Introduction

Top Abstract Introduction Materials & Methods Results Discussion References

Atorvastatin, [R-(R,R)]-2-(4-fluorophenyl)-,-dihydroxy-5-(1-methylethyl)-3-phenyl-4-[(phenylamino)carbonyl]-1H-pyrrole-1-heptanoic acid calcium salt (CI-981, AT¹) is a potent synthetic inhibitor of 3-hydroxy-3-methylglutaryl (HMG)-CoA reductase, the rate-limiting enzyme in cholesterol biosynthesis. Recently approved for human use, AT has a significant cholesterol lowering effect (Nawrocki et al., 1995). As part of AT development, absorption and disposition studies were conducted in rats and dogs. No apparent metabolic differences were observed between the two species; extensive presystemic metabolism with biliary excretion was the predominate route of radioactivity excretion (Michniewicz et al., 1994). Metabolites identified include ortho- and para-hydroxy-AT, -oxidized products, and a glucuronide conjugate of ortho-hydroxy-AT. Toxicokinetic investigations in mice and rats after repeated AT dose administration did not produce equivalent plasma concentrations, as measured by a nonspecific HMG-CoA inhibition assay (Black et al., 1995; Shum et al., 1996). In the mouse, daily doses ranging from 10 to 400 mg/kg AT produced a marked effect on plasma AT activity equivalent concentrations. C_max and AUC values were higher on day 1 than day 14 in mice, a result not observed in rats. The present studies were conducted to characterize the metabolism and excretion of AT in mice after single- and multiple-dose administration. In addition, we investigated the effect of multiple-dose administration on hepatic CYP450.

	Materials and Methods

Top Abstract Introduction Materials & Methods Results Discussion References

Chemicals. Radiolabeled atorvastatin. [¹⁴C]AT (specific activity, 81.2 mCi/mmol, chemical and radiochemical purity >98%) was synthesized with the carbon label at the 3-position of the pyrrole moiety. Ring-labeled [¹⁴C] was synthesized using an adaptation of the synthetic methodology of Baumann et al. (1992). The label was introduced as [¹⁴C]benzaldehyde, followed by sequential condensation with isobutyrylacetanilide, and the product was reacted with para-fluorobenzaldehyde to form a diketone intermediate. The diketone was conjugated with the protected chiral dihydroxyaminoheptanoic ester (Roth, 1987), and the product was treated with acid and base prior to calcium salt formation. Using a similar reaction sequence, ortho- and para-hydroxy-AT were synthesized and purified as sodium salts (Bjorge et al., 1995; Roth et al., 1991). Testosterone and [¹⁴C]lauric acid were obtained from Sigma. All other chemicals were reagent grade or better.

Animal Metabolism Studies. Mass balance and metabolic profile. Male B6C3F1 mice (N = 120, 9-14 weeks of age, 22-34 g) were divided into two groups for plasma profiling or mass balance studies. These groups were further subdivided into mice administered either a single 200 mg/kg [¹⁴C]AT (free acid) dose or a single daily dose of 200 mg/kg AT (free acid) for 13 days followed by a final dose of 200 mg/kg [¹⁴C]AT. All doses were administered by gavage. Radiolabeled doses were prepared by dissolving labeled and unlabeled material in dimethylacetamide (less than 10% of final volume) followed by the addition of 0.5% methylcellulose to produce a uniform suspension. Each mouse received approximately 10 µCi. Unlabeled AT was prepared as a suspension in 0.5% methylcellulose. Heparinized blood samples (1.0 ml) were collected by cardiac puncture from anesthetized mice, five to seven animals per time point, at predose, 1, 2, 4, 8, and 12 hr postdose for study days 1 and 14. The harvested plasma at each collection time was pooled and stored frozen until analysis. A control group of mice received no drug or vehicle. Mass balance studies used 15 animals per dose. Animals were housed in metabolism cages (five animals per cage) with access to food and water ad libitum. Urine and feces were collected predose and at 24-hr intervals for 6 days. After each radioactive dose, urine and feces from the same collection intervals were pooled. All samples were stored frozen at 20°C until analysis.

Enzyme Induction Study. Male B6C3F1 mice (N = 21, 9-14 weeks of age, 22-34 g) were divided into three groups: one group received daily single 1.0 mg/kg (free acid) oral doses of AT for 14 days, the second group received daily single 200 mg/kg (free acid) oral doses of AT for 14 days, and the third group served as a control and received daily single oral doses of vehicle (0.5% methylcellulose) for 14 days. Food and water were provided ad libitum. On day 15, animals were sacrificed, and the livers were removed and frozen in liquid nitrogen. Microsomes were isolated from pooled livers from each of the treatment groups (N = 7/group) (Guengerich, 1989). Cytochrome P450 and protein concentrations were determined by the methods of Omura and Bradford, respectively (Omura and Sato, 1964; Bradford, 1976).

Catalytic Assays. The catalytic activities of CYP1A, -2B, -3A, and -4A were measured with ethoxyresorufin (ER), pentoxyresorufin (PR), testosterone, and lauric acid, respectively. ER and PR dealkylase activity, expressed as pmol of resorufin/min/mg of microsomal protein, were determined by the method of Burke et al. (1985). Briefly, liver microsomes (0.5 mg/ml) from each group were incubated in duplicate in the presence of ER or PR, and the rate of resorufin formation was measured by fluorescence. The activity of CYP3A was measured by determining the rate of testosterone 6-hydroxylation. Liver microsomes (0.5 mg/ml) were incubated in triplicate with 200 µM testosterone (0.4 µCi) and 1.0 mM NADPH in 50 mM potassium phosphate buffer, pH 7.4, for 10 min. Sample aliquots were extracted with 2 × 5.0 ml of dichloromethane. The organic fractions were dried under nitrogen, and the samples were reconstituted in 50 µl of methanol. Testosterone and 6-hydroxytestosterone were separated on a Waters Spherisorb C-18 (250 × 4.6 mm) column coupled to a Packard Radiomatic flow detector CR (Downers Grove, IL). The mobile phase consisted of methanol:water (31:19, v/v) at a flow rate of 1.0 ml/min at 50°C. The activity of CYP4A was measured by determining the rate of lauric acid disappearance. Liver microsomes (0.5 mg/ml) were incubated in duplicate with 100 µM lauric acid (1.3 µCi) and 1.0 mM NADPH in 50 mM Tris-HCl buffer, pH 7.4, for 10 min. Sample aliquots were quenched with 300 µl of 3 N HCl and extracted with 8 ml of diethyl ether. Organic fractions were removed and dried under nitrogen; the samples were reconstituted in 60 µl of methanol. Samples were analyzed using the HPLC-radiochromatographic system mentioned above. The mobile phase consisted of 37:63 methanol:methanolic mixture (57.5/42/0.5, methanol/water/acetic acid) at a flow rate of 1.0 ml/min.

Radioactivity Assay. Radioactivity was measured by liquid scintillation counting (LCS) with external standardization for quench correction. Duplicate aliquots of plasma and urine were counted in 16 ml of Ready Safe (Beckman, Fullerton, CA) by a Packard TR2500 liquid scintillation spectrometer. After addition of water (9:1, v/w), fecal samples were homogenized, and duplicate 0.5-ml aliquots were removed for oxidation. Samples were air-dried and combusted in a Tri-Carb Sample Oxidizer (model 306, Packard) prior to LSC.

Sample Preparation and HPLC Radioactivity Profiles. Plasma samples (1.0 ml, 1 hr) and fecal homogenates (2.0 ml, 24 hr) were extracted with four volumes of acetone:acetonitrile (1:19, v/v). After centrifugation, the organic layer was placed in another tube and evaporated with nitrogen at room temperature. Residues were reconstituted in 50% acetonitrile:0.1 M ammonium acetate, pH 4, for radioactivity profiling. Recovery averaged 90%. Plasma and fecal extracts were profiled by gradient HPLC with radioactivity detection using a BioSil ODS-5S column (150 × 4.6 mm) (Hercules, CA) connected in a series with a Brownlee RP Spheri-5 guard cartridge (San Jose, CA). The mobile phase consisted of A (0.1 M ammonium acetate, pH 4.0) and B (acetonitrile). The gradient had an initial solvent composition of 75% A and 25% B that was held for 5 min. Over a 40-min period, solvent B increased to 32%, followed by another increase of solvent B to 40% for another 20 min. This mixture was held until 75 min into the run, and then solvent B increased to 80% over the next 15 min. This mixture was held until 110 min was reached (Michniewicz et al., 1994). Radioactivity was profiled with a Radiomatic Flo-One beta radioactivity detector (model CR, Packard) after the addition of Flo-Scint III (Packard) at 3 ml/min. Recovery of radioactivity applied to the HPLC column was 100%.

LC/MS Analysis. Two systems were used for separation and identification of radioactive components.

System 1. LC/MS with on-line radioactivity detection was performed using a Hewlett-Packard 1090L HPLC system (Atlanta, GA) in series with a Raytest Ramona 92 radioactivity detector (Pittsburgh, PA) coupled to a Fisons Quattro II triple quadrupole mass spectrometer (Manchester, UK). Chromatographic conditions were identical to those used in profiling, and samples were introduced using atmospheric pressure chemical ionization (APcI) in the positive and negative ion modes.

System 2. LC/MS was performed using an Isco DM 100 gradient system coupled to a VG Autospec Ultima-Q hybrid mass spectrometer (Manchester, UK). Sample were introduced using either APcI or electrospray ionization (ESI). Increasing the cone voltage from 20 V to 40 V resulted in source-induced fragmentation. A column identical to that above was used to separate analytes; however, the mobile phase consisted of acetonitrile (A) and water with 0.1% (v/v) glacial acetic acid (B). Components were eluted over 30 min using a linear gradient with A starting at 30% and ending at 90%. Analytes eluted at a flow rate of 1 ml/min (APcI) or 0.2 ml/min (ESI).

Data Analysis. Cumulative recovery of radioactivity in urine and feces is reported as per cent of administered dose. Plasma radioactivity concentrations are expressed as nanogram equivalents of AT free acid/ml. The per cent changes in total CYP450 and resorufin formation from ethoxy- and pentoxyresorufin are compared with values obtained from vehicle-treated control groups. The formation of 6-hydroxytestosterone from testosterone and the disappearance of lauric acid were also compared with those of vehicle-treated control groups.

	Results

Top Abstract Introduction Materials & Methods Results Discussion References

Plasma Radioactivity and Metabolite Identification. After oral administration of [¹⁴C]AT on days 1 and 14, radioactivity reached maximum concentrations at 1 hr and then declined rapidly (table 1). Maximum achieved plasma radioactivity concentrations were considerably lower after multiple-dose administration compared with single dose. The total plasma radioactivity AUC(_0-12) value of day 1 decreased by 45% after multiple-dose administration.

View larger version (14K): [in this window] [in a new window]   Fig. 1.   Gradient HPLC-radioactivity chromatogram of a 1-hr plasma sample. A, profile of plasma on day 1. B, profile of plasma on day 14.

View larger version (18K): [in this window] [in a new window]   Fig. 2.   CID mass spectra of AT eluting at 71 min. A, product ion spectrum of m/z 559 [M+H]+. B, product ion spectrum of m/z 557 [M-H].

View larger version (24K): [in this window] [in a new window]   Fig. 3.   CID mass spectra of a mouse plasma component eluting at 88.7 min. A, product ion spectrum of m/z 515 [M+H]+. B, product ion spectrum of m/z 513 [M-H].

View larger version (18K): [in this window] [in a new window]   Fig. 4.   ESI of a mouse plasma component with an LC retention time of 88.5 min. A, positive ion mass spectrum of m/z 497 [M+H]+. B, negative ion mass spectrum of m/z 495 [M-H].

Urine and Feces. The per cent of radioactivity recovered from urine and feces after administration of [¹⁴C]AT (table 3) shows feces to be the major route of excretion, with 98% of dose recovered within 48 hr. Under steady-state conditions, recovery of a higher percentage of the administered radioactivity was expected in feces within the first 24 hr compared with fecal recovery after a single dose of [¹⁴C]AT. Radioactive components were detected in day 1 and day 14 fecal extracts (table 4); however, the per cent of each component varied. Fecal radioactivity profiles were similar between treatment groups with most of radioactivity associated with metabolites. Many additional metabolite peaks present in feces were not found in plasma. Unchanged drug contributed less than 2% of the radioactivity after either treatment, indicating good absorption of AT.

View larger version (18K): [in this window] [in a new window]   Fig. 5.   ESI of a mouse feces component with an LC retention time of 68.5 min. A, positive ion mass spectrum of m/z 573 [M+H]+. B, negative ion mass spectrum of m/z 571 [M-H].

View larger version (16K): [in this window] [in a new window]   Fig. 6.   APcI mass spectrum (cone voltage of 40V) of a component in mouse feces (LC retention time of 83.5 min) of m/z 531 [M+H]+.

Enzyme Induction Investigations. Total spectral CYP450 concentrations increased by 5 and 30% in the 1.0 and 200 mg/kg treatment groups compared with control animals. Incubations with liver microsomes using ER, PR, testosterone, and lauric acid were employed as indicators of catalytic activity of CYP1A, -2B, -3A, and -4A, respectively. The results in table 6 indicate that CYP1A showed a minor dose-dependent decrease in ER activity at the 200 mg/kg AT dose. The catalytic activity of CYP2B was variable with an initial increase in activity at the lower AT dose and a decrease at the higher one. CYP3A was the only CYP450 that demonstrated a small increase in activity with increasing AT dose, i.e. 26% increase at 200 mg/kg of AT. CYP4A activity, as measured by the disappearance of lauric acid, decreased to 83 and 47% at the 1 and 200 mg/kg AT doses, respectively.

	Discussion

Top Abstract Introduction Materials & Methods Results Discussion References

This study was performed to evaluate in vivo AT metabolism, total CYP450 levels, and selected CYP450 activities in mice after an earlier multiple-dose toxicokinetic study demonstrated marked dose-related effects on plasma HMG-CoA reductase inhibitory activity. Comparison of plasma radioequivalent concentrations after [¹⁴C]AT on days 1 and 14 showed a dramatic decline. Prior studies (Shum et al., 1996) have shown that the dosing duration altered neither half-life nor the time to t_max. Recovery of radioactivity was essentially complete within 48 hr after either dosing regimen, with greater than 98% of the dose recovered in feces. In general, metabolite profiles of plasma and fecal extracts after both single- and multiple-dose administration were similar. Because of very low plasma concentrations, we only were able to examine the 1-hr plasma sample. After multiple doses of AT, two additional radioactive peaks were present that were not found after a single dose; however, whether these peaks represent new or additional metabolites could not be determined. Most metabolites present in plasma extracts were determined to be products of -oxidation of the seven-carbon side chain. Minor peaks had also undergone aromatic hydroxylation. Identified in plasma extracts were AT (a minor component), a -oxidized hydroxy metabolite of AT, -oxidized AT, and unsaturated analogs of -oxidized metabolites. All radioactive components identified in plasma extracts were found in fecal extracts. Additional fecal metabolites were ortho- and para-hydroxy AT and other unsaturated products including AT. Michniewicz et al. (1994) reported that ortho- and para-hydroxy AT were major metabolites in rat and dog bile; however, these metabolites were present in very small amounts.

In the present studies, we found only a 5 and 30% increase in total spectral CYP450 concentration at the 1.0 and 200 mg/kg AT levels, respectively. In addition, small and/or fluctuating changes were observed in the catalytic activities of CYP1A1, -2B, and -3A. Only CYP3A showed an increase in catalytic activity with an increasing AT dose. These minor increases in total CYP450 and CPY3A activity at the highest dose of AT seem insufficient to explain entirely the discrepancy observed in the plasma radioactivity concentrations between mice receiving single and multiple doses. In contrast, the CYP4A activity decreased dramatically to 47% of control at the 200 mg/kg AT dose level. More extensive experiments are needed to elucidate the mechanism of the decrease in CYP4A activity after multiple-dose AT administration.

Metabolism by the -oxidation pathway in rodents has been observed with other HMG-CoA reductase inhibitors. -Oxidation of the dihydroxy side chain (as the open ring form) plays a role in the metabolism of lovastatin, simvastatin, and pravastatin (Halpin et al., 1993, Komai et al., 1992, Vickers et al., 1990, Vyas et al., 1990). With the synthetically made fluvastatin, the extent of metabolism is greatest in the mouse; however, detailed examination of the metabolites was not provided (Tse et al., 1990). In the dog, -oxidation of lovastatin, simvastatin, and pravastatin is a very minor metabolic pathway (Vickers et al., 1990). In addition, after oral administration of [¹⁴C]AT to bile fistula rats and dogs, examination of the bile indicates that -oxidation is not a major pathway in these species (Michniewicz et al., 1994). The major metabolites identified in rat and dog resulted from aromatic hydroxylation to ortho- and para-hydroxy AT. In the mouse, although aromatic hydroxylation occurs, the major contribution is by -oxidation of these and other metabolites. That unsaturated AT and other unsaturated metabolites contribute to the majority of radioactivity in fecal extracts does not seem to be an artifact. These unsaturated metabolites have not been observed in similar extracts from other species.

The decrease in plasma concentrations after multiple-dose administration cannot be explained entirely by induction of the CYP450 enzymes examined. The expected shorter plasma half-life of AT inhibitory equivalents and radioequivalents, a pharmacokinetic feature of induction, was not detected after multiple-dose administration in the mouse. Other HMG-CoA reductase inhibitors have not shown a decrease in plasma-equivalent concentrations after multiple-dose administration. CYP3A in the gut wall does not seem to contribute significantly to AT metabolism in the mouse because most of the metabolites detected result from -oxidation rather than metabolism by CYP3A. The source of the effect seems more complicated than simple induction of hepatic cytochrome P450s; perhaps transport or -oxidation is involved. The induction of -oxidation, which can occur in mitochondria and peroxisomes (Osmundsen et al., 1991), was not examined in this study. Recently, other investigators have reported the effects of HMG-CoA reductase inhibitors on fatty acid metabolism in rodent liver. Studies in rats administered simvastatin indicate that this drug is not a peroxisomal proliferator; fatty acid distribution, however, was altered in microsomal phosphatidylcholines (Mercenne et al., 1991). Comparison of in vitro and in vivo effects of lovastatin on fatty acid metabolism in rat liver indicate an increased fatty acid-oxidative capacity (Guzman et al., 1993).

Metabolites detected after AT administration indicate that metabolism in the mouse is much more extensive (fig. 7) with a greater predisposition toward -oxidation than either rat or dog. AT disposition in mice is affected by multiple-dose treatment, something not explained by CYP450 levels or activity. Clearly, the -oxidation pathway is favored in the mouse, but more detailed metabolism studies in mice are needed to assess the role of multiple-dose administration on this pathway.

View larger version (17K): [in this window] [in a new window]   Fig. 7.   Atorvastatin and identified metabolites in mouse. a indicates metabolites found in plasma; b indicates those found in feces.