The cannabinoid receptor 2 (CB₂R) along with the cannabinoid receptor 1 (CB₁R) are the receptors of the endogenous cannabinoid system (ECS). Early research focused on elucidating the physiological role and therapeutic utility of CB₁R in neuropsychiatry since it was the first cannabinoid receptor identified in the central nervous system (CNS). However, the detection of CB₂R in neurons and glia in different brain regions has increased the number of investigations to elucidate its role at the central level, beyond its initial consideration as a “peripheral cannabinoid receptor”. This review summarizes the main findings highlighting the involvement of CB₂R in different physiological processes such as stress response, emotional reactivity, and cognitive processing that are affected in different psychiatric disorders (anxiety, depression, schizophrenia, and bipolar disorder), as well as the evidence supporting its therapeutic potential in the pharmacological approach to these pathologies.

ECS components are endogenous ligands, including anandamide (AEA) [1] and 2-arachidonoyl glycerol (2-AG) [2, 3], synthesis enzymes, such as N-acylphosphatidylethanolamine phospholipase D (NAPE-PLD) and diacylglycerol lipase (DAGL), metabolizing enzymes, such as fatty acid amide hydrolase (FAAH) [4, 5] and monoacylglycerol lipase (MAGL) [6–9], and cannabinoid receptors CB₁R [10], CB₂R [11] and, more recently, G protein-coupled receptor 55 (GPR55) [12, 13], known as non-CB₁ non-CB₂ cannabinoid-related receptor (Figure 1). Besides, cannabinoids act on transient receptor potential (TRP) channels, particularly on TRP vanilloid 1 (TRPV1), causing these receptors to be called ionotropic cannabinoid receptors [14, 15].

Contrary to most neurotransmitters, endogenous cannabinoids are not stored in synaptic vesicles but are synthesized and released on demand by postsynaptic neurons and accompanying glial cells in response to changes in neuronal activity [16–18]. Once released, they act on specific cannabinoid receptors, predominantly CB₁R and CB₂R.

CB₁R is widely distributed in the brain (cingulate cortex, entorhinal cortex, caudate nucleus and putamen, hippocampus, amygdala, thalamus, substantia nigra, medial hypothalamus, globus pallidus, nucleus tractus solitarius, cerebellum, and substantia gelatinosa of the spinal cord) [10, 19, 20]. This receptor has been proposed as a therapeutic target in the treatment of chemotherapy-associated nausea and vomiting [21, 22], in the treatment of anorexia in patients with acquired immunodeficiency syndrome [23, 24], chronic pain [25], spastic pain in multiple sclerosis [26], psychosis [27] and in anxiety disorders and depression [28–30].

The first attempts to identify the expression of CB₂R in the brain under normal conditions resulted unsuccessful [31–35]. In 2005, Van Sickle et al. [36], published the first information identifying the neuronal expression of CB₂R in the brainstem of rats, mice, and ferret under physiological conditions. These results served as a precedent for the full characterization of CB₂R expression in rat and mouse brains years later [37–39].

The wide distribution of CB₂R, both in neurons and glia, in various brain areas (striatum, cerebral cortex, amygdala, hippocampus, substantia nigra, olfactory tubercle, thalamic nuclei, ventromedial nucleus of the hypothalamus, paralemniscal nucleus, pontine nuclei, red nucleus, inferior colliculus, spinal nucleus, paratrochlear nucleus, and cerebellum), ended the idea that CB₂R presents an exclusively peripheral role, opening news avenues to research on its potential involvement in different brain functions and neuropsychiatric disorders [37, 39–44].

The presence of CB₂R in brain areas closely related to the response to stress and anxiety, such as the hippocampus and the amygdala, has led to the study of its possible involvement in these conditions [37, 39]. One of the main tools to evaluate the involvement of this receptor in anxiety has been the development of animal models that allow us to assess the response of rodents to anxiogenic stimuli. These models have been employed to evaluate whether genetic or pharmacological manipulation of CB₂R modifies the response to anxiogenic stimuli in different experimental paradigms.

Interestingly, our group showed that transgenic mice overexpressing CB₂R in the CNS (CB₂xP mice) present a resistant phenotype against anxiogenic stimuli in the light-dark box and the elevated plus-maze [43]. Complementarily, mice lacking the CB₂R gene (CB₂^–/– mice) develop higher levels of anxiety in both behavioral tests [44].

Another approach to study the role of CB₂R in anxiety has been to evaluate the effects of intracerebroventricular (icv) administration of an antisense probe designed against CB₂R messenger RNA in rodents [45]. In this study, the authors demonstrated that the blockade of the CB₂R gene is associated with a reduction in anxiety-like behaviors, in contrast to the results obtained with genetically modified mice. These discrepancies could be related to the antisense probe administered icv, which may affect different brain regions and in a different manner (more acutely) in comparison with the genetic deletion, in addition to the distinct genetic background of the strains of mice used in these studies [DBA/2, C57BL/6J, BALB/c, and institute of cancer research (ICR)] (Table 1).

The effects of the pharmacological modulation of CB₂R by agonists and antagonists were also evaluated. Acute administration of the CB₂R agonists GW405833 (100 mg/kg) and JWH015 (1–20 mg/kg) show anxiolytic effects on the rotarod [46] and the elevated plus-maze [38], respectively. However, depending on the agonist, the dose used, and the test employed, some authors find no differences [46, 47] or, on the contrary, anxiogenic effects are observed [45]. These discrepancies show that the effects would be influenced by several factors such as 1) the drug used, 2) the route and pattern of administration (acute vs. chronic) evaluated, 3) the behavioral test, and 4) the species and animal strains studied (rat, mouse; Table 1).

The effects induced by the blockade of CB₂R by antagonists, such as that produced by compound AM630, were also evaluated. Its acute administration (1, 2, or 3 mg/kg) in ICR mice induces anxiogenic effects in the LDB. Interestingly, these effects are not found in the previous administration of a CB₂R agonist (JWH133), supporting the involvement of CB₂R in anxiety [47]. The change of the administration pattern from acute to chronic also modifies the effects mediated by CB₂R agonist or antagonist compounds. Indeed, chronic administration of the antagonist AM630 (1, 2, or 3 mg/kg) exhibits anxiolytic effects in the light-dark box and the EPMs, while the agonist JWH133 (0.5, 1, or 2 mg/kg) presents anxiogenic actions. These effects are accompanied by increased CB₂R gene expression in the amygdala and cortex, key regions involved in emotional response, in AM630-treated mice. Moreover, chronic activation of CB₂R by the agonist JWH133 reduced CB₂R gene expression in both regions [47]. These findings agree with the results found in studies performed with genetically modified mice, in which CB₂R overexpression is associated with resistance to developing anxiety-like behaviors [43] and, on the contrary, its absence is associated with greater vulnerability [44].

In animal models of social defeat, paradigms of psychosocial anxiety, a reduction of CB₂R gene expression was found in different regions of the hippocampus [cornu ammonis-1 (CA1) and dentate gyrus] both in mice that develop submissive behavior towards the aggressor and in those that develop avoidance behavior [48]. Additionally, it has been noted that the absence of CB₂R produces increased aggressiveness in the social interaction test and the social defeat model [49]. In this same study, the acute administration of the agonist JWH133 (1, 2, or 4 mg/kg) reduces the aggressive behavior of wild-type mice exposed to this model. Taken together, these results support the involvement of the CB₂R in the regulation of adaptive responses to environmental factors.

Few clinical studies have been performed but with promising results. A relationship was found between the rs2070956 polymorphism of CB₂R and a poorer response to treatment in children with anxiety disorders [50]. Taking together, more studies in both mice and humans are needed to further investigate the role of CB₂R in anxiety disorders, especially from the perspective of its potential therapeutic suitability.

In this section, it is essential to highlight the findings of experimental animal and human studies that revealed the potential therapeutic role of CB₂R in depressive disorders. Genetic and pharmacological studies in rodents indicate that overexpression of CB₂R is associated with increased resistance to stimuli that promote depressive-like behaviors. CB₂xP mice (overexpressing CB₂R) showed reduced immobility time in acute models of depression (tail suspension and novelty-suppressed feeding tests) [39]. Similarly, mice lacking this receptor (CB₂^–/– mice) present a higher degree of vulnerability to develop depressive-like behavior in the tail suspension test [44]. The results found with the conditional knockout mouse lacking CB₂R in dopaminergic neurons (DAT-Cnr2^–/–) are similar, that is, they present increased immobility time in the tail suspension and forced swimming tests, suggesting increased vulnerability to stimuli promoting depressive-like behaviors in rodents [51] (Table 2).

Another important strength supporting the role of CB₂R in the development of depressive behaviors is provided by studies using the CMS model, the most relevant animal model of depression. This model induces behavioral and neurochemical alterations in rodents similar to the different signs and symptoms observed in depressive patients, such as anhedonia, reduced sexual activity, increased aggressiveness, reduced grooming, and detriment of neurogenesis processes, especially in the hippocampus [52, 53]. Exposure to this model does not induce any alteration in CB₂xP mice in any of the parameters evaluated, immobility time in the tail suspension test, and consumption of a 1% sucrose solution (anhedonia), both typical traits of depressive-like behaviors in rodents [39].

Additionally, several studies were conducted to evaluate whether pharmacological manipulation of CB₂R by the administration of a selective antagonist would reproduce a similar phenotype in wild-type mice to that observed in CB₂xP mice. This assumption is based on the evidence supporting that administration of receptor antagonists produces long-term up-regulation of the blocked receptor [54, 55]. To this end, the effects of chronic administration of the CB₂R antagonist AM630 [1 mg/(kg·12 h)] were evaluated in the CMS model. This study demonstrated how the administration of AM630 reversed depressive-like behavior [evaluated in the tail suspension test and the ingestion of a 1% sucrose solution (anhedonia)] after 4 weeks of treatment, an effect accompanied by an increase of CB₂R gene expression in the hippocampus [39]. In contrast to these data, it was previously shown that administration of AM630 [1 mg/(kg·24 h)] failed to exert any effect on sucrose consumption in CMS-exposed mice [41]. Again, these discrepancies could be due to notable differences between the two studies such as 1) the rodent strains employed (BALB/c, ICR), and 2) the dose and administration patterns evaluated [3 mg/(kg·24 h)] [41] in contrast to 1 mg/(kg·12 h) [39].

In humans, the research team led by Onaivi [41] showed that there is a high incidence of the Q63R polymorphism of CB₂R in Japanese depressive patients. Furthermore, a postmortem study performed in suicides revealed that there is a reduction of CB₂R gene expression in the dorsolateral prefrontal cortex and amygdala, key regions involved in decision making, impulsivity, and emotional reactivity, supporting the role that this receptor could play in several psychiatric diseases with an increased tendency to suicide, such as depression, schizophrenia, and bipolar disorder [56].

Together, these results highlight the involvement of CB₂R in the emotional response and its potential utility as a new target for the treatment of depressive disorders.

Currently, a large number of results support the involvement of CB₂R in psychotic disorders. Clinically, significant reductions in CB₂R gene expression were detected in peripheral mononuclear blood cells of schizophrenic patients in remission [57]. Also, an increased incidence of CB₂R polymorphisms rs12744386 and rs2501432 was found in Japanese schizophrenic patients [42], and of rs2501432 and rs22229579 CB₂R polymorphisms in the Han ethnic population in China [58]. The results of this latest study suggest that the T allele of the rs2501432 polymorphism would have a protective role, especially in men. Conversely, the T allele of the rs2229549 polymorphism would constitute a risk factor for the development of schizophrenia. However, further studies failed to establish a relationship between additional polymorphisms in CB₂R (rs6689530, rs34570472, Cnr2_ht1, Cnr2_ht2, and Cnr2_ht3) and schizophrenia in the Korean population, probably due to the small sample sizes included in them [59] (Table 3).

The use of different animal models of schizophrenia supported the involvement of CB₂R in this psychiatric disease. Early maternal deprivation constitutes an animal model for the study of schizophrenia by reproducing several symptoms (cognitive and attention alterations) observed in schizophrenic patients that translate into alterations in prepulse inhibition (PPI), startle reflex, and latent inhibition [60, 61]. These behavioral disturbances are accompanied by important neurobiological alterations in the hippocampus, increased number of astrocytes, neuronal degeneration, elevated corticosterone and 2-AG levels [62, 63], reduced CB₁R gene expression, and increased CB₂R gene expression [64].

One of the main tests used both in humans and rodents for the evaluation of attention-deficit associated with schizophrenic pathology is the PPI test in which reduced values are representative of the inability to filter the information presented by these patients. In this regard, the administration in rodents of methamphetamine or MK-801, a non-competitive N-methyl-D-aspartate (NMDA) glutamatergic receptor antagonist, are two highly accepted models of schizophrenia producing a reduction in PPI values similar to those observed in schizophrenic patients [65, 66]. While CB₂R blockade using the CB₂R antagonist AM630 (3 mg/kg and 30 mg/kg) does not trigger alterations per se in PPI, it exacerbates MK-801- or methamphetamine-induced alterations in C57BL/6J mice [42]. Another complementary study shows that the agonist JWH015 (1, 3, and 10 mg/kg) improves MK-801-induced impairment in PPI, without presenting any effects on its own [67]. The fact that this effect is blocked by AM630, but not by the CB₁R antagonist AM251, confirms the involvement of CB₂R.

Hyperlocomotion induced by psychostimulants or by MK-801 is another animal model of schizophrenia. The results obtained using this model have been promising. On the one hand, the administration of the agonist JWH133 dose-dependently reduces cocaine-induced hyperlocomotion in wild-type mice [68, 69]. On the other hand, AM630 at doses that do not affect motor activity (2 mg/kg) worsens MK-801-induced hyperlocomotion [70]. These results strongly suggest that the inhibition of CB₂R function potentially precipitates schizophrenia-like symptoms in rodents when combined with risk factors such as inhibition of glutamatergic transmission induced by NMDA receptor antagonists.

Studies carried out in genetically modified mice have confirmed the involvement of CB₂R in schizopsychotic disorders. CB₂^–/– mice present increased vulnerability to hyperlocomotion induced by acute cocaine administration, in addition to PPI and cognitive alterations [44, 71]. Moreover, treatment with the antipsychotic risperidone normalizes PPI values in CB₂^–/– mice. All these facts suggest that genetic deletion of CB₂R could be a useful tool as an animal model of schizophrenia [44].

Recently, it was shown that CB₂R is critical for acetylcholine muscarinic 4 (M4) receptor agonists, such as compound UV0467154, to exhibit their antipsychotic actions [72]. This study indicates that agonist-regulated stimulation of M4 receptors would increase the release of endogenous cannabinoids, in turn activating CB₂R, which would eventually inhibit dopamine release, a fact that would be involved in the antipsychotic efficacy of M4 agonists. Consequently, blocking CB₂R by the antagonist AM630 would prevent M4 agonists to present antipsychotic effects. This finding highlights that CB₂R is indispensable in the regulation of dopamine concentrations altered in this disease, pointing out its potential therapeutic utility in the treatment of psychotic disorders.

Taken together, these findings suggest that CB₂R is crucial in the regulation of symptoms related to psychotic disorders. Further pharmacological studies, such as those evaluating the additive effects of CB₂R agonist/antagonist combinations with antipsychotics (e.g., risperidone), are needed to clarify the role of pharmacological modulation of CB₂R in the treatment of psychotic disorders.

To date, few studies focused on investigating the possible link between CB₂R and bipolar disorder. There is only one study in which the incidence of three polymorphisms of the gene coding for CB₂R, rs2501432 (315>G; Arg63Gln), rs41311993 (524C>A; Leu133Ile) and rs2229579 (1073C>T; Tyr316His), was evaluated in 80 patients with bipolar disorder [73]. The results allowed to establish a direct relationship between the rs41311993 (524C>A) polymorphism of CB₂R and bipolar disorder. This polymorphism consists of the substitution of the amino acid leucine by isoleucine at position 133, involved in the structural stability of the receptor since it allows the formation of hydrophobic interactions between helix III with the rest of the 7 transmembrane domains (except helix I). Helix III, therefore, plays a primordial role in the regulation of CB₂R activity, participating in its interaction with the G protein-coupled to it. Consequently, the presence of this polymorphism could affect the structural stability of the receptor, as well as its interaction with the G protein. This finding is the first evidence to indicate that CB₂R may be involved in the etiology of bipolar disorder. Further studies, both clinical and preclinical, are needed to investigate the involvement of CB₂R in this psychiatric disorder.

CB₂R was identified in different cell types (neurons, glia, and astroglia) of the as well as its interaction with targets classically related to the response to stress, anxiety, depression, and schizophrenia, such as the hypothalamic-pituitary-adrenal (HPA) axis, processes of neurogenesis, and neuronal plasticity, and the GABAergic, dopaminergic and serotonergic neurotransmitter systems.

In summary, the findings included in this review highlight the potential therapeutic suitability of CB₂R in a wide variety of psychiatric disorders, including anxiety disorders, depressive disorders, schizophrenia, and disorders involving cognitive impairment. An additional value of CB₂R pharmacological manipulation is the fact that, contrary to CB₁R activation, does not induce psychotropic effects. The availability of agonist/ antagonist compounds of this receptor, along with recent allosteric modulators, represents a valuable tool for further experimental and preclinical studies to finally elucidate the role of the pharmacological regulation of CB₂R in psychiatry. Besides, the few clinical trials conducted to date showed that these compounds are well tolerated, at least at the doses and in the pathologies tested [109, 110]. Nonetheless, future studies are needed to evaluate safety in neuropsychiatric disorders.