Pre-incubation with OATP1B1 and OATP1B3 inhibitors potentiates inhibitory effects in physiologically relevant sandwich-cultured primary human hepatocytes

Abstract

Organic anion transporting polypeptides (OATP)1B1 and OATP1B3 are liver-specific transport proteins that express on the basolateral membrane of human hepatocytes and mediate hepatic uptake of many drugs, including statins. They are important determinants of transporter-mediated drug-drug interactions (DDIs). It has been reported that pre-incubation with some OATP1B1 and OATP1B3 inhibitors potentiates the inhibitory effects, yielding reduced IC₅₀ values. The US FDA draft guidance has recently recommended to use the lower IC₅₀ values after inhibitor-preincubation to assess OATP1B1 and OATP1B3-mediated DDIs. However, it remains unknown whether the potentiation effects of inhibitor-preincubation on IC₅₀ values occur in a physiologically relevant cell model. The current study was designed to determine the IC₅₀ values of OATP1B1 and OATP1B3 inhibitors everolimus (EVR), sirolimus (SIR), and dasatinib against OATP1B substrates in physiologically relevant primary human hepatocytes with or without inhibitor-preincubation and to compare the OATP-mediated DDI prediction using data from primary human hepatocytes and that reported previously in transporter-expressing cell lines. Primary human hepatocytes were cultured in a sandwich configuration. Accumulation of [³H]-CCK-8 (1 μM, 1.5 min), [³H]-rosuvastatin (0.5 μM, 2 min) and [³H]-pitavastatin (1 μM, 0.5 min) was determined in human sandwich-cultured hepatocytes (SCH) in the presence of vehicle control or an inhibitor with or without inhibitor-preincubation at designated concentrations, and was utilized to determine the IC₅₀ values for these inhibitors. R-value models were used to predict OATP-mediated DDIs. Pre-incubation with EVR at a clinically relevant concentration of 0.2 μM significantly reduced accumulation of [³H]-CCK-8 and [³H]-rosuvastatin even after washing. Reduced IC₅₀ values following inhibitor pre-incubation were observed for all three inhibitors using [³H]-CCK-8 and [³H]-rosuvastatin as substrates in human SCH. The IC₅₀ values after inhibitor-preincubation were lower or comparable in transporter-expressing cell lines compared with that in human SCH. For dasatinib, R-values from both cell lines and human SCH were greater than the US FDA cut-off value of 1.1. For EVR, R values from cell lines were 1.23 and were lowered to near 1.1 (1.08–1.09) in human SCH. For SIR, R values from either cell type were less than the cut-off values of 1.1. In conclusion, the current study is the first to report that pre-incubation with OATP1B inhibitors potentiates inhibitory effects in physiologically relevant primary human hepatocytes, supporting the rationale of the current US FDA draft guidance of including an inhibitor-preincubation step when assessing OATP-mediated DDIs in vitro. IC₅₀ values after inhibitor-preincubation in transporter-expressing cell lines may be used for DDI prediction for the purpose of mitigating false-negative OATP-mediated DDI prediction.

INTRODUCTION

Hepatic basolateral uptake transporters organic anion transporting polypeptides (OATP) 1B1 and OATP1B3 transport many important drugs, including lipid-lowering statins, from the blood into the liver (König et al., 2000b; König et al., 2000a). Because reduced OATP1B1 and OATP1B3 transporter activity due to drug-drug interactions (DDIs) or genetic variation of the transporters often leads to increased exposure of OATP1B1 and 1B3 substrate drugs, OATP1B1 and 1B3 are key determinants of transporter-mediated DDIs (Giacomini et al., 2010; Tweedie et al., 2013). In vitro pre-incubation with some OATP inhibitors has been reported to potentiate the inhibitory effects (Shitara and Sugiyama, 2017; US FDA, 2020). Reduced IC₅₀ values following inhibitor-preincubation have been reported for several OATP1B1/3 inhibitors, including cyclosporine (Cs) A (Amundsen et al., 2010; Gertz et al., 2013; Izumi et al., 2015; Pahwa et al., 2017), rifampicin (Pahwa et al., 2017), dasatinib (Pahwa et al., 2017), and mTOR inhibitors everolimus (EVR) and sirolimus (SIR) (Farasyn et al., 2019). For CsA and rifamipicin, the IC₅₀ values determined following pre-incubation in vitro (Gertz et al., 2013; Pahwa et al., 2017) are close to the estimated in vivo inhibition constant (Ki) values against the OATP transporters (Yoshikado et al., 2016). Recent US FDA draft guidance recommended adding an inhibitor pre-incubation step when determining the IC₅₀ values of investigational drugs against OATP1B1 and 1B3 in vitro, and to use the IC₅₀ values following inhibitor pre-incubation to assess the DDI potential using model approaches (US FDA, 2020).

OATP1B1 and OATP1B3 predominantly express on hepatocytes of human livers (König et al., 2000b; König et al., 2000a). To date, the potentiation effects of OATP inhibitor-preincubation on their inhibition potency have been primarily observed in transporter-overexpressing human embryonic kidney (HEK) 293 cell lines, which are not physiologically relevant (Amundsen et al., 2010; Gertz et al., 2013; Izumi et al., 2015; Pahwa et al., 2017; Farasyn et al., 2019), with limited reports in primary hepatocytes from animal species including cynomolgus monkey and rats (Shitara et al., 2009; Ufuk et al., 2018). For example, in cynomolgus monkey hepatocytes, pre-incubation with CsA reduced the IC₅₀ values against rosuvastatin and pitavastatin, and pre-incubation with rifampicin reduced IC₅₀ values against rosuvastatin (Ufuk et al., 2018). Another study (Shitara et al., 2009) reported that pre-incubation of rat hepatocytes with CsA reduced the IC₅₀ values of CsA against bromosulfophthalein (BSP), a substrate of Oatps (Cui et al., 2001; Kullak-Ublick et al., 2001). Currently, there is a lack of information regarding whether pre-incubation with OATP1B1/3 inhibitors also potentiate the inhibitory effects and yield reduced IC₅₀ values in physiologically relevant primary human hepatocytes.

The sandwich cultured human hepatocyte model is a well characterized in vitro model that is physiologically relevant (Swift et al., 2010). It has been utilized to assess hepatic OATP transport activity (Powell et al., 2014; Pahwa et al., 2017). The aims of the current study were twofold: 1) to determine the IC₅₀ values of OATP1B1 and OATP1B3 inhibitors everolimus (EVR), sirolimus (SIR), and dasatinib against OATP1B substrates in physiologically relevant primary human hepatocytes with or without inhibitor-preincubation and 2) to compare the R-values of OATP-mediated DDI prediction using data from primary human hepatocytes with that published in transporter-expressing cell lines.

MATERIALS AND METHODS

Materials.

[³H]cholecystokinin 8 (CCK-8) (specific activity 98.7.0 Ci/mmol) was purchased from Perkin Elmer Life Science (Waltham, MA). [³H]Pitavastatin (specific activity 5 Ci/mmol), unlabeled pitavastatin and [³H]rosuvastatin (specific activity 20 Ci/mmol) were purchased from American Radiolabeled Chemicals (St. Louis, MO). Unlabeled rosuvastatin was purchased from Toronto Research Chemicals, Inc. (North York, ON, Canada). EVR, SIR, and dasatinib were purchased from LC laboratories (Woburn, MA). Unlabeled CCK-8, bromosulfophtalein (BSP), rifampicin, and Dulbecco’s Modified Eagle Medium (DMEM) were purchased from Sigma-Aldrich (St. Louis, MO). Fetal bovine serum (FBS) was purchased from Hyclone Laboratories (Logan, Utah). Bio-Safe II liquid scintillation mixture was purchased from Research Products International (Mt. Prospect, IL). All other materials were purchased from Thermo Fisher Scientific (Waltham, MA), Sigma-Aldrich (St. Louis, MO), or VWR International (Radnor, PA).

Human SCH.

Primary human hepatocytes were purchased from Life Technology (Grand Island, NY) and Triangle Research Labs, LLC (Research Triangle Park, NC), with donor demographics shown in Supplemental Table S1. Hepatocytes were cultured in sandwich configuration as described previously (Swift et al., 2010; Alam et al., 2016; Pahwa et al., 2017). Briefly, on day 0, cells were seeded at a density of 3.5 ×10⁵ cells per well of 24-well Biocoat^™ culture plates in DMEM seeding medium containing 1% (v/v) MEM non-essential amino acid (NEAA), 1 μM dexamethasone, 4 μg/ml insulin, 5% (v/v) FBS, 2 mM L-glutamine, 100 units of penicillin G sodium/ml, and 100 g/ml of streptomycin sulfate. At 6 h after seeding when the cells were attached to the plate, cells were overlaid with Matrigel^™ at 0.25 mg/ml in DMEM feeding medium containing 0.1 μM dexamethasone, 1% (v/v) ITS+ premix (Corning, Inc., NY), 2 mM L-glutamine, 1% (v/v) MEM NEAA, 100 units/ml of penicillin G sodium, and 100 g/ml of streptomycin sulfate. Culture medium was replaced every 24 h thereafter. On day 2 of culture, transport assays were conducted.

Inhibition and Uptake Studies.

Accumulation of [³H]rosuvastatin (0.5 μM, 2 min), [³H]CCK-8 (1 μM, 1.5 min), and [³H]Pitavastatin (1 μM, 0.5 min) was determined as reported previously (Powell et al., 2014; Alam et al., 2016; Pahwa et al., 2017). Substrate incubation times were all within the linear uptake ranges as published previously (Powell et al., 2014; Alam et al., 2016; Pahwa et al., 2017) for [³H]CCK-8 and [³H]Pitavastatin, or determined in the present study for [³H]rosuvastatin (Supplemental Fig. S1). The substrate concentrations used in the present study were below the Michaelis constant (K_m) values for OATP1B1 [rosuvastatin - 0.8 μM (Kitamura et al., 2008), pitavastatin - 3 μM (Hirano et al., 2004)] and OATP1B3 [rosuvastatin – 14.2 μM (Kitamura et al., 2008)], pitavastatin - 3 μM (Hirano et al., 2004)], and CCK-8 – 3.8 μM (Hirano et al., 2004; Kitamura et al., 2008). Hence, the IC₅₀ values determined using these substrates are approximate to the inhibition constant (K_i) values of the transporters.

The inhibitor effects of EVR, SIR, and dasatinib on [³H]CCK-8, [³H]rosuvastatin, or [³H]pitavastatin uptake in human SCH were determined under pre-incubation, co-incubation, or co-incubation following a pre-incubation step (pre+co-incubation), similarly to those described previously (Pahwa et al., 2017; Farasyn et al., 2019). In the pre-incubation condition, substrate accumulation was determined after incubation with vehicle control or an inhibitor in feeding medium for 1 h and washing. In the co-incubation condition, the uptake assay was conducted in the presence of vehicle control or inhibitors without any pre-incubation. In the pre+co-incubation condition, cells were first pre-incubated with inhibitors or vehicle control for 1 h, washed, and then incubated in HBSS buffer containing inhibitors at the same concentrations utilized in the pre-incubation step. DMSO (≤ 0.1%, v/v) was used as the vehicle control. BSP and rifampicin, which are potent inhibitors of OATP1B1 and OATP1B3 (Annaert et al., 2010), served as positive controls. At the end of substrate incubation, after lysis with 0.5% Triton-X 100-PBS buffer, accumulation of radiolabeled substrates was determined by scintillation counting (LS6500 scintillation counter, Beckman Coulter, Brea, CA) and normalized to protein concentration determined via BCA assay (Pierce Chemical, Rockford, IL). Assay on a blank plate without cells was included to correct for nonspecific binding.

IC₅₀ values were determined using the method similar to that published previously (Pahwa et al., 2017; Farasyn et al., 2019). Substrate accumulation (expressed as percentage of the vehicle control) was plotted against the inhibitor concentrations. Nonlinear regression was used to fit the data according to Equation 1 using the three-parameter model with GraphPad Prism v.7.0 (GraphPad Software, La Jolla, CA), where the top value was fixed to 100%:

$$\text{E}=\text{Bottom}+\left(\text{Top}-\text{Bottom}\right)/\left(1+\left(\text{C}/\text{I}{\text{C}}_{50}\right)\right)$$ Eq. 1

DDI prediction using the R-value model.

R-values, which reflect the predicted area under the curve (AUC) ratio of a victim drug in the presence vs. absence of an investigational drug, were determined according to Eq. 2 following US FDA draft guidance for in vitro OATP1B1 and 1B3 DDI studies (US FDA, 2020),

$$\text{R}=1+\left({\text{f}}_{\text{u},\text{p}}\times {\text{I}}_{\text{in},\text{max}}/\text{I}{\text{C}}_{50}\right)$$ Eq. 2

The I_in,max is calculated as:

$${\text{I}}_{\text{in},\text{max}}={\text{I}}_{\text{max}}+\left({\text{F}}_{\text{a}}{\text{F}}_{\text{g}}\times {\text{k}}_{\text{a}}\times \text{Dose}\right)/\text{Qh}/\text{RB}$$ Eq. 3

Where f_u,p is the inhibitor fraction of unbound in the plasma, I_max is the maximum concentration of the inhibitor in the plasma of systemic circulation, I_in,max is the estimated maximum plasma concentration of the inhibitor at the inlet to the liver, F_a is the fraction absorbed, F_g is the intestinal availability, ka is the absorption rate constant, and Qh is the hepatic blood flow rate (1500 ml/min (Bradley et al., 1945). RB is the blood-to-plasma concentration ratio. FaFg = 1 and ka = 0.1/min were used as a worst-case estimate. The parameters were obtained from the literature and are summarized in the supplemental materials (Supplemental Table S2).

Statistics.

For the statistical analysis shown in Fig. 1 A, C and E, Fig. 2 A, C and E, and Fig. 4 and 5, fold changes and associated standard errors (SEs) vs. control were estimated using linear mixed models with a random effect (human hepatocytes donor) and a fixed group effect, allowing for group-specific variances, similarly as published previously (Pahwa et al., 2017; Farasyn et al., 2019). P-values were adjusted using Bonferroni’s method in cases of multiple comparisons. A two-sided p-value of ≤ 0.05 defines statistical significance. Student’s t-test was used to determine the statistical significance of IC₅₀ values between co-incubation and pre+co-incubations when n=3 human hepatocytes donors were used for both conditions. SAS software (version 9.3, Cary, NC) was used for statistical analyses.

Figure 1.

Effects of pretreatment with EVR on substrate transport in human SCH. Model-estimated fold change and associated SE of accumulation of (A) [³H]CCK-8 (1 μM, 1.5 min), (C) [3H]rosuvastatin (0.5 μM, 2 min), and (E) [³H]Pitavastatin (1 μM, 0.5 min) vs. CTL in human SCH in the absence of EVR after pretreatment with EVR at the indicated concentrations for 1 h followed by washing. *p<0.05 vs. CTL by mixed-effect model (n=3 hepatocytes donors at each concentration in triplicate). B, D, and F. The IC₅₀ values were determined by fitting dose-response curves to the data by nonlinear regression analysis in co-incubation and pre+co-incubation as determined in the Materials and Methods. Solid (co-incubation) and dashed lines (pre+co-incubation) represent the fitted lines. Data are expressed as percentage of vehicle control in co-incubation (closed circles) and pre+co-incubation (open circles) scenarios. Data represent mean ± SEM (n=3 hepatocytes donors in triplicate).

Figure 2. Effects of pretreatment with SIR on substrate transport in human SCH.

Model-estimated fold change and associated SE of accumulation of (A) [³H]CCK-8 (1 μM, 1.5 min), (C) [³H]rosuvastatin (0.5 μM, 2 min), and (E) [³H]Pitavastatin (1 μM, 0.5 min) vs. CTL in human SCH in the absence of SIR after pretreatment with SIR at indicated concentrations for 1 h, followed by washing. *p<0.05 vs. CTL by mixed-effect model (n=3 hepatocytes donors at each concentration in triplicate). B, D, and F. The IC₅₀ values were determined by fitting dose-response curves to the data by nonlinear regression analysis in co-incubation and pre+co-incubation as determined in the Materials and Methods. Solid (co-incubation) and dashed lines (pre+co-incubation) represent the fitted lines. Data are expressed as percentage of vehicle control in co-incubation (closed circles) and pre+co-incubation (open circles) scenarios. Data represent mean ± SEM (n=3 hepatocytes donors in triplicate for B and D), or mean ± range (n=2 hepatocytes donors in triplicate for F).

Figure 4. Effects of BSP on substrates transport in human SCH.

Model-estimated fold change and associated SE of accumulation of (A) [³H]CCK-8 (1 μM, 1.5 min), (B) [³H]rosuvastatin (0.5 μM, 2 min), and (C) [³H]pitavastatin (1 μM, 0.5 min) vs. CTL in human SCH. Co-incubation, the indicated substrate was co-incubated with vehicle control or BSP (100 μM). Pre-incubation, human SCH was preincubated with BSP (100 μM) or vehicle control for 1 h; after washing, substrate accumulation was determined in the absence of inhibitor BSP. All experiments were conducted in triplicate (n=2 and 3 hepatocytes donors in A and C, respectively; in B, n=2 and 3 hepatocytes donors for co-incubation and pre-incubation, respectively). *p<0.05 vs. CTL by mixed-effect model.

Figure 5. Effects of rifampicin on substrates transport in human SCH.

Model-estimated fold change and associated SE of accumulation of (A) [³H]CCK-8 (1 μM, 1.5 min) and (B) [³H]rosuvastatin (0.5 μM, 2 min) vs. CTL in human SCH. Cells were preincubated with 50 μM rifampicin for 1 h, then co-incubated with 50 μM rifampicin and substrate [³H]CCK-8 (1 μM, 1.5 min) (A) or [3H]rosuvastatin (0.5 μM, 2 min) (B). Experiments were conducted in triplicate. N=2 for CCK-8, N=3 for rosuvastatin. *p<0.05 vs. CTL by mixed-effect model.

RESULTS

Effects of EVR and SIR pre-incubation on accumulation of and IC₅₀ values against OATP1B substrates in human SCH.

For EVR and SIR, each human hepatocyte donor was assayed for both co-incubation and pre+co-incubation IC₅₀ values. Pre-incubation with EVR significantly decreased accumulation of [³H]CCK-8 (ranging 0.83 ± 0.05–0.65 ± 0.06 fold of CTL at 0.2–0.5 μM), [³H]rosuvastatin (0.8 ± 0.07–0.59 ± 0.07 fold of CTL at 0.2 and 0.8 μM), and [³H]pitavastatin (0.78±0.06 fold vs. CTL at 0.5 μM), respectively (Fig. 1 A, C, E). A one-hour pre-incubation followed by co-incubation (pre+co-incubation) with EVR significantly decreased the IC₅₀ values of EVR against [³H]-rosuvastatin and [³H]-CCK-8, respectively (p<0.05 by t-test, n=3 in triplicate) (Fig. 1 B and D, Table 1). Pre+co-incubation with EVR led to a decreasing trend of IC₅₀ values against pitavastatin (~10 fold). However, the trend did not reach a statistically significant level (p=0.057 by t-test, n=3 in triplicate) (Fig. 1 F and Table 1).

Table 1.

IC₅₀ and R-values of EVR, SIR, and dasatinib against rosuvastatin, pitavastatin, and CCK-8 in primary human hepatocytes. Previously published R-values in OATP1B1- or OATP1B3-expressing cell lines (Farasyn et al., 2019) are shown for comparison. IC₅₀ and R-values of inhibitors in each hepatocytes donor in the presence or absence of pre-incubation are shown in Table S3.

	Cell type	Substrates	Fold Change in IC50 (Co vs. pre+Co)	Co-Incubation		Pre+Co-Incubation
				IC50 (μM)	R	IC50 (μM)	R
EVER	Hepatocytes	Rosuvastatin	2.8	1.81 ± 0.31	1.02 (1.02–1.03)	0.64 ± 0.09*	1.07 (1.05–1.11)
	Hepatocytes	Pitavvastatin	10.3	5.16 ± 3.12	1.01 (1.01–1.02)	0.50 ± 0.16	1.09 (1.08–1.13)
	Hepatocytes	CCK-8	3.6	1.89 ± 0.35	1.02 (1.02–1.03)	0.53 ± 0.06*	1.08 (1.06–1.10)
	OATP1B1a	Rosuvastatin	8.3	1.58 ± 0.92	1.03 (1.02–1.09)	0.19 ± 0.07*	1.23 (1.10–1.42)
	OATP1B3a	CCK-8	2.7	0.51 ± 0.08	1.09 (1.07–1.11)	0.19± 0.02*	1.23 (1.21–1.25)
SIR	Hepatocytes	Rosuvastatin	2.5	1.90 ± 0.39	1.00 (1.00–1.00)	0.77 ± 0.07*	1.00 (1.00–1.00)
	Hepatocytes	Pitavvastatin	1.1	0.93 ± 0.67	1.00 (1.00–1.01)	0.87 ± 0.64	1.00 (1.00–1.01)
	Hepatocytes	CCK-8	7.6	2.72 ± 1.28	1.00 (1.00–1.00)	0.36± 0.07*	1.01 (1.00–1.01)
	OATP1B1a	Rosuvastatin	1.8	0.56± 0.30	1.00 (1.00–1.03)	0.32± 0.11	1.01 (1.01–1.03)
	OATP1B3a	CCK-8	2.7	0.97 ± 0.20	1.00 (1.00–1.01)	0.36± 0.05*	1.01 (1.00–1.01)
Dasatinib	Hepatocytes	Rosuvastatin	2.9	3.38 ± 1.42	1.09 (1.05–1.14)	1.17 ± 0.39	1.27 (1.15–1.51)
	OATP1B1b	E217βG	1.9	4.81 ± 1.7	1.07 (1.04–1.09)	2.33 ± 0.86	1.13 (1.09–1.20)
	OATP1B3b	E217βG	2.1	5.76 ± 1.1	1.05 (1.05–1.09)	2.75 ± 1.17*	1.11 (1.09–1.22)

IC₅₀ values are expressed as the mean ± SE from the model estimate. Determination of IC₅₀ and R-values were described in Materials and Methods. R>1.1 were bolded. t-test was conducted when IC₅₀ values are n=3 for both conditions.

^*indicates a statistically significant difference between pre+co-IC₅₀ and co-IC₅₀ values (p<0.05 by student t-test).

Pre-incubation with SIR at 0.6 μM significantly decreased accumulation of [³H]CCK-8 (ranging 0.68 ± 0.07 fold vs. CTL) and [³H]rosuvastatin (0.63 ± 0.08 fold vs. CTL) (Fig. 2 A and C), while it did not affect that of [³H]pitavastatin (Fig. 2 E). One hour of pre-incubation followed by co-incubation (pre+co-incubation) with SIR significantly decreased the IC₅₀ values of SIR against [³H]CCK-8 and [³H]rosuvastatin (Fig. 2 B and D, p<0.05 by t-test, n=3, Table 1), while it did not appear to affect IC₅₀ values against [³H]-pitavastatin (Fig. 2 F, n=2 Table 1). For EVR and SIR, the pre+co-incubation-IC₅₀ values were lower in transporter-expressing cell lines than in human SCH (Table 1). As shown in Table S3, for EVR and SIR, pre-incubation effects were similarly observed in all batches of hepatocytes.

Dasatinib IC₅₀ values against [³H]rosuvastatin in human SCH.

The IC₅₀ values of dasatinib under the pre+co-incubation-condition was 2.9 fold lower than that under the co-incubation condition (1.17 vs. 3.38 μM) (Fig. 3, Table 1). Different human hepatocytes donors were used in the co-incubation and pre+co-incubation IC₅₀ value determination (Table S3). The observed difference between co-incubation and pre+co-incubation for dasatinib could be attributed to the pre-incubation condition and/or difference in hepatocytes donors. The pre+co-incubation-IC₅₀ values for dasatinib in human SCH were comparable to that in transporter-expressing cell lines (Pahwa et al., 2017) ((Table 1).

Figure 3.

Inhibitory effects of dasatinib on [³H]rosuvastatin transport under co-incubation and pre+co-incubation in human SCH. [³H]rosuvastatin accumulation (0.5 μM, 2 min) was determined in the presence of dasatinib at the indicated concentration without (co) (A) or with (pre+co) a 1-h pre-incubation with dasatinib. IC₅₀ values were determined by fitting dose-response curves to the data by nonlinear regression analysis as described in the Materials and Methods. Solid (co-incubation) and dashed lines (pre+co-incubation) represent the fitted lines. Data are expressed as percentage of vehicle control in co-incubation (closed circles) and pre+co-incubation (open circles) scenarios. Data are percent of control expressed as mean ± SEM (n=3 hepatocytes donors in triplicate for pre+co-incubation) and mean ± range (n=2 hepatocytes donors in triplicate for co-incubation). A and B were plotted on different scales to indicate that different hepatocytes donors were used in co- and pre+co-incubation (Table S3).

Effects of BSP and rifampicin on accumulation of OATP1B substrates in human SCH.

Co-incubation with BSP (100 μM) and 1 h pre-incubation with BSP (100 μM) followed by washing both significantly decreased accumulation of [³H]CCK-8, [³H]rosuvastatin, and [³H]pitavastatin in human hepatocytes to 0.10 ± 0.01 vs. 0.13 ±0.01, 0.20 ± 0.09 vs. 0.16 ± 0.08, and 0.44 ± 0.05 vs. 0.48 ± 0.13 fold of control, respectively (p<0.05, model estimated mean ± associated standard error) (Fig. 4). Co-incubation following 1 h pre-incubation with rifampicin (pre+co-incubation) both at 50 μM significantly reduced accumulation of [³H]CCK-8 and [³H]rosuvastatin to 0.18 ±0.12 and 0.45 ± 0.15 fold of control, respectively (p<0.05, model estimated mean ± associated standard error) (Fig. 5).

R-value prediction of OATP-mediated DDIs of EVR, SIR, and dasatinib.

R-values for EVR, SIR, and dasatinib using OATP1B substrates rosuvastatin, pitavastatin, or CCK-8 in primary human hepatocytes are summarized in Table 1. Under co-incubation conditions, the R-values for EVR, SIR, and dasatinib using either substrate were less than the US FDA cut-off value of 1.1 in human hepatocytes or transporter-expressing cell lines. Under pre+co-incubation conditions, for dasatinib, R-values from both cell line and human SCH were greater than the US FDA cut-off value of 1.1, consistently predicting OATP-mediated DDIs. For EVR, R values from cell lines were 1.23 and were lowered to near 1.1 (1.08–1.09) in human SCH. For SIR, R values from either cell types were less than the cut-off values of 1.1.

DISCUSSION

Active transport processes are responsible for approximately 90% of total hepatic uptake clearance of rosuvastatin, as estimated from a suspension of human hepatocytes comparing rosuvastatin uptake between with and without rifamipicin treatment (Bi et al., 2013). BSP is an inhibitor of OATP1B1, OATP1B3, OATP2B1, and NTCP, with IC₅₀ or K_i values ranging 0.08–7.3 μM (Kim et al., 1999; Tirona et al., 2003; Annaert et al., 2010). BSP 100 μM was used in a previous study to inhibit OATP-mediated transport of rosuvastatin in human hepatocytes (Kock et al., 2013). In the current study, uptake of rosuvastatin human hepatocytes was inhibited by ~84% by BSP (100 μM) co-incubation (range 77–92% inhibition) (Fig. 4 B) and inhibited by ~55% by rifampicin pre+co-incubation (range 27–78% inhibition) (Fig. 5 B). The slightly lesser inhibition using rifampicin as inhibitor may be due to variability in transporter activity among hepatocyte donors. Overall, our data are comparable to a previous report of the primary contribution of active transport to the overall uptake of [³H]rosuvastatin in human hepatocytes (Bi et al., 2013). The pre-incubation effects on [³H]pitavastatin accumulation were less prominent than on [³H]rosuvastatin or [³H]CCK-8 using EVR, SIR, and BSP as the inhibitor (Fig. 1 E and 2 E and Fig. 4). In cynomolgus monkey hepatocytes, uptake clearance attributed to passive diffusion was greater for pitavastatin than for rosuvastatin (20% vs. 4%) (Ufuk et al., 2018). An incomplete inhibition of pitavastatin by OATP inhibitors was noted in cynomolgus monkey hepatocytes (Ufuk et al., 2018). Our finding of incomplete inhibition of [³H]pitavastatin uptake in human hepatocytes is also likely due to potential higher passive diffusion of pitavastatin than rosuvastatin.

OATP1B1 contributed primarily to pitavastatin uptake into hepatocytes (Hirano et al., 2004). OATP1B1, NTCP, and OATP1B3 contribute to ~50%, 35%, and 15% of active uptake of rosuvastatin into human hepatocytes, respectively (Ho et al., 2006; Kitamura et al., 2008). CCK-8 is a specific substrate of hepatic uptake transporter OATP1B3 (Ismair et al., 2001). Using CCK-8 as substrate, we were able to specifically dissect the inhibitor pre-incubation effects on IC₅₀ values against OATP1B3 in human hepatocytes. In human SCH, pre-incubation with EVR and SIR reduced IC₅₀ values against OATP1B3 by 3.6 (p=0.02) and 7.6 fold (p=0.03), respectively (Table 1). Using rosuvastatin as the substrate, of which OATP1B1 contributes the most to active hepatic uptake among basolateral uptake transporters, pre-incubation with EVR and SIR reduced IC₅₀ values against rosuvastatin by 2.8 (p=0.0007) and 2.5 (p=0.041) fold, respectively. This is the first indication that pre-incubation with OATP1B1 and OATP1B3 inhibitors potentiates their inhibitory effects against OATP1B in physiologically relevant human hepatocytes. It appears that rosuvastatin is a more sensitive substrate than pitavastatin when used to determine the inhibitor-preincubation effects in human SCH, as the EVR- or SIR-preincubation did not significantly affect IC₅₀ values against pitavastatin (Table 1). Such difference between rosuvastatin and pitavastatin may be due to higher variability of co-incubation IC₅₀ values among hepatocytes donors for EVR and/or incomplete inhibition of pitavastatin uptake in human hepatocytes.

The maximum total plasma concentrations of EVR and SIR at the inlet to the liver (I_{in,max,plasma}) are 0.17 and 0.03 μM, respectively (Supplement Table S2). Pre-incubation with EVR or SIR was conducted in hepatocyte culture media supplemented with ITS⁺ premix, which contains proteins, including bovin serum albumin. Hence, the pre-incubation concentrations of 0.2 and 0.05 μM in the current studies (Fig. 2 A, C and E) were considered within the range of clinically relevant concentrations for EVR and SIR, respectively. Pre-incubation with EVR for 1 h at concentrations as low as 0.2 μM significantly decreased accumulation of [³H]-CCK-8 and [³H]-rosuvastatin (Fig. 1 A and C). These pre-incubation effects are consistent with our previous report that pre-incubation of transporter-expressing cell lines in culture medium containing 0.2 μM EVR significantly reduced OATP1B1- and OATP1B3-mediated transport (Farasyn et al., 2019), suggesting that the pre-incubation effects of EVR on OATP1B substrate transport are likely to occur in vivo in the liver. Pre-incubation with SIR significantly reduced [³H]-CCK-8 and [³H]-rosuvastatin accumulation in human hepatocytes (Fig. 2 A and C) and OATP1B1- and OATP1B3-mediated substrate transport in cell lines (Farasyn et al., 2019). However, this reduction occurred at a total concentration (0.6 μM) much greater than the physiologically relevant concentration of 0.03 μM, suggesting that the pre-incubation effect for SIR is unlikely to occur in vivo.

Using the same mTOR inhibitor and substrate pair, we compared the pre-incubation effects on inhibitor IC₅₀ values between human hepatocytes in the current study (Table 1) and in the OATP1B1- or OATP1B3-expressing cell lines we published previously (Farasyn et al., 2019). Using [³H]rosuvastatin as the substrate, fold reductions in IC₅₀ values resulting from inhibitor pre-incubation were greater in OATP1B1-expressing cell lines than in human hepatocytes (8.3 vs. 2.8 fold) for EVR, while reductions were similar in these two systems for SIR against [³H]-rosuvastatin (1.8 vs. 2.5 fold). Using [³H]CCK-8 as the substrate, the pre-incubation effects on fold reduction in IC₅₀ values in OATP1B3-expressing cell lines and human hepatocytes were comparable (2.7 vs. 3.6 fold) for EVR, while the fold reductions were much greater in human hepatocytes than in OATP1B3-expressing cell lines for SIR (7.6 vs. 2.7 fold). Based on these findings, the transporter-expressing cell lines do not always yield greater pre-incubation effects on IC₅₀ reduction than do primary hepatocytes.

CsA has a greater pre-incubation effect in IC₅₀ reduction in cynomolgus monkey (c) OATP1B1- and cOATP1B3-expressing cell lines (Takahashi et al., 2016) than in cynomolgus monkey hepatocytes (up to 23 fold vs. 3.7 fold) (Takahashi et al., 2016; Ufuk et al., 2018). Pre-incubation with CsA in OATP1B1- and OATP1B3-expressing cell lines using various substrates caused up to 22-fold reduction in IC₅₀ values. The pre-incubation effects of CsA on its IC₅₀ values against OATP substrate in human hepatocytes have not been determined. Future studies to explore the possibility to extrapolate the OATP-mediated DDI potential using CsA as the calibrating compound from IC₅₀ data in human hepatocytes are warranted.

For dasatinib, the current studies used only a clinically relevant drug substrate, [³H]rosuvastatin, as the probe, since rosuvastatin appears to more sensitive than [³H]pitavastatin when assessing the inhibitor-preincubation effects based on data from EVR, SIR and BSP (Figs. 1, 2 and 4). The pre-incubation effects of dasatinib on [³H]pitavastatin and [³H]CCK-8, hepatic uptake of which is mediated primarily by OATP1B1 and OATP1B3, respectively, merit evaluation in future studies. BSP (Fig. 4) and rifampicin (Fig. 5), two potent inhibitors of OATP1B1 and OATP1B3, were used as positive controls to show the maximum inhibition of uptake of OATP1B substrate in hepatocytes. Since the inhibitory effects of BSP between co-incubation and pre+co-incubation were similar for each substrate determined (Fig. 4), we chose to use only one condition, pre+co-incubation, for rifampicin. Whether rifampicin has a similar inhibitory effects between co-incubation and pre+co-incubation on hepatic uptake of OATP1B substrate remains undetermined and merits future investigation.

We published previously that in transporter-expressing cell lines, the pre-incubation effects on both OATP1B1- and OATP1B3-mediated transport are similar among 10, 30, and 60 min (for EVR and SIR) (Farasyn et al., 2019) and between 30 and 60 min (for dasatinib) (Pahwa et al., 2017). Based on these studies, we arbitrarily chose 1 h as the pre-incubation period in the current study. It remains unknown whether different pre-incubation times affect their inhibitory effects in human hepatocytes. Regarding the potential underlying mechanism of the potentiation effects of inhibitor-preincubation on OATP1B1- and OATP1B3-mediated transport, a previous publication suggested that inhibition of OATP1Bs from the intracellular side by inhibitors may contribute to the pre-incubation effect (Shitara and Sugiyama, 2017). We have not measured the intracellular concentration of EVR, SIR, or dasatinib after 1 hour of pre-incubation in human hepatocytes. To the best of our knowledge, accumulation of these three drugs in human hepatocytes has not been reported. It is not known whether cellular accumulation of these inhibitors correlated with their inhibition potency against OATP1Bs following inhibitor preincubation. We previously reported that in transporter-expressing cell lines, the pharmacological effect of EVR and SIR as mTOR kinase inhibitors does not appear to be involved in the reduced OATP1B1- and OATP1B3-mediated transport after their pre-incubation ((Farasyn et al., 2019). Dasatinib is a multi-kinase inhibitor, and OATP1B1 and OATP1B3 are phosphorylated proteins (Powell et al., 2014; Crowe et al., 2019). Whether the pharmacological effects of EVR, SIR, and dasatinib as kinase inhibitors are involved in the pre-incubation effects in human hepatocytes warrants characterization in future studies.

In conclusion, the current study is the first report that pre-incubation with inhibitors of OATP1B1 and OATP1B3 reduces IC₅₀ values against OATP1B substrates in physiologically relevant primary human hepatocytes, supporting the rationale for the current FDA guidance recommending the inhibitor pre-incubation step. Because transporter-expressing cell lines seems to yield pre+co-IC₅₀ values that are lower than or comparable to those in human SCH, the pre+co-IC₅₀ values against OATP1B in transporter-expressing cell lines may be used to mitigate false-negative prediction of OATP-mediated DDIs.

Abbreviations:

AUC

area under the plasma concentration-time curve

CsA

cyclosporine A

DDI

drug-drug interaction

DMEM

Dulbecco’s Modified Eagle’s Medium

DMSO

dimethyl sulfoxide

EVR

everolimus

FBS

fetal bovine serum

HBSS

Hanks’ Balanced Salt Solution

IC₅₀

inhibitor concentration producing 50% inhibition

K_m

Michaelis constant

K_i

inhibition constant

mTOR

mammalian target of rapamycin

OATP

organic anion-transporting polypeptide

SIR

sirolimus