GWAS studies
With the aim of studying these genetic risk factors associated with dyslexia, numerous genome-wide association studies (GWAS) have been carried out on the genome of patients to better understand the genetic component of this disease and design new biomarkers that facilitate early diagnosis of the disorder.[20].
The largest GWAS study to date was conducted by Dous et al.,[20] which analyzed 51,800 participants (21,513 men, 30,287 women) who answered "yes" to the question "Have you been diagnosed with dyslexia?" (who are considered cases) and 1,087,070 participants (446,054 men, 641,016 women) who answered "no" to this question (who are considered controls).
After genotyping these patients and performing the GWAS study, 42 significantly genomic loci broadly associated (P<5x) with dyslexia were identified (which are represented in the Manhattan Diagram of Fig. 1). Additionally, 64 loci with suggestive significance (P<1x) were found.
Of these 42 genomic loci obtained, 17 have already been previously described in the literature, and of these, 15 are associated with cognitive and educational traits. Regarding these identified genes, the following stand out:
• - Gene AUTS2: has been linked to autism and intellectual disability. The variant discovered (rs3735260) represents the strongest AUTS2 SNP association identified with a neurodevelopmental trait to date. This gene has a fundamental role in neurodevelopment and in the migration processes of neurons.
• - Gene TANC2: has been linked to language delay and intellectual disability. This gene controls the connections between brain cells.
• - Gene GGNBP2: has been linked to neurodevelopmental delay and autism. This gene is required for the synaptic development of motor neurons.
The remaining 27 loci have not yet been described in the literature and were therefore considered new. These 27 loci are detailed in the following table:
However, because dyslexia is more prevalent in young people (5.34% in people aged 20 to 30 years) than in adults (3.23% in people aged 80 to 90 years), an additional age- and sex-specific GWAS study was carried out. The results of these subsample analyzes showed a high relationship with the main GWAS. To compare both GWAS (the main one and the age-sex-specific one), they estimated the genetic correlation between sexes and between ages using the linkage disequilibrium score (LDSC) regression method. The genetic correlation between sexes gave a value of 0.91; and the genetic correlation between young people and adults gave a value of 0.97. These values indicate a strong genetic concordance in dyslexia between different groups.
Regarding the GWAS carried out in women, 17 significant genetic variants were identified. Most of these variants (13 of 17) were also significant in the main GWAS. However, four variants (rs61190714, rs4387605, rs12031924, and rs57892111) were not significant in the main GWAS. An intergenic SNP (rs57892111) was also obtained, located between the TFAP2B and PKHD1 genes, which was not significant in the main analysis and has no evidence of association with other human traits in previous GWAS studies. This suggests that these variants could be specific to the female sex. The Manhattan diagram obtained in this female-specific GWAS study can be seen represented in Fig. 2..
Regarding the GWAS performed in men, 6 significant genetic variants were identified (present in the genes CADM2, STAG1, PET112, IER5L, DNMT3B and ARFGEF2). These variants were also significant in the main GWAS, so no specificity was observed in the male sex.
With respect to the regions where these genetic variants are found, the following stand out:
• - Variants in introns: 61% of the genetic variants.
• - Variants in non-coding transcripts: 10% of the genetic variants.
• - Variants in downstream genes: 8% of the genetic variants.
• - Variants in upstream genes: 7% of genetic variants.
• - Variants in intergenic regions: 3% of genetic variants.
Regarding the types of mutations that occur, the following stand out:
• - Nonsense mutation: 55% of genetic variants.
• - Synonymous mutation: 29% of genetic variants.
• - STOP codon gain mutation: 11% of genetic variants.
• - STOP codon loss mutation: 2% of genetic variants.
• - Frameshift mutation: 2% of genetic variants.
The heritability of the SNPs was estimated using the linkage disequilibrium score (LDSC) regression method, obtaining a value of =0.152, considering a prevalence of 5%; and =0.189, considering a prevalence of 10%.
Heritability of dyslexia SNPs showed significant enrichment in conserved regions as well as in H3K4me1 clusters and in genes expressed in frontal cortex, cortex and anterior cingulate cortex. However, no specificity was observed for any type of brain cell (neuron, astrocyte or oligodendrocyte). These findings support the crucial importance of the brain and specific brain regions in the development of dyslexia.
Since reading is a uniquely human trait, it was thought that evolutionary changes in the human lineage would influence the genetic architecture of reading ability. However, no enrichment of significant associations was found for annotations covering different periods of hominid prehistory.
During the study, they attempted to relate the genetics of dyslexia to 98 traits or genetic characteristics, including measures of reading and spelling, hearing difficulties, pain threshold, volumes of subcortical brain structures, surface area and cortical thickness, among others. Of these 98 traits studied, 63 showed significant genetic correlations with dyslexia, which we can see represented in Fig. 3. This suggests that there is a common genetic basis between these traits and dyslexia. Next, we see the most relevant associated characteristics:
• - Quantitative measures of reading and spelling: A negative genetic correlation was observed, meaning that as genes associated with dyslexia increase, reading and spelling skills tend to decrease.
• - Verbal reasoning and numerical reasoning: a negative genetic correlation was observed, which suggests that the greater the genetic risk, the lower your ability to understand and reason using concepts framed in words (verbal reasoning) and the lower your ability to understand the main components that involve a mathematical problem.
• - Intracranial volume: a negative genetic correlation was observed, which means that as intracranial volume decreases, the prevalence of dyslexia tends to increase. No evidence has been found that common genetic variations associated with dyslexia are linked to differences in brain structure, such as subcortical volumes or structural connectivity, in adults. Therefore, the phenotypic correlations observed between dyslexia and brain characteristics may be largely due to environmental factors, possibly related to reading.
• - Attention deficit hyperactivity disorder: a positive genetic correlation was observed, suggesting a genetic connection between dyslexia and ADHD.
• - Ambidextrous: a positive genetic correlation was observed, indicating that there is a genetic correlation between equal use of both hands and being dyslexic.
• - Hearing difficulties: a positive genetic correlation was observed, suggesting that hearing problems at an early age could affect the acquisition of phonological processing skills.
• - Pain threshold: a positive genetic correlation was observed, which could indicate a genetic connection between the experience of pain and dyslexia.
Finally, in this study the Polygenic Risk Score (PGS) of each variant obtained in the GWAS study was calculated. In general, higher PGS scores were observed to be associated with lower reading and spelling accuracy. Thus, PGS were correlated with poorer performance on reading and spelling tests in population-based and enriched samples with reading disorders, especially in made-up word reading, a measure of phonological decoding typically impaired in dyslexia. Therefore, PGS could become a valuable tool to identify children with a propensity for dyslexia, allowing educational support before the development of reading skills.