The overall goals of this project are to couple our molecular modeling capabilities with new methodologies for encapsulating membrane proteins in soluble Nanodisc structures and obtaining precise three-dimensional structures by solid-state NMR (SSNMR) to derive accurate predictions for membrane proteins. The proposed project pilots these approaches with an analysis of a limited number of cytochrome P450 monooxygenases (P450s) that play critical roles in synthetic and detoxicative functions in mammalian, insect, plant, fungal and bacterial systems. Expression of these membrane-bound proteins utilizes experimental Nanodisc systems that allow membrane proteins to exist in an active, native conformation within a stabile lipid-protein complex to provide monodisperse proteins for solid-state NMR analysis. Use of high-field (600-750 MHz 1H frequency) magic-angle spinning NMR spectroscopy and multidimensional dipolar recoupling pulse sequences will provide low-resolution structural backbone information for the ubiquitous and difficult class of membrane proteins that are not readily characterized at a structural level by many techniques applied to soluble proteins. We will, in a completely interdisciplinary fashion, use our molecular modeling capabilities to facilitate resolution of SSNMR spectra and our SSNMR-derived backbone structures to limit the range of molecular models requiring analysis and structural predictions for side-chain orientations. Automation of the interface between molecular modeling and SSNMR has significant potential for improving the capacity for defining structures on a wide range of proteins, not only membrane-bound P450 proteins but also the entire range of proteins capable of being expressed in prokaryotic (E. coli) and eukaryotic (baculovirus-infected insect cells, yeast) expression systems suitable for isotopic labeling. Structural comparisons of proteins within this very large superfamily of divergent P450 sequences will allow us to define the limits on current molecular modeling procedures, to contrast the evolutionary distinct solutions to common metabolic activities and to begin accurately predicting structures for P450s for which crystal structures may never be a possibility. These studies will also allow to characterize these proteins in a lipid-associated environment that should recapitulate interactions of these proteins with their native membranes and identify differences with available crystal structures derived for their truncated and soluble forms.
