Brain imaging studies point to functional differences between catechol-O-methyltransferase (COMT) Val¹⁵⁸Met carriers in the prefrontal and limbic areas related to emotional processing (Rasch et al., 2010; Shehzad et al., 2012). Whereas Met¹⁵⁸ allele carriers may have higher risk of developing mood and anxiety disorders (Hosák, 2007; Kocabas et al., 2010; Witte and Flöel, 2012), individuals carrying the Val¹⁵⁸ variant are thought to display more cognitive difficulties (Swart et al., 2011). However, these assumptions are not yet fully established and reverse findings have been reported (Amstadter et al., 2012). Indeed, some studies reported that the Met¹⁵⁸ allele is associated with increased limbic responsiveness to negative stimuli (Smolka et al., 2005, 2007), whereas others reported the opposite (Kempton et al., 2009; Domschke et al., 2012) or described null-results (Drabant et al., 2006). Further, the meta-analysis of Mier et al. (2010) provided evidence for a neural prefrontal cortical substrate of the pleiotropic behavioral effects of COMT Val¹⁵⁸Met (rs4680) genetic variation. This means that a single gene has an effect on the expression of two or more phenotypic traits; in this case the COMT Val¹⁵⁸Met gene that aligns with a differential impact on cognitive and emotional function. These opposing effects were found for executive cognition paradigms (favoring Met¹⁵⁸ allele carriers) and emotional paradigms (favoring the Val¹⁵⁸ variant). To summarize, at the brain level it remains poorly understood how these Val¹⁵⁸Met single nucleotide polymorphisms (SNP) interfere with cognitive and/or emotion associated neurocircuits.

Resting state fMRI (rsfMRI)—used to measure neuronal connections between distinct regions, referred to as resting-state functional connectivity (rsFC; Biswal et al., 1995)–may have the potential to gain more insight in the genetic influence on these neuronal processes (Fox et al., 2012). More in particular, rsFC has several practical and theoretical advantages over task based fMRI, including improved signal to noise, a reduced need for participant compliance, and the avoidance of task performance confounds (Fox and Greicius, 2010). However, until now, only two brain imaging studies applied rsFC together with the COMT Val¹⁵⁸Met gene typing. Tunbridge et al. (2013) showed that Val¹⁵⁸ homozygotes displayed greater rsFC than the Met¹⁵⁸ allele carriers between an executive control network and the ventrolateral parts of the prefrontal cortex during cognitive task performance (working memory). These findings concur with the observations of Liu et al. (2010) on default network connectivity where the Val¹⁵⁸ homozygotes compared to COMT Val¹⁵⁸ and Met¹⁵⁸ heterozygous individuals also showed poor cognitive performance, possibly through differential effects on prefrontal dopamine levels. Indeed, homozygote Val¹⁵⁸ carriers display higher enzymatic activity resulting in less prefrontal dopamine, whereas for the Met¹⁵⁸ variant carriers this is the reverse (Heinz and Smolka, 2006). Further, stress-related influences may affect COMT Val¹⁵⁸ and Met¹⁵⁸ allele carriers differently (Antypa et al., 2013). Under stress, working memory performance of Met¹⁵⁸ homozygotes was significantly worse compared to Val¹⁵⁸ homozygotes (Buckert et al., 2012). Carriers of the Val¹⁵⁸ allele of COMT (rs4680) were also more likely to quit a task measuring the level of distress intolerance than those without a Val¹⁵⁸ allele (Amstadter et al., 2012). This is in line with the assumption that the Val¹⁵⁸ allele is associated with cognitive inefficiency during tasks involving cognitive control. Bishop et al. (2006) found that, during the performance of a house-matching task under emotional distraction, the Val¹⁵⁸ load correlated positively with activity in parahippocampal regions.

Importantly, dopaminergic activity within the ventromedial prefrontal cortex (vmPFC), including the subgenual anterior cingulate cortex (sgACC), has been associated with subjective levels of psychosocial stress (Lataster et al., 2011). This ACC region is part of distributed corticolimbic neurocircuits implicated in “visceromotor” functions and in modulating affect, such as sadness and ruminative thought patterns (Disner et al., 2011; Smith et al., 2011; Davey et al., 2012). The sgACC also participates in cognitive functions that create sad moods and lead to pessimism (Meyer, 2012).

Because the sgACC is a key region in modulating emotional behavior (Drevets et al., 2008), our main research objective was to examine whether in the Val¹⁵⁸Met SNP (rs4680) differently influences sgACC rsFC with areas related to emotion and/or cognitive functioning. Because age (Ferreira and Busatto, 2013) and gender (Wang et al., 2014), may confound rsFC results, and given that age and gender influences have also been reported for the COMT Val¹⁵⁸Met gene (Harrison and Tunbridge, 2008; Witte and Flöel, 2012), all female participants were selected within a narrow age range and were documented never to have suffered from neuropsychiatric illnesses. The chosen sgACC seed Montreal Neurological Institute (MNI) coordinates were based on brain anatomical coordinates provided by a neuroimaging study of emotion processing and emotion regulation in women resilient or susceptible to the depressogenic effects of early life stress (Cisler et al., 2013).

In Val¹⁵⁸ carriers we hypothesized predominantly sgACC rsFC correlations with anatomical areas related with cognitive functioning. In Met¹⁵⁸ carriers we expected sgACC rsFC associations with neurocircuits implicated in emotional processing. Of note, at this stage we cannot draw hypotheses about the direction of these correlations.

From the 61 participants, 14 were Val¹⁵⁸ carriers, 37 Val/Met¹⁵⁸ heterozygotes, and 10 were Met¹⁵⁸ homozygotes. No volunteer stated to have fallen asleep during scanning. The calculation of the Hardy-Weinberg Equilibrium for two alleles showed no deviation of this assumption (X²_{(1, N = 61)} = 2.89, p = 0.09). A one-way ANOVA did not show age differences between the three groups (F_(2,58) = 0.62, p = 0.54). Significant one-way ANOVA sgACC rsFC clusters are displayed in Figure 1 for the 6 and 12 mm diameter seed sphere separately. For an overview of all significant clusters see also Table 1.

As hypothesized, we found different sgACC rsFC patterns in Val¹⁵⁸ and Met¹⁵⁸ carriers. Compared to the Met¹⁵⁸ homozygotes, COMT Val¹⁵⁸ and Val/Met¹⁵⁸ carriers were found to exhibit stronger sgACC rsFC associations with the left parahippocampus (PH), the ventromedial parts of the inferior frontal cortex, and the nucleus accumbens (NAc) and caudate nucleus. The involvement of hippocampal areas has been demonstrated earlier in different brain imaging studies examining the effect of the COMT Val¹⁵⁸Met polymorphism on cognitive processes, usually reporting that Val¹⁵⁸ carriers perform worse on memory tasks than Met¹⁵⁸ homozygotes (Bertolino et al., 2008; Heim et al., 2013); in particular in stressful situations (Buckert et al., 2012). As mentioned before, Val¹⁵⁸ carriers recruited the parahippocampal regions while performing a house-matching task during emotional distraction, a typical attentional control related task (Bishop et al., 2006). Although we did not observe differences between homo-and heterozygote Val¹⁵⁸ carriers and we did not include stress-related tasks (other than lying in the scanner with eyes closed), our findings support the hypothesis that COMT Val¹⁵⁸ and Met¹⁵⁸ carriers engage different neurocircuits while performing cognitive loaded tasks (Dennis et al., 2010). Indeed, because the sgACC serves as a facilitator of visceral responses during emotional processing–more in particular in stress-related conditions (Mayberg et al., 1999; Baeken et al., 2010)–the observed FC between the sgACC and the (left) PH in Val¹⁵⁸ carriers could imply a stronger “emotional” participation associated with memory and recollection processes which may interfere with task achievement outcomes (Eichenbaum et al., 2007, 2012). Val¹⁵⁸ carriers also displayed a stronger sgACC rsFC with the ventromedial IFG, which extended bilaterally to the NAc (which is part of the ventral striatum), and to the caudate nucleus (12 mm diameter sgACC seed), both areas comprising mostly dopaminergic neurons. This could be in line with the neurobiological action of the COMT Val¹⁵⁸Met gene where Val¹⁵⁸ carriers display higher enzymatic activity resulting in less prefrontal dopamine (Witte and Flöel, 2012). Again, as we did not include cognitive tasks and no direct dopaminergic measurements, these assumptions should be interpreted cautiously at this point. Because this is a post-hoc data driven interpretation on FC, additional data on the dopaminergic system will be crucial to verify that a stronger sgACC rsFC in Val¹⁵⁸ carriers is related to dopamine enzymatic activity. Indeed, in spite that functional connectivity is a unique powerful tool able to increase our knowledge on human brain organization, rsFC is based on an inherently ambiguous measure reflecting constraints both from static anatomical connectivity and from poorly understood functional coupling changes that are dynamic; physiological data from other sources may be required to confirm and interpret FC findings (Buckner et al., 2013).

Post-hoc t-tests showed that compared to Val¹⁵⁸ homozygotes, Met¹⁵⁸ and Val/Met¹⁵⁸ carriers displayed stronger sgACC rsFC with the MFG, more in particular the (left) orbitofrontal cortex (OFC: Brodmann area (BA) 11). Again, no differences between Met¹⁵⁸ homo- and heterozygote allele carriers were observed. Other researchers also found OFC involvement in Met¹⁵⁸ carriers related to emotional or arousal processing (Drabant et al., 2006; Montag et al., 2008; Dreher et al., 2009). Here, in contrast to the more lateral parts of the OFC, the medial OFC regions showed stronger connectivity with the default mode network, the limbic system, and areas related to autonomic processes. For a recent overview see Zald et al. (2014). Interestingly, it has recently been hypothesized by Falk et al. (2012) that Met¹⁵⁸ allele carriers may be more sensitive to cues that signal social reward or punishment and that carrying the Met¹⁵⁸ allele may be associated with greater neural activity to reward-related stimuli (Lancaster et al., 2012). This might predispose them to seek approval and thus exhibit more social conformity (Deuker et al., 2013). In line with these findings, a recent meta-analysis showed that medial OFC/vmPFC may subserve different roles in processing of reward magnitude (Diekhof et al., 2012). These dopaminergic neurons coming from the ventral tegmental area (VTA) are crucial for the recognition of rewards and their consumption or lack of consumption (Russo and Nestler, 2013). Although reverse findings have been reported (Hopkins et al., 2013), carrying the Met¹⁵⁸ allele is associated with increased activity in limbic areas and prefrontal cortex. Of note, Met¹⁵⁸homozygotes also exhibit greater anxiety (Montag et al., 2008; Baumann et al., 2013).

However, as our sample is relatively small, the interpretation of these results should be done cautiously. Due to the nature of this study, we can only draw conclusions regarding right-handed healthy young women. Although the selection of psychopathology-free female subjects within a narrow age-range can be considered as a major advantage of the study, including only healthy women within a certain age range does mean that we cannot generalize our findings to men, older women or subjects with any form of psychiatric illness. Because we did not a priori select our participants based on their genetic COMT Val¹⁵⁸Met profile, the three groups were unbalanced, which might have influenced our imaging results. To some extent this may explain the lack of sgACC rsFC differences between COMT Val¹⁵⁸Met homo- and heterozygotes. Another limitation of our study is the fact that no cardiac and respiratory data were collected during rsfMRI. Besides head movement, cardiac pulsation during scanning results in brain tissue movement and inflow effects leading to correlated signal fluctuations. Further, chest movement during breathing results in magnetic field changes that distort the MR image acquired of the brain (For an overview see Birn, 2012). Further, our hypothesis on sgACC involvement in the COMT Val¹⁵⁸Met polymorphism was determined a priori. By not examining other dedicated seeds important information may have been missed. However, our main research objective was to examine the COMT Val¹⁵⁸ Met gene in relation a specific node documented in women to be involved in emotion (dys)regulation in brain networks (Cisler et al., 2013), which makes the choice of the sgACC seed particularly relevant. Further, this seed based correlation analysis technique has been successfully applied by many research groups and has been proven useful in revealing the connectivity properties of similar seed areas (Cole et al., 2010). In addition, the test-retest reliability of such seed based correlation analysis has indicated moderate to high reliability to examine resting-state networks (Shehzad et al., 2009).

To summarize, in line with our hypotheses, we found distinct sgACC rsFC differences between Val¹⁵⁸ and Met¹⁵⁸ allele carriers. In spite that we did not use emotional or cognitive tasks, our sgACC rsFC results could represent specific neural substrates for the pleiotropic behavioral effects of COMT genetic variation between Val¹⁵⁸ and Met¹⁵⁸ allele carriers. Although our results corroborate an emotional influence on the COMT Val¹⁵⁸Met polymorphism, future imaging research examining this SNP, also with dopaminergic brain measurements, may do well to include a larger number of seeds implicated in emotional and cognitive functioning.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.