Convergent translational evidence of a role for anandamide in amygdala-mediated fear extinction, threat processing and stress-reactivity

Abstract

Endocannabinoids are released 'on-demand' on the basis of physiological need, and can be pharmacologically augmented by inhibiting their catabolic degradation. The endocannabinoid anandamide is degraded by the catabolic enzyme fatty acid amide hydrolase (FAAH). Anandamide is implicated in the mediation of fear behaviors, including fear extinction, suggesting that selectively elevating brain anandamide could modulate plastic changes in fear. Here we first tested this hypothesis with preclinical experiments employing a novel, potent and selective FAAH inhibitor, AM3506 (5-(4-hydroxyphenyl)pentanesulfonyl fluoride). Systemic AM3506 administration before extinction decreased fear during a retrieval test in a mouse model of impaired extinction. AM3506 had no effects on fear in the absence of extinction training, or on various non-fear-related measures. Anandamide levels in the basolateral amygdala were increased by extinction training and augmented by systemic AM3506, whereas application of AM3506 to amygdala slices promoted long-term depression of inhibitory transmission, a form of synaptic plasticity linked to extinction. Further supporting the amygdala as effect-locus, the fear-reducing effects of systemic AM3506 were blocked by intra-amygdala infusion of a CB1 receptor antagonist and were fully recapitulated by intra-amygdala infusion of AM3506. On the basis of these preclinical findings, we hypothesized that variation in the human FAAH gene would predict individual differences in amygdala threat-processing and stress-coping traits. Consistent with this, carriers of a low-expressing FAAH variant (385A allele; rs324420) exhibited quicker habituation of amygdala reactivity to threat, and had lower scores on the personality trait of stress-reactivity. Our findings show that augmenting amygdala anandamide enables extinction-driven reductions in fear in mouse and may promote stress-coping in humans.

Figure 1

The fatty acid amide hydrolase (FAAH) inhibitor AM3506 (5-(4-hydroxyphenyl)pentanesulfonyl fluoride) facilitates fear extinction. Systemic treatment with AM3506 (n = 3 per time point) produced profound and lasting brain FAAH inhibition, as measured by [³H]ethanolamine activity (a), but not monoacylglycerol lipase, as assayed via [³H]glycerol activity (b) (*P<0.05 vs 1 mg/kg/time point 0). (c) Systemic AM3506 treatment before extinction training reduced fear on a retrieval test (n = 10–11 per treatment). Systemic AM3506 did not affect fear during a retrieval test when administered before conditioning (n = 8 per treatment) (d), after extinction training (n = 12 per treatment) (e) or before fear memory reactivation (n = 8 per treatment) (f). Systemic AM3506 did not affect fear open field locomotor activity (n = 8 per treatment) (g), forced swim ‘depression-related’ behavior (n = 8 per treatment) (h) or fasting-induced feeding (n = 8 per treatment) (i). Cond = conditioning; Ext = extinction; Rect = reactivation; Ret = retrieval. *P<0.05. Data are means±s.e.m.

Figure 2

Extinction training and AM3506 (5-(4-hydroxyphenyl)pentanesulfonyl fluoride) increase anandamide levels in the basolateral amygdala. (a) Schematic of testing and tissue collection procedure. (b) Systemic AM3506 treatment and extinction training increased anandamide levels in the basolateral amygdala in fear-conditioned (CS–US) mice, as compared to non-conditioned (CS) controls (n = 10–11 per treatment). (c) Systemic AM3506 treatment and extinction training increased anandamide levels in the cortex and striatum, as compared to vehicle treated controls (n = 10–11 per treatment). (d) Low-frequency stimulation (LFS) of the external capsule was used to induce endocannabinoid-mediated long-term depression of inhibitory, GABAergic, transmission (LTDi) in the basolateral amygdala. (e) In slices treated with 2 μm AM3506, but not vehicle, there was a marked and lasting reduction in evoked inhibitory postsynaptic currents (eIPSCs) amplitude following LFS, relative to baseline (n = 7–8 per treatment). Example traces of eIPSC amplitudes are shown on the right. (f) Paired pulse ratio (PPR), measuring the peak change in eIPSC amplitude over two consecutive stimulations, was higher in AM3506-treated slices after stimulation, relative to pre-stimulation baseline. CS = conditioned stimulus; US = unconditioned stimulus. *P<0.05. Data are means±s.e.m.

Figure 3

AM3506 (5-(4-hydroxyphenyl)pentanesulfonyl fluoride) facilitates fear extinction via CB1 receptor (CB1R) in the basolateral amygdala. (a) The ability of systemic AM3506 treatment before extinction to reduce fear on a retrieval test was blocked by systemic treatment with the CB1R antagonist SR141716 (n = 9–12 per treatment). (b) The ability of pre-extinction systemic AM3506 treatment to reduce fear on a retrieval test was blocked by intra-amygdala infusion of SR141716 (n = 7 per treatment). (c) Cannula tip localization. (d) Pre-extinction infusion of AM3506 into the amygdala was sufficient to reduce fear on a retrieval test (n = 8–11 per treatment). (e) Cannula tip localization. VEH = vehicle; SR = SR141716; AM = AM3506. *P<0.05. Data are means±s.e.m.

Figure 4

The lesser-expressing human fatty acid amide hydrolase (FAAH) gene 385A variant is associated with fear-related endophenotypes and traits. (a) Both the right and left amygdala exhibited significant habituation to repeated viewing of threatening faces. (b) Individuals carrying at least one lesser-expressing FAAH 385A variant showed greater amygdala-mediated habituation to threatening stimuli in all three clusters from (a). Given the low occurrence of homozygous AA carriers (n = 3) in this sample, A/A and C/A carriers (n = 31) were combined and compared with C/C homozygotes (n = 50). Data are means±s.e.m. (c, d) AA homozygotes scored significantly lower on a personality endophenotype of Stress Reactivity, measured by the Multidimensional Personality Questionnaire. *P<0.05. Data are means±s.e.m.