Molecular modelling studies on G protein-coupled receptors: from sequence to structure?

In pharmaceutical research G protein-coupled receptors (GPCR) emerged as a superfamily of prominent drug targets. Extensive protein sequence analyses on GPCRs revealed a common protein topology consisting of a membrane-spanning seven-helix bundle, which is believed to accommodate the binding site for low-molecular weight ligands. Enormous efforts are undertaken to generate GPCR structural models by means of molecular modelling, since these have already been shown to aid the process of lead structure finding and optimisation in that they provide atomistic models for structure-based drug design approaches. One of the most critical steps in modelling the transmembrane domains of GPCRs is the assignment of the putative transmembrane sequence stretches from multiple protein sequence alignment analyses. This study focuses on the comparative evaluation of protein sequence analysis tools, such as periodicity analyses, multiple sequence analyses or directional helix descriptors, especially developed for modelling the 7TM domains of GPCRs. In this context we will demonstrate that from application of different methods contradictory results can be obtained for the identification of the putative transmembrane sequence stretches of peptide-binding GPCRs, as exemplified with a comprehensive protein sequence analysis study based on the most prominent members of that receptor class (angiotensin II, CCK/gastrin, interleukin 8, endothelin, etc.).