Introduction
Orthotopic liver transplantation, shock, abdominal surgery, and vascular embolism can cause liver ischemia-reperfusion (IR) injury.1 In addition, as the standard treatment method for end-stage liver diseases, liver transplantation may result in liver tissue ischemia. However, after blood perfusion is restored, organ function may deteriorate further, which seriously restricts the therapeutic effect of liver transplantation.2,3 Therefore, it is extremely important to find ways to pre-adapt the body and reduce injury based on the pathogenesis of IR.4
The mechanism of liver IR injury is complex, involving almost all types of hepatocytes and infiltrating neutrophils, macrophages, and platelets.5 Iron-dependent biological activities may also be involved the pathogenesis of IR injury.6,7 In 2012, Dixon et al.8 studied the mechanism of erastin inducing cell death through the RAS gene mutation and discovered ferroptosis, a new type of programmed cell death mediated by iron ions.9 Research has shown that ferroptosis is an important link in the pathophysiology of IR injury in various organs.10 In the liver, Wu et al.11 demonstrated that the level of ferroptosis increased in patients with hepatectomy and in mice with liver IR, and ferrostatin-1 (Fer-1) reduced liver injury induced by IR. The results showed that regulating ferroptosis to interfere with IR may provide new research ideas and treatment strategies for the treatment of liver IR injury.
Fucoidan, a sulfated polysaccharide from the sea, is mainly present in the cell wall matrix of various brown algae, is widely sourced, and has a variety of biological and pharmacological activities.12–17 Fucoidan does not have significant toxicity in rats, and, volunteers who ingested 3 g of 75% fucoidan capsules for 12 days experienced no side effects or toxicity.18–21 Studies have confirmed that fucoidan supplementation protects liver cells from ferroptotic damage caused by long-term alcohol exposure by inhibiting iron overload and liver oxidative damage.22 In a previous study, we found that fucoidan also had antioxidant activity, so we have reason to believe that it protects against IR-induced injury.14,23 In this study, we used a variety of molecular biological experiments to verify that the injury induced by ferroptosis was reduced by fucoidan activation of the Nrf2/HO-1/GPX4 axis.
Methods
Animals and cell culture
Six- to eight-week-old male Balb/C mice weighing 20–25 g were purchased from Shanghai SLAC Laboratory Animal Co. Ltd. (Shanghai, China). The mice were housed in clean cages with a 12 h alternating light/dark cycle at a constant temperature of 22–25°C and were given free access to food and water until the start of fasting 12 h before surgery. Normal human hepatocyte LO2 cells were purchased from the Cell Bank of Type Culture Collection (Chinese Academy of Sciences, Beijing, China) and cultured in RPMI 1640 medium (Sigma-Aldrich Corp., St. Louis, MO, USA) with 10% fetal bovine serum, 100 IU/mL penicillin and 100 mg/mL streptomycin in 5% CO2 and at 37°C in a humidified incubator. The experiment protocol was conducted according to the Declaration of Helsinki with the approval of the Shidong Hospital Research Ethics Committee (No. 2020-0002).
Mouse model of IR injury and experimental design
A 70% liver IR model was established as previouslydescribed.24,25 A total of 108 mice were randomly divided into six groups of 18 each:14,23 (1) Sham, an equivalent volume of normal saline as received by the IR group was delivered for 2 weeks by gavage before on-off of abdomen; (2) Fucoidan (Fu), fucoidan (40 mg/kg for 2 weeks) by gavage before on-off of abdomen; (3) IR, an equal volume of normal saline for 2 weeks by gavage before IR operation; (4) IR+Fu, fucoidan (40 mg/kg for 2 weeks) by gavage before IR operation; (5) IR+Fu+Erastin (Er), fucoidan (40 mg/kg for 2 weeks) by gavage and erastin (100 mg/kg, four times at 12 h intervals) by intraperitoneal injection before IR operation; and (6) IR+Fer-1, an equal volume of normal saline for 2 weeks) by gavage and Fer-1 (1 mg/kg, once at 48 h) by tail vein injection before IR operation (Fig. 1B). Six mice in each group were euthanatized at 4, 8, and 24 h after IR and serum and liver tissue were collected for further experiments.
Cell model of hypoxia-reoxygenation (HR) injury and experimental design
We simulated IR injury in vitro by treating normal hepatocytes with hypoxia and reoxygenation.26 We placed proliferating LO2 cells in a three-gas tank (oxygen concentration <1%) in glucose-free RPMI-1640 medium for 3 h, and then transferred them back to the normal incubator for 6 h before conducting subsequent experiments. The cell groups were: (1) Normal culture (NC) conditions; (2) Fu, fucoidan treatment (50 µg/mL for 2 h); (3) C-IR, HR treatment; (4) C-IR+Fu, fucoidan pretreatment (50 µg/mL for 2 h) before HR treatment; (5) C-IR+Fu+Er, fucoidan (50 µg/mL for 2 h) and erastin (10 µM for 1 h) pretreatment before HR treatment; (6) C-IR+Fer-1, Fer1 pretreatment (2 µM for 12 h) before HR treatment.
Chemicals and reagents
Fucoidan (Cat. No. F8190) was purchased from Sigma-Aldrich dissolved in normal saline or phosphate-buffered saline (PBS) and was prepared as a mother liquor at a concentration of 20 mg/mL and stored at 4°C refrigerator. Fer-1 (Cat. No. HY-100579), erastin (Cat. No. HY-15763) and MG-132 (Cat. No. HY-13259) were obtained from MedChemExpress (Monmouth Junction, NJ, USA). Alanine transaminase (ALT; Cat. No. C009-2-1), aspartate aminotransferase (AST; Cat. No. C010-2-1), Lactate dehydrogenase (LDH; Cat. No. A020-2-2), (GSH; glutathione, Cat. No. A006-2-1), malondialdehyde (MDA; Cat. No. A003-4-1) and tissue iron (Cat. No. A039-2-1) microplate assay kits were from Jancheng Biotech (Nanjing, China). Staining solutions for hematoxylin and eosin (HE, Cat. No. C0105), oil red O (Cat. No. C0157), reactive oxygen species (ROS; Cat. No. S0033), Mito-Tracker Green (Cat. No. C1048), and nuclear and cytoplasmic protein extraction kits (Cat. No. P0027) were purchased from Beyotime (Shanghai, China). FerroOrange (Cat. No. F374) and cell counting kit (CCK)-8 were obtained from Dojindo Laboratories (Tokyo, Japan). Antibodies, including ubiquitin (Cat. No. 10201-2-AP), Nrf2 (Cat. No. 16396-1-AP), HO-1 (Cat. No. 10701-1-AP), GPX4 (Cat. No. 67763-1-Ig), GAPDH (Cat. No. 60004-1-Ig) and Lamin B (Cat. No. 12987-1-AP) were all from Proteintech (Chicago, IL, USA). The secondary antibodies for western blots, IRDye 800CW goat anti-mouse and goat anti-rabbit IgG (H+L) (Cat. No. 926-32210, Cat. No. 926-32211) were purchased from LI-COR Biosciences (Lincoln, NE, USA). Kits for the reverse transcription of RNA, PCR (Cat. No. 12574018), SYBR Green PCR (Cat. No. 4309155), siRNA sequence synthesis and co-immunoprecipitation assays (Cat. No. 26149) were acquired from Thermo Fisher Scientific (Waltham, MA, USA).
Serum biochemical assays and histopathological evaluation
Serum was centrifuged and tested according to the manufacturer’s protocol and the final concentration was determined using a standard curve. Hematoxylin-eosin (HE) staining was performed on 5 µm paraffin thick sections. Representative pictures were obtained with a light microscope equipped with a digital camera (Leica Microsystems, Wetzlar, Germany). Six regions of three random sections were selected for evaluating the extent of necrosis, which was quantified using Image-Pro Plus 7.0 (Media Cybernetics, Rockville, MD, USA).
Measurement of MDA, GSH, and iron accumulation
Homogenized tissue and fragmented cell samples were prepared, and reaction reagents were added following the kit instructions. Using a standard curve as the control, the content of substances in cells or tissues was calculated for statistical analysis after correction for protein content.
ROS and lipid deposition determination
After warming frozen tissue sections to room temperature, 10 µM dihydroethidium (DHE) dye solution (λ excitation 510 nm, λ emission 590 nm) was added to cover the tissue and incubated for 60 m in a dark incubator at 37°C. After slightly drying the slices, 4′,6-diamidino-2-phenylindole (DAPI, λ excitation 350 nm, λ emission 420 nm) was added and incubated at room temperature in the dark for 10 m. Antifluorescence quenching sealing agent was used to seal and observe the sections. In vitro, treated adherent cells were incubated with 10 µM dihydroethidium diluted with serum-free culture medium for 30 m, and the cells were washed with serum-free culture medium three times to fully remove the staining solution that did not enter the cells. Fluorescence intensity was observed with a fluorescence microscope (Leica Microsystems).
Oil red O dye solution was diluted with triple-distilled water at 3:2, passed through filter paper and placed at room temperature for 10 m to produce a wine-red solution without sedimentation. Frozen tissue sections were placed in the oil red O dye solution in the dark for 15 m, washed with triple-distilled water, and then put into the hematoxylin dye solution for 5 m. The sections were quickly differentiated in 1% hydrochloric acid and alcohol for 5 s, slowly washed with running water for 10 m, and observed directly. The red intensity of Reactive oxygen species (ROS) and the fat drop area ratio after oil red O dyeing were quantitatively analyzed by Image-Pro Plus 7.0.
Quantitative real-time polymerase chain reaction (qRT-PCR)
Total RNA from tissues and cells was extracted with TRIzol reagent (Invitrogen Corporation, Carlsbad, CA, USA) and then reverse transcribed into cDNA after quantitative and quality testing. The PCR reaction system was prepared according to the kit instructions by successively adding loading buffer, primers, dNTP mixture, Tag polymerase, and water to a total volume of 20 µL. After a short centrifugation, gene expression was detected with a 7900HT fast real-time PCR system (Applied Biosystems, Foster City, CA, USA). Relative mRNA expression was calculated using 2−△△Ct values and normalized to expression of glyceraldehyde-3-phosphate dehydrogenase (GAPDH). The primer sequences used in the experiment are shown in Table 1.
Table 1Nucleotide sequences of primers used for qRT-PCR
Species | Gene | | Primer sequence (5′-3′) |
---|
Mouse | NRF2 | Forward | ACCAAGGGGCACCATATAAAAG |
| | Reverse | CTTCGCCGAGTTGCACTCA |
| HO-1 | Forward | CCTCGGACAAACTGGCACT |
| | Reverse | CCGACGCCTGACAACATCT |
| GPX4 | Forward | TGTGCATCCCGCGATGATT |
| | Reverse | CCCTGTACTTATCCAGGCAGA |
| GCH1 | Forward | GCCGCTTACTCGTCCATTCT |
| | Reverse | GAACAAGGTGATGCTCACACA |
| FTH1 | Forward | CAAGTGCGCCAGAACTACCA |
| | Reverse | ACAGATAGACGTAGGAGGCATAC |
| TFAP2A | Forward | ACGACCCCTACAGCCTGAAT |
| | Reverse | GGCGTGAGGTAAGGAGTGG |
| SREBF1 | Forward | TGACCCGGCTATTCCGTGA |
| | Reverse | CTGGGCTGAGCAATACAGTTC |
| SQSTM1 | Forward | GAGGCACCCCGAAACATGG |
| | Reverse | ACTTATAGCGAGTTCCCACCA |
| FANCD2 | Forward | CAAAATCAGCTAGGTGTGGATCA |
| | Reverse | CCAGGCCATTAACAAACTCTTCT |
| TFRC | Forward | GTTTCTGCCAGCCCCTTATTAT |
| | Reverse | GCAAGGAAAGGATATGCAGCA |
| PRKAA2 | Forward | AAGATCGGACACTACGTCCTG |
| | Reverse | TGCCACTTTATGGCCTGTCAA |
| ATF4 | Forward | CCTGAACAGCGAAGTGTTGG |
| | Reverse | TGGAGAACCCATGAGGTTTCAA |
| GAPDH | Forward | AATGGATTTGGACGCATTGGT |
| | Reverse | TTTGCACTGGTACGTGTTGAT |
Human | NRF2 | Forward | GGCGTTAGAAAGCATCCTTCC |
| | Reverse | GCAGAGGGCACACTCAAAGT |
| HO-1 | Forward | CTGTGCCACCTGGAACTGAC |
| | Reverse | TCTTGTGGGTCTTGAGCTGTT |
| GPX4 | Forward | GAGGCAAGACCGAAGTAAACTAC |
| | Reverse | CCGAACTGGTTACACGGGAA |
| GAPDH | Forward | TGTGGGCATCAATGGATTTGG |
| | Reverse | ACACCATGTATTCCGGGTCAAT |
Western blot and co-immunoprecipitation assays
Total protein and nuclear protein from tissues and cells were extracted and quantified following the kit instructions. LO2 cells were incubated with 100 nm MG-132 for 1 h and then treated with fucoidan. The specific protein Nrf2 was isolated with ubiquitin by immunoprecipitation, and the isolated protein was resolved by electrophoresis. The prepared samples were electrophoresed through the separation gel, and when the bromophenol blue dye front had moved to the bottom of the gel, the gel was transferred to a polyvinylidene fluoride membrane. After blocking nonspecific sites with 5% nonfat milk powder for 1 h, primary antibodies were added, and the membrane was incubated with gentle shaking overnight at 4°C. The membranes were washed three times before the secondary antibody was added (dilution, 1:2,000) at room temperature for 1 h. The membranes were scanned with an Odyssey two-color infrared laser imaging system (LI-COR Biosciences) that determined the intensity of the protein bands.
Immunohistochemistry and immunofluorescence staining
Dewaxed paraffin sections were rehydrated by successive immersion in decreasing ethanol concentrations. The paraffin sections were then placed in a microwave oven for 10 m until the solution boiled. Endogenous catalase was blocked with 3% hydrogen peroxide. Bovine serum albumen (BSA) antigen blocking solution (5%) was prepared in 0.1% PBT with Tween (PBST) and incubated over the sections at room temperature for 60 min. Primary antibody diluted in 0.1% PBS was added and incubated overnight at 4°C, followed by addition of working solution for the second antibody from the corresponding species labeled with horseradish peroxidase and incubation at room temperature for 60 m. Diaminobenzidine condensed chromo reagent was added dropwise to the dried paraffin tissue, the color development was observed under the microscope, and the staining was terminated by rinsing with tap water.
Treated-cell slides were fixed for 10 m, and after washing in PBS, 0.1% Triton X-100 in PBS was added for 10 m and the slides were then washed again with PBS. After blocking in 1% BSA, primary antibodies were added for overnight at 4°C. The following day, after washing with PBST three times, fluorescent second antibody was added in the dark for 1 h. Finally, an autofluorescence quenching sealing agent containing 4′,6-diamidino-2-phenylindole (commonly known as DAPI) was used to seal the slides, which were then observed by a fluorescence microscope equipped with a digital camera.
Transmission electron microscopy
Fresh samples were placed immediately in fixing solution overnight at 4°C. The samples were then rinsed and incubated with 1% osmium tetroxide for 1 h. After dehydration through a series of diluted alcohols and treated with acetone, the resin was embedded overnight and replaced. Ultra-thin sections were prepared and photographed with an electron microscope (JEM-1230; JEOL Ltd, Tokyo, Japan) to show mitochondrial morphology.
Cell viability assay and siRNA transfection
Cells were placed in a 96-well plate at a density of 2×104 cells/mL and treated as described above for the cell model. After treatment, CCK-8 reagent was added and cell viability was obtained with a microplate reader (Synergy H4; BioTek, Winooski, VT, USA) with an excitation wavelength of 450 nm. For siRNA transfection, six-well plates were used to plate cells and 50 µM siRNA-NC or siRNA-Nrf2 combined Lipofectamine 3000 was added to the cell culture for 8 h before HR according to the manufacturer’s protocol. PCR and western blot analysis were used to determine the knockout effect. The sequences of effective siRNA-Nrf2 were sense 5′-GCCUGUAAGUCCUGGUCAUTT-3′, and antisense 5′-AUGACCAGGACUUACAGGCTT-3′.
Iron level and mitochondria detection
Living cells treated with drugs were removed from the cell culture medium, and Mito-Tracker Green working solution (λ excitation 505 nm, λ emission 535 nm) at a concentration of 50 nM was incubated with the cells at for 30 m at 37°C. After removing the liquid, serum-free culture medium containing 1 mM FerroOrange (λ excitation 543 nm, λ emission 580 nm) was added and the cells were observed directly after reacting at room temperature for 1 h. The red intensity of Fe2+ was quantitatively analyzed by Image-Pro Plus 7.0.
RNA sequencing
Fresh liver tissue from the IR and fucoidan groups (n=3) was prepared as described for standardized sample processing and sent to the Shanghai Yanjiang Biotechnology Co., Ltd. (Shanghai, China) for high-throughput sequencing using an Illumina Hiseq sequencing platform. The pretreated data were subjected to quality control by FastQC software and then analyzed for differentially expressed genes (DEGs). If the fold–change was ≥2 and the corresponding corrected p-value was <0.05, the gene was considered to be differentially expressed. The Sangerbox tool (http://vip.sangerbox.com/home.html ) was used to complete gene ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) analysis to determine functional annotation and pathway enrichment. A heat map was used to display gene expression, and Venn maps were drawn to obtain the intersection of up-regulated genes in liver tissues and inhibitory genes from FerrDb (http://www.zhounan.org/ferrdb ).27
Statistical analysis
Results were reported as means±SDs and analyzed with SPSS 22.0 (IBM Corp., Armonk, NY, USA). The overall significance of data was determined through one-way analysis of variance with the Bonferroni correction for comparisons between groups. A p-value of <0.05 was considered statistically significant. Positively stained areas in this study were evaluated with Image-Pro Plus 7.0.
Results
Fucoidan protects against IR-induced liver injury by inhibiting ferroptosis
Fucoidan is an anionic sulfated polysaccharide with a molecular formula of C7H14O7S extracted from marine brown algae. The chemical structure shown in Figure 1A. Beginning 2 weeks before the IR operation, we administered fucoidan to mice at 40 mg/kg by gavage and injected erastin and Fer-1 as ferroptosis regulators to observe the extent of liver injury after IR (Fig. 1B). The results showed that the group receiving fucoidan alone had no obvious toxic or side effects in the liver at each time point from 4 to 24 h. The IR group had significant increases of ALT and AST, indicating that a model of IR injury had been successfully established. In addition, compared with the IR group, the liver enzyme levels of the fucoidan pretreatment group decreased significantly, similar to effect of Fer-1, while the ferroptosis inducer erastin reversed the suppressive effects of fucoidan on liver enzymes, most prominently at 8 h (Fig. 1C). Based on the pathological evaluation of HE staining in liver tissue, and consistent with the level of liver enzymes, the IR group exhibited cytoplasmic concentration and cell structure disorders 4 h after IR surgery. Fusion lamellar necrosis, nuclear pyknosis, and fragmentation began to appear at 8 h. In the fucoidan and Fer-1 treatment groups, necrosis was significantly reduced, but necrosis worsened after treatment with erastin. The change in necrotic area was statistically significant in erastin-treated animals (Fig. 1D, E). The serum LDH level also indicated ongoing cell necrosis. It was present at each time point, but the LDH level was significantly increased in the IR group, while treatment with fucoidan, and Fer-1 decreased the LDH level. Similarly, erastin significantly reversed the decrease in LDH produced by fucoidan or Fer-1 (Fig. 1F). The above results suggest that IR-induced liver injury was related to ferroptosis and ferroptosis inhibitors effectively protected liver function.
Fucoidan inhibits oxidative stress and lipid peroxidation
To further verify the inhibitory effect of fucoidan on ferroptosis, we evaluated the level of cellular oxidative stress products when the effect was most obvious at 8 h. Initially, the application of fucoidan alone in tissues did not cause significant changes in the accumulation of MDA, GSH, or iron, but IR significantly increased the accumulation of MDA and iron in cells and decreased GSH. Conversely, fucoidan treatment significantly inhibited the production of MDA and iron and promoted GSH to play a positive role (Fig. 2A). We used DHE and oil red O to stain frozen liver tissue to more clearly show ROS and lipid peroxidation levels in liver cells. The results showed that the fluorescent DHE staining in the IR group was significantly enhanced, and the area of positive oil red O staining increased, but staining decreased after fucoidan treatment (Fig. 2B, C). The results show that fucoidan effectively eliminated IR-induced iron accumulation, oxidative stress products and inhibited lipid peroxidation in vivo, and may be an effective inhibitor of ferroptosis.
Fucoidan regulates ferroptosis mediated by the Nrf2/HO-1/GPX4 axis
Bioinformatics analysis found 1887 DEGs after IR and fucoidan treatment, of which 1208 were up-regulated and 679 were down-regulated (Fig. 3A, B). GO functional annotation showed the biological process of DEG enrichment in programmed cell death, and the analysis of KEGG pathway enrichment found DEGs focused on the ferroptosis pathway (Fig. 3D). A total of 12 common genes were screened by FerrDb between the 1208 up-regulated DEGs and the 237 suppressor genes (Fig. 3C). To further confirm the relationship between fucoidan and ferroptosis, we screened the key pathway molecules of ferroptosis in the IR and fucoidan pretreatment groups, and found that Nrf2, HO-1, and GPX4, which are most closely related to oxidative stress, HAD the most significant changes (Fig. 3E).
Fucoidan inhibits ferroptosis through the Nrf2/HO-1/GPX4 pathway in vivo
We also verified the expression of the above genes in all groups. The genes decreased significantly in the IR group but increased significantly in the fucoidan pretreatment group (Fig. 4A). We extracted tissue protein for verification and found that the total protein level of the three above factors was significantly inhibited by IR, but the expression of Nrf2 increased in the fucoidan pretreatment group. However, Nrf2 mainly exerts its effect in the nucleus. We therefore isolated nuclear protein from cells for verification and found that the level of Nrf2 in the nucleus was consistent with that of total protein and its expression intensity relative to the internal reference gene was quantitatively counted (Fig. 4B). We then used immunohistochemical methods to visualize protein expression. We observed in liver tissue that the expression of Nrf2 and GPX4 in the fucoidan single-drug group was consistent with the control group but the induction of IR significantly inhibited the increase in Nrf2 and GPX4. After treatment of IR-injured animals with fucoidan, the proportion of Nrf2 in the nucleus increased, and GPX4 increased uniformly. Mitochondrial morphology observed by electron microscope is also an important sign of ferroptosis. We found that mitochondria in the IR group were significantly smaller, membrane density was increased, cristae were reduced or even absent, but morphological changes in the fucoidan-treated group were significantly improved (Fig. 4C). The results suggest that the mechanism by which fucoidan inhibited ferroptosis was related to the Nrf2/HO-1/GPX4 pathway, which is closely related to oxidative stress.
Fucoidan protects against hepatocyte injury induced by HR
In vitro, we simulated the in vivo IR model (C-IR) by HR, and used erastin and Fer-1 as regulators of ferroptosis. First, we tested cell survival with cell counting kit-8 (CCK-8) assays and found that at the same time point, the cell death rate in the IR model group increased, but treatment with fucoidan and Fer-1 partially maintained the survival of hepatocytes. Erastin reversed the protective effect of fucoidan, and the effect was most significant after 24 h, but the high death rate in the 48 h model group was not suitable for the study time point (Fig. 5A). Similarly, the levels of MDA, GSH, and LDH in cells are important markers of oxidative stress. Similar to the in vivo results, the levels of MDA and LDH increased and GSH levels decreased in a HR-dependent manner, which was reversed by fucoidan and Fer-1 treatment, whereas erastin exerted an effect consistent with induced in vivo injury (Fig. 5B). The level of lipid peroxidation in cells was also shown by the presence of red lipid droplets stained by oil red O, and their presence was consistent with increased concentrations of ROS and iron (Fig. 5C). The above fluorescence intensity and areas with red fat droplets differed significantly between the groups (Fig. 5D). In addition, we verified the protein level of related molecules in the pathway by western blot assays, which showed that Nrf2, HO-1, and GPX4 levels in both the cytosol and nucleus were significantly down-regulated by HR, but increased again after treatment with fucoidan and Fer-1. Erastin reversed that effect, which was significantly correlated with the ferroptosis level. The relative strength of the protein bands was quantified, and the results indicated that fucoidan protected hepatocytes from hypoxia and reoxygenation in vitro.
Fucoidan protects hepatocytes by inhibiting Nrf2/HO-1/GPX4 axis-mediated ferroptosis
To further verify the role of Nrf2 as the key link in ferroptosis, we synthesized siRNA to knock out Nrf2, and demonstrated that siRNA-Nrf2-2 had the best effect in regulating Nrf2-2 RNA (Fig. 6A). We then modeled the cells and treated them with fucoidan, followed by down-regulation of Nrf2 to verify the causal relationship between the drugs and the pathways. First, we double-stained intracellular iron and mitochondria and found that fucoidan significantly reduced the increase in iron induced by HR, Down-regulation of Nrf2 by siRNA reversed the effect of fucoidan, which was correlated with changes in mitochondria (Fig. 6B, C). The PCR results and protein levels of Nrf2, HO-1, and GPX4 also demonstrated that fucoidan simultaneously increased the gene and protein expression of these factors, and that down-regulating Nrf2 significantly weakened the protective effect of fucoidan, shown by quantitation of protein level intensity (Fig. 6D, E). To more clearly show changes in the expression of related molecules in the cell, we used cellular immunofluorescence to verify that hypoxia and reoxygenation significantly reduced the expression of Nrf2, and that fucoidan promoted the entry of Nrf2 into the nucleus. After down-regulation of Nrf2, its expression and rate of entry into the nucleus were significantly decreased. Similarly, the fluorescence intensity of GPX4 was weakened in the model group and was inhibited by siRNA-Nrf2 after being increased by fucoidan (Fig. 6G). Finally, increased ubiquitination and Nrf2 binding in LO2 cells in the model group were reduced by fucoidan, which further confirmed the molecular mechanism of fucoidan acted via Nrf2 (Fig. 6F). The results suggest that fucoidan acted to protect against hepatocyte injury induced by HR by inhibiting ferroptosis mediated by the Nrf2/HO-1/GPX4 pathway.
Discussion
At present, treatments are mainly intended to improve the metabolism of ischemic tissue, antioxidation, scavenging of free radicals, and other measures to mobilize endogenous adaptive protective mechanisms.3 In this study, we confirmed that ferroptosis occurred during of IR, and that fucoidan can inhibit iron accumulation in hepatocytes through the Nrf2/HO-1/GPX4 pathway. Thus, fucoidan was an effective inhibitor of ferroptosis, and is a potential drug to prevent IR injury.
Fucoidan is extracted from brown algae and some marine invertebrates. It has a wide range of applications in medicine, food, and industry owing to its antioxidant, antitumor, and other biological activities.13 Recent advances in extraction and preparation have overcome difficulties in mass production and have expanding its range of uses.28 Toxicological studies of fucoidan have shown that there are no significant adverse reactions in humans and mice at higher doses.19,29 We first divided mice into groups and carried out pretreatment with fucoidan in a mouse model of IR injury similar to liver transplantation. In those experiments, we used Fer-1 as a positive control. Erastin simulated the occurrence of ferroptosis. The results showed that fucoidan and Fer-1 significantly reduced the levels of serum ALT, AST, and LDH, and reduced the size of necrotic areas as shown in pathological sections, and erastin reversed the effect of fucoidan and damaged liver cells. The above effects were most significant after 8 h of treatment. This shows that inhibition of ferroptosis had a protective role in the liver. Fucoidan had the same protective effect as the inhibitor Fer-1, but its mechanism needs to be further clarified.
Sufficient evidence has been produced showing that ROS accumulation and lipid peroxidation occur in the process of IR injury, accompanied by the production of MDA and the consumption of GSH, and these factors are closely related to ferroptosis.30,31 In this study, we further examined the groups related to drug treatment at 8 h to detect the level of liver oxidative stress. It was found that the increase of MDA and decrease of GSH induced by IR were significantly improved by fucoidan supplementation, and the visual ROS level and lipid droplets stained by oil red O were also correspondingly reduced. The large amount of iron accumulated in the process was also down-regulated by fucoidan, which was in a stable state. Therefore, we have reason to further speculate that fucoidan eliminated harmful ROS and inhibited lipid peroxidation in the process, and also inhibited ferroptosis by regulating the level of iron, thus significantly improving cell function.
Studies of the regulation of iron death are currently focused on the metabolism of System Xc-, GSH metabolism, GPX4 activity, and ROS production. There are many related molecules, including fms-related tyrosine kinase 1(FLT1), FTH1, GPX4, fibroblast Specific Protein 1(FSP1), solute Carrier Family 7 Member 11(SLC7A11), Nrf2, TFRC, acyl-CoA synthetase long-chain family member 4(ACSL4), and others.32–36 Both Xu et al.37 and Qi et al.38 confirmed that brown rice extracts improved lipid peroxidation, cytotoxicity, and delayed proliferation induced by GPX4 knockdown, and that dimethyl fumarate protected the liver from IR injury by reducing ferroptosis mediated by Nrf2 and SLC7A11. To confirm the specific action pathway of fucoidan, we performed RNA sequencing and biological information analysis of liver tissue to screen for the major molecular players and found that changes in the expression of Nrf2, HO-1, and GPX4 closely related to GSH metabolism were the most significant. Multiple validation by PCR, western blotting, and immunohistochemical staining confirmed that IR inhibited the transcription and expression of the above genes in mice, by inhibiting the entry of NrF2 into the nucleus, and that fucoidan treatment reversed the above changes. Fucoidan pretreatment also lessened the mitochondrial changes caused by IR, including the reduction of mitochondrial cristae, rupture of the outer membrane, and shrinkage of mitochondria. Based on the above results, when IR-induced cells produce a large amount of lipid ROS, the GSH antioxidant system is activated, ubiquitinated Nrf2 entry into the nucleus is inhibited, and HO-1 and GPX4 are reduced, resulting in ferroptosis. Fucoidan promoted Nrf2 entry into the nucleus and transcription of HO-1 and GPX4 by clearing a large number of ROS products, thus directly inhibiting ferroptosis.39
Because of factors interfering with the in vivo experiment, we explored the mechanism in vitro. First, we established a hepatocyte model of hypoxia and reoxygenation. Similar to the in vivo experiments, we used ferroptosis regulators and fucoidan to treat cells synchronously and found that fucoidan protected against hepatocyte death, which is consistent with the effect of Fer-1. Fucoidan also reversed the increase in MDA, LDH, and GSH induced by IR and erastin, and was aided in the removal of ROS by inhibiting iron involvement and lipid droplet formation. Overall, the results indicate that in isolated hepatocytes, the antioxidant activity of fucoidan and regulation of iron homeostasis were prominent. The preliminarily verified pathway molecules also showed that fucoidan reduced the rates of ubiquitin and Nrf2 binding and promoted the entry of Nrf2 into the nucleus, resulting in increased expression of HO-1 and GPX4, which also inhibited ferroptosis. After knockout of Nrf2, fucoidan was no longer protective, resulting in serious damage to mitochondria, iron accumulation in cells, and simultaneous damage to liver function, suggesting that the Nrf2/HO-1/GPX4 axis played a key role in HR-induced liver cell, ferroptosis consistent with the in vivo results (Fig. 7).