Development and Validation of a Proteomic Correlation Profiling Technique to Detect and Identify Enzymes Involved in Metabolism of Drugs of Concern.

To predict the variation of pharmacological or toxicological effect caused by pharmacokinetic variance, it is important to be able to detect previously unknown and unsuspected enzymes involved in drug metabolism. We investigated the use of proteomic correlation profiling (PCP) as a technique to identify the enzymes involved in metabolism of drugs of concern. By evaluating the metabolic activities of each enzyme (including isoforms of cytochrome P450, uridine 5' diphospho-glucuronosyltransferase, and hydrolases, plus aldehyde oxidase and carbonyl reductase) on their typical substrates using a panel of human liver samples, we were able to show the validity of PCP for this purpose. R or Rs and P values were calculated for the association between the protein abundance profile of each protein and the metabolic rate profile of each typical substrate. For the 18 enzymatic activities examined, 13 of the enzymes reported to be responsible for the reactions had correlation coefficients higher than 0.7 and were ranked first to third. For the remaining five activities, the responsible enzymes had correlation coefficients lower than 0.7 and lower rankings. The reasons for this were diverse, including confounding resulting from low protein abundance ratios, artificially high correlations of other enzymes due to limited sample numbers, the presence of inactive enzyme forms, and genetic polymorphisms. Overall, PCP was able to identify the majority of responsible drug-metabolizing enzymes across several enzyme classes (oxidoreductase, transferase, hydrolase); use of this methodology could allow more timely and accurate identification of unknown drug-metabolizing enzymes. SIGNIFICANCE STATEMENT: Proteomic correlation profiling using samples from individual human donors was proven to be a useful methodology for the identification of enzymes responsible for drug-metabolism. This methodology could accelerate the identification of unknown drug-metabolizing enzymes in the future.