Genetics of Autism: Where
its been—where its going
Autism
has, for some time now, been recognized as a disorder with a strong genetic component.
However, the actual genes involved have, so far, been very difficult to identify. In 1991,
a gene called FMR1 on the X chromosome was identified. An FMR1 gene that carries a
mutation can lead to a mental retardation disorder known as Fragile X mental retardation. Because females have two X
chromosomes whereas males only have one, a girl carrying such a mutation on one of her X
chromosomes will have much milder symptoms than a boy with the same mutation, because her
other chromosome carrying the normal copy of the
FMR1 gene will partly compensate for the effect of the defective gene. Sometimes girls
with the FMR1 mutation and the milder form of mental retardation may be diagnosed with
autism. In 1999 another gene on the X chromosome, MECP2, was identified as the genetic
cause of Rett’s syndrome, another mental retardation disorder where some of the
symptoms may overlap with autism. In this disorder, only females are affected, because the
mutation in males is lethal before birth. Again, it is the fact that females have the
extra X chromosome that makes the symptoms less severe. Although the discovery of these
genes has been very enlightening, they only account for a very small percentage of
diagnosed autism cases. However, what these two examples do give us is a better
appreciation of some of the complexities of genetic diseases, particularly where a
disorder is seen predominantly in one gender over the other, as is the case in autism
where there are roughly 3 males diagnosed for every female.
A brief primer in genetics
Although autism is considered a disorder with a strong genetic component, the inheritance of autism differs greatly to that of other, better understood, genetic disorders such as cystic fibrosis or Huntington’s disease. For a new-born child with a sibling who has cystic fibrosis, a “recessive” genetic trait, the risk of also getting the disease is around 25%. This sibling risk for Huntington’s disease, a “dominant” trait, is even higher, at around 50%. In contrast, this sibling risk in autism may be as low as 3%, and the inheritance pattern is neither clearly recessive nor dominant. The sibling risk for autism is somewhat closer to that for schizophrenia, which is around 10%. While clearly recessive or clearly dominant disorders such as cystic fibrosis and Huntington’s disease are quite straightforward when it comes to actually identifying the main genetic cause, disorders with less obvious patterns of inheritance and lower sibling-risk values are much more complicated to figure out. Ironic then, that these “complex” disorders that include diabetes, late-onset Alzheimer’s , hypertension, schizophrenia, manic depression and autism, should be so much more prevalent than the “simple” genetic disorders. So what makes complex disorders so complex? The fact that more than one gene (possibly many genes), as well as non-genetic factors (for instance dietary factors in diabetes), and the fact that different combinations of genes and “environment” may cause the same disease, sometimes even in the same family.
So, what is being done to identify
genetic factors that cause autism?
In recent years many researchers have turned their attentions towards this problem. In 1998 the first “genome scan” for autism was reported. This is where all 22 chromosomes plus the X and Y sex chromosomes are screened for potential disease genes in families with more than one affected child. A region on chromosome 7 was identified as the most likely to harbour such a disease gene for autism. Subsequently, several other research groups performing genome-scans for autism also identified the same region of chromosome 7.
So, does that mean there is a gene for autism on chromosome 7? And how soon before the disease gene is identified?
In answering this, a little bit of history in the study of complex genetic traits might be useful. In 1988 the first genetic findings for schizophrenia were reported, with a potential disease gene on chromosome 5. At the same time a second group published the findings of their work suggesting that there was not a potential schizophrenia gene on chromosome 5. Since then, numerous studies have identified many different regions of the genome for schizophrenia. Some of these regions overlap, but so far very little consensus has been established, despite the efforts of a number of huge collaborative studies, and consequently no definitive susceptibility genes for schizophrenia have been found as yet. This is an extreme example of the limitations of finding genes for complex traits. One of the problems with schizophrenia is that it is very common, occurring in one in a hundred people, which means that families used for the genetic studies may be “contaminated” by individuals with the same disease but with different causes. Autism, in contrast, is much rarer, and so is much less likely to be affected by this type of problem. Another problem with studying schizophrenia and other complex genetic disorders is that the disease usually manifests itself either in late adolescence or in adulthood. By this time, the individual has been exposed to many external or environmental factors that may play a role in the development of the disorder. In contrast, autism is a juvenile onset disorder, and will have been exposed to fewer contributory external factors. This then would suggest that autism may stand a better chance than other complex disorders, but only time, and a lot more hard work, will tell whether we can find a gene for autism on chromosome 7.
So are we close to finding a gene for
autism on chromosome 7?
My answer would be “closer” but not “close”. Despite the continued efforts of a number of research groups, including our own, which are trying to narrow down the “autism region” on chromosome 7, this region does not seem to be getting any smaller, and may still encompass some 50 million letters of DNA sequence- a region spanning almost one third of the size of the whole of chromosome 7. On the plus side, the evidence for a gene there has not vanished (vanishing evidence for disease genes is a common hazard in the study of complex genetic disease) Also, with the fruits of the human genome project now being harvested, at least we have a much better idea of what genes are in that region (in fact, there are far fewer genes than was anticipated), and we can start trying to establish what the function of these genes are, and which of them may be candidates for involvement in autism. This is one approach that the autism project at The Hospital for Children in Toronto is employing. Another strategy that we are using to identify candidate genes for autism is to find rare cases of autism where the individual also has a structural abnormality on chromosome 7. We then attempt to “map” whereabouts on the chromosome the abnormality occurs, and then see if it disrupts any genes. So far we have successfully identified one such gene, which we are now studying in more detail to determine whether it plays any role in the cause of autism.
All this effort to identify a gene for autism on chromosome 7 should not distract us from the certainty that there will be other, perhaps more important, genes for autism elsewhere in the genome. Hence it is crucial to keep an open mind as far as different hypotheses for autism are concerned. If one cause of autism can be determined, either genetic or non-genetic, it should, by deducting those cases from our studies, make the identification of other genes simpler. Back to the X chromosome, where the FMR1 and MECP2 genes were identified, and indeed on any other chromosome, it seems a tantalizing possibility that in genes that may cause mental retardation, a much less severe mutation within the same gene may cause much milder and perhaps more autistic-like symptoms