Single-Cell RNA Sequencing Characterizes Microglial Phenotypic Heterogeneity Across Different Stroke Subtypes.

Microglia-mediated neuroinflammation is a key contributor to secondary brain injury across diverse cerebrovascular pathologies, yet the precise transcriptomic diversity and temporal evolution of microglial states across different stroke subtypes remain to be fully elucidated. We utilized single-cell RNA sequencing to profile the temporal transcriptomic landscape of the immune microenvironment in mouse models of ischemic stroke (IS), hemorrhagic stroke (HS), and chronic cerebral hypoperfusion (CCH) at 3, 7 and 14 days post-injury. Our data showed a "temporal mismatch" in immunological trajectories between acute and chronic insults. Specifically, IS and HS triggered an immediate, burst-like microglial activation as early as 3 days, the CCH model exhibited a significant latency window, maintaining a predominantly homeostatic profile until a profound transcriptomic transition toward active phenotypes occurred at 14 days. Notably, we identified a prominent S100A8⁺/CHIL3⁺ inflammatory cluster enriched during the hyperacute phase of HS, which may represent an early inflammatory signature associated with intraparenchymal hemorrhage. Interestingly, despite these early divergent responses, the 14-day microglial landscapes across the three models showed a putative shared chronic activation pattern characterized by SPP1⁺ DAM-like signatures and MHC-II-associated antigen-presentation features. Furthermore, we identified a pre-existing stress-responsive signature (Hspa1a⁺/ Hspa1b⁺) in steady-state white matter microglia. These findings provide a high-resolution roadmap of the divergent and convergent immune signatures across cerebrovascular disease spectra, offering hypothesis-generating insights for the development of stage-specific and etiology-tailored immunotherapeutic interventions. Overall, our pooled-library scRNA-seq analysis provides a descriptive and hypothesis-generating framework for stroke-associated microglial and supported by selected tissue-level validation.