Class-dependent sequence alignment strategy improves the structural and functional modeling of P450s.

Different procedures for obtaining homology models for P450s are investigated using various sequence alignments sharing various levels of sequence identity with available P450 crystal structures. In this analysis, we have investigated how well homology modeling can reproduce known crystal structures as well as how effectively these homology models can be used to reproduce known ligand-binding modes. Homology models obtained from sequence alignments that discriminate between Class I and Class II P450s are significantly closer to the experimental crystal structures and more closely reproduce known ligand's binding modes, than those obtained using sequence alignments that combine Class I and Class II P450s. The quality of the models is slightly improved by constructing hybrid-structure models that model three of the most variable regions of P450s independently from the rest of the protein: the B region that includes SRS1, the FG region that includes SRS2 and SRS3 and the beta4 region that includes SRS6.