Calcium Elevation at Fertilization Coordinates Phosphorylation of Xerp1/Emi2 by Plx1 and Camk II to Release Metaphase Arrest by Cytostatic Factor

View metadata, citation and similar papers at core.ac.uk brought to you by CORE

provided by Elsevier - Publisher Connector Current Biology, Vol. 15, 1458–1468, August 23, 2005, ©2005 Elsevier Ltd All rights reserved. DOI 10.1016/j.cub.2005.07.030 Calcium Elevation at Fertilization Coordinates Phosphorylation of XErp1/Emi2 by Plx1 and CaMK II to Release Metaphase Arrest by Cytostatic Factor

Junjun Liu and James L. Maller* somes and resembling that seen at meiosis II. Masui Howard Hughes Medical Institute and and Markert named this cytoplasmic activity “cyto- Department of Pharmacology static factor” (CSF). They further speculated that CSF University of Colorado School of Medicine was responsible for the meiotic arrest at metaphase II Denver, Colorado 80262 of vertebrate unfertilized eggs, and this hypothesis was supported by the finding that the CSF activity appeared during meiosis II and disappeared shortly after fertiliza- Summary tion in response to elevated free Ca2+ [1]. Recently, a substantial amount of evidence has sug- Background: Vertebrate oocytes are arrested at sec- gested that regulation of anaphase-promoting com- ond meiotic metaphase by cytostatic factor (CSF) while plex/cyclosome (APC/C) is responsible for CSF arrest. awaiting fertilization. Accumulating evidence has sug- The APC/C is a ubiquitin-protein isopeptide ligase (E3) gested that inhibition of the anaphase-promoting com- composed of at least nine conserved subunits, and, in plex/cyclosome (APC/C) is responsible for this arrest. combination with a ubiquitin-activating enzyme (E1) Xenopus polo-like kinase 1 (Plx1) is required for activa- and a ubiquitin carrier protein (E2 or UBC), it promotes tion of the APC/C at the metaphase-anaphase transi- ubiquitination and degradation of M phase regulators tion, and calcium elevation, upon fertilization/activation by the 26S proteasome at the metaphase-anaphase of eggs, acting through calmodulin-dependent kinase transition [2–4]. Of the three ubiquitination compo- II (CaMKII) is sufficient to activate the APC/C and termi- nents, the APC/C is the only one whose activity clearly nate CSF arrest. However, connections between the oscillates during the cell cycle. In vertebrates, the activ- Plx1 pathway and the CaMKII pathway have not been ity of the APC/C is regulated in several ways. One in- identified. volves regulation of Fzy/Cdc20, which binds to the Results: Overexpression of Plx1 causes CSF release in APC/C and activates it [5]. During arrest at metaphase the absence of calcium, and depletion of Plx1 from egg by the spindle assembly checkpoint, the spindle- extracts blocks induction of CSF release by calcium checkpoint protein Mad2 binds Fzy/Cdc20 to inhibit ac- and CaMKII. Prior phosphorylation of the APC/C inhibi- tivation of the APC/C by Cdc20 [6–8]. In Xenopus egg tor XErp1/Emi2 by CaMK II renders it a good substrate extracts, overexpression of Xenopus Mad2 blocks cal- for Plx1, and phosphorylation by both kinases together cium-induced CSF release, presumably by inhibiting promotes its degradation in egg extracts. The pathway APC/C-Fzy/Cdc20 activity [5], and endogenous Mad2 is enhanced by the ability of Plx1 to cause calcium- is required for establishment of CSF arrest by Mos [9]. independent activation of CaMKII. The results identify Another pathway contributing to CSF arrest is the the targets of CaMKII and Plx1 that promote egg acti- cyclin E/Cdk2 pathway [10, 11]. This pathway is inde- vation and define the first known pathway of CSF re- pendent of the Mos/MAPK pathway and appears to also involve negative regulation of Fzy/Cdc20 because lease in which an APC/C inhibitor is targeted for degra- overexpression of Fzy/Cdc20 overcomes APC/C inhibi- dation only when both CaMKII and Plx1 are active after tion by cyclin E/Cdk2 (B. Grimison and J.M., unpub- calcium elevation at fertilization. lished data). Previously, it was reported that Early mi- Conclusions: Plx1 with an intact polo-box domain is totic inhibitor 1 (Emi1), a vertebrate homolog of the necessary for release of CSF arrest and sufficient when Drosophila Rca 1 gene, is able to bind Cdc20, inhibit overexpressed. It acts at the same level as CaMKII in the activity of the APC/C, and cause a CSF-like arrest the pathway of calcium-induced CSF release by coop- [12–14]. Depletion of Emi1 from CSF-arrested extracts erating with CaMKII to regulate APC/C regulator(s), was reported to cause spontaneous CSF release. How- such as XErp1/Emi2, rather than by directly activating ever, this finding was challenged by a later report show- the APC/C itself. ing that Emi1 is not present in CSF-arrested cells at a level significant enough to participate in regulation of Introduction APC/C activity [15], and Emi1-mediated M-phase arrest is unaffected by Ca2+, making it distinct from CSF ar- During vertebrate oocyte maturation, the immature oo- rest. In addition, during arrest with Emi1, both cyclins cyte develops into a fertilizable gamete arrested at sec- A and B are stabilized [13], whereas during CSF arrest ond meiotic metaphase while awaiting fertilization. The with Mos or cyclin E/Cdk2, only cyclin B is stabilized mechanism of meiotic metaphase arrest has been [16–19]. A recent report identified a Xenopus Emi1- studied for over 30 yr. In early studies, by microinjecting related protein, XErp1, that contains a Cdc20 binding cytoplasm from a mature oocyte (unfertilized egg) into domain closely related to that present in Emi1 [20]. In- embryonic blastomeres, Masui and Markert found that activation of XErp1, also called Emi2 and referred to as the injected blastomere underwent cleavage arrest at Emi2 hereafter in this paper, led to premature APC/C the next mitosis [1]. Cytological analysis of the arrested activation and CSF release, and addition of excess blastomere revealed that the cell was arrested at meta- Emi2 to CSF extracts prevented release even in the phase, with a single metaphase spindle lacking centro- presence of calcium [20]. Subsequent work showed that some antibodies to Emi1 also coimmunoprecipi- *Correspondence: [email protected] tate Emi2 [21]. This suggests that Emi2, not Emi1, is the Plx1 and CaMK II Cooperate to Promote CSF Release 1459

relevant APC/C inhibitor involved in CSF arrest. Inter- Plx1 Causes CSF Release in the Absence of Calcium estingly, Xenopus Polo-like kinase 1 (Plx1) phosphory- in a Polo-Box-Dependent Manner lates Emi2 at sites that target it for degradation by the So far, all evidence based on loss of Plx1 function indi- SCF pathway, hence promoting CSF release [20]. cates that Plx1 is required for activating the proteolytic The involvement of Plx1 in regulation of CSF arrest machinery for mitotic exit, but it has not been deter- was initially suggested by the observation that the ad- mined whether enhanced expression of Plx1 is suffi- dition of kinase-dead Plx1 to CSF extracts blocked cal- cient to cause CSF release. In this study, active, highly cium-induced CSF release [22–24], indicating a require- purified Plx1 and the two Plx1 mutants were used in the ment for Plx1 activity in APC/C activation. This result Xenopus cell-free system to study the effects of Plx1 on was further supported by the finding that the meta- mitotic exit. To Xenopus egg extracts arrested in metaphase-anaphase transition was blocked by immunode- phase by CSF we added recombinant Plx1 to a level pletion of Plx1, and addition of recombinant Plx1 re- 10–15-fold over the endogenous level and found that stored the transition [23]. The effect of Plx1 on CSF without the addition of calcium, Plx1 alone caused mi- release requires an intact polo-box domain [24, 25]. It totic exit as judged by the degradation of cyclin B1 was suggested that rather than activating the APC/C (Figure 1A) and histone H1 kinase activity (data not directly, Plx1 might work by removing an inhibitor of the shown). The addition of the same amount of Plx1NA APC/C [23]. However, because Plx1 is fully active at did not cause release, indicating Plx1 kinase activity is metaphase [26], it is unclear why Plx1-dependent required for CSF release. It was noted that the CSF re- phosphorylation effects would be restricted to ana- lease caused by Plx1 was somewhat slower than that phase. Under physiological conditions, CSF arrest is re- induced by calcium (Figure 1A), a result similar to that leased upon fertilization, which causes a transient rise seen previously with release induced by CamCat in the in the level of free calcium within the cell. In Xenopus absence of calcium [28]. This is consistent with evi- cell-free extracts, the target of elevated calcium is cal- dence that several different pathways lead to APC/C modulin-dependent protein kinase II (CaMK II); specific inhibition and CSF arrest during oocyte maturation [31], inhibition of CaMK II prevented calcium-induced CSF and presumably at egg fertilization/activation, multiple release, and a constitutively active CaMK II mutant pathways act in concert to activate the APC/C and pro- (CamCat) caused CSF release in extracts even in the mote rapid CSF release. absence of calcium [27–29]. Here, we report that Plx1 We have previously shown that an intact PDB is re- is a necessary component in the pathway of calcium- quired for Plx1 to function properly at multiple points induced CSF release, and recombinant Plx1 protein is during mitosis [24]. In particular, Plx1NA blocked cal- sufficient to trigger CSF release even in the absence of cium-induced CSF release, whereas Plx1NA with a calcium. We also studied the relationship between Plx1 W408F mutation in its first polo box did not block re- and CaMK II because each is necessary for CSF release. This suggested that Plx1NA was competing with lease and sufficient when overexpressed. It was found endogenous Plx1 for PBD binding site(s), and this bind- that Plx1 and CaMK II require each other in order to ing might be required for Plx1 to phosphorylate its sub- induce CSF release in the absence of calcium, and strate(s) and activate the proteolytic machinery. These overexpression of Plx1 elevates calcium-independent results raise the question of whether Plx1-induced CSF activity of CaMK II. Further investigation showed that release is also polo box dependent. To address this is- most likely both Plx1 and CaMK II achieve activation of sue, we added Plx1WF to CSF-arrested Xenopus egg the APC/C and CSF release through dual phosphoryla- extracts; as predicted, the cyclin B1 level remained tion of Emi2 to promote its degradation. high even 80 min after the addition of Plx1WF (Figure 1B), whereas cyclin B1 was undetectable 40 min after Results the addition of wild-type Plx1, indicating the release of CSF activity. New cyclin synthesis restored cyclin B1 to For this study, we generated two Xenopus Plx1 mutants its original level 80 min after Plx1 addition, indicating in addition to wild-type Plx1 (see Figure S1 in the Sup- re-establishment of CSF arrest at the next M phase as plemental Data available with this article online) [24]: expected. This result and previous evidence indicate Plx1NA, a kinase-dead Plx1 whose kinase activity is that both inhibition and induction of CSF release by abolished by a N172A point mutation in its kinase do- Plx1NA and Plx1, respectively, require a functional main [26]; and Plx1WF, a kinase-active Plx1 with a PBD. W408F point mutation in its first polo box, a mutation The above results showed that Plx1 is sufficient to equivalent to the W414F mutation in Plk1, which has induce CSF release in a polo-box-dependent manner. been demonstrated to dramatically abolish the localiza- It is unclear, however, whether Plx1 is part of the path- tion capability of the polo-box domain (PBD) [30]. This way by which calcium and CaMK II induce CSF release. mutation also blocks the effects of Plx1 or Plx1NA on To determine whether Plx1 is required for calcium- multiple mitotic steps, including CSF release, in Xeno- induced CSF release, we first completely immuno- pus [24]. As described in the Experimental Procedures, depleted endogenous Plx1 from metaphase-arrested the Plx1, Plx1NA, and Plx1WF proteins were expressed Xenopus egg extracts (Figure 2A). Upon calcium addi- in insect cells and purified to homogeneity by a combi- tion to the Plx1-depeleted extracts, the cyclin B1 level nation of affinity (Talon beads) and ion-exchange (Mono remained high even 40 min after calcium addition (Fig- S) chromatography (see Supplemental Data). The re- ure 2A), whereas the same amount of calcium caused combinant proteins were highly concentrated, above 5 CSF release in mock-depleted extracts. It was noted mg/ml as measured by the Bradford assay and quantifi- that in the Plx1-depleted CSF extracts, the cyclin B1 cation of stained gels. level remained stable when no calcium was added, Current Biology 1460

Figure 1. Plx1 Induces CSF Release in the Absence of Calcium in a Polo-Box-Depen- dent Manner As a result of the activity of CSF, a Xenopus egg extract (CSF extract) is arrested at metaphase of meiosis II with a high level of cyclin B protein. Upon the addition of calcium, CSF arrest is released, anaphase commences, and B type cyclins are destroyed by the APC/C. (A) To CSF-arrested extracts recombinant Plx1 or Plx1NA was added to a level 15-fold over endogenous Plx1, and CSF release in the absence of calcium was monitored by the degradation of cyclin B1 (upper panel). Control extracts were incubated with or without calcium addition, and cyclin B degradation was monitored as indicated. Lower panel shows immunoblot of His-tagged recombinant Plx1 proteins added to the extracts. (B) In the upper panel, CSF extracts were supplemented with or without calcium, and the level of cyclin B1 was monitored. In the lower panel, recombinant Plx1 or Plx1WF was added to the CSF extracts in the absence of calcium, and the degradation of cyclin B1 was monitored by immunoblotting at the indicated times.

suggesting that Plx1 is not required for maintenance of elucidate the relationship between these two kinases CSF arrest. and, in particular, to determine whether they act in inde- To extend the above loss-of-function result and elimi- pendent pathways or whether one is upstream of the nate the potential codepletion of other important com- other in the pathway of CSF release. To address this ponents, we performed rescue experiments. In sepa- question, we first purified the constitutively active mu- rate experiments, we determined that the addition of tant of CaMK II (CamCat) and characterized its action recombinant Plx1 to CSF extracts to a level 2-fold over in CSF extracts. As shown in Figure S2, purified Cam- the endogenous level did not cause CSF release in the Cat was added to CSF extracts at two final concentra- absence of calcium (data not shown). Therefore, to tions as indicated. At both final concentrations, Cam- Plx1-depleted CSF extracts we added recombinant Cat was able to induce CSF release in the absence of Plx1 to 2-fold over the undepleted endogenous level calcium, consistent with a previous report [28]. For fu- and observed restoration of calcium-induced CSF re- ture experiments, a final concentration of 30 ng/␮l lease (Figure 2B). As controls, the same level of recom- CamCat was used. binant Plx1 did not cause CSF release in the absence Because both Plx1 and CaMK II are necessary for of calcium, and cyclin B remained stable in the Plx1- CSF release, it is possible that they act in the same depleted extracts even when calcium was added (Fig- molecular pathway to control CSF release. In this case, ure 2B). These results support the hypothesis [23] that if one kinase is upstream of another, depletion or inhibi- Plx1 is a necessary component in the pathway of cal- tion of that kinase might block the ability of the other cium-induced CSF release. kinase to induce CSF release. It is also possible that both kinases act at the same level but need the activity Plx1 and CamCat Require Each Other to Induce CSF of each other to pass the signal to downstream tar- Release in a Calcium-Independent Manner get(s), e.g., to phosphorylate their substrate(s). To dis- Previously, it has been reported that CaMK II is required tinguish between these possibilities, we first com- for calcium-induced CSF release [28] because inhibi- pletely immunodepleted Plx1 from CSF extracts (Figure tion of CaMK II blocks CSF release when calcium is 3A) and then added CamCat at a concentration that added [27, 28]. Moreover, a calcium-independent con- causes spontaneous CSF release in mock-depleted ex- stitutively active kinase fragment of CaMK II (CamCat) tracts (Figure 3A). We also added calcium to one ali- is sufficient to induce CSF release in a calcium-inde- quot of the Plx1-depleted extract; although in this ex- pendent manner [28]. Because either Plx1 or CaMK II periment some cyclin B degradation occurred, nuclear expression can induce CSF release, it is important to morphology verified that this was not sufficient to Plx1 and CaMK II Cooperate to Promote CSF Release 1461

Figure 2. Depletion of Plx1 Blocks Calcium- Induced CSF Release (A) Complete depletion of Plx1 from CSF extracts is shown in the left panel. The right panel shows the level of cyclin B1 at the indicated times in extracts depleted of wild-type Plx1 in the presence and absence of calcium. (B) The left panel shows depletion of Plx1. The right panel shows cyclin B1 levels in Plx1-depleted extracts with or without calcium addition and with or without an add- back of Plx1 to a final level 2-fold over the wild-type level. Similar results were obtained in several independent experiments.

cause CSF release and exit from M-phase, demonstrat- lated on Thr286 is known to exhibit activity even in the ing functional depletion of Plx1. It was found that Cam- absence of calcium [32], this suggests the possibility Cat could not trigger CSF release in Plx1-depleted ex- that Plx1 promotes phosphorylation of Thr286 to gener- tracts, indicating that CaMK II is not a downstream ate calcium-independent activity of CaMK II. The component of the pathway of Plx1-mediated CSF re- mechanism is not yet clear but is likely to be indirect lease. To extend this observation, we added CamCat to inasmuch as CaMK II is not a substrate for Plx1 in vitro a CSF extract containing Plx1NA to determine whether (data not shown). CamCat can overcome inhibition of CSF release by kinase-dead Plx1. In accordance with the Plx1 depletion Overexpression of Xenopus Mad2 Blocks CSF result, CamCat was unable to overcome inhibition of Release Induced by either Plx1 or CamCat CSF release by Plx1NA (Figure 3B). Overexpression of Mad2 blocks calcium-induced CSF Next, we employed the CaMK II-specific inhibitor release by binding the APC/C activator Fzy/Cdc20 peptide, 281–309, to inhibit endogenous CaMK II activ- and inhibiting Fzy/Cdc20-dependent activation of the ity [27] in order to evaluate whether Plx1 is able to trig- APC/C [6]. When overexpressed, both Plx1 and Cam- ger CSF release in the absence of CaMK II activity. As Cat are sufficient to induce CSF release in the absence shown in Figure 3C, the inhibitor peptide 281–309 not of calcium, indicating their ability to ultimately activate only blocked calcium-induced CSF release as reported the APC/C for degradation of mitotic regulators. There- previously [27] but also inhibited the ability of Plx1 to fore, it was important to determine whether overexpres- induce CSF release. This latter event was surprising be- sion of Mad2 would block CSF release triggered by cause it suggested that basal activity of CaMK II was Plx1 and/or CamCat because this would help clarify necessary for Plx1-induced CSF release. To investigate whether the APC/C itself was a direct target of Plx1 this point further, we performed CaMK II assays in ex- and/or CamCat. When recombinant Xenopus Mad2 tracts by using a specific peptide substrate that en- was added to CSF extracts, CSF release induced by compasses the sequence around Thr286 in the auto- calcium was completely blocked (Figure 5Ab). In addi- regulatory region of CaMK II, which is associated with tion, the amount of Plx1 and CamCat able to induce enzyme activation [32]. As shown in Figure 4, calcium CSF release in the absence of calcium (Figure 5Aa) addition to a CSF extract led to rapid CaMK II activa- failed to trigger CSF release in the presence of Mad2 tion, peaking at 1–2 min and returning to a sustained (Figure 5Ab). This result was also verified by sperm activity level above basal by 5 min. Studies of CaMK II nuclear morphology (Figure 5A, lower panel). This result in other systems have suggested the sustained activity strongly suggests that neither Plx1 nor CaMK II pro- of CaMK II long after a calcium stimulus reflects the motes CSF release by directly activating the APC/C; activity of CaMK II phosphorylated at Thr286 [32]. rather, they most likely regulate the APC/C through acti- These kinetics are similar to those reported previously vating regulator(s) such as Cdc20 or removing inhibi- for CaMK II during CSF release with a phosphospecific tors that affect Cdc20-dependent activation. This con- antibody to Thr286 [27]. Basal activity against the cept is consistent with evidence that antisense ablation CaMK II substrate (inhibitable by the 281–309 peptide) of Fzy/Cdc20 in maturing oocytes blocks egg activation was not detectable in a control extract, but remarkably, (CSF release) by the calcium ionophore A23187 [33] CaMK II activity was significantly generated upon addi- and that the addition of antibody to Fzy/Cdc20 to CSF tion of Plx1 (Figure 4). The increase was completely extracts blocks Ca2+-induced CSF release [5]. How- blocked by inclusion of the inhibitor peptide 281–309 ever, our study does not exclude the possibility that but was largely insensitive to the inclusion of BAPTA either kinase might phosphorylate subunits of the and EGTA, indicating elevation of the calcium-indepen- APC/C at sites that are insufficient for activation. dent activity of CaMK II. Because CaMK II phosphory- To further verify action through effects on Fzy/Cdc20, Current Biology 1462 Plx1 and CaMK II Cooperate to Promote CSF Release 1463

Combined Action of Plx1 and CamCat Overcomes Inhibition of CSF Release by Emi2 but Not by Mad2 If both Plx1 and CaMK II act in the same pathway on different components or on two independent pathways, then the combined action of both kinases might be sufficient to overcome arrest by either Mad2 or Emi2. We have shown that alone neither Plx1 nor CamCat is able

to overcome Mad2 inhibition of the APC/C (Figure 5Ab). Further work confirmed that Emi2 inhibits Fzy/Cdc20- APC/C activity in the presence of calcium, consistent with a previous report [20]. We found that Emi2 also inhibited CSF release induced by either Plx1 or CamCat (Figure 6A). However, as shown in Figure 6B, the combination of both kinases together did overcome the APC/C inhibition imposed by expression of Emi2 but not by Mad2, suggesting that no additional kinase activities are required to overcome inhibition of APC/C activation by Emi2. It has been demonstrated that the Plx1 phosphorylation-dependent degradation of Emi2 requires the PBD [20]. The PBD is thought to bind a phosphopeptide motif in target proteins under conditions in which a “priming kinase” has phosphorylated the substrate first to promote docking of the PBD, and this docking en- hances subsequent Plx1-dependent phosphorylation Figure 4. CaMK II Activity in Egg Extracts Is Increased by Plx1 [34, 35]. The fact that neither Plx1 nor CamCat alone was able to overcome the inhibition of Emi2 on the As indicated, CSF extracts were supplemented with either buffer, calcium, Plx1 + (281–309), Plx1 + (BAPTA + EGTA [10 mM each]), APC/C, whereas the combination of both kinases did or Plx1 alone, and endogenous CaMK II activity in the absence of overcome the inhibition, led us to suspect that CamCat added calcium was measured as described in the Experimental might act as a priming kinase for Plx1 to generate a Procedures. The level of CaMK II activity upon buffer addition was binding site for the PBD, leading to Plx1-dependent not reduced by inclusion of the 281–309 inhibitor peptide (data phosphorylation of Emi2 and facilitation of its degrada- not shown). tion. To test our hypothesis, we carried out a series of in vitro phosphorylation reactions. First, GST-Emi2 was found to be a good substrate for CaMK II, whereas GST we immunodepleted Fzy/Cdc20 from CSF extracts. As alone was only marginally phosphorylated (Figure 6C). shown in Figure 5B, there was no detectable Fzy/Cdc20 Next, GST-Emi2 coupled to Dynabeads was initially remaining in Fzy/Cdc20-depleted CSF extracts, and phosphorylated by CaMK II in an assay with unlabeled calcium addition to the Fzy/Cdc20-depleted extracts ATP; after several washes of the complex, the GST- did not cause CSF release (Figure 5Cb), demonstrating Emi2 was then incubated with Plx1 in the presence of the efficiency of Fzy/Cdc20 depletion. This result is [γ-32P]ATP. It was found that phosphorylation of GST- consistent with a previous report that depletion of Fzy/ Emi2 by Plx1 was clearly enhanced after pretreatment Cdc20 blocks CSF release induced by constitutively with CaMK II (Figure 6D, lower panel). At this level of active CaMK II [5]. As expected, addition of either Plx1 exposure, no phosphorylation of Emi2 by Plx1 is evi- or CamCat to the Fzy/Cdc20-depleted extracts also did dent in the absence of prior phosphorylation by CaMK not trigger CSF release (Figure 5Cb), whereas the same II. However, other experiments have confirmed that amount of Plx1 or CamCat caused CSF release in some Plx1-dependent phosphorylation of Emi2 can oc- mock-depleted extracts (Figure 5Ca). The result was cur without CaMK II treatment under some conditions, also confirmed by examining sperm nuclear morphol- as reported previously [20]. The increased autophos- ogy. This result further supports the hypothesis that phorylation of Plx1 in the presence of Emi2 that had both Plx1 and the CaMK II promote CSF release by reg- undergone priming phosphorylation by CaMK II (Figure ulating APC/C regulator(s) rather than by activating the 6D) may reflect loss of autoinhibition by the PBD when APC/C directly. it binds a phosphomotif [36]. Importantly, in egg ex-

Figure 3. Plx1 and CamCat Require Each Other to Induce CSF Release (A) Panel a shows depletion of endogenous Plx1. In (b), CamCat-induced release was monitored with or without Plx1 depletion, and the corresponding nuclear morphology is shown in the lower panel. (B) CSF release by CamCat was monitored in the presence of Plx1NA, as indicated. The corresponding nuclear morphology is shown in the lower panel at the indicated times. Note that the partial degradation of cyclin B in this experiment was insufficient to cause exit from M phase. (C) In the absence or presence of the CaMK II inhibitor 281–309 at a final concentration of 0.4 mM, CSF release induced by both calcium and Plx1 was analyzed by immunoblotting cyclin B1. The corresponding nuclear morphology is shown in the lower panel at the indicated times. Current Biology 1464

Figure 5. Mad2 Inhibits Fzy/Cdc20-Dependent CSF Release by Either Plx1 or CamCat (A) In (a), induction of CSF release in the absence of calcium by either CamCat or recombinant Plx1 protein was assessed by the level of cyclin B1 (upper panel) or changes in sperm nuclear morphology (lower panel). In (b), in the same experiment, recombinant Xenopus Mad2, which completely blocked calcium-induced CSF release, also blocked CSF release triggered by either the CamCat or recombinant Plx1 (upper panel). The middle panel shows the level of recombinant Mad2 added as blotted by anti-His antibody. Nuclear morphology is shown in the lower panel. (B) Fzy/Cdc20 was not detectable in Fzy/Cdc20-depleted CSF extracts. The lower bands demonstrate the specificity of the immunodepletion. (C) In (a), both CamCat and recombinant Plx1 were able to induce CSF release in the absence of calcium in mock-depleted CSF extracts. In (b), depletion of Fzy/Cdc20 from the extracts totally blocked CSF release induced by calcium, CamCat, or recombinant Plx1, as indicated by the stable cyclin B1 level and nuclear morphology. tracts, Emi2 was degraded only when both Plx1 and pendent pathways [31]. It is evident that regulation of CaMK II were active and able to cause CSF release APC/C activity is also a key element in the maintenance (Figure 6E). This result strongly supports a model in and release of CSF arrest. Plx1 has previously been which CaMK II acts as a priming kinase for Plx1 phos- shown to be required for mitotic exit. In this report, we phorylation of Emi2 at sites that target it for degrada- used the Xenopus cell-free system to demonstrate that tion and presumably release of Fzy/Cdc20 for activa- Plx1 is not only necessary but also sufficient when over- tion of the APC/C [20]. expressed to trigger the metaphase-to-anaphase transition and CSF release. Wild-type recombinant Plx1 and a Discussion constitutively active mutant of Plx1 have been added to Xenopus CSF extracts previously, but no effect on Several lines of evidence indicate that CSF activity is maintenance of CSF arrest was evident after the addi- established by inhibition of the APC/C by multiple inde- tion (Y.-W. Qian, J.L., and J.L.M., unpublished data). Plx1 and CaMK II Cooperate to Promote CSF Release 1465

Figure 6. Combined Action of Plx1 and Cam- Cat Overcomes Inhibition by Emi2 but Not Mad2 (A) CSF release was monitored by immunoblotting cyclin B1 at the indicated times in extracts supplemented with Plx1 or CamCat. (B) The experiment in (A) was repeated ex- cept that both kinases together were added, and release was monitored either in the presence of Emi2 or Mad2. (C) An equimolar amount of GST-Emi2 or GST alone was incubated with either buffer or CaMK II under the phosphorylation condi- tion described in the Experimental Pro- cedures. The GST-Emi2 is substantially phosphorylated by CaMK II, whereas GST is only marginally phosphorylated. An autora- diograph is shown. Similar results were obtained when CamCat was used instead of CaMK II (data not shown). (D) The upper panel shows loading of Emi2 or Plx1 where applicable. The lower panel shows Plx1-dependent phosphorylation of Emi2 or CaMK II-treated Emi2. (E) The upper panel shows CSF release induced by combined action of both Plx1 and CamCat as monitored by a cyclin B2 blot. The middle panel shows the degradation of overexpressed GST-Emi2 in the presence of both Plx1 and CamCat, whereas the GST- Emi2 is stable in extracts to which only calcium is added. Other experiments demonstrated that the addition of either CamCat or Plx1 alone did not cause degradation of overexpressed GST-Emi2 (data not shown). The lower panel shows the degradation of endogenous Emi2 upon calcium addition to the same extract.

Now it is clear that this was due to insufficient expres- However, this model raises a question. Because Plx1 sion levels, only about 3-fold over the endogenous activity is high during meiotic metaphase II arrest, why level. In this study, we added highly purified recombi- would a Plx1 substrate(s), e.g., Emi2, not be phosphor- nant Plx1 to CSF extracts at a final concentration of 10– ylated until after the addition of calcium? There are sev- 15-fold over the endogenous Plx1 level and achieved CSF eral possibilities. One explanation is that Plx1 and the release even in the absence of calcium. This result sug- substrate(s) are localized differentially, such that only gests that a substrate of Plx1 normally not accessible to when Plx1 is translocated to the site where the sub- Plx1 before fertilization or calcium addition was phos- strate(s) resides will the phosphorylation occur. Sup- phorylated by excess Plx1 because the kinase activity porting evidence for this scenario comes from the of Plx1 was required for CSF release. Moreover, an in- translocation of Plx1 during mitosis, at which time Plx1 tact polo box of Plx1 was also shown to be required for moves from centrosomes and spindle poles to the Plx1 function, indicating that a PBD-mediated interac- metaphase plate and finally to the midbody during cy- tion regulates phosphorylation of the substrate. These tokinesis [39, 40]. However, this cannot explain the results are consistent with the recent finding that Plx1 requirement for Plx1 for CSF release (APC/C activation) phosphorylates Emi2 and targets it for degradation, in extracts without added DNA, where a mitotic spindle thereby relieving inhibition of the APC/C and promoting does not exist. This suggests a second possibility, that CSF release [20]. Although the involvement of Emi1 in the polo-box-dependent interaction between Plx1 and CSF is controversial, it was also shown that Plx1 regu- its substrate(s) is dependent upon a third element. A lates the stability of Emi1 in a similar manner [37, 38]. plausible model involves a priming kinase. It has been In the case of Emi2, its interaction with Plx1 was shown reported that the PBD recognizes an optimal sequence to be polo box dependent, and depletion of Emi2 abol- containing Ser and Thr, and binding of the PBD to the ished the ability of Plx1NA to block calcium-induced motif may require phosphorylation by a priming kinase CSF release [20]. distinct from Plx1 itself [34, 35]. In support of this Current Biology 1466

model, it has been reported that the interactions between Plx1 and Cdc25C [41] and between Plx1 and the checkpoint mediator protein Claspin [42] are PBD- dependent and that binding of the PBD to Cdc25C or Claspin requires prior phosphorylation by a distinct priming kinase to generate a docking site for the PBD. In the case of regulation of CSF activity, it is possible that the phosphorylation of Plx1 substrate(s) also requires priming by a distinct kinase to generate a PBD binding site in the substrate itself (Figure 7) that brings the Plx1 catalytic domain into proximity with its substrate. Because the interaction between Plx1 and Emi2 is PBD-dependent and because CaMK II is required for Plx1 to induce CSF release, it seems likely that CaMK II may act as a priming kinase for phosphorylation of Figure 7. Dual Regulation of CSF Release by Plx1 and CaMK II Emi2 to generate a docking site for the PBD, thereby Emi2 inhibits the activation of the APC/C by Fzy/Cdc20. Upon the promoting the Plx1-dependent phosphorylation of addition of calcium, CaMK II is activated and phosphorylates Emi2 (priming), generating a docking site for the PBD of Plx1, which sub- Emi2 at sites that target it for degradation, leading to sequently binds to and further phosphorylates Emi2. The phos- activation of the APC/C and CSF release. This model phorylation by Plx1 targets Emi2 for degradation by the SCF (Skp1- gains strong support from the finding that CaMK II was Cullin F-box) ubiquitin ligase complex, leading to CSF release. able to phosphorylate Emi2 in vitro (Figure 6C) and that this phosphorylation in turn enhanced the ability of Plx1 to phosphorylate the APC/C inhibitor. Moreover, only vate calcium-independent CaMK II activity explains the combined action of both kinases in extracts was why the inhibitor peptide 281–309 was able to block able to cause degradation of Emi2 and overcome CSF Plx1-dependent CSF release, and it reinforces the con- arrest (Figure 6E). These results explain why Plx1 is cept in our model (Figure 7) that Plx1 and CaMK II re- needed for CSF release at anaphase even though the quire each other to cooperate in relief of APC/C inhibi- enzyme is fully active at metaphase during CSF arrest, tion by Emi2. In addition to activation of CaMK II by and they also identify a key target of CaMK II for cal- calcium, it is evident that there is an apparent positive cium-induced CSF release at fertilization. feedback loop between these kinases during CSF re- However, neither Plx1 nor CamCat alone or the com- lease, a loop in which Plx1 may promote elevated activ- bination of both kinases was able to overcome inhibi- ity of CaMK II, which then phosphorylates Emi2 to pro- tion of the APC/C by Mad2. Mad2 inhibits APC/C activ- mote PBD binding to Emi2 and elevation of Plx1 activity ity by binding to the positive APC/C regulator Fzy/ toward Emi2 (Figure 6). This positive feedback loop Cdc20. It has been reported that the binding of Mad2 may contribute to the rapidity of APC/C activation at to Fzy/Cdc20-APC/C is regulated by phosphorylation fertilization because Plx1 is already fully active at meta- [43], and phosphorylated Mad2 is unable to bind to phase [26]. Fzy/Cdc20. Although we found that both Plx1 and CaMK II were able to phosphorylate Mad2 in vitro to a Conclusions certain extent (data not shown), phosphorylation by Plx1 has been previously shown to be required for the both kinases in CSF-arrested extracts did not relieve metaphase-to-anaphase transition. Here, we have shown Mad2 inhibition of APC/C activation by Fzy/Cdc20, that overexpression of Plx1 alone is sufficient to trigger suggesting that another pathway is responsible for activation of the APC/C and release from CSF arrest. Mad2 phosphorylation/inactivation. It is notable that By studying the relationship between Plx1 and CaMK depletion of Mad2 does not trigger CSF release [9], II, a kinase previously demonstrated also to be suffi- whereas depletion of Emi2 does cause release [20], in- cient to trigger CSF release, we found that Plx1 and dicating that Emi2 is sufficient to inhibit Fzy/Cdc20 and CaMK II are mutually dependent upon each other to maintain CSF arrest in the absence of calcium. Another induce CSF release. It is therefore evident that both pathway involved in CSF arrest is cyclin E/Cdk2, whose Plx1 and CaMK II cooperate to promote activation of ability to inhibit the APC/C is removed by Tyr15 phos- the APC/C and the release of CSF arrest. It was demon- phorylation of Cdk2 after calcium addition [11]. It is strated that, rather than activating the APC/C directly, possible the partial cyclin B degradation seen in some both Plx1 and CaMK II activate the complex by regulat- extracts in the absence of Plx1 (Figure 3A), albeit not ing Fzy/Cdc20. A plausible model involves CaMK II as sufficient for M phase exit, reflects calcium-dependent a priming kinase to generate a docking site for PBD inactivation of the Mad2 or cyclin E/Cdc2 APC/C inhibi- binding, which enables Plx1 to phosphorylate its sub- tory pathway. strate(s) and activate the APC/C. One candidate sub- It is clear that calcium- and CaMK II-induced CSF strate is Emi2, whose interaction with Plx1 is PBD- release requires Plx1, demonstrating that Plx1 is proba- dependent and whose phosphorylation by Plx1 targets bly a physiological component of the CSF regulation it for ubiquitin-mediated degradation. Our data show machinery. In any case, a surprising result was that CSF that phosphorylation of Emi2 by Plx1 is substantially release in response to Plx1 overexpression still requires enhanced if the substrate is prephosphorylated by CaMK II because the CaMK II inhibitory peptide CaMK II, and the combination of Plx1 and CamCat can blocked the effects of Plx1. The ability of Plx1 to ele- overcome inhibition of CSF release by Emi2 by tar- Plx1 and CaMK II Cooperate to Promote CSF Release 1467

geting it for degradation. The dual phosphorylation of ml DAPI), and sperm nuclei morphology was monitored by fluores- Emi2 by Plx1 and CaMK II is further enhanced by the cence microscopy. One microliter of sample at each time point was ability of Plx1 to activate CaMK II activity in a calcium- mixed with 20 µl of Laemmli Sample Buffer (BioRad) and boiled for 5 min. For immunoblotting, 10 µl of each sample, equivalent to 0.5 independent manner. µl of extract, was used. Cyclins B1 and B2 were detected with sheep anti-cyclin B antibodies prepared as described previously Experimental Procedures [44], and Emi2 was monitored with rabbit anti His-Emi2 (1–374) antibodies that were affinity purified on a column of GST-Emi2. Constructs and Protein Purification Plx1 and its mutants Plx1NA and Plx1WF were generated as described previously [24]. A baculovirus encoding a constitutively Supplemental Data active fragment (amino acids 1-280) of CaMK II (CamCat) was a gift Supplemental Data include two figures and are available with this from M.D. Browning (Department of Pharmacology, UCHSC). The article online at: http://www.current-biology.com/cgi/content/full/ Plx1, Plx1NA, or Plx1WF proteins were purified with Talon beads, 15/16/1458/DC1/. and this was followed by Mono S chromatography as described previously [26]. CamCat bearing a T7 tag was purified with a T7 Acknowledgments Tag Affinity Purification Kit (Novagen) following the manufacturer’s protocol. Emi2 was cloned by PCR amplifying the DNA from a Xen- We thank Mike Browning (Department of Pharmacology, UCHSC) opus stage VI oocyte cDNA library (Stratagene) with the primers for providing a baculovirus encoding CamCat and Ulli Bayer (De- 5#-GACGACGACAAGATGGCAAATCTCTTAGAG-3# and 5#-GAGGA partment of Pharmacology) for helpful discussions about CaMK II. GAAGCCCGGTCTAGCTTCAAAGTCTC-3#, with subsequent clon- We are very grateful to Eleanor Erikson for help with chromato- ing into the pET41 ek/LIC vector (Novagen) following the manufac- graphic purification of recombinant proteins and for a critical read- turer’s protocol. This vector produces a recombinant protein having ing of the manuscript. DNA sequencing and protein expression in both GST and 6His tags. GST-Emi2 was purified with Glutathione Sf9 cells were performed by the University of Colorado Cancer Sepharose 4B beads (Amersham Biosciences). The pET41 GST Center Core facility (CA46930). was prepared by cleaving a portion of GST-Emi2 on the Glutathione beads with thrombin protease (Amersham Biosciences). After several washes with PBS, GST was eluted from the beads with 10 mM Received: May 18, 2005 glutathione (Sigma). Revised: June 30, 2005 Accepted: July 8, 2005 In Vitro Phosphorylation Published online: July 21, 2005 For in vitro phosphorylation of GST-Emi2, equal amounts of GST- Emi2 (0.3 ␮M) and pET41 GST (0.3 ␮M) were incubated with either References buffer or 50 ng of CaMK II (Upstate) in the presence of 0.5 mM 32 CaCl2,5␮M calmodulin (Calbiochem), and 100 ␮M[γ− P]ATP 1. Masui, Y., and Markert, C.L. (1971). Cytoplasmic control of (1500 cpm/pmole) at 30°C for 10 min. For in vitro phosphorylation nuclear behavior during meiotic maturation of frog oocytes. J. of either buffer-treated or CaMK II-treated GST-6His Emi2 by Plx1, Exp. Zool. 177, 129–145. an equal amount of Emi2 was bound to Dynabeads Protein G (Dy- 2. Peters, J.M. (1998). SCF and APC: The Yin and Yang of cell nal) precoupled with anti-His monoclonal antibody (Sigma) and in- cycle regulated proteolysis. Curr. Opin. Cell Biol. 10, 759–768. cubated as above in the presence of 100 ␮M cold ATP with either 3. Peters, J.M. (1999). Subunits and substrates of the anaphase- buffer or 50 ng of CaMK II in buffer containing calcium and calmod- promoting complex. Exp. Cell Res. 248, 339–349. ulin (Calbiochem). After washing the Dynabeads/Emi2 complex 5 4. Walsch, R., and Engelbert, D. (2005). Anaphase-promoting times with PBS, the beads were incubated with an equal amount complex-dependent proteolysis of cell cycle regulators and (100 ng) of Plx1 in the presence of 100 ␮M[γ−32P]ATP at 30°C for genomic instability of cancer cells. Oncogene 24, 1–10. 10 min. The reaction was stopped by addition of 4× SDS gel sample 5. Lorca, T., Castro, A., Martinez, A.M., Vigneron, S., Morin, N., buffer, electrophoresed on a 10% SDS-PAGE gel, and analyzed by Sigrist, S., Lehner, C., Doree, M., and Labbe, J.C. (1998). Fizzy autoradiography or immunoblotting. For CaMK II assay, the CaM is required for activation of the APC/cyclosome in Xenopus egg Kinase II Assay Kit (Upstate) was used, and the manufacturer’s pro- extracts. EMBO J. 17, 3565–3575. tocol was followed. The autocamtide 2 substrate (KKALRRQETV- 6. Fang, G., Yu, H., and Kirschner, M.W. (1998). The checkpoint DAL) is based on the sequence around CaMK II’s autoregulatory protein MAD2 and the mitotic regulator CDC20 form a ternary phosphorylation site (Thr286), which is associated with activation complex with the anaphase-promoting complex to control ana- of calcium-independent CaMK II activity [32]. Other experiments phase initiation. Genes Dev. 12, 1871–1883. demonstrated that Plx1 did not phosphorylate autocamtide 2 spe- 7. Li, Y., Gorbea, C., Mahaffey, D., Rechsteiner, M., and Benezra, cifically, and Plx1 activity toward casein was not inhibited by the R. (1997). MAD2 associates with the cyclosome/anaphase-pro- 281–309 peptide (data not shown). moting complex and inhibits its activity. Proc. Natl. Acad. Sci. USA 94, 12431–12436. Preparation and Manipulation of CSF Extracts, Sperm 8. Yu, H. (2002). Regulation of APC-Cdc20 by the spindle check- Nuclear Morphology, and Immunoblotting point. Curr. Opin. Cell Biol. 14, 706–714. CSF extracts were prepared as described previously [44]. These 9. Tunquist, B.J., Eyers, P.A., Chen, L.G., Lewellyn, A.L., and extracts are arrested at metaphase of meiosis II with a high level Maller, J.L. (2003). Spindle checkpoint proteins Mad1 and of cyclin B/Cdc2 activity as a result of the activity of CSF. Upon Mad2 are required for cytostatic factor-mediated metaphase the addition of calcium, the arrest is released, causing exit from arrest. J. Cell Biol. 163, 1231–1242. metaphase and proteolysis of mitotic cyclins by the APC/C. Each 10. Gabrielli, B.G., Roy, L.M., and Maller, J.L. (1993). Requirement individual reaction was made by transferring on ice 40 µl of extract for Cdk2 in cytostatic factor-mediated metaphase II arrest. Sci- into a 0.5 ml microcentrifuge tube containing about 8000 Xenopus ence 259, 1766–1769. sperm nuclei and then adding buffer or recombinant proteins to the 11. Tunquist, B.J., Schwab, M.S., Chen, L.G., and Maller, J.L. following final concentrations unless otherwise indicated: 150 ng/ (2002). The spindle checkpoint kinase bub1 and cyclin e/cdk2 µl for Plx1, Plx1NA, or Plx1WF; 30 ng/µl for CamCat; and 120 ng/µl both contribute to the establishment of meiotic metaphase ar- for Mad2. After a 10 min incubation on ice, CSF arrest was released rest by cytostatic factor. Curr. Biol. 12, 1027–1033. by the addition of 0.4 µl of 50 mM CaCl2 when applicable, and 12. Reimann, J.D., and Jackson, P.K. (2002). Emi1 is required for incubation was carried out at 21°C. Samples of 1 µl were taken at cytostatic factor arrest in vertebrate eggs. Nature 416, 850– the indicated times, mixed with 4 µl of fixation/stain buffer (48% 854. glycerol, 11% formaldehyde, 1× MMR [100 mM NaCl, 2 mM KCl, 13. Reimann, J.D., Gardner, B.E., Margottin-Goguet, F., and Jack-

1 mM MgSO4, 2 mM CaCl2, and 5 mM HEPES (pH 7.5)], and 1 µg/ son, P.K. (2001). Emi1 regulates the anaphase-promoting com- Current Biology 1468

plex by a different mechanism than Mad2 proteins. Genes Dev. 35. Elia, A.E., Rellos, P., Haire, L.F., Chao, J.W., Ivins, F.J., Hoepker, 15, 3278–3285. K., Mohammad, D., Cantley, L.C., Smerdon, S.J., and Yaffe, 14. Reimann, J.D., Freed, E., Hsu, J.Y., Kramer, E.R., Peters, J.M., M.B. (2003). The molecular basis for phosphodependent sub- and Jackson, P.K. (2001). Emi1 is a mitotic regulator that in- strate targeting and regulation of Plks by the Polo-box domain. teracts with Cdc20 and inhibits the anaphase promoting com- Cell 115, 83–95. plex. Cell 105, 645–655. 36. Jang, Y.J., Lin, C.Y., Ma, S., and Erikson, R.L. (2002). Functional 15. Ohsumi, K., Koyanagi, A., Yamamoto, T.M., Gotoh, T., and Kish- studies on the role of the C-terminal domain of mammalian imoto, T. (2004). Emi1-mediated M-phase arrest in Xenopus polo-like kinase. Proc. Natl. Acad. Sci. USA 99, 1984–1989. eggs is distinct from cytostatic factor arrest. Proc. Natl. Acad. 37. Hansen, D.V., Loktev, A.V., Ban, K.H., and Jackson, P.K. (2004). Sci. USA 101, 12531–12536. Plk1 regulates activation of the anaphase promoting complex 16. Whitfield, W.G., Gonzalez, C., Maldonado-Codina, G., and by phosphorylating and triggering SCFbetaTrCP-dependent Glover, D.M. (1990). The A- and B-type cyclins of Drosophila destruction of the APC Inhibitor Emi1. Mol. Biol. Cell 15, are accumulated and destroyed in temporally distinct events 5623–5634. that define separable phases of the G2-M transition. EMBO J. 38. Moshe, Y., Boulaire, J., Pagano, M., and Hershko, A. (2004). 9, 2563–2572. Role of Polo-like kinase in the degradation of early mitotic in- 17. Bastians, H., Topper, L.M., Gorbsky, G.L., and Ruderman, J.V. hibitor 1, a regulator of the anaphase promoting complex/ (1999). Cell cycle-regulated proteolysis of mitotic target pro- cyclosome. Proc. Natl. Acad. Sci. USA 101, 7937–7942. teins. Mol. Biol. Cell 10, 3927–3941. 39. Barr, F.A., Sillje, H.H., and Nigg, E.A. (2004). Polo-like kinases 18. den Elzen, N., and Pines, J. (2001). Cyclin A is destroyed in and the orchestration of cell division. Nat. Rev. Mol. Cell Biol. prometaphase and can delay chromosome alignment and ana- 5, 429–440. phase. J. Cell Biol. 153, 121–136. 40. Glover, D.M., Hagan, I.M., and Tavares, A.A. (1998). Polo-like 19. Geley, S., Kramer, E., Gieffers, C., Gannon, J., Peters, J.M., and kinases: A team that plays throughout mitosis. Genes Dev. 12, Hunt, T. (2001). Anaphase-promoting complex/cyclosome- 3777–3787. dependent proteolysis of human cyclin A starts at the begin- 41. Perdiguero, E., Pillaire, M.J., Bodart, J.F., Hennersdorf, F., ning of mitosis and is not subject to the spindle assembly Frodin, M., Duesbery, N.S., Alonso, G., and Nebreda, A.R. checkpoint. J. Cell Biol. 153, 137–148. (2003). Xp38gamma/SAPK3 promotes meiotic G2/M transition 20. Schmidt, A., Duncan, P.I., Rauh, N.R., Sauer, G., Fry, A.M., Nigg, in Xenopus oocytes and activates Cdc25C. EMBO J. 22, E.A., and Mayer, T.U. (2005). Xenopus polo-like kinase Plx1 reg- 5746–5756. ulates XErp1, a novel inhibitor of APC/C activity. Genes Dev. 42. Yoo, H.Y., Kumagai, A., Shevchenko, A., and Dunphy, W.G. 19, 502–513. (2004). Adaptation of a DNA replication checkpoint response 21. Tung, J.J., and Jackson, P.K. (2005). Emi1 Class of Proteins depends upon inactivation of Claspin by the Polo-like kinase. Regulate Entry into Meiosis and the Meiosis I to Meiosis II Cell 117, 575–588. Transition in Xenopus oocytes. Cell Cycle 4, 478–482. 43. Wassmann, K., Liberal, V., and Benezra, R. (2003). Mad2 phos- 22. Qian, Y.W., Erikson, E., and Maller, J.L. (1999). Mitotic effects phorylation regulates its association with Mad1 and the APC/ of a constitutively active mutant of the Xenopus polo-like ki- C. EMBO J. 22, 797–806. nase Plx1. Mol. Cell. Biol. 19, 8625–8632. 44. Izumi, T., Walker, D.H., and Maller, J.L. (1992). Periodic changes 23. Descombes, P., and Nigg, E.A. (1998). The polo-like kinase Plx1 in phosphorylation of the Xenopus cdc25 phosphatase regu- is required for M phase exit and destruction of mitotic regula- late its activity. Mol. Biol. Cell 3, 927–939. tors in Xenopus egg extracts. EMBO J. 17, 1328–1335. 24. Liu, J., Lewellyn, A.L., Chen, L.G., and Maller, J.L. (2004). The polo box is required for multiple functions of Plx1 in mitosis. J. Biol. Chem. 279, 21367–21373. 25. Liu, J., and Maller, J.L. (2005). Xenopus Polo-like kinase Plx1: A multifunctional mitotic kinase. Oncogene 24, 238–247. 26. Qian, Y.W., Erikson, E., Li, C., and Maller, J.L. (1998). Activated polo-like kinase Plx1 is required at multiple points during mitosis in Xenopus laevis. Mol. Cell. Biol. 18, 4262–4271. 27. Matsumoto, Y., and Maller, J.L. (2002). Calcium, calmodulin, and CaMKII requirement for initiation of centrosome duplica- tion in Xenopus egg extracts. Science 295, 499–502. 28. Lorca, T., Cruzalegui, F.H., Fesquet, D., Cavadore, J.C., Mery, J., Means, A., and Doree, M. (1993). Calmodulin-dependent protein kinase II mediates inactivation of MPF and CSF upon fertilization of Xenopus eggs. Nature 366, 270–273. 29. Lorca, T., Abrieu, A., Means, A., and Doree, M. (1994). Ca2+ is involved through type II calmodulin-dependent protein kinase in cyclin degradation and exit from metaphase. Biochim. Bio- phys. Acta 1223, 325–332. 30. Lee, K.S., Grenfell, T.Z., Yarm, F.R., and Erikson, R.L. (1998). Mutation of the polo-box disrupts localization and mitotic functions of the mammalian polo kinase Plk. Proc. Natl. Acad. Sci. USA 95, 9301–9306. 31. Tunquist, B.J., and Maller, J.L. (2003). Under arrest: Cytostatic factor (CSF)-mediated metaphase arrest in vertebrate eggs. Genes Dev. 17, 683–710. 32. Miller, S.G., Patton, B.L., and Kennedy, M.B. (1988). Sequences of autophosphorylation sites in neuronal type II CaM kinase that control Ca2+-independent activity. Neuron 1, 593–604. 33. Taieb, F.E., Gross, S.D., Lewellyn, A.L., and Maller, J.L. (2001). Activation of the anaphase-promoting complex and degradation of cyclin B is not required for progression from Meiosis I to II in Xenopus oocytes. Curr. Biol. 11, 508–513. 34. Elia, A.E., Cantley, L.C., and Yaffe, M.B. (2003). Proteomic screen finds pSer/pThr-binding domain localizing Plk1 to mitotic substrates. Science 299, 1228–1231.