Differential Tissue Abundance of Membrane-Bound Drug Metabolizing Enzymes and Transporter Proteins by Global Proteomics

Protein abundance data of drug-metabolizing enzymes and transporters (DMETs) are useful for scaling in vitro and animal data to humans for accurate prediction and interpretation of drug clearance and toxicity. Targeted DMET proteomics that relies on synthetic stable isotope-labeled surrogate peptides as calibrators is routinely used for the quantification of selected proteins; however, the technique is limited to the quantification of a small number of proteins. Although the global proteomics-based total protein approach (TPA) is emerging as a better alternative for large-scale protein quantification, the conventional TPA does not consider differential sequence coverage by identifying unique peptides across proteins. Here, we optimized the TPA approach by correcting protein abundance data by the sequence coverage, which was applied to quantify 54 DMETs for characterization of 1) differential tissue DMET abundance in the human liver, kidney, and intestine, and 2) interindividual variability of DMET proteins in individual intestinal samples (n = 13). Uridine diphosphate-glucuronosyltransferase 2B7 (UGT2B7), microsomal glutathione S-transferases (MGST1, MGST2, and MGST3) carboxylesterase 2 (CES2), and multidrug resistance-associated protein 2 (MRP2) were expressed in all three tissues, whereas, as expected, four cytochrome P450s (CYP3A4, CYP3A5, CYP2C9, and CYP4F2), UGT1A1, UGT2B17, CES1, flavin-containing monooxygenase 5, MRP3, and P-glycoprotein were present in the liver and intestine. The top three DMET proteins in individual tissues were: CES1>CYP2E1>UGT2B7 (liver), CES2>UGT2B17>CYP3A4 (intestine), and MGST1>UGT1A6>MGST2 (kidney). CYP3A4, CYP3A5, UGT2B17, CES2, and MGST2 showed high interindividual variability in the intestine. These data are relevant for enhancing in vitro to in vivo extrapolation of drug absorption and disposition and can be used to enhance the accuracy of physiologically based pharmacokinetic prediction of systemic and tissue concentration of drugs. SIGNIFICANCE STATEMENT: This study quantified the abundance and compositions of drug-metabolizing enzymes and transporters in pooled human liver, intestine, and kidney microsomes as well as individual intestinal microsomes using an optimized global proteomics approach. The data revealed large intertissue differences in the abundance of these proteins and high intestinal interindividual variability in the levels of cytochrome P450s (e.g., CYP3A4 and CYP3A5), uridine diphosphate-glucuronosyltransferase 2B17, carboxylesterase 2, and microsomal glutathione S-transferase 2. These data are applicable for the prediction of first-pass metabolism and tissue-specific drug clearance.

Antioxidant Peptides Derived from Cheese Products via Single and Mixed Lactobacillus Strain Fermentation

Probiotics are used in cheese fermentation to endow the product with unique functional properties, such as enhanced flavor and aroma development through proteolysis and lipolysis. In this study, two probiotic Lactobacillus strains, Lactobacillus plantarum A3 and Lactobacillus reuteri WQY-1, were selected to develop new probiotic cheeses in the form of single- and mixed-strain starters. The results demonstrated that the L. plantarum A3 single-strain group and the L. plantarum A3/L. reuteri WQY-1 mixed fermentation group exhibited superior product performance, particularly the release of functional hydrolysates during cheese ripening. Furthermore, Label-free quantitative proteomic analysis revealed 26 unique antioxidant peptides in the L. plantarum A3 single-strain group and 53 in the L. plantarum A3/L. reuteri WQY-1 mixed fermentation group. Among these, CMENSAEPEQSLACQCL (β-lactoglobulin), CMENSAEPEQSLVCQCL (β-lactoglobulin), and IQYVLSR (κ-casein) have been found to possess potential antioxidant properties both in vitro and in vivo. This confirmed that milk-derived protein peptides in cheese products exhibit potential antioxidant functions through the hydrolysis of probiotic strains.

Quantitative Characterization of Clinically Relevant Drug-Metabolizing Enzymes and Transporters in Rat Liver and Intestinal Segments for Applications in PBPK Modeling

Rats are extensively used as a preclinical model for assessing drug pharmacokinetics (PK) and tissue distribution; however, successful translation of the rat data requires information on the differences in drug metabolism and transport mechanisms between rats and humans. To partly fill this knowledge gap, we quantified clinically relevant drug-metabolizing enzymes and transporters (DMETs) in the liver and different intestinal segments of Sprague-Dawley rats. The levels of DMET proteins in rats were quantified using the global proteomics-based total protein approach (TPA) and targeted proteomics. The abundance of the major DMET proteins was largely comparable using quantitative global and targeted proteomics. However, global proteomics-based TPA was able to detect and quantify a comprehensive list of 66 DMET proteins in the liver and 37 DMET proteins in the intestinal segments of SD rats without the need for peptide standards. Cytochrome P450 (Cyp) and UDP-glycosyltransferase (Ugt) enzymes were mainly detected in the liver with the abundance ranging from 8 to 6502 and 74 to 2558 pmol/g tissue. P-gp abundance was higher in the intestine (124.1 pmol/g) as compared to that in the liver (26.6 pmol/g) using the targeted analysis. Breast cancer resistance protein (Bcrp) was most abundant in the intestinal segments, whereas organic anion transporting polypeptides (Oatp) 1a1, 1a4, 1b2, and 2a1 and multidrug resistance proteins (Mrp) 2 and 6 were predominantly detected in the liver. To demonstrate the utility of these data, we modeled digoxin PK by integrating protein abundance of P-gp and Cyp3a2 into a physiologically based PK (PBPK) model constructed using PK-Sim software. The model was able to reliably predict the systemic as well as tissue concentrations of digoxin in rats. These findings suggest that proteomics-informed PBPK models in preclinical species can allow mechanistic PK predictions in animal models including tissue drug concentrations.