INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Meiosis is required for the production of haploid gametes from diploid cells. A unique feature of meiosis is that two rounds of chromosome segregationmeiosis I and meiosis IIfollow DNA synthesis. In contrast, a single round of nuclear division follows each S phase in mitosis. Although many of the same genes are required for both mitosis and meiosis (Simchen, 1973), their activities must be modified to allow these distinct patterns of nuclear divisions to occur. The yeast gene SPO12 regulates the meiotic nuclear divisions and may play a role in mitosis as well. Mutations in SPO12 cause cells to undergo a single meiotic division resembling meiosis II (Klapholz and Esposito, 1980a,b). SPO12 mRNA is meiotically induced (Malavasic and Elder, 1990), peaking at the time of the meiotic nuclear divisions (Chu et al., 1998). SPO12 encodes a small protein of ~20 kDa (Malavasic and Elder, 1990) that is novel and has no homologies to proteins of known function.

SPO12 mRNA is also present in vegetative cells (Parkes and Johnston, 1992). During vegetative growth, SPO12 message is cell cycle regulated, peaking in M phase (Parkes and Johnston, 1992; Cho et al., 1998; Spellman et al., 1998). Although SPO12 is expressed in vegetative cells, deletion of SPO12 causes only a mild vegetative phenotype: mutants exhibit a slight G2/M delay (Parkes and Johnston, 1992). SPO12, however, does exhibit a number of intriguing genetic interactions with genes required for exit from mitosis. For example, deletion of SPO12 is synthetically lethal with deletion of DBF2 (Parkes and Johnston, 1992), which codes for a protein kinase believed to be required for mitotic exit (Toyn and Johnston, 1994). Based on genetic arguments, Toyn and Johnston (1993) hypothesized that Spo12p may be a regulatory subunit for the protein kinases Dbf2 and its relative Dbf20. SPO12 has also been isolated as a high-copy suppressor of a number of mutations affecting vegetative growth, e.g., cdc15, dbf2-2, cdc5-1, and tem1-3 (Parkes and Johnston, 1992; Shirayama et al., 1996; Jaspersen et al., 1998). All of these mutant strains arrest late in mitosis, with high Clb-Cdc28 kinase activity. These observations suggest that SPO12 can facilitate exit from mitosis when overexpressed, perhaps by antagonizing Clb-Cdc28 kinase activity. How SPO12 functions during mitosis and meiosis is not known. In this article, we describe studies on the level and localization of Spo12p during mitosis and meiosis, two-hybrid partners of Spo12p, high-copy suppression of spo12 meiotic defects, mutant interactions, and the effects of SPO12 on Clb-Cdc28 kinase activity during meiosis.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Media and Growth Conditions

Yeast were grown in YEPD (1% Bacto-yeast extract, 2% Bacto-peptone, 2% dextrose) (complete) or minimal (SD; 10.67% yeast nitrogen base without amino acids, 2% dextrose, 1× amino acids) medium, as described by Hicks and Herskowitz (1976). Cells were sporulated in liquid sporulation medium (0.3% potassium acetate and 0.02% raffinose; Kassir and Simchen, 1991) or on sporulation plates (1% potassium, 25 µg/ml zinc acetate, 2% washed agar, and complete amino acids as for SD). Sporulating cultures were synchronized according to the standard regimen described by Padmore et al. (1991) with the following modifications. To identify a clone of a particular strain with high sporulation efficiency, four colonies from the strain were grown in liquid YEPD and patched onto a sporulation plate. The sporulation efficiency of each patch was assessed microscopically after ~24 h. Overnight cultures corresponding to the most efficient sporulators were used to inoculate 200 ml of YPA (1% potassium acetate, 2% bactopeptone, 1% yeast extract) cultures (1:100 dilution). After ~15 h of growth in YPA, cells were collected by filtration, washed twice in 200 ml of double-distilled water, and resuspended in 200 ml of sporulation medium. Time zero in each time course is defined as the time of resuspension in sporulation medium.

Genetic and Molecular Biological Methods

Yeast genetic methods were used as described by Rose et al. (1990). DNA manipulations were performed as described by Sambrook et al. (1989)

Strains and Plasmids

Yeast strains used in this study are described in Table 1. pGal-SPO12, pGal-SIC1 (Jaspersen et al., 1998), pGPD-SPO12, and pGPD-SIC1 were all courtesy of S. Jaspersen (University of California, San Francisco).

Epitope Tagging of Spo12, Spo13, Sic1, and Clb4

Starting plasmids for each construct were the following: pRS306-SPO12, a 0.8-kilobase (kb) EcoRI fragment (from L. Johnston, National Institute for Medical Research, London, United Kingdom) cloned into pRS306; pRS306-SPO12(long), a 1.7-kb PvuII fragment (from S. Jaspersen) cloned into pRS306; pRS306-SPO13, a 2.6-kb BamHI/NarI fragment derived from BE806 (from R. Esposito, University of Chicago, Chicago, IL) cloned into pRS306; pRS306-SIC1, a 2.1-kb EcoRI/BglII genomic fragment (from S. Jaspersen) cloned into pRS306; and pRS306-CLB4, a 2.7-kb fragment cloned by PCR amplification with the use of oligonucleotides OMG66 (5'GAAAGCGGCCGCACCGGAATGACCTGGAACCG) and OMG67 (5'GAAAGAATTCGGATACCCAGCATATCAGG). This fragment was cloned into pRS306 with the use of the NotI and EcoRI sites introduced by PCR amplification. For all of the plasmids listed above, site-directed mutagenesis was used to introduce an in-frame BamHI (SPO13, SIC1, and CLB4) or BglII (SPO12) site. For N-terminal epitope tagging, the new site was introduced just downstream of the translational start. For C-terminal tagging of reading frames, the site was introduced immediately upstream of the translational stop codon. A BamHI fragment containing two copies of an 11-amino acid HA epitope (YPYDVPDYSAL; Wilson et al., 1984) or a multimerized Myc epitope (EQKLISEEDLN; Evan et al., 1985) was inserted into the introduced site.

Construction of CLB3HA

Starting with a pGal-CLB3HA clone constructed by S. Jaspersen, a BamHI/XbaI fragment containing the CLB3HA reading frame was cloned into pRS306. PCR was used to amplify a 1-kb 5' flanking sequence (with the use of oligonucleotides OMG60 [5'CCGGAAATTCACCACGTTTGACAC] and OMG61 [5'CCGGGGATCCAGTCCTGTGTTTAAG]) and a 1-kb 3' flanking sequence (with the use of oligonucleotides OMG62 [5' CCGGTCTAGAAAAGCCTCAGCTCGAGACAT] and OMG63 [5'GAAAGCGGCCGCGAACTAATCG-TCCTGTTACC]). These fragments were sequentially cloned upstream (5' flanking sequence) and downstream (3' flanking sequence) of the ORF in pRS306 with the use of the restriction sites introduced by PCR amplification. Finally, a 2.2-kb EcoRI/BglII fragment derived from genomic clone A5 (Fitch et al., 1992) was used to substitute for the equivalent fragment in the CLB3HA genomic clone; thus, only sequences 3' to the HA epitope tag were derived from PCR-amplified DNA.

Construction of CLB1HA in pRS306

A 2.5-kb fragment including the CLB1HA gene was PCR amplified with the use of oligonucleotides OMG55 (5'GAAAGAGCTCGATTTCCTCATTCGTCTTCC) and OMG56 (5'GAAAGTCGACGTCGAAGGAATTAGGAAGGC) from a strain containing CLB1HA (from M. Tyers, Mount Sinai Hospital, Toronto, Ontario, Canada). The resulting fragment was then cloned into pRS306 with the use of SpeI and SacI sites introduced by PCR amplification.

Disruption of SPO12

Starting with the spo12::TRP1 construct of Parkes and Johnston (1992; courtesy of L. Johnston), an XbaI/BamHI fragment containing the hisG-URA3-hisG cassette (Alani et al., 1987) was substituted for the XbaI/BamHI fragment of spo12::TRP1, which contained most of the TRP1 sequence, yielding spo12::hisG-URA3-hisG.

Disruption of BNS1

A 1.5-kb genomic fragment containing the BNS1 ORF and flanking sequences was PCR amplified from genomic DNA with the use of oligonucleotides OMG23 (5'CCGGGAATTCGTGCTTTAGGGCTTGGTGCC) and OMG24 (5'CCGGACTAGTTCGCCAGAGTTTACGAACGC) and cloned into pBluescript KS⁺ with the use of EcoRI and SpeI sites introduced in the oligonucleotides. Oligonucleotides OMG25 (5'GGCCGGATCCCGCAAATCTGTTACTTTTTTTTAGC-GC) and OMG26 (5'GGCCGGATCCTTTAACGGATGATGATCATAAATTGGC) were then used to amplify sequences 5' and 3' to the ORF as well as the pBluescript backbone. This fragment was then cut with BamHI (introduced in OMG25 and OMG26) and religated, resulting in a circularized plasmid in which a BamHI site was present in place of the BNS1 ORF. The resulting plasmid was then cut with BamHI, and the BamHI/BglII fragment of pNKY51 (Alani et al., 1987) containing the hisG-URA3-hisG cassette was inserted, resulting in the bns1::hisG-URA3-hisG construct.

Construction of pLexA-SPO12

A pRS306-SPO12 clone containing a BglII site introduced immediately downstream of the translational start (see above) was cut with Asp718 (filled in with the Klenow fragment of DNA polymerase) and BglII and cloned into pEG203 cut with XhoI (filled in with the Klenow fragment of DNA polymerase) and BamHI to yield pLexA-SPO12.

Immunofluorescence and DAPI Staining

Cells were processed for immunofluorescence as described by Sil and Herskowitz (1996). DAPI staining of sporulating cells was carried out by fixing cells in 70% ethanol or formaldehyde, rinsing once in 1× PBS, incubating for 12 min in 1 µg/ml DAPI, washing once in 1× PBS, and spotting on a slide for microscopic examination.

Cell Extracts and Immunoblot Analysis

Cell extracts were prepared for immunoblot analysis essentially as described by Sanders and Herskowitz (1996) with the following modifications. Samples were fractionated with 12.5% SDS-polyacrylamide gels and dry blotted with the use of a dry blot apparatus (E&K Scientific Products, Saratoga, CA) according to the manufacturer's instructions. Filters were blocked for 1-16 h in TBS containing 0.1% Triton X-100 and 2% nonfat dry milk and decorated with a 1:1000 dilution of HA11 or 12CA5 anti-HA antibodies (Babco, Richmond, CA). A 1:2000 dilution of goat anti-mouse immunoglobulin G heavy and light chain coupled to HRP (Bio-Rad, Richmond, CA) was used as the secondary antibody. Spo12-HA protein was visualized with the use of the Amersham (Arlington Heights, IL) ECL protein detection kit.

Synchronization of Cells with -Factor

A total of 200 ml of a SPO12-HA cells (YMG530) were grown to OD₆₀₀ = 0.3 at 30°C and arrested with 20 µg/ml -factor (Bio-Synthesis, Lewisville, TX) for 3-3.5 h. Cells were collected by filtration and washed twice with 200 ml of 30°C YEPD. Cells were resuspended in 200 ml of 30°C YEPD and allowed to reenter the cell cycle. Samples (10 ml) were taken every 10 min. One milliliter was fixed for microscopic examination of budding index, and the remainder was processed for immunoblot analysis.

Two-Hybrid Analysis

Two-hybrid analysis was performed as described by Gyuris et al. (1993). Yeast strain EGY48 containing the reporter plasmid pSH18-34 was cotransformed with pLexA-SPO12 and a library of Saccharomyces cerevisiae genomic plasmids in a pJG4-5 vector containing transcriptional activation domain fusions (courtesy of R. Brent, Massachusetts General Hospital, Boston, MA). Of 385,000 independent transformants screened, 886 colonies were identified that allowed growth on SD-LEU as a result of the activation of the LexAop::LEU2 reporter construct. These were colony purified and tested for the ability to activate Gal-inducible transcription of two reporter constructs (LexAop::LEU2 and pSH18-34; Gyuris et al., 1993) in the presence of pLEXA-SPO12. Fourteen clones were able to do so. The specificity of the interaction of these clones with pLexA-SPO12 was tested by analyzing the activation potential of these clones when coexpressed with an unrelated LexA DNA-binding domain fusion (pRFHM1; Gyuris et al., 1993). DNA sequencing of plasmids was done with the use of an Applied Biosystems (Foster City, CA) sequencing machine (Biomolecular Resource Center facility, University of California, San Francisco).

High-Copy Suppression Analysis

A library of S. cerevisiae genomic plasmids in the 2-µm vector YEp13 (Nasmyth and Reed, 1980) was transformed into a strain homozygous for spo12::hisG-URA3-hisG. Individual transformants were patched onto selective medium to select for the library plasmid and replica plated to sporulation medium. Cells from each patch of sporulating cells were microscopically examined for tetrads. Approximately 8,000 independent transformants were analyzed. Plasmids with suppressing activity were rescued from yeast and retransformed into a spo12 deletion mutant. All five positive clones were tested for the presence of SPO12 or BNS1 by PCR. For the two clones that did not contain SPO12 or BNS1, end sequencing was performed to map the genomic inserts. The inserts of these clones were found to overlap in a small region containing just two complete ORFs, SPO13 and ARD1. SPO12, BNS1, and SPO13 were each subcloned into pRS425 and tested for the ability to suppress spo12 or spo13 mutations.

Histone H1 Kinase Assays from Sporulating Cultures

Ten-milliliter aliquots were collected from 200-ml cultures of cells undergoing synchronous meiosis at 30- or 60-min intervals after resuspension in sporulation medium. One milliliter was fixed for DAPI analysis, and the remainder was processed for kinase assays as described below. Cells were collected by centrifugation, washed once with water, and frozen in liquid nitrogen until processing. Pellets were then allowed to thaw on ice and were resuspended in 500 µl of IP buffer (20 mM Tris, pH 7.4, 100 mM NaCl, 50 mM NaF, 50 mM -glycerophosphate, 5 mM EDTA, 0.2% Triton X-100, 1 mM PMSF, 2 µg/ml aprotinin, 2 mM benzamidine, 0.1 mM Na₃VO₄). Cells were lysed by bead beating twice for 1 min in a Biospec Products (Bartlesville, OK) mini bead beater according to the manufacturer's instructions. Supernatant was collected and centrifuged for 10 min at 4°C. Supernatant was then transferred to a fresh tube, and protein concentration was determined by the Bradford assay (Bio-Rad). ClbHA-Cdc28p kinase complexes were immunoprecipitated from 500 µg of total protein per sample with 1 µl of 12CA5 antibody (Babco) and 40 µl of a 1:1 slurry of protein A-CL4B Sepharose (Pharmacia, Piscataway, NJ) equilibrated in IP buffer by rocking for 1-3 h at 4°C. Immunoprecipitates were collected by centrifugation and washed three times in IP buffer. Immunoprecipitates were then washed once in 50 mM HEPES, pH 7.4, 1 mM DTT. Excess liquid was removed, and immunoprecipitates were resuspended in 20 µl of kinase assay reaction mixture (50 mM HEPES, pH 7.4, 2 mM MgCl₂, 100 µM ATP, 1 mM DTT, 5 µg of histone HI [Fluka Chemical, Buchs, Switzerland], 2.5 µCi [³²P]ATP) and incubated at room temperature for 20 min. Reactions were stopped with 10 µl of 4× sample buffer. Samples were boiled for 10 min and fractionated on a 12.5% SDS-polyacrylamide gel. Gels were fixed for 20 min in destain (25% methanol, 7.5% glacial acetic acid), dried, and exposed for ~16 h.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Localization and Levels of Spo12p during Meiosis

To characterize Spo12p during meiosis, we constructed both Myc and HA epitope-tagged versions of Spo12p, which were used for immunofluorescence and immunoblotting, respectively. Both epitope-tagged versions of Spo12p were able to complement the meiotic defect of a homozygous spo12 deletion mutant when integrated in single copy (our unpublished results).

Immunofluorescence was used to determine the cellular localization of Spo12p during meiosis. Progress through meiosis in these experiments was monitored by DAPI staining. By immunolocalization, we saw Spo12p staining predominantly in cells that were undergoing meiosis I or meiosis II (Figure 1A). Cells with small, round nuclei that exhibited Spo12p staining were occasionally observed (data not shown). It was not possible to determine by immunofluorescence whether these were sporulating cells or vegetative cells that had not yet entered the sporulation program (Spo12p is also expressed during the mitotic cell cycle; see below). In all cases, Spo12p colocalized with DAPI staining, indicating that it is predominantly if not exclusively nuclear. By the time the meiotic nuclear divisions were complete, Spo12p was no longer detectable.

View larger version (42K): [in this window] [in a new window]   Figure 1.   (A) Localization of Spo12p in meiotic cells. Cells expressing Spo12-(Myc)9 were sporulated and costained with the anti-Myc mAb 9E10 and DAPI to visualize nuclei. Spo12-(Myc)9 protein colocalized with DAPI staining and was visible only in cells undergoing meiotic divisions. (B) Time course of expression of Spo12p during meiosis. Cells homozygous for a gene replacement of Spo12-(HA)2 (YMG548) were induced to undergo synchronous meiosis. Aliquots were removed at hourly intervals and processed for immunoblot analysis and DAPI staining. The same amount of total protein was loaded per lane. Spo12-(HA)2 protein is indicated by an arrow. A background band present in all samples (including in untagged strains; data not shown) is indicated by an asterisk. The YPA sample was collected from cells grown overnight in YPA medium immediately before resuspension in sporulation medium. The 0-h time point was taken immediately after cells were resuspended in sporulation medium. Progress through meiosis was monitored by DAPI staining. % mononucleate indicates the percentage of cells that had not entered the meiotic divisions; % binucleate indicates the percentage of cells that were actively undergoing meiosis I or had completed meiosis I but had not entered meiosis II; % tetranucleate indicates the percentage of cells that were undergoing or had completed meiosis II. (C) Localization of Spo13p during meiosis. Cells expressing Spo13-(Myc)3 were costained with the anti-Myc mAb 9E10 and DAPI. Arrows point to the nuclei of three mononucleate cells in which Spo13p colocalized with DAPI staining. The arrowhead shows the nucleus of a cell undergoing meiosis II in which Spo13p is not detectable.

We used immunoblot analysis to determine the time course of Spo12p expression during meiosis (Figure 1B). For these experiments, we used the SKI strain background, which progresses through meiosis relatively synchronously (Padmore et al., 1991). Progress through meiosis was monitored by DAPI staining. Spo12p first became detectable as cells started to enter the meiotic divisions (~5 h) and was most prominent when the cells were in meiosis I or meiosis II. Once meiosis was complete, Spo12p was no longer detectable. The lack of Spo12p at the beginning of the meiotic time course and in cells arrested in YPA medium suggests that the cells observed by immunofluorescence with small round nuclei that contained Spo12p are likely to be residual vegetative cells in the sporulating culture.

Localization and Presence of Spo12p during Vegetative Growth

Although the phenotype of spo12 mutants is most striking in sporulating cells, SPO12 does have a role in vegetative growth; spo12 mutants exhibit a slight G2/M delay (Parkes and Johnston, 1992). SPO12 mRNA is present in vegetative cells, and its message is one of many genes known to exhibit cell cycle periodicity, peaking in M phase (Cho et al., 1998; Spellman et al., 1998). Using the same epitope-tagged constructs as in the meiotic studies, we determined the localization of Spo12p during vegetative growth and its expression throughout the mitotic cell cycle (Figure 2). As in meiosis, Spo12p was predominantly nuclear during vegetative growth. Staining was most prominent in large-budded cells that were going through mitosis, although staining was also observed in unbudded cells and in cells with small buds.

View larger version (67K): [in this window] [in a new window]   Figure 2.   (A) Localization of Spo12p in vegetative cells. An asynchronous population of cells expressing Spo12-(Myc)9 (strain YMG303) was costained with the anti-Myc mAb 9E10 and DAPI. The same field is shown with anti-Myc staining, DAPI staining, and a phase image. Spo12p protein was detectable in the nuclei of most but not all cells. The arrowheads point to a large-budded cell that exhibits Spo12p staining. (B and C) Time course of expression of Spo12p in vegetatively growing cells. Cells expressing Spo12-(HA)2 (YMG530) were arrested with -factor. After release from -factor arrest, samples were taken at 10-min intervals. Spo12-(HA)2 protein is indicated by an arrow. A background band present in all samples is indicated by an asterisk. Synchrony was monitored by counting budding index (B). In C, the same amount of total protein was loaded per lane. unt, untagged control; asy, sample from the asynchronous population of cells before -factor arrest; arr, cells arrested in -factor.

To obtain a more precise profile of the level of Spo12p throughout the mitotic cell cycle, we used -factor to synchronize a population of cells containing epitope-tagged Spo12p in G1. The cells were then released from -factor arrest, and aliquots were taken every 10 min and analyzed for Spo12p by immunoblotting. Progress through the cell cycle was monitored by scoring budding index (Figure 2B). We found that Spo12p was expressed with some periodicity, peaking as cells were in G2/M and at its lowest levels during G1 and S (Figure 2C). The pattern of protein expression parallels the cell cycle regulation of SPO12 mRNA observed by Parkes and Johnston (1992). In addition, Spo12p levels are at their highest at the same point in the cell cycle that seems to be affected in the spo12 deletion mutant, i.e., G2/M.

Two-Hybrid Analysis

Two-hybrid analysis was used to identify proteins that may physically interact with Spo12p. We constructed an in-frame fusion of the entire SPO12 ORF to the LexA DNA-binding domain and demonstrated that this construct (pLexA-SPO12) had no intrinsic transcriptional activity when transformed into strains containing reporter constructs regulated by LexA DNA-binding sites (not shown). To identify proteins that can interact with LexA-Spo12p to activate transcription, a library of S. cerevisiae genomic clones in the pJG4-5 vector (courtesy of R. Brent) was transformed into EGY48, which contained pLexA-SPO12 and appropriate reporter constructs. Because a genomic library was used in this study, in principle it should have been possible to identify both meiotic and vegetative partners of Spo12p. Of 385,000 independent transformants, 14 were able to activate transcription of two independent reporter constructs in the presence of pLexA-SPO12. These same clones were unable to induce transcription when coexpressed with an unrelated LexA DNA-binding domain fusion (pRFHM1; Gyuris et al., 1993), indicating that the interaction with LexA-SPO12 was specific. All 14 clones were sequenced to determine the nature of the library activation domain fusion proteins.

Three of the 14 clones contained very short fusion protein products that did not correspond to identified ORFs in the S. cerevisiae genome. The remaining 11 corresponded to novel or known genes (Table 2). Ten of the clones were isolated only once in the screen; the plasmid containing the fusion to YDR267C was isolated twice. Three of the genes identified in the two-hybrid screen are induced during meiosis (Chu et al., 1998). They are MEC3, which plays a role in a DNA damage checkpoint gene in mitotic cells (Weinert et al., 1994); YLR132C, a novel ORF; and FUR1, which encodes uracil phosphoribosyltransferase (Kern et al., 1990).

High-Copy Plasmid Suppression of a spo12 Mutation during Meiosis

To identify genes that can function in place of or downstream of SPO12 during meiosis, we carried out a screen for high-copy plasmids that could compensate for a complete deletion of SPO12. A library of S. cerevisiae genomic clones in the 2-µm vector YEp13 (Nasmyth and Reed, 1980) was transformed into a strain homozygous for a spo12 deletion (YMG270). Individual transformants were patched onto selective medium to select for the library plasmid and then replica plated to sporulation medium. Cells from each patch of sporulating cells were microscopically examined for the presence of tetrads, which indicated restoration of SPO12-like activity. Of ~8000 independent transformants, 7 clones allowed the spo12 deletion strain to form some tetrads (9-80% of total asci; Table 3). Homozygous spo12 mutants carrying YEp13 formed virtually no tetrads (<0.3%; Table 3). Plasmids with suppressing activity were rescued from yeast and retransformed into the spo12 deletion strain. Of these, five retained the ability to suppress the spo12 deletion at least partially. These clones were analyzed further to determine the sequences responsible for the suppressing activity.

View this table: [in this window] [in a new window]   Table 3.  High-copy suppression of spo12/spo12 and spo13/spo13 strains

The two clones with the strongest ability to suppress the deletion phenotype, 21-36 and 47-32, encoded SPO12 itself (Table 3). A third clone, 32-1, contained YGR230W, an ORF with homology to SPO12 (Figure 3). We have named this ORF BNS1 for Bypasses Need for Spo12p. In addition, two independent clones contained SPO13, another gene required for progression through the meiotic nuclear divisions (see below). Each of these three ORFs (SPO12, BNS1, and SPO13) was subcloned into another high-copy vector (pRS425) to determine whether they were responsible for the complementing activity. Suppression was observed in each case, but the magnitude of suppression was 1.5- to 3-fold lower than that of the original library clone. The reduced suppression may have been due to a slightly lower copy number of clones in pRS425 than in YEp13 or to the presence of additional ORFs on the original plasmids that contribute to the suppression, although this seems unlikely. In any case, it is clear that overexpression of SPO12, SPO13, and BNS1 allows a homozygous spo12 deletion strain to sporulate to form some tetrads instead of dyads.

View larger version (28K): [in this window] [in a new window]   Figure 3.   (A) Homology between Spo12 and Bns1. Lines indicate identical amino acids. Numbers correspond to the amino acid positions in each protein. (B) Domain of homology shared by Spo12 and other ORFs. Amino acids identical to those found in Spo12 are shaded black. Those in common among other ORFs but not in common with SPO12 are shaded gray. Numbers correspond to amino acid positions in each protein where known.

In an attempt to understand how BNS1 and SPO13 can partially compensate for the lack of Spo12p, we characterized these ORFs further. BNS1 mRNA is not induced during meiosis (Chu et al., 1998). BNS1 is predicted to encode a 137-amino acid protein of 15.8 kDa that is homologous to Spo12p (40% identical over 80 amino acids; Figure 3A). Like Spo12p, the protein encoded by BNS1 is very basic (predicted isoelectric point of 11.22 compared with an isoelectric point of 10.19 for Spo12p). Neither Spo12p nor Bns1p contains any recognizable structural motifs, nor are they similar to Spo13p. The highest degree of identity between Spo12p and Bns1p is a segment of 19 amino acids that is also homologous to at least three other hypothetical ORFs, one from Schizosaccharomyces pombe, one from Caenorhabditis elegans, and, intriguingly, an expressed sequence tag isolated from a human testis cDNA library (Figure 3B). The functional significance of this homology is not known.

To determine whether BNS1 plays a role during meiosis or vegetative growth, we replaced its coding sequence with the hisG-URA3-hisG cassette (Alani et al., 1987). Strains homozygous for the bns1 deletion were fully viable and sporulated as efficiently as wild-type strains. We also determined the phenotype of a spo12 bns1 double mutant. These strains were also viable and did not display any obvious vegetative growth defects. Furthermore, when sporulated, the spo12 bns1 double mutant formed dyads with the same efficiency as a spo12 single mutant. Thus, although BNS1 can partially compensate for a lack of SPO12 function when overexpressed, it does not appear to play any role in controlling meiotic nuclear divisions.

Analysis of Spo13p

SPO13, like SPO12, is clearly important for programming the meiotic nuclear divisions. A homozygous spo13 deletion mutant sporulates to form dyads with high efficiency, and as in spo12, the single meiotic division observed in spo13 most closely resembles meiosis II (Klapholz and Esposito, 1980a,b). To gain some additional information regarding when and where Spo13p might act, we epitope tagged the SPO13 ORF with the Myc epitope. This protein was able to complement the meiotic defect of a homozygous spo13 deletion mutant when integrated at single copy (our unpublished results). We found that like Spo12p, Spo13p is also predominantly nuclear (Figure 1C). However, Spo13p was detectable only before meiosis I and not during meiotic nuclear divisions. Spo13p was seen in mononucleate cells with large nuclei with diffuse DAPI staining. We hypothesize that these correspond to cells immediately before meiosis I, probably in the late stages of or after the completion of premeiotic DNA synthesis. Because we were unable to detect Spo13p in vegetative cultures, it is unlikely that the Spo13p-staining cells in sporulating cultures were contaminating vegetative cells. The temporal profiles of Spo13p and Spo12p during meiosis suggest that SPO13 acts upstream of SPO12.

Because of the similarity between Spo12p and Bns1p, we tested the hypothesis that SPO13 was able to high-copy suppress spo12 mutants by activating BNS1. If this were true, overexpression of SPO13 should not be able to suppress a spo12 bns1 double mutant. High-copy SPO13, however, was able to suppress the double mutant as efficiently as it suppressed spo12 single mutants (our unpublished results).

Because high-copy SPO13 was able to partially suppress the phenotype of a spo12 deletion mutant, we tested whether the reciprocal was also true, i.e., whether overexpression of SPO12 could suppress the phenotype of a spo13 mutant. High-copy plasmids encoding SPO12 or BNS1 were transformed into a homozygous spo13 deletion mutant, and the strains were sporulated (Table 3). Empty-vector and high-copy SPO13 plasmids were used as controls. High-copy SPO13 was, as predicted, the strongest suppressor. High-copy SPO12 could weakly suppress the spo13 mutant phenotype. Although the total number of tetrads seen in high-copy suppression of the spo13 mutant by SPO12 was comparable to that seen by high-copy suppression of spo12 by SPO13, the magnitude of the suppression was quite different (~3-fold versus ~37-fold; Table 3) as a result of differences in the background number of tetrads observed in spo12 versus spo13 mutants. In spo12 mutants, one virtually never sees any tetrads formed, whereas in spo13 mutants, we observed a background level of tetrads that could reach as high as ~8%. High-copy BNS1 did not cause statistically significant suppression of a spo13 deletion. Thus, overexpression of BNS1 can partially bypass the need for SPO12 but not SPO13 function.

Genetic Interactions of spo12 with hct1 and sic1

Overexpression of SPO12 can suppress the late mitotic arrest phenotype of a number of mutants that arrest with high Clb-Cdc28 kinase activity (Parkes and Johnston, 1992; Shirayama et al., 1996; Jaspersen et al., 1998), suggesting that SPO12 may play a role in exit from mitosis. To determine whether Spo12p may act to decrease Clb-Cdc28 kinase activity, we tested whether deletion of SPO12 was synthetically lethal with hct1 or sic1 mutants. HCT1 encodes a substrate-specific regulator of the anaphase-promoting complex (APC) and is required for degradation of Clb2p during exit from mitosis (Schwab et al., 1997; Visintin et al., 1997). SIC1 encodes a cyclin-dependent kinase inhibitor (CKI) that is expressed from late mitosis into G1 (Donovan et al., 1994). Although a strain containing deletion of HCT1 or SIC1 alone is viable, hct1 sic1 double mutants are inviable. This result has been interpreted to mean that as long as at least one of the two pathways for inactivating Clb-Cdc28 kinase activity is intact (through proteolysis mediated by Hct1p or by inhibition of Clb-Cdc28 by Sic1p), cells are able to progress through the cell cycle relatively normally. Inactivation of both pathways simultaneously, however, causes cells to arrest with large buds, presumably with high Clb-Cdc28 kinase activity (Schwab et al., 1997). We found that whereas spo12 sic1 double mutants were viable, spo12 hct1 double mutants were inviable (Table 4). Thus, if Spo12p inhibits Clb-Cdc28 kinase activity, it does so in an APC-independent manner and may facilitate the activity of a CKI, presumably Sic1p.

To investigate this hypothesis further, we found by immunoblot analysis that Spo12p and Sic1p show similar levels throughout the mitotic cell cycle (data not shown). We also tested whether the ability of SPO12 to suppress cdc15-2 when overexpressed (Jaspersen et al., 1998) was dependent on SIC1 function. Although overexpression of SPO12 did suppress the growth defect of a cdc15-2 mutant, it did not suppress the defect of a cdc15-2 sic1 double mutant (Figure 4A). In addition, the ability of SIC1 to suppress cdc15 when overexpressed (Mendenhall, 1993; Schwob et al., 1994; Toyn et al., 1997; Jaspersen et al., 1998) was dependent on having an intact copy of SPO12 in the genome (Figure 4B). These results suggest that Spo12p and Sic1p may act together in vegetative cells to facilitate exit from mitosis.

View larger version (80K): [in this window] [in a new window]   Figure 4.   (A) High-copy suppression of cdc15-2 but not cdc15-2 sic1-1::HIS3 by pGAL-SPO12. Strains of the genotypes sic1-1::HIS3, cdc15-2, and cdc15-2 sic1-1::HIS3 were transformed with pGAL-SPO12 or the vector control pRD53. They were then streaked for single colonies on SD medium or SGAL (0.67% yeast nitrogen base without amino acids, 1× amino acids, 2% galactose, 2% Bacto-agar) medium (to induce expression of GAL-SPO12), and plates were incubated at 25 or 37°C (the nonpermissive temperature for cdc15-2). Strains were YMG422, YMG484, and YMG485. (B) Partial high-copy suppression of cdc15-2 but not cdc15-2 spo12 by pGPD-SIC1. Strains of the genotypes cdc15-2, spo12::hisG-URA3-hisG, and cdc15-2 spo12::hisG-URA3-hisG were transformed with pGPD-SIC1 (the GPD promoter is constitutively active) or a pGPD vector control, streaked for single colonies, and grown at 25 or 37°C to test the ability of pGPD-SIC1 to restore growth to strains containing the cdc15-2 mutation. Strains were YMG536, YMG537, and YMG535.

Effect of Spo12p on Clb-Cdc28 Kinase Activity during Meiosis

We examined the levels of Sic1p during meiosis to test the hypothesis that Spo12p might interact with Sic1p during meiosis to modulate Clb-Cdc28 kinase activity and thus control progress through the meiotic nuclear divisions. We were unable to detect any Sic1p in meiotic cells, although high levels of Sic1p were found in cells arrested in YPA presporulation medium (Dirick et al., 1998; our unpublished results). Thus, if Spo12p interacts with a CKI during meiosis, it is unlikely to be Sic1p.

To test more directly the hypothesis that Spo12p modulates Clb-Cdc28 kinase activity during meiosis, we followed levels of kinase activity during meiosis in the wild type and in spo12 mutants (Figure 5). For this experiment, we constructed HA epitope-tagged versions of the five Clb proteins expressed during meiosisClb1p, Clb3p, Clb4p, Clb5p, and Clb6p (Grandin and Reed, 1993; Chu and Herskowitz, 1998; Stuart and Wittenberg, 1998)and introduced them individually as homozygous gene replacements into wild-type SK1 cells or homozygous spo12 deletion mutants in the SK1 strain background. These strains were then sporulated synchronously, and aliquots were taken at intervals of 30-60 min over a period of 9 h, at which point most of the cells had completed the meiotic divisions. Progress through meiosis was monitored by DAPI staining (Table 5). Clb-associated kinases were immunoprecipitated from the same amount of total protein from each culture, and in vitro kinase assays were performed with histone H1 as substrate. We were unable to detect any H1-associated kinase activity in wild-type or spo12 mutant strains containing Clb6-HA (data not shown). For the remaining strains, no significant differences between the wild type and the spo12 mutants in the timing or the level of activation of the specific Clb-Cdc28 kinases were detected (Figure 5).

View larger version (45K): [in this window] [in a new window]   Figure 5.   Histone H1 kinase activity in strains undergoing synchronous sporulation. Wild-type and spo12 strains homozygous for gene replacements of CLB1HA, CLB3HA, CLB4HA, and CLB5HA (CLB1/CLB1 [A], CLB1/CLB1 spo12/spo12 [B], CLB3/CLB3 [C],CLB3/CLB3 spo12/spo12 [D], CLB4/CLB4 [E], CLB4/CLB4 spo12/spo12 [F], CLB5/CLB5 [G], and CLB5/CLB5 spo12/spo12 [H]) were induced to undergo synchronous meiosis. Aliquots were collected at hourly or half-hourly intervals as indicated. Active Clb-Cdc28p kinase complexes were immunoprecipitated from the same amount of total protein in each lane with the use of anti-HA antibodies. Kinase activity of these complexes was measured by an in vitro kinase assay with histone H1 as a substrate. Progress through meiosis was monitored by DAPI staining (Table 5). In each case, the onset of histone H1 kinase activity corresponded to the onset of the meiotic divisions. Strains were YMG654, YMG655, YMG624, YMG659, YMG522, YMG632, YMG998, and YMG999.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Spo12p plays an important role in meiosis and appears to have a role in the mitotic cell cycle as well. We have made several observations on SPO12 that bear on its meiotic and mitotic roles.

Presence and Localization of Spo12p

We found that Spo12p is localized to the nucleus during meiosis and mitosis. Spo12p is present during the meiotic nuclear divisions but is not detectable by immunoblot or immunofluorescence at other times, making it unlikely that it is present but delocalized in these cells. SPO12 mRNA levels have a similar profile during meiosis (Malavasic and Elder, 1990; Chu et al., 1998). Immunoblot analysis of a synchronized vegetative cell population showed that Spo12p is present throughout the mitotic cell cycle and peaks in G2/M. SPO12 mRNA levels have also been shown to peak at this stage of the cell cycle (Parkes and Johnston, 1992; Cho et al., 1998; Spellman et al., 1998).

Proteins That Interact with Spo12p

Candidate proteins that interact with Spo12p were identified with the use of the two-hybrid system. At least three of the genes identified in this screen are induced to some level during meiosis (Chu et al., 1998) and clearly warrant further study. They are YLR132C, MEC3, and FUR1. YLR132C is a novel ORF with no known homologies. Its transcript exhibits an expression pattern during sporulation that is similar to that of SPO12 (Chu et al., 1998), suggesting that their proteins may be present at the same time during sporulation and have the opportunity to interact physically.

MEC3 is required for the DNA damage checkpoint response in mitosis and is probably required for the meiotic recombination checkpoint (Weinert et al. 1994; Lydall et al., 1996; Lydall and Weinert 1997). Transcription of MEC3 is induced twofold when cells enter the meiotic divisions (Chu et al., 1998). Perhaps interaction between Spo12p and Mec3p signals a meiotic cell that recombination is complete and that the cell can execute meiosis I.

FUR1 encodes uracil phosphoribosyltransferase, which produces UMP from uracil (Kern et al., 1990). Transcription of FUR1 is induced when cells enter meiotic divisions (Chu et al., 1998). Although a link between Spo12p and pyrimidine metabolism is not obvious, FUR1 is notable because of a previous functional connection between FUR1 and SPO12: overexpression of either gene can suppress mitotic arrest resulting from overexpression of stabilized Clb2 (Glotzer, 1993). Overexpression of SPO12 or FUR1 suppresses point mutations but not deletions in the CLB2 destruction box.

The other candidate partners for Spo12p might function with Spo12p during vegetative growth. Whether any of these proteins physically interact with Spo12p remains to be determined

Overexpression of BNS1 and SPO13 Can Partially Suppress the spo12 Meiotic Phenotype

We used high-copy suppression of a spo12 deletion to identify proteins that can drive the two meiotic nuclear divisions in the absence of Spo12p. The rationale was that these proteins might be downstream targets of Spo12p or might bypass the requirement of Spo12p for the meiotic divisions. We found two genes, BNS1 and SPO13, that could partially compensate for a complete lack of Spo12p during meiosis.

BNS1 is the only ORF in the yeast genome with significant homology to SPO12. Their proteins are 40% identical over ~80 amino acids, with one segment containing 14 of 19 identical residues and 4 of 19 conservative changes. Proteins containing this segment are also found in a number of other organisms, including S. pombe, C. elegans, and humans. The degree of similarity with Spo12p does not extend significantly beyond this region. In at least one other circumstance, BNS1 can function like SPO12: both BNS1 and SPO12 can suppress the arrest phenotype of cdc15-2 mutants when overexpressed (Jaspersen et al., 1998). Despite the similarity between SPO12 and BNS1, BNS1 does not appear to have a role in meiosis: BNS1 mRNA is not induced during meiosis (Chu et al., 1998), null mutants for bns1 exhibited no meiotic phenotype, and spo12 bns1 double mutants were viable with no obvious growth defects. Spo12p protein has been proposed to be a cofactor for the protein kinase Dbf2p (Toyn and Johnston, 1993). Perhaps Bns1p acts to facilitate the activity of the Dbf2-related protein kinase Dbf20p.

Although both Spo12p and Spo13p are nuclear, Spo13p is present before Spo12p. How then can suppression of spo12 by SPO13 be explained? Overexpression of SPO13 during meiosis has been reported to delay meiosis I (McCarroll and Esposito, 1994). spo12 mutants skip meiosis I and execute only meiosis II (Klapholz and Esposito, 1980a,b). Perhaps overexpression of SPO13 in spo12 mutants delays meiotic divisions long enough that a compensatory mechanism can trigger meiosis I. Another possibility is that overexpression of SPO13 in spo12 mutants somehow promotes a second round of nuclear division that follows the meiosis II-like division that normally occurs in spo12 mutants. Assessment of the patterns of chromosome segregation or spore viability under conditions of suppression might make it possible to distinguish between these two possibilities.

We have also observed that SPO12 can partially suppress spo13 mutants when overexpressed. This suppression can be explained if SPO12 acts downstream of SPO13, as is indicated by their temporal patterns of expression. The magnitude of suppression might be low because Spo13p activates targets other than Spo12p for efficient sporulation.

Synthetic Lethality of spo12 and hct1 Further Supports a Role for SPO12 in Exit from Mitosis

A critical event for exit from mitosis is the inactivation of Clb-Cdc28 kinase activity. Although the phenotype of spo12 mutants in vegetative cells is subtle, overexpression of SPO12 can suppress late mitotic arrest of several different mutants. All of these mutants arrest with high Clb-Cdc28 kinase activity, leading to the hypothesis that Spo12p inhibits Clb-Cdc28 kinase activity in some way (Shirayama et al., 1996). The fact that Spo12p is expressed in late mitosis and that spo12 mutants experience a G2/M delay indicates that SPO12 normally has a role at this time.

We observed that spo12 and hct1 are synthetically lethal. HCT1 is a component of the APC that is required for degradation of Clb2p and subsequent inactivation of Clb-Cdc28 kinase activity (Schwab et al., 1997; Visintin et al., 1997). Synthetic lethality of spo12 with hct1, a known component of the APC, suggests that if Spo12p acts to inhibit Clb-Cdc28 kinase activity, it does not do so via an APC-dependent mechanism.

One possibility is that Spo12p activates a CKI. Sic1p is present from late M into G1 and is required to keep Clb5- and Clb6-Cdc28 kinases inactive until cells are ready to replicate their DNA (Donovan et al., 1994). Sic1p may also play a role in exit from mitosis (Donovan et al., 1994; Toyn et al., 1997). Thus, one attractive possibility is that Spo12p may facilitate Sic1p activity during vegetative growth. Genetic observations indicate that this hypothesis may be correct. The ability of SPO12 when overexpressed to suppress cdc15-2, which arrests late in mitosis with high Clb-Cdc28 kinase activity, requires functional SIC1. In addition, the ability of SIC1 to suppress cdc15-2 requires functional SPO12. Therefore, these proteins may act together. For example, interaction between Spo12p and Sic1p may stabilize Sic1p and allow it to decrease Clb-Cdc28 kinase activity. Another possibility is that Spo12p acts to inhibit SCF, which is required for degradation of Sic1p (Feldman et al., 1997; Skowyra et al., 1997).

Spo12p Does Not Affect Clb1-, Clb3-, Clb4-, or Clb5-Cdc28 Kinase Activity during Meiosis

The experiments in mitotic cells described above lead to the possibility that Spo12p might regulate Clb-Cdc28 kinase activity during meiosis. Five of the six B-type cyclins are expressed during meiosis (all but CLB2), and deletion of the CLB genes either alone or in combination can lead to the reduction or complete absence of the meiotic nuclear divisions, depending on the mutant combination (Grandin and Reed, 1993; Dahmann and Futcher, 1995; Chu and Hersko-witz, 1998; Stuart and Wittenberg, 1998). Because Spo12p may inhibit Clb-Cdc28 kinase activity in vegetative cells, and because of the clear importance of the Clb proteins for meiosis, we hypothesized that Spo12p might modulate Clb-Cdc28 kinase levels during meiosis.

We first considered the possibility that Spo12p regulates Clb-Cdc28 kinase activity during meiosis through Sic1p; however, Sic1p was not detectable in meiosis after the very early stages and in particular was absent during the meiotic divisions, when Spo12p is believed to be required. Our observations provide no support for the hypothesis that Spo12p acts via Sic1p in meiosis. We then examined Clb1-, Clb3-, Clb4-, Clb5-, and Clb6-Cdc28 kinase activities directly in the wild type and in spo12 mutants. The levels and timing of specific Clb-Cdc28 kinase activities (as assayed by measuring histone H1 kinase activity in vitro) were indistinguishable for wild-type and spo12 mutant strains, indicating that Spo12p does not regulate these species.

So what is Spo12p doing? It remains an open question whether Spo12p regulates Clb-Cdc28 activity during meiosis. Perhaps Spo12p acts with Sic1p during mitosis, but during meiosis its relevant partner is an as yet unidentified meiotic CKI. Another possibility is that Spo12p alters Clb-Cdc28 kinase activity phosphorylation. We cannot exclude the possibility that Spo12p regulates meiotic divisions via a completely distinct mechanism. For example, perhaps Spo12p (and Spo13p) are required to ensure cosegregation of sister chromatids during meiosis I.

In summary, SPO12 is crucial for the proper regulation of the meiotic nuclear divisions and possibly exit from mitosis as well. The observations described here should provide information that will ultimately contribute to the understanding of the function of Spo12 protein.