G45 and V386 in Transmembrane Domains 1 and 8 are Critical for the Activation of OATP1B3-Mediated E17βG Uptake by Clotrimazole

Abstract

Organic anion transporting polypeptides (OATPs) 1B1 and 1B3 are two highly homologous transport proteins. However, OATP1B1- and 1B3-mediated estradiol-17β-glucuronide (E17βG) uptake can be differentially affected by clotrimazole. In this study, by functional characterization on chimeric transporters and single mutants, we find that G45 in transmembrane domain 1 (TM1) and V386 in TM8 are critical for the activation of OATP1B3-mediated E17βG uptake by clotrimazole. However, the effect of clotrimazole on the function of OATP1B3 is substrate-dependent as clotrimazole does not stimulate OATP1B3-mediated uptake of 4’,5’-dibromofluorescein (DBF) and rosuvastatin. In addition, clotrimazole is not transported by OATP1B3, but it can efficiently permeate the plasma membrane due to its lipophilic properties. Homology modeling and molecular docking indicate that E17βG binds in a substrate binding pocket of OATP1B3 through hydrogen bonding and hydrophobic interactions, among which its sterol scaffold forms hydrophobic contacts with V386. In addition, a flexible glycine residue at position 45 is essential for the activation of OATP1B3. Finally, clotrimazole is predicted to bind at an allosteric site, which mainly consists of hydrophobic residues located at the cytoplasmic halves of TMs 4, 5, 10, and 11.

Graphical Abstract

INTRODUCTION

Solute carriers (SLC) are a large superfamily of transport proteins, currently including 65 families with over 400 members.^1,2 The SLCO family is comprised of the organic anion transporting polypeptides (OATPs), which include 11 human members and are divided into 6 subfamilies based on amino acid identity.¹ Among them, OATP1B1 and 1B3 are specifically expressed in the liver under normal physiological conditions, mediating the uptake of endogenous substances and various drugs from blood into liver cells.³

OATP1B1 and 1B3 share 87% amino acid sequence similarity and possess 12 transmembrane domains (TMs).^4,5 Although they are highly homologous, the function of OATP1B1 and 1B3 can be differentially regulated by specific modulators. For instance, epigallocatechin gallate (EGCG) and quercetin 3-O-α-L-arabinopyranosyl(1 → 2)-α-L-rhamnopyranoside stimulate OATP1B3’s but inhibit OATP1B1’s transport of estrone-3-sulfate (E3S);^6,7 clotrimazole and 5-O-propylquercetin stimulate OATP1B3’s but inhibit OATP1B1’s transport of estradiol-17β-glucuronide (E17βG).^8,9 On the contrary, dioscin inhibits OATP1B3’s but stimulates OATP1B1’s transport of steviol glucuronide.¹⁰ On the basis of these results, it is conceivable that some TMs and amino acid residues play critical roles in the activation/inhibition of OATPs by modulators.

Accordingly, we have investigated structural domains and amino acid residues that are crucial for the activation of OATP1B3-mediated E3S uptake by EGCG in previous studies.^11,12 By functional analysis of a series of chimeras between OATP1B3 and 1B1, we successfully identified that TMs 1, 8, and 10 are critical for the activation of OATP1B3 by EGCG.^11,12 With site-directed mutagenesis, we further pinpointed that G45, V386 and F555 are the most important amino acid residues in TMs 1, 8, and 10, respectively, for the activation of OATP1B3.^11,12 By using similar strategies as well as molecular modeling, in the present study we have investigated the underlying molecular mechanism for the activation of OATP1B3-mediated E17βG uptake by clotrimazole.

MATERIALS AND METHODS

Chemicals.

Atorvastatin (≥98%), clotrimazole (≥98%), 4’,5’-dibromofluorescein (DBF) (≥95%), and tolbutamide (≥97%) were purchased from Sigma-Aldrich (St. Louis, MO, USA). Estradiol-17β-glucuronide (E17βG) (≥95%) and formic acid (≥98%) were from Aladdin (Shanghai, China). Rosuvastatin (≥98%) was from Adamas (Shanghai, China). HPLC-grade acetonitrile and methanol were from Merck (Darmstadt, Germany). High glucose Dulbecco’s Modified Eagle’s Medium (DMEM) was from ThermoFisher Biochemical (Beijing, China). Transfection reagent ExFect was purchased from Vazyme (Nanjing, China). Antibodies for detecting the six-His tag and the Na⁺/K⁺-ATPase α subunit were purchased from Proteintech (66005-1-Ig) and Abcam (ab76020), respectively. Other reagents were obtained from the same companies as previously indicated.¹²

Construction of Chimeras and Mutants of OATP1B3.

Chimeras and single-point mutants of OATP1B3 were constructed as previously described.^4,11,12 All constructs were verified by DNA sequencing.

Protein Expression in HEK293T Cells.

Human embryonic kidney 293T (HEK293T) cells were cultured in high glucose DMEM and transfected as previously described.¹³ Twenty-four hours later, transfected cells were used for expression and functional studies.

Transporter Surface Expression Assay.

Transporter surface expression was analyzed by cell surface biotinylation and immunoblot analysis with procedures as described previously.^12-14 In brief, plasma membrane proteins of transfected HEK293T cells were labeled with sulfo-N-hydroxysuccinimide-SS-biotin and isolated by streptavidin-agarose beads. Isolated surface proteins were then subjected to immunoblot analysis. OATPs and Na⁺/K⁺-ATPase were detected with anti-His tag and anti-Na⁺/K⁺-ATPase α subunit antibodies, respectively.

Transporter Functional Assay.

Transporter function was analyzed by a 1-min uptake assay with procedures as described previously.^4,11 DBF was quantified by its fluorescence at 485/528 nm.¹⁵ E17βG, clotrimazole and rosuvastatin were quantified by liquid chromatography-tandem mass spectrometry (LC-MS/MS). An Agela Venusil C₁₈ column (2.1 mm × 50 mm, 5 μm) was used for chromatographic separation and column temperature was set to be 40 °C. The total analysis time was 6 min.

For E17βG analysis, the flow rate was set to be 0.4 mL/min and the mobile phase was composed of 0.1% formic acid aqueous solution (A) and methanol with 0.1% formic acid (B). The elution gradient was as follows: 0-0.5 min, 20% B; 0.5-1.5 min, 95% B; 1.5-4.5 min, 95% B; 4.5-5.0 min, 20% B; and 5.0-6.0 min, 20% B. Mass spectrometry was operated in multiple reaction monitoring (MRM) mode with negative electrospray ionization (ESI). The ion transitions for E17βG and the internal standard (IS) tolbutamide were m/z 447.2 → 271.0, and m/z 269.1 → 169.1, respectively.

For clotrimazole analysis, the flow rate was 0.4 mL/min and the mobile phase was composed of 0.1% formic acid aqueous solution (A) and acetonitrile with 0.1% formic acid (B). The elution gradient was as follows: 0-1.0 min, 20% B; 1.0-3.0 min, 90% B; 3.0-4.5 min, 90% B; 4.5-4.7 min, 20% B; and 4.7-6.0 min, 20% B. Mass spectrometry was operated in MRM with positive ESI. The ion transitions for clotrimazole and tolbutamide (IS) were m/z 277.2 → 165.2 and m/z 271.2 → 172.2, respectively.

For rosuvastatin analysis, the flow rate was 0.3 mL/min and the mobile phase consisted of 0.1% formic acid aqueous solution (A) and methanol with 0.1% formic acid (B). The elution gradient was as follows: 0-0.5 min, 15% B; 0.5-2.5 min, 95% B; 2.5-4.5 min, 95% B; 4.5-4.6 min, 15% B; and 4.6-6 min, 15% B. Mass spectrometry was operated in MRM with positive ESI. The ion transitions for rosuvastatin and atorvastatin (IS) were m/z 482.2 → 258.1 and m/z 559.4 → 440.0, respectively.

Homology Modeling, Allosteric Site Prediction, and Molecular Docking.

First, a structural model for OATP1B3 in substrate bound state was constructed based on the cryo-EM structure of human OATP1B1 in complex with E3S (PDB entry: 8HND)¹⁶ by homology modeling incorporated in SWISS-MODEL (https://swissmodel.expasy.org/).¹⁷ The global and per-residue model quality was assessed using the QMEAN scoring function.¹⁸ The stereochemistry and geometry of the model was evaluated by MolProbity.¹⁹ Next, PASSer (https://passer.smu.edu) was used to identify potential allosteric sites in the structural model of OATP1B3 with Ensemble method.²⁰ Finally, E17βG and clotrimazole were respectively docked into the substrate binding and allosteric sites of OATP1B3 by AutoDock 4.2²¹ and a model for OATP1B3/E17βG/clotrimazole complex was generated. The interactions between E17βG/clotrimazole and OATP1B3 were analyzed by Ligplot.²²

Data Analysis.

Statistics were performed with Prism 8 (GraphPad Software, La Jolla, CA, USA). Differences between two groups were analyzed by two-tailed unpaired Student’s t test. Differences among three or more groups were analyzed by one-way analysis of variance (ANOVA) followed by Dunnett’s test. p < 0.05 was considered statistically significant.

RESULTS AND DISCUSSION

Clotrimazole, an imidazole derivative and antifungal agent, is an inhibitor of fungal cytochrome P450 51A1 (CYP51A1) as well as of the calcium pump, the Na⁺/K⁺-ATPase, the H⁺/K⁺-ATPase and the multidrug resistance-associated protein (MRP1).^23-27 In addition, clotrimazole is a substrate for Pdr5p, a yeast ortholog of human P-gp.²⁸ Our previous study showed that clotrimazole inhibits OATP1B1-mediated (40-80%) but stimulates OATP1B3-mediated (1.5-2 folds) E17βG uptake.⁸ In the present study, the exact molecular mechanism for the activation of OATP1B3-mediated E17βG uptake by clotrimazole has been elucidated by functional characterization of OATP1B3 chimeras/mutants and molecular modeling.

Effect of Clotrimazole on E17βG Uptake Mediated by OATP1B1, 1B3, and Chimeras 1B3-T1~T12.

As mentioned above, OATP1B3 is stimulated but 1B1 is inhibited by clotrimazole, and both of them comprise 12 TMs. To find out which TM(s) is(are) important for the activation of OATP1B3, we utilized the previously published 12 chimeric transporters of OATP1B3 (namely chimeras 1B3-T1~T12) where individual TMs of OATP1B3 were replaced by the respective TMs of OATP1B1.¹¹

First, OATP1B1, OATP1B3, and Chimeras 1B3-T1~T12 were transiently transfected and expressed in HEK293T cells. Surface biotinylation and immunoblot analysis showed that all of them can be successfully expressed on the plasma membrane, although 1B3-T3 and 1B3-T5 exhibited much lower surface expression than OATP1B3 (Figure 1A). As the function of each transporter will be checked by comparing its uptake for E17βG in the absence and presence of clotrimazole, the decrease or increase of surface expression of a transporter should not affect the outcome of its activation or inhibition.

Figure 1.

Surface expression of OATP1B1, 1B3, and the twelve OATP1B3-derived chimeric transporters and effect of clotrimazole on their transport of E17βG. (A) Surface expression of OATP1B1, 1B3, and the twelve chimeric transporters derived from OATP1B3. OATP proteins expressed on the cell surface were isolated by surface biotinylation and detected by immunoblot analysis with an anti-His tag antibody. The plasma membrane marker Na⁺/K⁺-ATPase α subunit was used as the protein loading control. (B) Effect of clotrimazole on E17βG uptake mediated by OATP1B1, 1B3, and twelve 1B3-derived chimeric transporters. Uptake of 5 μM E17βG in the absence and presence of 50 μM clotrimazole was measured at 37 °C for 1 min with empty vector- and OATP-transfected cells. The net uptake was obtained by subtracting the uptake of empty vector-transfected cells from the uptake of OATP-transfected cells. Data are presented as mean ± SD (n = 3). *p < 0.05, absence vs presence of clotrimazole, two-tailed unpaired Student’s t test.

Functional assays with transiently expressed OATP1B1 and 1B3 in HEK293T cells showed that OATP1B3-mediated uptake of E17βG is stimulated by clotrimazole, while OATP1B1-mediated uptake of E17βG is inhibited by clotrimazole (Figure 1B). These results are consistent with our previous observations, which were obtained with CHO cells stably expressing OATP1B1 and 1B3,⁸ and indicate that OATP1B1 and 1B3 have the same function in transient and stable expression systems.

We then examined the uptake of E17βG by chimeras 1B3-T1~T12 in the absence and presence of 50 μM clotrimazole, at which concentration clotrimazole exhibiting the largest stimulating effect on OATP1B3-mediated E17βG uptake.⁸ As shown in Figure 1B, the activation of chimeras 1B3-T1 and 1B3-T8 by clotrimazole was abolished, while the remaining 10 chimeras still could be significantly activated by clotrimazole. These results indicate that T1 and T8 are the most important structural segments for the activation of OATP1B3-mediated E17βG uptake by clotrimazole.

TM1 and TM8 Are Critical for the Activation of OATP1B3-Mediated E17βG Uptake by Clotrimazole.

T1 (aa 1-48) consists of the N-terminus (NT) (aa 1-24) and TM1 (aa 25-48), while T8 (aa 359-394) consists of the extracellular loop 4 (EL4) (aa 359-375) and TM8 (aa 376-394). As T1 and T8 are critical parts for OATP1B3’s activation, we further constructed chimeras 1B3-NT, 1B3-TM1, 1B3-EL4, and 1B3-TM8 to determine the key structural domains more precisely. As shown in Figure 2A, chimeras 1B3-NT and 1B3-TM1 are derived from 1B3-T1 by replacing NT and TM1 of OATP1B3 with the corresponding regions of OATP1B1, respectively. Chimeras 1B3-EL4 and 1B3-TM8 are derived from 1B3-T8 by replacing EL4 and TM8 of OATP1B3 with the corresponding regions of OATP1B1, respectively. Surface biotinylation and immunoblot assay showed that chimeras 1B3-NT, 1B3-TM1, 1B3-EL4, and 1B3-TM8 were properly expressed on the cell surface (Figure 2B).

Figure 2.

Schematic diagram of chimeras 1B3-NT, TM1, EL4, and TM8 derived from 1B3-T1 and T8, and their surface expression and activation by clotrimazole. (A) Chimeras 1B3-NT and 1B3-TM1 were derived from 1B3-T1 and constructed by replacing the N-terminus (NT) (aa 1-24) and transmembrane domain 1 (TM1) (aa 25-48) of OATP1B3 with the corresponding regions of OATP1B1, respectively. Chimeras 1B3-EL4 and 1B3-TM8 were derived from 1B3-T8 and constructed by replacing extracellular loop 4 (EL4) (aa 359-375) and TM8 (aa 376-394) of OATP1B3 with the corresponding regions of OATP1B1, respectively. (B) Surface expression of chimeras 1B3-NT, TM1, EL4 and TM8, together with OATP1B3, 1B3-T1, and T8. Chimeras expressed on the cell surface were isolated by surface biotinylation and detected by immunoblot analysis with an anti-His tag antibody. The plasma membrane marker protein Na⁺/K⁺ ATPase was used as a loading control. (C) Effect of clotrimazole on the uptake of E17βG mediated by chimeras 1B3-NT, TM1, EL4 and TM8. Uptake of 5 μM E17βG in the absence and presence of 50 μM clotrimazole was measured at 37 °C for 1 min with empty vector- and OATP-transfected cells. The net uptake was obtained by subtracting the uptake of empty vector-transfected cells from the uptake of OATP-transfected cells. Data are presented as mean ± SD (n = 3). *p < 0.05, absence vs presence of clotrimazole, two-tailed unpaired Student’s t test.

Functional experiments showed that chimeras 1B3-NT and 1B3-EL4 could still be activated by clotrimazole (Figure 2C). However, 1B3-TM1 and 1B3-TM8 could not be activated anymore. These results indicate that TM1 and TM8 are the key transmembrane domains for the activation of OATP1B3-mediated E17βG uptake by clotrimazole.

G45 in TM1 and V386 in TM8 Are the Key Amino Acid Residues for the Activation of OATP1B3-Mediated E17βG Uptake by Clotrimazole.

Amino acid sequence alignment showed that in both, TM1 and TM8, 5 amino acid residues were different between OATP1B3 and 1B1.^11,12 To determine which amino acid residues in TM1 and TM8 are important for the activation of OATP1B3 by clotrimazole, we constructed 5 single mutants for each of TM1 and TM8, by replacing each residue in OATP1B3 with its corresponding residue in OATP1B1. As shown in Figure 3A, all these 10 mutants had comparable surface expression levels as wild-type OATP1B3.

Figure 3.

Surface expression and activation of single mutants in TM1 and TM8. (A) Surface expression of five mutants (1B3)F27L(1B1), F36L, Y38F, A42T and G45A in TM1, and five mutants F376I, I380V, T385I, V386F and T388S in TM8. (B) Effect of 50 μM clotrimazole on the uptake of 5 μM E17βG mediated by OATP1B3 mutants in TM1 and TM8. Uptake of 5 βM E17βG in the absence and presence of 50 μM clotrimazole was measured at 37 °C for 1 min with empty vector- and OATP-transfected cells. The net uptake was obtained by subtracting the uptake of empty vector-transfected cells from the uptake of OATP-transfected cells. Data are presented as mean ± SD (n = 3). *p < 0.05, absence vs presence of clotrimazole, two-tailed unpaired Student’s t test.

Functional studies showed that mutants G45A in TM1 and V386F in TM8 almost completely lost their activation by clotrimazole, while the remaining 8 single mutants still could be activated (Figure 3B). Although the difference between G45A-mediated uptake in the absence and presence of clotrimazole was statistically significant, it was very small. These results demonstrate that G45 in TM1 and V386 in TM8 are the key amino acid residues for the activation of OATP1B3-mediated E17βG uptake by clotrimazole. Our previous studies showed that G45 and V386 are also important for the activation of OATP1B3-mediated E3S uptake by EGCG.^11,12 However, F555 in TM10 is important for OATP1B3’s activation by EGCG¹¹ but not critical for OATP1B3’s activation by clotrimazole (Figure 1B). These results indicate that the binding sites for E17βG/clotrimazole and E3S/EGCG partially overlap but are not identical.

Effect of Clotrimazole on the Kinetics of E17βG Uptake Mediated by OATP1B3, 1B3-TM1, 1B3-TM8, 1B3-G45A, and 1B3-V386F.

To further characterize the crucial roles of TM1, TM8, G45A, and V386F in the activation of OATP1B3, kinetic studies on E17βG uptake mediated by OATP1B3, 1B3-TM1, 1B3-TM8, 1B3-G45A and 1B3-V386F in the absence and presence of clotrimazole have been carried out. The kinetic parameters are summarized in Table 1.

Table 1.

Effect of clotrimazole on the kinetics of E17βG uptake mediated by OATP1B3, 1B3-TM1, 1B3-TM8, 1B3-G45A, and 1B3-V386F^a

	without clotrimazole		with 50 μM clotrimazole
	Km (μM)	Vmax (pmol/mg protein/min)	Km (μM)	Vmax (pmol/mg protein/min)
OATP1B3	10.7 ± 1.4	179.7 ± 8.4	6.0 ± 0.7b	250.2 ± 7.8b
1B3-TM1	8.3 ± 1.0	137.2 ± 5.5	6.5 ± 0.9	132.7 ± 5.4
1B3-TM8	1.8 ± 0.4	39.2 ± 1.8	1.3 ± 0.3	54.6 ± 2.3b
1B3-G45A	8.1 ± 1.3	187.2 ± 9.5	4.4 ± 0.6b	182.2 ± 6.0
1B3-V386F	15.7 ± 2.0	278.4 ± 13.6	8.0 ± 0.7b	162.1 ± 4.7b

^aUptake of increasing concentrations of E17βG in the absence and presence of 50 μM clotrimazole was measured at 37 °C for 1 min with empty vector and OATP-transfected cells. The net uptake was obtained by subtracting the uptake of empty vector-transfected cells from the uptake of OATP-transfected cells. Net uptake was then fit to the Michaelis–Menten equation to obtain K_m and V_max values.

^bp < 0.05 vs corresponding value in the group without clotrimazole, two-tailed unpaired Student’s t test.

For OATP1B3, clotrimazole significantly increased both its binding affinity and transport capacity for E17βG as its K_m value decreased from 10.7 ± 1.4 to 6.0 ± 0.7 μM and the V_max increased from 179.7 ± 8.4 to 250.2 ± 7.8 pmol/mg protein/min in the presence of clotrimazole (Table 1). Therefore, clotrimazole stimulates OATP1B3-mediated E17βG uptake at both low and high substrate concentrations. This is different from the effect of EGCG on OATP1B3-mediated E3S uptake, in which EGCG showed a stimulating effect at low substrate concentrations but an inhibitory effect at high substrate concentrations.^6,11,12 In addition, clotrimazole has no significant effect on OATP1B3-mediated E3S uptake, while EGCG has no effect on OATP1B3-mediated E17βG uptake.^6,8 Therefore, the mode of OATP1B3 activation by these two modulators is substrate-dependent.

For chimeras 1B3-TM1 and 1B3-TM8, their respective K_m values in the presence of clotrimazole were not significantly different from those in the absence of clotrimazole (Table 1). In addition, the V_max values of 1B3-TM1 in the presence and absence of clotrimazole were comparable. However, the V_max value of 1B3-TM8 in the presence of clotrimazole was significantly higher than that of 1B3-TM8 in the absence of clotrimazole. Therefore, at high substrate concentrations, clotrimazole showed a small stimulation on E17βG uptake mediated by 1B3-TM8.

For mutant 1B3-G45A, the K_m decreased while V_max was unchanged in the presence of clotrimazole as compared to the corresponding values without clotrimazole (Table 1). Therefore, at low substrate concentrations, clotrimazole showed a small stimulating effect on E17βG uptake mediated by 1B3-G45A, which is consistent with the result shown in Figure 3B. For mutant 1B3-V386F, K_m and V_max values both decreased in the presence of clotrimazole. Therefore, at high substrate concentrations, uptake of E17βG mediated by 1B3-V386F was slightly inhibited by clotrimazole.

Effect of Clotrimazole on the Function of Additional Mutants for G45 and V386.

To further investigate the roles of different amino acid residues at positions 45 and 386 in the activation of OATP1B3, the function of additional mutants, including 1B3-G45S (polar), G45D (negatively charged), G45K (positively charged), and G45F (hydrophobic) at position 45 and 1B3-V386I, V386A, and V386G (non-polar residues with different side chain sizes) at position 386 was measured. As shown in Figure 4A, all additional mutants at both positions were properly expressed on the cell surface.

Figure 4.

Surface expression and activation of additional mutants for G45 and V386 of OATP1B3. (A) Surface expression of additional mutants for G45 (G45S, G45D, G45K, G45F) and V386 (V386I, V386A, V386G). (B) Effect of 50 μM clotrimazole on the uptake of 5 μM E17βG mediated by additional mutants for G45 and V386. Uptake of 5 μM E17βG in the absence and presence of 50 μM clotrimazole was measured at 37 °C for 1 min with empty vector- and OATP-transfected cells. The net uptake was obtained by subtracting the uptake of empty vector-transfected cells from the uptake of OATP-transfected cells. Data are presented as mean ± SD (n = 3). *p < 0.05, absence vs presence of clotrimazole, two-tailed unpaired Student’s t test.

At position 45, all of four additional mutants (G45S, G45D, G45K, and G45F) were unable to rescue OATP1B3’s activation (Figure 4B), indicating that a flexible glycine residue at this position is essential for the activation of OATP1B3-mediated E17βG uptake by clotrimazole. This phenomenon was also observed for the activation of OATP1B3-mediated E3S uptake by EGCG.¹¹ Another interesting observation at position 45 is that when glycine was replaced by a negatively or positively charged residue (G45D and G45K), transport for E17βG in the absence of clotrimazole was significantly reduced as compared to that of OATP1B3 (Figure 4B). On the contrary, when G45 was replaced by a phenylalanine residue (G45F), the transport of E17βG was significantly increased. These results indicate that G45 plays quite different roles in substrate translocation and allosteric modulation.

Similar to position 45, at position 386, all three additional mutants (V386I, V386A and V386G) could not rescue OATP1B3’s activation (Figure 4B), indicating that a hydrophobic amino acid residue with a suitable side chain size is strictly required at position 386 for the activation of OATP1B3-mediated E17βG uptake by clotrimazole. This is different from the activation of OATP1B3-mediated E3S uptake by EGCG, in which smaller amino acids such as alanine and glycine at position 386 could restore OATP1B3’s activation by EGCG.¹²

As clotrimazole showed opposite effects on the function of OATP1B3 and G45F (Figure 4B), we further examined OATP1B3- and G45F-mediated uptake of E17βG in the presence of increasing concentrations of clotrimazole. As shown in Figure 5A, clotrimazole stimulated OATP1B3-mediated E17βG uptake at all concentrations tested with a maximal effect at 50 μM, which is consistent with our previous observation.⁸ On the contrary, clotrimazole inhibited G45F-mediated E17βG uptake in a concentration dependent manner (Figure 5B). These results further demonstrated that G45 is important for OATP1B3’s allosteric regulation by clotrimazole.

Figure 5.

Effect of clotrimazole on (A) OATP1B3- and (B) G45F-mediated E17βG uptake. Uptake of 5 μM E17βG in the presence of increasing concentrations of clotrimazole was measured at 37 °C for 1 min with empty vector- and OATP-transfected cells. The net uptake was obtained by subtracting the uptake of empty vector-transfected cells from the uptake of OATP-transfected cells. Data are presented as mean ± SD (n = 3).

Clotrimazole is Not Transported by OATP1B3 and 1B1.

Because clotrimazole stimulated OATP1B3 but inhibited OATP1B1 mediated uptake of E17βG, we wondered whether it would be a substrate for OATP1B1 and 1B3. As shown in Figure 6, 50 μM clotrimazole was not transported by OATP1B3 and 1B1. Similarly, mutants 1B3-G45A and V386F did not transport clotrimazole. However, clotrimazole is a substrate of the yeast ABC transporter Pdr5p,²⁸ and a previous study reported that clotrimazole is a substrate of OATP1B1 but not OATP1B3, based on a competitive counterflow assay, where clotrimazole stimulated the efflux of radiolabeled E17βG.²⁹

Figure 6.

Uptake of clotrimazole by empty vector, OATP1B1, OATP1B3, 1B3-G45A, and 1B3-V386F. Uptake of 50 μM clotrimazole in the presence of 5 μM E17βG was measured at 37 °C for 1 min with empty vector- and OATP-transfected cells. Data are presented as mean ± SD (n = 3). *p < 0.05 vs empty vector, one-way ANOVA followed by Dunnett’s test.

The competitive counterflow assay is more sensitive for very hydrophobic compounds when direct uptake of a substrate would be masked by unspecific binding to the membrane or by simple diffusion. Accordingly, this different result could be explained by the hydrophobicity of clotrimazole, which has a calculated logP of 5.47 by Molinspiration (https://www.molinspiration.com, Slovensky Grob, Slovakia). Thus, it can efficiently bind to and potentially permeate the cell membrane at the relatively high concentration used via simple diffusion, leading to a much higher amount of clotrimazole after a 1-min incubation compared to E17βG (Figures 1 and 6). Therefore, it seems likely that the stimulating effect of clotrimazole on OATP1B3 is not due to its extracellular to intracellular electrochemical gradient but more likely due to an allosteric effect of clotrimazole by binding to OATP1B3. Furthermore, it seems feasible that the action site of clotrimazole is not necessary at the extracellular portion of OATP1B3.

OATP1B3-H115A Mutant Can Still be Activated by Clotrimazole.

Cryo-EM structures of human OATP1B1 and OATP1B3 have been reported recently.^16,30 The structure of OATP1B3 reveals a binding pocket for bicarbonate which may serve as a counter-ion in the transport cycle.³¹ This pocket is lined by two conserved tryptophan residues (W258 and W259) in the OATP signature motif and a histidine residue H115.³⁰ It has been observed that the binding of bicarbonate into the pocket couples with allosteric changes of helix bundles in OATP1B3.³⁰ It was suggested that H115 may play a critical role in the binding and release of bicarbonate in OATP1B3 and thus would be important for the pH-dependent transport of OATPs.^30,32

To examine whether OATP1B3’s activation by clotrimazole is related to H115, we mutated H115 to an alanine residue and measured uptake of E17βG in the absence and presence of clotrimazole. As shown in Figure 7, H115A-mediated E17βG uptake could still be activated by clotrimazole, indicating that the action of clotrimazole on OATP1B3 is not via its interaction with the bicarbonate binding site. As clotrimazole is a lipophilic molecule, we speculated that clotrimazole would bind in a relatively hydrophobic site to allosterically regulate OATP1B3’s transport of E17βG.

Figure 7.

Effect of clotrimazole on E17βG uptake mediated by mutant H115A. Uptake of 5 μM E17βG in the absence and presence of 50 μM clotrimazole was measured at 37 °C for 1 min with empty vector- and OATP-transfected cells. The net uptake was obtained by subtracting the uptake of empty vector-transfected cells from the uptake of OATP-transfected cells. Data are presented as mean ± SD (n = 3). *p < 0.05, absence vs presence of clotrimazole, two-tailed unpaired Student’s t test.

Effect of Clotrimazole on the Uptake of Other Substrates Mediated by OATP1B3 and Chimeras 1B3-NT, 1B3-TM1, 1B3-EL4 and 1B3-TM8.

Given that clotrimazole can stimulate OATP1B3-mediated E17βG uptake, we wanted to know whether clotrimazole would have a similar effect on OATP1B3-mediated uptake of other substrates. Therefore, we tested two additional substrates, namely DBF and rosuvastatin. As shown in Figure 8, OATP1B3-mediated uptake of DBF and rosuvastatin was inhibited by clotrimazole. DBF uptake mediated by chimeras 1B3-NT, 1B3-TM1, 1B3-EL4, and 1B3-TM8 was also inhibited by clotrimazole. Rosuvastatin uptake mediated by chimera 1B3-TM1 was inhibited by clotrimazole, whereas its uptake mediated by chimeras 1B3-NT, 1B3-EL4, and 1B3-TM8 was not significantly affected by clotrimazole. These results further demonstrate the substrate-dependent effect of modulators on OATP1B3 as clotrimazole has opposite effects on OATP1B3-mediated E17βG and DBF/rosuvastatin uptake. In addition, it was reported that clotrimazole had no significant effect on OATP1B3-mediated E3S uptake.⁸

Figure 8.

Effect of clotrimazole on the uptake of DBF and rosuvastatin mediated by OATP1B3, chimeras 1B3-NT, TM1, EL4 and TM8. Uptake of 5 μM (A) DBF and (B) rosuvastatin in the absence and presence of 50 μM clotrimazole was measured at 37 °C for 1 min with empty vector- and OATP-transfected cells. The net uptake was obtained by subtracting the uptake of empty vector-transfected cells from the uptake of OATP-transfected cells. Data are presented as mean ± SD (n = 3). *p < 0.05, absence vs presence of clotrimazole, two-tailed unpaired Student’s t test.

Structural Modeling of the OATP1B3/E17βG/Clotrimazole Complex.

As the experimental structures for human OATP1B1 and OATP1B3 are currently available,^16,30 it is possible to try to elucidate the mechanism of OATP1B3 activation from a structural perspective by homology modeling and molecular docking. Although the structure of OATP1B3 has been solved (PDB entry: 8PG0), it is in the apo state, and direct docking into the structure is not possible.³⁰ Therefore, we first constructed a structural model for OATP1B3 in a substrate-bound state based on the cryo-EM structure of human OATP1B1 in complex with E3S (PDB entry: 8HND)¹⁶ by SWISS-MODEL.¹⁷ The sequence identity between OATP1B3 and the template is 79.74%. Model quality was assessed by the QMEAN scoring function and MolProbity.^18,19 The GMQE model quality score and the MolProbity score were 0.71 and 1.36, respectively. These data indicate that the obtained OATP1B3 model is structurally reasonable. In addition, our structural model is very close to the experimental apo structure of OATP1B3 (PDB entry: 8PG0) with a Cα RMSD of 1.414 Å between the two structures (Figure 9A), indicating that our model is reliable.

Figure 9.

(A) Superimposition of the modeled OATP1B3 structure in a substrate-bound state (in green) and its experimental structure in the apo state (in red, PDB entry: 8PG0). The two structures were aligned by their Cα atoms. (B) Allosteric sites in the OATP1B3 model predicted by PASSer (https://passer.smu.edu). Two sites were predicted with high probability. The yellow site overlaps with the binding site of E3S in OATP1B1 and thus is the substrate binding site. The red site is a potential allosteric site.

Because clotrimazole may allosterically modulate the function of OATP1B3, we employed the allosteric site prediction program PASSer to identify potential allosteric sites in OATP1B3.²⁰ As shown in Figure 9B, two sites were identified with high probability. The yellow site located in the middle of TMs 7, 8, 9, 10 and 12, overlapped with the binding site of E3S in OATP1B1³⁰ and we considered this the substrate binding site. The red site located at the cytoplasmic halves of TMs 2, 4, 5, 10 and 11, is a potential allosteric site. As the intracellular or intramembranous concentration of clotrimazole is likely high during the uptake process (Figure 6), the location of its binding site close to the cytoplasmic border of the membrane is possible. However, G45 and V386 are located close to the substrate binding site and not at the allosteric site.

To potentially elucidate the interaction modes of E17βG/clotrimazole with OATP1B3, we developed a complex model for OATP1B3/E17βG/clotrimazole by docking E17βG and clotrimazole into the substrate binding and allosteric sites, respectively. As shown in Figure 10A, E17βG binds in the substrate binding site with its negatively charged glucuronide group at the apex of the cavity, while its sterol scaffold points towards the center of the transporter. The glucuronide group forms an electrostatic interaction with R633 and hydrogen bonds with S355, Q541, and N544 (Figure 10B). This is similar to the interactions between the sulfate group of E3S and OATP1B1 as reported.³⁰ The sterol scaffold of E17βG forms extensive hydrophobic interactions with hydrophobic residues of OATP1B3, which include V386 (Figure 10B). This indicates that V386 may be important in the activation of OATP1B3-mediated E17βG uptake through hydrophobic contacts with the substrate. However, direct contact of G45 with E17βG was not observed. As TM1 is located in the central area of the substrate translocation pathway (Figure 10A), we speculate that the flexibility of TM1, which may be affected by G45, could be important for the activation of OATP1B3.

Figure 10.

(A) Structural model of OATP1B3/E17βG/clotrimazole complex. E17βG/clotrimazole are displayed in CPK style and their carbon atoms are colored in yellow and cyan, respectively. G45 and V386 are shown in ball-and-stick and colored in green. (B) Interactions between E17βG and OATP1B3. Dashed lines represent hydrogen bonds and spikes represent amino acid residues involved in hydrophobic contacts with E17βG. (C) Interactions between clotrimazole and OATP1B3.

Clotrimazole occupies the identified allosteric site without direct contact to E17βG (Figure 10A). Interaction analysis showed that the imidazole ring of clotrimazole forms a hydrogen bond with N213, and its three phenyl rings form hydrophobic interactions with V189, I193, L209, I556, and M572 (Figure 10C). Therefore, the binding site of clotrimazole is mainly composed of hydrophobic residues, which is compatible with the lipophilic property of clotrimazole, as mentioned above. Nevertheless, whether these residues are important for the activation of OATP1B3 remains to be verified experimentally. In addition, it needs to be emphasized that although we simultaneously docked E17βG and clotrimazole into the OATP1B3 model, in reality, it is also possible that E17βG and clotrimazole bind alternately to OATP1B3 during the transport cycles.

CONCLUSIONS

In summary, the present study investigated the molecular mechanism of clotrimazole’s activation on OATP1B3-mediated E17βG uptake by experimental studies and molecular modeling. The functional studies showed that G45 in TM1 and V386 in TM8 are critical for the activation of OATP1B3-mediated E17βG uptake by clotrimazole. However, the effect of clotrimazole on the function of OATP1B3 is substrate-dependent, as clotrimazole does not stimulate OATP1B3-mediated DBF and rosuvastatin uptake. In addition, clotrimazole is not transported by OATP1B3, but it can permeate the cell membrane by simple diffusion. Molecular modeling indicated that V386 interacts with the sterol scaffold of E17βG through hydrophobic interactions, and a nonpolar residue with a suitable size is strictly required at this position. A flexible glycine residue at position 45 in TM1 is essential for the activation of OATP1B3. As TM1 is located in the center of the substrate translocation pathway, the flexibility of TM1, which may be highly affected by G45, could be important for the activation of OATP1B3. Finally, clotrimazole is predicted to bind in an allosteric site by forming a hydrogen bond and hydrophobic interactions with amino acid residues located at the cytoplasmic halves of TMs 4, 5, 10, and 11.

ABBREVIATIONS

DBF

4’,5’-dibromofluorescein

DMEM

Dulbecco’s Modified Eagle’s Medium

E17βG

estradiol-17β-glucuronide

E3S

estrone-3-sulfate

EGCG

epigallocatechin gallate

EL

extracellular loop

ESI

electrospray ionization

HEK293T

human embryonic kidney 293T cells

IS

internal standard

LC-MS/MS

liquid chromatography-tandem mass spectrometry

MRM

multiple reaction monitoring

NT

N-terminus

OATP

organic anion transporting polypeptide

SD

standard deviation

SLC

solute carrier

TM

transmembrane domain