Neuromodulatory actions of endocannabinoids

Vincenzo Di Marzo

Endocannabinoid Research Group, Istituto per la Chimica di Molecole di Interesse Biologico, Consiglio Nazionale delle Ricerche, Via Toiano 6, 80072, Arco Felice, Naples, Italy.

vdimarzo@icmib.na.cnr.it

Considerable progress has been made during the last 40 years towards the understanding of the mechanism of action of marijuana’s inhibitory effects on ambulation, cognition, pain perception and other CNS functions. In the early 1960’s it was found that this illicit drug, used for over four millennia also for its medicinal properties, owes most of its psychotropic effects to one of its major chemical constituents, (-)-D⁹-tetra-hydrocannabinol (THC). THC, and some of its synthetic analogues, the cannabinoids, act on both central and peripheral tissues mostly via two membrane receptors, the cannabinoid receptors, of which two subtypes are known, the CB₁ and CB₂ receptors. They are G-protein-coupled proteins belonging to the “seven trans-membrane domain” family of receptors. While CB₂ receptors are restricted to immune tissues, CB₁ receptors are distributed throughout the mammalian body, with the highest concentrations in some brain areas, such as the basal ganglia, the cerebellum and the hippocampus, as well as in reproductive, gastrointestinal and cardiovascular tissues. The ultra-structural localization of CB₁ receptors in neurons, as well as its co-localization with classical and peptide neurotransmitters and their receptors, is being revealed and suggests that activation of these receptors by THC and synthetic cannabinoids plays a neuromodulatory function in the CNS. In the 1990’s endogenous ligands of cannabinoid receptors, named endocannabinoids, were discovered. They are derivatives of polyunsaturated fatty acids, two of which have been studied most thoroughly: N-arachidonoylethanolamine (anandamide, AEA) and 2-arachidonoyl-glycerol (2-AG).

The endocannabinoids are not stored in synaptic vescicles but are produced “on demand” by intact neurons on stimulation with membrane depolarising stimuli, such as Ca²⁺ ionophores, glutamate and electrical stimulation, and immediately released via facilitated diffusion through the neuronal membrane. The termination of the endocannabinoid signal is achieved by a two-step mechanism comprising: a) re-uptake by cells via carrier-mediated facilitated diffusion, and b) enzymatic hydrolysis of endocannabinoid amide or ester bonds. While the major enzyme catalysing endocannabinoid hydrolysis, the “fatty acid amide hydrolase” (FAAH), has been characterized and cloned, the membrane carrier responsible for endocannabinoid re-uptake, also known as the “anandamide membrane transporter” (AMT), has not been isolated yet, even though evidence exists in favour of its proteic nature. Selective inhibitors for both FAAH and AMT have been developed. The endocannabinoids are synthesized by neurons from membrane phospholipid precursors. AEA is obtained by the cleavage of N-arachidonoyl-phosphatidylethanolamine, catalysed by a phospholipase D, whereas 2-AG is produced from the hydrolysis of sn-1-acyl-2-arachidonoyl-glycerols, catalysed by a diacylglycerol lipase. Endocannabinoid precursors are products of membrane phospholipid remodelling, and no selective inhibitor of their biosynthesis has been developed.

Activation of CB₁ receptors by exogenous and endogenous cannabinoids is coupled, via Gi/o proteins, to a series of intracellular events: a) inhibition of stimulus-activated cAMP formation, b) inhibition of voltage-activated N, P and Q Ca²⁺ channels, c) activation of inwardly rectifying K⁺ channels, d) activation of mitogen-activated protein kinase, and e) activation of nitric oxide release. Through these effects, pre-synaptic CB₁ receptors can inhibit the vescicle-mediated synaptic release of neurotransmitters, reduce the likelihood of membrane depolarisation and of action potential generation, whereas post-synaptic CB₁ receptors can modulate slow neurotransmission and modulate synaptic plasticity. The action and, more often, release of several neurotransmitters, namely GABA, glutamate, dopamine, noradrenaline and acetylcholine, has been reported to be inhibited by stimulation of pre-synaptic CB₁ receptors, which can also enhance the activity of GABA, for example in the globus pallidus, by inhibiting its re-uptake by neurons.

Recently, CB₁ receptors were found to be associated with nerve fibers and axon terminals but not in neuronal somata. This pattern is consistent with the pre-synaptic inhibitory effects of cannabinoids on neurotransmitter release in the brain. CB₁-expressing cells in mouse forebrain can be divided into distinct neuronal subpopulations. The majority of the cells that highly express CB₁ are GABAergic neurons belonging mainly to the cholecystokinin-positive type of interneurons (basket cells). In the hippocampus, amygdala and entorhinal cortex area, CB₁ mRNA is present at low but significant levels in many non-GABAergic cells that can be considered as projecting principal neurons. These data are in good agreement with the observation that cannabinoids act on principal glutamatergic circuits as well as modulate local GABAergic inhibitory circuits by inhibiting glutamate and GABA release. Interestingly, CB₁ mRNA is found in striatonigral neurons that contain dynorphin and substance P and striatopallidal neurons that contain enkephalin. A similar co-localization pattern, if found also in other brain regions, may account for some of the effects of exogenous and endogenous cannabinoids on CNS functions, including inhibition of both short and long term memory, suppression of motor behaviour and induction of catalepsy, sedation and analgesia.

Several pharmacological studies have shown that endocannabinoids are involved in the control of pain perception. However, experiments carried out with cannabinoid receptor antagonists or CB₁ receptor knockouts have reported contrasting results. Hence the need to perform analytical investigations aimed at correlating the tissue levels of endocannabinoids with various nociceptive responses. Electrical stimulation of the periaqueductal grey (PAG) was shown to induce CB₁-mediated analgesia while leading to the release of AEA into microdialysates from this region of the brainstem. Also the injection of the chemical irritant formalin into the paw induced a nociceptive response concomitantly to the release of AEA from the PAG. An endocannabinoid tone may down-modulate pain perception via CB₁ receptors in another region of the brainstem, the rostral ventromedial medulla, through the same circuit previously shown to contribute to the pain-suppressing effects of morphine. Other studies have shown that blockade of the action or expression of spinal CB₁ receptors by a CB₁ receptor antagonist or a CB₁ anti-sense oligonucleotide, respectively, leads to hyperalgesia, thus suggesting the existence also of a spinal endocannabinoid tone down-modulating nociceptive responses. The capability of AEA to interact directly with a particular type of nociceptors, the VR1 vanilloid receptors, thereby inducing sensory neuron activation and hyperalgesia, possibly followed by desensitisation of the nociceptive response, should be taken into account when studying the analgesic actions of this endocannabinoid.

Finally, endocannabinoids, and AEA in particular, have been claimed to participate in the control of the sleep-wake cycle, possibly as intermediates of the sleep-inducing factor oleamide, another fatty acid derivative that accumulates in the CSF of sleep-deprived mammals, and induces sedation in rats. Oleamide is hydrolysed by FAAH, and, apart from directly modulating GABA_A receptors like an anesthetic, and activating 5-HT receptors, may also act by increasing AEA levels through substrate competition with FAAH. In fact, the sedative effects of oleamide, which does not activate CB₁ receptors, are significantly reduced with a CB₁ receptor antagonist.

Reviews for Reference: Di Marzo V et al., (1998) Trends Neurosci., 21, 521-528; Mechoulam R et al., (1998) Eur. J. Pharmacol., 359, 1-18; Di Marzo V & Deutsch D, (1998) Neurobiol. Dis. 5, 386-404; Szallasi A & Di Marzo V, (2000) Trends Neurosci., 23, 491-497; Pertwee RG, (2001) Prog. Neurobiol., 63, 569-611.