Evaluation of CendR peptides mined from protein databases as potential inhibitors of neuropilin-1 using an in-silico molecular modeling approach

Neuropilin 1 (NRP1) mediates lung branching angiogenesis and plays a vital role in cancer progression, especially in solid tumors. Its natural ligands, VEGFA and VEGFB, bind to the b1 binding pocket of NRP1 through hydrophilic c-terminus sites. Viral, bacterial and human proteins have evolved to interact with NRP1, facilitating entry into the cell, and therefore probing databases for hydrophilic CendR sequences could be a useful source of potential inhibitors of NRP1. Molecular docking was used to estimate the binding energies of 4185 naturally occurring CendR peptides to the b1 binding pocket of NRP1. The data predicts that proline, alanine, phenylalanine, leucine, tryptophan and tyrosine in the X positions of XXXXRXXR significantly strengthens the binding of the CendR peptides to the b1 binding pocket of NRP1 and that the presence of glutamic acid, lysine and serine tend to negatively impact binding. Three peptides predicted to have strong affinities and three peptides predicted to have weaker affinities were selected for molecular dynamic (MD) simulations. The data confirms that the interactions between the c-terminal arginine of these CendR peptides with several key residues, including Asp320, Thr316, Tyr297, and Tyr353, of the b1 binding site of NRP1 are particularly stable. During the MD simulations the three peptides predicted to be strong binders showed negative average free energies of interaction, while the three peptides predicted to bind more weakly showed positive free energies of interaction, providing validation of the docking results.