Elsevier

Life Sciences

Volume 78, Issue 5, 22 December 2005, Pages 527-531
Life Sciences

Minireview
Psychopharmacology of the hallucinogenic sage Salvia divinorum

https://doi.org/10.1016/j.lfs.2005.09.008Get rights and content

Abstract

At present, the Mexican mint Salvia divinorum is an unregulated hallucinogen. This has resulted in various on-line botanical companies advertising and selling S. divinorum as a legal alternative to other regulated plant hallucinogens. It is predictable that its misuse will increase rapidly. The active ingredient in S. divinorum is the neoclerodane diterpene, salvinorin A (1a), which has been shown to be a κ agonist both in vitro and in vivo. This review will cover the current state of research into the psychopharmacology of S. divinorum.

Introduction

Of the almost 1000 species of Salvia in the world, none has fired the imagination as much as Salvia divinorum Epling and Játiva-M (Reisfield, 1993). S. divinorum is a plant from the Sage family that has been used in traditional spiritual and ethnopharmacological practices by the Mazatec Indians of Oaxaca, Mexico to produce “mystical” or hallucinogenic experiences (Váldes III et al., 1983). The western world first learned of this species in 1962 when Epling and Játiva-M described the plant from specimens collected by Hofmann and Wasson (Hofmann, 1980). The plant was named S. divinorum after its use in divination by the Mazatec Indians. Other native uses for the plant include the treatment of diarrhea, headache, and rheumatism. In addition, the plant is used to treat a semi-magical disease known as panzón de barrego, or swollen belly, which is caused by a curse from an evil sorcerer (Váldes III et al., 1983, National D.I.C., 2003). Young adults have begun to smoke the leaves and leaf extracts of the plants recreationally to induce powerful hallucinations (Hazelden, 2004). Currently, S. divinorum is unregulated in most countries and available throughout the world over the internet. It is, however, listed as a controlled substance in Denmark, Australia, and Italy. To date, U.S. laws for controlled substances do not ban the use of S. divinorum or its active components. This has resulted in many internet companies advertising and selling S. divinorum-derived products as legal hallucinogens.

The active constituent in S. divinorum is the neoclerodane diterpene salvinorin A (1a) (Fig. 1) (Siebert, 1994, Váldes III, 1994). A smoked dose of 200 to 500 μg in humans produces profound hallucinations lasting approximately 1 h (Siebert, 1994, Váldes III et al., 2001). Thus, it has a potency in humans that is similar to the highly active synthetic hallucinogen LSD (Sheffler and Roth, 2003).

Curiously, 1a does not act at the presumed molecular target responsible for the actions of classical hallucinogens, the serotonin 5-HT2A receptor (Egan et al., 1998, Glennon et al., 1984, Nichols, 2004). An initial pharmacological screen using a NovaScreen™ protocol was unsuccessful in identifying the molecular target for the hallucinogenic activity of 1a (Siebert, 1994). A more recent screening protocol identified the molecular target of 1a to be the κ opioid receptor (Roth et al., 2002). Additional studies have shown 1a to be a potent and selective and highly efficacious κ agonist in vitro and in vivo (Butelman et al., 2004, Chavkin et al., 2004, Roth et al., 2002).

S. divinorum is a relatively rare plant and few chemical studies have characterized its components. The first compounds isolated from S. divinorum were the neoclerodane diterpenes salvinorin A (1a) and salvinorin B (1b) (Ortega et al., 1982, Váldes III et al., 1984). Prior to this report, Valdés III was working on the isolation and characterization of the psychoactive substance from S. divinorum (Váldes III, 1983). Infusions of the plant had been shown to possess psychotropic activity (Váldes III et al., 1983) but the component responsible for this activity and its mechanism of action were not known. Having ascertained the active component to be a terpenoid, efforts were initiated by Valdés III to identify the molecular target of this compound. These efforts were largely unsuccessful. A manuscript describing the isolation of the psychotropic terpenoid, divinorin A and its congener divinorin B, was then submitted to the Journal of Organic Chemistry. Comparison of the structures of divinorin A and divinorin B with 1a and 1b isolated by Ortega et al. found these compounds to be identical. Therefore, divinorin A and B are now called salvinorin A and B, respectively. Later work by Valdes isolated another bneoclerodane diterpene salvinorin C (2). However, it has been suggested that 2 has no psychotropic activity (Siebert, 2004). Additional neoclerodane diterpenes, salvinorin D–F (35) and divinatorins A–C (68) have been isolated (Bigham et al., 2003, Munro and Rizzacasa, 2003). Further phytochemical investigations in the author's laboratory have identified salvinicin A (9) and B (10) (Harding et al., 2005, Harding et al., in press). Furthermore, various neoclerodanes have been prepared semi-synthetically (Beguin et al., 2005, Harding et al., 2005, Munro et al., 2005).

Currently, there are no radioligands or pharmacological tools of the salvinorin A chemotype. However, a deuterium labeled analog of salvinorin A (1c) and its utility as an internal standard for the detection of 1a and its metabolites in biological fluids by LC-MS has been described (Tidgewell et al., 2004).

As mentioned earlier, 1a was found to be a potent κ agonist in vitro (Roth et al., 2002). Using a screen of 50 receptors, transporters, and ion channels, 1a showed a distinctive profile that was different than the classical hallucinogen, lysergic acid diethylamide (LSD). Functional studies also demonstrated that 1a is a potent and selective κ opioid agonist at both cloned κ opioid receptors expressed in human embryonic kidney-293 cells and native κ opioid receptors expressed in guinea pig brain.

There has been only one report of behavioral testing of 1a in nonhuman primates (Butelman et al., 2004). All subjects (n = 3) dose-dependently exhibited over ≥ 90% U69,593-appropriate responding after subcutaneous injection of 1a (0.001–0.032 mg/kg). Quadazocine (0.32 mg/kg), a κ selective opioid antagonist, blocked the effects of 1a. However, κ selective antagonist GNTI (1 mg/kg; 24 h pretreatment) did not cause significant antagonism of the effects of 1a (GNTI, under these conditions, was only effective as an antagonist in two of three monkeys). Therefore, 1a produces discriminative stimulus effects similar to those of a high efficacy κ agonist. Interestingly, ketamine (0.1–3.2 mg/kg) was not generalized by any subject. This work indicates that not all compounds that produce hallucinogenic or psychotomimetic effects in humans are generalized by subjects trained to discriminate U69,593.

Presently, few studies have been initiated to more fully understand the remarkable selectivity of 1a for κ opioid receptors. This is the first example of a nonnitrogenous opioid selective ligand. Given its unique structure, its mode of interaction with the κ opioid receptor is not clear. A recent report explored the role of the 2-acetyl group of 1a on affinity and selectivity for κ opioid receptors (Chavkin et al., 2004). Structural modification of this position resulted in a change in activity from a full agonism to partial agonism for inhibition of forskolin-stimulated cAMP production. In particular, 1a was found to be a full agonist while propionate 11a and heptanoate 11b were found to be partial agonists in this assay (Chavkin et al., 2004). Surprisingly, 1a was found to be more efficacious than the selective κ agonist U50,488 and similar in efficacy to the naturally occurring peptide ligand for κ receptors, dynorphin A.

Recently, several analogues of 1a were evaluated for affinity at κ opioid receptors (Harding et al., 2005, Harding et al., in press, Munro et al., 2005). Salvinorin C (2) was found to have 250-fold lower affinity compared to 1a (Ki = 1022 nM vs. Ki = 4 nM) whereas 3 and 4 were found to have no affinity for κ receptors (Ki > 10,000 nM) (Munro et al., 2005). Additional work found that reduction of the furan ring (12) (Fig. 2) reduced affinity for κ receptors compared to 1a (Ki = 156 nM vs. Ki = 4 nM). Removal of the lactone carbonyl (13) was found to have little effect on binding compared to 1a (Ki = 6 nM vs. Ki = 4 nM). Reduction of the ketone to an α-alcohol (14) reduced affinity over 250-fold compared to 1a (Ki = 1125 nM vs. Ki = 4 nM). Removal of the ketone (15) resulted in a 5-fold loss of affinity compared to 1a (Ki = 18 nM vs. Ki = 4 nM). Finally, the C-8 epimer of 1a (16) was found to have 41-fold lower affinity for κ receptors compared to 1a (Ki = 163 nM vs. Ki = 4 nM). Interestingly, benzoyl derivative 17 was found to be the first example of a nonnitrogenous a μ opioid receptor agonist (Harding et al., 2005, Harding et al., in press). In addition, mesylate 18 was found to be more potent as an agonist at κ receptors compared to 1a.

The pharmacological activity of 1a was compared to two other structurally distinct κ ligands, 3FLB (18) and TRK-820 (19) (Wang et al., 2005). Binding affinities using [3H]diprenorphine at κ receptors were in the order of 19 (Ki = 75 pM) > 1a (Ki = 7.9 nM) > (Ki = 248 nM). All compounds were found to be full agonists in the [35S]GTP-γ-S binding assay in the order of 19 (EC50 = 25 pM) >> 1a (EC50 = 2.2 nM) > 18 (EC50 = 73.6 nM). Interestingly, 1a was found to be 40-fold less potent in promoting internalization of the hKOR compared to U50,488 and showed little anti-scratching activity and no antinociception in mice (Wang et al., 2005). It has been speculated that the divergence between the in vivo and in vitro effects of 1a may be due to in vivo metabolism of 1a to metabolites that are inactive at the κ opioid receptor (Wang et al., 2005). Another possibility for the discrepancies is that 1a may be interacting with additional receptors, ion channels, and/or transporters.

Recently, systemic administration of 1a has been found to elevate intracranial self-stimulation levels (ICSS) in rats (Todtenkopf et al., 2004). This depressive-like effect was found to be qualitatively similar to the systemic administration of U69,593. Pretreatment with the selective κ opioid antagonist ANTI (5′-acetylamidinoethylnaltrindole) dose dependently blocked elevations in ICSS threshold effects. This finding then suggests that stimulation of κ receptors in rats triggers depressive-like signs in a behavioral model.

The potenital toxicity and metabolism of 1a has not been fully investigated in laboratory animals or humans. An initial study examined the potential toxicity of 1a in rodents (Mowry et al., 2003). This study showed that little to no toxicity associated with high doses of 1a in mice. However, the study was carried out for only two weeks. No significant histologic differences between the control mice and the ones treated with doses of 1a were found. However, this does not mean that potential toxicities do not exist.

Presently, the identity of the metabolites of 1a are not definitely known. It was suggested that 1b is a metabolite of 1a (Roth et al., 2004, Váldes III et al., 2001). However, there are few analytical methods to study the routes of metabolism of 1a in vitro or in vivo. One method for determining the concentration of 1a in human and rhesus monkey plasma, rhesus monkey cerebrospinal fluid, and human urine by negative ion LC-MS/APCI has recently been described (Schmidt et al., 2005a). The fully validated method had a lower limit of detection using FDA guidelines of 2 ng/mL for 0.5 mL plasma samples; the linear range was from 2–1000 ng/mL. Several derivatives in the salvinorin family can also be analyzed by this method. The method has been used to establish that 1b is the principal metabolite of 1a ex vivo. However, 1b was not found in significant amounts in plasma of nonhuman primates. A preliminary study indicated that the elimination half-life of 1a in nonhuman primates was found to be 56.6 ± 24.8 min for all subjects tested (Schmidt et al., 2005b).

Section snippets

Conclusion

S. divinorum is an unregulated hallucinogenic plant whose use is increasing. The active component of S. divinorum is the neoclerodane diterpene salvinorin A (1a). In vitro pharmacological studies have found 1a to be a potent and selective κ agonist in vitro. In vivo studies indicate that 1a produces discriminative stimulus effects similar to those of a high efficacy κ agonist. However, there are discrepancies between the in vitro and in vivo effects of 1a. Preliminary structure activity

Acknowledgements

The author wishes to thank Leander J. Valdés III for a critical reading of the manuscript and the Biological Sciences Funding Program of the University of Iowa and the National Institute on Drug Abuse for financial support of this work.

References (32)

  • E.R. Butelman et al.

    The plant-derived hallucinogen, salvinorin A, produces kappa-opioid agonist-like discriminative effects in rhesus monkeys

    Psychopharmacology

    (2004)
  • C. Chavkin et al.

    Salvinorin A, an active component of the hallucinogenic sage Salvia divinorum is a highly efficacious κ-opioid receptor agonist: structural and functional considerations

    Journal of Pharmacology and Experimental Therapeutics

    (2004)
  • C.T. Egan et al.

    Agonist activity of LSD and lisuride at cloned 5HT2A and 5HT2C receptors

    Psychopharmacology

    (1998)
  • W.W. Harding et al.

    Neoclerodane diterpenes as a novel scaffold for μ opioid receptor ligands

    Journal of Medicinal Chemistry

    (2005)
  • Harding, W.W., Tidgewell, K., Shah, K., Dersch, C.M., Synder, J., Parrish, D., Deschamps, J.R., Rothman, R.B.,...
  • Hazelden Foundation

    Drug Abuse Trends

    (2004)
  • Cited by (101)

    • A systematic review on the kappa opioid receptor and its ligands: New directions for the treatment of pain, anxiety, depression, and drug abuse

      2022, European Journal of Medicinal Chemistry
      Citation Excerpt :

      Salvinorin A (Fig. 2), a neo-clerodane diterpene, is the main active ingredient of the hallucinogenic plant Salvia divinorum that has been safely used by the Mazatec for centuries in religious rituals [63]. First isolated from the plant in 1982 by Ortega and coworkers [64], salvinorin A is one of the most potent, naturally occurring opioid agonists, with high selectivity and affinity toward KOR (Ki = 4 nM, EC50 in [35S]GTPγS binding assay = 2.2 nM) [65]. In 2002, Roth and coworkers discovered that salvinorin A targeted KORs expressed in both human embryonic kidney-293 cells (Ki = 16 nM) and guinea pig brain (Ki = 4.3 nM) [66].

    • Psychoactive plants and phytochemicals

      2021, Phytochemistry, the Military and Health: Phytotoxins and Natural Defenses
    • Toxicity of Drugs of Abuse

      2018, Veterinary Toxicology: Basic and Clinical Principles: Third Edition
    View all citing articles on Scopus
    View full text