Neutrophils are identified as the most active cellular population during the acute phase post-ischemic stroke and represent the initial immune cells infiltrating the brain parenchyma [1]. In the pro-inflammatory environment induced by ischemic stroke, these neutrophils undergo hyperactivation, leading to their rapid exponential proliferation [2]. This exacerbates the inflammatory response and leads to further neural tissue damage. Simultaneously, activated neutrophils consume significantly more oxygen, generating large quantities of oxygen free radicals that harm cells [3]. Microglia, the primary resident immune effectors of the central nervous system (CNS), exhibit rapid chemotactic migration toward sites of injury following activation, thereby initiating neuroinflammatory responses. Notably, these activated microglia actively participate in the phagocytic clearance of neutrophils within the ischemic brain tissue [2]; on the other hand, they secrete inflammatory mediators that exert a “double-edged sword” effect in modulating neuroinflammation [4]. Therefore, enhancing microglial phagocytic capacity represents a potential therapeutic strategy to mitigate inflammatory infiltration following cerebral ischemia-reperfusion injury.

Microglia require substantial energy to sustain diverse functions, with their phagocytic activity primarily relying on oxidative phosphorylation (OXPHOS) for ATP production [5]. OXPHOS-derived ATP is critical for maintaining microglial homeostasis and effector functions [6]. This metabolic pathway occurs within mitochondria and constitutes the principal route for energy generation in homeostatic microglia [7]. In Alzheimer’s disease, metabolic reprogramming of microglia modulates their phagocytic clearance of β-amyloid plaques [8]. The unbiased identification of cellular subclusters via single-cell RNA sequencing (scRNA-seq) enables a deeper understanding of their heterogeneous roles in tissue homeostasis and disease pathology. Additionally, pseudotime analysis and transcription factor analysis derived from scRNA-seq provide innovative tools for unraveling cellular differentiation. For example, researchers have employed scRNA-seq to investigate [9] and have uncovered novel cellular subclusters, elucidating the complex pathophysiological mechanisms underlying cerebral tissue injury during cerebral ischemia-reperfusion. scRNA-seq facilitates a meticulous elucidation of the molecular mechanisms underpinning cerebral ischemia-reperfusion injury and allows for a rigorous assessment of the efficacy of targeted therapeutic interventions.

Lrg1 has been identified as a potential molecule upregulated following cerebral ischemia-reperfusion [10]. A comparative analysis undertaken between Lrg1 knockout mice and their wild-type littermates, with a primary focus on assessing brain infarct volume, the temporal progression of cerebral edema, and neurological impairments induced by ischemia-reperfusion injury, demonstrated that Lrg1 deficiency bestows considerable neuroprotective benefits, thereby effectively preserving brain tissue from injury. Lrg1 has emerged as a critical mediator upregulated in response to cerebral ischemia-reperfusion. Comparative evaluation of Lrg1^-/- and wild-type mice following MCAO/R (Middle Cerebral Artery Occlusion/Reperfusion) revealed that Lrg1 knockout significantly attenuated brain infarct volume, reduced edema formation, and improved neurofunctional recovery. These findings establish Lrg1 as a promising therapeutic target whose inhibition confers neuroprotective effects against ischemic brain damage. Our previous scRNA-seq studies demonstrated that Lrg1 orchestrates multiple cellular responses critical to cerebral ischemia-reperfusion injury, including groundbreaking discoveries revealing novel microglia differentiation trajectories [11]. We identified 7 microglia subclusters differentiating along two pathways [11]. Similarly, there is limited literature on single-cell studies of neutrophils. Also, in addition, the clearance mechanism of neutrophils in ischemic brain tissue remains not fully understood. Given the crucial role of neutrophils in diseases, we aim to investigate the therapeutic potential of modulating Lrg1 to clear neutrophils that accumulate after cerebral ischemia-reperfusion.

In this research, using Lrg1 knockout mice, we examined how Lrg1 modulation influences post-ischemic neutrophil clearance in a cerebral ischemia-reperfusion model. We focused particularly on microglia and their metabolic regulatory pathways, employing a variety of techniques, including single-cell RNA sequencing, three-dimensional imaging, immunofluorescence, and western blotting. We aim to characterize the molecular mechanisms underlying Lrg1’s role in the regulation of neutrophils by microglia following cerebral ischemia-reperfusion.

In the immediate aftermath of ischemic stroke, rapid neutrophil infiltration is initiated, driven by blood-brain barrier disruption. This breach facilitates the entry of peripheral immune cells, predominantly neutrophils and macrophages, into the ventricular system. The infiltration reaches its zenith between 24 and 72 h post-stroke, marked by a substantial surge in N1 neutrophils compared to N2 neutrophils [16]. Thereafter, the neutrophil count begins to dwindle by days 4 and 7, with a dramatic decrease observed in the next 14 days. Throughout this inflammatory cascade, neutrophils trigger the production of ROS, which exacerbates hypoxic conditions and disrupts the oxidative phosphorylation process within the mitochondrial respiratory chain, leading to mitochondrial depolarization and elevated O₂ levels. Concurrently, the biosynthesis of cytokines such as TNF-α and IL-6 occurs across multiple brain regions [17]. Neutrophil infiltration persists throughout the acute, subacute, and regenerative phases of cerebral ischemia, underscoring its pivotal role in orchestrating the complex inflammatory response following ischemic stroke. The scRNA-seq results provide valuable insights into the cellular subsets and differentiation patterns in cerebral ischemia-reperfusion injury. Available literature evidence also hints at a greater degree of neutrophil heterogeneity in nervous system diseases, yet the mechanisms underlying the microenvironmental regulation of neutrophil status remain elusive [18]. In particular, by focusing on microglia, we examine their phagocytic role in clearing neutrophils from the brain parenchyma, thereby attenuating neuroinflammatory responses. Parallel investigations under this paradigm also dissect the molecular pathways potentially mediating this regulatory function.

Figs. 1, 2 and 5A suggest that after MCAO/R or Lrg1 knockout, the interactions among various cell types have undergone changes. We selected 2 representative cell types–endothelial cells and microglia, and found that there is immunofluorescence colocalization expression between endothelial cells/microglia and neutrophils, leading us to speculate that there may be interactions such as penetration or phagocytosis occurring. There are numerous reports detailing the effects of microglia on neutrophils. In aged brains post-stroke, 6 microglial subclusters exist, with MG5 being highly proliferative and MG6 mimicking neutrophils [19]. Microglial cell are the major cell type involved in debris removal, which chew up the debris, the pathogene, and the bacteria in response to ischemia damage [20]. The adherence of neutrophils to endothelial cells stands as a cornerstone mechanism in the initiation of inflammation during ischemia-reperfusion scenarios. This crucial interaction significantly contributes to the sequential chain of responses that hasten the commencement of inflammation subsequent to episodes of diminished blood flow and the ensuing re-establishment of circulation. When neutrophils adhere to the endothelial lining of blood vessels, they facilitate the infiltration of immune cells into tissues, which triggers a series of inflammatory reactions aimed at restoring tissue homeostasis but can also cause damage if unregulated. Therefore, the adhesion of neutrophils to endothelial cells constitutes a critical early step in the inflammatory process that characterizes MCAO/R injury [21,22]. Additionally, there are also reports that activated vascular smooth muscle cellular subcluster was enriched in neutrophil-mediated immune responses and exocytosis regulation, while specific subcluster 5, linked to type I interferon, had a reduced proportion in the MCAO group [23].

As the precious value of cellular heterogeneity, researchers found through scRNA-seq experiments that 4 neutrophil clusters were identified in the brain, with NEUT0, the most mature, highly expressing CXCL1 [23]. scRNA-seq revealed 6 neutrophil clusters in the blood [24], with age-related changes leading to atypical neutrophils exacerbating stroke. Aged mice had higher CXCL3 production by CD62^low neutrophils, linked to poorer reperfusion and outcome. Elevated plasma levels of neutrophil growth factors and proinflammatory mediators were also seen in aged mice. Yang et al. distinguished three neutrophil subsets in the brain, with NEUT1, composed of infiltrated neutrophils, increasing from 0.3% to 18.7% post-stroke. NEUT1 cells showed high activation, with NEUT2 categorized as immature [19]. Through dimensionality reduction, 5 subclusters of neutrophils were identified following ischemic stroke, with the highest cell count observed in MCAO+WT, indicating the infiltration of neutrophils after cerebral ischemia. Additionally, Fig. 4C reveals that NEUT1 accounts for the largest proportion, albeit with varying percentages among different groups. Further experimental evidence is required to investigate the functions of NEUT1. Through further comparison of differentially expressed genes, the differentially expressed genes Camp, Ngp, Lfitm6, Ltf, and Ly6g exhibit higher positive correlation in expression within the NEUT1 subclusters. Research has demonstrated that Ly6G is critically involved in the infiltration, recruitment, and migration of neutrophils during inflammatory responses [25]. With the onset of immunosenescence or aging, the functionality of innate immune cells, including neutrophils and macrophages, becomes compromised, potentially paving the way for pathogen invasion [26,27]. Moreover, a strong correlation has been found between the expression level of Ly6G and the immune functions of neutrophils in antibacterial immunity. These immune capabilities, such as phagocytosis, ROS production, NET formation, migration, and bacterial clearance, weaken with aging [28]. Here, we suggest that the proportion of NEUT1 decreases after MCAO modeling, but increases after Lrg1 knockout, which also indicates an increase in NEUT1. Our research findings have surpassed the conventional limitations of the defined polarization states of neutrophil, which are influenced by a multitude of disease conditions.

Under physiological conditions, homeostatic microglia primarily use oxidative phosphorylation (OXPHOS) for ATP production, but switch to glycolysis during inflammation to meet energy demands. This metabolic shift enables rapid responses to danger signals, critical under pathological conditions [29]. Glycolysis inhibits energy-intensive phagocytosis by providing rapid ATP, while Resolvin D1 reverses this reprogramming to sustain microglial clearance activity [2]. CD68 serves as a well-established marker for phagolysosomes within microglia and macrophages. It is particularly noteworthy that the phagocytic ability of microglia exhibits a substantial and statistically significant enhancement in the MCAO/R+WT and MCAO+Lrg1^-/- groups, as depicted in Fig. 6C,D [30]. In Alzheimer’s models, enhancing OXPHOS improves amyloid-β phagocytosis and cognitive outcomes [31]. Chronic Aβ exposure induces microglial immune tolerance via downregulated glycolysis/OXPHOS, impairing reactivity [29]. Conversely, IL-33 promotes glycolysis in primary microglia, activating AKT/NDRG1/mTOR pathways to modulate neurodevelopment [32]. Aging microglia shift to anaerobic glycolysis, increasing inflammatory cytokines/ROS and exacerbating damage [33]. Trem2 deficiency disrupts OXPHOS, altering respiratory gene expression and ATP synthesis, thereby affecting neuronal metabolism [34]. KEGG analysis of MCAO+Lrg1^−/− mice showed OXPHOS pathway enrichment, suggesting Lrg1 knockout modulates microglial bioenergetics. However, the molecular mechanisms of neutrophil phagocytosis by microglia require further study.

Glycolysis transpires in the cytoplasm, whereas oxidative phosphorylation unfolds within mitochondria. While glycolysis offers swift energy release, oxidative phosphorylation, albeit accompanied by the production of ROS, yields substantially more energy for the cell. Specifically, oxidative phosphorylation generates approximately 32 ATP molecules per glucose molecule metabolized, in contrast to glycolysis’s mere 2 ATP molecules. This intricate process mandates the participation of proteins situated within the mitochondrial inner membrane transport chain to facilitate the sequential transfer of electrons through a cascade of oxidation-reduction reactions: commencing with Complex I, advancing to Complex II, then moving on to Coenzyme Q (CoQ), followed by Complex III, Cytochrome c, and ultimately reaching its terminus at Complex IV [35]. Notably, the subunits of these five oxidative phosphorylation complexes are encoded by diverse genomes, such as Complex I encoded with Ndufb8 [36], Complex II encoded with Sdhb [36], Complex III encoded with Uqcrc1 [37], Complex IV encoded with Mtco2 [38], and Complex V encoded with Atp5a1 [39] following MCAO/R, a decrease is noted in the expression of signaling molecules related to oxidative phosphorylation. Conversely, when Lrg1 is knockout, a significant increase is observed in the expression of these oxidative phosphorylation signaling molecules, which is accompanied by an augmentation in the levels of ROS. This finding indicates that the absence of Lrg1 results in the enhancement of oxidative phosphorylation in microglia after MCAO/R.

Despite our findings, our study is constrained by several limitations. Notably, our analysis of microglia predominantly hinges on informatics outcomes stemming from single-cell sequencing, particularly in the context of elucidating intercellular communication and oxidative phosphorylation mechanisms. Although we have experimentally confirmed the regulatory mechanism of Lrg1 on oxidative phosphorylation in microglia, the literature on this topic remains sparse, necessitating further investigation into the differentiation process and underlying molecular mechanisms.

The aforementioned research demonstrates that Lrg1 regulates both microglial phagocytosis and the subsequent phenotypic switching of neutrophils in post-ischemic brain tissue. Lrg1 knockout results in a reduction in neutrophil count, and this regulation is associated with the oxidative phosphorylation pathway in microglia modulated by Lrg1 (Fig. 8). Given the correlation between neutrophils and inflammatory tissue infiltration after cerebral ischemia, Lrg1 emerges as a promising target for intervening in the inflammatory and immune responses in ischemic stroke.