Contribution of OATP1B1 and OATP1B3 to the disposition of sorafenib and sorafenib-glucuronide

Abstract

Purpose: Many tyrosine kinase inhibitors (TKI) undergo extensive hepatic metabolism, but mechanisms of their hepatocellular uptake remain poorly understood. We hypothesized that liver uptake of TKIs is mediated by the solute carriers OATP1B1 and OATP1B3.

Experimental design: Transport of crizotinib, dasatinib, gefitinib, imatinib, nilotinib, pazopanib, sorafenib, sunitinib, vandetanib, and vemurafenib was studied in vitro using artificial membranes (PAMPA) and HEK293 cell lines stably transfected with OATP1B1, OATP1B3, or the ortholog mouse transporter, Oatp1b2. Pharmacokinetic studies were conducted with Oatp1b2-knockout mice and humanized OATP1B1- or OATP1B3-transgenic mice.

Results: All 10 TKIs were identified as substrates of OATP1B1, OATP1B3, or both. Transport of sorafenib was investigated further, as its diffusion was particularly low in the PAMPA assay (<4%) than other TKIs that were transported by both OATP1B1 and OATP1B3. Whereas Oatp1b2 deficiency in vivo had minimal influence on parent and active metabolite N-oxide drug exposure, plasma levels of the glucuronic acid metabolite of sorafenib (sorafenib-glucuronide) were increased more than 8-fold in Oatp1b2-knockout mice. This finding was unrelated to possible changes in intrinsic metabolic capacity for sorafenib-glucuronide formation in hepatic or intestinal microsomes ex vivo. Ensuing experiments revealed that sorafenib-glucuronide was itself a transported substrate of Oatp1b2 (17.5-fold vs. control), OATP1B1 (10.6-fold), and OATP1B3 (6.4-fold), and introduction of the human transporters in Oatp1b2-knockout mice provided partial restoration of function.

Conclusions: These findings signify a unique role for OATP1B1 and OATP1B3 in the elimination of sorafenib-glucuronide and suggest a role for these transporters in the in vivo handling of glucuronic acid conjugates of drugs.

Figure 1. Transport of sorafenib by OATP1B transporters

A) [³H]-Labeled TKIs (0.1 μM) were incubated with HEK293 cells expressing OATP1B1, OATP1B3, or vector control (VC) for 30 min. Intracellular concentrations of drug were determined by liquid scintillation counting. Data represent the mean ± SE of two experiments with triplicate samples (n=6) (*, P<0.05; **, P<0.01; ***, P<0.001). B, C) T-Rex293 cells expressing OATP1B1*1A or VC were incubated with 0.1 μM for the indicated time (B) or the indicated concentration of [³H]sorafenib for 15 min (C). Data represent the mean sorafenib uptake (symbols) ± SE (error bars, shown when larger than symbol) of triplicate samples (n=3). VC represents the cells transfected with an empty vector, representing background transport. The OATP1B1-VC data points, representing the specific contribution of OATP1B1 to the observed transport of sorafenib, were determined by averaging the triplicate values for each group, then subtracting the uptake values obtained in cells transfected with VC from the uptake values obtained in cells expressing OATP1B1.

Figure 2. Influence of Oatp1b2-deficiency on sorafenib pharmacokinetics

A, B) Wildtype (WT) and Oatp1b2(-/-) female mice (n = 4/group) were given 10 mg/kg sorafenib via oral gavage. Plasma sorafenib (A) and sorafenib-glucuronide (B) concentrations were determined by LC-MS/MS. Data represent the mean ± SE (*, P<0.05; **, P<0.01).

Figure 3. Transport of sorafenib-glucuronide by OATP1B transporters

A) Ex vivo sorafenib-glucuronide formation was determined in intestine and liver microsomes from wildtype (WT) and Oatp1b2(-/-) mice. Microsomes (1 mg/mL) were incubated with 10 μM sorafenib for 60 min and sorafenib-glucuronide formation velocity was determined. Data represent the mean ± SE of 4 (intestine) or 16 (liver) samples. B, C) HEK293 cells expressing Oatp1b2, OATP1B1, or OATP1B3 were incubated with 10 μM [³H]sorafenib (B) or 10 μM sorafenib-glucuronide (C) for 15 min and intracellular concentrations were determined by liquid scintillation or LC-MS/MS, respectively. Data represent the mean ± SE drug uptake relative to vector control cells from 2 experiments performed with triplicate samples (n = 6). D) T-Rex293 cells expressing OATP1B1*1A (*1A), OATP1B1*5 (*5), or OATP1B1*15 (*15) were incubated with 0.1 μM [³H]sorafenib or 1 μM sorafenib-glucuronide for 15 min. Data represent the mean ± SE drug uptake of 6-15 replicates (*, P<0.05; **, P<0.01; ***, P<0.001).

Figure 4. Influence of Oatp1a/b-deficiency on sorafenib pharmacokinetics

A, B) Female wildtype (WT), Oatp1a/b(-/-), OATP1B1^tg, and OATP1B3^tg mice (n =4/group) were given 10 mg/kg sorafenib via oral gavage. Plasma (A) sorafenib and (B) sorafenib-glucuronide concentrations were determined by LC-MS/MS. Data represent the mean ± SE (Ø, P<0.001 for all groups compared to WT).