Molecular cloning and characterization of an adaptor protein Shc isoform from Xenopus laevis oocytes

transcript

Molecular cloning and characterization of an adaptorprotein Shc isoform from Xenopus laevis oocytes

Franck Chesnel a,*, Christophe Heligon b, Laurent Richard-Parpaillon c, Daniel Boujard d

a UMR 6061-CNRS « Génétique et développement », IFR 97 – Université de Rennes 1 – 35042 Rennes cedex, Franceb Institute of Pharmaceutical Sciences, Swiss Federal Institute of Technology, Winterthurerstrasse 190, 8057 Zürich, Switzerland

c Centre Européen des Sciences du Goût, CNRS, 15, rue Hugues-Picardet, 21000 Dijon, Franced UMR 6026-CNRS « Interactions Cellulaires et Moléculaires », IFR 98 – Université de Rennes 1 – 35043 Rennes cedex, France

Received 31 March 2003; accepted 19 May 2003

Abstract

In order to gain further insight into IGF-1 receptor signaling in Xenopus laevis oocytes and embryos, we have undertaken thecharacterization of the adapter protein Shc and studied its implication in oocyte maturation induced after IGF-1 receptor activation, especiallysince expression of this molecule has been indirectly evidenced in Xenopus oocytes, eggs and embryos. We report herein the cloning fromXenopus postvitellogenic oocytes of a complementary DNA encoding a protein of 470 amino acids which shows the higher identity with themammalian adaptor protein p52ShcA. Western blot analysis using homologous antibodies evidenced a 60-kDa protein, p60XlShc, that ispredominantly expressed in oocytes and in early embryos. We also demonstrate that, like p60XlShc, Grb2 and the guanine nucleotide exchangefactor Sos are expressed in oocytes throughout vitellogenesis and in early embryos and that overexpression of a dominant-negative form ofGrb2 specifically inhibits insulin-induced resumption of meiosis. We finally show that Grb2 binds to p60Shc in oocytes specifically uponinsulin treatment. Altogether, these results suggest that Shc and Grb2-Sos are implicated in ras-dependent Xenopus oocyte maturation inducedby insulin/IGF-1; they also indicate that inability of insulin/IGF-1 to activate the Ras-MAPK cascade in vitellogenic oocytes does not resultfrom an insufficient expression level of Shc, Grb2 and Sos.

Keywords: IGF-1 receptor; MAPK cascade; meiosis resumption; Xenopus oocytes

1. Introduction

The type 1 insulin-like growth factor receptor (IGF-1R), isa transmembrane tyrosine kinase receptor involved in cellproliferation, transformation, differentiation and apoptosis(Baserga et al., 1999). In the amphibian Xenopus laevis,IGF-1R RNA expression is detected throughout oogenesis,in eggs and in the early phase of embryogenesis with atransient decrease in embryos around the midgastrula stage(Groigno et al., 1996; Zhu et al., 1998). In vitro experimentsindicate that IGF-1 can stimulate vitellogenin uptake ingrowing oocytes (Wallace and Misulovin, 1978; Opresko andWiley, 1987), glucose uptake and glycogen metabolism inpostvitellogenic oocytes (Janicot and Lane, 1989; Thomaset al., 1997), and progression of oocyte from prophase I to

metaphase II during meiosis, also called “oocyte maturation(El-Etr et al., 1979; Maller and Koontz, 1981). It also partici-pates in the formation of the head during organogenesis (Peraet al., 2001; Richard-Parpaillon et al., 2002).

Even if the natural inducer of oocyte maturation in Xeno-pus is the steroid progesterone secreted by surrounding fol-licle cells in response to gonadotropins (Masui and Clarke,1979), much attention has been paid to the intracellularmechanisms underlying both insulin- and progesterone-induced maturation. Thus, meiotic progression triggered byinsulin or IGF-1 specifically requires Ras and Raf-dependentactivation of MAPK (Deshpande and Kung, 1987; Korn etal., 1987; Biocca et al., 1993; Chung et al., 1997) and activa-tion of the phosphatidyl 3-kinase (PI(3)K) pathway (Liu etal., 1995; Lopez-Hernandez and Santos, 1999), which ulti-mately lead to the conversion of the p34cdc2-cyclinB complexinto active M-phase promoting factor (MPF). Surprisingly,insulin mitogenic effect is only observed in postvitellogenic,

* Coresponding author.E-mail address: Franck.chesnel@univ-rennes1.fr (F. Chesnel).

Biology of the Cell 95 (2003) 311–320

www.elsevier.com/locate/bicell

stage VI, oocytes whereas IGF-1R is already present (Groi-gno et al., 1996) and partially functional in growing oocytessince insulin can induce ribosomal protein S6 phosphoryla-tion and a pH increase (Stith and Maller, 1984). This meioticincompetence results at least from the presence in smalloocytes of an active tyrosine kinase that strongly keeps newlyformed p34cdc2/cyclinB complexes in an inactive state (Rimeet al., 1995). We later confirmed such a defect upstream ofMPF activation and further demonstrated, in the specific caseof insulin-induced oocyte maturation, the inability of thepeptide to activate p21Ras in stage IV oocytes (Chesnel et al.,1997). In order to elucidate why insulin is unable to activateRas in growing oocytes, it was first necessary to clearlyidentify the components of the pathway linking IGF-1 recep-tor to p21Ras in meiotically competent oocytes.

Very little has been investigated about these early steps ofIGF-1R signaling in Xenopus oocytes. In mammalian cells,the predominant pathway which links IGF-1R to Ras succes-sively implies binding to a docking protein, like IRS (InsulinReceptor related Substrate) or Shc (Src homology/collagen).Tyrosine phosphorylation of these adaptor proteins canthereafter target the complex Grb2-Sos to the membrane toactivate p21Ras. It seems that a similar mechanism mayoccur in Xenopus. Indeed, the SH2/SH3 adapter protein Grb2has been cloned from gastrula embryos (GenbankAJ223061)and Grb2 overexpression in oocytes triggers oocyte matura-tion (Cailliau et al., 2001). Moreover, two reports indicate thepossible implication of Grb2 (Aroca et al., 1996) as well asthe guanine nucleotide exchange factor Sos (Chie et al.,1999) in mediating insulin-induced meiotic progression.Two Xenopus IRS-like proteins have been cloned so far froma oocyte cDNA library (Liu et al., 1995; Ohan et al., 1998).Overexpression of native or mutated forms of IRS respec-tively resulted in a specific facilitation or inhibition ofinsulin-induced oocyte maturation (Chuang et al., 1993;Yamamoto-Honda et al., 1996; Ohan et al., 1998). However,these IRS did not fulfil all the requirements such as in vivophosphorylation and complex formation with Grb2-Sos uponoocyte exposure to IGF-1R ligand. Therefore, the implica-tion of an other docking protein like Shc should be hypoth-esized especially with the increasing number of reportsshowing that signalling through Shc to MAPK plays a criticalrole in mediating IGF-I- and insulin-stimulated mitogenesisin mammalian cells (see Butler et al., 1998 for review). Inmammals, Shc gene codes for 46-, 52-, 66-kDa isoforms thatare ubiquitously expressed. p46Shc and p52Shc are both ex-pressed from the same mRNA transcript by alternate transla-tional initiation sites and thus differ in the extent of theiramino-terminal domain sequences (Pelici et al., 1992). Incontrast, p66Shc results from a distinct transcript of the samegene by alternative splicing (Migliaccio et al., 1997). Thethree Shc proteins share a carboxyl-terminal Src homology 2(SH2) domain, a central glycine/proline-rich region homolo-gous to the a1 chain of collagen (CH1), and an amino-terminal region containing a PTB domain (for phospho-tyrosine-binding domain). Interestingly, there are indirect

evidences of Shc expression in Xenopus, at least in eggs andembryos (Aoto et al., 1999; Dupont and Blancq, 1999),which led us to undertake the characterization of the XenopusShc isoform(s) in oocytes and their participation in IGF-1Rsignaling during oogenesis and early embryogenesis.

2. Results

Cloning of Xenopus Shc from postvitellogenic oocytes.To obtain a fragment of Xenopus Shc cDNA, the nucle-

otide sequences of human, mouse, rat, and drosophila Shcwere aligned and consensus-degenerate hybrid primers weredesigned in the most conserved regions using the CODEHOPsoftware. The cDNA from Xenopus stage VI oocytes wasamplified by PCR with various combinations of these prim-ers but only one pair (5'-GGACGTGATCGGCACCATHGG-NMARGC-3' and 5'-GGTCCGCACCACGCCYTCNGG-RTCNA-3') successfully produced a single PCR product of720 bp (data not shown). The PCR fragment was cloned,analyzed and designated as probe A (Figs. 1 and 2). ABLAST homology search of the polypeptide coded by thispartial nucleotide sequence revealed that human and mouseubiquitous p52shcA were the most identical proteins (68 and67%, respectively). Gene-specific primers were thus de-signed from probe A sequence to perform 5' and 3' RACE.Products of RACE were cloned, and three different clones ineach case were double-strand sequenced, which led to aconsensus sequence of each fragment. In the particular caseof 5'RACE, we obtained 3 clones of different size (936, 977and 1069 bp) because of extension in their 5'ends. We there-fore only considered the shortest sequence (936 bp), identicalin the three clones.

As shown in Fig. 1 (and deposited with GenBank datalibrary under accession number AY027793) , the resultantXenopus shc cDNA is 3979 nucleotides in length, including a1413-bp open-reading frame (Shcorf), preceded by a 109-bp5'untranslated region (5'UTR) and followed by a long2434-bp 3'UTR and a poly(A) tail. Indeed, from the two ATGin positions 110 and 125, only the first one likely representsthe initiation codon since the flanking nucleotides matchbetter the eukaryotic translation initiation consensus se-quence (A/GCCATGG; Kozak, 1986). The reconstitutedcDNA thus encodes a protein of 470 AA with a predictedmolecular weight of 52.3 kDa. When compared with itsDrosophila, fish or mammalian counterparts, this proteinshows an overall identity of 34 to 78%, the highest being withthe mammalian p52ShcA forms (Table 1). It shows a compa-rable modular organization which includes PTB, CH1 andSH2 domains. PTB and SH2, which are crucial for the inter-action of Shc with tyrosine kinase receptors are particularlywell conserved since they respectively share up to 94 or90-91% identity with mammalian ShcA. In the CH1 region,the identity falls to 52-55% with ShcA, 21-23% with ShcB orC or even 13% with drosophila Shc. Within this CH1 do-main, however, the consensus motifs HXYYN and YVNXwhich contain phosphorylable tyrosine residues are both

312 F. Chesnel et al. / Biology of the Cell 95 (2003) 311–320

conserved. Clearly, the cloned cDNA represents the Xenopuslaevis orthologue of mammalian ShcA and hereafter is re-ferred to as XlShc.

2.1. Characterization of Xl Shc mRNA and proteinexpression

Northern blot analyses using probe A (Fig. 2A) revealed asingle 4.2-kb mRNA expressed in postvitellogenic oocytes,

in embryos before (stage 4) and after midblastula transition(stage 10) as well as in adult liver (Fig. 2B). This RNA wasnot highly expressed since it was only detected from poly(A)+ RNA, not from total RNA. Of note, although we onlyloaded 1.2 µg oocyte poly(A)+ RNA versus 5 µg liverpoly(A)+ RNA, the band was darker in oocytes suggestingthat XlShc expression was much higher in oocytes.

In order to study XlShc protein expression as well as itsregulation during insulin-induced oocyte maturation, rabbit

Fig. 1. Nucleotide and predicted amino acid sequence of Xenopus Shc cDNA. Nucleotides and amino acids are respectively numbered on the left and the right.In bold characters are successively indicated the two putative initiation sites, the termination codon and the polyadenylation signal sequence. A 5'UTR in-frametermination codon is double-underlined. The putative PTB and SH2 domains are shown in shaded boxes. Within the CH1 domain, the phosphorylable tyrosineresidues are boxed. The sequence corresponding to probe A (with the SacI restriction site, used for subcloning, in bold italics) is underlined.

313F. Chesnel et al. / Biology of the Cell 95 (2003) 311–320

polyclonal antibodies were raised against a purified His6-tagged polypeptide, His6-XlShcp, containing the CH1 do-main and the NH2-terminal half of the SH2 domain (AA236-431; Fig. 1 and 2), and affinity-purified. In Western blotanalysis, these antibodies recognized a prominent bandaround 60 kDa (designated as p60Shc – Fig. 3A and B)accompanied sometimes with a band of much slighter inten-sity around 120 kDa in the Xenopus ovary extract (notshown). p60Shc was also detected in the testis (Fig.3A and B),in the heart and to a lesser extent in the liver (Fig. 3B).

Proteins with a different molecular mass were also specifi-cally revealed in the brain (around 40 and 42 kDa) and thetestis (54 and 66 kDa). These bands were specific to shc sincethey were not revealed in absence of our polyclonal primaryanti-Shc antibody (Fig.3A, first two lanes) and they disap-pear after blotting the membrane with primary antibody inpresence of 50 µg of the antigen His6-XlShcp in the incuba-tion medium (Fig. 3A, last three lanes). A same competitionexperiment allowed to evidence that the other bands stainedin heart, liver and muscle (*; Fig. 3B) were not specific to Shcand reflected aspecific binding of primary antibody or sec-ondary HRP-conjugated antibody (data not shown).

Finally, a p60Shc expression profile was established duringoogenesis and embryogenesis by Western blot (Fig. 3C): theprotein was detected in oocytes at least as soon as the stageIII of vitellogenesis and slowly accumulated up to the postvi-tellogenic stage. The protein level was then maintained dur-ing oocyte maturation and remained constant after fertiliza-tion at least up to the stage 28 of embryogenesis. The highestexpression level of p60Shc expression in the oocytes andembryos compared to the other blotted tissues therefore sug-gests a putative role of this protein during these two develop-mental processes.

2.2. Xl Shc implication in insulin-induced oocytematuration

As Liu et al. (1995) who failed to evidence insulin-induced phosphorylation of the other adaptor protein IRS, wehave not been able to show p60XlShc phosphorylation infully-grown oocytes upon insulin treatment (data not shown).We next investigated the implication of Grb2 and Sos duringinsulin-induced meiotic progression since both moleculesare known in mammalian cells to form an heterocomplexstimulating p21Ras after activation of IGF-1R. For this pur-pose, it was first necessary to assess Grb2 and Sos expressionin Xenopus oocytes.

We previously reported that inability of insulin to in vitroinduce meiotic progression before Xenopus oocytes reach thepostvitellogenic stage was partly due to a dysfunction ofcellular signaling upstream of ras activation (Chesnel et al.,1997). This may result from insufficient expression level(s)of Grb2 and/or Sos. To assess this hypothesis, oocyte extractswere analyzed by immunoblotting with anti-Grb2 and anti-Sos heterologous antibodies. Anti-Grb2 antibody stained aprotein with an apparent molecular weight of 28 kDa, very

Fig. 2. Identification and expression of XlShc transcripts. (A), positions ofthe sequence used in the Northern blot analysis (probe A) and of thesequence corresponding to the 6His-tagged fusion protein (designed to raisespecific antibodies against XlShc), relative to the human p66Shc amino acidsequence (or p52/p46, which are generated from alternate translationalinitiation sites). CH, collagen homologous domain; PTB, phosphotyrosinebinding domain; SH2, Src homology 2 domain. (B) Northern blot analysis.Total [T] or poly(A)+ [A+] RNA from Xenopus adult liver, oocytes orembryos were subjected to formaldehyde-agarose gel electrophoresis andtranferred to nylon membranes. Filters were hybridized with [32P]-labeledprobe (probeA), washed under stringent conditions and exposed to X-ray for3 weeks at -70°C. The amount (in µg) of RNA loaded per well is indicatedbelow each blot. Positions of molecular weight markers (in kilobases) areshown are indicated on both sides. P.I, prophase I; M.II, metaphase II; St.,stage.

Table 1Comparison of XlShc amino acid sequence with the corresponding sequences of Shc in other species (using ClustalX and Genedoc softwares)

ShcA ShcB ShcC Shchuman1 mouse2 human3 human4 mouse5 rat6 fugu7 droso.8

p52 p52 p54 p52 p55 p52 p70 p45tot. 78 78 46 49 49 49 38 34PTB 94 94 64 71 71 70 71 43CH1 55 52 21 22 22 23 21 13SH2 91 90 67 72 73 73 70 62

(Genbank accession numbers: 1, X68148; 2, U15784; 3, AAB46782; 4, D84361; 5, U46854; 6, AB001453; 7, AF163839; 8, U26445)

often revealed as a doublet in our experimental conditions(Fig. 4A; top panel). It likely corresponds to Xenopus Grb2as the expected size of the protein calculated from its se-quence (Genbank AJ223061) is at least 25kDa. On the otherhand, Sos migrated as a single 170-kDa band (Fig. 4A;bottom panel) as previously shown in Xenopus embryos(Ryan and Gillespie, 1994). This protein likely represents thefrog orthologue of mammalian p170Sos1.

Both proteins accumulated in oocytes from stage III tostage VI of vitellogenesis (Fig. 4A). Surprisingly, theiramounts became almost as abundant in meiotically incompe-tent (stage IV) as in meiotically competent (stage VI) oo-cytes.

To ascertain whether Grb2 and Sos associated in vivo inboth stage IV and stage VI oocytes, cell lysates were thenimmunoprecipitated with anti-Sos antibodies followed byimmunoblotting with anti-Grb2 antibodies (Fig. 4B). Grb2was indeed present in Sos precipitates in both stage IV andstage VI oocytes, indicating that this two molecules form acomplex in prophase-I arrested Xenopus oocytes.

The participation of Grb2 in insulin-induced oocyte matu-ration was then demonstrated by injecting in vitro-transcribed Grb2P49L RNA into oocytes 12 hours prior stimu-lation with saturating doses of insulin or progesterone (usedas a negative control since it is supposed not to require Rasactivation to induce oocyte maturation). This 12-hour prein-cubation was necessary for this dominant negative form ofGrb2 to be overexpressed (Fig. 4C, insert). This mutatedprotein dramatically inhibited insulin-induced meiotic pro-gression since GVB rates scored 13 hr after stimulationdecreased from 96.6% to 10% (Fig. 4C) whereasprogesterone-induced GVB remained unaffected.

Finally, to determine whether Shc and Grb2 associated invivo, postvitellogenic oocytes were exposed to insulin orprogesterone for 30 min and 2 hr and oocyte extracts wereprepared. Endogenous p60Shc was immunoprecipitated withaffinity-purified antibodies followed by immunoblottingwith anti-Grb2 antibodies. Results of these experiments arepresented in fig. 5. For similar amounts of immunoprecipi-tated p60Shc protein ascertained by immunoblotting Shc withthe same antibody (Fig. 5 top panel), the characteristic dou-blet indicative of Xenopus Grb2 was only present in the Shcprecipitates when oocytes were treated with 5 µM insulin(Fig. 5; lanes 3 and 5 – bottom panel). This was also observedwhen using crude anti-His6-XlShcp immune serum whereasno Grb2 was blotted when using preimmune serum (data notshown). This experiment thus indicates that insulin is able tospecifically induce p60Shc-Grb2 association in Xenopuspostvitellogenic oocytes.

3. Discussion

In mammals, two kinds of proximal IGF-1R intracellularsubstrates have been mainly described: the adapter proteinsIRS and Shc. Both appeared as good candidates for mediat-ing Xenopus oocyte maturation induced through IGF-1 re-

Fig. 3. Identification and expression of XlShc protein(s). Whole-cell lysateswere prepared from the indicated tissues (A, B) or from oocytes andembryos (C) and 50 µg total proteins were separated by 12.5% SDS-PAGE,transferred to nitrocellulose and immunoblotted with affinity-purified anti-XlShcp antibody. The primary antibody was then probed with horseradishperoxydase-linked antibody (HRP-Ig) and developed with Pierce chemilu-minescence reagent (A; B, top panel; C). After stripping as described by themanufacturer (Amersham), the blot was reprobed with a polyclonal anti-actin primary antibody (B, lower panel). To ascertain anti-shc specificstaining, membranes were probed with secondary antibody alone (A, twofirst lanes) or with anti-XlShcp antibody + 50 µg his-tagged XlShcp peptideadded in the incubation buffer (A, last 3 lanes).*, Aspecifically labelled bands. Positions of molecular mass markers (inkilodaltons) are indicated on the right. Ov., ovary; B, brain; T, testis; MII,metaphase II; st., stage.

ceptor activation. Two different IRS have been cloned so farfrom Xenopus oocytes and their overexpression, as well asthat of heterologous rat IRS-1, either promoted (Chuang etal., 1993; Yamamoto-Honda et al., 1996) or facilitatedinsulin-induced oocyte maturation (Liu et al., 1995; Ohan etal., 1998). A dominant-negative approach was also appliedbased on overexpression of IRS truncated forms (Ohan et al.,1998) or mutated IRS unable to activate PI(3)K or MAPKpathway (Yamamoto-Honda et al., 1996) into postvitello-genic oocytes. From these studies, it was concluded that IRSmediated the mitogenic effect of insulin, likely via mobiliz-ing PI(3)K rather than Grb2-Sos, which suggests the partici-pation of another adapter molecule to target Grb2-Sos to themembrane. In addition, the authors did not rule out thepossibility that overexpression of IRS domains or mutatedIRS may affect IRS-independent pathways by competingwith other adapter proteins for IGF-1R binding.

We therefore determined whether Xenopus oocytes doexpress the other proximal IGF1-R substrate, Shc. By per-forming classical RT-PCR followed by RACE-PCR, we ob-tained a 3979-bp clone, called XlShc, with an 1413-bp open-reading frame coding for a 470-aminoacid protein with apredicted molecular weight of 52.3 kDa. Three different 5'and 3' RACE-PCR clones were sequenced on both strands tocompensate for eventual misreading by the AdvanTaq™polymerase; the cDNA corresponding to the open readingframe was again cloned by RT-PCR from a new pool ofpostvitellogenic oocyte RNA, and sequenced so that one canbe totally confident with the XlShc cDNA sequence nowdeposited with GenBank data library under accession num-ber AY027793. This sequence unlikely corresponds to theentire full-length XlShc mRNA since we obtained 5'RACEfragments of different sizes and considered only the shortest,the sequence of which was common to all three clones.

Fig. 4. Expression and association of Grb2 ans Sos in oocytes throughoutvitellogenesis – Participation of Grb2 during oocyte maturation. (A) totalproteins (30 µg – vitellus excluded) of staged oocytes were separated by 15or 8% SDS-PAGE (for Grb2 and Sos immunodetection, respectively), trans-ferred to nitrocellulose, and immunoblotted with commercial anti-Grb2(upper panel) or anti-Sos antibodies (lower panel). The primary antibodieswere then probed with horseradish peroxydase-linked antibody and visuali-zed with chemiluminescence reagent (Amersham). Arrows show the speci-ficically recognized antigens. (B) protein lysates (50 stage IV or 30 stage VIeq. oocytes) were immunoprecipitated (IP) with anti-Sos antibodies. Theimmunoprecipitates were then separated by 8 or 15% SDS-PAGE andimmunoblotted (WB) either with anti-Grb2 (lower panel) or anti-Sos anti-bodies (upper panel), respectively. (C) Stage VI oocytes (50 in each group)were injected with H2O or in vitro-transcribed Grb2P49L RNA (20 ng).Twelve hours following microinjection, oocytes were separated into twogroups and treated with 3 µM progesterone (P) or 5 µM insulin (Ins). GVBDwas scored at the indicated times after addition of both hormones; inset,western blot analysis of Grb2 (endogenous or mutant P49L) expression incontrol oocytes (n.i.), and oocytes injected with H20 or Grb2P49L RNA (andfurther incubated for 12 hr) using anti-Grb2 antibody to ascertain that themutant protein is synthesized in oocytes by the time the hormones are addedin the culture medium. Results are representative of three independentexperiments in each case.

Fig. 5. In vivo association of p60Shc with Grb2 during insulin-inducedoocyte maturation. Postvitellogenic oocytes were treated either with 5 µMinsulin (I) or 3 µM progesterone (P) for the indicated time periods andprotein lysates (40 eq. oocytes) were immunoprecipitated (IP) with affinity-purified anti-XlShcp antibodies. The immunoprecipitates were then separa-ted by 12.5% SDS-PAGE and immunoblotted (WB) either with anti-Grb2(bottom panel) or anti-XlShcp antibodies (top panel). Positions of molecularmass markers (in kilodaltons) are shown to the right. Results are represen-tative of three independent experiments.

This cDNA thus codes for a protein closely related insequence to the mammalian ubiquitous p52ShcA isoform. Thetwo domains, PTB and SH2, known to cooperate in Shcbinding to the intracellular domain of autophosphorylatedtyrosine kinase receptors, were particularly well conserved,with the presence in the PTB of a critical residue, arginine178 (R175 in mammalian p52ShcA – Van der Geer et al.,1996). Although the CH1 domain is the less conserved, itcontains the three characteristic tyrosine residues found in allmammalian species and in fish with the sequences surround-ing Y242 and Y325 which fit the consensus motif for Grb2-SH2 domain recognition (ie, pYXNX).

The size of the XlShc mRNA obtained by Northern blotanalysis was ~4,2 kb in oocytes, embryos as well as in adultliver. This value was very consistent with the length of thereconstituted clone (3979 bp), which is likely underestimatedsince we could at least add the 133-nt 5'extension we ob-served in one of the three 5'RACE clones. It is in any caselonger than that found for the corresponding mammaliantranscripts encoding p52Shc (3 kb; Pelicci et al., 1992) onlybecause of a ~1-kb larger 3'UTR. Interestingly, although wedesigned a probe capable of detecting the spliced variant ofShcA transcript (coding for p66Shc) in addition to the primarytranscript, we could not detect its expression in tested cellsand tissues. The lesser sensitivity of the Northern blot tech-nique cannot account for this missing band since in every5'RACE clone sequenced, we systematically found an in-frame termination codon (position +77) which excludestranslation of a larger protein. These results thus confirm thatin Xenopus as in mammals, the spliced variant of the primaryShcA transcript, if any, is not ubiquitously expressed (Migli-accio et al., 1997). It would be of interest to search for itspresence particularly in testis, where a 66-kDa protein wasdetected with the anti-His6-XlShcp antibodies (Fig. 3A).

At the protein level, the affinity-purified antibodies (raisedagainst the CH1 domain and a part of the SH2 domain ofXlShc, and not against the entire protein to avoid cross-reactivity of the antibodies with Xenopus IRS since bothmolecules share similar motifs in the PTB domain) specifi-cally recognized a major 60-kDa protein essentially in ovaryand testis, in heart and to a lesser extent in liver. The addi-tional immunodetection of the 54- and 40-kDa proteins re-spectively in testis and brain extracts can be interpreted asfollows: The in-frame alternative initiation codon (AUG en-coding Met46) which generates the p44 Shc protein fromhuman and mouse p52ShcA transcripts is conserved in XlShctranscript (Met49 – Fig.1) and may be used preferentially intestis to synthesize the 54-kDa protein in addition to the otherisoforms. The 40-kDa protein identified in adult brain may bethe product of another Shc gene, ShcC (or N-Shc) as de-scribed in mammals (O’Bryan et al., 1996) where ShcAtranscripts are not expressed in the adult brain. Although theapparent molecular weight of the XlShc protein in oocytes ishigher than expected (60 vs 52,3 kDa), it is undoubtedlyencoded by the cDNA we cloned as the protein synthesizedin oocytes after injection of in vitro-transcribed XlShcorf

RNA did migrate at the same position (data not shown). Inaddition, using heterologous antiShc antibodies, Aoto et al.(1999) also identified a 58-kDa protein in Xenopus eggswhich was tyrosine phosphorylated following fertilization.Altogether, the high conservancy in Xenopus p60Shc of allmotifs important for mediating tyrosine kinase receptor func-tions and the preferential expression of this protein in oocytesthroughout oogenesis and in early embryos strongly impli-cate this molecule early in development.

In this report, the participation of Shc as well as thedownstream effectors, Grb2 and Sos, in insulin-induced oo-cyte maturation were next investigated. We first evidencedexpression and accumulation of Grb2 and Sos proteins inoocytes throughout vitellogenesis. To ascertain the role of theadapter protein Grb2 in mediating Ras-dependent insulin-induced maturation in postvitellogenic oocytes and insteadof using the SH2 domain of Grb2 that has been reported toexert a mild effect on oocyte maturation induced by insulin(Aroca et al., 1996), we chose to evaluate the effect ofoverexpressing the full-length Grb2 protein with a singlemutation (Proline 49 replaced by a Leucine) in the amino-terminal SH3 domain. Indeed, this mutant molecule effi-ciently inhibits either p21Ras-dependent induction of meso-derm by FGF in animal caps of Xenopus embryos (D.L. Shi,personal communication) or p21Ras activation by FGF inFGFR-expressing Xenopus oocytes (Browaeys-Poly et al.,2000). The large inhibition of insulin-induced maturationthat we observed was not due to a toxic effect of the overex-pressed mutant protein since maturation triggered by proges-terone was absolutely not perturbed in the same conditions.Together with the results of Chie et al. (1999) showing theparticipation of Sos during insulin-induced MAPK activationand oocyte maturation, our data thus demonstrate that Grb2-Sos recruitment following IGF1-R activation is required topromote oocyte maturation.

At first glance, the absence of p60XlShc phosphorylationin insulin-treated postvitellogenic oocytes (data not shown)does not favor a possible implication of this adapter proteinin insulin-induced oocyte maturation. This is neverthelessbalanced by our observation that Grb2 associated with en-dogenous p60Shc within 30 min after oocyte treatment withinsulin specifically, which implies phosphorylation of at leasttyrosine 325 of p60Shc (equivalent to Y317 in its mammaliancounterpart). Tyrosine phosphorylation of mammalian Shcisoforms is yet easily detectable in cells treated with growthfactors or cytokines, by immunoblotting with antiphosphoty-rosine antibodies but in most cases, phosphorylation is as-sayed after immunoprecipitating Shc from 10,000 to 1 mil-lion cells. When studying cell signaling in amphibianoocytes, it is technically difficult for many reasons to homo-geneize more than fifty cells per sample which greatly re-duces the amount of immunoprecipitated molecules of inter-est. This is likely the reason why Liu et al. (1995) also failedto demonstrate insulin-induced phosphorylation of the en-dogenous oocyte IRS-like molecule despite the existence of8 putative tyrosine phosphorylation sites versus 3 in Shc.

In conclusion, the results presented herein combined withthe data concerning Xenopus IRS (Liu et al., 1995;Yamamoto-Honda et al., 1996; Ohan et al., 1998) and Grb2(Cailliau et al., 2001) strongly suggest that activation ofIGF-1R in Xenopus postvitellogenic oocytes triggers MAPKand MPF activation by concomitantly recruiting two proxi-mal adapter proteins, p60Shc and xIRS-1 (and/or xIRS-u) thatstimulate the Ras signaling pathway (through Grb2-Sosbinding) and the PI(3)K pathway, respectively. Our presentresults also demonstrate that growing oocytes do expressp60Shc, Grb2 and Sos in sufficient amounts so that theirinability to resume meiosis in response to insulin ratherresults from a defect of the activated IGF-1R to addressShc-Grb2-Sos to the membrane and/or to activate the parallelIRS-dependent PI(3)kinase pathway.

4. Materials and Methods

4.1. Materials

Sexually mature Xenopus laevis females were reared inthe laboratory. Unless otherwise indicated, reagents werepurchased from Sigma (Saint Quentin Fallavier, France).

4.2. Oocyte and embryos treatments

Defolliculated oocytes were prepared and maintained inpotassium-free medium, MK/PVP, as previously described(Chesnel et al., 1997). Embryos were obtained by in vitrofertilization of laid eggs, dejellied in 1 x F1 pH 8.0 (31.25mM NaCl, 1.75 mM KCl, 1 mM CaCl2, 0.06 mM MgCl2,10 mM Hepes) containing 2% cysteine, and allowed to de-velop to the desired stage in 0.1 x F1 (Webb and Charbon-neau, 1987). Embryonic stages were determined according toNieuwkoop and Faber (1967).

In some experiments, stage VI oocytes were microinjectedwith approximately 20 ng capped RNA in a final volume of50 nl/oocyte. RNA were generated from Xlshcorf or Grb2

cDNA inserts in the expression vector pSP64TBX, using themMessage mMachine™ in vitro SP6 transcription kit (Am-bion – Clinisciences, Montrouge, France); Grb2

P49L was adominant negative form of Xenopus Grb2 (Browaeys-Poly etal., 2000) generously provided by Dr Shi. Fourty-eight hoursafter microinjection, oocytes were either treated with hor-mones or directly processed for Western blot analyses. Eachset of experiments was performed at least three times.

4.3. RNA isolation and reverse transcription – polymerasechain reaction (RT-PCR)

Oocyte and embryo RNA were extracted by small-scaletotal RNA preparation (Condie and Harland, 1987) whileliver RNA was prepared using Trizol® as recommended bythe manufacturer (Roche Diagnostics, Meylan, France).Poly(A)+ RNA was purified using the PolyATtract® mRNAisolation system (Promega,Charbonnieres, France). First

strand cDNA was synthesized from 10 µg total RNA using50 ng random hexamers and 200 units SuperScript reversetranscriptase (Roche Diagnostics). After a 1-hour incubationat 37°C, the RT product was directly used for cDNA ampli-fication by PCR with GoldStar DNA polymerase (Eurogen-tec, Seraing, Belgium) using either specific or consensus-degenerate hybrid oligonucleotide primers (“CODEHOP”primers; Rose et al., 1998). Following elution using standardprocedures, amplification products were ligated into thepGEM®-T vector for sequencing by the dideoxynucleotidechain termination method using the Sequenase version2.0 DNA sequencing kit (Amersham Biosciences, Orsay,France).

4.4. Rapid amplification of cDNA ends (RACE)

Stage VI oocyte poly(A)+ RNA (1 µg) was used for thefirst-strand cDNA synthesis, using primers and reagents fromthe SMART™ RACE cDNA Amplification Kit (Clontech –Ozyme, Saint Quentin, France), Powerscript™ MMLV re-verse transcriptase (Clontech) and following the manufactur-er’s protocol. The cDNA generated by this kit were directlyused in 5' and 3'-RACE PCR, without the need for second-strand synthesis and adaptor ligation. Sequencing DNA reac-tions were performed as described in the “PRISM* ReadyReaction Big Dye* Terminator cycle Sequencing Kit” (PEBiosystems, Courtaboeuf, France) using either universal orgene-specific primers. Fluorescent DNA fragments wereseparated and analysed with an automated ABI 310 se-quencer (PE Biosystems).

4.5. Northern blot analysis

Twenty µg total or 1-5 µg poly(A)+ RNA were electro-phoresed in denaturing conditions on 1.2% agarose – 6%formaldehyde gels, transferred to Nylon membranes (Hy-bond N, Amersham Biosciences), hybridized with the radio-labeled Shc probe (106 cpm/ml) overnight at 65°C aspreviously described (Church and Gilbert, 1984) and autora-diographed.

4.6. Expression and purification of recombinant Xenopuspartial shc (Xlshcp)

Induction of His6-XlShcp synthesis by isopropyl-b-D-thiogalactopyranoside (IPTG) and purification of the recom-binant His6-tagged protein on a Ni2+-NTA agarose columnwere done according to the manufacturer’s procedure(Qiagen, Courtaboeuf, France) . Protein concentration wasestimated using the bicinchoninic acid assay (Sigma) and thedialysate was aliquoted and stored at –70°C.

4.7. Production and purification of polyclonal antibodies

New Zealand rabbit was immunized seven times with100µg affinity-purified, recombinant His6-XlShcp. The im-mune serum was collected a week after the last immuniza-

tion. Purification of immunoglobulins specific to His6-XlShcp was performed as described elsewhere (Beranger etal., 1991).

4.8. Immunoprecipitation

Xenopus oocytes were lysed in RIPA buffer (50 mMHepes pH 7.5, 150 mM NaCl, 2.5 mM EDTA and 1% TritonX-100) containing 1 mM 4-(2-aminoethyl)-benzenesulfonylfluoride (AEBSF), 10 µg/ml leupeptin and aprotinin and1 mM sodium orthovanadate. After centrifugation at 15,000gfor 15 min at 4°C to remove insoluble material, the aqueoussupernatant was mixed with anti-Xlshcp antibodies (crudeimmune serum or affinity-purified antibodies: 10 or 20 µlrespectively) and incubated at 4°C for 2 hours. Affi-prepprotein A support beads (Bio-Rad, Marne la Coquette,France) were added and the resulting mixtures were rotatedovernight at 4°C. The beads were recovered by centrifuga-tion and washed six times with RIPA buffer. Proteins werefinally eluted from the beads with 2x sample buffer (100 mMTris-HCl, pH 7.5, 2 mM EDTA, 10%(v/v) 2-mercapto-ethanol, 2%(w/v) SDS, 0.1%(w/v) bromophenol blue,20%(v/v) glycerol), subjected to sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE), and im-munoblotted.

4.9. Western blot analysis

Xenopus oocytes, dejellied embryos or various tissueswere homogenised in ice-cold buffer (80 mMb-glycerophosphate, 20 mM EGTA, 15 mM MgCl2, 1 mMDTT, 20 mM Hepes, pH 7.5) containing 1 mM AEBSF,10 µg/ml leupeptin and aprotinin and 1 mM sodium ortho-vanadate. Proteins (from crude extracts or from immuno-precipitates) were separated on 12.5 or 15% SDS-polyacrylamide gels and electrotransferred to a 0.2-µmnitrocellulose membrane (Amersham Biosciences) as previ-ously described (Chesnel et al., 1997). Anti-Sos primaryantibodies were provided by Santa Cruz Biotechnology(Tebu, Le Perray en Yvelines, France), while anti-Grb2 andanti-actin antibodies were purchased from Transductionlaboratories and Sigma, respectively.

Acknowledgements

We are grateful to Dr. J.J. Eppig (The Jackson laboratory,Bar Harbor, Maine, USA) for critical reading of the manu-script, to Dr D.L. Shi for generously providing thepSP64TBX-xGrb2P49L construct, and to C. Ralliere for herassistance in sequencing the RACE-PCR products. Thiswork was supported by funds from the “Centre National de larecherche Scientifique”, the “Direction de la Recherche etdes Etudes Doctorales”, the Fondation Langlois and by agrant from the “Association pour la Recherche sur le Can-cer”.

References

Aoto, M., Sato, K., Takeba, S., Horiuchi, Y., Iwasaki, T., Tokmakov, A.A.,Fukami, Y., 1999. A 58-kDa Shc protein is present in Xenopus eggs andis phosphorylated on tyrosine residues upon egg activation. Biochem.Biophys. Res. Commun. 258, 265–270.

Aroca, P., Mahadevan, D., Santos, E., 1996. Functional interactions betweenisolated SH2 domains and insulin/Ras signaling pathways of Xenopusoocytes: opposite effects of the carboxy-and amino-terminal SH2domains of p85 PI 3-kinase. Oncogene 13, 1839–1846.

Baserga, R., Prisco, M., Hongo, A., 1999. IGFs and cell growth. In: Rosen-feld, R.G., Roberts Jr, C.T. (Eds.), The IGF System. Humana Press,Totowa, NJ, pp. 329–353.

Beranger, F., Tavitian, A., de Gunzburg, J., 1991. Post-translational process-ing and subcellular localization of the Ras-related Rap2 protein. Onco-gene 6, 1835–1842.

Biocca, S., Pierandrei-Amaldi, P., Cattaneo, A., 1993. Intracellular expres-sion of anti-p21ras single chain Fv fragments inhibits meiotic maturationof xenopus oocytes. Biochem. Biophys. Res. Commun. 197, 422–427.

Browaeys-Poly, E., Cailliau, K., Vilain, J.P., 2000. Signal transductionpathways triggered by fibroblast growth factor receptor 1 expressed inXenopus laevis oocytes after fibroblast growth factor 1 addition: role ofgrb2, phosphatidylinositol 3-kinase, src tyrosine kinase, and phospholi-pase c gamma. Eur. J. Biochem. 267, 6256–6263.

Butler, A.A., Yakar, S., Gewolb, I.H., Karas, M., Okubo, Y., LeRoith, D.,1998. Insulin-like growth factor-I receptor signal transduction: at theinterface between physiology and cell biology. Comp. Biochem. Physiol.121, 19–26.

Cailliau, K., Browaeys-Poly, E., Broutin-L’Hermite, I., Nioche, P., Gar-bay, C., Ducruix, A., Vilain, J.P., 2001. Grb2 promotes reinitiation ofmeiosis in Xenopus oocytes. Cell Signal. 13, 51–55.

Chie, L., Chen, J.M., Friedman, F.K., Chung, D.L., Amar, S., Michl, J.,Yamaizumi, Z., Brandt-Rauf, P.W., Pincus, M.R., 1999. Inhibition ofoncogenic and activated wild-type ras-p21 protein-induced oocyte matu-ration by peptides from the guanine-nucleotide exchange protein, SOS,identified from molecular dynamics calculations. Selective inhibition ofoncogenic ras-p21. J. Protein Chem. 18, 875–879.

Chesnel, F., Bonnec, G., Tardivel, A., Boujard, D., 1997. Comparativeeffects of insulin on the activation of the Raf/Mos-dependent MAPkinase cascade in vitellogenic versus postvitellogenic Xenopus oocytes.Dev. Biol. 188, 122–133.

Chuang, L.M., Myers Jr, M.G., Seidner, G.A., Birnbaum, M.J., White, M.F.,Kahn, C.R., 1993. Insulin receptor substrate 1 mediates insulin andinsulin-like growth factor I-stimulated maturation of Xenopus oocytes.Proc. Natl. Acad. Sci. USA 90, 5172–5175.

Chung, D., Amar, S., Glozman, A., Chen, J.M., Friedman, F.K., Robin-son, R., Monaco, R., Brandt-Rauf, P., Yamaizumi, Z., Pincus, M.R.,1997. Inhibition of oncogenic and activated wild-type ras-p21 protein-induced oocyte maturation by peptides from the ras-binding domain ofthe raf-p74 protein, identified from molecular dynamics calculations. J.Protein Chem. 16, 631–635.

Church, G.M., Gilbert, W., 1984. Genomic sequencing. Proc. NatlAcad. Sci.USA 81, 1991–1995.

Condie, B.G., Harland, R.M., 1987. Posterior expression of a homeoboxgene in early Xenopus embryos. Development 101, 93–105.

Deshpande, A.K., Kung, H.F., 1987. Insulin induction of Xenopus laevisoocyte maturation is inhibited by monoclonal antibody against p21 rasproteins. Mol. Cell. Biol. 7, 1285–1288.

Dupont, H., Blancq, M., 1999. Formation of complexes involving RasGAPand p190 RhoGAP during morphogenetic events of the gastrulation inXenopus. Eur. J. Biochem. 265, 530–538.

El-Etr, M., Schorderet-Slatkine, S., Baulieu, E.E., 1979. Meiotic maturationin Xenopus laevis oocytes initiated by insulin. Science 205, 1397–1399.

Groigno, L., Bonnec, G., Wolff, J., Joly, J., Boujard, D., 1996. Insulin-likegrowth factor I receptor messenger expression during oogenesis in Xeno-pus laevis. Endocrinol. 137, 3856–3863.

Janicot, M., Lane, M.D., 1989. Activation of glucose uptake by insulin andinsulin-like growth factor I in Xenopus oocytes. Proc. Natl Acad. Sci.USA 86, 2642–2646.

Korn, L.J., Siebel, C.W., McCormick, F., Roth, R.A., 1987. Ras p21 as apotential mediator of insulin action in Xenopus oocytes. Science 236,840–843.

Kozak, M., 1986. Point mutations define a sequence flanking the AUGinitiator codon that modulates translation by eukaryotic ribosomes. Cell44, 283–292.

Liu, X.J., Sorisky, A., Zhu, L., Pawson, T., 1995. Molecular cloning of anamphibian insulin receptor substrate 1-like cDNA and involvement ofphosphatidylinositol 3-kinase in insulin-induced Xenopus oocyte matu-ration. Mol. Cell. Biol. 15, 3563–3570.

Lopez-Hernandez, E., Santos, E., 1999. Oncogenic Ras-induced germinalvesicle breakdown is independent of phosphatidylinositol 3-kinase inXenopus oocytes. FEBS Lett. 451, 284–288.

Maller, J.L., Koontz, J.W., 1981. A study of the induction of cell division inamphibian oocytes by insulin. Dev. Biol. 85, 309–316.

Masui, Y., Clarke, H.J., 1979. Oocyte maturation. Int. Rev. Cytol. 57,185–282.

Migliaccio, E., Mele, S., Salcini, A.E., Pelicci, G., Lai, K.M., Superti-Furga, G., Pawson, T., Di Fiore, P.P., Lanfrancone, L., Pelicci, P.G.,1997. Opposite effects of the p52shc/p46shc and p66shc splicing iso-forms on the EGF receptor-MAP kinase-fos signalling pathway. EMBOJ. 16, 706–716.

Nieuwkoop, S.D., Faber, J., 1967. Normal table of Xenopus laevis (Daudin).North Holland Publishing Co., Amsterdam.

O’Bryan, J.P., Songyang, Z., Cantley, L., Der, C.J., Pawson, T., 1996. Amammalian adaptor protein with conserved Src homology 2 andphosphotyrosine-binding domains is related to Shc and is specificallyexpressed in the brain. Proc. Natl Acad. Sci. USA 93, 2729–2734.

Ohan, N., Bayaa, M., Kumar, P., Zhu, L., Liu, X.J., 1998. A novel insulinreceptor substrate protein, xIRS-u, potentiates insulin signaling: func-tional importance of its pleckstrin homology domain. Mol. Endocrinol.12, 1086–1098.

Opresko, L.K., Wiley, H.S., 1987. Receptor-mediated endocytosis in Xeno-pus oocytes. II. Evidence for two novel mechanisms of hormonal regu-lation. J. Biol. Chem. 262, 4116–4123.

Pelicci, G., Lanfrancone, L., Grignani, F., McGlade, J., Cavallo, F.,Forni, G., Nicoletti, I., Grignani, F., Pawson, T., Pelicci, P.G., 1992. Anovel transforming protein (SHC) with an SH2 domain is implicated inmitogenic signal transduction. Cell 70, 93–104.

Pera, E.M., Wessely, O., Li, S.Y., De Robertis, E.M., 2001. Neural and headinduction by insulin-like growth factor signals. Dev. Cell 1, 655–665.

Richard-Parpaillon, L., Heligon, C., Chesnel, F., Boujard, D., Philpott, A.,2002. The IGF pathway regulates head formation by inhibiting Wntsignaling in Xenopus. Dev. Biol. 244, 407–417.

Rime, H., Jessus, C., Ozon, R., 1995. Tyrosine phosphorylation of p34cdc2 isregulated by protein phosphatase 2A in growing immature Xenopusoocytes. Exp. Cell Res. 219, 29–38.

Rose, T.M., Schultz, E.R., Henikoff, J.G., Pietrokovski, S., McCal-lum, C.M., Henikoff, S., 1998. Consensus-degenerate hybrid oligonucle-otide primers for amplification of distantly related sequences. NucleicAcids Res. 26, 1628–1635.

Ryan, P.J., Gillespie, L.L., 1994. Phosphorylation of phospholipase Cgamma 1 and its association with the FGF receptor is developmentallyregulated and occurs during mesoderm induction in Xenopus laevis. DevBiol 1994 (166), 101–111.

Stith, B.J., Maller, J.L., 1984. The effect of insulin on intracellular pH andribosomal protein S6 phosphorylation in oocytes of Xenopus laevis. Dev.Biol. 102, 79–89.

Van der Geer, P., Wiley, S., Gish, G.D., Lai, V.K., Stephens, R., White, M.F.,Kaplan, D., Pawson, T., 1996. Identification of residues that controlspecific binding of the Shc phosphotyrosine-binding domain to phospho-tyrosine sites. Proc. Natl Acad. Sci. USA 93, 963–968.

Wallace, R.A., Misulovin, Z., 1978. Long-term growth and differentiation ofXenopus oocytes in a defined medium. Proc. Natl Acad. Sci. USA 75,5534–5538.

Webb, D.J., Charbonneau, M., 1987. Weak bases inhibit cleavage andembryogenesis in amphibians and echinoderms. Cell Differ. 20, 33–44.

Yamamoto-Honda, R., Honda, Z., Ueki, K., Tobe, K., Kaburagi, Y., Taka-hashi, Y., Tamemoto, H., Suzuki, T., Itoh, K., Akanuma, Y., Yazaki, Y.,Kadowaki, T., 1996. Mutant of insulin receptor substrate-1 incapable ofactivating phosphatidylinositol 3-kinase did not mediate insulin-stimulated maturation of Xenopus laevis oocytes. J. Biol. Chem. 271,28677–28681.

Zhu, L., Ohan, N., Agazie, Y., Cummings, C., Farah, S., Liu, X.J., 1998.Molecular cloning and characterization of Xenopus insulin-like growthfactor-1 receptor: its role in mediating insulin-induced Xenopus oocytematuration and expression during embryogenesis. Endocrinol. 139,949–954.

Molecular cloning and characterization of an adaptor protein Shc isoform from Xenopus laevis oocytes

Documents