Professors Guoqiang Bi and Beiming Liu of the University of Science and Technology of China, together with teams from the Hefei Comprehensive National Science Center's Institute of Artificial Intelligence and the Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences, have made a major advance in large-scale 3D microscopy of biological tissue. The team developed the world's fastest subcellular-resolution 3D imaging technology for small animals, enabling efficient mapping of the peripheral nervous system at fine scale. The results were published in Cell on July 10.
3D visualization of peripheral nerve structures in the mouse body
The peripheral nervous system, as the body's "Internet of Things", carries the bidirectional communication and regulation tasks between the brain and all organs: on the one hand, it transmits movement instructions and regulates critical functions such as breathing and heartbeat; on the other hand, it real-time feedbacks sensory signals such as pain and temperature to the central nervous system for processing, thereby coordinating the activities of various tissues and organs. Mapping the fine connectivity of the peripheral nervous system throughout the body is key to understanding its complex functional mechanisms and related disease mechanisms.
For a long time, scientists' understanding of the overall structure of the peripheral nervous system relied on millimeter-resolution anatomical studies. Although advances in three-dimensional optical microscopy over the past decade have enabled micron-scale mapping of mesoscopic neural networks across the brain, whole-body studies of the peripheral nervous system still face technical bottlenecks. Existing advanced imaging technologies struggle to achieve both high resolution and high speed. Even after whole-body tissue clearing, resolving the complex long-range pathways of the peripheral nervous system throughout a mouse at subcellular resolution remains challenging.
The research team previously developed a new synchronous fly-scanning technology (VISoR), which combines thick slicing and clearing of large biological samples for three-dimensional microscopic imaging. VISoRCombines high speed, high resolution, and scalability, enabling sub-micron resolution imaging of mouse whole-brain samples within 1.5 hours (Hao Wang, Qingyuan Zhu, et al., "National Science Review", 2019), and after further optimization, it achieved micron-resolution 3D imaging and single-neuron fiber tracing of the macaque whole brain for the first time (Fang Xu, Yan Shen, Lufeng Ding, Chaoyu Yang, et al., "Nature Biotechnology", 2021). However, this whole-brain imaging strategy, which involves slicing before clearing, is not suitable for whole-body mouse samples. Unlike the relatively dense and homogeneous brain, the whole-body tissues of mice have high heterogeneity, containing diverse tissue types and irregular structures, making them prone to tissue dislocation and loss during the slicing process, thus hindering complete reconstruction.
To address this technical challenge, the team proposedThe strategy of "sample in situ sectioning + sectioned surface 3D imaging" and the development of the blockface-VISoR imaging system integrated with a precision vibration sectioning device, and the supporting sample preparation process of whole-body clearing and hydrogel embedding for mice, ARCHmap. The core of this technical process is to perform 3D imaging only on the surface of the sample block to a depth of about 600 microns each time, then automatically excise the imaged 400-micron-thick sample, and cycle this process until the sample imaging is complete. Then, using an automated stitching algorithm, the overlapping areas of about 200 microns between adjacent slices are seamlessly stitched and reconstructed in 3D. Since the imaging depth for each scan is only a few hundred microns, the light scattering effect after tissue clearing is weak, so high-resolution imaging can be achieved. Based on this strategy, researchers have established an optimized technical process,Completed uniform subcellular resolution 3D imaging of the entire body of adult mice within 40 hours, acquiring about 70 TB of raw image data per channel. So far, data from dozens of mice have been collected, totaling over 4 PB.
blockface-VISoR imaging system
Because the sample-preparation method preserves fluorescence well, ARCHmap–blockface-VISoR is compatible with fluorescent proteins expressed by common transgenic and neurotropic viral labeling methods, as well as immunofluorescence and other labeling techniques. By combining these labeling and imaging methods, the researchers resolved the fine structures and single-fiber projection pathways of multiple types of peripheral nerves throughout the mouse body. For the first time, the study reveals segmental projection patterns of individual spinal neurons, clarifies organ-specific vascular innervation patterns of sympathetic nerves throughout the body, and maps both the overall projection architecture and complex single-fiber pathways of the vagus nerve.
Mouse whole body blockface-VISoR imaging process and surrounding nerve mesoscopic structure analysis
Researchers say this breakthrough technology helps establish new paradigms for studying peripheral nerve connectomics and understanding neural regulation structures. It also has important applications in developmental biology, systemic anatomy, and biomedicine. Moreover, there is still room for improvement and optimization. The next step is to use dual or multi-camera imaging for simultaneous multi-channel image acquisition to improve data collection efficiency and explore its application in imaging larger-scale biological samples.
Discussion between mentors and team members
Reviewers of the journal "Cell" highly praised this work, pointing out that "these analyses produced strikingly detailed data at both the population and single-cell levels. Importantly, new insights emerged from this initial exploration," and commented, "This is an interesting work, beautifully illustrated, and the methodology shows great potential."
Team members conducting experiments
Meiyu Shi, a specially appointed associate researcher at the University of Science and Technology of China, doctoral students Yuchen Yao, Miao Wang, and master's student Qi Yang are the co-first authors of the paper. Professor Guoqiang Bi is the corresponding author, with special researcher Xu Cheng, Professor Beiming Liu, and senior engineer Qingyuan Zhu as co-corresponding authors. The University of Science and Technology of China is the primary completion unit for this work, with the Hefei Comprehensive National Science Center's Artificial Intelligence Research Institute and the Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences as the second and third completion units, respectively. This work also received collaborative support from Professor Qu Lei of Anhui University, researcher Xu Fuqiang of the Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences, Professor Zhang Ming of the University of Otago, New Zealand, and Professor Yusuf Ozgur Cakmak. In addition, this research was funded by the National Key Research and Development Program, the National Natural Science Foundation, the Science and Technology Innovation 2030 - "Brain Science and Brain-like Research" major project, the Chinese Academy of Sciences' strategic priority research program, the University of Science and Technology of China Youth Innovation Fund, and the Anhui Provincial Natural Science Foundation.
Research team photo
(From left to right: Zhang Keming, Xu Cheng, Qingyuan Zhu, Guoqiang Bi, Beiming Liu, Shi Meiyu, Yuchen Yao, Miao Wang)
Paper Link:
https://www.cell.com/cell/fulltext/S0092-8674(25)00673-7
Links to experimental techniques and demonstration datasets:
https://mesoanatomy.org/mesomouse
Source: USTC News Network
Photography: Xinyu Zhou
Cover: Xingjiaozhe Lv
Editor: Siming Zeng