The role of the osteopontin–integrin αvβ3 interaction at implantation: functional analysis using three different in vitro models

STUDY QUESTION Does the interaction between integrin and its ligand osteopontin (OPN) mediate embryonic attachment to endometrial epithelium at implantation? SUMMARY ANSWER OPN of epithelial origin binds the receptor integrin αvβ3 at the maternal surface to support adhesion during the early stages of implantation. WHAT IS KNOWN ALREADY Integrin αvβ3 and OPN are both present in the endometrial luminal epithelium in the mid-secretory phase. STUDY DESIGN, SIZE, DURATION Microscopy of attachment sites of blastocysts (mouse, n = 151, human, n = 8) and OPN- or BSA-coated beads (n = 488) interacting with Ishikawa cell monolayers at 24 and 48 h. Levels of epithelial OPN or integrin αvβ3 were altered by siRNA-mediated targeting and the results compared with non-targeting siRNA or mock-transfected controls. PARTICIPANTS/MATERIALS, SETTING, METHODS In vitro modelling of early implantation with human endometrial cells (Ishikawa) and mouse or human embryos or ligand-coated beads. Immunolocalization of antigen around attached embryos was measured by image analysis with multiple repeats (n > 3), allowing a gradient of relative intensity to be detected. Attachment was quantified using a stability scale and protein expression documented by indirect immunofluorescence. Protein associations were probed by pulldown assays. MAIN RESULTS AND THE ROLE OF CHANCE Integrin and OPN levels were increased in epithelial cells near to attached embryos. The pulldown assay confirmed OPN-integrin αvβ3 binding (n > 3). Decreased attachment stability of mouse embryos observed after siRNA knock-down of integrin αvβ3 or OPN itself, or OPN-coated beads after knock-down of integrin αvβ3, was tested for significance using Kruskal–Wallis with Dunn's post hoc tests. LIMITATIONS, REASONS FOR CAUTION In vitro model. Attachment data using human embryos is limited by embryo availability. Mouse embryo attachment to human cells involves a species crossover so must be interpreted with caution. Ligand-coated beads allow specific molecular interactions mediating attachment to be probed, but obviously lack the adhesion and signaling repertoire of a live embryo. WIDER IMPLICATIONS OF THE FINDINGS Some of the literature identifies reduced integrin αvβ3 expression in infertile endometrium; these findings predict that embryo attachment stability will be reduced in vivo if integrin levels are low. We suggest that the robustness of the initial attachment of the embryo affects its ability to progress to the post-epithelial phase of implantation; some poorly attached embryos will be lost. STUDY FUNDING/COMPETING INTEREST(S) No external funds were used for this study, which was supported by funds from the Universities of Manchester and Oxford. None of the authors has any conflict of interest to declare. TRIAL REGISTRATION NUMBER N/A