Autism spectrum disorder: unbroken mirror neurons; rare copy number variants.

03Original abstract

The authors conducted a study of mirror neuron function in autism as a possible explanation of poorer imitation performance in subjects with ASD. Twenty subjects with ASD between the ages 11 and 26 and 20 controls who were age and full-scale IQ group-matched were studied. The ASD group included subjects with autistic disorder, Asperger disorder and PDD-NOS, diagnosed using DSM-IV criteria. The authors stated that the ADI-R confirmed the DSM-IV diagnosis. Full documentation and validation of the “in-house” translation of the ADI-R would be useful. Moreover, the figure shows two subjects with ASD with communication scores of 5 and one with a lower score, so it is not clear how the ADI-R was used for the diagnosis confirmation since only citation of the original ADI-R manuscript is provided and that version without adaptation had cut-offs for the communication domain, which were higher than 5. Baseline (watching a static central cross on the screen) mu electroencephalogram (EEG) power over the sensorimotor cortex was determined as the first condition, which was compared to: (1) watching a video of a human hand manipulating an object (hand); (2) viewing of dots on the screen that were to reflect the physical properties of the visual stimuli that corresponded to a human hand moving (dot) and (3) imitation of the hand movement that was observed (imitation). The order was counterbalanced for the four conditions. The change in mu EEG power over the sensorimotor cortex from baseline to each condition was measured on a log-transformed scale. Mu suppression of the cortex related to inhibition by mirror neurons would be indicated by more negative log ratios. Less mu suppression would be indicated by less negative log ratios. One strength was that an eye-tracker synchronized with EEG data collection verified equivalent visual attention to the stimuli between cases and controls. The authors found that mu suppression was related to condition, with the dot condition leading to significantly less mu suppression than the hand and imitation conditions. The hand and imitation conditions did not differ in mu suppression. There were no significant differences between the ASD and control groups. In addition, there was no significant group×condition interaction. This indicated intact mirror neuron function in the ASD group in this paradigm. The authors did find a correlation between lower mu suppression and more qualitative impairment in communication, as measured by the ADI-R. Given that the ASD group had a full-scale IQ mean of 110 with a standard deviation of 18.7, it is likely that all the ASD subjects were verbal. The authors emphasized the impact of heterogeneity on study of mirror neuron function. Such heterogeneity may indicate that a more communication-impaired group of ASD subjects might have mirror neuron dysfunction. However, it is also possible that the correlation is spurious, given that it was P=0.03 uncorrected for multiple comparisons and with an outlier omitted. (Details about the outlier were not provided.) The authors demonstrated poorer imitation of the hand movement by the group with ASD. This was expected and demonstrated a dissociation between poor motor imitation and mu suppression. The authors pointed out that this indicated that poorer imitation was not likely to be due to mirror neuron dysfunction and suggested caution in speculation about the role of mirror neuron dysfunction in ASD. The authors reported on copy number variants (CNVs, insertions or deletions of DNA segments) in 996 subjects with ASD and 1,287 controls, using a single nucleotide polymorphism (SNP) array with approximately 1 million SNPs. This study used the highest resolution method of CNV detection published to date in a study of ASD. Cases and controls were matched for specific European ancestry by SNP genotypes. As previously discussed, microarray methods allow the detection of smaller insertions and deletions than may be seen under a microscope. The emphasis was on copy number variants found rarely in controls. One question that was tested was whether the overall burden of CNVs was higher in cases with ASD compared to controls. There was not an increase in the number of CNVs per subject or in the estimated CNV size per subject. However, there was an increase in the number of genes impacted by CNVs in subjects with ASD compared to controls. A list was compiled of genes; (1) previously strongly implicated in intellectual disability (ID) not previously strongly implicated in ASD; (2) strongly implicated in ASD (may or may not have been previously identified as strongly implicated in ID); and (3) associated with autism, but not as strongly implicated as rare variants in (1) and (2). An example of a gene strongly implicated in ID but not ASD (before this study) would be SYNGAP1 and an example of a gene previously strongly implicated in ASD would be SHANK3. In the group with ASD compared to the control group, there were significantly more rare CNVs in genes previously implicated in either ID or ASD. This finding was strengthened if the ID and ASD groups were combined. Support was not found for an increase in rare CNVs for genes associated with ASD, but not yet strongly implicated (group 3 above). These CNV burden analyses are interesting tests of the overall relevance of rare CNVs to ASD. Additional information was provided about the specific CNVs and their function. One of the more interesting findings coming out of rare CNV studies in ASD is a convergence of findings around the synapse, moving from presynaptic neurexin, to post-synaptic neuroligin and other post-synaptic density (PSD) proteins linked in a protein complex through the protein PSD95. This study revealed two additional proteins in the previously implicated set of physically interacting proteins, including SHANK2 and DLGAP2 (DLGAP=SAPAP proteins mediate interaction between PSD95 and SHANK proteins). In addition, an analysis was conducted to determine functional groupings of the rare CNVs. Groups of functionally related gene sets included GTPase/Ras signaling and cell proliferation, among others. This extends the theme of a relatively large number of individually implicated genes that converge functionally. This study adds to the emerging literature of the relevance of rare copy number variation to autism susceptibility and contributes to the understanding of the pathophysiology of autism. Dr. Cook is a co-author of the paper by Pinto and colleagues.

Autism research : official journal of the International Society for Autism Research, 2010 · doi:10.1002/aur.149