Genetic Interaction between COMT and Dysbindin Effects Prefrontal Cortex Function During a BOLD fMRI Working Memory Paradigm

Background: Previous studies show that the catechol-O-methyltransferase (COMT) gene and the dysbindin (DTNBP1) gene impact prefrontal cortical function. Variations in these genes have been individually associated with prefrontal cortex (PFC) blood oxygenation level dependent (BOLD) fMRI during the N-back working memory task. To investigate epistasis suggested by animal studies, we used the COMT Val158Met (rs4680) polymorphism to condition a DTNBP1 SNP (rs9296985) that showed such an effect across several case-control schizophrenia cohorts (Papaleo et al., in preparation). Methodology: We studied 401 healthy Caucasians with valid COMT rs4680 and DTNBP1 rs9296985 genotypes and high-quality BOLD fMRI data from the 2-Back task at 3T. There were no significant differences in demographic variables or 2-back performance measures (accuracy/reaction time) across genotype groups. To model COMTxDTNBP1 interaction, in SPM5, we used full-factorial ANOVAs with age and gender as covariates of no interest. Results: In healthy subjects, we found evidence for an epistatic interaction between COMT-rs4680 and DTNBP1-rs9296985. We found that in subjects with COMT-Val158 homozygous backgrounds, those with rs9296985 T/T alleles were more inefficient than those with rs9296985 T/C alleles in PFC, specifically BA 9 & 46, FWE p<0.05 (corrected for a small volume; punc<0.000). We did not find this interaction for COMT-Met158 homozygotes Conclusion: By specifically modeling an interaction based on COMT-Val vs. COMT-Met conditioning, we found that DTNBP1-rs9296985 increased PFC inefficiency only for individuals homozygous for COMT-Val158. Where other data suggest gene x gene effects, it may be important to combine larger samples with statistically-derived models to fully capture epistatic interactions.