The Rho GTPases and their associated effector proteins are known to play diverse roles in the regulation of cytoskeletal remodelling, cellular adhesion and cell motility. The complex morphogenetic movements associated with gastrulation in Xenopus, including changes in the morphology and polarity of mesodermal cells, and their later mesolateral intercalation of these cells during convergent extension, are now known to be dependent upon RhoGTPase functions [5, 12, 14] (for review, see [35, 36]). Currently, however, there is little known regarding the identity or role of specific effector proteins that are utilized to convert changes in the GTP binding state of RhoGTPases into gastrulation-associated changes in cytoskeletal architecture, cell morphology, and cellular adhesion and migration.
The data presented here indicate a role for the Xenopus Cdc42 effector protein 2 (XCEP2) in gastrulation movements. XCEP2 is a member of the recently characterized CEP family of Rho GTPase effector proteins [29, 30], which include the previously characterized mouse protein, MSE55 . Effector proteins of the CEP family share a conserved expanded CRIB domain, which binds Rho GTPases, and two other highly conserved protein domains (CI and CII) [29, 30]. By analogy to other Rho GTPase effector proteins, it has been proposed that Cdc42 binding to the CRIB domain of CEP proteins leads to a conformational change that exposes the previously inaccessible CI or CII domains (reviewed in ). The exposed domains of the effector protein would then be free to interact with downstream components of the signaling/regulatory pathway. When overexpressed in cultured cells, members of this family of effector protein induce marked pseudopodial/ lamellipodial extensions, membrane ruffling, alterations in actin and vinculin organization, and a reduction of E-cadherin staining at adherens junctions [29, 30].
Consistent with a potential role for this class of proteins in embryonic morphogenesis, we have shown that XCEP2 expression is temporally regulated at gastrulation stages, when major modulations of cellular morphology, cytoskeletal organization, and cellular adhesion are occurring. mRNA for XCEP2 is present prior to mid-blastula transition, persists through mid gastrulation, and is strongly down-regulated by the time the blastopore closes and neurulation begins. This pattern would suggest that XCEP2 protein would be present through the period when active gastrulation movements are occurring. The diffuse spatial pattern of mRNA and, presumably, protein expression of XCEP2 may suggest that XCEP2 functions broadly in cells of the animal and equatorial regions. However, the observed broad spatial distribution of XCEP2 mRNA does not preclude the possibility that the XCEP2 protein may be functionally activated or inactivated at discreet times and locations during early embryonic development.
Furthermore, we show that antisense morpholino oligonucleotides capable of blocking translation of the XCEP2 message interfere with Xenopus gastrulation, delay the closure of the blastopore and inhibit embryonic elongation. The observed rescue with XCEP2 mRNA is strong evidence for the specificity of the antisense morpholino effect. These effects are not due to a loss of mesodermal induction, as brachyury and goosecoid expression do not change in response to antisense XCEP2 morpholinos. This is particularly relevant given recent reports demonstrating a direct link between brachyury expression and control of cellular migration .
The effects we report require relatively high, although not unprecedented, doses of morpholino antisense oligonucleotide. This dosage requirement may reflect the difficulty inherent in morpholino-mediated translational knockdown of maternally expressed genes. XCEP2 mRNA and (presumably) XCEP2 protein are present prior to midblastula transition. Given this situation, the timing morpholino induced protein downregulation is dependent both upon the effectiveness of translational blockade, and the half life of the protein in the cytoplasm. In this context, it may be essential to impose close to complete translational inhibition in order to reduce protein levels rapidly enough to affect early embryonic events, such as gastrulation.
Clearly, specific probes to assess endogenous XCEP2 protein expression will be necessary for fuller characterization of the role of this protein during gastrulation. For this reason, antibodies are currently being raised against the XCEP2 protein. These antibodies will be important in the characterization of the developmental time course of endogenous XCEP2 protein expression, assessment of the subcellular localization of the XCEP2 protein, isolation of potential XCEP2 binding partners, and in assessing and further optimizing the extent of protein down-regulation in morpholino injected embryos.
Currently, the specific mechanism by which XCEP2 exerts its role in gastrulation is unknown. However, our preliminary data suggest that XCEP2 may either contribute to a required "ground state" of cellular adhesion or play a role in modulations in the strength of cadherin-mediated cell-cell adhesion that are known to occur during gastrulation [22, 39, 40]. More detailed work will be necessary to clearly distinguish between these possibilities.
In embryos Wnt-mediated signals have been shown to activate Cdc42, a process that is required for normal gastrulation movements [7, 12]. In future work, it will be important to discern whether XCEP2 plays an important role in transducing these upstream signals into changes in cellular behaviour during gastrulation.
The known functional properties of the CEP class of effector proteins, and the characteristics of CRIB domain effector proteins in general, suggest some interesting possibilities relating to the control of cell adhesion during gastrulation. Consistent with the observed functions of the XCEP2 homologs in cultured cells, XCEP2 in embryos may impinge on regulatory circuits downstream of Cdc42 that control actin filament assembly, which in turn may affect diverse cellular processes, including assembly of adherens junctions. Alternatively, XCEP2 may more directly impinge upon cadherin functional activity, perhaps by influencing the association of IQGAP or other molecules with cadherin complexes. In future studies, it will also be important to establish whether embryonic activation of whether there are links between Wnt-mediated activation of Cdc42 and functional activation of XCEP2 and to characterize the mechanism(s) by which XCEP2 contributes to cell adhesion between cells of gastrulating embryos.