Abstract

Atherosclerosis and vascular injury are characterized by neointima formation caused by the aberrant accumulation and proliferation of vascular smooth muscle cells (VSMCs) within the vessel wall. Understanding how to control VSMCs would advance the effort to treat vascular disease. However, the response to treatments aimed at VSMCs is often different among patients with the same disease condition, suggesting patient-specific heterogeneity in VSMCs. Here, we present an experimental and computational method called HETEROID (Heterogeneous Spheroid), which examines the heterogeneity of the responses to drug treatments at the single-spheroid level by combining a VSMC spheroid model and machine learning (ML) analysis. First, we established a VSMC spheroid model that mimics neointima formation induced by atherosclerosis and vascular injury. We found that FAK-Rac/Rho, but not Cdc42, pathways regulate the VSMC spheroid formation through N-cadherin. Then, to identify the morphological subpopulations of drug-perturbed spheroids, we used an ML framework that combines deep learning-based spheroid segmentation and morphological clustering analysis. Our ML approach reveals that FAK, Rac, Rho, and Cdc42 inhibitors differentially affect the spheroid morphology, suggesting there exist multiple distinct pathways governing VSMC spheroid formation. Overall, our HETEROID pipeline enables detailed quantitative characterization of morphological changes in neointima formation, that occurs in vivo, by single-spheroid analysis of various drug treatments.