SSF1 and SSF2 are redundant essential yeast genes that, when overexpressed, increase the mating efficiency of cells containing a defective Ste4p Gβ subunit. To identify the precise function of these genes in mating, different responses to pheromone were assayed in cells that either lacked or overexpressed SSF gene products. Cells containing null alleles of both SSF1 and SSF2 displayed the normal transcriptional induction response to pheromone but were unable to form mating projections. Overexpression of SSF1 conferred the ability to form mating projections on cells containing a temperature-sensitive STE4 allele, but had only a small effect on transcriptional induction. SSF1 overexpression preferentially increased the mating efficiency of a strain containing a null allele of SPA2, a gene that functions specifically in cell morphology. To investigate whether Ssf1p plays a direct physical role in mating projection formation, its subcellular location was determined. An Ssf1p-GFP fusion was found to localize to the nucleolus, implying that the role of SSF gene products in projection formation is indirect. The region of Ssf1p-GFP localization in cells undergoing projection formation was larger and more diffuse, and was often present in a specific orientation with respect to the projection. Although the function of Ssf1p appears to originate in the nucleus, it is likely that it ultimately acts on one or more of the proteins that is directly involved in the morphological response to pheromone. Because many of the proteins required for projection formation during mating are also required for bud formation during vegetative growth, regulation of the activity or amount of one or more of these proteins by Ssf1p could explain its role in both mating and dividing cells.
IN the yeast Saccharomyces cerevisiae, the response to pheromones is transmitted through cell surface receptors coupled to a heterotrimeric G protein that activates a mitogen-activated protein (MAP) kinase cascade (Kurjan 1992; Sprague and Thorner 1992). Activation of this pathway results in many changes in cellular physiology, including cell cycle arrest, transcriptional induction of genes required for mating, and morphological changes that result in mating projection formation. Cells form projections at the site of highest pheromone concentration and are capable of reorienting to a new pheromone gradient (Segall 1993), indicating that occupied pheromone receptor initiates the morphological response. The signaling components involved in projection formation appear to form a large complex that transmits a signal from the receptor to proteins involved in cytoskeletal reorganization (Wittenberg and Reed 1996; Lebereret al. 1997a). This complex includes the Ste20p kinase, which is thought to transmit the pheromone response signal from the G protein to the MAP kinase cascade (Lebereret al. 1992). Ste20p contains a domain that binds Cdc42p (Peteret al. 1996; Lebereret al. 1997b), a small guanine nucleotide binding protein that is involved in the establishment of cell polarity (Adamset al. 1990; Zimanet al. 1991). The Cdc42p binding domain of Ste20p is necessary for its localization to the projection tip, but is not required for mating projection formation (Peteret al. 1996; Lebereret al. 1997b). Another component that has been proposed to be part of the morphology signaling complex is Bem1p, a protein involved in bud emergence during vegetative growth (Bender and Pringle 1991; Chenevertet al. 1992). Bem1p associates with Ste20p (Leeuwet al. 1995) and with Cdc24p (Petersonet al. 1994), the exchange factor for Cdc42p (Zhenget al. 1994). Bem1p has also been shown to associate with Far1p (Lyonset al. 1996), a protein that is required both for cell cycle arrest and for orientation of mating projections toward the highest concentration of pheromone (Chang and Herskowitz 1990; Valtzet al. 1995). Although projection formation does not require all of the interactions known to occur between these proteins, it seems likely that they exist as a large complex in the cell, and that this complex plays a role in the morphological response to pheromone. For the morphology signaling complex to effect changes in cell shape, it must alter the organization of the cytoskeleton. However, the link between the signaling complex and the cytoskeleton has not been clearly established. One protein that has been proposed to have this function is Bem1p, which appears to bind both the signaling molecules mentioned above and actin (Leeuwet al. 1995). Another protein that has been proposed to have this function is Bni1p, which appears to bind Cdc42p, actin, and actin-associated proteins (Evangelistaet al. 1997).
Mating projection formation involves polarized cell growth in response to an external cue. This change in morphology requires actin, which is present in the cortical region of the growing projection (Readet al. 1992). An early step in polarity establishment in yeast is localization of the Spa2p protein to a small region of the cell surface that is the presumptive site of cell growth (Snyder 1989). Spa2p localizes to the bud tip in dividing cells and to the projection tip in cells responding to pheromone. Localization of Spa2p is partially defective in cells containing mutations in the actin gene (Readet al. 1992) and is delayed in cells that do not contain actin filaments (Ayscoughet al. 1997). These results indicate that Spa2p localization does not strictly require actin, but that the efficient localization of Spa2p may be facilitated by the actin cytoskeleton. Although Spa2p appears to function in an actin-independent manner, it is also required for mating projection formation. Cells containing null alleles of SPA2 do not form normal mating projections and display a decrease in mating efficiency, demonstrating that Spa2p plays an important role in this process (Gehrung and Snyder 1990; Doreret al. 1995). spa2 mutant cells exposed to high concentrations of pheromone for long periods of time are capable of forming broad projections that result in peanut-shaped cells (Valtz and Herskowitz 1996), suggesting that they retain some residual ability to undergo polarized growth.
The SSF1 and SSF2 genes encode nearly identical proteins that were isolated in a screen for genes that could augment the mating efficiency of a strain that is compromised in its ability to respond to pheromone (Yu and Hirsch 1995). Overexpression of either SSF1 or SSF2 increases the mating efficiency of cells containing a temperature-sensitive allele of STE4, the gene that encodes the β-subunit of the G protein. Although cells containing null alleles of either SSF1 or SSF2 display no obvious phenotype, double Δssf1 Δssf2 mutants are inviable. Depletion of SSF gene products causes growth arrest and a significant decrease in mating efficiency. These results suggest that, like CDC24 (Chenevertet al. 1994; Zhaoet al. 1995), CDC42 (Adamset al. 1990; Simonet al. 1995), BEM1 (Bender and Pringle 1991; Chantet al. 1991; Chenevertet al. 1992), STE20 (Lebereret al. 1992; Ramer and Davis 1993; Cvrckovaet al. 1995), and SPA2 (Gehrung and Snyder 1990; Costiganet al. 1992), the SSF genes have an important function in vegetative growth as well as a potential function in mating. Here we show that the SSF function in mating acts predominantly on the morphological response to pheromone.
MATERIALS AND METHODS
Plasmid construction: pGAL-SSF1.14 and pSSF1-2 have been described previously (Yu and Hirsch 1995). The pfar1c-1::URA3 plasmid, which contains a URA3 disruption of the carboxyl-terminal domain of FAR1, was constructed using a version of the 1.2-kb HindIII fragment that contains the URA3 gene in which the HindIII sites have been converted to XbaI sites by filling them in with Klenow polymerase. The 1.2-kb XbaI fragment containing URA3 was cloned into the XbaI sites of a 2.4-kb XhoI-EcoRI fragment containing FAR1 from pJM306 (McKinney and Cross 1995), which had been cloned into pGEM-2 (Promega, Madison, WI). The far1c-1::URA3 allele is predicted to express the amino acids 1–536 of Far1p, which contains the region necessary for its G1 arrest function (Valtzet al. 1995).
The construct that fuses SSF1 to green fluorescent protein (GFP) gene was made as follows. First, a NotI site was inserted immediately before the stop codon in SSF1 by performing the polymerase chain reaction (PCR; Daughertyet al. 1991) with oligonucleotide primers oSSF1-1, 5′-CCGGATCCATCCAGATATAGCAGAC-3′, and oSHA3, 5′-CCGGATCCTAGCGGCCGCATTCGACCTCACTAA-3′ (SSF1 sequences are underlined) to generate a fragment containing the entire coding region of SSF1. This 1.4-kb fragment was cloned into the BamHI site of YCplac33 (Gietz and Sugino 1988), using the terminal BamHI sites present in the primers, to create pSSF1r-NotI. Then, a NotI site was inserted at the 5′ end of the GFP gene by performing PCR with oligonucleotides oGFP-1, 5′-TTAGCGGCCGCAGTAAAGGAGAAGAACTTTTC-3′, (GFP sequences are underlined) and the T7 primer, 5′-AATACGACTCACTATAG-3′, using as the template plasmid pBlueScript-S65T, which contains the version of GFP that has serine 65 changed to threonine (Heimet al. 1995). The 0.7-kb fragment generated by PCR was digested with NotI and cloned into the NotI site of pSSF1r-NotI to create pSF-GP1. The final construct, pSF-GP2, was created by cloning the 1.3-kb XbaI fragment from pSF-GP1, which contains the 3′ end of SSF1 fused to GFP, into the XbaI site of a plasmid containing a 3.4-kb HpaI-XbaI fragment, which includes the 5′ end of SSF1 and 2.8-kb of upstream flanking region, cloned into the SmaI-XbaI sites of YCplac33. YEpSFGP was constructed by cloning the 3.0-kb BamHI fragment from pSF-GP2 into the BamHI site of YEp351.
Strains and media: Strain H50-16C.Ba was derived from H50-16C (Yu and Hirsch 1995), which has the genotype MATa leu2-3,112 trp1-1 can1-100 ura3-1 ade2-1 his3-11,15 ssf1-1::HIS3 ssf2-3::TRP1 [pGAL-SSF1.14]. H50-16C.Ba was constructed by transformation of H50-16C with a 5.7-kb EcoRI-SalI fragment from pJGSST1 (Elionet al. 1993) to create sst1::URA3 followed by deletion of URA3 from the chromosome through recombination of the adjacent HisG sequences to create the sst1::HisG allele, as described (Elionet al. 1993). Strain JH8-3510a was constructed using a GAL-HO construct to switch the mating type of JH8-3510 (Yu and Hirsch 1995), which has the genotype MATα leu2 trp1 ura3 his4-519 can1-101 ste4-3510, to MATa. Strain JH8-3510a.sp was derived from JH8-3510a by disruption of the SPA2 gene using a 3.8-kb HindIII-SalI fragment from plasmid p210 to create the spa2- Δ3::URA3 allele (Gehrung and Snyder 1990). Strain JH8-3510a.fc was derived from JH8-3510a by disruption of the carboxyl-terminal region of the FAR1 gene using a 2.0-kb HindIII-EcoRI fragment from plasmid pfar1c-1::URA3 to create the far1c-1::URA3 allele. Strain JH111-1C was constructed by transformation of strain W3031A (obtained from R. Rothstein, Columbia University), which has the genotype MATa leu2-3,112 trp1-1 can1-100 ura3-1 ade2-1 his3-11,15, with a 3.8-kb SacI-SphI fragment from pJH111 (Yu and Hirsch 1995) to generate an ssf1-1::HIS3 allele. Strain H67-9D.Ba has been described previously (Couve and Hirsch 1996). Strain AC17-7B is a derivation of H67-9D.Ba in which the STE4 gene has been replaced with a ste4::HIS3 allele. Mating assays utilized strain W3031B (obtained from R. Rothstein), which has the genotype MATα leu2-3,112 trp1-1 can1-100 ura3-1 ade2-1 his3-11,15, as the mating partner. All strain constructions involving transformations were confirmed by Southern blot.
Strains were grown on YEPD (2% glucose) or YEP-Gal (3% galactose), and strains under selection were grown on synthetic dropout media, as described (Shermanet al. 1989).
Yeast methods: Yeast transformations were performed by the lithium acetate method (Itoet al. 1983) modified as described previously (Hirsch and Cross 1993). Yeast RNA was extracted from cells as described previously (Cross and Tinkelenberg 1991).
Northern blots: RNA was transferred to a nitrocellulose membrane after formaldehyde-agarose gel electrophoresis as described (Lehrachet al. 1977). The membranes were UV cross-linked using a Stratalinker UV box (Stratagene, La Jolla, CA). Prehybridization and hybridization were done at 65° in a buffer containing 0.9 m NaCl, 0.09 m sodium citrate, 0.1% Ficoll, 0.1% polyvinylpyrrolidone, 0.1% bovine serum albumin, 33 mm sodium pyrophosphate, 50 mm sodium phosphate monobasic. The probes used were gel-purified DNA restriction fragments 32P-labeled by random primer labeling using a Prime-It kit (Stratagene). The fragments used were: FUS1, a 1.4-kb EcoRI-HindIII fragment from plasmid pSL589 (McCaffreyet al. 1987); ribosomal protein gene TCM1 (Schultz and Friesen 1983), a 0.8-kb HpaI-SalI fragment from plasmid pAB309Δ; phosphoglycerate kinase gene PGK1, a 0.5-kb BamHI-XbaI fragment from pPGK1.
Mating assays: Quantitative mating assays were performed essentially as described previously (Guthrie and Fink 1991). Briefly, approximately 1.5 × 107 cells from log phase cultures of MATa strains were mixed with an equal number of MATα cells and filtered onto 0.45 mm nitrocellulose filters (Whatman, Kent, UK). The filters were placed onto YEPD plates and incubated for 5 hr at 30°, 34°, or 37°. Cells were resuspended in 4 ml of sterile water by vortexing, and the cell suspension was diluted and placed onto one type of selective plate to determine the number of MATa plus diploid cells and another type of selective plate to determine the number of diploid cells. Mating efficiency was calculated as the percentage of diploid cells divided by the number of MATa plus diploid cells.
Immunoblots: For immunoblotting, cells were grown to early log phase, pelleted, and washed once with TE (10 mm Tris-HCl and 1 mm EDTA). Cells were resuspended in synthetic medium and incubated at 30°, 34°, or 37° for 3 hr. For pheromone-treated samples, α-factor was added to a final concentration of 3 μmol and aliquots were removed at 1 and 6 hr after α-factor addition. Cell lysates were prepared by harvesting 10 ml of log phase cells, washing once with cold TE and resuspending in 350 μl of lysis buffer [50 mm TrisHCl (pH 8.0), 1% SDS, 1 mm PMSF, 1 μg of apoprotin, leupeptin, chymostatin, and pepstatin per ml]. The mixture was added to acid-washed glass beads (0.5 mm) and vortexed at high speed for 10 min. Glass beads and cell debris were separated from the lysate by centrifugation in a microfuge for 2 min. Protein concentration of the samples was determined using a bicinchoninic acid protein assay kit (Pierce Chemical, Rockford, IL) and equal amounts were loaded onto SDS polyacrylamide gels (10% polyacrylamide). Separated proteins were transferred to nitrocellulose and the blot was probed with anti-Ste4p rabbit polyclonal antiserum (Hirschmanet al. 1997) at a dilution of 1:1000. Donkey anti-rabbit immunoglobulin conjugated to horseradish peroxidase (Amersham, Arlington Heights, IL) was used at a dilution of 1:10,000 and immune complexes were detected with an enhanced chemiluminescence kit (Amersham).
Microscopy: For projection formation assays, cells were grown at 30° to log phase, pelleted, and washed once with 10 mm Tris pH 8.0, 1 mm EDTA. The cell pellet was resuspended in 5 ml selective medium to an OD600 of 0.3, and α-factor was added to a final concentration of 3 μmol for ste4ts cells or 6 μmol for ste4ts spa2 cells. After a 3-hr incubation, the cells were fixed for 1 hr with 3.7% formaldehyde. Cells were photographed with a 100× objective using differential interference microscopy on a Zeiss Axiophot microscope (Carl Zeiss, Inc., Thornwood, NY) using TMAX 3200 film.
For fluorescence microscopy, cells containing the Ssf1p-GFP fusion protein were grown at 30° and incubated in 2 μg/ml 4′,6-diamidine-2-phenylindole-dihydrochloride (DAPI) for 5 min at room temperature to stain DNA. Indirect immunofluorescence techniques were essentially as described (Pringleet al. 1989). For double-label immunofluorescence, cells were grown to early log phase and fixed by adding 0.5 ml of 37% formaldehyde directly to a 4.5-ml culture. After 30 min at room temperature, the cells were washed twice with 1 ml of 1.2 m sorbitol, 0.1 m phosphate buffer pH 7.5 (solution A). Cell walls were removed by resuspending the fixed cells in 1 ml solution A, adding 2 μl β-mercaptoethanol and 3 μl of 15 mg/ml zymolyase 20T (Seikagaku, Rockville, MD), and incubating the samples at 37° for 15 min. Spheroplasted cells were attached to polylysine-coated wells and further permeabilized by immersion in −20° methanol for 5 min. Cells were washed with PBS/BSA (PBS + 1% BSA) and incubated with a 1:1000 dilution of anti-GFP polyclonal antibody (kindly provided by J. Kahana and P. Silver, Dana-Farber Cancer Institute) and a 1:50 dilution of anti-Nop1p monoclonal antibody (Aris and Blobel 1988) for 2 hr. Cells were washed with PBS/BSA and then incubated with Texas Red-conjugated anti-mouse antibody (1:200) and FITC-conjugated anti-rabbit antibody (1:200) for 1 hr. In control experiments, the Texas Red-conjugated anti-mouse antibody showed no reactivity with the rabbit primary antibody and the FITC-conjugated anti-rabbit antibody showed no reactivity with the mouse primary antibody. Cells were incubated with DAPI (1 μg/ml) for 5 min at room temperature, then washed with PBS/BSA. Slides were mounted with 90% glycerol containing 1 mg/ml p-phenylenediamine and 25 mg/ml NaN3. Cells were photographed with a ×100 objective using FITC or UV filters on a Zeiss Axiophot microscope using TMAX 3200 film.
The SSF1 and SSF2 genes encode highly homologous proteins that appear to play roles in both mating and vegetative growth. To investigate their specific role in mating, the effect of depleting SSF gene products on different pheromone responses was determined. The two responses studied were pheromone-induced transcription, which is activated by the MAP kinase signaling cascade, and mating projection formation, which is activated by the morphology signaling complex.
Depletion of SSF gene products does not greatly impair FUS1 RNA induction: Activation of the pheromone response signal transduction pathway results in a number of changes in cellular physiology. One of these responses is transcriptional induction of genes such as FUS1, which are required for mating. To determine whether SSF gene products are necessary for this response, cells were depleted of Ssf proteins and assayed for pheromone-inducible transcription. A strain with the genotype Δssf1 Δssf2 carrying a plasmid with the SSF1 gene under the control of the GAL promoter was grown in galactose to log phase and then transferred to glucose for varying amounts of time. At each time point, an aliquot of cells was treated with α-factor and the amount of FUS1 RNA was determined. Induction of FUS1 RNA in cells that were depleted of SSF gene products for 4, 8, or 12 hr (Figure 1, lanes 3–5) appeared equal to that seen in the same cells before SSF depletion (Figure 1, lane 2). When the amount of signal was quantified by Phosphorimager analysis and normalized to the control PGK1 RNA, there was no significant decrease in the level of FUS1 RNA induction at any time point. Cells depleted of Ssf proteins for 12 hr display a ninefold decrease in mating efficiency (Yu and Hirsch 1995), indicating that some aspect of the pheromone response is defective. This decrease in mating efficiency, however, does not appear to be due to impaired pheromone-inducible transcription.
Depletion of SSF gene products affects mating projection formation: Exposure to pheromone produces a change in cellular morphology that results in the formation of a mating projection. This response appears to involve the Ste4p Gβ subunit, the Ste20p kinase, and other signaling components that are specific to projection formation. Ste4p and Ste20p are thus situated at a branch point in the signaling pathway that leads either to MAP kinase activation and transcriptional induction or to morphology complex activation and projection formation (Wittenberg and Reed 1996; Lebereret al. 1997). To determine whether SSF gene products are required for the morphological response, cells were depleted of Ssf proteins, treated with pheromone, and observed under the microscope to assess the degree of projection formation. The Δssf1 Δssf2 [pGAL-SSF1.14] strain grown in glucose for 12 hr was unable to form mating projections after exposure to α-factor for 3 hr (Figure 2A). When this strain was maintained in galactose and treated with α-factor, it was fully competent to form projections (Figure 2B). Cells lacking Ssf proteins continued to increase in size, indicating that they remained metabolically active. Some of these cells displayed a pointed tip, but none of them formed elongated projections. The mating defect of cells lacking Ssf proteins may therefore be a consequence of their failure to carry out the morphological response to pheromone.
ste4ts cells overexpressing SSF1 display a small increase in FUS1 RNA induction: Overexpression of SSF1 increases the mating efficiency of a strain containing a temperature-sensitive mutation in the STE4 gene (Yu and Hirsch 1995). It was therefore of interest to determine which aspect of the pheromone response is affected by increased levels of Ssf1p. To investigate the effect of SSF1 overexpression on pheromone-induced transcription, a ste4ts strain incubated at either the permissive or nonpermissive temperature was treated with α-factor and assayed for FUS1 RNA expression. At 30°, ste4ts cells carrying a multicopy SSF1 plasmid displayed a slight increase in FUS1 RNA abundance after treatment with α-factor compared to the same strain with vector alone (Figure 3, lanes 3 and 4). At 37°, there was a slightly larger difference between the FUS1 RNA level in cells with multicopy SSF1 compared to control cells (Figure 3, lanes 7 and 8). Quantification of the amount of FUS1 RNA normalized to the amount of control TCM1 RNA revealed that overexpression of SSF1 caused a 1.4-fold increase in FUS1 RNA induction at 30° and a 2.4-fold increase at 37°. This small increase in pheromone-inducible transcription conferred by SSF1 overexpression is unlikely to account completely for the 17-fold increase in mating efficiency seen in this strain (see Figure 5), suggesting that SSF1 overexpression affects another aspect of the pheromone response. Moreover, the observation that Ssf-depleted cells are unable to form mating projections suggests that the effect of SSF1 overexpression could also involve mating projection formation.
SSF1 overexpression promotes mating projection formation in a ste4ts strain: The effect of SSF1 overexpression on pheromone-induced morphological changes was evaluated in a ste4ts strain at different temperatures. Although this strain mates with fairly high efficiency at 30° and with extremely low efficiency at 37°, it does not form mating projections at either temperature. The ability of ste4ts cells to mate well at 30° though they do not form projections may be due to the fact that they retain an underlying cell polarity, which allows some localization of signaling complexes.
At 37°, overexpression of SSF1 had no effect on projection formation in ste4ts cells (data not shown). At 30°, however, ste4ts cells carrying a multicopy SSF1 plasmid displayed a significant increase in projection formation compared to control cells (Figure 4). Some cells overexpressing SSF1 displayed the characteristic morphology of normal projections (Figure 4, arrow), whereas other cells displayed the pear-shaped morphology seen in wild-type cells at early stages of the response (Figure 4, arrowhead). These observations indicate that SSF1 overexpression does not completely restore projection formation to normal levels. Based on the ability of SSF1 overexpression to promote projection formation at 30°, it is possible that excess Ssf1p is able to promote events that precede projection formation at 37°. These events could include assembly of signaling components at a particular site at the cell surface. Increasing the local concentration of such signaling components could contribute to the increased mating efficiency seen in cells overexpressing SSF1 at 37°.
SSF1 overexpression preferentially increases the mating efficiency of a ste4ts Δspa2 strain: The results presented thus far suggest that the SSF1 function affects mating projection formation to a greater degree than pheromone-induced transcription. To investigate this possibility further, the effect of SSF1 overexpression in cells containing mutations in genes that affect projection formation was determined. Projection formation is altered in different ways in cells containing mutations in the SPA2 and FAR1 genes. Whereas SPA2 null mutants are unable to form normal mating projections (Gehrung and Snyder 1990), cells containing carboxyl-terminal truncations of the FAR1 gene form projections that are not properly oriented toward the highest concentration of pheromone (Valtzet al. 1995).
Quantitative mating assays performed on a ste4ts strain incubated at the nonpermissive temperature of 37° showed that overexpression of SSF1 under these conditions increased mating efficiency by 17-fold as compared to the same cells containing vector alone (Figure 5). However, cells containing the ste4ts mutation and either a Δspa2 or far1-Δc allele displayed no detectable mating at 37° in the absence of SSF1 overexpression (data not shown), so it was not possible to obtain a value for the basal level of mating at this temperature. Therefore, it was necessary to assay the mating efficiency of these strains at a semipermissive temperature. At 34°, overexpression of SSF1 in a ste4ts strain caused only a fourfold increase in mating efficiency, presumably because this strain mates fairly well under these conditions even in the absence of multicopy SSF1. At this semipermissive temperature, the mating efficiency of the ste4ts Δspa2 and ste4ts far1-Δc strains was decreased ~1000-fold relative to the parent ste4ts strain (Figure 5). This finding suggests that the ste4ts allele causes cells to become more sensitive to morphological response defects, because the Δspa2 and far1-Δc mutations confer a modest 10-fold reduction in mating efficiency in a wild-type background (Gehrung and Snyder 1990; Chenevertet al. 1994; Doreret al. 1995; Valtzet al. 1995). Overexpression of SSF1 in the ste4ts far1-Δc strain caused a 17-fold increase in mating efficiency, the same relative increase seen in the parent strain at 37°. In contrast, overexpression of SSF1 in the ste4ts Δspa2 strain caused a 127-fold increase in mating efficiency. The finding that SSF1 has a greater effect on the mating efficiency of a Δspa2 strain indicates that excess Ssf1p compensates for the specific mating defect of this strain, which is the failure to produce mating projections.
SSF1 overexpression has an effect in the absence of SPA2: Because SPA2 null mutants do not form normal mating projections (Gehrung and Snyder 1990), it was of interest to determine whether overexpression of SSF1 could affect cell morphology in cells that lack Spa2p. ste4ts Δspa2 strains carrying either a multicopy SSF1 plasmid or control vector were treated with α-factor for 3 hr at 30° and observed under the microscope to assess morphological changes. In this genetic background, cells overexpressing SSF1 were more elongated than cells carrying vector alone (Figure 6). Moreover, cells overexpressing SSF1 frequently displayed a constricted region reminiscent of normal mating projections that was not seen in control cells. Therefore, suppression of the ste4ts Δspa2 strain mating defect by SSF1 overexpression is accompanied by an effect on projection formation. These results are consistent with the idea that the predominant effect of SSF1 overexpression is on cell morphology.
SSF1 overexpression in a ste4ts strain does not affect the abundance of Ste4p: The enhancement of mating projection formation seen in ste4ts strains that overexpress SSF1 is likely to involve proteins that are specific for the morphological response to pheromone. However, another possible explanation for the effect of SSF1 overexpression is that excess Ssf1p causes an increase in the abundance of the temperature-sensitive form of Ste4p. This possibility could be the case if a low level of Ste4pts is sufficient for the transcriptional response to pheromone, but a higher level is required for the morphological response. To test this idea, the abundance of Ste4pts was assayed by immunoblot using an anti-Ste4p antibody in strains carrying either a multicopy SSF1 plasmid or vector alone. A ste4ts strain incubated at either 37°, 34°, or 30° contained approximately equal amounts of Ste4pts in the presence or absence of SSF1 overexpression (Figure 7A, lanes 1–6). The temperature-sensitive form of Ste4p was somewhat difficult to detect because it was present at a much lower level than wild-type Ste4p (compare Figure 7A, lanes 5 and 6 with Figure 7B, lane 1). However, although there was some variation in the amount of Ste4pts observed, there was clearly no increase in abundance that correlated with overexpression of SSF1. In addition, there was no detection of a protein band corresponding to Ste4p in a strain containing a disruption of the STE4 gene (Figure 7B, lane 2), confirming that the species detected by the antibody in a ste4ts strain is Ste4pts. To test whether SSF1 overexpression affects the abundance of Ste4pts during the response to pheromone, ste4ts strains were incubated at 34°, treated with α-factor for 1 or 6 hr, and assayed for Ste4pts expression. Approximately equal amounts of Ste4pts were observed in pheromone-treated ste4ts cells carrying either a multicopy SSF1 plasmid or vector alone (Figure 7C, lanes 1–6). These results demonstrate that the effect of SSF1 overexpression on mating projection formation in a ste4ts strain is not due to stabilization of the Ste4pts protein, and are consistent with the idea that Ssf1p is specifically involved in the morphological response.
Ssf1p localizes to a nuclear compartment: The increased mating efficiency seen in ste4ts cells overexpressing SSF1 could be due to either a direct or indirect involvement in the morphological response to pheromone. One way to address this question is to determine whether the Ssf1p protein colocalizes with components of the signaling pathway. To identify the subcellular location of Ssf1p, the SSF1 gene was fused with the coding sequence for GFP (Chalfieet al. 1994). The SSF1-GFP fusion gene complemented the viability defect of a Δssf1 Δssf2 strain, demonstrating that the fusion gene retains SSF1 function (data not shown). Growing cells expressing the SSF1-GFP fusion were treated with DAPI to label nuclei and observed by fluorescence microscopy. The signal generated by Ssf1p-GFP was present as a small discrete focus in each cell (Figure 8A), which was clearly different from the general cytoplasmic localization seen in control cells expressing GFP alone (data not shown). The region of Ssf1p localization partially overlapped with the DAPI signal (Figure 8B), indicating that Ssf1p is present in a subcompartment of the nucleus. When viewed by confocal microscopy, the region of Ssf1p localization appeared as a crescent that caps one end of the nucleus (J. Kim, J. P. Hirsch and S. Kohlwein, unpublished data), suggesting that this subcompartment is the nucleolus.
To determine whether the localization of Ssf1p changes during mating projection formation, cells expressing the SSF1-GFP fusion were treated with α-factor for 3 hr before observing them by fluorescence microscopy. In cells that formed obvious projections, most of the Ssf1p-GFP signal did not localize to the projections but rather was concentrated in the main body of the cell (Figure 8D). However, essentially all cells that had projections displayed a more diffuse and expanded region of Ssf1p localization. Moreover, many cells that had Ssf1p concentrated in a small region outside the projection displayed a less intense area of signal that trailed off from the concentrated patch and extended into the projection (Figure 8D, arrow). This expanded area of localization was clearly different from the small discrete focus seen in cells not exposed to pheromone (compare Figure 8, A and D).
Ssf1p colocalizes with the nucleolar protein Nop1p: A double-label immunofluorescence experiment was performed to identify the precise subcellular location of Ssf1p. Cells expressing the SSF1-GFP fusion gene from a multicopy plasmid were fixed and processed for indirect immunofluorescence using an anti-GFP polyclonal and an anti-Nop1p monoclonal as the primary antibodies. Anti-Nop1p was chosen for the localization study because the region of Ssf1p-GPF staining in live cells was similar to that seen for proteins that are present in the nucleolus, and Nop1p is a yeast nucleolar protein (Aris and Blobel 1988). The staining patterns for Ssf1p-GFP and Nop1p were essentially identical and were present in a subregion of the area stained by DAPI (Figure 9, A–C). These results confirm the finding that Ssf1p is present in a subcompartment of the nucleus and identify that subcompartment as the nucleolus. Because proteins that participate directly in projection formation are found at the cell surface, the subcellular location of Ssf1p implies that its role in projection formation is not physically direct.
The mating function of the essential gene pair SSF1 and SSF2 was shown to act predominantly on mating projection formation based on the following observations. Depletion of SSF gene products eliminated the ability of cells to form projections and overexpression of SSF1 caused a significant increase in projection formation. Disruption or overexpression of SSF genes had little effect on pheromone-inducible transcription, indicating that the pheromone response pathway leading to transcriptional induction was signaling normally under these conditions. Moreover, SSF1 overexpression preferentially increased the mating efficiency of a strain containing a null allele of SPA2, a gene that functions specifically in cell morphology. Finally, overexpression of SSF1 enhanced projection formation in a ste4ts Δspa2 strain. These results are consistent with the idea that the function of Ssf1p and Ssf2p does not impinge on the MAP kinase cascade that leads to transcriptional induction, but rather affects the morphology signaling complex or the actin-based cytoskeleton. Similar effects are seen in cells that lack or overexpress BEM1, a gene that encodes a component of the morphology signaling complex. Cells containing mutant alleles of BEM1 do not form mating projections (Chenevertet al. 1992). BEM1 null mutants display a modest fivefold decrease in FUS1 induction and cells overexpressing BEM1 show a twofold increase in FUS1 induction (Lyonset al. 1996). The effect of BEM1 overexpression on FUS1 induction is thus similar to the effect of SSF1 overexpression, which also increases the level of FUS1 RNA by about twofold. Different alleles of genes that function mainly in cell morphology, such as BEM1, appear to have small effects on pheromone-inducible transcription. This observation suggests that a feedback mechanism exists that modulates transcriptional induction in response to ongoing morphological changes. One possible mechanism for this type of regulation is that an increase in clustering of morphological signaling components at a developing projection stimulates activation of the MAP kinase pathway due to the involvement of Ste20p in both processes.
Colocalization of Ssf1p-GFP with Nop1p demonstrated that Ssf1p is predominantly localized to the nucleolus. Examination of the Ssf1p protein sequence reveals a number of potential nuclear localization signals, such as the sequence KKQRKL at amino acids 399–404. The presence of such sequences within the Ssf1p protein is consistent with its experimentally determined location. Furthermore, the localization of Ssf1p to the nucleolus suggests that it is capable of binding nucleic acids, consistent with its pI of about 9.0 (Yu and Hirsch 1995).
The finding that Ssf1p is present in the nucleolus indicates that its role in the projection formation is likely to be physically indirect because the location of other components involved in this process is nonnuclear. For example, Cdc42p (Zimanet al. 1993), Ste20p (Peteret al. 1996; Lebereret al. 1997b), and Spa2p (Snyder 1989) are all localized to mating projections. Therefore, it seems likely that Ssf1p is involved in regulating the activity or amount of at least one protein that is a primary component of the morphology complex. This component cannot be Spa2p because the Ssf1p effect on morphology occurs in the absence of Spa2p. However, any of the other proteins that make up the morphology complex are candidates for the target of the Ssf1p function. All of these proteins, except Far1p, are involved in cell morphology during vegetative growth as well as during mating. The essential function of the SSF genes could therefore target the same protein as the mating function. Alternatively, it is possible that Ssf1p shuttles into and out of the nucleus and that its mating function is carried out in the cytoplasm or at the cell membrane. In that case, it would appear to be nuclear because its concentration in other cellular compartments would be too low to detect. This situation has been seen previously for proteins that shuttle between the nucleus and the cytoplasm (Piñol-Roma and Dreyfuss 1992).
The Ssf proteins are well conserved between yeast and plants (Yu and Hirsch 1995). In addition, a human homolog has recently been identified as a partial EST sequence (GenBank accession number N34073) that displays 41% identity to yeast Ssf1p over 116 amino acids. It is therefore likely that the human gene product represents the functional homolog of the yeast Ssf proteins. The regulation of cell morphogenesis is functionally conserved between yeast and mammalian proteins, consistent with the potential involvement of Ssf proteins in projection formation. Examples of other homologous proteins that play a role in this process include Cdc42 and Rho, members of the Ras-related GTPase family that display a high degree of sequence identity and are involved in polarized cell growth in both yeast and mammals (Chant and Stowers 1995). Although the involvement of SSF gene products in mating appears to be indirect, their conservation from yeast to mammals suggests that they function in a fundamental cellular process. Further studies of the defect conferred by the absence of these proteins should contribute to the elucidation of the Ssf function.
We thank M. Snyder, I. Karpichev, F. Cross and J. McKinney for providing plasmids used in this work, J. Hirschman and D. Jenness for providing anti-Ste4p antiserum, J. Kahana and P. Silver for providing anti-GFP antiserum, and John Hill for running the data base search that uncovered the human EST sequence with homology to the SSF genes. We also thank S. Kohlwein for confirming the nuclear localization of Ssf1p-GFP and S. Piñol-Roma for critical comments on the manuscript. This work was supported by National Institutes of Health grant GM-48808.
Communicating editor: D. Botstein
- Received October 13, 1997.
- Accepted December 22, 1997.
- Copyright © 1998 by the Genetics Society of America