Turning Neurons On Exposes Genetic Risk for Schizophrenia and Autism

A lot of the genetic code that drives complex mental health conditions sits completely dormant until a neuron fires.

In a new study, by forcing laboratory-grown human brain cells to fire, scientists have mapped the real-time genetic activity of the working brain. The results revealed thousands of genetic variants and structural changes that remain invisible when the brain is at rest.

Static tissue misses the drivers of neuropsychiatric disorders

To understand how conditions such as schizophrenia and autism develop, genomic studies have historically relied on tissue samples from postmortem brains; however, dead brain cells do not fire, communicate, or react. This means traditional functional genomic studies look at a static snapshot of a highly dynamic organ, missing the moments when genetic risk factors misbehave.

Many genetic variants tied to schizophrenia and autism are found in parts of DNA that only switch on during specific active brain states. When a brain is at rest or inactive, these variants remain silent and invisible to researchers.

“A major challenge is that most genome-wide association study risk variants are in noncoding regions and likely act in specific biological contexts; as such, the regulatory effects of these variants may be missed by studies that often use static postmortem brains,” said the authors of the new study.

The study investigated living, active systems, with the goal of forcing living brain cells into action to find the real-time genetic mechanisms behind these neuropsychiatric disorders.

Mapping real-time epigenetic shifts in neuropsychiatric disorders

The team cultured living neurons from 100 individual human stem cell lines, looking specifically at glutamatergic (excitatory) and GABAergic (inhibitory) neurons—the two main components that balance brain signaling. They used a chemical trigger to simulate active brain communication, forcing the cells to fire.

The team monitored the cells at baseline, one hour after stimulation, and six hours after stimulation. They tracked active gene expression alongside epigenetics.

“We found widespread, cell–type–specific activation–induced transcriptional and epigenomic changes,” said the authors.

The structural epigenetic changes captured more disease risk than standard, resting gene expression, accounting for ~25% of schizophrenia heritability, compared to the 10% found in resting cells.

The researchers also successfully mapped complex gene networks linked to autism. To prove their network predictions were accurate, they used CRISPR gene editing to disrupt specific genes and watch the downstream effects.

This approach also led them to find a clear biological defect related to schizophrenia. When activated, neurons derived from schizophrenia patients showed abnormal lipid and cholesterol metabolism, pinning down a functional malfunction that only appeared when the cells were working.

Designing state-dependent therapies

“This large-scale, context-specific single-neuron multiomics study demonstrates that neuronal activation unmasks functional effects of many neuropsychiatric disorders risk variants and genes obscured in resting states or postmortem brains,” said the authors.

The data revealed that the cells' structural gates remain open and accessible long after active gene signaling has faded out. This research also suggests that schizophrenia and autism risk genes distort dynamic processes such as cellular metabolism, which could lead to state-dependent, targeted therapies.

Instead of treating these conditions with broad, static medications, future treatments could target the specific metabolic pathways that misbehave during brain activity.

“Our observed expression increase of cholesterol synthesis genes in schizophrenia neurons suggests that cholesterol-targeting drugs may help improve clinical symptoms of neuropsychiatric disorders,” said the authors.

However, the study noted limitations. Simplified 2D laboratory cultures cannot mimic the complex 3D architecture of a living human brain, and chemical triggers are an imperfect substitute for natural synaptic inputs. Future research must expand these models to track a wider range of kinetic time points and validate these dynamic pathways within intact, living biological systems.

Reference: Liang L, Zhang S, Wang Z, et al. Single-cell multiomics of neuron activation reveals context-specific genetics of brain disorders. Science. 2026. doi: 10.1126/science.adw3949