A Meiosis-Specific Cyclin Regulated by Splicing Is Required for Proper Progression through Meiosis

Strains and media.All S. pombe strains used are isogenic to wild-type 972 h ⁻. The strains used in this work are listed in Table 1. Media were prepared as described previously (27), and S. pombe was transformed using the lithium acetate method (24). Flow cytometry was performed as described previously (3).

For 4′,6′-diamidino-2-phenylindole (DAPI) staining, 1 ml of cell culture (1 × 10⁷ to 2 × 10⁷ cells) was centrifuged briefly, fixed with 70% ethanol, and stored. Fifty microliters of fixed cells (∼5 × 10⁵ cells) were then added to 1 ml of water, centrifuged, and resuspended with 100 μl of water, spotted onto slides, and fixed by heating. Nuclei were stained with DAPI (5 μg/ml).

Fluorescence microscopy was carried out on a Nikon Eclipse E600 microscope at ×100 magnification. Images were captured by use of an Orca II dual-scan cooled charge-coupled device camera (Hamamatsu), using Hamamatsu AquaCosmos 2.0 software.

Synchronous meiosis.To obtain meiotic cultures in the diploid h ⁺/h ⁻ strains, single colonies were grown in YE (27) to early stationary phase. The cells were diluted in minimal medium supplemented with 100 μg/ml leucine and grown at 30°C with shaking to 1 × 10⁷ to 2 × 10⁷ cells/ml. Cultures were filtered through a Millipore membrane, washed with 2 volumes of medium without nitrogen, resuspended in medium without nitrogen containing 50 μg/ml leucine and 0.5% glucose, and placed in the shaker at 30°C. When indicated, hydroxyurea (20 mM) was added after 2 and 6 h in the medium without nitrogen.

For the pat1-114 strains, single colonies were grown in YE5S (27) to early stationary phase. The cells were diluted in minimal medium supplemented with 100 μg/ml leucine and grown at 25°C to 1 × 10⁷ to 2 × 10⁷ cells/ml. The cultures were filtered through a Millipore membrane, washed with 2 volumes of medium without nitrogen, and resuspended in medium without nitrogen containing 50 μg/ml leucine. The concentration of the cells was adjusted to 4 × 10⁶ to 6 × 10⁶/ml and incubated at 25°C overnight. NH₄Cl and leucine were added to the cultures at 500 μg/ml and 50 μg/ml, respectively, just before the cultures were shifted to 34°C to induce meiosis. When indicated, hydroxyurea (20 mM) was added at the time of the temperature shift and 4 h later.

Gene expression analysis.RNA was prepared by glass bead lysis in the presence of hot phenol, as described previously (7). Equal amounts of RNA (measured by optical density at 260 nm) were separated in formaldehyde agarose gels containing ethidium bromide (30), to confirm equal loading by visualizing rRNA, and transferred to GeneScreen Plus membranes (NEN Life Science Products). Hybridization and washes were performed as recommended by the manufacturer. rem1 and mei4 probes contained the open reading frame of these genes.

For the reverse transcription-PCRs (RT-PCRs), RNA was digested with DNase I for 30 min at 37°C, phenol extracted, and precipitated. Eight micrograms of total RNA was denatured at 65°C for 10 min and then chilled on ice. Reverse transcriptase reactions were carried out (60 min at 42°C, 30 min at 52°C, and 3 min at 94°C) following the manufacturer's guidelines (Promega) in the presence or absence of the enzyme. One microliter of the cDNA was used in the PCRs with primers JA71 (5′-CAAAGTCCTAATTGCAGTGTTTCGG-3′) and JA174 (5′-GGAGGATGGATCTCTTCATACTTGC-3′).

The transcription start point of rem1 was determined by primer extension using the oligonucleotide JA145 (5′-GCGCAACTCTTTTGTTGTTAGAGTTC-3′), labeled at its 5′ end with the fluorochrome 6-FAM, as described elsewhere. Briefly, the labeled primer, complementary to the mRNA template at a site 28 bp downstream from the ATG codon of the open reading frame of rem1, was hybridized with total mRNA from S. pombe diploid cells grown for 4 h in the absence of nitrogen. Extension of the primer with AMV reverse transcriptase (Life Technologies) was carried out. Analysis of the extended products was performed in an automatic DNA sequencer, loading together in the same well the primer extension reaction product and the DNA fragments generated in a sequencing reaction carried out with the same but nonlabeled oligonucleotide. The transcription start point was confirmed by rapid amplification of cDNA ends, cloning of the PCR product, and sequencing of five different clones.

Protein extraction, immunoprecipitation, and in vitro kinase assay.Extracts were prepared as previously described (2). For the Western Blots, 50 μg of extracts were resolved in 10% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (30:0.8 acrylamide:bisacrylamide), transferred to polyvinylidene difluoride membranes, and blotted with the indicated antibody.

For the in vitro kinase assay, protein extracts (50 μg) from cells expressing hemagglutinin (HA)-tagged Rem1 were immunoprecipitated with anti-HA or anti-Cig2 monoclonal antibodies. Immunoprecipitates were washed in NET-N buffer (20 mM Tris [pH 8.0], 100 mM NaCl, 1 mM EDTA, 0.5% NP-40, 1 mM dithiothreitol, 0.5 mM phenylmethylsulfonyl fluoride, 5 μg/ml leupeptin, and 5 μg/ml aprotinin) followed by kinase buffer (100 mM HEPES, pH 7.5, 20 mM MgCl₂, 4 mM EGTA, 2 mM dithiothreitol) and incubated with histone H1 (1 μg) and 10 μCi of [γ-³²P]ATP. After 20 min at 30°C, the reactions were stopped with sample buffer, and the proteins were separated by 11% sodium dodecyl sulfate-polyacrylamide gel electrophoresis.

rem1 mRNA is spliced during meiosis.Rem1 (SPBC16E9.17c) was investigated because it shows high sequence similarity with other S. pombe cyclins (Fig. 1A). rem1 mRNA was barely detectable during the mitotic cell cycle but was clearly accumulated at the onset of premeiotic DNA synthesis in diploid (h ⁻/h ⁻) pat1-114 cells (Fig. 1B; also see below). Nevertheless, Rem1 protein and its associated kinase activity were induced around the time of meiosis I onset (Fig. 1C and D). Furthermore, we have identified Cdc2 as the catalytic subunit of the complex, since this kinase coprecipitates with Rem1 (Fig. 1E).

• Open in new tab
• Download powerpoint

FIG. 1.

Rem1 is a meiosis-specific cyclin. (A) Sequence alignment of Rem1 with other S. pombe cyclins. (B) Synchronously induced diploid (h ⁻ /h ⁻) pat1-114 cells (JA95) were sampled and DNA content measured by flow cytometry. (C) Protein extracts were prepared from synchronous meiosis of pat1-114 cells (JA95), and Rem1-HA was detected by Western blotting using anti-HA antibodies. Anti-Cdc2 antibodies were used as a loading control (bottom panel). S, MI, and MII indicate S phase, meiosis I, and meiosis II, respectively. (D) In vitro Rem1-associated kinase activity (in arbitrary units) was measured in the same protein extracts, using histone H1 as the substrate. (E) Extracts (1 mg) prepared from synchronous meiosis in pat1-114 cells (JA95) at the time indicated at the bottom were immunoprecipitated with anti-HA (α-HA) antibody (12CA5). The presence of Cdc2 in the immunoprecipitate was analyzed in a Western blot. WCE, whole-cell extract. (F) RNA was isolated from the cells represented in panel B, and rem1, mes1, and his3 expression were monitored by RT-PCR. The faster-migrating form of rem1 corresponds to mature mRNA and the slow migrating form to unprocessed mRNA.

While investigating the regulation of rem1 transcription during meiosis in the same strain, we found that its mRNA undergoes meiosis-specific splicing before the onset of meiosis I (Fig. 1F) (t = 3.5 h), which would explain the delay between transcription and protein expression. Before this splicing occurs, functional Rem1 cannot be produced, because the intron encodes an in-frame stop codon. Interestingly, when the cDNA of rem1 was overexpressed in vegetatively growing cells using the weak nmt81 promoter, cells arrested in the G₁ phase of the cell cycle (Fig. 2A) and developed a typical “cut” phenotype, with cells dividing without completing DNA synthesis (Fig. 2B). Thus, Rem1 expression impairs the onset of mitotic S phase and leads to a failure in cell cycle checkpoint control, explaining why Rem1 has to be absent during the mitotic cell cycle.

• Open in new tab
• Download powerpoint

FIG. 2.

Rem1 induces a cell cycle arrest in mitotically growing cells. (A) Wild-type cells transformed with rem1 cDNA under the control of a weak nmt promoter (pREP81x) were grown to mid-exponential phase in the presence of thiamine (promoter off). Thiamine was washed out (promoter on), and DNA content was measured by flow cytometry at the times indicated at the right. (B) Micrograph of DAPI-stained cells from panel A after 16 h in the absence of thiamine. Arrowheads indicate cells with mitotic catastrophe.

To further investigate the meiosis-dependent regulation of rem1 splicing, we tested mutants that block meiotic progression at specific stages during meiosis, both in a pat1-114 background (data not shown) and in the more physiological h ⁺ /h ⁻ diploid background (Fig. 3). After nitrogen deprivation, cells harboring either a mei4 or a mes1 deletion did not complete meiosis I or II, respectively (Fig. 3A and C). In all the strains, rem1 transcription was induced 3 h after the cells were transferred to sporulating media, which correlates with the time at which premeiotic S phase takes place in this genetic background (Fig. 3B). However, in a wild-type or Δmes1 strain, rem1 mRNA was spliced, while splicing was not detected in Δmei4 cells blocked before meiosis I (Fig. 3D).

• Open in new tab
• Download powerpoint

FIG. 3.

Splicing of rem1 mRNA is dependent on Mei4. (A) In the absence of nitrogen, diploid cells transiently arrest in G₁ before proceeding into premeiotic S phase and the two nuclear divisions. Mei4 is required at the onset of meiosis I (MI), while Mes1 is required for meiosis II (MII) completion. (B) Synchronously induced diploid h ⁺/h ⁻ cells PN2304 (wt), JA175 (Δmei4), and JA174 (Δmes1) were sampled and DNA content measured by flow cytometry. (C) Meiotic progression of the same cultures was measured by nuclear counting of DAPI-stained cells. Closed circles are percentages of cells with one nucleus; open circles, two nuclei; and closed triangles, three or four nuclei. (D) RNA was isolated, and rem1 splicing was monitored by RT-PCR. The faster-migrating form of rem1 corresponds to mature mRNA and the slow-migrating form to unprocessed mRNA.

Even though lack of Mei4 blocks meiosis at the onset of meiosis I (15), we have observed that mei4 mRNA is already present during premeiotic S phase (see below). To determine whether the lack of rem1 mRNA splicing in the Δmei4 strain was caused directly by the absence of Mei4 or was due to cells being blocked in their meiotic progression, we examined splicing during a premeiotic S-phase block. pat1-114 and pat1-114 Δmei4 cells were nitrogen-starved to accumulate cells in G₁ phase, and the cultures were then shifted to the nonpermissive temperature to induce meiosis. At the same time, each culture was split in two, and hydroxyurea was added to one of the cultures, arresting cells before premeiotic S phase. The other culture progressed synchronously into meiosis, and premeiotic S phase took place 2 h after the cells were placed at the nonpermissive temperature (Fig. 4A). Aliquots of the cultures were taken every hour, and RNA was isolated and analyzed for rem1 and mei4 expression. As shown in Fig. 4B, in the absence of hydroxyurea, the mRNA of both genes was detected at the onset of premeiotic S phase in the pat1-114 cells, with the peak of accumulation occurring just before the first meiotic division (Fig. 4B, 4 h). However, when meiotic progression was blocked by hydroxyurea, transcripts of both genes were present throughout the time course (Fig. 4B, +HU). Furthermore, rem1 mRNA was fully spliced before meiosis I in the untreated pat1-114 culture (Fig. 4C, 4 h, −HU) and was also spliced concomitantly with the peak of transcript accumulation in cells arrested in S phase, although mRNA processing was only partial (Fig. 4C, +HU, 4 and 6 h). This result, together with the absence of splicing in a Δmei4 mutant, indicates that rem1 splicing was dependent on mei4 function and, unexpectedly, that such splicing can take place during S phase. It should also be noted that in the presence of hydroxyurea, splicing is not as effective as in its absence, since unprocessed mRNA could be detected throughout the time course, which is consistent with the fact that mei4 expression is also decreased in the presence of hydroxyurea (Fig. 4B). Similar results were obtained when meiosis was examined in diploid h ⁺ /h ⁻ cells (data not shown).

• Open in new tab
• Download powerpoint

FIG. 4.

rem1 is expressed and spliced during extended premeiotic S phase. (A) Synchronous meiosis was induced in haploid pat1-114 (JA247) and pat1-114 Δmei4 (JA416) cells in the presence (+HU) or absence (−HU) of 20 mM hydroxyurea, which was added when the culture was placed at the restrictive temperature and again 4 h later to ensure that DNA synthesis was blocked. DNA content of the cells was measured by flow cytometry. (B) RNA was isolated at the indicated times after shift to the restrictive temperature and analyzed for the expression of rem1 and mei4. (C) rem1 splicing was monitored by RT-PCR.

Rem1 is a cyclin required for normal levels of meiotic recombination.Next, we characterized the spores of zygotic asci (37, 38) from both wild-type and Δrem1 cells. Δrem1 zygotes produced four-spore asci with a level of viability similar to that of wild-type cells (>90%), but ade6 gene conversion was threefold reduced relative to that of wild-type cells (Fig. 5A). This decrease in the intragenic recombination was independent of the M26 hot spot, since a decrease was also observed in the absence of the ade6-M26 allele (lower left panel) and when ura1 gene conversion was measured (lower right panel). We observed a similar decrease in meiotic recombination (two- to fivefold) with other meiotic recombination mutants in S. pombe, including the swi5, dmc1, and rhp55 mutants (14; J. Malapeira and J. Ayté, unpublished data). We conclude that Rem1 is required for normal levels of meiotic intragenic recombination. Interestingly enough, Rem1 was not required for meiotic crossing-over when two different intervals (mat1-leu1 and ade6-arg1) were measured (Fig. 5B), indicating the presence of two pathways of recombination during meiosis, with and without crossing-over, with Rem1 having a role only in the latter. The existence of two different pathways was suggested before for budding yeast (1), and the isolation of a mutant (Δrem1) that shows an impairment only in meiotic gene conversion keeping wild-type levels of meiotic crossing-over confirms the existence of two distinct pathways.

• Open in new tab
• Download powerpoint

FIG. 5.

Rem1 is required for normal levels of meiotic recombination. (A) Intragenic ade6 or ura1 recombination (gene conversion) was measured in wild-type (wt) and Δrem1 strains with the indicated ade6 or ura1 alleles. Data are Ade⁺ or Ura⁺ prototrophs/10⁶ viable spores. (B) Frequencies of meiotic crossing-over of two different intervals (mat1-leu1 and ade6-arg1) in wild-type and Δrem1 cells.

To further investigate the function of Rem1, we deleted the gene in a pat1-114 background (25, 26, 29). In the absence of Rem1, pat1-114 cells completed meiosis but with a delay of 30 min in the onset of meiosis I and II (Fig. 6A). We next followed meiosis in diploid (h ⁺ /h ⁻) cells, which undergo meiosis less synchronously than pat1-114 cells but in a more physiological manner (Fig. 6B). When placed in sporulating media, diploid wild-type and Δrem1 cells transiently arrested in G₁ with a 2C DNA content proceeded into premeiotic DNA synthesis and then underwent meiosis I and II, with no noticeable difference between the two strains, probably due to the lack of enough synchrony of the diploid h ⁺ /h ⁻ diploid meiosis. Since no cyclins have been identified as being essential for premeiotic S phase in S. pombe, and to further characterize the phenotype produced by the absence of rem1, we tested if mutations in other cyclins might be synthetically lethal with rem1. cig2 has been described to have a role controlling premeiotic S phase (5), and we thought that this cyclin might genetically interact with rem1. This was established when the h ⁺ /h ⁻ Δrem1/Δrem1 Δcig2/Δcig2 double mutant was found to become blocked before premeiotic S phase (Fig. 6B), although the cells were committed to meiosis, since they had initiated at least part of the meiotic transcriptional program (i.e., the ste11, mat1, and rec8 genes were induced; J. Malapeira and J. Ayté, unpublished data). Furthermore, the cell cycle block could be rescued by a plasmid containing a genomic copy of either cig2 or rem1 (J. Malapeira and J. Ayté, unpublished data). Finally, since rem1 mRNA splicing depends on Mei4, a Δcig2 Δmei4 double mutant should have the same phenotype as a Δcig2 Δrem1 mutant. This was confirmed when an h ⁺/h ⁻ Δcig2/Δcig2 Δmei4/Δmei4 double mutant was also found to become blocked before premeiotic S phase (Fig. 7A), since these cells do not have Cig2 and are functionally depleted for mature Rem1 (Fig. 7B). We conclude that Rem1, in the absence of Cig2, is required for premeiotic S phase.

• Open in new tab
• Download powerpoint

FIG. 6.

Rem1 is required for proper meiotic progression. (A) Synchronous meiosis of pat1-114 (JA66) and pat1-114 Δrem1 (JA180) cells was sampled, and meiotic nuclear division was monitored by DAPI staining. Closed circles are percentages of cells with 1 nucleus; open circles, two nuclei; and closed triangles, three or four nuclei. The dotted line indicates the timing at which meiosis I takes place in pat1 cells. (B) Synchronously induced meiosis in diploid strains PN2304 (wt), Δrem1 (JA65), Δcig2 (JA64), and Δrem1 Δcig2 (JA112) was carried out, and DNA content was measured by flow cytometry.

• Open in new tab
• Download powerpoint

FIG. 7.

Δmei4 Δcig2 cells are blocked before premeiotic S phase. (A) Synchronous meiotic cultures of wild-type (wt) (PN2304), Δcig2 (JA64), Δmei4 (JA175), or Δmei4 Δcig2 (JM137) cells were sampled, and DNA content was analyzed. (B) rem1 splicing was monitored by RT-PCR. The faster-migrating form of rem1 corresponds to mature mRNA and the slow-migrating form to unprocessed mRNA.

Cig2 regulates the timing of rem1 splicing.We showed above that Δcig2 Δrem1 cells arrested before premeiotic S phase (Fig. 6B), even though Rem1 apparently was not expressed until after premeiotic S phase in wild-type cells (Fig. 1C and 3D). Given this observation, we examined the regulation of rem1 during meiosis in Δcig2 diploid cells. RNA was isolated from wild-type and Δcig2 cells undergoing synchronous meiosis (Fig. 6B), and rem1 mRNA splicing was analyzed. As shown in Fig. 8A, rem1 mRNA splicing was not detected until 4 h after wild-type cells were placed in sporulating media. In contrast, splicing in Δcig2 cells was induced earlier, just 1 h after cells were nitrogen starved. Furthermore, mature full-length Rem1 was not detected until 4 h after induction of wild-type meiosis (Fig. 8B, upper panel) but was detected as early as 1 h after induction of meiosis in Δcig2 cells (Fig. 8B, lower panel). mei4 expression was also induced earlier in Δcig2 cells (Fig. 8C). This observation may explain how regulated splicing of rem1 RNA can compensate for the lack of Cig2, because in the absence of Cig2 the earlier appearance of Rem1 allows the cells to proceed through premeiotic DNA synthesis.

• Open in new tab
• Download powerpoint

FIG. 8.

Rem1 is required for premeiotic S phase in the absence of Cig2. (A) rem1 splicing was monitored by RT-PCR in diploid strains PN2304 (wt) and JA64 (Δcig2) as they progressed into meiosis. (B) Protein extracts of wild-type and Δcig2 diploid cells were prepared from synchronous meiotic cultures at the indicated times. Rem1 was detected by Western blotting using Rem1 antibodies. (C) RNA from the same cultures was isolated, and the expression of mei4 was analyzed by Northern hybridization.