D‐mannose alleviates osteoarthritis progression by inhibiting chondrocyte ferroptosis in a HIF‐2α‐dependent manner

Abstract Objectives Chondrocyte ferroptosis contributes to osteoarthritis (OA) progression, and D‐mannose shows therapeutic value in many inflammatory conditions. Here, we investigated whether D‐mannose interferes in chondrocyte ferroptotic cell death during osteoarthritic cartilage degeneration. Materials and methods In vivo anterior cruciate ligament transection (ACLT)‐induced OA mouse model and an in vitro study of chondrocytes in an OA microenvironment induced by interleukin‐1β (IL‐1β) exposure were employed. Combined with Epas1 gene gain‐ and loss‐of‐function, histology, immunofluorescence, quantitative RT‐PCR, Western blot, cell viability and flow cytometry experiments were performed to evaluate the chondroprotective effects of D‐mannose in OA progression and the role of hypoxia‐inducible factor 2 alpha (HIF‐2 α) in D‐mannose‐induced ferroptosis resistance of chondrocytes. Results D‐mannose exerted a chondroprotective effect by attenuating the sensitivity of chondrocytes to ferroptosis and alleviated OA progression. HIF‐2α was identified as a central mediator in D‐mannose‐induced ferroptosis resistance of chondrocytes. Furthermore, overexpression of HIF‐2α in chondrocytes by Ad‐Epas1 intra‐articular injection abolished the chondroprotective effect of D‐mannose during OA progression and eliminated the role of D‐mannose as a ferroptosis suppressor. Conclusions D‐mannose alleviates osteoarthritis progression by suppressing HIF‐2α‐mediated chondrocyte sensitivity to ferroptosis, indicating D‐mannose to be a potential therapeutic strategy for ferroptosis‐related diseases.


| INTRODUC TI ON
Ferroptosis is a form of oxidative cell death characterized by the iron-dependent accumulation of lipid hydroperoxides to lethal levels. [1][2][3] Ferroptosis has emerged as a potent mechanism for preventing multiple neoplastic and degenerative diseases such as Alzheimer's disease, Parkinson's disease and kidney degeneration. 1,4 As one of the most common degenerative joint disorders, osteoarthritis (OA) is the leading cause of chronic disability in the geriatric population, imposing a huge burden on society. 5 Current medical care for OA focuses mainly on alleviating painful symptoms, but typically fails to prevent disease progression. Pathologically, degradation of articular cartilage is one of the most prominent features of progressive OA. 6 Articular cartilage is composed of chondrocytes, which maintain the integrity of the extracellular matrix by balancing its synthesis and degradation. 7 Evidence shows that anabolic/catabolic balance and survival of chondrocytes are crucial for articular cartilage homeostasis and osteoarthritic destruction, 8 underlining the importance of chondrocyte fate control. Recent evidence has indicated that chondrocyte ferroptosis contributes to the progression of OA. 9 Iron-overloaded mice exhibit increased cartilage destruction, and intracellular iron uptake is favoured in chondrocytes mimicking an OA phenotype. 10,11 These findings suggest that ferroptosis suppression is a novel candidate component for preventing OA progression. D-mannose, a C-2 epimer of glucose, is naturally present in many fruits and plants, and has been reported to be beneficial in many human disease states. 12 For example, it has been shown to be effective in tumour inhibition, immunopathology repression and infection treatment through anti-inflammatory and immunomodulatory activity. [13][14][15] Lin et al. 16 reported that D-mannose could suppress monosodium iodoacetate-induced OA development in rats, and it was reported that D-mannose inhibits LPS-induced production of IL-1β, which also mediates OA progression. 17 Evidence has verified the roles of D-mannose in glucose metabolism, anti-inflammation and T-cell immune response. 13,[17][18][19] To bring the role of chondrocyte ferroptosis in OA into focus, the relationship between D-mannose and chondrocyte ferroptosis needs elucidation.
Hypoxia-inducible factor 2α (HIF-2α) has been reported to play a vital role in cartilage development, OA progression and sensitizing cells to ferroptosis. [20][21][22] In embryonic development, HIF-2α insufficiency impairs not only chondrocyte hypertrophy but also the subsequent steps of endochondral ossification. 23 During OA, HIF-2α performs as a key catabolic transcription factor, inducing the expression of matrix-degrading enzymes and progressive cartilage damage under the regulation of NF-κB signalling. [23][24][25][26] Recently, work focussing on tumours has demonstrated an essential role for HIF-2α in regulating cellular iron homeostasis and ferroptosis susceptibility. 21,22 For example, HIF-2α was shown to stimulate the specific enrichment of polyunsaturated fatty acids in clear cell renal cell carcinoma, 27 whereas in colorectal cancers, HIF-2α activation potentiates oxidative cell death by increasing cellular iron. 21 Until now, it has remained unknown whether HIF-2α contributes to chondrocyte ferroptosis and its regulation in OA and how it is regulated.
Here, we employed an anterior cruciate ligament transection (ACLT)-induced OA mouse model and an in vitro OA microenvironment induced by IL-1β exposure. Our results demonstrated that D-mannose possesses chondroprotective effects against OA progression through inhibiting HIF-2α-potentiated chondrocyte ferroptosis. These findings not only provide valid evidence for Dmannose to be used in clinical interventions in OA, but also suggest the potential therapeutic prospects of the natural plant-derived component for many other ferroptosis-related diseases.

| Mice and surgery
C57BL/6 J mice (8 weeks old, female) were purchased from Dossy Experimental Animal Limited Company (Chengdu, China). All the mice were bred and maintained under specific-pathogen-free conditions on a 12-/12-h light/dark cycle. For D-mannose (Sigma-Aldrich, M2069) treatment, normal drinking water was exchanged for 20% mannose in drinking water (w/v) according to previous studies. 13,14 For surgery, mice were anaesthetized with pentobarbital sodium (100 mg/kg, injected intraperitoneally) and subjected to unilateral ACLT procedures. 28 The sham group received a skin incision and su-

| Histology and immunofluorescence
After fixation, mouse legs were decalcified in 0.5 M EDTA for 2 weeks, embedded in paraffin and sectioned at 4 μm. Slides were stained with safranin O/fast green (Solarbio) using standard protocol. The severity of OA-like phenotype was analysed using the OARSI scoring system by two blinded observers.

| Isolation and culture of mouse chondrocytes
Primary mouse chondrocytes were isolated from knee joint cartilage of 5-day-old C57BL/6 J mice as described previously. 29 Briefly, after dissected into pieces, cartilage tissue was digested by 2.5mg/ ml collagenase type II (Gibco) for 2 h and 0.5 mg/ml collagenase type II overnight at 37°C. The primary chondrocytes were resuspended and cultured in low glucose DMEM medium (Gibco) containing 10% foetal bovine serum (Gibco) and 1% penicillin-streptomycin at 37°C with an atmosphere of 21% O 2 (for normoxia) or 1% O 2 (for hypoxia), 5% CO 2 and 95% humidity.
Primary chondrocytes were identified with toluidine blue

| Small interfering RNA assays
siRNAs specific to Epas1 was designed with the coding sequences of mouse Epas1 shown in Table S1. Chondrocytes cultured for 3 days were transfected for 24 h with siRNA (100 nM) using Lipofectamine 3000 (Invitrogen). Non-silencing siRNA was used as a negative control.

| Epas1 adenovirus and infection
Adenovirus expressing mouse Epas1 and mock virus were produced from Genechem. For infection of primary chondrocytes, chondrocytes were cultured for 3 days, infected with 800 MOI of mock virus or Ad-Epas1 virus for 12 h, and incubated for additional 36 h.

| CCK-8 cell viability assay
Primary chondrocytes were transferred to 96-well plates at a concentration of 5000 cells/well in 100 μl of culture medium supplemented with 10 μl of CCK-8 reagent (MCE) and incubated at 37°C for 2 h following indicated treatments. Cell viability was evaluated using the absorbance values determined at 450 nm using microplate reader (Synergy H1; BioTek).

| Lipid peroxidation, ROS assay
To detect lipid peroxidation, cells were incubated with 5 µM of

| Western blotting
The cell lysates were extracted using RIPA lysis buffer (Beyotime,

| Statistical analysis
Results are presented as the mean ± standard error of mean of independent replicates (n ≥ 3). Statistically significant differences were evaluated using two-tailed Student's t tests for comparison between two groups or by one-way analysis of variance followed by the Tukey's test for multiple comparisons. NEJM formatting for p values was used (NS when no significant difference, * when p < 0.05, ** when p < 0.01, *** when p < 0.001). All statistical analyses were conducted using GraphPad Prism 8.  All quantified data are shown as mean ± SEM; NS, not significant, *p < 0.05, **p < 0.01, ***p < 0.001 by one-way ANOVA followed by the Tukey-Kramer test

| D-mannose protects osteoarthritic chondrocytes by attenuating sensitivity to ferroptosis
Recent evidence has indicated that chondrocyte ferroptosis contributes to chondrocyte homeostasis and osteoarthritic cartilage degeneration. 9 Similar to previous study, ferrostatin-1 (Fer-1), the inhibitor of ferroptosis, could attenuate the cytotoxicity of IL-1β to chondrocytes, indicating the existence of ferroptosis in IL-1β-treated chondrocytes ( Figure S2A). GPX4 is a canonical glutathione-based ferroptosis inhibitor and functions mainly through inhibiting the formation ACLT and D-mannose-treated ACLT (Sham + Man) groups using immunohistochemistry staining assay. As expected, compared with the sham group, the percentage of GPX4 + chondrocytes in the ACLT group was reduced significantly (from 46.7 ± 9.1% to 12.5 ± 6.2%). To our surprise, we observed that D-mannose treatment largely rescued ACLT-induced GPX4 + chondrocyte reduction ( Figure 3A,B). These in vivo results suggested that OA-induced chondrocyte ferroptosis could be suppressed significantly by D-mannose application.
To confirm the effect of D-mannose on chondrocyte ferroptosis, we next investigated whether D-mannose interfered in chondrocyte ferroptotic cell death during cartilage degeneration in vitro.
As evidenced by cell viability assay, D-mannose rendered the chondrocytes more resistant to IL-1β-potentiated cell death ( Figure 3C). C11-BODIPY staining showed that lipid peroxidation was significantly enhanced in response to IL-1β stimulation and inhibited by D-mannose ( Figure 3D). Moreover, D-mannose potently decreased the level of malondialdehyde (MDA), a by-product of lipid peroxidation ( Figure 3E). Also, the RNA and protein levels of the two key ferroptosis suppressors, Gpx4 and Slc7a11, were increased in Dmannose-treated chondrocytes compared with those treated with IL-1β ( Figure 3F,G). We tested reduced glutathione (GSH), which are essential for ferroptosis prevention, 27 and found that reduced GSH level was downregulated in response to IL-1β treatment and could be reversed by D-mannose ( Figure 3E). We also investigated the intracellular ROS level, which always increases significantly in ferroptosis. Expectedly, after D-mannose treatment, we observed a much lower ROS level in control chondrocytes than in IL-1β-treated ones upon DCFH-DA staining ( Figure S3A

| D-mannose alleviates cartilage degeneration by suppressing HIF-2α
HIF-2α is a key catabolic mediator regulated by NF-κB signalling in OA progression. 23,24,35 As is revealed by histological assay, ACLT strongly induced p65 phosphorylation while in contrast, phosphorylation of p65 in cartilages from D-mannose-treated ACLT group was significantly reduced ( Figure S5A,B). Consistently, in cultured chondrocytes, p65 nuclear translocation ratio was dramatically upregulated by IL-1β treatment so did the expression level, which were suppressed by Dmannose treatment ( Figure S5C-E). We next investigated the change of HIF-2α expression in response to D-mannose in cartilage degeneration models in vivo and in vitro. In vivo, compared with the sham group, the percentage of HIF-2α-positive chondrocytes in the ACLT group increased dozens of times, but reverted to a level comparable with that of the sham group when D-mannose was applied ( Figure 4A,B).  Figure 4I). Together, these data suggested that D-mannose alleviates cartilage degeneration by suppressing HIF-2α via NF-κB signalling.

| D-mannose decreases chondrocyte ferroptosis sensitivity via inhibiting HIF-2α expression
Since D-mannose impedes HIF-2α activation and cartilage degeneration, we tested whether HIF-2α is involved in  found that D-Mannose enhanced chondrocyte autophagy via the AMPK pathway indicating a potential relationship between mechanisms of D-mannose therapeutic effect in OA. In fact, increasing body of evidence reflects that there exists an intrinsic connection between autophagy and ferroptosis. On the one hand, some studies identified autophagy is as an upstream mechanism in the induction of ferroptosis by regulating cellular iron homeostasis and cellular ROS generation. 45,46 On the other hand, ferroptosis regulators are also involved in the control of autophagy dependent on the context. 47 Besides, recent studies revealed that HIF-2α and AMPK pathway participated in these two forms of cell death respectively. For example, Lee et al. 48 found that AMPK activation could inhibit ferro- It is well acknowledged that the activity of HIF-2α, an extensive regulator of OA, and its upstream regulator, NF-κB, increases during OA development. 20,23,24 So far, several mechanisms by which HIF-2α regulates OA have been demonstrated. On one hand, HIF-2α directly induces higher expression of catabolic factors, leading to articular cartilage destruction. 24 On the other hand, HIF-2α levels in human and mouse OA chondrocytes are highly associated with kinds of regulated cell death of chondrocytes. 50 Meanwhile, HIF-2α potentiates Fas-mediated chondrocyte apoptosis and decreases expression of autophagy during OA development. 49,51 Here, besides findings consistent with previous ones establishing that HIF-2α is a catabolic regulator of osteoarthritic cartilage destruction, we first reported a novel mechanism of HIF-2α during OA progression by which it potentiates cell death via lipid oxidation, ROS accumulation and ferroptosis regulators. Our work not only provides a novel finding that HIF-2α mediates osteoarthritic cartilage degeneration via creating a ferroptosis-susceptible cell state of chondrocytes, which broadens the understanding of the potential therapeutic role for HIF-2α in OA management, but also brings insights into the mechanisms underlying ferroptosis susceptibility within an OA microenvironment.
Notably, previous studies highlighted the involvement of Dmannose in immunomodulation at the cellular and molecular levels and confirmed that D-mannose can suppress the immunopathology of autoimmune diabetes, airway inflammation and lupus. 13,52 In macrophages, D-mannose interferes with glucose metabolism by raising intracellular mannose-6-phosphate levels to inhibit macrophage activation and IL-1β production. 14,17 Zhang et al. 13,18 reported that by upregulation of reactive oxygen species generated by increased fatty acid oxidation, D-mannose promotes TGFβ activation to stimulates Treg cell differentiation. It cannot be neglected that an activation of innate and adaptive immune responses contribute to the F I G U R E 7 Model of D-mannose ameliorates osteoarthritis progression by inhibiting chondrocyte ferroptosis in a HIF-2αdependent manner. D-mannose inhibited chondrocyte ferroptosis by suppressing HIF-2α during OA progression, leading to the prevention of cartilage degeneration initiation and sustaining of OA, during which immune cells infiltrate synovium and act as a main contributor for the release of diseasespecific cytokines and chemokines. 53,54 Typically, macrophages can stimulate the release of pro-inflammatory cytokines like IL-1β and TNFα and other catabolic and anabolic mediators involved OA pathology. Besides, T cells are responsible for increased infiltration of macrophages by inducing macrophage inflammatory protein-1γ and can interact with chondrocytes through cell surface molecules to upregulate MMPs release. 55 Considering the tight connection of D-mannose, immune and OA progression, we consider that beyond the aim of the present work, future investigations are foreseen to determine whether, and the extent to which, immunomodulation induced by mannose administration in mice might contribute to OA amelioration.
Collectively, our investigations have revealed the previously unrecognized role of D-mannose in attenuating OA progression via inhibiting HIF-2α-induced ferroptosis, indicating a novel, safe and effective therapeutic strategy for OA and many other ferroptosisrelated diseases.

CO N FLI C T O F I NTE R E S T
The authors declare no competing interests.

AUTH O R CO NTR I B UTI O N S
J.L. and J.W. conceived and designed the project. X.Z. and Y.Z. performed the experiments; W.S., Z.Z., J.L., W.Y., X.Y. and W.S. contributed to data acquisition. X.Z., Y.Z., Z.Z. and W.S. analysed the data; X.Z. and Y.Z. wrote and edited the manuscript; Y.Y., J.L. and J.W.
contributed to the manuscript revision. J.L. and J.W. supervised the research. All authors read and approved the final paper.