Phosphorylation of Maskin by Aurora-A Participates in the Control of Sequential Protein Synthesis during Xenopus laevis Oocyte Maturation*

At the end of oogenesis, Xenopus laevis stage VI oocytes are arrested at the G2/M transition (prophase) waiting for progesterone to release the block and begin maturation. Progesterone triggers a cascade of phosphorylation events such as a decrease of pKa and an increase of maturating-promoting factor activity. Progression through meiosis was controlled by the sequential synthesis of several proteins. For instance, the MAPK kinase kinase c-Mos is the very first protein to be produced, whereas cyclin B1 appears only after meiosis I. After the meiotic cycles, the oocyte arrests at metaphase of meiosis II with an elevated c-Mos kinase activity (cytostatic factor). By using a two-hybrid screen, we have identified maskin, a protein involved in the control of mRNA sequential translation, as a binding partner of Aurora-A, a protein kinase necessary for oocyte maturation. Here we showed that, in vitro, Aurora-A directly binds to maskin and that both proteins can be co-immunoprecipitated from oocyte extracts, suggesting that they do associate in vivo. We also demonstrated that Aurora-A phosphorylates maskin on a Ser residue conserved in transforming acidic coiled coil proteins from Drosophila to human. When the phosphorylation of this Ser was inhibited in vivo by microinjection of synthetic peptides that mimic the maskin-phosphorylated sequence, we observed a premature maturation. Under these conditions, proteins such as cyclin B1 and Cdc6, which are normally detected only in meiosis II, were massively produced in meiosis I before the occurrence of the nuclear envelope breakdown. This result strongly suggests that phosphorylation of maskin by Aurora-A prevents meiosis II proteins from being produced during meiosis I.

In Xenopus oocytes, progesterone induces meiotic maturation by triggering the conversion of pre-MPF, 1 stockpiled as an inactive form in prophase I oocyte, into MPF at the time of germinal vesicle breakdown (GVBD) (1,2). Stage VI of the first meiotic division is a step of growth and synthesis of mRNA that will be used later for progression of prophase I-arrested oocytes through meiosis I. After GVBD, continued protein synthesis is still necessary to successfully complete meiosis I and to allow the reappearance of the MPF at the onset of meiosis II.
The dormant mRNAs, which are synthesized and stored in the growing oocyte, become translationally activated during the completion of meiosis (oocyte maturation). During meiotic maturation, oocytes are transcriptionally repressed, and all necessary proteins are translated from pre-existing, maternally derived mRNAs (3). The translation of Mos mRNA is necessary to produce the serine/threonine kinase Mos, a key regulator of oocyte maturation. Mos is a direct activator of mitogen-activated protein kinase (MAPK) kinase, which in turn activates MAPK (4). Mos protein levels are tightly regulated in vivo during oocyte maturation (5,6) to induce the MAPK cascade that directly activates the MPF. Mos translational control is exerted through the regulation of Mos mRNA (A) tail in immature oocytes, to which further adenyl residues are added during oocyte maturation (7,8). This early translation of c-Mos mRNA was reported to be essential for the progesterone-stimulated maturation (9). However, it was recently demonstrated that c-Mos synthesis is not necessary for the entry in meiosis I but is required after meiosis I to proceed to meiosis II in order to induce the cytostatic factor arrest (10).
In addition to Mos, a number of other Xenopus maternal mRNAs are translationally regulated by cytoplasmic polyadenylation during progesterone-stimulated oocyte maturation. Furthermore, the control of temporal order of the mRNA translation is crucial for oocyte maturation to proceed normally (11,12). Maternal mRNAs are not translated at the same time but are selectively and sequentially elongated and translated all along oocyte maturation to ensure a timely progression through meiosis I and II and to arrest oocytes at metaphase of meiosis II (13)(14)(15)(16). The mRNAs encoding c-Mos, several cyclins (cyclin B1 and cyclin B4), Cdk, and the enzymes required for DNA replication and chromatin assembly are concerned by this translational control (17,18).
Prior to the onset of meiotic maturation, these mRNAs possess a relatively short poly(A) tail. Progesterone-stimulated translational control may be divided into three processes as follows: (i) progesterone-initiated signal transduction; (ii) signal amplification; and (iii) mRNA cytoplasmic polyadenylation and translation. After progesterone treatment, several of these messages are translationally activated by poly(A) elongation; the poly(A) tail elongates up to about 100 nucleotides, and the mRNAs can thereafter be recruited by the translational machinery (19,20). Dynamic changes in poly(A) tail length of some of these maternal mRNAs play a crucial role in their ribosome recruitment and temporally regulated pattern of translation (7). Cytoplasmic polyadenylation is directed by two types of sequence-specific elements in the 3Ј-untranslated region of these mRNAs: a uracil-rich cytoplasmic polyadenylation element (CPE), and the polyadenylation hexanucleotide sequence AAUAAA (3,7,(21)(22)(23)(24). Cytoplasmic polyadenylation elementbinding protein (CPEB) is a protein that binds CPE at the 3Ј end of the mRNA. CPEB interacts with maskin, a factor that also associates with the mRNA 5Ј cap binding factor eIF4E. The binding of maskin to CPEB and eIF4E induces a loop formation in the mRNA. In Xenopus, the CPEB-maskin-eIF4E complex prevents the translation of mRNAs such as cyclin B1, c-Mos, or Cdc6.
When masked, these mRNAs typically have short poly(A) tails. In immature oocytes, the CPE of mRNA such as cyclin B is bound to the CPEB, a zinc finger-and RRM-containing protein, and the poly(A) tail of the mRNA is typically short. The translation inhibition before maturation results from the interaction of maskin and eIF4E, and this last interaction inhibits the binding of eIF4G to eIF4E that would trigger the initiation of mRNA translation. The association of CPEB with maskin ensures that only CPE-containing mRNAs are repressed.
In Xenopus oocytes, polyadenylation is initiated when progesterone interacts with its receptor leading to the activation of Aurora-A kinase and the phosphorylation of CPEB. Recent studies suggested that this phosphorylation event provokes the recruitment of CPSF, resulting in an addition of a poly(A) tail to mRNA. The phosphorylation of CPEB induces CPEB to bind and recruit CPSF possibly by stabilizing it on the AAUAAA sequence which in turn attracts poly(A) polymerase to the end of the mRNA. The activation of mRNA translation needs the dissociation of the complex eIF4E, maskin, and CPEB (25,26). This conformation loop of the mRNA prevents the recruitment of eIF4G on eIF4E and the subsequent 40 S ribosomal subunit positioning at the mRNA 5Ј end, inhibiting the mRNA translation (27,28).
In a search for Xenopus laevis Aurora-A substrates using a two-hybrid screen, we identified maskin (Xenopus TACC) as an Aurora-A binding partner. We demonstrate that both proteins interact in vitro and in vivo and that Aurora-A phosphorylates maskin in vitro. Maskin is a phosphoprotein in which the phosphorylation is partially dependent on Aurora-A in vivo. We found that injection in Xenopus stage VI oocytes of peptides mimicking the maskin sequence phosphorylated by Aurora-A accelerates the appearance of GVBD (meiosis I) induced by progesterone by allowing early translation of proteins normally expressed only for meiosis II.

EXPERIMENTAL PROCEDURES
Materials-X. laevis adult females (CNRS, Rennes, France) were bred and maintained under laboratories conditions. They were anesthetized under 0.2% phenoxyethanol and surged in order to take half of an ovary to obtain stage VI oocytes.
Antibodies-Monoclonal antibodies against Xl-Aurora-A (1C1 and 6E3) were produced in the laboratory, and polyclonal antibodies against cyclin B1 and cyclin B2 were gifts from T. Hunt (Cancer Research UK, South Mimms, UK). The polyclonal anti-Xl-Aurora-A antibodies were the gift from T. Lorca (CRBM, Montpellier, France); the polyclonal antibodies against maskin were the gift from J. Raff (Cancer Research UK, Cambridge, UK), and the polyclonal antibodies anti-Cdc6 were the gift from M. Mechali (IGH, Montpellier, France). The polyclonal antibodies against the phospho-Ser-626 of the protein maskin was performed by Eurogentec (Searing, Belgium) using the following peptide ESVLRKQS(P)LYKFDP. The GST and ␤-tubulin antibodies were from Sigma (G-7781 and T-4028, respectively), and the anti c-Mos antibody was from Santa Cruz Biotechnology (SC086).
Detection of Aurora-A/Maskin Interaction through the Two-hybrid System-The HpaI/XhoI Xl-Aurora-A fragment cloned in pET21 (29) was subcloned in pGBT11. The cDNA Xenopus egg library was obtained from J. Moreau (CRBM, Paris 7, France). The PCR fragment was digested by BamHI and XhoI and then subcloned in a pGADGH vector in the BamHI/SalI restriction sites. Yeast strain SFY 526 (MATa,112, can r , gal4-54, gal80-538, URA3:GAL1-lacZ) was transformed by means of a pairwise combination of both two-hybrid vectors and grown on a medium without leucine or tryptophan. Galactosidase activities were assayed using 5-bromo-4-chloro-3-indolyl ␤-D-galactopyranoside staining on filter replicates according to Breeden and Nasmyth (30). Both pGADGH and pGBT9 vectors without insert were used as negative controls.
The mutation S626A was performed by PCR amplification in the presence of the following oligonucleotides for GST-p17 and GST-maskin: (Sense)GTGCTCAGAAAGCAGGCTCTCTATCTGA; (Antisense)T-CAGATAGAGAGCCTGCTTTCTGAGCAC. The second PCR performed to amplify the GST-p17 and GST-maskin mutated sequences was realized in the presence of the following oligonucleotides: GST-p17, (Sense)CGCGGATCCGTGGTTTTAAGTTATGCTGACC, and (Antisense)CGGGAATTCGGAGCTCGAAACTCGTCTTC; GST-maskin, (Sense)CGCGGATCCATGAGCCTTCA, and (Antisense)CCGCTCGAG-TCAGATCTTCCC. All constructs were sequenced in full (Genome Express).
Purification of Recombinant Proteins-All recombinant proteins were prepared from Escherichia coli strain BL21(DE3)pLysS. The bacteria induced to produce Aurora-A(His) 6 were lysed in the IMAC 5 buffer (20 mM Tris-HCl, pH 7.5; 500 mM NaCl; 10% glycerin, and 5 mM imidazole) containing 1 mg/ml lysozyme and 1 mM PMSF for 1 h at 4°C (29). The bacteria lysates were then centrifuged for 30 min at 12,000 ϫ g at 4°C (JA-20 rotor, Beckman Instruments), and the supernatants were filtered. The proteins were then purified by Ni-NTA-agarose affinity chromatography following the manufacturer's instructions (Qiagen). The beads were washed twice with 10 volumes of IMAC 5 and were then incubated in 5% BSA in IMAC 5 for 1 h at 4°C. The supernatants were incubated with the beads for 3 h at 4°C, and the beads were washed twice with 10 volumes of IMAC 5 and three times with 10 volumes of IMAC20 (20 mM Tris-HCl, pH 7.5; 500 mM NaCl; 10% glycerol; 20 mM imidazole). The His-tagged proteins were eluted with IMAC 250 (IMAC, 250 mM imidazole). The eluted fractions were controlled in concentration by Bradford analysis and for purity by electrophoresis onto a 12.5% SDS gel (31).
The recombinant GST-p17, GST-p17S626A, GST-maskin, and GSTmaskin S626A were prepared as described previously (32). The bacteria were lysed in PBS (138 mM NaCl; 2.7 mM KCl; 14 mM Na 2 HPO 4; 1.8 mM KH 2 PO 4; pH 7) with 1 mg/ml lysozyme and 1 mM PMSF for 1 h at 4°C. The glutathione-Sepharose 4B beads (Amersham Biosciences) were washed twice with PBS and then incubated in 5% BSA in PBS for 1 h at 4°C. The bacteria lysate was mixed with the beads for 3 h on a wheel at 4°C. The column was washed two times with PBS and three times with washing buffer (WB300: 50 mM Tris-HCl, pH 8; 300 mM NaCl; 0,02% Tween 20; 1 mM PMSF). The proteins were eluted with 10 mM glutathione in 50 mM Tris-HCl, pH 8. The eluted fractions were controlled in concentration by Bradford analysis and for purity by electrophoresis onto a 12.5% SDS-polyacrylamide gel (31).
Affinity Chromatography on a Nickel-NTA-Agarose; Ni-NTA Pulldown Assay-Fifty l of nickel-agarose beads were washed in IMAC 5 and saturated with 10 volumes of IMAC 5 containing 3% BSA for 1 h at 4°C. In parallel, 4 g of purified recombinant GST-p17 were incubated with 4 g of recombinant Aurora-A(His) 6 in 500 l of interaction buffer (IB: 50 mM Tris-HCl, pH 8; 100 mM KCl; 5 mM MgCl 2 ; 0.1% Triton X-100; 20% glycerol) for 3 h at 4°C on a wheel. The mixture was then loaded on the beads and incubated at 4°C for 3 h. The beads were washed twice for 10 min at 4°C in 10 volumes of IMAC 5 and three times in 10 volumes of IMAC 20.
Histidine-tagged proteins and proteins bound to histidine-tagged proteins were eluted with 20 l of IMAC 250. The samples were mixed with 20 l of 2ϫ Laemmli buffer, and the proteins were separated on a 12.5% SDS-polyacrylamide gel and submitted to a Western blot using the 1C1 antibody (diluted 1/200) and GST antibody (diluted 1/50,000).
Affinity Chromatography on a GST-4B-Agarose; GST-Pull-down Assay-50 l of glutathione-Sepharose 4B beads (Amersham Biosciences, catalog number 17-0756-01) were washed in PBS and saturated with PBS containing 3% BSA for 1 h at 4°C. Four g of purified recombinant GST-p17 were incubated with 4 g of recombinant Aurora-A(His) 6 in 500 l of binding buffer (BB: 50 mM Tris-HCl, pH 8; 50 mM NaCl; 0.02% BSA; 0.02% Tween 20; 1 mM PMSF) for 3 h at 4°C. The mixture was then loaded on the beads and incubated at 4°C for 3 h. The beads were washed five times with WB300 for 10 min at 4°C. The proteins were eluted with 50 mM Tris-HCl, pH 8, containing 10 mM glutathione, mixed with 20 l of 2ϫ Laemmli sample buffer, and separated on a 12.5% SDS-polyacrylamide gel. Nitrocellulose membranes were then submitted to Western blot with the 1C1 antibody (diluted 1/200).
In Vivo Co-immunoprecipitation-Ten l of dried Affiprep protein-A (Bio-Rad) were washed with 500 l of immunopure (A) IgG binding buffer (IBB, Pierce) and were incubated for 2 h at 4°C with 500 l of 1C1 antibody (diluted 1/20) in IBB (diluted 1/2), and then washed twice with 500 l of TBS (50 mM Tris-HCl, pH 7.5; 150 mM NaCl). The beads were then incubated with a 10 eq of oocytes extract (MII) for 2 h at 4°C on a wheel. The beads were washed once in 500 l of 0.5% NaCl and five times with 500 l of TBST (TBS, 0.05% Tween 20). Bound proteins were eluted in 10 l of 2ϫ Laemmli sample buffer, and the proteins were separated on a 12.5% SDS-polyacrylamide gel and then immunoblotted. The Western blot analyses were performed with the anti-GST antibody (diluted 1/50,000) and the 1C1 (diluted 1/200).
Protein Kinase Assay-Two g of Aurora-A(His) 6 were incubated in 15 l of the kinase buffer (50 mM Tris-HCl, pH 7.5; 25 mM NaCl; 1 mM dithiothreitol; 10 mM MgCl 2 ) in the presence of 5 Ci of [␥-32 P]ATP at 3,000 Ci/mmol and 10 g of GST-p17 or GST-p17 S626A, GST-maskin or GST-maskin S626A. The reactions were incubated for 10 min at 30°C, stopped by addition of 5 l of 5ϫ Laemmli sample buffer, separated on a 12.5% SDS-PAGE, and analyzed by autoradiography.
Phosphatase Treatment-A two-oocyte X. laevis extract was treated with 400 units of -PPase for 60 min at 30°C (New England Biolabs). The proteins were then separated onto a 12.5% polyacrylamide gel and transferred onto nitrocellulose membrane before Western blotting with the antibody against X. laevis maskin (diluted 1/5000).
In Vivo Analysis-Denuded oocytes (stage VI) were prepared in Merriam buffer (88 mM NaCl; 0.33 mM Ca(NO 3 ) 2 ; 1 mM KCl; 0.41 mM CaCl 2 ; 0.82 mM MgSO 4 ; 10 mM Hepes, pH 7.4) (33). Groups of 100 oocytes were injected with 35 ng of mp1 (35 nl at 1 mg/ml) or with the same volume of dilution buffer (2.5% Me 2 SO in H 2 O). Oocytes were stimulated for maturation with 3 ϫ 10 Ϫ6 M of progesterone at 21°C and GVBD was monitored by the appearance of the white spot. Every hour after stimulation, oocytes were collected by groups of 10, and the extracts were prepared with the extraction buffer (EB: 80 mM ␤-glycerophosphate, pH 7.5; 20 mM EGTA; 15 mM MgCl 2 ; 1 mM dithiothreitol) as described previously (34). Proteases inhibitors were added to the lysate, centrifuged for 30 min at 15,000 ϫ g, and frozen at Ϫ80°C until processed. The extracts were then used to measure the histone H1 kinase activity with 0.25 oocyte eq.

Aurora-A Interacts with Maskin in Vivo-
A search for proteins interacting with Xl-Aurora-A was performed by two-hybrid screen using Aurora-A as bait and a X. laevis oocyte cDNA library. The strongest binding partner found (Fig. 1A) was encoded by a 435-bp cDNA that shared 88% identity with a portion of the CPEB-associated factor maskin cDNA (nucleotides 1545-1980). The isolated cDNA encoded a 145-amino acid peptide corresponding to a part of the original clone, hereafter called p17 (Fig. 1B) (27). The p17 sequence was subcloned into a pGEX-4T-1 vector; the fusion protein GST-p17 was expressed in E. coli and then purified on a glutathione-Sepharose column. The recombinant GST-p17 protein runs as a 56-kDa protein in SDS-PAGE (Fig. 1C, lane 2, upper panel) and was recognized by a polyclonal antibody directed against GST tag (Fig. 1C, lane  1, lower panel). The recombinant GST-p17 protein (Fig. 1C, lane 2) and the endogenous full-length maskin from oocytes extracts (migration at 150 kDa but theoretically 110 kDa when calculated with the molecular weight of the amino acids) were both recognized with an anti-maskin antibody (Fig. 1C, lanes 2  and 3). Comparison of the p17 sequence with those of Xlmaskin and TACC from Drosophila or human revealed a consensus sequence . . . (R)X(S*)(L) . . . in which the serine residue (S*) is a site of phosphorylation for the TACC proteins (Fig.  1D). As the two proteins interact in the two-hybrid system, we asked whether maskin was phosphorylated in oocytes and whether it was an Aurora-A substrate.
Aurora-A and Maskin Interact in Vitro-To investigate whether the protein maskin was a partner of Aurora-A, we performed two pull-down assays using recombinant tagged proteins. In a first series of experiments, Aurora-A(His) 6 was incubated with or without GST-p17 or GST and then the mixture was loaded onto glutathione-Sepharose beads. Aurora-A(His) 6 was found to bind only to the GST-p17-containing beads ( Fig. 2A). In a second round of experiments, the protein GST-p17 previously incubated with or without Aurora-A(His) 6 was mixed with nickel-agarose beads (Ni-NTA-agarose) (Fig.  2B, lanes 1 and 2). GST-p17 was found to bind only the Aurora-A(His) 6 -containing column (Fig. 2B, lane 1). These results indicated that both proteins interacted directly in vitro.
Maskin Is Phosphorylated in Vivo-To test whether maskin was phosphorylated in vivo, protein extracts from oocytes at different stage of maturation (prophase, GVBD an metaphase arrest) were separated on SDS-PAGE, and we analyzed the phosphorylation status of the endogenous protein by Western blot with an antibody anti-maskin. The electrophoretic migration of maskin was compared in the extracts prepared from Xenopus oocytes arrested in prophase I (stage VI) and metaphase II (matured) before and after treatment with -PPase. The upper shift of the protein was not observed in the prophase extract, consequently the treatment with the -PPase did not induce any electrophoretic mobility shift suggesting that, in prophase I oocytes, maskin might not be phosphorylated (Fig.  2C, lanes 1 and 2). In contrast, an upward shift of maskin was observed in GVBD-and metaphase II-arrested compared with prophase I-arrested oocytes (Fig. 2C, compare lanes 1, 3, and 5). However, a modification of the mobility was observed in both GVBD and metaphase extract after treatment with the -PPase (Fig. 2C, lanes 4 and 6). The shift was because of protein phosphorylation demonstrating that maskin is phosphorylated in vivo.
As maskin is phosphorylated in vivo, we asked whether Aurora-A could participate in this phosphorylation. We therefore assessed whether the interaction between Aurora-A and maskin occurred in oocytes. When endogenous Aurora-A was immunoprecipitated from prophase and metaphase II Xenopus oocytes extracts using anti-Aurora-A 1C1 antibody, maskin was found to co-immunoprecipitate (Fig. 2D, lane 1, upper  panel). Maskin was not detected in an immunoprecipitate performed with a monoclonal antibody (6E3) directed against the recombinant protein Aurora-A(His) 6 that does not recognize the endogenous Aurora-A protein (Fig. 2D, lane 2). To determinate whether the interaction evolved during maturation, the presence of maskin in the Aurora-A immunoprecipitate was compared between prophase I, GVBD, and metaphase-arrested oocytes (metaphase II). In both extracts, although the amount of immunoprecipitated Aurora-A increased between prophase I and GVBD oocytes as described previously (35) (Fig. 2D, lanes  3-5, lower panel), the amount of maskin protein bound to Aurora-A did not increase, suggesting that both proteins remained associated throughout oocyte maturation, from prophase I to metaphase II, and that newly synthesized Aurora-A did not bind to maskin (Fig. 2D, lanes 3-5). As observed previously in Fig. 2C, maskin presented an uppershift electrophoretic mobility in the GVBD and metaphase II stage suggesting that Aurora-A might be responsible for this shift.
Aurora-A Phosphorylates Maskin in Vitro-Because we demonstrated a physical interaction between Aurora-A and maskin in vitro, we investigated whether the kinase could phosphorylate maskin. The Aurora-A(His) 6 protein was incubated with the recombinant GST-p17 protein in the presence of [␥-32 P]ATP. Aurora-A did not phosphorylate GST alone as reported previously (Fig. 3A, lane 3) (36). The incorporation of radioactivity in GST-p17 was observed in the presence of Aurora-A(His) 6 , indicating that the kinase phosphorylated p17 in vitro (Fig. 3A, lane 2, upper panel). This experiment shows that the recombinant protein Aurora-A phosphorylates a fragment of maskin in vitro.
Maskin Is an Activating Partner of Aurora-A-It has been reported that binding of Aurora-A to its physiological substrates, such as TPX2 and Ajuba, results in the activation of the kinase (37,38). We thus investigated whether maskin was also an Aurora-A activator. Phosphorylation reactions were performed in the presence of [␥-32 P]ATP with fixed amounts of Aurora-A(His) 6 and GST-H3 tail that contained the Ser (Ser-10) phosphorylated by Aurora-A in vitro (39,40). Increasing amounts of GST-maskin were added to the reaction, and the activity of Aurora-A was measured on GST-H3 tail (Fig. 3B, middle panel). As expected, incorporation of 32 P onto the GSTmaskin followed the increase of GST-maskin concentration in the reaction mix. Most interestingly, the phosphorylation of the GST-H3 tail was found to be more efficient as the amount of GST-maskin increased in the mixture (Fig. 3B, upper panel). The control performed in the presence of GST alone revealed that the GST protein is not able to amplify the kinase activity of the Aurora-A(His) 6 onto the substrate GST-H3 tail (data not shown). GST-maskin induced a stimulation of Aurora-A kinase activity on GST-H3 tail comparable with the stimulation observed in the presence of TPX2 (Fig. 3B, lane 10). A quantification of the radioactivity incorporated in GST-H3 indicated that maskin provoked a 7-fold increase in Aurora-A(His) 6 activity (Fig. 3C). Maskin would then appear to be a third substrate activator of Aurora-A as Ajuba and TPX2.
Aurora-A Phosphorylates Ser-626 in Maskin-The maskin fragment p17 sequence was aligned with the maskin fulllength sequence, human-TACC3 and D-TACC, two proteins previously shown to be phosphorylated by Aurora-A (41). A conserved Ser was found in a consensus sequence for Aurora kinase . . . (R)X(S)(L) . . . in the three proteins as well as in p17 and maskin (Fig. 1D) (42). In maskin, the Ser is located at position 626 (Fig. 1B).
We mutated the amino acid corresponding to the Ser-626 in the GST-p17 and GST-maskin sequence, and we asked whether Aurora-A could phosphorylate this mutant. When GST-p17-S626A protein was incubated in the presence of Aurora-A, the amount radioactivity incorporated was very low compare with the radioactivity incorporated in the wild-type protein (Fig. 4A,  lanes 2 and 3, upper panel). An antibody prepared against   6 and then bound to glutathione-Sepharose 4B beads. The presence of Aurora-A(His) 6 was tested with anti-1C1 antibody (diluted 1/200). B, Aurora-A(His) 6 (lane 1) or interaction buffer (BB) alone (lane 2) were incubated with GST-p17 and then bound to Ni-NTA-agarose beads. The presence of GST-p17 was tested with anti-GST antibody (diluted 1/50,000). C, maskin is phosphorylated in vivo during oocyte maturation. The protein maskin extracted from Xenopus oocytes arrested in prophase I (lanes 1 and 2) or in metaphase II (lanes 3 and 4) was analyzed by SDS-PAGE before (Ϫ) and after -PPase treatment (ϩ) and immunoblotted with an antibody against maskin (diluted 1/5,000). Maskin and Aurora-A interact in vivo. D, Aurora-A was immunoprecipitated from metaphase II oocyte extract with the 1C1 monoclonal antibody (lane 1) or with the 6E3 antibody (lane 2). Aurora-A and maskin were immunodetected with a polyclonal antibody against Aurora-A (diluted 1/200, lower panel) or with a polyclonal antibody against maskin (diluted 1/5,000, upper panel), respectively. Aurora-A was immunoprecipitated from prophase I oocyte extract (stage VI, phosphorylated Ser-626 revealed that the amino acid was fully phosphorylated in the GST-p17 protein, whereas it was not in the GST-p17S626A. We then synthesized two peptides of 18 amino acids mimicking the sequence surrounding maskin Ser-626. The peptide mp1 corresponded to the wild-type sequence with a Ser residue at position 626. The peptide Smp1 was the scramble peptide of mp1 (Fig. 4B). The two peptides were incubated with Aurora-A(His) 6 in the presence of [␥-32 P]ATP. The peptide mp1 was readily phosphorylated by Aurora-A(His) 6 , whereas Smp1 was not phosphorylated as shown by a low level of radioactivity incorporated in a 10-fold amount of substrate (Fig. 4C, histogram). These results indicated that Aurora-A phosphorylated the Ser in the mp1 peptide that mimics the sequence surrounding Ser-626 in the full-length sequence of maskin. There was no phosphorylation onto the Ser in the scramble peptide.
If Ser-626 was phosphorylated by Aurora-A in GST-p17, the addition of an excess amount of peptide in the kinase reaction should compete with the phosphorylation of GST-p17 or GSTmaskin. The inhibitory effect of both mp1 and Smp1 peptides was tested in vitro on the phosphorylation of GST-maskin by Aurora-A(His) 6 . Ten g of the different peptides were added to a phosphorylation reaction containing Aurora-A(His) 6 , GSTmaskin, and [␥-32 P]ATP (Fig. 5A, lane 1). Unlike the scramble peptide (Fig. 5A, lane 3), the mp1 peptide inhibited the incor-poration of the radioactivity into GST-maskin (Fig. 5A, lane 2). The phosphorylated form of GST-maskin was tested for the presence of phosphorylated Ser-626. The antibody detected Ser-626 phosphorylated onto GST-maskin incubated in the presence of Aurora-A(His) 6 alone or together with the Smp1 peptide (Fig. 5A, lanes 1 and 3, lower panel). In contrast, no signal was observed in the presence of the mp1 peptide (Fig.  5A, lane 2, lower panel). The same experiment was carried out with the GST-maskin-S626A (Fig. 5A, lanes 4 and 5). As expected the antibody directed against phospho-Ser-626 did not detect GST-maskin S626A when phosphorylated by Aurora-A; however, maskin mutated on Ser-626 remained heavily phosphorylated by Aurora-A indicating that the kinase phosphorylated several sites on maskin. In contrast, Aurora-A(His)6 could not phosphorylate GST-maskin wild type or GST-maskin-S626A in the presence of mp1, indicating that the peptide inhibited Aurora-A(His) 6 activity.
The Peptide mp1 Inhibits the Phosphorylation of Maskin in Vitro-In order to examine the effect of the mp1 synthetic peptide in the phosphorylation of the maskin full-length protein, we incubated the recombinant protein GST-maskin in a phosphorylation assay in the presence of a metaphase extract that contains Aurora-A active kinase, with or without the synthetic peptides mp1 and Smp1 (Fig. 5B). Upon incubation in the presence of [␥-32 P]ATP and metaphase oocyte extracts, we observed an incorporation of radioactivity in the recombinant GST-maskin (Fig. 5B, lanes 1 and 3, upper panel). The antibody directed against phospho-Ser-626 detected maskin on Western blot indicating that Ser-626 was phosphorylated. Incorporation of radioactivity was also detected when GST-maskin-S626A was incubated in metaphase extracts. It was only in the presence of mp1, by using an antibody directed against the phospho-Ser-626, that a decrease in Ser-626 phosphorylation was observed in wild-type maskin (Fig. 5B, lane 2).
As expected the same antibody did not detect the GSTmaskin-S626A mutant when phosphorylated by metaphase extracts. However, maskin mutated on Ser-626 remained heavily phosphorylated by metaphase extracts suggesting that one or several kinase phosphorylated maskin on residues other than Ser-626 (Fig. 5B, lanes 1 and 5, lower panel). However, a metaphase extract still phosphorylated GST-maskin-S626A in the presence of the mp1 peptide. The same level of radioactivity incorporation was detected, indicating that maskin was indeed phosphorylated on residues other than Ser-626 by kinases different from Aurora-A, considering that mp1 inhibited Aurora-A in vitro.
Our results suggest the following: 1) oocyte extracts phosphorylated maskin onto Ser-626; 2) addition of mp1 peptide specifically reduced the phosphorylation of the Ser-626 by Aurora-A in the extract; and 3) the Ser-626 was phosphorylated by Aurora-A even if the protein was phosphorylated onto other amino acids by other kinases present in oocyte extracts.
Injection of the Peptide mp1 in Oocytes Induces Early Maturation-We injected the peptide mp1 into prophase I-arrested Xenopus oocytes and then induced the oocyte maturation with progesterone. We monitored the oocyte maturation by analyzing the appearance of the white spot. No maturation was observed when the injected oocytes were not treated with progesterone (data not shown). The oocytes injected with the Ser-626 containing peptide mp1 reproducibly underwent GVBD earlier than the oocytes injected with the dilution buffer (Fig. 6A).
To gain a deeper insight into the understanding of mp1 effect on GVBD, we investigated the pre-MPF into MPF conversion by measuring histone H1 kinase activity (Fig. 6, B and C, 1st panels) and cyclin B2 phosphorylation (Fig. 6, B and C, 2nd panels). The MPF was activated 3 h earlier in the presence of mp1 than in the presence of the dilution buffer. The asterisk on the Fig. 6 indicated the GVBD time. Consistent with the premature resumption of meiosis in mp1-injected oocytes, an early synthesis of c-Mos protein was also observed (Fig. 6, B and C,  4th panels). When we measured the appearance of cyclin B1, a protein synthesized for meiosis II (13), we found an early synthesis of the protein in the oocytes injected with the mp1 peptide (Fig. 6, B and C, 3rd panels). To confirm this finding, we analyzed the appearance of Cdc6, a protein that is normally synthesized only after meiosis I. Just like cyclin B1, Cdc6 synthesis was detected much earlier in the mp1-injected oocytes (Fig. 4, B and C, 5th panels), although the appearance of both cyclin B1 and Cdc6 occurred at the same time after GVBD (3 h after GVBD). c-Mos, cyclin B1, and Cdc6 mRNAs contain CPE sequences (28,43) that are presumably used to regulate translation during oocyte maturation (16). Our results indicated that the injected oocytes enter metaphase II stage prematurely. DISCUSSION The present study explores the role of the interaction of Aurora-A with the post-transcriptional regulator factor "maskin" during Xenopus oocyte maturation. By using the twohybrid system analysis and recombinant proteins, we show evidence of a physical interaction between the two proteins in vitro. An in vivo analysis during oocyte maturation revealed that the interaction is necessary to prevent the entry of oocyte into the meiosis II before the completion of the meiosis I.
In Xenopus, oocytes maturation is regulated in part by mRNAs that are synthesized and stored in prophase-arrested oocytes. These mRNAs are not translated at the same time but at specific times along meiotic division. In immature oocytes, dormant CPE-containing mRNAs are bound to CPEB, itself bound to maskin, preventing the translation.
Recent papers have described the cloning in Drosophila and mammals of a centrosomal protein that co purify with microtubules. This protein called TACC (transforming acid coiled coil) protein has been shown to interact with centrosomes and microtubules. Among the three TACC proteins identified in mammals, each of them interacts in a different and specific way (44,45). In Xenopus, a TACC protein called maskin has been identified (27). Maskin is part of a complex that associates with the 3Ј-untranslated region via CPEB in order to regulate the translation of specific mRNAs. As described for human or Drosophila TACC, maskin is localized onto the mitotic apparatus (25). In Xenopus embryos, immunostaining studies clearly showed that maskin and Aurora-A co-localize. The same observation has been made in Xenopus somatic cells, where maskin, as Aurora-A, is localized to the centrosome and along the spindle during the metaphase (25). The picture is not as clear in Xenopus oocytes, although maskin has been detected as moderately concentrated at the animal pole, whereas Aurora-A was specifically localized on the spindle. Recent papers have de- FIG. 6. Effect of the injection of the peptide mp1 on the kinetic of oocyte maturation. A, prophase I oocytes were injected with 35 ng of mp1 or 2.5% Me 2 SO in H 2 O (dilution buffer) and were then stimulated with 3 M progesterone. Scoring the appearance of the white spot monitored GVBD. B and C, oocyte extracts were prepared from 10 injected oocytes every hour following progesterone addition. Stage VI oocytes were injected with either the buffer alone (B) or mp1 (C). The extracts were analyzed for their MPF kinase activity (histone H1 kinase activity) and for oocyte maturation markers behavior by Western blot (cyclin B2, cyclin B1, c-Mos, and Cdc6). The ␤-tubulin revealed by Western blot was used as loading control. scribed a potential role Aurora-A in the separation of chromosomes and the spindle rotation during meiosis in Xenopus oocytes (46). There is no information on these two proteins in Xenopus oocytes, and a role of maskin in the meiotic process has not been documented.
Recently, in Drosophila, a physical interaction between TACC and Aurora-A has been shown to be needed for the recruitment of TACC onto the mitotic apparatus (41). Consistent with these previous observations, we wondered whether maskin and Aurora-A interact in Xenopus oocytes and what is the physiological significance of this possible interaction. In a first round of experiments, we demonstrate for the first time in vitro that Xl-Aurora-A phosphorylates GST-p17, probably onto the amino acid Ser-626. These results lead us to hypothesize that Aurora-A could interact with and phosphorylate maskin in vivo in order to recruit maskin onto the centrosome (25). Immunoprecipitation experiments revealed that Aurora-A and maskin interact in vivo. The interaction is detectable in prophase I and metaphase II oocytes in accord with the accumulation of the two proteins throughout oocyte growth (25,35). We observed that the protein maskin is phosphorylated during meiosis, only after GVBD as shown by the differential action of the -PPase. This phosphorylation has never been reported before but concurs with the activation of Aurora-A at the same period (35). Although it is too premature to affirm that this phosphorylation is owed to Aurora-A, we can hypothesize that Aurora-A is partially responsible of the upper shift observed in our assays.
Such an hypothesis raises the question whether maskin is a physiological substrate for Aurora-A. Other physiological substrates of Aurora-A, such as TPX2 (Targeting protein for Xenopus kinesin like protein 2) and Ajuba, were identified as binding activator partners of Aurora-A (37,38). As supported by the in vitro phosphorylation assay in the presence of the substrate H3, maskin is also an Aurora-A-activating factor. In the absence of GST-maskin, Aurora-A phosphorylated moderately the GST-H3 tail, and this activity increased in a GST-maskin concentration-dependent manner. It remains possible that a priming phosphorylation of Aurora-A may occur in order to initiate a basal activity, then the association with the physiological substrate will mediate a hyper-activation of the kinase (47). As the detection of maskin in the two-hybrid screen was performed with the catalytic domain of Aurora-A, we could hypothesize that, as TPX2, maskin binds to the catalytic domain of Aurora-A. In contrast, Ajuba interacts directly with the N-terminal noncatalytic domain (38,47).
What is the role of the phosphorylation of maskin? The identification of maskin as a substrate of Aurora-A clearly places the kinase upstream of the translation of mRNA coding for proteins that are necessary for oocyte maturation. The mRNAs initiating the process of oocyte maturation are those encoding c-Mos or cyclin B. In Xenopus, c-Mos protein kinase is an efficient activator of the MAPK kinase pathway and is able to induce meiotic maturation when injected in prophase I oocyte (5). Xenopus Mos also induces oocyte maturation in the absence of progesterone, but it is not clear whether its synthesis is sufficient, and required, to initiate the meiotic maturation. The exact role of Mos during oocyte maturation is still a matter for debate (10,18,48). Injection of the morpholinos designed to inhibit the expression of the protein Mos, into prophase I oocyte, does not affect the entry into meiosis I, but the role of Mos in meiosis could be to prevent the mitotic cycle after meiosis I and to trigger the cell cycle arrest in oocytes, in order to attempt fertilization (10). On the other hand, injection of mRNA encoding mutant CPEB completely blocked endogenous c-Mos synthesis and oocyte maturation (9).
In Xenopus oocytes, the masking of mRNAs encoding c-Mos or cyclin B is mediated by the association of maskin with CPEB and eIF-4E making a loop that inhibits the association of the poly(A) polymerase onto the 3Ј-untranslated region (27). Progesterone stimulation of maturation induces the activation of Aurora-A and the phosphorylation of the protein CPEB on the Ser-174. This phosphorylation is the first step that controls the polyadenylation of mRNA encoding cyclin B1, which is necessary for the translation of the protein. This polyadenylation depends on the adenylation and translation of c-Mos (49). The phosphorylated CPEB protein recruits the CPSF protein and helps it to bind to the hexanucleotide and then recruits the poly(A) polymerase to the end of the mRNA (24).
Our study clearly demonstrates that Aurora-A phosphorylates maskin in vitro. Peptides mimicking the phosphorylated sequence of maskin act as competitors in the kinase assay and inhibit the phosphorylation of the substrate GST-p17. These peptides, which also inhibit Aurora-A-catalyzed phosphorylation of GST-H3 tail, or myelin basic protein, appear to be useful tools to inhibit Aurora-A kinase activity. The peptide mp1 is a substrate of Aurora-A. It is phosphorylated by Aurora-A and can inhibit by competition the phosphorylation of other substrates.
As demonstrated in vitro, we expected that, in vivo, the phosphorylation of maskin depends on Aurora-A. So in order to determine whether the phosphorylation of maskin by Aurora-A is involved in the oocyte maturation process, we studied the consequence of the injection of the mimicking peptides on the oocyte maturation. The mp1 peptide induced an early entry in metaphase I when injected in Xenopus oocytes before their stimulation with progesterone. The injection of a non-peptidic specific inhibitor of Aurora-A capable of preventing the in vitro phosphorylation of maskin and other substrates by Aurora-A (data not shown) had a similar effect on oocyte maturation. This strongly suggests that in vivo the peptide mp1 interfered with Aurora-A kinase activity. Our results clearly show that the inhibition of Aurora-A activity in oocytes stimulated with progesterone induces a precocious maturation by affecting mRNA translation.
How does the inhibition of Aurora-A lead to advanced progesterone-induced maturation of oocytes? Based on the appearance of the white spot on the oocytes, and confirmed by the premature synthesis of Mos, cyclin B1, and Cdc6, the inhibition of Aurora-A provokes an advance of GVBD. This advance never exceeds 2 or 3 h, which corresponds to the amount of time necessary for the oocytes to complete meiosis I after GVBD. Our hypothesis is that Aurora-A phosphorylates maskin to ensure that it stays associated with CPEB between GVBD and meiosis I and that, consequently, a specific population of mRNA remains masked. The inhibition of Aurora-A, which would prevent maskin phosphorylation, would eliminate the 2-h gap between GVBD and the end of meiosis II and trigger the translation of mRNAs that are normally translated in meiosis II, or at least when meiosis I is complete. It is thus possible that Aurora-A phosphorylates maskin to prevent synthesis of meiosis II proteins during meiosis I. Aurora activity would then protect meiosis I from protein required for meiosis II.
The exact role of Aurora-A in meiotic maturation is still not clear. On the one hand, the ectopic expression of the active kinase is known to accelerate GVBD in Xenopus (50). On the other hand, our results show that the inhibition of the kinase activity also induces an advance in oocyte maturation. 2 However, in contrast to control oocytes, the synthesis of both c-Mos and cyclin B1 is initiated at the same time after GVBD in mp1-injected oocytes. Also, Cdc6 that controls the competence of the oocyte to replicate its DNA must be repressed in meiosis I, and the repression must be removed in meiosis II to allow DNA replication after fertilization (26,51). This protein is synthesized earlier in mp1-injected oocytes also. Aurora-A exhibits a biphasic pattern of phosphorylation and activation that coincides with that of MPF, and the expression of a constitutively active Aurora-A appears to be incompatible with the meiosis I to meiosis II transition (52). In vivo, the phosphorylation of maskin by Aurora-A could protect the oocytes going through meiosis from entering the mitotic cycle without DNA synthesis. As the Aurora-A activity decreases at the meiosis I to meiosis II transition, the two proteins would temporarily dissociate. The consequence of this dissociation at the end of the meiosis I would be the loss of the masking function of the complex eIF4E-maskin-CPEB. Consequently, the polyadenylation and then translation of the mRNAs encoding c-Mos and cyclin B1 could happen.
In conclusion, we indicate the direct interaction of Aurora-A and maskin both in vitro and in vivo. We show that Aurora-A phosphorylates maskin in vitro on Ser-626, and we further demonstrate that maskin is a new substrate/activator of Aurora-A. The function of maskin phosphorylation by Aurora-A during oocyte maturation could be to ensure the masking of specific mRNAs to prevent, during meiosis I, the synthesis of proteins required solely for meiosis II and to prevent the oocyte from entering metaphase II prematurely.