A two amino acid switch in the anti-angiogenic VEGF165b isoform creates a novel agent for therapeutic angiogenesis in peripheral artery disease

Abstract: Currently, no medical therapies can improve blood flow to ~12 million PAD patients (in the US alone). VEGF-A, a potent angiogenic growth factor was tested unsuccessfully in PAD clinical trials. We recently showed that failure to account for alternatively spliced anti-angiogenic VEGF-A (VEGF165b) isoform expression and function is one of the contributing factors behind VEGF-A clinical trial failure. Alternate splicing in exon-8 C- terminus of VEGF-A isoform results in the formation of pro-angiogenic VEGF165a (V165a-WT) and anti-angiogenic VEGF165b (V165b-WT) isoforms. The only difference between these 2 isoforms is a 6 amino acid shift from ‘CDKPRR’ in V165a-WT isoforms to ‘SLTRKD’ in V165b-WT isoforms. We have recently shown that 1) the fraction of V165b-WT is 2.5X higher than V165a-WT in total VEGF-A in human PAD muscle biopsies compared to controls and 2) in endothelial cells (ECs), V165b-WT blocked V165a-WT induced R1 activation even when present at 10X lower levels than V165a-WT. The net consequence of V165b-WT being 10X more potent than V165a-WT and being 2.5X more abundant than V165a-WT is a 25 fold functional molar excess of anti-angiogenic vs. pro-angiogenic VEGF isoforms in ischemic muscle. V165b-WT inhibition using a monoclonal antibody allowed V165a-WT to bind to R1 and activate novel VEGFR1 (R1) signaling pathways in ischemic ECs and macrophages that promoted perfusion in preclinical PAD models. Hence, our central hypothesis states that ‘displacement of R1 bound V165b-WT is necessary to allow ligand-induced R1-autophosphorylation and downstream signaling to enhance perfusion recovery in PAD’. Molecular processes that regulate R1 silencing ability of V165b-WT are not yet clear. Key residue alterations between V165a-WT and V165b-WT are the replacement of highly positively charged arginine residues in V165a-WT (CDKPRR) with neutral lysine-aspartic acid acids in V165b-WT (SLTRKD). We hypothesized that due to a net neutral charge conferred by ‘KD’ residues, V165b-WT binding cannot induce a strong internal rotation in the intracellular domain of R1 that is necessary to dimerize, autophosphorylate, and activate downstream signaling. To test our hypothesis, we switched the ‘KD’ residues in the C-terminus of V165b-WT to ‘RR’ (V165bKD→RR) and examined V165bKD→RR induced changes in R1 activation in ischemic ECs in vitro. Our preliminary data showed that V165bKD→RR induced R1 activation even in conditions where V165b-WT is induced, while V165a-WT failed to induce R1 activation. Furthermore, V165bKD→RR induced ischemic EC angiogenic potential and survival significantly higher compared to V165a-WT indicating a potential therapeutic for PAD. Based on these data, In Aim-1, we will determine the molecular processes (including binding affinities, structural changes, and receptor dimerization processes) by which V165b-WT and V165bKD→RR regulate R1 activation in vitro. In Aim-2, we will determine the cell- specific R1 signaling induced by V165bKD→RR to regulate EC and macrophage phenotypes in vitro. In Aim-3, we will use VEGF-A deficient mice, type-2 diabetic mice and EC-specific R1 deficient mice in preclinical PAD models to establish whether the translational potential of V165bKD→RR is R1 dependent.