Mechanical stress-mediated ferritinophagy aggravates cartilage endplate degeneration via a PIEZO1-NCOA4-ZBP1 axis Authors: Wenchao Wang, Dingfei Qian, Shengyao Sun, Yawei Yao, Xiaoqing Yang, Weibo Liu, Zelong Song, Haichao Yu, Linhao Hou, Peihong Hou, Jiayan Wu, Zhihao Ma, Jinlong Ma, Fanqi Hu, Hua Wang, Xuesong Zhang Published: 2026 ABSTRACT ABSTRACTAbnormal mechanical stress is closely linked to intervertebral disc degeneration (IVDD). Iron homeostasis disorder occurs in various degenerative diseases, including IVDD. PIEZO1 serves as a mechanosensitive cation channel, involving in multiple physiological and pathological processes; however, its potential association with iron homeostasis and IVDD remain to be elucidated. Here, it is discovered that PIEZO1 accumulates in cartilage endplate (CEP) during IVDD, accompanied by intensive ferritinophagy. Specific activation of PIEZO1 or NCOA4 (nuclear receptor coactivator 4) aggravates CEP degeneration. Conversely, chondrocyte-specific knockout of Piezo1 mitigates CEP degeneration by restoring the labile iron pool and stabilizing the mitochondrial genome. Mechanistically, PIEZO1-mediated nuclear translocation of YAP1 (Yes1 associated transcriptional regulator) enhances NCOA4-dependent ferritinophagy by promoting extracellular Ca2+ influx under oxidative stress. Moreover, ferritinophagy induces in the accumulation and release of Z-form mitochondrial DNA (Z-mtDNA), resulting in the activation of ZBP1 (Z-DNA binding protein 1), ultimately leading to NFKB-dependent inflammatory cascade. Therapeutically, blocking PIEZO1-mediated calcium influx or suppressing YAP1 activation alleviates ferritinophagy. Additionally, Ncoa4 silencing attenuates Z-mtDNA-ZBP1-NFKB axis-driven IVDD. Collectively, our findings suggest that mechanical overload induces ferritinophagy-dependent CEP degeneration via PIEZO1 activation and subsequent upregulation of the Z-mtDNA-ZBP1-NFKB axis, which might furnish a therapeutic target for IVDD.Abbreviation: AAV: adeno-associated virus; ALP: alkaline phosphatase; ARS: alizarin red S; BV:TV: bone volume:total volume; CAMK2/CaMKII: calcium/calmodulin dependent protein kinase II; CEP: cartilage endplate; CEPCs: cartilage endplate chondrocytes; ChIP: chromatin immunoprecipitation; CKO: conditional knockout; CsA: cyclosporin A; Co-IP: co-immunoprecipitation; DHI: disc height index; ECM: extracellular matrix; EtBr: ethidium bromide; HIF: hypoxia inducible factor; IVDD: intervertebral disc degeneration; KD: knockdown; LAT: large tumor suppressor kinase; LSI: lumbar spine instability; MDA: malondialdehyde; Mito-ROS: mitochondrial reactive oxygen species; MRI: magnetic resonance imaging; mtDNA: mitochondrial DNA; NCOA4: nuclear receptor coactivator 4; PCBP: poly(rC) binding protein; ROS: reactive oxygen species; RT-qPCR: real-time quantitative reverse transcription; SOFG: safranin O and fast green; WWTR1/TAZ: WW domain containing transcription regulator 1; TEAD1: TEA domain transcription factor 1; TEM: transmission electron microscopy; YAP1: Yes1 associated transcriptional regulator; ZBP1: Z-DNA binding protein 1; Z-DNA: Z-form DNA; Z-mtDNA: Z-form mitochondrial DNA. Introduction IVDD represents a major etiology of lower back pain in the elderly [1], contributing significantly to global disability and health care burdens. Risk factors associated with this pathology include aging, chronic inflammation, osteoporosis, mechanical stress and metabolic disorders [2–5]. Howerer, the precise molecular mechanisms driving its pathogenesis remain largely unclear. The CEP serves as an vital component of the intervertebral disc, which is indispensable in maintaining its structure and function integrity [4]. Meanwhile, spinal imbalance resulting from conditions such as degenerative scoliosis and degenerative lumbar spondylolisthesis leads to uneven mechanical stress distribution within the CEP, thereby accelerating IVDD. PIEZO1, a mechanosensitive ion channel, exerts a prominent role in diverse biological regulation pathways [6]. Emerging studies have illustrated that PIEZO1 is critically involved in bone metabolism, regulating processes ranging from bone formation [7–10] to bone resorption [11,12]. Moreover, it has been reported that chondrocytes, the only resident cells in cartilage tissue, are also modulated by PIEZO1 during the progression of osteoarthritis [13,14]. Notably, PIEZO1 activation promotes apoptosis in both annulus fibrosus cells [15] and nucleus pulposus cells [16], contributing to IVDD progression. Nevertheless, the mechanism by which PIEZO1 participates in the degeneration of CEP induced by mechanical loading remains to be elucidated. Iron stands as an essential trace element in humans, and sustaining cellular iron homeostasis is critical for numerous biological processes [17]. Ferroptosis, an iron-dependent form of regulated cell death [18], has evolved into a prospective target for age-related diseases [19–21]. Ferritinophagy, a selective autophagic process mediated by NCOA4 that targets ferritin for lysosomal degradation [22], has been implicated in the pathogenesis of osteoarthritis [23]. Our previous studies have found that iron overload stands out as an independent risk factor for IVDD [24]. Furthermore, relevant reports have confirmed that autophagy contributes to IVDD by promoting nucleus pulposus cell senescence [25,26]. Significantly, mechanical overloading induces GPX4-regulated knee chondrocytes ferroptosis via PIEZO1 channel [27], while gain-of-function PIEZO1 expression gives rise to iron overload in macrophages [28]. However, it remains unknown whether PIEZO1 participates in the development of IVDD by inducing NCOA4-mediated iron overload in cartilage endplate chondrocytes (CEPCs). YAP1 and WWTR1/TAZ (WW domain containing transcription regulator 1) operate as nuclear transducers of mechanical cues within the Hippo signaling [29]. They perceive extracellular matrix (ECM) stiffness and translate mechanical inputs into biological responses, thereby regulating diverse cellular behaviors such as proliferation and differentiation. Recent research has confirmed that loss of YAP1 initiates lumbar spine instability (LSI)-induced CEP remodeling by recruiting and promoting osteoclast differentiation [30]. Conversely, study also showed that PIEZO1 activation facilitates the osteogenic differentiation in aortic valve interstitial cells via YAP1-dependent glutaminolysis [31]. Crucially, YAP1 also exhibits distinct roles in modulating ferroptosis. Some studies indicate that YAP1 enhances sensitivity to ferroptosis [32], whereas others suggest that YAP1 activation promotes ferroptosis resistance [33]. Hence, further exploration is needed to decipher the precise mechanism governing YAP1’s regulation of PIEZO1-mediated ferroptosis. ZBP1 as a sensor of endogenous nucleic acids such as Z-mtDNA [34,35], is involved in inflammatory multiple responses [36–40]. Reactive oxygen species (ROS) originating from iron metabolic imbalance in CEPCs promotes mitochondrial DNA (mtDNA) release, thereby exacerbating IVDD, as previously demonstrated by our team [41]. Moreover, the increase in mtDNA release was observed during PIEZO1-mediated osteoarthritis progression [42]. However, it remains unclear whether ZBP1 is implicated in CEP degeneration triggered by mechanical stress. In this study, we proposed the new pathogenic mechanisms that underlie CEP degeneration caused by abnormal mechanical tension. Firstly, the upregulation of PIEZO1 activates ferritinophagy in CEPCs through the YAP1-WWTR1 pathway. Secondly, iron homeostasis disruption in CEPCs promotes the efflux of Z-mtDNA, which activates the ZBP1-driven inflammatory cascade and aggravates IVDD. Finally, we suggest that targeting the PIEZO1-YAP1-NCOA4-ZBP1-NFKB signaling axis may offer a new therapeutic strategy for IVDD. Results Mechanical stress promotes endplate cartilage iron homeostasis disorder and IVDD To unravel mechanical stress-induced IVDD pathogenesis, the convex and concave CEP samples were collected from degenerative scoliosis patients with endplate osteochondritis (Figure 1(A) and S1A-B). As shown in Figure 1(B), the endplate cartilage of concave zone was characterized by vacated chondrocyte lacuna and reduced COL2, accompanied with increased iron deposition. In addition, LSI mice model was established via simulating unbalanced mechanical loading [43]. Strikingly, immunostaining analysis showed that the CEP after LSI surgery underwent consistent degeneration, featured by ossification enhancement, ECM decline and iron overload (Figure 1(C)). Moreover, compared with the convex CEP, the Fe2+ content in the concave CEP is significantly increased, while the Fe3+ content is decreased (Figure 1(D)). Motivated by these findings, we aimed to ascertain the potential relationship between mechanical stress and the iron homeostasis. Therefore, we exposed isolated CEPCs to the mechanical stress environment of 1 Mpa. Microarray (Figure 1(E) and western blot (Figure 1(F) and S1C) analyses revealed that mechanical strain induced changes in the expression of cellular iron metabolism genes (Tfr1, Steap3, Dmt1, Fpn, Pcbp1, and Pcbp2), ferritinophagy markers (Ncoa4 and Fth), ferroptosis markers (Gpx4, Slc7a11, Acsl4, and Fsp1). Meanwhile, HAMP-the key regulator of SLC40A1/ferroportin-also showed altered expression (Figure S1C). As shown in Figure S1D, mechanical stimulation also modulated the expression of metabolic and inflammatory markers. Furthermore, mechanical stress significantly increased intracellular malondialdehyde (MDA) and Fe2+ levels (Figure 1(G) and (H)). Since ferroptosis is driven by iron-dependent lipid peroxidation with MDA is its terminal product, C11-BODIPY staining was used to further confirm the increase in lipid peroxidation under mechanical stimuli (Figure 1(I)). Additionally, transmission electron microscopy (TEM) demonstrated that 1 MPa stimulation promoted CEP chondrocytes ferroptosis, manifested by mitochondrial cristae disorganization and iron ion deposition (Figure 1(J)). Therefore, DCFH-DA, JC-1 and MitoTracker Green assay were further utilized to evaluate mitochondrial function. The results showed that mechanical stress significantly aggravated ROS production (Figure 1(K)), accompanied by the reduction in mitochondrial membrane potential (Figure 1(L)) and the destruction of mitochondrial length (Figure 1M). However, DFO (the iron chelator) or Fer-1 (the ferroptosis inhibitor) intervention partially restored mitochondrial homeostasis. To determine whether iron overload is involved in IVDD due to mechanical stress, LSI mice were intraperitoneally injected with DFO and Fer-1. Not surprisingly, the Safranin O and fast green (SOFG) assay showed that DFO or Fer-1 administration significantly inhibited the CEP degeneration, with a corresponding decrease in histological score (Figure 1N and O). Consistently, the downregulation of COL2 in CEP of LSI mice were markedly attenuated by DFO or Fer-1 treatment (Figure 1P and Q). Together, these data demonstrate that iron homeostasis disorder, mediated by mechanical stress, participates in the pathological progression of IVDD. Figure 1.Mechanical stress disrupts iron homeostasis and promotes IVDD. (A) X-ray from patients with degenerative scoliosis. (B) Left, representative HE, COL2 and Perls staining of convex and concave CEP in degenerative scoliosis patients with endplate osteochondritis. Right, quantification (n = 8). (C) Left, representative HE, COL2 and Perls staining images of the CEP in mice with LSI or sham surgery for 3 months. Arrows indicate the osteoid area within the CEP, along with the surrounding degradation of COL2 and accumulation of hemosiderin. Right, quantification (n = 8). (D) Quantitative analysis of iron ion content in convex and concave human CEP (n = 6). (E-M) Exposure of CEPCs to 1 MPa mechanical stimulation for 24 h was followed by corresponding analyses. (E) Heatmap illustrating the differential gene expression in CEPCs. (F) Western blot analysis of iron metabolic and ferroptotic markers. Right, quantification (n = 3). (G) Quantitative analysis of MDA content of CEPCs (n = 6). (H) Detection of intracellular Fe2+ in CEPCs using FerroOrange (n = 3). (I) Lipid ROS in CEPCs using C11-BODIPY (n = 3). (J) Representative TEM images of CEPCs. Arrows indicate shrunken mitochondria. (K-M) Under 1 MPa mechanical stress conditions, CEPCs were treated with DFO (100 μM) or Fer-1 (1 μM) for 24 h. (K) Detection of intracellular ROS in CEPCs using DCFH-DA (n = 3). (L) Assessment of mitochondrial membrane potential in CEPCs using JC-1 staining (n = 3). JC-1 monomers (green, depolarized) and aggregates (red, polarized) are shown. (M) Representative changes in mitochondrial morphology stained with Mito-Tracker green (n = 3). (N) Representative SOFG staining of the CEP in LSI mice following a 3-month treatment with intraperitoneal DFO (100 mg/kg/w) or Fer-1 (5 mg/kg/w) (n = 8). (O) Endplate scores of the CEP. (P and Q) Representative immunofluorescent analysis of the COL2 in CEP (n = 4).The composite image illustrates the impact of mechanical stress on iron homeostasis and intervertebral disc degeneration. It includes various analyses and visualizations. An X-ray highlights convex and concave regions. Staining images show differences in HE, COL2 and Perls staining between convex and concave regions, with histological scores and relative staining areas quantified. Sham and LSI treatments are compared using similar staining methods. A graph displays relative iron content, comparing Fe superscript 2 plus and Fe superscript 3 plus in convex and concave regions. A heatmap shows differential gene expression under control and 1 Mpa conditions. Western blot analysis presents iron metabolic and ferroptotic markers, with quantification graphs. A graph illustrates relative MDA content under control and 1 Mpa conditions. Fluorescence images using FerroOrange and Hoechst and C11-BODIPY, show intracellular changes, with quantification graphs. TEM images reveal mitochondrial changes under control and 1 Mpa conditions. Fluorescence images using DCFH-DA and JC-1 staining indicate changes in mitochondrial membrane potential and ROS production, with quantification graphs. MitoTracker Green staining shows mitochondrial morphology changes. SOFG staining compares sham, LSI, LSI plus DFO and LSI plus Fer-1 treatments. Endplate scores and COL2 staining images are provided for different treatments, with quantification graphs showing COL2-positive area percentages. The image conveys the detrimental effects of mechanical stress on cellular structures and functions, emphasizing the role of iron homeostasis in degeneration processes.Composite image showing effects of mechanical stress on iron homeostasis and intervertebral disc degeneration. Mechanical stress disrupts iron homeostasis and promotes IVDD. (A) X-ray from patients with degenerative scoliosis. (B) Left, representative HE, COL2 and Perls staining of convex and concave CEP in degenerative scoliosis patients with endplate osteochondritis. Right, quantification (n = 8). (C) Left, representative HE, COL2 and Perls staining images of the CEP in mice with LSI or sham surgery for 3 months. Arrows indicate the osteoid area within the CEP, along with the surrounding degradation of COL2 and accumulation of hemosiderin. Right, quantification (n = 8). (D) Quantitative analysis of iron ion content in convex and concave human CEP (n = 6). (E-M) Exposure of CEPCs to 1 MPa mechanical stimulation for 24 h was followed by corresponding analyses. (E) Heatmap illustrating the differential gene expression in CEPCs. (F) Western blot analysis of iron metabolic and ferroptotic markers. Right, quantification (n = 3). (G) Quantitative analysis of MDA content of CEPCs (n = 6). (H) Detection of intracellular Fe2+ in CEPCs using FerroOrange (n = 3). (I) Lipid ROS in CEPCs using C11-BODIPY (n = 3). (J) Representative TEM images of CEPCs. Arrows indicate shrunken mitochondria. (K-M) Under 1 MPa mechanical stress conditions, CEPCs were treated with DFO (100 μM) or Fer-1 (1 μM) for 24 h. (K) Detection of intracellular ROS in CEPCs using DCFH-DA (n = 3). (L) Assessment of mitochondrial membrane potential in CEPCs using JC-1 staining (n = 3). JC-1 monomers (green, depolarized) and aggregates (red, polarized) are shown. (M) Representative changes in mitochondrial morphology stained with Mito-Tracker green (n = 3). (N) Representative SOFG staining of the CEP in LSI mice following a 3-month treatment with intraperitoneal DFO (100 mg/kg/w) or Fer-1 (5 mg/kg/w) (n = 8). (O) Endplate scores of the CEP. (P and Q) Representative immunofluorescent analysis of the COL2 in CEP (n = 4). PIEZO1 mediates mechanical stress-induced ferritinophagy in endplate cartilage PIEZO1, a mechanosensitive ion channel, detects and reacts to mechanical stress. Based on the results shown in Figure 1(E), mechanical stress altered the expression of PIEZO1 and ferritinophagy-related markers. Immunohistochemistry analysis showed that mechanical overloading was found to activate PIEZO1, concomitant with the upregulation of NCOA4 and downregulation of FTH (Figure 2(A) and (B)). To investigate the mechanism through which PIEZO1 induces iron homeostasis imbalance under oxidative stress, CEPCs were treated with H2O2 (the inflammatory stimuli), Yoda1 (the PIEZO1 agonist) and GsMTx4 (the PIEZO1 inhibitor). Western blot analysis revealed that Yoda1 treatment downregulated the expression of FTH and FPN, while upregulating NCOA4 and MAP1LC3/LC3-II:I levels compared to GsMTx4 treatment (Figure 2(C) and D). Notably, Yoda1 inhibited the expression of PCBP1 and PCBP2, which function as chaperones that deliver iron to ferritin, thereby exacerbating the depletion of FTH (Figure S2A and S2B). Interestingly, Yoda1 reduced HAMP levels, thereby promoting iron efflux (Figure S2A and S2B). Next, immunofluorescence analysis confirmed that Yoda1 enhanced NCOA4 expression, whereas GsMTx4 suppressed it (Figure 2(E)). As a cargo receptor in ferritinophagy, NCOA4 orchestrates the autophagic degradation of ferritin and serves a crucial role in iron homeostasis. Therefore, we sought to determine whether NCOA4-mediated ferritinophagy contributes to PIEZO1-induced iron overload under oxidative stress. Interestingly, Yoda1 treatment led to a significant accumulation of Fe2+ within lysosomes (colocalization of FerroOrange and LysoTracker). Conversely, intervention with GsMTx4 resulted in a noticeable reduction in lysosomal Fe2+ levels (Figure 2(G) and (H)). Immunofluorescence analysis also revealed an increased colocalization of endogenous FTH and LC3 in Yoda1-treated CEPCs, which was diminished by GsMTx4 treatment (Figure 2I and J). Next, we set out to examine the regulatory role of PIEZO1 in ferritinophagy in vivo. Consistently, in the LSI model, intraperitoneal administration of Yoda1 upregulated NCOA4 expression and downregulated FTH expression in CEP, whereas GsMTx4 treatment produced the opposite effect (Figure 2K-N and S2I-J). Furthermore, Yoda1 treatment concomitantly upregulated both PIEZO1 and p-CAMK2 (phosphorylation of Calcium/calmodulin dependent protein kinase II), which consequently were inhibited by either calcium deprivation or CAMK2 inhibitor KN-93 (Figure S2C-H). These findings demonstrate that mechanical stress activates PIEZO1 through a Ca2+-CaM-CAMK2 axis-mediated transcriptional positive feedback loop, which in turn promotes ferritinophagy and ultimately leads to iron overload in the endplate cartilage. Figure 2.PIEZO1 activation promotes ferritinophagy in the endplate cartilage under mechanical stress. (A) Left, representative PIEZO1, NCOA4 and FTH immunohistochemistry staining of convex and concave CEP in degenerative scoliosis patients with endplate osteochondritis. Right, quantification (n = 8). (B) Left, representative PIEZO1, NCOA4 and FTH staining images of the CEP in mice with LSI or sham surgery for 3 months. Right, quantification (n = 8). (C and D) Western blot analysis of iron metabolic, ferritinophagy and ferroptotic markers in CEPCs treated with H2O2 (100 μM), Yoda1 (1, 5, 10 μM) or GsMTx4 (0.1, 0.5, 1 μM) for 24 h (n = 3). (E-J) CEPCs were treated for 24 h with H2O2 (100 μM), Yoda1 (10 μM), or GsMTx4 (1 μM). (E and F) Representative immunofluorescent analysis of NCOA4 in CEPCs (n = 3). (G and H) Fluorescence colocalization analysis of FerroOrange and LysoTracker (n = 3). (I and J) Fluorescence colocalization analysis of FTH and LC3 (n = 3). (K-N) LSI mice were intraperitoneal treated with Yoda1 (3 mg/kg/w) or GsMTx4 (1 mg/kg/w) for 3 months. (K and L) Representative immunohistochemistry analysis of the NCOA4 in CEP (n = 4). (M and N) Representative immunofluorescent analysis of the FTH in CEP (n = 4).The composite image consists of multiple panels showing various analyses related to PIEZO1, NCOA4 and FTH. Image A shows immunohistochemistry staining of PIEZO1, NCOA4 and FTH in convex and concave regions, with a graph quantifying positive area percentages. Image B displays similar staining in sham and LSI conditions, with quantification. Image C presents western blot analysis of iron metabolic markers under different treatments, including H subscript 2 O subscript 2, Yoda1 and GsMTx4. Image D shows graphs of protein expression levels for various markers, including TFR1, NCOA4, DMT1, FPN, SLC7A11, GPX4, FTH and LC3, under these treatments. Image E provides immunofluorescent analysis of NCOA4, with Image F showing relative fluorescence quantification. Image G illustrates fluorescence colocalization analysis of FerroOrange and LysoTracker, with Image H showing Pearson's coefficient. Image I presents colocalization analysis of FTH and LC3, with Image J showing Pearson's coefficient. Image K shows immunohistochemistry analysis of NCOA4 in sham, LSI, LSI plus Yoda1 and LSI plus GsMTx4 conditions, with Image L quantifying positive area percentages. Image M displays immunofluorescent analysis of FTH in similar conditions, with Image N quantifying positive area percentages.Image: PIEZO1, NCOA4, FTH analysis via immunohistochemistry, western blot, fluorescence under treatments. PIEZO1 activation promotes ferritinophagy in the endplate cartilage under mechanical stress. (A) Left, representative PIEZO1, NCOA4 and FTH immunohistochemistry staining of convex and concave CEP in degenerative scoliosis patients with endplate osteochondritis. Right, quantification (n = 8). (B) Left, representative PIEZO1, NCOA4 and FTH staining images of the CEP in mice with LSI or sham surgery for 3 months. Right, quantification (n = 8). (C and D) Western blot analysis of iron metabolic, ferritinophagy and ferroptotic markers in CEPCs treated with H2O2 (100 μM), Yoda1 (1, 5, 10 μM) or GsMTx4 (0.1, 0.5, 1 μM) for 24 h (n = 3). (E-J) CEPCs were treated for 24 h with H2O2 (100 μM), Yoda1 (10 μM), or GsMTx4 (1 μM). (E and F) Representative immunofluorescent analysis of NCOA4 in CEPCs (n = 3). (G and H) Fluorescence colocalization analysis of FerroOrange and LysoTracker (n = 3). (I and J) Fluorescence colocalization analysis of FTH and LC3 (n = 3). (K-N) LSI mice were intraperitoneal treated with Yoda1 (3 mg/kg/w) or GsMTx4 (1 mg/kg/w) for 3 months. (K and L) Representative immunohistochemistry analysis of the NCOA4 in CEP (n = 4). (M and N) Representative immunofluorescent analysis of the FTH in CEP (n = 4). Autophagy flux inhibition alleviates PIEZO1 mediated IVDD As shown in Figures S3A and S3B, Yoda1 treatment upregulated ATG7 and ATG5 expression and promoted SQSTM1/p62 degradation compared with GsMTx4. To further determine whether autophagy levels influence ferritinophagy in CECPs and contribute to CEP degeneration, 3-MA (the autophagy inhibitor) and BafA1 (the autophagosome-lysosome conjugation inhibitor) were employed. Figures S3C and S3D showed that inhibition of autophagy, reflected by decreased ATG7, ATG5, and LC3-II:I and elevated SQSTM1 in response to 3-MA and BafA1, dampened the Yoda1-mediated increase in NCOA4. Besides, immunofluorescence analysis revealed that both 3-MA and BafA1 suppressed the colocalization of FTH with LC3 and of LC3 with LAMP1 (Figure 3(A) and S3E), resulting in the restoration of FTH levels (Figure 3(C,D)). Furthermore, LSI mice received intraperitoneal injections of 3-MA (10 mg/kg/w) or BafA1 (1 mg/kg/w) for 3 months. Immunohistochemical and western blot analysis confirmed that these inhibitors also reverse FTH expression in CEP (Figure 3(B) and S3F-G). Next, we verified the effect of autophagy on ECM. Western blot analysis exposed that autophagy inhibition restored the expression of the anabolic marker (COL2), while inhibiting catabolic markers (MMP13 and MMP3) as well as markers associated with endochondral ossification (COL10A1) (Figure 3(C),(D)). Alkaline phosphatase (ALP) and alizarin red S (ARS) staining assays were utilized to investigate the role of autophagy in the mineralization process of CEPCs. As illustrated in Figure 3(E), PIEZO1 activation enhanced calcium deposition and increased ALP activity. Conversely, autophagy inhibition attenuated these effects. To further assess the role of autophagy in CEP calcification and degeneration, Micro-CT, histological, and fluorescence analyses were performed on LSI mice model. As displayed in Figure 3(F), the PIEZO1-mediated reduction in disc height index (DHI) and increase in Bone volume:Total volume (BV:TV) of CEP were partially reversed by autophagy inhibition. Moreover, autophagy ablation attenuated the PIEZO1-induced replacement of endplate cartilage by bone tissue (Figure 3G and H), accompanied by elevated COL2 expression and reduced COL10A1 expression (Figure 3Iand J). Figure 3.Autophagy flux inhibition alleviates PIEZO1 mediated IVDD. (A, C, and D) CEPCs were treated with H2O2 (100 μM), Yoda1 (10 μM), 3-MA (5 mM) or BafA1 (100 nM) for 24 h. (A) Left, representative fluorescence colocalization analysis of FTH and LC3 in CEPCs. Right, quantification (n = 3). (B) Left, representative immunofluorescent analysis of FTH in CEP from the LSI model. Right, quantification (n = 4). (C and D) Representative western blot analysis of COL2, COL10A1, MMP13, MMP3 and FTH in CEPCs (n = 3). (E) Left, representative ALP and ARS staining images. Right, quantification (n = 3). (F) Bottom, representative 3D micro-CT images. Top, quantitative analysis of DHI and BV:TV (n = 3). (G and H) Representative SOFG staining images and endplate scores of the CEP (n = 8). (I and J) Representative immunofluorescent analysis of the COL2 and COL10A1 in CEP (n = 4).The image contains seven panels. The first panel shows fluorescence colocalization analysis of FTH and LC3 in CEPCs with different treatments, including control, hydrogen peroxide, Yoda1, 3-MA and BafA1. The second panel presents immunofluorescent analysis of FTH in CEP from the LSI model with various treatments. The third panel displays a western blot analysis of COL2, COL10A1, MMP13, MMP3 and FTH in CEPCs under different conditions. The fourth panel includes ALP and ARS staining images with quantification. The fifth panel shows 3D micro-CT images with quantitative analysis of disc height index and bone volume to total volume. The sixth panel presents SOFG staining images and endplate scores of the CEP. The seventh panel displays immunofluorescent analysis of COL2 and COL10A1 in CEP with different treatments. Each panel includes quantification graphs or data points for comparison.Seven panels: fluorescence, immunofluorescence, western blot, staining, micro-CT, histology, immunofluorescence. Autophagy flux inhibition alleviates PIEZO1 mediated IVDD. (A, C, and D) CEPCs were treated with H2O2 (100 μM), Yoda1 (10 μM), 3-MA (5 mM) or BafA1 (100 nM) for 24 h. (A) Left, representative fluorescence colocalization analysis of FTH and LC3 in CEPCs. Right, quantification (n = 3). (B) Left, representative immunofluorescent analysis of FTH in CEP from the LSI model. Right, quantification (n = 4). (C and D) Representative western blot analysis of COL2, COL10A1, MMP13, MMP3 and FTH in CEPCs (n = 3). (E) Left, representative ALP and ARS staining images. Right, quantification (n = 3). (F) Bottom, representative 3D micro-CT images. Top, quantitative analysis of DHI and BV:TV (n = 3). (G and H) Representative SOFG staining images and endplate scores of the CEP (n = 8). (I and J) Representative immunofluorescent analysis of the COL2 and COL10A1 in CEP (n = 4). Chondrocyte-specific deletion of Piezo1 alleviates mechanical stress-induced ferritinophagy and IVDD Building on the findings that PIEZO1 regulates ferritinophagy under oxidative stress and that suppression of autophagy mitigates IVDD, we further investigated whether chondrocyte-specific deletion of Piezo1 alleviates IVDD through NCOA4. CEPCs were isolated from piezo1fl/fl Col2 Cre+ (piezo1 conditional knockout, piezo1 CKO) and Piezo1fl/f Col2 Cre− (control) mice. The knockout efficiency and specificity of Piezo1 was verified by immunohistochemistry and western blot (Figure S4A-F). Next, immunofluorescence and western blot analysis revealed that, under 1 Mpa mechanical stress, piezo1 CKO suppressed the upregulation of NCOA4 (Figure 4A, C and S4I-J), leading to the restoration of intracellular Fe2+ (Figure 4(B, D) and S4G-H) and FTH (Figure S4I and S4J) levels. However, piezo1 CKO had no effect on TFR1, DMT1, or FPN (Figure S4I and S4J). Moreover, immunofluorescence assay indicated that piezo1 knockout also reduced lysosomal Fe2+ levels (Figure 4E and F) and colocalization of LC3 with LAMP1 (Figure S4K). Notably, co-immunoprecipitation (co-IP) demonstrated that piezo1 CKO weakened the interaction between NCOA4 and FTH (Figure S4L). The present findings indicate that genetic inhibition of Piezo1 attenuates NCOA4-dependent ferritinophagy. Next, we further examined the impact of piezo1 CKO on the cartilaginous phenotype. ALP and ARS staining assays showed that the piezo1 CKO attenuated the Yoda1-induced osteogenic effect on CEPCs under inflammatory conditions (Figure 4(G)). Meanwhile, immunofluorescence analysis confirmed that piezo1 CKO alleviated the mechanical stress-induced decrease in COL2 and increase in COL10A1 in vitro (Figure 4H and I). Subsequently, we performed in LSI model to evaluate the effects of piezo1 CKO on ferritinophagy and CEP degeneration. Consistently, Piezo1 deletion attenuated the mechanical stress-induced upregulation of NCOA4, degradation of FTH, and abnormal iron deposition in vivo (Figure 4(J) and S4M-N). Importantly, the deletion of Piezo1 promoted the phenotypic recovery of endplate cartilage in LSI mice, as evidenced by the decreased endplate score, elevated COL2 expression, and reduced COL10A1 level (Figure 4K and O-Q). Figure 4.A piezo1 CKO alleviates mechanical stress-induced ferritinophagy and IVDD. (A-F and H-I) CEPCs were treated with 1 Mpa mechanical stress for 24 h. (A and C) Representative fluorescence analysis of NCOA4 in CEPCs (n = 3). (B and D) Representative fluorescence analysis of FerroOrange in CEPCs (n = 3). (E and F) Fluorescence colocalization analysis of FerroOrange and lysosomes in CEPCs (n = 3). (G) Left, representative ALP and ARS staining images. Right, quantification (n = 3). (H and I) Fluorescence analysis of COL2, COL10A1 and F-actin (n = 3). (J and L-N) Representative histological staining analysis of NCOA4, FTH and Perls (n = 4). (K and O-Q) Representative histological staining analysis of SOFG, COL2 and COL10A1 (n = 4).The image A shows fluorescence analysis of NCOA4 in Piezo1 superscript fl/fl Col2 Cre superscript + and Piezo1 superscript fl/fl Col2 Cre superscript − under control and 1 Mpa conditions. The image B shows fluorescence analysis of FerroOrange in similar conditions. The image C presents a bar graph of relative fluorescence of NCOA4 with Piezo1 superscript fl/fl Col2 Cre superscript + and Piezo1 superscript fl/fl Col2 Cre superscript − under control and 1 Mpa conditions. The image D shows a bar graph of relative fluorescence intensity of FerroOrange under similar conditions. The image E displays fluorescence colocalization analysis of FerroOrange and lysosomes in Piezo1 superscript fl/fl Col2 Cre superscript + and Piezo1 superscript fl/fl Col2 Cre superscript − under control and 1 Mpa conditions. The image F shows a bar graph of Pearson's coefficient for colocalization. The image G presents ALP and ARS staining images and quantification under different conditions. The image H shows fluorescence analysis of COL2, COL10A1 and F-actin in Piezo1 superscript fl/fl Col2 Cre superscript + and Piezo1 superscript fl/fl Col2 Cre superscript − under control and 1 Mpa conditions. The image I presents a bar graph of relative fluorescence density of COL2 and COL10A1. The image J shows histological staining analysis of NCOA4, FTH and Perls in Piezo1 superscript fl/fl Col2 Cre superscript + and Piezo1 superscript fl/fl Col2 Cre superscript − under control and LSI conditions. The image K displays histological staining analysis of SOFG, COL2 and COL10A1 under similar conditions. The image L, M, N, O, P and Q show graphs of positive area percentages and endplate scores under different conditions.Nine panels showing fluorescence and histological analyses of Piezo1 CKO effects on CEPCs under mechanical stress. A piezo1 CKO alleviates mechanical stress-induced ferritinophagy and IVDD. (A-F and H-I) CEPCs were treated with 1 Mpa mechanical stress for 24 h. (A and C) Representative fluorescence analysis of NCOA4 in CEPCs (n = 3). (B and D) Representative fluorescence analysis of FerroOrange in CEPCs (n = 3). (E and F) Fluorescence colocalization analysis of FerroOrange and lysosomes in CEPCs (n = 3). (G) Left, representative ALP and ARS staining images. Right, quantification (n = 3). (H and I) Fluorescence analysis of COL2, COL10A1 and F-actin (n = 3). (J and L-N) Representative histological staining analysis of NCOA4, FTH and Perls (n = 4). (K and O-Q) Representative histological staining analysis of SOFG, COL2 and COL10A1 (n = 4). Ferritinophagy exacerbates PIEZO1-induced IVDD by enhancing oxidative stress damage Based on previous studies, we further investigated the role of ferritinophagy in endplate cartilage degeneration by directly modulating Ncoa4 expression via knockdown and overexpression approaches. To functionally validate the role of Ncoa4, siRNA-mediated knockdown was initially employed. Western blot and immunofluorescence analyses confirmed that si-Ncoa4 intervention effectively suppressed its expression (Figure S5A), leading to upregulated FTH expression (Figure 5(A-B) and S5B) and reduced intracellular Fe2+ levels (Figures 5(C, F) and S5C). C11-BODIPY staining and MDA content analyses demonstrated that under inflammatory conditions, si-Ncoa4 effectively attenuated the Yoda1-induced increase in lipid peroxidation (Figure 5(D-E), Iand S5D). Moreover, si-Ncoa4 treatment also decreased mitochondrial reactive oxygen species (Mito-ROS) levels (Figure 5(G-H) and S5E) and facilitated the recovery of mitochondrial cristae structure (Figure 5(J) and S5F). The above results demonstrated that inhibition of Ncoa4 attenuates ferritinophagy-mediated mitochondrial oxidative stress damage. Subsequently, in the Tail-looping model, we performed in vivo functional validation of Ncoa4 using adeno-associated virus (AAV)-mediated knockdown and overexpression. As shown in Figure S5G, AAV-mediated Ncoa4 knockdown (AAV-Ncoa4 KD) was applied to both normal and symmetrically bent caudal intervertebral spaces, with AAV-GFP serving as the control. Micro-CT imaging revealed abnormal bone formation at the concave edge, which was attenuated by AAV-Ncoa4 KD intervention (Figure 5(K)). SOFG staining indicated that bending stress induced degeneration of the concave CEP, evidenced by heterotopic ossification formation, compared to the convex side. AAV-Ncoa4 KD attenuated this degenerative response, thereby reducing the CEP score (Figure 5L and M). Furthermore, Piezo1 deletion partially alleviated IVDD induced by flexion stress. The protective effect of piezo1 CKO on the concave CEP was characterized by reduced abnormal bone formation (Figure 5N and O), lower degeneration scores (Figure 5(P)), increased COL2 expression, and decreased COL10A1 expression (Figure 5Q and R). However, AAV-mediated overexpression of Ncoa4 (AAV-Ncoa4 OE) abolished this improvement (Figure 5(N-R)). Figure 5.Ferritinophagy exacerbates PIEZO1-induced IVDD by enhancing oxidative stress damage. (A-J) CEPCs were transfected with si-Ncoa4 or si-NC for 72 h, followed by treatment with H2O2 (100 μM) or Yoda1 (10 μM) for 24 h. (A and B) Western blot analysis of NCOA4 and FTH expression in CEPCs (n = 3). (C and F) Representative immunofluorescence analysis of FerroOrange in CEPCs (n = 3). (D and E) Lipid ROS in CEPCs using C11-BODIPY (n = 3). (G and H) Mito-ROS in CEPCs using MitoSox and MitoTracker (n = 3). (I) MDA content of CEPCs (n = 3). (J) Representative TEM images of CEPCs. Arrows indicate mitochondria. (K) Representative micro-CT images of the Tail-looping model following 8 weeks of intervention with AAV-GFP or AAV-Ncoa4 KD. (L and M) Representative SOFG analysis of the CEP (n = 8). (N and O) Representative micro-CT images of the Tail-looping model following 8 weeks of intervention with AAV-GFP or AAV-Ncoa4 OE. Abnormal ossifications were observed in the areas marked by arrows (K, L and O). (P-R) Representative histological staining analysis of SOFG (n = 8), COL2 (n = 4) and COL10A1 (n = 4) in CEP.The image provides an analysis of CEPCs under different treatments, featuring a western blot for NCOA4 and FTH proteins with treatments like si-NC, H2O2 and Yoda1. Immunofluorescence images use FerroOrange to highlight CEPCs. Lipid ROS is examined with C11-BODIPY, shown in fluorescence images and bar graphs for oxidized and non-oxidized lipid levels. Mito-ROS is evaluated using MitoSox and MitoTracker, with related images and graphs. TEM images display mitochondrial structures. Micro-CT images of the Tail-looping model show AAV-GFP and AAV-Ncoa4 KD interventions, indicating bone formation changes. SOFG staining reveals histological differences in the CEP under various conditions. Additional micro-CT images and SOFG, COL2 and COL10A1 staining are shown for convex and concave regions, with bar graphs for COL2 and COL10A1 expression. The panels examine the effects of treatments on oxidative stress and cellular changes in CEPCs.CEPCs analysis: protein expression, oxidative stress and histological changes studied. Ferritinophagy exacerbates PIEZO1-induced IVDD by enhancing oxidative stress damage. (A-J) CEPCs were transfected with si-Ncoa4 or si-NC for 72 h, followed by treatment with H2O2 (100 μM) or Yoda1 (10 μM) for 24 h. (A and B) Western blot analysis of NCOA4 and FTH expression in CEPCs (n = 3). (C and F) Representative immunofluorescence analysis of FerroOrange in CEPCs (n = 3). (D and E) Lipid ROS in CEPCs using C11-BODIPY (n = 3). (G and H) Mito-ROS in CEPCs using MitoSox and MitoTracker (n = 3). (I) MDA content of CEPCs (n = 3). (J) Representative TEM images of CEPCs. Arrows indicate mitochondria. (K) Representative micro-CT images of the Tail-looping model following 8 weeks of intervention with AAV-GFP or AAV-Ncoa4 KD. (L and M) Representative SOFG analysis of the CEP (n = 8). (N and O) Representative micro-CT images of the Tail-looping model following 8 weeks of intervention with AAV-GFP or AAV-Ncoa4 OE. Abnormal ossifications were observed in the areas marked by arrows (K, L and O). (P-R) Representative histological staining analysis of SOFG (n = 8), COL2 (n = 4) and COL10A1 (n = 4) in CEP. YAP1, as a transcription factor, is involved in PIEZO1-mediated ferritinophagy To determine whether YAP1, a mechanosensitive transcription factor, is involved in PIEZO1-mediated ferritinophagy, CEP samples with osteochondritis were collected from degenerative scoliosis patients. Immunohistochemical analysis showed a significant upregulation of nuclear YAP1 expression in the concave side of CEP (Figure S6A). Immunofluorescence analysis indicated that under inflammatory conditions, Yoda1 treatment markedly enhanced YAP1 nuclear translocation, whereas GsMTx4 significantly suppressed it (Figure 6A and B). Besides, western blot and RT-qPCR analysis showed that the activation of PIEZO1 promoted the transcriptional function of Yap1, as reflected by elevated levels of YAP1:p-YAP1, TEAD1 (transcriptional enhanced associate domain 1), WWTR1, and downstream target genes (CYR61, CTGF, and ANKRD1). Conversely, GsMTx4 treatment reversed these effects (Figure 6C and S6C). In addition, piezo1 CKO inhibited the activation of the YAP1-TEAD1 axis in the LSI model (Figure 6E-F and S6B). Crucially, chromatin immunoprecipitation (ChIP)-qPCR assay revealed that TEAD1 bound significantly to the Ncoa4 promoter at the Primer1 site (Figure S6D and S6E). The above results suggest that PIEZO1 positively regulates the YAP1-WWTR1-TEAD1 axis. We therefore sought to determine whether YAP1 also participates in PIEZO1-mediated ferritinophagy. As demonstrated in Figure 6G-J and S6H-K, siRNA-mediated knockdown of Yap1 (si-Yap1) (Figure S6F and S6G) suppressed the upregulation of NCOA4 induced by H2O2 and Yoda1 treatment, leading to elevated FTH expression and reduced intracellular Fe2+ levels. In addition, AAV-mediated overexpression of Yap1 (AAV-Yap1 OE) reversed the suppression of ferritinophagy induced by piezo1 CKO, as evidenced by elevated NCOA4 expression in the CEP of Tail-looping model (Figure 6K and L). Consistently, treatment with verteporfin (the YAP1-TEAD1 axis inhibitor) suppressed Yoda1-induced NCOA4 upregulation, ultimately leading to reduced FTH expression and reduced abnormal iron deposition (Figure 6M-P). These findings imply that PIEZO1 initiates NCOA4-mediated ferritinophagy via modulation of the YAP1- WWTR1-TEAD1 signaling axis under oxidative stress. Figure 6.YAP1-TEAD1 signaling axis were involved in PIEZO1-mediated ferritinophagy. (A and B) Representative fluorescence analysis of YAP1 in CEPCs treated with H2O2 (100 μM), Yoda1 (10 μM), or GsMTx4 (1 μM) for 24 h (n = 3). (C and D) Representative western blot analysis of YAP1, p-YAP1, TEAD1, WWTR1, CYR61, CTGF and ANKRD1 expression in CEPCs treated with H2O2 (100 μM), Yoda1 (1, 5, 10 μM) or GsMTx4 (0.1, 0.5, 1 μM) for 24 h (n = 3). (E and F) Representative histological staining analysis of TEAD1 in CEP (n = 4). (G-J) CEPCs were transfected with si-Yap1 or si-NC for 72 h, followed by treatment with H2O2 (100 μM) or Yoda1 (10 μM) for 24 h. (G and H) Representative western blot analysis of NCOA4, FTH expression in CEPCs (n = 3). (I and J) Representative immunofluorescence analysis of FerroOrange in CEPCs (n = 3). (K and L) Representative histological staining analysis of NCOA4 in CEP of the Tail-looping model following 8 weeks of intervention with AAV-GFP or AAV-Yap1 OE (n = 4). (M and N) Representative western blot analysis of NCOA4 and FTH expression in CEPCs treated with H2O2 (100 μM), Yoda1 (10 μM) and verteporfin (1 μM) for 24 h (n = 3). (O and P) Representative histological analysis of NCOA4 and FTH in CEP of LSI model following a 3-month treatment with intraperitoneal Yoda1 (3 mg/kg/w) and verteporfin (20 mg/kg/w) (n = 4).The image contains multiple panels illustrating various analyses of YAP1, TEAD1 and NCOA4 in CEPCs under different treatments. The first panel shows fluorescence analysis of YAP1 in CEPCs treated with hydrogen peroxide, Yoda1, or GsMTx4, highlighting YAP1, F-actin and DAPI. The second panel presents a bar graph of relative fluorescence intensity of YAP1. The third panel displays western blot analysis of YAP1, p-YAP1, TEAD1, WWTR1, CYR61, CTGF and ANKRD1 expression in CEPCs. The fourth panel shows bar graphs of protein expression levels relative to GAPDH. The fifth panel illustrates histological staining of TEAD1 in different genetic models. The sixth panel provides a graph of TEAD1-positive area percentage. The seventh panel shows western blot analysis of NCOA4 and FTH expression with siRNA treatments. The eighth panel presents a bar graph of relative protein levels of NCOA4 and FTH. The ninth panel displays immunofluorescence analysis of FerroOrange in CEPCs. The tenth panel shows a bar graph of relative fluorescence intensity. The eleventh panel illustrates histological staining of NCOA4 in tail-looping models. The twelfth panel provides a graph of NCOA4-positive area percentage. The thirteenth panel shows western blot analysis of NCOA4 and FTH with verteporfin treatment. The fourteenth panel presents a bar graph of relative protein levels. The fifteenth panel illustrates histological analysis of NCOA4, FTH and Perl's staining in different treatment models. The sixteenth panel provides graphs of positive area percentages for NCOA4, FTH and Perl's staining.Panels analyze YAP1, TEAD1, NCOA4 in CEPCs via fluorescence, western blot, histology under treatments. YAP1-TEAD1 signaling axis were involved in PIEZO1-mediated ferritinophagy. (A and B) Representative fluorescence analysis of YAP1 in CEPCs treated with H2O2 (100 μM), Yoda1 (10 μM), or GsMTx4 (1 μM) for 24 h (n = 3). (C and D) Representative western blot analysis of YAP1, p-YAP1, TEAD1, WWTR1, CYR61, CTGF and ANKRD1 expression in CEPCs treated with H2O2 (100 μM), Yoda1 (1, 5, 10 μM) or GsMTx4 (0.1, 0.5, 1 μM) for 24 h (n = 3). (E and F) Representative histological staining analysis of TEAD1 in CEP (n = 4). (G-J) CEPCs were transfected with si-Yap1 or si-NC for 72 h, followed by treatment with H2O2 (100 μM) or Yoda1 (10 μM) for 24 h. (G and H) Representative western blot analysis of NCOA4, FTH expression in CEPCs (n = 3). (I and J) Representative immunofluorescence analysis of FerroOrange in CEPCs (n = 3). (K and L) Representative histological staining analysis of NCOA4 in CEP of the Tail-looping model following 8 weeks of intervention with AAV-GFP or AAV-Yap1 OE (n = 4). (M and N) Representative western blot analysis of NCOA4 and FTH expression in CEPCs treated with H2O2 (100 μM), Yoda1 (10 μM) and verteporfin (1 μM) for 24 h (n = 3). (O and P) Representative histological analysis of NCOA4 and FTH in CEP of LSI model following a 3-month treatment with intraperitoneal Yoda1 (3 mg/kg/w) and verteporfin (20 mg/kg/w) (n = 4). PIEZO1-mediated ferritinophagy requires extracellular Ca2+ influx under oxidative stress As an ion channel with high selectivity for Ca2+, PIEZO1 exerts a key function in mediating calcium influx in response to mechanical stimulation. Therefore, CECPs were treated with H2O2 (100 μM), Yoda1 (10 μM) or GsMTx4 (1 μM) for 24 h in calcium-free culture medium. Western blot analysis (Figure 7A-B and S7A-B) revealed that in the absence of extracellular Ca2+, Yoda1 treatment failed to induce ferritinophagy, as evidenced by downregulation of NCOA4, LC3-II:I, ATG7 and ATG5, along with upregulation of FTH and SQSTM1, in contrast to the results shown in Figure 2C and S3A-B. Furthermore, Yoda1 treatment suppressed the expression of TFR1, DMT1 and FPN, whereas GsMTx4 exerted the opposite effect (Figure 7A and B). Similarly, calcium ion depletion abolished the ability of PIEZO1 to downregulate PCBP1, PCBP2, and HAMP, as shown in Figure S7C and S7D (compare with Figures S2A and S2B). Subsequent western blot (Figure 7C-D and S7E-F) and RT-qPCR (Figure S6C) analysis elucidated that Yoda1 lost its activating effect on the YAP1-TEAD1 signaling axis under extracellular Ca2+ deficiency, resulting in no increase in the levels of TEAD1, WWTR1, CYR61, CTGF, and ANKRD1. Remarkably, the observed downregulation of YAP1 may be attributed to direct inhibition by AMPK via the Hippo pathway in the absence of Ca2+ (Figure S7I-K). Interestingly, Ca2+ depletion markedly reduced TEAD1 binding to the Ncoa4 promoter (Figure S6D and S6E). FerroOrange staining confirmed that treatment with H2O2 (100 μM) and Yoda1 (10 μM) increased intracellular Fe2+ levels in CEPCs. However, under extracellular Ca2+ deprivation, this increase was suppressed (Figure 7E and F). BAPTA-AM, a cell-permeable calcium chelator, was employed to further probe the role of Ca2+ in PIEZO1-mediated ferritinophagy. As shown in Figure 7G and H, treatment with BAPTA-AM (5 μM) for 24 h significantly suppressed the upregulation of NCOA4 and downregulation of FTH induced by H2O2 (100 μM) and Yoda1 (10 μM). Drawing on these findings, we conclude that extracellular Ca2+ influx is necessary for ferritinophagy mediated by PIEZO1. Figure 7.PIEZO1-mediated ferritinophagy requires extracellular Ca2+ influx under oxidative stress. (A-D) CEPCs were treated with H2O2 (100 μM), Yoda1 (1, 5, 10 μM) or GsMTx4 (0.1, 0.5, 1 μM) in calcium chloride-free medium for 24 h. (A and B) Western blot analysis of iron metabolic, ferritinophagy and ferroptotic markers in CEPCs (n = 3). (C and D) Western blot analysis of YAP1, p-YAP1, TEAD1, and WWTR1 in CEPCs (n = 3). (E and F) Representative immunofluorescence analysis of fluo-4 AM and FerroOrange staining in CEPCs after treatment with H2O2 (100 μM) and Yoda1 (10 μM) for 24 h (n = 3). (G and H) Western blot analysis of NCOA4 and FTH in CEPCs treated with H2O2 (100 μM), Yoda1 (10 μM) and BAPTA-AM (5 μM) for 24 h (n = 3). (I and J) Histological analysis of PIEZO1, NCOA4, YAP1 and Perls expression in human CEP tissue from convex and concave regions without osteochondritis (n = 8). (K) Quantitative analysis of iron ion content in concave human CEP with and without osteochondritis (n = 6). (L and O) Western blot analysis of PIEZO1, ATG7, NCOA4, SQSTM1, ATG5, FTH, LC3-1, and LC3-II in CEPCs treated with 10 μM Yoda1 for 0.5, 1, 4, 12 and 24 h (n = 3). (M and N) MRI, SOFG and Perls staining analysis of CEP after Tail-looping and puncture model for 8 weeks (n = 4).The composite image includes multiple panels detailing various analyses of CEPCs under different treatments. Panel A shows western blot analysis of proteins such as TFR1, NCOA4, DMT1, FPN, GPX4, FTH and LC3 in CEPCs treated with hydrogen peroxide, Yoda1 and GsMTx4 in calcium-free medium. Panel B presents graphs of protein expression levels, with the x-axis labeled with treatment conditions and the y-axis labeled as protein expression relative to GAPDH. Panel C shows western blot analysis of YAP1, p-YAP1, TEAD1 and WWTR1 under similar conditions. Panel D provides graphs of these protein levels, with similar axis labeling. Panel E displays immunofluorescence images of CEPCs stained with Fluo-4 AM and FerroOrange, showing cellular responses to treatments. Panel F includes graphs of relative fluorescence density, with treatment conditions on the x-axis and fluorescence density on the y-axis. Panel G shows western blot analysis of NCOA4 and FTH after treatment with hydrogen peroxide, Yoda1 and BAPTA-AM. Panel H presents graphs of relative protein levels. Panel I shows histological analysis of PIEZO1, NCOA4, YAP1 and Perls expression in human CEP tissue from convex and concave regions. Panel J provides quantitative analysis of positive case percentages in human CEP. Panel K shows relative iron content in concave human CEP with and without osteochondritis. Panel L displays western blot analysis of PIEZO1, ATG7, NCOA4, SQSTM1, ATG5, FTH and LC3 in CEPCs treated with Yoda1 over time. Panel M includes MRI and SOFG analysis of CEP after tail-looping and puncture model. Panel N presents graphs of relative protein levels and Perls-positive area, with treatment conditions on the x-axis and protein levels on the y-axis. Panel O provides graphs of protein expression levels over time, with similar axis labeling.Composite image: western blot, immunofluorescence, histology, MRI of CEPCs under different treatments.Interestingly, histological analysis of osteochondritis-free human CEP revealed upregulation of PIEZO1, NCOA4, and YAP1 on the concave compared to the convex side, whereas Perls staining revealed no significant discrepancy in iron deposition between the two regions (Figure 7I and J). The above findings indicate that inflammatory environment may be essential for PIEZO1-mediated ferritinophagy. Next, we quantified iron levels in concave endplate cartilage with and without osteochondritis. The results indicated that both Fe2+ and Fe3+ levels were elevated in endplate cartilage with osteochondritis compared to osteochondritis-free tissue (Figure 7K). Besides, when CEPCs were treated with Yoda1 in the absence of H2O2, western blot analysis (Figure 7Land O) revealed that although PIEZO1 and NCOA4 were activated, ferritinophagy was not induced, as indicated by upregulation of SQSTM1 and FTH, as well as downregulation of ATG7, ATG5, and LC3-II:I. This also explained why Yoda1 treatment alone, while sufficient to induce extracellular Ca2 + influx, did not lead to an increase in intracellular Fe2+ levels (Figure 7E and F). Notably, Yoda1 alone also failed to activate Yap1 transcriptional activity (Figure S7G and S7H). Meanwhile, the sole intervention with Yoda1 conferred ferroptosis resistance, as manifested by the upregulation of SLC7A11, GPX4, and FSP1, and the downregulation of ACSL4 and HAMP (Figure S8A-D). Moreover, puncture procedures were performed on the Tail-looping model to simulate oxidative stress conditions which were confirmed by the reduced T2-weighted magnetic resonance imaging (MRI) signal. As shown in Figure 7(M-N) and S8E-F, alterations in bending stress alone did not significantly affect the CEP. However, the combined presence of mechanical stress and oxidative stress induced CEP degeneration, as evidenced by an elevated endplate score and increased iron deposition. In conclusion, our results demonstrate that PIEZO1-mediated ferritinophagy requires both extracellular Ca2+ influx and the presence of oxidative stress environment. PIEZO1-mediated ferritinophagy requires extracellular Ca2+ influx under oxidative stress. (A-D) CEPCs were treated with H2O2 (100 μM), Yoda1 (1, 5, 10 μM) or GsMTx4 (0.1, 0.5, 1 μM) in calcium chloride-free medium for 24 h. (A and B) Western blot analysis of iron metabolic, ferritinophagy and ferroptotic markers in CEPCs (n = 3). (C and D) Western blot analysis of YAP1, p-YAP1, TEAD1, and WWTR1 in CEPCs (n = 3). (E and F) Representative immunofluorescence analysis of fluo-4 AM and FerroOrange staining in CEPCs after treatment with H2O2 (100 μM) and Yoda1 (10 μM) for 24 h (n = 3). (G and H) Western blot analysis of NCOA4 and FTH in CEPCs treated with H2O2 (100 μM), Yoda1 (10 μM) and BAPTA-AM (5 μM) for 24 h (n = 3). (I and J) Histological analysis of PIEZO1, NCOA4, YAP1 and Perls expression in human CEP tissue from convex and concave regions without osteochondritis (n = 8). (K) Quantitative analysis of iron ion content in concave human CEP with and without osteochondritis (n = 6). (L and O) Western blot analysis of PIEZO1, ATG7, NCOA4, SQSTM1, ATG5, FTH, LC3-1, and LC3-II in CEPCs treated with 10 μM Yoda1 for 0.5, 1, 4, 12 and 24 h (n = 3). (M and N) MRI, SOFG and Perls staining analysis of CEP after Tail-looping and puncture model for 8 weeks (n = 4). In addition, as TFR1 and DMT1 were downregulated irrespective of extracellular Ca2+ availability (Figures 2(C-D) and 7(A-B)), the upregulation of FTH (Figure 7A and L), PCBP1, and PCBP2 (Figure S8C and S8D) can be attributed to PIEZO1-mediated iron influx. This subsequently induced the elevated expression of SLC7A11 and GPX4 (Figure 7(A-B) and S8A-B), possibly as a compensatory mechanism to counteract ferroptosis. To further investigate the relationship between PIEZO1-mediated iron influx and mechanical stress-induced iron overload in the CEP, Iron-Dextran was administered via intraperitoneal injection to the LSI model. Immunohistochemistry analysis showed that compared to the group receiving Iron-Dextran alone, the CEP in the group treated with both Iron-Dextran (1 g/kg/w) and Yoda1 (3 mg/kg/w) exhibited greater iron deposition, suggesting that PIEZO1-mediated iron influx contributes to iron overload in the endplate cartilage (Figure S8G and S8H). PIEZO1-mediated ferritinophagy activates the Z-mtDNA-ZBP1 signaling axis Building on previous findings that PIEZO1-mediated mechanical stress induces mitochondrial dysfunction, and considering our team’s earlier confirmation that iron overload exacerbates CEP degeneration via mtDNA release, we focused our subsequent investigation on mtDNA. RNA sequencing analysis showed that 1 Mpa mechanical stimulation significantly affectes cytosolic DNA-sensing pathway and upregulates ZBP1 expression in CEPCs (Figure S9A and S9B). Next, western blot analysis revealed that ZBP1 expression increases in a dose-dependent manner with rising concentrations of Yoda1 under inflammatory conditions (Figure 8(A)). Besides, immunofluorescence analysis indicated that Yoda1 treatment enhances ZBP1 expression, whereas GsMTx4 suppressed it, suggesting that PIEZO1 regulates ZBP1 (Figure 8(B)). To determine whether ferritinophagy contributes to the regulation of ZBP1 by PIEZO1, we transfected CEPCs with siRNA targeting Ncoa4. As shown in Figure 8(C), Ncoa4 knockdown suppressed the PIEZO1-induced upregulation of ZBP1. Given that ZBP1 functions as a sensor for Z-form DNA (Z-DNA), immunofluorescence staining was conducted with a specific anti-Z-DNA antibody to confirm Z-DNA formation. The signals of Z-DNA and its colocalization with TOMM20 were significantly enhanced by H2O2 and Yoda1 treatment but abolished upon Ncoa4 knockdown, confirming that PIEZO1-mediated ferritinophagy promotes Z-mtDNA formation (Figure 8(D) and (E)). Furthermore, immunofluorescence confirmed that ferritinophagy-induced mitochondrial stress leads to the release of Z-mtDNA, which recognized by ZBP1 (Figure 8(F) and (G)). To further confirm that the Z-DNA colocalizing with ZBP1 was mitochondrial in origin, CEPCs were treated with cyclosporin A (the blocker of mtDNA release) and ethidium bromide (the inhibitor of mtDNA replication). Treatment with either cyclosporin A (CsA) or ethidium bromide (EtBr) reduced ZBP1 expression (Figure 8H) and suppressed both the formation and release of Z-mtDNA in CEPCs (Figure 8(I and J)). And CsA treatment also exhibited an inhibitory effect on ZBP1 in the LSI model (Figure 8(K)). Additionally, AAV-mediated knockdown of Ncoa4 (AAV-Ncoa4 KD) inhibited the bending stress-induced upregulation of ZBP1, while AAV-mediated overexpression of Ncoa4 (AAV-Ncoa4 OE) reversed the downregulation of ZBP1 caused by piezo1 knockout (Figure 8L and M). Collectively, these findings underscore the pivotal biological link between the Z-mtDNA-ZBP1 axis and PIEZO1-mediated ferritinophagy. Figure 8.PIEZO1-mediated ferritinophagy activates the Z-mtDNA-ZBP1 signaling axis. (A) Western blot analysis of ZBP1 expression in CEPCs treated with H2O2 (100 μM) for 24 h and Yoda1 (10 μM) for 0.5, 1, 4, 12, or 24 h (n = 3). (B) Immunofluorescence analysis of ZBP1 in CEPCs following treatment with H2O2 (100 μM), Yoda1 (10 μM) or GsMTx4 (1 μM) for 24 h (n = 3). (C) Western blot analysis of ZBP1 expression in CEPCs transfected with si-Ncoa4 or si-NC for 72 h, followed by treatment with H2O2 (100 μM) or Yoda1 (10 μM) for 24 h (n = 3). (D and E) Fluorescence colocalization analysis of Z-DNA and TOMM20 (n = 3). (F and G) Fluorescence colocalization analysis of Z-DNA and ZBP1 (n = 3). Arrows indicate colocalization. (H-J) Western blot analysis of ZBP1 expression and immunofluorescence colocalization analysis of Z-DNA with TOMM20 in CEPCs pretreated with CsA (5 μg/mL) or EtBr (1 μg/mL) for 12 h, followed by co-treatment with H2O2 (100 μM) or Yoda1 (10 μM) for 24 h (n = 3). (K) Immunohistochemistry analysis of ZBP1 expression in LSI model following 3-month treatment with Yoda1 (3 mg/kg/w) and CsA (20 μg/kg/w) (n = 4). (L and M) Immunohistochemistry analysis of ZBP1 expression in Tail-looping model after 8 weeks of intervention with AAV-GFP, AAV-Ncoa4 KD or AAV-Ncoa4 OE (n = 4).The composite image includes multiple panels showing various analyses related to ZBP1 expression and colocalization in CEPCs. Image A presents western blot analysis of ZBP1 expression in CEPCs treated with hydrogen peroxide and Yoda1 at different time intervals, with a graph showing ZBP1 to GAPDH ratio. Image B shows immunofluorescence analysis of ZBP1 in CEPCs treated with hydrogen peroxide, Yoda1 and GsMTx4, alongside a graph of relative fluorescence intensity. Image C displays western blot analysis of ZBP1 expression in CEPCs transfected with si-Ncoa4, with a graph of ZBP1 to GAPDH ratio. Image D illustrates fluorescence colocalization analysis of Z-DNA and TOMM20 in CEPCs, with a graph of Pearson's coefficient. Image E shows fluorescence colocalization analysis of Z-DNA and ZBP1, with a graph of Pearson's coefficient. Image F presents immunofluorescence analysis of ZBP1 expression and colocalization with Z-DNA in CEPCs treated with cyclosporin A and ethidium bromide, with a graph of Pearson's coefficient. Image G shows immunohistochemistry analysis of ZBP1 expression in LSI model following treatment with Yoda1 and cyclosporin A, with a graph of ZBP1 positive cells percentage. Image H displays immunohistochemistry analysis of ZBP1 expression in Tail-looping model after intervention with AAV-GFP and AAV-Ncoa4 KD, with a graph of ZBP1 positive cells percentage. Image I shows immunohistochemistry analysis of ZBP1 expression in Tail-looping model after intervention with AAV-GFP and AAV-Ncoa4 OE, with a graph of ZBP1 positive cells percentage.Composite image: ZBP1 expression and colocalization in CEPCs via blot, fluorescence and histochemistry. PIEZO1-mediated ferritinophagy activates the Z-mtDNA-ZBP1 signaling axis. (A) Western blot analysis of ZBP1 expression in CEPCs treated with H2O2 (100 μM) for 24 h and Yoda1 (10 μM) for 0.5, 1, 4, 12, or 24 h (n = 3). (B) Immunofluorescence analysis of ZBP1 in CEPCs following treatment with H2O2 (100 μM), Yoda1 (10 μM) or GsMTx4 (1 μM) for 24 h (n = 3). (C) Western blot analysis of ZBP1 expression in CEPCs transfected with si-Ncoa4 or si-NC for 72 h, followed by treatment with H2O2 (100 μM) or Yoda1 (10 μM) for 24 h (n = 3). (D and E) Fluorescence colocalization analysis of Z-DNA and TOMM20 (n = 3). (F and G) Fluorescence colocalization analysis of Z-DNA and ZBP1 (n = 3). Arrows indicate colocalization. (H-J) Western blot analysis of ZBP1 expression and immunofluorescence colocalization analysis of Z-DNA with TOMM20 in CEPCs pretreated with CsA (5 μg/mL) or EtBr (1 μg/mL) for 12 h, followed by co-treatment with H2O2 (100 μM) or Yoda1 (10 μM) for 24 h (n = 3). (K) Immunohistochemistry analysis of ZBP1 expression in LSI model following 3-month treatment with Yoda1 (3 mg/kg/w) and CsA (20 μg/kg/w) (n = 4). (L and M) Immunohistochemistry analysis of ZBP1 expression in Tail-looping model after 8 weeks of intervention with AAV-GFP, AAV-Ncoa4 KD or AAV-Ncoa4 OE (n = 4). ZBP1-mediated activation of the NFKB pathway exacerbates CEP osteochondritis and degeneration To elucidate the mechanism through which ZBP1 activation contributes to CEP degeneration, subsequent investigations focued on the downstream cascades initiated by ZBP1. As showed in Figure 9A, siRNA-mediated knockdown of Zbp1 (si-Zbp1) suppressed the activation of the NFKB signaling pathway, as indicated by an decreased p-NFKB expression. Additionally, we observed the knocking down Zbp1 also repressed Mito-ROS in CEPCs (Figure 9B and E). Next, the Tail-looping model was constructed to further explore the specific role of ZBP1 in mechanical stress-induced CEP degeneration. Immunohistochemistry analysis demonstrated that the silencing of Zbp1 caused by AAV (AAV-Zbp1 KD) attenuated PIEZO1-mediated inflammatory response, manifested by the lower expression of IL1B, IL6 and TNF in the CEP (Figure 9C-D and F-G). Then, we further assessed alterations in ECM metabolism using western blot (Figure 9H and I) and immunofluorescent assays (Figure 9J and K). Zbp1 knockdown upregulated the anabolic markers COL2 and SOX9, while downregulating the catabolic markers MMP13 and MMP3, as well as the osteogenic markers COL10A1 and RUNX2. Collectively, our findings indicated that ZBP1 inhibition alleviates mechanical overloading-induced osteochondritis and degeneration of the CEP. Figure 9.ZBP1 activates the NFKB pathway, thereby exacerbating CEP osteochondritis and degeneration. CEPCs were transfected with si-Zbp1 or si-NC for 72 h following H2O2 (100 μM) and Yoda1 (10 μM) induction for 24 h. The intervertebral spaces of the Tail-looping mouse model were injected with AAV-Zbp1 KD or AAV-GFP and evaluated after 2 months. (A) Western blot analysis of NFKB and p-NFKB expression in CEPCs (n = 3). (B and E) Immunofluorescence analysis of Mito-ROS levels in CEPCs (n = 3). (C, D, F and G) Immunohistochemical analysis of IL1B, IL6 and TNF expression in CEP (n = 4). (H and I) Western blot analysis of ECM metabolism markers (COL2, SOX9, COL10A1, RUNX2, MMP13, MMP3) in CEPCs (n = 3). (J and K) Immunofluorescence analysis of COL2 and COL10A1 expression in CEP (n = 4). (L) Schematic diagram of mechanical stress-induced iron homeostasis disorder and degeneration in the CEPCs.The composite image illustrates ZBP1's impact on CEP degeneration through various analyses. Panel A shows western blot results for NFKB and p-NFKB in CEPCs, with bar graphs of protein levels under different conditions. Panel B highlights Mito-ROS levels in CEPCs via immunofluorescence, showing treatment differences. Panel C presents immunohistochemical analysis of IL1B, IL6 and TNF in CEP, with positive areas and graphs in panels D, F and G. Panel H displays western blot analysis of ECM markers (COL2, SOX9, COL10A1, RUNX2, MMP13, MMP3) in CEPCs, with relative expression graphs in panel I. Panel J shows immunofluorescence of COL2 and COL10A1 in CEP, with positive area graphs in panel K. Panel L includes a schematic of mechanical stress-induced iron homeostasis disorder and degeneration in CEPCs, featuring a mouse model for injections and evaluations. The diagram outlines pathways involving PIEZO1, TFR1 and ZBP1, contributing to CEP degeneration through cellular processes.Image: ZBP1's impact on CEP degeneration via blot, fluorescence, histochemistry and diagram. ZBP1 activates the NFKB pathway, thereby exacerbating CEP osteochondritis and degeneration. CEPCs were transfected with si-Zbp1 or si-NC for 72 h following H2O2 (100 μM) and Yoda1 (10 μM) induction for 24 h. The intervertebral spaces of the Tail-looping mouse model were injected with AAV-Zbp1 KD or AAV-GFP and evaluated after 2 months. (A) Western blot analysis of NFKB and p-NFKB expression in CEPCs (n = 3). (B and E) Immunofluorescence analysis of Mito-ROS levels in CEPCs (n = 3). (C, D, F and G) Immunohistochemical analysis of IL1B, IL6 and TNF expression in CEP (n = 4). (H and I) Western blot analysis of ECM metabolism markers (COL2, SOX9, COL10A1, RUNX2, MMP13, MMP3) in CEPCs (n = 3). (J and K) Immunofluorescence analysis of COL2 and COL10A1 expression in CEP (n = 4). (L) Schematic diagram of mechanical stress-induced iron homeostasis disorder and degeneration in the CEPCs. Discussion Mechanical load is a prevalent contributor to lumbar degeneration [44]. Due to its role as a stress-concentrating region, the CEP is particularly vulnerable to abnormal mechanical stimulation [2]. PIEZO1, a mechanosensitive ion channel, has been proven to trigger iron overload through the regulation of erythrophagocytosis by macrophages [28], and our previous studies also demonstrated that iron overload accelerates IVDD by inducing ferroptosis in endplate chondrocytes [24]. Curiosity-driven investigation revealed abnormal iron deposition in the concave endplate cartilage of patients with degenerative scoliosis, as well as in the CEP of mice following LSI surgery. In both scenarios, the endplate cartilage was subjected to abnormal mechanical stress and characterized by upregulated PIEZO1 expression. Nevertheless, the mechanism by which the intervertebral disc, the largest avascular tissue in human, acquires excess iron remains unclear. To elucidate the underlying mechanism of excessive iron accumulation in the CEP caused by mechanical stress, CEPCs were treated with a PIEZO1 agonist (Yoda1) and inhibitor (GsMTx4) under inflammatory conditions (H2O2). From the iron metabolism perspective, Yoda1 treatment suppressed both iron intake (TFR1, DMT1) and excretion (FPN), which could not adequately account for the iron overload in CEPCs. From the ferroptosis perspective, Yoda1 treatment did not downregulate GPX4 or SLC7A11, contrary to the effects observed in CEPCs with 1 MPa mechanical stimulation. This discrepancy may be attributed to the differences between physical activation (mechanical stress) and chemical activation (Yoda1) of PIEZO1. Excitingly, Yoda1-induced ferritinophagy likely explains iron deposition in CEPCs, indicated by elevated NCOA4 expression, enhanced colocalization of FTH with LC3 and FerroOrange with lysosome, as well as decreased FTH levels. Furthermore, both GsMTx4 treatment and chondrocyte-specific piezo1 knockout have validated the positive role of PIEZO1 in regulating ferritinophagy in vitro and in vivo. NCOA4-mediated degradation of ferritin releases excess iron ions [22], leading to lipid peroxidation and ECM degradation, which has recently been identified as a critical mechanism underlying chondrocyte ferroptosis in osteoarthritis [23]. We further elucidated the detrimental effects of ferritinophagy on mechanical stress-induced CEP degeneration using autophagy inhibitors (3-MA, BafA1), Ncoa4-targeting siRNA (si-Ncoa4), and AAV (AAV-Ncoa4 KD and AAV-Ncoa4 OE). Distinctly, COL2, the primary collagen component of cartilage, is essential for effective tissue repair. In contrast, COL10A1-a hallmark of hypertrophic chondrocytes-is upregulated during matrix calcification and promotes the secretion of MMP3 and MMP13, which directly degrade COL2. Notably, inhibiting mechanical stress-induced ferritinophagy preserved chondrocyte phenotype stability. YAP1-WWTR1 serves as a mechanosensory hub [29] that transduces mechanical signals and is key to multiple PIEZO1-mediated pathological processes, including fracture healing [45], skeletal remodeling [11] and macrophage polarization [46]. Besides, as nuclear transducers of the Hippo signaling pathway, YAP1-WWTR1 is strongly implicated in the regulation of ferroptosis [47]. However, the mechanism through which YAP1 bridges PIEZO1 and iron overload remains unclear. Our findings suggest that PIEZO1 facilitates YAP1 nuclear translocation and activates the transcription factor TEAD1, leading to NCOA4-mediated ferritinophagy. Furthermore, verteporfin treatment attenuated mechanical stress-induced CEP degeneration by suppressing YAP1-TEAD1 binding. Interestingly, we discovered that the regulation of ferritinophagy by the PIEZO1-YAP1-TEAD1 axis depends on Ca2+ influx under oxidative stress conditions. Our study showed that, as an ion channel that transduces mechanical stimuli into calcium signaling, PIEZO1 cannot activate YAP1-TEAD1 and thereby initiate ferritinophagy in the absence of Ca2+, an effect likely attributable to PIEZO1-mediated activation of the Hippo signaling pathway and the direct inhibition of YAP1 by AMPK [48]. Excitedly, relevant studies have supported the regulatory role of Ca2+ in ferroptosis [49]. Moreover, under non-inflammatory conditions (without H2O2), Yoda1 upregulated YAP1 and NCOA4 yet failed to elevate Fe2+ levels in the LIP, likely resulting from impaired autophagy, as evidenced by a decreased LC3-II:I. Additionally, no abnormal iron deposition was detected in either concave human endplate cartilage lacking osteochondritis or in the CEP of the Tail-looping mouse model [50] without intradiscal needling. Nevertheless, despite the lack of ferritinophagy induction in either scenario, Yoda1-treated CEP chondrocytes exhibited upregulated expression of FTH, which attracted our attention. A recent study revealed that PIEZO1, a nonselective cation channel, mediates ferroptosis in nucleus pulposus cells by promoting iron ion influx [51]. We therefore proposed that in the absence of Ca2+ or inflammatory stimuli, CEP chondrocytes may respond to mechanical stress by permitting iron influx, while compensating through upregulation of FTH, as well as SLC7A11, GPX4, ACSL4 and FSP1. Furthermore, our results also confirmed that combined treatment with Yoda1 and iron dextran in the LSI model partly exacerbated mechanical load-induced iron accumulation in the CEP. The Fenton reaction caused by iron overload typically exacerbates Mito-ROS accumulation [52], and our previous research has demonstrated that disruption of iron homeostasis aggravates CEP degeneration via mtDNA release [41]. Strikingly, ROS-induced mtDNA instability facilitates its transition from B-form to Z-DNA [53]. Recent studies have emphasized the role of ZBP1 as a sensor for Z-mtDNA [34,35], which triggers receptor-interacting protein kinase 3-dependent ferroptosis in response to diquat poisoning [54]. Intriguingly, emerging evidence indicates that the mtDNA-ZBP1 axis exerts a pathogenic effect on the progression of osteoarthritis [55]. However, the link between ZBP1 and IVDD is fully unknown. Notably, our data reveals for the first time that PIEZO1 activation promotes the formation and release of Z-mtDNA, which subsequently binds to and activates ZBP1. Furthermore, AAV-mediated knockdown of Ncoa4 or CsA-induced suppression of Z-mtDNA release attenuated the ZBP1 response. Given that the Z-mtDNA-ZBP1 axis functions downstream of ferritinophagy, we evaluated the role of ZBP1 in CEP degeneration. Interestingly, we observed that ZBP1 promotes the release of inflammatory mediators in an NFKB-dependent manner, ultimately contributing to degenerative phenotypic changes in the CEP. This study has several limitations. Firstly, the evaluation of iron ion content in the CEP of patients with degenerative scoliosis was based on two classification criteria: convex versus concave endplate zone and the presence or absence of endplate osteochondritis. However, other clinically relevant parameters, the scoliosis Cobb angle and the severity of Modic changes, were not incorporated into the consideration. Secondly, owing to the limited number of clinical cases available, it was not feasible to evaluate the influence of Ca2+ on ferritinophagy-mediated CEP degeneration by stratifying patients with degenerative scoliosis according to blood calcium levels (hypocalcemia versus hypercalcemia). Thirdly, as confirmed in this study, PIEZO1 can mediate iron ion influx under calcium-deficient conditions, thereby exacerbating CEP degeneration. However, direct evidence of PIEZO1-mediated iron ion influx remains unavailable, since current patch-clamp techniques are incapable of detecting Fe2+ or Fe3+ ions. Furthermore, the current lack of appropriate low-calcium animal models precludes in vivo validation of whether PIEZO1 alone can induce IVDD via iron influx, and the future development of vertebral disc organoids may offer a promising strategy to investigate this mechanism. Fourthly, iron homeostasis is orchestrated by an integrated network of regulatory systems, including the iron regulatory protein/iron response element hub, and the HIF (hypoxia inducible factor) axis [56]. Critically, iron regulatory protein 2 is required for the initiation of ferritinophagy upon iron depletion, as evidenced by recent experimental findings [57]. Moreover, relevant data also show that HIF-1α is implicated in the reduction of ferritinophagy and the concomitant decline in autophagic flux under hypoxia [58]. However, within a mechanical stress context, whether the iron regulatory protein/iron response element system or the HIF axis plays a pivotal role in regulating ferritinophagy remains unclear and warrants further investigation. Fifth, our data provided evidence that the ferritinophagy links mechanical loading and the Z-mtDNA-ZBP1 axis during CEP degeneration. However, the direct regulatory effect of ferritinophagy on ZBP1 has not yet been validated in vivo, due to the absence of NCOA4 conditional knockout mouse models. Lastly, although we are the first to demonstrate that inhibiting the PIEZO1-YAP1-NCOA4-ZBP1-NFKB axis can delay IVDD, conventional treatment strategies often prove ineffective due to the unique avascular nature of the disc. Therefore, based on this mechanistic insight, we plan to develop novel piezoelectric materials for targeted therapeutic applications. In summary, our study unveiled that under oxidative stress, abnormal mechanical stress promotes YAP1 nuclear translocation via PIEZO1-mediated Ca2+ influx, thereby inducing NCOA4-dependent ferritinophagy. The resulting increase in Fe2+ levels facilitates the formation and release of Z-mtDNA, which is recognized by ZBP1 and triggers the NFKB-dependent inflammatory cascade, ultimately aggravating CEP degeneration. In addition, inhibition of the PIEZO1-YAP1-NCOA4-ZBP1-NFKB axis delayed IVDD progression. These findings suggest that this signaling pathway holds promise as a potential therapeutic target for IVDD (Figure 9(L)). Materials and methods Human samples Human endplate specimens and imaging data were collected from the surgical patients with degenerative scoliosis. Informed consent was obtained in advance from all participants, and the protocol was approved by the Ethics Committee of the the Fourth Medical Center of Chinese PLA General Hospital (2024KY0116-KS001). Mice Piezo1flox/flox and Col2 CreERT+/+ homozygous mice were established by and purchased from Cyagen (USA). The offspring were intercrossed to generate piezo1flox/flox Col2 CreERT+/- and Piezo1flox/flox Col2 CreERT-/- as the experimental group and the sham group respectively (referred to as piezo1fl/fl Col2 Cre+ and Piezo1fl/fl Col2 Cre− herein). Mice breeding strategy is shown in Figure S9C. Four-week-old piezo1fl/flCol2 Cre+ mice were intraperitoneally injected with tamoxifen (MCE, HY-13757A) at a dose of 120 mg/kg/2d for 5 times to obtain piezo1-conditional knockout (piezo1 CKO) mice. C57BL/6J wild-type (WT) mice were purchased from Beijing Vital River Laboratory Animal Technology. Eight-week-old male mice were selected for LSI or Tail-looping model. All animal experiments were approved by the Institutional Animal Care and Use Committee of the Laboratory Animal Center of Academy of Military Medical Sciences (IACUC-DWZX-2024–543). Genotyping The genotypes of the mice were determined by PCR assay of genomic DNA, which was extracted from three-week-old mouse tails using the One Step Mouse Genotyping Kit (Wengene, WG-MG-1000) with the following primers: Piezo1 forward, 5’-AGCAAGGCCAATGTAGTATCTGG-3’; Piezo1 reverse, 5’-CTATTGGTGCCTAGTTGGCAGAC-3’; Col2 CreERT forward, 5’-GGCTCTACTTCATCGCATTCCTTG-3’; Col2 CreERT reverse, 5’-CGCAAACAAGTCTCACAAAGGAG-3.’ Bands were captured by the dark-box ultraviolet analyzer (ZYBOT, BOT-III) after agarosegel electrophoresis using the 2% Agarose gels (Solarbio, AG00208B). Genotype identification results are shown in Figure S9D. Iron ion detection in human CEP tissue Human CEP Iron ion detection were performed using Iron Assay Kit (Dojindo, I291) according to the manufacturer’s instructions. Briefly, 100 mg fresh human CEP tissue were grinded using homogenizer at 4°C in microcentrifuge tubes containing assay buffer. The supernatant was collected after homogenate was centrifuged. Assay buffer or reducer solution was added in corresponding supernatant to test Fe2+ level or total Iron level. After 1-h incubation at 37°C, the Fe2+ and total Iron levels of CEP tissue were determined by absorbance at 593 nm, and the Fe3+ level were calculated by subtraction. Western blot Western blot was conducted on protein lysates extracted by RIPA (CWBIO, CW2333S) with 1% proteinase inhibitor cocktail (MCE, HY-K0010) and 1% phosphatase inhibitor cocktail (CWBIO, CW2383). After incubation with specific antibodies, the protein bands were visualized using Immobilon western HRP substrate Kit (Millipore, WBKLS0100). The antibodies were PIEZO1 (Proteintech, 15939-1-AP), TFRC/TFR1 (Abcam, ab84036), NCOA4 (Abcam, ab314553), SLC11A2/DMT1 (Proteintech, 20507-1-AP), SLC40A1/FPN (Proteintech, 26601-1-AP), SLC7A11 (Boster, BM5318), GPX4 (Abcam, ab125066), FTH (Abcam, ab183781), LC3A/B (Cell Signaling Technology, 12741), COL2/collagen II (Proteintech, 28459-1-AP), COL10A1/collagen X (Boster, BA2023), MMP13 (Abcam, ab39012), MMP3 (Abcam, ab52915), RUNX2 (Boster, PB0171), SOX9 (Boster, A00177-2), YAP1 (Boster, PB1021), p-YAP1 (Abcam, ab76252), WWTR1/TAZ (Proteintech, 66500-1-Ig), TEAD1 (Boster, M03263-1), CCN1/CYR61 (Proteintech, 26689-1-AP), ANKRD1 (Proteintech, 11427-1-AP), CCN2/CTGF (Boster, PB0570), ACSL4 (Proteintech, 22401-1-AP), AIFM2/FSP1 (Proteintech, 20886-1-AP), NFKB (Cell Signaling Technology, 8242), p-NFKB (Cell Signaling Technology, 3033), ZBP1 (Boster, A04739-3), PCBP2 (Proteintech, 15070-1-AP), PCBP1 (Proteintech, 14523-1-AP), HAMP/HEPCIDIN (Abcam, ab190775), ATG7 (Boster, A00346), ATG5 (Boster, BA3525-2), SQSTM1/p62 (Boster, PB0458), LATS1 (Proteintech, 17049-1-AP), LATS2 (Proteintech, 20276-1-AP), p-LATS1 (Cell Signaling Technology, 8654), MST1 (Proteintech, 22245-1-AP), STK3/MST2 (Proteintech, 66637-1-Ig), p-MST (Proteintech, 80093-1-RR), PRKAA/AMPK (Proteintech, 10929-2-A), p-PRKAA/AMPK (Proteintech, 80209-6-RR), p-CAMK2/CaMKII (Cell Signaling Technology, 12716), CAMK2 (Boster, A03241-2), GAPDH (Proteintech, 10491-1-AP), HRP Conjugated Goat Anti-rabbit/mouse secondary antibodies (Boster, BA1056). Immunofluorescence analysis Chondrocytes were fixed with 4% PFA (Servicebio, G1101) for 30 min, followed by permeabilization utilizing 0.1% Triton X-100 (Solarbio, T8200) for 10 min and then blocking with 5% BSA (Biosharp, BS114) for 1 h. Subsequently, the cells were incubated with the primary antibodies against NCOA4 (Abclonal, A5695), FTH (Boster, BM4487), LC3B (Cell Signaling Technology, 83506), LAMP1 (Cell Signaling Technology, 99437), COL2/collagen II (Proteintech, 28459-1-AP), COL10A1/collagen X (Boster, BA2023), YAP1 (Cell Signaling Technology, 14074), TOMM20 (Abcam, ab186735), ZBP1 (Boster, A04739-3) and Z-DNA (Novus, NB100-749) at 4°C overnight. Then, the cells were subjected to corresponding secondary for 1 h and DAPI (Solarbio, C0065) for 10 min at 37°C. Finally, images were captured utilizing the laser confocal fluorescence microscope (Nikon, CrestOptics X-Light V3). Real-time quantitative reverse transcription PCR (RT-qPCR) analysis Total RNA was extracted using the AG RNAex Pro RNA reagent (Accurate Biology, AG21101), and reverse transcription was performed with the HiScript III All-in-one RT SuperMix (Vazyme, R333-01). mRNA expression levels were quantified using TB Green Premix Ex Taq (TaKaRa, RR420A) on the CFX Duet Real-Time PCR System (Bio-Rad). The primer sequences are as follows: Ankrd1 forward, 5’-GGAACAACGGAAAAGCGAGAA-3’; Ankrd1 reverse, 5’-GAAACCTCGGCACATCCACA-3’; Ctgf forward, 5’-GGCCTCTTCTGCGATTTCG-3’; Ctgf reverse, 5’-GCAGCTTGACCCTTCTCGG-3’; Cyr61 forward, 5’-TAAGGTCTGCGCTAAACAACTC-3’; Cyr61 reverse, 5’-CAGATCCCTTTCAGAGCGGT-3’; Gapdh forward, 5’-TGTGTCCGTCGTGGATCTGA-3’; Gapdh reverse, 5’-TTGCTGTTGAAGTCGCAGGAG-3.’ Histology and immunohistochemical analysis After fixed and then decalcified, the harvested specimens were embedded, sectioned and stained with hematoxylin-eosin (HE) (Servicebio, G1005) and safranin O and fast green (SOFG) (Servicebio, G1053), performed to assess the morphology of cartilage and subchondral bone. The severity of CEP degeneration was evaluated by IVDD assessment scoring system.DAB-enhanced Perl’s Prussian blue stain (Servicebio, GP1069) was used to assess CEP iron deposition. According to the protocol, the sections were deparaffinized and rehydrated, followed by staining with Prussian blue solution for 30 min and then DAB chromogenic solution for 10 min. Fe3+ from incompletely degraded ferritin denatured polymer were showed as brown granules under microscope. For imunohistochemistry and immunofluorescent staining, after deparaffinization, rehydration, antigen retrieval and blocking, the sections were treated with primary antibodies to PIEZO1 (Proteintech, 15939-1-AP), NCOA4 (Abclonal, A5695), COL2/collagen II (Servicebio, GB11021), COL10A1/Collagen X (Abcam, ab260040), FTH (Servicebio, GB115535), YAP1 (Servicebio, GB113975), TEAD1 (Boster, M03263-1), IL1B (Servicebio, GB11113), IL6 (Servicebio, GB11117), TNF (Servicebio, GB115701). Images were recorded using a microscope (Nikon, ECLIPSE E100). Alizarin red S (ARS) and alkaline phosphatase (ALP) assay CEP chondrocytes were cultured in osteogenic differentiation medium containing 10% FBS (Gibco), 1% P/S (Gibco), 1% L-glutamine (Gibco), 10−7M dexamethasone (Sigma, D4902), 10 mM β-glycerophosphate (Sigma, 50020) and 50 μg/ml ascorbic acid (Sigma, A8960). Qualitative and quantitative evaluations of ALP (Beyotime, C3206, P0321) and ARS (Servicebio, G1038) were conducted 1 week and 3 weeks later respectively. Mitochondrial membrane potential The mitochondrial membrane potential was evaluated via JC-1 test kit (Beyotime, C2006). Multimeric JC-1 with red fluorescence and monomeric JC-1 with green fluorescence were recorded by a fluorescence microscope (Zeiss, Axio Observer A1) after staining. Micro-CT and magnetic resonance imaging (MRI) The lumbar vertebrae after LSI surgery and the coccygeal vertebrae after Tail-looping surgery were scanned by micro-CT (Scanco Viva-CT80, Switzerland) with the resolution of 14 μm, 70 kVp, and 114 μA. Three-dimensional images were visualized by built-in software. And for MRI, the tail regions underwent MRI (PharmaScan, USA) scanning to analyze disc signal intensity changes in coronal T2-weighted images (Echo time 33 ms/Repetition time 2500 ms; Field of view 36 mm × 36 mm; Slice Thickness 0.3 mm). Reactive oxygen species (ROS) and mitochondrial ROS (Mito-ROS) measurement The intracellular total ROS levels were evaluated by ROS Assay Kit (Beyotime, S0033). The chondrocytes loaded with 2′,7′-dichlorodihydrofluorescein diacetate (DCFH-DA) probe were analyzed with a fluorescence microscope (Zeiss, Axio Observer A1). For Mito-ROS, CEPCs were stained with Mito-Tracker Red CMXRos (Beyotime, C1035) and MitoSoxTM (M36005, ThermoFisher). Subsequently, the fluorescence signal was acquired using a laser confocal fluorescence microscope (Nikon, CrestOptics X-Light V3). Isolation of CEP chondrocytes (CEPCs) Primary WT chondrocytes were isolated from the CEP of one-week-old C57BL/6J mice, and piezo1fl/fl Col2 Cre+ and Piezo1fl/fl Col2 Cre− chondrocytes were extracted from seven-week-old mice after tamoxifen treatment. After digesting with 0.25% EDTA trypsin (Gibco) for 30 min and 0.2 mg/mL Type II collagenase (Sigma, V900892) for 6 h at 37°C, the primary CEPCs were collected and cultured. CEPCs at passage 3 were used in the study. LSI model and intraperitoneal intervention To create lumbar instability, the L4-5 bilateral facet joints, supraspinous and interspinous ligaments were excised. In sham operation, only L4-5 paraspinal muscles were dissected. Regarding drug treatment, postoperative mice received intraperitoneal injections of DFO (Sigma, D9533, 100 mg/kg), Fer-1 (Glpbio, GC10380, 5 mg/kg), Verteporfin (Glpbio, GC14482, 20 mg/kg), Yoda1 (Glpbio, GC18233, 3 mg/kg), GsMTx4 (Glpbio, GC38465, 1 mg/kg), Iron-Dextran (Sigma, D8517, 1 g/kg), 3-MA (MCE, HY-19312, 10 mg/kg), BafA1 (1 mg/kg) and CsA (MCE, HY-B0579, 20 μg/kg) respectively once a week. The mice were euthanized after 3 months treatment. Tail-looing model and adeno-associated virus (AAV) administration After excising the distal tail portion beyond the 17 th coccygeal vertebra and removing the lateral skin of the 8 th coccygeal vertebra, Tail-looping model of the end-to-side anastomosis was established. Then, 1 × 1012 v.g/mL AAV-GFP, AAV-Yap1 OE, AAV-Ncoa4 KD, AAV-Ncoa4 OE or AAV-Zbp1 KD, purchased from Shanghai Genechem, in a 2 μL volume was injected into intervertebral space. The mice were euthanized after 2 months treatment. Intracellular Fe2+ and Ca2+ levels assay Intracellular Fe2+ was stained by FerroOrange Kit (Dojindo, F374). After incubation with 1 μM FerroOrange probe for 30 min, chondrocytes were observed under a laser confocal fluorescence microscope (Nikon, CrestOptics X-Light V3) or collected for flow cytometry analysis (BD Biosciences, FACSCalibur). Similarly, chondrocytes loaded with Fluo-4 AM indicator (Beyotime, S1060) were detected to observe calcium influx. Lipid peroxidation level detection Malondialdehyde (MDA) content was measured using the MDA Assay Kit (Beyotime, S0131) and BCA Assay Kit (Solarbio, PC0020). In accordance with instructions, the chondrocytes lysis buffer was collected to detect the concentration of MDA and BCA via the absorbance at a wavelength of 532 or 562 nm respectively. C11-BODIPY 581/591 (ThermoFisher, D3861) was also used to detect lipid peroxidation. Briefly, chondrocytes were incubated with 10 μM C11-BODIPY probe for 30 min and stained with 1× Hoechst 33,342 (Beyotime, C1028) for 10 min at 37°C. After that, images were captured by the laser confocal fluorescence microscope (Nikon, CrestOptics X-Light V3). Cell transfection The negative control siRNA (si-NC), si-Ncoa4, si-Yap1 and si-Zbp1 (Ribo Bio, China) were used for chondrocytes transfection via Lipofectamine 3000 reagent (ThermoFisher, L3000015). After 72-h of transfection, the chondrocytes were harvested to verify silence efficiency and for further experiments. Cytoskeleton morphology staining For F-actin staining, the chondrocytes were stained with CL488-Phalloidin (Proteintech, PF00001) after fixed and permeabilized. Cytoskeleton was observed using the laser confocal fluorescence microscope (Nikon, CrestOptics X-Light V3). Lysosome staining LysoTracker Green kit (Beyotime, C1047) was used for tracing lysosomes. Images were obtained with a laser confocal fluorescence microscope (Nikon, CrestOptics X-Light V3). Mitochondrial morphology analysis To detect mitochondrial morphology, the chondrocytes were incubated with Mito-Tracker Green probe (Beyotime, C1048) and then Hoechst 33,342 (Beyotime, C1028). Images were obtained with a laser confocal fluorescence microscope (Nikon, CrestOptics X-Light V3). Transmission electron microscopy (TEM) CEP chondrocytes were fixed in 2.5% glutaraldehyde (Solarbio, P1126) and then 1% osmic acid (TED PELLA, 18,456), followed by embedding and dehydration. TEM (Jeol, JEM1400) was used to capture ultrastructural features. Co-immunoprecipitation (co-IP) assay Piezo1fl/fl Col2 Cre+ and Piezo1fl/fl Col2 Cre− CEPCs were subjected to mechanical stimulation at 1 MPa for 24 h. Anti-NCOA4 antibody and control IgG (Beyotime, A7016) were incubated with Protein A/G Magnetic Beads (MCE, HY-K0202) overnight at 4°C. Protein lysates prepared with IP lysis buffer (Beyotime, P0013) were then added to the beads and incubated for another 24 h at 4°C. The denatured protein samples were subsequently analyzed by western blot. Chromatin immunoprecipitation (ChIP) and ChIP-qPCR DNA fragments were immunoprecipitated using the ChIP assay kit (Beyotime, P2078) and next subjected to qPCR assay (Bio-Rad, CFX Duet). The primers were as follows: F1, CCTGCCTTTGGCCTGTTAGA; R1, TCACCCAGCCCTACTGCTCT; F2, TACAAGATTTCCCGTCCCAG; R2, ACATTCACCCAGCCCTACTG; F3, TACAAGATTTCCCGTCCCAG; R3, GAAACATTCACCCAGCCCTA. Transcriptome analysis The total cellular RNA was extracted using AG RNAex Pro RNA reagent (Accurate Biology, AG21101), and the transcriptome sequencing was conducted by Novogene (China). Mechanical loading To simulate physical mechanical stimulation, CEP chondrocytes were cultured in programmable air compress system (Cell and Force, China). Statistical analyses Each experiment was repeated independently at least three times. The difference between two groups was analyzed by Student’s t-test. Multiple group comparison was analyzed by One-way ANOVA. Correlation analysis was performed using Pearson correlation coefficients. All data are performed with GraphPad Prism 9 and expressed as the mean±SD. ns (no significance), *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001. Supplementary Material Supplementary Material.docx