miR‐34a in extracellular vesicles from bone marrow mesenchymal stem cells reduces rheumatoid arthritis inflammation via the cyclin I/ATM/ATR/p53 axis

Abstract Extracellular vesicles (Evs) participate in the development of rheumatoid arthritis (RA), but the mechanisms remain unclear. This study aimed to determine the mechanism by which microRNA‐34a (miR‐34a) contained in bone marrow mesenchymal stem cell (BM‐MSC)‐derived Evs functions in RA fibroblast‐like synoviocytes (RA‐FLSs). BM‐MSC‐derived Evs and an Evs inhibitor were extracted. A rat model of RA was established. miR‐34a gain‐ and loss‐of‐function experiments were performed, and the inflammation in rat synovial fluid and tissues was detected. The role of miR‐34a in RA‐FLSs was also measured in vitro. The target gene of miR‐34a was predicted using the online software TargetScan and identified using a dual‐luciferase reporter gene assay, and the activation of the ATM/ATR/p53 signalling pathway was assessed. BM‐MSC‐derived Evs mainly elevated miR‐34a expression, which reduced RA inflammation in vivo and inhibited RA‐FLS proliferation and resistance to apoptosis in vitro, while inhibited miR‐34a expression enhanced RA development. In addition, miR‐34a could target cyclin I to activate the ATM/ATR/p53 signalling pathway, thus inhibiting abnormal RA‐FLS growth and RA inflammation. Our study showed that miR‐34a contained in BM‐MSC‐derived Evs could reduce RA inflammation by inhibiting the cyclin I/ATM/ATR/p53 signalling pathway.


| INTRODUC TI ON
Rheumatoid arthritis (RA) is one of the most commonly diagnosed systemic autoimmune diseases worldwide, and it can affect several organ systems. 1,2 Persistent inflammation often contributes to erosive joint injury and dysfunction in most RA patients. 3 In addition to the inconvenience caused by the disease itself, RA patients also have an increased risk of comorbidities, including cardiovascular diseases, osteoporosis, myocardial infarction and malignancies; thus, RA patients have decreased quality of life and shortened life expectancy. 4 Approximately half of the risk of the occurrence of RA is caused by genetic factors, 5 and female gender and smoking are also risk factors for RA. 2 Currently, disease-modifying anti-rheumatic drugs (DMARDs), such as methotrexate and sulfasalazine that can decrease joint inflammation and improve overall function, are the primary treatments for RA. 6 In addition, the infiltration of T or B cells and macrophages and the secretion of inflammatory cytokines, including tumour necrosis factor-α (TNF-α) and interleukins (ILs), are observed in the synovial tissues of RA patients. 7 Biologic treatments that target these cytokines have enhanced the outcomes of RA patients who show little response to conventional DMARDs. 8 However, approximately 50% of RA patients do not respond to anti-TNF-α treatments, and patients whose inflammation is controlled cannot have their bone erosions repaired. 9 Thus, other novel effective therapies are urgently needed.

Recently, human multipotent mesenchymal stem cells (MSCs)
have attracted wide-ranging interest as a novel therapy for cartilage defects, and their therapeutic potential has been shown to involve the secretion of extracellular vesicles (Evs). 10 Evs are heterogeneous, small vesicles that include three common types: exosomes, microparticles and apoptotic bodies. 11 Evs are known for their function in intercellular communication and their ability to transport their contents to cells and cause multiple changes in cell proliferation and cell transcription. 12 Evs can modulate many types of pathological and physiological processes, including inflammation and immune responses, tumour growth and infection. 13 Ev-based therapies possess promising potential for inflammatory and autoimmune diseases and tumours. 14,15 Evs are now known to transfer microRNAs (miRs) between cells, transmitting their miRs to target cells through endocytosis or membrane fusion. 16 It has been demonstrated that miRs in Evs can be transferred and function in immune cells, cancer cells and endothelial cells. 17 miRs exhibit tissue-specific expression patterns and are linked to various biological events, including cell growth, differentiation, apoptosis, cancer progression and autoimmune arthritis. 18 Deregulation of miR expression has been extensively observed in human diseases, including autoinflammatory and autoimmune diseases. 19 Jian et al demonstrated that elevation of miR-34a can alleviate microglial inflammation and inhibit neuronal apoptosis, thus improving nerve recovery and motor function. 20 In the current study, we examined the mechanism of bone marrow MSC (BM-MSC)-derived Evs in the development of RA and the involvement of miR-34a and its downstream pathway.

| BM-MSC-derived Evs extraction
Well-established BM-MSC cultures of passage 3 were washed 3 times with sterile phosphate-buffered saline (PBS), and then, the remaining medium was discarded and serum-free medium was added.
After incubation at 37°C for 24 hours, the supernatant was collected and considered the conditioned medium. Next, the medium was successively centrifuged at 300 g for 10 minutes, 2000 g for 30 minutes and 10 000 g for 1 hour, and then, the Evs were extracted after the cell debris and large granules were excluded via a 0.22-µm filter. The conditioned medium from cells supplemented with GW4869, an EV inhibitor, was used as a control. Then, the protein concentration of the Evs was evaluated using a bicinchoninic acid (BCA) assay kit, and then, the Evs were frozen at −80°C. 21 The extracted Evs were la- Next, the Evs were resuspended and dropped onto a sealing film and observed under an 80-kV transmission electron microscope (TEM). The levels of the Evs marker proteins CD63, CD81 and calnexin were detected using Western blot analysis, and the particle sizes of the Evs were measured using nanoparticle tracking analysis.

| Establishment of the rat model of RA
A total of 30 male Wistar rats (150 ± 20 g in weight; 6 weeks old) were purchased from the Laboratory Animal Center of Chinese Academy of Sciences and were separately housed in class II clean rooms with free access to standard feed and water. A week later, 20 rats were randomly selected, and each of the rats was intradermally injected with Freund's complete adjuvant F5881 (Sigma-Aldrich Chemical Company) through the left hindlimb toes at a dose of 0.1 mL/100 g; these rats were then assigned to the model group. The remaining 10 rats were injected with equal amounts of normal saline and assigned to the normal group. After establishing the model, 5 rats in the model group were injected with 75 μg/mL Evs (Evs group), 5 rats in the NC group were injected with GW4869 as a negative control via the tail vein, 5 rats in the model group were injected with 75 μg/mL Evs-NC (as Evs-NC group), and 5 rats in the Evs-IN group were injected with Evs-IN.
Then, the rat tissues were isolated, and the location of the Evs in the tissues was observed under a fluorescence microscope.

| Evaluation of rat foot swelling
The volumes of the rat right hindlimb toes (non-injection side) were measured one day before injection and at weeks 1, 2, 3 and 4 after injection. The degree of foot swelling was measured as follows: degree of swelling = average volume after injection -average volume before injection.

| Scoring of the polyarthritis index
The arthropathy of the rats was observed and recorded at weeks 1, 2, 3 and 4 after injection, and the polyarthritis index was evaluated with the following scores: 0 = no swelling; 1 = swelling of toes; 3 = swelling of foot claws below the ankle; 4 = swelling of all foot claws including the ankle. The total scores of both limbs were calculated as the polyarthritis index, and the maximum value was 12. 21

| Evaluation of the spleen and thymus indexes
Four weeks after establishing the model, the rats were killed, and their spleen and thymus were extracted and weighed. The index of the spleen (or thymus) of the rats was calculated as follows: spleen (or thymus) index = spleen weight (or thymus weight)/body weight.

| HE staining
The extracted synovial tissues were fixed with 10% formalin for 24 hours, decalcified with 10% ethylenediaminetetraacetic acid (EDTA) for 30 minutes, dehydrated, embedded in paraffin and subjected to HE staining. Histological changes were observed under an optical microscope.

| Immunohistochemistry
The tissue sections were fixed with 4% paraformaldehyde at 4°C for 30 minutes and then at room temperature for 10 minutes. Then, the sections were treated with 3% H 2 O 2 for 15 minutes to inactivate the endogenous peroxidase, washed with 10% goat serum 3 times and sealed for 20 minutes at room temperature. Next, the sections were

| ELISA
The expression of IL-6, IL-1β and TNF-α in the rat synovial fluid was detected using ELISA kits (Shanghai Bogoo Biological Technology Co., Ltd., Shanghai, China). The sandwich approach was applied during the assay as per the instructions of the kits.

| qRT-PCR
Total RNA was extracted from the synovial tissues using RNAiso Plus (Takara Holdings Inc, Kyoto, Japan) and TRIzol LS Reagent (Takara).
Formaldehyde denaturing gel electrophoresis was used to determine the RNA quality for the subsequent experiments. Next, RNA reverse transcription PCR (RT-PCR) was conducted strictly according to the instructions of the Prime Script TM RT assay kit (Takara).
The quantification of the mRNA and miR levels was conducted using SYBR Premix Ex Taq (Takara)-based qRT-PCR with glyceraldehyde-3-phosphate dehydrogenase (GAPDH) as the internal mRNA reference and U6 as the internal miR reference. The amplification curve and dissolution curve were confirmed after the reaction. The relative expression of the genes was calculated by the 2 −ΔΔCt method.
The primer sequences are shown in Table 1.

| Western blot analysis
Total proteins were extracted using radioimmunoprecipitation assay (RIPA) buffer containing phenylmethylsulfonyl fluoride (PMSF, Beyotime Biotechnology Co., Ltd, Shanghai, China), and the protein level in the supernatant was confirmed using a BCA assay. Next, equal volumes of proteins (50 mg) were separated by 10% sodium dodecyl sulphate-polyacrylamide gel electrophoresis and transferred to polyvinylidene difluoride (PVDF) membranes (Millipore Corp, Billerica, MA, USA). Then, the membranes were successively incubated with tris-buffered saline tween (TBST, Boster Biological Technology Co., Ltd, Wuhan, Hubei, China) containing 5% skim milk to block non-specific binding and then incubated with the indicated primary antibodies (all purchased from Abcam) ( Table 2) at 4°C overnight. After the membranes were incubated with rat antimouse antibodies at room temperature for 1 hour, the bands were visualized using enhanced chemiluminescence and imaged using a BioSpectrum system (Bio-Rad, Inc, Hercules, CA, USA).

| Rat synoviocyte separation and culture
Rat synoviocytes were extracted from the tissues using trypsinization.
When the synoviocytes covered more than 80% of the bottom of the culture flask, they were passaged; the cells were passaged a total of three times. The cells harvested from the normal rats were assigned to the FLS group, whereas those from the model rats were assigned to the RA-FLS group. FLSs and RA-FLSs in the logarithmic growth phase were suspended using DMEM containing 10% FBS, seeded into 24well plates at a density of 1 × 10 5 cells per well and cultivated in a 37℃ incubator with 5% CO 2 and 95% humidity for 48 hours. The cells in the Evs-IN group were treated with the p53-specific activator Tenovin-6 (0.5 μg/mL) or an equal volume of PBS for 48 hours and then collected for further use.

| RNA sequencing
The miRNAs of the cells in the RA-FLS group and RA-FLS-Ev group were extracted using a miRcute miRNA isolation assay kit (DP150, Beijing TianGen Biotech Co., Ltd. Beijing, China). The sequences of three samples from each group were determined by Shanghai Sangon Biotech Co., Ltd. (Shanghai, China). The differentially expressed miRNAs in the RA-FLS cells before and after EV treatment were screened, and a fold change >2 and P < 0.05 served as the screening criteria.

| Immunofluorescence assay
Slides of cells in each group were fixed with 4% paraformaldehyde at 4°C for 15 minutes, treated with 0.5% Triton X-100 for 20 minutes and then incubated with primary antibodies (all purchased from Abcam) ( Table 3)  at 4°C overnight. The next day, the cell slides were washed with PBS, incubated with goat anti-rabbit at 37°C for 1 hour and then observed under a fluorescence microscope (Leica DM 3000, Solms, Germany).

| 3-(4,5-Dimethylthiazol-2-yl)-2,5diphenyltetrazolium bromide (MTT) assay
The viability of the cells in each group was detected strictly according to the instructions of the MTT assay kit (C0009, Beyotime Biotechnology Co., Ltd, Shanghai, China), and the optical density (OD) value was measured at 490 nm. Cell cycle: The cell suspension was treated with −20°C 70% methanol (v/v) for 24 hours and then treated with 50 μg/mL PI containing 10 μg/mL RNase for 20 minutes. Then, the cell cycle distribution was measured.

| Flow cytometry analysis of Annexin V-FITC/ propidium iodide (PI) staining
Cell apoptosis: The cell suspension was stained with Annexin V-FITC (5 μL) and PI (5 μL) in the dark for 10 minutes. Then, cell apoptosis was detected.

| Statistical analysis
The Statistical Package for the Social Sciences (SPSS) 21.0 (IBM Corp. Armonk, NY, USA) was used for data analysis. The normality test was conducted using the Kolmogorov-Smirnov method. The measurement data were normally distributed and expressed as the mean ± standard deviation. Differences between pairs of groups were evaluated using the t test, whereas differences among multiple groups were compared using one-way analysis of variance (ANOVA). Tukey's multiple comparisons test was used for pairwise comparisons after ANOVA. The P value was calculated using a two-sided test, and P < 0.05 was considered to show a statistically significant difference.

| Identification of BM-MSCs and BM-MSCderived Evs
To identify the quality of BM-MSCs, we detected biomarkers of BM-MSCs and found that the cells were positive for CD44, CD73, CD90 and CD105, whereas the cells were negative for CD14, CD34, CD45 and HLA-DR ( Figure 1A). The Alizarin Red staining and Oil Red O staining results suggested that the BM-MSCs were able to differentiate into osteoblasts and adipocytes ( Figure 1B). Thus, the BM-MSCs qualified for the following experiments.
The size of the Evs was identified by a nanoparticle tracker, and the results showed that the diameter of the Evs was approximately 93 nm ( Figure 1C). The protein levels of the Ev markers CD63, CD81 and calnexin were measured by Western blot analysis. The results suggested that the Evs were positive for CD63 and CD81, whereas the Evs were negative for calnexin ( Figure 1D). According to the TEM observations, the extracted Evs were uniformly distributed with different sizes ( Figure 1E).

| BM-MSC-derived Evs reduce inflammation in rats with RA
To explore the role of Evs derived from BM-MSCs in RA, we in- synovial hyperplasia and inflammatory infiltration in the model and normal rats, whereas synovial hyperplasia and inflammatory infiltration in the rats treated with Evs were significantly reduced compared with those in the model rats ( Figure 2C). Moreover, the levels of TNF-α, IL-6 and IL-8 were notably higher in the synovial tissues and synovial fluid from the RA rats than in those from the normal rats, while these factors were significantly decreased in the RA rats after Ev injection ( Figure 2D-E).

| BM-MSC-derived Evs inhibit RA-FLS proliferation
To verify the experimental results in vivo, we named the cells

| BM-MSC-derived Evs mainly increase miR-34a expression
EVs play a vital role in intercellular communication and regulate recipient cells by transferring their contents, including proteins and nucleic acids such as miRs. 22 Based on the results described above that BM-MSC-derived Evs induce RA inflammation, the focus shifted to determining the potential molecular mechanisms involved in these effects. RNA sequencing was used to evaluate the differentially expressed miRNAs between the RA-FLSs before and after Ev treatment and the untreated RA-FLSs. A total of 16 miR-NAs showed differential expression, with 9 miRNAs up-regulated and 7 miRNAs down-regulated ( Figure 4A). Among the 9 miRNAs enhanced in the FLS-Evs, the most differentially expressed miRNA was miR-34a, the log 2 fold change of which was 7.69 (Table 4). In addition, the expression of miR-34a, miR-1180 and miR-123b in the RA-FLSs, RA-FLS-Evs and tissues of the model group and Ev group was detected using qRT-PCR, and the results suggested that the expression of miR-34a changed the most, and the expression of miR-34a exhibited the same trend as the RNA sequencing results ( Figure 4B).
To further identify the correlation between Evs and miR-34a expression, we treated BM-MSCs with a sphingomyelinase inhibitor (GW4869) to suppress Ev production and then detected miR-34a expression in the cells. The results showed that miR-34a mainly existed in Evs, and inhibition of Evs led to inhibition of miR-34a expression ( Figure 4C).
BM-MSCs were transfected with an miR-34a inhibitor or miR-NC using a Lipofectamine TM 2000 kit. Next, qRT-PCR confirmed the transfection ( Figure 4D). After the extraction of Evs, the expression of miR-34a in the Evs was detected. After inhibiting the expression of miR-34a, the expression of miR-34a in the Evs significantly decreased ( Figure 4E). The concentration of Evs was adjusted to 75 μg/ mL, and the EVs were added to the RA-FLSs. miR-34a expression in the cells treated with RA-FLS-Evs-IN was significantly lower than that in the cells treated with RA-FLS-Evs ( Figure 4F), indicating that miR-34a was carried into cells by Evs.

| Reduction of miR-34a in BM-MSC-derived Evs weakens the protective role of BM-MSC-derived Evs in rats with RA
Rats with RA were treated with Evs, Evs-NC or Evs-IN, and the immunofluorescence of PKH-26 suggested that the Evs were absorbed by the synovial tissues ( Figure 5A). Next, qRT-PCR  Figure 5D). In addition, the mRNA ( Figure 5E) and protein levels ( Figure 5F) of TNF-α, IL-6 and IL-8 in the tissues showed the same trend. The results of caspase-3 immunohistochemistry showed that after inhibiting miR-34a expression, positive caspase-3 expression was decreased ( Figure 5G).

| miR-34a contained in BM-MSC-derived Evs inhibits RA-FLS proliferation and induces RA-FLS apoptosis
Immunofluorescence of PKH26, a specific marker of EVs, showed that Evs in the Ev group were absorbed by RA-FLSs ( Figure 6A).
The qRT-PCR results suggested that miR-34a expression was significantly reduced in RA-FLSs transfected with Evs-IN ( Figure 6B).

| miR-34a targets cyclin I and activates the ATM/ATR/p53 signalling pathway
The computer-based miR target detection program (http://www. targe tscan.org) predicted that miR-34a could directly bind to the 3' UTR of cyclin I ( Figure 7A). The binding relationship was also identified using a dual-luciferase reporter gene assay. It was shown that the relative luciferase activity showed little difference between the MT + miR mimics group and the MT + miR-NC group, but the luciferase activity was notably reduced in the cells transfected with WT + miR mimics (P < 0.01) ( Figure 7B). Hence, these results showed that miR-34a could negatively target cyclin I and then activate the ATM/ATR/p53 signalling pathway, thus inhibiting cell proliferation and cell apoptosis resistance.
F I G U R E 5 miR-34a suppresses inflammation in RA rats. A, immunofluorescence of PKH26 in synovial tissues; B, miR-34a expression in synovial tissues detected using qRT-PCR; C, evaluation of RA inflammation in rats; D, inflammation detected using HE staining; E&F, expression of TNF-α, IL-6 and IL-8 in rat synovial tissues detected using qRT-PCR (E) and synovial fluid detected using ELISA (F); G, caspase-3 expression measured using immunohistochemistry; n = 3; *compared with the NC group; *P < 0.05, **P < 0.01 F I G U R E 6 miR-34a inhibits RA-FLS proliferation and promotes RA-FLS apoptosis. A, Immunohistochemistry of PKH26 in RA-FLSs; B, miR-34a expression in RA-FLSs detected using qRT-PCR; C, cell proliferation detected using flow cytometry; D, cell viability detected using MTT assay; E, immunofluorescence of PCNA; F, flow cytometry used to detect apoptosis; G, immunofluorescence of cytochrome C; H, expression of Bax, PUMA and cleaved PARP detected using Western blot analysis, *compared with the blank group; n = 3; *P < 0.05, **P < 0.01; compared with the Evs group, #P < 0.05, ## P < 0.01, ###P < 0.001

| Cyclin I inactivates the ATM/ATR/p53 signalling pathway
To

| D ISCUSS I ON
RA is a common chronic autoimmune disease that leads to articular destruction and related vascular, metabolic and bone comorbidities. 23 Evs have been recognized as key mediators of intercellular communication, and EVs are capable of transmitting biological signals between cells and then regulating diverse biological processes. 15 Evs can carry a variety of miRNAs. 24 In this study, we established a The anti-inflammatory role of BM-MSC-derived Evs prompted us to further investigate the mechanisms involved in this event.
Importantly, the in vitro study suggested that miR-34a is the most abundant miR in BM-MSC-derived Evs. Inhibition of miR-34a led to significantly enhanced inflammation but decreased caspase-3 expression in RA rats, and it led to enhanced proliferation of RA-FLSs and decreased expression of cytochrome C, Bax, PUMA and cleaved PARP in RA-FLSs. miR-34a is a well-recognized potential tumour suppressor, and its anti-inflammatory attributes have been shown. 30 Moreover, miR-34a has been suggested to enhance the expression of apoptosis-related factors, including PUMA, thus leading to apoptosis. 31 Caspase-3, cytochrome C, Bax, PUMA and cleaved PARP are well-recognized apoptotic-related factors. 32,33 The secretion of cytochrome C further activates downstream caspases, such as caspase-3 and cleaved PARP, to initiate cell apoptosis. 34 Moreover, our study showed that miR-34a could directly bind to cyclin I. It has been suggested that miR-34a could target another member of the cyclin family, namely cyclin D1, in a previous study. 35 The suppression of miR-34a significantly elevated the expression of cyclin I and reduced the expression of p53, ATM and ATR in synovial tissues and RA-FLSs. miR-34a knockdown aggravated inflammation (elevated levels of TNF-α and IL-6), inhibited cell proliferation and enhanced apoptosis in an in vitro model of spinal cord injury. 36 Cyclin I is an atypical cyclin known for its prosurvival and anti-apoptosis functions by mediating the activation of cyclin-dependent kinases (CDKs), 37 while ATM/ATR is a cell cycle checkpoint signalling pathway whose activation could lead to cell cycle arrest and cell apoptosis by inactivating CDKs. 38,39 Furthermore, a previous study showed that the induction of cyclin D1 could inhibit the phosphorylation of ATM/ ATR targets, thus inactivating the ATM/ATR signalling pathway. 40 Thus, we concluded that miR-34a could negatively bind to cyclin I to activate the ATM/ATR/p53 signalling pathway, thus inhibiting the resistance of cell cycle arrest and reducing the inflammation of RA.
In summary, the current study provided evidence that miR-34a contained in BM-MSC-derived Evs could bind to the 3' UTR of cyclin I and further enhance the activation of the ATM/ATR/p53 signalling pathway, thus promoting RA-FLS apoptosis and reducing RA inflammation. These results provide novel insights into the development of new therapies for RA prevention. However, the present study is limited by a lack of experiments on the mechanism of miR-34a in Evs. Additionally, we will further study the effect of human BM-MSCs in a rat model of RA in the future. Hopefully, more studies will be carried out in this field to reveal more details on the role of BM-MSC-derived Evs in RA and to develop more strategies for RA prevention and treatment.

E TH I C A L S TATEM ENT
The current study was approved by the Clinical Ethical Committee of Shenzhen People's Hospital. All the experimental procedures were conducted in line with the ethical guidelines for studies involving experimental pain in conscious animals.

DATA AVA I L A B I L I T Y S TAT E M E N T
All the data generated or analysed during this study are included in this published article.