SciELO - Scientific Electronic Library Online

 
vol.41 issue1Combined GSTM1 and GSTT1 null genotypes are strong risk factors for atherogenesis in a Serbian populationGenetic and phenotypic variability of iris color in Buenos Aires population author indexsubject indexarticles search
Home Pagealphabetic serial listing  

Services on Demand

Journal

Article

Indicators

Related links

Share


Genetics and Molecular Biology

Print version ISSN 1415-4757On-line version ISSN 1678-4685

Genet. Mol. Biol. vol.41 no.1 Ribeirão Preto Jan./Mar. 2018  Epub Feb 19, 2018

http://dx.doi.org/10.1590/1678-4685-gmb-2017-0165 

Human and Medical Genetics

Dyslexia risk variant rs600753 is linked with dyslexia-specific differential allelic expression of DYX1C1

Bent Müller1 

Johannes Boltze2  3 

Ivonne Czepezauer1 

Volker Hesse4  5 

LEGASCREEN Consortium

Arndt Wilcke1 

Holger Kirsten1  6  7 

1Fraunhofer Institute for Cell Therapy and Immunology, Leipzig, Germany

2Fraunhofer Research Institution for Marine Biotechnology, Department of Medical Cell Technology, Lübeck, Germany

3Institute for Medical and Marine Biotechnology, University of Lübeck, Lübeck, Germany

4German Center for Growth, Development and Health Encouragement in Childhood and Adolescence, Berlin, Germany

5Charité-University Medicine Berlin, Institute for Experimental Paediatric Endocrinolgy, Berlin

6Institute for Medical Informatics, Statistics and Epidemiology, University of Leipzig, Leipzig, Germany

7LIFE - Leipzig Research Center for Civilization Diseases, University of Leipzig, Leipzig, Germany

Abstract

An increasing number of genetic variants involved in dyslexia development were discovered during the last years, yet little is known about the molecular functional mechanisms of these SNPs. In this study we investigated whether dyslexia candidate SNPs have a direct, disease-specific effect on local expression levels of the assumed target gene by using a differential allelic expression assay. In total, 12 SNPs previously associated with dyslexia and related phenotypes were suitable for analysis. Transcripts corresponding to four SNPs were sufficiently expressed in 28 cell lines originating from controls and a family affected by dyslexia. We observed a significant effect of rs600753 on expression levels of DYX1C1 in forward and reverse sequencing approaches. The expression level of the rs600753 risk allele was increased in the respective seven cell lines from members of the dyslexia family which might be due to a disturbed transcription factor binding sites. When considering our results in the context of neuroanatomical dyslexia-specific findings, we speculate that this mechanism may be part of the pathomechanisms underlying the dyslexia-specific brain phenotype. Our results suggest that allele-specific DYX1C1 expression levels depend on genetic variants of rs600753 and contribute to dyslexia. However, these results are preliminary and need replication.

Keywords: dyslexia; SNP; eQTL; differential allelic expression

Introduction

Dyslexia is a highly heritable disorder. The genetic component contributes by up to 60% to this disorder (Schulte-Körne, 2010) and several genes are suggested to affect the development of dyslexia (see Tables S1 and S2). Variants of well-validated genes such as DCDC2 (Doublecortin Domain Containing 2), KIAA0319, ROBO1 (Roundabout Guidance Receptor 1) and DYX1C1 (Dyslexia Susceptibility 1 Candidate 1) are believed to be involved in disturbed neuronal migration and axonal guidance (Carrion-Castillo et al., 2013) as well as differences of brain phenotypes such as alterations in white matter structure (Darki et al., 2012). In contrast to relatively well-established knowledge on the gene level, information regarding the molecular mechanisms of dyslexia candidate single nucleotide polymorphisms (SNPs) is still fragmentary.

The molecular mechanism exerted by a certain SNP can be of different nature. On the one hand, SNPs might affect the structure of a gene-derived protein. For dyslexia, however, only very few nonsynonymous SNPs affecting protein structure and function are known (see Table S2 for an overview). On the other hand, SNPs might influence the protein quantitatively, e.g., by altering gene expression levels, a phenomenon also referred to as expression quantitative loci (eQTL). eQTLs are commonly differentiated in cis-, as well as in trans-acting eQTLs. Trans eQTLs are located distant to the gene which expression is affected and cis eQTLs are located close to the affected gene.

Only few studies specifically analyzed the impact of dyslexia candidate SNPs on gene expression levels. Two groups (Tapia-Páez et al., 2008; Tammimies et al., 2012) reported a putative effect of rs3743205 on expression level of DYX1C1. Tammimies et al. (2012) observed that a CpG site results from the G-allele of this SNP. This might lead to methylation of a transcription factor binding site, and, consequently, in disturbed binding of transcription factors. Paracchini et al. (2006) observed reduced expression levels of KIAA0319 in carriers of the risk haplotype rs4504469-rs2038137-rs2143340. In a second study, Dennis et al. (2009) tested seven SNPs of the KIAA0319 promotor region and observed reduced KIAA01319 expression levels for the minor allele rs9461045-T. However, these experiments were all carried out in cells derived from donors without dyslexia background. Such approaches might miss effects resulting from a disease-specific genetic background. Previous results from dyslexia, but also from other complex diseases, motivate to analyze such effects. For example, Hannula-Jouppi et al. (2005) observed disease-specific gene-expression levels in dyslexia and Furney et al. (2011) identified an Alzheimer-specific effect for a SNP within ZNF292 on entorhinal cortical volume. A disease-specific molecular mechanisms can be understood as an effect which only emerges in affected individuals. This might be due to different regulatory networks present in the affected individuals. Consequently, certain molecular factor might be active in the affected individuals, only, e.g., certain transcription factors. If a certain SNP would alter the genomic binding site of such a disease specific factor, the effect of this SNP would be also disease-specific, i.e., observed only in the affected individuals (de la Fuente, 2010).

Another problem that may affect detection of an effect of a genetic variant on gene expression might result from general heterogeneity across samples from a variety of biological and technical sources, what can decrease study power. However, these limitations can be addressed by the direct measurement of cis-regulated allelic expression differences by differential allelic expression (DAE) (Serre et al., 2008). In this approach, expression differences resulting from the two different alleles of a SNP are analyzed within heterozygous individuals. Consequently, this method is rather robust to biological or technical batch effects among individuals.

In our study we conducted a stepwise approach to identify dyslexia-specific effects of SNPs on gene expression. We started with the identification of suitable dyslexia candidate SNPs having a potential effect on local gene expression levels by assigning functional properties (McLaren et al., 2010) (Tables S1 and S2). Subsequently, all these SNPs were genotyped in 10 cell lines derived from multiple members of a family in which dyslexia frequently occurred, thus providing a disease-specific background, and in 18 control cell lines. Disease-specific DAE was assessed in two replicates applying forward and reverse Sanger sequencing of reverse transcribed cDNA. Results were compared with publically available (dyslexia-unspecific) eQTL-data.

Materials and Methods

SNP selection, cell lines and characterization steps

For the identification of dyslexia candidate SNPs, we conducted a systematic screening using ‘PubMED’ and ‘Google Scholar’ for genetic candidate-studies related to dyslexia. The identified SNPs had to map to an exonic, 5’-UTR or 3’-UTR location to have the potential to affect local expression levels of the target genes. For each of these SNPs, a minimum of four heterozygous cell lines per group was required to maintain validity of our analyses (Serre et al., 2008), and the general (non-allele specific) expression of the SNP-corresponding transcripts was tested with cDNA-specific primers in the sample cells of interest.

In total, 28 Epstein-Barr virus (EBV) immortalized cell lines derived from B-cells were available. Ten cell lines were derived from a three generational German family, in which dyslexia segregation suggests a full-penetrance, autosomal dominant inheritance. A genome-wide linkage analysis revealed a haplotype of chromosome 12 co-segregating with language impairment (Addis et al., 2010). For a detailed description of the family see Addis et al. (2010). 18 cell lines that served as controls were derived from several families with more details available elsewhere (Burkhardt et al., 2012).

Extraction of genomic DNA (DNeasy Blood & Tissue Kit, Qiagen, Hilden, Germany) and subsequent genotyping was performed by the matrix-assisted laser desorption/ionization time-of-flight spectrometry system iPLEX (Agena, Hamburg, Germany). SNPs had to fulfill Hardy-Weinberg-Equilibrium criteria (HWE; p > 0.05 after Bonferroni correction), and to exhibit a SNP-wise call rate > 97%, as well as a minor allele frequency (MAF) > 0.05.

SNPs had to be heterozygous in at least four individual cell lines to be eligible for analysis. Six SNPs fulfilled this criterion and were considered for further analyses. Appropriate cDNA-specific primers were designed and tested for blood-specific expression. Gel electrophoresis demonstrated sufficient expression of four SNPs in B-cells and were therefore analyzed for DAE. Figure 1 illustrates the workflow.

Figure 1 Workflow of SNP characterization. Candidate SNPs used in this study had to be heterozygous in at least three immortalize B-cell lines originating from the dyslexia family and three cell lines originating from controls and expressed in the immortalized B-cells. 

Heterozygous samples were quantified on cDNA and gDNA level. cDNA was reverse-transcribed from RNA using Direct-zol RNA MiniPrep (Zymo Research, Irvine, CA, USA) and Oligo(dT)15 primer (Promega, Madison, Wisconsin, USA). Exonic, cDNA-specific PCR primers, and gDNA-specific, intronic PCR primers were designed as flanking the four SNPs. PCR was carried out with 45 rounds and 58 °C annealing temperature. For further details, see Wilcke et al. (2009) and Müller et al. (2016). To quantify DAE, genomic and coding PCR products harboring the SNP of interest were purified (Qiaquick PCR Purification Kit, Qiagen, Hilden, Germany) and Sanger-sequenced by Seqlab (Göttingen, Germany). Relative peak heights of the SNP of interest were quantified using R software version 3.2.4 (R Core Team, 2016) applying the add-on package sangerseqR version 1.4.0 (Hill et al., 2014). Allelic ratios were calculated and log-transformed (see Figure 2). For each SNP, the log-transformed allelic ratio was corrected in an assay-specific manner (forward or reverse) by subtracting the respective average transformed gDNA ratio. Primer sequences can be found in Table S3.

Figure 2 Formula for calculating the log-transformed allelic ratio. The allelic ratio is the difference between the natural logarithm-transformed ratios of the allele heights of the cDNA and the gDNA. 

Statistical analyses

To identify a genetic effect on gene expression, we used the Kruskal-Wallis test to analyze significant differences between the allelic ratios of three groups: (i) affected dyslexia family, (ii) controls, and (iii) gDNA. We used the pairwise Wilcoxon rank sum test as post-hoc test, applying the closed test procedure to account for multiple testing. Results from forward and reverse sequencing were analyzed separately as well as combined by averaging allelic log-ratios.

In silico characterization and comparison with reported eQTLs

SNPs were characterized in silico according to Ensembl annotations and prediction data (McLaren et al., 2010). Furthermore, SNPs were annotated with known and predicted regulatory elements including binding sites of transcription factors and promoter regions using RegulomeDB (Boyle et al., 2012). Publication-based (dyslexia-unspecific) eQTLs comparison was performed by screening 24 published eQTL datasets (Dixon et al., 2007; Myers et al., 2007; Veyrieras et al., 2008; Heinzen et al., 2008; Ding et al., 2010; Liu et al., 2010; Murphy et al., 2010; Zeller et al., 2010; Gibbs et al., 2010; Barreiro et al., 2011; Borel et al., 2011; Fehrmann et al., 2011; Grundberg et al., 2011; Kompass and Witte, 2011; Qiu et al., 2011; Innocenti et al., 2011; Xia et al., 2012; Zou et al., 2012; Kim et al., 2012; Kabakchiev and Silverberg, 2013; Westra et al., 2013; Ramasamy et al., 2014; Kirsten et al., 2015; GTEx Consortium, 2015). These publications cover a broad range of 63 different tissues, including brain and neuronal tissues, as well as cis and trans eQTL data.

The LD structure of the DYX1C1 locus was analyzed via local association plots using LocusZoom software (Pruim et al., 2011).

Results

Functional variant annotation and identification of eligible SNPs

We identified 12 suitable SNPs with reported associations with dyslexia related phenotypes that have the potential to affect local expression levels: rs934634-CYP19A1, rs10046-CYP19A1 and rs555879-MYO5B are located in the 3’-UTR, rs3743205- DYX1C1, rs2038137-KIAA0319 and rs3178-NRSN1 are located in the 5’-UTR. Rs600753-DYX1C1, rs17819126-DYX1C1, rs9467075-DCDC2, rs3734972-FLNC and rs4504469-KIAA0319 are exonic SNPs, and rs2143340-TDP2 is located in a non-coding exon. Genotyping of these SNPs was performed in all 28 cell lines in order to identify heterozygous cell lines.

In a second step, SNPs were only considered for analyses if a minimum of three heterozygous cell lines from the dyslexia family as well as control cell lines were available. Six of the preselected SNPs fulfilled this criterion (rs10046-CYP19A1 (6 dyslexia and 11 controls), rs934634-CYP19A1 (4 dyslexia and 9 controls), rs9467075-DCDC2 (5 dyslexia and 8 controls), rs600753-DYX1C1 (7 dyslexia and 10 controls), rs555879-MYO5B (7 dyslexia and 14 controls), rs2143340-TDP2 (4 dyslexia and 7 controls)).

In a third step, sufficient expression of the transcripts corresponding to the SNPs in EBV cells was tested by cDNA-specific PCR. Four SNPs fulfilled these three criteria (rs10046-CYP19A1, rs600753-DYX1C1, rs934634-CYP19A1, rs9467075-DCDC2) and, thus, were tested for dyslexia-specific effects on gene expression (Figure 1).

Finally, sequences must have passed quality control. Hence, for rs600753, data from up to six dyslexics and seven controls, were included in differential allelic expression analysis. For detailed numbers see Table S4.

Differential allelic expression

No genetic effects on gene expression were observed for variants rs10046-CYP19A1, rs934634-CYP19A1, rs9467075-DCDC2. In contrast, we observed a significant effect of rs600753 on DYX1C1 expression levels (Table 1 and Figure 3; raw-data is shown in Figure 4). In particular, an effect of rs600753 on the forward sequencing based measurement was observed (p=0.016). The post-hoc test revealed significant differences between the cDNA levels of the dyslexia family and the controls. The significant difference could be confirmed in data from reverse sequencing (p=0.013), as well as within the combined analyses of both approaches (p=0.021). This showed that the reported dyslexia risk-allele (rs600753-C) was expressed higher than the protective allele (T) in the dyslexia family. The control cell lines revealed the opposite effect, as the T-allele was higher expressed compared to the C-allele (Table 1).

Figure 3 Differential allelic expression (DAE) of four dyslexia-related SNPs. Shown are allelic log-ratios measured in heterozygous samples adjusted for the gDNA allelic ratio. a) Forward sequencing analysis of all four expressed SNPs stratified for cDNA allelic ratios for controls and the dyslexia affected family and the logarithm of the gDNA allelic ratio. b) Reverse sequencing based replication of rs600753. c) Analysis of both sequencing approaches together. *p < 0.05 (Wilcoxon rank sum test). Global testing for genetic effects on gene expression of rs600753-DYX1C1 applying Kruskal-Wallis test was always p < 0.05. For details see Table 1

Figure 4 Sequencing results of dyslexia family and controls. The rs600753 cDNA-sequences and the respective gDNA-sequences for six dyslexia family members (a) and seven controls (b). Arrows indicate position of rs600753. 

Table 1 Association statistics of the four dyslexia-related SNPs and DAE. 

SNP Sequencing direction (allelic ratio) Mean allelic ratios (SD) Global p-value (Kruskal-Wallis test) Post-hoc p-values
gDNA cDNA controls cDNA dyslexia cDNA control vs. cDNA dyslexia cDNA dyslexia vs. gDNA cDNA controls vs. gDNA
rs10046-CYP19A1 F (C/T) 0 (0.04) -0.07 (0.28) 0.02 (0.31) 0.937
rs600753-DYX1C1 F (C/T) 0 (0.11) -0.34 (0.33) 0.42 (0.64) 0.016* 0.035* 0.2672 0.0085*
rs934634-CYP19A1 F (G/A) 0 (0.06) 0.15 (0.35) 0.02 (0.06) 0.917
rs9467075-DCDC2 F (A/G) 0 (0.08) 0.05 (0.71) -0.16 (0.65) 0.906
rs600753- DYX1C1 R (G/A) 0 (0.10) -0.19 (0.44) 0.55 (0.48) 0.013* 0.013* 0.024* 0.115
rs600753- DYX1C1 F&R 0 (0.08) -0.25 (0.37) 0.41 (0.57) 0.027* 0.021* 0.123 0.065

‘F’ denotes forward sequencing, ‘R’ denotes reverse sequencing replication and ‘F+R’ denotes analysis of both approaches together. Shown are the logarithm of the cDNA allelic ratios for controls and the dyslexia affected family members, the logarithm of the gDNA allelic ratio and the order the ratios were formed. P-values are given for the Kruskal-Wallis test comparing transformed allelic ratios of three groups (allelic ratios of cDNA from cell lines originating from the dyslexia family members, allelic ratios from cDNA from cell lines originating from controls and allelic ratios from gDNA). P-values ≤ 0.05 are indicated by an asterisk.

Since a single previous study reported sex-specific association of rs600753 with dyslexia (Dahdouh et al., 2009), we stratified our DAE analysis of rs600753 for sex. However, we did not observe any sex-specific effect. The risk and non-risk individuals revealed a similar genetic effect on gene expression for both sexes.

Functional annotation

Rs600753 was annotated with regulatory elements using RegulomeDB which includes the identification of transcription factor binding sites and their disturbance by position weight matrix (PWM). PWM indicates the disturbance of binding sites the transcription factor Srf, Nanog, Mtf1 by rs600753 (Matys et al., 2006; Badis et al., 2009; Boyle et al., 2012). Furthermore, the publication-based annotation with eQTL-effects revealed an cis effect of rs600753 on the expression levels of CCPG1 and PIGB in blood derived cells (Xia et al., 2012; Westra et al., 2013; Kirsten et al., 2015), and DYX1C1 in fibroblasts (GTEx Consortium, 2015).

Discussion

This study analyzed dyslexia candidate SNPs with regard to their disease-specific effect on the expression levels of their respective gene. This was performed by applying a DAE analysis of cells from a dyslexia family and controls. SNP rs600753 indicated an effect on the expression level of DYX1C1 with the reported risk allele rs600753-C (Dahdouh et al., 2009; Matsson et al., 2015) being stronger expressed in cell lines derived from a dyslexia family as compared to cell lines from controls.

Association of rs600753 with dyslexia related phenotypes was first reported by Dahdouh et al. (2009) who identified an association of a haplotype spanning rs3743205, rs3743204 and rs600753 in females. Variant rs600753 efficiently tags this haplotype. Corroborating, a nominally significant single-marker association of SNP rs600753 with spelling was identified in German dyslexia families (Matsson et al., 2015). Our study further supports a role of rs600753 in dyslexia as we found disease-specific effects of rs600753 on expression levels of DYX1C1.

A distinction must be drawn between disease-specific effects, such as those investigated here, and general, non-disease-specific effects. The disease-specific effect of rs600753 identified in this study can be explained by the complex genetic background underlying the disease. Affected individuals might exhibit changes in regulatory networks. This may lead, e.g., to the activation of transcription factors that are not active in unaffected controls. If a SNP causes a differential allelic expression of a binding site of a such factor, the effect of the SNP can be disease-specific. To the best of our knowledge, there is only one report of a SNP affecting gene expression levels in a dyslexia-specific manner. Hannula-Jouppi et al. (2005) observed allele-specific expression of a SNP in the 3’-UTR of ROBO1 (6483T > A) in a dyslexic Finnish family. The expression of the A-allele was absent or attenuated in four individuals. However, the same group was unable to directly replicate this finding in a more recent study (Massinen et al., 2016), but the conclusion remained that adequate ROBO1 expression is a prerequisite for a normal crossing of the auditory pathway (Lamminmäki et al., 2012; Massinen et al., 2016). We aimed to control for such variations by analyzing both strands.

Other studies investigated an effect of dyslexia candidate SNPs on gene expression levels in non-dyslexic samples, only. For instance, two studies reported effects of rs3743205 on the expression levels of DYX1C1 (Tapia-Páez et al., 2008; Tammimies et al., 2012). Similar results were reported for KIAA0319. Reduced expression levels were observed for the haplotype rs4504469-rs2038137-rs2143340 (Paracchini et al., 2006) and for rs9461045 (Dennis et al., 2009). We tested the expression of KIAA0319 with rs2038137, yet this gene was not expressed in the available cell lines. Furthermore, risk allele frequency of rs3743205-DYX1C1 was not sufficient in our sample. Consequently, from our study we cannot draw conclusions concerning genetic effects on gene expression of these SNPs.

In cell lines from the dyslexia family we observed significantly increased expression associated with the reported risk variant rs600753-C. Thereby we observed individual differences in the effect size (Figures 2), which is potentially due to a number of reasons. These include environmental factors and the presence of additional genetic factors modulating DYX1C1 expression levels. It certainly would be of interest to see whether a generally increased DYX1C1 expression level is present in the investigated dyslexia family including the non-carriers of the rs600753 risk variants. However, this comparison is not available due to the low number of available homozygous individuals.

The impact of altered gene expression levels on neuronal function was repeatedly observed for the best replicated dyslexia candidate genes (DCDC2, DYX1C1 and KIAA0319). Knockdown experiments for these genes in rats revealed disrupted neuronal migration to the neocortex (Adler et al., 2013). Particularly, neurons from Dyx1c1 knockdown rats exhibited bimodal ectopic locations by remaining at the white matter border or migrating beyond their expected position (Currier et al., 2011). Similar ectopic neuronal locations were also observed in brains of dyslexic individuals (Galaburda et al., 1985). We speculate that rs600753-DYX1C1 is part of the pathomechanism underlying the characteristic dyslexia phenotype described by Galaburda et al. (1985): Expression levels of genes being relevant for neurogenesis need to be strictly controlled, and too low as well as too high expression can be deleterious (Francesconi and Lehner, 2014). Thus, allele-specific alterations of DYX1C1 expression levels linked to rs600753 might have the potential to disturb downstream effects of DYX1C1, such as neuronal migration and neuronalplacement, and thereby affecting functionality of the resulting neural networks.

PWM-assays support this hypotheses as they indicated a disturbance of binding sites of three different transcription factors (Srf, Nanog, Mtf1) by rs600753 (Matys et al., 2006; Badis et al., 2009; Boyle et al., 2012). Altered binding of these transcription factors might provide a molecular mechanism for the observed genetic regulation. Srf (OMIM 600589) is an ubiquitous nuclear protein known to be involved in cell growth, Mtf1 (OMIM 600172) is involved in metal homeostasis and Nanog (OMIM 607937) is involved in embryonic stem cell proliferation and renewal. Hence, among these three putative affected transcription factors we consider Nanog as the most interesting candidate in the context of the molecular pathomechanism of dyslexia as its function provides a direct link to early developmental processes critical in dyslexia.

We analyzed published eQTL-data of unaffected populations to obtain further insights into the observed rs600753-DYX1C1 effect. Rs600753 directly affects the expression levels of CCPG1 and PIGB in blood derived cells (Xia et al., 2012; Westra et al., 2013; Kirsten et al., 2015), and DYX1C1 in fibroblasts (GTEx Consortium, 2015). The reported effect direction is in line with the direction we observed for the control cell lines (higher expressed T-allele). This strengthens the hypothesis that the effect of rs600753 is dyslexia-specific since we observed a significant opposite effect direction in cells from the dyslexia family (higher expressed C-allele).

However, rs600753 is not the strongest reported eQTL at this locus (Figure S1), as reported effects of rs12324434 are stronger (p=4.510-16). This variant is in moderate linkage disequilibrium (R2=0.67) with rs600753 (GTEx Consortium, 2015). Notably, two studies analyzed an association of rs12324434 with dyslexia but found no association (Bates et al., 2010; Paracchini et al., 2011). Therefore, in contrast to rs600753, a putative relevance of rs12324434 for dyslexia remains to be shown.

Limitations

We investigated DAE in immortalized B-cells and not in neuronal cell lines. However, it is well known that most cis eQTL are ubiquitous, as typically more than 50% are replicable among tissues (Van Nas et al., 2010; GTEx Consortium, 2015). All investigated affected individuals originated from a single, large dyslexia family, which limits the generalizability of our observation. Hence, our findings should be considered as preliminary and provocative and should be replicated in larger numbers of affected and unaffected individuals. Nevertheless, this family was very well characterized for any medical conditions, and dyslexia was the primary characteristic. Hence, we expect that the described genetic effects on gene expression is likely of dyslexia-specific nature. In line with this, when stratifying our data of rs600753 for affection with dyslexia, an even higher DAE was observed in cells originating from family members with reported dyslexia compared with family members without reported dyslexia (Figure S2). Moreover, all investigated cell lines originated from individuals of Caucasian ancestry. Although this eliminates an important source of false positives due to population stratification, this limits at the same time the transferability of our findings to other ethnicities.

Conclusion

We identified allele-specific DYX1C1 expression levels related to rs600753 in dyslexics, a variant previously reported to be associated with dyslexia. Our findings are among the first for dyslexia candidate SNPs suggesting an effect on gene expression in a dyslexia-specific manner. The results are in line with reported eQTL data and provide further insights into the molecular pathomechanisms of dyslexia.

Acknowledgments

We would like to thank all members of the Legascreen consortium for their support during this study. The consortium consists of: Prof. Dr. Dr. h.c. Angela D. Friederici, Prof. Dr. Frank Emmrich, Dr. Jens Brauer, Dr. Arndt Wilcke, Dr. Nicole Neef, Prof. Dr. Dr. Johannes Boltze, Dr. Michael Skeide, Dr. Holger Kirsten, Dr. Gesa Schaadt, Dr. Bent Müller, Dr. Indra Kraft, Ivonne Czepezauer, and Liane Dörr. The Legascreen Project is funded by the Fraunhofer Society and the Max-Planck-Society as a project within the framework of the “Pakt für Forschung und Innovation”. The authors are grateful to Prof. Dr. Svante Pääbo, Birgit Nickel and Prof. Dr. Frank Albert for providing material and expert support regarding cell culture experiments.

References

Addis L, Friederici AD, Kotz SA, Sabisch B, Barry J, Richter N, Ludwig AA, Rübsamen R, Albert FW, Pääbo S et al. (2010) A locus for an auditory processing deficit and language impairment in an extended pedigree maps to 12p13.31-q14.3. Genes Brain Behav 9:545-561. [ Links ]

Adler WT, Platt MP, Mehlhorn AJ, Haight JL, Currier TA, Etchegaray MA, Galaburda AM and Rosen GD (2013) Position of neocortical neurons transfected at different gestational ages with shRNA targeted against candidate dyslexia susceptibility genes. PLoS One 8:e65179. [ Links ]

Badis G, Berger MF, Philippakis AA, Talukder S, Gehrke AR, Jaeger SA, Chan ET, Metzler G, Vedenko A, Chen X et al. (2009) Diversity and complexity in DNA recognition by transcription factors. Science 324:1720. [ Links ]

Barreiro LB, Tailleux L, Pai AA, Gicquel B, Marioni JC and Gilad Y (2011) Deciphering the genetic architecture of variation in the immune response to Mycobacterium tuberculosis infection. Proc Natl Acad Sci U S A 109:1204-1209. [ Links ]

Bates TC, Lind PA, Luciano M, Montgomery GW, Martin NG and Wright MJ (2010) Dyslexia and DYX1C1: Deficits in reading and spelling associated with a missense mutation. Mol Psychiatry 15:1190-1196. [ Links ]

Borel C, Deutsch S, Letourneau A, Migliavacca E, Montgomery SB, Dimas AS, Vejnar CE, Attar H, Gagnebin M, Gehrig C et al. (2011) Identification of cis- and trans-regulatory variation modulating microRNA expression levels in human fibroblasts. Genome Res 21:68-73. [ Links ]

Boyle AP, Hong EL, Hariharan M, Cheng Y, Schaub MA, Kasowski M, Karczewski KJ, Park J, Hitz BC, Weng S et al. (2012) Annotation of functional variation in personal genomes using RegulomeDB. Genome Res 22:1790-17977. [ Links ]

Burkhardt J, Kirsten H, Wolfram G, Quente E and Ahnert P (2012) Differential allelic expression of IL13 and CSF2 genes associated with asthma. Genet Mol Biol 35:567-574. [ Links ]

Carrion-Castillo A, Franke B and Fisher SE (2013) Molecular genetics of dyslexia: An overview. Dyslexia 19:214-240. [ Links ]

Currier TA, Etchegaray MA, Haight JL, Galaburda AM and Rosen GD (2011) The effects of embryonic knockdown of the candidate dyslexia susceptibility gene homologue Dyx1c1 on the distribution of GABAergic neurons in the cerebral cortex. Neuroscience 172:535-546. [ Links ]

Dahdouh F, Anthoni H, Tapia-Páez I, Peyrard-Janvid M, Schulte-Körne G, Warnke A, Remschmidt H, Ziegler A, Kere J, Müller-Myhsok B et al. (2009) Further evidence for DYX1C1 as a susceptibility factor for dyslexia. Psychiatr Genet 19:59-63. [ Links ]

Darki F, Peyrard-Janvid M, Matsson H, Kere J and Klingberg T (2012) Three dyslexia susceptibility genes, DYX1C1, DCDC2, and KIAA0319, affect temporo-parietal white matter structure. Biol Psychiatry 72:671-676. [ Links ]

de la Fuente A (2010) From “differential expression” to “differential networking” - identification of dysfunctional regulatory networks in diseases. Trends Genet 26:326-333. [ Links ]

Dennis MY, Paracchini S, Scerri TS, Prokunina-Olsson L, Knight JC, Wade-Martins R, Coggill P, Beck S, Green ED and Monaco AP (2009) A common variant associated with dyslexia reduces expression of the KIAA0319 gene. PLoS Genet 5:e1000436. [ Links ]

Ding J, Gudjonsson JE, Liang L, Stuart PE, Li Y, Chen W, Weichenthal M, Ellinghaus E, Franke A, Cookson W et al. (2010) Gene expression in skin and lymphoblastoid cells: Refined statistical method reveals extensive overlap in cis-eQTL signals. Am J Hum Genet 87:779-789. [ Links ]

Dixon AL, Liang L, Moffatt MF, Chen W, Heath S, Wong KCC, Taylor J, Burnett E, Gut I, Farrall M et al. (2007) A genome-wide association study of global gene expression. Nat Genet 39:1202-1207. [ Links ]

Fehrmann RSN, Jansen RC, Veldink JH, Westra H-JJ, Arends D, Bonder MJ, Fu J, Deelen P, Groen HJM, Smolonska A et al. (2011) Trans-eQTLs reveal that independent genetic variants associated with a complex phenotype converge on intermediate genes, with a major role for the HLA. PLoS Genet 7:e1002197. [ Links ]

Francesconi M and Lehner B (2014) The effects of genetic variation on gene expression dynamics during development. Nature 505:208-211. [ Links ]

Furney SJ, Simmons A, Breen G, Pedroso I, Lunnon K, Proitsi P, Hodges A, Powell J, Wahlund LL-O, Mecocci P et al. (2011) Genome-wide association with MRI atrophy measures as a quantitative trait locus for Alzheimer’s disease. Mol Psychiatry 16:1130-1138. [ Links ]

Galaburda AM, Sherman GF, Rosen GD, Aboitiz F and Geschwind N (1985) Developmental dyslexia: Four consecutive patients with cortical anomalies. Ann Neurol 18:222-233. [ Links ]

Gibbs JR, van der Brug MP, Hernandez DG, Traynor BJ, Nalls MA, Lai S-L, Arepalli S, Dillman A, Rafferty IP, Troncoso J et al. (2010) Abundant quantitative trait loci exist for DNA methylation and gene expression in human brain. PLoS Genet 6:e1000952. [ Links ]

Grundberg E, Adoue V, Kwan T, Ge B, Duan QL, Lam KCL, Koka V, Kindmark A, Weiss ST, Tantisira K et al. (2011) Global analysis of the impact of environmental perturbation on cis-regulation of gene expression. PLoS Genet 7:e1001279. [ Links ]

GTEx Consortium (2015) Human genomics. The Genotype-Tissue Expression (GTEx) pilot analysis: Multitissue gene regulation in humans. Science 348:648-660. [ Links ]

Hannula-Jouppi K, Kaminen-Ahola N, Taipale M, Eklund R, Nopola-Hemmi J, Kääriäinen H and Kere J (2005) The axon guidance receptor gene ROBO1 is a candidate gene for developmental dyslexia. PLoS Genet 1:0467-0474. [ Links ]

Heinzen EL, Ge D, Cronin KD, Maia JM, Shianna K V, Gabriel WN, Welsh-Bohmer KA, Hulette CM, Denny TN and Goldstein DB (2008) Tissue-specific genetic control of splicing: Implications for the study of complex traits. PLoS Biol 6:e1000001. [ Links ]

Hill JT, Demarest BL, Bisgrove BW, Su YC, Smith M and Yost HJ (2014) Poly peak parser: Method and software for identification of unknown indels using sanger sequencin of polymerase chain reaction products. Dev Dyn 243:1632-1636. doi: 10.1002/dvdy.24183. [ Links ]

Innocenti F, Cooper GM, Stanaway IB, Gamazon ER, Smith JD, Mirkov S, Ramirez J, Liu W, Lin YS, Moloney C et al. (2011) Identification, replication, and functional fine-mapping of expression quantitative trait loci in primary human liver tissue. PLoS Genet 7:e1002078. [ Links ]

Kabakchiev B and Silverberg MS (2013) Expression quantitative trait loci analysis identifies associations between genotype and gene expression in human intestine. Gastroenterology 144:1488-1496. [ Links ]

Kim S, Cho H, Lee D and Webster MJ (2012) Association between SNPs and gene expression in multiple regions of the human brain. Transl Psychiatry 2:e113. [ Links ]

Kirsten H, Al-Hasani H, Holdt L, Gross A, Beutner F, Krohn K, Horn K, Ahnert P, Burkhardt R, Reiche K et al. (2015) Dissecting the genetics of the human transcriptome identifies novel trait-related trans-eQTLs and corroborates the regulatory relevance of non-protein coding loci. Hum Mol Genet 24:4746-4763. [ Links ]

Kompass KS and Witte JS (2011) Co-regulatory expression quantitative trait loci mapping: method and application to endometrial cancer. BMC Med Genomics 4:6. [ Links ]

Lamminmäki S, Massinen S, Nopola-Hemmi J, Kere J and Hari R (2012) Human ROBO1regulates interaural interaction in auditory pathways. J Neurosci 32:966-971. [ Links ]

Liu C, Cheng L, Badner JA, Zhang D, Craig DW, Redman M and Gershon ES (2010) Whole-genome association mapping of gene expression in the human prefrontal cortex. Mol Psychiatry 15:779-784. [ Links ]

Massinen S, Wang J, Laivuori K, Bieder A, Tapia Paez I, Jiao H and Kere J (2016) Genomic sequencing of a dyslexia susceptibility haplotype encompassing ROBO1. J Neurodev Disord 8:4. [ Links ]

Matsson H, Huss M, Persson H, Einarsdottir E, Tiraboschi E, Nopola-Hemmi J, Schumacher J, Neuhoff N, Warnke A, Lyytinen H et al. (2015) Polymorphisms in DCDC2 and S100B associate with developmental dyslexia. J Hum Genet 60:399-401. [ Links ]

Matys V, Kel-Margoulis O V, Fricke E, Liebich I, Land S, Barre-Dirrie A, Reuter I, Chekmenev D, Krull M, Hornischer K et al. (2006) TRANSFAC and its module TRANSCompel: Transcriptional gene regulation in eukaryotes. Nucleic Acids Res 34:D108-D110. [ Links ]

McLaren W, Pritchard B, Rios D, Chen Y, Flicek P and Cunningham F (2010) Deriving the consequences of genomic variants with the Ensembl API and SNP Effect Predictor. Bioinformatics 26:2069-2070. [ Links ]

Müller B, Wilcke A, Czepezauer I, Ahnert P, Boltze J, Kirsten H and LEGASCREEN consortium (2016) Association, characterisation and meta-analysis of SNPs linked to general reading ability in a German dyslexia case-control cohort. Sci Rep 6:27901. [ Links ]

Murphy A, Chu JH, Xu M, Carey VJ, Lazarus R, Liu A, Szefler SJ, Strunk R, DeMuth K, Castro M et al. (2010) Mapping of numerous disease-associated expression polymorphisms in primary peripheral blood CD4+ lymphocytes. Hum Mol Genet 19:4745-4757. [ Links ]

Myers AJ, Gibbs JR, Webster JA, Rohrer K, Zhao A, Marlowe L, Kaleem M, Leung D, Bryden L, Nath P et al. (2007) A survey of genetic human cortical gene expression. Nat Genet 39:1494-1499. [ Links ]

Paracchini S, Ang QW, Stanley FJ, Monaco AP, Pennell CE and Whitehouse AJO (2011) Analysis of dyslexia candidate genes in the Raine cohort representing the general Australian population. Genes Brain Behav 10:158-165. [ Links ]

Paracchini S, Thomas A, Castro S, Lai C, Paramasivam M, Wang Y, Keating BJ, Taylor JM, Hacking DF, Scerri T et al. (2006) The chromosome 6p22 haplotype associated with dyslexia reduces the expression of KIAA0319, a novel gene involved in neuronal migration. Hum Mol Genet 15:1659-1666. [ Links ]

Pruim RJ, Welch RP, Sanna S, Teslovich TM, Chines PS, Gliedt TP, Boehnke M, Abecasis GR, Willer CJ and Frishman D (2011) LocusZoom: Regional visualization of genome-wide association scan results. Bioinformatics 27:2336-2337. [ Links ]

Qiu W, Cho MH, Riley JH, Anderson WH, Singh D, Bakke P, Gulsvik A, Litonjua AA, Lomas D a., Crapo JD et al. (2011) Genetics of sputum gene expression in chronic obstructive pulmonary disease. PLoS One 6:e24395. [ Links ]

R Core Team (2016) R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna, Austria. [ Links ]

Ramasamy A, Trabzuni D, Guelfi S, Varghese V, Smith C, Walker R, De T, Coin L, de Silva R, Cookson MR et al. (2014) Genetic variability in the regulation of gene expression in ten regions of the human brain. Nat Neurosci 17:1418-1428. [ Links ]

Schulte-Körne G (2010) The prevention, diagnosis, and treatment of dyslexia. Dtsch Arzteblatt Int 107:718-727. [ Links ]

Serre D, Gurd S, Ge B, Sladek R, Sinnett D, Harmsen E, Bibikova M, Chudin E, Barker DL, Dickinson T et al. (2008) Differential allelic expression in the human genome: A robust approach to identify genetic and epigenetic cis-acting mechanisms regulating gene expression. PLoS Genet 4:e1000006. [ Links ]

Tammimies K, Tapia-Páez I, Rüegg J, Rosin G, Kere J, Gustafsson J-Å and Nalvarte I (2012) The rs3743205 SNP is important for the regulation of the dyslexia candidate gene DYX1C1 by estrogen receptor β and DNA methylation. Mol Endocrinol 26:619-629. [ Links ]

Tapia-Páez I, Tammimies K, Massinen S, Roy AL and Kere J (2008) The complex of TFII-I, PARP1, and SFPQ proteins regulates the DYX1C1 gene implicated in neuronal migration and dyslexia. FASEB J 22:3001-3009. [ Links ]

Van Nas A, Ingram-Drake L, Sinsheimer JS, Wang SS, Schadt EE, Drake T and Lusis AJ (2010) Expression quantitative trait loci: Replication, tissue- and sex-specificity in mice. Genetics 185:1059-1068. [ Links ]

Veyrieras J-B, Kudaravalli S, Kim SY, Dermitzakis ET, Gilad Y, Stephens M and Pritchard JK (2008) High-resolution mapping of expression-QTLs yields insight into Human Gene Regulation. PLoS Genet 4:e1000214. doi: 10.1371/journal.pgen.1000214. [ Links ]

Westra H-J, Peters MJ, Esko T, Yaghootkar H, Schurmann C, Kettunen J, Christiansen MW, Fairfax BP, Schramm K, Powell JE et al. (2013) Systematic identification of trans eQTLs as putative drivers of known disease associations. Nat Genet 45:1238-1243. [ Links ]

Wilcke A, Weissfuss J, Kirsten H, Wolfram G, Boltze J and Ahnert P (2009) The role of gene DCDC2 in German dyslexics. Ann Dyslexia 59:1-11. [ Links ]

Xia K, Shabalin AA, Huang S, Madar V, Zhou YH, Wang W, Zou F, Sun W, Sullivan PF and Wright FA (2012) SeeQTL: A searchable database for human eQTLs. Bioinformatics 28:451-452. [ Links ]

Zeller T, Wild P, Szymczak S, Rotival M, Schillert A, Castagne R, Maouche S, Germain M, Lackner K, Rossmann H et al. (2010) Genetics and beyond - the transcriptome of human monocytes and disease susceptibility. PLoS One 5:e10693. [ Links ]

Zou F, Chai HS, Younkin CS, Allen M, Crook J, Pankratz VS, Carrasquillo MM, Rowley CN, Nair AA, Middha S et al. (2012) Brain expression genome-wide association study (eGWAS) identifies human disease-associated variants. PLoS Genet 8:e1002707. [ Links ]

Associate Editor: Mara H. Hutz

Received: May 31, 2017; Accepted: August 28, 2017

Send correspondence to Bent Müller. Fraunhofer Institute for Cell Therapy and Immunology, Perlickstr. 1, 04103 Leipzig, Germany. E-mail: bent.mueller@izi.fraunhofer.de.

Creative Commons License License information: This is an open access article distributed under the terms of the Creative Commons Attribution License (type CC-BY), which permits unrestricted use, distribution and reproduction in any medium, provided the original article is properly cited.