In Vitro and In Vivo Functional Characterization of the ENGRAILED 2 ASD-Associated Haplotype

Background: Risk for ASD is likely due to both genetic and non-genetic environmental factors, with epigenetic regulation of genes providing a possible interface between genetic and environmental factors. Our previous research has focused on the homeobox transcription factor, ENGRAILED 2 (EN2). The common alleles (underlined) of two intronic EN2 SNPs, rs1861972 (A/G) and rs1861973 (C/T), are significantly associated with ASD (518 families, P=.00000035). Subsequent association, LD mapping, and re-sequencing identified the associated rs1861972-rs1861973 as the most suitable candidate to test for function. Objectives: The goal of these studies is to determine: i) whether the ASD-associated A-C haplotype is functional, and ii) if epigenetic differences function in concert with the A-C haplotype to affect EN2 regulation. Methods: In vitro functional differences were determined by Electrophoretic Mobility Shift Assays and transient transfections of luciferase reporters into primary mouse cerebellar neurons. Affinity purification followed by Mass Spectroscopy was used to identify the proteins binding specifically to the A-C haplotype. qRTPCR was employed to measure transgene levels in vivo and EN2 mRNA levels in cell lines and post-mortem samples. Bisulfite sequencing determined the methylation status of the EN2 promoter. Results: In vitro studies using primary neuronal cultures determined the A-C haplotype results in an ~200% increase in expression (P<.0001) due to the preferential binding of transcription factors. Additional experiments demonstrated both the A and C alleles are necessary and sufficient for enhancer function. Proteomic experiments identified a seven-protein complex that binds specifically to the A-C haplotype. To extend our analysis to an in vivo system, transgenic mice were generated for the A-C and G-T haplotypes. The A-C haplotype results in ~215% increase in mRNA levels (P<.001). Environmental factors can also affect gene expression by altering epigenetic regulation. Six significant CpG islands flank human EN2, suggesting the gene may be regulated by differential methylation. We investigated this possibility by treating human neuronal cell lines with a methylation inhibitor (AZA) and a methyl donor (SAM). AZA treatment resulted in hypomethylation of the EN2 promoter and increased expression while SAM treatment results in hypermethylation and decreased levels (P<.001). EN2 levels were then quantitated in 90 human post-mortem samples. EN2 is significantly increased in affected individuals compared to controls (P<.01). Affected individuals with an A-C/A-C or A-C/G-T genotype express EN2 at the highest levels. We then investigated whether epigenetic differences could contribute to increased EN2 levels. Initial analysis indicates the EN2 promoter is methylated in unaffected individuals, while in affected individuals the same CpG dinucleotides are unmethylated. Conclusions: These data demonstrate the A-C haplotype is functional and consistently results in increased expression. EN2 is also epigenetically regulated and hypomethylation results in increased expression. Our post-mortem data demonstrate increased expression for EN2, and suggest the rs1861972-rs1861973 haplotype and promoter methylation function in concert to regulate EN2 levels.