Genome Scan of a Nonword Repetition Phenotype in F

Authors: Zoran Brkanac Nicola H Chapman Robert P Igo Mark M Matsushita Kathleen Nielsen Virginia W Berninger Ellen M Wijsman Wendy H Raskind

Publish Date: 2008/07/08

Volume: 38, Issue: 5, Pages: 462-475

PDF Link

Abstract

To understand the genetic architecture of dyslexia and identify the locations of genes involved we performed linkage analyses in multigenerational families using a phonological memory phenotype—Nonword Repetition NWR A genome scan was first performed on 438 people from 51 families DS1 and linkage was assessed using variance components VC Bayesian oligogenic BO and parametric analyses For replication the genome scan and analyses were repeated on 693 people from 93 families DS2 For the combined set DSC analyses were performed with all three methods in the regions that were identified in both samples In DS1 regions on chromosomes 4p 6q 12p 17q and 22q exceeded our initial threshold for linkage with 17q providing a parametric LOD score of 32 Analysis with DS2 confirmed the locations on chromosomes 4p and 12p The strongest VC and BO signals in both samples were on chromosome 4p in DSC with a parametric multipoint LODmax of 236 for the 4p locus Our linkage analyses of NWR in dyslexia provide suggestive and reproducible evidence for linkage to 4p12 and 12p in both samples and significant evidence for linkage to 17q in one of the samples These results warrant further studies of phonological memory and chromosomal regions identified here in other datasetsDyslexia or specific reading disability OMIM 127700 is a common neurocognitive disorder that affects 5–10 of school age children Shaywitz et al 1990 The International Dyslexia Association defines dyslexia as a disorder “characterized by difficulties with accurate and/or fluent word recognition and by poor spelling and decoding abilities These difficulties typically result from a deficit in the phonological component of language that is often unexpected in relation to other cognitive abilities and the provision of effective classroom instructions” Lyon et al 2003 Application of these deceptively simple definitions led different investigators to use different criteria to define what constitutes “difficulties” and “unexpectedness” and different psychometric measures to evaluate dyslexia and “the phonological component of language” that characterizes it This lack of definitional precision has added to the complexity of conducting rigorous research on dyslexiaMultiple lines of evidence support the contribution of genetic factors to dyslexia Fisher and DeFries 2002 Raskind 2001 Twin studies of dyslexia categorically defined have found higher concordance rates for monozygotic vs dyzygotic twins with resulting broadsense heritability h2 of 06 DeFries and Alarcon 1996 and family studies have shown a sibling recurrence rate of 40–50 Gilger et al 1994 Wolff and Melngailis 1994 which is significantly above the population rate of 5–10 Shaywitz et al 1990 Estimates of heritability of dyslexia quantitative traits are also high Twin studies of component phenotypes of reading estimate h2 of 071 for phonological decoding 060 for orthographic choice and 085 for word recognition Gayan and Olson 2001 2003 However in the general population dyslexia does not show mendelian patterns of inheritance and is best categorized as a complex genetic disorder Not surprisingly targeted and genomewide linkage analyses have identified numerous localizations for genes that contribute to dyslexia and related cognitive processes including chromosomes 1p34–36 DYX8 OMIM 608995 Grigorenko et al 2001 2p16–15 DYX3 OMIM 604254 Fagerheim et al 1999 3p12q13 DYX5 OMIM 606896 NopolaHemmi et al 2001 6p222 DYX2 OMIM 600202 Cardon et al 1994 6q13–162 DYX4 OMIM 127700 Petryshen et al 2001 11p155 DYX7 OMIM 127700 Hsiung et al 2004 15q21 DYX1 OMIM 127700 Grigorenko et al 1997 and 18p11 DYX6 OMIM 606616 Fisher et al 2002For several of these loci candidate genes for dyslexia have been proposed DYX1C1 a candidate gene for the DYX1 region is disrupted by a chromosomal translocation that cosegregates with intellectual impairment/dyslexia phenotype in a Finnish family Taipale et al 2003 Two candidate genes in the DYX2 region KIAA0319 Cope et al 2005 and DCDC2 Meng et al 2005 were identified by linkage disequilibrium and association studies ROBO1 a candidate gene for the DYX5 region is disrupted by a translocation in one dyslexic individual and a rare ROBO1 haplotype cosegregates with dyslexia in a large pedigree HannulaJouppi et al 2005 Causative mutations in these genes have not been identified making it difficult to elucidate how these genes contribute to dyslexia in the population In utero RNAi studies in rodents have implicated DYX1C1 KIAA0319 and DCDC2 in neuronal migration Meng et al 2005 Paracchini et al 2006 Rosen et al 2007 and ROBO1 was implicated in axon guidance and dendritic connections Kidd et al 1998 These observations suggest that global brain development disruption might play a role in dyslexia Galaburda et al 2006 McGrath et al 2006Learning to read requires a variety of different linguistic and nonlinguistic cognitive abilities Vellutino et al 2004 each of which may have some unique genetic components Phonological awareness the ability to recognize that words can be decomposed into constituent phonological segments is thought to be a key skill that is typically impaired in dyslexia The ability to deal explicitly with the phonemes the smallest sound units of speech underlies reading acquisition and conversion of written symbols orthography to speech sounds phonology Gathercole et al 1994 Wagner and Torgesen 1987 Additional phonological skills involved in reading include encoding and retrieving of phonological information from shortterm working and longterm memory Kamhi and Catts 1986 To further dissect phonological skills phonological processing of sounds can be isolated from cognitive skills that rely on the meaning of words by the use of tasks that utilize nonwords Nonwords can be pronounced but have no associated meaning and must be processed solely based on sound information they contain The Nonword Repetition NWR task Wagner et al 1999 requires orally presented nonwords to be coded into memory while sounds in them are processed and prepared in order to be repeated Impairments on NWRlike tasks have been found in dyslexia Children with dyslexia make significantly more word repetition errors and use more phonological processes then normal readers Kamhi and Catts 1986 Similarly in a group of children and young adults with dyslexia NWR was significantly impaired compared to normal readers Hulslander et al 2004 Deficits in NWR have also been described in adults with dyslexia compared to normal readers Ramus et al 2003 Szenkovits and Ramus 2005 thus further attesting to the importance of this phenotype in dyslexia Levels of impairment on the NWR measure have been shown to be stable across development Bishop et al 1996 making it additionally suitable for family genetics studiesTo examine the genetic architecture of dyslexia we have adopted a familybased strategy that includes indepth evaluation of theorybased and empirically validated quantitative phenotypes relevant to dyslexia followed by aggregation segregation and linkage analysis of individual phenotypes Berninger et al 2006 Berninger et al 2001 Our initial familial aggregation Raskind et al 2000 and segregation Wijsman et al 2000 analyses of dyslexia phenotypes were performed on 102 families a sample that overlaps with the 144 families presented here In aggregation studies amongst all the measures in our test battery family correlation patterns of performance on NWR and a measure of rate of nonword reading Pseudoword Decoding Efficiency PDE Torgesen et al 1999 gave strongest support for a genetic basis Narrowsense heritability of NWR derived from the correlation between parents and offspring and considering additive genetic variance only was estimated at 020 In the same study we found within individual probands pairwise correlations of 033 and 032 respectively between NWR and accuracy of word reading—Word Identification WID Woodcock 1987 or rate of word reading—Single Word Efficiency SWE Torgesen et al 1999 and 036 and 042 respectively between accuracy of nonword reading—Word Attack WA Woodcock 1987 or rate of nonword reading—PDE Torgesen et al 1999 These are moderate correlations compared to individual pairwise correlations of reading measures WID to SWE WA to PDA and SWE to PDI of 079 06 and 071 respectivelyTo further evaluate the genetics basis of NWR estimate the parameters for models of inheritance and evaluate the suitability of the measure for linkage analysis we performed segregation analyses This analysis provided evidence in support of a majorgene mode of inheritance with multiple contributing loci and in rough agreement with our aggregation analysis it estimated narrowsense heritability at ~030 Together our familial aggregation and segregation studies show significant evidence for heritability of NWR in our sample

Keywords: