Autism, Chromosome 15, and the Power of Timing

Autism is not one story. For a small group of people, it’s tied to changes in a specific stretch of DNA on chromosome 15. In his talk, neuroscientist Eric S. Levine, PhD, explained why this region matters, how timing shapes treatment, and why more than one gene may drive symptoms—especially seizures.

A simple map of a complex region

Levine focused on Angelman syndrome and Dup15q syndrome. Both involve the same neighborhood on chromosome 15 (15q11–q13) but differ in what’s missing or extra and which parent’s copy is affected. The reason is “genomic imprinting.”

As Levine put it, “Genomic imprinting refers to an epigenetic process where the expression of a gene only comes from one of the inherited alleles, and is epigenetically silenced on the other allele.” In brain cells, a key gene here—UBE3A—is active only from the mother’s copy. If that active copy is lost, Angelman syndrome results. If the region is duplicated on the maternal side, Dup15q often presents with autistic features and high seizure risk.

Notably, Levine said Dup15q “is highly penetrant for profound autism,” with “80 or 90% of the kids with the most severe form of the duplication” showing significant support needs.

Why timing matters

Restoring UBE3A early seems to help most. In mouse studies of Angelman syndrome, “you can get a restoration of function if you restore UBE3A, but it typically needs to be done prenatally.” Later treatment helps less. That reality shapes how scientists think about future therapies. “Correcting these things postnatally may require looking more downstream, because… UBE3A may not be an ideal target once you’re out beyond that critical period.”

Beyond one gene: the “co-drivers”

UBE3A is a major player, but it is not alone. “There’s good evidence that the other genes that are in that region of the duplication or deletion also contribute,” Levine said. His team studies several candidates, including GABA_A receptor subunits that control inhibitory “brake” signals in the brain.

In Angelman-model human neurons, Levine’s group saw weaker GABA signaling. When they partially lowered the β3 subunit (to mimic the larger deletion), inhibition dropped further. In Dup15q neurons, the flip side appeared: stronger-than-normal GABA currents. When the team used an antisense oligonucleotide to dial β3 back down to typical levels, those abnormal GABA signals normalized. The takeaway: “It provides some support for targeting Beta-3 in conjunction with UBE3A as a therapeutic target for Angelman syndrome,” and suggests a similar co-targeting logic for Dup15q.

The right model for the job

Standard mouse models often change UBE3A alone. That misses the broader 15q shifts. Levine’s lab turns patient blood or skin cells into induced pluripotent stem cells (iPSCs) and then into human neurons. These cells carry a person’s exact DNA, imprint correctly over time, and develop active synapses after weeks in culture. This lets the team test how single genes and combinations alter real human neuronal signaling—and which levers might be safest to pull.

What this means now

Two big ideas connect here:

1. When a gene acts matters. Early development is a moving target. Therapies that restore UBE3A may work best very early. Later on, downstream pathways may offer more leverage. 
2. It’s not just one gene. UBE3A sits in a noisy neighborhood. Other genes, like the GABA receptor β3 subunit, can push brain networks toward or away from seizures. Calibrating those “brakes” may reduce risk and improve life quality.
   

For families and clinicians, this adds nuance without removing hope. New genetics-based tools are coming. The smarter they get about timing and co-targets, the better they may serve people living with Angelman, Dup15q, and related conditions.

Sources

1. NJACE & RUCARES Annual Conference, September 16, 2025, Douglass Student Center, Rutgers University.
2. Speaker transcript: Eric S. Levine, PhD, University of Connecticut School of Medicine.