Skip to content
Frontiers
Search

Review

Frontiers in Molecular Biosciences ·

Volume 13 - 2026 | https://doi.org/10.3389/fmolb.2026.1973436

Lipid droplet–mitochondria crosstalk in osteoarthritis: linking lipid routing to metabolic mismatch

  • Gang Meng1
  • Ruijie Hou1
  • Ping Dong1*corresponding author
  • Qiannan Chen1
  • Xueyuan Li2
  • Kai Hu1
  • Zixuan Du1
  • Rong Wang3
  • Zhaoxin Zhong1
  1. 1Inner Mongolia Medical University, Hohhot, Inner Mongolia, China
  2. 2Changchun University of Chinese Medicine, Changchun, Jilin, China
  3. 3Shanxi University of Traditional Chinese Medicine, Jinzhong, Shanxi, China
Article metrics0 viewsView details
PDFCiteMetrics

Abstract

Osteoarthritis (OA) is a whole-joint disease characterized by metabolic reprogramming beyond structural degeneration. Increasing evidence indicates that lipid droplets (LDs) regulate lipid storage, mobilization and intracellular lipid routing, while alterations in mitochondrial metabolism determine the capacity of chondrocytes to adapt to metabolic stress. However, LD accumulation alone is insufficient to define metabolic dysfunction, as it may represent either adaptive buffering or maladaptive retention. This review proposes a lipid flux–mitochondrial capacity matching framework to integrate lipid handling with mitochondrial oxidative capacity, redox regulation and quality control. We discuss how excessive lipid flux relative to mitochondrial capacity may promote oxidative stress, lipid peroxidation, ferroptosis, senescence and extracellular matrix disruption, whereas balanced lipid utilization supports chondrocyte homeostasis. Current evidence on LD biology, fatty acid oxidation, mitophagy and whole-joint metabolic communication is summarized, with clear distinction between direct OA evidence and findings from other cellular systems. Although organelle contact sites such as PLIN5–FATP4, PLIN3–VDAC3 and ESYT1/2–VAPB represent potential regulators of lipid routing, their functional relevance in OA chondrocytes remains to be established. Future studies integrating isotope tracing, metabolic flux analysis and dynamic organelle imaging are required to clarify how lipid routing and mitochondrial capacity jointly regulate OA progression.

Keywords

1 Introduction

Osteoarthritis (OA) is a heterogeneous whole-joint disease involving articular cartilage, synovium, subchondral bone and associated stromal and immune compartments (Wang et al., 2022; Tang et al., 2024; Raut et al., 2025). Although traditionally considered a degenerative disorder driven by mechanical stress and extracellular matrix breakdown, increasing evidence indicates that OA progression is accompanied by extensive metabolic reprogramming, including alterations in lipid metabolism, mitochondrial function and cellular stress adaptation. Lipid dysregulation in OA therefore extends beyond simple lipid accumulation and involves coordinated changes in lipid uptake, synthesis, storage, mobilization and utilization. However, how intracellular lipid handling is integrated with mitochondrial metabolic capacity to determine chondrocyte fate remains insufficiently understood.

Lipid droplets (LDs) represent dynamic intracellular organelles that regulate lipid storage and mobilization rather than passive lipid deposits. Recent studies have demonstrated that OA-associated LD accumulation may arise from distinct mechanisms, including impaired lipid droplet trafficking and lysosomal clearance under cellular stress (Yu et al., 2025), enhanced fatty acid synthesis and altered lipid metabolic programming (Mei et al., 2025), and broader remodeling of intracellular lipid composition (Xu K. et al., 2026). Importantly, LD accumulation alone cannot determine whether cells undergo metabolic adaptation or lipotoxic injury, because LD formation may serve as a protective buffering mechanism by temporarily sequestering excess fatty acids, whereas defective turnover or inappropriate lipid routing may contribute to pathological stress. Therefore, understanding OA-associated lipid abnormalities requires consideration of lipid composition, turnover and metabolic fate in addition to LD abundance.

The functional consequence of altered lipid availability depends largely on mitochondrial capacity, including oxidative substrate utilization, redox buffering and mitochondrial quality-control mechanisms. Fatty acids mobilized from intracellular lipid pools may undergo mitochondrial oxidation, re-esterification, membrane lipid synthesis or signaling pathways, and their biological effects are determined by whether mitochondrial processing capacity can accommodate metabolic demand. In OA chondrocytes, enhanced fatty acid oxidation promotes acetyl-CoA/HADHA-dependent metabolic reprogramming, resulting in SOX9 destabilization and cartilage catabolic changes (Mei et al., 2025). Conversely, restoration of mitochondrial quality control through HCAR2/AMPK/PINK1/Parkin-mediated mitophagy alleviates cartilage degeneration (Zhuang et al., 2024). These findings suggest that mitochondrial dysfunction in OA should not simply be defined as reduced mitochondrial activity, but rather as impaired metabolic flexibility and insufficient adaptation to changing substrate availability.

Recent studies have further highlighted potential functional connections between LD metabolism and mitochondrial utilization. In non-chondrocyte systems, LD–mitochondria contact sites mediated by proteins such as PLIN5–FATP4 have been shown to facilitate fatty-acid transfer toward mitochondrial β-oxidation (Miner et al., 2023). However, whether similar contact structures exist, and whether they contribute to lipid routing in OA chondrocytes, remains unknown. Thus, LD–mitochondria interactions should be considered as a potential regulatory mechanism requiring direct validation in OA models rather than an established disease mechanism. Meanwhile, impaired mitochondrial renewal and ferroptosis-related lipid remodeling have emerged as important contributors to OA progression (Cao S. et al., 2025; Deng et al., 2025), further supporting the concept that lipid handling and mitochondrial competence are functionally interconnected.

In this review, we propose a lipid flux–mitochondrial capacity matching framework to integrate lipid storage, mobilization, intracellular routing and mitochondrial metabolic regulation in OA. Rather than interpreting LD accumulation or fatty acid oxidation as inherently pathological or protective, this framework emphasizes the balance between lipid supply and mitochondrial processing capacity. We distinguish direct OA evidence from mechanistic findings obtained in other cellular systems and discuss how metabolic mismatch may contribute to oxidative stress, lipid peroxidation, ferroptotic susceptibility, cellular senescence and extracellular matrix disruption. This perspective provides a conceptual framework for understanding how adaptive lipid metabolism may transition into metabolic dysfunction during OA progression.

2 Literature search and evidence selection

Relevant literature was identified through targeted searches of PubMed and Web of Science Core Collection, focusing primarily on studies published between January 2021 and August 2026. Earlier landmark studies were additionally considered when they provided fundamental mechanistic insights into lipid droplet biology, mitochondrial regulation, fatty acid metabolism, or organelle communication.

The search strategy combined controlled vocabulary and free-text terms related to osteoarthritis, lipid metabolism, and mitochondrial biology. Representative terms included (osteoarthritis OR OA) AND (lipid droplet OR lipid trafficking OR lipidomic OR fatty acid oxidation OR cholesterol OR oxysterol) AND (mitochondrial respiration OR mitochondrial metabolism OR mitophagy OR organelle contact site). Additional searches incorporated related concepts, including ferroptosis, senescence, stable isotope tracing, and metabolic flux analysis, to capture emerging mechanisms linking lipid handling with mitochondrial adaptation.

This review considered evidence from multiple experimental contexts, including human OA tissues, primary and established chondrocyte models, cartilage explants, experimental OA models, and defined non-chondrocyte systems. Findings from human tissues and OA-related models were considered direct evidence for disease relevance, whereas studies from other cellular systems, including fibroblasts, macrophages, hepatocytes, myocytes, and mesenchymal stromal cells, were interpreted as mechanistic references that may inform hypotheses but require validation in OA-specific contexts. Particular emphasis was placed on studies providing functional evidence through genetic or pharmacological manipulation, metabolic flux analysis, isotope tracing, or direct assessment of mitochondrial activity, while descriptive studies were used to support biological context and emerging concepts. Studies were screened based on title/abstract relevance followed by full-text assessment. Studies lacking mechanistic relevance to lipid handling, mitochondrial metabolism, or OA-associated metabolic regulation were excluded.

3 Lipid droplet homeostasis in osteoarthritis: from adaptive buffering to maladaptive retention

OA-associated LD accumulation is mechanistically heterogeneous, arising from increased lipid supply, enhanced de novo lipogenesis (DNL) or defective lipid handling (Figure 1). Importantly, OA-associated lipid remodeling extends beyond changes in LD abundance. Recent lipidomic analyses of human osteoarthritic cartilage have revealed disease-associated alterations across multiple lipid classes, including glycerolipids, cholesteryl esters, ceramides, sphingolipids and other complex lipid species, demonstrating that OA is characterized by broad remodeling of intracellular lipid composition rather than simple neutral lipid accumulation (Zhou et al., 2025). These findings highlight the importance of considering lipid composition, turnover and metabolic routing alongside LD abundance when interpreting OA-associated lipid alterations. In obesity-associated mouse cartilage, glucose and insulin drive ACC-dependent DNL (Liu et al., 2022), and loss of the PPARα–ACOT12 axis increases acetyl-CoA availability, promotes DNL and causes LD accumulation and cartilage catabolism in human OA chondrocytes and experimental OA (Park et al., 2022). LD changes in OA are therefore not a simple consequence of extracellular lipid supply.

Dynamic dysregulation of lipid droplet homeostasis in osteoarthritis chondrocytes.
Figure 1. Dynamic dysregulation of lipid droplet homeostasis in osteoarthritis chondrocytes.

LD abundance is also regulated at the LD surface. FGF8–FGFR1–p38 signalling upregulates PLIN1 and increases LD accumulation in cultured chondrocytes (Huang et al., 2025), indicating that OA-relevant signalling alters the capacity to package neutral lipids, although the in vivo contribution of the FGF8–PLIN1 axis remains untested.

LD accumulation alone cannot indicate lipotoxicity or protection from it. LDs transiently buffer free fatty acids, and in high-fat-diet-associated OA mice and cultured articular chondrocytes, PKCK2–STAMP2–FSP27-dependent LD formation confers resistance to free fatty acid-induced lipotoxicity (Lee et al., 2018). This protection is not universal: metabolic tracing in human and murine OA articular chondrocytes shows simultaneous fatty-acid synthesis and oxidation (Wang et al., 2025a). Because distinct pools can be routed to storage, membrane synthesis or oxidation, the LD–FAO relationship requires direct inter-compartmental tracing. LD abundance alone is thus insufficient to define the underlying metabolic state, lipid composition, turnover, mobilization and routing may provide greater mechanistic insight.

Impaired LD turnover may also promote persistent accumulation. In stressed chondrocytes, microtubule disruption impairs LD trafficking, LD–lysosome interactions and Hsc70–PLIN3–LAMP2A-dependent lipophagic clearance, and microtubule stabilization restores clearance and phenotype (Yu et al., 2025). This study directly links turnover defects to LD retention but did not compare them with increased lipid input. The relative contribution of defective turnover versus enhanced supply therefore remains unresolved.

Redox regulation and cholesterol handling intersect with LD homeostasis. Swimming exercise increased cartilage SOD3 in OA mice, and SOD3–C/EBPβ interaction suppressed APOE transcription and APOE-dependent cholesterol uptake, reducing LD load and free cholesterol (Wu et al., 2025). Beyond cholesterol uptake and storage, oxysterol-mediated nuclear receptor signaling represents an additional layer of lipid regulation in osteoarthritis. Oxysterols, including 25-hydroxycholesterol and related cholesterol oxidation products, function as endogenous ligands of liver X receptors (LXRs), thereby regulating cholesterol efflux, lipid transport and inflammatory transcriptional programs. In OA-related contexts, altered cholesterol handling and LXR activity may influence chondrocyte lipid homeostasis. However, whether specific oxysterol species directly regulate LD remodeling or mitochondrial lipid utilization in OA chondrocytes remains insufficiently defined (Zhang Y. et al., 2025). Therefore, oxysterol–LXR signaling should be considered as a potential regulatory pathway linking cholesterol metabolism with broader lipid remodeling rather than as an established driver of LD accumulation. Because this study primarily demonstrates altered cholesterol uptake rather than lipid transport or clearance, broader conclusions on coordinated supply–transport–clearance changes require further evidence.

LD accumulation should not be viewed as intrinsically pathological. When supply remains within handling capacity, esterification into neutral lipids buffers excess fatty acids, whereas defective DGAT1-dependent biogenesis exacerbates lipotoxic stress and ferroptosis in other systems (Wang Y.-T. et al., 2025). Sustained lipid overload or turnover failure converts transient buffering into maladaptive retention, so the biological readout of LD abundance depends on lipid composition, remobilization and routing relative to mitochondrial capacity.

4 Lipid droplet–mitochondria interfaces and lipid routing in osteoarthritis

Fatty acids mobilized from LDs may be re-esterified, incorporated into membrane phospholipids, converted into signalling lipids or oxidized, we focus on oxidation. Mobilization itself uses distinct pathways: ATGL/HSL/MGL-dependent lipolysis releases fatty acids from triacylglycerols, whereas lipophagy delivers intact LDs to lysosomes. Both replenish the cellular fatty-acid pool but are differentially regulated, and neither implies global conversion of triacylglycerols to free fatty acids. After activation by acyl-CoA synthetases, mitochondrial entry proceeds mainly through the carnitine shuttle (CPT1/CPT2) and does not. In principle, require LD–mitochondria contact, such contacts could nevertheless concentrate this flow, and microtubule-dependent trafficking plus PLIN-dependent anchoring determine which LDs are accessible. In non-chondrocyte systems, PLIN5–FATP4 contacts channel LD-derived fatty acids to mitochondrial β-oxidation.

Miner et al. showed in myoblasts of human and murine origin that PLIN5–FATP4 binding channels LD-derived fatty acids to mitochondria for β-oxidation (Miner et al., 2023), and Wang et al. described in HEK293 and COS-7 cells a VPS13D–TSG101/ESCRT-dependent remodelling of LD membranes that facilitates transfer at mitochondria–LD contacts (Wang et al., 2021). Contact sites thus function as active lipid-transfer platforms in these systems.

Proximity proteomics in cultured cells identified an ESYT1–ESYT2–VAPB complex at LD–mitochondria–endoplasmic reticulum interfaces, its loss reduced LD-derived fatty-acid oxidation without altering physical proximity, dissociating contact abundance from transfer function (Bezawork-Geleta et al., 2025). No equivalent data exist for OA chondrocytes.

Contact does not necessarily commit lipids to oxidation. Loss of ESYT1, ESYT2 or VAPB limited LD-derived fatty-acid oxidation in cultured cells and Esyt1/Esyt2-deficient mice (Bezawork-Geleta et al., 2025), whereas PLIN5 phosphorylation in hepatocytes promoted esterification and storage (Kang et al., 2026). LD–mitochondria contacts are thus context-dependent routers rather than simple oxidation gates.

Tethering follows a metabolic logic and is tissue-specific. Energy stress activates Rab8a–PLIN5 tethering in rat L6 myotubes (Ouyang et al., 2023), and glucose starvation activates an AMPK–ARF1–PLIN2 axis in C2C12 myoblasts and mouse skeletal muscle (Chen L. et al., 2025). Structural contact abundance alone cannot predict transfer efficiency or downstream fate.

No functional LD–mitochondria contact has yet been reported in OA chondrocytes. Contact frequency, architecture and molecular identity across OA stages are unknown, and tethering machinery is tissue-specific, for example, Mfn2–Hsc70 in cardiomyocytes (Hu et al., 2024). LD–mitochondria contacts in OA are therefore a hypothesis to be tested rather than an established mechanism, and no assumption of tissue homology is warranted.

If such contacts exist in OA, their relevance lies in routing efficiency relative to mitochondrial capacity. Impaired mitophagy and mitochondrial dysfunction promote chondrocyte degeneration (Long et al., 2025), whereas BNIP3 activation restores mitophagy and limits cartilage damage (Gou et al., 2025). The cellular outcome depends on whether mitochondrial oxidative and quality-control systems accommodate incoming lipid flux (Figure 2).

Context-dependent lipid routing at the lipid droplet–mitochondria interface in osteoarthritis.
Figure 2. Context-dependent lipid routing at the lipid droplet–mitochondria interface in osteoarthritis.

Initial validation in OA chondrocytes: of the candidate systems above, three are prioritized. PLIN5–FATP4 has the strongest direct evidence of lipid transfer capacity; PLIN3–VDAC3 combines proximity proteomics with PLIN3 processing already implicated in stressed chondrocytes [4]; ESYT1–ESYT2–VAPB provides the clearest example of contact-independent transfer. Expression, localization, contact occupancy and lipid-transfer assays should be assessed in human OA cartilage first; VPS13D–TSG101 and Rab8a–PLIN5 are informative but lower priority because their OA relevance is untested. The current evidence for LD–mitochondria crosstalk and its potential relevance to OA is summarized in Table 1.

TABLE 1. Evidence framework for lipid droplet–mitochondria crosstalk: established mechanisms and potential relevance to osteoarthritis.

Mechanistic featureRepresentative mechanismExperimental contextEvidence statusPotential relevance to OARepresentative references
Contact–transfer dissociationPhysical proximity between lipid droplets (LDs) and mitochondria does not necessarily indicate functional fatty-acid transferCultured cells; myoblasts; Esyt1/2-deficient mice (non-chondrocyte systems)Indirect mechanistic evidenceLD–mitochondria contact abundance should be distinguished from actual lipid transfer activity in OA chondrocytes(Bezawork-Geleta et al., 2025; Wang et al., 2021)
Context-dependent lipid routingLD-derived fatty acids may undergo mitochondrial oxidation, re-esterification, membrane phospholipid synthesis or lipid signaling depending on cellular demandHepatocytes; skeletal muscle; cultured cellsIndirect mechanistic evidenceOA-associated lipid fate cannot be inferred from LD abundance or FAO alone(Miner et al., 2023; Kang et al., 2026)
Tissue-specific tethering machineryPLIN5–FATP4, ESYT1/2–VAPB and other tethering proteins regulate LD–mitochondria coupling in a tissue-dependent mannerMyoblasts, hepatocytes, cultured cells and mouse modelsEstablished in non-cartilage systemsThese proteins represent candidate regulators requiring validation in OA chondrocytes(Miner et al., 2023; Wang et al., 2021; Bezawork-Geleta et al., 2025)
PLIN-dependent lipid transferPLIN5–FATP4 interaction facilitates LD-derived fatty-acid transport toward mitochondrial β-oxidationHuman and mouse myoblastsIndirect mechanistic evidenceProvides a candidate mechanism for lipid routing, but its relevance to OA cartilage remains unknown(Miner et al., 2023)
ESYT/VAPB-mediated contact regulationESYT1/2–VAPB complexes regulate LD–mitochondria–ER interfaces and influence lipid transfer efficiencyCultured cells and mouse modelsIndirect mechanistic evidenceSuggests that contact function rather than contact number may determine metabolic outcomesBezawork-Geleta et al., 2025
Candidate validation priorities in OAPLIN5–FATP4, PLIN3–VDAC3 and ESYT1/2–VAPB represent candidate LD–mitochondria regulatory systemsProposed validation in human OA chondrocytesHypothesis-generatingRequires spatial localization, genetic perturbation and direct lipid-transfer assays in OA models(Miner et al., 2023; Bezawork-Geleta et al., 2025)

5 Mitochondrial capacity and lipid flux in osteoarthritic chondrocytes

To distinguish physiological metabolic adaptation from pathological metabolic mismatch, lipid–mitochondrial states in healthy and osteoarthritic chondrocytes are summarized in Table 2.

TABLE 2. Comparison of lipid–mitochondrial metabolic states between healthy and osteoarthritic chondrocytes.

FeatureHealthy cartilageOsteoarthritic cartilageRepresentative evidence
Lipid droplet biologyDynamic lipid buffering, controlled neutral lipid storage and regulated turnover maintain intracellular lipid homeostasisIncreased lipid retention may result from altered lipid synthesis, trafficking or clearance; LD accumulation alone is insufficient to define the underlying metabolic state(Yu et al., 2025; Zhou et al., 2025)
Lipid composition and remodelingBalanced distribution of glycerolipids, cholesteryl esters, phospholipids and other lipid species supports membrane integrity and metabolic adaptationOA involves broad lipid remodeling, including alterations in glycerolipids, cholesteryl esters, sphingolipids and other complex lipid classes(Zhou et al., 2025)
Lipid routingCoordinated partitioning of fatty acids among neutral lipid storage, membrane phospholipid synthesis, signaling pathways and mitochondrial oxidation according to cellular demandAltered lipid partitioning may increase retention, redirect fatty acids toward oxidation or peroxidation, and disturb metabolic adaptation; the relative contribution of each route remains context-dependent(Mei et al., 2025; Zhou et al., 2025; Wang et al., 2025a)
Fatty-acid utilizationFatty-acid oxidation is matched with mitochondrial oxidative capacity and cellular energy requirementsEnhanced lipid flux or altered fatty-acid utilization may exceed mitochondrial processing capacity, contributing to metabolic stress and chondrocyte dysfunctionMei et al., 2025; Wang et al., 2025a)
Mitochondrial oxidative capacityEfficient substrate oxidation, ATP production and respiratory adaptation maintain metabolic flexibility under changing substrate availabilityReduced respiratory adaptability, impaired substrate utilization and decreased metabolic flexibility increase susceptibility to metabolic stress(Liu et al., 2026)
Redox-buffering capacityControlled ROS production with effective antioxidant systems maintains redox balanceIncreased mtROS accumulation, lipid peroxidation and impaired antioxidant responses contribute to oxidative damage and inflammatory signaling(Wu et al., 2025; Jiang et al., 2025)
Mitochondrial quality-control capacityCoordinated mitophagy, mitochondrial dynamics and clearance of damaged organelles preserve mitochondrial integrityDysregulated mitophagy and mitochondrial dynamics promote accumulation of dysfunctional mitochondria and impair stress adaptation(Zhuang et al., 2024; Fang et al., 2023; Hu et al., 2023)
Lipid flux–mitochondrial capacity relationshipLipid input, storage, mobilization and mitochondrial processing remain balanced, supporting metabolic adaptationLipid supply and intracellular routing may exceed mitochondrial oxidative and quality-control capacity, leading to redox imbalance, ferroptotic susceptibility and senescence(Mei et al., 2025; Cao S. et al., 2025; Liu et al., 2026; Zou et al., 2025)
Chondrocyte phenotype and ECM homeostasisMaintained SOX9/COL2A1/ACAN expression and extracellular matrix synthesis preserve cartilage phenotypeMetabolic mismatch promotes anabolic–catabolic imbalance, senescence, ferroptotic vulnerability and extracellular matrix degradation(Li H. et al., 2026; Jia et al., 2026; Fang et al., 2026)

5.1 Physiological baseline of lipid–mitochondrial homeostasis in healthy chondrocytes

In healthy chondrocytes, oxidative metabolism supports ATP production and redox balance while fatty-acid utilization stays matched to oxidative capacity, and quality control (mitophagy plus fusion–fission dynamics) removes damaged organelles; in OA this balance is disrupted. Consequently, higher FAO is not uniformly beneficial: FPH2-mediated inhibition of CPT1-dependent mitochondrial entry improves chondrocyte metabolism and cartilage regeneration in vivo (Chen et al., 2024), whereas PKM2 dimerization disrupts MFN1-dependent mitochondrial fusion and function (Liu et al., 2026). Whether lipid oxidation helps or harms therefore depends on mitochondrial competence rather than on the absolute rate of FAO.

5.2 Defining mitochondrial oxidative, redox and quality-control capacity in osteoarthritic chondrocytes

Mitochondrial function denotes the capacity to utilize substrates, produce ATP, handle redox and preserve organelle integrity; mitochondrial capacity comprises three measurable components, namely, (i) oxidative capacity (substrate oxidation, oxygen consumption rate, ATP production, respiratory reserve), (ii) redox-buffering capacity (mtROS generation and antioxidant systems maintaining redox balance), and (iii) mitochondrial quality-control capacity (mitochondrial dynamics, mitophagic flux, clearance of damaged mitochondria). Mitochondrial quality control specifically denotes mitophagy and dynamics; biogenesis and proteostasis lie outside our scope. Protein abundance, for example, PINK1, Parkin, SIRT3 or respiratory-chain subunits records regulation not function. Capacity must be measured by metabolic flux analysis, extracellular flux, ATP production, ROS/redox assessment and dynamic mitophagy readouts.

Recent studies in OA chondrocytes further indicate that mitochondrial dysfunction is not represented by a uniform reduction in mitochondrial activity. Altered mitochondrial homeostasis and impaired metabolic adaptation contribute to OA progression. Thus, mitochondrial dysfunction in OA should be interpreted as reduced adaptive capacity to accommodate changing substrate availability rather than simply decreased mitochondrial activity (Liu et al., 2026).

5.3 Lipid flux–capacity mismatch in osteoarthritic chondrocytes

Compensatory increases in oxidative metabolism help sustain ATP and redox balance early, but become inadequate as dysfunction progresses. PDK2 deficiency restored oxidative phosphorylation, NAD+/NADH and ATP while reducing ROS and senescence in experimental OA (Han et al., 2025), whereas loss of PDZK1 increased damaged mitochondria, ROS and senescence (Shao et al., 2024). Oxidative metabolism is protective only while energetic and quality-control capacity meet demand.

Quality control contributes to capacity. Impaired FUNDC1-mediated mitophagy aggravates mitochondrial dysfunction, ROS accumulation and degeneration, while restoring the PFKP–FUNDC1 axis improves mitophagic clearance (Fang et al., 2023). TBK1-dependent DRP1 phosphorylation suppresses excessive fission and protects mitochondrial function (Hu et al., 2023). Quality control thus determines whether chondrocytes sustain metabolic stress without accumulating dysfunctional mitochondria.

The outcome of mitophagy cannot be inferred from static markers: α-ketoglutarate enhances mitophagy, reduces ROS and preserves cartilage, and its protection is abolished by mitophagy inhibition (Liu et al., 2023), whereas Astragalus polysaccharides improved OA while suppressing NOTCH3–PINK1/Parkin–LC3-II-associated mitophagy (Xu H. et al., 2026). Although the latter is pleiotropic, together these data argue for flux-based assessment.

Reliable assessment of mitophagy requires dynamic flux measurement. Validated reporters such as mt-Keima and mito-QC visualize delivery of mitochondria to acidic lysosomal compartments, and time courses with lysosomal inhibition (bafilomycin A1, chloroquine) distinguish increased initiation from impaired degradation. Mitophagic flux - mitochondrial recognition, autophagosomal sequestration, lysosomal delivery and degradation - should therefore be quantified rather than inferred from PINK1, Parkin or LC3-II abundance.

Collectively, the consequence of FAO in OA depends on whether flux is accommodated by oxidative and quality-control capacity. Maintenance of respiratory-chain function through the SIRT3–COX4I2 axis improves bioenergetics and attenuates degeneration (Zhang et al., 2023), whereas Txnip-driven reprogramming accompanies mitochondrial dysfunction and matrix degradation (Cao X. et al., 2025). The critical variable is the dynamic balance among lipid supply, mitochondrial oxidative competence and quality control (Figure 3).

Lipid flux–capacity matching determines metabolic outcome.
Figure 3. Lipid flux–capacity matching determines metabolic outcome.

6 Metabolic mismatch drives chondrocyte fate remodeling and whole-joint communication

Prolonged mismatch may reshape chondrocyte fate through convergent outcomes. We consider ferroptosis first, because it is the outcome most directly coupled to lipid routing and peroxidation, then senescence and matrix destabilization. LD accumulation per se is not pro-ferroptotic: in human cancer cell lines, DGAT-dependent PUFA sequestration into triacylglycerols suppresses ferroptosis (Lee et al., 2024), but this protection is conditional, because LD-localized FSP1 maintains neutral-lipid integrity through CoQ10 until loss of FSP1 permits oxidation of PUFA-rich triacylglycerols and LD-initiated ferroptosis (Lange et al., 2025). The ferroptotic consequence of LDs depends on PUFA partitioning and the antioxidant competence of the storage compartment, not on abundance.

In OA, ACSL4-centred remodelling links PUFA metabolism to ferroptotic injury. ACSL4 and 4-HNE increase in human and experimental OA cartilage, and ferrostatin-1 attenuates IL-1β-induced lipid peroxidation (Zou et al., 2025). Sp1-driven ACSL4 in IL-1β-treated human chondrocytes increases ROS, MDA and Fe2+, suppresses COL2A1/ACAN and induces MMP13, with reversal by ACSL4 silencing or ferrostatin-1 (He et al., 2023). No study has yet traced a continuous LD–PUFA–ACSL4/LPCAT3–ferroptosis axis, so susceptibility should be interpreted through lipid-pool partitioning and antioxidant capacity.

Senescence follows when quality control and redox buffering become insufficient. SIRT4 loss suppresses PINK1-mediated mitophagy, causing dysfunctional-mitochondria accumulation and senescence, reversible by SIRT4 or PINK1 restoration (Lin et al., 2024); PINK1/Parkin enhancement similarly restores membrane potential and reduces senescence and degeneration (Jie et al., 2024). Antioxidant capacity adds control: SIRT3/SOD2 activation suppresses senescence and delays OA (Zhang et al., 2024), and SIRT3/SOD2-dependent redox restoration reduces mtROS and the senescence-associated secretory phenotype (SASP)(Jiang et al., 2025). Defective mitochondrial clearance and mtROS accumulation thus reinforce senescence, potentially intersecting with lipid peroxidation, although a continuous causal circuit in OA remains unestablished.

These disturbances converge on loss of chondrocyte phenotype and matrix homeostasis. Oleic-acid accumulation in obesity-associated OA promotes CaMK2-dependent SOX9 degradation in murine chondrocytes and human cartilage explants (Li H. et al., 2026). CILP–Keap1–Nrf2 restoration increases type II/I collagen and SOX9 while suppressing lipid peroxidation and ferroptotic injury in early OA (Jia et al., 2026), and P21-dependent regulation reduces oxidative stress, promotes anabolic markers and suppresses MMP13 and ADAMTS4 (Fang et al., 2026). ECM failure is better viewed as a convergent outcome of lipid routing, mitochondrial dysfunction and redox imbalance than as a direct effect of LD abundance.

Metabolic imbalance propagates across the joint. RCAN1+IL-1α+ senescent synovial fibroblasts stabilize ATF4 and secrete saturated fatty acids that induce chondrocyte senescence, and synovium-targeted RCAN1 silencing attenuates this crosstalk (Hu et al., 2025). In synovial macrophages, pathological mechanical stress suppresses YAP–TEAD1-dependent MFN1, disrupting fusion, membrane potential and redox homeostasis, whereas YAP–MFN1 restoration reduces mtROS and catabolic signalling (Cheng et al., 2026). Macrophage-specific PGAM5 targeting likewise alters inflammatory output and protects cartilage (Liu et al., 2024). Macrophage function is thus defined by mitochondrial dynamics and redox integration rather than an M1/M2 binary.

Mitochondrial transfer extends this crosstalk. Human umbilical-cord mesenchymal stromal cells transfer mitochondria to OA chondrocytes, raising ATP production, oxygen consumption and redox resistance (Court et al., 2024), with connexin-43 (GJA1-20k) promoting contact-dependent donation (Irwin et al., 2024), and intra-articular mitochondrial transplantation attenuating degeneration (Vega-Letter et al., 2025). Conversely, mitochondrial components can also propagate pathological signals between joint compartments. Recent evidence demonstrated that infrapatellar fat pad mesenchymal stem cell-derived extracellular vesicles containing mitochondrial DNA (mtDNA) are increased in OA and can be internalized by chondrocytes, activating cGAS-STING-dependent inflammatory signaling and promoting mitochondrial stress and cartilage degeneration (Li S. et al., 2026). Thus, intercellular mitochondrial communication may exert either protective or pathogenic effects depending on the transferred cargo and cellular context. Chondrocyte–endothelial–osteoblast communication is extensive, but bone-derived lipid signals or LD–mitochondrial mechanisms remain undefined (Liu Y. et al., 2025).

Chondrocyte fate emerges from the imbalance among lipid storage and mobilization, membrane phospholipid remodelling, mitochondrial oxidation and quality control, and across the joint these processes are modulated by synovial fibroblast-derived lipids, macrophage mitochondrial state and intercellular mitochondrial transfer. Mapped onto the framework (Figure 3), synovial fibroblast-derived saturated fatty acids primarily raise the lipid-flux axis (extracellular burden and availability), whereas macrophage mitochondrial remodelling and intercellular transfer act on the mitochondrial-capacity axis (function, redox balance, resilience); osteochondral signalling may influence both through nutrient exchange and inflammation. Whether these extra-chondrocytic inputs push chondrocytes toward mismatch therefore depends on their relative magnitude, a prediction testable by manipulating each axis independently. These metabolic disturbances and their implications for chondrocyte fate remodeling and whole-joint communication in OA are summarized in Table 3.

TABLE 3. Lipid–mitochondrial metabolic mismatch and chondrocyte fate remodeling in osteoarthritis.

Metabolic disturbanceMajor consequenceExperimental contextEvidence statusOA implicationRepresentative references
PUFA partitioning between TAGs and membrane phospholipidsLipid peroxidation and ferroptotic susceptibilityNon-chondrocyte systems (human cancer cell lines)Indirect mechanistic evidenceLD-mediated PUFA sequestration may influence ferroptosis susceptibility; however, OA-specific lipid-pool redistribution remains unresolved(Lee et al., 2024; Lange et al., 2025)
ACSL4-dependent phospholipid remodelingFerroptotic injury and extracellular matrix catabolismHuman OA cartilage; IL-1β-treated human primary chondrocytesDirect OA evidenceACSL4 links PUFA remodeling with ferroptotic responses in OA, whereas a continuous LD–PUFA–ACSL4/LPCAT3 pathway requires further validation(Zou et al., 2025; He et al., 2023)
Impaired mitochondrial quality controlmtROS accumulation, cellular senescence and SASP activationHuman OA chondrocytes; experimental OA modelsDirect OA evidenceDefective mitophagy and impaired antioxidant systems reduce chondrocyte tolerance to metabolic stress(Zhuang et al., 2024; Fang et al., 2023; Hu et al., 2023)
Metabolic convergence on chondrocyte phenotypeLoss of SOX9/COL2A1/ACAN expression and extracellular matrix degradationHuman cartilage explants; primary chondrocytes; experimental OA modelsDirect OA evidenceLipid remodeling, mitochondrial dysfunction and redox imbalance converge on anabolic–catabolic phenotype switching rather than LD accumulation alone(Li H. et al., 2026; Jia et al., 2026; Fang et al., 2026)
Synovial fibroblast-derived lipid signalingChondrocyte senescence and catabolic remodelingHuman OA synovium; experimental OA modelsDirect OA evidenceExtra-cartilaginous lipid signals increase metabolic burden on chondrocytes and primarily influence the lipid-flux axis(Hu et al., 2025)
Macrophage mitochondrial dysfunctionAltered inflammatory output and paracrine signalingSynovial macrophage OA modelsDirect OA evidenceMacrophage mitochondrial dynamics and redox regulation influence inflammatory signaling and joint homeostasis(Cheng et al., 2026; Liu et al., 2024)
Protective mitochondrial transferIncreased ATP production, oxidative phosphorylation and redox resistanceMSCs; OA chondrocytes; experimental OA modelsEmerging OA evidenceIntercellular mitochondrial donation may enhance recipient-cell metabolic capacity under stress conditions(Court et al., 2024; Irwin et al., 2024; Vega-Letter et al., 2025)
Pathogenic extracellular vesicle mitochondrial signalingmtDNA transfer, cGAS–STING activation and cartilage degenerationIPFP-MSC-derived extracellular vesicles; OA chondrocytesEmerging OA evidenceMitochondrial cargo transfer may propagate inflammatory signals depending on cargo composition and recipient-cell response(Li S. et al., 2026)
Bone–cartilage metabolic communicationWhole-joint metabolic remodelingHuman single-cell studies; OA modelsEstablished communication; lipid mechanisms unresolvedCommunication between cartilage, synovium and subchondral bone may influence both lipid flux and mitochondrial capacity, but LD–mitochondria mechanisms remain undefined(Wang et al., 2022; Raut et al., 2025; Schurman et al., 2026)
Lipid flux–mitochondrial capacity mismatchFerroptosis, senescence and extracellular matrix failureHuman OA chondrocytes; experimental OA modelsIntegrative framework supported by multiple evidence levelsDisease progression depends on whether mitochondrial oxidative capacity and quality-control systems can accommodate altered lipid flux(Mei et al., 2025; Cao S. et al., 2025; Liu et al., 2026; Zou et al., 2025)

7 Therapeutic strategies and research priorities

Therapy should restore coordination rather than directionally manipulate a single pathway. SESN2 restoration suppresses SREBP1-dependent lipogenesis (Liu Z. et al., 2025), SIRT5-dependent protein lysine malonylation controls chondrocyte metabolism under ageing and obesity (Liu H. et al., 2025), and the mitochondria-targeted antioxidant XJB-5-131 suppresses lipid peroxidation and ferroptosis (Sun et al., 2024). These are proofs of metabolic-node tractability, not of sufficiency. Beyond direct metabolic modulation, extracellular vesicle-based delivery systems have emerged as a strategy to coordinate multiple pathological processes within the OA microenvironment. A logic-based spatiotemporal dual extracellular vesicle platform has been developed to achieve sequential regulation of inflammatory responses and tissue repair, highlighting the potential of engineered vesicles to overcome the limitations of conventional single-target interventions (Li S. et al., 2024). Such approaches may provide a framework for delivering metabolic regulators, anti-inflammatory signals or organelle-modulating cargos in a temporally controlled manner.

LDs should not be depleted indiscriminately. Accumulation may reflect increased synthesis or defective turnover, as TGF-β2–TβRI–p-Smad3–FABP5 promotes neutral-lipid accumulation in chondrocytes (Wei et al., 2026), whereas APOE loss impairs lipid handling and pharmacological LXR agonism (RGX-104) partially restores homeostasis (Zhang Y. et al., 2025). For PUFAs, AMPK-driven LD dynamics redistribute LD-stored PUFAs toward mitochondria in melanoma cell lines, increasing peroxidation susceptibility (Motamedi et al., 2025), and linoleate accumulation promotes ferroptotic injury in male rats (Deng et al., 2026). The therapeutic aim is context-dependent restoration of storage, mobilization and membrane partitioning while preserving buffering.

Mitochondrial fatty-acid handling should be neither uniformly activated nor inhibited. Restricting CPT1-dependent entry improves chondrocyte fitness (Chen et al., 2024), yet sustained FAO drives acetyl-CoA accumulation, HADHA hyperacetylation and SOX9 destabilization (Mei et al., 2025), and resistin-driven CAP1–PKA–CREB signalling increases CPT1A and FAO in metabolic-syndrome synoviocytes, promoting catabolism (Ding et al., 2025). Restoring competence is preferable: cartilage-targeted SIRT3 activation enhances energy production and repair (Lu et al., 2025), fasting-induced OPTN-mediated mitophagy reduces dysfunctional mitochondria (Zhang M.-N. et al., 2025), and JP4-039 enhances PINK1/Parkin mitophagy while suppressing lipid peroxidation (Xie et al., 2025).

Static readouts cannot resolve lipid origin and fate. Time-resolved 13C pulse–chase tracing should quantify incorporation from LD-labelled triacylglycerols into acylcarnitines, acetyl-CoA and TCA-cycle intermediates (Chen S. et al., 2025), supported by LC–MS acyl-CoA profiling (Liu R. et al., 2025), pulse–chase imaging of LD-to-mitochondria trafficking (Miner and Cohen, 2024) and temporally controlled perturbation of ATGL, lipophagy and CPT1A, since LD-to-mitochondria transfer and lipophagy are differentially regulated (Xue et al., 2026).

Contact sites should be converted from descriptors into manipulable variables. Reversible split-fluorophore systems (FABCON) resolve dynamic mitochondria–LD contacts (Li X. et al., 2024), LD-membrane probes with structured illumination resolve contact remodelling and fatty-acid redistribution (Kong et al., 2024), and proximity proteomics (LipoID; PLIN3–VDAC3) links contact regulators to membrane potential, ATP, FAO and lipid profiles (Guo et al., 2026). Increased contact abundance must not be equated with increased transfer, causal tests require perturbation paired with direct lipid-transfer and flux measurements.

Lipophagy and mitophagy should be tracked together in matched systems. TfLiveDrop reports lipophagic flux in vivo (Gong et al., 2026), fluorescence approaches resolve sequential mitophagic intermediates (Wang Y. et al., 2025), and Rab GTPases act as selective-autophagy cues shared by lipid droplets and damaged mitochondria (Zhao et al., 2025). Whether these fluxes remain matched or dissociate across OA stages is unknown.

Whole-joint reconstruction should integrate single-cell and spatial transcriptomics with spatial lipidomics, metabolomics and isotope-resolved flux analysis across synovium, cartilage and subchondral bone. Spatial immune dynamics in synovium (Wang et al., 2025b), proteomic–metabolomic synovitis endotypes (Zhu et al., 2025), molecular gradients across the cartilage–subchondral bone interface (Schurman et al., 2026), spatial isotope deep tracing (Li et al., 2025) and lipid metabolic flux analysis (Wessendorf-Rodriguez et al., 2026) are now feasible. The goal is to locate lipid origin, recipient cells and processing pathways, which abundance maps cannot resolve, and to distinguish synthesis from uptake or inter-tissue transport.

Human validation requires multidimensional stratification: age, sex, obesity, insulin resistance and imaging or histopathological grade, supported by multimodal risk modelling (Nielsen et al., 2024) and evidence that OA is a molecular continuum with phenotype-enriched pathways (Perry et al., 2026). Mechanistic findings should then be tested in human explants and organoid models (Dönges et al., 2024) before animal and independent human validation.

8 Conclusion

LD–mitochondrial crosstalk in OA has no fixed direction. Its consequences depend on the coordination of lipid input, LD buffering and turnover, intracellular routing and mitochondrial oxidative and quality-control capacity. When matched, storage, mobilization and oxidation support adaptation; when flux exceeds capacity, redox imbalance, lipid peroxidation, ferroptotic susceptibility, senescence and matrix disruption follow. Therapy should restore this matching rather than target a single organelle or pathway, and research should trace lipid fate across space and time to test it.

Generative AI statement

The author(s) declared that generative AI was not used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Glossary

4-HNE 4-hydroxynonenal ACC acetyl-CoA carboxylase ACOT12 acyl-CoA thioesterase 12 ACSL4 acyl-CoA synthetase long-chain family member 4 AMPK AMP-activated protein kinase APOE apolipoprotein E ARF1 ADP-ribosylation factor 1 ATF4 activating transcription factor 4 ATP adenosine triphosphate BNIP3 BCL2 interacting protein 3 CAP1 adenylyl cyclase-associated protein 1 CILP cartilage intermediate layer protein CMA chaperone-mediated autophagy CoQ10 coenzyme Q10 COX4I2 cytochrome c oxidase subunit 4 isoform 2 CPT1 carnitine palmitoyltransferase 1 CPT1A carnitine palmitoyltransferase 1A CREB cAMP response element-binding protein DGAT1 diacylglycerol O-acyltransferase 1 DNL de novo lipogenesis DRP1 dynamin-related protein 1 ECM extracellular matrix ESCRT endosomal sorting complexes required for transport ESYT1/2 extended synaptotagmin 1/2 FABCON fluorogen-activated BiFC at contact sites FABP5 fatty acid-binding protein 5 FAO fatty acid oxidation FATP4 fatty acid transport protein 4 FGF8 fibroblast growth factor 8 FGFR1 fibroblast growth factor receptor 1 FSP1 ferroptosis suppressor protein 1 FSP27 fat-specific protein 27 FUNDC1 FUN14 domain-containing protein 1 GJA1 gap junction protein alpha 1 HADHA hydroxyacyl-CoA dehydrogenase trifunctional multienzyme complex subunit alpha HCAR2 hydroxycarboxylic acid receptor 2 Hsc70 heat shock cognate protein 70 IL-1β interleukin-1β ISCU iron-sulfur cluster assembly enzyme LAMP2A lysosome-associated membrane protein 2A LD lipid droplet LPCAT3 lysophosphatidylcholine acyltransferase 3 LXR liver X receptor MDA malondialdehyde MFN1/2 mitofusin 1/2 MMP13 matrix metalloproteinase 13 MSC mesenchymal stromal cell mtROS mitochondrial reactive oxygen species Nrf2 nuclear factor erythroid 2-related factor 2 OA osteoarthritis OXPHOS oxidative phosphorylation PDK2 pyruvate dehydrogenase kinase 2 PDZK1 PDZ domain-containing protein 1 PFKP phosphofructokinase platelet type PGAM5 phosphoglycerate mutase family member 5 PINK1 PTEN-induced kinase 1 PKA protein kinase A PKCK2 protein kinase casein kinase 2 PKM2 pyruvate kinase M2 PLIN1/2/3/5 perilipin 1/2/3/5 PPARα peroxisome proliferator-activated receptor α PUFA polyunsaturated fatty acid RCAN1 regulator of calcineurin 1 ROS reactive oxygen species SASP senescence-associated secretory phenotype SESN2 sestrin 2 SIRT3/4/5 sirtuin 3/4/5 SOD2/3 superoxide dismutase 2/3 SOX9 SRY-box transcription factor 9 SREBP1 sterol regulatory element-binding protein 1 STAMP2 six-transmembrane protein of prostate 2 TAG triacylglycerol TBK1 TANK-binding kinase 1 TCA tricarboxylic acid cycle TEAD1 TEA domain transcription factor 1 TGF-β2 transforming growth factor-β2 TSG101 tumor susceptibility gene 101 TβRI transforming growth factor-β receptor I VAPB VAMP-associated protein B VDAC3 voltage-dependent anion channel 3 VPS13D vacuolar protein sorting 13 homolog D YAP Yes-associated protein.

Footnotes

Edited by: Martin Van Der Laan, Saarland University, Germany

Reviewed by: Shiyu Li, Jinan University, China

Daniela Bispo, Lund University, Sweden

Statements

Conflict of interest

The author(s) declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Funding

The author(s) declared that financial support was received for this work and/or its publication. Inner Mongolia Medical University.

Author contributions

GM: Writing – original draft. RH: Writing – review and editing, Writing – original draft. PD: Funding acquisition, Supervision, Validation, Writing – review and editing. QC: Methodology, Writing – review and editing. XL: Conceptualization, Writing – original draft. KH: Software, Writing – original draft. ZD: Data curation, Writing – original draft. RW: Investigation, Writing – original draft. ZZ: Validation, Writing – original draft.

References

  1. Bezawork-Geleta A., Devereux C. J., Keenan S. N., Lou J., Cho E., Nie S. (2025). Proximity proteomics reveals a mechanism of fatty acid transfer at lipid droplet-mitochondria-endoplasmic reticulum contact sites. Nat. Commun. 16, 2135. doi:10.1038/s41467-025-57405-5
  2. Cao S., Wei Y., Xiong A., Yue Y., Yang J., Wang D. (2025). Paeonol inhibits ACSL4 to protect chondrocytes from ferroptosis and ameliorates osteoarthritis progression. J. Ortho. Trans. 50, 1 – 13. doi:10.1016/j.jot.2024.10.005
  3. Cao X., Yang X., Zhang P., Xu J., Zhao J., Yang E. (2025). Targeting Txnip-mediated metabolic reprogramming has therapeutic potential for osteoarthritis. Cell Death Disc. 11, 110. doi:10.1038/s41420-025-02394-z
  4. Chen Y., Yan Y., Tian R., Sheng Z., Li L., Chen J. (2024). Chemically programmed metabolism drives a superior cell fitness for cartilage regeneration. Sci. Adv. 10, eadp4408. doi:10.1126/sciadv.adp4408
  5. Chen L., Liu Y., Zhang J., Song T., Wu J., Ren Z. (2025). AMPK regulates ARF1 localization to membrane contact sites to facilitate fatty acid transfer between lipid droplets and mitochondria. Cell Death Dis. 16, 623. doi:10.1038/s41419-025-07957-7
  6. Chen S., Zhuo A., Tang R., Su S., Wen B., Wei W. (2025). A-to-I edited SNHG3 promotes non-small cell lung cancer metastasis by promoting fatty acid oxidation and resisting ferroptosis. Commun. Biol. 8, 1333. doi:10.1038/s42003-025-08776-4
  7. Cheng Y., Guo Z.-L., Gu Y.-F., Li W.-L., Zhang W.-B. (2026). YAP in synovial macrophages mitigates mechanically-induced osteoarthritis by preserving mitochondrial fusion. Cell. Signal. 143, 112430. doi:10.1016/j.cellsig.2026.112430
  8. Court A. C., Vega-Letter A. M., Parra-Crisóstomo E., Velarde F., García C., Ortloff A. (2024). Mitochondrial transfer balances cell redox, energy and metabolic homeostasis in the osteoarthritic chondrocyte preserving cartilage integrity. Theranostics 14, 6471 – 6486. doi:10.7150/thno.96723
  9. Deng Y., Hou M., Wu Y., Liu Y., Xia X., Yu C. (2025). SIRT3-PINK1-PKM2 axis prevents osteoarthritis via mitochondrial renewal and metabolic switch. Bone Res. 13, 36. doi:10.1038/s41413-025-00413-4
  10. Deng X., Xu H., Wu J., Pan C., Hou W., Zhang Y. (2026). Linoleic acid accelerates osteoarthritis progression in male rats by targeting iron-sulfur clusters to drive ferroptosis in chondrocytes. Nat. Commun. 17 (1), 8447. doi:10.1038/s41467-026-75513-8
  11. Ding L., Ren J.-Y., Huang Y.-F., Zhang J.-Z., Bai Z.-R., Leng Y. (2025). Resistin upregulates fatty acid oxidation in synoviocytes of metabolic syndrome-associated knee osteoarthritis via CAP1/PKA/CREB to promote inflammation and catabolism. Arthritis Res. Ther. 27, 99. doi:10.1186/s13075-025-03527-6
  12. Dönges L., Damle A., Mainardi A., Bock T., Schönenberger M., Martin I. (2024). Engineered human osteoarthritic cartilage organoids. Biomaterials 308, 122549. doi:10.1016/j.biomaterials.2024.122549
  13. Fang G., Wen X., Jiang Z., Du X., Liu R., Zhang C. (2023). FUNDC1/PFKP-mediated mitophagy induced by KD025 ameliorates cartilage degeneration in osteoarthritis. Mol. Ther. 31, 3594 – 3612. doi:10.1016/j.ymthe.2023.10.016
  14. Fang C., Ji Z., Pan Y., Li Z., Liu C., Feng X. (2026). Dihydroxyacetone nanoparticles ameliorate osteoarthritis via P21-mediated regulation of chondrocyte repair. J. Nanobiotechnol. 24, 692. doi:10.1186/s12951-026-04595-9
  15. Gong S., Zhang Q., Wu H., Chen X., Liu L., Chen Y. (2026). In vivo visualization of lipophagy dynamics using the tfLiveDrop reporter mice. J. Lipid Res. 67, 101033. doi:10.1016/j.jlr.2026.101033
  16. Gou Y., Wang C., Fu K., Su S., Zhou H., Bao C. (2025). Targeted activation on Bnip3 enhances mitophagy to prevent the progression of osteoarthritis. J. Orthop. Transl. 51, 242 – 255. doi:10.1016/j.jot.2025.01.012
  17. Guo H., Wan W., Huang Y., Zhao N., Wu C., Zhong B. (2026). LipoID profiles lipid droplet interactions and identifies interorganelle regulators. Nat. Chem. Biol. 22, 1 – 11. doi:10.1038/s41589-025-02127-4
  18. Han J., Kim Y. H., Han S. (2025). Increased oxidative phosphorylation through pyruvate dehydrogenase kinase 2 deficiency ameliorates cartilage degradation in mice with surgically induced osteoarthritis. Exp. Mol. Med. 57, 390 – 401. doi:10.1038/s12276-025-01400-9
  19. He W., Lin X., Chen K. (2023). Specificity protein 1-mediated ACSL4 transcription promoted the osteoarthritis progression through suppressing the ferroptosis of chondrocytes. J. Orthop. Surg. Res. 18, 188. doi:10.1186/s13018-023-03673-0
  20. Hu S.-L., Mamun A. A., Shaw J., Li S.-L., Shi Y.-F., Jin X.-M. (2023). TBK1-medicated DRP1 phosphorylation orchestrates mitochondrial dynamics and autophagy activation in osteoarthritis. Acta Pharmacol. Sin. 44, 610 – 621. doi:10.1038/s41401-022-00967-7
  21. Hu L., Tang D., Qi B., Guo D., Wang Y., Geng J. (2024). Mfn2/Hsc70 complex mediates the formation of mitochondria-lipid droplets membrane contact and regulates myocardial lipid metabolism. Adv. Sci. 11, 2307749. doi:10.1002/advs.202307749
  22. Hu J., Yu D., Wang J., Guo P., Chen M., Wu D. (2025). Targeting of hyperlipid-producing senescent synovial fibroblasts to ameliorate cartilage degeneration. Proc. Natl. Acad. Sci. 122, e2417708122. doi:10.1073/pnas.2417708122
  23. Huang M., Chen H., Wei J., Pi C., Duan M., Pu X. (2025). FGF8 promotes lipid droplet accumulation via the FGFR1/p-p38 axis in chondrocytes: FGF8 increases lipid droplet accumulation. Acta Biochimica Biophysica Sinica 58, 258 – 274. doi:10.3724/abbs.2025075
  24. Irwin R. M., Thomas M. A., Fahey M. J., Mayán M. D., Smyth J. W., Delco M. L. (2024). Connexin 43 regulates intercellular mitochondrial transfer from human mesenchymal stromal cells to chondrocytes. Stem Cell Res. Ther. 15, 359. doi:10.1186/s13287-024-03932-9
  25. Jia S., Hu Z., Li Z., Zhang W., Chen L., Niu C. (2026). CILP Inhibits hyaline cartilage fibrosis and chondrocyte ferroptosis via keap1-Nrf2 axis in early osteoarthritis exercise therapy. Cell. Mol. Life Sci. 83 (1), 185. doi:10.1007/s00018-026-06174-5
  26. Jiang H., Zhang Y., Ji P., Ming J., Li Y., Zhou Y. (2025). Surfactant protein D alleviates chondrocytes senescence by upregulating SIRT3/SOD2 pathway in osteoarthritis. Mol. Med. 31, 161. doi:10.1186/s10020-025-01221-6
  27. Jie L., Shi X., Kang J., Fu H., Yu L., Tian D. (2024). Protocatechuic aldehyde attenuates chondrocyte senescence via the regulation of PTEN-induced kinase 1/Parkin-mediated mitochondrial autophagy. Int. J. Immunopathol. Pharmacol. 38, 03946320241271724. doi:10.1177/03946320241271724
  28. Kang S. W. S., Brown L. A., Miller C. B., Barrows K. M., Golino J. L., Liu H. (2026). PLIN5 phosphorylation orchestrates mitochondria lipid-droplet coupling to control hepatic lipid flux and steatosis. Nat. Metab. 8, 587 – 603. doi:10.1038/s42255-026-01476-1
  29. Kong L., Bai Q., Li C., Wang Q., Wang Y., Shao X. (2024). Molecular probes for tracking lipid droplet membrane dynamics. Nat. Commun. 15, 9413. doi:10.1038/s41467-024-53667-7
  30. Lange M., Wölk M., Li V. W., Doubravsky C. E., Hendricks J. M., Kato S. (2025). FSP1-mediated lipid droplet quality control prevents neutral lipid peroxidation and ferroptosis. Nat. Cell Biol. 27, 1902 – 1913. doi:10.1038/s41556-025-01790-y
  31. Lee S. W., Rho J. H., Lee S. Y., Chung W. T., Oh Y. J., Kim J. H. (2018). Dietary fat-associated osteoarthritic chondrocytes gain resistance to lipotoxicity through PKCK2/STAMP2/FSP27. Bone Res. 6, 20. doi:10.1038/s41413-018-0020-0
  32. Lee H., Horbath A., Kondiparthi L., Meena J. K., Lei G., Dasgupta S. (2024). Cell cycle arrest induces lipid droplet formation and confers ferroptosis resistance. Nat. Commun. 15, 79. doi:10.1038/s41467-023-44412-7
  33. Li S., Zheng W., Deng W., Li Z., Yang J., Zhang H. (2024). Logic-based strategy for spatiotemporal release of dual extracellular vesicles in osteoarthritis treatment. Adv. Sci. 11, 2403227. doi:10.1002/advs.202403227
  34. Li X., Gamuyao R., Wu M.-L., Cho W. J., King S. V., Petersen R. (2024). A fluorogenic complementation tool kit for interrogating lipid droplet–organelle interaction. J. Cell Biol. 223, e202311126. doi:10.1083/jcb.202311126
  35. Li X., Zhu Y., Li T., Tu X., Zhu S., Wang L. (2025). Spatial isotope deep tracing deciphers inter-tissue metabolic crosstalk. Nat. Commun. 16, 7934. doi:10.1038/s41467-025-63243-2
  36. Li H., Fu J., Liu J., Zhang B., Cai C., Zhang Z. (2026). Obesity-induced oleic acid metabolic dysregulation may exacerbate osteoarthritis through the degradation of SOX9. Cell. Mol. Life Sci. 83 (1), 284. doi:10.1007/s00018-026-06261-7
  37. Li S., Yan Z., Zhi X., Zheng W., Zhang Z., Dai Z. (2026). Extracellular vesicles from IPFP-MSCs trigger osteoarthritis by transferring mtDNA. Bioact. Mater. 58, 252 – 273. doi:10.1016/j.bioactmat.2025.11.046
  38. Lin S., Wu B., Hu X., Lu H. (2024). Sirtuin 4 (Sirt4) downregulation contributes to chondrocyte senescence and osteoarthritis via mediating mitochondrial dysfunction. Int. J. Biol. Sci. 20, 1256 – 1278. doi:10.7150/ijbs.85585
  39. Liu H., Witzigreuter L., Sathiaseelan R., Agbaga M. P., Brush R. S., Stout M. B. (2022). Obesity promotes lipid accumulation in mouse cartilage—A potential role of acetyl-CoA carboxylase (ACC) mediated chondrocyte de novo lipogenesis. J. Orthop. Res. 40, 2771 – 2779. doi:10.1002/jor.25322
  40. Liu L., Zhang W., Liu T., Tan Y., Chen C., Zhao J. (2023). The physiological metabolite α-ketoglutarate ameliorates osteoarthritis by regulating mitophagy and oxidative stress. Redox Biol. 62, 102663. doi:10.1016/j.redox.2023.102663
  41. Liu Y., Hao R., Lv J., Yuan J., Wang X., Xu C. (2024). Targeted knockdown of PGAM5 in synovial macrophages efficiently alleviates osteoarthritis. Bone Res. 12, 15. doi:10.1038/s41413-024-00318-8
  42. Liu H., Binoy A., Ren S., Martino T. C., Miller A. E., Willis C. R. (2025). Regulation of chondrocyte metabolism and osteoarthritis development by Sirt5 through protein lysine malonylation. Arthritis Rheumatol. 77, 1216 – 1227. doi:10.1002/art.43164
  43. Liu R., Zhang Z., Kyaw A. K., Grabińska K. A., Shah H., Shen H. (2025). Cellular pan-chain acyl-CoA profiling reveals SLC25A42/SLC25A16 in mitochondrial CoA import and metabolism. Nat. Metab. 7, 1871 – 1888. doi:10.1038/s42255-025-01358-y
  44. Liu Y., Da W., Xu M.-J., Xiao C.-X., Deng T., Zhou S.-L. (2025). Single-cell transcriptomics reveals novel chondrocyte and osteoblast subtypes and their role in knee osteoarthritis pathogenesis. Signal Trans. Targ. Ther. 10, 40. doi:10.1038/s41392-025-02136-8
  45. Liu Z., Jiang H., Zhang Z., Jiang R., Wang R., Li J. (2025). SESN2 maintains cartilage homeostasis by SREBP1-mediated lipid metabolism during osteoarthritis progression. Iscience 28, 113097. doi:10.1016/j.isci.2025.113097
  46. Liu B., Liang Y., Wang C., Weng Z., Yang Y., Shi Y. (2026). Dimeric PKM2 in chondrocytes impairs mitochondrial homeostasis in osteoarthritis. Cell Death Dis. 17 (1), 370. doi:10.1038/s41419-026-08621-4
  47. Long D., Deng Z., Zhao X., Xu Y., Li W., Mo X. (2025). m7G-modified mt-tRF3b-LeuTAA regulates mitophagy and metabolic reprogramming via SUMOylation of SIRT3 in chondrocytes. Biomaterials 314, 122903. doi:10.1016/j.biomaterials.2024.122903
  48. Lu Y., Liu Y., Xia X., Wu Y., Deng Y., Yu C. (2025). Trojan horse-inspired spatiotemporal strategy augments cartilage regeneration by enhancing mitochondrial energy production. Innovation 6, 100913. doi:10.1016/j.xinn.2025.100913
  49. Mei Z., Yilamu K., Ni W., Shen P., Pan N., Chen H. (2025). Chondrocyte fatty acid oxidation drives osteoarthritis via SOX9 degradation and epigenetic regulation. Nat. Commun. 16, 4892. doi:10.1038/s41467-025-60037-4
  50. Miner G. E., Cohen S. (2024). Protocol for monitoring fatty acid trafficking from lipid droplets to mitochondria in cultured cells. Star. Protoc. 5, 103236. doi:10.1016/j.xpro.2024.103236
  51. Miner G. E., So C. M., Edwards W., Ragusa J. V., Wine J. T., Gutierrez D. W. (2023). PLIN5 interacts with FATP4 at membrane contact sites to promote lipid droplet-to-mitochondria fatty acid transport. Dev. Cell 58, 1250 – 1265.e1256. doi:10.1016/j.devcel.2023.05.006
  52. Motamedi S., Ravoet N., Dehairs J., Vanderhoydonc F., Escamilla-Ayala A., Sliwinska M. A. (2025). AMP-activated protein kinase-driven lipid droplet dynamics govern melanoma sensitivity to polyunsaturated fatty acid and iron-induced ferroptosis. Nat. Commun. 16, 11252. doi:10.1038/s41467-025-66113-z
  53. Nielsen R. L., Monfeuga T., Kitchen R. R., Egerod L., Leal L. G., Schreyer A. T. H. (2024). Data-driven identification of predictive risk biomarkers for subgroups of osteoarthritis using interpretable machine learning. Nat. Commun. 15, 2817. doi:10.1038/s41467-024-46663-4
  54. Ouyang Q., Chen Q., Ke S., Ding L., Yang X., Rong P. (2023). Rab8a as a mitochondrial receptor for lipid droplets in skeletal muscle. Dev. Cell 58, 289 – 305.e286. doi:10.1016/j.devcel.2023.01.007
  55. Park S., Baek I.-J., Ryu J. H., Chun C.-H., Jin E.-J. (2022). PPARα− ACOT12 axis is responsible for maintaining cartilage homeostasis through modulating de novo lipogenesis. Nat. Commun. 13, 3. doi:10.1038/s41467-021-27738-y
  56. Perry T., Deng Y., Hulley P., Maciewicz R., Mitchelmore J., Larsson S. (2026). Large-scale molecular endotype discovery in synovial fluid reveals osteoarthritis as a single biological continuum. Nat. Commun. 17, 4721. doi:10.1038/s41467-026-71632-4
  57. Raut R. D., Chakraborty A. K., Neogi T., Albro M., Snyder B., Schaer T. P. (2025). A multi-tissue human knee single-cell atlas identifies that osteoarthritis reduces regenerative tissue stem cells while increasing inflammatory pain macrophages. Commun. Biol. 8, 1146. doi:10.1038/s42003-025-08586-8
  58. Schurman C. A., Bons J., Woo J. J., Yee C., Liu Q., Tao N. (2026). Tissue and extracellular matrix remodeling of the subchondral bone during osteoarthritis of knee joints as revealed by spatial mass spectrometry imaging. Bone Res. 14, 14. doi:10.1038/s41413-025-00495-0
  59. Shao Y., Zhang H., Guan H., Wu C., Qi W., Yang L. (2024). PDZK1 protects against mechanical overload-induced chondrocyte senescence and osteoarthritis by targeting mitochondrial function. Bone Res. 12, 41. doi:10.1038/s41413-024-00344-6
  60. Sun W., Lv Z., Li W., Lu J., Xie Y., Wang P. (2024). XJB-5-131 protects chondrocytes from ferroptosis to alleviate osteoarthritis progression via restoring Pebp1 expression. J. Orthop. Transl. 44, 114 – 124. doi:10.1016/j.jot.2023.12.005
  61. Tang S. A., Yao L., Ruan J., Kang J., Cao Y., Nie X. (2024). Single-cell atlas of human infrapatellar fat pad and synovium implicates APOE signaling in osteoarthritis pathology. Sci. Transl. Med. 16, eadf4590. doi:10.1126/scitranslmed.adf4590
  62. Vega-Letter A. M., García-Guerrero C., Yantén-Fuentes L., Pradenas C., Herrera-Luna Y., Lara-Barba E. (2025). Safety and efficacy of mesenchymal stromal cells mitochondria transplantation as a cell-free therapy for osteoarthritis. J. Transl. Med. 23, 26. doi:10.1186/s12967-024-05945-7
  63. Wang J., Fang N., Xiong J., Du Y., Cao Y., Ji W.-K. (2021). An ESCRT-dependent step in fatty acid transfer from lipid droplets to mitochondria through VPS13D− TSG101 interactions. Nat. Commun. 12, 1252. doi:10.1038/s41467-021-21525-5
  64. Wang M., Tan G., Jiang H., Liu A., Wu R., Li J. (2022). Molecular crosstalk between articular cartilage, meniscus, synovium, and subchondral bone in osteoarthritis. Bone Jt. Res. 11, 862 – 872. doi:10.1302/2046-3758.1112.bjr-2022-0215.r1
  65. Wang C., Fang L., Shi M., Niu X., Li T., Li X. (2025a). NFIA regulates articular chondrocyte fatty acid metabolism and joint homeostasis. Sci. Transl. Med. 17, eadm9488. doi:10.1126/scitranslmed.adm9488
  66. Wang C., Zeng Z., Wang T., Xie Z., Zhang J., Dong W. (2025b). Unraveling the spatial and signaling dynamics and splicing kinetics of immune infiltration in osteoarthritis synovium. Front. Immunol. 16, 1521038. doi:10.3389/fimmu.2025.1521038
  67. Wang Y.-T., Moura A. K., Zuo R., Wang Z., Roudbari K., Hu J. Z. (2025). Defective lipid droplet biogenesis exacerbates oleic acid-induced cellular homeostasis disruption and ferroptosis in mouse cardiac endothelial cells. Cell Death Disc. 11, 374. doi:10.1038/s41420-025-02669-5
  68. Wang Y., Song P., Zhou H., Wang P., Li Y., Shao Z. (2025). An AI-assisted fluorescence microscopic system for screening mitophagy inducers by simultaneous analysis of mitophagic intermediates. Nat. Comm. 16, 5179. doi:10.1038/s41467-025-60315-1
  69. Wei J., Pi C., Qi Y., Yang W., Sun J., Xu X. (2026). TGF-β2 drives lipid droplet accumulation in chondrocytes through the TβRI/p-Smad3/Fabp5 axis. J. Biol. Chem. 302, 113317. doi:10.1016/j.jbc.2026.113317
  70. Wessendorf-Rodriguez K., Ruchhoeft M. L., Murray C. W., Huang Y., Ashley E. L., Galvez H. M. (2026). Modeling lipid homeostasis using stable isotope tracing and flux analysis. Cell Metab. 38, 460 – 473.e466. doi:10.1016/j.cmet.2026.01.020
  71. Wu Y., Hou M., Deng Y., Xia X., Liu Y., Yu J. (2025). Swimming exercise induces redox-lipid crosstalk to ameliorate osteoarthritis progression. Redox Biol. 81, 103535. doi:10.1016/j.redox.2025.103535
  72. Xie Y., Lv Z., Li W., Lin J., Sun W., Guo H. (2025). JP4-039 protects chondrocytes from ferroptosis to attenuate osteoarthritis progression by promoting Pink1/Parkin-dependent mitophagy. J. Ortho. Transl. 51, 132 – 144. doi:10.1016/j.jot.2025.01.001
  73. Xu H., Li Y., Wang J., Xin Y., Xu Q., Qin X. (2026). Astragalus membranaceus polysaccharides attenuate osteoarthritis in human primary chondrocytes and in animal model: mechanism study using RNA-seq and network pharmacology. Phytother. Res. 40, 2249 – 2269. doi:10.1002/ptr.70270
  74. Xu K., Li J., Wang G., Lin J., Li J. (2026). Lipid metabolic reprogramming in osteoarthritis: cartilage-centered mechanisms, whole-joint interactions, and therapeutic challenges. Front. Endocrinol. 17, 1825725. doi:10.3389/fendo.2026.1825725
  75. Xue H., Hu X., Yang Z., Cao Y., Zhu Q., Zhao H. (2026). PISD acts as a switch between lipophagy and fatty acid transfer to mitochondria. J. Cell Biol. 225, e202605090. doi:10.1083/jcb.202605090
  76. Yu J., Liu Q., Zhang Y., Xu L., Chen X., He F. (2025). Stress causes lipid droplet accumulation in chondrocytes by impairing microtubules. Osteoarthr. Cartil. 33, 351 – 363. doi:10.1016/j.joca.2024.08.015
  77. Zhang Y., Liu Y., Hou M., Xia X., Liu J., Xu Y. (2023). Reprogramming of mitochondrial respiratory chain complex by targeting SIRT3-COX4I2 axis attenuates osteoarthritis progression. Adv. Sci. 10, 2206144. doi:10.1002/advs.202206144
  78. Zhang Y., Jiang H., Du J., Ming J., Li Y., Ma Y. (2024). Nanoparticle-activated SIRT3/SOD2 axis involves amelioration of chondrocyte senescence to delay osteoarthritis progression. Chem. Eng. J. 497, 154799. doi:10.1016/j.cej.2024.154799
  79. Zhang M.-N., Duan R., Chen G.-H., Chen M.-J., Hong C.-G., Wang X. (2025). Fasting activates optineurin-mediated mitophagy in chondrocytes to protect against osteoarthritis. Commun. Biol. 8, 68. doi:10.1038/s42003-025-07541-x
  80. Zhang Y., Fu M., Zhou C., Wang X., Jiang Z., Jiang C. (2025). Cold exposure promotes the progression of osteoarthritis through downregulating APOE in cartilage. EMBO Mol. Med. 17, 2137 – 2162. doi:10.1038/s44321-025-00268-6
  81. Zhao P., Tian R., Song D., Zhu Q., Ding X., Zhang J. (2025). Rab GTPases are evolutionarily conserved signals mediating selective autophagy. J. Cell Biol. 224, e202410150. doi:10.1083/jcb.202410150
  82. Zhou Q., Ghorasaini M., Cornelis F. M., Assi R., De Roover A., Giera M. (2025). Lipidomics unravels lipid changes in osteoarthritis articular cartilage. Ann. Rheum. Dis. 84, 1264 – 1276. doi:10.1016/j.ard.2025.01.009
  83. Zhu L., Diao X., Yuan C., Lau C. M. L., Wang J., Wu W. (2025). Integrative spatially resolved proteomic and metabolomic imaging reveals synovitis endotypes implicated in osteoarthritis progression. Theranostics 15, 9729 – 9741. doi:10.7150/thno.117788
  84. Zhuang H., Ren X., Zhang Y., Li H., Zhou P. (2024). β-hydroxybutyrate enhances chondrocyte mitophagy and reduces cartilage degeneration in osteoarthritis via the HCAR2/AMPK/PINK1/Parkin pathway. Aging Cell 23, e14294. doi:10.1111/acel.14294
  85. Zou Z., Hu W., Kang F., Xu Z., Li Y., Zhang J. (2025). Interplay between lipid dysregulation and ferroptosis in chondrocytes and the targeted therapy effect of metformin on osteoarthritis. J. Adv. Res. 69, 515 – 529. doi:10.1016/j.jare.2024.04.012

Citation

Meng, G., et al. (2026). Lipid droplet–mitochondria crosstalk in osteoarthritis: linking lipid routing to metabolic mismatch. Frontiers in Molecular Biosciences, Volume 13. https://doi.org/
Received
Accepted
Published
Volume
13 - 2026

Copyright

Copyright © 2026 Meng, Hou, Dong, Chen, Li, Hu, Du, Wang and Zhong.

License: https://creativecommons.org/licenses/by/4.0/

*Correspondence: Ping Dong

Article metrics

Views
0
Downloads
0
Citations
0
Shares
0

Related articles

ReviewWhen absence becomes functional: lysine deserts in protein regulation and evolutionNatalia A. SzulcOriginal ResearchExplainable machine learning integrating CT-derived intratumoral heterogeneity for histological subtype prediction in pure ground-glass lung adenocarcinoma: multicenter validation and transcriptomic characterizationYing Deng, Jiajun Liu, Dinghua Li, Bingyi Zhang, Bin Xiang, Wanyin Qi, Sanhong ZhangOpinionOn the physics of chromatin molecular complex: from phase-separation to membrane adsorptionAndrea Fontana, Andrea M. Chiariello