Experimental Procedures

Materials

[³H]Taurocholate (specific activity, 2 Ci/mmol) was obtained from DuPont New England Nuclear (Boston, MA). [³H]U71038 (specific activity, 35 mCi/mg), unlabeled U71038 and [³H]EMD51921 (specific activity, 79 mCi/mg), and unlabeled EMD51921 were obtained from The Upjohn Company (Kalamazoo, MI) and E. Merck (Darmstadt, Germany), respectively. [³H]Propranolol (specific activity, 26.4 Ci/mmol), unlabeled S- and R-enantiomers of propranolol, and its racemate had been previously obtained from ICI Pharmaceuticals (Alderley Edge, UK). [α-³²P]dCTP (specific activity, 3000 Ci/mmol) was purchased from Amersham (Arlington Heights, IL). Angiopeptin was provided by the Henri Beaufour Institute (Washington, DC). Recombinant vaccinia virus containing the T7-RNA polymerase gene (vtf-7) and the pTM1 expression vector were gifts provided by Dr. Bernard Moss (National Institutes of Health, Bethesda, MD). Human liver samples were provided by Dr. F. P. Guengerich at our institution. All other chemical and reagents, unless stated otherwise, were obtained from Sigma-Aldrich Research (St. Louis, MO) and were of the highest grade available.

Interindividual Variability in Hepatic NTCP mRNA Expression.

Human liver samples (Nashville Regional Organ Procurement Agency, Nashville, TN) from eight Caucasian donors (age range, 10–49 years; mean age, 29 years; five females and three males) without any history of underlying liver pathology or prior medications (cause of death: motor vehicle accident, n = 5; subarachnoid hemorrhage, n = 2; and self-inflicted gunshot wound, n = 1) were used in the total RNA extraction using Trizol reagent (Life Technologies, Rockville, MD). A ribonuclease protection assay was carried out on the derived total RNA samples to quantify NTCP mRNA levels. A 203-bp NTCP cDNA for use in the antisense riboprobe synthesis was obtained by PCR of a full-length NTCP cDNA using the oligonucleotide primer 5′-ATTGGCTTTCTGCTGGGTTA-3′ and a reverse primer consisting of the T7-promoter sequence (underlined) and NTCP specific sequence 5′-GGATCCTAATACGACTCACTATAGGGAGGGGAGGGGAAAGAAGAAAAGTGGTC3′. Inclusion of the T7-promoter sequence in the primer bypassed the need for subcloning or linearization. The resulting PCR product (100 ng) was used in the synthesis of T7-RNA polymerase-mediated antisense RNA using the RPA II kit (Ambion, Austin, TX) according to the recommended protocol supplied by the manufacturer. As a reference RNA, cyclophilin mRNA levels were assessed using pTRI-cyclophilin-human cDNA template (Ambion) designed to detect a 103-bp fragment of human cyclophilin, according to the same protocol. Subsequent to RNase digestion, samples were loaded onto a denaturing 4% acrylamide gel for electrophoresis. On the completion of electrophoresis, the gels were dried and overlaid with radiograph film for visualization of the protected bands. For quantification, the density of the NTCP and the cyclophilin protected bands were analyzed using the PhosphorImager 445 SI (Molecular Dynamics, Sunnyvale, CA). The quantity of expressedNTCP mRNA was calculated by normalization of the detectedNTCP density to that of cyclophilin.

Cell Culture and Virus Preparation.

HeLa (American Type Culture Collection) cells were cultured in Dulbecco's modified Eagle's medium supplemented with 5% FBS and 1% penicillin/streptomycin. For preparation of viral stock of VTF7 virus, HeLa cells grown to near confluency in 25-cm tissue culture plates were infected with 1 plaque-forming unit (PFU)/10 cells. After an incubation period of 48 h at 37°C, the infected cells were pelleted, homogenized, and recovered through centrifugation, followed by titering of viral stock as described by Blakely et al. (1991).

NTCP Plasmid Construction.

The full open reading frame of human NTCP cDNA was obtained by polymerase chain reaction (PCR) using a combination of Taq and Deep Vent DNA polymerases from a cDNA library synthesized from human liver mRNA using oligonucleotide primers 5′-ATGGAGGCCCACAACGCGTCT-3′ as the forward and 5′-CTAGGCTGTGCAAGGGGA GCA-3′ as the reverse primers. A single PCR product of expected size was visualized on an ethidium bromide-stained, 2% agarose gel. An aliquot of the PCR product was ligated into the PCR II vector (InVitrogen, Carlsbad, CA). Subsequent to Escherichia coli transformation and growth, the NTCP cDNA was released from the PCRII vector by digestion with the restriction endonuclease EcoRI. The NTCP insert cleaved from the PCR II vector was then recovered through agarose gel electrophoresis and extraction from the gel using the DNA binding resin Qiaex (Qiagen, Chatsworth, CA). The recovered cDNA was then ligated into the pTM1 vector, previously linearized with EcoRI, and dephosphorylated using calf intestinal alkaline phosphatase. After transformation and growth in E. coli, individual colonies containing the pTM1/NTCP construct were identified. A pTM1-NTCP with the NTCP cDNA insert in the sense orientation downstream from the T7-promoter region was fully sequenced using an ABI automated sequencer (Applied Biosystems Inc., Foster City, CA), and found to fully match the published sequence (Hagenbuch and Meier, 1994). This clone was used in the expression studies.

Transport Studies Using Recombinant Vaccinia Virus.

HeLa cells grown in 12-well plates (∼0.8 × 10⁶cells/well) were infected with vaccinia (VTF-7) at a multiplicity of 10 PFU/cell in serum-free OptiMEM I medium (Life Technologies) and allowed to adsorb for 30 min at 37°C. Cells in each well were then transfected using the cationic lipid agent Lipofectin (Life Technologies) with either 1 μg of pTM1 plasmid vector containing the transporter cDNA insert or pTM1 vector lacking any insert. A lipofectin/plasmid DNA ratio of approximately 3:1 yielded best results. After the application of this mixture, the cells were allowed to incubate at 37°C for 16 h. Incubation periods of less than 12 h or more than 20 h resulted in reduced transport activity. On the day of the transport study, the cells were washed with 2 ml of a pH 7.4 buffer (transport buffer A) containing 100 mM NaCl, 2 mM KCl, 1 mM CaCl, 1 mM MgCl, and 10 mM HEPES-tris(hydroxymethyl) aminomethane. Cells were then preincubated for 15 min with 0.3 ml/well of this buffer. Then, 0.1 ml/well of the same buffer containing a 4× cocktail of desired concentrations of labeled taurocholate with or without unlabeled inhibitor/drug was added. Uptake was allowed to occur for a predetermined time at 37°C and then the transport activity was stopped by washing each well with 1 ml of ice-cold buffer (transport buffer A), repeated three times. Subsequently, the cells were lysed by the addition of 0.5 ml of 1% SDS to each well, followed by the measurement of lysate radioactivity with a Rackbeta liquid scintillation counter (Pharmacia-LKB Nuclear, Gaithersburg, MD). Protein concentrations were determined using a Coomassie Protein Assay Reagent (Pierce Chemical Co., Rockford, IL) with BSA as the standard.

Characterization of Transport Kinetics.

To determine the transport kinetics of NTCP-mediated taurocholate uptake, the cellular accumulation of labeled taurocholate over time was assessed by stopping the transport activity at 1, 2, 3, 4, 5, 10, 15, 30, and 60 min after the addition of the substrate. The 1-min time point chosen as the best single time at which the rate of uptake appeared linear. In the subsequent experiments, varying concentrations of unlabeled substrate were also added to the wells. Uptake was allowed to continue for 1 min under the conditions outlined previously. A Michaelis-Menten type nonlinear curve-fitting was carried out using Prism (GraphPAD, San Diego, CA) to obtain the estimate of maximal uptake rate,V_max, and the concentration at which half-maximal uptake occurred, K_m. Inhibition of NTCP-mediated taurocholate transport by drugs was assessed by coincubating a candidate drug inhibitor (100 μM), along with the labeled taurocholate (5 μM) for 30 min. Taurocholate uptake in the presence and the absence of the candidate inhibitor was compared in cells transfected with pTM1-NTCP or only the parental pTM1 plasmid. In addition, selected drugs with inhibitory properties were assessed further at multiple inhibitor concentrations to define the inhibition constant IC₅₀ by using the Hill equation (Prism). All experiments were carried out in triplicate, on at least three differing experiment days. Data are shown as mean ± S.E.

Statistical Analysis.

Mean ± S.E. values were compared using a Student's t test or Mann-WhitneyU test, and a value of P < .05 was taken as the minimum level of significance.

Results

Quantification of hepatic NTCP mRNA expression, using an RNase protection assay on total RNA samples derived from eight disease-free livers obtained from donors of European-American descent (Fig.1), showed that after normalization to their corresponding cyclophilin levels, the estimated levels of NTCP mRNA varied by 40-fold between the liver samples.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Interindividual differences in hepatic NTCP mRNA by an RNase protection assay; 35 μg of human liver total RNA (C) was loaded onto each lane, and the protected bands for NTCP (A) and cyclophilin (B) riboprobes are shown; ∗, female liver samples. Additional demographic data and cause of death are listed inExperimental Procedures.

Recombinant vaccinia-mediated expression of NTCP indicated that uptake of [³H]taurocholate (0.2 μM) was rapid and very efficient relative to cells transfected with plasmids lacking NTCP, which failed to show any significant [³H]taurocholate uptake (Fig.2). Replacement of sodium in the incubation buffer with potassium resulted in a marked (8-fold) reduction in NTCP-mediated taurocholate uptake (Fig. 2); however, a modest residual level of transport activity remained even in the absence of sodium at this tracer concentration. Interestingly, the sodium-independent component was no longer discernible at a higher substrate concentration (5 μM taurocholate, data not shown). The uptake of [³H]taurocholate, in the presence of buffer sodium, at differing concentrations of unlabeled taurocholate showed that NTCP-mediated uptake of taurocholate was a saturable process that could be described using the Michaelis-Menten equation. The apparent K_m value was 7.9 ± 2.0 μM, whereas V_max was 49 ± 3 pmol/mg protein/min (Fig. 2).

• Download figure
• Open in new tab
• Download powerpoint

Figure 2

Uptake of [³H]taurocholate after recombinant vaccinia virus-induced expression of NTCP in HeLa cells. A, HeLa cells transfected with pTM1/NTCP construct (●) or pTM1 alone (○) incubated with [³H]taurocholate (0.2 μM) in the presence buffer sodium and the effect of buffer NaCl substitution with KCl on pTM1/NTCP-mediated [³H]taurocholate (■). B, transport kinetics of NTCP-mediated taurocholate uptake. Data are mean ± S.E. of at least three separate experiments.

A large number of bile acids and drugs were assessed for their effect on expressed NTCP activity. As expected, bile acids (100 μM) such as cholic acid, glycocholic acid, chenodeoxycholic acid, deoxycholic acid, and ursodeoxycholic acid were inhibitors of [³H]taurocholate (5 μM) uptake (Fig.3). Bromosulfophthalein (100 μM), an organic anion, was also an inhibitor of taurocholate uptake (Fig. 3). Further concentration-dependent inhibition studies with cholic acid, ursodeoxycholic acid, and bromosulfophthalein revealed that all were potent inhibitors (Fig. 4) with IC₅₀ values of 3.6 to 7.3 μM (Table1).

• Download figure
• Open in new tab
• Download powerpoint

Figure 3

Effect of bile acids and xenobiotics on the modulation of NTCP-mediated taurocholate uptake transport expressed in HeLa cells. The uptake of [³H]taurocholate (5 μM), in the presence of various bile acids, bromosulfophthalein, drugs, oligopeptides, and dipeptide/tripeptide agents added to the transport buffer at 100 μM concentrations (unless noted otherwise), were compared with control samples containing only [³H]taurocholate (5 μM). *P < .05 and **P < .01, statistically significant difference from the control values.

• Download figure
• Open in new tab
• Download powerpoint

Figure 4

Dose-dependent inhibition of NTCP-mediated taurocholate uptake (5 μM) by bile acids and drugs added to transport buffer. A, bile acids ursodeoxycholic acid (●) and cholic acid (▴) and the organic anion bromosulfophthalein (▿). B, drugs that were assessed were (S)-propranolol (■), (R)-propranolol (▵), ketoconazole (▾), furosemide (♦), and cyclosporin (●). The degree of inhibition was derived by comparison to control studies carried out without any inhibitor agents. Data are mean ± S.E.

Although a variety of drugs, such as sodium valproate, nicotine, and salicylate, were without effect at the studied inhibitor concentration (100 μM), a number of drugs were found to significantly reduce NTCP-mediated taurocholate uptake; these agents included furosemide, bumetanide, ketoconazole (400 μM), progesterone, cyclosporin, and the β-adrenoceptor blocker propranolol. Ketoconazole, furosemide, cyclosporin, and racemic-propranolol were further studied over a range of inhibitor concentrations (Fig. 4), and their IC₅₀ values were calculated (Table 1). Additionally, oligopeptides, including the linear oligopeptide renin inhibitor agents U71038 and EMD 51921 (Fig. 4), the somatostatin analog angiopeptin, and the mushroom poison α-amanitin, were also capable of inhibiting taurocholate transport (Fig. 3). By contrast, dipeptides/tripeptides were without effect (Fig. 3).

Interestingly, a number of drugs were found to significantly enhance NTCP-mediated taurocholate uptake (Figs. 3 and5), such as bupivicaine, lidocaine, and quinidine, with the greatest enhancement being observed with quinidine. This effect of quinidine was concentration dependent and resulted in over 2-fold greater uptake compared with that in the absence of quinidine (Fig. 5). BSA-induced enhancement of taurocholate uptake was most notable at a low (1 μg/μl) concentration (Fig. 5); at concentrations above 20 μg/μl, BSA inhibited taurocholate uptake.

• Download figure
• Open in new tab
• Download powerpoint

Figure 5

Dose-dependent induction and inhibition of NTCP-mediated [³H]taurocholate (5 μM) uptake by quinidine (A) and BSA (B). *P < .05 and **P < .01, statistically significant different compared with control values.

Discussion

Large interindividual differences were found in the level of expressed NTCP mRNA obtained from livers of disease-free donors; the difference in expressed mRNA level between the sample exhibiting the lowest level (HL123, Fig. 1) and the highest (HL104, Fig. 1) was 40-fold. In studies of rats, a 10-fold increase in Ntcp mRNA level after prolactin treatment was associated with a 2-fold increase in theV_max value for taurocholate uptake (Liu et al., 1994). Clearly, the measurement of NTCP protein will be an important next step in delineating whether this large variability in NTCP mRNA levels is similarly reflected at the protein level. However, this is currently difficult to determine because a high-affinity antibody against NTCP is not available.

Recombinant vaccinia-mediated expression of NTCP showed the uptake of taurocholate to be rapid and predominantly sodium dependent (Fig. 2). The derived K_m value of 7.9 ± 2.0 μM for taurocholate uptake is in close agreement to theK_m value of 6.2 μM observed for NTCP expressed in X. laevis oocytes (Hagenbuch and Meier, 1994). As expected, bile acids and the organic anion bromosulfophthalein were potent inhibitors of NTCP-mediated taurocholate uptake. Interestingly, however, ursodeoxycholic acid appeared to be the most potent inhibitor (Figs. 3 and 4), with an estimated IC₅₀ value of 3.6 μM (Table 1). Ursodeoxycholic acid is currently used in a variety of liver diseases, particularly for the treatment of cholestatic liver diseases (Poupon and Poupon, 1995). It has been suggested that the mechanism of action of this dihydroxy bile acid involves competition for entry into the hepatocyte with other more hepatotoxic bile acids (Galle et al., 1990; Heuman, 1993). Also, it is generally considered that ursodeoxycholic acid uptake into hepatocyte is primarily passive, due to its lipophilicity (Scharschmidt and Lake, 1989), although more recent work carried out with isolated hamster hepatocytes suggests the involvement of a transporter for ursodeoxycholic acid at concentrations of less than 20 μM (Bouscarel et al., 1995). However, the data in this study suggest that ursodeoxycholic acid reduces the hepatic uptake of more polar and hepatotoxic bile acids through the inhibition of NTCP.

To define the extent of interaction between NTCP and nonbile acid drugs, a two-step approach was adopted. Initially, effects of drugs already known to alter the activity of sodium-dependent bile acid transport system were investigated to determine their relative potency. Subsequently, a number of structurally divergent drugs that are frequently used in the clinical setting were assessed for their effect on NTCP activity. Indeed, drugs previously identified as inhibitors of sodium-dependent bile acid uptake transport systems, including furosemide (Blitzer et al., 1982; Zimmerli et al., 1989), bumetanide (Zimmerli et al., 1989; Hagenbuch et al., 1991; Boyer et al., 1994;Hagenbuch and Meier, 1994), progesterone (Zimmerli et al., 1989), ketoconazole (Zimmerli et al., 1989), and cyclosporin (Stacey and Kotecka, 1988; Moseley et al., 1990; Azer and Stacey, 1993), were able to inhibit NTCP-mediated taurocholate uptake. However, the effects of ketoconazole (Fig. 3) were seen only at high concentrations (>100 μM). On the other hand, cyclosporin was a very potent NTCP inhibitor (Fig. 3), with an apparent IC₅₀ value of 1.0 μM (Table 1). Therapeutic trough plasma levels of cyclosporine in humans are approximately 150 to 400 ng/ml (0.125–0.33 μM), but it is likely that portal vein concentrations are much higher. Accordingly, inhibition of NTCP may be an additional mechanism for the observed cyclosporin-induced cholestasis (Arias, 1993). We also found that a number of other unrecognized drugs, including sodium valproate, nicotine, salicylate (Fig. 3), and the commonly prescribed β-adrenoceptor-blocking agent propranolol, were inhibitors of NTCP (Figs. 3 and 4). Both the (R)- and (S)-propranolol enantiomers appeared to possess similar inhibitory properties, with IC₅₀ values of 5.5 and 6.1 μM, respectively, despite the fact that propranolol is not a substrate of NTCP (data not shown). In addition, peptidomimetic agents, including the pseudohexapeptide renin inhibitor U71038, a pseudotetrapeptide EMD51921, and the somatostatin analog angiopeptin, were also found to be capable of inhibiting NTCP-mediated taurocholate uptake (Fig. 4). However, like propranolol, neither U71038 nor EMD51921 was a transport substrate.

Albumin (Fig. 4), a known enhancer of sodium-dependent bile acid transport (Blitzer and Lyons, 1985), was able to enhance NTCP-mediated taurocholate uptake at low concentrations. However, this effect was significant only at a very low concentration (1 μg/μl), and the further addition of BSA resulted in reduced transport activity (Fig.5). Human plasma albumin levels range from 30 to 40 μg/μl; therefore, it is unlikely the enhanced transport activity seen at the very low concentrations would have any in vivo clinical relevance. Surprisingly, a number of antiarrhythmic drugs, including bupivicaine, lidocaine, and quinidine, were found to enhance NTCP-mediated taurocholate uptake. In fact, quinidine exhibited a concentration-related enhancement in taurocholate uptake. This enhancement of bile acid uptake represents an interesting contrast to the well-recognized effect of quinidine as an inhibitor of ATP-dependent drug efflux transporter P-glycoprotein (Cardarelli et al., 1995; Leu and Huang, 1995; Fromm et al., 1999). The fact that only class I antiarrhythmic agents were found to possess this ability suggests that cell membrane electrical gradient or endogenous sodium/potassium channels present in HeLa cells may be altering NTCP transport activity. Also, it is possible that these drugs directly interact with NTCP at the level of the substrate binding domain and induce positive cooperativity. Regardless, delineation of mechanisms for the observed enhancing effects of such agents was beyond the scope of the current study.

In conclusion, data obtained in this study suggest that the key bile acid uptake transporter NTCP can interact with a variety of drugs in clinical use and not just with bile acids. Moreover, there may be a wide range in the expressed level of NTCP. Clearly, additional studies, particularly in humans, are needed to determine whether these findings reflect the in vivo situation. Nevertheless, the inclusion of NTCP as a potential etiological cause of cholestasis may prove to be therapeutically relevant.