Parsing Heterogeneity: Additive Effects of Oxytocin Receptor Gene Polymorphisms on Reward Circuitry in ASD

Background: Autism spectrum disorders (ASD) are genetically complex. Single nucleotide polymorphisms (SNPs), which individually confer a small risk for neurodevelopmental disorders, may act additively to significantly increase risk for ASD (Gaugler 2014). A recent meta-analysis of SNPs on the oxytocin receptor gene (OXTR) suggests that multiple loci on OXTR increase the likelihood of developing ASD (LoParo 2014). Animal models have suggested that OXTR risk loci impact biological circuits relevant to ASD symptomatology; presynaptic oxytocin receptors in the nucleus accumbens (NAcc) are required for mice to exhibit a socially conditioned place preference (Dolen 2013). Importantly, neuroimaging research in ASD suggests that the NAcc is hypoactivated during reward processing compared to typically-developing (TD) controls (Dichter 2010; 2011). Together, these studies suggest a link between genetic and biological correlates of the oxytocin receptor, atypical social behavior, and hypoactivity in reward-related brain regions.  Here, we explore these links using resting state functional connectivity magnetic resonance imaging (rs-fcMRI).

Objectives: 1) Examine the link between variability in number of OXTR risk polymorphisms across four SNPs and rs-fcMRI of the NAcc. 2) Investigate how altered reward-system connectivity relates to measures of ASD symptomatology.

Methods: DNA was extracted from saliva samples and genotyped for four ASD-associated OXTR SNPs (rs53576, rs237887, rs2254298, rs1042778). Participants were 29 children with ASD and 30 TD children ages 9-17. Children participated in a six-minute eyes-open resting-state fMRI scan. Data were preprocessed using standard methods followed by motion scrubbing (Power 2012). To assess connectivity, rs-fcMRI activity in the bilateral NAcc (defined using the Harvard-Oxford Atlas thresholded at 25% probability) was extracted and correlated with all other brain voxels. Single-subject whole-brain rs-fcMRI maps were combined and compared at the group level, modeling the number of OXTR risk alleles as a covariate of interest. All results were thresholded at z>2.3, corrected for multiple comparisons at p<0.05.

Results: In both ASD and TD participants, the bilateral NAcc showed connectivity with frontal cortex, anterior cingulate, and subcortical regions including caudate, putamen, thalamus, and amygdala. In the TD group greater numbers of OXTR risk alleles were associated with increased NAcc connectivity with the frontal pole and paracingulate gyrus; in contrast, in the ASD group risk status was not associated with increased connectivity. Furthermore, there were no regions in the TD group for which greater numbers of risk alleles were associated with decreased NAcc connectivity, whereas in the ASD group greater aggregate risk was associated with decreased connectivity with the insula, anterior cingulate, bilateral caudate, thalamus, pallidum, putamen, and amygdala. This decreased connectivity in the ASD group was associated with ASD greater symptomatology – for example, reduced connectivity between the NAcc and amygdala was correlated with more severe preoccupations and mannerisms scores on the ADI-R.

Conclusions: These findings indicate that multiple OXTR risk polymorphisms have an additive effect on intrinsic reward system connectivity in children with ASD. Furthermore, this modulatory effect of OXTR SNPs on NAcc connectivity was related to ASD behavioral deficits. This work suggests a mechanism by which to parse genetic, neural, and behavioral heterogeneity within ASD.