The author is a guest blogger, Ashley Scott-Van Zeeland, Ph.D. and she is a Dickenson Fellow at the Scripps Translational Science Institute in La Jolla.
This was the second year IMFAR hosted an Invited Educational Symposium on Imaging Genetics in autism spectrum disorders. Dr. Susan Bookheimer, a leader in brain imaging and one of the pioneers of imaging genetics, and Dr. Daniel Geschwind a world-renowned autism geneticist, chaired the session. The goals of the symposium were to introduce the principles behind imaging genetics and demonstrate methods by which various researchers are using imaging genetics to discover relationships between genetics and brain function.
The first speaker was Dr. Susan Bookheimer, Professor of Cognitive Neurosciences from UCLA. Dr. Bookheimer gave a broad overview of the use of brain images as quantitative measures for use in imaging genetics investigations. Essentially, neural measurements such as activation or metabolic response, structural volumes, or connectivity measures can be thought of as being one step closer to the mechanism of gene action than broad diagnostic classifications. Therefore, these types of ‘endophenotypes’ are more strongly associated with gene variants and should considerably increase the ability to identify brain-gene relationships.
Dr. Daniel Geschwind, Chair of Human Genetics and Professor of Neurology at UCLA followed with a general primer on genetics to introduce all clinicians, brain imagers, and non-geneticists to the types of genetic assays available. The goal of Dr. Geschwind’s talk was to encourage non-geneticists to consider what is “reasonable” from a genetics perspective when designing studies. He highlighted three main frameworks for imaging genetics studies: 1) Identify genetic factors that underlie normal brain structural and/or functional variation 2) use differences in brain imaging measures to identify disease-associated genes, or 3) use brain imaging as a way understand the neurobiological effects and neural mechanisms of disease-associated genes. He also re-emphasized the issue of statistical power in the search for disease-associated genes, which depends on many factors but can be increased by using brain imaging that likely has larger detectable gene effects than broader phenotypes such as autism diagnosis. He also stressed that the best imaging genetic approaches are hypothesis driven, with known implicated brain regions or neural effects to limit statistical issues caused by testing many genetic markers over many brain regions.
After the introductory talks by Drs. Bookheimer and Geschwind, there were three more talks that went into greater detail about current applications of imaging genetic approaches in the study of autism – the first being by Dr. Joshua Trachtenberg, Assistant Professor of Neurobiology at UCLA. Dr. Trachtenberg presented methods for imaging dendritic growth in animal models. Using cutting edge 2-photon microscopy techniques, Dr. Trachtenberg is able to examine the neurobiological effects of candidate autism risk genes in a living animal. His lab has been working on the PTEN gene, which has been associated with an enlarged brain. In order to determine how PTEN contributes to an enlarged brain, Dr. Trachtenberg is able to selectively alter the PTEN gene in mouse brain after birth and monitor the changes in neural architecture over time. In the PTEN mutant animals, Dr. Trachtenberg observed up to 50% extra growth of dendrites – neuronal structures that receive input from other cells – suggesting a mechanism by which this autism-associated gene contributes to larger brain size. Importantly, he also presented evidence that the dendrites involved in long-range communication circuits are more affected than those involved in short-range circuits – a story that is emerging as a consistent finding across many measures of brain connectivity in autism.
Next, Dr. Declan Murphy, Professor of Psychiatry and Brain Maturation and the Head of Department of Forensic and Neurodevelopmental Science at Kings College in London presented an overview of the state of the field for serotonin genes and imaging genetics. Serotonin is a neurotransmitter involved in both brain function and brain development, and has been associated with behaviors such as aggression, affiliative behaviors, and obsessional behaviors. There are two major serotonin genes that have been the focus of imaging genetics studies – MAOA and 5-HTTLPR. MAOA is involved in the breakdown of serotonin and 5-HTTLPR encodes a protein that transports serotonin into neurons. Both genes have been associated with many neuropsychiatric disorders, including autism. Dr. Murphy presented studies in which activity in the amygdala and fusiform gyrus, regions thought to play a role in autism, was associated with different versions of 5-HTTLPR. However, he noted that although there is some evidence for serotonin dysfunction in autism, the evidence is not conclusive and imaging studies reflect this as well. In structural MRI studies, investigators have found frontal lobe abnormalities associated with the serotonin transporter, though this has not consistently been replicated. Within this talk, Dr. Murphy emphasized that it remains important to search beyond simple gene effects and consider systems-based approaches, developmental effects, gene-gene interactions and pharmacogenetic effects in imaging genetics studies.
I had the great pleasure of wrapping up the session by describing a general framework for pursuing targeted imaging genetics studies in autism, using a recent study that explored frontal lobe connectivity and an autism risk gene, CNTNAP2. As imaging genetics studies can be quite difficult, it is important to begin with well-devised experimental paradigms and testable hypotheses for gene effects. Since CNTNAP2 encodes a protein that is involved in cell-cell interactions and synaptic transmission, we reasoned that CNTNAP2 might be related to neural connectivity. As mentioned by Dr. Bookheimer in the introduction, it is important to use brain imaging as a quantitative measure of brain function or structure. To this end we chose to use a functional paradigm that we knew would elicit activity in regions where CNTNAP2 is expressed during development. We observed differences in frontal lobe activity in both typical children and those with autism depending on which CNTNAP2 allele they carried, and also found differences in connectivity patterns with the frontal lobe such that children who carried the autism-associated risk gene had more local frontal connections and reduced long-range connections. In sum, by leveraging information about the role of CNTNAP2 in the brain , we were able to better understand the mechanism by which CNTNAP2 contributes to increased risk for autism using an imaging genetics approach.
Overall, the application of imaging genetics to autism spectrum disorders is beginning to reveal very interesting neurobiology and I expect we will see more of these types of studies at IMFAR 2012!