While the gene MECP2 has classically been linked to Rett Syndrome, primarily occurring in females, it is becoming increasingly clear that it plays a role in the etiology of autism and closely related developmental learning disorders in boys. Past sequencing studies have failed to find exonic mutations in autistic boys. A more recent study has found novel 3’UTR variants that were not detected in controls. Regulation of MECP2 expression has been shown not only to occur in the promoter and 3’UTR, but in flanking enhancers and silencers. The potential exists that autism spectrum causing mutations exist in regulatory elements of MECP2.
This study attempted base-pair inclusive sequencing of the MECP2 gene to detect both common and novel variants that may be enriched in autistic males. Additionally, novel statistical methods were developed to pinpoint genetic regions that associate with clinical and brain-imaging phenotypes among individuals with autism spectrum disorder (ASD).
Next-generation, paired-end sequencing using the Illumina GA2 was employed to thoroughly catalog variation at each base-pair in the entire 163 kb MECP2 region including exons, introns, promoter, 3’UTR and distal regulatory elements, in 73 ASD males compared to 70 (pooled) typically developing males. For comparison of common single nucleotide variants, our data was combined with the Autism Genetic Resource Exchange (AGRE) and HapMap sample. For comparison of common and rare indel events, data from the 1000 Genomes Project was used. Detailed phenotypic assessment on clinical and brain imaging data was obtained. Multivariate sequence-based distance matrix regression was used to detect within-case associations between combinations of MECP2 variants and phenotypes.
The frequency of common SNPs (rs6571303: p=.011) and a common MECP2 haplotype is significantly different in cases versus controls. Additionally, 6 novel 3’UTR variants (5 SNV, 1 indel), a previously discovered exonic synonomous variant, and many novel single nucleotide variants and indels were discovered in the intronic and intergenic regions that did not exist in multiple control datasets. A sliding-window, sequence-based association method was developed to examine potential consequences of combinations of variants on multiple clinical and brain-imaging phenotypes, with significant results consistent with the prior literature on MECP2. A specific 1kb segment in the upstream, promoter region of MECP2 consistently associated with scores on clinical (expressive language as assessed in the Mullen Scales of Early Learning) and brain imaging phenotypes (total brain and putamen volumes).
Conclusions: Both common and rare variants of MECP2 appear to play a role in the etiology of some autism cases, as previously thought. While it is unclear if these variants are causal and/or modifiers of ASD phenotypes, certain variants and haplotypes associate with specific ASD subphenotypes that share similarities with Rett Syndrome, where the cause is almost exclusively exonic mutations in MECP2, which are theorized to be intolerable in males. Consistent with this hypothesis, no deleterious exonic mutations were detected. Several variants in known regulatory elements including the promoter and 3’UTR were detected that were not found in controls, and these variants associate with clinical and imaging phenotypic measures.