Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

5-(2-Ethyl-2H-tetrazol-5-yl)-1-methyl-1,2,3,6-tetrahydropyridine (Lu 25-109) is a muscarine agonist that has been under development for the treatment of senile dementia of Alzheimer's type. In binding assays (in vitro), Lu 25-109 acts as a partial M1 agonist and an M2/M3 antagonist with low affinities for other receptor types. The pharmacological profile indicates that Lu 25-109 may be expected to act selectively on muscarinic receptors without displaying the side effects normally associated with muscarinic compounds (Meier et al., 1997). Lu 25-109 constitutes a bioisosteric form of arecoline or methylguvacine ethyl ester, the 2-ethyl-2H-tetrazol-5-yl group substituting the ethyl ester (Moltzen et al., 1994). To our knowledge, Lu 25-109 is the first compound containing a 2,5-disubstituted tetrazole that has been investigated with regard to the mammalian metabolism.

Investigation of the biotransformation of Lu 25-109 in human liver microsomes showed that metabolite production was NADPH dependent, suggesting that the metabolism of Lu 25-109 in human liver microsomes is primarily P-450 dependent. The P-450s involved were identified, except for the step leading to Lu 35-026. This major metabolite in human urine was always small in human liver microsome incubations and showed very little variation (Jensen and Dalgaard, 1999).

This report describes the metabolite profile in plasma and urine, and the isolation and identification of metabolites from mice, rats, dogs, and humans after p.o. administration of [¹⁴C]Lu 25-109 as tartrate or maleate salt. The salts are expected to be bioequivalent due to the high water solubility. The disposition of total radioactivity showed that the majority of the dose, 89%, was excreted in urine and only 5% in feces of healthy male volunteers (D. Kenworthy, unpublished results). Furthermore, in vitro investigations were performed to elucidate a possible metabolic pathway leading to the formation of Lu 35-026.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Chemicals. Radioactive material was synthesized at Amersham International plc (Buckinghamshire, UK) with the ¹⁴C label located in the carbon atom of the tetrazole ring (see Scheme 1). The specific activities of Lu 25-109 salts used in these studies were tartrate: 68 µCi/mg and maleate: 71 µCi/mg. The radiochemical purity was >99% by HPLC. Lu 25-109, Lu 25-077 (5-(2-ethyl-2H-tetrazol-5-yl)-1,2,3,6-tetrahydropyridine), Lu 32-181 (5-(2-ethyl-2H-tetrazol-5-yl)-1-methyl-1,2,3,6-tetrahydropyridine 1-oxide), Lu 31-126 (5-(1H-tetrazol-5-yl)-1-methyl-1,2,3,6-tetrahydropyridine), Lu 29-297 (3-(2-ethyl-2H-tetrazol-5-yl)-1-methylpyridinium iodide), Lu 31-190 (5-(1H-tetrazol-5-yl)-1,2,3,6-tetrahydropyridine), Lu 31-102 (3-(1H-tetrazol-5-yl)-pyridine), Lu 35-026 (5-(1H-tetrazol-5-yl)-1-methyl-2-pyridone), Lu 35-038 (5-(2-ethyl-2H-tetrazol-5-yl)-1-methyl-2-pyridone), and Lu 35-080 (5-(2-ethyl-2H-tetrazol-5-yl)-1-methylpyridinium 3-oxide) were prepared at H. Lundbeck A/S. The synthesis of Lu 25-109, Lu 25-077, Lu 29-297, and Lu 31-102 has been described elsewhere (Moltzen et al., 1994). From 5-cyano-1-methyl-1,2,3,6-tetrahydropyridine (Schenker and Druey, 1959), Lu 31-126 was prepared in 45% yield (m.p. 209-11°C, hydrogen chloride salt) by treatment with refluxing azidotrimethylsilane. From 1-(ethoxycarbonyl)-5-(1H-tetrazol-5-yl)-1,2,3,6-tetrahydropyridine (Moltzen et al., 1994), Lu 31-190 was prepared in 58% yield (mp 165-68°C, hydrogen bromide salt) by treatment with concentrated hydrogen bromide in acetic acid at room temperature. Lu 32-181 was prepared in 40% yield (mp 149-51°C, hydrogen bromide salt) by treatment of Lu 25-109 with 3-chloroperoxybenzoic acid in dichloromethane at room temperature. From 3-cyano-1-methyl-6-oxo-1,6-dihydropyridine (Bradlow and Vanderwerf, 1951), Lu 35-026 was prepared in 77% yield (mp 251-53°C, hydrogen chloride salt) by treatment with refluxing azidotrimethylsilane. Lu 35-038 was prepared from Lu 35-026 in 32% yield (mp 95-97°C) by refluxing in acetone with bromoethane and potassium carbonate for 2 days. Lu 35-080 was prepared by treatment of 3-chloro-5-hydroxypyridine (Umemura et al., 1997) with methyl iodide. The resulting 3-cyano-1-methyl-5-hydroxypyridinium iodide (1 g) was treated with hydrazoic acid in toluene/ethanol for 3 days in a closed vessel at 130°C. After workup and chromatography, 0.2 g of the desired compound contaminated with 40% of the 2-methyltetrazole analog was obtained as an oil. General purpose reagents and solvents were purchased from commercial suppliers and were of the purest grade available.

View larger version (15K): [in this window] [in a new window]   Scheme 1.   Scheme of [14C]Lu 25-109 metabolism. *, Denotes the site of the radiolabel. { } = not seen in vivo. AO, aldehyde oxidase. , metabolizing enzymes identified in Jensen and Dalgaard, 1999. Broken arrow, route of metabolism not proofed.

Study Conduct. Dosing, sample collection, determination of total radioactivity, and profiling in plasma were performed at Covance Laboratories (Harrogate, North Yorkshire, U.K.). Metabolite isolation, profiling in urine, and identification were performed at H. Lundbeck A/S, Copenhagen, Denmark. Samples were shipped in dry ice and stored at <15°C.

Animals. Male and female B6C3F1 mice weighing 20 to 27 g (males) and 17 to 19 g (females) on the day of dosing, male and female Sprague-Dawley rats weighing 178 to 298 g (males) and 163 to 301 g (females) at the start of treatment, and male and female beagle dogs weighing 7 to 12 kg at the start of treatment were supplied by Charles River Ltd., Margate, UK. Animals were fed a standard diet with water ad libitum. The diet was removed the evening before dose administration and returned ca. 4 h postdose.

Dose Preparation and Administration. For the animal studies the oral doses were prepared as a solution in water to provide a nominal dose volume of 5 ml/kg body weight. The oral doses were administered by gavage. For the study in human volunteers the dose was prepared as a drinking solution containing 30 mg of base in 100 ml of sterile water.

Dosing and Sample Collection.

Mice Three male and three female mice received a single oral dose of [¹⁴C]Lu 25-109 maleate, 40 mg base/kg (200 µCi/kg b.wt.), and were placed in groups of three animals in all glass metabolism cages with collection of urine 0- to 24-h postdose.

Rats. Four male and four female rats were given an oral dose of [¹⁴C]Lu 25-109 tartrate, 40 mg base/kg body weight (75 µCi/kg b.wt./day) for 8 days. The rats were placed individually in all glass metabolism cages. Urine collected for 24 h before the last dose was used for metabolite profiling and identification. Blood collected from a lateral tail vein in heparin tubes at 0.5 and 6 h after a single dose (40 mg base/kg b.wt.) was used for plasma analysis.

Dogs. [¹⁴C]Lu 25-109 tartrate was administered orally by gavage to four male and four female dogs at a dose level of 1 mg of free base/kg/day for 8 days (2.7 µCi/kg b.wt./day). Urine collected for 24 h before the last dose was used for metabolite profiling and identification. Blood was sampled from singly dosed animals (1 mg base/kg) via the jugular vein in heparin tubes at 0.5 and 2 h postdose were used for plasma analysis.

Humans. Six healthy male Caucasian subjects (20-48 years old, 57-81 kg b.wt.) participated in the study. The study was conducted according to International codes of Good Clinical Practice and the protocol approved by the Administration of Radioactive Substance Advisory Committee of the Department of Health (U.K.), and all subjects gave written informed consent. Each fasted volunteer received a single oral dose of [¹⁴C]Lu 25-109 maleate (30 mg base) and 100 µCi as a drinking solution in 100 ml of sterile water. Urine collected 0 to 6 h and feces collected 0 to 24 h after administration of the dose was used for metabolite profiling and identification. Blood collected at 0.75-, 3-, 6-, and 10-h postdose and sampled into heparin tubes was used for plasma analysis.

Metabolite Patterns in Plasma, Urine, and Feces by HPLC.

Chromatographic methods Separation was achieved at an ambient temperature on a YMC-Pack ODS AQ, 5 µm, 250 × 4.6 mm column (YMC Co. Ltd., Kyoto, Japan). Isocratic elution was performed with solvent A (acetonitrile/0.2 M ammonium acetate pH 5, 3:97) for 85 min at a flow rate of 1 ml/min. Chromatography was finalized with a linear wash gradient to 100% solvent B (acetonitrile/0.2 M ammonium acetate pH 5, 30:70, v/v) at 95 min, kept at 100% B for 5 min, and returned to 100% A for re-equilibration of the column for a minimum of 10 min. (method 1). Cation-exchange chromatography was performed on a Dionex OmniPack PCX-100, 8.5 µm, 250 × 4 mm column (Diouex Comp. Sunnyvale, CA) eluted at ambient temperature with isocratic 50 mM hydrochloric acid and 50 mM potassium chloride in water/acetonitrile, 96:4 (v/v) at a flow rate of 1 ml/min (method 2). Before ion-exchange HPLC, the "front" was isolated by semipreparative HPLC on a YMC-basic, 5 µm, 250 × 10 mm column. The mobile phase consisted of acetonitrile/0.2 M ammonium acetate pH 5, 5:95 (v/v), delivered at 3 ml/min at an ambient temperature (method 3).

Metabolite profiles in urine. Aliquots of urine were centrifuged and the supernatant was analyzed by HPLC (method 1) with UV detection at 229 nm and on-line liquid scintillation counting (LSC)¹. The nonretained "front" was collected using method 3 and the relevant fractions were concentrated and analyzed by method 2. The dose proportion of any individual metabolite in a urine sample was determined as the percentage of total radioactivity recovered in the respective chromatographic peak:
For the metabolites eluting in the front, the dose proportion was calculated from a cation-exchange chromatogram by normalization with the front-peak % of the actual sample:

Metabolite profiles in feces. Weighed aliquots were added 5 ml of 1% ammonia/methanol solution followed by vigorous agitation and centrifugation. The supernatant was removed and the extraction procedure repeated was twice more. The resulting supernatants were pooled by sample and evaporated to dryness. The dry extracts were reconstituted in solvent A (method 1), ultrasonicated, and analyzed by HPLC (method 1) with on-line LSC.

Metabolite profiles in plasma. Aliquots were mixed with 3 volumes of acetonitrile, vortexed, and centrifuged. The supernatant was reduced to dryness using centrifugal evaporation. Samples were reconstituted in an appropriate volume of solvent A (method 1) and spiked with a nonradiolabeled reference standard mixture (Lu 25-109, Lu 25-077, Lu 32-181, and Lu 29-297). HPLC (method 1) was carried out with UV detection at 229 nm and on-line LSC (animals) or off-line LSC (fractions of 0.5 ml). The concentration of each metabolite represented in the chromatogram (microgram equivalents per milliliter) was determined from the percentage of total radioactivity recovered in the respective chromatographic peak and the radioactive concentration of the sample.

Glucuronidase and Basic Hydrolysis. Glucuronidase (10 µl) from Helix pomatia (type HP-2, 98,700 U/ml; Sigma) was added to 2 ml of urine adjusted to pH 5 and 37^o. Aliquots (100 µl) were injected on the liquid chromatography-mass spectrometry (LC/MS) using chromatographic method 1 at 0, 4, and 22 h of incubation. Basic hydrolysis was carried out at approximately pH 11 and 37^o, and aliquots were analyzed as above.

Chromatography for Isolation.

Hydrophobic chromatography A glass column (XK 26/40 26 × 400 mm; Pharmacia) was prepared with Serdolite AD-2 (Serva, Heidelberg, Germany) suspended in water, the inlet tube connected to a peristaltic pump (Pump P-1; Pharmacia), and the column pre-equilibrated with water 3 ml/min for 24 h. Filtered urine (400 ml) was applied on the column. The column was washed with 300 ml of water followed by a gradient of acetonitrile from 0 to 100% at a flow rate at 3 ml/min.

Cation exchange chromatography. A glass column (Pharmacia XK 16/20, 16 × 200 mm) was prepared with a ca. 50-ml cation exchange resin (Dowex 50Wx2; Aldrich, Steinheim, Germany) in 0.1 M HCl and plugged with glass wool. Sample application and elution were performed manually. The sample was acidified with HCl before application. After application the column was washed with water until it had neutral pH and then with an additional 300 ml of water. Elution was performed with 2 M ammonia until all radioactivity was recovered.

Semi preparative HPLC (method 3). Semipreparative purification was performed on a YMC-basic, 5 µm, 250 × 10 mm column. The mobile phase consisted of acetonitrile/0.2 M ammonium acetate pH 5, 5:95 (for m/z 178) or 3:97 (for m/z 206 and m/z 398) delivered at 3 ml/min at ambient temperature for 50 min.

Analytical HPLC. The final purification was achieved with HPLC method 1.

Fraction collection and selection. For each chromatographic step, subsamples of collected fractions were analyzed by LSC and fractions of interest were selected by recovery of radioactivity followed by LC/MS. Fractions from the same peak were pooled and concentrated before the next purification step.

Radio Analysis. A Flo-One A-500 or 500RT beta counter equipped with a 500 µl flow cell (Packard Instruments, Meriden, CT) was used for on line HPLC detection. Ultima-Flo M (Packard) 3 ml/min was used as scintillation liquid. For off line detection, subsamples of chromatographic fractions were assayed directly in 3 ml of Ultima-Flo M (Packard). Radioactivity was measured using a Packard Tri-Carb liquid scintillation counter (Canberra Packard, Meriden, CT) and corrected for quenching with automatic external standardization.

LC/MS and Liquid Chromatography/Tandem Mass Spectrometry. A VG Quattro II mass spectrometer (Micromass, Altrincham, UK) equipped with electrospray and an 1050 HP HPLC system were used.

Loop injections. Before injection, 500 µl of each sample was diluted with 500 µl acetonitrile/water/acetic acid (500:500:2, v/v/v). The MS was operated in positive ion mode (ESP+) mode with a scan range of 30 to 600 m/z and a cone voltage of approx. 20 eV. Product ion spectra were obtained at a collision energy of approx. 20 eV and an argon gas pressure of 3.6 × 10³ mbar.

Deuterium exchange experiments. A modified HPLC method 1 with isocratic elution and a run time of 20 min was used. Two runs were compared, first with the mobile phase, consisting of 200 mM ammonium acetate pH 5.5/acetonitrile (90:10, v/v), and second with deuterated solvents: D₂O (99.97 atom % D) from Cambridge Isotope Lab., Andover, MA; 26% ND₃ in D₂O (99 atom % D), and CH₃COOD (98 atom % D) from Aldrich, Milwaukee, WI. The pH was adjusted disregarding the deuterium effect on the pH measurement. The HPLC column was allowed to equilibrate for 10 min with the deuterated mobile phase.

Isolation of Metabolites. "m/z 206" (Lu 35-080) was isolated from 400 ml of pooled dog urine collected during the last dosing interval using Serdolite AD-2, semiprep HPLC, and analytical HPLC. The isolated sample was dissolved in methyl-d₃-alcohol-d (99.8 atom % D) for NMR. "m/z 178" (Lu 35-026) was isolated from 400 ml of pooled human urine collected 0- to 6-h postdose using Serdolite AD-2, cation exchange, semiprep. HPLC, and analytical HPLC. The sample was dissolved in methyl-d₃-alcohol-d (99.8 atom % D) for NMR. m/z 398 was isolated from 100 ml of human urine collected 0- to 6-h postdose. Urine was vacuum concentrated and redissolved in 10 ml of mobile phase for semiprep. HPLC (method 3), followed by analytical HPLC (method 1). The isolated sample was dissolved in deuterium oxide (99.96 atom % D). m/z 400 was isolated from 400 ml of human urine adjusted to pH 3 with 0.1 M HCl, freeze-dried, and redissolved in approx. 20 of ml mobile phase (method 3). The sample was injected in 1-ml portions to the HPLC. The pooled fractions of interest were concentrated and 1-ml portions were injected in the analytical HPLC system (method 1). Finally, the pooled fractions from method 1 were injected in an HPLC system with a mobile phase consisting of 20 mM trifluoracetic acid, pH 1.78:acetonitrile (95:5, v/v). The pooled fractions were freeze-dried and treated with D₂O repeatedly before transfer to the NMR tube with 2 × 25 µl D₂O.

NMR Spectroscopy. "m/z 178" (Lu 35-026) and "m/z 206" (Lu 35-080). NMR spectra were recorded on a Bruker AC-250 NMR spectrometer operating at 250.13 MHz for ¹H and at 62.896 MHz for ¹³C. Methyl-d₃-alcohol-d (99.8 atom % D) was used as solvent and the methyl-d₂ residual signal was used as internal reference (3.35 ppm for ¹H and 47.07 ppm for ¹³C). A volume of 100 µl in a 2.5 mm NMR tube was measured in 5 mm probeheads (normal geometry for 1D spectra and inverse geometry for 2D CH-correlated spectra).

m/z 398. A volume of 100 µ in a 2.5 mm NMR tube was measured without spinning in 5 mm probehead with inverse geometry. Solvent suppression with irradiation of HOD was used. The spectra were run at 320 K to avoid signal overlapping between HOD and compound.

Incubations with Human Liver Microsomes/S9. Human liver microsomes (Hepatoscreen Test Kit Ver 3.0 and Pooled Hepatosomes) and pooled human liver S-9 preparations were obtained from Xenotech LLC (Cambridge, KS), shipped in dry ice, and samples were stored in a freezer at 80°C until use. Microsomal preparations from transfected human AHH-1 TK ± B-lymphoblastoid cell lines expressing human CYP1A2, CYP2A6, CYP2C9, CYP2C19, CYP2D6, CYP2E1, CYP3A4, as well as control microsomal preparations, were obtained from Gentest Corp. (Woburn, MA), shipped in dry ice, and samples were stored in a freezer at 80°C until use. Quinidine, tranylcypromine, coumarin, diethyldithiocarbamate (DDC), quercetin, allopurinol, and menadione were obtained from Sigma (St. Louis, MO); ketoconazole, 99% purity, from Oxford Biomedical Research Inc., (Oxford, MI); and sulfaphenazole and furafylline from Research Biochemicals International (Natick, MA).

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Isolation and Identification of Urinary Metabolites. The molecular ion, the product ion spectra, and the HPLC retention times of identified metabolites in human urine samples and synthetic standards showed the same pattern (Table 1). Useful product ion spectra could not be recorded for Lu 25-109, Lu 25-077, and Lu 31-190 in human urine because of the small amount present. The metabolites m/z 398 and m/z 400 were not synthesized, thus their proposed structures rely solely on the spectroscopic data and the enzymatic cleavage. ¹H data for isolated metabolites are shown in Table 2 with the structures shown in Scheme 1 or Fig. 1.

View larger version (17K): [in this window] [in a new window]   Fig. 1.   Glucuronic acid-conjugated metabolites of Lu 25-109.

m/z 178 (Lu 35-026). Material (0.3 mg) was isolated from human urine by Serdolite AD-2, cation-exchange purification, and semipreparative and analytical chromatography in 16% yield of the theoretical. The structure of the isolated material was identified by NMR spectroscopy in methyl-d₃-alcohol-d in combination with the molecular weight. The ¹H-NMR data (Table 2) suggested a deethylated tetrazole pyridine with an oxygen in the pyridine ring. The coupling pattern suggested an oxygen substitution at position 4 or 6 in the pyridine ring. Position 4 was ruled out by a difference nuclear Overhauser enhancement (NOE) experiment with saturation on the methyl singlet. NOE was only observed at the doublet at 8.36 ppm with the small long-range coupling constant (3 Hz), proving that no hydrogen is present at position 6. This structure was later verified by comparison with the reference standard Lu 35-026 (Table 1).

m/z 206 (Lu 35-080). This compound was isolated from dog urine by XAD, semipreparative, and analytical chromatography in a yield of 50% of the theoretical. The structure of the isolated material was identified by NMR spectroscopy in methyl-d₃-alcohol-d. The ¹H- (Table 2) and ¹³C-NMR data suggested an ethyl tetrazole (¹H signals at 1.75 and 4.80 ppm and ¹³C signals at 12.8 and ca. 47 ppm) attached to a methyl pyrimidinium ring because the ¹H signal from the methyl group was at 4.25 ppm. A 3,5-substitution is the only substitution pattern that will give rise to three singlets for a pyrimidine compound. Furthermore, the assignment was based on a NOE difference experiment (irradiation on the pyrimidinium methyl group) and a long-range CH-correlation (HMBC-experiment). The suggested structure was later verified by comparison with the reference standard Lu 35-080 (Table 1).

m/z 398. Sixty micrograms of material was isolated from human urine by semipreparative and analytical chromatography in a yield of 60% of the theoretical. The MS data, including a deuterium shift of +5, and disappearance of m/z 398 by glucuronidase hydrolysis (data not shown) suggested a monoglucuronide of a dihydroxylated pyridine. This was in good accordance with the ¹H spectrum in deuterium oxide. The coupling pattern in the aromatic region suggested two protons sharing a meta or para coupling. A NOE difference experiment with irradiation on the methyl singlet at 4.30 ppm gave enhancement of the signal at 8.75 ppm, proving that either position 2 or 6 has been substituted. Enhancement was also seen at the signal at 5.75 ppm (assigned to the anomeric glucuronide proton), but this may be due to coirradiation of the multiplet at 4.15 to 4.35 ppm. Irradiation at the signal at 5.75 ppm gave enhancement of the signal at 8.35 ppm. The only substitution pattern that will lead to these observations is a 4,6-dihydroxylation (Fig. 1). The complete lack of enhancement of the signal at 4.30 ppm on irradiation of the signal at 5.75 ppm indicated that the glucuronide may be at position 4, but because of internal relaxation in the methyl group this can not be considered a proof.

m/z 400. This compound was isolated from human urine by semipreparative and analytical chromatography in a yield of approx. 30% of the theoretical. The MS data, including a deuterium shift of +5 and disappearance of m/z 400 by glucuronidase hydrolysis (data not shown) suggested a monoglucuronide. The product ion spectrum of m/z 400 showed m/z (I%) 206 (67), 178 (26), 165 (29), 159 (51), 122 (100), and 79 (39). Prolonged standing showed the formation of four chromatographically different compounds, all with a molecular ion m/z 400, indicating that rearrangement/isomerization took place. The product ion spectra of two of these compounds showed m/z 206 (base peak) and m/z 224 (30-70%) and the other two exclusively m/z 224 (base peak). Both glucuronidase and basic hydrolysis resulted in an increase of the m/z 206 (Lu 35-080) chromatographic peak. The ¹H spectrum showed that the N-methyl group and the ethyl tetrazole group were intact. Three signals from the pyridine ring were found: A methine signal at 7.00 ppm (singlet) and a methylene signal (AB system at 4.45 and 4.55 ppm, J = 18 Hz), thus indicating a dihydropyridine structure. More signals for a glucuronic acid derivative were found. Long-range correlations are observed from the methine protons and the N-methyl protons to the same quarternary carbons at approx. 167 ppm and to the carbons from the methylene groups. On prolonged standing (approx. 2 months) at room temperature a partial rearrangement was observed. The NMR solution (in D₂O) after about half a year showed the same rearrangement pattern but with molecular ions of m/z 401 on LC/MS. 2D gradient heteronucleous single quantum coherence with echo/antiecho coherence selection and 2D gradient HMBC experiments were performed on the this sample, showing that the Lu 25-109 part of the molecule was unchanged, indicating that the rearrangements occurred in the glucuronic acid part of the molecule. These observations, together with the fact that the compound on treatment with base could be transformed into the known 5-hydroxy metabolite with m/z 206 (Lu 35-080), points at 5,6-dihydroxlated or 2,5-dihydroxylated structures (Fig. 1).

Metabolite Pattern in Urine and Feces. The reversed phase systems proved useful in separating most of the metabolites and with a complete recovery of radioactivity. The metabolites, which showed no retention on these stationary phases in spite of the low content of acetonitrile (3%), were separated on a cation exchange column.

View larger version (13K): [in this window] [in a new window]   Fig. 2.   Representative radiochromatograms of 0- to 6-h urine sample from human male volunteer (subject 2) after an oral dose of [14C]Lu 25-109 maleate (100 µCi, 30 mg, calculated as base). A, total sample (method 1); B, "front" (method 2).

Metabolite Pattern in Plasma. Chromatographic profiling of metabolites in plasma of single orally dosed subjects was performed in humans (0.75, 3, 6, and 10 h postdose), rats (0.5 and 6 h postdose), and dogs (0.5 and 2 h postdose). Reconstituted plasma extracts injected on the HPLC generally contained in excess of 80% of the radioactivity measured in the original sample. The concentration of major circulating metabolites was calculated by use of the specific activity of the dose. Several chromatographic peaks were tentatively identified by their retention time: Lu 25-109, Lu 25-077, Lu 32-181, and Lu 35-026. Furthermore, Lu 31-126 eluted with the chromatographic front and accounted for most of the radioactivity of the front. The plasma concentrations at the examined time points of Lu 25-109 and selected metabolites from human volunteers receiving 30 mg (calculated as base) are shown in Fig. 3. The maximal plasma concentration of Lu 25-109 was observed in the 0.75- or 3-h samples with concentrations of 14 to 51 ng eq/ml. At 10 h postdose the Lu 25-109 level decreased to 3 to 10 ng eq/ml. The concentration of Lu 25-077 was comparable to that of Lu 25-109 at the time points examined. In general, Lu 32-181 and Lu 35-026 were the major components in plasma at 0.75 h with concentrations of 43 to 161 and 44 to 112 ng eq/ml, respectively. This was followed by a fast decrease in concentrations at 3 and 6 h to below the detection limit at 10 h. For most subjects, C_max for Lu 31-126 was observed in the 3-h sample with concentrations of 25 to 124 ng eq/ml. In the later samples, Lu 31-126 was still the major component in plasma. Comparable metabolite concentration-versus-time patterns were seen in rats and dogs administered with 40 and 1 mg base/kg, respectively. In rats, the major metabolites at 0.5 h postdose were Lu 32-181 and Lu 31-126, the latter being the major component at 6 h postdose as well. In dogs, Lu 32-181 was the major component in plasma at 0.5 and 2 h postdose whereas Lu 31-126 was present in smaller concentrations than in humans. The plasma concentrations of Lu 25-109 and Lu 25-077 were lower in rats and dogs than in humans. Lu 35-026 was present in much smaller amounts relative to Lu 25-109 in rats than in humans and was not detected in dog plasma at all. (Data for the first time point are given in Table 4).

View larger version (24K): [in this window] [in a new window]   Fig. 3.   Time course of plasma concentration of Lu 25-109 and selected metabolites from human volunteers (subjects 1-6) after an oral dose of [14C]Lu 25-109 maleate (100 µCi, 30 mg, calculated as base).

Incubation with Human Liver Microsomes/S9. The identification of P-450s involved in the metabolism of Lu 25-109 was previously reported in (Jensen and Dalgaard, 1999) except for the step from Lu 38-038 to Lu 35-026 presented herein (Scheme 1). To elucidate possible metabolic pathway leading to the formation of Lu 35-026, several metabolites and possible intermediates were incubated in S-9 and microsomes. Formation of Lu 29-297 from Lu 25-109 was NADPH dependent and occurred mainly in microsomes as would be expected for a P-450 reaction, whereas formation of Lu 35-038 from Lu 29-297 was not NADPH dependent and occurred mainly in S-9. Thus, this reaction is not catalyzed by P-450 enzymes but most probably by cytosolic enzyme(s). Lu 35-038 was efficiently converted to Lu 35-026 in microsomes and was dependent on NADPH (data not shown). The reaction Lu 29-297 to Lu 35-038 was inhibited by menadione and quercetin, which are selective aldehyde oxidase inhibitors and not by allopurinol, a selective xanthine oxidase inhibitor (Sugihara et al., 1996; Rashidi et al., 1997), suggesting that aldehyde oxidase is responsible for the conversion (Table 5).

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

Metabolite Identification. Lu 25-109 is the first compound containing a 2,5-disubstituted tetrazole, which to our knowledge has been investigated with regard to the mammalian metabolism. Other tetrazoles that have been investigated appear to be metabolically stable. In literature, there are no reports on biotransformation with oxidation or ring cleavage of tetrazoles, except in tetrazolium salts (Zeevalk et al., 1995). 5-aryl-tetrazoles, which are bioisosteres of aromatic carboxylic acids, have been shown to undergo N-glucuronidation (Huskey et al., 1994), but this was not the case with Lu 25-109. The N-methyl-tetrahydropyridine group is prone to oxidative metabolism as could be expected from literature data of compounds containing similar groups. The deethylation of Lu 25-109 is assumed to proceed through a hydroxylation of the nitrogen-bearing carbon followed by loss of ethanol. N-methyl-tetrahydropyridine substituted in the 4 position (MPTP) is known to be metabolized to the pyridinium compound (MPP⁺) (Zuddas et al., 1994). Likewise, Lu 25-109 was oxidized to pyridinium compounds although Lu 25-109 is substituted in the 5 position. Lu 25-109 resembles more the muscarinic agonist xanomeline (Shipley et al., 1995) in that the two compounds both contain a N-methyl-tetrahydropyridine substituted with a heterocyclic ring in the 5 position. Xanomeline is metabolized by N-demethylation, N-oxidation ,and side chain oxidation (Shipley et al., 1995). Indeed, as shown in Scheme 1 the metabolites of Lu 25-109 are results of the same type of processes. Furthermore, Lu 25-109 is oxidized to mono-hydroxy-pyridinium (Lu 35-080) or pyridone (Lu 35-026) compounds, or glucuronides of a dihydroxy-pyridine (m/z 398). Also, a glucuronide of a dihydroxylated dihydro-pyridine compound (m/z 400) was isolated. However, the exact structure of m/z 400, present in only small amounts and relatively unstable, remains to be identified. Both glucuronides showed a m/z shift of +5 to m/z 403 and m/z 405, respectively, in accordance with the number of exchangeable hydrogens in the deuterium exchange experiment. In solution, m/z 400 rearranged to four chromatographically different compounds, all with m/z 400. Thus, it appears that the m/z 400 metabolite can rearrange to other m/z 400 isomers, which can loose glucuronic acid (194 amu) or at glucuronyl group (176 amu) by fragmentation in the mass spectrometer to the product ion m/z 206. Also, enzymatic or basic hydrolysis with subsequent loss of water lead to m/z 206 (Lu 35-080). After about half a year, the NMR solution containing D₂O rearranged in the same pattern (but showing m/z 401), indicative of a slowly exchangeable hydrogen in these molecules. The isomers remain to be identified, but are possibly due to isomerization of the glucuronic acid moiety and/or migration of the aglycone.

Metabolite Profile. The human urinary [¹⁴C]Lu 25-109 metabolite profiles reported are based on 67% of the radioactive dose collected 0 to 6 h postdose. By using 0- to 24-h urine samples the results could have been based on about 85% of the radioactive dose but at the same time the volume of the samples would have been doubled with lower absolute concentrations and difficulties in detection of the minor metabolites as a consequence. Except for m/z 400, all of the metabolites identified in human urine could be found in at least one of the three animal species investigated (Table 3).

Incubation with Human Liver Microsomes/S-9. In Jensen and Dalgaard (1999) the identification of human hepatic cytochromes P-450 enzymes involved in the in vitro metabolism of [¹⁴C]Lu 25-109 were investigated, but one major metabolite of Lu 25-109 (Lu 35-026) was not formed in microsomes and thus no enzymes could be assigned in that in vitro system (summarized in Scheme 1). To elucidate possible metabolic pathways leading to the formation of Lu 35-026 several metabolites and possible intermediates were incubated in S-9 and microsomes. The formation of the intermediate metabolite Lu 35-038 (not identified in vivo) from Lu 29-297 are dependent on human liver S-9 cytosol and independent of the presence of NADPH, suggesting involvement of a cytosolic enzyme. The reaction Lu 29-297 to Lu 35-038 was inhibited by selective aldehyde oxidase inhibitors, i.e., menadione and quercetin, and not by allopurinol, a selective xanthine oxidase inhibitor. Thus, this evidence suggests that aldehyde oxidase catalyzes the reaction Lu 29-297 to the intermediate metabolite Lu 35-038. The intermediate Lu 35-038 has not been identified in vivo as a metabolite of Lu 25-109 in any species, possibly because of rapid conversion to Lu 35-026. The identification of P-450s involved in the metabolism of Lu 35-038 to Lu 35-026 through studies using: 1) simple regression analysis with phenotyped human liver samples, 2) selective chemical inhibitors, and 3) microsomes containing cDNA-expressed enzymes, all indicated CYP1A2 as being involved in the metabolic step. The correlation study also indicated involvement of CYP3A4 but the phenotyped probes for CYP1A2 and CYP3A4 also correlated among each other. Because the inhibition studies indicated no involvement of CYP3A4 and no formation was seen in microsomes containing only c-DNA-expressed CYP3A4, the involvement of CYP3A4 seems unlikely. Inhibition studies also indicated some involvement of CYP2A6 but the small inhibitory response of coumarin inclusion was not confirmed by correlation studies nor in microsomes only containing c-DNA-expressed CYP2A6. Similarly, the formation of Lu 35-026 seen in only c-DNA-expressed CYP2C19 could not be confirmed by correlation or inhibition studies. Thus, the in vitro transformation of Lu 35-038 to Lu 35-026 was mainly catalyzed by CYP1A2 (see Scheme 1).