Dual-modal antioxidant and epigenetic synergy attenuates the self-perpetuating senescence cycle in osteoarthritis

Abstract

Osteoarthritis (OA) arises from chondrocyte senescence driven by intertwined oxidative stress and abnormal m6A methylation, with few treatments targeting both pathological pathways. Lycopene, an antioxidant, is limited by poor bioavailability, whereas Wilms tumor 1-associating protein (WTAP), a core m6A methyltransferase, has no specific inhibitors. Herein, we fabricated cartilage-targeted HPcLW nanoparticles (∼250 nm) via electrostatic self-assembly of human serum albumin (HSA) and poly-L-lysine (PLL) with collagen II-binding peptide, co-loaded with lycopene and WTAP siRNA. The targeting modification extended joint fluorescence retention to 10 days after intra-articular injection with good serum stability and biosafety. In aged mice and medial meniscus (DMM)-induced OA mouse models, intra-articular HPcLW attenuated cartilage degeneration, restored COL2 expression, and suppressed MMP13 levels. Mechanistically, WTAP siRNA suppressed m6A modification to downregulate PAI-1 expression, while lycopene scavenged ROS and protected siRNA integrity, cooperatively disrupting the WTAP/PAI-1 axis and alleviating mitochondrial dysfunction. By integrating antioxidant and epigenetic strategies, HPcLW counteracts the senescence loop, establishing the WTAP/PAI-1 axis as a therapeutic target and highlighting co-delivery nanomedicine for age-related joint diseases.

Graphical abstract

Highlights

• •A cartilage-targeting nanoplatform (HPcLW) is developed via HSA/PLL electrostatic self-assembly for co-delivery of lycopene and WTAP siRNA.
• •HPcLW achieves synergistic inhibition of the WTAP/PAI-1 epigenetic axis, combining antioxidant therapy with m6A epigenetic modulation.
• •Intra-articular HPcLW attenuates chondrocyte senescence, restores cartilage matrix homeostasis, and slows OA progression.
• •WTAP is identified as a critical m6A methyltransferase that drives PAI-1 expression via m6A modification.
• •The work establishes a new paradigm for targeting the oxidative-epigenetic loop in osteoarthritis using bioactive nanocarriers.

1. Introduction

Osteoarthritis (OA) is a prevalent degenerative joint disease characterized by progressive cartilage erosion, subchondral bone remodeling, and chronic synovial inflammation, representing a leading cause of disability in the elderly [1]. Despite advances in understanding OA pathogenesis, current clinical management is largely limited to symptomatic relief with analgesics or anti-inflammatory agents, necessitating total joint replacement for end-stage disease [2]. Current therapies offer only symptomatic relief, and no disease-modifying drugs exist, largely due to poor bioavailability and off-target distribution [2,3] (see Scheme 1).

Scheme 1.

Schematic illustration of the synthesis and application of HPcLW nanoparticles. (a) Synthesis of HPcLW. (b) Biological mechanism of HPcLW nanoparticles in alleviating chondrocyte senescence in OA. After intra-articular injection, WTAP siRNA downregulates PAI-1 expression by inhibiting WTAP-mediated m6A modification, while lycopene scavenges ROS and partially suppresses the WTAP regulatory axis. This synergistic action alleviates mitochondrial damage, suppresses senescence, and attenuates OA.

Accumulating evidence has established chondrocyte senescence as a central driver of OA progression, marked by the senescence-associated secretory phenotype (SASP) that promotes matrix degradation and spreads senescence to neighboring cells, establishing a vicious cycle [[4], [5], [6], [7]]. Within this pathological niche, oxidative stress and epigenetic dysregulation do not operate in isolation but rather converge to sustain this senescent phenotype [8,9]. Reactive oxygen species (ROS), a hallmark of the OA joint, serve as both inducers and executors of senescence, driving DNA damage, mitochondrial dysfunction, and SASP activation. Concurrently, aberrant epigenetic modifications, particularly N⁶-methyladenosine (m6A) RNA methylation, have emerged as critical regulators of senescence-associated gene expression [10,11]. Notably, oxidative stress modulates m6A modification, and aberrant m6A in turn amplifies oxidative damage, forming a reinforcing loop [[12], [13], [14]]. This interdependence implies that targeting a single node may be insufficient to dismantle the self-perpetuating senescence cycle in OA. Given this interconnected pathology, an ideal therapeutic strategy should simultaneously intervene in both oxidative stress and epigenetic dysregulation. Lycopene is a potent antioxidant that neutralizes ROS, stabilizes mitochondrial function, and attenuates senescence [[15], [16], [17], [18]]. However, its clinical translation is severely limited by poor water solubility, susceptibility to oxidation, and low bioavailability [15,19]. In intra-articular therapy, these limitations are further exacerbated by the avascular nature of cartilage, which restricts drug diffusion into chondrocytes [20].

On the epigenetic front, our previous research found WTAP is highly expressed in OA cartilage. According to the GEO dataset (GSE249509), we also confirmed that WTAP expression is markedly elevated in the cartilage of aged mice [21]. As a crucial m6A writer, WTAP facilitates the deposition of m6A modifications on target transcripts, modulating their expression and downstream signaling pathways [22]. Gene silencing of WTAP has been shown to reduce cartilage degeneration in animal models. Despite its potential as a therapeutic target, no specific small-molecule inhibitor against WTAP is currently available. Given the challenges of developing inhibitors for epigenetic “writer” proteins, RNA interference (RNAi) has gained increasing attention as an alternative strategy for targeted gene silencing. In OA, intra-articular delivery of small interfering RNA (siRNA) has been successfully employed to silence disease-associated genes, demonstrating therapeutic potential with acceptable safety profiles [23]. However, the efficacy of WTAP siRNA may be compromised by the ROS-rich OA microenvironment, which destabilizes nucleic acids and sustains downstream pro-senescence signaling [24]. This limitation underscores the need for a combinatorial strategy that not only delivers siRNA but also modulates the oxidative milieu to maximize therapeutic outcome.

Thus, co-delivery of WTAP siRNA and lycopene offers a synergistic strategy to simultaneously target the oxidative and epigenetic drivers of chondrocyte senescence. However, the translational realization of this dual-agent strategy faces formidable delivery barriers: lycopene suffers from poor aqueous solubility and chemical instability, while siRNA requires protection from enzymatic degradation and efficient intracellular delivery. These challenges are further compounded by the avascular nature of articular cartilage, which restricts passive diffusion into chondrocyte-rich regions [18]. To overcome these barriers, we developed a human serum albumin (HSA)-poly-L-lysine (PLL) nanocarrier, which offers biocompatibility, siRNA complexation, and lycopene loading [[25], [26], [27], [28]]. This platform offers mild, solvent-free fabrication, a critical advantage for preserving lycopene's antioxidant activity, and enables surface functionalization with targeting ligands for enhanced cartilage accumulation [26]. By conjugating a collagen II-binding peptide (CollBP) to the PLL-PEG backbone, we achieved cartilage matrix targeting, ensuring preferential accumulation in chondrocyte-rich regions. The resulting nanocarrier thus provides a versatile platform for the co-loading and targeted co-delivery of lycopene and WTAP siRNA, enabling us to test the central hypothesis that simultaneously mitigating oxidative stress and silencing WTAP-mediated m6A dysregulation can synergistically attenuate the self-perpetuating chondrocyte senescence cycle.

In this study, we constructed this cartilage-targeting nanosystem (termed HPcLW) and systematically evaluated the physicochemical properties, biosafety, and therapeutic efficacy in both naturally aged and surgically induced OA mouse models, as well as in senescent chondrocytes. By integrating antioxidant therapy with epigenetic modulation, this study presents a paradigm for OA treatment that synergistically targets the interconnected drivers of chondrocyte senescence, highlighting the broader potential of multipronged strategies designed to attenuate self-reinforcing pathological loops.

2. Results

2.1. Construction and Characterization of HPcLW nanoparticles

To simultaneously target the oxidative and epigenetic drivers of chondrocyte senescence, a cartilage-targeted co-delivery platform was developed for the synchronized delivery of lycopene and WTAP siRNA. Our previous research found that WTAP is highly expressed in OA cartilage.²¹According to the GEO dataset (GSE249509), we also confirmed that WTAP expression is markedly elevated in the cartilage of aged mice (Fig. S1). To address the interconnected delivery barriers, poor bioavailability of lycopene and the susceptibility of siRNA to enzymatic degradation, a cartilage-targeting nanoparticle system was first constructed via electrostatic self-assembly of HSA and PLL, incorporating PLL-PEG-CollBP for cartilage targeting, followed by crosslinking with dithiothreitol (DTT) at pH 9.0 [25,29]. The resulting nanoparticles (HPc) were then co-incubated with lycopene and WTAP siRNA overnight at low temperature to obtain the final HPcLW nanoparticles (Fig. 1a).

Fig. 1.

Construction and Characterization of HPcLW Nanoparticles. (a) Schematic illustration of the construction of cartilage-targeting nanoparticles loaded with lycopene and WTAP siRNA (HPcLW nanoparticles) based on HSA and PLL. (b) Photograph of free lycopene and HPcL nanoparticles after centrifugation. (c) Zeta potential of Blank, HPL, and HPcL nanoparticles (n = 3). (d) Zeta potential of HPcLW nanoparticles with different HSA (mg)/siRNA (μg) ratios (n = 4). (e, f) Encapsulation efficiency (e) and drug loading capacity (f) of Lycopene in HPL, HPcL, and HPcLW nanoparticles (n = 4). (g) Cumulative Lycopene siRNA release profile from the HPcLW at pH 7.4 and pH 5 (n = 6). (h, i) Representative TEM images (h) and size distribution (i) of Blank, HPL, HPcL, HPcW, and HPcLW nanoparticles. Scale bars, 100 nm.

As the key targeting component, PLL-PEG-CollBP was synthesized using a two-step method (Fig. S2a). First, CollBP, containing a free amino group (-NH₂), was conjugated to one distal NHS ester group of NHS-PEG₂₀₀₀-NHS via amide bond formation. Second, the resulting intermediate (NHS-PEG₂₀₀₀-CollBP) was coupled to PLL. Gel permeation chromatography (GPC) analysis showed that the copolymer exhibited a single symmetric peak with a number-average molecular weight (M_n) of 33855, a weight-average molecular weight (M_w) of 51738, and a polydispersity index (PDI, M_w/M_n) of 1.53, indicating high purity and a relatively narrow molecular weight distribution (Fig. S2b). The ¹H NMR spectrum also clearly exhibited characteristic proton signals of PLL, PEG, and CollBP segments in PLL-PEG-CollBP (Fig. S2c).

Based on the above nanoparticle design, a series of nanoparticles with defined compositions were prepared, including Blank nanoparticles (without PLL-PEG-CollBP, lycopene, or WTAP siRNA), HPL nanoparticles (lycopene-loaded), HPcL nanoparticles (lycopene and PLL-PEG-CollBP co-incorporated), HPcW nanoparticles (including PLL-PEG-CollBP and WTAP siRNA), and HPcLW nanoparticles (co-loaded with PLL-PEG-CollBP and WTAP siRNA) (Fig. S3). The successful incorporation of lycopene into HPcL nanoparticles was confirmed by centrifugation at 5000 rpm, where HPcL nanoparticles settled at the bottom of the tube, while the free lycopene remained in the supernatant (Fig. 1b). The surface charge of the nanoparticles was then characterized using zeta potential measurements. The Blank, HPL, HPcL, and HPcW nanoparticles exhibited potentials of 36.9 ± 3.2 mV, 30.3 ± 0.8 mV, 38.1 ± 1.3 mV, and 25.9 ± 2.0 mV, respectively (Fig. 1c and S4). The increased zeta potential of HPcL compared to Blank and HPL nanoparticles suggests successful surface presentation of the positively charged PLL-PEG-CollBP targeting moiety. Upon loading HPcL with WTAP siRNA, the zeta potential of HPcLW decreased to 22.3 ± 1.5 mV (Fig. 1d), indicating successful siRNA incorporation while maintaining a positive surface charge favorable for electrostatic interaction with negatively charged cell membranes.

The lycopene encapsulation efficiency and drug loading capacity of the nanoparticles were evaluated. HPL, HPcL, and HPcLW nanoparticles all exhibited encapsulation efficiencies exceeding 80% (Fig. 1e), with drug loading capacities of 24.7%, 27.5%, and 26.2%, respectively (Fig. 1f), as determined using a lycopene calibration curve at 475 nm (Fig. S5). To evaluate the release behavior under the weakly acidic OA microenvironment [30], the release profiles of lycopene from HPcLW nanoparticles were examined at pH 7.4 and 5.0. At pH 7.4, the cumulative release of lycopene reached 28.0% at 24 h and 40.7% at 72 h. In contrast, an accelerated release was observed at pH 5.0, with cumulative release values of 52.6% at 24 h and 69.9% at 72 h (Fig. 1g).

The loading efficiencies of WTAP siRNA in HPcW and HPcLW nanoparticles reached 92.3% and 95.3%, respectively, with a siRNA loading capacity of 47.6 μg per mg of HSA in HPcLW nanoparticles (Fig. S6a and b). Consistent with the release behavior of lycopene, the cumulative release of WTAP siRNA at 72 h was 42.6% at pH 7.4 and 83.8% at pH 5.0, demonstrating a substantially enhanced release under acidic conditions (Fig. S6c). Following RNase A treatment (100 ng/mL, 1 h), free siRNA was nearly entirely degraded, while 82.3% of the siRNA loaded in HPcLW nanoparticles was preserved, indicating that the nanoparticle formulation effectively protected siRNA from enzymatic degradation (Fig. S7).

Finally, the particle sizes of nanoparticles with different compositions were characterized. Transmission electron microscopy (TEM) and dynamic light scattering (DLS) analyses revealed that all prepared nanoparticles exhibited a relatively uniform size distribution (Fig. 1h and i). Specifically, HPcLW nanoparticles had an average diameter of approximately 250 nm. Collectively, these findings demonstrate the successful fabrication of HPcLW nanoparticles, providing a targeted co-delivery platform that simultaneously addresses the oxidative and epigenetic drivers of chondrocyte senescence.

2.2. Serum stability, biosafety, and targeting capability of HPcLW nanoparticles

The serum stability, biosafety, and cartilage-targeting capability of HPcLW nanoparticles were evaluated (Fig. 2a). To assess the serum stability, the size of HPcLW nanoparticles was monitored during incubation with 50% fetal bovine serum (FBS) over 24 h. The nanoparticles remained stable with no significant aggregation or size fluctuation, indicating good serum tolerance (Fig. 2b). Then, the biosafety of HPcLW nanoparticles was evaluated in vitro and in vivo. For the in vitro assessment, chondrocytes were treated with freshly prepared HPcLW nanoparticles or HPcLW nanoparticles stored at 25°C in the dark for 7 days for 24 and 48 h, and cell viability was determined using the CCK-8 assay. No significant cytotoxicity was observed for both freshly prepared (Fig. 2c and d) and stored HPcLW nanoparticles (Fig. S8a and b). To further examine in vivo toxicity, HPcLW nanoparticles were intra-articularly injected into mice at a lycopene-equivalent dose of 2.5 mg/kg. After two months of treatment, major organs (heart, liver, spleen, lungs, and kidneys) were collected for H&E staining, and no apparent morphological abnormalities or tissue damage were observed across groups (Fig. 2e). Additionally, blood routine analysis after two weeks of treatment revealed no significant differences in hematological parameters among all groups (Fig. S9). Collectively, these results confirmed the favorable biosafety of HPcLW nanoparticles for intra-articular administration.

Fig. 2.

Evaluation of Serum Stability, Biosafety, and Targeting Capability of HPcLW Nanoparticles. (a) Schematic illustration of the serum stability, safety, and targeting experiments of HPcLW nanoparticles. (b) Size distribution of HPcLW nanoparticles in 50% FBS at different time points (0 h, 3 h, 6 h, 12 h, and 24 h) (n = 4). (c, d) In vitro cytotoxicity of chondrocytes treated with HPcLW nanoparticles for 24 h and 48 h was evaluated by CCK8 assay (n = 6). (e) Representative H&E staining images of heart, liver, spleen, lungs, and kidneys of mice after 2 months of treatment with Blank, HPL, HPcL, HPcW, and HPcLW nanoparticles (lycopene 2.5 mg/kg, twice a week). Scale bars, 100 μm. (f, g) Uptake of HP@C6 (f) and HPc@C6 (g) nanoparticles by chondrocytes at 4 h and 12 h. Scale bars, 100 μm. (h) Competitive inhibition assay of HPc@C6 nanoparticle uptake in chondrocytes. Chondrocytes were pre-incubated with free CollBP (200 μg/mL) for 2 h, followed by incubation with fluorescent HPc@C6 for 12 h. Scale bars, 100 μm. (i) In vivo imaging of knee joints at different time points (Day 0, Day 1, Day 4, Day 8, and Day 10) after intra-articular injection of HPc@DiR and HP@DiR nanoparticles (DiR 2.5 mg/kg). (j) Fluorescence images of joint tissue sections at 48 h post-injection of HPc@C6 and HP@C6. Scale bars, 100 μm.

To evaluate the cartilage-targeting capability of the CollBP-functionalized nanoparticles, cellular uptake analysis using coumarin 6 (C6)-loaded HP@C6 or HPc@C6 nanoparticles (with or without the CollBP) demonstrated that HPc@C6 nanoparticles were internalized more efficiently by chondrocytes after 4 or 12 h of incubation, and co-staining with LysoTracker Red revealed efficient endolysosomal escape following uptake (Fig. 2f and g). Pre-incubation of chondrocytes with free CollBP (200 μg/mL) for 2 h inhibited the subsequent cellular uptake of HPc@C6 nanoparticles after 12 h incubation (Fig. 2h), which confirmed that the enhanced cellular internalization of HPcLW nanoparticles was dependent on CollBP-mediated targeting. In vivo fluorescence imaging revealed that HPc@DiR nanoparticles exhibited prolonged retention within the articular cavity compared to HP@DiR (Fig. 2i). On day 10, the fluorescence intensity in the HPc@DiR group was 1.67-fold higher than that in the HP@DiR group (Fig. S10a). Ex vivo imaging of dissected tissues on day 10 further confirmed that the residual fluorescence was localized predominantly within the knee joints of HPc@DiR-treated mice (Fig. S10b), indicating the superior joint-retention capability of the HPcLW formulation. Moreover, fluorescence imaging of joint tissue sections at 48 h post-injection of HP@C6 and HPc@C6 revealed no obvious fluorescence changes in the tissues surrounding the adjacent cartilage. In contrast, the HPc@C6 group exhibited markedly stronger fluorescence intensity in the cartilage layer, with penetration depth reaching approximately 60 μm (Fig. 2j), suggesting that HPcLW nanoparticles can effectively enrich in the deep cartilage region. Collectively, these results demonstrate that HPcLW nanoparticles combine excellent stability and biocompatibility with enhanced cartilage-targeting ability, positioning them as a promising platform for targeted osteoarthritis therapy.

2.3. HPcLW Nanoparticles Effectively Attenuate Cartilage Senescence and OA Progression

The therapeutic effects of HPcLW nanoparticles in vivo were evaluated in naturally aged mouse OA models, with assessments performed two months after treatment (Fig. 3a). As shown in Fig. 3b and c, Safranin O-Fast Green staining and OARSI score of the aging OA cartilage revealed that the cartilage damage in the HPcL-treated group was lower than that in the HPL group, and HPcLW treatment resulted in the lowest cartilage loss among all nanoparticle treatments. Immunofluorescence (IF) staining showed that HPcLW reduced the senescence marker p-p53 and the senescence-associated secretory factor IL6 (Fig. 3d, and Fig. S11). Importantly, HPcLW treatment promoted the expression of anabolic factors COL2 (Fig. 3e–g) and reduced extracellular matrix catabolism marker MMP13 (Fig. 3f–h) in aging OA cartilage.

Fig. 3.

HPcLW Nanoparticles Effectively Attenuate Cartilage Senescence and OA Progression. (a) Schematic illustration of the experimental design for HPcLW nanoparticle treatment in DMM-induced OA mice and aged OA mice. Male C57BL/6J mice (8 months old) subjected to DMM surgery or aged OA mice (18 months old) were treated with intra-articular injections of various formulations twice weekly for 8 weeks (n = 6 mice per group). The lycopene-equivalent concentration of nanoparticles in all treatment subgroups was maintained at 2.5 mg/kg (b, c) Representative Safranin O-fast green staining images of articular cartilage (b) and OARSI scores (c) in different treatment groups of the aged OA model (n = 6). Scale bars, 100 μm. (d) Representative immunofluorescence staining of p-p53 (Green) and IL6 (Red) in articular cartilage from different treatment groups of the aged OA model. Scale bars, 100 μm. (e, f) Representative immunohistochemical staining of COL2 (e, Scale bars, 100 μm) and MMP13 (f, Scale bars, 50 μm) in articular cartilage from different treatment groups of the aged OA model. (g, h) Quantitative analysis of COL2 (g) and MMP13 (h) immunohistochemical staining by Fiji indices across groups (n = 6). Asterisks above each bar in (c, g and h) indicate significant differences vs. Old group; horizontal lines denote pairwise comparisons among other groups. Statistical significance was analyzed by Student's t-test or one-way ANOVA. * p < 0.05, ** p < 0.01 and *** p < 0.001.

Given that elevated WTAP expression was also observed in the DMM-induced OA model in our preliminary findings [21], we further investigated the therapeutic potential of HPcLW in DMM-induced OA model. Consistently, HPcLW treatment exhibited marked chondroprotective effects in DMM-induced OA, as evidenced by Safranin O-Fast Green staining, as well as immunohistochemical analysis of COL2 and MMP13 (Fig. S12). These findings provide strong evidence that HPcLW treatment suppresses cartilage senescence and OA progression.

2.4. HPcLW Nanoparticle Treatment Suppresses the OA Phenotype in Senescent Chondrocytes

To further confirm the critical anti-OA role of HPcLW nanoparticles in vitro, a CCK-8 assay was performed. As shown in Fig. S13, HPL, HPcL, HPcW, and HPcLW nanoparticles all enhanced the survival capacity of chondrocytes compared to the H₂O₂ alone group, with HPcLW demonstrating the strongest effect. We verified that the expression levels of extracellular matrix anabolic genes (COL2A1 and ACAN) were up-regulated (Fig. 4a and Fig. S14a), whereas the expression levels of extracellular matrix catabolism genes (MMP13, MMP3, ADAMTS4, and ADAMTS5) (Fig. 4b and Fig. S14b–d) and genes associated with age-related inflammation (IL1β, TNF-α and IL8) (Fig. 4c, and Fig. S14e and f) were down-regulated by HPcLW nanoparticle treatment. As expected, the aforementioned nanoparticles reduced total intracellular ROS levels by DCFH-DA staining, with HPcL demonstrating superior efficacy compared to HPL, indicating that the addition of the CollBP enhanced the nanoparticles’ effect. Furthermore, ROS levels in chondrocytes treated with HPcLW were lower than those treated with HPcL and HPcW nanoparticles, suggesting that the loading of WTAP siRNA can also suppress intracellular ROS in chondrocytes (Fig. 4d and e). Moreover, immunofluorescence (IF) analysis confirmed that HPcLW treatment also reduced MMP13 expression (Fig. 4f and g) while promoting COL2 expression (Fig. S15a and b). These results support that HPcLW nanoparticles have superior cytoprotective effects against H₂O₂-induced chondrocyte senescence.

Fig. 4.

HPcLW Nanoparticle Treatment Suppresses the OA Phenotype in Senescent Chondrocytes. (a-c) mRNA expression levels of COL2A1 (a), MMP13 (b), and IL-1β (c) in chondrocytes after treatment with different nanoparticle formulations, determined by RT-PCR (n = 3). (d, e) Representative fluorescence images of chondrocytes stained with DCFH-DA probe (d) and fluorescence heatmap analysis using Fiji (e) after treatment with different nanoparticle formulations. Scale bars, 100 μm. (f, g) Representative immunofluorescence images of chondrocytes stained with DAPI (Blue) and MMP13 antibody (Red) (f) and corresponding fluorescence heatmap analysis using Fiji (g) following treatment with different nanoparticle formulations. Scale bars, 100 μm. Asterisks above each bar in (a-c) indicate significant differences vs. H₂O₂ group; horizontal lines denote pairwise comparisons among other groups. Statistical significance was analyzed by Student's t-test or one-way ANOVA. * p < 0.05, ** p < 0.01 and *** p < 0.001.

2.5. HPcLW Nanoparticle Treatment Mitigates Chondrocyte Senescence and Mitochondrial Damage

We next evaluated the anti-senescence efficacy of HPcLW nanoparticles. SA-β-Gal staining results showed that the HPcLW group had the fewest blue-positive cells compared to the HPL, HPcL, and HPcW groups (Fig. 5a). We also found a reduced number of p21-positive cells compared to the other nanoparticle-treated groups (Fig. 5b and c). Mitochondrial damage is a landmark event in the aging process [31]. We further assessed mitochondrial function. According to JC-1 staining results, a key indicator of mitochondrial membrane potential, H₂O₂-treated chondrocytes showed sharply reduced aggregate fluorescence (Red) and increased monomer fluorescence (Green), while HPcLW treatment reversed this process (Fig. S16). Consistent with this, MitoSOX Red staining revealed pronounced mitochondrial ROS accumulation in H₂O₂-induced senescent chondrocytes, which was markedly reduced by HPcLW compared with HPL and HPcL (Fig. 5d and e). Taken together, these results confirm that HPcLW nanoparticles have obvious anti-chondrocyte senescence and anti-mitochondrial damage effects.

Fig. 5.

HPcLW Nanoparticle Treatment Mitigates Chondrocyte Senescence and Mitochondrial Damage. (a) SA-β-gal staining of senescent chondrocytes after treatment with different nanoparticle formulations. Scale bars, 100 μm. (b, c) Representative immunofluorescence images of chondrocytes stained with DAPI (Blue) and p21 antibody (Red) after treatment with different nanoparticle formulations (b). Scale bars, 100 μm. Fluorescence heatmap analysis was performed using Fiji (c). (d, e) Representative fluorescence images of chondrocytes stained with Hoechst (Blue) and MitoSOX Red dye (Red) after treatment with different nanoparticle formulations. Scale bars, 50 μm. Fluorescence heatmap analysis was performed using Fiji (e).

The synergistic effects of HPcLW were evaluated in H₂O₂-treated chondrocytes by comparing HPcLW with blank nanoparticles plus an equivalent amount of free WTAP siRNA and free lycopene. HPcLW showed stronger anti-senescence and anti-oxidative activities than the free drug combination (Fig. S17a and b), confirming that nanoparticle-mediated co-delivery of lycopene and WTAP siRNA exerts a synergistic protective effect against chondrocyte damage.

2.6. Mechanism Analysis of Chondrocytes Following Lycopene Treatment and WTAP knockdown

To investigate the mechanism by which lycopene alleviates chondrocyte senescence, we performed gene sequencing on senescent chondrocytes treated with lycopene. Statistical analysis revealed that, following lycopene treatment, 1244 genes were downregulated and 2429 genes were upregulated in senescent chondrocytes (Fig. S18a and b). KEGG pathway analysis indicated that the differentially expressed genes were primarily enriched in the p53 signaling pathway and the cellular senescence signaling pathway (Fig. 6a). Moreover, GSEA analysis further confirmed the downregulation of these two pathways after lycopene treatment (Fig. S19a and b). Further, lycopene treatment upregulated the expression of chondrocyte extracellular matrix anabolic gene ACAN, while downregulating the catabolic genes ADAMTS4 and MMP11 (Fig. S20).

Fig. 6.

Mechanism Analysis of Chondrocytes Following Lycopene Treatment and WTAP Knockdown. (a) KEGG pathway enrichment analysis of differentially expressed genes in chondrocytes treated with Lycopene (10 μg/mL) under H₂O₂ stimulation for 24 h compared to the H₂O₂ group. (b) Heatmap analysis of the top 10 highly abundant genes among the downregulated differentially expressed genes in WTAP-knockdown chondrocytes compared to the NC group under H₂O₂ stimulation. (c, d) Heatmap showing the expression levels of WTAP, PAI-1, CDKN1A, CDK2, CDK6, and CCNB1 genes in Lyco + H₂O₂-treated versus H₂O₂-treated groups. (e, f) Heatmap showing the expression levels of WTAP, PAI-1, CDKN1A, CDK2, CDK6, and CCNB1 genes in WTAP-knockdown (WTAP-KD) versus negative control (NC) groups under H₂O₂ stimulation. (g) SA-β-gal staining of chondrocytes following WTAP knockdown or PAI-1 overexpression. Scale bars, 100 μm. (h) Representative DCFH-DA fluorescence images and corresponding heatmaps analysis using Fiji of ROS levels in chondrocytes after WTAP knockdown or PAI-1 overexpression. Scale bars, 100 μm. (i, j) mRNA expression levels of WTAP (i) and PAI-1 (j) in chondrocytes after WTAP knockdown or PAI-1 overexpression (n = 3). Statistical significance was analyzed by one-way ANOVA. * p < 0.05, ** p < 0.01 and *** p < 0.001.

To investigate the key genes regulated by WTAP in chondrocytes, we knocked down WTAP in chondrocytes and confirmed the knockdown efficiency using qRT-PCR (Fig. S21). Among the downregulated differentially expressed genes, we screened the top 10 with the highest abundance and found that PAI-1, a marker of aging [32], was downregulated following WTAP knockdown under H₂O₂ stimulation (Fig. 6b). Furthermore, interaction analysis showed that PAI-1 was associated with genes related to the chondrocyte OA phenotype (Fig. S22). Therefore, we hypothesize that inhibiting WTAP may suppress chondrocyte senescence and the OA process by inhibiting PAI-1.

Integrating transcriptomic analysis, we found that lycopene treatment downregulated WTAP, PAI-1 (also named SERPINE1), the p53 pathway downstream gene p21 (CDKN1A), and upregulated cell cycle-related genes (CDK2, CDK6, and CCNB1) (Fig. 6c and d). Strikingly, a similar expression of WTAP, PAI-1, p21, CDK2, CDK6, and CCNB1 was observed following WTAP knockdown in chondrocytes (Fig. 6e and f). Furthermore, we found that treatments with different concentrations of lycopene or WTAP knockdown could all reduce the number of β-gal-positive cells induced by H₂O₂, and also decrease the expression of the WTAP, PAI-1 and CDKN1A genes, while upregulating the expression of CDK2, CDK6 and CCNB1 (Figs. S23–26). However, 5 μg/mL lycopene was not effective in reducing the expression of WTAP. The 10 μg/mL treatment for 24 h only reduced the WTAP expression by 23.1% compared to the H₂O₂ group (Fig. S24a). After the WTAP knockdown, the intracellular level decreased by 63.3% compared to the H₂O₂ group (Fig. S26a). Therefore, we hypothesize that lycopene mainly exerts its effect through antioxidant action, while WTAP mainly alleviates the aging process of chondrocytes through m6A epigenetic modification. However, both can regulate the aging process of chondrocytes through the WTAP/PAI axis. To further verify the mechanism by which WTAP regulates PAI-1, we constructed chondrocytes with overexpressed PAI-1 or WTAP knockdown. As shown in Fig. 6g and h, overexpression of PAI-1 enhanced the number of β-gal-positive cells and the intracellular reactive oxygen species levels. WTAP knockdown rescued the chondrocyte aging, oxidative stress levels, and changes in p21 pathway-related genes (Fig. S27) caused by PAI-1 overexpression. RT-PCR results revealed that WTAP knockdown reduced the increase in PAI-1 induced by H₂O₂ (Fig. 6i), while overexpression of PAI-1 did not change the level of WTAP (Fig. 6j), indicating that WTAP is an upstream regulatory factor of PAI-1. These findings establish the WTAP/PAI-1 axis as a causal driver of chondrocyte senescence. In summary, we hypothesize that the two components in HPcLW nanoparticles can jointly regulate WTAP/PAI to affect cartilage aging and oxidative stress levels.

2.7. HPcLW Nanoparticles Ameliorate Chondrocyte Senescence and Osteoarthritic Pathology via jointly targeting the WTAP/PAI-1 Epigenetic Modification Axis

To further investigate how WTAP, as an m6A writer, regulates PAI-1 expression, we analyzed m6A sequencing data from clinical human OA cartilage samples and identified enriched m6A modification signals in the 3′UTR and exon regions of PAI-1 (Fig. 7a). Furthermore, m6A-RIP-qPCR confirmed that the levels of m6A-modified PAI-1 were reduced following WTAP knockdown (Fig. 7b). We also employed an m6A inhibitor (3-Deazaadenosine, DAA) to inhibit the m6A level of chondrocytes, as shown in Fig. 7c, DAA treatment reversed the upregulation of PAI-1 mRNA caused by WTAP overexpression, indicating that WTAP acts on PAI-1 in a m6A-dependent manner. Additionally, we conducted an mRNA degradation experiment on the chondrocytes knocked down by WTAP using actinomycin D (5 μg/mL). We found that H₂O₂ treatment prolonged the half-life of PAI-1 mRNA to 25.3 h. However, in WTAP-knockdown cells stimulated by H₂O₂, the half-life of PAI-1 decreased to 7.59 h (Fig. 7d). This indicates that the activation of WTAP by H₂O₂ can affect and promote the stability of this gene through the m6A modification of PAI-1, while knocking down WTAP reverses this process.

Fig. 7.

HPcLW Nanoparticles Ameliorate Chondrocyte Senescence and Osteoarthritic Pathology via Synergistically Targeting the WTAP/PAI-1 Epigenetic Modification Axis. (a) Expression and regional enrichment analysis of the top 50 upregulated differentially expressed genes with high-confidence m6A modifications in human osteoarthritic cartilage samples (n = 3) compared to non-injured cartilage samples (n = 2). (b) m6A-RIP-qPCR analysis of m6A-modified PAI-1 mRNA levels in chondrocytes following WTAP knockdown (n = 3). (c) mRNA expression levels of WTAP and PAI-1 in chondrocytes after WTAP overexpression (n = 3). (d) mRNA expression levels of PAI-1 in WTAP-knockdown chondrocytes after treatment with actinomycin D at different times (0 h, 2 h, 4 h and 6 h), with or without H₂O₂ treatment (n = 3). (e) Representative immunofluorescence images of chondrocytes stained with DAPI (Blue), WTAP antibody (Red), and PAI-1 antibody (Green) after treatment with different nanoparticle formulations. Scale bars, 100 μm. (f, g) Representative immunofluorescence staining and quantitative analysis was performed by Fiji (n = 6) of m6A (Green) and WTAP (Red) in articular cartilage from different treatment groups of the aged OA model. Scale bars, 100 μm. Asterisks above each bar in (b, c and g) indicate significant differences vs. Old group; horizontal lines denote pairwise comparisons among other groups. Statistical significance was analyzed by Student's t-test or one-way ANOVA. * p < 0.05, ** p < 0.01 and *** p < 0.001.

Furthermore, we verified this mechanism in the nanoparticle treatment. Immunofluorescence results for WTAP and PAI-1 in chondrocytes treated with different nanoparticles showed that H₂O₂ treatment enhanced the fluorescence of both WTAP, m6A and PAI-1 levels, while treatment with HPL, HPcL, HPcW, and HPcLW resulted in varying degrees of reduction in WTAP, m6A and PAI-1 fluorescence, with the most significant attenuation observed following HPcLW treatment (Fig. 7e, and Fig. S28, 29). Moreover, RT-PCR results indicated that H₂O₂ treatment promoted CDKN1A gene expression and downregulated cell cycle-related genes (CDK2, CDK6, and CCNB1). However, HPcLW nanoparticles reversed these changes to a greater extent than other nanoparticles (Fig. S30). Finally, we also verified this regulatory axis in mouse in vivo treatment. After treatment with different nanoparticles in vivo, WTAP, m6A and PAI-1 showed consistent changes as in vitro (Fig. 7f, g, and Fig. S31). These results confirm that in our HPcLW nanosystem, WTAP siRNA and lycopene act jointly to alleviate cartilage aging and the OA phenotype by blocking the WTAP/m6A-PAI-1 pathway and suppressing the p53-p21 axis.

Overall, the above data indicate that HPcLW nanoparticles can alleviate chondrocyte senescence and the progression of OA by jointly inhibiting the WTAP/PAI-1 axis.

3. Discussion

In this study, we successfully developed a cartilage-targeting nanoparticle platform, HPcLW, for the co-delivery of lycopene and WTAP siRNA to counteract chondrocyte senescence. This nanosystem was constructed via electrostatic self-assembly of HSA and cationic PLL, both of which are clinically used materials with established safety profiles, and synthesized PLL-PEG-CollBP to confer cartilage-targeting capability. HPcLW nanoparticles exhibited favorable physicochemical properties, including uniform size distribution (∼250 nm), high lycopene (>88%) and WTAP siRNA (>95%) encapsulation efficiency, excellent serum stability, and remarkable biosafety both in vitro and in vivo. Importantly, HPcLW nanoparticles effectively attenuated OA progression in both naturally aged and surgically induced OA mouse models. Mechanistically, our findings revealed that HPcLW nanoparticles synergistically suppressed the WTAP-m6A/PAI-1 epigenetic axis, thereby mitigating mitochondrial dysfunction and oxidative stress in senescent chondrocytes. Thus, our study constructed a novel nanodelivery system for targeting chondrocyte senescence and showed the important epigenetic m6A modification event critical for OA progression.

Cartilage repair remains challenging in osteoarthritis, and various biomaterial-based strategies have been explored. Bioactive glass-based approaches have demonstrated biocompatibility and chondrogenic potential through ion release, yet they often lack active targeting precision for cartilage-specific delivery [33]. A key obstacle in our formulation arose from the net positive charge of both the HSA-PLL core and the CollBP peptide, which prevented direct electrostatic adsorption, a common conjugation strategy. To overcome this, we synthesized a PLL-PEG2000-CollBP conjugate; the hydrophilic PEG spacer physically separates the positively charged moieties to eliminate repulsion, while also providing steric stabilization to enhance colloidal stability and ligand accessibility. The utility of PEGylation as a steric stabilizer is well documented: PEGylated mesoporous silica nanoparticles have achieved enhanced brain targeting via intranasal administration with improved stability and retention [34], and bioactive polyvinyl alcohol hydrogels have facilitated tissue regeneration through polymer matrix design [35]. Collectively, these reports support our rationale that the PEG spacer not only resolves electrostatic incompatibility but also serves as a critical design element for improving in vivo nanocarrier performance. In line with this design, our imaging and uptake data confirmed that the targeting moiety enhanced cartilage retention and chondrocyte uptake, with joint sections showing markedly stronger fluorescence in the cartilage layer (penetration depth ∼60 μm) and negligible signal in adjacent tissues. Thus, this conjugate offers a versatile platform for cartilage-targeted delivery.

Another major challenge is the delivery of lycopene, a highly hydrophobic hydrocarbon with poor aqueous solubility and limited systemic bioavailability [36]. Leveraging the natural transport function of human serum albumin (HSA), the most abundant protein in plasma, we utilized HSA as a nanocarrier for lycopene. HSA possesses multiple hydrophobic binding pockets, notably Sudlow's site I within subdomain IIA, which binds diverse lipophilic molecules [37]. Driven by hydrophobic interactions, lycopene partitions into these cavities, forming a stable complex via static quenching. This HSA-mediated encapsulation offers several advantages: it improves lycopene's aqueous solubility, protects it from oxidative degradation and enzymatic metabolism, and facilitates targeted delivery. In our system, HSA-PLL nanoparticles formed stable lycopene complexes, as confirmed by co-sedimentation, and the resulting HPcLW nanoparticles exhibited high serum stability and favorable biosafety profiles in vitro and in vivo. Furthermore, carotenoids complexed with HSA show enhanced quenching of singlet oxygen (¹O₂) compared to free carotenoids [38], an effect particularly relevant in osteoarthritis, where oxidative stress drives chondrocyte senescence and matrix degradation.

An additional intriguing finding is that HPcLW alleviates chondrocyte mitochondrial dysfunction and cellular senescence via inhibition of the WTAP/PAI-1 axis. HPcLW treatment restored mitochondrial membrane potential and reduced mitochondrial ROS, leading to attenuated senescence and an improved OA phenotype, evidenced by decreased ROS, SA-β-gal-positive cells, and senescence markers (p-p53, IL6, p21), alongside reversal of ECM catabolic/anabolic markers (reduced MMP13, increased COL2). Transcriptomic analysis revealed that WTAP knockdown suppressed PAI-1, which in turn inhibited the p53-p21 pathway, reducing p21 and upregulating cell-cycle-related genes (CDK2, CDK6, CCNB1). Similarly, lycopene alone downregulated WTAP and PAI-1, yielding comparable effects. The combination of both agents acting on the same axis achieved a synergistic regulatory outcome, representing a major advance in the functional design of HPcLW.

More importantly, we uncovered a previously unrecognized mechanism whereby WTAP drives pathological PAI-1 elevation via an m6A modification-dependent pathway. WTAP is highly expressed in both human and aged mouse OA cartilage. Knockdown of WTAP downregulated PAI-1, a senescence-associated marker known to activate the p53 pathway [39]. As a core component of the m6A methyltransferase complex, WTAP facilitates m6A modification on target transcripts, influencing their stability and translation [40]. m6A modifications preferentially occur on the DRACH motif (i.e., [AGU][AG]AC[ACU]) and are often enriched around stop codons and 3'untranslated regions (3′UTRs) [41], where they promote mRNA translation [42]. Through sequencing of human clinical samples, we identified candidate m6A modification sites in both the 3′UTR and an exon of PAI-1 in OA cartilage. m6A-RIP-qPCR and m6A inhibitor treatment further supported that WTAP regulates PAI-1 expression via m6A modification. Collectively, these findings point to a previously unrecognized WTAP/PAI-1 epigenetic axis that may contribute to chondrocyte senescence and OA progression.

Several limitations should be acknowledged. 1) This study was conducted in cellular and mouse models; translational validation in larger animals and human tissues is required. 2) The follow-up period was relatively short (≤8 weeks), leaving long-term safety and efficacy unknown. 3) While our analyses focused mainly on cartilage, OA is a whole-joint disease; future studies should evaluate synovitis, subchondral bone changes, and pain-related behaviors. 4) The H₂O₂-treated C28/I2 cell model does not fully recapitulate the complex OA microenvironment; validation in primary chondrocytes under inflammatory or mechanical loading is needed. 5) Although our data support a functional link between WTAP and PAI-1, direct evidence of specific m6A site regulation and the involvement of reader proteins warrants further investigation.

4. Conclusions

In conclusion, we have developed a cartilage-targeting HPcLW nanoparticle platform that effectively attenuates chondrocyte senescence and OA progression through synergistic inhibition of the WTAP/PAI-1 epigenetic axis and protection against mitochondrial dysfunction. By targeting the m6A epigenetic aging pathway, a mechanism distinct from currently available symptom-relief agents, HPcLW represents a novel disease-modifying strategy for OA. This study provides a foundation for developing targeted nanomedicine approaches for age-related joint diseases.

5. Methods

5.1. Chemicals and reagents

NHS-PEG2000-NHS (R-2002-2K) was provided by Ruixibio®. (China). CollBP (purity >95%) was custom-synthesized by ChinaPeptides (QYAOBIO). (China). Dimethyl sulfoxide (30072418), Acetone (40064460), Hydrogen peroxide (H₂O₂, 10011218), and Sodium hydroxide (10019719) were obtained from Sinopharm Chemical Reagent Co., Ltd. (China). Human serum albumin (A5843) and DL-DTT (D9779) were purchased from Sigma-Aldrich Co. (USA). Poly-L-lysine Hydrobromide (PLL, P8120), Lycopene (Lyco, SVL1005), Puromycin (IP12803), and Senescence-Associated β-Galactosidase (SA-β-Gal) Stain Kit (G1580) were purchased from Beijing Solarbio Science & Technology Co., Ltd. (China). Human and Mouse targeting WTAP siRNA was constructed by Shanghai Genechem Co., Ltd. Coumarin 6 (C6, B21363) was purchased by Shanghai Yuanye Bio-Technology Co., Ltd. (China). DiR (FD-CDR003) was provided by Guangzhou Biolight Biotechnology Co., Ltd. (China). Hoechst 33342 (C1027) and LysoTracker Red (Y060275) were purchased by Beyotime Biotechnology. (China). Cell Counting Kit-8 (CCK-8, HY-K0301), Actinomycin D (HY-17559), and JC-1 (HY-15534) were purchased from MedChemExpress. (USA). 2′,7′-Dichlorodihydrofluorescein diacetate (DCFH-DA, M9096) was purchased from AbMole. (USA). MitoSOX Red (M36009) was purchased by Thermo Fisher Scientific. (USA). C28/I2 chondrocyte cell line was purchased by Sigma-Aldrich (Shanghai) Trading Co., Ltd. (China).

5.2. Animals

Male C57BL/6 mice (8-week-old, 8-month-old, and 18-month-old) were purchased from the Model Animal Research Center of Nanjing University. All animal feeding and experimental procedures were approved by the Ethics Committee and the Institutional Animal Care and Use Committee of Drum Tower Hospital, Nanjing University Medical School (2026AE010236). For the naturally aged OA model, 18-month-old male C57BL/6 mice were randomly divided into five treatment groups: Blank, HPL, HPcL, HPcW, and HPcLW nanoparticles (n = 6 per group), with 8-month-old male mice serving as young controls (n = 6). Mice were housed under specific pathogen-free conditions with ad libitum access to food and water. Inclusion criteria required healthy mice free from visible joint abnormalities upon baseline gross examination. Exclusion criteria included signs of infection, severe weight loss (>20% of initial body weight), or poor general health during the treatment period. Animals were randomly assigned to groups using a computer-generated random number sequence. For the DMM-induced OA model, 8-week-old male mice underwent destabilization of the medial meniscus (DMM) surgery under general anesthesia as previously described. After surgery, the mice were randomly assigned to four treatment groups (Blank, HPL, HPcL, and HPcLW; n = 6 per group). All intra-articular injections were performed by an investigator blinded to treatment allocation.

The lycopene-equivalent concentration of nanoparticles in all treatment subgroups was maintained at 2.5 mg/kg, with the corresponding siRNA dose at∼142.45 μg/kg per injection. Intra-articular injections (25 μL per joint) were administered twice weekly for 8 weeks. Dosing volumes and schedules were consistent across all treatment groups. After 2 months of treatment, all mice were sacrificed. Knee joints, hearts, livers, spleens, lungs, and kidneys were collected for further analysis. For blood routine analysis, 8-week-old mice received intra-articular knee injections of Blank, HPL, HPcL, HPcW, or HPcLW nanoparticles (lycopene 2.5 mg/kg, twice a week) for two weeks. Following treatment, blood samples were collected and hematological analysis using a Mindray Veterinary Automatic Hematology Analyzer (BC-2800Vet) by Servicebio (Wuhan, China). Histological sections were coded prior to staining and scoring, and all OARSI scoring, immunohistochemical staining quantification, and immunofluorescence intensity analyses were performed by two independent observers blinded to the group assignments.

5.3. Synthesis of PLL30000-PEG2000-CollBP (PLL-PEG-CollBP) conjugate

Synthesis of PLL₃₀₀₀₀-PEG₂₀₀₀-CollBP was commissioned by Ruixibio®. NHS-PEG₂₀₀₀-NHS (100 mg) was dissolved in 2 mL of DMSO, followed by the addition of CollBP (1.0 equivalent) and triethylamine (3.0 equivalent). The mixture was stirred at room temperature for 4 h. Subsequently, PLL (Mw 30,000, 0.9 equivalent) was added, and the reaction was continued for another 4 h at room temperature. The reaction mixture was then transferred into a dialysis bag (MWCO 3500 Da) and dialyzed against purified water for 12 h. The dialysate was collected and lyophilized. The lyophilized product was precipitated in a large volume of ice-cold diethyl ether, collected by centrifugation, and dried under vacuum to obtain the final PLL-PEG-CollBP conjugate.

5.4. GPC and proton nuclear magnetic resonance (1H NMR) analysis

GPC analysis was performed on a Shimadzu LabSolutions GPC system with a refractive index detector. The sample was dissolved in the eluent at 1 mg/mL, filtered through a 0.22 μm membrane, and injected with a volume of 20 μL. The column was calibrated with appropriate standards, and the molecular weight parameters were calculated using Shimadzu LabSolutions software: number-average molecular weight (M_n) = 33855, weight-average molecular weight (Mw) = 51738, and polydispersity index (Mw/M_n) = 1.52824.

¹H NMR spectra were recorded on a Bruker Avance III 400 MHz NMR spectrometer at 25°C. Chemical shifts were reported in ppm, with D₂O (4.8 ppm) and residual DMSO (2.5 ppm) identified as solvent signals. The characteristic proton signals of PEG (3.5-3.7 ppm), PLL (3.0-3.2 ppm, 1.2-1.8 ppm), and CollBP (6.5-7.5 ppm, 4.0-4.5 ppm) were assigned to confirm the triblock copolymer structure, as labeled in the spectrum.

5.5. Preparation and characterization of HPcLW nanoparticles

Cartilage-targeting HPcLW nanoparticles were fabricated as follows: an aqueous solution containing HSA and PLL at a 1:1 mass ratio, supplemented with PLL-PEG-CollBP (the final concentration is 0.3 mg ml⁻¹) for cartilage targeting, was prepared. The pH was adjusted to 9 using 1 M NaOH. Following 5 min of equilibration, DTT (0.1 M, 20 μL) was added per 1 mL of the mixture, and the reaction was allowed to proceed for 1 h at room temperature. The observed increase in turbidity indicated the successful assembly of cartilage-targeting nanospheres. To load lycopene and WTAP siRNA, the nanospheres were centrifuged at 5000 rpm for 10 min, resuspended in ddH₂O, and then incubated overnight at 4°C with lycopene (final concentration 4 mM) and WTAP siRNA (final concentration 7.2 μM). For control formulations, Blank nanospheres were prepared by omitting PLL-PEG-CollBP, lycopene, and WTAP siRNA; HPL nanospheres were prepared by adding lycopene only; HPcL nanospheres were prepared by incorporating PLL-PEG-CollBP and lycopene; and HPcW nanospheres were prepared by incorporating PLL-PEG-CollBP and WTAP siRNA.

For in vivo imaging, HP@DiR nanoparticles were prepared by loading DiR onto blank nanoparticles, while HPc@DiR contained PLL-PEG-CollBP and DiR. For cell uptake experiments, HPc@C6 was prepared by loading C6, and HPc@C6 contained PLL-PEG-CollBP and C6.

The morphology of the prepared nanoparticles was observed by TEM (Hitachi HT7800, Japan). The size and zeta potential of the nanoparticles were characterized by a Malvern dynamic laser scattering instrument (Zetasizer NanoZSP, UK).

The encapsulation efficiency and drug loading capacity of lycopene in HPL, HPcL, and HPcLW nanoparticles were determined using a microplate reader (Molecular Devices ID5, China). A standard calibration curve was first established using lycopene standards at 475 nm. The encapsulation efficiency and drug loading capacity were calculated using the following formulas: Encapsulation efficiency (%) = (Amount of lycopene in nanoparticles/Total amount of lycopene added) × 100%; Drug loading capacity (%) = (Amount of lycopene in nanoparticles/Weight of nanoparticles recovered) × 100%. The in vitro release profiles of lycopene from nanoparticles was investigated under two different pH conditions (pH 7.4 and pH 5.0) at 37°C and were measured by a microplate reader at 475 nm, at different times (0 h, 3 h, 6 h, 12 h, 24 h, 36 h, 48 h, 60 h, 72 h), 100 μL samples were taken, centrifuged, and the pellets were dissolved in acetone for drug quantification, with volume maintained by replacing each sample with 100 μL fresh medium.

The encapsulation efficiency and drug loading capacity of WTAP siRNA in HPcW and HPcLW nanoparticles were determined using a Nanodrop (ND-ONE-W), Thermofisher, USA). The encapsulation efficiency and drug loading capacity were calculated using the following formulas: Encapsulation efficiency (%) = (Amount of WTAP siRNA in nanoparticles/Total amount of WTAP siRNA added) × 100%; Drug loading capacity = (Amount of WTAP siRNA in nanoparticles/Weight of HSA). Cumulative WTAP siRNA release was also quantified by Nanodrop. At different times (0 h, 3 h, 6 h, 12 h, 24 h, 36 h, 48 h, 60 h, 72 h), 1 mL HPcLW was centrifuged, and the supernatant was analyzed for released siRNA, with cumulative release determined against the initial siRNA content.

The protective effect of nanoparticles against siRNA degradation was assessed by treating 1 mL of HPcLW-loaded siRNA and equivalent free siRNA with RNase A (100 ng/mL) for 1 h. Residual siRNA was measured by Nanodrop, and the protection efficiency was calculated relative to the free siRNA control.

5.6. Cellular uptake

Chondrocytes seeded in confocal dishes (2 × 10⁴ cells/dish) were cultured for 24 h and then treated with HP@C6 or HPc@C6 nanoparticles loaded with C6 for 4 h or 12 h at 37°C. After washing with warm PBS, cells were stained with Hoechst 33342 and LysoTracker Red for 30 min at 37°C. Live-cell imaging was performed immediately using a fluorescence microscope (Leica, Wetzlar, Germany).

5.7. In vivo imaging experiments

8-week-old C57BL/6 mice were Intra-articular injected with 25 μL of HP@DiR and HPc@DiR. At different time points (Day 0, Day 1, Day 4, Day 8, and Day 10) after administration, mice were anesthetized, imaged and analyzed by an in vivo imaging system (AniView X, China) under (Excitation): 745 nm/Em (Emission):840 nm. Then, mice were sacrificed by isoflurane overdose. The knee joint, heart, liver, spleen, lung, and kidney were collected and imaged by an in vivo imaging system (AniView X, China).

5.8. Cell viability assay

Chondrocytes were seeded in 96-well plates and cultured overnight. For toxicity detection of nanoparticles, chondrocyte cells were treated with different concentrations of HPcLW (Lyco, 1.25 μg/mL,2.5 μg/mL, 5 μg/mL, and 10 μg/mL) and HPcLW (stored at 25°C in the dark for 7 days) for 24 h and 48 h. For the effects of H₂O₂ and nanoparticles, six treatment groups were used: Ctrl (untreated), H₂O₂ (400 μM), Blank (Blank/H₂O₂), HPL (HPL/H₂O₂), HPcL (HPcL/H₂O₂) and HPcLW (HPcLW/H₂O₂), with all groups treated with a consistent lycopene nanoparticle concentration of 10 μg/ml. Following the above treatments, 10 μL of CCK-8 reagent was added to each well, and the plates were incubated at 37°C for 2 h. The absorbance at 450 nm was measured by a microplate reader (Molecular Devices M3, USA). The viability of treated cells was normalized by Ctrl group.

5.9. SA-β-gal staining

Chondrocytes were seeded in 6-well plates at a density of 50-70%. After treatment, the cells were washed with PBS, followed by the addition of 2 mL of β-Gal fixative solution and fixation at room temperature for 15 min. The fixative was then removed, and the cells were washed three times with PBS. The staining working solution was prepared according to the manufacturer's instructions. Then, 2 mL of the staining working solution was added to each well. The plates were incubated at 37°C overnight and subsequently observed under a light microscope (Nikon, Japan).

5.10. Total ROS detection assay

To detect intracellular total ROS levels, chondrocytes were incubated with 5 μM DCFH-DA working solution at 37°C in the dark for 30 min. After incubation, the cells were washed three times with PBS and observed under a fluorescence microscope (Leica, Wetzlar, Germany).

5.11. Assessment of mitochondrial function

Mitochondrial membrane potential was assessed using JC-1, and mitochondrial ROS detection was performed using mitoSOX Red. Briefly, chondrocyte cells were treated with H₂O₂ and nanoparticles. Cells were stained with two dyes using working-concentration staining solutions specified in the manual, and images were captured by fluorescence microscopy (Leica, Wetzlar, Germany).

5.12. Histology, immunohistochemistry, and immunofluorescence analyses

For cellular immunofluorescence, chondrocytes were fixed with 4% paraformaldehyde, permeabilized with Triton X-100, and blocked with goat serum. The cells were then incubated overnight at 4°C with primary antibodies against P21 (Cat No. 10355-1-AP; Proteintech), MMP13 (Cat No. 18165-1-AP; Proteintech), COL2 (Cat No. BA0533; Boster), WTAP (Cat No. 60188-1-Ig; Proteintech), PAI-1 (Cat No. A6211; Abclonal) and m6A (Cat No. # 56593; Cell Signaling Technology). Following primary antibody incubation, the cells were incubated with Alexa Fluor-conjugated secondary antibodies (Thermo Fisher Scientific, Shanghai, China) and counterstained with DAPI (Cat No. P0131; Beyotime). Fluorescence images were captured using a microscope (Leica, Wetzlar, Germany) and analyzed with Fiji software.

For histological evaluation, knee joints and major organs were collected, fixed, decalcified, paraffin-embedded, and sectioned at a thickness of 5 μm. Hematoxylin and eosin (H&E) staining and Safranin O-Fast Green staining were performed according to standard protocols. The Osteoarthritis Research Society International (OARSI) grading system was used to evaluate the cartilage lesions. For immunohistochemistry (IHC) and immunofluorescence (IF), antigen retrieval was carried out using an antigen retrieval solution (Beyotime, Shanghai, China). Sections were blocked with 5% normal goat serum and incubated overnight at 4°C with primary antibodies against p-p53 (Cat No. 28961-1-AP; Proteintech), IL6 (Cat No. 66146-1-Ig; Proteintech), MMP13 (Cat No. 18165-1-AP; Proteintech), COL2 (Cat No. BA0533; Boster), PAI-1 (Cat No. A6211; Abclonal), WTAP (Cat No. 60188-1-Ig; Proteintech) and m6A (Cat No. # 56593; Cell Signaling Technology). For IHC, detection was performed using a horseradish peroxidase (HRP)-conjugated secondary antibody and 3,3′-diaminobenzidine (DAB) as the chromogen. For IF, sections were incubated with Alexa Fluor-conjugated secondary antibodies (Thermo Fisher Scientific, Shanghai, China) and counterstained with DAPI (Cat No. P0131; Beyotime). Images were acquired using a light microscope (VS200, Olympus, Japan) or a fluorescence microscope (Leica, Wetzlar, Germany). Immunofluorescence images were analyzed with Fiji software.

5.13. Cell transfection

Chondrocytes were seeded in 6-well plates at a density of 50-70%. Chondrocytes were transfected with WTAP interference virus (HBLV-h-WTAP shRNA1-ZsGreen-PURO) to construct WTAP-KD cells; NC was used as the control cells. For WTAP and PAI-1 overexpression, 70% confluent chondrocytes were transfected with the plasmid pCMV-WTAP-neomycin and pCMV-PAI-1-puromycin (GK Gene) and the FuGENE®HD transfection reagent (Promega, Madison, Wisconsin, USA) and incubated for 24 h at 37°C and 5% CO₂. Then, chondrocytes were screened using neomycin or puromycin (Meilunbio, Dalian, China) for one day. All transfection efficiencies were verified by qRT-PCR analysis.

5.14. Transcriptome and MeRIP sequencing and analyses

Transcriptome sequencing was performed on the Illumina platform. Library construction was carried out using the Illumina TruSeq RNA Sample Preparation Kit. Differential expression analysis was conducted using the DESeq2 R package. Gene Ontology (GO) enrichment, Gene Set Enrichment Analysis (GSEA), and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analyses were performed using the clusterProfiler package. Heatmaps of differentially expressed genes were generated using the Hiplot tool. Lyco + H₂O₂ vs. H₂O₂ samples transcriptome sequencing and bioinformatics analysis were carried out by Hangzhou Cosmos Wisdom Biotech Co., Ltd., and WTAP-KD + H₂O₂ vs. NC + H₂O₂ samples transcriptome sequencing and bioinformatics analysis were carried out by Annaroad Gene Technology (Beijing)Co, Ltd. PAI-1 and COL2A1, ADAMTS5, ADAMTS4, MMP3, and MMP12 interaction analysis by the String Database (https://cn.string-db.org/).

To investigate the m6A modification of PAI-1 in clinical human OA samples, we performed a secondary analysis of the MeRIP RNA sequencing data generated in our previous study [21], differential m6A peaks were screened with FDR <0.05 to control false positives, which were simultaneously performed (Genesky Biology).

5.15. Quantitative reverse transcription polymerase chain reaction (RT-PCR) and m6A-RIP-PCR

For RT-PCR, total RNA was isolated from chondrocytes using TRIzol reagent (Invitrogen, Carlsbad, CA, USA) following the manufacturer's instructions. Complementary DNA (cDNA) was synthesized from total RNA using the HiScript II QRT SuperMix for qPCR (+gDNA wiper) (Vazyme, Nanjing, China). Quantitative real-time PCR (qRT-PCR) was performed with the ChamQ SYBR Color qPCR Master Mix (Vazyme, Nanjing, China) using gene-specific primers. The primer sequences are shown in Supplementary Table 1.

For m6A-RIP-qPCR, total RNA was extracted from chondrocytes and treated with DNase (Sigma, St. Louis, MO, USA) to remove genomic DNA. After mRNA purification and fragmentation, the fragments were immunoprecipitated with an anti-m6A antibody using a Magna MeRIP™ m6A kit (IVDSHOW, Zhengzhou, China). The enriched m6A-modified mRNA was then quantified by qRT-PCR.

5.16. mRNA stability assay

WTAP knockdown (WTAP-KD) chondrocytes were used for the mRNA stability assay. After H₂O₂ treatment, actinomycin D (MedchemExpress) was added to all groups at a final concentration of 5 μg/mL (defined as time 0). Cells were harvested at 0 h, 2 h, 4 h, and 6 h after actinomycin D addition for RNA extraction. Half-life of PAI-1 mRNA analyzed by GraphPad Prism.

5.17. Statistical analysis

All data are presented as the mean ± standard deviation (SD). All experiments were performed with at least three biological replicates. For in vivo experiments, n indicates the number of mice per group; for in vitro experiments, n indicates the number of independent biological replicates. Statistical comparisons between two groups were performed using an unpaired, two-tailed Student's t-test. For comparisons involving multiple groups, one-way analysis of variance (ANOVA) followed by Dunnett's post-hoc test was applied. All statistical analyses were conducted using GraphPad Prism (version 10; GraphPad Software, San Diego, CA, USA). Significance was defined as *P < 0.05, **P < 0.01, and ***P < 0.001.

Data availability statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.

Ethics approval and consent to participate

The experimental protocol was established according to the ethical guidelines of the Helsinki Declaration and was approved by the Ethics Committee of the Drum Tower Hospital, Medical School of Nanjing University. The ethics approval number was 2026AE010236. The animals were treated in accordance with the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health, USA.

CRediT authorship contribution statement

Xueying An: Conceptualization, Formal analysis, Methodology, Resources, Validation, Visualization, Writing – original draft. Hantao Cai: Conceptualization, Formal analysis, Methodology, Resources, Validation, Writing – original draft. Wenshu Wu: Conceptualization, Formal analysis, Methodology, Resources, Validation, Writing – original draft. Minyi Cai: Data curation, Methodology, Visualization. Yu Ben: Data curation, Methodology, Visualization. Tao Shen: Methodology, Visualization. Pan Zhang: Methodology, Visualization. Jianmei Chen: Formal analysis, Methodology, Project administration, Resources, Validation. Zhihong Xu: Formal analysis, Methodology, Project administration, Resources, Validation. Qing Jiang: Conceptualization, Funding acquisition, Project administration, Supervision, Writing – review & editing.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Supporting Information

Supplementary figures offer additional experimental and analytical data that support the key conclusions in the main text, including bioinformatics analysis results, GPC and NMR spectra of synthesized materials, the characterization of materials, in vitro and in vivo functional evaluation data, in vitro immunohistochemical staining and analyses, and RT-PCR results. Supplementary tables detail the primer sequences and siRNA sequences used in this study.

Appendix A. Supplementary data

The following is the Supplementary data to this article.