Many journeys to the same location – methods that 21st century scientists are using to identify genes that cause autism
We have asked several scientists who gave presentations at the April 10-11 DAN! conference in Baltimore to share their research and perspectives from the meeting with you here on the blog. The following piece is from Dr. Simon G. Gregory, Ph.D. Dr. Gregory is an Associate Professor in the Center for Human Genetics, Duke University Medical Center His primary area of research involves identifying the complex genetic factors that give rise to cardiovascular disease and multiple sclerosis, Dr. Gregory’s group is also applying high-resolution genomic arrays to the discovery of chromosomal abnormalities and identification of epigenetic factors associated with autism.
For decades scientists have relied on traditional genetic approaches to identify genes that contribute to the development of autism. These approaches have included collecting extended, multi-generation families that contain one or more individuals with autism. Once a sufficient number of these families have been collected, hundreds of DNA markers that can highlight differences between us all at a genetic level (a bit similar to DNA fingerprinting used in CSI) are used to identify a region of the genome is linked with the disorder in the families. More recently, higher resolution technologies that benefitted from the completion of the human genome project have been developed to allow us to look at hundreds of thousands of markers simultaneously. This approach differs from the family-based approach in that it uses thousands of unrelated cases and age/sex/ethnicity matched controls. Together, both approaches have been successful in identifying genes, or at least regions of our chromosomes, that play a role in the development of autism. However, we know that many more genes will likely contribute to autism spectrum disorders (ASDs) and that not all mechanisms that disrupt the normal function of a gene will be based on a single DNA mutation or a variant present in the general population.
My group at the Duke Center for Human Genetics recently embarked on a journey to identify autism genes using an approach that looks at chromosome content, and yet we reached an interesting finding using another technique that doesn’t rely on the sequence of our DNA. For many years it has been known that ASDs can be caused by loss (deletion) or gain (duplication) of chromosomes. These genomic rearrangements can be large or small and some have been shown to cluster in particular regions of the genome, for example on the long arm of chromosome 15 (15q11-13). Our journey began by assessing the genomic content of 119 unrelated autistic children from families that we had collected over a number of years. We used genomic microarrays (microscope slides with printed content spanning our genome) to identify novel or common regions of deletion or duplication within one autistic child from each family. Although we identified a handful of novel regions and confirmed dozens of rearrangements that have increasingly been shown to exist within the ‘normal’ population, we decided to focus on a single deletion in one individual. The deletion, which contained five genes, sparked our interest because one of the genes, the oxytocin receptor (OXTR), had been previously implicated in autism. Previous studies using the traditional genetic approaches (mentioned above) had shown that OXTR was genetically associated with autism, other studies had shown that levels of oxytocin (a neurotransmitter that the receptor binds to OXTR) is important for pro-social behavior and that a knock-out of the receptor in mice can affect this behavior, and very recent studies have tested the supplementation of oxytocin in adults with autism.
After identifying the deletion in the one affected child we screened his family to see if it was novel and causative of his autism, or if he inherited the deletion from his parents. Somewhat disappointingly, his mother also had the deletion, but interestingly she had self reported obsessive compulsive disorder (OCD) for which OXTR has also been implicated. Although we could have stopped the journey at this point, we decided try a different approach. Previous studies have shown that OXTR is controlled in the liver by a chemical group (methyl –CH3) that sits on top of the DNA but which doesn’t change its sequence (click here for more information on epigenetic changes as they relate to autism). The phenomena of DNA methylation is a normal mechanism by which our cells silence parts of the genome that they don’t want to be active (mainly repeat and retro-viral sequences) but they also use it as a switch to turn on or turn off genes for protein production. Significant in our journey was that the child with autism who had the OXTR deletion also had an affected brother. This is significant because when we assessed the methylation status of the whole family we noticed that the affected brother has very high levels of methylation in five different CG bases (among the A’s C’s G’s and T’s of our genome where the methyl group resides) in a 1,600 base pair window that had previously been identified as being methylated in the liver OXTR.
Differential methylation isn’t a novel finding as a possible cause of autism. Previous studies in Rett syndrome, which has ASD like symptoms, have shown that mutations in the methyl CpG binding protein 2 (MECP2) gene not only contributes to the development of Rett syndrome but that these mutations affect the normal functioning of the gene which regulates the expression of other genes by altering their DNA methylation state. Additionally, imprinting (the silencing of a maternal or paternal copy of a chromosome by DNA methylation) has also been established as the basis for the development of Angelman and Prader-Willi syndromes that, again, have ASD like symptoms. Our next step was to validate the observation that increased methylation of OXTR in the affected brother may be a causal mechanism of autism. We expanded our analysis to include evaluation of the methylation status of OXTR in the peripheral blood of 20 controls and 20 individuals with autism, and in a second dataset of the post-mortem samples of temporal cortex tissue (a center for language and learning) in eight controls and eight children with autism. In both datasets we were pleased to see that the individuals with autism had higher levels of methylation than the controls. We next measured the level of gene expression of OXTR because we would expect that increased DNA methylation would result in decreased gene expression – which it did in a very small number of samples that we could test.
So you’re asking yourself, how, if all these other studies had pointed to OXTR being implicated in autism, is this discovery significant? Well, we hope that we have found a unique autism-associated DNA methylation signature in a receptor for a neurotransmitter that is important in social cognition in autism. The next more exciting journey will be to substantiate what we have found in a larger number of samples and to establish if oxytocin treatment can help