Vol. 10, No. 6
MOLECULAR AND CELLULAR BIOLOGY, June 1990, p. 2966-2972
Mutations in Cell Division Cycle Genes CDC36 and CDC39 Activate the Saccharomyces cerevisiae Mating Pheromone Response Pathway MIGUEL DE BARROS LOPES, JEONG-YAU HO,+ AND STEVEN
Department of Molecular Biology, Research Institute of Scripps Clinic, 10666 North Torrey Pines Road, La Jolla, California 92037 Received 20 November 1989/Accepted 22 March 1990
Conditional mutations in the genes CDC36 and CDC39 cause arrest in the Gl phase of the Saccharomyces cerevisiae cell cycle at the restrictive temperature. We present evidence that this arrest is a consequence of a mutational activation of the mating pheromone response. cdc36 and cdc39 mutants expressed pheromoneinducible genes in the absence of pheromone and conjugated in the absence of a mating pheromone receptor. On the other hand, cells lacking the G. subunit or overproducing the Ga, subunit of the transducing G protein that couples the receptor to the pheromone response pathway prevented constitutive activation of the pathway in cdc36 and cdc39 mutants. These epistasis relationships imply that the CDC36 and CDC39 gene products act at the level of the transducing G protein. The CDC36 and CDC39 gene products have a role in cellular processes other than the mating pheromone response. A mating-type heterozygous diploid cell, homozygous for either the cdc36 or cdc39 mutation, does not exhibit the Gl arrest phenotype but arrests asynchronously with respect to the cell cycle. A similar asynchronous arrest was observed in cdc36 and cdc39 cells where the pheromone response pathway had been inactivated by mutations in the transducing G protein. Furthermore, cdc36 and cdc39 mutants, when grown on carbon catabolite-derepressing medium, did not arrest in Gl and did not induce pheromone-specific genes at the restrictive temperature.
Saccharomyces cerevisiae exists in two allelic haploid mating types, a and at, which conjugate to produce a matingtype heterozygous diploid cell. The mating response is initiated by peptide pheromones, a-factor and a-factor, that are secreted from cells of one mating type and bind to surface receptors on cells of the opposite mating type. This interaction leads to a number of physiological responses which promote mating. Conjugating partners arrest in Gi of their respective cell cycles, exhibit a chemotropic response and orient toward cells of the opposite mating type, and induce transcription from a number of mating-specific genes (reviewed in references 7 and 41). Genetic and molecular analysis of the mating pheromone response suggest that signaling is initiated by G-proteincoupled receptors. The sequences of the receptor genes (STE2 and STE3) predict a topological structure for the encoded products similar to that of G-protein-coupled receptors in vertebrate cells, including seven potential transmembrane domains and a hydrophilic carboxyl terminus (2, 12, 30). Experimental evidence suggests the involvement of a single heterotrimeric G protein in transducing the pheromone signal in both mating types. Sequence analysis of the gene GPAI (SCGI) indicates that it codes for a Ga-like protein with 48% amino acid identity with the at subunits of mammalian G proteins (8, 15, 26). Loss of GPAI function confers a phenotype suggestive of a constitutive mating pheromone response (8, 15, 25). STE4 and STE18 encode homologs of the G, and Gy subunits of mammalian G proteins (49). Mutations in these genes result in an inability to respond to mating pheromone (6, 49). On the basis of these results and the paradigm developed from work with G proteins in mammalian cells (reviewed in references 11 and 43), it has been proposed that pheromone-receptor interac-
tion promotes guanyl nucleotide exchange on the Ga subunit (Gpal) of the G-protein heterotrimer. This, in turn, leads to dissociation of the ,-y moiety (Ste4 and Stel8), which is then able to interact with an as yet unknown effector to generate a response (8, 15, 49). Mutational analysis has identified other elements that are essential for conjugation (13, 19). Mutations in STE5, STE7, STEJJ, and STE12 are epistatic to mutations in the G protein (1, 29). The STE12 gene product has recently been shown to bind to DNA at pheromone response elements found in the promoter regions of inducible genes, suggesting that it may mediate the expression of pheromone-responsive genes (9, 10). STE7 and STEJ I encode protein kinases, and it has been proposed that these kinases may modify the Stel2 transcription complex and affect the levels of pheromone-specific transcripts (10, 44). The function of the STE5 gene product is unknown. Epistasis studies indicate that Ste5 is also downstream of Gpal (29) and Ste4 (1). The CDC36 and CDC39 genes were originally identified in a screen for conditional cell division mutants which arrest in Gl (32). cdc36 and cdc39 mutant strains are able to conjugate when arrested by a shift to the restrictive temperature. Furthermore, cdc36 and cdc39 mutants continue to enlarge and form asymmetric projections, morphologically resembling cells responding to mating pheromone. The CDC36 and CDC39 genes have been isolated and sequenced (31; unpublished data). The CDC36 gene has significant homology to another yeast cell division cycle gene, CDC4, and to the transformation-specific sequence of avian erythroblastosis virus E26, ets (31). In this paper, we demonstrate that cdc36 and cdc39 mutants constitutively activate the pheromone response pathway. Epistasis analysis with ste2, ste4, and stel8 null mutations and with GPAI overexpression indicate that the CDC36 and CDC39 gene products act downstream from the mating pheromone receptor and upstream from, or at the same level as, the transducing G protein.
* Corresponding author. t Present address: Department of Nematology, University of California, Davis, Davis, CA 95616.
CDC36 AND CDC39 IN S. CEREVISIAE PHEROMONE RESPONSE
VOL. 10, 1990
MATERIALS AND METHODS Media and strains. Yeast and bacterial culture media have been described previously (20, 38). For growth in carbon catabolite-derepressing medium, glucose was replaced by 2% ethanol and 2% glycerol. Propagation of recombinant plasmids took place in Escherichia coli JA226 (C600 hsdR hsdM leuB6 lopli thi recBC Str). Most of the yeast strains used in this study are congenic with strain 381G (MATa ade2-1 cryl his4-580 lys2 tyri trpl) (13). The cdc36 strains used for measuring FUSI expression were obtained from a cross between strains 381G and 15Du (MATa adel his2 leu2-3,112 trpl ura3A), which is congenic to strain BF264-15D (35). FUSI expression is essentially the same in strains 381G and 15Du when induced by either pheromone or cdc36 mutation. Mating efficiencies for the cdc36 ste double mutants are also essentially the same in both strain backgrounds. The cdc36 allele used in these studies was cdc36-16 (32). The cdc39 allele was cdc39-1 (32). The ste2, ste4, and stel8 null mutations have the LEU2 gene inserted into their coding sequences. ste2: :LEU2 was crossed into the cdc36 and cdc39 mutants with strain 3734-5-4 (381G MATct ste2::LEU2 leu2 ura3-52) (kindly provided by Anne Burkholder). Disruption of the STE4 and STE18 genes was performed by one-step gene replacement (6). Overexpression of GPAI was accomplished by transforming yeast with plasmid YEP13[GPAI]. DNA manipulations and transformation methods. Bacterial transformations, plasmid DNA preparations, and restriction digests were performed by standard procedures (20). Yeast transformations were performed by the alkali cation method (14). One-step gene replacement was performed as described by Rothstein (37). Yeast RNA analysis. Exponentially growing cultures were divided and then shifted to 38°C for 3 h or allowed to continue growth at room temperature. Cells were then harvested, and total RNA was prepared by methods described previously (33). Formaldehyde-agarose gels were run essentially as described by Maniatis et al. (20). Two micrograms of total yeast RNA was loaded per lane. The RNA was electrotransferred (4) to a Biotrans nylon membrane (ICN Pharmaceuticals Inc., Irvine, Calif.) in 25 mM sodium phosphate buffer (pH 6.5) for 2 h at 0.6 mA. After transfer, the wet nylon membrane was exposed to shortwavelength UV light for 10 min. At this point, the membrane was stained in 0.04% methylene blue-0.5 M sodium acetate to check for equal loading and transfer of RNA. A 0.6kilobase HindIlI fragment internal to the STE4 gene and a 1.7-kilobase EcoRI-BamHI fragment containing the GPAI sequence were radioactively labeled by random hexanucleotide priming (Boehringer Mannheim Biochemicals, Indianapolis, Ind.). Autoradiography of hybridized membranes was performed on XAR-5 film (Eastman Kodak Co., Rochester, N.Y.) for 1 to 3 days. Autoradiographs were quantified by using an Ultrascan XL laser densitometer (LKB Instruments, Inc., Rockville, Md.). j3-Galactosidase assays. ,-Galactosidase assays were done as described previously (24). Cells containing plasmid pSB231 (FUSI-lacZ) (47) were grown to an optical density at 600 nm (OD6.) of 0.5 in selective medium. Cultures were divided and then shifted to 38°C for 3 h, treated with 2 ,ug of a-factor per ml for 3 h, or allowed to continue growth at 23°C. Cells were then harvested and suspended in 0.8 ml of Z buffer (0.06 M Na2HPO4, 0.06 M NaH2PO4, 0.01 M KCI, 0.001 M MgSO4 [pH 7]) plus 20 ,ul of 0.1% sodium dodecyl
sulfate and 60 [li of chloroform. The cell suspension was incubated at 28°C for 10 min, and then 160 ,ul of o-nitrophenyl-p-D-galactoside (ONPG) was added to initiate a colorimetric assay. At appropriate times, 400 ,ul of 1 M Na2CO3 was added to stop the reaction, and the OD420 was measured. Units of ,-galactosidase activity are defined by the formula (1,000 x OD420)/(OD6w of culture x volume of culture x minutes of assay). Quantitative mating assays. Mating assays were performed essentially as described by Reid and Hartwell (36), except that conjugation mixtures consisted of 106 cells to be tested and 107 tester cells of strain 262a (MATa- thrl). Elevated temperature preincubations were done at 36°C for 2 h. Cells were collected on a nitrocellulose filter and then incubated on yeast extract-peptone-dextrose (YEPD) plates for 4 h at either 23 or 36°C. Mating efficiency is expressed as a percentage of wild-type levels. Microscopy. All micrographs were taken with a Zeiss Axiophot photomicroscope with differential interference contrast optics by using a 100x objective and Kodak TriX-pan 400 film.
RESULTS cdc36 and cdc39 mutations induce pheromone-specific transcripts. The pheromone response leads to induction of a number of genes, many of which are required for conjugation (42). One such gene, FUS1, is required for cell fusion during zygote formation (21, 47). Since the transcript level of FUSI increases rapidly as a direct consequence of pheromonereceptor binding, the induction of this gene can serve to monitor the mating pheromone response. The FUS1 gene has been linked in frame to the E. coli lacZ gene producing a reporter construct and thus allowing assay of FUSI induction by measuring 3-galactosidase activity (47). Since the morphology of cdc36 and cdc39 mutants resembles that of cells preparing for conjugation, we sought to determine whether the mutant phenotype in fact results from activation of the mating pheromone response by analyzing the expression of FUS1 in these mutants. A centromeric plasmid, pSB231 (kindly provided by G. Fink), containing the FUSJlacZ fusion was transformed into wild-type cells and cdc36 and cdc39 temperature-sensitive mutants. P-Galactosidase activity was then assayed under a variety of conditions. In cdc36 mutant cells, FUSI transcription, as estimated by 3-galactosidase activity, was 10-fold higher than in wild-type cells at the permissive temperature (23°C), indicating that the mating response pathway is activated constitutively in these cells in the absence of pheromone-receptor interaction (Table 1). When the mutant cells were raised to the restrictive temperature (38°C), FUSI expression was further induced. This level of FUSI expression was similar to the maximum induction observed when wild-type cells were treated with saturating amounts of a-factor. These results demonstrate that the cdc36 mutant gene product is partially inactivated at 23°C, even though the growth rate of cells carrying that mutation is close to that of wild-type levels. Shifting the cells to the restrictive temperature led to further activation of the pathway, as illustrated by increased FUSI transcription and cell cycle arrest. Similar results were obtained when the mRNA levels of a second pheromone-inducible transcript, sigma (48), were analyzed directly by RNA blot analysis (data not shown). cdc39 mutant cells also expressed FUS1 in the absence of pheromone. At the permissive temperature, however, there was no FUS1 induction, but when the cells were shifted to
MOL. CELL. BIOL.
DE BARROS LOPES ET AL.
TABLE 1. Induction of FUSI-lacZ in cdc36 and cdc39 mutants 3-Galactosidase activity (U) Relevant genotype'
CDC+ STE+ cdc36 STE' cdc39 STE+ CDC+ ste2A cdc36 ste2A cdc39 ste2A CDC+ ste4A cdc36 ste4A cdc39 ste4A CDC+ stel8A cdc36 stel8A cdc361cdc36 MA Ta/MA Ta
0.8 11.5 0.6 0.4 9.3 0.3