Targeting the collagen-von Willebrand factor-platelet receptor binding axis by rational drug design.

Korte inhoud:Targeting Protein-Protein Interactions (PPIs) has become more and more the focus of drug design projects during the last decade. This type of interactions differs from protein-ligand interactions by a rather flat, hydrophobic, and often discontinuous interaction surface, and by the absence of well-defined binding pockets, characteristics that make the development of small molecule inhibitors very difficult. Nevertheless, successful attempts to inhibit protein-protein interactions with Small Molecule Protein-Protein Interaction Inhibitors (SMPPIIs) have been reported, and a few of these compounds have already found their way into clinical trials.This thesis focuses on the molecular modelling and rational drug design as part of an interdisciplinary knowledge platform, which was set up with the aim to offer solutions to any protein-protein interaction problem with sufficient structural information available. The modelling work encompassed the characterization and analysis of the targeted PPIs, the creation of structure-based pharmacophore queries, virtual screening of commercially available small molecule databases, molecular docking of hits retrieved from virtual screening into putative binding pockets, and similarity searching for analogues of compounds that showed in vitro activity. During this research, we aimed to find lead compounds for SMPPIIs targeting PPIs that occur in the collagen-VWF-GPIbα binding axis, responsible for the onset of thrombosis at sites of high shear stress. Under these conditions, the VWF-A3 domain interacts with collagen exposed in the subendothelial matrix after injury. Once immobilized, the multimeric VWF unfolds, revealing a binding site for the platelet receptor GPIbα on domain VWF-A1, and binds the platelet. This interaction happens fast but is still reversible. However, it will cause platelets to slow down and roll over the VWF-coated artery surface, thereby causing collagen platelet receptors GPVI and integrin α2ß1 to establish a firm adhesion to the site of injury. The already adhered platelets will also activate circulating ones, which results in the recruitment of more and more platelets and eventually the formation of a haemostatic plug or even a pathological thrombus. For the collagen – VWF-A3 interaction, we started from the crystal structure of VWF-A3 complexed with 82D6A3, a monoclonal antibody that inhibits VWF-A3 binding to collagen by blocking its collagen binding site. After a detailed analysis of the interaction surface, pharmacophore queries were created and used in virtual screening of small molecule databases. Subsequent docking of pharmacophore hits and in vitro testing of selected compounds led to the identification of five small molecules (BML1022, BML1023, BML1058, BML1061, and BML1062) that were able to disrupt the collagen – VWF-A3 interaction, however only at concentrations in the mM range. The VWF-A1 – GPIbα interaction was targeted by first detecting putative small molecule binding pockets on the surfaces of the interaction partners by the aid of computational techniques such as Molecular Dynamics (MD) and site finding algorithms. Subsequently, pharmacophore screening and molecular docking was also applied for these targets. Surprisingly, we found a small GPIbα-binding molecule (G6) that activated binding of GPIbα to VWF-A1. Similarity searching identified three more compounds with the same activating property. In vitro tests confirmed the activating effect of G6, as well as its ability to bind to GPIbα. Structural information from the complex between the inhibitory small helical peptide OS1 and GPIbα was also applied for the creation of pharmacophore models. This approach led to the characterization of the small inhibiting molecule G7.

Jaar van publicatie:2012

Toegankelijkheid:Closed