Gut microbiota mediates trafficking of intestinal T cells to the brain to exacerbate neuroinflammation after ischemic stroke

Gut microbiota mediates trafficking of intestinal T cells to the brain to exacerbate neuroinflammation after ischemic stroke | 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 Biological Sciences - Article Gut microbiota mediates trafficking of intestinal T cells to the brain to exacerbate neuroinflammation after ischemic stroke Zhuang Li, Nannan Guo, Zhixin Li, Xiangyu Zuo, Jinlian Shao, Fabing Zhang, and 25 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-6952560/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Ischemic stroke disrupts neuroimmune homeostasis, leading to pronounced neurological impairment, yet the cellular mechanisms remain incompletely defined. Here, we construct a high-resolution spatiotemporal immune atlas of infarcted brain regions from post-stroke patients using imaging mass cytometry, revealing a marked perivascular accumulation of PD-1⁺ T cells. Parallel investigations in a murine model of middle cerebral artery occlusion (MCAO) recapitulate this central enrichment, with PD-1⁺ T cell infiltration peaking at three days post-ischemia and persisting for up to fourteen days. Spatiotemporal fate mapping with Kaede photoconvertible reporter mice demonstrates that intestinal-derived PD-1⁺ T cells infiltrate the injured brain in a gut microbiota–dependent manner and engage directly with brain-resident PD-L1⁺ microglia. Both genetic and pharmacological disruption of the gut–brain PD-1/PD-L1 axis markedly reduces the infiltration of gut-derived T cells and ameliorates neurological deficits. Single-cell RNA sequencing of PD-L1⁺ microglia identifies Notch1-driven inflammatory signaling as a key mediator of neuroinflammation. Notably, acute stroke patients exhibit a significant increase in circulating PD-1⁺ T cells, serving as a diagnostic indicator of central nervous system injury. Together, these findings define a previously unrecognized gut–brain immune circuit that orchestrates post-stroke neuroinflammation and highlight circulating PD-1⁺ T cells as a potential biomarker of neurological damage. Biological sciences/Immunology/Neuroimmunology Biological sciences/Neuroscience/Blood–brain barrier Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Main For much of contemporary scientific history, the central nervous system (CNS) was regarded as an immune-privileged compartment 1 , with resident microglia thought to solely maintain immune surveillance and defense, and peripheral immune cell entry presumed minimal in healthy brain tissue 2,3 . This longstanding paradigm has been fundamentally revised by advances in single-cell sequencing and multi-omic technologies, which have revealed the presence and active participation of peripheral immune cells within the CNS, not only in disease contexts—including Alzheimer’s disease 4 , Parkinson’s disease 5 , multiple sclerosis 6 , traumatic brain injury 7 , and most notably, ischemic stroke 8-10 —but also under physiological conditions 11 . Once within the CNS, these immune cells interact dynamically with resident populations, regulating inflammation and contributing to tissue homeostasis and repair. These insights have transformed our understanding of neuroimmune crosstalk and broadened the conceptual framework for investigating CNS disorders. Emerging evidence indicates that peripheral immune cells access the CNS via multiple anatomical routes, such as the blood–brain barrier, the meninges 12 , choroid plexus 13 , and meningeal lymphatic vessels 14,15 . These infiltrating populations arise from diverse immune reservoirs—including the bone marrow, spleen, and, importantly, the gut 16-18 . The gut, housing the body’s largest reservoir of immune cells and microbiota, exerts significant influence on systemic and tissue-specific immune responses. The gut microbiota regulates the activation, differentiation, and migratory capacity of immune cells, shaping the nature and extent of immune traffic to distal organs such as the brain，where gut-educated cells have been implicated in the pathogenesis and progression of a range of neurological disorders 19 . In the context of ischemic stroke, interruption of cerebral blood flow induces tissue damage and disrupts the blood–brain barrier, enabling the influx of peripheral immune cells—including T cells, B cells, monocytes, macrophages, and neutrophils—into the injured parenchyma 20 . While this process has largely been attributed to passive barrier disruption, the precise mechanisms that govern the recruitment, origin, spatial dynamics, and functional interactions of these infiltrating cells with brain-resident counterparts, particularly in the context of post-stroke neuroinflammation, remain incompletely understood. To address these outstanding questions, we used imaging mass cytometry to construct a spatially resolved atlas of immune cell heterogeneity in the post-ischemic human brain. We further dissected the migratory dynamics and molecular mechanisms underlying the migration of gut-derived T cells to the CNS, and delineated their contribution to neuroinflammatory responses after stroke. Our study reveals that gut-derived PD-1⁺ T cells infiltrate the post-stroke brain in a microbiota-dependent manner and directly engage with PD-L1⁺ microglia, establishing the gut–brain PD-1/PD-L1 immune axis as a pivotal driver of post-stroke neuroinflammation. Results Construction of a high-resolution immune atlas of infarcted brain regions in post-stroke patients To systematically characterize the immune landscape of the human brain following ischemic stroke, we performed imaging mass cytometry (IMC) using a 27-marker panel on post-mortem brain tissue from patients with acute cerebral infarction and matched healthy controls (Extended Data Fig. 1). Regions of interest were delineated by CD45 immunohistochemistry (Fig. 1a), and subsequent single-cell segmentation enabled comprehensive, high-dimensional phenotypic profiling. Unsupervised clustering and UMAP embedding resolved eleven distinct cellular populations, including infiltrating T cells, microglia, oligodendrocytes, macrophages, astrocytes, and endothelial cells (Fig. 1b, c). Spatial mapping of these populations demonstrated a marked increase in immune cell density within infarcted regions compared to controls (Fig. 1d), indicating profound disruption of immune homeostasis in the post-stroke brain. Further spatial analysis revealed that T cells were predominantly localized to perivascular regions (Fig. 1e), consistent with an infiltrative phenotype and suggestive of active recruitment from the periphery via the vasculature in response to ischemic injury. Subclustering of the T cell compartment identified a distinct subset co-expressing CD3, CD4, CD45RO, and high levels of PD-1 (Fig. 1f), characteristic of an infiltrating effector memory phenotype and implicating these cells in the orchestration of post-stroke immune responses. High-resolution pixel-level immunohistochemical analysis further confirmed the presence of PD-1⁺CD45RO⁺ T cells within the parenchyma, in close proximity to blood vessels (Fig. 1g, h), providing additional evidence for their migration from the circulation in response to injury. Collectively, these data establish a high-resolution spatial atlas of immune cell architecture in the post-stroke human brain and highlight the selective accumulation of PD-1⁺ T cells as a defining feature of the neuroimmune response to ischemic injury. Spatiotemporal immune cell dynamics in the brain of the MCAO mouse model To systematically map the spatiotemporal dynamics of immune cell infiltration following ischemic stroke, we employed a transient middle cerebral artery occlusion (MCAO) mouse model and analyzed brain tissues at 4 hours, 1 day, and 3 days post-reperfusion, corresponding to the acute, subacute, and stable phases of injury, respectively (Fig. 2a). CD45 immunohistochemistry revealed a progressive accumulation of CD45 high immune cells within the infarcted cortex (Fig. 2b, c), indicative of sustained recruitment of peripheral immune cells rather than expansion of CNS-resident populations. Building on the prominent enrichment of PD-1 + T cells observed in human post-stroke brain, we further profiled these populations in MCAO mice up to 14 days after reperfusion. Single-cell transcriptomic analysis revealed marked upregulation of Pdcd1 gene expression in infiltrating T cells during both the subacute (day 2) and stable phases (day 14) (Fig. 2d). Dimensionality reduction analyses of T cells showed that PD-1 expression was largely restricted to CD44 + CD62L - effector memory T cells (Fig. 2e, Extended Data Fig. 2), which peaked at 3 days and remained elevated for at least 14 days (Fig. 2f), suggesting persistent recruitment. Notably, the majority of these PD-1 + T cells exhibited an unconventional CD4 - CD8 - γδ T cell–like phenotype (Fig. 2g), implicating this subset as a potentially important regulator of neuroimmune responses following stroke. Multiplex immunohistochemistry further demonstrated that PD-1 + T cells were detectable within the infarct core, particularly in the hippocampus and peri-infarct border regions (Fig. 2h), highlighting their spatial association with areas of neuronal injury. Collectively, these findings delineate the temporal progression and spatial distribution of infiltrating PD-1 + T cells in the post-ischemic brain, providing a comprehensive spatiotemporal atlas of neuroimmune cell dynamics following stroke. The majority of brain-infiltrating T cells after stroke originate from the gut To clarify the origin of PD-1 + T cells accumulating in the brain after ischemic stroke, we first performed flow cytometric profiling of T cell populations from the blood, heart, liver, lung, small intestine, and spleen in both MCAO and sham-operated mice (Fig. 3a). Dimensionality reduction analysis revealed a strong phenotypic resemblance between brain-infiltrating and peripheral T cells, with both populations exhibiting consistent expression patterns of CD3, CD4, CD8a, CD25, CD62L, and PD-1 (Fig. 3b, c), suggesting that most brain-infiltrating T cells are peripherally derived. To validate this migratory route, we adoptively transferred ex vivo fluorescently labelled immune cells into the circulation of MCAO and sham mice. Progressive accumulation of labelled cells was detected in the brains of MCAO mice, whereas sham animals showed no such infiltration, confirming stroke-induced recruitment of peripheral immune cells to the brain (Extended Data Fig. 3a–d). Intriguingly, analysis of peripheral organs revealed a specific reduction of PD-1 + T cells in the small intestine of MCAO mice, with no significant changes observed in other tissues (Fig. 3d). This finding implicates the gut as the principal source of PD-1 + T cell depletion in the periphery following stroke, consistent with their active migration to the injured brain. To directly trace gut-derived T cells, we employed Kaede photoconvertible fate-mapping mice, in which immune cells in the small intestine were photoconverted from Kaede-Green to Kaede-Red fluorescence 12 hours prior to MCAO. Subsequent analysis at days 1, 3, and 7 post-stroke revealed a robust, time-dependent accumulation of Kaede-Red + PD-1 + T cells in the brain, peaking at day 3 (Fig. 3e–g). Together, these results provide direct evidence that PD-1 + T cells are actively recruited from the gut to the brain after ischemic stroke, highlighting the gut as a critical reservoir for neuroimmune responses and delineating the spatiotemporal dynamics of this process. Gut microbiota mediates trafficking of intestinal T cells to the brain after stroke Given that most brain-infiltrating immune cells originate from the gut and that the gut microbiota plays a critical role in stroke outcomes, we hypothesized that post-stroke alterations in the gut microbiota might contribute to the trafficking of immune cells to the brain. To test this, we depleted gut microbiota in mice using a broad-spectrum antibiotic cocktail (ABx) for 7 days prior to MCAO induction (Fig. 4a). t-SNE analysis of brain-infiltrating T cells at day 3 post-stroke revealed distinct T cell populations based on CD3, CD4, CD8a, CD62L, CD44, and PD-1 expression (Fig. 4b). Notably, ABx-treated MCAO mice exhibited a marked reduction in PD-1 + CD4 - CD8a - T cells, as well as lower frequencies of PD-1 + CD4 + and PD-1 + CD8a + subsets, compared with untreated controls (Fig. 4c), suggesting that the gut microbiota is involved in the migration of intestinal T cells to the brain. To further investigate whether stroke-associated dysbiosis contributes to the migration of gut-derived T cells into the brain, we transplanted fecal microbiota from either ischemic stroke patients (IS-FMT) or healthy controls (HC-FMT) into ABx-pretreated mice (Fig. 4d). Microbial profiling showed that IS-FMT resulted in a distinct and more diverse ileal bacterial community, characterized by increased richness and evenness (Extended Data Fig. 4a-c), with enrichment of Bacteroides and Fusobacterium and reduced levels of Akkermansia , Parabacteroides , Enterocloster , and Escherichia Shigella (Extended Data Fig. 4d). Importantly, after MCAO, mice receiving IS-FMT displayed significant increases in brain-infiltrating PD-1 + CD4 - CD8a - and PD-1 + CD8a + T cells, along with a trend towards higher PD-1 + CD4 + T cell frequencies (Fig. 4e, f). These findings indicate that gut microbial dysbiosis following stroke contributes to the migration of gut-derived T cells to the brain. Disruption of the gut–brain PD-1/PD-L1 immune axis ameliorates post-stroke neuroinflammation To investigate the central function of gut-derived PD-1 + T cells following stroke, we administered intravenous anti–PD-1 monoclonal antibody after MCAO to block peripheral PD-1 signaling (Fig. 5a). This intervention significantly reduced the frequency of PD-1 + T cells in the brain, as shown by flow cytometry, with total T cell numbers also trending lower (Fig. 5b, Extended Data Fig. 5a). Notably, peripheral PD-1 blockade resulted in markedly improved neurological outcomes post-MCAO compared to IgG controls (Fig. 5c), indicating that the recruitment of intestinal PD-1 + T cells contributes to neurological deficits after stroke. To further assess the role of central PD-1 signaling, we delivered anti–PD-1 antibody intracerebroventricularly post-MCAO (Fig. 5d). Central blockade significantly improved neurological scores post-stroke (Fig. 5e), demonstrating that PD-1 signaling in the CNS aggravates functional impairment and reinforcing the pathogenic role of intestinal PD-1 + T cell infiltration. To further elucidate the mechanisms by which PD-1 regulates central neuroinflammation, we performed bulk RNA sequencing of cerebral tissues from wild-type, Pdcd1 –/– (PD-1 knockout), and sham-operated mice following MCAO (Fig. 5f). Principal component analysis revealed distinct transcriptomic profiles between groups (Fig. 5g). Differential expression analysis identified 3,042 genes altered between WT-MCAO and WT-sham groups, and 4,829 genes between Pdcd1 –/– -MCAO and WT-MCAO samples (Extended Data Fig. 5b, c). Unsupervised clustering of 772 overlapping differentially expressed genes showed close alignment between Pdcd1 –/– -MCAO and WT-sham, with significant downregulation of genes linked to glial activation, proliferation, and neuroinflammation (such as S100a10 , Lama5 , Cspg4 , Tnc , and Mki67 ) in Pdcd1 –/– mice (Fig. 5h, Extended Data Fig. 5d). These findings suggest that PD-1 signaling contributes to post-stroke neuroinflammation primarily by activating resident glial cells. Given that PD-1 acts via its ligand PD-L1, predominantly expressed in CNS glial populations, we next assessed the contribution of brain-resident PD-L1. Administration of anti–PD-L1 antibody intracerebroventricularly before MCAO (Fig. 5i) significantly improved neurological function during the first three days post-ischemia (Fig. 5j), indicating that CNS PD-L1 blockade alleviates stroke-induced deficits. Flow cytometry revealed that PD-L1 expression was robustly upregulated in microglia, but not in astrocytes or oligodendrocytes after MCAO (Fig. 5k–m, Extended Data Fig. 5e), establishing microglia as the main PD-L1–expressing population responding to intestinal PD-1 + T cells. Gene set enrichment analysis of sorted PD-L1 + versus PD-L1 – microglia highlighted activation of immune pathways, notably the Notch-inflammatory signaling (Fig. 5n). Both Notch1 and Notch2 in gene and protein levels were upregulated after stroke (Fig. 5o, p), but only Notch1 was significantly reduced by anti–PD-L1 treatment (Fig. 5q). Additionally, anti–PD-L1 administration decreased expression of the pro-inflammatory cytokines IL-1β and IL-6 post-MCAO (Fig. 5r), both downstream of Notch1. These data indicate that Notch1-dependent signaling in PD-L1 + microglia mediates the neuroinflammatory response to intestinal PD-1 + T cell infiltration after stroke. Taken together, these results show that disruption of the gut–brain PD-1/PD-L1 axis, either by targeting PD-1 + T cells or microglial PD-L1, attenuates post-stroke neuroinflammation, at least in part through inhibition of microglial Notch1 signaling, thereby conferring neuroprotection after ischemic stroke. Peripheral PD-1⁺ T cells are markedly increased in stroke patients and correlate with neurological dysfunction Given that intestinal PD-1⁺ T cells can migrate via the bloodstream to the brain and shape post-stroke neuroinflammation, we assessed the clinical relevance of circulating PD-1⁺ T cells in ischemic stroke by performing flow cytometric profiling of whole blood samples from both healthy controls and patients with ischemic stroke (Fig. 6a). Unsupervised UMAP projection delineated canonical T cell subsets based on the expression of CD3, CD4, CD8α, CD45RA, CCR7, and PD-1 across the T cell compartment (Fig. 6b, Extended Data Fig. 6). Notably, the proportion of PD-1⁺ T cells was significantly elevated in IS patients compared to healthy individuals (Fig. 6c). Moreover, the frequency of PD-1⁺ T cells showed a positive correlation with clinical stroke severity, as determined by NIH Stroke Scale (NIHSS) scores (Fig. 6d), and exhibited significant diagnostic value (Fig. 6e). ROC curve analysis demonstrated that circulating PD-1⁺ T cell proportion could effectively distinguish between mild (0 < NIHSS ≤ 4), moderate (4 15) stroke cases, with high AUC values for mild and moderate stroke groups (AUC = 0.913 and 0.967, respectively; Fig. 6e). Collectively, these findings indicate that ischemic stroke is associated with a marked increase in peripheral PD-1⁺ T cells, which may serve as a biomarker reflecting post-stroke neurological dysfunction. Discussion Our study provides direct evidence that the gut–brain immune axis is a critical mediator of neuroinflammation after ischemic stroke. By integrating the Kaede photoconvertible fate-mapping system with high-dimensional imaging mass cytometry 21,22 , we show that gut-derived immune cells, particularly intestinal T cells, actively migrate into infarcted brain regions and interact dynamically with the neural microenvironment via the PD-1/PD-L1 pathway. This migration is not simply a consequence of blood–brain barrier disruption; blockade of the PD-1 pathway significantly reduces the central infiltration of PD-1 positive T cells, indicating the involvement of precisely regulated, tissue-specific molecular cues within the injured brain. These results support a model of orchestrated immune cell trafficking rather than passive leakage. Although accumulating evidence links the gut microbiota to neurological outcomes after stroke 23,24 , the mechanisms underlying this gut–brain crosstalk remain poorly understood. Our findings address this gap by showing that the gut microbiome not only shapes the repertoire and functional properties of gut-derived T cells but also actively facilitates their migration to sites of cerebral injury. Notably, the number of gut-derived PD-1⁺ T cells migrating out of the gut is significantly reduced following antibiotic depletion of the microbiota, while fecal microbiota transplantation from stroke patients increases the peripheral pool of PD-1⁺ T cells, suggesting the microbiota provides essential cues for immune cell mobilization. Recent studies have offered a mechanistic explanation, demonstrating that SFB-specific TCR7B8 T cells, upon activation in the gut, migrate to the CNS in a β7 integrin-dependent manner 19 . While our study establishes that these T cells act as amplifiers of neuroinflammatory responses and directly connect gut microbial dynamics to the immune milieu of the CNS after stroke, the precise microbial and molecular signals driving such selective immune cell trafficking warrant further investigation. Our study highlights the remarkable functional and phenotypic plasticity of infiltrating T cells within the CNS, showing that their roles depend not only on their intrinsic properties but also on direct interactions with resident neural cells. Building on previous findings that regulatory T cells accumulate in the injured brain and contribute to tissue repair and inflammation resolution 8,25 , we demonstrate that T cells can form reciprocal connections with microglia. These interactions enable T cells to modulate local immune signaling, influence glial cell activation, and shape neurological recovery or degeneration. Central to this neuroimmune regulation is the gut–brain PD-1/PD-L1 axis. We observed upregulation of PD-L1 on microglia together with the infiltration of intestinal PD-1⁺ T cells, and showed that the communication between these cells calibrates neuroinflammatory responses after stroke. Disrupting this axis reduces T cell infiltration and attenuates global neuroinflammatory signaling, highlighting the importance of intercellular crosstalk between peripheral and resident immune cells in governing CNS immune states. Our findings underscore that the ultimate function of infiltrating T cells in the brain is determined by both their own programming and their interactions with the neural microenvironment, emphasizing the need for detailed, context-dependent cellular and molecular profiling to fully understand neuroimmune regulation after stroke. Our results also offer new mechanistic understanding of chronic neuroinflammation after stroke. While traditional models have attributed prolonged inflammation primarily to local glial activation, our fate-mapping studies show that gut-derived T cells can persist within the post-ischemic CNS, maintaining sustained interactions with resident neural cells. This persistent crosstalk likely perpetuates inflammatory cascades, contributing to delayed neurodegeneration and altered neural plasticity or recovery. These data support a model in which long-lasting, cross-organ immune interactions drive chronic central nervous system inflammation and highlight the importance of extending therapeutic strategies beyond the acute phase to address ongoing pathological gut–brain communication. Nevertheless, our study has limitations. The Kaede fate-mapping system, while highly informative, does not fully capture the complexity of human immune responses, and the scarcity of clinical tissue limits translational prospects. Moreover, the specific microbial taxa, metabolites, and molecular cues responsible for directing immune cell migration require further clarification. Future research using advanced imaging, integrative multi-omics, and targeted interventions in both preclinical models and clinical studies will be necessary to refine these mechanistic insights and develop effective, clinically translatable therapies. In summary, our findings demonstrate that the gut microbiota orchestrates the migration and function of gut-derived T cells into the injured brain through the gut–brain PD-1/PD-L1 axis, ultimately shaping neuroinflammatory outcomes after stroke. We further show that circulating PD-1⁺ T cells serve as promising biomarkers closely linked to neurological severity. Together, these discoveries position gut-derived T cells as central players in neuroimmune interactions and identify cross-organ immune circuits as compelling targets for both therapeutic intervention and biomarker development in post-stroke neurological disorders. Declarations Acknowledgements This work was supported by grants from the National Key R&D Program of China (2022YFA0806400), Guangzhou Key Research Program on Brain Science (202206060001) and the National Natural Science Foundation of China (81925026 and 82130068) to HWZ, the National Natural Science Foundation of China (82200936) and the Guangdong Basic and Applied Basic Research Foundation (SL2023A04J02020) to ZL, the National Natural Science Foundation of China (82302608) and the Guangdong Basic and Applied Basic Research Foundation (SL2025A04J3774) to NNG, the National Natural Science Foundation of China (82300623 and 2025B1515020059) to SXH, and the National Natural Science Foundation of China (82302610) to CCQ. We thank all the patients and their families involved in this study. We sincerely thank Hao Sun team at Sun Yat-sen Memorial Hospital, Sun Yat-sen University, for their collaboration in the transgenic Kaede mouse experiments. We also acknowledge Infinity Scope Inc.for their support on IMC experiments. Author Contributions N.G., Z.L., X.Z., Z.L., and H.Z. conceived the study and wrote the manuscript (original draft). N.G. performed most experiments with the help of Z.L., X.Z., Y.X., H.Z., L.Z., Y.Z., Z.G.. Moreover, N.G. analyzed the data with the help of B.Z., Y.C., Q.D., J.S., F.Z., Y.L., C.C., S.C., L.L., Y.L., C.Q., collected clinical samples. J.L., G.L., B.L, H.Y., S.H., E.Z., G.W., C.Y., Y.Z., F.K. help review and edit the manuscript. All authors discussed the results and commented on the manuscript. Competing interest The authors declare no competing interests. Data availability Single-cell RNA-seq data is available via Gene Expression Omnibus accession code GSE225948. Imaging mass cytometry data are deposited at Mendeley Data (https://data.mendeley.com/preview/b2s2868gty). Source data are provided with this paper. References Medawar, P. B. Immunological tolerance. Nature 189 , 14-17 (1961). https://doi.org:10.1038/189014a0 Niederkorn, J. Y. See no evil, hear no evil, do no evil: the lessons of immune privilege. Nat Immunol 7 , 354-359 (2006). https://doi.org:10.1038/ni1328 Castellani, G., Croese, T., Peralta Ramos, J. M. & Schwartz, M. Transforming the understanding of brain immunity. Science 380 , eabo7649 (2023). https://doi.org:10.1126/science.abo7649 Su, W. et al. CXCR6 orchestrates brain CD8(+) T cell residency and limits mouse Alzheimer's disease pathology. Nat Immunol 24 , 1735-1747 (2023). https://doi.org:10.1038/s41590-023-01604-z Sulzer, D. et al. T cells from patients with Parkinson's disease recognize α-synuclein peptides. Nature 546 , 656-661 (2017). https://doi.org:10.1038/nature22815 Andreadou, M. et al. IL-12 sensing in neurons induces neuroprotective CNS tissue adaptation and attenuates neuroinflammation in mice. Nat Neurosci 26 , 1701-1712 (2023). https://doi.org:10.1038/s41593-023-01435-z Shechter, R. et al. Recruitment of beneficial M2 macrophages to injured spinal cord is orchestrated by remote brain choroid plexus. Immunity 38 , 555-569 (2013). https://doi.org:10.1016/j.immuni.2013.02.012 Ito, M. et al. Brain regulatory T cells suppress astrogliosis and potentiate neurological recovery. Nature 565 , 246-250 (2019). https://doi.org:10.1038/s41586-018-0824-5 Garcia-Bonilla, L. et al. Analysis of brain and blood single-cell transcriptomics in acute and subacute phases after experimental stroke. Nat Immunol 25 , 357-370 (2024). https://doi.org:10.1038/s41590-023-01711-x Levard, D. et al. Central nervous system-associated macrophages modulate the immune response following stroke in aged mice. Nat Neurosci 27 , 1721-1733 (2024). https://doi.org:10.1038/s41593-024-01695-3 Hughes, H. K. et al. Dysregulated gene expression associated with inflammatory and translation pathways in activated monocytes from children with autism spectrum disorder. Transl Psychiatry 12 , 39 (2022). https://doi.org:10.1038/s41398-021-01766-0 Wang, Y. et al. Early developing B cells undergo negative selection by central nervous system-specific antigens in the meninges. Immunity 54 , 2784-2794.e2786 (2021). https://doi.org:10.1016/j.immuni.2021.09.016 Baruch, K. et al. Aging. Aging-induced type I interferon response at the choroid plexus negatively affects brain function. Science 346 , 89-93 (2014). https://doi.org:10.1126/science.1252945 Herisson, F. et al. Direct vascular channels connect skull bone marrow and the brain surface enabling myeloid cell migration. Nat Neurosci 21 , 1209-1217 (2018). https://doi.org:10.1038/s41593-018-0213-2 Aspelund, A. et al. A dural lymphatic vascular system that drains brain interstitial fluid and macromolecules. J Exp Med 212 , 991-999 (2015). https://doi.org:10.1084/jem.20142290 Fitzpatrick, Z. et al. Gut-educated IgA plasma cells defend the meningeal venous sinuses. Nature 587 , 472-476 (2020). https://doi.org:10.1038/s41586-020-2886-4 Ginhoux, F. et al. Fate mapping analysis reveals that adult microglia derive from primitive macrophages. Science 330 , 841-845 (2010). https://doi.org:10.1126/science.1194637 Yoshida, T. M. et al. The subfornical organ is a nucleus for gut-derived T cells that regulate behaviour. Nature (2025). https://doi.org:10.1038/s41586-025-09050-7 White, Z. et al. Gut inflammation promotes microbiota-specific CD4 T cell-mediated neuroinflammation. Nature (2025). https://doi.org:10.1038/s41586-025-09120-w Iadecola, C., Buckwalter, M. S. & Anrather, J. Immune responses to stroke: mechanisms, modulation, and therapeutic potential. J Clin Invest 130 , 2777-2788 (2020). https://doi.org:10.1172/jci135530 Guo, N. et al. A 34-Marker Panel for Imaging Mass Cytometric Analysis of Human Snap-Frozen Tissue. Front Immunol 11 , 1466 (2020). https://doi.org:10.3389/fimmu.2020.01466 Morton, A. M. et al. Endoscopic photoconversion reveals unexpectedly broad leukocyte trafficking to and from the gut. Proc Natl Acad Sci U S A 111 , 6696-6701 (2014). https://doi.org:10.1073/pnas.1405634111 Xu, K. et al. Rapid gut dysbiosis induced by stroke exacerbates brain infarction in turn. Gut (2021). https://doi.org:10.1136/gutjnl-2020-323263 Singh, V. et al. Microbiota Dysbiosis Controls the Neuroinflammatory Response after Stroke. J Neurosci 36 , 7428-7440 (2016). https://doi.org:10.1523/jneurosci.1114-16.2016 Shi, L. et al. Treg cell-derived osteopontin promotes microglia-mediated white matter repair after ischemic stroke. Immunity 54 , 1527-1542.e1528 (2021). https://doi.org:10.1016/j.immuni.2021.04.022 Jackman, K., Kunz, A. & Iadecola, C. Modeling focal cerebral ischemia in vivo. Methods Mol Biol 793 , 195-209 (2011). https://doi.org:10.1007/978-1-61779-328-8_13 Funato, H. et al. Effects of a calcium antagonist, lacidipine, on experimental focal cerebral ischemia in rats. Jpn J Pharmacol 75 , 415-423 (1997). https://doi.org:10.1254/jjp.75.415 Bederson, J. B. et al. Rat middle cerebral artery occlusion: evaluation of the model and development of a neurologic examination. Stroke 17 , 472-476 (1986). https://doi.org:10.1161/01.str.17.3.472 Ijsselsteijn, M. E., van der Breggen, R., Farina Sarasqueta, A., Koning, F. & de Miranda, N. A 40-Marker Panel for High Dimensional Characterization of Cancer Immune Microenvironments by Imaging Mass Cytometry. Front Immunol 10 , 2534 (2019). https://doi.org:10.3389/fimmu.2019.02534 Greenwald, N. F. et al. Whole-cell segmentation of tissue images with human-level performance using large-scale data annotation and deep learning. Nat Biotechnol 40 , 555-565 (2022). https://doi.org:10.1038/s41587-021-01094-0 Zheng, G. X. et al. Massively parallel digital transcriptional profiling of single cells. Nat Commun 8 , 14049 (2017). https://doi.org:10.1038/ncomms14049 Methods Patients and specimens Brain tissue samples from IS patients were obtained during decompressive craniectomy performed for large-area cerebral infarctions with associated intracranial hypertension. The specimens consisted of small portions of ischemic and necrotic brain tissue that were surgically removed. Control brain tissues were collected from patients with intracerebral hemorrhage during the creation of the surgical access channel for hematoma evacuation, and these samples represented relatively normal brain regions adjacent to the hematoma site. Peripheral blood samples were collected from IS patients and healthy donors. All samples were obtained from the Department of Neurosurgery and Emergency Department, Zhujiang Hospital, Southern Medical University. Written informed consent was obtained from all participants. The study was approved by the Ethics Committee of Zhujiang Hospital, Southern Medical University (Approval NO. 2025-KY-012-02). Mice C57BL/6J strains of mice used in this study were obtained directly from the Zhuhai Baishitong Biotechnology Co., Ltd. (Zhuhai, China). Experiments were performed in young C57BL/6Smoc-Pdcd1 em3Smoc mice (Pd-1-KO mice, Cat. NO. NM-KO-190423) were purchased from Shanghai Model Organisms Center, Inc. (Shanghai, China). The source of B6. Cg-c/c Tg(CAG-tdKaede)15Utr mice is RIKEN BioResource Center. Mice were housed in individually ventilated cages under standard conditions at 22 °C with 40 ± 5% relative humidity and a 12-h light/12-h dark cycle. Water and a standard laboratory diet were available ad libitum, unless indicated otherwise. 8-12-week-old mice were used for all experiments. All mice were acclimatized for one week before the initiation of the experiments. All animal maintenance protocols and procedures performed were approved by the Ethics Committee of the Animal Experimental Center of Zhujiang Hospital, Southern Medical University (Approval NO. LAEC-2024-038; Guangzhou, China). Mice experiments Transient focal cerebral ischemia was induced in mice using the middle cerebral artery occlusion (MCAO) model as previously described 26 . Mice were anesthetized via intraperitoneal injection of tribromoethanol (1.25%, 0.02 ml/g body weight). A midline cervical incision was made to expose the right common carotid artery (CCA), internal carotid artery (ICA), and external carotid artery (ECA). The CCA and ECA were ligated, and a silicon-coated nylon filament was inserted through the ICA to occlude the origin of the middle cerebral artery (MCA). After 60 min of occlusion, the filament was carefully withdrawn to allow reperfusion. Throughout the procedure, body temperature was maintained at 37 °C using a heating pad equipped with a rectal probe. Sham-operated mice as control underwent the same surgical exposure of the carotid arteries without insertion of the filament. Neurological function was assessed using a composite scoring system incorporating three established behavioral evaluations: Bederson's postural reflex test, Longa's tail suspension test, and Feeney's balance-beam walking test 27,28 . The final neurological score was calculated as the sum of scores from these three assessments: Bederson's postural reflex test (0: No observable deficits; 1: Impaired straight-line ambulation; 2: Ipsilateral rotational behavior; 3: Lateralized postural collapse) , Longa's tail suspension test (0: Symmetrical forelimb extension; 1: Contralateral forelimb flexion during elevation; 2: Sustained contralateral flexion without rotation; 3: Spontaneous ipsiversive circling), Feeney's beam walking test (0: Stable quadrupedal stance; 1: Grasping beam edges with all limbs; 2: Unilateral limb disengagement; 3: Bilateral limb slippage (>60 s retention); 4: Beam maintenance >40 s before fall; 5: Beam maintenance >20 s before fall; 6: Immediate postural failure). For photoconversion, Kaede transgenic mice were anesthetized with a ketamine/xylazine mixture (10 mg/kg and 2 mg/kg, respectively, intraperitoneally) or with 3% isoflurane in 100% O₂ at a flow rate of 2 L/min. As previously described 22 , mice were placed on their backs with an aluminum foil blanket covering the surrounding tissue, the small intestines were exposed to violet light was shone (handheld 405-nm laser; peak power <5 mW) onto the exposed area for a total of 10 min with brief pauses every 3 min, Throughout the procedure, the exposed intestine tissue was kept moist using sterile PBS to prevent desiccation. As for intraventricular administration, mice were anesthetized with tribromoethanol (1.25%, 0.02 ml/g body weight, intraperitoneally) and positioned in a stereotaxic apparatus (UMP3 microinjection system; WPI, Sarasota, FL, USA). Anti-mouse PD-1 monoclonal antibody, or anti-mouse PD-L1 monoclonal antibody (BioXCell) was delivered into the lateral ventricle using a 10 μL microsyringe fitted with a 33-gauge needle (RWD, China). The injection coordinates relative to bregma were: +0.3 mm anterior posterior, -1.0 mm lateral and -2.5 mm ventral. Cell suspension preparation for multiple organs and peripheral blood of mice Mice were anesthetized with tribromoethanol (0.2 mL/10g body weight) and transcardially perfused with cold PBS. Following perfusion, the right cerebral cortex was visually inspected to confirm the presence of ischemic lesions. Subsequently, brain tissue, spleen, lung, heart, small intestine, liver, and peripheral blood were collected. For mice brains, brain from control-sham or ischemic mice were dissected free of the cerebellum and olfactory bulb, minced in cold RPMI 1640 supplemented with 2% FBS, and digested for 45 min at 37°C in 5 ml of digestion medium (RPMI 1640, 2% FBS, 2 mg/mL Collagenase type D, 2 mg/mL DNase I) with constant stirring at 200 rpm. The digested tissue was washed with PBS, sequentially filtered through a 70 μm cell strainer, centrifuged, and resuspended in PBS containing 1% FBS. Spleens were mechanically dissociated through a 70-μm cell strainer into RPMI 1640 medium supplemented with 2% FBS. Red blood cells were lysed using ammonium-chloride-potassium (ACK) lysis buffer for 3 min at room temperature, followed by washing with RPMI 1640. The resulting single-cell suspension was resuspended in PBS containing 1% FBS. Small intestines were flushed with cold PBS, and adipose tissue and intestinal contents were removed. Tissues were cut into small pieces, and incubated in extraction buffer (RPMI 1640, 2% FBS, 5 mM DTT, 1 mM EDTA) at 37°C for 15 min twice with stirring to remove intraepithelial lymphocytes. Remaining fragments were washed and digested in digestion buffer (RPMI 1640, 2% FBS, 1 mg/mL collagenase II, 0.5 mg/mL Dispase). The suspension was filtered through a 70 μm strainer, layered on between 40% and 80% Percoll, and centrifuged at 1,300 × g. Cells at the interface were collected and resuspended in PBS containing 1% FBS. Lung and heart tissues were harvested after perfusion and were disrupted into small pieces, then were digested in 10 ml digestion medium (RPMI 1640, 2% FBS, 1mg/ml Collagenase type I, 0.5mg/ml Dispase; all Gibco) for 30 min at 37 °C with stirring at 200 rpm. The tissues were neutralized with PBS, then filtered through a 70 µm strainer, centrifuged, and resuspended in PBS with 1% FBS. Liver tissues were thoroughly minced in a digestive solution containing 0.5 mg/mL collagenase IV, 150ug/mL DNase I, 2% FBS, RPMI 1640 and incubated for 40 min at 37 °C on a shaker at 180 rpm. Afterwards, samples were filtered through a 70 mm cell strainer to obtain final single cell suspensions. Fresh peripheral blood underwent red blood lysis using ammonium-chloride-potassium (ACK) lysis buffer for 3 minutes at room temperature, followed by washing with RPMI. The resulting single-cell suspension was resuspended in PBS with 1% FBS. Antibiotic-mediated microbiota depletion To deplete gut microbiota, male C57BL/6J mice (6-8 weeks old) were administered a broad-spectrum antibiotic cocktail consisting of vancomycin (100 mg/kg/day), ampicillin, neomycin and metronidazole (200mg/kg/day, Aladdin Reagent) by daily oral gavage (200 µl) for 7 consecutive days. Control mice received an equivalent volume of sterile PBS following the same procedure. Fecal microbiota transplantation (FMT) FMT was performed to assess the impact of gut microbiota from individuals with ischemic stroke and healthy controls. Following antibiotic treatment and a 2-day washout period, Recipient mice were then orally gavaged (200 µl) once daily for 7 days with fecal suspensions prepared from stool samples of patients with ischemic stroke (n = 6) or healthy controls (n = 6). A total of 3,000 mg of fresh feces was resuspended in 30 ml sterile saline supplemented with 20% (v/v) glycerol and 0.5 g/l cysteine-HCl (Sigma-Aldrich) under anaerobic conditions. Suspensions were homogenized, filtered through a 100 µm cell strainer, aliquoted, and stored at −80 °C until use. Imaging cell migration To track immune cell migration in vivo , splenocytes were isolated from male C57BL/6J mice. Briefly, spleens were dissected and mechanically dissociated through a 70μm cell strainer. Erythrocytes were lysed using ACK Lysis Buffer (Leagene, Beijing, China). Isolated cells were labeled with the Cy5-conjugated cholesteryl oligonucleotide probe 5′-Cholesteryl-AAAAAAAAAAA-3′-Cy5 (Sangon Biotech, Shanghai, China) via 5 min incubation. Subsequently, Recipient mice were administered by 100μL of labeled cell suspension through tail vein injection. Recipient mice underwent MCAO or sham surgery, followed by brain collection at 3-, 24-, 48-, and 72-hours post-procedure. Fluorescent signals were acquired using an IVIS Spectrum imaging system (PerkinElmer, Boston, MA, USA). Immunohistochemistry Formalin-fixed, paraffin-embedded tissue sections were deparaffinized with xylene and rehydrated through decreasing concentrations of ethanol, respectively. Endogenous peroxidase activity was quenched with 0.3% H 2 O 2 in methanol (Merck Millipore, Burlington, MA) for 15 min at RT. Sections were boiled in Tris-EDTA (10 mM/1 mM, pH 9) buffers for antigen retrieval followed by cooling down to RT. Then tissue sections were blocked with Superblock solution (Thermo Fisher Scientific, Waltham, MA, USA) and incubated overnight at 4°C with a primary antibody (anti-mouse CD45，anti-human CD45，anti-mouse Notch1 and anti-mouse Notch2 antibody, Cell Signaling Technology). After three PBS washes, sections were incubated with poly-HRP-conjugated secondary antibody (Immunologic, Duiven, Netherlands) for 1h at RT. Signal detection used DAB+ chromogen (DAKO, Agilent Technologies, Santa Clara, CA), and nuclei were counterstained with Mayer’s hematoxylin (Thermo Fisher Scientific). Data analysis was utilized by SlideViewer software (version 2.7.0.191696). Multiplex immunohistochemistry Tissue section preparation followed standard IHC procedures before the Opal multiplex antibody staining protocol. Tissue sections were circled with a hydrophobic barrier pen Primary antibodies were applied and incubated under optimized conditions. After washing in TBST (3 times × 2 min), slides were incubated with Opal Polymer HRP Ms+Rb for 10 min. Following another TBST wash, Opal fluorophore working solution was applied for 10 min. Slides were then rinsed in AR buffer, and antigen retrieval was repeated by microwave heating as described. These steps-including blocking, primary and secondary antibody incubation, fluorophore labeling, and antigen retrieval-were repeated iteratively for each target marker. For Opal Polaris 780, TSA-DIG working solution was applied, followed by antigen retrieval and incubation with Opal Polaris 780 fluorophore for 1 h. Nuclear counterstaining was performed with DAPI (5 min, RT), followed by washing and covers lipping with anti-fade mounting medium. Slides were either scanned immediately or stored at 4°C in the dark. Imaging mass cytometric immunostaining, acquisition and analysis Imaging mass cytometry (IMC) staining on formalin-fixed paraffin-embedded (FFPE) sections of human brain biopsies was performed as previously described 29 . FFPE tissue blocks were sectioned at 4 µm thickness. These sections underwent a heating process at 68°C for 1 h to facilitate dewaxing, followed by two 10-minute incubations in xylene at 68°C for complete paraffin removal. Rehydration was achieved through sequential immersions in ethanol solutions of decreasing concentrations (95%, 85%, and 75%), each for 5 min at RT. Antigen retrieval was conducted by heating sections in 10 mM sodium citrate buffer (pH 6.0) at 100°C for 30 min. After cooling to RT, sections were washed twice in PBS-TB (PBS with 0.5% Tween-20 and 1% BSA) for 5 min each. Blocking was performed using SuperBlock buffer (Thermo Fisher Scientific) for 30 min at RT. Sections were incubated with metal-tagged antibodies overnight at 4°C as indicated in Table 1. Post-incubation, samples were washed three times with PBS-TB. Nuclear labeling was performed with an Intercalator-Ir (Fluidigm) solution in PBS-TB (1.25 µM) for 30 min at RT, followed by two additional PBS-TB washes and a final rinse with deionized water. The tissue sections were ablated at 200 Hz on a Helios time-of-flight mass cytometer coupled to a Hyperion Imaging System (Fluidigm). Instrument calibration and autotuning were performed according to the manufacturer’s protocol, using a 3-element tuning slide (Fluidigm), as described in the Hyperion Imaging System User Guide. The regions of interest (ROIs) per sample were selected based on CD45 immunohistochemistry staining on consecutive slides. All raw data were analyzed for marker intensity based on the maximum signal threshold, defined at the 98th percentile of all pixels in a single ROI using Fluidigm MCD TM viewer (v1.0.560.2). For cell segmentation analysis of IMC datasets, we utilized the pre-trained TissueNet, as described previously 30 . The TissueNet requires two channels of imaging data for its operation. TissueNet requires two imaging channels: the nuclear channel, typically stained with Intercalator-Ir, to identify cell nuclei; and a second channel representing either membrane or cytoplasmic markers to delineate cell boundaries. In this study, membrane markers were used as the second input channel to improve boundary resolution. Marker expression data were transformed using a hyperbolic arcsine function and trimmed at the 1 st and 99 th percentiles to define the range. Each channel was subsequently normalized using min-max scaling. To correct for batch effects across samples, we applied the Harmony algorithm (version 0.1.0) from the R package. Subsequent cell clustering was performed using the Rphenograph (version 0.99.1), with the number of nearest neighbors set to 30. Clustering was guided by specific markers, including CD14, GFAP, CD8, CD31, S100β, HIF1α, CD45, CD7, CD163, MAP2, CD68, CD11c, IbaI, P2Y12, HLA-DR, CD3, Olig2, αSMA, Vimentin, CD16, CD4, CD11b, CD206 and CD45RO to identify major cell populations. The mean expression of each marker across clusters was visualized on a heatmap (Fig. 1c). This expression pattern was then utilized for further annotation. Resulted images and tSNE feature plots were exported by relative R packages. Flow cytometry Single cell suspension from different mice tissues or human whole blood cells were incubated with fluorochrome-conjugated antibodies and Fc block (BioLegend) for 30 min at 4 °C for surface staining. After washing, the stained cells were resuspended. Reference samples were incorporated and individually stained by UltraComp eBeadsTM Compensation Beads (ThermoFisher), whole blood cells from mice or human. After stained samples preparation, the samples were immediately acquired using a 5-laser Cytek® Aurora (Cytek® Biosciences). Data was analyzed to check quality with FlowJo software version 10.6 (Tree Star Inc). We utilized OMIQ to perform the dimensional reduction analysis for mice and human samples (https://www.omiq.ai/). Detailed information regarding antibody panels is listed in Table 2-5. For microglia sorting, brain cell suspension were stained with surface staining flow cytometry antibodies for 30 min at 4 °C in the dark, then purified on BD Influx cell sorter (BD Biosciences). PD-L1 + microglia were identified as CD45 dim CD11b + PD-L1 + ; PD-L1 - microglia were identified as CD45 dim CD11b + PD-L1 - . Detailed information regarding antibodies is listed in Table 6. Transcriptome sequencing experiment and analysis Brain tissues were collected immediately post-euthanasia, snap-frozen in liquid nitrogen, and stored at -80°C until processing. Total RNA was extracted and quantified via Bioanalyzer 2100 (Agilent Technologies, Santa Clara, CA). RNA was reverse-transcribed into cDNA for library construction. Library fragments were size-selected (150-200 bp) and purified with AMPure XP system (Beckman Coulter, Brea, CA). Indexed libraries were clustered on a cBot Cluster Generation System with the TruSeq PE Cluster Kit v3-cBot-HS (Illumina) following the manufacturer’s instructions, then the library preparations were sequenced on an Illumina NovaSeq platform, resulting sequencing data for all samples with a read length of 150 bp (paired-end). Raw sequencing data underwent quality control. Gene-level counts were generated with FeatureCounts v1.5.0-p3. Subsequently, Normalized FPKM values were calculated for gene expression quantification. Differential expression analysis between different groups was performed using DESeq2. Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis was performed using the clusterProfler R package to test the statistical enrichment of differential expression genes. Single-cell RNA sequencing Pdcd1 gene expression in mice brain subsets was from a publicly accessible interactive web portal of previous published scRNAseq datasets (https://anratherlab.shinyapps.io/strokevis/) 9 . Single, live, PD-1 + and PD-L1 - microglia in brain from MCAO mice were sorted using a FACSAria III flow cytometer (BD Biosciences). Single-cell RNA sequencing was performed as described previously 31 . In brief, single-cell suspensions were loaded onto a Chromium Single Cell Controller (10x Genomics) together with oil, reagents, and barcoded beads. Cell lysis and barcoded reverse transcription of polyadenylated mRNA were carried out within individual gel bead emulsions. cDNA libraries were generated in a single bulk reaction, and sequencing was performed on an Illumina HiSeq 4000 platform. Quantitative real-time PCR Total RNA was extracted from brain tissue samples using the TRIzol method. cDNA was synthesized from 500ng total RNA using the PrimeScript RT reagent Kit (TaKaRa, Japan). Quantitative real-time PCR was performed using the TB Green Kit (Takara, RR820A) on an ABI Step One RT-PCR system (Applied Biosystems, Foster City, USA) under the following conditions: 95°C for 1 min; 40 cycles of 95°C for 15 s; then 60°C for 1 min. Relative gene expression levels were calculated using the 2 -ΔΔCT method, with β-actin primers serving as the internal control. Primer sequences are provided in Table 7. Microbial 16S rRNA Gene Sequencing and Bioinformatic Analysis Ileal fecal contents from HC-FMT and IS-FMT MCAO mice were collected under sterile conditions as previously described. Microbial genomic DNA was extracted using the E.Z.N.A.® Soil DNA Kit (Omega Bio-tek), following the manufacturer’s protocol. The V3–V4 region of the 16S rRNA gene was amplified and sequenced on the Illumina NextSeq 2000 platform at Majorbio Bio-Pharm Technology Co. Ltd. (Shanghai, China). Raw reads were processed using fastp (v0.19.6) for quality filtering and merged using FLASH (v1.2.7). Operational taxonomic units (OTUs) were clustered at 97% sequence identity. Taxonomic assignment was performed using the RDP Classifier (v2.2) against the SILVA v138 database. Alpha diversity was assessed by calculating Chao1 and Shannon indices using Mothur (v1.30.1). Differences between groups were evaluated by unpaired Student’s t-test. Beta diversity was analyzed using principal coordinate analysis (PCoA) based on Bray–Curtis distance matrices, with significance assessed by PERMANOVA (Vegan, v2.5-3). The relative abundance of bacterial genera was visualized with bar plots generated using the Majorbio Cloud Platform. Statistical analysis Statistical analyses were performed using GraphPad Prism 10 and R. Bar graphs and dot plots display mean values, with error bars representing the standard error of the mean (s.e.m.). For pooled analyses, data from all mice within the same experimental group across independent repeats were aggregated, and statistical comparisons were conducted on the pooled dataset as for individual experiments. All measurements were obtained from distinct samples. Statistical significance was assessed using the Wilcoxon signed-rank test and Mann–Whitney U test, as appropriate. A two-tailed P value < 0.05 was considered significant; significance levels are indicated as follows: *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001. Tables Tables 1 to 7 are available in the Supplementary Files section Additional Declarations There is NO Competing Interest. Supplementary Files Tables.xlsx Table1-Table7 ExtendedFiguresandlegends0623.docx Extended Data 1 to Extended Data 6 Cite Share Download PDF Status: Posted Version 1 posted 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-6952560","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Biological Sciences - Article","associatedPublications":[],"authors":[{"id":476449784,"identity":"1b941e2f-1aac-4ddd-ba20-3b1109c051c4","order_by":0,"name":"Zhuang Li","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA4ElEQVRIiWNgGAWjYBACxmYgwQOkGxuYD0CEDhCvhS2BOC1gANLSwMBjQJwW5nbmZw/e1NyRbZ6R8/nTzTYGOb4bCYyfC/A6jM3ccM6xZ8aNPWe3See2MRhL3khglp6B3y9m0jxshxMb23u3MQO1JG64kcDGzINXC/s3aZ5/QC3NPI8/A7XUE6GFx0yatw1kSw8DyGEJBkRoKZOc23cY6JdjZtI55yQMZ5552CyNT4th//FtEm++HZbdOCP58eecMht5vuPJBz/j1dKAypBgAMcRPiCPwRgFo2AUjIJRgA4A/yROoZWhXlMAAAAASUVORK5CYII=","orcid":"","institution":"Southern Medical University","correspondingAuthor":true,"prefix":"","firstName":"Zhuang","middleName":"","lastName":"Li","suffix":""},{"id":476449785,"identity":"e20ec5b0-e336-4b8b-a94f-9c8215c119eb","order_by":1,"name":"Nannan Guo","email":"","orcid":"","institution":"Southern Medical University","correspondingAuthor":false,"prefix":"","firstName":"Nannan","middleName":"","lastName":"Guo","suffix":""},{"id":476449786,"identity":"32cddc0a-d07a-4c2e-bbcb-c2294945a0b9","order_by":2,"name":"Zhixin Li","email":"","orcid":"","institution":"Southern Medical University","correspondingAuthor":false,"prefix":"","firstName":"Zhixin","middleName":"","lastName":"Li","suffix":""},{"id":476449787,"identity":"e4b07e02-1daf-47f9-ab61-a263c83e2b5f","order_by":3,"name":"Xiangyu Zuo","email":"","orcid":"","institution":"Southern Medical University","correspondingAuthor":false,"prefix":"","firstName":"Xiangyu","middleName":"","lastName":"Zuo","suffix":""},{"id":476449788,"identity":"e4231309-1d7f-4184-88e6-226bdfca3efb","order_by":4,"name":"Jinlian Shao","email":"","orcid":"","institution":"Southern Medical University","correspondingAuthor":false,"prefix":"","firstName":"Jinlian","middleName":"","lastName":"Shao","suffix":""},{"id":476449789,"identity":"bc8b26a0-2a24-4800-9e6b-5a7269610613","order_by":5,"name":"Fabing Zhang","email":"","orcid":"","institution":"Southern Medical University","correspondingAuthor":false,"prefix":"","firstName":"Fabing","middleName":"","lastName":"Zhang","suffix":""},{"id":476449790,"identity":"2afd8b80-5df2-4450-9944-41231bb2f9d8","order_by":6,"name":"Yu Xiao","email":"","orcid":"","institution":"Southern Medical University","correspondingAuthor":false,"prefix":"","firstName":"Yu","middleName":"","lastName":"Xiao","suffix":""},{"id":476449791,"identity":"9542e1ec-afd0-477a-a15c-623b5b54f336","order_by":7,"name":"Huaying Zhang","email":"","orcid":"","institution":"Southern Medical University","correspondingAuthor":false,"prefix":"","firstName":"Huaying","middleName":"","lastName":"Zhang","suffix":""},{"id":476449792,"identity":"785a59d1-78a3-4c02-9446-9a2c980f1e7b","order_by":8,"name":"Lingling Zhang","email":"","orcid":"","institution":"Southern Medical University","correspondingAuthor":false,"prefix":"","firstName":"Lingling","middleName":"","lastName":"Zhang","suffix":""},{"id":476449793,"identity":"81148094-0046-4206-b65b-32439f2c0f8a","order_by":9,"name":"Yulong Zhang","email":"","orcid":"","institution":"Southern Medical University","correspondingAuthor":false,"prefix":"","firstName":"Yulong","middleName":"","lastName":"Zhang","suffix":""},{"id":476449794,"identity":"ecb95fe6-8882-4473-a10c-7d079d27115b","order_by":10,"name":"Boteng Zhang","email":"","orcid":"","institution":"Southern Medical University","correspondingAuthor":false,"prefix":"","firstName":"Boteng","middleName":"","lastName":"Zhang","suffix":""},{"id":476449795,"identity":"e0b98e09-c3e4-4e81-a5f4-d40ec6343937","order_by":11,"name":"Yu Chen","email":"","orcid":"","institution":"Southern Medical University","correspondingAuthor":false,"prefix":"","firstName":"Yu","middleName":"","lastName":"Chen","suffix":""},{"id":476449796,"identity":"8769e3ce-c7d3-4b22-884f-72ecc8acacfc","order_by":12,"name":"Qingyao Du","email":"","orcid":"","institution":"Southern Medical University","correspondingAuthor":false,"prefix":"","firstName":"Qingyao","middleName":"","lastName":"Du","suffix":""},{"id":476449797,"identity":"aec0046e-1a18-4ad7-9f0e-f08ab36c6640","order_by":13,"name":"Yancaho Liu","email":"","orcid":"","institution":"Southern Medical University","correspondingAuthor":false,"prefix":"","firstName":"Yancaho","middleName":"","lastName":"Liu","suffix":""},{"id":476449798,"identity":"ffa523bf-7dc7-471a-a52a-e7065f5e7e4e","order_by":14,"name":"Chunhui Cui","email":"","orcid":"","institution":"Southern Medical University","correspondingAuthor":false,"prefix":"","firstName":"Chunhui","middleName":"","lastName":"Cui","suffix":""},{"id":476449799,"identity":"26498fb2-d25a-42ae-baec-4c5d96695ffc","order_by":15,"name":"Cancan Qi","email":"","orcid":"","institution":"Zhujiang Hospital, Southern Medical University","correspondingAuthor":false,"prefix":"","firstName":"Cancan","middleName":"","lastName":"Qi","suffix":""},{"id":476449800,"identity":"9238fe88-f7d3-49c5-a293-5b896763cd2f","order_by":16,"name":"Shi Chen","email":"","orcid":"","institution":"Southern Medical University","correspondingAuthor":false,"prefix":"","firstName":"Shi","middleName":"","lastName":"Chen","suffix":""},{"id":476449801,"identity":"163f8107-1095-4832-a0a2-c23b31b5b3ce","order_by":17,"name":"Li Ling","email":"","orcid":"","institution":"Southern Medical University","correspondingAuthor":false,"prefix":"","firstName":"Li","middleName":"","lastName":"Ling","suffix":""},{"id":476449802,"identity":"b1b6a6cd-2bb1-44c0-91d1-45cd5c5c2ee4","order_by":18,"name":"Yundi Li","email":"","orcid":"","institution":"Southern Medical University","correspondingAuthor":false,"prefix":"","firstName":"Yundi","middleName":"","lastName":"Li","suffix":""},{"id":476449803,"identity":"cb4cb8db-f7d6-4ec0-b246-b711016f651f","order_by":19,"name":"Jia Yin","email":"","orcid":"","institution":"NanFang Hospital","correspondingAuthor":false,"prefix":"","firstName":"Jia","middleName":"","lastName":"Yin","suffix":""},{"id":476449804,"identity":"a511c954-25f6-41a6-b2de-9606aa0fa99a","order_by":20,"name":"Guo-Wang Lin","email":"","orcid":"","institution":"Zhujiang Hospital, Southern Medical University, Guangzhou","correspondingAuthor":false,"prefix":"","firstName":"Guo-Wang","middleName":"","lastName":"Lin","suffix":""},{"id":476449805,"identity":"a2c661d6-29fe-4e96-a5bf-a94922e220c6","order_by":21,"name":"Bin Liu","email":"","orcid":"","institution":"Southern Medical University","correspondingAuthor":false,"prefix":"","firstName":"Bin","middleName":"","lastName":"Liu","suffix":""},{"id":476449806,"identity":"099a61b4-ceb6-4971-a1e3-927e3d044d3a","order_by":22,"name":"Hengwen Yang","email":"","orcid":"https://orcid.org/0000-0002-6418-1985","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Hengwen","middleName":"","lastName":"Yang","suffix":""},{"id":476449807,"identity":"d335bbf5-f08c-4087-9358-399bec2cd4d2","order_by":23,"name":"Shixian Hu","email":"","orcid":"","institution":"UMCG","correspondingAuthor":false,"prefix":"","firstName":"Shixian","middleName":"","lastName":"Hu","suffix":""},{"id":476449808,"identity":"107e0327-40c7-4510-a065-3401e4e5d5fe","order_by":24,"name":"Enchen Zhou","email":"","orcid":"","institution":"University of California","correspondingAuthor":false,"prefix":"","firstName":"Enchen","middleName":"","lastName":"Zhou","suffix":""},{"id":476449809,"identity":"7113a560-ae68-4337-ba50-90dadd43ca6e","order_by":25,"name":"Gangqi Wang","email":"","orcid":"","institution":"Fudan University","correspondingAuthor":false,"prefix":"","firstName":"Gangqi","middleName":"","lastName":"Wang","suffix":""},{"id":476449810,"identity":"4085f4d1-23b1-4d29-874f-c92cca7fb429","order_by":26,"name":"Zizheng Gao","email":"","orcid":"","institution":"Beihua University","correspondingAuthor":false,"prefix":"","firstName":"Zizheng","middleName":"","lastName":"Gao","suffix":""},{"id":476449811,"identity":"a9b32dff-56f8-4b03-b5ea-0afa963c3920","order_by":27,"name":"Chun Yang","email":"","orcid":"","institution":"College of Basic Medicine","correspondingAuthor":false,"prefix":"","firstName":"Chun","middleName":"","lastName":"Yang","suffix":""},{"id":476449812,"identity":"92799ed6-913d-4443-b2b1-66d2a0b36aed","order_by":28,"name":"Yan Zhang","email":"","orcid":"","institution":"School of Public Health","correspondingAuthor":false,"prefix":"","firstName":"Yan","middleName":"","lastName":"Zhang","suffix":""},{"id":476449813,"identity":"213d2518-7638-440b-a527-de9ead729f14","order_by":29,"name":"Frits Koning","email":"","orcid":"","institution":"Leiden University Medical Center","correspondingAuthor":false,"prefix":"","firstName":"Frits","middleName":"","lastName":"Koning","suffix":""},{"id":476449814,"identity":"64a1954d-9d58-4403-b99c-cbefa5f91a36","order_by":30,"name":"Hongwei Zhou","email":"","orcid":"https://orcid.org/0000-0003-2472-8541","institution":"Zhujiang Hospital, Southern Medical University","correspondingAuthor":false,"prefix":"","firstName":"Hongwei","middleName":"","lastName":"Zhou","suffix":""}],"badges":[],"createdAt":"2025-06-23 04:15:12","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-6952560/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-6952560/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":85619543,"identity":"da3fd614-e47c-4dd9-86ad-37fe8042856a","added_by":"auto","created_at":"2025-06-29 15:02:15","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":2653872,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eSpatial mapping of immune cells in post-stroke human brain tissue reveals PD-1⁺ T cell infiltration adjacent to vasculature.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ea, \u003c/strong\u003eImages of CD45 immunohistochemistry (IHC) in brain tissue sections from healthy controls (HC) and ischemic stroke (IS) patients. Regions of interest (ROIs) were selected for imaging mass cytometry (IMC) analysis. Scale bars: 2.5 mm (HC) and 1.25 mm (IS).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eb\u003c/strong\u003e, After cell segmentation of IMC datasets, UMAP dimensionality reduction of single-cell data identified 11 clusters annotated as T cells, microglia, oligodendrocytes, macrophages, astrocytes, endothelial cells, and other cell types.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ec\u003c/strong\u003e, Heatmap of relative marker expression across the identified clusters.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ed\u003c/strong\u003e, Spatial mapping of immune and glial cell types in selected ROIs from HC and IS brain tissues. Scale bars of ROIs: 200 μm and 100 μm.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ee\u003c/strong\u003e, High-resolution spatial mapping shows clustering of infiltrating T cells (green) around vasculature and near other immune cells including microglia (grey) and macrophages (cyan). Scale bar: 50 μm.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ef\u003c/strong\u003e, UMAP plots showing expression of CD3, CD7, CD4, CD8a, CD45RO, and PD-1 across cell populations, highlighting a cluster of CD3⁺CD4⁺CD45RO⁺PD-1⁺ T cells.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eg\u003c/strong\u003e, ROI02 from IS patient in pixel level of IMC datasets showing MAP2⁺ neurons (green), αSMA⁺ vasculature (red), CD45⁺ immune cells (magenta), and DNA (blue). Scale bar: 200 μm.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eh\u003c/strong\u003e, Enlarged view of selected region in (\u003cstrong\u003eg\u003c/strong\u003e) showing detectability of perivascular localization of CD3⁺CD45RO⁺PD-1⁺ T cells. Scale bar: 50 μm.\u003c/p\u003e","description":"","filename":"1.png","url":"https://assets-eu.researchsquare.com/files/rs-6952560/v1/3d6d6f46800d92b12c50b3ba.png"},{"id":85619540,"identity":"09ef1acf-d732-46b0-941f-6abaaf933263","added_by":"auto","created_at":"2025-06-29 15:02:15","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":2001001,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eTemporal accumulation of PD-1⁺ T cells in the brain after ischemic stroke in mice.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ea\u003c/strong\u003e, Schematic of experimental timeline: mice underwent middle cerebral artery occlusion (MCAO), and brain tissues were collected at indicated time points (4 h, 1 d, and 3 d) for CD45 IHC staining.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eb\u003c/strong\u003e, Representative images of CD45 IHC staining at different time points of MCAO mice show infiltrating immune cells into the infarcted brain regions.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ec\u003c/strong\u003e, Quantification of CD45\u003csup\u003ehigh\u003c/sup\u003e infiltrating immune cells in different time points. Each dot represents an individual sample. (MCAO4h, n=3; MCAO1d, n=5; MCAO3d, n=4; ns, not significant; *\u003cem\u003eP\u003c/em\u003e \u0026lt; 0.05, one-way ANOVA with multiple Tukey’s test).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ed\u003c/strong\u003e, Dot plot showing gene expression of \u003cem\u003ePdcd1\u003c/em\u003e (encoding PD-1) across single-cell RNA-seq clusters from MCAO day 2, day 14, and Sham brain samples.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ee\u003c/strong\u003e, tSNE plots of brain-infiltrating T cells showing expression of CD3, CD4, CD8a, CD62L, CD44, and PD-1 markers. Each dot represents a single cell. Brain T cells were analyzed by OMIQ (https://www.omiq.ai/). Colors represent relative expression of indicated immune markers.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ef\u003c/strong\u003e, Longitudinal flow cytometric analysis of brain T cells after MCAO showing the percentage of PD-1⁺CD4⁺, PD-1⁺CD4⁻CD8a⁻, and PD-1⁺CD8a⁺ T cells from day 1 to day 14 (n = 3-5 per group).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eg\u003c/strong\u003e, Composition of PD-1⁺ T cell subsets in the brain over time.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eh\u003c/strong\u003e, Representative mIHC images showing visualization of PD-1\u003csup\u003e+\u003c/sup\u003e T cells in brain tissue from MCAO mice as indicated by white arrows. CD3\u003csup\u003e+\u003c/sup\u003e (red), PD-1\u003csup\u003e+\u003c/sup\u003e (green), and DNA (blue) Scale bar, 100 μm.\u003c/p\u003e\n\u003cp\u003e\u0026nbsp;\u003c/p\u003e","description":"","filename":"2.png","url":"https://assets-eu.researchsquare.com/files/rs-6952560/v1/b8904629703a89e95e4f85ef.png"},{"id":85620255,"identity":"8b8569cf-211e-4ad4-b176-c54e61fc237e","added_by":"auto","created_at":"2025-06-29 15:10:15","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":1448830,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eIntestinal PD-1⁺ T cells migrate to the brain following ischemic stroke.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ea\u003c/strong\u003e, Experimental design: mice were subjected to MCAO or sham surgery. At day 3 post-surgery, T cells from blood and peripheral organs (heart, liver, lung, small intestine (SI), and spleen) were analyzed by flow cytometry.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eb\u003c/strong\u003e, tSNE plots showing expression of T cell markers (CD3, CD4, CD8a, CD25, CD62L, and PD-1) in all T cells from pooled peripheral tissues and blood.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ec\u003c/strong\u003e, tSNE plots PD-1 expression in T cells from individual organs and blood.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ed\u003c/strong\u003e, Quantification of PD-1⁺ T cell subsets in peripheral tissues and blood. Frequencies of PD-1⁺CD4⁺ T, PD-1⁺CD4⁻CD8a⁻ T and PD-1⁺CD8a⁺ T cells between sham and MCAO groups (n = 5 per group; **P \u0026lt; 0.01, Mann-Whitney test with two-tailed for comparisons).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ee\u003c/strong\u003e, Schematic of photoconversion experiment using Kaede transgenic mice: mice intestines underwent photoconversion 12 h before MCAO surgery. At indicated time points post-stroke (day 1, 3, and 7), PD-1⁺ Kaede-Red⁺ T cells in the brain were analyzed by flow cytometry.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ef\u003c/strong\u003e, Quantification of Kaede-Red⁺PD-1⁺ T cell subsets (PD-1⁺CD4⁺ T, PD-1⁺CD4⁻CD8a⁻ T and PD-1⁺CD8a⁺ T cells) in the ischemic brain over time.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eg\u003c/strong\u003e, Proportional representation of Kaede-Red⁺PD-1⁺ T cell subsets in the brain at each time point.\u003c/p\u003e","description":"","filename":"3.png","url":"https://assets-eu.researchsquare.com/files/rs-6952560/v1/403c217e3e4ca70bf8c7cab4.png"},{"id":85619546,"identity":"1ce72e66-a3a1-42ea-85ca-5f647f09c104","added_by":"auto","created_at":"2025-06-29 15:02:15","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":957262,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eGut microbiota modulates PD-1⁺ T cell accumulation after ischemic stroke.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ea\u003c/strong\u003e, Schematic of experimental design. Mice were treated with PBS or antibiotics (ABx) from day 0 to 7, followed by MCAO on day 10; brains were collected for flow cytometric analysis on day 13.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eb\u003c/strong\u003e, tSNE plots showing marker expression (CD3, CD4, CD8a, CD62L, CD44, PD-1) in brain-infiltrating T cells from MCAO and Abx + MCAO groups.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ec\u003c/strong\u003e, Quantification of PD-1⁺CD4⁺ T, PD-1⁺CD4⁻CD8a⁻ T and PD-1⁺CD8a⁺ T subsets within PD-1\u003csup\u003e+\u003c/sup\u003e T cells (MCAO group, n = 6; ABx + MCAO group, n = 4; ns, not significant; *P \u0026lt; 0.05, Mann-Whitney test with two-tailed for comparisons).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ed\u003c/strong\u003e, Schematic of FMT experimental workflow, mice received ABx from day 0 to 7, then transplanted with feces from healthy controls (HC-FMT) or ischemic stroke (IS-FMT) donors from day 8 to 14, followed by MCAO on day 15 and brain harvest on day 18.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ee\u003c/strong\u003e, tSNE plots of brain T cells stained for CD3, CD4, CD8a, CD62L, CD44, and PD-1 from HC-FMT and IS-FMT groups.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ef\u003c/strong\u003e, Quantification of PD-1⁺CD4⁺ T, PD-1⁺CD4⁻CD8a⁻ T and PD-1⁺CD8a⁺ T subsets within PD-1\u003csup\u003e+\u003c/sup\u003e T cells (n = 9 per group; ns, not significant; *P \u0026lt; 0.05, ***P \u0026lt; 0.001, Mann-Whitney test with two-tailed for comparisons).\u003c/p\u003e","description":"","filename":"4.png","url":"https://assets-eu.researchsquare.com/files/rs-6952560/v1/62233e89d16e5d4893234a27.png"},{"id":85619547,"identity":"8a6af21c-64bd-4c9a-ba4e-4e7338f77df3","added_by":"auto","created_at":"2025-06-29 15:02:15","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":1075137,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003ePd-1⁺ t cells exacerbate neurological deficits after stroke by promoting Notch signaling activation in Pd-L1⁺ microglia\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ea\u003c/strong\u003e, Experimental schematic: mice were intravenously injected with IgG control or anti–PD-1 monoclonal antibody (mAb) at 12 h after MCAO. Neurological function was assessed on days 1, 2, and 3, followed by flow cytometric analysis on day 3.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eb\u003c/strong\u003e, Frequency and absolute counts of PD-1\u003csup\u003e+\u003c/sup\u003e T cells at day 3 post-MCAO (n = 4 per group; ns, not significant; *P \u0026lt; 0.05, ***P \u0026lt; 0.001, Mann-Whitney test with two-tailed for comparisons).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ec\u003c/strong\u003e, Neurological scores at indicated time points post-MCAO in mice treated with IgG or anti–PD-1 mAb intravenously (ns, not significant; *P \u0026lt; 0.05, Mann-Whitney test with two-tailed for comparisons).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ed\u003c/strong\u003e, Experimental schematic: mice received IgG or anti–PD-1 mAb i.c.v. 12 h prior to MCAO. Neurological function was assessed at days 1, 2, and 3 post-MCAO.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ee\u003c/strong\u003e, Neurological scores measured at indicated time points after MCAO for i.c.v. injection and control groups (*P \u0026lt; 0.05, Mann-Whitney test with two-tailed for comparisons).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ef\u003c/strong\u003e, Schematic of bulk RNA-seq experiment for WT and \u003cem\u003ePdcd1\u003c/em\u003e\u003csup\u003e−\u003c/sup\u003e/\u003csup\u003e−\u003c/sup\u003e mice: mice were subjected to Sham or MCAO surgery, and the brain tissue of experimental mice at day 3 for transcriptomic analysis.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eg\u003c/strong\u003e, Principal component analysis (PCA) of brain tissue transcriptomes shows clustering according to genotype and treatment (WT Sham, n = 6; WT MCAO, n = 4; \u003cem\u003ePdcd1\u003c/em\u003e\u003csup\u003e−\u003c/sup\u003e/\u003csup\u003e− \u003c/sup\u003eMCAO, n = 6).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eh\u003c/strong\u003e, Heatmap illustrating hierarchical clustering of 772 differentially expressed genes among the three groups, with representative gene names indicated above. Blue indicates lower expression, red indicates higher expression (scale: −1 to 1).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ei,\u003c/strong\u003e\u0026nbsp;Experimental schematic: mice received i.c.v. injection of IgG or anti–PD-L1 mAb at 12 hours before MCAO. Neurological function was evaluated on days 1–3 post-MCAO.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ej,\u003c/strong\u003e\u0026nbsp;Neurological scores at indicated time points following MCAO between IgG and anti–PD-L1 mAb i.c.v. treatment groups (*P \u0026lt; 0.05, **P \u0026lt; 0.01, Mann-Whitney test with two-tailed for comparisons).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ek,\u003c/strong\u003e\u0026nbsp;Workflow schematic: Mice were subjected to Sham or MCAO surgery, and brains were collected for flow cytometric analysis and sorting of PD-L1\u003csup\u003e+\u003c/sup\u003e\u0026nbsp;and PD-L1\u003csup\u003e\u003cstrong\u003e–\u003c/strong\u003e\u003c/sup\u003e\u0026nbsp;microglia at day 3.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003el-m,\u003c/strong\u003e\u0026nbsp;Representative flow cytometry plots of PD-L1 expression on microglia, astrocytes, and oligodendrocytes from Sham and MCAO mice, and quantification of percentage of PD-L1\u003csup\u003e+\u003c/sup\u003e cells among microglia, astrocytes, and oligodendrocytes (****P \u0026lt; 0.0001, Mann-Whitney test with two-tailed for comparisons).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003en,\u003c/strong\u003e\u0026nbsp;KEGG enrichment analysis of all significant upregulated genes in PD-L1\u003csup\u003e+\u003c/sup\u003e\u0026nbsp;microglia versus PD-L1\u003csup\u003e\u003cstrong\u003e–\u003c/strong\u003e\u003c/sup\u003e\u0026nbsp;microglia with highlighting the Notch signaling pathway.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eo,\u003c/strong\u003e\u0026nbsp;Relative expression of \u003cem\u003eNotch1\u003c/em\u003e and \u003cem\u003eNotch2\u003c/em\u003e gene in brains from MCAO and Sham mice, assessed by qPCR (Sham group, n = 7; MCAO group, n = 6; **P \u0026lt; 0.01, Mann-Whitney test with two-tailed for comparisons).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ep\u003c/strong\u003e, Representative images and quantification of Notch1 and Notch2 positive cells in Sham and MCAO groups. Images were captured at 20× magnification. (n = 3 or 4; *P \u0026lt; 0.05, Mann-Whitney test with two-tailed for comparisons).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eq-r,\u003c/strong\u003e\u0026nbsp;Relative expression of \u003cem\u003eNotch1, Notch2\u003c/em\u003e,\u003cem\u003e Il1b\u003c/em\u003e, and \u003cem\u003eIl6\u003c/em\u003e in brains from IgG versus anti–PD-L1 mAb i.c.v. treatment groups at 3 days post-MCAO (IgG + MCAO group, n = 4; anti–PD-L1 mAb i.c.v +MCAO group, n = 5; **P \u0026lt; 0.01, Mann-Whitney test with two-tailed for comparisons).\u003c/p\u003e","description":"","filename":"5.png","url":"https://assets-eu.researchsquare.com/files/rs-6952560/v1/b926aea9bb2a78b161451627.png"},{"id":85619545,"identity":"3d6a1b03-cb16-437d-afe2-37c510766a3f","added_by":"auto","created_at":"2025-06-29 15:02:15","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":526031,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eIncreased frequency of PD-1⁺ T cells in stroke patients and its association with NIHSS score.\u003c/strong\u003e\u003cbr\u003e\n \u003cstrong\u003ea\u003c/strong\u003e, Schematic of the study design showing whole blood collection from healthy controls (HC, n = 23) and ischemic stroke patients (IS, n = 37) followed by flow cytometric analysis.\u003cbr\u003e\n \u003cstrong\u003eb\u003c/strong\u003e, UMAP visualization of T cell subsets from human whole blood, colored by marker expression levels (CD3, CD4, CD8a, CD45RA, CCR7, PD-1).\u003cbr\u003e\n \u003cstrong\u003ec\u003c/strong\u003e, Quantification of PD-1⁺ T cell frequency in HC and IS cohorts. (*P \u0026lt; 0.05, Mann-Whitney test with two-tailed for comparisons).\u003cbr\u003e\n \u003cstrong\u003ed\u003c/strong\u003e, Correlation analysis between PD-1⁺ T cell frequency and NIHSS score in IS patients.\u003cbr\u003e\n \u003cstrong\u003ee\u003c/strong\u003e, ROC curves for PD-1⁺ T cell frequency discriminating stroke severity groups based on NIHSS scores. AUC values are shown for mild (0 \u0026lt; NIHSS ≤ 4), moderate (4 \u0026lt; NIHSS ≤ 15), and severe stroke (NIHSS \u0026gt; 15).\u003c/p\u003e","description":"","filename":"6.png","url":"https://assets-eu.researchsquare.com/files/rs-6952560/v1/fdbb1d6b41fe34a53d375edc.png"},{"id":90590827,"identity":"c55b6e3e-34ba-44cf-a1d3-807aec0fef8f","added_by":"auto","created_at":"2025-09-04 12:34:27","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":9503129,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-6952560/v1/9b02f391-ef75-4f2f-a834-8b5e8e60bf85.pdf"},{"id":85619541,"identity":"a0ddef78-7ba2-44d6-a386-2f191c49dd13","added_by":"auto","created_at":"2025-06-29 15:02:15","extension":"xlsx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":17772,"visible":true,"origin":"","legend":"Table1-Table7","description":"","filename":"Tables.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-6952560/v1/3c85a6e13b2af177dc36e8f9.xlsx"},{"id":85619542,"identity":"c5e1c53b-0af0-4e07-81c1-4e46b6f2aad1","added_by":"auto","created_at":"2025-06-29 15:02:15","extension":"docx","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":2217659,"visible":true,"origin":"","legend":"Extended Data 1 to Extended Data 6","description":"","filename":"ExtendedFiguresandlegends0623.docx","url":"https://assets-eu.researchsquare.com/files/rs-6952560/v1/7ba420155279f7078bcae198.docx"}],"financialInterests":"There is \u003cb\u003eNO\u003c/b\u003e Competing Interest.","formattedTitle":"Gut microbiota mediates trafficking of intestinal T cells to the brain to exacerbate neuroinflammation after ischemic stroke","fulltext":[{"header":"Main","content":"\u003cp\u003eFor much of contemporary scientific history, the central nervous system (CNS) was regarded as an immune-privileged compartment\u003csup\u003e1\u003c/sup\u003e, with resident microglia thought to solely maintain immune surveillance and defense, and peripheral immune cell entry presumed minimal in healthy brain tissue\u003csup\u003e2,3\u003c/sup\u003e. This longstanding paradigm has been fundamentally revised by advances in single-cell sequencing and multi-omic technologies, which have revealed the presence and active participation of peripheral immune cells within the CNS, not only in disease contexts\u0026mdash;including Alzheimer\u0026rsquo;s disease\u003csup\u003e4\u003c/sup\u003e, Parkinson\u0026rsquo;s disease\u003csup\u003e5\u003c/sup\u003e, multiple sclerosis\u003csup\u003e6\u003c/sup\u003e, traumatic brain injury\u003csup\u003e7\u003c/sup\u003e, and most notably, ischemic stroke\u003csup\u003e8-10\u003c/sup\u003e\u0026mdash;but also under physiological conditions\u003csup\u003e11\u003c/sup\u003e. Once within the CNS, these immune cells interact dynamically with resident populations, regulating inflammation and contributing to tissue homeostasis and repair. These insights have transformed our understanding of neuroimmune crosstalk and broadened the conceptual framework for investigating CNS disorders.\u003c/p\u003e\n\u003cp\u003eEmerging evidence indicates that peripheral immune cells access the CNS via multiple anatomical routes, such as the blood\u0026ndash;brain barrier, the meninges\u003csup\u003e12\u003c/sup\u003e, choroid plexus\u003csup\u003e13\u003c/sup\u003e, and meningeal lymphatic vessels\u003csup\u003e14,15\u003c/sup\u003e. These infiltrating populations arise from diverse immune reservoirs\u0026mdash;including the bone marrow, spleen, and, importantly, the gut\u003csup\u003e16-18\u003c/sup\u003e. The gut, housing the body\u0026rsquo;s largest reservoir of immune cells and microbiota, exerts significant influence on systemic and tissue-specific immune responses. The gut microbiota regulates the activation, differentiation, and migratory capacity of immune cells, shaping the nature and extent of immune traffic to distal organs such as the brain，where gut-educated cells have been implicated in the pathogenesis and progression of a range of neurological disorders\u003csup\u003e19\u003c/sup\u003e.\u003c/p\u003e\n\u003cp\u003eIn the context of ischemic stroke, interruption of cerebral blood flow induces tissue damage and disrupts the blood\u0026ndash;brain barrier, enabling the influx of peripheral immune cells\u0026mdash;including T cells, B cells, monocytes, macrophages, and neutrophils\u0026mdash;into the injured parenchyma\u003csup\u003e20\u003c/sup\u003e. While this process has largely been attributed to passive barrier disruption, the precise mechanisms that govern the recruitment, origin, spatial dynamics, and functional interactions of these infiltrating cells with brain-resident counterparts, particularly in the context of post-stroke neuroinflammation, remain incompletely understood.\u003c/p\u003e\n\u003cp\u003eTo address these outstanding questions, we used imaging mass cytometry to construct a spatially resolved atlas of immune cell heterogeneity in the post-ischemic human brain. We further dissected the migratory dynamics and molecular mechanisms underlying the migration of gut-derived T cells to the CNS, and delineated their contribution to neuroinflammatory responses after stroke. Our study reveals that gut-derived PD-1⁺ T cells infiltrate the post-stroke brain in a microbiota-dependent manner and directly engage with PD-L1⁺ microglia, establishing the gut\u0026ndash;brain PD-1/PD-L1 immune axis as a pivotal driver of post-stroke neuroinflammation.\u0026nbsp;\u003c/p\u003e"},{"header":"Results","content":"\u003cp\u003eConstruction of a high-resolution immune atlas of infarcted brain regions in post-stroke patients\u003c/p\u003e\n\u003cp\u003eTo systematically characterize the immune landscape of the human brain following ischemic stroke, we performed imaging mass cytometry (IMC) using a 27-marker panel on post-mortem brain tissue from patients with acute cerebral infarction and matched healthy controls (Extended Data Fig. 1). Regions of interest were delineated by CD45 immunohistochemistry (Fig. 1a), and subsequent single-cell segmentation enabled comprehensive, high-dimensional phenotypic profiling. Unsupervised clustering and UMAP embedding resolved eleven distinct cellular populations, including infiltrating T cells, microglia, oligodendrocytes, macrophages, astrocytes, and endothelial cells (Fig. 1b, c). Spatial mapping of these populations demonstrated a marked increase in immune cell density within infarcted regions compared to controls (Fig. 1d), indicating profound disruption of immune homeostasis in the post-stroke brain.\u003c/p\u003e\n\u003cp\u003eFurther spatial analysis revealed that T cells were predominantly localized to perivascular regions (Fig. 1e), consistent with an infiltrative phenotype and suggestive of active recruitment from the periphery via the vasculature in response to ischemic injury. Subclustering of the T cell compartment identified a distinct subset co-expressing CD3, CD4, CD45RO, and high levels of PD-1 (Fig. 1f), characteristic of an infiltrating effector memory phenotype and implicating these cells in the orchestration of post-stroke immune responses. High-resolution pixel-level immunohistochemical analysis further confirmed the presence of PD-1⁺CD45RO⁺ T cells within the parenchyma, in close proximity to blood vessels (Fig. 1g, h), providing additional evidence for their migration from the circulation in response to injury. Collectively, these data establish a high-resolution spatial atlas of immune cell architecture in the post-stroke human brain and highlight the selective accumulation of PD-1⁺ T cells as a defining feature of the neuroimmune response to ischemic injury.\u003c/p\u003e\n\u003cp\u003eSpatiotemporal immune cell dynamics in the brain of the MCAO mouse model\u003c/p\u003e\n\u003cp\u003eTo systematically map the spatiotemporal dynamics of immune cell infiltration following ischemic stroke, we employed a transient middle cerebral artery occlusion (MCAO) mouse model and analyzed brain tissues at 4 hours, 1 day, and 3 days post-reperfusion, corresponding to the acute, subacute, and stable phases of injury, respectively (Fig. 2a). CD45 immunohistochemistry revealed a progressive accumulation of CD45\u003csup\u003ehigh\u003c/sup\u003e immune cells within the infarcted cortex (Fig. 2b, c), indicative of sustained recruitment of peripheral immune cells rather than expansion of CNS-resident populations.\u003c/p\u003e\n\u003cp\u003eBuilding on the prominent enrichment of PD-1\u003csup\u003e+\u003c/sup\u003e T cells observed in human post-stroke brain, we further profiled these populations in MCAO mice up to 14 days after reperfusion. Single-cell transcriptomic analysis revealed marked upregulation of\u0026nbsp;\u003cem\u003ePdcd1\u003c/em\u003e gene expression in infiltrating T cells during both the subacute (day 2) and stable phases (day 14) (Fig. 2d). Dimensionality reduction analyses of T cells showed that PD-1 expression was largely restricted to CD44\u003csup\u003e+\u003c/sup\u003eCD62L\u003csup\u003e-\u0026nbsp;\u003c/sup\u003eeffector memory T cells (Fig. 2e, Extended Data Fig. 2), which peaked at 3 days and remained elevated for at least 14 days (Fig. 2f), suggesting persistent recruitment. Notably, the majority of these PD-1\u003csup\u003e+\u003c/sup\u003e T cells exhibited an unconventional CD4\u003csup\u003e-\u003c/sup\u003eCD8\u003csup\u003e-\u003c/sup\u003e \u0026gamma;\u0026delta; T cell\u0026ndash;like phenotype (Fig. 2g), implicating this subset as a potentially important regulator of neuroimmune responses following stroke. Multiplex immunohistochemistry further demonstrated that PD-1\u003csup\u003e+\u003c/sup\u003e T cells were detectable within the infarct core, particularly in the hippocampus and peri-infarct border regions (Fig. 2h), highlighting their spatial association with areas of neuronal injury. Collectively, these findings delineate the temporal progression and spatial distribution of infiltrating PD-1\u003csup\u003e+\u003c/sup\u003e T cells in the post-ischemic brain, providing a comprehensive spatiotemporal atlas of neuroimmune cell dynamics following stroke.\u003c/p\u003e\n\u003cp\u003eThe majority of brain-infiltrating T cells after stroke originate from the gut\u003c/p\u003e\n\u003cp\u003eTo clarify the origin of PD-1\u003csup\u003e+\u003c/sup\u003e T cells accumulating in the brain after ischemic stroke, we first performed flow cytometric profiling of T cell populations from the blood, heart, liver, lung, small intestine, and spleen in both MCAO and sham-operated mice (Fig. 3a). Dimensionality reduction analysis revealed a strong phenotypic resemblance between brain-infiltrating and peripheral T cells, with both populations exhibiting consistent expression patterns of CD3, CD4, CD8a, CD25, CD62L, and PD-1 (Fig. 3b, c), suggesting that most brain-infiltrating T cells are peripherally derived. To validate this migratory route, we adoptively transferred ex vivo fluorescently labelled immune cells into the circulation of MCAO and sham mice. Progressive accumulation of labelled cells was detected in the brains of MCAO mice, whereas sham animals showed no such infiltration, confirming stroke-induced recruitment of peripheral immune cells to the brain (Extended Data Fig. 3a\u0026ndash;d).\u003c/p\u003e\n\u003cp\u003eIntriguingly, analysis of peripheral organs revealed a specific reduction of PD-1\u003csup\u003e+\u003c/sup\u003e T cells in the small intestine of MCAO mice, with no significant changes observed in other tissues (Fig. 3d). This finding implicates the gut as the principal source of PD-1\u003csup\u003e+\u003c/sup\u003e T cell depletion in the periphery following stroke, consistent with their active migration to the injured brain. To directly trace gut-derived T cells, we employed Kaede photoconvertible fate-mapping mice, in which immune cells in the small intestine were photoconverted from Kaede-Green to Kaede-Red fluorescence 12 hours prior to MCAO. Subsequent analysis at days 1, 3, and 7 post-stroke revealed a robust, time-dependent accumulation of Kaede-Red\u003csup\u003e+\u003c/sup\u003ePD-1\u003csup\u003e+\u0026nbsp;\u003c/sup\u003eT cells in the brain, peaking at day 3 (Fig. 3e\u0026ndash;g). Together, these results provide direct evidence that PD-1\u003csup\u003e+\u003c/sup\u003e T cells are actively recruited from the gut to the brain after ischemic stroke, highlighting the gut as a critical reservoir for neuroimmune responses and delineating the spatiotemporal dynamics of this process.\u003c/p\u003e\n\u003cp\u003eGut microbiota mediates trafficking of intestinal T cells to the brain after stroke\u003c/p\u003e\n\u003cp\u003eGiven that most brain-infiltrating immune cells originate from the gut and that the gut microbiota plays a critical role in stroke outcomes, we hypothesized that post-stroke alterations in the gut microbiota might contribute to the trafficking of immune cells to the brain. To test this, we depleted gut microbiota in mice using a broad-spectrum antibiotic cocktail (ABx) for 7 days prior to MCAO induction (Fig. 4a). t-SNE analysis of brain-infiltrating T cells at day 3 post-stroke revealed distinct T cell populations based on CD3, CD4, CD8a, CD62L, CD44, and PD-1 expression (Fig. 4b). Notably, ABx-treated MCAO mice exhibited a marked reduction in PD-1\u003csup\u003e+\u003c/sup\u003eCD4\u003csup\u003e-\u003c/sup\u003eCD8a\u003csup\u003e-\u0026nbsp;\u003c/sup\u003eT cells, as well as lower frequencies of PD-1\u003csup\u003e+\u003c/sup\u003eCD4\u003csup\u003e+\u003c/sup\u003e and PD-1\u003csup\u003e+\u003c/sup\u003eCD8a\u003csup\u003e+\u003c/sup\u003e subsets, compared with untreated controls (Fig. 4c), suggesting that the gut microbiota is involved in the migration of intestinal T cells to the brain.\u003c/p\u003e\n\u003cp\u003eTo further investigate whether stroke-associated dysbiosis contributes to the migration of gut-derived T cells into the brain, we transplanted fecal microbiota from either ischemic stroke patients (IS-FMT) or healthy controls (HC-FMT) into ABx-pretreated mice (Fig. 4d). Microbial profiling showed that IS-FMT resulted in a distinct and more diverse ileal bacterial community, characterized by increased richness and evenness (Extended Data Fig. 4a-c), with enrichment of \u003cem\u003eBacteroides\u003c/em\u003e and \u003cem\u003eFusobacterium\u003c/em\u003e and reduced levels of \u003cem\u003eAkkermansia\u003c/em\u003e, \u003cem\u003eParabacteroides\u003c/em\u003e, \u003cem\u003eEnterocloster\u003c/em\u003e, and \u003cem\u003eEscherichia Shigella\u003c/em\u003e (Extended Data Fig. 4d). Importantly, after MCAO, mice receiving IS-FMT displayed significant increases in brain-infiltrating PD-1\u003csup\u003e+\u003c/sup\u003eCD4\u003csup\u003e-\u003c/sup\u003eCD8a\u003csup\u003e-\u003c/sup\u003e and PD-1\u003csup\u003e+\u003c/sup\u003eCD8a\u003csup\u003e+\u003c/sup\u003e T cells, along with a trend towards higher PD-1\u003csup\u003e+\u003c/sup\u003eCD4\u003csup\u003e+\u003c/sup\u003e T cell frequencies (Fig. 4e, f). These findings indicate that gut microbial dysbiosis following stroke contributes to the migration of gut-derived T cells to the brain.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eDisruption of the gut\u0026ndash;brain PD-1/PD-L1 immune axis ameliorates post-stroke neuroinflammation\u003c/p\u003e\n\u003cp\u003eTo investigate the central function of gut-derived PD-1\u003csup\u003e+\u003c/sup\u003e T cells following stroke, we administered intravenous anti\u0026ndash;PD-1 monoclonal antibody after MCAO to block peripheral PD-1 signaling (Fig. 5a). This intervention significantly reduced the frequency of PD-1\u003csup\u003e+\u003c/sup\u003e T cells in the brain, as shown by flow cytometry, with total T cell numbers also trending lower (Fig. 5b, Extended Data Fig. 5a). Notably, peripheral PD-1 blockade resulted in markedly improved neurological outcomes post-MCAO compared to IgG controls (Fig. 5c), indicating that the recruitment of intestinal PD-1\u003csup\u003e+\u003c/sup\u003e T cells contributes to neurological deficits after stroke. To further assess the role of central PD-1 signaling, we delivered anti\u0026ndash;PD-1 antibody intracerebroventricularly post-MCAO (Fig. 5d). Central blockade significantly improved neurological scores post-stroke (Fig. 5e), demonstrating that PD-1 signaling in the CNS aggravates functional impairment and reinforcing the pathogenic role of intestinal PD-1\u003csup\u003e+\u003c/sup\u003e T cell infiltration. To further elucidate the mechanisms by which PD-1 regulates central neuroinflammation, we performed bulk RNA sequencing of cerebral tissues from wild-type,\u0026nbsp;\u003cem\u003ePdcd1\u003c/em\u003e\u003csup\u003e\u0026ndash;/\u0026ndash;\u003c/sup\u003e (PD-1 knockout), and sham-operated mice following MCAO (Fig. 5f). Principal component analysis revealed distinct transcriptomic profiles between groups (Fig. 5g). Differential expression analysis identified 3,042 genes altered between WT-MCAO and WT-sham groups, and 4,829 genes between\u0026nbsp;\u003cem\u003ePdcd1\u003c/em\u003e\u003csup\u003e\u0026ndash;/\u0026ndash;\u003c/sup\u003e-MCAO and WT-MCAO samples (Extended Data Fig. 5b, c). Unsupervised clustering of 772 overlapping differentially expressed genes showed close alignment between\u0026nbsp;\u003cem\u003ePdcd1\u003c/em\u003e\u003csup\u003e\u0026ndash;/\u0026ndash;\u003c/sup\u003e-MCAO and WT-sham, with significant downregulation of genes linked to glial activation, proliferation, and neuroinflammation (such as\u0026nbsp;\u003cem\u003eS100a10\u003c/em\u003e,\u0026nbsp;\u003cem\u003eLama5\u003c/em\u003e,\u0026nbsp;\u003cem\u003eCspg4\u003c/em\u003e,\u0026nbsp;\u003cem\u003eTnc\u003c/em\u003e, and\u0026nbsp;\u003cem\u003eMki67\u003c/em\u003e) in\u0026nbsp;\u003cem\u003ePdcd1\u003c/em\u003e\u003csup\u003e\u0026ndash;/\u0026ndash;\u003c/sup\u003e mice (Fig. 5h, Extended Data Fig. 5d). These findings suggest that PD-1 signaling contributes to post-stroke neuroinflammation primarily by activating resident glial cells.\u003c/p\u003e\n\u003cp\u003eGiven that PD-1 acts via its ligand PD-L1, predominantly expressed in CNS glial populations, we next assessed the contribution of brain-resident PD-L1. Administration of anti\u0026ndash;PD-L1 antibody intracerebroventricularly before MCAO (Fig. 5i) significantly improved neurological function during the first three days post-ischemia (Fig. 5j), indicating that CNS PD-L1 blockade alleviates stroke-induced deficits. Flow cytometry revealed that PD-L1 expression was robustly upregulated in microglia, but not in astrocytes or oligodendrocytes after MCAO (Fig. 5k\u0026ndash;m, Extended Data Fig. 5e), establishing microglia as the main PD-L1\u0026ndash;expressing population responding to intestinal PD-1\u003csup\u003e+\u003c/sup\u003e T cells.\u003c/p\u003e\n\u003cp\u003eGene set enrichment analysis of sorted PD-L1\u003csup\u003e+\u003c/sup\u003e versus PD-L1\u003csup\u003e\u0026ndash;\u003c/sup\u003e microglia highlighted activation of immune pathways, notably the Notch-inflammatory signaling (Fig. 5n). Both Notch1 and Notch2 in gene and protein levels were upregulated after stroke (Fig. 5o, p), but only Notch1 was significantly reduced by anti\u0026ndash;PD-L1 treatment (Fig. 5q). Additionally, anti\u0026ndash;PD-L1 administration decreased expression of the pro-inflammatory cytokines IL-1\u0026beta; and IL-6 post-MCAO (Fig. 5r), both downstream of Notch1. These data indicate that Notch1-dependent signaling in PD-L1\u003csup\u003e+\u003c/sup\u003e microglia mediates the neuroinflammatory response to intestinal PD-1\u003csup\u003e+\u003c/sup\u003e T cell infiltration after stroke.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eTaken together, these results show that disruption of the gut\u0026ndash;brain PD-1/PD-L1 axis, either by targeting PD-1\u003csup\u003e+\u003c/sup\u003e T cells or microglial PD-L1, attenuates post-stroke neuroinflammation, at least in part through inhibition of microglial Notch1 signaling, thereby conferring neuroprotection after ischemic stroke.\u003c/p\u003e\n\u003cp\u003ePeripheral PD-1⁺ T cells are markedly increased in stroke patients and correlate with neurological dysfunction\u003c/p\u003e\n\u003cp\u003eGiven that intestinal PD-1⁺ T cells can migrate via the bloodstream to the brain and shape post-stroke neuroinflammation, we assessed the clinical relevance of circulating PD-1⁺ T cells in ischemic stroke by performing flow cytometric profiling of whole blood samples from both healthy controls and patients with ischemic stroke (Fig. 6a). Unsupervised UMAP projection delineated canonical T cell subsets based on the expression of CD3, CD4, CD8\u0026alpha;, CD45RA, CCR7, and PD-1 across the T cell compartment (Fig. 6b, Extended Data Fig. 6). Notably, the proportion of PD-1⁺ T cells was significantly elevated in IS patients compared to healthy individuals (Fig. 6c). Moreover, the frequency of PD-1⁺ T cells showed a positive correlation with clinical stroke severity, as determined by NIH Stroke Scale (NIHSS) scores (Fig. 6d), and exhibited significant diagnostic value (Fig. 6e). ROC curve analysis demonstrated that circulating PD-1⁺ T cell proportion could effectively distinguish between mild (0 \u0026lt; NIHSS \u0026le; 4), moderate (4 \u0026lt; NIHSS \u0026le; 15), and severe (NIHSS \u0026gt; 15) stroke cases, with high AUC values for mild and moderate stroke groups (AUC = 0.913 and 0.967, respectively; Fig. 6e). Collectively, these findings indicate that ischemic stroke is associated with a marked increase in peripheral PD-1⁺ T cells, which may serve as a biomarker reflecting post-stroke neurological dysfunction.\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eOur study provides direct evidence that the gut–brain immune axis is a critical mediator of neuroinflammation after ischemic stroke. By integrating the Kaede photoconvertible fate-mapping system with high-dimensional imaging mass cytometry\u003csup\u003e21,22\u003c/sup\u003e, we show that gut-derived immune cells, particularly intestinal T cells, actively migrate into infarcted brain regions and interact dynamically with the neural microenvironment via the PD-1/PD-L1 pathway. This migration is not simply a consequence of blood–brain barrier disruption; blockade of the PD-1 pathway significantly reduces the central infiltration of PD-1 positive T cells, indicating the involvement of precisely regulated, tissue-specific molecular cues within the injured brain. These results support a model of orchestrated immune cell trafficking rather than passive leakage.\u003c/p\u003e\n\u003cp\u003eAlthough accumulating evidence links the gut microbiota to neurological outcomes after stroke\u003csup\u003e23,24\u003c/sup\u003e, the mechanisms underlying this gut–brain crosstalk remain poorly understood. Our findings address this gap by showing that the gut microbiome not only shapes the repertoire and functional properties of gut-derived T cells but also actively facilitates their migration to sites of cerebral injury. Notably, the number of gut-derived PD-1⁺ T cells migrating out of the gut is significantly reduced following antibiotic depletion of the microbiota, while fecal microbiota transplantation from stroke patients increases the peripheral pool of PD-1⁺ T cells, suggesting the microbiota provides essential cues for immune cell mobilization. Recent studies have offered a mechanistic explanation, demonstrating that SFB-specific TCR7B8 T cells, upon activation in the gut, migrate to the CNS in a β7 integrin-dependent manner\u003csup\u003e19\u003c/sup\u003e. While our study establishes that these T cells act as amplifiers of neuroinflammatory responses and directly connect gut microbial dynamics to the immune milieu of the CNS after stroke, the precise microbial and molecular signals driving such selective immune cell trafficking warrant further investigation.\u003c/p\u003e\n\u003cp\u003eOur study highlights the remarkable functional and phenotypic plasticity of infiltrating T cells within the CNS, showing that their roles depend not only on their intrinsic properties but also on direct interactions with resident neural cells. Building on previous findings that regulatory T cells accumulate in the injured brain and contribute to tissue repair and inflammation resolution\u003csup\u003e8,25\u003c/sup\u003e, we demonstrate that T cells can form reciprocal connections with microglia. These interactions enable T cells to modulate local immune signaling, influence glial cell activation, and shape neurological recovery or degeneration. Central to this neuroimmune regulation is the gut–brain PD-1/PD-L1 axis. We observed upregulation of PD-L1 on microglia together with the infiltration of intestinal PD-1⁺ T cells, and showed that the communication between these cells calibrates neuroinflammatory responses after stroke. Disrupting this axis reduces T cell infiltration and attenuates global neuroinflammatory signaling, highlighting the importance of intercellular crosstalk between peripheral and resident immune cells in governing CNS immune states. Our findings underscore that the ultimate function of infiltrating T cells in the brain is determined by both their own programming and their interactions with the neural microenvironment, emphasizing the need for detailed, context-dependent cellular and molecular profiling to fully understand neuroimmune regulation after stroke.\u003c/p\u003e\n\u003cp\u003eOur results also offer new mechanistic understanding of chronic neuroinflammation after stroke. While traditional models have attributed prolonged inflammation primarily to local glial activation, our fate-mapping studies show that gut-derived T cells can persist within the post-ischemic CNS, maintaining sustained interactions with resident neural cells. This persistent crosstalk likely perpetuates inflammatory cascades, contributing to delayed neurodegeneration and altered neural plasticity or recovery. These data support a model in which long-lasting, cross-organ immune interactions drive chronic central nervous system inflammation and highlight the importance of extending therapeutic strategies beyond the acute phase to address ongoing pathological gut–brain communication.\u003c/p\u003e\n\u003cp\u003eNevertheless, our study has limitations. The Kaede fate-mapping system, while highly informative, does not fully capture the complexity of human immune responses, and the scarcity of clinical tissue limits translational prospects. Moreover, the specific microbial taxa, metabolites, and molecular cues responsible for directing immune cell migration require further clarification. Future research using advanced imaging, integrative multi-omics, and targeted interventions in both preclinical models and clinical studies will be necessary to refine these mechanistic insights and develop effective, clinically translatable therapies.\u003c/p\u003e\n\u003cp\u003eIn summary, our findings demonstrate that the gut microbiota orchestrates the migration and function of gut-derived T cells into the injured brain through the gut–brain PD-1/PD-L1 axis, ultimately shaping neuroinflammatory outcomes after stroke. We further show that circulating PD-1⁺ T cells serve as promising biomarkers closely linked to neurological severity. Together, these discoveries position gut-derived T cells as central players in neuroimmune interactions and identify cross-organ immune circuits as compelling targets for both therapeutic intervention and biomarker development in post-stroke neurological disorders.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eAcknowledgements\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis work was supported by grants from the National Key R\u0026amp;D Program of China (2022YFA0806400), Guangzhou Key Research Program on Brain Science (202206060001) and the National Natural Science Foundation of China (81925026 and 82130068) to HWZ, the National Natural Science Foundation of China (82200936) and the Guangdong Basic and Applied Basic Research Foundation (SL2023A04J02020) to ZL, the National Natural Science Foundation of China (82302608) and the Guangdong Basic and Applied Basic Research Foundation (SL2025A04J3774) to NNG, the National Natural Science Foundation of China (82300623 and 2025B1515020059) to SXH, and the National Natural Science Foundation of China (82302610) to CCQ. We thank all the patients and their families involved in this study. We sincerely thank Hao Sun team at Sun Yat-sen Memorial Hospital, Sun Yat-sen University, for their collaboration in the transgenic Kaede mouse experiments. We also acknowledge Infinity Scope Inc.for their support on IMC experiments.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthor Contributions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eN.G., Z.L., X.Z., Z.L., and H.Z. conceived the study and wrote the manuscript (original draft). N.G. performed most experiments with the help of Z.L., X.Z., Y.X., H.Z., L.Z., Y.Z., Z.G.. Moreover, N.G. analyzed the data with the help of B.Z., Y.C., Q.D., J.S., F.Z., Y.L., C.C., S.C., L.L., Y.L., C.Q., collected clinical samples. J.L., G.L., B.L, H.Y., S.H., E.Z., G.W., C.Y., Y.Z., F.K. help review and edit the manuscript. All authors discussed the results and commented on the manuscript.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCompeting interest\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors declare no competing interests.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eData availability\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eSingle-cell RNA-seq data is available via Gene Expression Omnibus accession code GSE225948. Imaging mass cytometry data are deposited at Mendeley Data (https://data.mendeley.com/preview/b2s2868gty). Source data are provided with this paper.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n \u003cli\u003eMedawar, P. B. Immunological tolerance. \u003cem\u003eNature\u003c/em\u003e \u003cstrong\u003e189\u003c/strong\u003e, 14-17 (1961). https://doi.org:10.1038/189014a0\u003c/li\u003e\n \u003cli\u003eNiederkorn, J. Y. See no evil, hear no evil, do no evil: the lessons of immune privilege. \u003cem\u003eNat Immunol\u003c/em\u003e \u003cstrong\u003e7\u003c/strong\u003e, 354-359 (2006). https://doi.org:10.1038/ni1328\u003c/li\u003e\n \u003cli\u003eCastellani, G., Croese, T., Peralta Ramos, J. M. \u0026amp; Schwartz, M. Transforming the understanding of brain immunity. \u003cem\u003eScience\u003c/em\u003e \u003cstrong\u003e380\u003c/strong\u003e, eabo7649 (2023). https://doi.org:10.1126/science.abo7649\u003c/li\u003e\n \u003cli\u003eSu, W.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e CXCR6 orchestrates brain CD8(+) T cell residency and limits mouse Alzheimer\u0026apos;s disease pathology. \u003cem\u003eNat Immunol\u003c/em\u003e \u003cstrong\u003e24\u003c/strong\u003e, 1735-1747 (2023). https://doi.org:10.1038/s41590-023-01604-z\u003c/li\u003e\n \u003cli\u003eSulzer, D.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e T cells from patients with Parkinson\u0026apos;s disease recognize \u0026alpha;-synuclein peptides. \u003cem\u003eNature\u003c/em\u003e \u003cstrong\u003e546\u003c/strong\u003e, 656-661 (2017). https://doi.org:10.1038/nature22815\u003c/li\u003e\n \u003cli\u003eAndreadou, M.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e IL-12 sensing in neurons induces neuroprotective CNS tissue adaptation and attenuates neuroinflammation in mice. \u003cem\u003eNat Neurosci\u003c/em\u003e \u003cstrong\u003e26\u003c/strong\u003e, 1701-1712 (2023). https://doi.org:10.1038/s41593-023-01435-z\u003c/li\u003e\n \u003cli\u003eShechter, R.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e Recruitment of beneficial M2 macrophages to injured spinal cord is orchestrated by remote brain choroid plexus. \u003cem\u003eImmunity\u003c/em\u003e \u003cstrong\u003e38\u003c/strong\u003e, 555-569 (2013). https://doi.org:10.1016/j.immuni.2013.02.012\u003c/li\u003e\n \u003cli\u003eIto, M.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e Brain regulatory T cells suppress astrogliosis and potentiate neurological recovery. \u003cem\u003eNature\u003c/em\u003e \u003cstrong\u003e565\u003c/strong\u003e, 246-250 (2019). https://doi.org:10.1038/s41586-018-0824-5\u003c/li\u003e\n \u003cli\u003eGarcia-Bonilla, L.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e Analysis of brain and blood single-cell transcriptomics in acute and subacute phases after experimental stroke. \u003cem\u003eNat Immunol\u003c/em\u003e \u003cstrong\u003e25\u003c/strong\u003e, 357-370 (2024). https://doi.org:10.1038/s41590-023-01711-x\u003c/li\u003e\n \u003cli\u003eLevard, D.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e Central nervous system-associated macrophages modulate the immune response following stroke in aged mice. \u003cem\u003eNat Neurosci\u003c/em\u003e \u003cstrong\u003e27\u003c/strong\u003e, 1721-1733 (2024). https://doi.org:10.1038/s41593-024-01695-3\u003c/li\u003e\n \u003cli\u003eHughes, H. K.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e Dysregulated gene expression associated with inflammatory and translation pathways in activated monocytes from children with autism spectrum disorder. \u003cem\u003eTransl Psychiatry\u003c/em\u003e \u003cstrong\u003e12\u003c/strong\u003e, 39 (2022). https://doi.org:10.1038/s41398-021-01766-0\u003c/li\u003e\n \u003cli\u003eWang, Y.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e Early developing B cells undergo negative selection by central nervous system-specific antigens in the meninges. \u003cem\u003eImmunity\u003c/em\u003e \u003cstrong\u003e54\u003c/strong\u003e, 2784-2794.e2786 (2021). https://doi.org:10.1016/j.immuni.2021.09.016\u003c/li\u003e\n \u003cli\u003eBaruch, K.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e Aging. Aging-induced type I interferon response at the choroid plexus negatively affects brain function. \u003cem\u003eScience\u003c/em\u003e \u003cstrong\u003e346\u003c/strong\u003e, 89-93 (2014). https://doi.org:10.1126/science.1252945\u003c/li\u003e\n \u003cli\u003eHerisson, F.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e Direct vascular channels connect skull bone marrow and the brain surface enabling myeloid cell migration. \u003cem\u003eNat Neurosci\u003c/em\u003e \u003cstrong\u003e21\u003c/strong\u003e, 1209-1217 (2018). https://doi.org:10.1038/s41593-018-0213-2\u003c/li\u003e\n \u003cli\u003eAspelund, A.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e A dural lymphatic vascular system that drains brain interstitial fluid and macromolecules. \u003cem\u003eJ Exp Med\u003c/em\u003e \u003cstrong\u003e212\u003c/strong\u003e, 991-999 (2015). https://doi.org:10.1084/jem.20142290\u003c/li\u003e\n \u003cli\u003eFitzpatrick, Z.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e Gut-educated IgA plasma cells defend the meningeal venous sinuses. \u003cem\u003eNature\u003c/em\u003e \u003cstrong\u003e587\u003c/strong\u003e, 472-476 (2020). https://doi.org:10.1038/s41586-020-2886-4\u003c/li\u003e\n \u003cli\u003eGinhoux, F.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e Fate mapping analysis reveals that adult microglia derive from primitive macrophages. \u003cem\u003eScience\u003c/em\u003e \u003cstrong\u003e330\u003c/strong\u003e, 841-845 (2010). https://doi.org:10.1126/science.1194637\u003c/li\u003e\n \u003cli\u003eYoshida, T. M.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e The subfornical organ is a nucleus for gut-derived T cells that regulate behaviour. \u003cem\u003eNature\u003c/em\u003e (2025). https://doi.org:10.1038/s41586-025-09050-7\u003c/li\u003e\n \u003cli\u003eWhite, Z.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e Gut inflammation promotes microbiota-specific CD4 T cell-mediated neuroinflammation. \u003cem\u003eNature\u003c/em\u003e (2025). https://doi.org:10.1038/s41586-025-09120-w\u003c/li\u003e\n \u003cli\u003eIadecola, C., Buckwalter, M. S. \u0026amp; Anrather, J. Immune responses to stroke: mechanisms, modulation, and therapeutic potential. \u003cem\u003eJ Clin Invest\u003c/em\u003e \u003cstrong\u003e130\u003c/strong\u003e, 2777-2788 (2020). https://doi.org:10.1172/jci135530\u003c/li\u003e\n \u003cli\u003eGuo, N.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e A 34-Marker Panel for Imaging Mass Cytometric Analysis of Human Snap-Frozen Tissue. \u003cem\u003eFront Immunol\u003c/em\u003e \u003cstrong\u003e11\u003c/strong\u003e, 1466 (2020). https://doi.org:10.3389/fimmu.2020.01466\u003c/li\u003e\n \u003cli\u003eMorton, A. M.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e Endoscopic photoconversion reveals unexpectedly broad leukocyte trafficking to and from the gut. \u003cem\u003eProc Natl Acad Sci U S A\u003c/em\u003e \u003cstrong\u003e111\u003c/strong\u003e, 6696-6701 (2014). https://doi.org:10.1073/pnas.1405634111\u003c/li\u003e\n \u003cli\u003eXu, K.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e Rapid gut dysbiosis induced by stroke exacerbates brain infarction in turn. \u003cem\u003eGut\u003c/em\u003e (2021). https://doi.org:10.1136/gutjnl-2020-323263\u003c/li\u003e\n \u003cli\u003eSingh, V.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e Microbiota Dysbiosis Controls the Neuroinflammatory Response after Stroke. \u003cem\u003eJ Neurosci\u003c/em\u003e \u003cstrong\u003e36\u003c/strong\u003e, 7428-7440 (2016). https://doi.org:10.1523/jneurosci.1114-16.2016\u003c/li\u003e\n \u003cli\u003eShi, L.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e Treg cell-derived osteopontin promotes microglia-mediated white matter repair after ischemic stroke. \u003cem\u003eImmunity\u003c/em\u003e \u003cstrong\u003e54\u003c/strong\u003e, 1527-1542.e1528 (2021). https://doi.org:10.1016/j.immuni.2021.04.022\u003c/li\u003e\n \u003cli\u003eJackman, K., Kunz, A. \u0026amp; Iadecola, C. Modeling focal cerebral ischemia in vivo. \u003cem\u003eMethods Mol Biol\u003c/em\u003e \u003cstrong\u003e793\u003c/strong\u003e, 195-209 (2011). https://doi.org:10.1007/978-1-61779-328-8_13\u003c/li\u003e\n \u003cli\u003eFunato, H.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e Effects of a calcium antagonist, lacidipine, on experimental focal cerebral ischemia in rats. \u003cem\u003eJpn J Pharmacol\u003c/em\u003e \u003cstrong\u003e75\u003c/strong\u003e, 415-423 (1997). https://doi.org:10.1254/jjp.75.415\u003c/li\u003e\n \u003cli\u003eBederson, J. B.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e Rat middle cerebral artery occlusion: evaluation of the model and development of a neurologic examination. \u003cem\u003eStroke\u003c/em\u003e \u003cstrong\u003e17\u003c/strong\u003e, 472-476 (1986). https://doi.org:10.1161/01.str.17.3.472\u003c/li\u003e\n \u003cli\u003eIjsselsteijn, M. E., van der Breggen, R., Farina Sarasqueta, A., Koning, F. \u0026amp; de Miranda, N. A 40-Marker Panel for High Dimensional Characterization of Cancer Immune Microenvironments by Imaging Mass Cytometry. \u003cem\u003eFront Immunol\u003c/em\u003e \u003cstrong\u003e10\u003c/strong\u003e, 2534 (2019). https://doi.org:10.3389/fimmu.2019.02534\u003c/li\u003e\n \u003cli\u003eGreenwald, N. F.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e Whole-cell segmentation of tissue images with human-level performance using large-scale data annotation and deep learning. \u003cem\u003eNat Biotechnol\u003c/em\u003e \u003cstrong\u003e40\u003c/strong\u003e, 555-565 (2022). https://doi.org:10.1038/s41587-021-01094-0\u003c/li\u003e\n \u003cli\u003eZheng, G. X.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e Massively parallel digital transcriptional profiling of single cells. \u003cem\u003eNat Commun\u003c/em\u003e \u003cstrong\u003e8\u003c/strong\u003e, 14049 (2017). https://doi.org:10.1038/ncomms14049\u003c/li\u003e\n\u003c/ol\u003e"},{"header":"Methods","content":"\u003cp\u003e\u003cstrong\u003ePatients and specimens\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eBrain tissue samples from IS patients were obtained during decompressive craniectomy performed for large-area cerebral infarctions with associated intracranial hypertension. The specimens consisted of small portions of ischemic and necrotic brain tissue that were surgically removed. Control brain tissues were collected from patients with intracerebral hemorrhage during the creation of the surgical access channel for hematoma evacuation, and these samples represented relatively normal brain regions adjacent to the hematoma site. Peripheral blood samples were collected from IS patients and healthy donors. All samples were obtained from the Department of Neurosurgery and Emergency Department, Zhujiang Hospital, Southern Medical University. Written informed consent was obtained from all participants. The study was approved by the Ethics Committee of Zhujiang Hospital, Southern Medical University (Approval NO. 2025-KY-012-02).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eMice\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eC57BL/6J strains of mice used in this study were obtained directly from the Zhuhai Baishitong Biotechnology Co., Ltd. (Zhuhai, China). Experiments were performed in young C57BL/6Smoc-Pdcd1\u003csup\u003eem3Smoc\u003c/sup\u003e mice (Pd-1-KO mice, Cat. NO. NM-KO-190423) were purchased from Shanghai Model Organisms Center, Inc. (Shanghai, China). The source of B6. Cg-c/c Tg(CAG-tdKaede)15Utr mice is RIKEN BioResource Center. Mice were housed in individually ventilated cages under standard conditions at 22 °C with 40 ± 5% relative humidity and a 12-h light/12-h dark cycle. Water and a standard laboratory diet were available ad libitum, unless indicated otherwise. 8-12-week-old mice were used for all experiments. All mice were acclimatized for one week before the initiation of the experiments. All animal maintenance protocols and procedures performed were approved by the Ethics Committee of the Animal Experimental Center of Zhujiang Hospital, Southern Medical University (Approval NO. LAEC-2024-038; Guangzhou, China).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eMice experiments\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTransient focal cerebral ischemia was induced in mice using the middle cerebral artery occlusion (MCAO) model as previously described\u003csup\u003e26\u003c/sup\u003e. Mice were anesthetized via intraperitoneal injection of tribromoethanol (1.25%, 0.02 ml/g body weight). A midline cervical incision was made to expose the right common carotid artery (CCA), internal carotid artery (ICA), and external carotid artery (ECA). The CCA and ECA were ligated, and a silicon-coated nylon filament was inserted through the ICA to occlude the origin of the middle cerebral artery (MCA). After 60 min of occlusion, the filament was carefully withdrawn to allow reperfusion. Throughout the procedure, body temperature was maintained at 37 °C using a heating pad equipped with a rectal probe. Sham-operated mice as control underwent the same surgical exposure of the carotid arteries without insertion of the filament.\u003c/p\u003e\n\u003cp\u003eNeurological function was assessed using a composite scoring system incorporating three established behavioral evaluations: Bederson's postural reflex test, Longa's tail suspension test, and Feeney's balance-beam walking test\u003csup\u003e27,28\u003c/sup\u003e. The final neurological score was calculated as the sum of scores from these three assessments: Bederson's postural reflex test (0: No observable deficits; 1: Impaired straight-line ambulation; 2: Ipsilateral rotational behavior; 3: Lateralized postural collapse) , Longa's tail suspension test (0: Symmetrical forelimb extension; 1: Contralateral forelimb flexion during elevation; 2: Sustained contralateral flexion without rotation; 3: Spontaneous ipsiversive circling), Feeney's beam walking test (0: Stable quadrupedal stance; 1: Grasping beam edges with all limbs; 2: Unilateral limb disengagement; 3: Bilateral limb slippage (\u0026gt;60 s retention); 4: Beam maintenance \u0026gt;40 s before fall; 5: Beam maintenance \u0026gt;20 s before fall; 6: Immediate postural failure).\u003c/p\u003e\n\u003cp\u003eFor photoconversion, Kaede transgenic mice were anesthetized with a ketamine/xylazine mixture (10 mg/kg and 2 mg/kg, respectively, intraperitoneally) or with 3% isoflurane in 100% O₂ at a flow rate of 2 L/min. As previously described\u003csup\u003e22\u003c/sup\u003e, mice were placed on their backs with an aluminum foil blanket covering the surrounding tissue, the small intestines were exposed to violet light was shone (handheld 405-nm laser; peak power \u0026lt;5 mW) onto the exposed area for a total of 10 min with brief pauses every 3 min, Throughout the procedure, the exposed intestine tissue was kept moist using sterile PBS to prevent desiccation.\u003c/p\u003e\n\u003cp\u003eAs for intraventricular administration, mice were anesthetized with tribromoethanol (1.25%, 0.02 ml/g body weight, intraperitoneally) and positioned in a stereotaxic apparatus (UMP3 microinjection system; WPI, Sarasota, FL, USA). Anti-mouse PD-1 monoclonal antibody, or anti-mouse PD-L1 monoclonal antibody (BioXCell) was delivered into the lateral ventricle using a 10 μL microsyringe fitted with a 33-gauge needle (RWD, China). The injection coordinates relative to bregma were: +0.3 mm anterior posterior, -1.0 mm lateral and -2.5 mm ventral.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCell suspension preparation for multiple organs and peripheral blood of mice\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eMice were anesthetized with tribromoethanol (0.2 mL/10g body weight) and transcardially perfused with cold PBS. Following perfusion, the right cerebral cortex was visually inspected to confirm the presence of ischemic lesions. Subsequently, brain tissue, spleen, lung, heart, small intestine, liver, and peripheral blood were collected.\u003c/p\u003e\n\u003cp\u003eFor mice brains, brain from control-sham or ischemic mice were dissected free of the cerebellum and olfactory bulb, minced in cold RPMI 1640 supplemented with 2% FBS, and digested for 45 min at 37°C in 5 ml of digestion medium (RPMI 1640, 2% FBS, 2 mg/mL Collagenase type D, 2 mg/mL DNase I) with constant stirring at 200 rpm. The digested tissue was washed with PBS, sequentially filtered through a 70 μm cell strainer, centrifuged, and resuspended in PBS containing 1% FBS.\u003c/p\u003e\n\u003cp\u003eSpleens were mechanically dissociated through a 70-μm cell strainer into RPMI 1640 medium supplemented with 2% FBS. Red blood cells were lysed using ammonium-chloride-potassium (ACK) lysis buffer for 3 min at room temperature, followed by washing with RPMI 1640. The resulting single-cell suspension was resuspended in PBS containing 1% FBS.\u003c/p\u003e\n\u003cp\u003eSmall intestines were flushed with cold PBS, and adipose tissue and intestinal contents were removed. Tissues were cut into small pieces, and incubated in extraction buffer (RPMI 1640, 2% FBS, 5 mM DTT, 1 mM EDTA) at 37°C for 15 min twice with stirring to remove intraepithelial lymphocytes. Remaining fragments were washed and digested in digestion buffer (RPMI 1640, 2% FBS, 1 mg/mL collagenase II, 0.5 mg/mL Dispase). The suspension was filtered through a 70 μm strainer, layered on between 40% and 80% Percoll, and centrifuged at 1,300 × g. Cells at the interface were collected and resuspended in PBS containing 1% FBS.\u003c/p\u003e\n\u003cp\u003eLung and heart tissues were harvested after perfusion and were disrupted into small pieces, then were digested in 10 ml digestion medium (RPMI 1640, 2% FBS, 1mg/ml Collagenase type I, 0.5mg/ml Dispase; all Gibco) for 30 min at 37 °C with stirring at 200 rpm. The tissues were neutralized with PBS, then filtered through a 70 µm strainer, centrifuged, and resuspended in PBS with 1% FBS.\u003c/p\u003e\n\u003cp\u003eLiver tissues were thoroughly minced in a digestive solution containing 0.5 mg/mL collagenase IV, 150ug/mL DNase I, 2% FBS, RPMI 1640 and incubated for 40 min at 37 °C on a shaker at 180 rpm. Afterwards, samples were filtered through a 70 mm cell strainer to obtain final single cell suspensions. \u003c/p\u003e\n\u003cp\u003eFresh peripheral blood underwent red blood lysis using ammonium-chloride-potassium (ACK) lysis buffer for 3 minutes at room temperature, followed by washing with RPMI. The resulting single-cell suspension was resuspended in PBS with 1% FBS.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAntibiotic-mediated microbiota depletion\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo deplete gut microbiota, male C57BL/6J mice (6-8 weeks old) were administered a broad-spectrum antibiotic cocktail consisting of vancomycin (100 mg/kg/day), ampicillin, neomycin and metronidazole (200mg/kg/day, Aladdin Reagent) by daily oral gavage (200 µl) for 7 consecutive days. Control mice received an equivalent volume of sterile PBS following the same procedure.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFecal microbiota transplantation (FMT)\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eFMT was performed to assess the impact of gut microbiota from individuals with ischemic stroke and healthy controls. Following antibiotic treatment and a 2-day washout period, Recipient mice were then orally gavaged (200 µl) once daily for 7 days with fecal suspensions prepared from stool samples of patients with ischemic stroke (n = 6) or healthy controls (n = 6). A total of 3,000 mg of fresh feces was resuspended in 30 ml sterile saline supplemented with 20% (v/v) glycerol and 0.5 g/l cysteine-HCl (Sigma-Aldrich) under anaerobic conditions. Suspensions were homogenized, filtered through a 100 µm cell strainer, aliquoted, and stored at −80 °C until use.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eImaging cell migration\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo track immune cell migration in \u003cem\u003evivo\u003c/em\u003e, splenocytes were isolated from male C57BL/6J mice. Briefly, spleens were dissected and mechanically dissociated through a 70μm cell strainer. Erythrocytes were lysed using ACK Lysis Buffer (Leagene, Beijing, China). Isolated cells were labeled with the Cy5-conjugated cholesteryl oligonucleotide probe 5′-Cholesteryl-AAAAAAAAAAA-3′-Cy5 (Sangon Biotech, Shanghai, China) via 5 min incubation. Subsequently, Recipient mice were administered by 100μL of labeled cell suspension through tail vein injection. Recipient mice underwent MCAO or sham surgery, followed by brain collection at 3-, 24-, 48-, and 72-hours post-procedure. Fluorescent signals were acquired using an IVIS Spectrum imaging system (PerkinElmer, Boston, MA, USA).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eImmunohistochemistry\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eFormalin-fixed, paraffin-embedded tissue sections were deparaffinized with xylene and rehydrated through decreasing concentrations of ethanol, respectively. Endogenous peroxidase activity was quenched with 0.3% H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2\u003c/sub\u003e in methanol (Merck Millipore, Burlington, MA) for 15 min at RT. Sections were boiled in Tris-EDTA (10 mM/1 mM, pH 9) buffers for antigen retrieval followed by cooling down to RT. Then tissue sections were blocked with Superblock solution (Thermo Fisher Scientific, Waltham, MA, USA) and incubated overnight at 4°C with a primary antibody (anti-mouse CD45，anti-human CD45，anti-mouse Notch1 and anti-mouse Notch2 antibody, Cell Signaling Technology). After three PBS washes, sections were incubated with poly-HRP-conjugated secondary antibody (Immunologic, Duiven, Netherlands) for 1h at RT. Signal detection used DAB+ chromogen (DAKO, Agilent Technologies, Santa Clara, CA), and nuclei were counterstained with Mayer’s hematoxylin (Thermo Fisher Scientific). Data analysis was utilized by SlideViewer software (version 2.7.0.191696).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eMultiplex immunohistochemistry\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTissue section preparation followed standard IHC procedures before the Opal multiplex antibody staining protocol. Tissue sections were circled with a hydrophobic barrier pen Primary antibodies were applied and incubated under optimized conditions. After washing in TBST (3 times × 2 min), slides were incubated with Opal Polymer HRP Ms+Rb for 10 min. Following another TBST wash, Opal fluorophore working solution was applied for 10 min. Slides were then rinsed in AR buffer, and antigen retrieval was repeated by microwave heating as described. These steps-including blocking, primary and secondary antibody incubation, fluorophore labeling, and antigen retrieval-were repeated iteratively for each target marker. For Opal Polaris 780, TSA-DIG working solution was applied, followed by antigen retrieval and incubation with Opal Polaris 780 fluorophore for 1 h. Nuclear counterstaining was performed with DAPI (5 min, RT), followed by washing and covers lipping with anti-fade mounting medium. Slides were either scanned immediately or stored at 4°C in the dark.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eImaging mass cytometric immunostaining, acquisition and analysis \u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eImaging mass cytometry (IMC) staining on formalin-fixed paraffin-embedded (FFPE) sections of human brain biopsies was performed as previously described\u003csup\u003e29\u003c/sup\u003e. FFPE tissue blocks were sectioned at 4 µm thickness. These sections underwent a heating process at 68°C for 1 h to facilitate dewaxing, followed by two 10-minute incubations in xylene at 68°C for complete paraffin removal. Rehydration was achieved through sequential immersions in ethanol solutions of decreasing concentrations (95%, 85%, and 75%), each for 5 min at RT. Antigen retrieval was conducted by heating sections in 10 mM sodium citrate buffer (pH 6.0) at 100°C for 30 min. After cooling to RT, sections were washed twice in PBS-TB (PBS with 0.5% Tween-20 and 1% BSA) for 5 min each. Blocking was performed using SuperBlock buffer (Thermo Fisher Scientific) for 30 min at RT. Sections were incubated with metal-tagged antibodies overnight at 4°C as indicated in Table 1. Post-incubation, samples were washed three times with PBS-TB. Nuclear labeling was performed with an Intercalator-Ir (Fluidigm) solution in PBS-TB (1.25 µM) for 30 min at RT, followed by two additional PBS-TB washes and a final rinse with deionized water.\u003c/p\u003e\n\u003cp\u003eThe tissue sections were ablated at 200 Hz on a Helios time-of-flight mass cytometer coupled to a Hyperion Imaging System (Fluidigm). Instrument calibration and autotuning were performed according to the manufacturer’s protocol, using a 3-element tuning slide (Fluidigm), as described in the Hyperion Imaging System User Guide. The regions of interest (ROIs) per sample were selected based on CD45 immunohistochemistry staining on consecutive slides. All raw data were analyzed for marker intensity based on the maximum signal threshold, defined at the 98th percentile of all pixels in a single ROI using Fluidigm MCD\u003csup\u003eTM\u003c/sup\u003e viewer (v1.0.560.2).\u003c/p\u003e\n\u003cp\u003eFor cell segmentation analysis of IMC datasets, we utilized the pre-trained TissueNet, as described previously\u003csup\u003e30\u003c/sup\u003e. The TissueNet requires two channels of imaging data for its operation. TissueNet requires two imaging channels: the nuclear channel, typically stained with Intercalator-Ir, to identify cell nuclei; and a second channel representing either membrane or cytoplasmic markers to delineate cell boundaries. In this study, membrane markers were used as the second input channel to improve boundary resolution. Marker expression data were transformed using a hyperbolic arcsine function and trimmed at the 1\u003csup\u003est\u003c/sup\u003e and 99\u003csup\u003eth\u003c/sup\u003e percentiles to define the range. Each channel was subsequently normalized using min-max scaling. To correct for batch effects across samples, we applied the Harmony algorithm (version 0.1.0) from the R package. Subsequent cell clustering was performed using the Rphenograph (version 0.99.1), with the number of nearest neighbors set to 30. Clustering was guided by specific markers, including CD14, GFAP, CD8, CD31, S100β, HIF1α, CD45, CD7, CD163, MAP2, CD68, CD11c, IbaI, P2Y12, HLA-DR, CD3, Olig2, αSMA, Vimentin, CD16, CD4, CD11b, CD206 and CD45RO to identify major cell populations. The mean expression of each marker across clusters was visualized on a heatmap (Fig. 1c). This expression pattern was then utilized for further annotation. Resulted images and tSNE feature plots were exported by relative R packages.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFlow cytometry\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eSingle cell suspension from different mice tissues or human whole blood cells were incubated with fluorochrome-conjugated antibodies and Fc block (BioLegend) for 30 min at 4 °C for surface staining. After washing, the stained cells were resuspended. Reference samples were incorporated and individually stained by UltraComp eBeadsTM Compensation Beads (ThermoFisher), whole blood cells from mice or human. After stained samples preparation, the samples were immediately acquired using a 5-laser Cytek® Aurora (Cytek® Biosciences). Data was analyzed to check quality with FlowJo software version 10.6 (Tree Star Inc). We utilized OMIQ to perform the dimensional reduction analysis for mice and human samples (https://www.omiq.ai/). Detailed information regarding antibody panels is listed in Table 2-5.\u003c/p\u003e\n\u003cp\u003eFor microglia sorting, brain cell suspension were stained with surface staining flow cytometry antibodies for 30 min at 4 °C in the dark, then purified on BD Influx cell sorter (BD Biosciences). PD-L1\u003csup\u003e+\u003c/sup\u003e microglia were identified as CD45\u003csup\u003edim\u003c/sup\u003eCD11b\u003csup\u003e+\u003c/sup\u003ePD-L1\u003csup\u003e+\u003c/sup\u003e; PD-L1\u003csup\u003e-\u003c/sup\u003e microglia were identified as CD45\u003csup\u003edim\u003c/sup\u003eCD11b\u003csup\u003e+\u003c/sup\u003ePD-L1\u003csup\u003e-\u003c/sup\u003e. Detailed information regarding antibodies is listed in Table 6. \u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eTranscriptome sequencing experiment and analysis\u003c/strong\u003e \u003c/p\u003e\n\u003cp\u003eBrain tissues were collected immediately post-euthanasia, snap-frozen in liquid nitrogen, and stored at -80°C until processing. Total RNA was extracted and quantified via Bioanalyzer 2100 (Agilent Technologies, Santa Clara, CA). RNA was reverse-transcribed into cDNA for library construction. Library fragments were size-selected (150-200 bp) and purified with AMPure XP system (Beckman Coulter, Brea, CA). Indexed libraries were clustered on a cBot Cluster Generation System with the TruSeq PE Cluster Kit v3-cBot-HS (Illumina) following the manufacturer’s instructions, then the library preparations were sequenced on an Illumina NovaSeq platform, resulting sequencing data for all samples with a read length of 150 bp (paired-end). Raw sequencing data underwent quality control. Gene-level counts were generated with FeatureCounts v1.5.0-p3. Subsequently, Normalized FPKM values were calculated for gene expression quantification. Differential expression analysis between different groups was performed using DESeq2. Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis was performed using the clusterProfler R package to test the statistical enrichment of differential expression genes.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eSingle-cell RNA sequencing\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cem\u003ePdcd1\u003c/em\u003e gene expression in mice brain subsets was from a publicly accessible interactive web portal of previous published scRNAseq datasets (https://anratherlab.shinyapps.io/strokevis/)\u003csup\u003e9\u003c/sup\u003e.\u003c/p\u003e\n\u003cp\u003eSingle, live, PD-1\u003csup\u003e+\u003c/sup\u003e and PD-L1\u003csup\u003e-\u003c/sup\u003e microglia in brain from MCAO mice were sorted using a FACSAria III flow cytometer (BD Biosciences). Single-cell RNA sequencing was performed as described previously\u003csup\u003e31\u003c/sup\u003e. In brief, single-cell suspensions were loaded onto a Chromium Single Cell Controller (10x Genomics) together with oil, reagents, and barcoded beads. Cell lysis and barcoded reverse transcription of polyadenylated mRNA were carried out within individual gel bead emulsions. cDNA libraries were generated in a single bulk reaction, and sequencing was performed on an Illumina HiSeq 4000 platform. \u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eQuantitative real-time PCR\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTotal RNA was extracted from brain tissue samples using the TRIzol method. cDNA was synthesized from 500ng total RNA using the PrimeScript RT reagent Kit (TaKaRa, Japan). Quantitative real-time PCR was performed using the TB Green Kit (Takara, RR820A) on an ABI Step One RT-PCR system (Applied Biosystems, Foster City, USA) under the following conditions: 95°C for 1 min; 40 cycles of 95°C for 15 s; then 60°C for 1 min. Relative gene expression levels were calculated using the 2\u003csup\u003e-ΔΔCT\u003c/sup\u003e method, with β-actin primers serving as the internal control. Primer sequences are provided in Table 7.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eMicrobial 16S rRNA Gene Sequencing and Bioinformatic Analysis\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eIleal fecal contents from HC-FMT and IS-FMT MCAO mice were collected under sterile conditions as previously described. Microbial genomic DNA was extracted using the E.Z.N.A.® Soil DNA Kit (Omega Bio-tek), following the manufacturer’s protocol. The V3–V4 region of the 16S rRNA gene was amplified and sequenced on the Illumina NextSeq 2000 platform at Majorbio Bio-Pharm Technology Co. Ltd. (Shanghai, China). Raw reads were processed using fastp (v0.19.6) for quality filtering and merged using FLASH (v1.2.7). Operational taxonomic units (OTUs) were clustered at 97% sequence identity. Taxonomic assignment was performed using the RDP Classifier (v2.2) against the SILVA v138 database. Alpha diversity was assessed by calculating Chao1 and Shannon indices using Mothur (v1.30.1). Differences between groups were evaluated by unpaired Student’s t-test. Beta diversity was analyzed using principal coordinate analysis (PCoA) based on Bray–Curtis distance matrices, with significance assessed by PERMANOVA (Vegan, v2.5-3). The relative abundance of bacterial genera was visualized with bar plots generated using the Majorbio Cloud Platform.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eStatistical analysis\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eStatistical analyses were performed using GraphPad Prism 10 and R. Bar graphs and dot plots display mean values, with error bars representing the standard error of the mean (s.e.m.). For pooled analyses, data from all mice within the same experimental group across independent repeats were aggregated, and statistical comparisons were conducted on the pooled dataset as for individual experiments. All measurements were obtained from distinct samples. Statistical significance was assessed using the Wilcoxon signed-rank test and Mann–Whitney U test, as appropriate. A two-tailed P value \u0026lt; 0.05 was considered significant; significance levels are indicated as follows: *P \u0026lt; 0.05, **P \u0026lt; 0.01, ***P \u0026lt; 0.001, ****P \u0026lt; 0.0001.\u003c/p\u003e"},{"header":"Tables","content":"\u003cp\u003eTables 1 to 7 are available in the Supplementary Files section\u003c/p\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-6952560/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-6952560/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eIschemic stroke disrupts neuroimmune homeostasis, leading to pronounced neurological impairment, yet the cellular mechanisms remain incompletely defined. Here, we construct a high-resolution spatiotemporal immune atlas of infarcted brain regions from post-stroke patients using imaging mass cytometry, revealing a marked perivascular accumulation of PD-1⁺ T cells. Parallel investigations in a murine model of middle cerebral artery occlusion (MCAO) recapitulate this central enrichment, with PD-1⁺ T cell infiltration peaking at three days post-ischemia and persisting for up to fourteen days. Spatiotemporal fate mapping with Kaede photoconvertible reporter mice demonstrates that intestinal-derived PD-1⁺ T cells infiltrate the injured brain in a gut microbiota–dependent manner and engage directly with brain-resident PD-L1⁺ microglia. Both genetic and pharmacological disruption of the gut–brain PD-1/PD-L1 axis markedly reduces the infiltration of gut-derived T cells and ameliorates neurological deficits. Single-cell RNA sequencing of PD-L1⁺ microglia identifies Notch1-driven inflammatory signaling as a key mediator of neuroinflammation. Notably, acute stroke patients exhibit a significant increase in circulating PD-1⁺ T cells, serving as a diagnostic indicator of central nervous system injury. Together, these findings define a previously unrecognized gut–brain immune circuit that orchestrates post-stroke neuroinflammation and highlight circulating PD-1⁺ T cells as a potential biomarker of neurological damage.\u003c/p\u003e","manuscriptTitle":"Gut microbiota mediates trafficking of intestinal T cells to the brain to exacerbate neuroinflammation after ischemic stroke","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-06-29 15:02:10","doi":"10.21203/rs.3.rs-6952560/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"894626f6-ffdf-4f7c-80e0-b0389e7a1ff3","owner":[],"postedDate":"June 29th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[{"id":50594673,"name":"Biological sciences/Immunology/Neuroimmunology"},{"id":50594674,"name":"Biological sciences/Neuroscience/Blood\u0026#x2013;brain barrier"}],"tags":[],"updatedAt":"2025-09-04T12:26:16+00:00","versionOfRecord":[],"versionCreatedAt":"2025-06-29 15:02:10","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-6952560","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-6952560","identity":"rs-6952560","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}