gene rearrangements in the evolution ofthe tryptophan ... · shared by all bacterial tryptophan...

BACTERIOLOGICAL REVIEWS, June 1975, p. 87-120 Vol. 39, No. 2Copyright 0 1975 American Society for Microbiology Printed in U.S.A.

Gene Rearrangements in the Evolution of the TryptophanSynthetic Pathway

IRVING P. CRAWFORDDepartment of Microbiology, Scripps Clinic and Research Foundation, La Jolla, California 92037

INTRODUCTION.88DESCRIPTION OF THE PATHWAY.88Reactions.88Regulation of Metabolite Flow.89

Feedback inhibition.89Repression of enzyme synthesis.89

Early Evolutionary Speculations.90ENTEROBACTERIACEAE.90Operon Structure. go

Genetic studies.90Messenger RNA studies.91Regulatory mutations.91Polarity and antipolarity.92

Deviations from Coordinacy.93Subunit Structure of Enzymes and Amino Acid Sequences ....... ............. 93Organisms Unable to Synthesize Tryptophan.95

OTHER GRAM-NEGATIVE BACTIERIA.95Pseudomonadaceae.95

Multiple operon trp gene arrangement.95Subunit structure of enzymes and amino acid sequences.96Regulation of enzyme synthesis.97

Acinetobacter.97Multiple operon arrangement of genes.97Subunit structure of enzymes.98Regulation of enzyme synthesis.98

GRAM-POSITIVE BACTERIA.............................. 99Bacillus and Clostridium.99

Chromosomal arrangement.99Subunit structure of enzymes and amino acid sequences.100Regulation of enzyme synthesis.100

Staphylococcus and Micrococcus.101Chromosomal arrangement.101Subunit structure and regulation of enzymes.101

Streptomyces.101Chromosomal arrangement.101

BACTERIA STUDIED NON-GENETICALLY.102Chromobacterium, Lactobacillus, and Blue-Green Bacteria.102

Subunit structure and regulation of enzymes.102EUKARYOTIC ORGANISMS.102Fungi.102

Chromosomal arrangement.102Subunit structure of enzymes.103Regulation of enzyme synthesis.104

Algae, Euglenoids, and Plants.105Subunit structure ofenzymes.105Regulation of enzyme synthesis.105

GENERAL CONCLUSIONS ABOUT EVOLUTION OF THEPATHWAY ........ 105Conservation of Reaction Mechansims.105Variation in Chromosomal Disposition of Genes.106

Translocation and gene fusion.106Regulatory mechanisms.106

Enzymes Shared with Other Pathways.107NATURAL RELATIONSHIPS REVEALED BY COMPARATIVE STUDIES ...... 107Amino Acid Sequences .....I...... . ......... . . ....I..................107Nucleotide Sequences and Nucleic Acid Hybridization .. . 108

87

on March 13, 2021 by guest

http://mm

br.asm.org/

Dow

nloaded from

http://mmbr.asm.org/

CRAWFORD BACTERIOL. REV.

Subunit Interchange.108Immunological Cross-Reactivity.109Regulatory Mechanisms.110

EVOLUTIONARY TRENDS.110Fused versus Separate Genetic Elements.110Dependent Enzyme Aggregates versus Independent Ones.110One Operon or Several.110Predictions Concerning Future Studies.110

LITERATURE CITED ............................................... 110

INTRODUCTIONKnowledge of the natural relationships of the

major bacterial families is clearly still in itsinfancy (119). It is not clear what effect, if any,the more recent molecular taxonomic ap-proaches, exemplified in higher organisms bysequence comparison of proteins like cyto-chrome c (144), may have in the maturation ofthis subject. Since bacteria are predominatelyhaploid and have a short generation time, it isnot at all obvious that a meaningful outline ofbacterial evolution will result from applicationof strategies that have been successfully em-ployed with eukaryotic organisms.This review is an attempt to summarize still

fragmentary data on the comparative biochem-istry and genetics of the tryptophan syntheticpathway, emphasizing work with various bacte-ria but not excluding eukaryotes. From theseresults one may develop a case for commonancestry for several of the major gram-negativeand gram-positive bacterial taxa. This in turnsuggests that the rate of accumulation of muta-tional substitutions and chromosomal rear-rangements during evolution of this pathwayhas not been too rapid to obscure the naturalaffinities of the main bacterial families.

DESCRIPTION OF THE PATHWAY

ReactionsThe de novo synthesis of tryptophan proceeds

from chorismic acid (73), the last intermediatecommon to the aromatic amino acids and vita-mins, through a series of reactions that has notvaried in those prokaryotes and eukaryotesstudied (Fig. 1). The first reaction is an enol-pyruvyl elimination accompanied by a gluta-mine amidotransfer and is catalyzed by theenzyme anthranilate synthase (AS; EC4.1.3.27). In all organisms so far examined thiscomplex enzyme contains at least two noniden-tical subunits (225). As with other glutamineamidotransferases, under certain conditionsammonia can replace glutamine in the reactiongiving anthranilate and pyruvate as products.This version of the reaction can usually be

catalyzed by a single subunit of the enzyme andwill be referred to as the chorismate "amina-tion" reaction.The second step in tryptophan synthesis is

the addition of the phosphoribosyl moiety of5-phosphoribosyl-1-pyrophosphate to the 3-position of anthranilate, catalyzed by anthrani-late phosphoribosyltransferase (PRT; EC2.4.2.18). Although certain members of theEnterobacteriaceae have the enzymes for thefirst and second steps associated in an enzymecomplex (11, 61, 106), this situation is not seenoutside the enteric family and is not true evenfor some species within the family (96, 125a,225).The third step is an Amadori rearrangement

of phosphoribosylanthranilate catalyzed byphosphoribosylanthranilate isomerase (PRAI),and the fourth step is a decarboxylation andring closure catalyzed by indoleglycerol phos-phate synthase (InGPS; EC 4.1.1.48). In all theenteric bacteria studied to date, the activitiesfor both these steps are associated with a singlepolypeptide chain (125a, 139). In other bacteriatwo smaller proteins coded by separate genescarry out these functions.The final step is the removal of the glycerol-

phosphate side-chain from indoleglycerol phos-phate (InGP) and its replacement by the alanylmoiety of L-serine, catalyzed by the complexenzyme, tryptophan synthase (TS; EC 4.2.1.20)(216). The earliest known microbial multicom-ponent enzyme, this protein has been exten-sively studied, especially in Escherichia coli,and often reviewed (41, 118, 216). In entericbacteria, each molecule consists of four non-covalently bound polypeptides, two a-chainsprimarily concerned with InGP aldolysis (pro-ducing indole and glyceraldehyde-3-phos-phate), and two ,B-chains, each binding a pyri-doxal-5'-phosphate cofactor molecule and pri-marily concerned with the synthesis of trypto-phan from indole and serine. The enzyme com-plex normally functions without releasing in-dole into solution (long arrow in Fig. 1), butgiven the proper conditions and choice of sub-strates, the half-reactions of aldolysis (TS-A,

88


http://mm

br.asm.org/

Dow

nloaded from


TRYPTOPHAN PATHWAY EVOLUTION

WCOON

PM~I0~~~~~~~~H

0 Hz

FIG. 1. Biosynthetic pathway for tryptophan. Ab-

breviations: AS, anthranilate synthase (amidotrans-

ferase reaction is shown); PRPP, 5-phosphoribosyl

pyrophosphate; PP, pyrophosphate; PRT, anthrani-

late phosphoribosyltransferase; PRA, N-phos-

phoribosylanthranilate; PRAI, phosphoribosylan-

thranilate isomerase; CDRP, 1-(o-carbQxy-

phenylamino)-1-deoxyribulose phosphate; InGPS, in-

doleglycerol phosphate synthase; InGP, indole-

glycerol phosphate; Gly-3-P, D-glyceraldehyde-

3-phosphate; TS, tryptophan synthase; TS-A and

TS-B, see text.

reversible arrows in Fig. 1) or tryptophan syn-

thesis from indole (TS-B, short arrow in Fig. 1)

can be observed at excellent rates. With minor

variations, these general characteristics are

shared by all bacterial tryptophan synthasesstudied.

One unusual feature of the tryptophan path-

way is that it contains both phophhorylated andnon-phosphorylated intermediates. Two of thelatter, anthranilate and indole, are freely

permeable to most bacterial and fungal cells.

Thus, tryptophan auxotrophs lacking only the

first enzyme of the pathway can grow on anthra-

nilate or indole in place of tryptophan. Simi-

larly, auxotrophs with a normal TS, or even anabnormal one lacking TS-A activity but retain-

ing TS-B activity, can grow on indole. (It may

be noted here that although indole is normallyan enzyme-bound intermediate, TS either invitro or in vivo utilizes it in preference to InGP

[46, 50], probably effecting significant meta-

bolic economies thereby).Auxotrophs can often be characterized by the

accumulation of compounds antecedent to their

defective enzyme. TS mutants may accumulate

InGP, indole-glycerol, indole, or a mixture of

these, depending on their ability to perform theTS-A half-reaction; InGPS mutants accumu-

late the dephosphorylated form of the substrate

of this reaction along with some anthranilate.

Only mutants in steps 2 and 3, the PRT and

PRAI reactions, cannot easily be distinguishedfrom each other without an enzyme assay, forthe lability of phosphoribosyl anthranilatecauses mutants in either step to accumulateonly anthranilate.

Regulation of Metabolite FlowFeedback inhibition. Tryptophan may par-

ticipate with other substances in the feedbackregulation of chorismic acid formation, by inhi-bition of 3-deoxy-D-arabino-heptulosonate-7-phosphate synthesis (71), and glutamine for-mation, by influencing glutamine synthase(213), but the major point of feedback controlfor the amino acid is on the enzyme catalyzingthe first tryptophan-specific reaction, AS(165a,196b). As this topic has been reviewed recently(71, 146, 225), little will be said about it hereexcept to note that notwithstanding the diver-sity in molecular architecture of this enzymeobserved throughout the prokaryotic and eu-karyotic kingdoms, all forms of the enzyme havein common an effective feedback inhibitionmechanism that operates equally well on theamidotransferase and amination reactions.Therefore, the allosteric binding site for trypto-phan must be on the large, chorismate-bindingsubunit (196b), and it has been evolutionarilywell conserved, all the way up to higher plants(14).Repression of enzyme synthesis. Very early

in the study of the tryptophan pathway inenteric bacteria, it was found that exogenoustryptophan represses the synthesis of TS (155).On the other hand, auxotrophs grown to trypto-phan limitation are derepressed more than10-fold above the wild-type level in the synthe-sis of any tryptophan pathway enzyme notablated by the mutation (214, 215). Together,repression and derepression allow the specificactivity of tryptophan enzymes in enteric bacte-rial extracts to vary over two orders of magni-tude. In these bacteria the structural genes forthe tryptophan pathway enzymes are clusteredin one operon. The five gene products varycoordinately (105), and corresponding changesin the level of the polycistronic trp messengerribonucleic acid (mRNA) are seen (101). TrpR,the unlinked structural gene for the tryptophanrepressor, has been well characterized (38, 159),and it is clear that tryptophan and not trypto-phanyl-transfer RNA (tRNA) is the effector inthis negative transcriptional control system(160, 197). Not all bacterial groups employ thisregulatory system, however. Some, such as thepseudomonads, have different controls for dif-ferent parts of the pathway (42). Others, such asChromobacterium violaceum, seem completelyunable to alter the rate of synthesis of the trpenzymes in response to end product scarcity orexcess (204). Repression of enzyme synthesis is,as we will see, a variable that is related tochromosomal location of the structural genes. It

89VOL. 39, 1975


http://mm

br.asm.org/

Dow

nloaded from


BACTERIOL. REV.

thus seems a better source of information onnatural relationships than the study of controlby feedback inhibition, at least as far as con-cerns the tryptophan pathway.

Early Evolutionary SpeculationsSoon after it was determined that the TS of

enteric bacteria dissociates into dissimilar sub-units whereas that of the ascomycete, Neuro-spora crassa, does not (51), Bonner et al. (17),impressed by the enzymological and geneticsimilarities shown by the two types of TS,speculated that gene fusion may have takenplace at some time during the evolution of thefungi from their presumed prokaryotic progeni-tor. They recognized that there were discordantaspects to the situation; even then it was knownthat the green alga, Chlorella ellipsoidea, has adissociable TS, and that the gene-enzyme rela-tionships of the early enzymes of the tryptophanpathway in Neurospora were not, and still arenot, wholly explainable on the basis of anevolutionary trend towards fusion of geneticelements of related function in eukaryotes.Nevertheless, their speculations concerninggene fusion now seem prescient in view of theresults of the subsequent ten years. As we willsee, gene fusions have certainly occurred duringevolution of the tryptophan pathway, thoughthe trend of evolution from prokaryotes to thehigher eukaryotes has not necessarily alwaysfavored the condensation of genes.

ENTEROBACTERIACEAE

Operon StructureGenetic studies. All members of the enteric

bacterial family studied so far seem to have thegenes for the tryptophan pathway clustered inone operon on the chromosome. The geneticevidence for this in E. coli (221) and Salmonellatyphimurium (49) was first obtained throughstudy of deletions and by co-transduction ofstructural gene point mutations with each otherand with neighboring cysB mutants. Less directevidence for gene clustering is available forSerratia marcescens (96), Aeromonas formicans(43), and several species of the genera Citrobac-ter, Enterobacter (Aerobacter), Erwinia, andProteus (125, 125a). Most of this indirect evi-dence rests on coordinate regulation of thetryptophan enzymes, evidence of polarity andantipolarity (see below), and indications of thesame low-level, operon-internal promoter foundin the E. coli and S. typhimurium operons (10,157). Although the number of enteric bacterialspecies examined so far is still small, the sampleseems representative, including as it does spe-

cies from each of the three major taxonomicsubdivisions, Escherichia-Salmonella-Shigella-Citrobacter (G+C content, 50 to 52%), En-terobacter-Serratia-Klebsiella-Erwina (G +Ccontent, 54 to 59%) and Proteus-Providencia(G+C content, 39 to 50%), as well as thetaxonomically controversial Aeromonas. Directgenetic confirmation of trp gene clusteringshould be available for many of these speciessoon, for techniques of genetic exchange in allthe enteric bacteria seem to be developingrapidly.The trp gene order appears to be identical in

all enteric bacterial trp operons. This is illus-trated in the top two lines of Fig. 2, by using a"universal" nomenclature (187) for the trpgenes rather than the varied ones that wereproposed originally for different organisms.TrpE is closest to the promoter-operator region,is transcribed first, and encodes the large sub-unit of AS. Auxotrophic mutants in trpE areunable to perform either the glutamine amido-transferase version or the amination version ofthe AS reaction. TrpD is a complex gene in bothE. coli and S. typhimurium, having its proximalthird devoted to the glutamine amidotransfer-ase AS function and its distal two thirds to the

BACTERIAL GROUP CHROMOSOMAL DISTRIBUTION OF trp GENES

Enteric Bacteria

Escherichio, elc.SerrOfi7, etc.

Other Gram NegativePseudomonas

Acdnetobacter

Gram PositiveBacillus

Staphy/ococcusMicrococcus

Streptomyces

* E (G)D . CIF B A

.4R G. DCD(F) B A .It

R E .D, CFB ,A,.=4 =.=$

D F

FIG. 2. Tryptophan gene distributions in bacteria.A universal nomenclature (187) is used; analogousgenes have the same letter designation in each orga-nism. The trp prefix for all gene symbols has beenomitted for simplicity. Gene-enzyme key: trpA, TSa-chain; trpB, TS f-chain; trpC, InGPS (except inenteric bacteria where the bifunctional PRAI-InGPSis produced); trpD, PRT (except in enteric bacteria ofthe Escherichia type where this activity and theglutamine amidotransferase function for AS are com-bined); trpE, AS large subunit; trpF, PRAI; trpG, ASsmall, glutamine amido-transferase subunit; trpR,tryptophan repressor. Not shown: trpS, tryptophanyl-tRNA synthase. In Acinetobacter the gene order inthe three-gene clusters is not certain; the most proba-ble order is shown. A question mark following a genedesignation indicates doubt concerning its true loca-tion.

mo CRAWFORD


http://mm

br.asm.org/

Dow

nloaded from



PRT reaction. There is reason to believe (see"Subunit Structure" below) that these twosegments of the trpD gene are two separatecistrons in Serratia and Aeromonas, so a moreprecise designation for the fused element inEscherichia and Salmonella might be trp(G)D.TrpC in all enteric bacteria is also a complexcistron, encoding in its proximal half the InGPSfunction and in its distal half the PRAI func-tion. Missense mutants normally lose only oneof the two activities, depending on their location(195). Other bacteria have separate cistrons forthese two activities. We have called the gene forPRAI in other organisms trpF, so an unambigu-ous designation for trpC in enteric bacteriamight be trpC(F). TrpB, coding for the ,-chainof TS, and trpA, coding for its a-chain, completethe operon. There are apparently no otherstructural elements associated with the enterictrp operon, for the messenger RNA size corre-sponds fairly closely to the length of deoxyribo-nucleic acid (DNA) required to encode thesefive polypeptide chains (180). There is, how-ever, a transcribed "leader" region of some 160base pairs lying between the operator and trpEthat appears to be of regulatory significance(21a, 110).The trp operon has opposite orientations in E.

coli and S. typhimurium (185), being tran-scribed in a counterclockwise direction in theformer organism and clockwise in the latter, asthe map is usually drawn. This difference re-sults from its being situated within an inversionnow estimated to involve about 10% of thechromosome (31). There is no evidence that thisinversion affects the transcription or translationof the trp genes in any significant way. Othermembers of the enteric bacterial group have notyet been examined for the presence of thisinversion, which extends from 24 to 34 min onthe E. coli map (31).mRNA studies. The isolation of specific

transducing lambdoid phages carrying part orall of the E. coli tryptophan operon on theirchromosome has permitted detailed study ofthe mRNA transcribed in vivo or in vitro fromthis gene cluster (101, 179). Regulation of thelevel of the tryptophan enzymes in the cell isaccomplished primarily by varying the numberof mRNA transcripts of the operon. The detailsof synthesis and degradation of this typicalpolycistronic mRNA molecule have been re-cently reviewed (100) and are beyond the scopeof this article. These details are not known todiffer in any fundamental way in the severalenteric bacterial species that have been studied.Of some interest is the occurrence of a second,low-level promoter within the distal third of the

trpD structural gene (10, 108, 157). The activityof this second promoter is revealed when theoperon is repressed and transcription beginningat the normal promoter is inhibited. Underthese conditions (but not normally, see refer-ence 105), enzyme production is not coordinate,and the gene products of the last three genes areproduced at five times the basal rate of the firsttwo genes. Whether this is of selective advan-tage to the organisms is unknown, but it iscommon to all enteric bacterial species in whichit has been sought (125). Other, high efficiencypromoters can be created within several of thestructural genes of the operon (24, 138, 158,207), frequently at the expense of the functionalintegrity of the polypeptide normally encodedthere (24, 138, 158).The hybrid phage-bacterial chromosomes

used as molecular probes in detection andmeasurement of mRNA molecules in E. coli canbe employed with other species as well. Whenthese E. coli probes are used with RNA fromfive members of the enteric group (55), mRNAmolecules of similar size but diminished hybrid-izing efficiency are found. Their weakened af-finity for E. coli DNA, obviously due to base-pair mismatching, can be quantitated and com-pared with the amino acid sequence differencesknown for one region of the operon, the gene forthe TS a-chain. When these measurements aremade (132) the results show, as expected, aninverse relationship between the strength ofhybridization and the number of amino aciddifferences. By calculation it appears, however,that at least twice as many base changes haveoccurred and been "fixed" as would be requiredto elicit the amino acid substitutions. Obviously"silent" base changes to synonymous codonscan be evoked in explanation, and this illus-trates how certain bacterial relatives having asimilar chromosome organization and closelyrelated amino acid sequences (see below) mayyet have quite different DNA base ratios. It alsoexplains why nucleic acid hybridizationmethods, useful though they may be in assign-ing various strains to certain taxonomic groups,are often too sensitive to be effective in measur-ing the relatedness of different bacterial fami-lies. Denney and Yanofsky (55) could see nosignificant hybridization between E. coli trpDNA and mRNA from Bacillus subtilis, forexample, though it is easy to recognize a rela-tionship in the amino acid sequence of their TSa-chains (134).Regulatory mutations. Selection for resist-

ance to tryptophan analogs such as 5-methyl-tryptophan very often results in the appearanceof constitutive or partially constitutive strains

VOL. 39, 1975 91


http://mm

br.asm.org/

Dow

nloaded from


BACTERIOL. REV.

that coordinately overproduce all the enzymesof the tryptophan pathway (8, 38, 84, 159).These mutants are usually found either in thetrpR locus near thr at 90 min on the E. coli mapor in the operator region (trpO) of the trypto-phan operon at 27 min (8, 84). As expected forthe structural locus for a repressor protein,mutations in trpR conferring constitutivity maybe either missense or nonsense and are recessiveto the wild-type form of the gene (159). Simi-larly, as one would predict, constitutive mu-tants located in the operator region of the trpoperon are cis-dominant in merodiploids (8, 84).Although it was first estimated by recombina-tional mapping that the operator region wasrather large (84), more recent results suggestthat it consists of a normal-sized repressorrecognition region plus the "leader" segmentmentioned earlier (21a, 110). Nucleic acid basesequence studies are under way in this regionand should be available in the near future.The protein product of the trpR gene has a

molecular weight of about 58,000 (194, 197),does not appear to require interaction withtryptophanyl-tRNA or tryptophanyl-tRNA syn-thase for function (160, 197), and in the absenceof its effector, tryptophan, binds to at least oneother operator site in addition to trpO, the oneaffecting the tryptophan-sensitive isozyme of3-deoxy-D-arabino-heptulosonate-7-phosphatesynthase at 32 min on the E. coli chromosome(71).Mutations in trpS, the gene for tryptophanyl-

tRNA synthase at 66 min on the E. coli chromo-some, indirectly produce a repressing effect onthe tryptophan operon because mutants thatpermit some growth and protein synthesis usu-ally accumulate a large intracellular pool oftryptophan (56, 71). A suggestion that the trpSgene product interacts with the trpR repressor(107) has not been confirmed in in vitro studies(197).

In addition to yielding constitutive mutationsat the trpR and trpO sites, selection for resist-ance to 5-methyltryptophan in E. coli mayproduce mutants having fairly normal repres-sion and derepression behavior; these mutantsare linked to argG at 61 min and have beendesignated mtr for methyltryptophan resistance(85). The mechanism of their resistance remainsunknown, though the three most interestingsites of interaction of tryptophan analogs withthe tryptophan synthetic pathway, false feed-back repression of AS, false repression of the trpoperon, and inhibition of tryptophanyl-tRNAsynthetase, are clearly unaffected in these mu-tants. In addition to the classes of tryptophananalog-resistant mutants described above, S.

typhimurium exhibits yet another class ofconstitutive mutants unlinked to either the trpoperon, the trpR gene, the trpS gene, argG, orthe aromatic amino acid permease (8). Thesemutants have been tentatively designated trpT,although their map position is unknown andthey may not even define a single locus.Recently a new tool for study of the regulation

of tryptophan genes in enteric bacteria has beenreported by Hershfield et al. (83). They wereable to study the function of a normal E. coli trpoperon inserted artificially into the DNA of theCol El plasmid. In this way they ascertainedthat the single trpR gene normally producessufficient repressor molecules to inactivateabout 30 trp operators. It will obviously be ofinterest to study the function of such "extra-chromosomal" trp operons in the presence ofsome of the regulatory mutations describedabove and to transfer operons from one speciesto another by using this technique.

Polarity and antipolarity. As very little isknown on this subject outside the Enterobac-teriaceae, the molecular mechanism is still indoubt (100), and as the subject has been compe-tently reviewed by Margolin both as a phenom-enon (145) and specifically with respect to thetrp operon (146), there would seem little neces-sity for an extensive coverage of polarity here. Itis a dramatic regulatory event, however, andcan serve to establish the direction of transcrip-tion and translation as well as the extent of anoperon. As such it could be useful in character-izing units functioning as operons when genes ofthe pathway are disassociated, as is the case inmost bacteria outside the enteric group. Theoccurrence of polarity suppressor mutations,such as suA in E. coli (189), having the effect ofabolishing the polarity phenomenon anywhereon the chromosome, opens up the possibilitythat this phenomenon may not be universal indifferent bacterial taxa, however. More infor-mation on this point would be very valuable.

Early studies of nonsense mutations in the E.coli trp operon showed that when these arelocated proximally in a structural gene (i.e., atthe end nearest the trp operator) they have adepressing effect on the amount of proteinproduced from all the more distal genes of theoperon (105, 150, 219). The polarity "gradient,"the intensity of this effect as a function ofposition of the mutation within the gene, hasbeen accurately determined for all of the rele-vant genes of the tryptophan operon, trpE,D, C,and B (218). Similar but less extensive dataexist for the trp operon of S. typhimurium (11,16). In each case polarity is more drastic withtrpE nonsense mutants than with mutants in

92 CRAWFORD


http://mm

br.asm.org/

Dow

nloaded from



other genes. Frameshift or deletion mutantsthat result in premature termination of transla-tion followed by an appreciable length of un-

translated message are as polar as analogousnonsense mutations, and polarity is associatedwith absence (100) or hyperlability (86) of thatportion of the polycistronic mRNA distal to thesite of the nonsense mutation. The precisemechanism of polarity and polarity suppressionremains controversial, however (100, 120).

Early studies of polarity in the trp operon alsorevealed an apparent decrease in function of thegene immediately proximal to that occupied bythe nonsense mutant as well as distal ones.

Unlike the distal polarity effects, this was

usually a moderate diminution, and it never

extended back more than one or at most twogenes (7, 105, 220). It was termed a "short rangeantipolar" effect, and was noticed with bothpoint mutants and deletions. The degree ofantipolarity seen with nonsense mutants intrpA was proportional to their location in thegene (220), just as is the case with polarmutants in the rest of the operon (218). Morerecent work has failed to demonstrate signifi-cant antipolarity from nonsense mutants in anygene but trpA, the last gene of the operon (218).When sought in the second and third genes ofthe nine-gene histidine operon of S. typhimuri-um, no antipolarity was found (68). Thus, thestatus of antipolarity as a regulatory phenome-non is anything but clear. Ito (104a) showedthat during tryptophan starvation genes nearestthe operator are preferentially expressed. Var-ied results obtained in different studies may

reflect variations in strains, growth conditions,and techniques used for derepression. Factorssuch as these have also been evoked as reasonsfor discrepancies in studies of polarity andmRNA production (100).Although the mechanisms bringing about

polar and antipolar effects are still in doubt, itshould not be assumed that these phenomenaare inconsequential. In the enteric bacteria a

polar mutation may drastically impair the abil-ity of a strain to use anthranilate or indole as a

precursor of tryptophan, though the genes forthe necessary enzyme are intact and can bereleased from their functional stranglehold byany of several direct or indirect genetic events(7, 24, 100, 120, 158, 207).

Deviations from CoordinacySeveral deviations from coordinate produc-

tion of the tryptophan operon gene productshave already been mentioned. Though the fivegene products are produced in approximate

equimolarity under conditions of moderate orextreme derepression, the last three enzymesare produced at about five times the basal rateof the first two during full repression (10, 157)because of the second, low-level promoter intrpD. Mutational events within the operon mayabolish coordinacy, as when a new, high-levelpromoter is created in an internal gene (138) ora polar nonsense mutation occurs (16, 101, 105,218, 219). An interesting additional example ofnoncoordinate synthesis is seen when an ordi-nary auxotroph such as a missense or nonpolarnonsense mutant is starved for tryptophan(104a, 196a, 219). Under these circumstancesthe trpE and trpA products, and to a lesserextent the trpD product, continue to be synthe-sized (albeit slowly) after the total cell massstops increasing due to tryptophan limitation,but the amount of trpC and trpB gene productsincreases little if at all. The reason for this lackof coordinacy is not known with certainty, but itis possible that the few molecules of tryptophanthat do become available under these circum-stances (from proteolysis, for example) are pref-erentially used by ribosomes translating theproximal genes. Since the a-chain of TS lackstryptophan, ribosomes beginning at the trpAinitiation site can synthesize the entire geneproduct, however, Some support for this hy-pothesis comes from the work of Brammer (21),who showed that a-chains containing trypto-phan at residue 212 as a result of mutation arenot synthesized at all during tryptophan starva-tion.

Subunit Structure of Enzymes and AniinoAcid Sequences

Primary sequence analysis of the a-chain ofTS has been accomplished for three entericbacterial species (76, 136, 137), and partial,amino-terminal sequences are available for twoothers (133, 135). These data are summarized inFig. 3. It is apparent that S. typhimurium andEnterobacter aerogenes both differ from E. coliin about 15% of their residues, and they differfrom each other to a similar extent. S.dysenteriae differs from E. coli in only one of thefirst 50 residues analysed, whereas S.marcescens is somewhat more distant from E.coli than S. typhimurium and E. aerogenes,with nine changes in the first 28 residues. It isworth noting in passing that none of theseamino acid substitutions affect the nine siteswithin the polypeptide chain where inactivatingmissense mutations have been found in E. coli(217). There is a noticeable clustering of evolu-tionary substitutions between residue 243 and

93VOL. 39, 1975


http://mm

br.asm.org/

Dow

nloaded from


A.

Residue 2 6 12 13 15 24 30 39 42 43 52 56 66 68 91 97 104 108

E. coli Gin Ser Lys Glu Lys Thr Ile Thr Glu Ala Ile Asp Asn Thr Lys Ile Asn Asn

S. typhimurium Glu Asn Asn Asp Arg + + + Asp + Val Asn + Asn Asn Val Lys SerE. aerogenes Gln Thr Lys Lys + Ile Thr Ala Glu Gly Ile + Gly Ala + Ile + +

109 113 114 117 119 120 134 144 148 166 180 189 194 197 198 201 204 208 216 218

Lys Glu Phe Gin Glu Lys Gin Leu Val Ile Ala Ala Asn Val Ala Lys Asn Pro Ala Asp

Pro + Lou Arg + Gin + + Ile Val Ser Gly His Ile Glu + His Ala Ser Glu

+ Ala Phe Gin Ala Arg Glu Met + Ile Ala Ala + Val + Ala + Pro Ala 4

221 225 243 244 245 246 247 249 250 253 254 256 257 260 261 262 266 267 268 269

Lys Asp Gin His Asn Ile Glu Glu Lys Ala Ala Lys Val Gin Pro Met Thr Arg Ser &Val Arg Lys Asn Lou Ala Ser Lys Gin + Glu Arg Ser Ser Ala + Ser + Ala iSer Asp Arg His + Asp Glu Gln Thr Asp + Lys Ala Gin Ser Leu Thr Lys Thr Ala

B.

Residue 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28

E. coli Met Gln Arg Tyr Glu Ser Leu Phe Ala Gin Leu Lys Glu Arg Lys Glu Gly Ala Phe Val Pro Phe Val Thr Leu Gly Asp Pro

S. dysenteriae Glu + + + + +4' +

S. marcescens Glu Gin Gin Lys Arg Glu Ser Asn Lys j IleP. putida A Ser Lou Glu 4' Arg Ala Glu Lys Ala Glu Gly Arg Ser i Lou Ile. Ala

B. subtilis A S A A a A Met Lys Lou Asp Lou Gin Ala Ser Glu Lys Lou Phe Ile Pro + Ile +4

Residue 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52

E. coli Gly Ile Glu Gin Ser Leu Lys Ile Ile Asp Thr Lou Ile Glu Ala Gly Ala Asp Ala Leu Glu Leu Gly IleS. dysenteriae + 1 4 4 4+ + + + +i X X

P. putida I Tyr Asp Ala Gin Lou Lys Gly Pro Ala Val Ile Gly X

B. subtilis Ser Pro Glu Val Val Lys Lou Ala Pro Ser Glu Val 4 Val Thr Lys Lou + Val

FIG. 3. Amino acid sequences for the a-chain of tryptophan synthase. (A) Positions of divergence in the totalsequences of E. coli (268 residues), Salmonella typhimurium (268 residues), and Enterobacter aerogenes (269residues). (B) Positions of divergence in the N-terminal portions of E. coli (52 residues shown), Shigelladysenteriae (50 residues known), S. marcescens (28 residues known), Pseudomonas putida (50 residues known),and B. subtilis (46 residues known). Symbols: x, unknown amino acid; A, absence of an amino acid; and 1'identical with the residue above.

the end of the chain, suggesting a lesser degreeof structural constraint near the carboxy-ter-minus of the protein.As yet only a portion of the primary structure

of the #-chain of E. coli TS is known, and no

sequences from other enteric bacteria are avail-able for comparison. It has been known for a

long time that a and :3 chains from E. coli, S.typhimurium, and S. marcescens are inter-changeable and active in any combination (6),though the affinity of the S. marcescens enzymeis somewhat less for a complementary subunitfrom E. coli or S. typhimurium than for its ownhomologous one (178). An immunological ap-proach to measuring the evolutionary diver-gence of various enteric bacterial TS moleculesgave results in accord with the sequences of thea-chain (161, 178).Although sequences are not yet available for

the trpC gene product, a single polypeptide of45,000 molecular weight with both PRAI andInGPS activities, the proteins from E. coli, S.typhimurium, and E. aerogenes have been puri-fied to near homogeneity (139). Trypsin pluschymotrypsin peptide patterns of these threemolecules were examined and found to be aboutas similar in appearance as corresponding mapsof the TS a-chains from the same species (139).

The enzyme from S. marcescens is also very likethe above in its activities and molecular weight(96).As described earlier, the AS and PRT activi-

ties reside in a single, large complex enzyme inE. coli, Salmonella sp., E. aerogenes, and Er-winia dissolvens, whereas they are separate inProteus sp., Erwinia carotovora, Erwinia haf-niae, Serratia sp., Enterobacter liquefaciensand A. formicans (125a). From genetic andbiochemical evidence, several authors (75, 226;reviewed in reference 225) have proposed thatthe E. coli type of enzyme has resulted fromfusion of two polypeptides that are separate inthe other species, a small one of 20,000 molecu-lar weight that binds to and provides the gluta-mine amidotransferase function for AS, and the45,000 molecular weight PRT enzyme. Directevidence for this hypothesis was provided by Liet al. (133a), who sequenced the amino-terminalregions of the fused protein (trpD gene product)from E. coli and S. typhimurium and the smallAS subunit from S. marcescens. The former twowere very similar, as expected, differing only inthree of the first 53 residues, but homology wasquite apparent in the latter also, which differedfrom the E. coli sequence in only 14 of the first61 residues. This work provides documentation

94 CRAWFORD BACTERIOL. REV.


http://mm

br.asm.org/

Dow

nloaded from



for the occurrence at least once during theevolution of the Enterobacteriaceae of a genefusion similar in result to that created in thelaboratory by the fusion of two his genes byYourno et al. (223).As we will see later, the trpC gene product

also appears to be a fusion product of twosmaller genes that are separate in other bacte-rial taxa. Since this fused, bifunctional poly-peptide is invariant in all theEnterobacteriaceae studied to date, fusionseems to have taken place in a precursor tomany or all of the surviving forms assigned tothe group. We will also see later that gene fusionhas apparently not been confined to the entericbacterial group, but can also be deduced to haveoccurred during the evolution of fungi.

Organisms Unable to Synthesize TryptophanFor over forty years it has been known that

most or all strains of Salmonella typhosa re-quire tryptophan for growth but can give rise totryptophan-independent variants (67). Thesame is apparently true for some strains ofSalmonella cholerasuis and Salmonellaparatyphi A (126). In all three Salmonellaspecies, this tryptophan requirement is oftenassociated with auxotrophy for the sulfur-con-taining amino acids cysteine or methionine(126). Many, perhaps most, strains of S.dysenteriae also require tryptophan, cysteine,and sometimes other amino acids (115). It wasthe observation of Fildes (66) that indole couldreplace tryptophan as a growth supplement forS. typhosa that led to the first use of tryptophanauxotrophs in various microbes as subjects forgene-enzyme investigations.The only relatively recent investigation of a

naturally occurring defect in the tryptophansynthetic machinery of an enteric bacterium isthat of Eisenstein and Yanofsky (62) usingthe avirulent S. dysenteriae strain Sh/S. Theywere able to show a partial requirement (brady-trophy) for tryptophan traceable to the firstgene of the operon; this requirement was satis-fied equally well, of course, by anthranilate orindole. This laboratory line of Shigella quiteobviously has undergone some adaptations.Like many of the auxotrophic pathogens men-tioned above, however, it gives rise to trypto-phan prototrophs by reversion. When its brady-trophic character was transduced into E. coli, itresulted in derepression during growth on mini-mal medium of TS and perhaps the trpD andtrpC gene products as well (62). In the parentalS. dysenteriae strain carrying this block, noderepression of these late pathway enzymes wasseen. When one considers the great similarity in

the genes and enzymes of S. dysenteriae and E.coli (135), this unexpected finding begs for anexplanation. Other interesting, unansweredquestions attract attention also. Do the auxo-trophic mutations in the Salmonella speciesalso reside in trpE, and are they polar? Docommensal enteric bacteria show an apprecia-ble frequency of auxotrophy or is this restrictedto pathogens? Does auxotrophy ever result froma deletion of all or part of the trp operon? Theseand other questions should not have to waitmuch longer before molecular biologists inter-ested in pathogens decide to study them.

OTHER GRAM-NEGATIVE BACTERIA

PseudomonadaceaeMultiple operon trp gene arrangement.

Pseudomonas strains were the first bacterialexceptions to the single operon trp gene configu-ration studied in any detail. In their pioneeringefforts, Fargie and Holloway (65), studyingmutants of Pseudomonas aeruginosa, notedthat certain genes which are clustered in theenteric bacteria, notably those for the histidine,leucine, and tryptophan biosynthetic pathways,are separate on the Pseudomonas chromosome.In these early studies by the Australian groupnot all the trp loci were represented or distin-guished; nevertheless it was clear that theremust exist at least three trp gene locationsunlinked to each other in transduction experi-ments. Studies in a closely related fluorescentpseudomonad, P. putida, soon established (77)the disposition of the trp genes shown in Fig. 2.(The assignment of trpG to the position shownis still somewhat conjectural, for no mutantsaffecting the small AS subunit have been ob-tained; however, strictly coordinate synthesis ofboth AS subunits has been reported [173]).A general description of the genetics of the

fluorescent pseudomonads is beyond the scopeof this review. The topic has been reviewed byHolloway (93), and a general chromosome maphas been formulated. At present there is nodiscernable similarity between the maps of theenteric species and the pseudomonads. Thethree trp gene clusters are assigned locations 15to 20 min apart on the approximately 70-minPseudomonas chromosome (93).The genetic structure differs and, similarly

the patterns of regulation of enzyme synthesisin Pseudomonas (e.g., in the pyrimidine [103]and isoleucine-valine [147] pathways), showlittle similarity to those seen in enteric bacteria.Considerable effort has been devoted to studyof the genetics and regulation of the extensivepathways of catabolism of diverse carbon and

95VOL. 39, 1975


http://mm

br.asm.org/

Dow

nloaded from


BACTERIOL. REV.

energy sources in strains of Pseudomonas. Incontrast to biosynthetic pathways, clustering offunctionally related dissimilatory genes is com-mon (127, 181), even though different membersof the gene cluster may show independentregulation. Genes for initiating the degradationof certain classes of organic compounds, includ-ing camphor, salicylate, and naphthalene, arefound on transmissible plasmids rather than inchromosomal clusters (35, 60, 175). In the caseof the camphor pathway, it appears that certainenzymes concerned with isobutyrate utilizationlate in the pathway may be duplicated, havingboth a chromosomal and plasmid-borne version(78). It has not escaped the attention of workersin this area that clustering of the genes fordegradative functions may be advantageous forthe pseudomonads, allowing an enormous rep-ertory of dissimilatory enzymes to be main-tained in one or another fraction of the popula-tion, to be amplified by plasmid replication orgenetic transmission when conditions favor it(127, 175). The nearly ubiquitous ability ofmembers of this genus to grow on mineral saltswhen supplied with an oxidizable carbon andenergy source suggests that an adequate biosyn-thetic apparatus is independently maintained.It is probably significant that all of the chromo-somal dissimilatory pathway gene clusters ap-pear confined to one segment representing only10 to 15% of the total chromosome (127).Although the chromosomal disposition of the

genes of several biosynthetic pathways appearssimilar in P. aeruginosa and P. putida (93), itcan not yet be assumed that their genomes areas uniform overall as those of many members ofthe enteric group, for the genetic analysis ofPseudomonas is at a rather primitive level.What follows concerning subunit compositionand regulation of the tryptophan enzymes hasproved fairly uniform in the few species studied,but does not justify the conviction that there isa monotonous genetic structure underlying thishighly varied taxon (166).Subunit structure of enzymes and amino

acid sequences. These aspects have been betterexamined in P. putida than in other species.This was the first organism found to havediscrete enzymes for the third and fourth stepsof the pathway. The sum of their sizes (molecu-lar weight of 39,000 for PRAI and 32,000 forInGPS; [63]) is greater than the size of thebifunctional enteric bacterial protein (molecu-lar weight 45,000). Amino acid sequences arenot yet available to determine their exact ho-mologies. Assuming that the PseudomonasPRAI and InGPS will be homologous with theamino-terminal and carboxy-terminal halves,

respectively, of the enteric bacterial protein,it will be interesting to see the nature and loca-tion of the extra amino acids in Pseudomonas.These may give a clue to whether the two Pseu-domonas proteins arose by a duplication fol-lowed by differentiation of an E. coli-like gene,or whether, as seems more likely a priori, theenteric bacterial gene arose as a fusion of thecompletely independent Pseudomonas genes,with loss of any unnecessary segments.P. putida was also the first bacterium found

to have a small, dissociable glutamine amido-transferase AS subunit (172). Both the large(trpE, molecular weight 63,400) and small(trpG, molecular weight 18,000) subunits havebeen obtained in pure form from P. putida(173); less purified preparations from P.aeruginosa, P. acidovorans, P. stuzeri, P. mul-tivorans, and P. testosteroni have also beenstudied. All six Pseudomonas species formactive hybrid AS molecules, but the nativecomplex from the first two has a molecularweight of 79,000, whereas that of P. acidovoransand P. testosteroni is above 150,000, suggestingaB and a2,h structures, respectively.At the moment there are two equally unsup-

ported and diametrically opposite reasons ad-vanced for the failure to find tryptophan auxo-trophs lacking the small AS subunit. The first isthat this gene product is also used as a gluta-mine amidotransferase in other pathways, somultiple metabolic defects would result from itsloss; in support of this possibility, the sharing ofa single small subunit for both p-aminobenzo-ate and anthranilate synthesis in Acinetobactercalcoaceticus and B. subtilis (see below) may beadduced. The other possibility is that activityin the chorismate amination reaction catalyzedby the large AS subunit alone is sufficient topermit growth of a small subunit mutant onminimal medium without tryptophan. It hasbeen found (72) that whole cells of an E. coliglutamine auxotroph can use ammonia as thenitrogen source for tryptophan synthesis.Whether either of these hypotheses is correctremains for future study.The possibility that a simple fusion of the

trpEGDC segment to one end of trpF and trpBAsegment to the other, creating a single operon ofthe enteric type, is a fair approximation of anevent that actually occurred in bacterial evolu-tion may someday be testable; some indirectsupport for it might come from amino acidsequence analyses showing homology betweenthe small AS subunit of Pseudomonas and thatof Serratia, known (133a) to be homologous tothe amino-terminal portion of the E. coli trpDgene product, or between the small Pseudo-

96 CRAWFORD


http://mm

br.asm.org/

Dow

nloaded from



monas InGPS and PRAI enzymes and thetwo halves of the enteric trpC gene product.The feasibility of such sequence studies isunderscored by the similarity shown betweenthe TS a-chains of P. putida and the entericbacteria (Fig. 3) (44). Their sequence homologyextends from residue 2 (the N-terminal methio-nine residue present in all enteric bacteriala-chains is absent in P. putida, probably be-cause of an intracellular aminopeptidase capa-ble of cleaving a methionylserine bond ratherthan a fundamental genetic nonhomology)through residue 50, the end of the knownsequence, without deletions or insertions. Thesequence seems best conserved in residues 21 to28 and 42 to 50, segments surrounding the tworesidues known to mutate to give inactivemissense proteins in E. coli and suspected,therefore, to be important for the active site (41,216).

Thirteen of the 24 amino acid differences inthe P. putida and E. coli a-chains require morethan a single-base change in the codon. Analysisof the sequence data shows that the P. putidasequence differs more from the known entericbacterial sequences than any enteric organism'sdiffers from another (135), but suggests that allmust have evolved from a common ancestralgene, despite the fact that the chromosomalorganization of the trp genes in Pseudomonasand the Enterobacteriaceae is so different.Regulation of enzyme synthesis. Three dis-

tinct and apparently unrelated regulatory situa-tions prevail in the tryptophan pathway of P.putida (42) and P. aeruginosa (23). Each chro-mosomal location seems associated with a

unique regulatory mechanism. The products ofthe trpE, G, D, and C genes respond to repres-sion by tryptophan, whether exogenously sup-plied or endogenously synthesized, much as theentire enteric bacterial operon does. Mutationsat an unlinked regulatory gene can be selectedthrough their resistance to 5-fluoroindole or

5-fluorotryptophan (152). These mutations re-

sult in constitutive overproduction of the twoAS subunits, PRT and InGPS, with no discerni-ble influence on enzymes in the remainder ofthe pathway. In general these tryptophan ana-

log-resistant mutants of Pseudomonas resembletrpR mutants of E. coli and S. typhimurium;whether in fact they affect a tryptophan re-

pressor protein interacting with an operatorregion for the trpEGDC cluster remains to beproved. The most meticulous study of regula-tion of this segment of the pathway was per-formed by Queener et al. (173), who followedderepression during tryptophan starvation of a

trpF mutant of P. putida. The two AS subunits

increase coordinately with each other but not inperfect unison with the trpD and C gene prod-ucts, for at the end of 3 h the former twopolypeptides had increased more than ninefoldover the repressed value while the latter two hadincreased only half as much. The fine structuremap of the mutants in this area shows anappreciable gap between the trpE and trpDmutants (77), but whether this space is real orartifactual and whether regulatory as well asstructural gene (trpG) elements might occupy itremains to be seen.The solitary trpF gene of P. putida and P.

aeruginosa does not respond to tryptophanscarcity or excess, nor is it affected by regula-tory mutations affecting the remainder of thepathway (23, 42, 152). That is not to say that allstrains and species have identical enzyme lev-els, but whatever level a given strain may have,it seems fixed and unable to respond to scarcityor excess of the end product. In this respect trpFresembles all the enzymes of the pyrimidinepathway in P. aeruginosa (103) and all thetryptophan enzymes of C. violaceum (204).

Probably the most interesting regulatory situ-ation in the Pseudomonas tryptophan pathway,one that is unique so far among the bacteria, isthe inducibility of tryptophan synthase by itssubstrate InGP (42). Both the a- and 13-chainsof this enzyme remain at a low, basal level in P.putida cells undergoing tryptophan starvation ifInGP is absent, and tryptophan even in largeexcess will not prevent induction of the enzymeif InGP is present. Similar results were found forP. aeruginosa (23). Mutations leading to con-stitutivity of the trpA and B genes always lie inclose proximity to this gene pair and frequentlyresult in loss of the enzyme activity associatedwith the trpA gene product (42, 77). This hasled to the hypothesis (167a) that the trpAB genepair exemplifies the phenomenon of autogenousregulation (74), with the a-chain forming atleast part of the repressor protein. Proof ordisproof of this hypothesis should be readilyobtainable in the near future as improvementsin genetic methodology become available for thePseudomonas group.

AcinetobacterMultiple operon arrangement of genes.

Members of the genus Acinetobacter aregram-negative, oxidase-negative, nonflagellate,short rods or diplobacilli found ubiquitously insoil. Modern taxonomy assigns them to theNeisseriaceae, a group of diplococci and di-plobacilli quite distinct from the pseudomonadsor the enteric bacteria (81). Among their rela-

97VOL. 39, 1975


http://mm

br.asm.org/

Dow

nloaded from


BACTERIOL. REV.

tives are the oxidase-positive Moraxella andsome organisms originally called Neisseria, es-pecially Neisseria catarrhalis, N. ovis and N.caviae (12, 13, 81). All members of theAcinetobacter genus can serve as DNA donorsfor transformation of certain receptor strains ofA. calcoaceticus (113); transduction also occursin this genus (82). Transformation is also fre-quent among Moraxella and Neisseria strains(20, 34). Though B6vre (18) and Catlin (33)demonstrated frequent "intergeneric" transfor-mation between certain Moraxella species andthe N. catarrhalis group, Acinetobacter DNAseemed capable of transforming only one trait,streptomycin resistance, into oxidase-positivespecies, and that only at low frequencies (19).Juni (112) confirmed this feeble exchange of thestreptomycin resistance marker between oxi-dase-positive Moraxella and oxidase-negativeAcinetobacter in the reciprocal direction, but hewas unable to show a similar transformability ofauxotrophic markers in the tryptophan andisoleucine-valine pathways. The present conclu-sion (81) is that Acinetobacter, represented by asingle species, A. calcoaceticus, is a distantrelative of the moraxellas and the "false" neis-serias (N. catarrhalis, N. ovis and N. caviae),and still more distantly related to Neisseriagonorrhoeae and N. meningitidis.The relationships just described in the Acine-

tobacter-Moraxella-Neisseria group discernedthrough conventional taxonomic (12, 13) andgenetic (19, 112) means are in agreement withthe results of nucleic acid homology studies(114).Although the general chromosome organiza-

tion in Acinetobacter and Moraxella is stillunknown, it is known that the six trp genes ofA.calcoaceticus are dispersed in three regions, notwo of which can be carried on the sametransforming DNA molecule (187). The assign-ment of the genes to each location is indicatedin Fig. 2. Even though exact gene order withinclusters has not been determined, it is clearfrom the distribution that the trp gene arrange-ment is quite unlike that in Pseudomonas.The gene for the large AS subunit is alone,

those for the small AS subunit, PRT andInGPS are linked, and the gene for PRAI islinked to the pair encoding the TS subunits.Mutations in trpG, the gene for the small ASsubunit, are remarkable for conferring an abso-lute requirement for p-aminobenzoate (or fo-late) in addition to tryptophan (188). It isinteresting that both in A. calcoaceticus and inB. subtilis, where the small AS subunit has alsobeen implicated in p-aminobenzoate synthesis

(116), the genes for the two AS subunits areunlinked.

If the trp genes of A. calcoaceticus are inthe order shown in Fig. 2, two translocations,with the correct fusion of trpC and trpF and theloss of unnecessary control regions, would becapable of creating the enteric bacterial trpoperon. To evolve from the Pseudomonas to theAcinetobacter gene order also involves a mini-mum of two translocations, so Occam's razordoes not suggest any one pathway of evolutionof the gram-negative bacterial families studiedso far.

Investigation of the tryptophan genes andenzymes in Moraxella osloensis has just begun.Preliminary results suggest that the chromo-somal trp gene disposition there is the same asin A. calcoaceticus (P. A. Buckingham and E.Juni, personal communication).Subunit structure of enzymes. Twarog and

Liggins (202), in the first study of the trypto-phan enzymes of A. calcoaceticus, determinedthat the five enzyme activities are readilyseparable. The PRAI and InGPS enzymes ap-pear to be smaller than 30,000 in molecularweight by gel filtration and zonal centrifugationin sucrose gradients; PRT is larger at 42,500 to51,000 and AS larger still at 80,000 or more.Subsequent to their work it was shown that theAS molecule can be dissociated into a large(trpE) subunit of 70,000 molecular weight and asmall one (trpG) of 14,000 (188); the molecularweight of the AS complex in 30% glycerol wasestimated to be 86,000, suggesting an af# struc-ture. Acinetobacter TS has not received muchattention so far. The little that has been learnedis compatible with the occurrence of a readilydissociating a21f2 molecule (187, 202).

In N. crassa and several other fungi, theenzymes of both the tryptophan and the com-mon aromatic amino acid synthetic pathwayshave associations quite unlike those of bacteria.A. calcoaceticus possesses, like N. crassa, ahydroaromatic degradation pathway leadingfrom quinic or shikimic acid to protocatechuicacid (200), and also possesses, as does N. crassa,two dehydroquinases, one inducible in the deg-radative pathway, the other constitutive andused for biosynthesis. Although the constitutivetype occurs in a five-enzyme complex in Neuro-spora, these same enzymes are separate in A.calcoaceticus, as is typical for all other bacteria(15).Regulation of enzyme synthesis. In view of

the unusual diversity of regulatory mechanismsfor the three trp gene clusters in Pseudomonas,it is obviously of considerable interest to know

98 CRAWFORD


http://mm

br.asm.org/

Dow

nloaded from



how the trp genes are regulated inAcinetobacter and Moraxella. The first investi-gations along this line led to the conclusion thatthe AS of A. calcoaceticus does respond totryptophan excess or scarcity, being derepressi-ble as much as 20-fold when certain auxotrophsare starved for tryptophan, but that the otherenzymes of the pathway are largely unrespon-sive (202). More recent work has forced arevision of this interpretation.As in B. subtilis, certain of the tryptophan

synthetic enzymes in A. calcoaceticus are un-stable in extracts but can be stabilized byadding glycerol. This is particularly true for ASand PRT. When all gene products are assayedin stabilized extracts of trp auxotrophs blockedat various points along the pathway, measura-ble derepression occurs for all seven polypep-tides (39). The increase found for the first fivegene products in the pathway ranges from 5- to15-fold, whereas that found for the trpA andtrpB gene products is only 1.5- to 3-fold. Inexperiments similar to those that easily demon-strated the inducibility of TS by InGP inPseudomonas, no obvious inducibility of any ofthe gene products by any of the pathwayintermediates was seen.The operators for trpE and the trpGDC

cluster appear to respond similarly to regulatoryperturbations, resulting in a similar level ofexpression for all four of these early gene prod-ucts. The situation in the trpFAB cluster ismore complex, however. When wild-type A.calcoaceticus is shifted from minimal mediumwithout tryptophan to tryptophan excess, onlythe trpF gene responds by a decreased level ofexpression (39). Apparently trpF is the only trpgene that is not maintained near its maximallyrepressed level when the cell is growing "nor-mally" and making its own tryptophan. More-over, when an auxotroph is starved for trypto-phan, the trpF gene product increases three- tosixfold, whereas the expression of trpA and trpBis augmented only 1.5- to 3-fold (39). This issuggestive evidence for the existence either oftwo operators for the trpFBA cluster or forsome other more complex regulatory situation.Among the tryptophan analog-resistant mu-

tants obtained to date, two types are note-worthy. One has constitutive, derepressed levelsfor all the enzymes of the pathway (39). It thusresembles trpR mutants in the enteric bacteriaand differs from the trpR mutants of P. putida,which affect only the gene cluster for the earlyenzymes (152). Another, showing resistance to5-methylindole in the presence of anthranilateand a mixture of the other major compounds

derived from chorismate (phenylalanine, tyro-sine, p-aminobenzoate, p-hydroxybenzoate, 2,4-dihydroxybenzoate, and vitamin K), has con-stitutive derepressed levels for only the prod-ucts of the trpGDC cluster. Whether or notthere is in fact a single repressor molecule(recognized at three or more spatially distinctand functionally independent regions of thechromosome) for all of the tryptophan genes inAcinetobacter is certainly open to further inves-tigation. The hypothesis is a reasonable specu-lation at present.

GRAM-POSITIVE BACTERIA

Bacillus and ClostridiumChromosomal arrangement. Notwithstand-

ing a long history of investigation of the trypto-phan genes and enzymes in the sporeformingbacteria, only recently has a reasonably com-plete picture become evident. The first studieswere done with B. subtilis 168 shortly after itwas learned that this strain is capable of DNA-mediated transformation (4). This strain carriesan auxotrophic requirement for tryptophan thathad been introduced earlier (22); the mutationproved to be in the trpC gene. Other tryptophanauxotrophs blocked in various parts of thepathway were obtained, and all were found tobe closely linked to each other by transforma-tion (4). Later a quantitative recombinationalmap of mutations affecting all the steps in thepathway was obtained and found to be remark-ably similar to the map of the trp operon inenteric bacteria (Fig. 2) (5, 25). As betterenzymological data appeared (89), it becameclear that mutations affecting PRAI and InGPSoccur in independent but adjacent genes. Andfinally, when it was shown that PRT and AS areunassociated in B. subtilis (89) and that AS hasa large and small subunit (117), it was obviousthat another trp gene must exist. Kane et al.(116) found a mutation in the gene affecting thesmall subunit, and showed that the mutationproduces a defect in both anthranilate andp-aminobenzoate synthesis and that it is un-linked to the original trp gene cluster (Fig. 2).Surprisingly, this mutant did not show anabsolute requirement for either folate or trypto-phan when grown in minimal medium. It wasisolated in a strain with low chorismate mutaseand hence high chorismate levels; when trans-ferred to a strain with normal chorismate mu-tase levels, this mutation does confer a require-ment for both tryptophan and p-aminobenzoate(J. F. Kane, personal communication). Re-cently Kane (personal communication) has

99VOL. 39, 1975


http://mm

br.asm.org/

Dow

nloaded from


BACTERIOL. REV.

shown that trpG lies close to another gene forp-aminobenzoate synthesis (pabA) on the B.subtilis chromosome, and that the pabA proteinserves the same function in p-aminobenzoatesynthesis that the trpE product does in anthra-nilate synthesis. This suggests that in B. sub-tilis trpG was originally a component of p-aminobenzoate synthase that was recruited intoservice in the tryptophan pathway at some timeduring evolution. The protein produced by trpD(estimated molecular weight of 57,000; [89]) issufficiently large to support speculation that avestigial but nonfunctional glutamine amido-transferase segment could remain appended tothe PRT portion.Subunit structure of enzymes and amino

acid sequences. Not only B. subtilis, but Bacil-lus pumilis, B. licheniformis, B. coagulans, B.macerans, and B. alvei all have low-molecular-weight AS molecules which dissociate into largeand small subunits (90, 171). Clostridiumbutyricum AS is similar with a molecularweight near 89,000 (9); it also has been inferredto have a large and small subunit. Although thePRT of C. butyricum was not detected inextracts (9), that of B. pumilus was found at45,000 molecular weight and that of B. alvei at54,000 (90), separate from AS in both cases, ofcourse. Though the AS molecules of the sixBacillus and one Clostridium species all showsimilar molecular weights (9, 32, 90, 117, 171),they differ considerably in their subunit affini-ties, the conditions permitting dissociation ofsubunits and their stabilization, the ratio ofamidotransferase to amination activities of thecomplex, and the apparent molecular weight ofthe large subunit (90, 171). In a group ofbacteria as large and diverse as the sporeform-ers, it would indeed be surprising if all theenzymes of the pathway were uniform structur-ally. In a later section, I will consider the use ofsubunit interchange to assess evolutionary re-latedness. Functional hybrid complexes can beformed with all six Bacillus species (171). Onesurprising result published recently is that hy-brid AS molecules containing one subunit fromB. subtilis and another from P. aeruginosa showappreciable activity in the glutamine-depend-ent AS reaction (170).The sequence of the first 46 amino acids of the

a-chain of B. subtilis TS has been determined(134) and is shown in Fig. 3 along with theanalogous regions for E. coli and P. putida.Although only 20 of the 47 residues are identicalwith the E. coli sequence, the homology isobvious, and it appears that the first six resi-dues of the gram-negative chain (assuming onlya single methionine is cleaved from the P.

putida version) are missing from the B. subtilisversion. Fifteen of the 27 amino acid substitu-tions between E. coli and B. subtilis require atwo-base change in the codon. As with thePseudomonas-Enterobacteriaceae comparison,two regions of the known sequence seem betterconserved than the rest, residues 19 to 28 and 48to 52 (E. coli numbering). These two conservedareas include the only two positions in the first52 residues of the E. coli a-chain known togive rise to inactive molecules by missensemutations, phenylalanine at position 22 andglutamic acid at position 49 (216, 217). Thesetwo residues are among those conserved inall bacterial TS a-chains so far studied (44,134, 135). The molecular weight of the B.subtilis TS a-chain is about 10% less than thatof E. coli (87). Thus the B. subtilis a-chainprobably will have additional segments deletedsomewhere in its amino acid sequence.Regulation of enzyme synthesis. Despite an

earlier report to the contrary (208), upon satis-factory stabilization of all the enzyme activities,production of the proteins of the trp operon inB. subtilis was seen to be coordinate (89). Theproduct of the unlinked trpG gene is alsocontrolled by tryptophan and appears to re-spond to the same repressor molecule as the trpoperon (116), though apparently it cannot bederepressed to the same extent as the rest of thetrp gene proteins. This fact explains a break inthe Ames-Garry plot for glutamine-dependentAS activity at high levels of derepression (89).The location of the B. subtilis mtr gene (namedfor methyltryptophan resistance), the putativestructural gene for the trp repressor, is knownwith some precision. It is carried on the sametransforming DNA molecule as the trp operon,but is not immediately adjacent to it (88a, 91,165, 208). Mutants lacking an effective trprepressor also show some derepression of a hisand tyr gene located distal to trpA beyond thetrp operon, as though some read-through wereoccurring (182).Mutants in the mtr locus may or may not

excrete appreciable amounts of tryptophan.This difference was found to depend partly onthe chorismate mutase level (91). Lorence andNester (131) showed that B. subtilis strains mayhave a chorismate mutase with low activity, onewith high activity, or both. Strains with lowchorismate mutase activity accumulate signifi-cant amounts of chorismate; this overcomes thefeedback inhibition of AS by tryptophan andallows a considerable overproduction and excre-tion of tryptophan (91). Interestingly enough,this excess tryptophan then exerts an inhibitoryeffect on the phenylalanine synthetic enzyme,

100 CRAWFORD


http://mm

br.asm.org/

Dow

nloaded from



prephenate dehydratase (174), resulting inthe paradoxical situation that some 5-methyl-tryptophan-resistant mtr mutants exhibit apartial requirement for phenylalanine. In addi-tion, these mutants show unexplained growthresponses to tyrosine, histidine, arginine, andproline (88a, 91).One mutant bearing a temperature-sensitive

tryptophanyl-tRNA synthase has been exam-ined for regulatory anomalies at permissive andelevated temperatures (197a). Although theauthor concluded that the trpS gene productinteracts in some way with the regulatory mech-anism, it should be noted that similar claimshave been made for the trpS gene product of E.coli (107). Judgment may be suspended untilsuitable in vitro regulatory assays are availablein B. subtilis to subject the hypothesis to acritical test.

Staphylococcus and MicrococcusChromosomal arrangement. Proctor and

Kloos (168, 169) used transduction to study agroup of Staphylococcus aureus tryptophanauxotrophs blocked at various points in thepathway. All their mutants were linked, and byquantitative recombination experiments theydetermined the gene order to be trpE, D, C, F,B, and A (Fig. 2). If a trpG locus exists, itsposition is unknown, for no mutants have beenobtained to define it. In an analogous studyfrom the same laboratory, Kloos and Rose(118a) used transformation to study a set oftryptophan auxotrophs of Micrococcus luteus.Here close linkage was demonstrated for trpE,C, B, and A mutants, and both quantitativemapping experiments and three-point tests in-dicated the genes in the cluster to be arrangedin that order (Fig. 2). Ten mutants were de-duced by growth and accumulation tests tolack either PRT or PRAI, hence, they musthave been either trpD or trpF mutants. Noneof the ten was linked by transformation to thetrpECBA cluster. Unfortunately it cannot bestated at present whether only one or both ofthese genes, trpD and trpF, are unlinked tothe main cluster; moreover, the existence andlocation of a trpG gene will remain speculativeuntil mutants defective in a glutamine amido-transferase AS subunit are found. Additionalgenetic and biochemical studies with thesegram-positive cocci may be awaited with in-terest.Subunit structure and regulation of

enzymes. In S. aureus the PRAI and InGPSenzymes are separable by gel filtration. Bothhave molecular weights below 30,000, indicating

that trpC and trpF are separate loci, not twoparts of a single gene (169). AS is not found in acomplex with PRT or any other enzyme of thepathway. Its estimated molecular weight of65,000 seems quite low, but not amazingly so inview of the values obtained with some of thebacilli, such as B. alvei, under certain condi-tions (32, 90, 171); this enzyme too is probably acomplex of a large and small subunit, similar tothe ones found in all other bacteria except theEscherichia-Salmonella-Enterobacter group.PRT is small in S. aureus at molecular weight

30,000. TS activities were labile in extracts, sothat enzyme has not yet received a carefulcharacterization in Staphylococcus. Little isknown of the subunit structure and molecularweights of the Micrococcus enzymes.

Proctor and Kloos (169) studied the trypto-phan enzyme levels in trpA and trpE auxo-trophs of S. aureus grown under conditions oftryptophan excess or limitation. They foundthat all five enzymes (both the A and B reac-tions of TS were studied, so six specific activi-ties were measured) were derepressed concur-rently and coordinately upon tryptophan star-vation. The highest level of derepression ob-served was 100-fold greater than the repressedlevel. Thus, the mode of regulation of thispathway in Staphylococcus seems quite analo-gous to that in the enteric bacteria and thebacilli, but unlike that in Pseudomonas orAcinetobacter. It will be interesting to learnhow the enzyme levels are controlled inMicrococcus where there are at least two trpgene clusters.

StreptomycesChromosomal arrangement. A large body of

genetic work has been performed by usingconjugation in Streptomyces coelicolor andsome closely related actinomycetes (for a reviewsee reference 94). These prokaryotes grow as amycelium and make spores asexually. Theirgenetic map shows a single circular chromo-some. Tryptophan auxotrophs are found at twodifferent sites on the map (64, 94, 195a). Eightmutants lacking either the A or the B reaction ofTS have been mapped at one site near twomutants lacking InGPS; one mutant lackingPRAI and 17 lacking PRT map at another site(195a). So far no mutants lacking AS have beenfound, so the number of trp gene clusters maybe more than two. There have been no reports ofenzyme associations in this group of organisms,but there are indications that S. coelicolorregulates the level of all the enzymes of thepathway except InGPS (195a).

101VOL. 39, 1975


http://mm

br.asm.org/

Dow

nloaded from


BACTERIOL. REV.

BACTERIA STUDIED NON-GENETICALLY

Chromobacterium, Lactobacillus, andBlue-Green Bacteria

Subunit structure and regulation ofenzymes. This section summarizes some en-zymological work done in organisms where a

method of genetic exchange has either not yetbeen found or has not yet been employed tostudy the chromosomal location of the trpgenes. Bacteria in the genus Chromobacteriumare gram-negative rods characterized by theproduction of an insoluble, dark blue pigment,violacein. This pigment is a rather simplemolecule formed enzymatically from two mole-cules of tryptophan (53). Mutants unable tosynthesize tryptophan make colonies that arecolorless on a medium with limiting tryptophanbut bright blue in tryptophan excess (204). Thisproperty makes the selection of tryptophanauxotrophs rather easy. Mutants blocked ineach step of the pathway have been studiedphysiologically and enzymatically (204). Intheir molecular weights and absence of aggrega-tion, the enzymes of the pathway resemblethose of B. subtilis. Chromobacterium is quitedifferent, however, in the apparent lack of a

mechanism of regulation of tryptophan enzymesynthesis. Feedback inhibition of AS by trypto-phan is very efficient in this organism (204), andas far as can be determined, it is this mecha-nism alone that keeps the flow of metabolitesdown the tryptophan pathway in balance withthe other uses of chorismate, yet adequate bothfor protein synthesis and pigment formation.Mutants resistant to 5-methyltryptophan were

isolated and studied (J. W. Hirst and 1. P.Crawford, unpublished data). None were foundto have elevated tryptophan enzyme levels.In the earliest investigation showing that

anthranilate can be used in place of tryptophan,Snell (196) used several species of lactic acidbacteria having a natural tryptophan require-ment. Among these, Lactobacillus casei andLactobacillus arabinosis 17-5 would accepteither anthranilate or indole in place of trypto-phan, whereas three other species would nottake either intermediate. The similarity of thissituation to that in enteric bacteria showing a

natural tryptophan requirement is further em-phasized by more recent work with L. casei(156) showing that spontaneous and mutagen-induced reversion of a lesion near the end of thepathway can occur. In this case no reversion ofthe natural AS block was seen. Clearly thelactobacilli are candidates for more detailedinvestigation of enzyme size and aggregation,

subunit interchange with other gram-positiveand even gram-negative forms, and mode ofregulation of enzyme synthesis.

Studies of the tryptophan pathway in theblue-green bacteria (formerly called blue-greenalgae) are at an early stage. In one species,Anabaena variabilis, association of the enzymeshas been studied. AS and PRT sediment inde-pendently, while PRAI and InGPS cosedimentat about 5.5S (99); the authors note that thispattern resembles that of the oomycetous fungi.Sakaguchi (184), on the other hand, found thatthe TS of the same organism dissociates into aand fl2 subunits like bacterial versions of theenzyme and unlike the known fungal ones. In adifferent blue-green bacterial species, Agmenel-lum quadruplicatum, a tryptophan auxotrophblocked in the TS-A reaction has been obtainedand used in studies of the regulation of enzymelevels (102). All the enzymes of the pathwayincrease in response to tryptophan deprivation,but the increase of the first four activities of thepathway is only two- to threefold while TS-Bactivity increases 20-fold. AS amination activ-ity was shown to be fully inhibited by 10 ILML-tryptophan; no AS glutamine amidotransferactivity was detected in extracts (102). Lest itbe thought from the above that the blue-greenbacteria are very close relatives of other bacte-rial taxa, it should be noted that they synthesizetyrosine by a novel pathway involving transami-nation of prephenate prior to aromatization,whereas all other groups of prokaryotic andeukaryotic organisms perform these steps inreverse order (198).

EUKARYOTIC ORGANISMS

FungiChromosomal arrangement. This survey of

the genetics of the tryptophan pathway ineukaryotes will be very cursory, primarily be-cause the topic has been adequately coveredelsewhere (51, 97). Some of the evolutionarymechanisms that have been observed in bacte-ria, including gene fusion, appear to have oc-curred in eukaryotic organisms as well, espe-cially among the fungi, necessitating a briefreview of the topic, however.Study of the fungal tryptophan genes and

their products has a long history, beginningwith the pioneering efforts of Tatum and Bon-ner in 1944 (199). N. crassa devotes four un-linked genes to the enzymes of the tryptophanpathway (Fig. 4). All mutations affecting TS arefound at a single locus, tryp-3. Mutations affect-ing PRT map at tryp-4. At the tryp-1 and tryp-2loci are found mutations having diverse pheno-

102 CRAWFORD


http://mm

br.asm.org/

Dow

nloaded from



types affecting AS, PRAI, and InGPS (51, 97).The tryp-1 and tryp-2 gene products combine toform a multimeric enzyme catalyzing thesethree reactions, as will be described in the nextsection.There is a fifth locus in N. crassa capable of

mutating to give rise to tryptophan auxotrophs(3), but this has been shown to be the structuralgene for tryptophanyl-tRNA synthase (163) so itis not, properly speaking, a component of thetryptophan synthetic pathway.The trp genes are also well characterized in

several other fungi. The gene-enzyme relation-ships in Aspergillus nidulans are certainly verysimilar to Neurospora (98, 177), and Coprinusradiatus is probably also similar (80). In theyeast Saccharomyces cerevisiae there are fivetrp pathway structural genes instead of four, allunlinked to each other (51, 57). The gene-enzyme assignments are similar to Neurosporaexcept for PRAI, which is encoded by a separategene in yeast (Fig. 4). Work has recently begunon the tryptophan gene-enzyme relationships inthe fission yeast Schizosaccharomyces pombe(193). Some details remain to be elucidated, butthe authors suggest that the PRAI and InGPSloci may be genetically distinct though adjacenton the chromosome, which is a situation inter-mediate between Saccharomyces and the fila-mentous fungi.Subunit structure of enzymes. Two multi-

meric proteins are known in the tryptophanpathway of N. crassa. The most recent evidencesuggests that one of them, TS, is a dimer ofmolecular weight 150,000 having two identicalsubunits (149, 200a). Earlier findings that thismolecule had an a1#2 structure, with a and ,chains both about 35,000 daltons in size (26),may have resulted from endogenous proteolyticattack. In both N. crassa (224) and yeast (142,183) there are inactivating proteases that co-purify with TS, as well as an inhibitor of thisprotease that may be removed from the prepa-ration at certain steps. Genetic evidencestrongly supports the homodimer model for theTS enzyme (17). There is a localization ofmutant phenotypes within the gene both in N.crassa (17, 121) and S. cerevisiae (143); muta-tions affecting one of the two TS half reactionsmap together at one end of the gene, separatefrom mutations affecting the other half-reac-tion. In both organisms mutations affectingonly one of the two TS half-reactions are in theminority. Some of the mutations leading tototal inactivity are missense, termed "pro-found" in N. crassa (17, 121), whereas some arenonsense. In yeast the effect of nonsense muta-tions in one small segment of the gene is

Organism

Neurospora crassa

Linkage group

Aspergillus nidulans

Linkage group

Enzyme

tryp-2 tryp-l tryp-4 tryp-3 tryp-S

VI III IV 11 V

trypA - trypC trypD trypB trypE

II Vill II I VI

Sacchareuyces cerevisiae trp-2- trp-3 trp-1 trp-4 trp-S

Linkage groQ V Xi IV IV VII -

FIG. 4. Tryptophan gene-enzyme relationships inthree fungal genera. The nomenclature is that inuse currently (51, 177) in each organism. Enzymeabbreviations are as in Fig. 1 and the text; TTS,tryptophanyl-tRNA synthase. Linkage groups arethose specified for each organism and are not uniform.The two trp genes on linkage group IV of S. cerevisiaeare widely separated. Products of the genes connectedby the double arrow - associate to form a depend-ent enzyme complex. All other gene products appearto be unassociated.

particularly interesting, for the nonsense frag-ments produced retain activity in the TS-Areaction, though their molecular weight deter-mined from gel filtration is only 35,000, aboutone-fourth of that of the native enzyme (141).The most reasonable interpretation of all thedata available for the tryptophan synthaselocus in N. crassa and S. cerevisiae is that itencodes a single long polypeptide chain havingthe activity of the bacterial TS-a chain at oneend and the #-chain at the other. The orienta-tion of the two segments of the locus is suchthat the a-chain portion is translated first.The second multimeric protein in the fungal

tryptophan synthetic pathway has three activi-ties in Neurospora (AS, PRAI, and InGPS) andtwo in Saccharomyces (AS and InGPS) (51,57). The most recent work suggests that inNeurospora it is a 310,000 molecular weightprotein with two 94,000 and two 70,000 molecu-lar weight subunits (95). This is in good agree-ment with the genetic evidence that two un-linked loci control this enzyme aggregate (51).Apparently one of the polypeptide chains isconcerned primarily with AS (tryp-2 product),the other with PRAI and InGPS (tryp-l prod-uct), but both must be present for apprecia-ble AS activity (54, 69). Earlier evidence sug-gesting a subunit molecular weight as low as40,000 for this protein (69) may have been dueto proteolysis (95) as suggested above for TS.It is conceivable that the tryp-2 gene productin Neurospora represents a fusion peptidecontaining the activities of the bacterial trpEand trpG gene products, and the tryp-1 geneproduce represents a fusion of the bacterialtrpC and trpF functions.The AS-InGPS complex of S. cerevisiae has

103VOL. 39, 1975


http://mm

br.asm.org/

Dow

nloaded from


CRAWFORD

received less study than the Neurospora tri-enzymic complex just described, but both in itsgenetics and enzymology it resembles the latterwithout the PRAI segment of the tryp-1 product(51, 57). In yeast, PRAI is a separate protein,smaller than the PRAI-InGPS bifunctional en-zyme in E. coli (51). Recently H. Henke and R.Hitter (personal communication) have foundthat in some basidiomycetes, such as Coprinuslagopus and Thanatephorus cucumeris, all theactivities of the pathway except TS can appearin a single, large enzyme complex. Additionalstudies in this group will be awaited withinterest.HUtter and DeMoss (99) performed an exten-

sive survey of the size and association of en-zymes of the tryptophan pathway in 22 conven-tional fungal species plus the myxomycetePhysarum polycephalum. The patterns found inthe filamentous ascomycetes and some phy-comycetes conform reasonably closely to thepattern just described for Neurospora. Two ofthe yeast genera, Dipodascus and Endomyces,share the Saccharomyces pattern describedabove; Schizosaccharomyces is now known to bequite similar also. One group of phycomycetes,the Oomycetes group, including Saprolegniaand Pythium, differs markedly from all otherfungi in having its AS unassociated with otherenzymes of the tryptophan pathway. In thisgroup the PRT is also an independent molecule,but the PRAI and InGPS cosediment at 3.5 S insucrose gradients (99). Whether these two activ-ities are associated on a single polypeptidechain or not remains undetermined. The uniquetaxonomic position of the Oomycetes groupvis-a-vis the rest of the fungi is reflected in othercharacters as well. They have cellulose insteadof chitin as the main cell wall constituent, andthey synthesize lysine by the diaminopimelicacid pathway as do bacteria, algae, and vascu-lar plants, rather than by the a-amino-adipateroute employed by ascomycetes, basidi-omycetes, chytrids, and euglenoids (203). Itwould be highly desirable to perform a geneticanalysis of the trp loci in a typical oomycete,and a study of the TS in this group to see if itconforms to the bacterial model would also bevery interesting.

Regulation of enzyme synthesis. This sub-ject has been intensively studied in fungi,principally Neurospora, and has recently beenreviewed (97, 154). The degree of derepressionattainable differs for different enzymes, but ingeneral is less than is seen in comparableexperiments with enteric bacteria, pseudomo-nads, or bacilli (57, 163). Since tryptophan is aneffective feed-back inhibitor of both AS and

PRT (52, 57), the increased enzyme synthesismay result either from induction by accumu-lated intermediates or from derepression bydecreased levels of tryptophan or charged tryp-tophanyl-tRNA. In fact, at least two signals areinvolved in the regulatory response of the tryp-tophan enzymes in Neurospora, for the accumu-lation of InGP has an inducing effect on TSsynthesis (201), but even mutants unable toform this intermediate respond to tryptophandeprivation (128, 129). Study of the N. crassamutant deficient in tryptophanyl-tRNA synthe-tase suggests that it is the charged tRNA ratherthan free tryptophan that is the effector in thelatter case (163). The increase in TS levelsbrought about by indoleacrylic acid is due toboth these mechanisms, for Matchett hasshown that this compound effectively inhibitsTS-B activity, resulting in the accumulation ofInGP, the inducer, and a decrease of chargedtryptophanyl-tRNA, the repressor (148a).The regulatory interrelation of the histidine

and tryptophan synthetic pathways suggestedlong ago by Hogness and Mitchell (92) has beendocumented and amplified recently. Imidazole-glycerol phosphate, an intermediate in the histi-dine pathway structurally similar to InGP, hasan inductive effect on TS similar to InGP (201).This effect is specific for the tryp-3 gene productand is not counteracted by exogenous trypto-phan. But Carsiotis and his co-workers haveshown that this is not the only point of interac-tion between these pathways, for histidine aux-otrophs of all types, even those blocked beforeimadazoleglycerol phosphate formation, in-crease the level of all the tryptophan enzymeswhen starved for histidine (28, 30). Moreover,exogenous tryptophan does not reverse the histi-dine starvation effect (28). This effect is recipro-cal, for tryptophan starvation also results in anincreased synthesis of enzymes of the histidinepathway (27). Two other amino acids haverecently been included in this regulatory inter-relationship in Neurospora, arginine (27, 29) andlysine (206). The present interpretation of thissituation is that regulatory signals, probablyrelated to the degree of charging of specifictRNA molecules with these four amino acids,are reflected in the rate of formation of theenzymes for the synthetic pathways and thatconsiderable cross-pathway response to thesesignals occurs (130, 205). The mechanism andfull significance of these relationships are stillnot entirely clear, but it is apparent that similarinterrelationships of the histidine, tryptophan,and arginine pathways occur in S. cerevisiae(191), and at least the histidine and tryptophanpathways are interrelated in Euglena (123).

104 BACTERIOL. REV.


http://mm

br.asm.org/

Dow

nloaded from



Algae, Euglenoids, and PlantsSubunit structure of enzymes. Genetic anal-

ysis of the tryptophan genes in these eukaryoticorganisms has not yet been accomplished, butbiochemical experiments have uncovered someproperties of the enzymes of the pathway. TheTS of the conventional, unicellular green algaChlorella ellipsoidea is a two-component, dis-sociating enzyme like that of the prokaryotesrather than a single-component one like that ofthe fungi (184). Sakaguchi reports that thisorganism's B ( = pl2?) subunit, as well as that ofthe blue-green bacterium A. variabilis, isweakly complemented by a-chains from E. colior S. typhimurium TS. The Chlorella B com-ponent, but not the Anabaena one, weaklycross-reacts immunologically with antiserumto N. crassa TS (184). Confirmation of theseresults with more purified enzyme and someamino acid sequence information would be ex-tremely interesting.The eukaryote Euglena gracilis is unique

among organisms studied so far in having all theenzymes of the pathway beyond AS combinedin a single aggregate of about 234,000 molecularweight (122). The AS of this organism has amolecular weight of 80,000 (C. Hankins and S.E. Mills, manuscript in preparation). The sub-unit structure of AS as well as that of the largetrp enzyme complex in E. gracilis remains asubject for future study. Even with the presentknowledge, however, it is clear that there aremajor differences in the tryptophan enzymes ofeuglenoids and the higher fungi, despite the factthat these two groups share the a-aminoadipicacid pathway of lysine formation (203).And finally, the tryptophan synthetic path-

way of higher plants has been the subject ofseveral enzymological investigations. It is clearthat the pathway and intermediates are thesame as in microorganisms (48, 211). TS fromseveral genera of higher plants is dissociableinto two unlike components (36, 47, 162). Themolecular weight of TS is about 140,000, closeto that of the microbial forms (36, 162). Theearly enzymes of the pathway all appear toreside on separate molecules; the molecularweights of these enzymes as found in corn andpeas are: AS, 90,000; PRT, 80,000; PRAI,26,000; and InGPS, 50,000 (C. Hankins and S.E. Mills, manuscript in preparation). TheInGPS is probably a dimer, for active materialat half that molecular weight is also present.Regulation of enzyme synthesis. Little is

known of the regulatory mechanisms employedby algae and euglenoids, and the subject iscertainly not closed in plants either. It is quiteclear that in plants feedback inhibition of AS

plays an important role in the regulation of theflow of metabolites down the tryptophan path-way (14). Widholm has demonstrated that themechanism of growth inhibition by tryptophananalogs such as 4- and 5-methyltryptophan and5- and 6-fluorotryptophan is false feedbackinhibition at this site (209). Mutants resistantto these analogs have a modified AS resistant tofeedback inhibition by either the normal aminoacid or its analogs (210). Whether, as in mostprokaryotes and fungi, this feedback inhibitionmechanism is ever supplemented by a modula-tion of enzyme synthesis remains an open ques-tion at present.

GENERAL CONCLUSIONS ABOUTEVOLUTION OF THE PATHWAYConservation of Reaction Mechanisms

It is always dangerous to pontificate, but itcan be safely stated that if a pathway ofbiosynthesis of tryptophan different from thatshown in Fig. 1 exists in living things, ratherwide-ranging investigations have so far failed touncover it. At least seven widely divergentprokaryotic groups and five eukaryotic oneshave been surveyed, and although representa-tives from groups with very different mech-anisms of tyrosine (198) and lysine (203) synthe-sis are present in the assemblage, no variationin the route of tryptophan synthesis was found.Even more remarkable, perhaps, is that cer-

tain nuances of the reaction mechanisms alsoseem to have been retained during evolution.The ability of AS to use either glutamine as anamino group donor at physiological pH or mo-lecular ammonia at a slightly higher pH hasbeen found wherever sought, from bacteria(225) through fungi to plants (210). Similarly,TS from all groups retains the ability to utilizeindole in place of InGP despite the fact thatindole is a "captive" intermediate in the physio-logical reaction (118, 216) and must rarely beencountered in the natural environment of someof the organisms studied.

If the reactions of the pathway present amonotonous uniformity, the associations be-tween the enzymes catalyzing them show abewildering variety. Discounting the aggregatesthat are covalently bound as a result of fusedgenetic elements, it is still a puzzle to under-stand the occurrence of AS, PRAI, and InGPSin a complex in most fungi and the last fourenzymes of the pathway in a complex in Eu-glena. Some of the puzzle may be resolved asmore structural information becomes available.It now seems reasonable that as a result of thefusion of the genes for the small AS subunit andPRT in the Escherichia-Enterobacter-Sal-

105VOL. 39, 1975


http://mm

br.asm.org/

Dow

nloaded from


ABACTERIOL. REV.

monella group of the enteric bacteria, AS andPRT exist in a relatively firm complex, whereasthey show no aggregation in Serratia where thisfusion has not occurred (225). Similar explana-tions must be sought in the other cases, but itmight also be remembered that certain advan-tages accrue to assembling the enzymes of onepathway into a single complex. Some examplesof catalytic "facilitation" by such molecular"organelles" in the tryptophan and the multiplearomatic pathways of Neurospora (70) andEuglena (14a) have been adduced. Perhapsmany organisms have their tryptophan pathwayenzymes associated intracellularly through non-covalent interactions of varied strength. Whencell-free extracts are made only the strongest ofthese interactions may be manifest in residualenzyme complexes.

Variation in Chromosomal Disposition ofGenes

Translocation and gene fusion. Nearlyevery major bacterial group studied has a dis-tinctive chromosomal arrangement of the genesconcerned with tryptophan synthesis (Fig. 2).Although less information is available fromeukaryotes, among the fungi also there is vari-ety in the number and location of the trp genes(Fig. 4). Differences in regulatory response seemcorrelated with chromosomal location. It seemsrather obvious that the structural genes for thetryptophan enzymes, many of them having thesame evolutionary origin, as indicated by resid-ual sequence similarities (Fig. 3), have beentranslocated to various positions on the chromo-some during evolution and are associated withdifferent trp gene partners, even responding todifferent regulatory stimuli. All this has proba-bly resulted from ordinary mechanisms of chro-mosomal rearrangement (i.e., reciprocal trans-locations, insertions, inversions, etc.), coupledwith a considerable plasticity of regulatoryelements capable of responding to selection tomaximize cellular economy under the prevail-ing conditions.An unexpected finding in the study of the trp

genes is the number of examples of fusion ofgenes encoding two polypeptides in the samepathway. In discussing fusions it will be as-sumed that the original state consisted of sixquite independent genes, trpE and trpG for AS,trpD for PRT, trpF for PRAI, trpC for InGPS,and trpA and trpB for TS. Fusion must haveoccurred twice in the lineage of the entericbacteria, once combining trpC and trpF to forma single bifunctional peptide, and later a fusionof trpG to trpD. The first may have occurred ina strain ancestral to the entire taxon, the second

in one common to only the Escherichia-Sal-monella-Enterobacter group. In both casesthere is suggestive evidence that the bifunc-tional polypeptide assumes a conformation con-sisting of two "primal" domains connected by alinking segment (45, 75, 124). In one case, atleast, the linkage appears to be protease sensi-tive (75).

In the fungi the prediction of Bonner et al.(17) concerning TS appears to have been borneout; the fungal enzyme consists of a fusion of a-and ,B-chain elements seen in all other orga-nisms (149 200a). Kaplan et al. have presentedimmunochemical evidence for two independentpolypeptide "domains" in the Neurosporaprotein (117a). It will be interesting to see if thesite(s) most sensitive to protease attack in thisenzyme are in a linkage region between twoindependently folding a and ,B domains. Thoughthis fusion may have occurred ancestrally to allthe higher fungi, fused trpC and trpF functionsare found only in the filamentous ascomycetesand the basidomycetes, not the yeasts (99).Whether in fact a fusion of the trpG and trpEfunctions has also occurred or whether trpGmay be fused to trpC in this interesting complexwill probably become clear with additionalstudy.

It is easy to note that all the examples of genefusion just described are reflected in enzymeaggregations (considering TS-A and TS-B to betwo activities). In those organisms where ge-netic analysis has not been performed, theexistence of multienzyme complexes might be-token still more examples of gene fusion. Aprime candidate for such an occurrence wouldbe Euglena with its large complex embodyingall the reactions of the pathway beyond AS(122).Regulatory mechanisms. Among the pro-

karyotic species studied an almost bewilderingrange of regulatory variations has been de-scribed. Where units of function occur theyseem to reflect chromosomal organization, andthe regulatory signal seems usually to be theamino acid itself or a pathway intermediaterather than the charged tRNA. Few other gener-alizations are possible, however, for segments ofthe pathway may be controlled by either induc-tion or repression over a wide or very narrowrange. There are even instances where the twopolypeptide chains of a given enzyme are notunder fully coordinate regulation, as in B.subtilis AS (89, 116).With the bewildering variety of bacterial

examples, it is perhaps no surprise that thefungi should have developed something quitedifferent for their regulation. Though more work

106 CRAWFORD


http://mm

br.asm.org/

Dow

nloaded from



remains to be done, the fungal mechanismappears to recognize charged tRNAs as effec-tors, and a given signal may influence thesynthesis of enzymes in several amino acidpathways. It is far too early to assume that theseregulatory differences embody a major eu-karyote-prokaryote dichotomy, however, Theimpression at present is that regulatory mech-anisms can be altered during evolution at leastas easily as chromosomal gene location andmuch more easily than the amino acid sequenceof an enzymatically active polypeptide.

Enzymes Shared with Other PathwaysThe sole example so far uncovered of a

tryptophan pathway element shared with an-other pathway is the use of the same glutamineamidotransferase subunit for both AS and p-aminobenzoate synthesis in B. subtilis (116)and A. calcoaceticus (187). The obvious similar-ity of these two reactions, having the samesubstrates and a product differing only in orthoor para placement of the amino group, leads oneto suspect that one enzyme may have evolvedfrom the other. Gene duplication followed byactive site modification is an evolutionarymechanism favored by theoreticians, and itmay well have occurred in this case.Rather than postulate duplication of the large

AS subunit gene without a corresponding du-plication of the small subunit in the ancestry ofB. subtilis and A. calcoaceticus, however, I willpropose a different possibility which recognizesthe fact that the "trpG" of A. calcoaceticus islinked to other trp genes, whereas that of B.subtilis is linked to a p-aminobenzoate pathwaygene. If one assumes that both organisms origi-nally possessed complete, two-subunit en-zymes, but that through deletion, insertion ofextraneous DNA, or mutation within the gene,A. calcoaceticus lost its amidotransferase sub-unit for p-aminobenzoate synthase, while B.subtilis lost the one for AS, then each organismmight have become dependent on the smallsubunit from the counterpart enzyme. Thissituation, which is not very bizarre either genet-ically or biochemically, proposes independentoccurrence in separate lineages of "mirror-image" inactivating events. Subsequent modifi-cation of the regulatory region for the p-aminobenzoate enzyme in B. subtilis, to bring itunder tryptophan control, would produce thepresent situation, as far as is known.Whether the hypothesis just stated is approx-

imately true or merely fanciful, it certainlybehooves investigators to be alert to the possi-bility of other elements of similar function thatcould be shared between distinct pathways in

specific organisms. Not only amidotransferases,but many other transferases and isomerases areobvious candidates. Such features may be asmuch a fixture of certain taxa as an alternatepathway, multiple enzymes for one reaction, orfused genes.

NATURAL RELATIONSHIPS REVEALEDBY COMPARATIVE STUDIES

Amino Acid SequencesPredictably, I begin this section on determi-

nation of natural relationships by pointing outthat pitfalls exist in any one method, so aconclusion based on many different lines ofevidence is more likely to stand the test of timethan one resting on a single support, howeverclear-cut and logical. Many ways exist to deter-mine relatedness by the degree of similarityshown within a large, natural group such as theenteric bacteria or the pseudomonads. Whenapplied with intelligence, most of thesemethods seem to give nearly identical answers.It is in the determination of natural relation-ships among different large groups that manyexisting taxonomic methods fall short whenapplied to prokaryotes. My preference for thispurpose lies with the comparison of amino acidsequences in homologous proteins, preferablythose of major importance to cellular economy.This preference is based on the experienceobtained so far with enzymes of the tryptophansynthetic pathway.Much of the relevant, admittedly limited

data is summarized in Fig. 3. There is acomforting conservation of sequences in theseveral enteric bacteria and a recognizable butmore distant relationship between these and thePseudomonas and Bacillus sequences. Some-what similar results involving other prokaryotictaxa have been obtained for the iron-sulfurproteins (222).Though it seems obvious that amino acid

sequence comparisons will provide highly relia-ble evidence of relatedness between large tax-ons, and considerable experience with a widerange of eukaryotes seems to bear this out, thereis a pecularity of prokaryotic genetics thatmight lead to misleading results. Plasmid trans-fer has been observed between bacteria of differ-ent groups, specifically between Pseudomonas,Vibrio, and enteric bacteria (79). Doubtless thisrange will soon be extended. Plasmids may onoccasion bring chromosomal elements withthem; these, if analyzed, could lead to falseestimates of similarity. Perhaps the best de-fense against such errors would be to chooseproteins of basic, nearly indispensable path-

107VOL. 39, 1975


http://mm

br.asm.org/

Dow

nloaded from


BACTERIOL. REV.

ways, to study their genes whenever possible toassure a conventional chromosomal location,and above all not to depend on the analysis of asingle polypeptide.

Nucleotide Sequences and Nucleic AcidHybridization

Protein sequence similarities are reflected insimilarities in sequence of the correspondingmRNA and DNA molecules. The developmentof sensitive DNA-DNA and DNA-RNA hybridi-zation methods has provided a methodology forassessing sequence similarities in nucleic acidswithout knowing the actual nucleotide se-quence. Careful measurements of the stabilityof RNA-DNA hybrids corresponding to the TSa-chain in three enteric bacteria led Li et al.(132) to the conclusion that nucleotide sequencehas diverged almost twice as rapidly as aminoacid sequence, i.e., that as many base changesto synonymous codons have been fixed in evolu-tion as base changes leading to amino acidsubstitutions. This fact limits the usefulness ofnucleic acid hybridization techniques to rela-tively closely related species, a conclusion borneout by the inability of Denney and Yanofsky(55) to demonstrate hybridization between themRNA of B. subtilis and the trp DNA of E. colidespite the recognizable conservation of theamino acid sequence of their TS a-chains (134).The above is not meant to denigrate the useful-ness of hybridization techniques for determina-tion of natural relationships within bacterialfamilies, or to deny that certain nucleotidesequences such as those concerned with ribo-somes may be conserved better than the trpones (37, 58, 59, 213a). Nevertheless, the tech-niques of protein purification and automaticEdman degradation have now progressed to thepoint where wide surveys of the prokaryotickingdom are quite feasible, and would be ex-pected to orient on a larger canvas the detailedrelationships revealed by numerical taxonomyand nucleic acid hybridization (for reviews seereferences 40 and 140).

Subunit InterchangePurification of proteins and determination of

their amino acid sequence has not always in thepast proved a simple or infallible procedure;therefore many schemes have been concocted tocircumvent it. Two of the more obvious oneswill be dealt with briefly in this and the nextsection. Where there are multimeric enzymeswith readily dissociable, unlike subunits, it islogical to assume that the affinity of thesesubunits obtained from different organisms

might closely approximate the degree of se-quence similarity, and hence the evolutionaryrelatedness of the sources. In the context of thetryptophan genes and enzymes of bacteria, thisexpectation has not been fully borne out.There are two multimeric enzymes in the

bacterial tryptophan pathway suitable for sub-unit exchange experiments, AS, when gluta-mine rather than ammonia is the amino groupdonor, and TS, where the low activity of thedissociated subunits versus the complex allowsactivity in any reaction to serve as a yardstickfor measuring complex formation.

It has been shown that the trpE product of E.coli will function with the trpD proteins of S.typhimurium and E. aerogenes quite readily(104), and can even accept the much smallertrpG product from S. marcescens for gluta-mine utilization (176). A similar complementa-tion with the small trpG subunit of E. hafniaehas been reported (124), but there are noindications that the large and small subunits ofS. typhimurium and P. putida can functioncooperatively (173). Functional hybrid AS mol-ecules involving six different species of Bacillushave been observed (171). More surprisingly,the large B. subtilis subunit complements wellwith the small P. putida one; the reversecombination is somewhat less catalytically ef-fective though not entirely inert (170). Large A.calcoaceticus AS subunits do not form veryeffect complexes with the small subunits ofB. subtilis or P. putida, and no activity at allwas found with the small S. marcescens subunit(188).TS complexes formed with the a and #2

subunits from different enteric bacterial speciesare quite active (6), though the affinity betweenS. marcescens #2 subunit and E. coli a-chains issomewhat less than that of the Shigella, Salmo-nella, and Enterobacter #2 subunits (178). Thereverse is not true, as a-chains from all four ofthese organisms have excellent affinity for theE. coli f2 subunit (161). P. putida a and #2subunits give little or no activity with E. coliones, and in fact no physical binding of P.putida #2 and E. coli a subunits is seen (151). B.subtilis #2 subunits are stimulated in the TS-Breaction by both E. coli and P. putida asubunits to 30% of the activity of the homolo-gous complex, however (88). This interaction isnot reciprocal, for B. subtilis a-chains have noeffect on E. coli or P. putida #l2 subunits (87).Within the Bacillus genus, B. pumilus a-chainscomplement B. subtilis #2 subunits perfectly,but B. alvei a-chains do not complement themat all (90). Weak stimulation by the B ( = #2?)subunits of Anabaena and Chlorella of the

108 CRAWFORD


http://mm

br.asm.org/

Dow

nloaded from



a-chains of E. coli and S. typhimurium in theconversion of indole to InGP was reported (184).The general conclusion to be drawn from

these admittedly scattered observations is thatsubunit interchange has not lived up to itspromise as a means of ascertaining intermediateand weak natural relationships. Perhaps this isnot really surprising, for the two requirementsfor establishing an enzymatically effective hy-brid molecule, heterologous subunit binding,and effective active site formation or facilita-tion, may both be drastically influenced by thechange of a few critical amino acid residues, yetmay be largely unaffected by many of thesubstitutions occurring elsewhere in the mole-cule during evolution. When it is further real-ized that most of these experiments are per-formed by mixing crude extracts of mutantswith lesions that may be missense or nonsenseand whose content of proteases or other inhibi-tors for the foreign subunits is unknown, it is nosurprise that the results seem somewhat irregu-lar.

Immunological Cross-ReactivityThere is a body of work (167, 186) with several

proteins from eukaryotes showing a correlationbetween strength of immunological cross-reac-tion and amino acid sequence resemblence,therefore evolutionary distance. This approachto the determination of prokaryotic naturalrelationships would appear to be less timeconsuming than the determination of manyamino acid sequences. Once proteins from sev-

eral different organisms have been purified andused to generate antisera, many other orga-nisms can be tested against these reference serafor the extent of cross-reaction of the homolo-gous protein in their crude extracts. As yet nosystematic interfamily survey of any of thetryptophan enzymes has been done, but scat-tered observations have been made. These willbe reviewed briefly, for they do lead to certainconclusions.The immunological cross-reactivities of the a

and #2 subunits of TS in the enteric bacteriahave been studied with some care (161, 178). Noantigenic determinants are shared by the trpAand trpB gene products. When the homologousreactions are considered, it can be seen that,with respect to the a-chains where the sequence

is known, the degree of cross-reaction demon-strated by complement fixation is in agreementwith sequence divergence and also agrees withimmunological studies of the unrelated proteinalkaline phosphatase from the same species(212). Enzyme neutralization tests performed

with the same sera showed a stronger cross-reac-tion than microcomplement fixation, suggestingthat this might be a more sensitive test fordetecting distant evolutionary relationships(161). When the same enteric bacterial specieswere studied systematically for the immunolog-ical relatedness of their TS #2 subunits, thesomewhat surprising result was that the differ-ences found, while in proportion to those of thea-chains, were much less (178). Enzyme neu-tralization results were rather similar for thetwo subunits, however. These results are sum-marized in Table 1.The other immunological results available

with this system represent more sporadic effortsto compare tryptophan synthases from differenttaxa. Weak neutralization and precipitationreactions were found when the P. putida #2subunit was tested with antiserum prepared tothe E. coli #2 subunit (151); no reciprocalneutralization of E. coli enzyme was found withP. putida antiserum, but this result deservesreinvestigation. B. subtilis #2 subunits areweakly neutralized by antisera prepared toeither E. coli or P. putidaf#2 subunits (88). Therespective neutralization titers are only 1.5 and3.5% that of the homologous system, however. Ithas also been reported that the B (= #2?)subunit of Anabaena is neutralized by S. typhi-murium antiserum at 2.9% of the homologoustiter, whereas the Chlorella enzyme does notcross-react with the enteric bacterial system atall (184). The significance of the extremely weak(0.05%) cross-reaction between the B (= 2?)subunit of the Chlorella two-component TS andantiserum prepared to the N. crassa single

TABLE 1. Immunological cross-reactivity of thetryptophan synthetase a and f2 subunits from five

enteric bacteria

Index of NeutralizationOrganism dissimilaritya efficiencyb

a 2 a 2

E. coli 1.0 1.0 1.0 1.0S. dysenteriae 1.1 1.0 1.0 1.02S. typhimuriu 2.2 1.4 1.25 1.49E. aerogenes 2.8 1.2 1.39 1.64S. marcescens 8.4 1.8 1.89 2.38

a Ratio of antiserum concentrations required withhomologous (E. coli) and cross-reacting antigens togive identical levels of complement fixation. Data fora subunits from (161) and for 02 subunits fromreference 178.

SRatio (homologous/heterologous antigen) of thenumber of enzyme units neutralized by a givenamount of antiserum made to the E. coli subunit.Data for a subunits from (161) and for 02 subunitsfrom reference 178.

VOL. 39, 1975


http://mm

br.asm.org/

Dow

nloaded from


CRAWFORD

component one (184) remains problematical,but is remiscent of the weak neutralization of B.subtilis fh subunits by N. crassa antiserumreported by Hoch (88).The conclusion to be drawn from these immu-

nological studies is that the easily quantifiablemethods of complement fixation and precipita-tion which have proven most useful in taxo-nomic studies of higher organisms (167, 186)may be too sensitive for determining the naturalrelationships between the major bacterialgroups. In this respect they resemble nucleicacid hybridization techniques; like them theywill be of considerable use in assignment ofindividual organisms to positions within theirgroup, however. The use of enzyme activityneutralization appears to hold promise for as-certaining group relationships, but must beapplied on a broader scale and more systemati-cally than has been done heretofore. Relation-ships determined in this way should also beconfirmed by amino acid sequencing, at least insome cases, for the possibility exists that low-level cross-neutralization may occur when thetwo enzymes in question really have very littleoverall sequence similarity but possess a similarorientation of cofactors and essential aminoacid side chains at the active site. In thisconnection it is worth noting that there are twounrelated families of proteases sharing the"active serine" type of active site; though thecoupled amino acid side chains that theseenzymes use in catalysis are present in identicalorientations, the remainder of the moleculeshows no sequence homology (148). With onlylimited data available, there is no evidence sofar for any similar convergent lines of evolutionfor any of the tryptophan pathway enzymes,though the circumstances surrounding thesmall, glutamine amidotransferase subunits ofAS and p-aminobenzoate synthase describedearlier deserve closer study.

Regulatory MechanismsAs with immunological and subunit inter-

change experiments, investigators have hopedto avoid purifying and sequencing many pro-teins in ascertaining relationships by studyingregulatory phenomena. To a certain extent thishas been successful. As described earlier, nearlyevery major taxonomic group investigated dif-fers from the others in some aspect of themechanism by which it controls, or does notcontrol, the synthesis of the enzymes of thetryptophan pathway. In a like fashion, themeans by which the various prokaryotes andfungi regulate the activity and synthesis of thefirst enzyme of the common aromatic pathway,

3-deoxy-D-arabino-heptulosonate 7-phosphatesynthase, has proved to be characteristic andquite varied (71, 111, 164). Although it mayprove possible to correlate these patterns withgenetic and physiological differences and toassign controversial organisms to one or anotherfamily by their use - as in the case of Aeromo-nas, whose tryptophan enzymes are those of atypical enteric bacterium rather than a pseudo-monad (43, 124) - it is not generally possible todetermine the natural relationships of differentfamilies in this way. Moreover, elucidation ofthe regulatory mechanisms used by a certaintaxon usually requires a time-consuming study,surveying under varied growth conditions thelevels of all the enzymes in auxotrophic andregulatory mutants as well as the wild type. Asimilar amount of effort devoted to the isolationof a protein and determination of its amino acidsequence is likely to produce information ofgreater taxonomic value.

EVOLUTIONARY TRENDS

Fused versus Separate Genetic ElementsIn their paper published in 1965, Bonner et al.

(17), extrapolating from the information availa-ble about the tryptophan enzymes of en-teric bacteria, Neurospora, and a few otherorganisms, suggested a progression in evolution-ary development from independent proteinsthrough dependent, multimeric enzymes tofused genetic elements. In the ensuing decade,even though the examples of fused geneticelements in the tryptophan pathway have mul-tiplied, the direction of evolution no longerseems so obvious.The examples of gene fusion pointed out in

the preceding sections are: the TS of yeasts andfilamentous fungi, where the a- and 3-chains ofthe dependent multimeric bacterial enzymehave fused; the PRAI-InGPS fused polypeptideof the enteric bacteria and Neurospora, thoughother organisms appear to have independentenzymes; the small AS subunit-PRT fusion incertain enteric bacteria, though in all otherorganisms these two elements are independent;and a few other less definite but suggestivecases such as the large and small subunits of ASin Neurospora and yeast. The occurrence ofinterdependent complexes such as the AS-PRAI-InGPS in Neurospora and the AS-InGPSof yeast, and possibly the large complex per-forming all the reactions beyond AS in Euglena,also deserves to be mentioned. The evolutionaryadvantages supposedly accruing at successivestages in this progression would be conservation

BACTERIOL. REV.110


http://mm

br.asm.org/

Dow

nloaded from



of diffusible intermediates in the case of de-pendent complexes and increased stability andregulatory simplicity in the case of fused genes(17). It is apparent that these factors may havebeen effective in a few instances, but organismsthought to be most highly evolved, such assporeforming bacilli and blue green bacteriaamong the prokaryotes and plants among theeukaryotes, have not responded to their pres-sure.Gene fusion has taken place at various times

during evolution of the tryptophan pathway,probably by the same mechanism observed inthe laboratory in the case of two adjacenthistidine genes (223), deletion of the translationstop and start codons. However, if these eventsconfer a decisive advantage, one wonders why,in the course of evolution, some particular formof the fusions found has not replaced the sepa-rate genes in all subsequent organisms. Perhapsthere is also an advantage, from time to time atleast, in the flexibility conferred by separategenes, because they may more easily interactwith other metabolic pathways in the cell andmay be controlled independently.

Dependent Enzyme Aggregates versusIndependent Proteins

If in fact there was once an independent formof AS, where glutamine was deamidated by oneenzyme and the ammonia formed was used byanother, or of TS, where indole formed in the Areaction by one enzyme diffused freely to an-other enzyme to be converted to tryptophan, theorganisms bearing them have not survived orstill remain to be found. In these instancesaggregation, by associating two unlike activesites in spatial proximity, may have triumphedearly over the more primitive state. This de-pendency could just as well have been extendedto other reactions in the pathway. In the higherfungi the apparent dependence of AS on associ-ation with another polypeptide coding for activ-ities further down the pathway has been noted,and other instances will doubtless come to lightin the future. Still it is fair to note that in mostof the bacteria and in the higher plants noincreased dependence on aggregation is seen. Infact, it has been suggested that the TS ofAnabaena and Chlorella (184) as well as that ofhigher plants (36, 47, 162) is less aggregationdependent than that of bacterial because its B(= p3?) subunit catalyzes the TS-B reactionequally well in the presence or absence of the A(= a?) component, though the latter must bepresent for the conversion of InGP to trypto-phan. A word of caution must be injected here,however, for the same claim was made earlier

for the B. subtilis enzyme (153, 192). Uponstabilizing and purifying the latter enzyme tohomogeneity, Hoch (87, 88) found little residualTS-B activity catalyzed by the pure fl2 subunit,though there was marked stimulation of thisactivity upon combination with the a subunit.Probably the earlier conclusions resulted fromsome a-chains remaining bound to the impure"B" fraction. Whatever the reason, these stud-ies underscore the universal preservation in theTS complex of the ability to utilize indole inplace of InGP, even in these highly evolvedorganisms. Similarly, the AS of all organismsappears to have preserved the ability to utilizeammonia in place of glutamine, albeit undersomewhat unphysiological conditions. In thefirst case, at least, a case can be made for theusefulness of this activity as a "salvage" opera-tion in cases where indole molecules becomeavailable.

One Operon or SeveralThere seem to be few organisms having all

their tryptophan genes in a single operon. Out-side the Enterobacteriaceae, only Staphylococ-cus appears to have a good chance to enjoy thesingle operon arrangement; even there, informa-tion on the small AS subunit is lacking and sothe B. subtilis situation may prevail. It is notclear why having all the genes for this ratheruncomplicated pathway in a single operon is notthe evolutionarily preferred solution. Two factsseem obvious, however. Translocation of singlegenes or multigenic portions of a trp genecluster are rather common events; recentlyJackson and Yanofsky (109) described a selec-tive system in E. coli in which duplications andtranslocations of multigenic segments of the trpoperon to another part of the chromosome couldbe sought and their frequency could be deter-mined. These events were found to be surpris-ingly frequent. Secondly, once genes have beentranslocated to a new region of the chromosomean organism seems to have little difficulty indesigning or converting regulatory elements fortheir control. This may account for the regula-tory variety seen up to now primarily amonggram-negative bacteria, but probably to beexpected throughout the prokaryotic and eu-karyotic kingdoms.

Predictions Concerning Future StudiesTwo sets of predictions can be made, one

concerning genetic and biochemical studies de-sirable to increase our understanding of thetryptophan synthetic apparatus, and the otherconcerning the use of the tryptophan genes and

VOL. 39, 1975 ill


http://mm

br.asm.org/

Dow

nloaded from


BACrERIOL. REV.

enzymes in determining taxonomic relation-ships.

In the former case, the dependent enzymeaggregates in this pathway, both in prokaryotesand eukaryotes, seem sure to be the objects ofcontinued study. Among the features makingthem attractive to students of enzyme mechan-ics are their relative simplicity and low molecu-lar weight, the fact that organisms can begenetically tricked into synthesizing them inlarge amounts, the availability of many mutantforms of them, and the opportunity to constructfunctioning complexes with subunits derivedfrom different organisms. The TS from E. colican be crystallized in the a2,,% complex (1) aswell as in individual subunit forms (2, 190),leading one to hope for eventual three-dimen-sional structural analysis of this multimericenzyme. Among its many attractive features arethe variety of amino acid substitutions availa-ble at certain sites in the a chain. At position211, for example, 15 different amino acids havebeen inserted by mutation or suppression, eightof which are functional and seven are not(160a). One totally unanswered question con-cerning TS is the rationale for its occurrence asan aA2r tetramer instead of an a3 dimer. Nocooperative interactions between the two identi-cal active sites have been detected, thoughthese have been sought by several techniques(41, 46, 50, 118, 216).The cross-pathway relationships described

for the glutamine amidotransferases of the tryp-tophan and folate pathways in Bacillus andAcinetobacter seem destined to be studied insome detail as examples of metabolic inter-dependence. They may be just an example ofmany similar events involving enzymes of re-lated function in different pathways. Similarly,the study of regulatory variations found in thetryptophan pathway of various organismsshould be extended in breath and depth toclarify the possibilities for control of gene ex-pression after translocation or duplication ofstructural gene elements.

In the second case the future is less easilypredicted. I have tried to make a case for the useof amino acid sequences in determining thepath of evolution in the prokaryotes, using thefragmentary sequence information presentlyavailable from the tryptophan enzymes as myvehicle. It seems clear to me that the resultsobtained so far encourage additional studies,and the wide variety of organisms with trypto-phan enzymes that have not been studiedsuggests that some amplification of effort isneeded. Even organisms with a natural require-ment for tryptophan may have preserved por-

tions of the pathway that can be studied. Quitefrankly, it must be admitted that other proteinsmay become more important than the trypto-phan pathway enzymes in this type of endeavorin the future, but this pathway should find aplace among the pioneer systems used for thestudy of prokaryotic evolution, just as it hasamong the early models for the study of enzymeaggregates.

ACKNOWLEDGEMENTSI am indebted to the National Science Foundation

(grants GB-4267, GB-6841, and GB-16388) and theNational Institutes of Health (grant AM-13224 fromthe National Institute of Arthritis, Metabolism, andDigestive Diseases, and grant GM-19098 from theNational Institute of General Medical Sciences) forfinding portions of our work worthy of grant support.

For that portion of this work emanating from mylaboratory, I am indebted to the students and col-laborators who shared in its joys and disappoint-ments. Their identity will be obvious from perusal ofthe reference list. My colleagues Sallie 0. Hoch,Junetsu Ito, and Stanley E. Mills gave generously oftheir time to read and criticize the manuscript.Lastly, I wish to take this opportunity to express heremy gratitude to two superb teachers, C. B. van Niel,who introduced me to the microbes, and CharlesYanofsky, who has helped me to discern the mysteriesof their methods for synthesizing tryptophan.

LITERATURE CITED1. Adachi, O., L. D. Kohn, and E. W. Miles. 1974.

Crystalline a2.2 complexes of tryptophansynthetase of Escherichia coli: a compar-ison between the native complex and thereconstituted complex. J. Biol. Chem.249:7756-7763.

2. Adachi, O., and E. W. Miles. 1974. A rapidmethod for preparing crystalline #2 subunit oftryptophan synthetase of Escherichia coli inhigh yield. J. Biol. Chem. 249:5430-5434.

3. Ahmad, M., A. Mozmadar, and S. H-endler.1968. A new locus in the tryptophan pathwayof Neurospora crassa. Genet. Res. 12:103-107.

4. Anagnostopoulos, C., and I. P. Crawford. 1961.Transformation studies on the linkage ofmarkers in the tryptophan pathway in Bacil-lus subtilis. Proc. Natl. Acad. Sci. U.S.A.47:378-390.

5. Anagnostopoulos, C., and I. P. Crawford. 1967.Le groupe de genes regissant la biosynthesedu tryptophan chez Bacillus subtilis. C. R.Acad. Sci. Paris 265:93-96.

6. Balbinder, E. 1964. Intergeneric complementa-tion between A and B components of bacterialtryptophan synthetases. Biochem. Biophys.Res. Commun. 17:770-774.

7. Balbinder, E., A. J. Blume, A. Weber, and H.Tamaki. 1968. Polar and anti-polar mutantsin the tryptophan operon of Salmonella typhi-murium. J. Bacteriol. 95:2217-2229.

112 CRAWFORD


http://mm

br.asm.org/

Dow

nloaded from



8. Balbinder, E., R. Callahan, P. P. McCann, J.C. Cordaro, A. R. Weber, A. M. Smith, and F.Angelosanto. 1970. Regulatory mutants of thetryptophan operon of Salmonella typhimu-rium. Genetics 66:31-53.

9. Baskerville, E. N., and R. Twarog. 1972. Regu-lation of the tryptophan synthetic enzymes inClostridium butyricum. J. Bacteriol. 112:304-314.

10. Bauerle, R. H., and R. Margolin. 1966. Thefunctional organization of the tryptophan genecluster in Salmonella typhimurium. Proc.Natl. Acad. Sci. U.S.A. 56:111-118.

11. Bauerle, R. H., and P. Margolin. 1966. A mul-tifunctional enzyme complex in the trypto-phan pathway of Salmonella typhimurium:comparison of polarity and pseudopolaritymutations. Cold Spring Harbor Symp. Quant.Biol. 31:203-214.

12. Baumann, P., M. Doudoroff, and R. Y. Stanier.1968. A study of the Moraxella group. I. GenusMoraxella and the Neisseria catarrhalis group.J. Bacteriol. 95:58-73.

13. Baumann, P., M. Doudoroff, and R. Y. Stanier.1968. A study of the Moraxella group. II.Oxidative-negative species (genus Acineto-bacter). J. Bacteriol. 95:1520-1541.

14. Belser, W. L., J. Baron-Murphy, D. P. Delmer,and S. E. Mills. 1971. End product control oftryptophan biosynthesis in extracts and intactcells of the higher plant Nicotiana tabacumvar. Wisconsin 38. Biochem. Biophys. Acta273:1-10.

14a. Berlyn, M. B., S. I. Ahmed, and N. H. Giles.1970. Organization of polyaromatic biosyn-thetic enzymes in a variety of photosyntheticorganisms. J. Bacteriol. 104:768-774.

15. Berlyn, M. B., and N. H. Giles. 1973. Organiza-tion of interrelated aromatic metabolic path-way enzymes in Acinetobacter calcoaceticus.J. Gen. Microbiol. 74:337-341.

16. Blume, A. J., A. Weber, and E. Balbinder. 1968.Analysis of polar and nonpolar tryptophanmutants by derepression kinetics. J. Bacteriol.95:2230-2241.

17. Bonner, D. M., J. A. DeMoss, and S. E. Mills.1965. The evolution of an enzyme, p. 305-318.In V. Bryson and H. J. Vogel (ed.), Evolvinggenes and proteins. Academic Press Inc., NewYork.

18. Bovre, K. 1963. Affinities between Moraxellaspp. and a strain of Neisseria catarrhalis asexpressed by transformation. Acta. Pathol.Microbiol. Scand. 58:528-529.

19. Bovre, K. 1967. Transformation and DNA com-position in taxonomy, with special reference torecent studies in Moraxella and Neisseria.Acta Pathol. Microbiol. Scand. 69:123-144.

20. B5vre, K., and S. D. Henriksen. 1962. An ap-proach to transformation studies in Moraxel-la. Acta Pathol. Microbiol. Scand. 56:223-228.

21. Brammer, W. J. 1973. Independent expression ofthe A gene of the tryptophan operon of Esche-richia coli during tryptophan starvation. J.Gen. Microbiol. 76:395-405.

21a. Bronson, M. J., C. Squires, and C. Yanofsky.1973. Nucleotide sequences from tryptophanmessenger RNA of Escherichia coli: the se-quence corresponding to the amino-terminalregion of the first polypeptide specified by theoperon. Proc. Natl. Acad. Sci. U.S.A.70:2335-2339.

22. Burkholder, P. R., and N. H. Giles. 1947. In-duced biochemical mutations in Bacillus sub-tilis. Am. J. Bot. 34:345-348.

23. Calhoun, D. H., D. L. Pierson, and R. A. Jensen.1973. The regulation of tryptophan biosynthe-sis in Pseudomonas aeruginosa. Mol. Gen.Genet. 121:117-132.

24. Callahan, R., and E. Balbinder. 1970. Trypto-phan operon: structural gene mutation creat-ing a "promoter" and leading to 5-methyltryp-tophan dependence. Science 168:1586-1589.

25. Carlton, B. C., and D. D. Whitt. 1969. Theisolation and genetic characterization of mu-tants of the tryptophan system of B. subtilis.Genetics 62:445-460.

26. Carsiotis, M., E. Appela, P. Provost, J. Germer-shausen, and S. R. Suskind. 1965. Chemicaland physical studies of the structure of trypto-phan synthetase from Neurospora crassa. Bio-chem. Biophys. Res. Commun. 18:877-888.

27. Carsiotis, M., and R. F. Jones. 1974. Cross-path-way regulation: tryptophan-mediated controlof histidine and arginine biosynthetic enzymesin Neurospora crassa. J. Bacteriol. 119:889-892.

28. Carsiotis, M., R. F. Jones, A. M. Lacy, T. J.Cleary, and D. B. Fankhauser. 1970. Histi-dine-mediated control of tryptophan biosyn-thetic enzymes in Neurospora crassa. J. Bac-teriol. 104:98-106.

29. Carsiotis, M., R. F. Jones, and A. C. Wesseling.1974. Cross-pathway regulation: histidine-mediated control of histidine, tryptophan, andarginine biosynthetic enzymes in Neurosporacrassa. J. Bacteriol. 119:893-898.

30. Carsiotis, M., and A. M. Lacy. 1965. Increasedactivity of tryptophan biosynthetic enzymesin histidine mutants of Neurospora crassa. J.Bacteriol. 89:1472-1477.

31. Casse, F., M. C. Pascal, and M. Chippaux. 1973.Comparison between the chromosomal mapsof Escherichia coli and Salmonella typhimu-rium. Length of the inverted region. Mol. Gen.Genet. 124:253-257.

32. Catena, A., and R. D. DeMoss. 1970. Physiologi-cal and kinetic studies with anthranilate syn-thetase of Bacillus alvei. J. Bacteriol.101:476-482.

33. Catlin, B. W. 1964. Reciprocal transformationbetween Neisseria catarrhalis and Moraxellanonliquefaciens. J. Gen. Microbiol. 37:369-379.

34. Catlin, B. W., and L. S. Cunningham. 1961.Transforming activities and base contents ofdeoxyribonucleates from various Neisseria. J.Gen. Microbiol. 26:303-312.

35. Chakrabarty, A. M. 1972. Genetic basis of thebiodegradation of salicylate in Pseudomonas.

113VOL. 39, 1975


http://mm

br.asm.org/

Dow

nloaded from


BACTERIOL. REV.

J. Bacteriol. 112:815-823.36. Chen, J., and W. G. Boll. 1972. Tryptophan

synthetase: partial purification and someproperties of the B-protein subunit from peaplants (Pisum sativum cv. Alaska). Can. J.Bot. 50:587-594.

37. Chilton, M. D., and B. J. McCarthy. 1969.Genetic and base sequence homologies inbacilli. Genetics 62:697-710.

38. Cohen, G., and F. Jacob. 1959. Sur la repressionde la synthese des enzymes intervenant dansla formation du tryptophane chez Escherichiacoli. C. R. Acad. Sci. Paris 248:3490-3492.

39. Cohn, W., and I. P. Crawford. 1975. Regulationof enzyme synthesis in the tryptophan path-way of Acinetobacter calcoaceticus. Abstr.Annu. Meet. Amer. Soc. Microbiol. 1975, K2,p. 147.

40. Colwell, R. R. 1973. Genetic and phenetic classi-fication of bacteria. Adv. App. Microbiol.16:137-175.

41. Crawford, I. P. 1974. Tryptophan synthetase, p.223-265. In K. Ebner (ed.), Subunit enzymes.Marcel Dekker, New York.

42. Crawford, I. P., and I. C. Gunsalus. 1966. Induci-bility of tryptophan synthetase in Pseudomo-nas putida. Proc. Natl. Acad. Sci. U.S.A.56:717-724.

43. Crawford, I. P., S. Sikes, and D. K. Melhorn.1967. The natural relationships of Aeromonasformicans. Arch. Mikrobiol. 59:72-81.

44. Crawford, I. P., and C. Yanofsky. 1971. Pseu-domonas putida tryptophan synthetase: par-tial sequence of the a subunit. J. Bacteriol.108:248-253.

45. Creighton, T. E. 1970. N-(5'-phosphoribosyl)an-thranilate isomerase-indole-3-ylglycerol phos-phate synthetase of tryptophan biosynthesis.Relationship between the two activities of theenzyme from Escherichia coli. Biochem. J.120:699-707.

46. Creighton, T. E. 1970. A steady-state kineticinvestigation of the reaction mechanism of thetryptophan synthetase of Escherichia coli.Eur. J. Biochem. 13:1-10.

47. Delmer, D. P., and S. E. Mills. 1968. Trypto-phan synthetase from Nicotiana tabacum.Biochem. Biophys. Acta. 167:431-443.

48. Delmer, D. P., and S. E. Mills. 1968. Tryptophanbiosynthesis in cell cultures of Nicotiana taba-cum. Plant Physiol. 43:81-87.

49. Demerec, M., and Z. Hartman. 1956. Trypto-phan mutants in Salmonella typhimurium p.5-13. In Genetic studies with bacteria. Car-negie Inst. Wash. Publ. 612.

50. DeMoss, J. A. 1962. Studies on the mechanismof the tryptophan synthetase reaction. Bio-chem. Biophys. Acta 62:279-293.

51. DeMoss, J. A. 1965. Biochemical diversity in thetryptophan pathway. Biochem. Biophys. Res.Commun. 18:850-856.

52. DeMoss, J. A. 1965. The conversion of shikimicacid to anthranilic acid by extracts of Neuro-spora crassa. J. Biol. Chem. 240:1231-1235.

53. DeMoss, R. D., and N. R. Evans. 1960. Incorpo-

ration of C'4-labeled substrates into violacein.J. Bacteriol. 79:729-733.

54. DeMoss, J. A., R. W. Jackson, and J. H.Chalmers. 1967. Genetic control of the struc-ture and activity of an enzyme aggregate inthe tryptophan pathway of Neurospora crassa.Genetics 56:413-424.

55. Denney, R. M., and C. Yanofsky. 1972. Detec-tion of tryptophan messenger RNA in severalbacterial species and examination of the prop-erties of heterologous DNA-RNA hybrids. J.Mol. Biol. 64:319-339.

56. Doolittle, W. F., and C. Yanofsky. 1968. Mu-tants of Escherichia coli with an altered tryp-tophanyl-transfer ribonucleic acid synthetase.J. Bacteriol. 95:1283-1294.

57. Doy, C. H., and J. M. Cooper. 1966. Aromaticbiosynthesis in yeast. I. The synthesis oftryptophan and the regulation of this path-way. Biochem. Biophys. Acta 127:302-316.

58. Dubnau, D., I. Smith, P. Morell, and J. Mar-mur. 1965. Gene conservation in Bacillus spe-cies. I. Conserved genetic and nucleic acidbase sequence homologies. Proc. Natl. Acad.Sci. U.S.A. 54:491-498.

59. DuBuy, B., and S. M. Weissman. 1971. Nucleo-tide sequence of Pseudomonas fluorescens 5 Sribonucleic acid. J. Biol. Chem. 246:747-761.

60. Dunn, N. W., and I. C. Gunsalus. 1973. Trans-missible plasmid coding early enzymes ofnaphthalene oxidation in Pseudomonas pu-tida. J. Bacteriol. 114:974-979.

61. Egan, A. F., and F. Gibson. 1966. Anthranilatesynthetase and PR-transferase from Aerobac-ter aerogenes as a protein aggregate. Biochem.Biophys. Acta 130:276-277.

62. Eisenstein, R. B., and C. Yanofsky. 1962. Tryp-tophan synthetase levels in Escherichia coli,Shigella dysenteriae, and transduction hy-brids. J. Bacteriol. 83:193-204.

63. Enatsu, T., and I. P. Crawford. 1968. Enzymes ofthe tryptophan pathway in Pseudomonas pu-tida. J. Bacteriol. 95:107-112.

64. Engel, P. P. 1973. Genetic control of tryptophanbiosynthesis in Streptomyces coelicolor, p.125-147. In Z. Vanek, Z. Hostalek, and J.Cudlfn (ed.), Genetics of industrial microor-ganisms: actinomyces and fungi, vol. 2. Aca-demia, Prague.

65. Fargie, B., and B. W. Holloway. 1965. Absenceof clustering of functionally related genes inPseudomonas aeruginosa. Genet. Res. 6:284-299.

66. Fildes, P. 1940. Indole as a precursor in thesynthesis of tryptophan by bacteria. Br. J.Exp. Pathol. 21:315-319.

67. Fildes, P., G. P. Gladstone, and B. C. J. G.Knight. 1933. The nitrogen and vitamin re-quirements of B. typhosus. Br. J. Exp. Pa-thol. 14:189-196.

68. Fink, G. R., and R. G. Martin. 1967. Translationand polarity in the histidine operon. II. Polar-ity in the histidine operon. J. Mol. Biol.30:97-107.

69. Gaertner, F. H., and J. A. DeMoss. 1969. Purifi-

114 CRAWFORD


http://mm

br.asm.org/

Dow

nloaded from



cation and characterization of a multienzymecomplex in the tryptophan pathway of Neuro-spora crassa. J. Biol. Chem. 244:2716-2725.

70. Gaertner, F. H., M. C. Ericson, and J. A.DeMoss. 1970. Catalytic facilitation in vitroby two multienzyme complexes from Neuro-spora crassa. J. Biol. Chem. 245:595-600.

71. Gibson, F., and J. Pittard. 1968. Pathways ofbiosynthesis of aromatic amino acids andvitamins and their control in microorganisms.Bacteriol. Rev. 32:465-492.

72. Gibson, F., J. Pittard, and E. Reich. 1967.Ammonium ions as the source of nitrogen fortryptophan biosynthesis in whole cells ofEscherichia coli. Biochem. Biophys. Acta136:573-576.

73. Gibson, M. I., and F. Gibson. 1964. Preliminarystudies on the isolation and metabolism of anintermediate in aromatic biosynthesis: choris-mic acid. Biochem. J. 90:248-256.

74. Goldberger, R. F. 1974. Autogenous regulation ofgene expression. Science 183:810-816.

75. Grieshaber, M., and R. Bauerle. 1972. Structureand evolution of a bifunctional enzyme of thetryptophan operon. Nature (London) NewBiol. 236:232-235.

76. Guest, J. R., G. R. Drapeau, B. C. Carlton, andC. Yanofsky. 1967. The amino acid sequenceof the A protein (a subunit) of the tryptophansynthetase of Escherichia coli. V. Order oftryptic peptides and the complete amino acidsequence. J. Biol. Chem. 242:5442-5446.

77. Gunsalus, I. C., C. F. Gunsalus, A. M. Chakrab-arty, S. Sikes, and I. P. Crawford. 1968. Finestructure mapping of the tryptophan genes inPseudomonas putida. Genetics 60:419-435.

78. Gunsalus, I. C., and V. P. Marshall. 1971.Monoterpene dissimilation: chemical and ge-netic models. Crit. Rev. Microbiol. 1:291-310.

79. Helinski, D. R. 1973. Plasmid determined resist-ance to antibiotics: molecular properties of Rfactors. Annu. Rev. Microbiol. 27:437-470.

80. Henke, H., J. L. Guerdoux, and R. Hiitter. 1973.Relations entre les genes et les enzymes de labiosynthese du tryptophane chez Coprinusradiatus Fr. ex Bolt. C. R. Acad. Sci. Paris276:2001-2004.

81. Henriksen, S. D. 1973. Moraxella, Acinetobac-ter and the Mimrae. Bacteriol. Rev. 37:522-561.

82. Herman, N. J., and E. Juni. 1974. Isolationand characterization of a generalized trans-ducing bacteriophage for Acinetobacter. J.Virol. 13:46-52.

83. Hershfield, V., H. W. Boyer, C. Yanofsky, M. A.Lovett, and D. R. Helinski. 1974. Plasmid ColEl as a molecular vehicle for cloning andamplification of DNA. Proc. Natl. Acad. Sci.U.S.A. 71:3455-3459.

84. Hiraga, S. 1969. Operator mutants of the trypto-phan operon in Escherichia coli. J. Mol. Biol.39:159-179.

85. Hiraga, S., K. Ito, T. Matsuyama, H. Ozaki, andT. Yura. 1968. 5-Methyltryptophan-resistantmutations linked with the arginine G marker

in Escherichia coli. J. Bacteriol. 96:1880-1881.86. Hiraga, S., and C. Yanofsky. 1972. Hyper-labile

messenger RNA in polar mutants of the tryp-tophan operon of Escherichia coli. J. Mol. Biol.72:102-110.

87. Hoch, S. 0. 1973. Tryptophan synthetase fromBacillus subtilis: purification and charac-terization of the a component. J. Biol. Chem.248:2999-3003.

88. Hoch, S. 0. 1973. Tryptophan synthetase fromBacillus subtilis: purification and charac-ization of the 02 component. J. Biol. Chem.248:2992-2998.

88a. Hoch, S. 0. 1974. Mapping of the 5-methyl-tryptophan resistance locus in Bacillus sub-tilis. J. Bacteriol. 117:315-317.

89. Hoch, S. O., C. Anagnostopoulos, and I. P.Crawford. 1969. Enzymes of the tryptophanoperon of Bacillus subtilis. Biochem. Biophys.Res. Commun. 35:838-844.

90. Hoch, S. O., and I. P. Crawford. 1973. Enzymesof the tryptophan pathway in three Bacillusspecies. J. Bacteriol. 116:685-693.

91. Hoch, S. O., C. W. Roth, I. P. Crawford, and E.W. Nester. 1971. Control of tryptophan bio-synthesis by the methyltryptophan resistancegene in Bacillus subtilis. J. Bacteriol. 105:38-45.

92. Hogness, D. S., and H. K. Mitchell. 1954.Genetic factors influencing the activity oftryptophan desmolase in Neurospora crassa.J. Biol. Chem. 150:325-333.

93. Holloway, B. W., V. Krishnapillai, and V. Stani-sich. 1971. Pseudomonas genetics. Annu. Rev.Genet. 5:425-446.

94. Hopwood, D. A., K. F. Chater, J. E. Dowding,and A. Vivian. 1973. Advances in Strep-tomyces coelicolor genetics. Bacteriol. Rev.37:371-405.

95. Hulett, F. M., and J. A. DeMoss. 1973. Thesubunit structure of the anthranilate synthe-tase complex of Neurospora. Abstr. Annu.Meet. Amer. Soc. Microbiol., P 47, p. 148.

96. Hutchinson, M. A., and W. L. Belser. 1969.Enzymes of tryptophan biosynthesis in Serra-tia marcescens. J. Bacteriol. 98:109-115.

97. Hutter, R. 1973. Regulation of the tryptophanbiosynthetic enzymes in fungi, p. 109-124. InZ. Vanek, Z. Hostalek, and J. Cudlin (ed.),Genetics of industrial microorganisms. Aca-demia, Prague.

98. Hutter, R., and J. A. DeMoss. 1967. Enzymeanalysis of the tryptophan pathway in Asper-gillus nidulans. Genetics 55:241-247.

99. Hiutter, R., and J. A. DeMoss. 1967. Organiza-tion of the tryptophan pathway: a phyloge-netic study of the fungi. J. Bacteriol.94:1896-1907.

100. Imamoto, F. 1973. Translation and transcriptionof the tryptophan operon, p. 339-407. In J. N.Davidson and W. E. Cohn (ed.), Progress innucleic acid research, vol. 13. Academic PressInc., New York.

101. Imamoto, F., N. Morikawa, S. Sato, S. Mi-shima, T. Nishimura, and A. Matsushiro.

115VOL. 39, 1975


http://mm

br.asm.org/

Dow

nloaded from


BACTERIOL. REV.

1965. On the transcription of the tryptophanoperon in Escherichia coli. II. Production ofthe specific messenger RNA. J. Mol. Biol.13:157-168.

102. Ingram, L. O., D. Pierson, J. F. Kane, C. vanBaalen, and R. A. Jensen. 1972. Documenta-tion of auxotrophic mutation in blue-greenbacteria: characterization of a tryptophanauxotroph in Agmenellum quadraplicatum. J.Bacteriol. 111:112-118.

103. Isaac, J. H., and B. W. Holloway. 1968. Controlof pyrimidine biosynthesis in Pseudomonasaeruginosa. J. Bacteriol. 88:273-278.

104. Ito, J. 1969. Hybrid anthranilate synthetasemolecules from enterobacterial sources. Na-ture (London) 223:57-59.

104a. Ito, J. 1970. Effect of tryptophan starvation onthe rate of translation of the tryptophanoperon in Escherichia coli. Biochem. Biophys.Acta 204:624-626.

105. Ito, J., and I. P. Crawford. 1965. Regulation ofthe enzymes of the tryptophan pathway inEscherichia coli. Genetics 52:1303-1316.

106. Ito, J., and C. Yanofsky. 1969. Anthranilatesynthetase, an enzyme specified by the trypto-phan operon of Escherichia coli: comparativestudies on the complex and its subunits. J.Bacteriol. 97:734-742.

107. Ito. K. 1972. Regulatory mechanism of the tryp-tophan operon in Escherichia coli: possibleinteraction between trpR and trpS gene prod-ucts. Mol. Gen. Genet. 115:349-363.

108. Jackson, E. N., and C. Yanofsky. 1972. Internalpromoter of the tryptophan operon of Esche-richia coli is located in a structural gene. J.Mol. Biol. 69:306-313.

109. Jackson, E. N., and C. Yanofsky. 1973. Duplica-tion-translocations of tryptophan operongenes in Escherichia coli. J. Bacteriol.116:33-40.

110. Jackson, E. N., and C. Yanofsky. 1973. Theregion between the operator and first struc-tural gene of the tryptophan operon of Esche-richia coli may have a regulatory function. J.Mol. Biol. 76:89-101.

111. Jensen, R. A., and D. S. Nasser. 1968. Compara-tive regulation of isoenzymic 3-deoxy-D-arabino-hepulosonate 7-phosphate synthe-tases in microorganisms. J. Bacteriol. 95:188-196.

112. Juni, E. 1972. Interspecies transformation ofAcinetobacter: genetic evidence for a ubiqui-tous genus. J. Bacteriol. 112:917-931.

113. Juni, E., and A. Janik. 1969. Transformation ofAcinetobacter calcoaceticus (Bacterium ani-tratum). J. Bacteriol. 98:281-288.

114. Johnson, J. L., R. S. Anderson, and E. J. Ordal.1970. Nucleic acid homologies among oxidase-negative Moraxella species. J. Bacteriol.101:568-573.

115. Johnson, R. B., and C. G. Mays. 1954. A strain ofShigella dysenteriae requiring proline. J. Bac-teriol. 67:542-544.

116. Kane, J. F., W. M. Holmes, and R. A. Jensen.1972. Metabolic interlock: the dual function of

a folate pathway gene as an extra-operonicgene of tryptophan biosynthesis. J. Biol.Chem. 247:1587-1596.

117. Kane, J. F., and R. A. Jensen. 1970. Themolecular aggregation of anthranilate syn-thase in Bacillus subtilis. Biochem. Biophys.Res. Commun. 41:328-333.

117a. Kaplan, S., S. E. Mills, S. Ensign, and D. M.Bonner. 1964. Genetic determination of theantigenic specificity of tryptophan synthe-tase. J. Mol. Biol. 8:801-813.

118. Kirschner, K., and R. Wiskocil. 1972. Enzyme-enzyme interactions in tryptophan synthetasefrom E. coli, p. 245-258. In R. Jaenekke (ed.),Protein-protein interactions, 23rd Colloquiumder Gesellschaft fur Biologische Chemie.Springer-Verlag, Berlin.

118a. Kloos, W. E., and N. E. Rose. 1970. Transfor-mation mapping of tryptophan loci in Micro-coccus luteus. Genetics 66:595-605.

119. Kluyver, A. J., and C. B. van Niel. 1936.Prospects for a natural system of classificationof bacteria. Zentralbl. Bakteriol. Abt. II94:369-403.

120. Kuwano, M., D. Schlessinger, and D. E. Morse.1971. Loss of dispensible endonuclease activ-ity in relief of polarity of suA. Nature (Lon-don) New Biol. 231:214-217.

121. Lacy, A. M. 1965. Structural and physiologicalrelationships within the td locus in Neuro-spora crassa. Biochem. Biophys. Res. Com-mun. 18:812-823.

122. Lara, J. C., and S. E. Mills. 1972. Tryptophansynthetase in Euglena gracilis strain G. J.Bacteriol. 110:1100-1106.

123. Lara, J. C., and S. E. Mills. 1973. Regulation ofL-tryptophan biosynthesis in Euglena gracilis.Abstr. Annu. Meet. Amer. Soc. Microbiol.1973, P5, p. 141.

124. Largen, M. T. 1972. Tryptophan biosynthesisin the Enterobacteriaceae: some phylogeneticimplications. Ph.D. thesis, University of Cali-fornia, Riverside.

125. Largen, M.T., and W. L. Belser. 1973. The ap-parent conservation of the internal low effi-ciency promoter of the tryptophan operon ofseveral species of Enterobacteriaceae. Ge-netics 75:19-22.

125a. Largen, M.T., and W. L. Belser. 1975. Trypto-phan biosynthetic pathway in the Enterobac-teriaceae: some physical properties of the en-zymes. J. Bacteriol. 121:239-249.

126. Lederberg, J. 1947. The nutrition of Salmonella.Arch. Biochem. 13:287-290.

127. Leidigh, B. J., and M. L. Wheelis. 1973. Theclustering on the Pseudomonas putida chro-mosome of genes specifying dissimilatoryfunctions. J. Mol. Evol. 2:235-242.

128. Lester, G. 1961. Some aspects of tryptophansynthetase formation in Neurospora crassa.J. Bacteriol. 81:964-973.

129. Lester, G. 1968. In vivo regulation of intermedi-ate reactions in the pathway of tryptophanbiosynthesis in Neurospora crassa. J. Bac-teriol. 96:1768-1773.

116 CRAWFORD


http://mm

br.asm.org/

Dow

nloaded from



130. Lester, G. 1971. Regulation of the tryptophanbiosynthetic enzymes in Neurospora crassa.J. Bacteriol. 107:193-202.

131. Lorence, J. H., and E. W. Nester. 1967. Multiplemolecular forms of chorismate mutase. Bio-chemistry 6:1541-1553.

132. Li, S.-L., R. M. Denney, and C. Yanofsky. 1973.Nucleotide sequence divergence in the achain-structural genes of tryptophan synthe-tase from Escherichia coli, Salmonella typhi-murium and Aerobacter aerogenes. Proc.Natl. Acad. Sci. U.S.A. 70:1112-1116.

133. Li, S.-L., G. R. Drapeau, and C. Yanofsky. 1972.Amino terminal sequence of the tryptophansynthetase a chain of Serratia marcescens. J.Bacteriol. 113:1507-1508.

133a. Li, S.-L., J. Hanlon, and C. Yanofsky. 1974.Structural homology of the glutamine amido-transferase subunits of the anthranilate syn-thetases of Escherichia coli, Salmonella typhi-murium and Serratia marcescens. Nature(London) 248:48-49.

134. Li, S.-L., and S. 0. Hoch. 1974. Amino-terminalsequence of the tryptophan synthetase a-chain of Bacillus subtilis. J. Bacteriol. 118:187-191.

135. Li, S.-L., and C. Yanofsky. 1972. Amino acid se-quences of fifty residues from the amino ter-mini of the tryptophan synthetase a chains ofseveral enterobacteria. J. Biol. Chem. 247:1031-1037.

136. Li. S.-L., and C. Yanofsky. 1973. Amino acid se-quence studies with the tryptophan synthe-tase a chain of Salmonella typhimurium. J.Biol. Chem. 248:1830-1836.

137. Li, S.-L., and C. Yanofsky. 1973. Amino acid se-quence studies with the tryptophan synthe-tase a chain of Aerobacter aerogenes. J. Biol.Chem. 248:1837-1843.

138. McPartland, A., and R. L. Sommerville. 1974.Isolation and properties of new high efficiencypromoters. Genetics 77:s43.

139. McQuade, J. F., and T. E. Creighton. 1970.Purification and comparisons of the N-(5'-phosphoribosyl) anthranilic acid isomerase-indole-3-glycerol phosphate synthetase oftryptophan biosynthesis from three species ofEnterobacteriaceae. Eur. J. Biochem. 16:199-207.

140. Mandel, M. 1969. New approaches to bacterialtaxonomy: prespectives and prospects. Annu.Rev. Microbiol. 23:239-274.

141. Manney, T. R. 1968. Evidence for chain termina-tion by super-suppressible mutants in yeast.Genetics 60:719-733.

142. Manney, T. R. 1968. Regulation of factors thatinfluence in vitro stability of tryptophan syn-thetase from yeast. J. Bacteriol. 96:403-408.

143. Manney, T. R., W. Duntze, N. Janosko, and J.Salazar. 1969. Genetic and biochemical stud-ies of partially active tryptophan synthetasemutants of Saccharomyces cerevisiae. J. Bac-teriol. 99:590-596.

144. Margoliash, E., and W. M. Fitch. 1967. Phyloge-netic trees based on a matrix of differences of

amino acid sequences. Science 155:279-284.145. Margolin, P. 1967. Genetic polarity in bacteria.

Am. Nat. 101:301-312.146. Margolin, P. 1971. Regulation of tryptophan

synthesis, p. 389-446. In H. J. Vogel (ed.),Metabolic pathways, vol. 5. Academic PressInc., New York.

147. Marinus, M. G., and J. S. Loutit. 1969. Regula-tion of isoleucine-valine biosynthesis in Pseu-domonas aeruginosa. H. Regulation of en-zyme activity and synthesis. Genetics 63:557-567.

148. Markland, F. S., and E. L. Smith. 1971. Sub-tilisins: primary structure, chemical andphysical properties, p. 561-608. In P. D. Boyer(ed.), The enzymes, 3rd ed., vol. 3. AcademicPress Inc., New York.

148a. Matchett, W. H. 1972. Inhibition of tryptophansynthetase by indoleacrylic acid. J. Bacteriol.110:146-154.

149. Matchett, W. H., and J. A. DeMoss. 1975. Thesubunit structure of tryptophan synthase fromNeurospora crassa. J. Biol. Chem. 250:2941-2946.

150. Matsushiro, A., S. Sato, K. Ito, S. Kida, and F.Imamoto. 1965. On the transcription of thetryptophan operon in Escherichia coli. I. Thetryptophan operator. J. Mol. Biol. 11:54-63.

151. Maurer, R., and I. P. Crawford. 1971. Propertiesand subunit structure of the B component ofPseudomonas putida tryptophan synthetaseArch. Biochem. Biophys. 144:193-203.

152. Maurer, R., and I. P. Crawford. 1971. Newregulatory mutation affecting some of thetryptophan genes in Pseudomonas putida. J.Bacteriol. 106:331-338.

153. Meduski, J. W., and S. Zamenhof. 1969. Studieswith tryptophan synthetases from variousstrains of Bacillus subtilis. Biochem. J.112:285-292.

154. Metzenberg, R. L. 1972. Genetic regulatory sys-tems in Neurospora. Annu. Rev. Genet.6:111-132.

155. Monod, J., and G. Cohen-Bazire. 1953. Theeffect of specific inhibitors in the biosynthesisof tryptophan desmolase in Aerobacteraerogenes. C. R. Acad. Sci. Paris 236:530-532.

156. Morishita, T., T. Fukuda, M. Shirota, and T.Yura. 1974. Genetic basis of nutritional re-quirements in Lactobacillus casei. J. Bacte-riol. 120:1078-1084.

157. Morse, D. E., and C. Yanofsky. 1968. Theinternal low-efficiency promoter of the trypto-phan operon of Eschericha coli. J. Mol. Biol.38:447-451.

158. Morse, D., and C. Yanofsky. 1969. A transcrip-tion-initiating mutation within a structuralgene of the tryptophan operon. J. Mol. Biol.41:317-328.

159. Morse, D. E., and C. Yanofsky. 1969. Ambermutants of the trpR regulatory gene. J. Mol.Biol. 44:185-193.

160. Mosteller, R. D., and C. Yanofsky. 1971. Evi-dence that tryptophanyl transfer ribonucleicacid is not the corepressor of the tryptophan

117VOL. 39, 1975


http://mm

br.asm.org/

Dow

nloaded from


BACTERIOL. REV.

operon of Escherichia coli. J. Bacteriol.105:268-275.

160a. Murgola, E. J., and C. Yanofsky. 1974. Selec-tion for new amino acids at position 211 of thetryptophan synthetase a chain of Escherichiacoli. J. Mol. Biol. 86:775-784.

161. Murphy, T. M., and S. E. Mills. 1969. Immuno-chemical and enzymatic comparison of thetryptophan synthetase a subunits from fivespecies of Enterobacteriaceae. J. Bacteriol.97:1310-1320.

162. Nagao, R. T., and T. C. Moore. 1972. Partialpurification and properties of tryptophan syn-thetase of pea plants. Arch. Biochem. Bio-phys. 149:402-413.

163. Nazario, M., J. A. Kinsey, and M. Ahmad. 1971.Neurospora mutant deficient in tryptophanyl-transfer ribonucleic acid synthetase activity.J. Bacteriol. 105:121-126.

164. Nester, E. W., and R. A. Jensen. 1966. Control ofaromatic amino acid biosynthesis in Bacillussubtilis: sequential feedback inhibition. J.Bacteriol. 91:1594-1598.

165. Nester, E. W., M. Schafer, and J. Lederberg.1963. Gene linkage in DNA transfer: a clus-ter of genes concerned with aromatic bio-synthesis in Bacillus subtilis. Genetics48:529-551.

165a. Novick, A., and L. Szilard. 1954. Experimentswith the chemostat on the rates of amino acidsynthesis in bacteria, p. 21-32. In Dynamics ofgrowth processes. Princeton University Press,Princeton, N.J.

166. Palleroni, N. J., R. Kunisawa, R. Contopoulou,and M. Doudoroff. 1973. Nucleic acid homolo-gies in the genus Pseudomonas. Int. J. Syst.Bacteriol. 23:333-339.

167. Prager, E. M., and A. C. Wilson. 1971. Thedependence of immunological cross-reactivityupon sequence resemblence among lysozymes.I. Microcomplement fixation studies. J. Biol.Chem. 246:5978-5989.

167a. Proctor, A. R., and I. P. Crawford. 1975. Au-togenous regulation of the inducible trypto-phan synthetase of Pseudomonas putida.Proc. Natl. Acad. Sci. U.S.A. 72:1249-1253.

168. Proctor, A. R., and W. E. Kloos. 1970. Thetryptophan gene cluster of Staphylococcusaureus. J. Gen. Microbiol. 64:319-327.

169. Proctor, A. R., and W. E. Kloos. 1973. Trypto-phan biosynthetic enzymes of Staphylococcusaureus. J. Bacteriol. 114:169-177.

170. Patel, N., W. M. Holmes, and J. F. Kane. 1973.Intergeneric complementation of anthranilatesynthase subunits. J. Bacteriol. 114:600-603.

171. Patel, N., W. M. Holmes, and J. F. Kane. 1974.Homologous and hybrid complexes of anthra-nilate synthase from Bacillus species. J. Bac-teriol. 119:220-227.

172. Queener, S. F., and I. C. Gunsalus. 1970. An-thranilate synthase enzyme system and com-plementation in Pseudomonas. Proc. Natl.Acad. Sci. U.S.A. 67:1225-1232.

173. Queener, S. W., S. F. Queener, J. R. Meeks,and I. C. Gunsalus. 1973. Anthranilate syn-

thase from Pseudomonas putida: purificationand properties of a two-component enzyme. J.Biol. Chem. 248:151-161.

174. Rebello, J. L., and R. A. Jensen. 1970. Metabolicinterlock: the multi-metabolite control of pre-phenate dehydratase activity in Bacillus sub-tilis. J. Biol. Chem. 245:3738-3744.

175. Rheinwald, J. G., A. M. Chakrabarty, and I.C. Gunsalus. 1973. A transmissible plasmidcontrolling camphor oxidation in Pseudomo-nas putida. Proc. Natl. Acad. Sci. U.S.A.70:885-889.

176. Robb, F., and W. L. Belser. 1972. Anthranilatesynthetase. In vitro complementation betweenSerratia marcescens and Escnerichia coli sub-units. Biochem. Biophys. Acta 285:243-252.

177. Roberts, C. F. 1967. Complementation analysisof the tryptophan pathway in Aspergillusnidulans. Genetics 55:233-239.

178. Rocha, V., I. P. Crawford, and S. E. Mills. 1972.Comparative immunological and enzymaticstudy of the tryptophan synthetase 2 subunitin the Enterobacteriaceae. J. Bacteriol.111:163-168.

179. Rose, J. K., C. L. Squires, C. Yanofsky, H. L.Yang, and G. Zubay. 1973. Regulation of invitro transcription of the tryptophan operonby purified RNA polymerase in the presence ofpartially purified repressor and tryptophan.Nature (London) New Biol. 245:133-137.

180. Rose, J. K., and C. Yanofsky. 1971. Transcrip-tion of the operator proximal and distal endsof the tryptophan operon: evidence that trpEand trpA are the delimiting structural genes.J. Bacteriol. 108:615-618.

181. Rosenberg, S. L., and G. D. Hegeman. 1969.Clustering of functionally related genes inPseudomonas aeruginosa. J. Bacteriol.99:353-355.

182. Roth, C. W., and E. W. Nester. 1971. Co-ordi-nate control of tryptophan, histidine and tyro-sine enzyme synthesis in Bacillus subtilis. J.Mol. Biol. 62:577-589.

183. Saheki, T., and H. Holzer. 1974. Comparison ofthe tryptophan synthetase inactivating en-zymes with proteinases from yeast. Eur. J.Biochem. 42:621-626.

184. Sakaguchi, K. 1970. The similarity of trypto-phan synthetases of Anabaena variablis andChlorella ellipsoidea with that of bacteria.Biochem. Biophys. Acta 220:580-593.

185. Sanderson, K. E., and C. A. Hall. 1968. F-primefactors of Salmonella typhimurium and aninversion between S. typhimurium and E.coli. Genetics 64:215-228.

186. Sarich, V. M., and A. C. Wilson. 1967. Immuno-logical time scale for hominid evolution.Science 158:1200-1203.

187. Sawula, R. V., and I. P. Crawford. 1972. Map-ping of the tryptophan genes of Acinetobactercalcoaceticus by transformation. J. Bacteriol.112:797-805.

188. Sawula, R. V., and I. P. Crawford. 1973. Anthra-nilate synthetase of Acinetobacter cal-coaceticus: separation and partial character-

118 CRAWFORD


http://mm

br.asm.org/

Dow

nloaded from



ization of subunits. J. Biol. Chem.248:3573-3581.

189. Scaife, J., and J. R. Beckwith. 1966. Mutationalalteration of the maximal level of lac operonexpression. Cold Spring Harbor Symp. Quant.Biol. 31:403-408.

190. Schultz, G. E., and T. E. Creighton. 1969.Preliminary X-ray diffraction study of thewild-type and a mutationally altered trypto-phan synthetase a subunit. Eur. J. Biochem.10:195-197.

191. SchUrch, A., J. Miozzari, and R. HUtter. 1974.Regulation of tryptophan biosynthesis in Sac-charomyces cerevisiae: mode of action of 5-methyl tryptophan and 5-methyl tryptophan-sensitive mutants. J. Bacteriol. 17:1131-1140.

192. Schwartz, A. K., and D. M. Bonner. 1964.Tryptophan synthetase in Bacillus subtilis:effects of high potassium ion concentration ona two component enzyme. Biochim. Biophys.Acta. 89:337-347.

193. Schweingruber, M. E., and R. Dietrich. 1973.Gene-enzyme relationships in the tryptophanpathway of Schizosaccharomyces pombe. Ex-perientia 29:1152-1153.

194. Shimizu, Y., N. Shimizu, and M. Hayashi. 1973.In vitro repression of the transcription of thetryptophan operon by trp repressor. Proc.Natl. Sci. U.S.A. 70:1990-1994.

195. Smith, 0. H. 1967. Structure of the trpC cistronspecifying indole-glycerol phosphate synthe-tase and its location in the tryptophan operonof Escherichia coli. Genetics 57:95-105.

195a. Smithers, C. M., and P. P. Engel. 1974. Gene-enzyme relations of tryptophan mutants inStreptomyces coelicolor A3(2). Genetics78:799-808.

196. Snell, E. E. 1943. Growth promotion on trypto-phan-deficient media by o-aminobenzoic acidand its attempted reversal with orthanil-amide. Arch. Biochem. 2:389-394.

196a. Somerville, R. L., and C. Yanofsky. 1964. Onthe translation of the A gene region of trypto-phan messenger RNA. J. Mol. Biol. 8:616-619.

196b. Somerville, R. L., and C. Yanofsky. 1965.Studies on the regulation of tryptophan bio-synthesis in Escherichia coli. J. Mol. Biol.11:747-759.

197. Squires, C. L., J. K. Rose, C. Yanofsky, H. L.Yang, and G. Zubay. 1973. Tryptophanyl-tRNA and tryptophanyl-tRNA synthetase arenot required for in vitro repression of thetryptophan operon. Nature (London) NewBiol. 245:131-133.

197a. Steinberg, W. 1974. Temperature-inducedderepression of tryptophan biosynthesis in atryptophanyl-transfer ribonucleic acid synthe-tase mutant of Bacillus subtilis. J. Bacteriol.117:1023-1034.

198. Stenmark, S. L., D. L. Pierson, and R. A.Jensen. 1974. Blue-green bacteria synthesizeL-tyrosine by the pretyrosine pathway. Na-ture (London) 247:290-292.

199. Tatum, E. L., and D. Bonner. 1944. Indole and

serine in the biosynthesis and breakdown oftryptophane. Proc. Natl. Acad. Sci. U.S.A.30:30-37.

200. Tresguerres, M. E. F., W. M. Ingledew, and J.C. Canovas. 1972. Potential competition for5-dehydroshikimate between the aromaticbiosynthetic route and the catabolic hy-droaromatic pathway. Archiv. Mikrobiol.82:111-119.

200a. Tsai, H., C.-Y. Yang, and J. H. J. Tsai. 1974.The subunit structure of Neurospora trypto-phan synthase: a reappraisal. Biochem. Bio-phys. Res. Commun. 61:1332-1339.

201. Turner, J. R., and W. H. Matchett. 1968. Altera-tion of tryptophan-mediated regulation inNeurospora crassa by indoleglycerol phos-phate. J. Bacteriol. 95:1608-1614.

202. Twarog, R., and G. L. Liggins. 1970. Enzymes ofthe tryptophan pathway in Acinetobacter cal-co-aceticus. J. Bacteriol. 104:254-263.

203. Vogel, H. J. 1965. Lysine biosynthesis andevolution, p. 25-40. In V. Bryson and H. J.Vogel (ed.), Evolving genes and proteins. Aca-demic Press Inc., New York.

204. Wegman, J., and I. P. Crawford, 1968. Trypto-phan synthetic pathway and its regulation inChromobacterium violaceum. J. Bacteriol.95:2325-2335.

205. Wesseling, A. C. 1974. Cross-pathway regulationof amino acid biosynthetic pathways in Neu-rospora crassa. Ph.D. thesis, University ofCincinnati, Cincinnati, Ohio.

206. Wesseling, A. C., and M. Carsiotis. 1974. Aminoacid cross-pathway regulation in Neurosporacrassa: involvement of nitrogen-rich aminoacids. Abstr. Annu. Meet. Amer. Soc. Micro-biol., 1974, P288, p. 192.

207. Wuesthoff, G. W., and R. H. Bauerle. 1970.Mutations creating internal promoter ele-ments in the tryptophan operon of Salmonellatyphimurium. J. Mol. Biol. 49:171-196.

208. Whitt, D. D., and B. C. Carlton, 1968. Non-coor-dinate regulation in 5-methyl-tryptophan re-sistant mutants of Bacillus subtilis. Biochem.Biophys. Res. Commun. 33:636-642.

209. Widholm, J. M. 1972. Tryptophan biosynthesisin Nictiana tabacum and Daucus carota cellcultures: site of action of inhibitory tryptophananalogs. Biochem. Biophys. Acta 261:44-51.

210. Widholm, J. M. 1972. Anthranilate synthetasefrom 5-methyltryptophan-susceptible and re-sistant cultured Daucus carota cells. Bio-chem. Biophys. Acta 279:48-57.

211. Widholm, J. M. 1973. Measurement of the fiveenzymes which convert chorismate td trypto-phan in cultured Daucus carota cell extracts.Biochem. Biophys. Acta 320:217-226.

212. Wilson, A. C., and N. 0. Kaplan. 1962. Enzymestructure and its relation to taxonomy, p.321-346. In C. A. Leone (ed.), Taxonomicbiochemistry and serology. The Ronald PressCo., New York.

213. Woolfolk, C. A., and E. R. Stadtman. 1964.Cumulative feedback inhibition in the multi-ple end product regulation of glutamine syn-

119VOL. 39, 1975


http://mm

br.asm.org/

Dow

nloaded from


BACTERIOL. REV.

thetase activity in Escherichia coli. Biochem.Biophys. Res. Commun. 17:313-319.

213a. Woese, C. R., M. L. Sogin, and L. A. Sutton.1974. Procaryote phylogeny. I. Concerning therelatedness of Aerobacter aerogenes to Esche-richia coli. J. Mol. Evol. 3:293-299.

214. Yanofsky, C. 1956. The enzymatic conversion ofanthranilic acid to indole. J. Biol. Chem.223:171-184.

215. Yanofsky, C., and I. P. Crawford. 1959. Theeffect of deletions, point mutations, rever-sions and suppressor mutations on the twocomponents of the tryptophan synthetase ofEscherichwa coli. Proc. Natl. Acad. Sci. U.S.A.45:1016-1026.

216. Yanofsky, C., and I. P. Crawford. 1972. Trypto-phan synthetase, p. 1-31. In P. D. Boyer (ed.),The Enzymes, 3rd ed., vol. 7. Academic PressInc., New York.

217. Yanofsky, C., and V. Horn. 1972. TryptophansYnthetase a chain positions affected by muta-tions near the ends of the genetic map of trpAof Escherichia coli. J. Biol. Chem. 247:4494-4498.

218. Yanofsky, C., V. Horn, M. Bonner, and S.Stasiowski. 1971. Polarity and enzyme func-tions in mutants of the first three genes of thetryptophan operon of Escherichia coli. Genet-ics 69:409-433.

219. Yanofsky, C., and J. Ito. 1966. Nonsense codons

and polarity in the tryptophan operon. J. Mol.Biol. 21:313-334.

220. Yanofsky, C., and J. Ito. 1967. Orientation ofantipolarity in the tryptophan operon. J. Mol.Biol. 21:313-334.

221. Yanofsky, C., and E. S. Lennox. 1959. Transduc-tion and recombination study of linkage rela-tionships among the genes controlling trypto-phan synthesis in Escherichia coli. Virology8:425-447.

222. Yasunobu, K. T., and M. Tanaka. 1973. Theevolution of iron-sulfur protein containing or-ganisms. Syst. Zool. 22:570-589.

223. Yourno, J., T. Kohno, and J. R. Roth. 1970.Enzyme evolution: generation of a bifunc-tional enzyme by fusion of adjacent genes.Nature (London) 228:820-824.

224. Yu, P. H., M. R. Kula, and H. Tsai. 1973.Studies on the apparent instability of Neuro-spora tryptophan synthetase: evidence for aprotease. Eur. J. Biochem. 32:129-135.

225. Zalkin, H. 1973. Anthranilate synthetase, p.1-39. In A. Meister (ed.), Advances in enzy-mology, vol. 38. J. Wiley and Sons, New York.

226. Zalkin, H., and L. H. Hwang. 1971. Anthranilatesynthetase from Serratia marcescens: on theproperties and relationship to the enzymefrom Salmonella typhimurium. J. Biol. Chem.246:6899-6907.

.120 CRAWFORD


http://mm

br.asm.org/

Dow

nloaded from


gene rearrangements in the evolution ofthe tryptophan ... · shared by all bacterial tryptophan...

Documents