洛桑大学Andrea Volterra小组揭示了星形胶质细胞通过专门的小叶结构域整合多个突触。2025年9月24日出版的《细胞》发表了这项成果。

通过结合体积高分辨率电子显微镜和双光子Ca²⁺成像，研究人员表征了与突触界面的星形胶质细胞小叶。这些直径≤250nm的卷曲结构来自星形细胞轴或细胞体，包含表达IP₃受体但不含线粒体的微小内质网小囊，并且通常通过间隙连接形成具有细胞质连续性的结构域。小叶包裹着细胞中90%的突触，单个只包裹10%。通过星形胶质细胞外周微体积的快速成像，研究人员确定了突触诱导的，IP₃R1介导的叶片特异性Ca²⁺事件，并且经常显示单独的起源合并成大的，持久的Ca²⁺升高。使用联合轴突-小叶Ca²⁺成像，该课题组表明这些复杂的事件反映了来自不同神经元的传入输入的整合。星形胶质细胞小叶组织可能通过Ca²⁺信号，在不同的时空尺度上协调多个突触和回路，执行不同于神经元的计算。

据介绍，星形胶质细胞Ca²⁺动力学控制突触回路和行为，但潜在的生物学仍然知之甚少。

附：英文原文

Title: Astrocytes functionally integrate multiple synapses via specialized leaflet domains

Author: Lucas Benoit, Ines Hristovska, Nicolas Liaudet, Pierre-Henri Jouneau, Arnold Fertin, Roberta de Ceglia, David Litvin, Maria Amalia Di Castro, Milica Jevtic, Ioannis Zalachoras, Toko Kikuchi, Ludovic Telley, Matteo Bergami, Yves Usson, Chihiro Hisatsune, Katsuhiko Mikoshiba, Karin Pernet-Gallay, Andrea Volterra

Issue&Volume: 2025-09-24

Abstract: Astrocyte Ca2+ dynamics control synaptic circuits and behavior, yet the underlying biology remains poorly understood. By combining volumetric high-resolution electron microscopy and two-photon Ca2+ imaging, we characterize astrocyte leaflets that interface with synapses. These convoluted structures with ≤250 nm diameter originate from astrocytic shafts or cell bodies, contain minuscule endoplasmic reticulum saccules expressing IP3 receptors but not mitochondria, and are often interconnected via gap junctions forming domains with cytosolic continuity. Leaflets enwrap 90% of synapses in clusters and only 10% individually. By fast imaging of astrocyte peripheral microvolumes, we identify leaflet-specific Ca2+ events that were synaptically induced, IP3R1-mediated, and often displayed separate originations merging into large, long-lasting Ca2+ elevations. Using combined axon-leaflet Ca2+ imaging, we show that these complex events reflect integration of incoming inputs from different neurons. The astrocyte leaflet organization may thus coordinate, via Ca2+ signals, multiple synapses and circuits active at different spatiotemporal scales, executing computations distinct from neurons.

DOI: 10.1016/j.cell.2025.08.036

Source: https://www.cell.com/cell/abstract/S0092-8674(25)01028-1