Simon Folkertsma1, Paula van Noort4, Joost Van Durme1, Henk-Jan Joosten1, Emmanuel Bettler5, Wilco Fleuren1, Laerte Oliveira2, Florence Horn3, Jacob de Vlieg4, and Gerrit Vriend1*
1 CMBI, University of Nijmegen, PO Box 9010, 6500 GL, Nijmegen, the Netherlands.
2 Escola Paulista de Medicina, UNIFESP, PO Box 20388, 04041-990, Sao Paulo, Brazil.
3 Department of Cellular and Molecular Pharmacology, UCSF, PO Box 2240, CA 94143-2240, San Francisco, USA.
4 Organon NV, PO Box 20, 5340 BH, Oss, The Netherlands.
5 IBCP, Passage du Vercors 7, 69367, Lyon, France.
* Corresponding author: email@example.com
J Mol Biol. 2004 Aug 6;341(2):321-35.
Literature studies, 3D structure data, and a series of sequence analysis techniques were combined to reveal important residues in the structure and function of the ligand-binding domain of nuclear hormone receptors. A structure-based multiple sequence alignment allowed for the seamless combination of data from many different studies on different receptors into one single functional model.
It was recently shown that a combined analysis of sequence entropy and variability can divide residues in five classes (1) the main function or active site, (2) support for the main function, (3) signal transduction, (4) modulator or ligand binding and (5) the rest. Mutation data extracted from the literature and intermolecular contacts observed in nuclear receptor structures were analyzed in view of this classification and showed that the main function or active site residues of the nuclear receptor ligand-binding domain are involved in cofactor recruitment. Furthermore, the sequence entropy-variability analysis identified the presence of signal transduction residues that are located between the ligand, cofactor and dimer sites, suggesting communication between these regulatory binding sites. Experimental and computational results agreed well for most residues for which mutation data and intermolecular contact data were available. This allows us to predict the role of the residues for which no functional data is available yet.i
This study illustrates the power of family-based approaches towards the analysis of protein function, and it points out the problems and possibilities presented by the massive amounts of data that are becoming available in the 'omics era'. The results shed light on a family of receptors involved in processes ranging from cancer to infertility, and that therefore is one of the more important targets in the pharmaceutical industry.Supplementary materials: