Autism gene acts as ‘switch’ for brain development genes

The autism-linked gene AUTS2 activates a group of genes that may be important for early brain development. And in a surprising twist, it does this by involving a group of proteins that normally inhibit gene expression. The findings, published 18 December in Nature, hint at AUTS2’s role in autism and other developmental disorders¹.  

“If you look at the genes that are targeted by AUTS2, most of those genes are involved in brain function,” says lead researcher Danny Reinberg, professor of biochemistry at New York University.

Researchers first reported AUTS2 mutations in 2002 in a pair of identical twins with autism². Since then, the gene has been implicated in a range of neurological disorders, but its function has been unclear.

“We have been in the dark about the role of AUTS2. We knew when it’s mutated you get a syndrome, but we didn’t know why,” says Erik Sistermans, head of genome diagnostics at Vrije University Medical Center in Amsterdam, who was not involved in the study.

Sistermans led a 2013 study of 50,000 people with intellectual disability that uncovered AUTS2 mutations in 24 individuals³. Of 17 people with well-characterized AUTS2 mutations, 7 have autism, strongly linking the gene to the disorder.

Knowing AUTS2’s role in the brain opens doors to understanding the effects of these mutations, Sistermans says. “It means that if you delete AUTS2, it will not only have an effect on AUTS2 — it will have an effect on a lot of neurological genes,” he says.

Gene switch:

Reinberg and his team stumbled on AUTS2 while studying the biochemistry of polycomb repressive complexes — groups of proteins that inhibit gene expression. When bound to AUTS2, however, one of these complexes instead boosts gene expression by 20-fold.

“There’s been no other report ever that any polycomb group of genes activates transcription,” says Reinberg.

The complexes normally inhibit gene expression by adding tags to histones — the spools around which DNA is tightly wound. AUTS2 recruits an enzyme called CK2 that prevents the complexes from tightening this DNA coil.

“We have been in the dark about the role of AUTS2. We knew when it’s mutated you get a syndrome, but we didn’t know why.”

Unexpectedly, AUTS2 also boosts gene expression. It binds an enzyme called P300 that adds another chemical tag that loosens the DNA coils around histones. AUTS2 also binds to DNA near the start sites of genes, possibly bringing P300 to specific points in the genome.

Reinberg and his team found that AUTS2 is primarily expressed in the cortex of fetal mice during late gestation, with lower levels in adult animals. This suggests that the gene may switch on a network of genes that is necessary for brain development. Studies have similarly found that expression of many autism genes in people peaks during mid-fetal development in the cortex.

Mice that are missing one or both copies of the gene in neurons are less likely to call out to their mothers when isolated as pups — a behavior reminiscent of the communication deficits seen in autism. The mice are also smaller than controls, and struggle to right themselves after being placed on their backs.  

“They did a nice job showing that mice lacking even one copy of AUTS2 have some neurodevelopmental symptoms,” says Nadav Ahituv, associate professor of bioengineering at the University of California, San Francisco. Ahituv studies the role of AUTS2 in the zebrafish brain but was not involved in the new work.

To further understand the role of AUTS2 in motor function, Reinberg and his team plan to engineer mice that lack the gene only in Purkinje cells. This family of neurons helps control brain signals in the cerebellum, a brain region that coordinates movement and has been implicated in autism.

Pinning down the effects of AUTS2 mutations during development is another important next step, says Ahituv. “The timing of expression of this gene is crucial,” Ahituv says. “What exactly is happening in the various tissues during that time when it’s expressed?”

References:

1: Gao Z. et al. Nature 516, 349-354 (2014) PubMed

2: Sultana R. et al. Genomics 80, 129-134 (2002) PubMed

3: Beunders G. et al. Am. J. Hum. Genetic. 92, 210-220 (2013) PubMed