As the COMT enzyme is primarily related to
As the COMT enzyme is primarily related to prefrontal dopamine degradation (e.g., Yavich et al., 2007), the present study supports the notion that prefrontal dopamine modulates retention of fear extinction rather than within-session extinction (Abraham et al., 2014). Consistent with other studies we found no COMT Val158Met effects for Day 1 extinction (Gruss et al., 2016, Norrholm et al., 2013, Raczka et al., 2011) but instead for Day 2 recall. One previous study in humans found stronger fear responses in Met homozygotes (Lonsdorf et al., 2009) during extinction 24 h after initial fear conditioning which was later interpreted as increased long-term fear retention (Lonsdorf & Kalisch, 2011). In line with the present result pattern, this supports a role of COMT Val158Met for memory retention rather than fear responding per se. As our results suggest generally more stable fear retention in Val/Val carriers rather than Met/Met carriers, follow-up studies may inform about possible determinants of the relationship between COMT Val158Met and long-term fear retention inherent to the experimental design. One advantage of the present study is the use of a fully crossed (i.e., 2 × 2) differential conditioning and differential extinction design, which allows to disentangle COMT Val158Met effects on long-term fear retention and long-term extinction retention. Taken together, the present results indicate that Val homozygotes show better fear and extinction retention than Met carriers in the LPP and HP, at least at a group level when individual differences in Day 1 extinction success are discarded. Although retention effects might stem from differences in both consolidation and/or recall of extinction memory, prefrontal dopamine – and thereby COMT enzyme angiotensin ii receptor blockers – might be particularly relevant for consolidation as indicated by studies on endogenous prefrontal dopamine release in rats after extinction training (Hugues, Garcia, & Léna, 2007) and the finding that L-DOPA administration after extinction training improves long-term fear extinction in humans (Haaker et al., 2013, Haaker et al., 2015). Future studies using COMT inhibitors could test directly to which extent COMT Val158Met modulates extinction consolidation and recall, respectively (Farrell et al., 2012, Giakoumaki et al., 2008). We suggest that COMT Val158Met influences long-term fear extinction via prefrontal dopamine, with relatively lower dopamine levels (i.e., Val/Val) promoting generally better consolidation across all subjects, independent of individual differences in Day 1 extinction success. Nevertheless, the underlying mechanisms might go beyond a linear and direct effect of prefrontal dopamine levels on long-term fear extinction success. First, relationships between prefrontal functioning and extracellular dopamine levels seem to be inversely U-shaped (Durstewitz and Seamans, 2008, Farrell et al., 2012, Mueller et al., 2014a, Mueller et al., 2011). Beyond the PFC, the relevant neural network likely extends to the ventral striatum which is involved in fear extinction via prediction error coding (Abraham et al., 2014). Moreover, concurrent dopaminergic activation of striatum and amygdala may be necessary for long-term learning (LaLumiere & Nawar, 2005). COMT Val158Met affects phasic dopaminergic firing in the striatum in a more subtle way than in the PFC (Bilder et al., 2004) and likely interacts with other dopaminergic genotypes (Felten et al., 2011, Raczka et al., 2011). Interestingly, the agency facet of extraversion, which has been consistently related to individual differences in dopamine in general and to COMT Val158Met in specific (Depue and Collins, 1999, Mueller et al., 2014b), was unrelated to both COMT Val158Met and Day 2 extinction recall in the present study. Although variations in agentic extraversion have been linked to differences in frontostriatal dopaminergic activity, different psychological processes (e.g., incentive motivation in agentic extraverts vs. extinction retention processes) may be mediated via different neuron populations or distinct frontostriatal connections (Bromberg-Martin et al., 2010, Salamone and Correa, 2012). Future studies might combine neuroimaging with pharmacological challenges and assess more genotypes in order to disentangle these complex mechanisms.