Identification and Validation of Microglia-Associated Genes in Ischemic Stroke Using Single-Cell and Bulk RNA-seq

Identification and Validation of Microglia-Associated Genes in Ischemic Stroke Using Single-Cell and Bulk RNA-seq | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Identification and Validation of Microglia-Associated Genes in Ischemic Stroke Using Single-Cell and Bulk RNA-seq Dongliang Qian, Shuangshuang Lu, Yuanyuan Hu, Bing Leng, Xuanfeng Qin This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-7276448/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 07 Dec, 2025 Read the published version in Molecular Brain → Version 1 posted 7 You are reading this latest preprint version Abstract Ischemic stroke (IS) is an acute cerebrovascular disease characterized by high incidence and mortality. The mechanism of microglia in the pathogenesis of IS remains unclear. This study aimed to explore the key genes related to microglia in IS and their molecular mechanisms in the pathogenesis. In this study, the transcriptome data of IS were retrieved from public databases. Subsequently, candidate genes were identified through the intersection of microglia-related genes (MGGs) obtained via single-cell annotation and High-Dimensional Weighted Gene Co-Expression Network Analysis (hdWGCNA) with differentially expressed genes (DEGs). Next, key genes were determined through protein-protein interaction (PPI) analysis and verification of expression levels. Afterwards, enrichment analysis, variation analysis, construction of regulatory networks, drug prediction, and molecular docking were performed to evaluate the role of key genes in the pathogenesis of IS. Ultimately, the quantitative real-time PCR (qRT-PCR) was applied to confirm the expression levels of DEGs in brain tissues between sham and transient middle cerebral artery occlusion (tMCAO) mice. A total of 1,407 DEGs intersected with 100 MGGs, yielding 51 candidate genes. Subsequently, 3 key genes (Cd14, Csf1, and Tlr2) were successfully obtained. The study revealed that these 3 key genes were co-enriched in 4 pathways, such as leishmania infection and ribosomal, and there were differences in the enriched pathways among groups. Notably, the expression of the 3 key genes was regulated by multiple factors, including 32 microRNAs (miRNAs), such as mmu-miR-3072-5p and mmu-miR-3970, and 7 transcription factors (TFs), such as Sp1 and Nfkb1. Meanwhile, these 3 key genes predicted 8 common drugs. Interestingly, Tlr2 and Adapalene exhibited a strong binding affinity (-9.73 kcal/mol). qRT-PCR analysis revealed significantly elevated mRNA expression levels of Cd14, Csf1, and Tlr2 in tMCAO mice compared to sham-operated controls (p < 0.01). This study identified and validated 3 key genes (Cd14, Csf1, and Tlr2) associated with IS, which may serve as novel targets for IS diagnosis and treatment strategies. Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Figure 8 Figure 9 1 Introduction Ischemic stroke (IS), an acute brain injury caused by interrupted cerebral blood flow, accounts for 87% of strokes and is a leading global cause of death and disability[ 1 ]. Its etiology associated with hypertension, diabetes, and atherosclerosis, while pathogenesis involves complex mechanisms including inflammation, oxidative stress, and blood-brain barrier (BBB) disruption[ 2 ]. Current reperfusion therapies, including thrombolysis and thrombectomy, are limited by narrow time windows (< 4.5 hours) and hemorrhagic complications[ 3 ], while neuroprotective strategies targeting oxidative stress or glutamate excitotoxicity have demonstrated limited clinical translatability due to off-target effects or insufficient efficacy[ 4 ]. Therefore, elucidating the diagnostic criteria, etiological drivers, and pathogenic mechanisms of IS, coupled with identifying diagnostic biomarkers linked to targeted therapeutic pathways, may advance precision medicine approaches to optimize clinical management and outcomes. Microglia, the CNS-resident immune cells, regulate neural homeostasis and neuroinflammation through dynamic polarization into pro-inflammatory (M1) or reparative (M2) phenotypes[ 5 ]. Following stroke, microglia rapidly activate in response to danger signals like ATP, glutamate, and high mobility group box 1(HMGB1), triggering pro-inflammatory responses via receptors such as purinergic (P2X, P2Y), TLRs, and inflammasomes such as NLRP3, which drive cytokine release and exacerbate neuronal damage[ 5 ]. Concurrently, they undergo metabolic shifts toward glycolysis and lipid droplet accumulation, supporting phagocytosis and proliferation but potentially leading to dysfunctional foam cells in chronic stages [ 6 , 7 ]. Despite advances, the mechanistic basis through which MRGs coordinate phenotypic transitions remains incompletely defined. This knowledge gap is compounded by the paucity of integrated multi-omic datasets that could enable systematic discovery of MRG-associated therapeutic targets and stage-specific disease biomarkers.Elucidating MRGs could unveil novel strategies to modulate microglia responses, improving IS outcomes. Single-cell RNA sequencing (scRNA-seq) has revolutionized IS research by resolving cellular heterogeneity and dynamic functional states at unprecedented resolution. This technology enables precise mapping of transcriptional profiles across diverse cell types, including microglia, neurons, and vascular endothelial cells, unveiling their roles in neuroinflammation, blood-brain barrier (BBB) disruption, and tissue repair. Recent studies employing scRNA-seq in aged murine IS models revealed dynamic shifts in brain immune cells post-stroke, identifying a potentially stroke-specific microglial subset (MG6) and functionally distinct myeloid subpopulations, providing critical insights for therapies targeting immune subsets[ 8 ]. For example, integrated spatial transcriptomics uncovered GALECTIN-mediated microglia-astrocyte crosstalk promoting recovery. Recent studies utilizing scRNA-seq have further elucidated synergistic neuroprotective mechanisms involving KBA and Z-GS in IS, identifying Spp1 as a pivotal mediator of their interaction, which may guide the development of targeted therapies against this key molecular hub[ 9 ]. However, challenges persist in translating single-cell insights into clinical applications, necessitating multi-omics integration and functional validation to bridge molecular discoveries with therapeutic innovation. In this study, publicly available murine single-cell datasets were used to identify microglia in IS. Cell-cell communication and pseudotime trajectory analyses were employed to map their interaction networks and activation dynamics. hdWGCNA screened MGGs, which were intersected with IS DEGs. Key genes were prioritized via PPI networks and expression validation, followed by functional enrichment, molecular regulatory network analysis, drug prediction, molecular docking, and qRT-PCR validation in tMCAO and sham mouse brain tissues. This approach provides novel insights into microglia-related mechanisms for IS diagnostic and therapeutic targets. 2 Results 2.1 A total of 18 types of cells were identified in single cells First, the raw data of the single-cell dataset GSE174574 was subjected to QC (Quality Control) processing for subsequent analysis. Figure 1 a-b showed the data before and after QC processing. Before processing, there were 18,676 genes and 58,523 cells. After QC, 18,676 genes and 58,025 cells were altogether selected for subsequent analysis. Second, the top 2,000 HVGs were identified. The top 10 genes with the greatest variability included S100a8, S100a9, Hbb-bs, Hbb-a1, and Hba-a2. (Fig. 1 c). Third, the PCA results indicated that there were no obvious outlier samples in the data. However, the significance decreased after PC = 21, and the curve in the PC scree plot tended to plateau at PC = 21. Therefore, 21 principal components were selected for further analysis (Fig. 1 d-e). After that, UMAP (Uniform Manifold Approximation and Projection) clustering divided the cells into 18 cell types (Fig. 1 f). On the basis of the expression intensities of marker genes (Fig. 1 g), the cell clusters in the disease group and the control group were annotated as 18 cell types: Vascular smooth muscle cells (SMC), Perivascular fibroblast-like cells (FB), Central nervous system (CNS), Central associated macrophages (CAM), Monocyte-derived cells (MdC), Venous endothelial cells (vEC), Capillary endothelial cells (capEC), Arterial endothelial cells (aEC), Pericytes (PC), Choroid plexus capillary endothelial cells (CPC), Ependymocytes (EPC), Microglia (MG), Neutrophils (NEUT), Astrocytes (ASC), Lymphocytes (LYM), Oligodendrocytes (OLG), Neural progenitor cells (NPC), and Red blood cells (RBC) (Fig. 1 h). 2.2 Pseudotime analysis of microglia and cell communication analysis Prior to pseudotime analysis, microglia were subdivided into 5 subpopulations (Fig. 2 a). Of the 5 subpopulations of cells, one of them accounted for the largest proportion in the control group, and the other 4 subpopulations accounted for the largest proportion in the disease group (Fig. 2 b). Pseudotime analysis showed that microglia developed sequentially into different branches from their developmental starting positions, and passed through a total of 7 different developmental stages (Fig. 2 c). Signal exchanges between subgroups of microglia and other cells were more frequent (Fig. 2 d), and the intensity of signal exchanges between subgroups MG5 and MG3 was higher than that of other subgroups (Fig. 2 e). 2.3 A total of 100 MGGs were obtained When the number of module genes was configured to 50, the optimal soft threshold of 3 was selected to construct a hierarchical clustering tree (Fig. 3 a-b). Subsequently, the main genes of 2 gene modules, which were significantly associated with microglia, were screened out. These 2 modules were the blue-colored module and the cyan-colored module respectively (Fig. 3 c). Finally, the correlations between the 2 gene modules and the other 18 modules were obtained (Fig. 3 d). The gene modules of Red blood cells (RBC) and Neural progenitor cells (NPC) both had relatively high average expression levels regarding module eigen-gene characteristics. Ultimately, 100 MGGs in the 2 gene modules were obtained. 2.4 Functional enrichment of 51 candidate genes Differential expression analysis revealed that there were 1,407 DEGs between the IS group and the control group. Among them, in the IS group, 1,086 genes were up-regulated genes, and 321 genes were identified as down-regulated genes (Fig. 4 a). The volcano plot marked the top 10 genes with the most significant up-regulation and down-regulation (ranked from high to low according to |log 2 FC|). In addition, a heatmap was used to illustrate the expression profiles of the above-mentioned genes (Fig. 4 b). Subsequently, an intersection analysis was performed on the 1,407 DEGs and the 100 MGGs, and finally 51 candidate genes were identified (Fig. 4 c). Subsequently, an enrichment analysis was carried out to understand the signaling pathways involved by the candidate genes. The candidate genes were significantly enriched in 571 GO terms (adj.p < 0.05), comprising 517 biological processes (BPs), 13 cellular components (CCs), and 41 molecular functions (MFs) (Fig. 4 d) ( Additional file 1 ). In particular, the top 3 BP terms included leukocyte migration, myeloid leukocyte migration, and leukocyte chemotaxis. Among the CCs, terms such as membrane raft, membrane microdomain, and endocytic vesicle played prominent roles. Meanwhile, regarding molecular functions (MFs), they were mainly enriched in cytokine activity, cytokine receptor binding, and chemokine activity. The GO analysis findings showed that the candidate genes were crucial for functions such as clearing necrotic tissue, initiating repair, and promoting nerve regeneration in IS disease. In addition, the KEGG enrichment analysis of the candidate genes demonstrated that among the top 15 significantly enriched pathways, pathways such as coronavirus disease-mus musculus (COVID-19), phagosome-mus musculus, cytokine-cytokine receptor interaction-mus musculus, and rheumatoid arthritis-mus musculus were significantly associated with the candidate genes (Fig. 4 e ) Additional file 2 ). This suggested that the candidate genes exerted crucial functions in the onset, progression, and prognosis of IS. 2.5Cd14, Csf1 and Tlr2 were identified as key genes for IS Subsequently, a PPI network consisting of 255 interaction relationships corresponding to 47 candidate genes was constructed (Fig. 5 a), and 4 genes formed isolated targets. Among the network, Tnf, Ccl2, Fcgr3, Ccl3, and Lgals3 had frequent protein-level interactions with other genes. After that, these candidate genes were integrated into four algorithms of the cytoHubba plugin. Cd14, Csf1, and Tlr2 were obtained from the intersection of the top 10 genes in each algorithm (Fig. 5 b). Then, in the IS groups and control groups of GSE58720 and GSE202659, the expression levels of Cd14, Csf1, and Tlr2 increased significantly (p < 0.05), and a consistent expression trend was witnessed in the two datasets (Fig. 5 c-d). This indicated their potential value in the diagnosis of IS, and they were regarded as key genes. 2.6 GSEA and GSVA of Cd14, Csf1 and Tlr2 Among the top 5 up-regulated and down-regulated pathways that were significantly enriched in Cd14, Csf1, and Tlr2, the pathways that were commonly enriched by the 3 genes included cytokine-cytokine receptor interaction, leishmania infection, ribosome, and toll-like receptor signaling pathway (Fig. 6 a-c). Tlr2 and Cd14 were co-enriched in the calcium signaling pathway, long-term potentiation, oxidative phosphorylation, and toll-like receptor signaling pathway. The co-enrichment suggested that Cd14, Csf1, and Tlr2 might participate in disease regulation in areas such as inflammation and immune regulation, energy and substance metabolism, and neural function regulation. GSVA analysis showed that among the pathways significantly and differentially enriched between the disease group and the control group were interferon gamma response, interferon alpha response, TNF-α signaling via NF-κB, IL-6 JAK-STAT3 signaling, inflammatory response, allograft rejection, hedgehog signaling, TGF-β signaling, and G2/M checkpoint, etc. These pathways play roles in the development of IS mainly through aspects including immune and inflammatory regulation, regulation of cell biological processes, and metabolic regulation (Fig. 6 d-f). 2.7 Cd14, Csf1 and Tlr2 were regulated by multiple factors Further research on the regulatory factors of Cd14, Csf1, and Tlr2 revealed a total of 7 TFs that regulate the key genes (Fig. 7 a). Among them, Sp1 was observed to target all 3 key genes. Additionally, Csf1 was regulated by 1,054 miRNAs, Cd14 by 35 miRNAs, and Tlr2 by 1 miRNA, such as mmu-miR-3072-5p, mmu-miR-3970, mmu-miR-6906-3p, mmu-miR-7041-3p, mmu-miR-7652-3p, mmu-miR-3572-5p, mmu-miR-7044-5p, and mmu-miR-1927. A regulatory network was constructed based on the TFs, key genes, and the 20 miRNAs with the lowest predicted binding p-values. The network showed that a total of 32 miRNAs and TFs jointly regulate the key genes, with 43 types of interactions among them (Fig. 7 b). This indicates that Cd14, Csf1, and Tlr2 were regulated by multiple factors. Meanwhile, genes with similar functions to the key genes were predicted. Different interaction patterns were formed among the 20 genes, and these interactions were associated with functions such as tumor necrosis factor production, tumor necrosis factor superfamily cytokine production, cellular response to molecule of bacterial origin, and pattern recognition receptor signaling pathway (Fig. 7 c). 2.8 Drug prediction and molecular docking of Cd14, Csf1 and Tlr2 The drug-prediction results showed that 6 drugs were predicted for Tlr2, such as Tuberculin purified protein derivative; 1 drug, Atibuclimab, was predicted for Csf1; and 1 drug, Bombesin, was predicted for Cd14 (Fig. 8 a). The molecular formulas and structures of the 8 drugs were shown in Table 1 . Among the 8 drugs, 6 drugs could not be downloaded in SDF format or failed in docking. Finally, only 2 drugs were subjected to molecular docking. The free-binding energy between Golotimod and Tlr2 was − 3.96 kcal/mol, and the free-binding energy between Adapalene and Tlr2 was − 9.73 kcal/mol. Therefore, Adapalene showed a very good binding affinity with Tlr2. The results of the molecular docking were shown in Fig. 8 b. Amino acid residues such as lysine (LYS) and glutamine (GLN) were likely to be involved in the interaction with the ligand. 2.9 Validation of Cd14, Csf1 and Tlr2 The gene expression difference was veriffed by qRT-PCR measurement of 6 samples of tMCAO and sham mice from Shang Hai Jiao Tong University and Fudan university. The results showed that Cd14, Csf1 and Tlr2 were highly expressed in tMCAO brain tissues. The results were consistent with the TCGA database (Fig. 9 ). 3 Discussion The pathogenesis of IS is intricately linked to microglial activation, which modulates neuroinflammation and neuronal damage through immune signaling pathways[ 2 ]. This study integrated transcriptomic datasets of tMCAO mice from the GEO database to explore the regulatory roles of MGGs. Through integrated analysis of single-cell and bulk RNA sequencing data, 51 candidate genes were identified. Subsequent functional enrichment analysis and PPI network construction pinpointed three key genes: Cd14, Csf1, and Tlr2. Furthermore, GeneMANIA analysis and GSEA elucidated the regulatory relationships of these key genes and identified potential drug targets, providing novel insights for the diagnosis and treatment of IS. The Cluster of Differentiation 14 (Cd14) gene is located on chromosome 18 in mice. Similar to its human counterpart, the murine Cd14 gene comprises multiple exons and introns[ 10 ]. This structural feature allows the gene to undergo alternative splicing during transcription, generating multiple mRNA isoforms that may translate into protein isoforms with distinct functional properties. As a pattern recognition receptor, Cd14 specifically recognizes pathogen-associated molecular patterns and plays a pivotal regulatory role in inflammatory responses. Upon recognition of pathogens by cell-surface Cd14, immune cells release inflammatory cytokines that initiate anti-pathogen inflammatory cascades[ 11 , 12 ]. Notably, Cd14 has also been implicated in IS pathogenesis. Cd14 upregulation on activated microglia drives neuroinflammation through MyD88/NF-κB-mediated pathways, inducing iNOS/NO overproduction and TNF-α/IL-1β release, which reciprocally enhance Cd14 activity and neuronal apoptosis in IS [ 13 ]. Beyond local injury, Cd14 + extracellular vesicles from activated monocytes/microglia propagate systemic inflammation, correlating with transient ischemic attack(TIA) diagnosis and cardiovascular risk [ 14 , 15 ]. Therapeutic targeting via Cd14 silencing attenuates neurotoxicity in preclinical models but risks impairing TLR4-TRIF-dependent bacterial clearance [ 12 ]. Precision strategies, such as temporal modulation or compartment-specific targeting (EV neutralization), combined with Tlr2/4 inhibitors, may balance anti-inflammatory efficacy and microbial defense. The Colony Stimulating Factor 1 (Csf1) gene is located on mouse chromosome 3. Through alternative splicing, it generates multiple mRNA isoforms encoding distinct Csf1 glycoprotein variants. These proteins exert their biological effects by specifically binding to the Csf-1 receptor (Csf-1R) on the cell surface[ 16 ]. In IS, Csf1 exhibits dual-phase roles: acutely, it recruits monocytes and activates microglia, driving neuroinflammation and oxidative injury [ 17 , 18 ]; however, delayed Csf1 administration enhances neuroprotection by suppressing pro-inflammatory cytokines and restoring CREB-dependent synaptic plasticity[ 19 , 20 ]. Csf1 also promotes neural repair through neural stem cell proliferation and angiogenesis, though elevated Csf1 levels correlate with stroke risk, potentially via pathological vascular interactions[ 18 , 21 ]. For instance, Csf1 overexpression exacerbates BBB disruption via sLOX-1 upregulation, reversible by dual inhibition[ 22 ]. Therapeutic Csf1R blockade (e.g., PLX5622) mitigates chronic neuroinflammation but risks impairing homeostatic microglial functions[ 17 ]. Precision strategies, such as temporally restricted inhibition (acute phase) or cell-specific targeting (e.g., neuronal Csf1), may dissociate detrimental synaptic pruning from beneficial tissue repair, balancing neuroprotection with physiological microglial activity[ 23 ]. The Toll-Like Receptor 2 (Tlr2) gene resides on mouse chromosome 4. Through alternative splicing during transcription, this gene generates multiple mRNA isoforms that may encode protein variants with distinct functional properties. Tlr2 encodes a transmembrane protein-Tlr2 capable of either specifically recognizing pathogen-associated molecular patterns or interacting with downstream signaling molecules to activate immune pathways. Tlr2 is a transmembrane receptor mediating neuroinflammation in IS via MyD88-dependent DAMP recognition[ 24 ]. Post-ischemia, Tlr2 activation on microglia and neurons triggers divergent pathways: (1) neuronal MyD88/JNK-AP1 signaling induces caspase-3-dependent apoptosis through calcium dysregulation[ 25 ]; (2) microglial Tlr2/Sphk1 axis amplifies IL-1β/TNF-α release via S1P-mediated feedback, exacerbating neuroinflammation[ 26 ]. Paradoxically, Tlr2 deficiency shifts ischemic cell death from apoptosis to necrosis, underscoring its dual role in death modality regulation [ 25 ]. Clinically, elevated blood Tlr2 mRNA correlates with severe neuroinflammation and poor prognosis[ 27 ]. Pharmacological Tlr2/4 inhibition (e.g., JLX001) attenuates NF-κB activation, microglial hyperreactivity, and oxidative stress, preserving neuronal survival[ 28 ]. However, complete Tlr2 blockade risks compromising antimicrobial defenses, necessitating spatiotemporal therapeutic strategies—early-phase inhibition (0–72 h) to suppress neuroinflammation versus later-phase restoration to support repair processes. Our findings revealed that the Cd14/Csf1/Tlr2 genes collectively demonstrated significant enrichment in six key pathways: cytokine-cytokine receptor interaction, Leishmania infection, ribosome biogenesis, Toll-like receptor signaling pathway, interferon gamma response, and Hedgehog signaling pathway. Cytokine-cytokine receptor interaction pathway serves as a nexus for amplifying ischemic damage through pro-inflammatory cascades. Tlr2 and Cd14 jointly potentiate cytokine release (e.g., TNF-α, IL-1β) via MyD88/NF-κB signaling, while Csf1 modulates macrophage polarization to regulate cytokine balance[ 29 , 30 ]. Notably, m6A hypomethylation in IS patients enhances cytokine receptor transcription, exacerbating neuroinflammation[ 29 ]. Targeting this axis may require dual inhibition of upstream DAMPs recognition (via Tlr2/Cd14) and downstream Csf1-driven macrophage recruitment. Ischemia induces dynamic ribosome reprogramming to prioritize stress-responsive protein synthesis. Upregulated ribosomal proteins (e.g., RPS23) facilitate translation of repair factors like HSP70 in astrocytes, while OGFOD1-mediated hydroxylation links ribosomal stress to the UPR activation[ 31 , 32 ]. Csf1 may indirectly regulate ribosomal activity by enhancing microglial phagocytosis of damaged neurons, thereby reducing proteotoxic burden. Therapeutic strategies targeting ribosome quality control (e.g., FG4592) could synergize with CSF1R inhibitors to restore proteostasis. Tlr2 and Cd14 form a co-receptor complex that amplifies DAMP recognition, driving MyD88-dependent neurotoxicity and Sphk1-mediated cytokine storms[ 33 , 34 ]. Csf1 intersects this pathway by sustaining microglial survival, which perpetuates Tlr2/Cd14 signaling. Pharmacological disruption of Tlr4/NF-κB (e.g., NXT capsules) shows promise in attenuating gut-brain axis inflammation, suggesting combinatory approaches targeting Tlr2-Cd14-Csf1 crosstalk may enhance efficacy[ 35 ]. IFN-γ exacerbates ischemic injury by skewing microglia toward M1 polarization, a process potentiated by Tlr2-mediated STAT1 activation and Csf1-dependent macrophage accumulation [ 36 ]. Paradoxically, IFN-γ also primes antigen presentation for tissue repair. Spatiotemporal modulation—suppressing early IFN-γ while promoting its delayed immunoregulatory effects—could exploit this duality. The Sonic Hedgehog (Shh) pathway counterbalances neuroinflammation by promoting M2 microglial polarization and angiogenesis[ 37 ]. Csf1 synergizes with Shh via PDGFA-mediated fibroblast activation, facilitating fibrotic scar formation. Tlr2 inhibition may further enhance Shh-driven repair by reducing pro-inflammatory TGF-β1 suppression[ 38 ]. Small-molecule Shh agonists (e.g., NBP) represent viable adjuvants to CSF1R-targeted therapies. While no direct evidence links this pathway to IS, its enrichment suggests shared immune evasion mechanisms. Our analysis delineates a regulatory triad involving miR-3072-5p, miR-3970, miR-1927, NF-κB (Nfkb1/Rela), which orchestrates vascular repair, neuroinflammation, and cellular stress responses in IS.miR-3072-5p acts as a brake on angiogenesis by targeting VEGF’s 3'UTR, suppressing its translation under physiological conditions [ 39 ]. Remote ischemic preconditioning (rIPC) alleviates spinal ischemia through miR-3072-5p downregulation, which derepresses VEGF to enhance perfusion and neuronal survival. This mechanism may extend to IS, where miR-3072-5p silencing could amplify the collateral circulation in penumbral regions. However, excessive VEGF elevation risks BBB disruption, necessitating spatiotemporal control via nanoparticle-delivered antagomirs.Ischemia-triggered IKK activation phosphorylates IκBα, liberating Rela (p65) to translocate into the nucleus and initiate transcription of TNF-α, IL-1β, and iNOS [ 40 , 41 ]. Pharmacological inhibitors (e.g., DZSM) and siRNA strategies (e.g., p65/CMI) effectively block Rela’s nuclear import, attenuating M1 microglial polarization and neuronal apoptosis[ 41 , 42 ]. Paradoxically, TLR2 activation by Pam2CSK4 engages MyD88-dependent NF-κB signaling to induce a "hybrid" microglial phenotype—simultaneously enhancing debris clearance (via IL-1β) and tissue repair (via IL-10) [ 43 ]. This duality suggests NF-κB’s role is phase-dependent: early inhibition mitigates acute inflammation, while timed activation in subacute phases may promote resolution. Despite their co-enrichment in IS-related pathways, no direct evidence links these miRNAs to IS pathophysiology. Given that miR-3970’s has predicted binding sites in NF-κB inhibitors (e.g., IκBα), and miR-1927’s potential regulation of TLR2, further studies should explore their roles in fine-tuning neuroinflammatory cascades. Our computational drug prioritization identified Adapalene as the top candidate targeting Tlr2, demonstrating superior binding affinity and multimodal neuroprotective potential. As a third-generation retinoid, Adapalene suppresses Tlr2/NF-κB-driven neuroinflammation by inhibiting lipoxygenase-mediated arachidonic acid metabolism while concurrently activating RAR-β to enhance neuronal antioxidant defenses, as evidenced in ALS models where its nanoformulation (Adap-NPs) crosses the blood-brain barrier; prolongs survival, and mitigates oxidative apoptosis[ 44 , 45 ]. This dual action—attenuating MyD88-dependent cytokine storms and stabilizing vascular integrity—aligns with ischemic stroke pathophysiology, particularly in subacute phases where controlled Tlr2 inhibition may balance inflammatory resolution and angiogenesis. While other candidates (e.g., Pam2CSK4) exhibit Tlr2 agonism, their thrombotic risks and lack of CNS specificity limit translational feasibility[ 46 ]. Future studies should optimize Adapalene’s spatiotemporal delivery using microglia-targeted nanocarriers and validate its efficacy in preclinical stroke models, particularly in combination with VEGF-enhancing strategies to synergize vascular repair and neuroprotection. Other compounds—S-(Dimethylarsenic) cysteine, Pam2CSK4, tuberculin purified protein derivative, recombinant OspA Lyme disease vaccine, Golotimod, Atibuclimab, and Bombesin—have been minimally studied in the context of neurological and vascular diseases, warranting further investigation. To corroborate our bioinformatic predictions, we performed qRT-PCR analysis on ipsilateral cortical tissues from C57BL/6 mice subjected to tMCAO with 3-day reperfusion. qRT-PCR results demonstrated significantly higher expression of Csf1, Tlr2, and Cd14 in ischemic brains compared to sham-operated controls, consistent with the bioinformatic analysis. This study employed bioinformatics approaches to comprehensively analyze IS and successfully identified three microglia-related key genes: Cd14, Csf1, and Tlr2, which were further validated in an animal model via qRT-PCR. While the current research provides novel bioinformatics perspectives and evidence for pivotal genes in IS, enhancing the credibility of these candidates necessitates subsequent experimental investigations in animal and cellular models to elucidate their specific biological roles and functional mechanisms in IS pathophysiology. 4 Materials and Methods 4.1 Data origin In this study, the datasets were retrieved from the Gene Expression Omnibus (GEO) database ( https://www.ncbi.nlm.nih.gov/geo/ ). The dataset GSE58720 (platform: GPL10787) was utilized for the training set, consisting of 3 brain tissue samples from middle cerebral artery occlusion (MCAO) mice and 3 samples from sham-operated mice (control group). The GSE202659 dataset (platform: GPL24247) served as the validation set, containing 3 MCAO mice brain tissue samples and 3 sham-operated control samples. Additionally, the single-cell dataset GSE174574 (platform: GPL21103) included 3 MCAO mice brain tissue samples and 3 sham-operated control samples. 4.2 Processing of the scRNA-seq data In the GSE174574 dataset, the Seurat package (v 5.1.0) was utilized for quality control (quality control criteria: 200 < nFeature_RNA < 6,000, percent.mt < 15%, nCount_RNA < 20,000, and filtered out cells with fewer than 200 genes and genes covered by fewer than 3 cells)[ 47 ]. Furthermore, the FindVariableFeatures function in the Seurat package (v 5.1.0) was used, and according to the variance stabilization transformation (vst) method, the top 2,000 genes with relatively high coefficients of variation between cells, namely highly variable genes (HVGs) were obtained, and the top 10 HVGs with the most variation were identified by LabelPoints function and labeled in this study. Similarly, principal components analysis of HVGs was performed using the Seurat package (v 5.1.0). RunPCA function, and the Elbowplot and JackStraw function were employed to plot the scree plot to show the contribution of the top-ranked principal components (PCs) to cellular variation and quantify the significance of the PCs, respectively (p < 0.05). After completing the PCA dimensionality reduction, the Seurat package (v 5.1.0) (resolution = 0.2) was utilized to cluster the cells utilizing the uniform manifold approximation and projection (UMAP) method. Subsequently, the FindAllMarkers function of the Seurat package (min.pct = 0.25, logfc.threshold = 0.25, test.use = auc) was employed to detect the differentially expressed genes in each cell cluster. The Marker genes reported in previous studies were used as the primary reference, and the CellMarker dataset ( http://117.50.127.228/CellMarker/ ) was used as the secondary reference to annotate the cell populations[ 48 ]. A bubble plot was drawn to visually display the representation of Marker genes in various cell types. After annotation, a UMAP plot of different cell type groups was drawn for visualization. 4.3 Pseudotime analysis and cellular communication analysis Aiming to understand the differentiation trajectory of microglia during the disease-onset process, the RunPCA, FindNeighbors, and FindClusters functions in the Seurat package (v 5.1.0) were sequentially used to perform dimensionality reduction and clustering on microglia, which were then annotated into different subgroups (p < 0.05, resolution = 0.2). Subsequently, the ggplot2 package (v 3.5.1) was employed to explore the proportional distribution of cells in each subgroup within disease samples and control samples[ 49 ]. Next, the Monocle2 package (v 2.28.0) was utilized to analyze the differentiation trajectory[ 50 ]. In addition, to analyze the communication between microglia and other cells in IS, based on the CellChatDB database ( http://www.cellchat.org/cellchatdb/ ), with the “CellChatDB.mouse” as a reference, the aggregateNet function of the CellChat package (v 1.6.1) was utilized to analyze the cell-to-cell communication of ligand-receptor complexes (mean = trimean, trim = 0.1)[ 51 ]. Moreover, based on functions such as identifyOverExpressedGenes, identifyOverExpressedInteraction, ProjectData, computeCommunProb, filterCommunication, and computeCommunProbPathway, potential ligand-receptor interactions were identified, and an interaction analysis was conducted on the cell subgroups within microglia (p ≤ 0.05, log 2 mean (Molecule 1, Molecule 2) ≥ 0.1). 4.4 Identification of MGGs Furthermore, to obtain genes associated with microglia in IS, for the GSE174574 dataset, the MetacellsByGroups function in the hdWGCNA package (v 0.4.00) was applied to construct metacells (50 cells) based on the microglia subgroups of each sample[ 47 ]. The TestSoftPowers function was employed to select an appropriate soft threshold, and based on this soft threshold, the ConstructNetwork function was utilized to create a weighted co-expression network. Subsequently, the hierarchicalCluster function was applied for hierarchical clustering (gene modules ≥ 50) to obtain co-expressed gene modules. The ModuleEigengenes function was used to calculate the gene correlations within each module, and the ModuleConnectivity function was applied to evaluate the connectivity between modules. Then, the gene expression levels of each module were compared between disease and control samples using Wilcoxon test (p < 0.05). Modules with significantly altered gene expression abundances were selected, and the genes within these modules were regarded as MGGs. 4.5 Uncovering and analyzing the functions of candidate genes To obtain genes that exhibited differential expression in the IS and control groups within the GSE58720 dataset, the limma package (v 3.58.1) was employed to discern DEGs(|log 2 Fold change (FC)| >0.5, adj.p < 0.05)[ 51 ]. Moreover, the volcano plot and heatmap of DEGs were produced via the ggplot2 package (v 3.5.1) and the ComplexHeatmap package (v 1.0.12), respectively[ 49 , 52 ]. The intersection of DEGs and MGGs was determined by means of the ggvenn package (v 0.1.10) ( https://CRAN.R-project.org/package=ggvenn ) to identify candidate genes associated with IS and microglia. Subsequently, the candidate genes were analyzed for Gene ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) analyses (adj.p < 0.05) via the clusterProfiler package (v 4.8.3), thus elucidating the biological functions of the candidate genes[ 53 ]. 4.6 Identification of key genes In order to investigate the protein-level interactions of candidate genes, we utilized the Search Tool for the Retrieval of Interacting Genes/Proteins (STRING) ( https://www.string-db.org ) to construct a PPI network (combined score ≥ 0.4). Subsequently, the PPI network was visualized by the Cytoscape package (v 3.10.2) [ 25 ]. Moreover, we employed the cytoHubba plugin in the Cytoscape package (v 3.10.2) to rank the candidate genes using 4 algorithms (Degree, Edge Percolated Component (EPC), Maximum Neighborhood Component (MNC), and Density of Maximum Neighborhood Component (DMN)). Then, the ggvenn package (v 0.1.10) was used to obtain the intersection of the top 10 genes selected by each of the 4 algorithms to acquire the candidate key genes. After that, the Wilcoxon test was used to contrast the expression abundances of candidate key genes among the IS group and the control group in the GSE58720 and GSE202659 datasets. Genes that had significantly different expression levels (p < 0.05) and consistent trends in both datasets were defined as key genes. 4.7 Gene set enrichment analysis (GSEA) and Gene set variation analysis (GSVA) The GSEA was conducted to elucidate the biological functions of key genes throughout the progression of IS. Initially, the Spearman correlation coefficients among each key gene and all other genes across all samples in the GSE58720 dataset were calculated using the psych package (v 2.4.3) ( https://CRAN.R-project.org/package=psych ), with the results then ordered in descending order. Subsequently, “c2.kegg.v7.5.1.entrez.gmt” was downloaded as a reference gene set from the Molecular Signatures Database (MSigDB) ( https://www.gsea-msigdb.org/gsea/msigdb/index.jsp ) using the msigdbr package (v 7.5.1) [ 54 ]. Following this, the GSEA was performed with the clusterProfiler package (v 4.8.3), with a threshold of |normalized enrichment score (NES)| >1, False discovery rate (FDR) < 0.25, and p < 0.05. The GSVA was performed to elucidate the enriched pathway differences between disease and controls. First, the “h.hallmark.v7.4.symbols.gmt” from the Molecular Signatures Database (MSigDB) ( https://www.gsea-msigdb.org/ ) was utilized as the reference gene set. Subsequently, GSVA was carried out with the GSVA package (v 1.42.0) with a threshold of |t| >2 and p < 0.05[ 55 ]. 4.8 Construction of regulatory networks and GeneMANIA analysis The regulatory mechanisms of key genes can also be probed based on the molecular level. The Transcriptional Regulatory Relationships Unraveled by Sentence-based Text mining (TRRUST) database ( https://www.grnpedia.org/trrust/ ) was utilized to predict the transcription factors (TFs) that regulate key genes. In addition, the miRWalk database ( http://mirwalk.umm.uni-heidelberg.de ) was employed to forecast the miRNAs of key genes (the top 20 miRNAs with the smallest binding p values). Subsequently, the Cytoscape package (v 3.10.2) was adopted to construct the TFs-mRNAs network and the TFs-key genes-miRNAs regulatory network. After that, to reveal the relationships among key genes and predict the functions they participate in, based on the GeneMANIA database ( https://genemania.org/ ), a co-expression network of key genes was developed to explore genes functionally related to key genes and their modes of functional interaction. 4.9 Drug prediction and molecular docking Drugs targeting key genes were predicted through the DrugBank database ( https://go.drugbank.com/ ). The molecular formulas and 2 dimensional (2D) structures of the drugs were retrieved from the PubChem ( https://pubchem.ncbi.nlm.nih.gov ). The drug-key gene network was presented via the Cytoscape package (v 3.10.2). Among the predicted drugs, the one that targets the largest number of key genes and has the highest interaction score was identified as the key drug. Subsequently, convert the 2D structure into a 3D structure in Chem3D package (v 22.0.0)[ 56 ]. Finally, after hydrogenation in Autodock ( http://autodock.scripps.edu/ ), it was saved in PDBQT format. In addition, the 3D configuration of the target protein of the key gene was retrieved using the IBDB database ( https://www.rcsb.org/ ). Subsequently, molecular docking was carried out using the Autodock ( http://autodock.scripps.edu/ ), and the results of molecular docking were visualized using PyMOL software (v 3.0.3) [ 57 ]. A molecular binding energy of ≤ − 5.0 kcal/mol was considered to indicate good binding activity. 4.10 tMCAO tMCAO procedure was performed in mice using an established method[ 58 ]. Briefly, anesthesia was induced and maintained with 1.5% isoflurane (RWD, Shenzhen, China), with body temperature kept normothermic throughout surgery. Following surgical exposure of the common carotid artery (CCA), external carotid artery (ECA), and internal carotid artery (ICA), a silicone-coated 6 − 0 monofilament (Covidien, St. Louis, MO) was inserted into the ECA. This filament was then advanced retrogradely to block the origin of the middle cerebral artery (MCA) for a duration of 90 minutes. Reperfusion commenced upon removal of the filament. Cerebral blood flow (CBF) was continuously assessed using laser Doppler flowmetry (Moor Instruments, Devon, UK). Occlusion was deemed successful if CBF decreased to ≤ 20% of pre-occlusion baseline levels, and reperfusion was confirmed by a return of CBF to ≥ 80% of baseline. 4.11 qRT-PCR For the study, tissue samples from tMCAO (n = 3) and normal samples (n = 3) were collected from Fudan University and were approved by the Institutional Animal Care and Use Committee of Fudan University. Total RNA was extracted from tissues using TRI Reagent (Sigma-Aldrich, St. Louis, MO) via phenol-chloroform phase separation. Reverse transcription was performed using the PrimeScript RT Master Mix (Takara Bio, Dalian, China). PCR amplification was conducted on a StepOnePlus Real-Time PCR System (Thermo Fisher Scientific, Waltham, MA) under the following conditions: initial denaturation at 95°C for 3 min, followed by 40 cycles of denaturation at 95°C for 15 sec and annealing/extension at 60°C for 30 sec. Gene expression levels were normalized to GAPDH, and primer sequences are listed in Additional file 3 . Reverse transcription was carried out with the TaqMan MicroRNA Reverse Transcription Kit (Applied Biosystems, Foster City, CA), followed by TaqMan probe-based quantification on a LightCycler 480 II platform (Roche, Basel, Switzerland). For every biological sample, 3 technical replicates were carried out. In the meantime, our study had been sanctioned via the Ethics Committee of Fudan University. The resulting data underwent statistical analysis and were depicted via Graphpad Prism (v 10.0) [ 59 ]. 4.12 Statistical analysis The statistical analyses were carried out with R software (v 4.3.1), and the notable group disparities were assessed by means of the Wilcoxon test. In qRT-PCR, the Ct values were compared using paired, two-tailed t-tests, which were performed using GraphPad Prism. A p-value < 0.05 was considered statistically significant. Abbreviations Abbreviation Full Form IS Ischemic Stroke MGGs Microglia-Related Genes DEGs Differentially Expressed Genes PPI Protein-Protein Interaction qRT-PCR Quantitative Real-Time PCR tMCAO Transient Middle Cerebral Artery Occlusion scRNA-seq Single-Cell RNA Sequencing hdWGCNA High-Dimensional Weighted Gene Co-Expression Network Analysis GEO Gene Expression Omnibus QC Quality Control HVGs Highly Variable Genes PCA Principal Component Analysis UMAP Uniform Manifold Approximation and Projection GO Gene Ontology KEGG Kyoto Encyclopedia of Genes and Genomes GSEA Gene Set Enrichment Analysis GSVA Gene Set Variation Analysis TFs Transcription Factors miRNAs MicroRNAs CNS Central Nervous System BBB Blood-Brain Barrier CCA Common Carotid Artery ECA External Carotid Artery ICA Internal Carotid Artery CBF Cerebral Blood Flow MCA Middle Cerebral Artery TLR Toll-Like Receptor NF-κB Nuclear Factor Kappa-Light-Chain-Enhancer of Activated B Cells TNF-α Tumor Necrosis Factor Alpha IL-1β Interleukin-1 Beta VEGF Vascular Endothelial Growth Factor Declarations Ethics approval and consent to participate The study was conducted according to the guidelines of the Declaration of Helsinki, and approved by the Institutional Review Board. (protocol code 2025-HSYY-279 and date of approval 2025-03-05). Consent for publication Not applicable. Availability of data and materials The datasets generated and/or analysed during the current study are available in the [GEO] repository, [http://www.ncbi.nlm.nih.gov/geo/] Competing interests The authors declare that they have no competing interests. Funding The research was generously supported by National Natural Science Foundation of China under grant agreement number 8217052080 and Minhang District Natural Science Foundation of Shanghai under grant agreement number 2022MHZ098. Authors' Contributions Conceptualization, DL. Q. and XF.Q.; methodology, DL.Q.; software, DL.Q; validation, DL.Q., SS.L. and YY.H.; formal analysis, DL.Q and SS.L.; investigation, DL.Q.; resources, DL.Q.; data curation, XF.Q.; writing—original draft preparation, DL.Q.; writing—review and editing, DL.Q.; visualization, DL.Q.; supervision, XF.Q.; project administration, XF.Q and B.L.; funding acquisition, XF.Q and B.L. All authors have read and agreed to the published version of the manuscript. Acknowledgments We would like to express our sincere gratitude to all individuals and organizations who supported and assisted us throughout this research. Special thanks to the following authors: Cao Shanshan, Xu Tongtong. In conclusion, we extend our thanks to everyone who has supported and assisted us along the way. Without your support, this research would not have been possible. References Hou S, Zhang Y, Xia Y, Liu Y, Deng X, Wang W, Wang Y, Wang C, Wang G. Global, regional, and national epidemiology of ischemic stroke from 1990 to 2021. Eur J Neurol. 2024;3112:e16481. DeLong JH, Ohashi SN, O'Connor KC, Sansing LH. Inflammatory Responses After Ischemic Stroke. Semin Immunopathol. 2022;445:625–48. Nguyen TN, Abdalkader M, Fischer U, Qiu Z, Nagel S, Chen HS, Miao Z, Khatri P. Endovascular management of acute stroke. Lancet. 2024;40410459:1265–78. Paul S, Candelario-Jalil E. Emerging neuroprotective strategies for the treatment of ischemic stroke: An overview of clinical and preclinical studies. Exp Neurol. 2021;335:113518. Prinz M, Jung S, Priller J. Microglia Biology: One Century of Evolving Concepts. Cell. 2019;1792:292–311. Planas AM. Role of microglia in stroke. Glia 2024, 726:1016–53. Guan X, Zhu S, Song J, Liu K, Liu M, Xie L, Wang Y, Wu J, Xu X, Pang T. Microglial CMPK2 promotes neuroinflammation and brain injury after ischemic stroke. Cell Rep Med. 2024;55:101522. Li X, Lyu J, Li R, Jain V, Shen Y, Del Águila Á, Hoffmann U, Sheng H, Yang W. Single-cell transcriptomic analysis of the immune cell landscape in the aged mouse brain after ischemic stroke. J Neuroinflammation. 2022;191:83. Liu T, Bai M, Liu M, Li T, Liao Y, Zhao C, Yao M, Wang J, Wen A, Ding Y. Novel synergistic mechanism of 11-keto-β-boswellic acid and Z-Guggulsterone on ischemic stroke revealed by single-cell transcriptomics. Pharmacol Res. 2023;193:106803. Sharygin D, Koniaris LG, Wells C, Zimmers TA, Hamidi T. Role of CD14 in human disease. Immunology. 2023;1693:260–70. Reed-Geaghan EG, Savage JC, Hise AG, Landreth GE. CD14 and toll-like receptors 2 and 4 are required for fibrillar A{beta}-stimulated microglial activation. J Neurosci. 2009;2938:11982–92. Regen T, van Rossum D, Scheffel J, Kastriti ME, Revelo NH, Prinz M, Brück W, Hanisch UK. CD14 and TRIF govern distinct responsiveness and responses in mouse microglial TLR4 challenges by structural variants of LPS. Brain Behav Immun. 2011;255:957–70. Zhou M, Wang CM, Yang WL, Wang P. Microglial CD14 activated by iNOS contributes to neuroinflammation in cerebral ischemia. Brain Res. 2013;1506:105–14. Burrello J, Bianco G, Burrello A, Manno C, Maulucci F, Pileggi M, Nannoni S, Michel P, Bolis S, Melli G, et al. Extracellular Vesicle Surface Markers as a Diagnostic Tool in Transient Ischemic Attacks. Stroke. 2021;5210:3335–47. Timmerman N, Waissi F, Dekker M, van de Pol QY, van Bennekom J, Schoneveld A, Klein Avink MJM, de Winter RJ, Pasterkamp G, de Borst GJ, et al. Pre-Operative Plasma Extracellular Vesicle Proteins are Associated with a High Risk of Long Term Secondary Major Cardiovascular Events in Patients Undergoing Carotid Endarterectomy. Eur J Vasc Endovasc Surg. 2021;625:705–15. De I, Nikodemova M, Steffen MD, Sokn E, Maklakova VI, Watters JJ, Collier LS. CSF1 overexpression has pleiotropic effects on microglia in vivo. Glia. 2014;6212:1955–67. Boland R, Kokiko-Cochran ON. Deplete and repeat: microglial CSF1R inhibition and traumatic brain injury. Front Cell Neurosci. 2024;18:1352790. Cao L, Liu D, Karhunen V, Ren Y, Ye D, Gao J, Gill D, Wang M. Circulating macrophage colony-stimulating factor levels and stroke: A Mendelian randomization study. J Stroke Cerebrovasc Dis. 2023;324:107050. Li L, Yerra L, Chang B, Mathur V, Nguyen A, Luo J. Acute and late administration of colony stimulating factor 1 attenuates chronic cognitive impairment following mild traumatic brain injury in mice. Brain Behav Immun. 2021;94:274–88. Luo J, Elwood F, Britschgi M, Villeda S, Zhang H, Ding Z, Zhu L, Alabsi H, Getachew R, Narasimhan R, et al. Colony-stimulating factor 1 receptor (CSF1R) signaling in injured neurons facilitates protection and survival. J Exp Med. 2013;2101:157–72. Shibahara T, Ago T, Tachibana M, Nakamura K, Yamanaka K, Kuroda J, Wakisaka Y, Kitazono T. Reciprocal Interaction Between Pericytes and Macrophage in Poststroke Tissue Repair and Functional Recovery. Stroke. 2020;5110:3095–106. Huang J, Jiang W, Pan J, Xu X. Correlation of sLOX-1 and CSF1 with the pathological progression of hypoxic-ischemic encephalopathy in neonatal rats. Cell Mol Biol (Noisy-le-grand) 2023, 6910:150–4. Wohleb ES, Terwilliger R, Duman CH, Duman RS. Stress-Induced Neuronal Colony Stimulating Factor 1 Provokes Microglia-Mediated Neuronal Remodeling and Depressive-like Behavior. Biol Psychiatry. 2018;831:38–49. Tajalli-Nezhad S, Karimian M, Beyer C, Atlasi MA, Azami Tameh A. The regulatory role of Toll-like receptors after ischemic stroke: neurosteroids as TLR modulators with the focus on TLR2/4. Cell Mol Life Sci. 2019;763:523–37. Josić Dominović P, Dobrivojević Radmilović M, Srakočić S, Mišerić I, Škokić S, Gajović S. Validation and application of caged Z-DEVD-aminoluciferin bioluminescence for assessment of apoptosis of wild type and TLR2-deficient mice after ischemic stroke. J Photochem Photobiol B. 2024;253:112871. Sun W, Ding Z, Xu S, Su Z, Li H. Crosstalk between TLR2 and Sphk1 in microglia in the cerebral ischemia/reperfusion-induced inflammatory response. Int J Mol Med. 2017;406:1750–8. Jaber Al-Mamoori A, Abdulameer Almustafa HMS, Alsaffar Y. Overexpressed Toll-Like Receptor 2 and the Influence on the Severity of Acute Ischemic Stroke. Arch Razi Inst. 2022;776:2379–84. Qiu Y, Yin Q, Fei Y, Li Y, Huang H, Fang W, Shen W, Liang B, Zhu X, Li Y. JLX001 Modulated the Inflammatory Reaction and Oxidative Stress in pMCAO Rats via Inhibiting the TLR2/4-NF-κB Signaling Pathway. Neurochem Res. 2019;448:1924–38. Zhu L, Liu S, Liao F, Yang J, Liang T, Yang Y, Huang X, Gu L, Su L. Comprehensive Analysis of Blood-Based m6A Methylation in Human Ischemic Stroke. Mol Neurobiol. 2023;602:431–46. Ji Y, Chen Y, Tan X, Huang X, Gao Q, Ma Y, Yang S, Yin M, Yu M, Fang C, et al. Integrated transcriptomic and proteomic profiling reveals the key molecular signatures of brain endothelial reperfusion injury. CNS Neurosci Ther. 2024;304:e14483. Xie J, Zhang Y, Li B, Xi W, Wang Y, Li L, Liu C, Shen L, Han B, Kong Y, et al. Inhibition of OGFOD1 by FG4592 confers neuroprotection by activating unfolded protein response and autophagy after ischemic stroke. J Transl Med. 2024;221:248. Shim B, Ciryam P, Tosun C, Serra R, Tsymbalyuk N, Keledjian K, Gerzanich V, Simard JM. RiboTag RNA Sequencing Identifies Local Translation of HSP70 In Astrocyte Endfeet After Cerebral Ischemia. bioRxiv 2024. Liu L, Xu TC, Zhao ZA, Zhang NN, Li J, Chen HS. Toll-Like Receptor 4 Signaling in Neurons Mediates Cerebral Ischemia/Reperfusion Injury. Mol Neurobiol. 2023;602:864–74. Zhang M, Zhang Y, Peng J, Huang Y, Gong Z, Lu H, Han L, Wang D. Gastrodin against oxidative stress-inflammation crosstalk via inhibiting mtDNA/TLR9 and JAK2/STAT3 signaling to ameliorate ischemic stroke injury. Int Immunopharmacol. 2024;141:113012. Li Y, Zhang J, Lei Y, Chang M, Xu J, Tang S. Multi-omics approaches reveal the therapeutic mechanism of Naoxintong capsule against ischemic stroke. J Ethnopharmacol. 2025;343:119435. Im JH, Yeo IJ, Park PH, Choi DY, Han SB, Yun J, Hong JT. Deletion of Chitinase-3-like 1 accelerates stroke development through enhancement of Neuroinflammation by STAT6-dependent M2 microglial inactivation in Chitinase-3-like 1 knockout mice. Exp Neurol. 2020;323:113082. Yang Q, Jiang P, Tang H, Wen J, Zhou L, Zhao Y, Wang L, Wang J, Yang Q. Shh regulates M2 microglial polarization and fibrotic scar formation after ischemic stroke. Neurochem Int. 2024;180:105862. Dai MJ, Gui XX, Jia SM, Lv ST, Dou H, Cui W. Dl-3-n-butylphthalide promotes angiogenesis in ischemic stroke mice through upregulating autocrine and paracrine sonic hedgehog. Acta Pharmacol Sin. 2023;4412:2404–17. Ueno K, Samura M, Nakamura T, Tanaka Y, Takeuchi Y, Kawamura D, Takahashi M, Hosoyama T, Morikage N, Hamano K. Increased plasma VEGF levels following ischemic preconditioning are associated with downregulation of miRNA-762 and miR-3072-5p. Sci Rep. 2016;6:36758. Xu H, Wang Y, Luo Y. OTULIN is a new target of EA treatment in the alleviation of brain injury and glial cell activation via suppression of the NF-κB signalling pathway in acute ischaemic stroke rats. Mol Med. 2021;271:37. Ganbold T, Bao Q, Zandan J, Hasi A, Baigude H. Modulation of Microglia Polarization through Silencing of NF-κB p65 by Functionalized Curdlan Nanoparticle-Mediated RNAi. ACS Appl Mater Interfaces. 2020;1210:11363–74. Ye J, Huang F, Zeng H, Xu X, Wu G, Tian S, Zhao J, Zhang W. Multi-omics and network pharmacology study reveals the effects of Dengzhan Shengmai capsule against neuroinflammatory injury and thrombosis induced by ischemic stroke. J Ethnopharmacol. 2023;305:116092. Stirling DP, Cummins K, Mishra M, Teo W, Yong VW, Stys P. Toll-like receptor 2-mediated alternative activation of microglia is protective after spinal cord injury. Volume 137Pt 3. Brain; 2014. pp. 707–23. Medina DX, Chung EP, Teague CD, Bowser R, Sirianni RW. Intravenously Administered, Retinoid Activating Nanoparticles Increase Lifespan and Reduce Neurodegeneration in the SOD1(G93A) Mouse Model of ALS. Front Bioeng Biotechnol. 2020;8:224. Rusu A, Tanase C, Pascu GA, Todoran N. Recent Advances Regarding the Therapeutic Potential of Adapalene. Pharmaceuticals (Basel) 2020, 139. Parra-Izquierdo I, Lakshmanan HHS, Melrose AR, Pang J, Zheng TJ, Jordan KR, Reitsma SE, McCarty OJT, Aslan JE. The Toll-Like Receptor 2 Ligand Pam2CSK4 Activates Platelet Nuclear Factor-κB and Bruton's Tyrosine Kinase Signaling to Promote Platelet-Endothelial Cell Interactions. Front Immunol. 2021;12:729951. Morabito S, Reese F, Rahimzadeh N, Miyoshi E, Swarup V. hdWGCNA identifies co-expression networks in high-dimensional transcriptomics data. Cell Rep Methods. 2023;36:100498. Zheng K, Lin L, Jiang W, Chen L, Zhang X, Zhang Q, Ren Y, Hao J. Single-cell RNA-seq reveals the transcriptional landscape in ischemic stroke. J Cereb Blood Flow Metab. 2022;421:56–73. Li Y, Zhang Q, Yang J, He W, Jiang Y, Chen Y, Wang Y. Metformin combined with glucose starvation synergistically suppress triple-negative breast cancer by enhanced unfolded protein response. Biochem Biophys Res Commun. 2023;675:146–54. Trapnell C, Cacchiarelli D, Grimsby J, Pokharel P, Li S, Morse M, Lennon NJ, Livak KJ, Mikkelsen TS, Rinn JL. The dynamics and regulators of cell fate decisions are revealed by pseudotemporal ordering of single cells. Nat Biotechnol. 2014;324:381–6. Liu Z, Sun D, Wang C. Evaluation of cell-cell interaction methods by integrating single-cell RNA sequencing data with spatial information. Genome Biol. 2022;231:218. Ding W, Goldberg D, Zhou W. PyComplexHeatmap: a Python package to visualize multimodal genomics data. Imeta 2023, 23. Wu T, Hu E, Xu S, Chen M, Guo P, Dai Z, Feng T, Zhou L, Tang W, Zhan L, et al. clusterProfiler 4.0: A universal enrichment tool for interpreting omics data. Innov (Camb). 2021;23:100141. Liberzon A, Birger C, Thorvaldsdóttir H, Ghandi M, Mesirov JP, Tamayo P. The Molecular Signatures Database (MSigDB) hallmark gene set collection. Cell Syst. 2015;16:417–25. Hänzelmann S, Castelo R, Guinney J. GSVA: gene set variation analysis for microarray and RNA-seq data. BMC Bioinformatics. 2013;14:7. Meng Z, Yamazaki T, Sode K. A molecularly imprinted catalyst designed by a computational approach in catalysing a transesterification process. Biosens Bioelectron. 2004;206:1068–75. Abu Hassan A, Hanževački M, Pordea A. Computational investigation of cis-1,4-polyisoprene binding to the latex-clearing protein LcpK30. PLoS ONE. 2024;195:e0302398. Li Y, Liu Z, Song Y, Pan JJ, Jiang Y, Shi X, Liu C, Ma Y, Luo L, Mamtilahun M, et al. M2 microglia-derived extracellular vesicles promote white matter repair and functional recovery via miR-23a-5p after cerebral ischemia in mice. Theranostics. 2022;127:3553–73. Mitteer DR, Greer BD. Using GraphPad Prism's Heat Maps for Efficient, Fine-Grained Analyses of Single-Case Data. Behav Anal Pract. 2022;152:505–14. Table Tables 1 is not available with this version. Additional Declarations No competing interests reported. Supplementary Files Additionalfile1.xlsx Supplementary Materials Additional file 1 Result of Gene Ontology (GO) enrichment analysis Additionalfile2.xlsx Additional file 2 Result of Kyoto Encyclopedia of Genes and Genomes(KEGG) enrichment analysis Additionalfile3.docx Additional file 3 qPR-PCR primer sequence Cite Share Download PDF Status: Published Journal Publication published 07 Dec, 2025 Read the published version in Molecular Brain → Version 1 posted Editorial decision: Revision requested 18 Sep, 2025 Reviews received at journal 18 Sep, 2025 Reviewers agreed at journal 09 Sep, 2025 Reviewers invited by journal 14 Aug, 2025 Editor assigned by journal 04 Aug, 2025 Submission checks completed at journal 04 Aug, 2025 First submitted to journal 02 Aug, 2025 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-7276448","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":501233628,"identity":"ec5ea426-2dba-4fe9-8c25-9a3fdac13bbc","order_by":0,"name":"Dongliang Qian","email":"","orcid":"","institution":"Huashan Hospital, Fudan University","correspondingAuthor":false,"prefix":"","firstName":"Dongliang","middleName":"","lastName":"Qian","suffix":""},{"id":501233629,"identity":"09326898-bd46-4869-854a-eca1529ce7c1","order_by":1,"name":"Shuangshuang Lu","email":"","orcid":"","institution":"Fudan University","correspondingAuthor":false,"prefix":"","firstName":"Shuangshuang","middleName":"","lastName":"Lu","suffix":""},{"id":501233630,"identity":"aaf3ab14-687e-43c2-8157-a7c456d7f01a","order_by":2,"name":"Yuanyuan Hu","email":"","orcid":"","institution":"Huashan Hospital, Fudan University","correspondingAuthor":false,"prefix":"","firstName":"Yuanyuan","middleName":"","lastName":"Hu","suffix":""},{"id":501233631,"identity":"5aee26ec-07a2-461d-8b1f-79e786f0ca4a","order_by":3,"name":"Bing Leng","email":"","orcid":"","institution":"Huashan Hospital, Fudan University","correspondingAuthor":false,"prefix":"","firstName":"Bing","middleName":"","lastName":"Leng","suffix":""},{"id":501233632,"identity":"e7e3deb7-1864-455b-a46c-2ced1518d71f","order_by":4,"name":"Xuanfeng Qin","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA/ElEQVRIiWNgGAWjYDACZgjFD8Tsv//+s+Hh528gToskSJ0ED1uajOSMA8RZBtNy2MagIQG/Ur7jvMekeWruSPBLt18wkOA5z2PAcIDxw8ccPKYf5ks25jn2TEJyzpmCBAOJ2zzmzA3MkjO34dZicJjH8DHQPXUGN3ISDiQY3OaxbDjAxsyLXwsQ/TssYX8jJ7HhQMI5HoMDCQS1GD7mbTssYSCRfpix4cABwlokD/MYG87tOywhcSOHjZmxIZlHcsbBZrx+4Tt/xkzizbfDEvwz0p8BtdjZ8/M3H/zwEY8WhgNwFjB4IYCxAY96FC3sD/CrHAWjYBSMghELAKIRUJAqMYRfAAAAAElFTkSuQmCC","orcid":"","institution":"Huashan Hospital, Fudan University","correspondingAuthor":true,"prefix":"","firstName":"Xuanfeng","middleName":"","lastName":"Qin","suffix":""}],"badges":[],"createdAt":"2025-08-02 07:53:02","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-7276448/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-7276448/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1186/s13041-025-01259-x","type":"published","date":"2025-12-07T15:58:29+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":89633344,"identity":"69eef705-1502-46ef-9b1b-5db6eeecbe78","added_by":"auto","created_at":"2025-08-22 06:57:47","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":5531015,"visible":true,"origin":"","legend":"\u003cp\u003eIS single-cell data preprocessing. \u0026nbsp;\u003cstrong\u003ea:\u003c/strong\u003e nFeature_RNA, nCount_RNA, and percent.mt plots of pre-QC dataset data, \u003cstrong\u003eb:\u003c/strong\u003e nFeature_RNA, nCount_RNA, and percent.mt plots of the dataset data after QC, the abscissa is different grouping information, the ordinate is the count information, and the black dots represent individual cells. \u003cstrong\u003ec:\u003c/strong\u003e Scatter plot of high variation genes, the horizontal axis represents the average expression level of the gene, the vertical axis represents the normalized variance, each dot in the graph represents a gene, and the color and shape of the dot indicate the variability of the gene, \u003cstrong\u003ed: \u003c/strong\u003ePCA Principal Component Analysis Diagram, \u003cstrong\u003ee: \u003c/strong\u003eIdentification scatter plots for available dimensions,\u003cstrong\u003ef:\u003c/strong\u003e UMAP plot of cell cluster classification. \u003cstrong\u003eg:\u003c/strong\u003eExpression of Marker genes in different cell types, the abscissa represents different genes, each gene occupies a row in the graph, and the ordinate represents the different clusters obtained by dimensionality reduction, each clusters occupy a column. \u003cstrong\u003eh:\u003c/strong\u003e UMAP annotation plot of cells of different cell types.\u003c/p\u003e","description":"","filename":"Figure1.png","url":"https://assets-eu.researchsquare.com/files/rs-7276448/v1/ba584d849af69bfdd947307b.png"},{"id":89634412,"identity":"b19e1068-2825-4aeb-93ae-4738e132aad8","added_by":"auto","created_at":"2025-08-22 07:05:47","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":6249611,"visible":true,"origin":"","legend":"\u003cp\u003eResults of the pseudo-time analysis.. \u003cstrong\u003ea:\u003c/strong\u003e Distribution of different subgroups of microglia in disease group and control group, \u003cstrong\u003eb: \u003c/strong\u003eDifference in the proportion of microglia in different samples,\u003cstrong\u003e c: \u003c/strong\u003eDifferentiation trajectory of microglia (from 1-3 to cell differentiation time trajectory, cell state trajectory, cell type trajectory), \u003cstrong\u003ed:\u003c/strong\u003e ligand-receptor interaction point diagram, \u003cstrong\u003ee:\u003c/strong\u003e Number and intensity of interactions between microglia subgroups,\u003c/p\u003e","description":"","filename":"Figure2.png","url":"https://assets-eu.researchsquare.com/files/rs-7276448/v1/603c7abfb31895d40efba839.png"},{"id":89633340,"identity":"934d54b0-50cf-4418-98c4-a6c5481c7478","added_by":"auto","created_at":"2025-08-22 06:57:47","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":2616866,"visible":true,"origin":"","legend":"\u003cp\u003eWCGNA's analysis results. \u003cstrong\u003ea:\u003c/strong\u003e Scale-free index and average connectivity plot for soft thresholds, \u003cstrong\u003eb: \u003c/strong\u003ehierarchical clustering tree, \u003cstrong\u003ec:\u003c/strong\u003e main candidate characteristic genes of different modules, \u003cstrong\u003ed: \u003c/strong\u003ecorrelation between candidate trait genes in the module,\u003c/p\u003e","description":"","filename":"Figure3.png","url":"https://assets-eu.researchsquare.com/files/rs-7276448/v1/055abc3692e514957fd21227.png"},{"id":89633346,"identity":"e2553a18-b26e-4411-8c78-c8435d75b008","added_by":"auto","created_at":"2025-08-22 06:57:47","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":3736986,"visible":true,"origin":"","legend":"\u003cp\u003eScreening and identification of candidate genes. \u003cstrong\u003ea:\u003c/strong\u003e Volcanic plot of differentially expressed genes distribution between IS and Control, abscissa Log2FC, ordinate -Log10 (adjP. Value), each dot represents a gene; The transverse reference line represents - Log10(0.05)=1.3, the longitudinal reference line represents log2FC = ±0.5, and the genes in the upper right corner are up-regulated differentially expressed genes (indicated by red), the genes in the upper left corner are down-regulated differentially expressed genes (represented by blue), and the rest of the genes are genes with no significant statistical significance (represented by gray). The genes labeled in the figure were the top 10 up-regulated genes and the top 10 down-regulated genes with the largest log2(FoldChange)||, i.e., the largest fold of difference. \u003cstrong\u003eb:\u003c/strong\u003e Differential gene expression heat map, annotation bar above, orange for disease samples, green for control samples; The ordinates in the heat map represent genes, with red being the most expressed genes and blue being the low expression genes. c\u003cstrong\u003e: \u003c/strong\u003eVenn diagram of candidate genes. d\u003cstrong\u003e: \u003c/strong\u003eGO enrichment analysis results. \u003cstrong\u003ee\u003c/strong\u003e: KEGG enrichment analysis results\u003c/p\u003e","description":"","filename":"Figure4.png","url":"https://assets-eu.researchsquare.com/files/rs-7276448/v1/e94969fe5408b8817c80fbec.png"},{"id":89633351,"identity":"9e196ab1-aa17-4093-b610-902a8edb420f","added_by":"auto","created_at":"2025-08-22 06:57:47","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":3476082,"visible":true,"origin":"","legend":"\u003cp\u003ePPI network construction and expression level verification of candidate genes. \u003cstrong\u003ea: \u003c/strong\u003ePPI network of candidate genes, the shade of color represents the degree value of the gene, the darker the color, the higher the degree value of the gene, the more genes that interact with the gene, the lighter the color, the smaller the degree value of the gene, and the fewer genes that interact with the gene. \u003cstrong\u003eb:\u003c/strong\u003e cytoHubba screening Venn diagram, \u003cstrong\u003ec:\u003c/strong\u003e Expression levels of candidate key genes in the IS group and control group in the training set. \u003cstrong\u003ed:\u003c/strong\u003e Expression levels of candidate key genes in the IS and control groups in the validation set\u003c/p\u003e","description":"","filename":"Figure5.png","url":"https://assets-eu.researchsquare.com/files/rs-7276448/v1/1e26fb5dfe61bcf2f97c7d3b.png"},{"id":89633354,"identity":"3a2f75e7-6100-4068-a37e-58f3419dde16","added_by":"auto","created_at":"2025-08-22 06:57:47","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":6066019,"visible":true,"origin":"","legend":"\u003cp\u003eEnrichment analysis of key genes. \u003cstrong\u003ea:\u003c/strong\u003e GSEA analysis of CD14. \u003cstrong\u003eb:\u003c/strong\u003e GSEA analysis results for Csf1. \u003cstrong\u003ec:\u003c/strong\u003e GSEA analysis of Tlr2, the upper part represents the enrichment fraction of the TOP5 gene pathway, the middle part represents the distribution of genes in the reference gene set in the training set, different colors represent different pathways, and the lower part represents the signal-to-noise ratio, where the larger the area, the greater the multiple of the difference. \u003cstrong\u003ed:\u003c/strong\u003e GSVA analysis results for CD14. \u003cstrong\u003ee:\u003c/strong\u003e GSVA analysis results for Csf1. \u003cstrong\u003ef:\u003c/strong\u003e GSVA analysis results of TLR2, the ordinate represents the pathway name of the differential pathway, the abscissa represents the differential pathway GSVA score, the green bar chart represents the down-regulated pathway, and the blue bar chart represents the up-regulated pathway.\u003c/p\u003e","description":"","filename":"Figure6.png","url":"https://assets-eu.researchsquare.com/files/rs-7276448/v1/47997c7747cd1f8069bdbaef.png"},{"id":89634414,"identity":"d028dc01-2de2-494a-b00e-8bdac0195951","added_by":"auto","created_at":"2025-08-22 07:05:47","extension":"png","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":11083011,"visible":true,"origin":"","legend":"\u003cp\u003eConstruction of regulatory networks of key gene-related molecules. \u003cstrong\u003ea:\u003c/strong\u003e Interaction network between key genes and co-expressed genes. \u003cstrong\u003eb:\u003c/strong\u003e Regulatory network of key genes and transcription factors. c\u003cstrong\u003e: \u003c/strong\u003eRegulatory network between biomarkers and TF and miRNAs, with lines indicating regulatory relationships between key genes and TF and miRNAs. Different colors represent different factors, with yellow rectangles for TFs, pink diamonds for miRNAs, and ovals for key genes.\u003c/p\u003e","description":"","filename":"Figure7.png","url":"https://assets-eu.researchsquare.com/files/rs-7276448/v1/e2a2b3d00e320ce7e8016fae.png"},{"id":89633365,"identity":"317cd118-c9e5-4238-a5fc-76e442e87531","added_by":"auto","created_at":"2025-08-22 06:57:48","extension":"png","order_by":8,"title":"Figure 8","display":"","copyAsset":false,"role":"figure","size":4435726,"visible":true,"origin":"","legend":"\u003cp\u003eDrug prediction of key genes. \u003cstrong\u003ea\u003c/strong\u003e: Network diagram of key genes and potential therapeutic drugs. \u003cstrong\u003eb1-2\u003c/strong\u003e: Molecular docking of the predicted drug to a potential target protein of a key gene\u003c/p\u003e","description":"","filename":"Figure8.png","url":"https://assets-eu.researchsquare.com/files/rs-7276448/v1/679095b2f37864aae2b8ce32.png"},{"id":89633353,"identity":"faae98ef-373a-4065-b532-4866974bc3d3","added_by":"auto","created_at":"2025-08-22 06:57:47","extension":"png","order_by":9,"title":"Figure 9","display":"","copyAsset":false,"role":"figure","size":548671,"visible":true,"origin":"","legend":"\u003cp\u003eqRT-PCR detection of key genes\u003c/p\u003e","description":"","filename":"Figure9.png","url":"https://assets-eu.researchsquare.com/files/rs-7276448/v1/7a1fa967e69ae3c19e773c3d.png"},{"id":97725010,"identity":"3de5d1b4-f0c6-426a-a3ce-1d6508ac42df","added_by":"auto","created_at":"2025-12-08 16:14:09","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":44102786,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7276448/v1/00ba893d-24f4-4d37-bf15-0848b92ec0ca.pdf"},{"id":89633341,"identity":"d4026fae-019f-40be-8cef-0dcb669d21df","added_by":"auto","created_at":"2025-08-22 06:57:47","extension":"xlsx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":60963,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eSupplementary Materials\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAdditional file 1\u003c/strong\u003e Result of Gene Ontology (GO) enrichment analysis\u003c/p\u003e","description":"","filename":"Additionalfile1.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-7276448/v1/89db91476579a517f7e8e032.xlsx"},{"id":89633339,"identity":"1bc546c3-6e48-4e91-956e-08f2e6272c8a","added_by":"auto","created_at":"2025-08-22 06:57:47","extension":"xlsx","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":13302,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eAdditional file 2 \u003c/strong\u003eResult of Kyoto Encyclopedia of Genes and Genomes(KEGG) enrichment analysis\u003c/p\u003e","description":"","filename":"Additionalfile2.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-7276448/v1/d28be330ed14fd180ad653f8.xlsx"},{"id":89635584,"identity":"884391b6-c57c-4e6f-8100-707cb8c1551e","added_by":"auto","created_at":"2025-08-22 07:13:47","extension":"docx","order_by":3,"title":"","display":"","copyAsset":false,"role":"supplement","size":11731,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eAdditional file 3\u003c/strong\u003e qPR-PCR primer sequence\u003c/p\u003e","description":"","filename":"Additionalfile3.docx","url":"https://assets-eu.researchsquare.com/files/rs-7276448/v1/60d130c3e7c3162401e3e8da.docx"}],"financialInterests":"No competing interests reported.","formattedTitle":"Identification and Validation of Microglia-Associated Genes in Ischemic Stroke Using Single-Cell and Bulk RNA-seq","fulltext":[{"header":"1 Introduction","content":"\u003cp\u003eIschemic stroke (IS), an acute brain injury caused by interrupted cerebral blood flow, accounts for 87% of strokes and is a leading global cause of death and disability[\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e]. Its etiology associated with hypertension, diabetes, and atherosclerosis, while pathogenesis involves complex mechanisms including inflammation, oxidative stress, and blood-brain barrier (BBB) disruption[\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e]. Current reperfusion therapies, including thrombolysis and thrombectomy, are limited by narrow time windows (\u0026lt;\u0026thinsp;4.5 hours) and hemorrhagic complications[\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e], while neuroprotective strategies targeting oxidative stress or glutamate excitotoxicity have demonstrated limited clinical translatability due to off-target effects or insufficient efficacy[\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e]. Therefore, elucidating the diagnostic criteria, etiological drivers, and pathogenic mechanisms of IS, coupled with identifying diagnostic biomarkers linked to targeted therapeutic pathways, may advance precision medicine approaches to optimize clinical management and outcomes.\u003c/p\u003e\u003cp\u003eMicroglia, the CNS-resident immune cells, regulate neural homeostasis and neuroinflammation through dynamic polarization into pro-inflammatory (M1) or reparative (M2) phenotypes[\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e]. Following stroke, microglia rapidly activate in response to danger signals like ATP, glutamate, and high mobility group box 1(HMGB1), triggering pro-inflammatory responses via receptors such as purinergic (P2X, P2Y), TLRs, and inflammasomes such as NLRP3, which drive cytokine release and exacerbate neuronal damage[\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e]. Concurrently, they undergo metabolic shifts toward glycolysis and lipid droplet accumulation, supporting phagocytosis and proliferation but potentially leading to dysfunctional foam cells in chronic stages [\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e, \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e]. Despite advances, the mechanistic basis through which MRGs coordinate phenotypic transitions remains incompletely defined. This knowledge gap is compounded by the paucity of integrated multi-omic datasets that could enable systematic discovery of MRG-associated therapeutic targets and stage-specific disease biomarkers.Elucidating MRGs could unveil novel strategies to modulate microglia responses, improving IS outcomes.\u003c/p\u003e\u003cp\u003eSingle-cell RNA sequencing (scRNA-seq) has revolutionized IS research by resolving cellular heterogeneity and dynamic functional states at unprecedented resolution. This technology enables precise mapping of transcriptional profiles across diverse cell types, including microglia, neurons, and vascular endothelial cells, unveiling their roles in neuroinflammation, blood-brain barrier (BBB) disruption, and tissue repair. Recent studies employing scRNA-seq in aged murine IS models revealed dynamic shifts in brain immune cells post-stroke, identifying a potentially stroke-specific microglial subset (MG6) and functionally distinct myeloid subpopulations, providing critical insights for therapies targeting immune subsets[\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e]. For example, integrated spatial transcriptomics uncovered GALECTIN-mediated microglia-astrocyte crosstalk promoting recovery. Recent studies utilizing scRNA-seq have further elucidated synergistic neuroprotective mechanisms involving KBA and Z-GS in IS, identifying Spp1 as a pivotal mediator of their interaction, which may guide the development of targeted therapies against this key molecular hub[\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e]. However, challenges persist in translating single-cell insights into clinical applications, necessitating multi-omics integration and functional validation to bridge molecular discoveries with therapeutic innovation.\u003c/p\u003e\u003cp\u003eIn this study, publicly available murine single-cell datasets were used to identify microglia in IS. Cell-cell communication and pseudotime trajectory analyses were employed to map their interaction networks and activation dynamics. hdWGCNA screened MGGs, which were intersected with IS DEGs. Key genes were prioritized via PPI networks and expression validation, followed by functional enrichment, molecular regulatory network analysis, drug prediction, molecular docking, and qRT-PCR validation in tMCAO and sham mouse brain tissues. This approach provides novel insights into microglia-related mechanisms for IS diagnostic and therapeutic targets.\u003c/p\u003e"},{"header":"2 Results","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e\u003ch2\u003e2.1 A total of 18 types of cells were identified in single cells\u003c/h2\u003e\u003cp\u003eFirst, the raw data of the single-cell dataset GSE174574 was subjected to QC (Quality Control) processing for subsequent analysis. Figure\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ea-b showed the data before and after QC processing. Before processing, there were 18,676 genes and 58,523 cells. After QC, 18,676 genes and 58,025 cells were altogether selected for subsequent analysis. Second, the top 2,000 HVGs were identified. The top 10 genes with the greatest variability included S100a8, S100a9, Hbb-bs, Hbb-a1, and Hba-a2. (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ec). Third, the PCA results indicated that there were no obvious outlier samples in the data. However, the significance decreased after PC\u0026thinsp;=\u0026thinsp;21, and the curve in the PC scree plot tended to plateau at PC\u0026thinsp;=\u0026thinsp;21. Therefore, 21 principal components were selected for further analysis (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ed-e). After that, UMAP (Uniform Manifold Approximation and Projection) clustering divided the cells into 18 cell types (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ef). On the basis of the expression intensities of marker genes (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eg), the cell clusters in the disease group and the control group were annotated as 18 cell types: Vascular smooth muscle cells (SMC), Perivascular fibroblast-like cells (FB), Central nervous system (CNS), Central associated macrophages (CAM), Monocyte-derived cells (MdC), Venous endothelial cells (vEC), Capillary endothelial cells (capEC), Arterial endothelial cells (aEC), Pericytes (PC), Choroid plexus capillary endothelial cells (CPC), Ependymocytes (EPC), Microglia (MG), Neutrophils (NEUT), Astrocytes (ASC), Lymphocytes (LYM), Oligodendrocytes (OLG), Neural progenitor cells (NPC), and Red blood cells (RBC) (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eh).\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec4\" class=\"Section2\"\u003e\u003ch2\u003e2.2 Pseudotime analysis of microglia and cell communication analysis\u003c/h2\u003e\u003cp\u003ePrior to pseudotime analysis, microglia were subdivided into 5 subpopulations (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ea). Of the 5 subpopulations of cells, one of them accounted for the largest proportion in the control group, and the other 4 subpopulations accounted for the largest proportion in the disease group (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eb). Pseudotime analysis showed that microglia developed sequentially into different branches from their developmental starting positions, and passed through a total of 7 different developmental stages (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ec). Signal exchanges between subgroups of microglia and other cells were more frequent (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ed), and the intensity of signal exchanges between subgroups MG5 and MG3 was higher than that of other subgroups (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ee).\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec5\" class=\"Section2\"\u003e\u003ch2\u003e2.3 A total of 100 MGGs were obtained\u003c/h2\u003e\u003cp\u003eWhen the number of module genes was configured to 50, the optimal soft threshold of 3 was selected to construct a hierarchical clustering tree (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ea-b). Subsequently, the main genes of 2 gene modules, which were significantly associated with microglia, were screened out. These 2 modules were the blue-colored module and the cyan-colored module respectively (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ec). Finally, the correlations between the 2 gene modules and the other 18 modules were obtained (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ed). The gene modules of Red blood cells (RBC) and Neural progenitor cells (NPC) both had relatively high average expression levels regarding module eigen-gene characteristics. Ultimately, 100 MGGs in the 2 gene modules were obtained.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec6\" class=\"Section2\"\u003e\u003ch2\u003e2.4 Functional enrichment of 51 candidate genes\u003c/h2\u003e\u003cp\u003eDifferential expression analysis revealed that there were 1,407 DEGs between the IS group and the control group. Among them, in the IS group, 1,086 genes were up-regulated genes, and 321 genes were identified as down-regulated genes (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ea). The volcano plot marked the top 10 genes with the most significant up-regulation and down-regulation (ranked from high to low according to |log\u003csub\u003e2\u003c/sub\u003eFC|). In addition, a heatmap was used to illustrate the expression profiles of the above-mentioned genes (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eb). Subsequently, an intersection analysis was performed on the 1,407 DEGs and the 100 MGGs, and finally 51 candidate genes were identified (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ec). Subsequently, an enrichment analysis was carried out to understand the signaling pathways involved by the candidate genes. The candidate genes were significantly enriched in 571 GO terms (adj.p\u0026thinsp;\u0026lt;\u0026thinsp;0.05), comprising 517 biological processes (BPs), 13 cellular components (CCs), and 41 molecular functions (MFs) (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ed) (\u003cb\u003eAdditional file 1\u003c/b\u003e). In particular, the top 3 BP terms included leukocyte migration, myeloid leukocyte migration, and leukocyte chemotaxis. Among the CCs, terms such as membrane raft, membrane microdomain, and endocytic vesicle played prominent roles. Meanwhile, regarding molecular functions (MFs), they were mainly enriched in cytokine activity, cytokine receptor binding, and chemokine activity. The GO analysis findings showed that the candidate genes were crucial for functions such as clearing necrotic tissue, initiating repair, and promoting nerve regeneration in IS disease. In addition, the KEGG enrichment analysis of the candidate genes demonstrated that among the top 15 significantly enriched pathways, pathways such as coronavirus disease-mus musculus (COVID-19), phagosome-mus musculus, cytokine-cytokine receptor interaction-mus musculus, and rheumatoid arthritis-mus musculus were significantly associated with the candidate genes (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ee\u003cb\u003e) Additional file 2\u003c/b\u003e). This suggested that the candidate genes exerted crucial functions in the onset, progression, and prognosis of IS.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec7\" class=\"Section2\"\u003e\u003ch2\u003e2.5Cd14, Csf1 and Tlr2 were identified as key genes for IS\u003c/h2\u003e\u003cp\u003eSubsequently, a PPI network consisting of 255 interaction relationships corresponding to 47 candidate genes was constructed (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ea), and 4 genes formed isolated targets. Among the network, Tnf, Ccl2, Fcgr3, Ccl3, and Lgals3 had frequent protein-level interactions with other genes. After that, these candidate genes were integrated into four algorithms of the cytoHubba plugin. Cd14, Csf1, and Tlr2 were obtained from the intersection of the top 10 genes in each algorithm (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eb). Then, in the IS groups and control groups of GSE58720 and GSE202659, the expression levels of Cd14, Csf1, and Tlr2 increased significantly (p\u0026thinsp;\u0026lt;\u0026thinsp;0.05), and a consistent expression trend was witnessed in the two datasets (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ec-d). This indicated their potential value in the diagnosis of IS, and they were regarded as key genes.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec8\" class=\"Section2\"\u003e\u003ch2\u003e2.6 GSEA and GSVA of Cd14, Csf1 and Tlr2\u003c/h2\u003e\u003cp\u003eAmong the top 5 up-regulated and down-regulated pathways that were significantly enriched in Cd14, Csf1, and Tlr2, the pathways that were commonly enriched by the 3 genes included cytokine-cytokine receptor interaction, leishmania infection, ribosome, and toll-like receptor signaling pathway (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003ea-c). Tlr2 and Cd14 were co-enriched in the calcium signaling pathway, long-term potentiation, oxidative phosphorylation, and toll-like receptor signaling pathway. The co-enrichment suggested that Cd14, Csf1, and Tlr2 might participate in disease regulation in areas such as inflammation and immune regulation, energy and substance metabolism, and neural function regulation. GSVA analysis showed that among the pathways significantly and differentially enriched between the disease group and the control group were interferon gamma response, interferon alpha response, TNF-α signaling via NF-κB, IL-6 JAK-STAT3 signaling, inflammatory response, allograft rejection, hedgehog signaling, TGF-β signaling, and G2/M checkpoint, etc. These pathways play roles in the development of IS mainly through aspects including immune and inflammatory regulation, regulation of cell biological processes, and metabolic regulation (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003ed-f).\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec9\" class=\"Section2\"\u003e\u003ch2\u003e2.7 Cd14, Csf1 and Tlr2 were regulated by multiple factors\u003c/h2\u003e\u003cp\u003eFurther research on the regulatory factors of Cd14, Csf1, and Tlr2 revealed a total of 7 TFs that regulate the key genes (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003ea). Among them, Sp1 was observed to target all 3 key genes. Additionally, Csf1 was regulated by 1,054 miRNAs, Cd14 by 35 miRNAs, and Tlr2 by 1 miRNA, such as mmu-miR-3072-5p, mmu-miR-3970, mmu-miR-6906-3p, mmu-miR-7041-3p, mmu-miR-7652-3p, mmu-miR-3572-5p, mmu-miR-7044-5p, and mmu-miR-1927. A regulatory network was constructed based on the TFs, key genes, and the 20 miRNAs with the lowest predicted binding p-values. The network showed that a total of 32 miRNAs and TFs jointly regulate the key genes, with 43 types of interactions among them (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003eb). This indicates that Cd14, Csf1, and Tlr2 were regulated by multiple factors. Meanwhile, genes with similar functions to the key genes were predicted. Different interaction patterns were formed among the 20 genes, and these interactions were associated with functions such as tumor necrosis factor production, tumor necrosis factor superfamily cytokine production, cellular response to molecule of bacterial origin, and pattern recognition receptor signaling pathway (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003ec).\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec10\" class=\"Section2\"\u003e\u003ch2\u003e2.8 Drug prediction and molecular docking of Cd14, Csf1 and Tlr2\u003c/h2\u003e\u003cp\u003eThe drug-prediction results showed that 6 drugs were predicted for Tlr2, such as Tuberculin purified protein derivative; 1 drug, Atibuclimab, was predicted for Csf1; and 1 drug, Bombesin, was predicted for Cd14 (Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003ea). The molecular formulas and structures of the 8 drugs were shown in \u003cb\u003eTable\u0026nbsp;1\u003c/b\u003e. Among the 8 drugs, 6 drugs could not be downloaded in SDF format or failed in docking. Finally, only 2 drugs were subjected to molecular docking. The free-binding energy between Golotimod and Tlr2 was \u0026minus;\u0026thinsp;3.96 kcal/mol, and the free-binding energy between Adapalene and Tlr2 was \u0026minus;\u0026thinsp;9.73 kcal/mol. Therefore, Adapalene showed a very good binding affinity with Tlr2. The results of the molecular docking were shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003eb. Amino acid residues such as lysine (LYS) and glutamine (GLN) were likely to be involved in the interaction with the ligand.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec11\" class=\"Section2\"\u003e\u003ch2\u003e2.9 Validation of Cd14, Csf1 and Tlr2\u003c/h2\u003e\u003cp\u003eThe gene expression difference was veriffed by qRT-PCR measurement of 6 samples of tMCAO and sham mice from Shang Hai Jiao Tong University and Fudan university. The results showed that Cd14, Csf1 and Tlr2 were highly expressed in tMCAO brain tissues. The results were consistent with the TCGA database (Fig.\u0026nbsp;\u003cspan refid=\"Fig9\" class=\"InternalRef\"\u003e9\u003c/span\u003e).\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003c/div\u003e"},{"header":"3 Discussion","content":"\u003cp\u003eThe pathogenesis of IS is intricately linked to microglial activation, which modulates neuroinflammation and neuronal damage through immune signaling pathways[\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e]. This study integrated transcriptomic datasets of tMCAO mice from the GEO database to explore the regulatory roles of MGGs. Through integrated analysis of single-cell and bulk RNA sequencing data, 51 candidate genes were identified. Subsequent functional enrichment analysis and PPI network construction pinpointed three key genes: Cd14, Csf1, and Tlr2. Furthermore, GeneMANIA analysis and GSEA elucidated the regulatory relationships of these key genes and identified potential drug targets, providing novel insights for the diagnosis and treatment of IS.\u003c/p\u003e\u003cp\u003eThe Cluster of Differentiation 14 (Cd14) gene is located on chromosome 18 in mice. Similar to its human counterpart, the murine Cd14 gene comprises multiple exons and introns[\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e]. This structural feature allows the gene to undergo alternative splicing during transcription, generating multiple mRNA isoforms that may translate into protein isoforms with distinct functional properties. As a pattern recognition receptor, Cd14 specifically recognizes pathogen-associated molecular patterns and plays a pivotal regulatory role in inflammatory responses. Upon recognition of pathogens by cell-surface Cd14, immune cells release inflammatory cytokines that initiate anti-pathogen inflammatory cascades[\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e, \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e]. Notably, Cd14 has also been implicated in IS pathogenesis. Cd14 upregulation on activated microglia drives neuroinflammation through MyD88/NF-κB-mediated pathways, inducing iNOS/NO overproduction and TNF-α/IL-1β release, which reciprocally enhance Cd14 activity and neuronal apoptosis in IS [\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e]. Beyond local injury, Cd14\u0026thinsp;+\u0026thinsp;extracellular vesicles from activated monocytes/microglia propagate systemic inflammation, correlating with transient ischemic attack(TIA) diagnosis and cardiovascular risk [\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e, \u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e]. Therapeutic targeting via Cd14 silencing attenuates neurotoxicity in preclinical models but risks impairing TLR4-TRIF-dependent bacterial clearance [\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e]. Precision strategies, such as temporal modulation or compartment-specific targeting (EV neutralization), combined with Tlr2/4 inhibitors, may balance anti-inflammatory efficacy and microbial defense.\u003c/p\u003e\u003cp\u003eThe Colony Stimulating Factor 1 (Csf1) gene is located on mouse chromosome 3. Through alternative splicing, it generates multiple mRNA isoforms encoding distinct Csf1 glycoprotein variants. These proteins exert their biological effects by specifically binding to the Csf-1 receptor (Csf-1R) on the cell surface[\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e]. In IS, Csf1 exhibits dual-phase roles: acutely, it recruits monocytes and activates microglia, driving neuroinflammation and oxidative injury [\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e, \u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e]; however, delayed Csf1 administration enhances neuroprotection by suppressing pro-inflammatory cytokines and restoring CREB-dependent synaptic plasticity[\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e, \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e]. Csf1 also promotes neural repair through neural stem cell proliferation and angiogenesis, though elevated Csf1 levels correlate with stroke risk, potentially via pathological vascular interactions[\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e, \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e]. For instance, Csf1 overexpression exacerbates BBB disruption via sLOX-1 upregulation, reversible by dual inhibition[\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e]. Therapeutic Csf1R blockade (e.g., PLX5622) mitigates chronic neuroinflammation but risks impairing homeostatic microglial functions[\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e]. Precision strategies, such as temporally restricted inhibition (acute phase) or cell-specific targeting (e.g., neuronal Csf1), may dissociate detrimental synaptic pruning from beneficial tissue repair, balancing neuroprotection with physiological microglial activity[\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e].\u003c/p\u003e\u003cp\u003eThe Toll-Like Receptor 2 (Tlr2) gene resides on mouse chromosome 4. Through alternative splicing during transcription, this gene generates multiple mRNA isoforms that may encode protein variants with distinct functional properties. Tlr2 encodes a transmembrane protein-Tlr2 capable of either specifically recognizing pathogen-associated molecular patterns or interacting with downstream signaling molecules to activate immune pathways. Tlr2 is a transmembrane receptor mediating neuroinflammation in IS via MyD88-dependent DAMP recognition[\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e]. Post-ischemia, Tlr2 activation on microglia and neurons triggers divergent pathways: (1) neuronal MyD88/JNK-AP1 signaling induces caspase-3-dependent apoptosis through calcium dysregulation[\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e]; (2) microglial Tlr2/Sphk1 axis amplifies IL-1β/TNF-α release via S1P-mediated feedback, exacerbating neuroinflammation[\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e]. Paradoxically, Tlr2 deficiency shifts ischemic cell death from apoptosis to necrosis, underscoring its dual role in death modality regulation [\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e]. Clinically, elevated blood Tlr2 mRNA correlates with severe neuroinflammation and poor prognosis[\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e]. Pharmacological Tlr2/4 inhibition (e.g., JLX001) attenuates NF-κB activation, microglial hyperreactivity, and oxidative stress, preserving neuronal survival[\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e]. However, complete Tlr2 blockade risks compromising antimicrobial defenses, necessitating spatiotemporal therapeutic strategies\u0026mdash;early-phase inhibition (0\u0026ndash;72 h) to suppress neuroinflammation versus later-phase restoration to support repair processes.\u003c/p\u003e\u003cp\u003eOur findings revealed that the Cd14/Csf1/Tlr2 genes collectively demonstrated significant enrichment in six key pathways: cytokine-cytokine receptor interaction, Leishmania infection, ribosome biogenesis, Toll-like receptor signaling pathway, interferon gamma response, and Hedgehog signaling pathway. Cytokine-cytokine receptor interaction pathway serves as a nexus for amplifying ischemic damage through pro-inflammatory cascades. Tlr2 and Cd14 jointly potentiate cytokine release (e.g., TNF-α, IL-1β) via MyD88/NF-κB signaling, while Csf1 modulates macrophage polarization to regulate cytokine balance[\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e, \u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e]. Notably, m6A hypomethylation in IS patients enhances cytokine receptor transcription, exacerbating neuroinflammation[\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e]. Targeting this axis may require dual inhibition of upstream DAMPs recognition (via Tlr2/Cd14) and downstream Csf1-driven macrophage recruitment. Ischemia induces dynamic ribosome reprogramming to prioritize stress-responsive protein synthesis. Upregulated ribosomal proteins (e.g., RPS23) facilitate translation of repair factors like HSP70 in astrocytes, while OGFOD1-mediated hydroxylation links ribosomal stress to the UPR activation[\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e, \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e]. Csf1 may indirectly regulate ribosomal activity by enhancing microglial phagocytosis of damaged neurons, thereby reducing proteotoxic burden. Therapeutic strategies targeting ribosome quality control (e.g., FG4592) could synergize with CSF1R inhibitors to restore proteostasis.\u003c/p\u003e\u003cp\u003eTlr2 and Cd14 form a co-receptor complex that amplifies DAMP recognition, driving MyD88-dependent neurotoxicity and Sphk1-mediated cytokine storms[\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e, \u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e]. Csf1 intersects this pathway by sustaining microglial survival, which perpetuates Tlr2/Cd14 signaling. Pharmacological disruption of Tlr4/NF-κB (e.g., NXT capsules) shows promise in attenuating gut-brain axis inflammation, suggesting combinatory approaches targeting Tlr2-Cd14-Csf1 crosstalk may enhance efficacy[\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e]. IFN-γ exacerbates ischemic injury by skewing microglia toward M1 polarization, a process potentiated by Tlr2-mediated STAT1 activation and Csf1-dependent macrophage accumulation [\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e]. Paradoxically, IFN-γ also primes antigen presentation for tissue repair. Spatiotemporal modulation\u0026mdash;suppressing early IFN-γ while promoting its delayed immunoregulatory effects\u0026mdash;could exploit this duality. The Sonic Hedgehog (Shh) pathway counterbalances neuroinflammation by promoting M2 microglial polarization and angiogenesis[\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e]. Csf1 synergizes with Shh via PDGFA-mediated fibroblast activation, facilitating fibrotic scar formation. Tlr2 inhibition may further enhance Shh-driven repair by reducing pro-inflammatory TGF-β1 suppression[\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e]. Small-molecule Shh agonists (e.g., NBP) represent viable adjuvants to CSF1R-targeted therapies. While no direct evidence links this pathway to IS, its enrichment suggests shared immune evasion mechanisms.\u003c/p\u003e\u003cp\u003eOur analysis delineates a regulatory triad involving miR-3072-5p, miR-3970, miR-1927, NF-κB (Nfkb1/Rela), which orchestrates vascular repair, neuroinflammation, and cellular stress responses in IS.miR-3072-5p acts as a brake on angiogenesis by targeting VEGF\u0026rsquo;s 3'UTR, suppressing its translation under physiological conditions [\u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e]. Remote ischemic preconditioning (rIPC) alleviates spinal ischemia through miR-3072-5p downregulation, which derepresses VEGF to enhance perfusion and neuronal survival. This mechanism may extend to IS, where miR-3072-5p silencing could amplify the collateral circulation in penumbral regions. However, excessive VEGF elevation risks BBB disruption, necessitating spatiotemporal control via nanoparticle-delivered antagomirs.Ischemia-triggered IKK activation phosphorylates IκBα, liberating Rela (p65) to translocate into the nucleus and initiate transcription of TNF-α, IL-1β, and iNOS [\u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e, \u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e]. Pharmacological inhibitors (e.g., DZSM) and siRNA strategies (e.g., p65/CMI) effectively block Rela\u0026rsquo;s nuclear import, attenuating M1 microglial polarization and neuronal apoptosis[\u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e, \u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e]. Paradoxically, TLR2 activation by Pam2CSK4 engages MyD88-dependent NF-κB signaling to induce a \"hybrid\" microglial phenotype\u0026mdash;simultaneously enhancing debris clearance (via IL-1β) and tissue repair (via IL-10) [\u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e43\u003c/span\u003e]. This duality suggests NF-κB\u0026rsquo;s role is phase-dependent: early inhibition mitigates acute inflammation, while timed activation in subacute phases may promote resolution. Despite their co-enrichment in IS-related pathways, no direct evidence links these miRNAs to IS pathophysiology. Given that miR-3970\u0026rsquo;s has predicted binding sites in NF-κB inhibitors (e.g., IκBα), and miR-1927\u0026rsquo;s potential regulation of TLR2, further studies should explore their roles in fine-tuning neuroinflammatory cascades.\u003c/p\u003e\u003cp\u003eOur computational drug prioritization identified Adapalene as the top candidate targeting Tlr2, demonstrating superior binding affinity and multimodal neuroprotective potential. As a third-generation retinoid, Adapalene suppresses Tlr2/NF-κB-driven neuroinflammation by inhibiting lipoxygenase-mediated arachidonic acid metabolism while concurrently activating RAR-β to enhance neuronal antioxidant defenses, as evidenced in ALS models where its nanoformulation (Adap-NPs) crosses the blood-brain barrier; prolongs survival, and mitigates oxidative apoptosis[\u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e44\u003c/span\u003e, \u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e45\u003c/span\u003e]. This dual action\u0026mdash;attenuating MyD88-dependent cytokine storms and stabilizing vascular integrity\u0026mdash;aligns with ischemic stroke pathophysiology, particularly in subacute phases where controlled Tlr2 inhibition may balance inflammatory resolution and angiogenesis. While other candidates (e.g., Pam2CSK4) exhibit Tlr2 agonism, their thrombotic risks and lack of CNS specificity limit translational feasibility[\u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e46\u003c/span\u003e]. Future studies should optimize Adapalene\u0026rsquo;s spatiotemporal delivery using microglia-targeted nanocarriers and validate its efficacy in preclinical stroke models, particularly in combination with VEGF-enhancing strategies to synergize vascular repair and neuroprotection. Other compounds\u0026mdash;S-(Dimethylarsenic) cysteine, Pam2CSK4, tuberculin purified protein derivative, recombinant OspA Lyme disease vaccine, Golotimod, Atibuclimab, and Bombesin\u0026mdash;have been minimally studied in the context of neurological and vascular diseases, warranting further investigation.\u003c/p\u003e\u003cp\u003eTo corroborate our bioinformatic predictions, we performed qRT-PCR analysis on ipsilateral cortical tissues from C57BL/6 mice subjected to tMCAO with 3-day reperfusion. qRT-PCR results demonstrated significantly higher expression of Csf1, Tlr2, and Cd14 in ischemic brains compared to sham-operated controls, consistent with the bioinformatic analysis.\u003c/p\u003e\u003cp\u003eThis study employed bioinformatics approaches to comprehensively analyze IS and successfully identified three microglia-related key genes: Cd14, Csf1, and Tlr2, which were further validated in an animal model via qRT-PCR. While the current research provides novel bioinformatics perspectives and evidence for pivotal genes in IS, enhancing the credibility of these candidates necessitates subsequent experimental investigations in animal and cellular models to elucidate their specific biological roles and functional mechanisms in IS pathophysiology.\u003c/p\u003e"},{"header":"4 Materials and Methods","content":"\u003cdiv id=\"Sec14\" class=\"Section2\"\u003e\u003ch2\u003e4.1 Data origin\u003c/h2\u003e\u003cp\u003eIn this study, the datasets were retrieved from the Gene Expression Omnibus (GEO) database (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.ncbi.nlm.nih.gov/geo/\u003c/span\u003e\u003cspan address=\"https://www.ncbi.nlm.nih.gov/geo/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e). The dataset GSE58720 (platform: GPL10787) was utilized for the training set, consisting of 3 brain tissue samples from middle cerebral artery occlusion (MCAO) mice and 3 samples from sham-operated mice (control group). The GSE202659 dataset (platform: GPL24247) served as the validation set, containing 3 MCAO mice brain tissue samples and 3 sham-operated control samples. Additionally, the single-cell dataset GSE174574 (platform: GPL21103) included 3 MCAO mice brain tissue samples and 3 sham-operated control samples.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec15\" class=\"Section2\"\u003e\u003ch2\u003e4.2 Processing of the scRNA-seq data\u003c/h2\u003e\u003cp\u003eIn the GSE174574 dataset, the Seurat package (v 5.1.0) was utilized for quality control (quality control criteria: 200\u0026thinsp;\u0026lt;\u0026thinsp;nFeature_RNA\u0026thinsp;\u0026lt;\u0026thinsp;6,000, percent.mt\u0026thinsp;\u0026lt;\u0026thinsp;15%, nCount_RNA\u0026thinsp;\u0026lt;\u0026thinsp;20,000, and filtered out cells with fewer than 200 genes and genes covered by fewer than 3 cells)[\u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e47\u003c/span\u003e]. Furthermore, the FindVariableFeatures function in the Seurat package (v 5.1.0) was used, and according to the variance stabilization transformation (vst) method, the top 2,000 genes with relatively high coefficients of variation between cells, namely highly variable genes (HVGs) were obtained, and the top 10 HVGs with the most variation were identified by LabelPoints function and labeled in this study. Similarly, principal components analysis of HVGs was performed using the Seurat package (v 5.1.0). RunPCA function, and the Elbowplot and JackStraw function were employed to plot the scree plot to show the contribution of the top-ranked principal components (PCs) to cellular variation and quantify the significance of the PCs, respectively (p\u0026thinsp;\u0026lt;\u0026thinsp;0.05). After completing the PCA dimensionality reduction, the Seurat package (v 5.1.0) (resolution\u0026thinsp;=\u0026thinsp;0.2) was utilized to cluster the cells utilizing the uniform manifold approximation and projection (UMAP) method. Subsequently, the FindAllMarkers function of the Seurat package (min.pct\u0026thinsp;=\u0026thinsp;0.25, logfc.threshold\u0026thinsp;=\u0026thinsp;0.25, test.use\u0026thinsp;=\u0026thinsp;auc) was employed to detect the differentially expressed genes in each cell cluster. The Marker genes reported in previous studies were used as the primary reference, and the CellMarker dataset (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://117.50.127.228/CellMarker/\u003c/span\u003e\u003cspan address=\"http://117.50.127.228/CellMarker/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e) was used as the secondary reference to annotate the cell populations[\u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e48\u003c/span\u003e]. A bubble plot was drawn to visually display the representation of Marker genes in various cell types. After annotation, a UMAP plot of different cell type groups was drawn for visualization.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec16\" class=\"Section2\"\u003e\u003ch2\u003e4.3 Pseudotime analysis and cellular communication analysis\u003c/h2\u003e\u003cp\u003eAiming to understand the differentiation trajectory of microglia during the disease-onset process, the RunPCA, FindNeighbors, and FindClusters functions in the Seurat package (v 5.1.0) were sequentially used to perform dimensionality reduction and clustering on microglia, which were then annotated into different subgroups (p\u0026thinsp;\u0026lt;\u0026thinsp;0.05, resolution\u0026thinsp;=\u0026thinsp;0.2). Subsequently, the ggplot2 package (v 3.5.1) was employed to explore the proportional distribution of cells in each subgroup within disease samples and control samples[\u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e49\u003c/span\u003e]. Next, the Monocle2 package (v 2.28.0) was utilized to analyze the differentiation trajectory[\u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e50\u003c/span\u003e]. In addition, to analyze the communication between microglia and other cells in IS, based on the CellChatDB database (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://www.cellchat.org/cellchatdb/\u003c/span\u003e\u003cspan address=\"http://www.cellchat.org/cellchatdb/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e), with the \u0026ldquo;CellChatDB.mouse\u0026rdquo; as a reference, the aggregateNet function of the CellChat package (v 1.6.1) was utilized to analyze the cell-to-cell communication of ligand-receptor complexes (mean\u0026thinsp;=\u0026thinsp;trimean, trim\u0026thinsp;=\u0026thinsp;0.1)[\u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e51\u003c/span\u003e]. Moreover, based on functions such as identifyOverExpressedGenes, identifyOverExpressedInteraction, ProjectData, computeCommunProb, filterCommunication, and computeCommunProbPathway, potential ligand-receptor interactions were identified, and an interaction analysis was conducted on the cell subgroups within microglia (p\u0026thinsp;\u0026le;\u0026thinsp;0.05, log\u003csub\u003e2\u003c/sub\u003e mean (Molecule 1, Molecule 2)\u0026thinsp;\u0026ge;\u0026thinsp;0.1).\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec17\" class=\"Section2\"\u003e\u003ch2\u003e4.4 Identification of MGGs\u003c/h2\u003e\u003cp\u003eFurthermore, to obtain genes associated with microglia in IS, for the GSE174574 dataset, the MetacellsByGroups function in the hdWGCNA package (v 0.4.00) was applied to construct metacells (50 cells) based on the microglia subgroups of each sample[\u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e47\u003c/span\u003e]. The TestSoftPowers function was employed to select an appropriate soft threshold, and based on this soft threshold, the ConstructNetwork function was utilized to create a weighted co-expression network. Subsequently, the hierarchicalCluster function was applied for hierarchical clustering (gene modules\u0026thinsp;\u0026ge;\u0026thinsp;50) to obtain co-expressed gene modules. The ModuleEigengenes function was used to calculate the gene correlations within each module, and the ModuleConnectivity function was applied to evaluate the connectivity between modules. Then, the gene expression levels of each module were compared between disease and control samples using Wilcoxon test (p\u0026thinsp;\u0026lt;\u0026thinsp;0.05). Modules with significantly altered gene expression abundances were selected, and the genes within these modules were regarded as MGGs.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec18\" class=\"Section2\"\u003e\u003ch2\u003e4.5 Uncovering and analyzing the functions of candidate genes\u003c/h2\u003e\u003cp\u003eTo obtain genes that exhibited differential expression in the IS and control groups within the GSE58720 dataset, the limma package (v 3.58.1) was employed to discern DEGs(|log\u003csub\u003e2\u003c/sub\u003eFold change (FC)| \u0026gt;0.5, adj.p\u0026thinsp;\u0026lt;\u0026thinsp;0.05)[\u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e51\u003c/span\u003e]. Moreover, the volcano plot and heatmap of DEGs were produced via the ggplot2 package (v 3.5.1) and the ComplexHeatmap package (v 1.0.12), respectively[\u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e49\u003c/span\u003e, \u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e52\u003c/span\u003e]. The intersection of DEGs and MGGs was determined by means of the ggvenn package (v 0.1.10) (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://CRAN.R-project.org/package=ggvenn\u003c/span\u003e\u003cspan address=\"https://CRAN.R-project.org/package=ggvenn\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e) to identify candidate genes associated with IS and microglia. Subsequently, the candidate genes were analyzed for Gene ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) analyses (adj.p\u0026thinsp;\u0026lt;\u0026thinsp;0.05) via the clusterProfiler package (v 4.8.3), thus elucidating the biological functions of the candidate genes[\u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e53\u003c/span\u003e].\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec19\" class=\"Section2\"\u003e\u003ch2\u003e4.6 Identification of key genes\u003c/h2\u003e\u003cp\u003eIn order to investigate the protein-level interactions of candidate genes, we utilized the Search Tool for the Retrieval of Interacting Genes/Proteins (STRING) (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.string-db.org\u003c/span\u003e\u003cspan address=\"https://www.string-db.org\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e) to construct a PPI network (combined score\u0026thinsp;\u0026ge;\u0026thinsp;0.4). Subsequently, the PPI network was visualized by the Cytoscape package (v 3.10.2) [\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e]. Moreover, we employed the cytoHubba plugin in the Cytoscape package (v 3.10.2) to rank the candidate genes using 4 algorithms (Degree, Edge Percolated Component (EPC), Maximum Neighborhood Component (MNC), and Density of Maximum Neighborhood Component (DMN)). Then, the ggvenn package (v 0.1.10) was used to obtain the intersection of the top 10 genes selected by each of the 4 algorithms to acquire the candidate key genes. After that, the Wilcoxon test was used to contrast the expression abundances of candidate key genes among the IS group and the control group in the GSE58720 and GSE202659 datasets. Genes that had significantly different expression levels (p\u0026thinsp;\u0026lt;\u0026thinsp;0.05) and consistent trends in both datasets were defined as key genes.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec20\" class=\"Section2\"\u003e\u003ch2\u003e4.7 Gene set enrichment analysis (GSEA) and Gene set variation analysis (GSVA)\u003c/h2\u003e\u003cp\u003eThe GSEA was conducted to elucidate the biological functions of key genes throughout the progression of IS. Initially, the Spearman correlation coefficients among each key gene and all other genes across all samples in the GSE58720 dataset were calculated using the psych package (v 2.4.3) (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://CRAN.R-project.org/package=psych\u003c/span\u003e\u003cspan address=\"https://CRAN.R-project.org/package=psych\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e), with the results then ordered in descending order. Subsequently, \u0026ldquo;c2.kegg.v7.5.1.entrez.gmt\u0026rdquo; was downloaded as a reference gene set from the Molecular Signatures Database (MSigDB) (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.gsea-msigdb.org/gsea/msigdb/index.jsp\u003c/span\u003e\u003cspan address=\"https://www.gsea-msigdb.org/gsea/msigdb/index.jsp\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e) using the msigdbr package (v 7.5.1) [\u003cspan citationid=\"CR54\" class=\"CitationRef\"\u003e54\u003c/span\u003e]. Following this, the GSEA was performed with the clusterProfiler package (v 4.8.3), with a threshold of |normalized enrichment score (NES)| \u0026gt;1, False discovery rate (FDR)\u0026thinsp;\u0026lt;\u0026thinsp;0.25, and p\u0026thinsp;\u0026lt;\u0026thinsp;0.05. The GSVA was performed to elucidate the enriched pathway differences between disease and controls. First, the \u0026ldquo;h.hallmark.v7.4.symbols.gmt\u0026rdquo; from the Molecular Signatures Database (MSigDB) (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.gsea-msigdb.org/\u003c/span\u003e\u003cspan address=\"https://www.gsea-msigdb.org/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e) was utilized as the reference gene set. Subsequently, GSVA was carried out with the GSVA package (v 1.42.0) with a threshold of |t| \u0026gt;2 and p\u0026thinsp;\u0026lt;\u0026thinsp;0.05[\u003cspan citationid=\"CR55\" class=\"CitationRef\"\u003e55\u003c/span\u003e].\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec21\" class=\"Section2\"\u003e\u003ch2\u003e4.8 Construction of regulatory networks and GeneMANIA analysis\u003c/h2\u003e\u003cp\u003eThe regulatory mechanisms of key genes can also be probed based on the molecular level. The Transcriptional Regulatory Relationships Unraveled by Sentence-based Text mining (TRRUST) database (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.grnpedia.org/trrust/\u003c/span\u003e\u003cspan address=\"https://www.grnpedia.org/trrust/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e) was utilized to predict the transcription factors (TFs) that regulate key genes. In addition, the miRWalk database (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://mirwalk.umm.uni-heidelberg.de\u003c/span\u003e\u003cspan address=\"http://mirwalk.umm.uni-heidelberg.de\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e) was employed to forecast the miRNAs of key genes (the top 20 miRNAs with the smallest binding p values). Subsequently, the Cytoscape package (v 3.10.2) was adopted to construct the TFs-mRNAs network and the TFs-key genes-miRNAs regulatory network. After that, to reveal the relationships among key genes and predict the functions they participate in, based on the GeneMANIA database (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://genemania.org/\u003c/span\u003e\u003cspan address=\"https://genemania.org/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e), a co-expression network of key genes was developed to explore genes functionally related to key genes and their modes of functional interaction.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec22\" class=\"Section2\"\u003e\u003ch2\u003e4.9 Drug prediction and molecular docking\u003c/h2\u003e\u003cp\u003eDrugs targeting key genes were predicted through the DrugBank database (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://go.drugbank.com/\u003c/span\u003e\u003cspan address=\"https://go.drugbank.com/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e). The molecular formulas and 2 dimensional (2D) structures of the drugs were retrieved from the PubChem (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://pubchem.ncbi.nlm.nih.gov\u003c/span\u003e\u003cspan address=\"https://pubchem.ncbi.nlm.nih.gov\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e). The drug-key gene network was presented via the Cytoscape package (v 3.10.2). Among the predicted drugs, the one that targets the largest number of key genes and has the highest interaction score was identified as the key drug. Subsequently, convert the 2D structure into a 3D structure in Chem3D package (v 22.0.0)[\u003cspan citationid=\"CR56\" class=\"CitationRef\"\u003e56\u003c/span\u003e]. Finally, after hydrogenation in Autodock (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://autodock.scripps.edu/\u003c/span\u003e\u003cspan address=\"http://autodock.scripps.edu/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e), it was saved in PDBQT format. In addition, the 3D configuration of the target protein of the key gene was retrieved using the IBDB database (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.rcsb.org/\u003c/span\u003e\u003cspan address=\"https://www.rcsb.org/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e). Subsequently, molecular docking was carried out using the Autodock (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://autodock.scripps.edu/\u003c/span\u003e\u003cspan address=\"http://autodock.scripps.edu/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e), and the results of molecular docking were visualized using PyMOL software (v 3.0.3) [\u003cspan citationid=\"CR57\" class=\"CitationRef\"\u003e57\u003c/span\u003e]. A molecular binding energy of \u0026le;\u0026thinsp;\u0026minus;\u0026thinsp;5.0 kcal/mol was considered to indicate good binding activity.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec23\" class=\"Section2\"\u003e\u003ch2\u003e4.10 tMCAO\u003c/h2\u003e\u003cp\u003etMCAO procedure was performed in mice using an established method[\u003cspan citationid=\"CR58\" class=\"CitationRef\"\u003e58\u003c/span\u003e]. Briefly, anesthesia was induced and maintained with 1.5% isoflurane (RWD, Shenzhen, China), with body temperature kept normothermic throughout surgery. Following surgical exposure of the common carotid artery (CCA), external carotid artery (ECA), and internal carotid artery (ICA), a silicone-coated 6\u0026thinsp;\u0026minus;\u0026thinsp;0 monofilament (Covidien, St. Louis, MO) was inserted into the ECA. This filament was then advanced retrogradely to block the origin of the middle cerebral artery (MCA) for a duration of 90 minutes. Reperfusion commenced upon removal of the filament. Cerebral blood flow (CBF) was continuously assessed using laser Doppler flowmetry (Moor Instruments, Devon, UK). Occlusion was deemed successful if CBF decreased to \u0026le;\u0026thinsp;20% of pre-occlusion baseline levels, and reperfusion was confirmed by a return of CBF to \u0026ge;\u0026thinsp;80% of baseline.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec24\" class=\"Section2\"\u003e\u003ch2\u003e4.11 qRT-PCR\u003c/h2\u003e\u003cp\u003eFor the study, tissue samples from tMCAO (n\u0026thinsp;=\u0026thinsp;3) and normal samples (n\u0026thinsp;=\u0026thinsp;3) were collected from Fudan University and were approved by the Institutional Animal Care and Use Committee of Fudan University. Total RNA was extracted from tissues using TRI Reagent (Sigma-Aldrich, St. Louis, MO) via phenol-chloroform phase separation. Reverse transcription was performed using the PrimeScript RT Master Mix (Takara Bio, Dalian, China). PCR amplification was conducted on a StepOnePlus Real-Time PCR System (Thermo Fisher Scientific, Waltham, MA) under the following conditions: initial denaturation at 95\u0026deg;C for 3 min, followed by 40 cycles of denaturation at 95\u0026deg;C for 15 sec and annealing/extension at 60\u0026deg;C for 30 sec. Gene expression levels were normalized to GAPDH, and primer sequences are listed in \u003cb\u003eAdditional file 3\u003c/b\u003e.\u003c/p\u003e\u003cp\u003eReverse transcription was carried out with the TaqMan MicroRNA Reverse Transcription Kit (Applied Biosystems, Foster City, CA), followed by TaqMan probe-based quantification on a LightCycler 480 II platform (Roche, Basel, Switzerland). For every biological sample, 3 technical replicates were carried out. In the meantime, our study had been sanctioned via the Ethics Committee of Fudan University. The resulting data underwent statistical analysis and were depicted via Graphpad Prism (v 10.0) [\u003cspan citationid=\"CR59\" class=\"CitationRef\"\u003e59\u003c/span\u003e].\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec25\" class=\"Section2\"\u003e\u003ch2\u003e4.12 Statistical analysis\u003c/h2\u003e\u003cp\u003eThe statistical analyses were carried out with R software (v 4.3.1), and the notable group disparities were assessed by means of the Wilcoxon test. In qRT-PCR, the Ct values were compared using paired, two-tailed t-tests, which were performed using GraphPad Prism. A p-value\u0026thinsp;\u0026lt;\u0026thinsp;0.05 was considered statistically significant.\u003c/p\u003e\u003c/div\u003e"},{"header":"Abbreviations","content":"\u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 109px;\"\u003e\n \u003cp\u003eAbbreviation\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 459px;\"\u003e\n \u003cp\u003eFull Form\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 109px;\"\u003e\n \u003cp\u003eIS\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 459px;\"\u003e\n \u003cp\u003eIschemic Stroke\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 109px;\"\u003e\n \u003cp\u003eMGGs\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 459px;\"\u003e\n \u003cp\u003eMicroglia-Related Genes\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 109px;\"\u003e\n \u003cp\u003eDEGs\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 459px;\"\u003e\n \u003cp\u003eDifferentially Expressed Genes\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 109px;\"\u003e\n \u003cp\u003ePPI\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 459px;\"\u003e\n \u003cp\u003eProtein-Protein Interaction\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 109px;\"\u003e\n \u003cp\u003eqRT-PCR\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 459px;\"\u003e\n \u003cp\u003eQuantitative Real-Time PCR\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 109px;\"\u003e\n \u003cp\u003etMCAO\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 459px;\"\u003e\n \u003cp\u003eTransient Middle Cerebral Artery Occlusion\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 109px;\"\u003e\n \u003cp\u003escRNA-seq\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 459px;\"\u003e\n \u003cp\u003eSingle-Cell RNA Sequencing\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 109px;\"\u003e\n \u003cp\u003ehdWGCNA\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 459px;\"\u003e\n \u003cp\u003eHigh-Dimensional Weighted Gene Co-Expression Network Analysis\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 109px;\"\u003e\n \u003cp\u003eGEO\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 459px;\"\u003e\n \u003cp\u003eGene Expression Omnibus\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 109px;\"\u003e\n \u003cp\u003eQC\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 459px;\"\u003e\n \u003cp\u003eQuality Control\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 109px;\"\u003e\n \u003cp\u003eHVGs\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 459px;\"\u003e\n \u003cp\u003eHighly Variable Genes\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 109px;\"\u003e\n \u003cp\u003ePCA\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 459px;\"\u003e\n \u003cp\u003ePrincipal Component Analysis\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 109px;\"\u003e\n \u003cp\u003eUMAP\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 459px;\"\u003e\n \u003cp\u003eUniform Manifold Approximation and Projection\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 109px;\"\u003e\n \u003cp\u003eGO\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 459px;\"\u003e\n \u003cp\u003eGene Ontology\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 109px;\"\u003e\n \u003cp\u003eKEGG\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 459px;\"\u003e\n \u003cp\u003eKyoto Encyclopedia of Genes and Genomes\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 109px;\"\u003e\n \u003cp\u003eGSEA\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 459px;\"\u003e\n \u003cp\u003eGene Set Enrichment Analysis\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 109px;\"\u003e\n \u003cp\u003eGSVA\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 459px;\"\u003e\n \u003cp\u003eGene Set Variation Analysis\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 109px;\"\u003e\n \u003cp\u003eTFs\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 459px;\"\u003e\n \u003cp\u003eTranscription Factors\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 109px;\"\u003e\n \u003cp\u003emiRNAs\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 459px;\"\u003e\n \u003cp\u003eMicroRNAs\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 109px;\"\u003e\n \u003cp\u003eCNS\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 459px;\"\u003e\n \u003cp\u003eCentral Nervous System\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 109px;\"\u003e\n \u003cp\u003eBBB\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 459px;\"\u003e\n \u003cp\u003eBlood-Brain Barrier\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 109px;\"\u003e\n \u003cp\u003eCCA\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 459px;\"\u003e\n \u003cp\u003eCommon Carotid Artery\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 109px;\"\u003e\n \u003cp\u003eECA\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 459px;\"\u003e\n \u003cp\u003eExternal Carotid Artery\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 109px;\"\u003e\n \u003cp\u003eICA\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 459px;\"\u003e\n \u003cp\u003eInternal Carotid Artery\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 109px;\"\u003e\n \u003cp\u003eCBF\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 459px;\"\u003e\n \u003cp\u003eCerebral Blood Flow\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 109px;\"\u003e\n \u003cp\u003eMCA\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 459px;\"\u003e\n \u003cp\u003eMiddle Cerebral Artery\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 109px;\"\u003e\n \u003cp\u003eTLR\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 459px;\"\u003e\n \u003cp\u003eToll-Like Receptor\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 109px;\"\u003e\n \u003cp\u003eNF-\u0026kappa;B\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 459px;\"\u003e\n \u003cp\u003eNuclear Factor Kappa-Light-Chain-Enhancer of Activated B Cells\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 109px;\"\u003e\n \u003cp\u003eTNF-\u0026alpha;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 459px;\"\u003e\n \u003cp\u003eTumor Necrosis Factor Alpha\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 109px;\"\u003e\n \u003cp\u003eIL-1\u0026beta;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 459px;\"\u003e\n \u003cp\u003eInterleukin-1 Beta\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 109px;\"\u003e\n \u003cp\u003eVEGF\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 459px;\"\u003e\n \u003cp\u003eVascular Endothelial Growth Factor\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n \u003c/table\u003e\n\u003c/div\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eEthics approval and consent to participate\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe study was conducted according to the guidelines of the Declaration of Helsinki, and approved by the Institutional Review Board. (protocol code 2025-HSYY-279 and date of approval 2025-03-05).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConsent for publication\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNot applicable.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAvailability of data and materials\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe datasets generated and/or analysed during the current study are available in the [GEO] repository, [http://www.ncbi.nlm.nih.gov/geo/]\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCompeting interests\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors declare that they have no competing interests.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe research was generously supported by National Natural Science Foundation of China under grant agreement number 8217052080 and Minhang District Natural Science Foundation of Shanghai under grant agreement number 2022MHZ098.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthors\u0026apos; Contributions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eConceptualization,\u0026nbsp;DL. Q.\u0026nbsp;and\u0026nbsp;XF.Q.; methodology,\u0026nbsp;DL.Q.; software,\u0026nbsp;DL.Q; validation,\u0026nbsp;DL.Q.,\u0026nbsp;SS.L. and\u0026nbsp;YY.H.; formal analysis,\u0026nbsp;DL.Q\u0026nbsp;and\u0026nbsp;SS.L.; investigation,\u0026nbsp;DL.Q.; resources,\u0026nbsp;DL.Q.; data curation,\u0026nbsp;XF.Q.; writing\u0026mdash;original draft preparation,\u0026nbsp;DL.Q.; writing\u0026mdash;review and editing,\u0026nbsp;DL.Q.; visualization,\u0026nbsp;DL.Q.; supervision,\u0026nbsp;XF.Q.; project administration,\u0026nbsp;XF.Q and B.L.; funding acquisition,\u0026nbsp;XF.Q and B.L. All authors have read and agreed to the published version of the manuscript.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAcknowledgments\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe would like to express our sincere gratitude to all individuals and organizations who supported and assisted us throughout this research. Special thanks to the following authors: Cao Shanshan, Xu Tongtong. In conclusion, we extend our thanks to everyone who has supported and assisted us along the way. Without your support, this research would not have been possible.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eHou S, Zhang Y, Xia Y, Liu Y, Deng X, Wang W, Wang Y, Wang C, Wang G. Global, regional, and national epidemiology of ischemic stroke from 1990 to 2021. Eur J Neurol. 2024;3112:e16481.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eDeLong JH, Ohashi SN, O'Connor KC, Sansing LH. Inflammatory Responses After Ischemic Stroke. Semin Immunopathol. 2022;445:625\u0026ndash;48.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eNguyen TN, Abdalkader M, Fischer U, Qiu Z, Nagel S, Chen HS, Miao Z, Khatri P. Endovascular management of acute stroke. Lancet. 2024;40410459:1265\u0026ndash;78.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003ePaul S, Candelario-Jalil E. Emerging neuroprotective strategies for the treatment of ischemic stroke: An overview of clinical and preclinical studies. Exp Neurol. 2021;335:113518.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003ePrinz M, Jung S, Priller J. Microglia Biology: One Century of Evolving Concepts. Cell. 2019;1792:292\u0026ndash;311.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003ePlanas AM. Role of microglia in stroke. Glia 2024, 726:1016\u0026ndash;53.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eGuan X, Zhu S, Song J, Liu K, Liu M, Xie L, Wang Y, Wu J, Xu X, Pang T. Microglial CMPK2 promotes neuroinflammation and brain injury after ischemic stroke. Cell Rep Med. 2024;55:101522.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLi X, Lyu J, Li R, Jain V, Shen Y, Del \u0026Aacute;guila \u0026Aacute;, Hoffmann U, Sheng H, Yang W. Single-cell transcriptomic analysis of the immune cell landscape in the aged mouse brain after ischemic stroke. J Neuroinflammation. 2022;191:83.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLiu T, Bai M, Liu M, Li T, Liao Y, Zhao C, Yao M, Wang J, Wen A, Ding Y. Novel synergistic mechanism of 11-keto-β-boswellic acid and Z-Guggulsterone on ischemic stroke revealed by single-cell transcriptomics. Pharmacol Res. 2023;193:106803.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eSharygin D, Koniaris LG, Wells C, Zimmers TA, Hamidi T. Role of CD14 in human disease. Immunology. 2023;1693:260\u0026ndash;70.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eReed-Geaghan EG, Savage JC, Hise AG, Landreth GE. CD14 and toll-like receptors 2 and 4 are required for fibrillar A{beta}-stimulated microglial activation. J Neurosci. 2009;2938:11982\u0026ndash;92.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eRegen T, van Rossum D, Scheffel J, Kastriti ME, Revelo NH, Prinz M, Br\u0026uuml;ck W, Hanisch UK. CD14 and TRIF govern distinct responsiveness and responses in mouse microglial TLR4 challenges by structural variants of LPS. Brain Behav Immun. 2011;255:957\u0026ndash;70.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eZhou M, Wang CM, Yang WL, Wang P. Microglial CD14 activated by iNOS contributes to neuroinflammation in cerebral ischemia. Brain Res. 2013;1506:105\u0026ndash;14.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eBurrello J, Bianco G, Burrello A, Manno C, Maulucci F, Pileggi M, Nannoni S, Michel P, Bolis S, Melli G, et al. Extracellular Vesicle Surface Markers as a Diagnostic Tool in Transient Ischemic Attacks. Stroke. 2021;5210:3335\u0026ndash;47.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eTimmerman N, Waissi F, Dekker M, van de Pol QY, van Bennekom J, Schoneveld A, Klein Avink MJM, de Winter RJ, Pasterkamp G, de Borst GJ, et al. Pre-Operative Plasma Extracellular Vesicle Proteins are Associated with a High Risk of Long Term Secondary Major Cardiovascular Events in Patients Undergoing Carotid Endarterectomy. Eur J Vasc Endovasc Surg. 2021;625:705\u0026ndash;15.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eDe I, Nikodemova M, Steffen MD, Sokn E, Maklakova VI, Watters JJ, Collier LS. CSF1 overexpression has pleiotropic effects on microglia in vivo. Glia. 2014;6212:1955\u0026ndash;67.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eBoland R, Kokiko-Cochran ON. Deplete and repeat: microglial CSF1R inhibition and traumatic brain injury. Front Cell Neurosci. 2024;18:1352790.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eCao L, Liu D, Karhunen V, Ren Y, Ye D, Gao J, Gill D, Wang M. Circulating macrophage colony-stimulating factor levels and stroke: A Mendelian randomization study. J Stroke Cerebrovasc Dis. 2023;324:107050.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLi L, Yerra L, Chang B, Mathur V, Nguyen A, Luo J. Acute and late administration of colony stimulating factor 1 attenuates chronic cognitive impairment following mild traumatic brain injury in mice. Brain Behav Immun. 2021;94:274\u0026ndash;88.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLuo J, Elwood F, Britschgi M, Villeda S, Zhang H, Ding Z, Zhu L, Alabsi H, Getachew R, Narasimhan R, et al. Colony-stimulating factor 1 receptor (CSF1R) signaling in injured neurons facilitates protection and survival. J Exp Med. 2013;2101:157\u0026ndash;72.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eShibahara T, Ago T, Tachibana M, Nakamura K, Yamanaka K, Kuroda J, Wakisaka Y, Kitazono T. Reciprocal Interaction Between Pericytes and Macrophage in Poststroke Tissue Repair and Functional Recovery. Stroke. 2020;5110:3095\u0026ndash;106.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eHuang J, Jiang W, Pan J, Xu X. Correlation of sLOX-1 and CSF1 with the pathological progression of hypoxic-ischemic encephalopathy in neonatal rats. Cell Mol Biol (Noisy-le-grand) 2023, 6910:150\u0026ndash;4.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eWohleb ES, Terwilliger R, Duman CH, Duman RS. Stress-Induced Neuronal Colony Stimulating Factor 1 Provokes Microglia-Mediated Neuronal Remodeling and Depressive-like Behavior. Biol Psychiatry. 2018;831:38\u0026ndash;49.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eTajalli-Nezhad S, Karimian M, Beyer C, Atlasi MA, Azami Tameh A. The regulatory role of Toll-like receptors after ischemic stroke: neurosteroids as TLR modulators with the focus on TLR2/4. Cell Mol Life Sci. 2019;763:523\u0026ndash;37.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eJosić Dominović P, Dobrivojević Radmilović M, Srakočić S, Mišerić I, Škokić S, Gajović S. Validation and application of caged Z-DEVD-aminoluciferin bioluminescence for assessment of apoptosis of wild type and TLR2-deficient mice after ischemic stroke. J Photochem Photobiol B. 2024;253:112871.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eSun W, Ding Z, Xu S, Su Z, Li H. Crosstalk between TLR2 and Sphk1 in microglia in the cerebral ischemia/reperfusion-induced inflammatory response. Int J Mol Med. 2017;406:1750\u0026ndash;8.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eJaber Al-Mamoori A, Abdulameer Almustafa HMS, Alsaffar Y. Overexpressed Toll-Like Receptor 2 and the Influence on the Severity of Acute Ischemic Stroke. Arch Razi Inst. 2022;776:2379\u0026ndash;84.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eQiu Y, Yin Q, Fei Y, Li Y, Huang H, Fang W, Shen W, Liang B, Zhu X, Li Y. JLX001 Modulated the Inflammatory Reaction and Oxidative Stress in pMCAO Rats via Inhibiting the TLR2/4-NF-κB Signaling Pathway. Neurochem Res. 2019;448:1924\u0026ndash;38.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eZhu L, Liu S, Liao F, Yang J, Liang T, Yang Y, Huang X, Gu L, Su L. Comprehensive Analysis of Blood-Based m6A Methylation in Human Ischemic Stroke. Mol Neurobiol. 2023;602:431\u0026ndash;46.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eJi Y, Chen Y, Tan X, Huang X, Gao Q, Ma Y, Yang S, Yin M, Yu M, Fang C, et al. Integrated transcriptomic and proteomic profiling reveals the key molecular signatures of brain endothelial reperfusion injury. CNS Neurosci Ther. 2024;304:e14483.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eXie J, Zhang Y, Li B, Xi W, Wang Y, Li L, Liu C, Shen L, Han B, Kong Y, et al. Inhibition of OGFOD1 by FG4592 confers neuroprotection by activating unfolded protein response and autophagy after ischemic stroke. J Transl Med. 2024;221:248.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eShim B, Ciryam P, Tosun C, Serra R, Tsymbalyuk N, Keledjian K, Gerzanich V, Simard JM. RiboTag RNA Sequencing Identifies Local Translation of HSP70 In Astrocyte Endfeet After Cerebral Ischemia. bioRxiv 2024.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLiu L, Xu TC, Zhao ZA, Zhang NN, Li J, Chen HS. Toll-Like Receptor 4 Signaling in Neurons Mediates Cerebral Ischemia/Reperfusion Injury. Mol Neurobiol. 2023;602:864\u0026ndash;74.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eZhang M, Zhang Y, Peng J, Huang Y, Gong Z, Lu H, Han L, Wang D. Gastrodin against oxidative stress-inflammation crosstalk via inhibiting mtDNA/TLR9 and JAK2/STAT3 signaling to ameliorate ischemic stroke injury. Int Immunopharmacol. 2024;141:113012.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLi Y, Zhang J, Lei Y, Chang M, Xu J, Tang S. Multi-omics approaches reveal the therapeutic mechanism of Naoxintong capsule against ischemic stroke. J Ethnopharmacol. 2025;343:119435.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eIm JH, Yeo IJ, Park PH, Choi DY, Han SB, Yun J, Hong JT. Deletion of Chitinase-3-like 1 accelerates stroke development through enhancement of Neuroinflammation by STAT6-dependent M2 microglial inactivation in Chitinase-3-like 1 knockout mice. Exp Neurol. 2020;323:113082.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eYang Q, Jiang P, Tang H, Wen J, Zhou L, Zhao Y, Wang L, Wang J, Yang Q. Shh regulates M2 microglial polarization and fibrotic scar formation after ischemic stroke. Neurochem Int. 2024;180:105862.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eDai MJ, Gui XX, Jia SM, Lv ST, Dou H, Cui W. Dl-3-n-butylphthalide promotes angiogenesis in ischemic stroke mice through upregulating autocrine and paracrine sonic hedgehog. Acta Pharmacol Sin. 2023;4412:2404\u0026ndash;17.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eUeno K, Samura M, Nakamura T, Tanaka Y, Takeuchi Y, Kawamura D, Takahashi M, Hosoyama T, Morikage N, Hamano K. Increased plasma VEGF levels following ischemic preconditioning are associated with downregulation of miRNA-762 and miR-3072-5p. Sci Rep. 2016;6:36758.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eXu H, Wang Y, Luo Y. OTULIN is a new target of EA treatment in the alleviation of brain injury and glial cell activation via suppression of the NF-κB signalling pathway in acute ischaemic stroke rats. Mol Med. 2021;271:37.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eGanbold T, Bao Q, Zandan J, Hasi A, Baigude H. Modulation of Microglia Polarization through Silencing of NF-κB p65 by Functionalized Curdlan Nanoparticle-Mediated RNAi. ACS Appl Mater Interfaces. 2020;1210:11363\u0026ndash;74.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eYe J, Huang F, Zeng H, Xu X, Wu G, Tian S, Zhao J, Zhang W. Multi-omics and network pharmacology study reveals the effects of Dengzhan Shengmai capsule against neuroinflammatory injury and thrombosis induced by ischemic stroke. J Ethnopharmacol. 2023;305:116092.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eStirling DP, Cummins K, Mishra M, Teo W, Yong VW, Stys P. Toll-like receptor 2-mediated alternative activation of microglia is protective after spinal cord injury. Volume 137Pt 3. Brain; 2014. pp. 707\u0026ndash;23.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMedina DX, Chung EP, Teague CD, Bowser R, Sirianni RW. Intravenously Administered, Retinoid Activating Nanoparticles Increase Lifespan and Reduce Neurodegeneration in the SOD1(G93A) Mouse Model of ALS. Front Bioeng Biotechnol. 2020;8:224.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eRusu A, Tanase C, Pascu GA, Todoran N. Recent Advances Regarding the Therapeutic Potential of Adapalene. Pharmaceuticals (Basel) 2020, 139.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eParra-Izquierdo I, Lakshmanan HHS, Melrose AR, Pang J, Zheng TJ, Jordan KR, Reitsma SE, McCarty OJT, Aslan JE. The Toll-Like Receptor 2 Ligand Pam2CSK4 Activates Platelet Nuclear Factor-κB and Bruton's Tyrosine Kinase Signaling to Promote Platelet-Endothelial Cell Interactions. Front Immunol. 2021;12:729951.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMorabito S, Reese F, Rahimzadeh N, Miyoshi E, Swarup V. hdWGCNA identifies co-expression networks in high-dimensional transcriptomics data. Cell Rep Methods. 2023;36:100498.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eZheng K, Lin L, Jiang W, Chen L, Zhang X, Zhang Q, Ren Y, Hao J. Single-cell RNA-seq reveals the transcriptional landscape in ischemic stroke. J Cereb Blood Flow Metab. 2022;421:56\u0026ndash;73.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLi Y, Zhang Q, Yang J, He W, Jiang Y, Chen Y, Wang Y. Metformin combined with glucose starvation synergistically suppress triple-negative breast cancer by enhanced unfolded protein response. Biochem Biophys Res Commun. 2023;675:146\u0026ndash;54.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eTrapnell C, Cacchiarelli D, Grimsby J, Pokharel P, Li S, Morse M, Lennon NJ, Livak KJ, Mikkelsen TS, Rinn JL. The dynamics and regulators of cell fate decisions are revealed by pseudotemporal ordering of single cells. Nat Biotechnol. 2014;324:381\u0026ndash;6.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLiu Z, Sun D, Wang C. Evaluation of cell-cell interaction methods by integrating single-cell RNA sequencing data with spatial information. Genome Biol. 2022;231:218.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eDing W, Goldberg D, Zhou W. PyComplexHeatmap: a Python package to visualize multimodal genomics data. Imeta 2023, 23.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eWu T, Hu E, Xu S, Chen M, Guo P, Dai Z, Feng T, Zhou L, Tang W, Zhan L, et al. clusterProfiler 4.0: A universal enrichment tool for interpreting omics data. Innov (Camb). 2021;23:100141.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLiberzon A, Birger C, Thorvaldsd\u0026oacute;ttir H, Ghandi M, Mesirov JP, Tamayo P. The Molecular Signatures Database (MSigDB) hallmark gene set collection. Cell Syst. 2015;16:417\u0026ndash;25.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eH\u0026auml;nzelmann S, Castelo R, Guinney J. GSVA: gene set variation analysis for microarray and RNA-seq data. BMC Bioinformatics. 2013;14:7.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMeng Z, Yamazaki T, Sode K. A molecularly imprinted catalyst designed by a computational approach in catalysing a transesterification process. Biosens Bioelectron. 2004;206:1068\u0026ndash;75.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eAbu Hassan A, Hanževački M, Pordea A. Computational investigation of cis-1,4-polyisoprene binding to the latex-clearing protein LcpK30. PLoS ONE. 2024;195:e0302398.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLi Y, Liu Z, Song Y, Pan JJ, Jiang Y, Shi X, Liu C, Ma Y, Luo L, Mamtilahun M, et al. M2 microglia-derived extracellular vesicles promote white matter repair and functional recovery via miR-23a-5p after cerebral ischemia in mice. Theranostics. 2022;127:3553\u0026ndash;73.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMitteer DR, Greer BD. Using GraphPad Prism's Heat Maps for Efficient, Fine-Grained Analyses of Single-Case Data. Behav Anal Pract. 2022;152:505\u0026ndash;14.\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"},{"header":"Table","content":"\u003cp\u003eTables 1 is not available with this version.\u003c/p\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"molecular-brain","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"mbrj","sideBox":"Learn more about [Molecular Brain](http://molecularbrain.biomedcentral.com/)","snPcode":"","submissionUrl":"https://www.editorialmanager.com/mbrj/default.aspx","title":"Molecular Brain","twitterHandle":"@molecularbrain","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"BMC/SO AJ","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-7276448/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-7276448/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eIschemic stroke (IS) is an acute cerebrovascular disease characterized by high incidence and mortality. The mechanism of microglia in the pathogenesis of IS remains unclear. This study aimed to explore the key genes related to microglia in IS and their molecular mechanisms in the pathogenesis. In this study, the transcriptome data of IS were retrieved from public databases. Subsequently, candidate genes were identified through the intersection of microglia-related genes (MGGs) obtained via single-cell annotation and High-Dimensional Weighted Gene Co-Expression Network Analysis (hdWGCNA) with differentially expressed genes (DEGs). Next, key genes were determined through protein-protein interaction (PPI) analysis and verification of expression levels. Afterwards, enrichment analysis, variation analysis, construction of regulatory networks, drug prediction, and molecular docking were performed to evaluate the role of key genes in the pathogenesis of IS. Ultimately, the quantitative real-time PCR (qRT-PCR) was applied to confirm the expression levels of DEGs in brain tissues between sham and transient middle cerebral artery occlusion (tMCAO) mice. A total of 1,407 DEGs intersected with 100 MGGs, yielding 51 candidate genes. Subsequently, 3 key genes (Cd14, Csf1, and Tlr2) were successfully obtained. The study revealed that these 3 key genes were co-enriched in 4 pathways, such as leishmania infection and ribosomal, and there were differences in the enriched pathways among groups. Notably, the expression of the 3 key genes was regulated by multiple factors, including 32 microRNAs (miRNAs), such as mmu-miR-3072-5p and mmu-miR-3970, and 7 transcription factors (TFs), such as Sp1 and Nfkb1. Meanwhile, these 3 key genes predicted 8 common drugs. Interestingly, Tlr2 and Adapalene exhibited a strong binding affinity (-9.73 kcal/mol). qRT-PCR analysis revealed significantly elevated mRNA expression levels of Cd14, Csf1, and Tlr2 in tMCAO mice compared to sham-operated controls (p\u0026thinsp;\u0026lt;\u0026thinsp;0.01). This study identified and validated 3 key genes (Cd14, Csf1, and Tlr2) associated with IS, which may serve as novel targets for IS diagnosis and treatment strategies.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e","manuscriptTitle":"Identification and Validation of Microglia-Associated Genes in Ischemic Stroke Using Single-Cell and Bulk RNA-seq","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-08-22 06:57:42","doi":"10.21203/rs.3.rs-7276448/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2025-09-18T08:19:55+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-09-18T04:15:57+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"324821142620555656741604829118790568707","date":"2025-09-09T04:20:00+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2025-08-14T07:36:53+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2025-08-04T13:48:56+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2025-08-04T13:48:54+00:00","index":"","fulltext":""},{"type":"submitted","content":"Molecular Brain","date":"2025-08-02T07:37:19+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"molecular-brain","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"mbrj","sideBox":"Learn more about [Molecular Brain](http://molecularbrain.biomedcentral.com/)","snPcode":"","submissionUrl":"https://www.editorialmanager.com/mbrj/default.aspx","title":"Molecular Brain","twitterHandle":"@molecularbrain","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"BMC/SO AJ","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"c7888c19-df38-43d0-8bb0-cd3e3d978f98","owner":[],"postedDate":"August 22nd, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[],"tags":[],"updatedAt":"2025-12-08T16:11:04+00:00","versionOfRecord":{"articleIdentity":"rs-7276448","link":"https://doi.org/10.1186/s13041-025-01259-x","journal":{"identity":"molecular-brain","isVorOnly":false,"title":"Molecular Brain"},"publishedOn":"2025-12-07 15:58:29","publishedOnDateReadable":"December 7th, 2025"},"versionCreatedAt":"2025-08-22 06:57:42","video":"","vorDoi":"10.1186/s13041-025-01259-x","vorDoiUrl":"https://doi.org/10.1186/s13041-025-01259-x","workflowStages":[]},"version":"v1","identity":"rs-7276448","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-7276448","identity":"rs-7276448","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}