Gas chromatography – mass spectrometry of JWH-018 metabolites in urine samples with direct comparison to analytical standards

Abstract

JWH-018 (1-pentyl-3-(1-naphthoyl)indole) is one of numerous potential aminoalkylindoles contained in products marketed as ‘K2’ or ‘Spice’. Investigation of the urinary metabolites from consumption of these compounds is important because they are banned in the United States and many European countries. An efficient extraction procedure and gas chromatography – mass spectrometry (GC-MS) method were developed for detection of ‘K2’ metabolites in urine from individuals suspected of using these products. Analytical standards were used to elucidate the structure-specific mass spectral fragmentations and retention properties to confirm proposed identifications and support quantitative studies. A procedure for the synthesis of one of these metabolites (5-hydroxypentyl JWH-018) was also developed. Results are comparable to existing LC-MS/MS methods, with the same primary metabolites detected. The specific metabolite hydrolysis products include 4-hydroxpentyl, 5-hydroxypentyl, and N-pentanoic acid derivatives.

1. Introduction

Since 2004, herbal mixtures under the brand names ‘Spice’, ‘K2’, and others have been sold via the internet and in ‘headshops’. Although these products are marketed as incense, the blends have been smoked in a manner similar to tobacco products giving users cannabis-like effects comparable to marijuana [1,2]. These psychoactive effects are a result of synthetic cannabinoids, including aminoalkylindole (AAI) and cyclohexylphenol (CP) compounds, added to the mixtures. One of the first and most commonly reported additives is JWH-018, an AAI with binding affinity to the CB₁ and CB₂ cannabinoid receptors [1–3].

Analysis of synthetic cannabinoids is relevant from both a clinical and a law enforcement perspective. Several studies have investigated their detection in seized material and in bodily fluids [1,3–11]. Reliable detection of AAI’s and metabolites in a variety of substrates is critical because the numbers of severe episodes of intoxication are increasing at healthcare facilities. In addition, five synthetic cannabinoids (JWH-018, JWH-073, JWH-200, CP-47,497, and cannabicylcohexanol) have been classified as Schedule 1 substances by the U.S. Drug Enforcement Administration (DEA) [12,13]. These compounds, and other JWH analogues, are also banned in many European countries [14].

Several methods have been reported for detection of JWH metabolites in urine samples using liquid chromatography and tandem mass spectrometry (LC-MS/MS) [6–9]. The development of a second technique, using a different approach, would be useful for confirmation. We elected to develop a GC-MS method to complement the existing LC-MS approach because it has higher specificity in both the chromatographic and the mass spectral detection steps, the obtained electron impact (EI) spectra are reproducible allowing them to be searched and matched in a library database, and the instrument is usually less expensive [15,16]. The GC chromatographic column provides an order of magnitude more theoretical plates for separation, and the mass spectral ionization step (EI) produces suppression-free spectra with higher structure specific information [15–17]. The advantage of the LC-MS/MS technique is that it is uniquely amenable to direct analysis of aqueous solutions and is better suited for non-volatile compounds [15, 17,18]. With suitable extraction of the problematic components beforehand, GC-MS should provide a powerful and complementary approach for the characterization of these metabolites after separation from the urine matrix. This approach has not yet been extensively studied due to a lack of method validation studies (i.e. detection limits, recovery efficiency, and quantification) and the difficulty in obtaining proper analytical standards [10,11]. These limitations permit only tentative or semi-quantitative assignments of suspected urinary metabolites by GC-MS. Additionally, identification of some metabolites is difficult because of the presence of isomeric compounds with similar mass spectral fragmentation patterns and retention properties. This can lead to differences in the reported identity of metabolites from the same parent compounds. For example, the main urinary metabolites of JWH-018 reported by Sobolevksy et al. involve monohydroxylation on the indole ring [10] whereas Grigoryev et al. identify monohydroxylation on the pentyl chain [11]. The synthesis of the appropriate analytical standards is probably the best approach to resolve such differences. Even with standards, however, differences for the main metabolites and their relative abundance have been reported by LC-MS/MS. Chimalakonda et al. identified three metabolites with the following abundance ranking: 4-hydroxypentyl > 5-hydroxypentyl > N-pentanoic acid derivatives [6]. Analysis of urine samples with a similar method by ElSohly et al. identified three main metabolites with the order as N-pentanoic acid > 5-hydroxypentyl > 6-hydroxyindole derivatives [19].

The aim of this present work was to develop a GC-MS method that is analogous and complementary to the existing LC-MS/MS method. A GC-MS method would be of significant value for confirmation of LC-MS/MS results with regard to the main urinary metabolites of JWH-018. This is especially important given the differences between detected metabolites from different studies as described above. Development of a GC-MS method also provides an alternative technology that may be better suited to the existing equipment in a particular laboratory. GC-MS is a fundamental tool in forensic toxicology and was the only permitted method for urine drug testing from 1988 to October 2010 under the Mandatory Guidelines for Federal Workplace Drug Testing Programs. New guidelines now permit alternative technologies (LC-MS, GC-MS/MS, and LC-MS/MS) to be used as long as the methods are scientifically validated [20–22]. In this current report, three metabolite hydrolysis products (4-hydroxypentyl, 5-hydroxypentyl, and a carboxylated derivative of JWH-018) and the native compound of JWH-018 were used as reference standards for comparison to urine samples from suspected ‘K2’ users. These particular metabolites were chosen because of their reported abundance in urine samples from ‘K2’ users [6–9]. A method meeting criteria established for proof of identity was developed based on chromatographic retention properties and monitoring four diagnostic ions for each compound [23,24]. The efficiency of extracting the metabolites from urine samples using solid phase extraction (SPE) and calculation of detection limits were also monitored.

2. Materials and methods

2.1. Reagents

The reagents and solvents were obtained from EMD chemicals (Gibbstown, NJ), Sigma Aldrich (St. Louis, MO), and TCI America (Portland, OR). N,O-Bis(trimethylsily) trifluoroacetamide + 10% trimethylchlorosilane (BSTFA + 10% TMCS) was purchased from Regis Technologies (Morton Grove, IL). SPE disposable cartridges (octadecyl C18) were manufactured by J.T. Baker (Phillipsburg, NJ). A certified negative control urine sample was obtained from Biochemical Diagnostics Inc. (Edgewood, NY). Because of the lack of availability at the time of analysis, the 5-hydroxypentyl JWH-0108 metabolite was synthesized as reported below. As the study progressed, two additional standards (4-hydroxypentyl and N-pentanoic acid JWH-018 derivatives) were made commercially available by Cayman Chemical (Ann Arbor, MI). JWH-018 was synthesized as reported in the literature [25].

2.2. Synthesis of 5-hydroxypentyl JWH-018 metabolite

This JWH-018 metabolite was synthesized as indicated in Scheme 1: The hydroxyl group of 5-bromopentanol (1) was protected with tert-Butyldimethylsilyl chloride (TBDMSCl) to give product 2 [26]. A nucleophilic substitution reaction was used to obtain product 3 by reaction of the protected bromopentane with indole [27]. Friedel-Crafts acylation of the protected pentylindole with 1-naphthoyl chloride and Me₂AlCl afforded product 4 [25]. Deprotection of product 4 with tetrabutylammonium fluoride (TBAF) in tetrahydrofuran (THF) produced product 5 in 71 % yield [26]. All molecular structures were confirmed using GC-MS and ¹H-NMR. Conditions for GC-MS analysis are discussed below. ¹H-NMR spectra were recorded on a Bruker 300 MHz spectrometer using CDCl₃ as the solvent with TMS as the internal standard.

Scheme 1.

Reagents and conditions for synthesis of 5-hydroxypentyl JWH-018 metabolite: (a) TBDMSCl, imidazole, DMF, 0°C to room temperature; (b) DMF, indole, KOH; (c) 1-naphthoyl chloride, CH₂Cl₂, Me₂AlCl, 0°C; (d) THF, TBAF.

2.2.1. 5-bromopentoxy-tert-butyl-dimethyl-silane

To a solution of imidazole (1.2 g, 17.6 mmol) in 10 mL DMF was added 5-bromopentan-1-ol (1) (2.0 g, 12.0 mmol) and 1 M TBDMSCl in THF (15 mL, 15 mmol) at 0 °C under N₂. Stirring was continued overnight at room temperature. The mixture was diluted with Et₂O and washed with 1 N HCl, water, and brine. The filtrate was then dried with Na₂SO₄ and concentrated. Product 2 (2.4 g, 71 %) was afforded after purification by silica gel column chromatography (petroleum ether/Et₂O, 50:1) as a colorless oil [26].

2.2.2. tert-butyl-(5-indol-1-ylpentoxy)-dimethyl-silane

To a solution of 2 (8.4 g, 30 mmol) in 50 mL DMF was added indole (1.2 g, 10.2 mmol) and ground KOH powder (0.6 g, 10.7 mmol). The mixture was stirred overnight at room temperature. Water (100 mL) was added, and the product was extracted into ether (3 × 50 mL). The ether extracts were washed with water and dried with MgSO₄. After concentrating the solution, product 3 (2.0 g, 62 %) was isolated through chromatography (petroleum ether followed by petroleum ether/ether, 10:1) as an oil [27].

2.2.3. [1-[5-[tert-butyl(dimethyl)silyl]oxypentyl]indol-3-yl]-(1-naphthyl)methanone

To a stirred solution of 3 (0.16 g, 0.50 mmol) in 1.5 mL dry CH₂Cl₂ at 0 °C under N₂ was added dropwise Me₂AlCl (1 M in hexanes, 0.75ml, 0.75 mmol). After stirring the mixture for 30 min at 0 °C, 1-naphthoyl chloride (0.12 g, 0.63 mmol) in 1.5 mL of CH₂Cl₂ was added. The reaction mixture was stirred at 0 °C until the reaction was complete as indicated by thin layer chromatography (TLC) analysis (approximately 1 hr). The mixture was poured into iced 1 M aqueous HCl and extracted with CH₂Cl₂ (3 × 50 mL). The extracts were washed with aqueous NaHCO₃ and then dried with MgSO₄. After evaporation of the solvent, chromatography (petroleum ether/ethyl acetate, 9:1) was used to obtain product 4 (0.17 g, 72 %) as an off-white solid [25].

2.2.4. [1-(5-hydropentyl)indol-3-yl]-(1-naphthyl)methanone

To a solution of 4 (4.3 g, 9.1 mmol) in 10 mL of THF was added 1 M TBAF in THF (18.4 ml, 18.4 mmol). The reaction mixture was stirred for 2 hr and then quenched with MeOH. The mixture was washed with water and brine then dried with Na₂SO₄. After concentrating the solution, chromatography was used to give product 5 (2.3 g, 71 %) as an off-white solid [26].

2.3. Urine samples of suspected ‘K2’ users

The urine samples analyzed in this study were collected from three individuals through a drug testing program operated by Employee Screening Management (Fayetteville, AR). No information regarding prior drug history or admittance to smoking any ‘K2’ products of the participants was provided. Samples were supplied with an assigned number with no personal information exchanged. Samples were first analyzed by the Arkansas Department of Health, Public Health Laboratory using their established LC-MS/MS method [7].

2.4. Preparation of urine samples

A 1-mL urine sample containing an internal standard of bisphenol A (BPA, 900 ng/mL) was evaporated to ~0.25 mL under N₂ at room temperature. Hydrolysis of glucuronic acid conjugates was completed by addition of 0.5 mL trifluoroacetic acid (TFA) to the dried residue and heating the sample for 40 min. at 100 °C. After cooling to room temperature, ammonium hydroxide (~4.5 mL, 28–30 %) was added to adjust the pH to ~9 as monitored with pH paper.

SPE cartridges (C18) were conditioned prior to analysis by rinsing the column with 3 mL methanol followed by 10 mL distilled water. The urine sample, prepared as described above, was then passed through the cartridge followed by rinsing with 10 mL distilled water. After discarding the above washing solutions, the analytes of interest were eluted with 4 mL methanol. The eluted methanol was evaporated to dryness at 60 °C (Centrivap Concentrator, Labconco, Kansas City, MO). The residue was then dissolved in 150 μL DMF and derivatized by addition of 150 μL BSTFA + 10 % TMCS. After heating for 25 min. at 70 °C, the samples were analyzed by GC/MS. To avoid contamination, a separate SPE cartridge was used for each sample.

2.5. Method parameters

2.5.1. Specificity

Five urine samples, collected over five consecutive days, were obtained from a healthy individual and prepared as described above. Samples were analyzed using GC-MS to determine if any interfering peaks at the retention time of the internal or analytical standards were present.

2.5.2. SPE extraction recovery

The efficiency of the SPE method was determined by evaluating a reference sample before and after extraction. A sample of synthesized metabolite (100 μg/mL), without any SPE preparation, was compared to a negative control urine sample spiked with the same concentration of metabolite. After extracting the urine sample according to the above procedure, both samples were derivatized and analyzed by GC-MS. The chromatographic areas of both samples were compared and % recovery was calculated as follows: (peak area of sample after extraction/peak area of sample before extraction) * 100.

2.5.3. Reproducibility

Reproducibly of the method was calculated by addition of synthesized metabolite (500 ng/mL) to three samples of negative control urine. Samples were then prepared and extracted using SPE as described above.

2.5.4. Detection limits

Detection limits were calculated using the slope of the calibration curve (m) and the standard deviation of the response (σ) based on repeated blank measurements (n=8). The limit of detection (LOD) was calculated using criteria of 3 σ/m and the limit of quantification (LOQ) as 10 σ/m [28].

2.6. GC-MS analysis of synthesized metabolite and urine samples

GC-MS analysis was performed on a Varian 450-GC coupled to a 320-MS triple quad mass spectrometer (Bruker Daltonics, Billerica, MA). Separation was achieved with a Phenomenex Zebron ZB-5HT Inferno column (30M × 0.25ID). The initial column temperature was set at 180 °C and was increased at a rate of 20 °C/min to 320 °C (held for 12 min). One μL injections were done with the injector at 310 °C in the split mode (10:1) with the transfer line temperature set at 320 °C. The mass spectrometer was operated under EI in both full scan and SIM modes. Four ions were monitored for each analyte and are discussed below.

3. Results and discussion

The metabolism of JWH-018 appears to be consistent with cytochrome P450 oxidation followed by transformation to glucuronic acid conjugates [29]. Because of the variability of metabolic processes and differences in the time frame of urinary retention among individuals, the free metabolites in urine may include both the conjugates and their hydrolysis products. Deliberate hydrolysis of samples with an enzyme or acid then yields complete conversion of the free non-conjugated metabolite and also produces chemical structures more amenable to GC separation. The addition of an acid (i.e. TFA) and neutralization with ammonium hydroxide works effectively for this process [30].

Synthesis of the 5-hydroxypentyl JWH-018 metabolite is illustrated in Scheme 1. The route of synthesis is similar to JWH-018 but includes substitution of bromopentane with 5-hydroxybromopentane and protection and deprotection of the hydroxyl end group. Each product was confirmed through ¹H-NMR and GC-MS analysis with assignments in agreement with those reported in the literature [6–10]. It is important to note that step d of the synthesis could be omitted as previous studies by the author’s (not shown) have indicated that silylation of the hydroxyl group is necessary for better detection limits using GC-MS. The OTBDMS compound (product 4) could then be used for analysis rather than derivatizing the final product.

SPE extraction recovery was completed using certified negative control urine samples spiked with the synthesized metabolite. Using the specified SPE procedure, a recovery of 91 % was achieved. It should be noted that this efficiency is based upon recovery of the free hydroxylated product and not the glucuronidated metabolite excreted from the body. Hydrolysis efficiency and recovery could be determined more precisely if glucuronidated JWH-018 compounds were more readily available. Efficiencies of > 90 % for acid and enzyme hydrolysis of other glucuronidated compounds (i.e. morphine-6-glucuronide to morphine) have been reported using similar conditions [30]. Addition of TFA to the metabolite hydrolysis product, however, demonstrates no decomposition or dehydration is observed. This is indicated by the high % recovery and observation of no additional products in the GC chromatogram. A reproducibility of 12 % (% CV, n = 3) was determined from recovery of synthesized metabolite in spiked samples. The LOD was calculated as 2.8 ng/mL and the LOQ as 9.24 ng/mL with linearity (R² = 0.9989) at 10 times the concentration expected in urine samples

A typical chromatogram containing native JWH-018 (N1), 4-hydroxypentyl (M1), 5-hydroxpentyl (M2), and N-pentanoic acid (M3) JWH-018 is shown in Fig. 1. Baseline resolution is achieved between compounds with elution of all analytes within 14 min. A mass spectrum for a TMS derivative of M2 is given in Fig. 2a with the structures and fragmentation for all metabolites in Fig. 2b. Ions at m/z 127 and 155 correspond to naphthalenyl and carbonylnaphthalenyl fragments. The ions at m/z 270 and 284 are produced from loss of the substituted butyl and pentyl side chains, respectively. These ions are characteristic for JWH-018 and the three metabolites analyzed in this study. Ions at m/z 127, 155, and 284 were then monitored for all urine samples of suspected ‘K2’ users in addition to the molecular ion for each analyte. The ion at m/z 341 was selected for N1, m/z 429 for M1 and M2, and the ion at m/z 443 for M3.

Fig. 1.

Chromatograms produced from a) native JWH-018 (N1, 5 μg/mL), b) 5-hydroxypentyl (M1), c) 4-hydroxypentyl (M2), and d) N-pentanoic acid (M3) JWH-018 metabolites (3 μg/mL).

Fig. 2.

a) Mass spectrum of 5-hydroxypentyl JWH-018 metabolite and b) the structure and resulting fragmentation for each of the analytical standards. Ions monitored during analysis are indicated by a rectangular box in the mass spectrum.

The chromatogram of a negative control urine sample is shown in Fig. 3a. Specificity of the method is demonstrated by the absence of interfering peaks at the retention times of the analytes of interest and the internal standard. A chromatogram of a urine sample (specimen 1) from a suspected ‘K2’ is presented in Fig. 3b. For clarification purposes, the chromatograms are expanded over a time range of 9–14 min. The internal standard, with a retention time of 5 min. is thus excluded. Retention times and mass spectra for this sample are consistent with those for the analytical standard (Fig. 1 and 2). Specimen 1 then contains M1, M2, and M3 but no N1. The absence of N1 has been observed in other studies and is therefore not useful as an indicator of JWH-018 consumption [6–11]. This pattern is consistent with all of the samples except specimen 3 which contains only M1 and M2. Quantification of metabolites from specimen 1 and 3 indicates that M1 is excreted in the highest concentration followed by M2 then M3 (Table 1). This ranking is consistent with other reports that measured metabolites in samples using LC-MS/MS [6]. In specimen 2, however, M2 is excreted in a higher concentration followed by M1 then M3. Differences in the metabolites are consistent with individual differences in metabolism. No other metabolites for JWH analogues were detected in any of the samples.

Fig. 3.

Chromatograms resulting from a) a negative control urine sample and b) a sample taken from an individual suspected of consuming ‘K2’ products (specimen 1).

Table 1.

Detection of JWH-018 metabolites in urine samples^a

	M1	M2	M3
Specimen 1	485	126	25
Specimen 2	81	123	10
Specimen 3	40	16	---

^aUrinary concentration in ng/mL after acid hydrolysis and SPE;

---, not detected.

4. Conclusions

In this report, a procedure was established for detection of three JWH-018 urinary metabolites using GC-MS. Acid hydrolysis followed by SPE extraction was used for preparation of samples. Using this method, three metabolites were detected in urine samples from individuals suspected of using ‘K2’ products. These analytes were confirmed using analytical standards for comparison. A procedure for synthesis of one of the standards (M2) was also described. Identification of the sites of hydroxylation as occurring at positions 4 and 5 on the pentyl chain, with detection of a carboxylic acid derivative, are in agreement with those in urine samples analyzed using LC-MS/MS by Chimalakonda et al [6]. Previous studies using GC-MS were only able to determine the location of hydroxylation as somewhere on the alkyl chain or indole ring because of no direct comparison to standards. Analysis of samples using the described GC-MS method indicates it is a suitable technique for the detection of JWH-018 metabolites in urine. When compared to LC-MS/MS, similar results are achieved for sensitivity and reproducibility with identification of the same metabolites reported. For example, Moran et al. report detection limits of ~ 2 ng/mL with an analytical precision of ~10 % using enzyme hydrolysis [7]. In contrast, a detection limit of 2.8 ng/mL and a precision of 12 % were calculated using acid hydrolysis and SPE in this report. Analysis of urine samples suspected of containing JWH-018 metabolites could then be analyzed by either GC-MS or LC-MS/MS. Because of structural similarities and thus common mass spectral fragmentation patterns, other JWH compounds and their metabolites could also be detected using this method. For example JWH-073 and JWH-019, the butyl and hexyl homologues of JWH-018, have been shown to have similar mass spectral fragments as JWH-018 [7,32]. Thus, by selecting common ions (m/z 127, 155, 284) these compounds could also be detected using the differentiating molecular ions and comparison to proper analytical standards. For the urine samples analyzed in this report, no other no other metabolites for JWH analogues were detected.