A common genetic variant in fatty acid amide hydrolase is linked to alterations in fear extinction neural circuitry in a racially diverse, nonclinical sample of adults

Abstract

Poor fear extinction learning and recall are linked to the development of fear-based disorders, like posttraumatic stress disorder, and are associated with aberrant activation of fear-related neural circuitry. This includes greater amygdala activation during extinction learning and lesser hippocampal and ventromedial prefrontal cortex (vmPFC) activation during recall. Emerging data indicate that genetic variation in fatty acid amide hydrolase (FAAH C385A; rs324420) is associated with increased peripheral endocannabinoid (eCB) levels and lesser threat-related amygdala reactivity. Preclinical studies link increased eCB signaling to better extinction learning and recall, thus FAAH C385A may protect against the development of trauma-related psychopathology by facilitating extinction learning. However, how this FAAH variant affects fear extinction neural circuitry remains unknown. In the present study, we used a novel, immersive-reality fear extinction paradigm paired with functional neuroimaging to assess FAAH C385A effects on fear-related neural circuitry and conditioned fear responding (US expectancy ratings, subjective units of distress, and skin conductance responding) in healthy adults from an urban area (Detroit, MI; N = 59; C/C = 35, A-carrier = 24). We found lesser amygdala activation in A-allele carriers, compared to C/C homozygotes, during early extinction recall. Likewise, we found lesser dorsal anterior cingulate cortex and greater hippocampus activation in early extinction learning in A-carriers compared to C/C homozygotes. We found no effects of FAAH C385A on vmPFC activation or behavioral fear indices. These data support and extend previous findings that FAAH genetic variation, associated with increased eCB signaling and subsequent enhanced fear extinction, may predict individual differences in successful fear learning.

Keywords: amygdala; aversive; endocannabinoids; genetic polymorphism; magnetic resonance imaging; prefrontal cortex.

FIGURE 1

Fear conditioning task. A novel Pavlovian fear conditioning paradigm within a 3D virtual environment using naturally fear-invoking stimuli. (a)Two contexts were utilized, a backyard garden scene during fear conditioning (CXT+) and a public fountain plaza scene during extinction learning and extinction recall (CXT−). (b) Three colored bags (CS) were shown during fear conditioning. Two of the CSs were paired with an aversive stimulus (US), which consisted of a virtual snake and hissing sound (CS+). The third CS was never paired with the aversive stimulus (CS−). Ten minutes later, extinction learning began, with two colored bags (CSs) shown, including one of the CS+ and the CS−. The CS+ was no longer paired with the US and served as the extinguished cue (CS+E). Participants underwent a test of extinction recall 24 hr later. During extinction recall, all three CSs were shown again in the absence of the US, including the CS+E, the CS−, and the unextinguished CS+ (CS+U)

FIGURE 2

US expectancy ratings across participants and phases. Vertical axis indicates the percentage of responses. The dashed horizontal line indicates chance and the * indicates a significant divergence from chance (33.3%, indicates divergence in response frequency of “Yes,” “I don’t know,” or “No”). Panels a–c show fear conditioning, extinction learning, and extinction recall, respectively. *p < 0.001

FIGURE 3

fMRI during fear acquisition showing a (a) main effect of stimulus in the dACC and (b) stimulus × time interaction in the right hippocampus. Across participants, (a) dACC activation was greater to the cue paired with the unconditioned stimulus (CS+) than the safety cue (CS−). (b) Right hippocampus activation was greater to the CS+ (> CS−) during early fear acquisition (> late). Results significant at p_FWE < 0.05; both images shown at p < 0.005. Boxplots show the median and upper and lower quartiles

FIGURE 4

fMRI during fear extinction learning showing a main effect of time in the (a) dACC, (b) vACC, (c) right hippocampus, and (d) left hippocampus. Across participants, (a) dACC activation was greater during early fear extinction (vs. late). In contrast, (b) a cluster within the vACC was activated more during late fear extinction (early < late). Likewise, (c) right and (d) left hippocampus activation is lesser during early fear extinction than late (early < late) across participants. Results significant at p_FWE < 0.05; all images shown at p < 0.005. Boxplots show the median and upper and lower quartiles

FIGURE 5

fMRI during extinction learning showing an interaction effect of (a) genotype × time in the right hippocampus and a (b) genotype × stimulus × time interaction in the dACC. (a) A-carriers exhibited greater early (> late) right hippocampus activation versus C/C counterparts. (b) A-carriers exhibited lesser dACC activation to CS+E (> CS−) during early extinction learning (> late). Results significant at P_FWE < 0.05; right hippocampus image shown at p < 0.005 while the dACC image shown at p < 0.05 for display purposes only. Boxplots show the median and upper and lower quartiles

FIGURE 6

fMRI during fear extinction recall showing a main effect of time in the dACC. Across participants, dACC activation was greater during early extinction recall (> late). Results significant at p_FWE < 0.05; image shown at p < 0.005. Boxplots show the median and upper and lower quartiles

FIGURE 7

fMRI during extinction recall showing FAAH C385A effect on the right amygdala activation. A-carriers displayed lesser right amygdala activity to a previously extinguished cue (CS+E > CS+U) during early extinction recall, compared to C/C counterparts. Results significant at p_FWE < 0.05; right amygdala image shown at p < 0.05 for display purposes only. Boxplots show the median and upper and lower quartiles