Genetics Of Dyslexia Research Paper

1. Definition Of The Phenotype

In the case of complex behavioral traits like dyslexia, careful cognitive analysis is often needed before the genetic dissection begins. Without such analysis, the phenotype utilized in genetic studies will be much coarser and more likely to be influenced by numerous genes and environmental influences. Genetic studies of dyslexia have built upon extensive research on the psychology of reading. This research has pinpointed which of the many cognitive components of skilled reading are most important for individual differences in reading proficiency. Some of the cognitive components of skilled reading include eye movements, visual perception, language processing, reasoning, and memory. In neuroimaging studies of skilled reading, many parts of the brain ‘light up,’ including parts associated with the cognitive components just listed. The key question then is whether all of these cognitive components are equally important for individual differences in learning to read or whether a cognitive explanation of most of reading difficulty can be provided by a much smaller set of components. If the former situation were the case, we would want to divide individuals with reading difficulty into many different cognitive subtypes before undertaking genetic studies. On the other hand, if the latter situation obtains, genetic studies would be more feasible.

Dyslexia was first described 100 years ago by Pringle-Morgan (1896), but real advances in our understanding of its cognitive phenotype have only come in the last 20 years (see Shaywitz 1996 and Snow et al. 1998 for reviews). Earlier theories of dyslexia focused on the reversal errors made by individuals with dyslexia, such as writing b for d or was for saw, and postulated a basic deficit in visual processing. Orton (1925) termed this deficit ‘strephosymbolia,’ which means ‘twisted symbols.’ The recent advances in our understanding of the cognitive phenotype in dyslexia have built upon basic research focused on both human speech and skilled reading. Although the goal of reading is comprehension, which depends on many higher cognitive processes such as reasoning and memory, it turns out that a very substantial proportion of the variation in reading comprehension can be accounted for by individual differences in the accuracy and speed of single printed word recognition. So, most poor readers have a primary problem with recognizing individual printed words, not with reading comprehension. Not too surprisingly, problems with printed word recognition are almost inevitably accompanied by problems with spelling, since both word recognition and spelling depend on understanding the relations between letters and sounds. Poor readers with these primary problems in word recognition and spelling that are not due to a vision or hearing problem, or a lack of instruction, are called dyslexics. The smaller subset of poor readers with a primary problem in reading comprehension and not in word recog nition are thus not included in the definition of dyslexia.

Single word recognition skill is highly related to phonological coding skill, the ability to pronounce letter strings that have never been seen before (usually measured by having a subject read pseudowords). A reader who can pronounce pseudowords demonstrates an implicit understanding of the quasi-regular correspondences that exist between written letters and spoken sounds. Without this implicit understanding, a reader would have to depend on pure rote memory to read or spell a very large number of seemingly unrelated items. Of course, the pronunciation and spelling of some exception words, such as yacht, have to be memorized. The ability to learn word-specific pronunciations and spellings is called orthographic coding skill. Extensive research has documented that dyslexics are less accurate in reading pseudowords (phonological coding) than even younger, normal readers matched on single, real word reading accuracy. While the majority of developmental dyslexics also have deficits in orthographic coding skill, it is not clear that there is a subtype of dyslexia with only a selective deficit in orthographic coding. So, the vast majority of dyslexics have a deficit in phonological coding that at least partially explains the key defining symptoms of their disorder, slower and less accurate single word recognition and spelling.

But what are these speech sounds that the phonological code represents? A skilled reader knows that the spoken word bag is spelled with three letters and that these letters correspond to the three ‘speech sounds’ in the spoken word. However, a sound spectrograph of this or other spoken words will not reveal discrete speech sounds because of the phenomenon of coarticulation: the articulatory gestures for each speech sound overlap in their production, allowing us to talk much faster than would be the case if we produced each speech sound separately. Linguists call these speech sounds ‘phonemes,’ explicit awareness of which is called ‘phoneme awareness,’ which can be measured by tasks which require a subject to count or manipulate the phonemes in a word. Interestingly, one can learn to talk without phoneme awareness, because spoken language can be processed in larger ‘chunks,’ such as words or syllables. But one cannot learn to read, unless the language one is learning to read is a syllabary, like Japanese Kana. Linguistic theory predicts, and extensive empirical research has confirmed, that the ability to segment these phonemes in spoken speech is a prerequisite for learning to read an alphabetic language like English. If the alphabet is a code and if the letters correspond to phonemes, then how could one understand that code without phoneme awareness? Thus, one might expect deficiencies in phoneme awareness to underlie dyslexia. Indeed, numerous studies have found that dyslexics across the lifespan have deficits in phoneme awareness, which appear to have their origin in sometimes subtle problems with early language development. However, deficits in phoneme awareness are not uniquely predictive of later dyslexia, since other measures of phonological development predict about as well (Snow et al. 1998). Finally, it should be acknowledged that the relation between reading and phoneme awareness is a reciprocal one; while phoneme awareness is a prerequisite for normal reading, reading experience also facilitates phoneme awareness.

In sum, contrary to intuition, individual differences in reading skill have more to do with speech than with vision, and at least as much to do with single word processing as with the processing of connected text. Once we focus on poor readers with primary problems in word recognition, the vast majority are found to have problems in phonological coding and phonological development. So, the cognitive explanation of most of reading difficulty involves many fewer cognitive components than those involved in skilled reading. These facts about the dyslexia phenotype have facilitated genetic studies which have built on this cognitive science understanding of normal and abnormal reading. If these linkage studies of dyslexia had used letter reversals, or even reading comprehension, as the phenotype, it is unlikely that as much progress would have been made.

It is important to emphasize that few of the complex behavioral phenotypes now being tested for linkage have an equally mature cognitive science analysis to build on. For example, other substantially heritable childhood disorders, such as autism and attention deficit hyperactivity disorder, and most adult psychiatric disorders, are less well understood than dyslexia at the cognitive level of analysis. How much of a barrier this will prove to be for genetic studies remains to be seen, but it could be a substantial one.

2. Genetic Mechanisms

In the last two decades, our understanding of the etiology of dyslexia has increased considerably due to advances in both behavioral and molecular genetics. For fifty years after it was first described by PringleMorgan (1896) evidence for recurrence in families was repeatedly documented in case reports, leading Hallgren (1950) to undertake a more formal genetic epidemiological study of a large sample of families. Besides conducting the first test of the mode of transmission, his comprehensive monograph also documented severa1 characteristics of dyslexia that have recently been rediscovered: (a) the sex ratio is nearly equal, about 1.5 M: F and (b) there is not a significant association between dyslexia and non-right handedness. Hallgren also documented that dyslexia co-occurs with other language disorders; however, the degree and basis of this comorbidity has not yet been fully worked out, although it fits with the view that dyslexia is a developmental language disorder.

Although Hallgren and his predecessors provided considerable evidence that dyslexia is familial, it has taken modern twin studies (see DeFries and Gillis 1993 for a review) to demonstrate that this familiality is substantially genetic and modern linkage studies to actually begin to locate the genes involved. Unlike the situation in Hallgren’s time, we now have very strong, converging evidence that dyslexia is both familial and heritable (see Pennington and Gilger 1996 for a review). We can also reject the hypotheses of classic, X-linked or simple recessive autosomal transmission, at least in the vast majority of cases. We also have evidence that dyslexia is genetically heterogeneous. Perhaps most importantly, we have evidence that supports Hallgren’s observation that what looks like autosomal dominant transmission occurs in many dyslexic families. So there do appear to be effects of loci with sizable effects, acting in a dominant or additive fashion, on the transmission of reading problems.

However, we can place several important constraints on Hallgren’s hypothesis of a single autosomal dominant gene influencing dyslexia. First, it is very unlikely to be one gene, because of the evidence for genetic heterogeneity. Second, it may not be a gene influencing dyslexia per se, since the familiality, heritability, and transmission results for normal variations in reading skill are not clearly different from those for dyslexia. Hence, the same loci may be involved in the transmission of both normal reading skill and dyslexia. If this is true, then dyslexics would just have more of the unfavorable alleles at these loci and or more environmental risk factors, such that their reading scores are pushed beyond the cutoff for dyslexia. In this case, the locus (or loci) is not a necessary ‘disease’ locus, but is instead better conceptualized as a susceptibility locus. A susceptibility locus, unlike a disease locus, is neither necessary nor sufficient to produce the disorder in question. If a susceptibility locus influences a continuous (as opposed to categorical) trait, then it is called a quantitative trait locus (QTL). Complex behavioral traits are more likely to be influenced by several QTLs than a single Mendelian locus.

So instead of a classic, autosomal dominant ‘disease’ gene, which is rare in the population and which is by itself necessary and sufficient to produce the disorder of dyslexia, we are likely dealing with several, more frequent quantitative trait loci (QTLs), which are involved in the transmission of both dyslexia and normal variations in reading skill. No one QTL is likely to be necessary to produce dyslexia. Whether one QTL has an effect size sufficient to produce dyslexia is an empirical question which only linkage methods can answer.

In summary, several hypotheses about the transmission of dyslexia can be rejected on the basis of available data reviewed in Pennington and Gilger (1996). (a) Dyslexia is not an X-linked disorder, and there is little evidence of parental sex effects on transmission. There is converging evidence for sex differences in penetrance, which would produce the slight preponderance of males (1.5 M: F) that is observed. (b) Simple polygenic multifactorial transmission can be rejected because there is a major locus effect in several samples. This major locus effect acts in an additive or dominant, but not in a recessive fashion. (c) A monogenic hypothesis can be rejected because dyslexia is genetically heterogeneous. (d) A necessary ‘disease’ allele hypothesis can be rejected because there is evidence for a major locus effect on the transmission of normal variation in reading skill. A remaining hypothesis that fits the empirical data considered here is that a small number of QTLs underlie the transmission of both dyslexia and normal variations in reading skill.

Given the strong possibility that the loci contributing to dyslexia are QTLs, and given the evidence for genetic heterogeneity, traditional linkage analysis (of large extended dyslexic families) is not the most appropriate method to identify these loci. Instead, sibling pair linkage analysis is more appropriate. By selecting sibling pairs in which at least one sibling has an extreme score, one can perform linkage analyses which screen for genetic loci influencing extreme scores on a continuous measure.

Using such a method, evidence for a QTL on the short arm of chromosome 6 was found across two independent samples of sib pairs and across two sets of genetic markers (Cardon et al. 1994). Each sample gave significant evidence of a quantitative trait locus located in a two centimorgan (2cM) region of 6p22.3-21.3.

This finding has now been replicated in three other samples and by two other laboratories (Fisher et al. 1999, Gayan et al. 1999, Grigorenko et al. 1997), although a fourth laboratory using a different phenotype definition did not replicate this result (Field and Kaplan 1998). The Grigorenko et al. study also found some evidence of differential genetic effects as a function of which cognitive component of reading was used to define the phenotype. However, Fisher et al. and Gayan et al. both found that the QTL on chromosome 6 was linked to deficits in both of the two main cognitive components of word recognition, phonological and orthographic coding. This result is consistent with behavioral genetic analyses of these two components, which have found that they have a large, although not complete, genetic overlap (Olson et al. 1999). Work is now under way to identify the actual gene in this region, to screen the remainder of the genome to identify other QTLs that influence dyslexia, and to test whether they differentially influence cognitive components of reading, such as phonological and orthographic coding.

3. Implications Of The Genetic Findings

These exciting linkage findings may eventually allow us to address how much variance in reading scores these genetic loci account for, and eventually how frequently unaffected siblings have unfavorable alleles at these loci. If a similar, sib pair linkage study were conducted using probands selected for extremely high reading scores, we could determine whether different alleles at the same loci influence exceptionally good reading. If so, we could conclude that the same quantitative trait loci are affecting reading scores across the whole distribution. If they are, then our speculation that the same genes are influencing normal and extreme individual differences in reading would be supported. If not, then we would have direct evidence that dyslexia is etiologically distinct. Once we have a better understanding of these genetic mechanisms, we can also conduct much more revealing studies of environmental factors, both risk factors and protective ones, which are also undoubtedly operative in the transmission of both abnormal and normal reading skill.

These linkage results also permit a direct test of the etiology of the comorbidity of dyslexia with other developmental disorders, such as attention deficit hyperactivity disorder and speech and language disorders. Most importantly, once these genes are clearly identified, we can begin to trace the dynamic, developmental pathway that runs from gene to brain to behavior.

Finally, it is important to discuss how the finding of moderate genetic influence on dyslexia and the approximate location of some of the genes involved influences how we treat this common disorder. Here it is important to emphasize that the issues of etiology and treatability are separate ones. Traits under a high degree of genetic influence can nonetheless be quite malleable because the reaction range of the genotype is broad. Many non-geneticists are resistant to genetic explanations because they fear that a genetic etiology means that the disorder is immutable. Instead, exactly the opposite is often the case; understanding the genetics of a disorder can lead to fundamental advances in its treatment. Although we do not currently know whether the linkage results for dyslexia will lead to changes in its identification or treatment, we already know that the phenotype is malleable and that training can significantly improve the educational outcome of children with dyslexia.

Bibliography:

1. Cardon L R, Smith S D, Fulker D W, Kimberling W J, Pennington B F, DeFries J C 1994 Quantitative trait locus for reading disability on chromosome 6. Science 266: 276–9
2. DeFries J C, Gillis J J 1993 Genetics of reading disability. In: Plomin R, McClearn G E (eds.) Nature and Nurture. American Psychological Association, Washington, DC
3. Field L L, Kaplan B J 1998 Absence of linkage of phonological coding dyslexia to chromosome 6p23-p21.3 in a large family data set. American Journal of Human Genetics 63: 1448–56
4. Fisher S E, Marlow A J, Lamb J M E, Williams D F, Richardson A J, Weeks D E 1999 A quantitative-trait locus on chromosome 6p influences different aspects of developmental dyslexia. American Journal of Human Genetics 64: 146–56
5. Gayan J, Smith S D, Cherny S S, Cardon L R, Fulker D W, Brower A M, Olson R K, Pennington B F, DeFries J C 1999 Quantitative-trait locus for specific language and reading deficits on chromosome 6p. American Journal of Human Genetics 64: 157–64
6. Grigorenko E L, Wood F B, Meyer M S, Hart L A, Speed W C, Shuster A, Pauls D L 1997 Susceptibility loci for distinct components of developmental dyslexia on chromosomes 6 and 15. American Journal of Human Genetics 60: 27–39
7. Hallgren B 1950 Specific dyslexia: A clinical and genetic study. Acta Psychiatrica Neurological Scandinavian (Supp) 65: 1–287
8. Olson R K, Datta H, Gayan J, DeFries J C 1999 A behavioral-genetic analysis of reading disabilities and component processes. In: Klein R M, McMullen P A (eds.) Converging Methods for Understanding Reading and Dyslexia. MIT Press, Cambridge, MA.
9. Pennington B F, Gilger J W 1996 How is dyslexia transmitted? In: C H Chase, G D Rosen, G F Sherman (eds.) Developmental Dyslexia: Neural, Cognitive, and Genetic Mechanisms. York Press, Parkton, MD, pp. 41–61
10. Pringle-Morgan W P 1896 A case of congenital word-blindness (inability to learn to read). British Medical Journal 2: 1543–4
11. Shaywitz S 1996 Dyslexia. Scientific American 277: 98–104
12. Snow C E, Burns M S, Griffin P 1998 Preventing Reading Difficulties in Young Children. National Academy Press, Washington, DC