IL-9^−/− mice in a Balb/c background were generated as previously described (8) and were provided by the Laboratory of Molecular Biology, Medical Research Council, Cambridge, UK. Wild-type male Balb/c mice were purchased from the Experimental Animal Center of Hubei province (Wuhan, China). All the animals were housed under standard pathogen-free conditions at the Experimental Animal Center (Tongji Medical College of Huazhong University of Science and Technology, Wuhan, China). The animal experiments were carried out according to the guidelines for the Care and Utilization of Laboratory Animals (Huazhong University of Science and Technology, China). And this study was approved by the Institutional Animal Care and Use Committee of Tongji Medical College, Huazhong University of Science and Technology according to the regulations for the administration of affairs concerning experimental animals in Hubei province of China and the constitution of the experimental animal ethics committee in Huazhong University of Science and Technology.

The CVB3 (3 m strain, CCTCC GDV115) titer determined by plaque-forming unit (PFU) assay in HeLa cells was 1 × 10⁷. IL-9^−/− and WT BALB/c mice aged 4 weeks were infected by an intraperitoneal (i.p.) injection of 0.2 mL of RPMI-1640 (Gibco) containing approximately 10⁵ PFU of CVB3 to establish the VMC models. The virus experiments were performed according to the general requirements for laboratory biosafety (GB 19489–2008) in China.

IL-9KO and WT BALB/c mice were divided into four groups randomly: (1) control group (n = 20) containing WT BALB/c mice treated with saline (300 μL per mouse); (2) WT group (n = 20) comprising WT BALB/c mice treated with 200 μL CVB3 and 100 μL saline; (3) IL-9KO group (n = 20) including IL-9^−/−Balb/c mice treated with 200 μL CVB3 and 100 μL saline; and (4) rIL-9 group (n = 20) consisting of mice treated with 200 μL CVB3 and rIL-9 (1 μg diluted in 100 μL saline per mouse, PeproTech). The intraperitoneal injections of CVB3 and saline were developed on day 0. The rIL-9 was administered intraperitoneally on days 0 and 3. All the animals in each group were euthanized on days 5 and 7. The blood and hearts were removed aseptically for further measurements.

Hearts of mice were minced into 1 mm³ sections and digested with 0.1% collagenase B (Roche Diagnostics GmbH) for 6 min four times in a 37°C water bath (9). Cell suspensions were obtained by filtering through a cell strainer (40 μm size, BD Falcon) and layered over Ficoll–Hypaque density gradient solution to separate mononuclear cells for flow cytometry. Intracardiac IL-9-producing leukocytes were measured by labeling the harvested cells with the following surface markers: PE anti-mouse CD45, FITC anti-mouse CD4, FITC anti-mouse CD49b, PE-cy7 anti-mouse CD11b, PE-cy7 anti-mouse CD8, or PE anti-mouse Gr-1 antibodies (eBioscience). After washing with PBS, these cells were stimulated with 1 μg/mL ionomycin, 20 ng/mL phorbol myristate acetate (PMA), and 2 μmol/L monensin (eBioscience) for 4 h under 5% CO₂ at 37°C in 24-hole culture plates (Costar). After washing, fixing, and permeabilizing according to the manufacturer’s instructions, the cells were stained with APC anti-mouse IL-9 antibody or isotype control antibody. The stained cells were measured and analyzed by FACScalibur flow cytometry (BD Biosciences).

The heart was fixed in 4% paraformaldehyde for 24 h, trimmed, and embedded routinely in paraffin. Longitudinal, 5-mm-thick sections of heart were obtained for staining with hematoxylin and eosin. The severity of impairment was assessed by the percentage of cardiac sections showing inflammation compared with the overall size of the heart sections, under a microscope eye piece grid (magnification 200×) according to the following scoring system: grade 0, none; grade 1, 25% cardiac inflammation; grade 2, 25–50%; grade 3, 50–75%; and grade 4, more than 75% (10). Two independent researchers scored the results in a blinded manner.

To further evaluate the cardiac expression of IL-9, the sections were heated in a microwave using 0.01% citrate buffer (pH = 9.0) and treated with 3% H₂O₂ for 10 min. After washing with PBS buffer three times and blocking with 3% bovine serum albumin (BSA) for 30 min, the sections were incubated with hamster anti-mouse IL-9 IgG (eBioscience) at 4°C overnight and washed with PBS buffer three times. After incubation with HRP-conjugated anti-hamster antibody for 45 min and washing adequately, diaminobenzidine solution was added, and the sections were counterstained by hematoxylin.

Serum levels of IL-9, IL-17a, TGF-β, IL-10, TNF-α, IFN-γ, IFN-α, and IFN-β were determined using sensitive mouse IL-9 (Biolegengd), TNFα/IFN-γ/TGF-β/IL-10/IL-17a (Neobioscience), IFN-α (eBioscience), and IFN-β (Pbl Assay Science) kits, according to the manufacturers’ instructions. No cross-reactivity was detected. Blood concentrations of serum cardiac troponin (cTn) T were measured using a quantitative rapid assay kit (Roche Diagnostics GmbH Elecsys, Shanghai, China) as previously described (11). All the samples were measured in triplicate.

A portion of the heart was weighed and homogenized in PBS. After three freeze-thaw cycles and centrifugation at 2000 rpm for 10 min, the supernatant was obtained and sequentially diluted 1:10 in RPMI 1640 medium. The HeLa cell monolayers were cultured in six-well plates with the supernatant for 1 h at 37°C, 5% CO₂. They were washed in PBS and covered with 2 mL 0.4% agar, RPMI 1640, and 10% FBS (Gibco). After 72 h of incubation, the number of plaques was counted. The viral titers were analyzed by standard plaque formation assay and expressed per organ weight (gram).

Neonatal cardiomyocytes were isolated as previously described (12). The ventricles obtained from 1–3 days BALB/c mice were removed rapidly into cold Hanks’ balanced salt solution (HBSS). After washing and mincing, tissues were digested in 0.05% trypsin (GIBCO) for 30 min at 4°C with rotation. The tissues were transferred into DMEM (GIBCO) containing 20% FBS (Fetal bovine serum, Gibco) to terminate the digestion. After washing with HBSS, the tissues were incubated with Liberase TH (0.1 U/mL, Roche, Germany) at 37°C for 5 min, and the dissociated cells were collected into 20% FBS DMEM. This procedure was not repeated until most of the cells were released. The isolated cells were incubated with 5% CO₂ at 37°C for 1 h. The unattached cardiomyocytes were seeded into fibronectin-coated 12-well tissue culture plates (Costar) and subsequent experiments were performed when the cardiomyocytes formed a confluent monolayer and beat in synchrony at 72 h.

The isolated neonatal cardiomyocytes were divided into three groups: (1) control group, including neonatal cardiomyocytes treated with 50 μL PBS; (2) CVB3 group, comprising neonatal cardiomyocytes incubated with CVB3 at 5 × 10⁵ PFU in 50 μL PBS; (3) IL-9 group, containing neonatal cardiomyocytes incubated with CVB3 at 5 × 10⁵ PFU in 25 μL PBS and 500 ng/mL IL-9 diluted in 25 μL PBS; and (4) IL-9 + TGF-β monoclonal antibody (mAb) group, administrating 10 μg/mL TGF-β mAb (eBioscience, 1 mg/mL) in the IL-9 group. After 48 h, the plaque-forming assay was developed as above. The protein was extracted for Western Blot test. The supernatant was tested for TGF-β, TNF-α, IFN-α, and IFN-β, using ELISA, as mentioned above.

The cardiomyocytes proliferated over the coverslips in 12-well culture plates and were incubated with CVB3 at 5 × 10⁵ PFU or CVB3 + 500 ng/mL IL-9, as described above. After 12 h, the RNA was extracted and subjected to real-time PCR for Bax/Bcl-2 analysis. The TUNEL assay was performed using an in situ cell death detection kit (Roche) according to the manufacturer’s protocol. The TUNEL-stained slides were washed with PBS and counterstained with α-SMA (Boster, Wuhan, China) and 4′,6-diamidino-2-phenylindole (DAPI; Beyotime, Shanghai, China). A laser confocal microscope (Olympus, Tokyo, Japan) was used to acquire the images. Nuclei, which were labeled with both TUNEL and DAPI, were considered TUNEL-positive.

Total proteins of the heart tissue or cardiomyocyte were extracted with the total protein extraction kit (Pierce/Thermo Scientific, USA). The BCA protein assay kit (Pierce) was used to determine protein concentrations. Samples containing 30 μg proteins were separated on a 10% SDS-PAGE and electrotransferred onto nitrocellulose membranes. The membrane was blocked for 2 h in TBST containing 5% skim milk and incubated with primary antibodies against IL-9 receptor (IL-9R, 1:500 dilution, Abcam), coxsackie and adenovirus receptor (CAR, 1:500 dilution, Santa Cruz), phosphorylated Erk1/2 (1:500 dilution, cell signaling technology), total Erk1/2 rabbit polyclonal antibody (1:1000 dilution, cell signaling technology), and beta-actin (1:1000 dilution, cell signaling technology) at 4°C over night. After washing, the membranes were incubated with HRP-conjugated secondary antibodies (1:3000) at 37°C for 2 h. The target bands were finally developed with super ECL reagent (ThermoScientific, USA), captured by Image Lab, and semi-quantitatively analyzed with densitometric methods.

Total RNA of heart tissue or cardiomyocyte was extracted with TRIzol reagent (Takara Biotechnology) following the manufacturer’s protocol and the PrimeScript RT reagent kit was used to reverse transcribe the RNA into DNA (Takara Biotechnology). The primers for CAR, Bax, Bcl-2, and GAPDH are listed: CAR (Sense: GCACCCGCTAAGGTAGCTG, Antis: ATAGACCCGTCCTTGCTCTGT), Bax (Sense: TGCAGAGGATGATTGCTGAC, Antis: GATCAGCTCGGGCACTTTAG), Bcl-2 (Sense: GTACCTGAACCGGCATCTG, Antis: GCTGAGCAGGGTC TTCAGAG), and GAPDH (Sense: CACGGCAAATTCAACGGC, Antis: TGATGA CCCTTTTGGCTCCA). After an initial denaturation step at 94°C for 3 min, a three-step cycle procedure (denaturation: 94°C, 30 s; annealing: 58°C, 30 s; and extension: 72°C, 30 s) was carried out for 40 cycles. The mRNA levels of target genes were quantified using SYBR Green Master Mix (Takara Biotechnology) with CFX connect real-time system (Biorad, USA). The relative level of gene expression was normalized to the level of GAPDH transcripts.

Data are presented as means ± SEM. Statistical analysis was performed by one-way ANOVA using SPSS 11.0, and P < 0.05 was considered statistically significant.

IL-9 protein expression in myocardium was enhanced on days 5 and 7 in WT and rIL-9 groups, compared with that of the control group (Figure 1A). Further, IL-9-secreting leukocyte (CD45⁺IL-9⁺) levels were significantly increased in WT and rIL-9 groups on days 5 and 7 compared with those in the control and IL-9 KO group (Figures 1B–D, all P < 0.01). These cells were higher in WT and rIL-9 groups on day 5 than on day 7. In addition, IL-9 protein and IL-9-secreting leukocytes in myocardium were almost not expressed in the IL-9 KO mice (Figures 1A–D). Thus, IL-9 expression was increased in myocardium of VMC mice.

To identify the leukocytes contributing to cardiac IL-9 secretion in VMC, we stained the cells with various surface markers. Most of the IL-9-secreting leukocytes were CD8-positive on day 5, and the majority of these cells were CD4-positive on day 7 (Figure 1E).

The HW/BW (the ratios of heart weight to body weight), the pathological scores of heart sections and cTNT levels in WT, IL-9KO, and rIL-9 groups were elevated significantly compared with those in the control group (all P < 0.05). However, these three indices for evaluation of VMC severity were enhanced in the IL-9KO group compared with those in the WT group (all P < 0.05). The severity of VMC was suppressed in rIL-9 group compared with WT group (all P < 0.05; Figure 2). From this, we found that IL-9 could attenuate the severity of VMC.

On days 5 and 7, the levels of cardiac CVB3 titers and CAR expressions in WT, IL-9KO, and rIL-9 groups were increased compared with those in the control group (all P < 0.05). The viral titers and CAR expressions in IL-9KO group were higher than those in WT group (P < 0.05). However, they were lower in rIL-9 group (P < 0.05). The control group was tested negative for cardiac CVB3 (Figure 3). These data proved that IL-9 inhibited cardiac viral replication and CAR expression in VMC mice.

Except for IL-10, the levels of serum IL-9, IL-17a, TGF-β, TNF-α, IFN-γ, IFN-α, and IFN-β in WT, IL-9KO, and rIL-9 groups were higher than in control mice on days 5 and 7 (all P < 0.05). However, the decreased IL-9 level in the IL-9KO group was accompanied by enhanced IL-17a levels and attenuated TGF-β levels compared with those in WT group (all P < 0.01). The opposite changes in IL-9, IL-17a, and TGF-β levels were detected in rIL-9 groups compared with those in WT group (all P < 0.01, Figure 4). This indicated that IL-9 downregulated IL-17a expression and upregulated TGF-β expression in VMC mice.

The direct effects of IL-9 on myocardial cells were investigated in vitro. The neonatal cardiomyocytes were isolated, infected with CVB3, and incubated with IL-9. The CVB3 titers in the CVB3 and IL-9 groups were higher than in the control group (P < 0.01). However, the viral titers in IL-9 group were lower than in the CVB3 group (P < 0.05). CVB3 was not detected in the control group (Figures 5A,B).

To clarify the role of IL-9 in CVB3 replication in the cardiomyocytes, we determined the changes in IL-9R and CAR levels of myocardial cells. We first found that cardiomyocytes express IL-9R. Subsequently, we found that the CAR protein levels in myocardial cells were increased in the CVB3 and IL-9 groups compared with those in the control group (P < 0.01). They were lower in the IL-9 group than in the CVB3 group (P < 0.01, Figure 5C). Thus, IL-9 could directly suppress CVB3 replication and CAR expression in cardiomyocytes.

The autocrine effect of TGF-β, TNF-α, IFN-α, and IFN-β from myocardial cells was detected after neonatal cardiomyocytes were infected with CVB3 and incubated with IL-9. The levels of the four cytokines were higher in the CVB3 and IL-9 groups than in the control group (all P < 0.05). However, only TGF-β levels were higher in the IL-9 group than in the CVB3 group (P < 0.05). No differences in TNF-α, IFN-α, and IFN-β were found between the CVB3 and IL-9 groups (Figure 6A).

The signal molecules associated with TGF-β production in myocardial cells were investigated. The phosphorylated ERK1/2 levels were increased after CVB3 infection (P < 0.01) and were attenuated after IL-9 treatment (P < 0.05; Figure 6B). The other signal molecule RhoA was not altered following CVB3 and IL-9 interventions (Figure 6B).

Then, the TGF-β mAb in combination with IL-9 were added to further clarify the mechanisms of IL-9 on CVB3 replication and CAR expression. The data showed that the CVB3 titers and CAR expression in IL-9 + TGF-β mAb group were higher than in the IL-9 group (P < 0.05; Figure 6C). It further indicated that IL-9 could suppress CVB3 replication and CAR expression by facilitating TGF-β autocrine effect in cardiomyocytes.

To test the effects of IL-9 on cardiomyocyte apoptosis after CVB3 infection in vitro, we carried out TUNEL in myocardial cells. As shown in Figures 7A,B, CVB3 infection remarkably increased the number of TUNEL-positive cardiomyocytes compared with the control groups (P < 0.05). However, no significant differences were found in the number of TUNEL-positive cells between IL-9 and CVB3 groups.

We also examined the concurrent expression of Bcl-2 family using real-time PCR. The Bax/Bcl-2 ratio was significantly increased in cardiomyocytes exposed to CVB3 (P < 0.01). However, no changes in Bax/Bcl-2 ratio were found between IL-9 and CVB3 groups (Figure 7C). Thus, IL-9 did not directly influence cardiomyocyte apoptosis.