Pyrimidine Metabolism in Microorganisms - NCBI

0 downloads 0 Views 10MB Size Report
... as dR; C, cyto- sine; CR, cytidine; CdR, 2'-deoxycytidine; 5- ..... glutamine was established by velocity-substrate ...... taking logarithms, we get log (v/V - v) =.

BACTERIOLOGICAL REVIEWS, Sept. 1970, p. 278-343 Copyright © 1970 American Society for Microbiology

Vol. 34, No. 3 Printed in U.S.A.

Pyrimidine Metabolism in Microorganisms GERARD A. O'DONOVAN AND JAN NEUHARD Department of Biochemistry and Biophysics, Texas A & M University, College Station, Texas 77843, and Enzyme Division, University Institute of Biological Chemistry B, S~lvgade 83, DK-1307, Copenhagen K, Denmark 279 INTRODUCTION ............................................. Abbreviations and Symbol (*) ............................................. 280 DE NOVO SYNTHESIS OF UTP AND CTP .................................... 280 Carbamyl Phosphate Synthetase ............................................... 280 281 Properties ............................................... Genetics (pyrA) ................... ............................ 283 Comparison of different microorganisms ...................................... 284 286 Aspartate Transcarbamylase ............................................... 286 Properties ............................................... 289 Genetics (pyrB) ............................................... Comparison of different midroorganisms ...................................... 291 Dihydroorotase and Dihydroorotate Dehydrogenase .............................. 293 293 Properties ............................................... Genetics (pyrC, pyrD) . ............................................. 294 295 OMP Pyrophosphorylase ............................................... 295 Properties ............................................... 295 Genetics (pyrE) ............................................... 295 OMP Dedarboxylase ............................................... 295 Properties ............................................... 296 Genetics (pyrF) ............................................... UMP Kinase and UDP Kinase ............................................. 296 CTP Synthetase ............................................... 296 296 Properties ............................................... ................................ 298 Genetics (pyrG*) ............. OVERALL REGULATION OF THE DE NOVO PATHWAY ..................... 299 299 Control of Enzyme Activity ............................................. 300 Control of Enzyme Synthesis ............................................... 301 PyrA ............................................... 302 PyrB. PyrC, pyrD, pyrE, pyrF, and pyrG ............................................. 302 Control in different microorganisms ............................................ 302 SUMMARY OF DE NOVO BIOSYNTHESIS ................................... 304 ................... 304 BIOSYNTHESIS OF DEOXYRIBONUCLEOTIDES .......... Ribonucleoside diphosphate reduction in E. coli .................................. 304 305 Thioredoxin ............................................. 305 Thioredoxin reductase ............................................... Ribonucleoside diphosphate reductase ......................................... 306 Ribonucleoside triphosphate reductase of L. leichmannii ........................... 308 310 Conclusion ............................................... 311 BIOSYNTHESIS OF dTTP ............................................... ...............................................

Thymidylate Synthetase ............................................... Properties ............................................... Selection of thymine-requiring mutants ....................................... .............................. Genetics (thyA) ............... Biosynthesis of dUMP ...............................................

From UDP ............................................. From dCTP ............................................... From deoxyuridine ............................................... Two Pathways for Endogenous dTTP Biosynthesis? .............................. Bacillus subtilis .............. ............................... Enterobacteria ............................................... INTERCONVERSIONS OF NUCLEOTIDES ...................................

278

311

311 311 312 312 312 313 314 314 314 314 315

VOL. 34, ,1970

279

PYRIMIDINE METABOLISM IN MICROORGANISMS

METABOLISM OF BASES AND NUCLEOSIDES ........................... .......................................................... Uracil Compounds. ...................................... UMP-pyrophosphorylase (upp*). Uridine phosphorylase (udp*) ............................................... Uridine kinase (udk*)...................................................... ...................................... Deoxyunridine metabolism. Cytosine compounds ...................................................... Cytosine dminse (cod*) ................................................. Cytidine (deoxycytidine) deaminase (cdd)..................................... Cytidine and deoxycytidine kinases ......................................... ................................................................. Permeases. Genetics ................................................................ Discussion................................................................. METABOLISM OF THYMINE AND THYMIDINE............................ Incorporation of Thymine and Thymidine into DNA in E. coli...................... Enzymes ................................................................. Thymidine pophorylase (tpp).............................................. Thymidine kinase (tdk) ....................................................

................................................................ Lactobacillh. ................................................................ Neurospora. THVYMINE-REQUIRING MUTANTS ........................................ Properties ................................................................ High and Low Thymine-Requiring Mutants.................................... Genetics and Regulation of Deoxyribonucleoside Catabolism...................... .............................................. PYRIMIDINE ANALOGUES General Principles.. ......................................................... 5-Fluoro Analogues ....................................................... ............................................................ Aza-Analogues. Cytosine Arabinoside ..................................................... APPENDIX ................................................................ LITERATURE CITED........................................................

316 316 316 316 317 317 317 317 318 318 318 319 319 320 320 320 320 321 321 322 322 322 323 323 324 324 325 325 333 333 334

including uptake, thymine-requiring mutants and 2'-deoxythymidine 5'-triphosphate synthesis A linkage map is included, localizing all the genes mentioned in the text (see Fig. 8). Two concentric circles, the outer one for E. coli and the inner for S. typhimurium, are employed, with the loci mentioned in the text underlined. An appendix appears at the end, listing the properties and genetics of all the enzymes mentioned in the text, in the belief that it will be useful for research workers in the area of pyrimidine metabolism. The review will not deal with the following areas. (i) Thymineless death; this vast and diffuse subject needs a review by itself and is not within the scope of this review. (ii) Pyrimidine metabolism in phage-infected cells; we will refer to relevant work with bacteriophages from time to time but we will not cover this rapidly expanding field in any detail. (iii) Sugar nucleotides; the metabolism of this group of nucleotide derivatives, mainly involved in polysaccharide synthesis, will not be treated. (For remary purpose. These areas are (i) the de novo views, see references 236, 237, 287, 298.) (iv) pathway, its regulation and genetics, (ii) the Methylated bases and pseudouridine; since so-called salvage pathways which are easily these derivatives, found in nucleic acids, most confused in vivo, and (iii) thymine metabolism, probably are synthesized at the polynucleotide

INTRODUCIION It has been some time since a review on pyrimidine metabolism has appeared. (For reviews see references 260 and 323.) Previous reviews were concerned mainly with the biosynthesis of pyrimidine nucleotides, and very little was then known about the regulation of pyrimidine metabolism. We feel that our present knowledge warrants a new review, and unlike previous reviews it will be oriented towards regulation [compare Stadtman (362) and Blakley and Vitols (53)]. Most of the material presented will concern Escherichia coli and Salmonella typhimurium. However, other genera of bacteria will be treated in some detail if a contribution to the understanding of pyrimidine metabolism can be so attained, and many recent experiments with Saccharomyces cerevisiae and Neurospora crassa will be described as an integral part of the whole picture. There are at least three areas which we consider need clarification and this will be our pri-

280

O'DONOVAN AND NEUHARD

level (see references 153 and 402), they will not be treated. Abbreviations and Symbols Ribonucleotide bases are identified as follows: A, adenine; C, cytosine; G, guanine; 0, orotate; U, uracil; and T, thymine. Thus, OMP = orotidine-5'-monophosphate, and UTP = uridine-5'-triphosphate. Deoxyribonucleotides are abbreviated as the ribonucleotides except that they are preceded by a "d" as follows: dCTP = 2'-deoxycytidine 5'-triphosphate and dTMP = 2'-deoxythymidine 5'-monophosphate. Other abbreviations: THFA, tetrahydrofolic acid; DHFA, dihydrofolic acid; PRPP, 5-phosphoribosyl-1-pyrophosphate; PPi, inorganic pyrophosphate; and CoB12, coenzyme B12. For charts and tables, nucleosides are abbreviated. Bases are identified as above, ribosyl group as R, and 2'-deoxyribosyl as dR; C, cytosine; CR, cytidine; CdR, 2'-deoxycytidine; 5FUdR, 5-fluorodeoxyuridine. Asterisks (*), appearing both in the table of contents and in the text, indicate new genotypes proposed in this review. DE NOVO SYNTHESIS OF UTP AND CTP The de novo synthesis of pyrimidine nucleotides has been studied in detail in bacteria (36, 178, 195, 401, 403-405), in fungi (77, 78, 225, 398), and in mammals (166, 279, 367). The pathway appears to be universal-the same sequence is followed in all organisms. The key to the elucidation of the biosynthesis of pyrimidine nucleotides was provided by microorganisms. It was found that orotate could satisfy the pyrimidine requirement in mutants of Lactobacilli (83), Streptococci (337), and Neurospora (254). Subsequent studies with mammalian tissues showed that not only could these tissues convert orotate to pyrimidine nucleotides (23, 193, 322, 327, 388), but they could also convert carbamyl aspartate via orotate to pyrimidine nucleotides (389). In studies with Lactobacillus bulgaricus, an orotate-requiring mutant (399, 400), it was shown that carbamyl aspartate could substitute for orotate and that both compounds could be converted to pyrimidine nucleotides by the organism. An enzyme capable of synthesizing carbamyl phosphate from C02, NH3, and ATP was discovered in Streptococcus faecalis, as well as the enzyme that condenses aspartate and carbamyl phosphate to carbamyl aspartate (204). Corresponding enzymes have since been found

BACTERIOL. REV.

in other organisms including E. coli and S. typhimuriwn. Enzymes which are capable to reversibly transforming carbamyl aspartate to orotate were discovered in the anaerobe Zymobacterium oroticum, isolated by enrichment culture on orotate by Lieberman and Kornberg (243). Extracts of the organism contained enzymes that catalyzed the reversible reactions carbamyl aspartate to dihydroorotate and dihydroorotate to orotate (242, 243, 334). Since Z. oroticum was cultivated in a medium with orotate as the sole carbon source, one might expect the enzymes dihydroorotate dehydrogenase and dihydroorotase, acting catabolically, to be induced in the presence of their substrate, orotate. If this was the case, the two enzymes, although present, may be unrelated to the normal biosynthetic pathway. It was necessary, therefore, to demonstrate that the reactions which convert carbamyl phosphate to orotate were necessary for pyrimidine biosynthesis. This demonstration was beautifully carried out by Yates and Pardee (403) who showed (i) that dihydroorotase, dihydroorotate dehydrogenase, and aspartate transcarbamylase were present in extracts of E. coli grown in the absence of orotate; (ii) that a mutant lacking dihydroorotate dehydrogenase did not make orotate, excreted dihydroorotate and carbamyl aspartate into the medium, and required orotate, uracil, or cytosine for growth; and (iii) that cell-free extracts of E. coli B grown in minimal media converted aspartate plus carbamyl phosphate to carbamyl aspartate, carbamyl aspartate to dihydroorotate, and dihydroorotate to orotate. Thus, it was shown in vitro and in vivo (403) that these reactions were part of the pyrimidine biosynthesis. The enzymes responsible for the conversion of orotate to UMP were discovered in yeasts (245, 246). A specific orotidine-5'-phosphate pyrophosphorylase condenses orotate and phosphoribosyl pyrophosphate (PRPP) to OMP, which is then decarboxylated to UMP by OMP decarboxylase (245, 246). The characterization of PRPP and its role in purine and pyrimidine biosynthesis was discovered in the course of these studies by Kornberg and his co-workers (219). UMP is converted to UTP in two successive steps involving ATP and specific kinases (38, 39, 247). UTP, is subsequently aminated to CTP (192, 240, 241). The entire sequence from C02, ATP, and glutamine is shown in Fig. 1.

Carbamyl Phosphate Synthetase Carbamyl phosphate, the product of the reaction catalyzed by carbamyl phosphate synthetase

VOL. 34, 1970 HC

CARBAMY PHOSP

ATP

~

CPSase exhibits cooperative effects for ATP binding as revealed by a sigmoidal curve in a velocity-substrate plot. The Km for the reaction

MgffKE

Pp

ATE

Pi

LOP

is 8 MrP

_

OkeAdd_

AOP

E

LP

*'-,E e

OM

FIG. 1. Pathway of pyrimidine Ibiosynthesis in Escherichia coli and Salmonella typhimturium. Genetic in? italics. symbols for the enzymes involved are sJ own

(CPSase), plays a dual role of b equally for arginine and pyrimidintbee biosyhequir biosynthesis (5). This fact contributes to many of the unique properties of the enzyme in that its activity is modulated by intermediates of boti arginine and pyrimidine biosynthesis. Pierard and Wiame (307) demonstrated the presence of a single species of glutamine-dependent enzLyme, CPSase, responsible for the synthesis of ca rbamyl phosphate in E. coli. Properties. Anderson and Meister (20) showed that highly purified CPSase froim E. coli B catalyzed the reaction shown in eqtuation I. K+ 2ATP + glutamine + HCO3- + H [20 [2° eing

required

Mg'+

(I)

carbamyl phosphate + Pi + Iglutamate + 2ADP

The enzyme has been purified 300-fold from E. coli B, and the reaction mechatiism has been worked out in detail by Andersor and Meister (20), who have shown that the ftallowing three steps are involved: i

(i) enzyme + ATP + HCO3(CO3PO3) + ADP

enzyme-

(ii) enzyme-(C03PO3) + glutamine + H20 = enzyme-(NH2CO2-) + glutamate + Pi (iii) enzyme-(NH2CO2-) + ATP enzyme + carbamyl phosphate + ADP Glutamine is the preferred amino donor although ammonia does work with a 200-fold higher Km than glutamine (205). Thus, ATP reacts with HC03-, producing an enzyme-bound carbonate phosphate anhydride (reaction i). This reacts with glutamine to produce a carbamate still bound to the enzyme (reaction ii). This carbamate is phosphorylated with a second molecule of ATP, yielding carbamyl phosphate (reaction iii).

X 10-4

substrates,

M. The other

bicarbonate

and glutamine, do not yield sigmoidal curves, but show hyperbolic curves with apparent Km of 12

10-4 M and 3.8

x

CTP

UTP

281

PYRIMIDINE METABOLISM IN MICROORGANISMS

x

10-4

M

respectively (20).

The enzyme activity is subjected to feedback inhibition by UMP and is activated by ornithine alone or IMP alone (18, 21, 305, 306). The mechanism by which UMP (a negative allosteric

effector) or IMP and ornithine (positive allosteric effectors) exert their respective allosteric effects is not known, but the presence of either positive effector (ornithine or IMP) decreases the ATP concentration required for half-maximal velocity; the presence of the negative effector (UMP) has the opposite effect. Thus, the affinity of

the

enzyme

for ATP

appears

to

be altered by

the various allosteric effectors. The vital importance of these allosteric effectors is further seen in sucrose density gradients, in which the rate of sedimentation (high

is dependent present

on

concentration used)

(18).

In the absence of

tion

enzyme

which allosteric effector is

constant,

creases as

in

any

the sedimentadensity gradients, in-

effectors

sucrose

the enzyme concentration is increased

(19; Fig. 2). As might be expected, the increase in

sedimentation constant is facilitated by the presence of IMP or ornithine (positive-effectors) and

by ATP and Mg2+ together;

increase is is present. Thus, a minimal sedimentation constant of about 8.5 is observed with no effectors, at low enzyme concentration; a maximal value of 15.3 is observed in the presence of IMP or ornithine at high enzyme concentration. A monomer-oligomer equilibrium is considered to account for these data (19; Fig. 3). observed when UMP

no

(negative effector)

Is is

z214

3_

-

d----BIATP."f

0'

z

2

12

Ia

W10 Wa

bi9

1o-2

I0-

ENZYME CONCENTRATION.

0

MO/MI

FiG. 2. Effect of carbamyl phosphate synthetase concentration on the sedimentation constant of the enzyme in the presence of the indicated effectors. Open triangles were obtained at 11 C, all the rest at 17 C. From Anderson and Marvin (19).

282

O'DONOVAN AND NEUHARD ORNITHINE

ATP

' [MMONOMER-Il] x [MONOMER-I] s UMP SH

[OLIGOMER]

SH

(MONOMER-I10 [OLIGOMER]

FIG. 3. Effect of allosteric effectors on the rate of inhibition of carbamyl phosphate synthetase by N-ethylmaleimide: a possible mechanism. From Anderson and Marvin (19).

In this very beautiful study, Anderson and Marvin (19) showed the complexity of the control of CPSase. By using N-ethylmaleimide, which inactivates the catalytic activity of CPSase, it was possible to examine under different concentrations of enzyme and substrate the effect of the various effectors on the rate of inhibition by Nethylmaleimide. In the absence of effectors the rate of inhibition was rapid, the rate being dependent on the enzyme concentration; at high enzyme concentration the rate was less. In the presence of UMP the rate of inhibition increased. At low enzyme concentration the rate of inhibition was the same with or without UMP. The addition of ornithine or IMP markedly reduced the inhibition rate. Quite likely the sites at which N-ethylmaleimide binds are not available in the oligomeric form. Thus, a high enzyme concentration in the presence of ornithine or IMP would afford complete protection from N-ethylmaleimide; complete protection would not be available even at high enzyme concentration if ornithine or IMP were missing because under these conditions the enzyme does not exist in toto as an oligomer (Fig. 2). Yet, at low enzyme concentrations the presence of ornithine decreases the rate of inhibition, indicating that the action of ornithine occurs even on the monomeric state. The effect of ATP on the rate of inhibition was extremely interesting. The rate of inhibition in the presence of ATP was considerably enhanced as if the site(s) for N-ethylmaleimide were exposed by ATP. The addition of ornithine along with ATP gave an even more striking result; it further enhanced the rate of inhibition. The addition of UMP together with ATP decreased the inhibition. This is further evidence that the apparent affinity of the enzyme for ATP is markedly affected by the presence of allosteric effectors. These data may appear contradictory at first, but a simple scheme is proposed by Anderson and Marvin (19) as a possible mechanism to explain the data (Fig. 3). The data suggest that in the presence of ornithine the enzyme exists in a conformational state that is different from that which exists in the

BAcrRioL. REV.-

presence of UMP. In the presence of ornithine, the enzyme can aggregate reversibly at high enzyme concentration, the enzyme activity is not appreciably inhibited by N-ethylmaleimide, and the effect of ATP on the rate of inhibition is enhanced. In the presence of UMP, the enzyme cannot aggregate at high enzyme concentration, enzyme activity is appreciably inhibited, and the effect of ATP is greatly decreased. These inhibition studies suggest that the enzyme exists in a conformational state in the presence of ATP plus ornithine which is quite different from that which predominates with ornithine alone or UMP alone. The binding of ATP to the enzyme is probably greatly altered by the presence of ornithine or UMP. In the scheme shown in Fig. 3, the reaction of N-ethylmaleimide is considered to be associated with SH groups. The enzyme is considered to exist in at least three conformational states, two of which are conducible for oligomer formation. In the presence of excess ornithine (binds preferentially to monomer-II), the equilibrium would be shifted in favor of monomer-II which at high enzyme concentration would tend to form an oligomer, merely by mass action. The effect of ATP on the inhibition by N-ethylmaleimide of the activity of the enzyme can be explained by assuming that the inhibitor reacts with an SH group "available" only in monomer-I and monomer-III and not available in the oligomer or monomer-II state. The enhancement of inhibition in the presence of both ornithine and ATP over that with ATP alone can thus be explained. ATP, binding only to monomer-II, makes available reactive group for N-ethylmaleimide. The addition of ornithine enhances this by producing more ATP-reactive monomerII; the addition of UMP has the reverse effect. The approximate molecular weight in the presence of ornithine is 390,000 (15.2S). The enzyme can be dissociated into two different subunits, the molecular weights of each being 50,000 and 160,000 respectively (19). CPSase from S. typhimurium has also been studied in detail. The enzyme has been purified 320-fold by Abd-El-Al and Ingraham (3). The following points can be made about its properties and regulation. The enzyme is inhibited by UMP, and in the absence of UMP it is activated by ornithine. Ornithine will also reverse the inhibition by UMP. A sigmoidal relationship in the case of ATP and a hyperbolic one in the case of glutamine was established by velocity-substrate kinetic studies.- ATP reverses the inhibition by UMP at low UMP concentrations, but UMP does not inhibit the enzyme activity at low concentrations of glutamine. As pointed out by Abd-EI-Al

VPYRIMIDINE METABOLISM IN MICROORGANISMS VOL. 34, 1970

and Ingraham, glutamine thus probably enhances the binding of UMP to the enzyme, no UMP being bound at low glutamine concentrations. In contrast to the E. coil enzyme about which it was reported that IMP activated CPSase (20), no such activation was detected with the enzyme from S. typhimurium (3). One further observation seems pertinent. Temperature has been demonstrated as an important variable in assessing the regulatory properties of allosteric proteins (19, 198, 289, 290, 303, 366). Such effects may account for possible changes of function at low temperature (198, 289). As was shown by Abd-El-Al and Ingraham ornithine was a more powerful activator at 20 C than at 37 C, whereas UMP inhibited equally well at both temperatures. Under approximately physiological conditions, one might anticipate 10-4 M UMP and 10-J M ornithine to be present simultaneously. If such relative conditions are attained, net activation will occur at 20 C, inhibition at 37 C. Genetics (pyrA). In enteric bacteria, carbamyl phosphate synthesis is catalyzed by a single enzyme (307) coded for by the gene pyrA. Its map position is close to 12 o'clock on the E. coli (368) and S. typhimurium linkage maps (see Fig. 8; 341). It is not linked to any of the other pyrimidine genes. Mutants in pyrA have been isolated in both E. coli (307) and S. typhimurium (401). For an enzyme which produces a compound shared by two pathways, one might expect to find different phenotypes among mutants affected in such an enzyme. Accordingly, five different phenotypes have been ascribed to mutants which map within the pyrA gene from E. coil and S. typhimurium. (i) The most common mutation observed is, of course, the one which creates the dual requirement for both arginine and uracil. Mutant forms of pyrA which produce no protein or enzymatically inactive protein will provide such a phenotype (306). (ii) A report by Pierard et al. (306) described E. coil mutants which produce a CPSase that is hypersensitive to UMP. These mutants grow in minimal media alone and in minimal medium plus arginine and uracil but not in minimal medium plus uracil. Although the mutant will grow in minimal medium plus arginine, it will not grow at the same rate as in minimal medium; in arginine the generation time is almost double (306). The strains have been termed uracil-sensitive [urs (306)]. To explain the phenotype of these mutants, Pierard et al. (306) suggested that they had lost the activation by ornithine. However, it is not necessary to assume that the enzyme has lost its activation by ornithine to explain the growth behavior of the urs mutant. If the mutant enzyme is more sensitive to UMP than the parent, a greater supply of ornithine would be required to

283

counteract the hyperinhibition. Thus, in minimal medium this implies a low intracellular level of arginine to allow for the increased rate of synthesis of ornithine. The addition of arginine would not alleviate this problem completely because it would reduce the level of the obligately required activator. This mutation has also been observed and mapped in S. typhimurium (401). (iii) As might be expected, the opposite phenotype has also been observed. Whereas a urs mutant is partially starved for arginine, an ars, or arginine-sensitive strain, is partially starved for uracil. J. Ishidsu (1968, Abst. Proc. XII Int. Congr. Genet., Tokyo, p. 16) reported the isolation of a mutant (ars) of S. typhimurium that grew normally in minimal medium but was strongly inhibited when arginine was added. Growth was normal in arginine plus uracil and in uracil alone. The ars locus was found by Eisenstark to map within the pyrA gene (123). (iv) There are two interesting reports from Ingraham's group which provide the last two phenotypes for mutations within the pyrA gene. The first of these reports describes the most unique of all the phenotypes. It was found in two strains of E. coli B/r which appeared to be typical arginine auxotrophs displaying properties of mutants lacking ornithine transcarbamylase (argF); i.e., citrulline but not ornithine satisfied the arginine requirement for growth. However, on the basis of the presence of ornithine transcarbamylase at wild-type levels, low CPSase activity, frequency of cotransduction of the arginine auxotrophic character and the pyrA gene, and growth stimulation by uracil it is clear that the mutation is in pyrA (5). (v) The fifth phenotype is an arginine cold-sensitive mutant (4) of S. typhimuriwn caused by a mutation in pyrA. The cold-sensitive mutant grows at 37 C but cannot synthesize arginine at 20 C. In fact, wild-type growth at 20 C can be attained only in the presence of arginine and uracil. When grown at 37 C in minimal medium, both aspartate transcarbamylase and ornithine transcarbamylase are derepressed. This would indicate a limiting supply of carbamyl phosphate. This was indeed the case, for after growth in minimal medium at 37 C the activity of CPSase was found to be only 13% of that of the parent strain. The enzyme has been purified 102-fold from this mutant (AA2) and its properties compared with those of the wild strain. Ornithine (10-8 M) activated the purified mutant enzyme 9.4-fold at 37 C and 4.1-fold at 20 C. Wild-type purified enzyme was activated only 2-fold and 2.5-fold at 37 and 20 C, respectively. Thus, the wild-type enzyme is activated to a lesser extent than the mutant enzyme and the temperature dependence is dramatically altered by mutation. The mutant enzyme is five-

284

O'DONOVAN AND NEUHARD

BAC rERIOL. REV.

mine-dependent CPSase (393) and argininespecific glutamine-dependent enzyme (115) have been characterized. Although the common intermediate, carbamyl phosphate, is normally channeled in Neurospora, a mutant in ornithine transcarbamylase (arginine biosynthesis) makes carbamyl phosphate available to the pyrimidine pathway, eliminating the effect of a pyrimidine requirement due to loss of the pyrimidine-specific CPSase. Likewise, a mutant lacking aspartate transcarbamylase (ATCase) makes carbamyl phosphate available for arginine biosynthesis, eliminating the arginine requirement created by a loss of the argininespecific CPSase (115). A single genetic region codes both for the pyrimidine-specific CPSase and ATCase (110, 116, 391, 397). Although the enzyme activities of each can be clearly separated, the two enzymes are nevertheless co-purified in a single enzyme complex. It has been possible to isolate mutants which lack either or both activities, all designated pyr-3 mutants (116, 129) but as might be expected, the frequencies of these three classes of mutants as obtained with different mutagens are not the organisms. In Neurospora there are two distinct CPSases same. With a mutagen which primarily gives (82, 111, 112, 114, 255, 331, 332), one specific for base-pair substitutions (e.g., nitrosoguanidine) pyrimidine biosynthesis and the other specific for pyr-3 mutants have been isolated which lack both arginine biosynthesis. The carbamyl phosphate activities (ATCase- and CPSase-) or which lack produced by one reaction is not freely available CPSase only (ATCase+ CPSase-). The third for the other pathway (113, 116). Due to this class (ATCase- CPSase+) is far less frequent. carbamyl phosphate channeling, mutants in the Of 27 ultraviolet (UV) induced pyr-3 mutants arginine-specific or pyrimidine-specific CPSases (116), 20 lacked both activities (ATCaseare auxotrophic for arginine or pyrimidine, CPSase-), 6 lacked CPSase activity (ATCase+ respectively. Thus, there exist two pathway- CPSase-), and only 1 lacked ATCase activity specific carbamyl phosphate pools which are (ATCase- CPSase+). Of 68 nitrosoguanidineindependently synthesized in wild strains of induced pyr-3 mutants studied by D. F. Caroline Neurospora. Both the pyrimidine-specific gluta- (Ph.D. Dissertation, Univ. of Michigan, 1968),

fold more sensitive to UMP inhibition at 20 than at 37 C. Whereas 10-4 M UMP causes 50% inhibition of activity at 37 C, it exerts nearly maximal inhibition at 20 C. No such temperature gradient was observed for UMP in the wild strain. Although ornithine reverses the inhibition of CPSase by UMP and strongly activates the mutant enzyme in the absence of UMP, nevertheless at various concentrations of inhibitor and activator the net result is always lower relative activity at 20 than at 37 C. At 10-4 M concentration of each, net inhibition occurs at 20 C, net activation at 37 C. The cold-sensitive phenotype is therefore a consequence of the altered regulatory properties of CPSase. In light of the recent studies of Anderson and Marvin (19), this is not a surprising result. Table 1 summarizes the properties of the five phenotypes. Comparison of different microorganisms. Even though a single CPSase provides carbamyl phosphate for both arginine and pyrimidine biosynthesis in the enteric bacteria (4, 21, 305) and probably also in Bacillus subtilis (201), this does not appear to be universally the case in micro-

TABLE 1. Phenotypes restluing from pyrA mutationsa Ability to grow i nb

Organism

Phenotype

Minimal medium

Double auxotroph

Plus arginine

S. typhimurium S. typhimurium E. coli

+

E. coli S. typhimurium

are from reference 4. b Symbols: +, growth; -, no growth.

aData

+ +

+

Arginine auxotroph Cold sensitive at 37 C Cold sensitive at 20 C

Plus arginine and uracil

Genotype

pyrA pyrA

S. typhimurium

E. coli

Arginine sensitive Uracil sensitive

Plus

uracil

ars urs urs

VOL. 34, 1970

PYRIMIDINE METABOLISM IN MICROORGANISMS

285

biosynthesis. In contrast to Neurospora, however, either enzyme system alone will permit growth, indicating that there is no complete channeling of carbamyl phosphate in yeasts. Strains which lack both enzyme activities require arginine and uracil for growth (226). As in Neurospora, the first two enzymes in pyrimidine biosynthesis in yeast, namely the pyrimidine-specific CPSase and ATCase, are coded for by one genetic region, ura-2 (225, 226). Both enzymes are associated in a single enzyme complex (256). A purification technique designed to isolate ATCase, in fact, produced a complex, wherein the activity for CPSase greatly exceeded that of ATCase. UTP is the feedback inhibitor of both activities (225, 256). The molecular weight of the undissociated complex in the presence of UTP, glutamine, and Mg2+ is 800,000. In the absence of UTP, a complex of molecular weight 380,000 is found displaying both activities. The CPSase was found to be highly sensitive to UTP; ATCase was much less sensitive. Omission of UTP, glutamine, and Mg2+ gives a CPSase enzyme of molecular weight 250,000 which was still feedback inhibited by UTP. (Lue and Kaplan, unpublished data; J. G. Kaplan, personal communication). Heat or chromatography on diethylaminoethyl-Sephadex gave the appearance of an ATCase peak at 140,000 molecular weight (256, 257). No inhibition by UTP nor any CPSase activity is observed in this fully dissociated enzyme. It is clear that the fully aggregated complex has both ATCase and CPSase activities as well as inhibition by UTP. In the fully dissociated state only ATCase activity is found; feedback inhibition and CPSase activity are lost. The enzyme complex under the control of ura-2 may be an operon with two, or probably three, different polypeptide chains being made. One polypeptide would have ATCase activity, one would have CPSase activity, and a third would bind the allosteric effector, UTP. This hypothesis, favored by Lacroute, is currently being tested. A very similar case has been studied by Cohen et al. (92) in the threonine pathway of E. coli. Two activities and a common regulatory site have been shown. As pointed out by Lacroute (225), even though the economy of having two enzymatic activities with CPSase+ ATCase+ CPSse one regulatory site is obvious, nevertheless freone of the enzymes present in the complex quently ATCaoS CPSos6 corresponds to an enzymatic activity present in duplicate in the cell. A possible role in the pref- -pyrY3 Q _.erential channeling of key metabolites is most likely involved (114, 115, 257). direction of translation Recently ATCase- CPSase+ mutants have been FiG. 4. Pyrimidine-3 genetic region in Neurospora shown to map at one end of the ura-2 genetic region. Double mutants (ATCase- CPSase-) crassa showing the three classes of mutants found. map throughout the rest of the region but not in From Radford (318).

13 lacked both activities (ATCase- CPSase-), 54 lacked CPSase (ATCase+ CPSase-), and again only 1 lacked ATCase (ATCase- CPSase+). The 27 UV-induced mutations have been mapped (365) and the 1 ATCase- CPSase+ was located at one end of the recombination map. Another similar mutant (ATCase- CPSase+) also mapped at that end (396). The other two classes lacking both activities (ATCase- CPSase-) or just the CPSase activity (ATCase+ CPSase-) were scattered throughout the pyr-3 genetic region. If a mutagen which primarily induces frameshifts is used (e.g., acridine mustard ICR-170), a very different result is obtained. Here, the class that was extremely rare with ultraviolet- or nitrosoguanidine-induced mutations (ATCaseCPSase+) was found to be quite common (317, 318). In fact, the most common class observed with base-substitution mutagens (ATCase+ CPSase-) was not found at all when ICR-170 was used as a mutagen. The other two classes (ATCase- CPSase+, ATCase- CPSase-) were quite frequent. Reversion studies have verified the existence of a frame-shift for those pyr-3 mutants which lack ATCase only [ATCase- CPSase+ (318)]. Furthermore, polarized complementation was shown with most of the ICR-induced mutants. In essence then, all ATCase- CPSase+ mutants map to the left of the pyr-3 genetic region (Fig. 4). Any frame-shift mutation in pyr-3 which destroys CPSase activity will also destroy ATCase activity. In Fig. 4, the direction of translation would be from right to left, suggesting that the predominant class of frame-shift mutants with one activity would be ATCaseCPSase+. This is exactly what is found. The proximal end of the genetic region controls CPSase activity; the distal end controls ATCase (318). In the yeast, Saccharomyces cerevisiae, results similar to those of Neurospora have been obtained. There are two distinct enzymes that produce carbamyl phosphate in yeasts (227), and mutations which affect each one separately have been found. As expected, one enzyme is related to pyrimidine biosynthesis, the other to arginine

IATCaso

286

O'DONOVAN AND NEUHARD

the ATCase end (M. Duphil-Denis and J. G. Kaplan, personal communication). Complementation studies indicate that ATCase- mutants (ura-2) complement CPSase- mutants (cpu); the cpu-ura-2 double mutants do not complement. Neither do any of the CPSase- (cpu) mutants complement another; certain ATCase- mutants can complement certain other ATCase- mutants at very low frequency. It can be concluded from genetic and enzymological data that the ura-2 region is an operon with at least two structural genes, one for ATCase and one for CPSase. Since feedback-resistant mutants have now been isolated (209), it will be possible to ascertain the genetic locus of these mutants and to determine whether a third polypeptide chain exists which binds UTP on its regulatory site (J. G. Kaplan, personal communication). Finally, the transcription (or translation) is from CPSase to ATCase, precisely as was shown in Neurospora.

Aswrtate Transcarbamylase ATCase catalyzes the first reaction unique to pyrimidine biosynthesis and is regulated by feedback inhibition in E. coli (404). The reaction is shown in equation II. Carbamyl phosphate + L-aspartate carbamyl aspartate + Pi (II) Properties. The native enzyme which has been purified from E. coli (146, 149, 348) has a molecular weight of 310,000. Using a specially constructed strain of E. colf, diploid for the region of the chromosome which contains the cistrons for ATCase (pyrB), it was possible to force this organism to make 8% of its total protein as ATCase (146). Five grams of enzyme is available from about 700 g of wet packed cells (146). The enzyme has been crystallized in a variety of forms, such as very large canoe-shaped crystals (200 um in length), small octahedra, and square plate

(392).

Purified ATCase contains two kinds of protein subunits (149), one responsible for all of the catalytic activity and accordingly designated the catalytic subunit; the other is responsible for the affinity for the feedback inhibitor and is designated the regulatory subunit (145). The properties of each of these proteins will be discussed separately below. However, the native enzyme is most probably a hexamer, composed of six regulatory and six catalytic polypeptide chains (387). The native enzyme from E. coil is feedback inhibited by CTP > CDP > CMP > cytidine. Cytosine, at the base position, is a prerequisite for

BAcrERIOL. REV.

inhibition; uridine derivatives do not inhibit the native enzyme. No inhibition is seen with the base cytosine. However, deoxycytidine derivatives are as effective as cytidine compounds; in fact, dCTP is the best natural inhibitor found (147). ATP activates the native enzyme, and both ATP and CTP appear to compete for a single site on the regulatory subunit (81). The kinetic properties of the native enzyme have been studied in great detail. The enzyme shows sigmoidal dependence on aspartate concentration when velocity is plotted as a function of substrate concentration (147). The Km for aspartate is 5.5 mm. It has been reported that the carbamyl phosphate saturation curve for ATCase is hyperbolic (147, 216, 328, 404). In these studies, however, sufficiently low carbamyl phosphate values were not tested to allow a precise determination of the true shape of the curve. In Kleppe's study in which he estimated a Km of 0.05 mm for carbamyl phosphate, the lowest carbamyl phosphate concentration was 0.3 mm. As pointed out by Bethell et al. (46), at this concentration the rate of reaction is still 80% of the maximal velocity. Using a more sensitive radioactivity assay (46) and using carbamyl phosphate values in the required range, the velocity-substrate plot for carbamyl phosphate was shown to be sigmoidal. A Lineweaver-Burk plot for the native enzyme for carbamyl phosphate is linear up to 0.14 mM; then the curve increases (becomes concave upwards) rapidly at low 'substrate concentration. The apparent Km is 0.2 mm [compare Km of carbamyl phosphate for catalytic subunit 0.05 mM (46)]. Thus, both aspartate and carbamyl phosphate alter the subunit interactions of native ATCase, creating cooperative binding effects and exhibiting sigmoid saturation kinetics. The addition of CTP, the feedback inhibitor, enhances the sigmoidicity for both carbamyl phosphate and aspartate. At high carbamyl phosphate concentrations, CTP decreases the affinity of ATCase for aspartate; at high aspartate concentrations, CI7P decreases the affinity of ATCase for carbamyl phosphate. When both substrates are saturating, very little inhibition can be seen (46, 147, 148). CTP is therefore a competitive inhibitor for ATCase and the inhibition is determined jointly by the concentration of both allosteric substrates, carbamyl phosphate and aspartate. These results are lucidly depicted by Bethell et al. (46). The effect of temperature on purified ATCase from E. coli has recently been studied. Whereas velocity-substrate plots always give sigmoid kinetics at high temperatures (> 20 C), this does not occur at low temperatures (< 10 C). At 4 C, a Michaelis-Menten hyperbola is observed in the

VOL. 34, 1970

PYRIMIDINE METABOLISM IN MICROORGANISMS

287

absence of CTP; sigmoid kinetics are observed aspartate dissociates first and phosphate dissoin the presence of CTP. As was seen for yeast ciates second. ATCase (208), inhibition decreases with decreasThe Km for aspartate is very much higher (2 x ing temperature (J. A. Fuchs and G. A. O'Dono- 10-2 M) than the Km for carbamyl phosphate van, Bacteriol. Proc. p. 132, 1970). (1 X 10' M). Because of this fact, the usual Treatment of the native enzyme from E. coli technique of varying the respective substrate conwith 1.0 M urea or high pH has been reported to centrations in their Km ranges was very difficult. cause a loss of sigmoidal kinetics although At low carbamyl phosphate and high aspartate sensitivity to CTP inhibition was not destroyed another problem arose-the rate of reaction was (390). By treating the native enzyme with mercu- too rapid for convenient measurement and the rials or mild heat, the enzyme is rendered insensi- usual procedure of using very low concentrations tive to CTP. This desensitization is accompanied of enzyme did not apply because of its instability, by the dissociation of native ATCase into sub- as mentioned above. For these reasons, a new and units, one of which contains all of the catalytic more sensitive assay scheme determining the activity, has no regulatory sites for CTP, and has kinetic parameters of the catalytic subunit was a sedimentation coefficient of 5.8S. It is termed worked out. [For discussion see Porter et al. the catalytic subunit. The other component from (310).] the mercurial treatment is a smaller noncatalytic Using this extremely sensitive assay scheme, protein with a sedimentation coefficient of 2.8S, the catalytic subunit was shown to exhibit simple bearing all the sites for regulation by CTP. It is Michaelis-Menten kinetics at low substrate termed the regulatory subunit. Both proteins will concentrations. A hyperbolic curve was noted now be treated in detail. when carbamyl phosphate was varied with After treatment of the native enzyme with p- limiting aspartate (46, 310). Similarly, no cooperhydroxymercuribenzoate, two classes of protein ativity was ever observed with velocity-substrate subunits are found. The larger of these possesses plots even when very low concentrations of all the catalytic activity of the native enzyme and aspartate were tested in the presence of saturating must therefore bear the active sites for the sub- (10-w M) and half-saturating carbamyl phosphate strates, carbamyl phosphate and aspartate. It is (2.5 x 10- M) (310). termed the catalytic subunit (145, 149). It is Several analogues of aspartate were examined insensitive to the feedback inhibitor CTP and for inhibition. As shown previously by Gerhart to the activator ATP. It has a molecular weight and Pardee (147), for the native enzyme, sucof about 100,000 and a sedimentation coefficient cinate and maleate were good competitive inhibiof 5.8S (145). Weber (387) estimated the molec- tors with aspartate. Since malonate is not nearly ular weight of a catalytic chain at 33,000 to 34,000, as effective as succinate, and glutarate does not so the catalytic subunit contains three binding inhibit, a good fit of a four-carbon dicarboxylic sites per 100,000 molecular weight and is a trimer acid into the aspartate site is required. Moreover, (K. Weber, personal communication). There are since fumarate does not effectively inhibit, two catalytic subunits per native enzyme (145, aspartate and succinate would seem to bind in 149), so there are most likely six identical poly- conformations akin to the cis-configuration of peptide chains in the catalytic subunit (387). maleate (310). This is in agreement with the existence of six Competitive inhibition was also observed for regulatory polypeptide chains in the native enzyme, analogues of carbamyl phosphate. All commaking ATCase a hexamer (387). pounds containing a phosphate or a phosphonate The catalytic subunit is stable for several dianion inhibited the enzyme, indicating that the months at 4 C when kept at concentrations above phosphate-binding site is easily accessible (310). 1 mg per ml (146). In dilute solutions, however, The isolated catalytic subunit at low aspartate the enzyme is slowly and irreversibly inactivated. concentrations is four times as active as the Below 1 ,g per ml, the specific activity decreased native enzyme, indicating that the affinity of the with decreasing enzyme concentration (46, 310). catalytic subunit for aspartate is greater than In electrophoresis, the catalytic subunit moves that of the native enzyme. The coaggregation of as a particle slightly more negatively charged than each catalytic subunit with another catalytic and the native enzyme, and it comprises 68% of the appropriate regulatory subunits causes a loss in total protein of ATCase (150). activity as if a metabolite became bound to the Recently, Stark and co-workers performed an subunit. As pointed out by Gerhart and Schachelegant series of kinetic studies on the catalytic man (149), the regulatory metabolite CTP may subunit of ATCase (99, 310, 344). The binding of be considered a co-inhibitor (compare co-repressubstrates by the subunit is ordered. Carbamyl sor) acting on an apo-inhibitor, the regulatory phosphate binds first, aspartate second; carbamyl subunit. Thus, the regulatory subunit is only a

288

BACTrERIOL. REV .

O'DONOVAN AND NEUHARD

The regulatory subunits of ATCase represent class of proteins produced by the cell not for catalytic activity but for the control of catalytic activity (149). Two features of this control seem pertinent: (i) a very large fraction of the total protein of ATCase (32%) is exclusively involved in this regulation, and (ii) the presence of the regulatory subunits seem to decrease the catalytic activity of the native enzyme. Table 2 summarizes some of the properties of native ATCase and its catalytic and regulatory subunits. The dissociation of native ATCase by mercurials is readily reversed by the addition of the sulfhydryl compound mercaptoethanol (149). The reconstituted enzyme is similar to the native enzyme from which it was dissociated, in molecular size (11.8S), in sensitivity to CTP, and in catalytic properties. This indicates that the native enzyme spontaneously aggregate from its respective sub-units. Recently, ATCase has been purified from a related genus, Salmonella, and its enzymatic properties have been studied (O'Donovan, Holoubek, and Gerhart, in preparation). Striking similarities exist for the two enzymes. ATCase was purified from a constitutive strain of S. typhimurium which produced about 2% of its protein as this enzyme. A cooperative saturation curve for aspartate, inhibition by CTP, and activation by ATP all resembled the properties found for the E. coli enzyme. However, the extent of CTP inhibition was clearly distinguishable from that found for the E. coli enzyme (Fig. 5A) in which it was possible to inhibit maximally only to about 80%; with the S. typhimurium enzyme up to 95% inhibition was attained. The protein has a molecular weight of about 300,000 and a sedimentation coefficient of about

partially active inhibitor of the catalytic subunit, the fully active inhibitor being the CTP-regulatory complex. The second kind of protein in native ATCase, isolated after treatment with p-hydroxymercuribenzoate, has no catalytic activity and is not necessary for the activity of catalytic subunit (149). The protein binds CTP and thus must contain all six regulatory sites found in the native enzyme. The term regulatory subunit is a misnomer if taken literally. We use it with caution and only because of its now almost universal use. It does not mean that the entire regulation of native ATCase is in the domain of the regulatory subunit to the exclusion of any regulatory role by the catalytic subunit. On the contrary, both types of subunits, taken together, produce complementary regulatory effects (149). The regulatory subunit is stable, being able to retain its affinity for the effectors CTP and ATP. When the dissociating agents (mercurials) are removed, the regulatory subunit does not aggregate. Its electrophoretic migration is that of a particle more positively charged than the native enzyme (149). Weber (387), in his studies on the primary structure of ATCase, worked out the amino acid sequence of the regulatory subunit as well as its molecular weight. After denaturation with sodium dodecylsulfate, he determined the carboxy termini and estimated the molecular weight of a regulatory chain to be 17,000. Since the value of the phydroxymercuribenzoate dissociated regulatory subunit has been reported to be about 30,000 to 36,000 (149, 150; K. Weber, personal communication), it is likely that there are two regulatory polypeptide chains per isolated subunit. The isolated subunit exists primarily as a dimer, tending towards a monomer at low dilution.

a

TABLE 2. Properties of native ATCase and its catalytic and regulatory subunitsa mercurials

Property

Native enzyme

Subunit structure ................... Molecular weight ................... Sedimentation coefficient Weight fraction of subunits Molecular weight of each polypeptide chain Number of polypeptide chains Total number of polypeptide chains ............................ Number of half-cystines Shape of saturation curve ........... Catalytic activity (binding of substrates) ....................... Binding of effectors ................. .............

a

Data are from Gerhart and Schachman

'mercaptoethanol

6C+6R

310,000 11.8S

2 (Catalytic)

3

(Regulatory)

2(3 C) = 6 C + 3(2 R) = 6 R 3 (34,000) 2 (100,000) + 2.8S 5.8S 0.68

0.32

33,000 3

17,000 2

Sigmoidal

2(3) = 6 8 Hyperbolic

+

+

+

-

12 32

+

3(2)

=

6

24-28 -

+

(149, 150) and Weber (387 and personal communication).

VOL. 34, 1970

PYRIMIDINE METABOLISM IN MICROORGANISMS

MOLARITY OF L-AS:ARTATE x 03 Fio. 5. (A) Velocity-substrate plots of native aspartate transcarbamylase from Escherichia coli (open squares) and Salmonella typhimurium (open circles). Closed squares and circles denote the respective results in the presence of 0.2 mm CTP. (B) Reconstituted native enzymes from E. coil and S. typhimurium are shown with open squares and circles (as above). Reconstituted hybrid enzymes: E. coli catalytic, S. typhimurium regulatory (A); S. typhimurium catalytic, E. coll regulatory (V).

12S. Moreover, it can be desensitized by treatment with mercurials in a reaction leading to dissociation. Thus, the S. typhimurium enzyme can be separated into its respective subunits which display activities similar to those observed for the E. coil subunits. The catalytic subunit has a sedimentation coefficient of about 6.OS and exhibits MichaelisMenten kinetics when velocity-substrate plots are made. It does not show inhibition by CTP or activation by ATP. It has been possible to reconstitute the native S. typhimurum enzyme from its dissociated subunits as a 12S aggregate displaying sensitivity to CTP, activation by ATP, and catalytic properties like the native enzyme of S. typhimurium (Fig. 5). The reason for the above study was not to observe the properties of ATCase in another organism but to use ATCase, in a manner similar to that of Bethell and Jones (45), for measuring

289

the evolutionary relatedness of different genera. A stringent test of evolutionary relatedness would certainly be shown by the ability of two different genera of bacteria to produce regulatory enzymes which could be dissociated into their respective subunits and then reconstituted to form functional hybrid enzymes. This has been achieved for the ATCase from E. coli and S. typhimuriwn. Purified ATCase from both genera was treated with a mercurial, and regulatory and catalytic subunits were produced. Homologous regulatory enzymes (catalytic from one genus with regulatory from the same genus) were reconstituted from both genera and used as controls. Heterologous subunits (catalytic from one with regulatory from the other) gave two classes of hybrid enzymes whose properties differed somewhat from their original native enzymes. Velocity-substrate curves for the E. coli and S. typhimuriwn native enzymes (and reconstituted native enzymes) are virtually identical (Fig. 5A). However, it was found that the hybrid enzymes both gave sigmoidal curves but the catalytic subunit from E. coil reconstituted with the regulatory subunit from S. typhimuriwn gave weaker cooperative effects (less sigmoid) than either native or other hybrid enzyme [catalytic from S. typhimurium and regulatory from E. coli (Fig. 5B)]. The significance of this study lies in the fact that it may be possible to reconstitute native ATCase from a wide variety of related and unrelated organisms and thus find some clue to the evolution of regulatory enzymes. It is not necessary to use pure enzyme for a gross estimate of this kind. Crude extracts will suffice in most cases. Recently, we have been able to obtain enzyme from B. subtilis which is unique in that it contains a noninhibitable ATCase (45) with a sedimentation coefficient of about 6.OS. It may be equivalent to the E. coil catalytic subunit. As mentioned elsewhere in this review, a coaggregate with the B. subtilis enzyme and either E. coli or S. typhimuriwn regulatory subunits has not been achieved. These and related studies are now in progress in our laboratory. Genetics (pyrB). ATCase is coded for by the gene pyrB and is located on the linkage map of E. coil and S. typhimurium at about 11 o'clock (see Fig. 8; 341, 368). The gene pyrB is unlinked to any of the other genes involved in pyrimidine biosynthesis in both E. coli and S. typhimurium. Unfortunately, very little is known about the genetics of this protein. Mutants have been isolated which lack ATCase in both E. coli (36) and S. typhimuriwm (401). However, although there exist many mutants which effect or eliminate the catalytic activity, so far no mutant has been shown which lacks only the regulatory subunit. Mutants which affect the catalytic subunit have an absolute

290

BAMrRIOL. REV.

O'DONOVAN AND NEUHARD

requirement for pyrimidines; mutants which lack only the regulatory subunit would not be expected to have a requirement for pyrimidines and would be relieved of feedback inhibition by CTP, since its binding site would be eliminated. Even though it has not been possible to isolate feedback-resistant mutants for ATCase, it has been possible to isolate mutants which are somewhat modified in their feedback inhibition by CTP. Among pyrimidine regulatory mutants, one would anticipate at least two classes of mutants: (i) feedback-resistant mutants and (ii) constitutive mutants. However, the conventional selection techniques do not work for the isolation of regulatory mutants in pyrimidine biosynthesis. The feedback inhibitor is a cytidine nucleotide and since phosphorylated compounds do not permeate the cell membrane in E. coli or S. typhimurium, only cytosine nucleosides can be used. Accordingly, many cytidine analogues have been tested in strains lacking the deaminases which convert cytosine compounds to uracil compounds. No feedback-resitant mutant has been isolated. Regulatory mutants of the pyrimidine pathway would be expected to over-produce pyrimidines and pyrimidine intermediates and excrete them into the medium. A similar observation has been made for the histidine biosynthetic pathway (347). Since there is no false feedback inhibitor

available which is effective in pyrimidine biosynthesis [compare 2-thiazolealanine for histidine, in the histidine pathway (347)] there is no selection available for the isolation of feedbackresistant mutants. However, by constructing a very sensitive indicator bacterium in a pyrA deletion able to use carbamyl aspartate, dihydroorotate, and orotate as well as uracil to satisfy its pyrimidine requirement, it was possible to identify excreting colonies by their cross-feeding of the indicator strain. It was necessary to starve the indicator bacterium for uracil prior to its incorporation on the plate, which contained ornithine to assure an unlimited supply of carbamyl phosphate and to counter-inhibit the probably large concentration of UMP that might occur in a pyrimidine regulatory mutant (see Fig. 9). About 400 different regulatory mutants from S. typhimurium were thus isolated by their producing zones of secondary growth (halos) about them. It was hoped that among these mutants some feedback-modified ATCase mutants would occur. To selectively screen for these mutants, transduction was employed in a manner similar to that used previously by Somerville and Yanofsky (360). Inherent in the selection is the assumption that the cistrons for the regulatory and catalytic polypeptides are adjacent on the chromosome of S. typhimurium. By transducing the regulatory mutants into a pyrB deletion (no

TABLE 3. Properties of parental, feedback-modified, and constitutive strains of Salmonella typhimuriuma Ratio of ATCase StnUracilb Strain Description | | Uracilb

ATCase | pactivityc specific

DHO dehase

activity'd specific

specific activity uracil

-

Per cent inhibition

CTP ofbyATCase

+ uracil

LT2 LT2 HD23 HD23 HD27 HD27 HD58 HD58 HD47 HD47

Wild strain

pyrC138

pyrC138

Feedback modified

+ _ + _

Feedback modified

+

Constitutive

Constitutive

+ _ + _

Pyrimidine auxotroph

+

Limit

0.06 0.25 0.05 0.06 0.05 0.06 6.4 6.3 7.9 6.0 0.12 1.00

0.11 0.16 0.09 0.14 0.14

4.1 1.2

0.23

1.2

0.38

1.0

0.65 0.62 0.12 0.31

0.8

0.41

8.3

87 85 35 30 38 35 88 90 88

88 80 85

Data are from O'Donovan and Gerhart, in preparation. bUracil (50 ,g/ml) was used where stated (+), except for pyrC138 which was grown in 6 jug of uracil per ml and then starved for 2 hr (limit). c ATCase activity was measured as previously described (147), using 5 X 108 cells per tube, at 30 C for 30 min. d DHO dehase (dihydroorotate dehydrogenase) activity was measured by using a modification of previously described methods (36, 371) with 5 X 108 cells per tube at 30 C for 30 min. e Inhibition was measured by using 5 mm aspartate and 0.2 mM CTP. a

VOL. 34, 1970

PYRIMIDINE METABOLISM IN MICROORGANISMS

ATCase activity; pyrimidine-requiring), selecting for pyrimidine prototrophy, and testing among the prototrophic transductants for pyrimidine overproduction on indicator plates, it was possible to isolate 70 feedback-modified mutants. By their selection all the mutants map within the ATCase cistrons, all excrete uracil into the medium, and all have reduced inhibition by CTP. Table 3 compares two such mutants (HD23 and HD27) with the wild strain. The interesting thing was that without exception each mutant had greatly repressed levels of ATCase, making it difficult to study enzyme kinetics. To study the kinetics of ATCase from these mutants in more detail, a constitutive mutation was transduced into some of the feedback-modified mutants. Further studies on these partially desensitized mutants are in progress. The second class of regulatory mutants isolated in this study was found to be unlinked to pyrB or to any of the other genes involved in pyrimidine biosynthesis. These are constitutive mutants. They produce extremely high levels of the pyrimidine enzymes in the presence or absence of uracil (Table 3). Unlike the wild strain and a pyrimidine-requiring mutant, no repression is seen when cells are grown in uracil; the ratio of ATCase enzyme activity in the absence to the presence of uracil is about one (Table 3). Though the activity of ATCase is very high in these strains, nevertheless it is feedback inhibited exactly like the parent strain. Comparison of different microornis. Even though the pyrimidine biosynthetic pathway has a common sequence for all organisms, the control of enzyme activity is extremely varied. Since ATCase is the first enzyme unique for pyrimidine biosynthesis, one would expect this enzyme to differ from organism to organism. This is exactly what is found and we shall briefly discuss these phenomena, first in bacteria, other than the enterics, and then in Neurospora and finally in Saccharomyces. As has been discussed previously ATCases of E. coli and S. typhimuriwn are very much alike (O'Donovan, Holoubek, and Gerhart, in preparation). Two previous comparative studies have been made on bacterial ATCases (45, 285). In their studies on Pseudomonas (P. aeruginosa and P. fluorescens), Jones and co-workers (45, 285) found that both UTP and CTP were effective feedback inhibitors and that the inhibition was competitive with carbamyl phosphate and not with aspartate. Using a different strain of P. aeruginosa (PA01), Isaac and Holloway found UTP to be an effective inhibitor but the inhibition was reported to be competitive with aspartate (199).

291

The velocity-substrate plots were both hyperbolic but the curve for carbamyl phosphate could be transformed to a sigmoidal shape by the addition of the inhibitor, UTP or CTP. It was not possible to separate the respective subunits of the Pseudomonas ATCases by the usual treatments which gave success in the case of the E. coli and S. typhimurium enzymes. Rather, these treatments (e.g., with p-hydroxymercuribenzoate) tended to decrease in parallel fashion, both catalytic and regulatory functions. Identical kinetics have been observed with ATCase from Azotobacter vinelandii. The enzymes from Serratia marcescens, Proteus vulgaris, Citrobacter freundii, and Rhodopseudomonas spheroides all appear to be quite similar in size and regulation to the E. coli and Salmonella typhimuriwn enzymes (45). The Serratia marcescens enzyme is inhibited by CMP and UMP and on denaturation markedly loses its sensitivity to pyrimidine nucleotides (285). M. R. Bethell (Ph.D. Dissertation, Brandeis University, 1967) found that the C. freundii enzyme was inhibited by CTP but found that the P. vulgaris was activated by the same concentration of CTP (1 mM). The enzyme from Aerobacter aerogenes has also been shown to be inhibited by CMP and UMP (285). A most interesting class of ATCase enzymes is certainly the one which appears to resemble closely the catalytic subunit of E. coli. The similarity includes both size and kinetics. Such enzymes have been observed in B. subtilis and Streptococcus faecalis. As might be expected, velocitysubstrate plots were found to be hyperbolic and no feedback inhibition was ever seen with pyrimidine nucleotides (45, 285). One other observation (45) seems pertinent. Extracts from C. freundii and P. vulgaris seem to contain both the E. colitype enzyme and the B. subtilis type enzyme. The dissociation of the native E. coli-type enzyme may be giving rise to the two classes found, but as pointed out by Bethell and Jones (45) there may be two distinct enzymes serving two different metabolic functions. The notion that an anabolic (E. coli-type) and a catabolic (B. subtilis-type; no inhibition) ATCase might coexist is intimated by a similar observation of Taylor et al. (372) for dihydroorotate dehydrogenase. ATCase has also been partially purified from an obligate halophile, Halobacterium cutirubrum, in which activity was dependent on a high salt concentration requiring up to 4 M NaCl or KC1. Feedback inhibition by CTP was observed, and the inhibition required a high salt concentration. The inhibition was quite extreme; 1 mM CTP gave 70 to 80% inhibition and 5 mm gave 90 to 96%. Velocity-substrate plots gave hyperbolic

292

O'DONOVAN AND NEUHARD

BACTERIOL. REV.

curves for aspartate and carbamyl phosphate plots are made for aspartate. No kinetic data are (248). available to us for the other substrate, carbamyl In the lactic acid bacterium L. leichmannii, no phosphate. The Km values for aspartate and feedback inhibition by pyrimidine nucleotides carbamyl phosphate, (compare the catalytic subwas observed for ATCase. However, a most unit of E. coli) are high for aspartate (7 mM, pH unusual pyrimidine pathway emerges from a 9.5) and low for carbamyl phosphate (0.7 mM, study of this organism. It appears that the guani- pH 9.2). The inhibition and Km values are very dino carbon of arginine replaces CO2 as a pyrimi- dependent on pH. In general the enzyme is dine precursor. Thus, the first step in pyrimidine inhibited by uridine compounds (UTP, UMP, biosynthesis in L. leichmannii is catalyzed by the uridine, thymidine, OMP) but not by cytidine enzyme arginine deiminase, which converts compounds (CTP, cytidine). Maximal inhibition arginine to citrulline. The citrulline then con- which has been attained is about 80% and the tributes carbamyl phosphate for the ATCase kinetics of the inhibition are such that it occurs reaction, which utilizes aspartate in the usual only when the enzyme is fully saturated with way. Deoxythymidylate (dTMP) is a feedback aspartate (119). This latter fact is exactly opposite inhibitor of the first enzyme, arginine deiminase to what is found with E. coli in which at high (195). aspartate concentrations (15 mM) virtually no Several useful criteria were examined by Bethell inhibition (by CTP) is observed. and Jones (45) to ascribe some possible taxonomic Studies of a purified preparation of the enzyme and evolutionary significance to the three different complex (CPSase-ATCase) of Saccharomyces classes of ATCases which they observed (i.e., cerevisiae indicated Michaelis-Menten kinetics P. aeruginosa > E. coil > B. subtilis). We have for both carbamyl phosphate and aspartate; now devised another method, namely the separa- Lineweaver-Burk plots were linear even at very tion of each ATCase into its respective subunits low substrate concentration. The Km for carbamyl followed by the reconstitution of hybrid enzymes phosphate was 4 X 10-3M and for aspartate 2.8 X containing heterologous regulatory and catalytic 10-2 M, both measured at pH 7.4. Higher pH subunits from different organisms. If such a values (8.0 to 8.5) did not seem to significantly hybrid enzyme exhibits activity and feedback change the Km for aspartate but it had a proinhibition, this is indeed a stringent test of the nounced effect on the Km for carbamyl phosphate closeness in evolution of two such organisms. As [1.8 X 10-4 M (208)]. mentioned above, such a study has been perUTP is the most potent inhibitor of both CPSformed with the two enteric genera E. coli and ase and ATCase of yeast (210, 225). ATP did not Salmonella typhimurium (O'Donovan, Holoubek, activate the enzyme, in contrast to what has been and Gerhart, in preparation). So far it has not observed with E. coli ATCase (148). The inhibibeen possible for us to reconstitute a full native tion by UTP decreased the maximal velocity and enzyme for B. subtilis. Neither aggregation nor remained constant at all carbamyl phosphate the appearance of feedback inhibition has been and aspartate concentrations (208). This is also found when regulatory subunits from E. coli and in contrast to the observation with the E. coli S. typhimurium have been added to the B. enzyme (148). In all cases, Hill plots were unity, subtilis enzyme (O'Donovan, unpublished data). indicating' the lack of cooperativity involving In both Neurospora crassa and Saccharomyces either substrate or inhibitor; kinetics were cerevisiae, several common features emerge. Both always Michaelis-Menten. fungi contain a single genetic region, probably an It has been observed that feedback inhibition operon, which controls the synthesis of both CPS- increases as temperature of assay is decreased ase and ATCase. Since both enzymes are feed- (289). The same effect was observed when ATCase back inhibited by UTP, a third polypeptide chain from yeast was assayed at 30 and 3 C. Inhibition bearing the regulatory site may also be present. was greater at 3 C, and the temperature gradient A unique feature of this genetic region is that it effect was more pronounced at low temperature. produces at least two enzyme activities which are It was never possible to get 100% inhibition with always co-purified and which share a common UTP (208), a common phenomenon for pyrimiregulatory site for UTP, the feedback inhibitor. dine nucleotide inhibition of all ATCases studied. Moreover both fungi have two separate CPSases, During derepression, ATCase was found to as previously discussed. become progressively less sensitive to feedback In Neurospora, UTP acts as a feedback inhibi- inhibition by UTP (208) when uracil-requiring tor of ATCase, the inhibition being noncompeti- cells were grown in limiting uracil. Wild-type tive with both substrates; the inhibition affects cells, or mutant cells grown in the presence of only the maximum velocity. Michaelis-Menten uracil (repressed, see below), did not show a loss kinetics are observed when velocity-substrate of inhibition (209). Heating, or high dilution in

VOL. 34, 1970

PYRIMIDINE METABOLISM IN MICROORGANISMS

vitro caused the loss of inhibition by UTP as well. Thus, it appears that UTP plays three unique roles in the regulation of pyrimidine biosynthesis in yeasts. (i) It acts as the feedback inhibitor of ATCase activity; (ii) it stabilizes the regulatory site on the enzyme; and (iii) it represses ATCase synthesis, as will be shown below (209). Mutants affected in their regulatory site have been isolated (208) by their resistance to the analogue 5-fluorouracil. They show little or no feedback inhibition by UTP and excrete uracil into the medium. When crude extracts of the 5fluorouracil-resistant strain were tested for feedback inhibition by UTP at high and low temperatures, a rather unique result was observed. Whereas no inhibition was found at 30 C, significant inhibition was observed at 3 C. This is a useful test, for (i) it clearly demonstrates the value of inhibition data obtained from assays at low temperatures, and (ii) it demonstrates in this particular case the presence of the regulatory site of the enzyme in the 5-fluorouracil-resistant strain. One might have suspected its absence from the lack of inhibition at normal temperatures (208). Dihydroorotase and Dihydroorotate Dehydrogenase Properties. Dihydroorotase catalyzes the reversible cyclodehydration of carbamyl aspartate to dihydroorotate: L-Carbamyl aspartate Zn2+ L-Dihydroorotate + H20 (III) The dihydroorotate so formed is converted to orotate by dihydroorotate dehydrogenase: L-Dihydroorotate + NAD+ = Orotate + NADH + H+ (IV) Thus, in the two reactions a dehydration gives rise to the first cyclic compound in pyrimidine biosynthesis, followed by a dehydrogenation giving rise to the first ultraviolet absorbing

pyrimidine. Dihydroorotase has been studied in detail in Z. oroticum by Lieberman and Kornberg (242). The enzyme was first shown to be involved in the utilization of orotate as a carbon source by the anaerobe Z. oroticum and by the aerobe Corynebacterium (242-244, 403). Dihydroorotase has also been identified in E. coli (370, 403). Yates and Pardee (403) showed that the enzyme was present in Z. oroticum whether the cells were grown in orotate or glucose; in orotate, however, much higher activities were found. The enzyme was not inducible in E. coli but was present regardless of carbon source. Cell-free extracts of

293

E. coli converted aspartate plus carbamyl phosphate to orotate. Thus, the enzyme was shown to be involved in pyrimidine biosynthesis (403). In most organisms studied, dihydroorotase is a very unstable enzyme (225), so it has never been extensively purified. The enzyme is highly specific and is active only with the L-isomers of carbamyl aspartate and dihydroorotate (243, 330). The enzyme from Z. oroticwn requires Zn2+ for maximal activity and has a Km for carbamyl aspartate of 8 X 10-4 M at pH 5.5 (340). Dihydroorotate dehydrogenase, just like dihyroorotase, was originally isolated from the obligate anaerobe Z. oroticum by Lieberman and Kornberg (242) by enrichment on orotate media. The enzyme has also been shown to be present in E. coli (370-372, 403, 404) and in a Pseudomonas strain (213, 268, 372). It has been reported (371) that the dihydroorotate dehydrogenase of E. coli is associated with the membrane portion of lysed spheroplasts and does not interact with pyridine nucleotides like the dehydrogenase of organisms grown on orotate as a carbon source (142, 376). Although it appears that the catabolic dihydroorotate dehydrogenases are pyridine nucleotidelinked, this is certainly not true of the biosynthetic enzymes which appear to link to oxygen or ferricyanide (377). Since E. coli or S. typhimurium cannot use orotate as the sole source of carbon and energy (164), a difference between the catabolic dehydrogenase and the biosynthetic one is to be expected. Recently, the dilemma was overcome by the isolation of two functionally different dihydroorotate dehydrogenases in a Pseudomonas strain (372). The pseudomonad, capable of growth on a minimal salts medium with glucose, glycerol, aspartate, or orotate as carbon sources, produced a particle-bound dihydroorotate dehydrogenase when so cultivated. Thus, regardless of carbon source, a particle-bound enzyme was produced which was exactly like the E. coli biosynthetic enzyme. Oxygen or ferricyanide were electron acceptors; pyridine nucleotides were not. The addition of orotate to the minimal medium as the sole or supplemental carbon source caused the synthesis of a new soluble enzyme which linked to nicotinamide adenine dinucleotide phosphate (NADP). Thus, when cultivated on orotate the organism produced both the biosynthetic E. coli-like enzyme and a soluble degradative enzyme which was NADP linked. From these results, the following useful criterion can be established. In aerobic assays (oxygen as the final electron acceptor), the biosynthetic dihydroorotate dehydrogenase is assayed and the enzyme is particle-bound and will donate electrons to oxygen. Thus, a better name for this enzyme

294

O'DONOVAN AND NEUHARD

might be dihydroorotate oxidase. If pyridine nucleotides are used as the electron acceptors, the degradative enzyme is the one which is assayed. This is frequently a soluble enzyme. In accord with the above, Eakin and Mitchell (121) recently isolated a mitochondrial dihydroorotate oxidase from Neurospora which uses molecular oxygen as the final electron acceptor. In most biological systems, the reaction dihydroorotate to orotate is reversible, but by coupling the reduction of molecular oxygen to the dehydrogenation of dihydroorotate, this Neurospora mitochondrial enzyme carries out an irreversible reaction in favor of orotate formation. There is but one such enzyme in Neurospora and its sole function is pyrimidine biosynthesis (121). The catabolic dihydroorotate dehydrogenase has been crystallized from Z. oroticum (141, 142) and shown to be a flavoprotein containing 2 moles of flavine adenine dinucleotide, 2 moles of flavine mononucleotide, 4 g atoms of iron, and 4 moles of acid-labile sulfide per mole of enzyme. Rather uniquely, this purified metalloflavoprotein contained only 1 atom of iron per molecule of flavine, and both flavine moieties seem to be involved in the catalytic activity of the enzyme (269). It is likely that each of the flavines is responsible for the reaction with one of the pairs of substrates, nicotinamide adenine dinucleotide/ reduced nicotinamide adenine dinucleotide (NAD+/NADH + H+) and dihydroorotate/ orotate. Different primary reductive steps have been shown by stopped-flow kinetics, in which the enzyme was made to react with either NADH + H+ or dihydroorotate (269). The physiological acceptor is probably NAD. Activation by cysteine is required for maximal rates of reduction of orotate or oxidation of dihydroorotate with the enzyme. Cysteine is not required however for the reduction of the enzyme by NADH or the aerobic oxidation of NADH. The enzyme has a molecular weight of 120,000. It is highly specific for L-orotate, no activity being shown with 5-methyl cytosine, cytosine, uracil, or thymine. Aleman and Handler (12) studied the substrate specificity with a series of pyrimidine derivatives which they tested for their ability to substitute for orotate as electron acceptor. They showed that both the ring hydroxyls as well as the carboxyl substituents were required for ability to act as substrate. Both 5-fluoro and 5-bromo derivatives, and to a lesser extent 5-iodoorotate, were reduced. The Km and Vmax values were increased when 5-halogenated compounds were used. The enzyme is inhibited by 5-methyl orotate (Ki = 2 x 10-3 M), the inhibition being competitive with orotate (140). The enzyme is also inhibited by 5-amino orotate and 5-nitroorotate

BACTERIOL. REV.

(12). The enzyme from E. coli B has been shown to be inhibited by dihydro-5-azaorotate (353). Genetics. Several mutants exist in S. typhimurium (401) and in E. coli (36, 403) for each of the genes coding for dihydroorotase (pyrC) and dihydroorotate dehydrogenase (pyrD), respectively. These mutants have an absolute requirement for pyrimidines. As shown in Fig. 8, the two loci are located on the linkage map of E. coli and S. typhimurium at approximately 3 o'clock (368, 401). Both loci (pyrC and D) are located very close together (401). The two genes are unlinked to any of the other pyrimidine loci. Mutants have been found in yeasts which specifically lack dihydroorotase (ura-4) and dihydroorotate dehydrogenase (ura-1; reference 225). As in the enterobacteria, these genes are unlinked to each other and to any of the other genes (ura-2, ura-3) coding for enzymes involved in the de novo pathway. In studies by Caroline (77) with Neurospora, a similar picture emerges. The two loci for dihydroorotase (pyr-6) and dihydroorotate dehydrogenase (pyr-1) are unlinked to each other. There have been several reports that mutants blocked in an early gene (pyrA, pyrB) can have their pyrimidine requirement satisfied by growth on carbamyl aspartate, dihydroorotate, or orotate. S. typhimurium strains blocked in an earlier step can be supplied carbamyl aspartate, dihydroorotate, or orotate from the medium (401); a similar observation was made for Serratia marcescens (178) and for E. coli (36, 118, 403, 404). None of these compounds readily permeates the bacterial membrane although it is possible to show cross-feeding between appropriately blocked mutants, and orotate can supply the pyrimidine requirement in Salmonella typhimurium and, with more difficulty, in E. coli. In Pseudomonas aeruginosa (199), Saccharomyces cerevisiae (225), Neurospora (77), and in almost all the E. coli (excepting a K-12 strain, CYA 288) and most Salmonella typhimurium strains in our laboratory, it is not possible to get reproducible results from such feeding experiments. However, it has been possible by using deletion mutants in pyrA or pyrB to isolate strains of S. typhimurium which take up pyrimidine intermediates more readily than the wild strain. It will be of interest to learn the genetic loci of these permeases since a similar study has been carried out by Lacroute (225) in yeasts. Using a uracil-requiring mutant lacking aspartate transcarbamylase (ura-2), Lacroute isolated strains which were able to incorporate carbamyl aspartate, dihydroorotate, and orotate by selecting for cells able to grow on these compounds. He found two kinds of mutations, one allowing the incorporation of carbamyl

VOL. 34, 1970

PYRIMIDINE METABOLISM IN MICROORGANISMS

aspartate (= ureidosuccinate), ure-1, and the other allowing the incorporation of both dihydroorotate and orotate, designated, oro-1. Neither ure-1 nor oro-1 was linked to each other or to any of the genes which control the de novo enzymes of pyrimidine biosynthesis.

OMP Pyrophosphorylase Properties. Using bacterial mutants (194, 254), it was shown that orotate could not be transformed to other pyrimidines at the free pyrimidine level. Instead, nucleosides and nucleotides were produced directly from orotate. Following the identification of PRPP (219), Kornberg and co-workers showed that yeast contained an enzyme, OMP pyrophosphorylase, that catalyzed the reversible formation of OMP from orotate and PRPP:

295

is likely, however, that some of the OMP decarboxylase mutants (ura-3) might simultaneously lack OMP pyrophosphorylase and OMP decarboxylase (225), a phenomenon commonly observed in mutant human cells (224). Even though orotate does not easily permeate the membrane in many organisms and even though the control of pyrimidine biosynthesis is highly regulated, the excretion of orotate into the medium has frequently been reported. In humans this is due to the lack of OMP pyrophosphorylase, causing the disease orotic aciduria (359). There is an interesting report (258) that describes the accumulation of orotate in cultures of some wild strains of E. coil K-12. Wild-type Salmonella typhimurium or E. coil B does not usually excrete orotate. The property of orotate excretion explains the recent observation in our laboratory that E. coil K-12 strains are resistant to high exogenous fluoride concentrations on minimal agar plates. On the other hand, S. typhimurium strains are sensitive to high fluoride in the medium, but the addition of orotate affords growth in the presence of fluoride. Moreover, a fluorideresistant S. typhimurium mutant excretes orotate into the medium and will allow the growth of sensitive wild-type cells adjacent to the resistant colony on plates of fluoride media. The fluorideresistant mutant has an elevated intracellular pool of OMP and a depressed pool of UMP. The fluoride-resistant mutation has been shown to cotransduce with pyrF (R. J. Stockert and G. A. O'Donovan, Bacteriol. Proc., p. 60, 1970). Orotate is produced by E. coil in cultures if 6azauracil is administered (36). Orotate is also excreted into the medium by pyrinidine-requiring mutants of Neurospora (271), Aerobacter aerogenes (69), E. coli (403), and Serratia marinorubra (37) when starved for pyrimidines. Bacterial mutants which produce large quantities of orotate have been used in the fermentations industry. One such mutant strain has been isolated from Brevibacterium ammoniagenes which excretes large quantities of orotate into the medium (191). The mutant is deficient in the enzyme OMP pyro-

Orotate + PRPP OMP + PPi (V) The equilibrium constant of the reaction is 0.1. OMP pyrophosphorylase is highly specific, no activity being found with adenine, dihydroorotate, orotidine, or uracil (compare equation XV) when substituted for orotate (134). The Km for orotate is 2.0 X 10-1 M. A similar Km is observed when 5-fluoroorotate is substituted for orotate. The other two reactants, namely PRPP and PPi, have Km values of 2 X 10-i and 1.2 X 10-4 M, respectively. Even though 5-fluoroorotate is an active substrate, no activity is seen with 5-bromo, 5-amino, 5-nitro, 5-chloro, or 5-methyl derivatives of orotate. The enzyme is also specific for PRPP and PPi (134). The enzyme is inhibited by OMP and by certain analogues of orotate wherein the carboxyl group is replaced by a sulfonate, sulfonamide, or sulfone, giving 4-uracilsulfonate, 4-uracilsulfonamide, and 4-uracil methyl sulfone, respectively. These analogues are powerful growth inhibitors when offered to microorganisms requiring orotate (67, 186, 323). The analogue 5-azaorotate inhibits the enzyme, the inhibition being competitive with orotate (89). Genetics. The gene which codes for OMP pyrophosphorylase (pyrE) has been located on phosphorylase (356). the E. coli (368) and S. typhimurium (401) linkage maps between 9 and 10 o'clock. It is not OMP Decarboxylase linked to any of the other genes involved in Properties. The next step in the biosynthesis of pyrimidine biosynthesis including upp (Fig. 8). OMP pyrophosphorylase mutants, showing an UTP and CTP is the irreversible decarboxylation absolute requirement for pyrimidines (uracil or of OMP to UMP, as shown in equation VI. The cytosine), have been isolated in E. coli (36), S. enzyme responsible for this reaction is highly typhimuriwn (401), Serrata marcescens (178), P. specific and has been purified from yeasts (246). aeruginosa (199), and Neurospora (pyr-2; reference Orotidine-5'-phosphate ( 77). In yeasts the genetic region affecting OMP (VI) pyrophosphorylase has not been identified. It uridine-5'-phosphate + CO2 It

296

O'DONOVAN AND NEUHARD

BAcTERjoL. REV.

The two previous steps in the biosynthesis (equations IV and V) were unfavorable for biosynthesis but by coupling with this reaction (equation VI), biosynthesis readily proceeds (323). The enzyme from yeast is inhibited by UMP (K, = 1.5 X 10-4 M), CMP, AMP, and GMP (106). The mammalian enzyme purified 600-fold is inhibited by CMP, UMP, CDP, UDP, and CTP in the descending order shown. The inhibition is competitive, OMP and CMP competing for the same catalytic site (22). The pyrimidine analogue 6-azauridine is a powerful inhibitor of OMP decarboxylase in microorganisms in vivo (172, 354, 355). It exerts its effect only after being converted to 6-azaUMP, inhibits OMP decarboxylase, and causes the accumulation of OMP, which subsequently inhibits OMP pyrophosphorylase; the net result is the excretion of orotate into the medium (173, 301). In accord with this, a partially purified OMP decarboxylase from yeast was found to be competitively inhibited by 6-aza-UMP. The inhibition is reversible and specific for the 6-azamononucleotide. 6-Azauracil, 6-azauridine, and the corresponding di- and triphosphates had no effect (173, 301). Genetics. The gene which codes for OMP decarboxylase (pyrF) has been located on the E. coli (368) and S. typhimurium (401) at about 4 o'clock (Fig. 8). It is unlinked to any of the other genes involved in pyrimidine biosynthesis (368, 401). Mutants which lack this enzyme have been isolated in E. coli (36), S. typhimurium (401), Serratia marcescens (178), P. aeruginosa (199), N. crassa (pyr-4; references 77, 271), and Saccharomyces cerevisiae (ura-3; reference 225). Mutants which lack OMP decarboxylase frequently excrete orotate into the medium. The pyr-4 mutants of Neurospora have been shown to excrete orotidine (267) as well as orotate into the medium (314). ARll mutants lacking OMP decarboxylase have an absolute requirement for pyrimidine.

UTP (246). Subsequently, it was found (247) that a 30-fold purified enzyme from yeast could carry out the transphosphorylation between adenosine, uridine, and guanosine nucleotides: YMP + ZTP = YDP + ZDP (IX) This enzyme was distinct from muscle myokinase which acted on adenosine nucleotides only (100). ATP was tested as a phosphate donor with a number of nucleoside monophosphates. AMP, UMP, GMP, and dAMP were active acceptors; no activity was seen with IMP, dCMP, and dGMP. An enzyme nucleoside diphosphokinase (38) which was first independently discovered in 1953 by Berg and Joklik (38) and Krebs and Hems (223) has been obtained in crystalline form from yeast (319). The reaction is as follows: N1TP + N2DP = N1DP + N2TP where N is any purine or pyrimidine base. The enzyme catalyzes the phosphorylation of ADP by a wide variety of ribo- and deoxyribonucleoside triphosphates. With ATP or dATP as phosphoryl donors, the diphosphates of cytosine, guanosine, inosine, and uridine are converted to their respective triphosphates (319). Thus, the enzyme seems to be quite unspecific towards the base moiety. Mutants which lack UMP kinase (nmk*) and UDP kinase (ndk*) have not been identified.

UMP Kinase and UDP Kinase UMP is next converted to UDP and then to UTP by kinases which appear to have no specificity for the nucleoside moiety of the nucleotides; uridine, cytidine, adenosine, and guanosine can therefore all react. The two reactions are shown in equations VII and VIII. UMP + ATP UDP + ADP (VII) UDP + ATP UTP + ADP (VIII) Extracts of yeast, capable of transforming orotate to UMP, are also capable of converting UMP to

Properties. CTP synthetase isolated and purified 300-fold (253) catalyzes the reaction:

CmP Synthetase The conversion of UTP to CTP with the utilization of ammonia and ATP was first described by Lieberman who showed that the reaction was carried out by a soluble enzyme from E. coli B (240, 241). Glutamine was shown to substitute for ammonia, and a guanosine nucleotide was required for maximal activity in crude extracts (80). Long and Pardee (253) showed with a purified enzyme that glutamine was an alternative nitrogen donor if GTP was present.

MG'+

UTP + ATP + glutamine Mg2l (X) CTP +ADP + Pi + glutamate The enzyme is a dimer with a molecular weight of 105,000, composed of presumably identical subunits. That there are four identical subunits in the active enzyme has also been confirmed. In the presence of its substrates, ATP, UTP, and Mg2+, the dimer aggregates to the enzymatically

PYRIMIDINE METABOLISM IN MICROORGANISMS

VOL. 34, 1970

active tetramer of 210,000 molecular weight (252). Recent studies by Levitzki and Koshland (in preparation) have identified a glutamylenzyme intermediate and have shown that the SH group of cysteine is the group which is acetylated by the glutamyl moiety. From these studies, and from studies with 180 and 32p exchange, the mechanism of the reaction has been shown to be as indicated in Fig. 6. The regulation of CI? synthesis is rather unique. This is not surprising when one considers the position ofthe reactionat the extreme end of the de novo pathway. Furthermore, both UTP (the substrate) and CT? (the product) are required equally for the biosynthesis of ribonucleic acid (RNA) and deoxyribonucleic acid (DNA),anda controlexerted solely by either one would be of little physiological use to the cell. The most common controls in biosynthetic pathways are inhibition by an end product and a activation by a metabolite which accumulates in response to the end product becoming limiting. At high substrate (UTP) concentrations, CTP acts as a competitive inhibitor with UTP, thus preventing its own synthesis. At subsaturating concentrations of GTP and glutamine, CTP activates the reaction. Activation control is also exerted by UTP and ATP. When either UTP or ATP is present at a subsaturating concentration, the synthesis of CTP is increased by an excess of the other (253). The kinetics studied for the glutamine reaction gave Michaelis-Menten hyperbolae when velocitysubstrate plots were examined. The Km values for UTP, ATP, GTP, and glutamine, calculated from linear Lineweaver-Burk plots, were 0.15, 0.20, 0.08, and 0.22 mm, respectively. Under saturation conditions, CTP interacted directly with the UTP site, having a Ki of 0.11 mm. When ATP and UTP were not at saturating concentrations, the kinetics of the glutamine

297

reaction no longer exhibited Michaelis-Menten behavior. At nonsaturating ATP concentrations, the UTP saturation curve was sigmoidal; at nonsaturating UTP, the ATP saturation curve was sigmoidal. This indicates cooperative substrate binding or interacting substrate sites. GTP at nonsaturating concentration did not influence cooperative UTP binding as did ATP. CTP inhibited the glutamine reaction when ATP and UTP were not saturating. The cooperative binding of ATP and UTP was not affected by CTP, however. The ammonia reaction is quite similar to the glutamine reaction in requiring ATP and Mg2+; however, glutamine and GTP are not required. Velocity-substrate plots give approximate Michaelis-Menten kinetics for UTP and ATP when all but one substrate are near saturation. However, cooperative binding for UTP and ATP existed even at high UTP and ATP concentration. At these near-saturating levels, CTP was a competitive inhibitor for UTP, noncompetitive for ATP. As was found for the glutamine reaction at subsaturating concentrations of UTP and ATP, sigmoidal kinetics were observed. The most basic difference between the glutamine and ammonia reactions was observed in the presence of CTP with nonsaturating concentrations of the substrates UTP and ATP. With ammonia a substrate, CTP showed no inhibition but, in fact, became an activator. GTP could partially replace CTP in this activating role. At nearsaturating concentration of substrates, it was possible to get some CTP activation of the glutamine reaction, however. In recent years much discussion has revolved around allosteric sites, conformational changes, subunits, Hill plots, positive and negative cooperativity. Frequently, it has been possible to treat data in slightly different ways with apparently quite different results. We will avoid any OH such interpretations here but we will attempt to -larify some definitions in the hope that a general use can be made of them. Since the de novo -N/ AT pathway of pyrimidine biosynthesis contains two of N rlb.A~( the most studied regulatory enzymes (ATCase, CTP synthetase) we feel that a short discussionisworthghutomine eouyve (I) gktanMI efl* + NH3 while. We will not deal with theoretical aspects of conformational changes or basic theories of N2 allostery since these are the subject of a beautiful treatment by Gerhart (Gerhart, in press) on (M) NH33+ these properties, as applied to ATCase. o 4i As recently pointed out by Koshland (238), HO e") it is necessary to consider and recognize the various types of curves that one would expect 91htamate (iv) .npe nzfye from various types of cooperativity. Figure 7 Pio. 6. Reaction mechanism of CTP synthelase from shows a velocity-substrate plot (A), a doubleEscherichia coli (Levitzki and Koshland, in preparation). reciprocal plot (B), and a Hill plot (C) for posi0P03

+

ATP

(I)

HO

N

Ibm

+

-

0P0

NS

b

glutamyl-

+

H20

+

298

O'DONOVAN AND NEUHARD

tive cooperativity, negative cooperativity, and Michaelis-Menten kinetics, respectively. The curves are all theoretical and are taken from the calculations of Levitzki and Koshland (238). Table 4 summarizes the information available from such curves. The Hill plot is derived from the Hill equation (179) v = V-Sn/(K + Sn). By transforming and taking logarithms, we get log (v/V - v) = n log S log K. A plot of log (v/V v) versus log S will have a slope of n, called the Hill coefficient. Ideally, a straight-line Hill plot is observed (Fig. 7), in which case n is considered to approximate the minimal number of binding sites for a given substrate. As mentioned above, the isolated CTP synthetase is a dimer (molecular weight, 105,000) of two identical polypeptide chains. The enzyme exhibits positive homotropic (cooperative) effects for ATP and UTP (substrates) with Hill coefficients of 3.8 and 3.4, respectively, indicating that the enzyme shows strong site to site interactions (253). Since Hill coefficients usually indicate the minimal number of sites (362), two ATP and two UTP sites per polypeptide chain (per subunit) would be needed for four sites per molecular weight of 105,000. Since from negative cooperativity data (see below) four glutamine sites were also predicted, a temporary dilemma was imminent. The finding that the dimer aggregates to a tetramer of 210,000 resolved the issue, each identical sub-

-

12 10

6

II

"

60:C ~~~~~40-

A

5 -8

4~~~~~~~~~~~046

020 0202466

8

s

0

02 1

2

3

tS]

181

4

5

1

0205

1.0

s

FIG. 7. Velocity-substrate (A), Lineweaver-Burk (B), and Hill (C) plots. Data are theoretical and are taken from Levitzki and Koshland (238).

BACrERIOL. REV.

unit being expected to contain one site each for ATP, UTP, and glutamine (252). The enzyme exhibits negative cooperativity with glutamine (substrate) and GTP (effector). The initial velocity of CTP synthesis exhibited pronounced deviations from Michaelis-Menten kinetics as a function of glutamine. Since GTP is also required along with glutamine (252), different fixed concentrations of GTP were used to give a family of curves whose deviations became less and less as GTP was increased (238). A double reciprocal plot indicated these deviations more clearly (see Fig. 7 and Table 4; 238). As negative cooperatively seems to be real and perhaps even a common phenomenon (42, 109, 235, 238, 295) one is forced to ask why such a control. With respect to a feedback inhibitor, one can possibly imagine an effect similar to that which occurs in a chemostat: the reaction is modulated but never completely shut off. This would be required if the product of the reaction was obligately required for some other metabolic scheme. Negative cooperativity involving an activator (or substrate) is by far the more common type observed As in this case with CTP synthetase, negative cooperativity concerns an activator. The possible function of negative cooperativity can be seen for the effectors GTP and glutamine with this enzyme: GTP and glutamine are allowed to accumulate extensively without interfering, too much with the activity of the enzyme. Thus, negative cooperativity may be an effective means to dampen fluctuations in enzyme activity at high effector concentrations. Genetics. As might be expected, mutants which lack CTP synthetase have a growth requirement for cytidine. However, because of the metabolic instability of cytidine, due to a highly active cytidine deaminase in the enteric bacteria, it was necessary first to construct strains lacking this enzyme before CTP synthetase mutants could be isolated. Such strains are

TABLE 4. Summary of kinetic data available from indicated plotsa Hill Characteristic Characteristic

l

saturation ~Shape Shaplot (vofversus [So)

[|S[SI atat 0.0.1 V, Vb

Shape of Lineweaver-Burk plot

coeffcient (log [v/V-v] versus

log

Positive cooperativity Michaelis-Menten (independent, noninteracting sites) Negative cooperativity a

b

[S])

Sigmoidal Hyperbolic (smooth)

1

Hyperbolic (broken)

>81

Concave downwards

OMP

-

6

pyrA

pyrB

pyrC

pyrD

pyrE

pyrF

pyr-3

pyr-3 ura-2

pyr-6 ura4

pyr-) ura-1

pyr-2

pyr4 ura-3

ura-2 (cpu)

UMP

a Enzymes: 1, carbamyl phosphate synthetase; 2, aspartate transcarbamylase; 3, dihydroorotase; 4, dihydroorotate dehydrogenase; 5, OMP pyrophosphorylase; 6, OMP decarboxylase.

would be manifested quickly. Indeed, as was pointed out by Abd-El-Al and Ingraham (3), the addition of 1 mm ornithine and 0.1 mm UMP, probable physiological concentrations, will activate the enzyme at 20 C but will inhibit the enzyme at 37 C. Complete loss of activation by ornithine, increase in the feedback inhibition by UMP, and overproduction of arginine or pyrimidines would all have an effect on CPSase, by virtue of the critical ornithine-to-UMP ratio that is required. Therefore, it is not surprising that a variety of different phenotypes have been observed in mutants of pyrA (3). ATCase of E. coli and S. typhimurium is feedback inhibited by CTP, a pyrimidine nucleotide; it is activated by ATP, a purine nucleotide. The inhibition by CTP is competitive for both substrates (46, 148) and is never complete (145, 148). The evolutionary choice of CTP may have been easy if one considers the following three points. (i) CTP is more abundant in exponentially growing enterobacteria than other cytosine derivatives. Its concentration in vivo has been estimated at 0.2 mm (282). This will inhibit 75% ATCase activity at 2 mm aspartate, a concentration that is physiologically feasible (147). (ii) At the nucleotide level, there are no deaminases to convert CTP or cytosine nucleotides to their noninhibitory uracil derivatives. (iii) Since UMP is required for the inhibition of CPSase, the previous enzyme with a different function, it is unlikely that a uridine nucleotide would regulate the activity of ATCase as well. The development of a regulatory site can be considered a late event in the evolution of an enzyme (94); indeed, different alternatives have evolved in different microorganisms for the regulation of the arginine and pyrimidine biosynthetic pathways. In Pseudomonas, Saccharomyces, and Neurospora, UTP (not CTP) is the feedback inhibitor of ATCase, so that one would expect the control of CPSase in these organisms to differ from that found for the enteric bacteria. In Saccharomyces

and Neurospora, ATCase and CPSase occur in an enzyme complex with a single regulatory site for UTP common for both. In both cases, a second pathway-specific CPSase has evolved for arginine biosynthesis. No information is available at present for the CPSase of Pseudomonas, but the possibility of two isoenzymes has not been ruled out. B. subtilis and Streptococcus faecalis have been shown to possess ATCases which are not feedback controlled (45, 285). The release from feedback inhibition of ATCase would present the cell with the difcult problem of supplying two uncontrolled pathways with a common intermediate, carbamyl phosphate. A possible way out of this dilemma would be to evolve two CPSases as in Neurospora in which the carbamyl phosphate produced by each enzyme is channeled to arginine and pyrimidines, respectively. Since B. subtilis probably contains only one CPSase (201) and no information is currently available for S. faecalis, further studies on the regulation of arginine and pyrimidine biosynthesis in these organisms are awaited with great interest. As can be seen from Fig. 9, CTP plays a dual rule in its own regulation. The position of CTP synthetase would suggest its unique regulation. Both UTP (the substrate) and CTP (the product) are required for nucleic acids and other cell components. Both UTP and CTP must therefore play a role in their' respective controls. CTP will competitively inhibit its own synthesis at high substrate concentrations; it can also act as an activator, ensuring its own synthesis under appropriate conditions. Thus, we can see the vital importance of CTP concentration for the control of enzyme activity in pyrimidine biosynthesis. Control of Enzyme Synthesis Initially, four of the six genes involved in pyrimidine biosynthesis were thought to be in an operon; the other two genes were known to be separated from each other and from the

of four. For this reason, the pyrimidine pathway of E. coli was regarded as an intermediate between an operon such as the histidine operon of S. typhimurium in which the nine enzymes controlling histidine biosynthesis were shown to be coordinately repressed (14, 15), and a single regulatory system controlling the expression of seven unlinked genes in the arginine pathway of E. coli (154, 380). Since the first two enzymes in pyrimidine biosynthesis, CPSase (pyrA) and ATCase (pyrB), were not linked and since it was felt that the other four genes might be adjacent, an ideal system was envisaged to check for (i) noncoordinate control of pyrA and pyrB and (ii) possible coordinate control of the final four enzymes coded for by genes pyrC,-D,-E, and-F, respectively, in the same pathway (36). Later it was shown that pyrE, which codes for OMP pyrophosphorylase mapped very far from the cluster (see Fig. 8; 369); since that time, the exact loci have been determined for all six genes of the pathway in E. coli and S. typhimurium. As can be seen from Fig. 8, no linkage occurs. Nevertheless, although all six genes are scattered, the original observation that the final four enzymes in the pathway, coded for by genes pyrC, -D, -E, and -F, are coordinately repressed and derepressed still holds true; the other two, the first and second enzymes in the pathway, are controlled but no group

coordinacy

301

PYRIMIDINE METABOLISM IN MICROORGANISMS

VOL. 34, 1970

is

seen.

PyrA. Not only is the activity of CPSase controlled by feedback inhibition but its synthesis in E. coli is repressed by both end products, arginine and a pyrimidine (306). The repression by both end products is additive, and either compound is equally effective (307). A similar result was found for S. typhimurium. The addition of arginine to the growth medium repressed CPSase; starvation of a uracil-requiring strain (DP18) gave sevenfold derepression (Table 6, reference 3). It was not possible to distinguish between a cytosineor uracil-containing compound, or both, as being responsible for the repression of CPSase under the conditions of growth employed (307). Indeed, in wild strains of S. typhimurium and E. coli there is no way to determine the repressor compound. Cytosine and cytidine compounds are rapidly transformed by specific deaminases to uracil and uridine, respectively, and at the triphosphate level UTP is converted to CTP. Abd-El-Al and Ingraham (3), using strains of S. typhimurium in which these known conversions of cytosine to uracil were blocked by mutations, investigated the effect of cytidine and uracil on the synthesis of CPSase. Added uracil did not repress strain DL38, a prototroph, lacking both deaminases, but added cytidine did repress CPSase (Table 6). This suggests that a cytosine compound is the active pyrimidine in

TABLE 6. Repression and derepression of CPSase and ATCase by arginine and pyrimidines in specially constructed mutants of Salmonella typhimuriuma Strain

DL38

Mutant genotype

cod cdd

DP18 DP45

pyrF

DP55

cdd udp pyrA, G

cod cdd pyrG

Reactions blocked

Cytidine uridine Cytosine * uracil

OMP -s UMP Cytidine uridine Cytosine - uracil UTP -- CTP NH3 + CO2 + ATP carbamyl P

Uridine + Pi uracil + ribose1-P Cytidine -- uridine UTP -. CI7P UTP --l CP a

Requirement for growth

None

Uracil

Cytidine Cytidine

to minimal Additionsmedia

None Arginine Uracil

Arginine + uracil Cytidine Arginine + uracil + cytidine Uracil (limiting) Cytidine (limiting)

Relative specific Derepresacivt CPSase

sionaof

100 65 104 56 39 22

705 340

5-fold

Cytidine (limiting) + uracil (un-

6-fold

Uracil (limiting) + cytidine (un-

17-fold

limited)

Uracil Arginine

|limited)

Data are from Abd-El-Al and Ingraham (3) and from Neuhard and Ingraham (283).

302

O'DONOVAN AND NEUHARD

CPSase repression. Now, by blocking the reaction UTP to CTP (creating a cytidine requirement) in a strain with neither deaminase (DP 45), starvation for cytidine derepresses CPSase about threefold (Table 6, reference 3). Furthermore, deprivation of this strain for cytidine diminishes the CTP pool to undetectable levels, whereas the UTP pool is increased 15-fold (282). Even now, with a huge UTP pool no repression is observed, in fact only derepression, strongly suggesting that only a cytosine compound is involved. Thus, the synthesis of CPSase in S. typhimurium is subject to repression by arginine and a cytosine compound (see Fig. 9). The amount of derepression of CPSase obtained is low (Table 6), but this is not surprising due to its tight control by feedback inhibition. One further point seems pertinent. Since both arginine and a cytosine compound are required for repression, any model suggested for the mechanism of repression of CPSase has to account for the existence of two corepressors. PyrB. Not only is the activity of ATCase feedback inhibited by CTP but also its synthesis is repressed by both cytidine and uridine nucleotides in S. typhimurium (283). In contrast to CPSase in which derepression was not very high, ATCase can be derepressed enormously by any means by which pyrimidines are made limiting (36). ATCase activity can be increased 200-fold if pyrimidine-requiring mutants are starved for pyrimidines (36). In a specially constructed partial diploid for ATCase, greater than 500-fold derepression was observed (146). For precisely the same reason as for pyrA, it is not possible to determine whether a uracil- or a cytosine-containing compound acts as the corepressor of the enzymes of pyrimidine biosynthesis. However, the availability of cytidine-requiring mutants in S. typhimurium, blocked between UTP and CTP in strains which do not convert cytosine compounds to their uracil derivatives (see pyrA of this section and Table 6), provided a means of distinguishing between cytosine or uracil compounds as the corepressors. By starving a cytidine-requiring strain for cytidine, a sixfold derepression of ATCase was found. Similarly, if a uracil-requiring cytidine-requiring double mutant is starved for uracil in presence of cytidine, a 17-fold derepression of ATCase is observed (283). Thus, starvation for either CTP or UTP (by limiting cytidine or uracil) derepresses ATCase, but not to any great extent since the absence of both corepressors is needed for full expression of derepression, and the absence of either one, in

BAcrER10L. REV.

a pyrG mutant, results in extensive accumulation of the other (282). Likewise, for maximal repression both cytidine and uridine nucleotides are required, indicating that also for pyrB two corepressors exist. Thus pyrA and pyrB cannot have the same regulator because pyrA responds to arginine and a cytosine compound, whereas pyrB responds to both cytosine and uracil

compounds. PyrC, pyrD, pyrE, pyrF, and pyrG. Though the locus for each of these four genes has been determined and found to be unlinked in E. coli and S. typhimurium, nevertheless the four enzymes which convert carbamyl aspartate to UMP (pyrC, -D, -E, and -F) do appear to exhibit coordinate repression and derepression, suggesting their control by a common repressor. The addition of uracil to the growth medium will repress the synthesis of each of these enzymes in a coordinate way (36). No coordinancy has been found for ATCase with the four final enzymes. The ability of ATCase to show rapid and enormous variation of activity under conditions of pyrimidine limitation (derepression) could confer a great advantage on the organism. Shifts from a pyrimidine-rich to a nonpyrimidine medium cause a rapid increase in ATCase activity. This will result in the synthesis of more carbamyl aspartate, which in turn may stimulate (induce?) the pathway to produce more UMP. The genes are unlinked, the enzymes are coordinately regulated, and the pathway is quite possibly induced. Frankly, we feel that this type of control is suggestive of induction identical to that observed by Lacroute in yeasts (225). Experiments to verify or exclude the induction hypothesis in S. typhimurium are now in progress in our laboratory. Recently, Dennis and Herman (118) studied the effect of pyrimidine limitation on the derepression of OMP pyrophosphorylase (pyrE) and OMP decarboxylase (pyrF). By growing a pyrB mutant on orotate, they were able to create a limitation of pyrimidines; they found under these conditions of derepression that the regulation of pyrE and pyrF was not strictly coordinate (118). Control in Different Microorganisms The regulation of pyrimidine biosynthesis has been studied in detail for the four final enzymes in Serratia marcescens (pyrC, -D, -E, and -F; reference 178). Starvation of pyrimidine-requiring mutants of this organism caused derepression of the last four enzymes in the pyrimidine pathway; growth in uracil similarly repressed the synthesis of the last four enzymes

VOL. 34, 1970

PYRIMIDINE METABOLISM IN MICROORGANISMS

in the pathway. Both the repression and derepression appeared to be coordinate (178). However, no genetic studies are possible in this organism. There appears to be very little regulation except feedback inhibition in operation in P. aeruginosa. No linkage was found among 27 mutants characterized for four of the five loci studied (199). No mutant was found for dihydroorotase, an extremely unstable enzyme both in vivo and in vitro. No repression or derepression was shown to occur despite conventional attempts to find it. Thus, pyrimidine biosynthesis in P. aeruginosa is constitutive. By growing P. aeruginosa on 6-azauracil, an inhibitor of OMP decarboxylase (174), it was possible to show a slight increase in dihydroorotate dehydrogenase and OMP pyrophosphorylase activities. No increase was seen in ATCase or dihydroorotase activities. We are at a loss to offer a good explanation for the apparent constitutive synthesis of pyrimidines in P. aeruginosa. However, a careful study of arginine biosynthesis in conjunction with CPSase might offer some interesting answers. The importance of pyrimidine nucleotides for ribonucleic acid (RNA), deoxyribonucleic acid (DNA), and other cell materials would seem to elicit a stringent control on biosynthesis, and we feel that the primary site of regulation in P. aeruginosa may be CPSase. It might be noted that no pyrA mutants have been found which require both arginine and uracil. A preferential draw of carbamyl phosphate for arginine biosynthesis might cause the evolution of constitutive synthesis of pyrimidines. There are several unusual features about the regulation of pyrimidine biosynthesis in L. leichmannii. In a study by Hutson and Downing (195) with an arginine-requiring strain of L. leichmannii, it was shown that the first step unique to pyrimidine biosynthesis was the deimination of arginine. The subsequent enzymes (only two tested, ATCase and dihydroorotase) in the pathway could be repressed fivefold by uracil, but the first enzyme, arginine deiminase, was derepressed twofold. Since arginine is converted first to citrulline and then to carbamyl phosphate, the derepression of the deiminase might be due to the limitation of carbamyl phosphate by an end product of the pyrimidine pathway. Since Lactobacillus can freely take up all of the intermediates of pyrimidine biosynthesis, including UMP, it might be a useful organism for studying induction of the pyrimidine pathway. As mentioned above, UTP simultaneously feedback inhibits the activities of CPSase and

303

ATCase in yeast (225, 226). Since there are two pathway-specific CPSases (226), naturally only the activity of the pyrimidine-specific enzyme is controlled by UTP. The regulation of pyrimidine biosynthesis has been beautifully worked out by Lacroute (225). The activity of the first enzyme ATCase is not only feedback inhibited by an end product, UTP, but its synthesis is also repressed by UTP. This is not unusual per se, but what makes it unusual is that in yeasts ATCase alone is repressed; the other enzymes in the pathway are sequentially induced in a manner first described by Stanier (363) for the catabolism of mandelate. By selecting for mutants in ATCase which could utilize carbamyl aspartate (ure-1), Lacroute showed that dihydroorotase of these mutants was induced by the exogenously supplied carbamyl aspartate. Similarly, a mutant lacking dihydroorotase (ura-4) when supplemented with exogenous dihydroorotate was shown to induce dihydroorotate dehydrogenase. In a 5-fluorouracilresistant strain, insensitive to feedback inhibition by UTP, both carbamyl aspartate and dihydroorotate were overproduced and, as expected, the levels of the enzyme dihydroorotase and dihydroorotate dehydrogenase were increased (227). It would appear, therefore, that the enzymes beyond ATCase are induced by dihydroorotate, but whether it is a truly sequential induction has not been determined, nor has the specificity of the induction yet been carefully worked out. A mutant blocked in OMP decarboxylase (ura-3) would be expected to overproduce the substrates for the enzymes dihydroorotase and dihydroorotate dehydrogenase, and in this strain one would expect these enzymes to be induced. This is exactly what was observed. By introducing an ATCase mutant (ura-2) into this strain lacking OMP decarboxylase, it was found that dihydroorotate dehydrogenase was low (like ura-2) rather than high (like ura-3), which it would be if derepression by limiting end product was in operation. A uracil prototroph which can take up carbamyl aspartate (ure-1) was grown in minimal medium, carbamyl aspartate, uracil, or uracil plus carbamyl aspartate. In minimal medium, wild-type enzyme levels were found but on carbamyl aspartate all the enzymes beyond this were induced and ATCase was repressed. Moreover, the simultaneous addition of uracil and carbamyl aspartate did not prevent the induction by carbamyl aspartate. The repression is probably by UTP or some other end product (225). Thus, there can be little doubt that the regulation found by Lacroute is a novel but probably

304

O'DONOVAN AND NEUHARD

not unique system. In the leucine biosynthetic pathway of Neurospora, an almost identical pattern emerges (163) in which the first enzyme is repressed but subsequent enzymes are induced by the product of the first enzyme, aisopropyl-malate. In human cells, the later enzymes involved in pyrimidine biosynthesis are induced by intermediates of the pathway (308, 309). Indeed, as pointed out by Lacroute, this may be a common scheme for higher organisms in which genes are scattered throughout the genome (225). There is no evidence available to suggest that the same is not true for E. coli. The first study of the regulation of pyrimidine biosynthesis in Neurospora was carried out by Donachie (119). The regulation of enzyme activity is by UTP, the feedback inhibitor, simultaneously on ATCase and CPSase. Uridine starvation of pyrimidine-requiring mutants in Neurospora invariably results in derepression of ATCase and dihydroorotate dehydrogenase and, occasionally, increased levels of dihydroorotase were also found. Pyrimidine-requiring mutants in Neurospora do not grow well on uracil or cytosine. However, some differences are noticed if the starved mutant bears a lesion in an early or late gene. In an early gene, very little increase in dihydroorotate dehydrogenase activity is observed; if the strain is mutated in a late gene, pyr-2 for example, excessive derepression is seen (78). The clear-cut case of induction shown by Lacroute (225) in yeasts is not quite as evident in Neurospora. However, the increase in activity of dihydroorotate dehydrogenase in uridinestarved mutants lacking OMP pyrophosphorylase (pyr-2) indicates that the dehydrogenase is induced by a precursor. That no such increase was observed in a pyr-3 (ATCase)-pyr-2 double mutant indicates that induction probably does occur.

BACTrERIOL. REV.

ATCase, induction of dihydroorotase and dihydroorotate dehydrogenase, and (iii) the probable induction of the other enzymes. If induction in Neurospora is true as we believe, this brings to three the number of biosynthetic induction schemes known. They include pyrimidine biosynthesis (i) in yeasts (225), (ii) in Neurospora (78), and (iii) the leucine biosynthetic pathway in Neurospora (163).

SUMMARY OF DE NOVO BIOSYNTHESIS From the previous section the following key points need to be stressed. (i) The pyrimidine pathway is common for all organisms. (ii) The genes which code for the enzymes involved in pyrimidine biosynthesis appear to be unlinked in all organisms. (iii) The enzymes in the pathway are all derepressable except CTP synthetase (pyrG). (iv) In addition to arginine (in the case of CPSase) there are at least two corepressors in E. coil and S. typhimurium, a uracil-containing and a cytosine-containing compound (Table 6). It has not been definitely established that induction does not occur in E. coli. In yeasts induction has been shown to occur. Even though a common pathway is found in all organisms, great differences have been observed for the regulation of the pathway as a whole, and especially for ATCase. The phenomenon of a universal pathway with one easily modifiable link, namely ATCase, together with the fact that hybrid enzymes can be made in vitro, may make ATCase an important tool for evolutionary studies. As pointed out earlier, it is certainly worthwhile to try to make a feedback-controlled enzyme by reconstituting the catalytic subunit from Bacillus or Streptococcus with other regulatory subunits. A summary of the regulation of the de novo pathway is shown in Table 7.

One of the problems which consistently arose in the Neurospora study (77, 78) concerned the BIOSYNTHESIS OF DEOXYRIBONUCLEOunstable dihydroorotase. Mutants which accuTIDES mulate carbamyl phosphate always showed Ribonucleoside Diphosphate Reduction in E. coli higher levels of dihydroorotase than wild strains. Ten years ago our knowledge of deoxyriboThis indicated the inducibility of dihydroorotase, nucleotide biosynthesis was based entirely on in as did the occasional elevation in ATCase(pyr-3) mutants. This latter increase in dihydro- vivo experiments in which whole cells were orotase activity, however, was not sufficiently offered radioactive ribonucleosides and the consistent to allow a clear-cut decision about its label appearing in the individual components of RNA and DNA was determined. These experiinduction. Only the first three enzymes were examined ments established that E. coli contains enzymes in detail in Neurospora (77, 78). However, it which are able to reduce ribonucleosides or seems likely that the regulatory pattern found ribonucleotides to the corresponding deoxyin yeast by Lacroute (225) is also in operation ribosyl compounds (24, 338, 381). During the in Neurospora: (i) feedback inhibition of ATC- last 10 years, the enzymatic mechanisms underase by UTP; (ii) repression or derepression of lying these reductions in E. coil have been

305J

PYRIMIDINE METABOLISM IN MICROORGANISMS

VOL. 34, 1970

TABLE 7. Summary of the genotypes and regulation of pyrimidine enzymes in different organisms

tAnsparbatyae trAns|arbamylase

Organism

Organism

Escherichia coli Salmonella typhimurium Serratia marcescens Pseudomonas

Dihydroorotase

Dihydroorotate

OMPpyoOM

dehydrogenase

phosphorylase

oM

e

pyrB; F, R, D pyrB; F, R, D

pyrC; R, D pyrC; R, D

pyrD; R, D pyrD; R, D

pyrE; R, D pyrE; R, D

pyrF; R, D pyrF; R, D

pyrB; F, R, D pyrB; F

pyrC; R, D

pyrD; R, D pyrD; I

pyrE; R, D pyrE; I

pyrF;

ura-2; F, R, D

ura-4; I

pyr-3; F, R, D R

pyr-6; I?, D? pyr-l; I, D

pyrF; R, D

aeruginosa

Saccharomyces

ura-1; I

ura-3; I

I

cerevisiae

Neurospora crassa Lactobacillus leichmannii Mammals a

pyr-2;

pyr-4;

R

D

I

I

Key: F, feedback inhibition; R, repression; D, derepression; I, induction.

1M5 ®2---

--o

n

INADP+ S 0

db*eW"b

O

a

FiG. 10. Reduction of ribonucleoside diphosphates deoxyribonucleoside diphosphates in Escherichia coil. "B" denotes uracil, cytosine, adenine, or guanine. to

clarified by Reichard and his group in Stockholm. Several recent reviews of this subject have appeared and should be consulted for a detailed account of these studies (232, 324, 325, 326). The overall reaction is shown in Fig. 10. The four naturally occurring ribonucleoside diphosphates are reduced to the corresponding 2'-deoxyribonucleoside diphosphates by the reduced form of thioredoxin. This reaction is catalyzed by one enzyme, ribonucleoside diphosphate reductase. The regeneration of the reduced form of thioredoxin by NADPH is catalyzed by the flavoprotein, thioredoxin reductase. Thiredoxin. As shown in Fig. 10, the flow of reducing equivalents from NADPH to the ribonucleoside diphosphates is mediated by a lowmolecular-weight, heat-stable protein, thioredoxin, which contains a single disulfide bridge (molecular weight, 11,657). The reduced form of the protein, thioredoxin-(SH)2, is the immediate hydrogen donor for ribonucleotide reduction (234). The complete primary structure of thioredoxin has been determined (187-190). Figure 11 shows the amino acid sequence around

S

FiG. 11. Amino acid sequence of the active center of thioredoxin from Escherichia coli B.

the active center. Only two amino acids separate the two half-cystines. The total amino acid sequence of the molecule indicates that the Nterminal half, including the active center, is acidic and hydrophilic in character, whereas the other half has a basic and hydrophobic character (188). Fluorescence emission spectra of oxidized and reduced thioredoxin, after irradiation with ultraviolet light at 280 nm, showed that on reduction the quantum yield of fluorescence increased threefold, whereas the wavelength of the emission maximum was unchanged (364). This indicates that reduction of thioredoxin is accompanied by conformational change involving tryptophan residues. Since other physicochemical properties such as fluorescence polarization, optical rotary dispersion, circular dichromism, and sedimentation behavior did not change on reduction of the protein, it was concluded that the conformational changes occurring when thioredoxin is reduced are localized in nature (364). In keeping with this, the primary structure of thioredoxin showed that the only two tryptophan residues of the entire molecule are located very close to the disulfide bridge (Fig. 11). It was also pointed out that these conformational changes might be of importance in determining the affinity of oxidized and reduced thioredoxin for thioredoxin reductase and ribonucleoside diphosphate reductase, respectively (Fig. 10). Thioedoxin reductse. Thioredoxin reductase

306

O'DONOVAN AND NEUHARD

of E. coli has been obtained as a homogeneous protein, containing two molecules of FAD per mole of enzyme and no bound metals (275, 373). As shown in Fig. 10, the enzyme catalyzes the reduction of the disulfide bridge of thioredoxin with NADPH as electron donor. The enzyme has a high specificity towards both hydrogen donor and acceptor (275). The hydrogen in #position of the reduced nicotinamide ring of NADPH is transferred during the reaction (233). The enzyme consists of two identical or very similar polypeptide chains held together by noncovalent bonds. Each polypeptide contains, besides one molecule of FAD, one intrachain disulfide and two free sulfhydryl groups. The latter are buried in the quaternary structure of the native protein (374). Titration of the enzyme under anaerobic conditions with the native substrate, NADPH, showed that two moles of NADPH were required for full reduction of 1 mole of enzyme-bound FAD, indicating that secondary redox-groups exist in the protein. These groups were identified as the two disulfides in the enzyme (374, 407). The data also indicate that these -S-S- groups are involved directly in the catalytic process. Thioredoxin reductase resembles two other flavoproteins in many of its physicochemical properties, namely, lipoyl dehydrogenase (264) and glutathione reductase (265). These enzymes contain a reactive disulfide group as part of the active center. Anaerobic addition of reduced pyridine nucleotides to lipoyl dehydrogenase and glutathione reductase results in the formation of stable 2 electron reduction intermediates, in which the flavine and the disulfide have each accepted one electron to give a flavine semiquinone, a sulfur-free radical, and a sulfhydryl. Such an intermediate is not formed under similar conditions with thioredoxin reductase (374, 407), indicating that the catalytic mechanism of thioredoxin reductase differs significantly from that of lipoyl dehydrogenase and glutathione reductase. Ribonucleoside diphosphate reductase. Ribonucleoside diphosphate reductase catalyzes the reduction of all four naturally occurring ribonucleoside diphosphates (ADP, GDP, UDP, and CDP) to the corresponding 2'-deoxyribonucleotides (Fig. 10). The reaction has an absolute requirement for Mg2+. The specificity towards the nucleotide substrate is determined by the presence of certain nucleoside triphosphates (see below). The reductant in vivo is thioredoxin-(SH)2. In vitro, certain other dithiols such as reduced lipoic acid or dithiothreitol will substitute, but much higher concentrations of these artificial reductants are required (232).

BACTrERIOL. REV.

To gain insight into the reaction mechanism, dCDP was synthesized enzymatically from CDP in presence of 3H20 or D20 with either reduced lipoic acid or thioredoxin-(SH)2 as reductants. (The hydrogen of the thiol groups exchanges rapidly with the protons of water.) The distribution of hydrogen isotopes in the deoxycytidine moiety of dCDP was determined chemically in the experiments with tritium and by nuclear magnetic resonance spectroscopy in experiments with deuterium. The results showed that only one hydrogen is transferred to CDP during the reaction and that this hydrogen is found exclusively in the 2'-position of the deoxyribose in trans-position to the base (120, 229). Thus, the reduction seems to occur with retention of configuration around the 2'-carbon, suggesting the intermediary formation of a carbonium ion at C-2' which in turn is reduced stereospecifically by a hydride ion (324). Recently, Brown et al. (71) showed that ribonucleoside diphosphate reductase contains bound inorganic iron, necessary for activity of the enzyme. As pointed out (326), this nonheme iron might be involved in the generation of the postulated hydride ion from reduced thioredoxin. Several nucleoside triphosphates are able to modify the activity or the specificity, or both, of the enzyme in a very complex manner. Both the Km values and the Vma. values for the different substrates may be affected (230, 231). Thus, dATP was found to be a general inhibitor at concentrations above 10- M while stimulating pyrimidine nucleotide reduction at 106 M. The dT`TP (101 M) is a general activator of the reduction of all four substrates, whereas dGTP (101 M) specifically stimulated purine nucleotide reduction. ATP (10-4 M) is a specific activator when pyrimidine nucleotides are substrates and, moreover, is able to overcome the inhibition of dATP. In the presence of two effectors the pattern becomes even more complicated, since the net effect is dependent not only on the effectors used but also on the relative concentrations employed. Thus, in the presence of stimulatory concentrations of dATP. dTTP activates the reaction further, whereas in the presence of inhibitory concentrations of dATP, the simultaneous addition of dTTP potentiates this inhibition (72). In Table 8, some of these findings are summarized. During purification of ribonucleoside diphosphate reductase from E. coli B, the protein dissociates into two nonidentical subunits, Bi and B2 (70). Both have been obtained in pure form and when tested separately are without catalytic activity. When mixed in presence of Mg2+, full enzymatic activity is restored (70). Equilibrium centrifugations of Bi and B2 in the presence of

VOL. 34, 1970

PYRIMIDINE METABOLISM IN MICROORGANISMS

307

TABLE 8. Effect of nucleoside triphosphates on the activity of ribonucleoside diphosphate reductase of Escherichia coliO Nucleotide (M)

Catalytic activity

ATP (2 X 10-3) dATP (10-6) dATP (10-4) dGTP (10-5) dTTP (10-5) dATP (10-4) + ATP (2 X 10-3) dATP (10-) + dTTP (10-4) dATP (10-6) + dGTP (10-4) ATP (10-) + dTTP (5 X 10-')

Stimulation Stimulation Inhibition

Stimulation Stimulation Stimulation Inhibition

Inhibition Inhibition

Specificity for base

Pyrimidines Pyrimidines

Purines + pyrimidines

Purines Purines + pyrimidines

Pyrimidines Purines + pyrimidines

Purines + pyrimidines

(Purines) + pyrimidines

v Data are from Brown and Reichard (73). Concentrations given are those at which the effect is fully developed.

6 M guanidine-hydrochloride and 0.1 M mercaptoethanol indicated that each of the proteins could be dissociated into two subunits (326). The B2 subunit has a very specific absorption spectrum showing a sharp peak at 410 nm and a broader peak around 360 nm. This is due to the presence of two molecules of firmly bound inorganic iron per molecule of B2. The iron can be removed reversibly from the protein by dialysis against 8-hydroxy-quinoline or by acid ammonium sulfate, resulting in loss of enzyme activity and disappearance of the two characteristic absorption peaks. Readdition of Fe2+ to the apoprotein fully restored enzyme activity and the original absorption spectrum of the protein (71). Reactivation of the B2 apoprotein in the presence of 59Fe2+ turned out to be a convenient way of labeling the protein for sedimentation studies (73). Hydroxyurea inhibits ribonucleoside diphosphate reductase of E. coil both in vitro (125, 221) and in vivo (281). The inhibition is specific for the B2 subunit, and addition of hydroxyurea to the B2 protein results in the disappearance of the characteristic absorption spectrum of the protein without removing bound iron from the molecule (71, 221). The role of iron for the catalytic process and its mode of binding to the protein are still unknown, but may turn out to fulfill a role in the E. coil enzyme similar to that of coenzyme B12 in the corresponding enzyme system of L. keichmannii. It may be mentioned at this point that the L. leichmannii reductase is not inhibited by hydroxyurea (125). Sucrose gradient centrifugations indicated that pure protein Bi (7.8S) and pure protein B2 (5.5S) when mixed in presence of Mg2+ gave rise to a catalytically active complex with a sedimentation coefficient of 9.7S. The complex is composed of Bi and B2 in a molar ratio of 1:1. Addition of positive effectors (see Table 8)

did not change the sedimentation value of the complex. However, addition of negative effectors (dATP) or any combination of effectors that resulted in inhibition of enzyme activity (ATP + dTlTP) invariably gave rise to a heavier complex with a sedimentation coefficient of 15.5S. The heavy complex contains equimolar amounts of B1 and B2 and was believed to be an inactive dimer of the catalytically active 9.7S species (72). Using a rapid dialysis technique (101), a detailed study of the binding of nucleotide effectors to ribonucleoside diphosphate reductase was made (73). Since the B2 protein does not bind any effector, whereas the B1 protein does, most of the data were obtained with pure samples of the B1 species. Each molecule of B1 contains four binding sites for effectors, two with a high affinity for dATP (h-sites) and two with a 10-fold lower affinity for dATP (i-sites). The h-sites are able to bind all four effectors (ATP, dATP, dGTP, and dTTP), whereas the I-sites are only able to bind dATP or ATP. By comparing binding data with enzyme activity measurements (Table 8), it appeared that binding of effectors to the h-sites influences the substrate specificity of the enzyme, whereas the 1sites seem to be involved in regulation of the overall activity of ribonucleoside diphosphate reductase (73). Although ribonucleoside diphosphate reductase shows hyperbolic saturation kinetics both in the presence and in the absence of effectors (230, 231), the fact that binding of effector to sites topographically different from the substrate site alters the specificity of the enzyme clearly classifies it as an allosteric enzyme. In Fig. 12, a schematic representation of the allosteric nature of protein B1 is shown (73). Control of deoxyribonucleotide synthesis is brought about not only by allosteric regulation of

O'DONOVAN AND NEUHARD

308 ATP Pyrimidine -Specif

J

dATP dATP
dTMP hydrofolate + dihydro- (XI) midylate synthetase activity. Due to the fact that thymine-requiring mutants undergo thyfolate mineless death when incubated in the absence During the transfer, the Cl-unit is reduced to of thymine (96), penicillin counter-selection the methyl group of dTMP, dihydrofolate being cannot be used for the isolation of thymine the second reaction product (54, 302, 383). auxotrophic mutants. This problem was overThus, in the reaction tetrahydrofolate acts both come when it was discovered that thymineas a carrier of the Cl-unit and as reductant requiring mutants of E. coli were readily ob(55, 302). The enzyme was shown to catalyze a tained as mutants resistant to the folate antagtetrahydrofolate-dependent exchange between onists aminopterin or trimethoprim in the prescarbon 5 of dUMP and water. The exchange was ence of thymine (292, 293, 361). A mechanism for only partially dependent on addition of Mg2+ this selective procedure that is in accord with

312

O'DONOVAN AND NEUHARD

existing experimental data has been proposed independently by Bertino and Stacey (44) and Wilson et al. (395). Aminopterin and trimethoprim are potent inhibitors of dihydrofolate reductase which catalyzes the formation of tetrahydrofolate from dihydrofolate (Fig. 14). Thymidylate synthetase is unique among enzymes that catalyze C1 transfer reactions in that tetrahydrofolate is consumed (oxidized) during the reaction. Thus, the continued action of thymidylate synthetase in the presence of the analogues will rapidly deplete the supply of tetrahydrofolate. Since tetrahydrofolate derivatives are required in a variety of biochemical pathways including purine biosynthesis, the metabolism of certain amino acids, and initiation of protein biosynthesis (349), treatment with aminopterin or trimethoptim in the presence of thymine results in cessation of RNA, DNA, and protein synthesis. Therefore cells are only able to grow in the presence of these analogues and thymine under conditions in which thymidylate synthetase activity is depressed. The continued growth of thymine-requiring mutants in the presence of aminopterin and thymine indicates that the inhibition of dihydrofolate reductase is sufficiently leaky to allow the synthesis of catalytic amounts of tetrahydrofolate, which are required under these conditions. Genetics (thyA). Aminopterin selection seems to be generally applicable and has been used to isolate thymine-requiring mutants in a variety of microorganisms: Salmonella typhimurium (124), Aerobacter aerogenes (176), Serratia

BACMERIOL. REV.

marcescens (200); B. megaterium (382), and B. subtilis (128). Genetic studies of thymine-requiring mutants (thy) showed that in E. coli and Salmonella typhimurium the thy locus is localized between 7 and 8 o'clock on the genetic maps (Fig. 8; references 124, 200, 215) in between and closely linked to the lysA and argA (in S. typhimurium, argB) loci (124, 368). Since all thy mutants investigated lack thymidylate synthetase (26, 124, 176, 395), it is believed that the thy locus is identical with the structural gene for this enzyme (thy A). In a genetic study of 134 thy mutants of E. coli, Alikhanian (13) found that they were distributed over more than 17 sites within the thy locus. However, 62 of the mutants were localized within one site, and all of these turned out to be temperature sensitive, i.e., prototrophic at 28 C but thymine requiring at 37 C (13). It was shown recently that complementation may occur between certain thy mutants of S.

typhimurium. Detailed studies of this phenomenon indicated that the complementation was intracistronic rather than intercistronic, suggesting that thymidylate synthetase consists of two or more identical subunits (A. J. Holmes, Ph.D. Dissertation, Kansas State Univ., 1969). Biosynthesis of dUMP From UDP. Because dUMP seems to be the immediate substrate for thymidine nucleotide biosynthesis, it is of great importance for the cell to be able to supply adequate amounts of this compound. UDP is reduced by ribonucleoside diphosphate reductase of E. coli to dUDP. The further conversion of dUDP to dUMP seems to occur via dUTP. Extracts of E. coli are able to phosphorylate dUDP to dUTP, which in turn is rapidly degraded to dUMP by a specific dUTP pyrophosphatase as shown in Fig. 15 (43, 159). An alternative pathway for dUMP synthesis con5

dCTP. 2

dUTP 2

^ dTTP

CDP -4 dCDP

dUDP * UDP

dTDP

\\3dCMP

t3 dTMP - 9 CdR

-

6

\

4

12

A

~

&3 dUMP 1

UdR

FIG. 15. Pyrimidine deoxyribonucleotide interconversions in Escherichia coli and Salmonella typhimurium. Symbols: solid lines, established reactions; broken lines, possible reactions; dotted lines, postulated reacTHFA FIG. 14. Thymidylate synthetase and its relation to folic acid metabolism. Abbreviations: El, thymidylate synthetase; EII, dihydrofolate reductase; THFA, tetrahydrofolate; DHF, dihydrofolate.

tions. Enzymes are: 1, ribonucleoside diphosphate reductase; 2, nucleoside diphosphate kinase; 3, deoxyribonucleotide kinase(s); 4, dUTP pyrophosphatase; 5, dCTP deaminase; 6, cytidine deaminase; 7, thymidine kinase; 8, 5-CH3-dCTP synthetase and 5-CH,-

dCTP deaminase; 9, thymidylate synthetase.

VoL. 34, 1970

PYRIMIDINE METABOLISM IN MICROORGANISMS

313

sisting of a direct hydrolysis of dUDP does not to be of significance in E. coli (158). The dUTP pyrophosphatase has been purified extensively from extracts of E. coli B by Haggmark (Abstract, 4th FEBS meeting, Oslo, p. 296, 1967). It catalyzes the reaction shown in equation XII. The enzyme is specific for dUTP; dCTP, dTTP, and dUDP will not serve as sub-

TABLE 11. Precursors for dUMP and UMP biosynthesis in cytidine-requiring mutants of Salmonella typhimuriuma

strates.

dTTP ....... dCTP ........ CTP ......... UMP dUMP..

seem

dUTP + water dUMP + PPi (XII) Besides its role in dUMP formation, it has been suggested that dUTP pyrophosphatase is necessary to prevent intracellular accumulation of dUTP, and thereby to prevent the incorporation of uracil into DNA (43). From dCTP. During a study of the metabolism of cytosine compounds in S. typhimurium (282), an enzyme catalyzing the deamination of dCTP was discovered. This dCTP deaminase has been purified 15-fold from extracts of S. typhimurium LT-2 and shown to be specific for dCTP; CTP, dCDP, CDP, dCMP, and deoxycytidine are not deaminated by the enzyme (Neuhard and Thomassen, in preparation). The sequential action of dCTP deaminase and dUTP pyrophosphatase would therefore constitute an alternative route for the generation of dUMP (Fig. 15). To test for this pathway in vivo, two specially constructed mutants were employed: DP-55 (pyrA, pyrG, udp, cdd) which requires cytidine and uracil for growth (283), and its thymine-requiring derivative KP-100. They were grown in the presence of 5-3H-cytidine, 12C-uracil, '2P-orthophosphate, and (in the case of KP-100) "2C-thymine, and the ratio 'H to 12P of the individual pyrimidine nucleotides in the acid-soluble pool was determined. The results of such an experiment are shown in Table 11. The difference in specific activity of tritium in UMP and dUMP in KP-100 shows that a large fraction of the dUMP in this strain is derived directly from a cytosine compound by a pathway that does not involve any intermediate formation of uridine nucleotides (Table 11). Since the tritium label of cytidine is on the 5-carbon of the pyrimidine ring, no label would be found in thymidine nucleotides. This may indicate that the dCTP deaminase pathway is operating under these conditions. Since a low d1TP results in accumulation of dCTP (280; Table 11), an increased supply of substrate for dCTP deaminase is available under such conditions. In KP-100, the surplus of dUMP will not influence the dT`TP directly since this strain lacks thymidylate synthetase. However, the apparently close connection between thymidine nucleotide metabolism and the dCT`P deaminase pathway suggests that --

DP-55

KP-100

Nucleotide

sizeb

Pool

Ratio 3H/uP

Pool

sizeb

Ratio 3H/nP

0.29 0.48 1.97

0.00 1.00 1.00

0.07 0.76 1.10 0.29 0.96

0.00 1.00 1.00 0.16 0.53

.........

-

a Strains used require cytidine and uracil for growth. In addition, KP-100 requires thymine. Genotypes are given in the text. Cells were grown in presence of 5-'H-cytidine, 12C-uracil, 32Porthophosphate, and, in the case of KP-100, 12C-thymine (Neuhard and Thomassen, unpublished data). b Values are expressed as micromoles per gram dry weight.

it may have a physiological function in the biosynthesis of dUMP also in wild-type cells. It is significant in this context that the affinity of the purified ribonucleoside diphosphate reductase of E. coli for UDP is 10-fold lower than it is for CDP, GDP, and ADP (230, 231). We were recently able to demonstrate that a mutant of E. coli, HD 1038 (288), lacks dCT`P deaminase, whereas the parent strain, JC 411, possesses the enzyme (Neuhard and O'Donovan, in preparation). HD 1038 is indistinguishable from JC 411 except for its 10-fold-elevated endogenous dCTP pool and its decreased dTTP pool [30 to 50% of JC 411 (288)]. Infection of E. coli with T-even phages creates a situation in which thymidine nucleotide metabolism may be stressed. After infection, a phage-specific dCTPase is induced which hydrolyzes dCTP and dCDP to dCMP (218, 408), thereby preventing the incorporation of dCMP into phage DNA and at the same time providing dCMP for the synthesis of 5-hydroxymethyldCMP (135). Thus, the formation of dUMP via the dCTP deaminase pathway is completely prevented. However, a highly regulated phagespecific dCMP deaminase (feedback inhibited by dTTP) is also induced after infection (136, 261, 345). Therefore, under these conditions the sequential action of dCTPase and dCMP deaminase (Fig. 16), in addition to UDP reduction, will ensure an adequate supply of dUMP for thymidine nucleotide biosynthesis. A dCTP deaminase has recently been found in B. subtilis infected with phage PBS1, a phage that contains uracil in place of thymine in its DNA

314

O'DONOVAN AND NEUHARD

(375). The enzyme was shown to be phage specific and to be inhibited noncompetitively by dTTP. Its possible physiological importance has been discussed (375). From deoxyuridine. Deoxyuridine may be phosphorylated to dUMP by thymidine kinase (Fig. 15). E. coli and S. typhimurium are unable to phosphorylate deoxycytidine to dCMP (282; 0. Karlstrom, in preparation), but possess a very active deoxycytidine deaminase. Thus, deoxycytidine may be converted to dUMP after deamination. The contribution of thymidine kinase to the overall metabolism of dUMP is probably of minor importance since deoxyuridine and deoxycytidine are not normally available to the cells. However, if they are added exogenously to the growth medium they may, under certain conditions, be able to contribute significantly to the dUMP supply of the cells (288).

Two Pathways for Endogenous dTTP Biosynthesis? Bacillus subtilis. Genetic studies of thyminerequiring mutants of B. subtilis by Wilson and co-workers (395) showed that the thy character resulted from a double mutation in two unlinked genes, thyA and thyB. The genotypes thyA+ thyB and thyA thyB+, both thymine prototrophs, could nevertheless be distinguished by their different sensitivity to aminopterin in the presence of thymine (Table 12). Enzyme studies identified the thyA mutation with lack of thymidylate synthetase activity. CDP

-d dCDP

'E, dCMP d'cmpk""

dCTP

5-HM-dCMP

dUMP

-

dTMP

UDP

FIG. 16. Biosynthesis o dUMP from deoxycytidine nucleotides in Escherichia coli infected with T-even phages.

BAcrERuOL. REV.

Table 12 shows that the aminopterin-resistant mutants, type thyA thyB+, in contrast to thyA+ thyB mutants, undergo thymineless death when treated with aminopterin in the absence of thymine. Thus, in such mutants the drug seems to inhibit thymidine nucleotide synthesis without affecting RNA and protein synthesis. This led the authors to propose (395) that the thyB locus codes for a second pathway for generation of thymidine nucleotides, a pathway that does not require tetrahydrofolate but a closely related coenzyme which is not involved in other C1-transfer reactions and which is inhibited by aminopterin. No direct biochemical evidence for such a pathway has yet been obtained (F. Rothman, personal

communication). Enterobacteria. It is not known whether a situation similar to that in B. subtilis exists in E. coli and S. typhimurium. However, no ultimate data have been reported that exclude two independent pathways for thymidine nucleotide biosynthesis. In the following paragraph we will discuss certain genetic and biochemical observations of relevance for this problem. We will use the symbols thyA and thyB to identify the loci for thymidylate synthetase and the second hypothetical pathway, respectively. In a detailed genetic study of a large number of thymine-requiring S. typhimurium mutants, Eisenstark and co-workers (124) ruled out the possibility of two unlinked loci for the thy character by showing that thymine requirement can be transduced into a thy+ recipient. This is in accord with mapping data which indicate that all thy mutations map within a narrow region on the chromosome in both E. coli (13) and S. lyphimurium (124). However, this does not rule out the existence of two closely linked loci, thyA and thyB, both of which have to mutate in order to give the thy- character. In favor of such a situation, Eisenstark and co-workers (124) observed that cotransduction of thy with a neighbor marker (lys or arg) was less frequent than cotransduction of the thy+ character with

TABLE 12. Properties of the four possible thy genotypes of Bacillus subtilisa Growth in presence of aminopterin

Genotype

Genotype

~

Thymine requirement

Thysnidylate

synthetase

With thymine

thyA+, thyB+ thyA+. thyB thyA, thyB+

No No No

Yes Yes No

No growth No growth

thyA, thyB

Yes

No

Normal growth

Data are from reference 395.

Normal growth

Without thymine

No growth No growth Unbalanced growth Thymineless death Unbalanced growth Thymineless death

VOL. 34, 1970

PYRIMIDINE METABOLISM IN MICROORGANISMS

the same markers. Furthermore, they found that some thy+ revertants remained aminopterin resistant, thus phenotypically similar to the thyA thyB+ class of B. subtilis (Table 12). Unfortunately, it is not known whether this class of revertants lacks thymidylate synthetase. Recent studies of the acid-soluble nucleotide pools of such trimethoprim-resistant thy+ revertants of S. typhimurium (Neuhard and O'Donovan, unpublished data) revealed that they have extremely low dTTP pools when grown in the absence of thymine, indicating that they are stressed in their ability to synthesize thymidine nucleotides endogenously. At this point it should be recalled that 40% of all thy mutants isolated by Alikhanian and coworkers in E. coli (13) were temperature sensitive. These temperature-sensitive mutants all mapped within one site of the thy locus to the left of all the rest, which were distributed over more than 17 sites. It would be of interest to determine whether there exists a relationship between the temperature-sensitive thy phenotype in E. coli and the aminopterin-resistant thy+ revertants of S. typhimurium. A short report by Forster and Holldorf (Abstract, 2nd FEBS meeting, Vienna, p. 146, 1965) described a new pathway for dTTP synthesis in extracts of E. coli. As shown in equations XIII and XIV, it consists of the conversion of dCTP to dTTP by the sequential action of a methylating enzyme and a deaminase. dCTP + 5,10-methylenetetrahydrofolate 5-methyl-dCTP + dihydrofolate (XIII) 5-methyl-dCTP -, dTTP + ammonia (XIV) The methylating enzyme catalyzing the first reaction (equation XIII) was separated from thymidylate synthetase on DEAE-cellulose and it was reported that thy mutants lacked both this enzyme and thymidylate synthetase. Whether the 5-methyl-dCTP deaminase is the same enzyme as the dCT`P deaminase described previously is not known at present. In his thesis, A. W. Holldorf (Albert-Ludwigs-UniversitAt, Freiburg, 1967) described the isolation of two classes of thy+ revertants, one containing only thymidylate synthetase activity and no dCTP methylating activity and a second in which the reverse combination of methylating enzymes was found. No phenotypic differences between the two classes were reported. Although favoring strongly the hypothesis of two mutations being necessary to obtain thy mutants, any discussion of Holldorf's data will have to wait until further details have been published. To study in more detail the actual pyrimidine

315

precursors for endogenous thymidine nucleotide biosynthesis, two independent series of experiments have been performed (212, 282). Cytidine deaminase-negative mutants (cdd) of E. coli B (212) and S. typhimurium LT-2 (282) were labeled differentially in their cytosine and uracil compounds in vivo. Analysis of the labeling pattern in the dTMP moieties of DNA (212) or the acid-soluble dTTP pool (282) showed that 75 to 80% of the thymidine nucleotides were derived from a cytosine compound without equilibration with the uridine nucleotide pool of the cells. These results seem to support the existence of a direct pathway from a cytidine or deoxycytidine nucleotide to dTTP, accounting for 80% of the d1TJ7P synthesized (compare equations XIII and XIV). However, they can also be interpreted in accordance with the existence of only the thymidylate synthetase pathway. In the latter case, the data indicate that 75 to 80% of the dUMP is derived directly from dCTP as outlined in Fig. 15, and only 20 to 25% via the direct reduction of UDP. At present no definite decision between these two alternatives can be made. Thus, it is not possible from our present knowledge to state whether one or two independent pathways are involved in thymidine nucleotide biosynthesis in E. coli and S. typhimuriun. However, if two loci thyA and thyB, are involved, the following characteristics may be predicted. (i) Loci thyA and thyB would be closely linked. (ii) Mutants of the genotypes thyA thyB+ and thyA+ thyB, as in B. subtilis, would be phenotypically thy+; the thy- character would be the result of a double mutation. It can be argued that if this is the case, thy mutants should be extremely rare; however, the special aspects of the aminopterin selection procedure may explain this phenomenon. The possibility that thyA thyB+ and thyA+ thyB are phenotypically thy- has been ruled out by complementation studies among a large number of thy mutants in S. typhimurium (A. J. Holmes, Ph.D Dissertation, Kansas State Univ., 1969). (iii) Both pathways would be inhibited in vivo by aminopterin and trimethoprim. (iv) Since addition of 5fluorodeoxyuridine results in thymine starvation and thymineless death (98, 280), both pathways would be inhibited by 5-fluoro-pyrimidine derivatives. In the case of thymidylate synthetase (thyA), this derivative seems to be 5-fluorodUMP.

INTERCONVERSIONS OF NUCLEOTIDES At this point we would like to make a few general statements about the interconversion of

316

O'DONOVAN AND NEUHARD

pyrimidine nucleotides which will help to integrate pyrimidine metabolism. (i) Although very little is known about the conversion of nucleoside monophosphates to the corresponding di- and triphosphates, e.g., NMP-+ NDP -- NTP or dNMP -- dNDP -- dNTP, nevertheless kinases exist which are capable of catalyzing any of these phosphorylations [N being uracil, cytosine, or thymine (39, 247)]. (ii) It is assumed that cells are able to convert any nucleoside triphosphate to the corresponding di- and monophosphate. Two enzymes involved in these conversions have been characterized: nucleoside diphosphate kinase (39, 247) which seems to be readily reversible and quite unspecific, and dUTP pyrophosphatase (Fig. 15; references 43, 159). The latter enzyme may be very significant in vivo in providing a source of dUMP as substrate for thymidylate synthetase. (iii) Since mutants lacking CTP synthetase (pyrG) have an absolute requirement for cytidine (283), only one enzyme exists for the conversion of uridine nucleotides to cytidine nucleotides, e.g. CTP synthetase. (iv) Pyrimidine-requiring mutants of S. typhimurium and E. coli blocked in cytosine deaminase (cod), cytidine deaminase (cdd), and CTP synthetase (pyrG) have an absolute requirement for both cytidine and uracil, indicating that no deaminases exist for cytosine ribonucleotides (282). However, at the deoxyribonucleotide level a specific dCTP deaminase has been found. As pointed out in paragraph ii above, if dUTP -- dUMP is a key pathway for dUMP synthesis, then a ready-made function for dCTP deaminase in dTMP synthesis is seen. In higher organisms, there exists a highly regulated dCMP deaminase (343), ensuring an adequate supply of dUMP for thymidine nucleotide biosynthesis. This enzyme is absent in the enterobacteria except after T-even infection. The only microorganism in which dCMP deaminase has been reported is L. acidophilus (350). (v) The dUMP is converted to dTMP by thymidylate synthetase (383). (vi) There may be a direct link between dCTP and dTTP if the enzyme system reported by Forster and Holldorf (Abstract, 2nd FEBS Meeting, Vienna, p. 146, 1965) in E. coli proves to be of general importance.

BACTERIOL. REV.

The fact that uracil, cytosine, uridine, cytidine, deoxyuridine, and deoxycytidine will satisfy the pyrimidine requirement of such mutants (93, 274) further indicates that considerable interconversions may occur among these compounds. In accord with this, isotope studies showed that any of the above mentioned pyrimidine compounds were equally effective as precursors for all the pyrimidine moieties of RNA and DNA (57, 58, 239, 351). Although enzymes involved in the metabolism of pyrimidine bases and nucleosides have been known for many years, very little is known about their properties or regulation. The recent isolation of mutants defective in these enzymes has made it possible to construct a more complete metabolic map of the pathways involved. Uracil Compounds UMP pyrophosphorylase (upp*). The genetic symbol upp* is proposed for this enzyme which catalyzes the formation of UMP from uracil and phosphoribosyl pyrophosphate as shown in equation XV (105). Mg2 +

U + PRPP

UMP +

PPi (XV)

Partial purification of the enzyme from L. bifidus separated it from the de novo enzyme OMP pyrophosphorylase. UMP pyrophosphorylase has been found in E. coli (68, 105, 272), and indirect evidence for its existence in S. typhimurium (282) and Saccharomyces cerevisiae has been presented (160). Cytosine will not act as substrate for this enzyme (89), although 5-fluorouracil is converted to 5-FUMP by extracts of E. coli containing the enzyme (68). Mutants resistant to 5-fluorouracil are readily obtained and were shown by Brockman and co-workers (68) to lack UMP pyrophosphorylase. Such mutants are still sensitive to 5-fluorouridine (Table 13). It seems quite likely that several other uracil analogues such as 6-azauracil and 5-azauracil are toxic only after they have been converted to the corresponding nucleotide by UMP pyrophosphorylase (85, 171). It has been suggested, based on indirect evidence, that the enzyme in Salmonella typhimurium is feedback controlled, probably by UTP (282). Recently, Molloy and Finch (272) observed that a crude preparation of UMP pyrophosphorylase from E. coli is strongly activated by GTP and METABOLISM OF BASES AND inhibited by UMP and UTP. NUCLEOSIDES Uridine phosphorylase (udp*) was first charThe existence of bacterial mutants with an acterized by Paege and Schlenck (299, 300; absolute requirement for pyrimidines shows that equation XVI). alternatives to the de novo pathway for the generation of pyrimidine nucleotides must exist. Uridine + Pi uracil + ribose-1-P (XVI)

317

PYRIMIDINE METABOLISM IN MICROORGANISMS

VOL. 34, 1970

TABLE 13. Sensitivity of Salmonella typhimurium mutants towards pyrimidine analoguesa Mutant genotype

5-FU

s-FUR

5-FUdR

5-FC

5-FCR

5-FCdR

Wild strain

S R S R S S R

S

S

S

S S S

S S R

S S S

S R S R R S R

S

S S R

S S R S S R

S S S S R R

UPP

udp upp, udk cod cdd upp, cdd, udk

a Key: R, resistant; S, sensitive.

Although several uridine analogues including 5-azauridine (85), 6-azauridine (85), and 5-fluorouridine (Neuhard, unpublished data) seem to be substrates for the enzyme, cytidine, orotidine, and deoxyuridine are not phosphorylyzed by partially purified enzyme preparations from E. coil (300, 321). The enzyme is induced by uridine (304) and cytidine (282) and is responsible for the ability of E. coil and S. typhimurium to grow on uridine or cytidine as the sole sources of carbon. This property has been used to select for udp mutants in S. typhimurium (283). The enzyme seems to be of widespread occurrence in microorganisms (197, 299). In B. subtilis and B. stearothermophilus, the enzyme is less specific. Both uridine and thymidine are substrates and able to induce it (342). Extracts of E. coli are able to catalyze a reversible transfer of ribosyl group from a purine ribonucleoside to uracil (217). Subsequently, transribosylations were shown to occur in many microorganisms and it was shown that they were dependent on the presence of both purine nucleoside phosphorylase and uridine phosphorylase. Thus, the transfer reaction is brought about by the coupled action of these two enzymes (197). Yeast (79) and L. pentosus (228, 384) catabolize pyrimidine nucleosides hydrolytically. In L. pentosus, both uridine and cytidine are hydrolyzed, whereas the yeast nucleosides seem to be specific for uridine (79). Mutants defective in uridine hydrolase have recently been isolated in Saccharomyces cerevisiae (160). Uridine kinase (udk*). Although uridine kinase (equation XVII) of higher organisms has been the subject of many studies (296, 357), very little is known about this enzyme in microorganisms.

been used to select for uridine kinase (udk) mutants in upp mutants of Salmonella typhimurium (Neuhard, unpublished data). Mutants of Saccharomyces cerevisiae lacking uridine kinase have likewise been isolated as mutants resistant to 5-fluorouridine (160). Uridine kinase of Streptococcus faecalis, Staphylococcus aureus, and E. coil has been reported to be feedback inhibited by UTP and CTP (16). Evidence for the operation of such a control system in vivo in mutants of Salmonella iyphimurium has been obtained (282). Deoxyuridine metabolism. Deoxyuridine may be phosphorylated to dUMP by thymidine kinase. That extensive catabolism of deoxyuridine may also occur is shown by the ability of E. coli and S. typhimurium to grow on deoxyuridine as the sole source of carbon (127). Thymidine phosphorylase-negative mutants (tpp) have been obtained as mutants unable to use thymidine as carbon source. Such mutants have simultaneously lost the ability to use deoxyuridine as sole energy source (127), indicating that thymidine phosphorylase is responsible for the catabolism of deoxyuridine. Since 5-fluorodeoxyuridine is a substrate for both thymidine kinase and thymidine phosphorylase, it may be predicted that a double mutation is required to get cells that are resistant to this analogue.

Cytosine Compounds Cytosine deaminase (cod*). Cytosine deaminase (cod*) was found in yeast and E. coil 45 years ago by Hahn and co-workers (167, 168; equation XVIII). Several other microorganisms have subsequently been shown to contain this enzyme (66, 177). Cytosine + water = uracil + NH3 (XVIII) Mg2+ Besides cytosine, the analogues 6-azacytosine ' UMP + ADP (XVII) (66, 249), isocytosine (249), and 5-fluorocytosine Uridine + ATP It has been mentioned (Table 13) that upp (282) are substrates for the enzyme. 5-Methylmutants, although resistant to 5-fluorouracil, cytosine is deaminated by the yeast enzyme are still sensitive to 5-fluorouridine. This has (222) but not by extracts of E. coli (93). 5-AzacyI

O'DONOVAN AND NEUHARD

318

tosine and 5-azauracil are inhibitors of the enzyme (88). Cytosine deaminase-deficient mutants are easily obtained since they are phenotypically 5-fluorocytosine resistant and 5-fluorouracil sensitive (282, 283; Table 13). The fact that cytidine-requiring mutants of S. typhimurium with the genotype pyrG, cdd, cod are unable to satisfy their cytidine requirement by cytosine shows that neither a CMP pyrophosphorylase nor a cytidine phosphorylase, acting in reverse, is operating in these cells (282; Fig. 17). Cytosine deaminase mutants have also been isolated in Saccharomyces cerevisiae (160). Cytidine (deoxycytidine) deaminase (cdd). Cytidine deaminase of E. coli and S. typhimurium (equation XIX) is extremely active. Cytidine (deoxycytidine) + water uridine (deoxyuridine) + NH3 (XIX) It permits the cells to grow with normal growth rates in minimal medium with cytidine or deoxycytidine as the sole source of nitrogen. The enzyme is induced by high concentrations of cytidine (304; J. Schaumburg, personal communication). It deaminates both cytidine and deoxycytidine, the latter being a better substrate than cytidine (385); 5-methyldeoxycytidine,

5-bromodeoxycytidine 5-chlorodeoxycytidine, (97), 5-azacytidine (87), and 5-fluorodeoxycytidine (211) will all serve as substrates. 5-Fluorodeoxycytidine is of special interest, since it has been used for selecting mutants lacking cytidine deaminase (cdd; reference 211). The cdd mutation makes the cells phenotypically 5-fluorodeoxycytidine resistant and 5-fluorodeoxyuridine sensitive (Table 13). Pyrimidinerequiring strains harboring the cdd mutation are unable to use deoxycytidine as a pyrimidine source, indicating that deoxycytidine cannot be converted directly to cytosine (Fig. 17). Because of the high activity of cytidine deaminase nor-

BACMPRIOL. REV.

mally present in E. coli and S. typhimurium, mutants lacking CTP synthetase (pyrG), i.e., cytidine-requiring mutants, can only be obtained if the parent strain used for the selection is cdtd (283; C. Beck, personal communication). Yeasts (160) and Lactobacilli (384) seem to be devoid of cytidine or deoxycytidine deaminase activity. Instead, Lactobacilli have acquired an ability to degrade cytidine hydrolytically to cytosine and ribose (228). (Deoxycytidine was not tried.) The implications of these differences in the metabolism of cytosine compounds by Lactobacilli are discussed later. Cytidine and deoxycytidine kinases. The mere existence of cytidine-requiring mutants (pyrG) in E. coli (Neuhard, unpublished data) and S. typhimurium (283) shows that these cells possess a cytidine kinase (212). This explains the toxicity of 5-fluorocytidine in cdd mutants (O'Donovan, unpublished data). During a selection for cytidine kinase mutants in a cdd,upp mutant of S. typhimurium with 5-fluorocytidine as the selective agent, it turned out that all 5-fluorocytidine-resistant mutants obtained had acquired resistance to 5-fluorouridine simultaneously (Table 13). This seems to indicate that uridine kinase and cytidine kinase in S. typhimurium, as in higher organisms (296), is one and the same enzyme. Studies of deoxycytidine uptake in cdd mutants of E. coli (Karlstrom, in preparation) and S. typhimurium (282) have indicated that these organisms do not contain a deoxycytidine kinase (Fig. 17). Permeases It has been established that pyrimidine nucleotides are not taken up as such by E. coil but are dephosphorylated (239). However, our knowledge of the uptake mechanism for bases and nucleosides is somewhat limited. A 6-aza-uracilmuitant %V1 of a..F L19, rnli unahle resistant ago un L"lV L%, Wto take L"9 sCzLLL JLLLULCL11L

skL%.;

O

V

me

has been isolated. exogenous that it ishas mutated in a uracilIt perbeen claimed'4C-uracil, i mease (uraP; R. Lavalle, personal communication, ll reference 368). Evidence for a specific transfitand aPP port system involved in uridine and cytidine \ uptake in E. coli has also been obtained (304). UR CdR dR URUUR vC Recently, the mechanisms responsible for II the uptake of exogenous pyrimidine comI 111I )t AL it pounds in yeast have been studied by Grenson \ \11 / { M/ PRPP (160). She was able to show that Saccharomyces \\ cerevisiae contains at least three distinct permeases, one specific for cytosine, one for uracil, -.-U debOs-I- T and one for uridine. By isolating mutants de.f.yri bases and FIG. 17. Metabolism cleosides in Escherichia coli and Salmonella typhimu- fective in each one of these transport systems, rium. Solid lines rep)resent established reactions; she found that the uracil and uridine permeases broken lines represent nzonexisting reactions. were subject to feedback inhibition by internal CMP

dCMP

X4,UMP\

dUMPAS

dTMP

/

-

\

\

n

nu

VOL. 34, 1970

319

PYRIMIDINE METABOLISM IN MICROORGANISMS

pyrimidines. The significance of such a control system may be to link the uptake of exogenous pyrimidines to the metabolism of the corresponding nucleotides as determined by net nucleic acid synthesis (122, 286). Genetics Very little is known about the genetics of the different enzymes shown in Fig. 17. Preliminary mapping data (C. Beck, personal communication) indicate that the udp and cdd, loci in Salmonella typhimurium are not linked to each other or to any of the pyr genes. The udp gene is cotransducible with metE and is therefore located around 122 min on the S. typhimurium linkage map (341). The cdd gene seems to be covered by the F'32 episome (133), indicating that it maps somewhere between araD and metG (Fig. 8). The uracil permease gene uraP has been localized quite precisely on the E. coli linkage map (368) very close to 50 min (Fig. 8). Genetic studies with mutants of Saccharomyces cerevisiae have shown that the loci for the enzymes involved in the metabolism of exogenous pyrimidine compounds in this organism are also unlinked. Discussion Figure 17 summarizes our present knowledge of the metabolism of pyrimidine bases and nucleosides in E. coli and Salmonella typhimurium. Concerning the physiological importance of these pathways, the following points can be made. (i) Two pathways for the synthesis of UMP from uracil exist (Fig. 17). Introduction of a upp mutation into a pyrimidine-requiring strain of S. typhimurium results in a mutant (pyr, upp) which cannot use uracil or cytosine as pyrimidine source, indicating that the direct conversion of uracil to UMP (equation XV) is the preferred route for utilization of exogenous uracil. However, uracil may serve as sole pyrimidine source for this mutant if a purine ribonucleoside is added simultaneously. Thus, although uridine phosphorylase normally is purely catabolic, probably due to further metabolism of ribose1-P, it may catalyze the synthesis of uridine from uracil in vivo if a surplus of ribosyl groups is available (Neuhard, unpublished data). (ii) Likewise, two pathways for the conversion of uridine to UMP exist (Fig. 17), one of which requires the prior conversion of uridine to uracil. In the pyr,upp mutant mentioned above, one of these pathways is blocked by mutation (upp). Such a mutant grows with normal growth

rate on uridine as a pyrimidine source, but the growth yield is extremely low (experiment 2, Table 14). Introduction of a udp mutation into this strain increases the growth yield on uridine significantly (Table 14). Thus, it appears that uridine kinase is sufficiently active to supply the cells with adequate amounts of UMP (compare growth rates in Table 14), but that the supply of uridine is rapidly converted to uracil by catabolic enzymes present in the cells. The fact that some catabolism of uridine occurs even in a udp mutant (experiment 3, Table 14) may indicate the existence of a nucleosidase that is able to hydrolyze uridine to uracil (Neuhard, unpublished data). (iii) Mutants lacking cytidine deaminase cannot use cytidine as a carbon source, whereas udp mutants have lost the ability to grow on either uridine or cytidine as the sole sources of carbon (Neuhard, unpublished data). Similarly, deoxycytidine will only serve as a carbon source if the cells contain both cytidine deaminase and

thymidine phosphorylase (127). Thus, cytidine and deoxycytidine are catabolized solely via the corresponding uracil compounds (212). (iv) Except for the existence of a cytidine kinase, which may be regarded as a uridine kinase with broad specificity, cytosine compounds must first be deaminated in E. coli and S. typhimurium to make them useful for the cells. Since both cytosine deaminase and cytidine deaminase have broad specificities for cytosine analogues, caution must be exhibited when the effects of such analogues are studied in cells containing these deaminases (see PYRIMIDINE ANALOGUES). (v) It is evident from this section and from Fig. 17 that extensive interconversions of pyrimidine bases and nucleosides occur in vivo. It is important, therefore, to be aware of these interconversions when these same compounds are used as expensive radioactive precursors of nucleic acids. TABLE 14. Growth yield of pyrimidine-requiring Salmonella typhimurium mutants on uridine as

pyrimidine

sourcee

Growth yield with 6

pg of uridine/ml

Expt

1 2 3

Mutant genotype

pyr

pyr, upp pyr, upp, udp

Growth

OD4uo (nm)

%

(doublings/ hr)

0.530 0.040 0.320

100 8 60

1.4 1.4 1.4

a Data are unpublished results from J. Neuhard and E. Thomassen.

320

O'DONOVAN AND NEUHARD

METABOLISM OF THYMINE AND THYMIDINE Incorporation of Thymine and Thymidine into DNA in E. coli Exogenous thymine is not incorporated into DNA in wild-type cells of E. coil (104, 351). However, the mere existence of thymine-requiring mutants shows that enzymes must exist for the conversion of thymine to dTMP. In contrast, exogenous thymidine is readily incorporated into DNA in wild-type cells, but the incorporation stops after a short time due to the rapid breakdown of thymidine to thymine by the inducible thymidine phosphorylase (315). Accordingly, mutants of E. coli lacking thymidine phosphorylase (tpp) were able to incorporate thymidine for extended periods of time (126, 127). By introducing a thy mutation into a tpp mutant, a thymidine-requiring mutant was obtained which was unable to grow on thymine (126). This shows that only one pathway exists for the conversion of thymine to dTMP, and that this pathway requires a functional thymidine phosphorylase (Fig. 17, 20). Addition of ribo- and deoxyribonucleosides promotes the incorporation of thymidine into DNA of wild-type cells (59). This was explained by Budman and Pardee (74), who found that deoxyadenosine and uridine inhibited thymidine phosphorylase competitively. A third way to promote thymidine incorporation is to use dTMP as a source of thymidine. Exogenous dTMP has to be dephosphorylated before it is utilized (239). Since the dephosphorylation seems to be rate limiting, dTMP addition will result in slow feeding of thymidine, thereby avoiding the induction of thymidine phosphorylase (64). As stated above, wild-type cells of E. coil do not incorporate exogenous thymine into their DNA. However, if a deoxyribonucleoside is added simultaneously, an extensive incorporation of thymine is observed (Fig. 18; references 59, 74, 206, 276). Purine deoxyribonucleosides have a greater effect than pyrimidine deoxyribonucleosides (276). Thus, it appears that the lack of incorporation of thymine is due to a lack of endogenous deoxyribosyl groups, probably deoxyribose-l-phosphate, necessary for formation of thymidine, and that they may be supplied from outside by addition of deoxyribonucleosides. This explanation is supported by the findings that deoxyadenosine fails to stimulate thymine incorporation in mutants defective in either purine nucleoside phosphorylase (pup) or thymidine phosphorylase

BAcrERIOL. REV.

15 0

0

*C(OXYACCNOSINE

100

200

PER CENT MASS INCREASE FIG. 18. Incorporation of exogenous (2-14C)-thymine into acid insoluble material by exponentially growing cells of Escherichia coil (wild type) in presence and absence of deoxyadenosine (0.2 mM) (A. MunchPetersen, personal communication).

(tpp; references 74, 278). The promoting effect of deoxyribonucleosides on thymidine incorporation, mentioned above, may therefore be due partially to recycling of thymine, formed by the degradation of thymidine.

Enzymes Thymidine phosphorylase (tpp). Thymidine phosphorylase (tpp; equation XX) is of widespread occurrence in microorganisms (197, 342). Lactobacilli, however, seem to be devoid of thymidine phosphorylase (384; W. Uerkvitz, personal communication). The enzyme from E. coil appears to be localized near the cell surface as judged by the fact that it is quantitatively released from the cells by osmotic shock (206, 276). Thymidine + Pi thymine + deoxyribose-l-P (XX) It has been partially purified and its specificity has been studied in great detail (320, 321). Deoxyuridine, thymidine, and the 5-halogensubstituted deoxyuridines are phosphorolyzed by the enzyme, whereas deoxycytidine, a-thymidine (the unnatural anomer), pyrimidine ribosides, and pyrimidine arabinosides are essentially inactive. Several ribo- and deoxyribonucleosides inhibit the enzyme competitively (30, 74, 277). Recently the enzyme from E. coil was obtained in an essentially pure form (M. Vilstrup, personal I

VOL. 34, 1970

PYRIMIDINE METABOLISM IN MICROORGANISMS

communication). Kinetic studies of this preparation indicate that the reaction is ordered and sequential, and that phosphate must bind to the enzyme prior to thymidine. Thymidine phosphorylase of E. coil and S. typhimurium are inducible (315), and 100-fold induced levels may be observed under certain conditions (M. Vilstrup, personal communication). The actual inducer seems to be a product of deoxyribonucleoside catabolism, probably deoxyribose-5-P. Thymidine phosphorylase mutants were isolated both in E. coil and S. typhimurium (10, 127), and their location on the genetic maps of both organisms were determined to be close to the thr locus around 12 o'clock (Fig. 8; references 10, 108). Thymidine kinase (tdk). Thymidine kinase (tdk) from E. coil was extensively purified and its properties were studied in great detail by Okazaki and co-workers (294, 295). Besides thymidine (equation XXI), deoxyuridine and the 5-halogenated deoxyuridines are active as substrates. dTIP inhibits the enzyme specifically, whereas dCDP, dCTP, and to a lesser extent dADP act as activators. dTMP Thymidine + ATP (XXI) + ADP Kinetic studies showed that in the absence of activators, ATP gave sigmoidal saturation kinetics, whereas they were hyperbolic in their presence. Addition of dCDP increases the Vmax and decreases the Km value for thymidine (295). Sucrose gradient centrifugation and gel filtration of the enzyme indicated that addition of effectors, regardless of whether negative (dTI-P) or positive (dCDP or dADP), resulted in a dimerization of the enzyme (202). The monomer appeared to be extremely temperature sensitive, the maximal reaction rate being obtained at temperatures below 30 C (depending on the concentration of thymidine). Higher temperature resulted in lower reaction rates. These changes were completely reversible. In the presence of effectors, the temperature sensitivity was abolished (203). Based on these results, Iwatzuki and ,Okazaki (203) proposed the following model. In the absence of effectors, the enzyme is a monomer which may exist in a catalytically active or less active form. Low substrate concentrations and high temperatures favor the less active form of the monomer. In the presence of effectors, the monomer is converted to a dimer which may assume an active or an inactive conformation, depending on the effector. Studies of the acid-soluble nucleotide pools of

321

E. coil (280) showed that cells which are limited in their supply of dTTP contain high levels of dCTP. Thus, according to the results presented above, thymidine kinase may be expected to be greatly stimulated under these conditions. E. coii mutants lacking thymidine kinase (tdk) have been isolated and characterized (180). The structural gene for the enzyme has been precisely located near the cluster of the tryptophan structural genes (196). Lactobacilli Deoxyribonucleotide metabolism in Lactobacilli deviates considerably from that of the majority of other bacteria. Most species of Lactobacillus have an absolute growth requirement for one deoxyribonucleoside (any deoxyribonucleoside) and a purine and pyrimidine. In some species, i.e. L. leichmannii and L. lactis, the deoxyribonucleoside requirement may be replaced by vitamin B12 (35). Since the ribonucleoside triphosphate reductase of L. leichmannii has an absolute requirement for coenzyme B12, the nutritional requirement of these cells indicates that in the absence of vitamin B12 they are capable of synthesizing the four deoxyribonucleoside triphosphates from the corresponding deoxyribonucleosides. Furthermore, the cells must have a considerable capacity for catalyzing deoxyribosyl transfer reaction to be able to synthesize all four deoxyribonucleosides from any one of them. In most bacteria, deoxyribosyl transfer between purines and pyrimidines is catalyzed by the coupled action of purine nucleoside phosphorylase and thymidine phosphorylase (197). However, Lactobacilli have acquired a specific enzyme, trans-N-deoxyribosylase, which catalyzes a phosphate independent transfer of deoxyribose moieties between a wide variety of purines and pyrimidines (equation XXII; references 35, 207, 259, 263, 270, 339). Baser-deoxyriboside + base2 base, + base2-deoxyriboside (XXII) The specificity and kinetics of a partially purified preparation of trans-N-deoxyribosylase from L. leichmannii (35) and a crystalline preparation from L. helveticus (W. Uerkvitz, personal communication) have been studied. Essentially identical results were obtained with these two enzymes. The synthesis of the enzyme has been reported to be derepressed during unbalanced growth caused by starving L. leichmannii for either vitamin B12 or deoxyribonucleosides (35). A prerequisite for the utilization of deoxyribonucleosides for deoxyribonucleotide synthesis by Lactobacilli is that they contain deI

O'DONOVAN AND NEUHARD

322

BAcrERIOL. REV.

studies indicated that uc-2 was blocked in the conversion of thymidine to thymine, uc-3 in the step between 5-hydroxymethyluracil and 5-formyluracil, and uc-4 in a step that prevented the utilization of uracil and thereby also the preceding compounds of the pathway (Fig. 19). Furthermore, the uc-2 mutant failed to incorporate 14C-thymidine into its DNA, indicating that N. crassa does not contain a thymidinephosphorylating enzyme. This is in accord with the finding (162) that extracts of Neurospora, Aspergillus, and Saccharomyces do not contain detectable amounts of thymidine kinase. Recently it was shown that cell-free extracts of N. crassa were able to oxidize thymidine to Neurospora ribosylthymine without cleavage of the glycosidic bond (346). The reaction is dependent on oxygen of to cultures '4C-thymidine of Administration N. crassa failed to label DNA preferentially. In and requires in addition Fe'+, a-keto-glutarate, contrast, the pyrimidine moieties of both RNA and ascorbate. Preliminary kinetic studies and DNA were labeled with equal efficiency. indicated that the extract initially converted If thymidine-methyl-3H was used, no label was thymidine to ribosylthymine which was subsefound in either RNA or DNA (130-132). These quently hydrolyzed to thymine and ribose (Fig. results suggested that Neurospora probably 19, dotted arrows). If this latter pathway turns cannot phosphorylate thymidine but that it can out to be responsible for the in vivo conversion convert thymidine to a pyrimidine which is of thymidine to thymine in N. crassa, the uc-2 utilized equally well for RNA and DNA syn- mutant mentioned above may be blocked in a thesis. Subsequent studies indicate that Neuros- "thymidine-2'-hydroxylase" rather than in a pora is capable of demethylating thymine to "pyrimidine deoxyribonucleosidase" as was uracil, 5-hydroxymethyluracil, 5-formyluracil, proposed by Williams and Mitchell (394). and uracil-5-carboxylic acid being intermediates THYMINE-REQUIRING MUTANTS (Fig. 19). The first three steps of this pathway Properties seem to be catalyzed by mixed-function oxidases, which in addition to oxygen require Fe +, a-ketoThe thymine requirement of thy mutants of glutarate, and ascorbate (1, 2, 386). E. coli may be satisfied by thymine, thymidine, Further evidence for the pathway shown in or 5-methyl deoxycytidine (25, 97). As has been Fig. 19 was obtained by Williams and Mitchell discussed, thymidine is an obligatory inter(394), who isolated three different mutants of mediate in the conversion of thymine to dTMP N. crassa (uc-2, uc-3, uc-4) blocked in different (Fig. 17). Uridine (cytidine) inhibits growth of steps in the conversion of thymidine into an thy mutants (74, 176), presumably by inhibiting RNA precursor. Nutritional and accumulation thymidine phosphorylase and thereby the uptake of thymine (74). Thymine starvation of thy dUMP dTMP UMP mutants results in unbalanced growth and 4 thymineless death (96, 107, 176, 358, 382). Removal of thymine is also accompanied by the following significant changes in the nucleotide C2 uc-4 UR TdR UdR metabolism of the cells: (i) increased differential rate of synthesis of ribonucleoside diphosphate uc-2 ? TR reductase (47) and thymidine phosphorylase I /1I< T ...... (60); (ii) intracellular accumulation of dATP, dCTP, and dCMP (280; Neuhard, unpublished 5-carboxyl-U deoxyribose--(P data); (iii) increased catabolism of pyrimidine t U deoxyribonucleotides (Neuhard, unpublished i 5-hdroxymethyl-U 5-formyl-U data); and (iv) excretion of uracil and other FIG. 19. Thymidine metabolism in Neurospora nucleic acid precursors into the medium (25, 60). crassa. Broken lines represent nonexisting reactions; These changes may be regarded as the cell's dotted lines represent possible alternative to thymidine attempts to counteract the lack of d1TP caused mutants in blocked different phosphorylase. Enzymes by removal of exogenous thymine. are indicated. oxyribonucleoside kinases. In view of the fact that neither E. coli (Karlstrom, in preparation) nor S. typhimurium (282) contains such activities (other than tdk), a study of the specificity and eventual regulation of these enzymes in Lactobacilli would be of great interest. In accord with the physiological role of deoxyribonucleosides in Lactobacilli, these cells differ from most other bacteria in their inability to catabolize deoxyribonucleosides. Thus, they do not contain thymidine phosphorylase, purine nucleoside phosphorylase, cytidine (deoxycytidine) deaminase, or adenosine (deoxyadenosine) deaminase (228, 384).

........u

--

?

VO.L. 34, 1970

PYRIMIDINE METABOLISM IN MICROORGANISMS

and Low Thymine-Requiring Mutants The thy mutants isolated by the aminopterin procedure (Fig. 14) require high concentrations of exogenous thymine for growth, i.e. 20 to 50 pg/ml (176). From such high thymine-requiring auxotrophs (thy, tlr+), low thymine-requiring thy mutants (thy, tlr-), capable of growing on 1 to 2 jLg of thymine per ml, are readily derived as the result of a second mutation (60, 176). The E. coli 15 T- strain originally studied by Cohen and co-workers (96) was a low thyminerequiring derivative of the strain isolated in 1947 by Roepke (335, 336). Two classes of tlrmutants may be distinguished. One is inhibited by high concentrations of thymidine or other deoxyribonucleosides (tds) and the other is resistant to these compounds (13, 278). The genetic loci for the tir character of both classes were closely linked but mapped separately from thyA, close to the thr locus (13, 124, 291; Fig. 8). The first indication that the tlr- phenotypes were related to deoxyribonucleoside catabolism was provided by Breitman and Bradford (60, 61), who observed, that low thymine-requiring thy mutants, in contrast to their high thyminerequiring parents, excreted large amounts of deoxyribose into the medium when starved for thymine. Subsequently, the mutations responsible for the two tar- phenotypes were identified as defects in phosphodeoxyribomutase, (drm) and deoxyriboaldolase (dra), respectively. The thy,dra genotype is identical to the thymidine-sensitive class (31, 62, 278). From Fig. 20, which shows the pathways involved in nucleoside catabolism in E. coli and S. typhimurium (184, 262), it appears that the tir mutations in both cases interfere with the catabolism of deoxyribose phosphates. The inability of wild-type cells to incorporate exogenous thymine is due to a lack of deoxyribosyl groups. Presumably any endogenously generated deoxyribose-l-P is immediately catabolized irreversibly via the glycolytic pathway (Fig. 20). Introduction of a thyA mutation, however, results in cells that are capable of

High

323

incorporating thymine, although only if high concentrations of this compound are present. The high thymine concentration is necessary to enable thymidine phosphorylase to compete effectively with phosphodeoxyribomutase for the available deoxyribose-l-P. A second mutation affecting either drm or dra would be expected to result in increased pools of deoxyribose phosphates, which in turn would promote the conversion of thymine to thymidine. The fact that thyA mutants, in contrast to wild-type cells, are able to utilize exogenous thymine indicates that the mere introduction of a thymine requirement results in an increased availability of deoxyribose-l-P, presumably as a result of an increased catabolism of deoxyribonucleotides. Two observations support this hypothesis. (i) Thymine starvation of thy mutants results in an increased catabolism of deoxyribonucleotides (60; Neuhard and Thomassen, in preparation). (ii) The dUMP pool of thy mutants, when grown with adequate amounts of exogenous thymine, is at least 20-fold higher than in wild-type cells (A. Munch-Petersen, Eur. J. Biochem., in press). The thymidine sensitivity of dra mutants has been correlated with an endogenous accumulation of deoxyribose-5-P in the presence of exogenous deoxyribonucleosides. However, the site of inhibition by deoxyribose-5-P has not yet been determined (31, 251).

Genetics and Regulation of Deoxyribonucleoside Catabolism The close connection existing between thymine metabolism and deoxyribonucleoside catabolism has led to careful studies of the four enzymes involved (Fig. 20). Two of them, thymidine phosphorylase and deoxyriboaldolase (181, 182), are specific for deoxyribosyl groups, whereas purine nucleoside phosphorylase (211) and phosphodeoxyribomutase (K. Hammer-Jespersen, personal communication) may act on both ribosyl and deoxyribosyl compounds. All four enzymes have been purified extensively and their properties have been studied (Chart A of Appendix). They are all localized close to the UR ,§41t T^, PERA3 cell surface, as indicated by the fact that they are released from the cells by osmotic shock (29, 206, 277, 278). DAA1 Recently, detailed genetic studies have shown U gre S-Fe that they map very close together and close to the thr locus (Fig. 8). In E. coli, the order of the genes is hsp, -dra, -tpp, -drm, -pup, -serB-thr A o10509~_ ~ r~huss5-(15 (10). The four enzymes are all induced by deoxyFIG. 20. Nucleoside catabolism in Escherichia coli ribonucleosides. Table 15 shows the induction and Salmonella typhimurium. ETK -(&

324

O'DONOVAN AND NEUHARD

BACTrERIOL. REV.

TABLE 15. Induction of deoxyriboaldolase, thymidine phosphorylase, phosphodeoxyribomutase, and purine nucleoside phosphorylase in different mutants of Escherichia coli K-12a strain

Inducer

Deoxyriboaldolase

Thymidine phosphorylase

Phosphodeoxyribomutase

Purine phosphorylase

Wild type

Thymidine Deoxyadenosine Adenosine

+

+

+

+

+

+

+

+

_

-

+

+

+ + _

+ + +

+ + +

+ +

+ +

Thymidine

dra

Deoxyadenosine Adenosine

tpp-

Thymidine

Deoxyadenosine Adenosine

drm

pUP-

- _ + -

Thymidine

-

Deoxyadenosine Adenosine

-

_

-

_

+ _

+ _

+

_

+

Thymidine

Deoxyadenosine Adenosine

_

_

+

a Data are irom K. i-ammer-Jespersen, P. Ny (gaard, andi M. viistrup (personal communications). Key: +, induction; - no, induction Cr-

1 X --:1ALAr_~vfB T.

pattern of the deoxyribonucleoside-catabolizing enzymes in different mutants of E. coli. It shows (i) that induction by thymidine depends on the presence of thymidine phosphorylase and phosphodeoxyribomutase, (ii) that the induction by deoxyadenosine requires a functional purine nucleoside phosphorylase (pup) and phosphodeoxyribomutase, and (iii) that a dra mutant is induced by both compounds. Thus, it may be concluded that the actual inducer of the four enzymes is deoxyribose-5-P. (9, 27, 30, 31, 63, 181, 182, 183, 251, 277, 278). Although these data are suggestive of one operon governing all four loci, two findings included in Table 15 argue against this. (i) In wild-type cells, adenosine induces only two of the enzymes, namely purine nucleoside phosphorylase and phosphodeoxyribomutase (277). (ii) Adenosine but not deoxyadenosine induces phosphodeoxyribomutase in a pup mutant (K. Hammer-Jespersen, personal communication). Recently, mutants constitutive for the synthesis of the four deoxyribonucleoside-catabolizing enzymes have been isolated in both E. coil (K. Hammer-Jespersen and A. MunchPetersen, personal communication) and S. typhimurium (P. Hoffee and B. C. Robertson, Abstract Federation meeting, Atlantic City, 1970). Genetic studies have revealed that in E. coli the locus for constitutivity is not closely linked

--

WsB......... :

to the structural genes for the four deoxyribonucleoside-catabolizing enzymes. (K. HammerJespersen, and R. H. Pritchard, personal communications).

PYRIMIDINE ANALOGUES In this section we will deal only with those analogues which have been found to be toxic for the growth of bacterial cells. We consider that it would be of no use to the reader to list an exhaustive number of known analogues. Instead, we will attempt to establish some general statements which are true for most analogues. After the treatment, we will discuss a few groups of analogues and point out the salient features of these in reference to the general statements.

General Principles (i) The analogue must be converted to the nucleotide level for expression of toxicity. (ii) Two general types of nucleotide analogue toxicity can be entertained: inhibition of one or more of the enzymes involved in pyrimidine metabolism, e.g., 6-aza-UMP inhibits OMP decarboxylase (pyrF) and incorporation of the analogue into nucleic acids after conversion to the triphosphate, e.g., 5-FUTP incorporation into RNA. (iii) Since the analogues are generally metabolized in the same manner, by the same

VOL. 34, 1970

PYRIMIDINE METABOLISM IN MICROORGANISMS

enzymes, as the natural bases and nucleosides, it is not surprising that addition of the natural compounds readily reverses the analogue toxicity, simply by competition with the analogue for conversion to the nucleotide level (compare the first principle). (iv) It is of vital importance when working with analogues to understand the interconversions of the natural bases and nucleosides (see Fig. 17) since many of the analogues are per se not inhibitory but become inhibitory only after being metabolized. For example, in E. coli and S. typhimurium, 5-fluorocytosine is toxic only after conversion to 5-fluorouracil (see Fig. 17 and Table 13). Thus one would anticipate the isolation of two unique classes of mutants resistant to pyrimidine analogues. The first class consists of mutants which are unable to convert the analogue to the nucleotide level (to the true inhibitor). Among the examples of this class are the following: 5-fluorouracilresistant mutants lacking UMP pyrophosphorylase, and 5-fluorocytosine-resistant mutants lacking cytosine deaminase; the latter are unable to convert 5-fluorocytosine to 5-fluorouracil, but are still sensitive to 5-fluorouracil. The second class consists of pyrimidine overproducers which are resistant by virtue of excreting natural pyrimidine bases and nucleosides which will compete with the analogues. For example, among 5-fluorouridine-resistant cells in S. typhimurium have been found pyrimidine regulatory mutants (O'Donovan and Gerhart, in preparation). 5-Fluoro Analogues Mutants unable to convert 5-fluorouracil to the nucleotide, the true inhibitor, are 5-fluorouracil resistant and lack the enzyme UMP pyrophosphorylase (68). Indeed, in micoorganisms this enzyme is instrumental in effecting the expression of toxicity of several uracil analogues (5-azauracil, 6-azauracil, 5-fluorouracil) by converting them to the corresponding nucleotides (85, 171). (For higher organisms, however, see reference

325

pathway (O'Donovan and Gerhart, in preparation) are simultaneously resistant to both 5-fluorouracil and 5-fluorouridine. However, since these strains have uridine phosphorylase, the presence of 5-fluorouridine alone gives the same effect. 5-Fluorouracil was found to be anabolized to a wide variety of compounds. (For review, see reference 67.) 5-Fluorodeoxyuridine exerts its toxic effect by being converted to 5-fluoro-dUMP by thymidine kinase. 5-Fluoro-dUMP strongly inhibits thymidylate synthetase. 5-Fluorodeoxyuridine is also a substrate for thymidine phosphorylase, thereby being catabolized to 5-fluorouracil. Accordingly, mutants which lack thymidine kinase are not resistant to the analogue. It probably requires a double mutant lacking both thymidine kinase and thymidine phosphorylase for 5-fluorodeoxyuridine resistance. Cytosine and deoxycytidine as such are uniquely inert in E. coli and S. typhimurium. They must first be converted to uracil derivatives by deaminases before being further metabolized (see Fig. 17). Accordingly, mutants resistant to 5-fluorocytosine or 5-fluorodeoxycytidine lack the enzymes cytosine deaminase and cytidine deaminase, respectively. These mutants are still sensitive to the corresponding 5-fluorouracil derivatives. The existence of cytidine-requiring mutants in E. coli (Neuhard, unpublished data) and in S. typhimurium (282, 283) indicates that these cells possess a functional cytidine kinase (212). This explains the toxicity of 5-fluorocytidine in mutants lacking cytidine deaminase (cdd; O'Donovan, unpublished data). Moreover, it indicates that not only 5-fluoro-UTP but also a 5-fluorocytidine nucleotide is toxic to the cells, although the exact nature and mode of action of such a compound is not known at present. Aza Analogues

Most of the work which elucidated the role of the aza-analogues has been done with micro329.) organisms (352). As always, when one tries to 5-Fluorouracil-resistant mutants are sensitive define the function of a compound the choice of to 5-fluorouridine, indicating the presence of a organism is very important. Thus, in the case of kinase which can convert 5-fluorouridine to 6-azathymine, most studies have been done with 5-fluoro-UMP. Thus, a 5-fluorouracil-resistant, microorganisms capable of using exogenous 5-fluorouridine-resistant double mutant would thymine. Certain strains of Lactobacilli prove to be expected to lack both UMP pyrophosphorylase be ideal for this task. Growth of L. leichmannii and uridine kinase. Such a mutant, (upp,udk) and Streptococcus faecalis are inhibited by 6-azahas been found (see Table 13). thymine (313) and even more effectively by the nuPyrimidine overproducers will also be re- cleoside 6-azathymidine (311). 6-Azathymidine is sistant to 5-fluorouracil because of the successful an effective competitive inhibitor of trans-Ncompetition of the natural substrate with the deoxyribosylase in L. leichmannii. The analogue analogue for the enzyme UMP pyrophosphoryl- inhibits transfer of deoxyribose to adenine from ase. In fact, constitutive mutants of the pyrimidine either thymidine, deoxyuridine, or deoxycyti-

1-

0

00

S:

%D

00

0

+

+

+

+

*3

0 U

34

'0 %D

2

C)

2

0

C14

f-iX2 1 F-0

1-4

U

CU-'

u.4>0

*0.14-

92

144

_4

Cd

z

z

048

_

~~~02

-

Co0

k.

N 00%

0 0

0>~~~~~~~~~~C 0'

,0

-4,~

>b

V

00

cd

f-v

W

UI

tr

f~ 0 CIS

O

0c0 oe

(O~~~~e

En ~~~~~'

0~ ~ ~ ~ ~ ~ ~ ~ ~D

0)

0

+

.0

._0

+

T

+

1

I 0

:)

UCU CtQ +

C) CU

C+, + +

N

++:

Cd- Cd 14C)14C, W ci a.

e z Go

E

tv ! +

326

uC

.0~~~~~~

~ ~ ~ ~

~

. ~~~~~~~~~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~~

Z~~~~~~~

.0. U

cis to

~~~~~4

~ ~ 41z 0

CA O

0

.0~

0

IA~~~~~

04 '0~~~~~~~~~~~~~~~~~~~~~~~~

+

0 I~~~~~~~~~~~~~~~~~~~~~~~~~0i

-"~~~~~~~~~~~~~-.

~

~

~

~

Z-0~ ~ ~ ~ ~

&~~~~~~~~~~~~~~~ 04~~~~~~~~~~~~~~~~ I +

0~~~~~~~~~~N ~ ~

0

0

00

4)~~~~~~~~~~~~~~0 0

4)

d

S

-

1

0~~~

04

03~~~~~~t-

02

W

0

N~~~~~~~~~c

40~~~~~

0

2

0 4) 04

"-~ ~ ~ ~ ~ .

N

' 6+ +

~~~~~~

o

NO

'*~~~~~~~~~~~~~~~~~=+J .

.

*+ +* : 042 04% ~00 .0 a< ~~ ~~~~0040 0 'o v

0

00~~~~~

+)

+ +

Ca2

S04

327

x>. .04

O

4)

,0~~~~~~~~~~~~~~~~~~~~~~04) N~~~~~'~)

4)

02

a

4F

B | ^

+~~~~~~

>

-~~~~~~~~~~~F 0-4 ~ ~ ~~

~

~

~

~

~

~

~

~~~4

.0 A''~~~~~~~~~~~~~~

It

00 ~ ~

~

~

~

~

N

S~~~~~~~~~~~~~~~~~~~~~~~~ S ~~~~~~~~~~~~~~~~~~~~~~~~~~~t

X

X.

0

*I

:~~

10

.4 4)).

o

i

z IL

o"U1o

0'0~~~Z-

in"U10

bi)4) g.y UU++

.

~

328

..

3t -0

+ +

+

+

S_~~~~~~~~~~~ .0

0

0)

0)~~~~~4)

N

4

00

S

it

.t

'

00 ~~~~~C0

+~~~~ 0

+

-o~t °; Ad N ~ ~

Co;XA0

Ao

c

_

0

AO

~ _e _O~ o U

U

0000*

N N

0r-

"~~~~~~~~

0

V

-

+g

o

8

-t+

91

Nif

-

0.~~~~~~~0)0 N -s N.t

.~~~~~5

329

00 M 4) u

6.41

'-0

oe X^ _M

-Oo

"

ff0

F

00

0

-1)4) 04

c3

, r~

-

I

'~~~~~~~~00 ~ ~ K'-

r^

m n r1C)r

t

M,-

.

r'~_ 0~~~~~~~~0 _ 0 r~

_

~

W t~ _ en0rI

tt

-

~

~

0 4)-Z

esi

-0

-

.

00

rmi6

00

| a)°.E b

W:

E ca=

0E 0 AX

4*C I 'c rA 0 9L)

zWc=Q 26. 5.

LW

< aE a .

0

'a2t 0a 0 "0o Uj

=

eoa E E r~

~

,0CUC~a)~0

aEE

c

I

,CU

-0~~~~~~~~~~~ o

"0

cr~

~~~~~~~~~~~~~,q

CU~~~~~o (U*V, "0. ~

V

a)r

UC CUC~0

U

01

01

m-

.0

V

co

00

.

~000

0

~~~Q

4L

LQ.

Q.

0~~~~~~

-~~~~~~~~~~~~~ Ca 0

I~~~~~~~~~~~4

0

0~~~~~~~~~~~~~~~~~~~~~~~ ts~~~~~~~~~~ 0 ~ Ln ~

1

P,

k

a

wI

Cd U 0

I-b4 a) (

co

CAUa)

0 0 = 4,) o CZ

c

U'W

CU

CU

.)

a)

a)

V

V)

CV >b

.*

,

CUU'

>16

u

U

330

U

U

V-(

en m e

-4en^

N

-

000 '

%e W)N lq N ^

M0

'e i t'- %)Om ^eo

.0'0

.

^

-

6 -

N

^

t- t

r

-

e

t

t

e at v6

in

-

i

c

en

W-

t-

~~~0!

0

00

'Nf N

en e

kn00

I,. t-:

C4 6,o

t-

T-

-

e

04) 4)"0 C,)

,D"0.

4)U -.

.00

.2,o

4)

-

U 4)*. .. .

"0..

4: cr~

~~"

to00 00

0)4. 0~~~~~~~~

4)4.0 0

E E E

S

E E

XP^

00

v

oNv

_

_

._ *

E:

*E0 E

E

.EC:

X

a, a

Z

0.

P N

0

I

~

~

E E

. @

00

00

I ~~~~~~~

~

X

. I

*~~~~~~~L.

X

. I

t)

a9EEEE a EE r~~~~~~~h%~~~~. a =E .

0

Oz

"0

0

0-

0

00

t-

00

X

^

0 t)0_3 C,=

.00

_I _

00

la

0)

0 ,0

4)

U)

0 ._

4

en

0

;

r~~~~~~ r . K&" - *.- E E a^Q . ~ E.rXr ~ r . F

00

4)

a

U)4-00

.4

"0.4

ci

S

00

~

IX

o

6o

~

~

~~

SW

O->

4),tO >-. 2

331

x0~

1~0 04) 0

rg

:3.

w :

>

an

Ci

av

^en

O

0 04

-

.'-5

cN

-

0 00

r

^

Zt

i

1

,

QQo

000

Z

.04 ,90.-

0d

"cd +

+ c

o

0

(A

o)

0 a

C oi

o

oS

o a

a

04

U1~

4)20 HH Z ~~~

.0

N

H

'A ~~~~~ 'AU

HZ

U:

0~~~

eq"

ci 00

E.S E ~

E

E

ci

'4

t 0w u

w

4)~~~~C4C:

P

-4 I.Q

4)

4)

Cd

0

Cd

4)

>b

o

co

04

._

>4

a

oA

v

0

4.)

0"0

.0

>4 v

osIE>6

a=H

E .,

A

Ez

332

6.004

4

0

)r

o1:04 L4

"0

VOL. 34, 1970

PYRIMIDINE METABOLISM IN MICROORGANISMS

dine, but does not itself serve as a deoxyribose donor for the enzyme (67, 263). Since 6-azathymidine also inhibits organisms which lack the enzyme trans-N-deoxyribosylase, this is obviously not the primary site of action of 6-azathymidine (67). Moreover, it has been shown that the analogue is incorporated into the DNA of S. faecalis, presumably in place of thymine (312). 6-Azauracil, 5-azauracil, and their nucleoside derivatives are converted to the corresponding aza-UMP compound by extracts of E. coli (84, 86, 171, 173, 175, 301, 354, 355), and 6-azauracil-resistant mutants are unable to convert uracil (6-azauracil) to UMP (6-aza-UMP) (17). In contrast to uridine, 6-azauridine was shown to be not readily degraded by such extracts (165). In the case of the 6-aza compounds, the true inhibitor, 6-aza-UMP, was found to be a competitive inhibitor of a partially purified OMP decarboxylase from yeast (173). Treatment in vivo with 6-azauracil results in OMP accumulation, which inhibits OMP pyrophosphorylase and results in orotate excretion into the medium. 6-Aza-UMP can be further anabolized to the triphosphate by bacteria, making possible other antimetabolic activities (67). 5-Azaorotate competes with orotate, the natural substrate, for the enzyme OMP pyrophosphorylase (89, 186). Recently, it has been shown that 5-azaorotate can be converted to 5-aza-UMP (89, 90), and one could anticipate its inhibitory effect on OMP-decarboxyiase (compare 6-aza-UMP in reference 91). As pointed out in General Principles, an organism may be resistant to an analogue by being unable to convert it to the nucleotide. This usually results from a loss of an enzyme that metabolizes the natural substrate. However, resistance may also be caused by the inability of the cell to take up exogenous pyrimidines Mutants of this class, lacking specific permeases, have been characterized in both E. coli and Saccharomyces. Cytosine Arabinoside Another pyrimidine analogue which has been studied extensively is cytosine arabinoside [for review see Cohen (95)]. Cytosine arabinoside is an analogue in the pentose moiety and is unique in that its toxicity is dependent on an intact cytosine moiety. The analogue is readily deaminated in the enterobacteria by cytidine (deoxycytidine) deaminase, which deaminates it to the corresponding nontoxic uracil derivative. Its toxicity in mammalian cells has been shown by Furth and Cohen (143) to be due to its conversion to the nucleoside triphosphate (ara-

333

CTP). They have shown that ara-CTP inhibits DNA polymerase and the inhibition is competitive with dCTP. On the other hand DNA polymerase of E. coli is not inhibited by ara-CTP. Cytosine arabinoside is converted by mammalian cells to the nucleotide level by deoxycytidine kinase (95). Since there is no deoxycytidine kinase in E. coli and S. typhimurium (211, 282), the conversion of cytosine arabinoside to the nucleotide level, if occurring, is presumably carried out by cytidine kinase. This may explain the fairly high concentrations of the analogue necessary to observe any inhibition in these organisms. Cytosine arabinoside is a feedback inhibitor of aspartate transcarbamylase of E. coli and S. typhimurium and seems to be the best available false feedback inhibitor of pyrimidine biosynthesis in specially constructed strains of E. coli and S. typhimurium (O'Donovan, unpublished data). Emerging from the analogue study of pyrimidine bases and nucleosides in E. coli and S. typhimurium come the following observations that may be added to the general statements made in the introduction to this section. (i) Cytosine bases and deoxyribonucleosides are singularly inert unless they are first deaminated to the corresponding uracil compounds. (ii) Cytidine analogues, on the other hand, may either be deaminated by cytidine deaminase or phosphorylated by cytidine kinase. The nucleotide derivatives, obtained by direct phosphorylation, may have a specific effect on metabolism, different from that of the corresponding uridine nucleotide analogue. APPENDIX In the two accompanying charts, we have included for summary purposes all of the key enzymes treated in the review. Chart A contains the known properties of the enzymes which are listed in alphabetical order. Chart B contains the known genetics of the same enzymes where relevant. The following abbreviations are used in charts A and B. E.C. number refers to Enzyme Commission Classification System. P indicates pure enzyme. PP indicates partially purified enzyme. CE indicates crude extract. An asterisk (*) indicates a new genotype proposed in the present review. 5-FUR and 5-FUS indicate resistance and sensitivity, respectively, to 5-fluorouracil. Numbers in parentheses refer to the bibliography list of the review. ACKNOWLEDGMENTS We express our thanks and appreciation to those who so generously contributed material in advance of publication, to our

334

O'DONOVAN AND NEUHARD

friends who read and criticized different sections of the manuscript, John Gerhart for his ever-helpful discussions, and to our respective collaborators at Texas A&M and Copenhagen for many hours of discussion. We are most grateful to John Womack who completely organized the bibliography. One of us (J. N.) acknowledges a Travel Grant from Rask 0rsted Fondet, Copenhagen. Some of this work was carried out during the tenure of a Dernham Fellowship (J-104), California Division, American Cancer Society, in the Department of Molecular Biology, University of California at Berkeley. Current research is supported by grants from the Robert A. Welch Foundation, Houston, Tex., and from the Research Council of Texas A&M to

University. LITERATURE CITED 1. Abbott, M. T., T. A. Dragila, and R. P. McCroskey. 1968. The formation of 5-formyluracil by cell-free preparations from Neurospora crassa. Biochim. Biophys. Acta 169:1-6. 2. Abbott, M. T., E. K. Schandl, R. F. Lee, T. S. Parker, and R. J. Midgett. 1967. Cofactor requirement of thymine 7-hydroxylase. Biochim. Biophys. Acta. 132:525-528. 3. Abd-El-AI, A., and J. L. Ingraham. 1969. Control of carbamyl phosphate synthesis in Salmonella typhimurium. J. Biol. Chem. 244:4033-4038. 4. Abd-EI-Al, A., and J. L. Ingraham. 1969. Cold sensitivity and other phenotypes resulting from mutation in pyrA gene. J. Biol. Chem. 244:4039-4045. 5. Abd-El-Al, A., D. P. Kessler, and J. L. Ingraham. 1969. Arginine-auxotrophic phenotype resulting from a mutation in the pyrA gene of Escherichia coll B/r. J. Bacteriol.

97:466-468. 6. Abeles, R. H., and W. S. Beck. 1967. The mechanism of action of cobamide coenzyme in the ribonucleotide reductase reaction. J. Biol. Chem. 242:3589-3593. 7. Abrams, R. 1965. Cytidine5Y-triphosphate as the precursor of deoxycytidylate in Lactobacillus lechmannit. J. Biol. Chem. 240:3697. 8. Abrams, R., and S. Duraiswami. 1965. Deoxycytidylate formation from cytidylate without glycosidic cleavage in Lactobacillus leichmannilextracts containing vitamin Bi2 coenzyme. Biochem. Biophys. Res. Commun. 18:409-413. 9. Ahmad, S. I., P. T. Barth, and R. H. Pritchard. 1968. Properties of a mutant of Escherichia colU lacking purine nucleoside phosphorylase. Biochim. Biophys. Acta 161:581583. 10. Ahmad, S.I., and R. H. Pritchard. 1969. A map of four genes specifying enzymes involved in catabolism of nucleosides and deoxynucleosides in Escherichia coli. Mol. Gen. Genet. 104:351-359. 11. Albrecht, A. M., F. K. Pearce, and D. J. Hutchison. 1966. Folate coenzymes in aminopterin-sensitive and -resistant strains of Streptococcus faecalls. Enzymatic formation and metabolic function. J. Biol. Chem. 241:1036-1042. 12. Aleman, V., and P. Handler. 1967. Dihydroorotate dehydrogenase. I. General properties. J. Biol. Chem. 242:40874096. 13. Alikhanian, S. I., T. S. Ijina, E. S. Kaliaeva, S. V. Kameneva, and V. V. Sukhodolec. 1966. A genetical study of thymineless mutants of E. coll K12. Genet. Res. 8:83-100.

14.

B. N., and B. J. Garry. 1959. Coordinate repression Ames, of the synthesis of four histidine biosynthetic enzymes by histidine. Proc. Nat. Acad. Sci. U.S.A. 45:1453-1461.

Ames, B. N., B. Garry, and L. A. Herzenberg. 1960. Thein genetic control of the enzymes of histidine biosynthesis Salmonella typhimurium. J. Gen. Microbiol. 22:369-378. 16. Anderson, E. P., and R. W. Brockman. 1964. Feedback inhibition of uridine kinase by cytidine triphosphate and uridine triphosphate. Biochim. Biophys. Acta 91:380-386. 17. Anderson, E. P., and L. W. Law. 1960. Biochemistry of

15.

cancer. Annu. Rev. Biochem. 29:577-608. 18. Anderson, P. M., and S. V. Marvin. 1968. Effect of ornithine,

IMP, and UMP on carbamyl phosphate synthetase from

BACTERIOL. REV.

Etscherichia coi. Biochem. Biophys. Res. Commun. 32:928-

934.

19. Anderson, P. M., and S. V. Marvin. 1970. Effect of allosteric effectors and adenosine triphosphate on the aggregation and rate of inhibition by N-ethylmaleimide of carbamyl

phosphate synthetase of Escherichia coi. Biochemistry 9:171-178. 20. Anderson, P. M., and A. Meister. 1965. Evidence for activated form of carbon dioxide in the reaction catalyzed by Escherichia colt carbamyl phosphate synthetase. Biochemistry 4:2803-2809. Escher21. Anderson, P. M., and A. Meister. 1966. Control of Ichia coli carbamyl phosphate synthetase by purine and pyrimidine nucleotides. Biochemistry 5:3164--3169. 22. Appel, S. H. 1968. Purification and kinetic properties of brain orotidine-5'-phosphate decarboxylase. J. Biol. Chem. 243:3924-3929. 23. Arvidson, H., N. A. Eliasson, E. Hammarsten, P. Reichard, H. V. Ubisch, and S. Bergstrom. 1949. Orotic acid as a precursor of pyrimidines in the rat. J. Biol. Chem. 179:169173.

24. Bagatell, F. K., E. M. Wright, and H. Z. Sable. 1959.Biosynthesis of ribose and deoxyribose inEscherichia coli. J. Biol. Chem. 234:1369-1373. 25. Barner, H. D., and S. S. Cohen. 1954. The induction of thymine synthesis by T2 infection of a thymine requiring mutant ofEscherichia coli. J. Bacteriol. 68:80-88. 26. Barner, H. D., and S. S. Cohen. 1959. Virus-induced acquisition of metabolic function. IV. Thymidylate synthetase in thymine-requiringEscherichia colt infected by T2 and

T5 bacteriophages.

J. Biol. Chem. 234:2987-2991.

27. Barth, P. T., I. R. Beacham,S. 1. Ahmad, and R. H. Pritchard. 1968. The inducer of the deoxynucleoside phosphorylases and deoxyriboaldolase inEscherichia colt. Bio-

chim. Biophys. Acta 161:554-557. 28. Batterham, T. J., R. K. Ghambeer, R. L. Blakley, and C. Brownson. 1967. Cobamides and ribonucleotide reduction. IV. Stereochemistry of hydrogen transfer to the deoxyribonucleotide. Biochemistry 6:1203-1208. 29. Beacham,I. R. 1969. A new assay for phosphodeoxyribomutase: surface localization of the enzyme. Biochim. Biophys. Acta 191:158-161. 30. Beacham, I. R., P. T. Barth, and R. H. Pritchard. 1968. Constitutivity of thymidine phosphorylase in deoxyribodependence on thymine requirealdolase negative strains: Biochim. Biophys. Acta 166:589ment and concentration. 592.

31. Beacham, I. R., A. Eisenstark, P. T. Barth, and R. H. Pritchard. 1968. Deoxynucleoside-sensitive mutants of Salmonella typhimurium. Mol. Gen. Genet. 102:112-127. 32. Beck, W. S. 1967. Regulation of cobamide-dependent ribonucleotide reductase by allosteric effectors and divalent cations. J. Biol. Chem. 242:3148-3158. 33. Beck, W. S., M. Goulian, A. Larsson, and P. Reichard. 1966.

Hydrogen donor specificity

of cobamide-dependent ribo-

nucleotide reductase and allosteric regulation of substrate specificity. J. Biol. Chem. 241:2177-2179. 34. Beck, W. S., and J. Hardy. 1965. Requirement of ribonucleotide reductase for cobamide coenzyme, a product of ribosomal activity. Proc. Nat. Acad. Sci. U.S.A. 54:286-293. 35. Beck, W. S., and M. Levin. 1963. Purification, kinetics and repression control of bacterial trans-N-deoxyribosylase. J. Biol. Chem. 238:702-709. 36. Beckwith, J. R., A. B. Pardee, R. Austrian, and F. Jacob. 1962. Coordination of the synthesis of the enzymes in the

pyrimidine pathway of E. coli. J. Mol. Biol. 5:618-634. 37. Belser, W. L. 1961. Uracil biosynthesis in Serratia marinorubra. Biochem. Biophys. Res. Commun. 4:56-60. 38. Berg, P., and W. K. Joklik. 1953. Transphosphorylation between nucleoside polyphosphates. Nature 172:1008-1009. 39. Berg, P., and W. K. Joklik. 1954. Enzymatic phosphorylation of nucleoside diphosphates. J. Biol. Chem. 210:657-672.

VOL. 34, 1970

PYRIMIDINE METABOLISM IN MICROORGANISMS

40. Berglund, 0. 1969. Identification of a thioredoxin induced by bacteriophage T4. J. Biol. Chem. 244:6306-6308. 41. Berglund,O., 0. Karlstrom, and P. Reichard. 1969. A new ribonucleotide reductase system after infection with phage T4. Proc. Nat. Acad. Sci. U.S.A. 62:829-835. 42. Bernofsky, C., and M. F. Utter. 1967. Secondary activation effects of mitochondrial isocitrate dehydrogenases from yeast. Biochim. Biophys. Acta 132:244-255. 43. Bertani, L. E., A. Haggmark, and P. Reichard. 1963. Enzymatic synthesis of deoxyribonucleotides. I. Formation and interconversion of deoxyuridine phosphates. J. Biol. Chem. 238:3407-3413. 44. Bertino, J. B., and K. A. Stacey. 1966. A suggested mechanism for the selective procedure for isolating thyminerequiring mutants of Escherichia coli. Biochem. J. 101:3233c. 45. Bethell, M. R., and M. E Tones. 1969. Molecular size and feedback-regulation characteristics of bacterial aspartate transcarbamylases. Arch. Biochem. Biophys. 134:352-365. 46. Bethell, M. R., K. E. Smith, J. S. White, and M. E. Jones. 1968. Carbamyl phosphate: an allosteric substrate for aspartate transcarbamylase of Fscherichia coli. Proc. Nat. Acad. Sci. U.S.A. 60:1442-1449. 47. Biswas, C., J. Hardy, and W. S. Beck. Release of repressor control of ribonucleotide reductase by thymine starvation. J. Biol. Chem. 240:3631-3639. 48. Blakley, R. L. 1963. The biosynthesis of thymidylic acid. IV. Further studies on thymidylate synthetase. J. Biol. Chem.

238:2113-2118. 49. Blakley, R. L. 1965. Cobamides and ribonucleotide reduction. I. Cobamide stimulation of ribonucleotide reduction in extracts of Lactobacillus leichmannii. J. Biol. Chem. 240:2173-2179. 50. Blakley, R. L. 1966. B12-dependent synthesis of deoxyribonucleotides. Fed. Proc. 25:1633-1638.

51. Blakley, R. L., and H. A. Barker. 1964. Cobamide stimulation of the reduction of ribotides to deoxyribotides in

Lactobacillus leichmannii. Biochem. Biophys.

Res. Com-

16:391-397. 52. Blakley, R. L., R. K. Ghambeer, T. J. Batterham, and C. Brownson. 1966. Studies with hydrogen isotopes on the mechanisms of action of cobamide-dependent ribonucleotide reductase. Biochem. Biophys. Res. Commun. 24:418426. 53. Blakley, R. L., R. K. Ghambeer, P. F. Nixon, and E. Vitols. 1965. The cobamide-dependent ribonucleoside triphosphate reductase of Lactobacilli. Biochem. Biophys. Res. Commun. 20:439-445. 54. Blakley, R. L., and B. McDougall. 1962. The biosynthesis of thymidylic acid. III. Purification of thymidylate synthetase and its spectrophotometric assay. J. Biol. Chem. 237:812818. 55. Blakley, R. L., B. V. Ramasastri, and B. M. McDougall 1963. The biosynthesis of thymidylic acid. V. Hydrogen isotope studies with dihydrofolic reductase and thymidylate synthetase. J. Biol. Chem. 238:3075-3079. 56. Blakley, R. L., and E. Vitols. 1968. The control of nucleotide biosynthesis. Annu. Rev. Biochem. 37:201-224. 57. Bolton, E. 1954. Biosynthesis of nucleic acid in Escherichia mun.

coli. Proc. Nat. Acad. Sci. U.S.A. 40:764-772. 58. Bolton, E. T., and A. M. Reynard. 1954. Utilization of purine and pyrimidine compounds in nucleic acid synthesis by Escherichia coli. Biochim. Biophys. Acta 13:381-385. 59. Boyce, R. P., and R. B. Setlow. 1962. A simple method of increasing the incorporation of thymidine into the deoxyribonucleic acid of Escherichia coli. Biochim. Biophys. Acta 61:618-620. 60. Breitman, T. R., and R. M. Bradford. 1964. The induction of thymidine phosphorylase and excretion of deoxyribose during thymine starvation. Biochem. Biophys. Res. Commun. 17:786-791. 61. Breitman, T. R., and R. M. Bradford. 1967. Metabolism of

thymineless

mutants

of Escherichia

coli.

I.

Absence of

62.

63.

64.

65.

66.

67.

335

thymidylate synthetase activity and growth characteristics of two sequential thymineless mutants. J. Bacteriol. 93:845-852. Breitman, T. R., and R. M. Bradford. 1967. The absence of deoxyriboaldolase activity in a thymineless mutant of Escherichia coli strain 15: a possible explanation for the low thymine requirement of some thymineless strains. Biochim. Biophys. Acta 138:217-220. Breitman, T. R., and R. M. Bradford. 1968. Inability of low thymine-requiring mutants of Escherichia coli lacking phosphodeoxyribomutase to be induced for deoxythymidine phosphorylase and deoxyriboaldolase. J. Bacteriol. 95:2434-2435. Breitman, T. R., R. M. Bradford, and W. D. Cannon, Jr. 1967. Use of exogenous deoxythymidylic acid to label the deoxyribonucleic acid of growing wild-type Escherichia coli. J. Bacteriol. 93:1471-1472. Brenner, S., and J. D. Smith. 1959. Induction of mutations in the deoxyribonucleic acid of phage T2 synthesized in the presence of chloramphenicol. Virology 8:124-125. Bresnick, E., S. Singer, and G. H. Hitchings. 1960. Mechanism of action of 6-azacytosine in bacteria. Biochim. Biophys. Acta 37:251-257. Brockman, R. W., and E. P. Anderson. 1963. Pyrimidine analogues, p. 239-285. In R. M. Hochster and J. H. Quastel (ed.), Metabolic inhibitors. Academic Press Inc., New York.

68. Brockman, R. W., J. M. Davis, and P. Stutts. 1960. Metabolism of uracil and 5-fluorouracil by drug-sensitive and by drug-resistant bacteria. Biochim. Biophys. Acta 40:22-32. 69. Brooke, M. S., D. Ushiba, and B. Magasanik. 1954. Some factors affecting the excretion of orotic acid by mutants of Aerobacter aerogenes. J. Bacteriol. 68:534-540. 70. Brown, N. C., Z. N. Canellakis, B. Lundin, P. Reichard, and L. Thelander. 1969. Ribonucleoside diphosphate reductase. Purification of the two subunits, proteins BI and B2. Eur. J. Biochem. 9:561-573. 71. Brown, N. C., R. Elliasson, P. Reichard, and L. Thelander. 1969. Spectrum and iron content of protein B2 from ribonucleoside diphosphate reductase. Eur. J. Biochem. 9:512518. 72. Brown, N. C., and P. Reichard. 1969. Ribonucleoside diphosphate reductase. Formation of active and inactive complexes of proteins BI and B2. J. Mol. Biol. 46:25-38. 73. Brown, N. C., and P. Reichard. 1969. Role of effector binding in allosteric control of ribonucleoside diphosphate reductase. J. Mol. Biol. 46:39-55. 74. Budman, D. R., and A. B. Pardee. 1967. Thymidine and thymine incorporation into deoxyribonucleic acid: inhibition and repression by uridine of thymidine phosphorylase of Escherichia coli. J. Bacteriol. 94:1546-1550. 75. Cannon, W. D., and T. R. Breitman. 1967. Control of deoxynucleotide biosynthesis in Escherichia coli. I. Decrease of pyrimidine deoxynucleotide biosynthesis in vivo in the presence of deoxythymidylate. Biochemistry 6:810816. 76. Cannon, W. D., and T. R. Breitman. 1968. Control of deoxynucleotide biosynthesis in Escherichia coli. II. Effect of deoxythymidylate on the biosynthesis of both deoxynucleotides and ribonucleotide reductase. Arch. Biochem. Biophys. 127:534-542. 77. Caroline, D. F. 1969. Pyrimidine synthesis in Neurospora crassa: gene-enzyme relationships. J. Bacteriol. 100:13711377. 78. Caroline, D. F., and R. H. Davis. 1969. Pyrimidine synthesis in Neurospora crassa: regulation of enzyme activities. J. Bacteriol. 100:1378-1384. 79. Carter, E. E. 1951. Partial purification of a non-phosphorylytic uridine nucleosidase from yeast. J. Amer. Chem. Soc. 73:1508-1510. 80. Chakraborty, K. P., and R. B. Hurlbert. 1961. Role of glutamnine in the biosynthesis of cytidine nucleotides in

Escherichia coli. Biochim. Biophys. Acta 47:607609.

336

BACrERIOL. REV.

O'DONOVAN AND NEUHARD

81. Changeux, J.-P., J. C. Gerhart, and H. K. Schachman. 1968. Allosteric interactions in aspartate transcarbamylase. I. Binding of specific ligands to the native enzyme and its isolated subunits. Biochemistry 7:531-537. 82. Charles, H. P. 1962. Response of Neurospora mutants to carbon dioxide. Nature 195:359-360. 83. Chattaway, F. W. 1944. Growth stimulation of L. casel E by pyrimidines. Nature 153:250-251. 84 Cihak, A., J. Skoda, and F. Sorm. 1964. Formation of 5-azauridine, ribosyl N-formalbiuret, ribosyl biuret, and their 5'-phosphates in Escherichia coll culture from 5'-azauracil. Collect. Cesk. Chem. Commun. 29:300-308. 85. Cihak, A., J. Skoda, and F. Sorm. 1964. Ribosylation and phosphoribosylation of 5-azauracil-2-4-14C in a cell-free extract of Escherichia colt. Collect. Cesk. Chem. Commun. 29:814-824. 86. Cihak, A., and F. Sorm. 1964. Inhibition by 5-azauracil of the uridine phosphorylase and deoxyuridine phosphorylase activities in a cell-free extract of mouse liver. Biochim. Biophys. Acta 80:672-674. 87. Cihak, A., and F. Sorm. 1965. Biochemical effects and metabolic transformations of 5-azacytidine in Escherichia coli. Collect. Cesk. Chem. Commun. 30:2091-2102. 88. Cihak, A., and F. Sorm. 1965. Inhibition of microbial cytosine deaminase by 5-azacytosine and 5-azauracil. Collect. Cesk. Chem. Commun. 30:2137-2140. 89. Cihak, A., and F. Sorm. 1965. Inhibitory effects of 5-azaorotate in Eacherichia coll. Collect. Cesk. Chem. Commun.

enzyme-dependent ribonucleotide reductase in Rhizobium species and the effects of cobalt deficiency on the activity of the enzyme. J. Bacteriol. 97:1460-1465. 104. Crawford, L. V. 1958. Thymine metabolism in strains of Escherichia coli. Biochim. Biophys. Acta 30:428-429. 105. Crawford, I., A. Kornberg, and E. S. Simms. 1957. Conversion of uracil and orotate to uridine 5'-phosphate by enzymes in Lactobacilli. J. Biol. Chem. 226:1093-1101. 106. Creasey, W., and R. E. Handschumacher. 1961. Purification

107.

108.

109.

110.

111. 112.

30:3513-3519. 90. Cihak, A., and F. Sorm. 1967. Metabolic transformations of 5-azaorotate: cause of marked inhibition of orotidine5'-phosphate decarboxylase. Biochim. Biophys. Acta 149:314-316. 91. Cihak, A., J. Vesely, and F. Sorm. 1968. Inhibition of pyrimidine biosynthesis by 5-azaorotate in mouse liver. Collect. Cesk. Chem. Commun. 33:1778-1787. 92. Cohen, G. N., J.-C. Patte, and P. Truffa-Bachi. 1965. Parallel modifications caused by mutations in two enzymes with the biosynthesis of threonine in Escherichia coli. Biochem. Biophys. Res. Commun. 19:546-550. 93. Cohen, S. S. 1953. Studies on controlling mechanisms in the metabolism of virus-infected bacteria. Cold Spring Harbor Symp. Quant. Biol. 18:221-235. 94. Cohen, S. S. 1963. On biochemical variability and innovation. Science 139:1017-1026. 95. Cohen, S. S. 1966. Introduction to the biochemistry of D-arabinosyl nucleosides. Progr. Nucl. Acid Res. Mol. Biol. 5:1-88. 96. Cohen, S. S., and H. D. Barner. 1954. Studies on unbalanced growth in Escherichla coli. Proc. Nat. Acad. Sci. U.S.A. 40:885-893. 97. Cohen, S. S., and H. D. Barner. 1957. The conversion of 5-methyldeoxycytidine to thymidine in vitro and In vivo. J. Biol. Chem. 226:631-642. 98. Cohen, S. S., J. G. Flaks, H. D. Barner, M. R. Loeb, and J. Lichtenstein. 1958. The mode of action of 5-fluorouracil and its derivatives. Proc. Nat. Acad. Sci. U.S.A. 44:10041012. 99. Collins, K. D., and G. R. Stark. 1969. Aspartate transcarbamylase. Studies of the catalytic subunit by ultraviolet difference spectroscopy. J. Biol. Chem. 244:18691877. 100. Colowick, S. P., and H. M. Kalckar. 1943. The role of myokinase in transphosphorylations. I. The enzymatic phosphorylation of hexoses by adenyl pyrophosphate. J. Biol. Chem. 148:117-126. 101. Colowick, S. P., and F. C. Womack. 1969. Binding of diffusible molecules by macromolecules: rapid measurement by rate of dialysis. J. Biol. Chem. 244:774-777. 102. Cowles, J. R., and H. J. Evans. 1968. Some properties of the ribonucleotide reductase from Rhizobian meliloti. Arch. Biochem. Biophys. 127:770-778. 103. Cowles, J. R., H. J. Evans, and S. A. Russell. 1969. B12 co-

113.

and properties of orotidylate decarboxylases from yeast and rat liver. J. Biol. Chem. 236:2058-2063. Cummings, D. J., and L. Mondale, 1967. Thymineless death in Escherichia cola: strain specificity. J. Bacteriol. 93:19171924. Dale, B., and G. R. Greenberg. 1967. Genetic mapping of a mutation in Escherkchia coll showing reduced activity of thymidine phosphorylase. J. Bacteriol. 94:778-779. Datta, P., and H. Gest. 1965. Homoserine dehydrogenase of Rhodospirillum rubrum. Purification, properties, and feedback control activity. J. Biol. Chem. 240:3023-3033. Davis, R. H. 1960. An enzymatic difference among pyr-3 mutants of Neurospora. Proc. Nat. Acad. Sci. U.S.A. 46: 677-682. Davis, R. H. 1961. Suppressor of pyrimidine-3 mutants of Neurospora and its relation to arginine synthesis. Science 134:470-471. Davis, R. H. 1962. Consequences of a suppressor gene effective with pyrimidine and proline mutants of Neurospora. Genetics 47:351-360. Davis, R. H. 1963. Neurospora mutant lacking an arginine-

specific carbamyl phosphokinase.

Science 142:1652-1654.

114. Davis, R. H. 1965. Carbamyl phosphate synthesis in Neurospora crassa. II. Genetics, metabolic position, and regulation of arginine-specific carbamyl phosphokinase. Biochim. Iiophys. Acta 107:54-68. 115. Davis, R. H. 1967. Channeling in Neurospora metabolism, p. 303-322. In H. J. Vogel, L. 0. Lampen, and V. Bryson

(ed.), Organizational biosynthesis.

Academic Press Inc.,

New York. 116. Davis, R. H., and V. W. Woodward. 1962. The relationship between gene suppression and aspartate transcarbamylase

activity

in pyr-3 mutants of Neurospora. Genetics 47:1075-

1083. 117. Denhardt, D. T. 1969. Formation of ribosylthymine in Escherkchia coli: studies on pulse labelling with thymine and thymidine. J. Biol. Chem. 244:2710-2715. 118. Dennis, P. P., and R. K. Herman. 1970. Pyrimidine pools and macromolecular composition of pyrimidine-limited Escherichia colt. J. Bacteriol. 102:118-123. 119. Donachie, W. D. 1964. The regulation of pyrimidine bio-

synthesis in Neurospora crassa. I. End-product inhibition and repression of aspartate carbamoyl-transferase. Biochim. Biophys. Acta 82:284-292. 120. Durham, L. J., A. Larsson, and P. Reichard. 1967. Enzymatic synthesis of deoxyribonucleotides. 11. The mechanism of hydrogen transfer of the ribonucleoside diphosphate reductase system from Escherichia coli studied with nuclear magnetic resonance. Eur. J. Biochem. 1:92-95. 121. Eakin, R. T., and H. K. Mitchell. 1969. A mitochondrial dihydroorotate oxidase system in Neurospora crassa. Arch. Biochem. Biophys. 134:160-171. 122. Edlin, G., and G. S. Stent. 1969. Nucleoside triphosphate pools and the regulation of RNA synthesis in E. coil. Proc. Nat. Acad. Sci. U.S.A. 62:475-482. 123. Eisenstark, A. 1967. Linkage of arginine-sensitive (ars) and uracil-arginine requiring (pyrA) loci of Salmonella

typhimurium.

Nature 213:1263-1264.

124. Eisenstark, A., R. Eisenstark, and S. Cunningham. Genetic analysis of thymineless (thy) mutants in

monella typhimurium. Genetics 58:493-506. 125. Elford, H. L. 1968. Effect of hydroxyurea on

1968. Sal-

ribonucleotide

reductase. Biochem. Biophys. Res. Commun. 33:129-135. 126. Fangman, W. L. 1969. Specificity and efficiency of thymi-

PYRIMIDINE METABOLISM IN MICROORGANISMS

VOL. 34, 1970

148. Gerhart, J. C., and A. B. Pardee. 1964. Aspartate transcarbamylase, an enzyme designed for feedback inhibition.

dine incorporation in Escherichia coli lacking thymidine phosphorylase. J. Bacteriol. 99:681-687. 127. Fangman, W. L., and A. Novick. 1966. Mutant bacteria showing efficient utilization of thymidine. J. Bacteriol.

Fed. Proc. 23:727-735.

91:2390-2394. 128. Farmer, J. L., and F. Rothman. 1965. Transformable thymine-requiring mutant of Bacillus subtilis. J. Bacteriol.

89:262-263. 129. Finck, D., Y. Suyama, and R. H. Davis. 1965. Metabolic role of pyrimidine-3 locus of Neurospora. Genetics 52: 829-839.

130. Fink, R. M., and K. Fink. 1961. Biosynthesis of radioactive RNA and DNA pyrimidines from thymidine-2-C04. Biochem. Biophys. Res. Commun. 6:7-10. 131. Fink, R. M., and K. Fink. 1962. Utilization of radiocarbon from thymidine and other precursors of ribonucleic acid in Neurospora crassa. J. Biol. Chem. 237:2289-2290. 132. Fink, R. M., and K. Fink. 1962. Relative retention of Ha and C14 labels of nucleosides incorporated into nucleic acids of Neurospora. J. Biol. Chem. 237:2889-2891. 133. Fink, G. R., and J. R. Roth. 1968. Histidine regulatory

typhirnurium. VI. studies. J. Mol. Biol. 33:547-557.

mutants in Salmonella

Dominance

134. Flaks, J. G. 1963. Nucleotide synthesis from 5-phosphoribosylpyrophosphate. III. Orotidine 5'-phosphate pyrophosphorylase. Methods in Enzymology 6:148-152. 135. Flaks, J. G., and S. S. Cohen. 1957. The enzymic synthesis of 5-hydroxymethyl-deoxycytidylic acid. Biochim. Biophys. Acta 25:667-668. 136. Fleming, W. H., and M. J. Bessman. 1967. The enzymology of virus-infected bacteria. IX. Purification and properties of the deoxycytidylate deaminase of T6-infected Escherichia coli. J. Biol. Chem. 242:363-371. 137. Follmann, H., and H. P. C. Hogenkamp. 1969. Ribonucleotide reductase. Studies with 15O-labeled substrates. Biochemistry. 8:4372-4375. 138. Friedkin, M., E. J. Crawford, E. Donovan, and E. J. Pastore. 1962. The enzymatic synthesis of thymidylate. III. The further purification of thymidylate synthetase and its separation from natural fluorescent inhibitors. J. Biol. Chem. 237:3811-3814. 139. Friedkin, M., and A. Kornberg. 1957. The enzymatic version of deoxyuridylic acid to thymidylic acid and the participation of tetrahydrofolic acid, p. 609-613. In con-

W. D. McElroy and B. Glass (ed.), A symposium chemical basis of heredity. Johns Hopkins Press,

on

the

Balti-

more.

140. Friedmann, H. C. 1963. Dihydroorotic dehydrogenase. Methods in Enzymology 6:197-203. 141. Friedmann, H. C., and B. Vennesland. 1958. Purification and properties of dihydroorotic dehydrogenase. J. Biol. Chem. 233:1398-1406 142. Friedmann, H. C., and B. Vennesland. 1960. Crystalline dihydroorotic dehydrogenase. J. Biol. Chem. 235:15261532. 143. Furth, J. J., and S. S. Cohen. 1968. Inhibition of mammalian DNA polymerase by the 5'-triphosphate of 1- --arabinofuranosylcytosine and the 5'-triphosphate of 9-

-D

arabinofuranosyladenine. Cancer Res. 28:2061-2067. 144. Gardner, R., and A. Kornberg. 1967. Biochemical studies of bacterial sporulation and germination. V. Purine nucleoside phosphorylase of vegetative cells and spores of Bacillus

cereus.

J. Biol. Chem. 242:2383-2388.

145. Gerhart, J. C. 1964. Subunits for control and catalysis aspartate transcarbamylase.

Brookhaven

Symp.

337

in

Biol.

17:222-231. 146. Gerhart, J. C., and H. Holoubek. 1967. The purification of aspartate transcarbamylase of Escherichia coll and separation of its protein subunits. J. Biol. Chem. 242:28862892. 147. Gerhart, J. C., and A. B. Pardee. 1962. The enzymology of control by feedback inhibition. J. Biol. Chem. 237:891896.

149. Gerhart, J. C., and H. K. Schachman, 1965. Distinct subunits for the regulation and catalytic activity of aspartate transcarbamylase. Biochemistry 4:1054-1062. 150 Gerhart, J. C., and H. K. Schachman. 1968. Allosteric interactions in aspartate transcarbamylase. II. Evidence for different conformational states of the protein in the presence and absence of specific ligands. Biochemistry 7:538-552. 151. Ghambeer, R. K., and R. L. Blakley. 1965. Factors influencing the level of cobamide-dependent ribonucleoside triphosphate reductase in Lactobacillus leichmannii. Biochem. Biophys. Res. Commun. 21:40-47. 152. Ghambeer, R. K., and R. L. Blakley. 1966. Cobamide and ribonucleotide reduction. III. Factors influencing the level of cobamide-dependent ribonucleoside triphosphate reductase in Lactobacillus leichmannii. J. Biol. Chem.

241:4710-4716. 153. Ginsberg, T., and F. F. Davis. 1968. The biosynthesis of pseudouridine in ribonucleic acids of Escherichia coli. J. Biol. Chem. 243:6300-6305. 154. Gorini, L., W. Gundersen, and M. Burger. 1961. Genetics of regulation of enzyme synthesis in the arginine biosynthetic pathway of Escherichia coli. Cold Spring Harbor Symp. Quant. Biol. 26:173-182. 155. Gottesman, M. M., and W. S. Beck. 1966. Transfer of hydrogen in the cobamide-dependent ribonucleotide reductase reaction. Biochem. Biophys. Res. Commun. 24:353359. 156. Goulian, M., and W. S. Beck. 1966. Purification and properties of cobamide-dependent ribonucleotide reductase from Lactobacillus lekichmannil. J. Biol. Chem. 241:4233-4242. 157. Goulian, M., and W. S. Beck. 1966. Variations of intracellular deoxyribosyl compounds in deficiencies of vitamin B12, folic acid, and thymine. Biochim. Biophys. Acta 129:

336-349. 158. Greenberg, G. R. 1966. New dUTPase and dUDPase activities after infection of Escherichia coli by T2 bacteriophage. Proc. Nat. Acad. Sci. U.S.A. 56:1226-1232. 159. Greenberg, G. R., and R. L. Somerville. 1962. Deoxyuridylate kinase activity and deoxyuridinetriphosphatase in Escherichia coli. Proc. Nat. Acad. Sci. U.S.A. 48:249-257. 160. Grenson, M. 1969. The utilization of exogenous pyrimidines

and the recycling of uridine-5'-phosphate derivatives in Saccharomyces cerevisiae, as studied by means of mutants affected in pyrimidine uptake and metabolism. Eur. J. Biochem. 11:249-260. 161. Griffin, C. E., F. D. Hamilton, S. P. Hopper, and R. Abrams. 1968. Stereospecificity of deoxycytidine triphosphate synthesis with the ribonucleotide reductase of Lactobaclllus leichmannii. Arch. Biochem. Biophys. 126:905-911. 162. Grivell, A. R., and J. F. Jackson. 1968. Thymidine kinase: evidence for its absence from Neurospora crassa and some other micro-organisms, and the relevance of this to the specific labelling of deoxyribonucleic acid. J. Gen. Microbiol. 54:307-317. 163. Gross, S. R. 1965. The regulation of synthesis of leucine biosynthetic enzymes in Neurospora. Proc. Nat. Acad. Sci. U.S.A. 54:1538-1546. 164. Gutnick, D., J. M. Calvo, T. Klopotowski, and B. N. Ames. 1969. Compounds which serve as the sole source of carbon or nitrogen for Salmonella typhimurium LT-2. J. Bacteriol. 100:215-219. 165. Habermann, V., and F. Sorm. 1958. Mechanism of the cancerostatic action of 6-azauracil and its riboside. Collect. Cesk. Chem. Commun. 23:2201-2206. 166. Hager, S. E., and M. E. Jones. 1967. A glutamine dependent enzyme for the synthesis of carbamyl phosphate for pyrimidine biosynthesis in fetal rat liver. J. Biol. Chem. 242: 5674-5680.

338

O'DONOVAN AND NEUHARD

167. Hahn, A., and W. Lentzel. 1923. tber das Verhalten von Pyrimidinderivaten in den Organismen. I. Einfluss von Hefe auf Pyrimidinderivate. Z. Biol. 79:179-190. 168. Hahn, A., and L. Schafer. 1925. tber das Verhalten von Pyrimidinderivaten in den Organismen. Z. Biol. 83:511514. 169. Hamilton, J. A. and R. L. Blakley. 1969. Electron spin resonance studies of ribonucleotide reduction catalyzed by the ribonucleotide reductase of Lactobacillus leichmannii. Biochim. Biophys. Acta 184:224-226. 170. Hamilton, J. A., R. L. Blakley, F. D. Looney, and M. E. Winfield. 1969. Formation of a cobamide containing divalent cobalt by the ribonucleotide reductase of Lactobacillus leichmannii. Biochim. Biophys. Acta 177:374-376. 171. Handschumacher, R. E. 1957. Studies of bacterial resistance to 6-azauracil and its riboside. Biochim. Biophys. Acta

23:428-430. 172. Handschumacher, R. E. 1958. Bacterial preparation of orotidine-5'-phosphate and uridine-5'-phosphate. Nature 182:1909-1910. 173. Handschumacher, R. E. 1960. Orotidylic acid decarboxylase: inhibition studies with azauridine 5'-phosphate. J. Biol. Chem. 235:2917-2919. 174. Handschumacher, R. E., and C. A. Pasternak. 1958. Inhibition of orotidylic acid decarboxylase, a primary site of carcinostasis by 6-azauracil. Biochim. Biophys. Acta 30:451-452. 175. Handschumacher, R. E., and A. D. Welch. 1956. Microbial studies of 6-azauracil, an antagonist of uracil. Cancer Res. 16:965-969. 176 Harrison, A. P. 1965. Thymine incorporation and metabolism by various classes of thymine-less bacteria. J. Gen. Microbiol. 41:321-333. 177. Hayaishi, O., and A. Kornberg. 1952. Metabolism of cytosine, thymine, uracil and barbituric acid by bacterial enzymes. J. Biol. Chem. 197:717-732. 178. Hayward, W. S., and W. L. Belser. 1965. Regulation of pyrimidine biosynthesis in Serratia marcescens. Proc. Nat. Acad. Sci. U.S.A. 53:1483-1489. 179. Hill, A. J. 1913. The combinations of haemoglobin with oxygen and with carbon monoxide. Biochem. J. 7:471480. 180. Hiraga, S., K. Igarashi, and T. Yura. 1967. A deoxythymidine kinase-deficient mutant of Escherichia coli. I. Isolation and some properties. Biochim. Biophys. Acta 145:

41-51. 181. Hoffee, P. 1968. 2-Deoxyribose gene-enzyme complex in Salmonella typhimurium. I. Isolation and enzymatic characterization of 2-deoxyribose-negative mutants. J. Bacteriol. 95:449-457. 182. Hoffee, P. A. 1968. 2-Deoxyribose-5-phosphate aldolase of Salmonella typhimurium: purification and properties. Arch. Biochem. Biophys. 126:795-802. 183. Hoffee, P. A., and B. C. Robertson. 1969. 2-Deoxyribose gene-enzyme complex in Salmonella typhimurium: regulation of phosphodeoxyribomutase. J. Bacteriol. 97:13861396. 184. Hoffmann, C. E., and J. 0. Lampen. 1952. Products of deoxyribose degradation by Escherichia coli. J. Biol. Chem. 198:885-893. 185. Hogenkamp, H. P. C., R. K. Ghambeer, C. Brownson, R. L. Blakley, and E. Vitols. 1968. Cobamides and ribonucleotide reduction. VI. Enzyme-catalyzed hydrogen exchange between water and deoxyadenosylcobalamin. J. Biol. Chem. 243:799-808. 186. Holmes, W. L. 1956. Studies on the mode of action of analogues of orotic acid: 6-uracilsulfonic acid, 6-uracilsulfonamide, and 6-uracil methyl sulfone. J. Biol. Chem. 223: 677-686. 187. Holmgren, A. 1968. Thioredoxin. IV. Amino acid sequence of peptide B. Eur. J. Biochem. 5:359-365. 188. Holmgren, A. 1968. Thioredoxin. V. Amino acid sequences

BACTERIOL. REV.

of the tryptic peptides of peptide A. Eur. J. Biochem. 6:467-474. 189. Holmgren, A., R. N. Perham, and A. Baldsten. 1968. Thioredoxin. III. Amino acid sequences of the peptic peptides from S-aminoethylated peptide B. Eur. J. Biochem. 5: 352-358. 190. Holmgren, A., and P. Reichard. 1967. Thioredoxin. II. Cleavage with cyanogen bromide. Eur. J. Biochem. 2: 187-196. 191. Honzova, H., B. Javurkova, J. Skoda, and J. Dyr. 1968. Quantitative evaluation of the mutagenic effect of ethyl methanesulfonate (EMS) in Brevibacterium ammoniagenes and qualitative character of mutants obtained. Folia Microbiol. (Prague) 13:125-128. 192. Hurlbert, R. B., and H. 0. Kammen. 1960. Formation of cytidine nucleotides from uridine nucleotides by soluble mammalian enzymes: requirements for glutamine and guanosine nucleotides. J. Biol. Chem. 235:443-449. 193. Hurlbert, R. B., and V. R. Potter. 1952. A survey of the metabolism of orotic acid in the rat. J. Biol. Chem. 195: 257-270. 194. Hurlbert, R. B., and P. Reichard. 1955. The conversion of orotic acid to uridine nucleotides in vitro. Acta Chem. Scand. 9:251-262. 195. Hutson, J. Y., and M. Downing. 1968. Pyrimidine biosynthesis in Lactobacillus leichmannii. J. Bacteriol. 96:12491254. 196. Igarashi, K., S. Hiraga, and T. Yura. 1967. A deoxythymidine kinase deficient mutant of Escherichia coil. II. Mapping and transduction studies with phage 4,80. Genetics 57:643-654. 197. Imada, A., and S. Igarasi. 1967. Ribosyl and deoxyribosyl transfer by bacterial enzyme systems. J. Bacteriol. 94: 1551-1559. 198. Ingraham, J. L., and 0. Maaloe. 1966. Cold-sensitive mutants and the minimum temperature of growth of bacteria, p. 297-309. In C. L. Prosser (ed.), Molecular mechanisms of temperature adaptation. Amer. Ass. Advan. Sci. Publ. 199. Isaac, J. H., and B. W. Holloway. 1968. Control of pyrimidine biosynthesis in Pseudomonas aeruginosa. J. Bacteriol. 96:1732-1741. 200. Ishibashi, M., Y. Sugino, and Y. Hirota. 1964. Chromosomal location of thymine and arginine genes in Escherichia coil and an F' incorporating them. J. Bacteriol. 87:554-567. 201. Issaly, I. M., A. S. Issaly, and J. L. Reissig. 1970. Carbamyl phosphate biosynthesis in Bacillus subtilis. Biochim. Biophys. Acta 198:482-494. 202. Iwatsuki, N., and R. Okazaki. 1967. Mechanism of regulation of deoxythymidine kinase of Escherichia coli. I. Effect of regulatory deoxynucleotides on the state of aggregation. J. Mol. Biol. 29:139-154. 203. Iwatsuki, N., and R. Okazaki. 1967. Mechanism of regulation of deoxythymidine kinase of Escherichia coli. It. Effect of temperature on the enzyme activity and kinetics. J. Mol. Biol. 29:155-165. 204. Jones, M. E., L. Spector, and F. Lipmann. 1955. Carbamyl phosphate, the carbamyl donor in enzymatic citrulline synthesis. J. Amer. Chem. Soc. 77:819-820. 205. Kalman, S. M., P. H. Duffield, and T. Brozowski. 1966. Purification and properties of a bacterial carbamyl phosphate synthetase. J. Biol. Chem. 241:1871-1877. 206. Kanmen, H. 0. 1967. Thymine metabolism in Escherichia coli. I. Factors involved in utilization of exogenous thymine. Biochim. Biophys. Acta 131:301-311. 207. Kanda, M., and Y. Takagi. 1959. Purification and properties of a bacterial deoxyribose transferase. J. Biochem. 46:725-732. 208. Kaplan, J. G., M. Duphil, and F. Lacroute. 1967. A study of the aspartate transcarbamylase activity of yeast. Arch. Biochem. Biophys. 119:541-551. 209. Kaplan, J. G., F. Lacroute, and I. Messmer. 1969. On the

VOL. 34, 1970

loss of feedback inhibition of yeast aspartate transcarbamylase during derepression of pyrimidine biosynthesis. Arch. Biochem. Biophys. 129:539-544 and Messmer. 1969. The combined effects 210. Kaplan, J. G., I. of temperature and dilution on the activity and feedback inhibition of yeast aspartate transcarbamylase. Can. J. Biochem. 47:477-479. 211. Karlstrom, 0. 1968. Mutants of Escherichia coil defective in ribonucleoside and deoxyribonucleoside catabolism. J. Bacteriol. 95:1069-1077. and A. Larsson. 1967. Significance of ribo212. Karlstrom, nucleotide reduction in the biosynthesis of deoxyribonucleotides in Escherichia coi. Eur. J. Biochem. 3:164170.

215. 216. 217.

218.

dihydroorotate dehydrogenase from a Pseudomonad. Can. J. Biochem. 45:1295-1307. Kirtley, M. E., and D. E. Koshland, Jr. 1967. Models for cooperative effects in proteins containing subunits. J. Biol. Chem. 242:4192-4205. Kitsuji, N. 1964. Thymineless mutation site on coli chromosome. J. Bacteriol. 87:802-808. Kleppe, K. 1966. Aspartate transcarbamylase fromEscherichia coil. I. Inhibition by inorganic anions. Biochim. Biophys. Acta 122:450-461. Koch, A. L. 1956. Some enzymes of nucleoside metabolism ofEscherichia coli. J. Biol. Chem. 223:535-549. Koerner, J. F., M. S. Smith, and J. M. Buchanan. 1960. dCTPase, an enzyme induced by bacteriophage infection.

Escherkchia

J. Biol. Chem. 235:2691-2694.

219. Kornberg, A., I. Lieberman, and E. S. Simms. 1955. Enzymatic synthesis and properties of 5-phosphoribosylpyrophosphate. J. Biol. Chem. 215:389-402. G. Nemethy, and D. Filmer. 1966. Comparison of experimental binding data and theoretical

220. Koshland, D. E., Jr.,

models in proteins containing subunits. Biochemistry 5:

365-385.

I. H., N. C. Brown, and P. Reichard. 1968. Inhi221. Krakoff, bition of ribonucleoside diphosphate reductase by hydroxyurea. Cancer Res. 28:1559-1565.

222. Kream, J., and E. Chargaff. 1952. On the cytosine deaminase of yeast. J. Amer. Chem. Soc. 74:5157-5160.

223. Krebs,

H. A., and R. Hems. 1953. Some reactions of adeno-

sine and inosine phosphates in animal tissues. Biochim. Biophys. Acta 12:172-180. 224. Krooth, R. S. 1964. Properties of diploid cell strain developed from patients with inherited abnormality of uridine biosynthesis. Cold Spring Harbor Symp. Quant. Biol. 29:189-212. 225. Lacroute, F. 1968. Regulation of pyrimidine biosynthesis in Saccharomyces cerevisiae. J. Bacteriol. 95:824-832. 226. Lacroute, F., A. Pierard, M. Grenson, and J. M. Wiame. 1965. The biosynthesis of carbamyl phosphate in Saccharomyces cerevisiae. J. Gen. Microbiol. 40:127-142. 227. Lacroute, F., and P. P. Slonimski. 1964. Etude physiologique de mutants resistant au 5-fluorouracile chez le levure. C. R. Acad. Sci. Paris 258:2172-2174. 228. Lampen, J. O., and T. P. Wang. 1952. The mechanism of action of Lactobacillus pentosus nucleosidase. J. Biol. Chem. 198:385-395. 229. Larsson, A. 1965. Enzymatic synthesis of deoxyribonucleotides. VII. Studies on the hydrogen transfer with tritiated water. Biochemistry 4:1984-1993. 230. Larsson, A., and P. Reichard. 1966. Enzymatic synthesis of

deoxyribonucleotides. IX. Allosteric effects in the reduction of pyrimidine ribonucleotides by the ribonucleoside diphosphate reductase system of Escherichia coll. J. Biol. Chem. 241:2533-2539. 231. Larsson, A., and P. Reichard. 1966. Enzymatic synthesis of deoxyribonucleotides. X. Reduction of purine ribonucleotides: allosteric behavior and substrate specificity of the

Escherichla coil

enzyme system from 241:2540-2549.

B. J. Biol. Chem.

232. Larsson, A., and P. Reichard. 1967. Enzymatic reduction of Acid Res. 7:303-347. ribonucleotides. Progr. 233. Larsson, A., and L. Thelander. 1965. The stereospecificity of thioredoxin reductase for triphosphopyridine nucleotide. J. Biol. Chem. 240:2691-2693. 234. Laurent, T. C., E. C. Moore, and P. Reichard. 1964. Enzymatic synthesis of deoxyribonucleotides. IV. Isolation and characterization of thioredoxin, the hydrogen donor from B. J. Biol. Chem. 239:3436-3444. H. B., and S. Jackson. 1968. Allosteric interaction 235. of a regulatory nicotinamide adenine dinucleotide-specific glutamate dehydrogenase from blastocladiella. J. Biol.

NMud.

Fscherichia coli 1eJohn,

Chem. 243:3447-3457.

213. Kerr, C. T., and R. W. Miller. 1967. Soluble NADP-linked 214.

339

PYRIMIDINE METABOLISM IN MICROORGANISMS

236. Leloir, L. F. 1965. The biosynthesis of polysaccharides. Proc. VI Congr. Biochem. N.Y. 33:15-35. 237. Leloir, L. F., and C. E. Cardini. 1960. p. 39. In P. Boyer, H. A. Lardy, and K. Myrback, (ed.), The enzymes, vol. 2.

Int.

Academic Press Inc., New York.

238. Levitzki, A., and D. E. Koshland, Jr. 1969. Negative cooperativity in regulatory enzymes. Proc. Nat. Acad. Sci. U.S.A. 62:1121-1128.

239.

J., H. D. Barner, and S. S. Cohen. 1960. The Lichtenstein, metabolism of exogenously supplied nucleotides by Escherichia coli. J. Biol. Chem. 235:457-465. of uridine triphos-

240. Lieberman,I. 1955.

to cytidine 2661-2662.

phate

Enzymatic amination triphosphate. J. Amer. Chem. Soc.

77:

241. Lieberman,I. 1956. Enzymatic amination of uridine triphosphate to cytidine triphosphate. Biol. Chem. 222:765-

J.

775.

Enzymic

synthesis 242. Lieberman, I., and A. Kornberg. 1953. and breakdown of a pyrimidine, orotic acid. I. Dihydroorotic dehydrogenase. Biochim. Biophys. Acta 12:223234.

243. Lieberman, I., and A. Kornberg. 1954. Enzymatic synthesis and breakdown of a pyrimidine, orotic acid. II. Dihydroorotic acid, ureidosuccinic acid, and 5-carboxy-methylhydantoin. J. Biol. Chem. 207:911-924. 244. Lieberman, I., and A. Komberg. 1955. Enzymatic synthesis and breakdown of a pyrimidine, orotic acid. III. Ureidosuccinate. J. Biol. Chem. 212:909-920. A. Kornberg, and E. S. Simms. 1954. Enzy245. Lieberman, matic synthesis of pyrimidine and purine nucleotides. II. Orotidine-5'-phosphate pyrophosphorylase and decarboxylase. J. Amer. Chem. Soc. 76:2844-2845. 246. Lieberman, I., A. Kornberg, and E. S. Simns. 1955. Enzymatic synthesis of pyrimidine nucleotides. Orotidine-5'phosphate and uridine-5'phosphate. J. Biol. Chem. 215:

I.,

403-415.

Simms.

I.,

1955. EnzyA Kornberg, and E. S. 247. Lieberman, matic synthesis of nucleoside diphosphates and triphosphates. J. Biol. Chem. 215:429-440. 248. Liebl, V., G. Kaplan, and D. Kushner. 1969. Regulation of a salt-dependent enzyme: the aspartate transcarbamylase of an extreme haolphile. Can. J. Biochem.

J.

J.

47:1095-1097.

249. Lisy, V., and J. Skoda. 1966. Deamination of some cytosine col extract. and cytidine analogues by an Collect. Cesk. Chem. Commun. 31:3020-3022. 250. Lomax, M. I. S., and G. R. Greenberg. 1967. An exchange between the hydrogen atom on carbon 5 of the deoxyuridylate and water catalyzed by thymidylate synthetase. J.

Escherichia

Biol. Chem. 242:1302-1306.

251. Lomax, M. S., and G. R. Greenberg. 1968. Characteristics of the deo operon: role in thymine utilization and sensitivity to deoxyribonucleosides. J. Bacteriol. 96:501-514. and D. E. Koshland, Jr. 1970. 252. Long, C. W., A. The subunit structure and subunit interactions of cytidine triphosphate synthetase. J. Biol. Chem. 245:80-87. 253. Long, C. W., and A. B. Pardee. 1967. Cytidine triphosphate

Levitzki,

340 254.

255.

256.

257

258. 259. 260.

261. 262. 263.

264. 265.

O'DONOVAN AND NEUHARD synthetase of Escherichia coli B. I. Purification and kinetics. J. Biol. Chem. 242:4715-4721. Loring, H. S., and J. G. Pierce. 1944. Pyrimidine nucleosides and nucleotides as growth factors for mutant strains of Neurospora. J. Biol. Chem. 153:61-69. Lou, M. F., and R. L. Hermann. 1967. Pyrimidine-specific carbamate kinase in Neurospora. Biochim. Biophys. Acta 139:199-201. Lue, P. F., and J. G. Kaplan. 1969. The aspartate transcarbamylase and carbamyl phosphate synthetase of yeast: a multi-functional enzyme complex. Biochem. Biophys. Res. Commun. 34:426-433. Lue, P. F., and J. G. Kaplan. 1970. Heat-induced disaggregation of a multifunctional enzyme complex catalyzing the first steps in pyrimidine biosynthesis in bakers' yeast. Can. J. Biochem. 48:155-159. Machida, H., and A. Kuninaka. 1969. Studies on the accumulation of orotic acid by Escherichia coli K-12. Agr. Biol. Chem. 33:868-875. MacNutt, W. S. 1952. The enzymically catalyzed transfer of the deoxyribosyl group from one purine or pyrimidine to another. Biochem. J. 50:384-397. Magasanik, B. 1962. Biosynthesis of purine and pyrimidine nucleotides, p. 295-334. In I. C. Gunsalus and R. Y. Stanier (ed.), The bacteria, vol. 3. Academic Press Inc., New York. Maley, G. F., and F. Maley. 1966. The significance of the substrate specificity of T2r+-induced deoxycytidylate deaminase. J. Biol. Chem. 241:2176-2177. Manson, L. A., and J. 0. Lampen. 1951. The metabolism of deoxyribose nucleosides in Escherkchia coll. J. Biol. Chem. 193:539-547. Marsh, J. C., and M. E. King. 1959. Purification of transN-glycosidase of Thermobacter acidophilus: inhibition of enzyme by 6-azathymidine. Biochem. Pharmacol. 2:146153. Massey, V., and C. Veeger. 1961. Studies on the reaction mechanism of lipoyl dehydrogenase. Biochim. Biophys. Acta 48:33-47. Massey, V., and C. H. Williams. 1965. On the reaction mechanism of yeast glutathione reductase. J. Biol. Chem.

240:4470-4480. 266 Mathews, C. K., and S. S. Cohen. 1963. Inhibition of phageinduced thymidylate synthetase by 5-fluorodeoxyuridylate. J. Biol. Chem. 238:367- 370. 267. Michelson, A. M., W. Drell, and H. K. Mitchell. 1951. A new ribose nucleoside from Neurospora: "orotidine." Proc. Nat. Acad. Sci. U.S.A. 37:396-399. 268. Miller, R. W., and C. T. Kerr. 1967. Particulate dihydroorotate oxidase from a pseudomonad. Can. J. Biochem. 45: 1283-1294. 269. Miller, R. W., and V. Massey. 1965. Dihydroorotic dehydrogenase. I. Some properties of the enzyme. J. Biol. Chem. 240:1453-1465. 270. Minghetti, A. 1959. Transdeoxyribosidations. Ital. J. Biochem. 8:224-230. 271. Mitchell, H. K., M. B. Houlahan, and J. F. Nyc. 1948. The accumulation of orotic acid by a pyrimidine-less mutant of Neurospora. J. Biol. Chem. 172:525-529. 272. Molloy. A., and L. R. Finch. 1969. Uridine-5'-monophosphate pyrophosphorylase activity from Escherichia coli. Fed. Eur. Biochem. Soc. Letters 5:211-213. 273. Monod, J., J.-P. Changeux, and F. Jacob. 1963. Allosteric proteins and cellular control systems. J. Mol. Biol. 6:306329. 274. Moore, A. M., and J. B. Boylen. 1955. Utilization of uracil by a strain of Escherichia coli. Arch. Biochem. Biophys. 54:312-317. 275. Moore, E. C., P. Reichard, and L. Thelander. 1964. Enzymatic synthesis of deoxyribonucleotides. V. Purification and properties of thioredoxin reductase from Escherichia coli B. J. Biol. Chem. 239:3445-3452. 276. Munch-Petersen, A. 1967. Thymidine breakdown and thy-

277.

278. 279.

280.

281. 282. 283.

BACTERIOL. REV.

mine uptake in different mutants of Escherichia col. Biochim. Biophys. Acta 142:228-237. Munch-Petersen, A. 1968. On the catabolism of deoxyribonucleosides in cells and cell extracts of Escherkchia coli. Eur. J. Biochem. 6:432-442. Munch-Petersen, A. 1968. Thymineless mutants of Escherichia coli with deficiencies in deoxyribomutase and deoxyriboaldolase. Biochim. Biophys. Acta 161:279-282. Nakanishi, S., K. Ito, and M. Tatibana. 1968. Two types of carbamyl phosphate synthetase in rat liver: chromatographic resolution and immunological distinction. Biochem. Biophys. Res. Commun. 33:774-781. Neuhard, J. 1966. Studies on the acid-soluble nucleotide pool in thymine-requiring mutants of Escherichia coli during thymine starvation. III. On the regulation of the deoxyadenosine triphosphate and deoxycytidine triphosphate pools of Escherichia coli. Biochim. Biophys. Acta 129:104-115. Neuhard, J. 1967. Studies on the acid-soluble nucleotide pool in Escherichia coli. IV. Effects of hydroxyurea. Biochim. Biophys. Acta 145:1-6. Neuhard, J. 1968. Pyrimidine nucleotide metabolism and pathways of thymidine triphosphate biosynthesis in Salmonella typhimurium. J. Bacteriol. 96:1519-1527. Neuhard, J., and J. Ingraham. 1968. Mutants of Salmonella typhimurium requiring cytidine for growth. J. Bacteriol.

95:2431-2433. 284. Neuhard, J., and A. Munch-Petersen. 1966. Studies on the acid-soluble nucleotide pool in thymine-requiring mutants of Escherkchia coli during thymine starvation. II. Changes in the amounts of deoxycytidine triphosphate and deoxyadenosine triphosphate in Escherichia coll 15 TAU. Biochim. Biophys. Acta 114:61-71. 285. Neumann, J., and M. E. Jones. 1964. End product inhibition of aspartate transcarbamylase in various species. Arch. Biochem. Biophys. 104:438-447. 286. Nierlich, D. P. 1968. Amino acid control over RNA synthesis: a re-evaluation. Proc. Nat. Acad. Sci. U.S.A. 60: 1345-1352. 287. Nikaido, H. 1968. Biosynthesis of cell wall lipopolysaccharide in gram-negative bacteria, p. 77-124. In F. F. Nord (ed.), Advances in enzymology and related areas of molecular biology, vol. XXXI. Interscience Publishers, Inc., New York. 288. O'Donovan, G. A. 1970. Nucleotide pool changes in mutants of Escherichia col. Biochim. Biophys. Acta 209: 589-591. 289. O'Donovan, G. A., and J. L. Ingraham. 1965. Cold-sensitive mutants of Escherkchia coli resulting from increased feedback inhibition. Proc. Nat. Acad. Sci. U.S.A. 54: 451-457. 290. O'Donovan, G. A., C. L. Kearney, and J. L. Ingraham. 1965. Mutants of Escherichia coli with high minimal temperatures of growth. J. Bacteriol. 90:611-616. 291. Okada, T. 1966. Mutational site on the gene controlling quantitative thymine requirement in Escherichia coli K-12. Genetics 54:1329-1336. 292. Okada, T., J. Homma, and H. Sonohara. 1962. Improved method for obtaining thymineless mutants of Escherichia coli and Salmonella typhimurium. J. Bacteriol. 84:602-603. 293. Okada, T., K. Yanagisawa, and F. J. Ryan. 1961. A method for securing thymineless mutants from strains of E. coil. Z. Vererb. 92:403-412. 294. Okazaki, R., and A. Kornberg. 1964. Deoxythymidine kinase of Eycherichia coli. I. Purification and some properties of the enzyme. J. Biol. Chem. 239:269-274. 295. Okazaki, R., and A. Kornberg. 1964. Deoxythymidine kinase of Escherichla coli. II. Kinetics and feedback control. J. Biol. Chem. 239:275-284. 296. Orengo, A. 1969. Regulation of enzymic activity by metabolites. I. Uridine-cytidine kinase of Novikoff ascites rat tumor. J. Biol. Chem. 244:2204-2209. 297. Orr, M. D., and E. Vitols. 1966. Thioredoxin from Lacto-

VOL. 34, 1970

PYRIMIDINE METABOLISM IN MICROORGANISMS

341

1964. Nucleoside triphosphate-nucleoside diphosphate bacillus leichmannii and its role as hydrogen donor for ribonucleoside triphosphate reductase. Biochem. Biophys. transphosphorylase (nucleoside diphosphokinase). I. Isolation of the crystalline enzyme from brewers' yeast. J. Res. Commun. 25:109-115. Biol. Chem. 239:301-309. 298. Osborn, M. J. 1969. Structure and biosynthesis of bacterial 320. Razzell, W. E., and P. Casshyap. 1964. Substrate specificity cell wall. Annu. Rev. Biochem. 38:501-538. and induction of thymidine phosphorylase in Escherichia 299. Paege, L. M., and F. Schlenk. 1950. Pyrimidine riboside metabolism. Arch. Biochem. Biophys. 28:348-358. coli. J. Biol. Chem. 239:1789-1793. 321. Razzell, W. E., and H. G. Khorana. 1958. Purification and 300. Paege, L. M., and F. Schlenk. 1952. Bacterial uracil riboside phosphorylase. Arch. Biochem. Biophys. 40:42-49. properties of a pyrimidine deoxyriboside phosphorylase from Escherichia coli. Biochim. Biophys. Acta 28:562-566. 301. Pasternak, C. A., and R. E. Handschumacher. 1959. The 322. Reichard, P. 1949. On the turnover of purines and pyrimibiochemical activity of 6-azauridine: interference with dines from polynucleotides in the rat determined with pyrimidine metabolism in transplantable mouse tumors. Ni5. Acta Chem. Scand. 3:422-432. J. Biol. Chem. 234:2992-2997. 323. Reichard, P. 1959. The enzymic synthesis of pyrimidines. 302. Pastore, E. J., and M. Friedkin. 1962. The enzymatic synAdvan. Enzymol. 21:263-291. thesis of thymidylate. II. Transfer of tritium from tetra324. Reichard, P. 1968. The biosynthesis of deoxyribonucleotides. hydrofolate to the methyl group of thymidylate. J. Biol. Eur. J. Biochem. 3:259-266. Chem. 237:3802-3810. 325. Reichard, P. 1968. The biosynthesis of deoxyribose. John 303. Paulus, H., and E. Gray. 1967. Multivalent feedback inhibition of aspartokinase in Bacillus polymyxa. J. Biol. Chem. Wiley & Sons, Inc., New York. 326. Reichard, P. 1969. Allosteric regulation of ribonucleotide 242:4980-4986. reductase from Escherichia coli, p. 44-59. In H. M. 304. Petersen, R. N., J. Boniface, and A. L. Koch. 1967. Energy Kalckar (ed.), The role of nucleotides for the function requirement, interactions and distinctions in the mechaand conformation of enzymes. Munksgaard, Copenhagen. nism for transport of various nucleosides in Escherichia coli. Biochim. Biophys. Acta 135:771-783. 327. Reichard, P., and S. Bergstrom. 1951. Synthesis of poly305. Pierard, A. 1966. Control of the activity of Escherichia coli nucleotides in slices from regenerating livers. Acta Chem. carbamyl phosphate synthetase by antagonistic allosteric Scand. 5:190-191. effectors. Science 154:1572-1573. 328. Reichard, P., and G. Hanshoff. 1956. Aspartate carbamyl 306. Pierard, A., N. Glansdorff, M. Mergeay, and J. M. Wiame. transferase from Escherichia coli. Acta Chem. Scand. 1965. Control of biosynthesis of carbamyl phosphate in 10:548-566. Escherichia coli. J. Mol. Biol. 14:23-36. 329. Reichard, P., and 0. Skold. 1959. Possible enzymic mecha307. Pierard, A., and J. M. Wiame. 1964. Regulation and mutanism for the development of resistance against fluorouracil tion affecting a glutamine-dependent formation of carin ascites tumours. Nature 183:939-941. in Escherichia Biochem. coli. bamyl phosphate Biophys. 330. Reichard, P., and 0. Skold. 1963. Pyrimidine synthesis and Res. Commun. 15:76-81. breakdown, p. 177-197. In S. P. Colowick and N. 0. R. S. Krooth. on and 1967. the control Studies 308. Pinsky, L., Kaplan (ed.), Methods in enzymology, vol. 6. Academic of pyrimidine biosynthesis in human diploid cell strains. Press Inc., New York. I. Effect of 6-azauridine on cellular phenotype. Proc. Nat. 331. Reissig, J. L. 1960. Forward and back mutation in the pyr-3 Acad. Sci. U.S.A. 57:925-932. regions of Neurospora. L. Mutations from arginine de309. Pinsky, L., and R. S. Krooth. 1967. Studies on the control pendence to prototrophy. Genet. Res. 1:356-374. of pyrimidine biosynthesis in human diploid cell strains. 332. Reissig, J. L. 1963. Spectrum of forward mutants in the pyr-3 II. Effects of 5-azaorotic acid, barbituric acid and pyrimiregion of Neurospora. J. Gen. Microbiol. 30:327-337. dine precursors on cellular phenotype. Proc. Nat. Acad. 333. Reyes, P., and C. Heidelberger. 1965. Fluorinated pyrimiSci. U.S.A. 57:1267-1279. dines. XXVI. Mammalian thymidylate synthetase: its 310. Porter, R. W., M. 0. Modebe, and G. R. Stark. 1969. Asmechanisms of action and inhibition by fluorinated nupartate transcarbamylase. Kinetic studies of the catalytic cleotides. Mol. Pharmacol. 1:14-30. subunits. J. Biol. Chem. 244:1846-1859. 334. Reynolds, E. S., I. Lieberman, and A. Kornberg. 1955. The 311. Prusoff, W. H. 1955. Studies on the mechanism of action of metabolism of orotic acid in aerobic bacteria. J. Bacteriol. 6-azathymine. I. Biosynthesis of the deoxyriboside. J. 69:250-255. Biol, Chem. 215:809-821. 312. Prusoff, W. H. 1957. Studies on the mechanism of action of 335 Roepke, R. R. 1967. Relation between different thymineless 6-azathymine. III. Relationship between incorporation mutants derived from Escherichia coli. J. Bacteriol. into deoxypentose nucleic acid and inhibition. J. Biol. 93:1188-1189. Chem. 226:901-910. 336. Roepke, R. R., and F. E. Mercer. 1947. Lethal and sublethal 313. Prusoff, W. H., and A. D. Welch. 1956. Studies on the mecheffects of X-rays on Escherichia coli as related to the yield anism of action of 6-azathymine. II. Azathymine deoxyof biochemical mutants. J. Bacteriol. 54:731-743. riboside, a microbial inhibitor. J. Biol. Chem. 218:929337. Rogers, H. J. 1944. Importance of pyrimidine derivatives in 939. the growth of group C (Lancefield) Streptococci upon a 314. Pynadath, T. I., and R. M. Fink. 1967. Studies of orotidine simplified medium. Nature 153:251. 5'-phosphate decarboxylase in Neurospora crassa. Arch. 338. Rose, I. A., and B. S. Schweigert. 1953. Incorporation of C14 Biochem. Biophys. 118:185-189. totally labelled nucleosides into nucleic acids. J. Biol. 315. Rachmeler, M., J. Gerhart, and J. Rosner. 1961. Limited Chem. 202:635-645. thymidine uptake in Escherichia coli due to an inducible 339. Roush, A. H., and R. F. Betz. 1958. Purification and propthymidine phosphorylase. Biochim. Biophys. Acta 49:

222-225.

316. Racker, E. 1952. Enzymatic synthesis and breakdown of deoxyribose phosphate. J. Biol. Chem. 146:347-365. 317. Radford, A. 1969. Polarized complementation at the pyrimidine-3 locus of Neurospora. Mol. Gen. Genet. 104:288294. 318. Radford, A. 1969. Information from ICR-170-induced mutations on the structure of the pyrimidine-3 locus in Neurospora. Mutat. Res. 8:537-544. 319. Ratliff, R. L., R. H. Weaver, H. A. Lardy, and S. A. Kuby.

erties of trans-N-deoxyribosylase. J. Biol. Chem. 233:261266. 340. Sander, E. G., L. D. Wright, and D. B. McCormick. 1965. Evidence for function of a metal ion in the activity of dihydroorotase from Zymobacterium oroticum. J. Biol. Chem. 240:3628-3630. 341. Sanderson, K. E. 1970. Current linkage map of Salmonella typhimurium. Bacteriol. Rev. 34:176-193. 342. Saunders, P. P., B. A. Wilson, and G. F. Saunders. 1969. Purification and comparative properties of a pyrimidine

342

O'DONOVAN AND NEUHARD

nucleoside phosphorylase from Bacillus stearothermophilus. J. Biol. Chem. 244:3691-3697. 343. Scarano, E., G. Geraci, and M. Rossi. 1967. Deoxycytidylate aminohydrolase. II. Kinetic properties. The activatory effect of deoxycytidine triphosphate and the inhibitory effect of deoxythymidine triphosphate. Biochemistry 6:192-201. 344. Schmidt, R. G., G. R. Stark, and J. D. Baldeschwieler. 1969. Aspartate transcarbamylase. A nuclear magnetic resonance study of the binding of inhibitors and substrates to the catalytic subunit. J. Biol. Chem. 244:1860-1868. 345. Scocca, J. J., S. R. Panny, and M. J. Bessmann. 1969. Studies of deoxycytidylate deaminase from T4-infected Escherlchia coll. J. Biol. Chem. 244:3698-3706. 346. Shaffer, P. M., R. P. McCroskey, R. D. Palmatier, R. J. Midgett, and M. T. Abbot. 1968. The cell-free conversion of a deoxyribonucleoside to a ribonucleoside without detachment of the deoxyribose. Biochem. Biophys. Res. Commun. 33:806-811. 347. Sheppard, D. E. 1964. Mutants of Salmonella typhimurium resistant to feedback inhibition by L-histidine. Genetics 50:611-623. 348. Shepherdson, M., and A. B. Pardee. 1960. Production and crystallization of aspartate transcarbamylase. J. Biol. Chem. 235:3233-3237. 349. Shih, A., J. Eisenstadt, and P. Lengyel. 1966. On the relation between ribonucleic acid synthesis and peptide chain initiation in Escherichfa colf. Proc. Nat. Acad. Sci. U.S.A. 56:1599-1605. 350. Siedler, A. J., and M. T. Holtz. 1963. Regulatory mechanisms in the deoxyribonucleic acid metabolism of Lactobacillus acidophilus R-26. Deoxycytidylate deaminase activity. J. Biol. Chem. 238:697-701. 351. Siminovitch, L, and A. F. Graham. 1955. Synthesis of nucleic acids in Escherlchla colf. Can. J. Microbiol. 1:721-732. 352. Skoda, J. 1963. Mechanism of action and application of azapyrimidines. Progr. Nucl. Acid Res. Mol. Biol. 2:197219. 353. Skoda, J. 1969. Inhibition of dihydroorotate dehydrogenase in cell-free extract of Escherkchia colf by dihydro-5-azaorotic acid. Collect. Cesk. Chem. Commun. 34:3189-3190. 354. Skoda, J., and F. Sorm. 1958. Accumulation of nucleic acid metabolites in Escherlchla colt exposed to the action of 6-azauracil. Biochim. Biophys. Acta 28:659-660. 355. Skoda, J., and F. Sorm. 1959. The accumulation of orotic acid, uracil, and hypoxanthine by Escherlchla colf in the presence of 6-azauracil, and the biosynthesis of 6-azauridylic acid. Collect. Cesk. Chem. Commun. 24:13311337. 356. Skodovi, H., and J. Skoda. 1969. Mechanism of overproduction of orotic acid by a mutant of Brevibacterlum ammoniagenes. Appl. Microbiol. 17:188-189. 357. Skold, 0. 1960. Uridine kinase from Ehrlich ascites tumor: purification and properties. J. Biol. Chem. 235:3273-3279. 358. Smith, D. W., and P. C. Hanawalt. 1968. Macromolecular synthesis and thymineless death in Mycoplasma laidlawil B. J. Bacteriol. 96:2066-2076. 359. Smith, L. H., C. M. Huguley, and J. A. Bain. 1966. Hereditary orotic aciduria, p. 739-758. In J. B. Stanbury, J. B. Wyngaarden, and D. S. Fredericson (ed.), The metabolic basis of inherited disease. McGraw-Hill, Inc., New York. 360. Somerville, R. B., and C. Yanofsky. 1955. Studies on the regulation of tryptophan biosynthesis in Escherichla coli. J. Mol. Biol. 11:747-759. 361. Stacey, K. A., and E. Simson. 1965. Improved method for the isolation of thymine requiring mutants of Escherichia col. J. Bacteriol. 90:554-555. 362. Stadtman, E. R. 1966. Allosteric regulation of enzyme activity. Advan. Enzymol. 28:141-154. 363. Stanier, R. Y. 1947. Simultaneous adaptation: a new technique for the study of metabolic pathways. J. Bacteriol. 54:339-348.

BACTERIOL. REV.

364. Stryer, L., A. Holmgren, and P. Reichard. 1967. Thioredoxin. A localized conformational change accompanying reduction of the protein to the sulfhydryl form. Biochemistry 6:1016-1020. 365. Suyama, Y., K. D. Munkres, and V. W. Woodward. 1959. Genetic analysis of the pyr-3 locus of Neurospora crassa: the bearing of recombination and gene conversion upon interallelic linearity. Genetica 30:293-311. 366. Taketa, K., and B. M. Pogell. 1965. Allosteric inhibition of rat liver fructose 1 ,6-diphosphatase by adenosine 5'monophosphate. J. Biol. Chem. 240:651-662. 367. Tatibana, M., and J. Ito. 1969. Control of pyrimidine biosynthesis in mammalian tissues. I. Partial purification and characterization of glutamine utilizing carbamyl phosphate synthetase of mouse spleen and its tissue distribution. J. Biol. Chem. 244:5403-5413. 368. Taylor, A. L. 1970. Current linkage map of Escherichia coli. Bacteriol. Rev. 34:155-175. 369. Taylor, A. L., J. R. Beckwith, A. B. Pardee, R. Austrian, and F. Jacob. 1964. The chromosomal location of the structural gene for orotidylic acid pyrophosphorylase in Escherichia coli. J. Mol. Biol. 8:771. 370. Taylor, W. H., and G. D. Novelli. 1964. Enzymes of the pyrimidine pathway in Escherichia coli. I. Synthesis by cells and spheroplasts. J. Bacteriol. 88:99-104. 371. Taylor, W. H., and M. L. Taylor. 1964. Enzymes of the pyrimidine pathway of Escherichla coli. II. Intracellular localization and properties of dihydroorotic dehydrogenase. J. Bacteriol. 88:105-110. ;72. Taylor, W. H., M. L. Taylor, and D. F. Eames. 1966. Two functionally different dihydroorotic dehydrogenases in bacteria. J. Bacteriol. 91:2251-2256. 373. Thelander, L. 1967. Thioredoxin reductase. Characterization of a homogenous preparation from Escherichia coll B. J. Biol. Chem. 242:852-859. 374. Thelander, L. 1968. Studies on thioredoxin reductase from Escherichia colt B. The relation of structure and function. Eur. J. Biochem. 4:407-422. 375. Tomita, F., and I. Takahashi. 1969. A novel enzyme, dCTP deaminase, found in Bacillus subtilis infected with phage PBS 1. Biochim. Biophys. Acta 179:18-27. 376. Udaka, S., and B. Vennesland. 1962. Properties of triphosphopyridine nucleotide-linked dihydroorotic dehydrogenase. J. Biol. Chem. 237:2018-2024. 377. Vitols, E., and R. L. Blakley. 1965. Hydrogen-donor specificity of ribonucleoside triphosphate reductase from Lactobacillus leichmannii. Biochem. Biophys. Res. Commun. 21:466-472. 378. Vitols, E., C. Brown, W. Gardiner, and R. L. Blakley. 1967. Cobamides and ribonucleotide reduction. V. A kinetic study of the ribonucleoside triphosphate reductase of Lactobacillus leichmannii. J. Biol. Chem. 242:3035-3041. 379. Vitols, E., H. P. C. Hogenkamp, C. Brownson, R. L. Blakley, and J. Connellan. 1967. Reduction of a disulphide bond of ribonucleotide reductase by the dithiol substrate. Biochem. J. 104:58-60c. 380. Vogel, H. J. 1961. Aspects of repression in the regulation of enzyme synthesis: pathway-wide control and enzymespecific response. Cold Spring Harbor Symp. Quant. Biol. 26:163-172. 381. Wacher, A., S. Kirschfeld, and L. Trager. 1959. Die Biosynthese der Desoxyribose bei Bakerien. Z. Naturforch. 14b: 145-150. 382. Wachsman, J. T., S. Kemp, and L. Hogg. 1964. Thymineless death in Bacillus megaterium. J. Bacteriol. 87:1079-1086. 383. Wahba, A. J., and M. Friedkin. 1962. The enzymatic synthesis of thymidylate. I. Early steps in the purification of thymidylate synthetase of Escherichia coli J. Biol. Chem. 237:3794-3801. 384. Wang, T. P., and J. 0. Lampen. 1951. The cleavage of adenosine, cytidine and xanthosine by Lactobacillus pentosus. J. Biol. Chem. 192:339-347.

VOL. 34, 1970

PYRIMIDINE METABOLISM IN MICROORGANISMS

385. Wang, T. P., H. Z. Sable, and J. 0. Lampen. 1950. Enzymatic deamination of cytosine nucleosides. J. Biol. Chem. 184:17-28. 386. Watanabe, M. S., R. P. McCroskey, and M. T. Abbott. 1970. The enzymatic conversion of 5-formyl uracil to uracil-5carboxylic acid. J. Biol. Chem. 245:2023-2027. 387. Weber, K. 1968. New structural model of E. coli aspartate transcarbamylase and the amino-acid sequence of the regulatory polypeptide chain. Nature 218:1116-1119. 388. Weed, L. L., and D. W. Wilson. 1951. The incorporation of C14-orotic acid into nucleic acid pyrimidines in vitro. J. Biol. Chem. 189:435-442. 389. Weed, L. L., and D. W. Wilson. 1954. Studies on precursors of pyrimidines of nucleic acid. J. Biol. Chem. 207:439-442. 390. Weitzman, P. D. J., and I. B. Wilson. 1966. Studies on aspartate transcarbamylase and its allosteric interaction. J. Biol. Chem. 241:5481-5488. 391. Westergaard, M., and H. K. Mitchell. 1947. Neurospora. V. A synthetic medium favoring sexual reproduction. Amer. J. Bot. 34:573-577. 392. Wiley, D. C., and W. N. Lipscomb. 1968. Crystallographic determination of symmetry of aspartate transcarbamylase. Nature 218:1119-1 121. 393. Williams, L. G., and R. H. Davis. 1968. Genetic and physical relationship between two early steps of pyrimidine synthesis. Genetics 60:238. 394. Williams, L. G., and H. K. Mitchell. 1969. Mutants affecting thymidine metabolism in Neurospora crassa. J. Bacteriol. 100:383-389. 395. Wilson, M. C., J. L. Farmer, and F. Rothman. 1966. Thymidylate synthesis and aminopterin resistance in Bacillus subtilis. J. Bacteriol. 92:186-196. 396. Woodward, V. W. 1962. Complementation and recombination among pyr-3 heteroalleles of Neurospora crassa. Proc. Nat. Acad. Sci. U.S.A. 48:348-356. 397. Woodward, V. W., and R. H. Davis. 1963. Coordinate

343

changes in complementation suppression and enzyme phenotypes of a pyr-3 mutant of Neurospora crassa. Heredity 18:21-25. 398. Woodward, V. W., K. D. Munkres, and Y. Suyama. 1957. Uracil metabolism in Neurospora crassa. Experientia 13:484-488. 399. Wright, L. D., C. S. Miller, H. R. Skeggs, J. W. Huff, L. L. Weed, and D. W. Wilson. 1951. Biological precursors of the pyrimidines. J. Amer. Chem. Soc. 73:1898-1899. 400. Wright, L. D., K. A. Valentik, D. S. Spicer, J. W. Huff, and H. R. Skeggs. 1950. Orotic acid and related compounds in the nutrition of Lactobacillus bulgaricus 09. Proc. Soc. Exp. Biol. Med. 75:293-297. 401. Yan, Y., and M. Demerec. 1965. Genetic analysis of pyrimidine mutants of Salnonella typhinurium. Genetics 52:643651. 402. Yarus, M. 1969. Recognition of nucleotide sequences. Annu. Rev. Biochem. 38:841-880. 403. Yates, R. A., and A. B. Pardee. 1956. Pyrimidine biosynthesis in Escherichia coli. J. Biol. Chem. 221:743-756. 404. Yates, R. A., and A. B. Pardee. 1956. Control of pyrimidine biosynthesis in Escherichia coli by a feedback mechanism. J. Biol. Chem. 221:757-770. 405. Yates, R. A., and A. B. Pardec. 1957. Control by uracil of formation of enzymes required for orotate synthesis. J. Biol. Chem. 227:677-692. 406. Yeh, Y. C., E. J. Dubovi, and I. Tessman. 1969. Control of pyrimidine biosynthesis by phage T4: mutants unable to catalyze the reduction of cytidine diphosphate. Virology 37:615-623. 407. Zanetti, G., and C. H. Williams. 1967. Characterization of the active center of thioredoxin reductase. J. Biol. Chem. 242:5232-5236. 408. Zimmerman, S. B., and A. Kornberg. 1961. dCDP and dCTP cleavage by an enzyme formed in phage-infected Escherkhia coli. J. Biol. Chem. 236:1480-1486.

Suggest Documents