Regulation of Meiotic S Phase by Ime2 and a Clb5,6-Associated Kinase in Saccharomyces cerevisiae

See allHide authors and affiliations

Science  18 Sep 1998:
Vol. 281, Issue 5384, pp. 1854-1857
DOI: 10.1126/science.281.5384.1854


Cyclin-dependent kinase (Cdk) mutations that prevent entry into the mitotic cell cycle of budding yeast fail to block meiotic DNA replication, suggesting there may be fundamental differences between these pathways. However, S phase in meiosis was found to depend on the same B-type cyclins (Clb5 and Clb6) as it does in mitosis. Meiosis differs instead in the mechanism that controls removal of the Cdk inhibitor Sic1. Destruction of Sic1 and activation of a Clb5-dependent kinase in meiotic cells required the action of the meiosis-specific protein kinase Ime2, thereby coupling early meiotic gene expression to control of DNA replication for meiosis.

In eukaryotes, both mitotic chromosomal DNA replication (S phase) and chromosome segregation (M phase) are triggered by cyclin-dependent kinases (Cdks). In budding yeast, specific B-type cyclins activate the principal Cdk, Cdc28, to control entry into S phase. Late in G1, a series of events leads to activation of the S-phase promoting factor (SPF), a complex of Cdc28 with either of two similar B-type cyclins, Clb5 and Clb6 (1–3). Before S phase, the Clb5,-6/Cdc28 kinase is maintained in an inactive state by its association with the CDK inhibitor (CDI) Sic1. Targeting of Sic1 for degradation requires activity of the Cdc28 kinase in conjunction with the G1-specific cyclins Cln1 and Cln2 (4–7). In mutants lacking Clb5 and Clb6, S phase occurs but is delayed, because other B-type cyclins (CLB1–4) that arise later in the cell cycle can also fulfill this function (8).

In meiosis, the G1- to S-phase transition is controlled differently. The temperature-sensitive CDK mutation that blocks mitotic S phase, cdc28-4, has no effect on meiotic S phase (9). Thus, Cdc28 might render mitotic DNA synthesis dependent on growth conditions, while being dispensable for meiotic S phase (10). One possibility is that the meiosis-specific kinase encoded by IME2, which has sequence similarity to Cdc28 and is required for meiotic S phase (11–13), replaces Cdc28 in an SPF-like function that is required in meiosis. To find out, we have reinvestigated genetic interactions that influence DNA replication in meiosis.

Yeast strains mutant for the G1-specific cyclinsCLN1,-2,-3 are blocked in mitosis at G1(14, 15) but form viable spores to the same extent as do wild-type (WT) cells (16, 17). Likewise, mutations eliminating components of transcription factors for the CLN1and CLN2 genes (SBF and MBF) cause mitotic G1 arrest (18) but fail to block meiosis (16, 17). These findings are consistent with a differential control of S phase between mitosis and meiosis, but the results of introducing mutations in CLB5 and CLB6(19) are not. Although a strain deleted for CLB6is fully proficient in sporulation, a clb5Δ homozygote fails to sporulate (16) (Fig. 1C). Because different genetic backgrounds vary in sporulation efficiency, we also testedclb5Δ in two other strain backgrounds (W303 and YK) and obtained similar results (16, 19). Meiotic S phase (20) in a clb5Δ homozygote was delayed about 3 hours relative to the WT (Fig. 1B) and meiotic progression was blocked either in G2 (60% of cells) or in meiosis I (40% of cells) (Fig. 1, A and C). The delay of S phase did not result from delayed initiation of meiosis, for the increase in transcription of both IME2 (an early meiotic gene) and CLB6occurred on schedule and to even higher levels than that in WT cells (Fig. 1D) (20). Assessment of intragenic recombination within the HIS4 locus revealed no His+ progeny (21), indicating either that CLB5 is required for recombination or that recombinants are inviable. We conclude thatCLB5 function is required for scheduled entry into S phase and that clb5Δ homozygotes, despite their capacity for extensive DNA replication, are unable to progress normally through subsequent stages of meiosis. These latter anomalies might reflect either qualitative defects in the newly replicated DNA or an additional requirement for CLB5 function in the transition from G2 to meiosis I. The finding that CLB5 is also activated by the G2-specific meiotic transcription factor Ndt80 (22) favors the latter explanation.

Figure 1

Phenotypes of clb5Δand clb6Δ mutants in meiosis. (A to C) Meiotic processes were scored for a/α strains L179 (WT), L180 (clb5Δ/Δ), L243 (clb6Δ/Δ), and L246 (clb5Δ/Δ clb6Δ/Δ) at the indicated times after transfer to sporulation medium (SPM). (A) Meiotic nuclear division: a single round nucleus (G1 + S + G2), two nuclear lobes (MI), and four nuclear lobes (MII). After t = 10 hours, nuclear staining ofclb5Δ/Δ andclb5Δ/Δ clb6Δ/Δcultures was too diffuse to score. (B) Percentage of cells with replicated DNA content (G2) by FACS analysis. (C) Percentage of asci (sporulated cells) in the same cultures. (D) Meiotic transcripts for the IME2and CLB6 genes in W303 WT (L51) andclb5Δ/Δ (L139) cells. IC, internal control (PC4 from J. Segal).

Because CLB6 is functionally redundant with CLB5in mitosis (1–3), we also tested strains deleted for both genes (16). In both genetic backgrounds tested, sporulation failed to occur and DNA replication was impaired to a greater extent than in cells lacking CLB5 alone (Fig. 1, B and C), which indicates that either CLB5 or CLB6can act to promote meiotic DNA synthesis. Flow cytometric profiles showed late and incomplete DNA replication, and further progression through the meiotic pathway did not occur (Fig. 1, A and B) (21).

Because both the Clb5 and Clb6 cyclins and the Ime2 kinase are required for meiotic DNA replication (Fig. 1) (11–13), they might constitute a single complex that promotes the G1-S transition in meiosis (Fig. 4B2). Alternatively, Ime2 might mediate entry into S phase indirectly by controlling removal of an inhibitor, such as Sic1 (Fig. 4B3). The target of this inhibitor would be an unspecified meiotic Cdk (designated X) that is activated by association with Clb5 and Clb6, and the role of Ime2 in meiosis would parallel that of Cln1,-2/Cdc28 in mitosis. In this case, deletion of SIC1might render IME2 dispensable for DNA replication. Indeed, deletion of SIC1 in ime2Δ cells restored DNA replication to the level and kinetics observed for the controlsic1Δ strain (Fig. 2). Thus, Sic1 acts by blocking entry into meiotic S phase, and Ime2, instead of promoting entry directly, may control it indirectly by mediating destruction of the SPF inhibitor Sic1.

Figure 2

Effect of SIC1deletion on meiotic DNA replication in an ime2Δ strain. FACS analysis of SK1 a/α strains bearingsic1Δ/Δ (L182),ime2Δ/Δ (L213), andime2Δ/Δ sic1Δ/Δ(L234). Because many sic1Δ cells remain blocked in G2 of mitosis after transfer to sporulation medium (27), the percentage of budded G2 cells was subtracted from the total percentage of cells with replicated DNA to yield the percentage of meiotic cells with G2 DNA content. Electron microscopy confirmed the presence of condensed chromosomes and other meiotic features in the unbudded cells of theime2Δ/Δ sic1Δ/Δculture (21).

If Ime2 regulates the amount of Sic1, ime2Δ cells should accumulate Sic1 protein and Clb5-associated kinase should be absent from ime2Δ cells undergoing meiosis. Protein immunoblot analysis of the meiotic cells (23) demonstrated that Sic1 was lost near the time of S phase in WT cells but accumulated in ime2Δ cells under identical conditions (Fig. 3A). Furthermore, histone H1 kinase assays (23) on extracts of meiotic cells showed a transient increase in Clb5-associated kinase in WT cells, but not inime2Δ cells, around the time of DNA replication (t= 6 to 10 hours; Fig. 3B). In ime2Δ cells, Sic1 persisted and Clb5 kinase activity was absent (Fig. 3, A and B). The presence of increased Clb5-associated kinase activity in the WT at late time points (up to 10 hours) might result from persistentCLB5 transcription in G2 (22), consistent with a possible role for this kinase in meiosis I.

Figure 3

Amounts of Sic1 protein (Sic1p) and Clb5-associated kinase in WT and ime2Δmeiotic cells. (A) (Top) Protein immunoblot analysis (23) with polyclonal antibody to Sic1 in WT (L179) and ime2Δ (L213) during a meiotic time course. (Bottom) FACS analyses of DNA replication. The sic1Δ control is L182. (B) Histone H1 kinase assay of anti-HA immunoprecipitates (23) from WT (L535) and ime2Δ (L545), both carrying two CLB5-HA3 alleles. Note the faint kinase activity due to remaining mitotic S-phase cells at t = 0 and 2 hours. YPD, samples from cycling mitotic cultures.

We have shown here that meiotic SPF has two predominant characteristics expected for a cyclin B/Cdc28 complex—dependence on Clb5,-6 (Figs. 1 and 3B) and inhibition by Sic1 (Figs. 2 and 3A). In addition, we have ruled out Ime2 as an essential component of meiotic SPF (Fig. 2). Lacking a cdc28 allele that is specifically defective in Clb5,-6 activation in vivo, we cannot exclude the possibility that enzyme X is Cdc28 itself. It has been shown that cdc2, the enzyme in Schizosaccharomyces pombe that is analogous to Cdc28, is required for meiotic S phase, but a relevant cyclin is unidentified (24). In vertebrates, Cdk functions needed for the G1/S transition are performed by several Cdks (25) and meiosis-specific activities remain unresolved.

How might meiosis have evolved from a mitotic cycle? Because the precondition in yeast for mitosis (growth conditions) differs so markedly from that for meiosis (starvation), the SBF/MBFtranscription program triggered by Cln3/Cdc28 for mitosis may have been supplanted by a meiosis-specific early transcription program triggered by IME1 (16, 26) (Fig. 4, A and B). However, the demonstrated conservation in meiosis of roles for the same cyclins (Clb5,-6) and their inhibitor (Sic1) as in mitosis strongly implicates Clb5,-6/Cdk complexes in the control of meiotic S phase. This suggests that functions of the cyclin B/Cdk complex in controlling DNA replication were conserved in the evolution of meiosis, whereas the role of Cln1,-2/Cdk complexes in Sic1 removal in the mitotic cycle was replaced by action of the Ime2 kinase in meiosis.

Figure 4

Comparative models for control of entry into S phase for mitotic and meiotic cell cycles. (A) Mitotic G1/S. Late G1-specific transcription factors SBF (Swi4/Swi6) and MBF (Mbp1/Swi6) stimulate expression not only of the G1 cyclins (CLN1,CLN2) that target Sic1 for proteolysis but also ofCLB5 and CLB6. S-phase B-type cyclins Clb5 and Clb6 combine with Cdc28 to form SPF, which promotes entry into S phase. (B) Meiotic G1/S. (B1) After transcription from Ime1/Ume6 (early meiotic transcription factors), the Ime2 (inducer of meiosis) kinase directly triggers entry into S phase. (B2) As in (B1) but showing Clb5 and Clb6 as potential activators of Ime2. (B3) Revised model. Like the Cln1,-2-associated kinase in model A, Ime2 promotes S phase indirectly by triggering destruction of Sic1, thus allowing increased activity of an unidentified kinase (X) associated with Clb5 and Clb6.


Stay Connected to Science

Navigate This Article