A team of researchers at Cold Spring Harbor Laboratory found what may be evidence for a genetic basis for autism. Using an engineered mouse model, Alea Mills and her colleagues show in a paper published in PNAS last month that the deletion of a 27-gene cluster on chromosome 16 can cause autism-like features, and that copy-number variations in this region can affect the architecture of the brain and the behavioral phenotype.
This study builds on work done in 2007 by Mills' colleague Michael Wigler and others, who discovered that some autistic children have deletions in the 16p11.2 chromosomal region. "I thought it was really an intriguing region because there was both loss and gain, and so there was a big question of what was going on," Mills says. "What we set out to do is model both the loss and gain and do a side-by-side comparison."
Using chromosome engineering technology and the Mutagenic Insertion and Chromosome Engineering Resource of vector sequences and targets specifically developed to make high-throughput manipulation of mouse genomes easier for researchers, Mills' team created mice with deletions in 16p11.2, and duplications of the same region. The technology allows researchers to create specific mouse models, using mouse embryonic stem cells in culture, and engineer specific changes in those cells. Mice with the deletion had some behavioral characteristics common in autism like hyperactivity, sleeping deficits, and repetitive behaviors.
Further, the mice that had been engineered with a duplication displayed different behaviors. "We saw reciprocal effects — loss did one thing and gain did the opposite — on either side of normal," Mills says. "For everything we've looked at, including gene expression, viability, behavior, and brain anatomy, however, the deletion was always more severe than the duplication."
The researchers also used MRI to look at the specific construction of the mouse brains, and saw alterations in their architecture, which Mills and her colleagues plan to explore further in future studies. "Right now we have a hypothesis. Our completely unbiased methodologies — MRI and behavioral analysis — point to the hypothalamus, and so we'd like to know if alterations in the hypothalamus are responsible for the behavioral phenotype," Mills says.
She and her team plan to delve more deeply into the genetic mechanisms of autism, specifically to determine which of the 27 genes in the 16p11.2 region might contribute to the behavioral phenotype. "We would like to subdivide the interval, so we can disrupt specific genes that we're interested in, and we can make sub-deletions or sub-duplications within the interval," Mills adds. "And we'll basically ask, now that we have these robust phenotypes for the full deletions, whether these sub-deletion models have the same phenotypes." Eventually, she says, it may be possible to determine biomarkers that signal if a patient is developing autistic features, before the disorder is fully realized.