Ferroptosis-Driven Senescence Loop as a Central Amplifier of Osteoarthritis Progression

Abstract

Osteoarthritis (OA) is a prevalent, chronic joint disorder characterized by cartilage degradation, synovial inflammation, and extracellular matrix (ECM) remodeling, yet disease-modifying therapies remain elusive. Emerging evidence implicates ferroptosis, an iron-dependent form of regulated cell death driven by lipid peroxidation, and cellular senescence, characterized by growth arrest and a senescence-associated secretory phenotype (SASP), as central contributors to OA pathogenesis. Ferroptotic chondrocytes release reactive lipid species and damage-associated molecular patterns (DAMPs) that induce paracrine senescence in neighboring cells, while senescent cells amplify oxidative stress and ferroptotic susceptibility, forming a self-perpetuating feed-forward loop that accelerates tissue degeneration. Histological, molecular, and in vivo studies demonstrate iron accumulation, lipid peroxidation, glutathione peroxidase 4 (GPX4) depletion, and SASP factor secretion in human OA cartilage, synovium, and animal models, linking these processes to ECM breakdown and joint inflammation. Targeted interventions, alone or in combination, can disrupt this pathological loop, preserve chondrocyte viability, reduce SASP-mediated inflammation, and mitigate cartilage damage. Integration of biomarker-guided patient stratification, advanced imaging, and spatial transcriptomic profiling may enable precision-targeted, disease-modifying therapies. Therefore, elucidating the crosstalk between ferroptosis and senescence offers a conceptual and translational framework for shifting OA management from symptomatic relief toward preservation of joint integrity and long-term disease modification.

INTRODUCTION

Osteoarthritis (OA) is a highly prevalent, chronic joint disorder and a leading cause of pain, disability, and loss of independence worldwide (Scheuing et al., 2023). Global Burden of Disease analyses estimate that OA affected approximately 595 million people by 2020, with an age-standardized prevalence exceeding 5% across all regions (Xu et al., 2025a). Population aging and growth, along with increasing prevalence of risk factors such as obesity and prior joint injury, suggest that the number of affected individuals could approach one billion by 2050 (Briggs et al., 2026). The disability burden of OA is substantial and rising; age-standardized years lived with disability (YLD) for total OA increased between 1990 and 2020, and OA remains among the leading causes of YLD in older adults (Wu et al., 2025a). Despite its major public health impact, there are currently no widely approved disease-modifying osteoarthritis drugs (DMOADs) capable of reliably halting or reversing structural progression (Scheuing et al., 2023). Although multiple candidate DMOADs and biologic approaches have entered clinical trials, consistent, reproducible disease-modifying effects remain elusive, reflecting the complex, multifactorial nature of OA pathophysiology (Briggs et al., 2026).

OA is a whole-joint disease involving chondrocytes, synoviocytes, subchondral bone cells, immune and stromal cells, extracellular matrix remodeling, mechanical stressors, metabolic factors, and neurogenic pain pathways, all interacting in space- and time-dependent ways (Yao et al., 2023). Key molecular regulators include inflammatory cytokines (IL-1β, TNF-α), matrix metalloproteinases, growth factors, epigenetic modulators, and neuropeptides, while cellular crosstalk within the synovial and cartilaginous microenvironments is increasingly recognized as central to disease initiation and progression (Harjacek, 2021). These multilayered interactions complicate target selection and translational predictability from preclinical models (Honkala et al., 2022). Addressing OA effectively therefore requires a multi-pronged research strategy: improved population surveillance and stratification, mechanistic studies that resolve cell-type and spatial heterogeneity (single-cell RNA sequencing and spatial transcriptomics) (Ajaz et al., 2025), development of better translational models and biomarkers of structural disease activity, and precision-targeted interventions that modulate key nodes in disease networks rather than single downstream mediators (Ripa et al., 2025). Such integrated approaches offer the most promising path toward effective DMOADs and strategies to reduce the projected global OA burden.

Ferroptosis is a unique form of regulated cell death that is mechanistically and morphologically distinct from apoptosis and necroptosis, with growing relevance to OA (Yang et al., 2022). Unlike apoptosis, characterized by caspase activation and chromatin condensation, or necroptosis, involving RIPK1/RIPK3-mediated membrane rupture, ferroptosis is driven by iron-dependent lipid peroxidation (Tkachenko and Havranek, 2025). Excess intracellular iron participates in Fenton reactions, producing highly reactive oxygen species that oxidize polyunsaturated fatty acids within cellular membranes (Su et al., 2019). Chondrocytes in osteoarthritic joints are particularly susceptible due to inflammatory and metabolic stress, disrupted iron homeostasis, and suppression of protective systems such as glutathione peroxidase 4 (GPX4), SLC7A11, and glutathione (Yan et al., 2025). Accumulation of lipid peroxides compromises membrane integrity, leading to cell death without classical apoptotic or necroptotic features (Yu et al., 2024b). Ferroptotic damage not only reduces chondrocyte viability but also releases oxidative byproducts and inflammatory mediators that amplify joint degeneration, making it a potential therapeutic target (Su et al., 2019; Yan et al., 2025).

Cellular senescence is now seen as a key factor in persistent OA inflammation, primarily via the senescence-associated secretory phenotype (SASP) (Lynch et al., 2025). Senescent cells, though permanently growth-arrested, remain metabolically active and secrete pro-inflammatory cytokines (IL-6, IL-8) (Sikora et al., 2014), chemokines, matrix-degrading enzymes (MMPs, ADAMTS) (Rentschler et al., 2022), growth factors (VEGF, TGF-β), and extracellular vesicles (Burton and Faragher, 2015). Chronic activity of NF-κB (Salminen et al., 2012), p38 MAPK, mTOR, and cGAS–STING, triggered by ongoing DNA damage, telomere dysfunction, oxidative stress, and mitochondrial abnormalities, maintains and amplifies SASP (Nadeem et al., 2025). Although temporary SASP can help tissue repair and recruit immune cells, the buildup of senescent cells leads to chronic, low-grade inflammation that accelerates matrix degradation, impairs tissue regeneration, and triggers paracrine senescence in neighboring cells (Xiao et al., 2023; Alqahtani et al., 2025). In joints, persistent SASP drives cartilage breakdown, synovial inflammation, and further disease progression (Han et al., 2024). Strategies to reduce SASP include senolytics, which remove senescent cells, and senomorphics, which inhibit SASP without inducing cell death, both of which offer potential for disease-modifying therapies (Gupta, 2025).

Emerging evidence indicates that ferroptosis and cellular senescence interact to drive joint degeneration. Hallmarks of ferroptosis, iron overload, and lipid peroxidation (Latunde-Dada, 2017) can induce or exacerbate chondrocyte senescence through DNA damage (Dixon and Stockwell, 2019), mitochondrial dysfunction, and activation of the p53/p21 or p16^INK4a pathways (Tarangelo and Dixon, 2018). In contrast, SASP factors disrupt redox homeostasis, increase ROS levels, and alter iron metabolism, thereby sensitizing neighboring cells to ferroptosis (Alqahtani et al., 2025). This bidirectional relationship forms a pathological feedback loop in which ferroptosis accelerates senescence, and SASP amplifies susceptibility to ferroptosis, collectively promoting inflammation and extracellular matrix breakdown (Coradduzza et al., 2023). Preclinical studies support this interplay: pharmacological inhibition of ferroptosis reduces senescence markers, while senolytic or SASP-suppressing interventions decrease lipid peroxidation and ferroptotic signaling (Vetuschi et al., 2022). These findings highlight the mechanistic convergence of ferroptosis and senescence as key drivers of OA progression and suggest the potential of dual-targeted therapies. This review highlights ferroptosis–senescence crosstalk in OA, its contribution to chronic inflammation and cartilage degeneration, and translational opportunities, including therapeutic targets and biomarkers for patient stratification.

PATHOGENIC ROLES OF FERROPTOSIS IN OA

Ferroptosis is a regulated form of cell death characterized by iron-dependent lipid peroxidation and oxidative membrane damage (Yang et al., 2022). Unlike apoptosis or necrosis, ferroptosis results from excess reactive oxygen species (ROS) and disturbed redox balance (Su et al., 2019). In OA, ferroptosis is a key process linking oxidative stress, inflammation, and cartilage loss (Al-Hetty et al., 2023). Its molecular markers include interconnected biochemical, genetic, and structural traits that distinguish it from other cell death pathways, such as apoptosis, necroptosis, or autophagy (Su et al., 2019). Ferroptosis is characterized by iron overload, lipid peroxidation, glutathione (GSH) depletion, GPX4 inactivation, and mitochondrial changes, all of which lead to irreversible oxidative membrane damage and cell death (Yu et al., 2024a).

A key marker of ferroptosis is iron overload, which drives the Fenton reaction and generates reactive hydroxyl radicals (•OH), which initiate lipid peroxidation (Dixon and Stockwell, 2019). Too much intracellular Fe²⁺ often results from faulty iron management, including raised transferrin receptor 1 (TFR1) (Gao et al., 2019a), increased divalent metal transporter 1 (DMT1) activity (Mims and Prchal, 2005), and ferritin breakdown via ferritinophagy by nuclear receptor coactivator 4 (NCOA4) (Xu et al., 2025b). This increases the pool of free iron, elevates oxidative stress, and promotes oxidation of polyunsaturated fatty acid (PUFA)-containing phospholipids (Mortensen et al., 2023).

Lipid peroxidation is another core hallmark of ferroptosis (Dixon and Stockwell, 2019). The process is primarily driven by enzymatic and non-enzymatic oxidation of PUFA-phospholipids, catalyzed by lipoxygenases (ALOXs) and promoted by Fe²⁺ (Latunde-Dada, 2017). The accumulation of lipid hydroperoxides (PUFA–OOH) disrupts cellular membrane structure, culminating in cell rupture (Islam et al., 2023). Acyl-CoA synthetase long-chain family member 4 (ACSL4) and lysophosphatidylcholine acyltransferase 3 (LPCAT3) are essential enzymes that incorporate PUFAs, such as arachidonic acid and adrenic acid, into phospholipids, thereby sensitizing cells to ferroptosis. Consequently, inhibition of ACSL4 or ALOX15 markedly reduces ferroptotic cell death (Kuwata and Hara, 2019).

A key regulatory hallmark of ferroptosis involves the GSH-GPX4 axis, which functions as the primary antioxidant defense system (Dar et al., 2024). GPX4 catalyzes the reduction of lipid hydroperoxides (L–OOH) to non-toxic lipid alcohols (L–OH) using GSH as a cofactor (Forcina and Dixon, 2019). Under ferroptotic conditions, depletion of intracellular GSH or direct inactivation of GPX4 leads to uncontrolled accumulation of lipid peroxides (Latunde-Dada, 2017). The cystine/glutamate antiporter system Xc⁻, composed of SLC7A11 and SLC3A2, plays an essential role in importing cystine for GSH synthesis (Lee and Roh, 2022). Inhibition of system Xc⁻ by agents such as erastin or by p53-mediated repression of SLC7A11 sensitizes cells to ferroptosis by reducing cystine uptake and impairing GSH biosynthesis (Lin et al., 2020).

Another molecular feature of ferroptosis is the suppression of antioxidant defense signaling, particularly through downregulation of nuclear factor erythroid 2-related factor 2 (NRF2) (Yan et al., 2023). NRF2 orchestrates the transcription of multiple antioxidant and iron-regulatory genes, including GPX4, SLC7A11, ferritin heavy chain (FTH1), and heme oxygenase-1 (HO-1) (Kerins and Ooi, 2018). Loss of NRF2 activity compromises cellular redox buffering capacity and enhances susceptibility to ferroptotic damage (Ren et al., 2021). Conversely, pharmacological activation of NRF2 has been shown to confer protection against ferroptosis.

At the ultrastructural level, mitochondrial morphological changes are considered a distinctive morphological hallmark of ferroptosis (Gao et al., 2019b). Ferroptotic cells exhibit shrunken mitochondria, increased membrane density, and reduced or vanished cristae, reflecting impaired oxidative phosphorylation and excessive ROS generation (Liu et al., 2023). Unlike apoptosis, ferroptosis does not involve nuclear condensation, chromatin fragmentation, or caspase activation (Ren et al., 2021).

Taken together, these molecular and morphological hallmarks, such as iron accumulation, lipid peroxidation, GSH depletion, GPX4 inactivation, impaired antioxidant defenses, and mitochondrial damage, define ferroptosis as a unique oxidative form of regulated cell death. Understanding these molecular signatures provides critical insights into how ferroptosis contributes to various pathological processes, including inflammation, neurodegeneration, and degenerative joint diseases such as osteoarthritis.

EVIDENCE FOR FERROPTOSIS IN OA TISSUES

Ferroptosis has emerged as a central pathogenic mechanism in OA, acting as a mechanistic nexus linking iron dysregulation, oxidative stress, chondrocyte death, extracellular matrix (ECM) degradation, and joint inflammation, with evidence spanning human tissues, animal models, and in vitro studies. Histopathologic analyses of human OA cartilage consistently reveal pronounced iron accumulation, particularly in chondrocytes and the superficial cartilage zone, which establishes a pro-ferroptotic environment (Cao et al., 2023). This iron overload catalyzes Fenton reactions, producing ROS that attack polyunsaturated fatty acids in lipid membranes, thereby elevating lipid peroxidation markers such as 4-hydroxynonenal (4-HNE) (Łuczaj et al., 2016) and malondialdehyde (MDA) (Zhang et al., 2023). At the molecular level, OA chondrocytes exhibit reduced expression of GPX4 (Wang et al., 2023a) and the cystine/glutamate antiporter subunit SLC7A11 (Ma et al., 2025), alongside depletion of intracellular GSH (Wang et al., 2023b), compromising their ability to detoxify lipid ROS and resist ferroptotic death. Transcriptomic and proteomic analyses further reveal dysregulation of iron metabolism, including upregulation of ferritin and transferrin receptor expression (Wu et al., 2025b), highlighting both intracellular iron sequestration and enhanced iron uptake. Synovial tissues exhibit similar ferroptotic signatures, with iron overload, increased lipid ROS, and impaired antioxidant defenses correlating with fibroblast activation, inflammation, and ECM remodeling (Huang et al., 2025; Sanchez-Lopez et al., 2022). Collectively, these observations indicate that OA joint tissues are highly susceptible to ferroptosis, which contributes to chondrocyte dysfunction, matrix degradation, and local inflammatory milieu.

Animal models of OA, particularly surgically induced post-traumatic models such as destabilization of the medial meniscus (DMM) in mice, recapitulate the human molecular and histopathologic signatures and provide causal evidence for ferroptosis in disease progression. In these models, chondrocytes demonstrate iron accumulation, elevated lipid peroxidation (4-HNE, MDA), downregulation of GPX4, and concomitant cartilage erosion with proteoglycan loss (Camacho et al., 2016; Culley et al., 2020; Loeser, 2006; Zhang et al., 2022). Experimental manipulation of iron homeostasis underscores the functional significance of ferroptosis: systemic iron loading or intra-articular iron injection exacerbates cartilage degeneration, increases chondrocyte lipid ROS, and upregulates ferroptotic mediators, including acyl-CoA synthetase long-chain family member 4 (ACSL4) and cyclooxygenase-2 (COX2) (Ding et al., 2023; Ru et al., 2024). Conversely, pharmacologic inhibition of ferroptosis using lipophilic antioxidants such as ferrostatin-1 and liproxstatin-1, or iron chelation with deferoxamine, preserves GPX4 expression, reduces lipid peroxidation, prevents chondrocyte death, and mitigates structural cartilage damage (Cao et al., 2023; Cheng et al., 2024; Zhang et al., 2025). These findings establish that ferroptosis is not merely an epiphenomenon but actively drives joint tissue degeneration.

In vitro studies provide mechanistic insights into the molecular pathways linking ferroptosis to ECM degradation and inflammation. Chondrocytes treated with ferroptosis inducers, such as erastin or RSL3, exhibit elevated lipid ROS, suppression of GPX4 and SLC7A11, and upregulation of ACSL4 and COX2, which are essential mediators of lipid peroxidation and ferroptotic signaling (Kuwata and Hara, 2019; Meng et al., 2025; Rosenblum, 2023; Yu et al., 2024a). These molecular alterations directly enhance the activity of ECM-degrading enzymes, including matrix metalloproteinase-13 (MMP-13) and a disintegrin and metalloproteinase with thrombospondin motifs 5 (ADAMTS5), linking ferroptotic signaling to cartilage matrix breakdown. Importantly, ferroptotic chondrocytes release lipid peroxidation products and damage-associated molecular patterns (DAMPs) that propagate oxidative stress and inflammatory signals to neighboring chondrocytes and synovial fibroblasts, establishing a self-amplifying feed-forward loop that exacerbates ferroptotic susceptibility, ECM degradation, and inflammation (Guan et al., 2024; Li et al., 2025). This cycle is further reinforced in subchondral osteoblasts, where iron-induced oxidative stress promotes the release of pro-inflammatory cytokines, alters bone remodeling, and contributes to changes in the joint microenvironment that accelerate cartilage degeneration.

The interplay between ferroptosis and ECM remodeling is bidirectional: lipid ROS generated during ferroptotic cell death not only triggers chondrocyte demise but also activates MMPs and ADAMTS5, accelerating collagen and aggrecan degradation, which in turn exposes remaining chondrocytes to additional oxidative stress and ferroptotic triggers (Huang et al., 2025; Yang et al., 2023). Moreover, ferroptosis in synovial fibroblasts and infiltrating immune cells amplifies local inflammatory signaling, enhancing the secretion of cytokines such as IL-1β and TNF-α, which further sensitize chondrocytes to oxidative damage and ECM catabolism. This integrative network, encompassing chondrocytes, synovial fibroblasts, and subchondral osteoblasts, creates a vicious cycle in which ferroptosis, oxidative stress, ECM degradation, and inflammation mutually reinforce one another, ultimately driving progressive joint destruction in OA.

Collectively, convergent evidence from human tissue analyses, preclinical models, and mechanistic in vitro studies establishes that ferroptosis is a central pathogenic process in OA, linking iron-dependent oxidative stress to structural joint degeneration. These insights underscore the potential of targeting ferroptotic pathways as a therapeutic strategy: interventions that restore iron homeostasis, enhance antioxidant defenses, or inhibit lipid peroxidation can disrupt the feed-forward loop, preserve chondrocyte viability, attenuate ECM breakdown, reduce local inflammation, and potentially slow OA progression (Cao et al., 2023; Cheng et al., 2024; Zhang et al., 2025). By positioning ferroptosis at the intersection of cellular metabolism, oxidative damage, and matrix catabolism, this integrative framework highlights its critical role in OA pathogenesis and its promise as a target for disease-modifying therapeutic interventions aimed at maintaining joint integrity and function.

SENESCENCE AND SASP IN OA

Cellular senescence in osteoarthritic joints represents a complex, multilayered biological process in which chondrocytes and synovial cells accumulate diverse forms of molecular damage and stress, ultimately adopting stable yet dysfunctional phenotypes that profoundly disrupt tissue homeostasis and drive the chronic, degenerative nature of osteoarthritis (Mobasheri et al., 2015). A defining hallmark of these senescent cells is the robust upregulation of the cyclin-dependent kinase inhibitors p16^INK4a and p21^CIP1, which impose irreversible cell-cycle arrest and prevent the proliferation of damaged cells, thereby limiting the joint’s intrinsic regenerative potential and contributing to the progressive depletion of functionally competent cell populations (Safwan-Zaiter et al., 2022; Stein et al., 1999; Romanov et al., 2012). This growth arrest is further reinforced by telomere attrition, a consequence of cumulative replication and heightened oxidative stress, that activates DNA damage responses and entrenches the senescent phenotype through replicative exhaustion (Von Zglinicki, 2000). Parallel to these cell-cycle and genomic alterations, senescent chondrocytes and synovial cells undergo profound metabolic reprogramming, characterized by enhanced glycolytic flux, mitochondrial impairment, and shifts in lipid and amino acid metabolism that promote cellular survival under persistent stress while fueling SASP synthesis (June et al., 2016). These metabolic changes not only amplify intracellular oxidative stress but also reshape the local microenvironment into a pro-inflammatory and pro-catabolic niche conducive to the progressive accumulation of dysfunctional cells (Ząbek et al., 2025).

Central to the pathological influence of senescence is the SASP, an expansive secretory program through which senescent chondrocytes and synoviocytes release numerous factors, including pro-inflammatory cytokines (IL-6 and IL-8) (Ortiz-Montero et al., 2017), chemokines (CXCL family members) (Takikawa et al., 2022), and matrix-degrading enzymes (MMPs and ADAMTS proteases) (Kim et al., 2022). Together, these mediators destabilize cartilage ECM homeostasis, exacerbate synovial inflammation, accelerate cartilage breakdown, and perpetuate disease progression by inducing senescence in surrounding cells, ultimately forming a self-reinforcing degenerative loop within the joint (Han et al., 2024). The scale, potency, and composition of SASP are tightly governed by upstream senescence hallmarks, including p16^INK4a/p21^CIP1-driven cell-cycle arrest, telomere shortening, persistent DNA damage signaling, and metabolic rewiring, which collectively potentiate the hypersecretory state characteristic of senescent cells (Romanov et al., 2012; Safwan-Zaiter et al., 2022). However, recent advances in single-cell transcriptomics and integrated multi-omics have revealed that senescence in osteoarthritis is not a monolithic state but instead comprises a spectrum of heterogeneous senescent subpopulations with unique transcriptional identities, stress sensitivities, metabolic dependencies, and functional contributions to disease (Li et al., 2025). Within this heterogeneous landscape, certain senescent subtypes predominantly express inflammatory cytokines (IL-6 and IL-8), thereby driving localized inflammatory amplification, whereas others specialize in ECM remodeling and degradation through high levels of MMPs and ADAMTS (Kim et al., 2022; Takikawa et al., 2022). Additional subpopulations exhibit distinct metabolic signatures, ranging from heightened glycolysis to pronounced mitochondrial dysfunction or altered lipid metabolism, that shape the magnitude and nature of their SASP output (Han et al., 2024; Li et al., 2025). These heterogeneous senescent populations interact dynamically within the joint, collectively creating a multi-dimensional inflammatory and catabolic microenvironment in which different senescent subsets reinforce one another’s pathological activity and contribute to the progressive remodeling, degeneration, and functional decline characteristic of osteoarthritis.

The growing recognition of this extensive cellular and functional heterogeneity fundamentally reshapes our understanding of senescence in OA, demonstrating that senescent chondrocytes and synovial cells differ not only in their secretory repertoires but also in their pathogenic roles, persistence, and responses to stress or therapeutic intervention (Romanov et al., 2012; Safwan-Zaiter et al., 2022). This complexity highlights why senescence is such a potent driver of OA progression: it simultaneously impairs tissue regeneration, enhances oxidative and inflammatory stress, disrupts ECM architecture, and sustains a chronic disease state through feedback loops that operate across cellular, tissue, and microenvironmental levels. Importantly, uncovering the molecular diversity and interdependence of these senescent subsets offers compelling opportunities for selective therapeutic strategies such as precision senolytics, which eliminate specific senescent subtypes, and senomorphics, which modulate SASP composition without removing senescent cells, providing a rational foundation for next-generation interventions aimed at halting or reversing osteoarthritis progression (Fig. 1).

Fig. 1.

Senescence and senescence-associated secretory phenotype (SASP) in osteoarthritis. SASP (senescence-associated secretory phenotype), OA (osteoarthritis), ECM (extracellular matrix), MMPs (matrix metalloproteinases), ADAMTS (a disintegrin and metalloproteinase with thrombospondin motifs).

FERROPTOSIS–SENESCENCE CROSSTALK: MECHANISTIC FRAMEWORK OF THE FEED-FORWARD LOOP

Ferroptotic chondrocytes act as potent amplifiers of joint damage by releasing lipid peroxidation products and DAMPs that can induce paracrine stress responses in neighboring cells (Zhang et al., 2022). Excess intracellular iron, a characteristic feature of osteoarthritic chondrocytes, catalyzes Fenton reactions that generate ROS, which attack polyunsaturated fatty acids in cellular membranes and produce highly reactive lipid species, including malondialdehyde (MDA), 4-hydroxynonenal (4-HNE), and oxidized phospholipids (Rentschler et al., 2022). Progressive membrane damage during ferroptosis facilitates the extracellular release of lipid peroxidation products, along with DAMPs such as HMGB1, ATP, heat shock proteins, and extracellular vesicles enriched in oxidized lipids, thereby creating a pro-oxidative and pro-inflammatory microenvironment (Romanov et al., 2012).

Neighboring chondrocytes and synovial fibroblasts sense these signals via pattern recognition receptors and redox-sensitive signaling pathways, leading to DNA damage, mitochondrial dysfunction, and activation of canonical stress-response cascades, including NF-κB, p53/p21, and p16^INK4a/Rb (Romanov et al., 2012; Safwan-Zaiter et al., 2022). Activation of these pathways promotes a senescence-associated phenotype characterized by stable cell-cycle arrest and acquisition of a SASP. The SASP includes pro-inflammatory cytokines (IL-6, IL-8), chemokines, and matrix-degrading enzymes such as MMPs and ADAMTS, which further exacerbate oxidative stress, inflammation, and extracellular matrix degradation within the joint (Ortiz-Montero et al., 2017; Rentschler et al., 2022). Dysregulation of iron-handling proteins, such as increased transferrin receptor–mediated iron import, reduced ferritin storage capacity, and impaired ferroportin-mediated export, further sustains intracellular iron accumulation, reinforcing redox imbalance and sensitizing cells to both ferroptosis and senescence (Levi et al., 2024). Iron-driven ROS production also enhances SASP factor expression in neighboring cells, amplifying tissue inflammation and matrix degradation (Wang et al., 2023a).

Support for this mechanistic framework is observed across multiple degenerative disease settings. In fibrotic disorders of the liver and lung, iron accumulation and lipid peroxidation coincide with ferroptotic cell death and senescent fibroblast activation, collectively driving chronic inflammation and tissue remodeling (Nadeem et al., 2025; Salminen et al., 2012). In neurodegenerative diseases, such as Parkinson’s and Alzheimer’s disease, iron-dependent oxidative stress promotes neuronal ferroptosis, while surrounding glial cells exhibit senescence-associated inflammatory phenotypes that exacerbate neuronal dysfunction (Mezzanotte and Stanga, 2024). Across these contexts, a recurring pattern emerges in which iron-mediated oxidative stress initiates ferroptosis, triggers paracrine senescence, and establishes a self-reinforcing cycle of tissue injury (Levi et al., 2024). Together, these observations provide a mechanistic rationale for a ferroptosis–senescence feed-forward loop, highly relevant to OA pathogenesis (Fig. 2).

Fig. 2.

Ferroptosis–senescence feed-forward loop in osteoarthritis. DAMPs (damage-associated molecular patterns), 4-HNE (4-hydroxynonenal), MDA (malondialdehyde), ROS (reactive oxygen species), ECM (extracellular matrix), MMPs (matrix metalloproteinases), ADAMTS (a disintegrin and metalloproteinase with thrombospondin motifs).

FERROPTOSIS–SENESCENCE FEED-FORWARD LOOP: OA-SPECIFIC EVIDENCE

Accumulating evidence from experimental OA models indicates that the interaction between ferroptosis and cellular senescence is not merely correlative but causally contributes to disease progression (Safwan-Zaiter et al., 2022). In surgically induced OA models, including DMM, chondrocytes exhibit iron accumulation, increased lipid peroxidation, and reduced expression of the lipid peroxide detoxifying enzyme GPX4, coinciding with cartilage erosion and proteoglycan loss (Fang et al., 2018; Meng et al., 2025). Experimental exacerbation of iron overload, through systemic iron loading or intra-articular iron administration, accelerates cartilage degeneration and enhances ferroptotic signaling, establishing iron-dependent lipid peroxidation as a driver rather than a bystander of OA progression (Xu et al., 2023; Yao et al., 2021). Conversely, pharmacological inhibition of ferroptosis provides direct evidence for causality. Treatment of OA models with lipophilic radical-trapping agents such as ferrostatin-1 or liproxstatin-1, as well as iron chelation using deferoxamine, preserves GPX4 activity, reduces lipid ROS accumulation, prevents chondrocyte death (Guo et al., 2022), and significantly attenuates structural joint damage in the DMM mouse model of OA (Cheng et al., 2024; Yao et al., 2021). These interventional studies demonstrate that suppression of ferroptosis is sufficient to mitigate OA pathology, confirming a causal role for ferroptotic signaling in joint degeneration. Beyond inducing chondrocyte death, ferroptotic stress has been shown to actively promote cellular senescence in OA-relevant contexts (Wang et al., 2023a). In vitro and ex vivo studies using primary chondrocytes demonstrate that ferroptosis induction by erastin or RSL3 leads to excessive lipid ROS accumulation and release of DAMPs, which activate p53/p21, p16^INK4a/Rb, and NF-κB signaling pathways (Jiaxin et al., 2025; Wang et al., 2023a). Activation of these pathways results in stable growth arrest and acquisition of a senescence-associated secretory phenotype. Importantly, ferroptotic chondrocytes induce paracrine senescence in neighboring chondrocytes and synovial fibroblasts, providing functional evidence that ferroptotic injury can spatially propagate senescence within the osteoarthritic joint.

Reciprocal causality is observed in the reverse direction, whereby senescence enhances susceptibility to ferroptosis in OA tissues (Sun et al., 2018). Senescent chondrocytes and synovial cells exhibit pronounced metabolic and redox alterations, including intracellular iron accumulation, impaired glutathione metabolism, reduced GPX4 activity, and elevated lipid peroxidation (Liao et al., 2022). These changes sensitize senescent cells and adjacent non-senescent cells exposed to the SASP to ferroptotic death. Functional support for this relationship is provided by OA models treated with senolytic or senomorphic agents, in which clearance of senescent cells or suppression of SASP signaling reduces oxidative stress, limits iron-driven lipid peroxidation, and attenuates cartilage degradation (Chin et al., 2023; Wang et al., 2022). Taken together, these OA-specific interventional studies establish a bidirectional causal relationship between ferroptosis and senescence. Ferroptotic stress induces senescence through lipid peroxidation– and DAMP-mediated signaling, while senescent cells exacerbate iron dysregulation and oxidative stress, thereby increasing ferroptotic vulnerability (Liao et al., 2022). This reciprocal reinforcement forms a feed-forward loop that accelerates extracellular matrix degradation, synovial inflammation, and cartilage erosion (Cheng et al., 2024). Framing ferroptosis–senescence crosstalk as a causally supported mechanism, rather than a descriptive association, provides a robust foundation for therapeutic strategies aimed at disrupting this loop to achieve disease modification in osteoarthritis.

THERAPEUTIC IMPLICATIONS

Ferroptosis represents a central mechanism driving chondrocyte death and OA progression, making it a promising therapeutic target (Yang et al., 2022). Pharmacological strategies aimed at inhibiting ferroptosis include activation of GPX4, which enhances cellular antioxidant defenses and directly inhibits lipid peroxidation, thereby protecting chondrocytes from ferroptotic cell death (Gao et al., 2025). Small-molecule radical-trapping agents, such as liproxstatin-1 and ferrostatin-1 (Table 1) (Cheng et al., 2024; Yao et al., 2021), prevent the propagation of lipid peroxyl radicals in cellular membranes, mitigating oxidative damage, while iron chelators, including deferoxamine and deferiprone, reduce intracellular labile iron pools, limiting Fenton reaction–mediated ROS production and subsequent lipid peroxidation (Guo et al., 2022; Liu et al., 2024). By preserving chondrocyte viability and reducing DAMP release, these interventions can attenuate paracrine senescence induction, potentially disrupting the ferroptosis–senescence feed-forward loop and slowing cartilage degeneration. Targeting cellular senescence offers a complementary approach to ferroptosis inhibition. Senolytic agents, such as Dasatinib combined with Quercetin or Navitoclax, selectively induce apoptosis in senescent cells, thereby reducing the burden of the SASP and its pro-inflammatory effects (Table 1) (Yang et al., 2020; Zhao et al., 2025). By eliminating senescent chondrocytes and synovial fibroblasts, these therapies can mitigate paracrine senescence, decrease ECM degradation, and alleviate local joint inflammation, whereas senomorphic agents, including p38 MAPK inhibitors, suppress SASP factor secretion without inducing cell death (Table 1), limiting the paracrine propagation of senescence and lowering the inflammatory milieu. Preclinical studies indicate that both senolytic and senomorphic interventions can reduce cartilage damage, preserve tissue homeostasis, and disrupt the ferroptosis–senescence feed-forward loop, highlighting their potential as disease-modifying therapies (Wang et al., 2022). Given the interdependent roles of ferroptosis and senescence in OA, combination strategies that simultaneously target both processes are increasingly recognized as promising. Co-administration of ferroptosis inhibitors, such as GPX4 activators (e.g., liproxstatin-1 or ferrostatin-1), with senolytic agents, such as Dasatinib plus Quercetin or Navitoclax, can synergistically disrupt the feed-forward loop, protecting chondrocytes from death while reducing SASP-mediated paracrine senescence (Fu et al., 2025b) (Table 1). Localized delivery methods, such as intra-articular injections, hydrogel-based depots, or tissue-targeted nanoparticles, achieve high local drug concentrations within the joint while minimizing systemic exposure and off-target toxicities (Kasiński et al., 2020). Temporal or sequential delivery strategies, where ferroptosis inhibition precedes senolytic treatment, may further optimize cellular protection by first stabilizing chondrocytes and then clearing existing senescent cells, breaking the self-perpetuating loop of joint degeneration (Fan et al., 2024b). Integration of these combination and localized therapies with biomarker-guided patient stratification, using circulating lipid peroxidation products (MDA, 4-HNE, oxidized phospholipids), iron metabolites (ferritin, transferrin saturation, labile plasma iron), and SASP-associated cytokines and chemokines (IL-6, IL-8, MCP-1, TNF-α), may enhance therapeutic efficacy and enable personalized treatment regimens (Yang et al., 2022). Additional emerging markers, such as extracellular vesicle–associated DAMPs and microRNA signatures linked to oxidative stress or senescence, could further refine patient stratification. Longitudinal monitoring of these biomarkers may serve as early indicators of therapeutic efficacy, allowing dynamic adjustment of treatment strategies and improving the translational potential of disease-modifying interventions in OA (Coleman et al., 2025) (Fig. 3).

Table 1.

Therapeutics strategies for OA

Therapeutic strategy	Drug/Agent	Dose/Concentration	Mechanism of action	Reference(s)
Ferroptosis inhibition	Ferrostatin-1	In vitro: 1 µM In vivo: 0.1-1 mg/kg	Prevents IL-1β or iron overload–induced chondrocyte death; restores collagen II, reduces MMP13; attenuates cartilage degradation in OA mouse model	Yao et al., 2021
	Liproxstatin-1	In vivo: 1 mg/kg	Prevents ferroptotic changes (reduced GPX4, increased ACSL4/p53) in cartilage; ameliorates cartilage degradation in TMJ-OA models	Cheng et al., 2024
	Deferoxamine	In vitro: 50-200 µM In vivo: 50-100 mg/kg	Reduces IL-1β/erastin-induced ROS, lipid ROS, MDA; preserves collagen II and suppresses MMP13; delays cartilage degeneration in DMM mouse OA model	Guo et al., 2022
	Baicalin	In vitro: 20 μM	Inhibits ferroptosis and ECM degradation	Liu et al., 2024
Senolytic therapy	Dasatinib + Quercetin	In vivo: 5+50 mg/kg	Selective clearance of senescent cells → reduces SASP burden	Zhao et al., 2025
	Navitoclax	In vivo: 0.25-5.0 mg/kg	Induces apoptosis in senescent cells; clearance of senescent chondrocytes or synovial fibroblasts → reduced local inflammation & ECM degradation	Yang et al., 2020
Senomorphic therapy	p38 MAPK inhibitors (SB203580, Losmapimod)	-	Suppresses SASP secretion without inducing cell death; decreased secretion of pro-inflammatory cytokines, reduced ECM degradation	Wang et al., 2022
Combination/Localized therapy	Ferroptosis inhibitor (Lip-1)+Senolytic (D+Q)	-	Dual inhibition of ferroptosis and clearance of senescent cells → disrupt feed-forward loop; synergistic protection of cartilage, reduced ECM breakdown, dampened inflammation	Fu et al., 2025a

Fig. 3.

Therapeutic strategies for osteoarthritis. GPX4 (glutathione peroxidase 4), MDA (malondialdehyde), ROS (reactive oxygen species), ACSL4 (long-chain-fatty-acid—coA ligase 4), SASP (senescence-associated secretory phenotype), OA (osteoarthritis), ECM (extracellular matrix), MMPs (matrix metalloproteinases).

CHALLENGES AND FUTURE DIRECTIONS

Although the ferroptosis–senescence feed-forward loop is well-supported by preclinical data, its precise role in human OA remains to be fully elucidated. Longitudinal studies in early-stage OA cohorts are critical to determine whether ferroptotic cell death and paracrine senescence act as initiators of cartilage degradation, or primarily exacerbate established disease (Yang et al., 2025). Such studies will provide insight into disease heterogeneity, identify patient subpopulations with high ferroptotic or senescent burden, and define optimal therapeutic windows (Liu et al., 2025). Coupling molecular profiling with clinical outcomes, including pain scores, joint function, and imaging-based cartilage assessment, will facilitate the development of predictive models that guide personalized treatment strategies (Khan et al., 2025). Additionally, multi-center cohort studies can help capture population-level variability in ferroptosis–senescence dynamics, supporting the generalizability of findings.

Translating ferroptosis- and senescence-targeted therapies from bench to bedside is challenging because achieving sustained, effective drug concentrations within the joint while avoiding systemic toxicity is difficult (Fu and Zhou, 2025). Cartilage’s avascular nature, the complex synovial microenvironment, and the rapid clearance of intra-articular agents limit the efficacy of conventional delivery methods (Huang et al., 2022). Emerging strategies, such as biodegradable hydrogels, sustained-release nanoparticles, liposomes, and intra-articular depots, offer promise but require careful optimization of drug release kinetics, tissue penetration, and biocompatibility (Lee et al., 2025b). Combining these approaches with molecular imaging or reporter systems could allow real-time monitoring of drug distribution and retention, further enhancing precision and safety (Willmann et al., 2008). Moreover, the development of minimally invasive, repeatable delivery systems will be crucial for chronic treatment regimens.

A deeper understanding of the ferroptosis–senescence axis in human joints will require integration of spatial multi-omics with advanced clinical imaging. Spatial transcriptomics, proteomics, metabolomics, and single-cell sequencing allow detailed mapping of ferroptotic and senescent cell populations, revealing cellular heterogeneity, localized microenvironmental stress, and cell–cell communication patterns (Wu et al., 2025b). When combined with MRI, PET, or emerging molecular imaging modalities, these approaches can link molecular signatures to structural and functional joint changes, enabling patient-specific therapeutic planning (Weissleder et al., 2016). Such integrative strategies will also allow dynamic monitoring of treatment responses, helping to determine whether interventions effectively disrupt the feed-forward loop, reduce SASP burden, or restore cartilage homeostasis (Larbie and Raza, 2024).

Manipulating ferroptosis and senescence in patients raises significant ethical and safety concerns (Liu et al., 2025). Both pathways are integral to normal tissue homeostasis, immune surveillance, and tumor suppression; inappropriate modulation could result in unintended systemic effects, increased tumorigenic risk, or interference with normal regenerative processes (Coradduzza et al., 2023). Rigorous preclinical safety testing, careful dose optimization, and long-term follow-up in clinical trials are essential. Ethical frameworks must also address equitable patient selection, informed consent, and transparent communication of risks versus benefits (Segun, 2024). Furthermore, societal and regulatory considerations, such as access to emerging biologics or nanoparticle-based therapies, must be carefully navigated to ensure responsible translation (Desai et al., 2025) (Fig. 4).

Fig. 4.

Integrated multi-omics and translational strategies for precision management of osteoarthritis. SASP (senescence-associated secretory phenotype), EV (extracellular vesicle), MRI (magnetic resonance imaging), PET (positron emission tomography).

Addressing these challenges requires interdisciplinary collaboration, combining molecular biology, bioengineering, imaging, pharmacology, and clinical expertise (Sun et al., 2011). Standardization of biomarkers, including circulating lipid peroxidation products, iron metabolites, SASP cytokines, and extracellular vesicle signatures, will support patient stratification and therapy monitoring (Lelièvre et al., 2020). Advancing delivery technologies, integrating spatial multi-omics with imaging, and incorporating ethical oversight will collectively facilitate safe, precise, and personalized interventions (Mohr et al., 2024). Ultimately, overcoming these hurdles has the potential to transform osteoarthritis management, shifting from symptom relief to true disease modification by targeting the ferroptosis–senescence axis, preserving cartilage integrity, and slowing or halting joint degeneration in a patient-specific manner (Fan et al., 2024a).

CONCLUSION

Ferroptosis-driven senescence is increasingly recognized as a central mechanism in OA, forming a self-amplifying feed-forward loop that accelerates cartilage degeneration, extracellular matrix breakdown, and synovial inflammation. Ferroptotic chondrocytes release reactive lipid species, DAMPs, and inflammatory mediators, which induce paracrine senescence in neighboring chondrocytes and synovial fibroblasts. This perpetuates oxidative stress and generates a pro-inflammatory microenvironment, further sensitizing cells to ferroptotic death. The resulting cycle amplifies tissue damage and highlights the close interplay between cell death and senescence in disease progression. Advances in single-cell and spatial multi-omics have revealed heterogeneity among senescent and ferroptotic cell populations, identifying discrete joint microdomains that serve as focal points of degeneration and inflammation. These insights offer a rationale for precision-targeted therapies, including ferroptosis inhibitors, senolytics, and combination strategies, which can be localized to affected regions to enhance efficacy while reducing systemic side effects. Integrating mechanistic understanding with biomarker-guided patient stratification, advanced imaging, and innovative delivery platforms provides a pathway toward disease-modifying interventions. Targeting the ferroptosis–senescence axis thus holds promise for shifting OA treatment from symptomatic management to preservation of joint integrity and long-term functional outcomes.