South Korean study reveals opposing molecular brain patterns in autism mouse models

Researchers in South Korea have identified two distinct molecular patterns in autism mouse models, offering new insights into the diverse genetic underpinnings of autism spectrum disorder and highlighting potential variations in drug responses.

Researchers in South Korea have identified two broad and opposing molecular patterns in autism mouse models, a finding that could help make sense of why autism spectrum disorder is linked to so many different risk genes. The work, led by Professor Kim Eunjoon at the Institute for Basic Science, analysed more than 1,000 mouse brain transcriptomes and suggests that genetically distinct models may converge on a smaller set of shared brain states rather than follow entirely separate biological paths.

According to the report, the team examined RNA sequencing data from the prefrontal cortex of 17 engineered mouse lines carrying autism-related mutations. They found that one group showed reduced activity in genes tied to communication between nerve cells alongside higher activity in genes involved in chromatin regulation and RNA processing, while the second group showed the reverse pattern. The same broad split appeared in several analyses, including gene expression, RNA splicing and networks of genes that tend to work together.

The researchers also found that the pattern was not fixed by genotype alone. In seven mouse lines, males and females with the same mutation fell into different groups, and some models shifted category as the animals developed. The distinction was also weaker in the hippocampus than in the prefrontal cortex. In parallel, single-cell analysis of about one million nuclei from 205 mice suggested that the opposing patterns reflected coordinated changes across multiple brain-cell types, rather than a single neuronal population.

The study further found that the two molecular groups responded differently to fluoxetine and lithium, two drugs that have been tested in autism-related mouse studies but are not established treatments for autism itself. In one group, the drugs moved some gene sets closer to control patterns, while the other showed more mixed effects. The researchers say the findings may help explain why experimental drug responses vary across models, although they caution that gene-expression changes do not necessarily translate into behavioural benefit.

To test whether the patterns might extend beyond mice, the team compared their findings with transcriptomic data from autistic and neurotypical people. They reported two human subgroups with opposing synaptic gene activity, though the human results were not identical and immune-related differences were more prominent. Earlier research cited in the background to the study has already pointed to converging molecular and network-level pathways in autism, including synaptic and immune signalling, and the new work adds another layer to that picture by suggesting that distinct mutations may still fall into a limited number of recurring brain states.