. 13: 1337–1346 (1997)

Lys80p of Saccharomyces cerevisiae, PreviouslyProposed as a Specific Repressor of LYS Genes, is aPleiotropic Regulatory Factor Identical to Mks1p

ANDREu FELLER, FERNANDO RAMOS, ANDREu PIEuRARD AND EVELYNE DUBOIS*

Laboratoire de Microbiologie, Faculte des Sciences, Universite Libre de Bruxelles and Institut de Recherches duCERIA, avenue E. Gryson 1, B-1070 Bruxelles, Belgium

Received 4 February 1997; accepted 5 May 1997

In Saccharomyces cerevisiae, an intermediate of the lysine pathway, á-aminoadipate semialdehyde (áAASA), acts asa coinducer for the transcriptional activation of LYS genes by Lys14p. The limitation of the production of thisintermediate through feedback inhibition of the first step of the pathway results in apparent repression by lysine.Previously, the lys80 mutations, reducing the lysine repression and increasing the production of lysine, wereinterpreted as impairing a repressor of LYS genes expression. In order to understand the role of Lys80p in thecontrol of the lysine pathway, we have analysed the effects of mutations epistatic to lys80 mutations. The effects oflys80 mutations on LYS genes expression were dependent on the integrity of the activation system (Lys14p andáAASA). The increased production of lysine in lys80 mutants appeared to result from an improvement of themetabolic flux through the pathway and was correlated to an increase of the á-ketoglutarate pool and of the levelof several enzymes of the tricarboxylic acid cycle. The LYS80 genes has been cloned and sequenced; it turned out tobe identical to geneMKS1 cloned as a gene encoding a negative regulator of the RAS-cAMP pathway. We concludethat Lys80p is a pleiotropic regulatory factor rather than a specific repressor of LYS genes. ? 1997 John Wiley &Sons, Ltd.

Yeast 13: 1337–1346, 1997.No. of Figures: 1. No. of Tables: 2. No. of References: 35.

— Saccharomyces cerevisiae; LYS80 gene; á-ketoglutarate; apparent repression; pleiotropic factor

INTRODUCTION

Among the various genes involved in thehomocitrate/á-aminoadipate pathway of lysinebiosynthesis in Saccharomyces cerevisiae (for areview see Bhattacharjee, 1992), at least six(LYS20, LYS21, LYS4, LYS2, LYS9 and LYS1)are regulated by lysine (Ramos and Wiame, 1985;Urrestarazu et al., 1985; Ramos et al., 1996). Thiscontrol is mediated by the Lys14p transcriptionalactivator according to the levels of the meta-bolic intermediate á-aminoadipate semialdehyde(áAASA) serving as a sensor of lysine availability(Ramos et al., 1988; Feller et al., 1994). We have

presented evidence that the activation function ofLys14p requires the presence of áAASA and isstrongly reduced when lysine is added to thegrowth medium, thus resulting in apparent repres-sion by lysine (Feller et al., 1994). A LexA-Lys14fusion protein is indeed able to activate the expres-sion of lacZ under the control of the lexA operatorwhereas high concentrations of lysine antagonizethe activation function.Since homocitrate synthase catalysing the first

step of lysine biosynthesis is feedback inhibited bylysine (Tucci and Ceci, 1972), high levels of lysinelead to low levels of intracellular áAASA whichreduce the transcriptional activation of LYS genesby Lys14p. Homocitrate synthase is thus a keyelement in the control of áAASA supply and ofLys14p activity. Recently, the genes LYS20 andLYS21 encoding the two isoforms of homocitrate

*Correspondence to: E. Dubois, Institut de Recherches duCERIA, 1 Av. E. Gryson, B-1070 Bruxelles, Belgium. Tel:32-2-5267277; fax: 32-2-5267273; email: fanarg@resulb.ulb.ac.be.Contract grant sponsor: Research Council of the UniversiteLibre de Bruxelles.

CCC 0749–503X/97/141337–10 $17.50? 1997 John Wiley & Sons, Ltd.

synthase have been characterized (Ramos et al.,1996; G. Volckaert, pers. comm.; F. Ramos, un-published results). A double lys20, lys21 mutantstrain is unable to grow in the absence of lysine.Lys14p is a 89 kDa DNA binding protein that

belongs to the Zn2Cys6 binuclear cluster family offungal transcriptional activators. The promoters ofthe co-regulated LYS genes contain a sequencewith CC and GG doublets separated by 3 bpwhich is required for Lys14p activation and lysineapparent repression (Becker, in preparation).It was reported previously that the repression

by lysine was dependent on a negative regulatorencoded by LYS80 (Ramos and Wiame, 1985).The lys80 mutations were obtained during thesearch for mutants defective in threonine catab-olism. A first lys80-1 mutation was present inaddition to the cha1 mutation (catabolism ofhydroxy amino acids) in a mutant selected for theabsence of growth on threonine as the nitrogensource. The cha1 gene encodes the catabolic-serine (-threonine) deaminase and is regulatedby transcriptional induction by serine or threo-nine (Ramos and Wiame, 1985; Petersen et al.,1988) mediated by the Cha4p, a transcriptionalactivator belonging to the Cys6 zinc cluster class(Holmberg and Schierling, 1996). Another muta-tion, lys80-2, was isolated independently of acha1 mutation and caused a reduction of thegrowth rate on threonine. The generation timewas 6 h instead of 4 h for a wild-type strain.These lys80 mutations led to a considerable in-crease in cellular lysine pools, to an increase ofLYS gene expression on minimal medium and toan impairment of LYS gene repression by lysine.These mutations were interpreted as affecting aspecific repressor of LYS gene expression andwere therefore called lys80. Recently, we haveshown that the antagonistic effect of lysine on theactivation ability of the LexA-Lys14 fusion pro-tein is strongly reduced in a lys80 mutant strain(Feller et al., 1994), a behaviour which seemsincompatible with that of a strain mutated in arepressor gene.The aim of the present work was to reinvestigate

the nature of the LYS80 gene product. Evidencewas obtained that the ability of lys80 mutations toincrease the lysine pool is correlated to an increaseof the cellular concentration of á-ketoglutarate, aprecursor of the lysine pathway. Moreover cloningof the LYS80 gene revealed that LYS80 is identicalto theMKS1 gene encoding a negative regulator ofthe RAS-cAMP pathway.

MATERIALS AND METHODS

Strains and media

The Escherichia coli strain used for plasmidmaintenance and for propagation was XL1-Bluefrom Stratagene (recA1 endA1 gyrA96 thi-1hsdR17 supE44 relA1 lac [F* proAB lacIqZÄM15Tn10 (Tetr)]).All strains of S. cerevisiae were derivatives of the

MATá wild-type strain Ó1278b (Bechet et al.,1970): 8761c (lys80-1), 8953b (lys80-2), 12T7c(ura3), 13T6d (ura3,lys80-1), 17T9c (ura3,lys80-1,lys2), 13T7d (ura3,ilv1), 13T9a (ura3,lms,ilv1),14TOd (ura3,lms,ilv1,lys80-1), 17T7a (ura3,lys2,lys9,lys80-1) and 14T3c (lys4 issued fromMG360, lys1 issued from MG762).All yeast strains were grown on minimal me-

dium containing 10 m-(NH4)2SO4, 3% glucose,vitamins and mineral traces as described pre-viously (Messenguy, 1976). Where specified, 66 ìgof -lysine per ml, 25 ìg of uracil per ml and 25 ìgof -isoleucine per ml were added to the minimalmedium.

Enzyme assays and measurement of metabolitepoolsAll the protein extracts were dialysed before

determination of the different enzymatic activities.Each value of enzyme specific activity and metabo-lite concentration is the mean of two of threeindependent measurements. The level of variabilityis about 10% for enzyme specific activity and 15%for metabolite concentration.Saccharopine dehydrogenase (NADP+, gluta-

mate forming; EC 1.5.1.10) was assayed asdescribed by Jones and Broquist (1965). Saccha-ropine dehydrogenase (NAD+ lysine forming; EC1.5.1.7) was assayed as described by Fujioka andNakatami (1970). Argininosuccinate lyase (EC4.3.2.1) was assayed as described by Delbecq et al.(1994) but the argininosuccinate concentration was2 m instead of 10 m. Citrate synthase (EC4.1.3.7), isocitrate dehydrogenase (EC 1.1.1.41, EC1.1.1.42), malate dehydrogenase (EC 1.1.1.37) andaconitase (aconitate hydratase; EC 4.2.1.3) wereassayed as described in Parvin (1969), Kornberg(1955) and Fansler and Lowenstein (1969)respectively.For lysine pool measurements, exponentially

growing cells were extracted with ice-cold 0·3 -HClO4. The resulting extract was neutralized withK3PO4, centrifuged to eliminate the cells and

1338 . .

? 1997 John Wiley & Sons, Ltd. . 13: 1337–1346 (1997)

acidified to pH 6·2 with HCl. The solution ob-tained was used in a bioassay for growing thelysine auxotroph 14T3c containing two mutations(lys4 and lys1) in order to avoid the selection ofprototrophic revertants.Glutamate and á-ketoglutarate pools were de-

termined as described in Delforge et al. (1975) andDubois et al. (1974) respectively.

DNA techniquesAll procedures for DNA manipulations were

carried out by standard procedures (Sambrooket al., 1989).All DNA sequences of double-stranded tem-

plates were determined by the method of dideoxychain termination (Sanger et al., 1977) usingSequenase DNA polymerase (USB), [35S]dATPand synthetic oligonucleotides as walking primers.

Cloning of the LYS80 geneA 4·9 kb SalI-SalI fragment bearing the LYS2

gene isolated from plasmid pDP6 (Fleig et al.,1986; a gift from P. Phillipsen) was inserted in theunique SalI restriction site of pFL38 plasmid(pUC19-ARS CEN-URA3; Bonneaud et al., 1991)yielding plasmid pBB17. To destroy the two ClaIrestriction sites flanking the ARS CEN region, thepBB17 was digested by ClaI and after ligation andE. coli transformation, a plasmid named pLAF58containing the ARS CEN region and having lostthe two ClaI restriction sites was selected. Theplasmid pLAF59 (pUC19-ARS CEN-URA3-LYS2) was obtained by inserting, in the uniquePstI site of pLAF58, an oligonucleotide bearingthe following restriction sites: PstI-MluI-ClaI-MluI-PstI.The genomic DNA of the wild-type Ó1278b was

isolated (Cryer et al., 1975) and partially digestedby TaqI. The genomic TaqI fragments were in-serted in the unique ClaI site of pLAF59 to give alow copy number library. Yeast strain 17T7a(ura3,lys2,lys9,lys80-1) was transformed by thisDNA library on minimal medium plus lysine bythe method of Ito et al. (1983). The pLAF67plasmid containing a 8 kb PstI-PstI fragment wasisolated from this library. The subcloning of thisinsert was performed by insertion of the followingfragments in pFL38 vector: the 4·7 kb BamHI-BamHI in the BamHI restriction site, the 2·6 kbEcoRI-EcoRI in the EcoRI restriction site, the3·1 kb NcoI-BamHI blunt-ended by the Klenow

enzyme in the SmaI restriction site, the 2·3 kbEcoRV-XbaI in the SmaI-XbaI restriction sites andthe 2·9 kb BamHI-HindIII in the BamHI-HindIIIrestriction sites.

Cloning and sequencing of lys80 mutantsTo determine the sequence of the LYS80 gene

and of the mutant loci lys80-1 and lys80-2, theLYS80 region was amplified by PCR using astemplate the chromosomal DNA of strains Ó1278b(wild-type), 13T2d (ura3,lys80-1) and 8953b(lys80-2) isolated by the method of Karier andAuer (1993). We used the oligonucleotides OA52(gcggatccCTCAAACTTGTGCAGATTGC) andOA53 (gcggatccGTGTGTCATTAGAAGGAACT) derived from the published sequence of theyeast genome and extended with a BamHI restric-tion site. These oligonucleotides are located re-spectively at 533 bp upstream from the ATG and113 bp downstream from the stop codon. The3·3 kb BamHI-BamHI fragments were insertedin the BamHI site of pFL44L vector (pUC19-2ì-URA3; Bonneaud et al., 1991) and entirelysequenced.

Deletion of the LYS80 geneThe PCR-strategy of Wach (1996) was followed

to disrupt the LYS80 gene. A disruption cassettecontaining the KanMX4 cassette conferring toyeast a resistance to geneticin (G418) with longflanking regions homologous to LYS80 was syn-thesized by PCR. The yeast strains, PCR protocolsand plasmids are identical to those described in theoriginal articles (Wach et al., 1994; Wach, 1996).The first two PCRs were made with oligonucle-otides P5*/P5*L and P3*/P3*L using genomic DNAas template. The P5* oligonucleotide (TCAACCGGGATCGTGTCATA) is located 450 bp upstreamfrom the start codon. The P5*L oligonucleotide(ggggatccgtcgacctgcagcCATTTTCAGGTTCCAGTCTCTC) is located at the start codon of LYS80.The P3* oligonucleotide (GTTGCAGTTGAGAATCTTGT) is located at 374 bp downstream fromthe stop codon. The P3*L oligonucleotide (cgagctcgaattcatcgatgaTGATTACTAAGTTAAATAAATCAG) is located at the stop codon of LYS80. Thelower-case letters are identical to the borders of thecassette. The second PCR was performed using astemplate NotI-digested pFA4-KanMX4 and eachproduct from the first PCR together with theoligonucleotide P3* and P5*.

1339 Lys80p

? 1997 John Wiley & Sons, Ltd. . 13: 1337–1346 (1997)

Haploid strain 12T7c (ura3) was transformedwith 1 ìg of PCR product according to the methodof Gietz et al. (1992) adapted for geneticin selec-tion (Wach, 1996). Integration of the KanMX4cassette at the target locus was confirmed by PCRusing three oligonucleotides located 304 bp up-stream from the LYS80 stop codon, 428 bp down-stream from the LYS80 stop codon and 125 bpupstream from the 3* end of the KanMX cassette.The 5* region was checked by PCR using threeother oligonucleotides located 142 bp downstreamfrom the LYS80 start codon, 550 bp upstreamfrom the LYS80 start codon and 126 bp down-stream from the 5* end of the KanMX cassette.

RESULTS

Suppression of lys80 effects on LYS genesexpressionSince we have evidence that repression of LYS

genes by lysine is not independent of activation byLys14p and áAASA, we have analysed the expres-sion of LYS genes in the Lys80,lys2 double mutantstrain in which there is no production of áAASA.As shown in Table 1, the lys2 mutation is epistaticto lys80 mutations, the impairment of LYS generepression by lysine in a lys80 strain being abol-ished in the lys80,lys2 strain. The effects of lys80mutation on LYS gene expression are conse-quently dependent on a functional activationsystem (Lys14p plus á-AASA).We have tried to select other mutations leading

to the suppression of the lys80 phenotype. Theselection was based on the following observations.An ilv1 mutant strain, lacking the anabolic threo-nine deaminase is unable to grow on minimalmedium without isoleucine whereas it is able togrow on serine as sole nitrogen source. Underthese conditions, the catabolic threonine/serinedeaminase is induced and converts intracellularthreonine into á-ketobutyrate, the precursor ofisoleucine. We did however observe that the pres-ence of a lys80 mutation strongly reduces thegrowth rate of a ilv1 mutant strain on serine (7 hinstead of 3 h). The difference in growth rateallowed us to select a spontaneous mutation con-ferring the ability to grow on serine to the ilv,lys80mutant strain. This new mutation was called lmsfor lys80 mutation suppressor. As shown in Table1, the lms mutation prevents not only the highproduction of lysine due to the lys80 mutation butalso strongly reduces that production as compared

to a wild-type strain. The lmsmutation is recessive,monogenic and not linked to lys80. In this strain,an extremely low lysine pool and a reduced gluta-mate pool are consequences of a six-fold reductionof the á-ketoglutarate pool. In contrast, in a lys80mutant, the á-ketoglutarate intracellular concen-tration is increased three- to six-fold as comparedto a wild-type strain. A correlation can conse-quently be established between the pools ofá-ketoglutarate and of lysine. The increasedproduction of á-ketoglutarate augments theentry of metabolites in the lysine pathway andconsequently the synthesis of lysine.The normal repression of the LYS gene by lysine

is restored in the lys80,lms double mutant strainsince the synthesis of saccharopine dehydrogenase(lysine forming) is six-fold repressed by lysine likein a wild-type strain (see Table 1). The higher levelon minimal medium is the result of derepression ofthe general control as indicated by a higher level ofargininosuccinate lyase, the product of the geneARG4 which was chosen as a reference enzymesince it only obeys the general control of aminoacid biosynthesis. This derepression is probablydue to lysine starvation but also to starvation foranother amino acid because the presence of lysinein the growth medium does not allow completerestoration of the repression of LYS1 expression.All our data indicate that the effect of the lys80

mutation on expression of LYS genes and onlysine production is a consequence of the improve-ment of the metabolic flux through the lysinepathway. This leads to an increase of áAASAconcentration and to a more efficient activation ofLYS genes by Lys14p.

Role of LYS80 and LMS gene products in theexpression of genes encoding the tricarboxylic acidcycle enzymesSince the á-ketoglutarate intracellular pool is

increased in the lys80 mutant and decreased in thelms mutant strain, we have measured the levels ofthe enzyme activities involved in the production ofá-ketoglutarate such as citrate synthase, aconitaseand NAD isocitrate dehydrogenase. The lys80mutation leads to a two- to three-fold increase ofthese different enzymatic activities whereas in thelms mutant strain, the synthesis of citrate synthaseis reduced four-fold and the level of aconitaseis two-fold weaker than in the wild-type strain(see Table 2). The synthesis of the other enzymesof the tricarboxylic acid cycle such as malate

1340 . .

? 1997 John Wiley & Sons, Ltd. . 13: 1337–1346 (1997)

Table 1. Effect of mutations in the LYS80 genes and of a suppressor of the lys80 mutation on LYS gene expression and lysine production.

Strain (genotype) Media

Specific activity (ìmol/h per mg protein) Pool (nmol/mg dry cells)

Saccharopinedehydrogenase

(glutamate forming)

Saccharopinedehydrogenase(lysine forming)

Arginino-succinatelyase Lysine

á-Keto-glutarate Glutamate

Ó1278b (w.t.) M.am 1·8 15·0 18 1·5 124M.am+lys 0·2 2·5

8761c (lys80-1) M.am 3·1 22·0 110 5·0 200M.am+lys 1·0 9·0

8953b (lys80-2) M.am 4·4 51·0 150 10·0M.am+lys 2·5 28·0

12T7cL80* (ura3,lys80::KanMX4) M.am 4·2 210 10·0M.am+lys 2·1

17T9c* (ura3,lys2,lys80-1) M.am+lys 0·1 3·0 6·013T7d* (ura3,ilv1) M.am 15·0 1·2 18 1·3

M.am+lys 3·013T9a* (ura3,ilv1,lms) M.am 26·0 3·0 1 0·25 20

M.am+lys 9·0 2·514TOd* (ura3,ilv1,lys80-1,lms) M.am 34·0 3·5 1 0·25

M.am+lys 6·0

*If required, uracil (25 ìg/ml) or isoleucine (50 ìg/ml) were added to the growth medium. M.am, minimal medium.

1341

Lys80p

?1997

JohnWiley

&Sons,

Ltd.

.13:

1337–1346(1997)

Table 2. Effect of mutations in the LYS80 genes and of a suppressor of lys80 mutation on the levels of enzymes involved in the production ofá-ketoglutarate. All specific activities are expressed in ìmol of formed product/h per mg protein.

Strain (genotype) Citrate synthase Aconitase

Isocitrate dehydrogenasesMalate

dehydrogenaseNAD NADP

Ó1278b (w.t.) 2·1 1·1 1·0 2·3 9·88761c (lys80-1) 4·3 2·0 2·6 3·3 9·28953b (lys80-2) 7·0 2·4 3·7 3·7 9·212T7cL80* (ura3,lys80::KanMX4) 10·713T7d* (ura3,ilv1) 2·3 1·3 1·8 4·313T9a* (ura3,ilv1,lms) 0·4 0·6 1·7 4·8

All the strains are grown on minimal medium.*If required, uracil (25 ìg/ml) or isoleucine (50 ìg/ml) were added to the growth medium.

1342.

.

?1997

JohnWiley

&Sons,

Ltd.

.13:

1337–1346(1997)

dehydrogenase and isocitrate dehydrogenase(NADP+) is not affected by the lys80 and lmsmutations (see Table 2). It is noteworthy that thesemutations are not linked to CIT1 (mitochondrialcitrate synthase), CIT2 (peroxisomal citratesynthase) or ACO1 (aconitase) loci.The products of LYS80 and LMS thus play

directly or indirectly a role in the synthesis ofenzymes leading to á-ketoglutarate production. Inorder to obtain more information about the func-tion of Lys80p, we have undertaken the cloningand sequencing of the LYS80 gene.

Cloning and sequencing of LYS80 geneWe chose to clone the LYS80 gene by comp-

lementation of a peculiar phenotype resulting fromthe lys80 mutation. We observed that the combi-nation of mutations lys80 and lys9 causes a loss ofcell viability, a behaviour which is not observedwhen the strains also carry a lys2 mutation. Thisprobably results from the fact that these strainspossess an increased metabolic flux that causesa toxic accumulation of áAASA (Zaret andSherman, 1985), this accumulation being avoidedin the presence of the lys2mutation which preventsthe production of áAASA.The cloning was achieved by transforming the

triple lys80,lys9,lys2 mutant strain with a ge-nomic library of S. cerevisiae inserted in a plas-mid bearing the wild-type LYS2 gene. To thisend, we have inserted the product of a partialTaq1 digestion of chromosomal DNA from thewild-type strain Ó1278b in the ClaI restrictionsite of plasmid pLAF59 (pUC19-ARS CEN-LYS2-URA3; see Materials and Methods for con-struction of this vector). The strain 17T7a(lys80,lys9,lys2,ura3) was transformed with thegenomic library and the transformants were se-lected on minimal medium+lysine. Since all theURA+ transformants also contain the LYS2gene, they are unable to grow on this medium.All the transformants except one contained aplasmid bearing the LYS9 gene. From the colonygrowing on minimal medium+lysine which is stilllys9, we recovered a plasmid (pLAF67) with aninsert of about 8 kb that was able to restore thecellular growth of the lys80,lys9,lys2,ura3 strain.The PstI-PstI fragment containing the completeinsert was introduced in the pFL38 vector(pUC19-ARS CEN-URA3), leading to the plas-mid pLAF70. In the lys80,ura3 strain (13T6d)transformed with the plasmid PLAF70, the pro-

duction of lysine and á-ketoglutarate and theexpression of LYS genes in the presence or not oflysine were similar to those of a wild-type strain(see Table 1).The nucleotide sequence of one extremity of

the insert was determined and compared to thesequence of the yeast genome in the MIPS database. This revealed that the cloned DNA frag-ment is located on the left arm of chromosomeXIV and contains four open reading framesYNL074c, YNL075w, YNL076w and YNL077w.This DNA fragment has been sequenced duringthe sequencing of the complete yeast genome butonly the YNL076w ORF has been previouslydescribed as the product of the MKS1 gene(Matsuura and Anraku, 1993). MKS1 encodes aprotein of 584 aa whereas the YNL076w is de-scribed as a protein of 599 aa. The difference inlength of the two ORFs results from the presenceof an intron not mentioned by Matsuura andAnraku (1993).Through a combination of subcloning exper-

iments (see Materials and Methods) and backtransformation of the lys80,ura3 strain (13T6d),the complementing activity was located on a3·1 kb NcoI-BamHI fragment containing theMKS1 gene (see Figure 1). Our data consequentlydemonstrate that the LYS80 and MKS1 loci areidentical.The complete coding sequence of LYS80 iso-

lated from the Ó1278b wild-type strain was deter-mined using synthetic oligonucleotides as primers.This sequence is 99% identical to the YNL076wsequence. There are 17 base substitutions, six ofthem being located in the intron and five of themleading to amino acid replacements (His90 by Tyr,Pro159 by Ala, Asp169 by Glu, Pro126 by Thr andGlu465 by Gly).Complete inactivation of the LYS80/MKS1

gene was achieved by replacing its coding regionwith the KanMX4 cassette (see Materials andMethods). The previously characterized lys80mutations and the lys80 deletion affected simi-larly the production of á-ketoglutarate and theexpression of the genes of lysine biosynthesis andof the tricarboxylic acid cycle. This result con-firms that Lys80p is not essential for cellulargrowth. The mutated lys80-1 and lys80-2 geneswere cloned and sequenced. Both mutations con-sisted of G]A transversions, yielding a stopcodon at amino acid position 347 in the caseof lys80-1 and the replacement of Ser73 by Asnfor lys80-2.

1343 Lys80p

? 1997 John Wiley & Sons, Ltd. . 13: 1337–1346 (1997)

DISCUSSION

The early inference that gene LYS80 encodes arepressor of LYS genes expression was essentiallybased on the observation that lys80 mutationsreduce the repression of LYS genes and increasethe production of lysine without affecting thefeedback inhibition of homocitrate synthase bylysine (Ramos and Wiame, 1985). This conclusionwas, however, difficult to reconcile with morerecent observations such as the influence of lys80mutations on the activation ability of a LexA-Lys14 fusion protein (Feller et al., 1994) or thesuppression of lys80 effects in the absence of afunctional activation system (Lys14p+áAASA).In this study, we show that the lys80 mutations,due probably to an increased production of severalcitrate cycle enzymes, lead to an increase of theá-ketoglutarate pool. Such mutations are conse-quently expected to augment the metabolic fluxthrough the lysine pathway and to result in anincreased concentration of áAASA. Therefore, ourdata demonstrate that Lys80p is not a repressor ofLYS genes and agree with the conclusion thatlysine does not work as a co-repressor but as amodulator of áAASA co-activator production.The products of LYS80 and LMS play directly

or indirectly a role in the synthesis of enzymes

involved in the production of á-ketoglutarate. InS. cerevisiae, the synthesis of citrate synthase andaconitase is repressed by glucose alone and, syner-gistically, by glucose and a glutamate source (Kimet al., 1986; Gangloff et al., 1990). However, thedecreased or increased levels of citrate cycle en-zymes in lms and lys80 mutant cells are indepen-dent of the carbon source, indicating that thesefactors are not involved in catabolite repression(unpublished results). In contrast to the RTG2gene which is a regulator of ACO1 expression onlyunder catabolite repression conditions (Velot et al.,1996), the LMS gene is required for maximalsynthesis of aconitase and citrate synthase in thepresence not only of glucose but also of galactoseor lactate.That Lys80p is not a specific regulator of lysine

biosynthesis is further supported by the demon-stration of the identity of the LYS80 gene with theMKS1 gene. This gene was identified for its abilityto inhibit growth of cells with diminished cyclic-AMP-dependent kinase activity. These data are infavor of a negative regulatory function of Mks1pin transcription of several genes controlled bycAMP (Matsuura and Ankaru, 1993). It is difficultto correlate the data aboutMKS1 and the effect oflys80 mutations on the expression of citrate cyclegenes; a positive or negative effect of cAMP on

Figure 1. Subcloning analysis of the genomic area around the MKS1/LYS80 locus. Subclones repre-sented by the black lines were constructed as described in Materials and Methods. All plasmids were testedfor lys80 complementing activity by transformation of strain 13T6d (lys80,ura3). Complementation wasscored by growth on minimal medium with 10 ìg/ml AEC. The open boxes represent the open readingframes present on the 8 kb PstI-PstI insert from plasmid pLAF70.

1344 . .

? 1997 John Wiley & Sons, Ltd. . 13: 1337–1346 (1997)

expression of these genes has never been reported.Moreover, analysis of the effects of cAMP onprotein synthesis showed that cAMP had only aminor effect on the protein pattern of cells growingexponentially on glucose (Boy-Marcotte et al.,1990). In contrast, deletion of the LYS80 geneaffects the production of enzymes of the tri-carboxylic acid cycle during growth on glucosemedium.Although the primary target of the Lys80p

remains unknown, it is clear that this 599 aaprotein presenting no feature of DNA bindingproteins is required for expression of a large setof genes. Moreover, it is noteworthy that thelys80 mutations were selected for a slower growthon threonine as sole nitrogen source. It wouldconsequently be interesting to study the expres-sion of the CHA1 gene encoding the catabolic-serine/-threonine deaminase in a lys80 deletedstrain.

ACKNOWLEDGEMENTS

We thank Fabienne Vierendeels for excellent tech-nical assistance in the sequencing of LYS80. Weare grateful to M. Crabeel for helpful discussionsand to P. Phillipsen for the gift of plasmid pDP6.This work was supported by the Research Councilof the Universite Libre de Bruxelles.

REFERENCES

Bechet, J., Grenson, M. and Wiame, J.-M. (1970).Mutations affecting the repressibility of arginine bio-synthetic enzymes in Saccharomyces cerevisiae. Eur. J.Biochem. 12, 31–39.

Bhattacharjee, J. K. (1992). Evolution of á-aminoadipate pathway for the synthesis of lysine infungi. In Mortlock, R. P. (Ed.), The Evolution ofMetabolic Function. CRC Press, New York, pp. 47–80.

Bonneaud, N., Ozier-Kalogeropoulos, O., Li, G.,Labouesse, M., Minvielle-Sebastia, L. and Lacroute,F. (1991). A family of low and high copy replicative,integrative and single stranded S. cerevisiae/E. colishuttle vectors. Yeast 7, 609–615.

Boy-Marcotte, E., Tadi, D., Perrot, M., Boucherie, H.and Jacquet, M. (1996). High cAMP levels antagonizethe reprogramming of gene expression that occurs atthe diauxic shift in Saccharomyces cerevisiae. Micro-biology 142, 459–467.

Cryer, D. R., Eccleshall, R. and Marmur, J. (1975).Isolation of yeast DNA. Meth. Cell. Biol. 12, 39–44.

Delbecq, P., Werner, M., Feller, A., Filipkowski, R. K.,Messenguy, F. and Pierard, A. (1994). A segment of

mRNA encoding the leader peptide of the CPA1 geneconfers repression by arginine on a heterologous yeastgene transcript. Mol. Cell. Biol. 14, 2378–2390.

Delforge, J., Messenguy, F. and Wiame, J.-M. (1975).The specificity of argR" mutations and the generalcontrol of amino acid biosynthesis. Eur. J. Biochem.57, 231–239.

Dubois, E., Grenson, M. and Wiame, J.-M. (1974). Theparticipation of the anabolic glutamate dehydrogen-ase in the nitrogen catabolite repression of arginasein Saccharomyces cerevisiae. Eur. J. Biochem. 48,603–616.

Fansler, B. and Lowenstein, J. M. (1969). Aconitasefrom pig heart. In Lowenstein, J. M. (Ed.),Methods ofEnzymology, vol. XIII. Academic Press, New York,pp. 26–30.

Feller, A., Dubois, E., Ramos, F. and Pierard, A. (1994).Repression of the genes for lysine biosynthesis inSaccharomyces cerevisiae is caused by limitation ofLys14-dependent transcription activation. Mol. Cell.Biol. 14, 6411–6418.

Fleig, U. N., Pridmore, R. D. and Philippsen, P. (1986).Construction of LYS2 cartridges for use in geneticmanipulations of Saccharomyces cerevisiae. Gene 46,237–245.

Fujioka, H. and Nakatani, Y. (1970). A kinetic study ofsaccharopine dehydrogenase reaction. Eur. J. Bio-chem. 16, 180–186.

Gangloff, S. P., Marguet, D. and Lauquin, G. J.-M.(1990). Molecular cloning of the yeast mitochondrialaconitase gene (ACO1) and evidence of a synergisticregulation of expression by glucose plus glutamate.Mol. Cell. Biol. 10, 3551–3561.

Gietz, D., St Jean, A., Woods, A. and Shiestl, R. J.(1992). Improved method for high efficiency transfor-mation of intact yeast cells. Nucl. Acids Res. 20,1425–1425.

Holmberg, S. and Schjerling, P. (1996). Cha4p ofSaccharomyces cerevisiae activates transcriptionvia serine/threonine response elements. Genetics 144,467–478.

Ito, H., Fukura, Y., Murata, K. and Timura, A. (1983).Transformation of intact yeast cells treated with alkalications. J. Bacteriol. 153, 163–168.

Jones, E. E. and Broquist, H. P. (1965). Saccharopine,an intermediate of the aminoadipic acid pathwayof lysine biosynthesis. II. Studies in Saccharomycescerevisiae. J. Biol. Chem. 240, 2531–2536.

Karier, P. and Auer, B. (1993). Rapid shuttle plasmidpreparation from yeast cells by transfer to E. coli.BioTechniques 14, 552.

Kim, K.-S., Rosenkrantz, M. S. and Guarente, L.(1986). Saccharomyces cerevisiae contains two func-tional citrate synthase genes.Mol. Cell. Biol. 6, 1936–1942.

Kornberg, A. (1955). Isocitric dehydrogenase of yeast.In Colowick, S. P. and Kaplan, N. O. (Eds),Methods

1345 Lys80p

? 1997 John Wiley & Sons, Ltd. . 13: 1337–1346 (1997)

in Enzymology, vol. I. Academic Press, New York,pp. 705–709.

Matsuura, A. and Anraku, Y. (1993). Characterizationof the MKS1 gene, a new negative regulator of theRas-cyclic AMP pathway in Saccharomyces cerevi-siae, Mol. Gen. Genet. 238, 6–16.

Messenguy, F. (1976). Regulation of arginine biosyn-thesis in Saccharomyces cerevisiae: isolation of acis-dominant, constitutive mutant for ornithinecarbamoyltransferase synthesis. J. Bacteriol. 128,49–55.

Parvin, R. (1969). Citrate synthase from yeast. InLowenstein, J. M. (Ed.), Methods in Enzymology,vol. XIII. Academic Press, New York, pp. 16–19.

Petersen, J. G. L., Kielland-Brandt, M. C., Nilsson-Tillgren, T., Bornaes, C. and Holmberg, S. (1988).Molecular genetics of serine and threonine catabolismin Saccharomyces cerevisiae. Genetics 119, 527–534.

Ramos, F. and Wiame, J.-M. (1985). Mutation affectingthe specific regulatory control of lysine biosyntheticenzymes in Saccharomyces cerevisiae. Mol. Gen.Genet. 200, 291–294.

Ramos, F., Dubois, E. and Pierard, A. (1988). Controlof enzyme synthesis in the lysine biosynthesis pathwayof Saccharomyces cerevisiae. Evidence for a regulatoryrole of gene LYS14. Eur. J. Biochem. 171, 171–176.

Ramos, F., Verhasselt, P., Feller, A., et al. (1996).Identification of a gene encoding a homocitrate syn-thase isoenzyme of Saccharomyces cerevisiae. Yeast12, 1315–1320.

Sambrook, J., Fritsch, E. F. and Maniatis, T. (Eds)(1989). Molecular Cloning. A Laboratory Manual.Cold Spring Harbor Laboratory Press, Cold SpringHarbor, New York.

Sanger, F., Nicklen, S. and Coulson, A. R. (1977). DNAsequencing with chain-terminating inhibitors. Proc.Natl. Acad. Sci. USA 74, 5463–5467.

Tucci, A. F. and Ceci, L. N. (1972). Homocitratesynthase from yeast. Arch. Biochem. Biophys. 153,742–750.

Urrestarazu, L. A., Borell, C. W. and Bhattacharjee,J. K. (1985). General and specific controls of lysinebiosynthesis in Saccharomyces cerevisiae. Curr. Genet.9, 341–344.

Velot, C., Haviernik, P. and Lauquin, G. J.-M. (1996).The Saccharomyces cerevisiae RTG2 gene is a regula-tor of aconitase expression under catabolite repressionconditions. Genetics 144, 893–903.

Wach, A., Brachat, A., Pohlmann, R. and Philippsen, P.(1994). New heterologous modules for classical orPCR-based gene disruption in Saccharomyces cerevi-siae. Yeast 10, 1793–1808.

Wach, A. (1996). PCR-synthesis of marker cassetteswith Long Flanking Homology regions for genedisruptions in Saccharomyces cerevisiae. Yeast 12,259–265.

Zaret, K. S. and Sherman, F. (1985). á-Aminoadipate asa primary nitrogen source for Saccharomyces cerevi-siae mutants. J. Bacteriol. 162, 579–583.

1346 . .

? 1997 John Wiley & Sons, Ltd. . 13: 1337–1346 (1997)