SPATIOTEMPORAL PROFILING OF POST-STROKE NEUROINFLAMMATION USING WHOLE-BRAIN 3D IMAGING OF MICROGLIA

Neuroinflammation is a hallmark of many neurological diseases, yet its spatial and temporal complexity remains poorly understood due to limitations in current analytical approaches. To address this, we developed a quantitative AI-assisted 3D imaging platform to analyse changes in microglial density and activation in intact mouse brains, applicable to characterising disease models and evaluating therapeutic effects.
Male mice underwent permanent middle cerebral artery occlusion (pMCAO), and brains were collected at 1-, 7-, and 14-days post-infarct (dpi). Whole brains were processed for IBA1 and SM22 immunolabelling or in situ hybridization (ISH) for Hexb and Trem2 mRNA. Samples were cleared, imaged using light-sheet fluorescence microscopy (LSFM), and analysed with a deep learning pipeline. SM22 labelling confirmed successful occlusion.
Using Hexb as a stable microglial marker and Trem2 as an inducible one, we characterised microglial activation over time. Hexb was broadly but sparsely distributed in healthy brain whereas Trem2 signal was largely absent. Robust increase in Hexb positive cells, also upregulating Trem2 was detected in peri-infarct regions at 7 dpi. The progressive microglia activation extended from the infarct zone into ipsilateral and contralateral white matter peaking at 7 dpi, whereas thalamic expression increased further from 7 to 14 dpi. Hexb-positive microglia closely mirrored patterns detected by IBA1 immunohistochemistry.
This study presents a 3D whole-brain imaging framework for spatiotemporal analysis of neuroinflammation. Our results demonstrate progressive microglial activation from the peri-infarct into connected brain regions. The established approach enables high-resolution, brain-wide quantification of dynamic neuroinflammatory responses, offering translational potential across multiple CNS disease models.