Finding gene variants associated with common neurological and developmental conditions, such as dyslexia, is turning out to be unexpectedly difficult. While every geneticist hopes to find a small number of genes that have strong links with a condition, in practice they are finding large numbers of weak links. The picture that is emerging suggests that common problems, even those that run in families, emerge from complex interactions among genes, and between genes and the environment.
Teasing out these interactions requires data from very large sample populations. Dr Silvia Paracchini and colleagues in the neurogenetics group at the WTCHG are among the first to be carrying out a major analysis of genetic and environmental data that is available from many of the 14,000 children in the Avon Longitudinal Study of Parents and Children (ALSPAC). Their Medical Research Council-funded project (which also includes collaborators from Oxford's Department of Experimental Psychology and groups from the University of Bristol and the Institute of Psychiatry in London) is focusing on problems with reading and language in these 'children of the 90s', all of whom were born in 1991 or 1992.
'Reading is specific to people, and requires instruction', says Paracchini. 'It's something that you learn. People with dyslexia have a specific problem to learn to read. A hundred years ago, when people were not expected to read, it would not be an issue in our society. Now it's been recognised, it's turned out to be very common.' Depending on how stringently you characterise the problem, between 5 and 10 per cent of people have the condition; a slightly smaller number has specific language impairment (SLI), which affects their ability to communicate through speech. The new study gives Paracchini and her colleague Dianne Newbury the opportunity to explore the overlap between the two.
Genetic and reading test data from the 'children of the 90s' might help to reveal the roots of dyslexia (Picture credit - ALSPAC)
Early work searching for a genetic link in families with more than one dyslexic member found linkages on chromosomes 1, 2, 6, 15 and 18. Not all of these were replicated between one study and another, and none of them is strong enough on its own to be claimed as the 'cause' of dyslexia.
Paracchini and her colleagues discovered that single nucleotide polymorphisms - single-letter variants in the spelling of the genetic code - in or near the gene KIAA0319 on chromosome 6 were linked to dyslexia in families with the condition in both the UK and the US. Other groups have now confirmed this link. It was when she wanted to find out if the gene also showed a link with reading in the general population that Paracchini first turned to the ALSPAC cohort. A large proportion of these children had had their reading assessed at age 7, and even more had donated DNA samples. Her paper confirming that variations in the genetic code in or near the KIAA0319 gene were significantly associated not just with dyslexia but with reading ability generally in the 'children of the 90s' came out in October 2008. At the same time, work on the biology of the KIAA0319 gene showed it is highly expressed during early brain development.
The variant associated with dyslexia has a reduced level of expression. But that is clearly not enough to cause dyslexia: in the ALSPAC study, this variant was only 1.2 times as common in the worst as in the best readers. 'It will interact with something else,' says Paracchini, and her visits to ALSPAC's Bristol headquarters had already set her thinking about how she might work with the study's researchers to explore those interactions.
'I realised that this sample had great potential', she says. 'They have measures of all sorts of things including pregnancy and birth, the mother's health, the environment - we can ask, for example, whether the children had access to books at home in their infancy. And such a large sample, even if it is not selected for any particular characteristic, gives you the power to detect even small effects.' The MRC grant will allow Paracchini and her colleagues to search the ALSPAC data for more than 40 genetic variants known to be associated with speech or language disorders, and to look for both genetic and environmental interactions. 'In ALSPAC we have the chance to check if a candidate gene associated with dyslexia might also be involved in language impairment, or if the genes are completely separate', she says. 'We can also see, if you have a predisposing genetic variant, what happens if you are exposed to an environment that's favourable or unfavourable for developing reading skills.'
The evidence is beginning to build that the preconditions for dyslexia are laid down in early development, under genetic influence. 'We always say it is a neurodevelopmental disorder,' says Paracchini. 'Something goes wrong while the brain is developing. You won't know you are dyslexic until you start to read, but it starts much earlier.' The findings of the ALSPAC study will shed light on the genetic interactions controlling this development. 'Genes can give you an idea of where to look in the brain', she says. 'You can look at an anatomical level at what happens when these genes are compromised.' Ultimately the goal would be to achieve early diagnosis so that therapists can intervene early and appropriately. 'Dyslexia is such a broad condition', she says. 'We would hope to be able to define subgroups and identify which is the best therapy to follow. You have to start early - the sooner the better.'
For more information, see www.well.ox.ac.uk/dyslexia-2