Inhibition of microRNA-23b prevents polymicrobial sepsis-induced cardiac dysfunction by modulating TGIF1 and PTEN

Biomedicine & Pharmacotherapy 103 (2018) 869–878

Contents lists available at ScienceDirect

Biomedicine & Pharmacotherapy journal homepage: www.elsevier.com/locate/biopha

Inhibition of microRNA-23b prevents polymicrobial sepsis-induced cardiac dysfunction by modulating TGIF1 and PTEN Haiju Zhanga,b, Yi Caudlea, Aamir Shaikha, Baozhen Yaob, Deling Yina, a b

T

⁎

Department of Internal Medicine, College of Medicine, East Tennessee State University, Johnson City, TN, United States Department of Pediatrics, Renmin Hospital of Wuhan University, Wuhan, China

A R T I C LE I N FO

A B S T R A C T

Keywords: microRNA-23b Late sepsis Cardiac dysfunction Cardiac fibrosis TGIF1 PTEN

Cardiovascular dysfunction is a major complication associated with sepsis induced mortality. Cardiac fibrosis plays a critical role in sepsis induced cardiac dysfunction. The mechanisms of the activation of cardiac fibrosis is unclarified. In this study, we found that microRNA-23b (miR-23b) was up-regulated in heart tissue during cecal ligation and puncture (CLP)-induced sepsis and transfection of miR-23b inhibitor improved survival in late sepsis. Inhibition of miR-23b in the myocardium protected against cardiac output and enhanced left ventricular systolic function. miR-23b inhibitor also alleviated cardiac fibrosis in late sepsis. MiR-23b mediates the activation of TGF-β1/Smad2/3 signaling to promote the differentiation of cardiac fibroblasts through suppression of 5′TG3′-interacting factor 1 (TGIF1). MiR-23b also induces AKT/N-Cadherin signaling to contribute to the deposition of extracellular matrix by inhibiting phosphatase and tensin homologue (PTEN). TGIF1 and PTEN were confirmed as the targets of miR-23b in vitro by Dual-Glo Luciferase assay. miR-23b inhibitor blocked the activation of adhesive molecules and restored the imbalance of pro-fibrotic and anti-fibrotic factors. These data provide direct evidence that miR-23b is a critical contributor to the activation of cardiac fibrosis to mediate the development of myocardial dysfunction in late sepsis. Blockade of miR-23b expression may be an effective approach for prevention sepsis-induced cardiac dysfunction.

1. Introduction Sepsis is identified as a systemic deleterious inflammatory response to infection or injury [1]. The severity of sepsis and septic shock is associated with high mortality rate, which mainly results from dysfunction and failure of vital organs [1,2]. Cardiovascular dysfunction is a major complication associated with sepsis induced mortality [3]. Survivors of severe sepsis also have high risk of cardiovascular events [4]. Cardiac fibrosis, an important hallmark of maladaptive hypertrophy, plays a critical role in sepsis induced cardiac dysfunction [5,6], which is characterized by the adverse accumulation of collagens and other extracellular matrix (ECM), resulting in myocardial stiffness, cardiac remodeling, and eventual heart failure [7]. However, the mechanisms inducing the activation of cardiac fibrosis remain unclarified in severe sepsis. The activation of transforming growth factor-β1 (TGF-β1) signaling mediates cardiac fibrosis by accumulation of fibroblasts and fibroblastto-myofibroblast transition (FMT) [8]. TGF-β1 promotes profibrotic signaling by activating Smad2/3 canonical pathway [9]. In addition, TGF-β1 signal can also orchestrate through non-canonical pathways

⁎

including PI3K/AKT, ERK1/2, and p38 MAPK, which can coordinate with the Smad-dependent canonical pathway to induce fibrosis [9,10]. Therefore, the negative regulators of the TGF-β1/Smad2/3 pathway are well-defined to protect against fibrosis [11]. 5′TG3′-interacting factor 1 (TGIF1) is the transcriptional repressor of TGF-β1 signaling via the Smad-dependent pathway [12]. Phosphatase and tensin homologue (PTEN) is a dual protein which dephosphorylates focal adhesion kinase (FAK) and suppresses the activation of PI3K/AKT signaling [13]. Sepsis is initiated by a hyperinflammatory reaction and shifts within a few days to a protracted state of anti-inflammation and immunosuppression, which is associated with increased production of immunosuppressive cytokines, including TGF-β1 [14]. Inhibition of TGF-β1/Smad2/3 signal may improve the host immunosuppression following sepsis [15], and deletion of Smad2/3 from cardiac fibroblasts similarly inhibited the gene program for fibrosis and extracellular matrix remodeling [16]. The complex pathogenesis of sepsis-induced cardiomyopathy involves a combination of dysfunction of cardiomyocytes, cardiac fibroblasts and/or endothelial cells [17]. Cardiac fibroblasts are important pathogenesis in inflammation and fibrosis in the heart during sepsis and lead to cardiac dysfunction that would affect the

Corresponding author at: Department of Internal Medicine, College of Medicine, East Tennessee State University, Johnson City, TN, 37614, United States E-mail address: [email protected] (D. Yin).

https://doi.org/10.1016/j.biopha.2018.04.092 Received 26 January 2018; Received in revised form 5 April 2018; Accepted 13 April 2018 0753-3322/ © 2018 Elsevier Masson SAS. All rights reserved.

Biomedicine & Pharmacotherapy 103 (2018) 869–878

H. Zhang et al.

USA), using a tidal volume of 0.3–0.5 ml and a respiratory rate of 110–120 breaths/min. After left thoracotomy, a microtip pressure-volume catheter was inserted into the left ventricular to measure the cardiac function. The signals were continuously recorded using an ARIA pressure-volume conductance system (Millar Instruments) connected to a Powerlab/4SPA/D converter (AD Instruments, Mountain View, CA, USA). All pressure-volume loop data were analyzed with PVAN3.4 (Millar Instruments). After the functional analysis, the hearts were harvested. Portions of the mid-ventricle were fixed for histologic examination and immunohistochemistry (IHC) detection.

outcome of sepsis by elevating adhesion to ECM and adhesive signaling [18]. microRNAs (miRNAs) are dominant players in different aspects of cardiac remodeling, including fibrosis [19,20]. microRNA-23b (miR23b) is emphasized recently as a multiple functional miRNA because of the prominent effects on immunology and inflammatory signal in sepsis and autoimmune diseases [21–23]. Iaconetti et al reported that miR23b as a regulator of vascular smooth muscle cells (VSMC) phenotypic switch and revealed that miR-23b suppressed urokinase-type plasminogen activator, SMAD family member 3, and transcription factor forkhead box O4 (FoxO4) expression in phenotypically modulated VSMCs [25]. However, the role of miR-23b in sepsis-induced cardiac dysfunction is not known yet. Knockdown of miR-23b cluster miRNAs in fetal and newborn liver can block or revert TGF-β1-induced liver fibrosis [26]. miR-23b-3p was significantly up-regulated in keloid fibroblasts (KFs) contributing to the etiology of keloids by affecting several pro-fibrotic signaling pathways [27]. These studies suggest that the upregulation of miR-23b is associated the activation of pro-fibrotic signal. Therefore, we hypothesized that the induction of miR-23b in polymicrobial sepsis might emerge as a modifier to regulate fibrotic remodeling in the heart.

2.4. Cell lines, transfection of plasmids, regent treatments and dualluciferase reporter assay HEK293T cell lines were obtained from the American Type Culture Collection (Rockville, MD) and human cardiac fibroblast cells (HCFs) were from Lonza Company. HEK293T cells were cultured in DMEM (Invitrogen) supplemented with 10% fetal bovine serum (FBS) and HCFs were cultured in DMEM/F12 (Invitrogen) with 10% FBS and 1% Hepes. All the cells were incubated in a humidified atmosphere containing 5% CO2 at 37 °C. The wild type (WT) 3′-UTR of TGIF1 and PTEN were cloned using primers and inserted into the SacI and XhoI sites of pmirGLO vector (Promega), as well as the mutant type (Mut) 3′-UTR plasmids of TGIF1 and PTEN (Genepharma, Beijing, China). Mut1 plasmid was mutant at binding site 1, and Mut2 at site 2. miRs and TGIF1/PTEN WT/Mut 3′UTR plasmids were co-transfected in HEK293T cells. Plate 3 wells for each condition to produce triplicate (conditions include negative controls for miRNA and UTR sequence). Validation of miR-23b binding to 3′- UTR was performed using Dual-Glo Luciferase assay system (Promega) as recommended by the manufacturer. The relative luciferase activities were determined by calculating the ratio of firefly luciferase activities over Renilla luciferase activities. All experiments were repeated three times in triplicate. HCFs were treated with different concentrations of recombinant human TGFβ1 from Humanzyme (Chicago, IL). They were also transfected mirVanaTM miR-23b mimics, negative control (NC) or miR-23b inhibitors (Genepharma, Beijing, China) with the final concentration of 100 nM using the transfection reagent HiPerfect® (QIAGEN, Germantown, MD) and stimulated with Lipopolysaccharides (LPS, 20 μg/ml) (Sigma-Aldrich) for 6 h. The supernatants and cells were collected for ELISA, quantitative RT-PCR or western blot examination.

2. Materials and methods 2.1. Animals and cecal ligation and puncture (CLP) induced polymicrobial late sepsis model Wild-type (WT) C57BL/6 mice were obtained from Jackson Laboratory (Bar Harbor, ME). 8–10-week old male mice were used for cardiac function analysis. All mice were maintained in the Division of Laboratory Animal Resources at East Tennessee State University (ETSU). The Animal experimental protocols were approved by the ETSU Committee on Animal Care. Polymicrobial late sepsis was induced by CLP as described previously [28]. Briefly, mice were anesthetized via 5.0% isoflurane inhalation with 100% oxygen in a closed chamber. A small anterior abdominal incision was made, and the cecum was ligated 1 cm proximal to the terminal of cecum, and then was punctured twice with a 23gauge needle. A small amount of feces content was extruded into the abdominal cavity. Sham-operated mice were processed identically, except without ligation and puncture. To create the late sepsis phenotype, mice were subcutaneously administered antibiotic (imipenem, 25 mg/ kg body weight) or an equivalent volume of 0.9% saline at 8 and 16 h after CLP. We and others have reported that the early sepsis is confirmed by elevated cytokine levels in the first 5 days after CLP, and the late/chronic sepsis (after day 5) was confirmed by reduced circulating pro-inflammatory cytokines [14,24]. In this study, 3 days after CLP represents for the early sepsis and 12 days for the late sepsis.

2.5. Quantitative PCR (q-PCR) assay of miRs miRs were isolated from heart tissues using mirVana™ miR isolation kit (Ambion) according to the manufacturer’s protocol. miR-23b levels were quantified by qPCR using specific primers TaqMan MicroRNA Assay (primer identification numbers:000,400 for mmu-miR-23b and 001,973 for snRU6) and TaqMan Universal PCR Master Mix (Applied Biosystems) on a Bio-Rad PCR instrument. miR-23b level was calculated using the 2−ΔΔCT cycle threshold method after normalization to the snRU6 as an internal control.

2.2. In vivo injection of miR-23b inhibitor by tail vein Both mirVanaTM in vivo ready miR-23b inhibitor and negative control (miR-Con) (Ambion, Carlsbad, CA) were complex with Invivofectamine 3.0 (Invitrogen) reagent according to the manufacturer’s protocol and our previous study [28], and were injected via the tail vein at the dose of 5 mg/kg in 100 μl volumes. Injection was performed 48 h after CLP to allow initiation of sepsis.

2.6. Cardiac tissue Masson’s trichrome staining and immunohistochemistry The hearts were fixed by immersion in 4% buffered paraformaldehyde overnight, embedded in paraffin, and cut into 5-μm-thick sections. Masson’s Trichrome staining: after deparaffinization, slides were stained with hematoxylin and eosin by standard methods. In addition to staining with hematoxylin and eosin, Azan staining was performed with aniline blue to identify fibrosis in the heart. IHC was performed with a Vector ® M.O.M.™ Immunodetection Kit (Vetor Laboratories). Paraffin heart sections (5 μm thick) were blocked in 1% Blocking Reagent for 1 h at 37 °C and then incubated with anti-intercellular adhesion molecule 1 (ICAM-1) (Santa Cruz) and anti- vascular cell adhesion molecule 1

2.3. Cardiac functional analysis Cardiac function was detected as described previously [28]. Heart rate (HR), ejection fraction (EF), left ventricular end- systolic pressure (Pes), cardiac output (CO) were calculated by the SPR-839 instrument (Millar Instruments, Houston, TX, USA). Briefly, each mouse was intubated with a 22-gauge soft catheter and ventilated with a rodent ventilator (Columbus Instruments International Corp., Columbus, OH, 870

Biomedicine & Pharmacotherapy 103 (2018) 869–878

H. Zhang et al.

Fig. 1. Sepsis induced microRNA-23b (miR-23b) expression in heart tissues. (A) Polymicrobial sepsis increased miR-23b levels in the heart tissues in early and late phases. (B) The levels of miR-23b in heart tissues were determined after 3 days or 12 days. Eight-to 10- weeks-old C57BL/6 male mice were subjected to cecal ligation and puncture (CLP) or Sham operation. miR-23b inhibitor or miR-Con (5 mg/kg) was injected through the tail vein 48 h after CLP. Heart tissues were harvested for assay of miR-23b by q-PCR (n = 5 per group). (C) The expression of miR-23b was determined in HCFs with different concentrations of TGF-β1 treatment (10 ng/ml and 20 ng/ml) by q-PCR. **p < 0.01, *** p < 0.001 compared with indicated groups.

TGT TGC TGA TGG-3′; mouse TGIF1 forward 5′-CTT GGT AGA TTA TTC ATA AGA TGG TTC CCA ATA TC-3′, TGIF1 reverse 5′-GAT ATT GGG AAC CAT CTC TTA TGA ATA ATC TAC CAA G-3′; mouse PTEN forward 5′-TTC TGC CAT CTC TCT CCT CC-3′, PTEN reverse 5-ATC CGT CTA CTC CCA CGT TC-3′; mouse GAPDH forward 5′-TCA ACA GCA ACT CCC ACT CTT CCA- 3′, GAPDH reverse 5′-ACC CTG TTG CTG TAG CCG TAT TCA-3′; human TGF-β1 forward 5′-CCC AGC ATC TGC AAA GCT C-3′, TGF-β1 reverse 5′-GTC AAT GTA CAG CTG CCG CA-3′; human IFN-γ forward 5′- TGA CCA GAG CAT CCA AAA GA -3′, IFN-γ reverse 5′-CTC TTC GAC CTC GAA ACA GC -3′; human TGIF1 forward 5′-GAA TTG TGC CAG TGT TTC TCT TTG-3′, TGIF1 reverse 5′-CGG CGC TGT CAG AGT GAG AGA GGC-3′; human PTEN forward 5′-CGA CGG GAA GAC AAG TTC AT-3′, PTEN reverse 5′-AGG TTT CCT CTG GTC CTG GT-3′; human GAPDH forward 5′-GAA GGT GAA GGT CGG AGT C -3′, GAPDH reverse 5′-GAA GAT GGT GAT GGG ATT TC -3′.

(VCAM-1) (Santa Cruz) respectively overnight at 4 °C. The signal was with DAB substrate and counterstained with hematoxylin. 2.7. Western blot Western blot was performed as described previously [28]. Samples containing equal amounts of protein extracted from heart tissue or cell lysis were loaded into 10–12% SDS-PAGE, and then transferred to a nitrocellulose membrane. After blocking the membrane for 1 h at room temperature with 5% BSA, the membrane was incubated overnight at 4 °C with the primary antibody. The signal was detected with ECL system (Amersham Biosciences) and quantified by Bio-Image Analysis System (Bio-Rad). Anti-phospho-AKT, anti-AKT, anti-phospho-Smad2/ 3, anti-Smad2/3, anti-N-Cadherin, anti-α-SMA, anti-MMP9, anti-PTEN and GAPDH antibodies were obtained from Cell Signaling Technology. Anti-TGIF1 antibody was obtained from Sant Cruz Technology.

2.11. Statistical analysis

2.8. Enzyme linked immunosorbent assay (ELISA) for cytokines

Data were analyzed using software Graphpad Prism 6. The data were analyzed by Student’s t-test for two-group comparison or one-way analysis of variance (ANOVA) or two-way ANOVA as appropriate. Survival curves were generated by the Kaplan-Meier method. All values are expressed as mean and standard deviations (SD). A value of p < 0.05 is considered statistically significant.

The levels of pro-fibrotic factor TGF-β1 and anti-fibrotic factor IFNγ in serum and supernatants of HCM were collected after cultivation were analyzed with mouse or human Quantikine® ELISA immunoassay kits (R&D systems, Minneapolis, MN) according to the manufacturer’s instructions. 2.9. Electrophoretic mobility shift assay (EMSA)

3. Results

Nuclear protein was isolated from heart tissue with NE-PER Nuclear and Cytoplasmic Extraction Reagents according to the manual. Smad2/ 3 binding activity was performed using a light shift chemiluminescent EMSA kit (Thermo Fisher Scientific) as previously described [29] in a 20 fmol double-stranded Smad2 and Smad3 consensus oligonucleotide, which was 5′-end-labeled with biotin and 15 ug nuclear proteins. The biotin end-labeled DNA was detected using the streptavidin-HRP conjugate and chemiluminescent substrate.

3.1. Expression miR-23b is induced and maintained in hearts during sepsis We induced sepsis by CLP and harvested hearts in the early (3 days after CLP) and late (12 days after CLP) periods for examination of miR23b by qPCR. As shown in Fig. 1A, polymicrobial sepsis increased the expression of miR-23b in myocardium both in early (5.6-fold compared to sham control) and late period (7.5-fold compared to sham control). To determine whether miR-23b is linked to cardiac dysfunction during late sepsis, miR-23b inhibitor was injected through the tail vein 48 h after CLP to allow the initiation of sepsis. As shown in Fig. 1B, the expression of miR-23b decreased in late sepsis, and approximately returned the baseline level after the injection of miR-23b inhibitor. We also tested the expression of miR-23b in HCFs with different concentration of TGF-β1 stimulation. As shown in Fig. 1C, miR-23b expression was up-regulated in a dose dependent manner. Taken together, these results indicate that miR-23b could be induced in myocardium by polymicrobial sepsis and might serve as a regulatory factor to pro-fibrotic signal during sepsis.

2.10. Real-time quantitative RT-PCR The quantitative RT-PCR detection technique was performed as described previously [28]. Total RNA in heart tissues and cultured cardiac myocytes was isolated with Qiagen RNeasy kit (Qiagen). Quantitative RT-PCR was performed using RT2 real-timeTM SYBR Green Fluorescien PCR Master Mix (Qiagen). GAPDH was used as internal control. The primers were as followings: mouse TGF-β1 forward 5′-CAA CAA TTC CTG GCG TTA CCT TGG-3′, TGF-β1 reverse 5′-GAA AGC CCT GTA TTC CGT CTC CTT-3′; mouse IFN-γ forward 5′-AGG AAC TGG CAA AAG GAT GGT GAC-3′, IFN-γ reverse 5′-TGA CGC TTA TGT 871

Biomedicine & Pharmacotherapy 103 (2018) 869–878

H. Zhang et al.

Fig. 2. Transfection of miR-23b inhibitor improves survival outcome and attenuates cardiac dysfunction in late sepsis. (A) After 48 h of CLP, mice were transfected with miR-23b inhibitor and then monitored for survival for up to 28 days. There were 12/group. (B) After 12 days of CLP, cardiac function measurements were analysis by pressure-volume loop hemodynamic parameters. (B) HR, heart rate. (C) EF, ejection fraction. (D) CO, cardiac output. (E)Pes, left ventricular end-systolic pressure. There were five mice per group. *p < 0.05, ** p < 0.01, *** p < 0.001 compared with indicated groups.

3.3. MiR-23b inhibitor alleviates cardiac fibrosis in late sepsis

3.2. Injection of miR-23b inhibitor improves late sepsis survival and attenuates cardiac dysfunction

Sham-operated and CLP septic mice did not substantially differ in cardiac chamber morphology (data not shown). However, as shown in Fig. 3A, representative photomicrographs of Masson-stained left ventricular sections from sham-operated, CLP, miR-23b inhibitor and miRCon treated CLP mice, perivascular and interstitial fibrosis were evident in the hearts of CLP-induced septic mice. The summarized data of cardiac fibrosis ratio showed that rapid progress of cardiac fibrosis seen in septic mice was significantly prevented by miR-23b inhibitor. Activated fibroblasts express α-smooth muscle actin (α-SMA) and are often referred to as myofibroblasts. Matrix metallopeptidase 9 (MMP-9) is an ECM molecule involved in cardiac remodeling, which was correlated with the severity and mortality of sepsis [6]. We determined the levels of α-SMA and MMP-9 in myocardium in late sepsis. The results showed α-SMA and MMP-9 levels elevated significantly in septic mice. In contrast, miR-23b inhibitor abrogated the expression of α-SMA, as well as the expression of MMP-9 (Fig. 3B), which indicated that miR-23b mediated the differentiation cardiac fibroblasts at least in

The Kaplan-Meier survival curve (Fig. 2A) indicated the survival was improved by 42% with the injection of miR-23b inhibitor compared with miR-Con and CLP group. Cardiac dysfunction plays a critical role in sepsis-induced mortality [6,28]. To define the role of miR-23b in cardiac dysfunction during late sepsis, hemodynamic parameters were examined by pressure-volume loop measurement on day 12 after sepsis. CLP induced significant cardiac dysfunction in late sepsis by decreased heart rate (26.6%) (Fig. 2B), EF% (41.5%) (Fig. 2C), cardiac output (48.3%)(Fig. 2D), and Pes (32.7%) (Fig. 2E), compared with baseline values. Whereas miR-23b inhibitor treatment dramatically relieved CLP-induced cardiac dysfunction, which increased heart rate by 21.9%, EF% by 36.7%, cardiac output by 44.2%, and Pes (28.6%) compared with the untreated CLP together. These data suggest inhibition of miR23b in myocardium attenuates sepsis -reduced cardiac output and enhances left ventricular function, which is highly related to the survival in late sepsis. 872

Biomedicine & Pharmacotherapy 103 (2018) 869–878

H. Zhang et al.

Fig. 3. Injection of miR-23b inhibitor attenuated cardiac interstitial fibrosis in CLP-induced septic mice. (A) Myocardial fibrosis content assessed by Masson’s Trichrome staining (400× magnification) and quantitation with left ventricular sections from sham, CLP, miR-23b treated CLP and miR-Con treated CLP mice. Perivascular (upper) and interstitial (lower) fibrotic areas are shown (blue). Fibrotic zones were analyzed and are shown as percentage of the fibrotic area under 120,000 μ m2 of sections (n = 5 per group). (B) The protein levels of α-smooth muscle actin (α-SMA) and matrix metalloproteinase-9 (MMP-9) were determined by western blot in heart tissue as the protocol (A). (C) The protein levels of TGF-β1 and IFN-γ in circulation were examined by ELISA. (D) Hearts were harvested for assay of cytokine TGF-β1 and IFN-γ by quantitative RT-PCR 12 days after CLP. There were 5 mice per group. *p < 0.05, ** p < 0.01, *** p < 0.001 compared with indicated groups.

3.5. miR-23b promotes the activation of TGF-β1/Smad2/3 signaling by suppressing TGIF1 to mediate cardiac fibrotic remodeling in late sepsis

part in late sepsis. IFN-γ exhibits potent anti-fibrotic activity in many tissues [30,31]. We examined the expression of pro-fibrotic cytokine TGF-β1 and antifibrotic cytokine IFN-γ. As expected, the serum concentration of TGF-β1 and mRNA level in myocardium were both increased in late sepsis, however, the expression of IFN-γ was decreased in serum and myocardium. Injection of miR-23b inhibitor restored the imbalance of proand anti-fabric factors in late sepsis (Fig. 3B). These results suggested that miR-23b contributed to cardiac fibrosis in late sepsis by promoting the activation of pro-fibrotic factors.

TGF-β1/Smad2/3 signaling is principal mediators of the fibrotic response in activated tissue-resident cardiac fibroblasts and regulates the expression of pro-fibrotic factors[16]. We examined the effect of miR-23b on Smad2/3 expression and binding activity in myocardium in late septic mice. As shown in Fig. 5A, the protein level and binding activity of Smad2/3 in myocardium both increased significantly compared with sham group in late sepsis. However, miR-23b inhibitor prevented sepsis-induced Smad2/3 expression and binding activity. TGIF1is transcriptional repressor of TGFβ1/Smad2/3 signal pathway [33]. The protein and gene level of TGIF1 was suppressed in septic mice, but miR-23b inhibitor restored the expression of TGIF1. In contrast, miR-Con treatment failed to upregulate TGIF1 protein and mRNA level in heart (Fig. 5B and C). Since sepsis induces the elevation of TGFβ1 and miR-23b. We also tested the expression of TGIF1 and Smad2/3 in HCFs after different dose of TGF-β1 stimulation. TGIF1 was downregulated and Smad2/3 was activated in a TGF-β1 dose-dependent manner (Fig. 5D). These results suggest that the induction of miR-23b in sepsis mediates the activation of TGFβ1/Smad2/3 signal by suppressing the expression of TGIF1.

3.4. Injection of miR-23b inhibitor blocks collagen deposition by suppressing adhesive signaling of myocardium in late sepsis Elevated adhesive molecules are deposited in the ECM and result in fibrotic remodeling [32]. We examined the effect of miR-23b on the adhesive molecules in the myocardium. As shown in Fig. 4A and B, there is more positive staining of ICAM-1 and VCAM-1 in the myocardium of late-septic mice compared with sham control. Moreover, injection of miR-23b inhibitor alleviated sepsis-induced accumulation of ICAM-1 and VCAM-1. miR-Con treatment had no influence. These results indicate that miR-23b promotes the activation of adhesive molecules to exacerbate the fibrotic remodeling in myocardium.

873

Biomedicine & Pharmacotherapy 103 (2018) 869–878

H. Zhang et al.

Fig. 4. Transfection of miR-23b inhibitor prevents the activation of adhesive molecules in myocardium in late sepsis. miR-23b inhibitor or miR-Con was injected through the tail vein 48 h after CLP. Hearts were harvested 12 days after CLP and sectioned for immunohistochemical staining to identify the activation of ICAM-1 (A) and VCAM-1 (B) in the myocardium. The bar graphs indicate the numbers of positive cells in five heart fields. There were 5 mice per group. ***p < 0.001 compared with indicated groups.

3.7. TGIF1 and PTEN are verified as the targets of miR-23b

3.6. miR-23b contributes to the activation of AKT/N-Cadherin axis by repressing PTEN to induce the deposition of ECM in myocardium in late sepsis

To define the mechanism how miR-23b promotes cardiac fibrosis in late sepsis, using online databases (TargetScan and microCosm), TGIF1 and PTEN were potential targets of miR-23b (Fig. 7A and C). To validate the putative binding sites, WT and mutant (MT) 3′-UTR of TGIF1and PTEN (mut1 mutants in binding site 1 and mut2 mutants in binding site 2) were cloned into the pmirGLO vector. The transcripts of the target genes were then assessed by dual luciferase reporter assays. When miR23b mimics were co-transfected with WT or MT pmirGLO vector in HEK293T cells, miR-23b reduced the luciferase activity 52% of TGIF1 WT-UTR, suggesting a direct interaction between miR-23b and 3′-UTR of TGIF1. Luciferase activities of TGIF1 Mut1-UTR decreased 47%, but Mut2-UTR hadn’t significant alteration in comparison with that of NC (Fig. 7B). These results suggest that miR-23b bounds to site 2 rather than site 1 of TGIF1. miR-23b reduced the luciferase activity 46% of PTEN WT-UTR. Whereas luciferase activities of PTEN Mut1-UTR were restored completely, Mut2-UTR decreased by 42% (Fig. 7D), suggesting that miR-23b bounds to site 1 rather than site 2. To further reveal the effect of miR-23b on myocardium, miR-23b mimic and inhibitor were transfected respectively to HCFs, then stimulating 6 h with LPS (20 μg/ml) and examined the expression of TGIF1 and PTEN. The protein levels of TGF-β1 and IFN-γ in

The activation of PI3K/AKT signaling pathway aggravates renal, pulmonary and cardiac fibrosis [34–36]. PTEN inhibits the activity of PI3K protein expression and AKT phosphorylation [37], which activates the expression of α-SMA to promote fibrosis process [34]. We evaluated PTEN/AKT signal in late sepsis. As shown in Fig. 6A and B, the protein level and mRNA expression of PTEN in myocardium both decreased significantly compared with sham control in late sepsis. However, miR23b inhibitor restored sepsis-inhibited PTEN expression. Being consistent with the blockade of PTEN, phosphor-AKT was activated in septic mice. Adhesive molecule N-Cadherin increased in myocardium in late septic mice. MiR-23b inhibitor remarkably alleviated the activation of AKT and N-Cadherin in late sepsis (Fig. 6C). Our in vitro experiment also demonstrated that PTEN decreased and phosphor-AKT was activated after TGF-β1stimulation (Fig. 6D). These results indicated the induction of miR-23b promoted cardiac fibrotic propagation by targeting PTEN/AKT/N-Cadherin signal axis in late sepsis.

874

Biomedicine & Pharmacotherapy 103 (2018) 869–878

H. Zhang et al.

Fig. 5. miR-23b mediates the activation of TGF-β1/Smad2/3 signal by suppressing TGIF1. miR-23b inhibitor or miR-Con (5 mg/kg) was injected through the tail vein 48 h after CLP. (A) Hearts were harvested for the determination of Smad2/3, phosphor-Smad2/3 by immunoblotting with specific antibodies and Smad2/3 binding activity by EMSA 12 days after CLP. (B) The protein level of TGIF1 in hearts was determined by immunoblotting. (C) The mRNA level ofTGIF1 in hearts was determined by quantitative RT-PCR 12 days after CLP. (D) The expression of TGIF1 and Smad2/3 was tested in HCFs with different doses (10 ng/ml and 20 ng/ml) of TGF-β1 stimulation by quantitative RT-PCR. The two lanes under each label were treated with the same reagent (repeated). **p < 0.01, ***p < 0.001 compared with indicated groups.

improved sepsis survival and attenuated cardiac dysfunction using the late-septic mice model. We also found that the augmentation of miR23b induced the activation of TGF-β1/Smad2/3 and AKT/N-Cadherin signal pathways to mediate cardiac fibrosis through the inhibition of TGIF1 and PTEN in late sepsis. TGIF1 and PTEN are verified to be the targets of miR-23b. These findings reveal that the blockade of miR-23b in late sepsis maybe an effective approach for prevention and treatment of sepsis-induced cardiac dysfunction. Cardiac fibrosis triggered by abnormal response of cardiac fibroblasts leads to cardiac dysfunction and compliance impairments [6]. Data from the present study revealed the imbalance of pro-fibrotic and anti-fibrotic factors in myocardium, as well as the elevation of fibrosis in perivascular and interstitial areas, suggesting that cardiac fibrosis process contributes to sepsis-induced cardiac dysfunction. Wu et al revealed that miR-23b regulate the vascular endothelial function after LPS induced septic shock [21]. Zhu et al reported that the mechanism of miR-23b suppressed IL-17 associated autoimmune diseases by regulating the inflammatory signal pathway [22,23]. In the current study, we found that the upregulation of miR-23b in late sepsis contributes to the cardiac fibrosis by activating pro-fibrotic signal pathways, which might be another pathogenesis of miR-23b in inflammatory disease. It has been reported that inhibition of miR-23b in glioma cell lines and orthotopic tumor mouse models results in a reduction in tumor malignancy by downregulation of HIF-1α, β-catenin, MMP2, MMP9 [38]. We also demonstrated that miR-23b inhibitor suppressed the activation of

supernatants and mRNA levels were also tested. TGIF1 and PTEN were inhibited after miR-23b mimic transfection (Fig. 7E) The mRNA levels of TGIF1 and PTEN showed the same trends observed at the protein level after the transfection of miR-23b mimic (Fig. 7F), which indicated that miR-23b repressed the expression of TGIF1 and PTEN by specifically binding with and subsequently inducing the degradation of mRNA. The mRNA levels of TGIF1 and PTEN in HCFs increased after miR-23b inhibitor transfection. There were amounts of TGF-β1 in protein and mRNA level after mimic transfection compared with NC and inhibitor treatment, but anti-fibrotic factor IFN-γ showed the converse trends with the treatments of miR-23b mimic or inhibitor (Fig. 7G, F). These data suggest that the suppression of TGIF1 and PTEN by miR23b may be important mechanisms by which induced cardiac fibrosis in late sepsis.

4. Discussion Cardiac depression is a well-recognized manifestation of organ dysfunction in sepsis. Although sepsis-induced myocardial dysfunction has been studied in clinical and basic research for more than 30 years, its pathophysiology is not completely understood, and no specific therapies for this disorder exist. In this study, depressed cardiac function was confirmed in late sepsis by reducing heart rate, cardiac output, EF% and Pes. We demonstrated that the injection of miR-23b inhibitor 875

Biomedicine & Pharmacotherapy 103 (2018) 869–878

H. Zhang et al.

Fig. 6. miR-23b promotes the activation of AKT/N-Cadherin signal by the inhibition of PTEN. miR-23b inhibitor or miR-Con (5 mg/kg) was injected through the tail vein 48 h after CLP. Hearts were harvested for the determination the expression of PTEN by immunoblotting (A) and mRNA level by quantitative RT- PCR (B). (C) AKT, phosphor-AKT, and N-Cadherin protein levels were examined by immunoblotting 12 days after CLP. (D) The expression of PTEN and AKT was examined in HCFs after different doses (10 ng/ml and 20 ng/ml) of TGF-β1 stimulation by immunoblotting. The two lanes under each label were treated with the same reagent (repeated). **p < 0.01, *** p < 0.001 compared with indicated groups.

α-SMA and MMP-9 in late sepsis. Over-expression of miR-23b may increase cardiomyocyte apoptosis and reduce cell growth under hypoxic conditions [39]. In our recent publication, we reveal that the up-regulation of miR-23b in late sepsis promotes splenic apoptosis to mediate immunosuppression [24]. The beneficial effect of miR-23b inhibitor in sepsis-induced cardiac dysfunction could be contributed to the prevention of cardiac fibrosis development. Whether inhibition of miR-23b reduces cardiomyocyte apoptosis partly to be responsible for the improvement of cardiac dysfunction will be studied in the future. Sustained activation of TGF-β1/Smad2/3 likely underlie inappropriate developmental remodeling and a wide range of fibrotic diseases, including heart failure. In the present study, elevated serum and cardiac TGF-β1 levels were proved in late sepsis, accompanying with the upregulation of α-SMA and MMP9. These results indicate that the activation of cardiac fibroblasts and elevated synthesizing ECM proteins in late sepsis. Smad2/3 can directly bind to the promoters of fibrosis and myofibroblast-dependent genes to modulate their expression characteristics [40]. TGIF1is a key effector that antagonizes TGFβ1-induced tropoelastin expression in fibroblasts differentiation[41], which can cause repression of TGFβ1-activated genes by direct competition with the coactivator p300/CBP for Smad2 interactions[42], and mediate the antifibrotic effect [33]. Our results revealed that TGFβ1/Smad2/3 signal was activated in CLP induced late sepsis, which had productive fibrosis in the heart by expressing fibrosis-mediating genes such as α-SMA, MMP-9 and N-Cadherin. However, injection of miR-23b inhibitor dramatically alleviated the activation of TGF-β1/Smad2/3 signal, meanwhile reduced the perivascular and interstitial fibrotic

area. Most of importance, the depressed cardiac function was improved in late sepsis with the miR-23b inhibitor treatment. TGIF1 was identified as the direct target of miR-23b and was inhibited in late sepsis. We can conclude that the upregulation of miR-23b in septic mice promotes TGF-β1/Smad2/3 activation to mediate cardiac fibrosis by suppressing TGIF1. Cell adhesion molecules are markers of inflammation and fibrosis [43]. Increased expression of adhesion molecules such as ICAM-1 and VCAM-1, play a critical role in mediating inflammatory cell infiltration into the tissues. They are also upregulated in fibrosis tissues. Activated high levels of ICAM-1 and VCAM-1are involved in the pathogenesis of pulmonary fibrosis and predict multi-organ failure (MOF) during sepsis [41]. In the early phase of hypertensive fibrosis heart, inflammation upregulation is associated with increased ICAM-1 and VCAM-1, which are as proinflammatory and profibrogenic mediators both on the inflammatory and fibrotic process [44]. The blockade of VCAM-1 and ICAM-1 abrogate LPS-induced cardiac dysfunction [28,45]. In our study, the activation of ICAM-1 and VCAM-1 was observed in late sepsis. Injection of miR-23b inhibitor significantly attenuated ICAM-1 and VCAM-1 expression. Thus, the suppression of ICAM1 and VCAM1 in miR-23b inhibitor treated mice is associated with both the attenuation of fibrotic process and inflammatory responses. Fibroblasts present in fibrotic lesions are characterized by elevated adhesion to ECM and adhesive signaling including FAK /PI3K/AKT, which induced cardiac fibrosis [46]. N-Cadherins mechanically link neighboring cells and are likely to contribute to fibrotic disease propagation. The activation of N-Cadherin also mediated cardiac myocyte 876

Biomedicine & Pharmacotherapy 103 (2018) 869–878

H. Zhang et al.

Fig. 7. TGIF1 and PTEN are verified as the targets of miR-23b. (A) The scheme showing the miR-23b putative binding sites in the 3′-UTR of TGIF1. (B) The relative luciferase activity was measured after miR-23b mimics and pmirGLO-TGIF1 -3′UTR vector were co-transfected in HEK293T cells. (C) Schematic representation of miR-23b putative binding sites in the 3′-UTR of PTEN. (D) The relative luciferase activity was determined after co-transfection of miR-23b mimics and WT or MT pmirGLO-PTEN -3′UTR vector in HEK293T cells. (E) HCFs were transfected with miR-23b mimics at 100 nM using Hiperfect reagent. NC oligonucleotides were as controls. The protein levels of TGIF1 and PTEN were determined by immunoblotting in harvested HCFs with LPS (20 μM) stimulation for 6 h. (F) The mRNA levels ofTGIF1 and PTEN were examined by quantitative RT-PCR after transfection miR-23b mimics (100 nM) or inhibitors (100 nM) in HCFs with LPS (20 μM) stimulation for 6 h. (G) TGF-β1 and IFN-γ protein levels were detected in supernatants of HCFs by ELISA as the protocol (F). (H)TGF-β1 and IFN-γ gene levels were examined by quantitative RT-PCR in HCFs as the protocol (E). n = 5 per group. * p < 0.05, ** p < 0.01, *** p < 0.001 compared with indicated groups.

activation of AKT/N-Cadherin signal, ICAM-1 and VCAM-1 to relieve the pro-fibrotic gene expression by the inhibition of PTEN, which was confirmed as the target of miR-23b. Taken together, the upregulation of miR-23 contributes to cardiac fibrosis to mediate cardiac dysfunction in late sepsis. The mechanism is involved in the activation of pro-fibrotic signal pathways, such as TGFβ1/Smad2/3 and AKT/N-Cadherin through the inhibition of the TGIF1 and PTEN. They are identified as the direct targets of miR-23b. Blockade of miR-23b expression may be an effective approach for prevention and/or treatment of sepsis-induced cardiac dysfunction.

remodeling [47]. Therefore, it is possible that miR-23b activates adhesive molecules by mediating AKT/N-Cadherin activation. Because of the broad consensus about septic pathophysiology (ie, initial inflammation/cytokine storm, and subsequent immunosuppression), specific interventions for the treatment of severe sepsis and septic shock in critically ill patients remain evasive [48]. The benefits in early sepsis survival have been forfeited to elevate long-term sepsis mortality. The cardiac activation of PI3K/AKT signaling has been demonstrated a causal relationship between the protection of myocardial function and survival in early sepsis [49]. However, the sustained activation of AKT/ N-Cadherin in late sepsis can increase the expression of collagen and fibronectin, which play critical role in the processing of fibrotic disease [50]. PTEN, as a negative regulator of PI3K/AKT, is down-regulated in lung and skin fibrosis as well as diabetic and ischaemic renal injury. Increased PTEN expression in vivo and in vitro plays protection through AKT inhibition and highlights AKT inhibitors as anti-fibrotic therapeutics [51]. PTEN loss initiates the induction of pro-fibrotic genes via PI3K/AKT signal [52]. Targeted inhibition of FAK/PI3K/AKT attenuates cardiac fibrosis and preserves heart function in adverse cardiac remodeling through inactivating the phenoconversion of fibroblasts and subsequent α-SMA expression [46]. MiR-23b inhibitor alleviated the

Potential conflicts of interest The authors have no financial conflicts of interest.

Acknowledgements This work was supported in part by NIH grants NIGM114716 and NIGM094740 to D. Yin. This research was also supported in part by NIH grant C06RR0306551. 877

Biomedicine & Pharmacotherapy 103 (2018) 869–878

H. Zhang et al.

References

[29] H. Zhang, Y. Caudle, C. Wheeler, et al., TGF-beta1/Smad2/3/Foxp3 signaling is required for chronic stress-induced immune suppression, J. Neuroimmunol. 314 (2018) 30–41. [30] C. Luck, V.G. DeMarco, A. Mahmood, et al., Differential regulation of cardiac function and intracardiac cytokines by rapamycin in healthy and diabetic rats, Oxid. Med. Cell. Longev. 2017 (2017), http://dx.doi.org/10.1155/2017/5724046 5724046. [31] F. Poosti, R. Bansal, S. Yazdani, et al., Interferon gamma peptidomimetic targeted to interstitial myofibroblasts attenuates renal fibrosis after unilateral ureteral obstruction in mice, Oncotarget 7 (34) (2016) 54240–54252. [32] S. Kolahian, I.E. Fernandez, O. Eickelberg, D. Hartl, Immune mechanisms in pulmonary fibrosis, Am. J. Respir. Cell Mol. Biol 55 (3) (2016) 309–322. [33] A. Sharma, N.R. Sinha, S. Siddiqui, R.R. Mohan, Role of 5'TG3'-interacting factors (TGIFs) in vorinostat (HDAC inhibitor)-mediated corneal fibrosis inhibition, Mol. Vis. 21 (8) (2015) 974–984. [34] J. Zhou, J. Zhong, S. Lin, et al., Inhibition of PTEN activity aggravates post renal fibrosis in mice with ischemia reperfusion-induced acute kidney injury, Cell. Physiol. Biochem. 43 (5) (2017) 1841–1854. [35] A. Guyard, C. Danel, N. Théou-Anton, et al., Morphologic and molecular study of lung cancers associated with idiopathic pulmonary fibrosis and other pulmonary fibroses, Respir. Res. 18 (1) (2017) 120. [36] H. Tao, J.G. Zhang, R.H. Qin, et al., LncRNA GAS5 controls cardiac fibroblast activation and fibrosis by targeting miR-21 via PTEN/MMP-2 signaling pathway, Toxicology 386 (7) (2017) 11–18. [37] S.K. Parapuram, K. Thompson, M. Tsang, et al., Loss of PTEN expression by mouse fibroblasts results in lung fibrosis through a CCN2-dependent mechanism, Matrix Biol. 43 (2015) 35–41. [38] L. Chen, K. Zhang, Z. Shi, et al., A lentivirus-mediated miR-23b sponge diminishes the malignant phenotype of glioma cells in vitro and in vivo, Oncol. Rep. 31 (4) (2014) 1573–1580. [39] W. He, H. Che, C. Jin, et al., Effects of miR-23b on hypoxia-induced cardiomyocytes apoptosis, Biomed. Pharmacother. 96 (2017) 812–817. [40] R.A. Bagchi, M.P. Czubryt, Synergistic roles of scleraxis and Smads in the regulation of collagen 1alpha2 gene expression, Biochim. Biophys. Acta 1823 (10) (2012) 1936–1944. [41] B. Amalakuhan, S.A. Habib, M. Mangat, et al., Endothelial adhesion molecules and multiple organ failure in patients with severe sepsis, Cytokine 88 (2016) 267–273. [42] D. Wotton, R.S. Lo, L.A. Swaby, J. Massagué, Multiple modes of repression by the Smad transcriptional corepressor TGIF, J. Biol. Chem. 274 (52) (1999) 37105–37110. [43] E. Granot, D. Shouval, Y. Ashur, Cell adhesion molecules and hyaluronic acid as markers of inflammation, fibrosis and response to antiviral therapy in chronic hepatitis C patients, Mediators. Inflamm. 10 (5) (2001) 253–258. [44] A.I. Othman, M.R. El-Sawi, M.A. El-Missiry, et al., Epigallocatechin-3-gallate protects against diabetic cardiomyopathy through modulating the cardiometabolic risk factors, oxidative stress, inflammation, cell death and fibrosis in streptozotocinnicotinamide-induced diabetic rats, Biomed. Pharmacother. 94 (2017) 362–373. [45] M. Gao, X. Wang, X. Zhang, et al., Attenuation of cardiac dysfunction in polymicrobial sepsis by microRNA-146a is mediated via targeting of IRAK1 and TRAF6 expression, J. Immunol. 195 (2) (2015) 672–682. [46] S. Phosri, A. Arieyawong, K. Bunrukchai, et al., Stimulation of adenosine A2B receptor inhibits endothelin-1-induced cardiac fibroblast proliferation and alphammooth muscle actin synthesis through the cAMP/Epac/PI3K/Akt-signaling pathway, Front. Pharmacol. 8 (6) (2017) 428. [47] A. Chopra, E. Tabdanov, H. Patel, et al., Cardiac myocyte remodeling mediated by N-cadherin-dependent mechanosensing, Am. J. Physiol. Heart Circ. Physiol. 300 (4) (2011) 1252–1266. [48] D. Deng, X. Li, C. Liu, et al., Systematic investigation on the turning point of overinflammation to immunosuppression in CLP mice model and their characteristics, Int. Immunopharmacol. 42 (1) (2017) 49–58. [49] H. Zhou, J. Qian, C. Li, et al., Attenuation of cardiac dysfunction by HSPA12B in endotoxin-induced sepsis in mice through a PI3K-dependent mechanism, Cardiovasc. Res. 89 (1) (2011) 109–118. [50] J.T. Park, M. Kato, H. Yuan, et al., FOG2 protein down-regulation by transforming growth factor-beta1-induced microRNA-200b/c leads to Akt kinase activation and glomerular mesangial hypertrophy related to diabetic nephropathy, J. Biol. Chem. 288 (31) (2013) 22469–22480. [51] D.F. Higgins, L.M. Ewart, E. Masterson, et al., BMP7-induced-Pten inhibits Akt and prevents renal fibrosis, Biochim. Biophys. Acta 1863 (12) (2017) 3095–3104. [52] R. Samarakoon, S. Helo, A.D. Dobberfuhl, et al., Loss of tumour suppressor PTEN expression in renal injury initiates SMAD3- and p53-dependent fibrotic responses, J. Pathol. 236 (4) (2015) 421–432.

[1] F. Venet, T. Rimmelé, G. Monneret, Management of sepsis-induced immunosuppression, Crit. Care Clin. 34 (1) (2018) 97–106. [2] F.B. Mayr, S. Yende, D.C. Angus, Epidemiology of severe sepsis, Virulence 5 (1) (2014) 4–11. [3] R. Sato, A. Kuriyama, T. Takada, et al., Prevalence and risk factors of sepsis-induced cardiomyopathy: a retrospective cohort study, Med. (Baltim.) 95 (39) (2016) e5031. [4] S. Yende, W. Linde-Zwirble, F. Mayr, et al., Risk of cardiovascular events in survivors of severe sepsis, Am. J. Respir. Crit. Care Med. 189 (9) (2014) 1065–1074. [5] C.K. Han, Y.C. Tien, H.D. Jine-Yuan, et al., Attenuation of the LPS-induced, ERKmediated upregulation of fibrosis-related factors FGF-2, uPA, MMP-2, and MMP-9 by carthamus tinctorius L in cardiomyoblasts, Environ. Toxicol. 32 (3) (2017) 754–763. [6] K. Tomita, M. Takashina, N. Mizuno, et al., Cardiac fibroblasts: contributory role in septic cardiac dysfunction, J. Surg. Res. 193 (2) (2015) 874–887. [7] G. Krenning, E.M. Zeisberg, R. Kalluri, The origin of fibroblasts and mechanism of cardiac fibrosis, J. Cell. Physiol. 225 (3) (2010) 631–637. [8] K.T. Weber, Y. Sun, S.K. Bhattacharya, et al., Myofibroblast-mediated mechanisms of pathological remodelling of the heart, Nat. Rev. Cardiol. 10 (1) (2013) 15–26. [9] S. Zhang, Q. Liu, J. Xiao, et al., Molecular validation of the precision-cut kidney slice (PCKS) model of renal fibrosis through assessment of TGF-beta1-induced Smad and p38/ERK signaling, Int. Immunopharmacol. 34 (5) (2016) 32–36. [10] A.K. Ghosh, S.E. Quaggin, D.E. Vaughan, Molecular basis of organ fibrosis: potential therapeutic approaches, Exp. Biol. Med. (Maywood) 238 (5) (2013) 461–481. [11] D. Pchejetski, C. Foussal, C. Alfarano, et al., Apelin prevents cardiac fibroblast activation and collagen production through inhibition of sphingosine kinase 1, Eur. Heart J. 33 (18) (2012) 2360–2369. [12] S. Itoh, D.P. Ten, Negative regulation of TGF-beta receptor/Smad signal transduction, Curr. Opin. Cell Biol. 19 (2) (2007) 176–184. [13] S.K. Parapuram, W.X. Shi, C. Elliott, et al., Loss of PTEN expression by dermal fibroblasts causes skin fibrosis, J. Invest. Dermatol. 131 (10) (2011) 1996–2003. [14] C. McClure, L. Brudecki, D.A. Ferguson, et al., MicroRNA 21 (miR-21) and miR181b couple with NFI-A to generate myeloid-derived suppressor cells and promote immunosuppression in late sepsis, Infect. Immun. 82 (9) (2014) 3816–3825. [15] Y.Y. Luan, C.F. Yin, Q.H. Qin, et al., Effect of regulatory T cells on promoting apoptosis of T lymphocyte and its regulatory mechanism in sepsis, J. Interferon Cytokine Res. 35 (12) (2015) 969–980. [16] H. Khalil, O. Kanisicak, V. Prasad, et al., Fibroblast-specific TGF-beta-Smad2/3 signaling underlies cardiac fibrosis, J. Clin. Invest. 127 (10) (2017) 3770–3783. [17] Y.C. Liu, M.M. Yu, S.T. Shou, Y.F. Chai, Sepsis-induced cardiomyopathy: mechanisms and treatments, Front. Immunol. 8 (2017) 1021. [18] D. Lagares, O. Busnadiego, R.A. García-Fernández, et al., Inhibition of focal adhesion kinase prevents experimental lung fibrosis and myofibroblast formation, Arthritis Rheum. 64 (5) (2012) 1653–1664. [19] E. van Rooij, E.N. Olson, MicroRNAs: powerful new regulators of heart disease and provocative therapeutic targets, J. Clin. Invest. 117 (9) (2007) 2369–2376. [20] A. Tuttolomondo, I. Simonetta, A. Pinto, MicroRNA and receptor mediated signaling pathways as potential therapeutic targets in heart failure, Expert Opin. Ther. Targets 20 (11) (2016) 1287–1300. [21] M. Wu, J.T. Gu, B. Yi, et al., microRNA-23b regulates the expression of inflammatory factors in vascular endothelial cells during sepsis, Exp. Ther. Med. 9 (4) (2015) 1125–1132. [22] S. Zhu, W. Pan, X. Song, et al., The microRNA miR-23b suppresses IL-17-associated autoimmune inflammation by targeting TAB2, TAB3 and IKK-α, Nat. Med. 18 (7) (2012) 1077–1086. [23] R. Hu, R.M. O’Connell, MiR-23b is a safeguard against autoimmunity, Nat. Med. 18 (7) (2012) 1009–1110. [24] H. Zhang, H. Li, A. Shaikh, et al., Inhibition of microRNA-23b attenuates immunosuppression in late sepsis through NIK, TRAF1 and XIAP, J. Infect. Dis. (2018), http://dx.doi.org/10.1093/infdis/jiy116. [25] C. Iaconetti, S. De Rosa, A. Polimeni, et al., Down-regulation of miR-23b induces phenotypic switching of vascular smooth muscle cells in vitro and in vivo, Cardiovasc. Res. 107 (4) (2015) 522–533. [26] C.E. Rogler, J.S. Matarlo, B. Kosmyna, et al., Knockdown of miR-23, miR-27, and miR-24 alters fetal liver development and blocks fibrosis in mice, Gene Expr. 17 (2) (2017) 99–114. [27] C. Li, Y. Bai, H. Liu, et al., Comparative study of microRNA profiling in keloid fibroblast and annotation of differential expressed microRNAs, Acta. Biochim. Biophys. Sin. (Shanghai) 45 (8) (2013) 692–699. [28] Y. Zhou, Y. Song, Z. Shaikh, et al., MicroRNA-155 attenuates late sepsis-induced cardiac dysfunction through JNK and beta-arrestin 2, Oncotarget 8 (29) (2017) 47317–47329.

878