- Top of page
- Conflict of interest
Salvinorin A is the main pharmacologically active ingredient of Salvia divinorum, a hallucinogenic plant that has been used for centuries for divination and shamanism by curanderos in Mexico and other areas (Roth et al., 2004). S. divinorum leaf preparations (occasionally fortified with extracted salvinorin A) are widely available in western Europe and the USA, notably on internet sites. The lipid-like salvinorin A molecule is chemically and structurally unique in that it represents the only known psychoactive diterpenoid, and was the first non-nitrogenous hallucinogen to be identified (Vortherms and Roth, 2006). Additionally, salvinorin A was reported to be the most potent naturally occurring hallucinogen, with an effective dose, when smoked, of 200–1000 μg in humans (Siebert, 1994; Sheffler and Roth, 2003).
Salvinorin A was found to be a strong and selective agonist of the κ-opioid receptor (KOR) (Roth et al., 2002; O'Connor and Roth, 2005). KOR-mediated salvinorin A effects in vivo include antinociceptive (McCurdy et al., 2006), antipruritic (Wang et al., 2005) and motor effects (Fantegrossi et al., 2005), inhibition of striatal dopamine levels in mice (Zhang et al., 2005) and induction of discriminative stimuli in rhesus monkeys (Butelman et al., 2004). We have recently shown that salvinorin A inhibited enteric cholinergic transmission in the isolated guinea-pig ileum (Capasso et al., 2006) and reduced motility in the croton oil model of intestinal inflammation in mice through activation of KORs (Capasso et al., 2008b). Moreover, croton oil-induced gut inflammation increased the potency of salvinorin A, but not of the reference KOR agonist U-50488, suggesting that salvinorin A may have target(s) other than KORs in the inflamed gut (Capasso et al., 2008b).
Because upregulation of cannabinoid CB1 receptors and increased expression of fatty acid amide hydrolase (FAAH, one of the enzymes responsible for endocannabinoid inactivation) were observed in the intestine of mice treated with croton oil (Izzo et al., 2001), and as salvinorin A exhibits rewarding effects that are attenuated by a cannabinoid CB1 receptor antagonist in zebrafish and rats (Braida et al., 2007, 2008), here we investigated the possible involvement of the endocannabinoid system, specifically, cannabinoid receptors, FAAH, monoacylglycerol (MAG) lipase (which is involved in the inactivation of the endocannabinoid 2-arachidonoylglycerol) and endocannabinoid membrane transporter (EMT), in salvinorin A-induced delay in motility during intestinal inflammation.
- Top of page
- Conflict of interest
Salvinorin A, a highly potent and selective KOR agonist, is the most potent naturally occurring hallucinogen known (Sheffler and Roth, 2003). Previous findings have shown that this diterpenoid may have target(s) other than KOR in the inflamed gut (Capasso et al., 2008b). Here, we have shown that salvinorin A reduced motility in a model of small bowel disease by a mechanism indirectly involving cannabinoid CB1 receptors.
Cannabinoid and opioid systems share neuroanatomical, neurochemical and pharmacological features (Corchero et al., 2004): both cannabinoids and opioids induce analgesia, catalepsy, hypothermia, motor depression, hypotension, immunosuppression, sedation, rewarding effects (Vigano et al., 2005) and, importantly, inhibition of gastrointestinal motility (Duncan et al., 2005). Cannabinoid CB1 and opioid receptors (μ-opioid receptor, δ-opioid receptor and KOR) are expressed presynaptically/prejunctionally and their activation elicits inhibition of excitatory transmission, which is associated with a decrease in acetylcholine release and a reduction in peristalsis (Coutts and Izzo, 2004; Wood and Galligan, 2004; Duncan et al., 2005; Izzo and Camilleri, 2008). Moreover, cannabinoid CB1 receptor colocalization has been demonstrated with KOR in porcine cultured myenteric neurons (Kulkarni-Narla and Brown, 2001). Evidence has been provided demonstrating a functional cross-talk between the cannabinoid and opioid systems in the central nervous system (Corchero et al., 2004; Fattore et al., 2004), although current evidence does not suggest a cross-talk between opioid and cannabinoid receptors in the digestive tract (Massa and Monory, 2006; Sanger, 2007). For example, Carai et al. (2006) showed that the cannabinoid CB1 receptor antagonist rimonabant failed to block the inhibitory effect of morphine and loperamide on transit in small intestine. However, it should be emphasized that morphine and loperamide inhibit intestinal motility through activation of μ-opioid receptor and δ-opioid receptor, but not KOR (Hurwitz et al., 1994). Thus, there have been no functional studies that have examined specifically KOR and cannabinoid CB1 interactions in the gastrointestinal tract to date, although an interaction between both KOR and the cannabinoid system on self-administration effects has been previously reported (Mendizábal et al., 2006).
In this study, we have shown that salvinorin A inhibited motility in the croton oil model of ileitis. The effect of salvinorin A was counteracted by both the selective KOR antagonist nor-binaltorphimine and the selective CB1 receptor antagonist rimonabant. U-50488, a synthetic KOR agonist, also inhibited motility in a KOR- and a CB1-antagonist sensitive manner. Collectively, such findings suggest a cross-talk between KORs and cannabinoid CB1 receptors in the regulation of intestinal motility under inflammatory conditions. Because rimonabant is known to exert pro-kinetic effects both in control and in croton oil-treated mice (Izzo et al., 2001), here, in the agonist/antagonist studies, we used a dose of rimonabant (0.1 mg kg−1) that, per se, did not affect intestinal motility. In addition, we have recently shown that this dose of rimonabant did not modify the inhibitory effect of loperamide (μ-opioid receptor and δ-opioid receptor agonist) on intestinal motility in croton oil-treated animals (Capasso et al., 2008a). Collectively, these results suggest that the interaction between cannabinoid CB1 receptors and KORs is specific and does not involve other opioid receptor subtypes. Interestingly, we have shown that the inhibitory effect on motility of the selective CB1 receptor agonist ACEA in croton oil-treated mice was not modified by the selective KOR antagonist nor-binaltorphimine, suggesting an ‘unidirectional’ cross-talk. This is possibly because KORs are not overexpressed in myenteric nerves of mice treated with croton oil (Pol and Puig, 2004), whereas CB1 receptors are overexpressed (Izzo et al., 2001). Indeed, the cannabinoid CB1/KOR cross-talk is restricted to pathophysiological states only, as the inhibitory effect of salvinorin A (as well as the inhibitory effect of U-50488) on motility was not modified by rimonabant under physiological conditions, that is, in non-croton oil-treated mice. Again, the increased expression of CB1 receptors in the small intestine of mice treated with oral croton oil (Izzo et al., 2001) could explain why the cannabinoid CB1/KOR interaction is observed in intestinal pathophysiological states only. The study of the mode of action of salvinorin A in reducing motility in control mice represents a separate aim of investigation, which falls outside the scope of this work. Salvinorin A has been shown to reduce motility through a non-CB1 non-KOR-mediated mechanism in control mice (Capasso et al., 2008b). Also, an involvement of CB2 receptors seems very unlikely, as these receptors are involved in the control of intestinal motility in pathophysiological states only (Izzo and Camilleri, 2008; Wright et al., 2008).
We have recently shown that gut inflammation increases the potency of salvinorin A; the increased potency was mediated by KOR, but it was not shared by the prototypical KOR agonist U-50488, suggesting that salvinorin A may have target(s) other than KORs in the inflamed gut (Capasso et al., 2008b). Because the endocannabinoid system undergoes adaptive changes in the inflamed gut (that is, increased expression of CB1 and CB2 receptors, increased levels of endocannabinoids) (Di Marzo and Izzo, 2006; Izzo, 2007; Wright et al., 2008) and because the rewarding effects of salvinorin A in zebrafish and rats involve cannabinoid CB1 receptors (Braida et al., 2007, 2008), we hypothesized one such target might be the endocannabinoid system. Thus, and in view of the results shown above, next we investigated the possibility that salvinorin A activates CB1 or CB2 receptors (which are known to reduce intestinal motility under inflamed conditions) (Di Marzo and Izzo, 2006; Izzo and Camilleri, 2008) either directly or by increasing the levels of endocannabinoids through inhibition of the mechanisms involved in their inactivation (that is, the putative EMT, FAAH and MAG lipase). Binding and enzymatic experiments revealed that salvinorin A had no effect on anandamide inactivation, nor any strong affinity for CB1 receptors, although it did exhibit a very weak affinity for CB2 receptors, which have been shown to be involved in the control of intestinal motility in the inflamed gut (Mathison et al., 2004; Duncan et al., 2008), including the croton oil model of intestinal inflammation (Capasso et al., 2008a). However, it is very unlikely that the weak affinity for CB2 receptors could contribute to the inhibitory effect of salvinorin A on motility because its effect on transit was not modified by the selective CB2 antagonist SR144528. The dose of SR144528 used in this study has been previously shown to counteract the inhibitory effect of the CB2 receptor agonist JWH 015 on intestinal transit in croton oil-treated mice (Capasso et al., 2008a). These findings strongly suggest that (1) the observation that the effect of salvinorin A in croton oil-treated mice is also antagonized by rimonabant is not due to its direct or indirect activation of CB1 receptors, as indicated also by the fact that rimonabant also antagonized the effect of U-50488, and (2) the higher potency of salvinorin A with respect to U-50488 under these inflammatory conditions is also not due to its additional activation of CB1 receptors.
Although the present experiments allow us to exclude the possibility that salvinorin A binds to cannabinoid receptors or influences endocannabinoid transport and enzymatic degradation, the explanation of why intestinal inflammation increased the potency of salvinorin A but not that of the synthetic KOR agonist U-50488 previously reported (Capasso et al., 2008b) still remains to be investigated. However, it should be considered that salvinorin A was significantly more efficacious than the KOR agonist U-50488 (Chavkin et al., 2004) and binds to and activates KOR through hydrophobic—and not ionic—interactions (Grundmann et al., 2007). Moreover, salvinorin A is unusual as a KOR agonist as it is 40-fold less potent in promoting internalization of the human KOR and causes less downregulation of surface receptors than U-50488 (Wang et al., 2005).
Finally, to test whether the interaction between cannabinoid CB1 receptors and KOR occurs at the level of enteric nerves, we performed experiments on the isolated mouse ileum. We found that salvinorin A inhibited EFS-induced contractions in this preparation. The IC50 value found in our study (3.2 × 10−9 M) was similar to that previously calculated in the isolated guinea-pig ileum (Capasso et al., 2006). However, compared with the guinea-pig ileum (Capasso et al., 2006), here we found a lower efficacy of salvinorin A at reducing EFS-induced contractions, as indicated by the Emax values (that is, 38% in the mouse and 56% in the guinea-pig). Nevertheless, these results are in line with the ability of KOR opioid agonists to produce a partial inhibition of electrically evoked acetylcholine release from guinea-pig myenteric neurons (Kojima et al., 1994). The inhibitory effect of salvinorin A on EFS-induced contractions was reduced by both the KOR antagonist nor-binaltorphimine and the cannabinoid CB1 receptor antagonist rimonabant (Figure 4). These results suggest that the CB1/KOR interaction occurs, at least in part, on enteric nerves. In contrast to the in vivo results, here we have found that the interaction between cannabinoid CB1 receptors and KORs in vitro occurs also under physiological conditions. In addition, we found that concentrations of rimonabant higher than 3 × 10−8 M reduced EFS-induced contractions in vitro. This unexpected result is not in agreement with similar studies in the isolated guinea-pig ileum (Pertwee et al., 1996; Izzo et al., 1998; Begg et al., 2002) and, notably, with in vivo studies on motility in mice, which consistently showed a pro-kinetic effect of rimonabant (Calignano et al., 1997; Izzo et al., 2000; Carai et al., 2006). Discrepancies between in vitro and in vivo actions of cannabinoids, which have been previously documented in the digestive tract (Coruzzi et al., 2006; Capasso et al., 2008a), need further investigation.
In conclusion, from the study of the inhibition of intestinal motility by the hallucinogenic compound salvinorin A, we have shown, for the first time, a selective functional interaction between cannabinoid CB1 receptors and KORs in the inflamed gut in vivo. The identification of the biochemical nature of this interaction will require additional ad hoc studies.