Abstract

Introduction

The endocannabinoid system is a ubiquitous cellular communication network including a family of endogenous signaling lipids (endocannabinoids) that function as agonists by engaging and activating at least one of two principal cannabinoid G protein-coupled receptors (GPCRs), designated CB1 and CB2 (De Petrocellis and Di Marzo 2009; Di Marzo and Petrosino 2007; Mackie 2008). Both CB1 and CB2 are expressed at key sites involved in pain perception and transduction (Mackie 2008; Vemuri et al. 2008). Small-molecule CB1 and non-selective CB1/CB2 agonists exert therapeutic analgesic effects in pre-clinical models of acute inflammatory, neuropathic, neoplastic, and chemothereapy-related pain (Cheng and Hitchcock 2007; Nackley et al. 2004; Pertwee 2009; Vemuri et al. 2008), which supports the conclusion that pharmacotherapeutic potentiation of cannabinergic signaling is a viable approach for pain management, a major medical need that remains unfulfilled due to the suboptimal efficacy and/or severe adverse effects associated with existing pain medicines (Stone and Molliver 2009). Unwanted psychoactivity and abuse liability mediated by central CB1 activation have hindered the clinical application of synthetic CB1-selective or non-selective CB1/CB2 agonists for pain relief (Cheng and Hitchcock 2007; Rahn and Hohmann 2009; Rahn et al. 2007). In contrast, the mainly peripheral localization of CB2 greatly reduces the risk of centrally-mediated side effects from selective CB2 agonists and offers the potential to separate the analgesic effects of cannabinergic agents from psychobehavioral liability (Anand et al. 2009; Malan et al. 2002; Whiteside et al. 2007).

The aminoalkylindole AM-1241, [(R,S)-(2-iodo-5-nitrophenyl)-[1-((1-methyl-piperidin-2-yl)methyl-1H-indol-3-yl]-methanone], (Fig.1), is a racemic CB2 ligand with low nanomolar affinity and high (up to 340-fold) selectivity for CB2 vs. CB1 (Beltramo 2009; Bingham et al. 2007; Mancini et al. 2009; Nackley et al. 2003; Quartilho et al. 2003; Yao et al. 2006). In cell systems over-expressing recombinant CB2, variable, species-dependent pharmacological functions have been attributed to AM-1241, leading some to characterize this agent as a “protean agonist” capable of changing its apparent mode of action (Beltramo 2009; Bingham et al. 2007; Mancini et al. 2009; Yao et al. 2006), whereas AM-1241 acts pharmacologically in native biological systems as a CB2 agonist (Beltramo 2009; Gutierrez et al. 2007; Hohmann et al. 2004; Ibrahim et al. 2005; Nackley et al. 2003; Nackley et al. 2004; Quartilho et al. 2003; Rahn et al. 2008). Although (R)-AM-1241 displays ~40-fold greater CB2 binding affinity than does (S)-AM-1241, the vast majority of investigations with AM-1241 in animal models have utilized the racemic compound (Beltramo 2009; Bingham et al. 2007; Rahn et al. 2008).

Figure 1

AM-1241 chemical structure

The primacy of AM-1241 as a research tool for interrogating CB2 pharmacology in health and disease shows no sign of abating (Beltramo 2009; Gutierrez et al. 2007). Across a broad range of rodent models, AM-1241 dose-dependently alleviates neuropathic pain and the allodynia/hyperalgesia induced by thermal, mechanical (e.g., nerve injury), or chemical (e.g., capsaicin, carageenan, chemotherapeutic drugs) provocation without apparent adverse side effects (Anand et al. 2009; Beltramo 2009; Bingham et al. 2007; Gutierrez et al. 2007; Rahn et al. 2010; Rahn et al. 2008; Whiteside et al. 2007). Therapeutic effects of AM-1241 have also been reported in murine models of inflammatory bowel disease (Storr et al. 2009), bone cancer (Lozano-Ondoua et al. 2010), and amyotrophic lateral sclerosis (Kim et al. 2006; Shoemaker et al. 2007). Additionally, AM-1241 has proven itself as a key research tool for defining the importance of CB1 vs. CB2 in psychobehavioral disorders, including substance abuse/drug addiction (Adamczyk et al. 2012; Gamaleddin et al. 2012a; Gamaleddin et al. 2012b).

Since AM-1241 is well established as the primary CB2 agonist, its thorough preclinical characterization would help optimize the pharmacology of this chemical class, assist in AM-1241 dose titration and interpreting dose-related activity across the diverse systems in which it continues to be tested, and inform the design of future CB2 agonists as potential medicines. These characteristics of AM-1241 have yet to be described, possibly due to the routine administration of AM-1241 in acute single-dose studies that evaluate only pathological and disease-related endpoints (Beltramo 2009). The apparent involvement of as yet undefined peripheral mechanisms in AM-1241’s analgesic action (Ibrahim et al. 2003; Rahn et al. 2010) further invites evaluation of AM-1241 biodistribution and biotransformation. This report constitutes the initial characterization of AM-1241 in vivo oral bioavailability, biodistribution and metabolism, and presents complementary in vitro studies of AM-1241 plasma stability, plasma-protein binding, predictive human oral and microsomal metabolism. A single-dose protocol was used to maintain congruency with the wealth of behavioral studies that have employed a single, acute AM-1241 administration. Although many of these studies used intraperitoneal (i.p.) or subcutaneous (s.c.) injection, our study focuses on the more clinically tractable intravenous and oral routes. The aggregate data offer the first insights into important pharmacological characteristics of AM-1241, the canonical CB2 agonist.

Materials and Methods

Results

Validation of AM-1241 HPLC quantification in biological samples

We first established conditions that allowed reliable AM-1241 recovery and quantification from plasma and various biological tissues that exhibit significant, therapeutically-relevant CB2 expression (Beltramo 2009; Cheng and Hitchcock 2007; Mackie 2008; Pertwee 2009; Rahn and Hohmann 2009; Whiteside et al. 2007). As determined from appropriately spiked samples in concentrations ranging from 5–100 µM, we achieved excellent AM-1241 recovery from all biological matrices examined: plasma, brain and spleen >90%, and from kidney, liver and intestine >70%. Representative chromatograms of AM-1241 alone, a blank mouse plasma extract, and a mouse plasma extract from an animal administered AM-1241 (Fig. 2), illustrate that the HPLC conditions we developed resolved the parent compound from all other ultraviolet-absorbing plasma constituents without interfering peaks. Comparable HPLC resolution of AM-1241 was achieved from the constituents of each tissue studied (data not shown).

Figure 2

Chromatograms illustrating the HPLC resolution of AM-1241 from endogenous constituents of mouse plasma: (a) AM-1241 in acetonitrile; (b) deproteinated blank mouse plasma; and (c) deproteinated mouse plasma sampled 10 minutes after i.v. administration …

AM-1241 biodistribution and clearance profiles

As calculated directly from the plasma area-under-the-curve (AUC) ratio between p.o. and i.v. administration, the AM-1241 oral bioavailability was ~21% (Table 1). The half-life of AM-1241 after p.o. administration was over 2-fold longer than after i.v. administration, and both the volume of distribution and rate of clearance were markedly greater after p.o. than after i.v. dosing. AM-1241 was present in plasma, brain, spleen, kidney, liver and intestine following i.v. administration; maximal concentrations within these compartments were observed within 5–15 minutes (Fig. 3). Thereafter, AM-1241 tissue concentrations decreased to become undetectable by 240 minutes in plasma, intestine, spleen and kidney; by 120 minutes in brain; and by 90 minutes in liver. Following p.o. administration, AM-1241 was detected at lower concentrations relative to plasma in spleen, kidney and intestine, with no obvious kinetic pattern, but not in brain or liver (Fig. 3). The fraction of the total AM-1241 dose in each analyzed tissue at the time of maximum observed concentration is given in Table 2. Among the organs examined, brain and kidney evidenced the greatest acute exposure to AM-1241 after i.v. administration, whereas kidney contained the greatest fraction of the total AM-1241 p.o. dose.

Figure 3

Distribution of AM-1241 among plasma and designated tissues after either i.v. or p.o. administration of 25 mg/kg AM-1241*HCl to mice. Values are the means ± standard deviation of three individual mice.

Table 1

AM-1241 biodistribution parameters in mice

Table 2

Tissue distribution of AM-1241 in mice

An in vitro metabolic-stability approach with mouse, rat and human liver microsomes was utilized to estimate the in vitro Cl`int of AM-1241(Mohutsky et al. 2006; Obach 1999). Table 3 summarizes the rate constant of substrate loss (kloss), the in vitro T1/2, the scaled Cl`int and the resulting predicted total hepatic clearance (ClH) from this analysis.

Table 3

Prediction of AM-1241 clearance

AM-1241 biotransformation

As quantified by analytical HPLC, the acute decrease in the liver and kidney contents of AM-1241 after i.v. administration (Fig. 3) was accompanied, within 15 minutes post-dosing, by the appearance of one (in kidney) or more (in liver) discrete, well-defined chromatographic peak(s) (Fig. 4, a–d). For kidney, a prominent new peak was observed at the retention time of ~13.0 minutes, whereas liver evidenced multiple new peaks in the 8–13 minute region, among which the peak at 12.4 minutes dominated. Since these peaks were not endogenous to either tissue itself, they were considered indicative of in vivo AM-1241 metabolites. The predominant AM-1241 metabolite in kidney reached highest levels at 15 and 45 minutes post-dose and was not detectable at 120 minutes, whereas the predominant liver metabolite reached a maximum at 10 minutes post-dose and was not detectable at 90 minutes post-dose (Fig. 4e).

Figure 4

HPLC chromatograms of tissue extracts suggestive of potential AM-1241 metabolites in kidney and liver 15 minutes after 25 mg/kg AM-1241*HCl i.v. administration to mice. Chromatograms of (a) kidney extract after AM-1241 administration, (b) kidney extract …

In order to gain some insight into the origin and identities of the putative in vivo AM-1241 metabolites, AM-1241 was reacted with rat and human liver microsomes, and the reaction mixtures were analyzed by LC-MS/MS. New molecular species appeared in the LC-MS/MS total-ion chromatogram that originated from the NADPH-dependent transformation of AM-1241. The product spectrum from the rat microsomal reaction mixtures evidenced a species with the m/z value indicative of the AM-1241 substrate itself (Fig. 5a) and two other species consistent with the formation of specific AM-1241-derived demethylation (Fig. 5b) and hydroxylation (Fig. 5c) products (Table 4). Thus, AM-1241 phase I microsomal metabolism occurred primarily on the piperadine moiety. Incubation of AM-1241 with human liver microsomes resulted in the appearance of the same metabolites, but the demethylated metabolite was even more prevalent than it was from the reaction of AM-1241 with rat liver microsomes (data not shown). Formation of AM-1241-derived hydroxylation and demethylation metabolites was independent of the sex of the rats from which microsomes were obtained (data not shown).

Figure 5

LC-MS spectra of: (a) AM-1241; (b) demethylated AM-1241 metabolite; (c) predominant hydroxylated AM-1241 metabolite. The spectra were obtained from LC-MS/MS analysis of organic extracts of AM-1241 incubated with rat liver microsomes for 30 minutes at …

Table 4

Rationalized fragments of AM-1241 used for metabolite identification

Discussion

Thorough preclinical characterization of GPCR orthosteric ligands displaying salutary pharmacological actions in vivo has helped establish GPCRs as the single most successful class of drug targets and GPCR ligands as important medicines (Congreve and Marshall 2010). A paramount example of a GPCR ligand with well-established preclinical therapeutic efficacy is the CB2 agonist AM-1241, which continues to be the most widely-used research tool for studying CB2 (patho)physiology, particularly the role of specific endocannabinoid-system receptors in pain transmission, antinociception and substance abuse/drug addiction (Adamczyk et al. 2012; Anand et al. 2009; Beltramo 2009; Gamaleddin et al. 2012a; Gamaleddin et al. 2012b; Gutierrez et al. 2007; Rahn et al. 2010; Rahn et al. 2008; Whiteside et al. 2007). Optimal use of AM-1241 as a pharmacological agent and template for CB2-agonist drug design necessitates appreciation of its in vivo distribution and metabolism. These properties of AM-1241 are of particular interest in light of the predominantly peripheral disposition of CB2 in vivo, and the contributions of both central and peripheral mechanisms to CB2-mediated pain relief (Cheng and Hitchcock 2007; De Petrocellis and Di Marzo 2009; Mackie 2008; Pertwee 2009; Rahn et al. 2010; Vemuri et al. 2008).

These considerations prompted us to conduct this initial characterization of AM-1241 biodistribution and biotransformation. Because rodents are routinely utilized in pre-clinical compound profiling, including study of AM-1241 pharmacological activity (Beltramo 2009; Gutierrez et al. 2007; Lozano-Ondoua et al. 2010; Rahn and Hohmann 2009; Rahn et al. 2007; Rahn et al. 2010; Rahn et al. 2008; Shoemaker et al. 2007; Storr et al. 2009), we selected the mouse as our in vivo system. We also employed a single-dose in vivo protocol, since the vast majority of AM-1241 studies in disease models have used an acute, single-dose paradigm(Beltramo 2009; Gutierrez et al. 2007; Ibrahim et al. 2003; Kim et al. 2006; Lozano-Ondoua et al. 2010; Nackley et al. 2004; Rahn et al. 2010; Shoemaker et al. 2007). Furthermore, a single-dose experimental paradigm constitutes a solid basis for designing future studies involving variations in administration protocol. We use a standardized dose of 25 mg/kg for both i.v. and p.o. AM-1241 administration. As summarized (Beltramo 2009), most studies on AM-1241 (and other CB2 agonists) in rodent pain models have administered AM-1241 i.p. as a single i.p. dose up to 50 mg/kg. Maximal AM-1241 analgesic efficacy in a standard hind paw model of thermal sensitivity was reached after local (intraplantar) administration of 3.3 mg/kg (Malan et al. 2001), although lower intraplantar and i.p. AM-1241 doses can elicit analgesia in models of carageenan- and capsaicin-induced pain (Beltramo 2009; Gutierrez et al. 2007; Hohmann et al. 2004; Nackley et al. 2003). Nonetheless, i.p. or s.c. analgesic doses of AM-1241 in the 10–30 mg/kg are commonly utilized in such models and in an incision model of post-operative pain (Bingham et al. 2007; LaBuda et al. 2005; Yao et al. 2008) and AM-1241 treatment at a daily dose of 20–40 mg/kg, i.p. was required to reduce experimental colitis in a murine model (Storr et al. 2009). Doses of AM-1241 up to 6 mg/kg i.v. and of other CB2 agonists up to 30 mg/kg p.o. have been reported efficacious in various preclinical pain models (Beltramo 2009; Beltramo et al. 2006; Ohta et al. 2008). These data support the conclusion that the 25 mg/kg AM-1241 dose we used is a pharmacological one. Indeed, the tissue distribution of GW405833, a CB2-selective aminoalkylindole agonist that exhibits maximal effective antihyperalgesic and analgesic activities at 1–3 mg/kg, i.p. (Beltramo 2009), has been characterized at 100 mg/kg, i.p. (Valenzano et al. 2005).

Whether administered orally or systemically (i.v.) at 25 mg/kg, AM-1241 was not toxic, for post-experiment necropsies were unremarkable. Following AM-1241 i.v., but not p.o., administration, transient hypomotility was noted. Although hypomotility is a behavioral response associated with cannabinoid agonists, their motor effects in rodents are variable and may exhibit a triphasic, dose-dependent pattern: lower doses induce transient hypoactivity similar to what we observed, whereas higher doses stimulate movement until the hallmark signs of muscular rigidity and fixed posture of catalepsy emerge (McLaughlin et al. 2005; Valenzano et al. 2005). Cannabinoid agonist-induced motor effects have been mainly attributed to central CB1 stimulation (McLaughlin et al. 2005; Monory et al. 2007). An acute “spillover” effect of AM-1241 on CB1 in the CNS might thus account for the transient hypomotility we observe after i.v. dosing, a conjecture congruent with our detection of AM-1241 in brain after i.v. (but not p.o.) AM-1241 administration. However, the appreciable (up to 340-fold) selectivity of AM-1241 for mouse CB2 vs. CB1 (Beltramo 2009; Hohmann et al. 2004; Nackley et al. 2003; Shoemaker et al. 2007) and the ability of two non-selective cannabinoid-receptor agonists, the aminoaklylindoles AM-2233 and WIN 55,212-2, to induce profound catalepsy at only 1.0 mg/kg− i.v. (Dhawan et al. 2006; Petitet et al. 1999) suggest that factors other than an off-target AM-1241 response are involved. Since CB1 stimulation by the endocannabinoid 2-arachidonoylglycerol (2-AG) is primarily responsible for hypomotility in mice, and 2-AG is a full agonist at both CB1 and CB2 (Long et al. 2009; McLaughlin et al. 2005; Monory et al. 2007; Morgan et al. 2009), AM-1241 associated hypomotility could reflect a shift in the intrinsic poise of 2-AG signaling at CNS synapses containing CB1 and CB2 consequent to AM-1241 CB2 engagement that favors CB1 stimulation by 2-AG. Two out of 29 mice receiving 25 mg/kg AM-1241 p.o. exhibited gaping, a surrogate response for nausea/emesis in rodents incapable of vomiting (Parker and Limebeer 2006). Since rodent gaping is commonly observed after ingestion of an aversive chemical (e.g., lithium) and can be induced by CB1 antagonists/inverse agonists, but not by CB1 neutral antagonists or CB2 agonists (Parker and Limebeer 2006; Salamone et al. 2007), the gaping we observed in only a few mice after AM-1241 administration p.o. likely reflects a negative taste reaction to the AM-1241 dosing solution.

We developed and validated extraction and analytical HPLC methods for accurate, reliable AM-1241 quantification in murine plasma and in tissues where CB2 has been associated with important (patho)physiological roles (Beltramo 2009; Cheng and Hitchcock 2007; Mackie 2008; Pertwee 2009; Rahn and Hohmann 2009; Whiteside et al. 2007). This direct approach is in contrast to reliance upon time-course studies using in vivo pharmacological effects as proxies for tissue distribution/bioavailability (Sink et al. 2009). Quantification of AM-1241 in plasma after p.o. and i.v. administration allowed determination that its oral bioavailability is ~21%. This value is consistent with the AM-1241 Papp value of 6.2 × 10−6 cm/s we measured in a Caco-2 cell system, since Papp values in the range of 1–10 × 10−6 cm/sec are indicative of moderate (i.e., 20–70%) in vivo oral bioavailability (Yee 1997). Among the very few CB2 agonists whose oral bioavailability has been determined experimentally in rodents, values of 0% for GW405833 (Valenzano et al. 2005), and 13–100% for a series of N-alkyldienearylcarboxamides (Ohta et al. 2007) have been reported at doses of 0.1–100 mg/kg.

The rapid AM-1241 hepatic clearance predicted from in vitro metabolic stability data with mouse, rat and human microsomes was substantiated by direct in vivo half-life profiling in the mouse. The half-life of AM-1241 after p.o. administration (82 minutes) is less than the 3–5 hour half-lives reported after p.o. administration of some N-arylamide-oxadiazole CB2 agonists (Cheng et al. 2008) and the CB2 agonist GW842166X (Giblin et al. 2007). AM-1241’s moderate oral bioavailability and short in vivo half-life would favor its i.p. or i.v. administration or injection at an injury/inflammation site, as well as its use in acute nociceptive tests, congruent with many published applications(Beltramo 2009; Gutierrez et al. 2007; Ibrahim et al. 2005; Malan et al. 2002; Nackley et al. 2003; Quartilho et al. 2003; Rahn et al. 2010; Whiteside et al. 2007). The half-life of AM-1241 may also provide a rationale for employing multiple i.p. administrations over several days in order to attain therapeutic efficacy against more chronic injury states, such as experimental colitis (Storr et al. 2009), as well as the need for sustained AM-1241 i.p. administration to reduce spontaneous and evoked pain and bone loss or fracture in a murine bone cancer model (Lozano-Ondoua et al. 2010). Given its demonstrated tissue-protective and nociceptive effects (Beltramo 2009; Gutierrez et al. 2007; Lozano-Ondoua et al. 2010; Ohta et al. 2008; Rahn et al. 2008; Shoemaker et al. 2007; Yao et al. 2008), AM-1241 may serve as a valuable template for judicious structural modification aimed at enhancing its oral bioavailability and/or extending its half-life. This proposition is especially attractive in light of the excellent plasma exposure after p.o. and (especially) i.v. AM-1241 administration (as indexed by the AUC), which is comparable to that reported for some lead N-arylamide-oxadiazole and N-alkyldienearylcarboxamide CB2 agonists (Cheng et al. 2008; Ohta et al. 2007).

AM-1241 is susceptible to extensive phase I cytochrome P450 metabolism to generate demethylated and hydroxylated products. Based on the fragmentation patterns we observed by LC-MS/MS when AM-1241 is incubated with rat or human microsomal systems, AM-1241 was predominantly metabolized at its piperadine moiety, an observation useful in assessing potential chemical modifications that might alter AM-1241’s metabolism. Although the P450 family plays a critical role in xenobiotic detoxification in mice, rats, and humans, it remains to be established whether there are significant, species-dependent differences in AM-1241 microsomal metabolism (Martignoni et al. 2006). The CB2 antagonist/inverse agonist AM-630 and the CB1/CB2 agonist WIN-55212-2 are metabolized at their morpholino moiety, which corresponds to AM-1241’s piperadine group (Zhang et al. 2002; Zhang et al. 2004), indicating that the P450 mediated biotransformation of all three aminoalkylindoles may be similar. The synthetic cannabinoid JWH-018 is likewise a substrate for phase I, P450 hydroxylation by human microsomes in vitro (Wintermeyer et al. 2010) and in vivo (Sobolevsky et al. 2010), as is the CB2 aminoalkylindole agonist JWH-015 in vitro (Zhang et al. 2006). These aggregate data suggest that P450 mediated hydroxylation may be inter-species metabolic route by which the lipophilicity of ligands that modify cannabinergic lipid signaling is reduced to facilitate compound elimination. The metabolites of AM-1241 may themselves prove to be very important pharmacologically: preliminary experiments indicate that the proposed demethylated metabolite of AM-1241 shares a similar selectivity for CB2 vs. CB1, but is somewhat less potent a CB2 agonist, as compared to the parent compound (data not shown).

Conclusion

The data presented constitute the initial characterization of the biodistribution and metabolism of AM-1241, the prototypic agonist used to gain insight into the (patho)physiological role of CB2-mediated signaling, as obtained from in vitro biochemical and cell-based systems and in vivo in mice. Our data show that AM-1241 has the following characteristics: a) short plasma half-life; b) limited oral bioavailability; c) extensive binding to plasma/albumin; d) subject to hepatic hydroxylation and demethylation; and e) moderate hepatic clearance. These results should aid in the design of detailed studies involving variations in AM-1241 dose and administration protocols, as well as the interpretation of already published efficacy data. The present work should also help inform the synthesis and optimization of future-generation CB2 agonists with therapeutic utility.

Acknowledgments

Footnotes

References