The causes of autism spectrum disorders (ASD) are highly complex, but abnormalities in neural circuits are a common feature. A joint research team led by Distinguished Research Fellow Yi-Ping Hsueh of the Institute of Molecular Biology and Associate Research Fellow Chien-Yao Wang of the Institute of Information Science at Academia Sinica has established the BM-auto (Brain Mapping with Auto-ROI correction) system. This system was used to analyze whole-brain circuit abnormalities in mouse models of autism, revealing the important role of the “olfactory cortex” in autism. The latest related findings have recently been published in the internationally renowned journal Molecular Psychiatry. The project was supported by the National Science and Technology Council’s Frontier Projects, Academia Sinica’s Investigator Award, and Thematic Research Programs.
Whole-brain neural circuits are highly complex, requiring the development of fast and precise whole-brain analysis techniques to determine whether and where abnormalities exist. To address this challenge, the team spent seven years developing the BM-auto system. From mouse brain sample processing to whole-brain fluorescence imaging, scanning, and quantification, the system enables rapid analysis of axonal projections and neuronal activity across the entire mouse brain, allowing researchers to understand the overall state of neural circuits. As early as 2024, the team successfully used this system to publish a whole-brain connectivity analysis of the amygdala in an ASD mouse model. To further optimize system performance, they incorporated manually corrected data accumulated over the past five years and introduced AI deep learning techniques to build an automated brain region identification system. This system can quickly and accurately analyze more than 500 brain regions in each mouse brain, generating reliable data on each region.
Using the BM-auto system, the team conducted quantitative analysis of whole-brain fluorescence imaging in three representative autism mouse models. By integrating data from the normal mouse database established by the Allen Institute for Brain Science in the United States, they mapped abnormalities in the whole-brain connectomes of these three autism models. They discovered a common pathology among all three: a significant reduction in specific projection neurons in the olfactory cortex. Further experiments confirmed that, although the three autism mouse models retained the ability to detect odors, they lost the ability to distinguish between different smells. Additionally, by using chemogenetic methods to suppress neuronal activity in the olfactory cortex of wild-type mice, the mice exhibited reduced social interaction (autism-like tendencies).
Moreover, analysis of the “functional connectivity” between the olfactory cortex and other brain regions revealed weakened “inter-regional connections” in autism mouse models. In particular, when exposed to specific odor stimuli, neuronal activity across various brain regions (including the olfactory cortex) in autism mice was generally lower than that in wild-type mice. This indicates that abnormalities in the olfactory cortex not only affect olfactory function but also disrupt information transmission and connectivity with other brain regions. These findings highlight the importance of the olfactory cortex and open new directions for future research.
The team not only elucidated the critical role of the “olfactory cortex” in the pathological mechanisms of autism, but also demonstrated the BM-auto system as a major contributor to this study. By overcoming bottlenecks in traditional whole-brain imaging processing, this system enables rapid and precise analysis of the entire brain. It serves as a powerful tool for autism research and shows broad potential for future applications in the analysis of various neurological disorders.
Media Contact：
Yih-Ru Chen
Program Manager
Department of Life Sciences
National Science and Technology Council
Phone: +886-02-2737-7461
E-mail: cyr202407@nstc.gov.tw
