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

This Article Abstract Full Text (PDF) Data Supplement All Versions of this Article: jpet.105.099887v1 318/3/992    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 Citation Map 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 HighWire Citing Articles via Google Scholar Google Scholar Articles by Kent, U. M. Articles by Hollenberg, P. F. Search for Related Content PubMed PubMed Citation Articles by Kent, U. M. Articles by Hollenberg, P. F. Pubmed/NCBI databases Compound via MeSH Substance via MeSH

METABOLISM, TRANSPORT, AND PHARMACOGENOMICS

Metabolism of Bergamottin by Cytochromes P450 2B6 and 3A5

Department of Pharmacology, University of Michigan, Ann Arbor, Michigan (U.M.K., H.L., K.R.N., P.F.H.); and Molecular Research Institute, Mountain View, California (D.L.H.)

Received December 20, 2005; accepted June 15, 2006.

	Abstract

Top Abstract Materials and Methods Results Discussion References

 
Cytochromes P450 (P450) 2B6 and 3A5 are inactivated by bergamottin (BG). P450 2B6 metabolized BG primarily to M3 and M4 and one minor metabolite (M1). The metabolites were analyzed, and the data indicated that M1 was bergaptol, M3 was 5'-OH-BG, and M4 was a mixture of 6'- and 7'-OH-BG. Because 6'- and 7'-OH-BG were the primary metabolites, it suggested that P450 2B6 preferentially oxidized the geranyloxy chain of BG. Metabolism of BG by P450 3A5 resulted in three major metabolites: [bergaptol, M3 (5'-OH-BG), and M5 (2'-OH-BG)], and two minor metabolites [M2 (6',7'-dihydroxy-BG) and M4 (6'- and 7'-OH-BG)]. Because bergaptol was the most abundant metabolite formed, it suggested that P450 3A5 metabolized BG mainly by cleaving the geranyl-oxy chain. Molecular modeling studies confirmed that docking of BG in the P450 2B6 active site favors oxidation in the terminal region of the geranyl-oxy chain, whereas positioning the 2'-carbon of BG nearest the heme iron is preferred by P450 3A5. Glutathione (GSH)-BG conjugates were formed by both P450. Each enzyme predominantly formed conjugates with m/z values of 662. Tandem mass spectrometry analysis of the GSH conjugates indicated that the oxidation forming a reactive intermediate occurred on the furan moiety of BG, presumably through the initial formation of an epoxide at the furan double bond. The data indicate that oxidation of the geranyl-oxy chain resulted in the formation of stable metabolites of BG, whereas oxidation of the furan ring produced reactive intermediates that may be responsible for binding to and inactivating P450 2B6 and 3A4.

Although the levels of P450 2B6 in liver are generally low (<2%) in the noninduced state, this isoform is responsible for the metabolism of >4% of drugs routinely used by patients such as buproprion, efaviranz, methadone, ifosfamide, and cyclophosphamide (Faucette et al., 2000; Huang et al., 2000; Ward et al., 2003; Gerber et al., 2004). BG, one of the major components in grapefruit juice, has been demonstrated to be, in part, responsible for the increased bioavailability of certain drugs in what has become known as the "grapefruit juice effect" (Bailey et al., 2000). BG and its hydroxylated product 6',7'-dihydroxy BG (DHBG) are mechanism-based inactivators of P450 3A4 (He at al., 1998). The ability of BG and DHBG to inactivate P450 3A4 is thought to be the major reason for the grapefruit juice-induced drug interactions that have been observed clinically (Paine et al., 2004). Recently, we have reported that BG also inactivated P450 3A5 and 2B6, further underscoring the potential problems associated with this component and its effect on the metabolism of clinically important drugs (Bumpus et al., 2005; Lin et al., 2005). Preliminary studies suggested that BG metabolism by P450 3A5 was more complex than that observed with P450 2B6. P450 3A5 generated five metabolites of BG that exhibited absorbencies at 310 nm, indicative of furanocoumarin-containing products. Two major metabolites were observed following metabolism of BG by P450 2B6.

Considerable information concerning the metabolism of furanocoumarins by P450 is available. 8-Methoxypsoralen is similar in structure to BG except that BG contains a 5-geranyl-oxy chain instead of a methoxy group at the 8-position. 8-Methoxypsoralen has been shown to undergo O-demethylation, hydroxylation of the phenyl ring, and oxidation of the furan ring with subsequent hydrolysis of the lactone ester (Mays et al., 1990). Subsequent studies established that one of the initial steps in the metabolism of furanocoumarins could involve oxidation of the furan double bond, leading to the formation of a furanoepoxide intermediate capable of combining either with glutathione to form the GS conjugate or with macromolecules (Koenigs and Trager, 1998). This route of metabolism by P450 2B1 and 2A6 resulted in the formation of dihydrodiol products (Koenigs and Trager, 1998; Lightning et al., 2000). In addition, 6',7'-epoxy-BG, a stable epoxide found in grapefruit peel, was shown to inhibit the activity of P450 3A4 (Wangensteen et al., 2003). This inhibitory effect was seen only with compounds containing the intact furanocoumarin core structure. Similar conclusions were drawn from studies using the furanopyridine L-754,394 that indicated that P450 3A4-dependent metabolism of this compound occurred via oxidation of the furan ring to an epoxide with the subsequent formation of the dihydrodiol (Sahali-Sahly et al., 1996).

The metabolism of BG by P450 3A4/3A5 and 2B6 results in a loss in the enzymatic activity of these P450 through a combination of modification of the heme moiety and covalent modification of the apoprotein (He et al., 1998; Lin et al., 2005). Bioactivation of BG by these P450 generated reactive intermediates that inactivated these enzymes. Small sulfhydryl-containing compounds such as glutathione and N-acetylcysteine have been used to trap reactive electrophiles generated by P450 for subsequent structural analysis of the S-substituted adducts using mass spectrometry (Jin and Baillie, 1997; Evans et al., 2004).

This investigation was designed to analyze the products that were formed by P450 3A5 and 2B6 during the metabolism of BG. Determination of the structure(s) of the BG-GS conjugate(s) in conjunction with the metabolite studies could then be used to elucidate the catalytic pathway leading to the production of the reactive BG intermediate responsible for the mechanism-based inactivation of P450 2B6 and 3A5.

	Materials and Methods

Top Abstract Materials and Methods Results Discussion References

 
Chemicals. Glutathione, catalase, NADPH, L--dilauroyl-phosphatidylcholine, L--dioleyl-sn-glycero-3-phosphatidylcholine, and L--phosphatidylserine were purchased from Sigma-Aldrich (St. Louis, MO). BG and bergaptol were obtained from Indofine Chemical Co., Inc. (Somerville, NJ). [¹⁴C]BG (55 mCi/mmol; 99% purity) was obtained from American Radiolabeled Chemicals (St. Louis, MO). DHBG was a gift from the Florida Department of Citrus (Lakeland, FL). 7-Ethoxy-4-(trifluoromethyl)coumarin was from Invitrogen (Carlsbad, CA). Accubond solid-phase extraction C18 cartridges were from Agilent Technologies (Palo Alto, CA). All other chemicals and solvents were of the highest purity available from commercial sources.

Purification of Enzymes. Expression plasmids for P450 2B6 and 3A5 were generous gifts from Dr. James R. Halpert (Environmental Toxicology, University of Texas Medical Branch, Galveston, TX). The enzymes were expressed in Escherichia coli Topp 3 cells and purified using Ni²⁺-agarose affinity chromatography as described previously (Domanski et al., 2001; Scott et al., 2001). Reductase was expressed and purified as described by Hanna et al. (1998). Cytochrome b₅ was purified from liver microsomes obtained from Long-Evans rats as described previously (Lin et al., 2002).

Enzyme Assays. P450 2B6 (1 µM) was reconstituted with 2 µM reductase on ice for 45 min. The primary reaction mixture contained catalase (110 units), 10 mM GSH, 1 mM ascorbate, 10 µM BG, and 100 mM potassium phosphate buffer, pH 7.5, in a total volume of 1 ml. Ascorbate was included to stabilize any catechols that may be formed. P450 3A5 (1 µM) was reconstituted with 2 µM reductase, 1 µM cytochrome b₅, and 60 µg of a 1:1:1 mixture of L--dilauroylphosphatidyl-choline, L--dioleyl-sn-glycero-3-phosphatidylcholine, and L--phosphatidylserine on ice for 45 min. The P450 3A5 reaction mixture received 110 units of catalase, 30 mM MgCl₂, 0.5 mM EDTA, 10 mM GSH, and 1 mM ascorbate in a final volume of 1 ml of 50 mM HEPES buffer, pH 7.5, containing 20% glycerol. Each of the reconstitution mixtures was divided into two samples and incubated at 37°C for 5 min. One sample received 1 mM NADPH, and an equal volume of water was added to the control sample. The mixtures were then incubated for 30 min at 37°C.

HPLC Analysis. In the initial experiment, the samples were reconstituted and incubated with radiolabeled BG in the presence or absence of NADPH. After incubating the samples for 30 min at 37°C, the resulting metabolites were extracted as described below. For all subsequent metabolite analyses, the samples were incubated for 30 min at 37°C with unlabeled BG and then spiked with 10 µM internal standard 6-methoxypsoralen and immediately extracted twice with 2 to 3 ml of nitrogen-purged methylene chloride. The organic phases were pooled and dried under N₂ gas. The dried samples were resuspended in 100 µl of 50% solvent B (solvent A: 0.1% acetic acid in water; solvent B: 0.1% acetic acid in acetonitrile). The metabolites and BG were resolved on a C8 column (5 µm, 2.1 x 150 mm, Zorbax XDB; MAC-MOD, Chadds Ford, PA) equilibrated with 35% solvent B. Samples were applied and washed with 35% solvent B for 8 min. The percentage of solvent B was raised to 45% over the next 17 min followed by an increase in the percentage of solvent B to 80% by 55 min at a flow of 1 ml/min.

LC-ESI-MS/MS Analysis of Metabolites. The same column and gradient conditions were used as described for the HPLC analysis. The column effluent was split, and 0.3 ml/min was diverted into the LCQ mass analyzer (Thermo Electron Corporation (Waltham, MA). The ESI conditions were as follows: sheath gas, 90 arbitrary units; auxiliary gas, 30 arbitrary units; capillary temperature, 170°C; and spray voltage, 4.5 V. Data were acquired in positive mode using the Excalibur software package (Thermo Electron Corporation) with one full scan from 100 to 1000 m/z followed by two data dependent scans of the second and third most intense ion. Alternatively, samples were analyzed in positive APCI mode using an LTQ mass analyzer equipped with a photodiode array detector (Thermo Electron Corporation) using the same chromatography conditions as described for the LCQ system. The analyzer conditions were optimized with BG, and the settings were as follows: APCI vaporizer temperature, 450°C; sheath gas, 55 arbitrary units; auxiliary gas, 15 arbitrary units; capillary temperature, 225°C; source voltage, 6 kV; source current, 5 µA; and tube lens offset, 51 V. Parent ion monitoring was carried out in positive APCI mode on a TSQ mass analyzer (Thermo Electron Corporation). The liquid chromatography conditions were as described for the LCQ. The mass analyzer settings were as follows: sheath gas, 40 arbitrary units; auxiliary gas, 10 arbitrary units; capillary temperature, 350°C; capillary offset, 35; discharge current, 4; and vaporizer temperature, 500°C.

LC-ESI-MS Analysis of GSH Conjugates. GSH conjugates were isolated after adding 60 µl of a 10% solution of trifluoroacetic acid in water to each 1 ml of sample and then applying the sample to a 1-ml Accubond solid-phase extraction ODS C18 cartridge previously washed with 1 ml of methanol followed by 2 ml of water. After the samples were loaded, the cartridges were washed sequentially with 2 ml of water, 2 ml of methanol, and 0.3 ml of acetonitrile. The organic phases were pooled and dried under N₂ gas. The dried samples were resuspended in 150 µl of 50% acetonitrile and 0.1% acetic acid. Samples were analyzed on a C18 reverse-phase column (3 µm, 100 x 4.6 mm, Luna; Phenomenex, Torrance, CA) using a gradient of 20 to 30% solvent B in 5 min followed by a linear increase to 40% solvent B by 15 min and to 90% solvent B by 30 min at a flow of 0.3 ml/min. The column effluent was directed into an LCQ mass analyzer. The ESI conditions were as follows: sheath gas, 90 arbitrary units; auxiliary gas, 30 arbitrary units; capillary temperature, 170°C; and spray voltage, 30 V. Data were acquired in positive mode using the Excalibur software package (Thermo Electron Corporation) with one full scan followed by two data-dependent scans of the most intense and the second most intense ions.

NMR Analysis of Metabolites. Approximately 20 to 30 nmol of each P450 was reconstituted with reductase and incubated with BG and NADPH as described above. The metabolites were extracted with nitrogen-purged CH₂Cl₂ and separated on a C8 reverse-phase column under the same conditions used for the HPLC analyses. Each injection contained the metabolites of BG derived from approximately 5 nmol of incubation mixture. M2 to M5 were collected manually and diluted with water to twice their initial volume. The metabolites were transferred into deuterated solvent by applying each of the diluted metabolite fractions to a 1-ml C8 SepPak solidphase extraction cartridge (Waters, Milford, MA) previously hydrated with methanol and equilibrated with water. After each diluted metabolite was loaded onto the C8 cartridge, the cartridge was washed with 2 ml of D₂O. The metabolites were then eluted with 1 ml of CD₃OD and concentrated under N₂ to approximately 300 µl. ¹H NMR spectra were collected on a Brucker Avance DRX 500 NMR (Brucker Instruments, Billerica, MA). Additional ¹H NMR spectra were collected after thoroughly drying the metabolite samples and resuspending them in CD₃Cl.

Molecular Modeling of P450 2B6 and 3A5 and BG Docking Studies. The recent solution of the crystal structures of P450 3A4 and 2B4 provide a basis for responsible construction of models of P450 2B6 and 3A5 based on high-sequence identity templates. P450 2B6 has a sequence identity of 72% with respect to 2B4, whereas P450 3A5 has a sequence identity of 80% with 3A4. Sequence alignments were performed using BLOSUM-30 scoring matrices and a gap penalty of 10 using ClustalW (Thompson and Higgins, 1994). Given the high-sequence identity and the lack of significant insertions of one sequence alignment with respect to the other, no manual realignment was necessary. Supplemental Figures 1 and 2 show the alignments with color coding of the secondary structural elements from the Dictionary of Protein Secondary Structure (Kabasch and Sander, 1983) scoring as reported in the crystal structures of P450 2B4 and 3A4 templates. Labels as to the positions of the helices in previously reported analyses of bacterial P450 and P450 2C5 templates are indicated. Underlined regions indicate conserved secondary structural regions deduced from past alignments including bacterial and mammalian P450. It is clear that the precise regions of structurally conserved helices and -sheets in P450 3A4 and 2B4 have a close correspondence with crystal structures of bacterial origins but deviate in the precise location and extent of those regions.

View larger version (27K): [in this window] [in a new window]   Fig. 1. HPLC analysis of metabolites of BG. P450 2B6 or 3A5 was incubated with BG as described under Materials and Methods. The samples were extracted with methylene chloride, and BG as well as the resulting metabolites were resolved by reverse-phase chromatography. A, HPLC chromatogram of BG and BG-derived metabolites generated by P450 2B6 monitored at 310 nm for absorbance of the psoralen ring (solid line) or for the 14C-labeled geranyl-oxy moiety of BG (dashed line). B, HPLC chromatogram of BG and BG-derived metabolites generated by P450 3A5 monitored at 310 nm for absorbance of the psoralen ring (solid line) or for the 14C-labeled geranyl-oxy moiety of BG (dashed line).

View larger version (28K): [in this window] [in a new window]   Fig. 2. APCI-LC-MS analysis of BG metabolites generated by P450 2B6 in the reconstituted monooxygenase system. P450 2B6 was reconstituted with reductase and incubated with BG in the presence or absence of NADPH. Metabolites were extracted and analyzed as described under Materials and Methods. A, HPLC chromatogram of BG and BG-derived metabolites that absorb at 310 nm. The insets show the photodiode array spectrum between 260 and 350 nm for each peak. BG eluted at 50.05 min, and the internal standard 6-methoxypsoralen (6-MOP) eluted at 17.82 min. B, extracted ion chromatogram at m/z 203 of the photodiode array profile shown in A. C, m/z of the ions corresponding to the peaks at 17.82 min (6-MOP), 33.43 min (M3), 34.64 min (M3), and 50.05 min (BG).

The initial models were then energy minimized using limited energy minimization (400 steps of steepest descents; 200 steps conjugate-gradient) merely to remove any bad contacts following rotamer exploration. Waters were then introduced using in-house software that places waters based on topological and energy metrics (Zhao et al., 1996). The resulting initial structure with modeled waters was then energy minimized for 150 steps of steepest descents minimization followed by 3000 steps of conjugate gradient minimization. The structure was then equilibrated at 300K for 600 ps using radial dielectric screening and a potential truncation of 12 Å. The final coordinates following molecular dynamic thermalization were then energy minimized for 2000 steps before use of the model in docking.

Prosa 2003 (Insoft Oy, Oulu, Finland) was used to assess the quality of the model/-fold commensurate with sequence (Sippl, 1993). The P450 3A5 model following equilibration and annealing had a normalized Prosa score of 0.88 compared with the P450 3A4 (template of 0.94), whereas the normalized Z-score was 0.91 before annealing. The Prosa scores indicate that use of limited equilibration resulted in no serious degradation of model folds but does improve facets such as the hydrogen bonding in the region of the binding site before model use in docking and molecular dynamics (Park and Harris 2003). Inspection of Prosa plots revealed no region of the model with positive Prosa energies. The P450 2B6 model had a normalized Z-score of 0.88 compared with the P450 2B4 crystal structure Z-score of 0.82. The two models are seen to have normalized Z-scores comparable with the templates from which they were derived. The summary assessments shown in Table 1 in the Appendix indicate that the deviations from "standard" bond lengths/angles/dihedrals do not result in a model that has substantial regions of geometric violations compared with high-quality crystal structures. This is reflected in the Procheck scores, which by virtue of having values more than –0.50 are deemed adequate. Comparison values for all measures are also indicated for the templates from which the models were derived.

Energy based docking was performed using Autodock3 (Morris et al., 1998), ICM (Molsoft Inc., San Diego, CA), and Glide (Schrodinger Inc., Portland, OR) docking modules to verify that similar lowest energy docked configurations were obtained irrespective of differences in configuration sampling algorithms and potential function/scoring functions used. Autodock uses a classical coulomb + 6-9-12 type of interaction, and in this instance a charge model of the heme (resting state or oxyferryl/compound I) was derived from density functional theoretical calculations of the electrostatic potential. ICM uses an ECEPP/4 force field supplemented by van der Waals, hydrogen bonding, electrostatic, and hydrophobic ligand/receptor interaction terms. Glide makes use of a Chemscore-like function in addition to classical interaction terms. Glide uses a series of hierarchical filters based on a combination of scoring functions based on active site and ligand shape, OPLS-AA potential function/receptor-flexible ligand Monte-Carlo docking followed by application of a Chemscore function (Friesner at al., 2004). Autodock3 protocols were as reported previously (Park and Harris, 2003), whereas in the case of ICM docking Monte Carlo based docking was performed into a region of the model within 12 Å of the model binding sites. Glide 4.02 docking was performed into a 12-Å box on the distal side of the heme. The lowest 100-energy configurations were saved from each docking study to assess the common low-energy compound I-BG interaction motifs. Molecular dynamics protocols for equilibration of docked configurations are as described previously (Park and Harris, 2003; Harris et al., 2004).

	Results

Top Abstract Materials and Methods Results Discussion References

 
HPLC Analysis of Metabolites of BG. P450 2B6 and 3A5 were incubated with BG that was radiolabeled on the C1 carbon of the geranyl-oxy chain. HPLC analysis of control samples incubated in the absence of NADPH revealed only one peak corresponding to BG that eluted at 50 min. This peak displayed an absorbance maximum at 310 nm indicative of the furanocoumarin moiety and also contained radioactivity due to the presence of the geranyl-oxy chain (data not shown). Incubation of BG with P450 2B6 in the presence NADPH resulted in the appearance of two major metabolites that both exhibited absorbencies at 310 nm as well as radiolabel and eluted at 34 min (M3) and 36 min (M4) (Fig. 1A). A minor third metabolite (M1) eluted at 10 min and was characterized by an absorbance at 310 nm indicative of an intact furanocoumarin moiety but an absence of counts, suggesting that this product was devoid of the geranyl-oxy chain. An additional smaller peak that also exhibited an absorbance at 310 nm without radioactivity eluted at 47 min. The BG metabolite profile obtained from P450 3A5 samples was more complex and exhibited five metabolites that eluted at 10 min (M1), 16 min (M2), 34 min (M3), 36 min (M4), and 39 min (M5). The amount of M1 that was generated by P450 3A5 was approximately 5-fold higher than that obtained from incubations with P450 2B6. Again, no counts were associated with M1, suggesting that metabolism of BG had resulted in the removal of the labeled geranyl-oxy moiety. Previous studies that compared the retention times under the same HPLC conditions of the authentic standards bergaptol and DHBG with the BG metabolite elution profile suggested that M1 and M2 may be bergaptol and DHBG, respectively. However, conclusive evidence as to the identity of M1 to M5 had not been obtained at this point. With all of the samples incubated with radiolabeled BG, an additional peak was seen with an absorbance at 310 nm that eluted at 23 min. This peak was only present in samples incubated with radiolabeled BG and NADPH and was not seen in samples incubated with nonradioactive BG in the presence of NADPH. Presumably this product was derived from metabolism of a radiolabeled breakdown product in the commercially obtained BG stock solution that occurred after long-term storage. Therefore, it was not analyzed further.

LC-APCI-MS and MS/MS Analysis of Metabolites Generated by Incubating P450 2B6 with BG. The data in Fig. 2 depict the HPLC chromatogram of metabolites from P450 2B6 samples incubated with BG in the presence of NADPH (Fig. 2A). The diode array spectra associated with the internal standard (6-methoxypsoralen), M3, M4, and BG are shown as insets to Fig. 2A. In the absence of NADPH, only BG and 6-methoxypsoralen were observed using diode array detection (data not shown). The total ion chromatogram traces for both the samples (incubated either in the absence or presence of NADPH) seemed similar and did not yield information as to the elution times of any metabolites (data not shown). If, however, an extracted ion chromatogram (XIC) with an [M + 1] of 203 for the 5-hydroxypsoralen moiety was monitored, the chromatogram depicted in Fig. 2B was observed. In the absence of NADPH, only the peaks eluting at 18.3 min (6-methoxypsoralen; m/z of 218) and at 50.2 min (BG; m/z of 339) were detected (data not shown). Metabolites of BG with m/z of 355 (M4; 34.4 min) and m/z of 357 (M3; 33.1 min) were observed in the presence of NADPH (Fig. 2C). In addition to these species, another product with an m/z of 203, suggesting the removal of the geranyl-oxy chain, was seen at 9.7 min (M1). Most of the parent ions were also observed as species with an increase in mass of 41 that was derived from the acetonitrile in the solvent. Because the metabolites could only be detected in the XIC at [M + 1] of 203, precursor ion monitoring was carried out on a triple quadrupole instrument to confirm the mass of the parent compound for each metabolite (data not shown). The subsequent MS/MS analysis of the NADPH-dependent metabolites is shown in Table 1. The mass increase of 16 Da for the most abundant metabolite, M4, indicated that this product arose from an insertion of one oxygen, presumably into one of the double bonds of BG. Tandem mass spectrometry of the M4 metabolite with an m/z of 354.8 indicated that the geranyloxy chain was the site of modification, because a strong signal for the molecular ion of the unmodified 5-hydroxypsoralen (m/z of 203) was observed. Metabolite samples for ¹H NMR analysis were recovered in CD₃OD, and the initial spectra were taken in this solvent to observe the aromatic region of the compounds. Because of interfering signals from residual water and methanol, the alkene region of the spectrum could not be completely interpreted. Therefore, the samples were dried thoroughly under N₂, resuspended in CD₃Cl, and reanalyzed. The combined ¹H NMR data obtained for M4 confirmed that the furanocoumarin moiety was unaffected. The signals in the aromatic region for M4 (: 8.29, 7.81, 7.25, 7.16, and 6.31 ppm) were the similar, and the peaks integrated to the expected number of protons as those seen in the ¹H NMR spectrum of authentic BG (: 8.3, 7.8, 7.2, 7.1, and 6.3 ppm) (Table 2). The signals originating from the protons located on the furanocoumarin ring and the geranyl chain of BG as well as the signal splitting when it was clearly observable is shown in Table 2. The ChemNMR ¹H Estimation software package from ChemDraw version 10.0 (CambridgeSoft Corporation, Cambridge, MA) was used to predict the ¹H NMR spectra of the metabolites. Comparisons of the metabolite spectra to spectra of authentic BG in conjunction with the ChemNMR predictions were used to arrive at the possible structures for the M3 to M5 metabolites (Table 2). The spectra of M4 most closely fit the estimations obtained for the ¹H NMR spectra of 6'-OH-BG and 7'-OH-BG. The proton signal originating from the 6'-location of BG at 5.06 ppm was missing and the significant shifts for the protons of the 6'-carbon to 3.80 and 1.05 ppm and the 8'- and 9'-methyl signals to 0.74 and 0.94 as well as the appearance of new 7'-signals at 1.05 and 3.65 ppm (OH) suggested a modification at or near the 6'- and the 7'-carbons on the geranyl chain of BG. Integrating all of the peaks and comparing the spectra obtained for M4 to the theoretical spectra derived for a 6'-OH and a 7'-OH product of BG indicated that M4 was a mixture of both.

The mass increase of 18.3 mass units to generate the 357.2 (M3) species from BG suggested that this metabolite arose from a hydroxylation of one of the three double bonds present in the BG molecule. The diode array spectrum of M3 as well as the mass spectrum (presence of the m/z 203) indicated that the 5-hydroxypsoralen ring of this metabolite was intact, suggesting again that the hydroxylation occurred on the geranyloxy chain. However, the tandem mass spectrum of M3 was complex and indicated the presence of two different hydroxylated species. The major metabolite exhibited product ions of 203, 215, and 339. The second, albeit much lower abundant metabolite with the same m/z of 357.3 that coeluted with M3 displayed unique ions with an m/z of 221 and 235 that might be observed if the hydroxylation had occurred on the furanocoumarin moiety of BG (data not shown). M3 was also analyzed by ¹H NMR in the same manner as M4. The furanocoumarin moiety of M3 was intact and the proton signals in the aromatic region for M4 corresponded to the expected number of protons for an unmodified furanocoumarin moiety. The signals arising from the 5'-protons at 2.09 ppm were missing. New peaks indicative of hydroxylation at the 5'-position were seen as multiplets arising from the 4'-protons in the 1.76 to 1.86 ppm region and a OH signal 3.95 ppm and a new multiplet at 3.7 to 3.9 ppm corresponding to the single 5'-proton. The spectra most closely resembled the predicted ¹H NMR spectrum of 5'-OH-BG. Together, the data indicate that M3 is 5'-OH-BG. The coeluting minor metabolite that was observed by LC-MS/MS was probably too low to be detected above the background of the NMR spectra.

LC-APCI-MS and MS/MS Analysis of Metabolites Generated by Incubating P450 3A5 with BG. A representative HPLC chromatogram depicting the elution profile of 6-methoxypsoralen, BG, and the BG-derived metabolites monitored at 310 nm from P450 3A5 incubated with BG in the presence of NADPH is shown in Fig. 3A. The photo diode array spectra between 250 and 340 nm of M1, M2, M3, and M5 are shown as insets in Fig. 3A. Because the total ion chromatograms of the samples incubated in the presence or absence of NADPH again were indistinguishable, the XIC with the [M + 1] at 203 was monitored (Fig. 3B). As was seen previously, BG with an m/z of 339 eluted at 50.2 min and the internal standard 6-methoxypsoralen with an m/z of 218 eluted at 18.2 min. Metabolites with an m/z of 203 (M1), 373 (M2), 357 (M3), or 355 (for M4 and M5) eluted at 9.6, 20.3, 33.1, 34.4, and 38 min, respectively, and were only seen in samples incubated with NADPH. M1 eluted at the same retention time and displayed the same m/z of 203 and MS/MS pattern as the authentic standard bergaptol.

View larger version (30K): [in this window] [in a new window]   Fig. 3. APCI-LC-MS analysis of BG metabolites generated by P450 3A5 in the reconstituted system. P450 3A5 was reconstituted with reductase and incubated with BG in the presence or absence of NADPH. Metabolites were extracted and analyzed as described under Materials and Methods. A, HPLC chromatogram of BG and BG-derived metabolites that absorb at 310 nm. The insets show the photodiode array spectrum between 260 and 350 nm of each peak. BG eluted at 50 min, and the internal standard 6-methoxypsoralen (6-MOP) eluted at 17.5 min. B, extracted ion chromatogram of the m/z 203 of the photodiode array profile shown in A. C, m/z of the ions corresponding to the peaks at 9.13 min (M1), 20.3 min (M2), 33.3 min (M3), and 38 min (M5).

M2 eluted at the same retention time as the authentic standard 6',7'-DHBG and displayed an m/z of 373 consistent with this dihydroxylated product. The fragmentation patterns observed for M1 and M2 were similar to those of the authentic standards for bergaptol or DHBG and to the patterns predicted by the Mass Frontiers 3.0 software (Thermo Electron Corporation). The amount of M2 recovered was too low to analyze by ¹H NMR.

M3 corresponded to 5'-OH-BG, the same M3 metabolite that was generated by P450 2B6, based on the HPLC retention time, the photodiode array spectrum, the XIC, the MS/MS fragmentation pattern, and the ¹H NMR spectrum. An additional product that may correspond to a furan-hydroxylated species was again observed coeluting in the M3 peak but in at least a 10-fold lower abundance. M4 corresponded to the 6'-OH and 7'-OH metabolites of BG that were produced by P450 2B6, except that P450 3A5 generated less of these metabolites than P450 2B6.

M5, the second most abundant metabolite seen in the P450 3A5 incubation mixtures, exhibited a weak parent ion signal with an m/z of 355 to 357, consistent with the addition of one oxygen or OH. The photodiode array spectrum and the XIC indicated that the furanocoumarin ring system was unaffected and that the modification had occurred on the geranyl chain. MS/MS analysis of the fragmentation pattern suggested a modification at the 2'- or 3'-carbon of BG. ¹H NMR analysis of M5 was carried out as was done for M4 and M3, and the results confirmed that the furanocoumarin ring was not modified. The proton ratios and signals that were observed at 8.32, 7.63, 7.23, 7.00, and 6.35 ppm were similar to what was seen for intact BG. The most striking feature in the M5 spectrum was the absence of the 2'-signal at 5.58 ppm and the appearance of several multiplets in the 2 to 5 ppm region and the appearance of an OH signal at 3.74 ppm. With this sample, it was also possible to irradiate the OH peak at 3.74 ppm, which caused a triplet at 3.88 ppm corresponding to the 2'-proton to collapse, lending further evidence for M5 being a 2'-OH product of BG.

View larger version (50K): [in this window] [in a new window]   Fig. 4. Energy-based docking of BG into models of P450 2B6 and 3A5: differential exposure of geranyl side chain atoms. A, energy-based docking distribution (Autodock3) for exposure of BG in the model of P450 2B6, indicating the frequency of exposure of geranyl side chain atoms within 4 Å of the oxyferryl (compound I) oxygen. Top, most frequently encountered configurations of BG in P450 2B6 to have exposure of the terminal atoms in the geranyl side chain with less exposure of the initial atoms at the 4-Å evaluation criteria. Bottom, structures show the lowest energy configuration of BG relative to the oxyferryl center without surrounding residues (left) and with nearest neighbor amino acid residues shown and BG shown in magenta (right). B, energy-based flexible ligand docking (Autodock3) of BG into the model for P450 3A5 with the heme in oxyferryl-heme form. Top left, histogram illustrating the frequency with which sites on the geranyl side chain were exposed within 4 Å of the oxyferryl (compound I) oxygen. Bottom left, one of the most frequently observed docked poses with the 1'- and 2'-centers exposed to the oxyferryl oxygen (indicated with an arrow). Right-most panel, space-filling illustration of a representative molecular dynamics configuration of BG interaction with compound I after 210 ps of molecular dynamics equilibration. C, docking of BG into the compound I heme model form of 3A5, using flexible-ligand/rigid protein docking (Glide 4.02), indicates that the configuration with 1'-exposure to compound I is "enriched" in P450 3A5 compared with P450 2B6.

ICM binding site locator was used to examine the binding site and closed cavity volumes near the substrate/inhibitor binding site of the models. 3A4(1TQN) had a binding site volume of 861 ų, whereas 3A4(1WOG) had a binding site volume of 919 ų, and P450 2B4 had a binding site volume of 594 ų. The P450 3A5 model based on the 3A4(1TQN) structure had a binding site volume of 347 ų, whereas P450 2B6 based on the P450 2B4 structure had a volume of 209 ų. By any measure, qualitative and quantitative, the binding site of P450 3A5 is larger near the heme than that of P450 2B6. Moreover the architecture of the binding sites differs. This should, in principle, alter the product distribution and was in agreement with what was observed in the metabolism studies.

Figure 4A illustrates that when BG is docked in the active site of P450 2B6, the P450 2B6/BG binding site interactions that result in the lowest energy-bound configurations have the 4'-, 5'-, and 6'-carbons of the geranyl-oxy chain carbon atoms exposed to the oxyferryl oxygen of compound I (see the BG structure in Fig. 4A for the numbering assignments of the geranyl carbons). This orientation would therefore predict the preferential oxidation at the terminal end of the geranyloxy chain of BG and supports the experimental findings for the metabolism of BG to primarily 6',7'-epoxy-BG by P450 2B6. In contrast to P450 2B6, the P450 3A5/BG binding site interactions that result in the lowest energy-docked configurations of BG occur when the 1'-, 2'-, and 9'-carbons have the most favorable interactions with the compound I center (Fig. 4B). The results are again consistent with the experimental findings that indicated that with P450 3A5, the cleavage of the geranyl-oxy chain at the 1'-carbon to form bergaptol is preferred, followed by oxidation of the geranyl chain to form M5 and M3 (Fig. 4C). Although, in general, the metabolism profile of a substrate is both a function of the configuration of the substrate (which sites are exposed to the active oxygen species) and kinetic factors (determining competitive transformation of sites simultaneously exposed), the docking results for BG in this study are consistent with the predominance of configuration factors in determining the differential profiles observed in the distribution of metabolites generated by P450 3A5 and 2B6.

LC-ESI-MS/MS Analysis of Glutathione Conjugates Obtained from Incubating P450 2B6 and 3A5 with BG in the Presence of GSH. For the P450 2B6 samples incubated with BG in the presence of NADPH and GSH, one major GS conjugate of BG with a mass of 661 Da (M + 1; m/z of 662) was observed (Fig. 5). This adduct eluted at 25 min and was only seen in samples incubated with NADPH. The mass of the GS conjugate corresponded to the mass of BG plus the mass of glutathione plus one oxygen atom. MS/MS analysis of the 662 ion [M + 1] resulted in two major daughter ions with m/z of 526, and 308 (Fig. 6A). The structure most consistent with these fragmentation products is also shown in Fig. 6A. In addition to the monooxygenated GS conjugate of BG a minor secondary adduct with a mass of 678 Da corresponding to a GS conjugate of BG with two oxygen atoms was seen eluting at 16.9 min (data not shown). Again, this ion was only present in samples incubated with NADPH. The appearance of the characteristic 526 and 308 ions in the MS/MS profile (Fig. 6B) of the doubly hydroxylated product suggests that it is structurally related to the monooxygenated GS conjugate of BG and that the second oxygen was introduced on the geranyl-oxy chain (Fig. 6B). Because P450 2B6 showed a preference for metabolism at the 6'- and 7'-carbon, it is conceivable that the 678 species arose from oxidation at that position, but oxidation at 5'-carbon may also be possible. The amount of GS conjugate that was formed was too low to further analyze these products by ¹H NMR.

View larger version (18K): [in this window] [in a new window]   Fig. 5. ESI-LC-MS analysis of BG-GS conjugates isolated from incubation mixtures with P450 2B6. Incubation conditions and extraction/analysis procedures were as described under Materials and Methods. A representative extracted ion chromatogram of the [MH+] ion with an m/z of 662 eluting at 25 min is depicted. The inset shows the mass spectrum of the peak at 25 min.

View larger version (23K): [in this window] [in a new window]   Fig. 6. LC-ESI-MS/MS analysis of the BG-GS conjugates isolated from incubation mixtures containing P450 2B6. Incubation conditions and extraction/analysis procedures were as described under Materials and Methods. A, MS/MS pattern of the ion with an m/z of 662 eluting at 25 min and the proposed structure of the conjugate. B, MS/MS pattern of the minor ion with an m/z of 678 eluting at 16.9 min and the proposed structure of the conjugate.

One major GS conjugate of BG with an m/z of 662 (Fig. 7A, inset) was also observed in P450 3A5 samples incubated with BG in the presence of NADPH and GSH. This adduct also eluted at 25 min and was only seen in samples incubated with NADPH and not in control samples (Fig. 7A; data not shown). The MS/MS analysis of the 662 ion [M + 1]⁺ revealed daughter ions with an m/z of 526, 355, and 308 (Fig. 7B). The structure that is most consistent with this observation is shown in the inset to Fig. 7B, identical to the structure shown in Fig. 6A. Two additional, albeit lower abundance GS conjugates of BG with masses of 678 Da [M + 1]⁺ were seen eluting at 16.9 and at 17.6 min (Fig. 8A). The MS/MS fragmentation patterns of these two products were similar to each other (Fig. 8B) and to the pattern seen with the deoxygenated product obtained with P450 2B6 (Fig. 7B). In each case, a prominent ion at m/z 526 was observed corresponding to a hydroxylated furanocoumarin moiety and oxidized glutathione, suggesting that the second oxidation event occurred on the geranyl-oxy chain. Whether the two products that were seen with P450 3A5 correspond to different isomers that were oxidized at the same position or to two different products with oxidation at the 6'- and 7'- or the 2'- and 3'-carbons of BG cannot be determined from this analysis. A previous report as well as observation in our laboratory suggests that these doublets could correspond to isomers of the same molecule (Shebley et al., 2006). However, the appearance of only one isomer at 16.9 min in the P450 2B6 samples suggests the possible formation of different GS conjugates and may be analogous to two of the possible sites of hydroxylation on the geranyl-oxy chain that were seen in the metabolites identified from P450 2B6 and 3A5. The structures believed to be most consistent with the ion fragmentation pattern are shown in Fig. 8C.

View larger version (19K): [in this window] [in a new window]   Fig. 7. LC-ESI-MS/MS analysis of BG-GS conjugates isolated from incubation mixtures with P450 3A5. Incubation conditions and extraction/analysis procedures were as described under Materials and Methods. A, representative extracted ion chromatogram of the [MH+] ion with an m/z of 662 eluting at 25 min. The inset shows the mass spectrum of the peak at 25 min. B, MS/MS pattern of the 662 ion eluting at 25 min showing the proposed structure of the conjugate in the inset.

View larger version (24K): [in this window] [in a new window]   Fig. 8. LC-ESI-MS/MS analysis of dihydroxylated BG-GS conjugates isolated from incubation mixtures with P450 3A5. A, representative extracted ion chromatogram of the [MH+] ions with an m/z of 678 eluting at 16.9 and at 17.6 min is shown. The insets shows the mass spectra of the peaks at 16.9 and 17.6 min. B, MS/MS pattern of the 678 ion eluting at 16.9 min and the MS/MS pattern of the 678 ion eluting at 17.6 min. C, proposed structure of the conjugates.

	Discussion

Top Abstract Materials and Methods Results Discussion References

 
BG is a mechanism-based inactivator of P450 3A5 and 2B6 (Lin et al., 2005). A combination of metabolite analyses, characterization of GS conjugates of BG, and molecular modeling have been used in this report to elucidate differences in the routes of metabolism of BG by these two enzymes and to understand which reactive BG intermediates may be formed and could be responsible for the mechanism-based inactivation.

The chemical structures of BG and 6-methoxypsoralen share the same furanocoumarin ring structure, and this suggested that some of the routes of metabolism of these two compounds by P450 of the 2B family might be similar. Studies with P450 2B1 indicated that the metabolism of 8-methoxypsoralen resulted in the formation of a monooxygenated metabolite of the furanocoumarin ring, the formation of a dihydrodiol furanocoumarin, and the oxidation of the double bond of the furan ring to generate a monooxygenated reactive intermediate that was responsible for inactivation and could be trapped by GSH (Koenigs and Trager, 1998). Studies with various coumarins and their dihydrodiols indicated that hydroxylation of the furan ring double bond resulted in compounds that displayed a UV shift from approximately 310 to 325 nm (Koenigs and Trager, 1998). P450 2B6 metabolized BG to M4, M3, and M1 that could be observed by their UV absorbance at 310 nm, a molecular ion at m/z 203, and five characteristic ¹H NMR signals at 8.29, 7.81, 7.22 6.31, and 6.28 ppm indicative of the presence of an intact furanocoumarin ring. The most abundant metabolite, M4, had a parent m/z of 355, suggesting the insertion of one oxygen atom into the geranyl-oxy chain of BG. MS/MS analysis was unable to conclusively determine the location of the oxygen addition. ¹H NMR studies were carried out, and the data indicated that the M4 peak was actually composed of two different metabolites with hydroxylations at either the 6'- or the 7'-carbon on the BG geranyl-oxy chain. The difference in two mass units between the observed mass obtained by API-MS for M4 and the theoretical mass cannot be explained at this time. Similar studies with the remaining two metabolites generated by P450 2B6 indicated that M3 was hydroxylated on the 5'-carbon of BG. The minor product M1 was distinguished by an absence of associated radioactivity (i.e., removal of the geranyl-oxy chain containing the radiolabel at the 1'-carbon) and an m/z of 203 and coelution with the authentic standard bergaptol. The ratios of the metabolites suggest that metabolism of BG by P450 2B6 occurs primarily on the geranyl-oxy chain. This observation was underscored by docking studies of BG into a model of P450 2B6. The model indicated that the most favorable binding site for BG/2B6 interactions leads primarily to exposure of the terminal part of the geranyloxy chain and in particular the 5'-, 6'-, and 4'-carbons to the oxyferryl oxygen of compound I. These sites would be the most favorable sites of metabolism since they were the primary sites exposed within 4 Å of the active oxygen species. Previous studies with 6-methoxypsoralen indicated that hydroxylation of the furan double bond by P450 2B1 resulted in the formation of a dihydrodiol product (Koenigs and Trager, 1998). With P450 2B6 (and 3A5), only a very minor monohydroxylated furan product was observed eluting at almost the same time as the M3 metabolite. Either P450 2B6/3A5 do not generate the dihydrodiol species from BG or one of the hydroxyl groups could have been lost as water during sample analysis in the mass spectrometer.

In contrast, metabolism of BG that involves the production of the reactive intermediate responsible for the loss in enzymatic activity seems to result from the metabolism at the furan ring. This observation supports previous studies that indicated that inactivation of P450 enzymes by psoralens required the presence of the furan ring double bond (Letteron et al., 1986; Cai et al., 1993). Tandem mass analysis of the glutathione conjugates of BG formed by P450 2B6 resulted in fragments that were most consistent with GS adduction to the furan ring analogous to what has been observed with 8-methoxypsoralen and P450 2B1 (Koenigs and Trager, 1998).

The HPLC profiles of the metabolites formed by incubating BG with P450 3A4 and 3A5 in the presence of NADPH were previously found to be identical (Lin et al., 2005). Although P450 3A4 seems to be the major P450 involved in the metabolism of BG, P450 3A5 may have additional and yet to be determined important physiological functions, because it is the only 3A family member found in the adrenal, prostate, and kidney and also shows differential levels of expression in children and African Americans (Kuehl et al., 2001). P450 3A5 metabolized BG to at least five different metabolites with bergaptol being the most abundant. Bergaptol could then in turn inactivate other P450 enzymes, such as P450 2B6 (Letteron et al., 1986). In addition to the 5'-OH-BG and 6'- and 7'-OH-BG seen with P450 2B6, a metabolite unique to P450 3A4/5 (M5) was observed. M5 exhibited a UV spectrum with an absorbance maximum at 306 nm, an m/z of 355 to 357, a prominent ion at m/z 203, and an ¹H NMR spectrum in the aromatic region that was indicative of a monooxygenated BG molecule with an intact furanocoumarin ring structure. ¹H NMR studies were required to locate the oxidation to the 2'-carbon on the geranyl-oxy chain of BG. The minor metabolite M2 was characterized by an absorbance maximum at 306 nm and an m/z of 371. The m/z 203 ion suggested the presence of an unmodified furanocoumarin ring system. The mass of this metabolite was most consistent with 6',7'-dihydroxy-BG. This result was supported by the observations that the authentic standard 6',7'-dihydroxy-BG exhibited the same LC, MS, and MS/MS characteristics. The observed metabolites were consistent with the BG-oxyferryl oxygen exposure probed by energy based docking of BG into a model of P450 3A5. The docking results indicated that in marked contrast to P450 2B6, the 1'-carbon in P450 3A5 was the most likely site for metabolism consistent with the formation of bergaptol as the primary metabolite generated by P450 3A5. The furan site was twice as probable as the 1'-site to be exposed to the oxyferryl center. Although the predicted metabolism is a function of both the propensity of exposure and the energy landscape of competitive metabolism of sites exposed to the active species center, the docking results are consistent with the observed metabolite distributions. Reactive intermediates that could be trapped by GSH were derived from oxidation of the double bond of the furan ring of BG. P450 3A5 generated at least two different GS conjugates of BG. The major conjugate exhibited an m/z of 662. The MS/MS fragmentation pattern of the GS conjugate of BG that was generated by P450 3A5 was identical to the one seen with P450 2B6 and indicated that GS was adducted to the furan ring of BG molecule. The most characteristic fragments were a 525 ion that could only arise from a GS adduct at the furanocoumarin ring and a 354 ion that was characteristic of a hydroxylated BG molecule. P450 3A5 also generated two additional GS conjugates of BG that 1) eluted approximately 5 min earlier than the 662 ion, 2) exhibited the same m/z of 678, and 3) eluted with approximately the same intensity within 1 min of each other. Tandem mass analysis of each of the ions indicated that GS was adducted to the furanocoumarin ring structure (M + 1; 525) and insertion of one oxygen atom at one of the geranyl-oxy double bonds (loss of M + 1 of 153 and M + 1 of 370).

View larger version (27K): [in this window] [in a new window]   Fig. 9. Proposed scheme for the metabolism of BG by P450 2B6 and 3A5.

	Acknowledgements

 
¹H NMR spectra were obtained by Scott Woehler in the NMR core facility (Department of Chemistry, University of Michigan, Ann Arbor, MI).

	Footnotes

 
This work was supported in part by National Institutes of Health Grant CA-16954 (to P.F.H.) and R43-DC-6925 (to D.L.H.).

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

doi:10.1124/jpet.105.099887.

ABBREVIATIONS: P450, cytochrome(s) P450; BG, bergamottin; M1, bergaptol; M2, 6',7'-dihydroxy-BG; M3, 5'-OH-BG; M4, mixture of 6'- and 7'-OH-BG; M5, 2'-OH-BG; DHBG, 6',7'-dihydroxybergamottin; GSH, reduced glutathione; GS, oxidized glutathione; HPLC, high-performance liquid chromatography; APCI, atmospheric pressure chemical ionization; LC, liquid chromatography; ESI, electrospray ionization; MS/MS, tandem mass spectrometry; MS, mass spectrometry; ICM, internal coordinate mechanics; XIC, extracted ion chromatogram; L-754,394, N-[2(R)-hydroxy-1(S)-indonyl]-5-[2(S)-(1,1-dimethyl ethylaminocarbonyl)-4-[(fluro[2,3-b]pyridin-5-yl)methyl]piperazin-1-yl]-4(S)-hydroxy-2-(R)-phenylmethylpentamide.

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

Address correspondence to: Dr. Paul F. Hollenberg, Department of Pharmacology, 2301 MSRB III, 1150 West Medical Center Dr., Ann Arbor, MI 48109-0632. E-mail: phollen{at}umich.edu

	References

Top Abstract Materials and Methods Results Discussion References

 

Bailey DG, Dresser GK, Kreeft JH, Munoz C Freeman DJ, and Bend JR (2000) Grapefruit-felodipine interaction: effect of unprocessed fruit and probable active ingredients. Clin Pharmacol Ther 68: 468–477.[CrossRef][Medline]

Blobaum AL, Harris DL, and Hollenberg PF (2005) P450 Active site architecture and reversibility: inactivation of cytochromes P450 2B4 and 2B4 T302A by tert-butyl acetylenes. Biochemistry 44: 3831–3844.[CrossRef][Medline]

Bumpus NN, Sridar C, Kent UM, and Hollenberg PF (2005) The naturally occurring cytochrome P450 (P450) 2B6 K262R mutant of P450 2B6 exhibits alterations in substrate metabolism and inactivation. Drug Metab Dispos 33: 795–802.[Abstract/Free Full Text]

Cai Y, Bennett D, Nair RV, Ceska O, Ashwood-Smith MJ, and DiGiovanni J (1993) Inhibition and inactivation of murine hepatic ethoxy- and pentoxyresorufin O-dealkylase by naturally occurring coumarins. Chem Res Toxicol 6: 872–879.[CrossRef][Medline]

Canutescu AA, Shelenkov AA, and Dunbrack RL Jr (2003) A graph-theory algorithm for rapid protein side-chain prediction. Protein Sci 12: 2001–2014.[Abstract/Free Full Text]

Ding X and Kaminsky LS (2003) Human extrahepatic cytochromes P450: function in xenobiotic metabolism and tissue-selective chemical toxicity in the respiratory and gastrointestinal tracts. Annu Rev Pharmacol Toxicol 43: 149–173.[CrossRef][Medline]

Domanski TL, Finta C, Halpert JR, and Zaphiropoulos PG (2001) cDNA cloning an initial characterization of CYP3A43, a novel human cytochrome P450. Mol Pharmacol 59: 386–392.[Abstract/Free Full Text]

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]

Faucette SR, Hawke RL, Lecluyse EL, Shord SS, Yan B, Laethem RM, and Lindley CM (2000) Validation of bupropion hydroxylation as a selective marker of human cytochrome P450 catalytic activity. Drug Metab Dispos 28: 1222–1230.[Abstract/Free Full Text]

Friesner RA, Banks JL, Murphy RB, Halgren TA, Klicic JJ, Mainz DT, Repasky MP, Knoll EH, Shelley M, Perry JK, et al. (2004) Glide: a new approach for rapid, accurate docking and scoring. 1. Method and assessment of docking accuracy. J Med Chem 47: 1739–1749.[CrossRef][Medline]

Gerber JG, Rhodes RJ, and Gal J (2004) Stereoselective metabolism of methadone N-demethylation by cytochrome P450 2B6 and 2C19. Chirality 16: 36–44.[Medline]

Hanna IH, Roberts ES, and Hollenberg PF (1998) Molecular basis for the difference in lidocaine binding and the regioselectivity of oxidation by cytochrome P450 2B1 and 2B2. Biochemistry 37: 311–318.[CrossRef][Medline]

Harris DL, Park JY, Gruenke L, and Waskell L (2004) Theoretical study of the ligand-CYP2B4 complexes: effect of structure on binding free energies and heme spin state. Proteins 55: 895–914.[CrossRef][Medline]

He K, Iyer KR, Hayes RN, Sinz MW, Woolf TF, and Hollenberg PF (1998) Inactivation of cytochrome P450 3A4 by bergamottin, a component of grapefruit juice. Chem Res Toxicol 11: 252–259.[CrossRef][Medline]

Hesselink DA, van Shaik RH, van der Heiden IP, van der Werf M, Gregoor PJ, and Lindemans J (2003) Genetic polymorphisms of the CYP3A4, CYP3A5, and MDR-1 genes and pharmacokinetics of the calcineurin inhibitors cyclosporine and tacrolimus. Clin Pharmacol Ther 74: 245–254.[CrossRef][Medline]

Huang Z, Roy P, and Waxman DJ (2000) Role of human microsomal CYP3A4 and CYP2B6 in catalyzing N-dechlorethylation of cyclophosphamide and ifosfamide. Biochem Pharmacol 59: 961–972.[CrossRef][Medline]

Ingelman-Sundberg M (2004) Pharmacogenetics of cytochrome P450 and its application in drug therapy: the past, present and future. Trends Pharmacol Sci 25: 193–200.[CrossRef][Medline]

Jin L and Baillie TA (1997) Metabolism of the chemoprotective agent diallyl sulfide to glutathione conjugates in rats. Chem Res Toxicol 10: 318–327.[CrossRef][Medline]

Kabasch W and Sander C (1983) Dictionary of protein secondary structure: pattern recognition of hydrogen-bonded and geometrical features. Biopolymers 22: 2577–2637.[CrossRef][Medline]

Kivistö KT, Niemi M, Schaeffeler E, Pitkälä K, Tilvis R, Fromm MF, Schwab M, Eichelbaum M, and Strandberg T (2004) Lipid-lowering response to statins is affected by CYP3A5 polymorphism. Pharmacogenetics 14: 523–525.[CrossRef][Medline]

Koenigs LL and Trager WF (1998) Mechanism-based inactivation of cytochrome 2B1 by 8-methoxypsoralen and several other furanocoumarins. Biochemistry 37: 13184–13193.[CrossRef][Medline]

Kuehl P, Zhang J, Lin Y, Lamba J, Assem M, Schuetz J, Watkins PB, Daly A, Wrighton SA, Hall SD, et al. (2001) Sequence diversity in CYP3A promoters and characterization of the genetic basis of polymorphic CYP3A5 expression. Nat Genet 27: 383–391.[CrossRef][Medline]

Letteron P, Descatoire V, Larrey D, Tinel M, Geneve J, and Pessayre D (1986) Inactivation and induction of cytochrome P-450 by various psoralen derivatives in rats. J Pharmacol Exp Ther 238: 685–692.[Abstract/Free Full Text]

Lightning LK, Jones JP, Friedberg T, Pritchard MP, Shou M, Rushmore TH, and Trager WF (2000) Mechanism-based inactivation of cytochrome P450 3A4 by L-754,394. Biochemistry 39: 4276–4287.[CrossRef][Medline]

Lin HL, Kent UM, and Hollenberg PF (2002) Mechanism-based inactivation of cytochrome P450 3A4 by 17-ethynylestradiol: evidence for heme destruction and covalent binding to protein. J Pharmacol Exp Ther 301: 160–167.[Abstract/Free Full Text]

Lin HL, Kent UM, and Hollenberg PF (2005) The grapefruit juice effect is not limited to P450 3A4: evidence for bergamottin-dependent inactivation, heme destruction and covalent binding to protein in P450 2B6 and 3A5. J Pharmacol Exp Ther 313: 154–164.[Abstract/Free Full Text]

Mays DC, Hilliard JB, Wong DD, Chambers MA, Park SS, Gelboin HV, and Gerber N (1990) Bioactivation of 8-methoxypsoralen and irreversible inactivation of cytochrome P 450 in mouse liver microsomes: modification by monoclonal antibodies, inhibition of drug metabolism and distribution of covalent adducts. J Pharmacol Exp Ther 254: 720–731.[Abstract/Free Full Text]

Mendes J, Nagarajaram HA, Soares CM, Blundell TL, and Carrondo MA (2001) Incorporating knowledge-based biases into an energy-based side-chain modeling method: application to comparative modeling of protein structure. Biopolymers 59: 72–86.[CrossRef][Medline]

Morris GM, Goodsell DS, Halliday RS, Huey R, Hart WE, Belew RK, and Olson AJ (1998) Automated docking using a Lamarckian genetic algorithm and an empirical binding free energy function. J Comput Chem 19: 1639–1662.[CrossRef]

Paine MF, Criss AB, and Watkins PB (2004) Two major grapefruit juice components differ in intestinal CYP3A4 inhibition kinetic and binding properties. Drug Metab Dispos 32: 1146–1153.[Abstract/Free Full Text]

Park J-Y and Harris DL (2003) Construction and assessment of models of CYP2E1: predictions of metabolism from docking, molecular dynamics and density functional theoretical calculations. J Med Chem 46: 1645–1660.[CrossRef][Medline]

Sahali-Sahly Y, Balani S, Lin JH, and Baillie TA (1996) In vitro studies on the metabolic activation of the furanopyridine L-754,394, an highly potent and selective mechanism-based inhibitor of cytochrome P450 3A4. Chem Res Toxicol 9: 1007–1012.[CrossRef][Medline]

Sali A and Blundell TL (1993) Comparative protein modeling by satisfaction of spatial restraints. J Mol Biol 234: 779–815.[CrossRef][Medline]

Scott EE, Spatzenegger M, and Halpert JR (2001) A truncation of 2B subfamily cytochromes P450 yields increased expression levels, increased solubility, and decreased aggregation while retaining function. Arch Biochem Biophys 395: 57–68.[CrossRef][Medline]

Shebley M, Jushchyshyn MI, and Hollenberg PF (2006) Identification of glutathione and N-acetylcystein adducts formed by cytochrome P450 2B enzymes during the metabolism of phencyclidine and 2-phenyl-2-(1-piperidinyl)propane. Drug Metab Dispos 34: 375–383.[Abstract/Free Full Text]

Sippl MJ (1993) Recognition of errors in three-dimensional structures of proteins. Proteins 17: 355–362.[CrossRef][Medline]

Thompson JD and Higgins DG (1994) CLUSTAL W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-specific gap penalties and weight matrix choice. Nucleic Acids Res 22: 4673–4680.[Abstract/Free Full Text]

Wangensteen H, Molden E, Christensen H, and Malterud KE (2003) Identification of epoxybergamottin as a CYP3A4 inhibitor in grapefruit peel. Eur J Clin Pharmacol 58: 663–668.[CrossRef][Medline]

Ward BA, Gorski JC, Jones DR, Hall SD, Flockhart DA, and Desta Z (2003) The cytochrome P450 2B6 (CYP2B6) is the main catalyst of efavirenz primary and secondary metabolism: implication for HIV/AIDS therapy and utility of efavirenz as a substrate marker of CYP2B6 catalytic activity. J Pharmacol Exp Ther 306: 287–300.