A non-protein-coding RNA may influence social and repetitive behaviors linked to autism, according to a new study.
The RNA, called PTCHD1-AS, is produced from a gene in a region of the X chromosome linked to autism and other neurodevelopmental conditions. Rather than encoding a protein, the new work suggests, PTCHD1-AS appears to regulate brain circuits in the striatum, a region involved in social and repetitive behavior.
“So far as we know, there is no other long noncoding RNA that has the same type of impact that PTCHD1-AS does,” says study investigator Lisa Bradley, research associate at the Hospital for Sick Children.
Noncoding RNAs are more abundant in the human brain than are protein-coding RNAs, but scientists know little about what most of them do, particularly during brain development, says Daniel Campbell, assistant professor of pediatrics and human development at Michigan State University, who was not involved in the study.
PTCHD1-AS is among the few noncoding RNAs with strong evidence linking it to autism, Campbell says. “It is fantastic that we are finally uncovering the mechanisms by which noncoding RNAs like PTCHD1-AS contribute to altered brain development,” he says.
B
radley and her colleagues analyzed whole-genome sequencing data from 9,349 autistic boys and 8,332 non-autistic controls, focusing on boys because PTCHD1-AS is X-linked. They identified 27 autistic boys with small deletions affecting PTCHD1-AS, which were associated with increased odds of autism.Two sets of mice—one engineered with a disruption of the PTCHD1-AS gene and the other with a partial deletion of that gene—showed structural differences in several brain regions linked with autism compared with control mice. The engineered mice also showed reduced sociability, the team found. Unlike control mice, they showed no preference for another mouse over an object, or for a new mouse over a familiar one. They also groomed themselves more—a potential proxy for repetitive behavior.
The mice did not show broad cognitive difficulties, though. Their learning, memory and synaptic function in the hippocampus were largely unaffected.
In control mice, PTCHD1-AS expression rose in the striatum after birth and remained elevated into young adulthood, the team found. And in mice with a disrupted PTCHD1-AS, synaptic plasticity in this brain area was altered—connections between neurons strengthened or weakened more readily than they did in control mice.
Striatal neurons and glial cells also showed changes in genes and proteins involved in synaptic signaling, the team reported. Female mice, which have another X chromosome copy, did not show obvious brain structural changes, Bradley says.
T
he work, which was published last month in Nature, provides a clear link between PTCHD1-AS and autism-related traits, and it changes how researchers should think about how non-protein-coding regions can shape neurodevelopment, says Michael Halassa, professor of psychiatry and biomedical engineering at Virginia Polytechnic Institute and State University, who was not involved in the work. But, he adds, “there are still many questions to answer.”Standard mouse behavior tests may not capture attention, working memory or other subtle cognitive functions, Halassa says. It also remains unclear how the striatal changes connect to the animals’ social behavior or repetitive actions, he says.
Still, the findings are starting to change clinical interpretation, says study investigator Stephen Scherer, chief of research at the Hospital for Sick Children. PTCHD1-AS is located near two protein-coding genes, PTCHD1 and DDX53, which have been linked to autism and other neurodevelopmental conditions.
Previously, PTCHD1-AS deletions were not reported as harmful unless they also involved DDX53, Scherer says. Now clinicians will recognize these deletions as pathogenic, he adds.
What’s more, having mouse models that separate autism-like behaviors from broader learning and memory difficulties can give researchers a way to probe the circuits underlying autism’s core traits, says study investigator Graham Collingridge, professor of physiology at the University of Toronto. “We’ve only just started to scratch the surface of the molecular basis of the phenotype,” he says.
