Injectable sustainable andrographolide-releasing hydrogel for long-lasting alleviation of osteoarthritis and regulation of chondrocyte autophagy via PRKCA/EGFR

Abstract

Osteoarthritis is one of the most prevalent age-related joint diseases, with chondrocyte inflammation and autophagy dysregulation serving as pivotal pathogenesis factors. Andrographolide (AD), a phytochemical identified in Andrographis paniculata, exhibits anti-inflammatory properties and regulates autophagy to safeguard cells from damage. Nevertheless, the precise mechanism underlying the influence of AD on autophagy in osteoarthritis (OA) chondrocytes remains unelucidated. Concurrently, sustained efficacy of andrographolide typically necessitates prolonged administration, posing a challenge for its clinical application. We engineered an injectable 4-arm PEG-Mix-Hydrogel/PF system capable of encapsulating lipophilic drugs and achieving sustained release over a period of up to 24 days, substantially reducing the frequency of medication. Our findings indicate that andrographolide augments chondrocyte autophagy via the PRKCA/EGFR pathway and modulates chondrocyte inflammation as well as extracellular matrix degradation. Subsequent experimentation revealed that the injectable 4-arm PEG-Mix-Hydrogel/PF@AD (PHPF@AD) exhibited excellent biocompatibility with chondrocytes, possessed a rapid in-situ gelation time, and a single injection was sufficient to alleviate joint degeneration, abnormal gait, and weakened chondrocyte autophagy in OA mice, while ameliorating inflammation, matrix degradation, and apoptosis levels, and maintaining a certain degree of bone mass around the joints. In summary, this injectable hydrogel with spontaneous andrographolide release is anticipated to be a promising therapeutic modality for OA.

Graphical abstract

Highlights

• •Experimental findings demonstrate that andrographolide (AD) promotes chondrocyte autophagy in osteoarthritis (OA) through the PRKCA/EGFR pathway. Simultaneously, it mitigates joint tissue damage and concurrently reduces inflammation, extracellular matrix degradation, and apoptosis. These results establish AD as a potentially valuable natural compound for OA treatment, thereby elucidating the hitherto unknown mechanism of AD in alleviating OA by enhancing autophagy.
• •The hydrogel-micelle composite creates an ideal microenvironment for the slow-release of hydrophobic AD. It prevents burst-type drug release, avoiding poor efficacy and side-effects. The injectable PHPF@AD system can sustain drug release for over 24 days, solving the problem of frequent AD administration in prior studies. Additionally, this system exhibits excellent biocompatibility and biosafety with cartilage tissue, essential for cartilage-related therapies.
• •The injectable PHPF@AD system, with a single administration, can provide continuous treatment for OA mice induced by destabilization of the medial meniscus (DMM) surgery over a period of 24 days. It significantly alleviates the abnormal gait and joint tissue damage in OA mice and, to a certain extent, reduces bone loss.

1. Introduction

Osteoarthritis (OA) is a disease characterized by an imbalance in the normal synthesis and degradation of articular chondrocytes, subchondral bone, and extracellular matrix, influenced by mechanical and biological factors. This imbalance results in degenerative changes of articular cartilage, fibrosis, osteophyte formation, synovial membrane inflammation, and subchondral bone sclerosis, ultimately leading to the principal pathological manifestations such as deformity, loss of function, mobility disorders, and joint pain [[1], [2], [3]]. However, the molecular mechanisms underlying OA remain inadequately understood, necessitating the identification of effective therapeutic agents based on the pathological mechanisms of OA.

Cartilage is composed of a limited number of chondrocytes embedded within a substantial extracellular matrix. The regenerative capacity of cartilage predominantly relies on chondrocyte proliferation, with research on pathological changes in cartilage focusing on inflammation, apoptosis, and autophagy in chondrocytes [4]. Recently, autophagy has garnered considerable attention due to its protective role in chondrocytes during OA [5]. This highly conserved lysosomal degradation process, prevalent in eukaryotes, is crucial for maintaining cellular homeostasis [6]. Previous studies have demonstrated that autophagy can suppress chondrocyte apoptosis and degradation [7,8], with decreased levels of autophagy observed in OA cartilage [9,10]. Moreover, impaired autophagy in chondrocytes significantly compromises cellular function, contributing to joint senescence and dysfunction, highlighting autophagy as a promising target for OA treatment [11]. However, the specific mechanisms of autophagy in OA remain understudied, prompting our interest in exploring candidates related to autophagy in OA.

Andrographolide (AD) has attracted significant interest due to its anti-inflammatory, antineoplastic, antioxidant, and immunological properties [12,13]. As the primary active ingredient isolated from the herb Andrographis paniculata, AD is recognised as a multi-targeted anti-inflammatory agent [14]. Accumulating evidence indicates that AD promotes chondrocyte proliferation and suppresses apoptosis, thereby exerting anti-arthritic effects on OA [15] and serving as a therapeutic agent to reduce synovial inflammation [16]. Notably, AD protects mouse astrocytes from injury by regulating and promoting autophagy, alleviating colitis in murine models [17,18]. However, whether AD can regulate OA progression via autophagy remains unknown. Based on these observations, we hypothesize that AD alleviates OA by inducing autophagy in chondrocytes.

Moreover, protein kinase C alpha (PRKCA), a member of the protein kinase C family, participates in multiple cellular processes, including differentiation, apoptosis, adhesion, migration, tumorigenesis, and inflammation [19]. Recent studies have indicated that PRKCA is one of the key genes aberrantly upregulated in osteoarthritis (OA), suggesting its potential as a therapeutic target for OA [20]. Additionally, PRKCA has been demonstrated to downregulate the expression of epidermal growth factor receptor (EGFR), and elevated EGFR levels have been shown to enhance chondrocyte autophagy in OA [21,22]. Consequently, we hypothesize that targeted downregulation of PRKCA via AD could alleviate OA by upregulating EGFR expression and inducing chondrocyte autophagy.

In previous investigations on the application of AD in OA treatment, continuous administration of high doses is required to maximize efficacy, attributable to the low bioavailability of AD due to its weak hydrophilicity and strong lipophilicity [[23], [24], [25], [26]]. Given that OA often necessitates long-term treatment, the high frequency of administration significantly increases the treatment burden, and the choice of joint injection therapy also raises the risk of joint infection, which is not conducive to further clinical research on AD-based OA treatment. In recent years, advancements in biomedical drug delivery materials have led to methods for enhancing the bioavailability of ester drugs, including the use of innovative nanoparticles carrying AD for drug administration, with promising experimental outcomes [27]. However, reducing the frequency of drug administration while maintaining therapeutic efficacy for OA remains a challenge. Conventional polymer-drug conjugation systems may encounter solubility issues due to the incorporation of hydrophobic drugs, whereas micelles can effectively encapsulate hydrophobic drugs and slow drug release, thereby significantly improving the bioavailability of lipophilic drugs. Micelles’ highly functional structure makes them well-suited for drug delivery systems [28]. Therefore, we selected micellar polymers to construct a conjugation system with AD. It was also observed that the micelle could release AD slowly under physiological conditions without exhibiting toxicity toward chondrocytes. To avoid potential cartilage wear and synovial irritation associated with simple micellar injections, a material carrier that enhances drug delivery efficiency and protects articular cartilage is necessary.

Among injectable biomaterials, hydrogels have demonstrated potential for biomedical applications and are frequently utilized in sustained drug release systems to achieve slow drug release. Moreover, the gelatinous properties of injectable hydrogels within joint cavities can reduce articular cartilage wear and lubricate joints, making hydrogels widely studied in orthopedic drug delivery [29]. Li et al. utilized injectable hydrogel to modulate the PI3K/AKT signaling pathway via continuous teriparatide release to mitigate OA progression [30]. Traditional injectable hydrogels may pose challenges such as difficult gelation transformation, prolonged gel formation time, loss of loaded active ingredients, and high viscosity [31]. Consequently, we prepared a non-irritating, dual mixture with in-situ chemical crosslinking and rapid gelation for injectable polyethylene glycol (PEG) hydrogel carrying AD-coated Pluronic F127 (PF) micelles (4-arm PEG-Mix-Hydrogel/PF@AD, PHPF@AD). This injectable hydrogel can gel rapidly in the human environment without additional stimulation and has been demonstrated not to affect mouse movement. The hydrogel in this study possesses non-toxicity, good biocompatibility, strong drug loading performance, mild and controllable preparation conditions, and considerable potential for clinical translation.

In this study, we proposed the injectable PHPF@AD system as a long-acting therapeutic solution for osteoarthritis (OA). This system ingeniously combines the benefits of andrographolide (AD), micelles, and polymer hydrogels, presenting an innovative strategy for OA treatment. This long-acting, sustained-release drug delivery system holds great promise in alleviating the treatment burden on clinical OA patients. By doing so, it can effectively mitigate the pain endured by patients during the treatment process. Moreover, it has the potential to reduce the risk of joint infection, which is of substantial significance for improving the quality of life of individuals suffering from OA.

2. Materials and methods

2.1. Cell treatment

Human chondrocytes (American Type Culture Collection, USA) were cultured in Dulbecco's Modified Eagle's Medium (DMEM) supplemented with 10 % foetal bovine serum (FBS; S9030; Solarbio, Beijing, China) and 1 % penicillin-streptomycin (15140122; Thermo Fisher Scientific, USA). The culture conditions were maintained at 5 % CO₂ and 37 °C. Chondrocytes at passage 2 or 3 were harvested for subsequent analyses.

To overexpress PRKCA, the full-length sequence of PRKCA was cloned into the pcDNA 3.1 vector (V79520; Thermo Fisher Scientific, USA) to construct the overexpression plasmid, with an empty vector serving as a control. For transfection, chondrocytes (1 × 10⁴ cells/well) were seeded in 96-well plates to achieve approximately 90 % confluence, followed by transfection with the constructed plasmids using Lipofectamine 3000 transfection reagent (L3000015; Thermo Fisher Scientific, USA). The plasmids and Lipofectamine 3000 reagent were diluted in Opti-MEM media and incubated for 10 min at 37 °C to form gene-lipid complexes. Finally, the chondrocytes were incubated for 48 h at 37 °C.

2.2. IL-1β-induced cellular OA model establishment and AD treatment

To simulate in vitro OA models, passaged chondrocytes (2 × 10⁵ cells/well) were seeded in six-well plates to achieve 90 % confluency. Following an overnight starvation period, the chondrocytes were stimulated with 1 ng/mL interleukin 1 beta (IL-1β) (SRP6169; Sigma Aldrich, St. Louis, MO, USA) dissolved in phosphate-buffered saline (PBS, ST476; Beyotime, Shanghai, China) and diluted in DMEM for 24 h. Subsequently, adenosine was dissolved in dimethyl sulfoxide (DMSO; W387520; Sigma Aldrich) and diluted in DMEM to final concentrations of 0, 1, 5, and 10 μM. The chondrocytes were then treated with AD for 2 h [32].

2.3. Network pharmacology and molecular docking analysis

Network pharmacology was employed to construct the chemical composition - target - pathway network and the interaction network, aiming to predict the potential targets of andrographolide (AD) for the treatment of osteoarthritis (OA).

First, the confirmation number of andrographolide was retrieved from the PubChem database (https://pubchem.ncbi.nlm.nih.gov/). Subsequently, target prediction was performed using the SwissTargetPrediction (http://www.swisstargetprediction.ch/), Super-Mr PRED (https://prediction.charite.de), and SEA Search Server (https://sea.bkslab.org/) databases.

For disease target retrieval, the GeneCards (https://www.genecards.org/) and OMIM (Online Mendelian Inheritance in Man) database (https://omim.org/) were utilized, with the search term “Osteoarthritis”. After setting the species, the genes of disease targets were obtained. To eliminate duplicate values, the intersection of disease targets retrieved from the databases was determined.

The selected compound targets and disease targets were imported into Venny 2.1, a software for generating Venn diagrams. Based on drugs, compounds, intersection targets, and enrichment pathways, the compound-disease-target network was constructed using Cytoscape 3.7.2 software.

The common targets of drug - related diseases were input into the STRING database (https://string - db.org/cgi/input.pl) to construct the protein - protein interaction (PPI) network, with the biological species set as “Homo sapiens”. A minimum required interaction score of 0.40 was set to obtain the PPI network.

The targets of drugs and diseases were imported into DAVID Bioinformatics Resources 6.8 (https://david.ncifcrf.gov/home.jsp) for gene ontology (GO) enrichment analysis, including biological processes (BP), cellular components (CC), and molecular functions (MF). The enrichment conditions were set as P-value cutoff = 0.05 and Q-value cutoff = 0.05, with the remaining parameters defaulting to the original settings. The top 10 significantly enriched entries were selected. According to the P - value, Q-value, and the number of genes in each entry, the GO histogram analysis was plotted using the Microletter platform (http://www.bioinformatics.com.cn/).

The common targets of drug - related diseases were also subjected to Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis using DAVID Bioinformatics Resources 6.8. Entries with a corrected P-value <0.05 were screened.

The compound name, molecular weight, and 3D structure of the active ingredient were determined from the PubChem database, and the 3D structure of the active ingredient was downloaded from the RCSB PDB database (http://www.rcsb.org/). Then, AutoDock software was used to prepare the ligands and proteins required for molecular docking. For the target protein, its crystal structure was processed by removing water molecules, adding hydrogens, modifying amino acids, optimizing energy, and adjusting force - field parameters to match the low - energy conformation of the ligand structure.

Finally, molecular docking was performed between the key target and the compound. The affinity (kcal/mol) value was used to represent the binding ability between them. A lower affinity value indicated a more stable binding between the ligand and the receptor. The docking results were analyzed and visualized using Discovery Studio software.

2.4. Immunofluorescence

The chondrocytes were fixed with 4 % paraformaldehyde (P0099; Beyotime, China) for 20 min at room temperature and subsequently permeabilized with 0.1 % Triton X-100 (93443; Sigma-Aldrich, USA) in phosphate-buffered saline (PBS; P1022, Solarbio, China) for 10 min. For antigen retrieval, the chondrocytes were heated in a stainless-steel pan with ethylenediaminetetraacetic acid (EDTA) retrieval solution (P0085, Beyotime, China). After blocking with normal goat serum (SL038, Solarbio, China) diluted in PBS for 30 min, the chondrocytes were incubated with Alexa Fluor 488-labeled Microtubule-associated protein 1A/1B-light chain 3 (LC3) antibody (ab225382, Abcam, Cambridge, UK) for 1 h at 37 °C. The chondrocytes were then incubated with 4’,6-diamidino-2-phenylindole (DAPI; C0065, Solarbio, China) for 15 min. Finally, the cells were mounted with neutral gum (96949-21-2, Solarbio, China), and fluorescence signals were detected using a CX31 microscope (Olympus, Tokyo, Japan) at × 200 magnification.

2.5. Bioinformatics

Potential targets of AD were predicted using SwissTarget Prediction (http://swisstargetprediction.ch/).

2.6. Quantitative real-time reverse transcription polymerase chain reaction

Quantitative real-time reverse transcription polymerase chain reaction (qRT-PCR) was used to analyse cartilage tissues and cells. Total RNA was extracted using the Trizol Reagent (15596026, Thermo Fisher Scientific, USA). After adding 1 mL Trizol to 0.02 g of tissue or 500 μL of cells, the samples were lysed for 3 min at room temperature. A Revert Aid First Strand cDNA Synthesis Kit (CW2569; Beijing Comwin Biotech, Beijing, China) was used for cDNA synthesis. Primer sequences were obtained from Sangon Biotech (Supplementary Table 1). DNA amplification and detection were performed with a fluorescent quantitative PCR instrument (PIKOREAL96, Thermo Fisher, USA), and the relative mRNA levels were assessed using the 2^-△△CT method with GAPDH and β-actin as reference genes.

2.7. Enzyme-linked immunosorbent assay

Chondrocytes transfected with or without overexpressed PRKCA were induced with IL-1β and treated with AD. The levels of IL-6, IL-1β, and tumor necrosis factor- α (TNF-α) in the culture supernatant were measured using the IL-6 enzyme linked immunosorbent assay (ELISA) kit (QK206, Biotechne, Minneapolis, MN, USA), IL-1β ELISA kit (QK201, Biotechne, USA), and TNF-α ELISA kit (QK210, Biotechne, USA), respectively. Following the manufacturer's instructions, 100 μL of cell culture supernatant was added to each well of a 96-well plate and incubated for 2 h at room temperature. Subsequently, 100 μL of diluted Detection Antibody in Reagent Diluent was added, and a new adhesive strip was applied to cover the wells, followed by another 2-h incubation. Subsequently, 100 μL of working dilution of Streptavidin-HRP was added, and the wells were covered again for 20 min. Afterward, 100 μL of Substrate Solution was added, followed by a 20-min incubation in the dark. Finally, after washing with 100 μL Stop Solution, the optical density (OD) at 450 nm was measured using an Infinite M200 microplate reader (30190085, Tecan, Männedorf, Switzerland).

2.8. MTT cell proliferation-toxicity test

Chondrocytes transfected with or without overexpressed PRKCA were induced with IL-1β, and cell viability was assessed using an MTT assay kit (M1020, Solarbio, China). According to the manufacturer's guidelines, 5 × 10³ chondrocytes were maintained in 96-well plates and incubated for 24 h in a 5 % CO₂ atmosphere at 37 °C. Following treatment with 0, 1, 5, and 10 μM AD, the cells were further incubated for 24 h. Then, 10 μL of MTT solution was added to each well, and the cells were incubated for an additional 4 h at 37 °C. Afterward, 110 μL of Formazan dissolving solution was added to each well, and the OD at 490 nm was recorded using an Infinite M200 microplate reader.

2.9. Flow cytometry

Chondrocytes transfected with or without overexpressed PRKCA were induced with IL-1β and treated with AD. Apoptosis was measured using an Annexin V-Fluorescein Isothiocyanate (V-FITC) apoptosis detection kit (P-CA-201; Procell, Wuhan, China). Following the manufacturer's guidelines, 5 × 10⁵ chondrocytes were centrifuged at 300×g for 5 min and resuspended in 500 μL of Annexin V Binding Buffer. Annexin V-FITC (5 μL) and Propidium Iodide (5 μL) were added, and the cells were incubated in dark conditions for 20 min at 37 °C. A flow cytometer (1026, Beamdiag, Changzhou, China) was used for immediate detection, and the data were analyzed using Cell Quest software (BD Biosciences, San Diego, CA, USA).

2.10. Western blotting

Western blotting (WB) was performed on both cells and cartilage tissues. For cells, 200 μL of RIPA lysis buffer was added to cells washed with ice-cold PBS. The cells were scraped off using a cell scraper, sonicated for 1.5 min, and lysed on ice for 10 min. After centrifugation at 12,000 rpm for 15 min, the supernatants were collected. For cartilage tissues, 300 μL of RIPA lysis buffer was added to 0.025 g of tissue, which was ground until no tissue fragments were visible. The lysates were incubated on ice for 10 min. After centrifugation at 12,000 rpm for 15 min, the supernatants were collected.

Protein concentrations were measured using the Bicinchoninic Acid (BCA) protein assay kit. Denatured proteins were loaded into each well, and electrophoresis was terminated once the gel reached the bottom. The transfer buffer was prepared using a gel, filter paper, and nitrocellulose membrane, and the membrane was transferred at a constant current. phosphate buffered saline tween 20 (PBST) containing 5 % skim milk was used to block the membrane for 1.5 h. Subsequently, the primary and secondary antibodies were added. Protein bands were visualized using a chemiluminescence imaging system, with β-actin as an internal reference. Details of the antibodies used are listed in Supplementary Table 2.

2.11. Synthesis of PF127@AD micelles

PF127 was procured from EFL (Engineering For Life, Suzhou, China), and PF127@AD micelles were synthesized via a one-step solid dispersion method, as detailed in reference [33]. In brief, the synthesis process involved dissolving 2 % wt of AD and the PF127 polymer in dehydrated alcohol under mild stirring conditions. Subsequently, the solution was subjected to condensation using a rotary evaporator at a temperature of 60 °C and the vacuum pressure is 200 mbar. The co-evaporated product was then dissolved in double-distilled water (ddH2O) at 40 °C, to facilitate self-assembly into micelles. Post self-assembly, the material was subjected to freeze-drying to yield the PF127@AD powder. The feed ratio of AD to PF127 polymer was maintained at 2:98.

2.12. Synthesis of hydrogels

4-arm PEG-NHS (Mw = 10 kDa) and 4-arm PEG-NH₂ (Mw = 10 kDa) were acquired from EFL. For the preparation of 4-arm PEG-Mix hydrogel (PH), the 4-arm PE-NHS and 4-arm PEG-NH₂ polymers were separately dissolved in PBS in two separate sample bottles at a concentration of 10 wt%, with an equal volume ratio of 1:1. The injectable PH hydrogel was subsequently prepared by mixing the two components using a dual syringe. To synthesize 4-arm PEG-Mix/PF@AD (PHPF@AD), the PF127@AD was first dissolved in distilled water at a concentration of 40 wt%. Subsequently, PF127@AD was mixed with 4-arm PEG NH₂ in advance at a predetermined volume ratio, yielding a final concentration of PF127@AD and 4-arm PEG NH₂ at 10 wt%, then mixing the two solutions using a dual syringe, with an equal volume ratio of 1:1.

2.13. Macroscopic observation

The overall appearance of the hydrogel was visually inspected. Rhodamine G was added to 4-arm PEG-NH2, and Sodium Fluorescein was added to 4-arm PEG-NHS solution. The precursor solutions were mixed in equal volumes within a transparent vial. The vial was inverted to observe the flow behavior of the mixture. The gel time was marked when the solution ceased to flow, indicating the formation of a stable hydrogel matrix.

2.14. Rheological characterization

Rheological tests of the PH and PHPF@AD hydrogels were conducted on a rotational rheometer (DHR-2, TA, USA) in oscillatory mode at 26 °C. the hydrogels were in the shape of a disk, with a diameter of approximately 10 mm and a height of 5 mm. For the frequency scanning, the formed disk hydrogels were placed on the plate and the frequency was scanned from 0.1 to 10 Hz with a strain of 1 %. The shear-thinning properties of the hydrogels were characterized by measuring the viscosity under a frequency sweep mode with a shear rate ranging from 0.1 to 1000 s-1, maintaining a frequency of 1 Hz and a strain of 1 %.

2.15. Microstructural morphology

Field emission scanning electron microscopy (FE - SEM, Zeiss, Germany) was employed to observe the microstructure of the PH and PHPF@AD hydrogels after gold sputtering. The accelerating voltage was 12 kV, the working distance was 8 mm, and the ambient temperature was 26 °C. The cross - sectional view of the dried samples was meticulously examined and documented through high - resolution imaging.

2.16. Mechanical properties

To evaluate the compressive mechanical properties of the hydrogels, a ZwickRoell universal testing machine was employed at 26 °C. Disk samples with a diameter of approximately 10 mm and a height of 5 mm were prepared in their equilibrium swelling state in PBS. These disks were then placed between the machine's jaws to determine the compressive moduli at 80 % deformation.

2.17. Swelling capacity

To assess the swelling capacity, the hydrogel was prepared into a cylindrical shape with dimensions of 5 × 10 mm, ensuring an equal volume. The hydrogel samples were initially weighed and recorded as W0. Subsequently, the samples were immersed in PBS at room temperature. At predetermined time intervals, the hydrogels were retrieved from the PBS solution, and excess surface moisture was carefully removed using filter paper. The weight of each swollen sample was then measured and recorded as Wt. The swelling degree was calculated using the formula: Swelling ratio = (Wt - W0)/W0 × 100 %.

2.18. In vitro degradation

The in vitro degradation of the hydrogels was conducted by incubating the samples in a simulated physiological environment-PBS at 37 °C with continuous shaking. The initial dry weight of the hydrogel (W0) was recorded before degradation, and the weight after degradation (Wt) was measured at each designated time point. The percentage of weight remaining was calculated using the following equation: Weight remaining (%) = (Wt/W0) × 100 %.

2.19. Drug release in vitro

The drug release characteristics of the PH and PHPF@AD hydrogels were evaluated in PBS for AD. For the PH hydrogel, AD was directly mixed into the premix solution, ensuring a consistent AD concentration with that in PHPF@AD. The fabrication method for PHPH@AD was employed as previously described. The drug released from the hydrogels was analyzed using a UV–Vis spectrophotometer at a wavelength of 225 nm (AD).

2.20. Live/dead staining and Cell Counting Kit - 8 (CCK8) assays

Hydrogel extracts were meticulously prepared. Chondrocytes were then seeded into 96 - well plates at an appropriate density and cultured separately for 1, 3, and 5 days. As a control group, chondrocytes were cultured in DMEM, supplemented with 10 % FBS.

For the quantification of cell viability, the Cell Counting Kit - 8 (CCK8) assay was employed. The optical density (OD) value was measured at a wavelength of 450 nm using a microplate reader. This measurement enabled the determination of the number of viable cells based on the reduction of the tetrazolium salt in the CCK8 reagent by mitochondrial dehydrogenases in living cells.

To further assess the viability of chondrocytes, Calcein - AM/PI double staining was carried out. Chondrocytes were seeded into 24 - well plates and incubated for 1, 3, and 5 days. Post - incubation, the cells were gently washed twice with pre - cooled phosphate - buffered saline (PBS). Subsequently, a pre - prepared Calcein - AM/PI staining solution was added to each well, and the cells were incubated for 15 min in the dark to prevent photobleaching. After the incubation period, the cells were washed again with PBS to remove any unbound staining reagents. Finally, the stained cells were visualized and photographed using a fluorescence microscope.

2.21. Establishment of OA rat models

Male C57BL/6J rats (8 weeks old) were housed under a 12-h light/dark cycle and allowed to acclimate for 7 d. Rats were randomly divided into two groups. The mice were anesthetized with isoflurane. Surgical incisions were made in the right knee capsules of rats in the control group (control, n = 6). In the OA group (n = 18), destabilization of the medial meniscus (DMM) was performed on the right knee. All rats were anesthetized with 2 % isoflurane during surgery. The OA mice were divided into 3 groups four weeks after operation. In the OA + AD group (OA + AD, n = 6), OA rats received a single articular injections of 100 μL AD. In the OA + PH group (OA + PH, n = 6), OA rats received a single articular injections of 100 μL PEG-Hydrogel. In the OA + PHPF@AD group (OA + PHPF@AD, n = 6), OA rats received a single articular injections of 100 μL PHPF@AD. In order to further explore the differences between PHPF@AD and traditional intraperitoneal injection of AD in the treatment of OA, we created additional OA mice n = 18. the AP low dose group with AP at 5 mg/kg/day (n = 6), the AP medium dose group with AP at 15 mg/kg/day (n = 6), and the AP high dose group with AP at 30 mg/kg/day (n = 6). The drugs were administered by intraperitoneal injection for four weeks [23]. All mice were euthanized and knee joints were collected 8 weeks after surgery for further investigation.

After modeling, all animals underwent gait tests to evaluate behavioral changes. Subsequently, the mice were euthanized, and knee joint samples were collected for Micro CT examination to assess the degree of knee joint degeneration. Knee joint samples were also collected for subsequent analysis. A portion of the obtained knee joints was fixed in 4 % paraformaldehyde for 48 h, and then decalcified in 10 % ethylenediaminetetraacetic acid (EDTA, pH 7.4) on a shaker (37 °C, 80 rpm) for 28 days. After gradient dehydration, the knee joints were embedded in paraffin wax and cut into 4 -μm - thick sections along the sagittal plane for histological staining, including hematoxylin - eosin (HE) staining, as well as safranin O and Fast Green (S - O) staining. After applying the modified Mankin and Synovitis scores, the remaining cartilage tissues were analyzed using Western blotting (WB) and quantitative real - time polymerase chain reaction (qRT - PCR).

2.22. In-vivo imaging system

Nine 8-week-old male C57BL/6 mice were randomly allocated into three experimental groups, each consisting of three animals. The mice were anesthetized using isoflurane. Subsequently, 1,1′-dioctadecyl- 3,3,3′,3′-tetramethylindotricarbocyanine iodide (DIR; sourced from Merck, Germany) was employed to label the AD-PBS (AD) solution, PEG-Hydrogel-AD (PH) solution, and PHPF@AD solution. These labeled solutions were then individually injected into the knee joints of the mice. The in - vivo localization of the injected substances within the mice was continuously monitored by an in - vivo imaging system (IVIS, Lumina Series III, manufactured by PerkinElmer, USA). At the pre - determined time point, under an excitation wavelength of 657 nm and an emission wavelength of 747 nm, the mice were imaged using the IVIS. The obtained images were analyzed to precisely determine the retention of the injected substances within the joint regions.

2.23. Micro Computed Tomography

Knee degeneration and bone changes were evaluated via a Micro Computed Tomography (Micro CT) scan on an OA model using a Quartus GX (PerkinElmer, Waltham, MA, USA). The scanning parameters were set as follows: a resolution of 9 mm, an aluminum filter of 1 mm thickness, a voltage of 70 kV, and a current of 120 mA. A 3D image was reconstructed using NRecon software. The entire subchondral bone of the tibial plateau was selected for analysis. The three-dimensional structural parameters analyzed included the bone volume/total volume (BV/TV) and bone mineral density (BMD). The samples were fixed in 4 % paraformaldehyde for 48 h and subsequently placed in the scanning tube of the machine for imaging.

2.24. Gait analysis

The paws of mice were inked by an examiner blinded to the surgical procedures. Immediately after the ink was applied, the mice were allowed to run on a 60 cm long, 6 cm wide Plexiglas track with white paper on the bottom. A dark chamber at the end of the track was used to entice the mice. After completion of the test, the paper was scanned at 300 dpi. The distance covered by the same paw between the two steps was designated the stride, whereas the horizontal distance between the left and right paws was defined as the stride width. Toe spread was also recorded. This method provides a simple yet effective measure of gait abnormalities in murine models [34].

2.25. Statistical analyses

In this study, we used GraphPad Prism 8.0 statistical software (GraphPad Software Inc., San Diego, CA, USA) to analyse all data expressed as mean ± standard deviation. Multiple groups were compared using a one-way analysis of variance (ANOVA). Statistical significance was showed as ∗p < 0.05; ∗∗p < 0.01; ∗∗∗p < 0.001; ∗∗∗∗p < 0.0001;^nsp ≥ 0.05.

3. Results

3.1. AD treatment neutralized the effects of IL-1β on inflammation, viability, matrix degradation and apoptosis in chondrocytes

We initially utilized IL-1β-induced osteoarthritis (OA) models to examine the effects of AD. The results from the ELISA indicated that the AD-0 μM group exhibited significantly elevated levels of IL-6, IL-1β, and tumor necrosis factor-α (TNF-α) in the cell culture supernatant compared to the control (CON) group (Fig. 1A–C, p < 0.001). Conversely, these inflammatory markers were reduced following treatment with 1, 5, or 10 μM AD in comparison to the AD-0 μM group (Fig. 1A–C, p < 0.001). Additionally, the MTT assay (Fig. 1D) and flow cytometry (Fig. 1E–F) revealed suppressed cell viability and enhanced apoptosis in chondrocytes from the AD-0 μM group (p < 0.001), while treatment with 1, 5, or 10 μM AD significantly increased cell viability and decreased apoptosis (p < 0.05). Moreover, according to the results of Western blot analysis, the AD-0 μM group showed higher expressions of Matrix Metalloproteinase-1 (MMP-1) and MMP-13 in chondrocytes compared to the CON group (Fig. 1G–I, p < 0.001). These markers were down-regulated following treatment with 1, 5, or 10 μM AD (Fig. 1G–I, p < 0.01).

Fig. 1.

Effects of AD on inflammation, viability, apoptosis, and expressions of MMP-1 and MMP-13 in IL-1β-induced cellular OA models.

(A-C) Chondrocytes were collected and induced with 1 ng/ml IL-1β for 24 h to establish IL-1β-induced OA models in vitro. Subsequently, the cells were treated with 0, 1, 5, or 10 μM AD for 2 h. The levels of IL-6, IL-1β, and TNF-α in the cell culture supernatants were measured using ELISA. (D) Chondrocyte viability was assessed using the MTT assay. (E-F) Apoptosis of chondrocytes was evaluated via flow cytometry. (G-I) The expressions of MMP-1 and MMP-13 in chondrocytes were assessed through Western blot analysis, with GAPDH serving as a loading control.

∗p < 0.05; ∗∗p < 0.01; ∗∗∗p < 0.001; ∗∗∗∗p < 0.0001;^nsp ≥ 0.05.

3.2. AD treatment neutralized the effects of IL-1β on the expression of autophagy in chondrocytes

LC3 is a biomarker of autophagy [35,36]. We further analyzed the changes in chondrocyte autophagy by assessing the number of cells with punctate LC3 using immunofluorescence staining (green indicates LC3). The AD-0 μM group exhibited a decrease in the number of LC3-positive cells compared to the CON group, while treatment with 1, 5, or 10 μM AD led to an increase in LC3-positive cells (Fig. 2A), confirming that AD treatment could promote IL-1β-suppressed autophagy in chondrocytes. Western blotting revealed that Beclin-1 expression and the ratio of LC3-II/LC3-I were reduced in the AD-0 μM group (Fig. 2B–D, p < 0.001), whereas these indicators were elevated following treatment with 1, 5, or 10 μM AD (Fig. 2B–D, p < 0.01).

Fig. 2.

Effects of AD on Chondrocyte autophagy in IL-1β-induced cellular OA models.

(A) Chondrocytes were induced with 1 ng/ml IL-1β for 24 h, followed by treatment with 0, 1, 5, or 10 μM AD for 2 h. The number of cells exhibiting punctate LC3 was analyzed through immunofluorescence staining (magnification: × 200, scale bar = 100 μm). (B-D) Western blot analysis was employed to evaluate the expressions of Beclin-1, LC3-II, and LC3-I in chondrocytes. GAPDH served as a loading control.

∗p < 0.05; ∗∗p < 0.01; ∗∗∗p < 0.001; ∗∗∗∗p < 0.0001;^nsp ≥ 0.05.

3.3. Analysis results based on network pharmacology

The selected compound targets and disease targets were imported into Venny 2.1, a software tool designed for generating Venn diagrams. As a result, 90 common targets were identified, and these were designated as the predicted targets of the compound's action on the disease for subsequent analysis (Fig. 3A) .

Fig. 3.

Analysis results based on network pharmacology.

(A) Visual analysis of compound (AD) - disease (OA) common target intersection. (B) Compound (AD) - Disease (OA) - target network. (C) GO biological process analysis. (D) KEGG pathway enrichment analysis. (E) Molecular docking model diagram (AD - PRKCA).

Cytoscape 3.7.2 software was utilized to construct the compound - disease - target network. In this network, the orange diamonds represent the compounds, the red diamonds symbolize the diseases, and the cyan circles denote the intersection targets (Fig. 3B).

Through comprehensive analysis, a total of 524 Gene Ontology (GO) terms were found to be enriched. Among them, 51 were related to cellular components (CC), with the majority being localized in the plasma membrane, cytoplasm, and karyoplasm. There were 354 biological processes (BP), which primarily involved the positive regulation of gene expression, chromatin remodeling, and the cellular response to hydrogen peroxide. Additionally, 119 molecular functions (MF) were identified, mainly manifested in the regulation of protein kinase activity, histone H3Y41 kinase activity, and protein lysine deacetylase activity (Fig. 3C).

Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis revealed significant enrichment in several key pathways, including EGFR tyrosine kinase inhibitor resistance, the Vascular Endothelial Growth Factor (VEGF) signaling pathway, growth hormone synthesis, secretion and action, the prolactin signaling pathway, and glioma - related pathways (Fig. 3D).

Based on this analysis, a total of 44 core targets with values exceeding the mean were identified (Supplementary Fig. S1, Supplementary Table 3, Supplementary Table 4). The binding energy between the compound and the targets is presented in Supplementary Fig. S3. It is commonly accepted that a binding energy value less than −4.25 kcal/mol indicates a certain degree of binding activity between the compound and the target, less than −5.0 kcal/mol suggests good binding activity, and less than −7.0 kcal/mol implies strong binding activity. These findings suggest that PRKCA may serve as a crucial target.

Molecular docking results demonstrated that AD forms conventional hydrogen bonds with LYS232 and ARG238 of the protein, an alkyl interaction with LEU200, and a Pi - alkyl interaction with PHE229 (Fig. 3E).

3.4. Overexpression of PRKCA increased inflammation and inhibited viability, autophagy, and EGFR expression and promoted apoptosis in IL-1β-induced chondrocytes, which was reversed by AD

Next, we analyzed the potential targets of AD using SwissTarget Prediction, which combined with target prediction results from network pharmacology (Supplementary Fig. S1, Supplementary Fig. S2, Supplementary Table 3), indicated that PRKCA was the most likely target gene for AD. Subsequent qRT-PCR results showed that the AD-0 μM group had increased PRKCA expression in chondrocytes compared to the CON group (Fig. 4A, p < 0.001). In contrast, treatment with 1, 5, or 10 μM AD resulted in down-regulation of PRKCA in IL-1β-induced chondrocytes (Fig. 4A, p < 0.001). Based on these findings, an overexpression plasmid for PRKCA (designated OE-PRKCA in Fig. 4B) was transfected into chondrocytes, as confirmed by the elevated PRKCA mRNA levels (Fig. 4B, p < 0.001). Subsequently, the evaluation of IL-6, IL-1β, and TNF-α levels in the cell culture supernatant through ELISA revealed that these levels were significantly increased in IL-1β-induced chondrocytes transfected with overexpressed PRKCA compared to those transfected with the negative control (NC) (Fig. 4C–E, p < 0.001). Additionally, IL-6, IL-1β, and TNF-α levels in IL-1β-induced chondrocytes treated with AD (10 μM) were lower compared to the CON group (Fig. 4C–E, p < 0.001). Importantly, overexpression of PRKCA reversed the effects of AD treatment on IL-6, IL-1β, and TNF-α levels in IL-1β-induced chondrocytes (Fig. 4C–E, p < 0.001), while AD treatment also mitigated the effects of overexpressed PRKCA on these inflammatory markers (Fig. 4C–E, p < 0.001). The results from the MTT assay (Fig. 4F), flow cytometry (Fig. 4G–H), and western blotting (Fig. 4I–J) indicated that IL-1β-induced chondrocytes transfected with overexpressed PRKCA exhibited suppressed viability, enhanced apoptosis, down-regulated expression of Beclin-1 and EGFR, and a reduced LC3-II/LC3-I ratio (p < 0.05) compared to those transfected with the NC. Conversely, treatment with 10 μM AD in IL-1β-induced chondrocytes led to increased viability, decreased apoptosis, upregulated Beclin-1 and EGFR expression, and an elevated LC3-II/LC3-I ratio (Fig. 4F–M, p < 0.001). Notably, overexpression of PRKCA was able to reverse the effects of AD treatment on these indicators in IL-1β-induced chondrocytes (Fig. 4F–M, p < 0.001), while AD treatment mitigated the effects of PRKCA overexpression (Fig. 4F–M, p < 0.05).

Fig. 4.

Effects of overexpressed PRKCA on inflammation, viability, apoptosis and EGFR expression in IL-1β-induced OA models.

(A) After establishing IL-1β-induced OA models and AD treatment (0, 1, 5, or 10 μM AD), qRT-PCR was utilized to measure PRKCA expression in IL-1β-induced chondrocytes. GAPDH served as a loading control. (B) Overexpression of PRKCA was achieved by transfecting it into chondrocytes, and its expression was evaluated via qRT-PCR. GAPDH was used as a loading control. (C-E) Following transfection, levels of IL-6, IL-1β, and TNF-α in the cell culture supernatant were analyzed using ELISA after establishing IL-1β-induced OA models and/or AD treatment (10 μM). (F) Chondrocyte viability was assessed using the MTT assay. (G-H) Flow cytometry was applied to evaluate chondrocyte apoptosis. (I-M) Following the transfection of overexpressed PRKCA, the expressions of Beclin-1, EGFR, LC3-II, and LC3-I were evaluated through Western blot analysis. GAPDH served as a loading control.

∗p < 0.05; ∗∗p < 0.01; ∗∗∗p < 0.001; ∗∗∗∗p < 0.0001;^nsp ≥ 0.05.

3.5. Fabrication and characterization of hydrogels

Firstly, we induced a chemical reaction between the two polymers, 4-arm PEG-NH₂ and 4-arm PEG-NHS, by mixing them. This reaction led to the formation of stable amide bonds, ultimately resulting in the generation of a 4-arm PEG hydrogel (Supplementary Fig. S4). Subsequently, as depicted in Fig. 5A, we employed the tilted vial method to comprehensively characterize the gelation process of the two hydrogel systems. In comparison with the PH hydrogel, the PHPF@AD hydrogel demonstrated a significantly shorter gelation time. This phenomenon can be attributed to the promotive effect of PF127 on hydrogel cross - linking. The reduced gelation time offers distinct advantages for in - situ injection. It not only effectively minimizes the risk of hydrogel leakage during the injection process but also ensures thorough and adequate filling of the joint cavity, which is crucial for the successful implementation of the treatment strategy.

Fig. 5.

Fabrication and characterization of hydrogels.

(A) Gross observation of the PH and PHPF@AD hydrogel formation process (a: 4-arm PEG-NHS, b: 4-arm PEG-NH₂, c: Hydrogel); (B) SEM images of lyophilized PH and PHPF@AD hydrogels.(scale bar = 200um); (C) Frequency sweeps of PH and PHPF@AD hydrogels at 1 % strain; (D) Viscosity measurements of PH and PHPF@AD hydrogels; (E) Mechanical properties: Compressive modulus of both hydrogels under 80 % strain; (F) Injectability and photograph of the hydrogel extruded using a syringe to write the letter “CY”; (G) In vitro swelling and (H) degradation properties of PH and PHPF@AD hydrogels in PBS at 37 °C; (I) AD release kinetics from PH and PHPF@AD in PBS; (J) Live/dead staining and (K) OD value of chondrocytes cultured in the medium of PH and PHPF@AD hydrogels(scale bar = 200um).

∗p < 0.05; ∗∗p < 0.01; ∗∗∗p < 0.001; ∗∗∗∗p < 0.0001;^nsp ≥ 0.05.

For the in - depth microscopic structural analysis, scanning electron microscopy (SEM) was employed, and the results are presented in Fig. 5B. The SEM images clearly revealed that both hydrogels possessed porous structures. These porous architectures play a vital role in maintaining high water - retention capacity and mechanical toughness, which are essential properties for hydrogels used in biomedical applications.

To precisely evaluate the mechanical properties of the hydrogels, a series of rheological tests were meticulously carried out. The frequency - sweep test results, shown in Fig. 5C, indicated that both hydrogel systems exhibited stable gel - like behaviors within the test frequency range. The shear - thinning experiments, as depicted in Fig. 5D, demonstrated that the apparent viscosity of the hydrogels decreased sharply with the increasing shear rates. This behavior is a characteristic feature of shear - thinning materials and is commonly observed in both hydrogel systems under investigation.

As a potential candidate material for knee joint implants, possessing adequate mechanical strength and toughness is of utmost importance. Compression tests were conducted, and the results are displayed in Fig. 5E. The findings revealed that both hydrogels exhibited excellent compressive properties. Notably, the hydrogels demonstrated remarkable toughness, with a maximum deformation of up to 80 %. Moreover, repeated compression tests did not result in significant degradation of their mechanical performance. Specifically, the PHPF@AD hydrogel showed slightly more superior mechanical properties at 80 % deformation, suggesting its better adaptability to the complex mechanical environment after implantation.

To quantitatively assess the injectability of the hydrogels, injection experiments were systematically performed, as shown in Fig. 5F. When extruded directly into water, the hydrogels formed continuous and slender threads. Furthermore, when writing “CY” using a 2.5 - ml medical syringe, uniform patterns were observed, which strongly indicates the material's suitability for uniform filling of the articular cavity, a key requirement for its application in joint - related therapies.

The swelling performance of the hydrogels was evaluated by directly immersing them in phosphate - buffered saline (PBS) to simulate the in - situ injection conditions, as presented in Fig. 5G. The results showed that the PHPF@AD hydrogel exhibited superior swelling properties compared to the other hydrogel, which may have implications for its functionality in the physiological environment.

The degradation rate of the hydrogels, which is critical for sustained drug release, was thoroughly evaluated in vitro, as shown in Fig. 5H. The PHPF@AD hydrogel exhibited a slower degradation rate compared to the PH hydrogel, which is highly beneficial for its role in providing prolonged drug release. Given the short - lived drug action associated with systemic or direct local administration, achieving sustained and on - demand drug release at the site of action is of great clinical significance.

To accurately evaluate the drug - loading efficiency, two systems were carefully prepared: one with AD directly mixed with PH and the other with AD encapsulated in micelles and then mixed with the hydrogel (PHPF@AD). The in - vitro drug - release results, as depicted in Fig. 5I, clearly demonstrated that the PHPF@AD hydrogel provided a more efficient and prolonged release profile of AD. Through these comprehensive experiments, it was determined that the PHPF@AD hydrogel exhibits excellent physicochemical properties, making it a promising candidate for intra - articular drug delivery in the treatment of osteoarthritis (OA).

However, the primary prerequisite for the practical application of any material in the biomedical field is safety and good biocompatibility. Using extracts of both hydrogel systems, live/dead staining was performed on mouse chondrocytes after 1, 3, and 5 days of culture, as shown in Fig. 5J. The results, further confirmed by CCK - 8 assays (Fig. 5K), indicated that both hydrogel systems showed no cytotoxicity. In conclusion, an injectable hydrogel material with outstanding physicochemical properties and excellent biocompatibility was successfully developed, holding great potential for future applications in the treatment of OA.

3.6. Evaluation of sustained-release drug properties of PHPF@AD by in vivo imaging

AD, PH, and PHPF@AD were labeled with DIR dye, and retention in the articular lumen was assessed with the IVIS spectroscopic system. The IVIS result is shown in Fig. 6. It can be seen from the IVIS image (Fig. 6A) that after intra-articular injection, the fluorescence intensity in the PHPF@AD group gradually decreased, and the rate of decrease was slower than that in the AD and PH groups. The fluorescence intensity of the joint cavity in the PHPF@AD group lasted for 24 days, while the fluorescence intensity in the AD and PH groups lasted only 5 and 9 days, respectively.

Fig. 6.

Evaluation of sustained-release drug properties of PHPF@AD by in vivo imaging.

(A) IVIS image of intraarticular injection of AD, PH, and PHPF@AD in mice over time. (B, C) Radiative efficiency and radiative residual ratio of the sample after intra-articular injection calculated by the IVIS imaging system. Data are expressed as the mean ± SD of at least three repeated experiments.

The IVIS results strongly confirm the fluorescence intensity efficiency (a.u) (Fig. 6B), and the residual fluorescence intensity efficiency (%) can be measured and calculated (Fig. 6C). As can be seen from Fig. 6B, the residual fluorescence intensity efficiency detected in AD and PH groups decreases more rapidly than PHPF@AD. Only small amounts of radiation were observed in the AD and PH groups on days 5 and 9, respectively, while those in the PHPF@AD group lasted for 24 days. The trend of residual fluorescence intensity efficiency (%) in Fig. 6B is similar to that in Fig. 6C.

3.7. PHPF@AD alleviates poor gait and radiographic bone degeneration in OA mice

After establishing OA in the mice via the DMM, there was no significant difference in body weight among all groups (Supplementary Fig. S5). Gait analysis was performed. Compared to the control group, OA mice exhibited reduced stride length, stride width, and toe spread, indicating that DMM surgery induced behavioural abnormalities that were partially reversed following PHPF@AD intervention (Fig. 7A–D). Micro-CT analysis revealed an increase in osteophytes and bone degeneration and a decrease in BV/TV and BMD in OA mice. Although PHPF@AD treatment partially reversed these effects, no statistically significant differences were obsereved in either BV/TV or BMD (Fig. 7E–G).

Fig. 7.

PHPF@AD alleviates poor gait and radiographic bone degeneration in OA mice.

(A) Representative images from the gait test for each group. Following PHPF@AD intervention, the poor gait induced by DMM was improved. (B) PHPF@AD intervention restored stride length levels reduced by DMM. (C) PHPF@AD intervention restored stride width levels decreased by DMM. (D) PHPF@AD intervention reversed the reduction in toe spread levels caused by DMM. (E) Representative Micro CT images for each group indicated that PHPF@AD alleviated osteophyte formation and bone erosion due to DMM. (F) PHPF@AD reversed the decrease in BV/TV levels caused by DMM. (G) PHPF@AD reversed the decrease in BMD levels caused by DMM.

∗p < 0.05; ∗∗p < 0.01; ∗∗∗p < 0.001; ∗∗∗∗p < 0.0001;^nsp ≥ 0.05.

3.8. PHPF@AD alleviates cartilage degeneration in OA mice

Histological assessments (HE and S-O staining) showed knee tissue injury was more severe in OA mice than in control mice, whereas PHPF@AD treatment significantly reduced this injury. Additionally, PHPF@AD decreased the Modified Mankin and synovitis scores in the OA mice (Fig. 8A–D). Compared with different doses of AD intraperitoneal injection intervention, the effect of PHPF@AD on joint protection of OA mice is close to that of high dose AD, and better than low and medium dose AD (Supplementary Fig. S6, Supplementary Fig. S7). qRT-PCR and western blot analyses revealed that levels of inflammatory factors IL-1β, IL-6, TNF-α, and matrix degradation factors MMP-1 and MMP-13 were upregulated in the knee cartilage of OA mice. Conversely, the anti-inflammatory cytokines IL-10 and IL-13 and matrix synthesis factors ACAN and COL2A1 were downregulated. Furthermore, the levels of Ki-67 and BCL-2 decreased, while those of the apoptosis factor Bax increased (Fig. 8E–S). These changes were significantly reversed following a single PHPF@AD injection intervention, however, these changes did not occur in OA + AD and OA + PH groups.

Fig. 8.

PHPF@AD alleviates cartilage degeneration in OA mice.

(A-B) Safranin O (S-O) and (C-D) Hematoxylin-Eosin (HE) staining were performed to assess tissue damage. PHPF@AD treatment significantly alleviated tissue damage and synovial lesions (HE and S-O, × 400, scale bar = 30 μm; × 100, scale bar = 100 μm). (E) The levels of IL-1β, IL-6, IL-10, IL-13, and TNF-α were measured via Western blot. (F) The expressions of ACAN, COL2A1, MMP-1, and MMP-13 were detected by Western blot. (G) The levels of Bax, Bcl-2, and Ki-67 were measured via Western blot. (H-S) Relative expressions of IL-1β, IL-6, IL-10, IL-13, TNF-α, ACAN, COL2A1, MMP-1, MMP-13, Bax, Bcl-2, and Ki-67 were evaluated using qRT-PCR.

∗p < 0.05; ∗∗p < 0.01; ∗∗∗p < 0.001; ∗∗∗∗p < 0.0001;^nsp ≥ 0.05.

3.9. PHPF@AD upregulates chondrocyte autophagy in OA through PRKCA/EGFR

Further analysis of cartilage tissues from experimental animals using qRT-PCR and western blotting demonstrated that PRKCA levels were upregulated and EGFR levels were downregulated in OA mice compared to those in the control group. Additionally, autophagy-related Beclin-1 and the LC3-II/LC3-I ratio were downregulated, whereas p62/SQSTM1 levels were upregulated, indicating weakened chondrocyte autophagy in OA mice. Following a single PHPF@AD injection intervention, PRKCA levels in OA cartilage were downregulated, EGFR expression was enhanced, Beclin-1 and LC3B levels were upregulated, and p62/SQSTM1 levels were decreased (Fig. 9A–F). This suggests that PHPF@AD enhances autophagy in the cartilage tissues of OA mice through the PRKCA/EGFR pathway. It is noteworthy that neither AD nor PH administration significantly reversed the autophagy changes caused by DMM.

Fig. 9.

PHPF@AD upregulates chondrocyte autophagy in OA through PRKCA/EGFR.

(A) The levels of PRKCA, EGFR, Beclin-1, LC3-I, LC3-II, and P62 were analyzed using Western blot. (B-F) Relative expressions of PRKCA, EGFR, Beclin-1, LC3B, and P62 were measured via qRT-PCR.

∗p < 0.05; ∗∗p < 0.01; ∗∗∗p < 0.001; ∗∗∗∗p < 0.0001;^nsp ≥ 0.05.

4. Discussion

Recently, inflammatory apoptosis and autophagy in articular chondrocytes have garned considerable attention. Autophagy exerts a cytoprotective effect on chondrocytes [5], and its effect is reduced in OA cartilage [9,10]. Notably, AD, a multitarget anti-inflammatory drug, shows a promising potential for OA treatment.

In this study, we utilized IL-1β to establish in vitro models of OA and investigated the effects of various AD concentrations on these models. We confirmed that AD alleviated IL-1β-induced damage to chondrocytes in vitro and promoted autophagy. Furthermore, considering the relationship between PRKCA and AD, particularly the influence of PRKCA on EGFR expression, we concluded that AD enhanced autophagy, thereby mitigating chondrocyte injury via the PRKCA/EGFR pathway and ultimately, providing relief from OA.

In this in vitro study, we focused on AD because of its anti-inflammatory, antineoplastic, antioxidant, and immunological properties [12,13,37], as well as its anti-arthritic effects in OA [15]. Importantly, its potential role in promoting autophagy in chondrocytes piqued our interest [17,18]. We hypothesised that AD alleviates OA by inducing autophagy in chondrocytes. Reportedly, inflammation of the synovial membrane contributes to OA progression, leading to cartilage matrix degradation via pro-inflammatory mediators [38,39]. Inflammatory mechanisms involved in the regulation of OA progression include the mediation of biomechanical disorders in chondrocytes. Various inflammatory mediators released by chondrocytes after joint injury serve as promising therapeutic targets in OA [40]. Notably, TNF-α, IL-1β, and IL-6 are recognised cytokines that correlate with the degree of the body's inflammatory response. Literature indicates that elevated levels of these inflammatory factors (TNF-α, IL-1β, and IL-6) are released during OA progression [41]. The expression of IL-1β is linked to joint damage [42], while IL-6 exacerbates both destruction and inflammation in the synovium [43]. Additionally, TNF-α correlates with joint destruction and inflammation in rheumatoid arthritis synovial fluid and has been shown to promote OA progression [39,44]. Our study confirmed that AD can attenuate inflammation in OA models, as evidenced by reduced TNF-α, IL-1β, and IL-6 levels, increased chondrocyte viability, and decreased apoptosis in IL-1β-induced models.

Notably, the imbalance in extracellular matrix (ECM) homeostasis is closely associated with cartilage injury in osteoarthritis [45]. Extracellular matrix (ECM) is an interconnected complex of collagen, proteoglycans, and water secreted and synthesized by chondrocytes, providing nutrients and support to these cells. The disruption of ECM homeostasis leads to an imbalance between cartilage matrix degradation and repair, often accompanied by the abnormal expression of certain proteases [46]. For instance, MMP-13 is known to cleave type II collagen and is considered a promising target for OA treatment [47]. Furthermore, MMP-1 may contribute to OA progression through cartilage degradation, with elevated levels in the synovial fluid or cartilage associated with knee OA pathogenesis [48]. Our findings demonstrate that AD can suppress the expression of MMP-1 and MMP-13 in OA models, further confirming their role in restoring ECM homeostasis. In animal experiments, AD alleviated cartilage damage and synovial inflammation in OA mice, downregulated inflammatory and ECM-degrading factors, and upregulated cell viability.

Autophagy in chondrocytes is closely linked to OA development, making it a critical target for OA treatment [11]. Recent findings indicate that autophagy can reduce inflammation, upregulate autophagy-related factors, and downregulate oxidative stress-related proteins (such as MMPs and iNOS) to protect chondrocytes, thereby alleviating OA onset [8]. Autophagy is a protective mechanism essential for maintaining normal cartilage function [49]. Previous studies have shown that autophagy levels in OA patients’ chondrocytes are significantly reduced, as evidenced by the decreased expression of ULK1, Beclin-1, and LC3 [50].ULK1, Beclin-1, LC3, and P62 are increasingly recognised as biomarkers of autophagy [24,26]. LC3 exists as LC3-I in the cytosol and as LC3-II in autophagosomal membranes [10]. Moreover, the downregulated expression of autophagy proteins has been observed in both mouse models and aged human cartilage [10], and emerging research indicates that increased chondrocyte apoptosis in OA is correlated with diminished levels of BECN1, LC3, and P62 [51]. We confirmed that AD could reverse IL-1β-induced suppression of autophagy in OA models in vitro, as indicated by elevated LC3 levels and increased Beclin-1 expression, along with a higher LC3-II/LC3-I ratio following AD treatment.

Next, we focused our study on investigating the mechanism by which AD influences the autophagy of chondrocytes in IL-1β-induced OA models. As previously mentioned, PRKCA gets abnormally upregulated in OA [20]. Furthermore, PRKCA downregulates EGFR expression [21], and elevated EGFR levels are associated with enhanced chondrocyte autophagy in OA [22]. Using overexpressed PRKCA, we confirmed that AD promotes autophagy, thereby mitigating IL-1β-induced chondrocyte injury through the PRKCA/EGFR pathway. This was validated by the neutralisation of overexpressed PRKCA, which affected inflammatory factor levels, cell viability, apoptosis, autophagy, and EGFR expression in IL-1β-induced OA models.

As previously elucidated, the orthodox modalities for the management of osteoarthritis (OA) utilizing andrographolide (AD) frequently mandate elevated dosages and uninterrupted administration. To surmount this constraint, we have fabricated a novel injectable hydrogel formulation with the objective of augmenting the bioavailability of AD and effecting its sustained liberation in vivo. This hydrogel was concocted by amalgamating 4-arm PEG-NH₂ and 4-arm PEG-NHS in equimolar proportions. Nonetheless, complications arose owing to the innate hydrophobicity of AD and the hydrophilic disposition of the hydrogel matrix. When AD was directly incorporated, it gave rise to substandard drug retention and diminished efficacy. To alleviate this predicament, we integrated PF127, a synthetic A-B-A triblock copolymer constituted of polyethylene glycol-polypropylene glycol-polyethylene glycol. The central polypropylene glycol segment of PF127 exhibits a relatively hydrophobic character, whereas the terminal polyethylene glycol segments are hydrophilic. This amphipathic property empowers PF127 to encapsulate AD and self - assemble into micelles.

Our PEG-Mix injectable hydrogel framework is characterized by its water-rich, loose, and porous structure, wherein the hydrophilic outer layer of the micelles promotes their long-term stability within the hydrogel matrix. This configuration allows for the sustained release of micelles, thereby maximizing the therapeutic effect of AD. As the PHPF@AD hydrogel gradually degrades, the micelles containing AD are released, significantly enhancing drug bioavailability and prolonging the duration of action. In vivo studies on OA-induced mice demonstrated that AD could be released slowly in the knee joint for over 24 days post-injection, enabling a single dose to achieve prolonged therapeutic efficacy. This approach markedly diminishes the medical burden on patients, obviates the risks associated with repeated injections, such as pain and infection, and reduces the economic costs of treatment.

We developed the injectable PHPF@AD hydrogel system under mild reaction conditions and validated its injectability. The hydrogel exhibits fluidity upon injection, rapidly spreading within the joint cavity and undergoing spontaneous gelation. Experimental validation confirmed that this hydrogel system has a reduced gelation time, significantly minimizing the risk of material leakage from the administration site. Additionally, our hydrogel system possesses superior mechanical toughness, withstanding repeated compression deformations without compromising joint integrity. In theory, this resilience ensures that the hydrogel can fully adapt to the dynamic motion patterns of joint tissues without causing damage. The gait test results of OA-induced mice demonstrated that both the OA + PH and OA + PHPF@AD groups exhibited improved gait abnormalities compared to the OA + AD group, underscoring the therapeutic potential of our design.

This structural innovation ensures the slow-release performance of the entire hydrogel system, an attribute that distinguishes it from other recently reported injectable hydrogel formulations [[52], [53]]. The injectable hydrogel system in this study is characterized by a short gelation time, no requirement for complex polymerization conditions, non-toxicity, and sustained drug release, thereby qualifying it for potential clinical translation.

Post-validation of the hydrogel's non-toxicity to chondrocytes, we applied the PHPF@AD system to mice with DMM-induced OA. Four weeks post-DMM induction, OA mice were administered a single knee injection of AD, PH, and PHPF@AD, each mixed with fluorescent dyes to facilitate real-time observation of drug presence within joint lumen tissues via in vivo imaging, a common method for evaluating sustained drug release [[31], [54]]. Experimental results indicated that PHPF@AD could sustain release in vivo for over 24 days, aligning with the expectations of this study. Micro-CT assessments, though not statistically significant, suggested that the sustained-release drug PHPF@AD could mitigate bone loss in OA mice. Histopathological staining of knee joints from each group revealed that a single administration of the PHPF@AD system could reverse DMM-induced knee joint tissue damage, a significant improvement over single AD administration.

Moreover, PHPF@AD effectively reduced the expression of inflammatory cytokines IL-1β, IL-6, and TNF-α induced by DMM, while simultaneously increasing the expression of anti-inflammatory cytokines IL-10 and IL-13 in mice, commonly assessed markers in animal OA models [55]. Furthermore, PHPF@AD was confirmed to alleviate chondrocyte matrix degradation and apoptosis induced by DMM. We also established the regulatory influence of AD on the PRKCA/EGFR axis and its role in promoting autophagy, findings consistent with prior cellular experiments. In summary, the PHPF@AD system demonstrates significant potential in the realm of OA treatment.

5. Conclusion

In summary, the natural compound AD exhibits efficacy in alleviating OA-associated chondrocyte injury, with the enhancement of chondrocyte autophagy by AD elucidated in this study. The innovative injectable PHPF@AD system efficiently loads and sustainably releases AD, enabling a single dose to achieve long-term therapeutic efficacy, thereby overcoming the application limitations of AD requiring extensive long-term administration. The injectable PHPF@AD exhibits a short in-situ gelation time, substantially reducing the risks of material leakage and drug loss. Our findings indicate that the injectable PHPF@AD system can enhance chondrocyte autophagy in OA, mitigate disease progression, reduce joint tissue injury and inflammation levels, characterized by a prolonged onset, shortened gelation time, and diminished medical burden. Additionally, the system's robust mechanical properties ensure uncompromised knee joint mobility. Given the complex pathogenesis of OA and the limited therapeutic options, the PHPF@AD system developed herein holds promise as a novel therapeutic strategy for osteoarthritis, leveraging these advantageous properties.

CRediT authorship contribution statement

Yang Chen: Writing – original draft, Data curation, Conceptualization. Peipei He: Writing – review & editing, Formal analysis, Data curation. Siyi Tao: Investigation, Funding acquisition, Data curation. Jintao Zhong: Software, Methodology. Kai Jiang: Validation, Software, Data curation. Yuching Hsu: Visualization, Validation, Resources. Guang Xia: Visualization, Software. Xinzhan Mao: Methodology, Investigation. Hongxun Sang: Writing – review & editing, Supervision, Funding acquisition. Ke Lu: Supervision, Project administration, Investigation, Funding acquisition.

Ethical approval

Ethical approval for the experimental procedures was obtained from the Animal Experiment Ethics Committee of Hunan University of Chinese Medicine (No: ZYFY20231110-71) and the Animal Experiment Ethics Committee of Hunan University of Chinese Medicine (No: 2024–2028).

Funding

This work was supported by the National Natural Science Foundation of China (82302757), Shenzhen Science and Technology Program (SGDX20201103095600002, JCYJ20220818103417037, KJZD20230923115200002),

Shenzhen Key Laboratory of Digital Surgical Printing Project (ZDSYS201707311542415),

Shenzhen Development and Reform Program (XMHT20220106001), Hunan Provincial Science and Technology Department, China (2018JJ2579, 2015JC3034).

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.

Appendix A. Supplementary data

The following are the Supplementary data to this article:

Data availability

The data that has been used is confidential.