Covalent organic framework based cytoprotective therapy after ischemic stroke

Abstract

Cytoprotection has emerged as an effective therapeutic strategy for mitigating brain injury following acute ischemic stroke (AIS). The sulfonylurea receptor 1-transient receptor potential M4 (SUR1-TRPM4) channel plays a pivotal role in brain edema and neuroinflammation. However, the practical use of the inhibitor glyburide (GLB) is hindered by its low bioavailability. Additionally, the elevated reactive oxygen species (ROS) after AIS exacerbate SUR1-TRPM4 activation, contributing to irreversible brain damage. To overcome these challenges, GLB and superoxide dismutase (SOD) were embedded in a covalent organic framework (COF) with a porous structure and great stability. The resulting S/G@COF demonstrated significant improvements in survival and neurological functions. This was achieved by eliminating ROS, preventing neuronal loss and apoptosis, suppressing neuroinflammation, modulating microglia activation, and ameliorating blood-brain barrier (BBB) disruption. Mechanistic investigations revealed that S/G@COF concurrently activated the Wnt/β-catenin signaling pathway while suppressing the upregulation of SUR1-TRPM4. This study underscores the potential of employing multi-target therapy and drug modification in cytoprotective strategies for ischemic stroke.

Graphical abstract

Highlights

• •We synthesized a covalent organic framework (COF) nanoparticle embedded with glyburide and superoxide dismutase.
• •Nanoscale S/G@COF penetrated the BBB and functioned as a cytoprotective agent against multiple aspects of the ischemic cascade.
• •Nanoscale S/G@COF targeting SUR1-TRPM4 channel and ROS exerted synergetic cytoprotection.

1. Introduction

Stroke stands as the leading cause of death and disability among adults worldwide, with acute ischemic stroke (AIS) constituting about 80% of cases [[1], [2], [3]]. The ischemic cascade after AIS is a complex pathophysiological process involving various cell types, mainly including energy failure, ion imbalance and excitotoxicity, oxidative stress, cell death (apoptosis or necrosis), inflammation, and immune response [[4], [5], [6]]. Therapeutic strategies against AIS typically include restoring blood flow by thrombolysis and endovascular thrombectomy, as well as enhancing the resilience of neurons to the ischemic cascade [4,7]. Compared to traditional neuroprotection the term cytoprotection is deemed more suitable as it underscores the safeguarding of all cells within the ischemic unit, encompassing neurons, astrocytes, microglia, and endothelial cells [4]. However, cytoprotective drugs face limitations due to their single-target nature, low solubility, short half-life, and various other factors [8]. Developing more effective cytoprotective strategies that encompass multiple aspects of the ischemic cascade is essential.

The Transient receptor potential M4 (TRPM4) channel, activated by intracellular calcium, is pivotal in the pathophysiology of ischemic stroke [9]. Its excessive activation leads to intracellular water accumulation, resulting in cell swelling and eventual death [[9], [10], [11]]. Following brain ischemia, TRPM4 undergoes significant upregulation, often triggered by factors like reactive oxygen species (ROS) [11,12]. This upregulated TRPM4 binds to sulfonylurea receptor 1 (SUR1), forming a novel SUR1-TRPM4 channel. Its association with SUR1 enhances sensitivity to sulfonylurea drugs like glyburide (GLB). The SUR1-TRPM4 cation channel is not constitutively present in brain tissues but is transcriptionally upregulated de novo under neuropathological conditions [13]. The newly upregulated SUR1-TRPM4 channels are expressed in various cells within the ischemic area, contributing to multiple aspects of the ischemic cascade, including cell swelling and death, neuroinflammation, and blood-brain barrier (BBB) disruption [9,14,15]. These multifaceted injuries, in turn, accelerate the production of ROS and promote the unchecked opening of the SUR1-TRPM4 channel, forming a vicious cycle and eventually leading to irreversible damage to the brain tissue. Therefore, it is reasonable to suggest that the simultaneous blockage of the SUR1-TRPM4 and elimination of ROS would effectively disrupt the ischemic cascade and provide synergetic cytoprotection [12].

Glyburide (GLB) is an FDA-approved sulfonylurea that has been proven to have cytoprotective properties in various neurological diseases, such as ischemic stroke, subarachnoid hemorrhage, and traumatic brain injury, mainly by its robust effect in alleviating brain edema and neuroinflammation by selectively targeting the SUR1-TRPM4 channel [16,17]. However, GLB is challenging to penetrate the complete BBB and is quickly cleared out of the brain [18,19], and frequent administration may cause side effects such as hypoglycemia [20]. Clearance of ROS is conducive to reducing oxidative stress damage and secondary ischemic injury and may also aid in inhibiting the up-regulation of the TRPM4. Superoxide dismutase (SOD) is a natural antioxidant enzyme with superior ROS scavenging ability. As the superoxide (O₂⁻) radical is the primary component of ischemic stroke-induced ROS, the co-delivery of GLB and SOD to the ischemic region would be reasonable and efficient to combat the ischemic cascade after an ischemic stroke.

Covalent organic frameworks (COFs) are porous organic polymers constructed from covalent bonds between two or more monomers. This structural feature leads to regular pores with substantial diameters, facilitating the loading of large molecules [[21], [22], [23], [24]]. Profiting from the inherent advantages of nanoparticles, COF can serve as an optimal carrier for GLB through the π-π stacking effect and function as an enzyme nanopocket for SOD. This imparts excellent stability, enabling prolonged presence in the ischemic region, thereby extending retention time and ensuring efficient catalysis of substrates by the enzyme in the ischemic area [23,25].

In this study, we developed a novel nanocomposite SOD/GLB@COF (S/G@COF) designed for the co-delivery of GLB and SOD by COF (Fig. 1). Upon intravenous administration, S/G@COF efficiently penetrated crossed BBB and functioned as a cytoprotective agent against multiple aspects of the ischemic cascade. This led to improved survival and enhanced functional recovery in mice subjected to transient middle cerebral artery occlusion (tMCAO). Mechanistically, the S/G@COF nanoparticle was capable of activating the Wnt/β-catenin signaling pathway and suppressing the up-regulation of SUR1-TRPM4, thereby regulating the BBB function and inflammatory responses.

Fig. 1.

Schematic illustration of the preparation of S/G@COF and its efficacy for cerebral ischemia-reperfusion injury.

2. Material and methods

2.1. Chemicals and materials

2,5-dimethoxyterephthaldehyde (DMTP), 1,3,5-tris (4-aminophenyl) benzene (TBP) and SOD were obtained from Macklin (Shanghai, China). Glyburide was sourced from Solabio (Beijing, China). Dulbecco's Modified Eagle's Medium (DMEM) and fetal bovine serum (FBS) were purchased from Gibco (USA). The Cell Counting Kit-8 (CCK-8) and the one-step TUNEL detection kit were purchased from Beyotime Biotechnology (Shanghai, China). 2,3,5-triphenyltetrazolium chloride (TTC) was purchased from Sangon Biotech (Shanghai, China). The bicinchoninic acid (BCA) kit were purchased from Sigma-Aldrich (St. Louis, MO, USA). The DCFH-DA assay kit and DHE Reactive Oxygen Species Assay Kit were purchased from Beyotime Biotechnology. Antibody against SUR1 (SAB5200528), β-catenin (ABE208), claudin5 (ABT45), and Cleaved Caspase-3 (AB3623) were purchased from Sigma-Aldrich. Antibody against TRPM4 was purchased from Alomone labs (ACC-044, Jerusalem, Israel). Antibody against iNOS (80517-1-RR) was sourced from Proteintech (MI,USA). Antibodies against CD206 (ab64693), Bcl-2 (ab182858), Bax (ab32503), and IBA1 (ab283319) were acquired from Abcam (Cambridge, UK).

2.2. Instrumentation

Transmission electron microscopy (TEM) was conducted using a JEM-2100 electron microscope (JEOL Japan Electronics Co., Ltd., Tokyo, Japan). The powder X-ray diffraction (PXRD) patterns were recorded with an X'pert Powder diffractometer (PANalytical, Holland). X-ray photoelectron spectroscopy (XPS) spectra were recorded under ultrahigh vacuum conditions using an Axis Ultra DLD diffractometer (KRATOS, UK). Dynamic light scattering (DLS) was conducted using the Zetasizer Nano ZS series instrument (Malvern Instruments Co., Ltd, UK) at room temperature. The Nicolet 6700 FT-IR spectrometer (Thermo Fisher, USA) was used to collect Fourier transform infrared spectra (FT-IR). The N2 adsorption/desorption curves were measured by ASAP2460 BET surface meter (Micromeritics, USA). Thermogravimetric analysis (TGA) was measured using a TG209 F3 Tarsus (Netzsch, Germany) at a heating rate of 10 K/min nitrogen protection. The brain blood flow was monitored by a laser Doppler flowmetry (Moor Instruments, Wilmington, USA). The movements of the mice in the Morris water maze were tracked by the TSE VideoMot2 tracking system (Bad Homburg, Germany). Sequential 6 μm-thick coronal sections of the brain were prepared by cryo-ultramicrotomy (CM1950, Leica, Germany). The fluorescence signal was acquired using fluorescence microscopy Axio Imager D2 (Zeiss) or fluorescence microscopy IX73 (Olympus, Japan).

2.3. Preparation of S/G@COF

The S/G@COF was prepared through the following two steps. Firstly, DMTP (984 mg, 2.8 mmol) and TBP (854 mg, 4.4 mmol) were mixed and stirred for 12 h under mild reaction conditions (CH₃CN, 25 °C, 12 h) with the aid of acetic acid (50 mL) and polyvinylpyrrolidone (PVP, M_w = 8000, 1 g). After quenching the reaction by benzaldehyde (40 μL, 0.4 mmol) and centrifugation, the particles were washed with acetonitrile three times to generate the solid TPB-DMTP-COF as a yellow powder.

Secondly, 20 mg TPB-DMTP-COF and 40 mg glyburide were dispersed in 80 mL of deionized water using ultrasonic treatment. Then, 10 mL SOD solution (1 mg/mL) was added to the solution. The mixture was stirred at room temperature for 24 h. Finally, the precipitate was obtained by centrifugation and washed with deionized water three times. A single loading with GLB or SOD yielded G@COF and S@COF, respectively.

Throughout the entire experimental duration, we consistently synthesized the study drugs, S@COF, G@COF, and S/G@COF, on multiple occasions to ensure reproducibility. Each synthesis was accompanied by thorough testing of drug loading efficiency. Notably, we observed a consistent drug loading efficiency of 43 ± 0.5% for GLB in G@COF and 5.9 ± 0.03% for SOD in S@COF across different batches. Similarly, for S/G@COF, the drug loading efficiencies remained stable, with 40 ± 0.5% for GLB and 5.4 ± 0.04% for SOD.

2.4. Primary culture of mouse neurons

The healthy neonatal mice were sterilized with 75% alcohol within 24 h after birth and decapitated under sterile conditions. Brain tissue was removed and placed in a dish containing cold D-Hank's solution. After the cerebellum, hippocampus, and medulla were removed, the cortex was isolated under sterile conditions. Then, the brain cortex was cut into blocks of approximately 1 mm³ using iris scissors, digested in 0.125% trypsin for 20 min at 37 °C, and shaken two to three times. The supernatant was discarded, and complete media was added to terminate digestion. The cell suspension was collected in a new centrifuge tube and centrifuged at 1000 rpm at 4 °C for 10 min. The supernatant was discarded. After being resuspended and filtered, cells were inoculated at a density of 8.0 × 10⁴ to 1.0 × 10⁵ mL^-1 with 500 μL per well into a 24-well plate pre-coated with poly-l-lysine. Cultured for 4–6 h in an incubator at 37 °C with 5% CO₂, the medium was changed to serum-free medium. At day 3 of culture, Ara-C working solution was added to inhibit the over-proliferation of non-neuronal cells and aspirated after 24 h treatment. Cells collected on day 7 of the in vitro culture were used for the experiment.

2.5. Oxygen and glucose deprivation/Re-oxygenation (OGD/R)

Briefly, primary neurons were seeded in culture plates or flasks at an appropriate cell density and adhered for 24 h. Then, the cells were washed with PBS 3 times and incubated with glucose-free DMEM. Subsequently, the cells were placed in a hypoxia chamber containing a mixture of nitrogen and carbon dioxide (95% N₂, 5% CO₂) and temporarily deprived of glucose and oxygen for 4 h in a temperature-controlled incubator. After OGD, the culture medium was replaced with high glucose DMEM, and the cells were returned to the normal incubator for another 24 h to simulate the reperfusion procedure.

2.6. Cytotoxicity

A cell counting kit (CCK-8 kit) was used to detect cell viability. Primary neurons were seeded in a 96-well culture plate at 1 × 10⁴ cells per well and cultured for 24 h to adhere. After each group's treatment procedure, 10 μL CCK-8 solution was added to each well, and the cell culture plate was incubated for 3 h. The absorbance at 450 nm was detected using a plate reader. Blank wells (containing culture media and CCK) and control wells (untreated cells, culture media, and CCK) were also detected.

2.7. Cellular uptake

Primary neurons were seeded in a 96-well culture plate at a density of 1 × 10⁴ cells per well and cultured for 24 h to adhere. Cultured neurons were then incubated with rhodamine-B labeled S/G@COF for 0, 30, and 60 min. After removing the medium, the fluorescence signal was detected by a fluorescence microscope.

2.8. Reactive oxygen species (ROS) detection by DCFH-DA assay

A DCFH-DA based ROS assay kit was used to test the reaction between H₂O₂ and S/G@COF. 20 μM freshly prepared DCFH-DA and 800 μM H₂O₂ were added to 300 μL of S/G@COF solution with varying concentrations and incubated at 37 °C for 30 min. The fluorescence signal of the oxidation product DCF was detected at 535 nm by the multimode microplate reader (Spectramax M5, Molecular Devices, USA) using a 488 nm excitation laser.

2.9. Detection of apoptosis

Cell damage and protection were assessed using a one-step TUNEL apoptosis assay kit (Beyotime, China), according to the manufacturer's protocol. The level of apoptosis was quantified as the number of TUNEL-positive cells/total number of cells per field × 100%.

2.10. Animals

All animal experiments in this study were approved by the Animal Care and Use Committee of Nanfang Hospital (IACUC-LAC-20230421-007), Southern Medical University (Guangzhou, China), and adhered to the National Institutes of Health Guide for the Care and Use of Laboratory Animals. Eight-week-old male C57BL/6 mice, weighing approximately 22 g, were kept in a 12 h light/dark cycle with unrestricted access to food and water. All animal experiments were performed in accordance with the ARRIVE (Animal Research: Reporting in vivo Experiments) guidelines [26].

2.11. Temporary middle cerebral artery occlusion (tMCAO)

The intraluminal filament model of focal ischemia was adopted. Briefly, under isoflurane anesthesia (induced with 3%, maintained with 1%), the right common carotid artery, external, and internal carotid artery were isolated from surrounding nerves and fascia through a midline incision. A monofilament nylon suture with a silicone-coated tip (silicone diameter: 0.22 ± 0.02 mm) was then inserted using arteriotomy of the external carotid artery and gently advanced into the internal carotid artery to the opening of the middle cerebral artery. After 1 h of occlusion, the suture was removed to allow reperfusion for 24 h. A heating pad (RWD Life Science, Shenzhen, China) was used to maintain the rectal temperature at 37 ± 0.2 °C during surgery and the anal temperature was monitored with a non-invasive detector. Regional cerebral blood flow (rCBF) was monitored in all stroke animals by a laser Doppler flowmetry (Moor Instruments, Wilmington, USA). Successful modeling of tMCAO was defined as a reduction of rCBF by more than 75% of baseline and complete blood flow recovery after reperfusion. Inadequate reperfusion referred to the failure of returning to more than 50% of initial blood flow after reperfusion. Sham surgery included exposure of the common, external, and internal carotid arteries with all ligations and transection; no occlusion occurred in the sham group. Animals that died or failed to display an rCBF reduction to 25% of baseline were excluded from further analysis. In the tMCAO model, the mortality rate was 16.2% (20 of total 123) and exclusion rate was 14.1% (14 of total 123 for inadequate reperfusion). Successfully modeled mice were randomly assigned to different groups to receive a single intravenous injection at 15 min after reperfusion.

In the first part, successfully modeled mice were randomly assigned to 5 groups: vehicle, GLB, COF, 2.5 μg/kg G@COF, and 5 μg/kg G@COF using Excel-generated random numbers. In animal studies, GLB exerts neuroprotection through a variety of dosing regimens, including 10 μg/kg daily in rats [27], a total amount of 0.6 μg in rats [28], 10 μg daily in mice [29], or an initial dose of 10 μg/kg followed by different maintenance doses in rats or mice [30,31]. Based on the above dosing regimen and to highlight the significance of COF as a drug carrier, we selected 5 μg/kg free GLB as a control in this study. Vehicle control mice were treated with the equivalent volume of saline. Regarding G@COF, the loading efficiency of GLB in G@COF was assessed using the HPLC system, yielding a result of 43 ± 0.5%. Consequently, the dosages administered to experimental mice receiving intravenous injections of G@COF were calculated to contain a dose-equivalent amount of 2.5 μg/kg and 5 μg/kg GLB post tMCAO. After calculation, 5.8 μg/kg and 11.6 μg/kg G@COF were designated as the major experimental groups, representing loading with 2.5 μg/kg and 5 μg/kg GLB, respectively. The relatively high dose of COF (11.6 μg/kg) was set as the control. Overall, the mice in this part received a single intravenous injection of GLB (5 μg/kg; dissolved in DMSO: PBS = 1:19, v/v at a concentration of 0.1 mg/mL), COF (calculated by the equivalent amount of COF in G@COF: 11.6 μg/kg), G@COF (a dose-equivalent amount of 2.5 μg/kg and 5 μg/kg GLB calculated by the loading capacity: 5.8 μg/kg and 11.6 μg/kg, respectively), or equal volumes of saline at 15 min after tMCAO.

In the second part, successfully modeled mice were randomly assigned to 4 groups: COF, S@COF, G@COF, and S/G@COF. The drug loading efficiency of GLB in G@COF and SOD in S@COF was 43 ± 0.5% and 5.9 ± 0.03% respectively. For S/G@COF, the drug loading efficiency of GLB and SOD was 40 ± 0.5% and 5.4 ± 0.04%. The total quantity for each formulation was determined relative to the loading capacity with 5 μg/kg GLB. Consistent with the approach in the first section, 11.6 μg/kg COF served as the control. After weighing the respective powder, an equivalent volume of saline was used to disperse it. The dosage of superoxide dismutase (SOD) was determined based on insights from our in vitro studies, as there is no standardized administration protocol for tMCAO mice. In the initial phase, the in vivo dosage of S/G@COF was carefully assessed for its efficacy in scavenging superoxide anion free radicals.

In the third part, we evaluated the protection of S/G@COF with a delayed administration 6 h after reperfusion. In this part, the dosing of S/G@COF was based on part 2, and vehicle and sham controls were set.

2.12. Measure of infarct volume

At 24 h after ischemia, the whole mice brains were extracted after euthanasia and sectioned into 7 contiguous coronal slices starting from the frontal pole with placement in a brain matrix. Then, all slices were incubated in 1% TTC for 30 min at 37 °C. After correction for edema, corrected infarct volume (%) was calculated by a researcher blinded to group allocation with Image J (NIH, Bethesda, United States) as follows: Corrected infarct volume (%) = [contralateral hemisphere volume - (ipsilateral hemisphere volume - infarct volume)]/contralateral hemisphere volume × 100%.

2.13. Morris water maze

Spatial learning and memory were evaluated in animals 8 days post-tMCAO using the Morris water maze. First, the mice were trained to reach the platform for 5 consecutive days with 4 trials each day. The movement of mice was tracked using the TSE VideoMot2 tracking system (Bad Homburg, Germany) to record the path and time taken to escape from 4 randomly assigned locations. Latency time to locate the hidden platform was evaluated across the groups. Following the acquisition trial, a probe trial was conducted on the subsequent day, during which mice were given 60 s to explore the area after the platform had been removed. The percentage of total time spent by mice in the target quadrant and the number of times they crossed the platform location were recorded and subjected to analysis.

2.14. Neurological test

A grip test and rotarod test were conducted to evaluate motor function. For the grip test, a length of rope (50 cm) was stretched taut between two vertical supports and raised 40 cm from a flat surface. The mouse was placed in the middle of the rope and rated on a scale of 0–5 (scoring scale: 0 = dropped; 1 = hanging on the string by one or both front paws; 2 = same as 1 and attempting to climb onto the rope; 3 = hanging on the rope by one or both hind paws plus one or both hind paws; 4 = hanging on the rope by the front and hind paws plus tail wrapped around the rope; 5 = escape (to the supports)).

The rotarod test was performed to evaluate motor coordination and balance. Mice were trained for 1 week before the induction of the tMCAO model. After 1, 3, and 7 days of ischemia, mice were placed on a rotarod apparatus. The rotating rod has a 3 cm diameter with a non-slippery surface. The rod was 30 cm long and placed at a height of 20 cm from the base. Each mouse was placed on the rod at a speed of 4 rotations per minute (rpm), accelerating to 40 rpm over 300 s. The duration of each mouse on the rod was recorded automatically. Each mouse underwent 3 consecutive trials with an interval of 15 min between each trial. Results were calculated as the average of three trials.

2.15. Nissl staining

After successful establishment of the tMCAO model, mice were perfused via the heart using 4% paraformaldehyde in PBS, and their brains were subsequently extracted. The brain tissues were fixed with 4% paraformaldehyde and dehydrated using 15% and 30% sucrose solutions. Following this, the brain tissues were sliced into 10-μm sections and stained with 0.04% cresyl violet dissolved in acetate buffer for 1 h. Subsequently, the sections were examined under a light microscope. Cell counting was conducted within the ischemic area of tMCAO mice using five sections from each brain tissue.

2.16. Immunofluorescence

Following 24 h of reperfusion, mice were euthanized, and their brains were transcardially perfused with saline. After overnight fixation with 4% PFA, the brains were immersed in 15% and 30% sucrose solutions at 4 °C. Whole brains were then embedded in OCT for the preparation of frozen sections. Coronal sections of the brain, 6 μm thick, were sequentially prepared using cryo-ultramicrotomy (CM1950, Leica, Germany). For co-localization staining, slices were simultaneously incubated at 4 °C overnight with two types of primary antibodies from different species: mouse primary antibody against SUR1 (1:100, Abcam) and rabbit primary antibody against TRPM4 (1:100, Sigma-Aldrich). Then the sections were incubated with secondary antibodies (1:200 goat anti-rabbit IgG H&L Alexa Fluor 594, goat anti-mouse IgG H&L Alexa Fluor 488). Image analysis was conducted using ImageJ software (National Institutes of Health).

2.17. Measurement of gene expression

The mRNA levels of SUR1, TRPM4, Il-1, Il-10, TNFα, and GAPDH were routinely measured by quantitative real-time polymerase chain reaction (qPCR). Briefly, total RNA was isolated using a total RNA/DNA isolation kit (Tiangen, Peking, China) and reverse transcribed to cDNA with PrimeScriptTM RT Master Mix Kit (TAKARA) according to the manufacturer's instructions. qPCR was performed using the SYBR Green master mixes (TAKARA) and Roche LightCycler480 System. Relative changes of mRNA expression were normalized to the level of GAPDH.

The Tested genes and primer sequences were listed as follows:

Abcc8: Forward primer 5′-3’: CATCCGGGTGAGGAGATACG;

Reverse primer 5′-3’: CAGGTTAACGAAGGGCTGCA.

Trpm4: Forward primer 5′-3’: TGATGAGCACACCACGGAGA;

Reverse primer 5′-3’: ATCCGTGCGATCAGACAGC.

Il-1: Forward primer 5′-3’: CCTACTTCAGCATCCTCTACTGG;

Reverse primer 5′-3’: AGGGTTTCTTGAGAAGGGGAC.

Il-10：Forward primer 5′-3’: CCCATTCCTCGTCACGATCTC;

Reverse primer 5′-3’: TCAGACTGGTTTGGGATAGGTTT.

Tnfa: Forward primer 5′-3’: CACTCTGGGTACGTGGGTG;

Reverse primer 5′-3’: CACAGGTGATAATGAGGACAGC.

Ccnd1: Forward primer 5′-3’:GCGTACCCTGACACCAATCTC;

Reverse primer 5′-3’: CTCCTCTTCGCACTTCTGCTC.

Nlrp3: Forward primer 5′-3’: ATTACCCGCCCGAGAAAGG;

Reverse primer 5′-3’: TCGCAGCAAAGATCCACACAG.

Gapdh: Forward primer 5′-3’: AGGTCGGTGTGAACGGATTTG;

Reverse primer 5′-3’: TGTAGACCATGTAGTTGAGGTCA.

2.18. Differentially expressed genes analysis

Following tMCAO, total RNA was extracted from the ischemic region. Transcriptomic analysis was conducted using the HISAT2-Stringtie-DESeq pipeline. Genes showing > 2-fold changes in expression with an adjusted P-value below 0.05 (Padj < 0.05) were identified as differentially expressed genes (DEGs). Enrichment Analysis, utilizing the Gene Ontology Database, was employed to annotate the gene functions of DEGs and identify the pathways in which these genes are implicated.

2.19. Western blotting

Cerebral tissue was resolved in RIPA solution (Beyotime Biotechnology, Shanghai, China) and restored at -80 °C until used. Denatured protein was separated by polyacrylamide gel electrophoresis and transferred to a polyvinylidene fluoride membrane (Millipore, Billerica, United States). The membrane was incubated overnight at 4 °C with antibodies involving anti-SUR1 (1:1000), anti-TRPM4 (1:1000), anti-Cleaved-Caspase-3 (1:1000), anti-Bcl-2 (1:1000), anti-Bax (1:1000), anti-GSK-3β (1:1000), and anti-β-actin (1:10000). Primary antibodies were detected with anti-mouse or anti-rabbit horseradish peroxidase-conjugated secondary antibodies. All signals were detected by enhanced chemiluminescence, followed by quantification with Image J and normalizing to β-actin levels.

2.20. ROS detection in brain tissue

To assess ROS production, the brain was carefully and quickly isolated, cut into 4.0 μm sections and placed on chilled microscope slides. The production of superoxide anions in brain tissue was examined using dihydroethidium (DHE) staining. Briefly, freshly prepared brain sections were incubated with 1 μM DHE for 10 min at room temperature. The fluorescence density was quantified using Image J software.

2.21. Statistical analysis

The mean and standard error of the mean (SEM) were used to express experimental data. Before assessing statistical significance, the Shapiro–Wilk normality test was used to analyze the normality of data. Statistical differences in multiple groups were compared using one-way ANOVA analysis followed by Tukey's multiple comparison test, while two groups were analyzed by unpaired t-tests. The difference in survival rate was analyzed by Kaplan-Meier analysis with the log-rank test. All statistical analyses were conducted using SPSS 25.0, and GraphPad Prism 8.0 was utilized for image creation and processing. Statistics were deemed significant at P < 0.05.

3. Results and discussion

3.1. The characterization of S/G@COF

Firstly, the imine-linked covalent organic frameworks (COF) were synthesized from 1,3,5-tris(4-aminophenyl)benzene (TPB) and 2,5-dimethoxyterephthalaldehyde (DMTP). Then, the superoxide dismutase (SOD) and glyburide (GLB) were co-encapsulated by COF to construct nanocomposite S/G@COF. It can be seen from transmission electron microscopy (TEM) that the COF (Fig. S1) and S/G@COF (Fig. 2A) were spherical nanoparticles with diameters of about 200 nm. The hydrated size of S/G@COF was consistent with the TEM result (Fig. 2B). In addition, the particle sizes of S/G@COF in PBS (pH 7.4, Fig. S2) and normal saline (Fig. S3) were basically unchanged over 72 h, indicating their prominent colloidal stability under physiological conditions. The zeta potentials of COF, S@COF, and S/G@COF were +31.3, +11.1, and +9.2 mV, respectively (Fig. 2C). The positive zeta potential could significantly enhance the interaction between the S/G@COF and the negatively charged BBB. The microcrystalline structures of COF and S/G@COF were confirmed by powder X-ray diffraction (PXRD). As shown in Fig. 2D, COF exhibited one intensive peak at 6.1°. The position of the relevant COF diffraction peaks remained unchanged in S/G@COF, implying that the framework was stable during the syntheses. In the Fourier transform infrared (FT-IR) spectroscopy (Fig. 2E), the C N stretching vibration at 1617 cm^-1 indicated that an imine linkage was present in COF and S/G@COF. After loading GLB, the observed characteristic stretching vibration of -CH at 2900 cm^-1 in S/G@COF demonstrated that the COF host framework successfully encapsulated GLB.

Fig. 2.

The characterization of nano-scale S/G@COF. A. The transmission electron microscopy (TEM) images of the S/G@COF; B. The DLS size profiles of COF, G@COF, and S/G@COF; C. The zeta potential of COF, G@COF, and S/G@COF; D. The PXRD pattern of the COF and S/G@COF; E. FT-IR spectroscopy of the COF and S/G@COF; F. XPS spectrum of C, N, O and other elements in COF and S/G@COF; G-H. The high-resolution spectrum of C element in COF and S/G@COF; I. The enzyme catalytic activity of S/G@COF.

The X-ray photoelectron spectroscopy (XPS) (Fig. 2F) showed that COF and S/G@COF were mainly consistent with C, N and O elements. Fig. 2G and H were the high-resolution spectrum of C elements in COF and S/G@COF, respectively. It was found that the binding energy for N-C O can be observed in S/G@COF but not in COF, which may be attributed to the loading of GLB. The porosity of COF was well demonstrated by the corresponding N₂ adsorption-desorption measurements (Figure S), while the porosity sharply decreased after SOD and GLB encapsulation. The thermogravimetric analysis experiments demonstrated the excellent thermal stability of COF and S/G@COF below 400 °C (Fig. S5). The antioxidant property of S/G@COF was investigated using superoxide anion radicals (O₂⁻) assay kit. In Fig. 2I, the absorbed intensity at 430 nm gradually decreased with the increase of S/G@COF concentration, implying the stable antioxidant activity of SOD. The above results demonstrated the successful construction of S/G@COF.

We quantified the GLB and SOD concentrations using high performance liquid chromatography (HPLC), the drug loading efficiency of GLB and SOD in S/G@COF was 40 ± 0.5%, and 5.4 ± 0.04%. The GLB concentrations in the 1 mg/mL S/G@COF is 0.4 mg/mL. To examine the drug release properties of S/G@COF, we performed the in vitro GLB release experiment in a saline solution containing H₂O₂ using a Liquid chromatography-tandem mass spectrometry (LC-MS/MS) system. S/G@COF released GLB rapidly before 8 h and reached a plateau after 24 h (Fig. S6).

3.2. The protective effects of S/G@COF on neurons with OGD/R

Oxygen glucose deprivation/reperfusion (OGD/R) is widely accepted to simulate ischemia-reperfusion injury in vitro, which produces significant toxicity to cells, leading to cellular apoptosis, oxidative stress, and inflammatory responses [32].The primary neurons isolated from the cerebral cortex were divided into six groups to assess the protective efficacy of S/G@COF (Fig. 3A). After treatment with different concentrations of S/G@COF without OGD/R, cell viability exceeding 80% was observed (Fig. 3B), indicating the negligible cytotoxicity of S/G@COF. Subsequently, primary neurons were exposed to rhodamine-B labeled S/G@COF for 0, 30, and 60 min, revealing a time-dependent increase in red fluorescence intensity (Fig. 3C), highlighting the efficient uptake of S/G@COF.

Fig. 3.

Application of S/G@COF alleviates neuronal apoptosis and scavenges oxygen free radical in vitro OGD model. A. Illustrative image of primary neurons culture; B. The cell viability of primary cortex neurons incubated with S/G@COF of different concentrations for 24 h; C. The uptake of S/G@COF by primary cortex neurons; D-E. Statistical analysis of ROS scavenge tested by DCFH-DA; F. mRNA expression of anti- and pro-inflammatory cytokines with OGD after incubation with different formulations; G-H. Representative images and statistical analysis of the anti- and pro-apoptotic protein content analyzed with Western blotting. I-J. Representative images of cell apoptosis with different formulations analyzed with Annexin/PI staining. K-L. Representative images and statistical analysis of cell apoptosis tested by TUNEL staining; Data are presented as means ± SD, n = 3; *P < 0.05, **P < 0.01, ***P < 0.001. C-casp3: Cleaved-caspase3; S@COF: COF embedded with SOD; G@COF: COF embedded with GLB; /S/G@COF: COF embedded with SOD and GLB. I: Control; II: OGD＋PBS; III: OGD＋COF; IV: OGD＋S/COF; V: OGD＋G@COF; VI: OGD＋S/G@COF.

To assess the ROS scavenging capacity in vitro, neurons were treated with a DCFH-DA probe for ROS detection following various treatments. At different time points after OGD/R, treatment with S/G@COF notably reduced ROS accumulation by 57% compared to PBS treatment, by 26% compared to S@COF, and by 18% compared to G@COF (Fig. 3D–E). Additionally, S/G@COF downregulated the mRNA expression of Abcc8, Trpm4, and typical pro-inflammatory cytokines (iNos, IL-1β, TNFα), while upregulating the expression of typical anti-inflammatory cytokines (IL-4, IL-10) (Fig. 3F). Furthermore, TUNEL assay and Annexin/PI staining suggested that S/G@COF significantly inhibited the OGD/R-induced apoptosis of the primary neurons (Fig. 3G-L) with significant up-regulation of the anti-apoptotic proteins (Bcl2) and down-regulation of the pro-apoptotic proteins (Bax, Caspase-3) (Fig. 3G–H, S7).

Collectively, these results demonstrated that S/G@COF had multifaceted protective effects on attenuating neuronal apoptosis, preventing neuroinflammation, and scavenging oxygen free radicals in an in vitro model of ischemic stroke.

3.3. BBB penetration and bio-safety of the nano-scale S/G@COF

The blood-brain barrier (BBB) consists of several structural components, including endothelial cells with tight junctions, pericytes, astrocytic endfeet, and extracellular matrix components [33,34]. These components work together to tightly regulate the passage of substances from the bloodstream into the brain capillaries. Tight junctions between endothelial cells form a seal that restricts the movement of molecules, while specialized transporter proteins selectively control the entry of essential nutrients into the brain. Astrocytic endfeet surrounding blood vessels help maintain BBB integrity by releasing signaling molecules that influence endothelial cell function and tight junction formation [33,34]. By collaborating through these mechanisms, the BBB establishes a robust barrier, safeguarding the brain from potentially harmful substances while permitting the passage of essential molecules necessary for proper brain function. While the BBB serves as a protective barrier, it also poses a significant challenge in the treatment of neurological disorders. Many drugs are unable to cross the BBB efficiently, limiting their effectiveness in neuroprotection.

However, during conditions such as stroke, the structure of the BBB is partially disrupted, leading to increased permeability to molecules and particles, which can facilitate the entry of nanoparticles without targeted modifications into the brain parenchyma [35]. Initially, the upregulation of caveolin expression in endothelial cells enhances the transcellular transport of nanoparticles, while the subsequent breakdown of tight junctions promotes particle crossing via the paracellular route [36]. This phenomenon has enabled many unmodified nanomedicines to passively accumulate in the brain following stroke [[36], [37], [38], [39]]. The positive zeta potential could significantly enhance the interaction between the S/G@COF and the negatively charged BBB.

To test the capability of S/G@COF in crossing the BBB, S/G@COF was labeled with rhodamine B, an easy-detectable staining fluorescent dye. At different time points post injection (0.5, 1, 3, 6, 9, 24 h), the in vivo fluorescence signals in the brain were recorded and analyzed. Fluorescence images of the back of the tMCAO mice illustrated that the fluorescence intensity in the brain slightly increased at 0.5 h after the injection of S/G@COF. Subsequently, the fluorescence signal peaked at 6 h and gradually declined to a minimum at 24 h (Fig. 4B–C), indicating that S/G@COF accessed the ischemic region in 0.5 h and persisted up to 24 h after a single use. Strikingly, in sham-operated mice with no BBB disruption and brain injury, S/G@COF also accessed the brain, further confirming its capability of BBB penetration (Fig. 4B–C). To provide more direct evidence, we detected the presence of GLB in the collected brain samples using LC-MS/MS. The concentration of GLB in brain tissue at 1 day after reperfusion was 7 ± 3 ng/mL (Fig. S8).

Fig. 4.

S/G@COF rapidly penetrates the intact BBB and is slowly cleared from the ischemic brain. A. Schematic image of the construction of tMCAO and the experimental arrangement. B. In vivo fluorescence imaging of tMCAO-bearing mice after injection of S/G@COF; C The relative fluorescence signal in the brain at different time points in tMCAO-bearing mice after injection of S/G@COF; D. The colocalization of S/G@COF and neurons detected by microscope on 1, 3 day post tMCAO; E. H&E staining of different organs after the injection of S/G@COF on 1, 3, 7 day post tMCAO.

It should be noted that the effectiveness of cytoprotective drugs relies on their successful uptake by neurons, astrocytes, and microglia in the ischemic region. Thus, we detected the uptake of S/G@COF by neuronal cells in the ischemic area using immunofluorescence co-localization on the brain section by neurons marked by NeuN and S/G@COF labeled by rhodamine B. As shown, on day 1 and 3 after injection, rhodamine B-modified S/G@COF continuously remained in neurons in the ischemic region (Fig. 4D), suggesting that S/G@COF could be effectively taken up by neurons in vivo. It should also be noted that some rhodamine B-positive cells co-localized with DAPI were not co-labeled with NeuN, suggesting that S/G@COF may also be taken up by other cells in the brain, such as microglia and astrocytes, and can also persist in these cells for at least 3 days.

The bio-safety of the nano-scale S/G@COF is indispensable in future applications. We further performed histological analyses of the principal organs by H&E staining to test the toxicity of S/G@COF in vivo. On 1, 3, and 7 days after the administration of S/G@COF, H&E staining did not identify detectable toxicity nor tissue damage in major organs isolated from the treated mice, including the heart, liver, spleen, lung, and kidney (Fig. 4E), suggesting that S/G@COF was safe for intravenous administration.

3.4. The in vivo therapeutic effects of G@COF on tMCAO mice

The neuroprotective qualities of GLB have been consistently demonstrated across various neurological diseases. This effect is attributed to its ability to inhibit SUR1-TRPM4 channels, which are highly expressed in all cell types within the ischemic unit [13]. By targeting these channels, GLB emerges as an ideal cytoprotective agent, showcasing its potential as a therapeutic intervention for neuroprotection in diverse neurological conditions. However, in a recently completed GAMES-RP clinical trial, intravenous injection of GLB failed to prevent decompressive craniectomy or improve functional outcomes [40]. An alternative explanation for the observed outcomes could be related to the challenges associated with GLB penetrating the blood-brain barrier (BBB) and its rapid clearance from the brain [18,19]. These factors may hinder its ability to entirely block the SUR1-TRPM4 channel within the central nervous system.

Therefore, we explored the potential of COF as a drug carrier for intravenous delivery of GLB for stroke treatment. Prior to the in vivo experiments, we successfully constructed the tMCAO mice model. In adherence with the STAIR guideline [26], regional cerebral blood flow (rCBF) was monitored before ischemia (baseline), during ischemia, 10 min after reperfusion, and 24 h after reperfusion. The observed reduction in rCBF by more than 75% of baseline following modeling, along with the complete recovery of blood flow after 24 h of reperfusion (Fig. 5A), signifies the successful construction of the tMCAO model. (Fig. 5A). The depicted data reveals that the intravenous administration of G@COF led to a dose-dependent reduction in infarct volume. Notably, at a dose of 5 μg/kg, there was a significant 37% decrease in infarct volume compared to the treatment with free GLB alone (Fig. 5B–C). These findings underscore the dose-dependent therapeutic impact of G@COF, highlighting its potential as an effective drug carrier for enhancing the efficacy of GLB in reducing infarct volume. The assessment of brain edema severity through water content measurements (Fig. 5D) and the evaluation of neurological deficits using neurological scores (Fig. 5E–F) provided additional confirmation of the therapeutic efficacy associated with the administration of 5 μg/kg G@COF. Nissl staining was employed to depict both the morphological and quantitative changes in neurons at 1 day post-treatment. Notably, G/COF administration mitigated the shrinkage of both cytoplasm and nucleus, leading to increased neuronal survival in the cortex (Fig. 5G–H).

Fig. 5.

Intravenous administration of G@COF provides significant neuroprotection in a mouse model of cerebral ischemia. A. Regional cerebral blood flow (rCBF) was monitored by laser speckle contrast imager before and during tMCAO and after reperfusion; B-C. Representative images and statistical analysis of infarct volume analyzed by TTC staining; D. Brain water content after tMCAO treated with different formulations; E. Grip test scores after tMCAO treated with different formulations; F. Rotarod test treated with different formulation after tMCAO; G-H. Neuronal survival in cortex after tMCAO treated with different dosages assessed by Nissl staining; I-K. Representative images and statistical analysis of the proliferation of microglia and polarization in cortex after tMCAO; Data are presented as means ± SD, n = 6; *P < 0.05, **P < 0.01.

Microglia exhibit a remarkable complexity within the central nervous system, stemming from their diverse functions, phenotypic variability, dynamic responses to environmental cues, and interactions with neighboring cells [41]. In our study, we opted to emphasize iNOS as a major marker for microglial activation, given its well-established role in inflammation. The proliferation and activation of microglia were detected by microglia marker (IBA1) and double labeling of pro-inflammatory marker (iNOS). As shown, G@COF significantly decreased the proliferation of microglia and the iNOS positive (iNOS+) microglia in the ischemic region (Fig. 5I–K). Additionally, G@COF demonstrated the ability to alleviated the release of pro-inflammatorty cytokinse (Fig. S9) and suppress astrocyte proliferation (Fig. S10).

These results collectively suggest the potential neuroprotective and anti-inflammatory effects of G@COF in the context of ischemic conditions with a lower dose than free GLB. This may be due to the following merits of the COF carrier: (1) the presence of inner hydrophobic pores with large diameters makes COF an ideal nanocarrier for loading, enhancing the biocompatibility of GLB; (2) the capability of BBB penetration and excellent stability of COF improved the accumulation of GLB, reduced the exposure of GLB to the circulatory system, and thus limited the risk of hypoglycemia; (3) COF is metal-free and highly biocompatible, effectively avoiding the potential toxicity caused by metal-containing nanomaterials.

3.5. Superior therapeutic effects of S/G@COF on tMCAO mice

While G@COF demonstrated a more pronounced therapeutic ability compared to free GLB, it is acknowledged that the efficacy of G@COF was still limited. This limitation may be attributed to the persistently heightened levels of ROS and the inflammatory microenvironment following tMCAO. These factors could potentially induce a secondary injury to surviving neurons and further stimulate the upregulation of SUR1-TRPM4. Recognizing these challenges underscores the complexity of the pathological processes involved and the need for comprehensive strategies to address both the primary and secondary injury mechanisms for optimal therapeutic outcomes. Indeed, superoxide dismutase (SOD) serves as a potent scavenger of reactive oxygen species (ROS), playing a crucial role in reducing ROS release and mitigating oxidative stress and neuroinflammation following cerebral ischemia. However, the practical application of SOD in vivo is constrained by its extremely short half-life. Utilizing the large-diameter internal hydrophobic hole in COF, we designed and constructed S/G@COF loaded with both superoxide dismutase (SOD) and GLB. Our objective was to investigate whether S/G@COF could offer superior cytoprotective therapy in the tMCAO mouse model (Fig. 6A). Compared to G@COF, S/G@COF was embedded with SOD at the same time, which consumed ROS by alternately catalyzing the dismutation of the superoxide (O⁻) radical into ordinary molecular oxygen (O₂) and hydrogen peroxide (H₂O₂). Using an array of well-established histological and behavioral tests, we evaluated in detail the effects of S/G@COF in short-term and long-term neurological function after stroke.

Fig. 6.

Targeted delivery of antioxidant and anti-edema combination therapy via S/G@COF for stroke treatment. A. Schematic image of the construction of tMCAO and the experimental arrangement. B-C. Representative images and statistical analysis of infarct volume analyzed by TTC staining; D. Brain water content after tMCAO treated with different formulations; E. Grip test scores after tMCAO treated with different formulations; F. Rotarod test treated with different formulations after tMCAO; G-I. Morris water maze after tMCAO treated with different formulations; J-K. Representative images and statistical analysis of the neuronal survival in hipppocampus after tMCAO treated with different formulations assessed by Nissl staining. Data are presented as means ± SD, n = 6; *P < 0.05, **P < 0.01.

In the acute phase after tMCAO, S/G@COF exhibited the most remarkable therapeutic ability in all groups labeled by TTC staining (Fig. 6B–C). The corrected infarct volume after S/G@COF treatment was 10.2% on average, which was far better than the S@COF and G@COF treatments, displaying a promising synergetic effect. In addition, the severity of brain edema assessed by water content (Fig. 6D) and neurological deficit evaluated by neurological scores (Fig. 6E–F) further confirmed the therapeutic ability of S/G@COF.

To further evaluate the long-term protection of S/G@COF, mice were subjected to the Morris water maze from day 8 to 12 after tMCAO. Compared with G@COF-treated tMCAO mice, S/G@COF-injected mice showed superior learning and memorial ability, revealing shortened latency in the training trial and increased crossing platform times in the probe trial (Fig. 6G–I). Subsequently, Nissl staining revealed more viable neurons in the hippocampus at 12 days after tMCAO (Fig. 6J–K).

In addition, we conducted additional experiments to evaluate the efficacy of delayed treatment of S/G@COF when administrated 6 h after reperfusion. The results showed that delayed S/G@COF treatment, when administrated 6 h after reperfusion, significantly improved 12-day survival, reduced infarction volume, and improved grip test scores at 24 h, compared to vehicle treatment (Figs. S11–12).

Overall, our data suggested that S/G@COF embedded with both GLB and SOD exhibited potent short-term and long-term neuroprotection after tMCAO, and its efficacy was significantly superior to that of COF embedded with GLB or SOD alone. Male mice are preferred in our study due to their consistent hormonal profiles compared to female mice, which undergo estrous cycles. The neuroprotection of S/G@COF is unknown regarding sex differences, and the exclusion of female mice is a limitation of our study that requires further investigation.

3.6. Cytoprotective effects of S/G@COF on ischemic cascade

As mentioned above, the ischemic cascade is a multifaceted process with a multitude of interconnected pathways featured by cell death (apoptosis or necrosis), neuroinflammation, and BBB disruption [33]. To examine the cytoprotective effects of S/G@COF, ROS scavenging, neuroinflammation, and BBB integrity in brain tissue were carefully studied.

DHE probe was utilized to quantify the ROS accumulation in the hippocampus. S/G@COF treatment significantly scavenged ROS accumulation, and the effect was better than that of S@COF and G@COF treatments (Fig. 7A–C). As the resident macrophages of the brain, microglia can monitor the microenvironment and initiate immune responses [42,43]. Immunomodulatory molecules produced by microglia, such as cytokines and chemokines, are closely associated with secondary brain damage or repair, respectively [42]. By using iNOS and CD206 together with IBA1 to label microglia, respectively, we found that S/G@COF significantly suppressed the overactivation of microglia, and switched the polarization of microglia from pro-inflammatory to anti-inflammatory phenotype (Fig. 7D–F). Prior research has indicated that Glyburide's protective effects in neuroinflammation are primarily linked to its ability to inhibit NLRP3 inflammasome activation and mitigate pro-inflammatory cytokine release [30]. Hence, we detected the mRNA level of Nlrp3 and several inflammatory cytokine (Il-1β, Tnfα, Il-4, Il-10) in ischemic region (Fig. S13). Our findings revealed an increase in anti-inflammatory cytokines and a decrease in pro-inflammatory cytokines following the application of S/G@COF, suggesting a potential mitigation of neuroinflammation.

Fig. 7.

Multifunctional S/G@COF scavenges the ROS accumulation and inhibits the neuroinflammation in lesion regions. A-C. Representative images and statistical analysis of ROS scavenging in hippocampus detected by DHE probe after the S/G@COF injection; D-F. The proliferation and polarization of microglia detected by pro-inflammatory marker (iNOS) and anti-inflammatory marker (CD206); G-H. GFAP staining in lesion regions; Data are presented as means ± SD, n = 6; *P < 0.05. **P < 0.01.

Claudin-5 is the most enriched tight junction protein and its downregulation is regarded as a marker of BBB disruption [33]. As observed, tMCAO significantly induced the downregulation of claudin-5, while S/G@COF but not G@COF restored this tendency (Fig. S14). To further evaluate the effect of S/G@COF on BBB integrity, Evans-blue leakage assay revealed that the treatment markedly reduced BBB leakage by 54% (Fig. S15). The above protection of S/G@COF treatment resulted in less astrocyte proliferation (Fig. 7G–H) in the ischemic region and less dendritic damage (Fig. S16) after tMCAO.

Together, these results presented the robust cytoprotective effects of S/G@COF on the ischemia cascade. Employing SOD to scavenge excessive ROS not only attenuated the oxidative stress injury but also cooperate with GLB to reduce downstream neuronal apoptosis, neuroinflammation, and BBB disruption.

3.7. Probable molecular mechanisms of cytoprotective effects of S/G@COF on tMCAO mice

As a promising cytoprotective therapy, S/G@COF exerted significant protection on two major fronts. First, S/G@COF enhanced the delivery of GLB and SOD to the brain without significant toxicity; Second, S/G@COF targeted multiple aspects of the ischemic cascade by preventing brain edema, oxidative stress, neuroinflammation, BBB disruption, and neuronal apoptosis, leading to improved survival and neurological functions after tMCAO. We thus explored the possible molecular mechanisms of cytoprotective effects of S/G@COF after ischemic stroke.

As previously mentioned, the TRPM4 channel is a non-selective cation channel activated by intracellular calcium (Ca²⁺) and permeable to monovalent cations, primarily Na⁺. When co-assembled with SUR1, TRPM4 gains sensitivity to sulfonylurea drugs such as GLB. The upregulation of TRPM4 protein triggered by ROS stimulation and the opening of this channel triggered by SUR1 binding are the two key targets in ischemic stroke therapy. We expected that S/G@COF would simultaneously be effective in reducing the expression and opening TRPM4. In the context of ischemic stroke therapy, the upregulation of TRPM4 protein triggered by reactive oxygen species (ROS) stimulation and the opening of this channel induced by SUR1 binding are critical targets. The expectation was that S/G@COF would be effective in simultaneously reducing the expression and opening of TRPM4. Consequently, the effects of different interventions on the expression of SUR1 and TRPM4 after ischemia and reperfusion were assessed. Consistent with in vitro findings indicating that S/G@COF significantly suppressed the mRNA levels of Abcc8 (the encoding gene of SUR1) and Trpm4, S/G@COF prevented the upregulation of TRPM4, as evidenced by Western blotting analysis. Immunofluorescence further illustrated a reduction in the number of SUR1 and TRPM4 positive cells (Fig. 8A–D). Since the inhibitory effect of GLB on SUR1-TRPM4 has been well-established [44], these results suggest that GLB combined with SOD can concurrently inhibit the expression of SUR1-TRPM4. This combined intervention has the potential to more comprehensively block the upregulated de novo channel in various cells within the ischemic brain. Finally, we explored the potential molecular mechanisms by which SOD plays a synergistic role with GLB in cytoprotection after ischemic stroke, in addition to its inhibition of SUR1-TRPM4 expression and scavenging of ROS. We conducted transcriptomic sequencing to compare differentially expressed genes between S/G@COF and G@COF treatment (Fig. 8E). A total of 321 DEGs (224 up- and 97 down-regulated) were found (Fig. 8F). The top five DEGs were verified by qPCR (Fig. 8G). Of note, the most up-regulated gene Lgr5 has been reported as a proliferative factor with the ability to regulate cell cycle proteins by promoting the activation of Wingless-Type MMTV Integration Site (Wnt) Family activating the Wnt/β-catenin signaling pathway [45]. Then, we verified the active protein level of the Wnt/β-catenin signaling pathway after the S/G@COF application. Results indicated that compared with the G@COF group, the expression of the active β-catenin and glycogen synthase kinase-3β (GSK-3β) was highly increased by over 2 folds in the S/G@COF group, suggesting the prominent activation of Wnt/β-catenin signaling pathway (Fig. 8H–J, Fig. S17). The serine–threonine kinase GSK-3β, which is a component of the adenomatous polyposis coli/axin/GSK-3β complex, plays an important role in the regulation of the phosphorylation and degradation of β-catenin [46]. GSK-3β facilitates the translocation of β-catenin to the nucleus, in which it binds to the transcription factors of the lymphoid enhancer (LEF)/T-cytokine family (TCF) to regulate gene transcription. In addition, we re-screened the transcriptome results concerning Wnt downstream regulators. Cyclin D1 was found up-regulated and verified again by qPCR (Fig. S18). Cylin D1 is a direct target gene of β-catenin in multiple cells. The upregulation of GSK-3β and β-catenin plays an important role in the canonical Wnt/β-catenin pathway. Moreover, in the GO enrichment analysis, the up-regulated DEGs were found enriched in the activation of the Wnt/β-catenin signaling pathway, the regulation of tight junction organization, and the extracellular matrix structural constitute (Fig. 8J), which might be responsible for the superior cytoprotection of the S/G@COF application.

Fig. 8.

Transcriptomic sequencing to explore the underlying mechanisms accounting for the significant therapeutic effects of S/G@COF treatment. A-B. Representative images and statistical analysis of the expression of TRPM4 protein in lesion regions detected by western blot; C-D. Representative images and statistical analysis of the expression of TRPM4 protein in lesion regions detected by microscope. E. Heat map of the transcriptomic sequencing between S/G@COF and G@COF; F. Volcano map of the DEGs between the S/G@COF and G@COF treatment; G. The top five DEGs verification by qPCR; H-I The protein level of Wnt/β-catenin signaling pathway detected by western blot; J. GO analysis of the up-regulated the DEGs; K. The illustrative image of the Wnt/β-catenin signaling pathway; Data are presented as means ± SD, n = 6; *P < 0.05.**P < 0.01.

The Wnt/β-catenin signaling pathway is pivotal in ischemic stroke, exerting neuroprotective effects by inhibiting apoptosis, mitigating oxidative stress, and counteracting excitotoxicity [47,48]. Moreover, it promotes neurogenesis by regulating the proliferation and differentiation of neural stem cells and progenitors, while also facilitating synaptic plasticity, crucial for restoring brain function after stroke-induced synaptic damage [47,48]. Inhibition of the Wnt/β-catenin signaling pathway in ischemic stroke would likely hinder these neuroprotective and repair mechanisms [47]. Further investigation is warranted to determine the potential synergistic effects. Our study showcases the efficacy of S/G@COF in reversing the suppression of Wnt/β-catenin signaling induced by ischemia/hypoxia (Fig. 8K). However, we recognize limitations in our mechanistic study, such as the absence of experiments utilizing inhibitors or gene editing to target the Wnt/β-catenin pathway. Addressing these limitations in future studies will allow us to thoroughly understand the intricate regulation of S/G@COF on the Wnt/β-catenin pathway.

To sum up, these results showed that S/G@COF activated the Wnt/β-catenin signaling pathway and suppressed the up-regulation of SUR1-TRPM4, which was functionally distinct from COF embedded with GLB alone in gene expression profiling, having more of an involvement with tissue homeostasis.

4. Conclusions

In summary, we prepared a COF via a facile synthetic approach under ambient conditions and used it as a dual-modal therapeutic platform with non-covalently encapsulation of GLB and SOD. The nanoscale S/G@COF targeting SUR1-TRPM4 channels and ROS exerts synergetic cytoprotection in preventing neuronal apoptosis, brain edema, neuroinflammation, microglia M1 polarization, and BBB disruption via a simple formulation. Our study introduces a novel avenue for cytoprotective therapy, demonstrating the modification of conventional drugs within a promising framework for the effective treatment of ischemic stroke.

CRediT authorship contribution statement

Yuqin Peng: Writing – original draft. Qingfan Ren: Methodology. Huanrong Ma: Formal analysis. Chuman Lin: Data curation. Mingjia Yu: Conceptualization. Yongchuan Li: Software. Jiancong Chen: Supervision. Haihao Xu: Investigation. Peng Zhao: Resources. Suyue Pan: Writing – review & editing, Funding acquisition. Jia Tao: Supervision, Formal analysis. Kaibin Huang: Writing – review & editing, Funding acquisition.

Declaration of competing interest

The authors declare they have no competing interests.

Appendix A. Supplementary data

The following is the Supplementary data to this article:

Data availability

Data will be made available on request.