Photoacoustic imaging-guided, synchronously targeted thrombolysis and neuroprotection of acute ischemic stroke

Summary

Acute ischemic stroke (AIS) is a major cause of disability and mortality worldwide, typically resulting from a thrombus that blocks cerebral blood flow. Current treatments focus on thrombolysis and neuroprotection, with recombinant tissue plasminogen activator (rt-PA) and edaravone (Eda) as key drugs. However, their therapeutic efficacy is constrained by poor targeting, leading to suboptimal drug delivery to the brain. Herein, we develop a biomimetic delivery system (PLEA, platelet membrane-liposome hybrid delivery system encapsulating Eda and rt-PA) by hybridizing platelet membranes with liposomes, leveraging natural thrombus-targeting ability of the platelet membrane to enhance drug delivery efficiency. PLEA effectively targeted thrombus sites, synchronously enhancing thrombolysis and providing neuroprotection in a mouse model of AIS. Photoacoustic (PA) imaging was used to monitor clot formation and assess changes in intracranial blood flow and oxygen levels. These findings indicate that PA imaging-guided PLEA treatment represents a promising approach for precise stroke treatment.

Graphical abstract

Highlights

• •PLEA co-delivers rt-PA and edaravone, targeting thrombi for enhanced drug accumulation
• •PLEA achieves synchronous thrombolysis and neuroprotection
• •Photoacoustic imaging enables non-invasive monitoring of cerebral blood flow
• •PLEA shows good biocompatibility with promising translational potential

Natural sciences; biological sciences; physiology

Introduction

Acute ischemic stroke (AIS) is a neurological emergency caused by a thrombus blocking cerebral blood flow, leading to oxygen and nutrient deprivation, neuronal injury, and a cascade of inflammatory responses that exacerbate brain damage and impair neurological function,^1,2,3 making AIS one of the leading causes of morbidity and mortality worldwide.^4,5 The clinical treatment of AIS mainly includes early reperfusion therapies with recombinant tissue plasminogen activator (rt-PA) and neuroprotective strategies with edaravone (Eda), which offers a promising synchronous approach to AIS management.^6,7,8 However, clinical practice has shown that the poor targeting of rt-PA and Eda in vivo often leads to insufficient accumulation at the lesion site.^9,10,11 This not only reduces thrombolytic efficiency but also increases the risk of bleeding in non-targeted areas and potential brain damage.^1,12,13 Additionally, their short circulation half-life of less than 5 min leads to rapid clearance, often necessitating higher doses that further elevate the risk of hemorrhagic complications.^14,15,16 Therefore, combining rt-PA and Eda, and enhancing their targeted effect could be a promising strategy for AIS treatment.

In recent years, biomimetic drug delivery systems (BDDSs) that utilize natural biological materials, such as cell membranes or proteins, to enhance the efficiency and precision of drug delivery have gained considerable attention due to their excellent biocompatibility and minimal immunogenicity.^17,18,19 For example, platelet (PLT) membranes inherently possess thrombus-targeting capabilities, probably due to their adhesion proteins and glycoprotein integrins, such as CD42, CD61, and GPIIb/IIIa.^17,20,21 These molecules enable interactions with damaged vascular endothelial cells and fibrin.²² Xu et al. prepared a BDDS that used PLT membranes to deliver rt-PA, which demonstrated targeted delivery, prolonged circulation time, and reduced cerebral damage compared with rt-PA alone.²¹ As a result, PLT membrane-decorated nanoparticles show great promise for delivering thrombolytic drugs. However, since platelets are derived from animal blood, producing sufficient quantities requires large amounts of blood, raising ethical concerns and limiting scalability. Liposomes, which are well established in nano drug delivery systems, have proven successful in clinical applications due to their spherical structure, uniformity, and ease of modification.^23,24 Hybridizing liposomal nanocarriers with PLT membranes could address these issues by reducing the need for large quantities of platelets, while preserving the intrinsic biological functions of platelet membranes and maintaining the uniform characteristics of liposomes. Given these advantages, hybrid platelet membrane-liposomes hold great promise for the treatment of AIS.

Photoacoustic (PA) imaging is an emerging non-invasive imaging technology that combines the advantages of optical and ultrasound imaging, offering high spatial and temporal resolution.^25,26 This technique involves exposing tissues to short laser pulses, where absorbed light energy induces thermoelastic expansion, generating acoustic waves.^27,28 These signals are then detected by ultrasound transducers and reconstructed into high-resolution images.^29,30 PA imaging is particularly sensitive to blood flow because hemoglobin, the primary oxygen-carrying molecule in blood, strongly absorbs laser light at specific wavelengths.^31,32 By leveraging this property, PA imaging can provide real time mapping of blood flow and oxygenation levels.^33,34 This sensitivity to vascular changes enables PA imaging to detect alterations in cerebral blood supply and oxygen saturation before significant tissue damage occurs, making it a valuable tool for early stroke diagnosis and timely intervention, potentially improving patient outcomes.

In this study, we developed a biomimetic delivery system: platelet membrane-liposome hybrid delivery system encapsulating Eda and rt-PA (PLEA). Specifically, Eda was loaded into the lipid bilayer of the hybrid, while rt-PA was encapsulated in the liposomal core (Figure 1). PLEA exhibited excellent biocompatibility and stability, with in vitro studies demonstrating its potent thrombolytic efficiency and antioxidant properties. In vivo fluorescence imaging confirmed its targeted accumulation at thrombus sites, supporting its effective targeted delivery. Additionally, PLEA demonstrated promising therapeutic effects in a mouse model of ischemic stroke, as evidenced by histological and pathological staining. To further investigate its efficacy, we employed PA imaging to assess the extent of ischemia in stroke models and to monitor changes in intracranial blood flow and blood oxygen levels following PLEA treatment. This allowed us to closely track the dynamics of thrombosis and the therapeutic impact of PLEA in real time. These findings suggest that PLEA holds significant potential as a BDDS for effectively targeting thrombotic events and treating acute stroke. Moreover, PA imaging presents itself as a promising and invaluable tool for monitoring brain thrombosis formation and evaluating treatment progress.

Figure 1.

Schematic of the synthesis and therapeutic mechanism of PLEA nanoparticles

(A) Preparation of PLEA nanoparticles. Lipid components (1,2-Dipalmitoyl-sn-glycero-3-phosphocholine [DPPC]), 1,2-Dipalmitoyl-sn-glycero-3-phosphate [DPPA], and 1,2-Distearoyl-sn-glycero-3-phosphoethanolamine-polyethylene glycol [DSPE-PEG]) and Eda were used to form a thin film, followed by rehydration with rt-PA to generate drug-loaded liposomes. Platelet membranes were extracted via freeze-thaw and sonication and subsequently coated onto liposomes through extrusion to obtain platelet membrane-mimetic dual-drug nanoparticles, PLEA.

(B) Illustration of the therapeutic mechanism of PLEA in a photothrombotic stroke model. After tail vein injection, PLEA nanoparticles target the thrombus site, promoting thrombolysis and crossing the disrupted BBB. The released Eda scavenges ROS, alleviating oxidative stress and protecting neurons. Photoacoustic imaging was used to assess the therapeutic effect of PLEA. This figure was generated using Figdraw.

PLEA, platelet membrane-liposome hybrid delivery system encapsulating Eda and rt-PA; BBB, blood-brain barrier; ROS, reactive oxygen species.

Results

Preparation and characterization of PLEA

To construct PLEA, LEA were firstly synthesized using thin-film hydration, followed by hybridization with platelet membranes via co-extrusion. As displayed in TEM images (Figure 2A), PLEA exhibited a uniform, monodispersed spherical shape, with an average diameter ranging from 180 to 200 nm. The absorbance spectra of PLEA demonstrated a characteristic peak corresponding to Eda, confirming its successful encapsulation within PLEA (Figure 2B). The loading efficiency and encapsulation efficiency of Eda in PLEA were calculated to be approximately 22.31% and 91.75%, respectively, while those for rt-PA were approximately 10.98% and 98.21%, respectively. To demonstrate the release property at the thrombus site, PLEA was incubated with PBS, PBS + bovine serum albumin (BSA), and PBS + BSA + activated platelets (aPLTs). As shown in Figures S1A and S1B, passive release of both Eda and rt-PA in PBS was minimal, confirming the stability of the PLEA formulation. The addition of BSA accelerated Eda release, likely because the hydrophobic pockets of BSA facilitate the solubilization of the drug. In contrast, rt-PA release was slightly reduced in the presence of BSA. It is presumed that the potential formation of a protein corona on PLEA might somewhat impede the diffusion of the encapsulated rt-PA, especially in protein-rich environments. Upon aPLT stimulation, both Eda and rt-PA were rapidly released. Notably, rt-PA release increased 3.8-fold within 2 h compared with the serum-only condition, suggesting a rapid therapeutic response at the site of ischemic stroke. The average diameter of PLEA was measured to be 203.4 ± 5.7 nm, slightly larger than that of LEA (177.1 ± 5.0 nm) (Figure 2C). Similarly, the zeta potential of PLEA was −13.37 ± 0.153 mV, compared to −10.48 ± 0.737 mV for LEA (Figure 2D). These minor differences might be attributed to the platelet membrane modification. To assess the stability of PLEA, the particle size was monitored in both 1× PBS and 0.5× FBS over 120 h (Figure 2E). The diameter of PLEA remained consistent for up to 120 h, indicating good physiological stability. To verify the successful fusion of the LEA with platelet membrane, LEA and platelet membranes were separately labeled with Dio and Dil dye, respectively. After the synthesis of PLEA, the nanoparticles were placed on a confocal dish coated with poly-L-lysine for 40 min. Fluorescence images of PLEA revealed the presence of both Dio and Dil signals, indicating the successful hybridization of platelet membranes with LEA (Figure 2F). To better characterize the fusion of liposomes and platelet membranes quantitatively, the fluorescence resonance energy transfer (FRET) assay was conducted before and after extrusion with platelet membranes. As illustrated in Figure S2A, LEA was labeled with the donor fluorophore DiO, which exhibits a fluorescence emission maximum at 508 nm. The platelet membrane was concurrently labeled with the acceptor fluorophore DiD, whose emission peak lies at 690 nm. Upon successful coating, energy transfer from DiO to DiD occurs, leading to a pronounced increase in the DiD emission signal (Figure S2A), thereby confirming the incorporation of platelet membranes onto the liposomal surface. Moreover, by varying the liposome-to-membrane mass ratio from 1: 0.01 to 1: 0.10, we observed a progressive increase in FRET efficiency (Figure S2B), indicating that a 1:0.10 mass ratio provides an optimal fusion between liposomes and platelet membranes. Further analysis of the protein composition in PLEA using SDS-PAGE demonstrated identical protein profiles to those of platelet membranes and rt-PA (Figure 2G), suggesting the retention of native platelet membrane proteins within PLEA, which could enhance thrombus-targeting capabilities. Western blot analysis confirmed the presence of key platelet surface markers, including CD41, CD62p, integrin β3, and glycoprotein Ⅵ (GPⅥ) (Figure 2H). To evaluate the thrombus-targeting ability of PLEA, artificial blood clots were prepared by incubating fresh blood with thrombin at 37°C for 2 h, followed by overnight storage at 4°C. To avoid fluorescence dilution caused by thrombolysis, rt-PA was not loaded during these experiments. The clots were treated with saline, indocyanine green (ICG)-labeled LE (liposomes containing Eda), or PLE (hybrid liposomes containing Eda) and incubated for 6 h. Notably, the clots incubated with PLE exhibited significantly higher fluorescence signals than those treated with LE (Figures 2I and 2J). This indicated the enhanced targeting ability of PLEA toward the thrombus.

Figure 2.

Characterization of PLEA nanoparticles

(A) TEM image of PLEA nanoparticles. Scale bars, 200 nm.

(B) UV-Vis absorption spectra of PLEA, PM, rt-PA, liposomes (Lipo), and Eda.

(C) Size distribution of LEA and PLEA, measured by DLS in deionized water.

(D) Zeta potential of LEA and PLEA (n = 3, error bars represent standard deviation [SD], mean ± SD).

(E) Stability of PLEA nanoparticles in PBS and 0.5× FBS over 5 days (n = 3, error bars represent SD, mean ± SD).

(F) Confocal fluorescence microscopy images of PLEA. Liposomes were labeled with Dio (green), and PMs were labeled with Dil (red). Scale bars, 20 μm.

(G) Protein profiles of PLT, PM, rt-PA, and PLEA analyzed by SDS-PAGE with Coomassie blue staining.

(H) Western blot analysis of PM proteins (CD41, CD62p, GPⅥ, and integrin β3) in PLT, PMs, and PLEA.

(I) Fluorescence images of artificial blood clots after incubation with PBS, LE, or PLE. Nanoparticles were labeled with indocyanine green (red signal).

(J) Quantitative analysis of fluorescence intensity of artificial thrombus sites at different time points. n = 3, error bars represent SD, mean ± SD; ∗∗∗p < 0.001.

TEM, transmission electron microscopy; PM, platelet membrane; Eda, edaravone; DLS, dynamic light scattering; PLT, platelet; GPⅥ, glycoprotein Ⅵ.

Antioxidant abilities of PLEA

To evaluate the in vitro antioxidant potential of PLEA, two commonly used free radical scavenging assays, ABTS and DPPH, were employed. As the concentrations of Eda and PLEA increased, a gradual enhancement in the free radical scavenging activity was observed (Figures 3A and 3B). Notably, no significant differences were found between the scavenging rates of PLEA and Eda, indicating that the PLEA successfully retained the intrinsic antioxidant properties of Eda. To further verify whether the free radical scavenging capability of PLEA can be effectively exerted in cellular environments, an in vitro reactive oxygen species (ROS)-mediated neuron cell (HT22) damage model was established. The intracellular ROS scavenging activity of PLEA was assessed using the fluorescent probe 2′,7′-dichlorodihydrofluorescein-diacetate (DCFH-DA). Tert-butyl hydroperoxide (TPBH), a widely used ROS generator, significantly enhanced the ROS signal in HT22 cells, as evidenced by the appearance of strong green fluorescence. In contrast, no green fluorescence was observed in the untreated control group (Figures 3C and 3D). Cells treated with either Eda + TPBH or PLEA + TPBH exhibited only faint green fluorescence, confirming the potent free radical scavenging capacity of PLEA in cellular environments. To assess the protective effects of PLEA under oxidative stress, HT22 cell viability was measured after intracellular ROS induction. As shown in Figure 3E, TPBH (200 μM) significantly reduced HT22 cell viability to 28.6%. However, treatment with PLEA effectively protected HT22 cells from TPBH-induced oxidative damage, as demonstrated by a marked improvement in cell viability. Additionally, flow cytometry was performed to investigate whether PLEA could mitigate TPBH-induced apoptosis. Exposure to TPBH alone resulted in significant cell apoptosis. However, the addition of PLEA to TPBH-treated cells markedly reduced apoptosis rates, indicating its protective effects against oxidative stress (Figures 3F and 3G). Collectively, these results demonstrate that PLEA possesses robust free radical scavenging capabilities and can effectively protect HT22 cells from oxidative damage due to its efficient free radical scavenging abilities.

Figure 3.

Antioxidant and cytoprotective effects of PLEA

(A and B) Free radical scavenging activity of Eda and PLEA at different concentrations, evaluated using ABTS and DPPH assays. n = 3, error bars represent SD, mean ± SD.

(C) Representative fluorescence images of intracellular ROS levels in different treatment groups, detected by DCFH-DA probe. Scale bars, 50 μm (for regional magnification: scale bars, 10 μm).

(D) Quantitative analysis of ROS fluorescence intensity with different treatments. n = 3, error bars represent SD, mean ± SD; ∗∗∗p < 0.001.

(E) Cell viability of TPBH-induced oxidative stress HT22 cells treated with PBS, Eda, and PLEA at different concentrations. n = 3, error bars represent SD, mean ± SD, ∗p < 0.05, ∗∗p < 0.01, ∗∗∗∗p < 0.0001.

(F) Flow cytometry analysis of apoptosis in different treatment groups.

(G) Apoptosis rate under oxidative stress conditions with different treatments. n = 4, error bars represent SD, mean ± SD, ∗∗p < 0.01; ns, no significance.

ABTS, 2,2'-Azinobis-(3-ethylbenzothiazoline-6-sulfonic acid); DPPH, 2,2-Diphenyl-1-picrylhydrazyl; DCFH-DA, 2′,7′-dichlorodihydrofluorescein-diacetate; Eda, edaravone; TPBH, tert-butyl hydroperoxide.

Ex vivo and in vivo thrombolysis efficacy of PLEA

The thrombolytic potential of PLEA was tested on ex vivo blood clots, prepared by mixing fresh mouse blood with thrombin. Clot dissolution was monitored by measuring the hemoglobin levels in the supernatant at different time points, and the percentage of mass loss after 8 h was used to quantify thrombolysis (Figure 4A). In the LEA group, the thrombolytic efficiency was slightly lower than that of free rt-PA, likely due to the sustained release of rt-PA from liposomes (Figure 4B). However, the difference was not statistically significant. In the PLEA group, the platelet membrane coating seemed to enhance the nanoparticles’ targeting ability, resulting in better thrombolysis. At the 4–8 h time point, the thrombus-dissolving capacity of PLEA was comparable to, or even slightly higher than, that of free rt-PA, indicating that the platelet membrane modification in PLEA did not impair the thrombolytic performance of rt-PA (Figures 4B and 4C). For in vivo evaluation, a carotid artery thrombus model was established in male C57BL/6J mice using FeCl₃ to induce clot formation. After thrombus formation, saline, rt-PA, Eda, LEA, or PLEA was administered via the tail vein (Figures 4D and 4E). Thrombus dissolution was monitored via Doppler ultrasound imaging. As shown in Figure 4E, FeCl₃-induced thrombi were clearly visible in the left common carotid artery (LCCA) prior to treatment. At 4 h post-treatment, Doppler ultrasound assessments of different groups revealed distinct anatomical details and imaging features: transverse views clearly depicted the right common carotid artery (RCCA) and LCCA (marked with circles), with the thrombus within the LCCA presenting as a slight hyperintense (Figure S3A). Meanwhile, longitudinal views outlined the vessel wall of the LCCA, and blood flow velocity in the LCCA of each group was simultaneously quantified using power Doppler. Notably, PLEA exhibited a markedly enhanced thrombolytic capacity compared to the other treatment groups, as reflected by improved blood flow velocity (Figure S3B). Finally, carotid arteries were harvested for histological analysis to accurately quantify the thrombus burden, expressed as the percentage of the arterial lumen occupied by thrombus (Figure 4E). Compared to saline group (91.51%), PLEA achieved the most significant thrombus reduction (56.12%), followed by LEA (64.29%), rt-PA (76.32%), and Eda (88.98%). These results confirm the potent thrombolytic capacity of PLEA. There are substantial differences between in vivo and in vitro environments. The encapsulation of rt-PA within liposomes, combined with platelet membrane modifications, allows for significantly prolonged circulation time in the bloodstream compared to the inherently short half-life of rt-PA. This extended circulation time enhances the therapeutic efficacy of rt-PA. Moreover, the platelet membrane modification further improves the nanoparticles’ targeting ability, thereby leading to a superior thrombolytic effect compared to rt-PA alone.

Figure 4.

Evaluation of the thrombolytic efficacy of PLEA in vitro and in vivo

(A) Representative images showing clot degradation at different time points (pre, 1 h, 2 h, 4 h, 6 h, and 8 h) following treatment with saline, rt-PA, Eda, LEA, or PLEA.

(B) Quantitative analysis of clot lysis over time, measured by OD values at 540 nm. n = 3, error bars represent SD, mean ± SD.

(C) Percentage change in clot weight after treatment with different formulations. n = 3, error bars represent SD, mean ± SD; ns, no significance.

(D) Representative images of carotid artery thrombus formation and treatment outcomes in different groups.

(E) H&E-stained cross-sections of thrombi from different treatment groups. Scale bars, 50 μm.

(F) Quantification of thrombus area as a percentage of arterial lumen following treatment. n = 3, error bars represent SD, mean ± SD; ∗∗p < 0.01; ∗∗∗∗p < 0.0001; Eda, edaravone.

In vivo targeting effect of PLEA in a thrombus mouse model

Encouraged by the promising in vitro targeting performance of PLEA, we further evaluated its in vivo thrombus-targeting ability using a carotid artery thrombus mouse model. To ensure an accurate assessment, rt-PA was not included in these experiments to avoid fluorescence dilution caused by thrombolysis. For real-time visualization, ICG-labeled LE and PLE were intravenously administered, and their distribution was tracked using IVIS spectrum. Additionally, to exclude potential interference from the surgical procedure, two sham groups were included. The carotid artery of these groups was isolated, which was washed by saline to keep the incision without thrombus formation (Figure 5A). Imaging results revealed minimal nanoparticle accumulation in the neck region of the sham groups, indicating that the surgical procedure had no significant impact on nanoparticle distribution. In contrast, PLE demonstrated rapid and pronounced accumulation at the thrombus site, with peak fluorescence intensity observed within 2 h, which remained stable thereafter (Figures 5A and 5B). In comparison, the LE group exhibited significantly lower signal intensity at the thrombus site, highlighting the superior targeting capability of PLE in vivo. To further validate the targeting efficiency of PLE, the mice were sacrificed, and bilateral carotid arteries of all groups were excised for ex vivo fluorescence analysis. As shown in Figures 5C and 5D, the carotid arteries of the PLE-treated group exhibited markedly stronger fluorescence signals than those of the LE-treated group. For non-thrombus regions in the thrombus-bearing mice, quantitative fluorescence measurement revealed that these regions exhibited fluorescence intensities comparable to the sham control, confirming the specific targeting effect of PLE. To investigate the biodistribution and clearance of PLEA, we analyzed its accumulation in major organs, including the heart, liver, spleen, lungs, and kidneys, at 24 and 48 h post-injection (Figures 5E and 5F). At the 24-h mark, fluorescence intensity in these organs had notably diminished, and by 48 h, the signals were almost undetectable, suggesting that PLEA was efficiently cleared from the system. Besides, at the 24 h time point, fluorescence was predominantly localized in the liver and kidneys, suggesting that PLEA is primarily metabolized and eliminated through these organs.

Figure 5.

In vivo targeting and biodistribution of PLEA

(A) Representative in vivo fluorescence imaging of thrombus-bearing mice at different time points (pre, 2 h, 4 h, and 6 h) after administration of LE or PLE. Sham PLE and LE groups serve as controls.

(B) Quantitative analysis of radiant efficiency over time. n = 3, error bars represent SD, mean ± SD, ∗∗∗p < 0.001.

(C) Ex vivo fluorescence imaging of ipsilateral (thrombotic) and contralateral (non-thrombotic) carotid arteries.

(D) Quantitative analysis of radiant efficiency of bilateral carotid arteries in different groups. n = 3, error bars represent SD, mean ± SD; ∗∗∗p < 0.001.

(E) Ex vivo fluorescence imaging of major organs at 24 and 48 h post-injections of LEA and PLEA.

(F) Quantification of radiant efficiency in major organs (heart, liver, spleen, lung, and kidney) at 24 and 48 h post-injection. n = 3, error bars represent SD, mean ± SD.

The in vivo therapeutic effects and photoacoustic assessment of the PLEA against brain infarction

Inspired by the promising results mentioned above, we further explored the protective effects of PLEA against stroke in an in vivo model. The stroke model was induced by photothrombosis, which involved the intravenous injection of Rose Bengal and subsequent laser irradiation to induce thrombus formation, as previously described.^32,35 As depicted in the experimental protocol in Figure 6A, the mouse stroke models were divided into various groups (group 1: sham; group 2: saline; group 3: rt-PA; group 4: Eda; group 5: LEA; and group 6: PLEA). After the respective treatments, the brains were collected to evaluate the permeability of the blood-brain barrier (BBB) through Evans blue staining. Additionally, infarct volume was determined using 2,3,5-triphenyltetrazolium chloride (TTC) staining. Compared to the rt-PA and Eda groups, the mice treated with PLEA showed a significant reduction in BBB permeability and a smaller cerebral infarction size (Figures 6B–6E). Likewise, hematoxylin & eosin (H&E) and Nissl staining confirmed that the infarct area in the PLEA group was considerably smaller than those in the other groups (Figures 6F and 6G). These findings indicate that the PLEA can protect neural cells from ischemic stroke. To obtain a more sensitive functional readout, an accelerating rotarod assay was incorporated (beginning at 4 rpm and ending at 40 rpm, accelerating speed at 5 rpm min⁻¹). Each mouse was subjected to three trials at 72 h post-stroke, and the latency to fall was averaged. The mice treated with PLEA showed a significantly prolonged latency to fall compared with the PBS group, indicating enhanced functional motor recovery (Figure S4). Next, ROS content in the brain tissue was determined by flow cytometry to investigate the free radical scavenging effect of PLEA in vivo. The cells from the stroke site of the brain were extracted and co-incubated with DCFH-DA. As shown in Figures 6H and 6I, increased ROS levels were observed in the saline group. After treatment, ROS concentration could be reduced to varying degrees in different groups, among which the PLEA treatment group exhibited significantly reduced ROS accumulation. Consistent with the results, we observed a higher ROS scavenging property of PLEA than the other groups in brain tissue homogenate (Figure 6J). Meanwhile, the inflammatory cytokine TNF-α was also tested, where the PLEA showed the best anti-inflammatory and brain protection effect (Figure S5), indicating that the as-designed PLEA holds great potential for the treatment of stroke. To further elucidate the protective mechanism of PLEA administration, RNA-seq of the infarct cortex tissue was performed. Compared with saline-treated controls, 426 genes were differentially expressed (adjusted p < 0.05, |log₂FC| > 1), of which 161 were upregulated and 265 were downregulated (Figure S6A). Hierarchical clustering of these differentially expressed genes (DEGs) clearly separated the PLEA-treated samples from the controls (Figure S6B), indicating that PLEA induces a distinct transcriptional program in the injured cortex. Functional enrichment analysis of the downregulated gene set revealed a pronounced over-representation of biological processes related to immune system regulation, cell-cell adhesion, and extracellular matrix organization (Figure S6C). Consistently, KEGG pathway enrichment showed that downregulated genes were significantly enriched in the inflammatory and adhesion-related pathways, including cytokine-cytokine receptor interaction, antigen processing and presentation, and cell adhesion molecules (Figure S6D). These molecular alterations are concordant with the phenotypic observations of reduced BBB leakage and smaller ischemic lesions in PLEA-treated animals. Earlier studies have shown that Eda acts as a potent free-radical scavenger to suppresses oxidative stress-induced activation of inflammatory cascades³⁶ and preserve BBB integrity by downregulating matrix-metalloproteinase expression,³⁷ which is consistent with our sequencing results. By suppressing genes that drive immune activation, leukocyte adhesion, and ECM remodeling, PLEA attenuates the post-stroke inflammatory cascade and reinforces vascular integrity, thereby contributing to neurovascular protection after cerebral ischemia.

Figure 6.

Evaluation of therapeutic effects in ischemic stroke mouse models

(A) Experimental scheme of photothrombotic stroke induction, treatment with saline, rt-PA, Eda, LEA, or PLEA at 2 h post-stroke, and multimodal assessments at 72 h post-treatment.

(B) Representative Evans blue-stained brain images in sham, saline-, rt-PA-, Eda-, LEA-, and PLEA-treated groups.

(C) Quantitative analysis of BBB permeability based on Evans blue staining. n = 3, error bars represent SD, mean ± SD; ∗∗p < 0.01, ∗∗∗∗p < 0.0001.

(D) Representative TTC-stained brain images across different treatment groups.

(E) Quantitative analysis of infarct volume from TTC staining. n = 3, error bars represent SD, mean ± SD; ∗∗∗∗p < 0.0001.

(F) Representative H&E and Nissl staining images of cerebral infarct area in sham-, saline-, rt-PA-, Eda-, LEA-, and PLEA-treated groups. Scale bars (black and red), 100 and 20 μm, respectively.

(G) Quantitative analysis of infarct volume based on H&E staining. n = 3, error bars represent SD, mean ± SD; ∗∗p < 0.01, ∗∗p < 0.0001.

(H) Flow cytometry analysis of ROS levels using DCFH-DA fluorescence intensity in brain tissue.

(I) Quantification of DCFH-positive cells in each group. n = 4, error bars represent SD, mean ± SD; ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001.

(J) Quantification of mean fluorescence intensity indicating intracellular ROS levels. n = 4, error bars represent SD, mean ± SD; ∗p < 0.05, ∗∗∗∗p < 0.0001.

(K) Representative photoacoustic (PA) and oxygen saturation (sO₂) imaging of brain cortex post-ischemia and post-treatment in sham, saline-, and PLEA-treated groups. Scale bars, 500 μm.

(L) Quantitative analysis of PA signal intensity, and relative sO₂ levels before and after treatment. n = 3, error bars represent SD, mean ± SD; ∗p < 0.05, ∗∗∗p < 0.001.

BBB, blood-brain barrier; TTC, 2,3,5-triphenyltetrazolium chloride; DCFH-DA, 2′,7′-dichlorodihydrofluorescein-diacetate; ROS, reactive oxygen species.

To further evaluate the therapeutic efficacy of PLEA, PA imaging was employed to assess cerebral hemodynamic changes and monitor the progression of ischemic stroke. As shown in Figures 6K and 6L, the sham group exhibited uniform PA signals, indicating intact cerebral perfusion and well-oxygenated brain tissue. In contrast, the saline-treated stroke group showed a marked reduction in PA signal intensity of the left hemisphere (indicated by the dashed box), with disrupted vascular structures and widespread hypoxia, as evidenced by the increased ischemic regions in the oxygenation maps. Notably, PLEA treatment led to a significant restoration of PA signals, reflecting improved blood flow and oxygenation in the ischemic brain. Compared to the saline group, the PLEA-treated mice displayed a substantial reduction in hypoxic areas and enhanced cerebral perfusion, suggesting its potential in promoting vascular recovery and mitigating ischemic damage. These findings highlight the effectiveness of PLEA in restoring blood flow and oxygen supply to the brain, further supporting its role as a promising therapeutic strategy for ischemic stroke.

Biosafety evaluation of PLEA

The cytotoxicity of PLEA in the neural cell line HT22 was initially evaluated. Following a 24 h incubation with PLEA at concentrations up to 200 μg/mL, cell viability remained above 80% and had no significant difference with control groups (Figure S7), suggesting a low cytotoxic property of PLEA. Hemolysis assays demonstrated favorable hemocompatibility, as evidenced by minimal hemolysis after 2 h of co-incubation with PLEA (Figure S8). Subsequently, in vivo safety evaluations were conducted in healthy mice. The mice were administered PLEA intravenously at a dose of 20 mg/kg, and after 3 weeks, they were humanely euthanized for a comprehensive hematological and histological analysis. Hematological analysis was performed on healthy mice and those treated with PLEA, and no significant differences were observed between the two groups (Figure S9). Additionally, histopathological evaluation of major organs (heart, liver, spleen, lungs, and kidneys) stained with H&E showed no pathological changes in either the healthy controls or the PLEA-treated mice (Figure S10).

Discussion

In summary, we developed PLEA, a synchronous biomimetic nanoparticle for targeted AIS therapy. PLEA was designed by encapsulating Eda in the lipid bilayer of platelet membrane-liposomes and loading rt-PA into the core. The incorporation of platelet membranes enabled PLEA to mimic natural thrombus-targeting mechanisms, enhancing its accumulation at clot sites and improving drug delivery efficiency. In vivo studies demonstrated that PLEA effectively reduced the infarct volume, preserved BBB integrity, and alleviated oxidative stress and inflammation. Recent AIS-targeted delivery platforms have demonstrated notable advances in loading capacity, thrombus-specific accumulation, and safety. Similar platelet-membrane-coated systems reported by Quan et al.³⁵ could achieve thrombus-specific binding and markedly reduced the intracerebral-hemorrhage risk. Stimulus-responsive nanocarriers, such as ROS-responsive PLGA nanostructures, provide controlled release across the BBB, with drug-loading capacities exceeding 10 wt% and can be functionalized with targeting ligands for precise lesion delivery.³⁸ When benchmarked against these state-of-the-art AIS delivery systems, PLEA exhibits a superior drug-loading capacity by encapsulating rt-PA in the core and embedding Eda within the platelet-membrane bilayer. The platelet membrane coating endows the nanoparticles with active thrombus-homing capability via native receptors (e.g., GPIbα, P-selectin, and integrin αIIbβ3), which is comparable to that of the other platelet membrane-based systems and superior to ligand-only nanocarriers, while also enhancing safety. Natural platelet membrane components reduce immune clearance, prolong circulation, and limit hemorrhagic transformation. Despite these strengths, PLEA also has limitations that warrant further investigation. Short-term biocompatibility was confirmed by routine hematology, coagulation assays, and organ histology up to three weeks post-administration, yet the long-term safety of the platform remains undefined. Systematic studies are needed to characterize biodistribution over extended periods and to evaluate the potential immunogenicity associated with repeated dosing. Beyond evaluating the nanocarrier itself, assessing therapeutic efficacy in a real-time, non-invasive manner is critical for translating AIS therapies into clinical practice, especially given the dynamic nature of cerebral perfusion and oxygenation during stroke progression. To address this need, we employed PA imaging to monitor the in vivo therapeutic performance of PLEA. PA imaging confirmed its ability to restore cerebral perfusion and oxygenation, providing a real-time, non-invasive method to assess treatment efficacy. This study highlights PLEA as a promising approach for stroke therapy, offering both thrombolytic and neuroprotective benefits, and PA imaging serves as a valuable tool for monitoring stroke progression and therapeutic outcomes.

Limitations of the study

Despite the promising findings of this study, several limitations warrant attention. First, the long-term safety profile of the PLEA nanosystem remains incompletely characterized (including potential accumulation in reticuloendothelial system and immunogenicity following repeated administration), given that only short-term (up to 3 weeks) biocompatibility was assessed. Second, the young, healthy rodent models employed do not fully recapitulate the complex pathophysiology of human AIS, such as the presence of comorbidities and age-related physiological changes, which may compromise translational applicability. Additionally, this study exclusively utilized male mice to model ischemic stroke; future investigations will incorporate female mice to explore potential gender-specific differences in therapeutic efficacy. Third, although PA imaging facilitated real-time preclinical monitoring, its clinical translation is hindered by the scarcity of specialized clinical-grade instruments and standardized imaging protocols tailored to stroke patients. Future studies are, therefore, warranted to further explore the clinical applicability of PA imaging for stroke diagnosis, which will be crucial for unlocking the full translational potential of the PLEA-PA integrated platform.

Resource availability

Lead contact

Further information and requests for resources, data, and codes should be directed to and will be fulfilled by the lead contact, Prof. Donghai Wang (drwangdonghai@sdu.edu.cn).

Materials availability

All materials and methods are presented in the paper or can be made available upon request from the lead contact.

Data and code availability

Data

All data generated in this study are presented in the main and supplementary figures and tables. RNA-seq data of infarct cortex tissue have been deposited in the Gene Expression Omnibus (GEO) database with the accession number GEO: GSE314676.

Code

This article does not report original code.

Other items

All additional materials, reagents, and experimental protocols relevant to this study are available from the lead contact upon reasonable request.

Acknowledgments

This work was supported by the Chinese Academy of Sciences Grant (Strategic Priority Research Program XDB0930000, Young Scientists in Basic Research grant no. YSBR-104, and Youth Innovation Promotion Association Y2023099, 2023374); National Key Research and Development Program of China (2023YFF0715300, 2023YFC2411700); National Natural Science Foundation of China Grant (82327805, 82172008); Shenzhen Science and Technology Innovation Grant (JCYJ20241202124916023, JCYJ20220818101403008, KJZD20240903101259001, KJZD20240903095714019, JCYJ20240813155900001, and JCYJ20220818101404009); Guangdong Provincial Key Laboratory of Biomedical Optical Imaging (2020B121201010); Shenzhen Key Laboratory for Molecular Imaging (SYSPG20241211173902013)); Shandong Provincial Natural Science Foundation (ZR2023MH023); Clinical Research Project of Shandong University (2021SDUCRCA009); National Health Commission Capacity Building and Continuing Education Center Nervous system and minimally invasive intervention project fund (1003); Crosswise tasks (contract number: 6010120062).

Author contributions

M.H., H.Z., and Y.Z. made equal contributions to this research; D.W., J.C., and L.Z. initiated the study concept; M.H., H.Z., Y.Z., and L.L. were responsible for material design, preparation, and characterization; the animal experiments were conducted by Z.X. and Y.H; data analysis, curation, and validation were performed by Y.R., Z.X., Y.T., and C.L; M.H. and H.Z. undertook the manuscript writing and editing, while L.Z. and J.C. provided critical review. All authors contributed to the manuscript and have approved the final version.

Declaration of interests

The authors declare no competing interests.

STAR★Methods

Key resources table

REAGENT or RESOURCE	SOURCE	IDENTIFIER
Antibodies
Rabbit Monoclonal Anti-CD41 antibody	Abcam	Cat#ab134131; RRID: AB_2732852
Rabbit Monoclonal Anti-P-Selectin (CD62P) antibody	Abcam	Cat#ab255822
Rabbit Monoclonal Anti-GPVI antibody	Abcam	Cat#ab289987
Rabbit Monoclonal Anti-Integrin beta 1 antibody	Abcam	Cat#ab179471; RRID: AB_2773020
Chemicals, peptides, and recombinant proteins
Rose Bengal	Sigma–Aldrich	Cat#330000
Tert-butyl hydroperoxide (TPBH)	Sigma–Aldrich	Cat#418064
2′,7′-dichlorodihydrofluorescein-diacetate (DCFH-DA)	Sigma–Aldrich	Cat#D6883
Ethylenedinitrilotetraacetic acid (EDTA)	Sigma–Aldrich	Cat#E9884
3,3′-dioctadecyloxacarbocyanine perchlorate (DiO)	Beyotime	Cat#C1038
1,1′-dioctadecyl-3,3,3′,3′-tetramethylindocarbocyanine perchlorate (Dil)	Beyotime	Cat#C1036
Edaravone	MedChemExpress	Cat#HY-B0099
Indocyanine green	MedChemExpress	Cat#HY-D0711
DSPE-PEG2000	Avanti Polar Lipids	Cat#793025
DPPA	Avanti Polar Lipids	Cat#830855
DPPC	Avanti Polar Lipids	Cat#850858
Critical commercial assays
Annexin V-FITC/PI Kit	Beyotime	Cat#C1062L
CCK-8 Kit	Beyotime	Cat#C0038
DPPH Free Radical Scavenging Capacity(Total Antioxidant Capacity-DPPH Method) Assay Kit	Solarbio	Cat#BC4755
ABTS Free Radical Scavenging Capacity(Total Antioxidant Capacity- ABTS Method) Assay Kit	Solarbio	Cat#BC4770
Human Tissue-Type Plasminogen Activator ELISA Kit	Shanghai Mlbio	Cat#ml105345
Deposited data
RNA–Seq dataset of saline and PLEA-treated mice brain tissues	This paper	GSE314676
Experimental models: Cell lines
HT22 (Mouse hippocampal neuronal cells)	Procell	Cat#CL-0697
Experimental models: Organisms/strains
C57BL/6J mice	Beijing Vital River Laboratory Animal Technology	–
Software and algorithms
GraphPad Prism 9.0	GraphPad Software	https://www.graphpad.com
ImageJ	NIH	https://imagej.nih.gov/ij/
FlowJo	BD Biosciences	https://www.flowjo.com/
MATLAB R2023a	MathWorks	https://www.mathworks.com/
Figdraw 2.0	Home for Researchers	https://www.figdraw.com/

Experimental model and study participant details

Mice

Wild-type male C57BL/6J mice (8–10 weeks, 22–25 g) were used. All mice were housed in a specific pathogen-free (SPF) environment with temperature 22°C–25°C, relative humidity 40%–60%, and 12-h light-dark cycle, with free access to standard rodent feed and sterile water. All animal procedures were approved by the Animal Care and Use Committee of the Shenzhen Institute of Advanced Technology, Chinese Academy of Sciences (SIATACUC, Approval No.: SIAT-IACUC-250324-YGS-CJQ-A2933) and complied with relevant guidelines.

Cells

HT22 mouse hippocampal neuronal cells were obtained from Procell Inc. (Catalog^#CL-0697). The cells were cultured in DMEM medium supplemented with 10% fetal bovine serum (FBS) and 1% penicillin/streptomycin solution, and maintained in a 37°C incubator with a humidified atmosphere containing 5% CO₂. All HT22 cells were authenticated by short tandem repeat profiling (performed by Procell Inc.) and tested negative for mycoplasma contamination using a PCR-based detection kit.

Method details

Preparation of PLEA

Liposomes were made via thin - film hydration and extrusion. Specifically, 18 mg of DPPC, 1,2 - distearoyl - sn - glycero - 3 - phosphoethanolamine - N - [poly(ethylene glycol)] - 2000 (DSPE - PEG2000), 1 mg of DPPA, and 10 mg of edaravone were dissolved in 5.0 mL of dichloromethane. After that, it was evaporated at 55°C using a rotary evaporating system to obtain a thin lipid film. Then, 5 mg of rt-PA dissolved in 5 mL of ddH₂O and added to the lipid film. The suspension was sonicated in an ice bath for 20min and then extruded through 0.4 and 0.2 μm polycarbonate membranes (Whatman, Maidstone, UK) at room temperature using an extruder system (Avanti Polar Lipids, Alabaster, AL, USA). The liposomes encapsulated with rt-PA were stored at 4°C until use.

Platelet membranes were prepared following previously established protocols.³⁹ Initially, whole blood from the orbital veins of male C57BL/6J mice was collected. The collected blood was anticoagulated with 5 mM ethylenediaminetetraacetic acid (EDTA). Next, the blood underwent centrifugation at 100g for 20 min. This process yielded platelet - rich plasma, which was then transferred to a separate tube. To avoid platelet activation, 2 μM of prostaglandin E1 was introduced prior to centrifuging the blood at 850g for 20 min to isolate the platelets. After the supernatant was carefully discarded, the platelets were re-dispersed in 1x PBS supplemented with 1 mM EDTA and a protease inhibitor. The platelet membranes were extracted by subjecting the suspensions to multiple cycles of snap-freezing in liquid nitrogen followed by thawing at room temperature for a minimum of three cycles. The suspension was then sonicated for 5 min using a 100 W, 42 kHz sonicator. Afterward, the membrane fractions were collected by centrifuging at 4000g for 15 min at 4°C, followed by three washes with PBS. The purified membranes were stored at −80°C. To prepare hybrid platelet-liposomes (PLEA), liposomes and platelet membranes were mixed at a 10:1 mass ratio, placed in a capped vial, and sonicated for 30 s. The size of the PLEA was then reduced to under 200 nm by extruder.

Characterization of PLEA

The structure of LEA and PLEA was analyzed using transmission electron microscopy (TEM) after negative staining with a 2% sodium phosphotungstate solution. UV-Vis-NIR absorption spectra were obtained with a Shimadzu UV-3600 spectrophotometer (Japan). Drug release was evaluated at 37°C in PBS. PLEA was suspended in PBS and incubated in a water bath at 37°C while being gently stirred with a magnetic stir bar (200 rpm/min) to maintain a uniform suspension. At 0.5, 1, 2, 4, 6, 8, 10, and 12 h, the suspension was ultracentrifuged for 1 h, and the supernatant was collected. The released rt-PA in supernatant was quantified using a Human Tissue-Type Plasminogen Activator ELISA Kit (Mlbio Co., Ltd.), strictly following the manufacturer’s instructions. In parallel, the released Eda was measured by its UV-Vis absorbance at 244 nm. To more accurately simulate drug release under protein-rich physiological conditions, PBS was supplemented with 3% bovine serum albumin (BSA). For the activated platelet assay, 200 μL of platelet suspension (1.0 × 10⁸ cells mL⁻¹) was added to a microplate, pre-activated by 20 μL thrombin (2 U mL⁻¹) and 8 nM CaCl₂, and incubated for at least 2 h. The release rate was calculated as the ratio of the amount of drug released to the initial administered dose. Particle size, polydispersity index, and zeta potential were determined by dynamic light scattering (DLS) on a Zetasizer Nano ZS90 (Malvern, U.K.). To assess membrane fusion, platelet membranes were stained with 10 μM Dil dye (Beyotime, China) and liposomes with 5 μM Dio dye (Beyotime, China) for 15 min at room temperature. Fluorescence signals were captured using confocal microscopy (Nikon A1, Japan). The protein components of the samples were measured through Coomassie blue staining, following the procedure outlined in a previous study.⁴⁰

In vitro evaluation of the antioxidant and neuroprotective ability of PLEA

The free radical scavenging ability of PLEA was evaluated through Total Antioxidant Capacity Assay Kit. 2,2′-azinobis(3-ethylbenzthiazoline-6-sulfonate (ABTS) and 2,2-diphenyl-1-picrylhydrazyl (DPPH) was used to evaluate the free radical scavenging ability of PLEA. In short, various concentrations of Eda or PLEA were introduced to the ABTS⋅+ or DPPH free radical solution, and the absorbance was monitored at 405 nm or 517 nm (Solorbio, China).

Intracellular ROS detection was assessed by 2′,7′-Dichlorodihydrofluorescein diacetate (DCFH-DA) (Beyotime, China). HT22 cells were pretreated with tert-butyl hydroperoxide (TPBH) (200 μM) for 12 h, followed by incubation with different concentrations of Eda or PLEA for 6 h. Subsequently, the cells were incubated with DCFH-DA for 30 min at room temperature. The blank control group did not receive any drug treatment, with all other conditions being the same as the experimental group. ROS generation in the cells was assessed using confocal microscopy.

Protective effect of PLEA against TPBH was assessed through CCK-8 assay and Annexin V-FITC/PI apoptosis detection kit (Beyotime, China). Briefly, diverse concentrations of PLEA/Eda and TPBH (200 μM) were incubated with HT22 cells for 24 h. Cell viability was then assessed using the CCK-8 assay. To assess the apoptosis rate, HT22 cells were firstly resuspended in Annexin V-binding buffer and incubated with 2 μM Annexin V-FITC in the dark for 15 min. Then, they were stained with propidium iodide for 10 min. The fluorescence of the cells was analyzed via flow cytometry (Beckman, USA).

In vitro and in vivo assessment of thrombolysis efficacy

The in vitro assessment of thrombolytic effect of PLEA was determined by thrombolysis of artificial blood clots. The artificial blood clots were weighed and transferred to centrifuge tubes containing 1 mL of saline. Then, they were placed into centrifuge tubes containing 1 mL saline. containing 1 mL of saline. Next, 100 μL of saline, rt-PA, Eda, LEA, or PLEA (all adjusted to the same concentration as rt-PA, 10 μg/mL) was added to the tubes, and the samples were incubated at 37°C. The effectiveness of thrombolysis was monitored by measuring the absorbance at 540 nm of the supernatants at different time intervals. After 12 h, the clots were reweighed, and the clot lysis percentage was calculated based on the change in mass.

For in vivo thrombolytic effect evaluation, carotid artery thrombosis was induced according to a previously described method. In brief, C57BL/6 mice were anesthetized and the left carotid artery was carefully expose. A piece of filter paper soaked in a 10% ferric chloride solution was then placed on the surface of the carotid artery for 3 min. Once the filter paper was removed, the thrombus formed in the carotid artery was observed, and the artery was washed with saline. Subsequently, 100 μL of saline, eda, rt-PA, LEA or PLEA was administered through tail vein. The incision was then sutured and reopened after 12 h to evaluate the degree of vascular recanalization. Finally, the mice were euthanized, and the carotid arteries were excised for histological analysis using hematoxylin and eosin (HE) staining to measure the clot area. Thrombolytic efficiency was quantified by the ratio of the vascular occlusion area to the total vascular area using ImageJ software.

In vivo thrombus targeting and biodistribution of PLEA

Following the carotid artery thrombosis was induced, ICG-LEA or ICG-PLEA was administered intravenously at a dose of 20 μg of ICG. Mice in the sham group had their arteries exposed and were treated with filter paper soaked in saline. Following the sham procedure, they were injected with ICG-LEA or ICG-PLEA. Over a 6-h period, the fluorescence signals at various time points in the carotid arteries were measured using an IVIS spectrum system. At the final time point, the carotid arteries were collected, and ex vivo imaging was carried out to verify the accumulation of PLEA. For biodistribution of PLEA, ICG labeled LEA or PLEA were injected intravenously. The mice were euthanized at 24 h or 48 h, and their major organs, including the heart, liver, spleen, lungs, and kidneys, were harvested. The distribution of ICG-labeled LEA or PLEA within the major organs was detected using IVIS. Fluorescence intensities were then analyzed.

Photochemically-induced stroke models

The acute ischemic stroke model, induced through photochemical methods, followed previously established protocols.^41,42 In brief, male C57/BL6J mice (8–10 weeks old, weighing 22-25g, from Beijing Vital River Laboratory Animal Technology Co., Ltd) were anesthetized with 1% isoflurane. A midline head incision was made, and the connective tissue was removed to expose the skull. Rose Bengal was intravenously at a dose of 100 mg/kg. A 250 mW 532 nm laser with a 1.5 mm diameter was directed at the left hemisphere and irradiated for 5 min. Afterward, the scalp was sutured, completing the induction of the stroke model.

In vivo assessment of BBB permeability, infarction volume, and antioxidant effect of PLEA

Acute ischemic stroke model mice were randomly divided into 5 group, saline, rt-PA, Eda, LEA or PLEA were administered intravenously. To assess infarction volume, mice were euthanized 72 h after treatment, and their brains were sectioned into 2 mm-thick slices. The slices were incubated in a 2% TTC solution for 20 min (10 min on each side) and then fixed in a 4% paraformaldehyde solution at room temperature for 30 min. To assess blood-brain barrier (BBB) permeability after treatment, Evans blue (2% w/v) was used. All sections were photographed using a digital camera, and the infarct sizes and BBB permeability were quantified using ImageJ software.

To further evaluate the ROS scavenging effect of PLEA in vivo, DCFH-DA was incubated with single-cell suspension isolated from brain tissues and kept at 37°C in the dark for 30 min. The intracellular ROS levels were then assessed using flow cytometry.

Quantification and statistical analysis

All data are expressed as mean ± standard deviation unless otherwise stated. For comparisons between two groups, unpaired two-tailed Student’s t test was performed. For multiple group comparisons, one-way analysis of variance (ANOVA) followed by Tukey’s post hoc test was applied. A p-value <0.05 was considered statistically significant. All statistical analyses were conducted using GraphPad Prism 9.0 Software (GraphPad Software, Inc., USA). Detailed statistical parameters, including exact n values and the definition are provided in the corresponding figure legends.

Published: January 22, 2026