Long non‐coding RNA MEG3 inhibits M2 macrophage polarization by activating TRAF6 via microRNA‐223 down‐regulation in viral myocarditis

Abstract Viral myocarditis (VMC) commonly triggers heart failure, for which no specific treatments are available. This study aims to explore the specific role of long non‐coding RNA (lncRNA) maternally expressed 3 (MEG3) in VMC. A VMC mouse model was induced by Coxsackievirus B3 (CVB3). Then, MEG3 and TNF receptor‐associated factor 6 (TRAF6) were silenced and microRNA‐223 (miR‐223) was over‐expressed in the VMC mice, followed by determination of ventricular ejection fraction (LVEF) and left ventricular fractional shortening (LVFS). Dual‐luciferase reporter assay was introduced to test the interaction among MEG3, TRAF6 and miR‐223. Macrophages were isolated from cardiac tissues and bone marrow, and polarization of M1 or M2 macrophages was induced. Then, the expressions of components of NLRP3 inflammatory body (NLRP3, ASC, Caspase‐1), M1 markers (CD86, iNOS and TNF‐α) and M2 markers (CD206, Arginase‐1 and Fizz‐1) were measured following MEG3 silencing. In the VMC mouse model, MEG3 and TRAF6 levels were obviously increased, while miR‐223 expression was significantly reduced. Down‐regulation of MEG3 resulted in the inhibition of TRAF6 by promoting miR‐223. TRAF6 was negatively correlated with miR‐223, but positively correlated with MEG3 expression. Down‐regulations of MEG3 or TRAF6 or up‐regulation of miR‐223 was observed to increase mouse weight, survival rate, LVEF and LVFS, while inhibiting myocarditis and inflammation via the NF‐κB pathway inactivation in VMC mice. Down‐regulation of MEG3 decreased M1 macrophage polarization and elevated M2 macrophage polarization by up‐regulating miR‐223. Collectively, down‐regulation of MEG3 leads to the inhibition of inflammation and induces M2 macrophage polarization via miR‐223/TRAF6/NF‐κB axis, thus alleviating VMC.


| INTRODUC TI ON
Myocarditis is an inflammatory disease that occurs in the myocardium, which results in significant alterations in cardiac physiological functions, including reduction in the pumping force of the cardiac muscles and abnormal heart rhythms. The incidence of myocardial inflammation is less than 10% of the population. 1 However, as the obvious clinical symptoms are not conspicuous at an early stage of myocarditis, the disease often progresses without treatment until the emergence of severe symptoms like chest pain, shortness of breath and body fatigue, which can even progress to chronic heart failure in worse condition. 2 Etiologies of myocarditis include pathogens, toxins, hypersensitivity and immunologic syndromes, among which viruses such as Coxsackievirus B3 (CVB3), hepatitis C virus (HCV), human immunodeficiency virus (HIV) and adenovirus have been identified as the prevalent factors. 3 In particular, CVB3 plays a crucial role in the pathogenesis of viral myocarditis (VMC) and can be used to induce a VMC animal model. 4,5 Following an infection, if the immune system fails to eliminate a viral infection, autoimmunity can be triggered against the myocardium, which entails the activation of infiltrated macrophages, autoreactive T cells, cytokine and production of cross-reacting antibodies. 6,7 Therefore, one can view the pathogenesis of VMC as an inflammatory process. Endo-myocardial biopsy, in combination with histological, immunological and molecular techniques aids the diagnosis of VMC. Unfortunately, an accurate diagnosis can be difficult to obtain in early disease stages of VMC, when only a few lymphocytic infiltration loci are presented. 6 Therefore, it is urgent to explore the molecular mechanism underlying VMC, which is beneficial for timely diagnosis and management of VMC.
Non-coding RNAs, including long non-coding RNAs (lncRNAs) and microRNAs (miRNAs), play critical roles in maintaining regular physiological processes as well as regulating pathological changes. 8,9 Maternally expressed gene 3 (MEG3) has been identified to be able to control cancer development, 10,11 angiogenesis 12 and differentiation of mesenchymal stem cells. 13 Indeed, MEG3 has been determined to be an inducer of cardiac fibrosis and diastolic dysfunction, which characterizes cardiac remodelling following cardiac injury. 14 MEG3 is reportedly associated with and down-regulation of miR-223, 15 which is a heart-enriched miRNA that protects against CVB3-mediated VMC damages through its regulation on M1/M2 macrophage polarization. 16 The imbalance of M1/M2 macrophage polarizations is a key factor in local inflammatory response and tissue repair. 17 Differential polarization of macrophages has been reported to may generate an opposite inflammatory response. 18 For example, M1 macrophages could aggravate myocarditis and M2 macrophages could alleviate myocardial inflammation. [19][20][21] M1 macrophages, which are activated by interferon γ (IFN-γ) or tumour necrosis factor (TNF), have a proinflammatory action to kill invading microbes. M2 macrophages, which are activated by interleukin (IL)-4, IL-10 or IL-13, play anti-inflammatory role to reduce autophagy and promote tissue repair. 22 M2 macrophages have been demonstrated to be related to protection against CVB3-induced myocarditis. 23 Moreover, a previous study has shown that the progression of VMC involves tumour necrosis factor receptor-associated factor number six (TRAF6) and nuclear factor κB (NF-κB) pathway. 24 Ge et al reported that TRAF6 promoted lipopolysaccharide (LPS)induced inflammatory injury in BV2 mouse microglial cells through the activation of the NF-κB pathway. 25 In addition, it has been proven that NF-κB activation is a pathological event that promotes local inflammation, 26 which implies the central role of the NF-κB pathway in VMC. However, the underlying mechanism of TRAF6 and the NF-κB pathway in regulating VMC remains little understood. In the current study, we explored the roles of MEG3 in the regulation of VMC and found that inhibition of MEG3 can alleviate VMC via up-regulation of miR-223 and down-regulation of the NF-κB pathway.

| Study objects
Male BALB/c mice (n = 105; aged 6 or 7 weeks) were purchased from Hunan Slac Jingda Laboratory Animal Co., Ltd. (Changsha, China). All mice were maintained in a specific pathogen-free (SPF) room with controlled temperature and humidity under a 12:12 hour light/dark cycle.

| Mouse model of VMC
CVB3 was purchased from American type culture collection (ATCC, Manassas, VA, USA), and the 50% tissue culture infectious dose (TCID50) was measured before the infection of HeLa cells. The VMC model was established by intraperitoneally injecting the BALB/c mice with 0.1 mL phosphate-buffered saline (PBS) containing CVB3 (1 × 10 3 TCID50). These mice were assigned randomly to several groups, which were subjected to different treatments (15 mice per group): VMC mice were infected with lentivirus expressing short hairpin MEG3 (shMEG3), miR-223 agomir or shTRAF6. In pre-experiments, we used three shRNAs to silence MEG3, and the shRNA (sh-MEG3-3) with best silencing efficacy under fluorescence microscope was selected for subsequent experiments ( Figure S1). The aforementioned lentivirus vectors were purchased from GeneChem (Shanghai, China). Green fluorescent protein (GFP) method was used to detect the expression of lentivirus vector in cells. Therefore, the titre was expressed as 'gene transfer unit (GTU)/mL' rather than 'plaque forming unit (PFU)/mL'. On the first and third day after CVB3 injection, mice were intraperitoneally injected with lentivirus at a dose of 5 × 10 4 gtu/mouse. The remaining part of the experiment was conducted in accordance with a previously described procedure. 27 On the seventh day following the viral infection, cardiac tissues or infiltrated macrophages in cardiac tissues were collected.
Subsequently, 5 mice in each group were used for histological analysis, and 10 mice were utilized for RT-qPCR and Western blot analysis.

| Isolation and culture of macrophages
Isolation of infiltrated macrophages from cardiac tissues was performed as previously described. 20 In brief, cardiac tissues were extracted aseptically, and cut into 1 mm 3 pieces, which were then digested with 0.01% hyaluronidase and 0.1% collagenase II for 2 hours.
Then, Ficoll density gradient separation was applied for the isolation of inflammatory cells in infiltrated heart. To harvest macrophages, inflammatory cells were stained with fluorescein isothiocyanate (FITC)-labelled anti-F4/80 monoclonal antibody (BD Bioscience, San Jose, CA, USA), followed by isolation using flow cytometry (FACS) (BD Biosciences, San Jose, CA, USA).
Subsequently, the isolation of bone marrow-derived macrophages (BMDMs) was carried out based on the outline as Bauerfeld described in a previous study. 28 In brief, femurs and tibias were dissected from normal mice and bone marrow was removed. BMDMs were cultured in Dulbecco's minimal essential medium (DMEM) supplemented with 30% L929-conditioned medium, 10% fetal bovine serum (FBS) and 2 mmol/L glutamine for 7 days. Next, the cells were cultured in a 6-well plate with 1 × 10 6 cells per well. After 1 day in culture, the cells were polarized by RPMI 1640 containing 5% FBS, 10 ng/mL lipopolysaccharide (LPS) and 20 ng/mL IFN-γ (for M1 macrophage polarization) or 20 ng/mL IL-4 (for M2 macrophage polarization).

| Transthoracic echocardiography
Following the operator's manufacturer, transthoracic echocardiography was carried out using an ultrasound imaging system (Acuson Sequoia C256, Siemens, Erlangen, Germany) and 13 MHz transducer Myocarditis grading (0-4) was evaluated as previously described 30 in a blinded fashion by 2 independent investigators.

| Immunohistochemistry (IHC)
The cardiac tissue sections were dewaxed with xylene and rehydrated with gradient ethanol, whereupon incubation was carried out with citrate solution for antigen retrieval with high pressure. Each section was incubated with 50 μL 3% H 2 O 2 at room temperature for 20 minutes.
Then, the sections were incubated with the primary rabbit anti-TRAF6 antibody (1:1000, ab227560, Abcam Inc, Cambridge, UK) overnight at 4, followed by incubation with 50 μL polymer reinforcing agent for 20 minutes at 37°C and 50 μL enzyme-labelled anti-rabbit polymer for 30 minutes at 37°C. Each section was then incubated with 100 μL diaminobenzidine (DAB) for 3-10 minutes. The specimens were counterstained with hematoxylin, dehydrated with gradient ethanol, mounted with neutral resin and observed under the microscope.

| Enzyme-linked immunosorbent assay (ELISA)
The level of IL-6, IL-β and IFN-γ in supernatant of heart homogenates was determined with the relevant ELISA kit (R&D Systems, Minneapolis, MN, USA) in accordance with the manufacturer's protocol.

| RNA isolation and quantitation
The total RNA was extracted from cells using TRIzol (Invitrogen,  (Table 1).

| Fluorescence in situ hybridization (FISH) assay
The subcellular localization of MEG3 was determined in accordance with the instructions of the FISH Tag™ RNA Green Kit (RiboBio Co., Ltd., Guangzhou, China). In brief, cells were inoculated into a 6-well culture plate and cultured for 1 day. When cell confluence reached 80%, the cells were rinsed with PBS and fixed with 1 mL of 4% polyformaldehyde at room temperature, followed by treatment with protease K (2 μg/mL), glycine and phthalide reagent. Next, 250 μL of pre-hybridization solution was added into the cells for incubation at 42°C for 1 hour. After removing the pre-hybridization solution, the cells were subsequently hybridized with 250 μL of hybridization solution containing 300 ng/mL probes at 42°C overnight. Following this, the cells were washed using phosphate-buffered saline/Tween (PBST) three times, and the cell nucleus was stained using PBST-

| NF-κB DNA-binding activity assay
NF-kB DNA-binding activity was determined using Nuclear Extract kit and Trans-Am NF-κB/p65 ELISA kit (Active Motif, Carlsbad, CA, USA) 31 in accordance with the manufacturers' instruction.

| Statistical analysis
The experimental data were analysed with SPSS 21.0 software (IBM Corp., Armonk, NY, USA) and described as mean ± SD. Unpaired t test was employed to compare data between two groups, while oneway analysis of variance (ANOVA) was utilized for testing differences among multiple groups, with Tukey's post hoc test conducted. The comparison of mouse weight at different time point was analysed with repeated measurement ANOVA, followed by Bonferroni post hoc test. Pearson's correlation was applied for correlation analysis.
P < 0.05 was considered as statistically significant.

| MEG3 was highly expressed in VMC mice
Firstly, we investigated whether MEG3 was up-regulated in VMC. After establishing the upon injection of CVB3, pathological changes of myocardial tissues were observed with H&E staining. As depicted in Figure 1A,B, VMC mice presented with severe myocarditis symptoms ( Figure 1A,B). Meanwhile, the body weights ( Figure 1C, P < 0.05) and survival rates ( Figure 1D) of VMC mice were significantly reduced. Echocardiography ( Figure 1E) showed that LVEF and LVFS of VMC mice were reduced ( Figure 1F,G). As shown by ELISA, the contents of IFN-γ, IL-6 and IL-1β in myocardial tissues of VMC mice were markedly increased ( Figure 1H). These findings were indicative of successful establishment of VMC mouse model. MEG3 expression was then examined by RT-qPCR, which showed that MEG3 was expressed at a higher level in myocardial tissues and cardiac infiltrated macrophages of the VMC mice than in normal mice ( Figure 1I, P < 0.05). Meanwhile, FISH assay was employed to determine the subcellular localization of MEG3, and the results are illustrated in Figure 1J, which revealed that MEG3 expressed both in nuclei and cytoplasm. These data demonstrated that MEG3 expressed at a high level in mice with VMC.

| Down-regulation of MEG3 alleviates myocarditis in mice
To explore the effects of MEG3 on VMC, MEG3 expression was

| MEG3 regulates TRAF6 expression through miR-223 inhibition
Previous studies have reported that MEG3 can bind to miR-223 and inhibit the expression of miR-223 expression and that miR-223 is down-regulated in mice with myocarditis. 15  in VMC, with logFC > 2.5 and P < 0.05 as threshold ( Figure S2A,B).
Then, Venny (v.2.1) tool was used to intersect the genes obtained from these two databases, which finally screened out 19 candidate target genes ( Figure S2C). Based on previous investigations, we found TRAF6 could participate in the development of VMC. [33][34][35] Therefore, we chose TRAF6 as the downstream gene of miR-223 for further study. By online prediction on Targetscan, a binding site between miR-223 and TRAF6 was predicted ( Figure 3C). Then, with an attempt to verify if TRAF6 was the target gene of miR-223, the dual-luciferase reporter assay was carried out ( Figure 3D). It was revealed that the luciferase activity of wt-TRAF6 was reduced by co-transfection with miR-223 mimic. However, the luciferase activity of mut-TRAF6 showed no significant difference between co-transfection with miR-223 mimic and NC mimic (P > .05). Next, based on RT-qPCR and Western blot analysis, we found that the mRNA and protein expressions of TRAF6 were significantly decreased by miR-223 mimic ( Figure 3E,F). On the other hand, miR-223 expression was significantly increased and TRAF6 expression was significantly reduced following treatment with sh-MEG3, whereas opposite effects were seen after treatment with overexpression (oe)-MEG3 ( Figure 3G-I). The aforementioned results suggested that MEG3 regulated the expression of TRAF6 through mediating miR-223.

| Decrease of MEG3 alleviates myocarditis via miR-223 and TRAF6
Based on the finding that MEG3 regulates TRAF6 by binding to miR-223, we wished to explore whether miR-233 and TRAF6 are involved in the mediation of myocarditis by MEG3. Results showed that miR-223 expression was decreased in VMC mice, which was reversed after MEG3 was down-regulated ( Figure 4A). Western blot analysis and IHC illustrated that the TRAF6 level was increased in VMC mice, which was blocked when MEG3 expression was reduced ( Figure 4B Figure 4N). Therefore, down-regulation of MEG3 alleviated myocarditis by up-regulating miR-223 and inhibiting TRAF6.

| Down-regulation of MEG3 inhibits the NF-κB pathway through miR-223/TRAF6 axis
To further investigate whether MEG3/miR-223/TRAF6 affected the NF-κB pathway, we heuristically examined the activity of the NF-κB pathway through probing the binding activity of NF-κB DNA as well as the phosphorylation levels of p65, IKBα, IKKα and IKKβ using western blot analysis. As Figure 5A displayed, NF-κB DNA binding activity was significantly increased in VMC mice, which was decreased following the down-regulation of MEG3. Meanwhile, the expression and phosphorylation levels of p65, IKBα, IKKα and IKKβ were increased obviously in VMC mice, which were declined after MEG3 was down-regulated ( Figure 5B). Therefore, down-regulation it was found that miR-223 overexpression or TRAF6 knockdown in VMC models led to a significant decline of NF-κB DNA binding activity ( Figure 5C) and an obvious reduce of p65, IKBα, IKKα and IKKβ expression and their phosphorylation ( Figure 5D). The aforementioned results led us to conclude that MEG3 down-regulation inhibits the NF-κB pathway through miR-223 over-expression and TRAF6 inhibition.

| Decreasing MEG3 suppresses M1 macrophage polarization but promotes M2 macrophage polarization via miR-223
Prior studies reported that M1 macrophages typically promoted myocarditis and tissue damage through generating proinflammatory cytokines, while M2 macrophages secreted anti-inflammatory cytokines relevant to tissue repair. 22,36 Based on a previous study that miR-223 inhibited M1 macrophage polarization and promoted M2 macrophage polarization, 16 which led to our hypothesis that MEG3 might mediate macrophage polarization by regulating miR-223.
Arginase-1, Fizz-1 and CD206 were shown to be up-regulated in M2 macrophages while the production of iNOS, TNF-α and CD86 was specific to M1 macrophages. 37 Thus, RT-qPCR and flow cytometry assay were carried out to detect the markers of M1-polarization (iNOS and TNF-α) and M2-polarization (Arginase-1 and Fizz-1) as well as CD86 and CD206, respectively. Results showed significantly increased levels of iNOS and TNF-α in macrophages in heart of VMC mice, which were blocked in response to MEG3 down-regulation ( Figure 6A). At the same time, the levels of Arginase-1 and Fizz-1 in macrophages in heart of VMC mice showed no significant difference between VMC mice and normal mice, but were increased after F I G U R E 4 Down-regulation of MEG3 alleviates myocarditis through inhibition of TRAF6 and increasing miR-223 expression. A, miR-223 expression after silencing MEG3 determined by RT-qPCR. B, TRAF6 expression after silencing MEG3 examined by Western blot analysis. C, TRAF6 expression after silencing MEG3 monitored by IHC (×200). *P < 0.05 vs control mice; # P < 0.05 vs viral myocarditis (VMC) mice injected with sh-NC. D, Correlation between expression of MEG3 and that of TRAF6 in myocardial tissues from VMC mice analysed by Pearson's correlation. E, Correlation between expression of miR-223 and expression of TRAF6 in myocardial tissues from VMC mice analysed by Pearson's correlation. F, TRAF6 expression after silencing TRAF6 or over-expressing miR-223 monitored by IHC (×200). G, Expression of NLRP3, ASC and Caspase-1 in cardiac tissues after silencing TRAF6 or over-expressing miR-223 detected by RT-qPCR. H, Change of mice weight after silencing TRAF6 or over-expressing miR-223. I, Survival rate of VMC mice after silencing TRAF6 or over-expressing miR-223. J, H&E staining of myocardial tissues (×200) after silencing TRAF6 or over-expressing miR-223. K, Evaluation of myocarditis after silencing TRAF6 or over-expressing miR-223. L, Transthoracic echocardiography to measure LVEF after silencing TRAF6 or over-expressing miR-223. M, Transthoracic echocardiography to examine LVFS after silencing TRAF6 or over-expressing miR-223. N, IFN-γ, IL-6 and IL-1β contents in myocardial tissues after silencing TRAF6 or over-expressing miR-223 tested by ELISA. *P < 0.05 vs VMC mice injected with NC agomir; # P < .005 vs VMC mice injected with sh-NC. The measurement data were described as mean ± SD. Statistical comparisons between two groups were analysed by unpaired t test, and repeated measurement ANOVA was used for comparing mice weight at different time point, followed by Bonferroni post hoc test. N = 5 or 10 down-regulation of MEG3 ( Figure 6B). Then, to explore whether MEG3 affected macrophage polarization, we isolated primary BMDMs and carried out M1 and M2 macrophage polarization induction experiments in vitro. Results suggested that after induction of M1 macrophage polarization, silencing MEG3 resulted in significant decrease in levels of iNOS and TNF-α, which was reversed in response to miR-223 inhibition ( Figure 6C). After inducing M2 macrophage polarization was induced, MEG3 knockdown increased the expression of Arginase-1 and Fizz-1, which was reversed by downregulation of miR-223 ( Figure 6D). Subsequently, results of the flow cytometry assay indicated that when M1 macrophage polarization was induced, CD86 level was declined markedly after MEG3 expression was decreased, which was reversed when miR-223 was down-regulated ( Figure 6E). Following M2 macrophage polarization induction, the CD206 level was increased when MEG3 was silenced, which was blocked by inhibition of miR-223 ( Figure 6E). In summary, silencing MEG3 inhibits M1 macrophage polarization but boosts M2 macrophage polarization by promoting miR-223. *P < 0.05 vs control mice; # P < 0.05 vs viral myocarditis (VMC) mice injected with sh-NC. C, NF-κB DNA binding activity after miR-223 over-expression and TRAF6 down-regulation. D, Expression of the NF-κB pathway components after miR-223 over-expression or TRAF6 knockdown determined by western blot analysis. *P < 0.05 vs VMC mice injected with NC agomir; # P < 0.05 vs VMC mice injected with sh-NC. All data were presented as measurement data and expressed as mean ± SD. Unpaired t test was made in comparisons. N = 5 or 10    Figure 7). Therefore, the MEG3/miR-233/TRAF6/NF-κB axis should be considered as a promising therapeutic target for VMC.

CO N FLI C T O F I NTE R E S T
The authors declare that they have no conflict of interest.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data sets used and/or analysed during the current study are available from the corresponding author on reasonable request.