Best Practices for Robust LC‐MS/MS Quantification of Drug Metabolizing Enzymes and Transporters to Predict Inter‐Individual Variability: Case Examples of Hepatic Cytosolic ADHs and ALDH1A1

Background Quantification of drug metabolizing enzymes (DMEs) and transporters in human tissues and biofluids offers unique advantage of predicting inter‐individual variability in drug disposition. While LC‐MS/MS quantification has distinct merits over conventional approaches, the methodology needs thorough understanding of factors that can affect quantitative reproducibility. Therefore, we established a robust protein quantification protocol, which was applied to quantify alcohol and aldehyde dehydrogenase (ADH and ALDH) in the human liver cytosol. Method Following variables involved in the LC‐MS/MS protein quantification were addressed; i) surrogate peptide quality, ii) ion suppression, iii) incomplete and inconsistent protein digestion, iv) sample‐to‐sample processing variability, v) inter‐day (or inter‐batch) variability, vi) tissue quality, vii) linearity and range, viii) correlation of expression and activity. Briefly, the best practices involved, the use of heavy internal standard, exogenous protein internal standard, optimized desalting methodology, incurred sample reanalysis as quality control, optimized data extraction and analysis protocols. This approach was applied to quantify ADH1A, ADH1B, ADH1C and ALDH1A1 in 137 pediatric and 57 adult human liver cytosolic samples. Protein expression data of cytosolic DMEs from various pediatric age groups (neonates, infancy, children, adolescents and adults) were compared using Kruskal‐Wallis test followed by Dunn's multiple comparison test. Results Excellent correlation of peak areas (r 2 >0.95) of various daughter ions across multiple samples confirmed accurate peak integration. Heavy labeled internal standards were used to address any ion suppression. Protein abundance calculated based on multiple peptides of a protein showed excellent correlation (r 2 >0.9) across samples indicating the reliability of surrogate peptide approach to quantify inter‐individual variability. An exogenous protein, bovine serum albumin, was added in each sample as an internal control to address sample‐to‐sample variability during sample processing or trypsin digestion. To address inter‐batch variability, data were further normalized by the incurred pool control included in each batch. Adult levels of ADH1A, ADH1B, ADH1C and ALDH1A1 were 3, 8, 148 and 3‐fold higher than the neonatal levels, respectively. For all proteins, the expression steeply increased during the first 2 years of life reaching adult levels during early childhood. ADH1A protein expression in adults (>18 years) was ~39% lower as compared to children and adolescents. Discussion A robust LC‐MS/MS protocol for quantification of DMEs and transporters in big cohorts of biological samples is critical to generate reproducible data. The systematic approach accurately measured age‐dependent ADHs and ALDH1A1 expression in human cytosol. These data are important for the development of physiologically based pharmacokinetic models to predict age‐dependent drug metabolism of ADH and ALDH1A1 substrates, e.g., 4‐hydroxy cyclophosphamide. Support or Funding Information NICHD R01HD081299‐02.