QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Data Supplement All Versions of this Article: dmd.107.019463v1 36/4/731    most recent Submit a response Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Xu, L. Articles by Harsch, A. Search for Related Content PubMed PubMed Citation Articles by Xu, L. Articles by Harsch, A. Pubmed/NCBI databases Compound via MeSH Substance via MeSH Hazardous Substances DB COLCHICINE

Identification of Novel Metabolites of Colchicine in Rat Bile Facilitated by Enhanced Online Radiometric Detection

Drug Metabolism and Pharmacokinetic Department, Merck Research Laboratory at Boston, Boston, Massachusetts (L.X., B.A., V.V.J.-M., G.K., A.H.); and Merck Frosst Canada Ltd., Kirkland, Quebec, Canada (L.T.)

(Received October 26, 2007; Accepted January 23, 2008)

	Abstract

Top Abstract Materials and Methods Results Discussion References

 
Three novel conjugation metabolites of colchicine were identified in rat bile facilitated by enhanced on-line liquid chromatography-accurate radioisotope counting. The known 2- and 3-demethylcolchicines (DMCs) underwent O-sulfate conjugation in addition to the previously described O-glucuronidation. 2-DMC was preferably O-glucuronidated, whereas 3-DMC predominantly yielded O-sulfation conjugates, indicating phase II conjugation regiopreferences. Moreover, M1 was identified as a novel glutathione conjugate and a possible biotransformation pathway for its formation was proposed. The known 2-DMC (M6), 3-DMC (M7), 2-DMC glucuronide (M4), and novel 3-DMC sulfate (M3) were confirmed as the major metabolites. Radiometric data were acquired by the XFlow liquid chromatography-accurate radioisotope counting (XFlow LC-ARC) system, a novel technology for dynamic control of both on-column and postcolumn high-performance liquid chromatography flow rates to maximize sensitivity and resolution of radiochromatograms. A comparative evaluation was also performed between the XFlow LC-ARC system and a conventional flow radiometric detection system using bile samples from an in vivo disposition study of colchicine in male Sprague-Dawley rats. Results demonstrated a 20-fold sensitivity improvement of the XFlow LC-ARC system in comparison with radioactivity detection by conventional flow scintillation analyzers. The dynamic flow mode also provided the best chromatographic resolution. Unambiguous metabolite identification was performed by high-resolution mass spectrometry and nuclear magnetic resonance analysis.

Hyphenation of high-performance liquid chromatography (HPLC) to radiometric detection has provided a powerful approach for in vitro and in vivo metabolite identification, specifically in the analysis of complex biological samples such as bile and plasma. On-line acquisition of ³Hor ¹⁴C radiochromatograms readily detects drug-related materials against the backdrop of massive matrix interferences. However, conventional FSA for on-line radioactivity measurement is relatively insensitive (Onisko, 2002; Nassar et al., 2003), especially in comparison with mass spectrometric analysis in which femtomole or attomole detection limits are routinely achievable (Göbel et al., 2004; Peterman et al., 2006). Therefore, comprehensive radiometric on-line detection of drug metabolism profiles can be difficult, especially when radiotracers with low specific activities are used or only limited quantities of radiolabeled samples are available. Identification of radiolabeled metabolites from human plasma or urine samples are particularly challenging as only low doses of radioactivity are generally administered because of regulatory restrictions and safety considerations. Hence, highly sensitive on-line radiometric detection is critical for rapid and effective metabolite profiling, especially for in vivo samples. Recent off-line microplate scintillation counting methodology has yielded improved sensitivity and enhanced sampling throughput. It has been demonstrated that microplate scintillation counting achieved accuracy and sensitivity comparable with that of traditional off-line liquid scintillation counting but offered higher throughput for ADME studies (Nassar et al., 2004; Bruin et al., 2006). This approach, however, is limited to nonvolatile compounds because samples are required to be dried down before radiometric counting.

The XFlow LC-ARC system is a novel analytical device for enhanced on-line radiometric detection (Nassar et al., 2003; Nassar and Lee, 2007). It controls the scintillation liquid flow as well as oncolumn and postcolumn LC flows. It has been demonstrated that dynamic liquid flow optimization, in conjunction with an improved flow cell design and a high-efficiency liquid scintillation cocktail, enhanced radiometric detection sensitivity and resolution without prolonging LC run times (Nassar and Lee, 2007). Integration of the XFlow LC-ARC system with mass spectrometric detection allows parallel acquisition of high-sensitivity radiometric and MS data and thus facilitates direct and effective metabolite identification in ADME studies.

A key objective of this study was to evaluate the XFlow LC-ARC system versus a conventional flow radiometric detection system using bile samples from an in vivo disposition study of colchicine in male Sprague-Dawley rats and to determine a comprehensive metabolic profile in this matrix. Colchicine is a highly soluble alkaloid, originally extracted from plants of the genus Colchicum. It has been used in gout and familial Mediterranean fever treatment (Wallace, 1974; Cerquaglia et al., 2005) and is also well known as an antimitotic agent by inhibiting microtubule polymerization through binding to tubulin (Morejohn and Fosket, 1991). Radiolabeled colchicine was used in the first preparations of purified tubulin (Correia, 1991). Its high toxicity, however, limits its clinical use. In previous drug metabolism studies in rats, colchicine has been shown to undergo extensive hepatic metabolism, and parent and metabolites are eliminated mainly by biliary excretion (Hunter and Klaassen, 1975; Sabouraud et al., 1992). Historic metabolic profiling of colchicine in rat bile was performed with low-sensitivity scintillation counting and ultraviolet spectroscopy; thus, re-elucidation of colchicine metabolism in rat bile was chosen as an excellent test case for high-sensitivity on-line radiometric detection.

	Materials and Methods

Top Abstract Materials and Methods Results Discussion References

 
Materials. [10-methoxy-³H]Colchicine was purchased from PerkinElmer Life and Analytical Sciences (Boston, MA) with a specific activity of 80.4 Ci/mmol. 2-DMC was a gift from Dr. K. H. Lee (Natural Products Research Laboratories, University of North Carolina, Chapel Hill, NC). 1-DMC was synthesized by regioselective demethylation of colchicine (Blade-Font, 1979), and 3-DMC was synthesized from colchicoside by treatment with 85% phosphoric acid (Rösner et al., 1981). All other chemicals were HPLC or analytical grade and were obtained from Fisher Scientific Co. (Suwanee, GA), unless specified otherwise.

Animal Studies and Sample Collections. All studies were reviewed and approved by the Merck Research Laboratories Institutional Animal Care and Use Committee. A dose of 50 µCi/kg (0.25 mg/kg) [10-methoxy-³H]colchicine was administered i.v. to three bile duct-cannulated male Sprague-Dawley rats purchased from Taconic Farms (Hudson, NY). Throughout the duration of the study, the animals were housed in metabolic cages (Nalgene, Pittsburgh, PA), and an artificial bile solution (100 mM taurocholic acid salt in 0.9% saline-0.5% KCl) was infused at 0.5 ml/h into the duodenum. Bile was collected between 0 and 2, 2 and 4, 4 and 6, 6 and 8, 8 and 10, 10 and 24, and 24 and 48 h over dry ice. Rats were euthanized with CO₂ after completion of the studies. The bile samples obtained were stored frozen at–80°C until analysis. The samples were filtered through 0.45-µm syringe filters (Millipore Corporation, Billerica, MA) before injection into the LC system.

Instrumentation. The XFlow LC-ARC–MS system consisted of two LC-10ADvp pumps (Shimadzu, Kyoto, Japan), an autosampler (CTC PAL; CYC Analytics, Swingen, Switzerland), an XFlow unit (AIM Research Co., Hockessin, DE), a 600TR Radiomatic detector (PerkinElmer, Meriden, CT), and a LTQ Orbitrap high resolution mass spectrometer (Thermo Fisher Scientific, San Jose, CA). The XFlow unit software controlled the entire system, including the high-performance liquid chromatograph, radioactivity detector, XFlow unit, and mass spectrometer.

For all chromatographic separations, a Luna analytical C₁₈(2) column (4.6 x 250 mm, 5 µm particle size; Phenomenex, Torrance, CA) was used. Mobile phases were 0.1% formic acid in water (solvent A) and 0.1% formic acid in acetonitrile (solvent B). The column was held at ambient temperature and the samples were eluted at a flow rate of 0.80 ml/min with a linear gradient from 5% to 30% solvent B over 50 min. The HPLC effluent was split into the radiometric detector fitted with a 1-ml radiochemical liquid flow cell (ARC flow cell; AIM Research Co.) and the mass spectrometer at a fixed 3:1 ratio controlled by the XFlow device. StopFlow AD cocktail (AIM Research Co.) was used as the scintillation liquid. The ARC instrument software provided full control over the radioactivity detector and the LC system. For comparative analysis, a 600TR Radiomatic detector with a built-in 0.5-ml radiochemical liquid cell and Ultima-Flo cocktail (PerkinElmer) was used as a conventional LC-FSA system. The cocktail flow rate during operation was 2.4 ml/min to maintain a 3:1 ratio of the cocktail to LC eluent. The LTQ Orbitrap mass spectrometer was operated in positive electrospray ionization mode. The heated capillary temperature was maintained at 350°C; the sheath gas and auxiliary gas flow rates were set to 60 and 15 units, respectively. The ion spray voltage, capillary voltage, and tube lens offset voltage were adjusted to 2, 9, and 85 V, respectively. The normalized collision energy was 30% during MS/MS acquisition and helium was used as the collision gas. High-resolution mass measurement was performed in Orbitrap mode with a resolution of 100,000. High-resolution data were expressed as five-decimal atomic mass units and mass difference (M) between measured mass and the calculated mass was expressed in parts per million units. Elemental compositions were calculated on the basis of high-resolution mass data and McLafferty's nitrogen rule.

For NMR analysis, the isolated metabolites were dissolved in hexadeuterodimethyl sulfoxide or deuterium oxide, and spectra were recorded with a Varian Inova 600 spectrometer with a cryogenic probe. A Shigemi advanced 3-mm NMR microtube (Sigma-Aldrich, St. Louis, MO) was used for M1 during NMR data acquisition. Chemical shifts () are expressed as parts per million downfield relative to tetramethylsilane, and coupling constants (J) are expressed in Hertz.

The preparative HPLC system contained two LC-10ADvp pumps, an SIL-10ADvp auto injector, and a SPD-M10Avp diode array detector (Shimadzu). The mobile phase was 2 mM ammonium acetate in acetonitrile-water (9:1) and 2 mM ammonium acetate in water, delivered at 5 ml/min. Separations were performed using a Luna semipreparative C₁₈(2) column (10 x 250 mm, 5-µm particle size, Phenomenex). The LC gradient conditions were individualized for different metabolites to achieve maximum isolation purity. UV chromatograms at 260 and 306 nm were used to guide manual fraction collections. UV spectra of metabolites were recorded with a SPD-M10Avp detector (Shimadzu).

Primary Hepatocyte Isolation and Incubations. Rat hepatocytes were isolated from male Sprague-Dawley rats according to the two-step collagenase perfusion method (Seglen, 1976). The liver was perfused via a catheter inserted into the inferior vena cava. The sequential retrograde perfusion was performed with two consecutive buffers: calcium- and magnesium-free salt solution (140 mM NaCl, 6 mM KCl, 10 mM HEPES, pH 7.4), followed by the same buffer with the addition of 4 mM CaCl₂ and 0.025 mg/ml Liberase Blendzyme 3 (Roche Applied Science, Indianapolis, IN). After digestion, hepatocytes were released with gentle shaking of the lobes, filtered through a 100 µm mesh and washed twice with Williams' medium E (50g for 5 min). Hepatocytes were purified from nonparenchymal cells through centrifugation with 90% isotonic Percoll (100g for 15 min). The resulting pellet was resuspended in Williams' medium E, and cell viability was determined by trypan blue exclusion. Hepatocyte preparations with viabilities greater than 90% were used for experiments.

Incubations were performed by suspending the rat hepatocytes in InVitro-GRO hepatocyte Media (In Vitro Technologies, Baltimore, MD) followed by the addition of aqueous solution of colchicine, 1-DMC, 2-DMC, or 3-DMC. The final substrate concentration in cell suspension was 10 µM in a volume of 1 ml at a cell density of 2 x 10⁶ cells/ml. Incubations proceeded for 2 h at 37°C in 20-ml glass vials under 95% O₂-5% CO₂.

Isolation and Identification of the Novel Metabolites. Preparative incubation of colchicine (50 µM) in the presence of 5 mM GSH was performed in the rat primary hepatocytes as described above, using a total of 200 incubation vials. After the reaction was quenched with trifluoroacetic acid, incubation mixtures from each vial were pooled and neutralized with 1 M ammonia to pH 6.5. The neutralized mixtures were sonicated for 5 min and centrifuged at 3000g for 10 min. The supernatants were lyophilized overnight, and the residues were dissolved in 10 ml of water. Each 5-ml aliquot was loaded on a Sep-Pak C₁₈ cartridge (10 g; Waters, Milford, MA) and subsequently eluted with 15 ml each of 0, 5, 10, 15, and 20% acetonitrile in water. Fractions containing glucuronide conjugates, sulfate conjugates, and the GSH adduct were pooled, lyophilized under vacuum, and subsequently reconstituted in 0.80 ml of water, respectively. The crude extract of each metabolite was repurified on the preparative HPLC system described above.

Analytical Data of M1, M2, and M3. 3-Demethylcolchicine GSH adduct (M1): UV_max (acetonitrile-water, 1:1), 260 and 380 nm. HRESI-MS [M + H]⁺: m/z 675.23332 (C₃₁H₃₉N₄O₁₁S, M 0.39 ppm). ¹H NMR (D₂O, , 600 MHz): 7.57 (d, 1H, 10.4 Hz, H-12), 7.32 (d, 1H, 10.4 Hz, H-11), 7.16 (s, 1H, H-8), 6.74 (s,1H, H-4), 4.62 (m, 1H, Hb), 4.31 (m, 1H, H-7), 3.78 (s, 3H, OCH₃-3), 3.73 (s, 3H, OCH₃-2), 3.60 (m, 2H, Ha), 3.52 (m, 1H, Hd), 3.51 (m, 1H, Hc), 3.48 (s, 3H, OCH₃-10), 3.25 (m, 1H, Hc'), 2.52 (m, 1H, H-5), 2.32 (m, 2H, Hf and Hf'), 2.22 (m, 1H, H-5'), 2.12 (m, 1H, H-6), 1.92 (m, 2H, He and He'), 1.88 (s, 3H, H₃COC), 1.82 (m, 1H, H-6').

View larger version (22K): [in this window] [in a new window]   FIG. 1. LC radiochromatograms acquired with a conventional radiometric detector (I and II) and with the XFlow system (III and IV). Experimental parameters are listed in Table 1.

View larger version (18K): [in this window] [in a new window]   FIG. 2. High-resolution product ion mass spectrum of M1 obtained by collision-induced dissociation of the [M + H]+ ion at m/z 675.20. Elemental composition of major product ions was calculated on the basis of high-resolution mass measurements (resolution 100,000).

	Results

Top Abstract Materials and Methods Results Discussion References

 
Sensitivity Comparison between Conventional LC-FSA and XFlow LC-ARC System. Bile samples (0–2 h) collected from [10-methoxy-³H]colchicine-treated rats were analyzed using four different instrument conditions (Table 1) to evaluate the XFlow LC-ARC system performance. The radioactivity chromatogram I (Fig. 1) was acquired on the conventional LC-FSA system after injection of 10 µl of filtrate, equivalent to 2.9 x 10^–2 µCi (6.5 x 10⁴ dpm). Colchicine and the major metabolites M3 and M4 were detected; however, the peak signal/noise ratio (PSNR) of colchicine was only 9. Peaks of metabolites M1 and M2 could barely be distinguished from the background noise level. Metabolites M5, M6, and M7 were not observed. In chromatogram II, acquired by the same conventional LC-FSA, the injection volume of filtrate was increased to 50 µl, quintupling the injected radioactivity. Under these conditions, the major metabolites were readily identified. The PSNR of colchicine reached 47 and the peaks of M1, M2, M3, and M4 were observed and resolved. The intensities of M1 and M2 peaks increased above the background. M5 was observed, but with low confidence because of poor achieved sensitivity. M6 and M7 were identified in the chromatogram but remained unresolved.

In comparison, the radioactivity chromatograms III and IV were acquired on the XFlow LC-ARC system after injection of 10 µl of filtrate. The XFlow system was operated in normal flow and dynamic flow modes. The normal flow mode mimicked the same functions as conventional FSA, providing a constant flow rate of the cocktail to maintain a fixed LC eluent/cocktail ratio. Results are shown in chromatogram III. The measured PSNR of colchicine was 209, representing a 20-fold PSNR increase versus chromatogram I. The radiometric peaks of metabolites M1 to M7 were detected. In chromatogram IV, the flow rates of cocktail and LC effluents were adjusted dynamically, depending on the appearance of a radiometric signal (dynamic flow mode). Under these experimental conditions, the radiometric peak of colchicine reached a PSNR of 185, similar to the observed value in normal flow mode (chromatogram III). Notably, metabolites M1 to M7 were now well separated. The retention times of M6 and M7 were 43.2 and 44.1 min, respectively, and were 2 min later than the respective retention times in III. However, the resolution between M6 and M7 improved. The total LC run times were identical in all radiochromatograms. In addition, the area ratio of each identified metabolite to the parent compound remained unaltered in chromatograms II, III, and IV, although different sensitivities and resolutions were observed.

View larger version (23K): [in this window] [in a new window]   FIG. 3. 1H NMR spectrum (600 MHz) of M1.

Metabolites M6, M7, and M8 displayed an [M + H]⁺ ion at m/z 386, indicating net loss of a methylene group from colchicine. M8 did not show a corresponding radiometric peak and eluted between M7 and colchicine (data not shown). The MS/MS spectrum of M8 revealed fragment ions at m/z 353, 344, 327, 312, 295/296, and 267, indicating a mass shift of 14 amu relative to the corresponding fragment ions of colchicine. This suggested that colchicine had undergone O-demethylation via loss of the radiolabeled methyl group at C10. In this biotransformation, the radiolabeled methyl group was released as tritiated methanol or formic acid and could be detected in the solvent front of the radiochromatogram. M8 was concluded to be a minor metabolite because only a modest radioactive signal was associated with the eluting solvent front. Observation of the relatively low peak area of M8 in both MS and UV spectra compared with colchicine, M6, and M7 supported this conclusion. Metabolites M6 and M7 had retention time of 43.2 and 44.1 min, respectively. MS/MS spectra of M6 and M7 exhibited the same fragment ion pattern as M8 (m/z 353, 344, 327, 312, 295/296, and 267), indicating that M6 and M7 were isomers of M8. Comparison of the LC retention times and MS/MS spectra of M6 and M7 against synthetic standards of 1-DMC, 2-DMC, and 3-DMC led to the unambiguous identification of M6 and M7 as 2-DMC and 3-DMC, respectively.

Metabolites M3 and M5 eluted at 36.3 and 38.2 min. Both metabolites yielded the same molecular ion [M + H]⁺ (m/z 466). MS/MS spectra of M3 and M5 showed a dominant fragment ion at m/z 386, consistent with a neutral loss of 80 amu from the molecular ion. This finding suggested that both metabolites were sulfate conjugates of a metabolic precursor with a molecular ion of m/z 386. Further MS/MS experiments (MS³) on this fragment ion (m/z 386) provided the typical product ion spectrum of DMC. Consequently, M3 and M5 were identified as sulfate conjugates of O-demethylated colchicine. Synthetic 1-, 2-, and 3-DMCs were incubated with rat hepatocytes, and the incubation supernatants were analyzed in LC-MS to unambiguously confirm the sulfate conjugate position. The retention times of direct sulfate conjugates from synthetic 2-DMC and 3-DMC in LC were identical to those observed from M5 and M3, respectively. Therefore, M3 and M5 were conclusively assigned as the 3-DMC sulfate and 2-DMC sulfate conjugates, respectively. M3 was detected as a major metabolite both in bile (Fig. 1) and in hepatocyte incubations and was isolated from the preparative in vitro incubation. The ¹H NMR spectrum was identical to the 3-DMC spectrum, except for the downfield shift of the H4 resonance from 6.47 ppm in 3-DMC to 6.63 ppm in 3-DMC sulfate.

The observed retention times for M2 and M4 were 24.2 and 37.1 min, respectively, and both metabolites had the same molecular ion [M + H]⁺ (m/z 562). Similar to the observations made for M3 and M5, the MS/MS spectra of M2 and M4 showed a dominant fragment ion at m/z 386 after neutral loss of 176 amu, suggesting glucuronide conjugates of the parent ion at m/z 386. The ion at m/z 386 was confirmed as the DMC moiety on the basis of daughter ions at m/z 353, 344, 327, 312, 295/296, and 267. Determination of the glucuronidation sites in M2 and M4 was achieved using the same experimental approach described for the sulfate conjugates. The retention times of glucuronide conjugates of synthetic 2-DMC and 3-DMC were identical to those for M4 and M2, respectively. This observation led to the assignment of M2 and M4 as the 3-DMC glucuronide and 2-DMC glucuronide, respectively. M2 was a major metabolite in hepatocyte incubation and was isolated by preparative LC. The proton chemical shifts of the colchicine moiety in the ¹H NMR spectrum were almost identical to 3-DMC, again with the only exception of a downfield chemical shift of H4 from 6.46 ppm in 3-DMC to 6.80 ppm in 3-DMC glucuronide (M2). The anomeric proton in the glucuronic acid moiety exhibited a chemical shift of 4.82 ppm with axial-axial coupling of 7.34 Hz, confirming that the glucuronide was in the β configuration.

M1 eluted at 24.2 min and its MS spectrum revealed a molecular ion [M + H]⁺ of m/z 675.2. The MS/MS spectrum acquired showed four major fragment ions at m/z 402.1, 528.2, 546. 2, and 600.2 (Fig. 2), which corresponded to the neutral loss of 273, 147, 129, and 75 amu, respectively. Typical neutral loss of 129 (glutamate moiety), 75 (glycine moiety), and 273 (GSH moiety) (Tang et al., 2005) inferred that M1 was a GSH conjugate. The high-resolution MS spectrum of M1 showed m/z 675.23332, which best matched a molecular formula of C₃₁H₃₈N₄O₁₁S(M 0.39 ppm). Subtraction of the GSH moiety led to a truncated molecular formula of C₂₁H₂₂NO₅, suggesting a net loss of CH₃O from the parent colchicine (C₂₂H₂₅NO₆).

The complete structural assignment of M1 was based on one- and two-dimensional ¹H NMR spectra. The ¹H NMR spectrum (Fig. 3) of isolated M1 showed three remaining methoxy groups at 3.78, 3.73, and 3.48 ppm, suggesting net loss of one methoxy group from colchicine. The resonance at 1.88 ppm indicated that the N-acetyl group was intact. Four aromatic protons (H4, H8, H11, and H12) in the tricyclic ring were found, and H4 was assigned as the singlet resonance at 6.74 ppm based on its correlation with an aliphatic proton at 2.52 ppm (H5) in the ROESY spectrum. The presence of a methoxy group at C3 was subsequently confirmed by the correlation between this particular methoxy group and H4 in the ROESY spectrum. H8 was assigned to another singlet resonance at 7.16 ppm based on its correlation with H7 at 4.31 ppm and the methyl group of the N-acetyl moiety (1.88 ppm). Two mutually coupled doublet resonances at 7.57 and 7.32 ppm were unique for the aromatic protons H11 and H12. H11 was tentatively assigned to the resonance at 7.57 ppm based on its relatively strong correlation with the methoxy group (3.48 ppm) at C10, which displayed an upfield resonance relative to the other methoxy groups on the six-membered aromatic ring. The chemical shifts of both H11 and H12 were observed downfield relative to the corresponding resonances in colchicine. The total correlation spectroscopy spectrum of M1 suggested that the molecule contained three spin systems, H5 to H7 in the tricyclic ring, Hd to Hf in glutamate, and Hb to Hc in cysteine.

Therefore, the loss of CH₃O atoms from colchicine could be narrowed down to methoxy groups at C1 or C2 or the oxygen at C9. The ROESY correlation between H7 and Hc (3.51 ppm) in cysteine suggested that the cysteine group in the GSH moiety was close to H7, and, thus, GSH could have substituted the oxygen at C9. A ROESY correlation between the 1-methoxy group and H12 is normally observed in colchicine and 3-DMC, but this correlation disappeared in M1, suggestive of a loss of the 1-methoxy group. A rearrangement of the conjugated double bonds in the tricyclic ring was proposed for M1 to satisfy valence requirements (Fig. 4). The observed absorption wavelength differences in the UV spectra of colchicine and M1 supported this double bond rearrangement (see Supplemental Data). Finally, the downfield shift of both H11 and H12 resonances in M1 could be caused by an anisotropic deshielding effect of the carbonyl group at C1, indirectly supporting the presence of a nearby carbonyl group.

View larger version (22K): [in this window] [in a new window]   FIG. 4. Proposed biotransformation pathways and structural variants of M1. A was excluded based on HRMS results.

	Discussion

Top Abstract Materials and Methods Results Discussion References

 
The metabolic profiling of radiolabeled colchicine discussed herein demonstrates the advantages of the XFlow LC-ARC system for online radiometric detection. In a comparative sensitivity analysis of radiolabeled colchicine, the PSNR of the colchicine peak using the XFlow LC-ARC system was approximately 20 times higher than that using conventional LC-FSA. The observed sensitivity improvement was consistent with previous published results (Nassar and Lee, 2007). The enhanced sensitivity and chromatographic resolution of this system enabled detection of three novel colchicine metabolites in rat bile samples.

View larger version (23K): [in this window] [in a new window]   FIG. 5. Metabolic pathway of colchicine in rat bile.

Previous disposition studies in rat revealed that colchicine and its metabolites were predominantly excreted into the bile (Hunter and Klaassen, 1975). 2-DMC, 3-DMC, 10-DMC, 2-DMC O-glucuronide, and 3-DMC O-glucuronide have been reported as major metabolites in bile (Sabouraud et al., 1992). The XFlow LC-ARC system allowed facile radiometric detection of the five known colchicine metabolites in the current study. O-Demethylation and subsequent O-glucuronidation were confirmed as the major metabolic pathway of colchicine. In addition, three hitherto unknown metabolites (M1, M3, and M5) were identified (Fig. 5). Novel sulfate conjugates M3 and M5 and the glutathione conjugate M1 suggested the existence of O-sulfation and glutathione phase II conjugation metabolic pathways. Metabolism of colchicine in vivo has not been fully investigated in humans (Niel and Scherrmann, 2006), but the in vitro experiments have demonstrated that colchicine undergoes same oxidative demethylation in both human and rat liver microsomes (Tateishi et al., 1997). The limited information regarding in vivo human colchicine metabolism could be partially due to low quantities of radiolabeled samples available from patients as therapeutic doses of colchicine are low (Cerquaglia et al., 2005). Therefore, it can be expected that application of XFlow LC-ARC system will overcome this limitation, and the system will become an important tool for in vivo human metabolism studies in the future.

MS/MS fragmentation spectra were insufficient for unambiguous determination of demethylation and conjugation sites. Synthetic 1-, 2-, and 3-DMCs were used to confirm the demethylated metabolites M6 and M7 and generate the glucuronide and sulfate conjugate metabolites in rat hepatocyte incubations. 2-DMC (M6) and 3-DMC (M7) were identified as the major O-demethylated metabolites, whereas 10-DMC was observed as a minor metabolite. 2-DMC O-glucuronide (M4) and 3-DMC O-sulfate (M3) were determined to be the two major conjugated metabolites. O-Glucuronide and sulfate conjugation did not significantly affect proton resonances of the DMC metabolic precursor in ¹H NMR spectra. UV spectra of the conjugated metabolites were identical to those of colchicine. The radioactive peak area of M4 was significant bigger than that for M5, suggesting that 2-DMC was preferably conjugated with glucuronic acid in vivo. In contrast, 3-DMC preferably underwent O-sulfation. Although the regioselectivity of cytochrome P450- and aldehyde oxidase-catalyzed reactions has been widely studied (Sheridan et al., 2007; Torres et al., 2007), the regiopreferences of glucuronidation and sulfation currently remain unknown. These regiodependent conjugation patterns of DMCs might serve as an interesting model for further investigation of geometric preferences in phase II biotransformations.

M1 was identified as an unexpected metabolite of colchicine. MS/MS data revealed a neutral loss of 129, indicating the presence of a GSH adduct in M1. Because O-demethylation has been established as the major phase I metabolic pathway for colchicine, the structural variant A was originally proposed (Fig. 4). It was rationalized that consecutive O-demethylation of two methoxy moieties yields a dihydroquinone intermediate. Under aerobic conditions, the dihydroquinone moiety is readily oxidized to form the ortho-quinone, which would then be susceptible to the GSH conjugation (Chen et al., 2002; Evans et al., 2004; Yu et al., 2004). The molecular formula of this proposed isobaric GSH conjugate A is C₃₀H₃₄N₄O₁₂S, with a calculated exact mass of 674.18939. However, actual measurement by HRMS of M1 showed a 52.7 ppm mass difference from the calculated mass of A, indicating that the proposed structure A was probably incorrect. The most probable elemental composition derived from M1 measured by HRMS was C₃₁H₃₈N₄O₁₁S, suggestive of a net loss of CH₃O from colchicine instead of two carbons and seven hydrogens as proposed in structure A. HRMS was critical in distinguishing the proposed isobaric molecular formulae. Subsequent ¹H NMR analysis confirmed that only one methoxy resonance had been lost in the spectrum of the isolated metabolite M1. Thus, M1 formation must have proceeded through a monodemethylation step and the biotransformation pathway (B), consistent with this observation, was proposed in Fig. 4. It is postulated that colchicine was first regioselectively demethylated to 1-DMC, followed by 1,2 addition of GSH to the carbonyl group at C9. Dehydration and rearrangement of the conjugated double bonds yielded M1. Synthetic 1-, 2-, and 3-DMCs were individually incubated with primary rat hepatocytes to examine whether 1-DMC was indeed an intermediate in this pathway. Unfortunately, M1 was not detectable in any of the incubation mixtures. Whether, in fact, O-demethylation and GSH addition occur simultaneously or as a two-step process remains to be determined. Currently 1,2 additions of GSH have been reported only in chemical reactions (Gebauer, 2007) or in the permeabilized recombinant yeast incubation (Liu et al., 1999). Further experiments to ascertain the mechanistic subtleties of this biotransformation are currently under consideration.

Endogenous GSH is a natural trapping agent for chemically reactive metabolites. Formation of the GSH conjugate M1 suggests that colchicine biotransformation may proceed through an electrophilic intermediate. Electrophilic activation of xenobiotics can lead to covalent binding of these molecules to biological macromolecules and is often associated with drug-induced idiosyncratic toxicities (Knowles et al., 2003; Evans et al., 2004; Walgren et al., 2005). Chronic colchicine therapy in patients causes adverse effects mainly in the gastrointestinal system (Cerquaglia et al., 2005; Niel and Scherrmann, 2006) and acute toxicity of colchicine is manifested in multiorgan failure. No toxicities associated with bioactivation have been reported so far. It is noteworthy to point out that idiosyncratic reactions are generally associated with drugs administrated at elevated doses (Knowles et al., 2003). Thus, idiosyncratic syndromes are unlikely to be observed in colchicine therapy, because very low doses (0.03 mg/kg/day) are given daily (Cerquaglia et al., 2005). The propensity for covalent protein binding of colchicine and its consequences for organ toxicity are currently under investigation.

	Acknowledgments

 
We thank Dr. D. Y. Lee for the support of the XFlow system and the laboratory animal research department at MRL-Boston for providing animal care and conducting the in vivo experiment. We also thank Dr. Ziping Yang for technical assistance.

	Footnotes

 
Article, publication date, and citation information can be found at http://dmd.aspetjournals.org.

doi:10.1124/dmd.107.019463.

ABBREVIATIONS: ADME, absorption, distribution, metabolism, and excretion; HPLC, high-performance liquid chromatography; FSA, flow scintillation analyzers; LC, liquid chromatography; ARC, accurate radioisotope counting; MS, mass spectrometry; MS/MS, tandem mass spectrometry; NMR, nuclear magnetic resonance; DMC, demethylcolchicine; GSH, glutathione; HRESI, high-resolution electrospray ionization; PSNR, peak signal/noise ratio; ROESY, rotating frame Overhauser effect spectroscopy; HRMS, high-resolution mass spectrometry.

The online version of this article (available at http://dmd.aspetjournals.org) contains supplemental material.

Address correspondence to: Dr. Lin Xu, DMPK, Drug Design and Optimization, Merck & Co., Inc., BMB4-110, 33 Avenue Louis Pasteur, Boston, MA 02115. E-mail: lin_xu2{at}merck.com

	References

Top Abstract Materials and Methods Results Discussion References

 


Aloj L and Morelli G (2004) Design, synthesis and preclinical evaluation of radiolabeled peptides for diagnosis and therapy. Curr Pharm Des 10: 3009–3031.[CrossRef][Medline]

Blade-Font A (1979) New chemistry of colchicine and related compounds. IV. Selective demethylations of colchicine with Lewis acids and the structure of Zeisel's dimethylcolchicinic acid. Afinidad 36: 329–331.

Bruin GJ, Waldmeier F, Boernsen KO, Pfaar U, Gross G, and Zollinger M (2006) A microplate solid scintillation counter as a radioactivity detector for high performance liquid chromatography in drug metabolism: validation and applications. J Chromatogr A 1133: 184–194.[CrossRef][Medline]

Burkey JL, Campanale KM, Barbuch R, O'Bannon D, Rash J, Benson C, and Small D (2006) Disposition of [¹⁴C]ruboxistaurin in humans. Drug Metab Dispos 34: 1909–1917.[Abstract/Free Full Text]

Cerquaglia C, Diaco M, Nucera G, La Regina M, Montalto M, and Manna R (2005) Pharmacological and clinical basis of treatment of familial Mediterranean fever (FMF) with colchicine or analogues: an update. Curr Drug Targets Inflamm Allergy 4: 117–124.[CrossRef][Medline]

Chen Q, Ngui JS, Doss GA, Wang RW, Cai X, DiNinno FP, Blizzard TA, Hammond ML, Stearns RA, Evans DC, et al. (2002) Cytochrome P450 3A4-mediated bioactivation of raloxifene: irreversible enzyme inhibition and thiol adduct formation. Chem Res Toxicol 15: 907–914.[CrossRef][Medline]

Correia JJ (1991) Effects of antimitotic agents on tubulin-nucleotide interactions. Pharmacol Ther 52: 127–147.[CrossRef][Medline]

Dalvie D (2000) Recent advances in the applications of radioisotopes in drug metabolism, toxicology and pharmacokinetics. Curr Pharm Des 6: 1009–1028.[CrossRef][Medline]

Evans DC, Watt AP, Nicoll-Griffith DA, and Baillie TA (2004) Drug-protein adducts: an industry perspective on minimizing the potential for drug bioactivation in drug discovery and development. Chem Res Toxicol 17: 3–16.[CrossRef][Medline]

Gebauer M (2007) Synthesis and structure-activity relationships of novel warfarin derivatives. Bioorg Med Chem 15: 2414–2420.[CrossRef][Medline]

Göbel A, McArdell CS, Suter MJ, and Giger W (2004) Trace determination of macrolide and sulfonamide antimicrobials, a human sulfonamide metabolite, and trimethoprim in wastewater using liquid chromatography coupled to electrospray tandem mass spectrometry. Anal Chem 76: 4756–4764.[Medline]

Hunter AL and Klaassen CD (1975) Biliary excretion of colchicine. J Pharmacol Exp Ther 192: 605–617.[Abstract/Free Full Text]

James JK, Nakamura M, Nakazato A, Zhang KE, Cramer M, Brunner J, Cook J, and Chen WG (2005) Metabolism and disposition of a potent group II metabotropic glutamate receptor agonist, in rats, dogs, and monkeys. Drug Metab Dispos 33: 1373–1381.[Abstract/Free Full Text]

Jang BS, Lim E, Hee Park S, Shin IS, Danthi SN, Hwang IS, Le N, Yu S, Xie J, Li KC, et al. (2007) Radiolabeled high affinity peptidomimetic antagonist selectively targets _vβ₃ receptor-positive tumor in mice. Nucl Med Biol 34: 363–370.[CrossRef][Medline]

Jigorel E, Le Vee M, Boursier-Neyret C, Bertrand M, and Fardel O (2005) Functional expression of sinusoidal drug transporters in primary human and rat hepatocytes. Drug Metab Dispos 33: 1418–1422.[Abstract/Free Full Text]

Knowles SR, Shapiro LE, and Shear NH (2003) Reactive metabolites and adverse drug reactions: clinical considerations. Clin Rev Allergy Immunol 24: 229–238.[CrossRef][Medline]

Liu Y, Hama H, Fujita Y, Kondo A, Inoue Y, Kimura A, and Fukuda H (1999) Production of S-lactoylglutathione by high activity whole cell biocatalysts prepared by permeabilization of recombinant saccharomyces cerevisiae with alcohols. Biotechnol Bioeng 64: 54–60.[CrossRef][Medline]

Marathe PHS, Shyu WC, Humphreys WG (2004) The use of radiolabeled compounds for ADME studies in discovery and exploratory development. Curr Pharm Des 10: 2991–3008.[CrossRef][Medline]

Morejohn LC and Fosket DE (1991) The biochemistry of compounds with anti-microtubule activity in plant cells. Pharmacol Ther 51: 217–230.[CrossRef][Medline]

Nassar AE, Bjorge SM, and Lee DY (2003) On-line liquid chromatography-accurate radioisotope counting coupled with a radioactivity detector and mass spectrometer for metabolite identification in drug discovery and development. Anal Chem 75: 785–790.[Medline]

Nassar AE and Lee DY (2007) Novel approach to performing metabolite identification in drug metabolism. J Chromatogr Sci 45: 113–119.[Medline]

Nassar AE, Parmentier Y, Martinet M, and Lee DY (2004) Liquid chromatography-accurate radioisotope counting and microplate scintillation counter technologies in drug metabolism studies. J Chromatogr Sci 42: 348–353.[Medline]

Niel E and Scherrmann JM (2006) Colchicine today. Joint Bone Spine 73: 672–678.[CrossRef][Medline]

Onisko BC (2002) Radiochemical detection for packed capillary liquid chromatography-mass spectrometry. J Am Soc Mass Spectrom 13: 82–84.[CrossRef][Medline]

Peterman SM, Duczak N Jr, Kalgutkar AS, Lame ME, and Soglia JR (2006) Application of a linear ion trap/orbitrap mass spectrometer in metabolite characterization studies: examination of the human liver microsomal metabolism of the nontricyclic anti-depressant nefazodone using data-dependent accurate mass measurements. J Am Soc Mass Spectrom 17: 363–375.[CrossRef][Medline]

Rösner M, Capraro HG, Jacobson AE, Atwell L, Brossi A, Iorio MA, Williams TH, Sik RH, and Chignell CF (1981) Biological effects of modified colchicines. Improved preparation of 2-demethylcolchicine, 3-demethylcolchicine, and (+)-colchicine and reassignment of the position of the double bond in dehydro-7-deacetamidocolchicines. J Med Chem 24: 257–261.[CrossRef][Medline]

Sabouraud A, Rochdi M, Urtizberea M, Christen MO, Achtert G, and Scherrmann JM (1992) Pharmacokinetics of colchicine: a review of experimental and clinical data. Z Gastroenterol 30 (Suppl 1): 35–39.[Medline]

Seglen PO (1976) Preparation of isolated rat liver cells. Methods Cell Biol 13: 29–83.[Medline]

Sheridan RP, Korzekwa KR, Torres RA, and Walker MJ (2007) Empirical regioselectivity models for human cytochromes P450 3A4, 2D6, and 2C9. J Med Chem 50: 3173–3184.[CrossRef][Medline]

Tang C, Subramanian R, Kuo Y, Krymgold S, Lu P, Kuduk SD, Ng C, Feng DM, Elmore C, Soli E, et al. (2005) Bioactivation of 2,3-diaminopyridine-containing bradykinin B1 receptor antagonists: irreversible binding to liver microsomal proteins and formation of glutathione conjugates. Chem Res Toxicol 18: 934–945.[CrossRef][Medline]

Tateishi T, Soucek P, Caraco Y, Guengerich FP, and Wood AJ (1997) Colchicine biotransformation by human liver microsomes: identification of CYP3A4 as the major isoform responsible for colchicine demethylation. Biochem Pharmacol 53: 111–116.[CrossRef][Medline]

Torres RA, Korzekwa KR, McMasters DR, Fandozzi CM, and Jones JP (2007) Use of density functional calculations to predict the regioselectivity of drugs and molecules metabolized by aldehyde oxidase. J Med Chem 50: 4642–4647.[CrossRef][Medline]

Walgren JL, Mitchell MD, and Thompson DC (2005) Role of metabolism in drug-induced idiosyncratic hepatotoxicity. Crit Rev Toxicol 35: 325–361.[Medline]

Wallace SL (1974) Colchicine. Semin Arthritis Rheum 3: 369–3814.[CrossRef][Medline]

Weeks JL, Blount MA, Beasley A, Zoraghi R, Thomas MK, Sekhar KR, Corbin JD, and Francis SH (2005) Radiolabeled ligand binding to the catalytic or allosteric sites of PDE5 and PDE11. Methods Mol Biol 307: 239–262.[Medline]

Yang ZP, Kwei G, Harsch A, and Xu L (2006) High sensitivity on-line radiometric detection: application for improved colchicine metabolism pathway analysis. Proceedings of Applied Pharmaceutical Analysis; 2006 Sept. 11; Boston Society for Advanced Therapeutics, Boston.

Yu L, Liu H, Li W, Zhang F, Luckie C, van Breemen RB, Thatcher GR, and Bolton JL (2004) Oxidation of raloxifene to quinoids: potential toxic pathways via a diquinone methide and o-quinones. Chem Res Toxicol 17: 879–888.[CrossRef][Medline]

This Article Abstract Full Text (PDF) Data Supplement All Versions of this Article: dmd.107.019463v1 36/4/731    most recent Submit a response Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Xu, L. Articles by Harsch, A. Search for Related Content PubMed PubMed Citation Articles by Xu, L. Articles by Harsch, A. Pubmed/NCBI databases Compound via MeSH Substance via MeSH Hazardous Substances DB COLCHICINE