Molecular basis of ligand recognition in the human glucocorticoid receptor

Johannes R.G. von Langen1, Stephan Diekmann2, Alexander Hillisch, IMB-Jena;, IMB-Jena

Introduction: The genome sequencing projects provided us with tremendous amounts of protein sequence information. However, to elucidate and understand the function of these proteins, structural information is essential. In general, protein structure is determined by x-ray crystallography and NMR spectroscopy. Nevertheless, for only about one percent of the known proteins an experimentally determined structure is available [1]. Homology modelling helps to fill this sequence-structure gap. The quality of the computed models strongly depends on the sequence identity between target and template. Highly accurate models can be built if this sequence identity exceeds 50 %. Methods: We have built a homology model of the human glucocorticoid receptor (hGR) ligand binding domain (LBD) based on a 53% sequence identity to the human progesterone receptor (hPR) and it’s crystal structure of the ligand binding domain [2]. Our goal is to get an in-depth understanding of the recognition of the endogenous ligand cortisol by the hGR in comparison with other similar steroids showing weaker or no binding to the hGR. In order to predict the mode and the dynamic of binding of the five steroids (estradiol, progesterone, testosterone, aldosterone and cortisol) to the hGR, we performed 5 separate 4 ns molecular dynamics (MD) simulations of the homology modelled protein-steroid complexes in aqueous environment. The steroids were placed in the hGR binding pocket similar to the binding mode of progesterone observed in the hPR [2]. Free energy of binding was simulated and predicted using the MM-PB/SA method. The conformation and the volume of the binding pockets of the hGR protein during the MD-simulation of all five protein-steroid complexes was analysed. Our homology model of the human glucocorticoid receptor is compared with the recently solved glucocorticoid receptor x-ray structure [3]. Results: During the MD simulation, cortisol binding changes slightly and adopts a binding mode which is in agreement with the x-ray structure. During the simulations of the other steroid-hGR complexes conformational changes in the global structure of the protein were observed, probably reflecting the lower binding affinity of those molecules. Using MD-simulations of hGR homology models, it was possible to discriminate good and bad binding ligands and recognise cortisol as the endogenous ligand of the hGR in silico. This paves the way for homology modelling of other members of the nuclear (orphan) receptor family. 1) A. Hillisch, R. Hilgenfeld, The role of protein 3D-structures in the drug discovery process, in: Modern Methods of Drug Discovery, eds. Hillisch A., Hilgenfeld, R., Birkhäuser Publishing Ltd., Basel, 2003. 2) S. P. Williams, P. B. Sigler, Atomic structure of progesterone complexed with its receptor (1998) Nature 393, 392-396. 3) R.K. Bledsoe, V.G. Montana, T.B. Stanley, C.J. Delves, C J. Apolito, D.D. McKee, T.G. Consler, D.J. Parks, E.L. Stewart, T.M. Willson, M.H. Lambert, J.T. Moore, K.H. Pearce and H.E. Xu, Crystal Structure of the Glucocorticoid Receptor Ligand Binding Domain Reveals a Novel Mode of Receptor Dimerization and Coactivator Recognition (2002) Cell 110, 93-105.