It can be expected that the ability of a drug or foreign compound to be metabolized by a particular isozyme will be a function of both the tendency of various sites within the molecule to undergo oxidation (electronic factors), and its binding characteristics (steric factors). We have developed a predictive model that estimates the tendency for a substrate to undergo P450 mediated hydrogen atom abstraction. We have tested this model by correlating the predicted oxidation rates and the toxicity of a series of nitriles. When hydrophobicity alone (pi and pi 2) was used to predict toxicity (1/LD50), a poor correlation was obtained (r = 0.59). However, a good correlation (r = 0.85) was obtained when toxicity was regressed on a combination of hydrophobicity and a rate constant corrected for the fraction of metabolism occurring at the alpha- position. After deleting two compounds for which acute toxicity is not thought to result from P450 oxidation and two compounds with olefinic oxidation sites, a minor improvement on the model was obtained (r = 0.90). Good correlations with other experimental data sets available in the literature suggest that theoretical models may be used to predict the electronic tendency for cytochrome P450 mediated oxidations. Stereochemical analysis of the metabolism of B[a]P reveals that seven human P450s, five rodent P450s and the bacterial P450cam all convert B[a]P to the most potent carcinogenic stereoisomer. Molecular dynamics was used to determine the specific amino acids responsible for the stereochemical outcome. The results suggest that a single helical region, that is likely to be conserved in all P450s, plays a primary role.

Agency
National Institute of Health (NIH)
Institute
National Cancer Institute (NCI)
Type
Intramural Research (Z01)
Project #
1Z01CP005676-03
Application #
3752724
Study Section
Project Start
Project End
Budget Start
Budget End
Support Year
3
Fiscal Year
1993
Total Cost
Indirect Cost
Name
Division of Cancer Epidemiology and Genetics
Department
Type
DUNS #
City
State
Country
United States
Zip Code