NQ was prepared by Guangdong Pharmaceutical University, and the botanical drugs in NQ were obtained from the Affiliated Traditional Chinese Medicine Hospital of Guangzhou Medical University, Musk was provided by Foshan Dezhong Pharmaceutical. Naoqing formula consists of drugs: Ligusticum striatum DC. [Ligusticum chuanxiong hort], Panax notoginseng (Burkill) F.H.Chen [Araliaceae; Panax notoginseng]and Mentha canadensis L. [Lamiaceae; dementholised mint oil], Musk, Dryobalanops aromatica C.F.Gaertn. [Dipterocarpaceae; Dipterocarpus camphora], Acorus tatarinowii Schott [Acori Tatarinowii Rhizoma]. Mentha canadensis L, at a ratio of 29:29:3.3:3.3:1:1 (Table 1). Oil and Acorus tatarinowii Schott Oil were extracted with triple-distilled water, Panax notoginseng (Burkill) F.H.Chen and Ligusticum striatum DC. were extracted with ethanol (separated and adsorbed with a macroporous resin), Musk and Dryobalanops aromatica C.F. Gaertn. were dissolved in ethanol. The drugs used in the in vitro cell assay were freeze-dried drugs, and the freeze-drying conditions were: refrigeration at 40°C, freezing at −20°C for 2 h, freezing at −10°C for 16 h, drying at 20°C for 36 h, and secondary drying at 35°C for 36 h. Weigh the freeze-dried NQ, dissolve it with DMSO to a final concentration of 10 mg/mL, centrifuge at 15000 rpm for 10 min, and filter the supernatant with a 0.22 μL filter membrane for later use.

Batch consistency was monitored by UPLC-QE-MS and GC-MS analysis of NQ major components and GC-MS quantitative analysis of muscone. Determination of components in NQ was performed on an UHPLC system (Vanquish, Thermo Fisher Scientific) with a Waters UPLC BEH C18 column (1.7 μm 2.1*100 mm). The sample injection volume was set at 5 μL. The flow rate was set at 0.5 mL/min. The mobile phase consisted of 0.1% formic acid in water (A) and 0.1% formic acid in acetonitrile (B). The multi-step linear elution gradient program was as follows: 0–11 min, 85%–25% A; 11–12 min, 25%–2% A; 12–14 min, 2%–2% A; 14–14.1 min, 2%–85% A; 14.1–15 min, 85%–85% A; 15–16 min, 85%–85% A. The main volatiles in NQ were detected using GC-MS. In SPME cycle of PAL rail system. Incubate temperature is 30°C; Preheat time in 15 min; Incubate time in 30 min; Desorption time is 4 min.GC-MS analysis was performed using an Agilent 7890 gas chromatograph system coupled with a 5977 B mass spectrometer. The system utilized a DB-Wax. Injected in Splitless Mode. Helium was used as the carrier gas, the front inlet purge flow was 3 mL min⁻¹, and the gas flow rate through the column was 1 mL min⁻¹. The initial temperature was kept at 40°C for 4 min, then raised to 245°C at a rate of 5 °C min⁻¹, kept for 5 min. The injection, transfer line, ion source and quad temperatures were 250, 250, 230°C and 150°C, respectively. The energy was −70 eV in electron impact mode. The mass spectrometry data were acquired in scan mode with the m/z range of 20–400, solvent delay of 2.37 min. The determination of each metabolite in NQ, quality control of muscone, fingerprinting methods for drugs were listed in Supplementary Material S1.

Male C57BL/6 mice weighing 21–25 g were housed in certified pathogen-free conditions, maintaining a light/dark cycle (25°C ± 2°C and 60% – 70% relative humidity). Adequate food and drink were provided. The Animal Laboratory of Guangzhou Topbioteeh Co. Ltd.’s Ethics Committee approved all pertinent operational research, which was overseen by the Animal Experiment Center (licence number: SYXK(YUE)2023-0315).

The mice were anaesthetised with 2% pentobarbital sodium and placed in a supine position. Sequential dissection of the right common carotid artery (CCA), external carotid artery (ECA) and internal carotid artery (ICA) was performed by carefully dissecting the arteries through a median incision in the mice’s anterior skin. The CCA and ECA were ligated step by step while simultaneously clamping the ICA with an arterial clamp. A silicone nylon monofilament was introduced through the CCA into the middle cerebral artery and secured temporarily. After 1 hour of ischaemia, the monofilament was removed, and the blood vessels were tied at the incision site. Surgical sutures were employed to close the neck wound. After 24 h of reperfusion, further experiments were conducted. Sham-operated mice underwent identical procedures without monofilament insertion. The cerebral cortex blood flux (CBF) was assessed using laser speckle contrast imaging/LSCI (RWD Life Science, China) before, during and after the operation to confirm MCA obstruction and blood flow changes.

In this study, the clinical dose of NQ administered was 14.26 mg/kg/day, equating to 1 mL of nasal spray per day. To translate this dose to mice, we employed the formula: equivalent dose in mice (mg/kg) = 9.01 × human dose (mg/kg), resulting in a dose of 128.34 mg/kg. Consequently, we adjusted the NQ doses for mice to 130 mg/kg/day (low), 260 mg/kg/day (medium), and 520 mg/kg/day (high). The mice were randomly assigned to seven groups: Sham operated, Sham + 520 mg/kg NQ, MCAO, 130 mg/kg NQ, 260 mg/kg NQ, 520 mg/kg NQ, 10 mg/kg Edaravone (EDA) (positive control drug), Nrf2 inhibitor ML385 (30 mg/kg), and Nrf2 inhibitor ML385 (30 mg/kg) combined with 520 mg/kg NQ. ML385 and was dissolved in a solution comprising 5% DMSO, 40% PEG300, 5% Tween80, and 50% ddH₂O. NQ was administered via nasal instillation, while ML385 and EDA were injected intraperitoneally. A total of seven administrations were given over 6 days before ischaemia/reperfusion, with the final dose administered 1 hour before the surgical procedure.

SH-SY5Y neuroblastoma cells were obtained from Pricella (CL-0208, China) and cultured in F12 medium (Gibco, Gaithersburg, United States) containing 15% FBS (ExcellBio, FSD500) and 1% penicillin-streptomycin (Gibco Gaithersburg, United States) at 37°C in a humidified 5% CO₂ environment. The ferroptosis metamorphosis model was induced using Erastin (Selleck).

For the remaining experiments in this study, GraphPad Prism 9.0 software was utilized for statistical analyses. The data presented feature a minimum of n = 6 per group, expressed as the mean with the standard error of the mean (SEM). Each experiment was replicated at least three times, ensuring the inclusion of at least three independent individuals per group. Additionally, a minimum of three independent replicate experiments, encompassing both Western blot (WB) and staining experiments, were conducted. For comparing means, an unpaired two-tailed t-test was employed, while a one-way ANOVA was used to assess differences in experiments involving multiple groups. Statistical significance was determined when p-values were less than 0.05, annotated as *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001. The sample size for each experimental condition is indicated using symbols in the accompanying figures. The remaining experimental methodology was described in Supplementary Material S2.

The chemical metabolites in NQ samples were analyzed using UHPLC-QE-MS and GC-MS, and the negative and positive ion chromatograms were shown in Supplementary Material S1. The main active metabolites in NQ were muscone, cholesterol, endo-borneol, isoborneol, ligustilide, beta-asarone, ginsenoside rg5 and (−)-menthone. The quality control of muscone content in NQ were shown in Supplementary Material S1. To verify the role of NQ in AIS, a MCAO-induced AIS model was used, and CBF was assessed in AIS mice. There was no significant change in CBF before surgery, after 60 min of ischemia and after 60 min of reperfusion in the Sham group and the Sham + NQ group, while the CBF of the MCAO group and the NQ group decreased at 60 min of ischemia, and the CBF recovered at 60 min of reperfusion but did not return to normal levels (Figures 1A, B), indicating that NQ did not cause adverse reactions to the blood flow of mice, and MCAO could successfully block the blood flow and cause AIS. Compared with the sham group, the CBF of the ischemic side and the contralateral side of the MCAO group decreased after 60 min of reperfusion (Figure 1C). Pretreatment with 520 mg/kg NQ reduced the degree of bilateral CBF infarction at 60 min of reperfusion compared with the MCAO group (Figure 1C), and EDA increased the CBF blood flow of the ischemic side at 60 min of reperfusion compared with the MCAO group, but the improvement was not as good as that of 520 mg/kg NQ (Figure 1C). Nissl staining showed that after MCAO treatment, a large number of neurons in the cerebral cortex and hippocampus (CA1, CA3 and DG) were disordered, the cells were dissolved, and the nuclei were shrunken; after pretreatment with 520 mg/kg NQ, the neurons were neatly arranged, the nuclei were centered, the degree of necrosis was reduced, the Nissl vesicles were well preserved, and the cell structure was better than that of the MCAO group. EDA protected the integrity of some Nissl bodies, but the protective effect was not as good as that of 520 mg/kg NQ (Figure 1D). HE staining also showed that in MCAO mice, the cells in the cerebral cortex and hippocampus were lysed and the nuclei were shrunken; in mice treated with 520 mg/kg NQ, the cells were arranged neatly, the nuclei were centered, and the degree of necrosis was reduced (Supplementary Material S3). We recorded neurological scores according to the Zea-longa neurobehavioral scoring criteria at the time of postoperative awakening in AIS mice. Behavioral scores were reduced in the NQ cohort compared to the MCAO group, suggesting that NQ promotes the recovery of neurological function in AIS mice (Figure 1E). Furthermore, the area of cerebral infarction induced by MCAO was significantly reduced following NQ pre-treatment, as visualised using TTC staining (Figure 1F). The results confirmer that NQ can effectively improve CBF, reduce cerebral infarction volume, and prevent and treat brain tissue damage and neurological function damage caused by AIS.

Our study found that the dose of 520 mg/kg NQ was optimal for the treatment of AIS in mice. To further investigate the protective mechanism of NQ against IS, we conducted a transcriptomic and proteomic analysis and analyzed the differentially expressed genes (DEGs) and differentially expressed proteins (DEPs) in the IR group and 520 mg/kg NQ group (7 samples in each group). DEGs and DEPs were considered as genes with Foldchange ≥ 2 or Foldchange ≤ 1/2 and p-value < 0.05. The expression patterns of DEGs in the IR group and NQ group are shown by principal component analysis (PCA), revealing differences between them (Figure 2A). NQ-vs-IR upregulated DEGs were enriched in GO, and “Lipid oxidation” was significant at the biological process level (Figure 2B). PCA was used to show the expression patterns of all DEPs, which showed significant differences between the IR group and the NQ group (Figure 2C). The GO enrichment analyzed top 20 histogram of molecular function showed that the DEPs were mainly enriched in “Iron ion binding” and “Oxidoreductase activity” (Figure 2D). Additionally, we also enriched the upregulated genes of NQ-VS-IR in Reactome, among which “(R-MMU-9759194) Nuclear events mediated by NFE2L2” was focused (Figure 2E). Nrf2 plays a crucial role in regulating both iron homeostasis and oxidative stress (Kerins and Ooi, 2018). Combining the results of transcriptomics and proteomics, we believe that NQ ameliorates AIS induced oxidative stress damage and ferroptosis by affecting the Nrf2 mediated ferroptosis.

Lipid peroxidation and iron accumulation are the key events in ferroptosis. To investigate the effect of NQ on ferroptosis, a series of ferroptotic events were examined. In AIS mice, an increase of ferroptotic events was found, including important marker of cellular oxidative damage ROS promotion (Figure 3A), product of lipid peroxidation MDA addition (Figure 3B), important antioxidant enzymes GSH and SOD depletion (Figures 3C, D), as well as Fe³⁺ and Fe²⁺ accumulation (Figures 3E, F). However, ROS, MDA, Fe³⁺, and Fe²⁺ were downregulated and the level of GSH and SOD were increased in ischemic brain tissue after being treated with NQ. Furthermore, the results of transmission electron microscope (TEM) revealed that the mitochondria in the MCAO group exhibited reduced size, increased membrane density and shorter, smaller cristae that moved towards the edge compared to the sham group, indicating that ferroptosis was significantly occurred in the neuronal cells of MCAO-induced AIS. Conversely, there was an improvement in mitochondrial volume crumpling in the NQ treated group (Figure 3G). These results confirmed that NQ inhibits AIS-induced ferroptosis. To further investigate the effect of NQ on ferroptosis, the Nrf2, xCT, and GPX4 were measured. Western blotting and immunohistochemistry staining showed that a significant decrease in xCT and GPX4 levels and an increase in Nrf2 levels was observed in AIS mice, whereas xCT, GPX4 as well as Nrf2 levels were significantly increased, particularly in the 520 mg/kg NQ group mice (Figures 3H, I). It is commonly acknowledged that Nrf2 functions as a pivotal transcription factor in the response to oxidative stress. In homeostatic conditions, Nrf2 is sequestered in the cytoplasm by its binding partner, KEAP1, where it undergoes ubiquitin-dependent degradation mechanisms, thus ensuring the maintenance of its low concentrations. Upon stress, there is a stress-induced increase and translocation of Nrf2 to the nucleus where it exerts antioxidant effects during the acute phase, which explains why there is an increase in Nrf2 in AIS mice. More interestingly, immunofluorescent colocalization indicated that Nrf2 was primarily expressed in the cytoplasm in the sham group, while mainly expressed in the nucleus and colocalizes with xCT in the MCAO and NQ groups (Figure 3J). Furthermore, the expression of Nrf2 and xCT were consistent with the results of immunohistochemistry and Western blotting. These results indicates that NQ potentially stimulates the activation and nuclear translocation of Nrf2, enhancing the expression of xCT and GPX4. This augmentation in expression leads to the exertion of antioxidant stress effects, thus affording protection against ferroptosis in AIS mice.

To further elucidate the role of NQ in ferroptosis, the ferroptosis inducer Erastin was used to incubate SH-SY5Y cells to specifically establish cell ferroptosis model. First, SH-SY5Y cells were exposed to different concentrations of NQ, ranging from 0 to 10 μg/mL, for 24 h. The obtained results showed that within this concentration range, no IC50 value was observed and surface NQ was not toxic to cells (Supplementary Material S3). We determined the optimal concentration of Erastin and NQ-treated cells using the cck8 assay. SH-SY5Y cells were treated with varying concentrations of Erastin ranging from 0 to 50 µM for 24 h. The concentration that inhibited 50% of cell viability, IC50, was achieved at 10 µM. The cells were treated with 10 µM Erastin for 24 h, followed by the addition of NQ. The cells exhibited an IC50 at an NQ concentration of 10 μg/mL, which corresponds to the maximum concentration at which NQ lyophilized powder can dissolve in DMSO (Figure 4A). The EdU fluorescence assay demonstrated decreased cell proliferation in the Erastin group compared to the control. In the NQ group, EdU (green) fluorescence intensity exceeded that of the Erastin group, indicating an increase in cell proliferation (Figure 4B). After Erastin (10 μM) treatment, significant ferroptosis was induced in SH-SY5Y cells with ROS increasing. Attractively, the production of ROS could be abolished in ferroptotic cell by 10 μg/mL NQ (Figure 4C). These findings suggest that NQ is effective in reducing ferroptosis. Additionally, quantifications of Nrf2, xCT, GPX4 and TFR1 protein expression was performed by immunofluorescence and Western blotting after Erastin incubation. The NQ group displayed a significant increase in Nrf2 in the nucleus, indicating the activation of Nrf2’s antioxidant capacity (Figures 4D, H). Moreover, Erastin blocked the xCT and GPX4 expression, and elevated TFR1 expression, implying the alteration of crucial protein levels in ferroptosis. Attractively, NQ completely revised xCT, GPX4 and TFR1 expression in the induction of ferroptosis (Figures 5E–H). Thus, based on the findings above, we suggest that NQ inhibits ferroptosis induced by Erastin by activating Nrf2/xCT/GPX4 pathway.

To investigate the role of Nrf2 in protection against AIS of NQ, Nrf2 inhibitor ML385 was used to inhibit the activation of Nrf2. The CBF decreased at 60 min of ischemia and increased but not to normal levels at 60 min of reperfusion in MCAO, MCAO + ML385, NQ and NQ + ML385 groups (Figures 5A, B). After 60 min of ischemia, both ipsilateral and contralateral CBF in the MCAO + ML385 group were lower than those in the MCAO group (Figure 5C). Following 60 min of reperfusion, the ipsilateral CBF in the MCAO + ML385 group remained lower than that in the MCAO group, indicating that ML385 can reduce cerebral blood flow in AIS mice (Figure 5D). Notably, at both 60 min of ischemia and 60 min of reperfusion, the CBF in the NQ + ML385 group was lower than that in the NQ group, suggesting that inhibition of Nrf2 expression can counteract the NQ-induced increase in CBF (Figures 5C, D). AIS mice exhibited signs of consciousness loss, contralateral paralysis, and inability to walk. However, post-administration of ML385, there was no significant change in the extent of limb damage in both MCAO and NQ mice, which may be because the severity of limb damage is related to whether the brain regions governing motor functions are affected, and less related to other interventional factors (Figure 5E). HE staining revealed that neuronal cytolysis and crumpled nuclei in in the cerebral cortex and hippocampal regions both in MCAO and MCAO + ML385 groups. Compared with the NQ group, the NQ + ML385 group displayed more crumpled nuclei (Figure 5F). Nissl staining revealed a reduction in Nissl vesicles in AIS mice following ML385 intervention in comparison to the MCAO group. Similarly, a reduction in Nissl vesicles was observed in the NQ + ML385 group compared with the NQ group (Figure 5G). The results suggest that inhibition of Nrf2 expression may exacerbate neuronal damage in AIS mice, and the neuroprotective effect of NQ appears to be mediated through the Nrf2 pathway.

To investigate the effect of NQ on ferroptosis after inhibiting Nrf2, a series of ferroptosis events were examined. Western blot showed that the total protein and nuclear protein levels of Nrf2 in AIS mice were significantly increased, while the cytoplasmic protein levels were low. The total Nrf2 protein and nuclear protein levels in the MCAO + ML385 group were significantly lower than those in the MCAO group, and the cytoplasmic protein levels were increased, indicating that activation of Nrf2 was inhibited by ML385 in AIS mice. Compared with the ML385 group, the ML385 + NQ group showed significantly higher levels of total Nrf2 protein and nuclear protein, and lower levels of cytoplasmic protein, suggesting that NQ was still partially able to activate Nrf2 after treatment with ML385 (Figures 6A–C). This also elucidates why NQ continues to exert a certain protective effect on AIS even after inhibition with Nrf2. ROS and Fe³⁺ in the ML385 group were higher than the MCAO group, while the NQ + ML385 group exhibited an accumulation of ROS and Fe³⁺ compared to the NQ group (Figures 6D, E). Immunohistochemical analysis and immunofluorescence techniques revealed a marked downregulation in the expressions of Nrf2, xCT, and GPX4 in AIS mice following ML385 intervention. Similarly, a significant reduction in the expression levels of Nrf2, xCT, and GPX4 was observed in the NQ + ML385 group in comparison to the NQ group alone (Figures 6F, G). These findings indicate that ML385 inhibited the activation of Nrf2, downregulated the expression of xCT and GPX4, and exacerbated oxidative stress and iron deposition induced by MCAO, and weakened NQ’s resistance to ferroptosis. Immunofluorescence co-localisation showed that Nrf2 was expressed in both the nucleus and cytoplasm in the MCAO group, while Nrf2 was mainly expressed in the cytoplasm in the ML385 group. Nrf2 was expressed in both the nucleus and the cytoplasm in the NQ group, while Nrf2 was mainly expressed in the cytoplasm in the NQ + ML385 group. These results indicated that ML385 inhibited Nrf2 activation in AIS mice (Figures 6F, G). In summary, the effect of NQ in attenuating AIS-induced ferroptosis is diminished after inhibition of Nrf2, suggesting that NQ inhibits the occurrence of ferroptosis in AIS mice via activating the Nrf2 pathway.