Isolation of a novel gene from Schizosaccharomyces pombe: stm1+ encoding a seven-transmembrane loop protein that may couple with the heterotrimeric Galpha 2 protein, Gpa2.

A putative seven transmembrane protein gene, stm1(+), which is required for proper recognition of nitrogen starvation signals, was isolated as a multicopy suppressor of a ras1 synthetic lethal mutant in Schizosaccharomyces pombe. Under nitrogen-deficient conditions, transcription of the stm1 gene was induced; deletion of stm1 was associated with early entry into G(1) arrest. Under nutritionally sufficient conditions, overexpression of Stm1 inhibited vegetative cell growth, resulted in decreased intracellular cAMP levels, increased the expression of the meiosis-specific genes ste11, mei2, and mam2, and facilitated sexual development in homothallic cells. However inhibition of vegetative cell growth and reduction of cAMP levels were not observed in a deletion mutant of the heterotrimeric G protein Galpha2 gene, gpa2, that is responsible for regulating intracellular cAMP levels, a key factor in determining the sexual development in S. pombe. Stm1 protein was shown to interact with Gpa2 through its C-terminal transmembrane domains 5-7. Mutation at Lys(199) in the C-terminal domain (stm1(K199A)) abolished the Stm1 overexpression effect on lowering cAMP levels. Induction of ste11, a meiosis-specific gene transcription factor, by Stm1 overexpression was enhanced in gpa2-deleted cells but was absent in a deletion mutant of sty1, a key protein kinase that links mitotic control with environmental signals and induces stress-responsive genes. Moreover, deletion of both stm1 and ras1 caused delayed entry into G(1) arrest in S. pombe when the cells were grown in a nitrogen-deficient medium. Thus we consider that the stm1 gene can function through Gpa2-dependent and/or -independent pathways and may play a role in providing the prerequisite state for entering the pheromone-dependent differentiation cycle in which heterotrimeric Galpha1 protein, Gpa1, and Ras1 play major roles. Stm1 could function as a sentinel molecule sensing the nutritional state of the cells, stopping the proliferative cell cycle, and preparing the cell to enter meiosis under nutritionally deficient conditions.

A putative seven transmembrane protein gene, stm1 ؉ , which is required for proper recognition of nitrogen starvation signals, was isolated as a multicopy suppressor of a ras1 synthetic lethal mutant in Schizosaccharomyces pombe. Under nitrogen-deficient conditions, transcription of the stm1 gene was induced; deletion of stm1 was associated with early entry into G 1 arrest. Under nutritionally sufficient conditions, overexpression of Stm1 inhibited vegetative cell growth, resulted in decreased intracellular cAMP levels, increased the expression of the meiosis-specific genes ste11, mei2, and mam2, and facilitated sexual development in homothallic cells. However inhibition of vegetative cell growth and reduction of cAMP levels were not observed in a deletion mutant of the heterotrimeric G protein G␣2 gene, gpa2, that is responsible for regulating intracellular cAMP levels, a key factor in determining the sexual development in S. pombe. Stm1 protein was shown to interact with Gpa2 through its Cterminal transmembrane domains 5-7. Mutation at Lys 199 in the C-terminal domain (stm1 K199A ) abolished the Stm1 overexpression effect on lowering cAMP levels. Induction of ste11, a meiosis-specific gene transcription factor, by Stm1 overexpression was enhanced in gpa2-deleted cells but was absent in a deletion mutant of sty1, a key protein kinase that links mitotic control with environmental signals and induces stress-responsive genes. Moreover, deletion of both stm1 and ras1 caused delayed entry into G 1 arrest in S. pombe when the cells were grown in a nitrogen-deficient medium. Thus we consider that the stm1 gene can function through Gpa2-dependent and/or -independent pathways and may play a role in providing the prerequisite state for entering the pheromone-dependent differentiation cycle in which heterotrimeric G␣1 protein, Gpa1, and Ras1 play major roles. Stm1 could function as a sentinel molecule sensing the nutritional state of the cells, stopping the proliferative cell cycle, and preparing the cell to enter meiosis under nutritionally deficient conditions.
Many of the cell membrane G protein-coupled hormone re-ceptors relay signals to the interior of the cell using several different classes of heterotrimeric G proteins: G s , G i , G o , G q , G 12 , and G 13 . These heterotrimeric G protein molecules specify intracellular signal pathways by either activating or inactivating different effector systems such as adenylyl cyclase or phospholipase C, which in turn initiate signals via second messengers such as cAMP or calcium (1)(2)(3)(4)(5)(6)(7)(8). On the other hand small G proteins, including Ras, modulate effector elements in response to external signaling, which can influence cell growth and differentiation (9 -11). Ras is capable of binding to different effector proteins such as Raf (12,13), phosphatidylinositol 3-kinase (14), and Ral (15,16). The interactions of different effectors with Ras can activate different downstream signaling pathways, including those that stimulate the cell division cycle, alter cell shape, or induce cellular differentiation (17). Thus signals emanating from the cell surface are delivered to diverse downstream effectors by switching G protein functions on and off. The integrated G protein signals are responsible for orchestrating a coherent specific biological response.
Fission yeast, Schizosaccharomyces pombe, possesses two genes encoding heterotrimeric G protein ␣-subunits, gpa1 and gpa2 (18,19), and one gene, gpb1, encoding a ␤-subunit (20). Gpa1 is responsible for pheromone-responsive sexual development, and Gpa2 relays the nutritional status information necessary to initiate the sexual differentiation of S. pombe. The G␤ subunit, Gpb1, functions as a negative factor in sexual development (20). Whereas only one ras gene, ras1 ϩ , has been identified in S. pombe, it is required for at least two distinct cellular functions: sexual differentiation, including conjugation and sporulation, and the control of cell morphology (21). Ras1 plays a major role in the initiation of meiotic differentiation when cells endure nutritionally unfavorable conditions (22). Unlike Saccharomyces cerevisiae, in which ras genes are essential for cell growth and modulate adenylate cyclase activity (16,(23)(24)(25)(26), S. pombe cells carrying a null mutation in the ras1 gene proliferate normally. These cells are deficient, however, in terms of sexual differentiation in that haploid cells are completely sterile and homozygous diploid cells have a very low frequency of sporulation. Mutated Ras1 does not affect the intracellular cAMP levels in S. pombe, indicating that Ras1 does not act primarily by modulating adenylate cyclase activity. Instead, when a mating pheromone binds to its serpentine receptor (Mam2 or Map3), Ras1 activates a mitogen-activated protein (MAP) 1 kinase module composed of Byr2, Byr1, and Spk1 protein kinases in concert with a heterotrimeric G protein ␣-subunit, Gpa1 (18,21,(27)(28)(29). This leads to the initiation of conjugation and sporulation. However, the activation of MAP kinases by Ras1 should be preceded by cell cycle arrest at G 1 . It has been observed that in S. pombe, when the nitrogen source is depleted, cell growth is arrested at G 1 first (30 -33), and then the heterotrimeric G protein ␣-subunit, Gpa2, which regulates adenylate cyclase in accordance with nutritional state of the cells (19), lowers the intracellular cAMP levels. This in turn triggers the induction of expression of genes such as mam2, ste11, and mei2, which are required in the initial stage of meiosis (22). Thus the decision to exit the cell cycle and initiate mating requires the integration of signal information concerning both the nutrient status and the presence of mating partners. The ras1 ϩ gene is considered to be involved in integrating these two signals by increasing the sensitivity to pheromone in the appropriate pathways when nutrients are in short supply (22). However, the link between the acceptance of the nutritional starvation signal and the activation of the differentiation cycle is poorly understood. In this report, we describe the isolation of the stm1 ϩ gene encoding a putative seven-loop transmembrane protein, which may function as a receptor that couples with the heterotrimeric G␣2 protein, Gpa2. Our studies indicate that Stm1 can interact with Gpa2 through the Cterminal domain. This interaction may provide the initiating signal that links signals triggered by nutritional deficiency and pheromone-dependent sexual differentiation, to allow Ras1 to activate sexual differentiation in S. pombe.

EXPERIMENTAL PROCEDURES
Strain Manipulation and Media-The S. pombe strains used in this study are listed in Table I. The newly isolated synthetic lethal mutant of ras1, KSC3, was derived from the KSC1 strain. S. pombe cells were grown in YEPD (0.5% yeast extract, 0.5% peptone, and 3% glucose), YES (0.5% yeast extract and 3% glucose with supplement), EMM (Edinburgh minimal medium) (34), EMM2 (EMM without supplements, CLONTECH), and ME (0.3% malt extract) media supplemented with adenine, uracil, and/or leucine as described previously (35,36). For detection of dead cells, phloxine B (Sigma) was added to the YES and EMM plates at a concentration of 20 mg/liter (37). Prior to analyzing the effects of nitrogen starvation on transcript levels of the genes of S. pombe, cells were grown in EMM overnight and resuspended at 10 7 cells/ml in EMM containing different nitrogen sources (0.5% NH 4 Cl, 0.5% proline, or no NH 4 Cl). For flow cytometry, cells were grown in EMM to log phase, harvested, washed, and then transferred to EMM devoid of a nitrogen source. Transformation of S. pombe cells was carried out using the lithium acetate method and standard yeast genetic techniques as described by Moreno et al. (38). Escherichia coli DH5␣ was used for subcloning.
Isolation of the stm1 Gene by Functional Complementation of a Synthetic Lethal Mutant of the ras1 Gene-To isolate a gene that may function in association with Ras1 in sexual differentiation of S. pombe, we used a mutant, KSC3, that exhibits a lethal phenotype when the Ras1 function is absent. This synthetic lethal mutant of Ras1 was isolated by UV mutagenesis from the strain KSC1, which contains the ras1 gene under the nmt1 promoter regulated by thiamine (nmt1-ras1 ϩ -ura4) (35,36). Approximately 2.5 ϫ 10 5 cells on EMM plates were treated with UV (dose: 254 nM, 115 V, 60 Hz, 0.16 A), which empirically is known to generate on average one point mutation/chromosome and leads to 20% survival of S. pombe cells. The mutagenized cells were incubated at 30°C for 6 days. The resulting 5ϫ10 4 colonies that survived were patched onto EMM plates and then replica-plated on EMM plates containing phloxine B (20 g/ml) with or without thiamine. KSC3 was selected from 50 candidate colonies that grew on EMM plates but did not grow on EMM containing thiamine. This mutant grew normally only when the ras1 gene under the nmt1 promoter was expressed in the absence of thiamine.
To clone a gene complementing this lethal phenotype, the mutant KSC3 strain was cultured in EMM in the absence of thiamine and was transformed with the genomic library constructed by Chung et al. (36). The transformed cells were plated onto EMM plates and incubated at 30°C for 4 days. The transformants grown in EMM were replica-plated onto EMM containing phloxine B and thiamine. Approximately 1ϫ10 5 transformants were screened, and 18 independent colonies that grew on EMM containing thiamine, where the Ras1 function was turned off, were selected. DNAs were recovered from these transformants, and the insert DNAs were subcloned and tested for complementation of the mutant phenotype of KSC3. The shortest insert DNA showing complementation was used for Northern blot and sequencing analysis.
Northern Blot Analysis-Total RNAs were prepared from S. pombe cells grown to log phase in different growth media as reported previously (20) and resolved on a 1% agarose gel containing 2.2 M formalde-fluorescent protein; kb, kilobase pair(s); PCR, polymerase chain reaction; MBP, maltose-binding protein; GST, glutathione S-transferase. hyde. RNAs were transferred to a nylon membrane and hybridized (39) with the probe DNAs radioactively labeled with a Redipriming kit (Amersham Pharmacia Biotech). Immunological Method and in Vivo Localization-To determine the location of Stm1 protein in cells, the HA (hemagglutinin A) tagging method (40) was used. An oligonucleotide encoding three copies of HA epitope sequences was synthesized and inserted at the 3Ј-end of the stm1 gene, which is under the nmt1 promoter sequence (pnmt1-stm1-HA). The cells containing the plasmid pnmt1-stm1-HA were grown at 2 ϫ 10 7 cells/ml in EMM in the absence of thiamine. 40 ml of the cells were harvested, washed with cold TE (10 mM Tris-HCl, pH 8.0, and 1 mM EDTA), and resuspended in 300 l of lysis buffer (20 mM Hepes, 150 mM NaCl, 1 mM dithiothreitol, 2 mM MgCl 2 , 1 mM phenylmethylsulfonyl fluoride, and 1 mg/ml each of aprotinin, leupeptin, and pepstatin). The cells were broken by a bead beater and centrifuged at 5,000 ϫ g for 3 min (41). The supernatant was centrifuged again at 100,000 ϫ g for 10 min, and this clarified supernatant was used as a cytosolic fraction. The pellet fraction was treated with 2% Triton X-100 and 1.5% Sarcosyl for 30 min at 4°C and centrifuged at 100,000 ϫ g for 10 min. The supernatant was used as a membrane fraction. Proteins (40 g from each fraction) were resolved on 12% SDS-polyacrylamide electrophoresis gels (PAGE), and electrotransferred to a nitrocellulose membrane (Schleicher & Schull). The proteins on the membrane were treated first with HA antibody (12CA5) according to the procedure described previously (42). The immune complexes retained on the membrane were visualized using anti-mouse IgG antibody as a secondary antibody and an alkaline phosphatase detection system (Promega). To determine the in situ location of Stm1, the coding region DNA of the green fluorescent protein gene (GFP) from the jellyfish Aequorea victoria (43,44) was fused in frame to the C terminus of the stm1 ϩ gene. The cells carrying the plasmid pnmt1-stm1-GFP were grown in minimal medium containing 25 nM thiamine, and Stm1-GFP fusion protein was detected by immunofluorescent microscopy.
Disruption and Overexpression of the stm1 ϩ Gene-The stm1 ϩ gene was disrupted using the one-step gene disruption method (38). The 0.9-kb HindIII-HpaI fragment containing the entire stm1 coding region DNA of genomic clone pWH031 ( Fig. 2A) was replaced with the ura4 ϩ gene. The resulting plasmid was treated with XbaI and HindIII, and the 3.2-kb XbaI-HindIII fragment so obtained was isolated and then used to replace one chromosomal copy of the stm1 ϩ gene in a diploid strain, SP286. After obtaining stable Ura4 ϩ transformants, genomic DNA was prepared from each transcript (38), and these were analyzed by Southern blotting with the probes of stm1 and ura4. The diploid strain containing one wild type and one disrupted copy of stm1 ϩ was selected. After sporulation, the resulting tetrads were analyzed, and the DNA content of the stm1-disrupted haploid cells that were cultured under nitrogen-starved conditions was analyzed using a Becton Dickinson flow cytometer (FACScan) as reported previously (46), with some modifications. The cells pre-grown in EMM were transferred to ammonium chloride-free EMM. Aliquots of the cells were withdrawn every 2 h and analyzed by FACScan. To examine the effects of overexpression of Stm1, the coding region sequence of stm1 ϩ was prepared by PCR and placed under the thiamine-regulated nmt1 promoter at the BamHI site of the pREP1 vector (45). The resulting plasmid, pnmt1-stm1, was introduced into either the heterothallic haploid (ED665 or various mutants strains) or the homothallic haploid strain (h 90 ). These transformants were initially allowed to grow in EMM containing thiamine. The cells were then transferred into thiamine-free fresh EMM and preincubated at 30°C for 12 h. These conditions allowed induction of stm1 gene transcription. 4 ϫ 10 6 cells/ml were then transferred into thiamine-free fresh EMM at a cell density of A 600 ϭ 0.2 and incubated at 30°C. Aliquots of the cells were removed at the designated times, and cell morphologies were examined under a phase contrast microscope (Carl Zeiss). The amounts of stm1 and sporulation-specific gene transcripts in each cell sample were analyzed by Northern blotting.
Two-hybrid Assay-To examine the interactions between Stm1 protein and other proteins of S. pombe, a yeast two-hybrid assay using GAL4 binding and activation domains was used. The DNAs in the coding region of the test protein genes were amplified by PCR from the genomic DNAs or cDNAs with the corresponding primers. The amplified DNAs were inserted at the 3Ј-end of the GAL4 binding or GAL4 activation domain sequences of the plasmids pGBT9 and pGAD424, respectively. The 1.2-kb fragment of the G protein ␣-subunit gpa1 gene was amplified with the primers 5Ј-CCGGATCCGGATGGGATGCATGTCGAGT-3Ј and 5Ј-CCGGATCCAAACAGCAAACTAGTGTC-3Ј (genomic and c-DNA sequences are underlined in all primers). The 1.06-kb fragment of the gpa2 gene was amplified from the cDNA using the primers 5Ј-CCGGATCCG-GATGACGATTTTTAATGGA-3Ј and 5Ј-CCGGATCCAAACATTCCCGC-TTCTTT-3Ј. These amplified fragments were cloned into the BamHI site of pGBT9 (CLONTECH) or pGAD424 to generate pGBT-Gpa1 and pGBT-Gpa2 or pGAD-Gpa1 and pGAD-Gpa2. The 0.96-kb fragment of the G protein ␤-subunit gene, gpb1, was amplified from the genomic DNA with the primers 5Ј-CCGAATTCATGATATTAGACGGAATACT-3Ј and 5Ј-CC-GTCGACCCCTGACGAAGACCAGAG-3Ј and cloned into the EcoRI and SalI sites of pGBT9 to generate pGBT-Gpb1. The 0.8-kb fragment of the stm1 gene was amplified from genomic DNA with the primers 5Ј-CCGG-ATCCGGATGTCTGTGATAGCTCCTTC-3Ј and 5Ј-CCGGATCCATAAT-GATTACGATACTTGATAA-3Ј and cloned into the BamHI site of pGAD424 or pGBT9 to produce pGAD-Stm1 or pGBT-Stm1, respectively. The 1.0-kb fragment of the mam2 gene of a pheromone receptor was amplified from genomic DNA with the primers 5Ј-CGCGAATTCATGAG-ACAACCATGGTGGAAA-3Ј and 5Ј-GCGGATCCGTCCACTTTTTAGTT-TCAGAT-3Ј, and cloned into the EcoRI and BamHI sites of pGAD424 or pGBT9 to generate pGAD-Mam2 or pGBT-Mam2, respectively. The pGBT and pGAD fusion plasmids were co-transformed into reporter strain SFY526, and these transformants were selected on appropriate synthetic minimal medium (CLONTECH). Transformants were assayed for ␤-galactosidase activity, and the colonies showing a blue color on an indicator filter were retested for ␤-galactosidase activity. As a positive control for specific protein interactions, plasmids pGBT-p53 (pVA3) and pGAD-SV40 large T-antigen (pTD1) were co-transformed. To exclude the possibility of nonspecific interactions, pGBT-Stm1 and pGAD424, or pGBT9 and pGAD-Gpa2, were also co-transformed and used as negative controls.
In Vitro Binding Assay-To investigate whether Stm1 binds to Gpa2 directly in vitro, GST or maltose-binding protein (MBP) fusion proteins were prepared by cloning stm1 into an E. coli expression vector, pGEX-3X, encoding isopropyl-␤-thiogalactopyranoside-inducible glutathione S-transferase (GST) and by cloning gpa2 into pMAL-c1, containing the MBP gene. The coding region DNA or each fraction of stm1 DNA or the coding region DNA of gpa2 was amplified by PCR under denaturing conditions of 95°C for 1 min, annealing at 55°C for 1 min, and polymerizing at 72°C for 2 min. The resulting DNAs were fused at the 3Ј-end of either the GST or MAL gene. The clones expressing proteins of the correct size were selected and used for in vitro binding experiments. Expression of GST-Stm1 or MBP-Gpa2 fusion proteins was induced in the presence of 0.5 mM isopropyl-␤-thiogalactopyranoside at 25°C as described previously (46).
In vitro binding of GST-Stm1 to MBP-Gpa2 was performed as follows. Harvested E. coli cells expressing GST-Stm1 or MBP-Gpa2 were suspended in NETN buffer (20 mM Tris-HCl, pH 8.0, 5 mM Na 2 EDTA, 100 mM NaCl, and 0.5% Nonidet P-40) containing 5 mM phenylmethylsulfonyl fluoride, sonicated, and centrifuged as described previously (46). The supernatant containing MBP-Gpa2 fusion protein was mixed with amylose-agarose beads, and the beads then were washed with NETN buffer. The purified MBP-Gpa2 on agarose beads was incubated with crude E. coli cell extract expressing GST-Stm1 fusion protein in NETN buffer for 1 h. After washing the agarose beads with NETN buffer, the proteins on the beads were subjected to electrophoresis on a 12% SDS-polyacrylamide gel. The proteins on the gel were transferred to a nitrocellulose membrane and analyzed by Western blotting with anti-GST (␣-GST) and anti-MBP (␣-MBP) antibodies.
Construction of Mutant Alleles of gpa2 and stm1-To generate a mutation in the GTPase domain of the gpa2 gene (gpa2 R176H ), we used a double-stranded site-directed mutagenesis kit (Stratagene) and followed the protocol described by the supplier. The sequences of the oligonucleotides used for construction of the R176H mutation in gpa2 were 5Ј-ACCCAGAGTACTATTGTGAGACCGGAGAATGTC-3Ј. The changed sequences are underlined, and this caused the elimination of an XbaI recognition site at amino acid position 176. After confirming the successful construction of the mutation in gpa2 by sequence analysis, we replaced the mutant allele in the chromosome by the method of Moreno et al. (38). A PUC19-based plasmid that contained the 0.85-kb EcoRI-SmaI fragment carrying the 5Ј-upstream element of gpa2, a 1.1-kb BamHI fragment of gpa2 coding region carrying the mutant allele, a 1.8-kb ura4 ϩ cassette with blunt ends, and an 0.6-kb SalI-HindIII fragment carrying the immediate 3Ј-downstream elements to the gpa2 open reading frame, was digested with EcoRI and HindIII. The 4.1 kb-fragment that spans the entire gpa2 region with the mutant allele and ura4 ϩ gene was replaced with the chromosomal copy of wild type gpa2. Altered sequences were confirmed by Southern blotting and by PCR. The oligonucleotides used for generating mutations at amino acid residues 197, 199, or both in stm1 were 5Ј-CGTATTCCTCAAGC-CATCAAGAATCACAAA-3Ј, 5Ј-ATTCCTCAAATCGCGAATCACAAAG-CA-3Ј, and 5Ј-CGTATTCCTCAAGCCATCGCGAATCACAAAGCA-3Ј, respectively. These oligonucleotides were used for site-directed mutagenesis as in the case of the mutation generated in gpa2. The changed nucleotide sequences, confirmed by sequence analysis, caused changes from isoleucine to alanine at position 197 and lysine to alanine at position 199.
Assay of Intracellular cAMP Levels-The method described by Mochizuki and Yamamoto (47) was used to measure cAMP levels in the cells. Cells carrying pnmt1-stm1 were grown first at 30°C in EMM containing thiamine to the mid-log phase. Then the cells were transferred to thiamine-free EMM and grown for 12 h. 4 ϫ 10 6 cells/ml were transferred into fresh thiamine-free EMM and incubated at 30°C. Aliquots of these cells were taken every 3 or 6 h or after 18 h and harvested on glass filters by rapid filtration. The cell samples were immediately soaked in acidic ethanol (0.01 N HCl in ethanol) and extracted using a bead beater. The debris was removed by centrifugation at 12,000 ϫ g for 3 min, and the amount of cAMP was determined using a cAMP assay kit according to the supplier's protocol (Amersham Pharmacia Biotech TRK432). The amount of protein was measured using the dye-binding method (48).
Nucleotide Sequence Deposition-The nucleotide sequence of the stm1 ϩ gene was submitted in 1995 and has been assigned GenBank TM accession number L49134.

Isolation of the stm1 ϩ Gene, Which Suppressed a Synthetic
Lethal Mutant of ras1-In an attempt to identify the elements that function in association with Ras1 in terms of delivering cell surface signals into the cell interior for the differentiation of S. pombe, we looked for novel genes that might function in linking poor nutritional status and pheromone signals with activated Ras1. We isolated one such clone that suppresses a mutant phenotype shown by a ras1 Ϫ synthetic lethal mutant, KSC3. This mutant had been isolated previously (35,36) and contains the chromosomal ras1 gene under the thiamine-regulated nmt1 promoter and a mutation outside of the ras1 gene. It grew normally when the ras1 gene was expressed in thiamine-free medium, but it showed an aggregated and severe growth-retarded lethal phenotype when ras1 gene expression was repressed in the presence of thiamine (Fig. 1, A and B). When we used this mutant to screen the genes that complement the lethal mutant phenotype in the Ras1 Ϫ condition, we identified six novel clones (36). One of these clones showed an insert DNA of 4.3 kb (pWH031). Subcloning of this DNA indicated that an 0.9-kb HindIII-HpaI fragment was necessary to suppress the lethal mutant phenotype of KSC3 under ras1repressed conditions ( Fig. 1B and pYC7 in Fig. 2A). Screening of an S. pombe cDNA library with this 0.9-kb HindIII-HpaI fragment probe identified several cDNAs. Nucleotide sequence analysis of the longest cDNA (1.3 kb) revealed an 813-base pair open reading frame sequence that encoded 271 amino acids with no intron sequences. The deduced amino acid sequence indicated a very hydrophobic protein with a molecular mass of 30.7 kDa. It showed seven transmembrane domains typical of heterotrimeric G protein-coupled receptors (Figs. 2B and 7E). We designated this gene as stm1 (Seven TransMembrane) and found two homologous sequences, YBR147w and YOL092w, in the genome data base of budding yeast S. cerevisiae (Fig. 2B). Sequence analysis of the stm1 locus in chromosomes of the KSC3 mutant revealed the same sequence as found in the isolated stm1 gene. This suggests that the stm1 ϩ gene we isolated by functional complementation of a synthetic lethal phenotype of KSC3 is not a cognate gene of the mutation in KSC3 but is rather a high copy suppressor of the mutation that is outside the ras1 gene.
Transcription of stm1 Is Induced by Nitrogen Starvation-Northern analysis of total RNAs probed with the coding region DNA of stm1 showed a 1.3-kb transcript. The transcript levels in the cells varied depending upon the medium in which the S. pombe cells were grown. As shown in Fig. 3A, the amount of stm1 transcripts in the cells grown in YEPD-rich medium was relatively low (lane 1). In the cells grown in EMM containing ammonium as a nitrogen source, transcription of stm1 increased (lane 2). This increment became higher in the cells grown in EMM containing proline, a poor nitrogen source (lane 3). Scanning and normalization of the stm1 transcript band against a ribosomal protein gene (rp) transcript in each lane showed that transcription of stm1 in ammonium (EMM/NH 4 ) or in proline (EMM/proline) was 5.7 or 13.3 times higher than that in YEPD, respectively. To test whether transcription of stm1 was regulated by depletion of the nitrogen source, the cells were grown first in EMM containing 0.5% ammonium chloride and then transferred to EMM free of ammonium chloride. Within 12 h, the transcript of stm1 increased markedly (Fig. 3B, lane 2), and levels remained high for at least 24 h (lane 3). In contrast to depletion of the nitrogen source, lowering of glucose concentration from 2% in EMM to 0.2% (starvation) did not elicit a change in the levels of the stm1 transcript (data not shown). These results indicate that transcription of the stm1 gene is induced during this nutritionally deficient physiological stage at a time when overall transcription declines dramatically (49,50). Nitrogen-deficient signaling may be the major regulator for this induction.
Western analysis of the HA epitope-tagged Stm1, which was produced from the plasmid pnmt1-stm1-HA3 in S. pombe cells, showed that Stm1 is present mainly in the membrane-enriched fraction rather than in the cytoplasmic fraction (Fig. 3C, lane  6). An in situ localization experiment in which Stm1 fused with jellyfish GFP was used revealed that Stm1 is likely to be associated with the plasma membrane (Fig. 3D), suggesting that Stm1 may function as a membrane protein.
Disruption of stm1 Facilitates G 1 Arrest under Nitrogendeficient Conditions-To explore Stm1 function, the stm1 ϩ gene was disrupted by a one-step gene disruption method using the ura4 ϩ gene as a marker gene (Fig. 4A). The entire coding region DNA of stm1, including a 100-base pair 5Ј-upstream and a 30-base pair 3Ј-downstream sequence was replaced with ura4. A diploid h ϩ /h ϩ strain that contained one wild type and one disrupted copy of stm1 was selected, converted into h ϩ /h 90 , and sporulated. All of the resulting tetrads were viable on YE plates and showed 2 ϩ :2 Ϫ segregation on Ura Ϫ plates. All Ura Ϫ spores were wild type. The stm1-disrupted haploid cells showed normal vegetative growth in nutritionally sufficient medium (YEPD or EMMϩN, Fig. 4D). However when the stm1-disrupted haploid cells were transferred to the nitrogen-depleted EMM (EMMϪN, Fig. 4C), they exhibited G 1 arrest much earlier than observed in the undisrupted wild type 972 cells. Flow cytometric analysis showed that half of the wild type cells progressed to G 1 4 h after transfer to the nitrogen-deficient medium (Fig. 4B, lane 1), whereas most of the stm1-disrupted cells shifted to G 1 phase within 2 h and remained arrested in the G 1 phase (Fig. 4B, lane 2). However the rapid G 1 arrest shown by the stm1-deleted cells under nitrogen-deficient conditions was not observed in the cells deleted in both stm1 and ras1 (stm1 Ϫ , ras1 Ϫ ) genes (Fig. 4B, lane 4). Instead, the stm1 Ϫ and ras1 Ϫ double-deleted cells showed continued cell cycle progression under nitrogen-deficient condition (Fig. 4B, lane 4,  and C, filled circles). We consistently observed that stm1 Ϫ ras1 Ϫ cells continued to grow at least for 24 h after transfer to nitrogen-deficient EMM (EMMϪN, Fig. 4C, filled circles). In EMM containing a sufficient nitrogen source (EMMϩN), the stm1 Ϫ , ras1 Ϫ double-disrupted cells showed a significant decrease in vegetative cell growth (Fig. 4D, filled circles), whereas the cells disrupted only in the stm1 gene grew like undisrupted wild type cells (Fig. 4D, open squares). This finding suggests that even though the stm1 gene itself is not essential for vegetative cell growth in nutritionally sufficient media, when ras1 is not functional, deletion of stm1 causes defects not only in vegetative cell growth under nutritionally sufficient conditions but also in G 1 arrest under nutritionally deficient conditions. It is possible that Ras1 can function cooperatively with Stm1 in regulating vegetative cell growth during the nitrogen deficient period.
Overexpression of Stm1 Inhibits Vegetative Cell Growth and Lowers Intracellular cAMP Levels-To investigate how stm1 gene expression drives S. pombe cells to enter either G 1 arrest or the G 0 phase under nitrogen-deficient conditions, we examined the effects of stm1 overexpression. The coding region DNA of stm1 ϩ was fused with the thiamine-regulatable nmt1 promoter (pnmt1-stm1). Expression of stm1 was examined by growing stm1 ϩ haploid cells containing this plasmid in the absence of thiamine. The cells were first grown in EMM containing thiamine that repressed expression of stm1. Then the cells were transferred to thiamine-free EMM, as indicated in Fig. 5A, arrow a, and grown for a further 12 h to remove residual thiamine in the cells. For stm1 induction, the cells were transferred again to fresh EMM containing no thiamine (Fig. 5A, arrow b), and the levels of the stm1 transcript, cell growth, and morphology were examined at every 3-6 h thereafter. As shown in Fig. 5B, transcription of stm1 increased markedly after 6 h of actual thiamine induction and continued to increase. DAPI and Calcofluor staining of wild type haploid cells (ED665) expressing Stm1 showed multisepta after 18 h of thiamine induction (Fig. 5C, panel 2). Vegetative cell growth in nitrogen sufficient EMM (EMMϩN) was inhibited as stm1 induction continued (Fig. 5D, open squares). Many of the cells were lethal after 24 h (Fig. 5E, lower panel). However, this severe growth inhibition and the multisepta-forming effects of Stm1 were not observed in the cells defective in the gpa2 gene, a heterotrimeric G protein ␣2 gene known to be responsible for relaying nutritional signal at the initial stage of sexual development in S. pombe (19) (Fig. 5C, panel 4, and D, open triangles). This indicates that Gpa2 function is required for Stm1 to be effective in altering cell growth under nutritionally sufficient condition. In a homothallic haploid strain, h 90 , overexpression of Stm1 caused facilitated sporulation to occur even in a nitrogen-rich medium in which sporulation is not normally observed (Fig. 6A, panel 2). The addition of 2 M cAMP and 5 mM caffeine, a treatment that regulates meiotic cell division in S. pombe and inhibits phosphodiesterase activity, respectively (19), suppressed the facilitated conjugation and sporulation phenotype caused by overexpression of Stm1 (Fig. 6A, panel 3).
To examine whether the hyper-sporulation effect of Stm1-overexpression in homothallic cells was due to changes in intracellular cAMP concentrations regulated mainly by the heterotrimeric G␣2 protein Gpa2 in S. pombe, the intracellular cAMP levels were monitored during stm1 induction. As shown in Table II (experiment 1), stm1 ϩ heterothallic haploid cells expressing Stm1 showed decreased cAMP levels as stm1 induction increased. Within 3 h after stm1 induction, intracellular cAMP levels did not change much. However after 6 h, when stm1 transcripts increased markedly (Fig. 5B), cAMP levels dropped to less than half of that observed in the uninduced state (2.5 versus 1.0). This value remained around 1.0 as stm1 transcripts continued to increase as thiamine induction progressed ( Fig. 5B and Table II, experiment 1). The increase in stm1 transcripts correlates with the decreases in cAMP levels down to a certain level. No further decrease in cAMP was observed despite further increases in the levels of stm1 transcripts. In homothallic cells we observed similar results (Table  II, (Table II, experiment 2, HD3, 1.3).
Deletion of the stm1 gene itself, or the ras1 gene that is required for initiation of sexual differentiation in S. pombe, did not elicit any change in intracellular cAMP levels (Table II, experiment 2, HD1 and HD8, 2.8 -2.5). Stm1 overexpression in the gpa2-deleted cells did not cause a further decrease in intracellular cAMP levels in gpa2 Ϫ cells (1.3 in both HD3 and KS3). To test whether this decrease in cAMP was because of modulation of cAMP-phosphodiesterase (Pde1), which breaks down cAMP, we examined cAMP levels in the pde1 null mutant expressing stm1. As shown in Table II, the pde1 null mutant cells (HD9), which are defective in cAMP hydrolysis, showed cAMP levels four times higher than those observed in pde1 ϩ cells (14 versus 3.3). This value decreased to 6.2 in the pde1 null mutant cells overexpressing Stm1 (KS7). In the cells carrying deletions for both pde1 and gpa2 genes (HD11), cAMP levels were slightly higher than observed in undisrupted gpa2 ϩ pde1 ϩ cells. These results support the hypothesis that not cAMP hydrolysis but cAMP production itself is affected by Stm1 overexpression and also supports earlier findings by Isshiki et al. (19) who showed that a null mutation in gpa2 inhibits cAMP production in pde1-deleted strains. In the strain that has a mutation in the GTPase domain in Gpa2 (gpa2 R176H ) we observed slightly higher levels of cAMP (HD13, 5.3). Overexpression of Stm1 in this strain did not alter this value very much (KS9, 4.9) as was observed in the gpa2 null mutant expressing Stm1 (HD3 and KS3, 1.3). Additional changes in the intracellular cAMP levels in gpa2 mutants overexpressing Stm1 were not observed. In the absence of functional Gpa2, expressed Stm1 could not affect intracellular cAMP levels (51). These results indicate that proper Gpa2 function is required for Stm1 to be effective in regulating intracellular cAMP and that gpa2 ϩ is epistatic to stm1 ϩ in cAMP regulation.
Stm1 Overexpression Leads to Increased Transcription of Meiosis-specific Genes-The finding that overexpression of Stm1 causes a reduction of intracellular cAMP levels and facilitates sexual development in homothallic cells (Table II, Fig. 5A. After 24 h, cell samples were removed and the degree of sporulation was determined under a DIC (differential interference contrast microscope). 2 M cAMP and 5 mM caffeine were added in thiamine-free EMM for panel 3. B, changes in the transcript levels of the meiosis-specific genes (ste11, mei2, and mam2) involved in the initial stages of sexual differentiation by overexpression of Stm1. wt, ED665 h Ϫ heterothallic or JY4 h 90 homothallic cells containing pnmt1-stm1 (KS1 and KS2). wt ϩcAϩCaf , KS2 cells grown in the presence of 2 M cAMP and 5 mM caffeine in thiamine-fee EMM. stm1 Ϫ , the stm1-deleted heterothallic h Ϫ haploid (stm1::ura4) or homothallic h 90 haploid (stm1::ura4) containing pnmt1-stm1 (KS11 and KS4). gpa2 Ϫ or sty1 Ϫ , the gpa2-deleted or sty1-deleted heterothallic haploid containing pnmt1-stm1 (KS3 or KS6). "ϩ" indicates the induction of stm1 expression, and "Ϫ" indicates the absence of stm1 induction from pnmt1-stm1. These RNAs were prepared from cells grown for 18 h after transfer to the fresh thiamine-free EMM. "rp" indicates the transcript of a ribosomal protein gene used as an internal control. Lanes 1-3, 4 -7, 8 -9, and 10 and 11 show results from four separate experiments. meiosis-specific genes necessary for sexual development. Therefore, we examined whether transcription of ste11, mei2, or mam2 genes, required for initiation of meiosis (52)(53)(54)(55), were induced by Stm1 overexpression. As shown in Fig. 6B, lanes 3 and 6, transcription of a meiosis-specific gene transcription factor, ste11, increased in both the heterothallic and homothallic haploid cells as stm1 transcript levels increased. The mei2 gene was also induced severalfold in the wild type cells expressing Stm1. These inductions were not observed in the cells grown in the presence of exogenous cAMP and caffeine (Fig. 6B,  lane 7). Mating factor receptor mam2 expression also increased in a similar manner (Fig. 6B, lane 3). Thus the suppression of hyper-sporulation phenotype of Stm1 overexpression by the addition of cAMP and caffeine (Fig. 6A, panel 3) in growth media correlates with the inhibition of meiosis-specific gene transcription (Fig. 6B, lane 7). In the gpa2-disrupted cells, the increase in stm1 transcripts caused a marked induction of the ste11 transcript and a fewfold induction of the mei2 transcript (Fig. 6B, lanes 8 and 9), whereas intracellular cAMP levels remained less than half that of gpa2 ϩ cells regardless of an increase of the stm1 transcript (Table II, HD3 and KS3). This showed that multiple copies of Stm1 affect transcription of ste11 responsible for the induction of pre-meiotic gene transcription even in the absence of Gpa2 function. It is already known that ste11 is regulated by stress-activated StyI protein kinase through Atf1 transcription factor in response to nitrogen limitation (64). The latter is a key signal that promotes sexual development in S. pombe, and so we examined whether high expression levels of the ste11 transcript induced by Stm1 was due to a modulation of the sty1-associated pathway. The marked induction of ste11 caused by Stm1 overexpression in the gpa2-disrupted cells was abolished when the sty1 gene was deleted (Fig. 6, lanes 9 and 11). These results suggest that unless a stress-activated pathway gene such as sty1 is active, increased transcription of meiosis-specific genes caused by overexpressing Stm1 is not possible. StyI function is required for Stm1 to be effective in the induction of meiosis specific-gene transcription. Thus it is likely that the stress-activated sty1 gene is epistatic to stm1 in meiosis-specific gene transcription. This confirms an earlier finding by Shiozaki and Russell (64) who showed that the stress-activated Wis1-StyI protein kinases are upstream regulators of meiosis-specific gene transcription.
Stm1 Is Likely to be Coupled with Gpa2, Which Is Involved in the Monitoring of Nutrition-Our observation that the stm1 transcript is induced under nitrogen-deprived conditions and that overproduction of Stm1 causes a reduction in internal cAMP levels led us to examine whether Stm1 can function through Gpa2, the mechanism by which intracellular cAMP production in S. pombe is known to be regulated in response to nutritional signals (19,47). First, we examined whether Stm1 interacts directly with Gpa2 and affects the ability of Gpa2 to modulate intracellular cAMP levels. Using yeast two-hybrid experiments with stm1 fused to the GAL4 binding domain (GAL4BD) and gpa2 fused to the GAL4 activation domain (GAL4AD), or vice versa, we observed ␤-galactosidase gene (lacZ) expression with both an X-gal filter assay and a quantitative liquid assay (Fig. 7A). In contrast, when the other G␣ protein gene, gpa1, which is responsible for transmitting the pheromone signal for sexual development (18), was fused with GAL4 and co-transformed with the GAL4-stm1 fusion construct, we did not observe any ␤-galactosidase expression. This

TABLE II
Overexpression of Stm1 lowers the cAMP level Cells were grown in EMM with thiamine and reinoculated in EMM without thiamine. All samples except in Experiment 1 were removed after 18 h, and cAMP level was measured at short interval. Two samples were assayed and the same results were obtained in separate experiments.

Strains
Relevant genotype Plasmid cAMP level pmol/mg of protein (2), inducible overexpression plasmid under the nmt1 or nmt2 promoter carrying LEU2 or ura4 selective marker, respectively.

FIG. 7. Interaction of Stm1 protein with the heterotrimeric G proteins.
A, yeast two-hybrid interactions between G␣ subunit proteins and Stm1. Stm1 fused to the GAL4 binding domain plasmid pGBT9 or the activation domain plasmid pGAD424 was co-transformed with either gpa1 or gpa2 fused to the GAL4 activation or binding domain plasmid. The degree of interaction between the two proteins was assessed by the amount of lacZ expression on the X-gal indicator filter and by a quantitative ␤-galactosidase liquid assay. The co-transformed cells were either patched on minimal plates or grown in 0.67% yeast nitrogen base, 2% glucose minimal medium to a cell density of A 600 ϭ 0.6 -0.8. ␤-Galactosidase activity in the cells was measured with permeabilized cells as described by Miller (69). B, domains of Stm1 protein responsible for the interaction with Gpa2. Each fraction of stm1 (D, ⌬C, and ⌬N) was amplified by PCR with the appropriate primers, cloned into the GAL4 binding domain plasmid pGBT9, and used to transform S. cerevisiae with full-length gpa2 fused at the GAL4 activation domain. Activation of ␤-galactosidase was examined using the X-gal filter assay. showed that Stm1 could interact with Gpa2 but not with Gpa1. Meanwhile, when the mam2 gene encoding a receptor for pheromone M factor was fused with GAL4 and co-transformed with GAL4-gpa1, activation of ␤-galactosidase expression was observed. However when GAL4-mam2 was co-transformed with GAL4-gpa2, it did not activate ␤-galactosidase expression (Fig.  7A). Therefore, the Mam2 receptor, specific for the pheromone signal, interacts only with the pheromone-responsive G␣ subunit, Gpa1, and not with the nutrient responsive G␣ subunit, Gpa2. The G␤ subunit gene, gpb1, which we had identified previously (20), induced ␤-galactosidase expression both with stm1 and mam2 (data not shown). These results suggest that Stm1 has the specific ability to interact directly with Gpa2 in S. pombe cells and also with the G␤ subunit, Gpb1. For this reason we examined which domains of Stm1 were directly involved in these interactions. Different regions of stm1, encoding the amino acids of the corresponding domains (Fig. 7B, D1-D4) and 3Ј-or 5Ј-deleted stm1 (Fig. 7B, ⌬C and ⌬N) were cloned separately into the Gal4 binding domain. These fusion constructs were used for yeast two-hybrid experiments with the entire gpa2 coding sequences fused to the Gal4 activation domain. As shown in Fig. 7B, the fragments D3 and D4 of Stm1, which contained amino acids 175-228 and 229 -271, respectively, showed activation of lacZ expression. Whereas deletion of the C-terminal 111 amino acid sequences, including the D3 and D4 regions, abolished lacZ expression (Fig. 7B,  ⌬C2), deletion of the N-terminal 142 amino acids did not affect the expression of the lacZ gene (Fig. 7B, ⌬N3).
The C-terminal Domain of Stm1 Is Required for the Possible Interaction with Gpa2-To confirm the binding ability of Stm1 to Gpa2, in vitro binding between Stm1 and Gpa2 was tested using the GST and MBP fusion proteins produced by E. coli. The full-length stm1 or its different regions, as shown in Fig.  7B, were fused to the 3Ј-end of the glutathione S-transferase gene (GST-stm1) and were expressed in E. coli. The amount of each domain protein of Stm1 expressed was verified by SDS-PAGE ( Fig. 7C-a, upper panel) and by Western blotting with anti-GST antibody ( Fig. 7C-a, lower panel). The entire coding region DNA of the gpa2 gene fused to the maltose-binding protein gene (MBP-gpa2) was expressed in E. coli and purified on agarose beads. The purified MBP-Gpa2 ( Fig. 7C-b, lower panel) was mixed with E. coli cell extracts expressing each domain protein as a GST-Stm1 fusion protein as shown in Fig.  7C-a. The proteins associated with MBP-Gpa2 were analyzed by SDS-PAGE and by Western blotting (Fig. 7C-b, upper panel). As found in the yeast two-hybrid experiments, the fulllength Stm1 showed direct binding to Gpa2 (Fig. 7C-b, lanes 3  and 9). When E. coli cell extracts expressing different regions of Stm1 as GST fusion proteins were mixed with the purified MBP-Gpa2, the proteins that were deleted in the N-terminal half were found not to affect the ability of Stm1 to bind Gpa2 ( Fig. 7C-b, lanes 12-14). However, deletion of a portion of the C-terminal region, domains 3 and 4 (⌬C2) abolished in vitro binding ability with Gpa2 ( Fig. 7C-b, lane 11). GST fusion containing these D3 or D4 regional sequences alone was enough for binding to Gpa2 (Fig. 7C-b, lanes 6 and 7). The other regions did not show this binding ability with Gpa2 (lanes 4 and 5). The region between the amino acid sequences 175 and 271 (D3ϩD4) seems to be necessary for the interaction with Gpa2. This region contains the third cytoplasmic loop (Fig. 7E), known to be required for coupling with G proteins in other well characterized G protein-coupled receptors (56,57). The amino acid sequences KHNKIIQ conserved in several G protein-coupled receptors, e.g. STE2, STE3, and GPR1 of S. cerevisiae and Mam2 and Map3 of S. pombe (58), are present in the third cytoplasmic loop region between amino acids 193 and 207 (Fig.   7, D and E). To test whether these conserved sequences are actually involved in interaction with Gpa2 and might affect some of Gpa2 function, one or both of the conserved sequences at amino acid 197 (Ile) and 199 (Lys) were changed to alanine by site-directed mutagenesis as described under "Experimental Procedures." The mutated stm1 was overexpressed in the absence of the chromosomal copy of stm1 ϩ , and the degree to which intracellular cAMP levels decreased with the overexpression of Stm1 was taken as a measure of the interaction between Stm1 and Gpa2. As shown in Table II, experiment 4, overexpression of the mutated Stm1 at the position 197 (HD1/ pnmt1-stm1 I197A ) caused a reduction of cAMP levels to less than half of that observed in the uninduced state (1.4 versus 2.8 -3.1). These levels were the same as those observed during overexpression of the un-mutated Stm1 (1.2, HD1/pnmt1-stm1). However overexpression of Stm1 mutated at the position 199 (HD1/pnmt1-stm1 K199A ) alone or at both positions 197 and 199 (HD1/pnmt1-stm1 I197A,K199A ) did not show a decrease in cAMP levels (2.4 -2.5). This indicates that among the sequences at the putative Gpa2 binding site, at the third cytoplasmic loop, lysine at position 199 in Stm1 is critical for Gpa2 function in regulating cAMP production. In addition, as shown earlier with the gpa2 null mutant or the GTPase mutant gpa2 R176H overexpressing Stm1 (Fig. 5C and Table II, KS9), the functional Gpa2 is also required if Stm1 is to be effective in modulating intracellular cAMP levels. Without functional Gpa2, Stm1 could not affect intracellular cAMP levels and/or inhibit cell growth. DISCUSSION Nutrient starvation in the fission yeast S. pombe leads to a number of physiological changes that accompany entry into the stationary phase or G o (59 -61). The cells in the stationary phase become differentiated in a way that allows the maintenance of viability for extended periods without added nutrients but enabling the yeast to retain the ability to resume growth promptly when appropriate nutrients become available (61). When mating partners are present the yeast undergoes sexual development. The decision whether to stay in the active growth cycle or to enter into the differentiation cycle is made by the coordinated interactions of complex molecules. For proper sexual differentiation, i.e., conjugation followed by meiosis and sporulation, the concerted action of a nutritional starvation signal, which brings S. pombe cells to arrest at G 1 , and a pheromone signal, which activates heterotrimeric G proteins and modulates MAP kinases via Ras1, are required. Integrated signals from the cell surface are required to arrest the ongoing cell cycle and facilitate entry into the differentiation cycle. Ras1 plays a critical role in this process by activating the MAP kinase pathway genes when the pheromone signal is accepted by a heterotrimeric G protein. Several elements are known to function together in association with Ras1, but before Ras1 can play its full part in sexual development, S. pombe cells must be in the physiological prerequisite state. This state can be initiated by nutritional starvation. The Ras1 function in activating the MAP kinase pathway genes such as byr2 ϩ and byr1 ϩ in response to pheromone is reasonably well characterized (12,27,33). In contrast, little is known about how the nutritional starvation signal is relayed to the downstream elements to execute the differentiation of S. pombe. In particular, the sensor capable of detecting nutritional deficiency and causing S. pombe cells to enter the meiotic cycle has yet to be identified. Moreover, the connection between nutritional deprivation and sexual development is poorly understood. The novel gene, stm1 ϩ , isolated in this study as a multicopy suppressor of a synthetic lethal mutant of ras1, encodes a molecule that may function as a sensor of changes in nutritional conditions and may signal the regulation of cell growth and/or differentiation.
The finding that the deletion of the stm1 gene itself was not detrimental to cell growth under nutritionally rich conditions but caused facilitated entry into G 1 arrest under nitrogenstarved conditions supports the notion that Stm1 is involved in regulating cell cycle progression under nitrogen-depleted conditions. Delayed entry into the G 1 arrest under nitrogen-depleted conditions in the stm1 and ras1 double-deleted mutant suggests that ras1 ϩ function is required for proper stm1 function in regulating the ongoing cell cycle at the initial stages of nitrogen starvation.
The following results support the hypothesis that Stm1 functions through the heterotrimeric G protein-coupled signaling pathway: (i) the seven-transmembrane domain structure of Stm1; (ii) the requirement for the C-terminal third cytoplasmic loop that contains conserved sequences shown in several G protein-coupled receptors for interaction with Gpa2; and (iii) no change in intracellular cAMP levels by overexpression of Stm1 mutated in this region (Lys 199 3 Ala). Even though we could not identify precisely the molecule that directly activates Stm1, it is possible that under nutritionally unfavorable conditions some nitrogenous metabolites may bind to Stm1 in a ligandreceptor-specific manner, changing its structure and activating Stm1 to interact with Gpa2 through its C-terminal region. This interaction may lead changes in adenylate cyclase activity resulting in regulatory changes in intracellular cAMP levels, supporting the earlier observation in S. pombe by Mochizuki and Yamamoto (47) that the cAMP level changes during transition from the exponential to the stationary phase or in response to a shift from nitrogen-rich to nitrogen-free medium. Therefore, although we could not totally exclude the possibility that Stm1 modulates cAMP levels independently of Gpa2, the putative G protein-coupled receptor protein gene stm1 ϩ may have a significant role in recognizing the nitrogen starvation signal and relaying this signal to modulate the Gpa2-associated pathway. The Gpa2 protein then mobilizes the elements necessary to distinguish whether S. pombe cells should remain in the vegetative cell cycle or switch to the differentiation cycle.
Meanwhile, because nitrogen starvation presents cells with a type of stress, it is possible that the stress-activated Wis1-StyI pathway of S. pombe (62)(63)(64)(65)(66)(67) could be mobilized by the nitrogen starvation signal accepted by Stm1. This pathway has already been known to regulate at least two different downstream targets. One branch proceeds by atf1 to cause G 1 arrest and induce expression of a meiosis-specific gene transcription factor, ste11, in response to nitrogen limitation (63). The other branch regulates the onset of mitosis. Our finding that transcript levels of ste11 were markedly increased by the overexpression of Stm1 in the gpa2-deleted cells, even though intracellular cAMP levels remained at a half of those observed in wild type cells (Fig. 6B, lane 9, and Table II), suggests that Stm1 could function through other pathways to induce ste11 transcription. During overexpression of Stm1 (Fig. 6B, lanes 10  and 11), the absence of induction of ste11 or mei2 gene transcription in the deletion mutant of sty1, a stress-activated protein kinase gene, strongly supports this idea. In S. pombe, because it is known that the stress-activated genes such as sty1 and wis1 activate transcription of meiosis-specific genes (66), it is possible that Stm1 could stimulate not only a Gpa2-associated pathway but also stress-activated pathways to synergistically activate the required gene functions for the initial stage of differentiation before the pheromone signal plays its part. The manner in which Stm1 senses intracellular nitrogen levels remains unknown, but there is a possibility that Stm1 senses this external nutritionally deficient status as a stress signal, modulates Gpa2, and at the same time activates a stressactivated pathway. The balanced function between the Gpa2-associated and StyI-associated pathways may be critical in sensing the nutritional states of the cells and functioning accordingly.
A novel G protein-coupled receptor, Stm1, that we characterized in this study may function as a pivotal molecule, acting upstream of Ras1, which may allow cross-talk between nutritional starvation and pheromone signals. The degree of cross-talk between the two signals may determine the fate of S. pombe cells as to whether to remain in the mitotic cell cycle or to undergo the pheromone-dependent Ras1-mediated meiotic differentiation cycle under nutritionally unfavorable conditions. We suggest that subscribers photocopy these corrections and insert the photocopies at the appropriate places where the article to be corrected originally appeared. Authors are urged to introduce these corrections into any reprints they distribute. Secondary (abstract) services are urged to carry notice of these corrections as prominently as they carried the original abstracts.