Summary: New research from the University of Minnesota Medical School finds that different genetic forms of autism can produce a common pattern of brain activity and related behavior. By recording neural activity across the entire brain in preclinical models, the team identified a shared neural signature: mice with autism-linked mutations showed difficulty updating expectations during decision-making, relied more on frontal brain regions, and showed reduced contribution from sensory areas.
These results suggest a potential bridge between diverse genetic causes of autism and the behavioral traits they produce, offering a clearer target for future studies that aim to connect genes, circuits, and behavior.
Key Facts:
- Shared brain signature: Different autism-associated genetic mutations produced similar patterns of brain activity across models.
- Sensory processing differences: Mutant models showed reduced flexibility in updating expectations based on sensory input.
- Frontal dominance: Decision-making in mutant models depended more on frontal cortical activity and less on sensory regions.
Source: University of Minnesota
Overview
Scientists at the University of Minnesota used brain-wide neural recordings and behavioral testing to examine whether mice carrying different autism-linked genetic mutations share common computational and neural features. The study, published in Nature Neuroscience, combined high-throughput psychophysics, detailed behavioral modeling, and single-cell extracellular recordings across the brain to compare three established genetics-based mouse models of autism.
Across models, the researchers detected a consistent computational anomaly: mice with mutations in genes linked to autism—Fmr1, Cntnap2, and Shank3B—showed a blunted ability to update their internal predictions, or priors, in response to new sensory information during decision-making tasks. In other words, these animals relied more heavily on long-term expectations and less on moment-to-moment sensory evidence.
Neurophysiological recordings revealed where this computational bias is reflected in the brain. Compared with control animals that flexibly adjusted expectations, mutant mice shifted the neural encoding of prior information away from sensory cortices and toward frontal brain regions. Frontal areas in the mutant models showed increased representation of deviations from the animals’ long-run expectations, while sensory areas failed to clearly differentiate expected from unexpected observations.
Functionally, this imbalance produced behavioral consequences: when sensory input was predictable or unpredictable, the mutant mice struggled to use incoming sensory cues to update their decisions, reducing behavioral flexibility. The team identified a feedback circuit projecting from frontal areas to visual cortex that appears to be central to these shared anomalies; follow-up work will probe that pathway in more detail.
Implications for autism research
Although autism spectrum disorder arises from many different genetic and developmental pathways, this study supports the idea that distinct genetic instantiations can converge on similar neural and computational outcomes. Identifying a common neural signature—an imbalance favoring frontal over sensory processing during expectation updates—helps explain a range of behavioral features and sets a concrete target for future mechanistic and translational studies. These findings may guide research aimed at developing circuit-level interventions and more precise behavioral assays.
Funding: This work was supported by grants and fellowships from the Wellcome Trust, the Simons Foundation, the National Institutes of Health (grant R00NS128075), a Simons Foundation Autism Research Initiative Pilot Grant, the University of Minnesota Clinical and Translational Science Institute, and a Sloan Research Fellowship.
About this genetics and Autism research news
Author: Alexandra Smith
Source: University of Minnesota
Contact: Alexandra Smith – University of Minnesota
Image: The image is credited to Neuroscience News
Original Research: Closed access. “A common computational and neural anomaly across mouse models of autism” by Jean-Paul Noel et al., published in Nature Neuroscience.
Abstract (summary)
A common computational and neural anomaly across mouse models of autism
Computational psychiatry indicates that individuals with autism spectrum disorder often update expectations inflexibly. Using high-yield rodent psychophysics, extensive behavioral modeling, and brain-wide single-cell extracellular recordings, this study evaluated whether mice carrying different genetic perturbations associated with autism display the same computational anomaly and which neurophysiological features are shared across genotypes.
Mice with mutations in Fmr1, Cntnap2, or Shank3B exhibited a reduced updating of priors during decision-making. Relative to animals that adjusted priors flexibly, these models showed a shift in the encoding of prior information away from sensory cortices and toward frontal cortices. Frontal brain regions in the autism models contained more neural units encoding deviations from long-run priors, while sensory responses did not differentiate expected from unexpected stimuli. These results indicate that distinct genetic causes of autism can converge on common neurophysiological and behavioral phenotypes.