Minocycline reverses IL-17A/TRAF3IP2-mediated p38 MAPK/NF-κB/iNOS/NO-dependent cardiomyocyte contractile depression and death

Abstract

Minocycline, an FDA-approved second-generation semisynthetic tetracycline, exerts antioxidant, anti-apoptotic and anti-inflammatory effects, independent of its antimicrobial properties. Interleukin (IL)-17A is an immune and inflammatory mediator, and its sustained induction is associated with various cardiovascular diseases. Here we investigated (i) whether IL-17A induces cardiomyocyte contractile depression and death, (ii) whether minocycline reverses IL-17A’s negative inotropic effects and (iii) investigated the underlying molecular mechanisms. Indeed, treatment with recombinant mouse IL-17A impaired adult cardiomyocyte contractility as evidenced by a 34% inhibition in maximal velocity of shortening and relengthening after 4 h (P<0.01). Contractile depression followed iNOS induction at 2 h (2.13-fold, P<0.01) and NO generation at 3 h (3.71-fold, P<0.01). Further mechanistic investigations revealed that IL-17A-dependent induction of iNOS occurred via TRAF3IP2, TRAF6, TAK1, NF-κB, and p38MAPK signaling. 1400W, a highly specific iNOS inhibitor, suppressed IL-17A-induced NO generation and contractile depression, where as the NO donors SNAP and PAPA-NONOate both suppressed cardiomyocyte contractility. IL-17A also stimulated cardiomyocyte IL-1β and TNF-α secretion, however, their neutralization failed to modulate IL-17A-mediated contractile depression or viability. Further increases of IL-17A concentration and the duration of exposure enhanced IL-1 β and TNF-α secreted levels, but had no impact on adult cardiomyocyte viability. However, when combined with pathophysiological concentrations of IL-1β or TNF-α, IL-17A promoted adult cardiomyocyte death. Importantly, minocycline blunted IL-17A-mediated deleterious effects, indicating its therapeutic potential in inflammatory cardiac diseases.

1. Introduction

The interleukin (IL)-17 family of proinflammatory cytokines contains six ligands (IL-17A, B, C, D, E, and F) and five receptors (IL-17RA, B, C, D and E) [1]. While the expression of ligands appears to be cell type-specific, the receptors are expressed ubiquitously, suggesting that many cell types are responsive to IL-17 via autocrine or paracrine mechanisms. Original reports suggested that IL-17A is expressed exclusively by a distinct subset of T cells, called Th₁₇. Later studies, however, demonstrated that various immune [2] and non-immune cells express IL-17A. Recently, bronchial epithelial cells were reported to constitutively express IL-17A, along with its key transcriptional regulator RORγt [3]. We and others have previously reported that adult mouse cardiac fibroblasts express IL-17A [4-6], indicating that IL-17A secreted from cardiac fibroblasts and other myocardial constituent and non-cardiac cells may affect cardiomyocyte phenotype and contractility in a paracrine manner.

Contractile depression is a hallmark of various cardiac diseases, and proinflammatory cytokines such as IL-1β and TNF-α exert cardiac depressant effects. IL-17A has been shown to induce these negative inotropes in cardiac cells, and the expression of IL-17, IL-1β and TNFαis upregulated during inflammatory cardiac diseases, including cardiac ischemic injury and myocarditis [7]. Further, both ischemic injury and myocarditis are characterized by increased oxidative stress, and IL-17 is known to induce oxidative stress and activate the redox-sensitive transcription factors NF-κB and AP-1 in myocardial constituent cells [4, 5]. IL-17-dependent oxidative stress involves, in part, upregulation of inducible isoform of nitric oxide synthase (iNOS, also known as NOS2; [8]) that is known to contribute to contractile depression through excess generation of NO. Indeed, iNOS activation elicits greater NO generation than the constitutively expressed endothelial (eNOS) and neuronal (nNOS) isoforms. During cardiac injury and inflammation, sustained pathological levels of NO depresses cardiomyocyte contractility[9], resulting in myocardial dysfunction. Therefore, we hypothesized that IL-17A induces cardiomyocyte contractile depression in an iNOS/NO-dependent manner. Since persistently increased NO levels also promote cell death [8], we further hypothesized that IL-17A induces cardiomyocyte death.

Mitogen-activated protein kinases (MAPKs) play a critical role in gene regulation and cardiomyocyte contractility [10]. We and others have previously reported that IL-17A activates all three MAPKs (ERK1/2, JNK and p38MAPK) in cardiac constituent cells [4, 11-13]. In general, while activation of ERK1/2 promotes cell survival, the role of JNK in cardiomyocyte contractility and survival is stimulus and context dependent. However, sustained activation of p38 MAPK is usually associated with cardiomyocyte contractile depression and sometimes death [14]. We have previously reported that IL-17A-dependent activation of MAPK kinases and NF-κB [15] involves physically association with the cytoplasmic adapter molecule TRAF3IP2 (TRAF3 Interacting Protein 2; also known as CIKS or Act1 [16, 17]). Since p38 MAPK plays a role in NF-κB activation and iNOS induction [18], we also hypothesized that IL-17A induces cardiomyocyte contractile depression and/or death via activation of the p38 MAPK/NF-κB/iNOS pathway. Critically, the FDA approved 2^nd generation tetracycline antibiotic minocycline (7-dimethylamino-6-dimethyl-6-deoxytetracycline) has been reported to exert antioxidant, anti-inflammatory and anti-apoptotic effects in vitro and in vivo, independent of its anti-microbial properties [19]. Accordingly, we further hypothesized that minocycline will inhibit IL-17A/TRAF3IP2-mediated proinflammatory signaling, cardiomyocyte contractile depression and/or death by targeting the iNOS/NO pathway.

Confirming the hypotheses, our mechanistic studies reveal that treatment of adult mouse cardiomyocytes with IL-17A alone induces contractile depression, but not death, via a p38MAPK/NF-κB/iNOS/NO pathway. However, in combination with inflammatory cytokines like IL-1β and TNF-α, IL-17A promotes cardiomyocyte death via activation of multiple pro-death pathways. Importantly, minocycline prevents IL-17A-induced contractile depression and death, indicating its therapeutic potential in inflammatory cardiac diseases.

2. Materials and methods

2.1. Reagents

Carrier-free recombinant mouse (rm) IL-17A (#421-ML-025), protein A or G purified monoclonal rat anti-mouse IL-17A IgG2A antibodies (#MAB4481; used in immunoblotting and neutralization), monoclonal rat anti-mouse IL-17RA/IL-17R IgG2A antibodies (#MAB4481-SP; used in immunob lotting and neutralization), Rat IgG2A Isotype control (#MAB006), antigen affinity-purified polyclonal goat anti-mouse IL-17RC IgG antibodies (#AF2270; used in immunoblotting and neutralization), monoclonal hamster anti-mouse IL-1β IgG antibodies used in immunoblotting and neutralization (#MAB4012; 1 μg/ml) and antigen affinity purified polyclonal goat anti-mouse TNF-α IgG antibodies (#AF-410-NA) were all purchased from R&D Systems (Minneapolis, MN). rmIL-17A contained <0.10 EU/1 μg of the protein as determined by the Limulus amebocyte lysate (LAL) assay (manufacturer’s technical data sheet). To rule out the potential contribution of trace amounts of endotoxin to the depressant effects of rmIL-17A, cardiomyocytes were pretreated with Polymyxin B sulphate (10 μg/ml in water for 2 h; #BP1028; Millipore Sigma) prior to IL-17A addition. Polymyxin B is a cyclic cationic decapeptide that binds lipid A in endotoxin and neutralizes its activity. Protein A/G purified normal hamster IgG Isotype control (#NBP2-21950) and rabbit polyclonal anti-mitofilin antibody (NB100-1919; 1:2000) were purchased from Novus Biologicals (Centennial, CO). Recombinant mouse (rm) IL-1β (#BMS332) and rmTNFα (#14-8321-63)were purchased from eBioscience/Invitrogen/Thermo Fisher Scientific (San Diego, CA). Both rmIL-1β and rmTNF-α contained trace amounts of endotoxin (<0.10 EU/1 μg of the protein as determined by the LAL assay; manufacturer’s technical data sheet) Antibodies against cleaved caspase-3 (#9661; 1:1000), NF-κBp65 (#3034; 1:1000), p38 MAPK (#9212; 1:1000), phospho- p38 MAPK (p-p38 MAPK Thr180/Tyr182, #9211; 1:1000), cleaved PARP (Asp214; #9544; 1:1000), and Bax (#2772; 1:1000) and (#2144; 1:1000) were purchased from Cell Signaling Technology, Inc. (Beverly, MA). Monoclonal antibody to Bcl-x_s/L that recognizes both long (Bcl-xL) and short (Bcl-xS) forms (#sc-70418; 1:200) was from Santa Cruz Biotechnology, Inc. (Dallas, TX). Rabbit polyclonal MyD88 antibody (#ab2064; 1:500), rabbit polyclonal anti-human TRAF2 antibodies that cross-react with mouse (ab244317; 0.25 μg/ml), rabbit monoclonal ASK1 antibody (#ab45178; 1:1000), rabbit polyclonal PARP1 antibody that recognized the proform (#ab227244; 1:5000), rabbit monoclonal cleaved PARP1 antibody (#ab32064; 1:5000), rabbit monoclonal anti-Cytochrome C antibodies (#ab133504; 1:5000), rabbit monoclonal anti-GAPDH antibody (#ab181602; 1:10000) and mouse monoclonal αTubulin antibodies (#ab7291; 1:5000) were purchased from Abcam (Cambridge, MA). Pierce™ BCA Protein Assay Kit (#23227) and SuperSignal® Westo Femto Maximum Sensitivity Substrate (#34096) were purchased from Thermo Fisher Scientific (Waltham, MA). The NO donor PAPA-NONOate (Z)-1-[N-(3-aminopropyl)-N-(n-propyl)amino]diazen-1-ium-1,2-diolate; Item #82140; 1 μM in water for 4 h) was from Cayman Chemical (Ann Arbor, MI). 1400W 2HCl (N-(3-(Aminomethyl)benzyl)acetamidine), a slow, tight binding, and highly selective iNOS inhibitor (#S8337; 1μM in water for 1 h) [20] and the potent p38 MAPK inhibitor SB239063 (#S7741; 10 μM in DMSO for 1h) were purchased from Selleckchem (Houston, TX). SNAP (S-Nitroso-N-acetyl-DL-penicillamine, #N3398; 1 mM in DMSO for 4 h), Doxorubicin hydrochloride (#D1515; 20 μM for 8 h) and all other chemicals were purchased from Millipore Sigma (St. Louis, MO).

2.2. Animals

This investigation conforms to the Guide for the Care and Use of Laboratory Animals, published by the National Institutes of Health. All protocols were approved by the Institutional Animal Care and Use Committees at UT Health, San Antonio TX and Tulane University School of Medicine, New Orleans, LA. Wild type C57Bl/6 mice (WT) and iNOS knockout mice (iNOS-null B6.129P2-Nos2^tm1Lau/J; Stock# 002609) were purchased from The Jackson Laboratory (Ann Harbor, MI). The TRAF3IP2-null mice (C57Bl/6 background) are previously described [12, 21]. Absence of TRAF3IP2 expression in cardiomyocytes isolated from TRAF3IP2-null mice was confirmed by immunoblotting (Fig. 1D, right hand lower panel).

Figure 1. IL-17A depresses cardiomyocyte contractility in part via TRAF3IP2.

A. IL-17A impairs shortening and re-lengthening of cardiomyocytes in a dose-dependent manner. Calcium-tolerant adult mouse cardiomyocytes cultured after a 12 h in 0.5% BSA were incubated rmIL-17A at the indicated concentrations for 4 h, and analyzed for maximal velocity of shortening (+dL/dt) and relengthening (−dL/dt) using IonOptix video-based edge-detection system. Eighteen individual cardiomyocytes were used/treatment. *P<0.01 vs. untreated (n=18). B, Endotoxin inactivation by polymyxin B fails to affect IL-17A-induced impairment in maximal velocity of shortening ((+dL/dt). Cardiomyocytes described as in A were incubated with polymyxin B (10 μg/ml in water for 2 h) prior to IL-17A addition (50 ng/ml for 4 h). *P<0.001 vs. untreated (n=18). C, Specificity of IL-17A. Specificity of IL-17A on contractile depression was verified by incubating cardiomyocytes with neutralizing IL-17A and IL-17R antibodies (10 mg/ml for 1 h) prior to IL-17A addition (50 ng/ml for 4 h). Normal rat IgG2 or normal goat IgG served as isotype controls. The maximal velocity of shortening was measured as in A. *P<0.01 vs. untreated, †P0.01 vs. IL-17A (n=12). D. TRAF3IP2 knockdown or gene deletion and minocycline each blunt IL-17A-induced impairment in maximal velocity of shortening. Cardiomyocytes from wild type mice (WT) transduced with adenoviral TRAF3IP2 shRNA or control GFP shRNA (100 moi for 36h) or cardiomyocytes isolated from TRAF3IP2-null mice or WT-cardiomyocytes treated with minocycline (10 μM in water for 15 min) were incubated with IL-17A (50 ng/ml for 4 h) and then analyzed for maximal velocity of shortening. Knockdown of TRAF3IP2 in WT cardiomyocytes or lack of TRAF3IP2 isolated from TRAF3IP2-null mice is confirmed by immunob lotting as shown on the right. ASK1 and MyD88 served as off-targets in shRNA-mediated knockdown experiments. *P< 0.01 vs. untreated, †P<0.01 vs. IL-17A (n=12). NS: Not significant.

2.3. Isolation of adult mouse cardiomyocytes (ACM)

Calcium-tolerant adult mouse ventricular myocytes (cardiomyocytes or CM) were isolated from 3 month-old wild type (WT) male C57Bl/6 mice, TRAF3IP2-null mice and iNOS knockout mice as previously described [12, 22]. In brief, mice were given heparin (1000 units/kg intraperitoneally) and then deeply anesthetized with intramuscular ketamine/xylazine. Following median sternotomy, the heart was rapidly excised and rinsed with physiologic saline. The aortic lumen was isolated and tied to an 18-gauge cannula, and the heart was perfused with oxygenated (95% O₂, 5% CO₂), Ca²⁺-free modified Tyrode's bicarbonate buffer (126 mM NaCl, 4.4 mM KCl, 1 mM MgCl₂, 18 mM NaHCO₃, 11 mM glucose, 4 mM HEPES, 10 mM 2,3-butanedione monoxime, 30 mM taurine, pH 7.35) at 37 °C for 5 min. Following this, the heart was perfused with 50 ml of digestion buffer (modified Tyrode's bicarbonate buffer with 0.25 mg/ml Liberase Blendzyme type 1 (Roche Applied Science), 0.14 mg/ml trypsin (Sigma), and CaCl₂ 2.5 μM) in a recirculating fashion for 12–15 min. The heart was then removed, and the left ventricle (LV) was separated and dissected with small, blunt forceps in 2–3 ml of digestion buffer. The minced tissue suspension was gently agitated by repeated pipette aspiration and transferred into myocyte stopping buffer 1 (modified Tyrode's bicarbonate buffer with 10% fetal calf serum and 12.5 μM CaCl₂) in a 50-ml conical tube and allowed to sediment for 10 min. The supernatant was transferred to another tube and centrifuged at 90 × g for 2 min. The sediment/pellet from both tubes were combined and resuspended in myocyte stopping buffer 2 (modified Tyrode's bicarbonate buffer with 5% fetal calf serum and 12.5 μM CaCl₂) in a 100-mm culture dish. Small aliquots of CaCl₂ were then added in a graded fashion at 4-min intervals to sequentially increase the Ca²⁺ concentration to 500 μM (five total steps). The suspension was then placed in a 50-ml conical tube and allowed to sediment for 10 min at 22 °C. As above, the supernatant was transferred to another tube and gently centrifuged at 90 × g for 2 min, and myocytes contained in both the sediment and the pellet were combined and resuspended in minimal essential medium (pH 7.35–7.45; catalogue number M1018; Sigma) containing 1.2 mM Ca²⁺, 12 mm NaHCO₃, 2.5% fetal bovine serum, and 1% penicillin/streptomycin. After examining under light microscopy, preparations containing ≥98% viable myocytes (trypan blue dye exclusion) were used in the proposed studies. Equal number of viable cells were plated onto 35-mm cell culture dishes pre-coated with 20 μg/ml mouse laminin in PBS with 1% penicillin/streptomycin for 1 h. Cardiomyocytes were maintained under resting conditions in the incubator for 12 h before experimentation. All studies were completed within 72 h after isolation.

2.4. Lenti and adenoviral transduction

Adenoviral vector expressing human TRAF3IP2 shRNA (Ad.GFP-U6-h-TRAF3IP2-shRNA) was custom generated at Vector Biolabs (Malvern, PA). Validated lentiviral shRNA against mouse TRAF6 (SHCLNV-NM_009424, TRCN0000040735), TAK1 (SHCKNV-NM_172688, TRCN0000022563) and p65 (TRCN 0000055346) were purchased from Sigma-Aldrich (St. Louis, MO, USA), and were previously described [23]. Lentiviral shRNA against IKKβ and GFP were purchased from Santa Cruz Biotechnology, Inc., and have been previously described [23]. Though transfection with adenoviral or lentiviral vectors slightly increased the number of rounded cells, no cell death was observed as determined by trypan blue dye exclusion. Only rod shaped cardiomyocytes were used in functional studies.

2.5. Cardiomyocyte contractility

Cardiomyocyte shortening (+dL/dt) and relengthening (−dL/dt) were measured on the stage of an inverted phase-contrast microscope (Nikon Eclipse TE200) using an optical-video system (IonOptix) in which the analogue motion signal was digitized and analyzed by computer [24, 25]. Cover slips with attached cardiomyocytes, cardiomyocytes in which TRAF3IP2, TRAF6, TAK1, IKKβ or p65 is silenced were placed in a temperature-controlled chamber at 370°C (total volume 260 μl), and superfused with a buffer containing IL-17A (50 ng/ml). Similarly, cardiomyocytes were incubated with neutralizing antibodies against IL-17A, IL-1β or TNF-α prior to the addition of IL-17A. p38 MAPK was targeted pharmacologically using SB239063. DMSO served as a solvent control, and never exceeded 0.1%. As described earlier [24, 25], shortening (+dL/dt) and relengthening (−dL/dt) were assessed based on peak shortening, time-to- peak shortening, time-to-90% relengthening, and maximal velocities of shortening and relengthening. IL-17A was added after achieving steady-state spontaneous contraction.

2.6. mRNA expression

IL-1β (Assay ID; Mm00434228_m1), TNF-α (Assay ID: Mm00443258_m1) and iNOS (Mm00440502_m1) mRNA expression was analyzed by real time quantitative PCR (RT-qPCR) using best coverage TaqMan® Gene Expression Assays. All data were normalized to corresponding 18S rRNA (Assay ID: Hs03003631_g1), and expressed as fold difference in gene expression compared to untreated controls.

2.7. Immunoblotting

Extraction of protein homogenates, Western blot analysis, autoradiography, and densitometry were performed as described previously [4, 5, 11, 23, 26]. Briefly, myocardial tissue was homogenized in a hypotonic lysis buffer containing 10 mM Tris (pH 7.5), 1 mM EDTA, 1 mM Na₃ VO₄, 10 mM sodium pyrophosphate, 10 mM β-glycerophosphate, 10 mM NaF, and 1% protease inhibitor cocktail (Sigma, St. Louis, MO). Protein concentrations were determined using Pierce™ BCA Protein Assay Kit and boiled in SDS sample buffer. Samples were then loaded onto hand-cast 10% polyacrylamide gels using a Bio-Rad Mini Protean 3 system. After electrophoresis, proteins were transferred on to polyvinylidene difluoride membrane and probed with primary antibodies diluted in 2% nonfat dry milk in Tris-buffered saline containing 0.05% Tween 20 (TTBS). After incubation overnight at 4°C, blots were rinsed in TTBS for 30 min and incubated in horseradish peroxidase-conjugated secondary antibodies in 5% milk in TTBS for 1 h. The blots were washed three times (5 min per wash) in TBST, and immunoreactive bands were developed using enhanced chemiluminescent substrate (SuperSignal West Femto Maximun Sensitivity). The antibodies used in the current studies and their dilutions are provided in the materials section. While a majority of proteins were detected using cleared whole cell lysates, the amount of Bax and Cytochrome C were analyzed in cytoplasmic and mitochondrial fractions as previously described [12, 27]. Tubulin, GAPDH, unphosphorylated respective proteins, and Mitofilin served as loading controls.

2.8. iNOS enzyme activity

iNOS (NOS2; Ca²⁺ -independent NOS) enzymatic activity was determined from the extent of conversion of L-³H-arginine to L-³H-citrulline in the presence of Ca²⁺chelators, and has been previously described [28]. After homogenizing cardiomyocytes in an ice-cold buffer containing 25 mM Tris.HCl, 1 mM EDTA and 1 mM EGTA, 1 mM dithiothreitol, 100 μg/ml PMSF, 10 μg/ml leupeptin, 10 μg/ml soybean trypsin inhibitor, and 2 μg/ml aprotinin at pH 7.4, the homogenate was spun at 100 000 g for 30 min, and the supernatant was used for the assay. Protein concentration was determined by the Pierce BCA Protein Assay Kit using bovine serum albumin (BSA) as a standard. The reaction was initiated by incubating 10 μg of supernatant with 80 nm³H-arginine (Amersham, Arlington Heights, IL) for 30 min at 25°C and pH 7.4 in a solution of 50 mM Tris.HCl, 1 mM EGTA, 1 mM DTT, 1 mM NADPH (freshly prepared), 100 μMBH₄ ((6R)-5,6,7,8-tetrahydrobiopterin) and 10 μM FAD. After the addition of 0.5 ml buffer containing 20 mM HEPES pH 5.5, and 2 mM EGTA, the mixture was applied to a 1-ml column (DOWEX AG 50 W-X8, Na⁺-form; BioRad, Hercules, CA) pre-equilibrated with the buffer containing 20 mM HEPES pH 5.5 and 2 mM EGTA. L-2,3,4,5-³H-citrulline was eluted twice with 0.5 ml of distilled water. Radioactivity of this 1 ml eluate was determined by liquid scintillation counting. The NOS2 enzyme activity was expressed in pmol/mg of protein/h.

2.9. Analysis of nitrite concentrations

Nitrite formation in culture supernatants served as an indicator of NO release by cardiomyocytes. After centrifuging culture supernatants at 3000 rpm for 5 min at 4°C to remove cell debris, nitrite levels were measured as previously described [29] by mixing 100 μl of medium with 1 ml Griess reagent (1% sulfanilamide and 0.1% naphthylethylenediamine in 5% H₃PO₄). The concentration of the resultant chromophore was determined spectrophotometrically at 543 nm using known concentrations of sodium nitrite as a standard. Protein concentration in the cells was measured by the BCA protein assay. Nitrite concentrations were expressed as nmol/mg cell protein.

2.10. Cell death analysis

To determine whether IL-17A treatment, transduction of viral vectors, and pharmacological inhibitors modulate cell viability, cell death was analyzed using the Cell Death Detection ELISA^PLUS, trypan blue dye exclusion, and microscopic visualization of cell shape and for cells floating in the media. In addition, the protein levels of Bcl-2, Bcl-X_L and X_s, cleaved PARP, and cleaved caspase-3 levels in cleared whole cell homogenenates and Bax and Cytochrome C levels in cytoplasmic and mitochondrial fractions were analyzed by immunoblotting. Doxorubicin (Dox., 20 μM in water) served as a positive control. All cell death assays and the use of Dox. have been previously described [27].

2.11. Statistical analysis

Results are expressed as the mean ± SE. Differences in outcomes were determined using one-way ANOVA with post-hoc Dunnett's t-tests were considered significant when P<0.05. The assays were performed three independent times, and are not replicates from the same mouse. Further, although representative immunoblots are shown in select figures, the intensity of immunoreactive bands from three independent experiments were semi quantified by densitometry and summarized in accompanying bar graphs at the bottom/side/inset of respective panels and displayed as ratios or fold changes from untreated or respective controls. The numbers at the bottom of each panel in figures denote lane numbers.

3. Results

3.1. Minocycline reverses IL-17A-mediated cardiomyocyte contractile depression

IL-17A is a proinflammatory cytokine belonging to a unique cytokine family. Since cardiomyocyte contractile dysfunction and death are part and parcel of adverse cardiac remodeling under various stress conditions, we investigated whether IL-17A depresses cardiomyocyte contractility. Indeed, analysis of cardiomyocyte mechanical properties revealed that IL-17A depressed maximal velocity of shortening (+dL/dt) and relengthening (−dL/dt) at 4h, with peak levels of impairment observed at 50 ng/ml at 4 h (Fig. 1A). Therefore, in all subsequent experiments, rmIL-17A was used at 50 ng/ml, and only data pertaining to shortening (+dL/dt) only is presented. rmIL-17A contained trace amounts of endotoxin (0.1 EU/μg of protein), and endotoxin is known Together depress cardiomyocyte contractility. Therefore, to rule out the potential contribution of endotoxin to IL-17A’s depressant effects, in a subset of experiments, cardiomyocytes were pretreated with Polymyxin B sulphate (10 μg/ml for 2 h) prior to IL-17A. This pretreatment did not alter the depressant effects of IL-17A, indicating no significant effect of the low levels of endotoxin present in the rmIL-17A preparations (Fig. 1B). Importantly, preincubation with antibodies against IL-17A, IL-17RA or IL-17RC, whose neutralizing properties we have previously demonstrated [26], markedly attenuated IL-17A - mediated cardiomyocyte depressant effects, demonstrating specificity of IL-17A (Fig. 1C). Since IL-17A predominantly signals via TRAF3IP2, we next determined whether IL-17A depresses cardiomyocyte contractility via TRAF3IP2 using two independent approaches: 1. Knockdown of TRAF3IP2 using an adenoviral vector expressing human TRAF3IP2 shRNA, and 2. Cardiomyocytes isolated from adult TRAF3IP2-null mice. The coding sequence of human and mouse TRAF3IP2 shows more than 90% homology, and our adenoviral vector caused more than 80% knockdown in TRAF3IP2 expression in adult mouse cardiomyocytes without inducing cell death (Fig. 1D, upper right hand panel). Lack of TRAF3IP2 expression in cardiomyocytes isolated from TRAF3IP2-null mice was confirmed by immunoblotting (Fig. 1D, lower right hand panel). Under both TRAF3IP2 silencing conditions, IL-17A-induced cardiomyocyte contractile depression was markedly blunted (Fig. 1D). Since minocycline exerts anti-inflammatory effects, we next investigated whether minocycline reverses IL-17A-induced contractile depression. Pretreatment with minocycline at 10 μM for 15 min prior to IL-17A addition prevented IL-17A’s depressant effects (Fig. 1D). Together, these results indicate that IL-17A is a cardiomyocyte contractile depressant, and its depressant effects are mediated via TRAF3IP2. Moreover, minocycline can prevent these depressant effects (Fig. 1).

3.2. Minocycline inhibits IL-17A-mediated TRAF3IP2, TRAF6, TAK1, IKK-dependent NF-κB activation

We have demonstrated that IL-17A depresses cardiomyocyte contractility in part via TRAF3IP2 (Fig. 1). Previously, Leonardi et al. demonstrated that TRAF3IP2 physically associates with IKKβ though the HLH domain, contributing to agonist-induced NF-κB activation [16]. Structurally, TRAF3IP2 contains two TRAF binding sites, and mediates NF-κB activation via TRAF6 and IKK signaling [30, 31]. TRAF6 is also known to activate NF-κB via TAK1 [32], indicating that IL-17/TRAF3IP2 interaction leads to NF-κB activation via multiple mechanisms. Therefore, we next investigated the roles of TRAF6, TAK1, IKKβ and p65 in IL-17A-mediated cardiomyocyte contractile depression. To that end, we demonstrated that IL-17A activates IKKβ, an effect markedly inhibited by an anti-IL-17A neutralizing antibody (Fig. 2A). Moreover, silencing TRAF3IP2, TRAF6 and TAK1 significantly inhibited IKKβ activation, as did pretreatment with minocycline (Fig. 2B). Further, IL-17A induced NF-κB activation, evidenced by a marked increase in phospho-p65 levels, was significantly inhibited by silencing TRAF3IP2, TRAF6, TAK1 and IKKβ (Fig. 2C), as well as by minocycline pretreatment (Fig. 2D). Furthermore, silencing TRAF6, TAK1, IKKβ and p65 reversed IL-17A-mediated cardiomyocyte contractile dysfunction (Fig. 2E). Together, these data indicate that IL-17A depresses cardiomyocyte contractility in part via TRAF6, TAK1, IKKβ and NF-κB (Fig. 2).

Figure 2. IL-17A impairs cardiomyocyte contractility in part via TRAF3IP2, TRAF6, TAK, IKKβ-dependent NF-κB activation.

A, IL-17A induces IKKβ activation. Cardiomyocytes were incubated with IL-17A neutralizing antibodies prior to IL-17A addition (50 ng/ml for 1 h). Total and phospho-IKKβ levels were analyzed by immunoblotting using activation-specific antibodies. Total IKKβ served as a loading control. Normal rat IgG2A served as the isotype control. *P<0.05 vs. untreated, †P<0.05 vs. IL-17A (n=3). B, IL-17A induces IKKβ activation in part via TRAF3IP2, TRAF6, and TAK1, and is inhibited by minocycline. Cardiomyocytes transduced with lentiviral shRNA against TRAF3IP2, TRAF6, and TAK1 (moi0.5 for 36 h) or treated with minocycline (10 μM in water for 15 min) prior to IL-17A addition (50 ng/ml for 1 h). IKKβ activation was analyzed as in A. Knockdown of TRAF6 and TAK1 was confirmed by immunoblotting, and is shown on the right. ASK1 served an off-target. *P<0.05 vs. untreated, †P<0.05 vs. IL-17A (n=3). C, D, IL-17A induces NF-κB activation in part via TRAF3IP2, TRAF6, TAK1 and IKKβ, and is inhibited by minocycline. Cardiomyocytes transduced with lentiviral shRNA against TRAF3IP2, TRAF6, TAK1 or IKKβ (moi0.5 for 36 h; C) or treated with minocycline (10 μM in water for 15 min; D) prior to IL-17A addition (50 ng/ml for 1 h) were analyzed for NF-κB activation by immunoblotting using antibodies that specifically detect phospho-p65. *P<0.05 vs. untreated, †P<0.05 vs. IL-17A (n=3). E, Silencing TRAF6, TAK1, IKKβ and p65 reverse IL-17A-mediated impairment in contractile function. Cardiomyocytes treated as in C and D and incubated with IL-17A (50 ng/ml for 4 h) were analyzed for maximal velocity of shortening. p65 knockdown was confirmed by immunoblotting and is shown on the right. *P<0.001 vs. untreated, †P< 0.01 vs. IL-17A (n=12). Bar graphs at the bottom of panels in A-D represent densitometric analyses from three independent experiments.

3.3. Minocycline inhibits IL-17A-mediated TRAF3IP2, TRAF6, TAK1, IKK-dependent p38 MAPK activation

We have demonstrated that IL-17A induces NF-κB activation in adult cardiomyocytes in part via TRAF3IP2, TRAF6, TAK1, and IKKβ (Fig. 2). In addition to NF-κB activation, TRAF6 and TAK1 also serve as upstream regulators of stress-activated protein kinases, including p38 MAPK activation. In fact, TGF-β1-TAK1-p38 MAPK signaling pathway in cardiomyocytes from non-infarcted cardiac tissue has been shown to contribute to left ventricular remodeling after myocardial infarction [33]. Therefore, we next investigated whether IL-17A induces p38 MAPK in adult cardiomyocytes. To that end, we demonstrate that IL-17A induces p38 MAPK activation (increased phospho-p38 at Thr180/Tyr182) in adult cardiomyocytes in a time dependent manner with peak phosphorylation detected at 30 min (Fig. 3A), an effect markedly attenuated by preincubation with anti-IL-17A neutralizing antibody and SB239063, a specific inhibitor of p38 MAPK (Fig. 3B). Moreover, pretreatment minocycline markedly suppressed IL-17A-induced p38 MAPK activation (reduced phospho-p38 levels; Fig. 3B). Since IL-17A signals mainly via TRAF3IP2, and as TRAF3IP2 physically interacts with TRAF6 to activate some downstream signaling intermediates, we next investigated whether IL-17A induced p38 MAPK activation via TRAF3IP2 and TRAF6. Accordingly, our data reveal that silencing TRAF3IP2 and TRAF6 each inhibited IL-17A-mediated p38 MAPK activation (Fig. 3C). Since TAK1 acts both upstream and downstream of p38 MAPK [34], we next silenced TAK1 and analyzed p38 MAPK activation. Our data show that TAK1 knockdown markedly suppressed IL-17A-mediated p38 MAPK activation (Fig. 3D), indicating that IL-17A induces p38 MAPK via TRAF3IP2, TRAF6 and TAK1 signaling. Importantly, inhibition of p38 MAPK with SB239063 partially reversed IL-17A-mediated cardiomyocyte contractile depression (Fig. 3E). Since IL-17A stimulates MMP1 expression in primary cardiac fibroblasts via p38 MAPK-dependent NF-κB and AP-1 activation and CRP expression in hepatocytes and smooth muscle cells via p38 MAPK-dependent NF-κB activation [4, 13], we next determined whether targeting p38 MAPK inhibits IL-17A-mediated NF-κB activation. Indeed pharmacological inhibition of p38 MAPK blunts IL-17A-mediated NF-κB activation (Fig. 3F). Together these data indicate that p38 MAPK is a critical signaling intermediate that contributes to IL-17A-mediated contractile depression, and minocycline inhibits IL-17A-mediated p38 MAPK activation (Fig. 3).

Figure 3. IL-17A impairs cardiomyocyte contractility in part via TRAF3IP2, TRAF6, and TAK-dependent p38 MAPK activation.

A, IL-17A induces p38 MAPK activation. Cardiomyocytes were incubated with IL-17A (50 ng/ml) for the indicated time periods. Phospho-p38 MAPK (p-p38) levels were analyzed by immunoblotting using activation-specific antibodies. Total p38 MAPK (p38) served as a loading control. *P< 0.05 vs. untreated (n=3). B, SB239063 and minocycline inhibit IL-17A-induced p38 MAPK activation. Cardiomyocytes were pretreated with neutralizing anti-IL-17A antibodies (10 mg/ml for 1 h) or p38 MAPK inhibitor SB239063 (10 μM in DMSO for 1h) or minocycline (10 μM in water for 15 min) prior to IL-17A addition (50 ng/ml for 1 h). p38 MAPK activation was analyzed as in A. *P<0.05 vs. untreated, †P<0.05 vs. IL-17A (n=3). C, D, IL-17A induces p38 MAPK activation in part via TRAF3IP2, TRAF6, and TAK1. Cardiomyocytes transduced with lentiviral shRNA against TRAF3IP2, TRAF6 (C), and TAK1 (D; moi0.5 for 36 h) prior to IL-17A addition (50 ng/ml for 1 h). p38 MAPK activation was analyzed as in A. Knockdown of TAK1 is shown on the right. ASK1 served an off-target. In TRAF6 knockdown experiments, TRAF2 served as an off-target. *P<0.05 vs. untreated, †P<0.05 vs. IL-17A (n=3). E, Targeting p38 MAPK reverses IL-17A-mediated impairment in contractile function. Cardiomyocytes were treated with SB239063 (10 μM in DMSO for 1h) prior to IL-17A addition (50 ng/ml for 4 h) were analyzed for maximal velocity of shortening. *P<0.001 vs. untreated, †P<0.01 vs. IL-17A (n= 12). F, IL-7A induces NF-κB activation in part via p38 MAPK. Cardiomyocytes were treated with SB239063 (10 μM in DMSO for 1h) prior to IL-17A addition (50 ng/ml for 1 h) were analyzed for NF-κB activation by immunoblotting using antibodies that specifically detect phosphor-p65 levels. Total p65 served as a loading control. *P<0.05 vs. untreated, †P<0.05 vs. IL-17A (n=3). Bar graphs at the bottom or right side of panels A-D and F represent densitometric analyses from three independent experiments.

3.4. Minocycline inhibits IL-17A-induced iNOS expression

Persistently increased nitric oxide generation detrimentally impacts cardiomyocyte function and survival. Among the three isoforms of nitric oxide synthase (NOS), iNOS, also known as NOS2, has been shown to contribute to excess generation of NO, and NO has been shown to reduce cardiomyocyte contractility [35]. IL-17A has been shown previously to induce cardiomyocyte death via Stat3-iNOS pathway [8]. Since IL-17A induced NF-κB and p38 MAPK activation, and as both NF-κB and p38 MAPK transcriptionally upregulate iNOS [36, 37], we next investigated whether IL-17A induces cardiomyocyte contractile dysfunction via iNOS-mediated NO generation, and whether minocycline blunts this effect. Our data reveal that, indeed, IL-17A induced iNOS RNA expression in a time-dependent manner with peak levels detected at 3 h (Fig. 4A). Similarly, IL-17A induced iNOS protein expression (Fig. 4B), enzyme activity (Fig. 4C) and nitrite levels (Fig. 4D). Further, silencing TRAF3IP2, TRAF6, TAK1, IKKβ, and p65 markedly inhibited IL-17A-induced iNOS mRNA upregulation. Moreover, pharmacological inhibition of p38 MAPK and minocycline both inhibited IL-17A-mediated mRNA expression (Fig. 4E). Importantly, pre-treatment with the iNOS-specific inhibitor 1400W reversed IL-17A-induced cardiomyocyte contractile depression (Fig. 4F) by significantly inhibiting NO generation (Fig. 4F, right hand panel). iNOS gene deletion similarly protected cardiomyocytes from IL-17A-induced impairment in contractility (Fig. 4F, green bar). As a proof-of-concept, we next analyzed whether NO donors mimic the contractile depression induced by IL-17A-iNOS-NO signaling. Accordingly, we incubated cardiomyocytes isolated from WT mice with the NO donors SNAP or PAPA-NONOate, and both NO donors at the 1 mM concentration [38] depressed cardiomyocyte contractility (Fig. 4G). However, unlike the positive control doxorubicin (Dox., 20 μM for 8 h), the NO donors did not alter cardiomyocyte viability, assessed by cleaved caspase-3 levels (Fig. 4G, right hand panel). Together, these results indicate that IL-17A depresses cardiomyocyte contractility in part via iNOS-mediated NO generation, and this effect is inhibited by minocycline (Fig. 4).

Figure 4. IL-17A impairs cardiomyocyte contractility in part via TRAF3IP2, TRAF6, TAK1, IKKβ, p65 and p38 MAPK-mediated iNOS expression and nitric oxide generation.

A, B, IL-17A induces iNOS mRNA and protein expression. Cardiomyocytes treated with IL-17A (50 ng/ml) for the indicated time periods were analyzed for iNOS mRNA expression by RT-qPCR using a validated Taqman® probe (A; n=6) and protein levels by immunoblotting using whole cell lysates (B; n=3). A, B, *P< 0,05 vs. untreated. C, IL-17A increases iNOS enzyme activity and nitric oxide production. Cardiomyocytes treated with IL-17A (50 ng/ml) for 3 h were analyzed for iNOS enzyme activity by the extent of conversion of L-[³H]arginine to L-[³H]citrulline in the presence of Ca²⁺chelators. *P<0.001 vs. untreated (n=6). D, IL-17A stimulates NO generation. Cardiomyocytes were treated with IL-17A (50 ng/ml) for 3 h. Nitrite levels in culture supernatants were analyzed using Griess reagent. *P<0.001 vs. untreated (n=6). E, IL-17A induces iNOS mRNA expression in part via TRAF3IP2, TRAF6, TAK1, IKKβ, p65 and p38 MAPK, and is inhibited by minocycline. Cardiomyocytes transduced with lentiviral shRNA against TRAF3IP2, TRAF6, TAK1, IKKβ or p65 (moi0.5 for 36 h) or incubated with SB239063 (SB; 10 μM in DMSO for 1h) or minocycline (10 μM in water for 15 min) prior to IL-17A addition (50 ng/ml for 3 h). iNOS mRNA expression was analyzed as in A. *P< 0.01 vs. untreated, †P< 0.05 vs. IL-17A (n=6). F, IL-17A impairs cardiomyocyte contractility via iNOS and NO. Cardiomyocytes isolated from WT mice treated with the selective iNOS inhibitor 1400W.2HCl (1μM in water for 1 h) prior to IL-17A addition (50 ng/ml for 4 h) and analyzed for contractile depression (n=12). Inhibition of NO generation by 1400W was confirmed by reduced nitrite levels, and is shown on the right (n=6). In a subset of experiments, cardiomyocytes isolated from iNOS-null mice incubated with IL-17A (50 ng/ml for 4 h) were analyzed for contractility (n=12). *P<0.001 vs. untreated, †P<0.01 vs. IL-17A. G, NO donors depress cardiomyocyte contractility. Cardiomyocytes isolated from WT mice were incubated with freshly prepared NO donors SNAP (1 mM in DMSO) or PAPA-NONOate(1 mM in water) for 4h, and analyzed for contractile depression (n=12). The right side panel shows cardiomyocyte viability at 8 h analyzed by immunoblotting for cleaved caspase-3 levels (n=3). Doxorubicin (Dox.; 20 μM for 8 h) served as a positive control. The immunoreactive bands were semiquantified by densitometry, and represented as a ratio of cleaved caspase-3 to respective total caspase-3 levels. *P< at least 0.01 vs. untreated.

3.5. Minocycline inhibits IL-17A-induced proinflammatory expression in cardiomyocytes

Proinflammatory cytokines such as IL-1β and TNF-α have been shown previously to depress contractile function in whole animals, isolated hearts and isolated cardiomyocytes [39-42]. Since IL-17A is known to induce inflammatory cytokine expression, we next investigated whether IL-17A-induced cardiomyocyte contractile depression is mediated via IL-1β and TNF-α. At first, we examined whether IL-17A induces their expression in cardiomyocytes. The results show that IL-17A induced IL-1β mRNA expression (Fig. 5A), protein expression (Fig. 5A, inset) and secretion (Fig. 5B), effects that were markedly inhibited by minocycline pretreatment. Similarly, IL-17A induced TNF-α mRNA expression (Fig. 5C), protein expression (Fig. 5C, inset) and secretion (Fig. 5D), effects that were also markedly inhibited by minocycline pretreatment. However, preincubation with neutralizing IL-1β or TNF-α antibodies failed to modulate IL-17A-induced iNOS mRNA expression (Fig. 5E), protein levels (Fig. 5F) and enzyme activity (Fig. 5G). Similarly, neutralizing IL-1β [4] and TNF-α [26] each failed to affect IL-17A-mediated increases in nitrite levels (Fig. 5H) and cardiomyocyte contractile depression (Fig. 5I). Further, as opposed to the positive control doxorubicin (Dox.), IL-17A failed to induce cardiomyocyte death, as evidenced by no significant changes in cleaved caspase-3 levels (Fig. 5J). Together, these results indicate that IL-17A stimulates IL-1β and TNF-α expression and secretion, but impairs cardiomyocyte contractility independent of their expression (Fig. 5).

Figure 5. IL-17A induces iNOS expression and impairs cardiomyocyte contractility independent of secreted IL-1β and TNF-α.

A-D, Minocycline inhibits IL-17A-induced IL-1β and TNF-α expression and secretion. Cardiomyocytes were treated with minocycline (10 μM water for 15 min) prior to IL-17A (50 ng/ml) addition. IL-1β and TNF-α mRNA expression (1h; A, C) was analyzed by RT-qPCR (n=6). Protein levels in whole cell lysates were analyzed by immunoblotting (1h; A, C, insets, n=3). Secreted cytokine levels were quantified by ELISA using equal amounts of culture supernatants (12 h; B, D, n=6). *P<0.01 vs. untreated, †P0.01 vs. IL-17A. E-H, IL-17A induces iNOS expression and NO generation independent of IL-1β and TNF-α. Cardiomyocytes were incubated with IL-1β or TNF-α neutralizing antibodies (10 mg/ml for 1 h) prior to IL-17A addition (50 ng/ml). Normal hamster IgG and normal goat IgG served as respective isotype controls. iNOS mRNA expression was analyzed by RT-qPCR (E; n=6). Protein levels in whole cell lysates were analyzed by immunoblotting (F). The immunoreactive bands from three independent experiments were semiquantified by densitometry, and are represented as fold change from untreated in the lower panel. *P<0.01 vs. Untreated. iNOS enzyme activity was analyzed by the extent of conversion of L-[³H] arginine to L-[³H]citrulline in the presence of Ca²⁺ chelators (G; n=6) . Nitrite levels in culture supernatants were analyzed using Griess reagent (H; n=6) *P< 0.01 vs. untreated. I, IL-17A impairs cardiomyocyte contractility independent of secreted IL-1β and TNF-α. Cardiomyocytes treated as in E, but for 4h with IL-17A were analyzed for cardiomyocyte contractility. *P<0.01 vs. untreated (n=12). J, IL-17A fails to induce cardiomyocyte death. Cardiomyocytes treated with IL-17A (50 ng/ml for 8 h) were analyzed for cleaved caspase-3 levels by immunoblotting. Doxorubicin HCl (Dox., 20 μM for 8 h) served as a positive control. The immunoreactive bands from three independent experiments were semiquantified by densitometry, and are represented as fold change from untreated in the lower panel. *P<0.01 vs. Untreated.

3.6. Minocycline inhibits cytokine cocktail (IL-17A/IL-1β/TNF-α)-induced cardiomyocyte death

We have demonstrated that IL-17A stimulates the production of IL-1β and TNF-α in cardiomyocytes, an effect inhibited by minocycline. However, neutralizing either cytokine failed to modulate IL-17A-induced iNOS expression, NO generation, or cardiomyocyte contractile depression, implying that the low levels of cytokines generated by IL-17A may not be sufficient to induce cardiomyocyte death. In fact, the role of cytokines, especially those generated by immune cells into culture media when added to cardiomyocytes, exerted variable effects [43-46], suggesting that the levels of cytokines have to reach a certain threshold to act either alone or in synergy to modulate cardiomyocyte contractility or viability. Since IL-17A induced contractile depression, we wanted to rule out the possibility of reduced survival contributing to the observed contractile depression. Therefore, cardiomyocytes were incubated with IL-17A at 50 ng/ml for 8 h. The results show that incubation of adult cardiomyoctes with IL-17A failed to suppress the expression of anti-apoptotic Bcl-2 or Bcl-X_L (Fig. 6A), or promote translocation of Bax from cytoplasm to mitochondria, or stimulate cytochrome C release, or activate PARP or caspase-3 (Fig. 6B and 6C). However, Dox., an anti-neoplastic drug with known cardiotoxic effects, markedly modulated their expression and translocation in favor of cell death (Fig. 6A-6C). In fact, ELISA of cytoplasmic extracts revealed increased levels of mono and oligonucleosomal fragmented DNA in Dox. but not IL-17A-treated cells (Fig. 6. D). However, when combined with high levels of IL-1β (100 ng/ml) or TNF-α (100 ng/ml) or IL-1β +TNF-α (100 ng each/ml), IL-17A induced cell death, as evidenced by an increase in levels of cleaved caspase-3 in whole cell lysates (Fig. 6E) and mono and oligonucleosomal fragmented DNA in cytoplasmic extracts (Fig. 6F). Notably, cell death induced by the cytokine cocktail was not modulated by preincubation of cardiomyocytes with polymyxin B, indicating that that the trace amounts of endotoxin present in the cytokine cocktail had no significant effect on cell viability (data not shown). Importantly, minocycline pre-treatment reversed cytokine cocktail-induced cardiomyocyte death, as indicated by reduced levels cleaved caspase-3 (Fig. 6G) and mono- and oligonucleosomal fragmented DNA (Fig. 6H). Together, these results indicate that treatment with IL-17A alone fails to induce cardiomyocyte death, but exerts pro-apoptotic effects when combined with pathophysiological levels of IL-1β and TNF-α, an effect inhibited by minocycline (Fig. 6).

Figure 6. IL-17A, in combination with pathophysiological levels of IL-1β and TNF-α, but not alone, induces cardiomyocyte death.

A-C, Treatment with IL-17A alone fails to modulate the expression or translocation of proteins that regulate cell death. Cardiomyocytes were treated with IL-17A (50 ng/ml for 4 h). Bcl-2, Bcl-xL, Bcl-xS, cleaved PARP and cleaved caspase-3 levels in whole cell homogenates, and Bax and Cytochrome C levels in cytoplasmic (Cyto) and mitochondrial (Mito) fractions were analyzed by immunoblotting (n=3). GAPDH and mitofilin served as loading controls. D, IL-17A fails to induce cell death. Cardiomyocytes treated as in A, but for 8 h, were analyzed for cell death by quantifying mono and oligonucleosomal fragmented DNA in cytoplasmic extracts. Doxorubicin (Dox.; 20 μM for 8 h) served as a positive control. *P< 0.001 vs. untreated (n=6). E, F, Treatment with IL-17A, when combined with IL-1β and TNF-α, induces cardiomyocyte death. IL-1β and TNF-α induces cell death. Cardiomyocytes treated with a cytokine cocktail containing IL-17A (50 ng/ml), IL-1β (100 ng/ml) and TNF-α (100 ng/ml) for 4h (E) or 8 h (F). Cleaved casopase-3 levels were analyzed by immunoblotting using whole cell lysates (E; n=3). Mono and oligonucleosomal fragmented DNA in cytoplasmic extracts was quantified by ELISA (F; n=6), *P< 0.01 vs. untreated. G, H, Minocycline inhibits cytokine cocktail-induced cell death. Cardiomyocytes were treated with minocycline (10 μM in water for 15 min) prior to incubation with a cytokine cocktail, and analyzed as in E and F. Doxorubicin (Dox., 20 μM for 8 h) served as a positive control. G, n=3, F, n=6. *P< 0.001 vs. untreated, †P<0.01 vs. Cytokine cocktail.

Treatment with higher levels of IL-17A for longer periods had no significant effect on cardiomyocyte survival

We have demonstrated that treatment with IL-17A alone at 50 ng/ml concentration induces cardiomyocyte contractile depression, but not death (Fig. 1 and Fig. 5J), suggesting that even at these high concentrations and for the duration of treatment (8h), IL-17A only acts as a negative inotrope. Since, IL-17A, at 100 ng/ml concentration, has been previously reported to induce apoptosis of neonatal mouse cardiomyocytes [8, 47], we hypothesized that exposure to higher concentrations of IL-17A for a longer duration, will further upregulate proinflammatory cytokine expression, and synergistically induce cardiomyocyte death. Therefore, adult cardiomyocytes were exposed to different concentrations of IL-17A, ranging from 50 ng/ml to 100 ng/ml. As shown in Fig. 7, IL-17A stimulated IL-1β (Fig. 7A) and TNF-α (Fig. 7D) mRNA expression and protein levels (Fig. 7A and Fig. 7D, insets; lh) at all concentrations tested, but the levels at 100 ng/ml were not statistically significant from those detected at 50 ng/ml. Similarly, though IL-17A stimulated IL-1β and TNF-α secretion (Fig. 7B and Fig. 7E; 12 h), the secreted levels were similar between 50 ng/ml and 100 ng/ml of IL-17A. Further, increasing the duration of exposure to IL-17A from 12 to 36 h, increased the secreted levels of IL-1β and TNF-α (Fig. 7C and Fig. 7F). However, extending the exposure time to IL-17A at 100 ng/ml for up to 36 h tended to increase cardiomyocyte viability, as evidenced by a slightly increase in the levels of mono- and oligoneucleosomal fragmented DNA in cytoplasmic extracts and cleaved caspase-3 levels in whole cell homogenates (Fig. 7G). Together, these results indicate that exposure to IL-17A, even at 100 ng/ml, has no significant effect on survival of adult mouse cardiomyocytes (Fig. 7).

Figure 7. Exposure to high levels of IL-17A for longer periods of time has no significant effect on survival of adult cardiomyocytes.

A-C, Dose- and time-dependent effects of IL-17A on IL-1β expression. Adult cardiomyocytes were treated with the indicated concentrations of IL-17A for 1 (A), 12 (B) or up to 36 h (C). IL-1β mRNA expression was analyzed by RT-qPCR (A; n=6). Protein levels in cleared whole cell lysates were analyzed by immunoblotting (1h; A, inset, n=3). Secreted IL-1β levels were quantified by ELISA using equal amounts of culture supernatants (12 h, B; up to 36 h, C, n=6). *P< at least 0.01 vs. untreated, †P<at least 0.01 vs. IL-17A. D-F, Dose- and time-dependent effects of IL-17A on TNF-α expression. Cardiomyocytes were treated with the indicated concentrations of IL-17A for 1 (D), 12 (E) or up to 36 h (F). TNF-α mRNA expression was analyzed by RT-qPCR (A; n=6). Protein levels in cleared whole cell lysates were analyzed by immunoblotting (1h; D, inset, n=3). Secreted TNF-α levels were quantified by ELISA using equal amounts of culture supernatants (12 h, E; up to 36 h, F, n=6). *P< at least 0.01 vs. untreated, †P<at least 0.01 vs. IL-17A. G, IL-17A, even at 100ng/ml, fails to significantly modulate cardiomyocyte survival. Cardiomyocytes were treated with IL-17A (100 ng/ml) for up to 36 h. Mono- and oligonucleosomal fragmented DNA in cytoplasmic extracts (n=6) was analyzed by ELISA and cleaved caspase-3 levels in cleared whole cell lysates by immunoblotting (inset; n=3). Doxorubicin (Dox., 20 μM for 8 h) served as a positive control. *P< 0.001 vs. untreated (n=6).

4. Discussion

Interleukin (IL)-17A belongs to the unique IL-17 family of cytokines. Increased IL-17A expression has been shown to play a pathological role in various cardiovascular diseases [48, 49]. We have previously reported that IL-17A induces matrix metalloproteinase-1 (MMP-1) expression in human cardiac fibroblasts directly via p38 MAPK- and ERK-dependent AP-1, NF-κB, and C/EBPβ activation [4], suggesting its causal role in cardiac fibrosis and adverse myocardial remodeling. Since cardiomyocyte contractile dysfunction and death contribute to fibroblast activation, proliferation, and adverse remodeling, here we investigated whether IL-17A contributes to cardiomyocyte contractile dysfunction and death. Using primary adult mouse cardiomyocytes, we show that IL-17A, at pathophysiological concentrations, induces cardiomyocyte contractile dysfunction, mediated in part via TRAF3IP2-dependent p38 MAPK/NF-κB/iNOS/NO signaling, and independent of IL-1β and TNF-α, two well-known proinflammatory cytokines with negative inotropic effects. Interestingly, though IL-17A by itself failed to induce cardiomyocyte death, it did promote cardiomyocyte death when combined with pathophysiological concentrations of IL-1β or TNF-α. Importantly, minocycline treatment reversed both IL-17A-induced contractile depression and IL-17A/IL-1β/TNFα-induced cardiomyocyte death by targeting TRAF3IP2 induction and TRAF3IP2/TRAF6/TAK1-dependent p38 MAPK and NF-κB activation, iNOS expression and iNOS-mediated NO generation (Fig. 8). These data suggest the therapeutic potential of minocycline in adverse myocardial remodeling and heart failure.

Figure 8. Schema showing the possible signal transduction pathways involved in IL-17A-mediated cardiomyocyte contractile depression and death.

IL-17A induces cardiomyocyte contractile depression and death in part via iNOS expression and NO generation. IL-17A induces iNOS expression in part via TRAF3IP2/IKKβNF-κB and TRAF3IP2/TRAF6/TAK1/p38 MAPK activation. A crosstalk between p38 MAPK and NF-κB is also observed. Though IL-17A induced IL-1β and TNF-α expression and secretion, it failed to induce cardiomyocyte death, suggesting that the secreted cytokine levels did not reach the threshold to act in synergy with IL-17A to induce cell death. However, when combined with pathophysiological levels of exogenous IL-1β and TNF-α, IL-17A induced cardiomyocyte death. Importantly, minocycline pretreatment markedly attenuated several of the second messengers activated by IL-17A, and reversing contractile depression and death, indicating its therapeutic potential in ischemic and inflammatory cardiac diseases. Purple diamond: NO donors. M: Minocycline. Dashed lines: Pathways reported in literature.

Minocycline (7-dimethylamino-6-dimethyl-6-deoxytetracycline) is an FDA-approved second-generation semisynthetic tetracycline antibiotic that acts against both gram-positive and gram-negative bacteria. Independent of its anti-bacterial effects, minocycline has been shown to exert antioxidant, anti-inflammatory and anti-apoptotic effects both in vitro and in vivo. It is highly lipophilic, absorbed rapidly, has a long half-life, shows excellent bioavailability, and importantly, accumulates in tissues at far higher levels compared to its systemic levels. Moreover, it has a good safety record even when administered for longer periods of time [50-52].

As an anti-inflammatory agent, minocycline inhibits various pro-inflammatory pathways and blunts progression of inflammatory diseases. As a phenolic antioxidant, it scavenges free radicals with similar efficiency to that of α-tocopherol (vitamin E), and independent of chelating ferrous ions [50]. For example, minocycline has been shown to significantly reduce oxidative stress and infarct size in a rat model of coronary artery ligation [53]. In that study, the authors demonstrated significant uptake of minocycline into both ischemic and the adjacent non-ischemic myocardium, increasing by ~50-fold in the ischemic region and ~24-fold in the adjacent normal heart. Supporting those in vivo observations, minocycline levels were also found increased by several fold in cultured neonatal and adult cardiac fibroblasts incubated with minocycline, as well as adult cardiomyocytes, demonstrating significant bioavailability. As an anti-apoptotic agent, it is known to inhibit both intrinsic and extrinsic cell death pathways. For example, it is shown to inhibit caspase activation and reactivation and increases in XIAP/Smac/DIABLO ratio [54]. Moreover, it is reported to reduce mitochondrial leakage of cytochrome c and Smac/DIABLO in cultured cardiomyocytes [54]. Together, these reports indicate that minocycline exerts pleiotropic cardio-protective effects that are independent of its antibacterial properties.

Here we show that therapeutic doses of minocycline inhibit IL-17/IL-17R-mediated TRAF3IP2 induction and its downstream inflammatory signaling. TRAF3IP2 is a cytoplasmic adapter molecule essential in IL-17 signaling [16, 17]. It physically associates with IL-17R via SEFIR-SEFIR interaction, and plays a role in the activation of multiple downstream signaling intermediates, including NF-κB, AP-1, c/EBP, p38 MAPK and JNK. Interestingly, its expression is also regulated by AP-1 and c/EBP [23, 55], suggesting that TRAF3IP2 acts not only as an upstream regulator of these proinflammatory cis regulatory elements and stress-activated protein kinases, but also as their downstream target, resulting in its perpetual induction and activation. Of note, oxidative stress is a well-known regulator of NF-κB, AP-1, c/EBP, p38 MAPK and JNK activation. In fact, IL-17 is previously shown to induce NADPH oxidase- and xanthine oxidase-dependent reactive oxygen species generation [56]. Oxidative stress is also shown to regulate IL-17R expression [57], suggesting amplification in IL-17-mediated oxidative stress and proinflammatory signaling. Our data show that minocycline inhibits IL-17-induced activation of key transcription factors and stress-activated protein kinases in cardiomyocytes, possibly due to its antioxidant properties.

NF-κB is a ubiquitous dimeric nuclear transcription factor present in cytoplasm in an inactive state due its binding to an inhibitory subunit IκB. Phosphorylation of IκB at serine residues 32 and 36 or tyrosine at residue 42 has been shown to result in its dissociation from the NF-κB dimer and degradation in cytoplasm [58]. Our data show that minocycline inhibits IL-17/TRAF3IP2-dependent NF-κB activation in cardiomyocytes, as evidenced by a marked reduction in phospho-p65 levels. IKKβ, a component of the IKK signalosome, is an upstream regulator of IκB phosphorylation and NF-κB activation that is also inhibited by minocycline. Our data show that minocycline also inhibits activation of IKKβ. Similar to minocycline, silencing TRAF6 and TAK1, which are downstream of TRAF3IP2, also strongly inhibited IL-17A-mediated IKKβ activation. Together, these data suggest that minocycline inhibits oxidative stress-responsive TRAF3IP2/IKKβ/NF-κB inflammatory signaling in cardiomyocytes, again suggesting its antioxidant properties and its therapeutic potential in ischemic and inflammatory cardiac diseases.

TRAF3IP2 is also an upstream regulator of stress-activated kinases [5, 12, 13, 16, 22, 26]. Sustained activation of stress-activated protein kinases contribute causally to cardiac injury and inflammation [10, 14]. Our data show that targeting TRAF3IP2 or pretreatment with minocycline each inhibited IL-17A-mediated p38 MAPK activation and p38 MAPK-dependent NF-κB activation. Also, silencing TRAF6 and TAK1 each inhibited IL-17A-mediated p38 MAPK activation, suggesting that IL-17A mediates p38 MAPK activation in cardiomyocytes via TRAF3IP2/TRAF6/TAK1 signaling. TAK1 is also a major regulator of IKK phosphorylation and activation. Previously, Ataie-Kachole P, et al. reported that minocycline targets the NF-κB nexus by suppressing TGF-β1/TAK1/IκB signaling in ovarian cancer cells [59]. Those authors demonstrated that minocycline inhibited TAK1 phosphorylation in a dose-dependent manner. Interestingly, minocycline is known to increase phosphorylation and nuclear export of some targets, like NFAT1 ([60]. How minocycline regulates phosphorylation still remains elusive, and is a potential area of further research.

Similar to its inhibitory effects on inflammatory mediators IL-1β and TNF-α, our data also show that minocycline pretreatment inhibited iNOS activation and iNOS-dependent increased NO generation in cardiomyocytes. It is well established that sustained iNOS induction and iNOS-mediated NO generation contribute causally to inflammatory cytokine-induced cardiomyocyte contractile dysfunction and death. In fact, both IL-1β and TNF-α have been shown to induce iNOS expression and activation in isolated whole heart, papillary muscle and cardiomyocytes ([61-63]. iNOS expression is mainly regulated at the transcriptional level, and several of the TRAF3IP2 downstream signaling intermediates, such as NF-κB, AP-1, c/EBP, p38 MAPK or JNK, have been shown to positively regulate its transcription and activation, and sustained generation of NO. Increased NO generation has been shown to elicit deleterious effects by enhancing oxidative stress due to the formation of peroxynitrite in the presence of superoxide, leading to nitration of tyrosine residues (3-nitrotyrosine) [64, 65]. Interestingly, cytokine-induced iNOS induction and NO generation has been reported to inhibit cardiomyocyte contractility, but in a reversible manner. Our data support those reports, and demonstrate that pathophysiological levels of IL-17A depress cardiomyocyte contractility via TRAF3IP2/NF-κB/p38 MAPK-dependent iNOS expression and iNOS-mediated NO generation. Of note, excessive generation of NO has been shown to inhibit cardiomyocyte contractility by decreasing myofilament [Ca²⁺]i responses [66], indicating that like proinflammatory cytokines, increased NO levels exert myocardial negative inotropic effects. Importantly, minocycline blunted this response, potentially due to its antioxidant and anti-inflammatory effects.

Interestingly, our data demonstrate that exposure to IL-17A alone induces cardiomyocyte contractile depression, but not death. This contrasts reports that IL-17A induces death of neonatal cardiomyocytes at a supraphysiological concentration of 100 ng/ml [8, 47]. Our results show that exposure of adult cardiomyocytes to 100 ng/ml for 36 h further increased secreted IL-1β and TNF-α levels, but did not induce cell death. Interestingly, a combination of 50 ng/ml of IL-17A, which induces cardiomyocyte contractile depression, with pathophysiological concentrations of IL-1β and TNFα, induced significant cell death. These results indicate that though IL-17A treatment stimulates IL-1β and TNFα secretion from adult cardiomyocytes, it appears that the secreted levels (picograms/ml) are not sufficient enough to synergize with IL-17A to cause cardiomyocyte death, suggesting that the levels of IL-17A, IL-1β and TNF-α need to reach a threshold to act in synergy and induce cardiomyocyte death.

The role of iNOS and iNOS-mediated sustained NO generation have been previously reported in cardiac conduction defect. In a transgenic mouse model, overexpression of iNOS specifically in cardiomyocytes has been shown to result in increased cardiac iNOS activity with progressive defects in cardiac conduction and the development of lethal bradyarrhythmias [67]. In that study, some of the transgenic mice developed overt heart failure, depressed cardiac hemodynamic function, cardiomyocyte damage, increased fibrosis, ventricular thickening, and chamber size. Those transgenic mice also expressed increased oxidative stress as evidenced by increased levels of 3-nitrotyrosine (3-NT). Together, these results support our in vitro studies, and demonstrate that increased iNOS and excessive NO generation resulting from the cytokine cocktail (IL-17A/IL-1β/TNFα) promotes cell death. Importantly, minocycline inhibited cytokine cocktail-induced cardiomyocyte death, potentially by blocking iNOS transcription, protein expression, and excess generation of NO and 3-NT levels, as has been previously reported [68].

A limitation of the present study is that it is an in vitro mechanistic study describing the potential antioxidant, anti-inflammatory and anti-apoptotic effects of minocycline on IL-17A-mediated cardiomyocyte contractile dysfunction and death. Though minocycline exerts beneficial effects in various inflammatory diseases involving multiple organs in both humans and pre-clinical animal models, it is primarily an antibiotic and its long-term use is associated with complications in some cases, specifically pigmentation of skin, nails, bones, sclera, thyroid gland, aortic valve and sinus of Valsalva [69, 70], which usually resolves following cessation of therapy. However, minocycline either alone or in combination with other antibiotics has proven effective in many diseases. Our in vitro data convincingly demonstrate that minocycline inhibits IL-17A-mediated cardiomyocyte contractile dysfunction and death, suggesting its therapeutic potential in ischemic and inflammatory cardiac diseases.

Highlights.

• Interleukin-17A impairs adult cardiomyocyte contractility, but not death

• IL-17A induces cardiomyocyte death when combined with pathophysiological levels of IL-1β and TNF-α

• IL-17 induces cardiomyocyte contractile depression via iNOS induction and NO generation

• Minocycline inhibits IL-17A-induced iNOS expression and NO generation

• Minocycline inhibits IL-17A-induced cardiomyocyte contractile depression and death

Abbreviations

AP-1

Activator Protein 1

ASK1

Apoptosis Signal Regulating Kinase 1

Bcl-2

B-Cell CLL/Lymphoma 2

Bax

BCL2 Associated X, Apoptosis Regulator

BH4

tetrahydrobiopterin

DMSO

Dimethyl sulfoxide

Dox

Doxorubicin

DTT

dithiothreitol

ERK1/2

extracellur regulated kinase ½

GFP

Green Fluorescent Protein

IKK

IκB kinase

IκB

Inhibitory kappa B

IL

Interleukin

IL-17RA

Interleukin-17 receptor A

JNK

c-Jun N-terminal kinase

NADPH

nicotinamide adenine dinucleotide phosphate

iNOS

inducible nitric oxide

NOS2

Nitric oxide synthase 2

nNOS

neuronal NOS

eNOS

endothelial NOS

NO

nitric oxide

CM

cardiomyocyte

MAPK

Mitogen Activated Protein Kinase

MyD88

Myeloid Differentiation Primary Response 88

NF-κB

Nuclear Factor kappa-light-chain-enhancer of activated B cells

PARP

poly(ADP-ribose) polymerase

SEFIR

Similar Expression to Fibroblast Growth Factor genes and Interleukin-17R domain

TAK1

Transforming Growth Factor-Beta-Activated Kinase 1

TRAF3IP2

TRAF3 Interacting Protein 2

TRAF

TNF Receptor-Associated Factor

TNF

Tumor Necrosis Factor

NK cells

Natural Killer cells

RORγt

Retinoid-related orphan receptor gamma t

XIAP

X-linked Inhibitor of Apoptosis

Smac/DIABLO

Second mitochondria-derived activator of caspase/direct inhibitor of apoptosis-binding protein with low pI

IAP

Inhibitor of Apoptosis Protein

shRNA

Small hairpin RNA

SNAP

S-nitroso-N-acetylpenicillamine

PAPA

1-propamine 3-(2-hydroxy-2-nitroso-1-propylhydrazine