Single-cell and spatial transcriptomics reveal mesenchymal stem cell–associated neuroimmune and neuronal signaling patterns after ischemic stroke

Abstract

Ischemic stroke induces coordinated yet spatially heterogeneous neuroimmune and neuronal responses that shape secondary injury and tissue remodeling. However, how intranasal mesenchymal stem cell administration is associated with these cell type-specific responses remains incompletely understood. Here, we integrated single-cell RNA sequencing of the ipsilateral cerebrum on days 1, 7, and 14 after middle cerebral artery occlusion/reperfusion with spatial transcriptomic profiling on day 7. A total of 55,520 high-quality single-cell transcriptomes were obtained from control, middle cerebral artery occlusion/reperfusion, and mesenchymal stem cell-treated samples. In this exploratory dataset, the middle cerebral artery occlusion/reperfusion samples showed relatively greater representation of inflammatory and antigen-presenting microglial states and elevated expression of selected microglial state-associated genes, including Slfn4 , Apoe, and Gpnmb. In the corresponding exploratory comparison, the mesenchymal stem cell-treated sample showed relatively greater representation of homeostatic-associated and Lgals3/Spp1 -enriched microglial states and lower expression of several inflammatory-associated genes. Spatial mapping further revealed regionally restricted expression of microglial and neuronal markers within the injured tissue. Excitatory neurons displayed preliminary mesenchymal stem cell-associated transcriptional patterns involving mitochondrial transcripts, oxidative phosphorylation, synaptic signaling, calcium-related functions, and apoptosis/autophagy-associated pathways. CellChat analysis computationally predicted relatively high levels of tumor necrosis factor, C–X–C motif chemokine ligand, and complement signaling in the middle cerebral artery occlusion/reperfusion sample, whereas the mesenchymal stem cell-treated sample had relatively high levels of predicted activity of neuregulin, pleiotrophin and selected semaphorin. Together, these integrated datasets provide a hypothesis-generating framework linking microglial state remodeling, neuronal metabolic and synaptic programs, and predicted neuroimmune communication patterns after ischemic stroke. Because each omics condition or time point was represented by one biological sample, all the transcriptomic findings should be considered preliminary and exploratory and should be validated in independently replicated cohorts.