Ime2 protein kinase activity is required for proper progression through the early stages of meiosis 
. However, critical substrates of this enzyme that specifically promote early meiotic progression have not been identified. The studies that we describe here were designed to define the position that Ime2 occupies in the pathway that leads to initiation of pre-meiotic DNA replication. This information will guide future studies aimed at characterizing mechanisms by which Ime2 directs early meiotic transitions.
We found that pre-meiotic DNA replication was delayed in ime2Δ
cells, as previously reported 
, and demonstrated that Sic1 disappeared in this same window. It has been shown that Sic1 becomes ubiquitylated in ime2Δ
cells, implying that Sic1 degradation occurs through the proteasome in the absence of Ime2 
. It has also been shown that delayed entry into pre-meiotic DNA replication in ime2Δ
cells is accompanied by delayed DNA polymerase alpha-primase complex phosphorylation, which occurs during S phase in WT cells engaged in the mitotic cell cycle or undergoing meiosis 
. Therefore, by different criteria the delayed pre-meiotic DNA replication observed in ime2Δ
cells resembles the WT process, suggesting that the normal mechanisms regulating DNA replication initiation occur in these cells, but with slower kinetics.
It is striking that the majority of Sic1 disappeared in ime2Δ
cells by 24 hours. Previous studies have shown that the same phosphorylation sites control Sic1 stability during the cell cycle and meiosis 
. However, the Cln/Cdk1 enzymes do not operate during meiosis 
. While we cannot be certain at this point that destabilization of Sic1 in ime2Δ
cells occurs through the same process as in WT cells, our results clearly indicate that the Sic1 steady state level can be decreased significantly in an Ime2-independent manner. Based on the phosphorylation sites that are required for Sic1 destruction during meiosis, we expect Sic1 phosphorylation to be catalyzed by a cyclin/CDK-like complex that normally requires Ime2 for activation but can eventually become activated in the absence of Ime2. Sic1 destabilization occurs in IME2+
cells devoid of Clb5 and -6, or when Cdk1 is inhibited, indicating that Clb5,-6/Cdk1 or other Cdk1-containing complexes are not involved 
. Future work will be aimed at identifying the responsible protein kinase(s).
Our examination of the Sic1 steady state level in ime2Δ
cells suggests that Ime2 lies further upstream in promoting Sic1 destruction than originally suspected. Support for this hypothesis comes from our global gene expression analysis indicating that Ime2 activates expression of many genes that are controlled during the cell cycle by SBF or MBF. While expression of these genes was not abolished in the ime2Δ
mutant, some redundancy could exist in controlling expression of these genes. This type of overlap has been well established in the mitotic cell cycle, as Bck2 induces transcription of many cell cycle-regulated genes regardless of cell cycle phase 
. Such a secondary pathway might eventually allow for the delayed progression through pre-meiotic DNA replication observed in ime2Δ
cells. From our demonstration that Ime2 significantly influences transcription of SBF and MBF targets, and from results that have been presented in the literature, we have arrived at the model depicted in . During the cell cycle, Cln3/Cdk1 lies upstream of SBF and MBF, while Cln1,-2/Cdk1 are downstream of Cln3/Cdk1 and respond to SBF and MBF 
. Therefore, rather than serving simply as a meiotic substitute for Cln1,-2/Cdk1, it appears that Ime2 has an earlier function in the meiotic pathway, perhaps operating like Cln3/Cdk1 or acting even further upstream. In this position, Ime2 activity could induce an increased level of a gene product that influences Sic1 structure and stability directly, e.g.
through Sic1 phosphorylation, or indirectly, e.g.
through activation of the responsible Sic1 kinase.
Our data indicate that upregulation of SBF and MBF targets, while similar in the cell cycle and during meiosis, is not identical, as subsets that respond during the cell cycle are not induced during meiosis. It is important to consider that cell cycle box function during meiosis, including the nature of trans-acting factors that bind to these sequences, has not been defined on a general level. Mutation analysis has revealed that MCB elements do operate during meiosis, specifically with regard to the CLB5
. However, this same study showed that Mbp1 is not required for normal meiotic upregulation of CLB5
. By contrast, meiotic regulation of other genes with MCB promoter elements has been shown to involve MBF components: Mbp1 is required for proper upregulation of RNR1
, and Swi6 is required for proper upregulation of RAD51
. (Note that RAD54
is not included as a cell cycle-regulated MBF target in the data set that we used for our analyses 
.) The CLB5
results suggest that meiotic versions of cell cycle box-binding factors could operate in some contexts. Strong support for this hypothesis comes from the report that cells lacking both Swi4 and Mbp1, which arrest in the mitotic cell cycle 
, can proceed through meiosis 
. Regulatory subunits might also differ between meiosis and the cell cycle, given our evidence that Whi5 does not lie downstream of Ime2 in the pathway leading to pre-meiotic DNA replication. Regardless of the exact components, our data indicate that IME2
acts during meiosis upstream of many genes controlled by SBF or MBF during the cell cycle. An important direction of our future research will be to precisely define the mechanism(s) by which Ime2 activity influences transcription of these genes.