PLATELETS DRIVE FIBROGENESIS THROUGH INDUCTION OF ENDOTHELIAL-MESEMCHYMAL TRANSITION (ENDOMT) IN ENDOMETRIOSIS.

Xishi Liu,Dingmin Yan,Sun-Wei Guo
DOI: https://doi.org/10.1016/j.fertnstert.2019.07.327
IF: 6.7
2019-01-01
Fertility and Sterility
Abstract:One conundrum in endometriosis arising from a recent study is that, while endometriotic epithelial and stromal cells supposedly co-exist, the two cellular components seem to take independent developmental trajectory. This is due to the finding that, while cancer-associated somatic mutations were found to be enriched in the epithelial component, the stroma does not carry much. Given that endometriotic lesions are fundamentally wounds undergoing repeated tissue injury and repair that ultimately progress to fibrosis, we hypothesized that the stromal component of endometriotic lesions may recruit other cells and turn them into mesenchymal cells. One possible candidate cell is endothelial cell. Essentially, endothelial cells in lesions transdifferentiate into mesenchymal cells, likely induced by transforming growth factor β1 (TGF-β1) released by activated platelets, contributing further to lesional fibrosis. This study was undertaken to test this hypothesis. Laboratory study using human tissues, in vitro experimentation using an human umbilical vein endothelial cell line HUVEC. Immunofluorescent analysis of 30 each ovarian endometrioma (OE) and deep endometriosis (DE) tissue samples, using antibodies against CD31 and fibroblast-specific protein 1 (FSP-1), was performed. Immunohistochemistry analysis of CD31, FSP-1 and α-smooth muscle acting (α-SMA) was also performed. Masson trichrome staining was used to evaluate the extent of lesional fibrosis. In in vitro experiments, we evaluated morphological changes, gene and protein expression levels, migratory and invasive propensity, cellular contractility, and collagen production for HUVEC co-cultured with vehicle, activated platelets or thrombin only. We used A83-01, a TGF-β1 inhibitor, to neutralize TGF-β1. Endometriotic lesions clearly exhibited signs consistent with EndoMT, especially in OE lesions. Activated platelets, through the induction of TGF-β1 signaling pathway, induced EndoMT in HUVECs, resulting in increased migratory and invasive propensity, cellular contractility, and collagen production. Prolonged exposure of HUVECs to activated platelets induced increased expression of α-SMA, desmin and F-actin suggesting further transdifferentiation into smooth muscle-like cells. Neutralization of TGF-β1 abolished these changes. OE lesions had significantly higher staining levels of CD31, but lower α-SMA and FSP-1 staining, concomitant with lower lesional fibromuscular content than that of DE lesions. The staining levels of CD31 correlated negatively with the staining levels of α-SMA, as well as the extent of lesional fibrosis. EndoMT contributes to fibrogenesis in endometriosis. Because of EndoMT, the endometriotic stroma is constantly replenished by endothelial cells and other cells. These cells generally have much lower mutation rates than that of the endometriotic epithelium. Thus, we provide an explanation for the above mentioned conundrum of apparent independent developmental trajectories taken by endometriotic epithelium and stroma.
What problem does this paper attempt to address?