Dihydroartemisinin ameliorates hemarthrosis-induced cartilage degeneration by suppressing chondrocyte senescence via activation of Keap1-Nrf2 signaling pathway

Abstract

Background

Joint bleeding (hemarthrosis) is a major manifestation of joint trauma, especially repeated and spontaneous in hemophilia patients. Hemarthrosis has been identified to induce the excessive reactive oxygen species (ROS) accumulation and permanent damage in articular cartilage. Dihydroartemisinin (DHA), a well-known clinical anti-malaria drug with few sides effects therapy, has been reported to possess anti-oxidative activity. This study was aimed at exploring the effect of DHA on blood-induced cartilage erosion and its underlying mechanisms.

Methods

Two distinct hemarthrosis models were constructed respectively by fresh blood joint injection in WT and joint needle puncture in F8^−/− mice, and then treated with DHA (10 or 20 mg/kg/day) for 4 weeks. In vitro chondrocytes treated with frozen-thaw blood and DHA (1, 5 or 10 μM) for 24 h. Histopathological, immunofluorescence and western blotting were investigated to demonstrate the effects of DHA on blood-induced chondrocyte senescence, ROS accumulation and extracellular matrix (ECM) degradation. Additionally, Nrf2 inhibitor (MLB385, 30 mg/kg for once a four days) and Nrf2-siRNA were used to investigate the relationship between DHA and Nrf2/Keap1 signaling in vitro and in vivo, respectively.

Results

DHA remarkably ameliorated the cartilage degeneration in both two hemarthrosis models. Similarly, in vitro experiments confirmed that DHA promoted the synthesis of ECM in blood-stimulated chondrocytes with a dose-dependent manner. DHA also effectively suppressed blood-induced chondrocyte senescence and ROS accumulation. Mechanistically, DHA activated the Nrf2 signaling by accelerating Keap1 ubiquitination and degradation. Furthermore, Nrf2 siRNA and antagonist abolished the anti-senescence and anti-oxidative functions of DHA, resulting the severe cartilage degeneration in bleeding joint of F8^−/− mice.

Conclusion

Our findings indicate that DHA effectively reduces chondrocyte senescence and mitigates cartilage destruction following hemarthrosis via activation of Nrf2/Keap1 signaling pathway.

The Translational potential of this article

On the one hand, this study highlights the important role of chondrocyte senescence in hemarthrosis-induced cartilage degradation, implying that inhibiting chondrocyte senescence may be a viable therapeutic strategy for blood-induced arthropathy. On the other hand, our findings demonstrate the remarkable chondroprotective effect of DHA in bleeding joint by modulating the Nrf2/Keap1 anti-oxidative signaling pathway, suggesting DHA may serve as a potential candidate drug for the therapy of blood-induced arthropathy.

Graphical abstract

DHA triggered Nrf2 anti-oxidation signaling by promoting keap1 protein degradation, consequently suppressing oxidative stress and chondrocyte senescence in vivo and in vitro.

The translational potential of this article: On the one hand, this study highlights the important role of chondrocyte senescence in hemarthrosis-induced cartilage degradation, implying that inhibiting chondrocyte senescence may be a viable therapeutic strategy for blood-induced arthropathy. On the other hand, our findings demonstrate the remarkable chondroprotective effect of DHA in bleeding joint by modulating the Nrf2/Keap1 anti-oxidative signaling pathway, suggesting DHA may serve as a potential candidate drug for the therapy of blood-induced arthropathy.

1. Introduction

The knee joint is susceptible to injury, and as a result, joint bleeding (hemarthrosis) or the presence of bloody effusion often occurs, leading to various immediate and long-term adverse effects for the patients [1]. In the short-term, hemarthrosis exerts joint pain, swelling and stiffness [2]. Regarding long-term implications, hemarthrosis increases the risk of early-onset arthropathy, characterized by the destruction of articular cartilage, loss of subchondral bone and synovial inflammation [[3], [4], [5]].

Generally, knee hemarthrosis can be classified based on traumatic and non-traumatic etiologies. Nontraumatic hemarthrosis is frequently observed in individuals with hemophilia affecting approximately 1.2 million people worldwide [1,6]. Approximately 44 % of hemophilia patients with previously untreated experience their initial joint hemorrhage in the first year of life [7]. About 90 % of patients suffer at least one joint hemorrhage by the age of 4.4 years [[8], [9], [10]]. Unfortunately, most people experience repeated episodes of hemarthrosis, sometimes starting before the previous episode subsides [11]. Despite the effective clotting factor replacement therapies, 90 % severe hemophilia patients continue to develop arthropathy due to breakthrough bleeding the development of inhibitors and the high cost of prophylaxis [12,13]. For traumatic hemarthrosis, the reported incidence is 4.7 per 10,000 inhabitants, which commonly occurs as a result of anterior cruciate ligament or patellar dislocation injury, and also as a postoperative comorbidity of common orthopedic procedures [[14], [15], [16], [17]]. Moreover, joints that suffer from traumatic hemarthrosis have a 50 % increased risk of developing arthropathy within 10–15 years after injury [15]. However, there is no consensus on the optimal treatment for traumatic hemarthrosis. Considering the potential requirement for anesthesia and the risk of infection in pediatric patients, the evacuation of hemarthrosis is usually not performed unless there is severe swelling, intractable pain or suspicion of infection [18]. For blood evacuation, doctor typically rely on the body's natural mechanisms to resolve the hemarthrosis without intervention [19]. Therefore, the impact of blood on joint homeostasis warrants more attention.

The vulnerability of cartilage to blood is evident, as even a single bleeding episode or only superficial cartilage touching blood can result in long-lasting alterations to cartilage matrix turnover, ultimately leading to its destruction [2,4,5]. Although degenerative changes in articular cartilage may not be observed clinically or radiologically until decades later, the underlying pathogenesis of articular cartilage destruction is thought to be caused by blood [1,20]. In vitro and in vivo models showed that short-term exposure of cartilage tissue to low concentrations of blood could lead to an increase in matrix metalloproteinase (MMPs) activation and inhibition of proteoglycan synthesis [4,[21], [22], [23]]. However, the molecular mechanisms by which hemarthrosis drive permanent damage in chondrocyte activity and cartilage matrix turnover are yet to be fully understood.

It has been shown that monocytes/macrophages and red blood cells, as present in blood, exert a direct deleterious impact on cartilage [22,24]. The iron derived from hemoglobin can catalyze the Fenton reaction to generate toxic hydroxyl radicals, leading to the reactive oxygen species (ROS) overproduction and oxidative stress in chondrocytes [25]. Following hemarthrosis, blood mononuclear cells in the joint cavity produce interleukin 1β (IL-1β) and tumour necrosis factor a (TNFa) independently of the inflammatory response [24]. These cytokines are known to disrupt cartilage matrix turnover and promote hydrogen peroxide production in chondrocyte [[26], [27], [28]]. Therefore, preventing free radical damage to chondrocytes may be a novel approach to the treatment of blood-induced joint damage.

Artemisinin, a well-known antimalarial drug, is a sesquiterpene lactone isolated from Artemisia [29]. Dihydroartemisinin (DHA) is a semisynthetic derivative of artemisinin and exhibits reduced side effects [30]. The antioxidant and anti-inflammatory properties of DHA have been demonstrated. Yang et al. showed that DHA exerts antioxidant effects through Nrf2 signaling pathway in bleomycin-induced pulmonary fibrosis [31]. Furthermore, DHA was associated with the maintenance of subchondral bone remodeling balance and chondrocyte extracellular matrix (ECM) homeostasis [32,33]. However, the potential therapeutic effects of DHA on blood-induced cartilage degeneration have not yet been investigated.

Consequently, in this study, we establish two different hemarthrosis models. We used fresh blood joint injection in WT mice to stimulate acute traumatic hemarthrosis, reflecting short-term cartilage damage caused by a single bleeding episode. The joint needle puncture in F8^−/− mice was conducted to stimulated chronic recurrent hemarthrosis in hemophilia. By using these two hemarthrosis models, we found that DHA attenuated the degeneration of cartilage matrix and subchondral bone abnormal remodeling. Subsequently, in vitro and in vivo experiments demonstrated that DHA suppressed ROS accumulation and chondrocyte senescence following blood stimulation. Mechanism studies showed that DHA induced activation of Nrf2 signaling through promoting the ubiquitination degradation of Keap1 protein, while blockade of Nrf2 abolished the chondroprotective effect of DHA. Therefore, DHA may serve as a potential candidate for blood-induced arthropathy therapy.

2. Methods

2.1. Reagents

DHA was derived from Chengdu Must Biotechnology Co., Ltd (Chengdu, China, CAS: 71939-50-9, 98 % purity). ROS inhibitor-Acetylcysteine was derived from Selleck (Shanghai, China, Catalogue No. S1623, 99.97 % purity). Nrf2 inhibitor-ML385 was derived from Selleck (Shanghai, China, Catalogue No. S8790, 99.92 % purity). Protein synthesis inhibitor-Cycloheximide was derived from MedChemExpress (New Jersey, United States, Catalogue No. HY-12320, 99.86 % purity).

2.2. Animal

Ten-week-old male wide type (WT, C57BL/6J) mice and FVIII gene knockout hemophilia (F8^−/−) mice were obtained from Shanghai Laboratory Animal Center of the Chinese Academy of Sciences. The mice participating in this study were maintained in a pathogen-free environment at the Animal Center of Zhejiang University of Traditional Chinese Medicine (Hangzhou, China). Each cage housed five mice and provided ad libitum access to food and water. All experiments were conducted with the approval of the Ethics Committee for Animal Experiments of Zhejiang University of Traditional Chinese Medicine. Ethical Permission for Animal Experiments (Approval Number.: 20220808-05).

2.3. Mice models construction

The two animal models were selected to mimic distinct different types of hemarthrosis. WT mice with fresh blood injection: this model simulated acute traumatic hemarthrosis, reflecting short-term cartilage damage caused by a single bleeding episode [23]. F8^−/− mice with needle puncture: this model stimulated chronic recurrent hemarthrosis in hemophilia, reflecting progressive joint damage due to prolonged blood exposure [34].

The Traumatic model was constructed as described by Li L et al. [23]. Briefly, whole blood was drawn from the hearts of WT mice and stored in sodium heparin tubes. Then, the joints cavity was injected with extracted blood (10 μL) for modeling.

The joint bleeding model in F8^−/− mice was established as described by Hakobyan et al. [34]. F8^−/− mice were anesthetized with 3 % isoflurane and maintained with 1 % isoflurane. Joint bleeding was modelled by puncturing the right knee joint capsule with a 30g needle.

2.4. Experimental grouping and drug administration

Mice were randomly divided into the control group, model group, DHA low-dose (DHA-L) group, DHA high-dose (DHA-H) group, DHA low-dose + ML385 (DHA + ML385) group, and ML385 group. The DHA dose for antimalarial treatment is 60 mg/day [35]. The equivalent dose ratio for humans to mice is 9.01; therefore, the calculated dose for mice is 60 mg/60 kg × 9.01 = 9.01 mg/kg. Therefore, two doses close to the estimated dose, 10 mg/kg and 20 mg/kg, were used. Starting from the day of modeling, mice in the DHA-L group (10 mg/kg/day), DHA-H group (20 mg/kg/day), DHA-L (10 mg/kg/day) + ML385 (30 mg/kg/day) group and ML385 group (30 mg/kg/day) were orally gavaged with related drugs. Specifically, ML385 was gavaged once a four days. DHA was gavaged continuously for 4 weeks. The mice in the control group and model group were administered an aliquot of 0.9 % saline by gavage daily.

2.5. Micro-computed tomography (μCT) analysis

Knee joint tissues, collected from 4 weeks mice after modeling, were stripped the surrounding muscle tissues, and then fixed with 4 % polyformaldehyde for 3 days. The fixed tissue samples were placed in the μCT mice bed in the sagittal direction for X-ray scanning. The μCT scanning (Skyscan1272, Bruker, USA) with the following parameters: the scanning software was SkyScan1176. The quantitative analysis of the regions of interest (ROI) was performed within the bone micromorphological analysis software CTAn. Finally, 3D image building was performed using the onboard CTvox graphics software.

2.6. Histomorphological staining

The samples were washed with running water to remove residual fixative from the surface, decalcified in 14 % EDTA solution for 2 week and embedded in paraffin. The paraffin tissues were sliced into 3 μm sections, and then stained with Safranin O/Fast green staining (SO/FG). Degradation of cartilage structure was scored by three uninformed observers following the recommendations of the Osteoarthritis Research Society International (OARSI) [36]. Quantitative analysis was performed by ImageJ software. Prussian blue staining was performed to evaluate the presence of hemosiderin deposits in the knee joints by using Prussian blue staining kit (Sbjbio, BP-DL121, Nanjing, China). The area of iron deposition in the cartilage and synovial tissue regions of each specimen by ImageJ. Hematoxylin-eosin (H&E) staining was used to evaluate the effects of DHA on the kidney and liver of mice.

2.7. Immunohistochemical Staining(IHC)

Sections were routinely dewaxed, rehydrated, and then soaked in 0.01 M citrate buffer (Solarbio, Beijing, China) at 60 °C for 4 h to thermally repair. Subsequently, the sections were soaked in endogenous peroxidase blocker for 20 min, and then incubated with primary antibody at 4 °C overnight. After incubated with homologous IgG polymer for 20 min, sections were stained with diaminobenzidine (DAB) solution and nuclear with hematoxylin. Immunohistochemistry was conducted with antibodies against Col2a1 (Abcam, Ab34712, Cambridge, UK; 1:200 dilution); Mmp13 (Abcam, Ab39012, Cambridge, UK, 1:200 dilution); Nox1 (Proteintech, 17772-1-AP, Wuhan, China, 1:200 dilution); Nox2 (Proteintech, 19013-1-AP, Wuhan, China, 1:200 dilution); Gpx4 (Abclonal, A1933, Wuhan, China, 1:200 dilution); P16 (Abclonal, A0262, Wuhan, China, 1:200 dilution); P21 (Abcam, Ab188224, Cambridge, UK, 1:200 dilution); Nrf2 (Abclonal, A25327, Wuhan, China, 1:200 dilution); Keap1 (Huabio, HA721525, Hangzhou, China, 1:200 dilution); HO-1 (Huabio, ER1802-73, Hanghzou, China, 1:200 dilution). The cells or areas of positive staining were determined upon secondary antibody treatment and color development using DAB chromogen (brown stain). We obtained the positive staining cells or areas ratio by calculating the ratio of the positive staining cells or areas to the whole tissue cells or areas by using ImageJ software. For immunofluorescence (IF) staining, the sections or cells were incubated with primary antibody at 4 °C overnight, and then incubated with Alexa Fluor™ 488 (Invitrogen, California, United States, 1:1000 dilution) for 1 h and nuclear stained with DAPI (Solarbio, Beijing, China, 1:1000 dilution) for 10 min. The positive cells were detected by fluorescence microscopy. In each biological repeat, three fields (50–100 cells/field) were randomly selected for the analysis of the mean fluorescence intensity by using ImageJ.

2.8. Cell culture and treatment

The mouse primary chondrocytes were obtained from two-week-old WT mice. Mice were sacrificed and disinfected in 75 % ethyl alcohol, followed by aseptic isolation of femoral capsule cartilages. Then, isolated cartilages were digested with 0.25 % collagenase P (Roche, Basel, Switzerland) for 4 h in cell culture incubator. The collected primary chondrocytes were cultured with F12/DMEM medium (Thermo Fisher Scientific, Massachusetts, United States) containing 10 % FBS (Thermo Fisher Scientific, Massachusetts, United States) and 1 % streptomycin/penicillin (Thermo Fisher Scientific, Massachusetts, United States) at 37 °C and 5 % CO₂.

The blood collection was performed according to the description of Coline Haxaire et al. [37]. Blood isolated from WT mice hearts by cardiac puncture was immediately adjusted to 7.5 mM EDTA and separated plasma and cells by centrifugation (10000g, 10 min, Eppendorf). Then, the cells were frozen at −80 °C, thawed, and added to chondrocytes (blood cells 2 × 10⁷ cells/mL) for 24h.

2.9. Cell viability assay

Primary chondrocyte viability was measured according to the manufacturer's instructions to use the Cell Counting Kit (CCK)-8 assay (Apexbio, Texas, USA). Briefly, cells were seeded into 96-well plates at a ratio of 5 × 10³ per well. After treatment with DHA (0, 1, 5, 10, 20, 50, 100 μM) for 24 or 48 h, the plates were incubated with 10 μL of CCK-8 solution at 37 °C for 2 h. Absorbance was calculated at 450 nm using a BioTek Synergy HT spectrophotometer (BioTek, Vermont, USA).

2.10. Real-time quantitative PCR (qPCR) assay

Primary chondrocytes were treated with blood for 24 h, then treated with DHA (0, 1, 5, 10 μM) for 24 h and then collected to isolate total RNA by using Trizol solution. mRNAs were reverse transcribed into cDNA using All-in-One cDNA Synthesis SuperMix (Takara, Beijing, China). The qPCR assay was performed with 2x SYBR Green qPCR Master Mix (Low ROX) (Takara, Beijing, China) on QuantStudio™ 7 Flex Real-Time PCR System (Thermo Fisher Scientific, Massachusetts, United States). Actb was used as the reference gene for quantitative analysis. Table 1 lists the primer sequences for the target genes.

Table 1.

Primer sequences for qPCR.

Primer name	Primer sequence(5’→3′)Age
Adamts5 Forward	CTGCCTTCAAGGCAAATGTGTGG
Adamts5 Reverse	CAATGGCGGTAGGCAAACTGCA
Mmp3 Forward	CTCTGGAACCTGAGACATCACC
Mmp3 Reverse	AGGAGTCCTGAGAGATTTGCGC
Mmp13 Forward	CTTCTTCTTGTTGAGCTGGACTC
Mmp13 Reverse	CTGTGGAGGTCACTGTAGACT
Col2a1 Forward	GGGAATGTCCTCTGCGATGAC
Col2a1 Reverse	GAAGGGGATCTCGGGGTTG
Aggrecan Forward	CCTGCTACTTCATCGACCCC
Aggrecan Reverse	AGATGCTGTTGACTCGAACCT
Sox9 Forward	CACACGTCAAGCGACCCATGAA
Sox9 Reverse	TCTTCTCGCTCTCGTTCAGCAG
P16 Forward	TGTTGAGGCTAGAGAGGATCTTG
P16 Reverse	CGAATCTGCACCGTAGTTGAGC
P21 Forward	TCGCTGTCTTGCACTCTGGTGT
P21 Reverse	CCAATCTGCGCTTGGAGTGATAG
P53 Forward	GCGTAAACGCTTCGAGATGTT
P53 Reverse	TTTTTATGGCGGGAAGTAGACTG
Actb Forward	GGCTGTATTCCCCTCCATCG
Actb Reverse	CCAGTTGGTAACAATGCCATGT

2.11. Western blot

Total proteins were extracted by RIPA buffer (Beyotime Biotechnology, P0013B, Shanghai, China) containing protease inhibitor (Phenylmethylsulfonyl fluoride) and phosphatase inhibitor (sodiμm pyrophosphate). The target proteins were separated by 10 % SDS-PAGE gel (20 μg/lane), transferred to PVDF membrane, which blocked with 5 % skimmed milk for 1 h. Subsequently, membranes were incubated in the primary antibodies at 4 °C overnight. The primary antibodies were as follows: Col2a1 (Abclonal, A1560, Wuhan, China, 1:1500 dilution); Mmp13 (Abclonal, A11755, Wuhan, China, 1:1500 dilution); Mmp3 (Abclonal, A23097, Wuhan, China, 1:1000 dilution); Adamts5 (Bioss, 3573R, Beijing, China, 1:1000 dilution); Aggrecan (Bioss, 11655R, Beijing, China, 1:1000 dilution); Sox9 (Abclonal, A19710, Wuhan, China, 1:1000 dilution); Nox1 (Proteintech, 17772-1-AP, Wuhan, China, 1:1000 dilution); Nox2 (Proteintech, 19013-1-AP, Wuhan, China, 1:1000 dilution); Gpx4 (Abclonal, A1933, Wuhan, China, 1:1000 dilution); P16 (Abclonal, A0262, Wuhan, China, 1:1000 dilution); P21 (Abcam, Ab188224, Cambridge, UK, 1:1000 dilution); P53 (Abcam, Ab131442, Cambridge, UK, 1:1000 dilution); Nrf2 (Immunoway, YT3189, Texas, USA, 1:1000 dilution); Keap1 (Huabio, HA721525, Hangzhou, China, 1:1000 dilution); HO-1 (Huabio, ER1802-73, Hanghzou, China, 1:1000 dilution). Membranes were then incubated with a homologous secondary antibody (Proteintech, SA00001-1, Wuhan, China, 1:5000 dilution) for 1 h. ECL substrate kit (4A Biotech, 4AW011-1000, Suzhou, China) was used to visualize protein bands and analyzed with a Bio-Rad scanner (Bio- Rad, California, United States). Nuclear and cytoplasmic proteins were extracted from chondrocytes using the Nucleus and Cytoplasmic Protein Extraction Kit (Beyotime Biotech, S0027, Shanghai, China) for the detection of Nrf2 protein. All subsequent procedures were as above described. ImageJ softwarewas used to calculate the grey scale value of the protein bands and analyze the expression of target proteins in each group. β-actin (proteintech, 20536-1-AP, Wuhan, China, 1:5000 dilution) and Lamin B1 (Huabio, ET1606-27, Hanghzou, China, 1:1000 dilution) served as loading controls.

2.12. Reactive oxygen (ROS) assay

Tissue sections were incubated with the ROS probe (1:500 dilution)for 1.5 h and then stained with DAPI (1:1000 dilutio) for 10 min, according to the manufacturer's instructions (BestBio, BB-470516, Shanghai, China). For primary chondrocytes, the ROS probe was incubated with primary chondrocytes for 20 min according to the manufacturer's instructions (Beyotime Biotechnology, S0033S, Shanghai, China). Finally, positive expression was detected by fluorescence microscopy. The quantitative analyses for positive staining area were performed by the software of ImageJ.

2.13. Superoxide dismutase (SOD) assay

The activity of SOD in primary chondrocytes in each group was measured by a total superoxide dismutase assay kit containing NBT (Beyotime Biotechnology, S0109, Shanghai, China), according to manufacturer's instructions.

2.14. Glutathione/oxidized glutathione disulfide (GSH/GSSG) assay

For primary chondrocyte GSH/GSSG measurements, GSH and GSSG were assayed using the GSH/GSSG Assay Kit (Beyotime Biotechnology, S0053, Shanghai, China), according to the manufacturer's protocol.

2.15. RNA sequencing (RNA-seq) for the chondrocyte

NovelBrain Cloud Analysis Platform was used to perform RNA-seq analysis of primary mouse chondrocytes. Total RNA was extracted from primary mouse chondrocytes with TRIzol reagent after 24 h of blood treatment. vAHTSTM Total RNA-seq (H/M/R) was used to construct cDNA libraries for each pooled RNA sample. TopHat and Cufflinks were used for transcriptional expression analysis. FPKM methods were used to determine gene expression. DESeq algorithm was used to identify differentially expressed genes. P-value <0.05 were used for significance analysis.

2.16. Aged-related β-galactosidase staining

Primary chondrocytes were fixed and stained using the Senescent β-Galactosidase Staining Kit (Beyotime Biotechnology, C0602, Shanghai, China) according to the manufacturer's protocol. Cells were placed in SA-β-gal staining solution and incubated overnight at 37 °C in a CO₂-free dry incubator, then observed under a microscope to assess the degree of blue color. The quantitative analyses for positive staining area were performed by the software of ImageJ.

2.17. Molecular docking

ChemBio3D (14.0.0.117) is used to convert 2D chemical structures into 3D structures and save them in MOL2 format from PubChem Compound. The target crystal structures were from the RCSB protein database. Two target proteins, Keap1(ID:1u6d) and Nrf2(ID:2lz1) were studied. AutoDockTools 1.5.6 was used to convert receptors and ligands from native format to pdbqt format. The structures were optimised by removing water molecules and adding hydrogen atoms. Autodock Vina was then used to perform molecular docking studies. All docking run options were set to default values based on the Genetic Algorithm. PyMoL was used to visualize the highest scoring docking results.

2.18. Protein stability assay

Primary Chondrocytes were treated with cycloheximide (CHX) for 0, 4, 8 and 12 h, and the cell protein samples were collected. The Keap1 protein expression level was detected by Western blot, and then the protein stability was analyzed.

2.19. Co-immunoprecipitation and ubiquitination assays

For the co-immunoprecipitation assay, cells were lysed with IP lysate containing PMSF (Beyotime Biotechnology, P0013, Shanghai, China). Total lysate (200 μg) was incubated with primary antibody (3 μg) or IgG (1 μg) overnight at 4 °C with gentle shaking. Then incubate with protein A/G beads (Santa Cruz Biotechnology, sc-2003, Texas, United States) for 4 h at 4 °C. The immunoprecipitated complexes were then boiled with 2 × Loading buffer for 10 min and analyzed by protein blotting. The Abs employed for Western blot containing Keap1 (Huabio, HA721525, Hangzhou, China) and ubiquitination (Ubiquitin, Abclonal, A0162, Wuhan, China, 1:1000 dilution).

2.20. Transfection of siRNA-Nrf2

Primary Chondrocytes were transfected with siRNA targeting Nrf2 (GenePharma, Shanghai, China) or negative control siRNA (scrambled; GenePharma, Shanghai, China) using the X-tremeGENE siRNA Transfection Reagent (Roche, XTG9-RO, Basel, Switzerland) according to the manufacturer's protocol. Transfection effects were determined by qPCR and Western blotting assays. Two validated siRNAs were used for subsequent experiments. After 24 h of transfection, cells were treated with blood for 24 h, followed by 10 μM DHA.

2.21. Statistical analysis

All experimental data of this study were analyzed by GraphPad Prism software version 8.0, and presented as mean ± standard deviation. One-way ANOVA with Tukey's multiple comparisons test or two-way ANOVA followed by Šidák post hoc test was used for comparison of multiple groups. Data based on grading systems were analyzed using nonparametric Mann–Whitney U test or Kruskal–Wallis test, respectivly for comparisons between 2 and >2 groups. The difference was statistically significant with P < 0.05.

3. Results

3.1. DHA alleviates hemarthrosis induced-cartilage impairment in both hemophilia and WT mice

To assess the effect of DHA on hemarthrosis-induced arthropathy, we constructed two distinct hemarthrosis mice models: (1) needle puncture knee injury in F8^−/− mice, and (2) fresh blood injection into joint cavity in WT mice, and then treated they with low (10 mg/kg/day) and high (20 mg/kg/day) doses of DHA for 4 weeks (Supplementary Figs. 1a–b). The SO/FG staining revealed the obvious degeneration of the articular cartilage in both two hemarthrosis models, accompanied by elevated OARSI score, in comparison to the control group (Fig. 1a and d). Subsequently, DHA treatment ameliorated the cartilage degradation and reduced OARSI score in both two models, but the high-dose DHA group exhibited higher OARSI score in comparison to the low-dose group (Fig. 1a and d).

Fig. 1.

DHA delays cartilage degradation in joint hemorrhage mice. (a) Representative images of SO/FG staining and the OARSI score for knee sections of F8^−/− mice treated with or without DHA at 4 weeks post-injury. Red arrow indicates the wear area. Representative images and corresponding quantitative analysis of IHC staining of (b) Col2a1 and (c) Mmp13 in F8^−/− mice treated with or without DHA at 4 weeks post-injury. Red arrow indicates the positive cells; scale bar: 100 μm. (d) Representative images of SO/FG staining and the OARSI score for knee sections of WT mice treated with or without DHA at 4 weeks following blood injection. Red arrow indicates the wear area; scale bar: 100 μm. Representative images and corresponding quantitative analysis of IHC staining of (e) Col2a1 and (f) Mmp13 in WT mice treated with or without DHA at 4 weeks following injury. Red arrow indicates the positive cells; scale bar: 100 μm; each point represents one biological repeats (n = 6) The data were presented as means ± SD and subjected to analysis by one-way ANOVA with Tukey's multiple comparisons test, ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001.

By using IHC staining, we observed a dramatic decrease in Col2a1 protein expression and increase in expression of Mmp13 protein in both two models (Fig. 1b–c and e-f). These negative alterations in cartilage matrix were reversed by DHA treatment, while the low-dose DHA showed better effects than high-dose DHA (Fig. 1b–c and e-f). Next, μCT analysis showed that compared to the control group, the hemarthrosis in F8^−/− mice induced subchondral bone loss (Supplementary Figs. 2a–d). However, in WT mice, hemarthrosis resulted in subchondral bone sclerosis (Supplementary Figs. 2e–h). These abnormal subchondral bone remodeling were relieved by DHA treatment, which was more pronounced with low-dose DHA (Supplementary Figs. 2a–h). In addition, DHA treatment failed to eliminate hemosiderin deposits in the affected synovium of F8^−/− mice (Supplementary Figs. 3a–b). We also examined synovial inflammation and found that compared with control group, both two model groups showed increased expression of macrophages marker F4/80 and fibroblasts marker FAPα in synovium, but no significant alleviation in synovial inflammation was detected following DHA treatment (Supplementary Figs. 3c–h). Taken together, these findings indicate the protective effect of DHA on cartilage integrity and subchondral bone remodeling following hemarthrosis.

3.2. DHA suppresses ECM degradation in blood-stimulated chondrocyte

We used frozen-thaw blood to stimulate primary mouse chondrocytes in vitro. As revealed by qPCR and Western blot, in blood-stimulated chondrocytes, the mRNA and protein levels of anabolism indicators (Col2a1, Aggrecan, Sox9) were significantly decreased, and expression levels of catabolism indicators (Mmp3, Mmp13, Adamts5) were dramatic up-regulated (Fig. 2a-m). Further we examined the impact of DHA on blood-stimulated chondrocytes in vitro. The result of the CCK-8 assay confirmed that DHA exhibited no cytotoxicity towards chondrocyte viability even at concentrations as high as 100 μM after treatment of 24 and 48 h (Supplementary Figs. 4a–b). The qPCR experiment was then conducted on blood-induced chondrocytes, and showed that the expression of Col2a1 was upregulated after treatment of DHA at concentrations of 1, 5, 10 μM, but no change was observed at a concentration of 20 μM (Supplementary Fig. 4c). We therefore chose 1, 5 and 10 μM DHA for subsequent experiments. The treatment of DHA significantly enhanced the mRNA and protein levels of anabolic genes (Col2a1, Aggrecan, Sox9) and reduced the expression levels of catabolic genes (Mmp3, Mmp13, Adamts5) in blood-stimulated chondrocytes in a DHA does-dependent manner (Fig. 2a-m). Meanwhile, IF analysis also demonstrated a progressive upregulation of Col2a1 protein level in response to increasing doses of DHA, while the protein level of Mmp13 gradually decreased (Fig. 2n-p). Together, the above data suggest that DHA exerts chondroprotective effects by promoting anabolism and inhibiting catabolism of ECM.

Fig. 2.

DHA inhibits blood-induced ECM degradation in chondrocytes. Quantification of mRNA levels for (a)Adamts5,(b)Mmp3, (c)Mmp13, (d)Col2a1, (e)Aggrecan, and (f)Sox9 in mouse chondrocytes treated with DHA (1, 5 and 10 μM); each independent biological repeat (n = 3–4) with 3 technical repeats. (g) Western blot and corresponding quantification analysis of proteins for (h) Adamts5, (i) Mmp3, (j) Mmp13, (k) Col2a1, (l) Aggrecan, and (m) Sox9 in mouse chondrocytes treated with DHA (1, 5 and 10 μM); each independent biological repeat (n = 4) with 3 technical repeats. (n) Representative immunofluorescence images of Col2a1 and Mmp13 expressions in mouse chondrocytes treated with DHA (1,5 and 10 μM). Corresponding quantification of (o) Col2a1 and (p) Mmp13 fluorescence intensity; scale bar: 100 μm, each independent biological repeat (n = 4) with 3 technical repeats. The above data were presented as means ± SD and subjected to analysis by one-way ANOVA with Tukey's multiple comparisons test, ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001.

3.3. DHA inhibits blood-induced oxidative stress in chondrocytes in vitro and in vivo

It is well known that chondrocytes exposed to blood are susceptible to the production of ROS, which lead to tissue or organ damage [38]. In both primary chondrocytes induced by blood and cartilages from hemarthrosis of F8^−/− mice, we observed the excessive accumulation of ROS, but DHA treatment effectively mitigated this production (Fig. 3a-b and m-n). In addition, DHA treatment significantly elevated critical antioxidants, as evidenced by enhanced SOD activity and up-regulated GSH level in blood-induced chondrocytes following DHA treatment (Fig. 3c and d). In vivo and in vitro chondrocytes, blood stimulation suppressed the protein expression of Gpx4 and induced expression of oxidative stress markers, namely Nox1 and Nox2 protein (Fig. 3e–l). As expected, DHA treatment alleviated these abnormal expressions, and these beneficial effects were in a dose-dependent manner in vitro experiment, but better in DHA-L group than in DHA-H group in vivo experiment (Fig. 3e-l). Similar effect of DHA on Nox1 expression was also observed in hemarthrosis model in WT mice (Supplementary Fig. 5a). In summary, these results suggest that DHA effectively attenuates blood-induced oxidative stress in chondrocytes.

Fig. 3.

DHA reduced blood-induced oxidative stress in chondrocytes in vitro and in vivo. (a) Representative images and (b) corresponding quantitative analysis of ROS staining in mouse chondrocytes treated with DHA (1, 5 and 10 μM); scale bar: 100 μm; n = 3. Quantitative analysis of (c) SOD and (d) GSH/GSSH ratio in mouse chondrocytes treated with DHA (1, 5 and 10 μM); each independent biological repeat (n = 3) with 3 technical repeats. (e) Western blot and corresponding quantification analysis of proteins for (f) Nox1, (g) Nox2, and (h) Gpx4 in mouse chondrocytes treated with DHA (1,5 and 10 μM); each independent biological repeat (n = 4) with 3 technical repeats. (i) Representative images and (j) corresponding quantitative analysis of IHC staining of Nox1 and Nox2 in F8^−/− mice treated with or without DHA at 4 weeks post-injury. Red arrow indicates the positive cells; scale bar: 100 μm. (k) Representative images and (l) corresponding quantitative analysis of IHC staining of Gpx4 in F8^−/− mice treated with or without DHA at 4 weeks post-injury. Red arrow indicates the positive cells; scale bar: 100 μm. (m) Representative images and (n) corresponding quantitative analysis of ROS staining in F8^−/− mice at 4 weeks post-injury. Scale bar: 100 μm. For the data i-n, each point represents one biological repeats (n = 6). The data were presented as means ± SD and subjected to analysis by one-way ANOVA with Tukey's multiple comparisons test, ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001.

3.4. DHA reduces chondrocyte senescence following blood stimulation in vitro and in vivo

Since oxidative stress is one of the major factors of cellular senescence [39], we explored the influence of blood in chondrocyte senescence. The RNA sequencing analysis revealed that blood significantly up-regulated senescence-promoting genes in chondrocytes, while down-regulated the expressions of senescence-suppressing genes (Fig. 4a). Both qPCR and Western blot experiments further confirmed that the expression levels of P16, P21, and P53, senescence-associated markers, were increased in chondrocytes following blood stimulation (Fig. 4d–j). Nevertheless, the blood-induced increase in P21 protein, was largely abrogated by ROS scavenger (S1623) (Fig. 4b and c), suggesting that blood-induced senescence in chondrocyte is attributed to ROS production. Furthermore, DHA treatment noticeably suppressed the expressions of senescence indicators (P16, P21, P53) in blood-induced chondrocytes in a dose-dependent manner (Fig. 4d–j). Besides, SA-β-gal was also dose-dependently attenuated by DHA treatment (Fig. 4k-l). Furthermore, in vivo experiments in F8^−/− and WT mice confirmed that DHA treatment alleviated hemarthrosis-induced chondrocyte senescence, as demonstrated by reduced expressions of P21 and P16 proteins (Fig. 4m-n, and Supplementary Fig. 5b). However, these in vivo results showed that the anti-senescence effect of low doses DHA was better than high dose (Fig. 4m-n, and Supplementary Fig. 5b). Altogether, above data suggest that DHA protects chondrocyte against senescence induced by joint bleeding.

Fig. 4.

DHA reduced blood-induced senescence in chondrocytes in vitro and in vivo. (a) Heatmap of mRNA-seq analysis showing the expressions of senescence-related genes in chondrocytes induced by blood. (b) Western blot and corresponding quantification analysis of protein for (c) P21 in mouse chondrocytes treated with S1623 (10 μM); each independent biological repeat (n = 4) with 3 technical repeats. (d) Western blot and corresponding quantification analysis of proteins for (e) P16, (f) P21, and (g) P53 in mouse chondrocytes treated with DHA (1, 5 and 10 μM); each independent biological repeat (n = 3–4) with 3 technical repeats. Quantification of mRNA levels for (h)P16,(i)P21, and (j)P53 in mouse chondrocytes treated with DHA (1,5 and 10 μM); each independent biological repeat (n = 3) with 3 technical repeats. (k) Representative images and (l) corresponding quantitative analysis of SA-β-gal staining in mouse chondrocytes treated with DHA (1,5 and 10 μM); scale bar: 100 μm; each independent biological repeat (n = 3–4) with 3 technical repeats. (m) Representative images and (n) corresponding quantitative analysis of IHC staining of P16 and P21 in F8^−/− mice treated with or without DHA at 4 weeks post-injury; red arrow indicates the positive cells; scale bar: 100 μm. For the data m-n, each point represents one biological repeats (n = 6). The data were presented as means ± SD and subjected to analysis by one-way ANOVA with Tukey's multiple comparisons test, ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001.

3.5. Nrf2 signaling mediates the anti-senescence effect of DHA in blood-stimulated chondrocytes

Due to the critical role of Nrf2 in the cellular antioxidant defense system, we further investigated the effect of DHA on Nrf2 signaling [40]. As expected, blood stimulation reduced the expressions of Nrf2 and HO-1 proteins and increased Keap1 protein expression in chondrocytes, whereas DHA treatment dose-dependently restored these aberrant expressions (Fig. 5a–d). Moreover, as shown in IF and western blotting analysis of cytosolic and nuclear fractions, the treatment of DHA promoted the translocation of Nrf2 into nucleus, which further confirmed the activation of Nrf2 signaling pathway (Fig. 5e–h). Keap1 is a key inhibitor of the Nrf2-mediated antioxidant response pathway and our molecular docking data showed higher binding affinity between DHA and keap1 protein (Fig. 5i and j). Thus, we examined the effect of DHA on Keap1 protein degradation by using CHX treatment assay, and found that DHA markedly accelerated the degradation of Keap1 (Fig. 5k-l). In addition, DHA significantly increased the level of ubiquitination of Keap1 in blood-induced chondrocytes (Fig. 5m) These findings suggest that Nrf2/Keap1 signaling pathway mediates the chondroprotective effects of DHA in blood-stimulated chondrocytes, which, at least in a significant part, is attributed to DHA-induced degradation of Keap1 protein.

Fig. 5.

DHA promotes Keap1 ubiquitination and degradation to suppress Nrf2 signaling in blood-induced chondrocytes in vitro. (a) Western blot and corresponding quantification analysis of proteins for (b) Nrf2, (c) Keap1 and (d) HO-1 in mouse chondrocytes treated with DHA (1, 5 and 10 μM); each independent biological repeat (n = 4) with 3 technical repeats. (e) Representative immunofluorescence images and (f) corresponding quantitative analysis of nuclear Nrf2 protein in mouse chondrocytes treated with DHA (1, 5 and 10 μM); scale bar: 100 μm; each independent biological repeat (n = 4) with 3 technical repeats. (g) Western blot analysis for Nrf2 protein in cytosolic and nuclear fractions, and (h) corresponding quantification analysis of in mouse chondrocytes treated with DHA (10 μM); each independent biological repeat (n = 4) with 3 technical repeats. (i–j) Molecular docking revealed the binding free energy of DHA to Keap1 protein. (k) Western blot and corresponding quantification analysis of proteins for (l) Keap1 in mouse chondrocytes treated with DHA (10 μM) and CHX (50 μM); each independent biological repeat (n = 4) with 3 technical repeats. (m) The Ubiquitination level of Keap1 was determined in the indicated group in chondrocytes; each independent biological repeat (n = 4) with 3 technical repeats. The data were presented as means ± SD and subjected to analysis by one-way ANOVA with Tukey's multiple comparisons test; ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001.

To better understand the role of Nrf2 in the antioxidant and anti-senescence effects of DHA in blood-stimulated chondrocytes, we performed Nrf2 knockdown assay by siRNA. The efficiency of Nrf2 knockdown in chondrocytes was confirmed at mRNA and protein levels (Supplementary Figs. 6a–c). Nrf2 knockdown largely diminished the antioxidant effects of 10 μM DHA, as evidenced by markedly reduction of SOD activity, GSH level and Gpx4 protein expression, as well as increase in ROS production, Nox1 and Nox2 protein expressions (Fig. 6a–i). Meanwhile, following Nrf2 knockdown, DHA treatment failed to suppress blood-stimulated expressions of P16, P21 proteins and SA-β-gal in chondrocytes, suggesting it exhibits anti-senescence effect by activating Nrf2 signaling (Fig. 6e and j-m).

Fig. 6.

Nrf2 mediated the effect of DHA on ROS and senescence in blood-induced chondrocytes. (a) Representative images and (b) corresponding quantitative analysis of ROS staining in blood-induced chondrocytes treated with or without DHA, siRNA-NC, siRNA-Nrf2-345, or siRNA-Nrf2-1565; scale bar: 100 μm; each independent biological repeat (n = 4) with 3 technical repeats. Quantitative analysis of (c) SOD and (d) GSH/GSSH ratio in blood-induced chondrocytes treated with or without DHA, siRNA-NC, siRNA-Nrf2-345 or siRNA-Nrf2-1565; each independent biological repeat (n = 4) with 3 technical repeats. (e) Western blot and corresponding quantification analysis of proteins for (f) Nrf2, (g) Gpx4, (h) Nox1, (i) Nox2, (j) P16, and (k) P21 in blood-induced chondrocytes treated with or without DHA, siRNA-NC, siRNA-Nrf2-345, or siRNA-Nrf2-1565; each independent biological repeat (n = 4) with 3 technical repeats. (l) Representative images and (m) corresponding quantitative analysis of SA-β-gal staining in blood-induced chondrocytes treated with or without DHA, siRNA-NC, siRNA-Nrf2-345, or siRNA-Nrf2-1565; scale bar: 100 μm; each independent biological repeat (n = 4) with 3 technical repeats. The data were presented as means ± SD and subjected to analysis by one-way ANOVA with Tukey's multiple comparisons test, ∗∗∗P < 0.001.

3.6. DHA restores the cartilage homeostasis in bleeding joint through activating Nrf2 signaling

To further clarify the underlying molecular mechanisms of DHA in the treatment of blood-induced cartilage degeneration, we first examined Nrf2/Keap1 signaling pathway in hemarthrosis mice models with or without DHA treatment. The protein expressions of Nrf2 and HO-1 were significantly reduced, but Keap1 protein was obviously increased in cartilages from hemarthrosis of F8^−/− mice compared with control (Fig. 7a–c). Interestingly, these inactivated changes in Nrf2/Keap1 signaling were effectively inhibited following DHA administration (Fig. 7a–c). Consistent with these results, in cartilages from hemarthrosis of WT mice, DHA also activating Nrf2/Keap1 signaling pathway by up-regulating the protein expressions of Nrf2 and HO-1 and down-regulating the protein expression of Keap1 (Supplementary Figs. 7a–c).

Fig. 7.

DHA activates Nrf2/Keap1 signaling in cartilage from F8^−/− mice following joint bleeding. Representative images and corresponding quantitative analysis of IHC staining of (a) Nrf2, (b) Keap1 and (c) HO-1 in F8^−/− mice treated with or without DHA at 4 weeks post-injury. Red arrow indicates the positive cells; scale bar: 100 μm. For the data a-c, each point represents one biological repeats (n = 6). (e) Representative images of SO/FG staining for knee sections of F8^−/− mice treated with or without DHA or ML385 at 4 weeks post-injury and the OARSI score. Red arrow indicates the wear area; scale bar: 100 μm. (f) Representative images and corresponding quantitative analysis of ROS staining in F8^−/− mice treated with or without DHA or ML385 at 4 weeks post-injury. Scale bar: 100 μm. Representative images and corresponding quantitative analysis of IHC staining of (g) P16 and (h) HO-1 in F8^−/− mice treated with or without DHA or ML385 at 4 weeks post-injury. Red arrow indicates the positive cells; scale bar: 100 μm. (i) Representative 3D reconstruction and corresponding quantification analysis of BV/TV (%) of the subchondral bone at F8^−/− mice treated with or without DHA or ML385 at 4 weeks post-injury. Scale bar: 100 μm. For the data e-i, each point represents one biological repeats (n = 3). The data were presented as means ± SD and subjected to analysis by one-way ANOVA with Tukey's multiple comparisons test; ns: no significance; ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001.

Subsequently, we used ML385 to block the activity of Nrf2 activity in hemarthrosis models of F8^−/− mice (Fig. 7d). Following ML385 intervention, DHA failed to mitigate the cartilage erosion, as shown in SO/FG staining (Fig. 7e). Additionally, compared with the DHA-treated group, ML385-DHA combination group exhibited a significant increase in ROS production and P16 protein expression, as well as the reduction of HO-1 protein expression (Fig. 7f–h). Besides, the alleviating effect of DHA on bone loss was also abolished by ML385 intervention (Fig. 7i). Therefore, DHA exerts the beneficial effects on cartilage matrix homeostasis partly through Nrf2/Keap1 signaling.

4. Discussion

The hemarthrosis is the most observed complication of patients with hemophilia, and it also occurs in young and active individuals following acute joint injury. Recurrent hemarthrosis, even a single bleeding episode, has the potential to cause long-term destruction of cartilage, and eventually result in crippling arthropathy. However, the prevention and treatment of blood-induced cartilage damage are limited, due to its unclear pathomechanisms. The excessive accumulation of ROS is a hallmark in chondrocyte following joint bleeding, resulting in persistent disturbance of the cartilage matrix turnover. Considering the great ability of DHA to inhibit oxidative stress, in this study, we administrated hemarthrosis models with DHA and observed its effective suppression of blood-induced accumulation of ROS and senescence in chondrocytes, as well as its therapeutic effect on cartilage degeneration. Our further mechanistic analyses demonstrated that DHA triggered Nrf2 anti-oxidation signaling by promoting keap1 protein ubiquitination and degradation, consequently suppressing oxidative stress and chondrocyte senescence. Collectively, these findings suggest the potential of DHA as a promising drug for maintaining cartilage integrity and homeostasis in patients with hemarthrosis.

DHA, as the first generation derivative of Artemisinin, not only enhances the antimalarial activity, but also serves as a pivotal starting point for synthesizing derivatives of Artemisinin, such as Artesunate, Artemether and Arteether [41,42]. The results of drug metabolism study indicated that DHA is the primary active metabolite of Artesunate, Artemether and Arteether, suggesting that these drugs primarily exert their effects in vivo through the production of DHA [43]. After the remarkable success of DHA in the treatment of malaria, Tu Youyou and other research groups began to explore new indications for DHA, leading to significant advancements in the investigation of systemic lupus erythematosus and cancer [44,45]. These findings suggest that DHA may possess additional potential applications beyond its anti-malarial effects. Here, we showed that both high and low dose DHA effectively inhibited the accumulation of ROS and cartilage degeneration following joint bleeding in both WT and F8^−/− mice after 4 weeks administration. Nevertheless, it should be noted that our in vivo data also indicated that high dose DHA (20 mg/kg/day) possessed the fewer therapeutic effects than low dose (10 mg/kg/day). In addition, our in vitro study confirmed that DHA (1, 5 or 10 μM) suppressed the expression of ECM catabolic genes and markers of oxidative stress, while promoted expression of ECM anabolic genes and antioxidant genes in blood-stimulated chondrocytes with a dose-depend manner. Our in vitro data showed that 20 μM DHA exhibited no effect on chondrocyte homeostasis following blood treatment. Jiang et al. also reported that low concentration of DHA (1 μM) without cytotoxicity, inhibited TNF-α-induced expression of Mmp-3 and -9, Adamts5 in chondrocytes by promoting autophagy via NF-κB pathway inhibition [46], and accumulating data suggest the role of autophagy in the ROS elimination in chondrocytes [47]. While they did not perform in vivo validation, their findings are highly consistent with ours, which together suggest that DHA has marked chondroprotective effects, most likely due its antioxidant properties.

Given the widespread clinical use of DHA as an antimalarial drug and fewer side effects [30], DHA is a feasible and rapid strategy for the clinical treatment of hemorrhagic arthropathy. For the prophylactic treatment of blood-induced arthropathy in hemophilia, DHA therapy is more convenient, cheaper and easier to promote clinically than clotting factors replacement.

The iron and pro-inflammatory cytokines, derived from whole blood, exert the lasting damage to chondrocytes and its extracellular matrix metabolism, despite short joint bleeding [4,22,48]. Particularly in hemophilia, recurrent joint hemorrhages often occur. Our Prussian blue stain showed that F8^−/− mice still showed hemosiderin deposits in affected joints 4 weeks after modeling. Thus, we administrated model mice with low (10 mg/kg/day) and high (20 mg/kg/day) doses of DHA for 4 weeks, and no damage was observed in kidney and liver (Supplementary Fig. 9). Moreover, in animal models of malaria and intestinal diseases, DHA treatment with the similar doses and times showed obvious therapeutic results and fewer side effects [35,49,50].

The Nrf2/Keap1 pathway serves as the primary antioxidant defense mechanism, supported by well-established role of Nrf2 in regulating the expression of genes associated with oxidative stress, such as HO-1, Gpx4, etc [51]. Moreover, Nrf2/HO-1 pathway exerts cyto-protective effect and plays a crucial role in the development of age-related disorders [52,53]. Previous in vitro study has reported that DHA induces the expression of Nrf2 in myeloid-derived suppressor cells. Here, we further demonstrated the activated function of DHA on Nrf2 signaling in blood-stimulated chondrocytes both in vitro and in vivo. Furthermore, Nrf2 inhibition significantly abrogated therapeutic effects of DHA on blood-induced cartilage degeneration. The exact molecular mechanisms underlying the activation of Nrf2 by DHA, however, remain undetermined. In the normal condition, Nrf2 is widespread present in the cytoplasm as a dimer with Keap1 (Nrf2-Keap1), whereas Keap1 inhibits Nrf2 activation. In our molecular docking test, we found that compared with Nrf2, keap1 showed higher affinity to DHA (Fig. 5i-j and Supplementary Figs. 8 a-b). Furthermore, the protein degradation experiment indicated that DHA promoted the ubiquitination and degradation of Keap1 protein. Thus, it is seeming that the satisfactory antioxidant activity of DHA in this research is attributed, at least in part, to DHA-induced degradation of keap1 protein.

Articular chondrocytes normally present a low level of proliferation and remain quiescent under normal conditions [54]. Following joint injury, chondrocytes increase their proliferation to repair cartilage and maintain tissue homeostasis. However, chondrocytes that re-enter the division cycle in this manner are more susceptible to cellular senescence [55,56]. To our knowledge, there is very limited research regarding the association of senescence to blood-induced cartilage degeneration, although the present of senescent cells are revealed in the cartilage of people with osteoarthritis [57]. In our hemarthrosis models, significant chondrocyte senescence was demonstrated through the increased expressions of P21 and P16 protein. Since ROS has been recognized as the most prominent risk factor of cellular senescence [39], we here scavenged ROS and detected a reduced senescence in blood-stimulated chondrocytes following blood stimulation. Furthermore, our in vitro and in vivo experiments both showed that DHA treatment effectively attenuates blood-induced chondrocyte senescence. Taken together, our findings not only imply a critical role of chondrocyte senescence in blood-induced arthropathy, but also suggest DHA as potential drug for other arthropathies accompanied with chondrocyte senescence.

Recently, it has been suggested that transplantation of senescent cells into the knee joint area of wild-type mice can induce an osteoarthritis-like condition [58]. Furthermore, in both 2D and 3D cell cultures, Jeon OH et al. revealed that the elimination of senescent cells from pathological chondrocyte cultures by UBX0101 increased the proliferation rate of the remaining chondrocytes [59]. Considering these findings, we hypothesized that the enduring damage to cartilage caused by blood might be attributed to chondrocyte senescence, even when a single episode of joint bleeding or only superficial chondrocytes directly contact with blood.

Additionally, in our F8^−/− mouse models, DHA attenuated the subchondral bone loss adjacent to the affected joint, which is another pathological feature in blood-induce arthropathy [37]. This result aligns with previous reports that DHA attenuated abnormal bone loss [32,60]. Dou et al. demonstrated that DHA mitigated lipopolysaccharide-induced osteoclastogenesis and bone loss via the mitochondria-dependent apoptotic pathways [60]. Ding et al. reported that DHA attenuates osteoclastogenesis via the inhibition of NF-κB/MAPK/NFATc1 in osteoarthritis [32]. Another study suggest that the impact of DHA on bone remodeling is contributed by the combination of its functions in angiogenesis and osteoclastgenesis [61]. Alternatively, we also thought that DHA protect subchondral bone from loss likely by maintaining cartilage integrity, as cartilage erosion leads to instability of mechanical force on subchondral bone and thus causes it abnormal remodeling [62]. Therefore, the cellular signaling pathways through which DHA exerts its protective effects on subchondral bone warrant further research. In summary, DHA have considerable therapeutic potential in blood-induced arthropathy.

With regard to the role of DHA in anti-inflammatory and immunomodulatory processes [52,60,63], we examined the effect of DHA treatment on synovitis following hemarthrosis, whereas no significant change was observed. Meanwhile, DHA treatment also failed to alleviate hemosiderin deposits in the affected synovium of F8^−/− mice, which might contribute to the less anti-inflammatory efficacy of DHA in synovial tissue.

5. Conclusion

In summary, our findings greatly advance our understanding of the crucial role of chondrocyte senescence in articular cartilage homeostasis following hemarthrosis, and suggests that inhibiting chondrocyte senescence may be a viable therapeutic strategy for blood-induced arthropathy. Notably, our study demonstrated the remarkable chondroprotective effect of DHA in bleeding joint by modulating the Nrf2/Keap1 anti-oxidative signaling pathway. Given DHA wide clinical use as an antimalarial drug, repurposing DHA as a joint protective agent might be a quick strategy for the development of drug targeting blood-induced arthropathy.

Lead contact

Further information and requests for resources and reagents should be directed to and will be fulfilled by the lead contact, Jiali Chen (chenjiali@zcmu.edu.cn).

Materials availability

This study did not generate new unique reagents.

Data and code availability

The data providing support for the results of this study are available from the lead contact, Jiali Chen (chenjiali@zcmu.edu.cn).

Author contributions

Qinghe Zeng: Conceptualization, In vivo experiments, Writing – original draft. Yongjia Feng: Conceptualization, In vitro experiments. Haipeng Huang: In vitro experiments. Kaiao Zou: In vitro experiments. Wenzhe Chen: In vitro experiments. Xuefeng Li: Data analysis. Yuliang Huang: Funding acquisition. Weidong Wang: Funding acquisition. Wenhua Yuan: Data analysis. Pinger Wang: Data analysis. Peijian Tong: Writing – review & editing. Hongting Jin: Supervision, Funding acquisition, Writing – review & editing. Jiali Chen: Supervision, Funding acquisition, Writing – review & editing.

Funding statement

This research was partially supported by the National Natural Sciences Foundation of China (Grant No. 82274280, 82305005, 82104891, 82074457), Zhejiang Provincial Natural Science Foundation (Grant No. LR23H270001). Medicine and Health Technology Program of Zhejiang Province (Grant No. 2023KY862). Zhejiang Provincial Vanguard and Leading Goose Science and Technology Program (Grant No. 2025C02177).

Declaration of competing interest

All authors declare that they have no conflict of interest.

Appendix A. Supplementary data

The following is the Supplementary data to this article.