Ketamine is metabolized to both active and inactive metabolites by cytochrome P4502B6 (CYP2B6), which is assisted by P450 oxidoreductase
CYP2B6 and by P450 oxidoreductase are highly polymorphic, with many genetic variants
Several genetic variants of CYP2B6 and P450 oxidoreductase have diminished ketamine N-demethylation activity
Variants do not have altered stereoselectivity of ketamine metabolism
Results suggest candidate polymorphisms of CYP2B6 and P450 oxidoreductase for clinical evaluation
Human ketamine N-demethylation to norketamine in vitro at therapeutic concentrations is catalyzed predominantly by the cytochrome P4502B6 isoform (CYP2B6). The CYP2B6 gene is highly polymorphic. CYP2B6.6, the protein encoded by the common variant allele CYP2B6*6, exhibits diminished ketamine metabolism in vitro compared with wild-type CYP2B6.1. The gene for cytochrome P450 oxidoreductase (POR), an obligatory P450 coenzyme, is also polymorphic. This investigation evaluated ketamine metabolism by genetic variants of human CYP2B6 and POR.
CYP2B6 (and variants), POR (and variants), and cytochrome b5 (wild-type) were coexpressed in a cell system. All CYP2B6 variants were expressed with wild-type POR and b5. All POR variants were expressed with wild-type CYP2B6.1 and b5. Metabolism of R- and S-ketamine enantiomers, and racemic RS-ketamine to norketamine enantiomers, was determined using stereoselective high-pressure liquid chromatography–mass spectrometry. Michaelis–Menten kinetic parameters were determined.
For ketamine enantiomers and racemate, metabolism (intrinsic clearance) was generally wild-type CYP2B6.1 > CYP2B6.4 > CYP2B6.26, CYP2B6.19, CYP2B6.17, CYP2B6.6 > CYP2B6.5, CYP2B6.7 > CYP2B6.9. CYP2B6.16 and CYP2B6.18 were essentially inactive. Activity of several CYP2B6 variants was less than half that of CYP2B6.1. CYP2B6.9 was 15 to 35% that of CYP2B6.1. The order of metabolism was wild-type POR.1 > POR.28, P228L > POR.5. CYP2B6 variants had more influence than POR variants on ketamine metabolism. Neither CYP2B6 nor POR variants affected the stereoselectivity of ketamine metabolism (S > R).
Genetic variants of CYP2B6 and P450 oxidoreductase have diminished ketamine N-demethylation activity, without affecting the stereoselectivity of metabolism. These results suggest candidate genetic polymorphisms of CYP2B6 and P450 oxidoreductase for clinical evaluation to assess consequences for ketamine pharmacokinetics, elimination, bioactivation, and therapeutic effects.
KETAMINE is anesthetic, analgesic at subanesthetic doses, widely used for acute, perioperative, and chronic pain, and sedation,1 and is a World Health Organization Essential Medicine. Ketamine more recently has been used for severe and treatment-resistant major depression, with rapid and sustained response.2–4 Ketamine can be administered by intravenous, oral, sublingual, intramuscular, intranasal, and rectal routes. Extensive first-pass metabolism limits oral bioavailability.
Ketamine is administered clinically as a racemic mixture of R- and S- enantiomers, although S-ketamine alone is used in Europe. Intranasal S-ketamine is in phase 3 clinical trials for depression5 and was evaluated for analgesia by inhalation.6 S-Ketamine is approximately fourfold more potent than R-ketamine as an analgesic, consistent with greater N-methyl-d-aspartate receptor affinity.1 S-Ketamine has faster systemic clearance and anesthetic recovery than R- and RS-ketamine.1,7
Ketamine is extensively metabolized hepatically via N-demethylation in vitro and in vivo to the predominant metabolite norketamine.1,7 Norketamine and the other primary metabolite hydroxyketamine are further metabolized to hydroxynorketamine. Norketamine is also further metabolized to dehydronorketamine. Norketamine and hydroxynorketamine are pharmacologically active, with enantiomeric selectivity, and may contribute to antidepressant and/or analgesic effects of ketamine.8–11 Thus, while ketamine N-demethylation was initially considered an elimination pathway, it is now also considered an important bioactivation pathway mediating one or more pharmacologic effects of ketamine. Hence variability in ketamine N-demethylation may have therapeutic importance.
Ketamine metabolism and disposition are stereoselective. In human liver microsomes, S-ketamine N-demethylation was 20% greater than R-ketamine and 10% greater than the racemate.12 There is also a metabolic enantiomeric interaction whereby one enantiomer may inhibit metabolism of the other enantiomer.12 This is also seen clinically, whereby R-ketamine inhibits systemic clearance of S-ketamine.13
Ketamine is metabolized by the P450 enzyme system, consisting of cytochrome P450 (CYP), NADPH–cytochrome P450 oxidoreductase (POR), and cytochrome b5.14,15 Cytochrome P450 functions in obligate partnership with POR, which transfers electrons from NADPH to cytochrome P450. Cytochrome b5 can also function electively in this capacity. The CYP2B6 isoform is the major catalyst in vitro of hepatic ketamine N-demethylation, norketamine hydroxylation, and ketamine metabolism overall, at clinically relevant concentrations,16–21 as well as clinically.22 CYP2B6-catalyzed ketamine N-demethylation is also the first step in putative bioactivation to a therapeutically significant metabolite.23
The CYP2B6 gene is highly polymorphic.24 At least 38 allelic variants have been described,25 25 are considered important, and eight are common in at least one racial/ethnic population.26 Several CYP2B6 variants encode proteins with altered activity.24 A common variant, CYP2B6*6, found mainly in African, African American, and some Asian populations, causes diminished hepatic CYP2B6 enzyme (CYP2B6.6) expression and catalytic activity toward several substrates, compared with wild-type CYP2B6.1. Liver microsomes from humans carrying CYP2B6*6 had lesser ketamine N-demethylation activity compared with CYP2B6*1 genotypes, consistent with lower activity of expressed CYP2B6.6.20 The influence of other CYP2B6 genotypes on ketamine metabolism is unknown.
The POR gene is also polymorphic, with at least 48 allelic variants described to date.27 Some POR enzyme variants have altered activity.28,29 The influence of POR variants on cytochrome P450 activity varies with the particular cytochrome P450 isoform, and sometimes with different substrates for the same P450. CYP2B6.1 activity has been reported to be decreased, unchanged, or increased by various POR enzyme variants in vitro and in vivo.29–31 The influence of any POR gene variant on ketamine metabolism is unknown. This investigation tested the hypothesis that one or more human CYP2B6 or POR gene variants result in altered ketamine N-demethylation activity.
Materials and Methods
S-Ketamine was purchased from Sigma-Aldrich (USA). R-Ketamine was purchased from Cayman Chemical (USA). Racemic RS-ketamine was from Spectrum Chemical Manufacturing Corp. (USA). Norketamine (d0 and d4) were from Cerilliant (USA). Spodoptera frugiperda (Sf9) cells and Sf-900 III SFM culture media were purchased from ThermoFisher (USA). Trichoplusia ni (Tni) cells, ESF AF culture media, and BestBac 2.0 baculovirus cotransfection kits were from Expression Systems (USA). BacPAK baculovirus rapid titer kit was from Clontech (USA). Strata-X 33u (30 mg) solid-phase extraction plates were from Phenomenex (USA). Other reagents were from Sigma-Aldrich.
Construction of Plasmid Transfer Vectors
Human genes for CYP2B6, POR, and b5 were amplified from the human liver Quick-Clone cDNA library (Clontech) and inserted individually into the transfer vector pVL1393 using the In-Fusion HD cloning system (Clontech). The plasmid carrying the genes of wild-type CYP2B6 and POR, pVL1393/2B6*1 and pVL1393/POR*1, were used as the template, and the polymorphic variants of CYP2B6 and POR (table 1) were generated by site-directed mutagenesis using Quik-Change II XL site-directed mutagenesis kit (Agilent, USA) according to the manufacturer’s instructions. The sequences of the forward primers are shown in the appendix with the coding introducing mutations underlined. The desired mutations were verified by DNA sequencing at the Protein and Nucleic Acid Chemistry Laboratory at Washington University (St. Louis, Missouri).
Generation of Recombinant Baculoviruses
BestBac 2.0 baculovirus cotransfection kits were used for generation of recombinant baculovirus. Point mutations were introduced into the wild-type baculovirus genome of the strain Autographa californica multiple nuclear polyhedrosis virus (AcMNPV) to generate novel restriction sites. Cleavage at these sites results in linearized defective DNA lacking an essential gene coded by orf1629. The linearized DNA is cotransfected into insect cells with transfer vector carrying the gene of interest and a copy of the missing sequence. Homologous recombination occurring inside the cotransfected cells introduces the gene of interest into the polyhedron locus and restores the essential gene (orf1629), leading to generation of a recombinant viral genome that can replicate and reproduce infectious recombinant virus. Sf9 cells were maintained in 125-ml Erlenmeyer shaker flasks in Sf-900 III SFM medium with shaking at 125 rpm and 27°C. Following the manufacturer’s instructions, 2 ml of Sf9 cells in log phase growth from the shaker flask were plated in each well of a six-well plate at 9 × 105 cells/ml for each cotransfection and allowed to adhere for 30 min. Meanwhile, the transfection reagent was prepared as follows: 100 µl of transfection medium was added to each of two sterile polystyrene tubes. Two µg of plasmid DNA of each transfer vector pVL1393/CYP2B6, pVL1393/POR, or pVL1393/b5 was combined with 0.5 µg of BestBac linearized DNA in one tube, and 6 µl of transfection reagent was added to the second tube. After incubation for 5 min at room temperature, the contents of both tubes were combined, and the resulting cotransfection mixture was further incubated for 20 min and then diluted by adding transfection medium to a total volume of 1 ml, which is ready for cotransfection in the next step. Culture media were removed from the six-well plate, and the diluted cotransfection mixtures were added to the wells. After incubation at 27°C for 4 h, 3 ml of Sf900 III SFM medium was added to each well, and the plate was incubated for 5 d at 27°C. At the end of the incubation, cells and cell debris were removed by centrifugation at 2,500 rpm for 5 min, and the supernatant was collected as p0 viral generation. For next viral generation p1, 0.5 ml of p0 was used to infect 40 ml of Sf9 cells in suspension in early log phase growth, and the viruses were harvested after 3 to 4 days of incubation at 27°C. The maximal viral generation is limited to p3 to minimized undesirable variants due to successive amplification. All viral titers were determined using the BacPAK baculovirus rapid titer kit (Clontech).
Expression of Recombinant Proteins in Insect Cells
All CYP2B6 variants were coexpressed in a single cassette with redox partners P450 reductase (POR) and cytochrome b5 in insect cells by triple infection. Tni cells were maintained in a 125-ml Erlenmeyer shake flask in ESF AF medium with shaking at 125 rpm and 27°C. 40 ml of Tni cells in early log phase growth were seeded at a cell density of 1 × 106 cells/ml. After overnight growth at 27°C with shaking at 125 rpm, the cells were infected with the recombinant baculoviruses carrying the genes of CYP2B6, POR, and b5, with the multiplicities of infection at ratio of 4:2:1 (CYP2B6:POR:b5), based on cell count and viral plaque-forming units measured from the titer assay. The heme precursors δ-aminolevulinic acid and ferric citrate were added at the time of infection (100 µM final concentration for each). After 48 to 72 h growth postinfection, cells were harvested by centrifugation for 15 min at 3,000 g and washed two times with phosphate-buffered saline followed by centrifugation in each wash step. The cell pellets were resuspended in 100 mM potassium phosphate buffer at pH 7.4 and stored frozen at −80°C. Frozen cells were thawed and lysed on ice in a Potter–Elvehjem tissue homogenizer. Fully cell disruption was achieved by the combination of one freeze–thaw cycle and 10 strokes in the glass Teflon Potter–Elvehjem pestle. Aliquots of 0.5 ml of homogenized cells were stored at −80°C.
Western Blot Analysis of Expressed Proteins
The expressed proteins were identified by Western blot analysis as described previously.32,33 Electrophoresis was carried out using a precast NuPAGE 4 to 12% Bis-Tris gel (Life Technologies, USA), and SeeBlue Plus2 prestained protein standard (ThermoFisher) as molecular weight markers. For protein identification, rabbit anti-CYP2B6 (H-110) antibody, rabbit anti-CTPOR (H-300) antibody, and mouse anti-cytochrome b5 (36) antibody (Santa Cruz Biotechnology, USA) were used as primary antibodies, and goat anti-rabbit IRDye 680 and goat anti-mouse IRDye 800CW (LI-COR, USA) were used as secondary antibodies.
Determination of P450 and Cytochrome b5 Contents and POR Activity
All assays for P450 content, b5 content, and POR activity were carried out using a Synergy MX microplate reader (Biotek, USA). Total protein concentrations were determined using Bio-Rad protein assay dye regent concentrate, which is based on the Bradford method. P450 concentration was determined by difference spectrum of ferrous-carbon monoxide complex in a carbon monoxide (CO)−binding assay (reduced vs. reduced CO) using the extinction coefficient Δε450–490nm of 91 mM−1 cm−1. Cytochrome b5 content was determined by difference spectrum of NADH-reduced and oxidized b5 using an extinction coefficient Δε424–410nm of 185 mM−1 cm−1. POR activity was measured by NADPH-cytochrome c reductase activity. To 200 µl of diluted cell lysate in 0.3 M potassium phosphate buffer at pH 7.7 containing 40 µM cytochrome c, 100 µM NADPH was added to initiate reduction of cytochrome c, and the reaction was followed at 550 nm and 23°C. The reaction rate was calculated using an extinction coefficient of ε550 nm of 21 mM−1 cm−1 for reduced cytochrome c. POR activity was converted to POR content based on the assumption that 3,000 nmol cytochrome c are reduced/min per nmol POR at 23°C.34 Because the cytochrome c reductase activities of POR variants are unknown, POR content could not be determined from cytochrome c reduction, and the content of POR variants was determined by Western blotting using POR.1 as a standard. All antibodies used were specific for CYP2B6, POR, and b5, and there was no cross-reactivity.
All incubations were carried out in triplicate in 96-well PCR plates with raised wells. Substrate S- or R-ketamine with concentrations of 0, 0.125, 0.25, 0.5, 1.25, 2.5, 5, 12.5, 25, 50, 125, 250, and 500 µM was mixed with the enzyme CYP2B6/POR/b5, and incubated at 37°C for 5 min. For RS-ketamine, each substrate concentration was doubled, and the highest 1,000 µM represents sum of S- and R-ketamine 500 µM each. Final CYP2B6 concentration was 2.5 pmol/ml, and the total reaction volume was 200 µl in 100 mM potassium phosphate (pH 7.4). The reaction was initiated by adding an NADPH regenerating system (final concentrations: 10 mM glucose 6-phosphate, 1 mM β-NADP, 1 unit/ml glucose-6-phosphate dehydrogenase, and 5 mM magnesium chloride, preincubated at 37°C for 10 min). After incubation for 10 min, the reaction was quenched by adding 40 µl of 20% trichloroacetic acid containing internal standard norketamine-d4 (final concentration, 90 ng/ml). The plate was centrifuged at 2,500 rpm for 5 min to remove precipitated proteins, and the supernatant was subjected to solid phase extraction and liquid chromatography–tandem mass spectrometry analysis. Predicate experiments showed that ketamine N-demethylation was linear with time and protein concentration (not shown).
High-pressure Liquid Chromatography/Tandem Mass Spectrometry
Norketamine concentrations were determined by enantioselective high-pressure liquid chromatography–tandem mass spectrometry using solid-phase extraction, as originally implemented for plasma analysis.7 To 200 µl of incubation samples was added 0.6 ml of 10 mM ammonium acetate in water (pH 9.5). Solid-phase extraction plates were conditioned with 1 ml of methanol, then water, and then 10 mM ammonium acetate in water (pH 9.5). Samples were loaded under soft (5 mmHg) vacuum at 0.5 ml/min, and the plate was washed with water and then completely dried under high vacuum (10 to 15 mmHg for 2 to 5 min). Analytes were eluted with 0.5 ml of methanol under soft vacuum (5 mmHg) and then evaporated to dryness under nitrogen at 40°C. For analysis, samples were resuspended in a 200-µl mobile phase (10 mM ammonium acetate, pH 7.6). Calibration standards contained 0, 2, 5, 10, 20, 50, 100, 200, 500, 1,000, and 2,000 ng/ml racemic norketamine in 100 mM potassium phosphate buffer (pH 7.4). Calibrators were processed identically to incubation samples.
High-pressure liquid chromatography–mass spectrometry analysis was performed on an ultra-fast liquid chromatography system (Shimadzu Scientific Instruments, USA) with a CMB-20A system controller, two LC-20ADXR pumps, DGU-20A3 degasser, SIL-20AC autosampler, FCV-11AL solvent selection module, and CTO-20A column oven (30°C) containing a ChiralPak AGP analytical column (100 × 2.0 mm, 5 µm) and AGP guard column (10 × 2.0 mm; Chiral Technologies, USA), coupled to an API-4000 QTrap triple quadrupole mass spectrometer (Applied Biosystems/MDS Sciex, USA). The mobile phase (0.22 ml/min) was 10 mM ammonium acetate in water, pH 7.60 (A) and isopropanol (B). The column was equilibrated with 4% B and maintained after injection for 0.1 min, and then a linear gradient to 16.8% B was applied for more than 12 min, then reverted back to 4% B for more than 0.5 min, and reequilibrated for 3.5 min. Total run time was 16 min. The injection volume was 10 µl. Under these conditions, approximate retention time for each compound was 5.37 and 6.86 min for S- and R-norketamine isomers. The mass spectrometer electrospray ion source was operated in positive-ion multiple reaction monitoring mode. The [M+H]+ transitions were optimized for norketamine as follows: m/z 224.0 → 125.0 and 228.2 → 129.1 for norketamine-d0 and d4. Mass spectrometer settings for the declustering potential (51 to 56 V), collision energy (33 V), entrance potential (10 V), and collision cell exit potential (20 V) were optimized. Optimized global parameters were as follows: source temperature, 350°C; ion spray voltage, 5,500 V; nitrogen (psig) curtain gas, 20; gas 1, 70; gas 2, 10; collision gas medium. Analytes were quantified using peak area ratios and standard curves. Control incubations lacking enzyme were included for all reactions to determine any background norketamine content, which was subtracted from all results.
Norketamine formation by enzyme variants was analyzed by ANOVA with post hoc Dunnet’s test (SigmaPlot 12.5; Systat, USA). The results are the means ± SD. Norketamine formation versus substrate concentration data were analyzed by nonlinear regression analysis. Ketamine demethylation was analyzed using a single-enzyme Michaelis–Menten model. Where appropriate, the equation was modified to reflect substrate inhibition. The results are reported as the parameter estimate ± standard error of the estimate. In vitro intrinsic clearance (Clint) was Vmax/Km.
CYP2B6 Variants in Insect Cells
P450, POR, and b5 contents for expression of each polymorphic variant are listed in table 2. All cytochrome P450 variants were expressed with wild-type POR and b5. All POR variants were expressed with wild-type CYP2B6 and b5. Figure 1 shows difference spectra of the ferrous–carbon monoxide complex for each CYP2B6 variant, showing expression of viable P450 protein. CYP2B6.19 and CYP2B6.26 were expressed mostly as P420 protein species (inactive enzyme) in regular expression medium, suggesting instability of expressed protein. However, when 5% glycerol was added to the medium 24 h postinfection, the active P450 content increased, while P420 content decreased (fig. 2). CYP2B6.16 and CYP2B6.18 were abundantly expressed, based on Western blot, but mostly as apoprotein, because P450 difference spectra were not detectable, and holoenzyme could not be quantified (table 2).
Ketamine Demethylation Catalyzed by
CYP2B6 can catalyze primary ketamine demethylation to norketamine and subsequent secondary norketamine conversion to both hydroxynorketamine and dehydronorketamine. Measurement of norketamine alone might theoretically underreport CYP2B6-dependent ketamine metabolism. A preliminary experiment therefore evaluated ketamine metabolism to hydroxynorketamine and dehydronorketamine by CYP2B6.1 at the highest substrate concentration used (1,000 µM RS-ketamine), and none was observed under the incubation conditions used (results not shown). Therefore only norketamine was routinely measured.
Norketamine formation from individual ketamine enantiomers and racemic ketamine at therapeutic concentrations catalyzed by coexpressed CYP2B6 (wild-type and variants), wild-type POR, and wild-type b5 is shown in figure 3. N-Demethylation of enantiomeric S-ketamine exceeded that of R-ketamine for all active CYP2B6 enzymes. For both ketamine enantiomers, activity of the variants was less than that of wild-type CYP2B6.1. CYP2B6.16 and CYP2B6.18 were comparatively catalytically inactive.
Norketamine formation catalyzed by wild-type CYP2B6.1 (fig. 4) and the catalytically active CYP2B6 variants (fig. 5) and coexpressed POR and b5 was determined as a function of R-ketamine, S-ketamine, and racemic RS-ketamine concentrations. For CYP2B6.1 and all CYP2B6 variant proteins, S-norketamine formation exceeded that of R-norketamine from individual enantiomers. Formation of R- and S-norketamine was less from the racemate than from individual ketamine enantiomers. Michaelis–Menten kinetic parameters were determined for all active enzymes (tables 3 and 4). There was a fivefold range in Vmax and Km parameters for both R- and S-ketamine, alone and in the racemate. CYP2B6.19 had the highest Vmax, which was more than twofold greater than that of CYP2B6.1, but the Km was also higher, so the net effect was a lower in vitro intrinsic clearance (Clint) compared with CYP2B6.1. Similarly, CYP2B6.6 Vmax was also greater than that of CYP2B6.1, but the Km was also higher, so Clint was lower than that of CYP2B6.1. The Clint for CYP2B6.1 was generally at least twofold higher than that of any CYP2B6 variant. Across all CYP2B6 isoforms, for both R- and S-ketamine enantiomers alone and in the racemate, Vmax and Km parameters averaged 25% less for R- than S-ketamine, and the intrinsic clearance was comparable. Across all CYP2B6 variants, Vmax averaged twofold greater for metabolism of the enantiomers alone compared with the racemate, for both R- and S-ketamine.
Ketamine Demethylation Catalyzed by POR Variants
Norketamine formation from individual ketamine enantiomers and racemic ketamine at therapeutic concentrations catalyzed by wild-type CYP2B6.1 coexpressed with P450 oxidoreductase variants POR.5, POR.28, and P228L, and wild-type cytochrome b5 is shown in figure 6. Enantiomeric S-ketamine N-demethylation exceeded that of R-ketamine for all POR isoforms. Relative activity of POR variants was POR.1 > P228L > POR.28 > POR.5.
Norketamine formation catalyzed by P450 oxidoreductase variants and wild-type CYP2B6.1 and cytochrome b5 was determined as a function of R-ketamine, S-ketamine, and racemic RS-ketamine concentrations (fig. 7). The results for wild-type POR are in figure 4. Michaelis–Menten kinetic parameters are in tables 3 and 4. Km values were similar for the three POR variants and wild-type reductase. However, some POR variants had lower Vmax values for enantiomeric S- but not R-ketamine. For racemic ketamine, Vmax values for POR variants were comparable to those for wild-type POR. Intrinsic clearances for POR variants were 56 to 98% of those for wild-type POR. The POR.5 variant had the lowest Vmax and Clint. POR polymorphisms did not affect the stereochemistry of ketamine N-demethylation (R vs. S) or the metabolic enantiomeric interaction.
Cytochrome P450, P450 reductase, and cytochrome b5 constitute a functional unit. POR is an obligatory redox partner, while b5 can increase, decrease, or have no influence on cytochrome P450 activity, depending on the substrate, cytochrome P450 isoform, and cytochrome P450 genetic variant. Cytochrome P450 can be expressed in bacterial, yeast, mammalian, and insect cells, using transient or stable transfection systems, with or without coexpressed POR and b5.24,35 Yeast and bacterial systems are easily used, and yield abundant protein, but require exogenous POR and b5. Effective bacterial expression of mammalian cytochrome P450 usually requires amino-terminal modification, which renders the protein different than native enzyme. Individual bacterially expressed cytochrome P450, POR, and b5 can be purified and reconstituted, along with crucial lipids, into a catalytically competent system, but this requires considerable extra effort, and protein–protein interactions may not fully recapitulate native interactions. Mammalian systems, particularly monkey kidney COS cells and human embryonic kidney HEK cells, allow easy cytochrome P450 expression and contain POR and b5, but cytochrome P450 expression levels and protein integrity can vary widely and thus influence apparent catalytic activity. Expression systems can influence apparent activity of cytochrome P450 polymorphisms.35 Baculovirus-mediated expression in insect cells is a mature technology and allows simultaneous and flexible expression of individual genetic variants, using separate or combination multiprotein virus cassettes. We used baculovirus systems and a single multiprotein construct to maximize reproducibility of expression and conservatively always included b5 to eliminate a potential confounder if metabolism is influenced by b5.
The results with coexpressed cytochrome P450, POR, and b5 show that every CYP2B6 genetic variant tested had diminished ketamine N-demethylation activity at clinically relevant concentrations compared with wild-type CYP2B6.1. Some variants (CYP2B6.16 and CYP2B6.18) were essentially catalytically incompetent, while others had activity ranging from 15 to 90% of CYP2B6.1. CYP2B6.9 had the lowest activity of the competent variants. For ketamine enantiomers and racemate, the general order of intrinsic clearance was CYP2B6.1 > CYP2B6.4 > CYP2B6.26, CYP2B6.19, CYP2B6.17, CYP2B6.6 > CYP2B6.5, CYP2B6.7 > CYP2B6.9 >> CYP2B6.16, CYP2B6.18. Stereoselectivity of ketamine metabolism by CYP2B6.1 was comparable to previous observations.12,16,18 Specifically, S-ketamine metabolism and Vmax exceeded that of R-ketamine, whether from individual enantiomers or the racemate. Cytochrome P450 allelic variation appeared to minimally influence CYP2B6 stereoselectivity. Also observed was that racemate metabolism was less than that of individual enantiomers, as occurred with human liver microsomes.12 This results from a metabolic enantiomeric interaction whereby one ketamine enantiomer inhibits the metabolism of the other enantiomer.12 Cytochrome P450 allelic variation appeared to also not influence the metabolic enantiomeric interaction, because metabolism was generally greater with S-ketamine compared with the racemate.
The present results can be compared with previous reports. The observed Clint for racemic ketamine metabolism by baculovirus-expressed CYP2B6.1 with coexpressed POR and b5 is comparable to a recent report.21 We observed that ketamine enantiomers metabolism at clinically relevant concentrations and the Clint for CYP2B6.6 was approximately half that for CYP2B6.1. In COS-1 cells, without additional POR or b5, Clint for CYP2B6.6 was also about half that of CYP2B6.1, for both ketamine enantiomers.20 Ketamine metabolism by other CYP2B6 variants has not been reported.
The influence of CYP2B6 polymorphisms on CYP2B6 activity is substrate-dependent, and generalizations about catalytic consequence or clinical implications are not possible. For example, CYP2B6.6 metabolized ketamine with a greater Vmax than CYP2B6.1, while CYP2B6.6 Vmax for methadone was less than CYP2B6.1.32,33 Nevertheless, for both ketamine and methadone, Clint was lower for CYP2B6.6 than CYP2B6.1. With other CYP2B6 substrates, CYP2B6.6 has been described as less active (efavirenz, bupropion, ifosphamide) or more active than CYP2B6.1.24,36,37 Ketamine metabolism by CYP2B6.4 had lower Vmax and Clint than CYP2B6.1, while methadone metabolism by CYP2B6.4 had much higher activity.32,38 CYP2B6.4 activity was lower than CYP2B6.1 for bupropion, cyclophosphamide, and ifosphamide, while it was unchanged or increased for efavirenz.24,37 For both ketamine and methadone, CYP2B6.5 and CYP2B6.9 were less active than CYP2B6.1.32 Indeed, for both drugs, CYP2B6.9 was the least active 2B6 variant studied. In contrast, for ifosfamide, CYP2B6.9 had threefold greater activity than CYP2B6.1.37 CYP2B6.16 and CYP2B6.18 were essentially inactive toward ketamine. Both have a common mutation of 983T>C (I328T). As shown in a computational docking and molecular dynamics simulation analysis, mutation of isoleucine 328, located in the J-helix, to threonine, resulted in structural changes in the C- and I-helices, which are directly involved in ligand/heme recognition, as well as other helices D, E, G, and H.39 The volume of the ligand-binding pocket was reduced significantly to 14 Å3 for CYP2B6.18 compared to 78 Å3 for wild-type CYP2B6.1. This suggests that the I328T mutation caused structural changes that disrupted heme binding; consequently, only inactive CYP2B6.16 and CYP2B6.18 apoproteins were expressed. These isoforms would consequently be expected to be catalytically impaired toward all CYP2B6 substrates.
In general, POR allelic variants had less influence than CYP2B6 variants on ketamine metabolism. Two POR variants were studied due to their high allele frequency, and one variant was studied due to more mechanistic implications. POR*5 (859G>C, A287P) is the most common POR mutation, has an allele frequency of 40% in Europeans,40 and is associated with Antley–Bixler syndrome.30 POR*28 (1508C>T, A503V) allele frequency ranges from 16% in Africans to 28% in Europeans and 39% in East Asians.29 POR 683C>T (P228L) is relatively rare (1 to 2% in Europeans), and is not found in isolation (and does not have an allele designation), but is a component of POR*36 (683C>T, 1508C>T), and was studied because the affected residue is located in the C-terminal side of helix F connecting to the pivotal hinge region. A structural restriction caused by proline residue inside the α helix might influence the connection between helix F and the hinge, which is important for the large conformational change needed for electron transfer.41 The alanine-to-proline mutation in POR.5 occurs at the flavin-adenine dinucleotide–binding domain and affects electron transfer from NADPH to flavin-adenine dinucleotide.40 Ketamine metabolism by CYP2B6.1 with coexpressed POR.5 was 30 to 40% lower than with wild-type POR.1. In cells coexpressing POR.5, CYP2B6.1 catalytic activity (bupropion hydroxylation) was also 30% lower than with wild-type POR.30 POR*5 was also reported to decrease the activity of CYPs 2C9, 3A4, and 17A1.29,30,42 Ketamine metabolism by CYP2B6.1 with coexpressed POR.28 was only 10 to 20% lower compared with wild-type POR. Effects of POR.28 on CYP2B6 activity have not been previously reported. POR.28 did modestly decrease in vitro activity of CYPs 1A2 and 2D6, but not CYPs 3A4 or 2C19, while increasing that of CYP2C9. Clinically, POR*28 had no influence on CYP2B6 activity, as measured by bupropion metabolism, in subject homozygous or heterozygous for CYP2B6*1 and *6.31,43 The influence of POR P228L on ketamine metabolism was similar to that of POR.28. Effects of POR P228L on CYP2B6 activity have not been reported, but POR P228L decreased in vitro activity of CYPs 1A2 and 2C19, but not 3A4.29
Expressed drug-metabolizing enzymes are used routinely to inform on genetic variants. However, there are limitations to the present investigation. Expressed enzymes report only on how amino acid substitutions change catalytic activity, but not the influence of polymorphisms on enzyme expression in the liver. Human liver microsomes may differ from expressed enzymes. For example, expressed CYP2B6.6 has been reported to have increased or decreased activity, but CYP2B6*6 results in erroneous splicing, and microsomes from CYP2B6*6 carriers have markedly decreased enzyme expression and less activity compared to those from wild-type individuals. Liver microsomes from carriers of CYP2B6 and POR variant alleles would be useful to evaluate, as would the carriers themselves.
Consequences of CYP2B6 genetics for clinical ketamine metabolism and pharmacokinetics and implications for patient care remain largely unknown. Ketamine metabolism may affect clinical clearance, inactivation, and/or bioactivation. Expressed CYP2B6.6 has lesser ketamine N-demethylation than CYP2B6.1, and liver microsomes from humans carrying the CYP2B6*6 allele had diminished ketamine N-demethylation compared with CYP2B6*1/*1 genotypes.20 However, clinical ketamine metabolism and clearance was reported to be diminished44 or unchanged7 in CYP2B6*6 carriers. It is, however, well established that CYP2B6*6 does reduce the clinical metabolism and clearance of methadone and efavirenz,38,45,46 and CYP2B6 pharmacogenetics does influence efavirenz dosing.
The CYP2B6 gene is highly polymorphic, causing considerable interindividual variability in protein expression and activity.24 At least 38 variants have been identified, and of all the hepatic cytochrome P450 isoforms, CYP2B6 has one of the highest cumulative population frequencies (nearly 50%) of variant alleles.26 In Africans and South Asians, approximately 20% and 17% of CYP2B6 alleles result in decreased enzyme activity.26 CYP2B6*6 was originally described as the most common variant,47 and, because of generally reduced activity, particularly toward drugs used against human immunodeficiency virus,46 has been the variant most widely studied clinically. More recently, however, a larger population analysis found that CYP2B6*5, *6, *9, *16, and *18 are common variants.26 CYP2B6*9 was the most abundant allele in all five populations, with frequencies ranging from 31% (Africans) to 16% (Europeans). CYP2B6*6 was common in South Asians (16%), while CYP2B6*5 was abundant in Europeans (13%). CYP2B6*16 and CYP2B6*18 occurred almost exclusively in Africans (6 to 7%). Selection of the CYP2B6 variants studied in this investigation was based on allele consequence and frequency. 516G>T (Q172H) and 785G>T (K262R) are common polymorphisms found in many variants. To evaluate their influence, we selected variants with various combinations of these mutations (2B6*4, *6, *7, *9, *16, *19, and *26), even though some have minor population frequency. The other three CYP2B6 variants (*5, *17, and *18) have relatively high allele frequency. Based on the CYP2B6 variants tested and the population allelic frequencies, the present results suggest CYP2B6*6, CYP2B6*9, CYP2B6*16, and CYP2B6*18 as the most useful polymorphisms to evaluate clinically for potential effects on ketamine disposition. Similarly, based on POR variants tested, POR*5 would be the most useful variant to evaluate clinically.
In summary, CYP2B6 and P450 oxidoreductase genetic variants have diminished ketamine N-demethylation activity, without affecting stereoselectivity of metabolism. The order of ketamine enantiomers and racemate metabolism, based on intrinsic clearance, was generally CYP2B6.1 > CYP2B6.4 > CYP2B6.26, CYP2B6.19, CYP2B6.17, CYP2B6.6 > CYP2B6.5, CYP2B6.7 > CYP2B6.9. CYP2B6.16 and CYP2B6.18 were inactive. For P450 oxidoreductase, the order was POR.1 > POR.28, P228L > POR.5. The consequences of CYP2B6 and POR polymorphisms for ketamine pharmacokinetics and therapeutic effects are unknown and would require clinical evaluation.
Supported by National Institutes of Health (Bethesda, Maryland) grant No. R01DA14211 and by the Department of Anesthesiology Russel B. and Mary D. Shelden fund, Washington University, St. Louis, Missouri.
Dr. Kharasch is the Editor-in-Chief of Anesthesiology. The authors declare no competing interests.