The cell adhesion molecule E-cadherin mediates the compaction process of mouse preimplantation embryos and is important for the maintenance and function of epithelial cell layers. To determine precisely the role of E-cadherin in epithelial biogenesis we monitored the developmental potential of embryos homozygously negative for E-cadherin that were derived from E-cadherin heterozygous transgenic mice. The homozygous negative embryos died around the time of implantation, although they did undergo compaction like their littermate controls, largely due to the presence of residual maternal E-cadherin. At the blastocyst stage, E-cadherin-negative embryos failed to form a trophectodermal epithelium or a biastocyst cavity. These results demonstrate the pivotal role of E-cadherin in one of the most basic morphogenetic events in the development of multicellular organisms, the biogenesis of an epithelium.Mouse preimplantation embryos represent a unique model for studying the biogenesis of an epithelium from a nonpolarized cell, the fertilized egg (1, 2). During preimplantation development, loosely attached blastomeres maximize their cell-cell contacts and form a compact morula as a prerequisite for the generation of two distinct cell types, those of the trophectoderm epithelium and those of the inner cell mass (ICM). The first morphological changes in spherical blastomeres occur at the 8-cell stage and result in cytocortical asymmetry, including apical membrane microvilli and polar-

Vertebrate beta-catenin and Drosophila Armadillo share structural similarities suggesting that beta-catenin, like Armadillo, has a developmental signaling function. Both proteins are present as components of cell adherens junctions, but accumulate in the cytoplasm upon Wingless/Wnt signaling. beta-Catenin has axis-inducing properties like Wnt when injected into Xenopus blastomeres, providing evidence for participation of beta-catenin in the Wnt-pathway, but until now no downstream targets for beta-catenin have been identified. Here we demonstrate that beta-catenin binds to the HMG-type transcription factor lymphoid enhancer factor-1 (LEF-1), resulting in a nuclear translocation of beta-catenin both in cultured mouse cells and after ectopic expression of LEF-1 in two-cell mouse embryos. LEF-1/beta-catenin complexes bind to the promoter region of the E-cadherin gene in vitro, suggesting that this interaction could regulate E-cadherin transcription. As shown for beta-catenin, ectopic expression of LEF-1 in Xenopus embryos caused duplication of the body axis, indicating a regulatory role for a LEF-1-like molecule in dorsal mesoderm formation.

Initial cell lineages that presage the inner cell mass and extra-embryonic trophectoderm are established when eight blastomeres compact to form polarized morulae in preimplantation mouse development. FILIA has been identified as a binding partner to MATER (maternal antigen that embryos require; also known as NLRP5), which is encoded by a maternal effect gene. Products of each gene are detected in growing oocytes and, although transcripts are degraded before fertilization, the cognate proteins persist in early blastocysts. The two proteins co-localize to the cytocortex of ovulated eggs, where the stability of FILIA is dependent on the presence of MATER. After fertilization, FILIA-MATER complexes become asymmetrically restricted in the apical cytocortex of twocell embryos due to their absence in regions of cell-cell contact. This asymmetry is reversible upon disaggregation of blastomeres of the two-and four-cell embryo. Each protein persists in cells of the preimplantation embryo, but the continuous cell-cell contact of 'inner' cells of the morulae seemingly precludes formation of the subcortical FILIA-MATER complex and results in cell populations that are marked by its presence ('outer') or absence ('inner'). Thus, the FILIA-MATER complex provides a molecular marker of embryonic cell lineages, but it remains to be determined if the molecular asymmetry established after the first cell division plays a role in cell fate determinations in the early mouse embryo. If so, the plasticity of the FILIA-MATER complex localization may reflect the regulative nature of preimplantation mouse development.

Trophectoderm epithelium formation, the first visible differentiation process during mouse embryonic development, is affected in embryos lacking the cell adhesion molecule E-cadherin. Here we analyze the developmental potential of such E-cadherin-negative embryos, focusing on the organization of cell junctions and the cytoskeleton. To do this we used antibodies directed against alpha-, beta-, or gamma-(plakoglobin)-catenin and junctional and cytoskeletal proteins including ZO-1 and occludin (tight junctions), desmoglein1 (desmosomes), connexin43 (gap junctions), and EndoA (cytokeratin intermediate filaments). Membrane localization of alpha- and beta-catenin, and ZO-1, as well as cortical actin filament organization were abnormal in E-cadherin-negative embryos, and the expression levels of alpha- and beta-catenin were dramatically reduced, all suggesting a regulatory role for E-cadherin in forming the cadherin-catenin complex. In contrast, the membrane localization of plakoglobin, occludin, desmoglein1, connexin43, and cytokeratin filaments appeared unaltered. The unusual morphogenesis in E-cadherin-negative embryos apparently reflects defects in the molecular architecture of a supermolecular assembly involving zonulae adherens, tight junctions, and cortical actin filament organization, although the individual structures still appeared normal in electron microscopical analysis.