Lab-grown brain models show patchy development in autism

A dish in a Singapore laboratory is standing in for a place researchers cannot inspect directly: the developing human brain. Teams at ASTAR Genome Institute of Singapore and ASTAR Bioinformatics Institute grew brain organoids from stem cells donated by people with and without autism spectrum disorder, then mapped their cells and their positions as the tissue assembled.

The contrast appeared in the organoids’ architecture. Models from donors without ASD formed the orderly layers typical of the cerebral cortex, the brain’s outer layer. In models from donors with ASD, that organization often broke down in distinct patches, while nearby areas developed normally. The location and extent of the disarray differed between individuals, suggesting that autism may involve different spatial patterns of development rather than one uniform process.

The team traced the problem to radial glial cells. These neural stem cells normally stay connected as they guide newly born neurons into their proper positions. In the ASD organoids, those connections were disrupted: the scaffold became disorganized, and neurons did not form the usual layered structure. The abnormalities remained visible at later stages of development.

The work combines single-cell gene activity mapping with spatial analysis and computational biology. That matters because the study adds the location and arrangement of cells to analyses of the genes and biological pathways involved in ASD. The study adds location and cell-to-cell arrangement to a field that has often focused primarily on the genes and biological pathways associated with ASD; researchers have identified more than a thousand genes linked to the condition.

And so what, concretely? Not a new diagnostic test or treatment today. The immediate result is a sharper laboratory model for comparing developmental patterns between individuals. A*STAR researchers say the framework could eventually improve how ASD is classified and help identify more targeted interventions, but those possibilities remain long-term prospects beyond the current organoid study.