Introduction

Top Abstract Introduction Experimental Procedures Results Discussion References

Proximal tubular cells take up a variety of organic anions from the peritubular interstitium, the first step of tubular secretion, via the basolateral p-aminohippurate (PAH) transporter (Moller and Sheikh, 1983; Pritchard and Miller, 1993; Ullrich and Rumrich, 1993; Ullrich, 1997). The basolateral PAH transporter has been a subject of extensive research especially from two perspectives: first, its substrate selectivity. The PAH transporter has a remarkably wide substrate selectivity; it interacts with and transports a number of anionic substances with unrelated chemical structures. The common structural characteristics of the substrates have been studied, and the prerequisite structures in the substrates of the PAH transporter have been considered to be an appropriately sized hydrophobic domain and a negatively charged site or sites (Ullrich, 1997). Second, the PAH transporter has been associated with the pharmacokinetics and toxicokinetics of anionic drugs. The substrates of the PAH transporter include a number of therapeutically important drugs, and the renal clearance of anionic drugs is closely related to their transport via the PAH transporter (Moller and Sheikh, 1983; Pritchard and Miller, 1993). In addition, several nephrotoxic drugs, such as cephaloridine (a nephrotoxic -lactam antibiotics), have been indicated to exert nephrotoxicity by virtue of their accumulation in the renal proximal tubular cells by the PAH transporter (Tune, 1997).

Recently, the PAH transporter was cloned from the rat kidney and designated as organic anion transporter 1 (OAT1/ROAT1) (Sekine et al., 1997; Sweet et al., 1997). The functional expression of rat OAT1 (rOAT1) allows examination of the definite substrate selectivity of the "multispecific transporter protein". To date, it has been reported that rOAT1 mediates the transport of anionic drugs such as nonsteroidal anti-inflammatory drugs (Apiwattanakul et al., 1999), -lactam antibiotics (Jariyawat et al., 1999; Takeda et al., 1999), methotrexate (Sekine et al., 1997), environmental substances such as a mycotoxin (Tsuda et al., 1999), and various endogenous organic anions, such as prostaglandins, dicarboxylates, cyclic nucleotides (Sekine et al., 1997), and folate (Uwai et al., 1998). Recently, cidofovir and adefovir, antiviral drugs that are nucleoside analogs with a phosphate group, were also demonstrated to be transported via OAT1 (Cihlar et al., 1999). Although the chemical structures of these compounds are diverse, all of these substrates possess a typical anionic moiety (mostly a carboxylic group or a phosphate group) and a hydrophobic core.

3'-Azido-3'-deoxythymidine (zidovudine, AZT) is a prototypical antiretroviral drug widely used for the treatment of human immunodeficiency virus (HIV) infection (de Miranda et al., 1989). AZT is a thymidine analog, and its pK_a value was reported to be 9.68 (Chatton et al., 1990). Because of the high pK_a value, several investigators regarded AZT as a cationic compound (Henry et al., 1988; Kornhauser et al., 1989). Acyclovir is an antiherpes virus drug (a purine nucleoside analog) with two pK_a values (2.27 and 9.25; Laskin et al., 1982). Neither of these drugs possesses a typical anionic moiety (Fig. 1). In addition, both of these drugs are not very hydrophobic. In the rat, AZT is mainly excreted in the urine, and the rate of excretion was reduced by probenecid, a typical inhibitor of renal organic anion transporter (Chatton et al., 1990; Mays et al., 1991). In humans, a reduction in renal clearance of ACV was observed after the oral administration of probenecid (Laskin et al., 1982). These results suggest that the renal organic anion transport system is responsible for the tubular secretion of ACV and AZT. Furthermore, involvement of the organic cation transport system was also suggested in the tubular secretion of AZT, because cimetidine (an organic cation) also reduced the renal clearance of AZT (Chatton et al., 1990). In the case of ACV, probenecid did not totally abolish its tubular secretion (Laskin et al., 1982). Thus, the elimination pathway for AZT and ACV remains to be elucidated.

View larger version (11K): [in this window] [in a new window]   Fig. 1.   Chemical structures of AZT and ACV. Neither AZT nor ACV possess typical anionic groups.

In this study, we investigated whether rOAT1 transports AZT and ACV. We also examined the transport of other nucleoside analogs such as zalcitabine, didanosine, lamivudine, stavudine, and trifluridine. The results demonstrated that rOAT1, heterologously expressed in Xenopus laevis oocytes, mediated the transport of nucleoside analog antiviral drugs.

	Experimental Procedures

Top Abstract Introduction Experimental Procedures Results Discussion References

Materials. [¹⁴C]AZT (2.04 GBq/mmol), [³H]zalcitabine (ddC; 1.85 TBq/mmol), [³H]didanosine (ddI; 2.08 TBq/mmol), [¹⁴C]stavudine (d4T; 2.07 GBq/mmol), [³H]lamivudine (518 GBq/mmol), [³H]ACV (333 GBq/mmol), [¹⁴C]foscarnet (1.92 GBq/mmol), and [¹⁴C]trifluridine (2.07 GBq/mmol) were purchased from Moravek Biochemicals Inc. (Brea, CA). [¹⁴C]PAH (2.00 GBq/mmol) was purchased from DuPont-New England Nuclear (Boston, MA). All other chemicals used in the study were purchased from Sigma Chemical Co. (St. Louis, MO).

cRNA Synthesis and Uptake Experiments Using X. laevis Oocytes. Capped cRNA for rOAT1 was synthesized in vitro using T7 RNA polymerase as described elsewhere (Sekine et al., 1997). Defolliculated oocytes were injected with 15 ng of rOAT1 cRNA and maintained in modified Barth's solution [88 mM NaCl, 1 mM KCl, 0.33 mM Ca(NO₃)₂, 0.4 mM CaCl₂, 0.8 mM MgSO₄, 2.4 mM NaHCO₃, 10 mM HEPES, and 50 µg/ml gentamicin, pH 7.4, and sterilized by filtration] at 18°C for 3 days. Noninjected oocytes were used as control, because preliminary experiments showed that there were no differences in the cell-associated count of radiolabeled PAH or antiviral agents between water-injected and noninjected oocytes.

View larger version (18K): [in this window] [in a new window]   Fig. 2.   Transport of [14C]AZT via rOAT1. A, uptake of 4 µM [14C]AZT by control oocytes or rOAT1-expressing oocytes was measured for 1 h in the presence and absence of 1 mM probenecid. B, uptake of 4 µM [14C]AZT by control oocytes and rOAT1-expressing oocytes was determined with or without the preincubation with 1 mM glutarate for 2 h. C, time-dependent uptake of [14C]AZT via rOAT1. The uptake of 4 µM [14C]AZT was measured for 3 h in control oocytes and oocytes expressing rOAT1. Values represent mean ± S.E. of 10 to 16 determinations.

View larger version (19K): [in this window] [in a new window]   Fig. 3.   Transport of [3H]ACV via rOAT1. Uptake of 150 nM [3H]ACV by control oocytes and oocytes expressing rOAT1 (A), the effect of preincubation with glutarate (B), and time-dependent uptake (C) were measured in the same manner as described in the legend for Fig. 1. Values represent mean ± S.E. of 10 to 16 determinations.

Statistical Analysis. The transport kinetics was estimated from the following equation: V = V_max × S/(K_m + S), where V and S are the rOAT1-specific transport rate and the substrate concentration, respectively, and V_max and K_m are the maximum uptake rate and the Michaelis-Menten constant, respectively. This equation was fitted to the uptake data sets by an iterative nonlinear least-squares method using the MULTI program to obtain the kinetic parameters (Yamaoka et al., 1981). Data are expressed as mean ± S.E. Statistical differences were determined using Student's unpaired t test.

	Results

Top Abstract Introduction Experimental Procedures Results Discussion References

Figure 1 depicts the chemical structures of AZT and ACV, both of which possess no typical anionic moiety. Figure 2A shows the transport of [¹⁴C]AZT mediated by rOAT1 expressed in the X. laevis oocytes. rOAT1-expressing oocytes showed a significantly higher uptake of AZT than control oocytes, and the rOAT1-mediated uptake of AZT was totally abolished by probenecid. When the oocytes-expressing rOAT1 were preincubated with glutarate, the rate of AZT uptake via rOAT1 was significantly enhanced (Fig. 2B). Figure 2C shows that the time-dependent uptake of AZT by rOAT1 increased progressively up to 3 h. Figure 3 shows the transport of [³H]ACV mediated by rOAT1. Similarly, the oocytes expressing rOAT1 showed probenecid-sensitive uptake of ACV, which was significantly enhanced by the preincubation with 1 mM glutarate (Fig. 3, A and B). rOAT1-mediated uptake of ACV increased up to 3 h.

In the experiment, the results of which are shown in Fig. 4, we examined the kinetics of AZT and ACV transport via rOAT1. rOAT1-mediated uptake of AZT and ACV showed saturable kinetics, and the Eadie-Hofstee plot yielded a straight line for both compounds (Fig. 4, inset). The estimated K_m values for AZT and ACV were 68.0 ± 7.2 and 242 ± 16 µM, respectively, and the corresponding V_max values were 42.4 ± 3.2 and 25.3 ± 0.8 pmol/oocyte/h.

View larger version (17K): [in this window] [in a new window]   Fig. 4.   Concentration dependence of rOAT1-mediated uptake of AZT (A) and ACV (B). The uptake of varying concentrations of AZT (2-200 µM) and ACV (30-3000 µM) was measured for 1 h, and the rOAT1-specific uptake was plotted as a function of the concentrations. Values represent mean ± S.E. of 9 or 10 determinations. Inset, Eadie-Hofstee plots for each compound.

Figure 5 shows the inhibitory effect of various antiviral agents on rOAT1-mediated uptake of [¹⁴C]PAH. AZT (1 mM) strongly inhibited rOAT1-mediated uptake of PAH. The inhibitory effect of 1 mM ACV and trifluridine was moderate; ddC, ddI, d4T, amantadine, and vidarabine showed only slight, although statistically significant, inhibitory effect. By contrast, foscarnet (a phosphate analog antiherpes virus drug) did not show any inhibitory effect.

View larger version (49K): [in this window] [in a new window]   Fig. 5.   Inhibitory effects of various antiviral agents on rOAT1-mediated [14C]PAH uptake. Oocytes were incubated with 2 µM [14C]PAH in the presence or absence of 250, 500, and 1000 µM of an antiviral agent for 1 h. Values are expressed as percentage of the PAH uptake by rOAT1-expressing oocytes in the absence of inhibitors (mean ± S.E.; 6-10 determinations of a representative experiment). The absolute value of 2 µM [14C]PAH uptake in the control oocytes and rOAT1-expressing oocytes was 0.31 ± 0.02 and 4.07 ± 0.12 mol/oocyte/h, respectively. *P < .05, **P < .01, ***P < .001 versus without inhibitor.

Figure 6 shows the time-dependent uptake of [³H]ddC, [³H]ddI, [³H]lamivudine, [¹⁴C]d4T, [¹⁴C]trifluridine, and [¹⁴C]foscarnet by rOAT1. rOAT1-expressing oocytes showed a significantly higher uptake of [³H]ddC, [³H]ddI, [³H]lamivudine, [³H]d4T, and [¹⁴C]trifluridine than control oocytes. The uptake rate of these compounds via rOAT1 increased linearly for up to 3 h. In contrast, the uptake rate of [¹⁴C]foscarnet by the oocytes expressing rOAT1 was the same as that by control oocytes.

View larger version (29K): [in this window] [in a new window]   Fig. 6.   Time-dependent uptake of antiviral agents via rOAT1. The uptake of 30 nM [3H]ddC (A), 30 nM [3H]ddI (B), 100 nM [3H]3TC (C), 4 µM [14C]d4T (D), 4 µM [14C]trifluridine (E), and 4 µM [14C]foscarnet (F) was measured for 3 h in control oocytes and oocytes expressing rOAT1. Values represent mean ± S.E. of 13 to 15 determinations.

	Discussion

Top Abstract Introduction Experimental Procedures Results Discussion References

Wide substrate selectivity is a prominent feature of the PAH transporter. Ullrich and colleagues have systematically analyzed the interaction of a number of chemicals with the renal organic transporter in terms of the inhibitory constant using the stopped flow peritubular capillary microperfusion technique (Ullrich and Rumrich, 1993; Ullrich, 1997). They concluded that an appropriately sized hydrophobic domain and a negatively (or partially negatively) charged group or groups in the substrate are essential for their binding to the PAH transporter (Ullrich, 1997). The substrate selectivity of cloned rOAT1 was almost similar to that reported for the PAH transporter (Sekine et al., 1997; Uwai et al., 1998; Apiwattanakul et al., 1999; Jariyawat et al., 1999; Takeda et al., 1999; Tsuda et al., 1999). The results of the present study are of particular interest considering the chemical structures of AZT and ACV. Neither drug possesses a typical anionic moiety (e.g., carboxylic, sulfate, or phosphate group; Fig. 1). AZT has a high pK_a value (9.68; Chatton et al., 1990), and ACV is a zwitterionic compound (pK_a = 2.27 and 9.25; Laskin et al., 1982). In addition, the hydrophobicity of the two compounds is not very high. Thus, neither AZT nor ACV fits well with the above-mentioned structures being accepted by rOAT1. The present study also showed that other nucleoside analogs (e.g., zalcitabine, didanosine, lamivudine, stavudine, and trifluridine), all of which lack a phosphate moiety, are transported by rOAT1. In contrast, foscarnet, a simple inorganic phosphate analog without the structure of nucleosides, does not interact with rOAT1. Taken together, pyrimidine and purine rings are suggested to interact with rOAT1. The binding of nucleoside analogs to OAT1 is probably accomplished by hydrogen bonding interactions via the carbonyl groups and/or nitrogen atoms of the heterocyclic ring, presumably as enolate anions. Weak hydrophobic interactions appear to reinforce the binding. Recently, it was reported that OAT1 mediates the transport of cidofovir and adefovir, other types of nucleoside analogs (Cihlar et al., 1999). However, unlike ACV and AZT, cidofovir and adefovir possess a phosphate moiety. Although these two compounds are classified in the same groups as ACV and AZT, the charge interaction of cidofovir and adefovir with OAT1 might be mediated via a phosphate moiety .

In humans, renal clearance accounts for 83% of the total clearance of ACV (Laskin et al., 1982). The renal clearance of ACV is three times higher than that estimated to occur by glomerular filtration alone, and the administration of probenecid significantly decreases its renal clearance, suggesting tubular secretion of ACV (Laskin et al., 1982). In rats and mice, ACV is predominantly excreted in the urine in the unchanged form (de Miranda et al., 1981). In the case of AZT, there is a notable species difference in its pharmacokinetics. In the rat, AZT is excreted rapidly in the urine in the unchanged form (de Miranda et al., 1990). At low plasma concentrations, the renal clearance of AZT is nearly equal to the renal plasma flow, which indicates that the tubular secretion of AZT is efficient. In humans, AZT is extensively metabolized in the liver and excreted in the urine as AZT-glucuronide (de Miranda et al., 1989). The extent of AZT glucuronidation, expressed as V_max/K_m, in the liver microsomes is five to six times higher in the human than that in the rat liver (Cretton et al., 1990). However, the importance of renal clearance of AZT and AZT-glucuronide in humans has also been indicated (Deray et al., 1988). Probenecid depressed the renal elimination of AZT-glucuronide as well as the hepatic glucuronidation of AZT (Kornhauser et al., 1989). Thus, in both humans and rats, the tubular secretory process is a major determinant of the pharmacokinetics of ACV and an important factor in that of AZT. This study indicates that the basolateral uptake of AZT and ACV is mediated, at least in part, by OAT1. So far, several OAT isoforms, that is, OAT2 (Sekine et al., 1998), OAT3 (Kusuhara et al., 1999), and OAT4 (Cha et al., 2000), have been reported. Among these, only OAT1 possesses the transport properties of the classic PAH/dicarboxylate exchanger, and rOAT1 appears to play a major role in the basolateral uptake of AZT and ACV. At the moment, however, the transport of antiviral drugs via other OAT isoforms remains to be elucidated, and the relative contribution of OAT1 in the renal handling of antiviral drugs should be addressed in future studies.

The tubular secretory process of antiviral drugs appears to be involved in the development of their nephrotoxicity. It was reported that i.v. administration of a high dose of ACV (500 mg/m² for 5 days) resulted in an elevation of the serum concentration of creatinine in 48% of the recipients (Bean and Aeppli, 1985), although ACV is considered to be well tolerated at lower doses. This may be due to the obstructive nephropathy caused by the intratubular precipitation of ACV as a consequence of extensive tubular secretion of the drug (Sawyer et al., 1988). The accumulation of ACV in the proximal tubular cells may exacerbate its nephrotoxicity. Cidofovir and adefovir are also potently nephrotoxic (Lalezari et al., 1997; Cihlar et al., 1999); the mechanism has been attributed to their accumulation in the proximal tubular cells via the renal organic anion transporter. Thus, the transport of antiviral drugs in the proximal tubules, presumably via OAT1, is closely related to the development of nephrotoxicity. Not only in relation to the tissue-specific toxicity, but also in relation to the drug-drug interactions of antiviral drugs, OAT1 is presumed to be one of the key molecules. Because OAT1 shows a remarkably wide substrate selectivity, the concomitant use of high-affinity substrates (or inhibitors) of OAT1 may reduce the renal elimination of antiviral drugs. Antiviral drugs are toxic at high concentrations in the plasma, causing serious untoward effects such as neurotoxicity and bone marrow suppression (Richman et al., 1987). Elucidation of the transport properties of OAT1, therefore, would provide essential information on the safe use of antiviral drugs in clinical medicine.

The distribution of nucleoside analogs in the brain is a critical issue. Many children with HIV infection have progressive encephalopathy, and about two-thirds of adult patients exhibit dementia (Wong et al., 1993). Herpes encephalitis is an infrequent but life-threatening complication. For the prophylaxis and successful treatment of these complications, the distribution of antiviral agents in the brain is essential. In the rat, the concentrations of ACV and AZT in the brain have been reported to be much lower than those in the plasma (de Miranda et al., 1981, 1990). This observation is in part explained by the anatomic characteristics of the blood capillary vessels and the surrounding glial cells [blood-brain barrier (BBB)]. In addition, the involvement of organic anion transporter in the elimination (efflux) of antiviral drugs from the brain has been indicated. Probenecid decreased the clearance of AZT from the cerebrospinal fluid and thalamus extracellular fluid and prolonged the half-life of AZT disappearance from the brain (Wong et al., 1993). The elimination of other anionic substances, such as -lactam antibiotics and PAH, from the blood-cerebrospinal fluid barrier and/or BBB by a transporter or transporters has been also suggested (Ogawa et al., 1994; Kakee et al., 1997). So far, several organic anion transporter proteins, such as oatp1 (Angeletti et al., 1997), oatp2 (Abe et al., 1998), and Mrp1 (Nishino et al., 1999), have been shown to be localized in the choroid plexus. rOAT1 is also expressed in the brain, and AZT, -lactam antibiotics, and PAH are all the substrates of rOAT1. Although the precise localization of rOAT1 in the brain has not been clarified, rOAT1 is a strong candidate for the molecule mediating the elimination of AZT and other organic anions from the brain. rOAT3 is also expressed in the brain, and its localization in the choroid plexus has been suggested (Kusuhara et al., 1999). Thus, OAT isoforms may mediate the efflux of anionic drugs, including antiviral drugs, across the blood-cerebrospinal fluid barrier or/and BBB.

Molecular information on the transporters is now being sought for the use in drug development. For example, the rate of intestinal absorption of ACV is very low (de Miranda et al., 1981); the L-valyl ester of ACV, a prodrug (valacyclovir), possesses good bioavailability compared with that of ACV. This increased intestinal absorption is explained by the transport of valacyclovir by Pept 1 (peptide transporter 1; Ganapathy et al., 1998; Han et al., 1998). Thus, information on OAT1 and its isoforms may be applied to the development of drugs with more favorable kinetic profiles, such as drugs that do not accumulate in the kidney or have good distribution in the brain.

In conclusion, we demonstrated that rOAT1 mediates the transport of nucleoside analog antiviral drugs without a phosphate group. This study demonstrates further extension of the substrate spectrum of rOAT1 and suggests its potential role in the renal elimination of antiviral drugs.