Mung bean-derived carbon dots suppress ferroptosis of Schwann cells via the Nrf2/HO-1/GPX4 pathway to promote peripheral nerve repair

Abstract

Schwann cells (SCs) can potentially transform into the repair-related cell phenotype after injury, which can promote nerve repair. Ferroptosis occurs in the SCs of injured tissues, causing damage to the SCs and exacerbating nerve injury. Targeting ferroptosis in SCs is a promising therapeutic strategy for effective repair; however, research on ferroptosis in the peripheral nervous system remains limited. In this study, we generated and characterized novel distinctive carbon dots, mung bean-derived carbon dots (MB-CDs). Our results demonstrated that MB-CDs have the advantages of low toxicity, good biocompatibility, high stability, the specific effect of ferric ions (Fe³⁺) on fluorescence, and antioxidant activity. We demonstrated that MB-CDs promoted functional recovery after peripheral nerve injury (PNI), preventing gastrocnemius atrophy. Further research indicated that MB-CDs boosted the repair-related phenotypes of SCs. We used lipopolysaccharide (LPS) to induce an inflammatory model of SCs and co-cultured them with MB-CDs. Then, we examined the effects of MB-CDs by dividing the cells into four groups: the control group (CTRL), MB-CD treatment group (CDs-SCs), LPS treatment group (LPS-SCs), and LPS and MB-CD treatment group (LPS-CDs). RNA sequencing of LPS-CDs and LPS-SCs indicated that LPS-CDs significantly upregulated heme oxygenase-1 (HO-1) expression. Furthermore, western blotting and immunofluorescence techniques demonstrated that MB-CDs suppressed the ferroptosis of SCs via the Nrf2/HO-1/GPX4 signaling pathway after PNI. Overall, this study further uncovered the connection between ferroptosis and the repair-related phenotypes of SCs, filling this gap in the existing knowledge; accordingly, they may be promising agents for treating PNI.

---

1. Introduction

PNI, one of the most common neurological disorders in clinical practice, affects approximately millions of patients worldwide each year.^1,2 Although the peripheral nervous system has a specific regenerative capacity after injury, functional recovery is often disappointing, with loss of motor and sensory function in the affected nerve area leading to temporary or lifelong disability.^3–5 Currently available strategies for improving the regenerative capacity after PNI include surgical^6,7 and non-surgical treatment methods.^8–12 However, these methods often result in high costs or unsatisfactory results.⁶ Therefore, novel therapies with lower cost-effectiveness and higher efficacy are urgently required.

SCs play an essential role in the repair of nerve injury. Our previous study showed that repair-related phenotypes of SCs promoted nerve regeneration, boosted and maintained by endothelial cell-derived exosomes.¹³ Evidence demonstrated a close link between oxidative stress (OS) in SCs and PNI.^14,15 OS involves a balance between oxidation and antioxidant capacities.¹⁶ Specifically, recent studies have shown that nuclear factor erythroid 2-related factor 2 (Nrf2) played a crucial role in PNI to prevent OS and inflammation damage.^14,17–19 Nrf2 was a primary transcription factor in defense mechanisms against OS.^20–22 HO-1 and glutathione peroxidase 4 (GPX4), the critical downstream target stress-inducible protein of Nrf2, also exerted antioxidant stress effects.^23–26 HO-1, an important inducible stress response protein, converts hemoglobin to CO, Fe²⁺, and biliverdin and reduces the above products to bilirubin, playing antioxidant roles.²⁶ GPX4, the key enzyme involved in the antioxidant defence against ferroptosis, reduces lipid hydroperoxides to alcohols in biological membranes.²⁷ Dysregulation of Nrf2 led to decreased antioxidants and detoxifying enzymes, resulting in an imbalance of redox homeostasis in the cell. Ferroptosis, an iron-dependent and lipid peroxidation-driven cell death cascade, has been implicated in the pathophysiology of PNI, occurring when redox homeostasis was imbalanced.^28,29 Research indicated that excess iron was detrimental to neuronal cells following PNI. This was primarily because elevated ferrous iron levels catalyzed the Fenton reaction, enhancing the production of reactive oxygen species (ROS) and lipid peroxidation.³⁰ These processes culminated in ferroptosis, which in turn led to secondary nerve damage, neuroinflammation, and neuropathic pain, ultimately exacerbating regenerative processes post-injury. The inhibition of ferroptosis can promote the repair of damaged peripheral nerves, reduce mitochondrial damage, and promote the recovery of neurological function.³¹ Guo et al. proved targeting mitochondria-dependent ferroptosis as a protective strategy for retinal ganglion cell injuries in optic neuropathies.³² Yan et al. proved that fibroblast growth factor 21 (FGF21) promoted peripheral nerve repair by inhibiting ferroptosis caused by mitochondrial dysfunction.³³ Dang et al. showed that edaravone ameliorated depressive and anxiety-like behaviors via the Sirt1/Nrf2/HO-1/Gpx4 pathway.³⁴ Consequently, targeting Nrf2-related signaling pathways to inhibit ferroptosis may be a potential therapy for treating PNI. However, the research on ferroptosis in the peripheral nervous system remains limited.

Carbon dots (CDs), a new type of low-toxicity, highly biocompatible fluorescent nanomaterial, have been widely used in biomedicine, environmental monitoring, and clinical diagnosis and treatment.^35,36 Recently, the raw materials for CD synthesis generally could be divided into two major categories, organic compounds and biomass.³⁷ Organic compounds included amino acids,³⁸ urea,³⁹ boric acid,⁴⁰ citric acid,⁴¹ and others. Our previous review revealed that biomass used for CD synthesis had significant advantages, including low cost, environmental friendliness, high biocompatibility, and natural biomaterial-based beneficial properties.³⁷ From a pharmacological point of view, working with crude extracts or fractions was interesting because of the multi-targeting effect of the various phytochemicals present in the plant extract. However, the lack of knowledge of the active compounds, poor stability, and low solubility of the bioactive compounds restricted the development of new products from medicinal plants.⁴² Interestingly, using this medicinal plant with good biological activity to develop CDs has recently been proposed as an alternative to solve these problems.³⁶

Mung bean, as a medicinal and edible homologous food, has been the focus of continuous attention in food and biomedicine.⁴³ Mung bean is rich in anthocyanins, alkaloids, coumarins, vitexins, isovitexins, phytosterols, and other bioactive substances. Researchers have isolated and purified important bioactive substances from mung beans.⁴⁴ For example, the seeds and sprouts of mung beans contain macronutrients such as proteins, peptides, oligosaccharides, and polysaccharides, and micronutrients such as flavonoids, phenolic acids, organic acids, sterols, triterpenes, and aldehydes.^45,46 Previous studies have indicated that extracts of mung beans have antioxidant, neuroprotective, and lipid-lowering substances.^47–49 Consequently, we speculated that MB-CDs could exhibit beneficial biological and pharmacological activities, implying their potential application in peripheral nerve repair.

Based on this, we prepared and characterized MB-CDs. The synthesized MB-CDs exhibited low toxicity, good biocompatibility, high stability, and a specific effect of Fe³⁺ on fluorescence. Moreover, MB-CDs have beneficial pharmacological and antioxidant activities. We subsequently evaluated the effects of MB-CDs on PNI using in vivo and in vitro experiments. These results illustrated that MB-CDs promoted functional recovery and suppressed the ferroptosis of SCs through the Nrf2/HO-1/GPX4 pathway after PNI. Overall, the results reported herein demonstrate that MB-CDs may be promising therapeutic agents for PNI.

2. Materials and methods

2.1 Preparation of MB-CDs

First, mung beans were washed and dried. Then, 4 g of mung beans was weighed and put into a 50 ml reactor liner, and 20 ml of distilled water was added. It was heated at 180 °C for 8 hours, and cooled to room temperature, to obtain a yellow clear solution by centrifugation, and further purified through a filter membrane. The yellow clear solution was dialyzed in a 500 Da dialysis bag for 8 hours. Finally, the dialyzed liquid was freeze-dried to obtain MB-CD solid powder.

2.2 Characterization of the MB-CDs

A fluorescence spectrometer was used to obtain the excitation spectrum of the sample, the emission spectrum at multiple wavelengths, and the lifetime decay curve, so as to evaluate the photoluminescence performance of MB-CDs. The experimental instrument used was an Edinburgh FLS1000 steady-state/transient fluorescence spectrometer. A UV spectrophotometer was used to determine the light absorption properties of the samples by detecting the UV-visible (UV-Vis) absorption spectrum of the samples. The experimental instrument used was a UV2000i spectrometer from Shimadzu Corporation of Japan. Transmission electron microscopy (TEM) was used to directly observe the structural characteristics such as morphology size, particle size distribution, and lattice spacing of the samples. Powder X-ray diffraction (XRD) is used to analyze the phase structure and composition of the samples. Through Fourier transform infrared spectroscopy (FT-IR) technology, the characteristic functional groups and molecular structures of the samples were determined based on the different absorption frequencies corresponding to the stretching vibration and bending vibration modes of atoms in different molecules. X-ray photoelectron spectroscopy (XPS) technology was used to analyze the elemental composition, group content and bonding state of the sample. Owing to the absence of an exact and well-defined chemical structure for carbon dots, it is infeasible to ascribe to them a purity value with the same level of precision and definitiveness as that of a chemically pure substance. Purification methods for carbon dots typically include filtration, dialysis, and column chromatography. In this work, we first employed filtration to separate carbon residues generated during the synthesis process, followed by purification using dialysis bags. Thin-layer chromatography was used to verify the purity of the obtained mung bean carbon dots. Chromatographic results indicated that the carbon dots were pure substances under appropriate polar eluents (Fig. S5†).

2.3 Cell culture

SCs were purchased from Procell Life Science & Technology Co., Ltd. SC cells were cultured in DMEM containing high glucose (Gibco, USA), and were supplemented with 10% FBS and 1% penicillin–streptomycin. The cells were expanded in tissue culture dishes and kept under a humidified atmosphere of 5% CO₂ at 37 °C. The medium was changed every other day. A confluent monolayer was detached with 0.5% trypsin and dissociated into a single-cell suspension for inoculation.

2.4 Cell proliferation and colony-forming assay

The SC and LPS-SC densities were adjusted to 2 × 10⁴ cells per mL and then inoculated into a 96-well plate. Different concentrations of MB-CDs (0, 25, 50, 100 and 200 μg mL⁻¹) were added to each well and were allowed to incubate with the cells for 48 hours. Then, 10 μL cell counting kit-8 (Dojindo, Japan) was added to each well and incubated for 2 hours. Cell proliferation was measured using a full wavelength microplate reader at 450 nm. We also used the CCK8 assay to study the growth curves of different groups of SCs. In addition, the 5-ethynyl-2-deoxyuridine (Edu) cell proliferation assay kit (Ribobio, China) was also used to measure cell proliferation in different groups of SCs. Manually count Edu labeled cells in three randomly selected fields from each well to calculate the percentage. In order to study the cloning ability of different groups of SCs, cells with different treatments were inoculated into a 6-well plate (1000 cells per well).

2.5 Migration assay

To investigate the role of MB-CDs in the migration of SCs, we co-cultured MB-CDs and SCs in a Transwell system. Firstly, the SCs and LPS-SCs were set to 1 × 10⁴ cells (100 μL) respectively, which were cultured in the upper chamber. Then, we added 200 μg ml⁻¹ MB-CDs or PBS into the cell culture medium of SCs or LPS-SCs to estimate the impacts of MB-CDs on SC migration. The procedures are as follows. First, the cells were incubated at 37 °C for 24 hours and the non-migrated cells were removed with a cotton swab. Then, the migrated cells were immobilized with 4% paraformaldehyde (PFA) (Solarbio, China) for 30 minutes. Next, the fixed cells were washed with PBS and then the cells were stained with 0.1% crystal violet staining solution for 30 minutes. After drying for 12 hours, the number of migrated cells were observed and counted.

2.6 Western blot analysis

Extracellular lysates and cell lysates were prepared by adding protease inhibitor cocktails (Yazyme, China) and phosphatase inhibitor cocktails (Yazyme, China) to RIPA lysis buffer (Yazyme, China). The total protein was separated from SDS polyacrylamide gel (Yazyme, China) and transferred to polyvinylidene fluoride (PVDF) membrane (Beyotime, China). Protein-free fast-blocking buffer (Yazyme, China) was used to block the membrane at room temperature for 30 minutes. The blocked membranes were incubated overnight at 4 °C with antibodies specific for the Nerve Growth Factor (NGF) (Abcam, 1 : 1000), Ciliary neurotrophic factor (CNTF) (Abcam, 1 : 1000), Glial cell derived Neurotrophic Factor (GDNF) (Abcam, 1 : 1000), Polymerase Gamma (POLG) (Bioss, 1 : 1000), 8-Oxoguanine DNA Glycosylase (OGG-1) (Bioss, 1 : 1000), GPX4 (Proteintech, 1 : 1000), Nrf2 (Abcam, 1 : 1000), HO-1 (Proteintech, 1 : 1000), β-tubulin (Bioss, 1 : 5000) and β-actin (Proteintech, 1 : 20 000). Then the membranes were washed and incubated with horseradish peroxidase (HRP)-coupled secondary antibodies (Solarbio, China). The blots were detected using an Amersham Imager 600. β-Actin or β-tubulin was used as the loading control, and the interested protein's relative intensity was normalized to that of the control group.

2.7 RNA sequencing

RNA was extracted from LPS-SCs treated with or without MB-CDs for 24 hours using RNAiso Plus (Takara, Japan), and the RNA integrity and total amount were analyzed using a 2100 bioanalyzer (Agilent, CA, USA). An RNA sequencing library was prepared and sequenced on an Illumina HiSeq 6000 (Illumina, CA, USA). The sequencing service was provided by Novogene (Beijing, China). The DESeq2 R software package (1.20.0) was used to analyze differentially expressed genes (DEG) between two groups. In addition, DEGs were screened based on thresholds of P value ≤0.05 and |log2FoldChange| ≥ 1. The Gene Ontology (GO), Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analysis, and Gene Set Enrichment Analysis (GSEA) of DEG were achieved through the ClusterProfiler R package (3.8.1). The datasets of ferroptosis-related genes were downloaded from the FerrDb database, which was assessed on June 20, 2024.⁵⁰

2.8 ROS detection

ROS generation in the cells was qualitatively analyzed using the DCFH-DA Reactive Oxygen Species Assay Kit (us Everbright, China), as per the manufacturer's instruction. ROS levels in nerves were detected using dihydroethylene glycol (DHE) probes. DHE can be oxidized by intracellular ROS, and the oxidation products would be bound to chromosomal DNA to produce fluorescence. In this way, by measuring the green fluorescence in living cells, the amount and variation of ROS content in cells can be determined. After the intervention of MB-CDs, 10 μL of Incubate M's DHE solution was added for 30 minutes in the dark, and then the excess DHE staining solution was washed off. The detection of the intensity and changes of green fluorescence in cells was observed by laser confocal microscopy and flow cytometry.

2.9 Assessment of mitochondrial membrane potential

Mitochondrial membrane potential (MMP, ΔΨ_m) is not only an important indicator of mitochondrial dysfunction, but it is also an early event of cell apoptosis. In this experiment, SC was labeled with JC-1, which is a cation fluorescent dye limited to mitochondria. Therefore, after treating the cells according to the above method, JC-1 was co-incubated with the cells in the dark for 15 minutes. In addition, after cell culture, fluorescence intensity was measured under a laser confocal microscope.

2.10 Cellular immunofluorescence

SCs were inoculated into a culture dish pre-placed with treated cover glass. Once the cells were close to growing into a monolayer, the cover glass was removed and washed twice with PBS. Then the slide was fixed with 4% PFA for 15 minutes, and sealed with room temperature antigen for 1 hour. Then, a sufficient amount of GPX4 (Abcam 1 : 200) was added to each slide and placed in a wet box, and incubated overnight at 4 °C. The secondary antibody used was Cora Lite 488-conjugated sheet anti rabbit IgG. The nucleus was stabilized with DAPI and the sections were observed under a confocal laser scanning microscope.

2.11 Animal model and delivery of MB-CDs

Adult male Sprague Dawley rats (300–400 g) were obtained from Vital River Experimental Animal Technology Co., Ltd (Beijing, China). The living conditions and experimental procedures complied with the guidelines of the National Institutes of Health. In addition, all animal experiments were approved by the Animal Experiment Ethics Committee of Zhengzhou University. Rats were raised with free access to food and water and under the conditions of constant temperature (23–24 °C), constant humidity (55 ± 5%) and 12 hours of light–dark cycle. The rats were randomly divided into three groups (n = 5 each group): the sham surgery group, PNI group, and MB-CD treatment group (PNI-CDs). After effective inhalation of ether, the rats were intraperitoneally injected with pentobarbital sodium (2 mL kg⁻¹, 2%). Then, the right nerve of the rat was exposed using the gluteal muscle dissection method. Firstly, Dumont No.5 forceps were used to establish a PNI model, clamping the nerve at the sciatic incision three times, 10 seconds each time with an interval of 10 seconds. By clamping the nerve, a 2 mm long translucent band was formed at the injury site and marked with 10-0 nylon sutures for future identification. Next, for the PNI-CD group, a mini syringe was used to inject 20 points under the outer membrane of the sciatic nerve with 10 μL of 50 μg mL⁻¹ MB-CDs. After each injection, the needle was left in place for 30 seconds to prevent leakage. Finally, the rats in the sham surgery group underwent the same surgery without any sciatic nerve damage.

2.12 Walking track analysis

In order to evaluate the motor function after nerve injury, walking trajectory analysis was conducted on the injured model rats 3 days before surgery, and 7, 14, and 28 days after surgery. In this experiment, Eosin Y solution (Solarbio, China) was applied to the surface of the soles of both hind paws of the rats, allowing them to walk along a narrow corridor with white paper at the bottom and a dark compartment at the end. The paw length (PL), the toe-spread distance between toes 1 and 5 (TS), and the toe-spread distance (IT) between toes 2 and 4 were recorded from the normal (N) and experimental (E) hind limbs. The sciatic functional index (SFI) was calculated using the following formula:

According to the analysis of walking trajectory, rat footprint measurement can evaluate the functional muscle state of the hind limbs. Usually, an SFI value of 0 indicates normal neurological function, while an SFI value of −100 indicates complete loss of motor function.

2.13 Histomorphological analysis of regenerative nerves

28 days after surgery, the sciatic nerve was removed, fixed overnight in 4% PFA, dehydrated in gradient-grade ethanol, and embedded in paraffin. After paraffin embedding, longitudinal and transverse sectioning (5 μm) was done and wax removal and hydration were performed. According to the manufacturer's instructions, the nerve slices were stained with hematoxylin eosin (HE) and Masson staining. Finally, the slide was fixed with neutral resin and covered. Images of the stained sections were obtained using an optical microscope (Olympus, Japan).

2.14 Histological assessment of muscle

On the 28th day after surgery, bilateral gastrocnemius muscles of rats were harvested and quickly weighed, and the weight ratio of the ipsilateral muscles was calculated to obtain the relative wet-weight ratio of the muscles. Then the experimental gastrocnemius abdominis was fixed, embedded in paraffin, and stained with H&E and Masson staining. Finally, representative images of the stained sections were observed using an optical microscope.

2.15 Immunofluorescence staining and immunofluorescence evaluation

28 days after surgery, nerve tissue was harvested and fixed with 4% PFA. Then longitudinal and transverse sections of neural tissue were prepared. In addition, S100β (Abcam, 1 : 200) was used for slicing and staining. The secondary antibodies used were: Alexa Fluor568-conjugated goat anti-rabbit IgG (Abcam), CoraLite594-conjugated sheep anti-mouse IgG (Proteintech, China), Cora Lite488-conjugated sheep anti-rabbit IgG, CY3 labeled goat anti-rabbit IgG (Servicebio, China), and Alexa Fluor594-conjugated goat anti-rabbit IgG (Abcam). In addition, we also used FITC Tyramide (Servicebio, China) and CY3 Tyramide to amplify the fluorescence intensity. The nucleus was stained with DAPI and the sections were observed under a confocal laser scanning microscope. The percentage of positive areas marked was calculated by dividing the density of the comprehensive options by the area of the selected area, and then multiplying by 100. All parameters were measured using ImageJ.

2.16 Statistical analysis

Statistical analysis was performed using GraphPad Prism 8.0 (GraphPad Software, La Jolla, CA, USA). The results were expressed as mean ± standard deviation. Single-factor analysis of variance was used for comparison within multiple groups, and the two-tailed unpaired student test was used for comparison between two groups. P values <0.05 were considered statistically significant.

3. Results and discussion

3.1 Preparation and characterization of MB-CDs

In this study, MB-CDs with blue emissions were prepared via a one-step hydrothermal reaction using mung beans as the carbon source. The properties of carbon dots can generally be categorized into two main aspects: optical properties and structural characteristics. We first examined the optical properties of carbon dots, including absorption, fluorescence emission, time-resolved spectroscopy, and photostability. At different excitation wavelengths, the MB-CDs demonstrated a clear excitation dependence (Fig. 1a). At the optimized excitation wavelength of 380 nm, the maximum emission wavelength of the MB-CDs was 462 nm (Fig. 1b). The UV-Vis absorption spectra of the MB-CDs exhibited three absorption bands at 225, 273, and 325 nm. The absorption peaks at 225 and 273 nm were assigned to π–π* transitions of C=C/C–C, and the absorption peak at 325 nm was classified as an n–π* transition between the oxygen/nitrogen moiety and the sp² structural domain. The time decay curves were obtained by two-photon fitting, and the average lifetime of the MB-CDs was 5.9 ns (Fig. 1c). Interestingly, the quantum yield of the MB-CDs was as high as 18.76%, which is currently the highest among the biomass carbon-dot-based nanomaterials. Subsequently, we investigated the optical stability of the MB-CDs. As displayed in Fig. 1d and f, the fluorescence intensity decreased slightly after continuous irradiation for 8 h in the emission spectra of MB-CDs under different illumination times. The fluorescence intensity of MB-CDs remained invariable due to the fluorescence intensity after storage at room temperature for three months (Fig. 1e and f). Overall, the above results indicated that MB-CDs we generated had high stability, which was critical for ion detection and biological applications.

Fig. 1 High stability of MB-CDs. (a) Fluorescence spectra of MB-CDs at different excitation wavelengths. (b) The UV–vis spectrum, the maximum excitation spectrum and the maximum emission spectrum of MB-CDs. (c) Fluorescence lifetime decay curves of MB-CDs. (d) Fluorescence emission spectra of MB-CDs under continuous UV light irradiation (0–8 h). (e) Fluorescence emission spectra of MB-CDs under different storage times (1, 2, 7, 14, 21, 30, 60, and 90 d). (f) Change curve of emission intensity (I/I₀) under different UV lamp irradiation (upper picture) and storage time conditions (lower picture).

Subsequently, we delved into the structural analysis of carbon dots. The structure of carbon dots can be divided into macroscopic and microscopic levels. We used TEM to detect the structure of MB-CDs, which demonstrated a homogeneously dispersed quasi-spherical structure with an average size of approximately 2.66 nm, as observed in Fig. 2a and b. Under high-resolution transmission electron microscopy, MB-CDs exhibited distinct lattice fringes (0.21 nm) corresponding to the (100) crystalline surface of graphite (inset of Fig. 2a). XRD data showed that MB-CDs had a broad peak near 25°, characteristic of the graphite plane. Furthermore, MB-CDs exhibited excellent graphitization (Fig. 2c). Subsequently, we investigated the structure of MB-CDs using FT-IR and X-ray photoelectron spectroscopy. Using FT-IR the types of functional groups on the carbon dot surface were further characterized and structural changes were revealed during reactions by analyzing band intensities. In the FT-IR spectra, the absorption peak was located at 3423 cm⁻¹ due to the O–H stretching vibration, and the absorption peak appeared at 2966 cm⁻¹ due to the C–H stretching vibration. The absorption peak at 1649 cm⁻¹ was related to the stretching vibration of the C=O bond. The absorption peaks at 1405, 1047, and 602 cm⁻¹ corresponded to the C–N stretching vibration, C–O–C stretching vibration, and O–H bending vibration, respectively (Fig. 2d). In the XPS spectrum, MB-CDs comprised three elements, C, N, and O, accounting for 71.57%, 7.25%, and 21.18%, respectively (Fig. 2e). The high-resolution C 1s spectrum exhibited three peaks at 287.38, 285.78, and 284.28 eV, corresponding to C=O, C–N/C–O–C, and C–C/C=C, respectively, as observed in Fig. 2f. The N 1s spectrum had two peaks at 399.46 and 401.18 eV, which could be assigned to C–N and N–H bonds, respectively (Fig. 2g). There were two peaks in O 1s at 531.28 and 532.68 eV, which were attributed to C=O and C–O, respectively, as displayed in Fig. 2h.

Fig. 2 Characteristics of MB-CDs. (a) TEM images, (b) size distributions, (c) XRD, (d) FT-IR, (e) XPS, and high-resolution (f) C 1s spectra, (g) N 1s spectra, and (h) O 1s spectra of MB-CDs.

Carbon dots generally consist of a carbon core and surface functional groups, where the core is usually sp² hybridized crystalline or sp³ hybridized amorphous carbon. As there were abundant functional groups on the surface of MB-CDs, we continued to explore the possibilities of fluorescent probes and biological applications. First, we investigated the pH stability by adjusting the pH value of MB-CD aqueous solution with hydrochloric acid and sodium hydroxide. As illustrated in Fig. 3a and d, the fluorescence intensity of MB-CDs dropped sharply in strongly acidic and alkaline environments. They exhibited significantly higher luminescence performance in a neutral environment, which indicated that the prepared MB-CDs were biocompatible. To validate the application of MB-CDs as fluorescent probes, we studied the fluorescence quenching effects of MB-CDs on various metal ions. We selected glutathione (GSH) and different ionic compounds, including AlCl₃, CuCl₂, FeCl₃, NiCl₂, CaCl₂, CoCl₂, MgCl₂, NaCl, and KCl. As displayed in Fig. 3b, GSH did not influence the fluorescence intensity of MB-CDs. The capability of different metal ions to quench the fluorescence of MB-CDs was determined by the type of metal ion. The fluorescence intensity of MB-CDs was reduced by 80% upon adding Fe³⁺.

Fig. 3 The specific effect of Fe³⁺ ions on the fluorescence of MB-CDs. (a) Fluorescence emission spectra and (b) fluorescence intensity variations of MB-CDs under different pH conditions. (c) Fluorescence emission spectra and (d) fluorescence intensity plots of MB-CDs after the addition of different metal ions. (e) Fluorescence emission spectra of MB-CD solutions with different concentrations of Fe³⁺ added, and (f) the linear relationship between the fluorescence intensity ratio of MB-CD solutions at 1–10 μm Fe³⁺ concentration and Fe³⁺ ion concentration.

In contrast, it was only slightly decreased in the presence of other ions, suggesting that MB-CDs have a higher specific response to Fe³⁺ (Fig. 3e). Subsequently, we studied the response of MB-CDs to Fe³⁺ ions. As the Fe³⁺ concentration increased from l to 100 μM, the fluorescence intensity gradually decreased (Fig. 3c). Within a certain Fe³⁺ concentration range, there was a good linear relationship between the fluorescence intensity of the MB-CDs solution and Fe³⁺ concentration (Fig. 3f, R² = 0.984). These results demonstrated that the specific effect of Fe³⁺ ions on the fluorescence of MB-CDs made MB-CDs potentially useful for selective sensing of Fe³⁺ ions.

3.2 MB-CDs exhibited antioxidant properties

Previous studies have indicated that mung bean has antioxidant properties.⁵¹ We further explored the antioxidant properties of MB-CDs based on the abundant surface functional groups and excellent optical properties of MB-CDs. As a strong oxidizing agent, potassium permanganate (KMnO₄) had a maximum characteristic absorption peak at 525 nm. After contact with MB-CDs, due to the redox reaction, the characteristic absorption peak of KMnO₄ gradually decreased as the concentration of MB-CDs gradually increased from 0 to 35 μg mL⁻¹ (Fig. 4a). The color of KMnO₄ solution gradually changed from pink to colorless (Fig. 4d). In addition, the reduction capability of MB-CDs was as high as 99.1%. To evaluate the free radical scavenging performance of MB-CDs, we used two free radical scavenging activities, 1,1-diphenyl-2-picrylhydrazyl (DPPH) and 2,2′-azinobis-(3-ethylbenzothiazoline-6-sulfonic acid) (ABTS). DPPH˙, a nitrogen-centered free radical, has a characteristic absorption peak at 556 nm, which decreases due to the formation of a reduced and stable DPPH-H complex after contact with a free radical scavenger. As the MB-CD concentration increased from 0 to 50 μg mL⁻¹, the absorption intensity of DPPH˙ continued to decrease (Fig. 4b). Simultaneously, the color changed from purple to colorless (Fig. 4e). The clearance rate was 47.7% from the calculation. In the ABTS system, ABTS˙⁺ exhibited a characteristic absorption peak at 730 nm. As the concentration of MB-CDs increased from 0 to 40 μg mL⁻¹, the absorption intensity of ABTS˙⁺ decreased (Fig. 4c). Simultaneously, the color changed from green to colorless, and the ABTS˙⁺ clearance rate of the MB-CDs was as high as 94.1% (Fig. 4f).

Fig. 4 Antioxidant properties of MB-CDs. UV-Vis absorption spectra of (a) KMnO₄, (b) DPPH˙, (c) ABTS˙⁺ exposed to different concentrations of MB-CDs. (d) Photographs of KMnO₄ at different concentrations of MB-CDs (from left to right: 0, 5, 10, 15, 20, 25, 30 and 35 μg mL⁻¹). (e) Photographs of DPPH˙ at different concentrations of MB-CDs (from left to right: 0, 5, 15, 20, 25, 35, 40 and 50 μg mL⁻¹). (f) Photographs of ABTS˙⁺ at different concentrations of MB-CDs (from left to right: 0, 5, 10, 15, 20, 25, 30 and 40 μg mL⁻¹).

It is generally believed that hydroxyl radicals (˙OH) and superoxide anions (˙O₂⁻) are the most important physiologically relevant highly reactive oxygen radicals, which can reduce physiological functions and cause various human diseases. Accordingly, we tested whether MB-CDs can scavenge ˙OH and ˙O₂⁻. First, the 3,3′,5,5′-tetra-methylbenzidine (TMB) chromogenic method was used to determine the scavenging activity of ˙OH produced by the Fenton reaction. After adding MB-CDs to the TMB solution mixed with ˙OH, the absorption intensity of TMB gradually decreased as the concentration of MB-CDs increased (Fig. 5a), and the reaction color gradually became lighter (Fig. 5c). The clearance rate of ˙OH of MB-CDs was as high as 95.8%. The colorimetric nitro-blue tetrazolium (NBT) enzymatic method was used to evaluate the ˙O₂⁻ scavenging activity of MB-CDs. After reacting with MB-CDs, the absorption intensity at 596 nm gradually decreased with increasing concentration of MB-CDs in the range of 0–50 μg mL⁻¹ (Fig. 5b), and the reaction color gradually became lighter (Fig. 5d), indicating that MB-CDs could remove ˙O₂⁻ and inhibit the reduction of NBT. Besides, we investigated the electron spin resonance (EPR) spectrum of the xanthine/xanthine oxidase reaction, revealing a specific signal for the DMPO-OOH adduct and indicating the successful generation of ˙O₂⁻. After adding different MB-CD concentrations to the system, the EPR signal intensity decreased to varying degrees (Fig. 5e). Adding different concentrations of MB-CDs also reduced the EPR amplitude of DMPO-OH (Fig. 5f), which further indicated the good ˙OH and ˙O₂⁻ scavenging capabilities of the MB-CDs. In summary, MB-CDs could effectively serve as potential candidates for scavenging free radicals, including DPPH˙, ABTS˙⁺, ˙OH, and ˙O₂⁻.

Fig. 5 MB-CDs had the ability to scavenge ˙OH and ˙O₂⁻. UV-Vis absorption spectra of (a) ˙OH and (b) ˙O₂⁻ exposed to different concentrations of MB-CDs. (c) Photographs of ˙OH at different concentrations of MB-CDs (from left to right: 0, 2, 4, 15, 6, 8, 10, 15 and 20 μg mL⁻¹). (d) Photographs of ˙O₂⁻ at different concentrations of MB-CDs (from left to right: 0, 10, 15, 20, 25, 30, 45 and 50 μg mL⁻¹). EPR spectra of (e) ˙OH and (f) ˙O₂⁻ scavenged by MB-CDs at different concentrations (50, 100, 200 μg mL⁻¹).

3.3 MB-CDs enhanced the repair-related phenotype of SCs in vitro

The repair-related phenotype of SCs was reported to be essential in the treatment of PNI, including (1) enhancing the proliferation and anti-apoptotic ability of damaged neurons, (2) enhancing the migration ability and formation of the Büngner band, (3) upregulating growth factors and neurotrophic factors for angiogenesis and nerve regeneration at nerve injury sites, including NGF, CNTF, and GDNF, (4) upregulating the expression of immune-related cytokines to clear myelin sheath fragments.¹³ To verify the above effects of MB-CDs on SCs, we labeled MB-CDs and observed their interaction with SCs using an inflammatory SC model mimicking the PNI microenvironment. First, we tested the ability of SCs to internalize MB-CDs in vitro. After incubation with SCs for 24 h, blue fluorescence of MB-CDs appeared in the cytoplasm of SCs, suggesting an effective ability to internalize the MB-CDs of SCs (Fig. 6a). Next, we evaluated the biosafety of MB-CDs on SCs. The results of the CCK8 assay indicated non-significant differences between the CDs-SCs and CTRL (Fig. 6b), suggesting low toxicity of MB-CDs on SCs.

Fig. 6 MB-CDs induced repair-related phenotypes of SCs. (a) SC was incubated with MB-CDs for 24 hours, and representative fluorescence images show that MB-CDs (blue) are delivered to SC. Scale bar, 100 μm. (b) Statistical evaluation of the survival percentage of SCs treated with different concentrations of MB-CDs. The data are expressed as mean ± SD (n = 3). (c) The protein levels of factors associated with nerve regeneration (CNTF, GDNF, and NGF) were detected by western blotting of LPS-SCs and LPS-CDs in each group. (d) Quantification of CNTF, GDNF, and NGF protein levels in each group. The data are expressed as mean ± SD (n = 3). (e) Statistical evaluation of the survival percentage of LPS-SCs treated with different concentrations of MB-CDs. The data are expressed as mean ± SD (n = 3). (f) The colony formation of SCs treated with different concentrations of MB-CDs for 10 days. (g) Statistical results of the colony formation in each group. The data are expressed as mean ± SD (n = 3). (h) Representative Edu staining images of the MB-CDs processing group within 24 hours. Scale bar, 100 μm. (i) Edu positive SC. The data is represented as mean ± standard deviation (n = 3). (j) The number of migrated SCs was counted and analyzed. The data are expressed as mean ± SD (n = 3). (k) Representative images of vertical migration of SCs with different treatments for 24 h. Scale bar, 100 μm. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.

Furthermore, we used LPS to induce an inflammatory model of SCs and co-cultured them with MB-CDs.⁵² The growth and neurotrophic factors involved in angiogenesis and nerve regeneration were tested.^53–55 As displayed in Fig. 6c and d, the expression and protein levels of NGF, CNTF, and GDNF in the LPS-CDs increased compared with the LPS-SCs. The cell death experiment indicated that MB-CDs inhibited LPS-induced SC death significantly, particularly at 200 and 100 μg mL⁻¹ (Fig. 6e). Based on these results, in the following experiment, we finally chose 200 μg mL⁻¹ MB-CDs as the experimental group in the following experiments. MB-CDs were incubated with SCs for 10 days to investigate their effects on SC colonies. We found that the MB-CD group aggregated more colonies than the control group treated with PBS (Fig. 6f and g). To further study the proliferation promotion effect of MB-CDs on SCs, we performed Edu staining to demonstrate the enhancing function of MB-CDs on the proliferation of SCs (Fig. 6h and i). As reported, the recruitment and migration of SCs to the injured nerve sites were crucial for forming Büngner bands and axonal regeneration. We co-cultured MB-CDs and LPS-SCs in a Transwell system (Fig. 6j and k), showing that MB-CDs significantly improved the migration ability of SCs. Overall, these results demonstrated that MB-CDs effectively enhanced the repair-related phenotype of SCs in LPS-induced nerve damage.

3.4 MB-CDs promoted functional recovery after PNI and prevented gastrocnemius atrophy in vivo

To evaluate the influence of MB-CDs on nerve functional recovery in PNI rat models, we tested and evaluated SFI 3 days before the operation, 7, 14 and 21 days following the operation using walking track analysis. As displayed in Fig. 7a, the PNI group experienced severe foot contractures; however, the footprints of the sham and PNI-CDs groups were clear and well-shaped, indicating successful functional recovery in the rats. The SFI value of the PNI-CDs was higher than that of the PNI group 21 days post-surgery (Fig. 7b). Moreover, we evaluated gastrocnemius muscle atrophy by comparing the relative weight and average fiber diameter of each group's gastrocnemius muscle atrophy. After a 21-day recovery, the gastrocnemius muscle in the PNI-CD group was larger and heavier than that in the PNI group, suggesting that MB-CD treatment could prevent gastrocnemius muscle atrophy (Fig. 7c and d). Moreover, HE and Masson staining were performed on longitudinal and transverse sections of the sciatic nerve to observe the morphology and histological changes of the injured nerve fibers. As displayed in Fig. 7e, compared with the PNI group the structure of the injured nerves treated with MB-CDs was more tightly ordered, with less edema and vacuolization. Moreover, H&E and Masson's trichrome staining results illustrated that the average diameter of the fibers in the gastrocnemius muscle was larger in the PNI-CD group than in the PNI group (Fig. 7f and g). These results demonstrated that MB-CDs could help protect the gastrocnemius muscle and prevent it from atrophy after sciatic nerve injury. To better understand the role of MB-CDs in promoting the regeneration process of the sciatic nerve, we also studied sensory function recovery. The mechanical and thermal pain thresholds of the rats were tested, and our analysis revealed that after MB-CDs injection, the pain thresholds began to recover gradually on the seventh day and displayed a significant difference from the control group on the 21^st-day post-injury (Fig. 7h and i). Finally, we explored the biosafety of MB-CDs in vivo. H&E staining was performed on the heart, liver, spleen, lungs, and kidneys of the rats. There were non-significant differences between control and MB-CD treatment groups (Fig. S3†).

Fig. 7 Injecting MB-CDs into the outer membrane of the sciatic nerve promotes nerve regeneration. (a) Representative images of the prosthetic foot, CTRL, PNI, MB-CDs group of footprints at 21 d after surgery. (b) SFI was analyzed for each group at 3 d preoperatively, 7 d, 14 d and 21 d postoperatively. Data are expressed as mean ± SD (n = 5). (c) Representative images of gastrocnemius muscles harvested from CTRL, PNI, and MB-CDs groups. (d) Statistical analysis of the weight ratio of gastrocnemius muscle in different groups. Data were expressed as mean ± SD (n = 5). (e) HE and Masson trichrome staining of nerves in each group. Scale bar, 20 μm. (f) HE and Masson trichrome staining of gastrocnemius muscle in each group. Scale bar, 20 μm. (g) Statistical analysis of the muscle fiber mean diameter. The data were expressed as mean ± SD (n = 5). (h) Changes in mechanical pain threshold in rats after different treatments. The data were expressed as mean ± SD (n = 5). (i) Changes in thermal pain threshold in rats after different treatments. The data were expressed as mean ± SD (n = 5). **p < 0.01, ***p < 0.001.

3.5 MB-CDs inhibited the production of ferroptosis in SCs

We performed RNA sequencing to identify underlying mechanisms in LPS-CDs and LPS-SCs. The result of the differential expression genes in the two groups was analyzed using Venn analysis (Fig. 8a). The volcano plot illustrated upregulated RNA (red) and downregulated microRNAs (green), as observed in Fig. 8b. The differentially expressed genes from the two groups were illustrated using heat maps (Fig. 8c). Gene ontology (GO) enrichment analysis demonstrated that GO terms were associated with oxidative stress, including the oxidation–reduction process and oxidative resistance activity (Fig. 8d–f).

Fig. 8 Transcriptome sequencing analysis. The mRNA in the treated SCs was analyzed through mRNA sequencing analysis. (a) Different mRNAs were detected in the LPS group and LPS-CDs groups on a co-expression Venn diagram. (b) Volcano plot of Differentially Expressed Genes (DEGs) between LPS-CPS and LPS groups. The x-axis and y-axis represent log2 fold-change differences between the compared samples and statistical significance as the negative logarithm of DEG p-values, respectively. The significantly up-regulated and down-regulated genes are indicated with red and green dots, respectively, while non-significant genes are shown as blue dots. (c) The heatmap illustrates the comparison of transcripts with higher differential expression values between the LPS-CPS and LPS groups. High values are represented in red color, while the lowest values are represented in green. (d) The top enriched biological process (BP) pathways of DEGs in GO. (e) The top enriched molecular function (MF) pathways of DEGs in GO. (f) The top enriched cellular component (CC) pathways of DEGs in GO. (g) Enrichment plots from GSEA show that the ferroptosis biological pathway (RN004216) was enriched in the LPS-CPS group, as sorted by normalized enrichment scores. (h) The expressions of DEGs in the ferroptosis KEGG pathway are visualized using a color scheme. In this scheme, red indicates up-regulated genes, while green represents down-regulated genes. (i) Differentially expressed genes and FRDEGs on a Venn diagram. (j) The expression heatmap of FRDEGs. (k) Protein–protein interaction networks. (l) Top 10 genes through the MCC algorithm.

Ferroptosis is a novel form of programmed cell death characterized by the excessive accumulation of lipid peroxides facilitated by iron ions. This process is primarily marked by the intracellular accumulation of iron ions, increased levels of ROS, and depletion of GSH.²⁷ As depicted in Fig. 3e, MB-CDs, which bind well to Fe³⁺, have demonstrated the potential to regulate ferroptosis by chelating free Fe³⁺ when combined with the oxidative stress process in cells’ molecular function. Therefore, we further analyzed the role of ferroptosis in the mechanism of MB-CD treatment of PNI. We first performed a GSEA using the R package “cluster profiler”, the results suggested that MB-CD treatment suppressed signaling pathways associated with ferroptosis (Fig. 8g). The Kyoto encyclopedia of genes and genome pathways related to ferroptosis displayed differentially expressed genes associated with ferroptosis (Fig. 8h). Second, we intersected the obtained differentially expressed genes with 426 ferroptosis-related genes to obtain 62 ferroptosis-related differentially expressed genes (FRDEGs) (Fig. 8i). The expression heatmap of FRDEGs is displayed in Fig. 8j. To identify hub genes, we used the STRING database to generate a PPI network (Fig. 8k). The top 10 genes through the MCC algorithm were selected, among which the first ranked was the HO-1 gene (Fig. 8l).

3.6 MB-CDs inhibited ferroptosis of SCs via Nrf2/HO-1/GPX4 signaling pathways

To investigate the anti-ferroptosis effects of MB-CDs on ROS in LPS-SCs, we treated LPS-SCs with 200 μg mL⁻¹ MB-CDs and PBS. Moreover, 200 μg mL⁻¹ MB-CDs and PBS were added to SCs as the control. The green fluorescence increased significantly in LPS-SCs and decreased significantly in LPS-CDs (Fig. 9a and b). Mitochondria are important mediators of cellular metabolism and generators and targets of ROS.⁵⁶ ROS disrupt the mitochondrial membrane, alter its permeability, and change the concentration of ions inside and outside the membrane, leading to a decrease in membrane potential.⁵⁷ To further investigate the inhibitory effect of MB-CDs on ROS production, we measured the mitochondrial membrane potential of LPS-SCs with JC-1 and found that the mitochondrial membrane potential decreased significantly (Fig. S1a and b†). We concluded that MB-CDs effectively inhibited LPS-induced ROS production in LPS-SCs. To verify the role of MB-CDs in vivo, we established a peripheral nerve crush injury model in rats and injected MB-CDs into the peripheral nerve epineurium. The rats were euthanized seven days after surgery, and the sciatic nerves were removed for pathological sectioning. The red fluorescence stained by ROS was apparent in the nerve injury group; however, this fluorescence was diminished when treating the injured models with MB-CDs, suggesting that MB-CDs could significantly inhibit the increase of ROS in SCs caused by injury (Fig. 9c and d).

Fig. 9 MB-CDs can inhibit ferroptosis of SCs via Nrf2/HO-1/GPX4 signaling pathways. (a) Images of SCs and LPS-SCs cells after treatment with MB-CDs for 48 h to analyze ROS production. Scale bar, 50 μm. (b) Statistical evaluation of the percentage of ROS-positive SCs. The data are expressed as mean ± SD (n = 3). (c) Statistical evaluation of the percentage of the ROS-positive rat sciatic nerve. The data are expressed as mean ± SD (n = 5). (d) Fluorescence images of the rat sciatic nerve after PNI for 7 days to analyze ROS production. Scale bar, 50 μm. (e) Expression levels of MDA, GPX, CAT, SOD, Fe²⁺ and total iron ions in SCs from different treatment groups (n = 5). (f) Western blotting and statistical analysis of OGG-1 and POLG. The data are expressed as mean ± SD (n = 3). (g) Western blotting and statistical analysis of GPX4, Nrf2 and HO-1. The data are expressed as mean ± SD (n = 3). (h) Western blotting and statistical analysis of GPX4, Nrf2 and HO-1. The data are expressed as mean ± SD (n = 5). (i) Nrf2 (purple) and S100β (green) double immunofluorescence staining of sciatic nerve longitudinal sections at 7 days following the operation. Nuclei were stained with DAPI. Scale bar, 50 μm. (j) HO-1 (white) and S100β (green) double immunofluorescence staining of sciatic nerve longitudinal sections at 7 days following the operation. Nuclei were stained with DAPI. Scale bar, 50 μm. (k) GPX4 (red) and S100β (green) double immunofluorescence staining of sciatic nerve longitudinal sections at 7 days following the operation. Nuclei were stained with DAPI. Scale bar, 50 μm. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.

Next, we examined the activity levels of malondialdehyde (MDA), glutathione peroxidase (GPX), catalase (CAT), superoxide dismutase (SOD), Fe²⁺ and total iron ions. Our findings indicate that MB-CDs could effectively inhibit elevated lipid oxidation and promote the elevation of GPX, CAT, and SOD in LPS-SCs. These results collectively suggest that MB-CDs could effectively inhibit the occurrence of intracellular ROS. Moreover, the reduction in total intracellular ferric and ferrous ions also inhibited cellular ferroptosis (Fig. 9e). Moreover, the expressions of DNA oxidative damage repair-related protein, OGG-1,⁵⁸ and the mitochondrial DNA repair enzyme POLG⁵⁹ were also increased by the addition of MB-CDs, suggesting that MB-CDs could promote DNA repair after oxidative stress (Fig. 9f). Chen et al. showed that POLG and OGG-1 were increased in D-gal-induced oxidative stress neurons, which is consistent with our results that the expression levels of POLG and OGG-1 also had a slight elevation in LPS-induced oxidative stress cells.⁶⁰

As illustrated in Fig. 8l, HO-1 and GPX4 were the top 10 hub genes in the FRDEGs. Accordingly, we performed further validation to explore whether MB-CDs inhibited the ferroptosis of SCs by regulating Nrf2/HO-1/GPX4 pathways. First, western blotting was performed. We found that the lower expression of GPX4, Nrf2 and HO-1 in SCs after LPS induction was reversed by co-culturing MB-CDs with LPS-SCs (Fig. 9g). Notably, a similar phenomenon was observed in the rat sciatic nerve after injury. After the injection of MB-CDs, the reduction of these proteins was mitigated (Fig. 9h). Similarly, immunofluorescence and immunohistochemistry were used to evaluate GPX4 protein expression in rat sciatic nerve sections. For crush injury models, the group treated with MB-CDs depicted a significant increase in Nrf2 compared to the group without MB-CDs (Fig. 9i). Moreover, HO-1 (Fig. 9j) and GPX4 (Fig. 9k), and the expression of OGG-1 and POLG (Fig. S2a–d†) were all increased after injecting MB-CDs into injured sciatic nerves. Taken together, these findings indicated that MB-CDs can suppress the ferroptosis of SCs through regulating Nrf2/HO-1/GPX4 pathways. While previous studies have either focused on ferroptosis or repair mechanisms in isolation, our work pioneers the exploration of the intricate relationship between them. This connection has far-reaching implications, as it provides a more comprehensive understanding of the cellular processes involved in nerve repair and offers a potentially transformative approach.

4. Conclusions

In this study, MB-CDs were synthesized as an alternative treatment for PNI. In vitro experiments demonstrated that MB-CDs could be internalized by SCs, and the biological safety of MB-CDs was reliable. Moreover, MB-CDs could significantly enhance the repair-related phenotype of SCs by promoting proliferation and migration and upregulating the expression of growth and neurotrophic factors. In vivo experiments indicated that MB-CDs promoted functional recovery after PNI and prevented gastrocnemius muscle atrophy. The possible mechanism was that MB-CDs could suppress the ferroptosis of SCs to promote peripheral nerve repair via Nrf2/HO-1/GPX4 signaling pathways. We hope this study will provide a valuable therapeutic strategy for PNI.

Abbreviations

PNI	Peripheral nerve injury
CDs	Carbon dots
MB-CDs	Mung bean-derived carbon dots
SCs	Schwann cells
OS	Oxidative stress
ROS	Reactive oxygen species
UV-Vis	UV-visible
TEM	Transmission electron microscopy
XRD	X-ray diffraction
FT-IR	Fourier transform infrared spectroscopy
XPS	X-ray photoelectron spectroscopy
NBT	Nitro blue tetrazolium
Edu	5-Ethynyl-2-deoxyuridine
DEG	Differentially expressed genes
FRDEGs	Ferroptosis-related differentially expressed genes
GO	Gene ontology
BP	Biological process
MF	Molecular functions
CC	Cellular component
KEGG	Kyoto Encyclopedia of Genes and Genomes
GSEA	Gene Set Enrichment Analysis
DHE	Dihydroethylene glycol
MMP	Mitochondrial membrane potential
PFA	Paraformaldehyde
HE	Hematoxylin eosin
DPPH	1,1-Diphenyl-2-picrylhydrazyl
ABTS	2,2′-Azinobis-(3-ethylbenzothiazoline-6-sulfonic acid)
TMB	3,3′,5,5′-Tetra-methylbenzidine
EPR	Electron spin resonance
NGF	Nerve growth factor
CNTF	Ciliary neurotrophic factor
GDNF	Glial cell derived neurotrophic factor
MDA	Malondialdehyde
GPX	Glutathione peroxidase
GSH	Glutathione
CAT	Catalase
SOD	Superoxide dismutase
OGG-1	8-Oxoguanine DNA glycosylase
POLG	Polymerase gamma
HO-1	Homox1
Nrf-2	Nuclear factor erythroid 2-related factor 2
GPX4	Glutathione peroxidase 4
SFI	Sciatic functional index

Author contributions

Conceptualization: F. Z., Y. Z., and N. Z. Methodology: F. Z., Y. Z., H. Z., and J. L. Investigation: F. Z., Y. Z., H. Z., J. L., and J. G. Visualization: F. Z., Y. Z., H. Z., and J. L. Supervision: S. L. and N. Z. Funding acquisition: Y. W. Project administration: X. Q., X. W., and Y. W. Writing–original draft: F. Z., Y. Z., and J. L. Writing–review and editing: H. Z. and N. Z. Resources: S. L. and N. Z. Data curation: Y. W. Validation: Y. W. Formal analysis: Y. W. Software: Y. W. All of the authors read and approved the final manuscript.

Ethics approval and consent to participate

The living conditions and experimental procedures complied with the guidelines of the National Institutes of Health. In addition, all animal experiments were approved by the Animal Experiment Ethics Committee of Zhengzhou University.

Consent for publication

All authors agreed to be published.

Data availability

The data supporting this article have been included as part of the ESI.†

The data and materials that support the findings of this study are available from the corresponding author upon reasonable request.

Conflicts of interest

The authors declare no potential conflicts of interest.

Acknowledgements

This work was supported by the National Natural Science Foundation of China (82372498), the Excellent Youth Science Foundation of Changchun (23YQ07), the Genetic precision medicine discipline development public welfare project of Peking Union Medical Foundation (PUMFO1010075), and the joint open project of Beijing Key Laboratory of Bone Malformation Genetics Research and Key Laboratory of Big Data for Spinal Deformities of Chinese Academy of Medical Sciences (BKJORT202401). We appreciate the technical help from the Academy of Medical Science, Zhengzhou University while performing experiments.

References

1. T. Huckhagel, J. Nüchtern, J. Regelsberger, R. Lefering and TraumaRegister DGU, Nerve injury in severe trauma with upper extremity involvement: evaluation of 49,382 patients from the TraumaRegister DGU® between 2002 and 2015, Scand. J. Trauma Resusc. Emerg. Med., 2018, 26, 76.
2. J. Noble, C. A. Munro, V. S. Prasad and R. Midha, Analysis of upper and lower extremity peripheral nerve injuries in a population of patients with multiple injuries, J. Trauma, 1998, 45, 116–122.
3. T. Gordon, Peripheral Nerve Regeneration and Muscle Reinnervation, Int. J. Mol. Sci., 2020, 21, 8652.
4. X. L. Chu, X. Z. Song, Q. Li, Y. R. Li, F. He, X. S. Gu and D. Ming, Basic mechanisms of peripheral nerve injury and treatment via electrical stimulation, Neural Regen. Res., 2022, 17, 2185–2193.
5. I. Allodi, E. Udina and X. Navarro, Specificity of peripheral nerve regeneration: interactions at the axon level, Prog. Neurobiol., 2012, 98, 16–37.
6. M. Modrak, M. A. H. Talukder, K. Gurgenashvili, M. Noble and J. C. Elfar, Peripheral nerve injury and myelination: Potential therapeutic strategies, J. Neurosci. Res., 2020, 98, 780–795.
7. Y. Dong and H. Lu, Editorial: Surgical treatment of peripheral neuropathic pain, peripheral nerve tumors, and peripheral nerve injury, Front. Neurol., 2023, 14, 1266638.
8. R. Sullivan, T. Dailey, K. Duncan, N. Abel and C. V. Borlongan, Peripheral Nerve Injury: Stem Cell Therapy and Peripheral Nerve Transfer, Int. J. Mol. Sci., 2016, 17, 2101.
9. Y. P. Liu, Z. R. Luo, C. Wang, H. Cai, T. T. Zhao, H. Li, S. J. Shao and H. D. Guo, Electroacupuncture Promoted Nerve Repair After Peripheral Nerve Injury by Regulating miR-1b and Its Target Brain-Derived Neurotrophic Factor, Front. Neurosci., 2020, 14, 525144.
10. X. C. Muniz, A. C. C. de Assis, B. S. A. de Oliveira, L. F. R. Ferreira, M. Bilal, H. M. N. Iqbal and R. N. Soriano, Efficacy of low-level laser therapy in nerve injury repair-a new era in therapeutic agents and regenerative treatments, Neurol. Sci., 2021, 42, 4029–4043.
11. A. W. English, K. Berglund, D. Carrasco, K. Goebel, R. E. Gross, R. Isaacson, O. C. Mistretta and C. Wynans, Bioluminescent Optogenetics: A Novel Experimental Therapy to Promote Axon Regeneration after Peripheral Nerve Injury, Int. J. Mol. Sci., 2021, 22, 7217.
12. N. Divac, M. Aksić, L. Rasulić, M. Jakovčevski, M. Basailović and I. Jakovčevski, Pharmacology of repair after peripheral nerve injury, Int. J. Clin. Pharmacol. Ther., 2021, 59, 447–462.
13. J. Huang, G. Zhang, S. Li, J. Li, W. Wang, J. Xue, Y. Wang, M. Fang and N. Zhou, Endothelial cell-derived exosomes boost and maintain repair-related phenotypes of Schwann cells via miR199–5p to promote nerve regeneration, J. Nanobiotechnol., 2023, 21, 10.
14. J. Qiu, X. Yang and L. Wang, et al., Isoquercitrin promotes peripheral nerve regeneration through inhibiting oxidative stress following sciatic crush injury in mice, Ann. Transl. Med., 2019, 7(22), 680.
15. L. S. Costa, F. J. Aidar and D. G. Matos, et al., Effects of Resistance Training and Bowdichia virgilioides Hydroethanolic Extract on Oxidative Stress Markers in Rats Submitted to Peripheral Nerve Injury, Antioxidants, 2020, 9(10), 941.
16. H. Sies, Oxidative stress: a concept in redox biology and medicine, Redox Biol., 2015, 4, 180–183.
17. H. F. Du, J. W. Wu, Y. S. Zhu, Z. H. Hua, S. Z. Jin, J. C. Ji, C. S. Wang, G. Y. Qian, X. D. Jin and H. M. Ding, Fucoxanthin Induces Ferroptosis in Cancer Cells via Downregulation of the Nrf2/HO-1/GPX4 Pathway, Molecules, 2024, 29, 2832.
18. A. A. El-Gazar, A. A. Soubh, D. M. Abdallah, G. M. Ragab and H. S. El-Abhar, Elucidating PAR1 as a therapeutic target for delayed traumatic brain injury: Unveiling the PPAR-γ/Nrf2/HO-1/GPX4 axis to suppress ferroptosis and alleviate NLRP3 inflammasome activation in rats, Int. Immunopharmacol., 2024, 139, 112774.
19. H. K. Tosyalı, E. S. Bora, O. S. Çınaroğlu and O. Erbaş, Oxytocin mitigates peripheral nerve damage via Nrf2 and irisin pathway, Eur. Rev. Med. Pharmacol. Sci., 2023, 27(23), 11340–11350.
20. H. Zhang, W. Zheng, X. Feng, F. Yang, H. Qin, S. Wu, D. X. Hou and J. Chen, Nrf2⁻ARE Signaling Acts as Master Pathway for the Cellular Antioxidant Activity of Fisetin, Molecules, 2019, 24, 708.
21. M. Dodson, R. Castro-Portuguez and D. D. Zhang, NRF2 plays a critical role in mitigating lipid peroxidation and ferroptosis, Redox Biol., 2019, 23, 101107.
22. S. Murakami, Y. Kusano, K. Okazaki, T. Akaike and H. Motohashi, NRF2 signalling in cytoprotection and metabolism, Br. J. Pharmacol., 2023, 1–14,  DOI:10.1111/bph.16246.
23. O. Adedoyin, R. Boddu, A. Traylor, J. M. Lever, S. Bolisetty, J. F. George and A. Agarwal, Heme oxygenase-1 mitigates ferroptosis in renal proximal tubule cells, Am. J. Physiol.: Renal Physiol., 2018, 314, F702–F714.
24. M. Xu, D. Zhang and J. Yan, Targeting ferroptosis using Chinese herbal compounds to treat respiratory diseases, Phytomedicine, 2024, 130, 155738.
25. Q. Zhang, J. Liu and H. Duan, et al., Activation of Nrf2/HO-1 signaling: An important molecular mechanism of herbal medicine in the treatment of atherosclerosis via the protection of vascular endothelial cells from oxidative stress, J. Adv. Res., 2021, 34, 43–63.
26. Y. Xie, R. Kang, D. J. Klionsky and D. Tang, GPX4 in cell death, autophagy, and disease, Autophagy, 2023, 19(10), 2621–2638.
27. E. Dai, X. Chen and A. Linkermann, et al., A guideline on the molecular ecosystem regulating ferroptosis, Nat. Cell Biol., 2024, 26, 1447–1457.
28. S. J. Dixon, K. M. Lemberg, M. R. Lamprecht, R. Skouta, E. M. Zaitsev, C. E. Gleason, D. N. Patel, A. J. Bauer, A. M. Cantley, W. S. Yang, B. Morrison 3rd and B. R. Stockwell, Ferroptosis: an iron-dependent form of nonapoptotic cell death, Cell, 2012, 149, 1060–1072.
29. D. Wu and L. Chen, Ferroptosis: a novel cell death form will be a promising therapy target for diseases, Acta Biochim. Biophys. Sin., 2015, 47, 857–859.
30. H. Ma, Y. Dong, Y. Chu, Y. Guo and L. Li, The mechanisms of ferroptosis and its role in alzheimer's disease, Front. Mol. Biosci., 2022, 9, 965064.
31. L. Huang, M. Bian, J. Zhang and L. Jiang, Iron Metabolism and Ferroptosis in Peripheral Nerve Injury, Oxid. Med. Cell. Longevity, 2022, 2022, 5918218,  DOI:10.1155/2022/5918218 , PMID: 36506935; PMCID: PMC9733998.
32. M. Guo, Y. Zhu, Y. Shi, X. Meng, X. Dong, H. Zhang, X. Wang, M. Du and H. Yan, Inhibition of ferroptosis promotes retina ganglion cell survival in experimental optic neuropathies, Redox Biol., 2022, 58, 102541.
33. Y. Yan, X. Ran, Z. Zhou, Y. Gu, R. Wang, C. Qiu, Y. Sun, J. Wang, J. Xiao, Y. Lu and J. Wang, FGF21 inhibits ferroptosis caused by mitochondrial damage to promote the repair of peripheral nerve injury, Front. Pharmacol., 2024, 15, 1358646.
34. R. Dang, M. Wang, X. Li, H. Wang, L. Liu, Q. Wu, J. Zhao, P. Ji, L. Zhong, J. Licinio and P. Xie, Edaravone ameliorates depressive and anxiety-like behaviors via Sirt1/Nrf2/HO-1/Gpx4 pathway, J. Neuroinflammation, 2022, 19(1), 41.
35. P. Zhu, S. Wang, Y. Zhang, Y. Li, Y. Liu, W. Li, Y. Wang, X. Yan and D. Luo, Carbon Dots in Biomedicine: A Review, ACS Appl. Bio Mater., 2022, 5, 2031–2045.
36. Y. Xu, B. Wang, M. Zhang, J. Zhang, Y. Li, P. Jia, H. Zhang, L. Duan, Y. Li, Y. Li, X. Qu, S. Wang, D. Liu, W. Zhou, H. Zhao, H. Zhang, L. Chen, X. An, S. Lu and S. Zhang, Carbon Dots as a Potential Therapeutic Agent for the Treatment of Cancer-Related Anemia, Adv. Mater., 2022, 34, 2200905.
37. M. Fang, B. Wang, X. Qu, S. Li, J. Huang, J. Li, S. Lu and N. Zhou, State-of-the-art of biomass-derived carbon dots: Preparation, properties, and applications, Chin. Chem. Lett., 2024, 35, 108423.
38. Y. Zhang, S. Zhang, B. Tan, L. Guo and H. Li, Solvothermal synthesis of functionalized carbon dots from amino acid as an eco-friendly corrosion inhibitor for copper in sulfuric acid solution, J. Colloid Interface Sci., 2021, 604, 1–14.
39. X. Tang, H. Yu, B. Bui, L. Wang, C. Xing, S. Wang, M. Chen, Z. Hu and W. Chen, Nitrogen-doped fluorescence carbon dots as multi-mechanism detection for iodide and curcumin in biological and food samples, Bioact. Mater., 2020, 6, 1541–1554.
40. S. Meng, D. Cheng, H. Gu, Y. Li, Y. Qin, J. Tan and Q. Li, Mechanical Force-Induced Color-Variable Luminescence of Carbon Dots in Boric Acid Matrix, Molecules, 2023, 28, 3388.
41. Q. Liu, Z. Dong, A. Hao, X. Guo and W. Dong, Synthesis of highly fluorescent carbon dots as a dual-excitation rationmetric fluorescent probe for the fast detection of chlorogenic acid, Talanta, 2021, 221, 121372.
42. H. H. Jing, F. Bardakci, S. Akgöl, K. Kusat, M. Adnan, M. J. Alam, R. Gupta, S. Sahreen, Y. Chen, S. C. B. Gopinath and S. Sasidharan, Green Carbon Dots: Synthesis, Characterization, Properties and Biomedical Applications, J. Funct. Biomater., 2023, 14, 27.
43. D. Liu, X. Guan, K. Huang, S. Li, J. Liu, W. Yu and R. Duan, Protective effects of mung bean (Vigna radiata L.) and pea (Pisum sativum L.) against high-fat-induced oxidative stress, Food Sci. Nutr., 2019, 7, 4063–4075.
44. S. Zhang, Y. T. Ma, Y. C. Feng, C. Y. Wang and D. J. Zhang, Potential effects of mung bean protein and a mung bean protein-polyphenol complex on oxidative stress levels and intestinal microflora in aging mice, Food Funct., 2022, 13, 186–197.
45. D. Tang, Y. Dong, N. Guo, L. Li and H. Ren, Metabolomic analysis of the polyphenols in germinating mung beans (Vigna radiata) seeds and sprouts, J. Sci. Food Agric., 2014, 94, 1639–1647.
46. T. Dongyan, D. Yinmao, L. Li, L. Yueheng, H. Congfen and L. Jixiang, Antioxidant activity in mung bean sprouts and safety of extracts for cosmetic use, J. Cosmet. Sci., 2014, 65, 207–216.
47. H. Xu, Q. Zhou, B. Liu, K. W. Cheng, F. Chen and M. Wang, Neuroprotective Potential of Mung Bean (Vigna radiata L.) Polyphenols in Alzheimer's Disease: A Review, J. Agric. Food Chem., 2021, 69(39), 11554–11571.
48. X. Shen, X. Jiang, L. Qian, A. Zhang, F. Zuo and D. Zhang, Polyphenol Extracts From Germinated Mung Beans Can Improve Type 2 Diabetes in Mice by Regulating Intestinal Microflora and Inhibiting Inflammation, Front. Nutr., 2022, 9, 846409.
49. L. A. R. Lopes, M. D. C. C. E. Martins, L. M. Farias, A. K. D. S. Brito, G. M. Lima, V. B. L. Carvalho, C. F. C. Pereira, A. M. Conde Júnior, T. Saldanha, J. A. G. Arêas, K. J. D. E. Silva and K. M. G. Frota, Cholesterol-Lowering and Liver-Protective Effects of Cooked and Germinated Mung Beans (Vigna radiata L.), Nutrients, 2018, 10, 821.
50. . FerrDb, https://www.zhounan.org/ferrdb, accessed 20 June 2024.
51. Y. Bai, J. Chang, Y. Xu, D. Cheng, H. Liu, Y. Zhao and Z. Yu, Antioxidant and Myocardial Preservation Activities of Natural Phytochemicals from Mung Bean (Vigna radiata L.), Seeds, J. Agric. Food Chem., 2016, 64, 4648–4655.
52. U. Andersson and H. Yang, HMGB1 is a critical molecule in the pathogenesis of Gram-negative sepsis, J. Intensive Med., 2022, 2, 156–166.
53. R. Li, D. H. Li, H. Y. Zhang, J. Wang, X. K. Li and J. Xiao, Growth factors-based therapeutic strategies and their underlying signaling mechanisms for peripheral nerve regeneration, Acta Pharmacol. Sin., 2020, 41, 1289–1300.
54. P. M. Richardson, Ciliary neurotrophic factor: a review, Pharmacol. Ther., 1994, 63, 187–198.
55. A. F. Cintron-Colon, G. Almeida-Alves, J. M. VanGyseghem and J. M. Spitsbergen, GDNF to the rescue: GDNF delivery effects on motor neurons and nerves, and muscle re-innervation after peripheral nerve injuries, Neural Regen. Res., 2022, 17, 748–753.
56. A. Singh, R. Kukreti, L. Saso and S. Kukreti, Oxidative Stress: A Key Modulator in Neurodegenerative Diseases, Molecules, 2019, 24(8), 1583.
57. A. Perelman, C. Wachtel, M. Cohen, S. Haupt, H. Shapiro and A. Tzur, JC-1: alternative excitation wavelengths facilitate mitochondrial membrane potential cytometry, Cell Death Dis., 2012, 3(11), e430.
58. Y. Nakabeppu, Regulation of intracellular localization of human MTH1, OGG-1, and MYH proteins for repair of oxidative DNA damage, Prog. Nucleic Acid Res. Mol. Biol., 2001, 68, 75–94.
59. M. A. Graziewicz, M. J. Longley and W. C. Copeland, DNA polymerase gamma in mitochondrial DNA replication and repair, Chem. Rev., 2006, 106, 383–405.
60. X. Chen, X. Zhao, H. Cai, H. Sun, Y. Hu, X. Huang, W. Kong and W. Kong, The role of sodium hydrosulfide in attenuating the aging process via PI3K/AKT and CaMKKβ/AMPK pathways, Redox Biol., 2017, 12, 987–1003.

---

Footnotes

† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d4bm01570c

‡ These authors contributed equally.

---