Epithelial–mesenchymal transition (EMT) is an important process of cell remodeling characterized by the gradual loss of the epithelial phenotype and progressive gain of a mesenchymal phenotype. EMT is not an all-or-nothing process, but instead a transition of epithelial to mesenchymal cells with intermediate cell states. Recently, EMT was described in endometriosis, and many EMT-specific pathways like Twist, Snail, Slug, Zinc finger E-box-binding homeobox 1/2 (ZEB1/2), E/N-cadherin, keratins, and claudins are involved. However, as pointed out in this review, a comparison of the eutopic endometrium of women with and without endometriosis yielded only subtle changes of these EMT markers. Furthermore, only very few alterations in cell–cell contacts could be found but without changes in the epithelial phenotype. This suggests only a partial EMT which is not a prerequisite for the detachment of endometrial cells and, thus, not critical for the first step(s) in the pathogenesis of endometriosis. In contrast, the majority of changes in the EMT-related marker expression were found in the ectopic endometrium, especially in the three endometriotic entities, ovarian, peritoneal, and deep infiltrating endometriosis (DIE), compared with the eutopic endometrium. In this review, we examine the most important EMT pathways described in endometriosis and propose that partial EMT might result from the interaction of endometrial implants with their surrounding microenvironment.

Claudins, as the major components of tight junctions, are crucial for epithelial cell-to-cell contacts. Recently, we showed that in endometriosis, the endometrial epithelial phenotype is highly conserved, with only minor alterations. For example, claudin-11 is strongly expressed; however, its localization in the endometriotic epithelial cells was impaired. In order to better understand the role of claudins in endometrial cell-to-cell contacts, we analyzed the tissue expression and localization of claudin-10 by immunohistochemistry analysis and two scoring systems. We used human tissue samples (n = 151) from the endometrium, endometriosis, and adenomyosis. We found a high abundance of claudin-10 in nearly all the endometrial (98%), endometriotic (98–99%), and adenomyotic (90–97%) glands, but no cycle-specific differences and no differences in the claudin-10 positive endometrial glands between cases with and without endometriosis. A significantly higher expression of claudin-10 was evident in the ectopic endometrium of deep-infiltrating (p < 0.01) and ovarian endometriosis (p < 0.001) and in adenomyosis in the cases with endometriosis (p ≤ 0.05). Interestingly, we observed a shift in claudin-10 from a predominant apical localization in the eutopic endometrium to a more pronounced basal/cytoplasmic localization in the ectopic endometria of all three endometriotic entities but not in adenomyosis. Significantly, despite the impaired endometriotic localization of claudin-10, the epithelial phenotype was retained. The significant differences in claudin-10 localization between the three endometriotic entities and adenomyosis, in conjunction with endometriosis, suggest that most of the aberrations occur after implantation and not before. The high similarity between the claudin-10 patterns in the eutopic endometrial and adenomyotic glands supports our recent conclusions that the endometrium is the main source of endometriosis and adenomyosis.

Purpose Claudins as the major components of tight junctions are important in maintaining cell–cell integrity and thus function as a barrier. Dysregulation of the claudins is often associated with loss of the epithelial phenotype, a process called epithelial–mesenchymal transition (EMT), which most often results in gain of migrative and invasive properties. However, the role of claudins in the endometrium or endometriosis has only rarely been examined. Methods In this study, we investigated localization of claudin-2 and claudin-3 in the eutopic and ectopic endometrium with immunohistochemistry. A detailed quantification with HSCORE was performed for claudin-2 and claudin-3 in endometrium without endometriosis and in cases with endometriosis compared to the three endometriotic entities: peritoneal, ovarian, and deep-infiltrating endometriosis. Results We found a preferential localization of both claudins in the glandular and the luminal epithelial cells in the endometrium with and without endometriosis. Quantification of localization of both claudins showed no differences in eutopic endometrium of control cases compared to cases with endometriosis. Furthermore, both claudins are localized highly similar in the ectopic compared to the eutopic endometrium, which is in clear contrast to previously published data for claudin-3. Conclusion From our results, we conclude that localization of claudin-2 and claudin-3 is highly stable in eutopic and ectopic endometrium without any loss of the epithelial phenotype and thus do not contribute to the pathogenesis of endometriosis.

No abstract