Carrier-Free Rapamycin-Loaded Nanoparticles with MRI and Near-Infrared Fluorescence Bimodal Imaging for Precise Diagnosis and Synergistic Therapy of Ischemic Stroke

Abstract

Background

Ischemic stroke (IS) management is challenged by low drug delivery efficiency and limitations in current imaging modalities. We aimed to assemble rapamycin (RAPA), indocyanine green (ICG), epigallocatechin-3-gallate (EGCG), and Gd³⁺ as neuroprotective drugs, near-infrared fluorescence (NIRF) imaging agents, biomimetic coatings, and MRI agents, respectively, into a carrier-free, dual-modal imaging-guided theranostic nanoplatform (IREGd NPs) for the integrated diagnosis and therapy of IS.

Methods

The physicochemical properties, cytotoxicity, drug release characteristics, and cellular uptake of IREGd NPs were characterized. Therapeutic efficacy was evaluated in vitro using oxygen-glucose deprivation/reoxygenation (OGD/R) models of cell injury and in vivo in a transient middle cerebral artery occlusion (tMCAO) rat model. Outcomes included sensorimotor function, cerebral infarct volume, brain edema, neuronal apoptosis, neuroinflammation, oxidative stress, and biodistribution. The bimodal imaging properties were verified in vitro and in vivo.

Results

IREGd NPs demonstrated good dispersibility and stability, with a RAPA loading efficiency of 65.9 wt%, enabling enhanced MRI and NIRF imaging capabilities with a longitudinal relaxation rate of 40.18 mM⁻¹·s⁻¹ for precise visualization of the ischemic region and drug biodistribution. In vitro experiments confirmed effective cellular uptake and superior neuroprotective effects in OGD/R-treated cells compared with RAPA or EGCG alone. Finally, in vivo experiments in rats with tMCAO confirmed that IREGd NPs passively accumulated in the ischemic hemisphere, enabling accurate bimodal imaging and exerting neuroprotective effects, as evidenced by improved sensorimotor function, reduced cerebral infarct volume and brain edema, decreased neuronal apoptosis, and attenuated neuroinflammation and oxidative stress, outperforming individual RAPA or EGCG treatment. Histology revealed no systemic toxicity.

Conclusion

IREGd NPs efficiently deliver RAPA, and the neuroprotective effects are exerted by RAPA in combination with EGCG. Dual-modal imaging enables precise visualization of the ischemic region and drug biodistribution. These results suggest the translational potential of the biomimetic nanotheranostic platform for the treatment of IS.

Introduction

Ischemic stroke (IS) is a neurological disorder characterized by neuronal death due to the occlusion of cerebral arteries, leading to high rates of disability and mortality.¹ Current therapeutic strategies primarily focus on early restoration of cerebral blood flow. However, they are constrained by a narrow treatment window and the risk of cerebral ischemia-reperfusion injury (CIRI).² Neuroprotective drugs have shown promise in mitigating brain damage and improving functional recovery both before and after cerebral ischemia-reperfusion. Despite this potential, the efficacy of conventional neuroprotective agents is often limited by poor blood-brain barrier (BBB) permeability, low solubility, rapid systemic clearance, and fast metabolism. For example, rapamycin (RAPA), a primary small-molecule inhibitor of mechanistic target of rapamycin (mTOR), exhibits multimodal benefits including anti-inflammatory activity, preservation of BBB integrity, maintenance of normal cerebral blood flow, and regulation of neuronal autophagy, which collectively contribute to its neuroprotective properties.³ Its capacity to target multiple pathways within the ischemic cascade and potentially extend the therapeutic time window confers significant advantages over many conventional neuroprotective agents in the treatment of IS.⁴ However, its poor aqueous solubility (2.6 μg/mL) and associated side effects have hindered further clinical applications.⁵ Thus, there is an urgent need to develop novel systemic delivery strategies for RAPA in the context of acute ischemic stroke (AIS), which could substantially improve therapeutic efficacy while minimizing drug-related adverse effects.

Advances in nanomedicine have provided various nanodrug delivery systems (NDDS) for CIRI, such as nanoparticles employing liposomes,⁶ mesoporous silica,⁷ cyclodextrin,⁸ or micelles⁹ as carriers. Although these carriers can enhance drug solubility and bioavailability, their clinical translation is often hampered by low drug-loading capacity, complex in vivo metabolic profiles of heterogeneous carriers, and potential toxicity. In recent years, carrier-free NDDS have attracted increasing attention due to their facile preparation process, avoidance of nanocarrier-associated side effects, improved solubility of hydrophobic drugs, and capacity for high drug-loading efficiency.¹⁰ Carrier free NDDS refers to nanoscale formulations formed entirely by self-assembly of active pharmaceutical ingredients, without any inert carrier materials such as lipids, polymers, or silica.¹¹ Natural bioactive drug molecule epigallocatechin-3-gallate (EGCG) is considered one of the most attractive candidates for constructing carrier free nano platforms. This is because EGCG possesses the characteristic of self-polymerization and can participate in nanoparticle assembly as a drug, being a part of nanoparticles.¹¹ Meanwhile, EGCG, the primary bioactive catechin in green tea, exerts anti-inflammatory, antioxidant, anti-apoptotic, and neuroprotective effects attributable to its polyphenolic structure. Studies have confirmed that EGCG exerts neuroprotective effects on the ischemic cerebral cortex during cerebral ischemia by suppressing inflammatory factor levels.¹² EGCG also promotes increased neuronal numbers and recovery of motor function after neural injury.¹³ With its favorable biocompatibility and cost-effectiveness, EGCG emerges as a promising agent for promoting neurobehavioral recovery following CIRI.¹⁴ Notably, current research on EGCG remains largely confined to monotherapeutic or surface-coating applications. The application of EGCG-related nanoparticles in IS deserves further exploration. Therefore, developing innovative EGCG-based nanoplatforms for combined diagnosis and treatment of IS holds considerable scientific and translational promise.

Furthermore, NDDS with imaging capabilities not only enable non-invasive, real-time monitoring of drug distribution in vivo but also allow assessment of therapeutic efficacy through visualization of the lesion area, which is crucial for rationally adjusting treatment plans during therapy to improve outcomes. Currently, some advanced NDDS can utilize multiple imaging techniques to locate lesions, such as positron emission tomography, ultrasound, magnetic resonance imaging (MRI), and near-infrared fluorescence (NIRF) imaging.^15,16 Among these, NIRF imaging offers high sensitivity and fast imaging speed, though its spatial resolution is relatively lower compared to MRI. For instance, lanthanide-doped nanoparticles developed by Zhang et al clearly visualized the ischemic area via NIRF imaging within 30 min after cerebral ischemia, demonstrating a higher sensitivity than MRI.¹⁷ Indocyanine green (ICG), an FDA-approved NIRF agent, has been used in clinical imaging and diagnosis. Recently, the ICG-loaded anti-apoptotic antibody (AntiTRAIL-ICG) has been engineered to monitor the extent of CIRI while simultaneously inhibiting ferroptosis and apoptosis in ischemic neurons.¹⁸ On the other hand, MRI, a cornerstone in stroke diagnosis and management, provides non-irradiating imaging with high spatial resolution.¹⁹ In a recent study, an MRI-guided NDDS (DMS) was designed to target and reduce reactive oxygen species (ROS) production following IS. The incorporation of the MRI contrast agent gadolinium (III) (Gd³⁺) not only enhanced the antioxidant capacity of DMS but also enabled MR imaging of the damaged brain tissue.²⁰ In our previous work, by leveraging the characteristics of the acidic environment in the ischemic brain region, we developed pH-responsive nanoprobes (RAPA/Gd³⁺@NPs) that achieved pH-triggered MR/NIRF imaging of the ischemic area. The results indicated that combining MRI and NIRF could provide more comprehensive information for the early diagnosis and treatment of AIS.¹⁶ Therefore, developing an NDDS capable of dual-modal MRI/NIRF imaging holds significant potential for early diagnosis and precise localization of cerebral infarction in IS patients.

Herein, we leveraged encapsulation properties and the metal chelation ability of EGCG to self-assemble a novel, carrier-free, and highly biocompatible ICG-RAPA-EGCG-Gd³⁺ nanoparticle (IREGd NPs) platform for theranostic application in AIS (Scheme 1). This system integrates the neuroprotective drug RAPA and EGCG, the NIRF imaging agent ICG, and the MRI contrast agent Gd³⁺ into a single nanoplatform. By combining the intrinsic neuroprotective properties of EGCG, such as anti-inflammatory and antioxidant effects, with the incorporated components, the IREGd NPs were designed for integrated diagnosis and treatment of AIS. Overall, the carrier-free IREGd NPs are expected to enhance the drug loading capacity of RAPA and provide a simplified alternative strategy for achieving dual-modal imaging and combined therapy for AIS.

Scheme 1.

Schematic illustration of the preparation and theranostic application of IREGd NPs. (A) Preparation process: free indocyanine green (ICG, Orange spheres), rapamycin (RAPA, green spheres), epigallocatechin-3-gallate (EGCG, red spheres), and gadolinium(III) ions (Gd³⁺, grey spheres) are self-assembled into IRE NPs, followed by chelation with Gd³⁺ to form IREGd NPs. (B) In vivo application: the red dotted circle indicates the tail-vein injection of IREGd NPs into a rat model of transient middle cerebral artery occlusion. The red dotted square highlights an enlarged view of the ischemic brain region, illustrating that IREGd NPs passively accumulate in the infarct area through the disrupted blood-brain barrier. Within this region, the nanoparticles simultaneously enable magnetic resonance imaging (MRI) and near-infrared fluorescence (NIRF) imaging for precise visualization of the ischemic region and drug biodistribution, while exerting combined neuroprotective effects.

Materials and Methods

Materials

RAPA, EGCG, ICG, GdCl₃·6H₂O, and methanol were purchased from Shanghai Aladdin Biochemical Technology Co., Ltd. RPMI 1640 medium and RPMI 1640 medium without glucose and FBS were purchased from Gibco, USA. Anti-Myc mycoplasma clearance agent was purchased from Wuhan Punoise Life Technology Co., Ltd. Trypsin was purchased from Beijing Solarbio Technology Co., Ltd. The CCK-8 assay kit was purchased from Dongren Chemical Technology Co., Ltd. Dimethyl sulfoxide (DMSO), 1× PBS, 0.9% NaCl solution, 4% paraformaldehyde, and anti-fluorescence quenching mounting medium (containing DAPI) were purchased from Biosharp Technology Co., Ltd. Hematoxylin and Eosin (HE) staining solution, TUNEL assay kit, EDTA antigen retrieval buffer, and Anti-NEUN antibody were purchased from Wuhan Savier Technology Co., Ltd. Lactate dehydrogenase (LDH)-1 assay kit and malondialdehyde (MDA) determining assay kits were purchased from Nanjing Jiancheng Bioengineering Institute (Nanjing, China). ELISA kits (IL-6, IL-1β, TNF-α, and TGF-β) were purchased from Shanghai Enzyme Biotechnology Co., Ltd.

All in vitro cellular studies of IREGd NPs were conducted using the rat pheochromocytoma cell line PC12 (highly differentiated) cells purchased from the Shanghai Cell Bank of the Chinese Academy of Sciences. The complete culture medium used for cytotoxicity and cellular uptake assays comprised 89% RPMI 1640 medium, 10% FBS, and 1% mycoplasma clearance agent. The culture medium used for OGD/R was RPMI 1640 medium without glucose.

SPF-grade adult SD rats (age: 8–12 weeks; weight: 280 ± 20 g; equal numbers of males and females) were purchased from Beijing VTLH Laboratory Animal Technology Co., Ltd. Rats were housed in a controlled environment with constant temperature (24°C ± 2°C), humidity, and a 12-hour light-dark cycle, and provided with standard laboratory water and chow. All animal experiments were conducted in accordance with the principles outlined in the Guide for the Care and Use of Laboratory Animals (National Institutes of Health, USA). All procedures followed the ARRIVE (Animal Research in Vivo) guidelines and were approved by the Animal Ethics Committee of Shantou University Medical College (Approval No.: SUMC2023-486). Rats were anesthetized using a mixture of isoflurane (4% concentration for induction, 2.5% concentration for animal modeling and MRI), oxygen, and air (gas flow rate: 0.5 L/min).

Preparation and Characterizations of IREGd NPs

For the preparation of nanoparticles loaded with RAPA, ICG, and Gd³⁺, RAPA, ICG, and EGCG were first dissolved in a molar ratio of 5:1:2 in 0.6 mL of methanol. Then, under vortexing conditions, the IRE methanol solution was added dropwise to deionized water. The solution was collected and transferred to a dialysis tube (MWCO 2500). The dialysis tube was then placed in deionized water to dialyze away unencapsulated RAPA, ICG, EGCG, and methanol. After 24 hours, the collected product was identified as IRE NPs. Next, an equivalent volume of 0.2 mM Gd³⁺ solution was added dropwise under stirring. The mixture was stirred at room temperature for 1 hour, collected, and transferred to a dialysis tube (MWCO 2500). Finally, the dialysis tube was placed in deionized water to dialyze away unencapsulated Gd³⁺. After 24 hours, the collected product was IREGd NPs. The resulting IREGd NPs solution was sterilized by filtration through a 0.45μm membrane before use.

The ICG content in the nanoparticles was determined using a multimode microplate reader (Thermo Scientific Varioskan LUX, Singapore). The RAPA loading in IREGd NPs was measured by LC-MS, while the Gd³⁺ loading was determined by ICP-MS (Agilent 7700). The morphology and size of IREGd NPs were observed using TEM. The particle size distribution and in vitro stability of IREGd NPs were measured using DLS. The surface zeta potential and in vitro stability of IREGd NPs were obtained using a Malvern NanoSizer (ZetaSizer, Malvern Instrument, England). Fluorescence intensity of ICG and IREGd NPs was obtained, and UV absorption peaks of five substances (PBS, EGCG, RAPA, ICG, and IREGd NPs) were determined using a multimode microplate reader.

In Vitro RAPA Release Profile of IREGd NPs

The most commonly used dialysis bag method was used to evaluate RAPA release from IREGd NPs.²¹ Specifically, 2 mL of the IREGd NPs solution was sealed into a dialysis bag with gentle mixing (200 rpm). At the designated time point, 100 μL of buffer solution was withdrawn, followed by replacing it with 100 μL of fresh buffer solution with the same pH. Then, the absorbance of the withdrawn solution and standards was measured at 295 nm using a multimode microplate reader. RAPA concentration was determined via standard curve correlation.

In vitro Study on the Dual-Modality MR/NIRF Imaging Properties of Nanoparticles

The synthesized IREGd NPs were diluted with deionized water to different Gd³⁺ concentrations (0.00834 mM, 0.00417 mM, 0.00209 mM, 0.00104 mM, and 0.00052 mM) for T1-weighted (T1WI) and T1 mapping scans. Subsequently, the longitudinal relaxation time (T1) of IREGd NPs was measured using a 7.0T Agilent MRI scanner (Agilent Technologies, Santa Clara, CA, USA), and the longitudinal relaxation rate (r1) was calculated. The r1 value of the commercially available T1-weighted MRI contrast agent Gd-DOTA (Gadoteric acid) was also assessed under identical conditions (7.0T) for comparison. T1WI scan parameters: TR = 800 ms; TE = 8.64 ms; segments/echo train length = 24/8; average = 3; slice thickness = 2.0 mm; slice spacing = 0.02 mm; Kzero = 1; field of view = 40×40mm; and imaging matrix size = 192×192. Scan duration: 59 seconds. T1 mapping scan parameters: TR = 6000ms; TE = 11.14ms; TI = 40ms, 80ms, 200ms, 500ms, 1000ms, 3000ms, 5000ms; AV = 1; Repetitions = 1; rare factor = 2; slice thickness = 2mm; FOV = 40mm × 40mm; matrix size = 64×64. Scan time: 1 minute 37 seconds.

Near-infrared fluorescence (NIRF) images were acquired using an IVIS Kinetic imaging system (PerkinElmer, USA), and the signal intensities of five substances (PBS, EGCG, RAPA, ICG, and IREGd NPs (4.5μg/mL)) were recorded separately. These intensities were expressed as radiative efficiency, with an excitation wavelength of 745 nm selected to match that of ICG.

In vitro Cellular Studies of IREGd NPs

Cytotoxicity Assessment of IREGd NPs

The biocompatibility of IREGd NPs toward PC12 cells under normoxic conditions in vitro was assessed using the CCK-8 assay. PC12 cells (5 × 10³ cells/well) were inoculated in 96-well plates and cultured for 24 hours. Complete medium was then replaced with complete medium containing different concentrations of IREGd NPs (NPs, 100, 80, 60, 40, 20, 10, 5 µg/mL) and incubated in a normoxic incubator. After 24 hours, the culture medium was discarded, and wells were washed three times with 100 μL/well of PBS. Subsequently, 100 μL/well of complete medium containing 10% CCK-8 was added and incubated for 1 hour. Finally, the absorbance of each well at 450 nm was measured using a multimode microplate reader. The biocompatibility of IREGd NPs with the cells was observed.

In vitro Cellular Uptake Experiments of IREGd NPs

PC12 cells (2×10⁴ cells/well) were seeded into confocal culture dishes and incubated for 24 hours. Then, 3 mL of complete medium containing IREGd NPs or free ICG (with an ICG concentration of 0.46 μg/mL) was added, and the cells were further cultured under normoxic conditions for 2 hours. Subsequently, the cells were washed three times with 3 mL of PBS to remove unabsorbed free drug. Cells were fixed with 4% paraformaldehyde solution for 20 minutes. Finally, anti-fluorescence quenching mounting medium (containing DAPI) was dispensed onto the center of the confocal culture dish. Following fixation, IREGd NPs were excited at 782 nm using a ZEISS inverted laser scanning confocal microscope (LSM880), and emission light was collected between 625–800 nm.

Cellular Protective Effects of IREGd NPs Against OGD/R-Induced Damage in PC12 Cells

The OGD/R experiment is an in vitro simulation of the cellular environment under IS conditions. PC12 cells (5 × 10³ cells/well) were inoculated in 96-well plates and cultured for 24 hours. Then, the cultured PC12 cells underwent a series of steps: three washes with PBS, switching to 100 μL/well of RPMI 1640 medium (glucose- and serum-free), and incubation in a hypoxic environment (1% O₂, 5% CO₂, 94% N₂) at 37°C for 2 hours. Subsequently, the culture medium was replaced with complete medium containing 1.2 µg/mL EGCG, 23.6 µg/mL RAPA, 35.8 µg/mL IREGd NPs, or an equal volume of PBS, and the cells were incubated in a normoxic incubator for 24 hours. Finally, PC12 cell viability following OGD/R treatment was assessed using the CCK-8 assay. To further evaluate OGD/R-induced cytomembrane damage, the culture supernatant was collected from each group 24 h after reoxygenation. LDH leakage, an indicator of cell membrane damage, was measured to assess cytotoxicity using a commercially available LDH assay kit according to the manufacturer’s instructions. Absorbance intensity was quantified at 440 nm. MDA was used to reflect the level of oxidative stress in OGD/R. MDA can condense with thibabituric acid to form a red product with a maximum absorption peak at 532 nm.

MRI and NIRF Imaging of IREGd NPs in the tMCAO Model

The transient middle cerebral artery occlusion (tMCAO) model was established by inducing focal cerebral ischemia through intraluminal occlusion of the middle cerebral artery, followed by reperfusion via the monofilament removal after 2 hours of ischemia (Supporting Information). To evaluate the imaging capability of IREGd NPs in the cerebral infarction region, SD rats subjected to left-sided tMCAO were examined using MRI and NIRF imaging. For this proof-of-concept imaging sub-study, a sample size of n = 4 per group was adopted based on large effect size (significant contrast-to-noise ratio (CNR) between-group differences at all time points) and the 3R principle. Eight rats were randomly assigned into two groups: the tMCAO + IREGd NPs group and the tMCAO + PBS group. Horizontal T1WI was performed before and at multiple time points after intravenous injection of IREGd NPs or PBS (every 2 hours for up to 6 hours) to monitor the enhancement effect and temporal changes in T1WI signal within the infarct area. In addition, tMCAO or sham group rats received the same dose of IREGd NPs via the tail vein; 6 hours post-injection, brain tissues were harvested for NIRF imaging to examine the distribution of fluorescence signals. Furthermore, NEUN staining was conducted on brain sections from the tMCAO + IREGd NPs group to further clarify the localization of the nanoparticles in the brain. The imaging parameters for both MRI and NIRF were consistent with those used for in vitro experiments. The CNR was calculated as: CNR = (S_inf − S_normal) / S.D._noise, where S_inf is the signal intensity in the ischemic region at each time point, S_normal is the signal intensity in the normal tissue, and S.D._noise is the standard deviation of noise.

Neuroprotective Effects and Biosafety Evaluation of IREGd Nanoparticles Following Cerebral Ischemia-Reperfusion

The neuroprotective effects and biosafety of IREGd NPs following cerebral ischemia-reperfusion in tMCAO rats were also evaluated (Supporting Information). Corresponding therapeutic agents (PBS, EGCG, RAPA, or IREGd NPs) were administered via tail vein injection at the onset of reperfusion (2 h after ischemia), and their effects on neurological function and histological changes were assessed post-treatment.

Grouping of the tMCAO Model

Neurological deficits were evaluated using the Zea-Longa scale. Only rats with scores of 1–3 after reperfusion, indicating stable and moderate injury, were enrolled for further study. A total of 65 rats underwent tMCAO surgery, 5 were excluded due to intraoperative death (n = 2) or failed occlusion (Zea-Longa score = 0 or no infarct on T2WI, n = 3). The remaining 60 rats were randomly assigned to the neuroprotective efficacy evaluation (n = 48) or the biosafety assessment (n = 12), using opaque sealed envelopes for allocation concealment. Of the 48 rats in the efficacy group, 24 were used for behavioral tests, imaging, immunofluorescence staining, and biochemical assays, while the other 24 were subjected to Evans blue (EB) extravasation assays. No additional exclusions occurred after randomization, and no deaths were observed during the 24-h observation period following treatment. All experiments were performed under double-blind conditions, and group allocation was revealed only after completion of behavioral and histological analyses.

The rats were divided into four groups: (1) PBS treatment (tMCAO + PBS), (2) EGCG treatment (tMCAO + EGCG), (3) RAPA treatment (tMCAO + RAPA), and (4) IREGd NPs treatment (tMCAO + IREGd NPs). The RAPA dose was standardized to 1 mg/kg based on previous studies; the EGCG dose (49 μg/kg) was chosen to match the EGCG content in the IREGd NPs formulation containing 1 mg of RAPA. Therapeutic efficacy was evaluated 24 h after administration.

Behavioral Assessment of tMCAO Rats after 24-Hour Treatment with IREGd NPs

Based on previously established methods, motor and sensory functional recovery in tMCAO rats across treatment groups was evaluated 24 hours post-treatment using the Zea-Longa score, forelimb placement test induced by tentacles, and foot-fault test (Supporting Information).

Imaging Evaluation of tMCAO Rats 24 Hours after Treatment with IREGd NPs

Twenty-four hours after treatment across the respective groups, the rat brains underwent axial T2-weighted (T2WI) to visualize and quantify the infarct volume in each animal. Additionally, T2 mapping was utilized to quantitatively assess the degree of cerebral edema. Finally, diffusion-weighted imaging (DWI) was performed on all rats, and the apparent diffusion coefficient (ADC) values within the infarct region were measured to evaluate cytotoxic edema. The scanning parameters were as follows:

T2WI: TR = 2000 ms; TE = 8.15 ms; Repetitions = 1; Averages = 3; Slice Thickness = 2 mm; Slice Gap = 0.02 mm; Kzero = 4; FOV = 45 mm × 45 mm; Matrix Size = 192×192. Scan time was 2 minutes 28 seconds.

B0 Map: TR = 40 ms; TE = 3, 3.5, 4 ms; Averages = 6; Slice Thickness = 2 mm; Number of Slices = 1; FOV = 45 mm × 45 mm; Matrix Size = 64 × 64; Flip Angle = 15°. Scan time was 46 seconds.

T2 Mapping: Scans were acquired at the same slice positions as the B0 map. TR = 2000 ms; TE = 8.34 ms; Averages = 1; Number of Echoes (NE) = 16; Slice Thickness = 2 mm; FOV = 45 mm × 45 mm; Matrix Size = 64×64. Scan time was 2 minutes 8 seconds.

DWI: DWI was performed using an EPI sequence with diffusion-sensitizing gradients applied along three orthogonal axes, scanning at the same positions as the B0 map. Amplitude = 24.13 G/cm; Duration = 4.00 ms; Separation = 16.00 ms; Target b-value = 1000 s/mm²; TR = 2000 ms; TE = 24.24 ms; Slice Thickness = 2 mm; FOV = 45 mm × 45 mm; Matrix Size = 96×96. Scan time was 2 minutes 47 seconds.

All image analyses, including infarct volume delineation, T2 mapping, and ADC measurements, were performed by two independent investigators who were blinded to the experimental groups. The final data were averaged from the measurements of the two observers. For infarct volume quantification on T2WI, the hyperintense ischemic region was manually delineated slice-by-slice on consecutive sections. The area of the region of interest (ROI) on each slice was multiplied by the distance between the centers of adjacent slices to obtain the volume per slice, and the total raw infarct volume (V_measured) was calculated by summing the volumes of all slices covering the entire lesion. To correct for post-ischemic edema, which can cause overestimation of infarct volume on T2WI at 24 h post-reperfusion, the edema-corrected infarct volume was calculated using the standard Reglodi method:²²

where V_{contralateral} and V_ipsilateral are the volumes of the contralateral and ipsilateral hemispheres, respectively, obtained by slice-wise summation of hemispheric areas.

For T2 mapping and ADC analyses, ROIs were manually drawn along the margin of the hyperintense ischemic core on the corresponding T2WI/DWI images and then copied to the co-registered T2 maps and ADC maps. The ROI was carefully placed to avoid ventricles and cerebrospinal fluid spaces to minimize partial volume effects. The transverse relaxation time (T2) was automatically fitted and quantified on the 7.0T MR system operator console using the T2-IR/SR method. The ADC value was calculated according to the Stejskal-Tanner equation:²³

where b is the target b-value (1000 s/mm²), S0 is the signal intensity at b = 0, and S is the signal intensity at the target b-value. All parametric maps and quantitative data were processed using MATLAB R2012b (MathWorks, Natick, MA, USA).

For quality control, all raw images were visually inspected before analysis; any slices with significant motion artifacts or signal loss were excluded. Data from a total of 24 animals were included in the final analysis, and no images were excluded due to poor quality.

Immunofluorescence Staining of tMCAO Rats 24 hours after Treatment with IREGd NPs

Glial fibrillary acidic protein (GFAP) staining was performed to evaluate astrocyte activation in the ischemic penumbra. Ionized calcium binding adaptor molecule-1 (Iba1)/CD206 immunofluorescence staining was used to assess microglial M2 polarization. Iba1 served as a pan-microglial marker, while CD206 was employed as a marker of the anti-inflammatory M2 microglial phenotype. NeuN/TUNEL co-localization staining was conducted to specifically label apoptotic neurons. For all staining procedures, nuclei were counterstained with DAPI. Images were acquired using a pathological slide scanner (3DHISTECH Pannoramic, Model: Pannoramic MIDI), and quantitative analysis was performed using CaseViewer 2.3 and ImageJ software. For microglial M2 polarization, Iba1-positive cells were visualized in red, and CD206-positive cells in green. For astrocyte activation, GFAP-positive astrocytes were visualized in red. For neuronal apoptosis, TUNEL-positive apoptotic cells were visualized in green, and NeuN-positive neurons in red. Quantification was performed by two independent investigators who were blinded to the experimental groups; for each rat, five random high-power fields in the ischemic penumbra were counted. Detailed experimental procedures are provided in the Supporting Information.

Quantification of MDA, ROS, and Cytokines in the Ischemic Cerebral Cortex of Rats

After 24 hours of treatment in the respective therapeutic groups, ischemic cortical tissues were extracted from tMCAO rats. A 10% homogenate was prepared in saline under ice-cold conditions and centrifuged at 4000 rpm for 10 minutes to collect the supernatant. Protein concentration was determined using the BCA method. The quantification of MDA is the same as in In Vitro Cellular Studies of IREGd NPs. ROS levels were measured using a DHE fluorescence method with excitation at 520 nm and emission at 605 nm. Pro-inflammatory cytokines (IL-1β, IL-6, TNF-α) and the anti-inflammatory cytokine (TGF-β1) in the samples were analyzed using enzyme-linked immunosorbent assay kits according to the manufacturer’s instructions.

Evaluation of Blood-Brain Barrier (BBB) Permeability by EB Extravasation Assay

BBB integrity was assessed by EB dye extravasation. Briefly, 2% EB solution in saline was freshly prepared and administered via the tail vein (4 mL/kg) at 24 h after reperfusion in each treatment group. After 90 min of circulation, rats were deeply anesthetized and transcardially perfused with saline to remove intravascular dye. The ischemic cortical tissue was harvested, weighed, and homogenized in dimethylformamide (100 mg tissue per 1 mL dimethylformamide), followed by incubation at 60°C for 24 h. After centrifugation (12,000 rpm, 10 min), the supernatant was collected and the absorbance at 630 nm was measured using a multimode microplate reader.

Biocompatibility Evaluation of Nanoparticles

To assess the biocompatibility of IREGd NPs in tMCAO rats, H&E staining was performed on sections of the heart, liver, spleen, lungs, and kidneys from rats treated with either PBS or IREGd NPs after one week of reperfusion. Tissue pretreatment at this stage was consistent with that used for TUNEL staining. After deparaffinization, rehydration, H&E staining, dehydration, and mounting of the paraffin-embedded tissue sections, images were acquired using a pathological slide scanner to examine potential differences in histomorphological features of various organs among the different groups. Beyond histomorphological examination, serum biochemical parameters (alanine aminotransferase (ALT), aspartate aminotransferase (AST), and creatinine (Cr)) and complete blood counts were measured at 7 days post-injection to further evaluate organ function.

Statistical Analysis

Statistical analyses were performed using SPSS Statistics, version 27.0 (IBM Corporation, USA). For continuous data with normal distribution (confirmed by the Shapiro–Wilk test and Q-Q plots) and homogeneity of variances (Bartlett’s test), comparisons between two groups were performed using unpaired Student’s t-test, while comparisons among three or more groups were performed using one-way analysis of variance (ANOVA) followed by Tukey’s HSD post-hoc test. Specifically, the CNR data were analyzed using two-way repeated measures ANOVA, followed by Bonferroni post-hoc test for multiple comparisons. Data are presented as mean ± SD. For ordinal data (Zea-Longa scores), count data with limited range (Forelimb placing scores) and continuous data violating normality assumptions, Kruskal–Wallis H-test followed by Dunn’s test with Bonferroni correction was applied. These data are presented as median [interquartile range]. Effect sizes were calculated as eta-squared (η²) for ANOVA, Cohen’s d (d) for pairwise comparisons and rank-biserial correlation (r) for non-parametric tests. Exact p values are reported unless p<0.001. All cellular assays were performed with three biological replicates, each including technical replicates (CCK-8: 5; LDH/MDA: 3) that were averaged before statistical analysis. For all biochemical assays in in vivo studies, biological replicates (individual rats, n = 6 per group) were analyzed in technical triplicate; data from technical replicates were averaged prior to statistical analysis. A p-value < 0.05 was considered statistically significant. Detailed statistical results for every assay, including exact p-values, effect sizes, and 95% CI, are provided in Tables S1 and S2.

Results and Discussion

Efficient RAPA Loading by IREGd NPs

The IREGd NPs were successfully synthesized via self-assembly and exhibited a spherical morphology with uniform size distribution, as confirmed by transmission electron microscopy (TEM) (Figure 1A). Quantitative TEM analysis of over 100 particles revealed a mean core diameter of 25.68 ± 4.07 nm (Figure S1A). Dynamic light scattering (DLS) measurements revealed an average hydrodynamic diameter of ca. 59 nm (Figure 1B) and a zeta potential of around –23 mV (Figure S1B), indicating favorable colloidal stability. The observed size difference between TEM and DLS mainly arises from the fundamental differences between the two techniques. The relatively small size of the NPs is advantageous for bypassing immune clearance, prolonging blood circulation, and enhancing penetration through the compromised BBB, particularly under ischemic conditions where vascular permeability increases. Previous studies have shown that nanoparticles below 100 nm can more effectively extravasate and accumulate in ischemic brain regions,²⁴ supported by the presence of endothelial pores exceeding 100 nm in diameter under pathological conditions.²⁵

Figure 1.

Characterization of IREGd NPs. (A) TEM image of IREGd NPs in PBS (pH=7.4). (Inset (a) is a photograph of IREGd NPs) (B) Hydrodynamic size distribution of IREGd NPs in PBS measured by DLS. (C) UV-vis absorbance spectra of PBS, EGCG, RAPA, ICG, and IREGd NPs. (D) Fluorescence spectra of ICG and IREGd NPs. (E) RAPA release profile from IREGd NPs in PBS at 37°C (n = 4). (F) The average particle size of IREGd NPs in PBS over 3 days (n = 3). (G) Zeta potential of IREGd NPs in PBS over 3 days (n = 3). (H) PDI of IREGd NPs in PBS over 3 days (n = 3). Data are expressed as the mean ± SD.

Abbreviations: TEM, transmission electron microscopy; PBS, phosphate-buffered saline; DLS, dynamic light scattering; EGCG, epigallocatechin-3-gallate; RAPA, rapamycin; ICG, indocyanine green; PDI, polydispersity index.

The assembly of IREGd NPs is attributed to multiple intermolecular interactions, including hydrogen bonding and π–π stacking among EGCG, RAPA, and ICG constituents.²⁶ In addition, Gd³⁺ was incorporated through chelation with the ortho-dihydroxyphenyl groups in EGCG, forming stable coordination complexes essential for its function as an MRI contrast agent.²⁷ UV-Vis spectroscopy showed characteristic absorption peaks corresponding to EGCG, RAPA, and ICG, confirming successful incorporation of all components into the nanoparticles (Figure 1C). Similarly, fluorescence spectroscopy demonstrated that the emission spectrum of IREGd NPs coincided with that of ICG (Figure 1D), indicating preserved functionality for NIRF imaging.

RAPA serves as an effective autophagy modulator that promotes neuronal survival under ischemic conditions through appropriate autophagy regulation and demonstrates multimodal benefits,⁹ including anti-inflammatory effects and BBB preservation.³ However, its poor aqueous solubility (2.6 μg/mL) and side effects have limited its clinical application.⁵ Conventional nanocarriers such as liposomes (<10% drug loading capacity, DLC),⁶ cyclodextrins (<30% DLC),⁸ micelles (<5% DLC),⁹ and cell membrane coatings (<20% DLC)²⁸ have shown limited loading efficiency for RAPA. Even our previous pH-sensitive amphiphilic block copolymer system (RAPA@NPs) achieved only about 17 wt% DLC.¹⁶ In contrast, the current EGCG-based carrier-free self-assembly approach significantly enhanced drug loading, with LC-MS analysis revealing a RAPA loading capacity of 65.9 wt% in IREGd NPs. Furthermore, ICP-MS measurements indicated a Gd³⁺ content of 3.7 wt%, enabling effective MRI contrast functionality.

In vitro release studies demonstrated that only 20.63% of RAPA was released from IREGd NPs within 24 h in phosphate-buffered saline (PBS), suggesting minimal premature leakage and good stability under physiological conditions (Figure 1E). Furthermore, the nanoparticles maintained consistent polydispersity index (PDI), zeta potential and size in PBS or 10% serum within 72 h (Figure 1F–H and Figure S2), indicating that EGCG can ensure IREGd NPs maintained favorable stability. This stability is critical for in vivo application, ensuring sustained functionality during circulation and accumulation at the ischemic brain tissue.

In Vitro Dual-Modal MR/NIRF Imaging by IREGd NPs

T1WI and T1 mapping of IREGd NPs at different concentrations were performed using a 7.0T MRI scanner (Figure 2A and B), and the T1 of each sample was measured. Figure 2C shows the linear relationship between the Gd³⁺ concentration in IREGd NPs and the reciprocal of T1 (1 / T1). As the concentration of Gd³⁺ increased, a corresponding enhancement in T1WI signal intensity was observed, accompanied by a progressive decrease in the measured T1 values on the T1 maps. A strong linear correlation was identified between 1/T1 and Gd³⁺ concentration, with a longitudinal relaxation rate (r1) of 40.18 mM⁻¹·s⁻¹ (95% confidence interval (CI): 36.58–43.78 mM⁻¹·s⁻¹) for the IREGd NPs. The linear regression equation was fitted as 1/T1 = 40.18 × [Gd³⁺] + 0.355 (R² = 0.978, n = 3 per concentration, fitted by the least-squares method) (Figure 2C). Compared to our previously reported RAPA-loaded nanocarriers, the IREGd NPs not only exhibited an improved RAPA loading capacity but also demonstrated a significantly enhanced T1 contrast ability, with a 3-fold increase in r1, showing concentration dependence.¹⁶ This enhancement is probably attributable to the chelation of Gd³⁺ on the periphery of EGCG, allowing direct interaction with water molecules, thereby markedly amplifying the signal. Notably, Figure S3 shows that the r1 value of IREGd NPs is better than that of the commercially available T1-weighted MRI contrast agent Gd-DOTA with r1 = 4.95 mM⁻¹·s⁻¹, which was measured under identical conditions. Therefore, IREGd NPs demonstrate high potential as a T1 contrast agent deserving further investigation.

Figure 2.

T1WI, T1 mapping, and NIRF imaging of IREGd NPs. (A and B) T1WI and T1 mapping images of solutions containing different concentrations (mM) of Gd³⁺ derived from IREGd NPs in PBS (pH = 7.4). (C) Linear regression analysis of the longitudinal relaxation rate (r1) of IREGd NPs. The reciprocal of the T1 (1/T1) was plotted against the Gd³⁺ concentration in IREGd NPs at 7.0 T. Data are shown as mean ± SD (n = 3 per concentration). The linear regression equation was fitted by the least-squares method as: 1/T1 = 40.18 × [Gd³⁺] + 0.355 with a coefficient of determination R² = 0.978. The slope represents the longitudinal relaxation rate (r1 = 40.18 mM⁻¹·s⁻¹). The area between the red dots represents the 95% confidence interval for fitting the regression line. (D and E) NIRF image and quantified fluorescence radiant efficiency of PBS, EGCG, RAPA, ICG, and IREGd NPs (4.5 μg/mL) at room temperature.

Abbreviations: T1WI, T1-weighted; NIRF, near-infrared fluorescence; T1, longitudinal relaxation time; PBS, phosphate-buffered saline; EGCG, epigallocatechin-3-gallate; RAPA, rapamycin; ICG, indocyanine green.

In addition to MRI, the NIRF imaging capabilities of IREGd NPs were evaluated. NIRF imaging offers advantages such as high sensitivity, rapid imaging speed, and operational simplicity, which can improve the sensitivity for detecting the ischemic penumbra following AIS onset. ICG, an excellent NIRF imaging agent, has been approved by the FDA for imaging and diagnosis in human diseases.¹⁸ As shown in Figure 2D and E, under 745 nm excitation, significant fluorescent emission was detected only in samples containing free ICG and IREGd NPs, confirming the successful incorporation and retained functionality of ICG within the nanoparticle system. No signal was observed in PBS, EGCG, or RAPA controls, underscoring the specificity and effective loading of ICG in the IREGd NPs.

In Vitro Cellular Studies of IREGd NPs

IREGd NPs Can Be Successfully Internalized by Cells into the Cytoplasm

For the in vitro cellular studies of IREGd NPs, PC12 (highly differentiated) cells were selected. This choice is based on the neuronal differentiation characteristics of PC12 cells and their advantage in maintaining cell line stability.²⁹ In their highly differentiated state, these cells exhibit a spindle-like morphology with multiple elongated processes resembling neuronal axons. Functional synaptic connections can be formed between cells via these neurite-like extensions. Furthermore, PC12 cells possess the ability to synthesize and release neurotransmitters such as dopamine.³⁰ Confocal laser scanning microscopy revealed that after 2 h of incubation, IREGd NPs were effectively internalized and localized within the cytoplasm of PC12 cells, as indicated by distinct green fluorescent signals (Figure 3A), while cell nuclei were counterstained with DAPI, appearing as blue fluorescence. Free ICG alone showed less intracellular fluorescence, confirming that IREGd NPs significantly improved cellular internalization, enhanced effective delivery of RAPA.⁹

Figure 3.

In vitro Cellular Studies of IREGd NPs. (A) Cellular uptake of IREGd NPs or free ICG by PC12 cells, with an ICG concentration of 0.46 μg/mL. Green fluorescence localized in the cytoplasm indicates internalized IREGd NPs or free ICG; nuclei are stained with DAPI (blue). The image surrounded by a red dotted line showed a higher magnification image (Scale bar = 5μm) of the red dotted square area (Scale bar = 20μm). (B) Cytotoxicity assessment of IREGd NPs. Cell viability after 24-hour incubation with gradient concentrations of IREGd NPs. (C) Evaluation of protective effects against OGD/R. Cell viability following OGD/R treatment in the presence of PBS, EGCG, RAPA, or IREGd NPs (at equivalent therapeutic concentrations). Quantify and compare the levels of lactate dehydrogenase (LDH) (D) and malondialdehyde (MDA) (E) in the culture supernatant across groups. Data are expressed as mean ± SD (n = 3 biologically independent experiments per group). Data comparisons are made using ordinary one-way ANOVA with Tukey’s HSD test; ***p<0.001 vs PBS group; ^&&&p<0.001 vs EGCG group; ^#p<0.05, ^###p<0.001 vs RAPA group. Statistical details are provided in Table S1.

Abbreviations: ICG, indocyanine green; PBS, phosphate-buffered saline; EGCG, epigallocatechin-3-gallate; RAPA, rapamycin; OGD/R, oxygen-glucose deprivation/reoxygenation. Group color coding: Blue, PBS; Red, EGCG; Green, RAPA; Purple, IREGd NPs; Black, Control (no OGD/R).

IREGd NPs Exhibit Good Biosafety and Mitigate OGD/R-Induced Cellular Injury

The potential cytotoxicity of IREGd NPs in PC12 cells was evaluated using the CCK-8 assay. No significant cytotoxicity was observed in PC12 cells treated with IREGd NPs at concentrations ranging from 0 to 100 µg/mL for 24 h (Figure 3B), confirming the favorable biosafety profile of the NPs.

To simulate ischemic-reperfusion injury in vitro, an oxygen-glucose deprivation/reoxygenation (OGD/R) model was employed to replicate the oxidative damage that ischemia-reperfusion inflicts on cultured cells.³¹ OGD/R-induced neuronal death is mediated by oxidative stress, where sustained deprivation of glucose and oxygen disrupts mitochondrial function.³² Upon reoxygenation, a massive generation of reactive oxygen species occurs, leading to oxidative stress, lipid peroxidation, DNA damage, and neuronal death.³³ OGD/R-induced cell death can be rescued by antioxidants both in vitro and in vivo. Moreover, these antioxidants can also ameliorate reperfusion-induced neuronal injury.³⁴ EGCG, a natural antioxidant, can help reduce neuronal damage and the upregulation of MMP-9 activity in the hippocampal CA1 and CA2 regions following transient global cerebral ischemia due to its antioxidative properties.³⁵

The viability of PC12 cells post-OGD/R treatment was assessed using the CCK-8 method. Results showed that the IREGd NPs group after OGD/R treatment exhibited a statistically significant difference compared with the PBS group (Figure 3C, Table S1). It is known that glucose and oxygen deprivation for over one hour can impair mitochondrial function.³² In this study, the OGD duration was set at two hours, leading to significant cell death due to oxidative stress. The results demonstrate that the IREGd NPs treatment group exerted a favorable anti-neuronal injury effect. Furthermore, IREGd NPs also showed statistically significant differences compared with either the EGCG group or the RAPA group. These findings suggest that the nanoparticle formulation may enhance therapeutic efficacy, potentially due to improved cellular uptake and the combined effects of the loaded components.³⁶ To confirm the cytoprotective effect of IREGd NPs observed in the CCK-8, LDH leakaged into the culture supernatant was quantified, with LDH serving as an indicator of membrane integrity. As shown in Figure 3D, OGD/R challenge dramatically elevated LDH leakage compared to the normoxic control, indicating severe cytomembrane damage. Treatment with IREGd NPs significantly reduced LDH leakage, which was markedly lower than that of the PBS, EGCG, and RAPA groups (Table S1). Furthermore, the MDA level in the supernatant, a marker of lipid peroxidation, was also significantly decreased by IREGd NPs treatment compared to the PBS, EGCG, and RAPA groups (Figure 3E, Table S1). These biochemical findings are fully consistent with the CCK-8 viability data, collectively demonstrating that IREGd NPs effectively protect PC12 cells against OGD/R-induced necrotic and oxidative injury. Previous studies have confirmed that EGCG can significantly reduce the levels of LDH and MDA after ischemia-reperfusion injury,³⁷ while RAPA can significantly reduce LDH levels,³⁸ thereby promoting cell survival. These results underscore the potential of IREGd NPs as a promising neuroprotective strategy under ischemic conditions.

Dual-Modal MRI/NIRF Nanoparticles for in vivo Visualization of Drug Distribution and Imaging of Ischemic Brain Tissue

The tMCAO rat model was used for in vivo studies of AIS and CIRI. As shown in Figure 4A, T1WI signal intensity in the infarct region of IREGd NPs-injected rats increased progressively over time compared to the pre-injection baseline, with gradually enhanced contrast against the contralateral normal tissue. This signal enhancement is attributed to the EGCG-chelated Gd³⁺ NP surfaces, consistent with a previous report on curcumin-Gd³⁺ NPs for tumor imaging.³⁹ In contrast, PBS-injected rats showed no notable signal changes throughout the observation period. As shown in the Figure 4B, the CNR values in the IREGd NPs group progressively increased over time, whereas the PBS group consistently showed low values. Two-way repeated measures ANOVA revealed that the CNR in the IREGd NPs group was significantly higher than that in the PBS group at 2 h (Cohen’s d = 8.61, p < 0.001), 4 h (Cohen’s d = 12.38, p < 0.001), and 6 h (Cohen’s d = 12.7, p < 0.001) after Bonferroni correction, with large effect sizes. These quantitative data indicate that IREGd NPs achieve progressive accumulation in the ischemic region during the acute phase (0–6 h post-reperfusion), which is likely attributed to the transiently disrupted BBB following cerebral ischemia, representing a classic passive targeting mechanism.

Figure 4.

In vivo MRI and NIRF Imaging of IREGd NPs. (A) MRI: Axial T1WI of the rat brain before and at specified time points (up to 6 hours) after tail vein injection of IREGd NPs or PBS in rats with transient middle cerebral artery occlusion (tMCAO). The T1 contrast of the ischemic area (red dotted circle) gradually increases over time. (B) Time-dependent contrast-to-noise ratio (CNR) changes determined by the T1WI of cerebral ischemic areas in different treatment groups of tMCAO rats. Data are expressed as mean ± SD (n = 4 per group). Data comparisons are made using two-way repeated measures ANOVA, followed by Bonferroni post-hoc test for multiple comparisons; ***p<0.001 vs PBS group. Group color coding: Purple: tMCAO + IREGd NPs; Blue: tMCAO + PBS. (C) Immunofluorescence staining of rat brain sections (NEUN: neuronal marker, red; ICG from nanoparticles, green) demonstrating efficient neuronal internalization of the nanoparticles. Scale bar = 20μm. (D) NIRF Imaging: Ex vivo NIRF images of brains harvested 6 hours after tail vein injection of IREGd NPs in tMCAO or sham-operated rats. Increased fluorescence intensity was observed only in the ischemic hemisphere of tMCAO rats. (E) Quantification of NIRF fluorescence intensity distribution in the left (ischemic) and right (normal) hemispheres across groups. Group color coding: Blue: Sham + IREGd NPs; Purple: tMCAO + IREGd NPs.

Abbreviations: T1WI, T1-weighted; NIRF, near-infrared fluorescence; PBS, phosphate-buffered saline; ICG, indocyanine green.

After MRI scanning, brains were harvested for NIRF imaging. As demonstrated in Figure 4D, only the ischemic hemisphere of tMCAO rats exhibited elevated fluorescence intensity, which was 4.2-fold higher than the contralateral side (Figure 4E). No changes in parenchymal fluorescence intensity were observed in the brains of sham-operated rats that also received IREGd NPs injection. This is because IREGd NPs can enter the infarct area for imaging via the BBB disruption caused by CIRI. These results demonstrate the feasibility of using IREGd NPs as a dual-modal MRI/NIRF contrast agent for in vivo imaging of ischemic brain tissue. By combining the high spatial resolution of MRI with the high sensitivity of NIRF imaging, IREGd NPs improve both the accuracy of infarct region identification and the sensitive visualization of drug distribution in vivo.

Given that IREGd NPs are loaded with RAPA drug, their dual-modal MRI/NIRF capability also enables visualization of RAPA distribution within the cerebral ischemic region. To investigate the distribution of IREGd NPs at the neuronal level in tMCAO rats, neuronal nuclei (NEUN) staining was performed to label neuronal cells in rat brains six hours after tail vein injection of IREGd NPs. As illustrated in Figure 4C, similar to the cellular uptake of IREGd NPs observed in PC12 cells in vitro, green fluorescence signals from IREGd NPs were visible in the cytoplasm of neurons within the ischemic left hemisphere. In contrast, no such signal was detected in the neuronal cytoplasm of the right (normal) brain hemisphere. This indicates that IREGd NPs can be effectively internalized by neurons in the ischemic region, released into the cytoplasm, and thereby realize efficient delivery of RAPA. Thus, the fluorescent properties of ICG within IREGd NPs enable visualizing the distribution of drug-loaded NPs at the cellular level.

Neuroprotective Effects and Biosafety Evaluation of IREGd NPs Following Cerebral Ischemia-Reperfusion

Improvement of Sensorimotor Function and Reduction of Infarct Volume by IREGd NPs in tMCAO Rats

As shown in Figure 5A–C, the IREGd NPs group exhibited significantly lower Zea-Longa scores compared with the PBS group (r = −0.679, p = 0.0026), a reduced percentage of errors with the right forepaw compared with the PBS (Cohen’s d = −5.27, p < 0.001), EGCG (Cohen’s d = −5.16, p < 0.001), and RAPA (Cohen’s d = −2.35, p = 0.02) groups, respectively, and a higher number of correct placements of the right forelimb compared with the PBS (r = 0.801, p = 0.0003), and EGCG (r = 0.624, p = 0.0067) groups, respectively. These results underscore the therapeutic potential of IREGd NPs in promoting the recovery of sensorimotor function after cerebral ischemia-reperfusion.⁴⁰

Figure 5.

Sensorimotor function and imaging evaluation 24 hours after treatment with PBS, EGCG, RAPA, or IREGd NPs administered concurrently with reperfusion in tMCAO rats. Quantitative analysis of the Zea-Longa score (A), percentage of right forepaw errors (B), and number of correct placements of the right forelimb (C). T2WI and quantitative comparison of infarct volume across groups (D and E). T2 mapping images and quantitative comparison of T2 values for assessing cerebral edema (D, F). DWI pseudocolor maps and quantitative comparison of ADC values in the ischemic region (D, G). Data are expressed as median [interquartile range] or mean ± SD (n = 6 biologically independent rats per group). For data comparisons, the Kruskal–Wallis H-test followed by Dunn’s test with Bonferroni correction was applied for Figures (A and C), while ordinary one-way ANOVA with Tukey’s HSD test was used for Figures (B, E, F, and G); **p<0.01, ***p<0.001 vs PBS group; ^&&p<0.01, ^&&&p<0.001 vs EGCG group; ^#p<0.05 vs RAPA group.

Abbreviations: PBS, phosphate-buffered saline; EGCG, epigallocatechin-3-gallate; RAPA, rapamycin; T2WI, T2-weighted; T2, transverse relaxation times; DWI, diffusion-weighted imaging; ADC, apparent diffusion coefficient. Group color coding: Blue, PBS; Red, EGCG; Green, RAPA; Purple, IREGd NPs.

To observe the extent of T2 hyperintensity in the ischemic region and calculate the cerebral infarction volume, T2WI was performed on the brains of tMCAO rats 24 hours post-treatment. As shown in Figure 5D, the infarct volume on T2WI was significantly smaller in the IREGd NPs group, while it was markedly largest in the PBS group. The infarct volume for each tMCAO rat post-treatment was calculated by manually delineating the T2 hyperintense ischemic area slice-by-slice on T2WI images. The inter-group differences obtained from these calculations were consistent with the changes observed in the T2WI images (Figure 5E). The cerebral infarct volume in the IREGd NPs group was significantly smaller than that in the PBS group (Cohen’s d = −6.33, p < 0.001), the EGCG group (Cohen’s d = −5.39, p < 0.001), and the RAPA group (Cohen’s d = −2.19, p = 0.0122).

Anti-Inflammatory Effects and Cerebral Edema Attenuation by IREGd NPs in tMCAO Rats

AIS involves complex pathophysiological processes that can lead to cerebral edema, which was assessed herein using two complementary imaging approaches. T2 mapping is sensitive to changes in tissue water content and can be used for the non-invasive assessment of cerebral edema.^41,42 Quantitative analysis of T2 revealed elevated values in ischemic regions, indicative of edema severity. As shown in Figure 5D and F, the IREGd NPs group shows significantly lower T2 values compared to PBS (Cohen’s d = −5.43, p < 0.001), EGCG (Cohen’s d = −4.01, p < 0.001), and RAPA (Cohen’s d = −2.30, p = 0.0447) groups. To further corroborate these findings, DWI was employed, which offers high sensitivity up to 92% in detecting AIS.⁴³ DWI hyperintensity accompanied by reduced ADC values reflects acute cytotoxic edema.⁴⁴ Quantitative ADC analysis in the ischemic brain region demonstrates significantly higher values in the IREGd NPs group than those in PBS (Cohen’s d = 3.33, p = 0.0003), EGCG (Cohen’s d = 2.88, p = 0.0029), and RAPA (Cohen’s d = 1.40, p = 0.0419) groups (Figure 5G). Together, these imaging results suggest that IREGd NPs substantially alleviate cerebral edema induced by ischemia-reperfusion.

Astrocyte-induced inflammatory responses and microglial polarization play key roles in the pathological alterations of CIRI. Immunofluorescence staining revealed that IREGd NPs treatment was associated with a significant reduction in GFAP⁺ astrocyte proliferation in the ischemic penumbra compared to other groups (Figure 6A–C, Table S2), indicating suppression of reactive astrogliosis. Concurrently, the Iba1/CD206 immunofluorescence staining demonstrated that IREGd NPs treatment correlated with a marked increase in the number of CD206⁺/Iba1⁺ M2-polarized microglia relative to other groups (Figure 6B, D, Table S2). In line with these cellular phenotypes, the IREGd NPs group exhibited significantly lower levels of pro-inflammatory cytokines (IL-1β, IL-6, and TNF-α) and higher levels of the anti-inflammatory cytokine TGF-β1 compared to other groups (Figure 6E–H, Table S2). Collectively, these observations demonstrate an association between IREGd NPs treatment and a shift toward an anti-inflammatory glial phenotype. Notably, the observed phenotypic and cytokine profiles are consistent with the known anti-inflammatory functions of RAPA and EGCG.^45–51 Previous studies have demonstrated that RAPA inhibits astrocyte proliferation and migration through mTORC1 blockade,⁴⁵ suppresses M1 microglial pro-inflammatory factor release,⁴⁶ and promotes M2 polarization to secrete anti-inflammatory cytokines.⁴⁷ Similarly, EGCG has been shown to exert anti-inflammatory effects by inhibiting astrocyte activation and reducing GFAP protein expression,⁴⁸ as well as promoting microglial M2 polarization with increased CD206 expression,⁴⁹ and continuously inhibit NF-κB activation, downregulating IL-1β, IL-6, TNF-α, iNOS, and COX-2 in tMCAO models.¹⁴ Therefore, the reduction in cerebral edema observed in the IREGd NPs group is likely a net consequence of these combined,⁵⁰ well-documented anti-inflammatory actions,⁵¹ although the present data do not directly dissect the individual molecular contributions within this complex in vivo system.

Figure 6.

IREGd NPs suppress astrocyte activation, promote microglial M2 polarization, and attenuate neuroinflammation in tMCAO rats. (A) Representative immunofluorescence images of GFAP (red) in the ischemic penumbra 24 h after treatment with PBS, EGCG, RAPA, or IREGd NPs. Nuclei were counterstained with DAPI (blue). Scale bar = 40μm. (B) Representative immunofluorescence images of Iba1 (red, pan-microglial marker), and CD206 (green, M2 anti-inflammatory phenotype) in the ischemic penumbra 24 h after treatment. Nuclei were counterstained with DAPI (blue). Scale bar = 40μm. (C) Quantitative analysis of GFAP⁺ fluorescence intensity in the ischemic penumbra. (D) Quantitative analysis of CD206⁺/Iba1⁺ cell ratios in each group. (E–H) Levels of pro-inflammatory cytokines IL-1β (E), IL-6 (F), TNF-α (G), and anti-inflammatory cytokine TGF-β1 (H) in ischemic brain tissue homogenates, measured by ELISA 24 h after treatment. Data are expressed as mean ± SD (n = 6 biologically independent rats per group). Data comparisons are made using ordinary one-way ANOVA with Tukey’s HSD test; ***p<0.001 v. PBS group; ^&&&p<0.001 vs EGCG group; ^#p<0.05, ^##p<0.01 vs RAPA group. Statistical details are provided in Table S2.

Abbreviations: tMCAO, transient middle cerebral artery occlusion; PBS, phosphate-buffered saline; EGCG, epigallocatechin-3-gallate; RAPA, rapamycin; GFAP, glial fibrillary acidic protein; Iba1, ionized calcium binding adapter molecule 1; CD206, cluster of differentiation 206; IL-1β, interleukin-1β; IL-6, interleukin-6; TNF-α, tumor necrosis factor-α; TGF-β1, transforming growth factor-β1; ELISA, enzyme-linked immunosorbent assay. Group color coding: Blue, PBS; Red, EGCG; Green, RAPA; Purple, IREGd NPs.

IREGd NPs Reduced Apoptosis in the Ischemic Penumbra of tMCAO Rats

Specific identification of apoptotic neurons through NeuN/TUNEL co-localization staining. The results demonstrated that IREGd NPs treatment significantly reduced the number of NeuN⁺/TUNEL⁺ double positive neurons in the ischemic penumbra compared with the PBS, EGCG, and RAPA groups (Figure 7A and B, Table S2), indicating that IREGd NPs administration was associated with reduced neuronal loss following ischemia-reperfusion. While these histological findings provide direct evidence of neuronal protection at the tissue level, they do not, by themselves, dissect the upstream molecular circuitry. Nevertheless, the observed reduction in neuronal apoptosis is consistent with the well-documented neuroprotective functions of both RAPA and EGCG reported in previous mechanistic studies. For instance, RAPA has been shown to promote neuronal survival via autophagy regulation and to inhibit ischemia-reperfusion-induced apoptosis, as reflected by reduced degenerating neurons and decreased Caspase-3 and Caspase-9 expression alongside increased ULK1, Beclin-1, and LC3 levels.⁵² Moreover, EGCG has been reported to modulate autophagy via the AKT/AMPK/mTOR phosphorylation pathway both in vivo and in vitro models of stroke,⁵³ suppress Caspase-3 activation and its substrate PKCε proteolytic cleavage in hypoxic cells,⁵⁴ and inhibit oxidative stress-initiated apoptosis.^55,56 Additionally, EGCG reduces inducible nitric oxide synthase and COX-2 expression in microglial cells,⁵⁷ and alleviates endoplasmic reticulum stress in cerebral ischemia models.⁵⁶ Thus, the anti-apoptotic outcome observed in the IREGd NPs group likely reflects the effect of these combined pharmacological actions, although the relative contribution of each pathway in the present in vivo setting cannot be quantitatively determined from the current dataset.

Figure 7.

IREGd NPs reduce neuronal apoptosis, attenuate oxidative stress, and preserve blood–brain barrier integrity in tMCAO rats. (A) Representative NeuN/TUNEL co-localization images in the ischemic penumbra 24 h after treatment with PBS, EGCG, RAPA, or IREGd NPs. NeuN-positive neurons are shown in red, TUNEL-positive apoptotic cells in green, and nuclei counterstained with DAPI (blue). Scale bar = 40μm. (B) Quantitative analysis of NeuN⁺/TUNEL⁺ double positive cells per field in the ischemic penumbra. (C–E) The levels of malondialdehyde (MDA), reactive oxygen species (ROS), and Evans blue (EB) in the homogenized ischemic brain tissue supernatant were quantified and compared across groups. Data are expressed as mean ± SD (n = 6 biologically independent rats per group). Data comparisons are made using ordinary one-way ANOVA with Tukey’s HSD test; ***p<0.001 vs PBS group; ^&p<0.05, ^&&&p<0.001 vs EGCG group; ^#p<0.05, ^###p<0.001 vs RAPA group. Statistical details are provided in Table S2.

Abbreviations: tMCAO, transient middle cerebral artery occlusion; PBS, phosphate-buffered saline; EGCG, epigallocatechin-3-gallate; RAPA, rapamycin. Group color coding: Blue, PBS; Red, EGCG; Green, RAPA; Purple, IREGd NPs.

To evaluate the redox status associated with treatment, MDA was quantified in the ischemic tissue. MDA levels in both the EGCG and IREGd NPs groups were significantly lower than those in the PBS group, suggesting a reduction in oxidative damage (Figure 7C, Table S2). In addition to MDA, quantitative detection of ROS levels in ischemic brain tissue revealed that compared with other groups, IREGd NPs significantly attenuated the ROS overproduction induced by ischemia-reperfusion (Figure 7D, Table S2). This direct evidence of ROS scavenging, together with the decreased lipid peroxidation, supports the interpretation that IREGd NPs mitigate oxidative stress in the ischemic penumbra. These biochemical findings are consistent with the established antioxidant properties of EGCG, which is known to inhibit the initiation of free radical-activated apoptosis.⁵⁵ However, as these oxidative markers represent downstream biochemical outcomes, the precise molecular pathways through which IREGd NPs coordinate antioxidant defense remain to be fully elucidated.

To evaluate the effect of IREGd NPs on BBB integrity, an EB extravasation assay was performed. As shown in Figure 7E, pronounced EB leakage (blue staining) was observed in the ischemic hemisphere of PBS-treated tMCAO rats, indicating severe BBB disruption. In contrast, IREGd NPs treatment significantly reduced EB extravasation compared to the PBS, EGCG, and RAPA groups, demonstrating effective preservation of BBB integrity. The protection of BBB integrity by IREGd NPs is likely a secondary consequence of the combined anti-inflammatory and antioxidant effects observed in this study, which collectively alleviate endothelial damage and reduce vascular permeability.

Biocompatibility of IREGd NPs

To evaluate the in vivo biosafety of IREGd NPs, H&E staining was performed on major organ sections (heart, liver, spleen, lung, and kidney) collected from rats one week after treatment with either PBS or IREGd NPs following reperfusion, and serum biochemical parameters (ALT, AST, and Cr) and complete blood counts were simultaneously measured at the same time point. As summarized in Figure S4, no obvious histopathological abnormalities or organ damage were observed in the IREGd NPs group, and all quantitative markers remained comparable to those of the PBS controls, with no statistically significant differences detected for any parameter (p > 0.05). These findings collectively demonstrate a favorable subacute safety profile of the IREGd NPs in vivo.

Limitations

Several limitations should be acknowledged. Firstly, the direct biochemical evidence of IREGd NPs mediated mTOR/autophagy pathway and apoptosis, including LC3-II/I, p62, p-mTOR, p-S6, and cleaved caspase-3, still needs to be fully elucidated. Secondly, the current IREGd NPs may rely on a passive targeting mechanism. Future work needs to focus on introducing active targeting ligands onto this nanoplatform to obtain the active targeting ability of the IREGd NPs. Finally, the long-term in vivo fate of chelated Gd³⁺, potential nephrogenic systemic fibrosis risk, and chronic reticuloendothelial system-induced inflammation require systematic evaluation in dedicated follow-up studies before clinical application.

Conclusions

In this study, novel, carrier-free, and highly biocompatible IREGd NPs were developed via a self-assembly strategy, integrating neuroprotective drug RAPA, EGCG, the NIRF imaging agent ICG, and the MRI contrast agent Gd³⁺ into a single nanoplatform. In vitro studies confirmed their high-efficiency loading of RAPA and excellent dual-modal MRI/NIRF imaging capabilities. The NPs were effectively internalized by cells into the cytoplasm and alleviated cell injury induced by OGD/R. In vivo, IREGd NPs served as an effective dual-modal contrast agent, enabling non-invasive visualization of cerebral ischemia-reperfusion injury by combining the high spatial resolution of MRI with the high sensitivity of NIRF imaging. This capability not only improved the accuracy of infarct delineation but also allowed visualization of drug distribution. Therapeutic assessments in a tMCAO rat model demonstrated that IREGd NPs outperformed individual treatments with EGCG or RAPA alone, yielding significant improvements in sensorimotor functional recovery, reduced infarct volume, attenuated cerebral edema, decreased neuronal apoptosis in the ischemic penumbra, and lowered neuroinflammation and oxidative stress. This nanoplatform provides a simple, efficient, and theranostic strategy for integrated imaging and treatment of AIS, offering substantial potential for clinical translation.

Funding Statement

This study was supported by grants from the National Science Foundation of China [grant number 82020108016, 82471974, 82411540241], the Guangdong Basic and Applied Basic Research Foundation (2023A1515220061; 2024A1515220021), and the STU Scientific Research Initiation Grant (270-09426004).

Data Sharing Statement

The datasets used and/or analysed during the current study are available from the corresponding author Prof. Renhua Wu on reasonable request.

Ethics Approval and Consent to Participate

Procedures that used animals were reviewed and approved by the Animal Ethics Committee of Shantou University Medical College (Approval No.: SUMC2023-486).

Disclosure

The authors declare no competing interests.