The 2-alkyl-2H-indazole regioisomers of synthetic cannabinoids AB-CHMINACA, AB-FUBINACA, AB-PINACA, and 5F-AB-PINACA are possible manufacturing impurities with cannabimimetic activities

Indazole-derived synthetic cannabinoids (SCs) featuring an alkyl substituent at the 1-position and l-valinamide at the 3-carboxamide position (e.g., AB-CHMINACA) have been identified by forensic chemists around the world, and are associated with serious adverse health effects. Regioisomerism is possible for indazole SCs, with the 2-alkyl-2H-indazole regioisomer of AB-CHMINACA recently identified in SC products in Japan. It is unknown whether this regiosiomer represents a manufacturing impurity arising as a synthetic byproduct, or was intentionally synthesized as a cannabimimetic agent. This study reports the synthesis, analytical characterization, and pharmacological evaluation of commonly encountered indazole SCs AB-CHMINACA, AB-FUBINACA, AB-PINACA, 5F-AB-PINACA and their corresponding 2-alkyl-2H-indazole regioisomers. Both regioisomers of each SC were prepared from a common precursor, and the physical properties, 1H and 13C nuclear magnetic resonance spectroscopy, gas chromatography–mass spectrometry, and ultraviolet–visible spectroscopy of all SC compounds are described. Additionally, AB-CHMINACA, AB-FUBINACA, AB-PINACA, and 5F-AB-PINACA were found to act as high potency agonists at CB1 (EC50 = 2.1–11.6 nM) and CB2 (EC50 = 5.6–21.1 nM) receptors in fluorometric assays, while the corresponding 2-alkyl-2H-indazole regioisomers demonstrated low potency (micromolar) agonist activities at both receptors. Taken together, these data suggest that 2-alkyl-2H-indazole regioisomers of AB-CHMINACA, AB-FUBINACA, AB-PINACA, and 5F-AB-PINACA are likely to be encountered by forensic chemists and toxicologists as the result of improper purification during the clandestine synthesis of 1-alkyl-1H-indazole regioisomers, and can be distinguished by differences in gas chromatography–mass spectrometry fragmentation pattern. Electronic supplementary material The online version of this article (doi:10.1007/s11419-016-0316-y) contains supplementary material, which is available to authorized users.


Introduction
Synthetic cannabinoids (SCs) are the most rapidly growing class of novel psychoactive substances (NPSs) [1]. SC ''designer drugs'' are intended to mimic the psychoactive effects of D 9 -tetrahydrocannabinol (D 9 -THC, 1, Fig. 1), the principal bioactive component of cannabis. Unlike D 9 -THC-a partial agonist at both cannabinoid type 1 (CB 1 ) M. Longworth and S. D. Banister contributed equally to this work.
Electronic supplementary material The online version of this article (doi:10.1007/s11419-016-0316-y) contains supplementary material, which is available to authorized users. and type 2 (CB 2 ) receptors-most SCs possess high efficacy agonist activities at both CB receptor subtypes. Since the discovery of CB 1 /CB 2 agonist JWH-018 (2) in consumer products in Germany, Austria, and Japan in 2008 [2,3], more than 130 SCs have been reported in Europe, with 30 identified in 2014 alone [4].
Recently, AB-CHMINACA, AB-FUBINACA, AB-PINACA, and 5F-AB-PINACA were shown to act as potent and efficacious agonists at human CB 1 and CB 2 receptors in vitro [31,32]. Moreover, AB-CHMINACA, AB-FUBINACA, and AB-PINACA exert potent cannabimimetic effects on locomotion, body temperature, heart rate, and nociception in mice and rats, as well as substituting for D 9 -THC in drug discrimination assays [31][32][33]. Taken together, these data indicate that a range of alkyl substituents are tolerated at the 1-indazole position in this class of SCs.
These SCs appear more toxic than earlier examples, and multiple overdoses and fatalities in the USA have been attributed to AB-CHMINACA, AB-FUBINACA, and AB-PINACA [34][35][36]. Due to the abuse potential and toxicity of these newer SCs, the Federal Government of the USA has used emergency scheduling laws to temporarily place AB-FUBINACA, AB-PINACA, and AB-CHMINACA into Schedule I, the most restrictive category of the Controlled Substances Act. Similar legislation has been enacted in Germany, Singapore, and elsewhere [37,38].
Isomerism may be used to generate new SC analogues [48][49][50], and the 2H-indazole regioisomer of AB-CHMI-NACA (AB-CHMINACA 2-isomer, 7) was recently detected in SC products in Japan [51]. It is currently unclear if this analogue is an impurity occurring as an unintended byproduct of AB-CHMINACA synthesis, or if the 2-isomer was willfully prepared as an intended cannabimimetic agent. Incidentally, the 2H-indazole analogue of APINACA was reported in China recently [52]. One study of bulk powders used as the active ingredients for SC products found purity to range from 78 to 96 %, consistent with the potential presence of manufacturing impurities in the raw materials [53].
We recently reported the synthesis, structural characterization, in vitro cannabinoid activity, and in vivo biotelemetry of several indazole synthetic cannabinoid designer drugs [31]. During the preparation of 5F-AB-PINACA, it was noted that alkylation of methyl indazole-3-carboxylate predominantly produced the desired 1-alkylated intermediate, with the 2-alkyl regiosiomer formed as a minor product. Failure to remove the unwanted 2-alkyl intermediate before the remaining synthetic steps would be expected to form 5F-AB-PINACA 2-isomer as an impurity. However, the cannabinoid activity of the 2H-indazole regioisomers of this class of SCs has never been reported, and AB-CHMINACA 2-isomer was detected in consumer products at a concentration similar to those reported for other SC products.

General chemical synthesis details
The synthesis of 3-10 is shown in Fig. 2. All reactions were performed under an atmosphere of nitrogen or argon unless otherwise specified. Anhydrous tetrahydrofuran (THF), methanol, acetonitrile, and dimethyl sulfoxide (DMSO) (Sigma Aldrich, St. Louis, MO, USA) were used as purchased. Commercially available chemicals (Sigma-Aldrich) were used as purchased. Analytical thin-layer chromatography was performed using Merck aluminumbacked silica gel 60 F254 (0.2 mm) plates (Merck, Darmstadt, Germany), which were visualized using shortwave (254 nm) UV fluorescence. Flash chromatography was performed using Merck Kieselgel 60 (230-400 mesh) silica gel. Melting point (m.p.) ranges were measured in open capillaries using a Stuart SMP10 m.p. apparatus (Bibby Scientific, Staffordshire, UK) and were uncorrected. NMR spectra were recorded at 300 K using either a Bruker AVANCE DRX400 (400.1 MHz) or AVANCE III 500 Ascend (500.1 MHz) spectrometer (Bruker, Bremen, Germany). The data are reported as chemical shift (d ppm) relative to the residual protonated solvent resonance, relative integral, multiplicity (s, singlet; br s, broad singlet; d, doublet; t, triplet; quin., quintet; sept., septet; dt, doublet of triplets; m, multiplet), coupling constants (J Hz) and assignment (for 13 C, quat., quaternary). Assignment of signals was assisted by correlation spectroscopy (COSY), distortionless enhancement by polarization transfer (DEPT), heteronuclear single quantum coherence (HSQC), and heteronuclear multiple-bond correlation (HMBC) experiments where necessary. Low-resolution mass spectra (LRMS) were recorded using electrospray ionization (ESI) recorded on a Finnigan LCQ ion trap mass spectrometer (ThermoFisher Scientific, Waltham, MA, USA). High-resolution mass spectra (HRMS) were run on a Bruker 7T Apex Qe Fourier Transform Ion Cyclotron resonance mass spectrometer equipped with an Apollo II ESI/APCI/MALDI Dual source by the Mass Spectrometry Facility of the School of Chemistry at The University of Sydney. IR absorption spectra were recorded on a Bruker ALPHA FT-IR spectrometer as solid or thin film from ethanol, and the data are reported as vibrational frequencies (cm -1 ). Please see the supplementary material for 1 H and 13 C NMR spectra and Fourier-transform infrared (FTIR) spectra of all final compounds.

General procedure D: preparation of indazole-3carboxamides
A cooled (0°C) solution of the appropriate 1-alkyl-1H-or 2-alkyl-2H-indazole-3-carboxylic acid (0.19 mmol, 1.0 equiv.), L-valinamide (31 mg, 0.20 mmol, 1.05 equiv.), and (benzotriazol-1-yl-oxy)tripyrrolidinophosphonium hexafluorophosphate (PyBOP Ò ) (106 mg, 0.20 mmol, 1.05 equiv.) in DMSO (3 mL) was treated dropwise with N,Ndiisopropylethylamine (DIPEA, 68 lL, 0.39 mmol, 2.0 equiv.) and stirred at ambient temperature for 2 h. The reaction was poured onto sat. aq. sodium hydrogen carbonate and extracted with diethyl ether (3 9 50 mL). The combined organic layers were dried (MgSO 4 ) and the solvent evaporated under reduced pressure. The crude products were purified using flash chromatography.     In vitro pharmacological assessment of 3-10 Mouse AtT-20 neuroblastoma cells stably transfected with human CB 1 or human CB 2 have been previously described [54] and were cultured in Dulbecco's modified Eagle's medium (DMEM) containing 10 % fetal bovine serum (FBS), 100 U penicillin/streptomycin, and 300 lg/mL G418. Cells were passaged at 80 % confluency as required. Cells for assays were grown in 75 cm 2 flasks and used at 90 % confluence. The day before, the assay cells were detached from the flask with trypsin/EDTA (Sigma-Aldrich) and resuspended in 10 mL of Leibovitz's L-15 media supplemented with 1 % FBS, 100 U penicillin/ streptomycin and 15 mM glucose (membrane potential assay and Ca5 calcium assay). The cells were plated in volume of 90 lL in black-walled, clear-bottomed 96-well microplates (Corning, Oneonta, NY, USA) which had been precoated with poly-L-lysine (Sigma-Aldrich). Cells were incubated overnight at 37°C in ambient CO 2 . Membrane potential was measured using a FLIPR membrane potential assay kit (blue) from Molecular Devices (Sunnyvale, CA, USA), as described previously [55]. The dye was reconstituted with assay buffer of composition (mM): NaCl 145, HEPES 22, Na 2 HPO 4 0.338, NaHCO 3 4.17, KH 2 PO 4 0.441, MgSO 4 0.407, MgCl 2 0.493, CaCl 2 1.26, glucose 5.56, and bovine serum albumin (0.1 mg/mL, pH 7.4, osmolarity 315 ± 5). Prior to the assay, cells were loaded with 90 lL/well of the dye solution without removal of the L-15, giving an initial assay volume of 180 lL/well. Plates were then incubated at 37°C at ambient CO 2 for 60 min. Fluorescence was measured using a FlexStation 3 (Molecular Devices) microplate reader with cells excited at a wavelength of 530 nm and emission measured at 565 nm. Baseline readings were taken every 2 s for at least 2 min, at which time either drug or vehicle was added in a volume of 20 lL. The background fluorescence of cells without dye or dye without cells was negligible. Changes in fluorescence were expressed as a percentage of baseline fluorescence after subtraction of the changes produced by vehicle Forensic Toxicol (2016) 34:286-303 295 addition, which was less than 2 % for drugs dissolved in assay buffer or DMSO. The final concentration of DMSO was always 0.1 %. Data were analyzed with PRISM (GraphPad Software Inc., San Diego, CA), using four-parameter nonlinear regression to fit concentration-response curves. In all plates, a maximally effective concentration (1 lM) of CP 55,940 (Cayman Chemical, Ann Arbor, MI, USA) was added to allow for normalization between assays.
The alkylation of methyl 1H-indazole-3-carboxylate (12) proceeded regioselectively to favour the formation of 1-substituted 1H-indazole intermediates as major products, entirely consistent with the known reactivity of indazoles towards nucleophilic substitution [56][57][58]. To quantify the regioselectivity of each alkylation method, the isolated yields of the regioisomeric products obtained following alkylation of 12 under each set of conditions for two representative bromoalkanes were determined are shown in Table 1. All reactions were performed in parallel, on the same scale, using identical reagents. Alkylation of 12 with 4-fluorobenzyl bromide or 1-bromopentane using potassium tert-butoxide as the base produced 14 and 15, respectively, as the major products with excellent regioselectivity for the 1-position (49:1 and 12.3:1, respectively). The corresponding 2-alkylated regiosiomers (18 and 19, respectively) were obtained as minor products in \10 % yield. While use of potassium carbonate as a base also furnished 14 and 15 as the major products, regioselectivity was reduced and appreciable quantities of 18 and 19 were obtained. The latter set of conditions (general procedure B) required flash chromatography to separate regiosiomeric intermediates of similar polarity and retention time; so it is unlikely that this method is used on an industrial scale to produce this class of SCs. In contrast, general procedure A proceeded with high regioselectivity, and it is feasible that incomplete purification of the crude product from this reaction allows 2-alkyl-2H-indazole intermediate to be carried through the synthesis, resulting in the production of regioisomeric SCs.
All 1-alkyl regioisomers failed to show molecular ion peaks, but similar fragmentation patterns were evident in each case. For AB-CHMINACA, AB-FUBINACA, AB-PINACA, and 5F-AB-PINACA, the most prominent fragment ions could be attributed to cleavage of the terminal carboxamide subunit of the pendant L-valinamide group (m/z 312, 324, 286, and 304 respectively), as well as C-N scission of the indazole-3-carboxamide bond (m/z 241, 253, 215, and 233). 5F-AB-PINACA also showed a peak at m/z 213, consistent with a protonated species arising from defluorination of fragment m/z 233.
In the cases of AB-CHMINACA, AB-PINACA, and 5F-AB-PINACA, a common fragment of 145 mass units was observed, likely due to the protonated species arising from indazole-3-carboxamide bond scission and loss of the  1-alkyl group. AB-FUBINACA was devoid of peak at m/z 145, but did feature the 4-fluorobenzyl fragmentation ion (m/z 109). The fragmentation patterns for the corresponding 2-alkyl-2H-indazoles 7-10 were distinct from their 1-alkyl regioisomers. Only in the case of AB-FUBINACA 2-isomer, the molecular ion (m/z 368) and fragment corresponding to cleavage of the terminal carboxamide group (m/z 324) were observed. Again, common fragment m/z 145 arising from dealkylated indazole-3-acylium ion could be seen in the spectra of 7, 9, and 10. A fragment corresponding to scission of the entire 3-substituent of indazoles 7, 9, and 10 (m/z 213, 187, and 205 respectively) was present as the base peak in these spectra. Fragmentation of the neighboring amide bond proximal to the indazole core was seen for 8, 9, and 10 (m/z 253, 215, and 233 respectively), but not for 7.
The UV-Vis absorption spectra of 3-10 are shown in Fig. 5, and peak absorbances are similar for each 1-alkyl-1H-indazoles and their regioisomers, indicating that ultraviolet spectroscopy is not a suitable method for differentiation of regiosiomeric indazole cannabinoids of this class.
The cannabinoid activities of indazole SCs 3-10 at CB 1 and CB 2 receptors are shown in Table 2. Murine AtT-20 neuroblastoma cells were stably transfected with human CB 1 or CB 2 receptors, and activities of CP 55,940 and 3-10 were evaluated using FLIPR membrane potential assays whereby endogenously expressed G protein-gated inwardly rectifying K ? channels (GIRKs) are activated by agonists at the coexpressed CB 1 or CB 2 receptors [55,59]. The maximum effects of 3-10 were compared to the high efficacy CB 1 /CB 2 receptor agonist CP 55,940, which produced a maximal decrease in fluorescence, corresponding to cellular hyperpolarization, of 14 ± 2 % in the AtT-20-CB 1 cells and 33 ± 2 % in the AtT-20-CB 2 cells.
In contrast, the 2-regioisomers 7-10 were not potent at stimulating CB 1 and CB 2 receptor coupling to GIRKs. EC 50 values could not be determined for 7 or 8 at either receptor, and 9 and 10 showed only micromolar potency for CB 1 (EC 50 = 4080 and 1930 nM, respectively) and CB 2 (EC 50 = 4120 and 2980 nM, respectively).
Of the 2-alkyl-2H-indazoles, 7, 9, and 10, displayed a maximum effect at CB 1 receptors comparable to 1 lM CP 55,940 at the highest concentration tested, 30 lM. The dramatic difference in potency and efficacy of the 1-alkyl-1H-indazoles 3-6 and 2-alkyl-2H-indazoles 7-10 at CB 1 and CB 2 receptors is shown in Fig. 6a, b respectively. The non-linear regression fit of CP 55,940 is included as a dashed line for comparison. The 1-alkyl compounds 3-6 (white data points) were all more potent and efficacious than CP 55,940 at CB 1 receptors, while 2-alkyl regioisomers 7-10 (black data points) were substantially less potent, with three reaching efficacy comparable to maximum CP 55,940 only at the highest concentration tested (Fig. 6a). At CB 2 receptors, 3-6 demonstrate potency and efficacy comparable to CP 55,940, but 7-10, with the exception of 9, show efficacy approximately 20-40 % that of a maximally efficacious concentration of CP 55,940 (Fig. 6b).

Conclusions
This study represents the first pharmacological characterization of the 2-alkyl-2H-indazole regioisomers of SC designer drugs AB-CHMINACA, AB-FUBINACA, AB-PINACA, and 5F-AB-PINACA. A general synthetic route to both 1-alkyl-1H-and 2-alkyl-2H-indazole-3-carboxamides was demonstrated, and may be of use to cannabinoid researchers for the preparation of related analogues. Additionally, characteristic differences in the fragmentation patterns of these regioisomers were described, enabling their differentiation by GC-MS. AB-CHMINACA, AB-FUBINACA, AB-PINACA, 5F-AB-PINACA, and the corresponding 2-alkyl-2H-indazole regioisomers 7-10 were evaluated for their activity at human CB 1 and CB 2 receptors in vitro using FLIPR membrane potential assays. AB-CHMINACA, AB-FUBINACA, AB-PINACA, and 5F-AB-PINACA are potent, efficacious agonists of CB 1 and CB 2 receptors, while the corresponding 2-alkyl-2H-indazole regioisomers possessed only low potency as CB 1 /CB 2 agonists. Having demonstrated the weak cannabimimetic properties of the 2-alky-2H-indazole regioisomers of AB-CHMINACA, AB-FUBINACA, AB-PINACA, and 5F-AB-PINACA, and the synthesis of both regioisomers from a common precursor, it is possible that 2-alky-2Hindazole-3-carboxamide SCs occur in recreational products as synthesis byproducts of 1-alky-1H-indazole-3-carboxamides rather than analogues produced deliberately as SCs in their own right. Fig. 6 Hyperpolarization of a CB 1 and b CB 2 receptors induced by 3-10 as a proportion of that produced by 1 lM CP 55,940. Membrane potential was measured using a fluorescent dye, as outlined in the text. Each point represents the mean ± standard error of at least five independent determinations, each performed in duplicate. Data was fitted with a four parameter logistic equation in GraphPad Prism Forensic Toxicol (2016) 34:286-303 301 Ethical approval This article does not contain any studies with human participants or animals performed by any of the authors.