Project 3. Rational GPCR compound library design

 

Key project members: Marijn Sanders, Jan Klomp, Sven van den Beld,  Stefan Verhoeven

 

Sponsor: Top Institute Pharma and Schering-Plough

 

Introduction

About 40% of all drugs on the market are targeting members of the GPCR protein family.

The GPCR family comprises a large variety of proteins sharing 7 transmembrane spanning helices(TM) as a common structural feature and transmits chemical signals into a wide array of different cell types. Natural ligands of GPCRs are extremely diverse, varying from lipids to protein, and as a consequence bind at different sites. Despite the large number of GPCRs and the great interest considering the substantial research investments of the scientific community and pharmaceutical industry, little detailed structure information is available. In this project structural features of different GPCRs, based on the alignment of the 7 TMs and the crystal structure of bovine rhodopsin and beta2adrenergic receptor are retrieved, and are subsequently translated into pharmacophoric properties for compound library design.

 

Project Goal

Goal of the project is to develop new methodologies to retrieve spatial information from protein structure and sequences alignments and translates this into pharmacophores which can be used to design low molecular weight compound libraries. The project is mainly focused on the G-protein coupled receptors family (GPCRs), however the bioinformatics technologies can be applied for other protein families as well. The final goal is to design innovative GPCR active compounds or compound libraries with different and more drug-like properties.

 

Approach

In the project structural features of different GPCRs, based on the alignment of the 7 TMs and the crystal structure of bovine rhodopsin and beta2adrenergic receptor are retrieved and translated into low-resolution pharmacophoric properties. From an analysis of non-olfactory rhodopsinlike GPCR sequences, 30 residue positions, located at the top of the helix bundle, were defined from which amino acids might interact with ligands and LMW compounds. Computational methods are developed to compare the micro-environments within the GPCR binding pocket on physico-chemical properties.

 

Results to date

  • New and less biased alignment of the TMs of 10 species is available, together with a method to add GPCRs from more species.
  • New clustering method is developed and the properties of this clustering are currently analyzed.
  • Geometric measure to predict interacting residues using low resolution information is developed and will be tested in the near future.
  • Methods to generate property profiles to deduce/propose multiple binding hypotheses (pharmacophores, key features), which can be used in virtual screening and GPCR focused rational library design. The CDD project team is working closely together with several (academic) groups. A key component is the synthesis of the designed compound libraries to validate and improve the computational methods (wet-dry cycle).