Introduction

Top Abstract Introduction Materials & Methods Results Discussion References

CPT, a plant alkaloid isolated from a tree found in China (Camptotheca accuminata), is a novel antitumor agent that exerts its activity exclusively by inhibition of topoisomerase I (Kim et al., 1992; Slichenmyer et al., 1993; Tanizawa et al., 1994). However, clinical evaluation of CPT was discontinued due to its unpredictably severe toxicity and poor water-solubility. Recently, several semisynthetic analogs of CPT have been developed (Miyasaka et al., 1981; Nagata, et al., 1987). Irinotecan (CPT-11) is a water-soluble analog of CPT that was discovered in an attempt to identify derivatives with greater water-solubility and antitumor activity than CPT (Hertzberg et al., 1989). CPT-11 acts as a prodrug that undergoes deesterification in vivo to yield SN-38, a metabolite that is 1000-fold more potent than the parent compound in vitro (Kawato et al., 1991; Kojima et al., 1993).

CPT-11 shows potent anticancer activity in many types of human tumor cells, including small-cell and non-small-cell lung cancers, malignant lymphoma, cervical cancer, ovarian cancer and colorectal cancer (Ohno et al., 1990; Negoro et al., 1991; Sasaki et al., 1995a). It has undergone phase I and II clinical trials in several countries. The major toxic effects of CPT-11 are myelosuppression and gastrointestinal toxicity, especially unpredictable, severe diarrhea (Araki et al., 1993). Such intestinal toxicity, however, exhibits large interpatient variability (Rothenberg et al., 1993; Rowinsky et al., 1994; Sasaki et al., 1995b), the mechanisms for which are currently unknown. One postulated mechanism for the toxicity of CPT-11 is related to the biliary excretion of its metabolites. After the administration of CPT-11, the active metabolite SN-38 is formed by deesterification. SN-38 is further conjugated to SN38-Glu in the liver and mainly excreted via the bile duct, followed by deconjugation by intestinal microflora to regenerate SN-38, which may cause diarrhea (Kaneda et al., 1990). However, the biliary excretion mechanism of CPT-11 and its metabolites has not been identified.

CPT-11 and its metabolites have an -hydroxy--lactone ring, which undergoes reversible hydrolysis at a rate that depends on many factors, including pH, ionic strength and protein concentration (Fassberg and Stella, 1992; Burke and Mi, 1994; Rivory and Robert, 1994) (fig. 1). The lactone form is essential for the activity of SN-38, and the lactone-hydrolyzed form (carboxylate form) exhibits only minimal topoisomerase I-inhibitory activity (Slichenmyer et al., 1993). At physiological pH, however, the lactone form is unstable and the equilibrium favors hydrolysis to open the lactone ring and yield the carboxylate form. Under acidic conditions, the reverse reaction, with formation of the lactone, is favored. Similar reactions also occur with CPT-11 and SN38-Glu (fig. 1). Among these six compounds, anionic charges are present on the carboxylate forms of CPT-11 and SN-38 and the carboxylate and lactone forms of SN38-Glu. Recently, it has been proposed that the hepatic cMOAT, which is expressed on the bile canalicular membrane, transports several types of organic anions into the bile as a primary active transport system (Ishikawa et al., 1990; Nishida et al., 1992; Sathirakul et al., 1993, 1994; Shimamura et al., 1994; Pikula et al., 1994; Yamazaki et al., 1996). Therefore, the four CPT-11-related compounds with an anionic charge may be recognized by the cMOAT. In this study, the biliary excretion mechanisms of both the carboxylate and lactone forms of CPT-11 and its metabolites were investigated in rats. EHBR derived from SD rats were used for this purpose; EHBR are genetically deficient with respect to cMOAT on the bile canalicular membrane (Fernandez-Checa et al., 1992; Takenaka et al., 1995a). The biliary excretion of CPT-11 and its metabolites was compared in SD rats and EHBR in vivo. The isolated bile CMV from SD rats and EHBR were also used for the transport study of CPT-11 and its metabolites.

View larger version (30K): [in this window] [in a new window]   Fig. 1.   Chemical structures of the lactone and carboxylate forms of CPT-11 and its metabolites SN-38 and SN38-Glu.

	Materials and Methods

Top Abstract Introduction Materials & Methods Results Discussion References

Materials

CPT-11, SN-38 and SN38-Glu were obtained from Daiichi Pharmaceutical Co. Ltd. (Tokyo, Japan) and the Yakult Honsha Co. Ltd. (Tokyo, Japan). The lactone and carboxylate forms of CPT-11, SN-38 and SN38-Glu were produced by dissolving the compounds in 50 mM phosphate buffer at pH 3.0 or 9.0 and leaving them overnight. The conversion of the lactones into carboxylates or carboxylates into lactones was virtually complete (>99%), as determined by HPLC. CPT, as an internal standard, was obtained from Sigma Chemical Co. (St. Louis, MO). [³H]DNP-SG (50.0 µCi/nmol) was synthesized according to the method described previously (Kobayashi et al., 1990). All other chemicals were commercial products and of analytical grade. Male SD rats weighing 250 to 300 g were purchased from Nisseizai (Tokyo, Japan), and male EHBR weighing 250 to 300 g were supplied by Eisai Laboratories (Gifu, Japan).

In Vivo Study

The SD rats and EHBR underwent cannulation of the femoral vein and artery using PE50 polyethylene tubing (inner diameter, 0.58 mm; outer diameter, 0.965 mm; Becton Dickinson & Co., Parsippany, NJ), and the bile duct was cannulated with PE10 polyethylene tubing (inner diameter, 0.28 mm; outer diameter, 0.61 mm; Becton Dickinson) under light anesthesia with ether. After the operation, the rats were kept in Bollman cages, with free access to food and water. The temperatures of the rats were monitored and found to remain constant during the experiment. CPT-11 was dissolved in distilled water with sonication and then diluted with 9% NaCl to give a final NaCl concentration of 0.9%. After i.v. injection of 10 or 40 mg/kg levels of the lactone form of CPT-11 (1 ml/250 g b.wt.), approximately 0.5 ml of blood was collected at 5 min, 30 min, 2 hr, 4 hr and 8 hr, transferred to centrifuge tubes containing 5 µl of 10 mM diisopropyl fluorophosphate dissolved in dimethylsulfoxide, kept on ice during the sampling period and finally centrifuged at 4°C. The plasma (0.2 ml) obtained was frozen under liquid N₂ immediately after harvesting. Bile samples were collected at 0 to 1, 1 to 2, 2 to 4, 4 to 6 and 6 to 8 hr; urine samples were collected spontaneously. To avoid conversion of the lactone and carboxylate forms of CPT-11 and its metabolites during sampling, bile and urine samples were collected every 2 hr, kept on ice during sampling and frozen immediately in liquid N₂ when sampling was complete. At 8 hr, the rats were sacrificed, and the livers were removed and immediately frozen in liquid N₂. Concentrations of both the lactone and carboxylate forms of CPT-11 and its metabolites were determined by HPLC, as described below. The urine samples obtained up to 8 hr were combined, and the total volume was determined before HPLC determination.

In Vitro Uptake by CMV

CMV were prepared from male SD rats and EHBR as described previously (Kobayashi et al., 1990). The CMV were resuspended in 250 mM sucrose containing 50 mM Tris-HCl (pH 7.4), frozen in liquid N₂ and stored at 100°C until use. The purity of the prepared CMV was evaluated by determining the activities of Mg⁺⁺-ATPase and alkaline phosphatase, according to the methods of Scharschmidt et al. (1979) and Yachi et al. (1989), respectively. The activity of prepared CMV was also checked by measuring the ATP-dependent uptake of standard substrates, [³H]taurocholate (1 µM) and [³H]DNP-SG (1 µM; 0.1 µM labeled and 0.9 µM unlabeled), at 37°C for 2 min.

Uptake of the carboxylate and lactone forms of CPT-11 and its metabolites. The incubation medium for the uptake studies contained 0.25 M sucrose, 10 mM Tris-HCl (pH 7.4), 10 mM MgCl₂, 5 mM ATP, 10 mM creatine phosphate and 100 µg/ml creatine phosphokinase. Control experiments were performed without the addition of ATP. The uptake study was performed at 37°C. After preincubation for 2 min, 5 µl of the carboxylate or lactone form of CPT-11 or its metabolites, which had been diluted with incubation medium to give a final pH of 7.4, was added to give a final ligand concentration of 50 µM. After preincubation for an additional 1 min, uptake was started by the addition of vesicles to give a final protein concentration of 1 mg/ml; the final incubation volume was 20 µl. At designated times, transport was terminated by addition of 1 ml of ice-cold stop solution, followed immediately by filtration through premoistened 0.45-µm HA Millipore filters (catalog no. HAWP 02500; Millipore Corp., Bedford, MA), which were subsequently washed twice with an additional 5.0 ml of ice-cold stop solution. The stop solution contained 10 mM Tris-HCl (pH 7.4), 0.25 M sucrose and 0.1 M NaCl. Filters were dried, cut into small pieces and frozen immediately under liquid N₂ for HPLC analysis. The nonspecific binding of drugs to the filters was also determined without CMV, and values for the CMV uptake were obtained by subtracting this nonspecific binding from the apparent uptake. Such nonspecific binding to the filter, normalized by the protein amount in the equivalent volume of incubation mixture with CMV, was 927 ± 57 and 893 ± 50 pmol/mg protein for the lactone form of CPT-11 in the presence and absence of ATP; 582 ± 9 and 554 ± 35 pmol/mg protein for the carboxylate form of CPT-11 in the presence and absence of ATP; 10.8 ± 5.5 and 9.09 ± 3.63 pmol/mg protein for the carboxylate form of SN-38 in the presence and absence of ATP; 19.5 ± 2.8 and 22.1 ± 0.7 pmol/mg protein for the lactone form of SN38-Glu in the presence and absence of ATP; 11.7 ± 0.9 and 12.4 ± 1.4 pmol/mg protein for the carboxylate form of SN38-Glu in the presence and absence of ATP; and 2.57 ± 0.55 and 2.47 ± 0.31 pmol/mg protein for [³H]DNP-SG in the presence and absence of ATP.

Effect of the lactone and carboxylate forms of CPT-11 and its metabolites on the uptake of [³H]DNP-SG. For transport studies involving the effect of the lactone and carboxylate forms of CPT-11 and its metabolites on the uptake of [³H]DNP-SG, 1.0 µM [³H]DNP-SG was used. After preincubation at 37°C for 2 min, the inhibitor was added and a further preincubation for 1 min was allowed. Inhibitor concentrations used were 0.03, 0.1, 0.3, 1, 10 and 500 µM for the lactone and carboxylate forms of SN38-Glu; 1, 50 and 500 µM for the lactone form of CPT-11; 0.3, 1, 10, 30, 50 and 500 µM for the carboxylate form of CPT-11; and 1, 10, 30, 50 and 500 µM for the carboxylate form of SN-38. After preincubation, the CMV obtained from SD rats was added to give a final protein concentration of 0.5 mg/ml. The final incubation volume was 20 µl. After incubation for 2 min, the reaction was terminated as described above. Radioactivity retained on the filter was determined using a liquid scintillation counter.

HPLC Analysis

The analysis of the carboxylate and lactone forms of CPT-11 and its metabolites was accomplished by HPLC, as described previously (Rivory and Robert, 1994; Sasaki et al., 1995b), with a modification that permitted the simultaneous determination of CPT-11, SN-38 and SN38-Glu. Briefly, for the analysis of plasma samples, 50 µl of ice-cold 50 mM phosphate buffer (pH 6.0) was added to 200 µl of plasma, followed by the addition of 50 µl of the ice-cold lactone form of CPT, as an internal standard, and 450 µl of methanol. The mixture was vortex-mixed for 10 sec and centrifuged at 4°C for 5 min. Then 70 µl of 50 mM phosphate buffer (pH 6.0) was added to 100 µl of supernatant, and the mixture was injected immediately onto the HPLC column, followed by determination of only the lactone forms, because of large interference peaks around the retention times of the carboxylate forms. Total (sum of lactone and carboxylate forms) drug was determined as the lactone in the same manner, except that 70 µl of 0.3 M HCl was added instead of 50 mM phosphate buffer, to ensure complete lactonization of the drug. The carboxylate form of the drug was determined by subtraction of the lactone form from the total drug concentration. Bile and urine samples were diluted with 50 mM phosphate buffer (pH 6.0), and liver samples (0.5 g) were homogenized with 3 ml of methanol/50 mM phosphate buffer (pH 6.0) (7:3, v/v) for the analysis of the lactone and total concentrations in the same way. Because there were no interfering peaks in the samples from the in vitro CMV uptake study, the lactone and carboxylate forms of the drugs were determined simultaneously by using the method for the lactone form determination described above. In the uptake study, the conversion of the lactone and carboxylate forms of the drug on the filter and in the medium during the experiment was <5%, as determined by HPLC. Therefore, only the total concentration of drugs was determined when analyzing filter and medium samples. Briefly, for the analysis of filter samples, 250 µl of ice-cold 50 mM phosphate buffer (pH 6.0) was added to the filter, followed by 50 µl of the ice-cold lactone form of CPT, as an internal standard, and 450 µl of methanol. The mixture was vortex-mixed for 1.5 min, to extract the drugs on the filter, and was centrifuged at 4°C for 5 min. Then, 70 µl of 0.3 M HCl was added to 100 µl of supernatant, and the mixture (50 µl) was injected onto the HPLC column, followed by determination of only the total form of the drugs. The medium samples were diluted with 50 mM phosphate buffer (pH 6.0) to give a total volume of 250 µl, followed by the addition of 50 µl of the ice-cold lactone form of internal standard and 450 µl of methanol. The mixture was vortex-mixed for 10 sec and centrifuged at 4°C for 5 min. Then, 70 µl of 0.3 M HCl was added to 100 µl of supernatant, and the mixture was injected onto the HPLC column for drug determination.

The HPLC system consisted of an Hitachi L-6300 pump, a Tosoh TSK Gel ODS-80Ts (150 × 4.6 mm inner diameter) column with a TSK Guardgel ODS-80Ts guard column, an Hitachi AS-4000 autosampler and an Hitachi F-1050 fluorescence detector. The excitation and emission settings for the in vivo analysis were 375 and 545 nm, respectively. The limits of detection, in terms of the amount of sample injected onto the HPLC column, were 0.2, 0.03 and 0.4 pmol for the lactone forms of CPT-11, SN-38 and SN38-Glu, respectively. In the CMV uptake study, the excitation and emission settings were 370 and 430 nm for CPT-11 and SN38-Glu and 380 and 556 nm for SN-38, respectively. The limits of detection were 0.0074, 0.01 and 0.0044 pmol for the lactone forms of CPT-11, SN-38 and SN38-Glu, respectively. The operation of the HPLC system was controlled by an Hitachi D-6100 HPLC manager. The mobile phase consisted of solvent A of acetonitrile/tetrahydrofuran/0.9 mM 1-heptanesulfonic acid sodium salt in 50 mM phosphate buffer (pH 6.0) (8:4:88) and solvent B of acetonitrile/5 mM 1-heptanesulfonic acid sodium salt in 50 mM phosphate buffer (pH 6.0) (30:70). The total elution time was 22 min, with 100% solvent A from 0 to 7 min, 100% solvent B from 7.1 to 16 min and 100% solvent A from 16.1 to 22 min. The flow rate was 0.9 ml/min.

Data Analysis

The AUC_(0-8hr) for the lactone and carboxylate forms of CPT-11 and its metabolites was calculated by the trapezoidal rule. CL_bile,p was calculated from the following equation:

(1)

where X_bile(0-8hr) is the cumulative amount excreted into bile over 0 to 8 hr. CL_bile,h, defined as biliary excretion rate divided by the hepatic concentration of the drug, was calculated from the following equation:

(2)

where V_bile(8hr) and X_liver(8hr) are the biliary excretion rate and hepatic concentration at 8 hr, with dimensions of milligram per minute per kilogram and milligram per gram of liver, respectively. V_bile(8hr) was approximated by the following equation:

(3)

where X_bile(6-8hr) is the amount of drug excreted into the bile duct from 6 to 8 hr. Thus, the CL_bile,h calculated has the dimensions of grams of liver per minute per kilogram. It was assumed that 1 g of liver is equivalent to 1 ml and, therefore, the CL_bile,h has the dimensions of milliliter per minute per kilogram.

C_bile/C_liver was calculated from the following equation:

(4)

where C_bile(8hr) is the drug concentration in bile at 8 hr and approximated by the following equation:

(5)

where V_bile(6-8hr) is the rate of drug excreted into the bile duct from 6 to 8 hr. Q_bile(6-8hr) is the bile flow rate from 6 to 8 hr.

The CL_r was calculated from

(6)

where X_urine(0-8hr) is the cumulative amount excreted into urine from 0 to 8 hr.

The inhibition constant (K_i) values for evaluating the inhibitory effect of CPT-11 and its metabolites on the uptake of [³H]DNP-SG by CMV from SD rats were obtained by fitting the following equation to the data:

(7)

where V_(+I) and V_(I) represent the transport velocity in the presence and absence of inhibitor, respectively; I is the inhibitor concentration. This equation was derived based on the assumption of competitive inhibition and the fact that the [³H]DNP-SG concentration (1 µM) is much lower than the K_m value (71 µM) (Akerboom et al., 1991) for DNP-SG uptake.

Statistical Method

The results are shown as means ± S.E. for the number of determinations. Dunnett's test was used to determine the significance of differences between the means of two groups, with P < .01 and P < .05 as the minimum levels of significance.

	Results

Top Abstract Introduction Materials & Methods Results Discussion References

Plasma concentration-time profiles of CPT-11 and its metabolites. The plasma concentration-time profiles for the lactone and carboxylate forms of CPT-11, SN-38 and SN38-Glu after i.v. injection of the lactone form of CPT-11 to SD rats and EHBR are shown in figures 2, 3 and 4, respectively. In the time period immediately after i.v. injection, CPT-11 was mainly present in plasma as the lactone form in both SD rats and EHBR; subsequently, the carboxylate form of CPT-11 appeared gradually in plasma (fig. 2). In SD rats, the AUC_(0-8hr) of the lactone form of CPT-11 exhibited nonlinearity with increasing dose, whereas minimal saturation could be observed in EHBR (table 1). The plasma concentration of the active metabolite SN-38 was much lower than that of the parent drug CPT-11 (fig. 3). Similarly, with CPT-11, the carboxylate form of SN-38 appeared gradually in plasma with time (fig. 3). At a dose of 10 mg/kg, the disappearance of the carboxylate form of SN-38 in plasma of SD rats was much faster than in EHBR, whereas the difference in the disappearance curve of the carboxylate form of SN-38 was minimal at 40 mg/kg for the SD and EHBR strains (fig. 3). The AUC_(0-8hr) for the carboxylate form of SN-38 differed markedly between SD rats and EHBR (0.0474 ± 0.0081 µg·hr/ml and 0.339 ± 0.078 µg·hr/ml for SD rats and EHBR, respectively) at 10 mg/kg but was very similar for the two strains at 40 mg/kg (0.249 ± 0.064 µg·hr/ml and 0.245 ± 0.060 µg·hr/ml for SD rats and EHBR, respectively) (table 2). The plasma disappearance of both the lactone and carboxylate forms of SN38-Glu was also slower in EHBR at 10 mg/kg, compared with that in SD rats (fig. 4). On the other hand, at the higher dose (40 mg/kg), similar concentration-time profiles were obtained for both the carboxylate and lactone forms of SN38-Glu with each strain (fig. 4). The difference in AUC_(0-8hr) for both the carboxylate and lactone forms of SN38-Glu between SD rats and EHBR was much more marked at the lower dose (10 mg/kg) than at the higher dose (40 mg/kg) (table 3).

View larger version (23K): [in this window] [in a new window]   Fig. 2.   Plasma concentration-time curves for the lactone (A and B) and carboxylate (C and D) forms of CPT-11 after i.v. administration of the lactone form of CPT-11 (10 and 40 mg/kg) in SD rats (A and C) and EHBR (B and D) with bile cannulation. Each data point represents the mean ± S.E. of three different rats. , lactone form at 10 mg/kg; , carboxylate form at 10 mg/kg; , lactone form at 40 mg/kg; , carboxylate form at 40 mg/kg.

View larger version (24K): [in this window] [in a new window]   Fig. 3.   Plasma concentration-time curves for the lactone (A and B) and carboxylate (C and D) forms of SN-38 after i.v. administration of the lactone form of CPT-11 (10 and 40 mg/kg) in SD rats (A and C) and EHBR (B and D) with bile cannulation. Each data point represents the mean ± S.E. of three different rats. , lactone form at 10 mg/kg; , carboxylate form at 10 mg/kg; , lactone form at 40 mg/kg; , carboxylate form at 40 mg/kg.

View larger version (24K): [in this window] [in a new window]   Fig. 4.   Plasma concentration-time curves for the lactone (A and B) and carboxylate (C and D) forms of SN38-Glu after i.v. administration of the lactone form of CPT-11 (10 and 40 mg/kg) in SD rats (A and C) and EHBR (B and D) with bile cannulation. Each data point represents the mean ± S.E. of three different rats. , lactone form at 10 mg/kg; , carboxylate form at 10 mg/kg; , lactone form at 40 mg/kg; , carboxylate form at 40 mg/kg.

Biliary excretion of CPT-11 and its metabolites. The biliary excretion of CPT-11 and its metabolites was measured up to 8 hr in SD rats and EHBR after i.v. injection of the lactone form of CPT-11 (figs. 5 and 6). The bile flow rate was not significantly affected by i.v. injection of CPT-11 in either SD rats or EHBR. The pharmacokinetic parameters obtained are listed in tables 1, 2 and 3. For the carboxylate form of CPT-11, with increasing dose, clear saturation of biliary excretion was observed in SD rats, whereas no apparent saturation was found in the EHBR (fig. 5, C and D). As for the lactone form of CPT-11, saturation was also observed but in both SD rats and EHBR (fig. 5, A and B). As shown in figure 6, SN-38 was mainly excreted into bile as the carboxylate form both in SD rats and EHBR. The cumulative amount of the carboxylate form of SN-38 excreted into bile exhibited almost no change even with increasing dose in SD rats (table 2), showing a clear saturation. On the other hand, such nonlinearity was not observed in EHBR (table 2). In SD rats, clear saturation of biliary excretion for both the lactone and carboxylate forms of SN38-Glu was observed with increasing dose (fig. 6). On the other hand, in EHBR, the biliary excretion of the lactone and carboxylate forms of SN38-Glu was not so clear, compared with SD rats (fig. 6).

View larger version (24K): [in this window] [in a new window]   Fig. 5.   Biliary excretion of the lactone (A and B) and carboxylate (C and D) forms of CPT-11 after i.v. administration of CPT-11 (10 and 40 mg/kg) in SD rats (A and C) and EHBR (B and D). Each data point represents the mean ± S.E. of three different rats. , lactone form at 10 mg/kg; , carboxylate form at 10 mg/kg; , lactone form at 40 mg/kg; , carboxylate form at 40 mg/kg.

View larger version (27K): [in this window] [in a new window]   Fig. 6.   Biliary excretion of the lactone (A and B) and carboxylate (C and D) forms of SN-38 and SN38-Glu after i.v. administration of the lactone form of CPT-11 (10 and 40 mg/kg) in SD rats (A and C) and EHBR (B and D). Each data point represents the mean ± S.E. of three different rats. , lactone form of SN-38 at 10 mg/kg; , carboxylate form of SN-38 at 10 mg/kg; , lactone form of SN-38 at 40 mg/kg; , carboxylate form of SN-38 at 40 mg/kg; , lactone form of SN38-Glu at 10 mg/kg; , carboxylate form of SN38-Glu at 10 mg/kg; , lactone form of SN38-Glu at 40 mg/kg; , carboxylate form of SN38-Glu at 40 mg/kg.

Estimation of CL_bile,p and CL_bile,h. The CL_bile,p value of both the lactone and carboxylate forms of CPT-11 exhibited nonlinearity in SD rats. Moreover, the saturation of CL_bile,p for the carboxylate form of CPT-11 was more marked than that of the lactone form (table 1). In contrast, no clear saturation could be observed for CL_bile,p of the lactone or carboxylate form of CPT-11 in EHBR. The CL_bile,h was reduced for the carboxylate form of CPT-11 as the dose increased only in SD rats; no nonlinearity could be observed in EHBR. The reduction in CL_bile,h for the lactone form of CPT-11 was minimal with increasing dose in both SD rats and EHBR. The absolute value of the CL_bile,h for the carboxylate form of CPT-11 was much larger than that of the lactone form in SD rats, especially at the lower dose (10 mg/kg). In addition, at 10 mg/kg, the CL_bile,h of the carboxylate form of CPT-11 in SD rats was larger than that in EHBR; at the higher dose, the difference was smaller (table 1). Similar results were also observed for the CL_bile,p and CL_bile,h of SN-38 (table 2). A reduction in CL_bile,h with increasing dose was not observed for the lactone form of SN-38 but was present for the carboxylate form in SD rats. The CL_bile,h value for the carboxylate form of SN-38 was larger in SD rats than in EHBR, especially at the lower dose. In contrast, no marked difference was found for the lactone form of SN-38 in CL_bile,h between SD rats and EHBR (table 2). For SN38-Glu, the CL_bile,h value of both the lactone and carboxylate forms was clearly lower in EHBR than in SD rats. In SD rats, the CL_bile,h value of the carboxylate form was larger than that of the lactone form at 10 mg/kg, and saturation was more marked for the carboxylate form than for the lactone form (table 3).

Urinary excretion of CPT-11 and its metabolites. Nonlinearity was observed in the CL_r of the carboxylate forms of CPT-11, SN-38 and SN38-Glu only in SD rats, whereas no clear saturation in the CL_r could be observed for those compounds in EHBR. In addition, the lactone forms of CPT-11, SN-38 and SN38-Glu showed a minimal reduction in CL_r with increasing dose in both SD rats and EHBR (tables 1, 2 and 3).

Characterization of CMV. The Mg⁺⁺-ATPase enrichment factor was 67.6 ± 9.1 (mean ± S.E., n = 3) and 72.7 (mean value, n = 2) in CMV from SD rats and EHBR, respectively, compared with the corresponding activity in liver homogenate. The alkaline phosphatase enrichment factor was 80.9 ± 7.8 (mean ± S.E., n = 3) and 76.4 (mean value, n = 2) in SD rats and EHBR, respectively. The ATP-dependent uptake of [³H]taurocholate was 186 ± 2 and 153 ± 10 pmol/2 min/mg protein (mean ± S.E. of three different experiments) and that of [³H]DNP-SG was 178 ± 19 and 0.1 pmol/2 min/mg protein in SD rats and EHBR, respectively.

Study of uptake of CPT-11 and its metabolites by CMV from SD rats and EHBR. The uptake by CMV from SD rats was markedly stimulated by ATP for the carboxylate and lactone forms of SN38-Glu and the carboxylate forms of CPT-11 and SN-38 (fig. 7). The uptake showed linearity over at least 2 min. No ATP-dependence was found in the uptake of the lactone form of CPT-11. Due to the low solubility of the lactone form of SN-38 and the detection limit for SN-38 (approximately 0.01 pmol), uptake of the lactone form of SN-38 could not be determined. The uptake by CMV prepared from EHBR was also examined (fig. 8). Compared with SD rats, the ATP-dependent uptake by CMV from EHBR was greatly impaired in the case of the lactone and carboxylate forms of SN38-Glu and the carboxylate forms of CPT-11 and SN-38. A minimal ATP-dependence was found for the uptake of the lactone form of CPT-11 by CMV from EHBR.

View larger version (30K): [in this window] [in a new window]   Fig. 7.   ATP-dependent uptake of the lactone and carboxylate forms of CPT-11 and its metabolites by CMV obtained from SD rats. CMV were incubated in the presence of ligand (50 µM) with () or without () ATP (5 mM) and an ATP-regenerating system. Data are means ± S.E. of three preparations. A, lactone form of SN38-Glu; B, carboxylate form of SN38-Glu; C, lactone form of CPT-11; D, carboxylate form of CPT-11; E, carboxylate form of SN-38.

View larger version (20K): [in this window] [in a new window]   Fig. 8.   ATP-dependent uptake of the lactone and carboxylate forms of CPT-11 and its metabolites by CMV obtained from SD rats (A) and EHBR (B). CMV were incubated in the presence of substrate (50 µM) with () or without () ATP (5 mM) and an ATP-regenerating system. The uptake time was 2 min. , ATP-dependent uptake for each compound. Data are means ± S.E. of three preparations. *P < .05, **P < .01 for the difference in uptake between SD rats and EHBR, by Dunnett's test.

Effect of the lactone and carboxylate forms of CPT-11 and its metabolites on the uptake of [³H]DNP-SG by CMV from SD rats. The effects of the lactone and carboxylate forms of CPT-11 and its metabolites on the ATP-dependent uptake of [³H]DNP-SG are shown in figure 9. DNP-SG uptake was significantly inhibited by the carboxylate and lactone forms of SN38-Glu and the carboxylate forms of SN-38 and CPT-11, whereas the lactone form of CPT-11 had little effect on the uptake of DNP-SG (fig. 9). The K_i values were 1.03 ± 0.05, 1.62 ± 0.05, 18.3 ± 3.6 and 96.6 ± 8.4 µM for the carboxylate form of SN38-Glu, the lactone form of SN38-Glu and the carboxylate forms of SN-38 and CPT-11, respectively.

View larger version (19K): [in this window] [in a new window]   Fig. 9.   Effect of the carboxylate and lactone forms of CPT-11 and its metabolites on ATP-dependent uptake of [3H]DNP-SG (1 µM) by CMV obtained from SD rats. Data are means ± S.E. of three preparations. , lactone form of CPT-11; , carboxylate form of CPT-11 (Ki = 96.6 ± 8.4 µM); , carboxylate form of SN-38 (Ki = 18.3 ± 3.6 µM); , carboxylate form of SN38-Glu (Ki = 1.03 ± 0.05 µM); , lactone form of SN38-Glu (Ki = 1.62 ± 0.05 µM).

	Discussion

Top Abstract Introduction Materials & Methods Results Discussion References

Although CPT-11 is a potent and novel anticancer drug, a severe side effect (diarrhea) has been observed in some patients during its clinical use. To understand the mechanism for this side effect, it is important to clarify the pharmacokinetics of CPT-11. The pharmacokinetics of CPT-11 in both humans and animals have been widely studied (Ohe et al., 1992; Rothenberg et al., 1993). Nonlinear pharmacokinetics of CPT-11 in rats were previously reported (Kaneda and Yokokura, 1990; Atsumi et al., 1991), showing that CPT-11 and its metabolites recovered from bile during the first 24 hr after drug administration accounted for about 50 to 60% of total drug disposition. Those data suggested that the elimination of CPT-11 and its metabolites in rats is mainly via biliary excretion. However, the biliary excretion mechanism of CPT-11 and its metabolites was unknown. In addition, most previous studies measured only total (sum of carboxylate and lactone forms) concentrations of CPT-11 and its metabolites, rather than considering the biliary excretion of the individual carboxylate and lactone forms of CPT-11 and its metabolites. In this study, we systematically examined the biliary excretion of the carboxylate and lactone forms of CPT-11 and its metabolites in normal rats and EHBR, to clarify the biliary excretion mechanism involved.

The biliary excretion mechanism for the carboxylate and lactone forms of CPT-11, SN-38 and SN38-Glu was investigated in both in vivo and in vitro CMV studies. Because the carboxylate and lactone forms of SN38-Glu and the carboxylate forms of CPT-11 and SN-38 have anionic charges, we hypothesized that the biliary excretion of these four compounds is mediated by cMOAT, which is known to be genetically deficient in EHBR. The observed results strongly support the validity of our hypothesis. First, in SD rats, the CL_bile,h values of the carboxylate forms of CPT-11, SN-38 and SN38-Glu showed obvious saturation when the dose was increased from 10 mg/kg to 40 mg/kg. Second, compared with SD rats, the CL_bile,h for the carboxylate and lactone forms of SN38-Glu and the carboxylate forms of CPT-11 and SN-38 in EHBR were markedly reduced, especially at the lower dose (10 mg/kg). Similarly to CL_bile,h, the C_bile/C_liver showed obvious saturation for the carboxylate forms of CPT-11, SN-38 and SN38-Glu in SD rats (tables 1, 2 and 3). The C_bile/C_liver for these three compounds and the lactone form of SN38-Glu in EHBR at a dose of 10 mg/kg was significantly lower than in SD rats. These results suggest that the biliary excretion of these four compounds is by carrier-mediated transport, which is deficient in EHBR. To examine our hypothesis directly, a study of the uptake of the carboxylate and lactone forms of CPT-11 and its metabolites by CMV isolated from SD rats and EHBR was performed. For the carboxylate and lactone forms of SN38-Glu and the carboxylate forms of CPT-11 and SN-38, the uptake by CMV in SD rats showed a clear ATP-dependence (fig. 7), whereas the ATP-dependent uptake was much lower in CMV obtained from EHBR (fig. 8). On the other hand, no ATP-dependence was found in the uptake of the lactone form of CPT-11 by CMV obtained from either SD rats or EHBR. These results indicate that there is a primary active transport system on the bile canalicular membrane specific for the four compounds with anionic charges and this transport system is deficient in EHBR. Finally, to add further support to our hypothesis, DNP-SG, which is well known as a representative substrate for cMOAT (Akerboom et al., 1991; Takenaka et al., 1995b; Ballatori and Truong, 1995; Yamazaki et al., 1996), was used to study the inhibition of DNP-SG uptake by these compounds. The results shown in figure 9 show that the uptake of DNP-SG by CMV from SD rats is greatly inhibited by the carboxylate and lactone forms of SN38-Glu and the carboxylate forms of CPT-11 and SN-38. However, no inhibitory effect was found for the lactone form of CPT-11. This result suggests that the carboxylate and lactone forms of SN38-Glu and the carboxylate forms of CPT-11 and SN-38 share, at least in part, the same transporter as DNP-SG. This finding, therefore, is consistent with the in vivo results with EHBR. Our in vivo and in vitro results, therefore, demonstrate that the biliary excretion of the carboxylate and lactone forms of SN38-Glu and the carboxylate forms of CPT-11 and SN-38 are mediated by the cMOAT located on the bile canalicular membrane, which is genetically deficient in EHBR. Comparison of the inhibitory efficiency for DNP-SG uptake by CMV, as shown in figure 9, suggests that both the carboxylate and lactone forms of SN38-Glu have a much higher affinity for the DNP-SG transporter than do the carboxylate forms of SN-38 and CPT-11, because their K_i values were much smaller (<2 µM) than those of the carboxylate forms of SN-38 (20 µM) or CPT-11 (100 µM). In addition, for SN38-Glu, the K_i value for its carboxylate form (1.0 µM) was smaller than that for the lactone form (1.6 µM), suggesting that the carboxylate form of SN38-Glu has a relatively higher affinity, compared with its lactone form. A similar finding was also obtained in the in vivo study. With increasing dose, saturation in the CL_bile,h for the carboxylate form in SD rats was more obvious than for the lactone form (table 3). That is, the CL_bile,h for the carboxylate form of SN38-Glu at 10 mg/kg was much greater than that at 40 mg/kg, whereas the difference in the CL_bile,h for the lactone form between 10 and 40 mg/kg was minimal. This result can be explained if the carboxylate form has a higher affinity (low K_m value) for the transporter than does the lactone form and, therefore, saturation of its biliary excretion can be easily observed. One possible factor in such a postulated difference in affinity is that, compared with the lactone form of SN38-Glu, there are two anionic charges present in the carboxylate form of SN38-Glu, and this might result in the higher affinity for the transporter than is the case with its lactone form.

As shown in figure 8, unlike the lactone and carboxylate forms of CPT-11, the ATP-dependent uptake of both lactone and carboxylate forms of SN38-Glu was partially retained in CMV from EHBR at a substrate concentration of 50 µM (fig. 8B). This suggests that a primary active transport system, other than the cMOAT deficient in EHBR, is expressed on the bile canalicular membrane in EHBR. Recently, we found that the ATP-dependent uptake of E3040 glucuronide could also be observed in CMV from EHBR (Niinuma, O. Takenaka, T. Horie, K. Kubayashi, Y. Kato, H. Suzuki and Y. Sugiyama, unpublished observations). Thus, it is possible that multiple transport systems for organic anions are located on the canalicular membrane. In addition, the K_i values obtained by evaluating the inhibitory effects of CPT-11 and its metabolites on the ATP-dependent uptake of DNP-SG were 1.0 and 1.6 µM for the carboxylate and lactone forms of SN38-Glu, respectively (fig. 9). On the other hand, ATP-dependence was clearly observed in the uptake of both the carboxylate and lactone forms of SN38-Glu at a substrate concentration of 50 µM (figs. 7 and 8). If these compounds share the same transporter (cMOAT) with DNP-SG, their own uptake by CMV at a concentration of 50 µM should be almost saturated and no clear ATP-dependent uptake should be observed. Accordingly, we suggest that there is a multiplicity of biliary excretion systems for both the lactone and carboxylate forms of SN38-Glu on the bile canalicular membrane. A primary active transporter, other than cMOAT, which is still maintained in EHBR, may be responsible for the residual ATP-dependent uptake of these compounds.

The CMV uptake of the carboxylate form of SN-38 also showed ATP-dependence in EHBR. Although the uptake in the presence of ATP (705 ± 31 pmol/mg protein) was significantly higher than the uptake in its absence (523 ± 49 pmol/mg protein), it is still possible that the primary active transporter for SN-38 (carboxylate form) might also be expressed on the bile canalicular membrane in EHBR, because such a difference in CMV uptake in the presence or absence of ATP was not so obvious (fig. 8B), nor was there any significant difference in uptake over 0.5 min in the presence or absence of ATP (data not shown). In addition, when the concentration-dependence of SN-38 uptake by CMV prepared from SD rats was examined, only a single component for the saturation of uptake was found (X.-Y. Chu, Y. Kato and Y. Sugiyama, unpublished observations). Therefore, it is likely that SN-38 (carboxylate form) is recognized only by cMOAT and the contribution of the other transporter is minor.

It has been reported that the diarrhea induced by CPT-11 also occurs in rats and is similar in nature to that in humans (Takasuna et al., 1995a,b). The possible mechanisms for the diarrhea caused by CPT-11 have received a great deal of attention. The diarrhea induced by enterocolitis has been suggested to be caused by high levels of SN-38 and/or CPT-11 retained for a long period in the intestine (Araki et al., 1993), and the biliary excretion of SN38-Glu may be contributory (Kaneda et al., 1990). One treatment approach based on this hypothesis would be to reduce the intestinal concentration of SN-38 by inhibiting the activity of -glucuronidase in the intestinal microflora, and baicalin, an inhibitor of -glucuronidase, has been reported to ameliorate CPT-11-induced diarrhea in rats (Takasuna et al., 1995b). A biliary index (the product of the relative area ratio of SN-38 to SN38-Glu multiplied by that of CPT-11 under the plasma concentration-time curve) has been defined and found to be correlated with the degree of diarrhea in humans (Gupta et al., 1994). This study is in agreement with our hypothesis if we consider that lowering the biliary excretion clearance of SN38-Glu consequently reduces the biliary index; however, we should consider the possibility that interpatient variance in glucuronidation activity might also affect the biliary index and side effects. Our study gives deeper insight into the biliary excretion mechanism of CPT-11 and its metabolites and also suggests that the biliary excretion of SN38-Glu is mediated by cMOAT; therefore, inhibitors of cMOAT may be candidates for ameliorating CPT-11-induced diarrhea. However, the mechanism for the diarrhea caused by CPT-11 is a complicated one. In addition to the possible mechanism discussed, there are other pathways that may also contribute to the diarrheal side effect. Indeed, the intestinal secretion of CPT-11 in rats has been observed (Kaneda and Yokokura, 1990). On the other hand, a recent study in two patients found that the total cumulative biliary and urine excretion of CPT-11 and its metabolites up to 48 hr ranged from 25 to 50% (Lokiec et al., 1995). It is possible that CPT-11 can be excreted by another, as yet unidentified, secretion pathway such as intestinal secretion or eliminated by conversion into other, as yet unidentified, metabolites in humans (Rivory and Robert, 1995). Therefore, it is possible that several other mechanisms may also be involved in the induction of the diarrheal side effect of CPT-11. As for the validity of our hypothesis and whether the mechanism for the toxicity of CPT-11 in rats is similar to that in humans, further investigation is needed. In conclusion, the excretion of the carboxylate and lactone forms of SN38-Glu and the carboxylate forms of CPT-11 and SN-38 across the bile canalicular membrane is mediated by the primary active transporter (cMOAT), which is deficient in EHBR.