Rich evidence has shown that cannabis products exert a broad gamut of effects on emotional regulation. The main psychoactive ingredient of hemp, Δ⁹-tetrahydrocannabinol (THC), and its synthetic cannabinoid analogs have been reported to either attenuate or exacerbate anxiety and fear-related behaviors in humans and experimental animals. The heterogeneity of cannabis-induced psychological outcomes reflects a complex network of molecular interactions between the key neurobiological substrates of anxiety and fear and the endogenous cannabinoid system, mainly consisting of the arachidonic acid derivatives anandamide and 2-arachidonoylglycerol (2-AG) and two receptors, respectively termed CB₁ and CB₂. The high degree of interindividual variability in the responses to cannabis is contributed by a wide spectrum of factors, including genetic and environmental determinants, as well as differences in the relative concentrations of THC and other alkaloids (such as cannabidiol) within the plant itself. The present article reviews the currently available knowledge on the herbal, synthetic and endogenous cannabinoids with respect to the modulation of anxiety responses, and highlights the challenges that should be overcome to harness the therapeutic potential of some of these compounds, all the while limiting the side effects associated with cannabis consumption.

Keywords: cannabis, anxiety, CB receptors, endocannabinoids, Δ⁹-tetrahydrocannabinol, cannabidiol

Anxiety is generally defined as an emotional state characterized by maladaptive and excessive emotional responsiveness to potentially dangerous circumstances. The pathological expression of anxiety leads to enduring emotional perturbations with a consistent apprehension towards the possibility of future, vaguely defined negative events [1]. According to the current classification of anxiety disorders in the fourth edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM-IV) [2], the main diagnostic entities in this category are:

• –
  generalized anxiety disorder (GAD), featuring general irritability, anxiety attacks, chronic apprehension/anxious expectation and secondary phobic avoidance.
• –
  panic disorder, characterized by brief (2-10 min) spells of overwhelming anxiety or fear, accompanied by somatic and cognitive symptoms;
• –
  social anxiety disorder (or social phobia), defined as extreme agitation in social contexts and avoidance of social situations;
• –
  obsessive-compulsive disorder (OCD), characterized by recurrent and intrusive anxiogenic thoughts (obsessions), and stereotyped behaviors (compulsions) aimed at the reduction of the distress caused by the obsessions.
• –
  post-traumatic stress disorder (PTSD), in which a prior intense trauma results in a long-lasting anxious response, with re-experiencing/flashback phenomena, avoidance and emotional numbing.

In keeping with their different clinical features and phenomenological presentations, these disorders are underpinned by divergent neurobiological alterations and respond to partially different pharmacotherapeutic strategies (outlined in Table 1). A fundamental contribution in our understanding of the neural bases of anxiety disorders and in the development of novel therapies has been afforded by animal models and testing paradigms for anxiety-like behaviors (summarized in Table 2).

Herbal cannabinoids

The three species included in the Cannabis genus (or sub-species, depending on the taxonomic classification; see [3], for a detailed discussion on the issue), sativa, indica and ruderalis, feature at least 85 unique terpenophenolic compounds, collectively named phytocannabinoids [4]. The main classes of phytocannabinoids are outlined in Figure 1. Quantitative analyses of cannabis constituents are usually performed by chromatographic techniques (generally Gas Chromatography, but also Thin-Layer Chromatography, or High-Performance Liquid Chromatography), often coupled with Mass Spectrometry. A detailed description of the instrumental methods used for classification and source tracing of Cannabis products (including DNA identification for forensic and intelligence purposes) is beyond the scope of this review, but can be found in [5–7].

The chemical fingerprinting of hemp products has revealed that the two most abundant phytocannabinoids are Δ⁹-tetrahydrocannabinol (THC, also named dronabinol) and cannabidiol (CBD):

The main psychoactive constituent of Cannabis, THC is a highly lipophilic alkaloid produced mainly in the leaves, flowers and glandular trichomes of the plant. Most of the pharmacological effects elicited by hemp products, including emotional and cognitive changes, analgesia, hypothermia and appetite stimulation, are considered to be reflective of the action of THC as a partial agonist of cannabinoid CB₁ and CB₂ receptors (see below). Additionally, THC has been shown to act as an acetylcholinesterase inhibitor [8–10].

In contrast with THC, CBD is not psychotropic, but has nevertheless been shown to play a role in the modulation of behavioral effects of cannabis [11]. In fact, the THC: CBD ratio is the main criterion to define different cannabis chemotypes [12] and has been posited to contribute to the variability in neurobehavioral outcomes of marijuana or hashish consumption [13,14]. Interestingly, most cannabis strains encountered in the illegal markets generally have elevated amounts of THC [15].

The different characteristics of THC and CBD are underpinned by their distinct mechanisms of action. Whereas THC has nanomolar affinity for both CB₁ (Ki = 25.1 nmol/L) and CB₂ (Ki = 35.2 nmol/L) receptors, CBD exhibits much lower affinity for either target [16–20]; however, the latter phytocannabinoid was recently found to act as a highly potent antagonist/inverse agonist of both CB receptors [21], possibly due to a non-competitive mechanism of receptor blockade [22]. Additionally, CBD has been shown to exert some of its actions through other receptors, including the vanilloid receptor VR₁ and the serotonin receptor 5-HT_1A (for a general overview of the topic, see [11]).

The other main phytocannabinoids, including cannabigerol (CBG), cannabichromene (CBC) and cannabinol (CBN) (Fig. 1) [4,23], have been shown to exert antibiotic and antiinflammatory properties, but have not been strongly associated with the behavioral effects of Cannabis; nevertheless, the recent discovery that CBG is a highly potent agonist for α₂ adrenoceptor and a blocker of serotonin 5-HT_1A receptor [24] underscores the potential importance of these and other alkaloids in the psychoactive profile of cannabis.

Synthetic cannabinoids

In addition to phytocannabinoids, several classes of synthetic CB receptor agonists have been developed; among these families, the best characterized are the synthetic analogs of THC – such as the biciclic compounds CP 47,497, CP 55,244, CP 55,940 and the benxopyrans HU-210 and nabilone (Fig. 2) – and the aminoalkylindole derivatives – including WIN 55,212-2, JWH-015, JWH-018, JWH-073, JWH-081 and JWH-398 (for a general review, see [23]). Of these agents, only nabilone has been approved for clinical use as an antiemetic treatment and an adjunct analgesic for neuropathic pain [25]. Other more potent synthetic cannabinoids, such as CP 47,497, HU-210 and most JWH compounds, have regrettably gained great popularity in the market of recreational substances during the last decade, under the generic brand names of “Spice” or “K2”. Unlike THC, which is a partial agonist of CB₁ receptors, these agents are full, high-potency CB₁ receptor activators [26,27], thereby eliciting greater psychotropic effects than THC (as CB₁ receptors are the key mediators of the psychotropic actions of cannabis). This characteristic, together with their legal status (recently revoked across most Western countries, including USA as of March 2011) and lack of available testing procedures for the detection of urinary metabolites, has unfortunately contributed to the great diffusion of “Spice” blends in Central and Western Europe, as well as Australasia.

CB receptors

Although CB₁ and CB₂ receptors only share 44% sequence identity (68% in the transmembrane domains), they are both coupled to G_i/o proteins [34] and activated by both anandamide and 2-AG. In line with their metabotropic nature, CB receptors mediate their intracellular response through a number of changes affecting signaling cascades, such as inhibition of adenylyl cyclase, activation of G-protein-activated inwardly rectifying potassium channels (GIRKs) and phosphorylation of extracellular signal-related kinases (ERKs) [35,36]. The distribution pattern of CB₁ and CB₂receptors is strikingly divergent, indicating diverse physiological functions: CB₁ is the most abundant metabotropic receptor in the brain, and is primarily distributed in the synaptic terminals of neurons across all the major structures that regulate emotional responsiveness, perception and memory, including prefrontal cortex, amygdala, septo-hippocampal system, striatum, thalamus, brainstem nuclei etc. [37–41]. CB₁ receptors are typically located on presynaptic terminals [42,43], but they have also been identified in postsynaptic locations [44,45]. Presynaptic CB₁ receptors are posited to serve critical functions for the regulation of synaptic plasticity and neurotransmitter release; in particular, they mediate the depolarization-induced suppression of inhibition (DSI) and depolarization-induced suppression of excitation (DSE), consisting in the reduction of γ-amino-butyric acid (GABA) or glutamate release, respectively, from presynaptic boutons following stimulation of the postsynaptic terminals [46–49]. In general, CB₁ activation has been shown to inhibit the neurotransmission of other mediators, including glycine, acetylcholine, norepinephrine and serotonin [50], but the underpinnings of these phenomena have not been completely elucidated. Additionally, CB₁ receptors have been implicated in short- and long-term synaptic depression, in relation to phasic or tonic endocannabinoid release (for a review on these topics, see [51]).

The function of CB₁ receptors may vary depending on the specific interactions that they entertain with other molecular targets. For example, CB₁ receptors have been found to associate with other G-protein complex receptors, such as dopamine D2, orexin Ox1, μ opioid and adenosine A_2a, to form heteromeric complexes (reviewed in [52,53])

The key role of CB₁ receptors as mediators of neurochemical homeostasis in the brain is maintained through a complex regulation of their expression. For example, these receptors are subjected to a rapid internalization (via clathrin-coated pits) following their binding with full agonists; on the other hand, the receptors are also recycled, with a process that requires endosomal acidification and dephosphorilation [54].

While CB₂ receptors are abundantly expressed in most peripheral organs (and particularly in immune cells, where they regulate cytokine secretion and modulate cell trafficking) [55], their distribution in the brain appears to be sparse and particularly confined to microglial cells; nevertheless, recent evidence has revealed the presence of CB₂ receptors in several areas of the brain [56–58]. Interestingly, a number of studies suggest that neuronal CB₂ receptors may be mainly located in postsynaptic terminals [58,59]; nevertheless, the functional role of these targets in the brain remains largely elusive and awaits further characterization.

The existence of cannabinoid receptors other than CB₁ and CB₂ has been postulated based on ample experimental evidence [60–62]. Interestingly, a number of investigations have pointed to GPR55 as a novel putative cannabinoid receptor [63,64]; nevertheless, evidence on the specificity of this receptor for endocannabinoid is still inconclusive [65].

Endocannabinoids

Both anandamide and 2-AG are derivatives of arachidonic acid, an unsaturated C20 fatty acid with 4 double bonds, which also serves as the precursor for synthesis of other eicosanoids, including prostaglandins and leukotriens. Anandamide is found in picomolar concentrations and acts as a high-affinity partial agonist for both CB₁ and CB₂ receptors. It is synthesized on demand by enzymatic hydrolysis of the membrane phospholipid N-arachidonoyl phosphatidylethanolamine (NAPE), a process catalyzed by several phospholipases [66–68]. Following release and activation of CB receptors, anandamide is rapidly removed from the synaptic cleft by a carrier-mediated system [69–72] and subsequently hydrolyzed by the membrane enzyme fatty acid amide hydrolase (FAAH) [73–75]. FAAH serves the catabolism of other substrates, including oleoylethanolamine (OEA) and palmitoylethanolamine (PEA). Both these compounds do not activate CB1 receptors [76], although they may reduce or slow down anandamide degradation by competing with it for FAAH activity.

In comparison with anandamide, 2-AG is much more abundant (occurring in nanomolar concentrations across most tissues) and acts as a full agonist of both CB receptors. It is produced from 1,2diacylglycerol (DAG) by diacylglycerol lipase (DAGL) [77] and degraded mainly by the cytosolic serine hydrolase monoacylglycerol lipase (MAGL) [78], although other enzymes are known to contribute to this process [79].

The divergent neurochemical profiles of anandamide and 2-AG underscore their different physiological roles. Although our current understanding of the different functions entertained by each endocannabinoid is still rudimentary, the development of FAAH and MAGL inhibitors [80,81] has been instrumental to elucidate the implication of each mediator in synaptic and neurochemical regulation. While 2-AG is known as the retrograde mediator of DSI [82,83] and DSE [84–87], a number of studies suggest that anandamide may serve as an activity-dependent regulator of monoaminergic transmission [88–90]. Recent evidence points to a potential biological antagonism between anandamide and 2-AG [91,92]; on the other hand, emerging evidence points to a similar role of anandamide and 2-AG in the regulation of anxiety (albeit in relation to different receptors) and pain [93]. The development of JZL195, a potent FAAH/MAGL inhibitor, has in turn revealed that the behavioral effects of CB₁ receptor agonists can be only recapitulated by the combination of both endocannabinoid-mediated functions [94].

Other lipids have been indicated as putative endocannabinoids, including 2-arachidonoylglycerylether (noladin ether) [95] and O-arachidonoylethanolamine (virodhamine) [96] (Fig. 3). Additionally, recent evidence has identified that CB receptors may be modulated by peptidic ligands, such as hemopressin and its derivatives [97,98].

Cannabis, THC and CB₁ receptor agonists

The employment of cannabis for its medicinal, relaxing and mood-enhancing properties has been documented across most ancient civilizations. Originally introduced in Chinese pharmacopoeia during the third millennium BCE [99,100], cannabis became a popular remedy throughout Asia and Europe in the following centuries [99,101]. The inclusion of cannabis in the medical treatises by Dioscorides and Galen secured the herb a stable reputation in the Roman Empire and the Arabic world [101]. Until the early 20^th century, the plant remained a valuable therapy for a large number of diseases [102]; however, growing concerns about the psychoactive and narcotic effects of cannabis led to a progressive restriction and ultimate ban of its usage in the United States and several European countries [100,103]. Despite its illicit status, cannabis remains one of the most popular recreational drugs, particular among adolescents and young adults, in view of its mood-enhancing and euphoriant characteristics [104–106].

Most psychological and behavioral effects of marijuana and other hemp products are induced by THC through activation of CB₁ brain receptors. In fact, although THC and most synthetic cannabinoids are known to activate both CB₁ and CB₂ receptors, their actions on anxiety-like behaviors and emotional regulation are efficiently countered by selective CB₁ receptor antagonists, such as rimonabant (see next section) [107].

The studies on the psychological effects of cannabis and THC have unfolded a highly complex and often contradictory scenario, fostering a long-standing debate on the potential harms and benefits of its products. An important aspect of this discussion (particularly in consideration of its legal aspects and the potential therapeutic applications of hemp derivatives), revolves around the distinction between use and misuse of cannabis. In particular, whereas the abuse and dependence liability of cannabis is generally well-recognized [108,109], the definition of these phenomena has been heavily criticized as reflective of political agendas rather than scientific bases. For instance, the diagnosis of substance abuse, according to the criteria listed by the DSM–IV TR, is based on the manifestation of at least one of four symptoms: interference with major professional or personal obligations; intoxication in hazardous settings; substance-related legal problems; continued use in the face of persistent social or interpersonal problems [110]. The applicability of some of these standards to marijuana and other cannabis derivatives, however, has been questioned [99], also in view of their lower potential to induce physical harm in comparison with other legal substances, such as alcohol and tobacco [111].

While the controversies surrounding cannabis are far from subdued (and are often permeated and masked by conflicting ideological credos), standardized studies on cannabinoids have highlighted that the psychological and behavioral outcomes of this substance are highly variable and range from relaxation, euthymia and heightened sociability to panic, paranoid ideation and psychosis [112–116]. A corollary of this observation is that the high comorbidity rate between cannabis use disorders and psychiatric conditions [100–105] may indicate that cannabis consumption is either a concurring cause or a “self-therapeutic” strategy for anxiety and mood disorders [117–123]. The latter interpretation is supported by the observation that anxiety-spectrum disturbances and traumas in early developmental stages are a strong predictor for later cannabis use disorders [124–127]; furthermore, several lines of evidence suggest that the anxiolytic effects of THC may partially account for the high prevalence of cannabis use in patients affected by PTSD [128–131] and OCD [132]. Accordingly, recent clinical studies have shown that THC elicits therapeutic effects in OCD [133] and trichotillomania, an impulse-control disorder characterized by compulsive hair-pulling [134]. Nevertheless, prospective analyses show that cannabis use and dependence increase the risk for development of panic disorder [135], suggesting that the effect of cannabis may vary with respect to the nosological entities within the spectrum of anxiety disorders. Of note, chronic consumption of cannabis has been hypothesized to exacerbate depressive or anxious manifestations, and reduce the therapeutic efficacy of anxiolytic agents [122,136–138]; an interesting theoretical implication of this finding is that long-term exposure to cannabinoid agents may lead to profound alterations of synaptic plasticity and neurochemical homeostasis and alter the pathophysiological trajectory of anxiety and mood disorders. Thus, while cannabis may be initially used as a self-therapy for certain anxiety disorders, the prolonged exposure to this substance in vulnerable individuals may in turn alter or aggravate the clinical course of these conditions and render the patients refractory to standard treatments.

The ability of cannabis to either exacerbate or attenuate emotional reactivity is highly influenced by numerous factors, including its chemotype, as well as the influence of genetic, developmental and contextual variables. Unfortunately, little is still known about the susceptibility factors that govern the behavioral outcomes of cannabis in patients affected by anxiety-spectrum disorders. Indeed, several components have been shown to play a role in this link, including genetic background, age, gender, environmental stress and concurrent use of other drugs; a detailed analysis of these determinants is outside the scope of the present work, but the interested reader should refer to [139].

Aside from the influence of vulnerability factors, the available evidence indicates that cannabis, THC and other CB₁ receptor agonists exercise a bidirectional influence on anxiety responses as a function of the dosage. The majority of users report that consumption of modest amounts of cannabis and CB₁ receptor agonists results in euphoria, relaxation, heightened perception, sociability and creativity, moderate to high doses have been reported to elicit phobia, agitation, panic, dysphoria, psychotic manifestations and cognitive impairments [112–116,124,140–143]. In line with these premises, early studies showed a robust anxiolytic efficacy of low-dose nabilone in comparison with placebo [144,145]. Additionally, the few available reports on the clinical outcomes of recreational cannabinoids show that a moderate consumption of “Spice” blends is generally associated with euphoria and disinhibition [146], but the abuse of these substances is conducive to high levels of anxiety, panic, paranoid ideation and mood disturbances [147–151].

The biphasic effects of cannabinoids on anxiety-related responses have been extensively documented in rodents. In agreement with human evidence, preclinical studies have elucidated that the acute administration of low doses of CB₁ receptor agonists elicits anxiolytic-like in approach/avoidance tasks [152–156]; conversely, high concentrations of the same compounds are generally associated with the opposite outcomes [157–162] (for complete reviews of the topic, see [163,164])

The bidirectional action of CB₁ receptors on anxiety responses may be related to the modulatory role of these targets on GABA and glutamate release across amygdala and other forebrain areas [41,165,166]. As these two major neurotransmitters affect anxiety in an opposite fashion, different doses of cannabinoids and synthetic CB₁ receptor agonists may indeed produce highly divergent effects, in relation to their ability to affect the homeostasis and the balance of GABA and glutamate (for a review on these issues, see [163]). Furthermore, CB₁ receptors have been shown to play critical roles in the regulation of most neurochemical substrates of anxiety, including the neurotransmitters serotonin, norepinephrine and acetylcholine, as well as stress hormones, colecystokynin and opioid peptides [50,163].

In line with this concept, the infusion in the periaqueductal grey of arachidonyl-2-chloroethylamide (ACEA), an anandamide synthetic analog with high CB₁ receptor selectivity, elicited anxiolytic-like effects in rats in an elevated plus maze, with a bell-shaped dose-response curve [167], the highest doses being associated to no significant behavioral change. Novel categories of compounds have been patented for potential efficacy as selective CB₁ receptor modulators, including sulfonyl-benzamides [168] and tetrasubstituted imidazole derivatives [169]. To the best of our knowledge, however, no findings on the action of these compounds in anxiety regulation have been reported to date.

CB₁ receptor antagonists/inverse agonists

The cannabinoid CB₁ receptor antagonists/inverse agonist rimonabant was introduced into clinical practice by Sanofi-Aventis in 2006 as a treatment for obesity [170] and smoking cessation [171]. The majority of preclinical studies found that these compounds are anxiogenic at high doses [158,159,172,173] and ineffective at low doses [174,175]. The anxiogenic properties of CB₁antagonists, were unequivocally confirmed by clinical data on the psychiatric side effects of rimonabant. The significant increase in anxiety, depression and suicidality in patients under treatment with rimonabant [176–179], in particular, led to the withdrawal of the drug from the European market in October, 2008. As a consequence, several pharmaceutical companies announced the interruption of their clinical research on CB₁ receptor antagonists, including taranabant (from Merck) and otenabant (from Pfizer), both in Phase 3 of development. Some of the anxiogenic properties of rimonabant and analogs have been speculated to be due to their activity as inverse agonists; as a result, the therapeutic use of newly-developed neutral CB₁ antagonists has been proposed, with the hypothesis that these compounds would not elicit the untoward psychological effects observed with rimonabant and its analogs [180,181]; this idea is supported by recent findings, showing that unlike CB₁ receptor inverse agonists, the neutral antagonists of this targets fail to facilitate the acquisition or consolidation of fear [182].

CB₂ receptor ligands

Few studies have actually evaluated the role of CB₂ receptor in anxiety and stress response. While this receptor was posited to be mainly expressed mainly in immune cells and peripheral areas, its identification in the brain under pathological conditions, such as Alzheimer’s disease, multiple sclerosis and amyotrophic lateral sclerosis spinal cord [183–185], led to a number of studies aimed at the assessment of its potential role in brain function and behavioral regulation. Some of these investigations indicated that the suppression of CB₂ receptor in the brain, through intracerebroventricular injection of antisense nucleotide sequences, elicited anxiolytic effects in rodents [186]. In contrast, Garcia-Gutierrez and Manzanares [187] recently described that the overexpression of CB₂ receptors reduced anxiogenic-related behaviors in the light-dark box and elevated plus maze. These premises point to the possibility that CB₂ receptor ligands may also play a role in the modulation of anxiety disorders. This hypothesis, however, awaits further examination with proper pharmacological tools.

CBD

Several studies suggest that THC and CBD may exert opposite actions on brain function and psychopathology [188], possibly in relation to the action of CBD as a potent CB₁ receptor antagonist/inverse agonist [21] (see above). Several lines of preclinical work have shown that CBD reduces the effects of THC on several behavioral functions [189–191]. In line with these data, CBD has been found to reduce the anxiety and improve the sensation of well being induced by an acute, high THC dose in healthy volunteers [192].

In contrast with these data, a number of studies have shown that CBD pretreatment potentiated the behavioral effects induced by THC [193–195]. These actions may signify the ability of CBD to inhibit cytochrome P450-mediated drug metabolism [196,197], which may increase THC blood and brain concentrations [193,195].

Notably, the behavioral outcomes of CBD do not appear to be only due to potential pharmacodynamic/pharmacokynetic competition with THC; indeed, recent studies have shown that CBD exerts inherent anxiolytic effects, both in rodent models [157,198–201] and, more recently, in patients affected by social phobia [202,203]. The anxiolytic action of CBD may be linked to 5-HT_1Areceptor, but not through benzodiazepine receptors [204]. Of note, the anxiolytic action of CBD also appears to be bidirectional, as only low to moderate doses, but not high doses, have been associated with exert anxiolytic effects [200,205].

The anxiolytic action of CBD do not appear to be mediated by benzodiazepine receptors [204], but rather by 5-HT_1A serotonin receptors in the bed nucleus of the stria terminalis [206], a critical component of the amygdaloid complex involved in the regulation of stress response.

Accordingly, CBD has been shown to reduce amygdalar responses to fearful stimuli [207]; this mechanism may be essential for the anxiolytic effects of this compound in social phobia [203]. Furthermore, CBD has been shown to elicit antipanic effects through the activation of 5-HT_1Areceptors in the dorsal periaqueductal gray, a critical area for the modulation of emotional reactivity to stress [208,209].

Endocannabinoid transport blockers

The systemic administration of the endocannabinoid transport blocker AM404 (Fig. 4) was shown to elicit anxiolytic-like behaviors in the elevated plus maze and defensive withdrawal in adult rats, as well as an attenuation of ultrasonic vocalizations in rat pups [175]. The same compound was shown to attenuate marble burying (a paradigm for compulsivity testing) in mice, suggesting that this compound may have some potential efficacy for OCD [206]. Interestingly, the anxiolytic effects of AM404 were shown to be contributed by both CB₁ and 5-HT_1A receptors [152,210], in a fashion similar to the potent CB₁ receptor agonist CP 55,940 [160]. Additionally, AM404 has been suggested to act as a FAAH inhibitor [211], although evidence in this respect is controversial [72]. Indeed, despite the identification of potential candidate endocannabinoid binding sites [212], no final evidence is currently available on the existence and/or molecular identity of the endocannabinoid transporter.

Although the possibility of targeting the endocannabinoid carrier for the development of anxiolytic compounds is appealing and has been targeted by a patent proposing these compounds as a pharmacological support for psychotherapy [213], the elusive molecular identity of the transporter itself has greatly limited the studies. Furthermore, preliminary data indicate that AM404 elicits reward in animals and is self-administered by squirrel monkeys [175,214], raising the possibility that endocannabinoid transport blockers may be addictive.

FAAH inhibitors

The prototypical FAAH inhibitor URB597 (Fig. 4) has been shown to reduce anxiety-like behaviors in rats, in a rimonabant-sensitive fashion [155,163,215–217]. In addition to its anxiolytic-like properties, URB597 was found to exert also antidepressant-like effects in several animal models with high face and predictive validity, such as the forced swim, tail suspension and chronic mild stress paradigms [89,210,216,218]. The anxiolytic action of FAAH inhibitors has been suggested to depend on the enhancement of anandamide in the dorsolateral periaqueductal gray [219]; interestingly, however, only low doses of URB597 in the prefrontal cortex were found to elicit anxiolytic-like effects, through CB₁ receptor activation. However, higher doses ceased to elicit anxiolysis, in view of their interaction with TPRV1 vanilloid receptors [220]. Furthermore, the anxiolytic and antidepressant actions of FAAH inhibitors were observed only under conditions of high environmental aversiveness, but not under normal conditions [163,218,221]. Importantly, the psychotropic effects of FAAH inhibitors are partially distinct from those associated with cannabinoids, in that they appear to fail to reproduce the hedonic and interoceptive states produced by CB receptor agonists [89] and to induce self-administration in squirrel monkeys [222]. Taken together, these data suggest that FAAH inhibitors may be promising tools in the therapy of anxiety and mood disorders with a safer profile than cannabinoid direct agonists. This idea has been recently endorsed by several authors in recent articles and patents, featuring novel categories of highly selective and potent FAAH inhibitors [223–225] [226]. However, it should be noted that recent data have recently shown that URB597 induce a number of side effects in rats, including social withdrawal, working memory deficits [227] and impairments in auditory discrimination and reversal of olfactory discrimination [228].

MAGL inhibitors

The role of 2-AG in emotional regulation has been difficult to ascertain until the recent development of highly selective monoacylglycerol lipase (MAGL) inhibitors [35,223]. Several lines of evidence have suggested that 2-AG plays a pivotal role in the pathophysiology of anxiety and defensive behaviors. The prototypical MAGL inhibitor, JZL184 (Fig. 4), has been shown to enhance the levels of 2-AG, but not anandamide; this effect is due to its extremely high selectivity for MAGL over FAAH and other brain serine hydrolases. Recent evidence has shown that this compound exerts anxiolytic-like effects in the elevated plus maze and in marble buyring, at doses that do not affect locomotor activity [93,229,230]. Similarly to the effects described for FAAH inhibitors (see above), the anxiolytic effects of this compound were observed in highly aversive (or anxiogenic) contextual settings [229]. The neurobiological role of 2-AG in anxiety is still poorly understood, although several studies have shown that environmental stressors alter its biosynthesis and degradation in key brain structures controlling emotional regulation, including periaqueductal grey, amygdala and hippocampus [231,232]. Interestingly, recent evidence has shown that the anxiolytic properties of JZL184 appear to be mediated by CB₂, rather than CB₁ receptors [93], pointing to a potential implication of this receptor in the role of 2-AG in anxiety regulation.

The present work was supported by the National Institute of Health grant R21HD070611 and the USC Zumberge Individual Research Grant (to MB).