The Development and Application of Computational Methods for the Prediction of G Protein-Coupled Receptor Structures

Citation

Bray, Jenelle Kiara (2010) The Development and Application of Computational Methods for the Prediction of G Protein-Coupled Receptor Structures. Dissertation (Ph.D.), California Institute of Technology. doi:10.7907/1655-ES74. https://resolver.caltech.edu/CaltechETD:etd-09032009-105125

Abstract

Computational methods for the prediction of G protein-coupled receptor (GPCR) structures were applied to serotonin receptors, and new methods were developed to predict an orphan GPCR structure. First, the MembStruk procedure was used to predict the structures of the serotonin 2b and 2c receptors. Ligand binding sites for agonists and antagonists were predicted for both receptors. In addition, the SAR data for a series of psilocybin analogs bound to serotonin 2c were predicted. There was good agreement with binding and mutagenesis experiments.

A new structure prediction procedure called SuperBiHelix was developed to predict an ensemble of low-lying structures. SuperBiHelix samples the tilt and sweep angles of the transmembrane helices along with the rotation of the helices along the helical axes. The procedure was validated on the β2-adrenergic receptor and A2A adenosine receptor crystal structures. This procedure was then used to predict the structure of GPR88, an orphan receptor. GPR88 has been identified as a novel target for psychiatric disorders. Three lipids were predicted to bind to GPR88. The head group of a lipid would bind to R113(3) and R116(3) at the extracellular side of the receptor. The lipid tail would bind in an aliphatic pocket in the TM2-TM3-TM6-TM7 region. The predicted bound complexes offer good suggestions for binding and mutagenesis experiments that could help validate the proposed structures.

Thesis Files