Comprehensive characterization of cytochrome P450 isozyme selectivity across chemical libraries

Abstract

The cytochrome P450 (CYP) gene family catalyzes drug metabolism and bioactivation and is therefore relevant to drug development. We determined potency values for 17,143 compounds against five recombinant CYP isozymes (1A2, 2C9, 2C19, 2D6 and 3A4) using an in vitro bioluminescent assay. The compounds included libraries of US Food and Drug Administration (FDA)-approved drugs and screening libraries. We observed cross-library isozyme inhibition (30-78%) with important differences between libraries. Whereas only 7% of the typical screening library was inactive against all five isozymes, 33% of FDA-approved drugs were inactive, reflecting the optimized pharmacological properties of the latter. Our results suggest that low CYP 2C isozyme activity is a common property of drugs, whereas other isozymes, such as CYP 2D6, show little discrimination between drugs and unoptimized compounds found in screening libraries. We also identified chemical substructures that differentiated between the five isozymes. The pharmacological compendium described here should further the understanding of CYP isozymes.

Figure 1. qHTS of cytochrome P450 isozymes

qHTS and potency distribution data for all five isozymes is shown. (a) Data for the entire 17K qHTS against all five isozymes. Dark blue or red represent data that fit to high confidence CRCs (inhibitors or activators respectively) and light blue or red are data that fit to low confidence CRCs (inhibitors or activators respectively). Inactive compounds that did show any concentration response are in grey. (b) The two rows show the potency distribution for category 1 CRCs (dark blue) and category 2 CRCs (light blue) for the biofocused (including the FDA drugs) and MLSMR sets. Scales are identical for both upper and lower graphs and are shown at left and bottom of the graphs.

Figure 2. Distribution and differences in CYP activity between MLSMR vs. FDA sets and comparison to published descriptions

(a) The distribution of compounds in terms of the number of active CYP isozymes that were found for the MLSMR subset (dark grey) and FDA (light grey) libraries. (b) The difference in the percentage of actives for the FDA and MLSMR libraries (left pie chart) as well as the distribution of metabolizing activity (middle pie chart) or hepatic expression (right pie chart), as reported in Shimada et al. are shown.

Figure 3. Clustering of CYP isozyme activity across the 17K compound collection

Self organizing maps are shown where each hexagon represents a cluster of compounds showing structural similarity. The heat map is colored so that red represents clusters that are enriched in compounds active against the CYP enzyme associated with the map and blue represents clusters that are deficient in this regard. A darker red or blue color indicates a higher level of enrichment or deficiency, respectively, in active compounds. For example, the group of blue hexagons that consistently appear in the top middle region of the SOMs indicates a group of structurally related compounds that tend to be inactive against all five CYP isozymes.

Figure 4. Fragment analysis of CYP activity

Selected organic functional groups found to be disproportionately distributed amongst the response classes relative to the overall testing set. The overall percentage of compounds assigned to the indicated response category for each isozyme (Active, white bars; Inactive, grey bars) for each of the five CYPs (in the order of CYP 1A2, 2C9, 2C19, 2D6 and 3A4) is shown. The entire dataset was 16,144 compounds with overall percentages of active and inactive compounds, respectively, of 36% and 42% (CYP 1A2), 25% and 50% (CYP 2C9), 36% and 42% (CYP 2C19), 15% and 66% (CYP 2D6), 32% and 44% (CYP 3A4). Data shown is the difference between the overall percentages and the distribution for the subset of compounds containing the indicated substructure. A shift toward CYP inactivity is reflected in the graphs if the bars show negative values for the active class and positive values for the inactive class. In contrast, a shift toward CYP activity is reflected by positive values in the active class and negative in the inactive class. Substructures are colored blue if these show shifts toward the pan-inactive class, and red if these show shifts toward the pan-active (inhibitor/substrate) class. Substructures showing isoform specificity (e.g. 4 and 5) are colored black.

Figure 5. Fragment analysis of CYP activity for more complex heterocycles

Selected organic functional groups found to be disproportionately distributed amongst the response classes relative to the overall testing set. The overall percentage of compounds assigned to the indicated response category for each isozyme (Active, white bars; Inactive, grey bars) for each of the five CYPs (in the order of CYP 1A2, 2C9, 2C19, 2D6 and 3A4) is shown. The dataset is as described in Figure 4 and the data shown is again the difference between the overall percentages and the distribution for the subset of compounds containing the indicated substructure. Substructures are colored blue if these show shifts toward the pan-inactive class, and substructures showing isoform specificity are colored black. Entries where this value is less (in magnitude) than 10 are shown with a (*).