Targeted knockdown of Piezo1 in synovial macrophages attenuates osteoarthritis development

Abstract

Objective

Osteoarthritis (OA) is a prevalent degenerative joint disease worldwide. Emerging therapies targeting the crosstalk between immune/inflammatory cells and chondrocytes have shown promise. Macrophage phenotypic reprogramming represents a potential therapeutic strategy, yet the molecular mechanisms by which mechanical signals regulate macrophage plasticity remain unclear. This study aimed to investigate the role of the mechanosensitive ion channel Piezo1 in synovial macrophage polarization and its contribution to OA pathogenesis.

Methods

Histological analyses were performed on synovial tissues from human OA patients and OA mouse models to assess Piezo1 expression in macrophages. Conditional Piezo1 knockout in macrophages was established in mice to evaluate its effect on OA progression. In vitro and in vivo experiments were conducted to explore the impact of Piezo1 deletion on macrophage polarization and chondrocyte metabolism. Mechanistic studies investigated the involvement of the DRP1-cGAS-STING axis in Piezo1-mediated inflammasome activation. Furthermore, mannose-modified liposomes carrying Si-Piezo1 were constructed to selectively target and inhibit Piezo1 expression in synovial macrophages.

Results

Piezo1 expression was significantly upregulated in synovial macrophages from OA joints compared to healthy joints. Macrophage-specific deletion of Piezo1 markedly alleviated OA symptoms and promoted chondrocyte anabolism. Mechanistically, Piezo1 facilitated M1 macrophage polarization by activating the NLRP3 inflammasome via the DRP1-cGAS-STING pathway, which in turn accelerated chondrocyte senescence and degeneration. Targeted delivery of Si-Piezo1 nanoparticles effectively suppressed Piezo1 expression in synovial macrophages, reduced the proportion of M1 macrophages, and alleviated OA progression in vivo.

Conclusion

Piezo1 plays a critical role in regulating synovial macrophage polarization through mechanotransduction, thereby promoting OA progression. Targeted inhibition of Piezo1 using mannose-modified nanoparticles provides a promising therapeutic strategy for OA treatment.

Translational potential

By offering experimental evidence on the role and mechanism of Piezo1 in OA synovium, this study underscores the potential of Man-LNP@Si-Piezo1 as a therapeutic strategy for OA.

Graphical abstract

1. Introduction

As is well documented, osteoarthritis (OA) is a prevalent chronic joint disorder characterized by progressive cartilage degeneration and secondary synovial inflammation [1]. Its clinical manifestations primarily include pain, swelling, and limited mobility [2]. Currently, its treatment predominantly relies on joint replacement surgery, which fails to precisely target the complex pathophysiological mechanisms [3]. Consequently, investigating the pathological mechanisms underlying OA has become a major focus and an area of intense interest in orthopedic research.

Earlier studies have established that OA is a disease driven by low-grade inflammation, with its pathological mechanisms largely mediated by chronic inflammatory cell infiltration, particularly macrophages [4,5]. The OA synovium exhibits infiltration of chronic inflammatory cells. This persistent inflammatory microenvironment leads to dysregulated macrophage phenotypic polarization. According to previous studies, the M1/M2 macrophage ratio in knee OA patients is positively correlated with Kellgren–Lawrence grades, suggesting that an imbalance in M1/M2 macrophages is strongly associated with OA severity [6,7]. Furthermore, the accumulation of M1 macrophages in OA synovial tissue exacerbates experimental OA and the proportion of M2 macrophages does not differ significantly between normal and OA synovium, highlighting the pivotal role of M1 macrophage polarization in OA progression [8,9].

The initial inflammatory response in OA is triggered by damage-associated molecular patterns (DAMPs), which are endogenous molecules released into the extracellular matrix following cartilage damage [10]. Persistent protease-mediated degradation of cartilage and meniscal tissues produces generates DAMPs, including glycosaminoglycans and tenascin-C. In turn, these molecules promote M1 macrophage polarization, initiating a vicious cycle of inflammation and tissue destruction [11,12]. The phagocytic function of macrophages is regulated by various factors, among which the dynamic reorganization of the cytoskeleton plays a critical regulatory role. Piezo1, a mechanosensitive calcium ion channel, plays a decisive role in regulating cytoskeletal remodeling, pseudopodia formation, and cellular stiffness in macrophages [13]. While its hyperactivation has been shown to enhance inflammation and accelerate disease progression in multiple pathological contexts, its role in arthritis models remains unexplored [14]. However, the mechanism by which Piezo1 regulates OA progression by influencing immune cells in the joint microenvironment, particularly macrophages, remains to be elucidated.

Mitochondrial dynamics, involving fusion and fission, modulate mitochondrial morphology and function to meet cellular metabolic demands. Imbalances in mitochondrial dynamics may promote M1 macrophage polarization, while mitochondrial dysfunction and associated catabolic effects drive OA progression [[15], [16], [17]]. Dynamin-related protein 1 (DRP1), a GTPase associated with mitochondrial fission, is tightly regulated by post-translational modifications, including phosphorylation, SUMOylation, and citrullination, which play vital roles in metabolic reprogramming, inflammation, and senescence [18,19]. However, the effects of Piezo1 in modulating mitochondrial dynamics and function in synovial macrophages during OA progression remains unclear.

In this study, we aimed to investigate whether Piezo1 modulates mitochondrial dynamics and macrophage polarization in OA. We hypothesized that Piezo1 activation promotes pro-inflammatory macrophage phenotypes by regulating DRP1-mediated mitochondrial fission and inflammatory signaling pathways.

2. Materials and methods

Ethical statement

The experimental procedures for animal care and use were performed in accordance with the Guidelines for the Care and Use of Laboratory Animals provided by the U.S. National Institutes of Health and were approved by the Animal Research Committee of Hebei Medical University（IACUC-Hebmu-2025046）. Human synovial tissues were collected from two groups: patients with osteoarthritis (OA group) and patients with tibial plateau fractures but without OA (control group). All procedures were conducted in accordance with the ethical standards of the institutional review board ( approval number: 2024-K-351-01), and informed consent was obtained from all participants.Freshly isolated synovial tissue was then immediately rinsed in PBS to remove residual blood and debris. Tissue was fixed, dehydrated, and embedded in paraffin.

2.1. Antibodies and reagents

Mdivi-1, KN-93, β-galactosidase staining kit, and JC-1 were purchased from Beyotime (Shanghai, China). EDTA solution, Safranin O-Fast Green (S&O), hematoxylin and eosin (H&E), Tween 80, toluidine blue staining solution, and lipopolysaccharide (LPS) were procured from Solarbio (Beijing, China). Dimethyl sulfoxide (DMSO) was acquired from Sigma Chemical Co. (St. Louis, MO, USA). MitoSox and MitoTracker Deep Red FM were obtained from Thermo Scientific (UT, USA). Secondary antibodies, goat anti-rabbit IgG H&L (Alexa Fluor ® 488) and goat anti-mouse IgG H&L (Alexa Fluor ® 594), were purchased from Abcam (Cambridge, MA, USA). HRP-conjugated goat anti-mouse IgG (H + L) and goat anti-rabbit IgG (H + L) were acquired from Zen Bioscience (Chengdu, China).

2.2. Animals

Piezo1 cKO mice were generated by crossing LysmCreERT2 transgenic mice with Piezo1^fl/fl mice. To activate Cre recombinase in macrophages, tamoxifen was fully dissolved in corn oil and intraperitoneally administered at a dose of 40 mg/kg. The tamoxifen-induced gene deletion was performed once daily for 5 consecutive days, with destabilization of the medial meniscus (DMM) surgery carried out after the third injection. Piezo1^fl/fl mice administered tamoxifen under identical conditions served as controls. Four-week-old male mice were typically used for in vitro experiments. Genetically modified Piezo1^fl/fl mice were obtained from Jackson Laboratory (Stock No. 029213). All mice were housed in a specific pathogen-free facility. The experimental subjects were 12-week-old male mice.In this study, knee OA was induced by DMM surgery, as described in a previous study [20].The tibial ligament of the medial meniscus in the right knee joint of each mouse was exposed via a medial parapatellar capsular incision and transected using a microsurgical scalpel. In the sham control group, the right knee joint underwent arthrotomy without transection of the medial meniscotibial ligament.

2.3. Micro-computed tomography (Micro-CT)

Samples were scanned using a Micro-CT system (100 kV, 98 μA; 12 μm resolution; Bruker Skyscan 1176, Luxembourg, Belgium). Image reconstruction was performed using DataViewer (v1.5, Bruker), data processing using CTAn (v1.9, Bruker), and 3D visualization using CTVox (v3.3, Bruker). Knee joint parameters were analyzed with a focus on the osteophyte region.

2.4. Histopathological analysis

Freshly isolated human synovial tissues and mouse knee joints were fixed in 4% paraformaldehyde (Sigma–Aldrich) for 24 h and then mouse knee joints decalcified in 10% EDTA solution for four weeks. Next, the samples were dehydrated and embedded in paraffin in the sagittal plane. 5-μm-thick sections were stained with S&O and H&E. Histological grading of OA in knee joints was performed according to the Osteoarthritis Research Society International (OARSI) system [21].

2.5. Tissue immunofluorescence staining

Sections of knee joints were deparaffinized, rehydrated in graded ethanol, and subjected to antigen retrieval overnight using EDTA buffer. Afterward, they were treated with 0.3% Triton X-100 for 15 min, blocked with 5% BSA for 1 h, and incubated overnight at 4 °C with primary antibodies against F4/80 (1:100, Invitrogen, 14-4801-82), iNOS (1:300, Servicebio, GB11119), and NLRP3 (1:200, Servicebio, GB114320). Subsequently, sections were incubated with secondary antibodies conjugated to Alexa Fluor 488 or 594 (1:500) at room temperature for 1 h, followed by DAPI staining for 1 min.Chondrocytes were seeded onto glass coverslips and cultured for 24 h. Mitochondria were labeled using MitoTracker Deep Red FM (100 nM; Thermo Scientific, UT, USA) by incubation at 37 °C for 30 min. After washing, cells were fixed with 4% paraformaldehyde, permeabilized with 0.1% Triton X-100 (Solarbio Science & Technology, China), and blocked with 5% goat serum at room temperature. Subsequently, cells were incubated overnight at 4 °C with the following primary antibodies: INOS (1:300), P21 (1:200), ADAMTS4 (1:200), and DRP1 (1:200). The next day, cells were incubated with Alexa Fluor® 488- or 594-conjugated secondary antibodies (1:400) for 45 min, followed by nuclear staining with DAPI (Solarbio) for 60 s at room temperature. Immunofluorescence images were captured using a fluorescence microscope or confocal laser scanning microscope (Olympus, Tokyo, Japan). Co-localization analysis was performed using ImageJ software (Fiji distribution).

2.6. Immunohistochemical (IHC) staining

IHC was performed to examine the expression and polarization of iNOS, Piezo1, and IL-6 in synovial macrophages. Additionally, ADAMTS4 and Aggrecan staining were performed to evaluate cartilage degeneration. Antibodies used included anti-ADAMTS4 (1:100, Servicebio, GB11807), anti-IL-6 (1:200, Servicebio, GB11117), anti-Aggrecan (1:200, Servicebio, 115687-100), anti-Piezo1 (1:300, ProteinTech, 15939-1-AP), and anti-iNOS (1:300, Servicebio, GB11119). IHC images were quantified using ImageJ software to calculate the integrated optical density (IOD).

2.7. Cell stimulation and Co-culture

Bone marrow cells were cultured in DMEM supplemented with 10% FBS and 20 ng/mL M-CSF for seven days to generate bone marrow-derived macrophages (BMDMs). Following PBS washes, cells were treated with 100 ng/mL LPS for 24 h to induce inflammation, and the supernatant was subsequently collected and co-cultured with chondrocytes for an additional 24 h.RAW264.7 macrophages were cultured in DMEM supplemented with 10% fetal bovine serum and 1% penicillin-streptomycin at 37 °C in a 5% CO₂incubator. To generate a Piezo1 knockout cell line, cells were transfected with a plasmid encoding Cas9 and guide RNAs targeting the Piezo1 gene. Piezo1-KO and wild-type(WT) RAW264.7 cells were subsequently stimulated with LPS (100 ng/mL) for 24 h before analysis.

2.8. Flow cytometry

BMDMs were induced with LPS under defined experimental conditions. M1 macrophages were identified by flow cytometry using the following antibodies: anti-CD86-eFluor 450 (1:100, Invitrogen, 48-0862-82) and anti-F4/80-Brilliant Violet™ 650 (1:100, Invitrogen, 416-4801-82). Cells were analyzed using a BD FACSCelesta Flow Cytometer (BD Biosciences, USA).

2.9. Western blot analysis

Total proteins from BMDMs were extracted using RIPA lysis buffer, and protein concentrations were determined using a BCA protein assay kit (Beyotime, China). Thereafter, the proteins were separated by SDS-PAGE and subsequently transferred onto PVDF membranes (Bio-Rad, USA). The membranes were blocked with 5% non-fat milk for 2 h, incubated with primary antibodies at 4 °C overnight, and then incubated with appropriate secondary antibodies (Bioworld Technology, USA) at room temperature for 2 h. Signal intensities were quantified using Image Lab 3.0 software (Bio-Rad). Information on primary antibody dilutions is shown in Table 2.

Table 2.

The detail information of primary antibodies.

Antibodies	Cat. No.	Company	Concentration for WB	Concentration for IF
Piezo1	15939-1-AP	Proteintech Group (Chicago, IL, USA)	/	1:100
Aggrecan	ab3778	Abcam (Cambridge, MA, USA).	1:1000	1:100
ADAMTS4	82744-2-RR	Proteintech Group (Chicago, IL, USA)	1:1000	1:200
GSDMD-N	DF12275	Affinity Biosciences (Jiangsu, China)	1:1000	/
P-CaMKII	AF3493	Affinity Biosciences (Jiangsu, China)	1:1000	/
P-DRP1(s616)	AF8470	Affinity Biosciences (Jiangsu, China)	1:1000	/
DRP1	DF7037	Affinity Biosciences (Jiangsu, China)	1:1000	1:100
NLRP3	381207	zen-bioscience (Chengdu, China)	1:1000	1:100
Collagen II	28459-1-AP	Proteintech Group	1:1000	/
MMP13	18165-1-AP	Proteintech Group	1:1000	1:100
β-actin	20536-1-AP	Proteintech Group	1:1000	/
Caspase1	sc-56036	Santa Cruz Biotechnology (CA, USA)	1:400	/
CaMKII	sc-5306	Santa Cruz Biotechnology (CA, USA)	1:400	1:100
IL-18	WL01127	Wanlei(Shenyang, China)	1:1000	/
IL-1β	WL00891	Wanlei(Shenyang, China)	1:1000	/
COXIV	WL02203	Wanlei(Shenyang, China)	1:1000	/
INOS	GB115703	Servicebio(Wuhan, China)	1:1000	1:100
TNF-α	GB11188	Servicebio(Wuhan, China)	1:1000	/
IL-6	GB11117	Servicebio(Wuhan, China)	1:1000	/
P21	GB115313	Servicebio(Wuhan, China)	1:1000	1:100
P16	28416-1-AP	Proteintech Group (Chicago, IL, USA)	1:1000	/
cGAS	84045-1-RR	Proteintech Group (Chicago, IL, USA)	1:1000	/
STING	19851-1-AP	Proteintech Group (Chicago, IL, USA)	1:1000	/

2.10. Co-immunoprecipitation (CO-IP)

CO-IP was performed using a Protein A/G Magnetic Beads Immunoprecipitation Kit (Beyotime) according to the manufacturer's instructions. Protein extraction was performed as previously described. Dynabeads conjugated with CaMKII or DRP1 primary antibodies were incubated overnight at 4 °C. Immunocomplexes were analyzed by Western blot.

2.11. Immunofluorescence staining

BMDMs were seeded onto glass coverslips and cultured for 24 h. Cells were stained with MitoTracker Deep Red FM dye (100 nM) and incubated at 37 °C for 30 min (Thermo Scientific). Cells were then fixed with 4% paraformaldehyde, permeabilized with 0.1% Triton X-100, and blocked with 5% goat serum. They were subsequently incubated with the primary antibodies (DRP1, MMP13, ADAMTS5, Aggrecan, Collagen II; 1:200 dilution) incubated overnight at 4 °C, followed by incubation with secondary antibodies (Alexa Fluor® 488 or Fluor® 594, 1:400 dilution) for 45 min at room temperature. Nuclei were counterstained with DAPI for 60 s. Fluorescence images were captured using an Olympus or confocal microscope and analyzed using ImageJ software.

2.12. Mitochondrial ROS and ΔѰm detection

Mitochondrial ROS levels and ΔѰm were assessed using MitoSox Red and JC-1 probes, respectively (Yeasen Biotech Co., Ltd), following the manufacturer's instructions. The results were visualized using a fluorescence microscope or flow cytometer for further analysis.

2.13. Mitochondrial morphology

Cells were seeded into confocal culture dishes and stained with Mitotracker™ Red (100 nM) at 37 °C for 30 min. The stained cells were analyzed using the Mitochondria Analyzer plugin in FIJI software. Mitochondrial fragmentation was quantified based on the form factor (FF and aspect ratio (AR).FF was calculated as (Perimeter²)/(4π × Area), reflecting mitochondrial branching complexity, while AR was defined as the ratio of the major axis to the minor axis, representing mitochondrial elongation.Morphologically, lower FF and AR values (close to 1) correspond to rounded and fragmented mitochondria, whereas increased FF and AR values indicate more elongated and fused mitochondrial networks.

2.14. Mitochondrial permeability transition pore (mPTP) opening detection

The mPTP opening was assessed using the mPTP Assay Kit (Beyotime) and confocal microscopy, following a previously published protocol [22]. Briefly, cells were incubated with 2 μM Calcein-AM and 100 nM MitoTracker Deep Red in the dark for 30 min. After washing with PBS three times, cells were treated with 2 mM CoCl2 for 15 min. Calcein-AM fluorescence, retained within the mitochondrial matrix, diminishes upon mPTP opening due to its release into the cytosol and subsequent binding to CoCl2. Images were captured using a confocal microscope, and the mPTP opening levels were quantified as the MitoTracker/calcein fluorescence ratio using ImageJ software.

2.15. Intracellular ATP levels

Intracellular ATP concentrations were quantified using an ATP Assay Kit (Beyotime) following the manufacturer's instructions. ATP levels were quantified using a fluorescence plate reader (PerkinElmer VICTOR Nivo, MA, USA) and normalized to the control group for comparison.

2.16. Transmission electron microscopy (TEM)

BMDMs were fixed with an electron microscopy fixation solution (Servicebio, Wuhan, China) at room temperature for 2 h. The fixed cells were subsequently embedded and stained with toluidine blue. Images were captured using a transmission electron microscope (Hitachi High-Technologies Corp, Tokyo, Japan).

2.17. Calcium imaging

BMDMs extracted from Piezo1^fl/fl and Piezo1 cKO mice were seeded onto glass-bottom culture dishes (NEST Biotechnology, Wuxi, China) and incubated with Fluo-4 AM (Beyotime) at 37 °C for 30 min. Real-time imaging was conducted using a confocal microscope at a frame rate of 1 frame per second for 5 min. Data were analyzed using NIS Elements software (Nikon, Tokyo, Japan). To investigate the effect of Piezo1 on calcium release, LPS or Yoda1 were added 1 min after imaging commenced.

Synthesis of mannose-modified liposomes (Man-LNP) @si-RNA Soybean phosphatidylcholine dioleoyl-3-trimethylammonium propane,1,2-distearoyl-sn-glycero-3- phosphoethanolamine-Polyethylene Glycol-mannose, and cholesterol were dissolved together in 3 mL of anhydrous ethanol and transferred to a pear-shaped flask. si-RNA was dissolved in a citrate buffer (50 mM, pH 4.0) containing 25% ethanol and slowly added to the lipid solution under gentle stirring. The mixture was incubated for 20 min to allow complex formation. The resulting solution was subjected to ultrasonic treatment, followed by extrusion through a liposome extruder equipped with a 100 nm polycarbonate membrane to obtain liposomes with uniform size.

2.18. SA-β-galactosidase staining

Cellular senescence in chondrocytes was evaluated using an SA-β-Galactosidase Staining Kit (Beyotime, Shanghai, China) according to the manufacturer's instructions. After washing with PBS, chondrocytes were fixed with 1 mL of fixation buffer for 15 min, washed three times with PBS, and incubated overnight with SA-β-galactosidase staining solution at 37 °C. Senescent cells displaying blue staining were observed under a light microscope (OLYMPUS, Japan).

2.19. Toluidine blue and safranin O staining

Cells were washed twice with PBS for 1 min each time, followed by staining with toluidine blue (Solarbio, China) for 15 min or Safranin O (Solarbio, China) for 30 min. Cells were washed three times with ddH₂O and then imaged.

2.20. Statistical analysis

Statistical analyses were carried out using Prism software (version 9.0; GraphPad Software). One-way ANOVA was applied for multigroup comparisons, while comparisons between two groups were performed using the Student's t-test. The results were presented as the mean ± standard deviation (SD), with a P-value <0.05 considered statistically significant.

3. Results

3.1. Increased Piezo1 expression in synovial macrophages of the OA synovium

To investigate the expression of Piezo1 in OA, human healthy and OA samples were collected for histological staining. We first summarized the baseline demographic and clinical information of the study participants. As shown in Table 1, a total of 5 patients with end-stage osteoarthritis and 5 trauma patients without osteoarthritis were included. HE staining confirmed the diagnosis of synovitis (Fig. 1A–C). Compared to control samples, qPCR and IHC revealed higher expression levels of Piezo1 and iNOS-positive cells in OA synovium (Fig. 1A–D-E Fig. S1). Similarly, consistent with human OA samples, HE and IHC staining of OA tissue sections from C57BL/6J mice subjected to DMM and anterior cruciate ligment transection (ACLT) surgery displayed aggravated synovitis, as well as increased expression levels of Piezo1 and an elevated proportion of iNOS-positive cells (Fig. 1B, C-E Fig. S2). As anticipated, Piezo1 protein levels were elevated in human OA synovium (Fig. 1F and G). Moreover, the proportion of double-positive cells for Piezo1 and the macrophage marker CD68 was significantly higher in human OA synovium compared to controls (Fig. 1H and I). Likewise, in murine synovial samples, immunohistochemical staining unveiled that the proportion of double-positive cells for Piezo1 and F4/80 was significantly higher in OA mice compared to controls (Fig. 1J and K).In parallel, immunofluorescence further demonstrated increased co-localization of Piezo1 with iNOS in OA synovium(Fig. S3), Taken together, these findings suggest that Piezo1 may serve as a potential regulator of macrophages in the synovium of OA.

Table 1.

The baseline characteristics of the OA and control groups.

Group	n	Age (years, mean ± SD)	Sex (M/F)	BMI (mean ± SD)	KL Grade	Major Comorbidities
OA	5	65.4 ± 7.3	2/3	26.2 ± 3.1	III–IV	Hypertension, Diabetes
Control	5	63.1 ± 8.2	2/3	24.9 ± 2.7	0	None reported

Fig. 1.

Increased expression of Piezo1 in OA synovial macrophages.(A) HE staining and immunohistochemical staining for iNOS and Piezo1 in normal and OA human synovium.(B) HE staining and immunohistochemical staining for iNOS and Piezo1 in normal and OA mouse synovium. (C–E) Quantification of synovitis scores, iNOS-positive cells, and Piezo1-positive cells as a proportion of total cells in normal and OA human and mouse synovium.(F–G) Western blot analysis and statistical comparison of Piezo1 expression in normal and OA human synovium.(H–K) Co-immunostaining of Piezo1 with CD68 in normal and OA human synovium and Piezo1 with F4/80 in normal and OA mouse synovium, along with quantitative analysis.Data are presented as mean ± SD. ∗P < 0.05; ∗∗P < 0.01; ∗∗∗P < 0.001; NS: not significant.

3.2. Piezo1 regulates osteoarthritis in mice by modulating macrophage polarization

To explore the role of macrophage Piezo1 in OA, a Lyz2-CreERT2 knock-in mouse model was constructed by crossing Piezo1^fl/fl mice with Lyz2-CreERT2 mice, yielding Piezo1^fl/fl-Lyz2-CreERT2 (Piezo1 cKO) mice to specifically eliminate Piezo1 in macrophages.The efficiency of Piezo1 deletion was validated by immunofluorescence staining of synovial tissues in vivo, which showed a markedly reduced Piezo1 signal in synovial macrophages from Piezo1 cKO mice compared with controls(Fig. S4). Knee joint tissues were harvested 6 weeks after DMM surgery for histological assessment,Micro-CT imaging revealed that the volume and number of osteophytes were lower in Piezo1 cKO mice compared to Piezo1fl/fl following DMM surgery(Fig. 2A–C). Additionally, HE and S&O staining demonstrated that Piezo1 cKO mice exhibited a cartilage layer with higher collagen content, intact cartilage surfaces, fewer synovial lining layers, lower stromal cell density, and reduced inflammatory infiltration levels compared to Piezo1^fl/fl mice. Similarly, synovitis and OARSI scores were significantly lower in Piezo1 cKO mice compared to Piezo1^fl/fl mice (Fig. 2D–F). At the same time, immunofluorescence staining of OA synovial tissues illustrated reduced macrophage aggregation and a lower proportion of iNOS-positive M1 macrophages in Piezo1 cKO mice (Fig. 2G–H,Fig. S5). Immunohistochemical analysis further indicated higher Aggrecan expression levels and lower ADAMTS4 expression levels in Piezo1 cKO mice compared to controls (Fig. 2I–K). Besides, the level of IL-6, an indicator of synovitis severity in OA, was significantly lower in Piezo1 cKO synovial tissues compared to controls (Fig. 2I–L) [23]. These in vivo results conjointly suggest that Piezo1 may mediate OA progression by regulating macrophage polarization.

Fig. 2.

Piezo1 regulates osteoarthritis in mice by modulating macrophage polarization.(A) 3D micro-CT images of the anterior and lateral views of the knee joints of Piezo1 cKO and Piezo1 fl/fl mice following DMM surgery.(B–C) Quantification of osteophyte number and volume in the regions of interest (ROIs), with ROIs marked in green.(D) H&E staining and Safranin O/Fast Green staining of joint cartilage in Piezo1 cKO and Piezo1 fl/fl mice.(E–F) Quantification of synovitis scores and OARSI scores in the synovium and cartilage of Piezo1 cKO and Piezo1 fl/fl mice.(G) Immunostaining for Piezo1 and F4/80 in the synovium of Piezo1 cKO and Piezo1 fl/fl mice with (H) quantitative analysis.(I–L) Immunohistochemical analysis of Aggrecan, ADAMTS4, and IL-6 in the cartilage of Piezo1 cKO and Piezo1 fl/fl mice, with (J–L) quantitative analysis.Data are presented as mean ± SD. ∗P < 0.05; ∗∗P < 0.01; ∗∗∗P < 0.001; NS: not significant.

3.3. Macrophage-specific Piezo1 knockout alleviates M1 macrophage polarization

LPS increases in synovial fluid and serum of OA patients and is vital in the pathogenesis and severity of structural abnormalities and symptoms of knee OA [24].To investigate the role of Piezo1 in LPS-induced M1 macrophage polarization, BMDMs were isolated from Piezo1 cKO and Piezo1^fl/fl mice and exposed to LPS for 24 h in vitro (Fig. 3A, Fig.S6). Western blot analysis uncovered that Piezo1 knockout down-regulated iNOS expression in LPS-treated macrophages compared to Piezo1^fl/fl controls (Fig. 3A and B), in agreement with the results of immunofluorescence staining (Fig. 3D and E). Similarly, Piezo1 knockout significantly down-regulated the LPS-induced expression of IL-1β, IL-6, and TNF-α (Fig. 3F and I, Fig.S7). Meanwhile, flow cytometry analysis showed that the proportion of F4/80⁺CD86⁺ M1 macrophages was significantly lower in Piezo1 cKO BMDMs compared to Piezo1^fl/fl controls (Fig. 3J and K). Additionally, CRISPR/Cas9 genome editing was employed to knock down Piezo1 in RAW264.7 macrophages (Piezo1-KO),PCR analysis revealed a marked reduction in Piezo1 expression, confirming efficient Piezo1 knockdown(Fig. S8). As expected, the LPS-induced upregulation of iNOS was significantly reversed in Piezo1-KO cells, as evidenced by the results of both Western blot analysis and immunofluorescence staining (Fig. 3L–O). Flow cytometry analysis further revealed a lower proportion of CD86⁺ cells in Piezo1-KO macrophages compared to WT macrophages (Fig. 3P and Q). These results strongly suggest that Piezo1 knockout mitigates M1 polarization in both BMDMs and RAW264.7 cells.

Fig. 3.

Knockdown of macrophage Piezo1 alleviates M1 macrophage polarization. Primary BMDMs were isolated and purity was determined by flow cytometry (A).Primary BMDMs were induced from Piezo1 cKO and Piezo1 fl/fl mice, followed by 24-h LPS stimulation.(B–C) Western blot and quantitative analysis of iNOS in BMDMs.(D–E) Immunofluorescence and quantitative analysis of iNOS in BMDMs.(F–I) Western blot and quantitative analysis of IL-6, TNF-α, and IL-1β in BMDMs.(J–K) Flow cytometric detection and quantification of F4/80+ CD86⁺ cells in BMDMs.RAW264.7 cells from WT and Piezo1-KO mice were stimulated with LPS for 24 h(L–M) Western blot and quantitative analysis of iNOS in RAW264.7 cells.(N–O) Immunofluorescence and quantitative analysis of iNOS in RAW264.7 cells.(P–Q) Flow cytometric detection and quantification of CD86⁺ cells in RAW264.7 cells.Data are presented as mean ± SD. ∗P < 0.05; ∗∗P < 0.01; ∗∗∗P < 0.001; NS: not significant.

3.4. Macrophage-specific Piezo1 knockout attenuates chondrocyte senescence and extracellular matrix (ECM) degradation

To elucidate the effects of macrophage-specific Piezo1 on cartilage tissues, BMDMs were isolated from Piezo1 cKO and Piezo1^fl/fl mice, treated with LPS for 24 h, and then co-cultured with chondrocytes isolated from neonatal WT mice for an additional 24 h. The chondrocyte phenotype was verified by Alcian blue staining and immunofluorescence staining for type II collagen (Fig. S9).Western blot analysis demonstrated that Piezo1 knockdown in macrophages limited cartilage degeneration and senescence in chondrocytes, as indicated by the altered expression of Aggrecan, ADAMTS4, p21, and p16 (Fig. S10A–F). Similarly, SA-β-gal staining showed a lower number of senescent chondrocytes after treatment with M1-CM from Piezo1 cKO macrophages compared to controls (Fig. S10G–H). Toluidine blue staining delineated that acidic polysaccharide levels in chondrocytes, reduced by LPS-treated macrophage-conditioned medium, were increased following Piezo1 knockout, leading to deeper purple staining (Fig. S10I). Safranin O staining indicated that LPS-treated CM reduced collagen content in chondrocytes, while Piezo1 knockout preserved collagen levels, leading to deeper red staining (Fig. S10I). Lastly, immunofluorescence analysis validated that Piezo1 knockout significantly down-regulated ADAMTS4 and p21 expression in chondrocytes, in agreement with Western blot results (Fig. S10J–L). Overall, these findings suggest that Piezo1 knockout inhibits M1-CM-induced chondrocyte senescence and ECM degradation.

3.5. Knockout of macrophage Piezo1 alleviates mitochondrial dysfunction

Earlier studies have established that mitochondrial function is closely associated with macrophage phenotype reprogramming [25]. Therefore, the effect of Piezo1 on mitochondria in macrophages was explored. MitoTracker and Calcein fluorescence labeling were employed to assess mPTP opening in LPS-induced macrophages. As depicted in Fig. S11A and B, mitochondrial calcein fluorescence intensity was lower in the LPS + Piezo1^fl/fl group, whereas Piezo1 knockout increased this intensity, suggesting that Piezo1 knockout minimizes excessive mPTP opening in macrophages. JC-1 assays revealed that LPS decreased mitochondrial membrane potential, which was significantly restored in Piezo1 cKO BMDMs (Fig. S11C and D). Similarly, mitochondrial ATP production was markedly enhanced in Piezo1 cKO BMDMs (Fig. S11E).

Mitochondrial structure is essential for its function. TEM revealed that mitochondria in both the Piezo1^fl/fl and Piezo1 cKO BMDM groups were elongated or oval-shaped with well-organized cristae. In contrast, mitochondria in the LPS + Piezo1^fl/fl group were round and swollen, accompanied by disrupted cristae structure. However, the LPS + Piezo1 cKO group exhibited a lower number of swollen mitochondria and relatively preserved cristae compared to the LPS + Piezo1^fl/fl group (Fig. S11F). Additionally, MitoSOX assays revealed that Piezo1 cKO significantly inhibited mitochondrial ROS production (Fig. S11G and H). These findings imply that Piezo1 knockout mitigates LPS-induced mitochondrial dysfunction.

3.6. Knockout of macrophage Piezo1 alleviates excessive mitochondrial fission

Furthermore, the role of Piezo1 in regulating mitochondrial fission in LPS-stimulated BMDMs was investigated. MitoTracker™ Red staining indicated that M1 macrophages exhibited mitochondrial fragmentation, with reduced form factor (FF) and aspect ratio (AR), suggesting significant mitochondrial fission following LPS stimulation (Fig. 4A–D). Western blot analysis demonstrated that Piezo1 knockout reduced LPS-induced phosphorylation of DRP1, decreased DRP1 accumulation in mitochondrial fractions, and increased DRP1 levels in cytosolic fractions (Fig. 4E–H).

Fig. 4.

Piezo1 regulates mitochondrial fission via DRP1 in LPS-stimulated BMDMs.(A–D) Mitotracker staining and quantification in BMDMs.(E–H) Western blot and quantification of P-DRP1 (Ser616), mito-DRP1, and cyto-DRP1 protein expression in BMDMs.(I–J) Immunofluorescence images and quantification of Mitotracker and DRP1 in BMDMs.(K–N) Western blot and quantification of mito-DRP1 and cyto-DRP1 in BMDMs treated with Mdivi-1 and Yoda1.Data are presented as mean ± SD. ∗P < 0.05; ∗∗P < 0.01; ∗∗∗P < 0.001; NS: not significant.

To further elucidate the role of Piezo1 in excessive mitochondrial fission, DRP1 protein levels were assessed in the presence or absence of Yoda1, a Piezo1 agonist. Western blot analysis revealed that Yoda1 promoted DRP1 translocation to mitochondria, which was attenuated by Mdivi-1, DRP1 inhibitor) (Fig. 4I–K).Consistently, immunofluorescence analysis showed that Yoda1 markedly increased MitoTracker-labeled DRP1 intensity, which was significantly reduced by Mdivi-1 treatment (Fig. S12).More importantly, Piezo1 knockdown significantly reduced MitoTracker-labeled DRP1 intensity and restored mitochondrial morphology, suggesting reduced DRP1 mitochondrial accumulation after LPS stimulation (Fig. 4L and M). In summary, these observations signal that Piezo1 knockdown effectively mitigates LPS-induced mitochondrial fission by inhibiting DRP1 mitochondrial translocation.

3.7. DRP1 regulates Piezo1-Mediated cGAS-STING-NLRP3 pathway in BMDMs

Mounting evidence suggests that DRP1 directly interacts with mPTP-associated proteins to promote excessive mPTP opening, eventually culminating in mtDNA leakage and triggering the cGAS-STING-NLRP3 axis inflammatory response [[26], [27], [28]]. However, the role of Piezo1 in this pathway remains unknown.

Western blot analysis indicated that LPS-treated Piezo1^fl/fl BMDMs exhibited significantly higher cGAS and STING protein levels compared to Piezo1 cKO BMDMs (Fig. 5A–C). dsDNA, including mtDNA released into the cytoplasm and oxidized mtDNA, is the primary activator of the cGAS-STING pathway. Quantitative PCR assessing mtDNA levels demonstrated significantly higher cytoplasmic mtDNA levels in Piezo1^fl/fl BMDMs compared to Piezo1 cKO BMDMs (Fig. 5D).Consistently, immunofluorescence staining of synovial tissues revealed increased 8-OHdG signals in OA mice, which were markedly reduced in Piezo1 cKO mice, indicating attenuated oxidative mtDNA damage in vivo (Fig. S13). In addition, immunofluorescence assays revealed lower NLRP3 expression levels in synovial macrophages of Piezo1 cKO DMM mice compared to Piezo1^fl/fl mice (Fig. 5E and F). Western blot analysis and q-PCR further validated that LPS-stimulated Piezo1 cKO BMDMs exhibited lower NLRP3 and cleaved caspase-1 expression levels compared to Piezo1^fl/fl BMDMs (Fig. 5G–I,Fig.S14).

Fig. 5.

DRP1 regulates the Piezo1-mediated cGAS-STING-NLRP3 pathway in BMDMs.(A–C) Western blot and quantitative analysis of cGAS and STING in BMDMs.(D) Determination of cytosolic mtDNA copy numbers via qPCR.(E–F) Immunofluorescence and quantification of F4/80 and NLRP3 in synovium.(G–I) Western blot and quantification of NLRP3, caspase1, and cleaved caspase1 in BMDMs.(J–U) Additional Western blot, flow cytometry, and quantitative analyses for Piezo1-mediated pathway regulation in BMDMs.Data are presented as mean ± SD. ∗P < 0.05; ∗∗P < 0.01; ∗∗∗P < 0.001; NS: not significant.

To examine the involvement of DRP1 in the Piezo1-mediated cGAS-STING-NLRP3 signaling pathway, Western blot analysis was performed, which revealed higher iNOS, NLRP3, cleaved caspase-1, TNF-α, IL-6, IL-1β, and CD86⁺ levels in LPS+Yoda1-treated Piezo1fl/fl BMDMs compared to the LPS -treated Piezo1fl/fl BMDMs. Moreover, DRP1 inhibition suppressed Piezo1-mediated inflammatory responses induced by LPS or LPS + Yoda1 in Piezo1fl/fl BMDMs (Fig. 5J–W). These findings suggest that DRP1 regulates the Piezo1-mediated cGAS-STING-NLRP3 pathway in BMDMs.

3.8. CaMKII mediates Piezo1-Induced DRP1 mitochondrial translocation

As a calcium ion channel, Piezo1 mediates extracellular Ca²⁺ influx, leading to intracellular Ca²⁺ overload and enhancing calcium-dependent signaling activation [29]. We hypothesize that Piezo1-mediated Ca²⁺ influx plays a paramount role in regulating DRP1 levels in mitochondria.

As displayed in Fig. S15A and C, Yoda1 activation of Piezo1 induced rapid calcium influx in WT BMDMs, a phenomenon not observed in Piezo1-deficient BMDMs. Notably, LPS-induced calcium influx was significantly lower in Piezo1 cKO BMDMs (Fig. S15B and D).

CaMKII, a key target protein in the calcium signaling pathway, regulates various cellular pathways and is activated by Piezo1 [30]. Piezo1 knockout attenuated LPS-induced CaMKII phosphorylation and mitochondrial accumulation, with a concurrent increase in cytosolic CaMKII expression levels (Fig. S15E–H). Immunoprecipitation and immunofluorescence assays revealed that LPS significantly promoted CaMKII-DRP1 colocalization in wild-type BMDMs, which was suppressed in Piezo1-deficient BMDMs (Fig. S15I–K). Furthermore, the CaMKII inhibitor KN-93 suppressed Yoda1-induced DRP1 mitochondrial translocation and reduced DRP1 phosphorylation(Fig. S15L–N,S12). Consistently, CaMKII knockdown by siRNA similarly attenuated DRP1 mitochondrial translocation in response to Yoda1 stimulation, supporting a CaMKII-dependent mechanism(Fig. S16).These results suggest that Piezo1 activation regulates CaMKII-DRP1 mitochondrial translocation via calcium influx.

3.9. Targeted knockdown of synovial macrophage Piezo1 with Man-LNP@Si-Piezo1 alleviates OA symptoms

Based on the role of Piezo1 in regulating macrophage polarization, mannose-modified liposomes (Man-LNP) were synthesized for targeted delivery of Si-Piezo1 to macrophages to inhibit Piezo1 expression. The schematic representation of Man-LNP@Si-RNA is shown in Fig. S17A, and the TEM structure of Man-LNP@Si-Piezo1 is depicted in Fig. S17B. The synthesized nanoparticles exhibited high gene silencing efficiency in Raw264.7 cells(Fig. S17C). Nanoparticle tracking analysis (NTA) revealed particle sizes ranging from 130 to 170 nm (Man-LNP@Si-NC: 155.75 ± 11.03 nm; Man-LNP@Si-Piezo1: 150.16 ± 11.12 nm, Fig. S17D–E).Immunofluorescence analysis demonstrated that over 90% of FAM-labeled siRNA signals co-localized with F4/80⁺synovial macrophages, with minimal uptake in chondrocytes, confirming efficient mannose-mediated macrophage targeting in vivo (Fig. S18A). Consistently, delivery of si-Piezo1 markedly reduced Piezo1 expression in synovial macrophages, validating effective in vivo gene silencing (Fig. S18B). Man-LNP exhibited sustained siRNA release under physiological conditions, with approximately 60% released within 12 h and reaching ∼85% by 48 h (Fig. S19). In addition, histological analysis of major organs, including the liver and kidney, revealed no apparent tissue damage or inflammatory changes compared with controls, indicating good biosafety (Fig. S20).

To evaluate the therapeutic effects, Man-LNP@Si-Piezo1 complexes were intra-articularly injected twice weekly into WT mice at the early stage of DMM-induced OA. Compared to the DMM and DMM + Man-LNP@Si-NC groups, injections of Man-LNP@Si-Piezo1 significantly alleviated OA symptoms, as evidenced by reduced osteophyte formation, decreased synovial inflammation, an increased Safranin O staining area, decreased ADAMTS4 expression, and enhanced Aggrecan expression (Fig. 6A–G, J-M). Notably, MCC950 treatment resulted in a comparable increase in SO staining, exhibiting a similar protective trend to that observed with Man-LNP@si-Piezo1 (Fig. S21).Furthermore, the number of F4/80+iNOS and F4/80+NLRP3 double-positive cells was significantly lower in the synovium of Man-LNP@Si-Piezo1-treated mice compared to controls (Fig. 6H–I, K-L,Fig. S5). These findings demonstrate that targeted knockdown of synovial macrophage Piezo1 using Man-LNP@Si-Piezo1 effectively alleviates OA symptoms.

Fig. 6.

Targeted knockdown of Piezo1 in synovial macrophages using Man-LNP@Si-Piezo1 alleviates OA symptoms.(A) 3D micro-CT images of the knee joints of different groups.(B–C) Quantification of osteophyte number and volume in ROIs.(D) H&E and Safranin O/Fast Green staining of joint cartilage.(E–F) Quantification of synovitis and OARSI scores.(G–M) Immunohistochemical and immunofluorescence analyses of Aggrecan, ADAMTS4, iNOS, and NLRP3 in cartilage and synovium, with quantification.Data are presented as mean ± SD. ∗P < 0.05; ∗∗P < 0.01; ∗∗∗P < 0.001; NS: not significant.

4. Discussion

The present study demonstrated that: (i) Piezo1 is predominantly localized in synovial macrophages in OA patients and murine OA models, indicating its potential role in OA pathogenesis; (ii) Conditional knockout of Piezo1 in macrophages inhibits M1 macrophage polarization, mitigates synovitis, and prevents cartilage degeneration during OA progression in the DMM model; (iii) In vitro inhibition of Piezo1 attenuates LPS-induced M1 macrophage polarization via the DRP1 signaling pathway and reduces chondrocyte senescence and cartilage degeneration; (iv) Piezo1-mediated Ca²⁺ influx enhances CaMKII phosphorylation, leading to DRP1 mitochondrial translocation and excessive mitochondrial fission; (v) Piezo1-induced DRP1 translocation drives mPTP overactivation, ΔΨm loss, and mitochondrial ROS (mtROS) accumulation; (vi) Piezo1 suppresses M1 macrophage polarization through the cGAS-STING-NLRP3 signaling pathway. To the best of our knowledge, this is the first study to explore the role of Piezo1 in regulating mitochondrial function and macrophage polarization in OA.

OA is a systemic joint disease with a multifactorial etiology, and synovial inflammation is a key pathological mechanism in its progression [31]. Macrophages, the predominant immune cell type in the synovium, are closely associated with synovial inflammation and OA severity [32]. Under physiological conditions, macrophages remain in a quiescent state; however, during OA, DAMPs activate macrophages, polarizing them into pro-inflammatory M1 macrophages. It is worthwhile emphasizing that previous studies have reported the accumulation of M1 macrophages in the synovial tissues of OA patients and animal models [33]. Of note, M1 macrophages promote the release of inflammatory cytokines, such as IL-6, IL-1β, and TNF-α, thereby exacerbating the inflammatory response. Simultaneously, they contribute to chondrocyte senescence and degeneration, thereby worsening OA progression [34]. In agreement with the findings of previous studies, our histological analyses revealed higher synovitis scores and infiltration of M1 macrophages in the synovium of OA patients and animal models compared to controls.

Previous studies have concluded that NLRP3-mediated inflammasomes are key regulators of OA progression [35]. In a study, the injection of specific NLRP3 inhibitors significantly alleviated inflammation and cartilage degeneration in an OA rat model [36]. NLRP3 inflammasomes play a central role in generating inflammatory factors from synovial macrophages (e.g., IL-1β and TNF-α), which collectively drive joint inflammation and accelerate cartilage degradation [37]. Additionally, pro-inflammatory factors released by M1 macrophages (e.g., TNF-α, IL-6, IL-1β) disrupt the metabolic balance of cartilage tissue, thereby accelerating its degradation through ADAMTS regulation [38]. Inhibition of NLRP3 significantly alleviates inflammation in cardiomyocytes, prevents M1 macrophage polarization, improves cardiac dysfunction, and maintains calcium homeostasis-related protein expression [39].Moreover, inhibition of TRPV4 mitigates OA by inhibiting M1 macrophage polarization through the ROS/NLRP3 signaling pathway [40].In the context of these findings, our results highlight Piezo1 as an upstream mechanosensitive regulator of the NLRP3 inflammasome. Piezo1 activation promotes calcium influx and mitochondrial stress, both of which are recognized triggers for NLRP3 activation. By conditionally deleting Piezo1 in macrophages, we observed a significant downregulation of the cGAS-STING-NLRP3 axis, reduced M1 macrophage polarization, and alleviated cartilage degeneration. These findings suggest that Piezo1 not only acts in previously established OA therapeutic targets (e.g., NLRP3) but also functions as a critical upstream node capable of modulating multiple inflammatory pathways simultaneously.

Existing evidence indicates that Piezo1 knockdown impairs the secretion of inflammatory factors such as IL-1β, TNF-α, and IL-6 by macrophages [41]. Furthermore, flow-induced shear stress activates NLRP3 inflammasomes in macrophages via Piezo1, whilst myeloid-specific Piezo1 deficiency modulates macrophage infiltration, limits M1 polarization, and attenuates inflammation in fibrotic livers [41,42]. Notwithstanding, the role of Piezo1 in OA-related M1 macrophage polarization remains unclear.

Therefore, a DMM model was constructed in transgenic mice with macrophage-specific Piezo1 knockout. The results demonstrated that Piezo1 cKO mice exhibited alleviated OA symptoms compared to controls, which was likely attributed to a reduction in the number of M1 macrophages within the synovium. Furthermore, Piezo1 cKO BMDMs promoted chondrocyte anabolism and minimized chondrocyte senescence in vitro. These findings demonstrate that Piezo1 Knockdown mitigates the LPS-induced and DMM-induced expression of M1 macrophage-related proteins, thereby preventing OA, which is consistent with our in vivo results. Therefore, inhibiting Piezo1-mediated NLRP3 inflammasome activation and M1 macrophage polarization may provide a promising therapeutic target for OA diagnosis and treatment.

Dynamic changes in mitochondrial fusion and fission play crucial roles in regulating macrophage function [43,44]. DRP1-dependent mitochondrial fission induces pro-inflammatory activation of macrophages, whereas Mdivi-1 inhibits M1 polarization in macrophages and microglia [45,46]. Noteworthily, DRP1-mediated mitochondrial fission leads to structural disruption and excessive mPTP opening and exacerbates mitochondrial and cellular dysfunction following hypoxia [22]. Inhibition of mitochondrial fission protects cells by preventing mPTP opening [46]. CaMKII phosphorylation at S616 on DRP1 mediates mitochondrial damage through mPTP overactivation [47]. Pharmacological inhibition of Piezo1 significantly alleviates matrix stiffness-induced mitochondrial fission and subsequent dysfunction [19]. However, the role of Piezo1 in macrophage mitochondrial dysfunction remains unclear.

Herein, Yoda1 treatment and LPS activation of Piezo1 enhanced DRP1 phosphorylation and its mitochondrial translocation. Additionally, MitoTracker and DRP1 were co-localized in LPS-treated macrophages. Piezo1 cKO BMDMs exhibited reduced mitochondrial fission, ΔΨm depolarization, MitoTracker-positive calcein intensity, and mtROS accumulation. Similarly, Mdivi-1 inhibited Yoda1-induced mitochondrial fission and subsequent dysfunction. These findings corroborate that Piezo1 is a critical regulator of macrophage mitochondrial fission and polarization. Furthermore, this study demonstrated that the Ca²⁺-CaMKII-DRP1 axis mediates mitochondrial fission induced by LPS or Yoda1 activation, suggesting that Piezo1 triggers mitochondrial dysfunction through CaMKII-DRP1 interactions.

To further elucidate the role of Piezo1 in macrophage polarization in OA, macrophage-related signaling pathways in Piezo1 cKO BMDMs were examined. The cytosolic escape of mtDNA from fragmented mitochondria triggers the activation of the STING signaling pathway, which senses changes in dsDNA levels and enhances type I interferon expression, as observed during dengue virus infection [48]. mtDNA is released from mitochondria via the mPTP, thereby activating the cGAS-STING pathway [49]. Herein, LPS-induced activation of the STING signaling pathway in BMDMs was associated with the release of cytosolic mtDNA, a process contingent upon DRP1-mediated mitochondrial fission. Downregulation of Piezo1 in BMDMs significantly reduced LPS-induced mtDNA release. These results suggest that, during Piezo1-DRP1-dependent mitochondrial fission, mtDNA is released via the mPTP and is essential for STING pathway activation.Finally our data indicate that Piezo1 regulates the cGAS-STING-NLRP3 axis in BMDMs, we acknowledge that activation of this pathway may also occur independently of Piezo1 under certain conditions.

SiRNA holds considerable promise for OA therapy owing to its high specificity in targeting and inhibiting complementary mRNA sequences [50,51]. Thus, a system was pioneered for delivering Piezo1-targeting siRNA to synovial macrophages for OA treatment. However, challenges such as enzymatic degradation, low transfection efficiency, nonspecific biodistribution, and uncontrolled release significantly limit its clinical applicability [[52], [53], [54]]. To address these challenges, a mannose-modified lipid nanoparticle system (Man-LNP) was developed for the targeted delivery of siPiezo1 to macrophages. Intra-articular injection of Man-LNP@Si-Piezo1 suppressed Piezo1 expression in macrophages, reduced M1 macrophage polarization, and significantly alleviated OA symptoms in mice.

This study has several limitations. First, although we focused on synovial macrophages, Piezo1 may also play important roles in other macrophage-derived lineages such as osteoclasts, which contribute to subchondral bone remodeling during OA progression. Future studies using lineage-specific models (e.g., CTSK-CreER mice) are warranted to clarify the role of Piezo1 in osteoclasts. Second, our findings were based primarily on the DMM model, which mimics mechanically induced OA. Due to resource constraints, we did not assess Piezo1 expression in aging-induced or spontaneous OA models, which would help validate the generalizability of our observations.Third, comprehensive in vivo biodistribution analyses of the nanoparticles were not performed, representing another limitation of this study.Finally, neonatal chondrocytes were used in our in vitro experiments. Although these cells provide a convenient and reproducible source, they may not fully recapitulate the phenotype and responses of adult chondrocytes in osteoarthritis. Addressing these aspects in future work will be important for developing targeted interventions aimed at both cartilage protection and subchondral bone preservation.

5. Conclusions

In summary, this study highlights the critical role of macrophage Piezo1 in the progression of OA. Myeloid-specific Piezo1 regulates macrophage infiltration, inflammation, and M1 polarization, thereby promoting chondrocyte senescence and degeneration in OA. These findings provide compelling evidence that Piezo1 is a key regulator of macrophage-mediated inflammatory responses. Targeted knockdown of Piezo1 in macrophages via Man-LNP@Si-Piezo1 significantly alleviated OA symptoms in mice. Beyond traditional pharmacological strategies and regenerative therapies, this immunotargeting approach not only promotes chondrocyte anabolism but also inhibits catabolism, presenting a novel intervention for OA.

Ethical statement

The experimental procedures for animal care and use were performed in accordance with the Guidelines for the Care and Use of Laboratory Animals provided by the U.S. National Institutes of Health and were approved by the Animal Research Committee of Hebei Medical University(IACUC-Hebmu-2025046). Human synovial tissues were collected from two groups: patients with osteoarthritis (OA group) and patients with tibial plateau fractures but without OA (control group). All procedures were conducted in accordance with the ethical standards of the institutional review board (approval number: 2024-K-351-01), and informed consent was obtained from all participants.Freshly isolated synovial tissue was then immediately rinsed in PBS to remove residual blood and debris. Tissue was fixed, dehydrated, and embedded in paraffin.

CRediT authorship contribution statement

Zijian Yan: Formal analysis, Investigation, Data curation, Methodology, Writing original draft, Writing review & editing.Dengying Wu: Formal analysis, Investigation, Data curation.Chengbin Huang: Investigation. Jianpeng Chen: Formal analysis, Writing review & editing.Haiyue Zhao: Formal analysis, Writing review & editing.Xuankuai Chen: Formal analysis.Yingze Zhang, Juan Wang: Supervision, Conceptualization, Formal analysis, Investigation, Writing original draft, Supervision, Funding acquisition.

Declaration of generative AI in scientific writing

No generative artificial intelligence (AI) or AI-assisted technologies were used in the preparation of this manuscript.

Funding

This study was partially supported by a research grant from the National Natural Science Foundation of China (Grant No. U23A6008; No. 91949203),Science and Technology Projects Xizang Autonomous Region, China (XZ202301ZY0046G)and USTC Research Funds of the Double First-Class Initiative (Grant No. YD9110002090)

Declaration of interests

The authors declare no competing interests.

Appendix A. Supplementary data

The following is/are the supplementary data to this article.

Data availability

Data will be made available on request.