The G2-M phase transition in eukaryotes is regulated by the synergistic and opposing activities of a cascade of distinct protein kinases and phosphatases. This cascade converges on Cdc2, a serine/threonine protein kinase required for entry into mitosis (reviewed in ref. 1). In the fission yeast Schizosaccharomyces pombe, inactivation of the Cdc2/cyclin B complex is achieved by phosphorylation of tyrosine 15 by Wee1 (refs 2,3). The action of the Wee1 kinase is opposed by the action of the Cdc25 phosphatase, which dephosphorylates Cdc2 on tyrosine 15, thereby activating the Cdc2/cyclin B complex. Much less is known about the regulatory signals upstream of cdc25 and wee1. Genetics indicate that the mitotic inducer nim1/cdr1 acts upstream of wee1, possibly as a negative regulator of wee1 (refs 10, 11). To characterize the nim1/cdr1 protein (Nim1), we have overproduced it in both bacterial and baculoviral expression systems. We report that Nim1 possesses intrinsic serine-kinase, threonine-kinase and tyrosine-kinase activities. Co-expression of the Nim1 and Wee1 kinases in insect cells results in the phosphorylation of Wee1 and therefore a shift in its electrophoretic mobility on SDS-polyacrylamide gels. When Wee1 is phosphorylated, its ability to phosphorylate Cdc2 on tyrosine 15 is inhibited; treatment with phosphatase restores this kinase activity. Furthermore, purified bacterially produced Nim1 kinase directly phosphorylates and inactivates Wee1 in vitro. These results show that nim1/cdr1 functions as a positive regulator of mitosis by directly phosphorylating and inactivating the mitotic inhibitor Wee1.

Previous work has established that activation of Mos, Mek, and p42 mitogen-activated protein (MAP) kinase can trigger release from G2-phase arrest in Xenopus oocytes and oocyte extracts and can cause Xenopus embryos and extracts to arrest in mitosis. Herein we have found that activation of the MAP kinase cascade can also bring about an interphase arrest in cycling extracts. Activation of the cascade early in the cycle was found to bring about the interphase arrest, which was characterized by an intact nuclear envelope, partially condensed chromatin, and interphase levels of H1 kinase activity, whereas activation of the cascade just before mitosis brought about the mitotic arrest, with a dissolved nuclear envelope, condensed chromatin, and high levels of H1 kinase activity. Early MAP kinase activation did not interfere significantly with DNA replication, cyclin synthesis, or association of cyclins with Cdc2, but it did prevent hyperphosphorylation of Cdc25 and Wee1 and activation of Cdc2/cyclin complexes. Thus, the extracts were arrested in a G2-like state, unable to activate Cdc2/cyclin complexes. The MAP kinase-induced G2 arrest appeared not to be related to the DNA replication checkpoint and not to be mediated through inhibition of Cdk2/cyclin E; evidently a novel mechanism underlies this arrest. Finally, we found that by delaying the inactivation of MAP kinase during release of a cytostatic factor-arrested extract from its arrest state, we could delay the subsequent entry into mitosis. This finding suggests that it is the persistence of activated MAP kinase after fertilization that allows the occurrence of a G2-phase during the first mitotic cell cycle.

The Ste20 family of serine/threonine kinases plays an important role in numerous cellular functions such as growth, apoptosis, and morphogenesis. We have identified a previously cloned but uncharacterized family member termed Stk10, which is a human homolog of murine Lok, a serine/threonine kinase highly expressed in lymphocytes. Northern analysis demonstrated that the Stk10 transcript is present in many tissues, although highest expression levels are seen in hematopoietic cells. Due to close sequence homology to human Slk and Xenopus laevis xPlkk1, two polo-like kinase kinases, we investigated whether Stk10 might also play a role as a Plk1 activator. Plk1 has been shown to be overexpressed in multiple tumor types, thus attracting high interest to its potential upstream regulators. We show here that Stk10 can associate with Plk1 in cells and furthermore can phosphorylate Plk1 in vitro. Engineered NIH-3T3 cell lines that overexpress a dominant negative version of Stk10 display an altered cell cycle phenotype characterized by increased DNA content, raising the possibility that expression of a dominant negative Stk10 may impinge upon Plk1 function in vivo; it has previously been shown that unregulated expression of Plk1 can result in a variety of nuclear defects. We suggest, therefore, that Stk10 is a novel polo-like kinase kinase that cooperates with hSlk to regulate Plk1 function in human cells.

Xenopus oocytes and eggs provide a dramatic example of how the consequences of p42 mitogen-activated protein kinase (p42 MAPK) activation depend on the particular context in which the activation occurs. In oocytes, the activation of Mos, MEK, and p42 MAPK is required for progesterone-induced Cdc2 activation, and activated forms of any of these proteins can bring about Cdc2 activation in the absence of progesterone. However, in fertilized eggs, activation of the Mos/MEK/p42 MAPK pathway has the opposite effect, inhibiting Cdc2 activation and causing a G2 phase delay or arrest. In the present study, we have investigated the mechanism and physiological significance of the p42 MAPK-induced G2 phase arrest, using Xenopus egg extracts as a model system. We found that Wee1-depleted extracts were unable to arrest in G2 phase in response to Mos, and adding back Wee1 to the extracts restored their ability to arrest. This finding formally places Wee1 downstream of Mos/MEK/p42 MAPK. Purified recombinant p42 MAPK was found to phosphorylate recombinant Wee1 in vitro at sites that are phosphorylated in extracts. Phosphorylation by p42 MAPK resulted in a modest ( approximately 2-fold) increase in the kinase activity of Wee1 toward Cdc2. Titration experiments in extracts demonstrated that a twofold increase in Wee1 activity is sufficient to cause the delay in mitotic entry seen in Mos-treated extracts. Finally, we present evidence that the negative regulation of Cdc2 by Mos/MEK/p42 MAPK contributes to the presence of an unusually long G2 phase in the first mitotic cell cycle. Prematurely inactivating p42 MAPK in egg extracts resulted in a corresponding hastening of the first mitosis. The negative effect of p42 MAPK on Cdc2 activation may help ensure that the first mitotic cell cycle is long enough to allow karyogamy to be accomplished successfully.