A Model for Predicting Likely Sites of CYP3A4-mediated Metabolism on Drug-like Molecules

Abstract

We have developed a rapid semiquantitative model for evaluating the relative susceptibilities of different sites on drug molecules to metabolism by cytochrome P450 3A4. The model is based on the energy necessary to remove a hydrogen radical from each site, plus the surface area exposure of the hydrogen atom. The energy of hydrogen radical abstraction is conventionally measured by AM1 semiempirical molecular orbital calculations. AM1 calculations show the following order of radical stabilities for the hydrogen atom abstractions: sp2 centers > heteroatom sp3 centers > carbon sp3 centers. Since AM1 calculations are too time intensive for routine work, we developed a statistical trend vector model, which is used to estimate the AM1 abstraction energy of a hydrogen atom from its local atomic environment. We carried out AM1 and trend vector calculations on 50 CYP3A4 substrates whose major sites of metabolism are known in the literature. A plot of the lowest hydrogen radical formation energy versus its sterically accessible surface area exposure for these 50 substrates shows that only those hydrogen atoms with solvent accessible surface area exposure > or = 8.0 A(2) are susceptible to CYP3A4-mediated metabolism. This approach forms the basis for our general model, which predicts sites on drugs that are susceptible to cytochrome P450 3A4-mediated hydrogen radical abstraction followed by a hydroxylation reaction. This model, in conjunction with specific enzyme site binding requirements, can aid in identifying possible sites of metabolism catalyzed by other cytochrome P450 enzymes.

Cytochrome P450 3A4-catalyzed Testosterone 6β-Hydroxylation Stereochemistry, Kinetic Deuterium Isotope Effects, and Rate-limiting Steps

“…Together the results provide a view of P450 3A4 different from that often presented in the literature about this enzyme, i.e. a protein with a large cavity in which substrates have considerable freedom to move and the oxidation reactions are dominated primarily by the inherent chemical reactivity at individual atoms (16,17).…”

“…However, the crystal structures have provided evidence that P450 3A4 has a large cavity available for substrate binding. As mentioned earlier, one view of the active site of P450 3A4 is that binding is relatively loose and the regioselectivity of oxidations is dominated by the ease of oxidation at individual atoms (16,17). However, the results of the experiments with the relatively unselective metalloporphyrin biomimetic models argue against this view.…”

“…P450 3A4 exhibits heterotropic and homotropic cooperativity in some but not all of its reactions (4,(7)(8)(9)(10), and several models have been proposed to explain such behavior (9,(11)(12)(13)(14)(15). A view frequently expressed for smaller substrates is that binding is relatively loose and the regioselectivity of oxidation of substrates is largely a function of the ease of C-H bond breaking (16,17).…”

Precursors for cytochrome P450 profiling breath tests from an in silico screening approach

“…Positions yielding a dE smaller than 30 kcal mol −1 are considered as likely SOMs. This empirical cutoff value is of similar magnitude as values discussed in literature for aliphatic carbon centers [40]. Additionally, we ensured accessibility of predicted sites by discarding potential SOMs buried within the energy optimized conformer.…”

Classification of Highly Unbalanced CYP450 Data of Drugs Using Cost Sensitive Machine Learning Techniques.

“…Therefore the in silico prediction of drug metabolism profiles of CYP450s has become one of the key technologies in early drug discovery support. [15][16][17][18][19][20][21][22][23][24] The primary aim of this paper is to demonstrate how to build useful classification models out of unbalanced [25][26][27][28] data sets. We consider a data set to be unbalanced if either the sizes of the two classes differ significantly, or the costs for a false negative classification are very high whereas a false positive is acceptable, or if both conditions hold.…”