Abstract

   Acute ischemic stroke is a leading cause of mortality and disability
   worldwide. Neuroinflammation following ischemia‐reperfusion plays a
   critical role in the disease's pathogenesis. Neutrophil aggregation and
   clearance within the brain parenchyma influence neuroinflammatory
   damage during ischemic stroke. Microglia‐mediated phagocytosis plays a
   pivotal role in mitigating neuroinflammation and promoting brain
   parenchyma recovery. However, the mechanisms underlying the cross‐talk
   between neutrophils and microglia remain poorly understood. Here, this
   study demonstrates that neutrophils can trigger microglial functional
   change to inhibit microglial phagocytosis and promote pyroptosis, which
   is regulated by neutrophil‐derived myeloid‐related protein 14.
   Additionally, interleukin‐1β released by pyroptotic microglia further
   upregulates myeloid‐related protein 14 expression and facilitates
   neutrophil mobilization from the bone marrow, establishing a
   self‐sustaining inflammatory loop. Therefore, neutrophils accumulate in
   the brain parenchyma and further exacerbate microglial
   neuroinflammation in the ischemic brain. These findings reveal a
   previously unknown interaction between neutrophils and microglia after
   acute ischemic stroke and suggest that targeting myeloid‐related
   protein 14 may provide a novel therapeutic strategy for ischemic stroke
   therapy.

   Keywords: ischemia‐reperfusion injury, microglial phagocytosis,
   microglial pyroptosis, neutrophil
     __________________________________________________________________

   These study underscores the pivotal role of neutrophil‐derived
   myeloid‐related protein 14 (MRP14) in the pathophysiology of acute
   ischemic stroke (AIS). MRP14 not only disrupts mitochondrial function,
   thereby inhibiting microglial phagocytosis of neutrophils, but also
   activates microglial pyroptosis, exacerbating neuroinflammation
   following AIS. Targeting MRP14 could offer a promising therapeutic
   strategy to mitigate neuroinflammation and improve outcomes for
   patients with AIS.

   graphic file with name ADVS-12-e03722-g005.jpg

1. Introduction

   Acute ischemic stroke (AIS) is a leading cause of mortality and
   disability worldwide.^[ [44]^1 ^] Extensive research has demonstrated
   that neuroinflammation, driven by activated immune cells following
   ischemia‐reperfusion, plays a critical role in the pathogenesis of
   ischemic stroke, contributing to secondary brain injury.^[ [45]^2 ,
   [46]^3 ^] Neutrophils, as key neuroinflammatory mediators, are among
   the first peripheral immune cells to infiltrate the brain, arriving
   within minutes to hours after stroke onset. Their rapid mobilization
   from the bone marrow and spleen is triggered by chemokines and
   inflammatory cytokines released by ischemic brain tissue.^[ [47]^4 ^]
   Within the ischemic microenvironment, neutrophils become hyperactivated
   and release proteolytic enzymes, which exacerbate the recruitment of
   additional peripheral immune cells and the activation of brain‐resident
   immune cells; this cascade further amplifies neuroinflammation.^[
   [48]^4 , [49]^5 , [50]^6 ^] Previous studies, including those from our
   laboratory, have demonstrated that elevated peripheral neutrophil
   counts during the early stages of AIS are associated with increased
   cerebral edema, larger infarct volumes, and poorer prognosis.^[ [51]^7
   , [52]^8 ^] Although experimental strategies aimed at reducing
   neutrophil migration and infiltration have shown promise in limiting
   brain injury, clinical trials have not yielded encouraging results,^[
   [53]^9 ^] primarily due to insufficient focus on the clearance of
   infiltrating neutrophils within ischemic brain tissue.

   Microglia, as the principal resident immune cells of the central
   nervous system, serve as primary responders following ischemic stroke.
   Microglia‐mediated phagocytosis plays a pivotal role in mitigating
   neuroinflammation and aiding brain parenchymal recovery. By engulfing
   apoptotic cells before they undergo secondary necrosis, which would
   otherwise release harmful intracellular components, microglia help
   modulate the overall inflammatory response.^[ [54]^10 ^] Research has
   shown that microglia can actively phagocytize both apoptotic and viable
   neutrophils recruited to the ischemic brain.^[ [55]^11 ^] However,
   ischemia significantly impairs microglial phagocytic function due to
   energy depletion and cellular atrophy, resulting in increased
   neutrophil accumulation and exacerbation of brain injury.^[ [56]^10 ,
   [57]^11 ^] Therefore, enhancing microglial phagocytosis of neutrophils
   presents a promising therapeutic strategy for reducing brain injury and
   improving outcomes in ischemic stroke.

   This study found novel interactions between neutrophils and microglia
   in both peripheral and intracranial immune responses after ischemic
   stroke. Our findings revealed that neutrophils can trigger microglial
   functional change to inhibit their phagocytic activity and promote
   pyroptosis, which is regulated by neutrophil‐derived myeloid‐related
   protein 14 (MRP14, also known as S100A9). In turn, interleukin‐1β
   (IL‐1β), released by pyroptotic microglia, further stimulates
   neutrophil mobilization from the bone marrow and upregulates MRP14
   expression, creating a self‐sustaining inflammatory loop. This
   accumulation of neutrophils and pyroptotic microglia amplifies the
   inflammatory cascade, ultimately exacerbating blood‐brain barrier (BBB)
   disruption in ischemic stroke.

2. Results

2.1. Neutrophil Mobilization was Associated with Cerebral Edema and
Microglial Function Changes after AIS

   A total of 414 AIS patients, who were consecutively enrolled from the
   Multicenter Clinical Trial of Revascularization Treatment for Acute
   Ischemic Stroke (TRAIS) cohort between January 1, 2019, and December
   30, 2021, were analyzed to assess the association between neutrophil
   counts and cerebral edema severity (clinical characteristics are shown
   in Table [58]S1, Supporting Information). As depicted in Figure [59]
   1A, peripheral neutrophil counts measured within 24 h of stroke onset
   demonstrated a positive correlation with cerebral edema grade. To
   further validate this association, we injected mice with an anti‐Ly6G
   antibody to deplete neutrophils before undergoing transient middle
   cerebral artery occlusion (tMCAO). Compared with vehicle group, the
   inhibition of neutrophils significantly reduced the ipsilateral
   hemisphere brain water content, indicating the alleviation of cerebral
   edema (Figure [60]1B).

Figure 1.

   Figure 1
   [61]Open in a new tab

   Neutrophil mobilization was associated with cerebral edema and
   microglial function changes after AIS. A) The number of peripheral
   neutrophils in patients with AIS with different cerebral edema grades.
   B) Brain water content of the contralateral cerebral hemisphere,
   ipsilateral cerebral hemisphere and cerebellum in the tMCAO mice after
   IgG or Anti‐Ly6G treatment (n = 6/group). C) Illustration of the
   penumbra and ischemic core in tMCAO mice. D‐F) The number of
   neutrophils and microglia in the penumbra and ischemic core of tMCAO
   mice after IgG or Anti‐Ly6G treatment (n = 6/group). G,H)
   Immunofluorescence staining showing the expression of CD68 in microglia
   in the penumbra and ischemic core of the tMCAO mice after IgG or
   Anti‐Ly6G treatment (n = 6/group). Data are presented as means ± SEM;
   one‐way ANOVA, two‐way ANOVA. **P < 0.01. Scale bar: 100 µm. AIS, acute
   ischemic stroke; tMCAO, transient middle cerebral artery occlusion.

   Next, we investigated the mechanisms by which neutrophils contribute to
   cerebral edema. Previous studies have demonstrated that, upon
   infiltrating the brain parenchyma, neutrophils interact with microglia
   and modulate their function following AIS. By performing
   double‐staining for neutrophils and microglia, we observed a higher
   density of neutrophils and a lower number of microglia in the ischemic
   core relative to the penumbra (Figure [62]1C‐F). Microglia within the
   core also exhibited signs of dystrophy and cell loss. Interestingly,
   neutrophil depletion led to an increase in microglial numbers in the
   ischemic core (Figure [63]1C‐F), suggesting a potential role for
   neutrophils in modulating microglial survival. In addition, we found
   three positional relationships between neutrophils and microglia in the
   ischemic brain: free, touching, and surrounded states (Figure
   [64]S1A,B, Supporting Information). The latter two patterns are
   indicative of microglial phagocytosis of neutrophils. Assessing
   phagocytic function, we found that neutrophil depletion significantly
   increased CD68 intensity in microglia in both the core and ischemic
   penumbra (Figure [65]1G,H), suggesting that neutrophils also influence
   microglial phagocytosis. Collectively, these results delineated a
   potential interaction between neutrophils and microglia that may
   contribute to the pathogenesis of cerebral edema after AIS.

2.2. Neutrophil MRP14 Expression is Elevated in Patients with AIS Exhibiting
Severe Cerebral Edema

   To determine the role of neutrophils in cerebral edema and microglial
   function, peripheral neutrophils were collected from six patients with
   AIS exhibiting either mild or severe cerebral edema. After matching for
   clinical characteristics, RNA sequencing was performed. A total of 1319
   differentially expressed genes (DEGs) were identified between the mild
   and severe cerebral edema groups (|Log2FC| > 1). REACTOME analysis of
   these DEGs identified “Neutrophil degranulation” as the most
   significantly enriched pathway (Figure [66] 2A). Proteomic analysis was
   conducted using plasma samples from the aforementioned two groups. This
   analysis revealed 63 differentially expressed proteins between the two
   groups (|Log2FC| > 1), with REACTOME enrichment highlighting “Immune
   system”, “Innate immune system”, and “Neutrophil degranulation” as the
   top three enriched pathways (Figure [67]2B). These findings highlight
   the critical role of neutrophil‐driven inflammation in cerebral edema.
   Among the DEGs and differentially expressed proteins, MRP14 and
   peptidoglycan recognition protein 1 (PGLYRP1) emerged as overlapping
   genes within the “Neutrophil degranulation” pathway identified in both
   transcriptomic and proteomic analyses (Figure [68]2C). Upon expanding
   the sample size, plasma levels of MRP14 were significantly higher in
   patients with severe cerebral edema compared with those with mild edema
   (Figure [69]2D), whereas PGLYRP1 levels did not differ significantly
   between the groups (Figure [70]2E). These results suggest that elevated
   plasma MRP14 levels are associated with severe cerebral edema in AIS.

Figure 2.

   Figure 2
   [71]Open in a new tab

   MRP14 expression is elevated in patients with AIS exhibiting severe
   cerebral edema and in the tMCAO mice. A) The REACTOME pathway
   enrichment analysis of differentially expressed genes in the peripheral
   neutrophils of patients with AIS exhibiting either mild or severe
   cerebral edema in the RNA‐sequencing. B) The REACTOME pathway
   enrichment analysis of differentially expressed proteins in the plasma
   proteomics of patients with AIS exhibiting either mild or severe
   cerebral edema. C) The Venn diagram showing the overlapped genes
   between the “Neutrophil degranulation” pathway in the RNA‐sequencing
   and proteomics. D,E) The ELISA test of MRP14 and PGLYRP1 level in the
   plasma of patients with AIS exhibiting either mild or severe cerebral
   edema (mild, n = 53; severe, n = 35). F,G) Western blot analysis of the
   MRP14 protein level in the ischemic brain of the tMCAO mice at
   different time points (n = 6/group). H) The mRNA level of MRP14 in the
   ischemic brain of the tMCAO mice at different time points (n =
   6/group). I) Quantification of the plasma MRP14 level by ELISA assay of
   the tMCAO mice at different time points (n = 6/group). J,K)
   Immunofluorescence staining showing the MRP14^+ neutrophils in the
   ischemic brain of the tMCAO mice at different time points (n =
   6/group). Data are presented as means ± SEM; unpaired t‐test, one‐way
   ANOVA. *P < 0.05, **P < 0.01. Scale bar: 25 µm. AIS, acute ischemic
   stroke; tMCAO, transient middle cerebral artery occlusion; ELISA,
   enzyme‐linked immunosorbent assay.

2.3. Neutrophil MRP14 Expression is Increased in the Ischemic Brain of tMCAO
Mice

   We further validated MRP14 expression after ischemic stroke using two
   independent datasets. The first was our own RNA sequencing dataset from
   peripheral neutrophils isolated from sham and tMCAO mice 24 h after
   stroke (Figure [72]S2A,B, Supporting Information). The second dataset
   ([73]GSE174440) consisted of single‐cell RNA sequencing (scRNA‐seq)
   data from sorted blood myeloid cells of mice at steady state and 24 h
   after tMCAO (Figure [74]S2C, Supporting Information). Both datasets
   showed elevated neutrophil MRP14 expression following ischemic stroke
   (Figure [75]S2B,C, Supporting Information).

   Next, we collected ischemic brain tissues and plasma from tMCAO mice at
   various time points to assess changes in MRP14 expression. Both protein
   and mRNA levels of MRP14 in the ischemic brain started to rise at 6 h
   post‐tMCAO, peaking at 24 h (Figure [76]2F‐H). In plasma, MRP14 levels
   started to increase at 6 h, reaching a peak at 12 h post‐tMCAO
   (Figure [77]2I). To further determine the cellular source of MRP14 in
   the ischemic brain, double immunostaining was performed using the
   neutrophil marker Ly6G and the microglial marker Iba‐1. The results
   demonstrated that MRP14 predominantly co‐localized with neutrophils
   (Figure [78]2J,K), with minimal expression observed in microglia at 3
   days post‐tMCAO (Figure [79]S3, Supporting Information). These results
   provide strong evidence that neutrophils are the primary source of
   elevated MRP14 in the ischemic brain following tMCAO.

2.4. Both paquinimod Treatment and MRP14 Knockout Attenuate BBB Disruption
after Ischemic Stroke

   The disruption of BBB integrity occurs 24–72 h post‐ischemic stroke,^[
   [80]^12 ^] corresponding to elevated intracranial levels of MRP14. To
   investigate whether increased MRP14 contributes to pathological BBB
   leakage, the MRP14 inhibitor paquinimod was intravenously administered
   into mice, to inhibit extracellular MRP14 function. BBB integrity and
   infarct area were assessed 1 day after tMCAO, and neurological deficits
   were assessed 1, 3, 7, 14, 21, and 28 days after tMCAO (Figure [81]
   3A). Mice subjected to tMCAO exhibited enhanced extravasation of Evans
   blue and fibrinogen, indicating BBB disruption, while paquinimod
   treatment significantly attenuated this leakage (Figure [82]3B,C,E,F).
   Additionally, expression of tight junction proteins (ZO‐1, Occludin,
   VE‐Cadherin) was reduced in tMCAO mice, which was counteracted by
   paquinimod treatment (Figure [83]3H‐K,O,P). Paquinimod also mitigated
   infarct area and neurological outcomes in these mice
   (Figure [84]3R,S,U). These results suggest that MRP14 inhibition can
   alleviate BBB breakdown following ischemic stroke. To further validate
   MRP14's role, MRP14 knockout (KO) mice were subjected to tMCAO (Figure
   [85]S4A‐D, Supporting Information). Compared with wild type (WT) mice,
   MRP14 KO mice exhibited significantly reduced BBB leakage, as evidenced
   by Evans blue and fibrinogen extravasation assays 24 h after tMCAO
   (Figure [86]3B,D,E,G). These mice also showed restored tight junction
   protein expression (Figure [87]3H,L‐O,Q), decreased infarct area and
   improved neurological outcomes (Figure [88]3R,T,V). These results
   underscore the pivotal role of MRP14 in mediating BBB disruption
   post‐stroke.

Figure 3.

   Figure 3
   [89]Open in a new tab

   Paquinimod treatment or MRP14 knockout attenuates BBB disruption after
   ischemic stroke. A) The illustration of the experimental schedule. The
   mice were first subjected to tMCAO procedure for 1‐h ischemia and
   paquinimod (10 mg kg^−1) was injected intravenously 10 min before
   reperfusion. The subsequent experiments were performed 24 h after
   reperfusion. B‐D) The Evans blue extravasation assay of the indicated
   groups (n = 6/group). E‐G) Immunofluorescence staining exhibiting the
   fibrinogen (green) leakage in the ischemic brain of the indicated
   groups (n = 6/group). H‐N) Western blot analysis of the expression of
   tight junction proteins, including ZO‐1, Occludin and VE‐Cadherin in
   the ischemic brain of the indicated groups (n = 6/group). O‐Q)
   Immunofluorescence staining showing the expression of VE‐Cadherin
   (green) on the blood vessels (red) in the ischemic brain of the
   indicated groups (n = 6/group). R‐T) The TTC staining showing the
   infarction area in the indicated groups (n = 6/group). U,V) The mNSS
   scores showing the neurological deficits in the indicated groups (n =
   6). Data are presented as means ± SEM; unpaired t‐test, one‐way ANOVA.
   *P < 0.05, **P < 0.01. Scale bar: 50 µm. BBB, blood brain barrier;
   tMCAO, transient middle cerebral artery occlusion; TTC, tri‐phenyl
   tetrazolium chloride; mNSS, Modified Neurological Severity Score; veh,
   vehicle; paq, paquinimod.

   To specifically elucidate the role of neutrophil‐derived MRP14 in BBB
   damage, neutrophils were isolated from the bone marrow of WT or MRP14
   KO mice, labeled with Molday ION Rhodamine B (MIRB), and transferred
   into MRP14 KO mice via tail vein injection prior to tMCAO (Figure
   [90]S5A, Supporting Information). MIRB‐labeled neutrophils were
   detected in the ischemic brain 24 h post‐tMCAO (Figure [91]S5B,
   Supporting Information), confirming successful neutrophil transfer.
   Mice receiving WT neutrophils exhibited significantly greater BBB
   leakage, as evidenced by both Evans blue and fibrinogen assays (Figure
   [92]S5C‐F, Supporting Information), than those receiving MRP14 KO
   neutrophils. Furthermore, WT neutrophil transfer led to decreased tight
   junction protein expression (Figure [93]S5G‐L, Supporting Information),
   larger infarct areas, and worsened neurological deficits (Figure
   [94]S5M‐O, Supporting Information). These results strongly indicate
   that neutrophil‐derived MRP14 is critical to BBB disruption following
   ischemic stroke.

2.5. MRP14 Triggers Microglial Functional Change after Ischemic Stroke

   Considering that MRP14, a member of damage‐related molecular patterns,
   triggers pro‐inflammatory immune responses,^[ [95]^13 ^] and that
   microglia and astrocytes serve as the primary immune cells in the
   brain, we isolated neutrophils from WT or MRP14 KO tMCAO mice, and then
   co‐cultured them with oxygen‐glucose deprivation/reoxygenation
   (OGD/R)‐treated primary microglia or astrocytes for 24 h. The
   supernatants were then added into a Transwell chamber, with endothelial
   bEnd.3 cells seeded in the upper compartment (Figure [96]S6A,
   Supporting Information). As shown in Figure [97]S6B,C (Supporting
   Information), supernatants from WT neutrophils reduced transendothelial
   electrical resistance (TEER) value and increased FITC‐dextran
   permeability in endothelial cells. These effects were alleviated by
   supernatants from MRP14 KO neutrophils, indicating that
   neutrophils‐derived MRP14 alone could lead to endothelial damage.
   Notably, these effects were amplified when endothelial cells were
   exposed to supernatants from WT neutrophil‐microglia co‐cultures
   compared with those from WT neutrophil‐astrocyte co‐cultures (Figure
   [98]S6B,C, Supporting Information). Furthermore, supernatants from
   MRP14 KO neutrophil‐microglia co‐cultures attenuated BBB leakage
   compared with those from WT counterparts (Figure [99]S6B,C, Supporting
   Information). Previous studies have indicated that MRP14 binds with
   high affinity to Toll‐like receptor 4 (TLR4) and the receptor for
   advanced glycation end‐products (RAGE),^[ [100]^14 , [101]^15 ^] which
   facilitate MRP14 signaling. We further found that only TLR4 inhibitor
   resatorvid, not RAGE inhibitor FPS‐ZM1, significantly reversed the
   effect of WT neutrophil‐microglia co‐cultures supernatants in the in
   vitro BBB model (Figure [102]S6D,E, Supporting Information). Together,
   these results suggest that MRP14 could induce more pronounced
   endothelial cell damage by stimulating microglia, which is mediated by
   TLR4 signaling.

   To further elucidate the underlying mechanisms, the scRNA‐seq analysis
   was conducted on microglia isolated from ischemic brain tissue of WT
   and MRP14 KO tMCAO mice. The cells were integrated using the “harmony”
   package in R. Microglia were identified by the markers “Trem2”,
   “Tmem119”, and “Siglech”. The uniform manifold approximation and
   projection (UMAP) visualization revealed six distinct microglial
   subpopulations (Figure [103] 4A), and a heatmap displayed the relative
   expression of the top 10 signature genes within each cluster
   (Figure [104]4B). Notably, clusters 0 and 1 were the dominant
   subpopulations. Compared with the WT group, the percentage of cluster 0
   significantly reduced and cluster 1 markedly increased in the MRP14 KO
   group (Figure [105]4C,D). Gene ontology (GO) and Kyoto encyclopedia of
   genes and genomes (KEGG) pathway analyses linked cluster 1 to
   microglial motility and energy metabolism (Figure [106]4E,F).
   Enrichment analysis of DEGs upregulated in cluster 1 between WT and
   MRP14 KO groups identified “Phagocytosis” and “Phagosome” as top
   pathways (Figure [107]4G), suggesting an association of cluster 1
   microglia with phagocytic function. Given the elevated proportion of
   this cluster in MRP14 KO mice, we assume that neutrophil‐derived MRP14
   may impair microglial phagocytosis following AIS. In addition to
   affecting phagocytosis, infiltrating neutrophils can also influence
   microglial viability. Previous results have demonstrated an inverse
   correlation between microglial counts and the number of infiltrating
   neutrophils in the ischemic brain (Figure [108]1D‐F). To further
   explore this relationship, area under the curve scores of cell death
   pathway activities were calculated for identified microglial clusters
   (clusters 0–5). Cluster 0, which proportion was most significantly
   reduced in the MRP14 KO group, exhibited the highest activity in the
   pyroptosis pathway (Figure [109]4H‐K), suggesting that MRP14 may
   contribute to microglial pyroptosis.

Figure 4.

   Figure 4
   [110]Open in a new tab

   Characteristics of microglia in the MRP14 KO mice 24 h after tMCAO. A)
   The UMAP plot displaying the 6 microglial clusters. B) Heatmap showing
   the top 10 signature genes in each cluster. C) The distribution of
   microglia in the WT and MRP14 KO groups in the UMAP plot. D) Bar graph
   showing the percentage of each microglial cluster in the WT and MRP14
   KO groups. E) The top 10 enriched GO terms of biologic process of
   cluster 1. F) The top 10 enriched KEGG terms of cluster 1. G) The top
   10 enriched KEGG terms of differentially expressed genes between the WT
   and MRP14 KO groups in the cluster 1. H‐K) The violin plots showing the
   AUC scores of different death pathway activities of each cluster. Data
   are presented as means ± SEM; one‐way ANOVA. **P < 0.01. tMCAO,
   transient middle cerebral artery occlusion; WT, wild type; KO,
   knockout; UMAP, Uniform Manifold Approximation and Projection; GO, Gene
   Ontology; KEGG, Kyoto Encyclopedia of Genes and Genomes; AUC, Area
   Under Curve.

   To further investigate the transformation among microglial clusters, a
   pseudotime trajectory analysis was performed (Figure [111]S7A,
   Supporting Information). We found that cluster 1, phagocytosis‐related
   microglia, was located at the start and middle stages of the
   developmental trajectory, while cluster 0, pyroptosis‐related
   microglia, appeared at the end stage (Figure [112]S7B, Supporting
   Information). This suggested a potential transition from cluster 1 to
   cluster 0 (Figure [113]S7C, Supporting Information). In line with this
   observation, the expression of phagocytosis‐related genes was decreased
   and pyroptosis‐related genes was increased along the pseudotime
   trajectory (Figure [114]S7D,E, Supporting Information).

2.6. MRP14 Inhibits Microglial Phagocytosis of Neutrophils and Promotes
Microglial Pyroptosis after Ischemic Stroke

   Previous results have suggested that activated microglia can engulf
   neutrophils recruited to the ischemic brain (Figure [115]S1A,B,
   Supporting Information). Thus, this study further investigated whether
   MRP14 regulates microglial phagocytosis of neutrophils. Microglial
   phagocytosis of neutrophils was quantified by calculating the
   percentage of neutrophils in touching and surrounded states in vivo. As
   shown in Figure [116] 5A,B, MRP14 KO mice exhibited a significantly
   higher percentage of neutrophils touching and surrounded by microglia
   compared with WT mice, 24 h post‐tMCAO. Additionally, the overall
   neutrophil count was reduced in the MRP14 KO group (Figure [117]5C). In
   addition, we also used flow cytometry to quantify the ability of
   microglia to engulf neutrophils. First, we labeled microglia with CD45
   and CD11b in the ischemic brain cells isolated from WT and MRP14 KO
   mice 24 h post‐tMCAO. Then, the cells were permeabilized and
   intracellularly stained with neutrophil marker Ly6G. Finally, the
   percentage of Ly6G^+ microglia, which indicates the microglia that have
   phagocytosed neutrophils, was analyzed using flow cytometry (Figure
   [118]S8, Supporting Information). Compared with the WT group, MRP14 KO
   mice exhibited a significantly higher percentage of Ly6G^+ microglia,
   suggesting an increase in phagocytosis‐related microglia
   (Figure [119]5D,E). To further assess changes in phagocytic capacity,
   the expression of CD68, a marker of active phagocytosis, was measured
   in microglia. Compared with WT mice, MRP14 KO mice showed enhanced
   microglial CD68 fluorescence intensity 24 h after tMCAO
   (Figure [120]5F,G). Together, these results suggested that MRP14
   prevents microglial phagocytosis of neutrophils after ischemic stroke.

Figure 5.

   Figure 5
   [121]Open in a new tab

   MRP14 inhibits microglia to phagocytize neutrophils and promotes
   microglial pyroptosis after ischemic stroke. A,B) Immunofluorescence
   staining showing the percentage of neutrophils (green) touching or
   surrounded by microglia (red) in the ischemic brain of WT and MRP14 KO
   groups (n = 6/group). The white arrow indicates neutrophils touching or
   surrounded by microglia. C) The neutrophil numbers per field in the
   ischemic brain of WT and MRP14 KO groups (n = 6/group). D,E) The flow
   cytometry analysis showing the percentage of Ly6G^+ microglia in the
   ischemic brain of the tMCAO mice from the WT and MRP14 KO groups (n =
   6/group). IgG was used as a negative control. F,G) Immunofluorescence
   staining showing the expression of CD68 (green) in the microglia (red)
   in the ischemic brain of WT and MRP14 KO groups (n = 6/group). H‐L)
   Western blot analysis of the expression of GSDMD, NLRP3, pro‐caspase 1,
   caspase 1, pro‐IL‐1β and IL‐1β in the ischemic brain of WT or MRP14 KO
   mice 24 h after tMCAO (n = 6/group). M,N) Immunofluorescence staining
   showing the expression of GSDMD (green) in the microglia (red) in the
   ischemic brain of WT or MRP14 KO mice 24 h after tMCAO (n = 6/group).
   Data are presented as means ± SEM; Mann‐Whitney test, unpaired t‐test,
   one‐way ANOVA. **P < 0.01. Scale bar: 25 µm (A and F); 50 µm (M).
   tMCAO, transient middle cerebral artery occlusion; WT, wild type; KO,
   knockout; veh, vehicle; paq, paquinimod; GSDMD, gasdermin D; NLRP3,
   NOD‐like receptor family, pyrin domain containing 3; IL‐1β,
   interleukin‐1β; Casp1, caspase 1.

   The role of MRP14 in promoting microglial pyroptosis was further
   assessed in vivo. Elevated expression of pyroptosis‐associated
   proteins, including GSDMD, NLRP3, caspase‐1 and IL‐1β, was observed in
   the ischemic brain tissue of tMCAO mice compared with the sham group
   (Figure [122]5H‐L). However, MRP14 deficiency significantly reversed
   these increases (Figure [123]5H‐L). Immunofluorescence staining further
   revealed a marked increase in GSDMD fluorescence intensity in microglia
   24 h post‐tMCAO (Figure [124]5M,N); this increase was similarly
   attenuated by MRP14 deficiency (Figure [125]5M,N). These results
   suggest that MRP14 may be involved in activating NLRP3
   inflammasome‐mediated microglial pyroptosis following ischemic stroke
   in vivo.

2.7. MRP14 Impairs Microglial Phagocytosis of Neutrophils by Triggering
Mitochondrial Dysfunction via TLR4 Receptor In Vitro

   The effect of MRP14 on microglial phagocytic activity was further
   explored in vitro. Neutrophils were isolated from the tMCAO mice, and
   microglia were exposed to OGD/R followed by stimulation with
   recombinant MRP14. After 24 h, neutrophils and microglia were
   co‐cultured to assess the effect of MRP14 on microglial phagocytosis
   (Figure [126]S9A, Supporting Information). As demonstrated in Figure
   [127]S9B,C (Supporting Information), OGD/R‐treated microglia exhibited
   increased CD68 expression and enhanced neutrophil phagocytosis, both of
   which were significantly attenuated by recombinant MRP14 treatment.
   Additionally, Western blot analysis revealed elevated MPO expression in
   OGD/R‐treated microglia, an effect that was reversed by recombinant
   MRP14 treatment (Figure [128]S9D,E, Supporting Information). We also
   investigated the role of TLR4 in mediating the effects of MRP14 on
   microglial phagocytosis. As shown in Figure [129]S9B‐E (Supporting
   Information), inhibition of TLR4 with resatorvid markedly reduced the
   inhibitory effect of recombinant MRP14 on microglial phagocytosis of
   neutrophils, underscoring the key role of TLR4 in this process.

   Efficient phagocytosis by microglia requires dynamic cytoskeletal
   reorganization, a process heavily dependent on adenosine triphosphatase
   (ATP).^[ [130]^16 ^] Prior studies have demonstrated that MRP14 induces
   mitochondrial dysfunction through TLR4 after myocardial infarction.^[
   [131]^17 ^] Consistently, KEGG pathway analysis of cluster 1 signature
   genes identified enrichment in pathways related to the tricarboxylic
   acid cycle, respiratory electron transport, and ATP synthesis
   (Figure [132]4F). These findings support the hypothesis that MRP14
   disrupts mitochondrial function in microglia, thereby compromising the
   energy supply necessary for phagocytosis. JC‐1 staining confirmed that
   OGD/R treatment significantly decreased the red/green fluorescence
   ratio in microglia, indicating mitochondrial membrane potential
   disruption, and recombined MRP14 treatment could further exacerbate the
   damage (Figure [133]S9F,G, Supporting Information). However, this
   damage was mitigated by TLR4 inhibition with resatorvid. Similarly, ATP
   concentration in OGD/R‐treated microglia were decreased, with further
   reductions observed following recombined MRP14 treatment. These ATP
   deficits were also mitigated by resatorvid‐mediated TLR4 inhibition
   (Figure [134]S9H, Supporting Information). Collectively, these results
   suggest that MRP14 may impair microglial phagocytosis of neutrophils by
   inducing mitochondrial dysfunction via the TLR4 receptor.

2.8. MRP14 Promotes Microglial NLRP3 Inflammasome Activation and Pyroptosis
Through TLR4/NF‐κB Signaling In Vitro

   The effect of MRP14 on microglial pyroptosis was also investigated in
   vitro. In primary microglia cultures, treatment with recombinant MRP14
   alone did not induce cleavage of caspase‐1 or IL‐1β, as shown by
   Western blot analysis (Figure [135]S10A‐C, Supporting Information).
   However, priming with recombinant MRP14 followed by ATP treatment, but
   not lipopolysaccharide (LPS), resulted in significant processing of
   caspase‐1 and IL‐1β (Figure [136]S10A‐C, Supporting Information). These
   findings indicate that MRP14 functions as a priming signal, similar to
   LPS, in the induction of microglial pyroptosis. In subsequent
   mechanistic studies, primary microglia were treated with both
   recombinant MRP14 and ATP. Additionally, primary microglia were further
   treated with recombinant MRP14 and/or the TLR4 inhibitor resatorvid.
   The results demonstrated that recombinant MRP14 significantly
   upregulated the protein expression of GSDMD, NLRP3, caspase‐1 and IL‐1β
   in microglia (Figure [137]S10D‐H, Supporting Information), and
   increased the release of IL‐1β into the supernatant (Figure [138]S10I,
   Supporting Information). Resatorvid effectively blocked these effects,
   indicating that the pro‐inflammatory actions of MRP14 are mediated
   through the TLR4 pathway (Figure [139]S10D‐I, Supporting Information).

   To further elucidate the downstream mechanisms of MRP14‐mediated
   pyroptosis, key signaling pathways were examined. Prior studies have
   shown that TLR4 predominantly activates the NF‐κB signaling pathway to
   facilitate intracellular functions.^[ [140]^18 ^] Our findings revealed
   that MRP14 significantly enhanced the phosphorylation of IκB and p65,
   and this effect was abolished by TLR4 inhibition (Figure [141]S11A‐D,
   Supporting Information). These results suggest that MRP14 regulates
   NF‐κB activation via TLR4. To confirm that MRP14 exerts its effects on
   microglia pyroptosis via the NF‐κB pathway, the NF‐κB inhibitor JSH‐23
   was employed to suppress its activity. Notably, NF‐κB inhibition
   significantly diminished the MRP14‐induced upregulation of GSDMD,
   NLRP3, caspase‐1 and IL‐1β, as well as the release of IL‐1β into the
   supernatant (Figure [142]S11E‐J, Supporting Information). Collectively,
   these results suggest that MRP14 promotes NLRP3 inflammasome‐associated
   pyroptosis in primary microglia through activation of the TLR4/NF‐κB
   signaling pathway.

2.9. Microglial TLR4 Knockout Reduces MRP14‐Induced Phagocytosis Impairment,
NLRP3 Inflammasome‐Associated Pyroptosis, and BBB Breakdown In Vivo

   To validate the involvement of microglial TLR4 in vivo,
   microglia‐specific TLR4‐deficient mice were generated by breeding
   TLR4^flox/flox mice with CX3CR1‐CreERT mice (Figure [143] 6A,B; Figure
   [144]S12A,B, Supporting Information). TLR4 expression was effectively
   reduced in TLR4^flox/flox; CX3CR1‐CreERT mice (Figure [145]6C,D).
   TLR4^flox/flox; CX3CR1‐CreERT mice, along with their TLR4^flox/flox
   littermates, were subjected to tMCAO followed by the administration of
   recombinant MRP14 protein.

Figure 6.

   Figure 6
   [146]Open in a new tab

   Conditional Knockout of microglial TLR4 alleviates MRP14‐induced
   impaired phagocytosis and NLRP3 inflammasome‐associated pyroptosis
   after ischemic stroke. A,B) The schematic diagram showing the breeding
   strategy for generating mice which were specifically lacking TLR4
   expression in microglia (TLR4^flox/flox; CX3CR1‐CreERT). C,D)
   Verification of conditionally knockout of TLR4 expression in
   TLR4^flox/flox; CX3CR1‐CreERT mice (n = 6/group). E,F)
   Immunofluorescence staining showing the percentage of neutrophils
   (green) touching or surrounded by microglia (red) in the ischemic brain
   of TLR4^flox/flox or TLR4^flox/flox; CX3CR1‐CreERT mice treated with
   vehicle or recombined MRP14 protein (0.6 mg kg^−1) 24 h after tMCAO (n
   = 6/group). G) The neutrophil numbers per high‐power field in the
   ischemic brain of TLR4^flox/flox or TLR4^flox/flox; CX3CR1‐CreERT mice
   treated with vehicle or recombined MRP14 protein (0.6 mg kg^−1) 24 h
   after tMCAO (n = 6/group). H‐L) Western blot analysis of the expression
   of GSDMD, NLRP3, pro‐caspase 1, caspase 1, pro‐IL‐1β and IL‐1β in the
   ischemic brain of TLR4^flox/flox or TLR4^flox/flox; CX3CR1‐CreERT mice
   treated with vehicle or recombined MRP14 protein (0.6 mg kg^−1) 24 h
   after tMCAO (n = 6/group). M,N) Immunofluorescence staining showing the
   expression of GSDMD (green) in the microglia (red) in the ischemic
   brain of TLR4^flox/flox or TLR4^flox/flox; CX3CR1‐CreERT mice treated
   with vehicle or recombined MRP14 protein (0.6 mg kg^−1) 24 h after
   tMCAO (n = 6/group). Data are presented as means ± SEM; unpaired
   t‐test, one‐way ANOVA. *P < 0.05, **P < 0.01. Scale bar, 50 µm. TLR4,
   Toll‐like receptor 4; veh, vehicle; TLR4^f/f, TLR4^flox/flox; GSDMD,
   gasdermin D; NLRP3, NOD‐like receptor family, pyrin domain containing
   3; IL‐1β, interleukin‐1β; Casp1, caspase 1.

   In TLR4^flox/flox mice, recombinant MRP14 reduced the percentage of
   neutrophils touching and surrounded by microglia, meanwhile the overall
   neutrophil count was increased (Figure [147]6E‐G). It also upregulated
   the expression of GSDMD, NLRP3, caspase‐1 and IL‐1β, and increased
   GSDMD fluorescence intensity in microglia (Figure [148]6H‐N). In
   contrast, these changes were absent in TLR4^flox/flox; CX3CR1‐CreERT
   mice (Figure [149]6E‐N). Furthermore, recombinant MRP14 protein
   significantly increased Evans blue and fibrinogen leakage post‐tMCAO in
   TLR4^flox/flox mice, while these effects did not occur in
   TLR4^flox/flox; CX3CR1‐CreERT mice (Figure [150] 7A‐D). These results
   firmly establish that microglial TLR4 is essential for MRP14‐induced
   impairment of phagocytosis, activation of pyroptosis, and BBB breakdown
   following cerebral ischemia.

Figure 7.

   Figure 7
   [151]Open in a new tab

   Conditional Knockout of microglial TLR4 alleviates MRP14‐mediated BBB
   breakdown in the tMCAO mice. A,B) The Evans blue extravasation assay of
   the TLR4^flox/flox or TLR4^flox/flox; CX3CR1‐CreERT mice treated with
   vehicle or recombined MRP14 protein (0.6 mg kg^−1) 24 h after tMCAO (n
   = 6/group). C,D) Immunofluorescence staining showing the fibrinogen
   (green) leakage in the ischemic brain of TLR4^flox/flox or
   TLR4^flox/flox; CX3CR1‐CreERT mice treated with vehicle or recombined
   MRP14 protein (0.6 mg kg^−1) 24 h after tMCAO (n = 6/group). E) The
   schematic illustration showing MRP14‐mediated interactions between
   neutrophils and microglia after ischemic stroke. Data are presented as
   means ± SEM; one‐way ANOVA. **P < 0.01. Scale bar, 50 µm. TLR4,
   Toll‐like receptor 4; BBB, blood brain barrier; tMCAO, transient middle
   cerebral artery occlusion; mNSS, Modified Neurological Severity Score;
   TLR4^f/f, TLR4^flox/flox; veh, vehicle.

2.10. Neutrophil MRP14 Expression can be Modulated by IL‐1β via
IL‐1R1/MAPK/C/EBP β Signaling Pathway

   Microglial pyroptosis results in the release of IL‐1β, which was
   reported to regulate the expression of pro‐inflammatory cytokines.^[
   [152]^19 ^] We found mRNA levels of IL‐1β receptor (IL‐1R1) were
   elevated in peripheral neutrophils from tMCAO mice (Figure [153]S13A,
   Supporting Information). Recombinant IL‐1β significantly increased
   MRP14 protein levels, and this effect was abolished by the IL‐1R1
   inhibitor anakinra (Figure [154]S13B,C, Supporting Information),
   indicating that IL‐1β might also interact with neutrophils to promote
   MRP14 expression. IL‐1β signaling has been shown to activate the NF‐κB
   and MAPK pathways.^[ [155]^19 , [156]^20 ^] To explore the downstream
   signaling pathway, we further performed KEGG pathway enrichment
   analysis of the upregulated DEGs identified in our previous RNA
   sequencing of peripheral neutrophils from sham and tMCAO mice (Figure
   [157]S2A, Supporting Information). The analysis showed the activation
   of the MAPK pathway (Figure [158]S13D, Supporting Information). Western
   blot analysis confirmed a significant increase in the p‐p38/p38 ratio
   in neutrophils treated with recombinant IL‐1β (Figure [159]S13E,F,
   Supporting Information). Pre‐treatment with the p38 inhibitor PD 169316
   effectively suppressed the IL‐1β‐induced elevation of MRP14 in
   neutrophils (Figure [160]S13G,H, Supporting Information), strongly
   supporting the role of the p38 MAPK pathway in IL‐1β‐induced MRP14
   expression in neutrophils.

   To identify transcription factors potentially regulating MRP14
   expression, bioinformatics analyses were conducted using ALGEEN, GTRD,
   and AnimalTFDB, with KnockTF employed to predict transcription factors
   enriched among the DEGs from RNA sequencing. There were five
   overlapping transcription factors among these four databases: C/EBP β,
   C/EBP α, YY1, STAT6, and PAX5 (Figure [161]S13I, Supporting
   Information). Previous studies have demonstrated that C/EBP β directly
   promotes MRP14 transcription^[ [162]^21 ^] and that its expression is
   modulated by the MAPK pathway.^[ [163]^22 ^] Thus, we performed a dual
   luciferase assay to assess whether C/EBP β could bind to the MRP14
   promoter and regulate its expression. The result demonstrated that only
   the co‐transfection group with MRP14 WT and C/EBP β plasmids can lead
   to an increase in luciferase activity (Figure [164]S13J,K, Supporting
   Information). Further experiments showed that recombinant IL‐1β induced
   C/EBP β expression in neutrophils, an effect significantly inhibited by
   PD 169316 (Figure [165]S13L,M, Supporting Information). These results
   conclusively suggest that the IL‐1R1/MAPK/C/EBP β signaling pathway
   contributes to the upregulation of MRP14 expression in neutrophils upon
   IL‐1β treatment.

   Previous study has reported that IL‐1β could induce neutrophil
   mobilization from the bone marrow in the model of myocardial
   infarction.^[ [166]^23 ^] Thus, we assumed that IL‐1β may similarly
   promote neutrophil mobilization from the bone marrow after ischemic
   stroke. Further analysis revealed significantly elevated mRNA levels of
   the IL‐1R1 in neutrophils isolated from the bone marrow of tMCAO mice,
   while CXCR4 expression was reduced, and CXCR2 levels remained unchanged
   (Figure [167]S14A‐C, Supporting Information). These results suggest
   that IL‐1β may facilitate neutrophil mobilization following ischemic
   stroke.

3. Discussion

   Neutrophil accumulation in the brain parenchyma is a key driver of
   neuroinflammation following ischemic stroke. These peripheral
   neutrophils infiltrate the ischemic brain as early as 15 min after
   stroke onset, with peak infiltration occurring between 24 and 72 h.^[
   [168]^4 ^] The balance between neutrophil accumulation and clearance in
   the ischemic brain critically influences the severity of
   neuroinflammatory damage. Impaired microglial phagocytosis facilitates
   neutrophil retention, thereby exacerbating brain injury.^[ [169]^11 ^]
   Our study identified a pivotal role for neutrophil‐derived MRP14 in
   exacerbating both neutrophil accumulation and neuroinflammation. MRP14
   not only serves as a “don't eat me” signal, inhibiting microglial
   phagocytosis of neutrophils, but also acts as a “kill you” signal,
   inducing microglial pyroptosis and neuroinflammation. Additionally,
   IL‐1β released from pyroptotic microglia further promotes MRP14
   expression in neutrophils and enhances their mobilization from the bone
   marrow. This cascade of neutrophil accumulation and neuroinflammation
   ultimately contributes to BBB disruption during ischemic stroke
   (Figure [170]7E). These results reveal previously unknown interactions
   between neutrophils and microglia within the brain parenchyma and
   support the therapeutic targeting of MRP14 in ischemic stroke.

   Despite promising results in pre‐clinical and phase II studies, no
   direct anti‐inflammatory drugs have demonstrated unequivocal clinical
   efficacy in phase III trials in stroke treatment.^[ [171]^24 ^] The
   failure of these drugs may be attributed to direct anti‐inflammation
   treatment act as a double‐edged sword in the context of acute ischemic
   stroke, which not only alleviates inflammatory damage and improves
   neurological function, but also compromises the body's immune defense,
   potentially leading to life‐threatening infections.^[ [172]^24 ^]
   Moreover, specific subsets of inflammatory cells, such as N2 and CD177⁺
   neutrophils,^[ [173]^4 , [174]^25 ^] may exert protective effects
   following ischemic stroke. Therefore, direct anti‐inflammatory
   therapies may inhibit the protective effects of these cells.
   Ultimately, these adverse effects may counteract their benefits and
   influence patient prognosis. In the current study, we found that
   selective MRP14 targeting effectively inhibits the inflammatory cascade
   within ischemic brain tissue while minimizing the adverse effects
   commonly associated with direct anti‐inflammatory treatments.

   Our study identified neutrophils as the primary source of MRP14 at 24 h
   after tMCAO. However, a recently published study demonstrated a role
   for microglia/macrophage‐derived MRP14 between 3 and 21 days after
   tMCAO.^[ [175]^26 ^] Notwithstanding, their findings were not
   inconsistent with ours. First, it is reported that MRP14 is
   predominantly expressed in neutrophils and monocytes in peripheral
   blood, with only limited expression in brain parenchymal microglia.^[
   [176]^27 ^] In peripheral blood, MRP14 constitutes ≈45% of the total
   cytosolic protein in neutrophils, while monocytes express ≈40‐fold
   lower levels.^[ [177]^28 ^] Second, neutrophils reach their peak
   infiltration 1–2 days post‐ischemia, corresponding with the increase in
   brain MRP14 expression observed in our study. In contrast, monocyte
   infiltration begins one day post‐ischemia, peaking at 4th day,^[
   [178]^29 ^] and occurs at levels significantly lower than those of
   neutrophils.^[ [179]^30 ^] Immunofluorescence staining in this study
   further confirmed that MRP14 predominantly co‐localizes with
   neutrophils, with minimal expression in microglia within ischemic brain
   tissue. Finally, bone marrow neutrophil transfer into MRP14 KO mice
   exacerbated BBB damage at 24 h after tMCAO, underscoring the role of
   neutrophil‐derived MRP14 in ischemia‐reperfusion injury. These results
   strongly demonstrate that neutrophils are the primary source of
   elevated MRP14 during the acute stage of ischemic stroke.

   Recent studies have shown that reactive microglia can phagocytose and
   clear infiltrating peripheral neutrophils recruited to the ischemic
   brain, thus playing a pivotal role in protecting against
   neutrophil‐induced tissue damage after stroke.^[ [180]^11 ^] Our
   findings suggest that MRP14 acts as a “don't eat me” signal, protecting
   neutrophils from microglial phagocytosis post‐stroke. Compared with WT
   mice, MRP14 KO mice exhibited significantly enhanced microglial
   phagocytosis after tMCAO. In vitro experiments further confirmed the
   inhibitory effects of MRP14 on microglial phagocytosis of neutrophils.
   Additionally, MRP14 induces mitochondrial dysfunction in microglia,
   which likely explains impaired phagocytic capacity following MRP14
   exposure. This observation is consistent with previous studies showing
   that elevated MRP14 levels can cause mitochondrial dysfunction in
   models of acute myocardial infarction and sepsis.^[ [181]^17 , [182]^31
   ^]

   In addition, MRP14 also functions as a “kill you” signal, promoting
   microglial pyroptosis and the subsequent release of inflammatory
   mediators, such as IL‐1β, which exacerbates BBB damage following AIS.
   Both pharmacological inhibition of MRP14 using paquinimod and genetic
   KO of MRP14 significantly attenuated NLRP3 inflammasome‐associated
   pyroptosis and mitigated BBB disruption after tMCAO. Furthermore, IL‐1β
   has been reported to enhance neutrophil mobilization from the bone
   marrow in conditions like myocardial infarction.^[ [183]^23 ^]
   Administration of the IL‐1R1 antagonist anakinra similarly reduced
   leukocyte counts, particularly neutrophils, in patients with stroke.^[
   [184]^32 ^] Our findings support these observations by showing
   significantly increased IL‐1R1 expression in bone marrow neutrophils
   post‐tMCAO, reinforcing the role of IL‐1β in promoting neutrophil
   mobilization. Moreover, IL‐1β induces MRP14 expression in neutrophils
   via the IL‐1R1/MAPK/C/EBP β signaling pathway, establishing a harmful
   feedback loop that amplifies MRP14‐induced damage.

   Previous studies have identified TLR4 and RAGE as receptors for
   extracellular MRP14.^[ [185]^27 ^] In our study, receptor‐blocking
   experiments demonstrated that MRP14‐induced microglial responses were
   dependent on TLR4 rather than RAGE. Using microglia‐specific
   TLR4‐deficient mice in the tMCAO model, it was confirmed that MRP14
   impairs microglial phagocytosis and promotes pyroptosis through TLR4.
   As a member of the pattern recognition receptor family, TLR4 plays a
   critical role in immune responses and phagocytic function.^[ [186]^33
   ^] Numerous studies have shown that TLR4 deficiency exerts
   neuroprotective effects by reducing post‐stroke inflammation.^[
   [187]^34 , [188]^35 ^] Additionally, our findings indicated that MRP14
   triggers mitochondrial dysfunction through TLR4, thereby reducing the
   energy supply needed for microglial phagocytosis of neutrophils
   following tMCAO. This mechanism aligns with previous studies linking
   MRP14 to mitochondrial impairment in myocardial ischemia‐reperfusion
   injury and sepsis.^[ [189]^17 , [190]^31 ^]

   This study has several limitations that warrant consideration. First,
   the precise mechanism by which ischemic stroke induces the release of
   MRP14 from neutrophils remains unclear. Second, whether
   paquinimod‐mediated inhibition of MRP14 can prevent AIS‐induced BBB
   damage in clinical settings remains unknown. Third, the potential
   adverse effects associated with MRP14 inhibition were not fully
   explored. Because intracellular MRP14 is involved in cytoskeletal
   regulation and arachidonic acid metabolism.^[ [191]^27 ^] Therefore,
   selectively targeting extracellular MRP14, possibly through antibodies
   or nanoparticles, could represent a novel adjunctive therapy for AIS,
   minimizing interference with essential intracellular processes.

4. Experimental Section

Patient Sample

   Patients for this study were drawn from the TRAIS cohort, which
   retrospectively compiled data from all consecutive patients with AIS
   admitted to 14 stroke centers in China (ChiCTR2000033456). All
   participants provided informed consent, and the study adhered to the
   ethical standards outlined in the Declaration of Helsinki. The local
   institutional review boards approved all study protocols (Wuhan, China,
   2020IEC‐0273).

   Plasma samples were collected from patients with AIS within 24 h of
   stroke onset. Edema was assessed using the SITS‐MOST grading system,
   where signs of focal brain edema typically involve the narrowing of
   cerebrospinal fluid spaces, such as the effacement of cortical sulci or
   ventricular compression.^[ [192]^36 , [193]^37 ^] Based on MRI or CT
   scans conducted within 24–72 h post‐stroke, patients were categorized
   as having no cerebral edema (cerebral edema‐0), cerebral edema
   affecting less than one‐third of the hemisphere (cerebral edema‐1),
   cerebral edema involving more than one‐third of the hemisphere
   (cerebral edema‐2), or cerebral edema resulting in midline shift
   (cerebral edema‐3). For the purposes of this study, patients classified
   as cerebral edema‐0 and ‐1 were grouped as having mild cerebral edema,
   while those with cerebral edema‐2 and ‐3 were grouped as having severe
   cerebral edema.

Animals

   C57BL/6J mice (8 weeks old) were obtained from Hubei Biont Biological
   Technology Co., Ltd. (Wuhan, China), and C57BL/6J‐MRP14^em1C /Cya
   (MRP14 KO, S‐KO‐16221) mice were sourced from Cyagen Biosciences Inc.
   (Suzhou, China). Additionally, B6.129P2(C)‐Cx3cr1^2.1 (cre/ERT2) Jung
   /J (CX3CR1‐CreERT, #020940) mice were purchased from Jackson Laboratory
   (Bar Harbor, USA), while C57BL/6‐Tlr4^em1 (flox) Smoc (TLR4^flox/flox,
   NM‐CKO‐200230) mice were acquired from Shanghai Model Organisms Center,
   Inc. (Shanghai, China). To generate microglia‐specific TLR4‐deficient
   mice, TLR4^flox/flox mice were crossed with CX3CR1‐CreERT mice.
   Six‐week‐old male TLR4^flox/flox; CX3CR1‐CreERT mice and littermate
   TLR4^flox/flox mice underwent tamoxifen (Beyotime Biotechnology, China)
   treatment to induce Cre recombinase expression (intraperitoneal
   injection, 75 mg kg^−1 for 5 consecutive days). All animals were housed
   under a 12‐h light‐dark cycle at controlled temperatures (20–25 °C) and
   humidity levels (40–60%) with ad libitum access to food and water. The
   medical ethics committee of Tongji Medical College, Huazhong University
   of Science and Technology, approved all animal protocols (Wuhan, China,
   2023IACUC‐4223), which were conducted in compliance with the Guide for
   the Care and Use of Laboratory Animals.

Establishment of the tMCAO Model

   The tMCAO model was performed on adult mice (22–25 g) as previously
   described.^[ [194]^38 ^] In brief, mice were anesthetized using
   isoflurane inhalation and a midline incision was made to fully expose
   the right common carotid arteries (CCA), internal carotid arteries
   (ICA) and external carotid arteries (ECA). Then, a 6‐0 silicon‐coated
   nylon suture (Yushun Biotech, China) was inserted into the ECA,
   advanced through the CCA, and finally reached the origin of the middle
   cerebral artery in the ICA to block the blood flow. One hour after
   ischemia, the suture was removed to achieve reperfusion to the ischemic
   brain. The sham group was exposed to the same procedures aside from
   middle cerebral artery occlusion. The temperature was controlled
   throughout the surgery and the cerebral blood flow was measured to
   ascertain ischemia and reperfusion. MRP14 inhibitor paquinimod
   (10 mg kg^−1, MedChemExpress, USA), or recombined MRP14 protein
   (0.6 mg kg^−1, Cloud‐Clone Corp, China) was intravenously injected into
   the tMCAO mice immediately after the removal of suture as previously
   described.^[ [195]^39 ^] The normal saline was administrated at the
   same volume as vehicle control.

Neutrophil Depletion

   To deplete neutrophils, anti–mouse Ly6G antibody (BioXCell, USA) were
   intraperitoneally injected into mice at a dose of 10 mg kg^−1 for
   consecutive two days before tMCAO procedure. Control mice were
   administrated with the same dose of IgG.

OGD/R

   OGD/R was conducted as previously described.^[ [196]^38 ^] Briefly, the
   culture medium was replaced with Dulbecco's Modified Eagle Medium
   (DMEM) without glucose. Then, the cells were transferred into the
   anaerobic incubator (95% N[2] and 5% CO[2]) at 37 °C. After 6 h of OGD,
   the cells were cultured with normal medium in the normoxic conditional
   incubator for 24 h at 37 °C to allow reperfusion.

Neutrophil Isolation and Transfer

   The bone marrow neutrophils were sorted from the mouse femurs and
   tibias and peripheral neutrophils were sorted from the mouse peripheral
   blood using Histopaque gradients (Sigma–Aldrich, USA).^[ [197]^40 ^] In
   brief, 3 mL Histopaque 1119 was added into the 15 mL tube, and then
   3 mL Histopaque 1077 was gently layered on the top of Histopaque 1119,
   followed by pipetting 1 mL cells suspensions. Then, the tube was
   centrifuged at 872 × g for 30 min at room temperature without brake.
   The cell ring at the interface of the Histopaque 1077 and Histopaque
   1119 layers was the neutrophils. After washed 3 times with phosphate
   buffer solution (PBS, Servicebio, China), neutrophils were suspended
   with RPMI (Gibco BRL, USA) containing 10% fetal bovine serum (FBS,
   Gibco BRL, USA) for the subsequent experiments. For tracing the
   isolated bone marrow neutrophils, MIRB (Biophysics Assay Laboratory,
   Inc. USA) was added into the culture at 12.5 µg mL^−1 concentration for
   24 h. Then, neutrophils were harvested and injected into recipient
   mouse via tail vein (5 × 10^6 neutrophils per mouse). Peripheral
   neutrophils were treated with recombined IL‐1β (MedChemExpress, USA) at
   a concentration of 10 ng mL^−1 for 4 h, and anakinra (10 µg mL^−1,
   MedChemExpress, USA) or PD 169316 (10 µM, MedChemExpress, USA) was
   added 30 min before IL‐1β treatment.

Primary Microglia and Astrocyte Cultures

   Primary microglia and astrocyte cultures were isolated from newborn
   mice pups (1–3 days old) as described previously.^[ [198]^41 ^] The
   cortices were dissected from the pups’ brains and the meninges were
   removed with microscope forceps. Then, the brain tissues were digested
   in 0.25% trypsin/EDTA solution for 15 min at 37 °C and prepared for
   cell suspension to seed onto the poly‐d‐lysine‐coated culture flasks in
   the medium containing 10% FBS and 1% penicillin/streptomycin. The
   cultures were maintained for 14 days and the medium was replaced every
   3 days to generate mixed glial system. In order to separate primary
   microglia from astrocytes, the culture flasks were placed on a gyratory
   shaker (250 rpm for 1 h) and then the detached microglia were collected
   and seeded on new culture plates for further studies. The remaining
   cells were used as primary astrocytes. The primary microglia were
   stimulated with recombined MRP14 (Abcam, China) protein at a dose of
   1 µg mL^−1 for 24 h, followed by ATP (Sigma–Aldrich, USA) treatment for
   30 min. The LPS (Sigma–Aldrich, USA) was added at 500 ng mL^−1 6 h
   before recombined MRP14 or ATP treatment. Resatorvid (the TLR4
   inhibitor, MedChemExpress, USA), FPS‐ZM1 (the RAGE inhibitor,
   MedChemExpress, USA) and JSH‐23 (the NF‐κB inhibitor, MedChemExpress,
   USA) were added into the microglial culture at 100 nM, 10 µM and 10 µM,
   respectively.

Phagocytosis Assay In Vitro

   First, primary microglia were subjected to OGD stimulus for 6 h and the
   recombined MRP14 protein was added at the beginning of reperfusion for
   another 24 h. At the same time, neutrophils were isolated from tMCAO
   mice. Then, primary microglia were co‐cultured with neutrophils for 12
   h (microglia: neutrophils, 1: 3) and washed with PBS to clear suspended
   neutrophils. The phagocytized neutrophils by microglia were measured
   using immunofluorescence staining and Western blot analysis.

ATP Measurements

   The intracellular ATP level was detected using the ATP Assay Kit
   (Beyotime Biotechnology, China). In brief, microglia were collected and
   incubated with 200 µL lysis buffer, followed by centrifugation at 12000
   x g for 5 min. Then, 20 µL supernatant was added into the 96 cell wells
   which contains 100 µL of the ATP reaction buffer. The luminescence of
   each cell well was determined by a microplate reader.

Measurement of Mitochondrial Membrane Potential

   The changes in mitochondrial membrane potential were detected using the
   enhanced mitochondrial membrane potential assay kit with JC‐1 (Beyotime
   Biotechnology, China). The treated microglia were collected and
   incubated with 1 mL of JC‐1 work solution for 20 min at 37 °C. Then,
   microglia were washed with staining buffer and analyzed on flow
   cytometry. The fluorescence change from red to green suggests the
   breakdown of mitochondrial membrane potential.

Single‐Cell Library Preparation and Sequencing

   Single‐cell suspensions were prepared as described previously.^[
   [199]^42 ^] Briefly, the ischemic brain was harvested from WT and MRP14
   KO mice 24 h after the tMCAO procedure, followed by preparation using
   the adult brain dissociation kit (Miltenyi Biotec, Germany). Cell
   suspensions with viability exceeding 85% were used for subsequent
   library preparation. The cells were then loaded onto a Chip A Single
   Cell Kit v2.0 (MobiDrop, cat. no. S050100201), and droplets were
   generated with a MobiNova‐100 (MobiDrop, cat. no. A1A40001). Each
   droplet encapsulated a single cell along with a gel bead coated with
   millions of oligonucleotides containing unique cell barcodes. The
   encapsulated droplets were photo‐cleaved using a MobiNovaSP‐100
   (MobiDrop, cat. no. A2A40001), allowing mRNA to bind to the
   oligonucleotide‐coated gel beads. After reverse transcription, the
   cDNAs were amplified to construct the library using the 3′ Single Index
   Kit (MobiDrop (Zhejiang) Co., Ltd., cat. no. S050300201) and the High
   Throughput Single Cell 3′ RNA‐Seq Kit v2.0 (MobiDrop (Zhejiang) Co.,
   Ltd., cat. no. S050200201), and the libraries were sequenced on the
   Illumina NovaSeq 6000 System.

Fundamental Analysis of scRNA‐seq Data

   Basic cell clustering analysis was performed using the Seurat 4.3.0 R
   package. Cells with gene numbers less than 200 or in the top 1%, or
   with mitochondrial gene content exceeding 25% were considered aberrant
   and excluded from further analysis. Dimensionality reduction was
   performed using principal component analysis (PCA), followed by
   identification of different clusters using the FindClusters function,
   which were visualized using UMAP.

Differentially Expressed Gene and Enrichment Analysis

   To identify DEGs in each cluster, the FindAllMarkers or FindMarkers
   function was employed. The top 20 marker genes from each cluster were
   subjected to GO, KEGG, and Reactome pathway enrichment analysis using
   KOBAS software with Benjamini‐Hochberg correction, and the results were
   visualized using the R package.

Calculation of Gene Set Module Score

   The AddModuleScore method was used to calculate the average expression
   levels of selected gene sets across clusters, with the results
   visualized as violin plots in Seurat (4.2.0). The MSigDB molecular
   signature database
   ([200]https://www.gsea‐msigdb.org/gsea/msigdb/index.jsp) was used to
   select gene sets related to each cell death pattern, as listed in Table
   [201]S2 (Supporting Information).

RNA Sequencing of Neutrophils

   Neutrophils were isolated from the peripheral blood of AIS patients
   with mild and severe cerebral edema, and sham and tMCAO mice. Total RNA
   was extracted using TRIzol reagent (Enzyme, China) according to the
   manufacturer's protocol and subsequently sent to Seqhealth Genomics
   (China) for RNA sequencing. In summary, mRNA was purified from total
   RNA and fragmented into smaller pieces, which were then used for cDNA
   library construction. Sequencing was conducted on the Illumina Hiseq
   platform, with ten replicates performed. DEGs were identified using
   DESeq2 (v1.4.5), with genes displaying a fold change greater than 1 and
   a p‐value of less than 0.05 considered significant. REACTOME pathway
   analysis was utilized to elucidate the altered intracellular signaling
   pathways in the various groups.

Flow Cytometry

   Flow cytometric analysis of the expression of MRP14 in the blood
   neutrophils and monocytes was conducted as previously described.^[
   [202]^43 ^] In brief, the following fluorescence‐conjugated antibodies
   were used to stain the single cell suspensions: APC Mouse Anti‐Human
   CD14 (M5E2, BD Pharmingen), PE‐Cy7 Mouse Anti‐Human CD16 (3G8, BD
   Pharmingen), BV421 Mouse Anti‐Human CD11b (ICRF44, BD Pharmingen), FITC
   Mouse Anti‐Human CD15 (HI98, BD Pharmingen), and PE Conjugated MRP14
   Rabbit mAb (D5O6O, Cell Signaling Technology). Then, the cells were
   analyzed with the flow cytometer (BD Pharmingen, USA) and FlowJo
   software (Treestar, USA).

Evans Blue Dye

   Evans blue dye (4%, 4 mL kg^−1, Sigma–Aldrich, USA) was injected into
   the mice via the tail vein. Four hours after circulation, the mice were
   sacrificed and transcardially perfused with PBS. Then, the mice brains
   were sliced and homogenized in 50% trichloroacetic acid (Sigma–Aldrich,
   USA) at 65 °C for 24 h. After centrifuged at 12, 000 × g for 20 min,
   the supernatant was collected and measured by spectrophotometry at
   absorbance of 620 nm to quantify the leakage of Evans blue dye.

Measurements of BBB Permeability

   To determine the permeability of the in vitro BBB model, the TEER value
   and FITC‐dextran trans‐endothelial permeability were measured. The
   endothelial voltohmmeter (EVOM, World Precision Instruments, USA) was
   used to calculate the TEER value following the manufacturer's
   instruction: TEER value (Ω × cm^2) = [(resistance of cells insert –
   resistance of blank insert) × superficial area of the insert]. For the
   FITC‐dextran trans‐endothelial permeability assay, 1 mg mL^−1 of
   FITC‐dextran (MW, 70, 000 kd, Sigma–Aldrich, USA) was first added into
   the upper chamber for 2 h and then the absorbance in the lower chamber
   was measured via a microplate reader at 450 nm.

Dual‐Luciferase Assay

   The MRP14 luciferase reporter plasmid, Renilla control reporter plasmid
   and C/EBP β plasmid were co‐transfected into the HEK293T cells. After
   48 h of transfection, HEK293T cells were lysed and the firefly and
   Renilla luciferase activity were measured using the Dual Luciferase
   Reporter Assay Kit (Vazyme, China) following the manufacturer's
   protocol. The Renilla control reporter was used as an internal control.

Western Blot Analysis

   Cell and tissue samples were lysed with the RIPA buffer (Beyotime
   Biotechnology, China) which contains 1% protease inhibitors (Beyotime
   Biotechnology, China) to extract proteins. Then, an equal amount of
   protein extract was added into the SDS‐PAGE gels and transferred onto
   the PVDF membranes (Millipore, USA). 5% skim milk buffer was used to
   block nonspecific binding and the transferred PVDF membranes was
   incubated with primary antibodies overnight at 4 °C: MRP14 (1:1000,
   73425, Cell Signaling Technology), Occludin (1:1000, 13409‐1‐AP,
   Proteintech), ZO‐1 (1:1000, 21773‐1‐AP, Proteintech), VE‐Cadherin
   (1:1000, AF1002, RD system), MPO (1:1000, 22225‐1‐AP, Proteintech),
   Caspase‐1 (1:50, A0964, Abclonal), IL‐1β (1:500, A16288, Abclonal),
   NLRP3 (1:500, [203]T55651, Abmart), GSDMD (1:1000, ab209845, Abcam),
   p65 (1:500, A19653, Abclonal), p‐p65 (1:500, TA2006, Abmart), IκB
   (1:500, A1187, Abclonal), p‐IκB (1:500, AP0707, Abclonal), Tubulin
   (1:1000, [204]M20005, Abmart), p‐p38 (1:1000, 4511, Cell Signaling
   Technology), p38 (1:1000, 8690, Cell Signaling Technology), C/EBP β
   (1:1000, 23431‐1‐AP, Proteintech), TLR4 (1:1000, 66350‐1‐Ig,
   Proteintech), β‐actin (1:1000, AC026, Abclonal), H3 (1:1000,
   17168‐1‐AP, Proteintech). The membranes were then incubated with
   secondary antibodies for 2 h at room temperature and visualized using
   ECL solution on the chemiluminescence system (UVP, USA). Image J
   software was used to detect the intensity of the bands.

Immunofluorescence

   The brain sections were first heat‐treated with citrate buffer solution
   to retrieve the antigen and then incubated with 0.3% Triton X‐100 and
   10% donkey serum for 1 h at room temperature. Next, the sections were
   incubated with the following primary antibodies overnight at 4 °C:
   MRP14 (1:50, 73425, Cell Signaling Technology), Ly6G (1:50, ab25377,
   Abcam), VE‐Cadherin (1:50, AF1002, RD system), Fibrinogen (1:50,
   ab118533, Abcam), Lectin (1:50, DL‐1177‐1, Vector laboratories), GSDMD
   (1:50, ab209845, Abcam), Iba‐1 (1:50, ab5076, Abcam), CD68 (1:50,
   [205]GB113109‐100, Servicebio). After washed with PBST buffer,
   fluorescein‐conjugated secondary antibodies were added. All brain
   sections were captured using the confocal microscope (Nikon, Japan).

Enzyme‐Linked Immunosorbent Assay (ELISA)

   After centrifuged at 400 × g for 10 min, the supernatant of cell
   culture medium and plasma samples were collected and stored at −80 °C
   for further measurement. The MRP14 and PGLYRP1 levels in the plasma of
   patients and mice, and the IL‐1β levels in the cell culture supernatant
   were quantified using the ELISA kits (Bioswamp, China) following the
   manufacturer's instructions.

Real‐Time Quantitative PCR

   The RNA extraction reagent (Vazyme, China) was used to extract total
   RNAs from neutrophils and brain tissues. Then, total RNA was
   reverse‐transcribed into cDNAs by HiScript III‐RT SuperMix (Vazyme,
   China) and amplified using SYBR Green Mix (Vazyme, China) according to
   the manuals. Primers for PCR were listed in Table [206]S3 (Supporting
   Information). The comparative 2‐ΔΔCt method against β‐actin was applied
   to quantify the gene expression.

Statistical Analysis

   GraphPad Prism 9.0 software was used to perform the statistical
   analyses presented in this study. All data were presented as mean ± SEM
   and the sample size (n) is indicated in the figure legends. The
   Shapiro‐Wilk test was used to assess normality. For normally
   distributed data, an unpaired two‐tailed Student's t‐test was applied
   for comparisons between two groups, while one‐way ANOVA followed by
   Tukey's post hoc test was used for comparisons involving more than two
   groups. For data not following a normal distribution, the Mann‐Whitney
   test was employed for two‐group comparisons, and Kruskal‐Wallis test
   followed by Dunn's test was used for nonparametric analysis involving
   more than two groups.

Conflict of Interest

   The authors declare no conflict of interest.

Supporting information

   Supporting Information
   [207]ADVS-12-e03722-s001.docx^ (11.5MB, docx)

Acknowledgements