Abstract
Verapamil (VP) is used as a racemate but shows stereoselective pharmacokinetics and pharmacodynamics. It undergoes extensive first-pass metabolism. Stereoselective first-pass metabolism in the intestine and liver was investigated in vivo and in vitro to determine its impact on the disposition of VP and its main metabolite, norverapamil (NVP). VP racemate was administered to rats i.v., p.o., and via the portal vein. The formation rates of the main metabolites of the VP enantiomers were estimated in an in vitro intestinal microsomal study. The hepatic bioavailability of VP showed saturable metabolism, and the hepatic bioavailability of R-VP was higher than that of S-VP. Conversely, the intestinal bioavailability of R-VP was lower than that of S-VP, resulting in a higher systemic bioavailability of S-VP. The pharmacokinetics of the NVP enantiomers was similar. These results suggest that the stereoselectivity of the total bioavailability of VP is determined by first-pass metabolism in the small intestine and liver, and that the NVP enantiomers observed in the systemic circulation after p.o. administration of VP racemate originate mainly from the liver in rats.
The first-pass metabolism of drugs in the small intestine and liver limits their bioavailability to the systemic circulation. Although the effects of first-pass metabolism in the liver after p.o. administration have been well studied, it is now known that many drugs also undergo first-pass metabolism in the small intestine. Reductions in the first-pass metabolism of these drugs, caused by drug-drug interactions or disease states, may occur to a different extent in the small intestine and liver. Therefore, it is very important to elucidate the mechanisms underlying first-pass metabolism, as well as the separate contributions of metabolism in the small intestine and liver, to be able to predict changes in oral bioavailability.
Verapamil (VP) is a calcium antagonist used clinically for the treatment of hypertension and for prophylaxis of supraventricular and ventricular arrhythmias. VP has a relatively narrow therapeutic plasma concentration range and shows relatively large interindividual variations in its pharmacokinetics and pharmacodynamics (Vogelgesang et al., 1984; Echizen et al., 1985a,b, 1988). Although VP is commercially available as a racemic mixture, its pharmacological effects and disposition have been reported to show stereoselectivity in both humans and animals. The antiarrhythmic effect of S-VP, as estimated by electrocardiograph, is 10 to 20 times higher than that of R-VP in humans (Echizen et al., 1985a,b). The oral bioavailability of S-VP is approximately 20% in humans, whereas that of R-VP is approximately 50% (Vogelgesang et al., 1984; Echizen et al., 1985a,b, 1988). The plasma protein binding of R-VP is higher than that of S-VP (free fractions: 7 and 12%, respectively) (Gross et al., 1988; Robinson and Mehvar, 1996). We have previously reported that VP binds enantioselectively to α1-acid glycoprotein and phosphatidylserine and shows an interaction in the binding between the enantiomers (Hanada et al., 1998b, 2000).
Norverapamil (NVP) is the main metabolite of VP. The area under the plasma concentration-time curve values for both VP (AUCvp) and NVP (AUCnvp) show enantiomeric differences; those for the R-isomer are greater than those for the S-isomer after p.o. administration in humans (Echizen et al., 1988). If S-VP is metabolized to a much greater extent than R-VP and the clearances of the NVP enantiomers are assumed to be the same, the AUCnvp for the S-isomer would be greater than that for the R-isomer. A similar relationship between VP and NVP is observed in rats, although the AUCvp and AUCnvp values for the S-isomer are greater than those for the R-isomer (Bhatti and Foster, 1997). These differences may be because of either stereoselective first-pass metabolism in the small intestine and liver or the disposition of NVP.
The contribution of the small intestine to first-pass metabolism after p.o. administration has not yet been directly estimated, and the pharmacokinetic interactions of the VP enantiomers remain unclear. Therefore, we investigated the stereoselective first-pass metabolism of VP in the small intestine and liver using previously reported in vivo dosing technique and in vitro metabolism experiments with intestinal microsomes. In particular, we aimed to elucidate the reason for the differences between the AUCvp and AUCnvp values for the two enantiomers on the basis of stereoselective first-pass metabolism in the small intestine and liver, and plasma protein binding. Although the relative relationships of the AUCs for the two enantiomers are opposite to those observed in humans, rats were used as the animal model for elucidating the mechanisms underlying the stereoselectivity of VP and NVP metabolism in the small intestine and liver.
Materials and Methods
Materials. Racemic VP hydrochloride and NVP hydrochloride were provided by Eisai Co. (Tokyo, Japan). N-Dealkylated VP hydrochloride and O-demethylated VP hydrochloride were provided by Knoll AG (Ludwigshafen, Germany). R-Propranolol hydrochloride was obtained from Aldrich Chem. Co. (Milwaukee, WI). [N-methyl-3H]VP hydrochloride (specific activity, 2.2 TBq/mmol) was obtained from DuPont-New England Nuclear (Boston, MA). The S- and R-enantiomers of VP were separated using a previously reported enantioselective high-performance liquid chromatography (HPLC) method (Hashiguchi et al., 1996; Hanada et al., 1998a,b), and their stereochemical purities were ascertained by stereospecific HPLC resolution (the stereochemical purities of S-VP and R-VP were 98.6 and 99.3%, respectively). All the other reagents used were of analytical grade unless stated otherwise.
Animals. The male Wistar rats (230–300 g) were handled in accordance with the Guide for the Care and Use of Laboratory Animals as adopted and promulgated by the U.S. National Institutes of Health and as approved by Meiji Pharmaceutical University Institutional Animal Care and Use Committees. The rats were kept under stable humidity and temperature conditions.
Blood to Plasma Partition Ratio. One hundred microliters of VP or NVP solution was added to freshly isolated rat blood (final concentrations of VP and NVP were 0.1, 0.5, and 2.0 μg/ml) and incubated for 1 h at 37°C. A sample of the incubated blood was taken for determination of the drug concentrations in whole blood, and a further aliquot was centrifuged at 1500g for 5 min at 37°C to produce plasma. The concentrations of the VP and NVP enantiomers in the blood and plasma samples were determined by the HPLC method described below.
Protein Binding. The protein binding of the VP enantiomers was evaluated using equilibrium dialysis (Hanada et al., 1998b, 2000). Briefly, 5 μl of 3H-labeled R- or S-VP (2.5 GBq/mmol) was added to 0.5 ml of freshly isolated serum. The solutions were dialyzed using a 5-cell equilibrium dialyzer (10 rpm; Spectrum Medical Industries, Inc., Houston, TX) against phosphate buffer (0.113 M Na2HPO4 and 0.017 M KH2PO4, pH 7.4) for 2 h at 37°C. The membrane used was a Spectra/Por-2, molecular mass cutoff 12,000 to 14,000 (Spectrum Chemical, Gardena, CA). Once equilibrium had been reached, the radioactivities of the solutions (inside and outside the dialysis tube) were determined using a liquid scintillation counter (Aloka Co., Ltd., Tokyo, Japan). Sample volume alteration during dialysis was corrected according to the change in the protein concentration. The concentration dependence of VP binding to rat serum protein was also studied (the final drug concentration in serum ranged from 0.06 to 3.0 μg/ml).
In Vivo Experiments. The in vivo experiments for determining separate intestinal and hepatic bioavailabilities were performed using a previously reported procedure (Mihara et al., 2001). The drug was administered via three different routes: infusion into the femoral vein (i.v.), into the portal vein, and into the duodenum as a substitute for p.o. administration. Male Wistar rats (230–250 g) maintained on standard food and water were fasted for 18 h before the experiment. Then i.p. sodium pentobarbital (50 mg/kg) was administered to each rat, and an abdominal incision was made to expose the distal ileum. Heparinized polyethylene tubes (PE-10) were inserted into the mesenteric vein and the femoral artery and vein. The three cannulations were carried out in all the rats for either drug administration or as a sham operation.
In the i.v. (1 mg/kg) and p.o. administration (5, 10 mg/kg) groups, racemic VP solution (5 mg/ml) was administered as a bolus via the femoral vein or duodenal cannula, respectively. In the portal vein administration group, racemic VP (0.31, 0.63, 1.25, 2.5, or 5.0 mg/ml) was infused for 20 min (corresponding to the normal absorption time) through the mesenteric cannula (total dose 0.62–10 mg/kg). Blood samples (approximately 0.2 ml) were obtained from the femoral artery at 1, 5, 15, 30, 60, 90, 120, and 180 min for the i.v. and p.o. administration groups, and at 10, 20, 30, 40, 60, 90, 120, and 150 min for the portal vein administration group.
To investigate the stereoselectivity of NVP pharmacokinetics, NVP racemate (2.5, 5, or 10 mg/kg) was administered i.v. to rats under sodium pentobarbital anesthesia. After administration, blood samples were withdrawn from the femoral artery at 5, 10, 15, 30, 45, 60, 90, and 120 min. All the blood samples were heparinized and were immediately centrifuged at 1500g for 5 min. The plasma samples were then stored at –20°C until required for analysis.
Preparation of Intestinal Microsomes. The rats were decapitated, and then the upper jejunums (30 cm from the pyloric valve) and livers were removed and immediately placed into ice-cold 0.9% (w/v) NaCl. The intestinal segments were then flushed with 50 ml of ice-cold 0.9% (w/v) NaCl to remove mucus and food residues (Mihara et al., 2001). Intestinal cells were obtained by shaving with a glass slide, placed in ice-cold 50 mM phosphate buffer, pH 7.4, and dispersed by sonicating (5 s, three times). Sixty microliters of phenylmethylsulfonyl fluoride (10 mg/ml in acetone) was added, and the mixture was homogenized (500 rpm, 6 strokes). The homogenate was centrifuged at 9000g for 30 min at 4°C. The supernatant was centrifuged at 105,000g for 65 min at 4°C. The final pellet was suspended in ice-cold 50 mM phosphate buffer, pH 7.4, and homogenized (500 rpm, 4 strokes).
VP Metabolism by Microsomes. The metabolism of VP by the microsomes was assessed using each enantiomer separately because of analytical difficulties when all the metabolites were mixed. Eight hundred microliters of 50 mM phosphate buffer, pH 7.4, 100 μl of the relevant VP enantiomer (0.05, 0.1, 0.5, 1, 1.5, 2, 3, and 4 mM), and 100 μl of the microsomal suspension (protein concentration was 1.0 mg/ml) were mixed. The reaction was started by adding an NADPH generating system. After incubation for 10 min at 37°C, the reaction was stopped by adding 1 ml of 5% trichloroacetic acid, and the concentrations of N-dealkylated VP, O-demethylated VP, and NVP were determined by HPLC as described below. The effect of rabbit antiserum against rat CYP1A, CYP2B, CYP2C, and CYP3A on the formation of these metabolites was also studied using a previously reported method (Mihara et al., 2001). The binding of the VP enantiomers to the microsomes was determined by equilibrium dialysis as described above.
Analytical Methods. The concentrations of the VP and NVP enantiomers in the plasma and blood were determined using a previously reported enantioselective HPLC method (Hashiguchi et al., 1996; Hanada et al., 1998a). The HPLC system consisted of a Shimadzu HPLC apparatus (Shimadzu, Kyoto, Japan), an LC-9A HPLC pump, and a C-R6A Chromatopac (Shimadzu) integrator. VP was detected by an RF-535 fluorescence detector, which was operated at excitation and emission wavelengths of 272 and 312 nm, respectively. The enantiomers were separated using a ChiralPak AD column (250 × 4.6 mm i.d.; Daicel, Tokyo, Japan) at 40°C. The mobile phase comprised hexane/isopropanol/diethylamine [94:6:0.1 (v/v)] at a constant flow rate of 1.2 ml/min. The drugs were extracted from the biological fluids using an organic solvent (ethyl ether) under alkaline conditions. With regard to analytical accuracy, the within- and between-day coefficients of variation were less than 9.2%.
The concentrations of N-dealkylated VP, O-demethylated VP, and NVP in the microsomal suspension were determined under the following HPLC conditions. The HPLC system consisted of a Shimadzu HPLC apparatus, an LC-6A HPLC pump, and a C-R6A Chromatopac integrator. These metabolites were detected by an RF-535 fluorescence detector that was operated at excitation and emission wavelengths of 272 and 312 nm, respectively. The metabolites were separated using an L-column octadecylsilane (250 × 4.6 mm i.d.; Chemical Inspection and Testing Institute, Tokyo, Japan) at 51°C. The mobile phase comprised 6 mM ammonium acetate/acetonitrile/methanol (70: 25:5, v/v; pH 4.2) at a constant flow rate of 1.2 ml/min. Two hundred microliters of 1 M NaOH and 6 ml of ethyl ether were added to 1 ml of microsomal suspension and shaken for 10 min. The organic layer was evaporated to dryness under a stream of nitrogen gas in a water bath at 50°C, the residue was reconstituted with 0.1 ml of mobile phase, and a 50-μl aliquot was injected into the HPLC system. With regard to analytical accuracy, the recovery was greater than 91%, and the within- and between-day coefficients of variation were less than 6.7%.
Data Analysis. The areas under the plasma concentration versus time curves after VP administration into the femoral vein (AUCiv), the portal vein (AUCpv), and the duodenum (AUCpo) were calculated by the trapezoidal and log-trapezoidal rules using the computer program WinNonlin (Pharsight, Mountain View, CA). Total bioavailability (Ft), hepatic bioavailability (Fh), and gastrointestinal bioavailability (Fg) were calculated using the following equations: where Fa is the fraction of absorption and assumed to be 1.
The total and hepatic clearances (CLtot and CLh, respectively) were calculated using the equations CLtot = Dose/AUCiv and CLh = Qh × Eh, where Qh and Eh are the hepatic blood flow rate and extraction ratio (1 – Fh) and a liver blood flow rate of 60 ml/min/kg was used (Davies and Morris, 1993). Extrahepatic clearance was estimated as CLtot – CLh.
In the microsomal experiments, data points for the formation velocities (v) of each metabolite at varying concentrations of the substrate VP (S) were fitted by the nonlinear least-squares regression program WinNonlin to a one-enzyme model yielding the maximum velocity (Vmax) and affinity constant (Km), using the following equation:
The differences between means for the enantiomers were tested using Student's paired t test or unpaired t test. A p value less than 0.05 was considered to indicate a significant difference.
Results
Both the blood to plasma concentration (B/P) ratio and the plasma protein binding of VP and NVP were concentration-independent within the concentration range studied (VP: 100-1000 ng/ml; NVP: 500-2000 ng/ml). The B/P ratio of R-VP was significantly higher than that of S-VP (Table 1). Similarly, the unbound fraction of R-VP was significantly higher than that of S-VP. After correction for the unbound fraction, the B/P ratio did not differ significantly between the enantiomers (S-VP: 8.3 versus R-VP: 7.6), indicating that the apparent stereoselectivity of distribution into the blood can be accounted for by plasma protein binding in rats. On the other hand, the B/P ratios of the NVP enantiomers were not significantly different.
The plasma concentrations of S-VP were higher than those of R-VP, and NVP was not detected after i.v. administration (Fig. 1). On the other hand, the concentrations of both enantiomers in whole blood were comparable. The plasma clearance was comparable with that obtained in our previously reported study using constant i.v. infusion (Hanada et al., 1998), indicating that the clearance of VP was linear at this dose. In this study, further data obtained were based on the concentrations in blood. The mean AUC values for S-VP and R-VP were 11,864 and 7619 ng/ml/min, respectively, after i.v. administration (1 mg/kg racemate).
The concentrations of S-VP and S-NVP were higher than those of the R-isomers after p.o. administration of VP racemate (Fig. 2), as has been reported previously (Bhatti and Foster, 1997). The oral clearance obtained at a dose of 5 mg/kg (S-VP 406, R-VP 936 ml/min/kg) was comparable with that at 10 mg/kg (S-VP 479, R-VP 965 ml/min/kg). The mean AUC values for S-VP and R-VP were 10,440 and 5181 ng/ml/min, respectively, after p.o. administration (10 mg/kg racemate). On the other hand, after intraportal vein administration of VP racemate, the concentration of R-VP was higher than that of S-VP, whereas the concentration of S-NVP was higher than that of R-NVP. The relationship between the dose administered via the portal vein and the AUC showed that the AUC increased disproportionately at racemate doses of more than 5.0 mg/kg (Fig. 3), indicating that first-pass metabolism in the liver is nonlinear within the dose range studied. To calculate each bioavailability value within the linear range, therefore, a similar AUC for VP (obtained by portal vein administration at dose of 3.0 mg/kg racemate) was used for calculation of hepatic bioavailability (Fh).
The Ft of S-VP was higher than that of R-VP (Table 2). However, the Fh of R-VP was almost double that of S-VP, and the Fg of S-VP was higher than that of R-VP. These results indicate that systemic bioavailability is determined by first-pass metabolism in both the small intestine and liver in rats, and that the stereoselectivity of Ft appears to result from the relative contributions of both organs to first-pass metabolism.
The calculated hepatic blood clearances of S- and R-VP were 52.2 and 47.0 ml/min/kg, respectively (Table 2). The calculated extrahepatic clearances of S- and R-VP were 21.7 and 31.1 ml/min/kg, respectively, suggesting that VP may also be eliminated by extrahepatic routes. The hepatic intrinsic clearance of S-VP was higher than that of R-VP.
Neither the blood clearance nor the volume of distribution at steady state of the NVP enantiomers was dose-dependent at the doses used (data not shown). The blood clearance and volume of distribution at steady state of S- and R-enantiomers were 257 ± 53 and 279 ± 54 ml/min/kg and 7409 ± 3450 and 6935 ± 3865 ml/kg, respectively (n = 4). These parameters did not differ significantly between the enantiomers, indicating that distribution and elimination of NVP are not stereoselective.
We performed an in vitro study of the metabolism of VP in the intestine to confirm the enantioselectivity of VP metabolism (Fig. 4). The intrinsic clearances of R-VP to N-dealkylated VP and NVP in the intestine were significantly higher than those of S-VP. However, intrinsic clearance to O-demethylated VP did not differ significantly between the enantiomers (Table 3). The effects of cytochrome P450 antibodies on the formation rates of N-dealkylated VP, O-demethylated VP, and NVP from S-VP in rat intestinal microsomes were also investigated. The formation rates of N-dealkylated VP and NVP were significantly depressed in the presence of anti-CYP3A2 in an antibody-concentration-dependent manner. On the other hand, the formation rates of O-demethylated VP were decreased in the presence of anti-CYP1A2 (data not shown).
Discussion
The partition of R-VP into blood cells was higher than that of the S-isomer, but this apparent stereoselectivity could be accounted for by a difference in plasma protein binding, as has been reported by others (Robinson and Mehvar, 1996). The hepatic clearances of both VP enantiomers were considerably limited by blood flow. The calculated extrahepatic clearance of VP was 22 to 31 ml/min/kg, suggesting that VP may be eliminated by extrahepatic metabolism and hepatic metabolism. Because the urinary excretion of both enantiomers is negligible (Eichelbaum et al., 1979), the small intestine may be one of the organs responsible for extrahepatic metabolism. Sandström et al. (1998) reported significant efflux of NVP into the lumen after i.v. administration of VP to rats and suggested that VP was metabolized in the enterocytes after i.v. administration.
In this experiment, the infusion time of VP into the portal vein was set at 20 min because the maximum plasma concentration of VP after p.o. administration occurs at about 20 min. However, the AUCpv showed dose dependence, so the bioavailability was calculated using the AUC obtained under apparent linear conditions. The Ft of S-VP was higher than that of R-VP. This result was comparable with that reported by others (Bhatti and Foster, 1997). However, the Fh of R-VP was almost double that of S-VP, whereas the Fg of S-VP was higher than that of R-VP. These results suggested that the stereoselectivity of the AUC of VP appears to be determined by first-pass metabolism in both the intestine and the liver (Fig. 5).
Efflux of NVP into the jejunum has been reported after i.v. administration of VP racemate to rats (Sandström et al., 1998). Both VP and NVP are known substrates for and inhibitors of P-glycoprotein (Sandström et al., 1998). Our preliminary in situ small intestinal perfusion study (an experimental procedure described previously by Mihara et al., 2001) showed that NVP was detectable in the outlet perfusate (data not shown). These results indicate that the NVP formed in the intestinal lumen and/or intestinal wall may be excreted into the lumen.
Although VP underwent first-pass metabolism in the intestine, the extent of its metabolism was smaller than that in the liver. The NVP metabolized in the intestine did not appear in the systemic circulation. Furthermore, the pharmacokinetics of NVP enantiomers was nonstereoselective. These results suggest that the reason why the AUC of S-NVP is higher than that of R-NVP, despite the AUC of S-VP being higher than that of R-VP after p.o. administration, is that hepatic first-pass metabolism of VP is the major factor determining the AUC of NVP (Fig. 5).
To confirm the stereoselective metabolism of VP in the intestine, we conducted an in vitro microsomal study. The concentrations of VP enantiomers used to investigate the kinetic parameters were corrected for the unbound fractions in the reaction mixture. The intrinsic clearances (Vmax/Km) of the R-isomer to N-dealkylated VP and NVP were significantly higher than those of the S-isomer. This stereoselectivity was corresponded to that of first-pass metabolism in the small intestine observed in the in vivo experiments (Table 2). The cytochrome P450 enzyme governing N-dealkylated VP and NVP formation was CYP3A2, whereas that responsible for O-demethylated VP formation was CYP1A2. Interestingly, S-VP was preferentially metabolized to NVP and O-demethylated VP in rat liver microsomes (Nelson and Olsen, 1988; Nelson et al., 1988), and this stereoselectivity also corresponded to the Fh values observed in our in vivo experiments (Table 2).
In previous studies, the AUCs of S- and R-VP in blood obtained after p.o. administration of VP (80 mg/kg) to healthy volunteers (Hashiguchi et al., 1996), corrected for the B/P ratio (Robinson and Mehvar, 1996), were 29 and 121 ng · h/ml, respectively, and the corresponding AUCs of the NVP enantiomers were 80 and 168 ng · h/ml. First-pass metabolism of VP in enterocytes has also been reported in humans (Sandström et al., 1999; von Richter et al., 2001). The extraction ratios of R- and S-VP by enterocytes were 0.49 and 0.68, respectively, and those by liver were 0.63 and 0.79, respectively (Sandström et al., 1999). These results indicate that S-VP undergoes extensive first-pass metabolism with the same direction of stereoselectivity in both organs. The species differences seen in the in vitro intestinal microsomal metabolic activity and the unbound fraction in blood also support this theory. Even if it is assumed that all the NVP formed in the intestine is excreted into the intestinal lumen and does not enter the systemic circulation, this does not explain the stereoselectivity and higher AUC level of R-NVP in humans. One possible explanation may be a difference in the elimination clearance of NVP because the unbound fraction of R-NVP in human blood was approximately half that of S-NVP (Robinson and Mehvar, 1996). Therefore, further studies will be required to investigate the species difference and stereoselectivity of first-pass metabolism in humans.
In summary, the Fh of R-VP was higher than that of S-VP. Conversely, the intestinal bioavailability of R-VP was lower than that of S-VP, resulting in a higher absolute bioavailability of S-VP. Therefore, the Ft appears to be determined by first-pass metabolism in both organs, and the NVP enantiomers observed in the systemic circulation after p.o. administration of VP racemate originate from stereoselective hepatic first-pass metabolism in rats.
Footnotes
-
This work was supported in part by a grant from the Japan Research Foundation for Clinical Pharmacology.
-
Article, publication date, and citation information can be found at http://dmd.aspetjournals.org.
-
doi:10.1124/dmd.107.020339.
-
ABBREVIATIONS: VP, verapamil; NVP, norverapamil; AUC, area under the drug concentration-time curve; HPLC, high-performance liquid chromatography; B/P, blood to plasma concentration.
- Received December 27, 2007.
- Accepted July 9, 2008.
- The American Society for Pharmacology and Experimental Therapeutics