High-Throughput Screening to Enhance Permissiveness of Cardiac Endothelial Cells to Viral Vectors

Endothelial cells (ECs) are key players in the process of new blood vessel formation, also named angiogenesis, during both development and adulthood. Impairment of their function leads to many pathological conditions with either increased or inadequate vascularization. Thus, modulation of EC biology represents an attractive target in both basic and applied research. For instance, the major obstacle in inducing therapeutic angiogenesis is inefficient gene transfer in cardiac ECs. This poor permissiveness to genetic modification, attributed to their primary barrier function, is a limiting step in studies of pathological conditions with inadequate and increased vascularization. To study the role of a specific gene and its potential implications in angiogenesis, the most straightforward approach would be to either overexpress or silence it in ECs. We and others have tested multiple protocols for either transfection and transduction of primary ECs in vitro and in vivo, with minimal success. Vectors based on adeno-associated virus (AAV) and lentivirus (LV) stand as ideal tools for gene delivery. However, they do not work in cardiac ECs. To identify compounds able to increase transduction efficiency and identify the major determinants inhibiting genetic modification of cardiac ECs, we performed high-throughput screening (HTS) using a library of FDA-approved drugs in combination with both AAV and LV vectors. Cardiac ECs were isolated from C57BL/6 mice and plated in 384 wells plates. After 48 hours, the FDA drug library (> 1500 compounds) was added together with AAV6dsRed and LVGFP. At 72 hours after transduction, cells were fixed with 4% PFA and stained for the EC marker ERG. Expression of dsRed, GFP, and ERG was assessed by automated, high-content microscopy. After two rounds of HTS at different drug concentrations, we obtained a list of candidate drugs able that increased adult cardiac EC transduction by either AAV6 or LV over 2-fold. Among these, our best hit increasing cardiac EC transduction by AAV6 was vatalanib, a tyrosine kinase inhibitor acting on multiple VEGFRs. Vatalanib significantly enhanced permissiveness to AAV6 and also changed the shape of cardiac ECs, which became elongated and similar to mesenchymal cells. This suggested ongoing endothelial to mesenchymal transition (EndMT), verified by decreased CD31 and increased a-SMA expression levels. To validate our hypothesis, we induced EndMT by different methods, such as siRNA-mediated silencing of 2 EC-specific genes (CD31, ERG) in cardiac ECs and treatment of cardiac ECs with transforming growth factor-β2 (TGF-ß2). They all promoted loss of endothelial and acquisition of mesenchymal markers. Interestingly, AAV6 invariably transduced cells with the lowest expression of EC markers. Since cardiac ECs undergo transient EndMT early after myocardial infarction (MI), we injected AAV9-Cre into the heart of the mT/mG Cre-reporter mouse. After three days post-MI, cardiac ECs exhibiting EndMT features within the fibrotic zone were selectively transduced by AAV. Collectively, our results point to transient EndMT as a powerful mechanism to improve the efficiency of gene transfer in cardiac ECs, paving the way to innovative strategies for the induction of therapeutic angiogenesis in the heart.