Dyslexia and language impairment associated genetic markers influence cortical thickness and white matter in typically developing children.

John D Eicher,Ondrej Libiger,David G Amaral,Walter E Kaufmann,Hauke Bartsch,Natacha Akshoomoff,Linda Chang,Donald J Hagler,Nicholas J Schork,Burcu F Darst,Timothy T Brown,Thomas Ernst,Cinnamon S Bloss,Terry L Jernigan,Anders M Dale,B J Casey,Tal Kenet,Elizabeth R Sowell,Jeffrey R Gruen,Brian Keating,Angela Montgomery ,Sarah Murray ,Jean A Frazier ,Stewart H Mostofsky ,David N Kennedy ,Joshua Kuperman

doi:10.1007/s11682-015-9392-6

Abstract

Dyslexia and language impairment (LI) are complex traits with substantial genetic components. We recently completed an association scan of the DYX2 locus, where we observed associations of markers in DCDC2, KIAA0319, ACOT13, and FAM65B with reading-, language-, and IQ-related traits. Additionally, the effects of reading-associated DYX3 markers were recently characterized using structural neuroimaging techniques. Here, we assessed the neuroimaging implications of associated DYX2 and DYX3 markers, using cortical volume, cortical thickness, and fractional anisotropy. To accomplish this, we examined eight DYX2 and three DYX3 markers in 332 subjects in the Pediatrics Imaging Neurocognition Genetics study. Imaging-genetic associations were examined by multiple linear regression, testing for influence of genotype on neuroimaging. Markers in DYX2 genes KIAA0319 and FAM65B were associated with cortical thickness in the left orbitofrontal region and global fractional anisotropy, respectively. KIAA0319 and ACOT13 were suggestively associated with overall fractional anisotropy and left pars opercularis cortical thickness, respectively. DYX3 markers showed suggestive associations with cortical thickness and volume measures in temporal regions. Notably, we did not replicate association of DYX3 markers with hippocampal measures. In summary, we performed a neuroimaging follow-up of reading-, language-, and IQ-associated DYX2 and DYX3 markers. DYX2 associations with cortical thickness may reflect variations in their role in neuronal migration. Furthermore, our findings complement gene expression and imaging studies implicating DYX3 markers in temporal regions. These studies offer insight into where and how DYX2 and DYX3 risk variants may influence neuroimaging traits. Future studies should further connect the pathways to risk variants associated with neuroimaging/neurocognitive outcomes.Electronic supplementary materialThe online version of this article (doi:10.1007/s11682-015-9392-6) contains supplementary material, which is available to authorized users.

Highlights

Neurocognitive and language traits are complex phenotypes with substantial environmental and genetic components.dyslexia and language impairment (LI) as well as quantitative performance in reading, language, and cognitive skills are heritable traits, with heritability estimates ranging from 45 to 84 % (DeFries et al 1987; Bishop and Hayiou-Thomas 2008; Pennington and Bishop 2009; van Soelen et al 2011)
Markers within ACOT13 and C6orf62 were associated with these traits but were in linkage disequilibrium (LD) with a previously identified risk haplotype in KIAA0319, leading us to hypothesize that the associations of ACOT13 and C6orf62 tagged variation in KIAA0319 (Eicher et al 2014; Francks et al 2004; Paracchini et al 2006)
The overall goal of this study is to examine the neuroimaging implications of DYX2 and DYX3 markers previously associated with reading, language, and/or IQ

Summary

Introduction

Neurocognitive and language traits are complex phenotypes with substantial environmental and genetic components.dyslexia ( known as reading disability or RD) and language impairment (LI) as well as quantitative performance in reading, language, and cognitive skills are heritable traits, with heritability estimates ranging from 45 to 84 % (DeFries et al 1987; Bishop and Hayiou-Thomas 2008; Pennington and Bishop 2009; van Soelen et al 2011). The DYX2 locus is the most replicated RD risk locus, with subsequent association studies identifying two well-established risk genes, DCDC2 and KIAA0319 (Meng et al 2005; Schumacher et al 2006; Harold et al 2006) Following their associations with RD, other studies have shown that genes within the DYX2 locus contribute to quantitative reading and language performance as well as other related neurocognitive and language traits including LI, overall cognition, and speech sound disorder (SSD) (Scerri et al 2011; Powers et al 2013; Eicher et al 2014; Smith et al 2005; Newbury et al 2011). Markers within ACOT13 ( known as THEM2) and C6orf were associated with these traits but were in linkage disequilibrium (LD) with a previously identified risk haplotype in KIAA0319, leading us to hypothesize that the associations of ACOT13 and C6orf tagged variation in KIAA0319 (Eicher et al 2014; Francks et al 2004; Paracchini et al 2006)

Objectives

Methods

Results

Discussion

Conclusion