Neural Activity in Mirror Neuron and Reward Circuitry While Viewing Emotional Expressions Relates to Core Deficits in Autism

Background: The social deficits observed in autism spectrum disorders (ASD) have been linked to dysfunction in the mirror neuron system (MNS; see Oberman & Ramachandran, 2007, for review). For instance, we previously showed that high-functioning children with ASD showed less MNS activity than typically-developing controls while observing emotional expressions, and that the level of mirror-related activity seen in children with ASD was negatively related to symptom severity in the social domain (as measured by both ADI-R and ADOS-G; Dapretto et al., 2006).  According to the social motivation hypothesis of autism, social stimuli such as faces may not be attended to by individuals with autism as they may not find them ‘rewarding.’ Surprisingly, activity in reward-related circuits in response to social stimuli has been largely unexplored in individuals with autism (Scott at al., under review).
Objectives: In this study, we used fMRI to further characterize the relationship between core autism deficits and activity in both MNS and reward circuitry. We used parental reports on the Social Responsiveness Scale (SRS), a validated measure of clinically significant autism traits, and examined how scores on this scale would relate to neural activity in a priori regions of interest including MNS (right IFG), limbic (amygdala, insula), and reward-related  areas (ventral striatum) as well as other ‘social brain’ regions (e.g., medial prefrontal cortex).
Methods:   Twenty high-functioning children with autism (19 males; mean age: 12.35) completed an fMRI task involving passive observation of faces displaying different emotions (angry, fearful, happy, sad, and neutral).  Using an event-related design, each face was presented for two seconds according to an optimized random sequence, which included null events (fixation crosses) and temporal jittering. Multiple regression analyses (controlling for the effects of IQ) were conducted to relate neural activity during this task to SRS total scores, as well as scores on the SRS treatment subscales (Social Motivation, Social Communication, Social Cognition, Social Awareness, and Autistic Mannerisms).
Results: Analyses revealed significant negative correlations between the SRS total score and activity in the right IFG, insula, and amygdala, as well as in the ventral striatum and medial prefrontal/anterior cingular cortex. Separate regressions with the treatment subscales of the SRS showed a similar pattern of results with the strongest correlations observed for the Social Motivation subscale. Somewhat surprisingly, the Autistic Mannerisms subscale also showed strong correlations within these networks, followed by weaker correlations with the Social Communication and Social Cognition subscales. Interestingly, no significant correlations were observed  for the Social Awareness subscale which taps sensory aspects of reciprocal social behaviors.
Conclusions: These findings provide further evidence of a relationship between deficits in the MNS and severity of autism symptomatology, including aspects that have not been previously tied to MNS dysfunction such as restricted interest and stereotypical behaviors.  Furthermore, the strong correlations observed between scores on the Social Motivation subscale and activity in both MNS and reward circuitry lend support to the social motivation hypothesis of autism and suggest that the social motivation deficits characteristics of autism might be related to MNS dysfunction.