Comprehensive Substrate Characterization of 22 Antituberculosis Drugs for Multiple Solute Carrier (SLC) Uptake Transporters In Vitro

The substrate potentials of antituberculosis drugs on solute carrier (SLC) transporters are not well characterized to date, despite a well-established understanding of their drug dispositions and pharmacokinetics. In this study, we investigated comprehensively the substrate potentials of the 22 currently available antituberculosis drugs for SLC family transporter-mediated uptake, using Xenopus laevis oocytes and stably transfected HEK-293 cells in vitro.

ABSTRACT

The substrate potentials of antituberculosis drugs on solute carrier (SLC) transporters are not well characterized to date, despite a well-established understanding of their drug dispositions and pharmacokinetics. In this study, we investigated comprehensively the substrate potentials of the 22 currently available antituberculosis drugs for SLC family transporter-mediated uptake, using Xenopus laevis oocytes and stably transfected HEK-293 cells in vitro. The result suggested that ethambutol, isoniazid, amoxicillin, and prothionamide act as novel substrates for the SLC transporters. In addition, in the presence of representative transporter inhibitors, the uptake of the antituberculosis drugs was markedly decreased compared with the uptake in the absence of inhibitor, suggesting involvement of the corresponding transporters. A cellular uptake study was performed, and the K_m values of ethambutol were found to be 526.1 ± 15.6, 212.0 ± 20.1, 336.8 ± 20.1, and 455.0 ± 28 μM for organic cation transporter 1 (OCT1), OCT2, OCTN1, and OCTN2, respectively. Similarly, the K_m of prothionamide was 805.8 ± 23.4 μM for OCT1, while the K_m values of isoniazid and amoxicillin for organic anion transporter 3 (OAT3) were 233.7 ± 14.1 and 161.4 ± 10.6 μM, respectively. The estimated in vivo drug-drug interaction indexes from in vitro transporter inhibition kinetics for verapamil, probenecid, and ibuprofen against ethambutol, prothionamide, isoniazid, and amoxicillin were found to show potential for clinical drug interactions. In conclusion, this is the first study that demonstrated 22 antituberculosis drug interactions with transporters. This study will be helpful for mechanistic understanding of the disposition, drug-drug interactions, and pharmacokinetics of these antituberculosis drugs.

INTRODUCTION

The World Health Organization has recommended four first-line antituberculosis (TB) drugs, isoniazid, rifampin, ethambutol, and pyrazinamide, as initial therapies for TB. Ten percent of TB patients have also been diagnosed with diabetes, and among the 9 million TB patients diagnosed in 2016, 40% of human immunodeficiency virus deaths were found to be due to TB (1–3 and http://www.who.int/en/news-room/fact-sheets/detail/tuberculosis#hiv). In addition to first-line anti-TB drugs, there are second-line drugs for the treatment of multidrug-resistant TB, such as linezolid, para-aminosalicylic acid (PAS), aminoglycoside, ethionamide, prothionamide, and cycloserine (4). Due to these multidrug regimens, several drug-drug interactions (DDIs) and unwanted pharmacokinetics/pharmacodynamics (PK/PD) effects have been reported in the literature, with case reports describing adverse events, nephrotoxicity, drug-induced liver injury, gastrointestinal disruption, teratogenicity, ocular toxicity, and neurotoxicity associated with the use of isoniazid, linezolid, rifampin, and ethambutol during anti-TB therapy (5–9).

To date, it has been well established that membrane transporters are significantly associated with both the clinical efficacy and the toxic events attributed to pharmaceuticals (10). It is also believed that membrane transporters are a major determinant in the PK/PD and safety profiles of therapeutic drug substances (11). The cellular uptake/influx of drugs is facilitated by the solute carrier (SLC) family of membrane transporters, in which over 400 human SLC transporter genes have been reported as a superfamily and classified into 46 subfamilies (12). There is a distinct pattern of transporter expression and localization in organs/tissues, such as the small intestine, liver, and kidneys, which is important for drug disposition (13). Within the SLC superfamily, SLC22 and SLCO family members, namely, the organic anion-transporting polypeptides (OATPs) OATP1B1, OATP2B1, and OATP1B3; organic cation transporters (OCTs) OCT1 and OCT2; and the organic anion transporters (OATs) OAT1 and OAT3, play major roles in the disposition of common therapeutic drugs (14). Drug-metabolizing enzymes are well known for their involvement in genotype-dependent DDIs, whereas very few genotype-dependent transporter-mediated DDIs have been reported to date. For example, cyclosporine has been reported to cause an increased area under the concentration-time curve (AUC) of repaglinide, a substrate of cytochrome P450 3A4 (CYP3A4) and OATP1B1; the extent of this interaction was 42% lower in healthy volunteers with the SLCO1B1 521TC genotype than in individuals with the high-activity 521TT reference genotype (15). Another genetic polymorphism in the OCT2 transporter, the SLC22A2 c.808G→T mutation, causes a significant reduction in metformin renal clearance compared with that of the wild-type OCT2. This indicates that transporter drug efficacy can be clinically affected by single nucleotide polymorphisms to produce interindividual variations in the response or PK/PD profile (16). Furthermore, transporters and enzymes both play a role in the PK/PD of a drug; however, in regard to personalized pharmacotherapy, transporters can play vital roles along with metabolizing enzymes (17).

Rifampin is a first-line drug of choice to treat TB and has a strong inhibitory potential against OATP-mediated uptake (18). We recently reported the inhibitory effects of anti-TB drugs on OATP-mediated uptake (19), which was consistent with a previous report that showed similar effects of ethambutol on OCT-mediated uptake (20). In addition, rifampin is a substrate of P-glycoprotein (21), OATP1B1 (22), and OATP1B3 (23) membrane transporters, and several DDI studies have reported that competitive inhibition of OATP1B1/1B3 by rifampin may lead to reduced hepatic uptake of substrates and may increase plasma exposure (24). Other known interactions between anti-TB drugs and membrane transporters include amoxicillin for peptide transporters (PepT1 and PepT2) (25); cycloserine for proton-coupled amino acid transporter 1 (PAT1) (26); moxifloxacin for OATP1A2 (27); ciprofloxacin for OATP1A2, OATP1A5 (28), OAT3 (29), and breast cancer resistance protein (BCRP) (30); and ethambutol for P-glycoprotein (31).

To date, characterization of the transporter-mediated uptake/efflux of the 22 currently marketed anti-TB drugs has not been fully realized. Therefore, in this study, we aimed to fully characterize the SLC transporter-mediated uptake/efflux of the 22 anti-TB drugs. Using Xenopus laevis oocytes and further confirmation with stably transfected HEK-293 cells, net cellular transport kinetic studies were performed. We believe that this study will be a helpful resource to understand the details of SLC membrane transporter interactions with anti-TB drugs, which may translate into clinical study and practice.

RESULTS

Characterization of cellular uptake using radiolabeled prototype substrates.

The uptake of the prototype substrates [³H]estron-3-sulfate ([³H]ES) for OATP1B1, OATP2B1, and OAT3; [³H]estradiol 17β-d-glucuronide ([³H]E2G) for OATP1B3; p-[³H]aminohippurate ([³H]PAH) for OAT1; [³H]methylphenylpyridinium ([³H]MPP⁺) for OCT1 and OCT2; [¹⁴C]tetraethylammonium ([¹⁴C]TEA) for OCTN1; and [³H]l-carnitine for OCTN2 are shown in Fig. S1 in the supplemental material. The cellular uptake levels of the prototype substrates were 31-, 42-, 26-, 16-, 13-, 16-, 12-, 19-, and 14-fold higher in HEK-OATP1B1, HEK-OATP2B1, HEK-OATP1B3, HEK-OAT1, HEK-OAT3, HEK-OCT1, HEK-OCT2, HEK-OCTN1, and HEK-OCTN2 cells, respectively, than those in the mock-transfected cells at 5 min of incubation, and they were at equilibrium (Fig. S1). Furthermore, the uptake kinetics of each prototype substrate (data not shown) estimated using nonlinear kinetics indicated that the transfected HEK-OATP1B1, HEK-OATP2B1, HEK-OATP1B3, HEK-OAT1, HEK-OAT3, HEK-OCT1, HEK-OCT2, HEK-OCTN1, and HEK-OCTN2 cells were suitable for use in uptake experiments. We also assessed the time-dependent uptake of the prototype substrates in transfected cells, showing linear uptake into transporter-overexpressing cells with respect to time (data not shown), which has been reported previously (32).

Characterization of uptake of anti-TB drugs by SLC transporters.

Twenty-two anti-TB drugs were tested using X. laevis oocytes for uptake screening. The results showed that ethambutol (Fig. 1D and E), PAS, isoniazid (Fig. 1G), amoxicillin (Fig. 1G), and prothionamide (Fig. 1D) intracellular concentration were markedly higher than that of the control (no transporter cRNA injected), identifying them as novel substrates for multiple SLC uptake transporters (Fig. 1). We then confirmed the transport properties of the anti-TB drugs that had been identified from the preliminary screening (i.e., ethambutol, amoxicillin, isoniazid, and prothionamide) using HEK-transfected cells (Fig. 2). The intracellular concentrations of the anti-TB drugs were markedly higher in the HEK-transporter-expressing cells than in the mock cells. The level of uptake was consistent with that in the preliminary screening with X. laevis oocytes. Furthermore, using stably transfected HEK-293 cells, we determined the uptake kinetics values (K_m and V_max) for the drugs. The K_m values of ethambutol were 526.1 ± 15.6, 212.0 ± 20.1, 336.8 ± 20.1, and 455.0 ± 28 μM for OCT1 (Fig. 3A), OCT2 (Fig. 3B), OCTN1 (Fig. 3C), and OCTN2 (Fig. 3D), respectively. The K_m value of prothionamide for OCT1 was 805.8 ± 23.4 μM (Fig. 3E). The K_m values of isoniazid and amoxicillin for OAT3 were 233.7 ± 14.1 μM (Fig. 3F) and 161.4 ± 10.6 μM (Fig. 3G), respectively. These data are summarized in Table 1.

FIG 1.

Uptake characterization of 22 anti-TB drugs. The uptake of 22 anti-TB drugs (100 μM) into Xenopus laevis oocytes expressed OATP1B1 (A), OATP2B1 (B), OATP1B3 (C), OCT1 (D), OCT2 (E), OAT1 (F), and OAT3 (G), respectively, normalized with % uptake into mock cells (control). Data represent as the mean ± SD of the results from triplicate experiments. Asterisks indicates significant differences compared to control uptake (*, P < 0.05; **, P < 0.01, ***, P < 0.001).

FIG 2.

Cellular uptake screening of anti-TB drugs for the SLC transporters. Ethambutol (20 μM), prothionamide (20 μM), isoniazid (20 μM), and amoxicillin (20 μM) uptake into the stably transfected HEK cells with SLC transporters and mock cells (control). Data are presented as the mean ± SD of the results from three or more independent experiments. Significant differences compared with the percent uptake for the control (mock cells) are indicated (***, P < 0.001).

FIG 3.

Uptake kinetics determined by stably transfected cells HEK-Mock, HEK-OCT1, HEK-OCT2, HEK-OAT1 and HEK-OAT3 for anti-TB drugs. Intracellular uptake kinetics of the ethambutol for OCT1 (A), OCT2 (B), OCTN1 (C), and OCTN2 (D), and prothionamide for OCT1 (E), isoniazid for OAT3 (F), and amoxicillin for OAT3 (G) derived from in vitro experiment with a concentration range were used and normalized with the control. Data are presented as the means ± SD of the results from three or more independent experiments.

TABLE 1.

Uptake kinetic parameters determined by stably transfected cells HEK cells for ethambutol, prothionamide, isoniazid and amoxicillin^a

Substrate	Transporter	Km (μM)b	Vmax (pmol/mg protein/min)c	CLint (μl/mg protein/min)d
Ethambutol	OCT1	526.0 ± 15.6	2,232	4.2
	OCT2	212.0 ± 20.1	3,289	15.5
	OCTN1	336.8 ± 20.1	3,174.0	9.5
	OCTN2	455.1 ± 28.0	3,993.1	8.7
Prothionamide	OCT1	805.8 ± 23.4	551.5	0.6
Isoniazid	OAT3	233.7 ± 26.6	746.5	3.2
Amoxicillin	OAT3	161.4 ± 16.2	209.6	1.3

^aIntracellular uptake kinetics was derived from in vitro experiment with a concentration gradient were used and normalized with control.

^bK_m, Michaelis-Menten constant. Data are presented as the means ± SDs from the results from three or more independent experiments.

^cV_max, maximum rate of uptake.

^dCL_int, intrinsic clearance per unit of time.

Effect of representative inhibitors on uptake of anti-TB drugs via SLC transporters.

We evaluated OCT1-, OCT2-, OCTN1-, OCTN2-, and OAT3-mediated ethambutol, prothionamide, isoniazid, and amoxicillin uptake in the presence of representative inhibitors. As expected, verapamil, ibuprofen, and probenecid, known inhibitors of these transporters, strongly attenuated ethambutol (Fig. 4A and B), prothionamide (Fig. 4C), isoniazid (Fig. 4D), and amoxicillin (Fig. 4E) uptake. Uptake transport was decreased to a greater extent in the presence of the corresponding inhibitor of the transporter. This result suggests that ethambutol, prothionamide, isoniazid, and amoxicillin are novel substrates of the tested SLC transporters. Furthermore, we evaluated the concentration-dependent inhibitory potential of the inhibitors on ethambutol, prothionamide, isoniazid, and amoxicillin uptake in vitro via OCT1-, OCT2-, OCTN1-, OCTN2-, and OAT3-mediated transport using stably transfected HEK-293 cells (Fig. 5). The estimated half-maximal inhibitory concentration (IC₅₀) values of verapamil on ethambutol uptake were 12.0, 21.4, 2.0, and 9.6 μM for OCTN1-, OCTN2-, OCT1-, and OCT2-mediated uptake, respectively (Fig. 5A). Similarly, the IC₅₀ value of verapamil was 3.9 μM for OCT1-mediated prothionamide uptake (Fig. 5B). In addition, the very well-known OAT3 inhibitor, probenecid, showed potent inhibition, with estimated IC₅₀ values of 14.4 and 12.5 μM for OAT3-mediated amoxicillin (Fig. 5C) and isoniazid (Fig. 5B) uptake, respectively. We also determined the IC₅₀ value of ibuprofen on isoniazid uptake to be 6.1 μM (Fig. 5D).

FIG 4.

Effect of representative inhibitors on the uptake of anti-TB drugs in stably transfected HEK-293 cells. Positive-control inhibitors for anti-TB drug uptake inhibition were verapamil at 10 μM for OCT1, 30 μM for OCT2, 40 μM for OCTN1, and 40 μM for OCTN2, and probenecid at 30 μM for OAT1 and mefenamic acid at 5 μM for OAT3, respectively. Data are presented as the mean ± SD of the results from three or more independent experiments. Significant differences compared with the percent uptake for the control (no inhibitor) are indicated (**, P < 0.01; ***, P < 0.001).

FIG 5.

Evaluation of the concentration-dependent inhibition of transporter-mediated anti-TB drugs uptake by the representative inhibitors. The kinetics of anti-TB drug uptake inhibition of ethambutol, prothionamide, isoniazid, and amoxicillin uptake into stably transfected HEK cells were estimated. The estimated IC₅₀ was calculated by nonlinear kinetics using WinNonlin software (version 5.1). Data are presented as the mean ± SD of the results from three or more independent experiments and are the level of uptake as a percentage of that for the control (no inhibitor).

Prediction of DDI index for transporter-mediated uptake inhibition.

We estimated the DDI indexes for the anti-TB drugs with the inhibitor of each transporter, using clinical plasma concentrations (Table S2) and a static model-based approach to assess the clinical relevance of this study. The DDI index values were found to be negligible to significant compared to the cutoff values (>0.1) in the FDA guidelines for in vivo DDIs. Among the inhibitors, probenecid showed a strong DDI index, with values of 12.5 and 14.4 for OAT3-mediated isoniazid and amoxicillin uptake inhibition, respectively (Table 2). Similarly, ibuprofen was found to be a potential DDI with isoniazid, with a higher value than the cutoff value. Similarly, as verapamil strongly inhibits OCT-mediated uptake in a concentration-dependent manner, the DDI indexes for OCT1-mediated ethambutol and prothionamide uptake were estimated to be 2.0 and 3.9, respectively, suggesting that they have in vivo DDI potential (Table 2). In contrast, verapamil was found to have negligible DDI potential to inhibit OCT2-, OCTN1-, and OCTN2-mediated ethambutol uptake, with DDI index values lower than the cutoff values (Table 2).

TABLE 2.

DDI indexes estimated from in vitro OCTN1-, OCTN2-, OCT1-, OCT2-, and OAT3-mediated ethambutol, amoxicillin, isoniazid, and prothionamide uptake inhibition kinetics by the inhibitors^a

Inhibitor	Substrate	Transporter	IC50 (μM)	[I]/IC50
Verapamil	Ethambutol	OCTN1	12.0 ± 3.2	0.04
		OCT1	2.0 ± 0.2	0.28b
		OCTN2	21.4 ± 5.2	0.02
		OCT2	9.6 ± 1.6	0.05
Verapamil	Prothionamide	OCT1	3.9 ± 0.4	0.14b
Probenecid	Isoniazid	OAT3	12.5 ± 0.88	19.5b
Ibuprofen	Isoniazid	OAT3	6.1 ± 0.8	16.7b
Probenecid	Amoxicillin	OAT3	14.4 ± 1.6	0.43b

^aThe DDI index values were determined using the inhibition constant (half-maximal inhibitory concentration [IC₅₀]) with the maximum plasma concentration (C_max; bound plus unbound) ([I]_max) of the inhibitor drugs following the regulatory guidelines described in the text. DDI index values represent means ± SDs obtained from the inhibition constants from three or more independent experiments.

^bThe result is significant according to FDA guidance for prototypical substrates, in which the DDI index values are greater than the corresponding cutoff values recommended by the regulatory authorities for an [I]/IC₅₀ value of ≥0.1. The cutoff value is expressed as the value according to the upper limit of the equivalence range suggested by FDA.

DISCUSSION

This study revealed the substrate potential of several anti-TB drugs on SLC transporter-mediated active uptake and their drug interaction possibility. Throughout this characterization study, we identified several anti-TB drugs (both first- and second-line) as a substrates of SLC uptake transporters (Fig. 1). Namely, ethambutol, amoxicillin, isoniazid, and prothionamide were the most key findings for multiple SLC uptake transporters (Table 1). The transport characteristics of these drugs were found to be close to similar to those we recently reported for PAS (33). Interestingly, ethambutol uptake was mediated by multiple cation transporters, such as OCT1, OCT2, OCTN1, and OCTN2 (Fig. 2). The transport characteristics of ethambutol for OCT1 and OCT2 were consistent with a recent report, except for the multidrug and toxic compound extrusion (MATE) transporters (34). However, our study revealed that MATE transporters are unlikely to be involved with ethambutol transport and elimination from the kidney. In our comprehensive screening, ethambutol was found to be a novel substrate of OCTN1 and OCTN2 (Fig. 4A). Previously, we also reported ethambutol inhibitory interactions with SLC transporters and drug-drug interaction potentials in vitro (35). Along with glomerular filtration, tubular secretion is also involved in ethambutol elimination from the kidney, as it has greater secretory clearance (36). From this study finding, we hypothesize that OCTN1 and OCTN2 play major roles in the secretory clearance of ethambutol. Similarly, prothionamide, amoxicillin, and isoniazid were found to be substrates of OCT1 and OAT3 in both X. laevis oocytes and HEK-293 cells (Fig. 2). As expected, the uptake of these drugs was decreased more than 85% by the representative inhibitors of the transporters, indicating that the involvement of these transporters is significant to the disposition and pharmacokinetics of the drugs. Amoxicillin was found to be a novel substrate of OAT3, and its uptake was greatly inhibited by probenecid (Fig. 4E), which has also been observed in clinical study. A similar interaction was observed for a fluoroquinolone analogue, ciprofloxacin, which was previously reported for OAT3-mediated active transport and clinical drug interactions with uptake inhibition (29). Similarly, prothionamide was found to be a substrate of OCT1, and verapamil greatly decreased its cellular uptake, identifying it as a novel substrate in this study. However, there are no pharmacokinetic data on the clinical interactions of prothionamide and OCT1 inhibitors, and despite metabolic clearance via bile canaliculi, renal elimination contributes poorly to its elimination. However, the involvement of OCT1 could be important for ethionamide penetration into the lung, because OCT1 is highly expressed in the lung epithelia. Additionally, isoniazid was found to be a substrate of OAT3, and ibuprofen and probenecid both significantly decreased its uptake into the HEK-OAT3 cells (Fig. 4D). Previously, nonsteroidal anti-inflammatory drugs were shown to inhibit intracellular accumulation of methotrexate via inhibition of OAT3-mediated uptake (37). Isoniazid-induced anion gap acidosis has been well reported, indicating the capability of isoniazid to interact with anion transporters to accumulate anion (38). We recently reported the substrate potential of PAS, including potential DDIs with nonsteroidal anti-inflammatory drugs and proton pump inhibitors (33). With the exception of PAS, ethambutol, isoniazid, prothionamide, and amoxicillin, the anti-TB drugs were found to have negligible uptake into transporter-overexpressing cells, indicating that they were not substrates for the tested SLC membrane transporters (Fig. 1). Additionally, the inhibition of specific transporters with potent inhibitors produced significant DDI indexes (Table 2). OCT1 and OAT3 were found to have significant DDIs involving ethambutol, prothionamide, isoniazid, and amoxicillin (Table 2). Probenecid was found to have the highest DDI index values for OAT-mediated amoxicillin and isoniazid uptake inhibition. This predicted DDI index is completely consistent with the known clinical drug interactions between amoxicillin and probenecid. In a clinical study among 10 healthy subjects, the amoxicillin serum concentration was increased almost 2-fold when amoxicillin was coadministered with probenecid; this clinical drug interaction has remained unsubstantiated to date, which can be explained by our novel finding (39).

This in vitro-to-in vivo extrapolation prediction can be helpful to understand complex DDIs and provide data for future mechanistic studies. Despite the SLC transporter-mediated substrate potential of these anti-TB drugs for ATP-binding cassette transporters still being unknown except for a few, this is currently the focus of a comprehensive investigation. Furthermore, with the knowledge that these anti-TB drugs are substrates of OCT and OAT transporters, it will now be necessary to investigate genetic polymorphisms in these transporters and their effects on anti-TB drug disposition. Previously, we published a clinical translational study in which metformin pharmacokinetics was greatly affected by polymorphisms in OCT1 and OCT2 among healthy Koreans (40). In the current study, we investigated only the parent anti-TB drugs, rather than their major metabolites (for those that have significant metabolism), because of simplicity, which can be considered a limitation of this study to clinical applications. In addition, the DDI index estimation considers only a single route of administration, but clinical DDIs can be impacted by several routes of administration. Overall, our study demonstrates a mechanistic view of SLC uptake transporter interactions with the 22 currently marketed anti-TB drugs. However, a number of new anti-TB drugs are currently under development.

To our knowledge, this is the first comprehensive characterization of 22 anti-TB drugs for membrane transport, in which ethambutol, prothionamide, isoniazid, and amoxicillin were revealed to be novel substrates of several SLC uptake transporters. These findings suggest a possible mechanism by which anti-TB drugs may interact with other therapeutic or endogenous compounds via membrane transporters. These in vitro studies could help to mechanistically understand the disposition and pharmacokinetics of anti-TB drugs for future pharmacotherapy against tuberculosis upon validation study.

MATERIALS AND METHODS

Chemicals and reagents used in this study.

The chemicals [³H]-estrone-3-sulfate ([³H]ES, 1.66 terabecquerel [TBq]/mmol), [³H]aminohippuric acid ([³H]PAH, 0.166 TBq/mmol), and [³H]N-methyl-4-phenylpylidinium acetate ([³H]MPP, 82.1 curie [Ci]/mmol) were purchased from PerkinElmer (Waltham, MA, USA). Verapamil (VPL), probenecid (PBA), fetal bovine serum (FBS), penicillin streptomycin (PS) and anti-TB drugs isoniazid (INH), ethambutol (EMB), pyrazinamide (PZA), rifampin (RIF), rifabutin (RFB), amikacin (AMK), kanamycin (KAN), streptomycin (STR), moxifloxacin (MXF), ciprofloxacin (CIP), levofloxacin (LVX), cycloserine (CS), para-aminosalicylic acid (PAS), prothionamide (PRO), ethionamide (ETA), linezolid (LZD), amoxicillin (AMX), clarithromycin (CLR), roxithromycin (RXM), clofazimine (CFZ), bedaquiline (BQN), and rifapentine (RFP), were purchased from Sigma-Aldrich. Other reagents used in this study were purchased from commercial suppliers offering the highest purity.

cRNA synthesis and transport study using Xenopus laevis oocytes.

The cRNA synthesis and uptake experiments were performed as described previously (41). In a brief, the capped OATP, OAT, and OCT cRNAs were synthesized in vitro. X. laevis oocytes were digested in ORII solution containing 82.5 mM NaCl, 2 mM KCl, 1 mM MgCl₂, 5 mM HEPES (pH 7.4), and 1.5 mg/ml collagenase for 90 min at room temperature. After digestion, defolliculated oocytes were injected with 50 ng of the capped cRNA and incubated at 18°C in Barth's solution [88 mM NaCl, 1 mM KCl, 0.33 mM Ca(NO₃)₂, 0.4 mM CaCl₂, 0.8 mM MgSO₂, 2.4 mM NaHCO₃, 10 mM HEPES (pH 7.4)]. After incubation for 2 days, uptake experiments were performed at room temperature in ND96 solution (96 mM NaCl, 2 mM KCl, 1.8 mM CaCl₂, 1 mM MgCl₂, 5 mM HEPES [pH 7.4]). The uptake reaction was initiated by replacing the ND96 solution containing anti-TB drugs in both control oocytes and oocytes expressing uptake transporters (i.e., OATPs, OATs, and OCTs) and terminated by the addition of ice-cold ND96 solution after 30 min for OATPs and OATs and 60 min for OCTs. Followed by three washes, the oocytes were sonicated with 120 μl of 100% acetonitrile two times at 3°C for 5 s and centrifuged at 13,000 × g for 10 min at 4°C. Aliquots of the supernatant were injected into a liquid chromatography-tandem mass spectrometry (LC-MS/MS) system for the detection of the anti-TB drugs.

Cell line preparation and culture for transport experiment.

The human embryonic kidney 293 HEK-293 cells were stably transfected according to a previously reported method (42), and recently, we have published studies using those stable cell lines (19, 35). In brief, the stably transfected HEK-OCT1, HEK-OCT2, HEK-OAT1, HEK-OAT3, HEK-OATP1B1, HEK-OATP2B1, HEK-OATP1B3, HEK-OCTN1, and HEK-OCTN2 cells were grown in tissue culture flasks in Dulbecco's modified Eagle medium supplemented with 10% fetal bovine serum, 1% nonessential amino acids, 2 mM l-glutamine, and 100 U/ml penicillin-streptomycin (Invitrogen) at approximately 37°C in atmosphere supplemented with 5% CO₂. Cells were harvested at confluence using trypsin-EDTA and resuspended in culture medium. This process was repeated as necessary to obtain sufficient cells for each experiment.

Uptake of radiolabeled probe substrate into the transporter-overexpressing cells.

Evaluation of the effect of transporter overexpression on cellular uptake of labeled transporter substrates was performed as described previously (43). To measure the cellular uptake of radiolabeled substrate, HEK-transfected cells were seeded in 24-well culture plates at a cell density of 2 × 10⁵ cells/well and incubated at 37°C in atmosphere supplemented with 5% CO₂ for 1 day. After cells reached 90% confluence, they were washed twice with phosphate-buffered saline (PBS) and incubated in medium containing 45 nM [³H]ES, 59 nM [³H]E2G, 25 nM [³H]MPP, 25 nM [³H]l-carnitine, 45 nM [¹⁴C]TEA, or 89 nM [³H]PAH for 5 min at 37°C. This medium was then aspirated, and cells were washed three times with ice-cold PBS and lysed in 1% Triton X-100 for 30 min at room temperature on a shaker. The uptake of substrate by the cells was quantitated by measuring the radioactivity in the lysate by using a liquid scintillation counter (PerkinElmer).

Uptake characterization of anti-TB drugs into transporter-overexpressing HEK cells.

Similar to prototypical transporter substrate uptake, cells were incubated in medium containing 100 μM anti-TB drugs for 5 min at 37°C, and then the medium was aspirated immediately. Cells were washed three times with ice-cold PBS, and lysates were prepared for LC-MS/MS by adding 100 μl acetonitrile (ACN; 70%) in a 1.5-ml tube, sonicating for 3 to 5 s, and centrifuging at 13,000 rpm for 10 min. An internal standard (IS) for LC-MS/MS was added to each supernatant. Supernatants were vortexed for 3 s and then transferred to LC-MS/MS vials for detection. A 120-μl aliquot of each lysate was analyzed using LC-MS/MS to quantify uptake. Positive controls were used in all experiments.

Evaluation of anti-TB drug uptake inhibition by representative inhibitors.

To characterize inhibitory effect of representative inhibitors on anti-TB drug transport via corresponding transporters, we used verapamil to inhibit OCT1, OCT2, OCTN1, and OCTN2, and probenecid and mefenamic acid to inhibit OAT1 and OAT3, respectively. For this experiment, control and transporter-expressing cells were incubated in medium containing 20 μM ethambutol, 20 μM prothionamide, 20 μM isoniazid, and 20 μM amoxicillin for 5 min at 37°C in the presence of 10 μM verapamil for OCTs, and 30 μM probenecid and 5 μM mefenamic acid for OAT3. Anti-TB drug uptake was then quantitated in LC-MS/MS, as described above. Kinetic studies of these inhibitors were performed using the same methods in the presence of inhibitors spanning a broad concentration range. The half-maximal inhibitory concentration (IC₅₀) of each inhibitor was determined by measuring the percent inhibition of OCT1-, OCT2-, OCTN1-, and OCTN2-mediated ethambutol, OCT1-mediated prothionamide, and OAT3-mediated isoniazid and amoxicillin uptake at a range of inhibitor concentrations. Half-maximal inhibitory concentrations (IC₅₀) were determined using the inhibitory effect model: E = E₀ × 1 − (C/[C + IC₅₀]), where E is the effect, E₀ is the baseline effect, and C is concentration. The values were determined by the use of the WinNonlin software (version 5.1; Pharsight Corp., Mountain View, CA).

Determination of anti-TB drug uptake kinetics into the HEK cells.

For the uptake kinetic study, anti-TB drug uptake into control and transporter-overexpressing cells was performed over a concentration range. Before these experiments, the linearity of cellular uptake over time was individually determined for each cell line (see the supplemental material; some data not shown). The cellular uptake rates were determined by normalization for incubation time and total protein content in the overexpressing cell lines. Net uptake rates were calculated as the difference in the uptake rate of the transfected and mock (without transporter DNA transfection) cells for each individual concentration. To ensure that the uptake inhibition is only by selected transporters, we used mock cells as a control and a positive control for all experiments. Kinetic parameters, K_m and V_max, were calculated using the Michaelis-Menten equation using the Phoenix software package (WinNonlin 8.0; Pharsight, USA); the equation is as follows: V = (V_max × S)/(K_m + S), where V indicates uptake velocity of the probe substrate (picomoles per minute per milligram protein), S indicates the substrate concentration in the experimental buffer (in micromolar), K_m is the Michaelis-Menten constant (in micromolar), and the maximum transport rate is V_max.

We used another equation to calculate kinetic parameters when an additional component was observed at a higher substrate concentration, and these were plotted using SigmaPlot 8.0 (SPSS, Inc., Chicago, IL, USA). The equation is V = (V_max × S)/(K_m + S) + P_dif × S, where P_dif indicates the nonsaturable uptake clearance (in milliliters per minute per milligram of protein).

Analytical methods.

The analysis of 22 anti-TB drugs was performed as described in our previous report, with minor modifications (44). A 6410 LC/MS system (Agilent Technologies, Santa Clara, CA) was used. The separation was performed using an Atlantic aC₁₈ column (2.1 by 50 mm, 3 μm; Phenomenex, Torrance, CA), with a mobile phase consisting of water and acetonitrile (55:45 [vol/vol]) containing 0.1% formic acid. The mobile phase was eluted using an Agilent Technologies 1200 series pump at a flow rate of 0.25 ml/min, and 2 μl was injected. Mass spectra were recorded by electrospray ionization with a positive mode. The turbo ion spray interface was operated at 4,000 V and 300°C. The operating conditions were optimized by flow injection of an analyte and were determined as follows: nebulizing gas flow, 20 lb/in²; curtain gas flow, 10 liters/min; and collision energy, 12 eV. Multiple-reaction-monitoring (MRM) mode using a specific precursor/product ion transition was employed for the quantification. Detection of the ion was performed by monitoring the m/z transition of each drug. The peak areas for all compounds were automatically integrated using MassHunter quantitative analysis (version B.1.4).

DDI index prediction using a static model.

To add clinical value to our study, we calculated the drug-drug interaction (DDI) index according to FDA guidance using maximum plasma concentration (C_max) and maximum unbound plasma concentration (C_max,u) (clinical pharmacokinetic data) (Table S2) for each drug at its IC₅₀ value of the inhibitor drugs in our in vitro study. Briefly, for the OCTs and OATs, the equation for the drug-drug interaction index was C_max/IC₅₀, where C_max,u indicates the maximum unbound concentration of the inhibitor present in the systemic circulation (45, 46).

Statistical analysis.

Two-sided Student's t tests were used to determine the statistical significance of differences between control and test data. The results are expressed as mean ± standard deviation (SD). P values of <0.05 are considered to indicate a statistically significant difference for the control, which was determined using the GraphPad Prism software (version 6; San Diego, CA).