Introduction

Top Abstract Introduction Materials Results Discussion References

The total body clearance of many NQ occurs mainly via metabolic elimination and urinary excretion, whereas the biliary clearance of newly developed quinolones such as GPFX and SPFX is greater than for other NQ (Matsunaga et al., 1991; Akiyama et al., 1995). As a first step in clarifying the mechanism that determines the degree of biliary clearance of NQ, we previously studied the hepatic uptake using isolated rat hepatocytes and found that NQ are taken up by the liver via a carrier-mediated active transport system (Sasabe et al., 1997). However, the biliary clearance was governed not only by the degree of hepatic uptake but also by the intrinsic ability of excretion into bile across the bile canalicular membrane. Therefore, in order to understand the mechanism for the effective hepatobiliary transport of GPFX, it is necessary to assess the transport system involved in the excretion process.

Some primary active transport systems driven by cellular ATP hydrolysis are known to mediate the biliary excretion of a number of compounds. Of these, 1) the first transporter is for bile acids (Inoue et al., 1984; Adachi et al., 1991; Nishida et al., 1991) such as TCA; 2) the second is for amphipathic organic cations and neutral compounds, including anticancer drugs (Kamimoto et al., 1989), called P-glycoprotein; 3) the third (cMOAT) is for organic anions, including a number of conjugates such as LTC₄ (Ishikawa et al., 1990; Sathirakul et al., 1994), an endogenous compound, and DNP-SG (Kobayashi et al., 1990), a glutathione conjugate, and some glucuronide conjugates of exogenous compounds (Shimamura et al., 1994; Takenaka et al., 1995a, 1995b). Multiplicity in transporters for organic anions has been recently investigated (Takenaka et al., 1995a; Yamazaki et al., 1996).

Mutant rats such as TR() (Jansen et al., 1985; Nishida et al., 1992a) and EHBR (Mikami et al., 1986), which have an inherited deficiency in their cMOAT, are a useful tool for investigating the transport mechanism involved in biliary excretion. Recently, using molecular biological techniques, the organic anion transporters from livers of Wistar and Sprague-Dawley strains of rat have been successfully cloned (Paulusma et al., 1996; Ito et al., 1997), and it has become clear that the transporter has two different ATP-binding regions. The origin of the molecular mutation in TR() and EHBR has also been identified (Mayer et al., 1995; Paulusma et al., 1996; Ito et al., 1997).

The biliary excretion of the glutathione conjugate LTC₄ and those of organic anions such as dibromosulfophthalein (DBSP) and the -lactam antibiotic cefodizime are markedly reduced in EHBR (Huber et al., 1987; Sathirakul et al., 1993). The studies using CMVs have directly demonstrated the ATP-dependent uptake of LTC₄ (Ishikawa et al., 1990; Fernandez-Checa et al., 1992), whereas there was no uptake of these compounds by CMV prepared from mutant rats, TR() and EHBR, although the ATP-dependent uptake of TCA remained normal (Takenaka et al., 1995a). These results suggest that biliary excretion of LTC₄ is mediated by cMOAT. Moreover, it has been reported that the glucuronides of some drugs are actively excreted into bile by cMOAT, although this transporter does not recognize the sulfate conjugates of these same drugs (Kobayashi et al., 1991; Shimamura et al., 1994; Takenaka et al., 1995a, 1995b).

The membrane vesicle experiments are very useful for transporter studies. For example, the vesicle studies of kidney and intestine indicated the possible presence of transporters that mediate the excretion of ofloxacin into urine and the absorption of enoxacin via the brush-border membrane of each organ (Okano et al., 1990; Iseki et al., 1992; Hirano et al., 1994). However, the excretion mechanism by which NQ are transported across the bile canalicular membrane has never been investigated until now.

In order to understand how the active transport system for organic anions mediates excretion of NQ from liver cells to bile, we selected GPFX, which has a greater hepatobiliary transport than other NQ, and performed the in vivo study and the in vitro study using CMV prepared from normal rats and EHBR. GPFX is known to undergo hydroxylation, cleavage of its piperadine ring and conjugation to the glucuronide or the sulfate in the liver (fig. 1). Because both parent and metabolites are excreted into bile (Akiyama et al., 1995), we also studied the biliary excretion mechanism of the 3-glucuronide, the main metabolite of GPFX, as well as that of GPFX itself.

View larger version (15K): [in this window] [in a new window]   Fig. 1.   Chemical structure of GPFX and its metabolic pathway.

	Methods and Materials

Top Abstract Introduction Materials Results Discussion References

Chemicals. [¹⁴C]GPFX (1.17 MBq/µmol, radiochemical purity 97.1%) was obtained from Amersham International (Buckinghamshire, UK). [³H]S-DNP-SG was synthesized from 1-chloro-2,4-dinitrobenzene and radiolabeled glutathione (GSH) in the presence of glutathione S-transferase by the method described by Kobayashi et al. (1990). [³H]S-GSH (1.62 MBq/µmol, 99.9%) and [³H]TCA, (128 MBq/µmol, 98.5%) were purchased from New England Nuclear Corp (NEN, Boston, MA). The [¹⁴C]3-glucuronide of GPFX was isolated from the bile of rats given i.v. [¹⁴C]GPFX at a dose of 3.7 MBq/3.2 µmol/rat (1.7 mg/rat). Unlabeled 3-glucuronide was obtained from the bile of rats receiving a constant infusion (57 µg/69.6 nmol/min/kg). Bile (1 ml) was applied to a pretreatment column (Bond Elut C18, 3 ml/500 mg, Varian, CA), eluted with methanol-water (15:85, v/v) and the eluate injected into an HPLC column to purify the 3-glucuronide. Unlabeled GPFX, SPFX, LFLX, CPFX, ENX and OFLX were synthesized or purified by Otsuka Pharmaceutical company (Tokyo, Japan). ATP, creatine phosphate and creatine phosphokinase were purchased from Sigma Chemical Co. (St. Louis, MO).

In vivo study. Male Sprague-Dawley rats (Nihon Ikagaku, Tokyo, Japan) and EHBR (SLC Japan, Shizuoka) weighing approximately 250 to 300 g were used throughout the experiments. Under light ether anesthesia, the femoral artery and the femoral vein were cannulated with polyethylene catheters (PE-50) for blood sampling and injection of GPFX, respectively. The bile duct was cannulated with a polyethylene catheter (PE-10) for bile collection; both ureters were also cannulated to collect urine. Radiolabeled GPFX was given i.v. at a dose of 5 mg/kg (788 KBq/13.9 µmol/kg). Bile was collected in preweighed test tubes at 10-min intervals until 1 hr after administration and then at 30 min intervals from 1 to 2 hr after administration. Urine was collected in test tubes at 30-min intervals throughout the experiment. Plasma was prepared by centrifugation of blood samples (10,000 × g, Microfuge, Beckman, Fullerton, CA).

TLC analysis. The metabolites of GPFX were purified from rat bile and human urine after p.o. administration of GPFX at a dose of 40 mg/kg and 400 mg/body, respectively. Then they were identified using mass spectrometry and ¹H-NMR spectrometry as described by Akiyama et al. (1995). The metabolites in rats and humans were the same except for one minor metabolite.

Preparation of CMVs. CMV were prepared from male SDR and EHBR using a slight modification of the method of Meier et al. (1984). After suspension of vesicles in 50 mM Tris-HCl buffer (pH 7.4) containing 250 mM sucrose, the vesicles were frozen in liquid N₂ and stored at 80°C until required. The transport activity of CMV used in this study was also checked by measuring the ATP-dependent uptake of standard substrates, [³H]TCA (1 µM) and [³H]DNP-SG (1 µM), for a 2-min incubation period at 37°C. Protein was determined by the method of Bradford (1976) with bovine serum albumin as a standard, using the Bio-Rad protein assay kit (Bio-Rad, Hercules, CA).

Uptake of [¹⁴C]GPFX and [¹⁴C]GPFX-glucuronide by CMV. The uptake of [¹⁴C]GPFX or [¹⁴C]GPFX-glucuronide was measured by the rapid-filtration technique described by Ishikawa et al. (1990). The transport medium (10 mM Tris-HCl buffer (pH 7.4)) containing 250 mM sucrose and 10 mM MgCl₂ contained 50 µM [¹⁴C]GPFX or [¹⁴C]GPFX-glucuronide, 5 mM ATP and ATP-regenerating system (10 mM creatine phosphate and 100 µg/ml creatine phosphokinase). The transport medium (final 40 µl) was mixed with 10 µl of vesicle suspension (20 µg protein) and incubated at 37°C. The uptake reaction was stopped by addition of 1 ml of ice-cold buffer containing 100 mM NaCl, 250 mM sucrose and 10 mM Tris-HCl (pH 7.4) at designated times. This reaction mixture (900 µl) was then filtered through a 0.45-µm HAWP filter (Millipore Corp., Bedford, MA) and washed twice with 5 ml of ice-cold buffer. In the uptake study of GPFX, the filter was steeped in 1% bovine serum albumin solution for one night before the experiment in order to reduce nonspecific adsorption of GPFX to the filter. Radioactivity retained on the filter was determined by liquid scintillation counting. The uptakes of these compounds by CMV were normalized with respect to both the amount of vesicles and the concentration of ligand in the reaction mixture. In the saturation study of the uptake of GPFX-glucuronide, the transport medium contained 5, 10, 20 and 50 µM [¹⁴C]GPFX-glucuronide. For higher concentrations, such as the final 100, 200 and 1000 µM, the ligand solution was prepared with radiolabeled (50 µM) and unlabeled GPFX-glucuronide.

Uptake of [³H]DNP-SG and [³H]TCA by CMV. The uptake of [³H]DNP-SG (1 µM) and that of [³H]TCA (1 µM) were examined by the same method as for GPFX-glucuronide except for a change in volume of the incubation mixture (final 20 µl) and amount of vesicles (10 µg protein). The uptake reaction was stopped at 2 min. For the inhibition of the uptake of [³H]DNP-SG, the reaction mixture was incubated in the presence of some NQ at concentrations of 0.1, 0.3, 1, 3 and 10 mM or of GPFX-glucuronide at concentrations of 0.1, 1, 10 and 100 µM and 1, 3 and 10 mM.

Estimation of kinetic parameters. For the ATP-dependent uptake of GPFX-glucuronide by CMV, which was obtained by subtracting the uptake in the absence of ATP from that uptake in its presence, the following equation was used.

(1)

Where V₀ is the initial uptake rate of the drug (pmol/min/mg protein), S is the drug concentration in the medium (µM), K_m1 and K_m2 are the Michaelis constants (µM), and V_max1 and V_max2 are the maximum uptake rates (pmol/min/mg protein). The foregoing equation was fitted to the uptake data set by an iterative nonlinear least-squares method using a MULTI program (Yamaoka et al., 1981) in order to obtain estimates of kinetic parameters. The input data were weighted as the reciprocal of the observed values, and the Damping Gauss Newton Method algorithm was used for fitting.

(2)

	Results

Top Abstract Introduction Materials Results Discussion References

Biliary excretion profiles of GPFX and GPFX-glucuronide in normal rats and EHBR. After a single i.v. administration of GPFX at a dose of 5 mg/kg to normal rats and EHBR, biliary excretion profiles were compared. The biliary excretion of GPFX in EHBR was 38% of that in normal rats (fig. 2A), whereas 3-glucuronide of GPFX, which is a main metabolite of GPFX, underwent very little excretion into the bile in EHBR (fig. 2B). Combined excretions of M-1 and M-2 with a cleaved piperadine ring exhibited a pattern similar to that of unchanged drug in terms of the comparison between normal rats and EHBR (fig. 2C). There was no difference in the biliary excretion of its 4-sulfate between normal rats and EHBR (fig. 2D).

View larger version (28K): [in this window] [in a new window]   Fig. 2.   Time profiles of the biliary excretion of GPFX and its metabolites after a single i.v. dose of [14C]GPFX to normal rats and EHBR. [14C]GPFX was given i.v. to normal rats () and EHBR () at a dose of 5 mg/kg including 0.79 mBq/kg. Parent (panel a), 3-glucuronide (panel b), M-1 and M-2 (panel c) and 4-sulfate (panel d) in bile samples were separated by TLC. Each point represents the mean ± S.E. of three rats. * P < .05, ** P < .01 (significantly different from the excretion over 120 min in normal rats using Student's t test).

Urinary excretion profiles of GPFX and GPFX-glucuronide in normal rats and EHBR. No difference in the urinary excretion of GPFX was observed between normal rats and EHBR (table 1). 4-Sulfate, M-1 and M-2 also showed no difference in urinary excretion between normal rats and EHBR, whereas the urinary excretion (0-120 min) of GPFX-glucuronide in EHBR increased 3.6-fold compared with that in normal rats.

Plasma concentration profiles of GPFX and GPFX-glucuronide in normal rats and EHBR. The plasma concentrations of unchanged GPFX in EHBR were slightly higher than those in normal rats beginning 15 min after administration, but no significant difference was observed between their AUCs from time 0 to 120 min (fig. 3A). The plasma concentrations of GPFX-glucuronide were higher in EHBR than in normal rats (fig. 3B). The biliary clearance of GPFX, based on the plasma concentrations, was 1.79 ± 0.05 and 0.52 ± 0.01 ml/min/kg in normal rats and EHBR, respectively. In addition, the biliary clearance of GPFX-glucuronide in normal rats and EHBR was 15.53 ± 0.90 and 0.09 ± 0.01 ml/min/kg, respectively (table 2). The urinary clearance of unchanged GPFX was 0.85 ± 0.41 and 0.51 ± 0.11 ml/min/kg in normal rats and EHBR, respectively, and that of the glucuronide was 0.86 ± 0.36 and 0.65 ± 0.13 ml/min/kg. Normal rats and EHBR did not exhibit any statistically significant difference in urinary clearance.

View larger version (14K): [in this window] [in a new window]   Fig. 3.   Time profiles of the plasma concentrations of GPFX and its glucuronide after a single i.v. dose of [14C]GPFX to normal rats and EHBR. [14C]GPFX was given i.v. to normal rats () and EHBR () at a dose of 5 mg/kg including 0.79 mBq/kg. Parent (panel a) and its glucuronide (panel b) in plasma samples were separated by TLC. The plasma samples from three rats were mixed in order to carry out the determination.

Uptake profiles of GPFX by CMV prepared from normal rats and EHBR. The studies were performed using CMV prepared from rat liver to clarify the transport mechanism for the biliary excretion of GPFX. The uptake of GPFX by CMV prepared from normal rats in the presence of ATP and ATP-regenerating system was significantly greater than that in their absence, which indicates that GPFX uptake involves ATP-dependent uptake (fig. 4A). However, no ATP-dependent uptake by CMV in EHBR was observed (fig. 4B).

View larger version (16K): [in this window] [in a new window]   Fig. 4.   Uptake profiles of GPFX by canalicular membrane vesicles prepared from normal rats (panel a) and EHBR (panel b). Uptake of GPFX was determined by incubating the vesicles (20 µg) at 37°C for 2, 5, 15 and 30 min after the addition of GPFX. The incubation mixture (40 µl) contained 50 µM [14C]GPFX with () or without () 5 mM ATP, 10 mM creatine phosphate and 100 µg/ml creatine phosphokinase. Uptake was calculated by dividing the amount in vesicles by the GPFX concentration in the medium. Each plot represents the mean ± S.E. of 6 to 23 determinations from four different experiments for normal rats and 15 determinations from three different experiments for EHBR. * P < .05, ** P < .01 (significantly different from the uptake in the corresponding ATP() group using Student's t test).

Inhibition of [³H]DNP-SG transport into CMV by NQ. The inhibition study involving the CMV uptake of [³H]DNP-SG, a typical substrate for cMOAT, was performed to differentiate the affinities of different NQ for cMOAT. Every NQ inhibited in a concentration-dependent manner the ATP-dependent transport of [³H]DNP-SG into CMV (fig. 5). GPFX and SPFX, which exhibit relatively greater hepatobiliary excretion, had greater inhibitory effects; they inhibited the [³H]DNP-SG uptake to 54.7 ± 1.7% and 60.0 ± 4.6% of the control, respectively, at a concentration of 3 mM.

View larger version (62K): [in this window] [in a new window]   Fig. 5.   Inhibitory effects of NQ on DNP-SG uptake by CMV prepared from normal rats. Uptake of DNP-SG was determined by incubating the vesicles (10 µg) at 37°C for 2 min after the addition of [3H]DNP-SG. The incubation mixture (20 µl) contained 1 µM [3H]DNP-SG and the quinolones as inhibitor, with or without 5 mM ATP, 10 mM creatine phosphate and 100 µg/ml creatine phosphokinase. Each column and vertical bar represent the mean ± S.E. from 3 to 7 determinations in two different experiments. * P < .05, ** P < .01 (significantly different from the uptake in the control group using Dunnett's test).

Uptake profiles of GPFX-glucuronide by CMV prepared from normal rats and EHBR. The uptake of GPFX-glucuronide by CMV prepared from normal rats exhibited marked ATP dependence and an overshoot phenomenon (fig. 6). The ATP-dependent uptake of GPFX-glucuronide by CMV prepared from EHBR was markedly reduced compared with CMV prepared from normal rats (table 3). The ATP-dependent uptake showed a saturable phenomenon. When the data were converted into Eadie-Hofstee plots, this saturable component included higher- and lower-affinity components (fig. 7). This uptake provided a K_m1 of 7.2 ± 2.4 µM and a V_max1 of 0.460 ± 0.110 nmol/min/mg protein for the high-affinity component and a K_m2 of 169 ± 110 µM and a V_max2 of 0.657 ± 0.102 nmol/min/mg protein for the low-affinity component.

View larger version (19K): [in this window] [in a new window]   Fig. 6.   Uptake profiles of GPFX-glucuronide by CMV prepared from normal rats. Uptake of GPFX-glucuronide was determined by incubating the vesicles (20 µg) at 37°C for 0.25, 2, 5, 15, 30, 60, 120 and 180 min after the addition of [14C]GPFX-glucuronide. The mixture (40 µl) contained 50 µM [14C]GPFX-glucuronide with () or without () 5 mM ATP, 10 mM creatine phosphate and 100 µg/ml creatine phosphokinase. Uptake was calculated by dividing the amount in vesicles by the GPFX-glucuronide concentration in the medium. Each plot represents the mean ± S.E. of three determinations.

View larger version (20K): [in this window] [in a new window]   Fig. 7.   Eadie-Hofstee plot of GPFX-glucuronide uptake by CMV prepared from normal rats. Uptake of GPFX-glucuronide was determined by incubating the vesicles (20 µg) at 37°C for 2 min after the addition of [14C]GPFX-glucuronide. The mixture (40 µl) contained 5, 10, 20, 50, 100, 200 and 1000 µM [14C]GPFX-glucuronide with () or without () 5 mM ATP, 10 mM creatine phosphate and 100 µg/ml creatine phosphokinase. ATP-dependent uptake () was calculated by subtracting ATP() uptake from the corresponding ATP(+) uptake. The solid line was obtained as follows. First, ATP-dependent uptake data were fitted to equation 1, and then the fitted line was converted to the V0/S vs. V0 form (Eadie-Hofstee plot). Each plot represents the mean ± S.E. of three determinations.

Inhibition of [³H]DNP-SG uptake by GPFX and GPFX-glucuronide. The GPFX-glucuronide inhibited the ATP-dependent uptake of [³H]DNP-SG by CMV in a concentration-dependent manner (fig. 8). The glucuronide in this inhibition study has a K_i of 9.2 ± 1.7 µM, which is comparable to its own K_m (7.2 µM) for the higher-affinity component of ATP-dependent uptake. The K_i of the parent drug obtained in the same study was 1.89 ± 0.80 mM, which demonstrates that the affinity of the glucuronide is approximately 200 times higher than that of the parent (fig. 8).

View larger version (20K): [in this window] [in a new window]   Fig. 8.   Effect of GPFX and GPFX-glucuronide on the ATP-dependent uptake of DNP-SG by CMV prepared from normal rats. Uptake of DNP-SG was determined by incubating the vesicles (10 µg) at 37°C for 2 min after the addition of [3H]DNP-SG. The incubation mixture (20 µl) contained 1 µM [3H]DNP-SG and GPFX-glucuronide () 0, 0.0001, 0.001, 0.01, 0.1, 1, 3 and 10 mM or GPFX () 0.1, 1, 3 and 10 mM with or without 5 mM ATP, 10 mM creatine phosphate and 100 µg/ml creatine phosphokinase. ATP-dependent uptake was calculated by subtracting ATP() uptake from the corresponding ATP(+) uptake. The solid line was obtained as follows. First data were fitted to equation 2, and then the fitted line was converted to the V0/S vs. I form. Each plot and vertical bar represent the mean ± S.E. of three determinations.

	Discussion

Top Abstract Introduction Materials Results Discussion References

It has been proved that the carrier-mediated active transporter contributes to the transport of organic anions from liver cells to bile. Detailed studies of the organic anion transport have been performed using mutant rat strains, TR() and EHBR, in which cMOAT is hereditarily defective with respect to biliary excretion (Ishikawa et al., 1990; Sathirakul et al., 1993, 1994; Yamazaki et al., 1996). In the present study, the biliary excretion of GPFX in EHBR was reduced to approximately one-third that in normal rats (fig. 2A). A main metabolite of GPFX, the 3-glucuronide, was scarcely excreted into bile in EHBR (fig. 2B), which indicates that cMOAT mediates the major portion of its biliary excretion. Such reduced biliary excretion in EHBR has also been observed for the biliary excretion of nonconjugated organic anion, DBSP, glutathion conjugates, LTC₄, BSP-GSH and glucuronides of bilirubin, liquiritigenin and E-3040, a dual inhibitor of 5-lipoxygenase and thromboxane A₂ (Huber et al., 1987; Jansen et al., 1987, 1993; Kobayashi et al., 1991; Nishida et al., 1992b; Sathirakul et al., 1993; Shimamura et al., 1994; Takenaka et al., 1995a,b).

The plasma concentration profile of parent drug in EHBR did not differ substantially from that in normal rats (fig. 3A), whereas GPFX-glucuronide concentration increased with time in EHBR (fig. 3B). This increase is thought to be due to back flux of the glucuronide produced from the liver cells to the plasma compartment, caused by a marked reduction in the biliary excretion of the glucuronide.

The urinary excretion of GPFX-glucuronide increased in EHBR (table 1). Inasmuch as the urinary excretion clearance of GPFX did not change in EHBR (table 2), this increase is thought to reflect the increase in the plasma concentration of GPFX-glucuronide that came from the reduction in its biliary excretion.

It may be that the reduction in the biliary excretion of the glucuronide in mutant rats is caused by a deficiency not only in biliary excretion ability but also in conjugating ability. Moreover, when some endogenous substances such as bilirubin glucuronide cannot be excreted into the bile in the mutant rats, and do accumulate in liver cells, the biliary excretion of exogenously administered anionic compounds may also be reduced by secondary effects, the endogenous substances inhibiting the transport. Jansen et al. (1987) and Elferink et al. (1989) have investigated this using a double-mutant rat derived from TR() and Gunn rats that has an inherited deficiency in its conjugation with glucuronic acid. Although the double-mutant rat did not exhibit cholestasis of conjugated bilirubin, the biliary excretion of DNP-SG and BSP was reduced in this rat, as in the TR() strain. These results rule out the possibility of secondary effects and suggest that the reduction in biliary excretion of these organic anionic compounds is directly caused by the deficiency in cMOAT in the mutant rat. Furthermore, ATP dependence was observed in the uptake of these compounds by CMV (Nishida et al., 1992a,b), and this ATP-dependent uptake was reduced markedly in CMV prepared from EHBR, which indicates that the biliary excretion is by primary active transport. Therefore, the reduction in the biliary excretion of GPFX and its glucuronide in EHBR strongly suggests that their excretion is mediated by cMOAT (fig. 2A,B).

To further clarify the mechanism for the biliary excretion of GPFX, we examined the uptake by CMV prepared from normal rats and EHBR. We observed ATP-dependent uptake only by CMV from normal rats and not from EHBR (fig. 4), which indicates that the biliary excretion of GPFX may be mediated at least partly by cMOAT if we consider the in vivo results (fig. 2A) together. However, this small ATP-dependent uptake indicates that GPFX may have a relatively weak affinity for cMOAT.

Differences in the contribution of biliary clearance to the total clearance of NQ may be due to variations in the affinity for the transport systems involved. Therefore, in order to differentiate the affinity of NQ for cMOAT, we tested the inhibitory effects of NQ on the CMV uptake of [³H]DNP-SG; this is a glutathione conjugate reported to be transported exclusively by cMOAT (Niinuma et al., 1997). GPFX with a K_i of 1.89 mM had the highest affinity for cMOAT among the NQ tested (fig. 5). Although the CMV uptake of NQ apart from GPFX has not been examined, these results may suggest that the other NQ could be substrates for cMOAT. GPFX and SPFX have been classified as to the biliary excretion type of NQ (Sekine, 1991; Akiyama et al., 1995). GPFX and SPFX had greater inhibitory effects than the urinary excretion type, LFLX, which suggests that the affinity of NQ for cMOAT may be one of the factors that determines the degree of biliary clearance. On the other hand, GPFX and SPFX also have a higher affinity for the transporter that mediates uptake into liver cells, whereas LFLX and CPFX have a lower affinity (Sasabe et al., 1997). The difference in the biliary clearance of NQ might thus stem from a differences in affinity for both uptake and excretion transporters among NQ.

GPFX is conjugated to glucuronide and sulfate at the 3-carboxyl group and 4-amino group in the piperazine ring, respectively (fig. 1). These conjugations are known as major metabolites of GPFX (Akiyama et al., 1995). Urinary excretion of the GPFX-glucuronide was much smaller than its biliary excretion (table 1), which suggests that the glucuronide produced is excreted mainly into bile. Biliary excretion of this 3-glucuronide of GPFX was scarcely observed in EHBR, whereas that of the 4-sulfate was similar to that in normal rats (fig. 2). These phenomena were also observed in earlier results from our laboratory involving the biliary excretion of E-3040 conjugates (Takenaka et al., 1995a,b). The glucuronide was excreted by cMOAT, whereas the sulfate was excreted by another transporter (Takenaka et al., 1995a,b). Moreover, the uptake of the GPFX-glucuronide by CMV consisted predominantly of ATP-dependent transport (fig. 6), which was not observed to any great extent in CMV from EHBR (table 3), which indicates that a large fraction of the biliary excretion was mediated by cMOAT. The K_i value (9.2 µM, fig. 8) of the glucuronide for the uptake of [³H]DNP-SG by CMV was comparable with the K_m value (7.2 µM) of the high-affinity component in its own ATP-dependent uptake (fig. 7). Therefore, GPFX-glucuronide and DNP-SG may share the same transporter, cMOAT. The K_i value (9.2 µM) of the glucuronide was less than 1/200 that of the parent (1.89 mM, fig. 8), which suggests that the glucuronide had a much higher affinity for cMOAT. Moreover, these results from the in vitro study correlated with those from the in vivo study in which 29% of the biliary excretion of the parent remained and minimal glucuronide was excreted into bile in EHBR (table 2).

However, small but significant ATP-dependent uptake of GPFX-glucuronide was observed in EHBR CMV (approximately one-fifth that in normal rats, table 3). We have already suggested that the glucuronide of E-3040 was taken up into CMV not only by cMOAT but also by the other transporter that exists in EHBR, an interpretation based on the results of mutual inhibition studies of DNP-SG and E-3040 glucuronide (Niinuma et al., 1997). Thus we propose that multiplicity exists in the biliary transporter for organic anions. The multiplicity may also be evident in this study using CMV in that the biliary excretion of GPFX-glucuronide may also be partially mediated by a transporter that was present in EHBR. However, this assumption seems to be inconsistent with the in vivo result that GPFX-glucuronide was scarcely excreted into bile in EHBR (table 2). This apparent discrepancy, however, may be resolved by assuming that some endogenous substance, such as bilirubin glucuronide, which accumulates in liver cells of EHBR, may inhibit excretion under in vivo conditions.

At least part of the biliary excretion of GPFX and a large portion of the main metabolite 3-glucuronide were shown to occur by primary active transport mediated by cMOAT with respect to hepatobiliary transport. Moreover, the affinity of NQ for cMOAT may be one of the factors that determines the degree of biliary clearance.