JOURNAL OF BACTFRIALOGY, Apr., 1966 Vol. 91, No. 4Copyright @ 1966 American Society for Microbiology Printed in U.S.A.

Expression of the Leucine OperonR. 0. BURNS,' J. CALVO,2 P. MARGOLIN, AND H. E. UMBARGER3

Cold Spring Harbor Laboratories of Quantitative Biology, Cold Spring Harbor, New York

Received for publication 23 December 1965

ABSTRACT

BuRNs, R. 0. (Cold Spring Harbor Laboratories of Quantitative Biology, ColdSpring Harbor, N.Y.), J. CALVO, P. MARGOLIN, AND H. E. UMBARGER. Expressionof the leucine operon. J. Bacteriol. 91:1570-1576. 1966.-The four genes whichspecify the structure of the three enzymes specifically involved in the biosynthesisof leucine in Salmonella typhimurium constitute a single operon. Three types ofcontrol mutants have been delineated on the basis of their location on the Sal-monella chromosome and the manner in which they coordinately affect the rates ofsynthesis of the pertinent enzymes. The three types of mutants correspond to opera-tor-negative, operator-constitutive, and regulator-negative. The rate of synthesis ofthe enzymes can also be altered by varying the amount of leucine made available tothe cell. Leucine can be effectively limited by limiting the supply of a-ketoisovalerate,but in doing so two of the three enzymes, a-isopropylmalate synthetase and iso-propylmalate isomerase, are labilized. This observation was correlated with an invivo diminution of the levels of the substrates of these enzymes and the fact thata-ketoisovalerate and a-isoporpylmalate protect the respective enzymes againstthermal inactivation in vitro. The functional association of the structural genes isalso illustrated by the presence of polarity mutations; that is, certain structural genemutations lower the rates of synthesis of the enzymes specified by genes locateddistally to the mutated gene and the operator segment of the operon.

A genetic analysis of the leucine biosyntheticpathway by Margolin (18), together with theidentification of the enzymes and intermediatesinvolved in the pathway (10, 15), led to theconcept that four genes (Table 3) specify thestructures of the three enzymes catalyzing theconversion of a-ketoisovalerate and acetate to a-ketoisocaproic acid and carbon dioxide. Jung-wirth et al. (14) have demonstrated that leucineauxotrophs of Salmonella typhimurium withlesions in cistron I are unable to condense a-ketoisovalerate and the acetyl group of acetyl co-enzyme A (CoA) to form a-isopropylmalate.Gross, Bums, and Umbarger (9) have demon-strated that leucine-requiring strains of S.typhimurium with lesion in cistrons III or IV areunable to effect the isomeration of a- and ,3-iso-propylmalate. Extracts obtained from leucineauxotrophs with lesions in cistron II are unable to

1 Present address: Department of Microbiology andImmunology, Duke University School of Medicine,Durham, N.C.

2 Present address: Department of Biochemistry,Cornell University, Ithaca, N.Y.

3 Present address: Department of BiologicalSciences, Purdue University, Lafayette, Ind.

convert 3-isopropylmalate to a-ketoisocaproate(3, 10).

It was recognized quite early in the analysis ofthis pathway that the syntheses of the threeenzymes were under control by the end product.For example, leucine auxotrophs cultured ongrowth-limiting concentrations of leucine alwayspossessed higher levels of leucine-synthesizingenzymes than did cultures obtained from mediumcontaining excess of leucine. Such observations,together with the fact that the genetic determi-nants for the three leucine enzymes are clusteredon the chromosome and were all rendered non-functional by a single point mutation (18), led tothe suggestion that the genes might comprise aregion of unit control or an operon (13).The organization into functional units of the

genes specifying the structures of enzymescatalyzing a sequence of reactions has beendescribed for a variety of systems (2). Manifesta-tions of such organization have been recognizedwhich can be used to ascertain whether or not aseries of enzymes are the products of a singleoperon. It is the purpose of this paper to illus-strate how the rate of synthesis of the threeenzymes specific for the synthesis of a-keto-

1570

on August 26, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

EXPRESSION OF THE LEUCINE OPERON

isocaproate, the immediate precursor of leucine,are affected by mutation and by the end product,leucine.

MATERIALS AND METHODSOrganisms and cultivation. The isolation and the

manner of characterizing the leucine auxotrophs of S.typhimurium LT2 were previously described (18).Strains C4, C-19, C-20, and C-6 were isolated asmutants resistant to 5', 5', 5'-trifluoroleucine. Thedetails of the isolation and a genetic analysis of thesestrains will be the subject of a subsequent publication(Calvo, in preparation). Strain C-6 ilva (an isoleucine-valine auxotroph) was isolated after treatment ofstrain C-6 with N-methyl-N-nitroso-N'-nitroguanidine(7). Strain ilva A-8 was obtained from M. Demerec,and is an isoleucine-valine auxotroph.

Cells were grown on Davis and Mingioli (4)medium modified by omitting the citrate and increas-ing the glucose to 0.5%. This medium was supple-mented as indicated in the text. All cultures weregrown at 37 C with aeration. The chemostat experi-ments were performed, in principle, according to theusual methods (19). Leucine was at a concentration(10 mg per liter) which yielded about 109 cells permilliliter. The growth volume was 500 ml and was re-placed over a period of 6 hr.Enzyme methods. Cell extracts were prepared by

use of a Branson 20-kc, 125-w sonifier. Extractsprepared in 0.1 M potassium phosphate (pH 7.0) wereassayed for a-isopropylmalate isomerase (formerlytermed 0-carboxy-,8-hydroxyisocaproate isomerase)and j3-isopropylmalate dehydrogenase (formerlytermed ,3-hydroxy-:3-carboxyisocaproate dehydro-genase) activities as previously described (3, 9). Thea-isopropylmalate synthetase (formerly ,B-carboxy-fl-hydroxyisocaproate synthetase) activity was deter-mined from extracts prepared in 0.05 M tris(hydroxy-methyl)amino methane-HCl (Tris-HCl), pH 7.2. Theextracts were assayed immediately after preparationin a reaction mixture containing in 0.5 ml: a-ketoiso-valeric acid, 25 ,umoles; potassium chloride, 100,umoles; Tris-HCI (pH 7.5), 100 umoles, acetyl co-enzyme A, 1 umole; and up to 0.5 units of enzyme.The reaction was started by the addition of acetylCoA 1 min after placing the reaction mixture at 37 C.The reaction was stopped by the addition of 3.0 ml ofmetaphosphoric acid (0.2% in saturated sodiumchloride). Immediately before reading at 540 myu, 0.5ml of 0.067 M sodium nitroprusside and 0.5 ml of 1.5M sodium carbonate in 0.067 M sodium cyanide wereadded to the mixture. A reaction mixture lacking a-

ketoisovalerate was assayed to correct for a phos-phate-stimulated sulfhydryl-releasing activity presentin the crude extracts. The synthetase activity assay waslinear with time and protein concentration, providedthat no more than one-half of the acetyl CoA had beencleared. The linear kinetics could be prolonged byincreasing the concentration of acetyl CoA in theassay. The enzyme requires monovalent cations formaximal activity. No divalent cation requirement wasobserved with crude extracts, and ethylenediamine-tetraacetic acid up to a concentration of 0.01 M did notinhibit the reaction. A unit of synthetase activity is

defined as the amount of enzyme releasing 1 pmole of-SH from acetyl CoA in 1 hr in the assay. Specificactivity is the number of units per milligram ofprotein. The nitroprusside reaction was standardizedwith reduced glutathione. Protein was detcrminedby the method of Lowry et al. (17).

Chemicals. Acetyl CoA was prepared according tothe method of Simon and Shemin (20). CoA, gluta-thione, and a-ketoisovaleric acid were obtained fromSigma Chemical Co., St. Louis, Mo. N-methyl-N-nitroso-N'-nitroguanidine was obtained from K &K Laboratories, Jamaica, N.Y. Glycyl-L-valine andL-leucine were purchased from Calbiochem. All otherchemicals were reagent grade.

RESULTSTable 1 shows the levels of the leucine-syn-

thesizing enzymes in S. typhimurium strain LT-2and several derivatives of it in which the levels ofthe enzymes in the pathway to leucine differedfrom those in the parent strain. In order that thedifferences in enzyme level would indicate clearlythe inherently different capacities of the variousstrains for enzyme formation, all cells (exceptstrain leu-500) were grown in minimal mediumand harvested during the logarithmic-growthphase. Strain leu-500 is a leucine auxotroph andwas grown in minimal medium containing 25 mgper liter of L-leucine; the leucine requirement ofthis strain is the consequence of a point mutationlocated to the left of the left-most mutation in thefirst cistron which is known (18) to have affectedthe structure of the corresponding enzyme (a-isopropylmalate synthetase). The results of anextensive genetic analysis led Margolin (18) toconclude that the lesion in strain leu-500 hasresulted in the condition 0 originally postulatedfor certain lac- mutants of Escherichia coli (13).The results given in Table 1 support this con-clusion. Strains C-4, C-19, and C-20 are resistantto 5', 5', 5'-trifluoroleucine, an inhibitor ofgrowth in the parent strain. The genetic lesionsin these strains are not linked by transduction

TABLE 1. Level of leucine biosynthetic enzymes invarious strains of Salmonella typhimurium*

Specific activityStrain

Synthetase Dehydrogenase Isomerase

LT-2.. 1.6 4.8 0.9C4 ...... 12.0 39.6 6.0C-6 ...... 22.8 55.0 13.0C-19 ...... 25.8 74.4 16.2C-20 ...... 4.2 12.0 2.7leu-500.... 0.0 0.02 0.0

* All of the strains were grown on minimalmedium as described in the text.

VOL. 91, 1966 1571

on August 26, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

BURNS ET AL.

with phage PLT-22 to the leucine region (Calvo,in preparation). It is not known whether or not thederepressed levels of the three leucine enzymes inthese three strains have resulted from mutationsin the same genetic region. Nevertheless, in termsof the Jacob and Monod model (13), these strainshave been tentatively assumed to have alteredregulatory genes which affect the activity of theleucine operon. Strain C-6 is also resistant tofluoroleucine, but its resistant character isgenetically linked to the region of the chromo-some specifying the structure of the leucine-synthesizing enzymes. Therefore, this strain hasbeen assumed to have an alteration in theoperator region of the leucine operon of the typecalled operator constitutive (00) (13).As shown in Table 1, the mutations which had

occurred in the various strains had differenteffects on the capacity of those strains to formthe enzymes of the leucine pathway. However, theeffect in any one strain was essentially the samefor all three enzymes. In other words, like themutations affecting the regulatory genes in fi-galactosidase in E. coli, these mutations affectedthe leucine-forming enzymes coordinately.With the finding that the rate of synthesis of the

enzymes of the leucine biosynthetic sequence canbe altered coordinately as a consequence ofcertain mutations, it was of interest to determinethe response of the "normally" controlled leucine-forming system to varying concentrations ofleucine. This determination could be effected bycultivating leucine auxotrophs in a chemostatcontaining media with growth-limiting con-centrations of leucine. Under such conditions,coordinate derepression would be expected if theleucine-forming enzymes are encoded by genes ina single operon (2). However, there are tworeasons why such a result might be more difficultto demonstrate in the case of the leucine operon:(i) only two enzyme activities could be measured,and (ii) it is known that certain mutations instructural genes within an operon are able toaffect the level of enzymes specified by otherstructural genes of the operon (2). The existenceof the latter phenomenon could lead to erroneousconclusions with regard to the possible unitaryresponse of the structural genes of the leucineoperon to the presence of leucine. A possiblealternative mechanism of effecting leucine limita-tion becomes evident if it is recalled that, in anorganism unable to synthesize valine, exogenouslysupplied valine is itself incorporated into proteinand is the source of four of the six carbons ofleucine. It was therefore reasonable to assumethat in such a system the mechanical limitation ofvaline would in fact result in a physiologicallimitation of leucine. This assumption was

affirmed by showing that cultures of strain ilvaA-8 obtained from a medium containing growth-limiting amounts of valine possessed higher levelsof a-isopropylmalate dehydrogenase than didcultures obtained from the same medium contain-ing excess leucine (6). The method describedabove therefore provides a means of limiting theamount of leucine available to the cell in spite ofthe prototrophy of the leucine biosynthetic path-way.When the rates of synthesis of the three enzymes

were followed throughout the growth of an iso-leucine-valine auxotroph in a medium contaiglimiting amounts of valine, the results shown inFig. 1 were obtained. Although the increases inthe levels of the enzymes were approximatelycoordinate during the early part of the growthcycle, a drastic decrease in specific activities of thesynthetase and isomerase was evident in the latterpart of growth. The growth increment during theperiod of enzyme decline is too small to accountfor this decrease in the specific activities by adilution of preformed enzyme. In other words, thedecrease in activity is not simply a cessation of

12

. '8e.25 -6cn

.i 4i21

5

w, 40

30

.72I

0.1 02 03 04 0.1 0.2 0.3 0.4Absorbancy of Culture

04

0l IWA

0.1 0.2 0.3 0.4Absorbancy of Culture

12345678Time (hrs.)

60 *50 gt0oS0

20 c

10 U.

0.4 .2c)

03

0.2 =0

0.1 °.0

FIG. 1. Effect of valine-mediated leucine limitationon the levels of the leucine-synthesizing enzymes.Strain ilva A-8 was grown on minimal medium supple-mented with 30 mg per liter of glycyl-L-valine and 50mg per liter of L-isoleucine. The culture was startedwith a relatively large inoculum of repressed cells. Theplots represent the specific activities (S.A.) of theenzymes obtainedfrom portions of the culture at varyingpoints in growth. The lower right-hand graph shows theincrease in the density of the culture with time; thearrows represent the points in growth at which sampleswere removed for analysis.

0-

I I

1572 J. BACTERIOL.

I I I

on August 26, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

EXPRESSION OF THE LEUCINE OPERON

synthesis of these two enzymes but is the result ofinactivation of the enzymes. [The use of glycyl-L-valine as the source of valine in this experimentwas necessary to obviate the problem of anapparent exclusion of valine by the relativelylarge amount of isoleucine required to satisfy thisrequirement of the double auxotroph. It is knownthat unfavorable ratios of isoleucine and valinewill not permit growth of a strain of S. thyphi-murium unable to synthesize these amino acids(Freundlich and Umbarger, unpublished data).]The correlation between the time of onset of the

instability observed above and the time at whichthe culture reached its terminal phase of growthprovided a clue to the reason for this instability.Owing to the fact that the termination of growthis caused by depletion of the leucine precursorsderived from valine, it seems likely that the inter-mediates formed in the sequence between valineand leucine protect the a-isopropylmalatesynthetase and a-isopropylmalate isomeraseagainst the observed in vivo inactivation. Morespecifically, a-ketoisovalerate and a-isopropylma-late might be required to stabilize the synthetaseand isomerase, respectively. As Fig. 2 shows, thesubstrates of these two enzymes do, in fact,stabilize the activities of these enzymes in vitro.According to these observations, it would be

expected that any interference with the synthesisof a-ketoisovalerate or a-isopropylmalate wouldlead to a decrease in the levels of synthetase orisomerase activity. It was of interest, therefore,to compare the effect of limiting valine in a strainof S. typhimurium in which the leucine operon wasnormally derepressed. Accordingly, the effect oflimiting valine was examined in strain C-6 ilva,in which there were lesions in the presumedoperator region of the leucine operon and in anonspecified ilva gene. As shown in Table 2,when this strain was grown with excess valine, allthree enzymes in the pathway to leucine wereconsiderably derepressed. When valine waslimiting, the activities of the synthetase andisomerase were again higher than those in thewild-type strain, LT-2, but much lower than thatfound in the same strain when grown withexcess valine. In contrast, the activity of thedehydrogenase, which in vitro has been observedto be quite stable, was even higher in cells grownin excess valine. [Although strain C-6 appears toresemble the operator-constitutive mutants de-scribed by Jacob and Monod (13) for the /3-galactosidase system, the levels of the enzymesare not completely refractory to regulation byexogenous leucine. This small degree of repres-sibility is shown for the prototrophic, derepressedstrain, C-6, in bottom of Table 2.]An alternative method of regulating the endog-

0-6

to._

2 3 4 5Time (hrs)

FIG. 2. Stabilization ofa-isopropylmalate synthetaseand isopropylmalate isomerase activities by substrates.A crude extract of strain C-6 was diluted to containabout 2 mg ofprotein per ml. One portion of the dilutedextract was made 10-3 M with respect to a-ketoiso-valerate and was placed at 25 C along with an unsupple-mented portion of extract. The synthetase activity wasdetermined at the times shown. The same extract andprocedure was used to test the stability of the isomerase,the addition being a-isopropylmalate and the tempera-ture 4 C.

enous supply of a-isopropylmalate is by takingadvantage of another type of control mechanismoperating in the leucine pathway. It is known thatthe activity of a-isopropylmalate synthetase issubject to inhibition by leucine (end-productinhibition). One might therefore expect that theaddition of excess quantities of leucine to aculture of S. typhimurium would result in adecrease in the endogenous level of a-isopropyl-malate accompanied by a diminution of thestabilizing effect on the isomerase. As shown inTable 2, this expectation was realized with strainLT-2, in which the isomerase is barely detectablein cells obtained from a medium containingleucine. However, in the derepressed strain C-6,the level of the isomerase is coordinated with theother enzymes, a condition that would obtain ifcells of this strain contained a significant level of

1573VOL. 91, 1966

on August 26, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

BURNS ET AL.

TABLE 2. Effect of limiting and excess amounits ofleucine and valine on the levels of the

leucine-synthesizing enzymes*

Specific activity

Strain Growth conditionSyn- dro-Isom-thetase genase erase

C-6 ilva Excess valine 26.0 76.2 9.0Limiting valine 8.2 114.0 1.0

LT-2 Minimal 1.4 6.0 0.72Minimal + leucine 0.3 1.2 0.01

C-6 Minimal 22.8 60.2 13.0Minimal + leucine 16.2 47.0 7.0

* Strain C-6 ilva was grown on minimal mediumsupplemented with (i) 100 mg per liter of L-valineand 50 mg per liter of L-isoleucine, and (ii) with30 mg per liter of glycyl-L-valine and 50 mg perliter of L-isoleucine. The cultures were harvested ata point in growth representing about three-fourthsof the maximal growth for the limited case. StrainsLT-2 and C-6 were grown on minimal mediumwith and without 100 mg per liter of L-leucine;the cultures were harvested midway through theexponential phase of growth.

a-isopropylmalate. Furthermore, strain C-6 (aswell as other derepressed strains) excretes leucineinto the culture medium (Calvo, in preparation).These observations indicate that, although thesynthetase of these strains is subject to end-product inhibition, that inhibition is not sufficientto inhibit completely the relatively high level ofthis enzyme found in strain C-6, and, therefore,enough a-isopropylmalate is formed to protectthe activity of the isomerase. Additional supportfor this view is found in the observation thatcultures of strain C-6 growing in the presence ofuniformly labeled C'4-glucose as sole carbonsource excrete C'4-leucine into the mediumwhether cultured in the presence or absence of 100mg per liter of L-leucine (Bums, unpublisheddata).The experience of Ames and Hartman (1) with

the histidine operon has led to the expectationthat, among a series of mutations affecting asingle gene in an operon, there exist some whicheffect the expression of intact genes in the sameoperon lying on the side of the mutated genedistal to the operator. This effect, which isbelieved to be due to a polarity in transcription ortranslation of the genetic message, has beendemonstrated in several other systems (11, 12,13). Since this may be another consequencespecifically due to a gene cluster functioning as anoperon, approximately 100 leucine auxotrophswere examined for the presence of such effects.A representative sampling of the results obtained

is presented in Table 3. For the purpose of thisanalysis, the various auxotrophs were grown in achemostat with leucine as the growth-limitingfactor; this method of cultivation obviates theproblem of partial derepression of the system andfacilitates comparison of the strains tested. Whengenetically unstable (i.e., those which reverted toleucine prototrophy) strains were analyzed, thecultures were initiated from a single colony, andthe culture present in the growth vessel after fourgenerations (the harvesting time) was assayed forthe total viable cells and also for the number ofleucine-independent organisms; this procedureinsured that the specific activities obtained werenot biased by the presence of prototrophicorganisms.

In Table 3, the relative levels of thethreeleucine-forming enzymes in a strain with no known intra-operonic mutation are indicated by the regulator-negative strain, C-19. Strain leu-124 possessed anormal level of dehydrogenase but a decreasedlevel of isomerase as compared with the control.This apparently discoordinate synthesis wouldappear to be due to the inability of strains lackingsynthetase activity to form a-isopropylmalate,the presumed stabilizer of isomerase activity.Strain leu-493 showed a drastic reduction of bothactivities. Approximately 30% of a group of 74group I auxotrophs analyzed showed a similar

TABLE 3. Effect of intraoperonic structuralmutations on the levels of the leucine-

synthesizing enzymes*IPM IPM IPM

SYNTHETASE DEHYDROGENASE --ISOMERASE-.0 I E mlm

500 124 499 126493 529 130

Specific activityStan Cistron-Stan affected Dehydro-

Synthetase genae- isomerase

leu-124 I 132 5leu-493 I 3 1leu499 II 21 30leu-529 II 19 4leu-126 III 18 110leu-130 III 20 119C-19 26 74 16

* The manner in which the cultures were grownis presented in the Materials and Methods. Thefigure above the table represents the relative posi-tions of the structural genes as well as the operatorregion and the locations of the mutations in thestrain analyzed. The mutant strains are placedaccording to the cistron affected, and the figureshould not be taken as representing the absolutelocation of the mutations within the cistron.

1574 J. BACTERIOL.

on August 26, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

EXPRESSION OF THE LEUCINE OPERON

effect (downward polarity). Some strains withlesions in the second cistron (affecting the de-hydrogenase) also showed downward polarity,but in these cases only the activity of the isom-erase was affected; a decrease in synthetaseactivity as a consequence of mutation in thesecond cistron was never observed. Strain leu-499had normal activities of the synthetase and isom-erase, whereas strain leu-529 had normalactivity of the synthetase but a decreased isom-erase activity. A total of eighteen group III andIV auxotrophs were examined, and all had normalsynthetase and dehydrogenase activities.When these results were analyzed with reference

to the orientation of the cistrons and the operatorsegment of the operon, it was evident that theability of an auxotrophic mutation to act pleio-tropically within the operon depends upon itsposition relative to the nonmutated cistrons.Furthermore, the lowering effects seem to be inone direction, i.e., only the activities of theenzymes specified by genes distal to both themutated gene and the operator segment of theoperon are affected, whereas enzymes specifiedby genes located between the operator segmentand the mutated genes are in normal levels.Although fewer genes could be examined, theresults obtained with the enzymes of the leucineoperon are precisely those to be expected if thepolarity effects observed by Ames and Hartman(1) constitute a general property of an operon.There remains one apparent inconsistency in

the data of Tables 1 and 3. The specific activitiesfor synthetase of leucine III auxotrophs are aboutone-sixth of those for dehydrogenase, whereasfor the constitutive control mutants and strainLT-2 the ratio is about one-third. In other words,the chemostat-grown cells have relatively lowerlevels of synthetase than do the cells obtainedfrom typical shake cultures. Although the precisereason for this observation is not clear, it mightreflect a difference in the conditions of growth inthat continuous cultivation might provide lesssuitable conditions for synthetase stability.

DISCUSSIONThe results reported are in accord with an

earlier conclusion that the leucine-synthesizingenzymes in S. typhimurium are the products of asingle operon (18). The existence of single-stepmutations which give rise to strains with coor-dinately altered levels of the three leucine-specificenzymes is strong evidence that the geneticorientation of the structural genes involved in thesynthesis of these enzymes corresponds to an areaof unified control on the S. typhimurium chromo-some. The significance of the fluoroleucine-resistant strains as related to the control of the

synthesis of the enzymes in point will be describedin a later publication. For the purpose of thisdiscussion, it should be sufficient to mention thatthese fluoroleucine-resistant strains appear to bearalterations in either the operator or the regulatorgenes, although a definitive conclusion in thisregard is dependent upon the results of thenecessary complementation tests. The primaryconcern here is that at least two types of muta-tions exist which are able to alter the rate of syn-thesis of the leucine-synthesizing enzymes. Thus,strain C-6, in which the fluoroleucine-resistantmarker is linked to the structural genes, possiblycontains an operator region which is desensitizedwith respect to the product of the regulator gene.Strains C-4, C-20, and C-19, in which the fluoro-leucine-resistant character is not linked to theleucine region, presumably bear mutations in theleucine regulator gene such that they produce arepressor which is altered with respect to itsability either to recognize the corepressor,leucine, or its presumably normal site of action,the operator. It is interesting to note that theleucine enzymes in strains C-4, C-19, and C-20encompass a rather broad range of constitutive-ness, suggesting that, if the aberrations in thesestrains are allelic, they must be due to alterationsof, rather than the absence of, the repressorsubstance, although a decreased rate of repressorsynthesis cannot be eliminated.The response of the wild-type leucine bio-

synthetic system to controlled amounts of leucinemight appear, at first, paradoxical. This observa-tion, however, is to be expected, owing to thefact that leucine, in quenching metabolic flow overthis pathway, prevents the formation of thecompound, a-isopropylmalate, which stabilizesthe isomerase. The results showing that cells withimpaired synthesis of a-isopropylmalate did yieldextracts with discoordinately lowered levels ofsynthetase activity, if viewed alone, could beinterpreted as an indication of induction of thisenzyme by its substrate. In such a case, organiza-tion of the genes might therefore involve twoclosely associated operons, one of which wouldspecify the synthetase and dehydrogenase and theother specifying the isomerase. The existence ofso-called regulator mutants for the leucine systemin which all three enzymes are synthesized atincreased rates cannot rule out this possibility,since these mutants presumably also would pro-duce increased levels of the postulated inducer, a-isopropylmalate; the same argument applies tothe operator-constitutive strain. It should bementioned here that, in the case of Neurosporacrassa in which the biosynthetic steps for thesynthesis of leucine appear to be identical withthose found in S. typhimurium and in which the

1575VOL. 91, 1966

on August 26, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

BURNS ET AL.

genetic determinants for the three enzymes arenot linked, a-isopropylmalate does appear toinduce the formation of the isomerase as well asthe dehydrogenase, and only the formation of thefirst enzyme in the pathway, the synthetase, iscontrolled by end-product repression (8).On the other hand, the finding that polarity is

exhibited by the cluster of genes concerned withleucine biosynthesis, like other clusters consideredto be operons (1, 5, 11, 13, 16), constitutes furtherevidence that this cluster is also a single operon.Whether the polarity effects of mutations involveperturbations of the translation of the messenger-RNA into specific polypeptide chains or someother mechanism is irrelevant in this respect. Theessential point is that there is no mechanismknown which would explain a polarity effectextending over two operons.

ACKNOWLEDGMENTSThis investigation was supported by Public Health

Service research grant GM 07615 from the Divisionof General Medical Sciences. A portion of thisinvestigation was conducted at Duke University andsupported by Public Health Service grant GM 12551.

LITERATURE CITED1. AMES, B. N., AND P. E. HARTMAN. 1963. The

histidine operon. Cold Spring Harbor Symp.Quant. Biol. 28:349-356.

2. AMES, B. N., AND R. G. MARTIN. 1964. Bio-chemical aspects of genetics. In Ann. Rev.Biochem. 33:235-258.

3. BURNS, R. O., H. E. UMBARGER, AND S. R. GROSS.1963. The biosynthesis of leucine. III. Theconversion of a-hydroxy-#3-carboxyisocaproateto a-ketoisocaproate. Biochemistry 2:1053-1058.

4. DAVIs, B. D., AND E. S. MINGIOLI. 1950. Mutantsof Escherichia coli requiring methionine orvitamin B12. J. Bacteriol. 60:17-28.

5. FRANKLIN, N. C., AND S. E. LURIA. 1961. Trans-duction by bacteriophage P1 and the propertiesof the loc genetic region in E. coli and S.desenteriae. Virology 15:299-311.

6. FREUNDLICH, M., R. 0. BURNS, AND H. E.UMBARGER. 1962. Control of isoleucine valineand leucine biosynthesis. I. Multivalent repres-sion. Proc. Natl. Acad. Sci. U.S. 48:1804-1808.

7. GAREN, A., as cited by L. Gorini, and E. Kataja.

1964. Phenotypic repair by streptomycin ofdefective genotypes in E. coli. Proc. Natl.Acad. Sci. U.S. 51:487-493.

8. GROSS, S. R. 1965. The regulation of synthesis ofleucine biosynthetic enzymes in Neurospora.Proc. Natl. Acad. Sci. U.S. 54:1538-1546.

9. GROSS, S. R., R. 0. BURNS, AND H. E. UMBARGER.1963. The biosynthesis of leucine. II. Theenzymatic isomerization of fl-carboxy-fl-hy-droxyisocaproate and a-hydroxy-i8-carboxyiso-caproate. Biochemistry 2:1046-1052.

10. GROSS, S. R., C. JUNGWIRTH, AND H. E.UMBARGER. 1962. Another intermediate inleucine biosynthesis. Biochem. Biophys. Res.Commun. 7:5-9.

11. HENNING, U., C. HERZ, AND K. SzoLGvoy. 1964.Polarization and disproportionate synthesis ofthe enzyme components of the pyruvate-dehydrogenase complex in mutants of Esch-erichia coli K-12. Z. Vererbungslehre 95:236-259.

12. ITO, J., AND I. P. CRAWFORD. 1965. Regulationof the enzymes of the tryptophan pathway inEscherichia coli. Genetics 52:1303-1316.

13. JACOB, F., AND J. MONOD. 1961. On the regulationof gene activity. Cold Spring Harbor Symp.Quant. Biol. 26:193-209.

14. JUNGWIRTH, C., S. R. GROSS, P. MARGOLIN, ANDH. E. UMBARGER. 1963. The biosynthesis ofleucine. I. The accumulation of j3-carboxy-3-hydroxyisocaproate by leucine auxotrophs ofSalmonella typhimurium and Neurospora crassa.Biochemistry 2:1-6.

15. JUNGWIRTH, C., P. MARGOLIN, AND H. E.UMBARGER. 1961. The initial step in leucine bio-synthesis. Biochem. Biophys. Res. Commun.5:435-437.

16. LEE, N., AND E. ENGLESBERG. 1962. Dual effectsof structural genes in Escherichia coli. Proc.Natl. Acad. Sci. U.S. 48:335-348.

17. LOWRY, 0. H., N. J. ROSEBROUGH, A. L. FARR,AND R. J. RANDALL. 1951. Protein measure-ment with the Folin phenol reagent. J. Biol.Chem. 193:265-275.

18. MARGOLIN, P. 1962. Genetic fine structure of theleucine operon in Salmonella. Genetics 48:441-457.

19. NovICK, A., AND L. SzILARD. 1950. Descriptionof the chemostat. Science 112:715-716.

20. SIMON, E. J., AND D. SHEMIN. 1953. The prepara-tion of S-succinyl coenzyme A. J. Am. Chem.Soc. 75:2520.

1576 J. BACTERIOL.

on August 26, 2020 by guest

http://jb.asm.org/

Dow

nloaded from