G-protein-coupled receptors (GPCRs) constitute the largest family of heptahelical transmembrane proteins. Essential for signal transduction across cell membranes they affect a broad variety of physiological processes. Rhodopsin is a prototypical GPCR which is responsible for scotopic vision in vertebrate species. It is also the only GPCR for which the crystal structure has been determined providing in atomic detail the structure of its ligand, 11-/cis/-retinal, and how it fits into the protein. Elucidation of the mechanism, by which ligand excitation is transformed into rhodopsin activation and coupling to the G-protein may become exemplary for the structure and function of other members of the GPCR group.Despite the information gained from the resting state of rhodopsin many of the ensuing intermediates of the visual cascade remain an enigma. For bathorhodopsin which is the intermediate directly following photoexcitation a crystal structure has been published, also for lumirhodopsin and metarhodopsin I.

The aim of the project is twofold: first, verify, and if necessary improve the reported structures, especially with respect to the ligand conformation, by employing high-quality ab initio methods supported, where necessary and possible, by classical force field methods; second, develop, based on the available structures and indirect evidence, structural models for the later intermediates.

Rhodopsin. The negative twist of the 11-cis-chromophore which had been predicted by us and was experimentally established by all reported X-ray structures so far has been confirmed by elaborate QM/MM modelling studies. Origin of this twist is the configuration at the covalent binding site of the chromophore to the apoprotein: the Lys296 side chain as well as the hydrogen-bond to wat2b exert a clockwise torque on the chromophore which is transmitted into the observed twists of the C13-C12 and the C12-C11-bonds. Likewise, the stabilization of the 6-s-cis conformation of the ß-ionone ring inthe binding pocket relative to the 6-s-trans conformation has been supported by these calculations. There is a large amount of mostly spectroscopic data concerning the conformation of retinal analogues inside the rhodopsin which still needs analysis in view of these new structural models.

Bathorhodopsin. Ab initio QM dynamics have been employed to study the photoisomerization reaction of rhodopsin to bathorhodopsin. In addition to the geometry proposed for the structure immediately following the S1 to S0 interconversion ("photorhodopsin") we plan to establish the complete trajectory of the reaction including the spectroscopic properties, until the relaxed bathorhodopsin intermediate is obtained. High on the agenda is the search for factors pertaining to the extra stability of rhodopsin which enables the ligand to act as an inverse agonist to pigment activation.

Intermediates further down the road. With the X-ray structures of lumirhodopsin and of metarhodopsin established QM and QM/MM studies are needed to trace the way of the protein towards the activated state reached in metarhodopsin II.