Proc. Nati Acad. Sci. USAVol. 79, pp. 2240-2244, April 1982Biochemistry

Cloning of a yeast gene coding for arginine-specific carbamoyl-phosphate synthetase

(transformation/Saccharomyces cerevisiae/recombinant DNA/cpall gene)

C. J. LuSTY AND JEFFREY LuDepartment of Biochemistry, The Public Health Research Institute ofThe City of New York, New York, New York 10016

Communicated by S. Ratner, January 8, 1982

ABSTRACT Several recombinant plasmids containing cpall,the gene that encodes the large subunit of yeast arginine-specificcarbamoyl-phosphate synthetase [carbamoyl-phosphate synthe-tase (glutamine-hydrolyzing), carbon-dioxide: L-glutamine amido-ligase (ADP-forming, carbamate-phosphorylating), EC 6.3.3.5],have been isolated. The plasmids were selected by transformationof a yeast strain with a mutation in the structural gene ofthe largesubunit of carbamoyl-phosphate synthetase. By using a recombi-nant pool with inserts ofyeast nuclear DNA of5-20 kilobase pairs,we obtained 13 transformants. Of five transformants studied,three have been found to have stable plasmid inserts. These plas-mids could be amplified in Ewcherichia coli and transferred backinto the yeast carbamoyl-phosphate synthetase-deficient strainswith concomitant complementation of the nuclear mutation. Plas-mids pJL2/T1 and pJL2/T5 contain identical nuclear DNA in-serts of 5.9 kilobase pairs. Although the insert of plasmid pJL2/T3 is also 5.9 kilobase pairs long, the sequence overlap with pJL2/TI and pJL2/T5 is only 4.5 Idlobase pairs long. The T3 insert hasan orientation in the vector opposite to that of the T1 and T5 in-serts. The recombinant plasmids with the yeast cpall gene fail tocross-hybridize with a cloned fragment ofE. coli DNA containingthe carA and carB genes for the bacterial carbamoyl-phosphatesynthetase.

Carbamoyl phosphate is essential to both pyrimidine and ar-ginine biosynthesis. As summarized in Table 1, in Escherichiacoli and most bacteria, carbamoyl phosphate is synthesized bya single enzyme, glutamine-dependent carbamoyl-phosphatesynthetase [carbamoyl-phosphate synthetase (glutamine-hy-drolyzing), carbon-dioxide: L-glutamine amido-ligase (ADP-forming, carbamate-phosphorylating), EC 6.3.3.5], which issubject to regulation by ornithine, IMP, and UMP (4). Fungi,including yeast and Neurospora crassa, and higher eukaryoteshave two separate enzymes for carbamoyl phosphate synthesis(5, 6). One enzyme (CPS-P) is specific to the pyrimidine path-way and is located in the nucleus (7); the second enzyme (CPS-A) functions in arginine biosynthesis and as a rule is a mito-chondrial constituent, although in certain organisms, such asSaccharomyces cerevisiae, it appears to be present in the cy-toplasm (8). Yeast carbamoyl-phosphate synthetase (arginine-specific) has been shown to be encoded by two separate genes,designated as cpal and cpaII (5). cpalI has been proposed (4,9) to be the structural gene of the larger subunit with a molecu-lar weight of 140,000. This subunit can catalyze the synthesis ofcarbamoyl phosphate from ammonia, bicarbonate, and ATP (4).The second subunit is required for carbamoyl phosphate syn-thesis when the substrate is glutamine. It has a molecularweightof 36,000 and is encoded by the cpaI gene (4, 9). An analogous

situation occurs in N. crassa (5, 10-13) although, in the latterorganism, the enzyme is located in the mitochondria (10, 11).Since prokaryotic carbamoyl-phosphate synthetase is composedoftwo subunits with molecular weights and catalytic propertiessimilar to those ofeukaryotic arginine-specific carbamoyl-phos-phate synthetase (4, 14, 15), an evolutionary relationship is im-plied. In contrast to the prokaryotic and fungal arginine-specificcarbamoyl-phosphate synthetases, those of mammals consist ofa single polypeptide that has a molecular weight of160,000 (16).It is not clear at present whether the mammalian enzyme arosefrom a gene fusion event or from some other evolutionarymodification.Our studies were undertaken with the general aim of estab-

lishing the relationship of the different carbamoyl-phosphatesynthetases ofE. coli, fungi, and mammalian organisms [whichhave a different form ofthe enzyme, carbamoyl-phosphate syn-thetase (ammonia) carbon dioxide:ammonia ligase (ADP-form-ing, carbamate phosphorylating) EC 6.3.4.16 (15)]. In this com-munication, we report successful cloning ofthe yeast cpalI genein yeast and E. coli. Hybridization studies indicate that the yeastcpall gene and the E. coli carA and carB genes are not suffi-ciently homologous to cross-hybridize even under conditionsof moderate stringency.

MATERIALS AND METHODSYeast and Bacterial Strains. Yeast strains with appropriate

markers for transformation were constructed by standard ge-netic methods (17). The yeast strains used for transformation,S. cerevuisae JL1 (a leu2-3 leu2-112 cpaII-3) and S. cerevisiaeJL2 (a leu2-3 leu2-112 cpu-2 cpall-3), were obtained from across of S. cerevnsiae MG701 (a cpu-2 cpall-3) (18) X S. cere-vsae LLL (a leu2-3 leu2-112) (C. Dieckmann, personal com-munication). E. coli strain 58.161 (recA carB-8 thr- metB) car-rying the recombinant plasmid pMC40 with the bacterial genes(carA and carB) coding for the small and large subunits of car-bamoyl-phosphate synthetase was provided by Marjolaine Cra-beel and Nicolas Glansdorff. E. coli strain RR1 (pro- leu- thi-lacY- hsdR- endA- rpsL2O ara-14 galK2 xyl-5 mtl-i supE44)was used for amplification of all recombinant plasmids.

Transformation of Yeast and E. coli. S. cerevisiae JL2 wastransformed with a recombinant plasmid pool of total yeast nu-clear DNA. The plasmid pool consisted ofa partial Sau3A digest(5- to 20-kilobase pair fragments) of nuclear DNA of yeast (19)ligated with the unique BamHI site of the yeast/E. coli vectorYEpl3 (20). Vector YEp13 is a construct of pBR322, the 2-,umyeast plasmid (sometimes called the "2,u" plasmid), and a frag-ment of yeast nuclear DNA having the leu2 gene. Conditionsfor yeast transformation were those described by Hinnen et al.(21) and Beggs (22), except that the cells were grown on 2%galactose/1% yeast extract/2% peptone medium (C. Dieck-

2240

The publication costs ofthis article were defrayed in part by page chargepayment. This article must therefore be hereby marked "advertise-ment" in accordance with 18 U. S. C. §1734 solely to indicate this fact.

Dow

nloa

ded

by g

uest

on

June

19,

202

1

Proc. Nad Acad. Sci. USA 79 (1982) 2241

Table 1. Carbamoyl-phosphate synthetases of bacteria, fungi,and mammalian liver

No. ofBiosynthetic polypeptides Polypeptide

Organism pathway in monomer Mr GeneE. coli andmost bacteria* Arginine and 130,000 carBCPS II pyrimidine 2 42,000 carA

S. cerevisiaeCPS-A Arginine 2 140,000 cpaI

36,000 cpalCPS-P Pyrimidine 1 Polyprotein cpu

N. crassaCPS-A Arginine 2 125,000 arg-3

45,000 arg-2CPS-P Pyrimidine 1 Polyprotein pyr-3

Rat (liver)CPS I Arginine 1 160,000CPS II Pyrimidine 1 Polyproteint

* Although most bacteria have-a single carbamoyl-phosphate synthe-tase, this is not true ofBacillus subtilis, whichmayhave two separatecarbamoyl-phosphate synthetases (1).

t See refs. 2 and 3.

mann, personal communication). Transformants were selectedon a 2% glucose/0. 67% yeast nitrogen base without amino acids(Difco)/1.2 M sorbitol/3% agar medium. This medium selectsmutants cotransformed for Leu+ and Cps' (where Cps' is des-ignated as carbamoyl phosphate-synthesizing ability).

Transformation of E. coli was carried out as described byPetes et aL (23) with E. coli RR1 serving as host. Transformantswere selected on L medium supplemented with ampicillin at20 jig/ml. The ampicillin-resistant colonies were replica plat-ed onto medium containing tetracycline (20 ktg/ml). Coloniesscored as ampicillin resistant, tetracycline sensitive were as-sumed to contain recombinant plasmids with intact nuclearDNA inserts at the BamHI site of the tetracycline gene.DNA Preparations. Plasmid DNA was purified from E. coli

by lysozyme/Triton lysis followed by phenol extraction andprecipitation with ethanol (24). The same protocol was used forthe preparation of plasmid DNA from yeast except that Zym-olase instead of lysozyme was used to prepare spheroplasts.These preparations were suitable for transformation of eitherE. coli or yeast. Plasmid DNA from E. coli was also preparedby the rapid alkaline extraction method of Birnboim and Doly(25). The latter preparations were used for restriction enzymeanalysis.

Restriction Endonuclease Analysis. Restriction enzymeswere obtained from New England BioLabs. Digestions withrestriction enzymes were carried out under conditions recom-mended by the supplier. The fragments were separated on 1%agarose slab gels in Tris base/boric acid/EDTA (26). The DNAfragments were stained with ethidium bromide, and their sizeswere determined by comparison with the known sizes of the

HindIII fragments of A DNA and the Hae III fragments of4X174 replicative form DNA.

Southern Blot Hybridization. Plasmid preparations, pJL2/T1 and pJL2/T3, containing the yeast cpaII gene were digestedwith EcoWl/Sal I. The digestion products were separated on1% agarose slab gels and transferred to diazobenzyloxymethylpaper according to the procedure of Alwine et aL (27), and theimmobilized DNA was hybridized with a nick-translated radio-active probe containing the carA and carB genes ofE. coli. Theprobe was isolated from a recombinant plasmid pMC40 shownto contain both the E. coli genes (M. Crabeel, personal com-munication). Plasmid DNA was digested with HindIII and the5.6-kilobase pair E. coli insert carrying the two genes was sep-arated from the pBR322 vector by electrophoresis on 1% low-melting-temperature agarose. The cloned DNA fragment wasextracted from heat-melted agarose with phenol and furtherpurified by precipitation with ethanol. carA and carB genes(28) were radioactively labeled by nick-translation (29) and usedfor Southern blot analysis. [a-32P]dATP (250 ,uCi; 400-600 Ci/mmol; 1 Ci = 3.7 x 1010 becquerels; Amersham) was used tolabel the probe to a specific activity of5 x 107 dpm/,ug ofDNA.

RESULTS

Selection of the cpall Gene by Transformation of Yeast.Yeast strain JL2, carrying mutations in cpu, cpalI, and a doublemutation in leu2, allowed selection of cotransformants for bothleu2+ and cpaII+ or cpu+. As transforming DNA we used a re-combinant plasmid pool constructed by insertion of 5- to 20-kilobase pair Sau3A fragments of wild-type yeast nuclear DNA(19) into the YEpl3 chimeric vector (20). The YEp13 vectorcontains pBR322, a fragment of the 2-,m yeast circle with theorigin of replication, and a fragment of yeast nuclear DNA withthe leu2 gene. In the recombinant plasmid, the Sau3A frag-ments were inserted at the BamHI site of pBR322 (19).

It is known that in yeast carbamoyl phosphate synthesizedby either CPS-P or CPS-A enters a common pool that can beused for either arginine or pyrimidine biosynthesis (5). Con-sequently, mutations in either gene do not prevent growth onminimal glucose medium. cpaII- or cpu- mutants can be dis-tinguished by their growth properties on minimal medium con-taining uracil. Thus, cpaII mutants still capable of derivingcarbamoyl phosphate from CPS-P are inhibited by uracil, dueto feedback inhibition of CPS-P by UMP (4). Although it hasbeen reported that growth of cpu mutants is inhibited by ar-ginine due to repression ofCPS-A by arginine (5), we have notobserved this effect on solid medium and therefore could notuse this test to distinguish cpaII and cpu mutants.

In a single transformation of S. cerevisiae JL2 with the YEpl3pool, 13 independent clones were obtained with selection, se-lection being carried out on minimal medium. As shown in Ta-ble 2, strains growing on this medium should be complementedfor the leu2 and either the cpall or the cpu mutations. It is un-likely that mutations in both cpaII and cpu would be comple-

Table 2. Growth requirements of yeast strains carrying mutations in leu2, cpu, and cpaIIAddition(s) to minimal medium

Gene(s) in Leucine/transforming Chromosomal Leucine/ arginine/

Strain plasmids genotype None Leucine Uracil uracil uracilJL1 None leu2-3 leu2-112 cpaII-3 - + - - +JL1 LEU2 CPAII leu2-3 leu2-112 cpaII-3 + + + + +JL2 None leu2-3 leu2-112 cpu-2 cpaII-3 - - - - +JL2 LEU2 CPU leu2-3 leu2-112 cpu-2 cpaII-3 + + - - +JL2 LEU2 CPAII leu2-3 leu2-112 cpu-2 cpaII-3 + + + + +

Biochemistry: Lusty and Lu

Dow

nloa

ded

by g

uest

on

June

19,

202

1

2242 Biochemistry: Lusty and Lu

Table 3. Segregation of yeast transformantsAdditions to

minimal glucoseColonies Arginine/

Transformant scored None Uracil uracil % stableJL2/T1 107 47 47 107 44JL2/T3 170 100 100 170 59JL2/T4 117 61 ND ND 52JL2/T5 179 82 82 179 46

Leu'Cps' transformants (in strain S. cerevisiae JL2) were grownunder nonselective conditions in YPD medium (1% yeast extract/2%peptone/2% glucose) for 24 hr at 300C and then diluted and spread forsingle colonies on solid YPD medium. The colonies were replica platedonto minimal glucose medium with various supplements and-scoredafter growth for 2 days at 300C. In all cases tested, colonies scored aspositive on minimal glucose also grew in the presence of uracil. Sta-bility of the transforming plasmid is expressed as percent of coloniesthat grew on minimal or minimal/uracil medium. ND, not done.

mented by a single plasmid since the two genes are known tobe genetically unlinked in yeast (5, 18). The 13 transformantswere tested for growth on minimal medium supplemented withuracil. None of the clones were inhibited by uracil, indicatingthat the mutation being complemented was most' likely in thecpaII gene.

Properties of the cpall Transformants. The Leu+Cps' phe-notype was verified to be due to complementation by autono-mously replicating plasmids for some of the transformed clones.Five clones, JL2/T1-JL2/T5, were grown in rich medium (1%yeast extract/2% peptone/2% glucose) to stationary phase.Cells from the cultures were diluted and spread on rich agar toproduce single colonies from isolated cells. The colonies formedafter several days were replica plated onto minimal mediumwith and without uracil. The results of this experiment (Table3) show that, under nonselective conditions, a significant per-centage of the population reacquires the original growth re-quirements ofthe host cells. The segregation ofthe transformedphenotype is consistent with the presence of wild-type leu2 andcpalI genes in a plasmid DNA.

Plasmid DNA was prepared from 7-ml cultures of each of thefive transformants (JL2/T1-JL2/T5). Part of the DNA prepa-rations were used to transform E. coli RR1. Since the nuclearDNA fragments in the YEp13 pool are inserted in the BamHIsite of the tet gene, E. coli transformants were selected for am-picillin resistance and tetracycline sensitivity. The tetracyclinesensitivity ensured that the transforming plasmid retained thenuclear DNA insert. All of the plasmid preparations except thatobtained from JL2/T2 transformed E. coli to ampicillin resis-tant, tetracycline sensitive. In the case of the transformationwith JL2/T2 DNA, for reasons unknown, no ampicillin-resis-tant transformants were obtained. Plasmid DNAs were pre-pared from the E. coli transformants by the rapid method ofDavis et al. (24) and used to back transform S. cerevisiae JL1(leu2-3 leu2-112 cpaII-3) and JL2 (leu2-3 leu2-112 cpu-2 cpaII-3). Transformants selected solely for complementation of theleu2 mutation were picked from plates and their growth prop-erties were tested on minimal medium alone and on minimalmedium supplemented with uracil. Of the four plasmid prep-arations used, three (pJL2/Tl, pJL2/T3, and pJL2/T5) cotrans-formed S. cerevisiae JL1 and JL2 for both the leu2 and cpaIImutations (Table 4). The fact that JL1, which lacks the cpumutation, was also transformed by the plasmid DNA confirmsthat the cloned gene is that of cpaII.

Restriction Enzyme Analysis of pJL2/TI, pJL2/T3, andpJL2/T5. The transforming plasmids pJL2/Tl, pJL2/T3, andpJL2/T5 were subjected to restriction analysis. Digestion ofthe

Table 4. Back transformation of S. cerevisiae JL1 and JL2'Growth of Leu2+

Recipient Transforming tranormntsstrain DNA Minimal Minimal/uracil

S& cerevisiae JL1 pJL2/T1 + +(leu2-3 leu2-112 pJL2/T3 + +cpaII-3) pJL2/T4 + -*

pJL2/T5 + +S. cerevisiae JL2

(leu2-3 leu2-112 pJL2/T1 + +cpu.2 cpalI-3) pJL2/T3 + +

pJL2/T4 - -*pJL2/T5 + +

PlasmidDNA from yeast transformants (JL2/T1-JL2/T5) was am-plified in E. coliand used for transformations of S. cerevisiae JLL andJL2. Transformants were selected for Leu+ on minimal medium con-taining uracil (25 ,ug/ml) and arginine (25 yg/ml). Fifteen transform-ants from each plate were picked and replica plated onto minimalmedium containing uracil and arginine. After 2 days at'30°C, theplates were replicated onto the indicated medium. Growth was scoredafter 2 days at 30°C. +, All 15 colonies picked grew on the medium.* Plasmid pJL2/T4 underwent deletion in the DNA insert, probablyduring subcloning in E. coli.

purified recombinant plasmid DNAs with BamHI or EcoRI/Sal I indicated that pJL2/T1 and pJL2/T5 had identical DNAinserts (Fig. 1). Based on the known sizes of the EcoRI/Sal Ifragments in the vector, the insert of pJL2/Tl (and T5) wasestimated to be 5.9 kilobase pairs. The insert in pJL2/T3 wasalso 5.9 kilobase pairs. Two EcoRI/Sal I fragments belongingto the insert have identical sizes in pJL2/Tl/T5 and pJL2/T3,suggesting an overlap in their DNA sequence (Fig. 1). This wasconfirmed by digestion of preparative BamHI fragments ofpJL2/Tl and pJL2/T3 with Hinfand HindIII. The BamHI frag-ments of the two different plasmids were cleaved into a largenumber ofsmaller fragments, some ofwhich had identical sizes.The BamHI, EcoRI, and Sal I sites of the two different insertswere mapped by digestion of the whole plasmid and ofthe pre-parative BamHI fragments with different combinations of re-striction enzymes. As shown by the derived restriction maps(Fig. 2) ofthe two inserts, the inserts in pJL2/Tl/T5 and pJL2/T3 have an overlap of4.5 kilobase pairs. The sizes offragmentsgenerated in digestions of the plasmids indicated that inserts

2 3 4

kbp

23 0-9.84.5-2.5 -

2123

5 6 7 8

kbp

4.5t-v 2.5

2.2-1.3 -I-

0.9---0.6-0.3 -

- v-v-v

FIG. 1. Restriction fragments of plasmids pJL2/T1, pJL2/T3, andpJL2/T5. Purified plasmid DNA was digested with BamHI and withEcoRI/Sal I, and the products were separated on a 1% agarose slab geland stained with ethidium bromide. Lanes: 1 and 5, mixture of aHihdM digest of A DNA and aHaem digest of 4X174 replicative formDNA; 2, BamHI digest of pJL2/T1; 3, BamIl digest of pJL2/T3; 4,BamHI digest of pJL2/T5; 6, EcoRI/Sal I digest of pBR322; 7, EcoRI/Sal I digest of pJL2/T1; 8, EcoRI/Sal I digest of pJL2/T3. kbp, Ki-lobase pairs of standard fragments; v, vector fragments.

Proc. Nad Acad. Sci. USA 79 (1982)

Dow

nloa

ded

by g

uest

on

June

19,

202

1

Proc. NatL Acad. Sci. USA 79 (1982) 2243

I kbpl-

E B

pJL2/TIlPJL2 /T5

pJL2/T3:

E SI I

S BS11

E BEI

E BE11. I, ,

FIG. 2. Restriction maps of the pJL2/Tl/T5 and pJL2/T3 inserts. BamHI (B), EcoRI (E), and Sal I (S) restriction sites are shown for both theinsert and the arms of the pBR322 component of the vector.-, Inserts [in both pJL2/Tl/T5 and pJL2/T3, these are -5.9 kilobase pairs (kbp) long];3, pBR322 arms. The left-arm boundary of the pJL2/T3 insert has a BamHI site that has been formed as a result of the ligation.

in pJL2/Tl/T5 have an opposite orientation to that of pJL2/T3 (Fig. 3).

Cross-Hybridization of the Cloned Yeast cpall and E. colicarB Genes. The availability ofthe cloned E. coli carA and carBgenes (28) in a single plasmid (pMC40) made it possible to ex-amine the question of sequence homology between the pro-karyotic and yeast genes. The EcoRI/Sal I and BamHI frag-ments of pJL2/Tl and pJL2/T3 were separated by electropho-resis on agarose and transferred to diazobenzyloxymethyl paper(27). The immobilized DNA fragments were hybridized to acloned fragment (5.6 kilobase pairs) of E. coli DNA containingthe adjacent carA and carB genes (28) for carbamoyl-phosphatesynthetase. The E. coli probe made radioactive by nick trans-lation failed to hybridize to any ofthe fragments originating fromthe nuclear inserts of the yeast gene even under conditions ofmoderate stringency (30% formamide at 45°C). Under both setsof conditions, good hybridization was observed with homolo-gous DNA.

DISCUSSION

The synthesis of carbamoyl phosphate is essential for both ar-ginine and pyrimidine biosynthesis. Although the different car-bamoyl-phosphate synthetases in prokaryotes and eukaryoteshave, in a number of instances, been well characterized (4,

E BE

S

B n

E, ~ n.1l P/Tl

9-14), their evolutionary relationships remain obscure. Thisquestion can be most directly answered by comparison of theprimary structures of the proteins. One of the distinguishingfeatures of the prokaryotic and eukaryotic arginine-specific car-bamoyl-phosphate synthetases is the unusually high molecularweight of the monomers. As a result, no attempt has been madeto determine the sequences of the bacterial, fungal, or mam-malian enzymes.

As an alternative approach, we have attempted to obtain theamino acid sequences of the prokaryotic and several eukaryoticcarbamoyl-phosphate synthetases through their gene nucleo-tide sequences. In this paper, we describe the cloning of thegene for the large subunit ofthe yeast arginine-specific enzyme.The clones isolated by the yeast transformation assay have beenshown to contain plasmids that complement the previously de-scribed mutation (5, 18) in the cpalI gene of the large subunit.The pJL2/Tl/T5 and pJL2/T3 plasmids are suitable for se-

quence analysis ofthe gene. The nuclear inserts in the two typesof plasmids have a common sequence of 4.5 kilobase pairs.Based on the reported molecular weight of the large subunit(140,000) (9), the region of overlap in the cloned nuclear DNAis only slightly larger (600 base pairs) than the expected lengthof the gene, assuming that there are no intervening sequences.The similarities in the subunit compositions, molecular

weights, sulfhydryl reactivities, and catalytic properties of the

E BI/ E

E / E

E ~~~~~~~~~~E

FIG. 3. Orientation of the nuclear DNA inserts in pJL2/Tl/T5 and pJL2/T3. Both recombinant plasmids are 16.6 kilobase pairs long. O,pBR322; M, 2-,um yeast DNA; *, leu2 gene (these three make up 10.7 kilobase pairs); -, insert; -+, orientation of the nuclear DNA insert withthe cpaII gene.

sI

B E

I m

Biochemistry: Lusty and Lu

Dow

nloa

ded

by g

uest

on

June

19,

202

1

2244 Biochemistry: Lusty and Lu

E. coli carbamoyl-phosphate synthetase and the yeast arginine-specific carbamoyl-phosphate synthetase (4, 14, 15) suggest thatthe two enzymes might also be homologous in their amino acidsequence. The failure to observe cross-hybridization ofthe twogenes indicates that, ifthe two enzymes are related, substantialchanges have occurred in their DNA sequences and probably,also, in their primary structures. This question will be resolvedwhen the nucleotide sequences of the genes are available.

We thank Drs. Maijolaine Crabeel and Nicolas Glansdorff for pro-viding us with yeast strain MG701 and the E. coli clones, pMC40 andpMC50, and Dr. Carol Dieckmann for her help in setting up the yeasttransformation system. The YEp13 plasmid bank ofyeast nuclear DNAwas constructed by Dr. Kim Nasmyth. This work was supported byGrant GM 25846 from the National Institutes of Health.

1. Potvin, B. & Gooder, H. (1975) Biochem. Genet. 13, 125-143.2. Coleman, P. G., Suttle, D. P. & Stark, G. R. (1977) J. Biol.

Chem. 252, 6379-6385.3. Davidson, J. N. & Patterson, D. (1979) Proc. Natl Acad. Sci. USA

76, 1731-1735.4. Pierard, A., Grenson, M., Glansdorff, N. & Wiame, J.-M. (1973)

in The Enzymes of Glutarmne Metabolism, eds. Pruisner, S. &Stadtman, E. R. (Academic, New York), pp. 483-503.

5. LaCroute, F., Pierard, A., Grenson, M. & Wiame, J.-M. (1965)J. Gen. Microbiol 40, 127-142.

6. Davis, R. H. (1967) in Organizational Biosynthesis, eds. Vogel,H. J., Lampen, J. 0. & Bryson, V. (Academic, New York), p.303.

7. Nagy, M., LaPorte, J., Penverne, B. & Herve, G. (1982) J. CellBioL 92, 790-794.

8. Urrestarazu, A., Vissers, S. & Wiame, J.-M. (1977) Eur. J.Biochem. 79, 473-481.

9. Pierard, A. & Schroter, B. (1978)J. Bacteriol 134, 167-176.10. Davis, R. H. (1972) Science 178, 835-840.11. Weiss, R. L. & Davis, R. H. (1973) J. BioL Chem. 248,

5403-5408.

12. Bernhardt, S. A. & Davis, R. H. (1969) Proc. Natl Acad. Sci. USA69, 1868-1872.

13. Davis, R. H., Ristow, J. L. & Hanson, B. A. (1980) J. Bacteriot141, 144-155.

14. Trotta, P. P., Pinkus, L. M., Wellner, V. P., Estis, L., Hasche-meyer, R. H. & Meister, A. (1973) in The Enzymes ofGlutamineMetabolism, eds. Pruisner, S. & Stadtman, E. R. (Academic,New York), pp. 431-482.

15. Jones, M. E. (1972) in Current Topics in Cellular Regulation,eds. Horecker, B. L. & Stadtman, E. R. (Academic, New York),Vol. 6, pp. 227-265.

16. Lusty, C. J. (1978) Eur. J. Biochem. 83, 373-383.17. Mortimer, R. K. & Hawthorne, D. C. (1969) in The Yeasts, eds.

Rose, A. H. & Harrison, J. S. (Academic, New York), pp.385-460.

18. Thuriaux, P., Ramos, F., Pierard, A., Grenson, M. & Wiame, J.-M. (1972)J. MoL Biol 67, 277-287.

19. Nasmyth, K. A. & Reed, S. I. (1980) Proc. NatL Acad. Sci. USA77, 2119-2123.

20. Broach, J. R., Strathern, J. N. & Hicks, G. R. (1979) Gene 8,121-133.

21. Hinnen, A., Hicks, J. B. & Fink, G. R. (1978) Proc. Natl Acad.Sci. USA 75, 1929-1933.

22. Beggs, J. D. (1978) Nature (London) 275, 104-109.23. Petes, T. D., Broach, J. R., Wensink, P. G., Hereford, L. M.

& Fink, G. R. (1978) Gene 4, 37-49.24. Davis, R. W., Thomas, M., Cameron, J., St. John, T. P.,

Scherer, S. & Padgett, R. A. (1980) Methods Enzymol, 65,404-411.

25. Birnboim, H. C. & Doly, J. (1979) Nucleic Acids Res. 7,1513-1523.

26. Peacock, A. C. & Dingman, C. W. (1968) Biochemistry 7,668-674.

27. Alwine, J. C., Kemp, D. J., Parker, B. A., Reiser, J., Renart, J.,Stark, G. R. & Wahl, G. M. (1979) Methods Enzymol, 68,220-242.

28. Mergeay, M., Gigot, D., Beckmann, J., Glansdorff, N. & Pier-ard, A. (1974) Mol Gen. Genet. 133, 299-316.

29. Jeffreys, A. J. & Flavell, R. A. (1977) Cell 12, 429-439.

Proc. Nad Acad. Sci. USA 79 (1982)

Dow

nloa

ded

by g

uest

on

June

19,

202

1