Hypoxia, cuproptosis, and osteoarthritis: Unraveling the molecular crosstalk

Abstract

Aberrant changes in the hypoxic microenvironment are implicated in osteoarthritis (OA) development. Cuproptosis, a unique copper-dependent form of regulated cell death that depends on the activity of key enzymes in mitochondrial metabolism, is also linked to oxygen levels. However, the crosstalk among hypoxic environment, cuproptosis, and OA development remains unclear. This study confirmed that oxygen levels in the OA model gradually increased during OA progression, which suppressed the expression of anabolic genes for articular cartilage extracellular matrix, upregulated catabolic genes, and increased the cell death rate of primary chondrocytes. Mechanistically, oxygen elevation upregulated the expression of solute carrier family 31 member 1 (SCL31A1), a transmembrane pump facilitating copper uptake, whereas it downregulated the expression of ATPase copper transporting beta (ATP7B), a copper chaperone facilitating copper efflux, leading to aberrant copper accumulation in the cells. Ultimately, this accumulation efficiently induces oligomerization of dihydrolipoamide S-acetyltransferase (DLAT), triggering cell death, known as cuproptosis. Hypoxia-inducible factor-1α (HIF-1α), induced under hypoxic conditions, is negatively correlated with DLAT and the severity of OA, as confirmed in human and rat cartilage. Furthermore, siRNA-mediated HIF-1α silencing sensitized primary chondrocytes to cuproptosis, whereas HIF-1α stabilization had protective effects. In summary, increased oxygen levels in the cartilage induce cuproptosis, thereby accelerating OA progression, whereas HIF-1α stabilization mitigates this process. These findings may provide novel therapeutic targets for OA treatment in clinical practice.

Keywords: HIF-1α; Osteoarthritis; cuproptosis; hypoxic environment.

Graphical abstract

Fig. 1

The oxygen level changes during the progression of OA. (A) Schematic diagram of the experimental procedure and experimental designs. (B) The images of distance from the left and right hind limbs to the midline (n = 6, mean ± s.d.). (C) The images of paw area and retention duration of the left and right hind paws, LH: Left hind paw, RH: right hind paw (n = 6, mean ± s.d.). (D) Data analysis of hind paw area and duration in rats (n = 6, mean ± s.d.). (E) Micro-CT images of sagittal and coronal of the knee joints at 0, 4, 8 weeks after modeling surgery (n = 6, mean ± s.d.). (F) The Tb·Th, BV/TV and BMD were analyzed by CT-AN software. (n = 6, mean ± s.d.). (G) Safranin-O Fast-Green (SO/FG) staining of joint sections at 0, 4, 8 weeks after modeling surgery, scale bar = 500 μm, 100 μm (n = 6, mean ± s.d.). (H) OARSI scores of different groups (n = 6, mean ± s.d.). (I) Immunofluorescence staining was used to evaluate the oxygen concentration in articular cartilage at different stages of OA, scale bar = 200 μm, 50 μm (n = 6, mean ± s.d.). (J) Quantitative analysis of immunofluorescence staining of Pimonidazole (n = 6, mean ± s.d.). ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001, ∗ means 0w vs. 4w or 8w groups.

Fig. 2

The effect of oxygen levels on the ECM, proliferation, and apoptosis of primary chondrocytes. (A) Alcian blue staining of primary chondrocytes at oxygen concentrations of 1 %, 3 %, 5 %, and 7 % in 1, 3, 7 days (1d, 3d and 7d), scale bar = 1000 μm (n = 3, mean ± s.d.). (B) Quantitative analysis of Alcian blue staining (n = 3, mean ± s.d.). (C) qRT-PCR was used to detect the levels of extracellular matrix synthesis gene markers Col2a1, Acan, and degradation gene markers Adamts5, Mmp13 (n = 3, mean ± s.d.). (D) Western blot assay of ACAN and MMP13 at different pO₂ (n = 3, mean ± s.d.). (E) Immunofluorescence staining of the expression level of MMP13, scale bar = 200 μm (n = 3, mean ± s.d.). (F) Quantitative analysis of IF staining of MMP13 (n = 3, mean ± s.d.). (G–H) EdU assay was used to detect the effect of different pO₂ on the proliferation of primary chondrocytes, scale bar = 200 μm (n = 3, mean ± s.d.). (I–J) Apoptosis staining of primary chondrocytes at different oxygen levels, scale bar = 200 μm (n = 3, mean ± s.d.). (K) Flow cytometry of apoptosis to detect the apoptosis rate of primary chondrocytes under different pO₂ (n = 3, mean ± s.d.). ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001, ∗ means 1 % O₂ vs. 3 % O₂, 5 % O₂ or 7 % O₂ groups.

Fig. 3

Elevated O₂ levels induce chondrocyte cuproptosis during OA progression. (A) Volcanic map of differentially expressed genes between blank and model rats (n = 3). (B) GO enrichment analysis of biological processes. (C) Heat map of genes related to carbohydrate metabolism process, including Glycolysis, TCA cycle and PDHC. (D–E) Immunohistochemical staining and quantitative analysis of RIPK3, Cleaved-Caspase3, LIAS and P16 in normal and OA rats, scale bar = 500 μm (n = 6, mean ± s.d.). (F) JC-1 staining was used to detect mitochondrial membrane potentials, scale bar = 200 μm, 40 μm (n = 3, mean ± s.d.). (G) DHE staining was used to detect superoxide anion level of primary chondrocytes, scale bar = 200 μm, 40 μm (n = 3, mean ± s.d.). (H) Cell copper colorimetric assay of Cu content detection in primary chondrocytes among 4 groups (n = 3, mean ± s.d.). (I) The aggregation of DLAT at four oxygen concentrations was observed by immunofluorescence staining, scale bar = 200 μm (n = 3, mean ± s.d.). (J) Western blot assay of soluble and insoluble DLAT of primary chondrocytes at different pO₂ after 24 h cultivation (n = 3, mean ± s.d.). (K) Western blot analysis of protein levels of ATP7B and SLC31A1 after 24 h cultivation (n = 3, mean ± s.d.). (L) Schematic diagram of obtaining and processing human cartilage. (M) Digital Radiography (DR) images of knee joint in OA patients at different grades by KL grading system (n = 6, mean ± s.d.). (N–O) Safranin-O Fast-Green (SO/FG) staining and OARSI scores of human cartilage sections, scale bar = 200 μm, 40 μm. (P–Q) Immunohistochemical staining and quantitative analysis of SLC31A1, scale bar = 200 μm, 40 μm (n = 6, mean ± s.d.). (R–S) Immunofluorescence staining and quantitative analysis of DLAT in cartilage of patients with different OA grades, and red foci represent protein aggregation, scale bar = 200 μm, 40 μm (n = 6, mean ± s.d.). ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001, ∗ means 1 % O₂ vs. 3 % O₂, 5 % O₂ or 7 % O₂ groups, and S0 vs. S1, S2, S3 or S4 groups.

Fig. 4

Inhibiting cuproptosis alleviated OA-like phenotypes in primary chondrocytes. (A) CCK-8 assay of primary chondrocytes treated with different concentrations of TTM (n = 10, mean ± s.d.). (B) Copper level in primary chondrocytes after TTM treatment (n = 3, mean ± s.d.). (C) The mRNA levels of Dlat, Dld, Pdha1 and Pdhb were detected by qRT-PCR (n = 3, mean ± s.d.). (D–E) Immunofluorescence staining and quantitative analysis of DLAT after TTM intervention, scale bar = 200 μm (n = 3, mean ± s.d.). (F–H) Detection of H₂O₂, MDA levels and SOD activity of primary chondrocytes after TTM treatment (n = 3, mean ± s.d.). (I–L) DHE staining and DCFH staining and quantitative analysis for detecting ROS levels in different groups, scale bar = 200 μm, 50 μm (n = 3, mean ± s.d.). (M–N) JC-1 staining of mitochondrial membrane potential in primary chondrocytes, scale bar = 200 μm, 50 μm (n = 3, mean ± s.d.). (O) Immunofluorescence staining of ATP7B, scale bar = 200 μm (n = 3, mean ± s.d.). (P) Western blot assay and quantitative analysis of protein levels of ATP7B and SLC31A1 (n = 3, mean ± s.d.) (Q) CCK-8 assay was used to detect the cell viability of primary chondrocytes (n = 3, mean ± s.d.). (R) EdU staining of primary chondrocytes after TTM treatment, scale bar = 200 μm, 50 μm (n = 3, mean ± s.d.). (S) The protein levels of ADAMTS5 and MMP13 were detected by Western blot (n = 3, mean ± s.d.). ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001, ^#P < 0.05, ^##P < 0.01, ^###P < 0.001, ∗ means 1 % O₂ vs. 7 % O_2,^# means 7 % O₂ vs. TTM.

Fig. 5

TTM attenuated cartilage degeneration in OA rat model. (A) Diagram of the experimental procedure and designs. (B) Immunohistochemical staining of SLC31A1 after TTM treatment in OA rats, scale bar = 500 μm, 100 μm (n = 6, mean ± s.d.). (C) Immunohistochemical staining of LIAS after TTM treatment in OA rats, scale bar = 500 μm, 100 μm (n = 6, mean ± s.d.). (D) The images of distance from the left and right hind limbs to the midline (n = 6, mean ± s.d.). (E) Data analysis of distance from the left and right hind limbs to the midline, hind paw area and duration in rats (n = 6, mean ± s.d.). (F) The curve graph of changes in paw area and duration time (n = 6, mean ± s.d.). (G) Data analysis of hind paw area and duration time in rats (n = 6, mean ± s.d.). (H) Data analysis of paw force and stride length (n = 6, mean ± s.d.). (I) Micro-CT images of sagittal and coronal of the knee joints after TTM treatment (n = 6, mean ± s.d.). (J) Safranin-O Fast-Green (SO/FG) staining was used to detect acidic polysaccharide levels on the surface of cartilage, scale bar = 500 μm, 100 μm (n = 6, mean ± s.d.). (K) Immunohistochemical staining of COL2A1 after TTM treatment in OA rats, scale bar = 500 μm, 100 μm (n = 6, mean ± s.d.). (L) Immunohistochemical staining of MMP13 after TTM treatment in OA rats, scale bar = 500 μm, 100 μm (n = 6, mean ± s.d.). ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001, ^#P < 0.05, ^##P < 0.01, ^###P < 0.001, ∗ means Ctrl vs. Model, ^# means Model vs. TTM.

Fig. 6

Knockdown of HIF-1α weakened the protective effect of hypoxia on cuproptosis in primary chondrocytes. (A) The mRNA level of HIF-1α was detected by qRT-PCR assay (n = 3, mean ± s.d.). (B) Immunofluorescence staining of HIF-1α in primary chondrocytes after siHIF-1α transfection, scale bar = 200, 40 μm (n = 3, mean ± s.d.). (C) Western blot and quantitative analysis of HIF-1α in different groups (n = 3, mean ± s.d.). (D) Immunofluorescence staining of ATP7B after silencing HIF-1α, scale bar = 200, 40 μm (n = 3, mean ± s.d.). (E) Western blot assay and quantitative analysis of the copper transporters ATP7B and SLC31A1 (n = 3, mean ± s.d.). (F) Cu content was evaluated by Cell copper colorimetric assay after siHIF-1α transfection, (n = 3, mean ± s.d.). (G) Immunofluorescence staining of DLAT, scale bar = 200, 40 μm (n = 3, mean ± s.d.). (H) The qRT-PCR assay of Pdha1, Pdhb, Dld and Dlat. (I–J) JC-1 staining was used to detect mitochondrial membrane potential, scale bar = 200, 40 μm (n = 3, mean ± s.d.). (K–L) EdU staining and quantitative analysis for proliferation of primary chondrocytes, scale bar = 200, 40 μm (n = 3, mean ± s.d.). (M) Cell viability of different groups was detected by CCK-8 assay (n = 3, mean ± s.d.). (N–O) Cell death staining of primary chondrocytes after HIF-1α knockdown, scale bar = 200, 40 μm (n = 3, mean ± s.d.). ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001, ns = no significance, ∗ means NC vs. siHIF-1α, ns means Ctrl vs NC. All experiments were conducted under 1 % O₂ conditions.

Fig. 7

Stabilization of HIF-1α inhibited cuproptosis under 7 % O₂. (A) CCK-8 assay was used to detect cell activity and screen for the optimal drug concentration after treatment of DMOG (n = 6, mean ± s.d.). (B) qRT-PCR was used to detect mRNA level of HIF-1α (n = 3, mean ± s.d.). (C) The expression level of HIF-1α after DMOG treatment was performed by Western blot (n = 3, mean ± s.d.). (D) Immunofluorescence staining of HIF-1α, scale bar = 200 μm, 40 μm (n = 3, mean ± s.d.). (E) The mRNA expression of Atp7b and Slc31a1 (n = 3, mean ± s.d.). (F) Western blot assay of ATP7B and SLC31A1 in 3 groups after DMOG treatment (n = 3, mean ± s.d.). (G) Immunofluorescence staining of ATP7B, scale bar = 200 μm, 40 μm (n = 3, mean ± s.d.). (H) Comparison of copper levels in primary chondrocytes of different groups after DMOG treatment. (I) The qRT-PCR assay of Pdhb, Pdha1, Dld and Dlat (n = 3, mean ± s.d.). (J) Immunofluorescence staining and quantitative analysis of DLAT, scale bar = 200 μm, 20 μm (n = 3, mean ± s.d.). (K) JC-1 staining and quantitative analysis of mitochondrial membrane potential, scale bar = 200 μm, 40 μm (n = 3, mean ± s.d.). (L) EdU staining and quantitative analysis of primary chondrocytes after DMOG intervention, scale bar = 200 μm, 40 μm (n = 3, mean ± s.d.). (M) Primary chondrocyte death rate was detected through Live/Dead cell assay and quantified by imageJ software, scale bar = 200 μm, 100 μm (n = 3, mean ± s.d.). (N–O) Immunofluorescence staining and quantitative analysis of ACAN and COL2A1, scale bar = 200 μm, 40 μm (n = 3, mean ± s.d.). ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001, ^#P < 0.05, ^##P < 0.01, ^###P < 0.001, ∗ means 1 % O₂ vs. 7 % O_2,^# means 7 % O₂ vs. DMOG.