Vol. 157, No. 1JOURNAL OF BACTERIOLOGY, Jan. 1984, p. 277-2820021-9193/84/010277-06$02.00/0Copyright © 1984, American Society for Microbiology

Identification of the Structural Gene and Nonsense Alleles forAdenylate Cyclase in Saccharomyces cerevisiae

KUNIHIRO MATSUMOTO,' ISAO UNO,2 AND TATSUO ISHIKAWA2*

Department of Industrial Chemistry, Tottori University, Tottori 680,1 and Institute of Applied Microbiology, University ofTokyo, Tokyo 113,2 Japan

Received 23 March 1983/Accepted 2 October 1983

Tetraploid strains of Saccharomyces cerevisiae carrying different dosages of the CYR]+ gene have beenconstructed. Adenylate cyclase activity observed in these tetraploid strains was proportional to the dosageof the active CYRJ+ gene. Of the 57 mutants requiring adenosine 3',5'-monophosphate for growth at 35°C,two allelic temperature-sensitive cyri mutants produced thermolabile adenylate cyclase. Crude extract andplasma membrane fraction of cyri mutant cells had no adenylate cyclase activity when assayed with GTP or5'-guanylyl imidodiphosphate in the presence of Mn2+ or Mg2 . Plasma membrane and Lubrol-solubleplasma membrane fractions obtained from the temperature-sensitive cyri mutant were thermolabilecompared with those from the wild-type strain. Three cyri mutants carried nonsense mutations susceptibleto ochre (UAA) suppressors, SUP3 and SUP-o, and had no detectable level of adenylate cyclase activity. Itis concluded that the cyri mutants carry lesions in the structural gene for adenylate cyclase.

Adenosine 3',5'-monophosphate (cAMP) has been shownto be involved in the control of a variety of metabolicprocesses in eucaryotes as well as procaryotes (15). In anattempt to determine the physiological roles of cAMP inyeast, we isolated cyrl mutants which were deficient inadenylate cyclase activity (11). Growth of cells carrying thecyri mutation was arrested at the Gl phase of the cell cyclein the absence of cAMP (11). The cyri mutation wassuppressed by the bcyl mutation, which resulted in thedeficiency of the regulatory subunit of cAMP-dependentprotein kinase and the production of a high level of a cAMP-independent one (11). Thus, we have presented evidencethat cAMP is an essential factor for yeast cells to proceedthrough the cell cycle via the activation of protein kinase.

It has been reported that the mammalian adenylate cyclaseis composed of at least two protein components in additionto hormone receptors, neither of which was catalyticallyactive when assayed independently (16). One component, athermolabile protein, was retained in membranes of a mouseS49 lymphoma clone that is phenotypically deficient inMg2+-dependent adenylate cyclase (17). An active adenylatecyclase is found in the plasma membrane of yeasts (4, 7).Although it is known that the membrane-bound enzymerequires Mn2+ and is stimulated by GTP (G. F. Casperson etal., J. Biol. Chem., in press), the molecular nature of thisenzyme is not yet clear.This communication reports the isolation of temperature-

sensitive mutants and ochre nonsense mutants at the cyrilocus and indicates that the CYR] locus is the structural genefor adenylate cyclase.

MATERIALS AND METHODSYeast strains. Genotypes and sources of strains of Saccha-

romyces cerevisiae used in this experiment are listed inTable 1.

Media. General use and composition of minimal, yeast-peptone-glucose-containing (YPGlu), and sporulation mediahave been described previously (10, 14). YPGlu/cAMP medi-um was YPGlu medium supplemented with filter-sterilizedcAMP (1 mM) where necessary. To test auxotrophic mark-

* Corresponding author.

ers, omission media were prepared from a basal medium ofyeast nitrogen base without amino acids.

Genetic techniques. The methods used for genetic analyseshave been described by Nogi et al. (14).

Construction of tetraploids. Four haploid strains were usedto construct tetraploid strains: AL224-2C (ot CYRI+ leul),AL224-4A (a CYRJ' leul), AM45-2C (a cyrl-J leul), andAM45-3B (ca cyrl-J leul). Diploids homozygous for matingtype (a/a and a(/ox) were derived from regular a/a diploids by a1-min UV light irradiation to induce mitotic recombination.Mitotic recombinants which were homozygous for a or awere scored by their ability to yield prototrophic triploidswhen crossed with a or at haploid strains containing comple-mentary auxotrophic markers. Tetraploid strains were con-structed by mating a/a diploids with ax/a diploids.

Preparation of enzyme samples. All experiments werecarried out at 4°C. Cells were harvested by centrifugationand suspended in buffer T containing 50 mM Tris-hydrochlo-ride buffer (pH 7.4), 1 mM EDTA, 1 mM mercaptoethanol,and 0.5 mM phenylmethylsulfonyl fluoride. The suspensionobtained was homogenized with an Aminco French pressurecell (J5-598A) at 10,000 lb/in2. The resulting homogenate wascentrifuged at 1,000 x g for 10 min. The supernatant fluidwas used as a crude extract. The 105,000 x g precipitate wasobtained by centrifugation of the crude extract at 105,000 xg for 60 min.

Preparation of plasma membrane fraction. Plasma mem-brane fractions were prepared by a modification of themethod of Duran et al. (2). Cells were grown in YPGlumedium (10 ml) to 1 x 108 to 2 x 108/ml, harvested bycentrifugation, washed with 1 M sorbitol containing 20 mMpotassium phosphate buffer (pH 7.0) (solution A), and resus-pended in 1 ml of solution A. Zymolyase 60000 was added toa final concentration of 0.4 mg/ml, the mixture was incubat-ed at 30°C for various lengths of time with monitoring of thepercentage of spheroplasts formed, and then 4 ml of chilledsolution A was added. The spheroplasts were collected bycentrifugation and resuspended gently in 1 ml of 0.8 Msorbitol solution containing 10 mM MgC92, 1 mM CaCl2, 1mM MnCl2, 0.1 mM EDTA, and 50 mM Tris-hydrochloridebuffer (pH 7.5) (solution B). An equal volume of concanava-lin A (0.5 mg/ml in solution B) was added, the mixture was

277

on May 29, 2021 by guest

http://jb.asm.org/

Dow

nloaded from

278 MATSUMOTO, UNO, AND ISHIKAWA

TABLE 1. Strains usedStrain Genotype Source or reference

FP-1A a wild type Segregant from FlOD (a wild type [9])x P-28-24C (a wild type) cross

G185-7D a SUP3 9G211-2A a gal7-2 9G427-1C a wild type 19P-28-24C a wild type 9AL224-2C a leul Haploid clone from the same family as

AL224-4AAL224-4A a leul 6AM3-4B a ade6 adelO 12

ampl camicam2 cam3

AM7-11D a cyrl-J 11AM18-5C a cyrl-1 Segregant from AM7-11D X FlOD crossAM45-2C a cyrl-l leul Segregant from AM18-5C x AL224-4A

crossAM45-3B a cyrl-l leul Haploid clone from the same family as

AM45-2CAM69-7D a bcyl leul Segregant from AM9-10A (a cyrl-1

bcyl [11]) x G435-1A (a leul); segre-gant from G98-2D (a wild type [19])x AL224-4A cross

AM76-2C a cyrl -l SUP3 Segregant from AM18-5C x G184-3B (agal4-2 SUP3); segregant from MG2(a gaI4-2 SUP3 [9]) x P612C (a gaI4-2 [9]) cross

AM87-1C a bcyl Segregant from AM9-8B (a bcyl [11]) xFlOD cross

AM88-1A a cyrl-3 Segregant from AM85-6C (a cyrl-3 [thisstudy]) x AM3-4B cross

AM91-1B a cyrl-3 hisS-2 Constructed by repeated crosses andIysl-l trp5- tetrad dissections among X2310-12D48 (a hisS-2 IysJ-J canl-100 trp548

ade2-1 ura3-1 SUPS; Yeast GeneticsStock Center, University of Califor-nia), P-28-24C, and AM88-1A

AM91-4B a cyrl-3 his5-2 Haploid clone from the same family asIysJ-I trpS- AM91-1B48

incubated at 30°C for 10 min, and the spheroplasts werecollected by centrifugation. The spheroplasts were lysed bythe addition of 5.5 ml of 25 mM piperazine-N,N'-bis(2-ethanesulfonic acid) buffer (pH 6.2) containing 1.1 mMMnC12, 0.1 mM EDTA, and 1 mM phenylmethylsulfonylfluoride (solution C), followed by homogenizing in a Potter-Elvehjen homogenizer. The crude plasma membrane frac-tion was collected by centrifugation at 20,000 x g for 45 minand resuspended gently in 1 ml of solution C.To solubilize the membrane-bound adenylate cyclase,

Lubrol was added to the crude plasma membrane fraction(about 5 mg of protein per ml) at a final concentration of1.0% and kept for 60 min at 4°C. The Lubrol-soluble fractionwas obtained by centrifugation for 60 min at 105,000 x g.

Adenylate cyclase assay. Adenylate cyclase was assayed asdescribed by Uno et al. (20). One unit of the enzyme activitywas defined as the amount which produced 1 pmol of cAMPin 1 min at 30°C.cAMP assay. cAMP content was measured by using the

cAMP assay kit as described by Uno et al. (20).Protein measurement. Protein was measured by the meth-

od of Lowry et al. (8) with bovine serum albumin as thestandard.

Chemicals. [3H]cAMP and cAMP assay kits were pur-chased from Amersham International, Amersham, UnitedKingdom; ATP, GTP, concanavalin A, 5'-guanylyl imidodi-

phosphate [Gpp(NH)p], and phenylmethylsulfonyl fluoridewere from Sigma Chemical Co., St. Louis, Mo.; Zymolyase60000 was from Kirin Brewery Co., Gumma, Japan; yeastnitrogen base without amino acids was from Difco Labora-tories, Detroit, Mich.; and Lubrol was from Nakarai Chemi-cals, Ltd., Tokyo, Japan.

RESULTSEffect of CYRI' gene dosage on adenylate cyclase activity.

To test the effect of CYR]' gene dosage on adenylatecyclase activity, five tetraploid strains were constructedbearing from zero to four wild-type alleles and from four tozero cyrl-l alleles as described above. These tetraploid cellswere grown in YPGlu or YPGlu/cAMP medium at 30°C, andthe specific activity of adenylate cyclase in the 105,000 x gprecipitate was determined. Tetraploid cells carrying noCYR] + gene produced no detectable amount of adenylatecyclase. Proportionality between specific activity of adenyl-ate cyclase and gene dosage was very good (Fig. 1). Eachwild-type allele (CYRI +) contributed approximately 2.3 U ofadenylate cyclase activity per mg of protein.

Isolation and characterization of temperature-sensitive cyrimutants. Temperature-sensitive cyri mutants were isolatedfrom AM3-4B cells by the method described previously (11),except that the cAMP requirement was tested at 35°C. Fifty-seven mutants that required cAMP for growth at 35°C wereisolated, and two mutants (AM3-4BM-13 and AM3-4BM-20)among them could grow without cAMP at 25°C. These twomutants were crossed with the wild-type strain (G427-1C).The resulting diploids grew on YPGlu medium at both 25 and35°C, indicating that these mutations are recessive to thewild-type counterpart. The diploids were sporulated, andfour-spored asci were dissected. All the asci tested, 11 and12, respectively, showed a 4+:0- segregation on YPGlu

- 1 0C

0

EI._

6U)n

04

2U

<0_ t + + +

_- + + +

- + +

+

Gene dosageFIG. 1. Adenylate cyclase activity of tetraploid strains bearing

various combinations of cyrl-J and wild-type alleles. Cells of eachtetraploid were grown in YPGlu or YPGlu/cAMP medium at 30°C. +and - symbols below the abscissa indicate the wild-type and mutantalleles, respectively.

J. BACTERIOL.

on May 29, 2021 by guest

http://jb.asm.org/

Dow

nloaded from

STRUCTURAL GENE FOR YEAST ADENYLATE CYCLASE 279

medium at 25°C, whereas they showed a 2+:2- segregationon the same medium at 35°C. This fact indicates that eachmutant bears a single chromosomal mutation. Haploid segre-gants from the above crosses bearing a mutant allele werethen crossed with the cyrl-J mutants, AM7-11D (a) andAM18-SC (a). The resulting diploids were tested for theirability to grow on YPGlu medium at 35°C. Two mutantsfailed to complement the cyrl-l mutant. We concluded thatthese two mutants, AM3-4BM-13 and AM3-4BM-20, aretemperature-sensitive cyri mutants and designated the mu-tant allele cyrl-2(ts). From the diploid constructed by theAM3-4BM-20 x G427-1C cross, AM26-2A [a cyrl-2(ts)] andAM26-2B (a wild type) were selected for further analyses.The cyrl-2(ts) cells grown on YPGlu/cAMP medium at

35°C had no detectable adenylate cyclase activity, but thosegrown on YPGlu medium at 25°C had about 6% of the wild-type activity (Table 2). The cAMP level found in cyrl-2(ts)mutant cells grown at 25°C was about 19% of that for wild-type cells. To examine the cAMP level in cyrl-2(ts) mutantcells at 35°C without interference by cAMP added in the.culture medium, the double mutant cyrl-2(ts) bcyl was

constructed by a cyrl-2(ts) (AM26-2A) x bcyl (AM69-7D)cross. In our previous report, the bcyl mutant was found tobypass the cyri defect without restoring adenylate cyclaseactivity (11). When diploids constructed by a cyrl-2(ts) xbcyl cross were analyzed, three ascus types showing2+:2-, 3+:1-, and 4+:0- segregations on YPGlu mediumat 35°C appeared in 5, 7, and 1 asci, respectively, but at 25°Call 13 asci showed 4+:0- segregation on the same medium.Adenylate cyclase activity and cAMP level of the cyrl-2(ts)bcyl segregant (AM81-1C) from the above cross were at thesame level as those of the cyrl-2(ts) mutant at 25°C (Table 2).

Characterization of wild-type and cyri mutant adenylatecyclase. Adenylate cyclase activity in the wild-type crudeextract and plasma membrane fraction was high when Mn2+was the supplement but significantly low when Mg2+ was

used in place of Mn2+, whereas the cyrl-J crude extract and

plasma membrane fraction showed no signs of adenylatecyclase activity in the presence of either Mn2+ or Mg2+(Table 3). The wild-type enzyme activity was stimulatedabout twofold in the presence of GTP (10 mM) or Gpp(NH)p(10 mM) but not by NaF (10 mM) or CaC12 (0.1 mM) (Table3). However, no adenylate cyclase activity was observed inthe cyrl-J enzyme samples, even when GTP or Gpp(NH)pwas supplemented (Table 3).Evidence that the cyrl-2(ts) mutation results in altered

adenylate cyclase was obtained by incubating crude enzymesamples of mutant and wild type at 450C for various lengths

of time. The wild-type enzyme activity was stable for at least10 min at 45°C, but the enzyme activity of the cyrl-2(ts)mutant was inactivated rapidly at 45°C. The inactivation rateof mixture of the wild-type and mutant enzymes was inter-mediate between those of both enzymes, indicating theabsence of inhibitory material for the enzyme at 45°C.Plasma membrane fractions prepared from wild-type andcyrl-2(ts) strains were preincubated at 45°C, and the adenyl-ate cyclase activity of these preparations was measured. Asshown in Fig. 2, adenylate cyclase of cyrl-2(ts) mutant wasmore thermolabile than that of wild-type strain. To excludethe possibility that the alteration of plasma membrane causesthe thermal instability of mutant enzyme, adenylate cyclasewas solubilized by Lubrol from the plasma membrane frac-tion, and its thermal inactivation was examined. About 70%of adenylate cyclase activity in plasma membrane fractionswas solubilized under these conditions. The enzyme in theLubrol-soluble plasma membrane fraction of the wild-typestrain was stable, but that of cyrl-2(ts) mutant was thermola-bile, as observed for those of plasma membrane fractions.

Isolation of mutants bearing a nonsense cyri mutation. Of57 mutants isolated as described above, each of 55 tempera-ture-insensitive mutants was crossed to the wild-type strain(G211-2A), and the resulting diploids were tested for theircAMP requirement. Most diploids thus constructed (53 of55) grew on YPGlu medium, indicating that these mutationsare recessive to their wild-type counterparts.Complementation tests between 53 recessive mutants and

the cyrl-l (AM7-11D) mutant showed that all these mutantswere allelic and were classified as cyri mutants. The 53 cyri

mutants thus identified were tested for susceptibility to adominant ochre (UAA) suppressor, SUP3, as follows. Thecyrl-l SUP3 strain (AM76-2C) obtained from a cross be-tween the cyrl-J strain (AM18-SC) and the SUP3 strain(G185-7D) required cAMP for growth, indicating that SUP3did not suppress cyrl-l. The cyrl-J SUP3 strain was crossedwith each of the 53 cyrl mutants, and the resulting diploidswere tested for growth on YPGlu medium. Three suchdiploids were able to grow on YPGlu medium. The resultindicated that the cyri alleles involved in these diploids areochre nonsense mutations; they were designated cyrl-3.One of these nonsense mutants (AM3-4BM-31) was

crossed with the wild-type strain (G211-2A). The diploidobtained was sporulated, and four-spored asci were dissect-ed. Tetrad analysis of the diploid showed a 2+:2- segrega-tion on YPGlu medium at 30°C in the 21 asci tested. This factindicated that the nonsense mutant bears a single chromo-somal mutation.

TABLE 2. Adenylate cyclase activity and cAMP level in crude extracts of wild-type and mutant strains

AdenyatecclasecAMPStrain Genotypea Growth conditions Adenylate cyclase (pmol/mg ofprotein)

AM26-2B a YPGlu at 25°C 5.3 2.1YPGlu/cAMP at 35°C 6.6 NTb

AM26-2A a cyrl-2(ts) YPGlu at 25°C 0.3 0.4YPGlu/cAMP at 35°C <0.1 NT

AM81-1C a cyrl-2(ts) bcyl YPGlu at 25°C 0.3 0.3YPGlu at 35'C <0.1 <0.1

AM86-1C (x YPGlu at 30'C 7.1 3.1AM85-9A a cyrl-3 YPGlu/cAMP at 30'C <0.1 NTAM86-1D a cyrl-3 SUP3 YPGlu at 30°C 1.4 1.5AM89-2B a cyrl-3 bcyl YPGlu at 30'C <0.1 <0.1AM91-lBR-1 a cyrl-3 SUP-o YPGlu at 30°C 2.4 1.7

a Genotypes for the auxotrophic markers were omitted.b NT, Not tested.

VOL. 157, 1984

on May 29, 2021 by guest

http://jb.asm.org/

Dow

nloaded from

280 MATSUMOTO, UNO, AND ISHIKAWA

TABLE 3. Effects of divalent cations on adenylate cyclase activity of crude extracts and plasma membrane fractions from wild-type andcyri mutant strains

Enzyme sourcea Addition Mn2+-dependent Mg2+-dependentenzyme activityb enzyme activityb

Wild-type None 4.7 0.5crude extract GTP (10 p.M) 8.2 0.7

Gpp(NH)p (10 ,uM) 9.0 0.8NaF (10 mM) 4.7 0.4CaCl2 (0.1 mM) 4.6 0.4

Wild-type None 6.4 0.6membrane fraction GTP (10 F.M) 17.2 1.2

Gpp(NH)p (10 ,uM) 15.1 1.4NaF (10 mM) 6.2 0.6CaCl2 (0.1 mM) 6.4 0.7

cyrl-l None <0.1 <0.1crude extratt GTP (10 ,uM) <0.1 <0.1

Gpp(NH)p (10 ,LM) <0.1 <0.1NaF (10 mM) <0.1 <0.1CaCl2 (0.1 mM) <0.1 <0.1

cyrl-l None <0.1 <0.1membrane fraction GTP (10 ,uM) <0.1 <0.1

Gpp(NH)p (10 p.M) <0.1 <0.1NaF (10 mM) <0.1 <0.1CaCl2 (0.1 mM) <0.1 <0.1

a Crude extracts and membrane fractions prepared from wild-type (P-28-24C) and cyri-J mutant (AM18-5C) cells were used.b Mn2 -dependent enzyme activity (units per milligram of protein) was assayed in the presence of 2.5 mM MnCI2 under the standard

conditions except for the additives indicated, whereas Mg2+-dependent enzyme activity (units per milligram of protein) was assayed in thepresence of 2.5 mM MgCl2 under the same conditions.

Tetrad analysis of a diploid constructed by a cross be-tween one of the cyrl-3 segregants (AM85-9A) from theabove cross and the SUP3 strain (G185-7D) showed 2+:2-,3+ :1-, and 4+ :0- segregations for growth on YPGlu medi-um in a ratio of 5:4:3. The result indicated that the cyrl-3mutation was susceptible to the SUP3 suppressor. Twosegregants from this cross, AM86-1D (cyrl-3 SUP3) andAM86-1C (wild type), were used for further analysis. When adiploid constructed by a cross between a cyrl-3 (AM88-1A)and a bcyl strain (AM87-1C) was subjected to tetrad analy-sis, three ascus types showing 2+:2-, 3+:1-, and 4+:0-segregations on YPGlu medium appeared in 4, 12, and 4 asci,respectively, indicating that the cyrl-3 mutation was sup-pressed by the bcyl mutation. One of segregants from thiscross, AM89-2B (cyrl-3 bcyl), was used for further analy-ses.The levels of adenylate cyclase activity and cAMP

were compared among the wild type (AM86-1C), cyrl-3(AM85-9A), cyrl-3 SUP3 (AM86-1D), and cyrl-3 bcyl(AM89-2B) (Table 2). The cyrl-3 mutant cells grown inYPGlu/cAMP medium at 30°C had no detectable level ofadenylate cyclase activity. The cyrl-3 SUP3 mutant cellsgrown in YPGlu medium at 30°C produced significantamounts of adenylate cyclase activity and cAMP, but thespecific activity of adenylate cyclase found in mutant cellswas about 29% of that in wild-type cells. On the other hand,the cyrl-3 bcyl mutant cells were able to grow in YPGlumedium at 30QC but produced no detectable amounts ofadenylate cyclase and cAMP.To obtain conclusive evidence that the cyrl-3 allele in-

volves an ochre mutation, we conducted reversion experi-ments with strain AM91-1B (a cyrl-3 his5-2 lysJ-1 trpS48)in which all three auxotrophic markers are known to beUAA nonsense mutations (3). Cells of AM91-1B grown in

100l

>. 50-Wua- 20.0-

.C10c4-C

'- 5a.

2

10 5 10 15 20

Incubation time (min)FIG. 2. Thermal inactivation of adenylate cyclase in plasma

membrane and Lubrol-soluble plasma membrane fractions obtainedfrom cyrl-2(ts) mutant and wild-type strains. Plasma membrane andLubrol-soluble fractions were prepared from cyrl-2(ts) and wild-type cells grown in YPGlu medium at 25°C and used as enzymepreparations. Each enzyme preparation was kept at 45°C for theindicated period and then immediately cooled in an ice bath, and theremaining enzyme activity was assayed at 30°C. Symbols: 0,plasma membrane fraction from the wild type (AM26-2B); O,plasma membrane fraction from a cyrl-2(ts) mutant (AM26-2A); *,Lubrol-soluble plasma membrane fraction from the wild type; and*, Lubrol-soluble plasma membrane fraction from cyrl-2(ts).

J. BACTERIOL.

on May 29, 2021 by guest

http://jb.asm.org/

Dow

nloaded from

STRUCTURAL GENE FOR YEAST ADENYLATE CYCLASE 281

TABLE 4. Tetrad data indicating that cyrl-3 is a nonsense mutationNumber of tetrads

Cross' Cvr+ His' LYS+ TIp+ Phenotype of Phenotype ofCyr- His- Lys- Trp Cyr' spore clone Cyr- spore clone

4 3 2 4 3 2 4 3 2 4 3 2 His Lys Trp His Ls Trp0 12 012 012 0 1 2 +:- +:- +:- +:- +:- +:-

AM91-1BR-1 x FP-1A 5 9 1 1 6 8 0 8 7 0 11 4 38:11 38:11 41:8 0:11 0:11 0:11AM91-1BR-1 x AM91-4B 0 0 7 0 0 7 0 0 7 0 0 7 14: 0 14: 0 14:0 0:14 0:14 0:14

a Genotypes of strains used: AM91-1BR-1, a cyrl-3 hisS-2 lysJ-l trp548 SUP-o; FP-1A, a wild type; AM91-4B, a cyrl-3 hisS-2 lysJ-l trp5-48.

YPGlu/cAMP medium were harvested, washed, and spreadonto minimal medium and minimal medium supplementedwith appropriate amounts of histidine and lysine (mediumA), histidine and tryptophan (medium B), or lysine andtryptophan (mnedium C). After these plates were incubated at30°C for 6 to 8 days, 44 colonies (16 from minimal medium, 6from medium A, 18 from medium B, and 4 from medium C)were selected. It was confirmed that all of the isolated clonescould grow on medium A, B, or C and also on minimalmedium. The mutation was further characterized in one ofthe isolates, AM91-1BR-1. It was crossed with a wild-typestrain (FP-1A), diploids were sporulated, and four-sporedasci were dissected. Two or more spore clones showed thepositive phenotypes for all of the markers in all of the ascitested. The two spore clones always showed the positivephenotypes for all of the traits in each ascus (Table 4). Thesefacts strongly suggest that the reversion was caused by asuppressor mutation. When AM91-1BR-1 was crossed withAM91-4B (rxcyrl-3 hisS-2 lysJ-l trpS-48), the diploid segre-gated two spore clones showing the positive phenotypes,whereas the other two spore clones showed the negativephenotypes for all of the traits in all asci tested (Table 4).These results clearly indicate that the suppressor (SUP-o) iseffective against all of the markers, including cyrl-3. Sincethe diploid is heterozygous for the suppressor and couldgrow on minimal medium, the suppressor is dominant overthe wild-type counterpart. Since the SUP-o suppressor iseffective against the UAA nonsense mutations, hisS-2,Iysl-J, and trp548, the cyrl-3 aliele is also a UAA nonsensemutation. The levels of adenylate cyclase activity and cAMPobserved in a cyrl-3 SUP-o strain (AM91-lBR-1) were lowerthan those of the wild type, as observed in cyrl-3 SUP3(Table 2).

DISCUSSIONThe isolation and characterization of mutations affecting a

structural gene play a vital role in defining the function of theenzyme responsible for the complex physiological activity.A direct relationship betweeti gene dosage and amount of thecorresponding enzyme activity is generally assumed to exist,as indicated in the structural gene loci of yeasts for galacto-kinase (13), tryptophanase (1), proteinase B (21), and alka-line phosphatase (6). We infer that cyri is the structural genefor adenylate cyclase of yeasts because (i) specific activitiesof adenylate cyclase in extracts of various tetraploids show adosage effect, and (ii) temperature-sensitive cyri mutationcauses thermolabile adenylate cyclase activity that is distin-guishable from that of the wild type. Successful isolation ofnonsense mutants at the cyri locus clearly indicates that theproduct of the cyri gene must be a protein. The observationof dosage effect and production of thermolabile enzyme atthe cyri gene locus imply that the cyri mutations do notaffect a regulatory protein for adenylate cyclase synthesis or

a posttranslational modification enzyme which may activateadenylate cyclase.

It could be argued that the adenylate cyclase defect in thecyri mutant cells would be a reflection of an altered mem-brane structure. This is what has been found for the frostmutation (fr) of Neurospora crassa affecting adenylate cy-clase activity (18). In fr cultures, the mutant enzyme wasthermolabile, and supplementation of agar medium withpolysaturated fatty acids such as linolenic acid or phosphodi-esterase inhibitors such as theophylline resulted in an elevat-ed cAMP content and a wild-type-like morphology. It wassuggested that the cAMP deficiency in thefr mutant resultedfrom a membrane defect that affected adenylate cyclasefunction. We consider this unlikely in the cyri mutants ofyeasts, because adenylate cyclase activity found in theenzyme sample obtained from the plasma membrane fractionof the temperature-sensitive cyri mutant by Lubrol treat-ment was still temperature sensitive. In addition, we con-firmed that the cyri mutants of yeasts did not respond toexogenous polysaturated fatty acids or phosphodiesteraseinhibitors (data not shown).Growth of cyri mutant cells was arrested at the Gl phase

of the yeast cell cycle in the absence of cAMP, and it hasbeen indicated that cAMP is an essential factor for yeastcells to proceed through the cell cycle via the activation ofprotein kinase (11). On the other hand, the cyri mutationspermitted the initiation of meiosis, indicating that the pro-duction of cAMP is inhibitory for the initiation of meiosis(lOa). The successful isolation of nonsense mutations at thestructural gene for adenylate cyclase ensures that all thesephenotypes of cryl mutants really resulted from the loss ofadenylate cyclase. Of the 53 cyri mutations we isolated, 6%were ochre mutations. This frequency is comparable tofrequencies found for other structural gene loci of yeasts (5,22).Ross et al. (17) found that the HC-1 hepatoma cells

derived from the mouse HTC line was a clonal line that wasdevoid of assayable catalytic adenylate cyclase activity, andAC- variants of S49 cells were deficient in the regulatorycomponent of the enzyme. Membrane fractions of AC-mutants contain only as Mn2+-dependent adenylate cyclaseactivity that was stimulated by GTP or Gpp(NH)p. Thepresent results indicate that the cyri enzyme in yeasts, wasdeficient in adenylate cyclase activity even with Mn2+ andGTP or Gpp(NH)p, indicating that the cyrl mutant may havedefective catalytic protein of adenylate cyclase. Furtherwork is required to show a possible existence of subunits ofadenylate cyclase in yeasts as has been shown in mammaliancells.

ACKNOWLEDGMENTSWe thank A. Toh-e for valuable discussions and J. Ishiyama and

H. Mori (Central Research Laboratory of Kikkoman Co., Noda,

VOL. 157, 1984

on May 29, 2021 by guest

http://jb.asm.org/

Dow

nloaded from

282 MATSUMOTO, UNO, AND ISHIKAWA

Japan) for their gift of cAMP. We also thank S. Swamy for readingthe manuscript and K. Adachi for preparing the illustrations.

LITERATURE CITED

1. Ciferri, O., S. Sora, and 0. Tiboni. 1969. Effect of gene dosageon tryptophan synthetase activity in Saccharomyces cerevisiae.Genetics 61:567-576.

2. Duran, A., B. Bowers, and E. Cabib. 1975. Chitin synthetasezymogen is attached to the yeast plasma membrane. Proc. Natl.Acad. Sci. U.S.A. 72:3952-3955.

3. Hawthorne, 0. C., and U. Leupold. 1974. Suppressor mutationsin yeast. Curr. Top. Microbiol. Immunol. 64:1-47.

4. Jaynes, P. K., J. P. McDonough, and H. R. Mahler. 1982.Properties and possible functions of the adenylate cyclase inplasma membranes of Saccharomyces cerevisiae. Mol. Cell.Biol. 2:1481-1491.

5. Jones, E. W. 1972. Nonsense mutations in the ade3 locus ofSaccharomyces cerevisiae. Genetics 71:217-232.

6. Kaneko, N., A. Toh-e, and Y. Oshima. 1982. Identification of thegenetic locus for the structural gene and a new regulatory genefor the synthesis of repressible alkaline phosphatase in Saccha-romyces cerevisiae. Mol. Cell. Biol. 2:127-137.

7. Liao, H., and J. Thorner. 1980. Yeast mating pheromone afactor inhibits adenylate cyclase. Proc. Natl. Acad. Sci.U.S.A. 77:1898-1902.

8. Lowry, 0. H., N. J. Rosebrough, A. L. Farr, and R. J. Randall.1951. Protein measurement with the Folin phenol reagent. J.Biol. Chem. 193:265-275.

9. Matsumoto, K., Y. Adachi, A. Toh-e, and Y. Oshima. 1980.Function of positive regulatory gene gaI4 in the synthesis ofgalactose pathway enzymes in Saccharomyces cerevisiae: evi-dence that the GAL81 region codes for part of the gaI4 protein.J. Bacteriol. 141:508-527.

10. Matsumoto, K., A. Toh-e, and Y. Oshima. 1978. Genetic controlof galactokinase synthesis in Saccharomyces cerevisiae: evi-dence for constitutive expression of the positive regulatory genegaI4. J. Bacteriol. 134:446-457.

10a.Matsumoto, K., I. Uno, and T. Ishikawa. 1983. Initiation ofmeiosis in yeast mutants defective in adenylate cyclase andcyclic AMP-dependent protein kinase. Cell 32:417-423.

11. Matsumoto, K., I. Uno, Y. Oshima, and T. Ishikawa. 1982.Isolation and characterization of yeast mutants deficient inadenylate cyclase and cyclic AMP-dependent protein kinase.Proc. Natl. Acad. Sci. U.S.A. 79:2355-2359.

12. Matsumoto, K., I. Uno, A. Toh-e, T. Ishikawa, and Y. Oshima.1982. Cyclic AMP may not be involved in catabolite repressionin Saccharomyces cerevisiae: evidence from mutants capable ofutilizing it as an adenine source. J. Bacteriol. 150:277-285.

13. Nelson, N. M., and H. C. Douglas. 1963. Gene dosage andgalactose utilization by Saccharomyces tetraploids. Genetics48:1585-1591.

14. Nogi, Y., K. Matsumoto, A. Toh-e, and Y. Oshima. 1977.Interaction of super-repressible and dominant constitutive mu-tations for the synthesis of galactose pathway enzymes inSaccharomyces cerevisiae. Mol. Gen. Genet. 152:137-144.

15. Robison, G. A., R. W. Butcher, and E. W. Sutherland. 1971.Cyclic AMP. Academic Press, Inc., New York.

16. Ross, E. M., and A. G. Gilman. 1980. Biochemical properties ofhormone-sensitive adenylate cyclase. Annu. Rev. Biochem.49:533-564.

17. Ross, E. M., A. C. Howlett, K. M. Ferguson, and A. G. Gilman.1978. Resonstitution of hormone-sensitive adenylate cyclaseactivity with resolved components of the enzyme. J. Biol.Chem. 253:6401-6412.

18. Scott, W. A. 1976. Adenosine 3':5'-cyclic monophosphate defi-ciency in Neurospora crassa. Proc. Natl. Acad. Sci. U.S.A.73:2995-2999.

19. Uno, I., K. Matsumoto, and T. Ishikawa. 1983. Characterizationof cyclic AMP-requiring yeast mutants altered in the regulatorysubunit of protein kinase. J. Biol. Chem. 257:14110-14115.

20. Uno, I., H. Nyunoya, and T. Ishikawa. 1981. Effects of 2-deoxy-D-glucose and quinidine on the fruiting body formation inCoprinus macrorhizus. J. Gen. Appl. Microbiol. 27:219-228.

21. Zubenko, G. S., A. P. Mitchell, and E. W. Jones. 1979. Septumformnation, cell division, and sporulation in mutants of yeastdeficient in proteinase B. Proc. Natl. Acad. Sci. U.S.A.76:2395-2399.

22. Zubenko, G. S., A. P. Mitchell, and E. W. Jones. 1980. Mappingof the proteinase B structural gene PRBJ, in Saccharomycescerevisiae and identification of nonsense alleles within thelocus. Genetics 96:137-146.

J. BACTERIOL.

on May 29, 2021 by guest

http://jb.asm.org/

Dow

nloaded from