Introduction

Abstract Introduction Results Discussion References

Classical antipsychotic drugs of the phenothiazine and butyrophenone classes have been established as effective agents for the treatment of schizophrenia, a mental disorder estimated to afflict 1% of the world population. However, these drugs are not effective in all patients or against all symptoms (1, 2). Secondly, an unacceptably high incidence of extrapyramidal symptoms limits the usefulness of these drugs (3). Laboratory and clinical findings have suggested that antagonism of serotonin 5-HT₂¹ receptors in the brain limits the undesirable motor side effects associated with dopamine receptor blockade and improves efficacy against the negative symptoms of schizophrenia (4-9).

ZIP is the hydrochloride salt of a benzisothiazolylpiperazine analog structurally related to the atypical antipsychotic drug tiospirone (10, 11). It was developed during a structure-activity investigation to find a compound that potently blocks dopamine D2 receptors while binding with even greater affinity to cerebral serotonin 5-HT₂ receptors (12-14). ZIP has extremely potent central serotonin 5-HT₂ (K_i = 0.42 nM) and potent dopamine D2 (K_i = 4.8 nM) receptor antagonistic properties. The ratio between these two affinities (5-HT₂:D2 = 11.4) is greater than observed for other antipsychotic drugs (12-14). In phase II clinical studies, ZIP has demonstrated good tolerability, particularly with regard to extrapyramidal symptoms, at doses that are associated with efficacy in patients with schizophrenia (15). The oral bioavailability of ZIP in human is 59% and its elimination half-life is 4 hr (16). ZIP is well tolerated in animals at doses producing effective blockade of dopaminergic behaviors and has a favorable separation between activities predictive of efficacy and side effect liability. ZIP is partially absorbed in rats and dogs, with absolute oral bioavailability of 39-60%. The elimination half-life is about 1 hr in rats and 2.3 hr in male dogs (17).

As with all new drugs, it is necessary that the metabolic fate of ZIP be evaluated in vivo. Metabolic studies conducted in living animals provide the ultimate information regarding metabolic pathways and disposition of the molecule that can be used to correlate or interpret efficacy and toxicity data. In addition, the identified metabolites can be evaluated for pharmacological activity. The objective of the present study was to investigate the metabolism and excretion of ZIP in LE rats after administration of an oral dose of a mixture of ³H- and ¹⁴C-labeled ZIP (fig. 1). The use of two labels greatly facilitated the tracing and identification of metabolites formed through cleavage of ZIP, as observed for the structurally related drug tiospirone (18, 19).

View larger version (8K): [in this window] [in a new window]   Fig. 1.   Structures of 14C- and 3H-labeled ZIP.

Materials and Methods

General Chemicals. Commercially obtained chemicals and solvents were of HPLC or analytical grade. -Glucuronidase (from Helix pomatia, type H-1, with sulfatase activity) was obtained from Sigma Chemical Co. (St. Louis, MO). YMC basic columns were obtained from YMC (Wilmington, NC). Ecolite-(+) scintillation cocktail was obtained from ICN (Irvine, CA). Carbsorb and Permafluor scintillation cocktails were purchased from Packerd Instrument Co. (Downers Grove, IL). Diazomethane was generated just before use from N-methyl-N-nitroso-p-toluene sulfonamide obtained from Aldrich (Milwaukee, WI). Silica gel was obtained from Fisher Scientific (Springfield, NJ).

General Instrumentation. The structures of all synthetic compounds were supported by their ¹H-NMR (Bruker AM-300) and FAB mass spectra. Melting points were determined on a Thomas-Hoover capillary melting point apparatus and are uncorrected. FAB mass spectra of synthetic standards were obtained on a VG-70/70 mass spectrometer, using glycerol as a matrix.

Radiolabeled Drug and Reference Compounds. [¹⁴C]ZIP, specifically labeled at C2 of the ethyl group attached to the piperazinyl nitrogen, was synthesized by Friedel-Crafts acylation of 6-chlorooxindole with [¹⁴C]chloroacetyl chloride, followed by triethylsilane reduction of the aryl carbonyl and coupling with BITP (20). [¹⁴C]ZIP showed a specific activity of 9.0 mCi/mmol (21.8 µCi/mg) and a radiochemical purity of 98%, as determined by radio-HPLC. [³H]ZIP, specifically labeled at the C7-position of the benzisothiazole ring, was synthesized by reduction of 7-bromo-ZIP with tritium gas (20). [³H]ZIP showed a specific activity of 20.66 Ci/mmol (50.0 µCi/mg) and a radiochemical purity of 99% (by radio-HPLC). BITP, BITP-SO, BITP-SO₂, BITP-SO₂-lactam, OX-Cl, and OH-ZIP were prepared as described previously (20, 21). ZIP-N-oxide was prepared by reaction of ZIP with m-chloroperbenzoic acid.

ZIP-SO. A mixture of BITP-SO (75 mg, 0.32 mmol), OX-Cl (73 mg, 0.32 mmol), Na₂CO₃ (100 mg, 0.96 mmol), and KI (5 mg, 0.03 mmol) in 1.0 ml of acetonitrile was stirred and heated to reflux for 32 hr. After cooling, the solvent was removed in vacuo and the residue was redissolved in 5 ml of anhydrous CH₃OH, filtered through a pad of diatomaceous earth, and concentrated in vacuo to a brown oil. Chromatography on silica gel (40 mm) eluted with CH₂Cl₂/CH₃OH (96:4) gave the desired product as a pale yellow solid [20 mg; m.p. 215-218°C; FAB mass spectrum: m/z 431 (³⁷Cl), 429 (MH⁺); ¹H-NMR (dimethylsulfoxide-d₆):  10.45 (s, 1H, N), 8.23 (d, 1H, BIT C7-), 8.05 (d, 1H, BIT C4-), 7.75 (m, 2H, BIT C5,6-), 7.25 (s, 1H, indole C7-), 6.82 (s, 1H, indole C4-), 3.95 (m, 4H, NC), 3.45 (s, 2H, indole C3-), 2.80 (m, 2H, ArC), 2.70-2.50 (m, 6H, NC)].

ZIP-SO₂. A mixture of BITP-SO₂ (81 mg, 0.32 mmol), OX-Cl (73 mg, 0.32 mmol), Na₂CO₃ (100 mg, 0.96 mmol), and KI (5 mg, 0.03 mmol) in 1.0 ml of acetonitrile was stirred and heated to reflux for 32 hr. Work-up of the reaction mixture as described for ZIP-SO gave the desired product as a pale yellow solid (22 mg). It was converted to the hydrochloride salt by stirring with 3 N HCl, filtering the solids, triturating with acetonitrile, and drying under vacuum at 70°C overnight to a pale pink solid [m.p. 295-297°C; FAB mass spectrum: m/z 447 (³⁷Cl), 445 (MH⁺); ¹H-NMR (dimethylsulfoxide-d₆):  10.45 (s, 1H, N), 8.22 (dd, 1H, BIT C7-), 8.04 (dd, 1H, BIT C4-), 7.74 (m, 2H, BIT C5,6-), 7.23 (s, 1H, indole C7-), 6.78 (s, 1H, indole C4-), 3.95 (m, 4H, NC), 3.45 (s, 2H, indole C3-), 2.84 (m, 2H), 2.70-2.50 (m, 6H, NC)].

Animals. LE rats were purchased from Charles River Laboratories (Stoneridge, NY). Animals were quarantined for a minimum of 7 days before treatment and were maintained on a 12-hr light/dark cycle. Animals were fed food and water ad libitium and maintained with United States Department of Agriculture guidelines for the care and use of laboratory animals.

Urinary, Biliary, and Fecal Excretion Studies. A group of three male and three female LE rats (240-260 g) were implanted with a PE-10 cannula into the common bile duct under anesthesia (sodium phenobarbital via ip injection). The animals were housed individually in stainless steel metabolic cages and were allowed to recover overnight before drug administration. The animals were orally administered 10 mg/kg radiolabeled ZIP. The dose was composed of a mixture of ³H- and ¹⁴C-labeled ZIP diluted with unlabeled drug to specific activities of 3.72 mCi/mmol and 2.82 mCi/mmol, respectively, and was administered as a suspension in 0.5% methyl-cellulose at a concentration of 1.5 mg/ml. Each animal received ~23 µCi of ³H-labeled and ~17 µCi of ¹⁴C-labeled material. All animals were fed at 2.5 hr after the dose and received electrolyte Krebs-Ringer solution throughout the study. Urine, bile, and feces from each animal were quantitatively collected for 7 days (168 hr after the dose) at 0-8, 8-24, 24-48, 48-72, 72-96, 96-120, 120-144, and 144-168 hr after the dose; the first feces sample was collected at 0-24 hr after the dose. The volumes of urine and bile samples were recorded, and all of the biological samples were stored at 20°C until analysis.

Plasma Time Course Study. Another group of LE rats (N = 14/gender, 270-320 g) were dosed by gavage with a 10 mg/kg dose of a mixture of ³H- and ¹⁴C-labeled ZIP. Blood was collected in heparinized tubes, by decapitation of two male and two female rats, at 0, 0.5, 1, 2, 4, 6, 8, 12, and 24 hr after the dose. The blood samples were centrifuged at 1000 g for 10 min to obtain the plasma. Samples were transferred to clean tubes and stored at 20°C until analysis.

Determination of Radioactivity. Total radioactivity in urine, bile, and plasma was quantitated by counting sample aliquots (20-50 µl) using a ³H/¹⁴C "dual-label" program with a Packard 2500 TR liquid scintillation counter. Ecolite-(+) scintillation cocktail (5 ml; ICN) was used for determination of the radioactivity in the samples. Quench curves were prepared separately for [³H]ZIP and [¹⁴C]ZIP. Fecal samples were lyophilized overnight and homogenized on a paint can shaker (Red Devil, model 5410), using chrome-plated 3/8 and 7/16 ball bearings. A small aliquot (20-60 mg) was combusted using a Packard Tricarb oxidizer (Irvine, CA). The liberated ¹⁴CO₂ and ³H₂O were trapped, and the radioactivity in the trapped samples was determined by counting in the liquid scintillation counter. Combustion efficiencies were determined by combusting ¹⁴C and ³H standards in an identical manner. The samples obtained at 0 hr after the dose were used as controls and counted to obtain a background count rate.

Pharmacokinetic Analysis. Plasma concentrations of the unchanged ZIP were determined by a validated HPLC/MS/MS assay (22). Pharmacokinetic parameters were determined by standard methods. The AUCs were calculated from plasma concentrations of ZIP and total radioactivity, using a trapezoidal approximation of area and using 0 as the time 0 concentration. The t_max value was the time of the first occurrence of the maximal plasma concentration.

Extraction of Metabolites from Biological Samples. Urine (1 ml) from each animal was centrifuged, and small aliquots (50 µl) were injected into the HPLC system without further purification. Bile (1 ml) was diluted with 4 ml of acetonitrile, and the precipitated protein was removed by centrifugation. The pellet was washed with an additional 2 ml of acetonitrile, and the two supernatants were combined. Small aliquots of supernatant and pellet were counted. The supernatant was concentrated and dissolved in 1 ml of mobile phase, and an aliquot (50 µl) was injected into the HPLC system.

HPLC. HPLC was carried out on a system that consisted of a Rheodyne injector for manual injections, a LDC/Milton Roy constametric CM4000 gradient pump, a Waters Lambda-Max model 481 UV detector, a -RAM, and a SP 4200 computing integrator. Chromatography was carried out on a YMC basic HPLC column (4.6 mm × 250 mm, 5 µm) with a binary mixture of 20 mM ammonium acetate (pH 5.0, solvent A) and methanol (solvent B). The mobile phase initially consisted of solvent A/solvent B at 90:10 for 10 min. It was then linearly programmed to solvent A/solvent B at 20:80 over 50 min. Chromatography was carried out under isocratic conditions for 7 min and then programmed back to the starting solvent mixture over 8 min. The system was allowed to equilibrate for approximately 10 min before the next injection was made. The retention times of radioactive peaks were compared with those of the synthetic standards.

Quantitative Assessment of Metabolite Excretion. Quantification of the metabolites was carried out by measuring radioactivity in the individual peaks that were separated on HPLC, using a -RAM. The -RAM provided an integrated printout in dpm and percentage of the radiolabeled material, as well as peak representation. The -RAM was operated in the homogeneous liquid scintillation counting mode, with addition of 4 ml/min Ecolite scintillation cocktail to the eluent after UV detection. For simultaneous monitoring of ³H- and ¹⁴C-labeled compounds, efficiencies of 37% for ³H and 55% for ¹⁴C were used, with a compensation for ¹⁴C spill-over into the ³H window of 31%. These parameters were determined through separate injections of singly labeled standards.

Enzyme Hydrolysis. Pooled rat bile and urine samples (0-24 hr, 0.5 ml each) were adjusted to pH 5 with sodium acetate buffer (0.1 M) and treated with 2500 units of -glucuronidase/sulfatase. The mixture was incubated in a shaking water bath at 37°C for 12 hr and diluted with acetonitrile. The precipitated protein was removed by centrifugation. The pellet was washed with an additional 2 ml of acetonitrile, and the two supernatants were combined. The supernatant was concentrated and dissolved in 0.5 ml of mobile phase, and an aliquot (50 µl) was injected into the HPLC system. Incubation of bile and urine samples without the enzyme served as a control.

Derivatization. One polar metabolite, M4, was isolated by HPLC and methylated with diazomethane. The purified metabolite M4 (100-200 ng) was dissolved in methanol (100 µl), and freshly prepared ethereal diazomethane (200 µl) was added. After standing for 15 min at room temperature, the solvent was removed with a stream of nitrogen and the residue was dissolved in the HPLC mobile phase.

MS. Analysis of the metabolites was performed on a Perkin-Elmer (Norwalk, CT) SCIEX API III HPLC/MS/MS system using ion spray. The effluent from the HPLC column was split and about 50 µl/min was introduced into the atmospheric ionization source via an ion spray interface. The remaining effluent was directed into the flow cell of the -RAM. The -RAM response was recorded in real time by the mass spectrometer data system, which provided simultaneous detection of radioactivity and MS data. The delay in response between the two detectors was about 0.2 min, with the mass spectrometric response being recorded first. The ion spray interface was operated at 6000 V, and the mass spectrometer was operated in the positive mode. CID studies were performed using argon gas at a collision energy of 25-28 eV and a collision gas thickness of 3.5 × 10¹⁴ molecules/cm².

	Results

Abstract Introduction Results Discussion References

Excretion. Preliminary studies showed that, at 24 hr after po administration of 10 mg/kg [³H]ZIP to rats, excretion of tritiated water amounted to only 0.2% of the dose (data not shown). The plasma sample obtained 24 hr after the dose had only 0.2% of the radioactive dose in the recovered water. These findings suggest that the tritium label was on a metabolically stable position.

View larger version (23K): [in this window] [in a new window]   Fig. 2.   Mean cumulative percent recovery of radioactivity in urine, bile, and feces after oral administration of 14C- and 3H-labeled ZIP to male (N = 3) (a) and female (N = 3) (b), bile duct-cannulated, LE rats.

Circulation. The mean plasma concentration-time curves for ZIP and total radioactivity for male and female rats are shown in fig. 3. Absorption of ZIP was rapid in both male and female rats, as indicated by the early appearance of radioactivity in plasma after oral administration. The plasma concentration of ZIP and metabolites reached a peak of 1.92 µg equivalents/ml and 2.42 µg equivalents/ml for male and female rats, respectively, at 1 hr after the dose. Based on AUC (0-12 hr) values, 41 and 48% of circulating radioactivity, respectively, was attributable to parent drug (table 2).

View larger version (24K): [in this window] [in a new window]   Fig. 3.   Mean plasma concentration-time curves for total radioactivity and ZIP in male (N = 2) (a) and female (N = 2) (b) rats.

Identification of the Major Metabolites. Metabolite Profile in Urine. An HPLC method was developed for the separation of ZIP and its metabolites. Baseline separation of all metabolites was obtained using this system. Representative profiles of the metabolites in urine samples (0-24 hr) from a male rat after oral administration of [³H]ZIP and [¹⁴C]ZIP, with on-line radioactivity monitoring of ¹⁴C and ³H, are given in fig. 4. Treatment of the urine samples with -glucuronidase/arylsulfatase did not result in any changes in the HPLC profile of the sample, indicating that few, if any, of the metabolites were present as conjugates. A total of 11 radioactive peaks were detected in the chromatograms from both male and female rats. The radio-HPLC data are presented with the peak height normalized to that of the largest peak. Metabolites that did not undergo cleavage at the piperazinyl nitrogen are expected to be dual-labeled and are represented by HPLC peaks with similar heights and retention times in the two traces. Metabolites that underwent cleavage are expected to be singly labeled. Each of the cleaved metabolites is represented by a HPLC peak in one of the traces and a smaller peak representing spillover in the other trace. The metabolites were quantified with on-line integration of the radiochromatographic peaks. The percentages of the metabolites excreted in relation to the total radioactivity in urine are presented in table 3. Judging from the quantities of cleaved metabolites, there was a gender-related difference in the amounts of cleaved metabolites excreted. Approximately 45% of the total radioactivity in male rat urine and 26.9% of the total radioactivity in female rat urine was attributable to the cleaved products (table 3).

View larger version (24K): [in this window] [in a new window]   Fig. 4.   HPLC radiochromatograms of ZIP metabolites in male rat urine.

View larger version (15K): [in this window] [in a new window]   Fig. 5.   CID product ion spectra of ZIP m/z 413 (35Cl) (top) and m/z 415 (37Cl) (bottom).

Urinary Peak II (M1). Peak II had a retention time of 7.30 min on HPLC and was detected only in the ³H chromatogram, indicating that it was a cleaved product. It showed a protonated molecular ion at m/z 252, 32 Da higher than BITP, and its CID product ion spectrum showed fragment ions at m/z 235 and 209. M1 coeluted with authentic BITP-SO₂ on HPLC and had an identical CID product ion spectrum. Based on these data, M1 was identified as BITP-SO₂.

Urinary Peak III (M2). Peak III had a retention time of 8.02 min and was detected only in the ³H chromatogram, indicating that it was also a cleaved product. It showed a protonated molecular ion at m/z 236, 16 Da higher than BITP, suggesting that an oxygen atom had been added to the BITP moiety. CID product ions of m/z 236 showed intense fragment ions at m/z 177, 150, 134, and 85. M2 coeluted with authentic BITP-SO on HPLC and had an identical CID product ion spectrum. Based on these data, M2 was identified as BITP-SO.

Urinary Peak IV (M3). Peak IV showed a protonated molecular ion at m/z 266, 46 Da higher than BITP, suggesting the addition of three oxygen atoms with subsequent loss of two hydrogen atoms. The presence of only ³H label in the chromatogram further suggested that it was a cleaved product. The CID product ion spectrum of m/z 266 showed an intense fragment ion at m/z 209. It coeluted with authentic BITP-SO₂-lactam on HPLC and had an identical CID product ion spectrum. Based on these data, M3 was identified as BITP-SO₂-lactam.

Urinary Peak V (M4). Peak V had a retention time of 22.1 min on HPLC and it had only ¹⁴C label, indicating that this metabolite was a cleaved product and contained an oxindole moiety. The metabolite did not show any distinct pseudomolecular ion, probably due to a low molecular weight. However, the reconstructed selected ion profile of peak V indicated an ammoniated adduct ion at m/z 243. The metabolite was, therefore, isolated by HPLC and was reacted with diazomethane. Treatment of M4 with diazomethane resulted in the disappearance of the peak coinciding with M4 and appearance of a new peak that had the retention time of 36.4 min on HPLC (fig. 6). The derivatized metabolite M4 showed an intense ion at m/z 257 (M+NH₄)⁺, 14 Da higher than that of the parent metabolite, indicating that the COOH functionality was present in the molecule. This was further supported by the fact that the CID product ion spectrum of the derivatized metabolite (m/z 257) showed intense ions at m/z 240 (MH)⁺ and 180, showing loss of a CH₃COOH molecule from the molecular ion (fig. 6). Based on these data, M4 was identified as 6-chloro-2-oxo-(2,3-dihydro-1H-indol-5-yl)acetic acid.

View larger version (21K): [in this window] [in a new window]   Fig. 6.   -RAM response (a), reconstructed selected ion monitoring response (b), and CID product ion spectrum of m/z 257 (M+NH4)+ (c) for ZIP metabolite M4 after treatment with diazomethane.

Urinary Peak VI (M5). Peak VI had a retention time of 32.1 min and was found to be a mixture of two metabolites (M5 and M6). One of the metabolites, M5, showed a protonated molecular ion at m/z 220, and its CID mass spectrum exhibited the fragment ion at m/z 177 (loss of 43), the same as BITP. M5 coeluted with synthetic BITP and had an identical CID product ion spectrum. Based on these data, M5 was identified as BITP.

Urinary Peak VI (M6). The second metabolite in peak VI showed a protonated molecular ion at m/z 445, 32 Da higher than the parent drug, suggesting that the molecule had undergone two oxidations. Its CID product ion spectrum (m/z 445) showed a fragment ion at m/z 166, suggesting the presence of two oxygen atoms on the benzisothiazole moiety. The presence of fragment ions at m/z 194, 263, and 280 in its CID spectrum further suggested that the oxidations had occurred at positions remote from the oxindolylpiperazine part of the molecule. M6 did not coelute with synthetic ZIP-SO₂ standard on HPLC. Based on these data, M6 was identified as the intact molecule with the addition of two oxygen atoms at the benzisothiazole moiety. The exact sites for oxidation were, however, not established by the mass spectral data.

Urinary Peak VII (M7). M7 showed a protonated molecular ion at m/z 447, 34 Da higher than the parent drug, suggesting that the molecule had undergone monooxidation and addition of a molecule of water. Its CID product ion spectrum (m/z 447) showed a fragment ion at m/z 168, indicating the presence of an oxygen atom and a water molecule on the benzisothiazole moiety (fig. 7). This was further supported by the presence of other characteristic fragment ions at m/z 194 and 280 in its CID product ion spectrum. Based on these data, M7 was tentatively identified as 2-{4-[2-(6-chloro-2-oxo-2,3-dihydro-1H-indol-5-yl)ethyl]piperazine-1-carbonyl}benzenesulfinic acid amide.

View larger version (17K): [in this window] [in a new window]   Fig. 7.   CID product ion spectrum of m/z 447 for ZIP metabolite M7.

Urinary Peak VIII (M8). M8 showed a protonated molecular ion at m/z 445, 32 Da higher than the parent drug, suggesting that the molecule had undergone two oxidations. Its CID product ion spectrum (m/z 445) showed an intense fragment ion at m/z 194, indicating that the oxidations had occurred at positions remote from the oxindole part of the molecule. M8 coeluted with authentic ZIP-SO₂ on HPLC and had an identical CID product ion spectrum (fig. 8). Based on these data, M8 was identified as ZIP-SO₂.

View larger version (18K): [in this window] [in a new window]   Fig. 8.   CID product ion spectra of m/z 445 for ZIP metabolite M8 and synthetic standard.

Urinary Peak IX (M9). M9 exhibited a protonated molecular ion at m/z 429, 16 Da higher than the parent drug, suggesting that it was a monooxidation product of ZIP. Its CID product ion spectrum (m/z 429) showed fragment ions at m/z 194, 280, and 150, suggesting the presence of an oxygen atom on the benzisothiazole moiety. The retention time of M9 was different from those of the synthetic ZIP-SO and OH-ZIP standards. Based on these data, M9 was identified as the intact molecule with the addition of one oxygen atom at the benzisothiazole moiety, other than the sulfur. These data could not establish the exact sites for the oxidation.

Urinary Peak X (M10). Peak X also indicated a protonated molecular ion at m/z 429, 16 Da higher than the parent drug, suggesting that a single atom of oxygen had been added to the molecule and it was an isomer of M9. Its CID product ion spectrum showed the fragment ions at m/z 194 and 232, indicating that the oxindole moiety was unchanged. The characteristic fragment ion at m/z 99 (piperazine-CH₂⁺) further suggested that the oxidation had occurred at the benzisothiazole moiety. M10 coeluted with synthetic ZIP-SO on HPLC and had an identical CID product ion spectrum (fig. 9). Based on these data, peak X was identified as ZIP-SO.

View larger version (21K): [in this window] [in a new window]   Fig. 9.   CID product ion spectra of m/z 429 for ZIP metabolite M10 and synthetic standard.

Urinary Peak XI (M13, Unchanged Drug). Peak XI showed the pseudomolecular ion at m/z 413, similar to ZIP. Its CID product ion spectrum showed fragment ions at m/z 194 and 159, indicating that the oxindole ring was unchanged. The fragment ions at m/z 220 and 177 suggested that there was no substitution on the BITP moiety. M13 coeluted with synthetic ZIP on HPLC and had an identical CID product ion spectrum. Based on these data, peak XI was identified as unchanged drug.

Metabolites in Bile. Metabolite Profile in Bile. A major portion of the radioactivity in bile was recovered from 0-8 and 8-24 hr. Therefore, bile samples from 0-8 and 8-24 hr were pooled, and the pooled sample was used for the identification of metabolites. Pooled bile samples were diluted with 4 volumes of acetonitrile, and the precipitated proteins were removed by centrifugation. Approximately 89% of the total radioactivity was recovered in the supernatant and 9% remained in the pellet. Representative profiles of the metabolites in bile (0-24 hr) from a male rat, with on-line radioactivity monitoring of ¹⁴C and ³H, are given in fig. 10. The relative amounts of metabolites excreted in the bile, in relation to the total radioactivity, are presented in table 3. There was no substantial gender-related difference in the profiles of the biliary metabolites. Only the unchanged drug was excreted 3 times more in the bile of female rats than male rats. ZIP and five radioactive peaks, accounting for 66% of the total radioactivity, were identified in bile. Five of these metabolites were identical to those identified from the urine, i.e. peaks I (M6), II (M7), III (M8), IV (M9), and VI (M13). Only one additional metabolite was identified from the bile.

View larger version (28K): [in this window] [in a new window]   Fig. 10.   HPLC radiochromatograms of ZIP metabolites in rat bile.

Biliary Peak V (M12). Biliary peak V showed a protonated molecular ion at m/z 447, 34 Da higher than the parent drug, suggesting that the molecule had undergone monooxidation and addition of a water molecule. The CID product ion spectrum of peak V showed three characteristic fragment ions, at m/z 212, 298, and 150 (fig. 11). The fragment ions at m/z 212 and 298 suggested that a molecule of water had been added to the oxindole moiety. The fragment ion at m/z 150 indicated that the monooxidation had occurred on the benzisothiazole moiety. Based on these data, this metabolite was tentatively identified as (2-amino-4-chloro-5-{2-[4-(1-oxobenzisothiazol-3-yl)piperazin-1-yl]ethyl}phenyl)acetic acid.

View larger version (17K): [in this window] [in a new window]   Fig. 11.   CID product ion spectrum of m/z 447 for ZIP metabolite M12.

Metabolites in Feces. Metabolite Profile in Feces. Representative HPLC profiles of the fecal metabolites from a male rat after oral administration of [³H]ZIP and [¹⁴C]ZIP, with on-line radioactivity monitoring of ¹⁴C and ³H, are given in fig. 12. There was no gender-related difference in the profiles of the metabolites. Most of the radioactivity (93%) in feces was attributable to the parent drug (table 3). Three minor metabolites were detected in feces of both male and female rats. These metabolites accounted for approximately 7% of the total radioactivity in feces. All three metabolites were due to oxidation on the BITP moiety, and two of these metabolites were identical to those identified from the urine and bile, i.e. peaks I (M9) and II (M10).

View larger version (18K): [in this window] [in a new window]   Fig. 12.   HPLC radiochromatograms of ZIP metabolites in rat feces.

Fecal Peak III (M11). Fecal peak III showed a protonated molecular ion at m/z 429, 16 Da higher than the parent drug, indicating the addition of an oxygen atom to the molecule. Its CID spectrum showed a fragment ion at m/z 194, suggesting that the oxindole moiety was unchanged. The fragment ion at m/z 232 suggested the presence of oxygen on the BITP moiety. M11 coeluted with synthetic ZIP-NO on HPLC and had an identical CID spectrum (fig. 13). Based on these data, M11 was identified as ZIP-NO.

View larger version (19K): [in this window] [in a new window]   Fig. 13.   CID product ion spectra of m/z 429 for ZIP metabolite M11 and synthetic standard.

	Discussion

Abstract Introduction Results Discussion References

A mixture of ZIP labeled with ¹⁴C at C2 of the ethyl group attached to the piperazinyl nitrogen and ZIP labeled with ³H at the C7-position of the benzisothiazole moiety was administered orally to LE rats. The use of two labels at these positions not only facilitated the tracing of metabolites that were formed through cleavage of ZIP but also aided in their identification. ZIP was extensively metabolized before excretion in bile and urine. Only a small percentage of the unchanged drug was found in the bile. Approximately 41% of the radioactivity was excreted in the urine and bile of rats after administration of ¹⁴C- and ³H-labeled ZIP. The major portion of the unabsorbed dose recovered in feces was attributable to the parent drug.

The profiles of metabolites before and after treatment with -glucuronidase were similar, suggesting that the elimination of the drug was due to only phase I metabolism and that phase II metabolic pathways did not play a significant role in the metabolism of ZIP. In addition to the minor amounts of drug, a total of 12 metabolites were identified in rats. The structures of eight metabolites were unambiguously confirmed by comparison of their chromatographic and mass spectral fragmentation properties with those of synthetic standards, and four metabolites were tentatively identified based on their fragmentation patterns. The major metabolites found in bile and feces were also found in urine, although there were significant quantitative differences (table 3). Metabolites were identified by ion spray LC/MS/MS, using parent ion and product ion scanning techniques, with simultaneous radioactivity monitoring. These techniques have proved to be very useful for rapid screening and identification of drug metabolites (23-25). ZIP and its synthetic metabolites gave very intense protonated molecular ions. CID product ion spectra of protonated molecular ions provided structurally significant fragment ions.

In an attempt to establish the exact sites of oxidation, synthesis of some potential metabolites (ZIP-SO, ZIP-SO₂, and ZIP-NO) was undertaken. Preliminary attempts to oxidize ZIP directly at the sulfur atom, using published procedures (21), gave a mixture of several products. This may be the result of instability of the oxindole moiety under acidic conditions. The metabolites ZIP-SO and ZIP-SO₂ were, therefore, synthesized by oxidation of BITP and then condensation of the intermediates with OX-Cl.

The metabolism of ZIP occurred by four major and two minor routes (fig. 14). The major routes included N-dealkylation of the ethyl side chain attached to the piperazinyl nitrogen, oxidation at sulfur (resulting in the formation of sulfoxide and sulfone), oxidation on the benzisothiazole moiety (other than sulfur), and hydration of the CN bond and subsequent oxidation at the sulfur of the benzisothiazole moiety. The minor routes involved N-oxidation at the piperazine ring and hydrolysis of the oxindole moiety. The metabolite patterns in male and female rats were qualitatively very similar. However, there were some gender-related quantitative differences in the excretion of urinary metabolites. N-Dealkylation was the major route of metabolism in male rats. In female rats, the major metabolites were due to oxidation at the sulfur of the benzisothiazole ring.

View larger version (20K): [in this window] [in a new window]   Fig. 14.   Proposed biotransformation pathways for ZIP in rats.

The first major route, N-dealkylation, was analogous to those observed for the structurally related drugs (18, 26, 27). The N-dealkylation products of ZIP were identified as BITP-SO₂ (M1), BITP-SO (M2), BITP-SO₂-lactam (M3), 6-chloro-2-oxo-(2,3-dihydro-1H-indol-5-yl)acetic acid (M4), and BITP (M5). The metabolites M1, M2, M3, M4, and M5, resulting from N-dealkylation, accounted for 42-45% of the radioactivity in male rats and only 22-27% of the radioactivity in female rats. On the other hand, the metabolites resulting from oxidation on the benzisothiazole ring represented 40 and 60% of the total radioactivity present in urine of male and female rats, respectively (table 3). It is not clear whether metabolites M1 and M2 were derived by N-dealkylation of ZIP-SO and ZIP-SO₂, respectively, or by oxidation of BITP. However, the formation of M3 could be speculated to be by oxidation of BITP-SO₂. Based on these data, it could be concluded that the N-dealkylation of ZIP formed BITP, which was further oxidized to form metabolites M1, M2, and M3.

The formation of sulfoxide and sulfone is common with sulfur-containing drugs (18, 19, 28), and this suggests that the benzisothiazole moiety is more susceptible to oxidation at sulfur than to aromatic hydroxylation. Oxidation at the -carbon to the nitrogen, resulting in the formation of a lactam, has been observed previously for nicotine (29), encainide (30), flecainide (31), and several other drugs (32). Oxidation on the benzisothiazole moiety other than sulfur was, however, not observed for tiospirone and similar drugs (18, 19, 28). The novel metabolite M7, formed by the addition of water and subsequent oxidation of the benzisothiazole moiety, was tentatively identified as 2-{4-[2-(6-chloro-2-oxo-2,3-dihydro-1H-indol-5-yl)ethyl]piperazine-1-carbonyl}benzenesulfinic acid amide. The proposed structure of M7 did not immediately suggest the routes for its formation. Obviously there may be some intermediate compounds that were not detected as the precursors of this metabolite. A possible reaction sequence is given in fig. 15. Hydration of the CN double bond of ZIP could afford the carbinoldiamine, which could be hydrolytically rearranged by a reaction identical to the second step of N-dealkylation (33, 34). Either of the two C---N bonds could be cleaved to yield [(2-piperazin-1-yl)ethyl]-6-chloro-1,3-dihydroindol-2-one and hydrated ZIP. Because only metabolite M7 was detected, it could be speculated that the hydrolytic cleavage was selective. The results of this study do not indicate whether hydration of the benzisothiazole moiety occurs before or after oxidation at the sulfur. Because ZIP-SO was detected as a metabolite, the only conclusion that can be drawn is that sulfur oxidation is a prerequisite for hydration to occur.

View larger version (10K): [in this window] [in a new window]   Fig. 15.   Proposed mechanism for the formation of metabolite M7. OX-P, [(2-piperazin-1-yl)ethyl]-6-chloro-1,3-dihydroindol-2-one.

Two of the oxygenated metabolites (M6 and M9) detected in the urine and bile appear to be modified on the benzisothiazole moiety other than at the sulfur atom. There are two possible sites for such oxidation, i.e. oxidation at the nitrogen of the benzisothiazole ring to form ZIP-thiazole-N-oxide and aromatic hydroxylation. Metabolites derived from these oxidations could then undergo subsequent oxidation at sulfur to form the dioxygenated metabolites ZIP-N-oxide-S-oxide and ZIP-OH-S-oxide. Monooxygenated metabolite M9 did not coelute with synthetic OH-ZIP, suggesting that hydroxylation had not occurred at the 5-position of the benzisothiazole ring. However, the formation of other regioisomeric hydroxy metabolites could not be ruled out. The N-oxide of BIT has been prepared by the reaction of 1,2-benzisothiazolium chloride with hydroxylamine (35). In addition, the oxidation of two heterocyclic atoms in an aromatic ring was reported earlier by Offen et al. (36). They identified 4,5-dimethylthiazol-N-oxide-S-oxide, a metabolite of chloromethiazole, in humans (36). Attempts to synthesize a thiazole-N-oxide derivative of ZIP and to methylate metabolites M6 and M9 were unsuccessful. Therefore, metabolites M6 and M9 were only partially identified. Complete structural characterization of partially identified metabolites requires further work to establish the exact sites of oxidation.

In summary, the clearance of ZIP in rats was mainly by phase I oxidative metabolism, followed by elimination. Phase II metabolism did not play any significant role in the elimination of ZIP in rats. The use of two labels was very helpful in identifying the cleaved metabolites. The pharmacological properties of the major identified metabolites are currently under investigation. The present study will contribute to our understanding of the metabolism of other benzisothiazolylpiperazine drugs. In addition, the study of ZIP metabolism in human subjects will be facilitated by these data.