In humans, contradictory findings have been reported regarding involvement of the cannabinoid system in anxiety disorders. Thus, although experienced users report a mild relaxation after marijuana smoking, anxiety is the most common unpleasant side effect of occasional cannabis use and seems a reason for discontinuation of use (Hall and Solowij, 1998). A double-blind study gave evidence of a large increase in the level of anxiety in healthy volunteers after ingestion of A9-THC (Zuardi et al., 1982). However, other double-blind placebo-controlled studies showed that chronic treatment with nabilone, a synthetic cannabinoid, decreased anxiety score (Hamilton Anxiety Scale) in anxious patients (Fabre and McLendon, 1981; Ilaria et al., 1981). Similar discrepancies have been described in animals. Indeed, cannabinoids are able to display both anxiogenic- and anxiolytic-like effects depending upon doses, animal models, specific test conditions, and strains (Onaivi et al., 1990; Onaivi et al., 1996; Rodriguez de Fonseca et al., 1996). Overall, low doses of cannabinoid agonists usually induce an anxiolytic-like effect, whereas higher doses cause the opposite response. Some results suggest that the endocannabinoid system is involved in the control of emotional behavior via CBj receptors. Neuroanatomical studies showed that this receptor is expressed at high levels in brain regions involved in the control of fear and anxiety, such as the basolateral amygdala, the anterior cingulated cortex, the prefrontal cortex, and the paraventicular nucleus (PVN) of the hypothalamus (Mailleux and Vanderhaeghen, 1992; Tsou et al., 1998). Furthermore, both exogenous and endogenous cannabinoids have been found to activate the hypothalamic-pituitary-adrenal (HPA) axis, the neuroendocrine system implicated in responses to emotional stress (Weidenfeld et al., 1994; Wenger et al., 1997). Acute intracerebroventricular (i.c.v.) administration of anandamide or A9-THC stimulates the release of adreno-corticotropin hormone (ACTH) and corticosterone, probably via a central mechanism, which involves the secretion of corticotropin releasing factor (CRF). Consistent with the fact that the CRF antagonist D-Phe CRF12-41 prevented anxiogenic-like effects of HU210, a synthetic cannabinoid agonist (Rodríguez de Fonseca et al., 1996), these results suggest that HPA activation plays an important role in the mediation of cannabinoid-induced anxiogenic patterns. However, a recent study pointed out a mechanism of rapid glucocorticoid feedback inhibition of the HPA involving a release of endocannabinoid and the activation of CB1 receptors in the PVN (Di et al., 2003). Moreover, mice lacking CB1 receptors (KO) exhibited an increased basal level of anxiety (Haller et al., 2002; Maccarrone et al., 2002; Martin et al., 2002), and the highly selective CB1 receptor antagonist rimonabant (SR141716) induced anxiety-like responses in rats subjected to the defensive withdrawal test (Navarro et al., 1997) and the elevated plus maze (Arévalo et al., 2001). These data suggest the existence of an intrinsic anxiolytic tone mediated by endogenous cannab-inoids. In agreement with this hypothesis, the blockade of anandamide hydrolysis by URB597 and URB532, two fatty acid amide hydrolase (FAAH) inhibitors, induced anxiolytic-like effects in rats subjected to the elevated zero maze and in rat pups in isolation-induced ultrasonic vocalizations. These effects were prevented by rimonabant (Kathuria et al., 2003). FAAH inhibitors may modulate anxiety-related behaviors by enhancing the tonic action of anandamide on a subset of CB1 receptors, which may normally be engaged in controlling emotions and stress.
However, some conflicting data have been reported, suggesting possible species difference in the role of endocannabinoid system and CB1 receptors in mechanisms of anxiety. Although rimonabant exerted anxiogenic-like effects in rats, as mentioned in the previous paragraph (Navarro et al., 1997; Rodríguez de Fonseca et al., 1997), as well as in DBA/2 mice (Akinshola et al., 1999), surprisingly, it reduced anxiety-like behavior in ICR and C57Bl6 mice subjected to the light-dark box and the elevated plus maze (Akinshola et al., 1999; Haller et al., 2002). Interestingly, Rodgers et al. (2003) reported that rimonabant produced some anxiolytic-like effects in maze-experienced mice but not in naive mice, emphasizing the crucial importance of the animals' experimental history in the consequences of a blockade of CB1 receptors. These results suggest that initial exposure to the plus maze fails to recruit the endocannabinoid system. In contrast, the reexposure to the maze may activate the endocannabinoid substrates, which should induce a qualitatively different basal anxiety via an inhibition of GABA release (see Schlicker and Kathmann, 2001). Consistent with a role of endocannabinoids in memory and learning processes, Marsicano et al. (2002) demonstrated that reexposure of mice to learned fear cues selectively increased endocannabinoid levels in the basolateral amygdala. Thus, during the second plus-maze exposure, fear cues acquired during the first trial would activate the endocannabinoid system, producing an elevation of basal anxiety level. Under these conditions, a blockade of CB1 receptors would result in an apparent anxiolytic profile.
Interestingly, it has been reported that four-month-old CB1 KO mice display decreased anan-damide levels in the hippocampus compared with young (one-month-old) KO or age-matched wildtype mice, an effect correlated with a mild reduction of anxiety-related behavior in the light-dark box (Maccarrone et al., 2002). This suggests that CB1 receptors are involved in age-dependent adaptive changes of endocannabinoid metabolism, which appear to correlate with waning of anxietylike behavior exhibited by young CB1 KO mice. However, consistent with the weak strength of this effect, it appears that anxiety-related behaviors may be less dependent on an activation of CB1 receptors than previously thought. Furthermore, a recent study demonstrated that rimonabant was able to reduce anxiety in the elevated plus maze in both wild-type and CB1 KO animals (Haller et al., 2002). This suggests that a novel rimonabant-sensitive non-CB1 neuronal cannabinoid receptor would be involved in anxiety mechanisms (Di Marzo et al., 2000; Breivogel et al., 2001). The existence of two different central cannabinoid receptors, having opposing influences on the expression of anxiety-like behavior, may explain the contradictory literature.
Schizophrenia is a complex psychiatric disorder characterized by symptoms such as delusions, hallucinations, deterioration of social functioning, and cognitive deficits. Although dopaminergic and glutamatergic neurotransmissions are thought to be abnormal, the neurobiological and neuro-biochemical bases of this disease are still poorly understood. Several clinical findings suggest that schizophrenia may be associated with functional anomalies in the endocannabinoid signaling system. First, the prevalence of regular or problematic cannabis use is higher in schizophrenic patients than in the general population (Bersani et al., 2002). Second, similarities between some effects of cannabinoid intoxication and some symptoms of schizophrenia, especially regarding cognitive disturbances, hallucinations, perceptual distortion, and paranoia, have been shown in many reports (Wylie et al., 1995; Hall and Solowij, 1998; Leweke, Schneider, et al., 1999; Skosnik et al., 2001; Fergusson et al., 2003). Third, heavy abuse of cannabis can be considered as a factor eliciting relapse in patients with schizophrenia and possibly a premorbid precipitant (Linszen et al., 1994; Linszen et al., 1997). Clinical data indicate that cannabis consumption might place vulnerable subjects at particular risk of developing psychotic symptoms and, perhaps, lasting psychotic disorders (Zammit et al., 2002; Degenhart, 2003). Otherwise, a recent genetic study on CBj receptor gene polymorphisms reported that certain alleles or genotypes of this gene might confer a susceptibility to the hebephrenic type of schizophrenia (Ujike et al., 2002). Finally, elevated levels of anandamide and palmitylethanolamide (PEA), another endogenous cannabinoid, have been measured in the cerebrospinal fluid (CSF) of ten schizophrenic patients compared with controls (Leweke, Giuffrida, et al., 1999).
Although an effect of antipsychotic medication cannot be excluded, it seems that such functional abnormalities in endogenous cannabinoid signaling may participate in the pathogenesis of schizophrenia. Leweke, Giuffrida, et al. (1999) considered several mechanisms that could account for this elevation of brain anandamide and PEA concentrations. An in vivo microdialysis study showed that activation of DA D2-like receptors by quinpirole increased anandamide release in rat dorsal striatum (Giuffrida et al., 1999). Thus, the high levels of anandamide found in schizophrenics' CSF might result from an overstimulation of D2-like receptors, due to the activation of DA neurotransmission in these patients. Alternatively, increased CSF anandamide levels may reflect a primary "hypercan-nabinergic" state, which may occur in schizophrenic patients. This dysregulation of endocannabinoid systems may then modify glutamate and DA neurotransmission. Indeed, evidence exists that CB1 receptor activation could reduce glutamate release in rat striatum (Gerdeman and Lovinger, 2001; Huang et al., 2001). Other findings suggested that complex interactions exist between DA and cannabinoid systems. In rats, autoradiography and in situ hybridization studies showed that CB1 receptors are highly expressed in basal ganglia, limbic structures (hippocampus, olfactory bulbs, and septum), and cerebellum (Herkenham et al., 1991; Mailleux and Vanderhaeghen, 1992).
In human brain, high densities of CB1 receptors are also found in basal ganglia, limbic system, and the cerebral cortex (Glass et al., 1997), providing an opportune anatomical substrate for functional interactions between endocannabinoid and dopaminergic systems. In mice, a recent study demonstrated that systemic administration of the synthetic cannabinoid agonists CP55940 and WIN55212-2 induced Fos expression within A10 DA neurons. This effect, probably mediated by CB1 receptors because it was prevented by rimonabant, also appeared to be dependent upon an activation of noradrenergic neurotransmission (Patel and Hillard, 2003). These results are consistent with earlier works showing that cannabinoids increased the firing rate of DA neurons in the ventral tegmental area (VTA) and the substantia nigra (SN) (French et al., 1997) and enhanced DA release in the medial prefrontal cortex (mPFC) and the nucleus accumbens (NAcc) (Chen, Paredes, Li, et al., 1990; Chen, Paredes, Lowinson, et al., 1990). Interestingly, an in vivo microdialysis study in rats indicated that rimonabant significantly increased the efflux of DA selectively in the mPFC but not in the NAcc (Tzavara et al., 2003). As a decrease in PFC function has been proposed to contribute to the physiopathology of schizophrenia (Grace, 1991; Knable and Weinberger, 1997),
CBi receptor antagonists, by enhancing mesocortical dopaminergic neurotransmission, might have some therapeutic potential in psychoses.
However, such a proposal is challenged by recent results obtained in rats subjected to the prepulse inhibition (PPI) procedure, claimed to model the sensorimotor gating deficit found in schizophrenics. Indeed, in this paradigm, the startle response to an acoustic stimulus is significantly blunted by preexposure to a priming stimulus presented a few milliseconds earlier, and this effect is disrupted by psychotomimetics such as amphetamine or apomorphine. Distinct to a variety of clinically effective antipsychotic drugs, rimonabant failed to reverse the PPI disruption induced by apomorphine, MK 801, and amphetamine (Martin et al., 2003). This result, together with the failure of CP55940 to disrupt PPI at doses producing no severe decrease of startle amplitude (Mansbach et al., 1996; Martin et al., 2003), suggests that CB! receptors exert only a modest control over DA and glutamate pathways involved in sensory motor gating mechanisms. On the other hand, Poncelet et al. (1999) demonstrated that rimonabant, like clozapine, antagonized the hyperactivity induced in habituated gerbils by drugs such as cocaine, amphetamine, and WIN55212-2, known to produce or exacerbate schizophrenic symptoms, and suggested that the endocannabinoid system might play a role in the psychostimulant effect of these drugs. Nevertheless, data from a phase Ha metatrial performed in schizophrenic patients indicated that, unlike haloperidol, rimonabant was not different from placebo on any efficacy endpoint (Arvanitis et al., 2001).
It clearly appears that if the endocannabinoid system really plays a role in the physiopathology of schizophrenia, the activation of CB, receptors does not seem to be a crucial point. Interestingly, CB! is not the sole CB receptor subtype expressed in the brain. Some studies using CBi KO mice supported the existence of G-protein-coupled non-CBi and non-CB2 cannabinoid receptors, sensitive to endocannabinoids, synthetic agonists, and rimonabant (Di Marzo et al., 2000; Breivogel et al., 2001; Fride et al., 2003). In addition, anandamide has been found to activate both CBi and vanilloid VRi receptors (for review see Di Marzo et al., 2002). However, the localizations and functions of such types of receptors are unclear, and their putative role in schizophrenia remains to be demonstrated. Although cannabis use appears to be neither a sufficient nor a necessary cause for psychosis, the current status of research on cannabis-associated psychosis (Leweke et al., 2004) is inconclusive, and further research is needed to understand the mechanisms by which cannabis is associated with psychosis.
Whereas several preclinical findings support cannabinoids' role in anxiety-related behaviors (as mentioned earlier) the implication of EPCS in the regulation of affective disorders seems more difficult to establish. In humans, the comorbidity of cannabis abuse and depression is relatively common in clinical and community populations (Angst, 1996; Agosti et al., 2002). However, the existence of a causal association between cannabis use and depression remains controversial (Degenhart et al., 2001; Chen et al., 2002). Some longitudinal studies in adults reported that cannabis abuse may increase the risk of developing depressive symptoms (Weller and Halikas, 1985; Angst, 1996; Bovasso, 2001; Degenhart, 2003). In addition, it has been found that early marijuana use in childhood or adolescence was related to major depressive disorders in adulthood (late 20s) (Green and Ritter, 2000; Brook et al., 2002; Patton et al., 2002; Fergusson et al., 2003). These later findings suggest that depression may follow cannabis abuse rather than vice versa. In contrast, other findings suggest that preexisting depressive symptoms might raise the likelihood of cannabis use through a mechanism of self-medication. Thus, anecdotal reports indicated that cannabis would induce an antidepressant effect and would be also useful in the treatment of bipolar disorders (Gruber et al., 1996; Grinspoon and Bakalar, 1998). In line with these findings, in the oncology population, A9-THC could have positive effects to enhance mood in addition to the recognized antinausea and analgesic action (Walsh et al., 2003). Finally, to the best of our knowledge, a possible role of the endocannabinoid system in the physiopathology of depression has never been suggested in humans. The preliminary findings from the laboratory of Onaivi et al., 2004 (unpublished reports) indicate that CB2 receptors can be found in the brain of naïve mice and that the expression of the CB2 receptors is enhanced in the brain of mice subjected to chronic mild stress.
Animal studies also provided contradictory results. In the mouse forced-swimming or tail-suspension tests, two procedures predictive of an antidepressant activity (Porsolt et al., 1977; Stéru et al., 1987), rimonabant and AM251, a CB1 receptor antagonist/inverse agonist, induced an antidepressant-like reduction of the time spent immobile, at doses that did not affect locomotor activity (Shearman et al., 2003; Tzavara et al., 2003). Consistent with the CBj receptors' role in such effects, the AM251-induced reduction of immobility did not occur in CB1 KO mice subjected to the forced-swimming test (Shearman et al., 2003). Moreover, in vivo microdialysis experiments also showed that rimonabant increased the efflux of norepinephrine (NE), DA, and 5-HT in the mPFC (Tzavara et al., 2003), an effect which has been proposed to play a role in the antidepressant-like effect of clinically effective drugs such as desipramine and fluoxetine (Bymaster et al., 2002). However, opposite results have been reported with CB1 KO mice; compared to their wild-type counterparts, mutant mice exhibited higher depressive-like responses in the chronic unpredictable mild stress (CMS) procedure (Martin et al., 2002). This paradigm has been proposed as a valid animal model of depression, especially of anhedonia (Willner, 1997). Defined as a loss of interest or pleasure, anhedonia is one of the core symptoms of a major depressive episode (DMS-IV-TR, 2000). CMS has been shown to cause an antidepressant-reversible reduction of consumption of sucrose solutions, which is hypothesized to reflect a decrease in the perception of the rewarding value of sucrose. In this procedure, CB1 KO mice showed an enhanced sensitivity in developing a depressive-like state. Together, these results suggest that an endogenous cannabinoid tone may contribute to the maintenance of mood, probably through a modulation of monoaminergic pathways. However, as shown previously (Chaperon and Thiébot, 1999), the perception of the motivational value of positive reinforcers seems to require the stimulation of CB1 receptors. Although baseline levels of sucrose intake were similar in CB1 KO and wild-type mice (Tzavara et al., 2003; Martin et al., 2002), it cannot be excluded that the larger reduction of sucrose consumption observed in stressed mutants reflected a lowered perception of the rewarding value of sucrose rather than an increased sensitivity to the chronic stress regimen per se. Therefore, further studies are necessary to elucidate the exact relationship between the endocan-nabinoid system and mood control.
Aggressive behavior is not a unitary phenomenon and accordingly is more complex and difficult to analyze and interpret than several other behaviors. Many diverse definitions of aggression have been proposed, particularly as the term is applied to human behavior. Only a relatively few of these definitions have been applied to animal research paradigms. Initial studies demonstrated that cannabinoids might alter aggressiveness in both humans and animals. However, the numerous work performed during the 1970s and 1980s have shown some inconsistent results. In normal animals, it seems that acute administration of cannabis or A9-THC might reduce aggressive behavior, probably due to a suppressant effect on locomotor activity and a depression of general motivation (for a review, see Abel, 1975; Frischknecht, 1984). For example, in pigeons, the rate of pecking maintained by access to a stuffed target bird that could be attacked was reduced by A9-THC (0.5 and 1 mg/kg), at doses which did not affect key pecking intermittently reinforced with food (Cherek et al., 1980). Moreover, Miczek (1978) showed that when a resident animal (mouse, rat, or squirrel monkey) is confronted by an intruder conspecific, A9-THC (0.25 to 2 mg/kg) decreased species-specific attack behavior. More recently, a study by Sulcova et al. (1998) demonstrated that singly housed timid male mice exhibited aggressive reactions toward a nonaggressive intruder following a 7-d chronic treatment with anandamide at a very small dose (0.01 mg/kg.d). A larger dose (1 mg/kg.d) produced no noticeable effect, whereas in similar conditions, anandamide (10 mg/kg.d) reduced agonistic behavior and enhanced defensive conducts in otherwise spontaneously aggressive mice, an effect perhaps linked to motor deficits induced by such a high dose. This biphasic pattern of effect seems to be a feature of cannabinoids and especially of anandamide. However, it seems that under particularly stressful experimental conditions, A9-THC can also facilitate aggressive behavior (Abel, 1975). For instance, when administered to rats already stressed by a 4-d REM sleep deprivation, cannabis extract and A9-THC provoked irritability and aggressiveness and impaired defensive-submissive behavioral pattern (Carlini, 1977). In isolated rats, food deprived for 22 h and then fed ad libitum for a 3-h period, a single injection of A9-THC (11 mg/kg) induced mouse-killing (muricidal) behavior with enhanced aggressiveness, as indicated by the dramatic increase in the number of attacks on the dead mouse until it was completely torn in pieces (Bac et al., 1998). These authors also showed that A9-THC, at doses (2, 4, and 8 mg/kg) inactive to induce muricidal behavior in control rats, became efficient in rats suffering a magnesium (Mg2+) deprivation for 6 weeks. A severe Mg2+ deficiency (50-ppm diet) induced killing behavior by itself, and A9-THC exacerbated further attacks on the dead mouse. A moderate Mg2+-deficient diet (150-ppm) alone did not produce muricidal behavior, but all the rats became mouse killers when given A9-THC, whatever the dose. These results suggest a potentiation between both treatments to elicit aggressiveness. A9-THC would act as a trigger to induce aggression in Mg2+-deficient rats and reciprocally Mg2+ deficiency would reveal the potential neurotoxicity of a low dose of A9-THC (Bac et al., 2002). Surprisingly, an increase in aggressive responses has been observed in CBj KO mice subjected to the resident-intruder test, but this effect occurred exclusively during the first interaction session (Martin et al., 2002). This suggests that an absence of CBj receptor-mediated endocannab-inoid tone may lead to emergence of aggressive reactions. However, an enhanced basal level of anxiety (as measured in the light-dark test) in CB1 KO mice can also facilitate the increase of aggressiveness during the first session of the test.
Clinical assays devoted to study the effects of oral A9-THC or smoked marijuana on human aggression have also produced inconsistent results. Using a procedure in which the subjects received intense provocation (electric shock administered by an opponent), Myerscough and Taylor (1985) showed that a low dose (0.1 mg/kg) of oral A9-THC tended to increase aggressive responses. In contrast, subjects receiving a larger dose (0.4 mg/kg) behaved in a relatively nonaggressive manner throughout the experimental session. On the other hand, using the point-subtraction aggression paradigm, Cherek and Dougherty (1995) found that smoked marijuana reduced the enhanced rate of aggressive responding induced by a shift from low to high level of provocation.
The mechanism involved in the alterations of aggressive behavior by CB receptor ligands, exclusively in certain stress-forming situations, is not yet precisely understood. A link between central 5-HT levels and aggressive behavior has been established in humans and animals (for review see Chiavegatto and Nelson, 2003). High aggression in humans is correlated with low CSF concentrations of the 5-HT metabolite 5-hydroxyindoleacetic acid (5-HIAA), suggesting a diminution of 5-HT turnover. In aggressive laboratory animals, brain 5-HT turnover is also reduced, and pharmacological manipulations of the serotoninergic system substantiate a negative correlation between 5-HT neurotransmission and aggressive behavior. Accordingly, mice lacking the 5-HT transporter (5-HTT KO) were found to be less aggressive than wild-type controls (Holmes et al., 2002).
Interestingly, the CB receptor agonist CP55940 inhibited the production of nitric oxide (NO), an effect probably mediated through CB1 receptors, because it was reversed by rimonabant (Waks-man et al., 1998). NO modulates many behavioral and neuroendocrine responses, and its synthetic enzyme NO synthase has been found in high densities within emotion-regulating brain regions (Nelson et al., 1997). Recently, the excessive aggressive and impulsive traits of neuronal NO synthase knockout mice (nNOS--) have been attributed to reductions in 5-HT turnover (Chiavegatto and Nelson, 2003). This suggests that possible modifications of the endogenous cannabinoid system would contribute, via NO-related mechanisms, to local changes in 5-HT levels leading to aggressive behavior. Furthermore, REM sleep deprivation and diet restriction have been shown to reduce 5-HT levels in the medial medullary reticular formation and in the hypothalamus and hippocampus, respectively (Hao et al., 2000; Blanco-Centurion and Salin-Pascual, 2001). Anandamide also decreased the concentration of 5-HT in the hippocampus, and rimonabant was found to increase brain 5-HT levels and turnover (Hao et al., 2000; Darmani et al., 2003; Tzavara et al., 2003). Together these data suggest that some categories of stress (diet restriction, REM sleep deprivation) combined with an alteration of the endocannabinoid control system could lead to a reduction in 5-HT transmission, which may favor the development of aggressive behavior. Unfortunately, the possible involvement of endocannabinoids in aggression has not been investigated so far, although this may represent an exciting research field. In particular, it would be very interesting to study if, whether or not in humans, a latent dysfunction of the endocannabinoid system could explain the propensity of reputedly violent individuals to react aggressively in many stressful circumstances (aggression as trait characteristic of an individual).
Was this article helpful?
Who Else Wants To Sleep From Lights Out 'Til Sunrise Without Staring At The Ceiling For Hours Leaving You Feeling Fresh And Ready To Face A New Day You know you should be asleep. You've dedicated the last three hours in the dark to trying to get some sleep. But you're wide awake.