Spatiotemporal transcriptome and scRNA sequencing analysis reveals that IKFZ1‐mediated microglia underlies a therapy for intracerebral hemorrhage

Dear Editor, After intracerebral hemorrhage (ICH), significant inflammatory response and neuronal damage are evident.1 Inflammatory factors induce neuronal apoptosis, which is the main cause of neuronal dysfunction.2 Although exosomes (Exo) are being studied as a treatment for neuronal dysfunction, their crucial mechanisms have remained unknown.3, 4 This study investigated the essential mechanisms by which Exo alleviates neuronal dysfunction post-ICH. First, we confirmed that post-Exo, the neuronal dysfunction and brain inflammation of ICH-rats were significantly improved (Figure S1). Subsequently, brain tissues were collected (Figure S2A–C) for single-cell sequencing (scRNA-Seq) and spatial transcriptome (ST). Following the annotation of ST (Figure 1A,B), pseudo-temporal analysis of the peri-hematoma area (Figure 1C,D) revealed a decrease in the expression of pyroptosis-related genes (V) post-Exo. Subsequent KEGG analysis of different time-points ST (Figure 1E; Table S1) indicated that post-Exo, the enrichment scores of immune-activations and apoptosis-related pathways decreased, while those of synaptic plasticity-related pathways increased. The scRNA-Seq results identified 12 cell types (Figure 1F; Table S2). KEGG and GO analysis (Figure 1G,H; Tables S3–S4) showed that Exo reduces immune activation and boosts neural protection pathways. Immune changed most significantly post-Exo (Figure S3C,D). Moreover, the integration of ST and scRNA-Seq indicated a notable clustering of immune in the peri-hematoma area (Figure 1I). Subtype analysis of scRNA-Seq identified 29 subtypes (Figure S4A, Table S5). Immune is categorized as MDM, DC, microglia, and neutrophils (Figure S4B,C). Microglia were categorized as MPtn, MIkzf1, MResident, and Mfew (Figure S4D–E). KEGG analysis showed significant enrichment of neurodevelopment pathways in the MResident and MPtn, with immune regulation pathways enriched in the MIkzf1 (Figure S4F, Table S6). MIkzf1 and MPtn predominated in the ICH and ICH+Exo groups, respectively (Figure 2A,B). Subpopulations of microglia in the brain were analyzed using immunofluorescence and flow cytometry (Figure 2C–E) to assess the impact of Exo. An increased co-expression of Ptn-Iba1&CD11b post-Exo, whereas a heightened co-expression of Ikzf1&Iba1&CD11b was observed in the ICH group. To elucidate the primary driving genes in microglia, we employed SCENIC analysis (Figure 2F) and constructed a gene-regulatory network specific to the MIkzf1 based on the top five transcripts (Figure S5A, Table S7). Through GO analysis of these transcripts (Figure S5B–E), Ikzf1 and its target genes exhibit a notable correlation with immune cell differentiation and activation (Figure S5F, Table S8). These findings emphasize the pivotal role of Ikzf1 in governing the immune response facilitated by microglia. Pseudo-time analysis indicates that the developmental trajectories of MIkzf1 and MPtn split into two branches from MResident (Figure 2G; Figure S5G). Genes on developmental trajectories with distinct trends are divided into four groups and subjected to GO analysis. The results show that genes biased towards branch 1 are enriched in various immune-related pathways. Genes biased towards branch 2 are enriched in neuroprotective function pathways (Figure 2H; Table S9). Moreover, the expression of pathways related to cell differentiation and activation increases along branch 1, while it decreases along branch 2 and is predominantly enriched in the MIkzf1 (Figure S5H,I). These findings suggest that the MIkzf1 and MPtn delineate distinct pathways of microglial post-ICH, with the MIkzf1 displaying a heightened affinity for immune-related pathways. To identify key target genes for enhancing the ICH prognosis through post-Exo, we analyzed the miRNA in Exo and identified 178 miRNAs targeting 346 genes (Figure S6A–C, Table S10). GO analysis enriched in pathways related to myeloid cell differentiation all contain the Ikzf1 (Figure S6D, Table S11), indicating that Exo may modulate the transition of microglia subclasses by influencing critical genes like Ikzf1. Annotate subclasses in peri-hematoma area on STs and analyze cell-chat (Figure 3A,B; Figure S7A). The cell-chat in scRNA-Seq (Figure 3C,D; Figure S7B) and STs (Figure 3E; Figure S7C) indicate a notable elevation in the PTN pathway post-Exo, with particular emphasis on the PTN-PTPRZ1 axis. Gene co-expression and immunofluorescence (Figure 3F–G) reveal an enhanced co-localization of PTN-PTPRZ1 post-Exo. The impact of Ikzf1 on the microglial state was explored by a virus to silence Ikzf1 in the brain and Bv-2 cells (Figure S8A,B). Immunofluorescence revealed the co-localization of Iba1&Ikzf1 was reduced, whereas the co-localization of Iba1&Ptn increased (Figure S8C,D), post-Si-Ikzf1. Activation of Nlrp3-inflammasomes triggers pyroptosis and the secretion of pro-inflammatory cytokines.5-7 Si-Ikzf1 treatment markedly decreased the co-expression of Iba1&Gsdmd (Figure 3H). Furthermore, it upregulated the expression of Ptn in brain and Bv-2 cells while simultaneously downregulating the levels of Nlrp3, IL-18, and IL-1β proteins (Figure 3I; Figure S8E). The expression of Nlrp3 inflammasome-related proteins, including Asc, cleaved-caspase-1, Gsdmd, IL-1β, and IL-18, was significantly inhibited (Figure 3J,K; Figure S9, S10). These findings indicate that Si-Ikzf1 suppresses the excessive activation of the Nlrp3 inflammasome, consequently alleviating inflammation in the ICH model. Microglia pyroptosis can result in cell membrane perforation, elevation of inflammatory factors in the microenvironment while triggering neuronal apoptosis.2, 8 TEM revealed a reduction in the formation of Gsdmd-N pores in Bv-2 cells and the brain (Figure 4A,B) post-Si-Ikzf1. To assess the influence of pro-inflammatory factors released from pyroptosis-microglia on neuronal apoptosis, the conditioned medium from LPS-BV-2 cells was utilized in primary-mouse neurons and Sh-Sy5y cells (Figure S11). Flow cytometry (Figure 4C; Figure S12) and apoptosis staining (Figure 4D) demonstrated induction of apoptosis by this conditioned medium, which was subsequently reversed post-Si-Ikzf1. Furthermore, Si-Ikzf1 treatment decreased the ratio of Bax/Bcl-2 and the cleaved caspase-3, caspase-9, and PARP1 in Sh-Sy5y cells and the brain (Figure 4E,F; Figures S13, S14). TUNEL and co-immunofluorescence test (Figure S15A; Figure 4G) supported the findings. These data indicate that Si-Ikzf1 not only promotes the transition from MIkzf1 to MPtn but also significantly reduces neuronal apoptosis in the microenvironment. MRI revealed a significant reduction in the volume of the peri-hematoma area post-Si-Ikzf1 (Figure 4H). The mNSS and Foot-fault testing indicated that Si-Ikzf1 partially reversed the sensorimotor deficits post-ICH (Figure S15A,B). MWM test results confirmed that Si-Ikzf1 alleviated cognitive impairments in ICH-rats (Figure 4I–L). TEM showed that Si-Ikzf1 improved neuronal demyelination (Figure S14D,E). Patch-clamp recordings revealed that Si-Ikzf1 enhanced long-term potentiation in the hippocampus CA3-CA1 of ICH-rats (Figure 4M, Table S14). The Ikzf1-SIRT1 axis regulates macrophage activation through AMPK signalling.9 Our study showed that Si-Ikzf1 reduced Ikzf1 expression while increasing SIRT1 and p-AMPK expression in LPS-Bv-2 cells and ICH-brain (Figure S15F,G). These results suggest that the Ikzf1-SIRT1 axis influences microglial function via the AMPK pathway. Overall, downregulating Ikzf1 promotes the differentiation of post-ICH microglia toward a neuroprotective phenotype, reducing inflammation, pyroptosis, and neuronal apoptosis. Ikzf1 represents a promising therapeutic target for ICH, providing new insights into neuroinflammation and neuronal injury treatment. Ruxiang Xu designed the experiments. Wenqiao Qiu, Jie Mei, and Jie Tian performed bioinformatic analysis. Mingjun Gao, Lili Guo, and Tao Xu performed confocal microscopy. Wei Liu, Jianwei Zhu, and Yi Zhang analyzed histological data. Huan Xiong, Yao Xie, and Xiaolin Hou performed miRNA sequencing and scRNA-sequencing. Xinda Li, Yangyang Wang, Anguo Wu, and Mingjun Gao performed brain MRI experiments. Wenqiao Qiu and Jie Mei wrote the manuscript and edited by Lulin Huang. All authors read and approved the manuscript. The biospecimens for this study were prepared at the Basic Medical Experimental Platform of the University of Electronic Science and Technology of China, ensuring compliance with animal experimental ethics. The authors declare no conflict of interest. This work was supported by the National Natural Science Foundation of China (project nos. 82171355, 81971295, 81671189, 82271105). National Key Research and Development Program of China (2023YFF1204200, 23ZYZYTS0271). Special project of Sichuan Province for central guidance of local scientific and technological development (2023ZYD0059). Ethics approval was obtained from the Medical Ethics Committee of Sichuan Provincial People's Hospital (approval number: 2022–154). All sequencing data are available from the corresponding author (R.X.) upon a reasonable request. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.