Abstract NDP52, a constituent of the selective autophagy receptors (SARs), was recognized for its involvement in facilitating substrate degradation via autophagic bridging. However, its autonomous function apart from autophagy remained largely unexplored. Here, we reported that NDP52 was down-regulated in degenerated chondrocytes. Besides, NDP52 deficiency promoted the extracellular matrix (ECM) degradation, inflammation, cell apoptosis and senescence via its autophagy-independent functions. The absence of NDP52 disrupted the flow of electron respiration chains and led to the production of intracellular mitochondrial reactive oxygen species (mtROS). Subsequent mechanistic investigations revealed that the downregulation of NDP52 upregulated the expression levels of mitochondrial complex Ⅰ by modulating MTIF3 expression, leading to reverse electron transport (RET) and mtROS production. Our research highlights the significance of NDP52 in facilitating chondrocyte degeneration and osteoarthritis, and provides insights into the distinctive mechanism by which autophagy receptors NDP52 induce intracellular mitochondrial ROS dysregulation via non-canonical pathways. Keywords: Osteoarthritis, Chondrocyte, NDP52, Mitochondria reactive oxygen species (mtROS), Reverse electron transport (RET), MTIF3 Graphical abstract Figure abstract. Illustration describing mechanism by which NDP52 deficiency leads to chondrocyte degeneration via RET-induced mtROS. [49]Image 1 [50]Open in a new tab 1. Introduction Selective autophagy is a cellular process in which specific intracellular components are targeted for degradation by being imported into the autophagosome and subsequently delivered to the lysosome [[51]1]. This process relies on several selective autophagy receptors (SARs) such as p62, NDP52, OPTN, NBR1 and TAX1BP1, which contain a LC3 interaction region (LIR) [[52]2] domain that enables them to bind to LC3 and facilize the degradation to implement the degradation process of intracellular components. Among these SARs, NDP52 is capable to recognize and bind to various intracellular components like RNA granules, bacteria, damaged mitochondrial and tau protein [[53][3], [54][4], [55][5], [56][6]] through its ubiquitin-binding domains (UBD) or SKIP carboxyl homology (SKICH) [[57]7] to activate autophagy. Andre Franke et al. reported that NDP52 takes center stage in Crohn's disease by targeting TLR signaling adaptors for degradation via autophagy [[58]8]. In addition, NDP52 also exhibits non-autophagy functions, for instance, NDP52 has been shown to degrade hepatitis B virus through Rab9-dependent lysosome [[59]9]. However, the autophagy-independent role of NDP52 in disease are largely unexplored. Osteoarthritis, one of the most prevalent forms of arthritis, poses a substantial health challenge on a global scale by inducing joint deterioration and persistent pain that significantly hinders individuals' daily functioning and occupational performance [[60]10]. The development of OA is linked to multiple joint components such as the meniscus, ligament, cartilage, and bone, with cartilage degradation being identified as the primary factor in OA pathogenesis [[61]11,[62]12]. Specifically, as the sole cell type present in cartilage tissue, the degeneration of chondrocyte plays a crucial role in the advancement of OA [[63]13]. Alongside this, several mechanisms, such as ECM degradation, inflammation, cell apoptosis, cellular senescence and metabolic abnormalities [[64][14], [65][15], [66][16], [67][17]], have been identified as contributing factors to chondrocyte degeneration in the advancement of osteoarthritis. These processes are regulated by a multitude of signaling pathways or signals, including reactive oxygen species (ROS) [[68]18]. Despite this, the precise mechanisms underlying the progression of OA are still not fully understood. Mitochondrial reactive oxygen species (mtROS) have been implicated in accelerating OA progression by promoting ECM degradation, inflammation, cell apoptosis and cellular senescence [[69][19], [70][20], [71][21], [72][22]]. It has been observed that mtROS primarily originate from electron leakage within the mitochondrial electron transport chain, particularly in complex Ⅰ, leading to the generation of superoxide anion as an initial form of mtROS in normal cells [[73]23]. Reverse electron transport (RET) is considered as a significant contributor to the substantial production of pathogenic mtROS generation [[74]24]. During RET, electrons reverse their flow back to complex Ⅰ and subsequently move in reverse from ubiquinone to NADH, resulting in a significant amount of electron leakage [[75]25]. The upregulation of complex Ⅰ plays a crucial role in facilitating RET [[76]26,[77]27]. The overexpression of complex Ⅰ could induce high NAD^+/NADH rate, high mitochondrial membrane potential, increased level of ubiquinone and a proportional imbalance between mitochondrial complexes [[78]25,[79]28], which subsequently drive RET. The induction of mtROS by RET has been implicated in various pathological processes, such as macrophage activation in tuberculosis, myocardial ischemia-reperfusion, tumor stem cell activation and aging [[80][29], [81][30], [82][31]]. While NDP52 is recognized as one of mitophagy receptors and mtROS may serve as a signal for mitophagy [[83]32], the potential interaction between mtROS, NDP52, and mitophagy in the progression of osteoarthritis remains uncertain. Our study indicated that a deficiency in NDP52 leads to the production of mtROS in chondrocytes, resulting in the degeneration of chondrocytes and the acceleration of osteoarthritis progression. This deficiency also leads to the overexpression of complex Ⅰ through an autophagy-independent mechanism, which subsequently triggers a cascade of events leading to the generation of reactive oxygen species and mitochondrial dysfunction. These findings offer a unique perspective on the role of autophagy receptors NDP52 in disease pathogenesis. 2. Material and method 2.1. Human specimen collection Human joint specimen collection protocols were approved by the Ethical Committee of the Sir Run Run Shaw Hospital Zhejiang University School of Medicine (Approval NO.: 1080 for ethics review at the Sir Run Run Shaw Hospital in 2023). 15 sets of cartilage tissue were obtained from patients diagnosed with knee OA who had undergone total knee replacement surgery. The OA tissues were collected from the damaged media side of the tibia plateau, whereas the normal tissues were obtained from the comparatively intact lateral side of the same patient. 2.2. Animal studies The animal studies protocols were approved by the Animal Research Ethical Committee of Zhejiang University (Approval NO.: ZJU20230242). The mice with Ndp52 conventional knockout in C57BL/6 background were constructed using the CRISPR-Cas9-based homology recombination method by Saiye Inc. (S–KO-14951, Suzhou, China). Simply, a single guide RNA (sgRNA) was engineered and high-throughput electrofertilization was employed to induce targeted disruption of 2–6 exons, ultimately leading to the generation of Calcoco2 gene knockout mice. For the DMM surgery, 10-week-old male mice underwent medial meniscotibial ligament transection surgery in the knee joint. Sham surgery was performed in control mice. 8 weeks after surgery, all mice were euthanized. The Adeno-associated viral vector (AAV) utilized in this study was synthesized and provided by Hanbio Tech (Shanghai, China). AAV-2 with CMV promoter was used. Virus (1 × 10^12 VG/ml) injections were performed using a 10 μl microsyringe (Hamilton Company, Reno, NV, USA) with a 34G needle, with a total of 10 μl AAV injected into knee joint of mice. The injection commenced 2 weeks prior to the DMM surgery. Hot-plate and rotarod assays were used for animal behavioral experiments in accordance with established protocols [[84]33]. For imaging examination, the mice knee joints were prepared with fixing after soft tissue removed and then scanned by high-resolution μCT scanner (Skyscan1275, Bruker, Belgium). Subsequent 3D reconstruction was performed post-scanning. 2.3. Histological staining and immunofluorescence Human or mouse joints were fixed in 4 % polyformaldehyde, then embedded in paraffin after decalcification and dehydration, finally sectioned at 5 μm. The sections were stained with safranin O (S2255, Sigma-Aldrich, USA) and fast green (F7252, Sigma-Aldrich). Histological score was assessed by two blind experienced scorers based on OARSI grade [[85]34,[86]35], osteophyte formation score and Mankin score [[87]36]. Human OARSI score ranks from 0 to 24 based on both the severity (grade) and extent (stage) of OA in the articular cartilage [[88]34]. And mouse OARSI score ranks from 0 to 6 based on the extent of cartilage damage [[89]35]. For immunofluorescence, sections were permeabilized in Triton X-100 in PBS. After being blocked with 5 % BSA, sections were incubated with corresponding primary antibodies at 4 °C overnight, followed by incubated with secondary antibodies [Alexa Fluor 488] for 30 min. The sections were finally scanned using KFBIO digital scanner (KFBIO, China). The positive cell counting was performed by two blinded counters and then the average was taken. 2.4. Western blotting Human and mouse cartilage tissues were ground into powder in liquid nitrogen and lysed with strong radioimmunoprecipitation assay (RIPA) lysis buffer (FD009, Fudebio, China) to extract total protein. Cells were treated with RIPA to extract total protein. Equal amounts of protein were separated on SDS-PAGE and then transferred to 0.22 μm polyvinylidene difluoride (PVDF) membranes. The membranes were blocked with 5 % milk, incubated with primary antibody at 4 °C overnight and continued with secondary antibody for 1 h. Finally, the bands were detected using Amersham Imager (GE, UK) by a chemiluminescence kit (FD8030, Fudebio). The antibodies used in this study were listed in [90]Supplemental Table 1. 2.5. Cell lines and cell C28/I2 normal chondrocyte cells line was purchased from Sigma-Aldrich and cultured in Dulbecco's modified Eagle's medium (DMEM, CR12800-S, Cienry, China) containing 10 % FBS (NFBS2500a, Noverse, Australia) and 1 % penicillin/streptomycin. For mouse chondrocytes extraction, 5-day-old wild type and Ndp52-KO mice were used. The femoral condyle and tibial plateau of mice were sectioned and digested by 0.2 % type II collagenase (9001-12-1, Sigma-Aldrich). Cells were collected after supernatant being removed and then cultured in DMEM supplemented with 10 % FBS and 1 % penicillin/streptomycin. The cells were cultured in an incubator with 5 % CO[2] at 37 °C. For human primary chondrocytes, cells were isolated from cartilage samples of tibial plateaus and femoral condyles and dissociated enzymatically with collagenase overnight, then cultured in DMEM supplemented with 10 % FBS and 1 % penicillin/streptomycin. The primary chondrocytes were cultured in an incubator with 5 % CO[2] and 1 % O[2] at 37 °C. 2.6. Gene editing materials For RNA interference, the siRNA was synthesized by Gene Pharma (Shanghai, China). The sequences of siRNA were listed in [91]Supplemental Table 2. For cell NDP52 deleting, CRISPR plasmid (sc-416970-NIC) was purchase from Santa Cruz Biotechnology. For NDP52 and Flag-AOX overexpression, plasmid was designed and synthesized by Hanbio Tech (Shanghai, China) and Tsingke (Beijing, China). 2.7. Micromass culture and senescence β‐galactosidase (SA‐β‐Gal) staining Approximately 2 × 10^6 cells were suspended in 10 μL of DMEM medium and subsequently seeded in the central region of 24-well plates for micromass culture. Following a 14-day incubation period, the cells were fixed with 4 % paraformaldehyde. Alcian blue solution (G1560, Solarbio, China) were used to staining the extracellular matrix within micromass. Cellular senescence was detected using SA‐β‐Gal staining kit (C0602, Beyotime, China). Cells were stained according to manufacturer's instructions. 2.8. Cell counting kit-8 assay 5000 chondrocytes were seeded in the 96-well plate and cultured with assay performed at 0, 24, 48, and 72-h time points. The Cell Counting Kit (CK04, Dojindo, Japan) was utilized in accordance with the manufacturer's guidelines, and the absorbance at 450 nm was examined by microplatereader. 2.9. Flow cytometry Cell apoptosis was assessed through the use of Annexin V-FITC/PI apoptosis Kit (556547, BD Biosciences, USA) and Annexin V-APC/7-AAD apoptosis Kit (E-CK-A218, Elabscience, China). Cells were stained according to manufacturer's instructions and subsequently analyzed using a BD LSRFortessa flow cytometer (BD Biosciences). 2.10. ELISA Approximately 2 × 10^5 cells were seeded in the 6-well plate and cultured for 3 days. The extracellular fluid was collected and Human Interleukin 6 (IL-6) ELISA Kit (E-EL-H6156, Elabscience) was used to measure the amount of IL-6 in the extracellular fluid according to manufacturer's instructions. The absorbance at 450 nm was examined by microplatereader. 2.11. RNA extraction and real-time quantitative PCR Total RNA extraction and reverse transcription were performed using Cell/Tissue Total RNA Kit (19221ES50, Yeason, China) and PrimeScript RT Reagent (AG11705, AG, China) respectively. After the cDNA was mixed with primers and Hieff qPCR SYBR mix (11201 ES, Yeason) according to manufacturer's instructions, the amplification reaction was carried out and measured using QuantStudio 6 Flex Real-Time PCR System (Thermo Fisher Scientific). Primers were supplied by Tsingke (Beijing, China). β-actin was used for reactions normalization. 2.12. 4D label-free proteomics analysis and untargeted metabolomics The 4D label-free proteomics analysis and untargeted metabolomics were supported by Applied Protein Technology (China). Briefly, negative control and NDP52-KO C28/I2 cells were collected for label-free proteomics analysis and untargeted metabolomics. For bioinformatic analysis, differentially expressed proteins and metabolites were defined as having |log[2]FC (NC/KO)|>1 and p-value<0.05. Gene Ontology (GO), Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis and Gene set enrichment analysis (GSEA) were conducted using R “clusterProfiler” package and visualized by Clue GO apps in Cytoscape and R “ggplot2” package. Ingenuity pathway analysis (IPA) was supported by QIAGEN. Protein-protein interaction (PPI) analysis was performed using STRING database ([92]https://www.string-db.org/) and visualized by Cytoscape. Heatmap of the correlation network of combined analysis of proteomics and metabolomics was performed by OmicStudio ([93]https://www.omicstudio.cn/tool). 2.13. Mitochondrial reactive oxygen species (mtROS), mitochondrial complex Ⅰ activity, aconitase (ACO) activity and NAD^+/NADH rate detection MitoSOX Red ([94]M36008, Thermo Fisher Scientific) was used for cell staining to detect mtROS, while MitoTracker Green (M7514, Thermo Fisher Scientific) was used for labeling mitochondria. Cells were stained and analyzed by a BD LSRFortessa flow cytometer (BD Biosciences) or photographed using a confocal microscope (Olympus). Cell Mitochondrial Complex Ⅰ (NADH-CoQ Reductase) Activity Assay Kit (E-BC-K834-M, Elabscience), NAD+/NADH Colorimetric Assay Kit (WST-8) (E-BC-K804-M, Elabscience) and Aconitase (ACO) Activity Assay Kit (Cat:BC4485, Solarbio, China) were used according to manufacturer's instructions. 2.14. JC-1 detection Mitochondrial Membrane Potential Assay Kit (with JC-1) (E-CK-A301, Elabscience) was used to detect mitochondrial membrane potential. Cells were stained according to manufacturer's instructions and then photographed by a confocal microscope (Olympus). 2.15. Cellular OCR assay with Seahorse OCR was assayed, as reported previously [[95]37]. Primary human chondrocytes with or without NDP52 knockdown were seeded at 1 × 10^5 cells/well in 24-well Seahorse XF plates. One medium-only well per row served as a background control. Cells were washed and equilibrated with 500 μl of glucose-supplemented XF assay buffer per well, then incubated for 30 min in a CO[2]-free environment. Metabolic modulators (FCCP, oligomycin, rotenone, and antimycin A) were sequentially injected via designated ports according to the experimental protocol. Real-time OCR values (pmol O[2]/min) were measured using the Seahorse XF-24 analyzer. Non-mitochondrial OCR was determined by subtracting post-inhibition OCR (after rotenone/antimycin A treatment) from total OCR under basal conditions or after oligomycin/FCCP treatment. 2.16. Blue-Native Page Blue-Native Page was performed as previously described [[96]38]. Briefly, following the extraction of mitochondria from cells, mitochondrial membrane protein complexes were isolated and solubilized using nonionic detergents. Subsequently, the samples were mixed with the anionic dye, Coomassie G-250, and separated using 4–13 % Blue Native Page gel (BL1437A, Biosharp, China). Finally, the gels were stained with Coomassie Blue or subjected to in-gel activity staining. 2.17. Statistic analysis SPSS 20.0 was used to perform statistical analysis. Results were visualized by GraphPad Prism 8. All data were shown as the mean ± standard deviation (SD). Paired or unpaired two-tailed t-test and one-way analysis of variance (ANOVA) with a post-hoc Turkey test were used to test significance of differences between two groups comparisons and multiple comparisons respectively. P values represented statistical significance and were indicated in figures (∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ns p > 0.05). 3. Results 3.1. NDP52 is down-regulated in degenerated articular chondrocytes While previous studies have indicated a connection between autophagy and OA, whether SARs function in OA remains unclear. Firstly, samples from patients undergoing total knee replacement surgery were collected ([97]Fig. S1a). Immunofluorescence analysis demonstrated that all SARs were present in cartilage, with NDP52 exhibited the highest protein expression among the five SARs ([98]Fig. 1a). Subsequent analysis of data from [99]GSE114007 revealed that NDP52 was the sole down-regulated SAR in OA cartilage samples ([100]Fig. 1b, c and [101]Fig. S1b–e). The immunofluorescence analysis revealed a decrease of NDP52 protein level in degenerated area compared to relatively normal area within same samples ([102]Fig. 1d and e). Western blotting of tissue total protein also indicated a decreased NDP52 level in human degenerated cartilage ([103]Fig. 1f and g). Additionally, samples from mice undergoing DMM surgery in one limb and sham surgery in another were analyzed, demonstrating a decrease of Ndp52 protein level in DMM group compared to the sham group ([104]Fig. 1h–k). Fig. 1. [105]Fig. 1 [106]Open in a new tab The level of NDP52 decreases in OA. (a) Immunofluorescence and analysis of NDP52, p62, NBR1, OPTN and TAX1BP1 of normal human cartilage. n = 3 per group. Scale bar, 200 μm. (b) Heatmap of SARs gene expression in [107]GSE114007. (c) Normalized gene expression of NDP52 in GEO dataset [108]GSE114007. n = 18 in normal group and n = 20 in OA group. (d) The immunofluorescence of normal and OA cartilage of human. Scale bar, 50 μm. (e) Positive cell rate of cartilage. n = 7 per group. (f and g) Immunoblot detection (f) and analysis (g) of NDP52 and GAPDH in human normal and OA cartilage tissue. n = 5 per group. (h) The immunofluorescence of joints of mice that underwent sham or DMM surgery. Scale bar, 50 μm. (i) Positive cell rate of joint. n = 7 per group. (j and k) Immunoblot detection (j) and analysis (k) of Ndp52 and Gapdh in mice joints. n = 5 per group. One-way analysis of variance (ANOVA) with a post-hoc Turkey test (a), unpaired (c, g and k) and paired (e and i) two-tailed Student's t-test were used to analyze data, which was presented as the means ± SD. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ns p > 0.05. 3.2. Ndp52 deficiency exacerbates overload-induced OA in mice Ndp52 knockout mice in C57/B6L background were generated to investigate the impact of Ndp52 on OA. The in vivo experimental process was illustrated in [109]Fig. 2a. Ndp52 KO mice did not manifest conspicuous differential alterations in systemic physiology and in the synovium and subchondral bone of the joints in 18 weeks ([110]Fig. S2a–c). Histological analysis suggested that Ndp52 knockout resulted in more severe damage to the articular joint cartilage following DMM surgery ([111]Fig. 2b). Consistent results were obtained from the histological score assay ([112]Fig. 2c–e). Immunofluorescence analysis further confirmed Ndp52 expression in WT and Ndp52-KO mice ([113]Fig. 2f and g). Furthermore, the protein expression of Acan was notably decreased in Ndp52 knockout mice ([114]Fig. 2f–h) while p16^Ink4a expression was significantly increased ([115]Fig. 2f–i). These findings suggested that the absence of Ndp52 may contribute to extracellular matrix degradation and cellular senescence. At the same time, an increased number of osteophytes were observed in Ndp52 knockout mice following DMM surgery ([116]Fig. 2j). These mice exhibited decreased pain threshold and mobility in hot-plate and rotarod assays post-surgery ([117]Fig. 2k and l). Taken together, Ndp52 was found to play a crucial role in the progression of OA. Fig. 2. [118]Fig. 2 [119]Open in a new tab Knockout of Ndp52 accelerates OA progress. (a) Schematic diagram of experiment design to evaluate function of Ndp52 in vivo. (b) The safranin O-fast green staining of joints of WT and Ndp52-KO mice that underwent sham or DMM surgery. Scale bar, 50 μm (c, d, and e) OARIS (c), osteophyte formation (d) and Mankin (e) score of joints of mice. n = 4 per group. (f, g, h and i) Immunofluorescence and analysis of Ndp52 (g), Acan (h) and p16^Ink4a (i). n = 4 per group. Scale bar, 50 μm. (j) Reconstructed μCT image of mice. Arrows indicate osteophytes. (k and l) The hot plate (k) and rotarod (l) assays of the mice. n = 4 per group. One-way analysis of variance (ANOVA) with a post-hoc Turkey test was used to analyze data, which was presented as the means ± SD. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ns p > 0.05. 3.3. NDP52 knockout promoted chondrocytes degeneration via autophagy-independent way To further understand the function roles of NDP52 in chondrocytes, we performed NDP52 knockout and knockdown in chondrocyte cell line C28/I2. Chondrocytes from WT and Ndp52-KO mice were also used. Firstly, through Alcian blue staining, C28/I2 chondrocytes with NDP52 knockout and knockdown showed a diminished coloration which means a reduced ECM accumulation ([120]Fig. 3a and [121]Fig. S3a). Subsequent CCK8 assay showed a decreased cell survival rate in NDP52 knockout and knockdown C28/I2 chondrocytes ([122]Fig. 3b and [123]Fig. S3b) and flow cytometry indicated an increased apoptosis rate ([124]Fig. 3c and [125]Fig. S3c). Furthermore, NDP52 knockout and knockdown induced C28/I2 chondrocytes senescence, as demonstrated by elevated levels of IL-6 in extracellular fluid of C28/I2 chondrocytes ([126]Fig. 3d and [127]Fig. S3d) and SA-β-gal staining assay ([128]Fig. 3e, f and [129]Fig. S3e and f). Western blotting further confirmed chondrocytes senescence by its marker (p16^INK4A, p21^CIP1) in C28/I2 chondrocytes with NDP52 knockout and knockdown and primary chondrocytes with NDP52 knockdown ([130]Fig. 3g and [131]Fig. S3g). Western blotting also verified ECM degradation by increased expression of ECM catabolism markers (MMP13, ADAMTS5) and decreased anabolism marker ACAN in NDP52 knockout and knockdown C28/I2 chondrocytes and primary chondrocytes ([132]Fig. 3g and [133]Fig. S3g). Given the association between cellular senescence and inflammation [[134]39], we next examined the expression level of inflammatory markers. The results revealed increased mRNA expression level of inflammatory factor (TNF-α, IL-1β, IL-6 and IL-8) by qPCR analysis in NDP52 knockout and knockdown C28/I2 chondrocytes ([135]Fig. 3h and fig. S3h). Meanwhile, mRNA expression of ECM metabolism and senescence marker (MMP13, ADAMTS5, ACAN, p16^INK4A, p21^CIP1) were also detected ([136]Fig. 3h and fig. S3h). Likewise, mice chondrocytes with Ndp52-KO also showed increased ECM degradation, cell apoptosis, inflammation and cell senescence ([137]Fig. S4). Fig. 3. [138]Fig. 3 [139]Open in a new tab Knockout of NDP52 in chondrocytes results in degeneration via non-autophagy way. (a) Micromass culture with alcian blue staining of NC and NDP52-KO C28/I2 chondrocytes. (b) Cell survival curves of treated chondrocytes. n = 3 per group. (c) Flow cytometry of cell apoptosis of treated chondrocytes. (d) ELISA analysis of IL-6 of treated chondrocytes. n = 3 per group. (e and f) SA-β-gal staining (e) and analysis (f) of treated chondrocytes. n = 3 per group. Scale bar, 100 μm. (g) Immunoblot detection of NDP52, ACAN, MMP13, ADAMTS5, p16^INK4A, p21^CIP1 and GAPDH in treated chondrocytes. (h) The mRNA level of NDP52, ACAN, MMP13, ADAMTS5, TNF-α, IL-1β, IL-6, IL-8, p16^INK4A and p21^CIP1 in treated chondrocytes. n = 3 per group. (i) Illustration of full-length NDP52. (j) Micromass culture with alcian blue staining of NDP52-KO C28/I2 chondrocytes overexpressing CTL, NDP52-WT and NDP52-mut. (k) Cell survival curves of treated NDP52-KO chondrocytes. n = 3 per group. (l) ELISA analysis of IL-6 of treated NDP52-KO chondrocytes. n = 3 per group. (m) Flow cytometry of cell apoptosis of treated NDP52-KO chondrocytes. (n and o) SA-β-gal staining (n) and analysis (o) of treated NDP52-KO chondrocytes. n = 3 per group. Scale bar, 100 μm. (p) Immunoblot detection of NDP52, ACAN, MMP13, ADAMTS5, p16^INK4A, p21^CIP1 and GAPDH in treated NDP52-KO chondrocytes. (q) The mRNA level of ACAN, MMP13, ADAMTS5, TNF-α, IL-1β, IL-6, IL-8, p16^INK4A and p21^CIP1 in treated NDP52-KO chondrocytes. n = 3 per group. Unpaired two-tailed Student's t-test (b, d, f and h) and one-way analysis of variance (ANOVA) with a post-hoc Turkey test (k, l, o and q) were used to analyze data, which was presented as the means ± SD. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ns p > 0.05. Since NDP52 is a member of SARs, we proceeded to evaluate the LC3 protein levels, a marker of autophagy. Both of the western blotting and mCherry-GFP-LC3 fluorescent staining indicated the autophagic flux remains unaltered in NDP52 knockout C28/I2 chondrocytes ([140]Fig. S5a–c). Furthermore, NDP52 is a multi-domain protein, with the LIR domain being responsible for its interaction with LC3 during autophagic process ([141]Fig. 3i). To further investigate the impact of autophagy on degeneration of chondrocytes induced by NDP52 knockout, we generated a mutant form of NDP52(V136S) with a modified LIR motif. Previous study had proved that NDP52(V136S) could inhibit NDP52-mediated selective autophagy process [[142]40]. And both wild-type and mutant forms of NDP52 were overexpressed in NDP52-KO C28/I2 chondrocytes as demonstrated by Western blotting ([143]Fig. 3p). Alcian blue staining demonstrated that both wild-type and mutant forms were able to rescue the ECM accumulation ([144]Fig. 3j). Moreover, the increase of cell apoptosis and cellular senescence resulting from NDP52 knockout was ameliorated through the overexpression of wild-type and mutant forms of NDP52 ([145]Fig. 3k–o). Subsequent Western blotting also verified rescued ECM metabolism and senescence by wild-type and mutant forms of NDP52 ([146]Fig. 3p). The evaluation of mRNA level of ECM metabolism markers, senescence and inflammatory markers also yielded consistent findings ([147]Fig. 3q). In a word, NDP52 modulates extracellular matrix, apoptosis, cellular senescence, and inflammation in chondrocytes during degeneration through an autophagy-independent mechanism. 3.4. Chondrocytes with NDP52 knockout showed mitochondrial dysfunction To elucidate the process by which NDP52 knockout led to chondrocyte degeneration, a 4D label-free quantitative proteomic analysis was conducted using wild-type and NDP52 knockout C28/I2 chondrocytes ([148]Fig. S6a). Subsequent analysis utilizing GO, KEGG, GSEA and IPA analysis revealed the enrichment of several OA-related pathways including necroptosis, glycosamino-glycan degradation and reactive oxygen species ([149]Fig. 4a and [150]Fig. S6b and c). Further investigation utilizing GO clue analysis highlighted the importance of oxidative phosphorylation and reactive oxygen species metabolic process in this particular scenario ([151]Fig. S6d), suggesting a potentially crucial role for mitochondrial function. Various pathways related to mitochondria including mitochondrial dysfunction, the electron transport chain, and the mitochondrial inner membrane were found to be enriched ([152]Fig. 4b, c and [153]Fig. S6b and c). These findings suggest that mtROS and the respiratory complex might be significant contributors to mitochondrial dysfunction, ultimately leading to the degeneration of NDP52-KO chondrocytes. The heatmap analysis of mitochondrial-related proteins also revealed alterations in the mitochondrial respiratory complex ([154]Fig. 4d). Given the pivotal role of mitochondria in cellular metabolism net, untargeted metabolomics was performed. An intergraded omics analysis highlighted modifications in arginine biosynthesis ([155]Fig. S6e), a process related to mitochondria and OA [[156]41,[157]42]. The correlation network heatmap further demonstrated these observations ([158]Fig. 4e). Fig. 4. [159]Fig. 4 [160]Open in a new tab NDP52 knockout induces chondrocytes mitochondrial dysfunction. (a) Ridgeline plots of GSEA enrichment analysis of whole proteins of NC and NDP52-KO chondrocytes. (b and c) GSEA analysis of electron transport chain (b) and Complex Ⅰ (c) based on the proteomic profiles. (d) Heatmap of mitochondrial-related differentially expressed proteins. (e) Heatmap of the correlation network of combined analysis of proteomics and metabolomics. (f) Mitochondrial TEM of NC and NDP52-KO chondrocytes. Scale bar, 200 nm. (g) Immunoblot detection of NDP52, NDUFV1, ABAT, SOD2 and GAPDH of treated chondrocytes. (h, i and j) Flow cytometry analysis (h), confocal microscopy photograph (i) and analysis (j) of MitoSOX staining in NC and NDP52-KO chondrocytes. n = 3 per group. Scale bar, 25 μm. (k) Flow cytometry analysis of MitoSOX staining in NDP52-KO chondrocytes with overexpression of CTL, NDP52-WT and NDP52-mut protein. Unpaired two-tailed Student's t-test was used to analyze data, which was presented as the means ± SD. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ns p > 0.05. Mitochondrial transmission electron microscope (TEM) was utilized to assess the impact of NDP52 knockout on mitochondrial integrity in chondrocytes, which showed an obscureness of swelling mitochondria, crista disappearance in NDP52-KO chondrocytes ([161]Fig. 4f). Western blot analysis revealed alterations in the expression of mitochondrial protein NDUFV1, metabolism-related protein ABAT [[162]43] and ROS-related protein SOD2 [[163]44] ([164]Fig. 4g). After MitoSOX staining, through flow cytometry and confocal microscope, we further confirmed an increase in mtROS levels in NDP52 knockout and knockdown chondrocytes ([165]Fig. 4h–j and [166]Fig. S7a). Considering the chondrocytes’ oxidative phosphorylation differs in different oxygen condition, we detected mtROS levels in NDP52 knockout chondrocytes in different oxygen concentration ([167]Fig. S7b). We further conducted experiments using primary cells under physiological oxygen conditions. NDP52 knockdown increased MitoSOX staining, which was inhibited by PEG-SOD [[168]45] ([169]Fig. S7c), and suppressed mitochondrial aconitase activity ([170]Fig. S7d). These results confirm that NDP52 knockout elevates mtROS levels. Seahorse analysis showed a reduction in oxidative phosphorylation, indicating diminished mitochondrial respiratory capacity and potential mitochondrial dysfunction ([171]Fig. S7e–i). And the increase of mtROS level in NDP52 knockout chondrocytes could be ameliorated by the overexpression of NDP52 WT and mut ([172]Fig. 4k) as well as MitoQ, a scavenger targeting mtROS ([173]Fig. S8a–c). To sum up, NDP52 knockout-induced elevation of mtROS levels resulted in mitochondrial dysfunction and subsequently degeneration of chondrocytes. 3.5. Reverse electron transport induced mitochondrial reactive oxygen species generation in NDP52-KO chondrocytes Given that mtROS has the potential to induce mitochondrial disorders and disrupt cellular function, our inquiry focused on the mechanisms underlying its generation in NDP52-KO chondrocytes. Our prior findings indicated that the mitochondrial respiratory complex, specifically the electron transport chain, may play a role in this process ([174]Fig. 4a–c and [175]Fig. S6b and c). A most recent study showed that enhanced complex Ⅰ activity could induce RET which caused mtROS generation and played a vital role in the pathogenic process of diseases [[176]46]. Coincidentally, analysis of protein-protein interactions revealed increased interactions among proteins of complex Ⅰ ([177]Fig. 5a). The proteins of complex Ⅰ exhibited greater differential expression ([178]Fig. 4d and [179]Fig. S6a) and played a more important role in enrichment analysis ([180]Fig. 4a). This observation aligns with previous studies indicating that complex Ⅰ was prominent site of electrons leakage [[181]23]. Western blotting further confirmed elevated expression of complex Ⅰ ([182]Fig. 5b), accompanied by an increase in its activity ([183]Fig. 5c). Moreover, NDP52-KO chondrocytes displayed enhanced mitochondrial membrane potential ([184]Fig. 5d and e), suggesting suggests the possibility of reverse electron transfer driven by a high mitochondrial membrane potential, leading to electron leakage. Fig. 5. [185]Fig. 5 [186]Open in a new tab NDP52 knockout induces mtROS bursting via RET in complex Ⅰ. (a) PPI of mitochondrial-related differentially expressed proteins. (b) Immunoblot detection of ATP5A1, UQCRC2, SDHB, MTCO2 and NDUFB8 in NC and NDP52-KO C28/I2 chondrocytes. (c) Complex Ⅰ activity detection of NC and NDP52-KO chondrocytes. n = 3 per group. (d and e) Mitochondrial membrane potential detection by JC-1 staining (d) and analysis (e) of treated chondrocytes. n = 3 per group. Scale bar, 50 μm. (f and g) Illustrations of FET (f) and RET (g) in Complex Ⅰ with mtROS production. ΔΨ, membrane potential; IMM, inner mitochondrial membrane; zigzag arrows, induction; red blunted arrows, inhibition. (h, i and j) Flow cytometry analysis (h), confocal microscopy photograph (i) and analysis (j) of MitoSOX staining in NC and NDP52-KO chondrocytes treated with DMSO or Rotenone. n = 3 per group. Scale bar, 25 μm. (k) Illustrations of AOX functioning in electron chain. (l) Immunoblot detection of Flag-AOX and GAPDH in NC and NDP52-KO chondrocytes expressing AOX or not. (m, n and o) Flow cytometry analysis (m), confocal microscopy photograph (n) and analysis (o) of MitoSOX staining in NC and NDP52-KO chondrocytes expressing AOX or not. n = 3 per group. Scale bar, 25 μm. (p) BN-PAGE followed by Coomassie Blue or in-gel activity staining of mitochondrial membrane protein isolated from NC and NDP52-KO chondrocytes. Unpaired two-tailed Student's t-test (c and e) and one-way analysis of variance (ANOVA) with a post-hoc Turkey test (j and o) were used to analyze data, which was presented as the means ± SD. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ns p > 0.05. In previous studies, it was often assumed that electrons leakage occurred in forward electron transport (FET) ([187]Fig. 5f). It was reported that the dysregulated expression of complex Ⅰ results in disrupted mitochondrial respiratory chain function, leading to a significant rise in mitochondrial reactive oxygen species (mtROS) levels and the reverse transport of electrons from ubiquinol (CoQH[2]) to NADH ([188]Fig. 5g) [[189]30]. Rotenone, a complex Ⅰ inhibitor, has been shown to impede electron transport, resulting in elevated mtROS levels during FET but reduced mtROS during RET [[190]23,[191]30] ([192]Fig. 5f and g). MitoSOX staining reflected that heightened mtROS levels observed in NDP52 knockout C28/I2 chondrocytes and NDP52 knockdown primary chondrocytes were mitigated by treatment with Rotenone ([193]Fig. 5h–j and [194]Fig. S9a). Two other widely used RET inhibitors, Diphenyleneiodonium (DPI) [[195]47] (inhibits flavins) and dimethyl malonate (DMM) [[196]48] (inhibits entry from complex II), also mitigated heightened mtROS levels observed in NDP52 knockout C28/I2 chondrocytes and NDP52 knockdown primary chondrocytes ([197]Fig. S9b–d). To further substantiate the assertion that RET was responsible for the increase in mtROS in NDP52-KO chondrocytes, Ciona intestinalis alternative oxidase (AOX) was introduced into the chondrocytes. AOX enzymes are exiting in plants, fungi, and certain invertebrates, but notably absent in vertebrates [[198]49]. AOX has the ability to catalyze the conversion of ubiquinol (CoQH[2]) to ubiquinone (CoQ) while transferring electrons from ubiquinol (CoQH[2]) to the electron terminal acceptor H[2]O, thereby preventing the occurrence of RET ([199]Fig. 5k), and subsequent elevation of mtROS levels [[200]50]. The expression of AOX in chondrocytes was confirmed through Western blotting ([201]Fig. 5l), which effectively mitigated the rise of mtROS observed in NDP52 knockout C28/I2 chondrocytes and NDP52 knockdown primary chondrocytes ([202]Fig. 5m–o and [203]Fig. S9e). To validate RET, we measured the NAD^+/NADH ratio. An increased ratio indicated enhanced Complex I function ([204]Fig. S9f), creating conditions for RET [[205]51,[206]52]. During the detection of mitochondrial membrane potential changes under various inhibitors, Rotenone (Complex Ⅰ inhibitor) had no significant effect, while DMM (Complex II inhibitor) and AOX reduced the membrane potential increase caused by NDP52 knockdown ([207]Fig. S9g–i). Previous studies have demonstrated that the mitochondrial electron transport chain is organized into supramolecular assemblies, known as supercomplexes, in order to maintain normal function [[208]53]. These supercomplex (SC) have been recognized for their involvement in the regulation of mtROS production by modulating electron flow in complex Ⅰ. The Blue Native PAGE was utilized and the reduction in SCs and an increase in complex Ⅰ were observed ([209]Fig. 5p). 3.6. NDP52 knockout specifically increasing complex Ⅰ level through MTIF3 Previous studies have indicated that mitochondrial translation initiation factor 3 (MTIF3) is essential for the coordinated assembly of electron transport complexes [[210]54], with the absence of MTIF3 leading to a specific reduction in complex Ⅰ subunits. Notably, MTIF3 expression was observed to be elevated in NDP52-KO chondrocytes ([211]Fig. 6a). Subsequently, Western blotting and immunofluorescence analyses confirmed an upregulation of MTIF3 in NDP52-KO C28/I2 chondrocytes and Ndp52-KO mice, respectively ([212]Fig. 6b–d). Moreover, MTIF3 expression was found to be elevated in the joints of DMM mice ([213]Fig. S10) and human degenerated cartilage ([214]Fig. 6e and f). Additionally, analysis of RNA count data from GEO dataset [215]GSE114007 for OA patients revealed a negative correlation between MTIF3 expression and NDP52 expression ([216]Fig. 6g). Subsequent immunofluorescence analysis indicated that MTIF3 expression was inversely correlated with the expression of NDP52 in both mice ([217]Fig. 6h and i) and human cartilage ([218]Fig. 6j and k). Besides, Western blotting showed a reduction in NDUFV1, a subunit of complex Ⅰ, in chondrocytes with the knockdown of MTIF3 ([219]Fig. 6l), indicating MTIF3's role in regulating mitochondrial complex Ⅰ. The following MitoSOX staining reflected that heightened mtROS levels observed in NDP52-KO chondrocytes were mitigated by MTIF3 knockdown ([220]Fig. 6m–o). Hence, it is hypothesized that the deletion of NDP52 induces an upregulation of MTIF3 expression, which in turn elevates the levels of complex Ⅰ. This elevation likely contributes to the degradation of SCs and initiation of complex I RET, thereby facilitating the generation of mtROS. Fig. 6. [221]Fig. 6 [222]Open in a new tab NDP52 knockout specifically promotes complex Ⅰ expression via increasing MTIF3. (a) MTIF3 showed in the volcano plots of differentially expressed proteins between NDP52-KO and NC chondrocytes. (b) Immunoblot detection of MTIF3 and GAPDH in chondrocytes. (c and d) Immunofluorescence (c) and analysis (d) of Ndp52 and Mtif3 of joint from WT and Ndp52-KO mice. n = 4 per group. Scale bar, 50 μm. (e and f) The safranin O-fast green staining and immunofluorescence (e) and analysis (f) of normal and OA cartilage of human. n = 5 per group. Scale bar, 200 μm (safranin O-fast green) and 50 μm (immunofluorescence). (g) Spearman correlation analysis between NDP52 and MTIF3 RNA counts of OA patients from GEO dataset [223]GSE114007 (n = 20). (h and i) Immunofluorescence (h) and positive cell rate Spearman correlation analysis (i) of Ndp52 and Mtif3 in joints of mice that underwent sham or DMM surgery. n = 4 per group. Scale bar, 50 μm. (j and k) Immunofluorescence (j) and positive cell rate Spearman correlation analysis (k) of NDP52 and MTIF3 of normal and OA cartilage of human. n = 5 per group. Scale bar, 100 μm. (l) Immunoblot detection of MTIF3, NDUFV1 and GAPDH in chondrocytes with MTIF3 knockdown. (m, n and o) Flow cytometry analysis (m), confocal microscopy photograph (n) and analysis (o) of MitoSOX staining in NC and NDP52-KO chondrocytes with MTIF3 knockdown or not. n = 3 per group. Scale bar, 25 μm. Unpaired two-tailed Student's t-test (d and f) and one-way analysis of variance (ANOVA) with a post-hoc Turkey test (o) were used to analyze data, which was presented as the means ± SD. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ns p > 0.05. 3.7. Inhibition of mtROS and RET attenuates the degeneration in NDP52-KO chondrocytes Subsequently, MitoQ and AOX were utilized to further investigate the involvement of mtROS and RET in the degeneration mediated by NDP52-KO. Alcian blue staining showed that MitoQ and AOX enhanced ECM accumulation in NDP52-KO C28/I2 chondrocytes ([224]Fig. 7a and b). Additionally, MitoQ and AOX were shown to inhibit cell apoptosis induced by NDP52 knockout ([225]Fig. 7c–f), as well as prevent cellular senescence of NDP52-KO chondrocytes ([226]Fig. 7g–l). Through Western blotting, we further proved that MitoQ and AOX was able to inhibit catabolic processes and promote anabolic process of ECM, as well as inhibit cellular senescence of NDP52-KO chondrocytes ([227]Fig. 7m and n). Subsequent evaluation of mRNA level of ECM metabolism markers, senescence and inflammatory markers also yielded consistent findings ([228]Fig. 7o and p). Taken together, these findings suggested that RET and mtROS induced by NDP52 knockout contributed to chondrocytes degeneration, a process that could be mitigated by the administration of MitoQ and AOX. Fig. 7. [229]Fig. 7 [230]Open in a new tab MitoQ and AOX rescue NDP52-KO chondrocytes degeneration. (a and b) Micromass culture with alcian blue staining of NC and NDP52-KO C28/I2 chondrocytes treated with DMSO/MitoQ (a) or expressing AOX or not (b). (c and d) Cell survival curves of treated chondrocytes. n = 3 per group. (e and f) Flow cytometry of cell apoptosis of treated chondrocytes. (g and h) ELISA analysis of IL-6 of treated chondrocytes. n = 3 per group. (i, j, k and l) SA-β-gal staining (i and k) and analysis (j and l) of treated chondrocytes. n = 3 per group. Scale bar, 100 μm. (m and n) Immunoblot detection of NDP52, ACAN, MMP13, ADAMTS5, p16^INK4A, p21^CIP1 and GAPDH in treated chondrocytes. (o and p) The mRNA level of ACAN, MMP13, ADAMTS5, TNF-α, IL-1β, IL-6, IL-8, p16^INK4A and p21^CIP1 in treated chondrocytes. n = 3 per group. One-way analysis of variance (ANOVA) with a post-hoc Turkey test was used to analysis data, which was presented as the means ± SD. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ns p > 0.05. 3.8. AOX alleviate overload-induced OA progression in NDP52 knockout mice For AOX enzymes was absent in vertebrates, to further confirm mechanism of RET in vivo, AOX AAV was injected into articular joint of Ndp52- KO mice 2 weeks prior DMM surgery to ectopic express AOX ([231]Fig. 8a). The expression of AOX was detected by immunofluorescence analysis ([232]Fig. 8f and g). Notably, AOX AAV treatment partially mitigated cartilage damage observed 8 weeks post-DMM surgery ([233]Fig. 8b–e). Immunofluorescence analysis revealed a significant increase in the protein expression of Acan and a decrease in p16^Ink4a ([234]Fig. 8f–h and i). Moreover, AOX treatment not only reduced the formation of osteophytes in Ndp52-KO mice following DMM surgery ([235]Fig. 8j) but also improved pain sensitivity and mobility impairment ([236]Fig. 8k and l). In conclusion, our findings suggested that the knockout of Ndp52 induced cartilage degeneration in vivo through the RET pathway. Fig. 8. [237]Fig. 8 [238]Open in a new tab AOX rescues Ndp52-KO-induced OA in vivo. (a) Schematic diagram of experiment design to evaluate function of AOX in vivo. (b) The safranin O-fast green staining of joints of Ndp52-KO mice overexpressing AOX or not that underwent sham or DMM surgery. Scale bar, 50 μm. (c, d, and e) OARIS (c), osteophyte formation (d) and Mankin (e) score of joints of mice. n = 4 per group. (f, g, h and i) Immunofluorescence and analysis of Flag-AOX (g), Acan (h) and p16^Ink4a (i). n = 4 per group. Scale bar, 50 μm. (j) Reconstructed μCT image of mice. Arrows indicate osteophytes. (k and l) The hot plate (k) and rotarod (l) assays of the mice. n = 4 per group. One-way analysis of variance (ANOVA) with a post-hoc Turkey test was used to analysis data, which was presented as the means ± SD. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ns p > 0.05. 4. Discussion Our study indicated that NDP52 deficiency elevated the expression of MTIF3, and subsequently increased complex Ⅰ level. Furthermore, we demonstrated for the first time that NDP52 deficiency in chondrocytes driven RET in mitochondrial complex Ⅰ, thereby leading to pathological mtROS production, ultimately contributing to chondrocytes degeneration and progression of OA. While previous studies have indicated that some proteins are degraded by SARs in age-related OA [[239]55,[240]56], the role of SARs in chondrocytes remains unexplored. Our work demonstrated that NDP52 deficiency led to ECM degradation, inflammation, cell apoptosis and cellular senescence in chondrocytes, ultimately resulting in chondrocyte degeneration. Additionally, prior studies have shown the impact of NDP52 on various diseases such as Alzheimer's Disease [[241]6], Crohn's disease [[242]57], Type 2 Diabetes [[243]58] through its involvement in selective autophagy and regulation of tau protein degradation, TLR signaling activation and proinsulin maturation. NDP52 was found to operate via non-autophagic mechanisms, such as the degradation of hepatitis B virus through lysosome [[244]9] and the alteration of DNA conformation [[245]59] etc. In our study, the NDP52 LIR mut exhibited similar functions to the wild-type NDP52, suggesting that NDP52 may function in chondrocytes independent of autophagy. Through proteomic and validation, we confirmed that NDP52 deficiency could induce mtROS generation in chondrocytes, a previously unreported finding. Furthermore, our subsequent work demonstrated that mtROS played a center role in chondrocytes degeneration induced by NDP52 deficiency. Mitochondrial ROS has been identified as a contributing factor of mitophagy [[246]32], while NDP52, a mitophagy receptor, has been demonstrated to function as a sensor for ROS [[247]60]. Studies have indicated that disruption of mitophagy can lead to the long-term existence of damaged mitochondria and an increase in mtROS levels [[248]61], while excessive mitophagy can also induce elevated mtROS production [[249]62]. Nevertheless, the precise mechanism by which NDP52 deficiency induces mtROS generation in chondrocytes remains elusive. Previous study has identified NDP52 as a mitophagy receptor [[250]63]. While mitophagy mediated by NDP52 can occur independently of LC3 [[251]64], it has been confirmed that LIR mutation of NDP52 can hinder mitophagy [[252]40,[253]65] by potentially affecting autophagosome formation. Our findings indicated that the introduction of a mutation in the LIR domain of NDP52 can still rescue chondrocyte degeneration in a manner similar to wild-type NDP52. Moreover, mitophagy is typically induced in response to stress and various stimuli, and the deletion of receptors dose not impact mitophagy flux [[254]63]. Therefore, we propose that NDP52 may regulate mtROS generation via non-mitophagy pathway. As shown in previous research, RET is considered as a primary contributor to the substantial production of pathogenic mtROS generation [[255]66]. Our findings indicate that NDP52 deficiency led to an increase in mitochondrial complex Ⅰ levels, resulting in an elevation of the NAD^+/NADH rate, which creating conditions for RET similar to those from NDI1 overexpression [[256]51,[257]52,[258]67]. The reduction of mitochondrial membrane potential by DMM and AOX further support the hypothesis that elevated membrane potential promotes reverse electron transfer in Complex Ⅰ [[259]68]. The rescue of mtROS by rotenone, DMM, DPI and AOX further validated the existence of RET, aligning with prior studies on the subject [[260]30]. The studies demonstrated that RET is modulated by Notch via interaction with complex Ⅰ subunits [[261]51], by succinate excess leading to elevated mitochondrial membrane potential [[262]29] and by complex Ⅰ incorporation into supercomplexes [[263]26]. Our research confirmed that AOX functions as alternative electrons receptor, inhibiting RET and mtROS in vivo and in vitro. Furthermore, AOX was found to effectively rescue chondrocyte degeneration induced by NDP52 deficiency, suggesting its potential as a therapeutic intervention of OA. Nevertheless, the root causes of the increase in complex Ⅰ levels due to NDP52 knockdown had not been investigated in prior research. A previous study demonstrated that accuracy of mitochondrial translation initiation could impact the makeup of the electron chain [[264]54], with the knockout of MTIF3 specifically decreasing the levels of complex Ⅰ subunits. Our study delved into the elevation in MTIF3 levels following NDP52 knockout, providing an insight into the mechanism behind the observed increase in complex Ⅰ levels. Overall, we identified NDP52 as a regulator of MTIF3, with implications for the modulation of complex Ⅰ subunit levels. Deficiency of NDP52 was found to increase complex Ⅰ levels, promoting RET and consequently mtROS generation. The main limitation of this study was that the direct mechanism linking NDP52 and MTIF3 had not been explored. Secondly, the specific ways in which NDP52-mediated mitochondrial dysfunctions affected chondrocytes metabolism remained unclear. 5. Conclusion In this study, it was determined that NDP52, acting through MTIF3, modulated the expression of mitochondrial complex Ⅰ in chondrocytes in an autophagy-independent manner. NDP52 deficiency resulted in the upregulation of complex Ⅰ, leading to the initiation of RET and subsequent elevation of mtROS levels, ultimately contributing to chondrocyte degeneration. This study represents the initial exploration of the roles of NDP52 and RET in OA pathogenesis, offering novel perspectives and potential therapeutic targets for the treatment of this condition. CRediT authorship contribution statement Yutao Zhu: Writing – original draft, Visualization, Validation, Software, Methodology, Investigation, Data curation, Conceptualization. Yaohan Xu: Writing – original draft, Validation, Data curation. Dinqi Xie: Writing – original draft, Methodology. Nengfeng Yu: Validation, Methodology. Jiaxin Chen: Investigation, Funding acquisition. Jiechao Xia: Methodology, Formal analysis. Zixuan Mei: Software, Methodology. Yang Jin: Validation, Methodology. Chuan Hu: Writing – review & editing, Methodology. Pan Tang: Visualization, Methodology. Sicheng Jiang: Visualization, Methodology. Chao Jiang: Writing – review & editing, Supervision, Investigation. Honghai Song: Writing – review & editing, Methodology, Funding acquisition, Conceptualization. Zhijun Hu: Writing – review & editing, Supervision, Resources, Funding acquisition, Conceptualization. Data and materials availability All data needed to evaluate the conclusions in the paper are present in the paper and/or the Supplementary Materials. Funding This work was supported by Natural Science Foundation of Zhejiang Province (LY22H060003, LQ23H060008 and LQN25H250001), Innovative talent project in medical field of Zhejiang Province, Youth Fund of the National Natural Science Foundation of China (82402842). Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. Acknowledgments