Mechanism for inactivation of the mitotic inhibitory kinase Wee1 at M phase

Kengo Okamoto and Noriyuki Sagata
PNAS March 6, 2007 104 (10) 3753-3758; https://doi.org/10.1073/pnas.0607357104

  1. Edited by Tony Hunter, The Salk Institute for Biological Studies, La Jolla, CA, and approved January 5, 2007 (received for review August 24, 2006)

Abstract

Wee1, the inhibitory kinase of cyclin B/Cdc2, undergoes a phosphorylation-dependent catalytic inactivation at M phase of the mitotic cell cycle, but the precise mechanism for this inactivation is not known. Using Xenopus egg and extract systems, we show here that the kinase activity of Xenopus somatic Wee1 (XeWee1B) is regulated by its N-terminal, small, well conserved region, termed here the Wee-box. The Wee-box is essential for the normal kinase activity of XeWee1B during interphase, acting positively on the C-terminal catalytic domain, which alone cannot efficiently phosphorylate Cdc2. Significantly, a Thr-186-Pro (TP) motif within the Wee-box is phosphorylated by Cdc2 at M phase and specifically binds the cis/trans prolyl isomerase Pin1. This Pin1 binding is required for the inactivation of XeWee1B at M phase, presumably causing isomerization of the phospho-TP motif and thereby impairing the function of the Wee-box. These results provide important insights into the mechanism of Wee1 inactivation at M phase.

In eukaryotic cells, Wee1 phosphorylates Cdc2 on Tyr-15 and inhibits its kinase activity, thereby preventing entry into mitosis (1, 2). At the mitotic entry and during M phase, Wee1 needs to be down-regulated to allow activation of Cdc2 by the Wee1-antagonizing Cdc25 phosphatase. Indeed, Wee1 seems to be destabilized at the late G2 and M phases during the normal cell cycle in diverse organisms (3–6). Specifically, human somatic Wee1 undergoes a phosphorylation- and SCFβ-TrCP-dependent degradation at M phase, offering one explanation for the mechanism of Wee1 down-regulation at M phase (7). However, degradation of human somatic Wee1 also occurs in interphase and is not complete even at normal M phase (4); other Wee1 homologs, such as Swe1 in budding yeast, are also not completely degraded at M phase (6, 8). Furthermore, and notably, Xenopus embryonic Wee1 is very stable at M phase in mature oocytes (9, 10), and mouse embryonic (but not somatic) Wee1 is also stable at M phase when expressed in Xenopus oocytes (11). Thus, there must be an additional mechanism(s) by which Wee1 kinases are down-regulated at M phase.

Indeed, it has long been known that the catalytic activity of Wee1 undergoes a phosphorylation-dependent inhibition at M phase. In fission yeast, Nim1/Cdr1 phosphorylates Wee1 in its C-terminal catalytic domain (CD) and inhibits its activity (12, 13). Also, in both Xenopus eggs and human somatic cells, Wee1 homologs are hyperphosphorylated in their N-terminal regulatory domains (NRDs) and inactivated at M phase (3, 4, 9, 14). Specifically, it was reported that Xenopus embryonic Wee1 can be phosphorylated and inhibited by Cdc2 in vitro, suggesting that Wee1 inactivation at M phase may be caused, at least in part, by a positive feedback loop (9). However, similar phosphorylation of human somatic Wee1 by Cdc2 in vitro cannot directly inhibit its kinase activity, questioning the feedback loop between Cdc2 and Wee1 (4). Besides Cdc2, other unknown kinases have been implicated in Wee1 inactivation at M phase (9, 15). Furthermore, a recent study shows that Akt/PKB phosphorylates human somatic Wee1 at G2/M phase and inhibits its kinase activity (16). Thus, multiple kinases, including Cdc2, seem to be involved in the inactivation of Wee1 homologs at M phase. So far, however, the general mechanism (if any) of Wee1 inactivation at M phase has not been known.

In this study we have explored the mechanism of Wee1 inactivation at M phase by using Xenopus egg and extract systems. We show that Xenopus somatic Wee1 (XeWee1B) has an N-terminally located, small key box that is essential for normal Wee1 kinase activity and that Cdc2 phosphorylation and prolyl isomerase Pin1 binding of this key box plays an important role for inactivation of the kinase at M phase. This mechanism may occur fairly generally, because the key box is evolutionarily well conserved in other Wee1 homologs and is functional in Xenopus embryonic Wee1 as well.

Results

Identification of a Key Box, Wee-Box, That Regulates Wee1 Activity.

Wee1 protein consists of a large, highly divergent NRD (≈200–550 residues) and a well conserved C-terminal CD (≈350 residues) (10, 17) (Fig. 1 A). To identify a region(s) in the NRD that would regulate Wee1 kinase activity, first we ectopically expressed a series of N-terminal truncation mutants of XeWee1B (10) at comparable levels in Xenopus oocytes and compared their activities to inhibit progesterone-induced germinal vesicle breakdown [(GVBD), a hallmark for entry or progression into meiotic M phase] (10). These analyses revealed that at least one region (residues 160–200) of the NRD was important for XeWee1B to inhibit GVBD or M phase progression (Fig. 1 B). Interestingly, this region contains a short sequence (termed here the Wee-box), 180(Val)-Asn-Ile-Asn-Pro-Phe-Thr-Pro-(Asp)188, that is exceptionally well conserved in the highly divergent NRDs of various Wee1 homologs (Fig. 1 C) (14). In this study we tentatively divided the Wee-box into two motifs, Asn-Ile-Asn (NIN) and Thr-Pro (TP), and introduced alanine mutations into these motifs (NIN→AAA or 3A and TP→AP). Compared with WT XeWee1B, the 3A mutant inhibited GVBD very weakly, whereas the AP mutant did so very strongly (Fig. 1 D), implying that the NIN and TP motifs acted for Wee1 activity (to inhibit M phase progression) positively and negatively, respectively. To test whether these two motifs had mutually opposite effects on XeWee1B kinase activity itself, we performed in vitro kinase assays of recombinant XeWee1B proteins (WT or mutants, isolated from insect Sf9 cells) using kinase-dead cyclin B/Cdc2 as substrate. The 3A mutant phosphorylated Cdc2 on Tyr-15 significantly (≈5-fold) less efficiently than WT XeWee1B, whereas the AP mutant did so to the same extent as WT XeWee1B (Fig. 1 E). Essentially similar results were obtained by using XeWee1B proteins that were expressed in and isolated from immature oocytes (data not shown). Thus, these results indicate that, whereas the TP motif does not act negatively for XeWee1B kinase activity in vitro (despite its negative action in vivo), the NIN motif does act positively for it (consistent with its action in vivo). These results would imply that the Wee-box as a whole acts positively for normal XeWee1B kinase activity mainly because of its NIN motif.

Fig. 1.
Fig. 1.

Identification of a small region that regulates XeWee1B kinase activity. (A) A schematic representation of XeWee1B protein. For the Wee-box, see C. (B) Thirty immature oocytes were left uninjected (None) or injected with 100–200 pg (depending on the sizes) of mRNA encoding the indicated N-terminal truncation mutants of XeWee1B (each tagged with a Myc epitope), cultured for 12 h, treated with progesterone, cultured further for 6–8 h (depending on the batches of oocytes), and scored for the percentage of GVBD inhibition. Expression levels of the various mutants just before progesterone treatment were determined by immunoblotting with anti-Myc antibody (Lower). All values are means ± SD of four independent experiments. (C) Conservation of the Wee-box (and the NIN and TP motifs therein) in Wee1 kinases from various organisms. (D) GVBD-inhibiting activity of the indicated XeWee1B mutants was analyzed as in B, except that the amount of mRNA injected was 50 pg per oocyte. (E) Recombinant XeWee1B protein (WT, kinase-dead K277R, or the indicated mutants) isolated from Sf9 cells was incubated with recombinant kinase-dead cyclin B/Cdc2 complexes for the indicated times (see Materials and Methods for details), and Tyr-15 phosphorylation of Cdc2 was analyzed by immnoblotting (IB) with anti-phospho-Tyr-15 antibody (pY15) and quantitated by using the Image Gauge (FujiFilm, Tokyo, Japan).

A Role for the Wee-Box in Normal XeWee1B Kinase Activity.

A recent crystal structure analysis suggests that the CD of human somatic Wee1 has an active conformation (18). Consistent with this, the NIN→3A XeWee1B mutant isolated from Sf9 cells (and from oocytes; data not shown) had essentially the same (auto)phosphorylation activity as WT XeWee1B (Fig. 2 A), although it could not efficiently phosphorylate Cdc2 on Tyr-15 (Fig. 1 E). Very interestingly, however, the CD alone (of XeWee1B) also had a very weak Cdc2 Tyr-15-phosphorylating activity in vitro, similar to the 3A mutant (Fig. 2 B). [Note also that the Δ240 XeWee1B mutant, which is equivalent to the XeWee1B CD, had a very weak activity to inhibit GVBD in oocytes (Fig. 1 B).] These results could imply that the NIN motif or the Wee-box is somehow required for the (constitutively active) CD of XeWee1B to properly phosphorylate Cdc2. If so, the NIN motif might function for XeWee1B to physically interact with cyclin B/Cdc2. To test this possibility, we expressed a GST-fused 3A mutant or its WT version in artificially activated eggs (for the advantages of using this egg system, see Materials and Methods). In this experiment, both the WT and the 3A mutant were, in fact, kinase-dead Lys-277→Arg forms, not to affect endogenous Cdc2 activity. GST pulldown assays, however, showed that the 3A mutant could bind to endogenous cyclin B/Cdc2 to the same extent as WT XeWee1B (Fig. 2 C). Thus, the NIN motif was not likely to be involved in the physical interaction of XeWee1B with cyclin B/Cdc2.

Fig. 2.
Fig. 2.

Analysis of the role for the NIN motif in XeWee1B kinase activity. (A) Autophosphorylation (32P) of the indicated XeWee1B constructs (isolated from Sf9 cells) was analyzed by using [γ-32P]ATP. See Materials and Methods for details. KR, kinase-dead. (B) In vitro kinase assays of the indicated XeWee1B constructs were performed as in Fig. 1 E. (C) Activated eggs were injected with 5 ng of mRNA encoding GST alone (−) or GST-tagged XeWee1B constructs (WT or 3A, both being kinase-dead K277R forms) and then incubated for 2 h. After GST pulldown (GST-PD), endogenous Cdc2 and cyclin B2 were analyzed by immunoblotting. (D) Recombinant His6-NRD protein (WT or 3A) was incubated with recombinant GST-CD protein (see Materials and Methods) and subjected to GST pulldown for immunoblotting with anti-His6 antibody. In both C and D, the pulldown and input were on the same gel. (E) Activated eggs were injected with 10 ng of mRNA encoding GST-tagged NRD constructs (WT or 3A) and/or with 5 ng of mRNA encoding Myc-tagged full-length XeWee1B (FL) or CD, incubated for the indicated times, and analyzed by immunoblotting with both anti-GST and anti-Myc antibodies or with anti-Cdc2 phospho-Tyr-15 antibody. (F) Recombinant full-length XeWee1B protein or CD protein (3 nM) was incubated with or without 10 nM NRD protein (WT or 3A) and with 10 nM cyclin B2/Cdc2(KR) complexes for the indicated times, and Cdc2 Tyr-15 phosphorylation was analyzed by immunoblotting. All (tagged) proteins were from Sf9 cells. (G) The indicated recombinant proteins at the indicated concentrations were incubated in different combinations with 10 nM cyclin B2/Cdc2(KR) for the indicated times, and Cdc2 Tyr-15 phosphorylation was analyzed.

We then suspected that the NIN motif might function in an intramolecular fashion so that the CD of XeWee1B can properly phosphorylate Cdc2 on Tyr-15. If this is the case, the NRD (which contains the NIN motif) could interact with the CD and the NIN motif in it might act positively on the CD. When incubated together, recombinant NRD protein (whether WT or 3A) could appreciably and specifically bind to recombinant CD protein in vitro (Fig. 2 D). Furthermore and intriguingly, when coexpressed with the WT but not 3A NRD fragments in activated eggs, the CD fragment could efficiently induce Tyr-15 phosphorylation of endogenous Cdc2 (if not comparably to singly expressed full-length WT XeWee1B) (Fig. 2 E), indicating that the NRD could stimulate CD's kinase activity toward Cdc2 in an NIN motif-dependent manner. Importantly, essentially similar results were obtained in vitro, or simply by incubating their recombinant proteins (including kinase-dead Cdc2 protein as substrate) in the presence of ATP (Fig. 2 F), suggesting that the NRD fragment directly affected CD's kinase activity in an NIN motif-dependent manner. In addition, in these in vitro experiments, the presence of an excess of either the 3A NRD or the kinase-dead CD fragments could inhibit the activity of the CD (plus the WT NRD), but not of full-length XeWee1B (Fig. 2 G), suggesting a preformed intramolecular NRD-CD interaction in XeWee1B protein. Together, these results strongly suggest that the NRD interacts directly with the CD, and the NIN motif (or the Wee-box) in it positively acts on the CD so that this domain can properly phosphorylate Cdc2 Tyr-15.

Mitotic Phosphorylation of the TP Motif by Cdc2.

The TP→AP mutant, another Wee-box mutant of XeWee1B, had essentially the same kinase activity as WT XeWee1B in vitro (Fig. 1 E), but inhibited M phase progression significantly more strongly than WT XeWee1B (Fig. 1 D). These results could imply that, on entry into and during M phase, XeWee1B undergoes some inhibitory feedback regulation, perhaps by phosphorylation at the Thr-186-Pro motif, thereby allowing normal progression through M phase. As expected, when expressed in M phase egg extracts and analyzed by immunoblotting, WT XeWee1B, but not the AP mutant, could be recognized by anti-phospho-Thr-186 antibody in a manner sensitive to phosphatase treatment [see supporting information (SI) Fig. 5]. Furthermore, and importantly, when expressed in cycling egg extracts, WT XeWee1B (kinase-dead in this case) was recognized by the anti-pThr186 antibody only at the time when cyclin B was accumulated and Cdc2 Tyr-15 was dephosphorylated (Fig. 3 A), indicating Thr-186 phosphorylation of XeWee1B at M phase. As Cdc2 is a proline-directed M phase kinase, this kinase might phosphorylate Thr-186 at M phase. Consistently, a noninhibitable (T14A/Y15F; AF), but not kinase-dead (K33R), form of Cdc2 could efficiently phosphorylate full-length WT XeWee1B on Thr-186 in vitro (Fig. 3 B). Furthermore, upon treatment of M phase egg extracts (containing kinase-dead XeWee1B) with the Cdc2 inhibitor roscovitine, the XeWee1B protein was rapidly (or within 15 min) dephosphorylated on Thr-186 (while Cdc2 was phosphorylated on Tyr-15 probably by endogenous Wee1) (Fig. 3 C). In contrast to these, ERK-MAPK, another proline-directed kinase, could not phosphorylate XeWee1B on Thr-186 either in vitro or in vivo (data not shown). Thus, these results strongly suggest that Thr-186 of XeWee1B undergoes a Cdc2-dependent phosphorylation at M phase. Although we obtained these results by ectopically expressing XeWee1B in egg extracts, we confirmed that endogenous XeWee1B can also be mitotically phosphorylated on Thr-186 in somatic cells (see SI Fig. 6).

Fig. 3.
Fig. 3.

Thr-186 phosphorylation by Cdc2 and its requirement for XeWee1B inactivation at M phase. (A) Cycling extracts were added with mRNA encoding Myc-tagged (kinase-dead) XeWee1B, incubated for 50 min, and then sampled at 10-min intervals for immunoblotting with the indicated antibodies. Interphase (I) and M phase (M) were determined by microscopic observation of the morphology of added sperm nuclei. (B) Recombinant full-length XeWee1B protein was incubated with recombinant cyclin B/Cdc2 (KR or AF) and analyzed by immunoblotting with anti-pThr186 or anti-XeWee1B antibodies. For detailed methods, see Materials and Methods. (C) Cycling extracts were added both with mRNA encoding Myc-tagged (kinase-dead) XeWee1B and mRNA encoding nondegradable cyclin B2, incubated for 2 h (or until M phase arrest), treated with 0.5 mM roscovitine, and sampled at the indicated times for immunoblotting with the indicated antibodies. (D) Activated eggs were injected with 5 ng of mRNA encoding GST-XeWee1B (KR, WT, or AP), together with (for M phase) or without (for interphase, I) 5 ng each of mRNAs encoding nondegradable cyclin B2 and Cdc2(AF). Two hours later, GST-XeWee1B proteins were isolated by GST pulldown and incubated with kinase-dead cyclin B/Cdc2 (as substrate) for the indicated times; Cdc2 was then analyzed for Tyr-15 phosphorylation. (E) Cycling extracts (50 μl) were first added with 1 μg of mRNA encoding nondegradable cyclin B2 and incubated for 1.5 h (to cause M phase arrest); the extracts were then added with 1 μg of mRNA encoding XeWee1B (KR, WT, or AP), incubated for the indicated times, and analyzed for Tyr-15 phosphorylation of endogenous Cdc2. (F) Recombinant XeWee1B protein was first incubated with recombinant cyclin B/Cdc2(AF) and/or (activated) Plk1 protein, isolated, and further incubated with recombinant cyclin B/Cdk2(KR), which was then analyzed for Tyr-15 phosphorylation. For details, see Materials and Methods.

Dependence of Mitotic Inactivation of XeWee1B on Thr-186 Phosphorylation.

We next asked whether XeWee1B could be inactivated at M phase and, if so, whether this inactivation depended on Thr-186 phosphorylation. For this, we expressed WT or AP XeWee1B proteins (tagged with GST) in activated eggs, together with or without a nondegradable cyclin B mutant and Cdc2(AF) to make either M phase or interphase. We then isolated the (GST-)XeWee1B proteins by GST pulldown and subjected them to in vitro kinase assays using kinase-dead cyclin B/Cdc2 as substrate. In these experiments, both the WT and AP XeWee1B proteins from M phase eggs were somewhat (≈30%) less abundant than those from interphase eggs (Fig. 3 D Upper) presumably because of their M phase-specific instability (4, 10). Under these conditions, however, the kinase activity of WT XeWee1B from M phase eggs was <10% of that from interphase eggs, whereas the activity of the AP mutant from M phase eggs was >60% of that from interphase eggs (Fig. 3 D Lower). [When their kinase activities were assayed by using the same amount, the AP mutant had ≈6-fold more activity than the WT at M phase (data not shown).] Furthermore, when expressed at appropriate and comparable levels in M phase egg extracts [containing nondegradable cyclin B but not Cdc2(AF), and hence being arrested at anaphase], the AP mutant, but not WT XeWee1B, was able to induce Tyr-15 phosphorylation of endogenous Cdc2 (Fig. 3 E). Thus, these results strongly suggest that XeWee1B can be inactivated at M phase in a Thr-186 phosphorylation-dependent manner.

We then asked whether Thr-186 phosphorylation could directly inhibit XeWee1B kinase activity. For this, we first incubated recombinant WT XeWee1B protein with bead-bound cyclin B/Cdc2(AF) (which fully phosphorylated the XeWee1B protein, as judged by its mobility upshift in Fig. 3 B and F), isolated it, and then further incubated it with kinase-dead cyclin B/Cdc2 as substrate. These analyses showed that phosphorylation of Thr-186 (and also of other sites) by Cdc2 in vitro could not appreciably reduce the kinase activity of XeWee1B (Fig. 3 F). Addition of Plk1, another mitotic kinase of Wee1 (19), could not affect XeWee1B activity either (Fig. 3 F). Thus, under the present conditions, Cdc2 phosphorylation of Thr-186 could not directly inhibit XeWee1B kinase activity, although it was evidently required for XeWee1B inactivation in vivo.

Binding of Pin1 to the pThr186-Pro Motif.

The cis/trans peptidyl-prolyl isomerase Pin1 binds to phosphorylated Ser/Thr-Pro motifs of target proteins and alters the properties of the proteins by changing the proteins' conformation via prolyl isomerization at the pSer/Thr-Pro bonds (20, 21). Interestingly, Pin1 can bind to mitotically phosphorylated (Xenopus embryonic) Wee1, but neither its binding site(s) nor the role for this binding is known (22). We therefore tested whether Pin1 could bind to the pThr186-Pro motif of XeWee1B at M phase. When coexpressed with GST-Pin1 in cycling egg extracts and analyzed by GST pulldown assays, (kinase-dead) XeWee1B was shown to be associated with Pin1 at M phase, but not at interphase (Fig. 4 A). Notably, however, (kinase-dead) AP mutant was not appreciably associated with Pin1 at M phase (or at interphase; data not shown) (Fig. 4 B), suggesting that Thr-186 phosphorylation was required for XeWee1B to bind Pin1 at M phase. To test whether the pThr186-Pro motif could indeed bind Pin1, we incubated pThr186-Pro-containing peptides (pT186-pep; residues 179–191) or unphosphorylated peptides (T186-pep), each coupled with agarose beads, with either interphase or M phase egg extracts, and then performed pulldown/immunoblotting assays for endogenous Pin1. These analyses revealed that pT186-pep, but not T186-pep, was able to bind Pin1 not only in M phase but also in interphase extracts (Fig. 4 C). Thus, these results suggest that Pin1 binds to the pThr186-Pro motif of XeWee1B at M phase and this motif is the major (perhaps unique) Pin1 binding site in XeWee1B.

Fig. 4.
Fig. 4.

Pin1 binding to pThr186 and its requirement for XeWee1B inactivation at M phase. (A) Cycling extracts were added both with mRNA encoding GST alone or GST-Pin1 and mRNA encoding Myc-tagged (kinase-dead) XeWee1B, incubated for 1 h (for interphase or I, as determined by the morphology of added sperm nuclei) or for 1.5 h (for M phase), subjected to GST pulldown, and immunoblotted with anti-Myc or anti-GST antibodies. (B) Binding of Pin1 to the AP mutant at M phase was tested in the same way as in A. (C) Either pThr186-containing peptides (pT186-pep) or unphosphorylated peptides (T186-pep), each coupled to agarose beads, were incubated with either interphase or M phase extracts (prepared as in A but without added mRNA), pulled down (PD), and immunoblotted for endogenous Pin1. For details, see Materials and Methods. (D) Thirty immature oocytes were injected with 50 pg of mRNA encoding Myc-tagged XeWee1B constructs (WT or AP) and, 3 h later, with or without 20 ng of mRNA encoding Myc-tagged Pin1 constructs (W34A/C109A or C109A), cultured for 12 h, treated with progesterone, and, 6–8 h later (depending on the batches of oocytes), scored for the percentage of GVBD inhibition. The values are means ± SD of five independent experiments. Expression levels of XeWee1B and (both endogenous and exogenous) Pin1 constructs just before progesterone treatment were also determined; the bands marked with asterisks are degradation products of Myc3-Pin1.

Requirement for Pin1 Binding in XeWee1 Inactivation at M Phase.

Given these results, Pin1 binding at the pThr186-Pro motif might be responsible, at least in part, for XeWee1B inactivation at M phase. This idea would be very attractive because the Thr-186-Pro motif is located within the Wee-box or just adjacent to the essential NIN motif. Pin1 consists of an N-terminal substrate-binding domain and a C-terminal catalytic domain (both of which recognize pSer/Thr-Pro motifs) (20). We therefore generated a dominant-negative Pin1 mutant (C109A) (which can bind to substrates but is catalytically inactive) (23) as well as a control Pin1 mutant (W34A/C109A) [which cannot bind to substrates and is inactive). We ectopically expressed either of these Pin1 mutants (≈20-fold over endogenous levels; see Fig. 4 D Bottom) together with either WT XeWee1B or the AP mutant in immature oocytes [which lack endogenous Wee1 protein (10)], and then compared the activities of the WT and the AP mutant to inhibit GVBD (or progression through M phase), as described earlier (Fig. 1 D). In the presence of control Pin1(W34A/C109A), WT XeWee1B inhibited GVBD significantly more weakly than the AP mutant, essentially as in the absence of ectopic Pin1 (Fig. 4 D Top; see also Fig. 1 D). Importantly, in the presence of dominant-negative Pin1(C109A), however, WT XeWee1B reproducibly inhibited GVBD strongly, or comparably to the AP mutant (Fig. 4 D Top), suggesting that WT XeWee1B retained a high M phase-inhibiting activity without endogenous Pin1 activity. In these experiments, WT XeWee1B and the AP mutant were expressed at comparable levels (see Fig. 4 D Middle), and, in contrast to WT XeWee1B, the (high) activity of the AP mutant was not significantly affected by the presence or absence of Pin1(C109A). Furthermore, in the absence of ectopic XeWee1B, GVBD inhibition was not appreciably affected by the presence or absence of Pin1(C109A), the inhibition degree remaining very low. These results do suggest that, in the present experimental systems, Pin1 specifically targeted pThr186-phosphorylated WT XeWee1B (rather than endogenous substrates) to regulate GVBD, and that in the absence of Pin1, WT XeWee1B was not inactivated, retaining a high M phase-inhibiting activity similar to the AP mutant. Furthermore, given the molecular property of the dominant-negative Pin1(C109A) mutant, these results suggest that not only Pin1 binding but also its activity was required to regulate XeWee1B activity. Taken together, these results strongly support the idea that Pin1 binding at the pThr186-Pro motif is required for XeWee1B inactivation at M phase.

Discussion

Although phosphorylation-dependent catalytic inactivation of mitotic Wee1 was shown many years ago, surprisingly little is known about its molecular mechanisms (1, 24). Here we find that a small, well conserved region (termed here the Wee-box) in the NRD of XeWee1B is required for both the normal kinase activity and mitotic inactivation of the kinase.

Despite its location (in primary structure) far from the CD, the Wee-box or the Asn-Ile-Asn (NIN) motif in it is essential for the normal kinase activity of XeWee1B toward Cdc2 Tyr-15. The CD alone also cannot efficiently phosphorylate Cdc2 on Tyr-15 either in vivo or in vitro. Although XeWee1B can physically interact with cyclin B/Cdc2, the NIN motif is not required for this interaction. Importantly, however, the NRD fragment (whether it contains the WT or the mutated NIN motif) can bind to the CD fragment in vitro, and it can stimulate the CD fragment to phosphorylate Cdc2 Tyr-15 in an NIN motif-dependent manner both in vivo and in vitro. These results, together with other results, strongly suggest that, under normal conditions, the NRD of XeWee1B interacts intramolecularly with the CD, and the NIN motif (or the Wee-box) in it positively acts on the CD so that this domain can properly phosphorylate Cdc2 on Tyr-15. These results would also imply that, although the activation segment of the CD may have an active conformation (18), the whole CD itself has a conformation that cannot efficiently phosphorylate Cdc2 Tyr-15. Thus, it is conceivable that the NIN motif or the Wee-box might act to complement or affect the conformation of the CD for this domain to properly recognize Cdc2 Tyr-15. The CD of Wee1 has an extra C-terminal region (17), and this region also plays an important role in normal XeWee1B kinase activity (10). It is possible, therefore, that the Wee-box functions cooperatively with the C-terminal region to promote CD's recognition of Cdc2 Tyr-15. Future studies are needed to elucidate exactly how the Wee-box acts to regulate XeWee1B kinase activity.

The Thr-186-Pro (TP) motif, located within the Wee-box and immediately downstream of the NIN motif, negatively acts for XeWee1B kinase activity during progression through M phase. Thr-186 is phosphorylated by Cdc2 at M phase, and this phosphorylation is required for the inactivation of XeWee1B at M phase, consistent with a feedback regulation of mitotic Wee1 by Cdc2 (9, 25). However, Thr-186 phosphorylation alone is not sufficient for the inactivation of XeWee1B. Instead, the cis/trans prolyl isomerase Pin1 binds specifically to the pThr186-Pro motif at M phase, and this binding and activity of Pin1 is required for XeWee1B inactivation at M phase. Given these results, Pin1 could isomerize XeWee1B at the pThr186-Pro motif, thereby inhibiting its kinase activity. This possibility is plausible because Pin1 isomerization of the pThr186-Pro motif would surely impair the function of the essential Wee-box (in which the motif is located). However, XeWee1B could also be inactivated by phosphorylation at some other Ser/Thr-Pro motif(s) (whether dependent or not on Pin1 activity), because our preliminary results show that Thr-138-Pro, like its equivalent site (Thr-104-Pro) in Xenopus embryonic Wee1 (14), is also required for XeWee1B inactivation at M phase. Thus, although we showed here the important mechanism for XeWee1B inactivation at M phase, further work is needed to fully understand the phosphorylation-dependent inactivation of XeWee1B.

Human somatic Wee1 is degraded, albeit not completely, at G2/M phase (4, 7), offering one explanation for the mechanism of Wee1 inactivation at M phase. Furthermore, a recent study in budding yeast shows that Swe1 is inhibited by dissociation from Cdc2 at M phase (26). Notably, however, Xenopus embryonic Wee1 (XeWee1A) is very stable at M phase and is stably complexed with Cdc2 (9, 10); therefore, this kinase must be inactivated at M phase by another mechanism. Interestingly, Kim et al. (14) recently showed a Wee-box (or its pThr150-Pro)-mediated catalytic inactivation of XeWee1A during mitotic entry in egg extracts, although they did not address the precise mechanism for the inactivation as we did for XeWee1B. Also, our additional data show that inactivation of XeWee1A by Thr-150 phosphorylation is essential for the meiotic cell cycle in oocytes (see SI Fig. 7). Thus, given these results as well as the strong conservation of the Wee-box in other Wee1 homologs (Fig. 1 C), the Wee-box/Pin1-mediated inactivation may be a fairly general mechanism for Wee1 inactivation at M phase.

Materials and Methods

Oocytes and Eggs.

Oocytes and eggs were prepared, cultured, and microinjected as described (10). Unfertilized eggs were treated with the calcium ionophore A23187 (1 μg/ml) to artificially induce egg activation, which mimics fertilization. The artificially activated eggs (which do not cleave) have very low levels of Cdc2 Tyr-15 phosphorylation after 1.5 h of activation, as do normally fertilized embryos, and are a very good system (for their single-cell nature) to express exogenous proteins (by mRNA injection) and, thereby, to analyze protein–protein interactions as well as regulation of Cdc2 Tyr-15 phosphorylation (27, 28).

Cell-Free Egg Extracts.

Cycling egg extracts were prepared as described (29); routinely, the time of M phase was confirmed by the morphology of added sperm nuclei. In some experiments, M phase extracts were prepared by expression of nondegradable cyclin B2 together with or without noninhibitable Cdc2(AF).

cDNAs and Baculoviral Recombinant Proteins.

See SI Materials and Methods .

Antibodies and Immunoblotting.

Anti-XeWee1B phopho-Thr-186 antibody was raised in rabbits against the peptides (CQVNINPFpTPDSLE) and affinity-purified by standard methods. See SI Materials and Methods for other antibodies. Routinely, proteins equivalent to one oocyte or egg were analyzed by immunoblotting, as described (10).

In Vitro Kinase Assays.

For in vitro XeWee1B kinase assays, 10 ng of His6-XeWee1B proteins isolated from Sf9 cells was incubated with 20 ng of GST-cyclin B/His6-Cdc2(KR) complexes (isolated from Sf9 cells and bound to glutathione beads) in 50 μl of a buffer [20 mM Tris·HCl (pH 7.5), 10 mM MgCl2, 50 mM KCl, 1 mM DTT, 1 mM Na3VO4, and 1 mM NaF] supplemented with 0.3 mM ATP for 5–20 min at 23°C, and the cyclin B/Cdc2 complexes were pulled down and analyzed by immunoblotting with anti-Cdc2 phospho-Tyr-15 antibody. In vitro kinase assays of His6-XeWee1B CD protein in the presence of other proteins were performed similarly, but by using specified amounts of the proteins. For autophosphorylation activity of XeWee1B, His6-XeWee1B protein was incubated in the above-described buffer in the presence of 10 μM ATP and 5 μCi of [γ-32P]ATP for 20 min at 23°C and then analyzed by SDS/PAGE followed by autoradiography. For XeWee1B phosphorylation by Cdc2 as well as sequential (Cdc2→XeWee1B→Cdc2) kinase assays, see SI Materials and Methods .

NRD-CD Binding Assays.

Recombinant His6-NRD protein (1 μM) was incubated with recombinant GST-CD protein (1 μM) in a buffer [50 mM Hepes·KOH (pH 7.7), 1 mM EDTA, 50 mM NaCl, and 0.2% Tween 20] for 20 min at 23°C and then subjected to GST pulldown for immunoblotting with anti-His6 antibody.

GST Pulldown Assays.

GST-fusion proteins were pulled down from egg extracts by using glutathione-Sepharose beads (GE Healthcare, Uppsala, Sweden), essentially as described (27) (see SI Materials and Methods for details). Co-pulled-down proteins (routinely from extracts equivalent to 10 eggs) were then analyzed by immunoblotting with appropriate antibodies.

Phospho-Peptide-Pin1 Binding Assays.

See SI Materials and Methods .

Acknowledgments

We thank members of the N.S. laboratory for discussions and K. Gotoh for typing the manuscript. This work was supported by a scientific grant from the CREST Research Project of the Japan Science and Technology Agency (to N.S.).

Footnotes

  • ‡To whom correspondence should be addressed. E-mail: nsagascb{at}mbox.nc.kyushu-u.ac.jp

  • Author contributions: K.O. and N.S. designed research; K.O. performed research; K.O. and N.S. analyzed data; and N.S. wrote the paper.

  • The authors declare no conflict of interest.

  • This article is a PNAS direct submission.

  • This article contains supporting information online at www.pnas.org/cgi/content/full/0607357104/DC1.

  • Abbreviations:
    CD,
    catalytic domain;
    GVBD,
    germinal vesicle breakdown;
    NRD,
    N-terminal regulatory domain;
    XeWee1B,
    Xenopus somatic Wee1.

References

  1. ↵
    1. Coleman TR ,
    2. Dunphy WG
    (1994) Curr Opin Cell Biol 6:877–882.
    OpenUrlCrossRefPubMed
  2. ↵
    1. Morgan DO
    (1995) Nature 374:131–134.
    OpenUrlCrossRefPubMed
  3. ↵
    1. McGowan CH ,
    2. Russell P
    (1995) EMBO J 14:2166–2175.
    OpenUrlPubMed
  4. ↵
    1. Watanabe N ,
    2. Broome M ,
    3. Hunter T
    (1995) EMBO J 14:1878–1891.
    OpenUrlPubMed
  5. ↵
    1. Aligue R ,
    2. Wu L ,
    3. Russell P
    (1997) J Biol Chem 272:13320–13325.
    OpenUrlAbstract/FREE Full Text
  6. ↵
    1. Sia RA ,
    2. Bardes ES ,
    3. Lew DJ
    (1998) EMBO J 17:6678–6688.
    OpenUrlAbstract/FREE Full Text
  7. ↵
    1. Watanabe N ,
    2. Arai H ,
    3. Nishihara Y ,
    4. Taniguchi M ,
    5. Watanabe N ,
    6. Hunter T ,
    7. Osada H
    (2004) Proc Natl Acad Sci USA 101:4419–4424.
    OpenUrlAbstract/FREE Full Text
  8. ↵
    1. Asano S ,
    2. Park JE ,
    3. Sakchaisri K ,
    4. Yu LR ,
    5. Song S ,
    6. Supavilai P ,
    7. Veenstra TD ,
    8. Lee KS
    (2005) EMBO J 24:2194–2204.
    OpenUrlCrossRefPubMed
  9. ↵
    1. Mueller PR ,
    2. Coleman TR ,
    3. Dunphy WG
    (1995) Mol Biol Cell 6:119–134.
    OpenUrlAbstract/FREE Full Text
  10. ↵
    1. Okamoto K ,
    2. Nakajo N ,
    3. Sagata N
    (2002) EMBO J 21:2472–2484.
    OpenUrlCrossRefPubMed
  11. ↵
    1. Han SJ ,
    2. Conti M
    (2006) Cell Cycle 5:227–231.
    OpenUrlCrossRefPubMed
  12. ↵
    1. Parker LL ,
    2. Walter SA ,
    3. Young PG ,
    4. Piwnica-Worms H
    (1993) Nature 363:736–738.
    OpenUrlCrossRefPubMed
  13. ↵
    1. Wu L ,
    2. Russell P
    (1993) Nature 363:738–741.
    OpenUrlCrossRefPubMed
  14. ↵
    1. Kim SY ,
    2. Song EJ ,
    3. Lee KJ ,
    4. Ferrell JE, Jr
    (2005) Mol Cell Biol 25:10580–10590.
    OpenUrlAbstract/FREE Full Text
  15. ↵
    1. Tang Z ,
    2. Coleman TR ,
    3. Dunphy WG
    (1993) EMBO J 12:3427–3436.
    OpenUrlPubMed
  16. ↵
    1. Katayama K ,
    2. Fujita N ,
    3. Tsuruo T
    (2005) Mol Cell Biol 25:5725–5737.
    OpenUrlAbstract/FREE Full Text
  17. ↵
    1. Hanks SK ,
    2. Quinn AM ,
    3. Hunter T
    (1988) Science 241:42–52.
    OpenUrlAbstract/FREE Full Text
  18. ↵
    1. Squire CJ ,
    2. Dickson JM ,
    3. Ivanovic I ,
    4. Baker EN
    (2005) Structure (London) 13:541–550.
    OpenUrl
  19. ↵
    1. Watanabe N ,
    2. Arai H ,
    3. Iwasaki J ,
    4. Shiina M ,
    5. Ogata K ,
    6. Hunter T ,
    7. Osada H
    (2005) Proc Natl Acad Sci USA 102:11663–11668.
    OpenUrlAbstract/FREE Full Text
  20. ↵
    1. Lu KP
    (2004) Trends Biochem Sci 29:200–209.
    OpenUrlCrossRefPubMed
  21. ↵
    1. Wulf G ,
    2. Finn G ,
    3. Suizu F ,
    4. Lu KP
    (2005) Nat Cell Biol 7:435–441.
    OpenUrlCrossRefPubMed
  22. ↵
    1. Shen M ,
    2. Stukenberg PT ,
    3. Kirschner MW ,
    4. Lu KP
    (1998) Genes Dev 12:706–720.
    OpenUrlAbstract/FREE Full Text
  23. ↵
    1. Yeh E ,
    2. Cunningham M ,
    3. Arnold H ,
    4. Chasse D ,
    5. Monteith T ,
    6. Ivaldi G ,
    7. Hahn WC ,
    8. Stukenberg PT ,
    9. Shenolikar S ,
    10. Uchida T ,
    11. et al.
    (2004) Nat Cell Biol 6:308–318.
    OpenUrlCrossRefPubMed
  24. ↵
    1. Kellogg DR
    (2003) J Cell Sci 116:4883–4890.
    OpenUrlAbstract/FREE Full Text
  25. ↵
    1. Pomerening JR ,
    2. Kim SY ,
    3. Ferrell JE, Jr
    (2005) Cell 122:565–578.
    OpenUrlCrossRefPubMed
  26. ↵
    1. Harvey SL ,
    2. Charlet A ,
    3. Haas W ,
    4. Gygi SP ,
    5. Kellogg DR
    (2005) Cell 122:407–420.
    OpenUrlCrossRefPubMed
  27. ↵
    1. Uto K ,
    2. Inoue D ,
    3. Shimuta K ,
    4. Nakajo N ,
    5. Sagata N
    (2004) EMBO J 23:3386–3396.
    OpenUrlCrossRefPubMed
  28. ↵
    1. Inoue D ,
    2. Sagata N
    (2005) EMBO J 24:1057–1067.
    OpenUrlAbstract
  29. ↵
    1. Murray AW
    (1991) Methods Cell Biol 36:581–605.
    OpenUrlCrossRefPubMed
View Abstract
Back to top
Article Alerts
Email Article
Citation Tools
Request Permissions
Mechanism for inactivation of the mitotic inhibitory kinase Wee1 at M phase
Kengo Okamoto, Noriyuki Sagata
Proceedings of the National Academy of Sciences Mar 2007, 104 (10) 3753-3758; DOI: 10.1073/pnas.0607357104
del.icio.us logo Digg logo Reddit logo Twitter logo CiteULike logo Facebook logo Google logo Mendeley logo
Proceedings of the National Academy of Sciences: 116 (7)
Current Issue

Submit

Jump to section

You May Also be Interested in

Nonmonogamous strawberry poison frog (Oophaga pumilio). Image courtesy of Yusan Yang (University of Pittsburgh, Pittsburgh).
Putative signature of monogamy
A study suggests a putative gene-expression hallmark common to monogamous male vertebrates of some species, namely cichlid fishes, dendrobatid frogs, passeroid songbirds, common voles, and deer mice, and identifies 24 candidate genes potentially associated with monogamy.
Image courtesy of Yusan Yang (University of Pittsburgh, Pittsburgh).
Active lifestyles. Image courtesy of Pixabay/MabelAmber.
Meaningful life tied to healthy aging
Physical and social well-being in old age are linked to self-assessments of life worth, and a spectrum of behavioral, economic, health, and social variables may influence whether aging individuals believe they are leading meaningful lives.
Image courtesy of Pixabay/MabelAmber.

More Articles of This Classification

Related Content

  • No related articles found.

Cited by...

Similar Articles