## Neuroanatomical substrates involved in cannabinoid modulation of defensive responses

F.A. Moreira, D.C. Aguiar, L.B. Resstel, S.F. Lisboa, A.C. Campos, F.V. Gomes and F.S. Guimaraes

Journal of Psychopharmacology, 2012, 26, (1), 40–55

DOI: 10.1177/0269881111400651

### Abstract

Administration of Cannabis sativa derivatives causes anxiolytic or anxiogenic effects in humans and laboratory animals, depending on the specific compound and dosage used. In agreement with these findings, several studies in the last decade have indicated that the endocannabinoid system modulates neuronal activity in areas involved in defensive responses. The mechanisms of these effects, however, are still not clear. The present review summarizes recent data suggesting that they involve modulation of glutamate and GABA-mediated neurotransmission in brain sites such as the medial prefrontal cortex, amygdaloid complex, bed nucleus of the stria terminalis, hippocampus and dorsal periaqueductal gray. Moreover, we also discuss results indicating that, in these regions, the endocannabinoid system could be particularly engaged by highly stressful situations.

Keywords : Amygdala, anxiety, bed nucleus of the stria terminalis, cannabidiol, endocannabinoid, hippocampus, medial prefrontal cortex, periaqueductal gray

### Introduction

Fear and anxiety could be seen as emerging properties of interacting brain regions that mediate defensive responses in animals exposed to threatening stimuli (Gray and McNaughton, 2000; McNaughton and Corr, 2004; Morgane et al., 2005). These responses depend on factors such as distance (proximal versus distal), environment (escape availability) and nature (potential versus real, innate versus learned) of the stimulus (Gray and McNaughton, 2000; McNaughton and Corr, 2004; Sandford et al., 2000). They are organized by at least partially distinct and hierarchical brain systems that include a behavioural inhibition system, responsible for the suppression of behaviours that could enhance danger, and an antipredator defensive system, involved in immediate responses to threatening stimuli. The hierarchy of these systems has been confirmed by recent neuroimaging studies in humans showing that initial detection of a potential threat engages mostly forebrain areas, such as the ventromedial prefrontal cortex, anterior cingulate cortex, amygdala, hippocampus and hypothalamus. These areas seem to be
involved in early threat responses, including the assignment and control of fear (Mobbs et al., 2009). When the chance of a predatory attack is high, however, midbrain structures such as the dorsolateral periaqueductal gray (PAG) would be preferentially engaged, resulting in fast, active defensive strategies
(Mobbs et al., 2009). Additional networks could also be responsible for responses such as avoidance (Graeff, 1994; McNaughton and Corr, 2004).

Dysfunctions in these defensive brain areas have been related to pathological anxiety in humans. For example, several neuroimaging studies have shown abnormalities in the prefrontal cortex of anxiety disorder patients, with decreased neuronal activity in disorders characterized by intense fear, such as panic, post-traumatic stress disorder (PTSD) and phobias, and increased activity in disorders that involve worry and rumination such as generalized anxiety and obsessive- compulsive disorder (Milad and Rauch, 2007). In addition to prefrontal cortex changes, patients with PTSD also have a smaller hippocampus volume and increased activity in the amygdala (Quirk and Mueller, 2008). Panic patients,
on the other hand, show, when treated with panic symptomsinducing drugs, increased activation of the parahippocampal gyrus, the superior temporal lobe, the anterior cingulate, cerebellar vermis, insula, temporal poles, the hypothalamus, and PAG (Boshuisen et al., 2002; Javanmard et al., 1999).

Diverse neurotransmitters have been shown to modulate these defensive responses. As will be discussed in this review, several pieces of evidence suggest that, in addition to classical neurotransmitters such as gamma-aminobutyric acid (GABA), glutamate and serotonin, endocannabinoids (eCBs) may also play an important role in the modulation of behavioural responses to threatening stimuli. eCBs act as neurotransmitters in several brain regions.

They have been characterized after studies with D9-tetrahydrocannabinol (THC) and synthetic related substances. THC is the main constituent that accounts for the psychotomimetic effects of the herb Cannabis sativa. Despite its ancient use either as a medicine or as a drug of abuse, it was only in
the middle of the last century that the chemistry of this plant started to be elucidated (Mechoulam, 1970). In addition to THC, several other compounds, such as cannabidiol (CBD) and cannabigerol, have also been identified in this plant and referred to as phytocannabinoids (Izzo et al., 2009). The characterization of these phytocannabinoids has rendered it possible to develop synthetic compounds that could mimic typical effects of THC. This, in turn, fostered pharmacological studies on this area.
For decades it remained unclear how cannabinoids would exert their effects. However, in 1988, a binding site for these compounds was finally detected in the mouse brain (Devane et al., 1988). The identification of this cannabinoid receptor, now named CB1, led to the hypothesis that an endogenous ligand should exist in mammals. Indeed, a cannabinoid receptor agonist was isolated from the porcine brain in 1992
(Devane et al., 1992). This substance, arachidonoyl-ethanolamide, was named anandamide (AEA), after ananda, the Sanskrit word for ‘bliss’. Later on, other endogenous cannabinoid receptor agonists, or endocannabinoids, were isolated.

They are all arachidonic acid derivates, and include 2-arachidonoyl glycerol (2-AG), N-arachidonoyl dopamine, noladin ether and virodhamine (Howlett et al., 2002). A few years later a CB2 receptor was also described (Munro et al., 1993). Both CB1 and CB2 are metabotropic receptors, with the former probably being the main receptor responsible for the behavioural effects of cannabinoids.

A particular feature of the CB1 receptor is its predominant location in presynaptic terminals. Indeed, contrary to classical neurotransmitters, eCBs act in retrograde fashion, as they are produced in and released from the postsynaptic neuronal membrane, acting presynaptically to decrease neurotransmitter
release (Wilson and Nicoll, 2001). Anandamide actions terminate after an internalization process followed by hydrolysis by an enzyme called fatty acid amid hydrolase (FAAH) in the postsynaptic neuron (Cravatt et al., 1996). 2-AG, on the other hand, is degraded by the enzyme monoacylglycerol lipase (MAGL) (Dinh et al., 2002). Anandamide can also target the transient receptor potential vanilloid type-1 channel (TRPV1), an ion channel permeable to calcium that, contrary to CB1, could facilitate glutamate release (Xing and Li, 2007).

Natural or synthetic cannabinoids often yield complex responses in experimental models of anxiety. As extensively reviewed elsewhere (Moreira et al., 2009a, 2009b; Viveros et al., 2005), these drugs produce bell-shaped or biphasic dose–response curves in several animal models including the elevated plus-maze (EPM), elevated T-maze and the zero maze, the light–dark test, and the Vogel conflict test. Usually the anxiolytic-like effects occur at low doses, whereas anxiogenic-like or no effects are observed with high doses (Moreira and Lutz, 2008; Moreira et al., 2009a, 2009b; Viveros et al., 2005). Similar results have also been observed in humans (Zuardi et al., 2006). The reasons for these complex effects remain unknown. CB1 receptors and other components of the eCB system are present in brain regions associated with defensive responses, including the medial prefrontal cortex (mPFC), amygdala, hippocampus, hypothalamus and PAG (Howlett et al., 2002). Cannabinoids are able to modulate the release of
several neurotransmitters that mediate or modulate those  responses, including glutamate (Jin et al., 2007), GABA (Szabo et al., 1998), glycine (Jennings et al., 2001), serotonin (Nakazi et al., 2000) and cholecystokinin (Schlicker and Kathmann, 2001). Among them, GABA and glutamate usually
play a pivotal and opposite roles, with the former inhibiting and the latter facilitating anxiety (Moreira et al., 2009a).

The contributions of these different neurotransmitters on cannabinoid effects in specific behavioural situations are poorly understood and could depend on several variables, including animal species and brain region (Haller et al., 2007; Lafeneˆtre et al., 2007; Lutz, 2009). Studies employing intra-cerebral injections into specific structures, therefore, could help to elucidate the role of cannabinoids on anxiety. In the last few years a number of studies have employed this approach to investigate the effects of cannabinoids in brain areas related to defensive behaviours.

(…)

Neuroanatomical_substrates_involved_in_c