ASD Associated Promoter Variants in the CNTNAP2 Gene Modulate Gene-Expression and Language Development

Background: Autism spectrum disorders (ASD) are mostly genetically determined and marked by aberrant social and reciprocal communication, language development and behavior. Recurrent findings have suggested variants within the CNTNAP2 gene as risk factors for ASD especially associated with delayed language development. Although CNVs within the relevant gene region reported for individuals with ASD reduce CNTNAP2 expression levels, studies so far have not focused on the promoter region, i.e. the regulator for gene expression.

Objectives: In this work we investigated if CNTNAP2 promoter variants are associated with ASD. Furthermore we characterized their functional impact on the transcriptional efficiency.

Methods: Direct sequencing of the CNTNAP2 promoter region was performed on a detection sample of 247 families with ASD. Variants with a genotype frequency > 1% were genotyped using RFLP in additional 356 families. Transmission disequilibrium testing was performed by UNPHASED. Impacts on transcriptional efficiency were investigated applying a luciferase-reporter gene assay.  

Results: In our detection sample we identified three annotated variants with a genotype frequency over 1% (rs150447075, rs34712024 and rs71781329) and six undescribed variants in single subjects. A preliminary family-based association study of the three annotated SNPs showed a significant association for rs34712024 (OR=0.200; CI95=0.044-0.913; p=0.0158) and rs71781329 (OR=inf.; CI95=inf.; p=0.0177), respectively. The power was too low to replicate the finding in the additional set and we thus combined the two samples and identified a significant association for rs34712024 (OR=0.409; CI95=0.1884-0.884; p=0.0177). Luciferase assay tests showed that all three variants significantly increased transcriptional efficiency. Prediction of transcription factor binding sites suggested that rs150447075 and rs34712024 lie within the binding region of the transcriptional repressor NRSF. The trimeric insertion rs71781329 is predicted to generate an additional TF binding site for EGR1, explaining increased transcription.

Conclusions: Given the findings that i) the variants analyzed here increased the transcriptional efficiency and may thus lead to an elevated level of CNTNAP2 and ii) the OR for the minor allele of rs34712024 is smaller than 1, it can be assumed that increased CNTNAP2 expression is protective for ASD. This is supported by the finding in a CNV study that a decreased CNTNAP2 expression may contribute to the risk for ASD. In conclusion, we could show that CNTNAP2 promoter variants play a crucial role in the ASD etiology. CNTNAP2 provides an important factor to be functionally characterized to better understand ASD related biological processes.