Multiplex and Label-Free Relative Quantification Approach for Studying Protein Abundance of Drug Metabolizing Enzymes in Human Liver Microsomes Using SWATH-MS.

We describe a sequential windowed acquisition of all theoretical fragment ion mass spectra (SWATH-MS) based method for label-free, simultaneous, relative quantification of drug metabolism enzymes in human liver microsomes (HLM; n = 78). In-solution tryptic digestion was aided by a pressure cycling method, which allowed a 90 min incubation time, a significant reduction over classical protocols (12-18 h). Digested peptides were separated on an Acquity UHPLC Peptide BEH C18 column using a 60 min gradient method at a flow rate of 0.100 mL/min. The quadrupole-time-of-flight mass spectrometer (ESI-QTOFMS) was operated in positive electrospray ionization mode, and data were acquired by data-dependent acquisition (DDA) and SWATH-MS mode. A pooled HLM sample was used as a quality control to evaluate variability in digestion and quantification among different batches, and inter-batch %CV for various proteins was between 3.1 and 7.8%. Spectral library generated from the DDA data identified 1855 distinct proteins and 25 681 distinct peptides at a 1% global false discovery rate (FDR). SWATH data were queried and analyzed for 10 major cytochrome P450 (CYP) enzymes using Skyline, a targeted data extraction software. Further, correlation analysis was performed between functional activity, protein, and mRNA expression for ten CYP enzymes. Pearson correlation coefficient (r) between protein and activity for CYPs ranged from 0.314 (CYP2C19) to 0.767 (CYP2A6). A strong correlation was found between CYP3A4 and CYP3A5 abundance and activity determined using midazolam and testosterone (r > 0.600, p < 0.001). Protein-to-activity correlation was moderate (r > 0.400-0.600, p < 0.001) for CYP1A2, CYP2A6, CYP2B6, CYP2C9, and CYP2E1 and significant but poor (r < 0.400, p < 0.05) for CYP2C8, CYP2C19, and CYP2D6. The findings suggest the suitability of SWATH-MS based method as a valuable and relatively fast analytical technique for relative quantification of proteins in complex biological samples. We also show that protein abundance is a better surrogate than mRNA to predict the activity of CYP activity.