yet known [49].

Fermentations usually give rise to acidic end-products, consequently as the fermentation proceeds, the pH of the medium eventually reaches a value below which no further growth is possible. The highly basic amines formed as the result of decarboxylase activity tend to counteract this fall in pH, and Gale has therefore proposed that the decar- boxylases may be regarded as a type of 'neutralization mechanism' [27]. Since the partial pressure of CO 2 in an acidic medium is low, another possible function of the decarboxylases is to make this essential metabolite available inside the cell. The observation that H. parainfluenzae is exacting towards putrescine, the amine formed by the decarboxylation of ornithine, implies that the amines them- selves may be of some significance in intermediary meta- bolism [31].

REFERENCES

1. Adler, E., Gunther, G. and Everett, J. E. (1938), Z. physiol.

Chem., 255, 27

2. Hellstrom, V., Gunther, G. and von Euler, H. (1938),

Z. physiol. Chem., 255, 14

3. Barker, H. A. (1937), Enzvmologia, 2, 175

4. Volcani, B. E. and Cardon, B. P. (1948), J. biol. Chem.,

173. 803

5. and Wiken, T. (1948), Arch. Biochem., 17, 149

30 NITROGEN METABOLISM

6. Bellamy, W. D. and Gunsalus, I. C. (1945), J. Bact., 50, 95

7. Bender, A. E. and Krebs, H. A. (1950), Biochem. J., 46, 210

8. Binkley, F. (1943), J. hiol. Chem., 150, 261

9. Boyd, W. L. and Lichstein, H. C. (1951), J. Bact., 62, 711

10. Burton, K. (1952), Biochem. J., 50, 258

11. Cardon, B. P. and Barker, H. A. (1946), J. Bact., 52, 629

12. and Barker, H. A. (1947), Arch. Biochem., 12, 165

13. Castor, J. G. B. and Gugmon, J. F. (1952), Science, 115, 147 14a. Chargaff, E. and Sprinson, D. B. (1943), J. biol. Chem., 151,

273 146. Christman, J. F. and Williams, V. R. (1952), J. Bad., 63, 107

15. Clifton, C. E. (1940), y. Bact., 39, 485

16. Cohen-Bazire, G., Cohen, G. N. and Prevot, A. R. (1948),

Ann. Inst. Past., 75, 291

17. Cook, R. P. and Woolf, B. (1928), Biochem. J., 22, 474

18. Dewey, D, L. and Work, E. (1952), Nature, 169, 533

19. Douglas, H. C. (195 1), J. Bact., 62, 517

20. Elsden, S. R. (1952), in The Enzymes, 2 (ii), Chap. 65

21. Erkama, J. and Virtanen, A. I. (1951), The Enzymes, i (ii),

Chap. 42

22. Fildes, P. and Richardson, G. M. (1935), Brit. J. exp. Path.,

16, 326

23. Fincham, J. R. S. (195 1), jf. gen. Microbiol., 5, 793

24. Fromageot, C. (195 1), in The Enzymes, i (ii), Chap. 41

25. Gale, E. F. (1940), Bact. Rev., 4, 135

26. (1945), Brit. J. exp. Path., 26, 225

27. (1946), Advances in Enzymology, 6, i

28. and Stephenson, M. (1938), Biochem. J., 32, 392

29. Happold, F. C. (1950), Advances in Enzymology, 10, 51

30. Hayaishi, O. and Stanier, R. Y. (1951), J. Bact., 62, 691

31. Herbst, E. J. and Snell, E. E. (1948), J. biol. Chem., 176, 989

32. Hills, G. M. (1940), Biochem. J., 34, 1057

33. Hoogerheide, J. G. and Kocholaty, W. (1938), Biochem. J.,

32, 949, 437

34. Jebb, W. H. H. and Tomlinson, A. H. (195 1), J. gen.

Microbiol., 5, 951

35. Johns, A. T. (1952), y. gen. Microbiol., 6, 123

36. Klein, J. R. (1940), y. biol. Chem., 134, 43

37. Knight, S. G. (1948), y- Bact., 55, 401

38. Knivett, V. A. (1952), Biochem. y., 50, xxx; (1953), 55, x

39. Korkes, S., Stern, J. R., Gunsalus, I. C. and Ochoa, S, (1950),

Nature, 166, 439

40. Kotake, Y. and Otani, S. (1933), Z. physiol. Chem., 214, i

41. Lichstein, H. C. (1949),^. biol. Chem., 177, 125

42. and Christman, J. F. (1948), y. biol. Chem., 175, 649;

(1949), J- ^acf., 58, 565

43a. McGilvery, R. W. and Cohen, P. P. (1948), y. biol. Chem.,

174, 813 436. Metzler, D. E. and Snell, E. E. (1952), J. biol. Chem., 198, 363

AMINO-ACID CATABOLISM 3I

44. Nisman, B. and Mager, J, (1952), Nature, 169, 243

45. Raynaud, M. and Cohen, G. N. (1948), Arch. Biochem.,

16, 473

46. and Vinet, G. (1949), Ann. Inst. Past., 77, 277

47. and Vinet, G. (i95o), Ann. Inst. Past., 78, 115

48a. Ohigashi, K., Tsunetoshi, A,, Uchida, M. and Ichihara, K.

(1952), J. Biochem., Japan, 39, 211 486. Pickett, M. J. (1943), J' biol. Chem., 151, 203

49. Proom, H. and Woiwod, A. J. (195 1), jf. gen. Microbiol., 5, 930

50. Raistrick, H. (1917), Biochem. jf., 11, 71

51. Rosenberg, A. J. and Nisman, B. (1949), Biophys. Biochim.

Acta, 3, 348

52. Slade, H. D. and Slamp, W. C. (1952), J. Bact., 64, 455

53. Sloane-Stanley, G. H. (1949), Biochem. J., 44, 373

54. Stanier, R. Y. (1950), Bact. Rev., 14, 179

55. Hayaishi, O. and Tsuchida, M. (195 1), J. Bact., 62, 355

56. Stickland, L. H. (1934), Biochem. J., 28, 1746; (1935), 29, 288,

889

57. Stokes, J. L., Larsen, A. and Gunness, M. (1947), J. Bact.,

54, 219

58. Stumpf, P. K. and Green, D. E. (1944), J. biol. Chem., 153,

387

59. Tabor, H., Mehler, A. H., Hayaishi, O. and White, J. (1952),

y. biol. Chem., 196, 121

60. Wickremasinghe, R. L. and Fry, B. A. (1954), Biochem. jf.,

58, 268

61. Winzler, R. J., Burk, D. and du Vigneaud, V. (1944), Arch.

Biochem., 5, 25

62. Wood, W. A. and Gunsalus, I. C. (1949), J. biol. Chem., 181,

171

63. Gunsalus, I. C. and Umbreit, W. W. (1947), jf. biol.

Chem., 170, 313

64. Woods, D. D. (1935), Biochem. y., 29, 640

65. (1936), Biochem. y., 30, 1934

66. and Clifton, C. E. (1937), Biochem. y., 31, 1774; (1938),

32, 345

67. Woolf, B. (1929), Biochem. y., 23, 472

68. Wright, L. D., Cresson, E. L. and Skeggs, H. R. (1950), Proc,

Soc. exptl. Biol. N.Y., 74, 334

69. Cresson, E. L., Skeggs, H. R., Reck, R. L., Wolf, D. E.,

Wood, T. R., Valiant, J. and Folkers, K. (195 1), Science, 114, 635

70. Yanofsky, C. (1952), y. biol. Chem., 198, 343

71- and Reissig, J. L. (1953), J. biol. Chem., 202, 567

CHAPTER III

NITRIFICATION AND DENITRIFICATION

Nitrification

Nitrosomonas and Nitrohacter are two genera of strictly aerobic chemosynthetic autotrophs which respectively obtain their energy by the oxidation of ammonia to nitrite (NOo), and nitrite to nitrate (NO3 ).

Nitrosomonas, NH3+iJ02=HN02+H20 Nitrohacter, N02+i02=N07

These organisms play an important role in the formation of NO"^ from NH3 and organic nitrogen compounds, a pro- cess occurring in soil and the beds of sewage purification works [7] and known as nitrification. Most organic nitrogen compounds are only nitrified by Nitrosomonas an^ Nitro- hacter after they have been degraded by heterotrophs to

NH3.

The first evidence in favour of Pasteur's suggestion that nitrification was due to micro-organisms came from the classic experiments of Schloessing and Muntz [27]. After sewage effluent had percolated through a column of sand and chalk for about twenty days, they noted that NH3 was being converted almost quantitatively into NO"^. This con- version was completely stopped if the column was subjected to conditions injurious to life, e.g. heat or chloroform, but commenced again after washing with non-sterile water derived from soil. Of the organisms known at that time, none could oxidize NH3 to NO 7, and the first attempts to isolate the causative agents by plating soil on nutrient gelatin media all ended in complete failure. The lack of success was traced to the inhibitory effect of organic sub- stances on the growth of nitrifying organisms and resulted in the introduction of media containing only inorganic

32

NITRIFICATION 33

salts [33]. After inoculating such media with soil, Wino- gradsky readily obtained excellent growth accompanied by nitrification, and the cultures were easily maintained through successive transfers in a mineral medium containing potassium phosphate, NH4CI, MgS04 and K2CO3 . Using the enrichment culture technique, the isolation of organisms oxidizing NH3 was facilitated by repeated subculture in an inorganic medium containing NHt but no added NOT: similarly, a medium containing N07 and no NH^ was used for organisms oxidizing NOT. Winogradsky solved the problem of how to obtain discrete colonies on a solid yet completely inorganic medium by using silica gel as the matrix for the mineral salts [34]. The colonies are minute (diam. 200 [.i) and micro-manipulators are sometimes used to pick out and transfer those required for inoculating sub- cultures. Winogradsky was the first to obtain indisputably pure cultures of organisms which specifically oxidized either NH3 (Nttrosomonas europaed) or NOT {Nitrohacter wino- gradsky). About the same time, Warington [31] and the Franklands [8], working independently, obtained cultures which oxidized NH3 to NOT and which appeared to be pure as judged by microscopic examination and the absence of growth on gelatin plates. The isolation of pure cultures of the nitrifiers is difficult because their slow growth favours the appearance of heterotrophs which grow rapidly even if only traces of organic matter are present in the medium. Great care is therefore required in assessing the purity of any culture of nitrifying bacteria and no growth should be detectable after incubating plates of nutrient agar streaked with such cultures. Lees has found that adequate aeration of the medium greatly facilitates the isolation and culturing oi Nttrosomonas [14].

Some strains of the nitrifiers are actively motile whilst others appear to be habitually associated with surfaces, e.g. they readily adhere to granules of CaCOg [cf. 19]. Many of these organisms prefer a slightly alkaline environment [27, 18], and for this reason CaCOg , K2CO3 or MgC03 are fre- quently added to the medium [34, 20, 12]. Such substances

34 NITROGEN METABOLISM

serve as a source of COg and at the same time prevent the H"^ produced during nitrification from lowering the pH to a value unsuitable for growth [20]. The optimum pH for the growth of a particular strain tends to be related to the pH of the soil from which it was isolated, e.g. one from a peaty soil will nitrify in relatively acid conditions [11]. Meyerhof observed that the growth of Nitrosomonas and Nitrohacter is inhibited by high concentrations of their respective substrates. Furthermore, high concentrations of NHt also inhibited the oxidation of NO^, and the higher the pH of the medium the greater the inhibition, both of growth and of nitrite oxidation. Both NHt and NO^ in excessive concentrations are known to be injurious to cells in general, and the effect of pH may be explicable on the basis that conditions of high pH favour the formation of unionized NH3 which may enter the cells more easily than the ammonium ion [20].

The autotrophic nature of the nitrifiers, together with their apparent inability to grow in organic media, led Wino- gradsky to conclude that they neither required nor utilized organic nutrients. There has been much controversy as to whether all organic materials are toxic to these organisms and whether they can in fact assimilate at least some of these substances. There is now a certain amount of evidence that their growth is stimulated by small amounts of peptone, by yeast extracts and by a partial hydrolysate of egg albumin [9]: moreover, some strains of Nitrosomonas can grow in the presence of high concentrations of glucose (10%) [11, cf. 12]. Some evidence that the nitrifiers may be able to obtain energy in a manner typical of heterotrophs, namely by the degradation of organic compounds, has been obtained by Bomeke. In manometric experiments with thick suspen- sions of Nitrosomonas and Nitrohacter y he found that in the absence of exogenous nutrients, these organisms absorbed O2 , indicating that they could obtain energy by the oxida- tion of endogenous reserve materials [6]. Whilst there have been many experiments purporting to show that organic media support the growth of the nitrifiers, the majority of

NITRIFICATION

35

these investigations have been discounted on the grounds that the cultures were contaminated with heterotrophs [35,

percolating fluid drawn up tube by suction

capillary tube

to suction pump «.

air or gas inlet and sampling tube

FIG. 3.1. — Lees' soil percolation apparatus, as modified by Audus [3]

12]. Quastel and Scholefield have pointed out that organic inhibitors fall into two groups, those which inhibit in con- centrations of 0-0 1 M. or less and those which inhibit only

36 NITROGEN METABOLISM

in very high concentrations [22]. But it must also be emphasized that the inhibitory action of any substance is determined by the organism concerned and also by the physical properties of the environment, e.g. peptone is far less inhibitory in the presence of sand than in ordinary liquid cultures [32].

Owing to their slow growth and the difficulty of isolating pure cultures, there have been few comprehensive investi- gations of the metabolism of Nitrosomonas and Nitrohacter. Meyerhof has dealt with the effect of substrate concentra- tion, inorganic ions and various inhibitors on the rate of nitrification in liquid cultures [20]. Nitrification under con- trolled conditions in soil has been studied more recently by Lees and Quastel using a technique based on the continuous repercolation of fluid through a column of soil [16]. This was accomplished by a simple apparatus which ensured that the soil was nearly saturated with water and was adequately aerated (cf. Fig. 3.1).

When NHt was added to the percolating fluid there was a lag period after which the rate of nitrate formation gradually increased and at the end of five days became linear. During the first two days small quantities of NO "2 were detected. A graph of the progress of nitrification is reminiscent of that expressing the rate of growth of a bacterial culture and it appears that the soil ultimately becomes saturated with nitrifying bacteria. At this stage the behaviour of the system is analogous to that of a washed cell suspension, i.e. it is metabolically active but the organ- isms are not actively dividing. When a fresh solution con- taining NH| was percolated through such a 'saturated soil', nitrification occurred at a linear rate from the beginning (Fig. 3.2). Consequently when the 'saturated soil' is treated with a substance which is nitrified directly, NO 7 formation should commence at once and at a maximal and linear rate. A lag period implies that the substance can only be nitrified after being converted into another compound and that time is required for the formation of the appropriate adaptive enzymes in the existing bacteria or for the growth of new

NITRIFICATION 37

organisms. For example, NO^ is immediately nitrified by soil which has been previously percolated with NHt, but there is a lag in the nitrification of NHt by soil previously treated with NO^. Hence the 'saturated soil' technique is of value in testing substances believed to be intermediates in the nitrification process. Thus hydroxylamine (NHgOH) has long been postulated as a possible intermediate in the conversion of NH3 to HNO2 , but there is little direct

i20-| f^ r-l-9

FIG. 3.2. — Course of nitrification in soil percolated with o-oi M.- NH4CI: A, first percolation of fresh soil; B and C, linear rela- tionship between time and logio /xg. N07-N/ml. formed respectively in the first (B) and second (C) percolations [22]

evidence [15]. Free NHgOH is toxic to bacteria and soon stops nitrification. However, when combined, as in pyruvic oxime, it is nitrified by soil enriched with nitrifiers, but only after a lag. This is due to the development of hetero- trophs, species of Archromohacter and Corynebacteria, which convert the NHgOH into NOT without the intermediate formation of NH3 [23]. The percolation technique is also valuable for determining the effects of various substances on nitrification under conditions simulating those found in

38 NITROGEN METABOLISM

nature. Most natural organic compounds, such as amino- acids, are not inhibitory, methionine being a notable excep- tion. Although potassium chlorate in low concentrations (e.g. io~^ M.) prevents the proliferation of Nitrohacter, it did not affect either the growth of Nitrosomonas or the oxidation of NO 2^ by an established culture of Nitrohacter. Chloromycetin, an antibiotic containing a nitro group, is especially active against organisms oxidizing NO 2" [22].

The rate of nitrification was found to be a function of the amount of NH 4 adsorbed by the base exchange complexes in the soil and could be increased by increasing the soil's base exchange capacity. No such effect was produced by the addition of materials (e.g. sand) whose only effect was to increase the available surface area. Moreover, the presence of ions such as Ca++, which can displace NHt, depressed the rate of nitrification. Few bacteria were found in the per- colating fluid itself, and Lees and Quastel deduced that the nitrifying organisms grow on the surface of soil particles [cf. 16] around receptor areas which combine with or adsorb NHt- All these areas are occupied in a 'saturated soil' and further growth of the bacteria is restricted to replacing dead cells [16]. Quastel and Scholefield have developed a tech- nique whereby the Warburg manometer can be used in studies of soil metabolism and they showed that following the addition of NHt or NO 7 to soil saturated with nitri- fying bacteria, the Og uptake was equal to that required for complete oxidation to NO 7 [22].

Nothing is known about the mechanisms which enable the organisms to utilize the energy made available by the oxidation of NH3 and NO^. Ammonia is oxidized even when the cells are unable to grow, e.g. in the absence of CO 2 [21c], and the suggestion has been made that a metal- activated enzyme may be involved in this process [15]. With regard to energy relationships and the efficiency of energy utilization, the best data are those of Baas-Beck- ing and Parks, who calculated the changes in free energy (AF) taking place at 25° C. in conditions shown by Meyer-

DENITRIFICATION 39

hof [4] to be Optimal for nitrification (NH 4" =0-005 ^-J H"^ = io-^ M.; N07=3-o3 m.).

NHt + ii02=N07+H20 + 2H+, AF=-66-5 kg. cal. N02+i02=N03, AF=- 17-5 kg. cal.

The amount of carbon assimilated as the result of the oxida- tion of a known amount of NH^ or NO 2^ is determined experimentally, and assuming that the reduction of i gram mole of CO 2 to CHgO (the generalized formula for cell sub- stance) requires 118 kg. cal. of free energy, it can be calcu- lated that the energy released in the oxidations is used with an efficiency of S'9% by Nitrosomonas and 7*8% by Nitro- bacter [4]. These values are only approximate, and in the case of Nitrosomonas it is known that the efficiency falls as the culture grows older [lo].

Reduction of nitrate: denitrification

In contrast to the limited number of organisms capable of oxidizing ammonia and nitrite, several species accom- plish the reverse process, namely, the reduction of nitrate and nitrite to ammonia, nitrous oxide (N2O) and molecular nitrogen. Whilst ammonia may be retained in the organism or in the medium, gaseous products such as Ng and NgO pass readily into the atmosphere with the result that the overall nitrogen content of the organism's immediate environment is decreased. In the latter instance the bio- logical reduction of NO 7 and NOT is often termed de- nitrification. Examples of organisms known to reduce NO 7 include Ps. fluorescens, Ps. denitrificans, Ps. stutzeri^ Mia-ococcus denitrificans, various spore-forming bacilli, Thiobacillus denitrificans, N. crassa, Hafisenula anonialay Azotobacter agilis, Esch. coli and CI. welchii. The latter five species appear to be incapable of taking the reduction as far as NgO or Ng .

With some organisms, NO"^ and NO^ serve as sources

of nitrogen for the synthesis of organic nitrogen compounds:

alternatively, or in addition, they may function as H-

acceptors in reactions concerned with the organism's energy

4

40 NITROGEN METABOLISM

metabolism. Since it is generally believed that only inorganic nitrogen in the forms of NH3 can be incorporated into organic molecules, it is probable that in both cases the metabolic pathways have at least the initial steps in common. The end-result is different in that one leads to the assimila- tion of nitrogen, whilst in the other the products are excreted into the medium. The ability to reduce NO7 enables certain organisms to grow anaerobically in media which would otherwise only support their growth in the presence of O, , and in such cases NO 7 may be regarded as replacing oxygen as the ultimate acceptor of metabolic hydrogen [25]. For example, Esch. colt cannot grow anaerobically on lactic acid as the sole source of carbon unless the medium also contains a suitable H-acceptor, and nitrate is only one of several sub- stances which can fulfil this function. Serratia marcescens and Pr. vulgaris behave similarly [2, 4], but other organisms are known which specifically use NO 7 and are unable to grow anaerobically in its absence, even though the medium contains NH3 . In anaerobic conditions, the chemosynthetic autotroph, Thio. denitrificans, can obtain energy only by the oxidation of sulphur compounds at the expense of reducing NOI:

6KN03+5S + 2H20=K2S04+4KHS04+3N2+energy

A number of aerobic spore-forming bacilli related to Bacillus subtilis can live anaerobically only in the presence of NO 7, and they have been isolated from anaerobic enrich- ment cultures in media containing a high concentration of KNO3 (8-10 per cent) [5, 30].

The first step in the reduction of NO 7 involves its con- version to NO 7 by an enzyme system which is adaptive in nature, and is known as nitratase [29]. The nitratase of Neurospora has a prosthetic group of adenine flavin di- nucleotide and will use reduced TPN to reduce nitrate to nitrite [zib]. Some strains oi Esch. coli are unable to reduce NO 7 any further [24], but others reduce both NO 7 and NO^ quantitatively to NH3 in the presence of a suitable H-donor such as glucose [2, 36]. Organisms able to reduce

DENITRIFICATION 4I

NO 7 and which also possess hydrogenase can use molecular hydrogen as a H-donor in these reductions [28]. Woods [36] thus demonstrated that washed suspensions of CI. zvelchii reduced NOJ, N07 and NH2OH to NH3 , the H2 uptake being in accordance with the following equations:

NH^OH+H. = NH3+H.O HNO,+3H2 = NH3+2H.,0 HNO3+4H, = NH3+3H2O

The Hg uptake in the third equation is that expected on theoretical grounds if NH2OH and NOT are in fact inter- mediates in the reduction of NO 7- During the early stages of the reduction of NO^", a transient accumulation of NO^ was observed.

The recent investigations of Verhoeven [30] have done much to confirm and extend the observations and hypotheses of earlier workers in this field [13]. He found that the reduc- tion of NO "3 by the aerobic spore-forming bacilli resulted in the production of NO 7, Ng and sometimes NH3 ; NO 7 could be replaced as the H-acceptor by NOT or by NgO. Strains producing large amounts of ammonia did not form appreciable amounts of gaseous end-products, and the con- verse was also true. A detailed study with one strain demon- strated that the reduction process took place in two stages. During the first stage NO 7 was converted to NOT and then to NgO; whilst in the second, the gas evolved was mainly N2 , indicating that NgO is the precursor of Ng . On two occasions, Verhoeven detected NHgOH in denitrifying cul- tures, and thus provided some evidence in support of the contention of Blom that this compound is an inter- mediate in the reduction of NO J. Working with Ps. stutzeri^ Allen and van Niel have come to the conclusion that, at least in this organism, although NgO was reduced, it was not a natural intermediate. They believe that nitramide (NO2.NH2) is a possible intermediate since Ng was formed from a preparation of this compound but not from hypo- nitrous acid [i]. In the presence of a H-donor, cell-free ex- tracts of Ps. stutzeri and B. subtilis reduce nitrate to Ng {Ps.

42

NITROGEN METABOLISM

stutzeri) or appreciable quantities of NgO {B. subtilis). Ex- tracts of both organisms convert nitrite to nitric oxide and nitric oxide to Ng [aifl]. Further advances in the elucidation of the pathways of the biological reduction of nitrate await the collection of more information concerning the chemical properties of nitroxyl (HNO), hyponitrous acid (HgNgOg), nitramide and other similar compounds of nitrogen, and the development of unequivocal methods for their identifi- cation and estimation. The scheme given below is based on those proposed by Kluyver [cf. 30] and by Allen and van Niel: it will be realized that there is no direct evidence that nitroxyl or hyponitrous acid or nitramide is a natural intermediate.

(R.NHo)

(NO2.NH2) t N2O

HNOa

HNO,

N,

>NH3

(HNO)

(H2N2O2)

-^N^O

Hypothetical Pathways of Nitrate Reduction in Micro-organisms

Routes 'a' and *6' may be operative in Ps. denitrificans, and 'c' in the aerobic spore-forming bacilli [30]. Allen and van Niel postulate that in Ps. stutzeri^ NO"^ enters into organic combination and is reduced to an amino compound (R.NH2) which then reacts with another substance, perhaps NO"^ itself, to yield nitramide (route 'd').

The presence of oxygen tends to suppress the reduction of no's, the degree of inhibition being determined both by the partial pressure of Og and the organism concerned, e.g. Esch. coli [29] is more sensitive than Ps. denitrificans [26]. It is interesting to note that restricted aeration of cultures of the denitrifying bacilli resulted in the production of large amounts of NH3 , even by those strains which in anaerobic

DENITRIFICATION 43

conditions produced large amounts of Ng and NgO. How- ever, no such effect was obtained with Pseudomonas organ- isms; they all produced Ng and no NHg . If cultures of the Bacillus and Pseudomonas organisms were sufficiently aerated there was no reduction of NO 7-

The ability to reduce nitrate has proved to be of value in the classification and identification of micro-organisms, e.g. in the yeasts, the genus Hansenula is differentiated from Pichia on the basis that only the former can grow on NO 7 as a source of N. The gaseous products or nitrite formed by the activities of nitrate-reducing organisms are responsible for troublesome and unwelcomed consequences in certain industries. Thus N07 and N07 are commonly used for curing and preserving meat products and spoliation is often due to denitrifying bacteria. Some workers believe that denitrifying micro-organisms convert an appreciable amount of fertilizers such as (NH4)2S04 and KNO3 into gaseous products and thus significantly decrease the amount of nitrogen available for plant growth [cf. 32]. After feeding on oat hay, which frequently contains large amounts of NOs"* sheep may show signs of methaemoglobinaemia (oat hay poisoning) and this condition is due to the absorption of NO 2" formed by micro-organisms in the rumen [17]. If acidic conditions arise during the commercial production of alcohol by the fermentation of sugar molasses, large amounts of nitrogen peroxide are sometimes evolved. This is due to the acidic decomposition of nitrites which were formed from nitrates by micro-organisms during the processing of the molasses [32].

REFERENCES

1. Allen, M. B. and van Niel, C. B. (1952), J. Bact., 64, 397 •

2. Aubel, E. (1938), C.R. Soc. Biol., Paris, 128, 45

3. Audus, L. J. (1946), Nature, 158, 419

4. Baas-Becking, L. G. M. and Parks, G. S. (1927), Physiol. Rev.

7,85

5. Beijerinck, M. W. and Minkman, D. C. J. (1910), Centr.

Bakt. Parasiteiik, (2. Abt.), 25, 30

6. Bomeke, H. (1939), Arch. MikrobioL, 10, 385

44 NITROGEN METABOLISM

7. Chick, H. (1906), Proc. Roy. Soc, 77B, 241

8. Frankland, P. F. and Frankland, G. (1890), Phil. Trans. Roy.

Soc, 181B, 107

9. Fred, E. B. and Davenport, A. (1921), Soil Sci., 11, 389

10. Hofman, T. and Lees, H. (1952), Biochem. J., 52, 140

11. Jensen, H. L. (1950), Nature, 165, 974

12. Kingma-Boltjes, T. Y. (1935), Arch. MikrohioL, 6, 79

13. Korsakova, M. P. (1927), Bull. Acad. Sci. U.S.S.R., Ser. 6,

1221

14. Lees, H. (195 1), Nature, 167, 355 / 15. (1^52), Biochem. y., 52y 134

J ~ 16. and Quastel, J. H. (1946), Biochem. J., 40, 803, 815, 824

17. Lewis, D. (1951), Biochem. J., 48, 175; 49, 149

18. Meek, C. S. and Lipman, C. B. (1922), J. gen. Physiol., 5, 195

19. Meiklejohn, J. (1950), J. gen. Microbiol., 4, 185

20. Meyerhof, O. (1916), Pfliig. Arch. ges. Physiol., 164, 353; 165,

229; (1917), 166, 240 2.1a. Najjar, V. A. and Allen, M. B. (1954), J. hiol. Chem., 206, 209 2.1b. Nason, A. and Evans, H. J. (1953), ^ biol. Chem., 202, 655 21C. Nelson, D. H. (193 1), Zbl. Bakt. (2. Abt.), 83, 280 22. Quastel, J. H. and Scholefield, P. G. (195 1), Bact. Rev., 15, i , 23. Scholefield, P. G. and Stevenson, J, W. (1952), Bio- chem. J., 51, 278

24. and Stephenson, M. (1925), Biochem. J., 19, 660

^ 25. Stephenson, M. and Whetham, M. D. (1925), Biochem.

y-, 19, 304

26. Sachs, L. E. and Barker, H. A. (1949), J. Bact., 58, 11

27. Schloessing, T. and Muntz, A. (1877), C.R. Acad. Sci., Paris,

84, 301; 85, 1018

28. Stephenson, M. and Stickland, L. H. (1931), Biochem. y., 25,

205

29. Stickland, L. H. (1931), Biochem. y., 25, 1543

30. Verhoeven, W. (1952), Thesis: Delft, Aerobic sporeforming

nitrate reducing bacteria

31. Warington, R. (1891), J. chem. Soc, 59, 484

32. Wimmer, G. (1904), Z. Hyg. Infaktkr., 48, 135

33. Winogradsky, S. (1890), Ann. Inst. Past., 4, 213, 257, 760

34. (1891), Ann. hist. Past., 5, 32, 577

35. (1933), Ann. Inst. Past., 50, 350

36. Woods, D. D. (1938), Biochem. y., 32, 2000

CHAPTER IV

THE FIXATION OF NITROGEN

For several centuries it has been a matter of common observation and agricultural practice that soil impoverished by the growth of cereals can be revitalized by allowing the land to lie fallow or by growing leguminous plants, yet not until the end of the nineteenth century were adequate explanations forthcoming as to why these procedures caused such beneficial effects. They are in fact due to micro- organisms which, either themselves or when in association with leguminous plants, possess the ability to use the atmosphere as a source of nitrogen. The conversion of mole- cular nitrogen (N2) into nitrogenous compounds which can be assimilated by the organisms concerned is termed nitrogen fixation. The ability to fix Ng appears to be restricted to micro-organisms and even amongst them it occurs in but a few genera.

Priestley's claim that green plants absorbed N2 as well as CO 2 and the suggestion of Sir Humphrey Davy that their nitrogen might be derived from the atmosphere by the agency of 'mushrooms and funguses' were the cause of much controversy and stimulated several investigations designed to test these statements by experiment [41]. Boussingault in 1838 was the first to show that, when they were grown in sand, the nitrogen content of clover plants increased yet that of wheat did not. Although in the years that followed, the swellings or nodules invariably found in the root systems of leguminous plants were frequently com- mented upon, half a century was to elapse before the classical experiments of Hellriegel and Wilfarth established that they were the site of the agency which enabled legumin- ous plants to fix atmospheric Ng . They found that nodules were formed only in non-sterile environments and that in contrast to cereals the growth of leguminous plants was normal even in the absence of fixed nitrogen (i.e. nitrate,

45

46 NITROGEN METABOLISM

NHt, etc.) provided they were able to develop these struc- tures. By this time nodule formation was known to be the outcome of bacterial invasion of the root tissues [cf. 41, 35] and the organisms living in them were described by Frank under the name oi Bacterium radicicola [18]. Contrary to the ideas of previous workers, Hellriegel and Wilfarth suggested that the bacteria were not parasites but lived in symbiotic association with the plant and endowed it with the ability to grow at the expense of atmospheric Ng and thus be inde- pendent of an exogenous source of fixed nitrogen. Pure cultures of the nodular organisms, now placed in the genus Rhizobium, were first isolated by Beijerinck, who also found them free-living in the soil [3]. Like later workers, he was unable to demonstrate that these organisms fixed Ng in the absence of the host plant, and the mechanism of Na-fixation by the symbiotic system still awaits elucidation. All the strains of a given species of Rhizobium induce nodule forma- tion in a restricted number of leguminous plants, termed a cross-inoculation group, and it is on this basis that the Rhizobium are classified into species, each species being specific for one cross-inoculation group [41, i]. Although nodules may be formed, they are not always effective, i.e. capable of fixing nitrogen.

Isolation of free-living N ^-fixing organisms

During the latter half of the nineteenth century, Jodin and Berthelot provided evidence that certain free-Hving micro-organisms fixed atmospheric Ng , and pure cultures of bacteria exhibiting this property were eventually isolated by Winogradsky and Beijerinck. Each of these eminent bacteriologists used the enrichment culture technique with media which, apart from inorganic salts, contained only a substance such as glucose or mannitol as a source of carbon and energy: no nitrogenous compound was added. After being inoculated with soil the cultures were incubated in an atmosphere of air or nitrogen. Winogradsky thus isolated the anaerobe Clostridium pasteurianum which fermented glucose to acetic and butyric acids together with Hg and CO 2 [49]. A few years later Beijerinck, using media

FIXATION OF NITROGEN 47

containing mannitol, isolated the two aerobes Azotohacter chroococcutn and Azotohacter agilis (extremely motile) [4]. Apart from Az. indiciwi, Azotohacter spp. in general do not fix Ng in an acidic environment, consequently their isolation is facilitated by the use of a neutral or slightly alkaline medium, e.g. one containing a buffer or CaCOg. Further- more, the incorporation of a small amount of sodium molybdate (5x10"® per cent) is frequently advantageous since molybdenum appears to be of especial significance in organisms which fix Ng .

Organic compounds, other than those which serve as sources of carbon and energy, retard the growth of Azoto- hacter but not Rhizohium. It is, however, unlikely that the concentration of organic matter in the soil is ever great enough to affect the fixation of Ng by Azotohacter in its natural habitat. Of the thirty-five compounds tested only aspartic acid, asparagine, glutamic acid, urea and adenine replaced molecular Ng as a source of N for Azotohacter [20]. On the other hand, the growth of Rhizohium spp. was sup- ported by any one of thirty-two organic nitrogen compounds and was luxuriant in rich organic media [30]. The optimal growth of fast-growing, but not slow-growing, strains of Wiizohium is dependent on an exogenous supply of a sub- stance originally named coenzyme R and later identified as biotin [cf. 40].

Detection of N ^-fixation

Although from time to time the power to fix N 2 has been attributed to many other organisms, the majority of these claims must now be discounted on the grounds of faulty experimentation. It is very difficult, if not impossible, to eliminate all nitrogenous compounds from the materials used to make media, consequently growth in what is believed to be, apart from molecular Ng , a N-free medium is not a sufficient criterion for stating that an organism can fix nitrogen and, indeed, there is often no mention of control cultures incubated in the absence of molecular Ng . It is therefore relevant to consider the techniques used in the detection and quantitative study of Ng-fixation. Apart from

48 NITROGEN METABOLISM

growth on a N-free media, these techniques involve gaso- metric analysis, nitrogen estimations by Kjeldahl proce- dures, manometry or the use of the nitrogen isotope N^^.

Gasometric analysis is used to detect whether there has been a decrease in the amount of gaseous Ng in a closed system, whilst increases in fixed nitrogen can be determined by the use of a Kjeldahl technique. Although often used successfully [22], gasometric analysis is tedious and the elimination of experimental errors is not easy. The Kjeldahl method, which at first sight appears to be an ideal and technically simple procedure, has yielded many erroneous results, the reasons for which have been cogently assessed by Wilson [41]. The samples taken for analysis do not have a homogenous composition, and whilst a particular Kjeldahl procedure will estimate with precision the nitrogen content of one or more related compounds, it does not necessarily follow that the nitrogen in other compounds is estimated with a comparable degree of accuracy; e.g. the mere addition of water to dry seeds appeared to increase their nitrogen content as measured by the Kjeldahl procedure being used. None the less, provided such sources of error are borne in mind, the Kjeldahl technique has been, and is, of great value.

A more convenient, though indirect, method for study- ing Ng-fixation by free-living organisms is based on the use of the Warburg manometer, organisms being grown in the flask of the apparatus and the gas changes consequent on growth being followed in the usual manner. An obvious com- plication with aerobic organisms is that superimposed on any uptake of Ng , there is a concomitant uptake of Og due to respiration. Indeed, Azotobacter and Rhizohium possess the highest respiration rates known, the Qq^ on glucose being of the order 1,000 to 2,000. Where large O2 uptakes are expected, the usual Brodie manometric fluid is often re- placed by one of greater density [e.g. Hg] in order that the manometer readings will be of convenient dimensions. If the rate of Ng-fixation is the factor limiting growth, then fixation by aerobic organisms can be followed in terms of an increase in the respiration rate, since the latter is directly

FIXATION OF NITROGEN 49

proportional to the mass of cytoplasm present in the system. With Azotohacter, the Ng uptake is insignificant compared with the O2 uptake, and is therefore neglected in calculating the results. Direct proof that such assumptions are justified was provided by Burk and Meyerhof, the originators of this technique [29, 7]. In their experimental conditions, irrespective of whether growth was followed by nephelometry, dry-weight measurements or total cell counts, the grov^h curves were all reasonably superimposable on the graph relating respiration rate with age of culture. Manometry has since been applied to the study of Na-fixa- tion by anaerobic bacteria, such as Clostridia, Desulpho- vibrio and photosynthetic organisms [34].

Unequivocal evidence as to whether a system can fix N2 is provided by the use of N2 enriched with the non- radioactive isotope N^^, a technique first introduced by Burris [9] and which later proved to be of value in the elucidation of the route of N2-fixation. The requisite control experiments showed that non-enzymic exchange reactions between molecular N2^^ and compounds containing nitrogen were insignificant, and that in vivo there was no preferential uptake of one isotope over another, i.e. the ratio of the isotopes one to another in the gas-phase did not change throughout the course of the experiment. To detect Ng- fixation, the experimental material is placed in a closed system through which N2 enriched with N^^ is circulated by a Urey pump. The nitrogen-containing compounds of the experimental material may be fractionated prior to being converted into N2 for analysis in the mass spectrograph. The technique is very sensitive and an increase of o-oi atom per cent N^^ above the normal distribution of 0-36 is regarded as being significant, and in certain conditions it is possible to detect the fixation of o-ooi //g. N [9, 10].

Organisms fixing N2

In addition to CI. pasteurianuni and Azotobacter spp., there is now adequate evidence, in many cases based on experi- ments with N^^, that Ng-fixation is a property of several other species of Clostridia [32], various photosynthetic

50 NITROGEN METABOLISM

bacteria (species of Rhodospirillum, RhodopseudomonaSy Rhodomicrobium, Chromatium and Chlorohacterium) [19, 26 27, 28], Desulphovibrio [34], blue-green algae of the family Nostocaceae [15, 17, 43], and Calothrix [43]. Claims for fixation by yeasts and other fungi are as yet unsubstantiated. The successful demonstration that nodules detached from roots fix N2 was dependent on the use of nodules from leguminous plants grown in the field rather than the green- house, performing the experiments immediately after detachment from the roots and subjecting only the soluble nitrogen compounds of the nodules to isotopic analysis [2]. The Leguminoseae is not the only family in which nodules are formed in consequence of microbial invasion, and the root nodules of the Alder {Alniis) and Coriaria japonica may also be concerned with the fixation of Ng.

Factors affecting Nz-fixation

There are no substantiated experiments in which Ng- fixation has been divorced from growth of the experimental material [24], consequently care is required in assessing whether the factor being studied directly affects the fixation mechanism: the observed effects may be no more than the result of interference with processes, such as the production of energy, which are essential to metabolism in general. For example, the influence of the partial pressure of oxygen (pOa) on fixation by Azotohacter and by clover plants appears to be entirely explicable in terms of its effects on respiration, and in consequence, the availability of energy [7, 41]. Similarly, although molybdenum, iron, calcium and strontium have all been implicated in the fixation mechanism, it is difficult to decide whether this is their primary function. These difficulties cannot be circum- vented by performing experiments in the presence of fixed nitrogen, since the latter induces a quicker rate of growth, and in all probability growth is then limited by different factors from those operative when the organisms are fixing molecular Ng . Whilst molybdenum undoubtedly influences the growth of Azotohacter^ Nostoc, CI. pasteurianum and leguminous plants, there is no direct evidence that it is

FIXATION OF NITROGEN 5I

specifically concerned with the fixation mechanism (for discussion and references concerning trace element nutri- tion, see 8, 44, 21, 6).

Burk was the first to demonstrate that the rate of Ng- fixation by Azotobacter was a function of the partial pressure of Ng (pN2)» the relationship between the two conforming to that expressed by the Michaelis-Menten equation for an enzymic reaction. The pNg at which fixation occurred at half the maximal rate (i.e. K^) was 0-21 atm., an unex- pectedly high value for a gaseous substrate [7]. In order to prepare gas mixtures containing different pNg , Burk had used Hg as a diluent gas since at that time Hg was thought to be inert in biological systems. Later experiments, first with red clover plants [41, 47] and subsequently with Azotobacter [50], revealed that Hg inhibited Ng-fixation and increasing the pHg caused a progressive decrease in the rate of growth: Hg had no effect on growth in the presence of fixed nitrogen, e.g. (NH4)2HP04 . Helium and argon exhibited no such inhibitory action, and in the absence of Hg , fixation by Az. vinelandii was maximal at a pNg of 0-15 atm. and half maximal at 0-02 atm. (Fig. 4.1). Hence, Wilson suggested that in natural conditions the pNg is not a factor limiting fixation. By comparing the rate of grov^1:h in air with that in gas phases containing, in addition to Og , either 0-15 atm. Ng or 0-3 atm. Ng , together with different amounts of Hg , Wyss and Wilson deduced that Hg inhibited fixation in a competitive manner. Unlike Azotobacter, Hg is a normal product in fermentations by CI. pasteurianum, and although the pHg did not influence the rate of Ng-fixation by this organism it did affect the amount of Ng fixed and the efficiency of the fixation process (i.e. mg. N fixed/g. glucose fermented) [32]. Wilson has noted that all the free- living organisms shown to fix Ng also possess the enzyme hydrogenase [cf. 42]. He suggests that the latter may be involved in the fixation mechanism, since the hydrogenase activity of Azotobacter is greatly reduced, even in the presence of Hg , w^hen growth is no longer dependent on the fixation of molecular Ng [25]. Hydrogenase has not been detected in Rhizobium either free-living or symbiotic, or in

52 NITROGEN METABOLISM

nodular tissues [45]. Nitrogen-fixation by Azotobacter, Nostoc and leguminous plants is inhibited in an apparently non-competitive manner by carbon monoxide in concentra- tions which cause little, if any, inhibition of respiration [48, 16]. Carbon monoxide is an isostere of Ng , and it may therefore replace Ng on the surface of the enzyme concerned

100-	^	r^	X.		'0	>^-^	-0
?80- <	/		y	/^
2	J	^/	/^
Ui 1- < ^ 60- 0		/
^	J
^	^ /				1
|40- a.	I
FIXATION 0 1	/
	/ 0-7 / 1	0-6 1		o-s		0-4 pH2 or Inert gaJ	_!
0-	/	o.f.		0)3		0.4 P^zC'" at'"-)	_J

FIG. 4.1. — Nitrogen fixation by Azotobacter as a function of the partial pressure of nitrogen (pNo) in the presence of different partial pressures of hydrogen (6), helium ( X ), or argon (A), or a partial vacuum (D)"- partial pressure of oxygen in all experiments, 0*2 atm. [50]

with the primary fixation reaction. Alternatively the fixation mechanism may involve a metal ion, perhaps combined in the prosthetic group of an enzyme, whose activity is lost on combination with CO. In this connection it may be of significance that the hydrogenase of Az. vinelandii is also inhibited by CO [44].

Role of ammonia in N^-jixatioyi

The idea has arisen that there is a key substance in

FIXATION OF NITROGEN 53

nitrogen-fixation, one which can be regarded both as the product of the fixation mechanism and as the substrate for the reactions by which inorganic nitrogen is incorporated into organic compounds. During the last decade, the eluci- dation of the nature of this substance has occupied the attention of two groups of workers, one associated with Virtanen in Finland [36] and the other with Burris and Wilson in the U.S.A. By using the isotope N^^, the American workers have obtained convincing evidence of the key role of NH3 in the fixation of N2 , a concept first advanced by Winogradsky [49]. In the isotope experiments, the period of exposure to the substance enriched with N^^ was not long enough for equilibrium to be established, and prior to isotopic analysis, the nitrogenous components of the experi- mental system were separated into fractions to facilitate location of the compounds with the highest concentration of N^^. For example, after acid hydrolysis, various amino- acids were isolated by classical precipitation procedures, or more recently, by the use of columns of ion exchange resins or starch [51, 43]. It was argued that if the assimilation of nitrogen involved a number of intermediates, then, before equilibrium was established, the concentration of the iso- tope would be the greater the nearer the intermediate to the substance initially enriched with N^^. Moreover, if a sub- stance on the fixation pathway was supplied instead of, or in addition to, molecular Ng , it should be utilized not only as a source of nitrogen, but also in preference to its pre- cursors (cf. simultaneous adaptation, p. 13) [11, 44].

The resultant distribution of the isotope amongst the constituents of Az. vinelandii was the same irrespective of whether nitrogen was derived from Ng^^ or N^^Ht, and of the cells' amino-acids, glutamic acid followed by aspartic acid, contained the highest concentration of N^^ [12]. The establishment of such a distribution, even when the period of contact with N^^Ht was short (3 min.), indicated that the enzymes responsible for the assimilation of NHg already existed in the bacteria. Moreover, the fact that NH3 gave rise to the same distribution of N^^ as N2 ^ implied that it was either itself on the pathway of Ng-fixation or at least

54 NITROGEN METABOLISM

closely related to one of the natural intermediates. The high concentration of N^^ in glutamic acid was indicative of its importance in the pathway of Na-fixation. Such a conclusion was in keeping with the results of previous work with other organisms in which it had been established that glutamic acid occupies a key position in amino-acid metabolism and could be synthesized from NHg and a-ketoglutarate by the glutamic acid dehydrogenase system (cf. Chap. II).

Winogradsky was the first to note, during his experiments with CI. pasteurianum, that ammonium salts inhibited N2- fixation. Azotobacter responds in a similar fashion, both to ammonium salts and to compounds (urea, asparagine) which it can convert to NH3 [46]. Since nitrate and nitrite were only inhibitory after a lag period, it was inferred that they were not utilized by Azotobacter until the appropriate enzymes had been formed by adaptation, and in fact there was no lag period with cells which had been grown in media containing NO7 and NO^ • Organic nitrogen compounds, e.g. aspartic and glutamic acids, were only moderately inhibitory, perhaps not unexpectedly since they are probably more concerned with intermediary metabolism rather than with the initial steps of the fixation mechanism.

Further isotope experiments provided additional evidence that the fixation of N 2 by CI. pasteurianum, Chromatium and Nostoc muscorum is accomplished by essentially the same route as in Az. vinelandii [43]. In each of these organisms after exposure to Ng-^^ or N^^H4 , the dicarboxylic amino- acids, and in particular glutamic, contained the highest concentration of N^^. Furthermore, the results of the ex- periments with CI. pasteurianum. [51] proved to be in some respects comparable with those of Virtanen with nodulated leguminous plants. Under certain conditions, the fixation of N 2 by CI. pasteurianum and by the symbiotic system was accompanied by the excretion of nitrogenous compounds into the environment; the anaerobe excreted mainly NH3 whilst the plant excreted aspartic acid, ^-alanine and a small amount of an oxime, identified as oximinosuccinic acid. These substances are regarded as being products, not of catabolism, but of the processes directly concerned with

FIXATION OF NITROGEN 55

Ng-fixation. After exposing cultures of CI. pasteurianum in the log phase of growth to Ng^^, the concentration of N^^ in the NH3 isolated from the medium was greater than that of the amide nitrogen of the cell protein which was in turn greater than the average level of the isotope in the pro- teins as a whole. This indicated that the excreted NH3 was a product of the fixation processes and did not arise by deamination of amino-acids [51]. The excretion of NH3 only occurs when the organisms are grown in certain media, and it is apparently an expression of a deficiency, probably of suitable organic acceptors, since supplementing the medium with biotin, ^-aminobenzoic acid and a-keto- glutarate completely stopped the excretion of NH3 and yet had little effect on the rate of N2-flxation [43].

Role of hydroxylamine in N2-fixation

Because leguminous plants apparently excreted only aspartic acid and substances related to it [38, 5], Virtanen originally concluded, especially in view of the presence of the oximino-compound, that hydroxylamine (NHgOH) was the key product of the fixation mechanism. He proposed that NH2OH condensed with oxaloacetate, a substance known to be present in nodulated roots, thus forming oximinosuccinic acid which was then reduced to aspartic acid. Many other workers were unable to repeat these experi- ments, and following a successful visit to Virtanen's labora- tory, Wilson concluded that excretion was observed only when the rate of photosynthesis was not sufficient to supply enough materials for the utilization of all the products of the fixation mechanism [41]. The later discovery [39] that glutamic acid was present amongst the excreted substances implied that many of the results of the leguminous plant experiments could also be interpreted in favour of the importance of NH3 in the fixation mechanism. Whilst it seems most probable that nitrogen enters into organic com- bination in the form of NH3 and not NHgOH, the possi- bility that NHgOH is a precursor of the NH3 , and perhaps, in certain circumstances, reacts directly with an organic acceptor, has not yet been excluded [cf. 37]. 5

56 NITROGEN METABOLISM

A series of well-controlled experiments has failed to find any evidence for the participation of NHgOH in the fixation of N 2 by Azotohacter and CI. pasteurianum [31, 33]. Hydrox- ylamine in concentrations greater than about 2 /<g./ml. was toxic, and growth, when it occurred, could be accounted for in terms of NH3 produced by the spontaneous decomposi- tion of NHgOH. Even with Ni^HgOH the results were diffi- cult to interpret because the amount of nitrogen involved was too small to permit significant determinations. The oximes of a-ketoglutarate, oxaloacetate and pyruvate were slowly utilized by CI. pasteurianum but not by Azoto- hacter [34].

Pathways of N^-fixation

There has been little support in recent years for the fixation of nitrogen being explained in terms of an oxidative pathway involving compounds such as nitrous oxide and hyponitrous acid. There is now a considerable amount of evidence that it is probably a process involving a number of reductive mechanisms which terminate in the formation of NH3 . Whilst the steps between molecular Ng and NH3 are completely unknown some hypothetical pathways are shown in Fig. 4.2 (details of other schemes will be found in reference 44). Although the fixation of Ng by cell-free systems has yet to be confirmed, past failures [24] may be attributed, at least in part, to the harsh procedures employed in the preparation of the extracts.

In spite of their diverse nature, bacteria, blue-green algae and leguminous plants all apparently fix nitrogen by mechanisms which, if not identical, have at least many features in common [43]. They all respond in a similar manner to changes in the partial pressure of Ng, Hg and CO and their K^ values for Ng are of the same order. In view of the repeated failures to demonstrate Ng-fixation by free-living cultures of Rhizobium, interest with regard to the symbiotic system is centred around the question why association with the leguminous plants should endow one or other of the symbionts with this ability. A noticeable feature of the association is the presence of a red haemo-

FIXATION OF NITROGEN

57

globin-like pigment ('leghaemoglobin') in the effective nodules [44]. The pigment is only formed after the nodules have been established and concurrent with the change to a green colour, their ability to fix N2 declines. The pOg in the nodules is relatively low and the pigment may serve to store and transport Og to the Rhizobia, which, it will be remem-

•NEN

+ H

nitrous oxide

+ H2P

nitrogen

+ HP / (NOH), HO. N.N. OH hyponitrous acid /H H

nitrogen

+ H

NH^OH_ hydroxylomine

+ COCOOH

I

CH^COOH oxaloacetic acid

hooc.ch2c(noh) cooh

oximinosuccinic acid

H00C.CH2CH(NH2.)C00H ospgrtic acid

(CH2)2C00H +

COCOOH oC-ketoglutaric ocid

(CHp)2COOH

ch(nh2)cooh

glutamic acid

FIG. 4.2. — ^Hypothetical pathways for the fixation of molecular nitrogen

bered, possess a high rate of respiration. Some workers have suggested that the haemoglobin participates directly in the ifixation mechanism. For example, NHgOH is decomposed by haemoglobin (Hb) in vitro, thus [14]:

NH20H+2Hb+++H20 -> 2MetHb++++NH3+20H- 2NH20H+2MetHb+++ -> 2Hb+++2H20+N2+2H+ and it is possible that the fixation of nitrogen is accom- plished by the reverse of these reactions. However, claims for the natural occurrence of methaemoglobin in the nodules

58 NITROGEN METABOLISM

have been seriously challenged [23]. The addition of haemo- globin increases the respiration of i^feo^mw, but this appears to be an indirect effect [13] and not connected with the transport of O2 .

It will be evident that there are still many aspects of nitrogen fixation to be explored, and apart from their bio- chemical interest, their economic importance should not be underestimated. The growth of plants, the leaching effects of rain-water and the activities of denitrifying organisms all tend to remove from the soil the nitrogenous compounds which are essential to the continued existence of most forms of plant life. This loss is in part restored naturally by the decomposition of plants and animals, and artificially by the application of inorganic or organic fertilizers. Fixed nitrogen compounds produced commercially probably account for no more than 15% of the nitrogen returned annually to the soil. By far the greatest proportion is due to Na-fixation by bio- logical agents, and it has been estimated that symbiotic systems and free-living organisms are responsible respectively for returning to the soil 5-46x106 and 4-37x10^ tons of nitrogen per year, yet, even allowing for this, there appears to be an annual overall loss in soil nitrogen [42].

REFERENCES

1. Allen, E. K. and Allen, O. N. (1950), Bad. Rev., 14, 273

2. Aprison, M. H. and Burris, R. H. (1952), Science, 115, 264

3. Beijerinck, M. (1888), Bot. Zbl, 39, 356

4. (1901), Zbl. Bakt., II, 7, 561

5. Billen, D. and Lichstein, H. C. (1949), J. Bad., 57, 267

6. Bortels, H. (1940), Arch. Mikrobiol., 11, 155

7. Burk, D. (1934), Ergeb. Enzymforsch., 3, 23

8. and Burris, R. H. (1941), Ann. Rev. Biochem., 10, 587

9. Burris, R. H. (1942), J. biol. Chem., 143, 509

10. Eppling, F. J., Wahlin, H. B. and Wilson, P. W. (1943),

J. biol Chem., 148, 349

II. and Wilson, P. W. (1946), J. Bad., 52, 505

12. (1946), J. biol. Chem., 165, 595

13. • (1952), Biochem. J., 51, 90

14. Colter, J. S. and Quastel, J. H. (1950), Arch. Biochem., 27, 368

15. De, P. K. (1939), Proc. Ro\. Soc. Lond., 127B, 121

16. Ebersole, E. R., Guttentag,"C. and Wilson, P. W. (1944), Arch.

Biochem., 3, 399

FIXATION OF NITROGEN

59

7. Fogg, G. E. (1947), Endeavour, 6, 172 ^. Frank, B. (1879), Bot. Ztg., 37, 377, 393 ). Gest, H., Kamen, M. D. and Bregoff, H. M. (1950), J. biol.

Chem., 182, 153 3. Horner, C. K. and Allison, F. E. (1944), ^ Bad., 47, i

[. Burk, D., Allison, F. E. and Sherman, M. S. (1942),

J. agric. Res., 65, 173 I. Hurwitz, C. and Wilson, P. W. (1940), Ind. Eng. Chem. Anal.,

ed. 12, 31 ;. Keilin, D. and Smith, J. D. (1947), Nature, 159, 692 .. Lee, S. B., Burris, R. H. and Wilson, P. W. (1942), Proc. Soc. exp. Biol. N.Y., 50, 96

. and Wilson, P. W. (1943), J. biol. Chem., 151, 377

.. Lindstrom, E. S., Burris, R. H. and Wilson, P. W. (1949),^. Bact., 58, 313

. Lewis, S. M. and Pinsky, M. J. (1951), J. Bact., 61, 481

. Tove, S. R. and Wilson, P. W. (1950), Science, 112, 197

. Meyerhof, O. and Burk, D. (1928), Z. phys. Chem., 139A, 117

. Nielsen, N. (1940), C.R. Lab. Carlsberg, 23, 115

. Novak, R. and Wilson, P. W. (1948), Jf. Bact., 55, 517

. Rosenblum, E. D. and Wilson, P. W. (1949), J. Bact., 57, 413;

(1950), 59, 83; (1951), 61, 475 . Segal, W. and Wilson, P. W. (1949), jf- Bact., 57, 55 . Sisler, F. D. and Zobell, C. E. (195 1), Science, 113, 511 . Thornton, H. G. and Nicol, H. (1936), Nature, 137, 494 . Virtanen, A. L (1947), Biol. Rev., 22, 239

. and Hakala, M. (1949), Acta chem. scand., 3, 1044

. and Laine, T. (1939), Biochem. J., 33, 412

. Linkola, H., Hakala, M. and Rautanen, N. (1946),

Suomen Kemistilehti, 19B, 83 . Wilson, J. B. and Wilson, P. W. (1942), J. Bact., 43, 329 . Wilson, P. W. (1940), The Biochemistry of Symbiotic Nitrogen Fixation, Wisconsin U.P., Madison, U.S.A.

. (195 1), in Bacterial Physiology, Chap. 14

. (1952), Advances in Enzymology, 13, 345

. and Burris, R. H. (1947), Bact. Rev., ii, 41

. Burris, R. H. and Coffee, W. B. (1943), J. biol. Chem.,

147. 475

. Hull, J. F. and Burris, R. H. (1943), Proc. Nat. Acad.

Sci., Wash., 29, 289

Lee, S. B. and Wyss, O. (1941), jf. biol. Chem., 139, 91

and Lind, C. J. (1943), jf. Bact., 45, 219

Winogradsky, S. (1893), C.R. Acad. Sci., Paris, 116, 1385;

(1894), 118, 353 Wyss, O., Lind, C. J., Wilson, J. B. and Wilson, P. W. (1941),

Biochem. jf., 35, 845 Zelitch, L, Rosenblum, E. D., Burris, R. H. and Wilson, P. W. ..•dil>: (1951), jf- biol. Chem., 191, 295; (195 1), J- Bact., 62, 747

I smbioj reptctive i:*toiis(

P W.I IV

,!r«.r.

S8

NITROGEN METABOLISM

have been seriously challenged [23]. The addition of ha ho- globin increases the respiration oiRhizobium, but this ap :ars to be an indirect effect [13] and not connected witjthe transport of Og .

It will be evident that there are still many aspe( of nitrogen fixation to be explored, and apart from theii io- chemical interest, their economic importance should ri be underestimated. The growth of plants, the leaching effe 5 of rain-water and the activities of denitrifying organisr all tend to remove from the soil the nitrogenous compc nds which are essential to the continued existence of most i ^ms of plant life. This loss is in part restored naturally b the decomposition of plants and animals, and artificially b the application of inorganic or organic fertilizers. Fixed nit gen compounds produced commercially probably account i \ no more than 15% of the nitrogen returned annually to th oil. By far the greatest proportion is due to Ng-fixation b^ bio- logical agents, and it has been estimated that sym t)tic systems and free-living organisms are responsible respec *»ely for returning to the soil 5-46x106 and 4-37x106 to i of nitrogen per year, yet, even allowing for this, there ap iars to be an annual overall loss in soil nitrogen [42].

REFERENCES

1. Allen, E. K. and Allen, O. N. (1950), Bad. Rev., 14, 2',

2. Aprison, M. H. and Burris, R. H. (1952), Science, 115,

3. Beijerinck, M. (1888), Bot. Zbl., 39, 356

4. (1901), Zbl. Bakt., II, 7, 561

5. Billen, D. and Lichstein, H. C. (1949),^. Bad., 57, 26*

6. Bortels, H. (1940), Arch. MikrobioL, 11, 155

7. Burk, D. (1934), Ergeb. Enzymforsch., 3, 23

8. and Burris, R. H. (1941), Ann. Rev. Biochem., 10,

9. Burris, R. H. (1942), J. biol. Chern., 143, 509

II. 12.

13. 14.

15- 16.

Eppling, F. J., Wahlin, H. B. and Wilson, P. W. ( J. biol Chem., 148, 349

and Wilson, P. W. (1946), J. Bad., 52, 505

(1946), J. biol. Chem., 165, 595

(1952), Biochem. J., 51, 90

Colter, J. S. and Quastel, J. H. (1950), Arch. Biochem., 2. De, P. K. (1939), Proc. Roy. Soc. Lond., 127B, 121 Ebersole, E. R., Guttentag, C. and Wilson, P. W. (1944), Biochem., 3, 399

7 ^3),

368

'Ch.

FIXATION OF NITROGEN 59

17. Fogg, G. E. (1947), Endeavour, 6, 172

18. Frank, B. (1879), Bot. Ztg., 37, 377, 393

19. Gest, H., Kamen, M. D. and Bregoff, H. M. (1950), jf. biol.

C/iem., 182, 153

20. Horner, C. K. and Allison, F. E. (1944),^. Bad., 47, i

21. Burk, D., Allison, F. E. and Sherman, M. S. (1942),

y. agric. Res., 65, 173

22. Hurwitz, C. and Wilson, P. W. (1940), Ind. Eng. Chem. Anal.,

ed. 12, 31

23. Keilin, D. and Smith, J. D. (i947), Nature, 159, 692

24. Lee, S. B., Burris, R. H. and Wilson, P. W. (1942), Proc. Soc.

exp. Biol. N.Y., 50, 96

25. and Wilson, P. W. (1943), J. biol. Chem., 151, 377

26. Lindstrom, E. S., Burris, R. H. and Wilson, P. W. (1949), J.

Bact., 58, 313

27. Lewis, S. M. and Pinsky, M. J. (1951), J. BacL, 61, 481

28. Tove, S. R. and Wilson, P. W. (1950), Science, 112, 197

29. Meyerhof, O. and Burk, D. (1928), Z.phys. Chem., ly^P^, J 17

30. Nielsen, N. (1940), C.R. Lab. Carlsberg, 23, 115

31. Novak, R. and Wilson, P. W. (1948), J. Bad., 55, 517

32. Rosenblum, E. D. and Wilson, P. W. (1949), J. Bad., 57,413;

(1950), 59, 83; (1951), 61, 475

33. Segal, W. and Wilson, P. W. (1949), J- Bad., 57, 55

34. Sisler, F. D. and Zobell, C. E. (195 1), Science, 113, 511

35. Thornton, H. G. and Nicol, H. (1936), Nature, 137, 494

36. Virtanen, A. L (1947), Biol. Rev., 22, 239

37. and Hakala, M. (1949), Acta chem. scand., 3, 1044

38. and Laine, T. (1939), Biochem. J., 33, 412

39- Linkola, H., Hakala, M. and Rautanen, N. (1946),

Suomen Kemistilehti, 19B, 83

40. Wilson, J. B. and Wilson, P. W. (1942), J. Bact., 43, 329

41. Wilson, P. W. (1940), The Biochemistry of Symbiotic Nitrogen

Fixation, Wisconsin U.P., Madison, U.S.A.

42. (195 1 ), in Bacterial Physiology, Chap. 14

43- (1952), Advances in Enzymology, 13, 345

44- and Burris, R. H. (1947), Bact. Rev., ii, 41

45- Burris, R. H. and Coffee, W. B. (1943), J. biol. Chem.,

147. 475

46. Hull, J. F. and Burris, R. H. (1943), Proc. Nat. Acad.

Sci., Wash., 29, 289 47- Lee, S. B. and Wyss, O. (1941),^ biol. Chem., 139, 91

48. and Lind, C. J. (1943), jf. Bact., 45, 219

49. Winogradsky, S. (1893), C.R. Acad. Sci., Paris, 116, 1385;

(1894), 118, 353

50. Wyss, O., Lind, C. J., Wilson, J. B. and Wilson, P. W. (1941),

Biochem. J., 35, 845

51. Zelitch, L, Rosenblum, E. D., Burris, R. H. and Wilson, P. W.

(1951), J- biol. Chem., 191, 295; (1951), J- Bad., 62, 747

CHAPTER V SYNTHESIS OF AMINO-ACIDS

From the vast amount of information now available it is clear that glutamic acid and aspartic acid occupy a key position in amino-acid metabolism, and, of the two, the former is the more important. Glutamic acid can be synthe- sized from NH3 and a-ketoglutaric acid by the glutamic dehydrogenase system (p. 12) and aspartic acid from NH3 and fumaric acid by aspartase (p. 23). The direct addition of an inorganic compound of nitrogen to the appropriate carbon skeleton does not appear to be a general route for the synthesis of amino-acids, and the importance of the dicarboxylic amino-acids is in part due to the fact that they contain nitrogen in a form which can be transferred to suit- able acceptors and thus utilized for the synthesis of other nitrogenous compounds, e.g. the conversion of citrulline to arginine (p. 70) and the synthesis of amino-acids by trans- amination. Moreover, it will become clear from the following paragraphs that several amino-acids may be derived from other preformed amino-acids — for example, proline and ornithine can be synthesized from glutamic acid (pp. 69, 70). Many organisms use an inorganic form of nitrogen, such as molecular N2 , NH3 and NOJ, as a source of this element, and it is now generally held that the first steps in the utilization of molecular Ng and nitrate for this purpose involves their conversion to NH3 (see Chaps. IV and III). It is therefore worthy of note that the glutamic dehydro- genase system and aspartase are mechanisms which enable inorganic nitrogen in the form of NH3 to be incorporated into an organic molecule.

Transamination and amino-acid synthesis

A transaminase catalyses the reversible transfer of the amino group of one amino-acid to the a-keto acid corre- sponding to another amino-acid:

60

SYNTHESIS OF AMINO-ACIDS 6l

RiCH(NH,)COOH+R2COCOOH ^

RiCOCOOH+R2CH(NH2)COOH (i)

Reactions of this type were first discovered in animal tissues by Braunstein and Kritzmann, who concluded that the amino groups of several amino-acids could be transferred in such a manner provided one of the participants in the system was a dicarboxylic acid, i.e. aspartic, glutamic, oxaloacetic, or a-ketoglutaric acid [4]. Cell-free preparations of what were believed to be two distinct transaminases were obtained, one catalysed reactions ii and iii, and the other reaction iv:

glutamate+ pyruvate ?^ a-ketoglutarate+ alanine (ii)

glutamate+oxaloacetate ^^ a-ketogIutarate+ aspartate (iii)

aspartate + pyruvate ?=i oxaloacetate+ alanine (iv)

Later workers isolated two enzyme systems, specific for re- actions ii and iii respectively, and Kritzmann 's second trans- aminase was shown to be an artifact and due to a mixture of these two enzymes with catalytic amounts of glutamate functioning as a carrier between the two systems [23].

Similar transaminase systems were later found in bacteria, yeasts [39^] and Neurospora [17^, Z>]. Washed cell suspensions of various bacteria (staphylococci, streptococci, pneumo- cocci, enterobacteria, Az. agilis, and Ps. pyocyanea) cata- lysed the transfer of the amino group of glutamic acid to oxaloacetic acid (reaction iii), and cell-free preparations of Strep, faecalis accomplished both reactions ii and iii, the equilibrium being in favour of the synthesis of alanine and aspartic acid respectively [34, 35]. Like the animal trans- aminases, the bacterial enzymes possess a prosthetic group of pyridoxal phosphate. The glutamic-aspartic trans- aminase of Strep, faecalis had been partially resolved and activity was restored by the addition of synthetic pyridoxal phosphate or natural codecarboxylase. The advent of paper chromatography has facilitated the detection and identifica- tion of small quantities of amino-acids and has been used to demonstrate that several amino-acids can transfer their amino groups to a-ketoglutarate. These experiments have been performed with Esch. coli, Ps.fluorescens, B. suhtilis [14]

62 NITROGEN METABOLISM

and Lb. arahinosus [36a] and also with animal tissues [26]. Unequivocal proof that such results are due to transamina- tion awaits the isolation of the appropriate enzymes, and, indeed, a glutamic-tyrosine transaminase and a glutamic- phenylalanine transaminase have been isolated from Esch. coli and shown to contain prosthetic groups of pyridoxal phosphate [14]. All known transaminases are specific for the L-isomers of the amino-acids.

Our conceptions of the mechanism operative in biological transamination are based on analogy with in mtro systems. Transamination is a true transfer process and there is no evidence at all that the amino group becomes free as NH3 at any stage in the reaction. Herbst has proposed that a type of Schiif's base is formed by condensation between the amino and keto groups of the two substrates and that this is followed by a molecular rearrangement involving the alpha hydrogen atom of the amino donor, after which the base is ruptured by hydrolysis [25]. Incubation of alanine containing deuterium in the alpha position with a-keto- glutarate and a mammalian glutamic-alanine transaminase resulted in the rapid appearance of deuterium in the water of the experimental system and is evidence in favour of such a mechanism: Ri R2 Ri R2

I I I I ■ CH(NH,)+CO ^HC— N = C ^

I " I II COOH COOH COOH COOH

Ri R2 Ri R2

I I I 1

C = N— CH ^ CO + CH(NH,)

I I I I "

COOH COOH COOH COOH

It is to be remembered that pyridoxal phosphate itself con- tains a carbonyl group and Snell, on the basis of non- enzymic chemical experiments, has made the following proposal [42]:

amino-acid A+pyridoxal phosphate ^

keto acid A+pyridoxamine phosphate pyridoxamine phosphate + keto acid B ^^

amino-acid B+pyridoxal phosphate

t

SYNTHESIS OF AMINO-ACIDS 63

Synthetic pyridoxamine phosphate activated some prepara- tions of partially resolved bacterial transaminases [48], and though first reported not to restore activity to a resolved preparation of the pig heart glutamic-aspartic transaminase, later workers showed that it was as active as pyridoxal phos- phate provided it was incubated with the apoenzyme for 30-60 minutes before adding the substrates [366]. Gunsalus and Tonzetich have recently demonstrated that prepara- tions oiEsch. coli catalysed transamination reactions between pyridoxal and glutamic acid and between pyridoxamine and a-ketoglutarate. Hence, while there is some reason to believe that the prosthetic group of the transaminases can function a£ a carrier of amino groups, direct proof that it does so has still to be obtained.

Owing to the lack of essential data, it is not yet possible to assess whether transamination is of key importance in amino-acid synthesis. The available information, admittedly extremely incomplete, indicates that the transaminases are widely distributed in micro-organisms, at least in those non- exacting to amino-acids. Whether or not there is a complete series of enzymes capable of reacting with the keto acids corresponding to all the natural amino-acids is still not known. If transamination is involved in the synthesis of an amino-acid, then it follows that the organism must be able to synthesize the appropriate keto acid. Except for pyruvate, oxaloacetate and a-ketoglutarate, virtually nothing is known about the synthesis of these compounds, and although the two former acids are known to be produced in the inter- mediary metabolism of many organisms, conclusive evidence of the ability to synthesize a-ketoglutarate is only available for Ps. fluorescens, Az. agilis and Sac. cerevisiae. Although in recent years it has been thought that glutamic acid was the only amino-acid able to transfer an amino group to a wide variety of keto acids, there is now some evidence that aspartic acid can transaminate with acids other than a-keto- glutarate, e.g. there is an aromatic amino-acid — aspartic acid transaminase in Esch. coli [39^]. Furthermore, trans- amination reactions are now known in which dicarboxylic acids do not participate, e.g. Brucella abortus appears to

64 NITROGEN METABOLISM

possess a leucine-alanine transaminase [ic]. Transamination can therefore be visualized as being of some significance in amino-acid synthesis provided the organism can either syn- thesize one or more amino-acids from ammonia by a direct route (e.g. the glutamic acid dehydrogenase system or aspartase), or obtain suitable amino group donor amino- acids from the environment.

Metabolite analogues and the elucidation of metabolic pathways Many enzyme systems are inhibited in a competitive manner by substances similar in chemical structure to their normal substrates: such substances are known as metabo- lite analogues and have been used, for example, to study the synthesis of tryptophan by Salmonella typhosa [15]. Although a reaction inhibited in this fashion may be essential for grow^th, nevertheless, growth may be possible provided the medium is fortified in some way, i.e. by the addition of substances antagonizing the effects of the inhibitor. Ideally, the substrates of the inhibited reaction act as competitive antagonists, whereas the products act in a non-competitive manner. The grow^th of freshly isolated strains of Salm. typhosa is dependent on tryptophan or indole, but non- exacting strains frequently arise. The inhibitory effects of /?-indoleacrylic acid on the growth of a non-exacting strain were completely overcome by the addition of tryptophan. Indole accumulated in the media of non-exacting strains growing in the presence of limited amounts of tryptophan and sub-lethal amounts of indoleacrylic acid, and was pre- sumed to be the substrate of the reaction blocked by the inhibitor. Serine was a powerful antagonist and high con- centrations of this amino-acid also decreased the accumula- tion of indole. Fildes therefore concluded that tryptophan was synthesized by the condensation of serine with indole and that indoleacrylic acid inhibited this reaction. The work of Snell and Schweigert had already indicated that an- thranilic acid might be an intermediate in the synthesis of tryptophan by Lb. casei and Lb. arabinosus, and it was later found that irrespective of their tryptophan requirements, strains of Salm. typhosa secreted anthranilic acid into the

SYNTHESIS OF AMINO-ACIDS 65

culture medium [40]. Since the amount of anthranilic acid produced by an exacting strain was greater than the amount of indole utilized, it was unlikely that the former was derived from the latter. The inhibitory effect of 4- or 5-methylan- thranilic acid on the growth of the non-exacting strain was reversed by anthranilic acid, indole or tryptophan [16]. These results are summarized in the following scheme, vertical arrows denoting the site of action of inhibitors:

4-methyl- indoleacrylate, methyl-

anthranilate 4-methylindole tryptophan

anthranilic ] >■ indole"! |

acid Y j *■ tryptophan '

serineJ y

'*" protein

Another example of antagonism in amino-acid metabolism comes from studies of the nutrition of Bacillus anthracis and serves to emphasize that growth may be dependent not only on the presence of certain amino-acids but also on their relative concentrations [19]. Although B. anthracis would not grow on a synthetic and complete amino-acid medium from which valine, leucine or isoleucine had been removed, growth did occur in the absence of all three of these amino- acids. When valine was present, no growth was possible unless a suitable amount of leucine had also been added, whilst growth in the presence of isoleucine was conditional on the addition of both valine and leucine. Since these amino-acids are of comparable chemical structure, Glad- stone concluded that they are synthesized by similar, if not identical, routes and that the addition of only one of the acids resulted in the competitive inhibition of one or more of these routes. Several examples of this type of effect have been found during experiments with Neurospora mutants [cf. 45]-

Mutants

Much valuable information concerning the routes of bio- logical synthesis of natural compounds has been derived by the use of mutants, i.e. from organisms genetically different from the parent strains. If this difference results in inability

66 NITROGEN METABOLISM

to synthesize a compound essential for life, the mutant is unable to grow unless it can obtain from the environment at least one of the products of the reaction which it is unable to accomplish. Such a reaction is often referred to colloqui- ally as a 'genetically blocked reaction'. Beadle and Tatum predicted that it should be possible to deduce the sequence of reactions in biosyntheses from the range of compounds which replace the substances required by nutritionally- exacting mutants. The organisms used in these studies include Penicillium notatum, Aspergillus niger. Asp. nidulans, Ophiostoma, Esch. coli and B. subtilis, but the ascomycetes Neurospora crassa and N. sitophila still remain the most suitable if precise genetic data is also required. Most natural (i.e. 'wild type') strains of Neurospora grow on a simple medium containing mineral salts, biotin, an inorganic source of N and an organic source of C and energy (e.g. sucrose, sorbitol) and must therefore possess the wide variety of enzymes required for the synthesis of all the normal constituents of cytoplasm. The vegetative phase reproduces asexually by conidia and micro-conidia; sexual reproduction is only possible between gametes from parents of opposite mating types. Much is known concerning the genetics of Neurospora [see 5] and since the vegetative phase is haploid, there are in contrast with diploid organisms, no problems concerning the dominance of one character over another. Strains whose nutritional requirements are different from the parent type may arise naturally by spontaneous mutation, but such mutations are often few in number and natural selection does not favour their survival. In order to increase the chance of isolating such mutants, the mutation rate is artificially increased by exposing the conidia to ultra- violet light. X-rays or chemical mutagens, e.g. mustard gas. The conidia are then transferred to the protoperithecia of wild type Neurospora of opposite mating type, and in conse- quence asci develop, each ascus containing eight spores. One spore from each ascus is transferred to a solid medium con- taining the minimal requirements for growth (minimal medium), and after incubation the colonies that have developed will be of the wild type. Their position is noted

SYNTHESIS OF AMINO-ACIDS 67

and agar containing known additional nutrients is layered over the original plate. Any new colonies which develop after further incubation are derived from mutant spores, and after being subcultured their nutrition can be studied in more detail. Special techniques may be required in order to promote the formation of discrete colonies, particularly with organisms whose growth tends to spread (many strains of Neurospora). A more laborious method of isolation is to subculture each ascospore on a rich medium, i.e. one con- taining amino-acids and gro-wth factors, and then transfer to a minimal medium. If there is no growth on the latter, the nutrition of the parent colony is then examined further. Other more efficient, technically easy and less laborious methods for selecting mutants of various organisms have been described [18, 31, 32].

Using such techniques, a number of mutants have been obtained which are exacting towards a particular amino- acid, growth factor, purine or pyrimidine. Many appear to be unable to perform reactions expected to take place in one step, e.g. the amination of inosine to form adenine, and Beadle and Tatum have advanced the hypothesis that each enzyme is controlled by a specific gene, and any change in the latter is reflected by an alteration in the enzyme's activity. Although the evidence is indirect and has been criticized by Delbruck [11] an analysis of the available infor- mation shows that at least 73% of the genes of Neurospora have only one function, and there are no indisputable examples of genes with two or more functions [30]. It must be stressed that mutation may involve modification rather than complete loss of the gene and the corresponding enzyme, e.g. a Neurospora mutant unable to synthesize tryptophan from indole and serine still possessed the requi- site condensing enzyme, though in an inactive state [22]. (Cf. also the synthesis of pantothenic acid [51]).

In order to show whether mutants requiring the same factor are due to mutation of either the same gene or com- pletely different genes, one of two tests may be applied. In organisms with a sexual cycle (e.g. Neurospora), each mutant is mated with a parent of known genetic composition and the

68 NITROGEN METABOLISM

percentage recombination of characters obtained in the pro- geny of each cross will indicate whether or not the mutants are the outcome of mutation at the same gene locus. The cells of fungi are multinucleate and it frequently happens that hyphae from two parents will fuse and thus form a structure known as a heterocaryon which contains nuclei derived from each of the parent hyphae. Consequently a heterocaryon formed between two genetically identical mutants will have the same growth requirements as the two parent mutants. But the heterocaryon from genetically different mutants will grow in the complete absence of such compounds since the metabolic deficiencies of the nuclei derived from one parent will be complemented by the activities of the nuclei from the other parent.

Primarily because of the ease with which they can be isolated, mutants of penicillin-sensitive bacteria have been used in many recent investigations [9]. Mutation is brought about by irradiation with ultraviolet light in doses sufficient to kill 99*9% of the organisms and then the suspension is cultured in a rich medium, washed and transferred to mini- mal medium containing penicillin. The non-mutants grow and are consequently killed by the penicillin. The suspen- sion is then plated on to a rich medium and the colonies appearing are usually those of nutritionally exacting mutants.

Having isolated a number of mutants exacting towards a particular substance, other substances likely to be inter- mediates in its synthesis are then tested for their ability to promote grov^h, since theoretically the product of the blocked reaction or any compound derived from it should be active in this respect, provided the organism is permeable to such substances. The sequence of intermediates in a biosynthesis may then be deduced by arranging the mutants in order according to the range of compounds supporting grov^h. It is to be expected that the further the blocked reaction from the end-product, the greater the number of compounds utilized by the mutant. For example, in the synthesis of D from A by:

PLATE I. — Syntrophism among arginine requiring mutants of Esch. coli. Mutant O responds to ornithine, citrulline or arginine; C to citrulline or arginine; A to arginine only. Photograph of growth after 48 hr. at 37° C. on a medium containing suboptimal amounts of required nutrients. Note the enhanced growth of C and O due to the secretion by A of a substance (citrulline?) which can be utilized by C and O. Similarly the enhanced growth of O due to the secretion by C of a substance (ornithine?) utilized by O [7]

SYNTHESIS OF AMINO- ACIDS 69

the growth of a mutant incapable of reaction a will be sup- ported by B, Cor D\ if incapable of b, by C or D\ and if in- capable of c, the mutant will only grow in the presence of D.

For convenience, mutants are described either by adding the suffix '-less' to the substance required for growth, e.g. arginineless denotes exacting towards arginine, or by using the term auxotroph, e.g. an arginine auxotroph.

Growth in the presence of suboptimal amounts of the product of a blocked reaction may result in the excretion of the precursor of this reaction into the medium. Consequently when two related mutants are streaked near one another on a solid medium containing suboptimal amounts of required

HOOC(cH2)2CH(nH2)COOH NH2CONhCcH2)3Ch(nH2)cOOH

Qlutomic acid \ / citrulline

NH2(cH2)3Ch(nH2)cOOH ^ ornithine

proline

CH,— CH, / rjjHj NH2

CO ^CNH(cH2)3Ch(nH2)COOH

CHj^CHCOOH NM2 UU arginine

NH urgQ

FIG. 5.1. — The arginine cycle and the route of arginine synthesis in Neurospora crassa and Penicillium notatum

nutrients, one may secrete a substance which is used directly by the other or changed by it into a form which can be utilized by the former mutant. The enhanced growth which results is readily visible and this phenomenon of 'syn- trophism' has been widely exploited by Davis [7] (Plate I).

Arginine synthesis

Seven genetically distinct arginineless mutants of A^. crassa were isolated and of these, four grew on ornithine, citruUine or arginine; two on citrulline or arginine and one on arginine only. The organism also possessed arginase and urease and it was concluded that there is an 'arginine cycle' in Neuro- spora (Fig. 5.1) comparable to that described by Krebs in mammalian liver. It can be deduced frorn the genetic data that there are at least four steps in the synthesis of ornithine

70 NITROGEN METABOLISM

and two in the conversion of ornithine to citruUine. In Hver, the synthesis of citrulUne from ornithine, CO 2 and NH3 proceeds by a mechanism utiUzing metaboHc energy and with carbamylglutamic acid as an essential co-factor. Contrary to expectation, the latter does not function by transferring the carbamyl group directly to ornithine. In the presence of ATP and Mg"^"^, the citrulline combines with aspartic acid and the product subsequently undergoes hydrolysis to yield arginine and malic acid [38]. Comparable systems have not yet been described in micro-organisms. In Lb. arabinosus glutamine appears to play an essential role in arginine synthesis, the amide group being used in the formation of citrulline from ornithine [37^].

A mutant of P. notatum grew on arginine, citrulline, ornithine or proline, whilst another grew on either of these amino-acids or glutamic acid. A third was known to grow only on proline, indicating that proline is not itself on the direct route of arginine synthesis but is probably related to a precursor of ornithine [3, also cf. 43]. The possible relationships between glutamic acid, proline and the ar- ginine cycle are shown in Fig. 5.1. Studies of the nutri- tion of naturally occurring strains of lactobacilli [50] and mutants of Esch, coli provided evidence that the mechanism of arginine synthesis in bacteria is the same as in the fungi. Using mutants of Esch. coli, Davis and his colleagues [8] have shown that proline is formed by the reduction of A^-pyrroline-5-carboxylic acid (PCA), a compound formed from the y-semialdehyde of glutamic acid (GSA).

CH2 — CH.2 CH2 — Crl2

II II -H,0

HOOC CH.COOH >CHO CH.COOH

/ /

NH2 NH2

Glutamic acid (GSA)

CH2 — Cri2 CHg — Cri2

II +2H I I

CH CH.COOH > CH2 CH.COOH

\/ \/

N NH

(PCA) Proline

SYNTHESIS OF AMINO-ACIDS

71

The relationships of these compounds to ornithine has still to be fully elucidated [lyb]. The conversion of glutamic acid to ornithine probably proceeds via N-acetylglutamic acid — > N-acetylglutamic acid y-semialdehyde — > a-N- acetylornithine — > ornithine [see id].

Tryptophan and nicotinic acid

The interrelationships betv^een tryptophan, the other aromatic amino-acids and nicotinic acid, as revealed by experiments with N. crassa and Esch. coli, are summarized in Fig. 5.2. Initially, two tryptophanless mutants oi Neurospora (10575 ^^^ 40008) were isolated, both utilized indole in place of tryptophan but only one (40008) utilized anthranilic

COCH2Ch(nH2)cOOH

p-amirtobcnzoic acid

phenylolanine

tyrosine

nicotinic ecid

FIG. 5.2. — The tryptophan cycle and the synthesis of nicotinic acid in Neurospora crassa. Compounds A and B are hypothetical intermediates, and the details of the relationship between tryptophan and the other aromatic amino-acids and PAB are not yet known (see pp. 72-3) [24] 6

72 NITROGEN METABOLISM

acid. Growth of 10575 i^ the presence of limiting amounts of tryptophan resulted in the appearance in the medium of a substance supporting the growth of 40008. This material, presumably the substrate of the blocked reaction, was later isolated and identified as anthranilic acid. The reaction sequence is therefore anthranilic acid — >■ indole — > trypto- phan. The rate of uptake of indole by 10575 was found to be a function of the concentration of L-serine in the medium, and after growth had ceased tryptophan was excreted into the medium. The enzyme forming tryptophan by the con- densation of serine with indole was studied in cell-free homogenates of Neurospora mycelia and shown to con- tain a prosthetic group of pyridoxal phosphate [49]. The mechanism by which anthranilic acid is converted into indole remains unknown. By using isotopes it has been shown that the carboxyl group of the former does not give rise to any of the carbon in the indole nucleus of trypto- phan [37«].

A detailed investigation of Neurospora mutants able to grow on tryptophan or nicotinic acid has confirmed the conclusion drawn from animal nutrition experiments that the metabolism of these two compounds is interrelated. Furthermore, Haskins and Mitchell have proposed that, at least in Neurospora, there is a 'tryptophan cycle' [24]. By using the appropriate mutants and growth conditions, 3-hydroxyanthranilic acid, kynurenine and quinolinic acid have all been isolated from culture filtrates. Like Lb. arabinosus, the growth of some, but not all, nicotinic acid auxotrophs of Neurospora is supported by quinolinic acid, although only in high concentrations [21]. It is therefore possible that quinolinic acid is a by-product derived from the substrate of a blocked reaction, rather than a direct intermediate in the synthesis of nicotinic acid.

Certain mutants of A^. crassa and Esch. colt require tryp- tophan, phenylalanine, tyrosine and ^-aminobenzoic acid (PAB), all in large amounts, before there is even slow growth, indicating that these four compounds may be derived from a common precursor and that the synthesis of anthranilic acid and tryptophan is connected with the

SYNTHESIS OF AMINO-ACIDS

73

metabolism of the other aromatic amino-acids. As a result of a suggestion made by Stanier, Davis found that this quadruple requirement could be replaced by shikimic acid, an alicyclic compound known to occur in plants. This acid has now been isolated from the culture filtrate of a mutant of Esch. coli and unequivocally characterized. A precursor of shikimic acid (SKA) has recently been isolated and iden- tified as 5-dehydroshikimic acid (DSKA) which is in turn probably derived from 5-dehydroquinic acid (DQA) [8]. Since growth in the presence of the four aromatic acids was slow and became maximal on the addition of shikimic acid or filtrates from wild-type cultures, Davis deduced that the mutants required at least one other aromatic substance, and one of these has been identified as ^-hydroxybenzoic acid

(DQA)

(POB). Shikimic acid is only utilized by mutants exacting to at least four aromatic compounds and this multiple requirement is probably due to the mutation of a single gene. It is not yet possible to state whether shikimic acid is in fact a simple precursor of all these aromatic nitrogen compounds. Davis has suggested [7, 10] that the apparent complexity in growth requirements is the result of inter- ference with the synthesis of a key substance which is responsible for the integration of various parallel and related pathways of biosynthesis (cf. valine-isoleucine, pp. 65, 76). Unlike mammals, Neiirospora and Esch. coli cannot convert phenylalanine to tyrosine.

Cysteine and methionine

Sulphur is found in organic combination in the amino- acids cysteine and methionine, and most - organisms can utilize inorganic forms of sulphur at any oxidation level as

74

NITROGEN METABOLISM

a complete source of this element. Mutants exacting towards various sulphur compounds are the easiest to produce and isolate. Of four methionineless mutants of A^. crassa, only- one specifically required methionine, homocysteine was just as effective for two of the mutants, whilst the other grew on cysteine, homocysteine or methionine. The culture filtrate

l:

NH2 CH2

COOH CH.NHj COOH cystathionine

CHjSH

— yCHj _

CH.NHj

COOH

hornocyitaine

CHjSCHj

(fHjSH

syi

CHjOH

CH.NH, I ^

COOH

COOH methionine

r3

CHOH I CH.NHj

COOH

(jiHjS.SOaH

CH.NH2

COOH

cysteine-s-iul phonate

homoierine «/- ominobutyric ocid threon in e

. ' ll(

CHOH ». y

CHg^HjCHa ^ C.OH CHOH

COOH o(-keto.i3 hydroxybutyric acid

CHjOH HjSjOg^— H^SOjV-HzSOav- H2Sq4

CH.NHj

COOH *—

^>A=1

FIG. 5.3.

HCOOH

CH.NHj COOH glycine

-Pathways for the synthesis of cysteine and isoleucine

COOH di hydroxy-^- ethyl butyric ac'i?"

i CHg^CHjCHa CH

CH,NH2 COOH

isoleucine

, methionine

of one homocysteine auxotroph was found to contain a sub- stance capable of supporting the growth of the other homo- cysteine autotroph and also of the mutant which would grow on cysteine. This substance was isolated and identified as cystathionine [29], and the suggested biosynthetic sequence is shown in Fig. 5.3. Methionine is probably synthesized by methylation of the homocysteine produced by the cleavage of cystathionine (see p. 149).

SYNTHESIS OF AMINO- ACIDS 75

A Neurospora mutant requiring threonine grew poorly unless the medium was also supplemented with methionine, cystathionine or homocysteine, indicating that all these com- pounds were derived from a common precursor. The latter is probably homoserine since this amino-acid can replace both threonine and the sulphur-containing amino-acids [46]. Moreover, if homoserine can couple with cysteine the product would be cystathionine. As regards the incorpora- tion of inorganic forms of sulphur into organic compounds, it has been suggested that this proceeds via the synthesis of cysteine by cysteine desulphurase, an enzyme known to decompose cysteine into HgS, NH3 and pyruvic acid. There is, however, no evidence that the enzyme can catalyse the reverse reaction. In Asp, ntger, Asp. nidulans and P. Tiotatum [27], thiosulphate, and not sulphide, may be an important intermediate in the synthesis of cysteine. The sulphur requirements of thiosulphate auxotrophs of Asp. 7iidulans were satisfied by cysteine- S-sulphonate (a thiosul- phate derivative of serine) and growth was extremely luxuri- ous in the presence of serine and thiosulphate [28]. Hocken- huU therefore proposed that the route of sulphur utilization involves reduction of sulphate to sulphite which is then converted to thiosulphate, perhaps with the intermediate formation of sulphoxylate, and that finally thiosulphate is condensed with serine or some other Cg-compound. A different sequence has been proposed for Ophiostoma multi- annulatum [see 5] and N. crassa:

Cs-compound (alanine?) plus S07~ — >- cysteic acid — >

cysteine sulphinic acid — >- cysteine

Lysine and threonine

The route of lysine synthesis in Neurospora appears to be different from that in Esch. coli. Lysine auxotrophs oi Neuro- spora were able to utilize a-aminoadipic or e-hydroxy- a-aminocaproic acid but not a-ketoadipic, a,a'-diamino- adipic or a,£-diaminopimelic acid. The biosynthetic se- quence in N. crassa is believed to be:

a-aminoadipic acid — >-

€-hydroxy-a-aminocaproic acid — >■ — > — > lysine [20]

^6 NITROGEN METABOLISM

Unlike Neurospora, Esch. coli contains a,e-diaminopimelic acid (DAP) and possesses a specific enzyme converting it into lysine and CO 2 (p. 29). This decarboxylase is not present in those lysineless mutants which accumulate large amounts of DAP in their culture media. The lysine require- ments of other mutants were satisfied by DAP but not by a-aminoadipic or a-amino-£-hydroxycaproic acid, and the conclusion was reached that DAP is the immediate pre- cursor of lysine in Esch. coli. Moreover, DAP and threonine may be derived from a common precursor, and a-amino- butyric acid may be an intermediate in the synthesis of threonine from homoserine (Fig. 5.3) [9].

Valine, isoleucine and threonine

A mutant (161 17) of N. crassa would only grow when provided with both L-valine and L-isoleucine, yet as far as could be ascertained it diifered in only one gene from the wild type parent. For optimal growth, the ratio of L-valine to L-isoleucine was critical (7:3) and increasing the concen- tration of either acid adversely affected growth. The a-keto acids corresponding to valine and isoleucine ('ketovaline' and 'ketoisoleucine') supported the growth of other mutants, and 161 17 would grow in the presence of isoleucine and 'ketovaline' but not valine and 'ketoisoleucine'. Moreover, 'ketoisoleucine' inhibited the growth of another mutant requiring only 'ketovaline'. Bonner has suggested that 16117 is unable to synthesize isoleucine and that the sub- strate of the blocked reaction ('ketoisoleucine'?) accumulates and competitively inhibits the synthesis of valine (from 'ketovaline'?), with the result that the mutation of one gene appears to bring about the blocking of two reactions [3]. After 161 17 has grown in the presence of isoleucine and valine, the medium contains a,^-dihydroxy-^-ethylbutyric acid, a substance replacing isoleucine for an auxotroph of Esch. coli; no 'ketoisoleucine' could be detected [ib]. Iso- leucine auxotrophs of N. crassa, B. suhtilis and Esch. coli can be divided into three groups according to the com- pounds which they can utilize: (i) only isoleucine, (2) a,j5-dihydroxy-/5-ethylbutyric acid or isoleucine, (3) iso-

SYNTHESIS OF AMINO-ACIDS 77

leucine, a,y5-dihydroxy-/?-ethylbutyric acid, a-aminobutyric acid, a-ketobutyric acid or threonine. Hence, the metabo- lism of isoleucine, the sulphur amino-acids and threonine is closely related and may proceed from a common C4- precursor. Isoleucine may arise from the latter by the addition and reduction of an acetyl group [ib] (see Fig. 5.3). The keto-acids of valine and isoleucine have been identified in the culture filtrate of an isoleucinless mutant of Esch. coli [47].

Isotopes

Isotopes of carbon are proving useful in tracing the origin of the various carbon atoms of amino-acids, and their application to the study of syntheses in micro-organisms (Torulopsis utilis, Esch. coli, N. crassa) is mainly due to Ehrensvard and his colleagues [6, 12, 13]. By using acetate as a sole source of carbon and labelling the carbon in the two positions with different isotopes (C^^HgC^^OOH), it is possible to determine whether a particular C atom is derived more or less directly from the CH3 or the COOH group. With the yeast T. utilis adapted to growth on acetate, it was found that two acetate carboxyl groups were liberated as respiratory CO 2 for every methyl group. After hydro- lysing the yeast with acid, the amino-acids were isolated by means of electro-dialysis and chromatography on ion exchange resins. The ratio of C^* to C^^ in the carboxyls of most of the amino-acids was the same as that in the respiratory CO 2 , thus demonstrating that CO 2 fixation had taken place. In general, all the alpha-C atoms and many of those in the side chains were derived from the methyl group of the acetate. It was concluded that glutamic acid is a precursor of arginine (cf. p. 70) and that lysine is synthe- sized by the head to tail condensation of acetyl radicals. When NHa-CHgC^^OOH was used as a sole N source the C^* was finally located mainly in the glycine, serine and proline of the proteins. Thus in Torulopsis, as in bacteria (p. 149) and animals [cf. 14], glycine is a precursor of serine. By growing cultures of Sac. cerevisiae, T. utilis and Ps. fluorescens in the presence of HC^*OOH, it has been shown

78 NITROGEN METABOLISM

that the C in position 2 of the imidazole ring of histidine is derived exclusively from formate [33, 44].

Isotopically labelled compounds such as amino-acids can be readily prepared by isolating them from organisms which have been grown in media containing substances enriched with the appropriate isotope [2]. Non-exacting organisms are especially useful since they utilize simple substances (HC07,NHt, SOI") as sources of C, Nand S and such substances enriched with the appropriate isotopes are readily available.

REFERENCES

la. Abelson, P. H. (1954), jf. hiol. Chem., 206, 235 \h. Aldelberg, E. A. (1951), J. Bact., 61, 365 \c. Altenbern, R. A. and Housewright, R. D. (1953), J. hiol. Chem., 204, 159

2. Aquist, S. E. (195 1), Acta chem. Scand., 5, 103 1

3. Bonner, D. (1946), Cold Spr. Harb. Symp., 11, 14

4. Braunstein, A. E. and Kritzmann, M. G. (1937), Enzymologia,

2, 129

5. Catcheside, D. G. (195 1), The Genetics of Micro-organisms

(Pitman, London)

6. Cutinelli, C, Ehrensvard, G., Reio, L., Saluste, E. and

Stjernholm, R. (195 1), Acta chem. Scand., 5, 353

7. Davis, B. D. (1950), Experientia, 6, 41

8. (1952), Intermediates in the Biosynthesis of Proline, etc.,

in Biological Systems, 2nd Internat. Congress of Biocheni.

9. (1952), Nature, 169, 534

10. (1952), J. Bact., 64, 729, 749

11. Delbruck, M. (1946), Cold Spr. Harb. Symp., 11, 22

12. Ehrensvard, G. (1952), On the Origin of Aromatic Structures

in Biological Systems, 2nd Internat. Congress of Biochem.

13. Reio, L., Saluste, E. and Stjernholm, R. (1951), J. hiol.

Chem., 189, 93

14. Feldman, L. I. and Gunsalus, I. C. (1950),^. biol. Chem., 187,

821

15. Fildes, P. (1945), Brit. J. exp. Path., 26, 416

16. and Rydon, H. N. (1947), Brit. J. exp. Path., 28, 211

i7<3. Fincham, J. R. S. (1950), J. biol. Chem., 182, 61

176. (1953), Biochem. J., 53, 313

18. Fries, N. (1947), Nature, 159, 199

19. Gladstone, G. P. (1939), Brit. J. exp. Path., 20, 189

20. Good, N., Heilbronner, R. and Mitchell, H. K. (1950), Arch.

Biochem., 28, 464

21. Gordon, M., Haskins, F. A. and Mitchell, H. K. (1950), Proc.

Nat. Acad. Set., Wash., 36, 427

SYNTHESIS OF AMINO- ACIDS 79

22. Gordon, M. and Mitchell, H. K. (1950), Genetics, 35, no

23. Green, D. E., Leloir, L. F. and Nocito, V. (1945), J. biol.

Chem., 161, 559

24. Haskins, F. A. and Mitchell, H. K. (1949), Proc. Nat. Acad.

Sci., 35, 500

25. Herbst, R. M. (1944), Advances in Enzymology, 4, 75

26. Hird, F. R. J. and Rowsell, E. V. (1950), Nature, 166, 517

27. Hockenhull, D. J. D. (1948), Biochem. J., 43, 498

28. (1949), Biochem. Biophys. Acta, 3, 326

29. Horowitz, N. H. (1947), J. biol. Chem., 171, 255

30. (1950), Advances in Genetics, 3, 33

31. Lederberg, J. and Tatum, E. L. (1946), J^. biol. Chem., 165, 381

32. Lein, J., Mitchell, H. K. and Houlahan, M. B. (1948), Proc.

Nat. Acad. Set., 34, 435

33. Levy, L. and Coon, M. J. (1951), jf. biol. Chem., 192, 807

34. Lichstein, H. C. and Cohen, P. P. (1945), J. biol. Chem., 157,

35. Gunsalus, I. C. and Umbreit, W. W. (1945), jf. biol.

Chem., 161, 311

26a. Meister, A. (1952), jf. biol. Chem., 195, 813

366. Sober, H. A. and Peterson, E. A. {1952), J. Amer. chem.

Soc, 74, 2385 37a. Nye, J. F., Mitchell, H. K., Leifer, E. and Langham, W. H.

(1949), J. biol. Chem., 179, 783 376. Ory, R. L., Hood, D. W. and Lyman, C. M. (1954), J. biol.

Chem., 207, 267 38. Ratner, S. (1949), Fed. Proc, 8, 603 29a. Roine, P. (1947), Thesis, Helsinki, On the formation of primary

amino-acids in protein synthesis in yeast 29b. Rudman, D. and Meister, A. (1953), J. biol. Chem., 200, 591

40. Rydon, H. N. (1948), Brit. J. exp. Path., 29, 48

41. Shemin, D. (1949), Cold Spr. Harb. Symp., 14, 161

42. Snell, E. E. (1945), J. Amer. chem. Soc, 67, 194

43. Srb, A. M., Fincham, J. R. S. and Bonner, D. (1950), JJ^. Amer.

Bot., 37, 533

44. Tabor, H., Mehler, A. H., Hayaishi, O. and White, J. (1952),

J. biol. Chem., 196, 121

45. Tatum, E. L. (1949), Fed. Proc, 8, 511

46. Teas, H. J., Horowitz, N. H. and Fling, M. (1948), J. biol.

Chem., 172, 651

47. Umbarger, H. E. and Magasanik, B. (1951), J. biol. chem.,

189, 287

48. Umbreit, W. W., O'Kane, D. E. and Gunsalus, L C. (1946),

y. Pact., 51, 576

49. Wood, W. A. and Gunsalus, L C. (1946), J. biol. Chem.,

165,731

50. Volcani, B. E. and Snell, E. E. (1948), X biol. Chem., 174, 893

51. Wagner, R. P. (1949), Proc. Nat. Acad. Sci., 35, 185

CHAPTER VI

ABSORPTION OF AMINO-ACIDS BY MICRO-ORGANISMS

The amino-acids required for growth are either synthesized by the organism itself or derived from its environment. Consequently the mechanisms controlling the absorption of amino-acids are of fundamental importance, especially to exacting organisms who must perforce rely entirely on their environment for supplies of compounds which they are un- able to synthesize. The factors influencing the passage of amino-acids into and out of micro-organisms, particularly those exacting towards amino-acids, have been investigated by Gale and his colleagues [6]. Whilst using the bacterial decarboxylases [4] (p. 27) for the analysis of the amino-acid composition of bacterial proteins [3], Gale found that Gram-positive bacteria contained large amounts of free amino-acids. This discovery arose from the fact that hydro- lysates of whole cells, and not isolated proteins, were being analysed, and it was realized that serious errors would result if there were appreciable amounts of free amino-acids in the experimental material prior to hydrolysis. Such amino- acids might be adsorbed on to the cell surface or be present inside the cells. In order to determine the significance of these possibilities, washed cells of Strep, faecalis were dis- integrated by shaking with glass beads, a procedure which ruptured the cells but caused no significant degradation of the proteins. Analysis of the disintegrated cells revealed the presence of relatively high concentrations of certain amino- acids (cf. columns h and a—hm Table 6.1). Although the latter were also found in disintegrated cells of Staph, aureus, there were no significant amounts in those of two Gram- negative organisms, Esch. colt and Aerobacter aerogenes [5]. Further experiments showed that the greater proportion of these amino-acids was located inside the cells.

There was therefore an indication that only Gram- 80

ABSORPTION OF AMINO-ACIDS 8l

positive bacteria contained free amino-acids. Moreover, since many amino-acid exacting organisms are also Gram- positive, their presence might be the outcome of mechanisms enabUng such organisms to absorb essential amino-acids

	TABLE	6.1
THE AMINO-ACID COMPOSITION OF STREP. FAECALIS	CELLS [5;
Results expressed	in terms of amino-	acid N as	0/ /o	total N
Amino-acid	Acid hvdrolysate	Disintegrated cells ib)		Combined amino-acid {a-b)
L(+)-Lysine L(+)-Argimne L(+)-Glutamic acid L(-)-Histidine L(+)-Omithine l(— )-Tyrosine	11-30 5-19 5-60 2-84 1-48 o-8i		2-90 010 1-74 0-70 1-27 o-o		8-40 5-09 3-86 2-14 0-2I o-8i

from their environment. In view of the potential signifi- cance of these observations, Gale decided to undertake a series of more detailed investigations. Apart from the intrinsic value of the results these experiments provide a valuable example of one type of approach to microbiological problems.

Procedure for assaying internal amino-acids

A thick suspension (about 30 mg.dry \vt./ml.) of washed cells of the organism is prepared and the amount of external amino-acid is determined by adding the appropriate decar- boxylase preparation to a sample of this suspension. The total amount of free amino-acid, i.e. the amount inside the cells plus that outside, is determined by adding the decar- boxylase to a sample of the suspension which has been previously heated at 100° C. for 15 minutes in order to dis- rupt the cells. The amount of amino-acid inside the cells is then readily calculated by subtracting the former result from the latter. Concentrations are expressed in terms of either the amount of amino-acid in a specified dry weight of cells, or the amount per millilitre of cell volume or 'internal free-space'. The volume occupied by the cells was

82 NITROGEN METABOLISM

obtained by centrifuging samples of the whole cell suspen- sion in graduated tubes and the volume of the cellular con- stituents was calculated by assuming that they were mainly proteins whose specific volume is 0-70. Subtraction of the volume of the cellular constituents from the volume of an equivalent amount of whole cells yields a value for the 'internal free-space', i.e. the internal environment [5].

The amino-acid decarboxylases are enzymes of high specificity (p. 27), consequently this procedure will esti- mate only amino-acids which are initially free in the cells or which are liberated from compounds that are easily decom- posed during the assay procedure, e.g. by heat or by other enzymes present in the decarboxylase preparations. Whilst enzymic decomposition during the assay procedure has not been ruled out, it is unlikely that they arise by the decompo- sition of heat-labile compounds since the same amount of amino-acid is found in cells disrupted by heat as in those made permeable by treatment with tyrocidin or phenol [13]. In the absence of evidence to the contrary the internal amino-acids are regarded as being 'free' in the sense that they are chemically uncombined.

Internal amino-acids in Gram-positive organisms

Washed suspensions of twenty-seven organisms com- prising thirteen genera were prepared from cultures which had been grown on a medium rich in amino-acids. The experiments were restricted to the six amino-acids for which specific decarboxylases were then known. Free amino-acids w^ere found only in Gram-positive organisms, yeasts as well as bacteria, and none were detected in Gram-negative bacteria. The yeasts contained high concentrations of all six of the amino-acids whereas the bacteria contained only lysine and glutamic acid in appreciable amounts [18].

Absorption of amino-acids by washed cells

The factors controlling the absorption of lysine and glutamic acid have been studied in experiments with washed cell suspensions of Strep, faeca lis and Staph, aureus {Micro- coccus pyogenes var. aureus), both of which are exacting

ABSORPTION OF AMINO-ACIDS 83

towards amino-acids. Cells which initially contain negligible amounts of amino-acids (i.e. 'amino-acid deficient cells') are especially suitable for this type of experiment and they were obtained from cultures grown on a liquid medium in which the concentration of amino-acids was just sufficient for growth. Such cultures were harvested near to the cessation of active cell division since the ability to absorb amino- acids was observed to decline appreciably in the stationary phase. The amino-acid deficient cells were suspended in an inorganic salt medium (pH 7-2) to which other substances (amino-acids, glucose, etc.) were added as required. At the end of the experimental period the cells were collected by centrifuging, washed, made into a thick suspension and the internal concentration of the amino-acid determined. Unless stated to the contrary the results given below apply to Strep, faecalis.

Uptake of lysine [5]

As soon as lysine was added to the experimental system it began to pass into amino-acid deficient cells of Strep, faecalis, and after about 15 minutes equilibrium was reached and there was no further increase in the internal concen- tration. This uptake of lysine occurred rapidly at 37° C. and was still appreciable at 4° C. The rate of appearance of lysine inside the cells w^as approximately directly propor- tional to its concentration in the external medium (Fig. 6.1), and the Q^q over the range of 20°-30° C. was i -4. From these facts it may be deduced that lysine is entering the cells by a process of diffusion. However, this uptake did not repre- sent simple equilibration betw^een lysine-deficient cells and a lysine-rich environment, since at equilibrium the internal concentration of the amino-acid was from three to twenty times greater than the external concentration, the ratio of internal to external concentration being inversely related to the latter. In other words, the cells were accumulating lysine against a concentration gradient (Fig. 6.2). The rate of lysine absorption increased as the hydrogen-ion concen- tration of the salt medium was decreased to pH 9-5, and since the isoelectric point of lysine is 9*47, it is feasible that

84 NITROGEN METABOLISM

lysine most readily enters the cells as the zwitterion. When the cells fermented glucose, the uptake of lysine was de- pressed but could be restored, although not completely, if glutamic acid was also added to the system. The factors influencing the outward migration of lysine were also of importance, if only to show that washing the cells prior to

1 1 1

O 5 10 15

EXTERNAL CONG. OF AMINO- ACID (a' mole/mi.)

FIG. 6.1. — Effect of external concentration of L-Iysine (#) and L-glutamic acid (O) on the rate of entry of the amino-acid into Strep, faecalis

assay did not remove any of the internal amino-acids. Cells containing large amounts of lysine, glutamic acid and probably several other amino-acids were obtained from cul- tures grown in the presence of a tryptic digest of casein. When these 'amino-acid rich cells' were incubated in an amino-acid free salt medium, there was no outward migra- tion of either lysine or glutamic acid unless the cells were fermenting glucose.

ABSORPTION OF AMINO- ACIDS 85

More recent experiments have served to emphasize that the previous history of the cells and the presence of other amino-acids profoundly influence the inward as v^ell as the outward migration of lysine and glutamic acid. Cells con- taining only lysine in large amounts were obtained by

50-1

4-0-

3-0-

«/)2-0-

r20

-16

hi2

d

z o o

z cr

1 £- ^

EXTERNAL CONG. (/* mole/m.l.)

?i

-4.

20

FIG. 6.2. — Effect of external concentration of lysine on (a) internal concentration of lysine (#) and (b) the ratio of internal con- centration of lysine to external concentration: in ratio calcu- lations internal concentrations expressed in terms of volume of intact cells (A) and volume of intact cells less that of solid debris (^). Strep, faecalis suspended in lysine solutions for 3 hr. at 4° C. [5]

incubating a washed suspension of amino-acid deficient cells in the presence of lysine [16]. When these cells were transferred to an amino-acid free environment (o'i5 M.-NaCl or Na2HP04 at 37° C.) lysine migrated out into the sur- rounding medium and, in contrast to the previous experi- ments with amino-acid rich cells, the migration occurred in

86 NITROGEN METABOLISM

the absence of glucose fermentation. It may be concluded that the presence of high concentrations of other amino-acids in the internal environment promotes the retention of lysine inside the cells and that their influence is overcome by glucose fermentation. The converse is also true, since the presence of acidic or basic amino-acids in the external environment retarded the uptake of lysine by amino-acid deficient cells in the absence of glucose. The effects of orthophosphate on the outw^ard migration of lysine have been investigated [i6], but the results are difficult to inter- pret and the reader is referred to the original paper.

From all these results, it may be deduced that the accumu- lation of lysine by Strep, faecalis involves a diffusion mech- anism which is independent of metabolic energy and may be due to the establishment of a Donnan type of equilibrium. Such a hypothesis is supported by the observation that, when suspended in an amino-acid free medium, lysine migrated out of cells which contained only this amino-acid in large amounts. However, it must not be assumed that the mechanism of lysine absorption is the same in all Gram- positive organisms. Although similar results were obtained with Staph, aureus, the uptake of lysine by Sac. cerevisiae was dependent on metabolic energy [6] and is therefore comparable to the uptake of glutamic acid by Strep, faecalis and Staph, aureus.

Uptake of glutamic acid [5]

Irrespective of the pH of the inorganic salt medium there was no absorption of glutamic acid by the amino-acid de- ficient cells unless glucose or a tryptic digest of casein was also added. Hence, unlike lysine, the uptake of glutamic acid appeared to be an endergonic process utilizing meta- bolic energy made available by the fermentation of glucose or by the metabolism of amino-acids. Arginine promoted the absorption of glutamic acid, although not as efficiently as glucose, and the fact that the organisms possess arginine dihydrolase (p. 27) may be of some significance. In the presence of glucose, glutamic acid was not only absorbed but also concentrated in the cells, and the relationships

ABSORPTION OF AMINO-ACIDS 87

between the internal and external concentrations at equili- brium were similar to those found for lysine (cf. Fig. 6.2). But, in contrast to lysine, the rate of uptake was not directly proportional to the external concentration (Fig. 6.1) and the relationship was reminiscent of that between the rate of an enzyme reaction and the concentration of the sub- strate. The Qjo over the range 2o°-30° C. w^as 1-94, which is close to 2-0 and therefore indicative of a chemical and presumably of an enzymic reaction.

Various inhibitors of intermediary metabolism, such as cyanide and iodoacetate, have no effect on the uptake of lysine, but any substance inhibiting fermentation also inhibits the uptake of glutamic acid. However, it is possible to separate the processes of energy production from those of energy utilization by preferential inhibition of the latter. This may be accomplished by using a substance such as 8-hydroxyquinoline which in low concentrations inhibits the uptake of glutamic acid without affecting fermentation, although higher concentrations inhibit the latter as well [7]. Since 8-hydroxyquinoline is a chelating agent, the conclu- sion was drawn that cations played an important role in glutamic acid absorption. Staphylococcus aureus was used in these experiments and the problem was further investigated by growing the organism in the amino-acid poor medium from which one or more cations had been removed. The ability to absorb and concentrate glutamic acid was seriously impaired only in cells harvested from media deficient in Mg"^"^ or Mn"*""^. Whilst there is insufficient data for deciding which of these ions is the natural activator of the glutamic acid absorption mechanism, it is pertinent to note that both of these cations are frequently found as co-factors of enzymes associated with phosphorylation.

In view of the dependence of the process on metabolic energy it is conceivable that glutamic acid passes through the cell wall in the form of a compound whose synthesis is endergonic (e.g. as glutamine, glutathione, or a phosphory- lated derivative), and having passed through the cell wall, this compound is reconverted to the free acid. Glutamine, glutathione and glutamylglutamic acid all failed to enter 7

58 NITROGEN METABOLISM

the cells unless glucose was also added to the experimen- tal system. On the other hand, a,}/-diethylglutamate and N-phosphorylglutamic acid were absorbed in the absence of glucose and gave rise to internal glutamic acid [8]. Hence it is possible that glutamic acid passes through the cell wall in a phosphorylated form, and irrespective of whether the phosphate group is on the amino group or on one or both of the carboxyl groups, the synthesis of such a compound would be endergonic. But, the evidence in favour of this hypothesis is far from conclusive. Many cells are known to be impermeable to phosphorylated compounds and the Q^q for the uptake of N-phosphorylglutamic acid was inter- mediate between that for free diffusion and for an enzymic process.

When amino-acid deficient cells of Staph, aureus were incubated with glutamic acid and glucose together with mixtures of other amino-acids, the accumulation of free glutamic acid was reduced and sometimes ceased. This was the outcome of a marked reduction in the amount absorbed and also an increase in the cellular combined glutamic acid. If the concentration of glutamic acid in the external medium was very much greater than that of the other amino-acids, assimilation into cell substance was suppressed and some accumulation of the free acid did take place. Evidently whether glutamic acid entered the cells and accumulated as the free acid or whether it was converted into a combined form, depended on the ratio of glutamic acid to other amino- acids in the external medium. The presence of single acids such as aspartic, cysteine, glycine, serine and alanine pro- duced a marked decrease in the rate of glutamic acid accu- mulation. Valine, leucine and isoleucine caused a small increase in the latter, whilst all the other amino-acids examined had no effect. Aspartic acid acted as a competitive inhibitor, but this explanation did not apply to cysteine and alanine, the addition of which led to the synthesis of extra- cellular peptides containing glutamic acid [lo, ii].

There was a small leakage of glutamic acid when cells of Staph, aureus containing large amounts of several amino- acids were incubated in an amino-acid free medium: the

ABSORPTION OF AMINO-ACIDS 89

addition of glucose completely suppressed this outward migration. On the other hand, there was no outward migra- tion of glutamic acid from Strep, faecalis unless glucose was present [5].

The uptake of histidine and aspartic acid by Strep, faecalis is also dependent on metabolic energy [5] and in Sac. cerevisiae this applied to all of the amino-acids investigated [19, 6].

Mechanism of amino-acid absorption

Whilst there are unaccountable differences even between somewhat similar organisms such as Strep, faecalis and Staph, aureus (cf. the action of inhibitors in 6, 9), there is sufficient data to warrant a discussion of the mechanisms which may be operative in the absorption of amino-acids. This topic is of particular interest since the physical and chemical structure of the barrier separating the interior of the bacterial cell from the external environment is now being actively studied by several workers [15, 17, 21]. Moreover, amino-acids exist in solution as ions, consequently their mode of absorption may be only one aspect of the general problem of ion transport across cellular membranes [cf. 20].

The normal electrochemical properties of a cell will be markedly altered by the absorption of ions of one species unless there is simultaneously an equivalent migration of ions either of opposite charge into the cell or of like charge out of the cell. Many experiments with plant and animal cells show that the uptake of one type of ion usually involves the concurrent transport of other ions, and it is therefore legitimate to expect that the uptake of amino-acids is like- wise associated with the movement of other ions. Whether this does in fact happen has not yet been investigated, and indeed little is known about the migration of ions into and out of micro-organisms. Several plant and animal cells con- tain higher concentrations of particular inorganic ions than the extracellular fluids, e.g. the K"^ content of mammalian cells is greater than that of the plasma. One explanation of the unequal distribution of diffusible ions between cells and their environment has been advanced by Donnan, who

90 NITROGEN METABOLISM

proposed that such a distribution in a system in equilibrium is the natural outcome of the presence of intracellular non- diffusible ionic substances such as proteins [2].

The experiments with Strep, faecalis, Staph, aureus and Sac. cerevisiae have shown that certain amino-acids can pass into the organisms against the concentration gradient and at equilibrium their internal concentration may be, and often is, greater than their concentration in the external environment. Furthermore, two distinct mechanisms appear to be operating in the absorption of amino-acids, one involving simple diffusion, e.g. the uptake of lysine by Strep, faecalis and Staph, aureus, whilst the other involves the utilization of metabolic energy, e.g. the uptake of glu- tamic acid. These facts may be accounted for by one of four explanations:

(i) That as a result of electrostatic attraction, the amino- acid becomes associated inside the cell with a non- diffusible ion of opposite charge, thus establishing a type of Donnan equilibrium [2].

(ii) That once inside the cell, the amino-acid is con- verted into a compound which cannot itself pass through the cell wall.

(iii) That the cell is permeable, not to the amino-acid itself, but to a derivative whose synthesis is endergonic.

(iv) That the migration of the amino-acid results from the movement of another ion whose formation or transport is endergonic.

Mechanism of lysine absorption

Najjar and Gale suggested that the absorption and accu- mulation of lysine in Strep, faecalis is due to the establish- ment of a type of Donnan equilibrium [16]. In physio- logical conditions of pH lysine carries an overall positive charge, consequently if only lysine is absorbed it is reason- able to suggest that this is accompanied either by the excre- tion of an equivalent amount of another cation such as H or K"*", or by the absorption of an equivalent amount of

ABSORPTION OF AMINO-ACIDS 9I

anion. Even though the latter takes place, there may be another mechanism which excretes an equivalent amount of either the same or a different anion together with cations to replace the lysine absorbed. It may be recalled that fer- menting yeast absorbs K"^ from a medium containing KCl and replaces it by H"^ [see 20]; there is apparently no uptake of Cl~. A possible explanation here is that metabolic energy is used to form H"^, which is then secreted into the medium in exchange for K"^, thus making it appear that the uptake of K"^ is an active process, i.e. dependent on metabolic energy. It may therefore be suggested that lysine is absorbed by Strep, faecalis in exchange for cellular K"^ or another cation which does not have to be formed at the expense of metabolic energy. However, recent work dealing with the accumulation of lysine showed that in Strep, faecalis it was accompanied by a gain in cellular K^ (with no significant change in cellular Na"*") whereas with Staph, aureus it had no apparent effect on either the K"^ or Na"*" content of the cells. In Saccharomyces fragilis, the uptake of lysine was dependent on glucose fermentation and was accompanied by the loss of Na"^ and K"*" from the cells [ib] (in none of these experiments was the migration of 0H~ and H studied).

Mechanism of glutamic acid absorptio?i

Having regard to the second explanation (ii) advanced above, it might be suggested that the 'free glutamic acid' of the cell is in the form of glutamine, a substance whose synthesis is endergonic and which is assayed by the decar- boxylase preparation as though it is the free acid: moreover, as glutamine cannot freely diffuse into cells, it is possible that it cannot itself pass through the cell wall [5]. But only a small part of the glutamic acid of the internal environment of streptococci is in fact in this form [14]. Hence the energy associated with the uptake of glutamic acid may be used in the manner outlined in the last two explanations, iii and iv. The former of these proposes that active transport, i.e. trans- port dependent on metabolic energy, involves chemical

92 NITROGEN METABOLISM

reactions between cellular constituents and the substance being transported across the cell membrane [20]. This implies that the latter contains a substance or carrier which reacts with the transported substance to form a relatively stable product which traverses the cell membrane and then either undergoes chemical decomposition or takes part in an exchange reaction, thus liberating the transported substance into the internal environment. Energy may be required for the formation of either the compound traversing the mem- brane or the substance which takes part in the final exchange reaction. Davies and Krebs [id] have shown theoretically how metabolic energy may be utilized for the production of H"^ or 0H~ from water. By analogy with their hypotheses concerning ion transfer in brain cells it may be proposed that metabolic energy is used to form an excess of 0H~> which may perhaps combine with CO 2 to form HCO 3", and that the bacterial cell wall contains a basic ion exchange complex (X). If the cell contains a high concentration of an anion such as 0H~ or HCO^ and the external environment contains glutamate ions, and if it is assumed that the ion exchange complex X can move in the cell wall and thus come into contact with the external and internal environ- ments, the following reactions can be expected to take place, the equilibrium being towards the right:

internally: X+ + OH" ^ X-OH

externally: X-OH + glutamate ^ X-glutamate + 0H~

internally: X-glutamate + 0H~ ^ X-OH + glutamate

Glutamate ions will thus be transported into the cell and will be replaced in the external environment by hydroxyl ions. Eventually a steady state will be established in which exchange is still taking place, but there is no further overall increase in the internal concentration of glutamic acid. Specificity in ion transport may be due to different ions being transported by diiferent ion-exchange complexes whose specificity properties are comparable with enzymes. The observation by Britten that intracellular glutamic

ABSORPTION OF AMINO-ACIDS 93

acid will exchange with extracellular isotopically labelled glutamic acid in the absence of metabolic energy can be explained if there is a carrier mechanism of this type in the cell membrane. All these theories are purely speculative and their acceptance or rejection awaits the results of further experiments. One fact of which account must now be taken is that the accumulation of glutamic acid within ferment- ing cells of Staph, aureus, Strep, faecalis and Sac. fragilis is accompanied by an increase in cellular K"^, the increase appearing to be of the order i gram atom K'^/mole glutamic acid [lb].

The full significance of the ability to absorb and accumu- late amino-acids still awaits complete evaluation since only three organisms and a restricted number of amino-acids have so far been investigated. It may be remarked that a decrease in the nutritional requirements of Staph, aureus is accompanied by a decreased ability to accumulate amino- acids [12], but it must also be noted that Sac. cerevisiae is not exacting towards amino-acids and yet accumulates many of these compounds [19]. Moreover, washed suspensions of a large variety of organisms, Gram-positive as well as Gram- negative, decompose several amino-acids, the implication being that unless the appropriate enzyme is in the cell surface the amino-acid enters the cell by free diffusion.

REFERENCES

la. Davies, R. E. and Krebs, H. A. (1952), Biochem. Soc. S.ymp.,

8, Chap. 6 lb. Davies, R., Folkes, J. P., Gale, E. F. and Bigger, L. C. (1953),

Biochem. y., 54, 430

2. Donnan, F. G. (1924), Chem. Rev., i, 73

3. Freeland, J. G. and Gale, E. F. (1947), Biochem. J., 41, 135

4. Gale, E. F. (1947), Biochem. J., 41, vii

5. (1947), J- gen. Microbiol, i, 53

6. (1948), Bull. Johns Hopkins Hosp., 83, 119

7. (1949), J' gen. Microbiol, 3, 369

8. (1950), y. gen. Microbiol, 4, v

9. (195 1), Biochem. y., 48, 286

10. (195 1), Biochem. y., 48, 290

11. and van Halteren, M. B. (1951), Biochem. y., 50, 34

94 NITROGEN METABOLISM

12. Gale, E. F. and Rodwell, A. W. (ig^g), jf. gen. Microbiol., 3, 127

13. and Taylor, E. S. (1947), J. gen. Microbiol., i, 77

14. Mcllwain, H., Roper, J. A. and Hughes, D. E. (1948),

Biochem. J., 42, 492

15. Mitchell, P. and Moyle, J. (195 1), J. gen. Microbiol, 5, 966,

981

16. Najjar, V. A. and Gale, E. F. (1950), Biochem. J., 46, 91 17 Salton, M. R. J. (1952), Biochim. Biophys. Acta, 8, 510

and Home, R. W. (195 1), Biochim. Biophys. Acta, 7, 19,

177

18. Taylor, E. S. (1947), J. gen. Microbiol, i, 86

19. (1949), y. gen. Microbiol, 3, 211

20. Ussing, H. H. (1949), Physiol. Rev., 29, 127

21. The Nature of the Bacterial Surface: Soc. Gen. Microbiol.

Symp., ed. Miles, A. A. and Pirie, N. W., Blackwell, Oxford, England

CHAPTER VII

PEPTIDES AND PROTEINS

Modern concepts of protein structure are founded mainly on the results obtained by subjecting proteins of animal ori- gin to procedures involving partial and complete hydrolysis, amino-acid analysis, ultracentrifugation, electrophoresis and X-ray diffraction analysis. From the limited data available there is, however, no reason to believe that the structure and general physical properties of the proteins of micro-organ- isms are in any way different from those of animals and plants. A few representative microbial proteins have been purified and in some cases crystallized, e.g. the enzymes alcohol dehydrogenase, catalase, amylase and an extracellular proteinase from Sac. cerevisiae, M. lysodeiktiais, B. suhtilis and Strep, haemolyticus respectively and the toxins of CI. botulinum and CI. tetani. Analyses of acid hydrolysates of such proteins and of whole cells by classical precipitation procedures, microbiological assay [41], the bacterial amino- acid decarboxylases (p. 27), and chromatography on ion exchange resins or paper have provided adequate evidence that the proteins of micro-organisms are composed of the L-stereoisomers of the same a-amino-acids as are those of more complex multicellular organisms.

Whilst there is no conclusive proof that D-amino-acids are constituents of proteins and it is probable that any found in acid or alkaline hydrolysates have arisen by racemization, they are not metabolically inert, and indeed several of them can be utilized by many micro-organisms [36]. The first step in their metabolism may involve deamination by a D-amino- acid oxidase (p. 11): alternatively, direct conversion to the corresponding L-isomer may be accomplished by a racemase, and at the present time two are known, specific for alanine [46] and glutamic acid [31] respectively. Both possess a pros- thetic group of py ridoxal phosphate and the alanine racemase

95

g6 NITROGEN METABOLISM

is widely distributed [46]. Several bacterial peptides, especi- ally those secreted into the environment, contain D-amino- acids and, in addition, amino-acids not yet found in peptide combination in animals and plants (Table 7.1). Whether these unusual amino-acids are also present in bacterial pro- teins remains to be proved. On the basis of their biological activity, the peptides associated with the metabolism of micro-organisms can be arranged into three groups:

(i) Peptides which are or may be co-factors in inter- mediary metabolism, e.g. glutathione, the folic acid factors, biocytin (p. 24), glutamine and aspara- gine. Though the latter three compounds are not true peptides, they can be regarded as containing the 'peptidic bond', — CO — NH— [25].

(ii) Peptides which serve as a source of amino-acids essential for growth. In natural environments, these peptides arise as the result of autolysis and the action of extracellular proteases (pp. 11 5-1 6).

(iii) Extracellular peptides. Among the bacteria, members of the Bacillaceae are especially active in the forma- tion of this type of peptide, the majority of which possess antibiotic activity, an exception being the polypeptides of D-glutamic acid which may be attached to the organism in the form of a capsule {B. anthracis and B. mesentericus) or free in the medium {B. suhtilis and B. mesentericus) [2]. The capsular material of B. anthracis consists of chains of glutamic acid residues linked by gamma peptide bonds together with chains linked by alpha peptide bonds. Such a capsule may confer immunity from attack by proteases of the infected host [15].

Co-factors containing peptide bonds

The available evidence indicates that glutamine and as- paragine play important roles in intermediary metabolism and it is to be noted that several co-factors are known which, like glutamine, contain the y-glutamyl radical. Glutamine may be essential for initiating the growth of certain organ- isms, but in many instances it is replaceable by glutamic

<

z >

o .t;

> o

.iri a •J "-l-l

-S ?• »

.s !->

g

Is o o

^'^

Ti • - c t5

sill

S-g 8 8 "g,

&.. aa«-2

o a

Be

08 O

X *' X r 4)

r; O U iJ a ^ (J o ^ —'

•^^ I II

+ +

00

Q C

,^ <U *^

^ .t: •- 3 c S

.S :^

i2 1 o

c; ° H

g 22 c«

.B "^

•r O iS

a-S

■t=Oc

= "- la g..

°^:^'8^.§|:B|

» o ^TS ^.t;

.Sfi-5

. u .^

:H t^ a^ 5^

a, -^

^ at: a2 au Uxw

a a a c a^ 2 5i -^

O O O 3 <U

a a a c a

I I

O O

o o o

+

o

=5 -:^

9 R -2

>• TJ

I I

o o

=2 >>

: <^ a

= s 2

<3 a Dqoq

Is

V (3

CO CO

'^ ^ (^

O^ O ?

O 4>

.2 S !=

C/2

c ^ S

is

^ U3 C

1 1

.2 S

98 NITROGEN METABOLISM

acid, though usually only if much larger amounts are sup- plied. When small inocula are used, the glutamine require- ments of Strep, haemolyticus (groups A and C) are absolute and cannot be replaced in this manner [10]. Streptococci decompose glutamine to glutamic acid and ammonia only when they are fermenting glucose, and the presence of small amounts of glutamine was observed to stimulate the fer- mentation of glucose by washed cell suspensions of all the streptococci examined, irrespective of their glutamine re- quirements during growth [27]. This stimulation was far greater than that produced by an equivalent amount of ammonium glutamate, and the more dilute the suspension, the greater the stimulation. Glutamine here appears to func- tion by restoring the intracellular concentration of a dif- fusible co-factor to an optimal value. Similar results were later obtained with Ln. mesenteroides, Lb. arabinosus [45] and CI. tetani [24]. Unlike the streptococci, Pr. morganii [26], Esch. colt, and CI. welchii [18] are able to hydrolyse the amide group of glutamine in the absence of glucose fermentation. Asparaginase — the enzyme system catalysing the hydrolysis of the amide group of asparagine — is widely distributed in fungi, yeasts and bacteria [cf. 50], and for this reason aspara- gine is frequently incorporated in media as a convenient source of readily available carbon and nitrogen. Asparaginase has been found, for example, in autolysates of P^. pyocyanea, Esch. colt, B. siibtiUs and Pr. vulgaris. Asparagine is an essential nutrilite for some strains of Ln. mesenteroides and Strep, lactis.

Glutamic acid is also a constituent of the folic acid factors, substances essential for the growth of Strep, faecalis R and Lb. casei and of key importance in the metabolism of all organisms. These factors contain a pterin linked to ^-amino- benzoic acid (PAB) which is in turn coupled through the amino group to one or more residues of glutamic acid (one in synthetic folic acid, three in the fermentation Lb. casei factor and seven in vitamin B^ conjugate). The linkages be- tween PAB and glutamic acid and between the various glut- amic acid molecules probably involve the y-carboxyl groups of the amino-acid. Whilst the enzyme systems in which folic

PEPTIDES AND PROTEINS 99

acid participates have yet to be isolated, there are good reasons for beUeving that it is impUcated in the synthesis ofcertainamino-acids, purines and pyrimidines (pp. 146-51). Glutathione (GSH), the first peptide to be assigned the function of a co-factor in intermediary metabolism, was dis- covered and isolated by Hopkins from yeast and various animal tissues. Little is known about the distribution of this peptide in bacteria [4] and its isolation from these organisms has not yet been reported. The chemical synthesis of GSH by Harington and Mead provided conclusive proof that it was y-glutamylcysteinylglycine. Ever since GSH was known to contain a thiol group, it has been postulated that GSH entered into cellular oxido-reduction reactions:

2GSH —^ OS— SG + 2H

In the presence of glutathione reductase, an enzyme found in yeast, plants and animals, oxidized glutathione will accept hydrogen from reduced TPN but not DPN: the reverse re- action has not yet been demonstrated [6]. The activity of many enzymes is inhibited by substances which react with or oxidize thiol groups, and it is therefore possible that GSH is part of the mechanism whereby these enzymes are main- tained in or brought into an active state in vivo. Glutathione takes an active part [34] in the conversion of methylglyoxal to lactic acid, a reaction catalysed by the enzyme system glyoxalase, found for example in Esch. colt and Sac. cere- visiae. Racker and Krimsky [35] have shown that GSH is tightly bound to the enzyme triosephosphate dehydrogenase, and they suggest that a thiol ester of 3-phosphoglyceric acid is an essential intermediate stage in the formation of 1:3- diphosphoglyceric acid (cf. the role of Co. A in the pyruvic oxidase system). Certain reactions in which GSH partici- pates as a substrate rather than as a co-factor have recently aroused great interest because of their potential significance in the synthesis of peptides and proteins in vivo. Working with cell-free preparations of sheep kidney, Hanes, Hird and Isherwood [16] have demonstrated that the y-glutamyl group of GSH and other y-glutamyl peptides (but not glutamine) can be transferred to peptides or to amino-acids,

lOO NITROGEN METABOLISM ■

i.e. the carboxyl moiety of a peptide bond can be transferred to a suitable amino acceptor. Thus incubation of GSH with phenylalanine or tyrosine resulted in the formation of y- glutamylphenylalanine and y-glutamyltyrosine respectively. Evidence that new peptides had been synthesized was first obtained by paper chromatography, and some of these com- pounds have now been isolated and characterized. The term transpeptidation has been applied to such transfer reactions and similar results were later obtained with Pr. vulgaris [37], Hanes et al. noted that prolonged incubation tended to pro- duce complete hydrolysis of all the peptides in the experi- mental system: it is therefore possible that these transpep- tidation reactions are catalysed by the intracellular proteases and indeed various proteases, like several other hydrol)rtic enzymes, are known to be capable of performing transfer reactions (pp. 104-5).

Utilization of peptides by micro-organisms

Information concerning the utilization of peptides comes mainly from the response of nutritionally exacting organisms to peptides containing an amino-acid essential for growth. The majority of these studies have been performed with non-proteolytic species and provide evidence that hydro- lysis by extracellular proteases is not an obligatory step in the utilization of simple peptides. Using four mutants of Esch. coli, exacting towards phenylalanine, tyrosine, proline and leucine respectively, Fruton and Simmonds compared growth in the presence of simple dipeptides containing the required amino-acid with that in the presence of the free acid [12]. Peptides of phenylalanine or tyrosine were as eifective as equimolecular amounts of the uncombined acids and there was little or no difference in the growth curves. It was concluded that prior to utilization, these peptides were hydrolysed by intracellular peptidases at a rate which did not limit growth. Lactobacillus arabinosus and Ln. mesen- teroides likewise utilize dipeptides of glutamine as readily as free glutamine or glutamic acid [45]. The leucineless Esch. coli mutant grew at the expense of peptides containing leucine, but although the rate of grov^^h in the log phase and

PEPTIDES AND PROTEINS lOI

the total amount of growth were equivalent to those on un- combined leucine, the length of the lag period was increased in proportion to the concentration of the peptide in the medium. This effect was not due to time being required for the formation of an adaptive enzyme. In experiments with Lactobacillus delbruckii, Lb. casei and Strep, faecalis, other workers have observed that the utilization of di- and tri- peptides serving as sources of valine or leucine was affected by the position of the amino-acid in the peptide, the nature of adjacent amino-acids, and in some instances the composi- tion of the medium [i, 23]. Unlike the examples described so far, the total amount of growth of the proline requiring Esch. coll mutant was greater when the amino-acid was sup- plied in the form of a dipeptide, yet the rate of growth was not affected. A possible explanation is that enzymes which decompose amino-acids are unable to attack those bound in peptides, consequently if as a result of peptidase activity an amino-acid gradually becomes available over a period of time, a greater proportion will be used in anabolic systems than if it is all initially present in the free state. Two other observations support such a conclusion. With Strep, faecalis , an organism with an active arginine dihydrolase system (p. 26), the same amount of growth was produced by appre- ciably less arginine when it was supplied in the form of small peptides. Similarly if the organism developed an active tyro- sine decarboxylase, dipeptides of tyrosine evoked greater growth than an equivalent amount of free tyrosine [20]. In a medium containing D-alanine in place of pyridoxin. Lb. casei becomes exacting towards dipeptides containing L- alanine, because the D-isomer inhibits the normal utilization of uncombined L-alanine, another essential nutrilite. High concentrations of glycine prevent the utilization of D-alanine, but in the presence of pyridoxin, neither glycine nor D-ala- nine is inhibitory and there is no requirement for L-alanyl peptides [21].

Unidentified growth factors believed to be peptides

The growth of several nutritionally exacting bacteria appears to be dependent on, or is stimulated by, unidentified

102 NITROGEN METABOLISM

substances which are beHeved to be peptides. Before at- tempting to decide whether such requirements are absolute it should be noted that in many instances the nature of the response is determined by the composition of the medium and the period for which the cultures are incubated [22]. Strepogenin is the name given to acid-labile material pre- sent in enzymic digests of proteins and required for the growth of certain streptococci and Lh. casei in synthetic media. The strepogenin activity of protein digests cannot be explained solely in terms of their content of glutamine or asparagine since neither of these substances replaced the digest factor for a strain of Strep, faecalis [48] and the res- ponse of Lb. casei to glutamine was different from that to strepogenin [47]. Moreover, the activity of both glutamine and GSH, unlike strepogenin, was destroyed by autoclaving. From their experiments with digests of crystalline insulin, Sprince and Woolley concluded that glutamic acid and glycine are two components of strepogenin, and of a large number of synthetic peptides, only tripeptides exhibited any activity, serylglycylglutamic acid being the most effective, though none was as active as the protein digest factor(s). Strepogenin was antagonized by peptides containing as- partic acid and also by lycomarasmin, a peptide secreted by Fusarium ly coper sici and responsible for the wilting of tomato plants. Lycomarasmin is composed of asparagine, glycine and a-hydroxyalanine, with the two latter sharing a common nitrogen atom, but the detailed structure is not known. From these various experiments, it seems likely that strepogenin contains a y-glutamyl residue and perhaps func- tions as a stable source of such groups for the synthesis of various co-factors. It is worthy of note that organisms res- ponding to strepogenin also readily decompose glutamine, and this may be another example of an essential nutrient being utilized more efficiently when it is supplied in the form of a peptide. The connection between strepogenin and the dicarboxylic amino-acids and their amides is, however, far from clear [cf. 42], and it is not yet possible to account for all the experimental results. Gravis and intermedius strains of Corynehacterium diphtheriae also require peptide

PEPTIDES AND PROTEINS I03

growth factors of unknown structure, and some similarities with strepogenin are indicated [5].

Synthesis of peptides and proteins

Owing to the paucity of available information the syn- thesis of peptides and proteins can only be discussed in general terms. Several suggestions have been made as to the mode of formation of peptide bonds and the three most likely mechanisms are those concerned with (i) the direct utilization of metabolic energy, (ii) transfer reactions and (iii) the reversal of proteolysis.

Calculations based on the synthesis of dipeptides in water and on the hydrolysis of dipeptides to ionic products have shown that the free energy associated with the peptide bond is in the range 420 to 3,000 calories [8, 25]. Though the pre- cise amount may be the subject of dispute, it is clear that the de novo synthesis of a peptide bond is endergonic, and it is reasonable to suggest that in biological systems ATP functions as a source of energy for the synthesis of peptide bonds. Glutathione, glutamine, acetylsulphanilamide and hippuric acid are all simple compounds containing peptide or peptidic bonds and the synthesis of each of these sub- stances was first observed in actively respiring preparations of animal tissues. Any condition inhibiting respiration and in consequence the production of energy, also inhibited synthesis. The recognition of ATP as a biological carrier of energy enabled the experimental systems to be greatly sim- plified and studies with non-respiring cell-free preparations revealed that synthesis was dependent on the presence of ATP. It was observed that a new co-factor, termed Coen- zyme A, played an important role in the synthesis of acetyl- sulphanilamide. Coenzyme A is now known to contain /5-mercaptoethylamine and to function in vivo as a carrier of acetyl and other acyl groups by virtue of its ability to form thiol esters. Studies with cell-free systems have shown that in addition to being synthesized from acetate in the presence of ATP, acetyl-Co.A can also be formed directly from an acetyl donor of suitable potential without the intervention of ATP, e.g. from pyruvate by the pyruvic oxidase system

104 NITROGEN METABOLISM

1

[see 32]. Dried cell preparations of Clostridium kluyveri cata- lyse the acetylatioh of amino-acids by acetylphosphate, a reaction which, although analogous to the synthesis of acetylsulphanilamide, only occurs in the presence of o-i m. cyanide. Acetylated amino-acids contain a peptidic bond, and it is feasible that peptides can be formed by transfer reactions in which the acetyl group is replaced by an amino- acid [40]. Cell-free extracts of Staph, aureus [9] and a number of other bacteria [13] and Sac. cerevisiae catalyse the syn- thesis of glutamine from glutamic acid and ammonia in the presence of ATP and Mg"^"^ or Mn"^"^. For each mole of amide synthesized, one mole of inorganic orthophosphate is liberated, and if ammonia is replaced by NHgOH the pro- duct is y-glutamylhydroxamic acid. By analogy with the role of Co. A as a carrier of acetyl groups in the synthesis of acetylsulphanilamide, it is tempting to suggest that glut- amine synthesis involves the formation of the corresponding y-glutamyl compound, but all attempts to obtain supporting evidence have failed. Though both glutamine and gluta- thione contain the y-glutamyl radical, there is no proof that their synthesis involves a common enzyme system or that the former participates in the synthesis of the latter. When incubated with ATP, K"^, Mg"^"*", phosphate and hexose diphosphate, cell-free extracts oiEsch. co/z synthesized GSH from glutamic acid, glycine and cysteine [37]. Whilst earlier and similar experiments with preparations of rat liver showed that the enzymes which synthesized GSH are dis- tinct from those catalysing hydrolysis, the individual steps in the biosynthesis are not yet known; the first one may be the formation of y-glutamylcysteine [38].

In addition to synthesis at the direct expense of metabolic energy, new peptide bonds may be formed by transfer (transpeptidation) reactions of the type:

XCO.NH.R+NH2R1 ^ X.CO.NH.Ri+NHoR

where XCOOH, R.NHg and R^.NHa represent amino-acids or peptides. Such reactions do not result in an overall in- crease in the number of peptide bonds, and since the type

PEPTIDES AND PROTEINS IO5

and number of bonds in the products is the same as in the reactants, they proceed with Httle overall change in free energy and are therefore independent of the availability of metabolic energy. Several typical proteol}1;ic enzymes are known to catalyse transpeptidation reactions, and further- more, to lengthen a peptide chain by the direct coupling of peptides. Thus, with chymotrypsin [ii]:

benzoyl-L-tyrosinamide+glycinamide ^

benzoyl-L-tyrosylglycinamide+NHs benzoyl-L-tyrosine+glycinamide ?^

benzoy 1-L-ty rosylgly cinamide + H 2O

It is therefore possible, as suggested many years ago, that the action of the proteases is reversible and that in the appropriate conditions they catalyse the synthesis and not the hydrol^'sis of peptides and proteins. Transpeptidation reactions involving GSH have already been described (p. 99). Because cysteinylglycine is readily hydrolysed by cellu- lar enzymes and yet is relatively stable when combined, as in glutathione, Hanes and his colleagues proposed that the attachment of a y-glutamyl radical to a peptide confers re- sistance to hydrolysis by intracellular proteases, and in con- sequence, synthesis is favoured and the peptide chain can be gradually lengthened by successive transfer reactions. Several bacteria catalyse exchange reactions between the amide group of asparagine or glutamine and hydroxylamine or isotopically labelled NHt: with NHgOH, such reactions lead to the formation of aspartyl- and glutamyl-hydroxamic acid respectively [45]. Whether the amide group can likewise be replaced by an amino-acid or peptide is not known. Pro- tein synthesis may therefore be visualized as a stepwise process beginning with glutathione, and possibly glutamine, as a source of peptide bonds synthesized at the expense of energy derived from ATP, and by means of transpeptidation reactions the amino-acid components of such bonds are subsequently altered so as to form peptides from which specific proteins are synthesized by further transfer and coupling reactions [11].

I06 NITROGEN METABOLISM

The mechanisms operative in the formation of peptide bonds are only one aspect of protein synthesis; it is also necessary to consider (i) the means whereby the correct sequence of amino-acids is attained in a peptide chain, (ii) if the protein molecule comprises more than one peptide chain, how such chains are linked together and (iii) the spatial arrangement of the amino-acids and the peptide chains. Major advances in solving these problems await the determination of the structure of specific proteins (cf. Sanger's recent elucidation of the amino-acid sequence in the peptide chains of insulin). Investigations of protein syn- thesis in animal tissues have been mostly confined to an examination of the conditions in which isotopically labelled amino-acids are incorporated into material precipitated by trichloracetic acid, i.e. presumably bound in proteins or polypeptides. Such experiments have proved little except that incorporation is associated with the utilization of meta- bolic energy, and great care is required in making deduc- tions from the observed results [49]. A more direct approach is provided by studies of the synthesis of a specific protein, e.g. an enzyme whose activity can be estimated and used as an index of concentration. Evidence has gradually accumu- lated to the effect that at least in certain cases the adaptive formation of enzymes is the outcome of de novo protein syn- thesis rather than the mere subtle modification of existing proteins, i.e. enzymes or 'enzyme precursors' [30]. If this is true, then the formation of adaptive enzymes would appear to offer a most promising field for studying protein synthesis in micro-organisms. Concentrations of 2:4-dinitrophenol and azide which, although not affecting respiration and the fermentation of carbohydrates, inhibit the uptake of inor- ganic phosphate and in consequence the synthesis of energy- rich phosphate bonds, also inhibit adaptive enzyme forma- tion and the incorporation of isotopically labelled amino- acids into peptides and proteins.

Certain observations indicate that the assembling of the constituent amino-acids is a preliminary step in the synthesis of a protein. For example, the synthesis of the adaptive enzyme nitratase in washed cells of Esch. coli [33] and of

PEPTIDES AND PROTEINS ' IO7

amylase by pigeon pancrease is enhanced by the addition of amino-acids, and the greater the number of amino-acids, the greater their effect. Gale has recently studied the effect of other amino-acids, and of purines and pyrimidines on the absorption, accumulation and further metabolism of glu- tamic acid by Staph, aureus, and obtained evidence that an increase in cellular combined glutamic acid is indicative of the synthesis of new protein [14]. Protein synthesis only occurred when the cells were suspended in a medium which contained, in addition to glutamic acid and glucose, all the amino-acids to which Staph, aureus is exacting.

Studies of actively dividing embryonic cells and cells engaged in rapid protein synthesis led Caspersson and inde- pendently Brachet to propose that protein synthesis is pre- ceded by the synthesis of pentose nucleic acids and that these substances then participate in and control the synthesis of proteins. By using the ultraviolet light microscope tech- nique (p. 130), Malmgren and Heden measured the nucleic acid content of cells at various stages during the growth of cultures of Esch. coli and Bacillus cereus. Their results indi- cated that the lag phase was a period of intense nucleic acid synthesis and in consequence the cellular concentration of nucleic acid reached a maximum during the early part of the lag phase: thereafter it gradually declined and became mini- mal during the stationary phase. Malmgren and Heden con- cluded that Caspersson's and Brachet's hypothesis also applied to bacteria and that a culture only passed out of the lag phase when a critical intracellular concentration of nucleic acid has been attained [28]. Other workers using Staph, aureus have provided further evidence in support of this conclusion [14, 29]. The rate of protein synthesis by washed cells of Staph, aureus can be directly correlated with the nucleic acid content of the cells at the time of harvesting [14]. During the growth of bacterial cultures it is the pentose nucleic acid content of the cell which alters: the desoxy- pentose nucleic acid content remains approximately con- stant [29]. It is interesting to note that after irradiation with ultraviolet light, bacteria are unable to develop the usual adaptive increase in activity when they are incubated in the

I08 NITROGEN METABOLISM

presence of the specific substrate [44], and that the action spectrum for light of different wavelengths resembles the absorption spectra of the nucleic acids [43].

Such experiments have naturally focused attention on the possible role of the nucleic acids in protein synthesis and so far two theories have been proposed: one suggests that the bond energy of the nucleic acid phosphate groups is used for the synthesis of peptide bonds, whilst the other regards the nucleic acids as being the fundamental components of the organized systems controlling the sequence in which amino-acids are joined together. Proteases appear to possess well-defined specificity with regard to the peptide bonds they attack (p. 113), and if they can in fact function syntheti- cally the same specificity is to be expected in the reverse reactions. If such enzymes can be organized so that they act in a predetermined sequence, a mechanism can be envisaged which possesses the ability to synthesize a peptide chain with the required serial arrangement of amino-acid residues. Since nucleic acids differ in their composition and structure and readily form complexes with proteins, it is possible that different nucleic acids combine specifically with different proteins. Hence an organized system of nucleic acids may provide a framework on to which enzymes are adsorbed in a particular order, with the result that they then direct the synthesis of a specific substance [17]. It is also feasible that the nucleic acids are structures on to which amino-acids rather than enzymes are adsorbed, and that differences in nucleic acid structure give rise to different sequences of amino-acids [3]. If there is any truth in such speculations, it is to be expected that disruption of nucleic acid meta- bolism will immediately result in the cessation of protein synthesis. It could therefore be argued that the effects of ultraviolet light described above are due to disruption or degradation of the organized nucleic acid systems directing adaptive enzyme synthesis. Bacteriophage are composed of nucleoprotein and they can infect and reproduce in irradi- ated cells of Esch. coli, although such cells are unable to form the adaptive enzyme ^^-galatosidase [19], and this may be construed to mean that irradiated cells can still synthesize

PEPTIDES AND PROTEINS I09

proteins provided the appropriate nucleic acid framework is made available.

By using materials enriched with radio-active phosphorus (P^2), several workers have measured the turnover rate of nucleic acid phosphate groups and the effect thereon of variations in the rate of protein synthesis. All the results reported so far are based on isotopic analyses of impure preparations of nucleic acid and must therefore be treated with caution since Davidson and his colleagues have shown that unequivocal results are obtained only if these substances are rigorously freed from contaminating materials before de- termining their P^^ content [7]. Spiegelman and Kamen [39] have found that the fermentation of glucose by washed sus- pensions of yeast previously grown in the presence of in- organic phosphate enriched with P^^ did not cause any decrease in the concentration of P^^ in the nucleic acid fraction of the cells. However, if NHt was added to the system, protein synthesis and budding occurred and the P^^ content of the nucleic acids decreased. A similar decrease took place during the adaptive formation of maltase. The proposal was therefore advanced that the energy in the phos- phate bonds of the nucleic acid could be utilized in transfer reactions for the synthesis of peptide bonds.

From this brief survey it will be apparent that whilst there is some experimental evidence that nucleic acids mediate protein synthesis, our present conceptions of the cellular organization controlling and bringing about such syntheses are purely speculative (see [51] for a critical analysis of pos- sible mechanisms for the synthesis of proteins).

REFERENCES

1. Agren, G. (1948), Acta chem. Scand., 2, 611

2. Bovarnick, M. R, (1942),^. biol. Chem., 145, 415

3. Caldwell, P. C. and Hinshelwood, C. (1950), J. chem. Soc,

3156

4. Callow, A. B. and Robinson, M. E. (1925), Btochem. y., 19, 19

5. Chattaway, F. W., Dolby, D. E., Hall, D. A. and Happold,

F. C. (1949), Biochem. J., 45, 592

6. Conn, E. E. and Vennesland, B. (195 1), Nature^ 167, 976

no NITROGEN METABOLISM

7. Davidson, J. N. and Smellie, R. M. S. (1952), Biochem.J., 52,

594

8. Dobry, A., Fruton, J. S. and Sturtevant, J. M. (1952), J. hiol.

Chem., 195, 149

9. Elliott, W. H. and Gale, E. F. (1948), Nature, 161, 129

10. Fildes, P. and Gladstone, G. P. (1939), Brit. jf. exp. Path., 20,

334

11. Fruton, J. S., Johnston, R. B. and Fried, M. (195 1), jf. bioL

Chem., 190, 39

12. and Simmonds, S. (1949), Cold Spr. Harb. Symp., 14, 55

13. Fry, B. A. (1954). In the press

14. Gale, E. F. and Folkes, J. P. (1953), Biochem. J., 53, 483

15. Hanby, W. E. and Rydon, H. N. (1946), Biochem. J., 40, 297

16. Hanes, C. S., Hird, F. J. R. and Isherwood, F. A. (1952),

Biochem. J., 51, 25

17. Hokin, L. E. (1952), Biochim. Biophys. Acta, 8, 225

18. Hughes, D. E. (1949), Biochem. J., 45, 325

19. Jacob, F., Torriani, A. N., and Monod, J. (195 1), C.R. Acad.

Sci., Paris, 233, 1230

20. Kihara, H., Klatt, O. A. and Snell, E. E. (1952), J. biol. Chem.,

197, 801

21. and Snell, E. C. (1952), jf. biol. Chem., 197, 791

22. Kodicek, E. and Mistry, S. P. (1952), Biochem.J., 51, 108

23. Krehl, W. A. and Fruton, J. S. (1948), J. biol Chem., 173, 479

24. Lerner, E. M. and Mueller, J. H. (1949), J- biol. Chem., 181, 43

25. Lipmann, F. (1949), Fed. Proc, 8, 597

26. Mcllwain, H. (1948), jf. gen. Microbiol., 2, 186

27. Roper, J. A. and Hughes, D. E. (1948), Biochem. J., 42,

492

28. Malmgren, B. and Heden, C. (1947), Acta path, microbiol.

Scand., 24, 472

29. Mitchell, P. and Moyle, J. (195 1), J. gen. Microbiol, 5, 421

30. Monod, J. and Cohn, M. (1952), Advances in Enzymology, 13,

67

31. Narrod, S. A. and Wood, W. A. (1952), Arch. Biochem., 35, 462

32. Ochoa, S. and Stern, J. R. (1952), A?in. Rev. Biochem., 21, 547

33. Pollock, M. R. and Wainwright, S. D. (1948), Brit. J. exp.

Path., 29, 223

34. Racker, E. (1951),^. biol. Chem., 190, 685

35. and Krimsky, I. (1952), Nature, 169, 1043

36. Rydon, N. N. (1948), Bioche?n. Soc. Symp., i, 40

37. Samuels, P. J. (1952), 2nd Internat. Biochem. Congress:

Abstracts, p. 93

38. Snoke, J. E. and Bloch, K. (1952), ^ biol. Chem., 199, 407

39. Spiegelman, S. and Kamen, M. D. (1946), Science, 104, 581

40. Stadtman, E. R., Katz, J. and Barker, H. A. (1952), jf. biol.

Chem., 195, 779

41. Stokes, J. L. and Gunness, M. (1946), J. Bact., 52, 195

PEPTIDES AND PROTEINS III

42. Stokes, J. L., Koditschek, L. K., Rickes, E. L. and Wood,

T. R. (1949), J. biol. Chem., 178, 93

43. Swenson, P. A. (1950), Proc. Nat. Acad. Sci., 36, 699

44. and Giese, A. C. (1950), J. cell. comp. Physiol., 36, 369

45. Waelsch, H. (1952), Advances in Enzymology, 13, 237

46. Wood, W. A. and Gunsalus, I. C. (1951), J. biol. Chevi., 190,

403

47. Woolley, D. W. (1948), J^. hiol. Chem., 172, 71; 176, 1291, 1299

48. Wright, L. D. and Skeggs, H. R. (1944), J- Bad., 48, 117

49. Zamecnik, P. C. and Frantz, I. D. (1949), Cold Spr. Harb.

Symp., 14, 199

50. Zittle, C. A. (195 1), The Enzymes, i (ii), Chap. 26

51. The Nature of Virus Multiplication, Soc. Gen, Microbiol.

Symp., ed. Fildes, P. and van Heyningen, W. E., Cambridge University Press, G.B.

CHAPTER VIII

PROTEOLYTIC ENZYMES

Putrefaction of the remains of dead animals and plants in natural environments is due in part to autolysis, i.e. to dis- integration of cellular components by the organism's own enzymes [27], and in part to the activities of the mixed population of micro-organisms which rapidly becomes established in such conditions. Because of their insolubility or molecular size, many cellular materials must be degraded to simpler substances before they can be utilized by micro- organisms as sources of carbon, nitrogen and energy. De- gradation may be accomplished either by enzymes present in the cell wall of the micro-organism, in which case the organism must itself come into physical contact with the substrate (e.g. the digestion of cellulose by species of Cyto- phaga), or by extracellular enzymes produced by the micro- organism yet acting independently of the parent cell. One group of extracellular enzymes especially important in putrefaction is the proteases, enzymes which hydrolyse peptide bonds and thus make available the amino-acid con- stituents of proteins.

For several years bacteriologists have used physical mani- festations of proteolytic activity in the classification and identification of micro-organisms, e.g. the liquefaction of gelatin or the clotting and digestion of milk. Moreover, the unwelcome effects of some pathogenic bacteria are now known to be due to extracellular toxins which possess en- zymic activity against certain proteins in the susceptible host. Mainly because of the influence of industry and medi- cine, the proteolytic activity of bacteria has been studied almost exclusively in terms of natural substrates such as casein, collagen (or gelatin) and fibrin, and more from the viewpoint of the bacteriologist rather than the biochemist. On the other hand, there have been considerable advances in knowledge with regard to animal proteases, several of

112

PROTEOLYTIC ENZYMES II3

which have been crystallized, and a summary of the more important results of this work provides a background against which the limited information concerning microbial pro- teases can be considered.

Hydrolysis of proteins and peptides by animal enzymes [36]

Until 1935 the animal proteases were classified according to molecular size of substrate and pH for optimum activity, i.e. the emphasis was on physical properties. Proteins were regarded as being the substrates of the proteinases (pepsin, chymotrypsin and trypsin), whilst peptides, substances with a relatively small number of peptide bonds, were the sub- strates of the peptidases (aminopolypeptidases, carboxypoly- peptidases and dipeptidases). The introduction by Berg- mann of a new and relatively simple method for the chemical synthesis of small peptides of known composition stimulated a detailed inquiry into the specificity of enzymes capable of hydrolysing peptide bonds. By using these synthetic sub- strates it was shown that the main factors affecting whether a proteolytic enzyme hydrolysed a peptide bond were, firstly, the nature of the amino-acids linked together by the bond, and secondly, the absence or presence of free amino or carboxyl groups in the vicinity of susceptible bonds. Bergmann proposed that the proteases be grouped into the endopeptidases and the exopeptidases according to whether they hydrolysed peptide bonds remote from or near to the ends of peptide chains in natural substrates. Endopeptidases were typified by pepsin, trypsin and chymotrypsin: the activity of pepsin and trypsin is inhibited by free amino groups near to susceptible bonds whilst chymotrypsin, and possibly trypsin as well, is inhibited by neighbouring car- boxyl groups. The exopeptidases comprise the dipeptidases and the amino- and carboxy-polypeptidases. The latter two groups of enzymes attack peptide bonds adjacent to, and in some cases penultimate to, terminal amino-acid residues with free amino groups and free carboxyl groups respec- tively. Whereas both proteins and peptides can serve as sub- strates for the endopeptidases, the exo_peptidases only attack peptides. The endopeptidases, unlike some of the

114 NITROGEN METABOLISM

exopeptidases, cannot hydrolyse bonds involving D-amino- acids. Bergmann and Fruton's investigations appeared to show that the specificity of the endopeptidases with regard to the amino-acid composition of the bonds they attacked was rela- tively high, and this was particularly so in the case of pepsin. But as more peptides have been synthesized and tested [see 2], it has become clear that these enzymes are not so specific as the results of the earlier work might suggest. Though few of them have yet been purified, it is evident that there are a number of dipeptidases, amino- and carboxy-polypepti- dases and it is possible that each one has different specificity requirements.

In addition to the extracellular enzymes, there are a num- ber of intracellular animal proteases known as kathepsins, of which four types are known, analogous to pepsin, trypsin, amino-polypeptidases and carboxy-polypeptidases respec- tively as regards their specificity. Unlike the extracellular enzymes, the kathepsins are only active in the presence of an activator such as HgS, cyanide, cysteine or glutathione, and in this respect they are very similar to some of the intracellular proteases of micro-organisms.

Great interest has been taken in the mammalian extra- cellular proteases known to be secreted in the form of an inactive precursor, a zymogen, the crystalline form of which is different from the corresponding active enzyme. Pepsino- gen is converted to pepsin by treatment with acid or with pepsin itself, whilst trypsinogen is activated by enterokinase or by trypsin. Activation of the precursor may involve the removal of a small peptide (as with pepsinogen) or the open- ing of a small number of peptide bonds (chymotrypsinogen and trypsinogen).

There is at present conflicting evidence as to the manner in which proteins and peptides are hydrolysed by the endo- peptidases to smaller units. The enzyme may attack all the susceptible peptide links in any one peptide chain simul- taneously and thus release the component amino-acids and peptides concurrently (the 'all or none' hypothesis). Alterna- tively, it may attack the peptides in a random manner, hydro- lysing only one bond at a time in any one peptide, until the

PROTEOLYTIC ENZYMES II5

system eventually contains no more susceptible bonds. Tiselius, on the basis of an electrophoretic analysis of the reaction mixture obtained by the treatment of egg-white with pepsin, supported the former concept [39]. However, chromatographic analysis has revealed that though the sys- tem ultimately contained mostly tripeptides, a few dipep- tides, no free amino-acids and still some undigested pro- tein, in the initial stages of the reaction, deca- and higher peptides were present [31].

Proteases of micro-organisms

Only a small number of the proteases of micro-organisms have been purified and only one has been crystallized [14]. Apart from showing that they produce certain effects, e.g. the liquefaction of collagen and gelatin, there have been few attempts to express activity in terms of the hydrolysis of peptide bonds, and virtually none concerned with their specificity with respect to the bonds attacked [32]. Most of the experiments with synthetic substrates have been con- fined to the peptidases and no bacterial proteinase has been examined in such detail as mammalian pepsin. In addition to intracellular proteases, some micro-organisms also pos- sess extracellular proteolytic enzymes and these enable pro- teins in the environment to be degraded at least to small peptides, if not to amino-acids, and the products are then absorbed and may be further degraded by intracellular enzymes.

The biological activity of extracellular substances pro- duced by micro-organisms can be detected by incorporating the appropriate substrate in the culture medium or, more usually, by testing the activity of the culture filtrate. The test may be applied directly to the latter, or to a concentrate of the active principle prepared from the culture filtrate by evaporation in vacuo^ by precipitation with ammonium sul- phate or ethanol, by freeze drying, or by an adsorption technique. Enzymic activity of the experimental material is detected by incubating with the appropriate substrate, toxic activity by injecting the material into animals or by pre- cipitation reactions with antisera, and antibiotic activity by

Il6 NITROGEN METABOLISM

incorporating the material into culture media. The occur- rence of enzymic activity in culture filtrates must not be taken as unequivocal evidence that the enzyme is extra- cellular, particularly if high activity is dependent on pro- longed incubation of the cultures: in such circumstances it is natural to suspect that the organisms may have undergone partial autolysis and thus released intracellular enzymes into the medium. If a culture filtrate contains more than one protease, they may be separated from one another by the usual procedures employed in the purification of enzymes. Very active extracellular proteases are produced by species of Proteus, Clostridium, Bacillus, Pseudomonas and Micro- coccus', less active enzymes are formed by the streptococci and staphylococci whilst the lactobacilli and Enterobacteri- aceae (with the exception of Proteus spp.) apparently pro- duce none.

Collagenase and gelatinase activity

The ability of certain bacteria to liquefy collagen and the material derived from it, gelatin, was soon discovered fol- lowing the introduction of gelatin as a means of solidifying culture media [12]. One of the methods employed in the quantitative determination of gelatinase activity makes use of an Ostwald viscometer [13], the method being based on the assumption that changes in the viscosity of a solution of gelatin are an index of proteolytic activity. Such a procedure is indirect, and the observed changes in viscosity may not necessarily be due to the hydrolysis of peptide bonds. It is therefore desirable that viscometric methods should be com- pared with those that are based more directly on the results of proteolytic activity, and in particular on the expected appearance of free amino and carboxyl groups in the system. Suitable methods of this type employ the van Slyke appar- atus for the determination of amino groups and the Sorensen titration procedure for carboxyl groups. Indeed, it is evident from the work of Gorini and Fromageot that changes in viscosity may bear no relation to the appearance of free amino groups (Fig. 8.1), but such a correlation has not often been attempted.

PROTEOLYTIC ENZYMES II7

The most detailed studies of the proteolytic enzymes in bacterial culture filtrates are those of Maschmann [29], who isolated from the culture filtrates of the invasive Clostridia {CI. histolyticum, CI. welchii, CI. septicum) an enzyme that attacked gelatin and collagen but had no action on casein, peptone or egg albumin. Even filtrates from young cultures

600-t

h^SOO-

5400-

0 300-

^lOO-

'l8hr.)

K

(18 hr.)

O

(20 min)

X

-24-0

-160 >.

«n O

■120 o> > Ui

> hQO P

/

cor

(Ihr.)

-40

I 1 \

^ 0-5 I'D , 1-5 20

ENZYME CONC. (ml. soln.)

FIG. 8.1. — Hydrolysis of gelatin by extracellular proteinase of Micrococcus lysodeikticus as a function of enzyme concentra- tion. Activity measured by appearance of amino nitrogen (solid line) and by changes in viscosity (broken line). Reaction times shown in parentheses. (From Gorini, L. and Fromageot, C. (1950), Biochim. Biophys. Acta, 5, 524; Elsevier Publishing Co. Inc., New York and Amsterdam)

contained this specific gelatinase, and it was therefore re- garded as being a truly extracellular enzyme. The invasive properties of certain Clostridia can be ascribed to their ability to secrete this enzyme since Maschmann could not detect any specific gelatinase in filtrates from the non-invasive Clostridia {CI. botidinum, CI. tetani), and later workers showed that the culture filtrates of the invasive Clostridia digested the collagen supporting material of muscle whereas an enzyme such as trypsin liquefied the muscle fibrils and

Il8 NITROGEN METABOLISM

left the connective tissue intact [34, 15]. The collagenase of CI. welchii type A, also known as the kappa (k) toxin, has been purified considerably [11]. Although on a weight basis this material was less toxic than the lecithinase (a-toxin) of CI. welchii, it was ten to fifty times more toxic than the endotoxins of Gram-negative bacteria. A curious feature of the purified collagenase is that after exposure to mild alkali or to heat, it still attacked gelatin but not collagen [9]. This treatment may have altered the structure of the enzyme such that it can combine only with gelatin, and this may also be the reason why there is a change in the optimum pH for gelatinase activity. Another possible, though perhaps un- likely, explanation is that exposure to heat or alkali de- natured the collagenase and at the same time activated a specific gelatinase precursor [9]. Some confusion exists in the naming of the enzymes which attack gelatin and collagen and it is sometimes assumed that the terms gelatinase and collagenase are synonymous. However, Evans and Wardlaw have evidence that though the formation of collagenase by various species of bacilli is always accompanied by gelatinase activity, some organisms, e.g. B. subtilis, produce a gelatinase and yet have no apparent action on collagen [16].

Other proteinases of the Clostridia

Apart from the specific gelatinase, Maschmann found three other types of proteolytic enzymes in various bacterial culture filtrates (Table 8.1). One type was active against pro- teins and peptones, and like the specific gelatinase, was found in young as well as old cultures and activity was unaffected by the presence of Og: gelatin was also digested, but usually not as rapidly as casein. The two other types of proteases, one active against proteins, the other against peptides, ap- peared in the medium after the cultures had stopped grow- ing, and it is therefore possible that at least in some instances they were intracellular enzymes released by autolysis. Their activity was depressed by O2 and was only maximal in the presence of reducing substances (HgS, cysteine or gluta- thione) whilst in addition, the peptidases also required di- valent cations (Mg+"^, Fe++ or Mn++). The Og-labile

PROTEOLYTIC ENZYMES

119

proteinases were found in cultures of CI. welchii, CI. histo- lyticum, CI. botulinum and CI. septicum; peptidases were present in cultures of all the organisms named in Table 8.1. There may be strain differences with regard to the enzymes formed by a particular species. For example, the proteinase of CI. histolyticum isolated by Maschmann w^as distinguished from that isolated by later workers in not being activated by cysteine. Further investigations revealed that some strains of CI. histolyticum produced both of these proteinases [23].

TABLE 8.1

PROTEOLYTIC ENZYMES OF ORGANISMS STUDIED BY MASCHMANN

Organism	Specific gelatinase	Proteases, attacking casein and peptone, inhibited by normal serum
Enzyme stable in oxygen	Enzymes stable in oxygen	Enzymes labile in oxygen
CI. botulinum CI. feseri CI. histolyticum CI. septicum CI. sporogenes CI. tetani CI. ivelchii Ps. aeruginosa Ps. fluorescens Serratia marcescens	+ + + +	+ 1 + 1 + + 1 + + +	+ 1 +++ 1 + 1 1 1

The one activated by Fe"^"^ and thiol compounds attacked casein and clupein as well as gelatin, and liberated more carboxyl groups than amino groups, indicating that peptide bonds involving the amino group of proline had been hydro- lysed [28]. The lambda (X) toxin of CI. welchii, purified by Bidwell, is a proteolytic enzyme capable of hydrolysing gelatin, casein, haemoglobin and hide pow^der, but it has no effect on native collagen [10]. It is strongly inhibited by cysteine, and, though there is no evidence that it is activated by Mg"^"*" or Mn"^"^, by citrate. 9

120 NITROGEN METABOLISM

Maschmann observed that the proteinases, but not the specific gelatinase, were inhibited by normal (i.e. not neces- sarily immune) sera. A more recent careful kinetic study of the proteolytic activity of several bacteria disproved the idea that this effect wsls due to the same substance that inhibits trypsin [13]. The trypsin inhibitor is found only in the albumin fraction of the serum proteins, whilst the labile antibacterial protease factor is in the globulin fraction: furthermore, the bacterial enzymes are not inhibited by the trypsin inhibitors present in soya bean, the pancreas and ovomucoid. The antibacterial serum factor inhibited all the bacterial proteases examined, even the specific gelatinases to a small extent. These investigations of Duthie and Lorenz also confirmed that the ability to clot milk is restricted to certain bacteria, and showed that although inhibited by the globulins of normal sera, the rate of clotting bore no relation- ship to the protease activity of the culture filtrates.

Factors affecting the formation and activity of extracellular proteinases The composition of the growth medium profoundly affects the degree of proteinase activity that is ultimately detectable in the culture filtrate. Three factors appear to be of special significance, (i) inorganic ions, (ii) fermentable carbohydrate and (iii) the organism's source of nitrogen. A previous observation [30] that protease production by a species of Proteus depended on the presence of Ca and Mg"^"^ was investigated in more detail by Hanes, using the organisms Bacillus subtilis, B. mesentericus, Pr. vulgaris, Ps. fiuorescens and Ps. aeruginosa [22]. The addition of Ca^"^ had little effect on grov^h, but markedly increased the pro- duction of gelatinase, whilst Mg"^"*", although increasing growth, caused no increase in the gelatinase activity of the cultures. The recent work of Gorini has served to emphasize the important role of cations such as Ca"^^ [17]. Optimal extracellular protease activity of Micrococcus lysodeikticus, B. megatherium, Proteus, Ps. pyocyanea and B. mesentericus was dependent on growth at a low temperature (26° C.) in

PROTEOLYTIC ENZYMES 121

well-aerated media in the presence of Ca"^"^. The amount of glucose added to the medium had to be such that the pH at the end of growth had not fallen below pH 7. Gorini con- tends that as well as activating the proteases, Ca"*" "*" also has a stabilizing influence and that in the absence of Ca"*"^ these enzymes are inactivated as fast as they appear in the medium [19]. This view is opposed to that of previous workers who believed that Ca"^ "*" stimulated the actual pro- duction of the enzymes. When proteins and polypeptides were used as a source of nitrogen, the growth of the culture was dependent on its proteolytic activity, and in such condi- tions Ca"*"^ was indispensable for growth [i8]. Sodium fluoride and citrate, substances capable of combining with Ca"^"^, inhibited these proteases, and although Mg"^"^ pro- tected them against such inhibitors and from denaturation by heat, Mg"^"^ could not replace Ca"^"^ as the cationic acti- vator [17]. There is some evidence that the gelatinase activity of cultures of B. suhtilis is dependent on the Mn"^"^ content of the medium [38].

Although several workers have regularly demonstrated activity in filtrates from cultures grown in media containing glucose [30], other workers have reported that the presence of fermentable carbohydrate inhibits the formation of pro- teolytic enzymes [3]. Such effects are probably to be attri- buted to the acidic products of carbohydrate catabolism causing the pH of the medium to fall to a value which does not favour the formation of proteases. In buffered media, or in those where the pH does not become acid [cf. 17, 19] the presence of fermentable carbohydrate has no effect [3].

Several attempts have been made to ascertain whether the growth of proteolytic bacteria is supported by pure native proteins, or only by those which have been denatured or partially degraded. Such bacteria failed to grow when sub- cultured into a medium containing inorganic salts and a pure protein as a source of carbon and nitrogen [i]. This might be explained on the basis that the synthesis and excretion of an extracellular enzyme involves the utilization of energy,

122 NITROGEN METABOLISM

which in these experiments could only be derived from amino-acids after they have been made available by enzymic hydrolysis of the protein. Even if any extracellular enzyme is carried over in the inoculum it is possible that it is diluted in the subculture to such an extent that its activity is no longer significant. If a small amount of peptone, presumably containing some small peptides or amino-acids, was added to the medium, certain proteolytic bacteria, e.g. Pr. vulgaris^ grew rapidly and crystalline egg albumin and serum pro- teins were then readily degraded and utilized [37]. Hence it appears that the formation of extracellular enzymes is de- pendent on the medium containing sources of carbon, nitrogen and energy that can be utilized immediately with- out having to be first broken down into smaller units by extracellular enzymes. In any event, proteoses and peptones are apparently not attacked by members of the Bacteriaceae or by Staph, aureus or Strep, faecalis [33].

Specific proteinases of the streptococci and staphylococci

The culture filtrates of some Lancefield group A strepto- cocci contain an Og-labile papain-like enzyme which apart from hydrolysing fibrin, casein, gelatin and benzoyl-L- arginamide, also attacked the M antigen, one of the antigens used in typing group A streptococci [14]. The M antigen is usually absent in those strains capable of producing this enzyme, though it may be present if the cultures are grown at a low temperature (22° C.). After passage through mice, formation of the enzyme ceased and the organisms became more virulent and possessed the M antigen. The latter two effects are not directly related, since the M antigen is also present in some avirulent strains. Of some clinical impor- tance is the fibrinolytic activity of haemolytic streptococci (groups A and C), staphylococci and gas-gangrene Clostridia. These bacteria produce an enzyme which converts a pre- cursor (plasminogen) in the globulin fraction of human sera into an active enzyme (plasmin) which digests the fibrin of clotted blood. The mode of action of the bacterial enzyme, named streptokinase in the case of streptococci, is compar- able with the activation of chymotrypsinogen by trypsin.

PROTEOLYTIC ENZYMES I23

Infection with Strep, haemolyticus soon results in the forma- tion of an antibody which completely antagonizes strepto- kinase.

Peptidases of bacteria, aspergilli and yeast [4, 25, 36]

A number of observations have provided evidence for the occurrence of peptidases in micro-organisms similar to those in animals and plants, and in general their activity has been studied using simple substrates, di- and tri-peptides of glycine, alanine and leucine. Many, but not all, of these enzymes are activated either by thiol compounds or by di- valent cations: some require both types of activator (Table 8.2). The peptidases of the Clostridia exhibit poor activity

TABLE 8.2

ACTIVATION OF MICROBIAL PEPTIDASES

The activators listed below are those which have been found to increase the activity of various preparations of peptidases from the organisms shown in the table. Which of these substances are the most effective activators for particular enzyme preparations de- pends on the peptidase concerned and sometimes on the substrate being tested.

Organism Activators

Aspergillus parasiticus Zn , cysteine

Bacillus megatherium Zn , Mn , Fe , cysteine

Clostridiu?n histolyticum Mn , Fe , cysteine

Escherichia coli Mg"^"*", Mn"^"^

Leuconostoc mesenteroides TjXi , Mn , Cd , Pb , cysteine

Phytomonas tumifaciens Mg"^"*") Mn , cysteine

Proteus vulgaris Mg""""*", Mn"*"*"

Pseudomonas fluorescens Mg > Mn "^

Saccharomyces cerevisiae Zn"*" , Mn , Fe , Cl~, Br", NO7

except in the presence of cysteine together with a cation such as Fe"*"^ or Mn"*"*" (concentrations of the order io~^ to io~^ M.). For the hydrolysis of leucyl peptides, the best acti- vator was Fe"*"^, for alanyl peptides, Mn"'"'' was better [29]. Maschmann suggested that the active enzyme was formed by combination of the cation wdth the reduced form of the

124 NITROGEN METABOLISM

'apoenzyme', and that the cation acted as a bridge joining the enzyme to the substrate. Smith [36] has proposed that the cation chelates with the substrate, and thus causes a redistribution of electrons with the result that the suscep- tible peptide bond becomes unstable and easily broken:

— NH.CH(R).C NH.CH(R')COO—

II O

e.g. for a carboxypeptidase,

Enzyme protein

The co-ordination is believed to be between the — CO — group of the peptide bond and the terminal free carboxyl group for carboxypeptidases and the — NH — group and the free amino group for aminopeptidases. Examples are known in which co-ordination between cations and peptides is extremely specific, and this may explain why peptidases attacking different substrates are activated by different metal ions. The aerobic bacteria Ps. aeruginosa, Ps. fluores- cens and Serratia marcescens contain a peptidase activated by Mg"*"^ and comparable in specificity with the leucine aminopolypeptidase of animal tissues [5, 7]. Dipeptidases, or enzymes with activity against dipeptides, have been found in preparations from Mycobacterium tuberculosis [35], Leuco- nostoc mesenteroides [6], Phytomonas tumifaciens [5], 5. mega- therium [8], Sac. cerevisiae [20, 21] and Aspergillus para- siticus [7, 26]. A polypeptidase from Sac. cerevisiae and Asp. parasiticus was activated by Zn"^"*", and both the di- and poly-peptidase of yeast also required chloride ions [24] (Table 8.2). In general, the optimum pH for peptidase activity is in the range 8-9, but some peptidases of Ln. mesenteroides, propionibacteria and lactobacilli are most active at an acid pH (5-5-6-0), and they are not activated by divalent cations [5, 6].

PROTEOLYTIC ENZYMES I25

REFERENCES

1. Bainbridge, F. A. (19 11), J. Hyg., 11, 341

2. Baker, L. E. (195 1), J. biol. Chem., 193, 809

3. Berman, N. and Rettger, L. F. (1918), J. Bact., 3, 389

4. Berger, J., Johnson, M. J. and Peterson, W. H. (1937). Enzymo-

logia, 4, 3 1

5. (1938), J. Bact., 36, 521

6. (1938), y. biol. Chem., 124, 395

7. (1939), J- biol. Chem., 130, 641

8. ■ (1940), J. biol. Chem., 133, 157, 639

9. Bidwell, E. (1949), Biochem. J., 44, 28

10. • (1950), Biochem. J., 46, 589

11. and van Heyningen, W. E. (1948), Biochetn. J., 42, 140

12. Brunton, T. L. and MacFadyen, A. (1889), Proc. Roy. Soc,

46B, 542

13. Duthie, E. S. and Lorenz, L. (1949), Biochem. jf., 44, 167, 173

14. Elliott, S. D. (1945), J- exp. Med., 81, 573; (1950), 92, 201

15. Evans, D. G. (1947), J. gen. Microbiol., i, 378

16. and Wardlaw, A. C. (1953), jf- gen. Microbiol., 8, 481

17. Gorini, L. (195 1), Biochim. Biophys. Acta, 6, 237

18. and Audrian, L. (195 1), Biochim. Biophys. Acta, 6, 477

19. and Fromageot, C. (1950), Biochim. Biophys. Acta, 5, 524

20. Grassmann, W. (1927), Z. physiol. Chem., 167, 202

21. and Dyckeroff, H. (1928), Z. physiol. Chem., 179, 41

22. Hanes, R. B. (1931), Biochem. J., 25, 1851; (1932), 26, 323;

(1933), 27» 466

23. van Heyningen, W. E. (1940), Bioche?n. jf., 34, 1540

24. Johnson, M. J. (1941), J. biol. Chem., 137, 575

25. and Berger, J. (1942), Advances in Enzymology, 2, 69

26. and Peterson, W. H, (1935), J. biol. Chem., 112, 25

27. Jones, A. S., Stacey, M. and Webb, M. (1949), Biochem.

Biophys. Acta, 3, 383

28. Kocholaty, W. and Krejci, L. E. (1948), Arch. Biochem., 18, i

29. Maschmann, E. (1943), Ergeb. Enzvmforsch., 9, 155

30. Merrill, A. T. and Mansfield-Clark, W. (1928), J. Bact., 15,

267

31. Moring-Claesson, I. (1948), Biochifn. Biophys. Acta, 2, 389

32. Mycek, M. J., Elliott, S. D. and Fruton, J.' S. (1952), J. biol.

Chem., 197, 637

33. Rettger, L. F., Berman, N. and Sturges, W. S. (1916),^ Bact.,

I, 15

34. Robb-Smith, A. H. T. (1945), Lancet, ii, 362

35. Roulet, F. and Zeller, E. A. (1948), Helv. Chim. Acta, 31, 191 5

36. Smith, E. L. in The Enzymes, 1 (ii), Chap. 23

37. Sperry, J. A. and Rettger, L. F. (i9i5),.7. biol. Chem., 20, 445

38. Stockton, J. R. and Wyss, O. (1946), ^ Bact., 52, 227

39. Tiselius, A. and Eriksson-Quensel, I. -B." (1939), Biochem. jf.,

33, 1752

CHAPTER IX

NUCLEOTIDES AND NUCLEIC ACIDS

It requires but a short acquaintance with biochemistry to reahze that the importance of nucleotides and nucleic acids is comparable with that of amino-acids and proteins. A nucleotide consists of a heterocyclic nitrogen compound, such as nicotinamide, a purine or a pyrimidine, joined to a sugar to which one or more orthophosphate groups are attached. The sugar is either a pentose (e.g. ribose) or a deoxypentose, and the dephosphorylated form of a nucleo- tide is known as a nucleoside. (The latter is really a N- glycoside and a nucleotide is therefore a phosphorylated N-glycoside.) Natural pyrimidines and purines include

	/ 4 \ N3 5CH			N3 5C
pyrimidine,	CH2	II 6CH	and	purine,	1 II CH2 6C ^n/\i^	sCH

cytosine (4-amino-2-ketopyrimidine), uracil (2:4-diketo- pyrimidine), thymine (5-methyluracil), adenine (4-amino- purine) and guanine (2-amino-4-ketopurine).^ The corre- sponding ribose nucleosides are known as cytidine, uridine, adenosine and guanosine, and the nucleotides as cytidylic, uridylic, adenylic and guanylic acid respectively. Plants and animals also contain 5-methylcytosine, but contrary to earlier reports, this pyrimidine is apparently not present in micro-organisms [51, 57].

Nucleotides are constituents of nucleic acids and also of the prosthetic groups and co-factors of many enzyme sys- tems concerned with transfer reactions (e.g. ATP in energy transfer; DPN, TPN, FMN and FADN in hydrogen transfer;

^ New system of numbering recommended by the Chemical Society {J. Chern. Soc, 5064 (1952) )•

126

NUCLEOTIDES I27

Co. A in acyl group transfer). The recently discovered nucleotide, uridine-diphosphate-glucose, is a co-factor in the enzymic conversion of galactose- 1 -phosphate to glucose- I -phosphate [9]. With the exception of FMN, in which the glycone is ribitol, all these co-factor nucleotides contain /i-ribose in the furanose form and phosphorylated on €'-5. In co-factors composed of two nucleotides the internucleo- tide bond is between the two 5-phosphate groups.

Nucleic acids are composed of large numbers of nucleo- tides and are consequently compounds of high molecular weight (of the order 5 x 10^ to i x lo^), but although the indi- vidual units are relatively simple, the determination of the detailed structure of a nucleic acid entails the solving of problems comparable in difficulty with those encountered in the elucidation of the structure of a protein (cf. p. 106). Depending on whether the component nucleotides contain either ribose or 2-deoxyribose, nucleic acids have been divided into two types, the ribose nucleic acids (RNA) and the deoxyribose nucleic acids (DNA). In only three instances has a derivative of ribose or 2-deoxyribose been isolated and characterized, and in the absence of such evidence some workers [14] prefer the terms pentose nucleic acid (PNA) and deoxypentose nucleic acid (DPNA). Adenine, guanine and cytosine are constituents of all the known nucleic acids, and in addition a PNA contains uracil, whereas a DPNA contains thymine. In consequence of their large numbers of phosphate radicals, nucleic acids are highly acidic and readily form salt-like compounds with bases. However, at least in bacteria, the proteins associated with nucleic acids in nucleoproteins are not necessarily of the basic protamine or histone type [15].

Structure of nucleotides and nucleic acids

Only in a few instances is there adequate proof of the structure of the component nucleotides of nucleic acids, but by analogy it is assumed that they all follow the same general pattern. The sugar is present as the ^-isomer and in the furanose form, with a glycosidic linkage between the reduc- ing group (C'-i) and N-7 of the purines and N-i of the

128 NITROGEN METABOLISM

pyrimidines. Position C'-3 is phosphorylated in the deoxy- pentose nucleotides, and though it is probably the same in the pentose nucleotides, position C'-2 is a possible alterna- tive. Since hydrolysis in certain conditions yielded four types of nucleotides, apparently in equimolecular proportions, all nucleic acids were at first thought to be polymers of units each of which contained the four nucleotides arranged in a straight chain or in a cyclic tetrad. An alternative suggestion was that the overall equivalent proportions did not neces- sarily imply such a regular arrangement but were due simply to a statistical mean. Such theories required that the ratio of purine-N to pyrimidine-N should be 2:1, but with the accumulation of precise quantitative data came the realiza- tion that few nucleic acids contained the four nucleotides, or even purines and pyrimidines, in equimolecular propor- tions. For example, three types of DPNA have been isolated; in the one found in animals, yeast and most bacteria, adenine and thymine predominate (AT type), in another, found in only a few bacteria, guanine and cytosine predominate (GC type), whereas in the third type isolated from strains of Esch. colt, the bases are in equimolecular proportions [13]. Although discovered over eighty years ago, detailed in- vestigations of structure have been confined to only two nucleic acids, a RNA from yeast and a DRNA from the thymus, and of these most is known about the former. Markham and Smith [38] believe that yeast RNA is a mix- ture of comparatively short, straight chains of nucleotides in which the internucleotide linkage is between the phos- phate at C'-3 (or C'-z) of one nucleotide and the hydroxyl group at C'-5 in the adjacent nucleotide. They have also presented evidence that some of the chains terminate in cyclic nucleotides, i.e. nucleotides in which the phosphate group forms a bridge between C'-2 and €'-3. By using ion exchange resins, previous workers had shown that each of the four nucleotides in alkaline hydrolysates of yeast RNA could be separated into two isomers, the 'a' and *b' nucleo- tides, which were regarded as being the nucleoside 2'- and 3 '-phosphates. It is now evident that only one of these isomers occurs naturally, and that the other is formed by

NUCLEOTIDES I29

the hydrolysis of cyclic nucleotides which are either present initially in the nucleic acid or are produced during the hydrolysis procedure [38]. The products obtained by the digestion of RNx\ with ribonuclease are mainly mono- and di-nucleotides together with some polynucleotide material which, because it w^ould not dialyse, was at first thought to be the 'core' of the nucleic acid and to be of high molecular weight. Markham and Smith have now shown that the 'core' consists of relatively small polynucleotides whose dialysis depends on the concentration of salt in the system [38].

Estimation of nucleic acids and their components

The procedures for estimating total nucleic acid are based on either ultraviolet spectrophotometry or on determinations of orthophosphate or sugar (pentose and deoxypentose). In the latter chemical methods the cells are first extracted with cold trichloracetic acid (TCA) and a fat solvent to remove acid-soluble compounds and phospholipoids. After the nu- cleic acids have been released from the nucleoproteins by heating the extracted cells with 5% TCA at 90° C. for 15 minutes, pentose is estimated colorimetrically by the orcinol method and deoxypentose by the Dische-diphenyl- amine method [48]. In the Schmidt and Thannhauser pro- cedure [47] the extracted cells are treated with N.-KOH at 37° C. to hydrolyse the PNA to free nucleotides, and then the undegraded DPNA and protein are precipitated by making the hydrolysate normal to HCl. The DPNA and PNA content of the original material is calculated from the organic and inorganic phosphate in the hydrolysate and the organic phosphate in the DPNA fraction. This method assumes that the only cellular acid-insoluble phosphorus compounds are nucleic acids and phosphoproteins — an assumption now known to be invalid [42]. In both the sugar and phosphate methods, the results are expressed in terms of nucleic acid by using conversion factors based on thymus DNA and yeast RNA as standards.

A characteristic property of purines and pyrimidines is that they strongly absorb in the ultraviolet region of the spectrum, with a peak absorption in the region of 260 m^«.

130 NITROGEN METABOLISM

It is therefore natural that spectrophotometric techniques should have been developed for the quantitative estimation of the free bases and the various substances in which they are constituents. The total nucleic acid content of organisms spread in a film on a slide can be determined by ultraviolet spectrophotometric microscopy [10], whilst for cell suspen- sions, there is a technically simpler procedure based on the use of a standard ultraviolet spectrophotometer [40]. By making various assumptions, an average value for the ab- sorption coefficient of a typical nucleic acid can be deduced and thus the absorption measurements interpreted in terms of nucleic acid. The originators of these techniques are well aware that reliable results are only obtained if due account is taken of a large number of variables and that many of the basic premises may not be strictly valid [cf. 10, 40]. Never- theless, even if the results are not entirely accurate in terms of precise quantitative values, they are extremely useful, particularly when the data are used in a comparative manner. The identification and quantitative determination of the various bases and nucleotides present in a nucleic acid necessarily involves hydrolysis of the nucleic acid, and chemical hydrolysis without degradation of one or more of the nitrogenous bases presents some difficulty. Depending on the conditions employed, the products are nucleotides, nucleosides, the free bases or mixtures of these substances. Thus subjecting a PNA to mildly acidic conditions, e.g. N.-HCl at 100° C. for i hour, liberates the purines in the free state, but the pyrimidines still remain in nucleotide combination from which they are freed by more vigorous hydrolysis, e.g. formic acid at 175° C. [see 25]. On the other hand, alkaline hydrolysis of a PNA yields a mixture of the four nucleotides. The deoxypentose nucleic acids behave differently on chemical hydrolysis, and degradation to nucleosides and nucleotides is best achieved by the use of enzymes. In the past, separation of the end-products ob- tained by hydrolysis or enzymic digestion of nucleic acid was accomplished by precipitation as the phosphotungstate or as the uranium or silver salt, or by simply adjusting the pH of the system. More recently, ionophoresis on paper [38]

PLATE II. — Separation and identification of the purines and pyrim- idines in yeast ribonucleic acid (YNA): YNA hydrolysed for 2 hr. in 70 %HC104 at 100° C: solvent system isopropanol (65% v/v) and 2N-HC1 (final concentration): G, A, C, and U pure samples of guanine, adenine, cytosine and uracil. Photograph obtained by placing chromatogram over reflex document paper and exposing to ultraviolet light. ST. de- notes starting line

NUCLEOTIDES I3I

and chromatography, both on paper [11] and on ion ex- change resins [18], have provided more convenient and pre- cise techniques for the separation, preparation [19] and identification of nucleotides, nucleosides and the bases. When a paper chromatogram is exposed to light of wave- length 260 m^w, the areas occupied by purines, pyrimidines, nucleosides or nucleotides appear as dark spots on a light blue fluorescent background. A permanent photographic record can be obtained by placing the chromatogram over reflex copying paper [37], and the areas in the chromato- gram containing compounds absorbing ultraviolet light will appear as white spots on a dark background in the developed photograph (Plate II). The appropriate areas of the chro- matogram are then cut out, the compounds eluted and estimated spectrophotometrically.

Microbial nucleic acids

Micro-organisms, particularly bacteria, are richer in nucleic acid than most of the cells of other organisms, and Belozersky has calculated that 15-30% of the dry weight of bacteria is nucleic acid and 50-80% is nucleoprotein [4]. Prior to isolating a nucleic acid, soluble nucleotides and phospholipoids are first removed by successively extracting the cells with cold TCA and a fat solvent. The residue is then treated with dilute solutions of an alkali, e.g. 0-2% NaOH or NagCOg in order to extract the nucleic acids, the details of the procedure varying according to the nucleic acid required. The isolation of microbial nucleoproteins, especially in an undegraded or 'native' state, presents many difficulties and, apart from low yields, it is doubtful if any of the present techniques are ideal. The most favoured method is to extract the cells with neutral solutions of NaCl at a concentration depending on the type of nucleic acid it is desired to isolate: even so, the deoxypentose nucleopro- teins adhere strongly to the cell structure and are only re- moved with difficulty [4]. When an aqueous solution of a nucleoprotein is shaken with chloroform, the protein be- comes denatured and collects at the interface, whilst the liberated nucleic acid remains in the water phase. The

132 NITROGEN METABOLISM

sodium salt of the nucleic acid can then be precipitated by the addition of ethanol, or, alternatively, the free nucleic acid by the addition of acidified ethanol or glacial acetic acid. The staining procedure introduced by Christian Gram to reveal the presence of bacteria in animal tissues was subse- quently developed into an empirical technique for dividing bacteria into two groups. After the fixed organisms have been stained with a basic dye (crystal violet or methyl violet) at pH 7-8, they are treated with a mordant, usually 1 2 in KI, and then washed with ethanol or acetone. If the stain is quickly removed, the organism is said to be Gram-negative; whereas if the dye remains, it is regarded as being Gram- positive. After the ethanol or acetone treatment, it is now common practice to counterstain with a red dye, with the result that in the final preparation Gram-positive organisms are stained blue whilst Gram-negative organisms are red. Though the mechanism of this staining reaction is still a matter of dispute [see 3, 41], retention of the basic dye by Gram-positive bacteria appears to be due to the presence of a Mg'^'^-PNA complex in the peripheral layers of the cells. The evidence for this belief rests on the observation that after treatment of the dead cells with ribonuclease — an enzyme depolymerizing PNA (p. 1 34) — or with a detergent such as bile salts. Gram-positive organisms stain as though they were Gram-negative. The action of the detergent is to liberate PNA from the cells and this material has been iso- lated by Henry and Stacey, who found that in the presence of Mg"*"^ and a reducing substance, the isolated PNA, or indeed yeast RNA, would restore Gram-positive stain- ing properties to the appropriate Gram-negative cytoskele- tons [21], but not to truly Gram-negative organisms. Though the ratio of PNArDPNA was reported to be 8:1 in Gram- positive as opposed to 1-3:1 in Gram-negative bacteria, these figures are disputed by Mitchell and Moyle who claim that the ratio is about 4:1 in all bacteria, irrespective of their staining properties [41]. Moreover, the presence of a peri- pheral layer containing PNA may not be a complete ex- planation of the structure responsible for the Gram-staining

PLATE III. — Photographs of fixed cells of a colon bacillus stained by the Giemsa technique: i, untreated cells; 2, fixed cells treated with ribonuclease prior to staining; 3, treated with ribo- nuclease and deoxyribonuclease. Ribonuclease removes cyto- plasmic PNA responsible for the overall staining in i and enables the presence of chromatinic (nuclear) bodies to be shown (2). The latter are composed of DFNA and are not present in the cells treated with deoxyribonuclease (3)

NUCLEOTIDES I33

reaction, since Gram-positive but not Gram-negative bac- teria contain large amounts of the phosphates of various polyalcohols, in particular, glycerol. If a suspension of Staph, aureus is shaken with minute glass beads, the cells are dis- rupted, but the outer layer of the cell — the 'cell wall' or 'cell envelope' — remains intact, and it is with this that the major portion of the polyol phosphates is associated [42]. However, no direct evidence has yet been presented that these phos- phates participate in the Gram-staining reaction. The recent work of Bartholomew and Mittwer has provided further support for the view that a layer immediately internal to the cell wall is the site of the Gram-staining reaction [3].

The nuclei of plant and animal cells are rich in nucleic acids, mostly of the DPNA type, and it is now generally accepted that transmission of hereditable characters is associated with these substances. Whether or not bacteria possess a nucleus has been the subject of endless and incon- clusive discussion, but they do undoubtedly contain struc- tures composed of DPNA and known as chromatinic bodies. The latter can be demonstrated in living cells by dark ground phase contrast microscopy [50] and in dead cells by the Feulgen or Giemsa staining technique. The Feulgen tech- nique is based on the fact that after being subjected to acid hydrolysis, deoxypentose nucleic acids, but not nucleic acids of the pentose type, restore the colour of Schiff's reagent, consequently cellular structures composed of DPNA be- come stained magenta. Since any aldehyde is capable of giving a positive reaction, the results should not be accepted without confirmatory evidence. If an observed staining re- action is due to PNA or DPNA, it should no longer be given by material previously treated with the appropriate enzyme, ribonuclease or deoxyribonuclease (p. 134), and then washed (Plate III). Caution is also required in interpreting the re- sults obtained in this manner since it may not be justifiable to assume (i) that the enzyme can penetrate the experimental material and thus come into contact with the substrate and (ii) that the enzyme preparation is specific in its activity.

Although the evidence is often indirect, , there are good reasons for believing that in micro-organisms, as in the

134 NITROGEN METABOLISM

more complex forms of life, the inheritance of specific char- acters is controlled by the deoxypentose nucleic acids, and in this respect a group of natural substances, known as trans- forming factors, are of particular interest. A transforming factor induces a susceptible cell to acquire a particular here- ditable and characteristic property of the cell from which the factor emanates. Once a cell has been 'transformed', the acquired feature is transmitted through all subsequent generations. The most thoroughly investigated example of this phenomenon is provided by the pneumococci whose virulence is associated with the possession of a capsule of polysaccharide material. Differences in the composition of the latter have enabled the pneumococci to be grouped into more than thirty serologically distinct types. Griffith ob- served that living non-encapsulated avirulent type II pneu- mococci were changed into virulent encapsulated type III pneumococci by passage of the former together with heat killed cells of the latter through mice. Such transformations can be brought about in vitro in certain well-defined condi- tions and later Avery and his colleagues obtained convincing evidence that the agents responsible for the transformation of pneumococcal types were deoxypentose nucleic acids, each acid being specific for one type of transformation [see i]. More recently, other transformations dealing with capsulation, resistance to penicillin and the ability to fer- ment particular sugars have been demonstrated with certain strains of Haem. influenzae, Esch. coli, Shigella paradysen- teriae, B. anthracis and Pr. vulgaris; as in the pneumococci, the factors accomplishing these transformations appear to be deoxypentose nucleic acids [i]. Pentose nucleic acids have been implicated in the formation of stretolysin S, the O2- stable haemolytic exotoxin of Strep, haemolyticus [7].

Enzymic degradation of nucleic acids [31]

The enzymic degradation of nucleic acids commences with disruption of the internucleotide linkages by ribo- nuclease or deoxyribonuclease, enzymes specific for the pentose and deoxypentose nucleic acids respectively: the nucleic acid is thus reduced to a mixture of mono-, di- and

NUCLEOTIDES I35

a few oligo-nucleotides. Whilst the properties of the nu- cleases of mammals and yeast have been studied in some detail, the corresponding bacterial enzymes have received comparatively little attention. Streptococcus haemolyticus (Group A) secretes both types of nuclease into the medium [34, 8], and ribonuclease is particularly active during the initial stages of bacterial autolysis [cf. 23]. At least as far as ribonuclease is concerned, depolymerization is due to hydrolysis of the bond linking C'-5 of one nucleo- tide to the phosphate group attached to C'-3 (or C'-2) of an adjacent pyrimidine nucleotide. Analogous linkages be- tween a pyrimidine and a purine nucleotide or between two purine nucleotides are stable to ribonuclease, and the enzyme may therefore be regarded as being a highly specific phosphodiesterase [38]. Nuclease activity can be determined by using a solution of the appropriate nucleic acid and fol- lowing (i) changes in viscosity, (ii) the appearance of acid- soluble phosphorus compounds, (iii) the appearance of acidic groups or (iv) the decrease in the absorption of ultra- violet light of wavelength 300 m/<.

After nucleotides have been dephosphorylated by appar- ently non-specific phosphatases, the resultant nucleosides may be attacked in one of three ways [28]. One type of nucleosidase catalyses their phosphorolytic decomposition to the free base and a pentose phosphate by transferring the sugar moiety to inorganic orthophosphate (reaction a). The phosphorolytic nucleosidases of Esch. colt exhibit specificity with regard to the nitrogenous base in that they attack either purine compounds or pyrimidine compounds, but not both, and yet no specificity is shown towards the sugar part of the substrates. Lactobacillus helveticus [33] and Esch. coli [28] possess another type of transferase, specific for deoxyribo- sides and utilizing not inorganic phosphate but a nitrogen- ous base as the acceptor (reaction b).

(a) uracil-riboside+H3P04 ^^ uracil+ribose-i -phosphate (6) uracil— deoxyriboside+thymine ^^

th>Tnin? —deoxyriboside + uracil

The hydrolytic decomposition of nucleosides to the free

136 NITROGEN METABOLISM

base and sugar in the complete absence of phosphate and other nitrogenous bases has been observed with prepara- tions from bakers' yeast [12] and Lb. pentosus [29].

The removal of amino groups from the nitrogenous bases is accomplished by highly specific deaminases, but apart from the cytosine deaminase of Esch. coli and yeast [cf. 14], the adenosine deaminase oi Neurospora [35] and the cytidine deaminase of brewers' yeast and Esch. coli [53], little is known about the occurrence and properties of these enzymes in micro-organisms.

Oxidation and fermentation of purines and pyrimidines

Application of the enrichment culture technique has led to the isolation of organisms capable of utilizing purines and pyrimidines as sole sources of carbon, nitrogen and energy. Thus CI. cylindrosporum ferments uric acid to NH3, CO 2 and acetic acid, while CI. acidi urici, in addition to uric acid, can also utilize xanthine, guanine and hypoxanthine [2]. Since neither of these species attacks allantoin, it has been suggested that they degrade uric acid by a route diiferent from that found in animal tissues, and there is some evidence that glycine is an important intermediate in these fermenta- tions. Unlike the Clostridia, Micrococcus aerogenes attacks neither uric acid nor glycine, and lactic acid is the main acidic end-product in the fermentation of adenine or gua- nine [54]. On the other hand, Micrococcus lactilyticus is unable to utilize adenine, guanine or uric acid but ferments hypoxanthine and xanthine to H2, COg, NH3 and urea, together with propionic and acetic acids [55]. A number of aerobic bacteria (species of Nocardia, Corynehacterium^ Mycobacterium and an unidentified soil organism) obtain carbon, nitrogen and energy by the oxidation of pyrimi- dines [30, 52]. Uracil and thymine are both oxidized to barbituric acid, which is then split into urea and malonic acid: the urea is subsequently decomposed by urease into CO 2 and NH3, but the further steps in the metabolism of malonic acid are unknown. The oxidation of thymine pro- ceeds by way of 5-methylbarbituric acid rather than by direct demethylation to uracil.

NUCLEOTIDES I37

Synthesis of purines

Though the synthesis of the nitrogenous bases and the pentose sugars is here, for convenience, considered separ- ately, such a division does not imply that the synthesis of nucleotides necessarily proceeds by the phosphorylation of a nucleoside formed by the joining together of the pre- formed base and the pentose sugar. There is some experi- mental evidence that glycosidation of a precursor precedes completion of the heterocyclic ring systems found in purines and pyrimidines. Moreover, the possibility must always be borne in mind that the routes of synthesis are not the same in diiferent species. Experiments with substances labelled with isotopes have revealed that Esch. coli and yeast resemble mammals in that they synthesize purines and pyrimidines from relatively simple precursors, namely CO 2, NH3, for- mate and glycine. In the synthesis of guanine by yeast, C-4 is derived from CO 2, C-2 and C-8 from formate or sub- stances which give rise to formate (e.g. serine, the methyl group of methionine), C-6, C-5 and N-7 from the carboxyl, methene and amino-N of glycine respectively, and the re- maining nitrogen atoms from NH3 [20]. The utilization of other compounds for purine synthesis is indirect and in- volves their prior conversion to CO 2, NH3, glycine or for- mate. Only one of the intermediates between these simple precursors and the completed purine is known. When Esch. coli is grown in the presence of sub-bacteriostatic concentra- tions of sulphanilamide or the folic acid analogue N-10- methylpteroylglutamic acid, an amine identified as 4- amino-imidazole-5-carboxamide (AIC) accumulates in the medium [49]. This substance only requires the addition of one carbon atom (corresponding to position 2) to complete the purine ring, and since it can be utilized by yeast, Lh. arahinosus and purine auxotrophs of Esch. coli and Ophiostoma, it is reasonable to suggest that it is a natural intermediate in purine synthesis. If glycine, a known pre- cursor in purine synthesis, is added to the medium, produc- tion of the amine is increased [45], whereas it is reduced by methionine, especially if trace amounts of PAB are present, and also by vitamin B;^,- Moreover, in the absence of purines,

138 NITROGEN METABOLISM

B12, but not PAB or folic acid, enhanced the utiHzation of AIC by mutants of Esch. coli exacting towards purines [6]. These resuhs are in accord with previous observations that the methyl groups of methionine serve as a source of one carbon units in intermediary metabolism and that they are related to formate, a substance known to be a precursor of purine carbon in position 2. Furthermore, Bj^g bas already been implicated in the metabolism of labile methyl groups in mammals, and it is to be noted that this growth factor had no effect on the utilization of the formyl derivative of AIC, 4-formamino-imidazole-5-carboxamide, by the Esch. coli mutants. Incubation of AIC with Esch. coli resulted in the formation of a substance, possibly a pentoside, which was five times more effective than AIC itself in supporting the growth of the purine auxotrophs [5]. Though the routes by which the various purine bases are synthesized may be quite distinct, it is most likely that they have several steps in common, and it is possible that one is formed directly from another. The interconvertibility of the purines shows species variation; thus whilst guanine and adenine are freely interconvertible in Esch. coli [27] and Lh. casei, yeast is only able to convert adenine to guanine. On the other hand, Tetrahymena gelei, a protozoon exacting towards guanine, can change guanine into adenine [26].

CO NH COOH CO

NH2 C \ CH NH CH

II CH II I II

C y C.COOH CO C.COOH

/ \ / / \ /

NH2 N NH2 NH

Amino-imidazolecarboxamide Anmiofumaric acid Orotic acid

Synthesis of pyrimidines

Although the pyrimidine ring system is also part of that present in purines, it is apparent that the synthesis of these two groups of nitrogenous bases proceeds by entirely inde- pendent routes. For example, in experiments with yeast, carbon from formate or glycine was incorporated into

NUCLEOTIDES I39

guanine and adenine but not into uracil orcytosine. Further- more, although carbon from isotopically labelled lactate was found in both types of compound, the resultant distribution of the isotope suggests that it entered the pyrimidines via oxaloacetate, and the purines via glycine [20]. Certain pyri- midine auxotrophs of Neurospora grow much better on uridine or cytidine than on the free bases, and this again suggests that the latter are not natural intermediates in nucleoside synthesis. Some of the mutants used orotic acid (uracil-6-carboxylic acid) to the same extent as uracil itself [32], whilst others, for which this replacement was not possible, accumulated orotic acid in the medium [39]. Though oxaloacetate, aminofumarate and aminofumaric acid diamide supported the growth of two of the mutants, they were only one tenth as effective as uracil. These and other observations led Houlahan and Mitchell to propose that in Neurospora, the biosynthetic sequence was: oxalo- acetate — >- a-A^-pentosylaminofumaric acid diamide — >- A — > B — > pyrimidine nucleoside: they also suggested that orotic acid is not a true intermediate but is related to the precursors A and B. In this connection it is interesting to note that a glycoside of orotic acid, probably the riboside, has been isolated from the mycelium of a uridine requiring mutant of Neurospora, and it is possible that this substance is a natural intermediate [43]. Incubation of Sac. cerevisiae with isotopically labelled orotic acid results in the appear- ance of the isotope in the uracil, but not the guanine, of the nucleic acids. Orotic acid has also been implicated in the synthesis of pyrimidines by streptococci, Lh. casei and Lh. bulgaricus [16]; the orotic acid requirements of the latter organism can be replaced by ureidosuccinic acid, a sub- stance related to aminofumaric acid [56].

Synthesis of deoxyribose, ribose, nucleosides and nucleotides

A partially purified preparation of an aldolase obtained from Esch. coli has been shown to condense glyceraldehyde- 3 -phosphate with acetaldehyde and thus form deoxyribose- 5-phosphate, which is converted by an isomerase, present in crude extracts, to deoxyribose- 1 -phosphate [44]. Whilst

140 NITROGEN METABOLISM

ribose- 5 -phosphate may likewise be synthesized from phosphoglyceraldehyde and glycolaldehyde, Esch. colt and Sac. cerevisiae are known to be capable of performing the series of reactions, glucose — > glucose-6-phosphate — >- 6-phosphogluconate — > 3(2)-ketogluconic acid — > ribulose- 5-phosphate — > ribose-5-phosphate — > ribose- 1 -phosphate [22, 36]. In view of the reversible nature of the nucleosidases, it is reasonable to suggest that transfer reactions between the pentose- 1 -phosphates and purines or pyrimidines are a possible route of nucleoside synthesis. There is, however, no proof that this is the natural pathway and mention has already been made of some evidence, admittedly indirect, which indicates that glycosidation precedes completion of the heterocylic rings present in purines and pyrimidines. With regard to the phosphorylation of nucleosides to form nucleotides, all of our present knowledge concerns nucleo- tides known to be constituents of coenzymes rather than of nucleic acids. The following syntheses, catalysed by enzyme preparations obtained from yeast {c, d, f) and from liver \e, g) have been described [24, 46]:

(c) adenosine+ATP — > adenosine-5-phosphate+ADP

Id) riboflavin+ATP — > FMN+ADP

{e) nicotinamide-riboside+ATP — >■

nicotinamide-riboside-5-phosphate+ADP (/) FMN+ATP — > FADN+pyrophosphate

(^) Nicotinamide-riboside-5-phosphate+ATP — >

DPN +pyrophosphate

Effect of bacteriophage on nucleic acid metabolism of host cell [17] All viruses consist essentially of nucleic acids and pro- teins, and, in addition, the more complex types infecting animals also contain fatty material. Deoxypentose nucleic acids and PNA are found respectively in bacteriophages and plant viruses, and although most animal viruses contain only one type of nucleic acid, a few, e.g. the influenza virus, appear to contain both DPNA and PNA. When purified fully infective preparations of various viruses have been examined for the presence of known enzymes, no activity

NUCLEOTIDES I4I

has been found except in those of the more complex animal viruses. Interest has therefore centred around the mechan- ism whereby a virus assumes control of the metabolic activities of the host cell and reorganizes them for the syn- thesis of identical virus particles. The virus-host-cell system investigated in most detail is that concerned with the infec- tion of Esch. coli B with a phage designated as the T phage, of which seven types are known. Electron microscopy has revealed that Ti and T5 are more or less spherical, whereas the other five T phages are club-shaped and have a distinct head, containing most of the DPNA, and a tail. Phage repro- duction involves three phases: (i) adsorption on to and invasion of the susceptible cell; (ii) multiplication in the host cell, and (iii) liberation from the host cell, a process usually accompanied by lysis, though liberation and lysis are not always coincident. Unless stated to the contrary, the follow- ing description applies to phage T2 and the details are not necessarily the same for other viruses.

The adsorption of T2 by Esch. coli B is a rapid and reversible process dependent on the presence of tryptophan in the medium. It is soon followed by an irreversible process in which part of the virus, mainly DPNA and a little pro- tein, enters the cell, an event which immediately results in a marked disturbance of the host cell's nucleic acid meta- bolism, the first visible signs being degradation of the chromatinic bodies. The inability of virus-infected cells to grow and divide is presumably an outw^ard expression of this destruction of nuclear material. Nucleic acid synthesis is at first completely halted, but after a little while the synthesis of purines and pyrimidines commences and is soon followed by the formation of DPNA. Pentose nucleic acid synthesis is completely suppressed in Esch. coli B infected with T2, but in other virus-host-cell systems, the synthesis of both types of nucleic acid may take place. The amount of DPNA synthesized is equivalent to the total amount of both types of nucleic acid synthesized by uninfected cells, and this indicates that in the infected cells the units normally incorporated into PNA are being diverted to the synthesis of DPNA. In attempting to discover the reason

142 NITROGEN METABOLISM

for this diversion Cohen has found that infection with phage causes a marked reduction in the abiUty of the cell to form ribose-5-phosphate from glucose-6-phosphate (p. 140), and this presumably may explain why such cells are unable to synthesize RNA. The synthesis of deoxyribose- 5 -phosphate from glucose-6-phosphate is unimpaired and consequently deoxypentose-nucleotide synthesis is unaffected.

Infection with phage has no visible effect on protein synthesis in the host cell, and although the amount of pro- tein synthesized is in excess of that found in the number of virus particles eventually liberated, it is not yet known whether all the excess is viral protein. The origin of the various substances present in the phage has been determined by using cells whose cellular constituents have been enriched with isotopes and also by suspending the infected cells in media containing the appropriate compounds labelled with isotopes. As far as T2 is concerned, no more than 25% of the phosphorus, protein nitrogen, pyrimidines and purines in the liberated phage is derived from substances present in the host cell at the time of infection. How the virus particle gains control of the anabolic activities of the host cell is still unknown. It is however clear that the term 'self- duplicating particle' should not be used indiscriminately since, for example, the synthesis of protein and nucleic acid in infected cells proceeds at a linear rate, whereas if the system were self-duplicating, one might expect it to be auto- catal5rtic. Moreover, there is evidence that the virus is radically changed during invasion of the host cell, since much of the phosphorus and some of the protein nitrogen of the infecting phage particle soon appears in the medium, and not for an appreciable time after invasion is it possible to isolate infective virus from the host cell. The assembling of the various units into completed virus appears to occur only a short time prior to liberation. In the T2 system, the first recognizable structures to be found in the infected cells are the collapsed heads of the phage containing only a little DPNA, the rest of the latter and the tails being added later.

NUCLEOTIDES I43

REFERENCES

1. Austrian, R. (1952), Bad. Rev., 16, 31

2. Barker, H. A. and Beck, J. V. (1942), J?^. Bact., 43, 291 '

3. Bartholomew, J. W. and Mittwer, T. (1952), Biol. Rev., 16, i

4. Belozersky, A. N. (1947), Cold Spr. Harb. Symp., 12, i

5. Ben-Ishai, R., Bergmann, E. D. and Volcani, B. E. (1951),

Nature, 168, 11 24

6. Bergmann, E. D., Volcani, B. E. and Ben-Ishai, R. (1952),^.

hiol. Chem., 194, 521, 53 1

7. Bernheimer, A. W. (1949), jf. exp. Med., 90, 373

8. (1953), Biochem. J., 53, 53

9. Caputto, R., Leloir, L. F., Cardini, C. E. and Paladini, A. C.

(1950), J' biol. Chem., 184, 333

10. Caspersson, T. (1947), Soc. exp. Biol. Symp., i, 127

11. Carter, C. E. (1950), j^. Amer. chem. Soc, 72, 1466

12. (i95i)> J- Amer. chem. Soc, 73, 1508

13. Chargaff, E. (195 1), Fed. Proc, 10, 654

14. and Kream, J. (1948),^. hiol. Chem., 175, 993

15. and Saidel, H. F. (1949), J. biol. Chem., 177, 417

16. Christman, A. A. (1952), Physiol. Rev., 32, 303

17. Cohen, S. S. (1951), Bad. Rev., 15, 131

18. Cohn, W. E. (1950),^. Amer. chem. Soc, 72, 1471

19. and Carter, C. E. (1950), J. Amer. chem. Soc, 72, 2606

20. Edmunds, M., Delluva, A. M. and Wilson, D. W. (1952), J.

biol. Chem., 197, 251

21. Henr\', H. and Stacey, M. (1946), Proc. Roy. Soc, 133B, 391

22. Horecker, B. L. and Smyrniotis, P. Z. (195 1), J. biol. Che7?i.,

I93» 371, 383

23. Jones, A. S., Stacey, M. and Webb, M. (1949), Biochim.

Biophys. Acta, 3, 383

24. Kearney, E. B. and Englard, S. (1951),^^. biol. Chem., 193, 821

25. Kerr, S. E., Seraidarian, K. and Wargon, M. (1949), J. biol.

Chem., 181, 761

26. Kidder, G. W. and Dewey, V. C. (1948), Proc Nat. Acad. Sci.,

34, 566

27. Koch, A. L., Putnam, F. W. and Evans, E, A. (1952), jf. biol.

Chem., 197, 105

28. Lampen, J. O. (1952), Bad. Rev., 16, 211

29. and Wang, T. P. (1952), J. biol. Chem., 198, 385

30. Lara, F. J. S. (1952), J. Bact., 64, 271, 279

31. Laskowski, M. in The Enzymes, 1 (ii), Chap. 29

32. Loring, H. S. and Pierce, J. G. (1944), J- ^iol. Chem., 153, 61

33. MacNutt, W. S. (1952), Biochem. y., 50, 384

34. McCartv, M. (1948),^. exp. Med., 88, 181

35. McElrov, W. D. and Mitchell, H. K. (1946), Fed. Proc, 5, 376

36. McNair-Scott, D. B. and Cohen, S. S. (1951), J. biol. Chem.,

144 NITROGEN METABOLISM

37. Markham, R. and Smith, J. D. (1949), Biochem. J., 45, 294 38. (1952), Biochem. y., 52, 552, 558, 565

39. Mitchell, H. K., Houlahan, M. B. and Nye, J^ F. (1948), J.

biol. Chem., 172, 525

40. Mitehell, P. {1950), jf. gen. Microbiol., 4, 399

41. and Moyle, J. (1950), Nature, 166, 218

42. and Moyle, J. (195 1), jf. gen. Microbiol., 5, 966, 981

43. Mitchelson, A. M., Drell, W. and Mitchell, H. K. (195 1),

Proc. Nat. Acad. Sci., 37, 396

44. Racker, E. (1952), J. biol. Chem., 196, 347

45. Ravel, J. M., Eakin, R. E. and Shive, W. (1948), J', biol. Chem.,

172, 67

46. Rowen, J. W. and Romberg, A. {1951), J- biol. Chem., 193, 497

47. Schmidt, G. and Thannhauser, S. J. (1945), jf. biol. Chem.,

161, 83

48. Schneider, W. C. (1945), 3^- biol. Chem., 161, 293

49. Shive, W., Ackermann, W. W., Gordon, M., Getzendaner,

M. E. and Eakin, R. E. (1947), jf. Amer. chem. Soc, 69, 725

50. Stempen, H. (1950),^. Bad., 60, 81

51. Vischer, E., Zamenhof, S. and Chargaff, E. (1949), jf. biol.

Chem., 177, 429

52. Wang, T. P. and Lampen, J. O. (1952), jf. biol. Chem., 194,

775, 785

53. Sable, H. Z. and Lampen, J. O. (1950), jf. biol. Chem.,

184, 17

54. Whiteley, H. R. (1952),^. BacL, 63, 163

55. and Douglas, H. C. (195 1), jf. Bact., 61, 605

56. Wright, L. D., Miller, C. S., Skeggs, H. R., Huff, J. W.,

Weed, L. L. and Wilson, D. W. (195 1), J. Amer. chem. Soc, 73, 1898

57. Wvatt, G. R. (1951), Biochem. J., 48, 581, 584

CHAPTER X

MODE OF ACTION OF CHEMOTHERAPEUTIC AGENTS

In recent years, the dramatic success of the sulphonamides and antibiotics in the treatment of bacterial and viral infec- tions of man has focused much attention on the mode of action of these substances and the factors contributing to their activity. Substances interfering with the continued normal existence of an organism may owe their properties to (i) denaturation of proteins, (ii) disruption of cellular membranes and in consequence the leakage of essential metabolites, or (iii) more specifically, inhibition of particular enzymes. Substances bringing about the first two of these effects are of little chemotherapeutic importance, because of their adverse effect on the host as well as the pathogen. The key to successful chemotherapy is contained in the phrase selective toxicity, i.e. the drug is more toxic to the pathogen than to the host. If a chemotherapeutic agent acts by inhibiting an enzymic system of fundamental importance to the pathogen, then this system must be absent from the host, or, if present, it is for some reason less susceptible to the drug, or not essential. (For a detailed exposition of the principles of chemotherapy and selective toxicity, references 31 and I are recommended.) One of the aims of contem- porary research is to establish a rational basis for chemo- therapy, and with this in mind, the purpose of the following paragraphs is to survey very briefly the observed effects of the sulphonamides and antibiotics on the metabolism of nitrogenous compounds.

It must be emphasized that great care is required in deciding whether the observed effect produced by a drug is due to direct inhibition of the reaction leading to that end result, or whether it is a secondary effect arising out of the inhibition of some other, perhaps unknown, reaction. Even if the drug inhibits cell-free preparations of an enzyme this

145

146 NITROGEN METABOLISM

does not constitute proof that its mode of action against the intact organism is necessarily expUcable in such terms. Before the latter can be attempted, precise and quantitative information is required deaUng with the significance of that enzyme in the general economy of the cell (e.g. turnover numbers). Moreover, ideally the observed results should be produced by drug concentrations of the same order as those used therapeutically. But, since the concentration of cells in washed suspension experiments is often many times greater than those in growing cultures or infected animals, it has been argued that this proviso can be ignored [14]; furthermore, it is also feasible that the drug enters growing and dividing cells more readily than resting cells [10].

Sulphonamides

The first major contribution towards understanding the mode of action of the sulphonamides was made by Woods who prepared from yeast an aromatic carboxylic amine which competitively antagonized the action of sulphanila- mide in preventing the growth of Strep, haemolyticus. Fildes' suggestion that a chemotherapeutic agent might function by virtue of its chemical structure being such that it was adsorbed on to an enzyme in place of the natural substrate led Woods to infer that the isolated material was ^-amino- benzoic acid (PAB) and he showed that the latter did in fact antagonize the sulphonamides in a competitive manner [28]. In other words, the biological activity of the sulphonamides could be explained on the grounds that they were non- utilizable analogues of a natural metabolite, namely, PAB. Until that time the importance of PAB in intermediary metabolism had not been suspected, but evidence soon became available that PAB was a growth factor for certain organisms, and that a group of substances containing PAB — the folic acid factors required by certain other organisms — ^were of universal importance. The various folic acid factors differ in the number of glutamic acid radicals in the molecule, the degree of reduction of the pterin and the presence or absence of a formyl group attached to one of the nitrogen atoms of the pterin or PAB. For example, the

CHEMOTHERAPEUTIC AGENTS I47

*folic acid' requirements (citrovorum factor or folinic acid) of Leuconostoc citrovorum are replaceable by a synthetic material, A^-5-formyl-5, 6, 7, 8-tetrahydropteroylglutamic acid, but not by pteroylglutamic acid itself (synthetic folic acid). The experimental evidence favours the view that organisms utilize PAB solely for the synthesis of 'folic acid' but the detailed constitution of the natural coenzyme con- taining combined PAB is still unknown [29].

The role of PAB in intermediary metabolism was ex- plored further by a comprehensive survey of the substances other than folic acid which were capable of antagonizing the growth inhibitory action of the sulphonamides. This proce- dure is based on the hypothesis that if PAB or a derivative functions catalytically in the synthesis of substances essential for growth, and if the sulphonamides act by preventing the normal functioning of PAB, then growth should be resumed if these substances are supplied exogenously. In a sense such cultures can be regarded as being deficient in PAB or folic acid, and the principles involved are the same as in the growth factor replacement technique in which an attempt is made to replace a grov^rth factor either by simpler substances from which it can be synthesized, or by sub- stances whose synthesis the growth factor is suspected to mediate. Apart from folic acid, the natural antagonists of the sulphonamides fall into three groups, (i) amino-acids, in particular methionine and serine, (ii) purines such as xanthine and (iii) thymine and thymidine [22, 27]. For example, the addition of methionine to the medium decreased the amount of PAB required to overcome the inhibitory effects of sulphanilamide on the growth of Esch. colt. The PAB requirement was further reduced if xanthine was also included and diminished still further, if the medium contained methionine, xanthine and serine. It was abolished altogether when thymine was added in addition to these three substances. Similarly with a PAB auxotroph of Esch. coll growth was possible in the absence of PAB, provided the medium contained methionine, a purine and thymine. Analogous experiments have been done with other organisms and the same three groups of substances

148 NITROGEN METABOLISM

are also active in antagonizing the growth inhibitory pro- perties of analogues of folic acid, e.g. A:-methylfolic acid.

Previous work with animal tissues and Neurospora indi- cated that serine was synthesized by the addition of formate

p-aminobenzoic acid

HOOC

precursors

glycine + formate

f

serine

precursors

valine

lysine

threonine

histidine

homocysteine methionine

H©

deoxyribosides

of guanine, adenine

and cytosine

citrovorum

factor

(f clinic acid)

combined PAB Folic acid ?

CHzCOOH

ch.nh.oc-^~\nh .CH2-«|

COOH

pteroylglutamic acid (synthetic folic acid)

-t

»""=

precursors

thymidine thymine

precursors

amino- imidazole carboxamide

MS)

vitamin ^2

purines

FIG. 10. 1. — Role of ^-aminobenzoic acid and vitamin B12 in the synthesis of amino-acids, purines and pyrimidines

to glycine, and since glycine, unlike serine, did not antagonize sulphanilamide, it was concluded that the latter was an inhibitor of this synthesis. Woods and his colleagues have now shown that though Ln. mesenteroides is exacting towards serine, growth occurred in the absence of the amino-acid

CHEMOTHERAPEUTIC AGENTS I49

provided pyridoxin, PAB, CO 2 and glycine were available. Further evidence that pyridoxin as well as PAB partici- pated in the synthesis of serine from glycine was obtained by using cells grown in the absence of these two growth factors in a medium containing all the other known grov^h fac- tors together with amino-acids, purines and pyrimidines. Washed suspensions of these cells would synthesize serine only if they were provided with glycine, formate, glucose, pyridoxin and PAB (or folinic acid); synthesis was com- pletely inhibited by sulphanilamide. Similar experiments have been done with Strep. faecaliSy Lb. hifidus and Sac. cerevisiae. It appears that both PAB and vitamin B^g play a role in the synthesis of methionine, the most active amino- acid antagonist of the sulphonamides. A vitamin B^2 auxo- troph of Esch. coli grew in the absence of B^g if the medium contained methionine [4], while cells of a mutant requiring PAB and grown in the absence of this factor (i.e. in a medium containing amino-acids, purines and pyrimidines) only synthesized methionine from homocysteine in the presence of PAB and glucose. The simultaneous addition of B12 stimulated synthesis threefold, and recent work suggests that the methyl group used in the methylation is derived from serine. There is some evidence that the role of PAB in methionine synthesis cannot be explained solely on the basis that it is required for the synthesis of B12 and serine [29].

The connection between PAB and the synthesis of the group of substances now designated as vitamin B12 (the cobalamins) began with the observation that the anaemia produced by feeding an animal large amounts of sulpha- thiazole was like pernicious anaemia and could be relieved by large doses of synthetic folic acid or concentrates of sub- stances isolated from the liver of normal animals. These liver substances were not of the folic acid type and functioned as growth factors for Lb. leichmannii and Lb. lactis. The key compound, vitamin B^g, active both as a growth factor and in the treatment of pernicious anaemia, has now been iso- lated in the crystalline state and is composed of 5:6-di- methylbenzimidazole - 1 - a - D -ribofuranoside-3 -phosphate

150 NITROGEN METABOLISM

combined with an unidentified organic molecule contain- ing cobalt. Though the B^g requirements of a micro-organ- ism cannot be replaced by any of the known folic acids, either synthetic or natural, they are replaceable by thymi- dine and also in most organisms, by the deoxyribosides of purines. This is to be contrasted with the fact that though thymidine replaces part of the PAB or folic acid require- ments of an organism, all other deoxyribosides are inactive. Vitamin B^g contains an aromatic ring, yet it did not stimu- late the growth of mutants with a multiple requirement for aromatic compounds. Davis therefore suggested that it was either derived from one of the aromatic substances required by such mutants or else it was synthesized by a totally different route or from an intermediate prior to the genetically blocked reaction. Vitamin B^g exerts a sparing effect on the PAB requirement of a PAB auxotroph of Esch. coli, and since the amount of vitamin B12 required is only one fiftieth of the amount of PAB, it is conceivable that the ring of PAB is used directly in the synthesis of B^g [3]. On the other hand, folic acid is known to be associated with the synthesis of purines, which, like B^gj contain an imidazole ring, and this may be the reason for the close relationship between these two co-factors. At the present time, the details of the relationship are but vaguely understood and have only been explored in one species, Esch. coli.

From these and other studies, it has become evident that the ultimate co-factor form of PAB participates in the methylation of homocysteine, the introduction of carbon into position 2 of the purine ring (pp. 1 37-8), and the synthesis of serine and thymine; in other words, this co-factor is con- cerned with the intermediary metabolism of one carbon units (cf. Co.A, the ultimate co-factor form of pantothenic acid, and its function in the metabolism of acyl units [20]). There is also reason to believe that Bjg is involved in at least some of these reactions, e.g. the synthesis of methionine and purines, and it is to be noted that the presence of B^g increases the amount of sulphanilamide required to induce bacteriostasis of Esch. coli growing in the presence of one or more of the sulphonamide antagonists methionine,

CHEMOTHERAPEUTIC AGENTS I5I

xanthine, serine and thymine [21]. Other effects of the sulphonamides have been reported, but whether they are also the outcome of interference with reactions involving PAB is not always known, e.g. sulphathiazole appears to interfere with protein synthesis in Staph, aureus, but since protein synthesis involves a complex metabolic sequence of reactions, it is not possible to deduce whether this is a direct effect or simply due to disturbances in the synthesis of amino-acids or nucleotides [5]. It has been shown that the even- but not the odd-numbered T phages are unable to multiply in Esch. coli growing in media containing sul- phanilamide together with methionine, serine, xanthine and thymidine [19]. The results of these and other experiments require to be interpreted with care, e.g. Pfiffner and his co- workers have isolated from bacteria compounds of the vitamin B^g group which contain adenine instead of di- methylbenzimidazole, and Davis later found that these substances (pseudo-Big) replaced B^g i^i all respects for Esch. coli mutants requiring this vitamin or methionine.

Antibiotics

Though the precise details are still unknown, it is highly probable that the sulphonamides owe their activity to the fact that they are metabolic analogues. On the other hand, no such simple hypothesis is available to explain the activity of any of the antibiotics. There is no conclusive evidence that penicillin inhibits enzyme systems concerned in respira- tion or fermentation, and it is therefore unlikely that any of the results described below are attributable to direct interference with energy metabolism. By way of contrast, aureomycin resembles 2:4-dinitrophenol and sodium azide in that it may act as an uncoupling agent and prevent the production of energy-rich phosphate groups [15]. If peni- cillin (o-i-io Oxford Units /ml.) was added to growing cultures of Staph, aureus, then within an hour of contact with the antibiotic the ability of the organism to accumulate amino-acids and synthesize protein progressively declined [8]. Penicillin had no effect on the uptake, of glutamic acid and lysine by washed suspensions of normal cells, hence it II

152 NITROGEN METABOLISM

would appear that this antibiotic prevents the synthesis of the systems responsible for the absorption of amino-acids and does not affect the functioning of those systems once they have been established. When washed suspensions of cells grown for a short time with penicillin were incubated with glucose and glutamic acid, little of the latter accumu- lated in the cells, though extracellular peptides of glutamic acid appeared in the system [7]. These results are analo- gous to those of Hotchkiss, who used a different species of staphylococcus and different experimental conditions. Washed suspensions incubated with a mixture of various amino-acids and glucose synthesized protein, but in the presence of penicillin there was no increase in cellular com- bined amino-acids though the number of free amino-groups in the medium decreased. As in Gale's experiments, the latter was correlated with the appearance of extracellular peptides, and Hotchkiss suggested that penicillin inhibited protein synthesis and that these peptides were either inter- mediates in this process or were derived from them [13]. Nevertheless, it is difficult to believe that penicillin inter- feres directly with the synthesis of all proteins since it has no effect on the formation of adaptive enzymes, a process now regarded as being associated with the synthesis of new protein. Streptomycin, aureomycin, chloramphenicol and terramycin inhibit adaptive enzyme formation [11], an effect which is possibly the outcome of interference with energy metabolism.

By examining a number of strains of Staph, aureus^ Gale found that increased resistance to penicillin could be correlated with a decline in ability to accumulate glutamic acid. It will be recalled that Gram-negative, unlike Gram- positive, organisms do not concentrate amino-acids in the cells (p. 82), and the most highly resistant variants of Staph, aureus obtained by successive subculture in increas- ing concentrations of penicillin were in fact Gram negative. Moreover, these organisms were no longer cocci but rod- shaped and had lost the ability to utilize certain sugars and grow anaerobically. Several workers have noted one or more of these effects (i.e. changes in morphology. Gram-staining

CHEMOTHERAPEUTIC AGENTS 153

properties and ability to metabolize various substances, and also a preference for aerobic growth) with other species growing in the presence of penicillin. By studying whether the development of resistance is accompanied by overall changes in the metabolism of the organism, it may be possible to gain valuable information concerning the mode of action of the agent being considered, and furthermore, if resistance to other drugs is acquired simultaneously, i.e. cross resistance, it is conceivable that the biological effects of all these substances is explicable in the same terms [see 31]. Another as yet unexplained observation that penicillin inter- feres with amino-acid metabolism concerns an unidentified Gram-negative organism which when growing on L-leucyl- glycine in a mineral salt medium was relatively insensitive to penicillin. Leucine and glycine, either singly or together, also supported growth, but in the presence of uncombined glycine the organism was very sensitive to penicillin (i to 10 units/ml.) [23].

The reports that the training of Staph, aureus to a high degree of resistance to penicillin resulted in a loss of Gram- positive staining properties could be taken to indicate that the biological effects of penicillin were the outcome of primary disturbances in nucleotide metabolism. In normal cultures of Staph, aureus the rate of cell growth appears to be controlled by the amount of pentose nucleic acid in the cells, and the cellular concentration of soluble nucleotides is inversely proportional to the rate of PNA synthesis [17]. If penicillin is added to a culture in the log phase of growth, then before there is any visible change in the growth-rate, the concentration of soluble nucleotides increases and the ratio of soluble nucleotides to total nucleic acid soon changes from o-i to 0-2 (Fig. 10.2). The percentage by weight of nucleic acid at first appears to increase, not because synthesis is stimulated but because there is a decrease in the rate of synthesis of some other substances (protein?) contributing to the dry weight of the cell. Though large amounts of nucleic acid are normally present in young cells, penicillin causes their concentration to fall rapidly to the low value characteristic of old cells in the stationary phase of growth,

154

NITROGEN METABOLISM

Q ■ • NORMAL CULTURE

•-C-- CULTURE PLUS

\fj^. PENICILLIN/ML.

8 20 22

T 1 I I n T

6 8 10 12 14 16 TIME (in hours)

FIG. I0.2. — Effect of penicillin on nucleic acid metabolism of Staph, aureus. %NA=percentage by weight of total nucleotide, %Nt= percentage by weight of nucleotide extractable with 5% trichloracetic acid, NA-Nt=total nucleic acid, DW— dry weight of organism in /xg./ml. [17]

CHEMOTHERAPEUTIC AGENTS 155

and the ability to adsorb glutamic acid is not impaired until this process is taking place. From such results it may be deduced that penicillin interferes not with nucleotide syn- thesis but with their polymerization to nucleic acid. Park and Johnson have noted that the gro\sth of Staph, aureus in the presence of penicillin (i unit/ml.) leads to the accu- mulation of intracellular compounds containing uridine- 5 -pyrophosphate and an unidentified amino carboxylic sugar: a peptide of D-glutamic acid, D-valine and DL-lysine is a component of one of these compounds, whilst another contains L-alanine [i8]. Synthesis of such nucleotides only occurs for as long as the cells are viable, and Park suggests that they are natural intermediates whose utilization is inhibited by penicillin. Maass and Johnson have shown that part of the penicillin absorbed by a cell is irreversibly bound within it, and they postulate that the antibiotic combines with and thus inhibits the natural functioning of a cellular constituent which is normally present in small amounts and controls the processes of cell division [16].

Penicillin is lethal to most Gram-positive organisms and is effective against only a few Gram-negative species, whereas streptomycin, chloramphenicol and aureomycin are active against a wide variety of organisms, and the two latter are also valuable in the treatment of diseases due to viruses and rickettsiae. Streptomycin had no effect on the accumu- lation of glutamic acid by Staph, aureus, but in concentra- tions markedly greater than those inhibiting growth, pre- vented protein synthesis. Aureomycin and chloramphenicol inhibited the absorption of glutamic acid and protein syn- thesis, the latter being especially sensitive [6, 7]. Of the many enzyme systems examined, only the diamine oxidase activity of whole cells of Mycobacteria, Ps. aeruginosa and Staph, aureus was inhibited by streptomycin, and there was some evidence that inhibition of this oxidase resulted in the cessation of growth [32]. Streptomycin contains basic guanidine groups and a possible explanation was that it was absorbed on to the enzyme in place of the natural basic substrate. However, cell-free preparations of the oxidase were but little affected by streptomycin, hence the observed

156 NITROGEN METABOLISM

result is not due to direct inhibition of this enzyme [9]. The oxidase activity of whole cells of streptomycin resistant variants of Myc. smegmatis was very much less sensitive and like the mammalian enzyme, only inhibited by high concen- trations [32]. Other experiments indicate that this antibiotic interferes with the entry of pyruvate into a terminal pathway responsible for its oxidation [12, 25], and that this pathway does not involve conversion to acetate and con- densation of acetate with oxaloacetate to form citrate. Streptomycin-resistant strains of Esch. coli do not possess this pathway, and although it is also present in mammalian mitochondria, permeability barriers prevent streptomycin from having any effect [26]. Chloramphenicol inhibits esterases in the cell-free state, but had no effect on forty other enzymes examined [24]. There is some evidence that it interferes with the metabolism of aromatic amino-acids, e.g. with Esch. coli, the addition of phenylalanine, tyrosine or tryptophan overcame the growth inhibitory effects of low concentrations of chloramphenicol [30], and in Esch. coli, as in Salm. typhosa, it appears to prevent the conversion of anthranilic acid to indole [2].

Although the above account is very incomplete, it serves to illustrate that much has still to be discovered before a precise statement can be made concerning the mode of action of the sulphonamides and antibiotics. Nevertheless, apart from their potential value in the development of new chemotherapeutic agents, such studies have made and can make valuable contributions to the general pool of bio- chemical knowledge.

REFERENCES

1. Albert, A. (1951), Selective Toxicity, Methuen, G.B.

2. Bergmann, E. D. and Sicher, S. (1952), Nature, 170, 931

3. Davis, B. D. (1951), 3^. Bact., 62, 221

4. and Mingiolo, E. S. (1950), J. Bad., 60, 17

5. Gale, E. F. (1947),^. gen. Microbiol, i, 327

6. and Folkes, J, P. (1953), Biochem. J., 53, 493

7. and Paine, T. F. (195 1), Biochem. J., 48, 298

8. and Taylor, E. S. (1947), J. gefi. Microbiol., i, 314

9. Geronimus, L. H. (195 1), Bad. Proc, 128

CHEMOTHERAPEUTIC AGENTS I57

10. Gros, F., Beljanski, M. and Macheboeuf, M. (1951), Bull. Soc.

Chim. Biol., Paris, 33, 1696

11. Hahn, F. E. and Wisseman, C. L. (195 1), Proc. Soc. exp. Biol.

Med., 76, 533

12. Henry, J., Henr}% R. J., Housewright, R. D. and Berkman, S.

(1948), 7 Bact., 56^527

13. Hotchkiss, R. D. (1950), 3^. exp. Med., 91, 351

14. Krampitz, L. O. and Werkman, C. H. (1947), Arch. Biochem.,

12, 57

15. Loomis, W. F. (1950), Science, iii, 474

16. Maass, E. A. and Johnson, M. J. (1949), J. Bad., 58, 361

17. Mitchell, P. (1949), Nature, 164, 259

18. Park, J. T. (1952), J. biol. Chem., 194, 877, 885, 897

19. Rutten, F. J., Winkler, K. C. and de Haan, P. G. (1950), Brit.

y. exp. Path., 31, 369

20. Shive, W. (1950), Ann. N.Y. Acad. Set., 52, 1212

21. (1951), Vitamins and Hormones, 9, 75

22. and Roberts, E. C. (1946), jf. biol. Chem., 162, 463

23. Simmonds, S. and Fruton, J. S. (1950), Science, iii, 329

24. Smith, G. N., Worrel, C. S. and Swanson, A. L. (1949), J.

Bact., 58, 803

25. Umbreit, W. W., Smith, P. H. and Oginsky, E. L. (195 1), J.

Bact., 61, 595

26. and Tonhazy, N. E. (1949), J. Bact., 58, 769

27. Winkler, K. C. and de Haan, P. G. (1948), Arch. Biochem., 18,

97

28. Woods, D. D. (1940), Brit.y. exp. Path., 21, 74

29. (1952), Internat. Congress of Biochem., Microbial Meta- bolism, p. 86

30. Woolley, D. W. {1950), J. biol. Chem., 185, 293

31. Work, T. S. and Work, E. (1948), The Basis of Chemotherapy,

Oliver and Boyd, G.B.

32. Zeller, E. A., Owen, C. A. and Karlson, A. G. (195 1), J. biol.

Chem., 188, 623

INDEX

Acetylsulphanilamide,synthesis,i03 Adaptation, 13 Adenine, deamination, 136 — , in nucleic acids, 126-7 Adenosine, deamination, 136 — , triphosphate, 3 Adenyhc acid, co-factor for deami- nases, 23-5 D- Alanine and Lb. casei, loi — , replacement of pyridoxin, 28,

lOI

L-Alanine, as H-donor, 17, 18, 21

— , deamination, 11, 17

— , fermentation, 21

— , from kynurenine, 14

— , racemase, 95

— , transamination, 60-4

Alcohols from amino-acids, 22

Aldolase, 139

Algae, nitrogen-fixation, 50

Amines, from amino-acids, 27

D-Amino-acids, in antibiotics, 96-7

— , in capsules, 96

- — , in peptides, 96-7, 155

— , oxidase, 11

— , racemase, 95

— , utilization, 95

L- Amino-acids, acetylation, 104

— , active transport of, 91

-^, antagonism, 65

— , as H-acceptors, 18

— , as H-donors, 17

— , as sources of carbon and nitro- gen, 20-2

— , assay by decarboxylases, 28

— , decarboxylases, 27-9

— , fermentation, 17-22

— , free in Gram-positive bac- teria, 82

— , in bacterial proteins, 95

— , oxidase, lo-ii

— , synthesis, 60-78

— , transamination, 60-4

— , uptake, 80-93

— , — , by Staph, aureus^ 86-9

— , — , by Strep, faecalis, 83-9

— , — , by yeast, 89

— , — , glutamic acid, 86-9, 91-3

— , — , lysine, 83-6, 90-1

— , — , mechanism, 89

^-Aminobenzoic acid, in folic acid, 146

, sulphonamide antagonist, 146

Aminofumaric acid, 139

, diamide, 139

4-Aminoimidazole-5-carboxamide,

137-8 Aminopolypeptidases, 113 Ammonia, and biotin-deficient

yeast, 25 — , as source of nitrogen, 4 — , excretion by CI. pastenrianum,

.54 . . — , inhibition of nitrate reduction,

34

— , — of nitrogen-fixation, 54

— , intermediate in nitrogen- fixa- tion, 52

— , liberation by deaminases, 10-26

— , oxidation, 32

Anaerobic a-deaminases, 22

Anaerobic metabolism, amino- acids as H-acceptors, 17

, nitrate as H-acceptor, 39

Anthranilic acid, excretion by Salm. typhosa, 64

, 3-hydroxy-, 72

, in tryptophan synthesis, 64,

. .71

Antibiotic polypeptides, composi- tion, 97

Antibiotics, 151

Arginine, as essential amino-acid,26

— , as possible energy source, 27

— , cycle, 69

— , decarboxylase, 27

— , dihydrolase, 26

— , exacting mutants, 69

Asparaginase, 98

Aspartase, 22, 60

Aspartic acid, deamination, 23

, in arginine synthesis, 70

, in nitrogen-fixation, 55

, transamination, 61-3

Aspergillus nidulans, use of mut- ants, 66, 75

Aspergillus niger, L-amino-acid oxidase, 11

, synthesis of cysteine, 75

, use of mutants, 66

159

i6o

NITROGEN METABOLISM

Aspergillus parasiticus, peptidases,

124 Athiorhodaceae, 5 Aureomycin, 152, 155 Autotrophs, nitrifiers, 32 — , nutrition, 4 Auxotroph, definition, 69 Azotobacter agilis, a-ketoglutarate,

63

, nitrate reduction, 39

, nitrogen fixation, 47

Azotobacter chroococcum, 47 Azotobacter indicum, 47

B12 and pseudo-Bi2, 137, 149, 151 Bacillus, extracellular peptides, 96 — , nitrate reduction, 39 — , proteases, 120 Bacillus anthracis, peptide cap- sule, 96

, transforming factors, 134

Bacillus megatherium, peptidases,

Bacillus mycoides, deamination, 1 1 Bacillus subtilis, asparaginase, 98

, gelatinase, 1 20-1

, mutants, 66

, transamination, 61

, tryptophan oxidation, 15

Bacteriophage, synthesis, 140 Bacterium cadaveris, anaerobic

deaminases, 24 Barbituric acid, 136 Biocytin, 24 Biotin, in ammonia assimilation

by yeast, 25 — , in deaminase activation, 24 Brucella abortus, transamination, 63

Calcium, in formation of extra- cellular proteinases, 120

— , in gelatinase production, 120

— , in nitrification, 33

— , in nitrogen-fixation, 47

Calothrix, nitrogen-fixation, 50

Carbamylglutamic acid, in citrul- line synthesis, 70

Carbon monoxide, inhibition of nitrogen-fixation, 52

Carbon source, 4

■ , carbon dioxide as, 4

Carboxypeptidases, 113

Chloramphenicol (Chloromycetin), 152, 155, 156

Chromatinic bodies, 133

CitruUine, from arginine, 26

Citrulline, from ornithine, 69-71

Clostridium, amino-acid decar- boxylases, 27

— , nitrogen-fixation, 49

— , proteases, 116-20

Clostridium acidi-urici, purine fer- mentations, 136

Clostridium botulinum, extracellu- lar proteinases, 119

Clostridium cochlearum, amino- acid fermentations, 22

Clostridium cylindrosporum, uric acid fermentation, 136

Clostridium histolyticum, extracel- lular proteinases, 117, 119

Clostridium kluyveri, acetylation of amino-acids, 104

Clostridium pasteurianum, isola- tion, 46

, nitrogen-fixation, 51, 54-5

Clostridium propionicum, amino- acid fermentations, 21

Clostridium septicum, arginine di- hydrolase, 26

, proteinases, 117, 119

Clostridium sporogenes, arginine dihydrolase, 26

• , glutamic dehydrogenase, 1 2

, Stickland reaction, 17

Clostridium tetani, amino-acid fer- mentations, 22

, and glutamine, 98

Clostridium tetanomorphum, amino- acid fermentations, 21

Clostridium welchii and glutamine, 98

, collagenase (< toxin), 118

, A toxin, 119

, nitrate reduction, 39

Cobalamins, see B12

Codecarboxylase, see Pyridoxal phosphate

Coenzyme A, 3, 103-4

Collagenase, 116

Corynebacterium, pyrimidine fer- mentations, 136

Corynebacterium diphtheriae, pep- tides for growth, 102

Cystathionine, 74

Cysteine, desulphurase (deami- nase), 23

— , sulphinic acid, 75

— , -S-sulphonate, 75

— , synthesis, 75

Cytosine, 126

— , 5-methyl-, 126

INDEX

;6i

Deaminases, amino-acids, aerobic,

10-17 — , — , anaerobic, 22-6 — , purines, 136 5-Dehydroquinic acid, 73 Denitrification, economic import- ance, 43 — , effect of oxygen, 42 — , organisms, 39 — , pathways, 42

Deoxypentose nucleic acids, in chromatinic bodies,

133 , in transforming factors,

134

■ , in viruses, 140

2-Deoxyribose, 127, 139 Desaturases, see Deaminases, an- aerobic Desulphovibrio, nitrogen-fixation,

50 Diamine oxidase, 155 a,e-Diaminopimelic acid, decar- boxylase, 29

, in lysine synthesis, 75

a,;?-Dihydroxy-iS-ethylbutyricacid,

76. Dipeptidases, 124 Diplococcus glycinophilus, 20 Drugs, mode of action, 145

Energy sources, 4, 17, 20 Energy transfer in biological sys- tems, 3 Enrichment cultures, 20, 46 Enzymes, adaptive, 13 ■ — , control of formation by genes,

^7 . Escherichia coli, amino-acid de- carboxylases, 27

, aspartase, 23

, cysteine desulphurase,

23

, cytidine deaminase, 136

, cytosine deaminase, 136

, deamination of amino-acids,

II

, — of histidine, 22

, dephosphorylation of

nucleotides, 135

, glutathione synthesis, 104

, infection by bacteriophage,

140 , mutants, utilization of

amino-acids in peptides,

100

Escherichia coli, nitratase, 40

, — , synthesis of, 106

, purine interconversion, 138

, purine synthesis, 137

, serine and threonine de- aminases, 23

, transamination, 60

, transforming factors, 134

, tryptophanase, 15

Extracellular products, anti- biotics, 96 — ■ — , enzymes, 112, 115

, peptides, 96-7

• , proteinases, 115

Fermentation, amino-acids, 20

— , purines, 136

— , pyrimidines, 136

Folic acid factors, structure, 98,

Folinic acid (citrovorum factor),

147 Formate, in purine synthesis, 137 Formylase, 13 Formylkynurenine, 13 Fusarium lycopersici, 102 Fusel oil, 22

Gelatinase, 116

Glucose, effect on tryptophanase formation, 15

D-Glutamic acid in extracellular peptides, 96

L-Glutamic acid, conversion to ornithine, 71

, — to proline, 70

, dehydrogenase, 12, 54, 60,

64

, in folic acid factors, 98

, in histidine catabolism, 22

, in nitrogen-fixation, 53

, in strepogenin, 102

, in transamination, 61

, racemase, 95

, uptake by cocci, 86

Glutamine, in citrulline synthesis, 70

— , initiation of growth, 96^ 98

— , synthesis, 104

Glutathione, activation of deami- nases, 24

— , glyoxalase, 99

— , occurrence, 99

— , reductase, 99

— , structure, 99

— , synthesis, 104

1 62

NITROGEN METABOLISM

Glutathione, transpeptidation re- actions, 99

— , triose phosphate dehydro- genase, 3, 99

Glycine, and Lb. easel, loi

— , fermentation, 20

— , in purine synthesis, 137

— , oxidation, 11

Gram stain, 132

Growth, and nucleic acid content, 107

Growth conditions, effect on amino-acid decarboxy- lases, 28

, deaminase activity, 25

, extracellular proteases,

120

, tryptophanase, 15

Guanine, 126

Haemoglobin, in legume root nodules, 57

Haemophilus influenzae, X and V factors, 12

, transforming factors, 134

Haemophilus parainfluenzae, and putrescine, 29

, glutamic dehydrogenase, 12

Haemophilus pertussis and gluta- mic acid, 12

Hansenula, differentiation from Pichia, 43

Hansenula anomala, nitrate reduc- tion, 39

Heterocaryosis, 68

Histidine, a-deaminase, 21

— , fermentation, 21

— oxidation by Ps. fluorescens, 22 Homoserine, 75

Hydrogen acceptors, i Hydrogen donors, i Hydrogenase, and nitrate reduc- tion, 41

— and nitrogen-fixation, 5 1

— and Stickland Reaction, 19 p-Hydroxybenzoic acid, 73 Hydroxylamine, in denitrification,

— , in nitrification, 37 — , in nitrogen-fixation, 55, 57 Hyponitrous acid, in denitrifica- tion, 42

Indole, from tryptophan, 15

• — in tryptophan synthesis, 64, 71

Indoleacrylic acid, 64

Inhibitors, metabolite analogues,

64 Intracellular proteases, 114 Ion transport, 89 Isoleucine and B. anthracis^ 65 — , synthesis, 76 Isotopes and study of amino-acid

synthesis, 77 Isotopic nitrogen and pathways of

nitrogen-fixation, 53

a-Ketoglutaric acid, synthesis, 63 Kynureninase, 14 Kynurenine, in tryptophan de- gradation, 13 Kynureninic acid, 13

Lactobacillus, amino-acid decar- boxylases, 27

— , peptidases, 124

Lactobacillus arabinosus and gluta- mine, 98

and quinolinic acid, 72

, transaminases, 62

, utilization of peptides, 100

Lactobacillus bifidus, serine syn- thesis, 149

Lactobacillus casei, and aspartase,

, folic acid requirement, 98

, purine interconversion, 138

, strepogenin, 102

— — , utilization of peptides, loi

Lactobacillus delbriickii, utiliza- tion of peptides, loi

Lactobacillus helveticus, nucleo- sidases, 135

Lactobacillus lactis and Bic, 149

Lactobacillus pentosus, nucleosi- dases, 136

Leghaemoglobin, see Haemoglo- bin

Leucine and B. anthracis, 65

Leuconostoc mesenteroides, and as- paragine, 98

, and glutamine, 98

, peptidases, 124

, serine synthesis, 148

, utilization of peptides, 100

Lycomarasmin, 102

Lysine, synthesis, 75

— , uptake by cocci, 83

— , — by yeast, 86

Magnesium and extracellular pro- teases, 120

INDEX

163

iVIanganese, activation of pepti- dases, 123

Metabolite analogues, 64

Metals, activation of proteases, 120, 123

Methionine, in purine svnthesis, 137-8

— , sulphonamide antagonist, 147

— , synthesis, 74, 149

Micrococcus aerogenes, purine fer- mentations, 136

Micrococcus anaerobius, glycine fermentation, 20

Micrococcus denitrificans, nitrate reduction, 39

Micrococcus lactilyticus, purine fermentations, 136

Micrococcus variabilis, glycine fer- mentations, 20

Molybdenum and nitrogen-fixa- tion, 50

Mutants, methods of isolation, 66-7

— , production, 66

— , spontaneous, 66

— , use of penicillin, 68

Mycobacterium, pyrimidine fer- mentations, 136

— , streptomycin and diamine oxidase, 155

Mycobacterium tuberculosis, pepti- dases, 124

Neiirospora, adenosine deaminase, 136

— , arginine synthesis, 69

— , nitratase, 40

— , reproductive cycles, 66

— , transaminases, 61

— , use of mutants, 66

Neiirospora crassa, D-amino-acid ^ oxidase, 11

r , L-amino-acid oxidase, 1 1

, L-glutamic acid dehydro- genase, 12

, nitrate reduction, 39

Nicotinic acid, relation with trj'p- tophan, 72

Nitramide, in denitrification, 41

Nitratase, 40

Nitrate and nitrogen-fixation, 54

• — from nitrite, 32

■ — reduction, 39

Nitric oxide in denitrification, 42

Nitrification, effect of organic compounds, 34

Nitrification, effect of Chloro- mycetin, 38

— , — of potassium chlorate, 38

— , energy relationships, 39

— , hydroxylamine in, 37

— , in soil, 36

— , isolation of organisms, 33

— , soil percolation apparatus, 35

Nitrite, oxidation to nitrate, 32

Nitrobacter zvinogradsky , 33

Nitrogen cycle, 6

Nitrogen-fixation by blue-green algae, 50, 56

by Clostridium, 49

by Desulphovibrio, 50

by leguminous plants, 45

by photosynthetic organ- isms, 50

detection of, 47

effect of pCO, 52

of pHa, 51

of pNa, 51

of trace elements, 50-1

isolation of organisms, 46

pathways, 56

Nitrogen sources, 4

Nitrosomonas europoea, 33

Nitrous oxide, in denitrification, 42

Nitroxyl, 42

Nocardia, pyrimidine fermenta- tion, 136

Nodules, nitrogen-fixation by, 50

Nostoc and nitrogen-fixation, 50

Nucleic acids, chromatinic bodies, 133

, components, estimation of,

129

, composition, 127

, Gram stain, 132

, hydrolysis of, chemical, 130

, enzymic, 134

, microbial, 131

, protein synthesis, 107

, structure, 127

, transforming factors, 134

Nuclecproteins, 127

Nucleosidases, 135

— , in nucleotide synthesis, 140

Nucleosides, structure, 126

Nucleotides, structure, 126-8

Nutrition, 3

Ophiostoma, mutants, 66, 75, 137 Orotic acid, 139 Oxido-reduction reactions, i

164

NITROGEN METABOLISM

Pantothenic acid in Co. A, 150 Penicillin, 151

— and amino-acid uptake, 151 Penicillium notatum, L-amino acid

oxidase, 11

, arginine synthesis, 70

Pepsin, 113

Peptidases, activation of, 123

— , endo-, 113

— , exo-, 113

— , intracellular, 114

— , microbial, 123

Peptide bond synthesis, 103

Peptides, as growth factors, 10 1

— , microbial, composition, 96

— , source of essential amino acids, 100-3

— . synthesis, 103

Phenylalanine and tryptophanase formation, 16

Phosphate compounds, energy- rich, 2

— of nucleic acids and protein

synthesis, 109

Photosynthetic organisms, nitro- gen-fixation, 50

Phytotnonas tumifaciens, pepti- dases, 124

Pichia, differentiation from Han- senula, 43

Plasmin, 122

Plasminogen, 122

Pneumococcus transforming fac- tors, 134

Polyol phosphates in Staph, aureus,

Propionibactena, peptidases, 124

Protein synthesis, 103

Proteins, decomposition, ii2

— , microbial, 95

— , synthesis, 106

Proteolytic enzymes of bacteria,

115

of Clostridia, 116

of mammals, 113

of staphylococci, 122

of streptococci, 122

Proteus, amino-acid decarboxy- lases, 27

— , proteases, 120

— , tryptophanase, 15

Proteus morganii, glutamine de- composition, 98

Proteus vulgaris, L-amino-acid oxi- dase, II

, anaerobic deaminases, 23

Proteus vulgaris, asparaginase, 98

, aspartase, 23

, cysteine desulphurase, 23

, growth on proteins, 122

, nitrate reduction, 40

, nutrition, 4

, transforming factors, 134

, transpeptidation reaction,

100

Pseudomonas, adaptation, 13

— , proteases, 120

— , tryptophan oxidation, 13

Pseudomonas aeruginosa (Ps. pyo~ cyanea), asparaginase, 98

, peptidases, 124

, serine and threonine

deaminase, 23

, transaminases, 61

Pseudomonas denitrificans, nitrate reduction, 39

Pseudomonas fluorescens, aspar- tase, 23

, a-ketoglutarate, 63

, nitrate reduction, 39

, peptidases, 124

, transaminases, 61

Pseudomonas stutzeri, nitrate re- duction, 39, 41

Purines, fermentation, 136

— , oxidation, 136

— , synthesis, 137

Putrefaction, 112

Putrescine, as growth factor, 29

Pyrimidines, fermentation, 136

— , oxidation, 136

— , synthesis, 138

Pyridoxal phosphate in amino- acid decarboxylases, 28

in cysteine desulphurase, 25

in racemases, 95

in D and l serine deami- nases, 25

in transaminases, 61

in ti-yptophan synthesis, 72

in tryptophanase, 16

Pyridoxamine phosphate in trans- amination, 62

Pyrocatechase in Pseudomonas, 14

Quinic acid, 5-dehydro-, 73 Quinolinic acid, 72

Reduction of amino-acids, 17

— of nitrate, 39

— of nitrite, 40

INDEX 165

Rhizobiiwi, cross-inoculation

groups, 46

— , isolation, 46

Rhodomicrohium and nitrogen- fixation, 50

Rhodopseudomonas and nitrogen- fixation, 50

Rhodospirillum and nitrogen-fixa- tion, 50

Ribonuclease, 134

Ribose, 127, 140

Saccharomyces cerevisiae, biotin and ammonia assimila- tion, 25 , glutamic acid dehydroge- nase, 12

, a-ketoglutarate, 63

, peptidases, 124

Salmonella paratyphi, histidine de- composition, 22 Salmonella typhosa, nutrition, 4

, tryptophan synthesis, 64

Serine, fermentation by CI. pro-

pionicum, 21 — , in tryptophan synthesis, 64, 72 — , synthesis, 149 Serine deaminases, 23-5 Serratia marcescens, aspartase, 23

, nitrate reduction, 40

, peptidases, 124

Sewage purification, 32 Shigella paradysenteriae, histidine decomposition, 22

, transforming factors, 134

Shikimic acid, 73

, 5-dehydro-, 73

Simultaneous adaptation, 13 Staphylococcus, serine and threo- nine deaminase, 23 Staphylococcus aureus and penicil- lin, 151

arginine dihydrolase, 26

glutamine synthesis, 104

internal free amino-acids, 82

, polyol phosphates, 133

, protein synthesis, 107

Stickland reaction, 17 Strepogenin, 102

Streptococcus, amino-acid decar- boxylases, 27 — , glutamine decomposition, 98 — , Gp. A, proteinase, 122 Streptococcus faecalis, arginine di- hydrolase, 26 , biotin and aspartic acid, 24

Streptococcus faecalis, folic acid requirement, 98

, transaminases, 61

, uptake of amino-acids, 82

, utilization of peptides, loi

Streptococcus haemolytiais and glutamine, 98

nucleases, 135

streptokinase, 122

Streptococcus lactis, asparagine as

growth factor, 98 Streptokinase, 122 Streptomycin, 152, 155 Sulphonamides, mode of action, 1 46 Synthesis, alanine, 61 — , amino-acids, 60 — , arginine, 69 — , aspartic acid, 60-1 — , bacteriophage, 140 — , cysteine, 73 — , deoxyribose, 139 — , glutamic acid, 60-1 — , isoleucine, 76 — , leucine, 65 — , lysine, 75 — , methionine, 73 — , nucleosides, 140 — , nucleotides, 140 — , ornithine, 71 — , phenylalanine, 73 — , proline, 70 — , purines, 137 — , pyrimidines, 138 — , ribose, 140 — , serine, 149 — , threonine, 75 — , tryptophan, 64, 71 — , tyrosine, 72 — , valine, 76 Syntrophism, 69

Tetrahymena gelei and purines, 138

Thiohacillus denitrificans, nitrate reduction, 40

Thiorhodaceae, 5

Thiosulphate in cysteine syn- thesis, 75

Threonine, 75

— , fermentation, 21

Thymine, 126-7

Tonilopsis utilis, isotopes and amino-acid synthesis, 77

Transaminases, 60

— , bacteria, 6i

— , Neurospora, 61

— , yeasts, 6i

1 66

NITROGEN METABOLISM

Transforming factors, 134 Transpeptidation, 99, 104 Triose phosphate dehydrogenase,

3» 99 Tryptophan and nicotinic acid, 71

— and Salm. typhosa, 64

— cycle, 72

— decomposition by Pseudomo-

nas spp., 13

, aromatic pathway, 14

, quinohne pathway, 14

— peroxidase in Pseudomonas, 13

— synthesis, 64, 71

, inhibitors, 64

Tryptophanase, 15 — , mechanism, 16 Tyrosine, decarboxylase, 27

— , effect on tryptophanase forma- tion, 16

Ultraviolet light, absorption by purines and pyrimidines, 129

, induction of mutation, 66

Ultraviolet light, inhibition of adaptation, 107

, use in photographing chro-

matograms, 131 Uracil, 126

Ureidosuccinic acid, 139 Uridine diphosphate glucose, 127 Uridine-5-pyrophosphate pep- tides, 15s Urocanic acid, 21

Valine and Bacillus anthracis, 65 Vibrio, tryptophanase, 15

Yeast, alcohols from amino-acids,

22 — , aspartase, 23 — , glutamine synthesis, 104 — , nucleosidase, 136 — , purine synthesis, 137 — , transaminases, 61

Zinc, activation of serine deami- nase, 25