35] Structural alignments of P450s and extrapolations to the unknown

Abstract

To obtain a sequence alignment that allows model building and identification of potential active site residues or redox partner-binding residues, one should first generate a computer alignment of ones favorite P450 with one of the structurally defined P450s, usually with P450BM-P as it is the most eukaryotic-like P450 of the three available structures. The C-terminal half of the generated alignment will align fairly well due to the large number of sequence similarities and conserved regions in this half of the molecule, i.e., the I helix consensus, the J and K helix, the meander, the heme-binding region, and the L helix. One should then justify the C-terminal half of the P450 with the potential turns found in β-sheets and between other structural elements of P450BM-P. After optimizing the alignment of the C-terminal half of the molecule, attention should be turned toward the more variable N terminus, starting the alignment with conserved residues in the C helix, in the G helix, and in β-strands 1-1 and 1-2. Using these sequences to initiate the N-terminal alignments, one should further refine the remaining half by identifying potential turns at the beginnings and ends of structural elements, and then by identifying hydrophobic residues in the amphipathic helices to determine the orientation of the helix and its length. However, no matter how many times the same sequence is aligned, one should never feel that ones work is done until a crystal structure has been determined for the protein!