Experimental Procedures

Materials.

[¹⁴C]Tetraethylammonium (TEA; 5 mCi/mmol) was obtained from PerkinElmer Life Science Products (Boston, MA). [³H]Quinidine (QD; 20 Ci/mmol) was purchased from American Radiolabeled Chemicals (St. Louis, MO). [¹⁴C]Rocuronium (54 mCi/mmol) and unlabeled rocuronium were kind gifts of Organon International BV (Oss, The Netherlands).N-(4,4-Azo-n-pentyl)-21-deoxy[21-³H]ajmalinium (APD-ajmalinium, APDA; 1.2 Ci/mmol),N-(4,4-azo-n-pentyl)-quinuclidine (APQ; 2.5 Ci/mmol), and unlabeled APQ were synthesized as described (Müller et al., 1994). [³H]Tributylmethylammonium (TBuMA; 85 Ci/mmol) was synthesized according to Neef et al. (1984). Unlabeled TBuMA was obtained from Fluka (Buchs, Switzerland). [³H]N-Methyl-quinine (NMQ, 85 Ci/mmol) and [³H]N-methyl-quinidine (NMQD; 85 Ci/mmol) were synthesized and characterized as described (van Montfoort et al., 1999). [³H]Azidoprocainamide methoiodide (APM; 85 Ci/mmol) and unlabeled APM were synthesized according to Mol et al. (1992). Radiochemical purity of the not commercially available substrates was determined by thin-layer chromatography and exceeded 99%. All other chemicals were of analytical grade and readily available from commercial sources.

Uptake Studies in Xenopus laevis Oocytes.

rOCT1 cDNA was subcloned into the pRSSP vector (Busch et al., 1996) and linearized with Mlu I. hOCT1 cDNA was subcloned into the pOG1 vector (Gorboulev et al., 1997) and linearized withNotI. cRNA was synthesized with the Ambion mMessage mMachine In Vitro Transcription kit using SP6 RNA polymerase for rOCT1 and T7 RNA polymerase for hOCT1 (Ambion, Austin, TX). In vitro synthesis of Oatp2- and OATP-A-cRNA was performed as described (Kullak Ublick et al., 1995; Noe et al., 1997). X. laevis oocytes were prepared (Hagenbuch et al., 1996) and cultured overnight at 18°C. Healthy oocytes were microinjected with 50 nl of water without and with 10 ng of rOCT1-, 10 ng of hOCT1-, 5 ng of Oatp2-, or 2.5 ng of OATP-A-cRNA and cultured for 3 days in a medium containing 88 mM NaCl, 2.4 mM NaHCO₃, 1 mM KCl, 0.3 mM Ca(NO₃)₂, 0.41 mM CaCl₂, 0.82 mM MgSO₄, 0.05 mg/ml gentamycin, and 15 mM HEPES (pH 7.6). Tracer uptake studies were performed in a NaCl-containing uptake medium (100 mM NaCl, 2 mM KCl, 1 mM CaCl₂, 1 mM MgCl₂, and 10 mM HEPES-Tris, pH 7.5 or 10 mM 4-morpholinoethanesulfonic acid-Tris, pH 6.0). The oocytes were washed in the uptake medium and then incubated at 25°C in 100 μl of the uptake medium containing the indicated substrate concentrations. Water-injected oocytes were used as controls for unspecific uptake of the substrate. After the indicated time intervals, uptake was stopped by addition of 6 ml of ice-cold uptake medium. The oocytes were washed twice with 6 ml of ice-cold uptake medium. Subsequently, each oocyte was dissolved in 0.25 ml of 10% SDS and 4 ml of scintillation fluid (Ultima Gold; Canberra Packard, Zürich, Switzerland) and the oocyte-associated radioactivity determined in a Tri-Carb 2200 CA liquid scintillation analyzer (Canberra Packard). Determination of kinetic uptake parameters was performed by a nonlinear curve-fitting program (Systat 8.0; SPSS Inc., Chicago, IL) using a simple Michaelis-Menten model:v = (V_max · [S])/(K_m + [S]).

Statistical Analysis.

Uptake results are given as means ± S.D. Statistical significance of transport differences between the various oocyte groups was determined by the Student's ttest (Systat 8.0; SPSS Inc.).

Results

To identify rOCT1 as the type I organic cation uptake system (system1) and Oatp2 as the type II organic cation uptake system (system2) of rat liver, we first performed a series ofcis-inhibition studies in rOCT1- and Oatp2-expressingX. laevis oocytes using TBuMA as a type I and rocuronium as a type II substrate. As illustrated in Fig.1A, rOCT1-mediated TBuMA uptake was inhibited by APM (a type I compound) and rocuronium (a type II compound), but not by taurocholate, ouabain, and K-strophantoside. Thiscis-inhibition pattern is characteristic for the rat liver organic cation uptake system1 (Steen et al., 1992), which therefore corresponds to rOCT1. The latter was also inhibited by QD as previously described (Koepsell et al., 1999) as well as by its methylated derivative NMQD (Fig. 1A). In contrast, Oatp2-mediated rocuronium uptake was insensitive to the type I compounds TBuMA and APM, but was strongly inhibited by taurocholate, ouabain, and K-strophantoside (Fig.1B). This cis-inhibition pattern is characteristic for the rat liver organic cation uptake system2 (Steen et al., 1992), which therefore corresponds to Oatp2. No significant inhibition of Oatp2-mediated rocuronium uptake was found for QD and NMQD (Fig. 1B), which is consistent with the previous observation that NMQD is not transported by Oatp2 (van Montfoort et al., 1999).

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

cis-Inhibition pattern of TBuMA and rocuronium uptake mediated by rOCT1 (A) and Oatp2 (B), respectively.X. laevis oocytes were injected with 10 ng of rOCT1 or 5 ng of Oatp2 cRNA. After 3 days in culture, uptakes of [³H]TBuMA (35 μM) by rOCT1 (A) and of [¹⁴C]rocuronium (25 μM) by Oatp2 (B) were measured at 30 min (under Experimental Procedures) in the absence (control) and presence of the indicated compounds (all 200 μM, except ouabain 2 mM). Uptake rates of water-injected oocytes were subtracted from cRNA-injected oocytes for each substrate. Data are expressed as means ± S.D. of uptake measurements into 10 oocytes. ∗, significant uptake inhibition (Student's t test,p < 0.01) compared with controls.

Next, we evaluated whether the characteristic cis-inhibition pattern of rOCT1 and Oatp2 is also valid for the human transporters hOCT1 and OATP-A. As illustrated in Fig.2A, hOCT1-mediated TBuMA uptake showed qualitatively a similar cis-inhibition pattern as rOCT1-mediated TBuMA uptake (Fig. 1A), indicating that hOCT1 and rOCT1 represent orthologous gene products with similar type I organic cation transport properties in rat and human liver. In contrast, OATP-A-mediated rocuronium uptake was inhibited by APM, taurocholate, K-strophantoside, QD, and NMQD, but not by TBuMA and ouabain (Fig. 2B). This cis-inhibition pattern is clearly different compared with Oatp2 (Fig. 1B). While part of these differences can be explained by distinct transport activities [e.g., NMQD is a transport substrate of OATP-A, but not of Oatp2 (van Montfoort et al., 1999)] and/or substrate affinities [e.g., K_m values for ouabain transport: OATP-A, ∼5.5 mM; Oatp2, ∼470 μM (Bossuyt et al., 1996; Noe et al., 1997)] of OATP-A and Oatp2, the inhibitory effect of the type I organic cation APM on OATP-A-mediated rocuronium uptake is surprising since APM has been shown to be not transported by OATP-A (van Montfoort et al., 1999). Hence, the data indicate that the rat-based classification of type I and type II organic cations and its association with distinct organic cation carriers cannot be applied to the human situation without modification.

• Download figure
• Open in new tab
• Download powerpoint

Figure 2

cis-Inhibition pattern of TBuMA and rocuronium uptake mediated by hOCT1 (A) and OATP-A (B), respectively.X. laevis oocytes were injected with 10 ng of hOCT1 or 2.5 ng of OATP-A cRNA. After 3 days in culture, uptakes of [³H]TBuMA (35 μM) by hOCT1 (A) and of [¹⁴C]rocuronium (25 μM) by OATP-A (B) were measured at 30 min (under Experimental Procedures) in the absence (control) and presence of the indicated compounds (all 200 μM, except ouabain 2 mM). Uptake rates of water-injected oocytes were subtracted from cRNA-injected oocytes for each substrate. Data are expressed as means ± S.D. of uptake measurements into 10 oocytes. ∗, significant uptake inhibition (Student's t test,p < 0.01) compared with controls.

Based on their physicochemical properties as permanently charged bulky quaternary ammonium compounds it has been suggested that NMQ and NMQD represent new type II organic cations (van Montfoort et al., 1999). However, the overall structure of these compounds shows a spatial separation of the single cationic group (surrounded by aliphatic moieties as in TBuMA and PAEB) from an aromatic structure as has been considered as being typical for type I compounds (Meijer et al., 1990;Groothuis and Meijer, 1996). Even APDA exhibits these structural features to some extent. The latter consideration is supported by the findings that NMQD inhibits rOCT1- and hOCT1-mediated TBuMA uptake rather than Oatp2-mediated rocuronium uptake (Figs. 1 and 2). Therefore, we tested next organic cation transport in rOCT1- and hOCT1-expressing oocytes. As illustrated in Fig.3, and consistent with thecis-inhibition data, rOCT1 and hOCT1 mediated not only transport of the type I compounds TBuMA, APM, and APQ but also of the previously supposed type II substrates NMQ, NMQD, and also but to a lesser extent, ADPA. The only type II organic cation not significantly transported by rOCT1 and hOCT1 was rocuronium. Time course experiments showed linear uptake rates for NMQ, NMQD, TBuMA, and APM for at least 30 min (data not shown). Therefore, kinetic uptake measurements were performed at 15 min. They showed saturation kinetics as indicated in Fig. 4 for hOCT1. Similar kinetic features were also obtained for rOCT1 with all apparentK_m values given in Table1. These data confirm the affinity of rOCT1 and hOCT1 for the established type I organic cations TBuMA, APM, and APQ. However, since rOCT1 and hOCT1 also transport the rather bulky organic cations NMQ, NMQD, and APDA, the data indicate that this physicochemical property alone should not be used to predict the types of carrier involved in hepatic organic cation uptake. In this study, only rocuronium showed entirely the expected behavior as a type II organic cation, i.e., exclusive and taurocholate/cardiac glycoside inhibitable transport by Oatp2 and OATP-A, respectively. Other examples of type II compounds are d-tubocurarine, metocurine, hexafluronium, and vecuronium. A general feature of these agents is that 1) the cationic groups are not clearly separated from the aromatic moieties or other bulky ring structures, and 2) that they contain a second quaternary or tertiary amine structure and consequently can form bivalent cationic molecules.

• Download figure
• Open in new tab
• Download powerpoint

Figure 3

Comparison of rOCT1- and hOCT1-mediated uptake of [³H]NMQ, [³H]NMQD, [³H]APDA, [¹⁴C]rocuronium, [³H]TBuMA, [³H]APM, and [³H]APQ. X. laevis oocytes were injected with 10 ng of rOCT1 (A) or 10 ng of hOCT1-cRNA (B). After 3 days in culture, uptakes of the indicated substrates were measured at 2 μM (except rocuronium at 14 μM) for 30 min in Na⁺ buffer (under Experimental Procedures). The standard substrate TEA was used as positive controls to test the expression of rOCT1 and hOCT1 in the cRNA-injected oocytes. The uptake rates of the water-injected control oocytes were subtracted from the cRNA-injected oocytes for each substrate. Data represent the mean ± S.D. from 11 to 15 oocyte uptake measurements. ∗, significantly different from water-injected control oocytes (Student's t test, p < 0.001).

• Download figure
• Open in new tab
• Download powerpoint

Figure 4

Saturation kinetics of NMQ, NMQD, TBuMA, and APM in hOCT1-cRNA-injected X. laevis oocytes. X. laevis oocytes were injected with 10 ng of hOCT1-cRNA. After 3 days in culture, uptakes of [³H]NMQ (A), [³H]NMQD (B), [³H]TBuMA (C), and [³H]APM (D) were measured at 15 min in Na⁺buffer (under Experimental Procedures) containing increasing substrate concentrations. The dotted lines represent fitted uptake differences between hOCT1-cRNA (●) and water-injected oocytes (○). Data are expressed as mean ± S.D. of 10 to 12 oocyte uptake measurements.

The quaternary ammonium compounds used in this study are permanently positively charged model compounds and with the exception of rocuronium not used as drugs. Therefore, we also investigated transport of the drug quinidine, which is a tertiary amine and shown to inhibit rOCT1, hOCT1, and OATP-A (Figs. 1 and 2). Quinidine is a base with a pK_a of 8.5 (Notterman et al., 1986), which means that at pH 7.5 about 10% of the quinidine molecules are protonated, while at pH 6.0 almost all molecules are positively charged. It can be seen in Fig. 5 that there is no significant quinidine transport by OATP-A-, rOCT1-, or hOCT1-cRNA-injected X. laevis oocytes at pH 7.5. At pH 6.0, however, transport of quinidine becomes detectable since the unspecific uptake into water injected oocytes is markedly reduced. These findings suggest that quinidine molecules need a positive charge to be transported by OATP-A, rOCT1, and hOCT1. In the unprotonated form, quinidine most probably enters the oocytes by passive diffusion as can be seen by the large unspecific uptake into water-injected oocytes at pH 7.5 (Fig. 5).

• Download figure
• Open in new tab
• Download powerpoint

Figure 5

pH dependence of OATP-A-, rOCT1-, and hOCT1-mediated uptake of [³H]quinidine. X. laevis oocytes were injected with 2.5 ng of OATP-A, 10 ng of rOCT1, or 10 ng of hOCT1-cRNA. After 3 days in culture, uptake of the quinidine was measured at 2 μM for 30 min in Na⁺ buffer at pH 7.5 (▨) or pH 6.0 (▪) (under Experimental Procedures). NMQD was used as positive control to test the expression of OATP-A, rOCT1, and hOCT1 in the cRNA-injected oocytes. Data represent the mean ± S.D. from 10 oocyte uptake measurements. ∗, significantly different from water-injected control oocytes (Student's t test,p < 0.001).

Discussion

The present study identifies rOCT1 as the type I organic cation uptake system (system1) and Oatp2 as the type 2 organic cation uptake system (system2) in rat liver. This conclusion is based on the typicalcis-inhibition of rOCT1-mediated TBuMA uptake by type I and type II cations, but not by taurocholate and cardiac glycosides (Fig.1A) (Steen et al., 1992), and of Oatp2-mediated rocuronium uptake by taurocholate and cardiac glycosides, but not by type I cations (Fig.1B) (Steen et al., 1992). While the transport properties of hOCT1 were similar to rOCT1 (Fig. 2A), the cis-inhibition patterns of Oatp2- and OATP-A-mediated rocuronium uptake were different in part in that OATP-A, but not Oatp2, was inhibited by the type I organic cation APM, QD, and NMQD (Fig. 2B). Thus, the clear-cut associations between transport of type I and type II organic cations by rOCT1 and Oatp2, respectively, is not entirely valid for hOCT1 and OATP-A, which supports the concept that the human OATP-A is not the orthologous gene product of the rat Oatp2 (Kullak-Ublick et al., 2001). Furthermore, studies in isolated human hepatocytes showed that rocuronium uptake could be inhibited by K-strophantoside and by the type I organic cation PAEB, but not by taurocholate (Olinga et al., 1998). These findings are an additional indication for species differences between organic cation uptake systems of rat and human liver. Since none of the so-far-identified human hepatic OATPs [i.e., OATP-B (SLC21A9), OATP-C (SLC21A6), and OATP8 (SLC2A8)] was able to transport organic cations, a human ortholog of rat Oatp2 remains to be identified (Kullak-Ublick et al., 2001). It might also be possible that the transport functions of rat and human OATPs have evolved differently since in contrast to rat Oatp2 no human OATP can transport both rocuronium and digoxin. In humans, rocuronium is transported by OATP-A (van Montfoort et al., 1999), while digoxin is transported by OATP8 (Kullak-Ublick et al., 2001).

The concept of rOCT1 corresponding to uptake system1 and Oatp2 to uptake system2 is valid for the same substrates and inhibitors that were used originally in rat hepatocytes (Steen et al., 1992). Other substrates have to be classified with caution. NMQ, NMQD, and APDA could, at first sight, be viewed upon as type II organic cations because of their bulky structure (van Montfoort et al., 1999). According to the proposed concept their uptake into rat liver is supposed to be mediated by Oatp2 rather than by rOCT1. However, while Oatp2 solely mediates the transport of the real type II organic cation rocuronium, APDA is only to some extent accommodated by OATP-A and Oatp2, respectively, whereas NMQ and NMQD even seem pure rOCT1-mediated type I compounds (van Montfoort et al., 1999). Therefore, a positive charge and a bulky structure alone are not sufficient to predict the putative uptake system. As NMQ and NMQD are mainly taken up by rOCT1, which corresponds to uptake system1, they should be reclassified as type I organic cations despite their bulky structure. It is possible that rOCT1 and hOCT1 recognize a single cationic group spatially separated from a flat (aromatic) ring structure that can be recognized in APDA, NMQ, NMQD, and PAEB. The type I agents TBuMA and APQ also have a single cationic group but lack an aromatic moiety. The only true type II substrate used in this study that is consistently transported only by Oatps is the steroidal muscle relaxant rocuronium. It shares a potential dicationic nature with previously categorized type II compounds (Meijer et al., 1990) in combination with a sufficient lipophilicity. The latter aspect and the presence of a permanently charged ammonium group as well as a second tertiary amine function may render APDA a mixed type I/type II substrate.

While there were considerable differences in the substrate specificity of Oatp2 and OATP-A in this study, rOCT1 and hOCT1 showed qualitatively similar substrate specificities and inhibition patterns (Figs. 1-3), although some quantitative differences were observed in the inhibition of TBuMA uptake by APM (rOCT1 > hOCT1) and NMQD (rOCT1 < hOCT1). In addition, the apparent K_mvalues for NMQ, NMQD, TBuMA, and APM were comparable (Table 1). hOCT1 showed slightly higher K_m values, which is in agreement with previous findings where hOCT1 exhibited a lower affinity than rOCT1 for several other ligands (Koepsell et al., 1999). Furthermore, it has been shown that hOCT1 transports largern-tetraalkylammonium compounds such as tetrapropylammonium and tetrabutylammonium at higher rates than rOCT1 (Dresser et al., 2000). The kinetics of TBuMA and APM has also been determined in isolated rat hepatocytes (Steen et al., 1991; Mol et al., 1992) where for both substrates the presence of high- and low-affinity uptake systems has been suggested with apparentK_m values of ∼1.2 and ∼107 μM for TBuMA (Steen et al., 1991) and of ∼3 and ∼100 μM for APM, respectively (Mol et al., 1992). In the present study, rOCT1 showed apparent K_m values of 34 μM for TBuMA and of 54 μM for APM. If these values, as obtained in the oocyte system, can be extrapolated to the hepatocyte, they could correspond to the low-affinity system found in isolated rat hepatocytes. In that case, candidate carriers for the supposed high-affinity systems should be identified in the future.

rOCT1, hOCT1, and OATP-A are not only inhibited by the permanently positively charged NMQD but also by the base quinidine (Figs. 1 and 2). However, at pH 7.5 where only about 10% of the quinidine molecules are protonated no carrier-mediated transport of quinidine could be detected (Fig. 5). In contrast, at pH 6.0, where almost all quinidine molecules are positively charged, unspecific diffusion into the oocytes was markedly reduced and uptake of the protonated quinidine by rOCT1, hOCT1, and OATP-A became evident (Fig. 5). This finding is an indication that quinidine is only transported by rOCT1, hOCT1, and OATP-A, when it carries a positive charge. The mechanism of quinidine inhibition of rOCT1, hOCT1, and OATP-A at pH 7.5 is not clear. It has been shown that quinine, which is a diastereomer of quinidine, is a noncompetitive inhibitor of rOCT1 when added at the extracellular site, suggesting an allosteric binding site (Koepsell, 1998). Recent experiments with macropatches of rOCT2 cRNA-injected X. laevis oocytes in which quinine could also be added from the cytoplasmic site showed competitive inhibition, suggesting that quinine inhibits rOCT2 from the intracellular site after crossing the lipid bilayer in its uncharged form (Budimann et al., 2000).

In conclusion, the present study demonstrates that rOCT1 corresponds to the predicted uptake system1 for type I organic cations in rat liver, whereas Oatp2 corresponds to uptake system2 for type II organic cations. However, the results cannot be transferred to human liver without modifications. While the transport properties of rOCT1 and hOCT1 are similar, considerable differences exist between Oatp2 and OATP-A, indicating further that rat Oatps and human OATPs have different transport properties and do not represent orthologous gene products.