Abstract

Quantitative proteomic methods require optimization at several stages, including sample preparation, liquid chromatography–tandem mass spectrometry (LC-MS/MS), and data analysis, with the final analysis stage being less widely appreciated by end-users. Previously reported measurement of eight uridine-5′-diphospho-glucuronosyltransferases (UGT) generated by two laboratories [using stable isotope-labeled (SIL) peptides or quantitative concatemer (QconCAT)] reflected significant disparity between proteomic methods. Initial analysis of QconCAT data showed lack of correlation with catalytic activity for several UGTs (1A4, 1A6, 1A9, 2B15) and moderate correlations for UGTs 1A1, 1A3, and 2B7 (Rs = 0.40–0.79, P < 0.05; R² = 0.30); good correlations were demonstrated between cytochrome P450 activities and abundances measured in the same experiments. Consequently, a systematic review of data analysis, starting from unprocessed LC-MS/MS data, was undertaken, with the aim of improving accuracy, defined by correlation against activity. Three main criteria were found to be important: choice of monitored peptides and fragments, correction for isotope-label incorporation, and abundance normalization using fractional protein mass. Upon optimization, abundance-activity correlations improved significantly for six UGTs (Rs = 0.53–0.87, P < 0.01; R² = 0.48–0.73); UGT1A9 showed moderate correlation (Rs = 0.47, P = 0.02; R² = 0.34). No spurious abundance-activity relationships were identified. However, methods remained suboptimal for UGT1A3 and UGT1A9; here hydrophobicity of standard peptides is believed to be limiting. This commentary provides a detailed data analysis strategy and indicates, using examples, the significance of systematic data processing following acquisition. The proposed strategy offers significant improvement on existing guidelines applicable to clinically relevant proteins quantified using QconCAT.