Kruppel-like factor 10 protects against acute viral myocarditis by negatively regulating cardiac MCP-1 expression

Abstract

Viral myocarditis (VMC) is a cardiac disease associated with myocardial inflammation and injury induced by virus infection. Cardiomyocytes have recently been regarded as key players in eliciting and modulating inflammation within the myocardium. Kruppel-like factor 10 (KLF10) is a crucial regulator of various pathological processes and plays different roles in a variety of diseases. However, its role in VMC induced by coxsackievirus B3 (CVB3) infection remains unknown. In this study, we report that cardiac KLF10 confers enhanced protection against viral myocarditis. We found that KLF10 expression was downregulated upon CVB3 infection. KLF10 deficiency enhanced cardiac viral replication and aggravated VMC progress. Bone marrow chimera experiments indicated that KLF10 expression in nonhematopoietic cells was involved in the pathogenesis of VMC. We further identified MCP-1 as a novel target of KLF10 in cardiomyocytes, and KLF10 cooperated with histone deacetylase 1 (HDAC1) to negatively regulate MCP-1 expression by binding its promoter, leading to activation of MCP-1 transcription and recruitment of Ly6C^high monocytes/macrophages into the myocardium. This novel mechanism of MCP-1 regulation by KLF10 might provide new insights into the pathogenesis of VMC and a potential therapeutic target for VMC.

Introduction

Viral myocarditis (VMC) is a cardiac disease associated with inflammation and injury of the myocardium caused by infections and autoimmune diseases. It has been identified as a major cause of sudden death in young adults. In the early stage, VMC usually shows mild symptoms. However, along with VMC progression, patients develop dilated cardiomyopathy (DCM) and even heart failure.^1,2 Type I interferon (IFN) has a certain therapeutic effect in acute VMC infection. However, neither VMC vaccines nor specific therapies have been available until now.^2,3 More than 20 viruses, including adenoviruses, influenza viruses, and cytomegaloviruses, have been identified as inducing myocarditis in both humans and mice. Enteroviral species, such as coxsackievirus B3 (CVB3) and B4 (CVB4), are the dominant viruses found in VMC patients. The features of CVB3 inoculation in mice resemble those of inflammatory cardiomyopathy in humans.⁴ Thus, CVB3-induced myocarditis has been extensively studied. The related mouse model is well established and widely used in clinical and basic studies.⁵ In addition to cardiomyocyte damage upon early viral infection, immune cell infiltration into the myocardium has been recognized as the most relevant pathological stress factor contributing to VMC progression.^6,7

Although it has been clarified that various immune cells, as well as proinflammatory cytokines produced by these cells, are involved in VMC progression, cardiac-resident cells were recently identified as key players in the initiation and regulation of the inflammatory response.^8,9 Recent studies have suggested that cardiac-resident cells trigger immune responses at the initial phase of myocardial injury. In the early phase of ischemic heart disease, cardiac fibroblasts are the dominant inflammatory producers, while endothelial cells release inflammatory factors after sensing a pressure overload.^10–12 In the case of VMC, cardiomyocyte signals are mainly required for recruiting leucocytes to the infected myocardium. Previously, we explored the expression pattern of chemokines in CVB3-induced VMC and found that the chemokines were immediately secreted by the infected cardiomyocytes and that elevated monocyte chemoattractant protein-1 (MCP-1) is the key chemokine that stimulates the migration of mononuclear cells.¹³ There is strong evidence that MCP-1 plays a major role in several cardiac diseases, including myocarditis and ischemic injury, and in cardiac repair.¹⁴ Disruption of the MCP-1/CCR2 pathway by antibody depletion or gene therapy effectively alleviated VMC progression.^15,16 In summary, in the initial phase of CVB3-induced VMC, cardiomyocytes primarily contribute to establishing a proinflammatory environment by releasing massive amounts of cytokines and chemokines, which recruit immune cells into the infected myocardium. Therefore, we sought to deeply explore the mechanism underlying the regulation of inflammation in cardiomyocytes.

Kruppel-like factors (KLFs) are a family of transcriptional factors containing highly conserved C2H2 zinc finger motifs capable of binding to GC-rich sites. Although KLFs have significantly similar sequences, the members play different roles in biological processes, and some of them have been reported to be involved in the development or activation of immune cells.^17,18 KLF4 is one of the best-known KLFs because of its key regulatory roles in monocyte differentiation and M2 macrophage polarization.¹⁹ Another family member, KLF10, was first identified in TGF-β-induced normal human fetal osteoblasts, and it regulates osteogenic differentiation by enhancing the TGF-β signaling pathway. Thus, this member is also called transforming growth factor-β (TGF-β)-inducible early gene-1 (TIEG1).^20,21 Subsequent studies further demonstrated its functions in various cellular processes, including cell proliferation, differentiation, and inflammation.¹⁸ Previously, we showed that KLF10 repressed IL-12p40 transcription in M-CSF-induced bone marrow-derived macrophages (M-BMMs) but not GM-CSF-induced BMMs.²²

Several studies have explored the functions of KLF10 in cardiovascular diseases and the underlying mechanisms. In early 2007, the Spelsberg group generated KLF10 knockout mice and found that 4- to 16-month-old male knockout mice developed typical features of cardiac hypertrophy and identified pituitary tumor-transforming gene-1 (PTTG-1) as the target gene, but this phenotype was not observed in female mice.²³ Nevertheless, Cen et al. demonstrated that KLF10 deficiency significantly protected against myocardial infarction in 6- to 8-week-old mice by modulating the Pten/Akt signaling pathway.²⁴ Therefore, the functions of KLF10 in the heart tend to be disease specific and should be more fully elucidated. In this study, we report that cardiac KLF10 confers enhanced protection in viral myocarditis via transcriptionally repressing MCP-1 to decrease Ly6C^high monocyte/macrophage (Mo/Mφ) infiltration into the myocardium.

Materials and methods

Mice and virus

C57BL/6 mice were purchased from Shanghai Slac Animal Inc. (Shanghai, China). KLF10-deficient mice were originally from the laboratory of Dr. Thomas Spelsberg (Mayo Clinic, MN, USA). CVB3 (Nancy strain) was a kind gift from Professor Yingzhen Yang (Key Laboratory of Viral Heart Diseases, Zhongshan Hospital, Shanghai Medical College of Fudan University). The viral titer was determined by a TCID₅₀ assay on HeLa cell monolayers with standard methodology as previously reported.²⁵ The animal experiments were performed in accordance with the Soochow University institutional guidelines, and the study was approved by the Ethics Committee of Soochow University in written form. Euthanasia of mice was performed by carbon dioxide inhalation with minimal fear, anxiety and pain.

Murine VMC

Six-week-old male KLF10-deficient or WT mice were intraperitoneally infected with 100 μL of PBS containing a 1500 TCID₅₀ dose of CVB3 to establish an acute viral myocarditis model. Mouse survival and body weight were recorded until the termination of the experiment. Heart tissue was collected for measurement of the viral titer and for flow cytometry, histological analysis, and echocardiology. For the in vivo MCP-1 blocking VMC model, 200 μg (10 mg/kg) of anti-MCP-1 (Clone 2H5, Bioxcell) and the isotype control Abs in 100 μL of PBS were intravenously injected 1 day (day −1) and 2 days (day 2) post CVB3 infection. Individual experiments were performed at least three times with 7–10 mice per group. The levels of creatine kinase (CK) and creatine kinase isoenzyme (CKMB) in serum were detected using a fully automatic biochemical analyzer following the manufacturer’s instructions (Thermo, USA).

Cell culture

Briefly, ventricular myocardial tissue from collected from mice within 24 h of birth were minced in a nominally Ca²⁺- and Mg²⁺-free Hanks’ balanced solution. Cardiomyocytes were dispersed by 0.625 mg/ml collagenase (type II) at 37 °C for 40 min. The isolated cells were preplated for 90 min to remove noncardiomyocytes. The cardiomyocytes were plated in M199 medium containing 10% fetal calf serum in 35-mm Petri dishes precoated with 1% gelatin. Cells were incubated at 37 °C in the presence of 5% CO₂. The HeLa and HEK293 cell lines were cultured in Dulbecco’s modified Eagle’s medium supplemented with 2 mmol/L glutamine, 100 units/ml penicillin and 100 μg/ml streptomycin sulfate and 10% heat-inactivated fetal bovine serum (FBS) (GIBCO) at 37 °C in the presence of 5% CO₂.

RNA isolation and Q-PCR

Total RNA was extracted by RNAiso reagent (Takara, Cat. No. 9109). Reverse transcription (Takara, Cat. No. DRR063A) and Q-PCR (Takara, Cat. No. DRR041A) were performed according to the manufacturer’s instructions. Q-PCR was performed on a 7500 Real-Time PCR System (Applied Biosystems, Carlsbad, CA). Gene expression was normalized to the expression level of GAPDH. For detecting viral RNA, one part of the extracted RNA was reverse transcribed with oligo-dT primer for GAPDH detection, while another part was reverse transcribed with a primer specific to the CVB3 positive RNA strand. Viral cDNA expression was normalized to the expression level of GAPDH. The reverse transcription primer for the CVB3 positive strand was 5′-CACCGGATGGCCAATCCA-3′.

Western blot analysis

Cells were lysed in complete lysis-M buffer (Roche, Switzerland), and heart tissue was ground by liquid nitrogen, lysed in complete lysis-M buffer and then heated for 10 min at 90 °C. The concentration was determined by the bicinchoninic acid (BCA) protein assay (Thermo, Cat. No. 155209). Equal amounts of protein were subjected to SDS-PAGE, transferred to nitrocellulose membranes, and hybrid blotted. Anti-KLF10 antibody was purchased from Abcam (1:2,000, Cat. No. ab73537), and anti-VP1 antibody was purchased from DAKO (1:2,000, Denmark, Clone 5-D8/1). Peroxidase-coupled secondary antibodies against rabbit or mouse immunoglobulin (Pierce) were diluted 1:10,000. Detection of the peroxidase-coupled antibodies was performed with Super Signal West Pico Chemiluminescent Substrate (Pierce).

IP and Immunoblotting

HEK293 cells were cotransfected with the indicated expression plasmids. After 24 h, cell lysates were collected using radioimmunoprecipitation assay (RIPA) lysis buffer with protease inhibitors (1 mM phenylmethylsulfonyl fluoride, Roche complete protease inhibitor), followed by immunoprecipitation with anti-Flag or anti-KLF10 beads (Santa Cruz, sc-2003). Proteins were eluted from the beads after washing six times with PBS. The protein bound to the beads was subjected to Western blot analysis with anti-KLF10 or anti-Flag (1:1,000, CST 8146S).

Preparation of mouse cardiac mononuclear cells

Fresh hearts were rinsed and minced into small pieces in sterile petri dishes containing 10 ml of digestion buffer containing 800 μg/ml type II collagenase (Sigma) and 5 μg/ml hyaluronidase (Roche) in RPMI 1640 medium supplemented with 10% fetal calf serum (FCS). Tissues were subjected to two digestions at 37°C while stirring for 1 hour. Then, the minced hearts were passed through 70-μm mesh strainers (BD Falcon). Individual cell suspensions from three mice were pooled and centrifuged. Cell pellets were resuspended in 40% Percoll (GE Healthcare). Single-cell leukocytes were isolated by 40–70% Percoll followed by centrifugation for 30 min at 800×g and resuspended in RPMI 1640 medium with 10% FCS.

Flow cytometry

All antibodies were purchased from BD Pharmingen (San Diego, CA): CD45.2-APC/Cy7 (clone 104), CD11b-FITC (clone M1/70), Ly6G-PerCP/Cy5.5 (clone1A8), Ly6C-PE/Cy7 (clone AL-21), CD3-PerCP (clone 145-2C11), CD4-FITC (clone RM4-5), and CD8-PE (clone 53–6.7). The isotype control Abs and unstained samples were included in the analysis. Cells were incubated in FACS buffer (PBS, 2% FCS, 2 mM EDTA) containing an anti-mouse Fc receptor blocking reagent (Miltenyi) for 20 min at 4 °C. Then, the cells were harvested and incubated with the relevant antibodies for 30 min on ice, washed, and analyzed on a Canto II flow cytometer (BD Bioscience). The data were analyzed using FlowJo v10.0 software (Tree Star). Reported numbers were normalized to the total weight of the hearts to obtain the number of corresponding cell fractions per mg of tissue.

TUNEL Staining

Primary cardiomyocytes were seeded and cultured in eight-well chamber slides and infected with CVB3 for 24 h. Then, the cells were washed with PBS, fixed with 4% paraformaldehyde for 30 min on ice, and permeabilized with PBST at room temperature. Cardiomyocytes were stained by the TUNEL agent using the TMR (red) in situ Apoptosis Detection Kit (catalog number #12156792910, Roche Diagnostic, Mannheim, Germany). Cells were counterstained with DAPI to detect the nucleus and examined by fluorescence microscopy. Five photomicrographs were randomly captured at each time point under high-power fields (200× magnification), and one representative image is shown.

ELISA of cytokines

Heart tissue was ground in liquid nitrogen, lysed in phosphate-buffered saline (PBS) with PMSF protease inhibitor (Roche, Switzerland) and centrifuged at 1800×g for 20 min at 4 °C. The supernatants were extracted and immediately frozen at −80 °C and centrifuged at 3000×g for 20 min to remove cellular debris before use. Blood serum was obtained by centrifugation and stored at −80 °C. Cytokines were measured by enzyme-linked immunosorbent assay (ELISA) kits according to the manufacturer’s instructions (eBioscience, San Diego, USA).

Mouse echocardiology

Murine transthoracic echocardiography was performed using a Vevo 2100 high-resolution microultrasound system (Visualsonics, Toronto, Canada). Two-dimensional guided M-mode echoes were obtained from short- and long-axis views at the level of the largest left ventricle (LV). LV wall dimensions at the end of the systole and diastole were measured from the M-mode image. LV fractional shortening (FS) and ejection fraction (EF) were calculated from the measured ventricle dimensions.

Histopathology

Histopathology was performed to assess inflammatory infiltration and myocardial lesions. Mice were evaluated for the development of acute viral myocarditis after infection. Briefly, the apical parts of the hearts were fixed in 10% phosphate-buffered formalin, embedded in paraffin wax, sectioned at 5 μm and stained with hematoxylin-eosin (H&E). Stained sections were used for image analysis with a Nikon Eclipse TE2000-S microscope, and five images were randomly captured under high-power fields (100 or 200× magnification). Myocarditis was graded in a blinded manner by two independent investigators based on the following semiquantitative scale: 1 = small foci of inflammatory cells between myocytes or inflammatory cells surrounding individual myocytes; 2 = 5–10% of a myocardial cross-section involved; 3 = 10–20% of a myocardial cross-section involved.

IHC

Immunohistochemistry was performed to assess the cardiac levels of CD11b or cleaved caspase-3 according to the manufacturer’s instructions for the Mouse and Rabbit Specific HRP/DAB Detection IHC Kit (ab64264, Abcam). Briefly, paraffin sections of hearts were incubated with primary antibodies against CD11b (1:250, ab133357, Abcam) or cleaved caspase-3 (1:250, Cat. No. 9664, Cell Signaling Technology) at 4 °C overnight. After washing, sections were stained with biotinylated goat anti-polyvalent antibodies for 10 min at room temperature. After washing, the sections were incubated with streptavidin peroxidase and DAB substrate. Stained sections were imaged with a Nikon Eclipse TE2000-S microscope, and five images were randomly captured under high-power fields (×100 or ×200 magnification).

Luciferase reporter assay

The MCP-1 promoter luciferase reporter vector containing the full-length gene was generated by amplifying the corresponding fragments into the XhoI and Hind III sites of the PGL3 enhancer vector (Promega, Madison, WI). The primer sequences were as follows: 5′-CTCGAGCCTTGGCCACTTCCTTTTATTTC-3′ and 5′-AAGCTTGGTGGTGGAGGAAGAGAGAG-3′. The mutant sequence was synthesized by GENEWIZ Biotech Company (Suzhou, China). HEK293 cells were cotransfected with 100 ng of luciferase reporter plasmid, 10 ng of a thymidine kinase promoter Renilla luciferase reporter plasmid, and a KLF10 overexpression or control plasmid. Forty-eight hours later, the luciferase activities were determined by the Dual-Luciferase Reporter Assay System (Promega, Cat. No. E10910) according to the manufacturer’s instructions.

ChIP assay

Chromatin immunoprecipitation assays were performed essentially according to the manufacturer’s protocol for MAGnify^TM CHIP (Thermo Fisher Scientific). Briefly, antibodies against KLF10 or IgG were coupled to Protein A/G Dynabeads, and cardiomyocytes (3 × 10⁶) were fixed for 10 min using 1% formaldehyde. Cells were then collected and lysed in 50 μL of lysis buffer. Equal amounts of lysates were used for immunoprecipitation of sheared chromatin with Dynabead-coupled anti-KLF10 and IgG antibodies (Santa Cruz). Chromatin DNA was extracted with a DNA purification column according to the manufacturer’s instructions (Qiagen) and was then analyzed by semiquantitative PCR and Q-PCR. Data were normalized to the level of IgG in each sample.

Recombinant AAV and infection of mice

The control virus AAV9-GFP and KLF10-overexpressing virus AAV9-KLF10 were designed and ordered from Shanghai Genechem Co., Ltd., and the titers were ~2 × 10¹² viral genomes per ml (vg/ml). In the AAV9-overexpressing VMC model, each mouse from the WT or KLF10-deficient group received a single tail vein injection of 3 × 10¹⁰ genome copies of AAV-GFP control or AAV9-KLF10. After 4 weeks, mice were intraperitoneally injected with CVB3 and subjected to evaluation of the progression of VMC.

Generation of BM chimera mice

Six-week-old male WT or KLF10-deficient recipient mice were irradiated with two doses of 600 rad within 4 h. Twelve hours later, they were administered 5 × 10⁶ WT or KLF10-deficient donor BM cells (i.v.). Animals were allowed to reconstitute for 8 week before CVB3 infection.

Statistical analysis

Data are presented as the mean ± SEM, and statistical analysis was performed by GraphPad Prism 5.0 (Graph Pad Prism Software Inc., San Diego, CA, USA) software. For two-group comparisons, statistical significance was determined by an unpaired two-tailed Student’s t test. For multiple groups, statistical significance was determined using ANOVA followed by Bonferroni post hoc analysis. Survival curves were estimated by the Kaplan–Meier procedure with the log-rank test to compare survival among groups. P values < 0.05 were considered to be statistically significant and are indicated as follows: *0.05 ≥ P > 0.01; **0.01 ≥ P > 0.001; ***P ≤ 0.001.

Results

KLF10 expression is significantly inhibited in the hearts of VMC mice

To investigate whether KLF10 was involved in acute myocarditis, we first established a CVB3-induced acute VMC model using C57BL/6 mice. Infected mice showed massive inflammatory infiltration in hearts at day 7 p.i. The viral titers in the myocardium increased along with infection, peaked at day 3 p.i., and then declined at day 7 p.i. (Fig. 1a, b). Next, we evaluated KLF10 expression by Q-PCR and Western blot analysis. As shown in Fig. 1c, KLF10 mRNA levels in heart tissues were significantly decreased and sustained at low levels, and KLF10 protein expression was also decreased after CVB3 infection (Fig. 1d). CVB3 infection also caused acute pancreatic injury (Fig. S1B). Pancreases from CVB3-infected mice showed decreased KLF10 expression (Fig. S1A), similar to heart tissue. Thus, these results indicated that CVB3 infection induced cardiac and pancreatic KLF10 downregulation.

Fig. 1.

KLF10 is downregulated in the hearts of mice with CVB3-induced VMC. Six-week-old male C57BL/6 mice were intraperitoneally injected with a 1500 TCID₅₀ dose of CVB3, and heart tissue was collected at the indicated time points. a At days 0, 3, and 7 p.i., hearts were harvested and sectioned for HE staining to analyze inflammation (indicated by black arrows). b At days 0, 3, and 7 p.i., heart tissue was collected and subjected to the TCID₅₀ assay for detection of the viral titer. c Heart tissue was collected at the indicated time points for RNA extraction, and KLF10 mRNA levels were analyzed by Q-PCR. d Protein samples from heart tissue were prepared, and the expression of KLF10 was analyzed by western blotting. Data were normalized to GAPDH expression and presented as the mean ± SEM of three representative independent experiments, with n = 6 per group. Data were analyzed by an unpaired two-tailed Student’s t test

KLF10 deficiency aggravates CVB3-induced myocardial injury

Then, we established a VMC model in KLF10-deficient and WT mice. As shown in Fig. 2a, b, the survival rate and body weight decreased significantly in KLF10-deficient mice. At day 7 p.i., WT mice lost 18.2% of their body weight, while more severe weight loss (approximately 28.7%) was observed in KLF10-deficient mice (Fig. 2a). In total, 60% of KLF10-deficient mice rapidly succumbed to infection, while only 20% of WT mice died within 7 days (Fig. 2b). In addition, KLF10 deficiency significantly increased the activities of serum creatine kinase (CK) and CK-MB (Fig. 2c). Echocardiographic measurements showed that the left ventricular ejection fraction (LVEF) and left ventricular fractional shortening (LVFS), which reflected cardiac function, were markedly decreased in KLF10-deficient mice (Fig. 2d). Then, we detected cardiac viral copies in CVB3-infected mice, and higher viral copy numbers were observed in KLF10-deficient mice (Fig. 2e). Immunohistochemical analysis showed that KLF10 deficiency enhanced the expression of cleaved caspase 3 in cardiomyocytes (Fig. 2f). In addition to impaired cardiac function, KLF10-deficient mice developed more severe pancreatic destruction than WT mice (Fig. S1B). Furthermore, we evaluated the role of KLF10 in the chronic phase and the subsequent cardiac fibrosis. KLF10 deficiency also decreased CVB3-induced myocardial injury at the late phase of infection (Fig. S2 A–C). Six weeks later, infected KLF10−/− mice exhibited much more severe cardiac fibrosis than WT mice (Fig. S2D). All the results above indicated that KLF10 deficiency decreased CVB3-induced myocardial injury.

Fig. 2.

KLF10-deficient mice aggravate CVB3-induced myocardial injury. Six-week-old male WT and KLF10−/− mice were infected (i.p.) with a 1500 TCID₅₀ dose of CVB3. a, b The body weight and survival rate were monitored daily until day 7 p.i. c Serum CK activity and CK-MB activity as indicators of cardiac injury were detected at days 0 and 4 post infection. d Transthoracic echocardiography was performed on each mouse at day 7 p.i. EF and FS were calculated from measured ventricle dimensions. e At day 3 p.i., the RNA copy level of CVB3 in hearts was analyzed by Q-PCR. f At day 7 p.i., hearts were sectioned and subjected to the IHC assay with a primary antibody against cleaved caspase 3. Five images were captured for each section, and a representative image is shown. Data are presented as the mean ± SEM of three representative independent experiments, with n = 6–8 per group. Data in d were analyzed by the unpaired two-tailed Student’s t test. Data in c, e were analyzed using one-way ANOVA followed by Bonferroni post hoc analysis

Enormous Ly6C^highinflammatory monocytes infiltrate the myocardium of KLF10-deficient mice with VMC

In addition to the early myocyte damage directly induced by viruses, host immune and inflammatory responses synergistically contribute to CVB3-induced myocarditis. As shown in Fig. 3a, KLF10-deficient mice developed severe myocarditis with diffuse inflammatory infiltration. We further analyzed the profile of infiltrating immune cells in the hearts of KLF10-/- and WT VMC mice. Immunohistochemical staining showed much higher CD11b expression in KLF10−/− mice than in WT controls (Fig. 3b). Flow cytometry data further confirmed this hypothesis. After CVB3 infection, KLF10−/− mice showed elevated CD45+ cell levels, which indicated high immune cell infiltration and inflammation (Fig. 3c). Then, we analyzed immune cell populations within CD45+ cells. More CD11b+ cells were found in KLF10-deficient mice (Fig. 3c), while no significant difference was observed in CD4⁺ or CD8⁺ T cells (Fig. S3). Thus, KLF10 deficiency mainly led to profound changes in the infiltrating myeloid cells. The proportion of Ly6G⁺CD11b⁺ neutrophils was comparable between KLF10−/− and WT mice (Fig. 3d). However, KLF10−/− mice had significantly more Ly6C^high populations and fewer Ly6C^low populations within the Ly6G^-CD11b⁺ Mo/Mφ compartment (Fig. 3e). Interestingly, this shift in Mo/Mφ populations was restricted to only the heart, as the levels of Ly6C^high and Ly6C^low monocytes in the blood were comparable between KLF10−/− and WT mice (Fig. S4). Thus, these data suggest that VMC aggravation in KLF10 knockout mice is due to the infiltration of massive CD11b⁺ myeloid cells in the myocardium, principally the Ly6C^high Mo/Mφ subsets. Consistent with this, the levels of cardiac proinflammatory cytokines, including IL-1β, TNF-α, IL-6, and IFN-γ, were significantly elevated in KLF10-deficient mice (Fig. 3f). These results demonstrate that KLF10 deficiency significantly aggravates the progression of viral myocarditis.

Fig. 3.

KLF10 deficiency deteriorates virus-induced myocarditis and alters the composition of heart-infiltrating cells. Six-week-old male WT and KLF10-/- mice were infected (i.p.) with a 1500 TCID₅₀ dose of CVB3. a At day 7 p.i., heart sections were subjected to HE staining for the evaluation of viral myocarditis. Five images were captured for each section, and a representative image is shown. Pathological scores of the hearts are shown. b At day 7 p.i., hearts were sectioned and subjected to the IHC assay with a primary antibody against CD11b. Five images were captured for each section, and a representative image is shown. c At day 5 p.i., infiltrating inflammatory cells in heart tissues were isolated and analyzed by flow cytometry. Representative flow diagrams showing immune cell populations by staining anti-CD45, CD11b, Ly6C, and Ly6G. d Mean values ± SEM for absolute numbers of CD45⁺, CD45⁺CD11b⁺, CD45⁺CD11b⁺Ly6G⁺ neutrophils, and CD45⁺CD11b⁺Ly6G⁻Ly6C⁺ Mo/Mφ are shown. e The ratio of Ly6C^high to Ly6C^low within CD45⁺CD11b⁺Ly6G⁻ cells was calculated from individual mice and is shown as the mean  ± SEM. f Levels of cytokines, including IL-1β, TNF-α, IL-6, IFN-γ, and IL-4, from heart homogenates of mice were detected by ELISA at day 5 p.i. Data are presented as the mean ± SEM of three representative independent experiments, with n = 6–8 per group. The data in a, e, and f were analyzed by an unpaired two-tailed Student’s t test. The data in d were analyzed using one-way ANOVA followed by Bonferroni post hoc analysis

KLF10 expression in nonhematopoietic cells is required for the progression of VMC

KLF10 is expressed in both infiltrating hematopoietic cells, such as macrophages and T cells, and cardiac-resident cells, such as cardiomyocytes and cardiac fibroblasts. To determine whether the aggravated VMC caused by KLF10 deficiency was correlated with its functions in infiltrating hematopoietic cells, we generated bone marrow chimeras. As shown in Fig. 4a, bone marrow from KLF10−/− or WT mice was transferred into lethally irradiated WT or KLF10−/− mice, respectively, and syngeneic transfer was also performed as a control to exclude the effects of bone marrow reconstruction itself. Strikingly, KLF10−/− recipient mice that lacked KLF10 in nonhematopoietic cells, regardless of the genotype of the bone marrow donors, were similarly susceptible to CVB3-induced myocarditis, showing much higher weight loss and viral copy numbers than WT recipient mice (Fig. 4b, c). Histopathological analyses revealed that chimeras in which KLF10 expression was deficient in nonhematopoietic cells developed severe inflammatory infiltration in the myocardium (Fig. 4d). Furthermore, the proportions of Ly6C^high Mo/Mφ and inflammatory cytokines were much higher in KLF10-/- recipient mice than in WT recipient mice (Fig. 4e, f). To summarize, these results suggest that KLF10 expression in nonhematopoietic cells contributes to the development of VMC.

Fig. 4.

KLF10 expression in nonhematopoietic cells is required for the progression of VMC. a Schematic of the generation of BM chimeras. Bone marrows were transferred from WT or KLF10−/− donor mice into lethally irradiated 6-week-old male KLF10−/− or WT recipient mice. After 8 weeks, chimeric mice were infected (i.p.) with a 1500 TCID₅₀ dose of CVB3. b The body weight of the four groups was monitored daily until day 7 p.i. c At day 3 p.i., the RNA copy level of CVB3 in hearts was analyzed by Q-PCR. d At day 7 p.i., heart sections were subjected to HE staining for evaluation of viral myocarditis. Five images were captured for each section, and a representative image is shown. Pathological scores of the hearts of chimeric mice are shown. e Representative flow cytometry plots of cardiac mononuclear cells stained with CD45, CD11b, and Ly6C are shown. The ratio of Ly6C^high to Ly6C^low within CD45⁺CD11b⁺ cells was calculated from individual mice and is shown as the mean ± SEM. f Levels of proinflammatory cytokines, including IL-1β, TNF-α and IL-6, from heart homogenates of mice were detected by ELISA at day 7 p.i. The data are presented as the mean ± SEM of three representative independent experiments, with n = 6–8 per group. The data were analyzed by unpaired two-tailed Student’s t test

AAV9-mediated KLF10 gene overexpression ameliorates VMC

It has been reported that recombinant adeno-associated virus (AAV) has been used to efficiently and stably deliver gene expression in mammalian tissues, which do not encode viral proteins and cannot induce immune responses. AAV9 is the most highly cardiotropic serotype in mice and can be used for the delivery of cardiomyocyte genes in vivo. Therefore, we constructed an AAV9 vector expressing full-length mouse KLF10 fused with GFP to establish a KLF10-overexpressing mouse model. First, we examined the delivery efficiency by measuring GFP expression. Ninety percent of the heart tissue samples were GFP positive at week 4 post the injection of the AAV9-ctrl virus (3 × 10¹⁰ vg per mouse) via the tail vein (Fig. S5A). To exclude the possibility that AAV infection affected subsequent CVB3 infection, cardiac CVB3 replication was evaluated in negative-ctrl and AAV9-ctrl mice intraperitoneally infected with CVB3. Q-PCR analysis showed that there was no difference in viral replication between the two groups (Fig. S5B), indicating that AAV infection had an undetectable effect on CVB3 infection. Then, WT or KLF10-/- mice were injected with the AAV9-ctrl or AAV9-KLF10 virus, followed by CVB3 infection after 4 weeks. KLF10 was successfully overexpressed in heart tissue (Fig. S5C), and VMC progression was monitored. As shown in Fig. 5a, AAV9-mediated KLF10 overexpression markedly improved weight loss induced by CVB3 infection in both WT and KLF10−/− mice. Viral replication and the activities of serum CK were significantly decreased in AAV9-KLF10-transduced WT or KLF10−/− mice (Fig. 5b, c). Consistent with this, histological analysis showed significantly decreased myocarditis with much fewer foci of inflammation in AAV9-KLF10-infected WT and KLF10−/− mice than in AAV9-ctrl-infected mice (Fig. 5d). These data demonstrated that AAV9-mediated KLF10 cardiac overexpression could efficiently alleviate virus-induced myocarditis, which further suggested that the expression of KLF10 in nonhematopoietic cells might be responsible for VMC progression.

Fig. 5.

AAV9-mediated KLF10 gene overexpression ameliorates VMC. Each mouse from the 6-week-old male WT or KLF10-deficient group received a single tail vein injection of 3 × 10¹⁰ genome copies of AAV-GFP control or AAV-KLF10. After 4 weeks, mice were infected (i.p.) with a 1500 TCID₅₀ dose of CVB3. a The body weight change of the four groups was monitored daily. b At day 3 p.i., the RNA copy level of CVB3 in hearts was analyzed by Q-PCR. c Serum CK activity as an indicator of cardiac injury was detected at day 4 post infection. d At day 10 p.i., heart sections were subjected to HE staining for the evaluation of viral myocarditis. Five images were captured for each section, and a representative image is shown. The data are presented as the mean ± SEM of three representative independent experiments, with n = 6–8 per group. The data in b, c were analyzed using one-way ANOVA followed by Bonferroni post hoc analysis

KLF10 expression is downregulated upon CVB3 infection but not directly involved in CVB3 replication and virus-induced cell death

Cardiomyocytes (CMs) and cardiac fibroblasts (CFs) are the main responsive nonhematopoietic cells during VMC. We isolated primary CMs from neonate mice, and more than 90% of the cells were successfully stained with anti-cardiac troponin T antibody and exhibited spontaneously beating (unpublished data). Then, we assessed KLF10 expression in vitro using isolated CMs. As shown in Fig. 6a, both mRNA and protein levels were significantly downregulated upon CVB3 infection, which was in accordance with observations in the hearts of VMC mice (Fig. 1c, d). We also isolated primary CFs from neonate mice and confirmed their purity by detecting α smooth muscle actin (α-SMA). Unlike in CMs, KLF10 expression was not altered in CFs after CVB3 infection (Fig. 6b), indicating that KLF10 expression in CMs could contribute to VMC progression. Several studies, including our previous research, identified KLF10 as an NF-κB-targeted gene in macrophages. Mitra et al. reported that CVB3-induced NF-κB activation is independent of the phosphorylation of the upstream IκB-α protein but through a PI3K/Akt-dependent pathway.^5,13 To explore the underlying signaling pathway that regulates inflammation, we pretreated CMs with the IκB-α inhibitor BAY11-7082 or the PI3K inhibitor LY294002 and then challenged them with CVB3. The results showed that the upregulation of the inflammatory cytokine IL-6 was abolished following PI3K inhibitor treatment. However, the decreased KLF10 expression after CVB3 infection could be partly rescued by LY294002 but was not altered by BAY11-7082 (Fig. 6c, d and Fig. S6). These results indicate that KLF10 may participate in regulating inflammation in CMs through the PI3K pathway during CVB3 infection.

Fig. 6.

KLF10 expression is downregulated upon CVB3 infection but is not directly involved in CVB3 replication and virus-induced cell death. Cardiomyocytes and myocardial fibroblasts were isolated from neonate mice and treated with CVB3 at an MOI of 10 for the indicated time. a, b The expression level of KLF10 was detected by Q-PCR and western blotting. c, d Cardiomyocytes were pretreated with or without 5 μmol/L LY294002 and then infected with CVB3 at an MOI of 10 for the indicated time. The mRNA levels of IL-6 and KLF10 were detected by Q-PCR, and the protein level of KLF10 was subjected to western blotting. e, f RNA copy levels of CVB3 in cardiomyocytes were analyzed by Q-PCR. The protein level of VP1 in cardiomyocytes was detected by Western blotting. g Primary cardiomyocytes were seeded and cultured in eight-well chamber slides and infected with CVB3 for 24 h. Cardiomyocytes were stained by the TUNEL agent using the TMR (red) In Situ Apoptosis Detection Kit. Cells were counterstained with DAPI to detect the nucleus and examined by fluorescence microscopy. Five photomicrographs were randomly captured at each time under high-power fields (200× magnification), and one representative image is shown. Experiments were repeated independently at least two times. The data in a and b were analyzed by an unpaired two-tailed Student’s t test. The data in c, e were analyzed using one-way ANOVA followed by Bonferroni post hoc analysis

Next, to explore whether KLF10 expression in cardiomyocytes was directly involved in CVB3 replication and virus-induced cell death, freshly isolated primary cardiomyocytes from KLF10−/− and WT neonate mice were infected with CVB3 in vitro. Then, viral replication was investigated. As shown in Fig. 6e, f, the levels of viral RNA and capsid protein VP-1 showed no difference between KLF10−/− and WT cardiomyocytes upon CVB3 infection, suggesting that KLF10 did not directly participate in CVB3 replication. We further investigated the possible effect of KLF10 on virus-induced myocyte death and found that KLF10 deficiency did not alter the proportion of CVB3 infection-induced cell death (Fig. 6g). Therefore, these results indicated that KLF10 expression in cardiomyocytes was inhibited by CVB3 infection, but it was not directly involved in the processes of CVB3 replication and virus-induced cardiomyocyte death.

Loss of KLF10 promotes the production of MCP-1 in the heart tissue of mice and in cardiomyocytes upon CVB3 infection

Cardiomyocytes are regarded as an early source of inflammatory cytokines and chemokines, so we detected the expression of 10 cytokines and chemokines in CMs from CVB3-infected KLF10−/− and WT mice. Among them, MCP-1 expression in CMs was the highest at the basal level, and this was also the most obviously upregulated gene upon CVB3 infection (Fig. 7a), which is consistent with a previous report.¹³ Importantly, we found that KLF10-deficient CMs produced more MCP-1 after CVB3 infection, but this phenomenon was not observed in CFs (Fig. 7b, c). Thus, we concluded that loss of KLF10 promotes MCP-1 production in CVB3-infected CMs but not in CFs, suggesting that CMs are the target cells of KLF10 and that MCP-1 might be a novel target of KLF10.

Fig. 7.

KLF10 deficiency promotes the production of MCP-1 in cardiomyocytes and heart tissue upon CVB3 infection. Cardiomyocytes from WT and KLF10−/− neonatal mice were infected by CVB3 at an MOI of 10 for the indicated time. a The gene expression levels of certain cytokines and chemokines were assessed by Q-PCR, and the β-actin mRNA level in these cells was set as 1. b, c Cell supernatant was collected to analyze the protein level of MCP-1 by ELISA. The data are presented as the mean ± SEM of three representative independent experiments. The data in b, c were analyzed using one-way ANOVA followed by Bonferroni post hoc analysis

Blockade of MCP-1 abolishes the deteriorative effect of KLF10 deficiency on VMC

To test the hypothesis that the deteriorative effect of KLF10 deficiency on VMC was dependent on the elevated level of MCP-1 in the myocardium, anti-mouse CCL2 antibody 2H5 or hamster IgG isotype control was injected into KLF10−/− or WT mice 1 day before and 2 days after CVB3 infection (Fig. 8a). Throughout the 7 days postinfection, KLF10−/− mice injected with the isotype Ab lost more body weight than WT mice, while comparable weight loss was observed between KLF10−/− and WT mice injected with anti-MCP-1 Ab, which was obviously better than both isotype Ab-treated groups (Fig. 8b). Consistent with a previous report,¹⁶ MCP-1 blockade in WT mice inhibited CVB3 replication within the myocardium but abolished the differences between KLF10−/− and WT mice (Fig. 8c). The myocarditis on day 7 p.i. in isotype Ab-treated KLF10−/− mice, as assessed by histology, was more severe than that in isotype Ab-treated WT mice. Although anti-MCP-1 Ab administration could significantly alleviate myocardial inflammation, no significant difference was observed in anti-MCP-1 antibody-treated KLF10−/− and WT mice (Fig. 8d). Consistent with Fig. 3, the number of infiltrating CD11b⁺Ly6C^high myeloid cells and the ratio of Ly6C^high to Ly6C^low Mo/Mφ were significantly increased in KLF10−/− mice with the isotype Ab compared to those in WT mice with the isotype Ab. However, blocking MCP-1 signaling eliminated this difference caused by KLF10 deficiency (Fig. 8e). These results illustrate that the deteriorative effect of KLF10 deficiency on VMC progression is dependent on the significant upregulation of MCP-1 levels.

Fig. 8.

Blocking the MCP-1 pathway abolishes the aggravated VMC mediated by KLF10 deficiency. a Schematic of the study strategy. Six-week-old male WT and KLF10−/− mice were treated (i.v.) with isotype or anti-MCP-1 (2H5, 10 mg/kg) Abs on days −1 and 2 and infected (i.p.) with a 1500 TCID50 dose of CVB3 on day 0. b The body weight change of the four groups was monitored daily until day 7 p.i. c At day 3 p.i., the RNA copy level of CVB3 in hearts was analyzed by Q-PCR. d At day 7 p.i., heart sections were subjected to HE staining for the evaluation of viral myocarditis. Five images were captured for each section, and a representative image is shown. The pathological scores of the hearts are shown. e Representative flow cytometry plots of cardiac mononuclear cells stained with CD45, CD11b, and Ly6C are shown. The ratio of Ly6C^high to Ly6C^low within CD45⁺CD11b⁺ cells was calculated from individual mice. The data are presented as the mean ± SEM of three representative independent experiments, with n = 6–8 per group. The data were analyzed by an unpaired two-tailed Student’s t test

KLF10 interacts with HDAC1 to cooperatively inhibit MCP-1 transcription by binding to the MCP-1 promoter

KLF family members act as transcriptional factors. They have been reported to bind to target genes to regulate the transcriptional activities and expression of these genes. To explore whether MCP-1 is the target gene of KLF10 in CMs, we analyzed the MCP-1 promoter sequence and found a CACCC site therein that is highly conserved in mammals (Fig. 9a, b). This is in accordance with our previous report showing that KLF10 targets the IL-12p40 promoter and regulates its transcription in mouse macrophages.²² Next, the chromatin immunoprecipitation (ChIP) assay was performed with the KLF10 antibody in CMs. We designed one pair of primers directed at the CACCC site, and the results of both semiquantitative PCR and Q-PCR showed that binding of the IgG control to the MCP-1 promoter was hardly observed, but the anti-KLF10 antibody pulled down more MCP-1, suggesting that KLF10 may directly bind to the MCP-1 promoter (Fig. 9c). To further investigate the regulatory effect of KLF10 on MCP-1 transcription, we constructed a series of luciferase reporter plasmids encoding the full-length MCP-1 promoter and a mutant with 2 site mutations in the CACCC region. As shown in Fig. 9d, KLF10 significantly inhibited the transcriptional activity of the MCP-1 promoter but did not affect the luciferase activity of the mutant construct. Therefore, these data suggest that KLF10 directly binds to the CACCC site located at the MCP-1 promoter and inhibits the MCP-1 transcriptional activity.

Fig. 9.

KLF10 binds to the MCP-1 promoter and interacts with HDAC1 to cooperatively inhibit MCP-1 transcription. a Putative and verified transcription factor-binding sites in the mouse MCP-1 promoter region. The putative binding sequences are shown in bold. The transcription initiation site is defined as +1. b High conservation of the CACCC-binding site in the MCP-1 promoter in mammals. c Cardiomyocytes were isolated from WT mice and subjected to the ChIP assay with the antibody against IgG or KLF10. The immunoprecipitated DNA was subjected to semiquantitative PCR and Q-PCR. d HEK293 cells were transfected with a full-length MCP-1 promoter luciferase reporter (top) or an MCP-1 luciferase reporter with a 2-bp mutant within the CACCC site, together with plasmids encoding full-length KLF10 or empty vector (mock). Forty-eight hours later, the luciferase activities were determined. e HEK293 cells were transfected with a full-length MCP-1 promoter luciferase reporter together with plasmids encoding full-length KLF10 or empty vector (mock) in the presence or absence of 100 nM trichostatin A (TSA). Forty-eight hours later, the luciferase activities were determined. f, g HEK293 cells were cotransfected with the indicated expression plasmids, and after 24 h, cells were harvested, lysed and subjected to immunoprecipitation by the indicated antibodies. Immunoprecipitation (IP) and whole cell lysates (input) were subjected to Western blotting with the indicated antibodies. h HEK293 cells were transfected with a full-length MCP-1 promoter luciferase reporter together with plasmids encoding KLF10, HDAC or both as well as empty vector. Forty-eight hours later, the luciferase activities were determined. i Cardiomyocytes were isolated from WT and KLF10−/− neonatal mice and subjected to the ChIP assay with an antibody against IgG or HDAC1. The immunoprecipitated DNA was subjected to Q-PCR. The data are presented as the mean ± SEM of three representative independent experiments. The data in c and h were analyzed by an unpaired two-tailed Student’s t test. The data in d, e, and i were analyzed using one-way ANOVA followed by Bonferroni post hoc analysis

Recently, KLF10 was reported to repress C/EBPα transcription by interacting with histone deacetylase 1 (HDAC1) and decreasing the enrichment of acetylated histone H4 on the C/EBPα promoter.²⁶ Moreover, HDAC1 binds to a specific regulatory region of the MCP-1 promoter and inhibits its expression.²⁷ Thus, we further explored whether HDAC1 participated in the regulation of MCP-1 transcription. As shown in Fig. 9e, treatment with the HDAC inhibitor trichostatin A (TSA) blocked the inhibitory effect of KLF10 on MCP-1 transcription. Next, the interaction between KLF10 and HDAC1 was determined by a coimmunoprecipitation assay in 293T cells transiently transfected with plasmids encoding KLF10 and/or FLAG-HDAC1. The results showed that the anti-FLAG antibody pulled down the KLF10 protein, and the anti-KLF10 antibody pulled down FLAG-HDAC1, verifying the direct interaction between KLF10 and HDAC1 (Fig. 9f, g). Furthermore, a luciferase assay showed that the inhibition of MCP-1 promoter activity was much more obvious upon overexpressing KLF10 and HDAC1 simultaneously than upon expressing either one alone (Fig. 9h). Next, to investigate the role of KLF10 in recruiting HDAC1 to the MCP-1 promoter, cardiomyocytes were isolated from neonate WT or KLF10−/− mice and subjected to a ChIP assay using an anti-HDAC1 antibody. KLF10 knockout inhibited the binding of HDAC1 to the MCP-1 promoter in cardiomyocytes (Fig. 9i). Taken together, these data indicate that KLF10 interacts with HDAC1 and promotes the recruitment of HDAC1 to the MCP-1 promoter, leading to the suppression of MCP-1 transcription.

Discussion

Recently, cardiomyocytes have been regarded as key players in eliciting and modulating inflammation within the myocardium in response to infectious (virus, bacterial or protozoan) and noninfectious causes, such as autoimmunity.^8,10 In the early stage of VMC, cardiomyocytes produce cytokines and chemokines to promote the extravasation of immune cells and their subsequent recruitment to infection sites.¹³ Monocyte chemotactic protein-1 (MCP-1) is one of the best-studied chemokines secreted by infected cardiomyocytes, and it binds to CC chemokine receptor 2 (CCR2), resulting in the accumulation of monocytes to inflammation sites. There is overwhelming evidence regarding the essential roles of MCP-1 in cardiovascular diseases.¹⁴ However, the mechanism underlying the regulation of its expression in cardiomyocytes remains poorly understood. Here, we report that KLF10 acts as a novel transcriptional regulator to inhibit MCP-1 expression in cardiomyocytes and protects against CVB3-induced myocarditis.

KLF10 was originally identified as an early gene induced by TGF-β and was also named TGF-β inducible early gene 1 (TIEG1). Subsequent studies demonstrated that KLF10 is widely expressed in skeletal and smooth muscle cells, glial cells, fibroblasts, and cardiac cells, as well as in epithelial cells, myeloid cells and various carcinoma cells. Furthermore, except for TGF-βs, the expression of KLF10 can be induced by bone morphogenetic protein-2, estrogen, nerve growth factor, and so on.²⁸ In this study, we found that both the mRNA and protein levels of KLF10 were significantly inhibited after CVB3 infection not only in cultured cardiomyocytes but also in the hearts and pancreases of mice with VMC (Fig. 1 and Fig. S1). However, we found no significant difference in TGF-β expression between KLF10−/− and WT mice (data not shown). In addition, it has been reported that the rapid downregulation of KLF10 in macrophages induced by LPS is abrogated by BAY11-7082, which is an IκB-α inhibitor, suggesting that KLF10 is an NF-κB target gene.²² However, treatment with BAY11-7082 in cardiomyocytes did not alter the decreased expression of KLF10 induced by CVB3 infection (Fig. S6). There is evidence showing that CVB3 infection activates the NF-κB pathway through a PI3K/Akt-dependent mechanism but not IκB-α phosphorylation.²⁹ As shown in Fig. 6, the PI3K inhibitor LY294002 could partly rescue the downregulation of KLF10 expression in cardiomyocytes. This is in accordance with a previous report showing that KLF10 is the target gene of the NF-κB pathway. However, in addition to direct viral replication in cardiomyocytes, we could not exclude the possibility that virus-induced early inflammatory cytokines might also trigger NK-κB or other unknown pathways to regulate KLF10 expression in infected or noninfected cells. Therefore, our current data suggested that virus infection could downregulate KLF10 expression in a PI3K/Akt-dependent manner; however, the direct mechanism underlying the regulation of this gene by virus infection needs to be further explored.

Although T lymphocyte-mediated immune responses are crucial for VMC, CD11b⁺ Mo/Mφs constitute a majority of accumulated leucocytes, especially at the early stage. They have a vital role in VMC progression.³⁰ Currently, it is widely accepted that Ly6C^high monocytes possess a more obvious proinflammatory phenotype than the Ly6C^low subset. Most studies demonstrated that the depletion of Ly6C^high monocytes by the anti-MCP-1 antibody or siRNA-targeted CCR2 could significantly alleviate acute viral-induced myocarditis.^15,16 However, in inflammatory dilated cardiomyopathy, Wu et al. reported that the depletion of CD11b⁺LyG^-Mo/Mφs by clodronate-loaded liposomes or by decreasing the ratio of Ly6C^high to Ly6C^low Mo/Mφs by PBS-loaded liposomes did not affect the severity of myocarditis but protected mice from fibrosis and DCM development.³¹ Here, we found that anti-MCP-1 Ab depletion could obviously decrease the early recruitment of monocytes and alleviate acute VMC progression, and the disruption of MCP-1 signaling abolished the pathological differences in the severity of myocarditis between KLF10-deficient and WT mice (Fig. 8). Bone marrow chimeric mouse experiments further indicated that the expression of KLF10 in nonhematopoietic cells, not in immune cells, was responsible for alleviating VMC. We did not observe a direct effect of KLF10 on CVB3 replication and myocyte death (Fig. 6). Our previous study has shown that disruption of the MCP-1 pathway in vivo inhibits cardiac viral replication.¹⁶ Therefore, the enhanced virus replication in KLF10-/- mice could be an indirect consequence of the elevated MCP-1 level within the myocardium.

Most KLFs share a common sequence, three C2H2 zinc finger motifs, in the C-terminal region, and this sequence can recognize and bind to a CACCC element or a GC box sequence to regulate the transcription of target genes. We further found one highly conserved CACCC element in the promoter of MCP-1. ChIP and luciferase reporter assays showed that KLF10 directly bound to the CACCC element and inhibited the transcription of MCP-1 (Fig. 9). In addition to zinc finger DNA-binding domains, the KLF10 protein contains several SH3 binding domains and unique repression domains in the N-terminus, which are reported to interact with the HDAC complex in regulatory T cells and in 3T3-L1 preadipocytes.²⁶ p50, which is a transcriptional repressor, recruits HDAC1 to a specific regulatory region of the MCP-1 promoter and decreases the enrichment of acetylated histone H4 on its promoter.²⁷ In this report, we also tested this possibility and demonstrated that in cardiomyocytes, KLF10 interacts with HDAC1 and then promotes KLF10 recruitment to the promoter of MCP-1, leading to the downregulation of MCP-1 transcription (Fig. 9).

In summary, we explored the role of KLF10 in the progression of VMC and identified MCP-1 as a new target of KLF10 in cardiomyocytes. As illustrated in Fig. S7, our data indicated that KLF10 cooperates with HDAC1 and inhibits MCP-1 transcriptional activity in resting cardiomyocytes to sustain myocardium homeostasis, while CVB3 infection suppresses KLF10 expression, leading to the activation of MCP-1 transcription and the recruitment of Ly6C^high monocytes/macrophages into the myocardium, which might provide new insights into the pathogenesis of VMC and a potential therapy target for VMC.

Author contributions

M.L. designed the study, analyzed the data and wrote the manuscript. J.Y., H.K.Z., X.L.W., Y.H.S. and Y.L. performed the experiments. J.G., L.W. and Y.X.Z. revised the paper. W.X., L.W., M.S., and T.C.S. supervised the study and reviewed the paper. All authors read and approved the final version of the paper.

Competing interests

The authors declare no competing interests.

Supplementary information

The online version of this article (10.1038/s41423-020-00539-x) contains supplementary material.