Background: Arterial thrombosis causes heart attacks and strokes and constitutes one of the leading causes of morbidity and mortality in the world and few therapies are available for its treatment. Thus, new therapeutic approaches in the prevention and treatment of arterial thrombosis are needed. Protein disulfide isomerase (PDI) has been shown to be expressed on vascular cells following injury and to be involved in regulating thrombus formation in vivo. Since inhibition of PDI prevents platelet accumulation and fibrin generation, it makes it a valuable target for the development of new antithrombotics. Rutin, a flavonol glycoside derivative of Quercetin, was previously described for displaying decent potency against PDI, and it inhibited the agonist-induced platelets aggregation in vivo, however its utility is limited by its low solubility and its off-target activity. Rutin was recently reported to bind specifically to the b’ domain of PDI affecting protein flexibility which results in the inhibition of its reductase activity. To investigate Rutin inhibitory mechanism we used docking and molecular dynamics simulation and we observed that Rutin binds to a specific hydrophobic pocket of the b’ domain which reduces PDI flexibility. Methods: In an attempt to identify more potent, soluble and specific PDI inhibitors, we established an in silico approach based on similarity search in Zinc Drug-like library composed of more than 17 million compounds satisfying Lipiniski’s rule of five. A KNIME workflow was established for selecting Rutin-similar compounds based on Tanimoto coefficients. Then, a virtual screening of selected compounds was performed using Autodock Vina on PDI target pocket. In order to select PDI specific probes, a counter-screen was run to eliminate hits binding Erp57 thioredoxin active site. Hits were then submitted to druglikeness prediction using Quantitative Estimate of Druglikeness (QED). A total of 5 compounds were selected and submitted to re-docking with Autodock Vina. Complexes were subject to Molecular Dynamics simulation using NAMD. Results and Discussion: a total of 4 compounds were shown to form stable complexes with PDI binding pocket and then could constitute promising candidates for lead optimization. In conclusion, our in silico approach lead to the identification of potential novel PDI inhibitors that may form suitable candidates for Arterial thrombosis drug discovery.
Machine Learning and Network Analysis of Molecular Dynamics Trajectories Reveal Two Chains of Red/Ox-specific Residue Interactions in Human Protein Disulfide Isomerase
