Multispecific Substrate Recognition of Kidney-Specific Organic Anion Transporters OAT-K1 and OAT-K2

Experimental Procedures

Materials.

[3′,5′,7′-³H(N)]Methotrexate, disodium salt (991.6 GBq/mmol); [3′,5′,7′,9′-³H]folic acid, diammonium salt (1.23 GBq/mmol); [methyl-³H]zidovudine (558.7 GBq/mmol); and [³H(G)]ochratoxin A (547.6 GBq/mmol) were obtained from Moravek Biochemicals (Brea, CA). [³H(G)]Taurocholic acid (111 GBq/mmol); [1,2,6,7-³H(N)]dehydroepiandrosterone (DHEA) (2220 GBq/mmol); [1,2,6,7-³H(N)]DHEA sulfate, sodium salt (2220 GBq/mmol); [estradiol-6,7-³H(N)]estradiol 17β-d-glucuronide (1628 GBq/mmol); [6,7-³H(N)]estrone sulfate, ammonium salt (1961 GBq/mmol); L-3,5,3′-[¹²⁵I]triiodothyronine (28.8 TBq/mmol); l-[¹²⁵I]thyroxine (35.9 TBq/mmol); [1-¹⁴C]tetraethylammonium bromide (88.8 MBq/mmol); and [³H(G)]digoxin (703 GBq/mmol) were from PerkinElmer Life Science Products (Boston, MA). [N-methyl-³H]Cimetidine (814 GBq/mmol) was from Amersham Pharmacia Biotech UK, Ltd. (Little Chalfont, Buckinghamshire, UK).

Unlabeled methotrexate, taurocholic acid, triiodothyronine, tetraethylammonium, digoxin, and cimetidine were purchased from Nacalai Tesque (Kyoto, Japan). DHEA was from Tokyo Kasei (Tokyo, Japan). Ochratoxin A, zidovudine, thyroxine, estrone 3-sulfate, DHEA 3-sulfate, and DHEA 3-acetate were obtained from Sigma Chemical Co. (St. Louis, MO). All other chemicals used were of the highest purity available.

Cell Culture and Transfection.

The MDCK cells stably expressing OAT-K1 and OAT-K2, designated MDCK-OAT-K1 and MDCK-OAT-K2, were constructed and maintained as described previously with some modifications (Masuda et al., 1999a).

Transport Studies by Cell Monolayers.

The cellular uptake of radiolabeled drugs was measured using monolayer cultures grown in 12-well microplates as described previously (Takeuchi et al., 2000). The incubation medium for uptake experiments was Dulbecco's phosphate-buffered saline (137 mM NaCl, 3 mM KCl, 8 mM Na₂HPO₄, 1.5 mM KH₂PO₄, 1 mM CaCl₂, and 0.5 mM MgCl₂; pH 7.4) containing 5 mM d-glucose (uptake buffer). Experimental procedures were performed as described previously (Takeuchi et al., 2000). Unlabeled triiodothyronine, thyroxine, DHEA, and ochratoxin A were dissolved in dimethyl sulfoxide, the final concentration of which in uptake buffer was 0.5%. After incubation, the cells were lysed in 0.5 N NaOH solution, and then the radioactivity in aliquots was determined in 5 ml of ACSII (Amersham Pharmacia Biotech UK, Ltd.). The protein content of the solubilized cells was determined by the method of Bradford (Bradford, 1976), by using Bio-Rad Protein Assay kit (Bio-Rad, Hercules, CA) with bovine γ-globulin as a standard.

Functional Expression of oatp2 in X. laevisOocytes.

Aliquots of 25 ng of capped RNA transcribed in vitro fromNotI-linearized oatp2 cDNA with T7 RNA polymerase were injected into X. laevis oocytes. Injected oocytes were maintained in modified Barth's medium [88 mM NaCl, 1 mM KCl, 0.33 mM Ca(NO₃)₂, 0.4 mM CaCl₂, 0.8 mM MgSO₄, 2.4 mM NaHCO₃, and 10 mM HEPES] containing 50 μg/ml gentamicin at 18°C for 3 days. The uptake study was initiated by incubating oocytes in 500 μl of uptake buffer (96 mM NaCl, 2 mM KCl, 1.8 mM CaCl₂, 1 mM MgCl₂, and 5 mM HEPES; pH 7.4) containing [³H]digoxin at 25°C in the presence or absence of various inhibitors for 1 h. Unlabeled triiodothyronine and digoxin was dissolved in dimethyl sulfoxide, and its final concentration in uptake buffer was 0.5%. At the end of incubation, the oocytes were washed five times with 2 ml of ice-cold uptake buffer. After washing, each oocyte was transferred to a single vial and dissolved with 500 μl of 10% sodium lauryl sulfate. The radioactivity of each solubilized oocyte was determined in 5 ml of ACSII.

Statistical Analysis.

Data were analyzed statistically using one-way analysis of variance followed by Fisher's t test.

Results

Uptake of Various Compounds by MDCK-OAT-K1 and MDCK-OAT-K2 Cells.

First, we examined the uptake of various compounds by MDCK-OAT-K1 and MDCK-OAT-K2 cells, in comparison with untransfected MDCK cells as a control. As shown in Table1, uptake of structurally unrelated anionic compounds such as [³H]methotrexate, [¹²⁵I]triiodothyronine, [¹²⁵I]thyroxine, [³H]DHEA 3-sulfate, [³H]estradiol 17β-d-glucuronide, [³H]estrone 3-sulfate, [³H]taurocholic acid, and [³H]ochratoxin A was significantly enhanced in MDCK-OAT-K1 and MDCK-OAT-K2 cells compared with untransfected MDCK cells. In addition, the uptake of the antiretroviral drug [³H]zidovudine, which possesses no typical anionic moiety, was also greater in MDCK-OAT-K1 and MDCK-OAT-K2 cells than that in MDCK cells. However, the uptake of unconjugated steroid [³H]DHEA and typical substrates for organic cation transporters, [¹⁴C]tetraethylammonium and [³H]cimetidine, was comparable among MDCK-OAT-K1, MDCK-OAT-K2, and untransfected MDCK cells.

Then, we determined the kinetic parameters of the newly identified substrates for OAT-K1 and OAT-K2. The chemical structures of substrates for OAT-K1 and OAT-K2 tested are illustrated in Fig. 1. As summarized in Table2, OAT-K1 and OAT-K2 had comparable affinities for structurally unrelated substrates [apparent Michaelis constant (K_m) values of 2 to 80 μM].

Mutual Inhibition between Substrates via OAT-K1 and OAT-K2.

We then examined whether structurally unrelated substrates caused mutual inhibition on the uptake activity of OAT-K1 and OAT-K2. Figure2 shows the effects of various compounds on [³H]methotrexate and [³H]taurocholic acid uptake via OAT-K1 and OAT-K2. [³H]Methotrexate uptake via OAT-K1 and OAT-K2 was markedly inhibited by unlabeled methotrexate but showed little or no inhibition by structurally unrelated substrates at the concentration of 100 μM. On the other hand, [³H]taurocholic acid uptake was significantly inhibited by taurocholic acid, DHEA, its conjugated anionic compounds DHEA 3-sulfate and DHEA 3-acetate, and ochratoxin A, although methotrexate and zidovudine did not have any effect at the concentration of 100 μM. Interactions of methotrexate and thyroid hormones with OAT-K1 and OAT-K2 are shown in Fig.3. Inhibition of [³H]methotrexate uptake by unlabeled thyroid hormones at a concentration of 50 μM, which was higher than the apparent K_m value of each compound, was relatively weak or was not observed. In addition, 2.5 mM unlabeled methotrexate, which was about 1000-fold higher than the apparentK_m values for OAT-K1 and OAT-K2, inhibited only about 20% of the control uptake of¹²⁵I-thyroid hormones. [³H]Zidovudine uptake via OAT-K1 and OAT-K2 was significantly inhibited by unlabeled zidovudine. However, unlabeled methotrexate, tetraethylammonium, and cimetidine at the concentration of 100 μM did not inhibit OAT-K1- or OAT-K2-mediated [³H]zidovudine uptake (Fig.4).

Trans-Stimulation Effects of Intracellular Methotrexate on Uptake of Various Labeled Compounds via OAT-K1 and OAT-K2.

To obtain more information about the characteristics of substrate recognition of OAT-K1 and OAT-K2, we examined thetrans-stimulation effects of intracellular unlabeled methotrexate on the uptake of various labeled substrates via OAT-K1 and OAT-K2. Although the uptake of [³H]folic acid was stimulated in the presence of the counterdirected transmembrane gradient of unlabeled methotrexate, the uptake of [³H]DHEA, [³H]DHEA 3-sulfate, [¹²⁵I]triiodothyronine, and [¹²⁵I]thyroxine by OAT-K1 and OAT-K2 was not stimulated (Fig. 5).

Inhibition Study by Using oatp2-Expressing Oocytes.

To examine whether these phenomena were unique features of OAT-K1 and OAT-K2, or were observed for other transporters belonging to theoatp-gene family, we performed inhibition experiments by using oatp2-expressing oocytes. To compare the inhibition potencies of unlabeled digoxin, triiodothyronine, and taurocholic on the oatp2-mediated uptake of [³H]digoxin, the dose dependence of the inhibition was examined (Fig.6). The [³H]digoxin uptake via oatp2 was inhibited by unlabeled digoxin, triiodothyronine, and taurocholic acid in this order of inhibitory potency. The estimated inhibition constant (IC₅₀) values for [³H]digoxin uptake were 0.81 μM for unlabeled digoxin, 50.4 μM for triiodothyronine, and 1.48 mM for taurocholic acid.

Discussion

Over the past several years, considerable progress has been made in the molecular identification and characterization of organic anion transporters in the kidney, providing useful information for pharmaceutical research (Inui et al., 2000). At present, organic anion transporters expressed in the kidney are classified as belonging to two distinct gene families, OAT and oatp/OAT-K (Inui et al., 2000; Dresser et al., 2001). OAT-K1 and OAT-K2 are members of the oatp-gene family, and their roles in the renal excretion of methotrexate have been clarified (Saito et al., 1996; Masuda et al., 1999a,b; Takeuchi et al., 2000, 2001).

In the present study, we expanded the substrate spectrum of OAT-K1 and OAT-K2, by using MDCK cells stably transfected with these transporter cDNAs. This broad substrate spectrum of OAT-K1 and OAT-K2 includes methotrexate, ochratoxin A, and various endogenous organic anions (thyroid hormones, taurocholic acid, and conjugated steroids) (Tables 1and 2). We previously reported that OAT-K1 specifically transports methotrexate and folic acid, but not taurocholic acid, by using LLC-OAT-K1 cells and OAT-K1-expressing oocytes (Masuda et al., 1996,1999a). These discrepancies are probably related to the possibility that membrane-sorting mechanism and the proteolytic processing and/or excision of OAT-K1 found in renal tubular cells are retained in MDCK-OAT-K1 cells, but not in LLC-OAT-K1 cells (Inui et al., 2000). By proteolytic processing and/or excision, OAT-K1 protein in the MDCK-OAT-K1 cells and in the renal tubular cells may lose some structural features, which determine the strict substrate recognition in LLC-OAT-K1 cells and in OAT-K1-expressing oocytes. The present study indicated that both OAT-K1 and OAT-K2 possess the properties of multispecific substrate recognition in the brush-border membranes of the renal tubular cells.

Among the newly identified substrates for OAT-K1 and OAT-K2, ochratoxin A, thyroid hormones, taurocholic acid, and conjugated steroids are overlapping substrates for other members of the oatp-gene family, oatp1, oatp2, and oatp3 (Jacquemin et al., 1994; Kontaxi et al., 1996; Noé et al., 1997; Abe et al., 1998; Reichel et al., 1999). However, OAT-K1 and OAT-K2 are unique in mediating the transport of methotrexate. In addition, OAT-K1 and OAT-K2 transported the antiretroviral drug zidovudine, which possesses no typical anionic moiety (Fig. 1). Montfoort et al. (1999) reported that oatp1 and oatp2 mediate the uptake of cationic compoundsN-(4,4-azo-n-pentyl)-21-deoxyajmalinium and rocuronium. These results indicated that although the majority of substrates for members of the oatp-gene family are amphipathic organic anions, these transporters can also transport cationic compounds as characteristics common among members of theoatp-gene family. Some of the substrates of OAT-K1 and OAT-K2, methotrexate, ochratoxin A, and zidovudine, also overlap with those of another organic anion transporter, rat OAT1 (rOAT1), which belongs to a distinct gene family from OAT-K1 and OAT-K2 (Inui et al., 2000; Dresser et al., 2001). rOAT1 is expressed in the basolateral membranes of the proximal tubules and exhibits the transport characteristics of a classic organic anion, p-aminohippuric acid, at the basolateral membranes of the proximal tubules (Sekine et al., 1997; Sweet et al., 1997). By considering the membrane localization of these transporters, rOAT1 at the basolateral membranes and OAT-K1 and OAT-K2 at the brush-border membranes may be responsible for the vectorial transport of several organic compounds via the proximal tubular cells. We previously demonstrated that both OAT-K1 and OAT-K2 mediate bidirectional transport of substrates, methotrexate and folic acid for OAT-K1, and taurocholic acid for OAT-K2, respectively (Masuda et al., 1999a,b; Takeuchi et al., 2000). Most recently, the role of OAT-K1 and OAT-K2 in the excretion of methotrexate from the proximal tubules into urine has been clarified using the rat model of renal failure (Takeuchi et al., 2001). Therefore, a wide variety of newly identified substrates of OAT-K1 and OAT-K2 may interact with these transporters from both inside and outside of the brush-border membranes. However, further characterization of these transporters, such as evaluation of substrate recognition from inside as well as outside of the brush-border membranes, is needed to elucidate their distinct roles in the renal handling of organic compounds.

If several substrates interact with a transporter at the same site in its molecule, the expected inhibition constant values should be comparable with the K_m values. Interestingly, the apparent inhibition constant value of taurocholic acid for [³H]methotrexate uptake via OAT-K1 is 183 μM (Takeuchi et al., 2000), and that of methotrexate for [³H]taurocholic acid uptake via OAT-K2 is 252 μM (A. Takeuchi, S. Masuda, H. Saito, and K. Inui, unpublished data), which are about 100-fold higher than the apparent K_m value of each compound, respectively (Table 2). The present study confirmed this phenomena. That is, uptake of structurally unrelated substrates via OAT-K1 and OAT-K2 showed little or no mutual inhibition even when concentrations of inhibitors were much higher than their apparentK_m values (Figs. 2-4). Moreover, not only OAT-K1 and OAT-K2 but also oatp2 had unique characteristics in substrate recognition. Although apparentK_m values of triiodothyronine and taurocholic acid for oatp2 were reported to be 5.87 and 35 μM, respectively (Reichel et al., 1999), the IC₅₀value of triiodothyronine for [³H]digoxin uptake via oatp2 was 50.4 μM, and that of taurocholic acid was 1.48 mM, which were about 9- and 40-fold higher than the apparentK_m value of each compound, respectively (Fig. 6). There is a discrepancy betweenK_m value and IC₅₀ value if these substrates interact at the same substrate recognition site in the transporter molecules. One possible explanation that accounts for this discrepancy is that there may be several substrate recognition sites in the OAT-K1, OAT-K2, and oatp2 molecules. The existence of several substrate recognition sites was further confirmed by the trans-stimulation experiment. Intracellular unlabeled methotrexate trans-stimulated the uptake of [³H]folic acid, but not of structurally unrelated substrates ¹²⁵I-thyroid hormones and [³H]DHEA 3-sulfate (Fig. 5). This was comparable with our previous report describing thetrans-stimulation effects of extracellular folic acid derivatives but not taurocholic acid on [³H]methotrexate efflux from MDCK-OAT-K1 cells (Takeuchi et al., 2000). Therefore, OAT-K1, OAT-K2 and oatp2 were suggested to transport a wide variety of structurally diverse organic compounds, not via a single interaction site in their molecules, but via several different interaction sites, preserving the transport activity even when one substrate occupies one interaction site. Such multispecificity of substrate recognition of OAT-K1, OAT-K2, and oatp2 may contribute to renal and hepatic transport systems in the elimination of a wide range of compounds. If these various compounds interact with a transporter at the same single interaction site, competition between substrates with the transporter should occur frequently. Therefore, the multiple substrate recognition sites in the OAT-K1, OAT-K2, and oatp2 molecules may minimize the severe drug-drug or drug-endogenous compound interactions via the transporter, reducing adverse effects on the body such as elevated plasma concentration or delayed excretion of drugs, toxins, and metabolites. Although the precise characteristics determining which kinds of substrates inhibit the transport activities of these transporters and which do not are not yet fully understood, the present results provide useful information for the comprehension of the pharmacokinetics of drugs in the kidney and the liver.

In conclusion, OAT-K1 and OAT-K2 were suggested to serve as multispecific organic anion transporters, mediating transport of a wide variety of xenobiotics and their metabolites, presumably via several substrate interaction sites.