Membrane-camouflaged metal-phenolic nanomedicines for the treatment of ischemic stroke via relieving oxidative stress and neuroinflammation

Abstract

Ischemia-reperfusion injury (IRI) following thrombolytic therapy significantly influences the ischemic stroke outcomes. Here, we develop M@EFE NPs, a biomimetic nanomedicine, to alleviate the reperfusion injury. This nanomedicine is constructed by encapsulating Edaravone into the metal-phenolic nanoparticles self-assembled from epigallocatechin gallate (EGCG) and iron ions (Fe³⁺), and further coating with macrophage membranes. This design integrates the antioxidant and anti-lipid peroxidation properties of the EGCG-Fe³⁺ nanoparticles with the inflammatory targeting capacity of macrophage membranes. Following systemic administration, M@EFE NPs are able to penetrate the blood-brain barrier and target the ischemic regions, thereby inhibiting oxidative damage, protecting neurons, and ultimately improving stroke outcomes. In vitro, the nanomedicine can effectively scavenge reactive oxygen species, inhibit lipid peroxidation, and enhance cell survival. In murine middle cerebral artery occlusion models, M@EFE NPs could reduce infarct volume, attenuate neuronal apoptosis and neuroinflammation, and improve motor recovery and long-term survival. This study presents a promising combinatorial nanotherapeutic strategy for ischemic stroke, providing insights into biomimetic nanomedicine for treating cerebral IRI.

Graphical abstract

Highlights

• •A biomimetic nanomedicine has been designed to alleviate ischemia-reperfusion injury.
• •The nanomedicine can target the ischemic regions and inhibit oxidative damage.
• •The nanomedicine reduces infarct volume and improves long-term survival of animals.

1. Introduction

Ischemic stroke (IS), a devastating neurological disorder, remains a leading cause of global mortality and long-term disability [1], [2]. Prompt restoration of cerebral blood flow through reperfusion therapy is essential for salvaging ischemic brain tissue [3]. Intravenous recombinant tissue plasminogen activator (rt-PA), an FDA-approved thrombolytic agent, facilitates early vascular recanalization [4], [5], [6]. However, clinical and experimental evidence indicates that reperfusion therapy can trigger destructive cascades known as ischemia-reperfusion injury (IRI), which exacerbates tissue damage [7], [8]. Therefore, alleviating reperfusion injury is the key target for improving the prognosis of cerebral ischemia.

Cerebral IRI triggers a destructive pathological cascade characterized by excessive reactive oxide species (ROS) production [9], lipid peroxidation [10], ferroptosis [11], and uncontrolled neuroinflammation [12]. Edaravone (EDV), one of the few clinically approved free radical scavengers for acute ischemic stroke [13], [14], suffers from a short half-life, rapid metabolism, and inadequate blood-brain barrier (BBB) penetration [15], [16], which collectively compromise its therapeutic efficacy.

Nanotechnology offers promising strategies to overcome these limitations [17], [18], [19]. Recent advances in nanomedicine have proposed various strategies to improve edaravone delivery for cerebral IRI. ROS-responsive nanoparticles that enable pathological-activated release of EDV in the ischemic penumbra were developed, minimizing off-target ROS scavenging while enhancing neuroprotection [20]. Another innovative approach utilized intranasal administration of EDV in ionic liquid form, which significantly improved brain delivery efficiency and ameliorated cerebral IRI in transient middle cerebral artery occlusion (MCAO) rat models by avoiding the BBB [21]. Especially, meta-polyphenol network nanoparticles, which integrate the catalytic functions of metal ions with the inherent antioxidant capacity of polyphenols, have demonstrated potential in anti-inflammatory and antioxidant therapies [22], [23], [24]. Epigallocatechin gallate (EGCG), a naturally occurring polyphenol, exhibits strong antioxidant and anti-inflammatory properties [25], [26], making it a compelling candidate for IRI therapy. However, the clinical translation of EGCG is hampered by its instability and low bioavailability in vivo [27]. Notably, EGCG can coordinate with iron ions to form Fe-EGCG network nanoparticles [28], [29], serving as an ideal carrier for free drugs. Such coordination not only enhances the stability of EGCG but also augments its ROS scavenging efficacy. Furthermore, EGCG-mediated iron chelation helps regulate iron homeostasis and suppress ferroptosis, a regulated cell death pathway implicated in neuroinflammation [30]. Despite these advantages, their application in neurological disorders remains limited, primarily due to poor pharmacokinetics, insufficient BBB penetration, and lack of targeted accumulation at disease sites.

In recent years, biomimetic nanotechnology, particularly cell membrane coating, has emerged as a promising strategy to overcome the limitations of conventional nanomedicine [31], [32]. This approach involves cloaking synthetic nanoparticles with natural cell membranes, such as red blood cell membranes [33], platelet membranes [34], cancer cell membranes [35], [36], or neutrophil membrane [37], [38], thereby endowing the nanoparticles with complex biological functions inherited from the source cells. Among various cell membrane coating strategies, macrophage membrane-coated nanoparticles have garnered particular interest for treating inflammatory diseases such as IRI [39], [40]. Compared to other membrane types, macrophage membranes offer several unique advantages for cerebral IRI therapy. Firstly, macrophage membranes are intrinsically equipped with a broad array of chemokine receptors (like CCR2 and CCR5) and adhesion molecules (integrins such as α4β1 and αLβ2) that enable active chemotaxis toward inflamed endothelium and ischemic brain regions. Besides, macrophages exhibit remarkable phenotypic plasticity, and membranes derived from resting or M2-polarized macrophages can provide “don't eat me” signals that prolong systemic circulation while avoiding excessive immune activation. Moreover, the relatively large membrane surface area and robust mechanical properties of macrophages facilitate efficient coating of nanoparticle cores without compromising the integrity of the nanoparticle structure. While platelet and neutrophil membranes have shown efficacy in targeting IRI, their targeting mechanisms are largely restricted to specific molecular interactions with damaged vasculature or inflamed endothelium, respectively, potentially limiting their versatility.

In this study, we developed a biomimetic nanomedicine (designated M@EFE NPs) through the encapsulation of EDV into EGCG-Fe³⁺-based network nanoparticles followed by macrophage membrane coating (Scheme 1a). This design enables the nanoplatform to mimic macrophage-like inflammatory targeting, facilitating efficient drug delivery to ischemic brain regions and promoting synergistic therapeutic outcomes. At the cellular level, M@EFE NPs effectively scavenged ROS, suppressed lipid peroxidation, and consequently reduced apoptosis while enhancing cell viability. In mouse MCAO models, the M@EFE NPs exhibited enhanced brain-targeting capability (Scheme 1b, c). Through behavioral tests and histological staining, we found that this integrated nanoplatform synergistically mitigated oxidative stress, ferroptosis, and neuroinflammation, thereby attenuating ischemic brain injury. Our findings indicate that M@EFE NPs not only promote targeted accumulation in the ischemic hemisphere but also significantly reduce infarct volume and improve functional recovery. This work presents an ingenious combinatory nanotherapeutic strategy for ischemic stroke and establishes a foundational framework for the application of biomimetic nanomedicine in cerebral IRI.

Scheme 1.

Scheme of the neuroprotective mechanisms of M@EFE NPs. (a) Synthesis of M@EFE NPs. Edaravone was encapsulated into the EGCG-Fe³⁺ nanoparticles and further coated with macrophage membrane. (b) Coating with macrophage membranes, M@EFE NPs can effectively cross the BBB and accumulate in ischemic regions after systemic administration. (c) M@EFE NPs can effectively scavenge ROS, inhibit lipid peroxidation, and consequently reduce apoptosis and enhance cell viability.

2. Material and methods

2.1. Synthesis of EFE NPs

To prepare EDV-loaded Fe³⁺-EGCG nanoparticles (EFE NPs), EGCG (10 mg/mL) and FeCl₃ (40 mM) were dissolved into DI water, and EDV (50 mg/mL) was dissolved into DMSO. EGCG (500 μL), FeCl₃ (500 μL) and EDV (20 μL) solution were mixed in 5 mL of PBS and stirred for 6 h. Subsequently, the solution was centrifuged to obtain EFE NPs (15000 rpm, 5 min), and the precipitation was washed with DI water several times to remove the free drugs.

2.2. Membrane extraction

The macrophage membrane was extracted using ultrasonic fragmentation method. Briefly, macrophages were cultured in Dulbecco's Modified Eagle Medium (DMEM) medium at 37 °C with 5% CO₂ until the cell density reached approximately 70-80% confluence in 10 × 15 cm culture dishes. The cells were then harvested, resuspended in hypotonic lysis buffer, and disrupted using an ultrasonic cell disruptor (300 W power, with 3 s pulses applied over a total duration of 20 min). The resulting suspension was then centrifuged at 10000 g for 20 min at 4 °C. The supernatant was collected and further centrifuged at 100000 g for 1 h at 4 °C. The final pellet was resuspended in deionized water to obtain the macrophage membrane. Membrane protein content was quantified using a BCA assay, and the prepared membranes were stored at −80 °C.

2.3. Synthesis of M@EFE NPs

To prepare membrane-coated EFE NPs (M@EFE NPs), EFE NPs were mixed with macrophage membranes at a 1:1 mass ratio and sonicated at 180 W for 2 min. The suspension was then centrifuged at 500 g for 5 min at 4 ^°C to remove free membranes.

2.4. Cellular uptake

To evaluate cellular uptake of the nanoparticles, EFE and M@EFE NPs were labeled with fluorescence probe 1,1′-Dioctadecyl-3,3,3′,3′-Tetramethylindodicarbocyanine Perchlorate (DiD). HT22 cells were seeded in confocal dishes and cultured for 24 h, followed by treatment with DiD-labeled EFE or M@EFE NPs (50 μg/mL) for different durations (1 h, 2 h, 4 h, 8 h, 12 h, and 24 h). After treatment, cell nuclei were stained with DAPI, and internalization of the nanoparticles was visualized using a CLSM.

For flow cytometry analysis, HT22 cells were plated in 12-well plates. Following drug administration at different time points, cells were collected, washed, and analyzed by flow cytometry to determine the fluorescence intensity associated with nanoparticle uptake.

2.5. OGD/R model construction

HT22 cells were exposed to OGD/R to mimic cerebral IRI. Briefly, the complete medium was replaced with glucose- and serum-free DMEM. The cells were then sealed in anaerobic gas-producing bags and incubate at 37 °C for 12 h to induce OGD. To model reperfusion, the cells were returned to normoxic conditions, the deprivation medium was replaced with complete DMEM (containing the drug or vehicle control), and incubation continued for 4 h at 37 °C in a 5% CO₂ atmosphere.

2.6. ROS detection

Following 12 h of OGD/R treatment, HT22 cells were incubated with PBS, EDV (5 μg/mL), EFE NPs (50 μg/mL), or M@EFE NPs (50 μg/mL) for 4 h, respectively. The cells were then washed with PBS and incubated with DCFH-DA for 30 min. Intracellular ROS levels were detected by CLSM and flow cytometry.

2.7. Mitochondrial membrane potential detection

HT22 cells were seeded in 12-well plates and treated with various formulations for 4 h after 12-h OGD/R. To measure changes in mitochondrial membrane potential, the cells were collected and washed with PBS for MT1 assay. Following the MT1 kit instructions, flow cytometry and confocal microscopy were performed to assess mitochondrial membrane potential.

2.8. Lipid peroxidation staining

HT22 cells were seeded in confocal dishes and treated with various formulations for 4 h. The cells were then washed with PBS and stained with BODIPY for 1 h. The cellular fluorescence was observed by CLSM. Average oxidation ratio was measured by Image J.

2.9. In vitro BBB penetration

To establish the in vitro BBB model, Transwell system was used in which bEnd.3 cells were seeded in the upper chamber and HT22 cells in the lower compartment. The bEnd.3 cells were first treated with EFE NPs or M@EFE NPs for 4 h. After treatment, the Transwell inserts were placed into the lower plates containing HT22 cells and co-cultured for 2 or 6 h. Subsequently, the HT22 cells were collected from the lower chamber, and the intracellular fluorescence intensity was quantified by flow cytometry.

2.10. MCAO model construction

To establish MCAO model, C57BL/6 mice (8 weeks old, 18-22 g) were anesthetized with 1.5-2% isoflurane. After shaving and disinfecting the neck, a midline neck incision was made to expose the right common carotid artery, internal carotid artery, and external carotid artery. A suture was used to ligate the distal end of the external carotid artery 4 mm from the bifurcation of the common carotid artery. Another suture was threaded through the external carotid artery and tied in a slipknot near the bifurcation. The common carotid artery was clamped with an arterial clamp. Make a small incision in the external carotid artery 3 mm from the bifurcation. Insert a pre-tapered 0.33 mm diameter nylon thread through the incision into the internal carotid artery, advancing it into the middle cerebral artery. The insertion depth of the nylon thread was approximately 16 ± 1 mm from the bifurcation. After 60 min of ischemia, remove the suture plug. Ligate the proximal end of the external carotid artery with silk suture. Close the neck wound with silk sutures. Disinfect the wound with povidone-iodine. Place the mouse on a heating pad. Once fully awake, transfer it to a temperature-controlled incubator for recovery.

2.11. In vivo brain-targeting

The in vivo brain-targeting capability was assessed in a mouse MCAO model. One hour after ischemia induction, the mice received an intravenous injection of free DiD, DiD-labeled EFE NPs, or M@EFE NPs (50 μg per mouse) via the tail vein. Twenty-four hours post-injection, the mice were euthanized, and their brains were excised for immediate ex vivo imaging using an IVIS system. Subsequently, the brains were snap-frozen at −80 ^°C, sectioned, and the fluorescence distribution within the brain sections was examined under a fluorescence microscope.

2.12. TTC staining

To quantify infarct volume after ischemia-reperfusion, MCAO was induced in C57BL/6 mice. One hour after occlusion, EDV, EFE, or M@EFE (50 μg per mouse) was administered via tail vein injection. Twenty-four hours later, the mice were euthanized, and the brains were harvested. The brains were briefly frozen at −20 ^°C for 30 min and then sectioned coronally. The sections were immersed in 1% (w/v) TTC solution pre-warmed to 37 ^°C and incubated for 10 min at 37 ^°C with periodic gentle agitation to ensure uniform staining. Viable brain tissue-stained brick red, while the infarcted areas remained pale.

2.13. Suspension test

Mice were placed on the wire mesh (1 m above the ground). The mesh was then slowly inverted to initiate the test. After several practice sessions, a formal trial was performed: each mouse was placed on the mesh, which was subsequently inverted while a timer was started. The latency to fall was recorded for each animal.

2.14. Rotarod test

Mice (3-6 per group) were first trained on a rotating rod apparatus with the speed gradually increasing to 40 rpm over 3-5 training sessions. During formal testing, the apparatus was cleaned with 75% ethanol before each trial to eliminate olfactory cues. The rotation speed was gradually increased to 40 rpm and then maintained at this constant speed. The latency to fall from the rod was recorded for each mouse.

2.15. Open-field test

Prior to testing, mice were placed in an open field room for 6 h to acclimate to the environment. During the formal experiment, mice were positioned at the center of a 1 m × 1 m open field. A thermal imaging camera recorded each mouse's movement trajectory within a 5-min period. Analysis software was used to generate movement trajectory maps, calculate the total distance traveled by each mouse, and determine the distance traversed within the central area.

2.16. Pathological evaluation

The brain was harvested and fixed in 4% paraformaldehyde solution. Following dehydration with sucrose solution, tissue was embedded in OCT and sectioned into 10-μm cryosections. Nissl staining was performed to assess infarct area.

3. Results and discussion

3.1. Synthesis and characterization of M@EFE NPs

To prepare EFE NPs, EDV was encapsulated into the metal-phenolic network nanoparticles self-assembled from EGCG and Fe³⁺. Subsequently, macrophage membrane was coated onto the EFE NPs to form M@EFE NPs. Transmission electron microscopy (TEM) images showed that both EFE and M@EFE NPs assembled into well-defined nanostructures (Fig. 1a). The coordination between EGCG and Fe³⁺ was confirmed by X-ray photoelectron spectroscopy (XPS). XPS spectra displayed a characteristic Fe-O peak, suggesting the coordinate interaction between Fe³⁺ and phenolic hydroxyl groups of EGCG (Fig. 1b). Besides, UV-vis spectroscopy showed that EFE NPs exhibited a characteristic absorption peak of Edaravone at 244 nm, confirming successful drug encapsulation (Fig. 1c). Dynamic light scattering (DLS) measurements indicated an increase in the hydrodynamic diameter of the nanoparticles after membrane coating, with EFE NPs and M@EFE NPs measuring approximately 208 nm and 247 nm, respectively (Fig. 1d). To further confirm the membrane coating, sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) was performed. M@EFE showed similar protein bands with macrophage membrane, confirming successful membrane coating (Fig. S1). Furthermore, western blot analysis was performed to specifically identify the membrane proteins. Characteristic macrophage markers, including CD11b, CD47, and CD68, were clearly detected in the macrophage, macrophage membrane, and M@EFE groups, whereas no corresponding bands were observed in the bare EFE group. These results further verified that the functional membrane proteins were successfully preserved on the M@EFE surface (Fig. 1e). The stability tests over 3 days revealed no significant changes in the hydrodynamic diameters of either EFE or M@EFE NPs. Both groups maintained the polydispersity index (PDI) of around 0.15, confirming their narrow size distribution and excellent colloidal stability (Fig. S2). Collectively, these results indicated the successful synthesis of M@EFE NPs. Using UV spectrophotometry, the drug loading rate of Edaravone in M@EFE NPs was determined to be 32.37% (Table S1).

Fig. 1.

Characterization of M@EFE NPs. (a) TEM image of EFE and M@EFE NPs. Scale bar, 100 nm. (b) XPS spectra of M@EFE NPs. (c) UV-vis spectra of free EGCG, FeCl₃, EDV, and M@EFE NPs. (d) Hydrodynamic diameter of EFE and M@EFE NPs (n = 3). (e) Western blot analysis of macrophage membrane proteins in different groups. (f) DPPH scavenging ability of M@EFE NPs. (g) ABTS scavenging ability of M@EFE NPs. (h) ESR spectra ESR spectra assessing the •O₂⁻ and •OH scavenging capacity of M@EFE NPs. Data are presented as means ± SD.

ROS generation is a key mechanism contributing to neuronal damage during IRI, making the scavenging of ROS essential for neuroprotection [41], [42]. The ROS scavenging ability of M@EFE was first evaluated using the 1,1-diphenyl-2-picrylhydrazyl (DPPH) radical assay, a widely adopted method for the evaluation of antioxidant activity. DPPH shows a strong absorption peak at 517 nm, which decreases as electrons are paired upon reaction with antioxidants [43]. As shown in Fig. 1f, treatment with M@EFE led to a notable reduction in the characteristic absorption of DPPH. Quantitative results indicated a scavenging rate of 58.4%, revealing the strong ROS-scavenging capability of M@EFE. In addition, the antioxidant activity was further examined using the 2, 2′-azino-bis(3-ethylbenzothiazoline-6-sulfonic acid) (ABTS) assay. The ABTS radical displays characteristic absorption at 405 nm, and its absorbance decreases in the presence of antioxidants [43]. UV spectrum revealed a marked reduction in ABTS absorption after treatment with M@EFE (Fig. 1g), corroborating its effective ROS scavenging performance. Consistently, the 3, 3′, 5, 5′-tetramethylbenzidine (TMB) assay also yielded similar results (Fig. S3), further supporting the robust antioxidant activity of M@EFE. Furthermore, the ROS elimination ability of M@EFE NPs was further confirmed by electron spin resonance (ESR) spectra, evidenced by the decrease of •O₂⁻ and •OH signals (Fig. 1h). Additionally, driven by the inherent CAT-like nanozyme activity, M@EFE NPs effectively triggered the catalytic decomposition of H₂O₂. As continuously monitored by a dissolved oxygen meter, the rapid generation of O₂ not only verified the clearance of pathological ROS but also indicated its potential to ameliorate local hypoxia (Fig. S4).

3.2. In vitro BBB penetration and cellular uptake

To evaluate the therapeutic potential of the nanomedicine, it is essential to determine its ability to cross the BBB and accumulate in the ischemic region. Prior to investigating the BBB penetration of M@EFE NPs in vitro, we first assessed the cytotoxicity of the nanoparticles using the CCK-8 assay. Results showed that M@EFE NPs had negligible impact on the viability of HT22 cells, indicating good biocompatibility (Fig. S5). Subsequently, the BBB penetration capacity of M@EFE NPs was evaluated using an in vitro BBB model based on a bEnd.3 monolayer and a co-culture system. Compared to EFE NPs, the membrane-coated nanoparticles exhibited significantly higher cumulative permeation (Fig. 2a), suggesting their enhanced potential to traverse the BBB.

Fig. 2.

In vitro BBB penetration and cellular uptake. (a) BBB penetration of different nanoparticles in Transwell-mediated BBB models (n = 3). (b) Fluorescence images of HT22 cells incubated with EFE or M@EFE NPs at different time points. Scale bar, 100 μm. (c) Flow cytometry analysis and quantitation of (d) cellular uptake of nanoparticles by HT22 cells at different time points. (e) Fluorescence images of M@EFE NPs colocalization with lysosomes in HT22 cells at different time points. Scale bar, 20 μm. (f) Quantitative analysis of the colocalization coefficients of M@EFE NPs and lysosomes. Data are shown as the mean ± SD. Significance between each group was calculated using two-way ANOVA with Tukey's post hoc test (a, d) and one-way ANOVA with Tukey post hoc test (f). ∗P < 0.05, ∗∗∗P < 0.001, ∗∗∗∗P < 0.0001, ns, no significance.

Besides, the cellular uptake of nanoparticles was studied with confocal laser scanning microscope (CLSM) and flow cytometer. As shown in Fig. 2b, the nanoparticles were gradually internalized by HT22 cells after 2 h of incubation. As time prolonged, more nanoparticles can enter into the cells. Notably, more M@EFE NPs were taken up by the cells compared to EFE NPs, suggesting that the macrophage membrane coating enhances cellular internalization (Fig. 2c and d). This enhanced uptake can be attributed to two mechanisms: On the one hand, macrophage membranes are intrinsically enriched with endocytic receptors, including scavenger receptors and integrins, that facilitate active, receptor-mediated endocytosis. On the other hand, the specific lipid composition of the macrophage membrane promotes better membrane affinity and fusogenicity with target cell membranes, further enhancing internalization. Collectively, these macrophage membrane-derived properties confer a significant advantage for cellular internalization compared to uncoated synthetic nanoparticles.

To further confirm which cell types the macrophage membrane-coated NPs target, cellular uptake experiments were performed on two other cells, including bEnd.3 cells (endothelial cells) and BV2 cells (microglia). Consistent with the observations in HT22 cells, both BV2 and bEnd.3 cells exhibited a significantly higher uptake of M@EFE NPs compared to bare EFE NPs (Fig. S6–S7). The quantitative fluorescence data further corroborated these results, confirming that the macrophage membrane biomimetic strategy effectively enhances nanoparticle internalization across various brain-related cell lines.

Furthermore, CLSM was employed to determine the lysosomal escape capability of the nanoparticles. Fluorescence images revealed clear colocalization of M@EFE NPs with lysosomes at 2 h and 4 h post-incubation. By 8 h, however, M@EFE NPs were predominantly observed outside the lysosomal compartments, demonstrating successful lysosomal escape (Fig. 2e and f).

3.3. In vitro antioxidant activity

Subsequently, to evaluate the in vitro antioxidant activity of the nanomedicine, an oxygen-glucose deprivation/reperfusion (OGD/R) model was established in HT22 cells to simulate IRI (Fig. 3a). Intracellular ROS levels were detected using the fluorescent probe 2′, 7′-dichlorofluorescein diacetates (DCFH-DA). DCFH-DA is hydrolyzed by cellular esterases to DCFH, which is then oxidized by ROS into highly fluorescent DCF [44]. As shown in Fig. 3b, PBS-treated cells exhibited intense fluorescence, indicating substantial ROS generation. In contrast, treatment with free EDV, EFE NPs, or M@EFE NPs all markedly reduced ROS-associated fluorescence to comparable levels, with no statistically significant differences observed among these three treatment groups. This finding suggested that the edaravone payload and the EGCG-Fe³⁺ network inherently possess potent ROS-scavenging capacity, and that macrophage membrane coating does not impede this intrinsic antioxidant activity in vitro. Notably, the equivalent performance of EDV, EFE, and M@EFE in this cell-based assay is expected, as the membrane coating primarily enhances in vivo pharmacokinetics and targeting rather than direct radical-scavenging chemistry. The free drug (EDV) and uncoated nanoparticles (EFE) can freely access cells in the culture system without biological barriers, allowing their antioxidant components to fully exert their effects. Besides, flow cytometry analysis was also performed to determine the ROS scavenger ability. As shown in Fig. S8, M@EFE NPs significantly reduced the fluorescence of ROS, indicating superior antioxidant capacity.

Fig. 3.

In vitro antioxidant activity. (a) Schematic diagram of the establish of OGD/R model. (b) Fluorescence images of intracellular ROS levels in normal HT22 cells or in HT22 cells undergoing OGD/R after different treatments. Scale bar: 200 μm. of Confocal microscopy observation of ROS levels in different treatment groups after OGD/R or normal HT22 cells. Scale bar: 200 μm. (c) Fluorescence images and quantitative analysis (d) of lipid peroxidation in HT22 cells undergoing OGD/R after different treatments for 4 h. Scale bar, 200 μm. (e) Fluorescence images and quantitative analysis (f) of Δψm in HT22 cells undergoing OGD/R after different treatments for 4 h. Scale bar, 100 μm. JC-1 red/green ratios were normalized to the control group. (g) Apoptosis of injured HT22 cells via Annexin V-FITC/PI double-staining assay. (h) CLSM images of Calcein-AM (live)/PI (dead) double-staining of HT22 cells suffered from OGD/R after different treatments for 4 h. Scale bar, 100 μm. Data are shown as the mean ± SD. Significance between each group was calculated using ANOVA with Tukey post hoc test. ∗P < 0.05, ∗∗∗P < 0.001, ∗∗∗∗P < 0.0001. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

In IRI, excessive ROS can induce lipid peroxidation of cell membranes [45]. Therefore, we investigated whether M@EFE could also suppress this process. Under normal condition, cells appeared in the reduction state. In OGD/R model, cells in reduction state significantly decreased, with a corresponding shift toward the oxidized state (Fig. 3c and d). Treatment with M@EFE NPs significantly reversed this trend, decreasing the ratio of cells in oxidation state. Besides, lipid peroxidation can directly impair mitochondrial membrane integrity and function [45]. JC-1, a fluorescent probe sensitive to mitochondrial membrane potential (Δψm), was used to assess mitochondrial health. Under physiological conditions, cells displayed a balance of JC-1 aggregates (indicative of normal Δψm) and monomers. In the OGD/R model, PBS-treated cells showed a pronounced decrease in JC-1 aggregates, reflecting loss of membrane potential. In contrast, cells in M@EFE-treated group maintained significantly higher levels of JC-1 aggregates, suggesting preservation of mitochondrial function (Fig. 3e and f). Flow cytometry analysis also showed that M@EFE NPs effectively reduced the proportion of JC-1 monomers, indicating the restoration of mitochondrial function (Fig. S9). These results indicated that M@EFE NPs not only lowered intracellular ROS but also effectively inhibits lipid peroxidation.

Given that mitochondrial dysfunction often initiates apoptosis, we further evaluated cell death using annexin V-FITC and propidium iodide (PI) staining. Flow cytometry analysis showed that M@EFE treatment markedly reduced the proportion of apoptotic cells compared to the control (Fig. 3g), indicating suppression of ischemia-induced apoptosis and improved cell survival. This anti-apoptotic effect can be attributed to multiple interconnected mechanisms. First, M@EFE NPs preserved mitochondrial membrane potential (ΔΨm), preventing the release of cytochrome c and subsequent activation of the intrinsic apoptotic pathway. Second, the potent ROS-scavenging capacity of M@EFE NPs removes oxidative stress signals that would otherwise trigger mitochondrial permeability transition and apoptotic signaling. Third, inhibition of lipid peroxidation prevents the generation of pro-apoptotic aldehydes such as 4-HNE, which can modify and inactivate anti-apoptotic proteins. Collectively, these mechanisms converge to maintain mitochondrial integrity and suppress the apoptotic cascade at multiple levels. Consistent with this, live/dead staining also revealed a significant decrease in dead cells after M@EFE treatment (Fig. 3h), demonstrating the neuroprotective effect.

In summary, at the cellular level, M@EFE demonstrates the ability to scavenge ROS, inhibit lipid peroxidation, maintain mitochondrial membrane potential, and thereby attenuate apoptosis and promote cell survival.

3.4. In vivo therapeutic effect

To further evaluate the neuroprotective efficacy of the nanomedicine under pathological conditions, MCAO models were established in C57BL/6 mice. First, nanoparticle distribution was analyzed to validate brain delivery capability. One hour after modeling, mice were intravenously administered free DID, DID-labeled EFE NPs, or M@EFE NPs via the tail vein. Twenty-four hours later, the mice were sacrificed, and brain tissues were collected. Compared to the EFE NPs and free probe groups, the M@EFE group exhibited significantly stronger fluorescence signals in the brain, particularly in the ischemic hemisphere (Fig. 4a and b and Fig. S10). Quantitative analysis further indicated that the fluorescence intensity in M@EFE NPs group was 1.76-fold higher than that of the free probe group. Moreover, fluorescence imaging revealed co-localization of M@EFE NPs with CD31, demonstrating that the nanomedicine can cross the BBB, accumulate in periventricular regions, and partially penetrate the vascular endothelium (Fig. S11). These findings confirmed that membrane coating significantly enhances the brain-targeting ability of the nanomedicine, establishing a foundation for subsequent neuroprotective effects.

Fig. 4.

In vivo therapeutic effect. (a) IVIS images of the brains in MCAO/R mice after intravenous injection with different nanoparticles or free probe for 24 h. (b) Fluorescence images of brain sections at 24 h post-injection (n = 6). (c) Illustration of the treatment regimen in MCAO/R mice. (d) Photographs and (e) quantitative analysis of TTC staining of the brains in MCAO/R mice after different treatments (n = 5). (f, g) Percentage change in body weight of mice after treatment in different groups (n = 6). (h) 14-day survival period for mice in different treatment groups (n = 8). Data are shown as the mean ± SD. Significance between each group was calculated using ANOVA with Tukey post hoc test. ∗∗P < 0.01, ∗∗∗∗P < 0.0001. Statistical significance for survival analysis was determined by Log-rank test with Bonferroni correction. ∗P < 0.0167, ∗∗P < 0.01, ∗∗∗P < 0.001 compared with the M@EFE group.

Subsequently, TTC staining was performed to evaluate the acute protective effect of the nanoparticles in cerebral ischemic injury (Fig. 4c). In the control group, extensive infarct areas were observed in the ischemic cerebral cortex and striatum, whereas treatment with free drug, EFE, and M@EFE alleviated ischemic damage (Fig. 4d). As demonstrated by the quantitative analysis, the M@EFE treatment yielded the lowest infarct volume ratio compared to the free drug and bare EFE groups (Fig. 4e), indicating that M@EFE effectively mitigates brain tissue damage induced by IRI.

We further investigated the long-term impact of nanoparticle treatment on animal outcomes. In terms of body weight changes, mice in the control group exhibited significant postoperative weight loss with limited recovery. In contrast, mice treated with M@EFE showed less weight reduction, with weight beginning to recover by postoperative day 4 and approaching normal levels by day 11 (Fig. 4f, g and Fig. S12). The results revealed a significantly higher survival rate in the M@EFE group compared to the EFE and free drug groups, suggesting an improved overall physiological condition and prolonged survival (Fig. 4h). These findings validated the protection effect of M@EFE against IRI, positioning it as a promising nanotherapeutic strategy for ischemic stroke treatment.

3.5. Evaluation of neuroprotective mechanisms

To verify the in vivo antioxidant effect of M@EFE NPs, ROS levels in the ischemic brain of mice were evaluated with the DHE fluorescent probe. Compared to the control, a significant increase in red fluorescence was observed in PBS-treated group. Notably, M@EFE NPs treatment effectively quenched the red fluorescence, exhibiting stronger ROS-scavenging capability than both free EDV and bare EFE NPs (Fig. S13a). Quantitative analysis confirmed that M@EFE NPs yielded the lowest ROS accumulation in the damaged hemisphere (Fig. S13b). This robust ROS scavenger performance establishes a crucial foundation for its subsequent anti-lipid peroxidation and anti-apoptotic functions.

Subsequently, we performed comprehensive histological analysis of brain tissues from MCAO mice to elucidate the neuroprotective mechanisms of M@EFE NPs against IRI. As is well-known, 4-hydroxynonenal (4-HNE) is a highly reactive and toxic aldehyde generated during lipid peroxidation, which is a well-established biomarker of oxidative stress and plays a causal role in mediating oxidative damage. Elevated 4-HNE levels are characteristic of ischemia-reperfusion injury and contribute to neuronal cell death. First, oxidative stress levels were evaluated by 4-HNE staining. As shown in Fig. 5a, sham-operated mice showed negligible 4-HNE expression, while MCAO mice exhibited widespread 4-HNE distribution, indicating severe oxidative damage. In contrast, M@EFE NP treatment markedly attenuated 4-HNE expression (Fig. 5b), indicating effective suppression of lipid peroxidation in the ischemic brain.

Fig. 5.

Neuroprotective mechanisms. (a) Immunohistochemical staining images and (b) quantitative analysis of 4-HNE in ischemic regions after different treatments (n = 5). Scale bar, 100 μm. (c) MDA contents in ischemic regions after different treatments (n = 3). (d) The expression of GPX4 and ACSL4 in ischemic regions after different treatments. Quantitative analysis of expression of (e) GPX4 and (f) ACSL4 (n = 3). (g) GPX-4/NeuN double staining images in ischemic regions after different treatments. Scale bar, 200 μm. Quantitative analysis of (h) NeuN and (i) GPX4 positive cells (n = 5). (j) IBA-1/GFAP double staining images in ischemic regions after different treatments. Scale bar, 50 μm. Quantitative analysis of (k) IBA-1 and (l) GFAP fluorescence intensity (n = 5). Data are shown as the mean ± SD. Significance between each group was calculated using ANOVA with Tukey post hoc test. ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001, ∗∗∗∗P < 0.0001, ns, no significance.

Excessive intracellular iron accumulation and fatal lipid peroxidation are two major hallmarks of ferroptosis. To elucidate whether M@EFE NPs could confer neuroprotection by inhibiting ferroptosis, the levels of iron ions was measured in the ischemic brain tissue of mice. An increase of iron content was detected in PBS-treated group compared to the control, which was profoundly reversed by M@EFE NPs treatment (Fig. S14). Meanwhile, the level of malondialdehyde (MDA), an essential indicator of ferroptosis lipid damage, was significantly reduced in the M@EFE group (Fig. 5c). Combined with the previously observed ROS-scavenging and 4-HNE-reducing capacities, these findings strongly suggest that M@EFE NPs efficiently alleviate neuronal damage by suppressing iron-dependent lipid peroxidation and early ferroptosis. The anti-ferroptotic capability was further validated by western blot analysis. Following cerebral ischemia, the depletion of GPX4 (an anti-ferroptotic enzyme) and overexpression of the ACSL4 (a pro-ferroptotic enzyme) were observed in the model group. Strikingly, M@EFE NPs treatment successfully reversed these pathological trends. It robustly restored the expression of GPX4 while simultaneously downregulating ACSL4 levels, exhibiting the most prominent regulatory effect among all experimental groups (Fig. 5d–f). These molecular findings, perfectly correlating with the biochemical clearance of iron and MDA, provide definitive evidence that M@EFE NPs block the ferroptotic cascade to preserve neuronal viability.

Further analysis with GPX4/NeuN double staining showed that M@EFE NPs significantly enhanced neuronal survival and restored GPX4 expression (Fig. 5g). Quantitative analysis revealed a 3.0-fold increase in NeuN positive cells and a 16.2-fold increase in GPX4 positive cells in M@EFE-treated mice compared to the controls (Fig. 5h and i). Notably, M@EFE exhibited superior efficacy in preventing neuronal apoptosis in the ischemic area compared to both free drug and EFE treatments, highlighting the most effective neuroprotective effect.

To evaluate cellular damage, we performed H&E, TdT-mediated dUTP Nick-End Labeling (TUNEL), and Nissl staining on the brain tissue of mice. H&E staining revealed a substantial number of necrotic cells in the ischemic hemisphere of MCAO mice, which were markedly reduced following treatment with M@EFE NPs (Fig. S15). Besides, both TUNEL and Nissl staining demonstrated that M@EFE NPs significantly attenuated apoptosis and neuronal injury (Fig. S16–S17). To further elucidate the underlying molecular mechanism of this anti-apoptotic effect, the expression of key apoptosis-regulatory proteins was evaluated with western blotting. Consistent with the histological observations, M@EFE NPs treatment robustly rescued the expression of the anti-apoptotic protein BCL-2 and simultaneously suppressed the pro-apoptotic protein BAX, resulting in a markedly elevated BCL-2/BAX ratio compared to the control group (Fig. S18). Taken together, these findings indicate that M@EFE NPs exert robust neuroprotective effects, thereby effectively mitigating reperfusion injury after ischemic stroke.

Furthermore, immunohistochemical staining of microglial marker IBA-1 and astrocytic marker GFAP showed that M@EFE NPs most effectively suppressed the activation of both microglia and astrocytes among all treatment groups (Fig. 5j–l), indicating potent inhibition of post-ischemic neuroinflammation. To quantitatively evaluate the inflammatory milieu, the levels of IL-1β, IL-6, TNF-α, and IL-10 were measured by ELISA. Consistent with the immunostaining results of GFAP and IBA-1, M@EFE NPs treatment most effectively suppressed the secretion of pro-inflammatory cytokines (IL-1β, IL-6, and TNF-α) while simultaneously promoting the expression of the anti-inflammatory factor IL-10 (Fig. S19). This optimized cytokine balance further confirms the potent immunomodulatory capacity of M@EFE NPs in resolving neuroinflammation and promoting tissue recovery.

Collectively, these findings established that M@EFE NPs exert robust neuroprotection through multi-faceted mechanisms: scavenging ROS, inhibiting lipid peroxidation, preserving neuronal integrity, reducing apoptosis, and attenuating neuroinflammation, thereby effectively mitigating ischemic reperfusion injury.

3.6. Behavioral test

Building upon the established evidence that the nanomedicines can mitigate tissue damage, suppress oxidative stress, and inhibit inflammatory responses, we further evaluated their impact on neurological recovery in ischemic mice. In the hanging test, control mice exhibited significantly reduced grip strength, whereas the nanomedicine-treated group demonstrated longer grip durations, indicating markedly improved motor function (Fig. 6a and b).

Fig. 6.

Behavioral test. (a) Schematic diagram of the hanging test of mice. (b) Suspension time of mice after treatment in different groups (n = 8). (c) Time spent by mice on the rotarod fatigue test after treatment in different groups. (d) Representative images of the movement paths of mice during the open field test after different treatments. (e) Total distance in the open field after different treatments (n = 8). (f) Distance in the center of the open field after different treatments (n = 8). (g) Longa scores after different treatments (n = 8). Data are shown as the mean ± SD. Significance between each group was calculated using ANOVA with Tukey post hoc test. ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001, ∗∗∗∗P < 0.0001, ns, no significance.

In the rotarod fatigue test, control mice exhibited significantly reduced dwell time with impaired balance and coordination, whereas mice receiving the nanomedicine exhibited significantly prolonged dwell times (Fig. 6c). Furthermore, in the open field test, control mice showed decreased spontaneous activity and shorter locomotor trajectories, while the nanomedicine-treated group exhibited significant recovery in central locomotor distance, total locomotor distance, and exploratory behavior (Fig. 6d–f). Simultaneously, Longa neurological function scores were assessed in mice from different groups. Compared with the control group, the M@EFE group exhibited significantly lower scores, demonstrating favorable neurological recovery (Fig. 6g).

Behavioral tests demonstrated that nanomedicines can not only improve structural and molecular-level damage in brain tissue but also significantly promote neural functional recovery, further validating their potential therapeutic value in the treatment of ischemic stroke.

In summary, in vivo studies confirmed that the nanomedicine effectively accumulated in the ischemic hemisphere, significantly reduced acute infarct volume, and promoted weight recovery and survival rates in long-term observations. Histological and immunofluorescence analyses demonstrated that the nanomedicine attenuates lipid peroxidation, restores GPX4 expression, diminishes neuronal apoptosis, and suppresses the overactivation of microglia and astrocytes. Behavioral assessments further revealed substantial improvements in motor function, coordination, and balance. These in vivo findings aligned well with in vitro results, collectively supporting the neuroprotective mechanism of the nanomedicine system: mediated through alleviation of oxidative stress, attenuation of inflammatory responses, and enhancement of neuronal survival.

3.7. In vivo biosafety evaluation

Following the demonstration of the outstanding neuroprotective efficacy and underlying mechanisms of M@EFE NPs against ischemic stroke, evaluating their systemic biosafety is an essential prerequisite for future clinical translation. Firstly, the blood compatibility of the nanoparticles was assessed via a hemolysis assay. As shown in Fig. S20, both EFE and M@EFE NPs exhibited negligible hemolysis rates across various concentrations (50, 100, and 200 μg/mL), which were well below the widely accepted safe threshold of 5%, indicating excellent hemocompatibility for intravenous administration.

Furthermore, histological analysis and serum biochemistry were conducted to evaluate the potential systemic toxicity in vivo. Hematoxylin and eosin (H&E) staining of major organs (heart, liver, spleen, lung, and kidney) revealed that, compared to the Sham and PBS-treated groups, no noticeable pathological abnormalities, inflammatory lesions, or tissue necrosis were observed in the EDV, EFE, or M@EFE treatment groups (Fig. S21). Consistently, serum biochemical parameters related to liver function ALT and AST and kidney function UREA and CREA in all nanoparticle-treated groups remained strictly within the normal reference ranges, showing no significant differences from the Sham group (Fig. S22). These results collectively demonstrate the highly favorable short-term biocompatibility and negligible systemic toxicity of M@EFE NPs.

Although our preliminary in vivo evaluations demonstrated the excellent short-term biosafety of M@EFE NPs, assessing the potential long-term safety of iron-containing nanomedicines is crucial for their rigorous clinical translation. Theoretically, as an essential endogenous trace element, the Fe³⁺ released from the gradual degradation of the metal-polyphenol network can be metabolized and cleared through normal physiological pathways (e.g., ferritin storage and transferrin transport). However, the potential risks of iron overload or chronic toxicity resulting from cumulative dosing or delayed clearance cannot be entirely ruled out at this stage. Therefore, comprehensive long-term pharmacokinetic, biodistribution, and chronic toxicity studies in large animal models will be systematically investigated in our future work to fully guarantee their translational feasibility.

4. Conclusions

In this study, we developed a biomimetic nanomedicine based on an EGCG/Fe³⁺ complex, encapsulated with Edaravone and coated with macrophage membrane. Systematic evaluation in an ischemic stroke model confirmed its multifunctional efficacy. The nanomedicine exhibited excellent cellular uptake, lysosomal escape, and antioxidant capacity in vitro, along with the ability to efficiently cross the BBB. In vivo, it demonstrated precise brain targeting, significantly reduced acute cerebral infarction volume, and promoted long-term functional recovery and survival. Mechanistic investigations revealed that its neuroprotective effects are primarily mediated by scavenging ROS, attenuating lipid peroxidation, preserving mitochondrial integrity, and mitigating neuroinflammation. Collectively, this biomimetic nanoplatform represents a promising therapeutic strategy for ischemic stroke.

CRediT authorship contribution statement

Bingyuan Liu: Conceptualization, Data curation, Formal analysis, Methodology, Writing – original draft. Xinyi Chen: Data curation, Investigation, Methodology. Luyun Tu: Data curation, Investigation. Lixin Yin: Formal analysis, Validation. Hongxia Huang: Investigation, Validation. Ye Wu: Investigation, Validation. Hui Cai: Funding acquisition, Supervision, Writing – review & editing. Yiliang Li: Conceptualization, Funding acquisition, Supervision. Sheng Hong: Conceptualization, Funding acquisition, Methodology, Writing – review & editing.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary data

The following is/are the supplementary data to this article:

Data availability

Data will be made available on request.