Objectives: The goal of this study was to identify differentially methylated genes in brain samples from individuals with autism that may be relevant to the neuropathology of autism. To accomplish this, we performed global methylation profiling of post-mortem tissues from the frontal cortex (BA9/10) and cerebellum of 7 and 8 male individuals with autism, respectively, and compared the methylation profiles with that of age- and sex-matched controls.
Methods: Genomic DNA was extracted from the brain tissues using Qiagen’s DNeasy columns and protocols. DNA enriched for methylated regions was isolated using Epigentek’s Methylamp Methylated DNA Capture (MeDIP) Kit. Affymetrix Human Promoter 1.0R GeneChips were used to analyze differentially methylated promoter regions in the enriched DNA (normalized by input DNA). Partek GS 6.6 beta software was used to analyze the intensities across the promoter regions using the workflow for methylation analyses. The data normalization procedures included adjusting for probe sequence, RMA background correction, quantile normalization, and log(base 2) transformation. Two-way ANOVA was used to determine differences in hybridization to specific probes between the cases and age-matched controls. The MAT (model-based analysis of tiling arrays) peak-seeking algorithm was used to detect enriched regions in the respective brain regions of cases vs. control samples. Multi-experiment Viewer (MeV) software was used for additional statistical analyses. A MAT score cutoff of ≥ |4.0| with p-value ≤ 0.05 was used to focus on the most differentially methylated genes for functional and pathway analyses using Ingenuity Pathway Analysis (IPA) network prediction software.
Results: Over 4000 promoter elements representing over 2000 unique genes were found to be differentially methylated in both frontal cortex and cerebellum of individuals with autism relative to the age-matched controls. There was an overlap of 754 differentially methylated genes between the two brain regions, including a number of previously identified autism candidate genes. Pathway analysis of these overlapping genes showed significant enrichment in genes involved in axon guidance, melatonin signaling, semaphorin signaling, and synaptic long term potentiation. Application of Pavlidis template matching to the respective sets of differentially methylated genes further reduced the set of genes to 63 cortical and 96 cerebellar genes whose methylation profiles completely separated cases from controls, as demonstrated by principal components analyses. Among these genes, S-phase kinase-associated protein 2 (SKP2) was found to be differentially methylated in both the frontal cortex and cerebellum. SKP2 is thus identified as a novel autism susceptibility gene which has been shown to be essential for the proliferation and differentiation of neuronal precursor cells.
Conclusions: Our results show global changes in the brain methylome of individuals with autism relative to that of age-matched controls. The differentially methylated genes in the frontal cortex and cerebellum are involved in pathways that are known to be disrupted in autism.