Synovial macrophage-targeted hierarchical microspheres achieve cartilage preservation and pain relief in osteoarthritis Authors: Guojiang Zhang, Laibo Zhang, Xiangdong Wen, Guangxian Liu, Dengju Li, Chuanhong Qin, Senbo An, Yawei Sun, Shui Sun Published: 2026 ABSTRACT AbstractPain, the hallmark symptom of osteoarthritis (OA), drives most patients to seek medical care. The current analgesics only provide partial pain relief, while fail to modify disease progression and introduce adverse effects with long-term use. Therefore, elucidating the underlying causes of OA-related pain and developing etiology-targeted therapeutics are urgently needed. In this research, we discovered that M1 macrophages and Netrin-1 are highly co-located in the synovium of both OA patients and anterior cruciate ligament transection (ACLT) mice. Moreover, depletion of macrophages in the knee of ACLT mice significantly alleviated OA pain. Thus, it is hypothesized that modulating synovial macrophage polarization may alleviate OA pain by regulating Netrin-1. Based on these findings, we designed Tof@HA-TAGel—an injectable hydrogel microsphere that selectively targets synovial macrophages and enables the sustained release of both tannic acid (TA) and tofacitinib (Tof). Experimental results demonstrated that Tof@HA-TAGel scavenges reactive oxygen species (ROS), suppress M1 macrophage polarization and attenuates cytokine storm through downregulating the JAK–STAT pathway, thereby reducing Netrin-1 secretion and inhibiting sensory nerve ingrowth. In ACLT-induced OA models, the microsphere significantly attenuated cartilage degradation. Behavioral assessments (Gait analysis, Von Frey tests) confirmed pain alleviation and functional recovery. Immunofluorescence further revealed suppressed M1 infiltration, diminished Netrin-1, and reduced Calcitonin Gene-Related Peptide (CGRP) nerve ingrowth in synovium. Collectively, Tof@HA-TAGel integrates anti-inflammatory, antioxidant, and analgesic mechanisms. This multifunctional system mitigates OA pain by inhibiting synovial macrophage polarization and subsequent sensory nerve ingrowth while protecting cartilage concurrently. It positions a promising disease-modifying biomaterial for etiology-aware OA therapy.Keywords: Osteoarthritis, Pain, Macrophage polarization, Microsphere, Synovium-targeting Introduction As a leading cause of chronic pain and physical disability,OA impacts an estimated 250 million individuals globally [1]. The growing prevalence of OA, driven by demographic aging, causes the substantial individual and socioeconomic burden and presents a critical challenge to global public health systems [2]. OA is a whole-joint disorder involving multiple articular tissues, driven by systemic and local factors. Its pathogenesis is driven by a multifaceted network of mechanisms encompassing mechanical stress, inflammatory responses, aging, and metabolic dysregulation, ultimately resulting in the hallmark triad of joint structural breakdown, functional decline, and chronic pain [3]. As the dominant clinical symptom, pain remains the key driver for medical consultation. Current first-line analgesics, such as NSAIDs, acetaminophen, and opioids, offer limited efficacy due to unresolved pain mechanisms, while long-term use poses risks of significant adverse effects and toxicity [4,5]. These limitations highlight an urgent need to develop disease-modifying OA drugs (DMOADs) that target the underlying mechanisms of pain and can concurrently provide sustained analgesia, promote structural improvement, and mitigate disease progression. OA is pathologically characterized by cartilage degeneration, while the avascular and aneural nature of cartilage excludes it as a direct pain source. Instead, pain likely originates from richly innervated peri-articular structures, including synovium, fat pad, ligaments, joint capsule, and subchondral bone [6]. The synovium consists of two distinct tissue layers responsible for producing and retaining synovial fluid within the joint. The inner layer, adjacent to the joint cavity, is composed of macrophages and fibroblast-like synoviocytes. The outer layer, known as the sublining, comprises fibrous connective tissue and blood vessels, with sparse populations of lymphocytes and macrophages. The imbalance between pro-inflammatory M1 and anti-inflammatory M2 phenotypes is one of the most important characteristics in the intimal lining of the OA synovium [7]. Growing evidence links synovitis to symptomatic pain and structural progression of OA [8]. Synovial inflammation is implicated in driving peripheral and central sensitization in OA. Given this link to pain severity and the transition to chronicity, targeting synovial inflammation to subsequently modulate sensitization represents a promising, disease-modifying strategy to alleviate OA pain [9]. Macrophage polarization is orchestrated by a complex network of signaling pathways, including Notch, JAK/STAT, MAPK, and NF-κB [10]. Tof, a novel JAK inhibitor, specifically targets intracellular JAK enzymes that mediate cytokine signaling. By potently inhibiting JAK1 and JAK3 activity, this agent effectively blocks M1 macrophage polarization and the production of pro-inflammatory cytokines such as IL-6 and TNF-α, thereby disrupting the inflammatory cascade [11]. Additionally, it was also observed that the application of Tof conferred a modest degree of pain relief. Clinical evidence supports the efficacy of Tof in alleviating pain in patients with rheumatoid arthritis (RA) and ankylosing spondylitis [12,13]. Based on this observation, we sought to determine if dysregulated macrophage polarization directly contributes to OA pain—a central hypothesis that also examines Tof's potential to exert analgesic effects via polarization modulation. Netrin-1 was originally identified as a diffusible factor released by neural floor plate cells in the developing spinal cord that regulates axonal outgrowth [[14], [15], [16]]. Recently, studies have expanded to peripheral organs where netrin-1 modulates cell development, migration, nerve regeneration and inflammatory [17,18]. It is reported that netrin-1 may be involved in the pathogenesis of various inflammatory bone diseases, such as OA and RA [19]. Our prior work identified that osteoclast-derived Netrin-1 promotes subchondral sensory innervation, correlating with pain perception in osteoarthritis [20,21]. However, whether Netrin-1 mediates neural infiltration in synovial tissue remains unknown. Notably, evidence from multiple studies establishes the coexistence of netrin-1 and CD68 (a macrophage marker) in various inflammatory milieus, including atherosclerotic plaques [22], adipose tissues [23] and inflamed aortas [24]. This colocalization implies macrophage-derived Netrin-1 secretion under inflammatory conditions. Building on this paradigm, we propose a macrophage-targeted Tof delivery system to investigate the interplay between macrophage polarization, netrin-1 secretion, and netrin-1 mediated neurite growth, thereby aiming for dual therapeutic efficacy against both structural degeneration and pain in osteoarthritis. In this study, we commenced by investigating the association between M1 macrophage and Netrin-1 expression in the synovium, to elucidate the critical role of macrophages in OA pain. Based on this finding, the injectable microspheres (Tof@HA-TAGel) composed of TA and gelatin methacryloyl (GelMA) were prepared for delivering Tof. Hyaluronic acid (HA) surface modification was used for precise M1 macrophage targeting [25,26]. The preparation and mechanism of the Tof@HA-TAGel are described in Scheme 1. The in vitro and in vivo studies verified that the composite microspheres Tof@HA-TAGel achieved efficient M1 macrophage and synovium targeting, downregulate the expression of Netrin-1 via JAK1/STAT3 pathway and subsequent suppression of sensory nerve ingrowth, as well as notably ameliorate mechanical allodynia and a concomitant improvement in motor coordination of OA mice, following an 8-week regimen of Tof@HA-TAGel. Overall, the macrophage-targeting microspheres present a promising minimally-invasive strategy for OA, exerting therapeutic effects by simultaneously modulating M1 polarization, synovitis, pain, and cartilage degradation. Schematic diagram of hydrogel microspheres (Tof@HA-TAGel). (A) The preparation process of Tof@HA-TAGel microspheres. (B) Following intra-articular injection into the knee joint, Tof@HA-TAGel combats arthritic pain and mitigates osteoarthritis progression through sustained, site-specific targeting of synovial M1 macrophages, local antioxidative effects, and downregulation of the JAK1/STAT3 pathway, thereby suppressing synovitis, decreasing Netrin-1 secretion, and inhibiting sensory nerve ingrowth. Experimental section Materials TA (16201) was purchased from Sigma-Aldrich; Gelatin methacryloyl (GelMA; WT001) was obtained from Engineering for Life; Tof (HY-40354AR) was sourced from MedChemExpress. Cell culture reagents, including Dulbecco's Modified Eagle Medium/Nutrient Mixture F-12 (DMEM/F-12; C11330500BT), Neurobasal Medium (21103049), fetal bovine serum (FBS; 10099-141C), GlutaMAX (35050061), B-27 neuronal supplement (A3653401), and penicillin/streptomycin (15140122), were obtained from Gibco. Recombinant NGF (256-GF-100) was procured from R&D Systems. Synthesis of Tof@HA-TAGel microsphere Microspheres were fabricated using a high-throughput homogenization-based organoid gel microsphere preparation chip (YX-XP-WQ1001). The oil phase consisted of mineral oil containing 2% (w/v) Span 80, flowing at 150 μL/min. The aqueous phase was GelMA (40 mg/ml)and Tof (50 nM) solution, delivered at 30 μL/min. The emerging microdroplets were immediately exposed to UV light (405 nm, 25 mW/cm2) at the tubing outlet for photopolymerization. The collected microspheres were then sequentially processed by filtering through a cell strainer to remove the oil phase, followed by washing once with 50% ethanol and twice with pure water. The resulting purified microspheres were lyophilized. Finally, The Tof@HA-TAGel microspheres were obtained by sequentially soaking the lyophilized microspheres first in TA (2 mg/ml) and then in HA (2 mg/ml) aqueous solutions at a 20:1:1 mass ratio, which was based on preliminary experiments that showed this composition could produce microsphers with a high loading efficiency, uniform size and homogeneous porosity. Characterization of Tof@HA-TAGel The Tof@HA-TAGel microspheres were prepared using microfluidic technology. Morphological characterization and size distribution were first determined statistically from optical micrographs (Olympus BX53) of over 100 individual microspheres. To probe the internal microstructure, the samples were subsequently lyophilized, gold-sputtered, and then imaged by scanning electron microscopy (SEM, SU-8010) at a 5 keV accelerating voltage. The lubricating performance was further evaluated through ball-on-plate tribological tests (Bruker UMT-5) under a 12 N normal load, while thermal stability was assessed by thermogravimetric analysis (Mettler Toledo) with heating from 50 °C to 800 °C at 10 °C min−1. Material chemistry was analyzed by Fourier-transform infrared (FTIR, Nicolet iS20). Freshly prepared microspheres dispersions were centrifuge-filtered with a 10 kD membrane filter (4000 rpm, 20 min) to separate non-encapsulated drug (in the filtrate).For TA analysis, the absorbance of the resulting solution was determined using an UV-Vis spectrophotometer (Cary 5000, Agilent) at 278 nm. The mass of TA in the filtrate was calculated based on a calibration curve. The TA loading efficiency was then calculated: Here, Mf and Mt are the masses of TA in the filtrate and in the total system, respectively. The Tof loading efficiency was determined using a high performance liquid chromatography (HPLC) method [27]. The Tof loading efficiency was calculated using the following equation: Here, Mf and Mt are the masses of Tof in the filtrate and in the total system, respectively. For the TA release study, samples in saline (3 mL) were incubated at 37 °C in the dark, and the solution was periodically sampled (1 mL) for concentration measurement using an UV-Vis spectrophotometer against a standard curve. Cell culture RAW 264.7 cell culture The macrophage model used in this study was the RAW 264.7 cell line (Procell CL-0190), obtained from Procell Life Science&Technology Co,.Ltd. Cells were maintained in DMEM supplemented with 10% FBS, 100 U/mL penicillin, and 100 μg/mL streptomycin at 37 °C in a 5% CO2 atmosphere. Macrophages were stimulated with 500 ng/mL LPS for 24 h to induce polarization to an M1 phenotype. Cell growth was monitored via bright-field microscopy, and subculturing was performed at a 1:4 to 1:5 ratio when cells reached 80–90% confluence. RAW 264.7 cells exhibit semi-adherent growth and were gently dislodged during passaging before inoculation into fresh medium. Due to their rapid proliferation, the cells were passaged every 2–3 days. Dorsal root ganglion (DRG) obtained and cultured Dorsal root ganglion (DRG) were extracted from 8-week-old postnatal mice under a stereomicroscope. They were cultured in serum-free culture medium containing neurobasal medium, 50 ng/mL NGF, B27 neuronal supplement, penicillin/streptomycin, and Glutamax at 37 °C with 5% CO2. After 24 h of in vitro culture, DRG explants were subsequently incubated for 48 h with conditioned media collected from RAW264.7 macrophage cultures. These macrophages had been pre-stimulated for 24 h under one of the following four conditions: LPS alone, LPS + Tof, LPS + Tof@HA-TAGel, or LPS + colivelin. Following 48 h of co-culture, the cells were immunostained for Anti-Neurofilament 200 (NF-200). Axonal morphology, including length and density, was analyzed utilizing the Simple Neurite Tracer plugin in ImageJ to trace individual branches from the cell body. Cell viability assay The cytotoxicity of the materials was assessed using a Cell Counting Kit-8 (CCK-8) assay. RAW264.7 cells were seeded in 96-well plates at 5000 cells per well and incubated for 12 h. The medium was then replaced with fresh medium containing TAGel, HA-TAGel, or Tof@HA-TAGel at concentrations ranging from 0 to 200 mg/mL. After 24 h of incubation, cell viability was measured by adding a 10% CCK-8 solution, incubating for 2 h, and recording the absorbance at 450 nm with a microplate reader (TECAN SPARK). Additionally, a Calcein-AM/PI double staining kit was used to visualize live and dead cells according to the manufacturer's protocol. ROS-scavenging activity Intracellular ROS levels were assessed using the DCFH-DA probe. Following a 24 h treatment with LPS (500 ng/mL) and the respective materials, RAW 264.7 cells (seeded at 5 × 104 cells/well in 12-well plates) were washed with PBS and incubated with DCFH-DA for 20 min. Fluorescence microscopy and ImageJ analysis were employed to measure the resulting green fluorescence intensity. DPPH assay The DPPH radical quenching capacity of the materials was assessed by combining 200 μL of a 100 μmol/L DPPH/ethanol solution with 20 μL of material solutions in a 96-well plate. After incubation for 30 min at room temperature in the dark, the absorbance at 517 nm was recorded. Immunofluorescence Following 24 h co-culture with LPS in the extracts of Tof, TAGel, HA-TAGel, or Tof@HA-TAGel, cells on coverslips were fixed with 4% paraformaldehyde for 15 min, permeabilized with 0.5% Triton X-100, and blocked with 3% BSA. Subsequently, the cells were incubated overnight at 4 °C with primary antibodies against TNF-α (11948, CST) and NF200 (N4142, Sigma-Aldrich), followed by incubation with appropriate Alexa Fluor® 488-conjugated secondary antibodies at room temperature. Nuclei were counterstained with DAPI, and fluorescence images were acquired using a ZEISS Celldiscoverer 7 microscope. Flow cytometry For flow cytometry analysis, RAW 264.7 cells were firstly seeded into 6-well plates (1 × 105 cells/well). After 12 h, cells were stimulated for 24 h with LPS (500 ng mL−1) alone or together with the indicated compounds (Tof, TAGel, HA-TAGel, Tof@HA-TAGel). Subsequently, cells were harvested, washed with PBS, and stained with a CD80 antibody (Biolegend, USA) for 40 min at 4 °C. Finally, cells were resuspended in flow staining buffer and analyzed using a Attune NxT flow cytometer. Western blot After a 24 h treatment with LPS in the extracts of Tof, TAGel, HA-TAGel, or Tof@HA-TAGel, cells were lysed, and the protein concentrations were quantified via BCA assay. Proteins were resolved by 10% SDS-PAGE, transferred to PVDF membranes, and blocked with 5% skim milk. The membranes were probed with specific primary antibodies overnight at 4 °C, primary antibodies included anti-TNF-α (11948, CST), anti-iNOS (13120, CST), anti-IL-10 (ab9969, Abcam), anti-Netrin-1 (AF1109-SP, R&D system), anti-CD206(24595, CST), anti-JAk1(ab47435, Abcam), anti-p-JAk1(ab138005, Abcam), anti-STAT3(4904T, CST), anti-p-STAT3 (9145S, CST). Then with HRP-conjugated secondary antibodies for 1 h at room temperature. Blots were visualized with chemiluminescent substrate on a ChemiDoc™ Touch imaging system, and band intensity was analyzed with ImageJ (version 1.51). ELISA The levels of inflammatory factors were measured by ELISA. RAW264.7 cells were seeded in 12-well plates (8.0 × 104 cells/well). After 12 h, the culture medium was replaced with DMEM containing LPS (500 ng/mL) alone (LPS group) or LPS combined with Tof, TAGel, HA-TAGel, or Tof@HA-TAGel, respectively. Following 24 h of incubation, the concentrations of TNF-α, IL-6, and IL-4 in the supernatants were quantified using commercial ELISA kits. RNA extraction and quantitative real-time PCR (qRT-PCR) Gene expression was analyzed by qRT-PCR. According to manufacturer protocols, total RNA was isolated with TRIzol® Plus and reverse-transcribed into cDNA. Following cDNA synthesis, qPCR amplifications were carried out in technical duplicates and three biological replicates using Fast SYBR® Green Master Mix. Relative transcript levels were calculated by the 2–ΔΔCt method using GAPDH as the endogenous control. RNA-seq analysis of RAW264.7 cells Total RNA was extracted using RNAiso Plus (Takara Bio) and assessed for purity (NanoDrop 2000) and integrity (Agilent 2100 Bioanalyzer). Sequencing libraries were prepared with the VAHTS Universal V6 RNA-seq Library Prep Kit, and transcriptome sequencing was performed by OE Biotech Co., Ltd. (Shanghai). Differential expression analysis was conducted with DESeq2, applying thresholds of |log2FC|>1 and Q value < 0.05 to identify significant DEGs. Hierarchical clustering and KEGG pathway enrichment analysis of DEGs were performed using R. Gene Set Enrichment Analysis (GSEA) was also carried out to evaluate the enrichment of predefined gene sets. Osteoarthritis animal model All animal experiments were conducted in accordance with protocols approved by the Ethics Committee of Shandong Provincial Hospital Affiliated to Shandong First Medical University (No. 2025-032). An osteoarthritis model was induced in 40 male C57BL/6 mice (8 weeks old) via transection of the right anterior cruciate ligament. Sham-operated mice underwent joint exposure and suture without ligament transection. The mice were randomly divided into six groups: sham, PBS, and four treatment groups (Tof, TAGel, HA-TAGel, Tof@HA-TAGel). Treatment groups received weekly intra-articular injections of the respective materials 10 μL [[28], [29], [30]], while the Sham and PBS groups received equal volumes of PBS, for a total of 8 injections. Upon completion of the treatment regimen, mice were euthanized, and the right knee joint was harvested for subsequent experiments. Mechanical sensitivity (Von Frey) test Mechanical allodynia was assessed by measuring the 50% paw withdrawal threshold (PWT) using Von Frey filaments. After a 30 min acclimation in cages with a wire mesh floor, filaments were applied to the hind paw plantar surface from below. The 50% PWT, defined as the force at which a 50% withdrawal response probability occurs, was calculated. An increased PWT denotes attenuated mechanical allodynia, indicating an analgesic effect [31]. Gait analysis Gait parameters of mice were quantified using the CatWalk XT system (Noldus). Briefly, each mouse was allowed to traverse a glass walkway in a darkened room. As the paws contacted the glass, the resulting illuminated footprints were recorded by a camera beneath the walkway and analyzed using the CatWalk software (v7.1). The following parameters were analyzed: stride length, paw print area and paw pressure. IVIS imaging Mice subjected to ACLT surgery were randomly allocated into three groups and received intra-articular injections of Cy5, Cy5-Tof@TAGel, or Cy5-Tof@HA-TAGel. The in vivo fluorescence distribution within the knee joints was monitored non-invasively at predetermined time points (days 0, 2, 4, 6, 8, 10, 12, 14) using an IVIS imaging system (IVIS Lumina XRMS Series III, PerkinElmer). For precisely biodistribution assessment, a separate cohort of mice was injected, and major organs (heart, liver, spleen, lung, and kidney), blood, and joint tissues (knee synovium (KS), femoral condyle (FC), meniscus (M), and tibial plateau (TP)) were harvested 24 h post-injection for ex vivo imaging and quantification of radiant efficiency. ROS imaging was performed using the chemiluminescent probe L-012. Briefly, following the respective material treatments, the L-012 probe was intra-articularly injected. The resulting fluorescence signal was then imaged in anesthetized mice and quantified to determine the ROS levels. Micro-CT analysis The knee joints, fixed and stripped of soft tissue, were scanned by micro-CT (SkyScan 1176) at a resolution of 9 μm (80 kVp, 124 μA). A region of interest encompassing the entire medial tibial subchondral bone was defined using Data Viewer software (Version 1.5). 3D models were reconstructed in CTVol (Version 2.0), and structural parameters—including bone volume fraction (BV/TV), trabecular number (Tb.N), and trabecular separation (Tb.sp)—were quantified from the datasets using CTAn (Version 1.8) software. Histological assessment Human osteoarthritic synovium was harvested from patients undergoing total knee arthroplasty, with normal synovial tissue obtained postmortem from non-OA donors. All procedures were approved by the institutional ethics committee (NSFC: NO.2024-680), and informed consent was obtained. The tissue samples were processed through fixation in 4% paraformaldehyde (48 h), paraffin embedding, and sectioning at a thickness of 6 μm. For histological examination, mouse tissue sections were subjected to Hematoxylin and Eosin (H&E) and Safranin O/Fast Green (S/F) staining. Osteoarthritis severity was subsequently quantified using the Osteoarthritis Research Society International (OARSI) scoring system. For immunofluorescence, frozen sections were incubated overnight at 4 °C with primary antibodies against iNOS (sc7271, Santa Cruz), Netrin-1 (AF1109-SP, R&D system), and CGRP (ab81887, Abcam). Subsequently, Alexa Fluor-conjugated secondary antibodies were applied for signal detection. Images were acquired using a TissueFAXS Plus S laser confocal microscope. Statistical analysis Data are from at least three independent experiments and presented as mean ± SD. Statistical analyses were performed using GraphPad Prism 8, with Student's t-test for two-group comparisons and one-way ANOVA followed by Tukey's test for multiple comparisons. Statistical significance is indicated as ∗p < 0.05, ∗∗p < 0.01 and ∗∗∗p < 0.001. Results and discussion Relationship between macrophages and pain in knee osteoarthritis To investigate the association between synovial macrophage polarization in OA related pain, human synovial tissues were acquired during total knee arthroplasty from OA patients (OA group), with control tissues obtained postmortem from donors with no recorded arthritic pathology (control group). Dual-label immunofluorescence was performed on tissue cryosections to concurrently detect netrin-1, a key pain-associated mediator [32], and iNOS, a canonical marker of M1 macrophages [33] (Fig. 1A). Results showed a significant increase in iNOS-positive M1 macrophages in OA synovium (80.4% ± 2.8%) (Fig. 1C), consistent with previous reports [34]. Concurrently, enhanced green fluorescence intensity indicated upregulated Netrin-1 expression (77.6% ± 3.2%) (Fig. 1D). Colocalization analysis revealed strong overlap between the two signals (Fig. 1H), suggesting a close correlation between Netrin-1 expression and M1 macrophage infiltration. Similar patterns were observed in ACLT-induced OA mouse models (Fig. 1B, E, 1F and 1G). These findings imply that M1 macrophage may exacerbate arthritic pain through Netrin-1 upregulation. To further investigate the role of macrophages in arthritic pain, we depleted macrophages in the joints of OA mice via intra-articular injection of clodronate liposome-a potent macrophage depleting agent [35,36] and assessed the pain threshold (50% PTW) pre-injection and post-injection (Fig. 1I). As shown in Fig. S1, the fluorescence signal of iNOS was markedly reduced following intra-articular injection of clodronate liposomes, indicating successful macrophage depletion. It's worth noting that macrophages depletion significantly alleviated pain sensitivity in OA mice, as evidenced by an increase in the 50% PWT from 0.56 ± 0.08 g to 1.49 ± 0.15 g (p < 0.05) (Fig. 1J). These results strongly indicate that synovial M1 macrophages play a critical role in the modulation of pain sensitivity in OA,highlighting the therapeutic potential of targeting suppress synovial macrophage M1 polarization for OA pain management. Correlation between M1 macrophages and pain in knee osteoarthritis. (A) Representative immunofluorescence images of iNOS and Netrin-1 in normal and OA human synovial tissues. (B) Representative immunofluorescence images of iNOS and Netrin-1 in the synovium of sham-operated and ACLT-induced OA mice. (C) Quantification of the percentage of M1 macrophages (iNOS + macrophages) in normal and OA human synovial tissues. (D) Quantification of the percentages of Netrin-1-positive cells in normal and OA human synovial tissues. (E) Quantification of the percentage of M1 macrophages (iNOS + macrophages) in the synovium of sham-operated and ACLT-induced OA mice. (F) Quantification of the percentages of Netrin-1-positive cells in the synovium of sham-operated and ACLT-induced OA mice. (G) Correlation analysis of the Netrin-1 expression and M1 macrophages in mouse synovium. (H) Correlation analysis of the Netrin-1 expression and M1 macrophages in human synovium. (I) Schematic of the experimental timeline showing clodronate liposome administration and Von Frey behavioral testing in the ACLT-induced OA mouse model. (J) 50% PWT of OA mice before and after administration of the macrophage-depleting agent. Data are presented as mean ± SD (n = 6), ns = not significant, ∗p < 0.05, ∗∗p < 0.01, and ∗∗∗p < 0.001. Characterization of Tof@HA-TAGel Tof@Gel microspheres with a hydrodynamic diameter of around 200 μm (Fig. 2B and C) were prepared by mixing Tof with GelMA monomer as the aqueous phase in microfluidic technology (Fig. 2A). The obtained Tof@Gel microspheres were immersed in TA solution for 12h, yielding TA-coated microspheres designated as Tof@TAGel. Based on the work of Liang et al., 2 mg/mL TA demonstrated favorable stability, antioxidant properties, biocompatibility, and high encapsulation efficiency when incorporated into the hydrogel matrix. Therefore, this concentration was selected for the synthesis of Tof@TAGel [37]. As shown in Fig. 2D and E, the particle size of Tof@TAGel microspheres was approximately 194.1 ± 2.6 μm, which was smaller than that of Tof@Gel, an effect that could be attributed to the cross-linking action of TA. To enable intra-articular administration, microspheres were formulated using 2 mg/mL HA to ensure stability and appropriate rheological properties [38,39]. ADH was coupled to the carboxyl group of HA to obtain HA-ADH (Figs. S2 and S3). The HA-ADH was mixed with Tof@TAGel to prepare Tof@HA-TAGel microspheres. With the surface graft of HA, the average hydrodynamic diameter of Tof@HA-TAGel increased to 213 ± 1.1 μm (Fig. 2F and G), double larger than that in the oil phase (Fig. S4). The loading efficiency of TA and Tof in the microspheres were 43.2 ± 1.1% and 78.8 ± 2.6%. FTIR spectroscopy was employed to characterize the chemical structure of the composite microspheres (Fig. 2H). The peak at 1653 cm− 1 was attributed to the stretching vibration of the amido bond in HA-ADH. TA showed stretching vibrations at 1695, 1178, and 752 cm−1, attributed to the C=O, C–O, and C–H in benzene ring, respectively. The C–O stretching vibration of amide I and the N–H deformation of amide II from Gel were located at 1633, and 1533 cm−1, respectively [40]. The phenol in TA could be oxidized to quinone and reacted with amine groups through a Michael-type reaction. The product of phenoxyl radical-related coupling reaction was confirmed at 966 cm−1, confirmed the successful grafting of the HA onto the microspheres [41]. Further thermogravimetric analysis was performed to evaluate the thermal degradation characteristics of the microspheres (Fig. 2I and J). The thermal weight loss of TA mainly originated from decarboxylation of the outermost layer of gallic acid units when the temperature was within the range of 230 °C to 400 °C. Upon heating above 600 °C, crosslinking of aromatic rings were occurred in TA, and futher condensated form carbonaceous char [42]. The thermal weight loss of gelatin was associated with protein degradation and thermal decomposition of the gelatine networks [43]. The thermogravimetric temperature of TA in TAGel microspheres decreased significantly, while that of gelatin increased. It might be due to the hierarchical structure of TAGel microspheres increase the thermal sensitivity of TA, and the TA-based cross-linked shell formed on the microsphere surface serves as a protective barrier for the gelatin core, thereby enhancing its structural stability. Their morphology and porous structure were investigated by SEM. As shown in Fig. 2K, the freeze-dried Tof@Gel microspheres possessed a 3D interconnected network with large pores, displaying a pore diameter of 13 ± 2.5 μm on average. A noteworthy observation was that the Tof@HA-TAGel microspheres developed a more homogeneous and dense porous structure (Fig. 2L), resulting from the cross-linking between TA and gelatin. To assess the lubrication performance of the TAGel microspheres, the friction test was conducted using a high-frequency reciprocating friction and wear tester for 1200 s. As shown in Fig. 2M, the physically blended TA&Gel exhibited a lower initial coefficient of friction (COF). However, its COF increased continuously due to the disruption of the hydration layer. In contrast, the COF of TAGel microspheres remained stable throughout the entire friction process. This stability is attributed to their inherent elasticity and rolling performance, which collectively enhance the load-bearing capacity and wear resistance. The in vitro release characteristics of TA in Tof@HA-TAGel is presented in Fig. 2N, calculated using the standard curve of TA (Fig. S5). The cumulative release of TA from Tof@HA-TAGel was assessed by ultraviolet-visible spectra. TA showed a faster release within 48 h, followed by a moderate and controlled release profile. Following 20 days of degradation, the Tof@HA-TAGel microspheres were lyophilized and their morphological changes were examined using SEM. Fig. 2O shows the disintegration of the microsphere architecture. The internal structure was observed at higher magnification, small holes appeared attributed to erosive degradation. Characterization of Tof@HA-TAGel. (A) The synthesis of Tof@HA-TAGel microspheres. The bright-field microscopic image (B) and diameter distribution (C) of Tof@Gel microspheres. The bright-field microscopic image (D) and diameter distribution (E) of Tof@TAGel microspheres. The bright-field microscopic image (F) and diameter distribution (G) of Tof@TAGel microspheres. (H) FTIR spectra of different compositions. (I) TGA and (J) DTG curves of TA, Gel and TAGel. SEM images of Tof@Gel (K) and Tof@HA-TAGel (L). (M) Friction curve of TA&Gel and TAGel. (N) Release behaviors of TA from Tof@HA-TAGel. (O) Representative SEM image of the degraded Tof@HA-TAGel. The red-boxed area in the high-magnification view reveals pores formed by water erosion. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.) Cytocompatibility in vitro The cytocompatibility of the materials was evaluated using CCK-8 assays. Based on the CCK-8 assay results (Fig. S6), a Tof concentration of 50 nM was selected for microsphere synthesis. RAW264.7 cells were cultured alongside TAGel, HA-TAGel, and Tof@HA-TAGel at a series of concentrations (0,5,10,25,50,100 and 200 μg/mL) for 24h (Fig. 3A,B and 3C). At concentrations below 25 μg/mL, Tof@HA-TAGel maintained cell viability above 98% ± 1.81% (Fig. 3C), demonstrating favorable cytocompatibility. However, the viability reduced when concentrations exceeding 50 μg/mL, and the viability decreased to 85.31% ± 2.04% in the group with 200 μg/mL Tof@HA-TAGel (Fig. 3C). The favorable cellular compatibility at 25 μg/mL of Tof@HA-TAGel was further confirmed by live/dead staining (Fig. 3E). Quantitative analysis revealed no significant differences in viability between the control group (99.61% ± 0.50%) and cells treated with TAGel, HA-TAGel, or Tof@HA-TAGel at 25 μg/mL, which exhibited viability rates of 98.07% ± 0.71%, 98.88% ± 1.05%, and 98.99% ± 0.43%, respectively (Fig. 3D). Based on these results, Tof@HA-TAGel at 25 μg/mL was selected for subsequent experiments. Cytocompatibility in vitro. CCK-8 assay of RAW264.7 incubated with various concentrations of TAGel (A), HA-TAGel (B) and Tof@HA-TAGel (C). (D) Quantification of the livie/dead cell proportion with Calcein-AM/PI staining. (E) Fluorescence images of RAW264.7 cells stained with Calcein-AM (green, live cells) and PI (red, dead cells) after incubation with various materials for 24h. Data are presented as mean ± SD (n = 3), ns = not significant, ∗p < 0.05, ∗∗p < 0.01, and ∗∗∗p < 0.001. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.) Selective targeting and retention of Tof@HA-TAGel in vitro and in vivo Macrophage affinity of Tof@HA-TAGel in vitro RAW264.7 cells, with or without LPS stimulation, were co-cultured with Tof@ TAGel or Tof@HA-TAGel microspheres. As shown in Fig. 4A-F, after 24 h of incubation, substantial cell adhesion was observed on the microsphere surfaces. Notably, Tof@HA-TAGel microspheres exhibited enhanced recruitment to LPS-induced M1 macrophages (Fig. 4F). Quantitative assessment revealed a statistically significant elevation in the cell adhesion rate in the Tof@HA-TAGel + LPS group relative to the other two groups (Fig. 4G), indicating stronger affinity of Tof@HA-TAGel toward M1-polarized macrophages. This enhanced targeting capability can be attributed to the ligand-receptor interaction between HA and CD44 [25,26]. It is reported that CD44 is preferentially overexpressed on M1 macrophages compared to M0 and M2 subtypes [26]. Selective targeting and retention of Tof@HA-TAGel in vitro and in vivo. Representative bright-field images of RAW264.7 with LPS recruitment by Tof@TAGel (A and D) or Tof@HA-TAGel (C and F). (B and E) Representative bright-field images of RAW264.7 without LPS recruitment by Tof@HA-TAGel. (G) Quantification of cell recruitment rates among different groups. (H) Representative fluorescence images of OA mice knees for 12 days after an intra-articular injection of Cy5, Cy5-Tof@TAGel, and Cy5-Tof@HA-TAGel. (I) Quantitative analysis of the time course radiant efficiency within knee joints following intra-articular injection of Cy5, Cy5-Tof@TAGel, and Cy5-Tof@HA-TAGel. (J) AUC of fluorescence intensity profiles. Biodistribution of Cy5 in joint components (M) and quantitative analysis (K) of radiation efficiency. Biodistribution of Cy5 in major organs and blood samples (N) and quantitative analysis (L) of radiation. KS: Knee synovium; FC: Femoral condyle; M: Meniscus; TP: Tibial plateau. Data are presented as mean ± SD (n = 6), ns = not significant, ∗p < 0.05, ∗∗p < 0.01, and ∗∗∗p < 0.001. Targeting and longevity of Tof@HA-TAGel in vivo Cy5-labeled microspheres were used as tracers to monitor the retention and distribution of Tof@HA-TAGel in the knee joint of male C57 mice subjected to ACLT surgery. Mice received a single intra-articular injection of free Cy5, Cy5-Tof@TAGel, or Cy5-Tof@HA-TAGel. Fluorescence imaging of the knee region at multiple time points are shown in Fig. 4H. The fluorescence signal of free Cy5 became nearly undetectable by day 6 post-injection. In contrast, the Cy5-Tof@HA-TAGel retained measurable signal intensity up to day 12, with an approximately 75% decay observed by day 6 (Fig. 4I). Based on this, we administered intra-articular injections to the mice once weekly to maintain a stable drug concentration. The AUC of the Cy5-Tof@HA-TAGel exceeded that of the free Cy5 by more than two-fold (Fig. 4J). These results demonstrate that the HA-TAGel microsphere significantly prolongs the intra-articular retention of the payload, likely due to its enhanced stability and sustained-release characteristics in joint tissues. Biodistribution of the microspheres in joint tissues and major organs was analyzed 24 h post-injection. Fluorescence signals were detected in multiple articular structures—including KS, FC, M, and TP (Fig. 4K). The higher fluorescence intensity across all joint components were observed in Cy5-Tof@HA-TAGel group compared to free Cy5, consistent with IVIS image observed in prior experiments. Notably, the KS displayed the strongest fluorescence accumulation following Cy5-Tof@HA-TAGel administration (Fig. 4M), whereas free Cy5 and Cy5 -Tof@TAGel showed no tissue-specific distribution. This preferential synovial enrichment likely results from the abundant M1 macrophages present in OA synovium, which facilitate targeted accumulation of the microspheres via ligand-receptor interactions. These in vivo findings further corroborate the M1 macrophage-targeting capability of Tof@HA-TAGel, aligning with the in vitro observations. Additionally, obvious fluorescence signals were observed in the liver across all groups (Fig. 4N). This distribution pattern suggests elimination pathways involving synovial fluid diffusion, capillary permeation, and lymphatic drainage, culminating in hepatic metabolism and detoxification. Notably, the fluorescence intensity in the liver were lower in two microsphere-based groups (Tof@TAGel and Tof@HA-TAGel) than that in Cy5 group (Fig. 4L). This phenomenon is likely attributable to the particle size of the microspheres, which may have restricted synovial fluid diffusion and capillary permeation, thereby enhancing drug retention within the joint cavity to a greater extent. These findings are consistent with our previous results. Therapeutic mechanisms of Tof@HA-TAGel demonstrated through RNA-seq This work established Tof@HA-TAGel as an exceptionally compelling candidate in the field of OA pain. To elucidate the ability and mechanistic basis of Tof@HA-TAGel in treating OA, mRNA sequencing was performed on RAW264.7 cells from Tof@HA-TAGel and LPS groups (n = 3). Comparative analysis identified 748 differentially expressed genes (DEGs), comprising 461 upregulated and 287 downregulated transcripts (Fig. 5A). As shown in Fig. 5A, the upregulated gene set showed strong associations with anti-inflammatory (Socs1, Il-10, Foxo3), antioxidant (Oxr1, Gpx4, Cyp27), and analgesic processes (Scn9a, Frzb, Ccn3). Conversely, downregulated genes were linked to inflammatory (Il-6, Tnf-α, Usp18), pro-oxidative (Nos2, Cybb, Edn1), and pain-related pathways (Sema4b, Ptafr), along with bone/cartilage destruction markers (Fcgr1, Bcl3, Irf9). Gene Ontology (GO) enrichment analysis revealed that Tof@HA-TAGel promoted expression of autophagy- and apoptosis-related genes while suppressing macrophage activation genes, indicating its role in limiting M1 macrophage-driven inflammation. Meanwhile, these downregulated genes were associated with oxidative stress (NO biosynthesis, monooxygenase activity), pain (chemokine-mediated signaling), and osteoclast proliferation/differentiation (chemotaxis and lipid response) (Fig. 5B). KEGG pathway analysis showed DEG enrichment in pain modulation and inflammatory microenvironment regulation (IL-17, TNF, JAK/STAT signaling), alongside pluripotency regulation pathways (osteoclast differentiation, apoptosis) (Fig. 5C). Gene set enrichment analysis (GSEA) further confirmed suppression of oxidative stress (Fig. 5D) and inflammatory responses (Fig. 5E) after Tof@HA-TAGel treatment. Based on these findings, we hypothesize that Tof@HA-TAGel could alleviate osteoarthritic inflammation and pain through coordinated mechanisms involving oxidative stress mitigation and immune microenvironment modulation. Next, we aimed to verify the efficacy and mode of action of Tof@HA-TAGel through a combination of in vitro and in vivo studies. Transcriptional analysis of RAW264.7 by RNA-seq. (A) Volcano plot of genes differentially expressed in Tof@HA-TAGel and LPS groups. (∗p < 0.05, |log2(fold change) |> 1). (B) GO terms of Tof@HA-TAGel versus LPS groups. (C) KEGG pathway enrichment analysis of the Tof@HA-TAGel versus LPS groups. Gene set enrichment analysis of oxidative stress (D) and inflammation-related genes (E). Antioxidant capacity of Tof@HA-TAGel in vitro Pathologically, the accumulation of reactive oxygen species (ROS), notably, superoxide (O2−•)and hydrogen peroxide (H2O2), is positively related with OA progression [44]. The highly expressed oxidative stress triggers mitochondrial dysfunction, extracellular cartilage matrix degradation, and increased inflammation [45]. Consequently, addressing oxidative stress is of great importance for slowing OA progression [46]. TA has been widely documented to enhance the activity of various antioxidant enzymes and exhibits potent free radical-scavenging capacity due to its multiple phenolic hydroxyl groups [47,48]. It is hypothesized that Tof@HA-TAGel could effectively eliminate excessive free radicals in the osteoarthritic microenvironment. The in vitro antioxidant activity was first assessed using a DPPH assay. DPPH, a stable radical compound with a characteristic purple color and absorption peak at 517 nm, can be reduced to a yellow product by antioxidants, serving as an ideal indicator for free radical scavenging. As shown in Fig. 6A, TAGel, HA-TAGel and Tof@HA-TAGel effectively preserved the intrinsic DPPH quenching ability of TA, as evidenced by the color transition from purple to light yellow, whereas free Tof showed negligible effect. Quantitative analysis at 517 nm further confirmed these results (Fig. 6B). To evaluate the modulation of intracellular oxidative stress, LPS-stimulated RAW 264.7 cells were treated with the extracts of the microspheres. Intracellular ROS levels were monitored using the fluorescent probe DCFH-DA via confocal laser scanning microscopy (CLSM) (Fig. 6D). A signally higher population of green fluorescence positive cells and stronger fluorescence intensity was observed in LPS-treated cells, whereas that in Tof@HA-TAGel group showed markedly reduced. Quantitative analysis was shown in Fig. 6C, confirmed a statistically significant decrease in ROS levels followed the intervention. Antioxidative function of Tof@HA-TAGel in vitro. (A) DPPH solution supplemented with Tof, TAGel, HA-TAGel and Tof@HA-TAGel. (B) DPPH elimination rates of Tof, TAGel, HA-TAGel and Tof@HA-TAGel. Data were presented as mean ± SD (n = 3). ∗∗∗p < 0.001, compared with the Tof group. (C) The relative intensity of intracellular ROS revealed by DCFH-DA in different groups. (D) Fluorescence images of DCFH-DA in RAW264.7 incubated with or without LPS and various materials. Data were presented as mean ± SD (n = 3). ∗∗∗p < 0.001, compared with the LPS group. Tof@HA-TAGel modulates macrophage polarization and immune microenvironment During the acute phase of OA, infiltrated M1 macrophages trigger a “cytokine storm” via massive release of inflammatory mediators, thereby exacerbating disease progression [33]. To evaluate the capacity of Tof@HA-TAGel to suppress M1 polarization, the expression of TNF-α and CD80 in macrophages were assessed using immunofluorescence and flow cytometry. As shown in Fig. 7A, LPS stimulation significantly upregulated TNF-α level in RAW 264.7 cells, whereas reduced green fluorescence intensity was observed in all treatment groups (Tof, TAGel, HA-TAGel, and Tof@HA-TAGel), with the most pronounced decrease observed in the Tof@HA-TAGel group (p < 0.001; Fig. 7B). Flow cytometric analysis also corroborated these findings, showing the lowest proportion of CD80+ cells in the Tof@HA-TAGel treated group (Fig. 7C and D). In addition, qRT-PCR analysis corroborated that LPS stimulation elevated TNF-α expression to 1.8 ± 0.8 fold of baseline (p < 0.001), which was markedly reversed by all therapeutic formulations. Notably, Tof@HA-TAGel mediated the most significant normalization of TNF-α expression (p < 0.001; Fig. 7E). A similar trend was observed for CD86 gene expression (Fig. 7F). Conversely, genes associated with M2 macrophages, CD206 (Fig. 7G) and Arg-1 (Fig. 7H), were upregulated to varying degrees across treatment groups. Meanwhile, western blot (Fig. 7M) analysis further demonstrated that Tof@HA-TAGel treatment downregulated M1-related proteins iNOS (Fig. 7I) and TNF-α (Fig. 7J), while upregulating M2 markers CD-206 (Fig. 7K) and IL-10 (Fig. 7L) compared to the LPS group. The capacities of Tof@HA-TAGel to inhibit M1 macrophage polarization and attenuate local “cytokine storm”. (A) Representative confocal images of LPS-activated RAW 264.7 macrophages in each group. (B) Relative fluorescence intensity of TNF-α in each group. (C) Detection of CD80+ macrophages by flow cytometry. (D) The ratios of CD80+ macrophages in each group. TNF-α (E), CD86 (F), CD206 (G) and Arg-1 (H) gene expression levels measured by qRT-PCR in each group after LPS activation. Western blot bands (M) and quantification of the M1 marker iNOS (I), TNF-α (J) and M2 marker CD206 (K), IL-10 (L). TNF-α (N), IL-6 (O) and IL-4 (P) measured by ELISA in each group after LPS activation. Data were presented as mean ± SD (n = 3). ∗∗p < 0.01, ∗∗∗p < 0.001, compared to the LPS group, respectively. The inhibition of M1 macrophage polarization and the restoration of the M1/M2 balance are expected to reduce the production of key inflammatory factors, including TNF-α and IL-6. Consequently, the concentrations of TNF-α, IL-6, and the M2-associated anti-inflammatory cytokine IL-4 in the cell culture supernatants were quantified by ELISA to confirm this effect. LPS stimulation markedly increased TNF-α secretion from 72 ± 10 pg mL−1 to 1650 ± 88 pg mL−1 (p < 0.001) (Fig. 7N). Following drug intervention, both TNF-α and IL-6 levels were reduced to varying degrees, with the Tof@HA-TAGel group exhibiting the lowest concentrations of these pro-inflammatory cytokines (Fig. 7N and O). The elevated IL-4 levels were observed in the TAGel, HA-TAGel, and Tof@HA-TAGel groups compared to LPS group (Fig. 7P). These multi-level findings consistently verify that Tof@HA-TAGel effectively regulate the balance of M1/M2 polarization, remodels the immune microenvironment, mitigates inflammatory cytokine release, and ultimately alleviates the local “cytokine storm." Tof@HA-TAGel suppress nerve ingrowth through the JAK1/STAT3/netrin-1 pathway Capitalizing on the pharmacological activity of Tof and the experimental rationale established above, this study aims to determine whether Tof@HA-TAGel could regulate the express of netrin-1 and subsequent modulation of sensory nerve ingrowth through JAK1/STAT3 pathway. Western blotting was performed to examine the expression of JAK1 and STAT3 proteins (Fig. 8A). Compared with the control group, LPS stimulation significantly upregulated the levels of p-JAK1and p-STAT3 (Fig. 8C and D), The phosphorylation of both JAK1 and STAT3 downregulated markedly after treated with Tof@HA-TAGel. Strikingly, the alteration in netrin-1 levels paralleled the changes observed in p-JAK1 and p-STAT3. Tof@HA-TAGel reduced the expression of Netrin-1 significantly (Fig. 8B and E). To validate the critical role of the JAK1/STAT3 pathway in mediating the inhibitory effect of Tof@HA-TAGel on Netrin-1 secretion, cells were co-treated with colivelin, a STAT3 agonist [49]. As shown in Fig. 8E, colivelin reversed the suppressive effect of Tof@HA-TAGel on Netrin-1 expression in LPS-induced RAW264.7 cells. Consistent results were obtained by ELISA (Fig. 8F) and RT-qPCR (Fig. 8G). A key step in the activation of the JAK1/STAT3 pathway is the nuclear translocation of p-STAT3 [50,51]. Therefore, we performed IF staining for p-STAT3 in mouse knee joint synovium and quantified the fluorescence intensity of p-STAT3 within the nucleus. As illustrated in Fig. 8H, pronounced nuclear translocation of p-STAT3 was observed in the synovium of OA mice. In contrast, the Tof@HA-TAGel group exhibited significantly reduced nuclear translocation compared to the PBS group (Fig. 8I), achieving levels comparable to those in the sham group. Furthermore, IF staining for synovium revealed a marked increase in iNOS-positive cells and enhanced Netrin-1 fluorescence signal in the PBS group (Fig. 8L). Following treatment, particularly with Tof@HA-TAGel, a reduction in M1 macrophage polarization was observed. Tof@HA-TAGel intervention decreased the percentage of M1 macrophages from 73% ± 2.8% to 24% ± 3% (Fig. 8J). Additionally, weakened Netrin-1 fluorescence signals were detected in the treatment groups, with the most pronounced effect seen in the Tof@HA-TAGel group (Fig. 8K). These findings collectively demonstrate that Tof@HA-TAGel reduces the secretion of the Netrin-1 from macrophages by downregulating the JAK1/STAT3 pathway. Tof@HA-TAGel inhibits sensory nerve growth through downregulation of JAK1/STAT3 pathway and Netrin-1 secretion. (A and B) Representative images of JAK1/p-JAK1, STAT3/p-STAT3 and Netrin-1 protein by western blotting of RAW264.7 cells in different groups. Western blot analysis of JAK1/p-JAK1 (C), STAT3/p-STAT3 (D) and Netrin-1 (E). (F) Netrin-1 measured by ELISA in each group. (G) Netrin-1 gene expression levels measured by qRT-PCR in each group. Data were presented as mean ± SD (n = 3). ∗∗∗p < 0.001, compared to the LPS group. (H) IF staining of p-STAT3 in mouse synovium in different groups. Yellow arrows indicate nuclear translocation of p-STAT3. (I) Quantitative analysis of p-STAT3 fluorescence intensity in the nucleus. (L) Representative IF staining images of iNOs and Netrin-1 in the ACLT mice with different treatments. (J) Relative iNOS and (K)Netrin-1 expression were analyzed in each group. Data were presented as mean ± SD (n = 6). ∗∗∗p < 0.001, compared to the PBS group. (M) Schematic diagram of the co-culture of DRGs with conditioned medium from LPS, LPS + Tof, LPS + Tof@HA-TAGel and LPS + Tof@HA-TAGel + colivelin treated RAW264.7 cells. The length (O) and density (P) of axons were quantified using ImageJ software. (N) IF images of the NF200 positive axons in DRGs. Data were presented as mean ± SD (n = 3). ∗∗∗p < 0.001, compared to the LPS group. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.) The DRG, which contains the soma of primary afferent neurons, serves as a critical hub for pain signal transduction [52]. However, the complex in vivo environment presents challenges for isolating specific cellular mechanisms. To enable a precise and controlled analysis, DRG were harvested from mice to delineate the impact of macrophage polarization on sensory neurite outgrowth. Conditioned media from RAW264.7 cells treated with LPS, LPS + Tof@HA-TAGel or LPS + Tof@HA-TAGel + colivelin were collected and added to the DRG culture medium (Fig. 8M). Following a 48h incubation in conditioned medium, the DRG were processed for NF-200 immunofluorescence analysis. IF staining of the DRG revealed that the conditioned medium from the LPS + Tof@HA-TAGel group significantly suppressed axonal outgrowth (Fig. 8N). Quantitative analysis further demonstrated that the axonal length of DRG in the Tof@HA-TAGel group (199.3 ± 3.8 μm) was substantially shorter than that in the LPS group (401.4 ± 3.6 μm) (Fig. 8O). Furthermore, the axonal density of DRG in the Tof@HA-TAGel group was 50.93 ± 3.5, which lower than that in the LPS group (96.97 ± 3.2) (Fig. 8P). Moreover, the colivelin significantly reversed the declining trend observed in the Tof@HA-TAGel group, with both axonal length and density increasing significantly. Together, these data provide evidence that Tof@HA-TAGel effectively inhibits sensory nerve growth by downregulating the JAK1/STAT3 signaling pathway. Therapeutic effects of Tof@HA-TAGel in vivo The therapeutic efficacy of Tof@HA-TAGel were further evaluated in ACLT-induced OA mice model. Beginning immediately after surgery, 8-week-old C57BL/6 male mice received weekly intra-articular injections of PBS, Tof, TAGel, HA-TAGel, or Tof@HA-TAGel for 8 weeks (Fig. 9A). Therapeutic outcomes were systematically assessed using Gait analysis, Von Frey test, infrared thermography, H&E staining, Safranin O-Fast Green staining, micro-CT and etc. Tof@HA-TAGel alleviates OA pain in vivo. (A) Schematic diagram of OA treatment and the timeline for the animal model. (B) Gait anlysis (pink footprints represent the right hindpaw (RH), blue the right forepaw (RF), green the left hindpaw (LH), and yellow the left forepaw (LF)). (C) The print area of mouse. Quantitative analysis of stride length (D), print area (E) and max intensity at (F). (G) Schematic of the Von Frey Test. (H) 50%PWT at 0, 2, 4, 6 and 8 weeks following ACLT surgery in the different groups. Data were presented as mean ± SD (n = 6). ∗∗p < 0.01, ∗∗∗p < 0.001, compared to the PBS group. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.) Tof@HA-TAGel alleviates osteoarthritic pain As OA pain represents the most prominent symptom of osteoarthritis, leading to functional limitations and reduced quality of life, we conducted behavioral tests to evaluate the therapeutic effects on pain perception. Gait analysis was performed using the CatWalk XT system to record movement trajectories across different groups (Fig. 9B). Typical vedios of mice in the sham, PBS and Tof@HA-TAGel group were showed in supplementary vedio1, vedio 2, and vedio 3. Stride length, a key parameter reflecting joint pain, is defined as the distance between two consecutive paw placements of the same limb. The increased values of stride length indicate improved knee function [53]. Quantitative analysis revealed a stride length of 6.8 ± 0.3 cm in the sham group, whereas the PBS treated group showed a significant reduction to 3.1 ± 0.2 cm, which attributed to pain-induced mobility impairment (Fig. 9D). Treatment with Tof, TAGel, HA-TAGel, and Tof@HA-TAGel resulted in stride lengths of 5.1 ± 0.2 cm, 3.6 ± 0.1 cm, 5.4 ± 0.2 cm, and 6.4 ± 0.3 cm, respectively, demonstrating varying degrees of functional recovery. Notably, the mice in Tof@HA-TAGel group exhibited the greatest improvement, closely approximating normal level. Additionally, hind paw footprint area and pressure were evaluated (Fig. 9C). Joint pain typically leads to limping or toe-walking, resulting in reduced contact area and pressure [54]. The footprint area in Tof@HA-TAGel group was the largest among all treatment groups, nearly tripling that of the PBS group (p < 0.001) (Fig. 9E). Similar improvements were observed in hind paw pressure measurements (Fig. 9F). These behavioral results collectively indicate that Tof@HA-TAGel effectively alleviates OA-related pain and restores locomotor function. In pain situations, the animals are sensitive to mechanical stimuli (such as light touch, light pressure, or brushing), which correspond with an increase in mechanical allodynia [55]. Von Frey test is a commonly used behavioral test to assess pain sensitivity or tactile sensory thresholds to mechanical stimuli in rodents [56]. Here, the von Frey test was performed to evaluate the mechanical allodynia of the OA-affected animals after different treatments (Fig. 9G). Two weeks after ACLT surgery, a significant reduction in 50% PWT was observed in mice, with values recorded at 0.37 g ± 0.08 g (PBS), 0.54 g ± 0.02 g (Tof@HA-TAGel), and 1.45 g ± 0.01 g (sham). Following intra-articular injections, the pain thresholds exhibited a gradual recovery. The Tof@HA-TAGel group demonstrated a markedly enhanced recovery compared to the PBS group. By week 4, the sham group's pain threshold closely approximated baseline levels (1.9 g ± 0.02 g). At the 8-week endpoint, the Tof@HA-TAGel group achieved near-complete functional recovery (1.86 g ± 0.02 g), showing comparable efficacy to the sham group and significant improvement over the PBS group (0.58g ± 0.03 g; Fig. 9H). This significant recovery indicates effective reversal of mechanical allodynia and reduced pain sensitivity. Overall, our results verified that Tof@HA-TAGel effectively alleviates pain and improves physical abilities in OA. Tof@HA-TAGel reduce inflammation and ROS level, inhibit sensory nerve ingrowth Local inflammation in OA is frequently accompanied by elevated knee temperature due to synovitis [57]. We therefore employed infrared thermography to assess thermal changes in response to different treatments. The PBS group exhibited the highest knee temperature whereas all treatment groups showed varying degrees of reduction (Fig. 10A). Notably, the Tof@HA-TAGel group achieved the lowest temperature among all treatment groups (p < 0.001; Fig. 10C), indicating superior anti-inflammatory efficacy. Histopathological evaluation of synovial inflammation was performed using H&E staining and scored according to OARSI guidelines. PBS group displayed characteristic synovitis, including intimal hyperplasia and inflammatory cell infiltration (Fig. 10B), resulting in significantly higher pathological scores at 8 weeks compared to sham group. In contrast, Tof@HA-TAGel treatment substantially attenuated these inflammatory changes. Consistent with this finding, the Tof@HA-TAGel group showed the lowest OARSI scores among all treatment groups (p < 0.001; Fig. 10D). Therapeutic effects of Tof@HA-TAGel in vivo. (A) Representative thermographic images of healthy knees or OA knees with different treatments. (B) H&E staining of inflammatory cell infiltration in the synovial tissue after different treatments. (C) Quantification of temperature in knees with different intervention. (D) OARSI score of rat mouse knee. In vivo luminescence images (G) and the relative L-012 luminescence intensity (E) of healthy knees and OA knees in mice after different treatments. (F)Relative CGRP expression were analyzed in each group. (H) Representative images of immunofluorescence staining of CGRP-positive sensory nerve fibers in the synovium of the mice after ACLT surgery. Data were presented as mean ± SD (n = 6). ∗∗p < 0.01, ∗∗∗p < 0.001, compared to the PBS group. To assess the in vivo ROS-scavenging efficacy of Tof@HA-TAGel, the luminescent probe L-012 was intra-articularly administered to OA mice, with subsequent imaging of ROS distribution by IVIS imaging [58]. A marked increase in knee fluorescence was observed in the PBS group when compared with the sham group. (Fig. 10G). All treatment groups (Tof, TAGel, HA-TAGel, and Tof@HA-TAGel) showed reduced fluorescence intensities, with Tof@HA-TAGel demonstrating the most potent scavenging effect (p < 0.001, Fig. 10E). CGRP, a potent vasodilator implicated in pain sensitization [59,60]. Previous studies demonstrate that knee osteoarthritis pain in mice closely correlates with the ingrowth of CGRP-positive neurons [21]. As illustrated in Fig. 10F and H, the CGRP fluorescence intensity in the knee synovium of the Tof@HA-TAGel treated group was substantially reduced compared to that in the PBS group (p < 0.01). Together, our findings establish that Tof@HA-TAGel mitigates OA inflammation and oxidative stress while reducing synovial sensory nerve ingrowth. Joint protective effect of Tof@HA-TAGel Progressive cartilage destruction in OA primarily results from the gradual loss of extracellular matrix (ECM) [61]. The cartilage and ECM integrity were examined using H&E and Safranin O-Fast Green staining (Fig. 11A). The mice from sham group maintained well-preserved cartilage architecture with robust proteoglycan retention evidenced by homogeneous Safranin O staining, whereas OA cartilage displayed marked surface fibrillation and profound proteoglycan depletion, confirming inflammatory-driven tissue degeneration. Treatment with Tof, TAGel, HA-TAGel, or Tof@HA-TAGel alleviated these degenerative changes. Notably, the cartilage morphology in Tof@HA-TAGel group closely approximated to that in the sham groups, featuring smooth surfaces and robust proteoglycan preservation, demonstrating significant mitigation of ECM degradation. Joint protection effect of Tof@HA-TAGel in vivo. (A) H&E staining and Safranin O-fast green staining. (B) Representative micro-CT images of knee joints in normal mice or OA mice after different treatments. Quantitative analyses of BV/TV (C), Tb.Sp (D) and Tb.N (E) of subchondral bone in different groups. Data were presented as mean ± SD (n = 6). ∗∗p < 0.01, ∗∗∗p < 0.001, compared to the PBS group. (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 of subchondral bone morphology and microstructure was performed to further validate the therapeutic effect of Tof@HA-TAGel (Fig. 11B). Bone structural parameters, including BV/TV, Tb.N and Tb.Sp were obtained through morphometric analysis. The mice in PBS group exhibited severe bone and cartilage erosion alongside extensive osteophyte formation, accompanied by significantly reduced BV/TV and Tb.N and increased Tb.Sp. Collectively, it is indicated an imbalance state between bone resorption and bone formation. In contrast, all treatment knees (Tof, TAGel, HA-TAGel, and Tof@HA-TAGel) showed attenuated degenerative changes and reduced osteophyte formation. Importantly, the best-preserved subchondral bone architecture was displayd in the Tof@HA-TAGel group among all treatment groups, with significantly higher BV/TV (p < 0.001; Fig. 11C) and Tb.N (p < 0.001; Fig. 11D), and the lowest Tb.Sp (p < 0.001; Fig. 11E). These results demonstrate that Tof@HA-TAGel effectively inhibits subchondral bone destruction and protects articular cartilage from structural deterioration. In all, Tof@HA-TAGel showed the excellent prominent advantage in protecting cartilage and alleviating OA progression. Biocompatibility of Tof@HA-TAGel in vivo The ideal intra-articular injectable biomaterial must combine therapeutic efficacy with high biocompatibility. Histopathological assessment of major organs (heart, liver, spleen, lungs, and kidneys) by H&E staining confirmed that Tof@HA-TAGel induced no significant abnormalities across all treatment groups, supporting its excellent in vivo safety profile (Fig. S7). ConClusion Here, we elucidated the relationship between macrophage polarization and Netrin-1-guided sensory nerve ingrowth. Accordingly, macrophage-targeting microspheres (Tof@HA-TAGel) were designed to inhibit M1 polarization and alleviate OA associated pain. The microspheres enable controlled release of the TA and the inner encapsulated Tof. The surface modification of HA enhanced the targeting specificity of microspheres towards M1 macrophages and synovium. Both in vitro and in vivo experiments demonstrated that Tof@HA-TAGel can inhibit M1 polarization, further downregulate inflammatory factors and the axonal guidance factor Netrin-1. Meanwhile, the JAK1/STAT3 pathway was identified as a critical regulator of Netrin-1-mediated nerve ingrowth. In vivo evaluation confirmed the efficacy of Tof@HA-TAGel in protecting inflamed articular cartilage, alleviating OA pain, and slowing OA progression. Collectively, this work positions the M1 macrophage-targeted microsphere as a promising therapeutic paradigm for OA, achieving simultaneous pain alleviation and disease modification through minimally invasive modulation of the inflammatory niche. Funding This research was partially supported by National Natural Science Foundation of China (82102600), China; National Natural Science Foundation of China (82272485), China;and Shandong Provincial Natural Science Foundation (ZR2025MS1391), China. Draft of Scheme 1 was created with BioRender.com. CRediT authorship contribution statement Guojiang Zhang: Conceptualization, Data curation, Formal analysis, Writing – original draft, Writing – review & editing. Laibo Zhang: Conceptualization, Data curation, Formal analysis, Writing – original draft, Writing – review & editing. Xiangdong Wen: Data curation, Formal analysis. Guangxian Liu: Formal analysis, Methodology. Dengju Li: Methodology. Chuanhong Qin: Methodology. Senbo An: Conceptualization, Funding acquisition, Project administration, Supervision, Validation, Writing – review & editing. Yawei Sun: Conceptualization, Funding acquisition, Project administration, Supervision, Validation, Writing – review & editing. Shui Sun: Conceptualization, Funding acquisition, Project administration, Supervision, Validation, Writing – review & editing. Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influencethe work reported in this paper. Footnotes Supplementary data to this article can be found online at https://doi.org/10.1016/j.mtbio.2026.103178. Contributor Information Senbo An, Email: ansenbo@sdfmu.edu.cn. Yawei Sun, Email: yaweisun@tju.edu.cn. Shui Sun, Email: sunshui@sdfmu.edu.cn. Supplementary data The following are the supplementary data to this article: Multimedia component 1 Multimedia component 2 Multimedia component 3 Multimedia component 4 Data availability Data will be made available on request. References Associated Data Supplementary Materials Multimedia component 1 Multimedia component 2 Multimedia component 3 Multimedia component 4 Data Availability Statement Data will be made available on request.