SNT1/FRS2 Mediates Germinal Vesicle Breakdown Induced by an Activated FGF Receptor1 in Xenopus Oocytes*

The docking protein SNT1/FRS2 (fibroblast growth factor receptor substrate 2) is implicated in the transmission of extracellular signals from the fibroblast growth factor receptor (FGFR), which plays vital roles during embryogenesis. Activating FGFR mutations cause several craniosynostoses and dwarfism syndromes in humans. Here we show that the Xenopus homolog of mammalian FRS-2 (XFRS2) is essential for the induction of oocyte maturation by an XFGFR1 harboring an activating mutation (XFGFR1act). Using a dominant-negative form of kinase suppressor of Ras, we show the Mek activity is required for germinal vesicle breakdown (GVBD) induced by co-expression of XFGFR1act and XFRS2, but this activity is not required for progesterone-induced GVBD. Furthermore, Mek/MAPK activity is critical for the induction and/or maintenance of H1 kinase activity at metaphase of meiosis II in progesterone-treated oocytes. An activated XFGFR1 containing a mutation in the phospholipase Cγ binding site (XFGFR1actY672F) displayed a reduced ability to induce cell-cycle progression in oocytes, suggesting phospholipase Cγ may not be necessary but that it augments XFGFR signaling in this system. Oocytes co-expressing XFGFR1act and XFRS2 showed substantial H1 kinase activity, but this activity was blocked when the oocytes were treated with the phosphatidylinositol 3-kinase inhibitor LY294002. Although phosphatidylinositol 3-kinase activity is essential for XFGFR1act/XFRS2-induced oocyte maturation, this activity is not required for maturation induced by progesterone. Finally, ectopic expression of Xspry2, a negative regulator of XFGFR signaling, greatly reduced MAPK activation and GVBD induced by the expression of either XFGFR1act plus XFRS2 or activated Ras (H-RasV12). In contrast, Xspry2 did not prevent GVBD induced by an activated form of Raf1, suggesting that Xspry2 exerts its inhibitory function upstream or parallel to Raf and downstream of Ras.

Fibroblast growth factor (FGFs) 1 represent a large family of at least 22 different growth factors. FGFs play a critical role in the control of multiple biological processes, including mitogenesis, angiogenesis migration, differentiation, and mesoderm induction (1,2). Upon activation, the FGF receptor dimerizes and undergoes rapid autophosphorylation on numerous tyrosine residues. Autophosphorylation of the cytoplasmic domain recruits SH2-containing molecules, such as phospholipase C␥ (PLC␥), Crk, and possibly Src, whereas a docking protein, SNT1/FRS-2 (here after referred to as FRS2), associates with the FGF receptor (FGFR) in a tyrosine phosphorylation-independent manner (2,3).
Upon tyrosine phosphorylation, FRS2 (3,4) engages the SH2 domains of Grb2 and Shp2 (5). In response to FGF stimulation in cultured cells, SNT1/FRS-2 and Gab1 associate indirectly via Grb2, resulting in tyrosine phosphorylation of Gab1 and activation of the PI 3-kinase/Akt pathway (6,7). FRS-2 has been shown to bind and link FGFR, TrkA, and the Ret receptors to MAP kinase (8 -14). The FGF receptor signaling pathway has been extensively studied in cultured cells, whereas the role of associated signaling components during development are beginning to emerge.
Targeted disruption of FGFs and FGF receptors indicates a crucial role for these polypeptides during mouse development. Mice lacking the FGFR1 or FGF-8 have severe gastrulation defects (15)(16)(17), and mice lacking FGF4 or the docking protein FRS2 exhibit early embryonic lethality (18,19). Mice null for FGF-9 display mesenchymal defects resulting in lung hypoplasia and neonatal death (20). Loss of the FGFR3 gene causes defects in skeletal morphogenesis (21), and FGFR2 null mice demonstrate that FGFR2 is required for trophectoderm development, lung-branching morphogenesis, and limb outgrowth (22,23).
Mutations in the human FGFRs have also helped elucidate the developmental role of FGFRs. For example, mutations in FGFR3 are associated with various forms of dwarfism (for review, see Ref. 24), and those in the FGFR1 and FGFR2 have been associated with craniofacial and limb deformities (24). In the mouse system, creation of gain-of-function mutations in FGFR2 and -3 cause similar phenotypes to the human syndromes (24,25). These genetic mutations provide evidence that FGFRs are involved in bone and limb growth as well as morphogenesis (24). The Xenopus embryo system has also provided valuable information regarding the role of FGFs and their receptors during development. Members of the FGF family control mesoderm production and maintenance as well as morphogenetic movements during gastrulation in Xenopus embryos. In ectodermal explant tissue, overexpression of FGF or treatment with FGF induces mesoderm (26,27). In embryos, a carboxylterminally truncated FGFR (XFD) has been demonstrated to inhibit the formation of most mesoderm and cause gastrulation defects (28,29). In contrast, constitutively activated forms of FGFR1 induce mesoderm in ectodermal explants (animal caps) (30).
Studies examining the role of FGFR-associated signaling molecules reveal that Grb-2 and Shp2 play critical roles in FGF-induced mesoderm tissue in explants as well as formation of posterior structures in embryos (31,32). In contrast, the ability to bind PLC␥ or mediate a signal through Crk is not required for mesoderm induction in explants (31,32). A Shp2 mutant blocks FGF-induced mesoderm as well as MAP kinase activation (32), which is an essential pathway for mesoderm induction by FGF (33)(34)(35)(36). Recently, PI 3-kinase activity has also been found to be critical for mesoderm induction and implicated to function in a pathway parallel to the Raf/MAP kinase pathway (37). A Xenopus homolog of FRS2 has recently been isolated. In embryos, proper XFRS2 function is required for development of trunk and posterior structures (38,39). Moreover, loss of XFRS2 function disrupts muscle and notochord formation and inhibits FGFR-induced MAP kinase activation (38). The Xenopus oocyte system has also proved useful for examining the events of signal transduction. Fully grown Xenopus oocytes are arrested in prophase of meiosis I and are induced to mature upon exposure to progesterone. Progesterone stimulates the synthesis of the Mos protooncogene product, pp39mos, and leads to the activation of maturation-promoting factor (MPF), an activity responsible for coordinating the biochemical events of meiosis I and II (40 -46). During meiosis I, the rise in MPF activity is coincident with phosphorylation and activation of MAPK (47)(48)(49).
Several RTKs have been shown to induce Xenopus oocyte maturation when activated constitutively or in response to growth factors. For example, germinal vesicle breakdown (GVBD) is induced by receptors for insulin-like growth factor (42,50), epidermal growth factor (51), nerve growth factor (52), and activated forms of hepatocyte growth factor (Tpr-Met) (53) and the glial cell-derived neurotrophic factor (Ret) (54). Although signal transduction by progesterone and RTKs results in the synthesis of Mos and the induction of MPF, the two pathways differ in several aspects. Although maturation induced by Tpr-Met and IGF-1 requires the stimulation of a phosphodiesterase, maturation induced by progesterone does not (42,53,(55)(56)(57)(58)(59). Furthermore, maturation-signaling cascades induced by IGF-1, but not those induced by progesterone, require the specific involvement of p21ras, GAP, and protein kinase C (60 -62).
To further investigate the signaling events elicited by an activated Xenopus FGFR1 harboring mutations identified in human dysmorphic syndromes, we expressed these activated mutant receptors in Xenopus oocytes. Our results indicate that the activated receptor can induce cell-cycle progression only when co-expressed with FRS2. Furthermore, Mek activity and PI 3-kinase activity are essential for GVBD mediated by an activated XFGFR1 and XFRS2, whereas PLC␥ assists but is not required for this event. Moreover, we show that the natural inducer of oocyte maturation, progesterone, does not require Mek activity or PI 3-kinase for GVBD. Finally, overexpression of Xsprouty2, a negative regulator of RTK signaling, can block GVBD and MAPK activation by the activated FGFR1 but not progesterone. This block appears to influence a point upstream or parallel to Raf and downstream of Ras.

EXPERIMENTAL PROCEDURES
Frogs, Oocytes, Constructs, RNA, and Injections-Xenopus laevis females were purchased from Nasco (Fort Atkinson, WI) and prepared as described previously (63). All XFGFR1 mutants were generated as previously described (30). These mutants were inserted into SP64T3 for this study, except XFGFR1K562E was inserted in pCS2ϩ. XFRS2 (38) and Xspry2 cDNAs were inserted into the pCS2ϩ plasmid. RafY340D, DnKSR (gift from Deborah Morrison), and H-RasV12 cDNAs are inserted into the SP64T plasmid. All capped mRNA was made using the SP6 mMessage mMachine kit as specified by the manufacturer (Ambion). 18 h after oocyte isolation, microinjections were performed using the attocyte injector (ATTO Instruments, Rockville, MD) with an injection volume of 30 nl. Specific amounts of various RNAs were injected into each oocyte: XFGFR1, XFGFR1act, XFGFR1kd, and XFGFR1actY672F at 14 ng, XFRS2 RNA at 7 ng, Tpr-met RNA at 0.5 ng, H-RasV12 RNA at 2 ng, and Raf1act RNA at 7 ng. Experiments employing XFGFR1act contained the activating mutation K562E or C289R. Experiments using DnKSR RNA or Xspry2 RNA were performed by microinjection of these RNAs (14 ng/oocyte) and culturing the oocytes for 2-4 h in 50% L-15 media with 1% bovine serum albumin before the second set of RNAs was injected. The experiments involving these proteins were performed in this manner to ensure adequate translation of these proteins. Oocytes were generally scored 16 -20 h later for GVBD, as evidenced by the appearance of a white spot at the animal pole. This observation was verified in many cases by manual dissection of oocytes after fixation in 8% trichloroacetic acid.
Histone H1 Assays-Crude MPF extracts were prepared and assayed as described previously (63). Samples were resolved by SDS-PAGE (16% polyacrylamide gel), and phosphorylated histone H1 was detected by autoradiography.

RESULTS
Although several RTKs can induce cell-cycle progression in fully grown Xenopus oocytes (42, 50 -54) that are arrested at the G 2 /M border, expression of an activated form of the XFGFR1 fails to induce the G 2 /M transition (data not shown). The endogenous FGF receptor1 transcripts are stored as stable untranslated maternal mRNA in fully grown oocytes until translationally activated at meiotic maturation (64). We reasoned that a signaling molecule specific to the XFGFR may be present at limiting levels in immature oocytes, since other activated RTKs can induce oocyte maturation. Thus, we expressed Xenopus FRS2 alone or with activated (XFGFR1act, wild-type (XFGFR1), and kinase-dead (XFGFR1kd) forms of the XFGFR1 in stage 6 oocytes. Injection of XFRS2 RNA (7 ng) alone or with XFGFR1kd (14 ng) had no observable effect on the oocytes (Table I), but when XFRS2 was co-injected with the activated XFGFR1 (XFGFR1act) RNA (14 ng), GVBD was ini-tiated in 76% of oocytes (Table I, Fig. 1A). Oocytes displayed the hallmark "white spot" on the animal pole, and H1 kinase assays performed on extracts from these oocytes confirmed the morphological data. H1 kinase activity is only present in extracts from oocytes co-injected with XFRS2 and XFGFR1act RNA (Fig. 1B). Western analysis demonstrates that approximately equivalent levels of the normal and mutant forms of XFGFR1 are expressed in the oocytes (Fig. 1C). These results suggest that the expression of XFRS2 is sufficient to allow only activated forms of XFGFR1 to induce oocyte maturation.
These data are in contrast to a previous study (65) showing that expression of the Pleurodeles FGFR1 in Xenopus oocytes induces maturation upon exposure to FGF-1 and when a platelet-derived growth factor receptor (PDGF) extracellular domain/Pleurodeles FGFR1 fusion protein is expressed and activated by PDGF. It may be possible that the differences in results may be attributed to species (Pleurodeles or Xenopus) differences, receptor expression levels, or possibly due to an intrinsic activity difference in XFGFR1 receptors that are activated as a result of mutations found in human disorders. To explore whether an intrinsic difference between the wild-type and activated XFGFR1 exists in the oocyte system, we examined whether the wild-type or mutant Xenopus FGFR1 can induce oocyte maturation upon ligand stimulation in the absence or presence of Xenopus FRS2. Oocytes were injected with RNA encoding XFGFR1 or XFGFR1act alone (14 ng) or coinjected with XFRS2 (7 ng) and either left untreated or treated with basic FGF (200 ng/ml). Oocytes co-expressing XFGFR1 and XFRS2 underwent GVBD (80% ; Table II) after stimulation with bFGF, whereas the unstimulated oocytes or bFGF-stimulated oocytes that were only injected with XFGFR1 RNA or XFGFR1kd plus XFRS2 RNA did not mature (Table II). As expected, oocytes injected with XFGFR1act and XFRS2 displayed GVBD regardless of whether the oocytes were treated with bFGF (93% Ϯ bFGF; Table II), suggesting that there is no obvious intrinsic difference between the constitutively active or ligand-activated XFGFR1 regarding induction of oocyte maturation.
FGFR-induced tyrosine phosphorylation of FRS-2 causes activation of the Ras/MAP kinase pathway via recruitment of Grb-2, SOS, and Shp-2. It has also been recently shown that in response to FGF stimulation, FRS2 and Gab1 associate indirectly via Grb2, resulting in tyrosine phosphorylation of Gab1 and activation of the PI 3-kinase/Akt pathway (7,19). We tested whether expression of XFRS2 allowed the XFGFR1act or ligand-stimulated XFGFR1 to activate MAPK in the oocyte. Western analysis demonstrates that roughly equivalent levels of XFGFR1act or XFGFR1 are present in the extracts (Fig. 2). Phosphorylated MAPK, which correlates with MAPK activity, was also examined. A basal level of MAPK phosphorylation is detected upon bFGF stimulation of oocytes expressing XFGFR1 or XFGFR1act in the absence of XFRS2 and when XFGFR1 and XFRS2 are co-expressed in oocytes in the absence of bFGF (Fig.  2). However, a dramatic induction of MAPK phosphorylation and, thus, MAPK activity is observed when the FGF receptor is activated by mutation or ligand in the presence of XFRS2 (Fig.  2). This induction may, in part, be due to the increased MAPK activity maintained by MPF (cyclinB/cdc2). These data show that in oocytes, XFGFR1act can induce a low level of MAPK  To determine whether the Mek/MAPK pathway was necessary for XFGFR1act-induced GVBD, we employed a dominantnegative form of kinase suppressor of Ras (DnKSR) (66). DnKSR has been shown to block MAPK activation by preventing the appropriate localization of Mek, an activator of MAPK (67). DnKSR RNA (14 ng/oocyte) was introduced into oocytes. Two hours later, oocytes were injected with XFGFR1act (14 ng) plus XFRS2 RNA (7 ng) or Tpr-met RNA (0.5 ng) or treated with progesterone (2 g/ml). Although XFGFR1act plus XFRS2 RNA-injected oocytes displayed GVBD (82% ; Table III), the co-expression of DnKSR almost completely blocked maturation (1% GVBD; Table III) as evidenced by the presence of the germinal vesicle (Fig. 3A). DnKSR also reduced the percentage of oocytes displaying GVBD induced by another tyrosine kinase, Tpr-met, from 97 to 3% (Table III). In contrast, expression of DnKSR did not block progesterone-treated oocytes from undergoing GVBD (93%) ( Table III, Fig. 3A).
Western analysis of extracts prepared from these oocytes demonstrates that substantial phospho-MAPK can only be observed in extracts from oocytes either injected with XFGFR1act plus XFRS2 RNA or Tpr-met RNA or treated with progesterone (Fig. 3B). Expression of DnKSR was able to effectively block MAPK activation from both the FGFR1 pathway and the progesterone pathway. Western analysis of the extracts shows that XFGFR1act is expressed at equivalent levels in the presence or absence of DnKSR (Fig. 3C). H1 kinase activity was examined to provide biochemical evidence regarding the induction or inhibition of GVBD in these oocytes (Fig. 3D). Because H1 kinase activity is necessary for both the induction of GVBD during meiosis I and for the metaphase arrest of oocytes during meiosis II, groups of progesterone-treated oocytes were collected at both GVBD and at least 6 h post-GVBD for H1 kinase assays. H1 kinase activity was readily detected in oocytes injected with XFGFR1act plus XFRS2 RNA or Tpr-met RNA or treated with progesterone. However, this activity was dramatically reduced when DnKSR RNA was also injected (Fig. 3D). H1 kinase activity was extremely low in the oocytes where GVBD had been blocked by DnKSR expression (XFGFR1act/ XFRS2-or Tpr-met-expressing oocytes). However, progesterone-treated oocytes displayed abundant H1 kinase activity at GVBD regardless of whether DnKSR was co-expressed. In contrast, H1 kinase activity was nearly absent in DnKSR-expressing oocytes that were treated with progesterone and collected several hours after GVBD, suggesting that initiation and/or maintenance of a metaphase II arrest was blocked (Fig. 3D). Collectively, these data demonstrate that Mek/MAPK signaling is essential for FGF receptor-induced but not progesteroneinduced GVBD in oocytes.
PLC␥ associates with the activated FGFR1 via phosphorylated tyrosine 766 and becomes tyrosine-phosphorylated and activated, leading to hydrolysis of phosphatidylinositol 4,5bisphosphate (2). The biological significance of this activation is not completely clear. To test whether PLC␥ binding and activation play a role in XFGFR1act-induced cell cycle progression, we generated an equivalent PLC␥ binding mutation in the context of the XFGFR1act protein (XFGFR1actY672F). To confirm that this mutation inhibits the association of the XFGFR1 with PLC␥, RNA encoding XFGFR1act (14 ng), FGFR1KD (14 ng), and XFGFR1actY672F (14 ng) was injected into oocytes, and the XFGFR1 was immunoprecipitated using XFGFR1specific antibodies. The immune complexes were subjected to Western analysis using anti-PLC␥ antibodies. PLC␥ coimmunoprecipitated with the XFGFR1act, but not the XFGFR1actY672F or the XFGFR1KD, indicating that PLC␥ signaling from the XFGFR1actY672F was abrogated (Fig. 4A). Western analysis of the oocyte extracts shows that comparable expression of the various XFGFR1 mutants was obtained (Fig. 4B).
The XFGFR1actY672F mutant was expressed in oocytes either alone or in combination with XFRS2 in three separate experiments. Although co-injection of XFGFR1act and XFRS2 RNA resulted in GVBD in 78% (56/71) of oocytes, co-expression of XFGFR1actY672F and XFRS2 yielded GVBD in 13% (12/91) of oocytes. Although these data demonstrate that PLC␥ binding is not essential for GVBD induced by the activated XFGFR1, PLC␥ association enhances the signals necessary for cell cycle progression in this system. This observation was biochemically confirmed by performing H1 kinase activity assays on the oocyte extracts. H1 kinase activity was not observed in extracts prepared from oocytes that did not display the white spot on the animal pole but had been injected with RNA encoding XFGFR1actY672F and XFRS2 (Fig. 4C). In contrast, H1 kinase activity was observed in the small percentage of oocytes displaying the white spot (Fig. 4C). MAPK phosphorylation was examined by Western analysis using anti-phospho-MAPK antibodies. A low level of MAPK phosphorylation was observed in extracts  from immature oocytes expressing XFGFR1actY672F alone or in conjunction with XFRS2, but a high level of MAPK activity was detected in extracts from those oocytes that underwent GVBD (Fig. 4D). Western analysis shows that the XFGFR1actY672F protein was appropriately expressed in all cases (Fig. 4E). The docking protein FRS2 is a major downstream effector that links FGF and nerve growth factor receptors with the Ras/MAPK signaling cascade. Recently, it has been demonstrated that FRS2 also plays a pivotal role in FGF-induced recruitment and activation of PI 3-kinase (7). To determine whether PI 3-kinase activity was necessary for XFGFR1actinduced GVBD, XFGFR1act RNA was injected alone or with XFRS2 RNA and either treated with the PI 3-kinase inhibitor LY294002 (100uM) or left untreated. As controls, a group of oocytes was also treated with progesterone (2 g/ml) in the absence or presence of LY294002 (100 M). Three independent experiments demonstrated that the PI 3-kinase inhibitor completely blocked GVBD induced by XFGFR1act plus XFRS2, reducing GVBD from 82% (49/60) to 0% (0/70). LY294002 had no effect on GVBD in the progesterone-treated oocytes (93% GVBD, 28/30 progesterone alone; 90% GVBD, 27/30 progesterone plus LY294002). As expected, extracts from oocytes treated with progesterone in the presence or absence of LY294002 displayed significant H1 kinase and MAPK activity (Fig. 5). Oocytes expressing XFGFR1act plus XFRS2 showed substantial H1 kinase and MAPK activity, but almost no activity was observed when these oocytes were treated with LY294002 (Fig.  5). These results are consistent with the study using Pleurodeles FGFR1 (65) and suggest that PI 3-kinase activity is essential for XFGFR-induced oocyte maturation but is dispensable for progesterone-induced maturation.
Finally, because the previous experiments were designed to determine the contribution made by FRS2-dependent and -independent positive signaling components to FGFR1-induced oocyte maturation, we thought it necessary to examine an inhibitory component of FGFR1 signaling. One such inhibitor of FGF signaling, termed Sprouty (Spry), was originally identified in Drosophila (68). Spry mutations in Drosophila lead to excessive tracheal branching (68), which is a similar phenotype to that observed with inappropriate activation of FGF signaling. Genetic evidence indicates that Spry is an inhibitor of the FGF pathway, but it is unclear where it may function in this pathway.
We first examined the influence that Xenopus Spry may have on XFGFR1/XFRS2-induced GVBD in oocytes. One group of oocytes was injected with XFGFR1act plus XFRS2 RNA alone or co-injected with RNA encoding a Myc-tagged version of Xenopus Spry2. Another group of oocytes was either left uninjected or injected with Xspry2-Myc-tagged RNA or treated with the Mek inhibitor U-0126 and subsequently treated with progesterone 4 h later. When Xspry2 was expressed at high levels, it effectively blocked GVBD in the oocytes expressing XFGFR1act and XFRS2 (72% GVBD was reduced to 8%, Table  IV). In contrast, the vast majority of oocytes treated with progesterone underwent GVBD regardless of whether they were treated with the Mek inhibitor or expressed Xspry2 (90% GVBD and 93% GVBD, respectively; Table IV). Western analysis demonstrated that Xspry2 substantially reduced the phosphorylation of MAPK in oocytes expressing XFGFR1act and XFRS2, whereas Xspry2 appeared to have no effect on the phosphorylation of MAPK induced by progesterone (Fig. 6A). Control oocytes treated with both progesterone and the Mek inhibitor (U-0126) displayed a low level of phosphorylated MAPK, similar to the levels displayed in the Xspry2/ XFGFR1act/XFRS2-expressing oocytes (Fig. 6A). Western analysis with anti-XFGFR1 antibodies or anti-Myc antibodies demonstrated that XFGFR1act and Myc-tagged Xspry2 were appropriately expressed (Fig. 6, B and C, respectively). These results indicate that Xspry2 overexpression is able to dramat-ically reduce GVBD and MAPK activity induced by XFGFR1 signaling but not that induced by progesterone.
It is still unclear at what point in the FGFR-signaling pathway Spry exerts its influence. Xspry2 has been shown to inhibit Ca 2ϩ mobilization independent of the MAPK block (69). There are also studies suggesting that Spry inhibits FGFR signaling by blocking Ras activation (70,71) and other studies suggesting that Spry functions downstream by inhibiting Raf (72,73). In an effort to define where Xspry2 may function in the Ras/ MAPK pathway, Xspry2 was co-expressed with an activated form of Ras or Raf1 (Fig. 7). Oocytes were either left alone or injected with Xspry2 RNA. Four hours later, these oocytes were either treated with progesterone or injected with RNA encoding XFGFR1act plus XFRS2 or H-RasV12 or Raf1Y340D. Oocytes were scored for GVBD the following day. GVBD was observed in the majority of oocytes expressing either the activated XF-GFR1 plus XFRS2 or H-RasV12 (75 and 93% GVBD, respectively; Table V), but few displayed GVBD when Xspry2 was co-expressed (5 and 15% GVBD, respectively; Table V). In contrast, oocytes expressing the activated form of Raf1 (RafY340D) exhibited GVBD regardless of whether Xspry2 was also expressed (77 and 71% GVBD, respectively; Table V). The same result was observed for oocytes expressing an activated form of Mek (data not shown).
We next examined whether H1 kinase activity and MAPK activity was also affected by Xspry2 in these oocytes. Western   blot analysis (Fig. 7A) and the H1 kinase assays (Fig. 7F) were performed on the oocyte lysates. Both phospho-MAPK and H1 kinase activity are greatly reduced when Xspry2 is co-expressed with either XFGFR1act plus XFRS2 or activated Ras (H-RasV12) but not when co-expressed with activated Raf1 (Raf1Y340D) or treated with progesterone (Fig. 7, A and F). Western analysis of the oocyte lysates shows that Xspry2, XFGFR1act, RasV12, and RafY340D were all expressed at appropriate levels in the indicated oocytes (Fig. 7, B-E). Collectively, these data suggest that Xspry2 blocks XFGFR1-in-duced maturation and MAPK activity upstream or parallel to Raf and downstream of Ras. DISCUSSION The Xenopus oocyte system was used to examine the contributions made by FRS2 to FGFR1 signaling. Here we show that Xenopus FRS-2 is essential for the induction of oocyte maturation by an FGFR1 harboring activating mutations (XFGFR1act) that are associated with skeletal abnormalities in humans.
We also show that ligand-activated exogenously expressed wild-type XFGFR1 also is dependent upon exogenous XFRS2. Using a dominant-negative form of kinase suppressor of Ras (KSR), we show that Mek/MAPK activity is required for GVBD induced by co-expression of XFGFR1act and XFRS2. Collectively, these data suggest that XFRS2 may be limiting in the oocyte, and it is required for appropriate transmission of FGFR1 signals leading to cell cycle progression.
The role of MAPK during oocyte maturation has been an area of great interest. Our data show that H1 kinase activation in response to FGFR signaling absolutely requires Mek/MAPK activation. In contrast, Mek/MAPK activation is not essential for H1 kinase activity initiated by the natural inducer progesterone during meiosis I. Mek/MAPK activity leads to the activation of Rsk, which plays an important role in both GVBD and arrest in meiosis I (40, 74 -76). Although it is generally assumed that Mek/MAPK activity is necessary for progesteroneinduced GVBD (77), a few studies provide evidence questioning this concept (63,76,78). One study showed that inhibition of the MAPK pathway by the specific Mek inhibitor, U-0126, allowed oocytes to enter meiosis I after progesterone treatment but failed to form metaphase I spindles and entered S phase rather than meiosis II (79). The data presented here provide clear evidence that Mek/MAPK is not necessary for the induction of GVBD in meiosis I but is critical for the induction and/or maintenance of H1 kinase activity during meiosis II in Xenopus oocytes.
Activation of the FGFR initiates autophosphorylation at a carboxy-terminal tyrosine residue that leads to binding and activation of PLC␥, which in turn results in activation of the protein kinase C and calcium signaling (2). In vitro functional assays using a PLC␥ binding mutant of FGFR1 have failed to show a requirement for this pathway in any FGFR1-induced response. Recently, a mouse harboring a mutation of the tyrosine 766 of the FGFR1 was generated and shown to have alterations in A-P patterning of the vertebral column in a direction opposite to hypomorphic alleles (80). From this study it was suggested that phosphorylation of Tyr-766 may play a role in the negative regulation of FGFR1 activity in vivo (80). In another study, use of the PLC␥ inhibitory peptide on oocytes expressing FGFR1 or the stimulation by platelet-derived growth factor (PDGF)-BB of oocytes expressing a PDGF receptor/FGFR1 chimeric protein mutated on the PLC␥ binding site prevented GVBD and MAPK phosphorylation (65). In our study, an activated XFGFR1 containing a mutation in the PLC␥ binding site (XFGFR1actY672F) displayed a reduced ability to induce cell-cycle progression in oocytes, suggesting that PLC␥ may not be required but does augment FGFR1 signaling in this system.
The role of PI 3-kinase in progesterone-induced maturation is unclear. PI 3-kinase is activated in response to both progesterone and insulin (81)(82)(83)(84). One report shows that a dominant negative form of PI 3-kinase inhibits progesterone-induced oocyte maturation (81), whereas another indicates that the SH2containing inositol phosphatase (SHIP) inhibits the induction of oocyte maturation by overexpressed PI 3-kinase and insulin but does not inhibit progesterone-induced maturation (84).  Furthermore, the PI 3-kinase inhibitor wortmannin has been reported to block insulin-induced maturation but not progesterone-induced maturation (82,83). Most recently, the Xenopus progesterone receptor has been shown to co-precipitate with active PI 3-kinase (85). However, the PI 3-kinase inhibitor, wortmannin, only delayed progesterone-induced maturation while completely blocking the insulin-dependent maturation (85). In our study, oocytes co-expressing XFGFR1act and XFRS2 showed substantial H1 kinase activity, but this activity was blocked when the oocytes were treated with the PI 3-kinase inhibitor LY294002. These data strongly indicate that although PI 3-kinase activity is essential for XFGFR1act/ XFRS2-induced oocyte maturation, this activity is not required for GVBD induced by progesterone. Finally, we examined whether a negative regulatory protein in FGF signaling, Spry, could influence XFGFR1act/XFRS2induced oocyte maturation. It has been recently shown in Xenopus oocytes that Xspry2 acts downstream of the activated XFGFR1 to inhibit calcium mobilization but does not block Ras/MAPK activity (69). It is worth mentioning that in our studies, high levels of Xspry2 expression were required to inhibit GVBD and MAPK activity induced by XFGFR1act plus XFRS2, and this inhibition may reflect a higher level of expression than in the previous report (69). In either case, Nutt et al. (69) show that Ca 2ϩ mobilization induced by an activated FGFR1 is effectively blocked by Xspry2 expression, but the mechanism is not known. Activated PLC␥ leads to the release of calcium stores from the endoplasmic reticulum, whereas DAG and Ca 2ϩ activate PKC (2). Xspry2 does not inhibit the recruitment of PLC␥ to activated FGFR1, a step critical to Ca 2ϩ efflux induced by FGFR. 2 Thus the mechanism by which Xspry2 inhibits Ca 2ϩ release remains to be determined.
Recent studies suggest that Sprouty functions either upstream or downstream of Ras (69 -73). In our study, we demonstrated that overexpression of Xspry2 greatly reduced MAPK activation of and GVBD induced by the activated XFGFR1 plus XFRS2 or an activated Ras (H-RasV12) but not an activated Raf1. The present data are in agreement with a recent report indicating that Spry2 inhibits MAPK activation by inhibiting activation of Raf1 (73). Because Spry2, but not Spry1 and 4, inhibits Raf1 activation in transiently transfected 293T cells (73), it will be of interest to determine whether Spry1 and Spry4 have any effect on GVBD and MAPK activation in Xenopus oocytes. It is still formally possible that Xspry2 is a scaffolding protein that mediates and regulates Ras/Raf signaling from FGFR1 and that when this protein is overexpressed it acts as an inhibitory protein perhaps in a manner similar to Kermit (86) or KSR (66). Regardless, our results suggest that when overexpressed in oocytes, Xspry2 can exert an inhibitory function upstream or parallel to Raf and downstream of Ras. Thus Xenopus oocytes provide a useful model to dissect the functions and site of action of Spry and perhaps other novel inhibitors of the FGFR and Ras pathways.