I-5: Molcular and Cellular Interactions in UterineReceptivity for Implantation

Background: Though plausible candidate adhesion systems have been identified, current knowledge of embryo-maternal attachment in human is limited by the inability to conduct well-controlled functional investigations. We have sought a viable medium-throughput model for the identification and functional assessment of molecular markers in the initial epithelial phases of implantation. An ideal model should bypass the scarcity of human embryos and well known difficulties in obtaining, establishing and maintaining confluent, normal, well-differentiated endometrial epithelium in vitro. Materials and Methods: Blastocysts flushed from day 4 pregnant superovulated mice were transferred to confluent human Ishikawa cell monolayers. Cells were untreated or steroid primed (E2 followed by MPA and E2) and characterised using a panel of differentiation markers. Embryo attachment was tracked by phase contrast microscopy, and weakly and stably attached embryos identified. Apically displayed cell surface glycoproteins were biotinylated using periodate/hydrazide, recovered by avidin affinity and analysed using a proteomics protocol.Results: After 48h of co-culture, 85% of blastocysts had attached loosely, but only 40% attached stably to the epithelial cell surface. In contrast, 95% of embryos attached stably to tissue culture plastic. This demonstrates that weak attachment of a majority of embryos is followed by stronger adhesion of a smaller proportion. Initial attachment is efficient either in the presence or absence of hormone, but steroid priming (E2 followed by MPA and E2) increased stable attachment from 40 to 70%. Initially, stable attachment occurred without disruption to the integrity of the epithelial monolayer, but clearance of surface MUC1 occurred beneath and adjacent to attachment sites. Later, lateral spreading of embryonic cells was accompanied by displacement of subjacent epithelial cells.Biotinylation was demonstrated to be highly vectorial, with label confined to the apical epithelial surface. The proteomic protocol led to the identification of approximately 30 species, about half of which are already recognised as endometrial cell surface glycoproteins. We tested endometrial tissue by immunofluorescence to confirm the presence of novel markers in luminal epithelium of midsecretory phase endometrium. Surprisingly, adhesion molecules present predominantly in the lateral membrane domain are also detectable in smaller amounts at the apical surface. siRNA knock-down reduces expression, but this leads to impairment of lateral cell adhesion and epithelial monolayer integrity, with implications for epithelial penetration by the implanting embryo.Conclusion: The model shows sufficient resemblance to key features of human and mouse implantation to be useful as a first line of discovery for cellular and molecular hypothesis building. Based on the observations, implantation in vivo may arrest when embryos fail to progress from initial to stable attachment. Novel candidate adhesion systems of possible importance for embryo-maternal interaction in vivo are being test