Oppositional COMT Val158Met effects on resting state functional connectivity in adolescents and adults

Abstract

Prefrontal dopamine levels are relatively increased in adolescence compared to adulthood. Genetic variation of COMT (COMT Val158Met) results in lower enzymatic activity and higher dopamine availability in Met carriers. Given the dramatic changes of synaptic dopamine during adolescence, it has been suggested that effects of COMT Val158Met genotypes might have oppositional effects in adolescents and adults. The present study aims to identify such oppositional COMT Val158Met effects in adolescents and adults in prefrontal brain networks at rest. Resting state functional connectivity data were collected from cross-sectional and multicenter study sites involving 106 healthy young adults (mean age 24 ± 2.6 years), gender matched to 106 randomly chosen 14-year-olds. We selected the anterior medial prefrontal cortex (amPFC) as seed due to its important role as nexus of the executive control and default mode network. We observed a significant age-dependent reversal of COMT Val158Met effects on resting state functional connectivity between amPFC and ventrolateral as well as dorsolateral prefrontal cortex, and parahippocampal gyrus. Val homozygous adults exhibited increased and adolescents decreased connectivity compared to Met homozygotes for all reported regions. Network analyses underscored the importance of the parahippocampal gyrus as mediator of observed effects. Results of this study demonstrate that adolescent and adult resting state networks are dose-dependently and diametrically affected by COMT genotypes following a hypothetical model of dopamine function that follows an inverted U-shaped curve. This study might provide cues for the understanding of disease onset or dopaminergic treatment mechanisms in major neuropsychiatric disorders such as schizophrenia and attention deficit hyperactivity disorder.

Keywords: Adolescents; Catechol-O-methyltransferase; Cognition; Dopamine; Functional neuroimaging; Magnetic resonance imaging.

Fig. 1

a Graph displays working memory performance in dependence of the positioning of COMT Val158Met genotypes along the hypothetical inverted U-shaped curve for adults and adolescents based on previous reports. b Graph displays resting state functional connectivity for COMT Val158Met genotypes (black bars mean and 95 % CI) in adolescents and adults between amPFC and peak regions (left vlPFC, left PHG, left dlPFC, right PHG), controlled for main effects. Please note the similarity between assumptions on behavioral level (a) and resting state functional connectivity data (b). c Interaction effect of COMT Val158Met × developmental stage (age) controlled for gender. Positive effects indicate a stronger coupling with the seed region for adult Val homozygotes and a weaker coupling for adolescent Val homozygotes compared to Met homozygotes. Results of the seed in the anterior medial prefrontal cortex (amPFC) are shown on the lateral view. Results shown on the medial view are the vertex-wise smallest interaction effect of four lateral seeds to illustrate the extend of the “dorsal nexus” within the DMN. d Centered peak coordinates of significant clusters in the left vlPFC, the left PHG, the left dlPFC and the right PHG (p < 0.05 corrected). Results were mapped on an averaged anatomical template with a threshold of p < 0.001 in line with the family-wise multiple comparison correction (volumetric view, z-values). amPFC anterior medial prefrontal cortex, vlPFC ventrolateral prefrontal cortex, dlPFC dorsolateral prefrontal cortex, PHG parahippocampal gyrus, DMN default mode network

Fig. 2

a Network representation visualizes marginal (simple Pearson) correlations between peak regions of significant clusters from resting state functional connectivity analyses assessing differences between COMT Val158Met effects in adolescents and adults. As straightforward network translation of the seed-based functional connectivity results, these effects are not necessarily driven by direct connections between each pair of regions. b Network representation visualizes partial correlations subserving as estimator for the “true network”. The remaining effect for most connections indicates that oppositional COMT Val158Met network findings in adolescents and adults are robust and not driven by a subset of regions. It is noteworthy, however, that the connection to the left PHG is pronounced when focusing on direct connections in this post hoc analysis (threshold p < 0.05, *p < 0.05 FDR corrected, line thickness and numbers indicate z-values of the interaction effect). This suggests a relative prominent role of the PHG in the context of this study