Targeted S100A8 PP-Cas9@PLGA-apt microparticles reduced cartilage degradation and subchondral bone isomerism in osteoarthritis treatment through JAK/STAT3 pathway modulation

Abstract

Synovial inflammation caused by osteoarthritis (OA) results in the release of numerous pro-inflammatory factors that promote cartilage degradation and pathological changes of subchondral bone. Nowadays, S100A8 has been recognized as a critical factor in the progression of inflammatory diseases, but its role in OA still needs to be confirmed. At the same time, the gene editing technology has emerged as a novel therapeutic approach for OA, such as clustered regularly interspaced short palindromic repeats (CRISPR/Cas9) technology, but application in inflammatory gene therapy still requires advanced delivery systems to ensure cell-specific targeting and biosafety.

In this study, S100A8 was confirmed as a key mediator perpetuating JAK/STAT3 pathway activation in OA progression by integrated RNA bioinformatics and synovial proteomic analyses. Based on it, we developed a polyamidoamine (PAMAM)-poly (lactic-co-glycolic acid) (PLGA) (PP) nanocore electrostatically complexed with Cas9-S100A8, encapsulated within an aptamer (apt)-grafted PLGA shell structure. This multifunctional nanocarrier could reduce dendrimer toxicity to cells and protein degradation, and enhance cellular targeting and endocytic capacity. PP-Cas9-S100A8@PLGA-apt exhibited 64.4 % S100A8 knockout efficiency (p < 0.001) and sustained mRNA release (71.5 % retention at 48 h), high cell viability (>80 %), and synovium-specific uptake (98.8 % at 0.8 μg/mL), inhibiting the JAK/STAT3 pathway. In OA-induced mice, this inhibition reduced pro-inflammatory responses, cartilage degradation, and attenuated osteophyte volume.

Our findings first established PP-Cas9-S100A8@PLGA-apt as an efficient and safe Cas9 delivery tool, advancing studies of JAK/STAT3 pathway inhibition and the clinical translation of gene therapy for OA.

Graphical abstract

The S100A8 preciously edited by Cas9-S100A8 from PP-Cas9-S100A8@PLGA-apt MPs resulted in the decreased phosphorylation of JAK/STAT3 pathway. Afterward, the inflammatory cytokines such as MMP13, were decreasingly released. The decreased secretion of inflammatory cytokines in joints effectively mitigated synovial inflammation, cartilage degeneration, and subchondral bone disarray in osteoarthritis.

1. Introduction

Osteoarthritis (OA), a prevalent degenerative joint disease, manifests clinically through pain, stiffness, and functional impairment of affected joints. Synovial inflammation plays a critical role in the progression of OA, manifesting as an abnormal immune response within the intimal layer of joints. It leads to joint swelling, localized hyperthermia, and tenderness, ultimately driving irreversible degradation of articular cartilage and subchondral bone [1]. This inflammation can be triggered by various factors, including aging, obesity, trauma, and mechanical load, establishing inflammatory modulation as a critical therapeutic target for decelerating OA progression [1]. Current treatments for synovial inflammation in OA include exercise, steroid injections, and usage of anti-inflammatory medications such as nonsteroidal anti-inflammatory drugs (NSAIDs) or disease-modifying antirheumatic drugs (DMARDs). When less invasive options fail, surgical interventions may be undertaken, such as synovectomy or joint replacement surgery. While synovectomy can alleviate pain and prevent disease progression, it carries risks including hemorrhage, postoperative pain and infection, and anesthesia-related complications [2]. Similarly, joint replacement surgery, the solution for severe arthritis unresponsive to conservative treatments, also presents potential complications such as infection and deep vein thrombosis (DVT) [3]. These challenges underline the urgent need to develop new therapeutic strategies that are both effective and safe.

CRISPR/Cas9 gene editing has emerged as a transformative therapy for OA, targeting pathological genes such as MMP13 [4] and IL-16 [5], to reduce pain and slow joint deterioration. However, the clinical translation of this technology is constrained by limitations in gene delivery efficiency, the transient windows of gene edits, and rapid clearance from the in vivo environment [5]. These issues have accelerated the development of innovative materials and delivery systems, such as polyamidoamine (PAMAM) dendrimers. While possessing potential to improve CRISPR/Cas9 component delivery, PAMAM is encumbered with significant drawbacks [6,7]. These include cytotoxic effects [[8], [9], [10], [11]] from higher-generation dendrimers with surface charges, possible immunogenic responses, insufficient RNA protection from enzymatic degradation, and diminished duration of gene editing effects in vivo environments [12]. Most critically, the non-specific PAMAM dendrimer-mediated delivery carries a risk of off-target genetic modifications [13]. This underscores the need for developing enhanced targeting delivery to achieve clinically viable OA gene therapy.

To alleviate the inherent cytotoxicity of PAMAM dendrimers, strategic surface modifications are employed by attaching poly (lactic-co-glycolic acid) (PLGA) molecules to neutralize surface charges [14]. As a biodegradable polymer esteemed in drug delivery, PLGA can not only diminish cytotoxicity, but also offer controlled degradation and sustained release properties optimal for gene delivery vectors by integration [15]. Nevertheless, the lack of specific targeting capability in PAMAM-PLGA (PP) MPs may affect the delivery of drugs to the intended pathological sites, thereby undermining therapeutic efficacy. This drawback results in the use of potentially higher drug dosages that increase the likelihood of adverse reactions and considerable therapeutic agent wastage due to systemic dispersion. Additionally, this indiscriminate distribution can provoke immune responses, exacerbate the complexity of treatment and impact safety profiles.

To identify potential gene targets for OA therapy, we performed an integrative analysis combining proteomic profiling of OA synovial tissues with transcriptomic data from the GEO database (GSE82107). Among the 13 overlapping genes identified, S100A8 emerged as the target consistently and significantly upregulated across both datasets. In addition, previous studies have demonstrated that S100A8 is a key pro-inflammatory mediator in OA, promoting synovitis and cartilage degeneration [16,17]. These findings collectively support the rationale for selecting S100A8 as a promising therapeutic target for gene editing–based intervention.

In conclusion, we pioneered the synthesis of a PP core designed to bind the S100A8 gene, a target identified through meticulous synovial cell database screening and proteomic analysis for its role in OA. This core was subsequently encapsulated in PLGA and surface-functionalized with an aptamer that targeted OA synovial cells. This innovative approach was poised to endow the therapeutic construct with the requisite specificity, substantially improving the drug delivery to OA lesions and potentially leading to a safer and more effective treatment.

2. Methods

2.1. Analysis of S100A8 gene in human synovial tissue of knees

Synovial tissues were collected from 3 patients undergoing arthroscopic meniscal repair (normal group) and 3 OA patients undergoing total knee arthroplasty (OA group), with ethical approval (Approval No.SYSEC-KY-KS-2021-243). Samples were homogenized, centrifuged, and quantified via BCA assay. Proteins were then digested with trypsin, labelled with iTRAQ tags, purified (C18 column), fractionated (high-pH reverse-phase), analyzed by LC-MS/MS, and differentially expressed proteins (DEPs) was identified based on iTRAQ signal intensity. For integrative analysis, the proteomics data were cross-validated with the GEO dataset GSE82107. Pathway analysis was conducted using GSEA. DEPs from the OA group were compared with differentially expressed genes (DEGs) from GSE82107 to identify overlapping RNA-protein targets in OA joints.

For immunohistochemical analysis, synovial tissues were fixed in 10 % neutral-buffered formalin, paraffin-embedded, and sectioned (5 μm). Then, sections were subjected to heat-induced epitope retrieval in citrate buffer (pH 6.0, 95 °C, 15 min), blocked with 3 % H₂O₂ and 5 % BSA. The sections were then left to incubate overnight with primary antibodies against S100A8 at a temperature of 4 °C, followed by 1 h of incubation with secondary antibodies at room temperature. A diaminobenzidine (DAB) substrate kit (Solarbio) was used to induce color development.

OA and normal synovial tissues were lysed in RIPA buffer containing protease/phosphatase inhibitors, homogenized, and centrifuged (12,000×g, 15 min, 4 °C). Protein concentrations were quantified via BCA assay. Equal protein (30 μg) was separated by SDS-PAGE, transferred to PVDF membranes, and blocked with 5 % non-fat milk (TBS-T, 1 h, room temperature). Membranes were probed overnight at 4 °C with primary antibodies against S100A8, and β-actin (loading control). HRP-conjugated secondary antibodies (1 h, room temperature) and ECL detection were used for visualization. Band intensities were quantified with ImageJ (normalized to β-actin).

The role of S100A8 was determined using Gene Ontology (GO) analyses. In OA progression, the elevation of IL-6 can be observed in the OA joints, leading to subsequent JAK/STAT3 pathway overactivation, and proliferation of synovial cells [[18], [19], [20]]. As a result, synovial cells obtained from patients with meniscal injury undergoing arthroscopy were treated with IL-6 (10 ng/ml) to mimic the OA status. Then, these cells were treated with or without S100A8, and si-S100A8, with vehicle-treated cells as controls. After treatment, cells were lysed and protein was collected. Western blotting was performed to analyze the expression of several key proteins in the JAK/STAT signaling pathway. Details were shown in supplementary material.

2.2. Synthesis of CRISPR/Cas9 mRNA targeting S100A8 gene knockout and T7E1 enzyme detection

The CRISPR/Cas9 mRNA system (RIBOBIO, China) was transfected into synovial cells (50–70 % confluency in 24-well plates) using riboFECT™ mRNA reagent. Briefly, 25 μL Opti-MEM™ was mixed with 2 μL riboFECT™ reagent (room temperature, 10 min). Separately, 25 μL Opti-MEM™ was combined with riboEDIT™ Cas9 mRNA (500 ng/μL, 2 μL), tracrRNA (10 μM, 1 μL), and 1 of 5 candidate crRNAs (10 μM, 1 μL) in a 1:1 M ratio. The ribonucleoprotein (RNP) complex was incubated with the transfection reagent (5 min, RT), added to cells, and cultured for 48 h. Gene-editing efficiency was assessed via riboEDIT™ T7EI cleavage assay using extracted genomic DNA. Relevant details were shown in supplementary material.

2.3. Development of Cas9 mRNA-encapsulated MPs (PP-Cas9-S100A8@PLGA)

Dendrimers, such as PAMAM, can bind DNA and RNA via electrostatic adsorption and facilitate their cellular delivery [[21], [22], [23]]. The PLGA-PAMAM (PP) copolymer was synthesized by conjugating PLGA-COOH (Mw 2000) with NH2-PAMAM G3 via EDC/NHS chemistry [24,25]. Structural confirmation was achieved by ¹H NMR (600 MHz, DMSO/D₂O). PP-mRNA complexes were prepared by mixing the copolymer with Cas9-S100A8 mRNA in DMSO. To enhance stability, PP-mRNA cores were encapsulated into PLGA microspheres (50:50, Mw 20k) using an O/W emulsion method: 100 mg PLGA in DCM was emulsified with 10 g/L PP-mRNA, homogenized in 2 % PVA (85–124k Da), vortexed (2000 rpm, 30 min), and centrifuged. Microspheres were washed 3 × with PBS and lyophilized.

2.4. Conjugation of synovial cell aptamer to the surface of PLGA MPs (PP-Cas9-S100A8@PLGA-apt MPs)

SA.2 aptamers (5′-aminated) targeting synovial cells in OA joints were synthesized and selected according to our previous patent [26]. Then they were conjugated to carboxylate groups on PP-Cas9-S100A8@PLGA MPs, following the protocol described by Li et al. [27]. Briefly, MPs (10 μg/μL) were activated with 400 mM EDC/100 mM NHS (30 min, RT, stirring), washed (ultrafiltration, nuclease-free H₂O), and incubated with aptamers to form PP-Cas9-S100A8@PLGA-apt bioconjugates. Unbound aptamers were removed by ultrafiltration. The PP-Cas9-S100A8@PLGA-apt MPs were stored in nuclease-free H₂O at 4 °C.

2.5. Characterization of PP-Cas9-S100A8@PLGA-apt MPs

MPs size and zeta potential were analyzed by dynamic light scattering (DLS; Malvern Instruments, UK, triplicate measurements). Fourier-transform infrared (FTIR) spectra were carried out on a Thermo Nicolet 670 FT-IR instrument (Thermo Nicolet Corporation, USA) operating at the wavelength range of 4000–400 cm⁻¹. SEM imaging was used to evaluate MP morphology. Encapsulation efficiency of PP/Cas9-S100A8 complexes was assessed via agarose gel retardation: naked mRNA and PP/mRNA complexes (weight ratios 1:1–1:40) were electrophoresed (1 % agarose/TAE, 100 V, 30 min) and visualized under UV. Retarded migration indicated mRNA encapsulation, with naked mRNA as a migration control.

2.6. Controlled release of Cas9-S100A8 from encapsulated MPs in vitro

FAM-labelled mRNA release kinetics from MPs were evaluated under physiological conditions (PBS, pH 7.4, 37 °C). MPs were dialyzed in 1 mL PBS, with 1 mL aliquots sampled and replaced with fresh PBS at days 1–8. Released mRNA was quantified via absorbance spectroscopy (480 nm, FAM label). Three independent biological replicates were performed.

2.7. PP-Cas9-S100A8@PLGA-apt MPs targeted internalization in different cells

Synovial cells, chondrocytes, meniscal cells, and HepG2 cells (5 × 10⁴/well in 24-well plates) were cultured and treated with PP-Cas9-S100A8@PLGA-apt MPs (0.8 μg/mL, 24 h, each condition included 3 technical replicates per plate). Cells were PBS-washed, fixed with 4 % PFA, and stained with Alexa Fluor 488-WGA (cell membrane) and DAPI (nuclei). Confocal imaging (Zeiss LSM 710) assessed MP internalization via co-localization analysis of MPs with membrane/nuclear markers. High-resolution images confirmed effective intracellular MP localization across cell types.

2.8. Endosomal escape capacity of PP-Cas9-S100A8@PLGA-apt MPs

Synovial cells were incubated with CY5-labelled PP-Cas9-S100A8@PLGA-apt MPs (0.8 μg/mL) for 8, 12, and 24 h. Receptor-mediated endocytosis and endosomal escape capacity were evaluated via co-localization analysis of CY5-labelled MPs with the early endosome marker EEA-1 antibody using fluorescence microscopy. Three independent biological replicates were performed. Endosomal escape capacity was assessed by tracking MP dissociation from EEA1-positive vesicles over time.

2.9. The optimal concentration of PP-Cas9-S100A8@PLGA-apt MPs cellular uptake by synovial cells

Synovial cells (2 × 10⁵/well in 6-well plates) were cultured and treated with PP-Cas9-S100A8@PLGA-apt MPs (0.1–0.8 μg/mL ± Lipofectamine 2000 [Lip2K]) for 24 h. Cells were PBS-washed, trypsinized, and analyzed via flow cytometry (BD FACSCanto II, ≥10,000 events/sample). Three independent biological replicates were performed. MP internalization rates were quantified using FlowJo.

2.10. Migration of synovial cells treated with different MPs

In scratch assay, synovial cells (6 × 10⁵/well in 6-well plates) were cultured to create a confluent monolayer and a 200 μL pipette tip was used to make a vertical scratch in the center of the wells. Next, each well was subjected to serum-free medium with PAMAM-Cas9-S100A8, PAMAM-Cas9-S100A8@PLGA, PP-Cas9-S100A8@PLGA, or PP-Cas9-S100A8@PLGA-apt (0.8 μg/mL) for 24 h after gently removing the dislodged cells. Wells only added with serum-free medium were the control group. Three independent experiments were performed (each with 3 technical replicates). The cellular migration under different treatments was imaged by fluorescence microscope (Olympus IX73).

In matrigel transwell assay, synovial cells were treated with PAMAM-Cas9-S100A8, PAMAM-Cas9-S100A8@PLGA, PP-Cas9-S100A8@PLGA, or PP-Cas9-S100A8@PLGA-apt (0.8 μg/mL) respectively in serum-free medium and seeded in upper chambers (4 × 10⁴/well in 24-well plates) with 8-μm pores (Corning). Cells only treated with serum-free medium were the control group. Lower chambers contained 600 μL of 0.1 % FBS medium. After 24 h incubation, migrated cells infiltrated into the bottom compartment were fixed and stained using 0.4 % crystal violet, imaged by fluorescence microscopy (Olympus IX73). Three independent experiments were performed (each with 3 technical replicates). The stained cell images were used for cell invasion analysis, and quantified using ImageJ v1.53.

2.11. Cytotoxicity assay

Synovial cells (8 × 10³/well in 96-well plates, each condition included 3 technical replicates per plate) were treated with PAMAM-Cas9-S100A8, PAMAM-Cas9-S100A8@PLGA, PP-Cas9-S100A8@PLGA, or PP-Cas9-S100A8@PLGA-apt (0.8 μg/mL) for 24, 48, and 72 h. Cell viability was assessed using CCK-8 (Dojindo) at each time point: medium in each well was replaced with 100 μL of fresh medium containing 10 μL of CCK-8, followed by 4 h incubation. Absorbance measured at 450 nm (Thermo Fisher). Medium-only served as control.

2.12. MPs uptake in synovial cells

Synovial cells (1 × 10⁵/well in 12-well plates) were pre-cultured for 24 h and treated with FAM-labelled MPs: PAMAM-Cas9-S100A8, PAMAM-Cas9-S100A8@PLGA, PP-Cas9-S100A8@PLGA, or PP-Cas9-S100A8@PLGA-apt. After 24, 48, and 72 h of incubation, uninternalized MPs were removed by washing three times with PBS. Three independent experiments were performed (each with 3 technical replicates). FAM fluorescence intensity (indicative of MP uptake) was quantified via confocal microscopy (Olympus IX73).

2.13. Vivo fluorescence imaging analysis

Male C57 mice (12-week-old, ∼30 g) were anesthetized with 1.5 % sodium pentobarbital (80 mg/kg) and received intra-articular injections of CY7-labelled Cas9-S100A8 mRNA-loaded MPs (10 μL, 0.8 μg/mL) via the patellar tendon, approved by the Ethics Committee of Sun Yat-sen University (Approval NO.SYSU-IACUC-2021-B0932). According to different MPs injections, mice were divided into four groups: PAMAM-Cas9-S100A8, PAMAM-Cas9-S100A8@PLGA, PP-Cas9-S100A8@PLGA, and PP-Cas9-S100A8@PLGA-apt (3 mice per group). Fluorescence signals were monitored immediately (0 h) and 96 h post-injection using a FluorVivo imaging system (INDEC BioSystems, USA).

2.14. Animal model and treatment

Male C57 mice (12-week-old; Approval NO.SYSU-IACUC-2021-B0932) were randomized into an anterior cruciate ligament transection (ACLT) group (n = 80) and a sham group (n = 40). The ACLT group underwent anterior cruciate ligament transection to induce post-traumatic osteoarthritis, while the sham group received a surgical incision without ligament transection. Postoperatively, ACLT mice were subdivided into PBS control (n = 40) and PP-Cas9-S100A8@PLGA-apt treatment (n = 40) groups. Mice from all groups underwent gait analysis and were euthanized at 2, 4, 8, and 16 weeks postoperatively for tissue collection and comparative subsequent analysis.

2.15. Immunohistochemistry and immunofluorescence staining of knees of mice after treatment

Mice knee sections (5 μm) were deparaffinized, rehydrated, and a subset of sections were stained with Safranin O/Fast Green (S-O). The remaining sections were subsequently subjected to antigen retrieval (citrate buffer, pH 6.0). Endogenous peroxidase was blocked with 3 % H₂O₂, followed by 5 % BSA blocking. Synovial tissues sections (S100A8, 1:200; JAK 1:500; STAT3, 1:500) and cartilage sections (Aggrecan, 1:200; Collagen II, 1:500; MMP13, 1:500; Endomucin, 1:500; CD31, 1:500) were incubated with corresponding primary antibody, respectively (4 °C, overnight). S100A8 detection followed the HRP-conjugated secondary antibodies, developed with DAB substrate, and counterstained with hematoxylin. For rest of sections, corresponding fluorophore-conjugated secondary antibodies (1:500) were were applied and incubated (1 h, room temperature, dark). Nuclei were counterstained with DAPI, and slides were mounted with antifade medium. All images were acquired using the light/fluorescence microscopes (Olympus, Tokyo, Japan).

2.16. Western blot analysis of synovial tissue and cartilage of mice after treatment

Synovial tissues and cartilage were lysed in RIPA buffer containing protease/phosphatase inhibitors, homogenized, and centrifuged (12,000×g, 15 min, 4 °C). Protein concentrations were quantified via BCA assay. Equal protein (30 μg) was separated by SDS-PAGE, transferred to PVDF membranes, and blocked with 5 % non-fat milk (TBS-T, 1 h, room temperature). Membranes were probed overnight at 4 °C with primary antibodies against JAK, STAT3, S100A8, MMP3, MMP13, Aggrecan, Collagen II (all at a dilution of 1:1000), and β-actin (loading control). HRP-conjugated secondary antibodies (1 h, room temperature) and ECL detection were used for visualization. Band intensities were quantified with ImageJ (normalized to β-actin).

2.17. Micro-CT and gait analysis

Mouse knee joints were scanned using a micro-CT system (Scanco Medical) to quantify subchondral bone volume and osteophyte formation. Images were reconstructed and analyzed with built-in software. Subchondral bone parameters were measured in regions of interest (ROIs) at 2, 4, 8, and 16 weeks post-surgery and compared to sham-operated controls.

The gait of the mice was analyzed by collecting footprints from a walk at 2, 4, 8, and 16 weeks post-surgery and compared to sham-operated controls. The fore paws of the mice were marked red and the hind paws were marked blue. Mice walked freely inside a dark enclosure on a recording sheet from one side to the other without any external interference. Each mouse underwent three repeated experiments, and three consecutive footprints were selected from the recording sheet for statistical analysis.

2.18. Statistics

Statistical analysis and graphing were performed using R (version 4.3.2; R Foundation for Statistical Computing, Vienna, Austria). The two-tailed Student's t-test with Welch's correction (to stabilize the variance between different groups) was used to compare two experimental groups, and one-way Anova with Tukey's post hoc analysis was used for comparing more than two groups. A value of p < 0.05 was considered statistically significant.

3. Results

3.1. Comparative analysis of protein profiles and overlapping gene expression in normal and OA human synovium

Fig. 1A systematically presented the protein profiles and overlapping gene expression features between healthy human synovium and OA-affected synovium. The left panel schematically illustrated the synovium isolation for protein profiles analysis using iTRAQ technology. Subsequent bioinformatics analysis integrated data from the GSE82107 database to identify differential proteins. The right panel featured two pie charts: the left pie chart illustrated 558 differential genes identified in the GSE82107 database, while the right pie chart illustrated 163 differential proteins identified in our proteomic analysis. Intersectional analysis yielded 13 common genes, including S100A8, and highlighted their potential significance as the key mediators in OA pathogenesis. Previous studies have verified that S100A8 has pro-inflammatory functions, and especially promotes OA progression [16,17]. As a result, we chose S100A8 as the targeted gene in our further study. Fig. 1B demonstrated a significant difference in S100A8 gene expression between OA and normal synovium, strongly underscoring its involvement in OA pathogenesis [28]. The volcano plot in Fig. 1C revealed additional genes potentially associated with the pathogenic effects of S100A8 in OA, providing novel mechanistic insights into OA and offering potential therapeutic targets.

Fig. 1.

S100A8 identification through integrated bioinformatics and experimental validation in human OA. A, schematic overview of the experimental design and bioinformatics analysis. Human normal (n = 3) and OA (n = 3) synovium samples were collected, followed by protein extraction and trypsin digestion. RNA analysis was conducted in parallel. The samples underwent LC-MS/MS analysis, and data were integrated using bioinformatics tools. A Venn diagram showed the overlap between gene expression data (GSE82107) and protein analysis results, identifying key differentially expressed proteins. B, the box plot showing a significant upregulation of S100A8 in human OA synovium samples compared to the normal samples. C, the volcano plot of differentially expressed proteins in human OA versus normal synovium samples. S100A8 was highlighted as one of the significantly upregulated proteins in OA. D, immunohistochemical staining of S100A8 in normal and OA synovium samples at 4X and 40X magnification. The images demonstrated a higher intensity of S100A8 staining in OA samples compared to normal samples. E, quantification of S100A8 positive cells in normal and OA synovium samples. The bar graph showed a significantly higher percentage of S100A8 positive cells in OA samples. F, Western blot analysis of S100A8 protein levels in normal and OA synovium samples. The blot showed increased S100A8 expression in OA samples (OA1 to OA4) compared to normal synovium, with actin used as a loading control. G, densitometric analysis of the Western blot results. The bar graph indicated a significant increase in S100A8 expression in OA samples compared to normal controls, confirming the upregulation observed in the proteomic and immunohistochemical analyses. ∗∗∗, p < 0.001, ∗∗∗∗, p < 0.0001. OA, osteoarthritis.

3.2. Differential expression of S100A8 and JAK/STAT3 pathway activation in human osteoarthritic synovium

S100A8 exhibited robust immunoreactivity in OA synovial tissue (Fig. 1D), with significantly higher expression compared to controls (p < 0.0001, Fig. 1E). Western blot confirmed elevated S100A8 levels in OA synovium (p < 0.0001, Fig. 1F and G), establishing its pathogenic role.

GO analysis of GSE82107 highlighted S100A8-associated enrichment in inflammatory and immune pathways, so S100A8 was further chosen to elucidate the role in the JAK/STAT3 pathway of osteoarthritis synovial cells, for further investigation and potential therapeutic targeting (Supplementary Fig. 1A). In synovial cells, compared with only IL-6 stimulation to mimic the OA status, S100A8 silencing with IL-6 co-treatment suppressed JAK/STAT3 phosphorylation (p < 0.001), while S100A8 supplementation failed to augment IL-6-induced phosphorylation (p > 0.05). Silencing S100A8 in IL-6 and S100A8-stimulated cells further reduced p-JAK/p-STAT3 (p < 0.05), confirming S100A8 as a critical upstream modulator of JAK/STAT3 signaling in OA (Supplementary Fig. 1B–G).

3.3. Validation of CRISPR-Cas9 mediated S100A8 gene editing efficiency via T7 endonuclease I assay

To evaluate the efficiency of CRISPR-Cas9 mediated gene editing, the riboEDIT T7E1 Enzymes assay was utilized to detect mutations induced by 5 crRNAs targeting the S100A8 gene (Supplementary Fig. 2A). The T7EI assay demonstrated that crRNAs 1, 2, and 3 facilitated effective editing of the target DNA, while crRNAs 4 and 5 failed to show significant editing activity, as indicated by cleavage bands (Supplementary Fig. 2B). Based on these results, crRNA 1 was selected for synthesizing the core CRISPR-Cas9 system to knock out the S100A8 gene.

3.4. Synthesis and Verification of the PAMAM-PLGA (PP) copolymer

To use surface positivity and mitigate the cytotoxicity of cationic PAMAM dendrimers [29], surface amine groups were amidated with PLGA-COOH via carboxyl coupling, forming a PAMAM-PLGA (PP) copolymer according to Kolhatkar et al. [30]. This modification reduced toxicity while retaining nucleic acid-binding capacity through residual cationic charge. ¹H NMR (DMSO/H₂O) confirmed successful copolymerization: PLGA protons (4.90 ppm: CH₂; 5.19 ppm: CH) and PAMAM Ha protons (2.19 ppm) were observed, with a conjugation efficiency of 73.76 %. The PP copolymer exhibited enhanced colloidal stability (zeta potential: +47 mV, Supplementary Fig. 3) and was utilized in subsequent experiments.

3.5. Characterization of PP-Cas9-S100A8@PLGA-apt MPs

Fig. 2A schematically illustrated the synthesis process and physical images of PP-Cas9-S100A8@PLGA-apt MPs. The sequential synthesis involved: (1) Core encapsulation of Cas9-S100A8 complexes within PAMAM-PLGA (PP) matrices, (2) Secondary PLGA coating, and (3) Surface conjugation of OA synovium-targeting aptamers. Zeta potential measurements were used to monitor the surface charge changes throughout the synthesis stages (Fig. 2B). PAMAM dendrimers displayed a highly cationic surfaces (+55.2 to +61.3 mV), while PP copolymer formation reduced surface charge (+39.0 to +55.4 mV) due to carboxyl group neutralization. After binding the Cas9-S100A8 complex, the zeta potential showed markedly the charge inversion (−32.9 to −26.8 mV), confirming successful nucleic acid complexation. Subsequent PLGA encapsulation attenuated the negative zeta potential (−26.6 to −21.6 mV), suggesting charge shielding through PLGA-protein electrostatic interactions. Final OA-targeting aptamer functionalization further decreased zeta potential (−29.8 to −24.9 mV), confirming the attachment of the negatively charged aptamer. Spherical particles with size exceeding the former particles during formulation, indicating additional components were added successfully (Fig. 2C).

Fig. 2.

Synthesis and characterization of PP-Cas9-S100A8@PLGA microparticles. A, schematic illustration of the synthesis process for PP-Cas9-S100A8@PLGA microparticles. B, zeta potential monitoring of different microparticle formulations. C, particle size distribution of the synthesized microparticles. D, FTIR spectra of the different microparticle formulations. E, SEM images of the synthesized microparticles. F, agarose gel electrophoresis of mRNA encapsulation efficiency. G, analysis of different microparticles release over 8 days. SEM, Scanning electron microscopy. FTIR, Fourier-transform infrared.

FTIR spectroscopy further demonstrated additional spectral features of PP-Cas9-S100A8@PLGA-apt MPs (Fig. 2D), which confirmed the successful attachment of the aptamer compared to other MPs. No significant spectral difference (p > 0.05) was found among three MPs, indicating structural consistency during synthesis. Scanning electron microscopy (SEM) images (Fig. 2E) showed PP-Cas9-S100A8@PLGA-apt MPs were uniform.

Agarose gel retardation assays (Fig. 2F) demonstrated formulation-dependent Cas9-S100A8 mRNA binding efficiency. Naked mRNA migrated freely towards the positive electrode (lanes 1 and 2), indicating their electrophoretic mobility. PP copolymer achieved complete nucleic acid retardation at the ratio of 1:5 and 1:10 (PP: Cas9-S100A8, lanes 4 and 5), with the latter showing a stronger retardation effect, while mRNA degradation was shown at low ratio (lane 3). Higher ratios of PP copolymer (lanes 6–8) provided no additional retardation. Therefore, we chose 1:10 as the optimal ratio for stable binding.

The PP nanocores exhibited sustained mRNA release over 8 days (Fig. 2G), achieving significantly higher cumulative release at later timepoints (days 7–8) compared to PAMAM-Cas9-S100A8@PLGA (p < 0.001). This result indicated that PP shell ensured particle stability, thereby enabling sustained therapeutic efficacy in cells.

3.6. Functional effects on synovial cells of PP-Cas9-S100A8@PLGA-apt MPs

Confocal microscopy revealed that synovial cells exhibited strong dual red and green fluorescence, despite chondrocytes showing weak red and green fluorescence, and meniscal as well as HepG2 cells displaying only minimal green fluorescence. This strongly confirmed the specificity of PP-Cas9-S100A8@PLGA-apt binding to synovial cells, although it might be affected by cellular non-specific endocytosis or extracellular environment, leading to the inherent less-than-100 % specificity (Fig. 3A). Both FAM-tagged Cas9-S100A8 and the CY5.5-tagged aptamer were taken up significantly by synovial cells, indicating its potential for targeted therapeutic applications, particularly in OA synovium.

Fig. 3.

Functional effects on synovial cells of PP-Cas9-S100A8@PLGA-Apt microparticles.

A, confocal microscopy images showing the internalization of PP-Cas9-S100A8@PLGA-Apt microparticles in synovial cells, chondrocytes, meniscal cells, and HepG2 cells. B, confocal microscopy images of LysoTracker staining to assess lysosomal localization of Cy5-mRNA and PP-Cas9-S100A8@PLGA-Apt microparticles. a, cells were treated with naked Cy5-mRNA. b, c, and d, PP-Cas9-S100A8@PLGA-Apt microparticles at different time points (8h, 12h, 24h). The red fluorescence indicated Cy5-mRNA, and the green fluorescence indicated lysosomes. Co-localization (yellow) confirmed lysosomal trafficking of the microparticles. C, flow cytometry analysis of synovial cellular uptake of PP-Cas9-S100A8@PLGA-Apt microparticles with different concentrations (0.1 μg/ml to 0.8 μg/ml). The histograms displayed the percentage of cells that have taken up the microparticles, with increased uptake observed at higher concentrations and in the presence of Lip2K. D, scratch assay and analysis depicting migration of synovial cells treated with different microparticles (magnification = 4 × ). Scale bar: 200 μm. Representative of three independent experiments (each with 3 technical replicates). E, matrigel transwell assay and related analysis of synovial cells treated with different microparticles. (magnification = 4 × ). Scale bar: 200 μm. Representative of three independent experiments (each with 3 technical replicates). F, cell viability assay of synovial cells treated with different microparticles over 24, 48, and 72 h. Representative of three independent experiments (each with 3 technical replicates). ∗, p < 0.05, ∗∗, p < 0.01, ∗∗∗, p < 0.001, ∗∗∗∗, p < 0.0001, ns, not significant. Lip2K, Lipofectamine 2000. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

Lysosomal colocalization studies revealed that naked Cy5-mRNA controls exhibited persistent lysosomal sequestration (Fig. 3B: a). In contrast, the distinct separation of the Cy5 signal from the Lysotracker by 24 h was observed, indicating that PP-Cas9-S100A8@PLGA-apt MPs achieved significant endosomal escape after initially colocalizing with lysosomal compartments (Fig. 3B: b-d).

Flow cytometry (Fig. 3C) showed that 99.8 % synovial cells were unstained in the control group. In the contrast, PP-Cas9-S100A8@PLGA-apt MPs exhibited concentration-dependent uptake in synovial cells: 13.2 % at 0.1 μg/ml, 17.9 % at 0.2 μg/ml, followed by a sharp rise to 67.7 % at 0.4 μg/ml, culminating in near-saturating uptake (98.8 %) at 0.8 μg/ml. Co-administration with Lip2K at 0.8 μg/mL yielded marginal enhancement (99.4 %), establishing this concentration as optimal for maximal transfection efficiency. These findings established 0.8 μg/ml as the threshold concentration for maximal in vitro transfection efficiency.

Additionally, to assess the potential impact of MPs on the biological activity of synovial cells, we performed a wound healing scratch assay and transwell migration assay. The results showed significant difference in migration rates between synovial cells treated with PP-Cas9-S100A8@PLGA-apt and those treated with other MPs (Fig. 3D and E). These suggested that PP-Cas9-S100A8@PLGA-apt could strongly reduce intra-articular and transendothelial migration capacity of synovial cells.

Longitudinal cytotoxicity assessment (72 h) revealed PP-Cas9-S100A8@PLGA-apt MPs maintained >80 % synovial cell viability across all timepoints (Fig. 3F). In contrast, PAMAM-Cas9-S100A8 exhibited a significant reduction in cell viability after 24 h of incubation, showing that PLGA encapsulation and PP core synergistically mitigated cytotoxicity and preserved cell viability, supporting in vivo applications.

3.7. PP-Cas9-S100A8@PLGA-apt MPs sustained in vitro and vivo bioavailability

In in vitro experiment, FAM-labelled tracking demonstrated that PP-Cas9-S100A8@PLGA-apt achieved 64.39 % uptake efficiency at 24 h, significantly outperforming other formulations (10.31 %, 16.99 %, and 43.67 % for PAMAM-Cas9-S100A8, PAMAM-Cas9-S100A8@PLGA, and PP-Cas9-S100A8@PLGA, respectively). Moreover, sustained retention (71.46 % at 48 h, 39.7 % at 72 h) highlighted aptamer-enhanced cargo delivery (Fig. 4A and B).

Fig. 4.

In vitro and in vivo evaluation of FAM and CY7-labelled Cas9-S100A8 microparticles. A, immunofluorescence staining of FAM-labelled Cas9-S100A8 microparticles in synovial cells was performed to detect FAM-labelled microparticles (green) and nuclei (blue). The images showed successful delivery and expression of the labelled microparticles in treated cells, with a notable increase in fluorescence intensity. B, quantitative analysis of mean fluorescence intensity of FAM-labelled microparticles in synovial cells, representing FAM-labelled Cas9-S100A8 expression levels. Data represent 3 biological replicates (each with 3 technical replicates). C, quantitative analysis of bioluminescence intensity in mice treated with different CY7-labelled Cas9-S100A8 microparticles at different time points. Data represent 3 biological replicates. D, in vivo bioluminescence imaging of mice treated with different CY7-labelled Cas9-S100A8 microparticles in left knee immediately and 96h after injection. ∗, p < 0.05, ∗∗, p < 0.01, ∗∗∗, p < 0.001, ∗∗∗∗, p < 0.0001, ns, not significant. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

In vivo fluorescence of different Cy7-labelled Cas9-S100A8 MPs administered via intra-articular injection in mice revealed distribution and retention over time (Fig. 4C and D). PP-Cas9-S100A8@PLGA-apt MPs demonstrated the highest efficiency in internalization, with an initial mean fluorescence intensity of 26.92 immediately after administration, compared to PP-Cas9-S100A8@PLGA (mean intensity ∼14.93), PAMAM-Cas9-S100A8@PLGA (mean intensity ∼8.83), and PAMAM-Cas9-S100A8 (mean intensity ∼10.70). At 96 h post-administration, PP-Cas9-S100A8@PLGA-apt MPs maintained the highest fluorescence intensity with a mean of approximately 16.15, although a decrease from the initial measurement was observed. In contrast, the other MPs exhibited reduced intensities over time: PP-Cas9-S100A8@PLGA had a mean intensity of ∼8.13, PAMAM-Cas9-S100A8@PLGA ∼3.41, and PAMAM-Cas9-S100A8 ∼2.37. These findings highlighted the enhanced cellular uptake and prolonged presence of PP-Cas9-S100A8@PLGA-apt MPs in vivo, indicating their potential for sustained therapeutic delivery.

3.8. Longitudinal analysis of inflammatory markers and osteophyte formation in mice OA models following treatment with PP-Cas9-S100A8@PLGA-apt MPs

Intra-articular administration of PP-Cas9-S100A8@PLGA-apt MPs (0.8 μg/mL, 10 μL) in ACLT-induced OA mice demonstrated its therapeutic efficacy by immunohistochemical analysis (Fig. 5A). Following intra-articular injection of MPs, acute post-surgical inflammation triggered upregulation of S100A8 at 2 weeks (30.64 % vs. sham 13.29 %, p < 0.0001), which progressively declined at 4 weeks (21.02 %, p = 0.004) and further decreased by 16 weeks without statistical significance (Fig. 5B). JAK and STAT3 expression followed a similar trend, peaking at 2 weeks after MPs injection (p < 0.0001 vs. sham) but attenuating at 4 weeks with sustained suppression at 8 and 16 weeks, confirming synovitis suppression (Fig. 5C and D). The lack of significant difference of expressions of these inflammatory markers between 4 and 8 weeks (p = 0.3002) suggested a transient inflammatory plateau, while marked reductions at these time points highlighted therapeutic onset.

Fig. 5.

Multifaceted impact of PP-Cas9-S100A8@PLGA-apt microparticles on synovitis markers of OA in ACLT mice model. A, immunohistochemical staining for S100A8, JAK, and STAT3 and formation of osteophytes (highlighted in red) at 2, 4, 8, and 16 weeks post-ACLT. B-D, immunohistochemical analysis of S100A8, JAK, and STAT3 in synovial tissues from different treatment groups. The bar graphs represented the densitometric analysis of the immunohistochemical staining results, indicating relative expression levels. E-F, quantitative analysis of subchondral bone volume and osteophyte formation in the knee joints of different treatment groups. G, representative Western blots of S100A8, JAK, STAT3, and MMP13 displayed protein expression at the each post-ACLT time points, with Actin as a loading control. H, densitometric quantification of Western blot bands relative to Actin, with corresponding statistical analysis. ∗, p < 0.05, ∗∗, p < 0.01, ∗∗∗, p < 0.001, ∗∗∗∗, p < 0.0001, ns, not significant. ACLT, anterior cruciate ligament transection. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

Micro-CT analysis demonstrated that PP-Cas9-S100A8@PLGA-apt MPs significantly attenuated osteophyte formation in ACLT-induced OA mice compared to the blank group (p < 0.0001), although osteophyte volume still increased in the treatment groups (Fig. 5E). Treated mice exhibited lower osteophyte scores than blank controls at all timepoints (from 2 to 8 weeks: p < 0.001; 16 weeks: p < 0.0001, Fig. 5F), confirming sustained structural disease modification.

Western blot analysis revealed that the therapeutic effects of PP-Cas9-S100A8@PLGA-apt MPs in ACLT mice were significant in the late phase (Fig. 5G and H). S100A8 expression was significantly suppressed at 8 and 16 weeks post-surgery (p < 0.0001), with no early difference (at 2 and 4 weeks) versus blank controls. Expression of JAK, STAT3, and MMP13, initially comparable to those in blank controls, declined significantly by 16 weeks (p < 0.0001, p = 0.0163, and p < 0.0001, respectively), reflecting gradual normalization. These time-dependent biomarker shifts underscored the treatment's ability to uncouple acute inflammation from chronic degeneration.

3.9. Therapeutic modulation of cartilage homeostasis and subchondral bone density in osteoarthritis via PP-Cas9-S100A8@PLGA-apt MPs

Building upon previous results, the therapeutic potential of PP-Cas9-S100A8@PLGA-apt MPs on articular cartilage was further assessed in mouse OA models (Fig. 6A). Longitudinal OARSI scoring demonstrated that PP-Cas9-S100A8@PLGA-apt MPs mitigated cartilage degradation in ACLT-induced OA mice. The treatment group had acute cartilage loss at 2 weeks (OARSI score 4.2, p < 0.0001 vs. sham) but showed no significant difference compared to the blank group (p = 0.1927 vs. blank). Progressive chondroprotection emerged from 4 to 16 weeks (p < 0.0001 vs. blank), confirming sustained therapeutic efficacy (Fig. 6B).

Fig. 6.

Evaluation of cartilage and subchondral bone in mouse OA model treated with PP-Cas9-S100A8@PLGA-apt microparticles. A, safranin-O staining, immunofluorescence and micro-CT reconstructions of cartilage and subchondral bone of knee joint sections from sham, blank, and PP-Cas9-S100A8@PLGA-apt-treated ACLT groups at 2, 4, 8, and 16 weeks post-ACLT. B, OARSI scoring of joint cartilage degradation of structural changes. C-E, Aggrecan, MMP13, and Collagen II intensity quantification. F, BV/TV ratio determination from micro-CT images of volumetric changes in the subchondral bone. G, analysis of Tb.Th. in subchondral bone post ACLT. H, Western blot of protein expression levels of MMP3, Aggrecan, MMP13, and Collagen II in cartilage. I, quantitative analysis of protein expression in osteoarthritis synovial cells post-treatment corresponding to Western blot analysis. ∗, p < 0.05, ∗∗, p < 0.01, ∗∗∗, p < 0.001, ∗∗∗∗, p < 0.0001, ns, not significant. OA, osteoarthritis. ACLT, anterior cruciate ligament transection. Tb.Th., trabecular thickness.

Aggrecan expression on articular cartilage remained low and stable during early OA stages (ACLT 2–4 weeks) but was markedly upregulated at 8–16 weeks (p < 0.0001 vs. blank, Fig. 6C). Collagen II, initially reduced at ACLT 2 weeks (p < 0.0001 vs. sham), rebounded at ACLT 4–8 weeks (p < 0.0001, Fig. 6D). MMP13-mediated degradation peaked acutely at ACLT 2 weeks (p < 0.0001) but progressively attenuated thereafter (p < 0.0001, Fig. 6E). Subchondral bone density (BV/TV ratio) transiently increased at ACLT 2 weeks (p < 0.0001 vs. sham), declined at ACLT 4–8 weeks (p < 0.0001), and normalized to sham levels by ACLT 16 weeks (p = 0.9970), reflecting treatment-mediated bone remodeling (Fig. 6F). Trabecular thickness (Tb.Th.) restoration (Fig. 6G), which reflected subchondral bone repair, exhibited delayed therapeutic effects. Significant improvement emerged at ACLT 8 weeks (p < 0.05), with complete normalization to sham levels by ACLT 16 weeks (p = 0.9645), despite no early-phase differences (ACLT 2–4 weeks). These findings indicated that PP-Cas9-S100A8@PLGA-apt MPs restored trabecular architecture during OA progression.

Western blot analysis further revealed that PP-Cas9-S100A8@PLGA-apt MPs reprogrammed OA cartilage metabolism after ACLT (Fig. 6H and I). MMP3 was consistently downregulated across all timepoints, with maximal suppression at 16 weeks (p < 0.0001). Aggrecan exhibited sustained upregulation (p < 0.0001) through 16 weeks, while MMP13, initially elevated post-ACLT, declined significantly by ACLT 16 weeks (p < 0.0001). Collagen II levels progressively increased, peaking at ACLT 16 weeks (p < 0.0001 vs. blank). These coordinated shifts—suppressed catabolism (MMP3/MMP13) and enhanced anabolism (Aggrecan/Collagen II)—highlighted the capacity of PP-Cas9-S100A8@PLGA-apt MPs to redirect cartilage degeneration toward regeneration.

3.10. Restorative H-type vascular channels in subchondral bone and gait analysis following ACLT and PP-Cas9-S100A8@PLGA-apt treatment

Immunofluorescence analysis of CD31+/EMCN+ revealed PP-Cas9-S100A8@PLGA-apt MPs restored physiological H-type vascular density in subchondral bone (Fig. 7A and B). The blank group exhibited pathological vascular escalation (p < 0.0001 vs. sham) shown by increased CD31+ and EMCN+ colocalization, while treated mice showed progressive normalization: acute-phase reduction at 2 weeks, sustained attenuation at 8 weeks (p < 0.0001 vs. blank), and near-sham levels by 16 weeks (p = 0.207). These findings indicated therapeutic reversal of OA-associated angiogenic dysregulation.

Fig. 7.

A, immunofluorescence of H-type vascular channels in subchondral bone. B, analysis of H-type vascular channels within the subchondral bone, as determined by immunofluorescence staining for CD31+ and Endomucin+, over a 16-week period following ACLT. ∗, p < 0.05, ∗∗, p < 0.01, ∗∗∗, p < 0.001, ∗∗∗∗, p < 0.0001, ns, not significant. C and D, gait analysis and quantification of stride length of mice in different groups (n = 10 per group). Red print: fore paw; blue print: hind paw. ACLT, anterior cruciate ligament transection. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

Furthermore, we performed gait analysis to assess the recovery of joint motor function. The results demonstrated that the stride length significantly improved after treatment with PP-Cas9-S100A8@PLGA-apt MPs and, over time, became comparable to that of the sham group (Fig. 7C and D). These findings indicated that PP-Cas9-S100A8@PLGA-apt MPs markedly alleviated claudication in OA mice.

4. Discussion

The principal finding of our study was the validation of S100A8 as a central mediator of OA pathogenesis, with CRISPR-Cas9-mediated gene silencing demonstrating potent anti-inflammatory and chondroprotective effects. The engineered PP-Cas9-S100A8@PLGA-apt microparticles represented a paradigm shift in gene editing delivery, combining PAMAM dendrimers' nucleic acid condensing capacity with PLGA's biodegradability and aptamer-mediated synovial targeting. This layer-by-layer design achieved 64.39 % transfection efficiency in synovial cells (Fig. 4B), while maintaining endosomal escape enhancement and over 80 % cell viability through surface charge modulation (Fig. 3B and F). This formulation effectively mitigated synovial inflammation, cartilage degeneration, subchondral bone disarray, and gait impairment in OA by modulating the JAK/STAT3 signaling pathway (Fig. 5, Fig. 6, Fig. 7), underscoring the therapeutic potential of precision-targeted gene editing in ameliorating the pathological manifestations of OA.

Extensive literature has established the significance of the S100A8 protein in the inflammatory pathology. S100A8, often in a complex with S100A9, is recognized as a critical factor in periodontal inflammatory diseases, serving as a potential biomarker for diagnosis and the assessment of treatment efficacy [31]. Its role extends to neuroinflammatory conditions, particularly in cases of traumatic brain injury and age-associated disorders, where it is implicated in the inflammatory amyloid cascade [32]. Moreover, the S100A8/A9 complex has emerged as an essential biomarker in SARS-CoV-2 infection, displaying a dual function by both eliciting inflammatory responses and inhibiting HIV replication in macrophages [33]. Collectively, the pleiotropic regulatory capacity of S100A8 in orchestrating inflammatory cascades through its modulation of alarmin signaling has been delineated across various inflammatory diseases. Similarly, in OA pathogenesis, S100A8 has been identified as a potential biomarker and therapeutic target through a comprehensive approach that combined database interrogation with proteomic analysis [34]. Elevated S100A8 levels were associated with synovitis severity (Fig. 1D–G), cartilage remodeling, and osteophyte formation in OA [16]. It is also integral to inflammatory pathways, interfacing with NF-κB [35], TLR4, and RAGE [36], highlighting it as a potential target for treating inflammatory joint conditions. Further investigation has demonstrated that S100A8 inhibition could attenuate synovial inflammation in OA and suppress phosphorylation of JAK and STAT3 in the pathway (Supplementary Fig. 1B–G). It revealed the involvement of the JAK/STAT3 pathway and suggested that the alarmin S100A8 might play an intermediary role in modulating inflammation.

Emerging evidence has demonstrated the therapeutic potential of S100A8/A9 heterodimer inhibition across diverse pathophysiological contexts and the knockout of the S100A8/A9 complex leading to notable changes in disease pathology and cellular responses. For instance, Li et al. [37] demonstrated that the knockout of S100A9, a component of the S100A8/A9 complex, markedly decreased cardiomyocyte death and improved heart function in myocardial ischemia-reperfusion injury, suggesting a detrimental role of S100A8/A9 in cardiac health. Additionally, in neuroinflammatory disorders, Ma et al. [38] reported that S100A8/A9 promotes the secretion of proinflammatory cytokines in microglial cells, implicating its role in neuroinflammation. Similarly, osteoarthritis studies by van Lent et al. [39] found that the absence of S100A8/A9 reduced synovial activation and cartilage destruction, indicating its pathogenic role in joint disease. Given the crucial role of S100A8/A9 inhibition in disease modulation and its therapeutic potential, we designed the delivery system of CRISPR-Cas9-mediated S100A8 gene knockout successfully. The downregulation of molecules inducing inflammation in the JAK/STAT3 pathway through mice OA models was aligned with the vitro experiments after treatment of PP-Cas9-S100A8@PLGA-apt microparticles (Fig. 5 and Supplementary Fig. 1). These findings not only validated the therapeutic efficacy of CRISPR-Cas9-mediated S100A8 ablation but also reaffirmed its pivotal role as a molecular linchpin orchestrating inflammatory cascades in OA pathogenesis, positioning it as a critical regulatory node across inflammatory arthropathies.

However, the generation of clonal knockout populations in mammalian systems still faces persistent technical hurdles, including incomplete gene ablation and functional intraclonal heterogeneity arising from stochastic CRISPR activity [40]. Moreover, dynamic transcriptional regulation further complicates precise genomic editing outcome, particularly for inflammation-responsive genes like S100A8 governed by transcription factors and epigenetic modifications [41]. While corroborating the findings of Yuen et al. [42] and highlighting significant differences in the knockout efficiency of various sgRNA design sequences, the efficiency of CRISPR-Cas9 mediated gene editing in our study achieved a 65 % knockdown efficiency of S100A8 using sgRNA-1(Supplementary Fig. 3). Furthermore, the specificity of the CRISPR-Cas9 system and the potential off-target effects must be considered, as these remain paramount concerns. [43]. To solve this, PP-Cas9-S100A8@PLGA-apt microparticles were conjugated with synovial-targeting aptamers for selective uptake by synovial cells (Fig. 3A).

While achieving high S100A8 knockdown efficiency via sgRNA-1 demonstrated substantial gene-editing capability, the study aimed to enhance therapeutic outcomes by using a PP core and PLGA encapsulation, thereby exploring novel therapeutic potential. To address cytotoxicity from higher-generation PAMAM dendrimers, we modified them with PLGA to balance biocompatibility with delivery efficacy. The integration of PAMAM dendrimers with FDA-approved PLGA to form PP copolymers yielded a composite that enhanced endolysosomal escape (Fig. 3B) and enabled sustained gene release (Fig. 4), as demonstrated in gene transfection assays [44]. Encapsulation in PLGA leveraged its biocompatibility and biodegradability, protecting CRISPR-Cas9 from degradation while ensuring controlled release. Aptamer functionalization further enabled synovium-specific targeting, achieving selective uptake by synovial cells (Fig. 3A) and prolonging in vivo retention, with 60 % residual fluorescence remaining at 96 h post-injection (Fig. 4C). Studies by Catuogno et al. [45] and others validate the aptamer-functionalization strategy, highlighting its efficiency in enhancing the delivery and imaging of therapeutic drugs. Therefore, the PP-Cas-S100A8@PLGA-apt not only increased targeted uptake by synovial cell, but also enhanced the knockout efficiency in these cells.

While this study provided important insights into the role of the S100A8 gene in OA and demonstrated the therapeutic potential of targeted gene editing, it still had some limitations. Firstly, although high knockdown efficiency of S100A8 was achieved using sgRNA-1, incomplete ablation of targeted gene could still necessitate some conventional therapies in treating OA. Secondly, while the PP-Cas9-S100A8@PLGA-apt MPs demonstrated high targeting specificity toward synovial cells by using aptamer-enhanced delivery systems, achieving absolute cell-type selectivity remained challenging in complex biological environments since they could lead to unintended genetic modifications with unknown biological consequences. Additionally, differences in the delivery and expression of CRISPR-Cas9 components among various cell types within the synovial membrane might lead to heterogeneity in editing efficiency, requiring further research. Moreover, interspecies divergence in OA progression and patterns might limit direct clinical extrapolation. The translational potential of these findings will require rigorous clinical testing to ensure safety, efficacy, and therapeutic relevance in human patients.

5. Conclusion

This study elucidating the role of S100A8 in OA via CRISPR-Cas9-mediated gene editing demonstrated the potential of this therapeutic intervention. Use of the PP-Cas9-S100A8@PLGA-apt delivery system showed promising results, including synovial-targeting specificity, reduced cytotoxicity, and improved editing efficiency, which are crucial for alleviating OA-related inflammation and degeneration. This research enhanced the therapeutic potential and provided new insights for precision medicine to treat OA and related inflammatory diseases.

CRediT authorship contribution statement

Zhong Chen: Writing – review & editing, Writing – original draft, Formal analysis, Data curation, Conceptualization. Miaomiao Zhou: Writing – original draft, Methodology, Formal analysis, Data curation, Conceptualization. Jialong Luo: Writing – review & editing, Writing – original draft, Formal analysis, Data curation, Conceptualization. Changchuan Li: Software, Methodology, Formal analysis. Shixun Li: Resources, Methodology, Data curation. Sipeng Lin: Resources, Methodology. Xinghao Deng: Software, Resources, Methodology, Data curation. KeLong Zhu: Writing – original draft, Validation, Supervision, Resources, Project administration, Conceptualization. Yue Ding: Writing – review & editing, Writing – original draft, Visualization, Validation, Supervision, Project administration, Funding acquisition, Conceptualization.

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 is the Supplementary data to this article:

Data availability

The authors are unable or have chosen not to specify which data has been used.