Epigenetic and Neural Correlates of the Broad Autism Phenotype

Background: Oxytocin plays an important role in social and affiliative behaviors, leading to questions of the hormone’s involvement in autism spectrum disorders. Attempts to pinpoint single nucleotide polymorphisms (SNPs) of the oxytocin receptor gene (OXTR) involved in autism have identified potential risk alleles; however these results are inconsistent (LoParo & Waldman, 2014). The current research addresses limitations of the SNP approach by considering a novel epigenetic modification of OXTR, DNA methylation, that impacts the expression of the gene (Kusui et al., 2001) without changing its underlying structure.

Critically, OXTR methylation is variable in the population and preserved in both peripheral blood and neural tissue (Gregory, Connelly et al., 2009). Previous work by our group demonstrated that higher OXTR methylation is associated with autism (Gregory, Connelly et al., 2009) and variability in OXTR methylation is associated with neural systems supporting social perception. (Puglia, Morris & Connelly, submitted). The current work extends these findings by examining the relationship between OXTR methylation, neural correlates of social perception, and the broad autism phenotype (BAP). The BAP is considered a wider continuation of the autism spectrum that extends to the general population and consists of milder, but qualitatively similar traits as those understood to be defining features of autism (Piven, et al., 1997). Our approach combines functional neuroimaging and epigenetic methods allowing us to explore a wider range of phenotypic variability and provide a shift away from traditional, categorical distinctions when studying genetic contributions to social cognitive abilities.

Objectives: To establish relationships between BAP traits, OXTR methylation, and neural activity during social perception.

Methods: Forty-three neurotypical adults (21 males) aged 18-25 provided a blood sample for assessment of OXTR methylation, participated in a passive-viewing biological motion perception fMRI task, and completed the Autism Spectrum Quotient Questionnaire (AQ, Baron-Cohen et al., 2001). fMRI blocks consisted of point-light walkers displaying biological (BIO) or random (RAND) motion. BOLD activity for the BIO>RAND contrast was modeled with AQ score as a regressor, whole-brain cluster corrected p<0.01. Additional analysis included sex as a nuisance variable.

Results: A methylation by sex interaction approached significance, revealing a negative association between methylation and AQ score for males (b=-0.20) and a positive association for females (b=0.23, p=0.063). There was a significant positive main effect of AQ score on BIO>RAND activity within ventromedial prefrontal cortex, dorsal anterior cingulate gyrus, left middle temporal gyrus, left supramarginal gyrus, and left angular gyrus. Analysis with sex as a nuisance variable revealed a significant positive main effect of AQ score within ventromedial prefrontal cortex, frontal pole and left middle temporal gyrus.

Conclusions: Expanding upon earlier work showing a relationship between OXTR methylation and autism, we reveal that OXTR methylation is differentially associated with the occurrence of BAP traits across sexes. Furthermore, we find significant increased neural activity associated with increased BAP traits occurring in regions supporting mentalizing. The over-recruitment of theory-of-mind regions during a simple social perception task for individuals with more BAP traits may reflect a compensatory mechanism associated with increased effort to interpret the biological motion.