Distances in the Protein Pocket

Kenworthy et al. [45] proposed a model with three subsites in order to explain the binding of testosterone (TS) and diazepam (DZ). One site binds diazepam, another binds testosterone, and the third is capable of binding either diazepam or testosterone. They found that testosterone caused extensive activation of diazepam metabolism, whereas diazepam caused inhibition of testosterone metabolism. Diazepam is present in the inhibitor database used to obtain the pharmacophoric model and its inhibitory activity is well explained by the model (y-residual = 0.27 log units). The model with multiple binding sites proposed by Kenworthy helped to explain the results of the obtained pharmacophoric model if one considers that competition will occur in the catalytic site that can bind either DZ or TS and that 

With the solving of the two crystal structures of 2B4 [47, 48], an overlap of 3A4 with the 2B4 closed and open forms was performed. This was done simply to highlight the flexibility of the enzyme as well as the potential size of a molecule or multiple molecules that could fit into the enzyme. The overlap of CYP3A4 and CYP2B4 structures together with the pertinent (corresponding to the pharmaco-phoric elements) MIFs are shown in Fig. 9.5. 

Many authors [49-51] have demonstrated that inhibition of catalysis by CYP3A4 is substrate dependent and that interactions observed with one CYP3A4 probe may not be representative of those observed with other CYP3A4 substrates. The proposed model could help to explain these data considering that different probes could bind to different sites and therefore show different competition patterns. 

For Pgp it was shown that, for a diverse set of substrates, a pharmacophore could be identified. This pharmacophore played a large role in explaining the variance in PGP inhibition data from a Calcein AM fluorescent assay, suggesting that interacting with the protein is important to inhibition to the efflux transporter. 

Finally, two general 3D-QSAR models superposition independent for datasets with and without ionizable nitrogen were obtained for FIERG inhibitors. 

---