Understanding the molecular mechanism of ligand binding to GPCRs

Strategic Research Priority: Industrial biotechnology and bioenergy
The G Protein Coupled Receptor (GPCR) family are an important drug target, which can be attributed to their ubiquitous nature in most cell types; 40 percent of currently marketed pharmaceuticals target a GPCR. Despite this sizable market share, 90 percent of GPCRs have no pharmaceutical that target them. This is part due to an inability of drug discovery to produce ligands that can bind in vivo with high efficacy. This problem may be due to the simplification of binding in drug discovery; binding affinity was first considered in terms of free energy alone, Kd. It has come to surface that kinetics of binding also have an effect on binding in vivo. The inverse value of the rate of unbinding is known as residence time, which now is also considered in the drug discovery progress. Rationalisation of residence time, also known as SKR (Structural Kinetic Relationship), has proven to be very complex and time consuming. This is likely due to the interlinked nature of binding affinity and binding kinetics. Due to the complexity and timescales of SKR studies, they are not suited to drug discovery. Computational methods have been used to predict ligand binding free energy and recently the rate of binding, kon, on a timescale suitable for drug discovery, however no workflows have been developed which can predict residence time or the rate of unbinding, koff. This project will uncover the molecular mechanism of binding by investigating and comparing different computational methods that can predict binding free energy and/or residence time. Steered Molecular Dynamics (SMD) is one of those methodologies that will be used to predict those binding parameters. Methodologies based on Umbrella SMD and adaptive biasing force will also be developed. These workflows will be tested in the example case of the adenosine receptors, a subclass of the GPCRs. This subfamily was chosen as there is a wealth of kinetic and free energy binding values for its ligands. These experimental binding values will be summarised in a database. In order to increase the size of that database, radioligand binding assays will be performed on a range of adenosine receptor ligands. These values will be used as both training and test data for the computational workflows. Successful development of computational workflows that can predict both residence time and free energy of binding will be of great help to drug discovery as well as personalised medicine, which currently can only conceivably be achieved using in-silico drug testing.