Possible Inhibitors of Endothelial-to-Mesenchymal Transition Targeting Endocardial Fibroelastosis

Objectives: The triggers for endothelial-to-mesenchymal transition (EndMT), the underlying mechanism of endocardial fibroelastosis (EFE), are still unknown. We used in vitro endothelial cell culture and ex vivo isolated heart models to investigate possible triggers and the potential use as targets to inhibit EFE.

Methods: Endothelial cells in culture (n = 3/group) were exposed to 10% uniaxial static stretch for 8 hours (stretch). Additionally, one group was stretched and either BMP7 (stretch + inhibitor 1) or SB431542 (stretch + inhibitor 2) was added on a gelatin-based drug carrier (drug delivery). Cells incubated with media only without stretch served as controls (control). Immature and mature isolated perfused rat hearts (n = 7/group) were stretched to 30% of their original length for 3 hours (stretch) and either treated with BMP7 (stretch + inhibitor-1) or SB431542 (stretch + inhibitor-2) through a localized drug delivery system onto the endocardial surface of the LV or perfused only (control). Histological analysis by immunofluorescent staining for VE-cadherin/CD-31 (endothelial marker), αSMA (mesenchymal marker), and SLUG/SNAIL (EndMT transcription factors) was performed. Double staining with both endothelial and mesenchymal markers or nuclei staining positive for SLUG/SNAIL was indicative of EndMT. Statistical analysis was performed with analysis of variance and Bonferroni post hoc analysis. Data are expressed as mean ± SEM.

Results: The results showed significantly more double-stained endothelial cells in the stretch groups in culture (46 ± 13% of total cell count) and in immature whole hearts (22 ± 4% of total cell count) compared with controls (cells: 7 ± 2%; heart: 6 ± 2; p < 0.05 compared with stretch). Among the mature hearts, there was no significant difference between the stretch (15.2% ± 9.1) and control (24.1% ± 2.9) groups (p > 0.05). In the stretch + inhibitor groups, significantly less double-stained cells were observed compared with stretch and even had similar counts as the control groups (cells: BMP7 = 7 ± 2%; heart: BMP7 = 2.9 ± 0.1%, SB431542 = 5.2 ± 1.3%, p < 0.001 compared with stretch).

Conclusion: This study could show that mechanical strain imposed on immature endocardium induces EndMT. Using BMP7 and SB431542 as inhibitors for EndMT together with positive staining for SLUG/SNAIL, points toward a TGF-β mediated mechanism. Localized inhibition of EndMT was demonstrated by administration of BMP7 and SB431542 pushing for prenatal and local interventions to inhibit EFE.