Forecasting a clear vision: How QSAR modeling catalyzes small molecule discovery in the field of corneal endothelial regeneration.

The corneal endothelium executes its crucial role in the maintenance of corneal transparency by regulating the corneal hydration levels. Dysfunction of the endothelial layer leads to corneal edema, i.e. bullous keratopathy, for which an endothelial transplantation is currently the only available treatment option. Nevertheless, the limited availability of donor tissue encourages the development of alternative treatment strategies. This study addresses a pharmacological-based alternative approach focusing on the comprehension and proof-of-concept validation of a hybrid small molecule screening approach. In this regard, a weighted voting-based quantitative structure-activity relationship (QSAR) model was developed for the in silico identification of small molecule building blocks promoting corneal endothelial regeneration. The established QSAR model was applied to the virtual Enamine® Hit Locator Library, composed of 460,160 compounds. Ultimately, the four most effective compounds predicted to stimulate corneal endothelial cell growth in silico, were selected for in vitro validation. All four selected compounds enhanced cell proliferation at nanomolar concentrations. Their effects were comparable to Y-27632, a ROCK inhibitor commonly used as a positive control to stimulate corneal endothelial cell proliferation and migration, supporting their potential as low-dose small molecule alternatives for corneal endothelial regeneration. These findings validate the predictive performance of our initial QSAR model and highlight the relevance for future screening of other hit locator library compounds, while feeding new biological data into the model for further refinement and accuracy improvement.