Altered metabolism of synthetic cannabinoid JWH-018 by human cytochrome P450 2C9 and variants

Abstract

Synthetic cannabinoids (SCBs), synonymous with ‘K2’, ‘Spice’ or ‘synthetic marijuana’, are psychoactive drugs of abuse that frequently result in clinical effects and toxicity more severe than those classically associated with Δ⁹-tetrahydrocannabinol such as extreme agitation, hallucinations, supraventricular tachycardia, syncope, and seizures. JWH-018 is one of the earliest compounds identified in various SCB products, and our laboratory previously demonstrated that JWH-018 undergoes extensive metabolism by cytochromes P450 (P450), binds to, and activates cannabinoid receptors (CBRs). The major enzyme involved in the metabolism of JWH-018 is CYP2C9, a highly polymorphic enzyme found largely in the intestines and liver, with *1 being designated as the wild type, and *2 and *3 as the two most common variants. Three different major products have been identified in human urine and plasma: JWH-018 (ω)-OH, JWH-018 (ω-1)-OH(R), and JWH-018 (ω-1)-OH(S). The (ω-1)-OH metabolite of JWH-018 is a chiral molecule, and is thus designated as either (ω-1)-OH(R) or (ω-1)-OH(S). Here, in vitro enzyme kinetic assays performed with human recombinant CYP2C9 variants (*1, *2, and *3) revealed that oxidative metabolism by CYP2C9*3 resulted in significantly less formation of (ω)-OH and (ω-1)-OH metabolites. Surprisingly, CYP2C9*2 was roughly 3.6-fold more efficient as the CYP2C9*1 enzyme based on V_max/K_m, increasing the rate of JWH-018 metabolism and allowed for a much more rapid elimination. These results suggest that genetic polymorphisms of P450 enzymes result in the production of varying levels of biologically active JWH-018 metabolites in some individuals, offering a mechanistic explanation for the diverse clinical toxicity often observed following JWH-018 abuse.

1. Introduction

SCBs, present in abused products known as ‘K2’ or ‘Spice’, were originally synthesized to investigate potential therapeutic utility of cannabinoid receptor (CBR) ligands. Since the early 2000s, more than 150 SCBs have been synthesized by clandestine labs and marketed to vulnerable populations as safe and legal alternatives to marijuana. Numerous serious adverse effects have been associated with these compounds such as seizures, severe tachycardia, psychosis, and even death [1–5], and the consequences of acute and chronic SCB abuse have been examined in studies ranging from basic science reports to clinical cases. Although production and usage of SCB products have significantly increased over the years, few studies have delineated mechanisms of toxicity or established direct mechanisms responsible for the increased toxicity of these high-affinity, high-efficacy purported ‘marijuana substitutes’ at CB1 and/or non-CB1-cannabinoid receptor targets [6–8].

CBRs are G-coupled protein receptors (GCPRs) are responsible for modulating many important physiological processes and are classified into two receptor subtypes; CB1 and CB2. CB1 receptors are found primarily in the central nervous system, whereas CB2 receptors are present in highest abundance in immune cells [8]. Ligands for these receptors are called cannabinoids. SCBs present in K2/Spice products, and many of their associated metabolites, exhibit high affinity and activity for both CBRs [6,7]. For example, the oxidative metabolism of JWH-018 results in a number of metabolites that retain biological activity towards CBRs, including JWH-018 (ω)-COOH, JWH-018 (ω)-OH, JWH-018 (ω-1)-OH(S), and JWH-018 (ω-1)-OH(R) (Fig. 1) [2,9].

Fig. 1.

Schematic representation of JWH-018 and oxidized metabolites. CYP2C9 metabolized JWH-018 to JWH-018 (ω)-OH, JWH-018 (ω-1)-OH(R), JWH-018 (ω-1)-OH(S) and JWH-018 (ω-1)-COOH. The structure of THC was included for comparison.

There is very little information concerning the mechanism of action of SCBs and how these compounds interact with CBRs. In addition, whether Phase I and/or II metabolism of SCBs contribute to toxicity observed following SCB use [10] is unknown. SCBs undergo extensive oxidation reactions by P450s, whereby oxidized metabolites are formed and then serve as substrates for a second metabolic phase, namely glucuronidation and/or sulfation [6,11].

Since SCBs undergo extensive metabolism to form multiple active metabolites, genetic polymorphisms of P450 and UDP–Glucuronosyltransferase (UGT) enzymes may contribute to SCB toxicity. CYP2C9 is one of the major cytochrome P450 enzymes responsible for JWH-018 metabolism. Importantly, CYP2C9 has over 35 known polymorphisms, makes up 10% of the expressed P450 enzymes, and is responsible for the metabolic clearance of up to 15–20% of all drugs undergoing Phase I metabolism; including acenocoumarol, candesartan, celecoxib, fluvastatin, glyburide, ketamine, methadone, phenytoin, and tolbutamide [12–15].

Several clinically relevant polymorphisms that affect the activity of CYP2C9 have been reported. Two of the most common variations are the phenotypes CYP2C9*2 (a cysteine substitutes for arginine at residue 144) and CYP2C9*3 (a leucine substitutes for isoleucine at residue 359); both reduce enzymatic activity [16]. Substrate affinity is not affected substantially by the *2 haplotype, but the maximum rate of metabolism (V_max) is reduced to approximately 50% of that of the WT [12]. The catalytic activity of the *3 haplotype is significantly reduced for most CYP2C9 substrates due to an increase in K_m and a reduction in V_max [12]. It is possible that individuals carrying variants, such as CYP2C9*3, will be at a higher risk of toxicity from SCBs than those carrying other genotypes.

Consequently, we hypothesize that genetic variability in oxidative metabolism catabolized by CYP2C9 can be responsible for toxic effects observed in some SCB users. Although it is difficult to correlate the poor metabolism of SCBs by P450s with increased potential to abuse by the users, we can speculate that in those individuals with genetic variations in CYPs, the native JWH-018 stays in the body not detoxified by metabolic enzymes, therefore it can activate CBR receptors for a much longer time. Consequently, this prevents the receptors from accepting endogenous CB ligands and alters normal activity of those receptors. This can potentially be related to the abnormal, self-destructive behavior of the users. More research is needed to correlate behavior of poor metabolizers to abusive activity, but it is a logical assumption. Moreover, the present studies will provide essential direction for pharmacokinetic and pharmacogenetic investigations of current and future classes of these virtually unknown compounds.

2. Materials and methods

2.1. Materials

All chemicals used in this study were at least reagent grade. JWH-018 ((1-pentyl-1H-indol-3-yl)-1-naphthalenyl-methanone) was purchased from Cayman Chemical (Ann Arbor, MI). Unless otherwise specified, all chemicals and reagents were purchased from either Sigma-Aldrich (St. Louis, MO) or Thermo Fisher Scientific (Waltham, MA). Ethyl alcohol (100%) was purchased from AAPER (Shelbyville, KY). Recombinant CYP2C9*1, CYP2C9*2 and CYP2C9*3 were expressed in baculovirus-infected insect cells and purchased from BD Biosciences (San Jose, CA).

2.2. Steady-state enzyme kinetic assays

Incubation conditions were optimized for time and protein concentration, and all reactions were performed within the linear range of metabolite formation with less than 5% of the substrate being consumed during the course of the reaction. Other than substrate concentrations and incubation times, reaction mixture compositions and analytical methods were identical to those described for the above screening assays. Kinetic parameters for CYP2C9*1, CYP2C9*2 and CYP2C9*3 were carried out with proteins (2 pmol each) in the presence of various concentrations of the substrate (0.05–100 μM) for 10 min at 37 °C [8].

2.3. LC-MS/MS analysis

Quantification of analytes of interest was performed using positive ion electrospray ionization LC-MS/MS methods developed previously for chiral analysis of synthetic cannabinoid metabolites [17]. For these studies, analytical standards (0.50–100 ng/ml) were matrix-matched in the final in vitro reaction mixture (ethanol-water-0.1 M phosphate buffer, pH 7.4, 50:30:20%). Analytes of interest were chromatographically separated using a Phenomenex Lux 3 μm Cellulose-3 analytical column (150 × 2.0 mm). Mobile phases consisted of 60% A (20 mM ammonium bicarbonate) and 40% B (acetonitrile) ramped to 5% mobile phase A over 10 min. The conditions were held for an additional 2 min before returning to initial conditions over 3 min and holding constant for 1 min for column reequilibration. The TurboIonSpray source voltage was 2500 V, and source temperature was maintained at 600 °C. Nitrogen gas pressures for the GS1 and GS2 source gases, curtain gas, and collision gases were 55.0 cm/s, 55.0 cm/s, 35.0 cm/s, and “high,” respectively. All metabolic products were confirmed by matching retention times of analytical standards and deuterated internal standards. Mass spectra and NMR data for each metabolite have been reported previously [8,17,18].

2.4. Kinetic data and statistical analysis

Curve-fitting and statistical analyses were performed using GraphPad Prism (version 4.0b; GraphPad Software Inc., San Diego, CA). Kinetic constants were obtained by fitting experimental data to kinetic models using the nonlinear regression (Curve Fit) function. The fit of the data for each model was assessed from the S.E., 95% confidence intervals, and R² values. Kinetic curves were also analyzed as Eadie-Hofstee plots to support kinetic models. Kinetic constants are reported as the mean ± S.E. of triplicate experiments. Kinetic constants were obtained by fitting experimental data to the following kinetic models using the nonlinear regression (curve fit) function: 1. Michaelis-Menten equation for one-enzyme model

$$\mathrm{v}=\mathrm{Vmax}\times [\mathrm{S}]$$

$${\mathrm{K}}_{\mathrm{m}}+[\mathrm{S}]$$

2. Hill equation, which describes sigmoidal autoactivation kinetics, where S₅₀ is the substrate concentration at 50% V_max (analogous to K_m in Michaelis-Menten kinetics) and n is the Hill coefficient, which can be considered to be a measure of autoactivation and reflects the extent of cooperativity among multiple binding sites [19].

3. Results

3.1. Steady-state enzyme kinetic assays

The kinetic constants of the recombinant P450 enzymes were analyzed, and as shown in Fig. 2, the kinetic profiles followed classical Michaelis-Menten kinetics. The V_max value (Table 1) for the JWH-018 (ω)-OH metabolite was almost 2-fold greater for the CYP2C9*2 variant than for the WT enzyme (101.9 ± 3.24 and 52.58 ± 3.29 pmol/min/nmol respectively), while the V_max value for the CYP2C9*3 variant was significantly less at 6.02 ± 0.31. This trend was observed with the JWH-018 (ω-1)-OH(S) and (R) metabolites as well. The JWH-018 (ω-1)-OH(S) metabolite exhibited a V_max value of 14.61 ± 0.97 pmol/min/nmol for CYP2C9*1, 26.62 ± 1.03 for CYP2C9*2 and 2.75 ± 0.23 for CYP2C9*3. Metabolism of the JWH-018 (ω-1)-OH(R) metabolite revealed a V_max of 19.63 ± 1.54 pmol/min/nmol for CYP2C9*1, 38.54 ± 1.49 for CYP2C9*2 and 2.62 ± 0.45 for CYP2C9*3. V_max values resulting from metabolism by the CYP2C9*3 variant were significantly lower than the other 2 variants, as observed by an almost 2-fold increase for CYP2C9*2 as compared to CYP2C9*1. Reactions resulting in the JWH-018 (ω)-COOH metabolite were not detected.

Fig. 2.

Michaelis-Menten values of hydroxylation of JWH-018 by recombinant CYP2C9 and its variants. Steady-state enzyme kinetics assays with A) CYP2C9*1, B) CYP2C9*2 and C) CYP2C9*3 were measured by incubating recombinant proteins with increasing concentrations of JWH-018) at a constant concentration of the NADPH regenerating system. Curve fits and kinetic constants were determined using GraphPad Prism 4 software. Kinetic constants were reported as the mean ± S.E. of triplicate experiments.

Table 1.

Oxidation kinetics for JWH-018 metabolites.

	Metabolites	Vmax (pmol product/min/nmol CYP)	Km (μM)	Vmax/Km (μl/mg protein/min)
CYP2C9*1	(ω)-OH	52.58 ± 3.29	0.90 ± 0.32	58.42
	(ω)-COOH	ND	ND	ND
	(ω-1)-OH(S)	14.61 ± 0.97	0.89 ± 0.33	16.40
	(ω-1)-OH(R)	19.63 ± 1.54	1.15 ± 0.49	17.07
CYP2C9*2	(ω)-OH	101.9 ± 3.24	0.48 ± 0.10	212.29
	(ω)-COOH	ND	ND	ND
	(ω-1)-OH(S)	26.62 ± 1.03	0.58 ± 0.14	45.90
	(ω-1)-OH(R)	38.54 ± 1.49	0.65 ± 0.15	59.29
CYP2C9*3	(ω)-OH	6.02 ± 0.31	0.68 ± 0.21	8.85
	(ω)-COOH	ND	ND	ND
	(ω-1)-OH(S)	2.75 ± 0.23	0.43 ± 0.24	6.40
	(ω-1)-OH(R)	2.62 ± 0.45	2.61 ± 2.52	1.00

Parameters were determined from the fit of initial velocities to a Michaelis-Menten kinetic model. All statistical calculations were performed using GraphPad Prism. Kinetic constants were reported as the mean ± S.E. of triplicate experiments.

K_m values (Table 1), however, did not follow this same trend. For the JWH-018 (ω)-OH metabolite, the K_m value for the CYP2C9*1 variant was almost twice that of the CYP2C9*2 variant (0.90 ± 0.32 μM and 0.48 ± 0.10 μM respectively), while the value for the CYP2C9*3 variant was 0.68 ± 0.21 μM. The JWH-018 (ω-1)-OH(S) metabolite exhibited K_m values of 0.89 ± 0.33 μM, 0.58 ± 0.14 μM and 0.43 ± 0.24 μM for CYP2C9 variants *1, *2 and *3. For JWH-018 (ω-1)-OH(R), the values were 1.15 ± 0.49 μM for CYP2C9*1, 0.65 ± 0.15 μM for CYP2C9*2 and 2.61 ± 2.52 μM with CYP2C9*3, surpassing the other two variants. As mentioned previously, reactions resulting in the JWH-018 (ω)-COOH metabolite were not detected.

4. Discussion

This study provides additional insight into the potential importance of genetic variation in the CYP2C9 enzyme for metabolism of JWH-018, an SCB present in abused K2/Spice products. One possibility that might contribute to the significant differences in reported clinical effects and toxicity of SCBs relative to marijuana is genetic variation found in essential drug metabolizing enzymes such as P450s.

JWH-018 is an aminoalkylindole that was originally synthesized to study the function of CB1 and CB2 receptors, and it exhibits high affinity for both receptor subtypes [8]. Since it was one of the initial SCBs identified in K2/Spice products, and a resurgence in its use is occurring in Europe, understanding the metabolism of this compound is of vital importance. Reported cases of human toxicity following exposure to SCBs may represent the combined actions of a complex mixture of multiple SCBs and their active metabolites [6,8]. On the other hand, some effects and/or the duration of drug effects observed in certain SCB users could be related to a genetic variant of CYP2C9 metabolism subsequent adverse psychoactive and physiological consequences such as stroke, convulsions and death [3–5].

CYP2C9 is a highly polymorphic enzyme, expressed in highest abundance in the intestine and the liver, with *1 designated as the wild type and *2 and *3 being two of the most common variants. CYP2C9 is responsible for metabolizing various compounds including procarcinogens, alcohols and many different drugs such as glipizide, ibuprofen and warfarin [13,20,21], and there is evidence for drug-drug interactions occurring and being influenced by these variants [20]. The known polymorphisms of CYP2C9 appear to contribute to a genetic phenotype, characterized by increased susceptibility of certain ethnic groups to drugs metabolized by this P450 enzyme [20,22]. In the present study, extreme differences in the rate of metabolism, designated by V_max values, were observed among the enzymes examined, as indicated by a wide variation in the ability of the CYP2C9 variants to metabolize JWH-018. When exposed to JWH-018, CYP2C9*2 biosynthesized oxidized metabolic products, including JWH-018 (ω)-OH, JWH-018 (ω-1)-OH(R), JWH-018 (ω-1)-OH(S) and JWH-018 (ω)-COOH, almost twice as quickly as CYP2C9*1; however, the individual rates for each of the different metabolites varied. This increased the rate of JWH-018 metabolism and allowed for a much more rapid elimination (Fig. 2, Table 1).

At this time, we don’t have an explanation for this phenomenon; moreover, this occurrence has not been found with other substrates evaluated for the hydroxylation by this variant [22–24]. On the other hand, we have observed a similar phenomenon for UGT1A10 variants involved in glucuronidation reactions of phenols and several derivatives of estrogens [25,26]. The detailed analysis of enhanced activity and affinity of F93A, a mutant of UGT1A10, with a variety of substrates led to the conclusion that the single amino acid substitution resulted in the modification of the binding site and is responsible for this enhancement in activity and affinity for this variant catalyzing glucuronidation. As mentioned before, CYP2C9*2 is generated by the substitution of cysteine for arginine. It is possible that the addition of an amino acid with a positive charge could be responsible for its increased activity and affinity with the JWH-018 substrate observed in the present work [27].

In contrast, CYP2C9*3 lowered the rate of JWH-018 metabolism, as demonstrated in Fig. 2, Table 1, by the very low levels of JWH-018 metabolites. Based on these observations, it might be predicted that drug abusers possessing the CYP2C9*3 variant would have minimal ability to metabolize JWH-018, resulting in an extraordinary susceptibility to the toxicological effects of this compound due to a marked increase in half-life in the body along with the presence of biologically active metabolites [2,9].

Surprisingly, the affinity of the enzyme for the substrate, or the K_m, was also altered between the CYP2C9*1 and CYP2C9*2 variants (Fig. 2, Table 1); K_m values for CYP2C9*2 were consistently almost half of those for CYP2C9*1, suggesting an increased affinity to JWH-018. This resulted in the metabolites being excreted from the body faster. However, the CYP2C9*3 variant exhibited a relatively high K_m, indicative of a decreased affinity of JWH-018 to the enzyme controlling its metabolism. As such, not only do metabolites remain in the body longer due to a slowed metabolism, these compounds are also unable to bind to the enzyme that is largely responsible for xenobiotic metabolism, leading to an increase in toxicity for people carrying the CYP2C9*3 variant. Such increased toxicity is especially concerning considering that the metabolites of JWH-018 have been demonstrated to maintain equivalent biological activity when compared to the parent compound.

It is also important to take into consideration that polymorphism both in CB1 and CB2 might be responsible for toxicity associated with use of SCBs. Single nucleotide polymorphism in both receptors (references), as well as occurrence of CB1 variants was described in the literature. The following references are related to the polymorphism of CBs and investigation of the target receptor pharmacological polymorphism, represents a very interesting, novel area of investigation [28]. Understanding the mechanisms and pathways involved in the metabolism of a compound is essential to discovering a potential antidote or treat associated adverse effects. Current treatment for SCB toxicity is purely supportive, with a lack of clinically available efficacious antidotes for serious life-threatening situations. Although much useful information has been gained concerning SCB metabolism, unfortunately, many important unsolved issues remain. For example, why do some people appear to be more susceptible to the propsychotic actions of SCBs than others? Data presented here suggest that the large variability in response to SCB abuse in the population may be largely due to genetics. For example, the amount of each enzyme expressed may vary significantly from person to person, affecting the ability to metabolize a wide variety of drugs at varying dosages. Mutations in these enzymes can affect this ability even more. SCB abuse is a growing global problem, with abuse increasing at an alarming rate among adolescents and young adults.

Although scheduling of SCBs has been a priority by the DEA since 2010, the inability for legislation to stay ahead of the production of novel SCBs continues to fail in preventing the ongoing abuse of these very dangerous and occasionally deadly drugs. Due to their varied and potentially severe clinical effects and lack of an antidote or other specific treatment intervention, understanding the cause of their toxicity has become a pressing issue that must be resolved.

Abbreviations:

Δ⁹-THC

(−)-Δ⁹-Tetrahydrocannabinol

P450

cytochrome P450

CYP

specific P450

UGTs

uridine diphosphate (UDP)-glucuronosyltransferases

LC-MS/MS

Liquid chromatography-tandem mass spectrometry

SCB

synthetic cannabinoid

CBR

cannabinoid receptor

CB1

cannabinoid receptor 1

CB2

cannabinoid receptor 2