BMSC-Derived Exosomes Ameliorate Osteoarthritis by Inhibiting Pyroptosis of Cartilage via Delivering miR-326 Targeting HDAC3 and STAT1//NF-κB p65 to Chondrocytes

Background In the past decade, mesenchymal stem cells (MSCs) have been widely used for the treatment of osteoarthritis (OA), and noncoding RNAs in exosomes may play a major role. Aim The present study is aimed at exploring the effect and mechanism of miR-326 in exosomes secreted by bone marrow mesenchymal stem cells (BMSCs) on pyroptosis of cartilage and OA improvement. Methods Exosomes from BMSCs (BMSC-Exos) were isolated and identified to incubate with OA chondrocytes. Proliferation, migration, specific gene and miR-326 expression, and pyroptosis of chondrocytes were detected. BMSCs or chondrocytes were transfected with miR-326 mimics or inhibitors to investigate the effect of miR-326 in BMSC-Exos on pyroptosis of chondrocytes and the potential mechanism. Finally, a rat OA model was established to verify the effect and mechanism of miR-326 in BMSC-Exos on cartilage of pyroptosis. Results Incubation with BMSC-Exos could significantly improve the survival rate, migration ability, and chondrocyte-specific genes (COL2A1, SOX9, Agg, and Prg4) and miR-326 expression of OA chondrocytes and significantly inhibit pyroptosis of chondrocytes by downregulation of the levels of inflammatory cytokines, Caspase-1 activity, and pyroptosis-related proteins such as GSDMD, NLRP3, ASC, IL-1β, and IL-18 (P < 0.01). PKH26 labeling confirmed the uptake of BMSC-Exos by chondrocytes. Incubation with exosomes extracted from BMSCs overexpressing miR-326 can significantly repress the pyroptosis of chondrocytes, while knockdown of miR-326 had the opposite effect (P < 0.01). The same result was also demonstrated by direct interference with the expression level of miR-326 in chondrocytes (P < 0.01). In addition, we found that the overexpression of miR-326 significantly inhibited the expression of HDAC3 and NF-κB p65 and significantly promoted the expression of STAT1, acetylated STAT1, and acetylated NF-κB p65 in chondrocytes (P < 0.01). The targeted relationship between miR-326 and HDAC3 was verified by dual-luciferase reporter assay. Animal experiments confirmed the mechanism by which miR-326 delivered by BMSC-Exos inhibits pyroptosis of cartilage by targeting HDAC3 and STAT1/NF-κB p65 signaling pathway. Conclusion BMSC-Exos can deliver miR-326 to chondrocytes and cartilage and improve OA by targeting HDAC3 and STAT1//NF-κB p65 to inhibit pyroptosis of chondrocytes and cartilage. Our findings provide a new mechanism for BMSC-Exos to treat OA.


Introduction
Osteoarthritis (OA) is the most common joint disease and the leading cause of degeneration, destruction, and osteogenesis of articular cartilage [1,2]. Increased prevalence of OA brings an enormous societal burden along with the aging population [3]. The pathogenesis of OA is complex and has a heritable tendency [4]. Due to the interaction of mechani-cal and biological factors, the imbalance of articular chondrocytes and extracellular matrix (ECM) are the main causes of OA [5]. At present, clinical nondrug therapy, drug therapy, and surgical treatment cannot fundamentally delay the progressive degeneration of articular cartilage during OA. In view of this, a new mesenchymal stem cell-(MSC-) based OA therapy has attracted increasing attention. MSCs can repair cartilage tissue and inhibit the secretion of inflammatory cytokines by chondrocytes and can be directed to differentiate into chondrocytes in vivo. A series of studies confirmed that MSCs can effectively treat OA from different aspects [6][7][8]. Clinical application has preliminarily proven that MSCs may be the best method to treat traumatic bone and cartilage defects [9]. However, the risk of tumor formation, ethical issues, and transplant rejection remain obstacles to the further clinical application of stem cells [10].
Exosomes are tiny vesicles released into the ECM upon the fusion of multivesicular bodies with the cytoplasmic side of the plasma membrane, which are shaped like discs and have a diameter of 30-150 nm [11,12]. Exosomes contain a variety of proteins, lipids, and nucleic acids with important functions that can be transferred to other cells to affect the function of the recipient cells. A growing body of data suggests that MSCs can inhibit the development of disease by secreting exosomes carrying proteins, miRNAs, long noncoding RNAs, or other small molecules [13]. Studies have found that exosomes derived from MSCs can promote chondrocyte proliferation, inhibit apoptosis, enhance chondrogenesis, inhibit cartilage degradation, promote cartilage repair, and relieve osteoarthritis by balancing the synthesis and degradation of chondrocyte ECM [14][15][16][17]. MSCderived exosomes are also widely recognized for their role in protecting cartilage degeneration and promoting cartilage repair and regeneration and the treatment of osteoarthritis [18][19][20][21]. A recent study showed that bone marrow mesenchymal stem cell-(BMSC-) derived exosomes protected cartilage damage and alleviated knee pain in osteoarthritis model rats [22]. In addition, BMSC-derived exosomes from congenital polydactyly tissue alleviate osteoarthritis by promoting chondrocyte proliferation [23]. However, the specific mechanism of BMSC-derived exosomes against osteoarthritis remains unclear.
Pyroptosis, including caspase-dependent programmed cell death and proinflammatory changes, is a combined process of apoptosis and necrosis [24,25]. Unlike other forms of programmed cell death, pyroptosis is closely associated with inflammation. Pyrolysis is associated with risk factors for OA and is involved in cartilage degeneration, synovial changes, and OA-induced pain and plays an important role in OA [26]. There are also evidences that inhibition of pyroptosis of chondrocytes can ameliorate cartilage degeneration and arthritis [27,28]. MSCs have been shown to interfere with pyroptosis during cerebral ischemia/reperfusion [29] and acute liver failure [30]. Moreover, it has been reported that exosomes of MSCs can protect cardiomyocytes from hypoxia/reoxygenation-induced pyroptosis [31]. However, in OA-related studies, there has been no report on the effect of MSC-or BMSC-derived exosomes on pyroptosis of chondrocytes.
Exosomes secreted by a variety of different cells have therapeutic or mitigating effects through miR-326 in inflammatory bowel disease [32], relapsing-remitting multiple sclerosis [33], and hepatocellular carcinoma [34] and are associated with inflammation and NF-κB pathway. MiR-326 has also been proved to be involved in the regulation of pyroptosis in Parkinson's disease [35]. In this study, we will explore the effect of miR-326 in BMSC-derived exosomes (BMSC-Exos) on pyroptosis of chondrocytes and the potential mechanism and discuss the possibility of BMSC-Exos in the treatment of OA.

Materials and Methods
2.1. Cell Culture and Treatment. Bone marrow was taken from the rat femur and washed with α-minimum essential medium (α-MEM, Gibco, USA) to isolate bone marrow mesenchymal stem cells (BMSCs). BMSCs cultured with α-MEM contained 10% fetal bovine serum, 100 U/mL penicillin (Solarbio, China), and 10 mg/mL streptomycin (Solarbio) at 37°C with 5% CO 2 . Cartilage tissue was collected from the knee joints of rats, washed with phosphate-buffered saline (PBS), cut into pieces, and digested with collagenase II (col II) to obtain chondrocytes. Isolated mouse chondrocytes and BMSCs were cultured in Dulbecco's Modified Eagle Medium (DMEM) supplemented with 10% fetal bovine serum (Gibco, USA) and 100 U/mL penicillin at 37°C with 5% CO 2 . Subculture was carried out when the cell confluence rate was close to 80%. Chondrocytes was stimulated by 10 ng/mL interleukin-1β (IL-1β) for 24 h to construct the osteoarthritis model as previously described [22].

Exosomes Collection and Detection.
BMSCs were cultured in exosome-free media in the initial. Media from cells and PBS-washed cells was collected and together centrifuged for 5 minutes, and then, the supernatant was centrifuged for 20 minutes. Exosomes were isolated with the ExoQuick-TC™ system (System Bioscience, USA) according to the protocol. Exosomal protein was measured via a BCA assay, and 10 μg exosomes were placed on a copper grid. Exosomes were skillfully separated to form a thin layer before a thin layer of 2% uranyl acetate is added. Grids were allowed to dry overnight, and transmission electron microscopy (TEM) was performed the next day. The size distribution of the exosomes was measured using Nanosizer™ technology (Malvern Instruments, UK), and was analysed using Zetasizer software (Malvern). After OA modeling, chondrocytes were treated with 10 μg/mL BMSC-Exos for 12 h at 37°C [22].

Exosome Labeling.
Exosome suspension (100 μL) was mixed with 1 mL of dilution C-diluted PKH67 (Sigma, USA) for incubation at room temperature for 4 min. The staining was terminated by adding 1 mL of 0.5% Bovine Serum Albumin (BSA, Gibco), and the exosomes were reextracted. The observation under a fluorescence microscope showed that the exosomes were stained by PKH67 (red). Chondrocytes were stained with DAPI (Sigma) in blue. PKH67-labeled exosomes were incubated with chondrocytes for 12 h before locating the exosomes.  2.8. Wound Healing Assay. Before the experiment, the medium was changed to serum-free medium. Pipette tip was used to create cell scratches, and the cell surface was washed once with serum-free medium. The cells were observed under a microscope (Olympus, Japan) and photographed, and the positions of cells in the photos were recorded, denoted as 0 h. Instead, serum-free culture was conducted in an incubator at 37°C and 5% CO 2 for 24 h. Take pictures and calculate the wound closure of cells using Image J software.
2.9. Immunofluorescence Assay. The articular cartilage was embedded in paraffin and sliced into 5 μm slices. Cells or tissues were fixed with methanol and permeabilized with 0.1% Triton X-100 in PBS for 20 min. The tissues were incubated with GSDMD (ab209845, Abcam) or cleaved Caspase-1 (cl-Caspase-1, af4005, Affinity Biosciences, USA) and Collagen II (Col II, MA5-12789, Invitrogen, USA) primary antibody (Cell Signal Technology), and cells were incubated with GSDMD or cleaved Caspase-1 primary antibody overnight at 4°C and then incubated in fluorochrome-conjugated or normal secondary antibodies for 2 h at room temperature. DAPI was used to counterstain the cell nuclei. Images were captured with a laser confocal microscope (Olympus, Japan) in five different fields for each sample.
2.10. Caspase-1 Activity Assay. Caspase-1 activity assay kit was purchased from Beyotime Biotechnology Company (China). The cells or tissues were lysed with lysate, and the lysate was collected and centrifuged for 10 min. After the supernatant was taken to determine the protein concentration, the same amount of samples were placed in a 96-well microtitration plate, treated with 20 ng Ac-DEVD-pNA at 37°C and incubated overnight. The absorbance at 405 nm was measured with a microplate analyzer to obtain the content of pNA generated, and the activity of Caspase-1 contained in the unit weight protein was calculated.
2.11. Enzyme-Linked Immunosorbent Assay (ELISA). Concentrations of IL-1β, IL-6, IL-18, and TNF-α in cell culture medium or rat articular cartilage were measured with rat ELISA Kits purchased from Sigma-Aldrich (USA) according to supplier's protocols. Absorbance was measured at 450 nm wavelength.
2.12. NF-κB p65 DNA-Binding Assay. NF-κB p65 DNAbinding activity in chondrocytes or cartilage nuclear extracts was determined with a transcription factor binding assay colorimetric ELISA kit (Cayman Chemical, USA) according to manufacturer's protocol as previously described [36]. Briefly, NF-κB p65 present in the nuclear extract binds specifically to the NF-κB response element (a double-stranded DNA sequence containing the NF-κB response elements immobilized to the wells of a 96-well plate), and the DNAprotein complex was detected by addition of a specific 3 Mediators of Inflammation primary antibody directed against NF-κB p65. A secondary antibody conjugated to HRP was added, and the absorbance was read at 450 nm.
2.13. Dual-Luciferase Reporter Assay. Wide sequence (WT-HDAC3) and mutant sequences (MUT-HDAC3) were designed and synthesized by GenePharma. Sequences were inserted into a luciferase reporter vector (pGL3-Basic). Vectors were cotransfected with miR-326 mimic or miRNA mimic NC into HEK293T cells. Cells were lysed with 100 μL of lysis buffer in a shaking bed at room temperature for 20 min. Cell suspension was incubated with luciferase solution (Promega, USA) before the intensity of firefly luciferase was determined. Stop&Glo reagent I from Promega was fully mixed with cell suspension, and then, the intensity of Renilla luciferase was measured. The activity of Renilla luciferase was considered as an internal control. The relative luciferase activity shall be the ratio of firefly luciferase intensity and Renilla luciferase intensity. Triplicate wells were set for each group.
2.14. Construction of the Animal OA Model and Grouping. All procedures were performed with the approval of the Animal Ethics Committee of Anhui Medical University, in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals. Twenty-four adult male Sprague Dawley rats (weighing 200~220 g) were anesthetized with 2.5% isoflurane. Eighteen rats were performed the unilateral intra-articular injection with 8% sodium iodoacetate (Sangon Biotech, China) to produce the OA change of the knee, and the rats in the sham group were injection with normal saline. One week later, intra-articular injection of BMSC-Exos (40 μg/100 μL) was performed in the OA rats once a week [22] (n = 6, OA+BMSC-Exo group and OA +BMSC-miR-326-Exo group). At 6 weeks after surgery, the knee joint specimens of rats were collected for histologic analysis. Gross scores of cartilage repair were made according to the International Society for Chondroprosthesis (ICRS) criteria.
2.15. Histologic Analysis. At 6 weeks after treatment, the rats were sacrificed, and articular cartilage samples were collected. After fixation with paraformaldehyde and decalcified in 10% EDTA, the tissues were embedded in paraffin and sectioned into a 5 μm thick section. Sections were made perpendicular to the defect area. The selected sections were deparaffinized in xylene, rehydrated through a graded series of ethanol washes, and followed by hematoxylin and eosin (H&E) and Safranin O/Fast Green staining (Solarbio). Then, the sections were stained with freshly prepared Weigert solution for 5 min and differentiated with acid differentiation solution for 15 s. The sections were immersed in Fast Green and Safranin O staining solution for 5 min and 2 min, respectively, and washed with distilled water for 1 min. The sections were washed with weak acid solution for 1 min and washed with distilled water for 1 min. The sections were dehydrated and transparent and finally observed under a light microscope.
2.16. Statistical Analysis. All data are presented as the means ± standard deviation (SD). The data statistical significance analyses and statistical mapping were performed using Prism GraphPad 8.0 software. Data were analyzed using one-way ANOVA and Student's t-test. P < 0:05 was considered statistically significant.

BMSC-Exos Inhibit Pyroptosis and Maintain
Chondrocyte Homeostasis of Rat OA Chondrocytes. To investigate the effect of BMSC-derived exosomes (BMSCs-Exos) on chondrocytes, exosomes derived from BMSCs were isolated and identified by nanoparticle tracking analysis (NTA), TEM, and Western blot. The diameter distribution of exosomes was analyzed by NTA, and the particles ranged around 123 nm in diameter (Figure 1(a)). Isolated exosomes were examined for size and structure via transmission electron microscopy ( Figure 1(b)). Subsequently, the expressions of CD9, CD63, and CD81 exosome markers in the isolated exosomes were detected by Western blot (Figure 1(c)). Based on these, it was concluded that exosomes of BMSCs were successfully isolated and purified.
After that, isolated BMSC-Exos were used to incubate OA chondrocytes to explore the effect of BMSC-Exos on pyroptosis of chondrocytes. Compared with normal chondrocytes, the survival rate and migration ability of OA chondrocytes decreased significantly, and BMSCs-Exos incubation can reverse the decrease of proliferation and migration ability of OA chondrocytes (P < 0:01, Figures 2(a) and 2(b)). The mRNA expression levels of COL2A1, SOX9, Agg, and Prg4 showed that the expression levels of chondrogenic specific genes in OA chondrocytes were significantly decreased compared with normal chondrocytes, and BMSCs-Exos incubation could significantly increase the level of chondrogenic specific genes in OA chondrocytes (P < 0:01, Figure 2(c)).
To investigate the effect of BMSCs-Exos on pyroptosis of OA chondrocytes, we detected the expression levels of the key genes of pyroptosis by immunofluorescence, ELISA, and Western blot. The levels of IL-1β, IL-18, IL-6, and TNF-α in cell culture medium were detected by ELISA, and all proinflammatory cytokines in OA chondrocyte culture medium were increased and significantly decreased after BMSCs-Exos were incubated (P < 0:01, Figure 2(d)). Results of immunofluorescence showed that OA significantly increased the levels of GSDMD and cleaved Caspase-1 in chondrocytes, and BMSCs-Exos incubation significantly decreased the levels of them in chondrocytes (P < 0:01, Figure 2(e)). The same trend was also shown in Caspase-1 activity assay, that is, OA increased the Caspase-1 activity of chondrocytes, and BMSC-Exos incubation reversed the OA-induced Caspase-1 activation (P < 0:01 , Figure 2(f)). Western blot detected the expression levels of pyroptosis-related proteins in each group and found that the levels of NLRP3, ASC, GSDMD, Caspase-1, IL-1β, and IL-18 were significantly decreased after BMSC-Exos were incubated (P < 0:01, Figure 2(g)). Mediators of Inflammation and incubated with chondrocytes for 12 h. The nuclei were located by DAPI staining, and fluorescence microscopy showed that exosomes could be ingested by chondrocytes ( Figure 3(a)). RT-PCR showed that miR-326 expression level in OA chondrocytes was significantly decreased compared with normal chondrocytes and increased significantly after BMSC-Exos incubation (P < 0:01, Figure 3(b)). These results suggest that BMSC-Exos deliver miR-326 into chondrocytes.

BMSC-Exos Inhibited
To verify the regulatory effect of miR-326 in BMSC-Exos on chondrocyte pyroptosis, miR-326 mimics and inhibitors were constructed and transfected into BMSCs. Exosomes from transfected BMSCs were extracted and used to treat OA chondrocytes. We found that the overexpression of miR-326 exosomes enhanced the proliferation and migration of OA chondrocytes, while the inhibition of miR-326 exosomes weakened (P < 0:01, Figures 4(a) and 4(b)). RT-PCR was performed to detect the level of miR-326 in each group of chondrocytes, and it was found that treatment with BMSCs-Exos that promoted or inhibited miR-326 could correlative promote or inhibit the level of miR-326 in chondrocytes (P < 0:01, Figure 4(c)), confirming that BMSCs-Exos could deliver miR-326 into chondrocytes. In addition, the mRNA expression level of chondrocyte-specific genes was significantly increased by the overexpression of miR-326 in BMCS-Exos, while the expression levels of chondrocytespecific genes (COL2A1, SOX9, Agg, and Prg4) were signif-icantly decreased by the inhibition of miR-326 (P < 0:01, Figure 4(c)). The detection of inflammatory cytokine levels in the cell culture medium of each group showed that incubation of BMSCs-Exos with overexpression of miR-326 significantly reduced the level of proinflammatory cytokines (IL-1β, IL-18, IL-6, and TNF-α) in the extracellular fluid of chondrocytes, while silence of miR-326 significantly increased the level of proinflammatory cytokines (P < 0:01, Figure 4(d)). The levels of GSDMD and cleaved Caspase-1 were detected by immunofluorescence, and it was found that overexpression of miR-326 significantly reduced the levels of GSDMD and active Caspase-1, while silencing miR-326 was the opposite (P < 0:01, Figure 4(g)). Detection of Caspase-1 activity also showed the same results, and Caspase-1 activity in chondrocytes was negatively correlated with miR-326 level in BMSCs, from which exosomes were extracted (P < 0:01, Figure 4(f)). The detection of the expression level of pyroptosis-related proteins in cells revealed that the overexpression of miR-326 in BMSCs-Exos significantly reduced the expression level of pyroptosis-related proteins in OA chondrocytes, whereas silencing miR-326 had the opposite effect (P < 0:01, Figure 4(i)).
The above results suggest that BMSC-Exos affect chondrocyte homeostasis and pyroptosis through miR-326. We attempted to explore downstream factors that might be targeted by miR-326, and we detected the expression levels of HDAC3 and STAT1/NF-κB p65 in each group. BMSC-     suggesting that miR-326 inhibits chondrocyte pyroptosis by inhibiting HDAC3 expression. The dual luciferase gene report confirmed the targeted interaction between miR-326 and HDAC3 (P < 0:01, Figure 5(d)), which supported the above conjecture.
We also treated chondrocytes with miR-326 inhibitors and inhibitors of HDAC3 or STAT1 signaling pathways, confirming the targeted regulation of miR-326 on HDAC3 and STAT1/NF-κB p65 from another perspective. As shown in Figure 6, compared with the control, transfection with miR-326 inhibitor significantly increased the expression levels of pyroptosis-related proteins such as GSDMD, NLRP3, and ASC, as well as the activity and expression level of Caspase-1, and the levels of inflammatory factors IL-1β, IL-18, IL-6, and TNF-α (P < 0:01). Compared with the transfection of miR-326 inhibitor, the addition of HDAC3 or STAT1 inhibitor effectively alleviated the pyroptosis induced by miR-326 inhibitor, that is, decreased the protein expression levels of GSDMD, NLRP3, and ASC; Caspase-1 activity and expression; and levels of inflammatory factors IL-1β, IL-18, IL-6, and TNF-α (P < 0:01). These results confirmed that miR-326 regulates pyroptosis by targeting HDAC3 and STAT1 in chondrocytes.

21
Mediators of Inflammation normal BMSCs and BMSCs overexpressing miR-326, respectively. After 6 weeks of modeling, the cartilage sections of the knee joint of rats were taken for ICRS score, HE staining, and Safranin-O/fast green staining (Figures 7(a)-7(c)). The results showed that the ICRS score of articular cartilage of OA rats was significantly reduced (P < 0:01), and the surface of cartilage was rough. Absence of layers, reduced cell numbers, unclear and interrupted tide lines, and increased clusters of cells were observed in the OA articular cartilage. The staining with Safranin-O/Fast Green showed uneven, light, or dis-stained cartilage matrix in the OA articular cartilage. BMSC-Exos-treated cartilage showed significant ICRS score increase, cartilage layer thickness, increased cell number and orderly arrangement, relatively clear tide line, and deepened Safranin-O staining. It indicates the existence of abundant cartilage matrix and suggests cartilage tissue regeneration. The BMSC-Exos with overexpression of miR-326-treated OA articular cartilage showed better tissue regeneration.

Discussion
OA is the most frequent form of arthritis, characterized by progressive degeneration in articular cartilage and chronic joint pain [37]. Owing to the limited understanding of the pathogenesis of OA, an urgent improvement of prevention and therapy is required in clinical practice. As pluripotent stem cells, BMSCs have played important roles in repairing and mitigating tissue damage in a variety of tissue damage models [38][39][40].
In recent years, increasing evidence has suggested that stem cells and precursor cells may play important roles in promoting tissue regeneration by activating surrounding cells through paracrine mechanisms [41,42]. Most cells produce exosomes and play crucial roles in intercellular communication [43]. Moreover, exosomes have biological functions similar to those of the cells from which they were derived, and direct use of these nanoparticles has no obvious adverse effects, such as immune rejection or tumorigenicity [44]. In addition, exosomes can also promote the proliferation of chondrocytes in vitro [45]. Therefore, this study selected BMSC-Exos as the object to explore its regulatory effect on OA chondrocytes and the potential mechanism.
Our study found that incubation with BMSC-Exos can reverse the reduction of chondrocyte proliferation and migration caused by OA, as well as the decrease of chondrocyte-specific gene expression levels. The promoting effect of BMSC-Exos on the proliferation and migration of chondrocytes has also been supported in other studies [22,23]. SOX9 is an important early transcription factor for chondrogenesis and chondrocyte differentiation, and it is a sensitive factor in articular cartilage tissue engineering, which plays an important role in chondrogenesis [46,47]. Collagen type II (COL2A1), as a downstream target protein of Sox9, is an important component of the cartilage matrix and is considered the basis of structural strength to outside pressure [48]. Aggrecanases (Agg) belonging to the "A Disintegrin And Metalloproteinase with ThromboSpondin motifs" (ADAMTS) family of proteinases play a significant role in aggrecan depletion in osteoarthritic cartilage [49]. Aggrecanase-mediated aggrecan degradation is a significant event in early-stage OA. Proteoglycan 4 (Prg4), also known as Lubricin, is a surface-active mucus glycoprotein secreted by synovial joints, which acts as a lubricant in joints [50]. Maintenance of Prg4 expression prevents OA progression, which indicates that Prg4 plays an important role in maintaining the integrity of articular cartilage [51]. Our results showed that BMSC-Exos incubation significantly reversed the reduction of COL2A1, SOX9, Agg, and Prg4 expression levels induced by OA, which suggested that BMSC-Exos can protect chondrocyte homeostasis from osteoarthritis by promoting chondrogenesis and chondrocyte differentiation, cartilage matrix construction, cartilage degeneration, and synovial integrity. In vivo animal experiments have confirmed the role of BMSC-Exos in protecting cartilage integrity, maintaining the number of chondrocytes and cartilage matrix composition, and promoting cartilage tissue regeneration. Our results demonstrate the therapeutic effect of BMSC-Exos on OA, but its clinical application needs to be further studied.
Pyroptosis, a type of regulated cell death, occurs when pattern recognition receptors (PRRs) induce the activation of cysteine-aspartic protease 1 (caspase-1) or caspase-11, which can trigger the release of the pyrogenic cytokines interleukin-1β (IL-1β) and IL-18 [26]. Inflammasome is a polymeric protein complex and plays an important role in the pathogenesis of inflammation as a platform for caspase activation. Caspases are cysteine proteases that initiate or execute cellular programs and can induce inflammation or cell death, depending on its function [52]. Caspase-1 is the most fully characterized and is closely regulated by 22 Mediators of Inflammation inflammasomes to process cytokines [53]. Activated caspase-1 is required for proteolysis and the release of cytokines IL-1β and IL-18 [54]. When damage is caused by endogenous stress or microbial infection, caspases release hazard-associated molecular patterns (DAMPs) or pathogen-associated molecular patterns (PAMPs), which trigger pattern recognition receptors (PRRs) [55]. Nucleotide-binding domain and leucine-rich repeatcontaining (NLR) protein are one of the intracellular PRRs [56]. Activated caspase-1 triggers a typical inflammasome pathway leading to cell pyroptosis. Assembly of the NLRP3 inflammasome in response to PAMPs or DAMPs can lead to caspase-1-dependent release of proinflammatory cytokines and gasdermin D-(GSDMD-) mediated pyroptotic cell death. NLRP3 inflammasome activation requires upregulation of inflammasome components including NLRP3, Caspase-1, and pro-IL-1β by increasing the transcriptional activity of NF-κB [57,58]. Next, NLRP3 is activated by multiple upstream stimuli induced by a large number of PAMPs or DAMPs and then forms a complex with ASC and caspase-1. The complex cleaves the GSDMD and forms the GSDMD-N pores on the membrane to induce pyrolysis. We confirmed the inhibitory effect of BMSC-Exos on pyroptosis in OA by detecting changes in the expression levels of pyroptosisrelated proteins including GSDMD, Caspase-1, NLRP3, ASC, IL-1β, and IL-18 in chondrocytes and cartilage, which is similar to the action pathway of other drugs [27,28,59].
In terms of mechanism, we found that BMSC-Exos treatment increased the expression level of miR-326 in chondrocytes, and the level of miR-326 directly affected the pyroptosis of chondrocytes. In this study, we used PHK26 to label exosomes and changed the level of miR-326 in exosomes by RNA interference techniques, confirming the role of miR-326 in exosomes entering chondrocytes. However, whether exosomes trigger endogenous secretion of miR-326 in chondrocytes remains to be further explored. MiR-326 has been proved to be involved in the regulation of pyroptosis in Parkinson's disease [35]. Histone deacetylase-3 (HDAC3) is ubiquitously expressed and conserved in a wide range of species [60]. HDAC3 can regulate various signaling pathways, including NF-κB activity, and affect cell apoptosis [61]. It has been reported that miR-326 and HDAC3 can form a feedback loop to regulate the invasion and tumorigenic and angiogenic response to anticancer drugs [62]. Our experiment confirmed the targeted interaction between miR-326 and HDAC3, and miR-326 promoted the pyroptosis of chondrocytes by inhibiting HDAC3 and activating STAT1/NF-κB p65 pathway. IL-4-STAT6 has polarization-specific epigenomic characteristics, and HDAC3-dependent macrophage response to inflammatory stimuli is reduced; inflammatory corpus activation and pyroptosis are reduced [63]. The role of HDAC3dependent STAT1 and NF-κB p65 pathways in the regulation of inflammatory response has also been demonstrated [64]. Valproic acid treatment inhibited HDAC3 expression and activity, enhanced STAT1, and acetylated NF-κB p65 after spinal cord injury (SCI) [64]. The acetylation status of NF-κB p65 and the NF-κB p65 and STAT1 complex inhibit the transcriptional activity of NF-κB p65 and atten-uate the microglia-mediated central inflammatory response after SCI. Pyroptosis involves the release of intracellular proinflammatory factors (including IL-18 and IL-1β), and pyroptosis-regulated cell death plays a key role in the pathogenesis of various neurological diseases including SCI [65]. Transcription factor activity of NF-κB is a key factor in activating NLRP3 inflammasome and pyroptosis. A melatonin and adipose study determined that nuclear factor κB (NF-κB) signaling is essential in GSDMD-mediated pyrophosis of adipose cells and that melatonin alleviates inflammatory body-induced adipose tissue spontaneous combustion in mice by blocking NF-κB/GSDMD signaling [66]. Besides, NEK7 interacts with NLRP3 to modulate NLRP3 inflammasome activation, therefore, modulating the pyroptosis in mice [67]. Polyphyllin VI induced caspase-1mediated pyroptosis via the induction of the ROS/NF-κB/NLRP3/GSDMD signal axis in non-small-cell lung cancer [68]. Retinoic acid-inducible gene I (RIG-I) agonist therapy activates proinflammatory transcription factors STAT1 and NF-κB and triggers external apoptotic pathways and pyroptosis, a highly immunogenic form of cell death in breast cancer cells [69]. Studies on murine Norovirus (MNV) suggest that protective STAT1 signaling controls viral replication and pathogenesis and that NLRP3 inflammasome activation and subsequent GSDMD-driven pyrodeath are contributors to MNV-induced immunopathology in susceptible STAT1-deficient mice [70]. A large number of studies all above have confirmed the key role of HDAC3 and STAT1/NF-κB p65 in pyroptosis, and our results provide evidence that miR-326 plays a direct role in chondrocytes, and HDAC3 and STAT1/NF-κB p65 affect pyroptosis. On the whole, miR-326 targeting HDAC3 activates the STAT1/NF-κB p65 signaling pathway, inhibits the occurrence of pyroptosis, and protects chondrocytes from osteoarthritis.

Conclusion
BMSC-Exos can deliver miR-326 to chondrocytes and cartilage and ameliorate OA by targeting HDAC3 and STAT1//NF-κB p65 to inhibit pyroptosis of chondrocytes and cartilage. Our findings provide a new mechanism instruction for BMSC-Exos in OA treatment.

Limitations
There are still some limitations to this study. First, due to the limited conditions of experimental equipment, we failed to directly observe and compare the morphological changes of chondrocyte pyroptosis and confirm the role of BMSC-Exos on pyroptosis. In addition, the therapeutic effect of BMSC-Exos remains to be validated in a clinical trial. All these limitations should be addressed in the future study.

Data Availability
The original contributions presented in the study are included in the paper and supplementary material; further inquiries can be directed to the corresponding author.

Ethical Approval
All animal procedures in this study were approved by the Animal Ethics Committee of Anhui Medical University and carried out in accordance with the guidelines of the National Institutes of Health Guide for the Care and Use of Laboratory Animals.

Consent
Consent is not necessary.

Conflicts of Interest
The authors declare that they have no conflict of interest.