Insulin stimulation of CPE-dependent polyadenylation. (A) In vitro synthesized radiolabeled RNA, containing or lacking a CPE, was injected into stage VI oocytes, some of which were incubated in the presence of progesterone or insulin. When a white spot was visible at the animal pole of these stimulated oocytes, the RNA was extracted and analyzed for polyadenylation by denaturing polyacrylamide gel electrophoresis. (B) Extracts prepared from unstimulated and stimulated (progesterone or insulin) oocytes were supplemented with purified recombinant histidine-tagged CPEB and γ³²P-ATP in an in vitro kinase assay. Following the kinase reaction, phospho-CPEB was analyzed by SDS-PAGE and autoradiography. Mos was detected by a Western blot, and MPF activity by an H1 kinase reaction. Extracts from insulin-stimulated oocytes were supplemented with recombinant wild-type CPEB, or recombinant CPEB with S174A and S180A mutations (see text for details), together with γ³²P-ATP. The resulting phosphorylated CPEB proteins were subjected to 2D phospho-peptide mapping. The arrow indicates the location of the peptide containing phospho-serine 174.

PI3 kinase and PKC-ζ are required for insulin-stimulated polyadenylation. (A) Oocytes were incubated with wortmannin for 2 h and then in some cases, injected with CPE-containing RNA, which was followed by stimulation with progesterone or insulin. MPF activity, mos protein levels, CPEB phosphorylation, and cytoplasmic polyadenylation were then assessed as described in Fig. 1. (B) Oocytes were incubated with rapamycin for 2 h, injected with CPE-containing RNA, and the same parameters as in panel A were assayed. (C) Oocytes were injected with PKC-ζ inhibitor peptide, PKC-α inhibitor peptide, or mock-injected, and induced to mature with insulin or progesterone. The same parameters as in A were assayed.

Aurora A is differentially phosphorylated during oocyte maturation. (A) E. coli-expressed phosphatase-treated Aurora A was incubated in extracts from immature or progesterone-stimulated oocytes primed with γ³²P-ATP. The resulting phosphorylated Aurora proteins were resolved by 2D phospho-peptide mapping. The top panel shows a short exposure; the bottom shows a longer exposure. Solid arrows indicate phospho-peptides that diminished during maturation; the open arrow denotes a phospho-peptide that increased during this time. (B) Identical phospho-peptide mapping procedures were performed with wild-type or S349A Aurora A proteins. The arrow denotes the phospho-peptide that contains S349.

GSK-3 mediated Aurora A phosphorylation decreases during oocyte maturation. (A) A Western blot (top panel) verifies that GSK-3 can be immunoprecipitated from immature extracts (lane 2), but not when extracts were incubated with protein A beads alone (lane 1). Lane 3 shows GSK-3 in unfractionated extracts. The bottom panel shows that immunoprecipitated GSK-3 phosphorylates a wild-type (WT) GSK-3 substrate peptide containing a “priming” phosphate at S21 (derived from glycogen synthase), but not a peptide with an S21A mutation, which removes the priming phosphorylation. The site of GSK-3 phosphorylation on the peptide (at S17) is noted. (B) A peptide derived from Xenopus Aurora A was phosphorylated with PKA and then used as a substrate for GSK-3 that had been immunoprecipitated from extracts derived from immature oocytes. The relative amount of peptide phosphorylation was quantified by a phosphorimager (left). This same peptide, as well as one with S290A/S291A mutations, was used as a substrate for GSK-3 that had been immunoprecipitated from control, progesterone-treated, or insulin-treated oocytes (the boxed Western blot shows equal GSK-3 immunoprecipitation). As above, the extent of phosphorylation was quantified by a phosphorimager (right). (C) Recombinant Xenopus wild-type Aurora A, or an Aurora with T294/295A mutations were Western blotted and probed with anti-human Aurora A phospho-specific T295 antibody. The gel was also stained with Coomassie blue to show equal loading of the gel (left). Protein from immature, progesterone-treated, and insulin-treated oocytes was Western blotted and probed with both the Aurora A phospho-specific T295 antibody and conventional Aurora A antibody.

Aurora-stimulated CPEB phosphorylation is required for insulin-induced maturation. (A) Oocytes were injected with a mos antisense oligonucleotide and then stimulated with progesterone or insulin. Some of the oocytes were also injected with CPE-containing radiolabeled RNA. Mos, MPF, and polyadenylation were analyzed as described in Fig. 1. (B) Oocytes were injected with in vitro synthesized mRNAs encoding myc-tagged wild-type CPEB or a CPEB with S174A and S180A mutations, some of which were then stimulated with insulin. The oocytes were then scored for maturation (GVBD), and the expression levels of the injected CPEB mRNAs were determined by ³⁵S-methionine labeling and immunoprecipitation with myc antibody, followed by SDS-PAGE and autoradiography (inset). Mos protein and MPF activity were assayed as described in Fig. 1. (C) A peptide corresponding to CPEB residues 163–181, and another containing the same sequence but with S174A, S180A changes were coupled to ovalbumin, and are referred to as Ova-WT and Ova-AA, respectively. The coupled peptides were subjected to in vitro kinase assays with E.coli-expressed Aurora A and analyzed by SDS-PAGE and autoradiography. Coomassie blue staining of the gel shows equal loading of peptide conjugates (peptide phosphorylation). Extracts from insulin-stimulated oocytes were either mock-depleted with pre-immune serum or immunodepleted with Aurora A antibody; the depleted fractions were then subjected to in vitro kinase assays with Ova-WT peptide, which was analyzed by SDS-PAGE and autoradiography. Coomassie blue staining of the gel shows equal loading of the peptide conjugate (immunodepletion of Aurora).

GSK-3 inhibits CPE-mediated translation in both insulin and progesterone signaling pathways. (A) Oocytes were injected with mRNA encoding a myc-tagged wild-type GSK-3 (Pierce and Kimelman 1995), some of which were then stimulated with insulin or progesterone. The oocytes were then scored for maturation (GVBD). The expression level of the injected GSK-3 mRNA was determined by a Western blot using myc antibody (inset). Polyadenylation, CPEB phosphorylation, mos protein levels, and MPF activity were assayed as in Fig. 1. (B) Oocyte extracts were immunoselected with GSK-3 antibody, or pre-immune serum, and then Western blotted and probed with Aurora A antibody (top). Purified kinase-dead Aurora A with a K169R mutation (to eliminate possible autophosphorylation) was incubated with recombinant GSK-3 and γ³²P-ATP; phospho-Aurora was then analyzed by both SDS-PAGE and 2D phospho-peptide mapping. Several truncated Aurora A proteins, as well as a protein with S290/291A mutations, were expressed in E. coli, subjected to phosphorylation by recombinant GSK-3 in vitro, and analyzed by SDS-PAGE (bottom panel).

GSK-3 inhibits Aurora A kinase activity. (A) Aurora A was incubated with or without recombinant GSK-3, purified, then used to phosphorylate the CPEB-derived peptide substrate. Top panel: the radiolabeled CPEB peptide (Ova-WT peptide); middle panel: a Coomassie blue-stained gel showing equal loading of Ova-WT; lower panel: a Western blot of Aurora A showing equal amounts used in the assay. The histogram quantifies the amount of phosphorylated peptide from three experiments (P < 0.05, Student's t-test). (B) Phosphatase-treated Aurora A incubated with GSK-3 and γ³²P-ATP was analyzed by SDS-PAGE and 2D phospho-peptide mapping. The arrow refers to an Aurora A autophosphorylated phospho-peptide that was stimulated by recombinant GSK-3. (C) Aurora A (wild-type or proteins with S349A, S349D, S290A/S291A, or S290/291D mutations), some of which were pre-incubated with recombinant GSK-3, were used to phosphorylate the CPEB-derived peptide linked to ovalbumin. Top panel: radiolabeled Ova-WT peptide; middle and bottom panels: Coomassie blue staining of gels showing equal loading of Ova-WT and Aurora A. Note that the Aurora A proteins containing aspartate (D) substitutions have slightly faster electrophoretic mobilities than wild-type or alanine (A)-containing proteins. (D) E. coli-expressed phosphatase-treated Aurora A wild-type and S290A/S291A mutant were incubated in extracts from immature extracts supplemented with γ³²P-ATP. The resulting phosphorylated Aurora proteins were resolved by 2D phospho-peptide mapping as in Fig. 5. The arrow denotes the S349-containing phospho-peptide. (E) To determine the effects of Aurora A S290A/S291A and S349A on CPE-mediated mRNA polyadenylation, oocytes were co-injected with recombinant Aurora A proteins (wild-type or mutant) and in vitro synthesized radiolabeled CPE-containing RNA (cf. Fig. 1). The oocytes were incubated with progesterone (0.01 μg/mL), harvested 0, 2, or 4 h later, and processed for polyadenylation analysis (lower panel).

Proposed signaling pathways used by progesterone and insulin to stimulate oocyte maturation. Whereas insulin signals through insulin receptor substrate 1 (IRS-1), PI3 kinase, and PKC-ζ, progesterone uses none of these components. Both pathways, however, converge on and inactivate GSK-3. In the immature oocyte, GSK-3 phosphorylates Aurora A S290/S291, which in turn stimulates an Aurora A inhibitory autophosphorylation on S349. GSK-3 inactivation leads to the dephosphorylation of these residues, which is required for Aurora A activation. Other components, such as TPX2, may be involved in Aurora A activation at this time. It is noteworthy that T295 phosphorylation, which is important for kinase activity, is observed in both immature and mature oocytes. Active Aurora A then phosphorylates CPEB S174, which induces polyadenylation and translation of mos mRNA. Mos protein subsequently stimulates the Map kinase cascade, MPF activation, and oocyte maturation.