Human neuroblastoma SH-SY5Y cell line was purchased from the Metson Cell Bank (M23SJ3004, MeisenCTCC, Jinhua, Zhejiang, China), with Short Tandem Repeat (STR) identification supplied by its partner Genetic Testing Biotechnology (Service Order No. 230524B; Suzhou, Jiangsu, China). SH-SY5Y cells were cultured in proliferative medium composed of DMEM/F12 (A4192001, Thermo Fisher Scientific, Waltham, MA, USA) supplemented with 1% penicillin-streptomycin (M23SA1004; 100×, MeisenCTCC) and 12% fetal bovine serum (A5256701, Thermo Fisher Scientific). Cells were incubated at 37 °C in a humidified incubator with a 5% CO₂. The medium was refreshed every 1–2 days to sustain optimal cell growth. Following seeding, the cells were cultured for 2–3 days to reach appropriate confluence, and the entire experimental workflow was completed within 5–7 days. The SH-SY5Y cells used in this study have been validated and tested for mycoplasma contamination. The test was performed using Hoechst DNA stain (indirect) method, Agar culture (direct) method, PCR-based assay, and the result was negative.

After 24 h of cell seeding, cells that had achieved 30% confluence were added to the culture medium. Transfection with the lentiviral strains LV-p62-OE and LV-OE-Control (H29906 and GL181; Heyuan Biotechnology, Wuhan, Hubei, China), and LV-p62-shRNA and LV-shRNA-Control (69956-11 and CON313; Shanghai Gene Chem Co., Ltd., Shanghai, China) was performed, as directed by the manufacturer. The cell culture medium was changed 16 h post-transfection to encourage growth.

The SH-SY5Y cells were cultured for 2–3 days until reaching approximately 70% confluence before OGD/R induction. Prior to OGD treatment, cells were rinsed with phosphate-buffered solution (PBS) (10010001, Thermo Fisher Scientific) and transferred to serum-free glucose-free DMEM (11966025, Thermo Fisher Scientific) with no additional supplements. The cells were then placed in a tri-gas incubator (51033720, Thermo Fisher Scientific) and cultured for 2 h, with 95% N₂ and 5% CO₂ to induce OGD. This 2-h OGD duration and subsequent 24-h reoxygenation period are consistent with established protocols [6, 14, 15, 16] and effectively mimic the ischemic-reperfusion pathological microenvironment in cerebral ischemia, inducing ferroptosis and neuronal damage in SH-SY5Y cells that recapitulate in vivo CIRI features. After 2 h of OGD treatment, the culture medium was renewed with complete DMEM/F12, supplemented with 1% penicillin-streptomycin and 12% fetal bovine serum, for full reoxygenation and glucose restoration.

The experiment involved eight groups: treatment details are presented in Table 1. For drug administration, inhibitors and activators were added twice: the first addition was 24 h prior to OGD/R induction, then the second addition was 2 h after OGD/R establishment. Cells in the lentiviral intervention groups were transfected 72 h before OGD/R modeling as previously described.

Cells were plated in 96-well plates at 5 $\times$ 10³ cells/well and subjected to synchronized cell culture, OGD/R, and drug treatment. Cell counting kit 8 (MA0218-1; Dalian Meilun Biotech Co., Ltd., Dalian, Liaoning, China) and a lactate dehydrogenase (LDH) kit (060319190906; Shanghai Beyotime Biotechnology Co., Ltd., Shanghai, China) were used according to the manufacturers’ guidelines to determine cell viability and the rate of LDH leakage, respectively.

Quantification of relative reactive oxygen species (ROS) levels was performed with a ROS assay kit (Cat# 120920210208, 100 assays, Shanghai Beyotime Biotechnology). After capturing images from three random, non-overlapping fields per group, fluorescence levels were analyzed using ImageJ (version 1.54s, NIH, Bethesda, MD, USA) to quantify ROS production, and the values were calculated and normalized to those of the sham group.

The kit of Image-iT® lipid peroxidation (C10445; Thermo Fisher Scientific) was used, following the manufacturer’s protocol, to quantify the relative intracellular lipid peroxidation levels of the cells. In each group, images of three randomly selected non-overlapping fields of view were analyzed using ImageJ software. The green-red fluorescence ratio was used for calibration, and the fluorescence value was calculated relative to that measured in the sham group.

Cells were seeded in 24-well plates. Before subsequent detection, they were fixed with 4% paraformaldehyde (P0099-500mL, Shanghai Beyotime Biotechnology), and blocked with Immunol Staining blocking buffer (P0102; Shanghai Beyotime Biotechnology) at 37 °C for 1 h. Primary antibodies (Nrf2: AF0639; GPX4: DF6701; ACSL4: DF12141; Affinity Biosciences, Cincinnati, OH, USA) were added at a dilution of 1:200 and incubated overnight at 4 °C. Then the secondary antibodies (anti-rabbit FITC: P0186; anti-rabbit Alexa Fluor 555: P0179; Shanghai Beyotime Biotechnology) were added at a 1:200 dilution and incubated for 1 h in the dark at 37 °C, followed by DAPI (P0131; Shanghai Beyotime Biotechnology) staining for 10 min. Images were captured via inverted fluorescence microscope (Axio Vert.A1, Carl Zeiss, Jena, Germany), and target protein fluorescence ratio to DAPI was quantified via ImageJ, normalized to the sham group.

After blocking, slides were co-incubated with p62 (Rabbit; 23214s; Cell Signaling Technology, Danvers, MA, USA) and Keap1 (Mouse; MA5-17106; Thermo Fisher Scientific) primary antibodies diluted 1:100 at 4 °C overnight. Following three rinses with PBS, species-specific fluorophore-conjugated secondary antibodies (FITC-anti-rabbit IgG, Cat. No. S0008, and Cy3-anti-mouse IgG, Cat. No. S0012; Affinity Biosciences; 1:200) were added at a 1:200 dilution. The fluorescence intensities of Keap1 and p62 in the cells were quantified using ImageJ software, and the fluorescence intensity ratio of p62 to Keap1 was estimated.

Cells were rinsed with pre-cooled PBS, lysed with pre-cooled RIPA buffer (P0013B, Beyotime Biotechnology) (1% Triton X-100, 150 mM NaCl, 50 mM Tris-HCl pH 7.4, 0.1% Sodium Dodecyl Sulfate (SDS), 1 mM Ethylenediaminetetraacetic Acid (EDTA), 1% sodium deoxycholate, 1$\times$ protease inhibitor cocktail (P1005; Beyotime Biotechnology) on ice for 5 min, and centrifuged at 14,000 $\times$g for 10 min. Supernatant protein concentration was measured via BCA assay (P0012S, Beyotime Biotechnology). 500 µL lysate (50 µL reserved as input) was incubated with 1–5 µg antibody at 4 °C for 24 h, mixed with pre-treated protein A agarose beads, and incubated overnight at 4 °C with shaking. After centrifugation at 3000 rpm for 3 min at 4 °C, 15 µL 2$\times$ SDS sample buffer was added, and proteins were denatured at 100 °C for 5 min. SDS-PAGE and Western blot were performed, and bands were analyzed via ImageJ grayscale analysis.

One-hundred and eighty male SPF Sprague–Dawley rats (220–240 g; Hunan Slake Jingda Experimental Animal Co., Ltd, Changsha, Hunan, China) were housed in SPF facility (12 h light/dark, 25 $\pm$ 1 °C, 60% humidity, 0.3 m/s wind speed) for 1 week acclimation; animals were fasted for 12 h before the procedure with free access to water. Ethical approval was obtained from the Animal Ethics Committee of Hunan University of Chinese Medicine. Rats were euthanized with pentobarbital sodium (P3761, Sigma-Aldrich, Merck) at 160 mg/kg via intraperitoneal injection, in accordance with the 2020 AVMA Guidelines for the Euthanasia of Animals.

Rats were randomly allocated into 6 groups: sham, MCAO/R, p62 negative control (LV-p62-NC), p62 overexpression (LV-p62-OE), sulforaphane (Nrf2 activator), and ferrostatin-1 (Fer-1, ferroptosis inhibitor, 5 µM) (HY-100579; MedChemExpress, Monmouth Junction, NJ, USA; 2 mg/kg) groups.

MCAO/R surgery was performed based on a previously established protocol [17]. Rats were anesthetized with 2% pentobarbital sodium (40 mg/kg, i.p.), a 1 cm neck midline incision was made, and right common, external, and internal carotid arteries were exposed. A monofilament (2636A2; Beijing Cinontech, Beijing, China) was inserted via external carotid artery to occlude middle cerebral artery for 2 h, followed by 24 h reperfusion. The sham group underwent neck incision and vessel isolation without occlusion.

Sulforaphane (4478-93-7; MedChemExpress, Monmouth Junction, NJ, USA; 5 mg/kg) was administered intraperitoneally to rats in the corresponding group 20 min following surgery. Fer-1 (HY-100579; MedChemExpress, Monmouth Junction, NJ, USA; 2 mg/kg) was intraperitoneally administered to rats in the Fer-1 group 2 h prior to the operation; equivalent saline volumes were administered to sham and MCAO/R controls.

The Sqstm1 (NM_175843) lentiviral vector was produced and assembled by Shanghai Gene Chem Co., Ltd. (GENE_ID: 113894; carrier name: GV707; component sequence: CMV enhancer-MCS-EF1a-ZsGreen1-T2A-puromycin). Rats were anesthetized with pentobarbital sodium (P3761; Sigma-Aldrich, Merck; 40 mg/kg, i.p.) and fixed in a stereotaxic frame (DW-2000D, Chengdu Techman Software Co., Ltd., Chengdu, Sichuan, China). LV-p62-OE was injected slowly at the injection point (DV = –3.5 to 4.0, ML = +2.0, AP = –1.0) in accordance with the atlas, at 1 µL/min with the Paxinos and Watson Atlas brain maps; the needle remained in position for 10 min. The LV-p62-OE group received a 5 µL injection of p62 lentivirus into the lateral ventricle 3 weeks before surgery, while the LV-Control group received a 5 µL injection of unloaded lentivirus at the same time. p62 lentiviral injection was evaluated 1, 2, and 3 weeks before surgery; 3 weeks was selected as the optimal time point.

The Garcia neurological score was selected as it is a well-validated, widely used tool for rodent cerebral ischemia models, comprehensively evaluating CIRI-related functional domains with high inter-rater reliability to objectively quantify neurological recovery [18].

Following 24 h of reperfusion, Garcia neurological scores were assessed by well-trained investigators who were blinded to the experimental grouping and not involved in subsequent data processing. Details of Garcia neurological function scoring are provided in Table 2.

Before surgery, all rats underwent open field test training twice daily for 3 days. All animals were placed in the bottom central compartment of the box and monitored for 5 min. To eliminate the scent of the prior animal, the interior of the box was disinfected with 75% alcohol between test subjects. Video recordings were used to measure the overall distance covered by the rat in its horizontal movement and the total number of times it entered the core area.

Brains were harvested post-euthanasia, frozen at –20 °C for 30 min, sectioned into 1 mm coronal slices, incubated in 2,3,5-triphenyltetrazolium chloride (TTC) (BCBP3272V, Sigma-Aldrich, Merck, St. Louis, MO, USA) solution for 30 min, and fixed in 4% paraformaldehyde at 4 °C overnight. Slices were photographed sequentially.

GSH, Fe²⁺, MDA, and 4-HNE levels were detected using ferrous iron colorimetric assay kit (FU0262ZV0943; Elabscience, Houston, TX, USA), MDA assay kit (A003-1-2, Nanjing Jiancheng Bioengineering Institute, Nanjing, Jiangsu, China), GSH assay kit (A006-2-1, Nanjing Jiancheng Bioengineering Institute), and 4-HNE ELISA kit (H268-1-1; Nanjing Jiancheng Bioengineering Institute) per manufacturers’ instructions. Absorbance was detected via enzyme marker, and contents were calculated accordingly.

Tissues and cells were lysed, proteins were quantified and denatured. 30 µg cell protein and 70 µg tissue protein were loaded onto SDS-PAGE gels. Nuclear extracts were validated with Lamin B1 (#4777S; Cell Signaling Technology, Danvers, MA, USA; 1:1000), cytoplasmic extracts with $\beta$-actin (1:10,000; A5441, Sigma-Aldrich, St. Louis, MO, USA). Membranes were incubated with primary antibodies (p62: 1:1000; 23214s; Cell Signaling Technology; Keap1: 1:1000; AF5266; Affinity Biosciences; Nrf2: 1:1000; AF0639; Affinity Biosciences; Cincinnati, OH, USA; GPX4: 1:1000; ab125066; Abcam; Cambridge, UK; $\beta$-actin: 1:10,000; A5441; Sigma-Aldrich) at 4 °C overnight, followed by secondary antibodies (goat anti-mouse: 1:10,000; AP130P; Merck Millipore, Billerica, MA, USA; goat anti-rabbit: 1:10,000; AP132P; Merck Millipore) at room temperature for 1 h. Bands were visualized via ECL, and relative expression was calculated as target protein IOD/$\beta$-actin IOD via ImageJ.

All data are presented as mean $\pm$ standard deviation. IBM SPSS Statistics software for Windows version 26 (IBM Corp., Armonk, NY, USA) was used for statistical analyses. One-way analysis of variance was used to compare means across several groups for normally distributed variables. The least significant difference test was used to compare the means among groups that showed the homogeneity of variance. Tamhane’s T2 test was applied when the homogeneity of variance assumption was violated, while non-parametric tests were used for non-normally distributed variables. Statistical significance was set at p 0.05.

To clarify the statistical comparisons across all figures, the following standardized symbols are used consistently throughout the manuscript:

*p 0.05, **p 0.01 versus the sham group;

#p 0.05, ##p 0.01 versus the OGD/R or MCAO/R group;

&p 0.05, &&p 0.01 versus the groups receiving Nrf2 activator tBHQ or Nrf2 inhibitor ML385 intervention.

Exposure of SH-SY5Y cells to OGD/R followed by reperfusion markedly diminished cell viability (Fig. 1A), concomitant with elevated lactate dehydrogenase release (Fig. 1B) and intracellular Fe²⁺ accumulation (Fig. 1C). Additionally, GSH levels were decreased (Fig. 1D), while those of MDA, ROS, and lipid peroxidation increased (Fig. 1E–I). Administering the ferroptosis inhibitor ferrostatin-1 (Fer-1, 5 µM) and the ferroptosis activator RAS-selective lethal 3 (RSL3, 50 nM) produced the opposite effects, indicating that OGD/R triggered ferroptosis in SH-SY5Y cells.

Fig. 1.

Nrf2 inhibits OGD/R-induced ferroptosis of SH-SY5Y cells. (A) Cell viability and (B) LDH release levels. Data are presented as mean $\pm$ standard deviation (n = 3). (C) Fe²⁺, (D) GSH, and (E) MDA contents (n = 4). (F) Intracellular ROS levels, as quantified by DCFH-DA staining (n = 4). (G) Representative images of DCFH-DA staining (n = 4). (H) Representative images of lipid peroxidation in SH-SY5Y cells (n = 3). (I) Quantitative analysis of lipid peroxidation levels (n = 3). **p 0.01 versus sham; *p 0.05 versus sham; ##p 0.01 versus OGD/R; #p 0.05 versus OGD/R; &&p 0.01 versus tBHQ and ML385; &p 0.05 versus tBHQ and ML385. LDH, lactate dehydrogenase; GSH, glutathione; MDA, malondialdehyde; ROS, reactive oxygen species; DCFH-DA, dichloro-dihydro-fluorescein diacetate; tBHQ, tert-butylhydroquinone. Scale bar = 20 µm.

Pharmacological activation of Nrf2 with tert-butylhydroquinone (tBHQ, 40 µM) (HY-100489, MedChemExpress, Monmouth Junction) markedly enhanced cell survival and glutathione content, concomitant with diminished LDH release, Fe²⁺ accumulation, MDA generation, ROS production, and lipid peroxidation relative to OGD/R controls. Conversely, administration of the Nrf2 antagonist ML385 (20 µM) (HY-100523, MedChemExpress, Monmouth Junction) elicited opposing effects across all measured parameters (Fig. 1A–I), implicating Nrf2 as a negative regulator of ferroptotic cascades in this cellular model.

Notably, the ferroptosis inducer RSL3 partially attenuated the cytoprotective effects of tBHQ, whereas the ferroptosis inhibitor ferrostatin-1 partially abrogated the exacerbated injury phenotype resulting from ML385 treatment. These observations collectively support the notion that Nrf2 confers protection against OGD/R-induced neuronal injury through suppression of ferroptotic execution.

To further explore the mechanism by which Nrf2 modulates ferroptosis under OGD/R, we studied the key regulatory proteins level involved in ferroptosis. OGD/R treatment resulted in a remarkable reduction in GPX4 protein expression and a concomitant elevation in ACSL4 protein levels in SH-SY5Y cells, as shown by WB and immunofluorescence measurements. However, Fer-1 (5 µM) and RSL3 (50 nM) (HY-100218A, MedChemExpress, Monmouth Junction, NJ, USA) counteracted and intensified these alterations, respectively. Following treatment with tBHQ (40 µM), ACSL4 levels were decreased and GPX4 levels increased. In contrast, treatment with ML385 (20 µM) had the opposite effects. Furthermore, the combination of ML385 + Fer-1 and tBHQ + RSL3 attenuated these effects (Fig. 2A–G). The original western blotting images can be found in the Supplementary Materials.

Fig. 2.

Effect of Nrf2 on ferroptosis-related protein levels in SH-SY5Y cells treated with OGD/R. (A) Western blotting was performed to detect the expression levels of GPX4 and ACSL4 (n = 3). (B,C) Quantitative analysis of GPX4 and ACSL4 expression, normalized to the internal control $\beta$-actin. Data are presented as mean $\pm$ standard deviation (n = 3). (D,E) Representative immunofluorescence images and corresponding quantitative analysis of GPX4 (n = 3). (F,G) Representative immunofluorescence images and corresponding quantitative analysis of ACSL4 (n = 3). (H,I) Representative images and quantitative analysis of Nrf2 nuclear translocation and fluorescence intensity (n = 3). **p 0.01 versus sham; ##p 0.01 versus OGD/R; #p 0.05 versus OGD/R; &&p 0.01 versus tBHQ and ML385; &p 0.05 versus tBHQ and ML385. ACSL4, acyl-CoA synthetase long-chain family member 4. Scale bar = 20 µm.

Through a series of biological processes including nuclear translocation, interaction with antioxidant response elements, activation of downstream antioxidant gene transcription, and regulation of critical proteins implicated in ferroptosis, Nrf2 inhibits ferroptosis and attenuates cellular injury. Therefore, in the OGD/R model, we performed immunofluorescence staining to analyze the nuclear translocation of Nrf2. The findings revealed that OGD/R promoted the nuclear translocation of Nrf2. This was further intensified by tBHQ treatment, but inhibited by ML385 (Fig. 2H,I). Together, these findings indicate that Nrf2 inhibits ferroptosis in OGD/R-treated SH-SY5Y cells by up-regulating GPX4 and down-regulating ACSL4 after entering the nucleus.

Based on the fluorescence images of the lentivirus (LV)-infected cells, our multiplicity of infection (MOI) pre-experimental data, and previous findings [16], LV-p62-OE with MOI = 40 and LV-p62-shRNA with MOI = 20 were selected as optimal conditions for infection efficiency (Fig. 3A). The significant overexpression of p62 by LV-p62-OE and the knockdown efficiency of p62 by LV-p62-shRNA were confirmed via immunofluorescence and Western blotting (Fig. 3A–C).

Fig. 3.

Validation of p62 lentivirus transfection of SH-SY5Y cells. (A) EGFP autofluorescence in SH-SY5Y cells after lentiviral transfection with p62 with different MOIs for 72 h. (B) Protein expression of p62 detected using western blotting. (C) Quantitative analysis of p62 protein expression levels, normalized to the internal control $\beta$-actin. Data are presented as mean $\pm$ standard deviation (n = 3). **p 0.01 versus sham. Scale bar = 20 µm.

Some research [12, 13] has demonstrated that p62 interacts with Keap1 directly to control the synthesis of Nrf2 in response to oxidative stress. Consequently, we investigated whether p62 inhibits ferroptosis by activating Nrf2, which helps reduce damage. Cell viability assays revealed that LV-p62-OE and tBHQ (40 µM) increased cell viability after OGD/R treatment, while LV-p62-shRNA and ML385 (20 µM) diminished it (Fig. 4A). Additionally, treatment with ML385 and tBHQ partially counteracted the effects of LV-p62-OE and LV-p62-shRNA, respectively. LV-p62-OE enhanced intracellular GSH and GPX4 levels and decreased Fe²⁺, MDA, and ACSL4 levels after OGD/R, while ML385 (20 µM) had the opposite effects. Further, treatment with ML385 partially blocked the effects of LV-p62-OE, suggesting that Nrf2 mediated the influence of p62 in the p62/Nrf2/Keap1 pathway and that overexpression of p62 inhibited ferroptosis by controlling Nrf2 after OGD/R (Fig. 4B–G).

Fig. 4.

p62 interacts with Keap1 to activate Nrf2 and inhibit ferroptosis. (A) Cell viability. Values are shown as mean $\pm$ standard deviation (n = 3). **p 0.01 versus sham; ##p 0.01 versus OGD/R; &&p 0.01 versus tBHQ and ML385. (B) GSH, (C) Fe²⁺, and (D) MDA levels (n = 3). Quantitative analysis of (E) GPX4 and (F) ACSL4 relative to $\beta$-actin protein expression levels. Values are shown as mean $\pm$ standard deviation (n = 3). **p 0.01 versus sham; ##p 0.01 versus OGD/R; #p 0.05 versus OGD/R; &&p 0.01 versus ML385; &p 0.05 versus ML385. (G) Expression of GPX4 and ACSL4 proteins was detected using WB. (H) Representative images of Nrf2 nuclear localization were captured by immunofluorescence, and (I) the fluorescence intensity was quantitatively analyzed. Values are shown as mean $\pm$ standard deviation (n = 3). **p 0.01 versus sham; #p 0.05 versus OGD/R; &&p 0.01 versus OGD/R+LV-p62-OE. (J) Representative immunofluorescence double staining of p62 and Keap1 and (K) quantitative ratio analysis of the immunofluorescence intensity of p62 and Keap1. Values are shown as mean $\pm$ standard deviation (n = 3). **p 0.01 versus sham; ##p 0.01 versus OGD/R; &&p 0.01 versus OGD/R+LV-p62-OE. (L) Interaction between p62 and Keap1 detected using co-immunoprecipitation. Scale bar = 20 µm.

Next, we explored the possible interaction between p62 and Keap1, as well as the consequent activation of Nrf2. Immunofluorescence results revealed that the nuclear translocation of Nrf2 following OGD/R was markedly enhanced after LV-p62-OE transfection and notably reduced by ML385 treatment (Fig. 4H,I). Furthermore, immunofluorescence double staining revealed a decrease in p62 expression following OGD/R, along with a reduced fluorescence intensity ratio of p62 to Keap1. In contrast, the fluorescence ratio of p62 relative to Keap1 was significantly increased following LV-p62-OE transfection (Fig. 4J,K), suggesting that overexpression of p62 promoted its binding to Keap1 following OGD/R.

Co-immunoprecipitation (Co-IP) experiments demonstrated an enhanced association between p62 and Keap1 after OGD/R, which was intensified further after LV-p62-OE transfection (Fig. 4L). Conversely, ML385 suppressed the interaction between Keap1 and p62 and reversed the effects of LV-p62-OE. Overexpression of p62 may activate Nrf2 via binding to Keap1, thereby suppressing ferroptosis and thus mitigating cell injury; Nrf2 may mediate the activity of p62 through a positive feedback loop of the p62/Keap1/Nrf2 signaling pathway. These findings show that p62 overexpression promotes p62 binding to Keap1, activating Nrf2, which inhibits cellular ferroptosis.

To investigate how the p62/Keap1/Nrf2 pathway regulate ferroptosis in rats with CIRI, an MCAO/R animal model was established. This pathway was activated by the intraperitoneal injection of Fer-1 and sulforaphane, and injection of LV-p62-OE into the lateral ventricle.

Western blotting and immunofluorescence revealed a marked increase in p62 protein expression levels in the brain tissue of SD rats following LV-p62-OE transfection (Fig. 5A–C). Protein expression analysis revealed that MCAO/R decreased levels of cytoplasmic Nrf2 and p62 and increased those of Keap1 and total Nrf2. Conversely, LV-p62-OE transfection and treatment with sulforaphane resulted in elevated expression of total Nrf2 and p62, along with decreased expression of Keap1 and cytoplasmic Nrf2 (Fig. 5D–J). These findings corroborate the outcomes of the in vitro experiments and reveal that p62 overexpression can boost the p62/Keap1/Nrf2 pathway.

Fig. 5.

Effect of MCAO/R and sulforaphane, and LV-p62-OE treatment on the p62/Keap1/Nrf2 pathway. (A) Representative images of lentivirus autofluorescence captured three weeks after the transfection of LV-p62-OE into the lateral ventricle. (B) Expression of p62 after transfection as detected by western blotting. (C) Quantitative analysis of p62 relative to $\beta$-actin protein expression levels. Values are shown as mean $\pm$ standard deviation (n = 3). **p 0.01 versus sham. (D) Protein expression of p62 in the sham, MCAO/R, LV-Control, and LV-p62-OE groups detected using western blotting. (E) Quantitative analysis of p62 relative to $\beta$-actin protein expression levels. Values are shown as mean $\pm$ standard deviation (n = 3). **p 0.01 versus sham; #p 0.05 versus MCAO/R. (F) Expression levels of Keap1, total Nrf2, cytoplasmic Nrf2, and nuclear Nrf2 detected using western blotting. Nuclear extracts were validated using Lamin B1 as the nuclear-specific loading control, and cytoplasmic extracts were validated using $\beta$-actin as the cytoplasmic-specific loading control to confirm fractionation purity. Quantitative analysis of the protein expression levels of (G) Keap1, (H) total Nrf2, (I) cytoplasmic Nrf2, and (J) nuclear Nrf2 relative to Lamin B1. Values are shown as mean $\pm$ standard deviation (n = 3). **p 0.01 versus sham; *p 0.05 versus sham; ##p 0.01 versus MCAO/R; #p 0.05 versus MCAO/R. Scale bar = 100 µm.

Following treatment with MCAO/R in the brain injury test, TTC staining showed large white infarcts in the right region of the coronal section of the rat brains (Fig. 6A,B). Garcia neurological function scores declined dramatically, and the total distances traveled by rats in the open field experiments decreased. Fewer entries into the central area were also recorded. In contrast, interventions involving LV-p62-OE, sulforaphane, and Fer-1 led to a substantial decrease in TTC infarct size, together with improvements in both the Garcia score and total distance traveled (Fig. 6C–E). The activation of the p62/Keap1/Nrf2 pathway may, therefore, improve neurological dysfunction and attenuate CIRI in rats.

Fig. 6.

Activation of the p62/Keap1/Nrf2 pathway inhibits ferroptosis and alleviates cerebral ischemia-reperfusion injury in rats. (A) 2,3,5-triphenyltetrazolium chloride staining detects infarcts in the brain slices; the white area indicates the infarct area. (B) Quantitative analysis of cerebral infarction volume (n = 3). (C) Garcianeural function score. Values are shown as mean $\pm$ standard deviation (n = 6). (D) Total distance and (E) number of times entering the center area in the open field test (n = 6). Detection of (F) GSH, (G) Fe²⁺, (H) MDA, and (I) 4-HNE levels in rat brain tissue (n = 3). (J) Expression of GPX4 and ACSL4 detected using western blotting. Quantitative analysis of (K) GPX4 and (L) ACSL4 relative to $\beta$-actin protein expression levels (n = 3). **p 0.01 versus sham; ##p 0.01 versus MCAO/R; #p 0.05 versus MCAO/R. Scale bar = 1 cm.

Finally, in the ferroptosis assay, GSH and GPX4 levels were reduced dramatically after MCAO/R, while Fe²⁺, MDA, 4-HNE, and ACSL4 levels increased. These effects were reversed by LV-p62-OE, sulforaphane, and Fer-1 (Fig. 6F–L), suggesting that activation of the p62/Keap1/Nrf2 pathway is capable of inhibiting ferroptosis and mitigating cellular damage.

In summary, p62 overexpression activated the pathway, suppressed ferroptosis in rat brain tissue following MCAO/R, and lessened brain injury.