The observation25 that similar changes in AMPAR GluA1 subunits occur in the amygdala with opioid addiction26 and fear conditioning,43–45 coupled with evidence of epigenetically-mediated alterations in gene expression that ensue in both processes after environmental exposures,43,46 provided the rationale for examining amygdala habituation to threat-related facial expressions (a reliable intermediate phenotype linked to psychopathology).46,47 After accounting for sex and the presence of a DSM-IV disorder, rs10799590 A allele carrier status predicted right (stand Beta = 0.147; ΔF1,308=6.93, p < .009, ΔR2=0.022; Figure 2), but not left (stand Beta = 0.031; ΔF1,308=0.302, p > .582, ΔR2<0.001), amygdala habituation. G allele homozygotes (n=102) had blunted right amygdala habituation (0.045±0.374) relative to A allele carriers (n=210; 0.164±0.371). We identified 3 genotyped SNPs (rs1369848, rs12730234, rs1965776) in moderate LD with rs10799590 that tagged SNP blocks; however, none was more strongly associated with amygdala habituation than rs10799590 (Supplementary Table 7). Follow-up analyses revealed that genotype groups did not differ in initial right amygdala responses to stimuli (stand Beta = 0.067; ΔF1,308=1.42, p > .234, ΔR2=0.004).
