An injectable dihydroartemisinin nanocomposite hydrogel for dual-targeting PANoptosis and inflammation to treat osteoarthritis

Abstract

PANoptosis, a distinct type of inflammation-associated programmed cell death that integrates features of apoptosis, necroptosis and pyroptosis, is closely related to the pathogenesis and progression of articular cartilage degeneration in osteoarthritis (OA). Meanwhile, excessive reactive oxygen species (ROS), pro-inflammatory pathways, and inflammatory cytokines in the OA microenvironment mutually reinforce one another, exacerbating synovial inflammation. To simultaneously target these interconnected pathological drivers, we developed an injectable nanocomposite thermosensitive hydrogel (DHA-PRO@NMs@TSH). This hierarchical system was constructed by integrating a ROS-responsive dihydroartemisinin prodrug (DHA-PRO) into Soluplus/TPGS-based nanomicelles (NMs), which were subsequently embedded within a Poloxamer-based thermosensitive hydrogel (TSH). This hierarchical and smart drug delivery system enables stable dihydroartemisinin (DHA) delivery and demonstrates a graded, ROS-triggered release profile, achieving sustained DHA retention in OA joints. The formulated system demonstrated excellent injectability, thermosensitivity, and physicomechanical stability, achieving sustained drug retention in rat joints for over 7 days, as confirmed by in vivo imaging. In vitro studies demonstrated that DHA-PRO@NMs significantly suppressed PANoptosis in chondrocytes by downregulating key markers (e.g., Bax/Bcl-2, RIPK3/p-RIPK3, NLRP3/Caspase-1/GSDMD) and inhibited inflammation by blocking the NF-κB pathway and subsequent cytokine production (e.g., TNF-α, IL-1β). In a rat OA model, treatment with DHA-PRO@NMs@TSH robustly attenuated disease progression, as evidenced by near-complete restoration of gait function, marked inhibition of cartilage degradation, a remarkable nearly 90% reduction in the Mankin score, and a comprehensive suppression of PANoptosis and inflammation pathological markers. This work not only presents a promising translational strategy for OA but also pioneers a therapeutic paradigm of leveraging advanced drug delivery system to dual-targeting PANoptosis and inflammation.

Graphical abstract

Supplementary Information

The online version contains supplementary material available at 10.1186/s12951-025-03990-y.

Introduction

Population aging has emerged as a pressing global issue, accompanied by a growing burden of age-related diseases [1]. Osteoarthritis (OA), the most prevalent degenerative joint disorder, is now a leading cause of disability among the elderly worldwide [2, 3]. In clinical practice, a stepwise treatment regimen is generally adopted according to disease severity [4]. Among these options, intra-articular injection is widely applied in patients with mild to moderate OA owing to its simplicity, direct local delivery, and minimally invasive nature [5, 6]. Current injectable agents include hyaluronic acid, corticosteroids, platelet-rich plasma, and mesenchymal stem cells, among others [7, 8]. However, these approaches remain insufficient due to variable efficacy, procedural complexity, high costs, and risks of cartilage damage upon repeated administration [9–13]. This stark clinical reality underscores the pressing need for safe, effective, and convenient intra-articular therapies capable of fundamentally modifying the destructive course of OA and alleviating the suffering of OA patients.

The pathogenesis of OA remains incompletely defined, but cartilage degeneration is considered the central pathological feature, while synovial inflammation is the primary source of pain in affected patients [14, 15]. On one hand, programmed death of chondrocytes is a hallmark of cartilage degeneration. In 2019, Malireddi et al. introduced the concept of PANoptosis, an inflammatory programmed cell death pathway mediated by the PANoptosome complex, which integrates features of apoptosis, necroptosis, and pyroptosis, and cannot be classified by a single mode of cell death [16, 17]. Since chondrocyte death in OA involves multiple death pathways, PANoptosis provides a strong theoretical basis for therapeutic intervention [18]. On the other hand, excessive reactive oxygen species (ROS) in the OA microenvironment not only induce oxidative damage to cells and tissues but also act as second messengers to activate pro-inflammatory pathways such as Nuclear Factor kappa-light-chain-enhancer of activated B cells (NF-κB), which drive the release of inflammatory cytokines including interleukin-1β (IL-1β) and tumor necrosis factor-α (TNF-α) [19]. These inflammatory mediators, in turn, stimulate mitochondrial ROS production, forming a self-amplifying positive feedback loop that aggravates synovial inflammation [20]. Therefore, therapeutic strategies that simultaneously target PANoptosis and inflammation hold great promise for OA treatment.

Artemisinin is a renowned antimalarial drug that has saved millions of lives [21]. Dihydroartemisinin (DHA), potent derivative of artemisinin, exhibits broader therapeutic potential, particularly in cancer and inflammatory diseases [22, 23]. Accumulating evidence has demonstrated that DHA possesses strong anti-inflammatory and antioxidant properties across diverse inflammatory conditions, including OA [24]. Furthermore, recent studies suggest that artemisinin and DHA can delay the progression of degenerative disorders such as Alzheimer’s and Parkinson’s diseases by protecting against cellular degeneration and death, highlighting its potential as a multi-target agent against PANoptosis [25–30]. Given that OA is a degenerative disease driven by chronic inflammation, it is reasonable to hypothesize that DHA may provide therapeutic benefits in this context. However, the clinical translation of DHA is severely hampered by its poor aqueous solubility, low stability, and rapid clearance from the joint space, and its intra-articular administration also remains largely neglected [31–33]. Advances in drug delivery systems (DDS) offer a compelling solution to these pharmaceutical limitations. However, current systems also face their own challenges. Simple drug-loaded hydrogels often lack specificity for intracellular release, while some nano-systems suffer from rapid clearance unless incorporated into a sustained-release depot [34–36]. Furthermore, few existing DDS are designed to explicitly respond to the OA-specific oxidative microenvironment.

To overcome these layered challenges, we engineered a hierarchical and smart DDS that integrates rational design at the molecular, nano-scale, and macro-scale. Our system, DHA-PRO@NMs@TSH, is constructed through: (1) Synthesizing a ROS-responsive DHA prodrug (DHA-PRO) to ensure stable delivery and intracellular ROS-responsive release. (2) Encapsulating DHA-PRO into Soluplus/TPGS-based nanomicelles (NMs) to enhance solubility, stability, and sustained release capability. (3) Incorporating DHA-PRO@NMs into a Poloxamer 407/188-based thermosensitive hydrogel (TSH) to form an injectable depot that guarantees prolonged joint residence and provides sustained release. The advantage of each step in the design lies in: (1) At the molecular level, the conjugation of one cinnamaldehyde (CA) molecule to two dihydroartemisinin (DHA) molecules via a thioacetal linker achieves a high drug payload. This small-molecule prodrug strategy features a straightforward and reliable synthetic route with excellent reproducibility [37, 38]. (2) The nano-encapsulation employs a dual-micelle system of Soluplus and TPGS, both widely investigated amphiphilic polymers. Soluplus provides exceptional solubilization capacity by encapsulating the highly hydrophobic prodrug within its hydrophobic core, yielding a stable micellar solution [39]. TPGS, a water-soluble derivative of natural vitamin E, enhances cellular uptake and offers excellent biocompatibility and safety [40]. This combined strategy ensures not only superior solubilization and stability but also significantly increased intracellular drug concentration. (3) At the macroscopic level, The Poloxamer 407/188 hydrogel demonstrates exceptional thermosensitivity, enabling facile injection and spontaneous formation of a stable drug depot at body temperature [41–43]. Furthermore, under alternating low-strain and high-strain stress conditions, its remarkable self-recovery capability ensures structural integrity and sustained drug release under dynamic mechanical stress in the joint cavity [44–46].

The application and therapeutic journey of the hierarchical DDS, from administration to pharmacological action, is designed as a multi-stage process as shown in the Graphical Abstract: Step 1 (Injectable depot formation). Upon intra-articular injection, the thermosensitive DHA-PRO@NMs@TSH system undergoes a rapid sol-to-gel transition at body temperature, forming a stable depot within the joint cavity. This macroscopic level of control primarily addresses the challenge of rapid clearance, ensuring long-term residence of the formulation. Step 2 (Stable delivery of nanomicelles). Nanomicelles loaded with DHA-PRO (DHA-PRO@NMs) are gradually released over an extended period from the eroding hydrogel matrix. At the microscopic level, these nanomicelles are safely delivered from the hydrogel to inflamed tissues and cells due to their excellent water solubility, nanoscale dimensions, and the properties of the TPGS/Soluplus matrix. Additionally, the nanomicelles also exhibit a certain degree of sustained release capability. Step 3 (Graded delivery targeting inflammation site). The nano-size of the nanomicelles and the ROS-triggered release mechanism creates a self-amplifying loop: Areas with higher ROS levels (more severe inflammation) trigger faster drug release, leading to accelerated consumption of DHA-PRO@NMs and accelerated cellular uptake. Based on these, our system holds promise for achieving pathology-dependent graded delivery: patients with severe OA demonstrate faster uptake and utilization of DHA compared to those with mild OA, and inflamed regions within the same knee joint exhibit greater DHA uptake and utilization than unaffected areas. Step 4 (Intracellular activation of the prodrug). Once inside the inflammatory chondrocytes, the DHA-PRO, which is stable under normal physiological conditions, encounters the elevated intracellular ROS levels characteristic of the OA microenvironment. This high ROS concentration triggers the cleavage of the thioacetal linker, leading to the controlled release of the active parent drug, DHA. Step 5 (Dual-targeting pharmacological action). The released DHA then executes its dual-therapeutic function. It suppresses the PANoptosis pathway by inhibiting key markers of apoptosis, necroptosis, and pyroptosis, thereby preventing chondrocyte death. Meanwhile it scavenges residual ROS, inflammatory cytokines, and inhibits the NF-κB signaling pathway, effectively breaking the cycle of oxidative stress and inflammation. In summary, this hierarchical design orchestrates a coordinated sequence of events: joint retention and sustained release (TSH), stable and graded delivery (NMs), ROS-responsive release (DHA-PRO), dual-targeting therapy (DHA). This sophisticated strategy ensures efficient, targeted, and sustained drug delivery directly to the pathogenic core of OA, which single-modal delivery systems fail to achieve.

In short, we present the development and comprehensive evaluation of this novel system (DHA-PRO@NMs@TSH). A distinctive feature of this work is the ternary composite: comprising a prodrug, nanomicelles, and a thermosensitive hydrogel, seamlessly merged into a single platform for intra-articular delivery. More importantly, it is the first to demonstrate a pharmacological strategy that simultaneously targets the PANoptosis pathway and the classic inflammatory axis for OA treatment. In this study, we systematically demonstrated and validated the synthesis, characterization, and efficacy of this system through a series of in vitro and in vivo experiments. Notably, we validated the marked mitigation of cartilage degeneration and joint damage in OA rats through multi-modal assessments encompassing behavioral tests, micro-CT, and histopathology, while demonstrating excellent biosafety. The specific workflow is shown in Fig. 1. This study is likely to provide a reliable reference for the rational design of intra-articular injection formulations and opens up new avenues for the treatment of OA.

Fig. 1.

Design and therapeutic mechanism of the Dihydroartemisinin (DHA) delivery system. (A) Illustrates the fabrication process of the DHA delivery system, from prodrug (DHA-PRO) to nanocomposite (DHA-PRO@NMs) and finally the thermosensitive hydrogel system (DHA-PRO@NMs@TSH). (B) Illustrates the DHA delivery and release process, its dual-targeting molecular mechanism, alongside therapeutic effect in vitro and in vivo

Experimental section

Reagents and materials

Dihydroartemisinin (DHA), cinnamaldehyde (CA) and 3-mercaptopropionic acid were purchased from Shanghai McLean Biochemistry Co., Ltd (Shanghai, China). Soluplus was purchased from Beijing Fengli Jingqiu Pharmaceutical Co., Ltd (Beijing, China). Vitamin E polyethylene glycol succinate (TPGS) was obtained from Sigma-Aldrich Co., Ltd (St. Louis, USA). Poloxamer 407 (P407) and Poloxamer 188 (P188) were purchased from Shanghai Aladdin Biochemical Technology Co., Ltd (Shanghai, China). Cell Counting Kit-8 (CCK-8) were purchased from Solarbio Co., Ltd (Beijing, China). Reactive oxygen species (ROS) Kit and YO-PRO-1/Propidium Iodide (YP1/PI) Kit were purchased from Shanghai Beyotime Biotechnology Co., Ltd (Shanghai, China). Coumarin 6 (C6) and Infrared-808 (IR808) iodide were obtained from Heowns Biochemical Technology Co., Ltd. (Tianjin, China). Primary antibodies against β-actin, Bcl-2, Bax, RIPK-3, p-RIPK-3, NLRP3, caspase-1, GSDMD, NOX4, SOD2, p-IKKβ, IKKβ, p-IκBα, IκBα, IL-1β, collagen II, aggrecan, MMP13, as well as goat anti-rabbit IgG and goat anti-mouse IgG secondary antibodies were obtained from Proteintech Group, Inc (Wuhan, China). Primary antibodies against p65, p-p65, TNF-α were obtained from ABclonal Biotech Co., Ltd. (Wuhan, China). Monosodium iodoacetate (MIA) were provided by Macklin Co., Ltd. (Shanghai, China). Enzyme-linked immunosorbent assay (ELISA) kits for IL-1β and TNF-α were purchased from Elabscience Biotechnology Co., Ltd. (Wuhan, China). Assay kits for alanine aminotransferase (ALT), aspartate aminotransferase (AST), blood urea nitrogen (BUN), creatinine (CRE), and lactate dehydrogenase (LDH) were acquired from Nanjing Jiancheng Bioengineering Institute (Nanjing, China).

Synthesis and ROS-Responsive release of DHA-PRO

Based on a previously established synthesis method from our laboratory, a dihydroartemisinin prodrug (DHA-PRO) was directly synthesized through optimization of the reactant stoichiometry [47]. The detailed procedure is as follows: CA (6.49 g, 49.1 mmol) and 3-mercaptopropionic acid (11.44 g, 108.02 mmol) were dissolved in ethyl acetate (20 mL) with trifluoroacetic acid as catalyst. The reaction mixture was stirred at 0 °C for 24 h in the dark. The resulting precipitate was collected via filtration, washed with cyclohexane, and used directly in the next step without further purification. The intermediate product (326 mg, 1.0 mmol), DHA (568 mg, 2.0 mmol), 1-ethyl-3-(3-dimethylaminopropyl) carbodiimide hydrochloride (EDCI) (383 mg, 2.0 mmol), and 4-dimethylaminopyridine (DMAP) (catalytic amount) were dissolved in anhydrous dichloromethane (DCM) (20 mL) under an argon atmosphere. The reaction was stirred at room temperature overnight. The mixture was then washed with dilute hydrochloric acid (20 mL), and the organic layer was dried over anhydrous Na₂SO₄. After removal of the solvent under reduced pressure, the crude product was purified by preparative thin-layer chromatography (PTLC) using a mobile phase of dichloromethane/methanol (30:1, v/v) to obtain DHA-PRO. The structure of DHA-PRO was then determined through HR MS and ¹H NMR. The in vitro release profile of DHA from DHA-PRO was evaluated using a dialysis method under different hydrogen peroxide (H₂O₂) concentrations. DHA-PRO (DHA-equivalent concentration = 100 µg/mL) was prepared in release medium (10% ethanol and 0.5% SDS, pH 7.4), and 1 mL of this solution was placed in a dialysis bag. The bag was immersed in 5 mL of release medium containing H₂O₂ at concentrations of 0 mM, 3 mM, 5 mM, or 10 mM and incubated at 37 °C with shaking. Samples (500 µL) were collected at 3, 6, 9, 12, and 24 h, with an equal volume of fresh medium replenished after each sampling. Each sample was extracted with ethyl acetate (1 mL), and 800 µL of the supernatant was collected, evaporated to dryness, and reconstituted in 500 µL of mobile phase for analysis. DHA concentration was quantified using high-performance liquid chromatography-tandem mass spectrometry (HPLC-MS/MS) on an API 3200 system (Applied Biosystems, Foster City, CA). Chromatographic separation was achieved using a mobile phase consisting of 0.1% formic acid in acetonitrile and 10 mM ammonium acetate (65:35, v/v) at a flow rate of 350 µL/min and a column temperature of 25 °C. Electrospray ionization was operated in negative ion mode with multiple reaction monitoring (MRM). The mass spectrometric parameters were set as follows: ion spray voltage (IS), 5.4 kV; curtain gas, 35 psi; collision gas, 10 psi; ion source temperature, 280 °C; nebulizer gas (GS1), 40 psi; and heater gas (GS2), 45 psi. The de-clustering potential (DP) and entrance potential (EP) were 75 eV and 10.0 eV, respectively. The collision energy (CE) and collision cell exit potential (CXP) were set to 10.0 eV and 22.0 eV. The mass transition ion-pair for DHA was selected as m/z 307.0→261.0.

Preparation and characterization of DHA-PRO-Loaded nanomicelles (DHA-PRO@NMs)

DHA-PRO-loaded nanomicelles (DHA-PRO@NMs) were prepared using the thin-film hydration method [47]. Briefly, DHA-PRO (1 mg), Soluplus (16 mg), and TPGS (4 mg) were dissolved in anhydrous ethanol (5 mL). The organic solvent was removed under reduced pressure using a rotary evaporator, and the resulting film was further dried overnight under vacuum in a desiccator to eliminate residual ethanol. The dried film was hydrated with purified water under stirring at room temperature for 30 min to form a micellar dispersion. The hydrodynamic size (SIZE), polydispersity index (PDI), and zeta potential (ZETA) of the resulting nanocomposites were measured using a Malvern Zeta Sizer Nano instrument (Nano ZS, Malvern Instruments, UK) at 25 °C. Morphological examination was performed by transmission electron microscopy (TEM, JEOL, JEM-2000EX, Japan). Subsequently, comprehensive Fourier transform infrared spectroscopy (FTIR) analysis was performed on DNA, DHA-PRO, lyophilized blank nanomicelles (NMs), and lyophilized DHA-PRO@NMs (Thermo Fisher Nicolets50). Drug loading (DL) and encapsulation efficiency (EE) were determined by HPLC-MS/MS. A batch of DHA-PRO@NMs was prepared using 1 mg DHA-PRO, 16 mg Soluplus, and 4 mg TPGS in 10 mL of final dispersion. A 100 µL aliquot was diluted to 1.0 mL with release medium (10% ethanol and 0.5% SDS, pH 7.4), sonicated for 30 min to disrupt the nanomicelles and precipitate DHA-PRO, treated with excessive H₂O₂, and incubated at 37 °C to facilitate DHA release. 1 mL ethyl acetate was added, after full vortex extraction, 500 µL supernatant was taken, nitrogen dried, redissolved with mobile phase. DL and EE were calculated using the following equations:

DL (%) = (mass of DHA-PRO in DHA-PRO@NMs)/(total mass of DHA-PRO@NMs) × 100%.

EE (%) = (mass of DHA-PRO in DHA-PRO@NMs)/(mass of DHA-PRO used in DHA-PRO@NMs preparation) × 100%.

The stability of DHA-PRO@NMs was evaluated by diluting 1 mL of the nanocomposite solution (DHA-equivalent concentration = 1 mg/mL) into 3 mL of various media: purified water, normal saline (NS), phosphate-buffered saline (PBS), and Dulbecco’s Modified Eagle Medium (DMEM) (pH 5.5 and pH 7.4). The samples were stored at 37 °C for 7 days, and SIZE and PDI were measured daily to assess the stability. The in vitro release profile of DHA from DHA-PRO@NMs in different concentrations of H₂O₂ release medium was investigated using a dialysis method. 1 ml DHA-PRO@NMs (DHA-equivalent concentration = 100 ug/mL) were placed in a dialysis bag and immersed in 5 mL of the release medium containing H₂O₂ at concentrations of 0, 3, 5, or 10 mM. The system was maintained at 37 °C under gentle shaking. Aliquots (500 µL) were collected at 3, 6, 9, 12, and 24 h, with an equal volume of the above medium replenished after each sampling. Add 1 mL ethyl acetate for extraction, take 800 µL supernatant, blow dry and redissolve with 500 µL mobile phase. DHA concentration was determined using HPLC-MS/MS.

Fabrication of the thermosensitive hydrogel system (DHA-PRO@NMs@TSH)

To achieve sustained retention in OA joints, DHA-PRO@NMs-loaded thermosensitive hydrogel (DHA-PRO@NMs@TSH) was prepared using a physical mixing method [48]. Briefly, predetermined amounts of P407 and P188 were weighed into a tube, followed by the addition of a DHA-PRO@NMs solution to achieve a total mass of 10 g. The mixture was stored at 4 °C for 24 h to facilitate complete polymer swelling and hydrogel formation, yielding DHA-PRO@NMs@TSH with varying P407/P188 mass ratios. The DHA concentration in all prepared DHA-PRO@NMs@TSH formulations was standardized to 1 mg/mL. Next, perform formulation screening for the mass ratio of P407/P188, which was first conducted by evaluating the sol-gel transition temperature. The gelation temperature of blank thermosensitive hydrogel (Blank@TSH) and DHA-PRO@NMs@TSH with varying mass ratios of P407/P188 was determined using the test tube inversion method. Briefly, 1.5 mL of each hydrogel formulation was placed in a 5 mL glass vial and immersed in a thermostatic water bath, ensuring the liquid level remained 1 cm below the water surface. The temperature was increased at a rate of 1 °C/min. At each 1 °C interval, the vial was inverted to assess flowability. The temperature at which no flow was observed for 30 s was recorded as the gelation point. Subsequently, formulation screening was performed based on the sol-gel transition time at physiological temperature. The gelation time-a critical parameter reflecting in situ gelling behavior-was measured at 37 °C using the test tube inversion method. Specifically, 1.5 mL of DHA-PRO@NMs@TSH was loaded into a 5 mL glass vial and immersed in a water bath maintained at 37 °C, with the liquid level kept 1 cm below the water surface. The timer was started immediately upon immersion. The vial was inverted every 5 s, and the gelation time was defined as the time point at which no flow was observed for 30 s after inversion, indicating complete gel formation under physiological conditions.

Morphological and rheological characterization of DHA-PRO@NMs@TSH

For morphological analysis, lyophilized samples of DHA-PRO@NMs@TSH were imaged using scanning electron microscopy (SEM) (JEOL, JSM-IT300, Japan). The rheological properties of DHA-PRO@NMs@TSH were evaluated using an Anton Paar MCR 301 rheometer. Both Blank@TSH and DHA-PRO@NMs@TSH (DHA-equivalent concentration = 1 mg/mL), prepared with P407/P188 at a mass ratio of 20%/3% (w/w), were loaded onto the rheometer plate. Rheological characterization was conducted by monitoring the storage modulus (G′) and loss modulus (G″) under two temperature conditions: ambient temperature (23 °C) and physiological temperature (37 °C). Measurements were performed at a fixed angular frequency of 1 Hz and 1% strain. To precisely determine the sol–gel transition temperature of DHA-PRO@NMs@TSH, temperature-sweep tests were carried out from 10 °C to 55 °C at a heating rate of 1 °C/min using a Discovery Hybrid Rheometer (DHR, TA Instruments, USA), during which changes in G′, G″, and viscosity (η) were recorded [49]. Next, we evaluated the physicomechanical stability of DHA-PRO@NMs@TSH in a simulated knee joint environment. The time-sweep test was conducted at 37 °C, a fixed frequency of 1 Hz, and a strain of 1% to monitor the evolution of G′ and G″ over 60 min. The frequency-sweep test was carried out at 37 °C and a strain of 1%, with frequency ranging from 0.1 to 100 Hz, to determine the dependence of G′ and G″ on deformation rate. The amplitude-sweep test was performed at 37 °C and a fixed frequency of 1 Hz, while the strain was logarithmically increased from 0.1% to 100%, to assess the yield point and the viscoelastic behavior under increasing deformation. To further investigate the self-recovery capability of the DHA-PRO@NMs@TSH under simulated joint movement, a repeated step-strain test was employed. This test consisted of 10 alternating cycles between a small stress (10 Pa, simulating the resting state) and a large stress (1000 Pa, simulating walking or running), with the corresponding G′ and G″ being recorded throughout the cycles.

In vitro Stability, Erosion, and release kinetics of DHA-PRO@NMs@TSH

The stability of DHA-PRO@NMs@TSH was evaluated based on its sol-gel transition temperature and rheological properties. First, the gelation temperature of freshly prepared DHA-PRO@NMs@TSH was measured on day 1 using the test tube inversion method. After storage at 37 °C for 7 days, the gelation temperature was reassessed under identical conditions to determine thermal stability. Subsequently, rheological stability was investigated by measuring the η of freshly prepared DHA-PRO@NMs@TSH on day 1 using an Anton Paar rheometer. The measurement was repeated after storing the hydrogel at 37 °C for 7 days under the same operational parameters to evaluate changes in viscosity.

The in vitro erosion behavior of DHA-PRO@NMs@TSH was evaluated using a membrane-free method [49]. Pre-weighed empty penicillin bottles were filled with 4 g of DHA-PRO@NMs@TSH and equilibrated in a thermostatic water bath shaker at 37.0 ± 0.1 °C for 10 min to facilitate complete gelation. Subsequently, 3 mL of PBS was added to each bottle, and the samples were incubated under agitation (100 rpm) at 37.0 ± 0.1 °C. At predetermined time points each day, the entire release medium was carefully decanted. The remaining gel was re-equilibrated in the water bath for 10 min before the addition of 3 mL of fresh PBS. This process was repeated until the residual gel mass was less than 5% of the initial weight. All experiments were conducted in triplicate. The erosion rate (Q) at each time point was calculated as follows:

Q (%) = [(m₀ - mₜ)/m₀] × 100% (m₀ = initial mass of DHA-PRO@NMs@TSH. mₜ = remaining mass of DHA-PRO@NMs@TSH at time t).

The in vitro release profile of DHA-PRO@NMs from DHA-PRO@NMs@TSH was investigated using a dialysis method [50]. Briefly, transfer 1 mL of DHA-PRO@NMs@TSH (DHA-equivalent concentration = 1 mg/mL) into a dialysis bag. Place the dialysis bag into a beaker containing 100 mL of release medium (10% ethanol and 0.5% SDS, pH 7.4) and shake at 37 °C at a constant speed (100 rpm). At predetermined time points each day, remove 1 mL of sample and replace it with an equal volume of the aforementioned release medium. The sample undergoes ultrasonication and H₂O₂ treatment, followed by ethyl acetate extraction and mobile phase redissolution. DHA concentration is determined using LC-MS/MS. Calculate the cumulative release of DHA at that time point based on the measured concentration. The cumulative release rate of DHA-PRO@NMs is indirectly reflected by the cumulative amount of DHA released, whitch was calculated using the formula:

Cumulative release rate (%) = (mₜ/m₀) × 100% (m₀ = the total amount of DHA loaded in DHA-PRO@NMs@TSH. mₜ = the cumulative amount of DHA released at time t).

Cells

Immortalized mouse chondrogenic ATDC5 cells were procured from Wuhan SUNNCELL Biotechnology Co., Ltd (Lot Number: 240304HJ). ATDC5 cells were cultured in DMEM medium supplemented with 10% fetal bovine serum and 1% penicillin/streptomycin in a 5% CO₂ incubator at 37 °C. Cells were trypsinized when they reached approximately 90% confluence and used for subsequent experiments.

Cytotoxicity assay

ATDC5 cells were seeded in 96-well plates at a density of 5 × 10³ cells per well in 100 µL of culture medium and incubated for 24 h. Cells were then treated with various concentrations (0.5, 1, 2, 4, 8, 16, and 32 µM, DHA-equivalent concentration) of the following compounds: DHA, a physical mixture of DHA and CA at a 2:1 molar ratio (hereafter referred to as DHA + CA), DHA-PRO, and DHA-PRO@NMs. After 24 h of exposure, 10 µL of CCK-8 reagent was added to each well, and the plates were incubated for an additional 1–4 h. The absorbance (A) was measured at 450 nm using a microplate reader. Cell viability was calculated according to the following formula:

Cell viability (%) = [(A_e−A_b)/(A_c−A_b)] × 100% (A_e: absorbance value of experimental group. A_b: absorbance value of blank group. A_c: absorbance value of the Control group, no drug added).

Cellular uptake

C6-loaded DHA-PRO@NMs were prepared by co-dissolving appropriate amounts of coumarin-6 (C6), DHA-PRO, Soluplus, and TPGS in ethanol, followed by rotary evaporation at 37 °C under light-protected conditions. The resulting film was dried overnight and subsequently rehydrated with sterile water to form C6&DHA-PRO@NMs. ATDC5 cells were seeded in 6-well plates and allowed to adhere by incubating for 24 h. Cells were then stimulated with or without 100 µM H₂O₂ for 48 h. After stimulation, the medium was replaced with fresh medium containing C6&DHA-PRO@NMs (DHA-equivalent concentration = 2 µM, C6: DHA-PRO = 1:10, mass ratio) and incubated for 1, 3, or 6 h. Following incubation, cells were washed three times with PBS and fixed with 4% paraformaldehyde for 15 min. Then, stained with 4′,6-diamidino-2-phenylindole (DAPI) for 10 min at room temperature. Cellular fluorescence was visualized and recorded using an inverted fluorescence microscope.

Intracellular Apoptosis/Necroptosis and ROS level detection, and ELISA analysis

Cell treatments for all the three experiments were as follows: ATDC5 cells were seeded in 12-well plates and allowed to adhere by incubating for 24 h. Cells were then stimulated with 100 µM H₂O₂ for 48 h. After stimulation, the medium was replaced with DHA, DHA + CA, DHA-PRO, or DHA-PRO@NMs (each DHA-equivalent concentration = 2 µM) for 24 h [51]. Intracellular Apoptosis/Necroptosis detection: cells were stained with 500 µL of YP1/PI working solution per well and incubated at 37 °C for 5–20 min. YP1-positive (green fluorescence) cells were considered apoptosis or necroptosis, while PI-positive (red fluorescence) cells indicated necroptosis. Fluorescence signals were captured using an inverted fluorescence microscope. Intracellular ROS detection: cells were incubated with 2′,7′-dichlorodihydrofluorescein diacetate (DCFH-DA) working solution (diluted 1:1000 in serum-free medium) at 37 °C for 60 min in the dark. Cells were then washed twice with serum-free medium and visualized under an inverted fluorescence microscope. ELISA analysis: Cell culture supernatants were collected from each group and centrifuged to remove debris. Levels of IL-1β and TNF-α were quantified using commercial ELISA kits according to the manufacturer’s instructions.

Western blot analysis

ATDC5 cells in each group were prepared according to the previously described method, while rat tissues in each group were obtained from subsequent animal experiments. First, cells or tissue were crushed and lysed using Radioimmunoprecipitation Assay (RIPA) buffer supplemented with phenylmethylsulfonyl fluoride (PMSF) and a protease inhibitor cocktail. The supernatant was collected by centrifugation at 12,000 ×g for 15 min at 4 °C, and the protein concentration was determined by Bicinchoninic Acid Assay (BCA, Solarbio) method. Equal amounts of protein were separated by sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS-PAGE) and transferred to polyvinylidene fluoride (PVDF, Millipore, Bedford, MA, USA) membranes. The membranes were blocked with 5% non-fat milk in Tris-buffered saline containing 0.1% Tween 20 (TBST) for 2 h at room temperature. Subsequently, the membranes were incubated with primary antibodies overnight at 4 °C, followed by incubation with horseradish peroxidase (HRP)-conjugated secondary antibodies for 90 min at room temperature. Protein bands were visualized using an enhanced chemiluminescence (ECL) kit (Millipore, USA) according to the manufacturer’s instructions.

Establishment of OA rat model

Male Sprague-Dawley (SD) rats (200 ± 20 g, 6–8 weeks old) were obtained from Liaoning Changsheng Biotechnology Co., Ltd. (China). Animals were housed under controlled conditions (temperature: 25 ± 2 °C, 12 h light/dark cycle) with ad libitum access to food and water. All experimental procedures were approved by the Institutional Animal Ethics Committee of Dalian Medical University and conducted in accordance with the Guide for the Care and Use of Laboratory Animals, License number: AEE21084. OA was induced in the left knee joint via intra-articular injection of MIA (2 mg in 50 µL normal saline) using a 26.5 G needle under anesthesia [51]. Seven days post-induction, rats were randomly assigned to 6 groups (n = 8 per group): the Control group: Healthy rats treated with normal saline (NS), the MIA group: OA rats treated with NS, the DHA + CA group: OA rats treated with DHA + CA, the DHA-PRO group: OA rats treated with DHA-PRO, the DHA-PRO@NMs group: OA rats treated with DHA-PRO@NMs, the DHA-PRO@NMs@TSH group: OA rats treated with DHA-PRO@NMs@TSH. Treatments were administered via intra-articular injection once every 7 days for a total of 4 doses. After the final evaluation, rats were euthanized, and the left knee joints, serum, and major organs (heart, liver, spleen, lungs, and kidneys) were collected. A subset of knee joints was fixed in 4% paraformaldehyde (PFA) for histological analysis, while the remaining joints were stored at − 80 °C for further studies.

In vivo imaging system (IVIS) analysis

The near-infrared probe IR808 and DHA-PRO were co-incorporated into the nanomicelles (NMs) and thermosensitive Hydrogel (TSH) for real-time tracking. Briefly, appropriate amounts of IR808, DHA-PRO, Soluplus, and TPGS were dissolved in anhydrous ethanol. The solution was subjected to rotary evaporation at 37 °C under light-protected conditions and dried overnight. The resulting film was reconstituted in normal saline to obtain the IR808&DHA-PRO@NMs solution. This solution was physically mixed with P407 and P188 (20%/3%, w/w), and the mixture was stored at 4 °C for 24 h to form IR808&DHA-PRO@NMs@TSH. To evaluate the in vivo retention time, OA rats received intra-articular injections (50 µL) of the following formulations: IR808 alone, IR808&DHA-PRO@NMs, or IR808&DHA-PRO@NMs@TSH (each IR808 = 0.2 mg/mL, each DHA-equivalent concentration = 1 mg/mL). Imaging was performed using an IVIS Spectrum system (PerkinElmer) with excitation and emission wavelengths set at 745 nm and 800 nm, respectively. Fluorescence intensity was quantified at predetermined time points to evaluate joint retention time.

Micro-CT analysis

Following euthanasia, the left knee joints in each group were dissected, and surrounding soft tissues were carefully removed. Specimens were fixed in 4% paraformaldehyde for 24 h and subsequently scanned using a high-resolution micro-CT system (Skyscan 1276, Kontich, Belgium). Scanning parameters were set as follows: voxel size 10 μm, medium resolution, voltage 85 kV, current 200 µA, aluminum filter 1 mm, and integration time 384 ms. Quantitative morphometric analysis was performed to evaluate the following parameters: Bone mineral content (BMC), Bone mineral density (BMD), Bone surface density (BS/TV), Percent bone volume (BV/TV), Trabecular number (Tb. N), and Trabecular separation (Tb. Sp). Two-dimensional (2D) image stacks were reconstructed into three-dimensional (3D) models to facilitate structural visualization and assessment of joint integrity.

Immunofluorescence and histopathological analysis

For immunofluorescence analysis, sections were sequentially subjected to deparaffinization, rehydration, antigen retrieval, and blocking. They were then incubated with primary antibodies against collagen II, aggrecan, and Matrix Metalloproteinase 13 (MMP13) overnight at 4 °C, followed by incubation with appropriate fluorescently labeled secondary antibodies. Finally, images were captured using a fluorescence microscope. For histopathological analysis, the left knee joints from each group were fixed in 4% paraformaldehyde for 24 h, decalcified in 10% ethylenediaminetetic acid (EDTA), embedded in paraffin, and sectioned sagittally at a thickness of 4 μm. Sections were stained with hematoxylin and eosin (H&E), toluidine blue (TB), and safranin O-fast green (SO-FG) for histopathological evaluation. Three experienced pathologists conducted the Mankin score based on the three staining results.

Gait analysis

To minimize stress-related variability, all rats were acclimatized to the walkway environment for three days prior to formal gait assessment. On day 35 post-treatment, quantitative gait analysis was performed using the VisuGait system (Xinruan Scientific, Shanghai, China). The following parameters were evaluated to characterize locomotor function of rats in each group: Paw touchdown area, Paw touchdown pressure and Paw touchdown time [52].

In vivo biosafety evaluation

Following euthanasia, serum and major organs (heart, liver, spleen, lungs, and kidneys) were collected from each rat. Serum levels of ALT, AST, BUN, CRE, and LDH were measured using commercial assay kits according to the manufacturers’ instructions. Additionally, the collected organs were fixed, paraffin-embedded, sectioned, and stained with H&E for histopathological examination to further assess systemic toxicity.

Statistical analysis

All statistical analyses were conducted using GraphPad Prism 10 software (GraphPad Software, USA). Data obtained from at least three independent experiments are presented as mean ± standard deviation (mean ± SD). Comparisons among multiple groups were performed using one-way analysis of variance (ANOVA) followed by Tukey’s test. For comparisons between two groups, Student’s t-test was applied. A p-value of less than 0.05 was considered statistically significant.

Results and discussion

Synthesis and characterization of the ROS-Responsive prodrug (DHA-PRO)

Building upon the established therapeutic potential of dihydroartemisinin (DHA) in osteoarthritis (OA) [24, 53], we sought to leverage the elevated reactive oxygen species (ROS) levels in the OA joint microenvironment to develop a targeted prodrug strategy [54]. To this end, we designed a ROS-responsive DHA prodrug. In preliminary screening, we evaluated two candidates, DHA-S-CA1 and DHA-S-CA2 (Fig. S1), both synthesized by conjugating DHA with cinnamaldehyde (CA) via a thioacetal linker. DHA-S-CA2 was selected for this study due to its superior pharmaceutical profile: it provided a more sustained release of DHA under identical ROS conditions and, owing to its molecular structure, yielded a higher molar amount of active DHA per equivalent mass compared to DHA-S-CA1 [47]. We then optimized the synthesis process to directly obtain DHA-S-CA2, which we designated as DHA-PRO (synthetic route shown in Fig. 2A). The core design of DHA-PRO features a thioacetal bond, which serves a dual function: first, it remains stable in normal tissues but is specifically cleaved in ROS-rich inflammatory environments, enabling controlled drug release; second, the cleavage process itself consumes excess ROS, providing an additional therapeutic effect. This mechanism not only mitigates the inherent instability of native DHA but also ensures site-selective activation within diseased joints. The chemical structure of DHA-PRO was unequivocally confirmed by High-Resolution Mass Spectrometry (HR-MS) and ¹H Nuclear Magnetic Resonance (¹H NMR) (Fig. S2 and Fig. 2B). To evaluate its responsiveness, DHA release from DHA-PRO was examined in media containing different H₂O₂ concentrations (0, 3, 5, and 10 mM). As shown in Fig. S3, DHA-PRO released ~ 85%, ~ 65%, and ~ 45% of DHA within 24 h in 10 mM, 5 mM, and 3 mM H₂O₂, respectively, while < 3% was released under ROS-free conditions. These results demonstrate that DHA-PRO enables rapid and ROS-dependent DHA release, with release kinetics positively correlated with H₂O₂ concentration. Importantly, this suggests that DHA-PRO may release faster in the joints of patients with severe OA and slower in those with milder disease, while preferentially releasing more DHA in inflamed regions than in normal tissue. Such pathologically and spatially differentiated release behavior supports the potential of DHA-PRO for achieving graded and on-demand DHA delivery in OA.

Fig. 2.

Preparation and characterization of DHA-PRO and DHA-PRO@NMs. (A) Synthetic scheme of DHA-PRO.(B) The structures of DHA-PRO were determined by ¹H NMR. (C) Macroscopic appearance and (D) Tyndall effect of DHA-PRO@NMs in aqueous solution. (E) Representative TEM image of DHA-PRO@NMs. (F, G, H) Hydrodynamic size (SIZE), polydispersity index (PDI), and zeta potential (ZETA) of DHA-PRO@NMs. (I) FTIR spectra of DHA, DHA-PRO, Blank nanomicelles (NMs), and DHA-PRO@NMs. Data are shown as mean ± SD (n = 3)

Preparation and characterization of DHA-PRO-Loaded nanomicelles (DHA-PRO@NMs)

Despite its therapeutic potential, the synthesized DHA-PRO suffered from poor aqueous solubility. To overcome this limitations, we developed a DHA-PRO-loaded nanomicelles (DHA-PRO@NMs) using Soluplus and TPGS as carriers via thin-film hydration, with an optimized ratio of DHA-PRO: Soluplus: TPGS = 1:16:4 (w/w/w) [47]. As shown in Fig. 2C, DHA-PRO@NMs dissolved well in water and were uniformly dispersed, while a clear Tyndall effect further confirmed the colloidal nature of the nanocomposites (Fig. 2D). Transmission electron microscopy (TEM) imaging revealed spherical particles with an average diameter of ~ 80 nm, smooth morphology, and uniform size distribution (Fig. 2E). Dynamic light scattering (DLS) confirmed an average hydrodynamic size (SIZE) of 70.48 nm with a narrow distribution (PDI = 0.066) and a zeta potential (ZETA) of −9.72 mV, indicating good stability (Fig. 2F-H). As shown in Fig. 2I, compared to DHA, the Fourier transform infrared spectroscopy (FTIR) spectrum of DHA-PRO exhibits the disappearance of the hydroxyl group at 3366 cm⁻¹ (-OH stretching vibration peak), the appearance of an ester carbonyl group at 1748 cm⁻¹ (C = O stretching vibration peak), and the emergence of a new disulfide bond at 763 cm⁻¹ (S-S stretching vibration peak), all indicating the successful synthesis of DHA-PRO. The FTIR spectra of NMs and DHA-PRO@NMs exhibit high similarity, suggesting that the characteristic peaks of DHA-PRO are largely masked by the nanocarrier matrix. This spectral behavior confirms that DHA-PRO has been successfully physically encapsulated within the NMs. The drug loading (DL) and encapsulation efficiency (EE) were 4.41 ± 0.02% and 92.18 ± 0.41%, respectively, as determined by validated HPLC-MS/MS. Stability tests demonstrated that DHA-PRO@NMs maintained clear appearance without precipitation for 7 days in H₂O, NS, PBS, and DMEM at both physiological (pH 7.4) and inflammatory (pH 5.5) conditions (Fig. S4A). SIZEs remained between 60 and 150 nm, and all PDIs were < 0.15 (Fig. S4B-E), confirming favorable colloidal stability. ROS-responsive release studies showed that within 24 h, DHA-PRO@NMs released ~ 70%, ~ 50%, and ~ 30% of DHA under 10 mM, 5 mM, and 3 mM H₂O₂, respectively, whereas negligible release occurred without H₂O₂ (Fig. S4F). These results confirm that DHA-PRO@NMs retain the ROS-responsive release properties of DHA-PRO, with release rates positively correlated with H₂O₂ concentration. Furthermore, compared with DHA-PRO, which released ~ 85% DHA in 10 mM H₂O₂, DHA-PRO@NMs released ~ 70% (Fig. S4G), indicating a slower but more sustained release profile. This sustained release may prolong drug action time, potentially reducing injection frequency. In summary, DHA-PRO@NMs successfully inherits the graded delivery and responsive release characteristics of DHA-PRO. Moreover, it simultaneously addresses the poor aqueous solubility of DHA-PRO, provides stable colloidal properties, and possesses demonstrated sustained release properties.

Fabrication and characterization of the thermosensitive hydrogel system (DHA-PRO@NMs@TSH)

Given that the highly water-soluble DHA-PRO@NMs are likely to be rapidly metabolized from OA joints. To prolong intra-articular retention time and reduce dosing frequency, we incorporated a thermosensitive hydrogel (TSH) delivery system in animal studies. TSH remains an injectable solution at low temperatures and undergoes a sol–gel transition at physiological temperature, forming a stable depot for sustained drug release [55, 56]. Poloxamer 407 (P407) and Poloxamer 188 (P188) were selected as the TSH matrix materials owing to their well-established thermoresponsive behavior (Fig. 3A) and good biocompatibility [42, 57]. We first performed formulation screening of P407/P188 to identify the optimal composition for DHA-PRO@NMs@TSH intended for intra-articular administration. The sol–gel transition temperature is critically influenced by the P407/P188 ratio. We aimed for an optimal transition range of 27–32 °C to balance injectability at room temperature with rapid gelation at body temperature. Initial screening with blank TSH (Blank@TSH) identified several candidate ratios (Supplementary Table 1). Subsequent screening using DHA-PRO@NMs as the solvent (Supplementary Table 2) yielded four formulations that met the temperature criterion: 18%/0% (28 °C), 18%/2% (30 °C), 20%/3% (28 °C), and 20%/5% (30 °C). As blended P407/P188 hydrogels exhibit greater mechanical strength and lower burst release than P407 alone, the 18%/0% group was excluded [46, 58]. Gelation time measurements further showed that the 20%/3% (85 s) and 20%/5% (75 s) groups gelled faster than the 18%/2% group (110 s), which would minimize post-injection spreading, the 18%/2% group was excluded. To reduce excipient load while maintaining satisfactory performance, the 20%/3% formulation was selected for subsequent studies. Needle flowability tests confirmed its good injectability. Notably, replacing water with DHA-PRO@NMs as the solvent only slightly lowered the transition temperature (by ~ 1 °C), indicating minimal impact on the thermosensitivity of the Poloxamer matrix. The final DHA-PRO@NMs@TSH formulation (designated as DHA-PRO@NMs@TSH) remained a free-flowing sol at 23 °C and rapidly transitioned to a non-flowing, homogeneous gel without precipitation at 37 °C (Fig. 3B). Overall, the selected formulation exhibited a sol–gel transition at 28 °C with a gelation time of 85 s, demonstrating high suitability for intra-articular injection. Lyophilized DHA-PRO@NMs@TSH samples were then examined by scanning electron microscopy (SEM), which revealed a uniform, porous three-dimensional network structure suitable for drug encapsulation and controlled release (Fig. 3C).

Fig. 3.

Characterization of DHA-PRO@NMs@TSH. (A, B) Sol-gel transition behavior of Blank@TSH and DHA-PRO@NMs@TSH. (C) SEM image of the lyophilized DHA-PRO@NMs@TSH. (D) Storage modulus (G′) and loss modulus (G″) of Blank@TSH and DHA-PRO@NMs@TSH at 23 °C and 37 °C. (E, F) Temperature-dependent rheological properties: G′, G″, and viscosity (η) during a temperature-sweep from 10 °C to 55 °C. (G) Time-dependent rheological properties: G′ and G″ during a time-sweep in 60 min. (H) Frequency-dependent rheological properties: G′ and G″ during a frequency-sweep from 0.1 Hz to 100 Hz. (I) Amplitude-dependent rheological properties: G′ and G″ during an amplitude-sweep from 0.1% stain to 100% stain. (J) Repeated step-strain test: G′ and G″ during a repeated step-strain test (10 cycles), alternating between low strain (﹤0.1% stain, 10 pa) and high strain (﹥100% stain, 1000pa)

The rheological properties of DHA-PRO@NMs@TSH were characterized using a rotational rheometer. In the rheological analysis, the viscoelastic properties and viscosity (η) of the hydrogel were evaluated. The viscoelastic behavior was assessed by the storage modulus (G′) and the loss modulus (G″). G′ reflects the elastic, solid-like behavior and indicates the hydrogel’s ability to recover from deformation, while G″ represents the viscous, liquid-like behavior and the energy dissipation under applied stress. A higher G′ indicates greater elasticity, whereas a higher G″ suggests more dominant flow behavior. When G′ < G″, the material behaves as a liquid-like sol with low moduli, while G′ > G″ indicates a solid-like gel with markedly increased moduli. η was also monitored as an indicator of network compactness. A low η corresponds to a loose network facilitating injection, while a high η reflects a dense, cross-linked structure conferring mechanical stability and prolonged drug release. The sol–gel transition temperature was defined as the point where G′ = G″, which typically coincided with a sharp increase in η [59–61]. Measurements were performed at an angular frequency of 1 Hz and 0.1% strain. At 23 °C, both Blank@TSH and DHA-PRO@NMs@TSH exhibited G′ < G″ with low modulus values (< 10 Pa), confirming a flowable liquid state suitable for injection. At 37 °C, both G′ and G″ increased sharply to approximately 10,000 Pa, with G′ > G″, indicating a non-flowing gel state (Fig. 3D). Incorporation of DHA-PRO@NMs caused only minimal changes in the rheological parameters compared to Blank@TSH, confirming compatibility with the TSH system. These results demonstrated that DHA-PRO@NMs@TSH possessed ideal rheological characteristics for intra-articular administration: low viscosity and good injectability at room temperature, and rapid formation of a stable, high-viscosity gel at body temperature conducive to sustained drug release. To precisely determine the sol–gel transition temperature of DHA-PRO@NMs@TSH, a temperature sweep rheological test was conducted from 10 °C to 55 °C while monitoring G′, G″, and η. As shown in Fig. 3E and F, at temperatures below 27 °C, G′ < G″ and η remained low, indicating a flowable liquid state. At 27 °C, G′ = G″, marking the sol–gel transition point, and a sharp increase in η was observed. At temperatures above 27 °C, G′ > G″ and η remained elevated, confirming a stable gel state. The sol–gel transition temperature was thus determined to be 27 °C, which differed by only 1 °C from the 28 °C measured by the test tube inversion method. This minor discrepancy was likely due to methodological differences: rheometry provides a quantitative, mechanics-based measurement, whereas the inversion method is qualitative and observation-based. Despite this variation, both values fell within the predefined optimal range of 27–32 °C, confirming that DHA-PRO@NMs@TSH exhibited suitable thermoresponsive behavior for intra-articular injection and sustained drug release. To further evaluate the physicomechanical stability of DHA-PRO@NMs@TSH, we conducted time-sweep, frequency-sweep, amplitude-sweep, and repeated step-strain test. The results showed that both time-sweep and frequency-sweep tests confirmed the remarkable long-term and structural stability of DHA-PRO@NMs@TSH, as evidenced by the constant and dominant G′ over 60 min and across a wide frequency range (Fig. 3G and H). Amplitude sweep testing identified a distinct yield stress point, beyond which the hydrogel structure exhibited temporary failure (G′ < G″) (Fig. 3I). The results indicate that DHA-PRO@NMs@TSH maintains a high modulus (G′ > G″) from 0.1% to 6% strain. It can thus be concluded that DHA-PRO@NMs@TSH exhibits solid-like properties and mechanical strength when the knee joint is at rest or in non-weight-bearing areas. However, when the knee joint is in motion during activities such as walking or running (strain > 6%), DHA-PRO@NMs@TSH in the weight-bearing areas often reaches its yield point [62–64]. To determine whether this system possesses “gel-yield-recovery” properties, we proceeded to conduct a repeated step-strain test. We conducted a repeated step-strain test (10 cycles), alternating between low strain (below yield, < 0.1% strain, simulating rest) and high strain (above yield, > 100% strain, simulating walking or running) to directly simulate the cyclic loading in a moving joint. The results confirm that during each low-strain phase, DHA-PRO@NMs@TSH instantaneously and fully recovered its solid-like structure (G′ > G″) (Fig. 3J). The perfect reproducibility over 10 cycles provides robust evidence for its exceptional self-recovery capability and mechanical resilience.

The stability of DHA-PRO@NMs@TSH was evaluated by monitoring its sol–gel transition temperature and η over 7 days at 37 °C. The transition temperature, measured using the test tube inversion method, remained constant at 28 °C on both day 1 and day 7. Consistently, rheological measurements showed that η also remained stable after 7 days of storage, with values consistently below 1 Pa·s (Fig. S5A). These findings confirmed that the formulation maintained good stability under physiological temperature conditions, supporting its suitability for subsequent in vivo studies. The erosion behavior of DHA-PRO@NMs@TSH was examined using a membrane-free dissolution method. The hydrogel exhibited gradual and uniform erosion over 8 days, with a total mass loss of 98.5% and a nearly constant daily erosion rate (Fig. S5B). This homogeneous erosion behavior was attributed to the robust mechanical properties of the 20%/3% P407/P188 blend and the enhanced structural integrity imparted by DHA-PRO@NMs, which increased viscosity and resistance to deformation. Drug release kinetics were evaluated by quantifying DHA in the dissolution medium daily via HPLC–MS/MS. The cumulative release profile of DHA-PRO@NMs closely paralleled the erosion profile, with over 90% of DHA-PRO@NMs released by day 8 at a steady rate (Fig. S5B). The close correlation between erosion and release indicated that drug release was primarily governed by hydrogel erosion, which is consistent with the established mechanism of matrix-type drug delivery systems [65]. The nearly superimposable erosion and release curves further suggested that DHA-PRO@NMs was uniformly distributed throughout the hydrogel matrix and released in a controlled, erosion-dependent manner.

DHA-PRO@NMs attenuated H₂O₂-induced PANoptosis and inflammation in ATDC5 cells

This study employed H₂O₂-induced ATDC5 cells as an in vitro model of inflammatory chondrocytes to evaluate the efficacy of an intelligently delivered DHA formulation, DHA-PRO@NMs [66]. Initially, we determined the safe concentration ranges of DHA, a physical mixture of DHA and CA at a 2:1 molar ratio (hereafter referred to as DHA + CA), DHA-PRO, and DHA-PRO@NMs using a CCK-8 assay. ATDC5 cells were treated with each compound at DHA-equivalent concentrations ranging from 0.5 to 32 µM for 24 h. While DHA, DHA + CA, and DHA-PRO did not induce significant cytotoxicity up to 32 µM, DHA-PRO@NMs reduced cell viability to approximately 90% at 2 µM and 80% at 4 µM (Fig. 4A). Therefore, a concentration of 2 µM (DHA-equivalent) was selected for subsequent experiments to ensure both safety and efficacy. To assess cellular uptake, we labeled DHA-PRO@NMs with C6 and incubated ATDC5 cells for 1, 3, and 6 h. Fluorescence microscopy revealed time-dependent accumulation of C6&DHA-PRO@NMs in the cytoplasm. Furthermore, in H₂O₂-pretreated (100 µM, 48 h) ATDC5 cells, the fluorescence intensity was significantly higher at each time point compared to untreated cells, suggesting enhanced cellular uptake under inflammatory conditions, possibly due to elevated intracellular ROS levels (Fig. 4B and C). This suggests that DHA-POR@NMs may be taken up more extensively in diseased areas within the same joint compared to healthy regions. This uptake characteristic facilitates the potential of DHA-PRO@NMs for achieving graded DHA delivery in OA.

Fig. 4.

DHA-PRO@NMs attenuate H₂O₂-induced PANoptosis and inflammation in ATDC5 cells. (A) Cell viability after 24 h treatment (mean ± SD, n = 6). (B, C) Cellular uptake of C6-labeled DHA-PRO@NMs over time in cells with or without H₂O₂ pretreatment: (B) representative images (scale bar = 20 μm) and (C) quantitative analysis. (D, E) Intracellular Apoptosis/necroptosis levels assessed by YO-PRO-1 (YP-1, green) and PI (red) staining: (D) representative images (scale bar = 100 μm) and (E) quantification analysis. (F, G) Intracellular ROS levels detected by DCFH-DA staining: (F) representative images (scale bar = 100 μm) and (G) quantification analysis. Data are shown as mean ± SD (n = 3), and *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001, ns means not significant

We next evaluated the Apoptosis/Necroptosis level using a YO-PRO-1/Propidium Iodide (YP1/PI) dual staining assay. PANoptosis-a coordinated cell death process involving pyroptosis, apoptosis, and necroptosis-was markedly induced by H₂O₂ treatment. The level of Apoptosis/Necroptosis can, to some extent, reflect the overall activity of PANoptosis. The results showed that all DHA formulations reduced the level of Apoptosis/Necroptosis, with DHA-PRO@NMs exhibiting the most potent protective effect (Fig. 4D, E). Notably, no significant difference was observed between the DHA group and the DHA + CA group, indicating that CA alone does not enhance anti-PANoptosis activity. Given the mutually reinforcing relationship between ROS and inflammation [67], we measured intracellular ROS levels using a DCFH-DA assay, indirectly reflected the level of inflammation. H₂O₂ stimulation increased ROS levels approximately fourfold, which was effectively mitigated by all treatment groups. DHA-PRO@NMs demonstrated the strongest ROS-scavenging capacity, nearly restoring basal levels (Fig. 4F, G). Again, the DHA group and the DHA + CA group showed comparable effects, confirming that CA does not contribute independently to ROS reduction. We further quantified the levels of key inflammatory cytokines IL-1β and TNF-α using Enzyme-linked immunosorbent assay (ELISA). H₂O₂ significantly elevated both cytokines, while all treatments reduced their secretion, with DHA-PRO@NMs being the most effective (Fig. S6). The similarity between the DHA group and DHA + CA group further supports that CA serves primarily as a prodrug constituent rather than an active anti-inflammatory agent. In summary, these results preliminarily demonstrate that DHA-PRO@NMs exert robust anti-PANoptosis and anti-inflammatory effects in inflammatory chondrocytes, outperforming other formulations. At the same time, we can observe that the inclusion of CA in the prodrug design facilitates assembly but does not enhance pharmacological activity. This may be attributed to the negligible content of CA. Therefore, subsequent experiments will focus on DHA-PRO@NMs without further inclusion of the DHA-alone group.

To further elucidate the protective effects of DHA-PRO@NMs mediated through the suppression of PANoptosis, we analyzed the expression of key proteins associated with the three constituent cell death pathways: apoptosis, nec4optosis, and pyroptosis [68–70]. First, the proteins B-cell lymphoma-2 associated X protein (Bax) and B-cell lymphoma-2 (Bcl-2) are classical markers of apoptosis and play key roles in the process of apoptosis. Bax is a pro-apoptotic protein that initiates the apoptotic program by releasing intracellular apoptosis-associated factors, such as cytochrome C, into the cytoplasm of the cell, mainly by increasing the permeability of the mitochondrial membrane. Bcl-2 is an apoptosis-inhibiting protein that protects cells from apoptosis primarily by inhibiting the activity of pro-apoptotic factors such as Bax. Therefore, increased ratio of Bax/Bcl-2 can reflect upregulation of apoptosis. Then, Receptor-Interacting Protein Kinase 3 (RIPK3) is a serine/threonine kinase, the phosphorylated form of which (p-RIPK3) is a central marker of activation of necroptosis signaling pathway. During necroptosis, the expression of both RIPK3 and p-RIPK3 is usually increased. Activated RIPK3 executes necroptosis by phosphorylating mixed-lineage kinase domain-like (MLKL). Therefore, increased expression of RIPK3 and p-RIPK3 proteins can reflect upregulation of necroptosis [71]. Finally, when cells are stimulated by injuries such as inflammation, the NOD-like receptor thermal protein domain associated protein 3 (NLRP3) is activated and binds to apoptosis-associated speck-like protein containing a CARD (ASC) and Pro-caspase-1 to form inflammasome. Within the inflammasome, the clustering of Pro-caspase-1 molecules leads to their auto-cleavage and activation. Caspase-1 is a pivotal executor of cellular pyroptosis. Activated Caspase-1 cleaves the N-terminal sequences of Gasdermin D (GSDMD), which binds to the cell membrane to create membrane pores, leading to cellular pyroptosis. Usually the expression of proteins NLRP3, Pro-caspase-1, Caspase-1, Full-length GSDMD (GSDMD-F), and N-terminal sequence of GSDMD (GSDMD-N) increases significantly during this process. Therefore, increased expression of these proteins can reflect upregulation of pyroptosis [72]. Western Blot (WB) analysis revealed that H₂O₂ stimulation significantly upregulated the expression of apoptosis-associated protein (BAX/Bcl-2), necroptosis-associated proteins (RIPK3 and p-RIPK3), and pyroptosis-associated proteins (NLRP3, pro-caspase-1, Caspase-1, GSDMD-F, and GSDMD-N). Treatment with DHA + CA, DHA-PRO, or DHA-PRO@NMs can partially reverse these changes, with the greatest degree of reversal in the DHA-PRO@NMs group (Fig. 5A-K). These results showed that DHA-PRO@NMs had the strongest inhibitory effect on H₂O₂-induced PANoptosis in ATDC5 cells.

Fig. 5.

DHA-PRO@NMs attenuate H₂O₂-induced PANoptosis and inflammation in ATDC5 cells. (A, B) Representative Western Blot (WB) images and quantification analysis of apoptosis-related proteins (Bax and Bcl-2). (C-E) Representative WB images and quantification analysis of necroptosis-related proteins (RIPK3 and p-RIPK3). (F-K) Representative WB images and quantification analysis of pyroptosis-related proteins (NLRP3, Caspase-1, Pro-caspase-1, GSDMD-F, and GSDMD-N). (L-N) Representative WB images and quantification analysis of ROS metabolism-related proteins (NOX4 and SOD2). (O-T) Representative WB images and quantification analysis of NF-κB-related proteins (p-IKKβ, IKKβ, p-P65, P65, p-IκBα, and IκBα), followed by representative WB images and quantification analysis of inflammatory cytokines (IL-1β and TNF-α). Data are shown as mean ± SD (n = 3), and *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001, ns means not significant.

Next, we further elucidate the protective mechanism of DHA-PRO@NMs by inhibiting inflammation. Given the reciprocal reinforcement between ROS and inflammation, we evaluated the expression of NADPH oxidase 4 (NOX4), a major ROS generator, and superoxide dismutase 2 (SOD2), a key mitochondrial antioxidant enzyme [73]. H₂O₂-induced increase in NOX4 and decrease in SOD2 were effectively counteracted by all treatments, with DHA-PRO@NMs exhibiting the strongest restorative effect (Fig. 5L–N). This suggests that DHA-PRO@NMs effectively inhibited ROS over-generation and potentially possessed better anti-inflammatory capacity. The NF-κB pathway and inflammatory cytokines play central roles in OA-related inflammation [74]. In resting cells, NF-κB (p65-p50) is sequestered in the cytoplasm by inhibitor of nuclear factor κB α (IκBα). Inflammatory stimuli activate inhibitor of nuclear factor κB kinase β (IKKβ), leading to IκBα phosphorylation and degradation, allowing NF-κB translocation to the nucleus, where it promotes the transcription of pro-inflammatory genes, including IL-1β and TNF-α. Phosphorylation of RelA protein (p65) further enhances its transcriptional activity [75]. H₂O₂ induction significantly increased the ratios of p-IKKβ/IKKβ, p-IκBα/IκBα, and p-p65/p65, indicating NF-κB pathway activation. All treatments reduced these ratios and downstream IL-1β and TNF-α expression, with DHA-PRO@NMs again showing the greatest inhibitory effect (Fig. 5O–T). The above results confirmed that DHA-PRO@NMs could inhibit the activation of the NF-κB signaling pathway and the overproduction of the inflammatory cytokines IL-1β and TNF-α, showing the strongest anti-inflammatory capacity.

DHA-PRO@NMs@TSH attenuated MIA-induced OA in vivo

Building upon the demonstrated anti-PANoptosis and anti-inflammatory efficacy of DHA-PRO@NMs in H₂O₂-induced ATDC5 cells, we next sought to achieve sustained intra-articular delivery using our TSH system, DHA-PRO@NMs@TSH. The animal experiment flowchart is shown in Fig. 6A. Firstly, to evaluate the retention behavior of DHA-PRO@NMs@TSH in OA joints, IR808-labeled formulations (free IR808, IR808&DHA-PRO@NMs, and IR808&DHA-PRO@NMs@TSH) were intra-articularly administered into Monosodium iodoacetate induced OA rats. In vivo imaging system (IVIS) results revealed that fluorescence intensity declined over time in all groups, with free IR808 detectable for ~ 2 days, IR808&DHA-PRO@NMs for ~ 3 days, and IR808&DHA-PRO@NMs@TSH for ~ 7 days (Fig. 6B). These results indicate that the NMs encapsulation modestly prolonged retention, while incorporation into the TSH system significantly extended joint residence time to approximately 7 days. Therefore, we speculate that DHA, delivered via DHA-PRO@NMs@TSH, can achieve at least 7 days of retention in OA joints. Based on this, a 7-day dosing interval was selected for subsequent animal experiments. Animal experiments were then initiated, OA was induced in the SD rats via intra-articular injection of Monosodium iodoacetate (MIA). Given the potent anti-PANoptosis effects demonstrated by the aforementioned DHA formulation in cellular experiments, and considering that poloxamer, being similar to hyaluronic acid, is only expected to provide lubrication without the capacity to prevent the death of articular chondrocytes, we decided against using Blank@TSH alone for the treatment of OA [76, 77]. The rats were then treated with normal saline (NS), DHA + CA, DHA-PRO, DHA-PRO@NMs, or DHA-PRO@NMs@TSH, respectively. All animal experiments were conducted in accordance with the established experimental protocol. One week after the final treatment, a comprehensive set of evaluations and analyses was performed on each group of rats (Fig. 6A).

Fig. 6.

DHA-PRO@NMs@TSH attenuated MIA-induced OA in vivo. (A) Schematic of animal experiments timeline. (B) In vivo fluorescence imaging showing the joint retention of free IR808, IR808-labeled DHA-PRO@NMs, and IR808-labeled DHA-PRO@NMs@TSH over time (n = 3). (C) Representative images of touchdown area, the left hind paw was magnified. (D) Representative images of touchdown pressure, the left hind paw was demarcated purple color. (E) Representative images of touchdown time, the left hind paw was magnified and demarcated purple color. (F-H) Quantitative analysis of (F) touchdown area, (G) touchdown pressure, and (H) touchdown time of the left hind paw in each group. Data are shown as mean ± SD (n = 6), and *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001, ns means not significant

OA leads to pain, joint dysfunction, and gait abnormalities. Gait analysis serves as an important behavioral assessment method to quantify motor impairments, indirectly reflecting OA severity [78]. Using an automated gait analysis system, we evaluated touchdown area (Fig. 6C), touchdown pressure (Fig. 6D), and touchdown time (Fig. 6E) of paws in each group during platform traversal. Results revealed that the OA-affected left hind paw exhibited significant reductions in touchdown area, pressure, and time compared to normal rats. These deficits were ameliorated following treatment with DHA + CA, DHA-PRO, DHA-PRO@NMs, or DHA-PRO@NMs@TSH. Although no statistically significant difference in touchdown time was observed between the DHA-PRO@NMs group and the DHA-PRO@NMs@TSH group, the combined improvements in touchdown area and pressure indicated that DHA-PRO@NMs@TSH produced the most substantial restorative effect (Fig. 6F–H). These behavioral results confirm that DHA-PRO@NMs@TSH effectively attenuates OA progression.

Morphological observations within the joint can reflect the severity of OA to a certain extent. Gross examination of joint morphologyrevealed severe cartilage erosion, subchondral bone sclerosis, synovitis, and osteophyte formation in the MIA group, all of which were ameliorated to varying degrees by treatments, with DHA-PRO@NMs@TSH exhibiting the most protective effect (Fig. 7A).

Fig. 7.

DHA-PRO@NMs@TSH attenuated MIA-induced OA in vivo. (A) Macroscopic appearance (external and cross-sectional views) of rat knee joints. (B) Representative micro-CT images showing subchondral bone details (top), 3D reconstructions (middle), and 2D coronal sections (bottom). (n = 2) (C) Immunofluorescence staining of extracellular matrix components (collagen II, aggrecan) and catabolic enzyme MMP13. Nuclei (DAPI, blue), collagen II/aggrecan (FITC, green), MMP13 (Cy3, red). Scale bar = 100 μm. (D) Histological evaluation by hematoxylin and eosin (H&E), toluidine blue (TB), and safranin O-fast green (SO-FG) staining. Scale bar = 2 mm. (n = 2)

The therapeutic efficacy of DHA-PRO@NMs@TSH was further assessed using micro-CT, which provides superior resolution and cross-sectional imaging capabilities compared to conventional radiography for evaluating OA structural damage [79]. Micro-CT enables detailed detection of characteristic OA pathologies, including joint space narrowing, subchondral bone sclerosis, subchondral cyst formation, and osteophyte proliferation at joint margins-all key indicators of disease severity [80]. Results showed that compared to the Control group, the MIA group exhibited significant joint space narrowing, pronounced subchondral bone sclerosis, and evident periarticular osteophyte formation. These pathological structural changes were markedly attenuated following treatment with DHA + CA, DHA-PRO, DHA-PRO@NMs, or DHA-PRO@NMs@TSH. Among all treatment groups, the DHA-PRO@NMs@TSH group demonstrated the most substantial reversal of these imaging abnormalities, indicating its superior capacity to mitigate OA-associated joint destruction (Fig. 7B). Quantitative analysis of bone architectural parameters provided further evidence of treatment efficacy. MIA induction significantly reduced Bone mineral content (BMC), Bone mineral density (BMD), Bone surface density (BS/TV), Percent bone volume (BV/TV), Trabecular number (Tb. N), while increasing Trabecular separation (Tb. Sp). Administration of DHA + CA, DHA-PRO, DHA-PRO@NMs, or DHA-PRO@NMs@TSH partially reversed these deteriorations in bone microstructure, with the most pronounced improvements observed in the DHA-PRO@NMs@TSH group (Fig. S7A-F). In summary, these micro-CT findings demonstrate that all formulated treatments exert beneficial effects on OA progression, with DHA-PRO@NMs@TSH showing the greatest potential in alleviating structural deterioration of OA joints.

Next, the extracellular matrix (ECM) serves as both the structural scaffold and functional regulatory network of joint tissues. Its homeostatic imbalance is a key pathological feature of OA [81]. We evaluated the protein expression of collagen II and aggrecan-major components of articular cartilage ECM-using immunofluorescence (IF) staining across groups. Higher fluorescence intensity indicated greater protein expression. Results demonstrated that DHA + CA, DHA-PRO, DHA-PRO@NMs, and DHA-PRO@NMs@TSH all elevated collagen II and aggrecan expression in OA cartilage, with DHA-PRO@NMs@TSH being the most effective. Concurrently, we assessed the expression of Matrix Metalloproteinase 13 (MMP13), a pivotal enzyme in ECM degradation. Lower fluorescence intensity corresponded to reduced ECM breakdown. All treatment groups effectively suppressed MMP13 expression, with DHA-PRO@NMs@TSH exhibiting the strongest inhibition. These findings collectively indicate that DHA-PRO@NMs@TSH provides the most potent protection against ECM degradation in OA articular cartilage (Fig. 7C).

Finally, we performed hematoxylin and eosin (H&E), toluidine blue (TB), and safranin O-fast green (SO-FG) staining in each group for histopathological examination [82]. H&E staining offers a clear view of articular cartilage and surrounding bone structure. TB staining specifically highlights proteoglycans in the cartilage matrix, and the intensity of blue staining is positively correlated with matrix integrity, providing an intuitive reflection of OA severity. SO-FG staining assesses metabolic imbalance in OA by simultaneously labeling proteoglycans (red) and collagen fibers (green). Reduced red intensity indicates proteoglycan loss, while abnormal green distribution reflects collagen network disruption. H&E staining results revealed that the Control group exhibited smooth articular cartilage with clear layers, absent subchondral bone sclerosis, and no significant osteophyte formation. In contrast, the MIA group showed rough cartilage surfaces with fissures, subchondral sclerosis, and prominent osteophytes. TB staining results demonstrated that the articular cartilage in the Control group exhibited intense, uniform blue, indicating abundant proteoglycans. the MIA group exhibited faint staining, suggesting severe proteoglycan loss. SO-FG staining results showed strong red safranin O binding in the Control group cartilage, indicating proteoglycan retention. After MIA induction, red staining was markedly reduced, reflecting proteoglycan loss across cartilage layers, while green collagen staining increased and appeared disorganized, highlighting an imbalance between proteoglycan and collagen. Treatments with DHA + CA, DHA-PRO, DHA-PRO@NMs, or DHA-PRO@NMs@TSH ameliorated joint damage to varying degrees, with the most pronounced improvement observed in the DHA-PRO@NMs@TSH group (Fig. 7D). These results strongly suggest that DHA-PRO@NMs@TSH effectively reverses histopathological changes in MIA-induced OA.

To quantify OA progression, we employed the modified Mankin scoring system: a histological tool evaluating cartilage structure, chondrocyte status, matrix staining, and tidemark integrity [83]. Higher scores indicate more severe OA. the MIA group received the highest Mankin score, while the DHA-PRO@NMs@TSH group received the lowest, indicating its potent inhibitory effect on OA progression histopathologically (Fig. S7G).

DHA-PRO@NMs@TSH attenuated MIA-induced OA via anti-PANoptosis and anti-inflammation

Having established the therapeutic effects of DHA-PRO@NMs@TSH in OA rats, we further investigated its underlying molecular mechanisms in vivo. First, we evaluated the expression of PANoptosis-related proteins, encompassing markers of apoptosis (BAX, Bcl-2), necroptosis (RIPK3 and p-RIPK3), and pyroptosis (NLRP3, pro-caspase-1, caspase-1, GSDMD-F, and GSDMD-N). WB analysis revealed significant upregulation of the three death pathways level following MIA modeling, which was markedly reversed following treatment with DHA + CA, DHA-PRO, DHA-PRO@NMs, or DHA-PRO@NMs@TSH. The most substantial reduction was observed in the DHA-PRO@NMs@TSH group (Fig. 8A–K), indicating its potent anti-PANoptosis activity. We next examined proteins related with inflammation, including ROS-related markers (NOX4 and SOD2), key components of the NF-κB pathway (p-IKKβ, IKKβ, p-IκBα, IκBα, and p-p65, p65), and critical inflammatory cytokines (IL-1β and TNF-α). All treatment groups reversed the inflammation-related changes induced by MIA, with DHA-PRO@NMs@TSH exhibiting the strongest inhibitory effect (Fig. 8L–T). Furthermore, ELISA confirmed elevated levels of IL-1β and TNF-α in the MIA group, which were significantly reduced following treatment. The lowest cytokine levels were observed in the DHA-PRO@NMs@TSH group (Fig. S8), supporting its superior anti-inflammation efficacy. In summary, these results demonstrate that DHA-PRO@NMs@TSH attenuates OA progression by concurrently suppressing PANoptosis and inflammation.

Fig. 8.

DHA-PRO@NMs@TSH attenuated MIA-induced OA via anti-PANoptosis and anti-inflammation. (A, B) Representative WB images and quantification analysis of apoptosis-related proteins (Bax and Bcl-2). (C-E) Representative WB images and quantification analysis of necroptosis-related proteins (RIPK3 and p-RIPK3). (F-K) Representative WB images and quantification analysis of pyroptosis-related proteins (NLRP3, Caspase-1, Pro-caspase-1, GSDMD-F, and GSDMD-N). (L-N) Representative WB images and quantification analysis of ROS metabolism-related proteins (NOX4 and SOD2). (O-T) Representative WB images and quantification analysis of NF-κB-related proteins (p-IKKβ, IKKβ, p-P65, P65, p-IκBα, and IκBα), followed by representative WB images and quantification analysis of inflammatory cytokines (IL-1β and TNF-α). Data are shown as mean ± SD (n = 3), and *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001, ns means not significant

In vivo biosafety of DHA-PRO@NMs@TSH

Finally, we evaluated the in vivo biosafety profile of DHA-PRO@NMs@TSH. Serum biochemical results indicated that alanine aminotransferase (ALT), aspartate aminotransferase (AST), blood urea nitrogen (BUN), creatinine (CRE), and lactate dehydrogenase (LDH) levels remained within normal ranges across all groups (Fig. 9A–E). Furthermore, H&E staining of major organs (heart, liver, spleen, lungs, and kidneys) revealed no evident histopathological abnormalities in any group (Fig. 9F). Together, these findings demonstrate that DHA-PRO@NMs@TSH exhibits good biocompatibility at the administered therapeutic dose, supporting its potential for future clinical translation.

Fig. 9.

In vivo biosafety evaluation of DHA-PRO@NMs@TSH. (A-E) The serum biochemical parameters of rats in each group: alanine aminotransferase (ALT), aspartate aminotransferase (AST), blood urea nitrogen (BUN), creatinine (CRE), and lactate dehydrogenase (LDH). (F) Representative H&E staining images of major organs (heart, liver, spleen, lung, and kidney) of rats in each group. Scale bar = 200 μm. Data are shown as mean ± SD (n = 3)

Conclusion

In conclusion, we successfully engineered a hierarchical and smart DHA delivery system (DHA-PRO@NMs@TSH) that integrates a ROS-responsive prodrug, nanomicelles encapsulation, and a thermosensitive hydrogel. This drug delivery system orchestrates stable DHA delivery, grants sustained intra-articular retention, and enables graded and ROS-triggered DHA release within the OA joint. Our comprehensive in vitro and in vivo studies demonstrate that DHA-PRO@NMs@TSH effectively attenuates OA progression by concurrently suppressing the newly identified PANoptosis pathway and key inflammatory signaling. The formulation exhibited excellent biosafety, underscoring its clinical translation potential. Beyond presenting a promising therapeutic strategy for OA, this work establishes a new treatment paradigm that leverages advanced drug delivery systems to simultaneously target multiple pathological drivers in complex diseases.

Author contributions

**Qingchen Li: ** Conceptualization, Data curation, Investigation, Methodology, Software, Writing – original draft. **Lixue Chen and Changyuan Wang: ** Resources, Validation, Visualization. **Yi Sun, YiLun Sun, and Kangjun Lu: ** Formal analysis, Data curation. **Lei Li, Huijun Sun, and Mozhen Liu: ** Funding acquisition, Project administration, Resources, Supervision, Writing – review and editing.

Funding

This work was supported in part by the National Natural Science Foundation of China (Grant No. 82073851), the Foundation from the educational department of Liaoning province (LJ232410161020), and the Dalian Medical Peak Program (Grant No. DF2023007). The authors gratefully acknowledge the financial support from these organizations.

Data availability

The data are available from the corresponding author upon reasonable request.

Declarations

Ethics approval and consent to participate

All animal experiments were conducted in compliance with the Guide for the Care and Use of Laboratory Animals and were approved by the Institutional Animal Ethics Committee of Dalian Medical University (License number: AEE21084).

Consent for publication

All the authors have approved the manuscript to publish.

Competing interests

The authors declare no competing interests.