![[Feedback]](/icons/banner/feedbackACT.gif){kind=icon}
![[For Subscribers]](/icons/banner/subscriberACT.gif){kind=icon}
![[Archive]](/icons/banner/archiveACT.gif){kind=icon}
![[Search]](/icons/banner/searchACT.gif){kind=icon}
![[Contents]](/icons/banner/tocACT.gif){kind=icon}
|
|
|
|
|
||
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
ej Slana
ek PerlíkLaboratory of Metabolism, Center for Cancer Research, National Cancer Institute, National Institutes of Health, Bethesda, Maryland (Y.Z., K.W.K., C.C., F.J.G.); Institute of Clinical Pharmacology, University of Bern, Switzerland (J.S., J.R.I.); Department of Pharmaceutical Sciences, Oregon State University, Corvallis, Oregon (K.L.M., T.M.Z.); and Institute of Pharmacology, 1st Faculty of Medicine, Charles University, Prague, Czech Republic (O.S., F.P., J.R.I.).
(Received December 12, 2005; Accepted June 12, 2006)
 |
Abstract |
|---|
 Top Abstract Materials and MethodsResultsDiscussionReferences |
|---|
) and immediately went into clinical trials as an anti-hypertensive agent (Talbot, 1965; Rosendorff et al., 1968; Somers et al., 1968; Blechman et al., 1969). Surprisingly, at the time of its launch, little, if anything was known about the metabolic disposition of debrisoquine. Workers at Hoffmann-La Roche UK (Allen et al., 1975) reported on a study in which 14C-labeled debrisoquine was administered to rats (50 mg/kg) and a single hypertensive patient (2.6 mg, on top of 15 mg q.d.s. therapeutic dose). Debrisoquine was excreted unchanged in urine of both human and rat, together with 4-hydroxydebrisoquine as the major metabolite, and traces of the phenolic metabolites 5-, 6-, 7-, and 8-hydroxydebrisoquine. In addition, both rats and humans excrete 10 to 15% of the dose as two ring-opened acidic metabolites, presumed to arise from hydroxylation of debrisoquine in positions 1 and 3 (Fig. 1A). The nature of these metabolites was subsequently confirmed (Allen et al., 1976; Eiermann et al., 1998). For the rat, 70% of the administered radioactivity could be accounted for as debrisoquine and these seven aforementioned metabolites. In the single patient who was studied, 66% of the radio-activity was similarly accounted for (Allen et al., 1975).
|
). This same group performed a second 14C-debrisoquine study in four human volunteers, for which 71.3 to 76.7% of the radioactivity was recovered in the urine within 24 h (Angelo et al., 1976). These authors identified debrisoquine and 4-hydroxydebrisoquine in human urine, and tentatively identified the phenolic metabolites and the carboxylic acid metabolites, as previously reported (Allen et al., 1975), together with a dihydroxy metabolite, which was later reported by another laboratory (Eiermann et al., 1998). Of interest was the report that one of the subjects excreted 60.8% of the dose as debrisoquine and only 5.4% as 4-hydroxydebrisoquine. This would correspond to a metabolic ratio (Mahgoub et al., 1977) (MR) of 60.8/5.4 = 11.3. Subsequent studies showed this volunteer to be a phenotypic poor metabolizer (PM) (Mahgoub et al., 1977). The other three subjects in this study had MR values of 0.5 to 1.4, typical of what is regarded as extensive metabolizers (EMs) (Mahgoub et al., 1977). This was perhaps the first report containing evidence of the existence of two drug oxidation phenotypes, EM and PM.
There were few further qualitative insights added to the debrisoquine metabolism knowledge base at that time, with the exception of a single 14C-debrisoquine study (Idle et al., 1979). These authors confirmed the excretion of debrisoquine, 4-, 5-, 6-, 7-, and 8-hydroxydebrisoquine, and also reported that neither the debrisoquine urea derivative nor N-hydroxydebrisoquine (see Fig. 1A) was excreted in human urine after administration by mouth of 40 mg of debrisoquine sulfate (Idle et al., 1979). However, none of these early studies accounted for the chemical nature of all the debrisoquine metabolites.
Subsequently, debrisoquine 4-hydroxylation took on a special significance as the first drug oxidation pathway to display genetic polymorphism (Mahgoub et al., 1977). This reaction is mediated by the polymorphic cytochrome P450 CYP2D6 (Gonzalez et al., 1988), but CYP1A1 may also play a role (Granvil et al., 2002), A wide variety of allelic variants of CYP2D6 give rise to four phenotypes for CYP2D6-mediated metabolism: ultrarapid metabolizer (UM), EM, intermediate metabolizer (IM), and PM phenotypes, which have been estimated to occur in 5 to 10%, 65 to 80%, 10 to 15%, and 5 to 10% of populations, respectively (Zanger et al., 2004). In addition to the archetypal polymorphically metabolized substrates, debrisoquine and sparteine, CYP2D6 substrates include at least 50 common pharmaceutical agents (Zanger et al., 2004), together with four endogenous 5-methoxyindolethylamines (Yu et al., 2003b), and also progesterone and tyramine (Niwa et al., 2004). PMs have an impaired hydroxylation capacity, and this may lead to adverse effects (Yu et al., 2004; Zanger et al., 2004). The current practice for the identification of PMs is to screen for common CYP2D6-inactivating alleles using polymerase chain reaction (Muller et al., 2003; Schaeffeler et al., 2003) or microarray-based methods (Chou et al., 2003). However, genotypic analysis provides only a forecast of the extent of metabolism of such well studied test substrates as debrisoquine and sparteine. Part of the problem is that the UM, EM, and IM phenotypes are both broad and overlapping. Despite this, individual MR values for debrisoquine and sparteine are stable over many years (Crothers et al., 1986; Zanger et al., 2004), and phenotype testing with debrisoquine or sparteine still provides a cheap, reliable, and direct evaluation of the CYP2D6 metabolic phenotype.
The factors that contribute to the intraphenotype variance in CYP2D6-mediated metabolism are poorly understood. For example, in the case of polymorphic debrisoquine 4-hydroxylation, most homozygous extensive metabolizers have MR values in the range 0.1 to 1.0, corresponding to 50 to 91% 4-hydroxylation of the drug (Daly et al., 1991). The occurrence of multiple variant CYP2D6 alleles explains the broad interphenotype variation in debrisoquine 4-hydroxylation but does not account for the wide intraphenotype variation in metabolism. Other explanations must be sought for this variability and, thus, the metabolism of debrisoquine was reexamined in both genotypic EM and PM subjects. In this study, a novel human urinary metabolite of debrisoquine, 3,4-dehydrodebrisoquine, was uncovered. The data also indicate that both debrisoquine and 4-hydroxydebrisoquine undergo glucuronidation. These novel metabolites could have a significant impact upon the debrisoquine metabolic ratio.
|
Materials and Methods |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
Subjects. The study was approved by the ethics committee of the General Faculty Hospital Prague and the Office of Human Subjects Research, National Institutes of Health, and was performed according to Good Clinical Practice Guidelines. All volunteers gave their written informed consent to participate in the study. The volunteers comprised 14 healthy Caucasians of Czech or Slovak nationality, primarily students of Charles University 1st Faculty of Medicine, and were aged 23.5 ± 1.0 years (mean ± S.E.M.). They were selected according to CYP2D6 genotype and availability from a pool of 250 volunteers that had been previously genotyped as EM or PM by the determination of CYP2D6 alleles *2, *3, *4, *5, *6, *7, *8, *9, *10, *11, *12, *13, *14, and *15 and gene duplications (Sachse et al., 1997). They comprised nine *1/*1 EM subjects (6 M, 3 F) and four *4/*4 PM subjects (3 M, 1 F), with mean Body Mass Index (±S.E.M.) of 23.4 ± 0.7 (male) and 20.5 ± 1.8 (female). All subjects were normotensive with mean (±S.E.M.) sitting blood pressures of 123 ± 2/79 ± 2 (male) and 115 ± 4/69 ± 4 (female). The volunteers were nonsmokers, with the exception of one female EM who smoked 5 cigarettes per day and one male PM who smoked 10 cigarettes per day. Alcohol consumption was 0 to 3 units/week for males and 0 to 2 units/week for females. The volunteers were drug-free for the previous month, with the exception of one male EM who had taken 800 mg of ibuprofen at –7 days and 10 mg of zolpidem [a CYP3A4 substrate (Pichard et al., 1995)] at –10 days, and one female EM who had taken 1200 mg of ibuprofen at –11/–12 days. Cigarette smoking (Vincent-Viry et al., 2000), alcohol consumption (Vincent-Viry et al., 2000), and ibuprofen (Hamman et al., 1997) and zolpidem (von Moltke et al., 2002) administration are unlikely to alter debrisoquine metabolism by CYP2D6. However, the effect of these potential confounders on non-P450 metabolism is at present unknown.
Debrisoquine Administration. After an overnight fast, each subject took a single tablet containing 12.8 mg of debrisoquine hemisulfate, equivalent to 10 mg of debrisoquine (gift of Professor Ann K. Daly, University of Newcastle, UK), taken from a newly opened bottle, from which five tablets had been taken, ground, suspended in water, shaken, centrifuged, and assayed by gas chromatography-mass spectrometry (GCMS) (see below) to ensure the quantity and quality of their debrisoquine content. After a period of 2 h, volunteers were allowed food and soft drinks ad libitum. They collected bulked 0- to 8- and 8- to 24-h urines which were taken promptly to the Clinical Pharmacology Unit, General Teaching Hospital in Prague, where they were measured, and 10-ml aliquots were frozen at –20°C before transport of a single set of samples by courier on dry ice to the National Institutes of Health, Bethesda, MD for analysis. These samples were coded in such as way that they could not be identified in Bethesda with particular volunteers. The same urine samples were analyzed for creatinine excretion (millimoles per 24 h) and a blood sample was taken for the determination of serum creatinine (micromoles per liter). Creatinine clearance (milliliters per second) and glomerular filtration rate (milliliters per second) were calculated using the Cockcroft-Gault formula (Cockcroft and Gault, 1976).
In Vitro Metabolism Studies. Long-term (24-h) incubations to obtain cytochrome P450 activity were carried out using pre-prepared human liver microsomes and rat liver microsomes under conditions we have described previously (Yu et al., 2003a). Both debrisoquine hemisulfate (7 µg/ml; 31 µM) and 4-hydroxydebrisoquine (20 µg/ml; 105 µM) were used as substrates. The incubations were assayed for their content of debrisoquine, 4-hydroxydebrisoquine, and 3,4-dehydrodebrisoquine by GCMS with single ion extraction, as described below for urine samples.
Analysis of Debrisoquine and Its Hydroxylated Metabolites by GCMS. The following calibration curves were constructed using blank human urine: debrisoquine hemisulfate and 4-hydroxydebrisoquine, from 0.2 to 20 µg/ml; 6- and 8-hydroxydebrisoquine, from 0.5 to 5.0 µg/ml. All urines were treated with ß-glucuronidase to deconjugate phenolic metabolites, as follows. To urine (0.5 ml) was added 1 M phosphate-buffered saline (pH 4.8) and ß-glucuronidase solution (54 µl; Sigma-Aldrich) to give a final pH of 5.0, and heated in a shaking water bath at 37°C for 24 h. Distilled water (0.44 ml) was then added and the samples with 1.00-ml volume were then treated like the diluted urine samples below. Standards and volunteer urines, with and without ß-glucuronidase treatment, were treated with 1,1,1,5,5,5-hexafluoro-2,4-pentanedione (hexafluoroacetylacetone) (Sigma-Aldrich) to create bis(trifluoromethyl)pyrimidine derivatives (Erdtmansky and Goehl, 1975) as follows. Duplicate urine samples or standards (0.5 ml) were diluted with distilled water (0.5 ml), unless already diluted after deconjugation, mixed with both saturated aqueous sodium hydrogen carbonate (0.2 ml) and internal standard solution (0.1 ml) containing 7 µg/ml 7-methoxyguanoxan, and then hexafluoroacetylacetone (200 µl) and high-performance liquid chromatography-grade toluene (1 ml) were added. The samples were heated in sealed septum-cap vials in a heating block at 90°C for 1 h and cooled, 1 M sodium hydroxide (1 ml) was added, and the samples were vortexed for 10 s and centrifuged at 1800g for 5 min. Upper organic layers were transferred to autoinjector vials and capped. Samples (1 µl) were injected in splitless mode from an Agilent 7683 Automatic Liquid Sampler (Agilent, Palo Alto, CA) into an Agilent 5890N Network GC system coupled to an Agilent 5973 Network Mass Selective Detector and a G1701 DA Enhanced ChemStation. Chromatography was performed on a 0.25-µm film thickness HP-5MS capillary column (28 m x 0.25 mm i.d.; Agilent, Palo Alto, CA). Carrier gas was helium (linear velocity 39 cm/s) with a temperature program of 10 min at 120°C, then 10°/min to 250°C, then 20°/min to 300°C, and then held at 300°C for 3 min (total run time 28.5 min). Injector and interface temperatures were 230 and 280°C, respectively. The mass spectrometer was operated in the electron ionization mode (69.9 eV). A mass range of 100 to 550 atomic mass units was scanned, only for the retention period of 16.8 to 24.0 min, and this permitted acquisition of 3.25 scans/s. Data sharing was accomplished by file transfer direct from the ChemStation to a file transfer protocol (ftp) server in Prague, where the appropriate analytical software was located. Peaks were identified from the retention times and mass spectra of authentic standards. For quantitation, peak area ratios (ratio of peak area of analyte/peak area of internal standard) were calculated for all peaks of interest using single ion-extracted chromatograms with the following ions: debrisoquine (347 m/z), 4-hydroxydebrisoquine (344 m/z), 3,4-dehydrodebrisoquine (344 m/z), 6-hydroxydebrisoquine (362 m/z), 8-hydroxydebrisoquine (362 m/z), and 7-methoxyguanoxan (internal standard) (379 m/z).
Analysis of Conjugated Metabolites by Ultra-Performance Liquid Chromatography-Coupled Time-of-Flight Mass Spectroscopy (UPLC-TOFMS). Urine samples were mixed with an equal volume of acetonitrile and centrifuged at 14,000 rpm to remove particles and protein. A 200-µl aliquot of the supernatant was transferred to an autosampler vial for UPLC-TOFMS analysis (Wilson et al., 2005). Urine samples (5 µl/injection) were separated on a 50 x 2.1 mm ACQUITY 1.7-µm C18 column (Waters Corp., Milford, MA) using an ACQUITY UPLC system (Waters) with a gradient mobile phase composed of A, 0.1% formic acid; and B, acetonitrile containing 0.1% formic acid. A 0.6 ml min–1 flow rate was maintained in a 10-min run. The eluent was directly introduced into the mass spectrometer. Mass spectrometry was performed on a Waters Q-TOF Premier operating in positive ion mode. The desolvation gas flow was set to 600 l h–1 at a temperature of 350°C, with the cone gas set to 50 l h–1 and the source temperature set to 120°C. The capillary voltage and the cone voltage were set to 3000 and 60 V, respectively. Leucine-enkephalin was used as the lock mass (m/z 556.2771) for accurate mass measurement. In MS scanning, data were acquired in centroid mode from 100 to 950 m/z. As for MS/MS fragmentation of target ions, collision energy ranging from 15 to 30 V was applied. After data acquisition, UPLC-MS chromatograms and spectra were further analyzed by MassLynx application software (Waters).
Synthesis and Characterization of 3,4-Dehydrodebrisoquine. 4-Hydroxydebrisoquine (6 mg; 31.4 µmol) was added to 18 M sulfuric acid (100 µl; 1.8 mmol) and dimethyl sulfoxide (200 µl) in a screw-topped vial, and the mixture was vortexed and then heated at 90°C for 20 min. An aliquot was taken, neutralized with saturated sodium hydrogen carbonate, and derivatized with hexafluoroacetylacetone, as described above. GCMS analysis established the absence of 4-hydroxydebrisoquine at 19.60 min, but the appearance of a single large peak at 17.81 min (see Results). These observations were taken as evidence of near 100% conversion of 4-hydroxydebrisoquine to 3,4-dehydrodebrisoquine. The resulting reaction product was diluted and used to construct a calibration curve for 3,4-dehydrodebrisoquine from 0.46 to 4.6 µg/ml. For purposes of authentication of synthetic 3,4-dehydrodebrisoquine, iterative experiments were performed with 4-hydroxydebrisoquine and 18 M sulfuric acid alone, and the products were monitored by GCMS after derivatization with hexafluoroacetylacetone, to determine the best conditions for the isolation of solid product for further analysis. It was determined that 4-hydroxydebrisoquine (50 mg; 262 µmol), when reacted with 1 ml of 18 M sulfuric acid (18 mmol) at 35 to 40°C for 5 min, in the absence of dimethyl sulfoxide, yielded sufficient off-white product (20 mg; 44% yield) after neutralization with saturated sodium hydrogen carbonate and sample clean-up using SepPak C18 cartridges (Waters). The methanol eluates were combined and reduced to dryness in a Speed Vac evaporating centrifuge (Savant Instruments, Farmingdale, NY). The resulting solid material was subjected to LC-MS/MS analysis, as described above, and to 1H NMR spectroscopy in D2O using a Bruker (Billerica, MA) Avance DPX 300 spectrometer.
Statistical Analyses. The 0- to 8-h, 8- to 24-h, and 0- to 24-h urinary excretion of debrisoquine and its metabolites was expressed as mean ± S.E.M. Interphenotype differences in the excretion of metabolites were evaluated statistically using the Mann-Whitney U test with one-tailed P values.
|
Results |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
|
Calibration curves for debrisoquine (retention time 17.46 min) and 4-hydroxydebrisoquine (retention time 19.60 min) were determined. The correlation coefficients (r2) were 0.999 for both debrisoquine and 4-hydroxydebrisoquine, and limits of sensitivity for debrisoquine and 4-hydroxydebrisoquine were 0.02 µg/ml and 0.05 µg/ml, respectively. Interassay variation was in the range 3 to 5%. Calibration curves for 8-hydroxydebrisoquine (retention time 20.39 min) and 6-hydroxydebrisoquine (retention time 20.56 min) were determined. Coefficients (r2) were 0.974 and 0.918 for 6-hydroxydebrisoquine and 8-hydroxydebrisoquine, respectively, and the limit of sensitivity for both compounds was 0.2 µg/ml. Interassay variation was in the range 5 to 10%. Calibration curves for 3,4-dehydrodebrisoquine (retention time 17.36 min), and an unknown peak (retention time 17.52 min) that was always present in solutions of 3,4-dehydrodebrisoquine, except immediately after synthesis, were determined (Fig. 2). This unknown compound had a mass spectrum almost identical to that of 3,4dehydrodebrisoquine. As will be seen, it was present in urine samples. The nature of this unknown compound is dealt with below. The relative concentrations of peaks A and B were determined from the relative total ion abundances for each peak. The correlation coefficients (r2) for both 3,4-dehydrodebrisoquine and its unknown derivative were 0.996, and the limits of sensitivity for 3,4-dehydrodebrisoquine and its unknown derivative were 0.4 µg/ml and 0.06 µg/ml, respectively.
The Nature of the Unknown Derivative of 3,4-Dehydrodebrisoquine. An extracted-ion chromatogram (344 m/z), with two peaks running at 17.33 min (peak A) and 17.48 min (peak B) in a solution of 5 µg/ml authentic 3,4-dehydrodebrisoquine derivatized with hexafluoroacetylacetone, is displayed in Fig. 3. Peak A corresponds to 3,4-dehydrodebrisoquine, with the difference in retention time to the synthetic material, shown in Fig. 2, due wholly to analysis on different GC columns. Peak B has an almost identical mass spectrum dominated by the M–1 transition (loss of H radical) shown in Fig. 2. No type of isomerism is possible for 3,4-dehydrodebrisoquine, neither around the 3,4-double bond nor the heterocyclic nitrogen atom. Therefore, peak B must represent a precursor of 3,4-dehydrodebrisoquine, and we propose that it is 3-hydroxydebrisoquine formed by the mechanism shown in Fig. 1B, specifically via the equilibrium A and the addition of OH– in reaction B. This relatively unstable carbinolamine, 3-hydroxydebrisoquine, is expected to dehydrate to give 3,4dehydrodebrisoquine in the mass spectrometer [as the respective bis(trifluoromethyl)pyrimidine derivatives] via reaction C. Indeed, this reaction occurs to some extent for 4-hydroxydebrisoquine itself, which also has a prominent ion at 344 m/z. Therefore, it would appear that a proportion of synthetic 3,4-dehydrodebrisoquine is converted to 3-hydroxydebrisoquine in vitro.
|
Determination of Debrisoquine Metabolism in Human Volunteers. First, the 0- to 8-h and 8- to 24-h excretion of debrisoquine, 4-, 6-, and 8-hydroxydebrisoquine was determined, together with the excretion of 3,4-dehydrodebrisoquine (sum of peaks A and B; see above). These determinations were performed on neat urine and on urine treated with ß-glucuronidase. Table 1 shows the 0- to 8-h urinary excretion of debrisoquine, 4-hydroxydebrisoquine, and 3,4-dehydrodebrisoquine without ß-glucuronidase treatment. As expected, EMs excreted significantly more 4-hydroxydebrisoquine (20.9 ± 3.0% dose) than did PMs (0.5 ± 0.2% dose; U = 36.0, P = 0.0014), and PMs excreted significantly more debrisoquine (33.9 ± 2.1% dose) than did EMs (8.3 ± 1.0% dose; U = 36.0, P = 0.0014) in this traditional phenotyping period (Mahgoub et al., 1977). Moreover, all EMs excreted 3,4-dehydrodebrisoquine as a metabolite of debrisoquine in urine (7.8 ± 2.2% dose), and three of four PMs also excreted this metabolite (0.9 ± 0.4% dose; U = 35.0, P = 0.0028). Table 2 shows that this pattern of metabolite excretion was mirrored in the 8- to 24-h urines also. 4-Hydroxydebrisoquine excretion (percentage of dose) was significantly greater in EMs (7.2 ± 1.4) than in PMs (0.2 ± 0.1; U = 36.0, P = 0.0014) and debrisoquine excretion (percentage of dose) was significantly greater in PMs (19.5 ± 5.3) than in EMs (3.7 ± 0.9; U = 36.0, P = 0.0014). The excretion of 3,4-dehydrodebrisoquine in the 8- to 24-h urine was 2.5 ± 0.9% of dose for EMs. Only one of four PMs excreted this metabolite in the 8- to 24-h urine. The interphenotype difference was barely statistically significant (U = 31.0, P = 0.025). Finally, Table 3 shows that, for the combined 0- to 24-h urine, debrisoquine metabolism displayed clear polymorphism based upon the excretion of debrisoquine itself (12.0 ± 1.5% of dose in EMs, 53.3 ± 7.0% of dose in PMs; U = 36.0, P = 0.0014), 4-hydroxydebrisoquine (28.1 ± 3.4% of dose in EMs, 0.7 ± 0.3% of dose in PMs; U = 36.0, P = 0.0014), and, indeed, for 3,4-dehydrodebrisoquine (10.3 ± 2.6% of dose in EMs, 1.0 ± 0.5% dose in PMs; U = 36.0, P = 0.0014). This observation adds weight to the proposition that 3,4-dehydrodebrisoquine is formed from 4-hydroxydebrisoquine. Of intense interest is the observation that one EM subject excreted inordinately large amounts of 3,4-dehydrodebrisoquine. EM subject 6 excreted 23.6% of dose as 3,4-dehydrodebrisoquine in 0 to 8 h and 27.6% in 0 to 24 h. However, another subject (3) excreted only 2.0% and 3.1% in 8 h and 24 h, respectively. These subjects are both homozygous for the CYP2D6*1 allele.
|
|
|
The excretion of 6- and 8-hydroxy-debrisoquine was also determined, after deconjugation with ß-glucuronidase. Unlike previous studies (Allen et al., 1975; Idle et al., 1979) that used packed GC columns, in our case, on a relatively short and nonpolar capillary column, these metabolites did not require further derivatization of the phenol group as methyl or trimethylsilyl derivatives. In all published gas chromatograms of the debrisoquine phenols (Allen et al., 1975; Idle et al., 1979), these metabolites eluted in the order 8-, 5-, 7-, and 6-hydroxy-debrisoquine. Although other peaks were observed to elute between the 8- and 6-hydroxy-debrisoquine peaks in our urine analyses, no attempt was made to assign these to 5- and 7-hydroxydebrisoquine, in the absence of authentic standards. The excretion of 6-hydroxydebrisoquine was found to be 1.8 ± 0.5 and 0.1 ± 0.1% of dose in EM 0- to 8-h and 8- to 24-h urine, respectively (Table 4). The excretion of 8-hydroxydebrisoquine was found to be 0.5 ± 0.2 and 0 ± 0% of dose in the EM 0- to 8-h and 8- to 24-h urine, respectively. Neither phenolic metabolite was detected in any urine from a PM volunteer. The total 6- and 8-hydroxy-debrisoquine excretion for EMs in 0 to 24 h was 2.4% of dose, which was significantly greater than in PMs (U = 34.0, P = 0.006).
|
Finally, the concentration in all urine samples of debrisoquine and 4-hydroxydebrisoquine, both before and after ß-glucuronidase treatment, was determined for 24 h. In all duplicate urine samples, we saw a small but consistent rise in these concentrations, which could not be explained by evaporation because the tubes were well sealed during incubation at 37°C. Moreover, concomitant elevation in 3,4-dehydrodebrisoquine was not observed. We interpret these findings as being due to conjugation of both debrisoquine and 4-hydroxydebrisoquine with glucuronic acid. In the case of debrisoquine, this must be as a ß-glucuronidase-labile N-glucuronide. With 4-hydroxydebrisoquine, an O4-glucuronide would be formed. Glucuronide formation from secondary alcohols, including benzylic alcohols (such as 4-hydroxydebrisoquine), has been known for many years (Williams, 1959). Neither of these metabolites has been described previously (Allen et al., 1975; Idle et al., 1979). Definitive identification of O4-glucuronide was achieved using UPLC-TOFMS, which enables exact mass determination (Fig. 4). Indeed, a 368.14 m/z ion, corresponding in Mr to the hydroxydebrisoquine O4-glucuronide, was found at 1.76 and 2.23 min of the chromatogram and, upon fragmentation, yielded a 192.1 m/z ion that corresponds to the Mr of hydroxylated debrisoquine. Although it is assumed that this is the O4-glucuronide, glucuronides derived from other minor hydroxylated metabolites, i.e., 6-, 7-, and 8-hydroxydebrisoquine, cannot be excluded by this analysis. In any case, debrisoquine glucuronide accounted for 0.4 ± 0.2, 1.5 ± 1.1, 0.3 ± 0.1, and 1.3 ± 0.8% of dose in EM 0- to 8-h, PM 0- to 8-h, EM 8- to 24-h, and PM 8- to 24-h urines, respectively. Interestingly, PMs excreted more of this glucuronide than did EMs, but the difference did not reach statistical significance. 4-Hydroxydebrisoquine glucuronide accounted for 2.5 ± 0.9, 0.1 ± 0.03, 1.5 ± 0.5, and 0.1 ± 0.1% of dose in EM 0- to 8-h, PM 0- to 8-h, EM 8- to 24-h, and PM 8- to 24-h urines, respectively. EMs excreted more of this glucuronide, in contrast to the debrisoquine glucuronide, and the difference was statistically significant in 0- to 8-h (U = 30.0, P = 0.032), 8- to 24-h (U = 32.5, P = 0.01), and 0- to 24-h (U = 33.0, P = 0.01) urines.
|
|
Formation of 3,4-Dehydrodebrisoquine in Vitro. Debrisoquine hemisulfate was incubated with both human and rat liver microsomes. Debrisoquine was converted to 4-hydroxydebrisoquine, as expected, with 6.6 and 2.1% substrate conversion in 24 h, by rat and human liver microsomes, respectively. However, debrisoquine was also converted to 3,4-dehydrodebrisoquine, with 6.3 and 4.4% substrate conversion in 24 h, by rat and human liver microsomes, respectively. Moreover, 4-hydroxydebrisoquine was converted to 3,4-dehydrodebrisoquine and a late running peak at 22.40 min, by both rat and human liver microsomes. The conversion of 4-hydroxydebrisoquine to 3,4-dehydrodebrisoquine in 24 h was 33 and 25.4% for rat and human liver microsomes, respectively. No metabolite was formed from boiled microsomes or in the absence of NADH or NADPH. The concentration of 3,4-dehydrodebrisoquine produced in these microsomal incubations with 4-hydroxydebrisoquine was sufficient, for both rat and human preparations, to obtain a full mass spectrum to confirm the identity of the 3,4-dehydrodebrisoquine peak (Fig. 6). The production of 3,4-dehydrodebrisoquine from 4-hydroxydebrisoquine was dependent upon the presence of microsomes, and depended on either NADH or NADPH, suggesting involvement of a non-P450 enzyme, such as a dehydratase.
|
) of percentage dose as debrisoquine/percentage dose as 4-hydroxydebrisoquine and a revised metabolic ratio of percentage dose as debrisoquine/(percentage dose as 4-hydroxydebrisoquine + percentage dose as 3,4-dehydrodebrisoquine). A strong and highly significant correlation was found (r = 0.969; P < 0.001) for which the revised metabolic ratio was 33% lower than the traditional metabolic ratio (paired Student's t test; t = 3.43, P = 0.0089). All homozygous EMs had a revised metabolic ratio <1. For PMs, who did not excrete much 3,4-dehydrodebrisoquine (0.9%), the effect on metabolic ratio was nevertheless considerable, with traditional and revised metabolic ratios of 111 ± 37 and 66.3 ± 43, respectively [paired Student's t test; t = 1.71, P = 0.19 (NS)].
|
There was no statistically significant correlation found between debrisoquine 4-hydroxylation and 3,4-dehydrodebrisoquine formation in the nine subjects homozygous for CYP2D6*1 (Fig. 7B). This is not surprising, since the formation of 3,4-dehydrodebrisoquine requires not only initial CYP2D6 activity to produce 4-hydroxydebrisoquine, but also a second enzyme activity, independent of CYP2D6, to desaturate the 3,4-carbon-carbon bond. This dehydratase activity may vary from person to person.
The question remains whether or not failure to account for the formation of 3,4-dehydrodebrisoquine in debrisoquine phenotyping tests using 0- to 8-h urine collections might have led to erroneous assignment of CYP2D6 phenotype. Many thousand of such tests have been performed, and it is not possible to answer this question here, based upon the study of only nine homozygous EMs, four PMs, and no heterozygotes. The consistent reduction by 33% of the metabolic ratio in our volunteers suggests that this may be a possibility. One way to answer this question would be in a large population study of debrisoquine metabolism versus CYP2D6 genotype.
|
Discussion |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
), and 0.6% and 0.4% of dose of 6-hydroxydebrisoquine and 8-hydroxydebrisoquine, respectively, were reported in the 0- to 8-h urine of six Nigerian PMs (Mbanefo et al., 1980). It is possible that our findings reflect interethnic variation; however, they are more likely to be due to the deployment in this study of the highly specific analytical tool of GCMS.
A second novel finding of the current study is the observation that both debrisoquine and 4-hydroxydebrisoquine appear to form ß-glucuronidase-labile glucuronides, as confirmed by UPLC-TOFMS analysis of urine metabolites. Furthermore, the pattern of glucuronide excretion was highly genotype-dependent, reducing the likelihood that it was due to artifact. A metabolic ratio constructed as percentage of dose eliminated as debrisoquine glucuronide/percentage of dose eliminated as 4-hydroxydebrisoquine glucuronide had a mean value of 0.2 for EMs and 15 for PMs.
Further studies revealed that 3,4-dehydrodebrisoquine is produced from 4-hydroxydebrisoquine, both in vivo and in vitro. Its formation in vitro, using both human and rat liver microsomes, was dependent on either NADPH or NADH, and was, therefore, probably not mediated by a cytochrome P450. Therefore, not only is 3,4-dehydrodebrisoquine a major and variable metabolite, but its production may also depend upon the activity of a secondary enzyme system, possibly a dehydratase. Thus, the intergenotype difference in excretion of 3,4-dehydrodebrisoquine may be largely due to the CYP2D6 polymorphism, whereas the considerable intragenotype variation in production of this metabolite may arise from intersubject variation in the 4-hydroxydebrisoquine desaturation reaction. A subject who has a high 4-hydroxydebrisoquine dehydratase activity may excrete relatively little 4-hydroxydebrisoquine, compared with a subject (with the same CYP2D6 genotype) who has a low 4-hydroxydebrisoquine dehydratase activity and thus excretes relatively larger amounts of 4-hydroxydebrisoquine. Owing to the contribution of CYP1A1 to debrisoquine 4-hydroxylation (Granvil et al., 2002), coupled with the variation introduced by glucuronidation of debrisoquine and 4-hydroxydebrisoquine, described herein, we can begin to comprehend why the debrisoquine metabolic ratio is so highly variable within a single genotype such as CYP2D6*1/CYP2D6*1. The discovery in recent years of a large number (close to 90) of CYP2D6 variant alleles (http://www.imm.ki.se/cypalleles/cyp2d6.htm) has explained the existence of the CYP2D6 UM, EM, IM, and PM phenotypes and some of the ethnic variability associated with them. However, these genetic data have failed to explain why the archetypal drug debrisoquine is metabolized quite differently in persons with identical genotypes. Apart from other influential constitutional and environmental factors, such as age, diet, lifestyle, and liver and kidney blood flow, it is the operation of additional mechanisms, both metabolic and nonmetabolic, that surely dictate the pattern of urinary metabolites of a drug such as debrisoquine. The effect of a second enzyme system on debrisoquine metabolism was not anticipated because debrisoquine has been very well studied with respect to the P450 system. It was not immediately apparent that the subject of investigation should be 4-hydroxydebrisoquine, in addition to debrisoquine, and only then was the participation of a second enzyme system, the 4-hydroxydebrisoquine 3,4-dehydratase, uncovered. The metabolism of debrisoquine in relation to the CYP2D6 polymorphism is now more completely understood.
Since the discovery of the debrisoquine 4-hydroxylation polymorphism almost three decades ago (Mahgoub et al., 1977), there have appeared a host of reports that correlated some aspect of drug disposition with numerical values of the debrisoquine metabolic ratio, with greater or lesser success; for example, with nortriptyline (Mellstrom et al., 1981; Woolhouse et al., 1984), amitriptyline (Mellstrom et al., 1986), phenformin (Oates et al., 1982), sparteine (Inaba et al., 1983), tolbutamide (Peart et al., 1987), and ifosfamide (Philip et al., 1988). It is now clear that the classical metabolic ratio, as used by these various investigators, may not have been the best tool with which to correlate the metabolism of a second substrate. However, these types of investigations have been superseded by genotype-based studies, which have their own advantages and disadvantages.
The ring-opened metabolites of debrisoquine ([7] and [8] in Fig. 1A) are believed to represent a significant proportion of debrisoquine metabolism in rats, dogs, and humans (Allen et al., 1975, 1976; Eiermann et al., 1998). It was first proposed in 1976 that these two ring-opened metabolites were formed as a consequence of 1- and 3-hydroxylation of debrisoquine (Allen et al., 1976). More recently, this concept has been reinforced after the detection of the two ring-opened metabolites using LCMS (Eiermann et al., 1998). These authors also reported that the excretion of these two metabolites correlated highly, across a broad range of metabolic capacity, with debrisoquine 4-hydroxylation, as judged by the traditional metabolic ratio (debrisoquine/4-hydroxydebrisoquine). However, no one has ever performed an experiment that demonstrated 1- or 3-hydroxylation of debrisoquine; simply detection of metabolites [7] and [8] (Fig. 1A) is taken as evidence for these hydroxylation reactions. An alternative scenario, that is based both upon our observation of the formation of 3,4-dehydrodebrisoquine from 4-hydroxydebrisoquine and upon some older literature, is proposed. Examination of the various published correlations between debrisoquine 4-hydroxylation and individual phenol metabolite formation, together with the data presented herein, is enlightening. The data presented in Tables 3 and 4 permit correlations to be made, within the homozygous CYP2D6*1 genotype, among 4-, 6-, and 8-hydroxylation of debrisoquine. The correlation coefficients (r) for 4-hydroxylation versus 6-hydroxylation, 4-hydroxylation versus 8-hydroxylation, and 6-hydroxylation versus 8-hydroxylation are 0.059, 0.055, and 0.993, respectively. 4-Hydroxylation does not correlate with either 6- or 8-hydroxylation, even though all three reactions are CYP2D6-dependent (see Tables 1, 2, 3, 4). Moreover, published data from 10 Nigerian EMs (Mbanefo et al., 1980) permit correlations to be made among 4-, 5-, 6-, 7-, and 8-hydroxylations of debrisoquine. The correlations with debrisoquine 4-hydroxylation and 6-, 7-, or 8-hydroxylation are virtually zero, and there is a weakly significant correlation with 5-hydroxylation (r = 0.642, P < 0.05). Therefore, there is no reason to expect 4-hydroxylation to correlate with 1- or 3-hydroxylation, should these latter reactions occur.
We propose that debrisoquine 4-hydroxylation also leads to 3,4-dehydrodebrisoquine in considerable amounts; as high as 27.6% of the dose was obtained in one subject. In addition, the possibility exists that 3,4-dehydrodebrisoquine not only is in equilibrium with the 3-dehydro-isoquinolinium compound in Fig. 4, but also rearranges to the 1-dehydro-isoquinolinium compound, because of the resonance between the 1,2-double bond and the aromatic system and the guanidine lone pairs. As shown in Fig. 1C, the desaturation of 4-hydroxydebrisoquine, subsequent rearrangement of the 3,4-dehydrodebrisoquine, and reaction with hydroxyl ions leads to 1- and 3-hydroxydebrisoquine, which then becomes further oxidized to yield, ultimately, the two ring-opened metabolites. In this proposed scheme, all but the phenolic metabolites arise via 4-hydroxylation of debrisoquine by CYP2D6.
In summary, this study has shown that debrisoquine forms a novel metabolite, 3,4-dehydrodebrisoquine, both in vivo and in vitro, from 4-hydroxydebrisoquine by an unknown dehydratase activity. The existence of this new metabolite has several consequences. First, the total flux of 4-hydroxylation is, on average, at least 30% greater in CYP2D6*1 homozygotes than previously thought. This has the effect of lowering the metabolic ratio, but does so in a predictable manner. In turn, this may have implications for studies in which the metabolic ratio is correlated with the metabolism of a second drug or, indeed, is used in case-control studies of cancer (Ayesh et al., 1984), Parkinson's disease (Barbeau et al., 1985), and other clinical entities. Second, the process of CYP2D6-mediated hydroxylation followed by desaturation may not be restricted to debrisoquine. It is possible that this double reaction plays a role in sparteine metabolism. Finally, we propose a metabolic scheme whereby all the major metabolites of debrisoquine arise via the 4-hydroxylation pathway. Neither the extent of metabolism nor the ratio of metabolites to each other can be readily predicted from CYP2D6 genotyping studies alone. Genotype determines an average metabolic performance, but the variability about this mean may be due to a range of factors, including alternative non-CYP2D6 metabolic pathways.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Supported by the National Cancer Institute Intramural Research Program (F.J.G.). J.R.I. is grateful to U.S. Smokeless Tobacco Company for a grant for collaborative research.
ABBREVIATIONS: MR, metabolic rate; EM, extensive metabolizer; IM, intermediate metabolizer; P450, cytochrome P450; PM, poor metabolizer; UM, ultrarapid metabolizer; GCMS, gas chromatography-mass spectrometry; LC-MS/MS, liquid chromatography-tandem mass spectrometry; UPLC-TOFMS, ultra-performance liquid chromatography-coupled time-of-flight mass spectrometry.
Address correspondence to: Professor Jeffrey R. Idle, Institute of Pharmacology, 1st Faculty of Medicine, Charles University, Albertov 4, 128 00 Prague 2, Czech Republic. E-mail: jeff{at}cyp2d6.com
|
References |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  R. E. Savage, A. N. Tyler, X.-S. Miao, and T. C. K. Chan Identification of a Novel Glucosylsulfate Conjugate as a Metabolite of 3,4-Dihydro-2,2-dimethyl-2H-naphtho[1,2-b]pyran-5,6-dione (ARQ 501, {beta}-Lapachone) in Mammals Drug Metab. Dispos., April 1, 2008; 36(4): 753 - 758. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
 |
 |
|
 |
 |
 |
5.6 to 
