In OA, factors such as synovial inflammation contribute to elevated levels of free fatty acids (FFAs) in the joint. Under normal conditions, autophagy can degrade these excess lipids, maintaining cellular homeostasis. However, when autophagic function is impaired, FFAs accumulate within chondrocytes, where they are oxidized by mitochondria into toxic lipid metabolites like ceramides. These metabolites trigger mitochondrial dysfunction, causing increased ROS production, which further induces cytochrome C release and activates the caspase pathway, ultimately leading to chondrocyte apoptosis (Zhang, 2015). Studies have shown that in mice fed a high-fat diet or chondrocytes exposed to FFAs, autophagic activity is suppressed, resulting in the accumulation of interferon-induced gene stimulator 1 (STING1). This activation leads to the stimulation of the STING1-TBK1-IRF3 and MAP2K3/MKK3-MAPK/p38 signaling pathways, both of which trigger ECM degradation in cartilage (Kang et al., 2025). Furthermore, cholesterol homeostasis plays a crucial role in skeletal development, and dysregulated cholesterol levels have been linked to the progression of cartilage diseases, including OA. When autophagic function is impaired, the ability to clear intracellular cholesterol is reduced. Excess cholesterol becomes oversaturated, forming tiny cholesterol crystals that can be recognized by the NOD-like receptor pyrin domain-containing 3 (NLRP3). The activation of NLRP3 promotes the release of pro-inflammatory cytokines, creating a local inflammatory environment that accelerates ECM degradation and chondrocyte death (Papathanasiou et al., 2021). Research has demonstrated that five cholesterol-lowering drugs can enhance cholesterol efflux, effectively preventing cartilage degeneration. This suggests that restoring proper lipid and cholesterol homeostasis through targeted therapies could offer a promising approach to mitigating the progression of OA (Lee et al., 2025).

Pathological cartilage calcification plays a crucial role in the progression of OA. The final step of the autophagic process involves the fusion of autophagosomes with lysosomes to form autolysosomes, where degradation occurs. However, when fusion fails or lysosomal function is impaired, undegraded autophagosomes accumulate within the cell. Research has shown that these stalled autophagosomes are rich in calcium, phosphate ions, and osteogenic differentiation markers (Pei et al., 2022). Upon cell death, these contents are released into the extracellular matrix (ECM), becoming the core of pathological calcification and promoting the formation of hydroxyapatite crystals, which further contribute to the mechanical wear of cartilage. In the early stages of OA, microtubule destabilization by histone deacetylase six hinders the fusion of autophagosomes with lysosomes, causing calcium-containing vesicles labeled by LC3 to be released into the matrix. This process triggers pathological calcification and accelerates cartilage degradation (Yan et al., 2022). Moreover, normal autophagy inhibits the abnormal differentiation of chondrocytes by degrading runt-related transcription factor 2 (Runx2). When autophagy is suppressed, proteins like Runx2 accumulate in the cell, driving the pathological phenotypic transformation of chondrocytes. These transformed cells exhibit osteoblast-like characteristics, such as high expression of alkaline phosphatase (ALP) and secretion of type I collagen rather than type II collagen, actively promoting subchondral bone sclerosis and cartilage matrix calcification. This disruption undermines the mechanical and lubricating properties unique to cartilage (Komori, 2022).

Aging significantly impacts autophagy-mediated chondrocyte vitality. Studies have shown that autophagy-related proteins are highly expressed in normal chondrocytes, whereas their expression is substantially reduced in elderly populations, leading to a marked decrease in autophagic activity (Caramés et al., 2010). Concurrently, with advancing age, the number of lysosomes and their enzymatic activity decline, severely impairing the fusion efficiency of autophagosomes with lysosomes, as well as the degradation of autophagic substrates. Even when autophagosomes form properly, effective recycling cannot be achieved (Fujimaki et al., 2019). The diminished autophagic activity results in the accumulation of damaged organelles and macromolecules within the cells, ultimately compromising chondrocyte vitality and triggering age-related OA (Bouderlique et al., 2016).

Under mechanical stimulation, moderate stress can upregulate the expression of autophagy-related proteins, inducing protective autophagy and aiding in the clearance of damaged molecules produced during the mechanical stress (Vinatier et al., 2018). However, excessively high-intensity stimuli suppress autophagic activity, leading to cellular apoptosis (Cecconi and Levine, 2008). Additionally, when prolonged stress exceeds the cell’s compensatory capacity, mitochondrial damage occurs in chondrocytes, resulting in a significant increase in ROS and ultimately triggering OA (López de Figueroa et al., 2015). Furthermore, studies have reported that during physiological activities, the expression of the cartilage-specific transcription factor SOX9 in circulating mesenchymal progenitor cells, associated with autophagy, is elevated, suggesting the beneficial role of physical exercise in protecting cartilage formation (Dalle Carbonare et al., 2019).

Clinical and epidemiological evidence increasingly associates circadian rhythm disturbances with heightened OA risk and severity, positioning circadian disruption as a modifiable environmental and genetic risk factor. Shift work chronically disrupts endogenous rhythms and correlates with a higher prevalence of knee OA and accelerated joint degeneration (Berenbaum and Meng, 2016). Sleep disorders such as insomnia and obstructive sleep apnea, often accompanied by circadian misalignment, are common in OA patients and may exacerbate pain and functional impairment (Sánchez Romero et al., 2022). Additionally, genetic polymorphisms in core clock genes (BMAL1, CLOCK) have been linked to OA susceptibility and progression in human cohorts, underscoring the translational relevance of circadian biology in OA (Yu et al., 2019). Mechanistically, disruption of circadian rhythms impairs the temporal orchestration of autophagy, leading to dysfunctional autophagic flux in chondrocytes. Core clock genes, including BMAL1 and CLOCK, directly regulate the transcription of several autophagy-related genes and key autophagy regulators such as TFEB and FOXO3, whose expression exhibits circadian oscillations (Dudek et al., 2016). BMAL1, in particular, binds to enhancer box sequences in the promoters of various ATG genes, driving their rhythmic expression (Dudek et al., 2016). In cartilage-specific BMAL1 knockout mice, loss of rhythmic transcriptional activation of downstream autophagy genes reduces autophagic activity, leading to accelerated cartilage degeneration (Dudek et al., 2016). Circadian misalignment dampens these transcriptional rhythms, reduces basal autophagic activity, and impairs the clearance of damaged organelles and protein aggregates. Moreover, circadian disruption alters the circadian gating of mTOR and AMPK signaling pathways, further compromising autophagy induction and progression (Zhang W. et al., 2022). Other core circadian components also modulate autophagy: the CLOCK: BMAL1 heterodimer drives the expression of period (PER) and cryptochrome (CRY) genes, which feedback to inhibit CLOCK: BMAL1 activity. PER/CRY complexes interact with autophagy regulators; for instance, CRY1 stabilizes the autophagy-initiating ULK1 complex, and its loss leads to premature autophagy activation (Lane et al., 2023). Conversely, REV-ERBα/β repress the transcription of several autophagy and lysosomal genes by binding to RORE elements in their promoters, and pharmacological activation of REV-ERBα suppresses excessive autophagy and ameliorates tissue damage in inflammatory conditions (Lee and Kim, 2013). In chondrocytes, coordinated expression of these circadian genes ensures temporally controlled autophagy, optimizing cellular cleanup and energy homeostasis. Disruption of any component can lead to autophagic imbalance, contributing to OA pathogenesis through the accumulation of cellular debris, enhanced oxidative stress, and accelerated chondrocyte senescence and apoptosis, thereby driving cartilage degradation (Dudek et al., 2017).

MiRNAs are single-stranded non-coding RNA molecules consisting of 19–24 nucleotides that mediate post-transcriptional gene silencing by targeting mRNA, playing pivotal roles in various physiological processes (Diener et al., 2022). Studies have shown that many miRNAs participate in the regulation of autophagy in OA chondrocytes, thereby influencing cartilage degeneration (Table 1). For instance, miR-302d-3p and miR-155 target ULK1, inhibiting the initiation of autophagy in chondrocytes and promoting chondrocyte apoptosis, thus exacerbating cartilage degeneration (Wang et al., 2019). Additionally, several miRNAs influence the progression of cartilage degeneration by regulating ATG proteins. For example, miR-155 targets ATG14, ATG5, and ATG3, leading to defects in chondrocyte autophagy (D’Adamo et al., 2016). miR-128 inhibits the expression of ATG12, suppressing autophagy in chondrocytes and accelerating cartilage degeneration (Lian et al., 2018). miR-375 targets ATG2B, inhibiting chondrocyte autophagy, promoting ERS and chondrocyte apoptosis, and thereby aggravating OA (Li et al., 2020). miR-378 reduces the expression of ATG2A and SOX6, inhibiting autophagy in chondrocytes, thus suppressing chondrogenesis and worsening OA (Shi J. et al., 2025). In summary, miRNAs directly regulate autophagy-related proteins or signaling pathways involved in processes such as autophagy initiation and autophagosome formation, contributing to the regulation of autophagy in OA chondrocytes.

Furthermore, miRNAs can also modulate the autophagy regulatory network centered on mTOR and AMPK, influencing the process of cartilage degeneration. Studies have shown that miR-20 targets ATG10 while simultaneously promoting the PI3K/AKT/mTOR signaling pathway, thereby inhibiting autophagy in OA chondrocytes (He and Cheng, 2018). Similarly, miR-103-3p and miR-7 also affect the PI3K/AKT/mTOR signaling pathway (Li J. et al., 2025). In IL-1β-induced mouse chondrocytes, the upregulation of miR-103-3p suppresses the PI3K/AKT/mTOR signaling pathway, promoting chondrocyte autophagy and inhibiting ECM degradation. In contrast, both in vitro and in vivo studies have shown that miR-7 expression is reduced in OA rats, leading to the activation of the PI3K/AKT/mTOR signaling pathway, inhibiting chondrocyte autophagy, and accelerating cartilage degeneration (Zhou et al., 2020). Xu found that miR-31-5p promotes IL-1β-induced chondrocyte autophagy and proliferation in rats by inhibiting the SOX4/mTORC1 axis (Xu et al., 2022). AMPK, a positive regulator of autophagy, and mTOR, a negative regulator, also play a critical role in autophagy regulation. In OA cartilage samples, miR-34a-5p is significantly upregulated while sestrin-2 (SESN2) is notably decreased; this correlates with downregulated AMPK levels and upregulated mTORC1 levels. This suggests that miR-34a-5p inhibits SESN2, suppressing the AMPK pathway and activating the mTOR signaling pathway, thereby inhibiting chondrocyte autophagy (Wang J. et al., 2024). HIF-1α, an upstream protein of AMPK, is upregulated by miR-411, promoting autophagy in chondrocytes (D’Adamo et al., 2016). FOXO3, a downstream protein of AMPK, is suppressed by the upregulation of miR-155, inhibiting autophagy in chondrocytes (Yang et al., 2020). Additionally, in vitro cell experiments have demonstrated that miR-142-3p targets high mobility group protein (HMGB1), while miR-146b-5p targets tumor necrosis factor receptor-associated factor 6 (TRAF6), both promoting chondrocyte autophagy and proliferation (Wang et al., 2021; Lou et al., 2023). Upregulation of miR-140-5p and miR-149 targets fucosyltransferase 1 (FUT1), stimulating chondrocyte autophagy (Wang et al., 2018). In conclusion, miRNAs have been clearly shown to participate in the regulation of autophagy in OA chondrocytes, holding potential as new therapeutic targets for OA treatment. However, the specific mechanisms and development of their clinical application still require further investigation.

LncRNAs are RNA molecules longer than 200 nucleotides, characterized by relatively low sequence conservation but high tissue/cell specificity in the expression. Their role is more complex than that of miRNAs, as they can regulate gene expression at various levels-pre-transcriptional, transcriptional, and post-transcriptional - and are involved in a wide range of diseases (Herman et al., 2022). As research into lncRNAs deepens, several have been shown to regulate autophagy in OA chondrocytes, and some have been implicated in the development of OA by interacting with miRNAs to influence the process of cartilage degeneration (Table 2). For instance, in an anterior cruciate ligament transection (ACLT) rat model of OA, the expression of lncRNA-H19 was upregulated. This lncRNA promotes chondrocyte autophagy by targeting miR-148a, a mechanism that helps alleviate cartilage degradation. Conversely, in OA cartilage from human patients, the expression of lncRNA-HOTAIR was also found to be upregulated, but it inhibits autophagy by targeting miR-130a-3p, contributing to cartilage deterioration (Zhou et al., 2025). These examples illustrate how lncRNAs can function through complex interactions with miRNAs to either promote or inhibit autophagy in chondrocytes, highlighting their potential as therapeutic targets in OA. The precise roles of lncRNAs and their mechanisms of action in OA still require further exploration, but their potential is promising for modulating autophagy and influencing cartilage homeostasis (He and Jiang, 2020).

Furthermore, numerous lncRNAs influence the signaling pathways involving miRNAs that regulate autophagy in chondrocytes. In a mouse model of OA induced by ACLT, lncRNA-NEAT1 promotes chondrocyte autophagy and proliferation, inhibits apoptosis, and alleviates OA symptoms by modulating the miR-122-5p/Sesn2/Nrf2 signaling pathway (Zhang and Jin, 2022). Another study demonstrated that in a rat OA model induced by ACLT, lncRNA-GAS5 inhibits chondrocyte apoptosis, promotes chondrocyte autophagy, and accelerates cartilage degeneration through the miR-144/mTOR axis (Ji et al., 2021). In both in vitro and in vivo studies, lncRNA-POU3F3 promotes chondrocyte autophagy, suppresses apoptosis, and reduces inflammation by regulating the POU3F3/miR-29a-3p/FOXO3 axis (Shi et al., 2022). Moreover, lncRNA-PVT1 inhibits chondrocyte autophagy in an in vitro cellular model via the miR-27b-3p/TRAF3 axis (Lu et al., 2020). Several studies have also revealed that lncRNA-SNHG1 activates the PI3K/AKT signaling pathway both in vivo and in vitro, suppressing chondrocyte autophagy and accelerating cartilage degeneration (Wang Q. et al., 2024). Elevated levels of lncRNA-GCH1 were found in the cartilage of OA patients, where it activates the PTEN 1/Parkin signaling pathway, promotes mitochondrial autophagy, improves mitochondrial function, and inhibits cartilage degradation (Zeng et al., 2025). Consequently, the regulation of these genes may hold significant potential for gene therapy in OA.

CircRNAs are a class of non-coding RNA molecules that form a closed circular structure through back-splicing of pre-mRNA. Due to the lack of a 5′ cap structure and a 3′ poly(A) tail, circRNAs are resistant to degradation by exonucleases, leading to a much longer half-life compared to linear mRNA, making them highly stable within the cell (Xu et al., 2024). Furthermore, many circRNAs exhibit tissue specificity and disease-relatedness, making them a prominent focus of research in the ncRNA field in recent years. CircRNAs can function as competing endogenous RNAs (ceRNAs), competing with miRNAs to influence the stability or translation of target RNAs, thereby regulating gene expression at the transcriptional level. For example, in the cartilage tissue of OA patients and LPS-induced OA chondrocytes, the expression of circ-MSR is upregulated and interacts with miR-761 to inhibit autophagy in chondrocytes (Xu et al., 2024). Similarly, circ-Pan3 modulates the miR-667-5p/ghrelin axis, inhibiting autophagy in chondrocytes and accelerating cartilage degeneration (Zeng et al., 2021). Additionally, both in vivo and in vitro studies have confirmed that upregulated ciRS-7 activates the PI3K/AKT/mTOR signaling pathway, inhibiting autophagy in chondrocytes and exacerbating cartilage degeneration (Zhou et al., 2020). Moreover, circRNAs regulate autophagy through their interactions with autophagy-related proteins. Circ-SPG21 interacts with Parkin to promote mitochondrial autophagy (Zhang et al., 2025c), while circ-TBCK affects LC3 to inhibit autophagy, thereby accelerating cartilage degeneration. Other circRNAs involved in the regulation of autophagy in OA chondrocytes are listed in Table 3.

Exosomes derived from mesenchymal stem cells serve as natural nanocarriers, characterized by excellent biocompatibility, low immunogenicity, and inherent cartilage-targeting properties, making them an ideal platform for delivering autophagy modulators in OA therapy. On one hand, exosomes are rich in miRNAs that can regulate autophagy; for instance, miR-98-5p carried by exosomes activates protective autophagy by targeting transcription factor 6 (Xu and Zhang, 2025). On the other hand, exosomes function as delivery systems, capable of loading small molecule autophagy drugs or mRNA via techniques such as electroporation, ultrasound, or co-incubation, thereby enabling targeted drug delivery. Moreover, surface modification of exosomes with cartilage-targeting peptides allows for the active, targeted delivery of drugs to the affected cartilage, significantly enhancing local drug concentration while minimizing systemic exposure (Kuang et al., 2025).

Intelligent hydrogels or nanofiber scaffolds can serve as carriers for autophagy modulators. By encapsulating small molecule autophagy activators within responsive biomaterials, drugs can be released on demand at the site of pathology. When inflammation or an acidic microenvironment arises within the OA joint cavity, the material degrades more rapidly, releasing the drug to activate autophagy, thereby suppressing inflammation and protecting chondrocytes (He et al., 2025). The properties of the material itself can also indirectly influence cellular autophagy. Studies have shown that biomaterials with specific topological structures or mechanical properties can affect autophagy levels through mechanotransduction signals (Feng et al., 2025). By utilizing such materials, a synergistic effect between the materials and the loaded drugs can be achieved, promoting chondrocyte anabolic metabolism and maintaining homeostasis more effectively.

The traditional drug development model, when applied to the complex and dynamic process of autophagy, significantly increases both the cost and time required for research. Artificial intelligence (AI) and machine learning (ML), by mining vast biomedical datasets, have the potential to accelerate the development of autophagy modulators. Initially, deep learning algorithms can quickly identify lead compounds with potential regulatory activity by analyzing the chemical structures and bioactivity data of known autophagy modulators. Furthermore, AI can design novel molecules with entirely new structures based on specific targets, thereby enhancing the efficiency of drug discovery. Additionally, considering the heterogeneity of OA patients, AI algorithms can identify distinct OA subtypes, aiding in the selection of the most suitable patient populations for clinical trials, thus advancing personalized medicine (Zitnik et al., 2018; Stokes et al., 2020). Therefore, the future of OA research requires the establishment of a more comprehensive multi-omics database for autophagy and the development of more interpretable AI models to better understand the biological logic behind their predictions.