A3 Differential tissue expression of ADME proteins in humans

Marc Vrana , Pharmaceutics, University of Washington, Seattle, WA
Dale Whittington , Medicinal Chemistry, University of Washington, Seattle, WA
Jashvant D. Unadkat , Pharmaceutics, University of Washington, Seattle, WA
Bhagwat Prasad , Pharmaceutics, University of Washington, Seattle, WA
Purpose: Physiologically based Pharmacokinetic (PBPK) modeling serves an important role in determining the potential in vivo PK characteristics of a new molecular entity (NME). Whether determining the interindividual variability in ADME of drugs, first in human dosing, the potential for drug-drug interactions, or the best formulation characteristics of a drug, PBPK modeling can provide a valuable source of preliminary data with minimal cost or risks. To enhance the capabilities of PBPK models, it is important that the expression of various ADME proteins, i.e., drug metabolizing enzymes and transporters, is known in human tissues important in ADME of drugs. Therefore, the specific goals of this study were: 1) to create a database of in silico selected surrogate peptides for LC-MS/MS protein quantification of 275 ADME proteins; 2) to develop and validate the protein quantification methods for 110 important ADME proteins, and 3) to apply the validated methods to quantify inter-tissue variability in protein expression of ADME genes in human tissues.  

Methods: MRM quantifiable surrogate peptides for 275 human ADME proteins known to be expressed at mRNA level were selected using in silico tools. From this list, the quantification methods for 110 important ADME proteins were refined and validated by experimental analysis of synthetic heavy labeled peptides. The selected heavy peptides were scanned through an iterative process where the best responsive fragment ions were selected, while collision energy and retention time were optimized. Once the methods were optimized for heavy peptides, the corresponding predictive parameters for the light peptides were added into the methods.  110 ADME proteins using 2-3 unique surrogate peptides per protein were quantified in human livers (n=6) and kidneys (n=6) after trypsin digestion and LC-MS/MS analysis as described in our previous publications (1, 2). The relative expression of each protein was normalized to the total amount of tissue protein.

Results: Of the 110 proteins, 29 were detected in both the liver and the kidney, 27 were specific to the liver, and 17 were specific to the kidney. Those detected only in the liver included major CYPs, GSTM1, SULTs, UGT1A4/10 UGT2B15/2B4, ABCB11, SLC10A1, SLC22A1, SLC22A3 and SLCOs. Major GSTs and SLC22s were primarily present in the kidney. Furthermore, amongst the proteins that are expressed both in the liver and the kidney, the expression of ADH1A/1C and CES1 was 100-fold higher in the liver than in the kidney. Similarly, UGT1A1 and EPHX1 expression was >10-fold higher in the liver than in the kidney. In contrast, the expression of GSTP1, UGT1A9 and SLC47A1 was ~10-fold higher in the kidney than in the liver.

Conclusions: The localization and expression of ADME proteins indicates the respective importance of both the human liver and kidney with regards to drug elimination and potential drug-drug interactions. A database of ADME protein quantification methods created in this study will be used to determine the relative and absolute expression of these proteins in various human tissues including the intestine and blood-brain barrier.