Cardioprotective cytokine interleukin‐33 is up‐regulated by statins in human cardiac tissue

Abstract Interleukin (IL)‐33 is a member of the IL‐1 family and is able to act cardioprotective. The aim of this study was to investigate the regulation of IL‐33 by 3‐hydroxy‐3‐methylglutaryl‐coenzyme‐A (HMG‐CoA) reductase inhibitors (statins) and bisphosphonates (BPs) in human cardiac tissue. The lipophilic fluvastatin, simvastatin, atorvastatin, and lovastatin as well as the nitrogenous BPs alendronate and ibandronate, but not hydrophilic pravastatin increased IL‐33 mRNA and intracellular IL‐33 protein levels in both human adult cardiac myocytes (HACM) and fibroblasts (HACF). Additionally, fluvastatin reduced soluble ST2 secretion from HACM. IL‐33 was also up‐regulated by the general inhibitor of prenylation perillic acid, a RhoA kinase inhibitor Y‐27632, and by latrunculin B, but statin‐induced IL‐33 expression was inhibited by mevalonate, geranylgeranyl pyrophosphate (GGPP) and RhoA activator U‐46619. The IL‐33 promoter was 2.3‐fold more accessible in statin‐treated HACM compared to untreated cells (P = 0.037). In explanted hearts of statin‐treated patients IL‐33 protein was up‐regulated as compared with the hearts of non‐statin‐treated patients (P = 0.048). As IL‐33 was previously shown to exert cardioprotective effects, one could speculate that such up‐regulation of IL‐33 expression in human cardiac cells, which might happen mainly through protein geranylgeranylation, could be a novel mechanism contributing to known cardioprotective effects of statins and BPs.

organisation of cardiac tissue and, as a consequence, in myocardial hypertrophy.
BPs represent another group of drugs targeting the mevalonate (MVA) pathway and are used for the treatment of osteoporosis and other diseases involving bone resorption. 10 BPs also exert beneficial functions in the cardiovascular system: they attenuate cardiac hypertrophy as well as fibrosis and improve endothelial function in hypertensive rats and mice, and cause regression of atherosclerosis. [11][12][13][14] However, the mechanisms underlying the pleiotropic effects of statins and BPs in myocardial tissue remodelling are not yet fully clarified.
IL-33 belongs to the IL-1 family of cytokines. 15 Depending on the experimental model and the pathophysiological context IL-33 can act either pro-or anti-inflammatory. 3,[16][17][18][19][20] In the heart, IL-33 is believed to be cardioprotective as it was shown to possess antihypertrophic and anti-apoptotic properties in animal models of pressure overload or myocardial infarction, respectively. [21][22][23][24] In this study, we aimed to explore if statins and BPs can influence the expression of IL-33 in human adult cardiac myocytes and fibroblasts, and to reveal the responsible mechanisms. Additionally, we compared IL-33 expression in explanted hearts from statin-treated versus non-statin-treated patients with end-stage HF undergoing heart transplantation.

| Cell culture
Primary HACM and HACF were isolated from myocardial tissue obtained from explanted recipients' hearts after heart transplantation and cultivated as described by our group. 8,26,27 For this study, cells were isolated from the hearts of 10 different donors (eight persons >18 years old and two children, mean age 40.5 ± 20.8; six males and four females) suffering from different pathologies (three patients with ischaemic cardiomyopathy, five patients with dilated cardiomyopathy (in one case perinatal), one patient with constrictive pericarditis, and one child with an atrioventricular septal defect).
HACM were isolated from ventricular tissue and HACF from the papillary muscle. For HACM isolation, ventricular tissue was grinded into very small fragments in PBS (pH 7.4) without enzymatic digestion. Red blood cells were removed by rinsing with PBS twice and debris was removed by passing through 40 mm cell strainer (Gibco-Life Technologies, Paisley, UK). The filtrated suspension was centrifuged at 335 g for 10 minutes. Next, the pellet was washed twice with PBS and resuspended in Dulbecco's Minimal Essential Medium (DMEM) supplemented with 10% fetal calf serum (FCS), 100 U/mL penicillin and 100 mg/mL streptomycin (all Cambrex, East Rutherford, NJ, USA). Cells were seeded into a Petri-dish and incubated for 60 minutes at 37°C in a humidified atmosphere of 5% CO 2 :95% air in order to separate myocytes from fibroblasts by pre-plating. Afterwards the non-attached cells were removed, centrifuged, and washed twice with PBS. The cell pellet was resolved in DMEM supplemented with 10% FCS, 100 U/mL penicillin, 100 mg/mL streptomycin, 10 mg/mL transferrin (Sigma), and 10 mg/mL insulin (Sigma). HACM were seeded at the density of around

| Human tissue
Human heart tissue was obtained from the left ventricle of explanted hearts from 10 patients (mean age 53.5 ± 9.3; nine males; five with statin treatment) undergoing heart transplantation for endstage HF. Specimens were paraffin-embedded and cut into 5 μm sections for further analysis. All human material was obtained and processed according to the recommendations of the hospital's Ethics Committee.

| Total RNA purification and cDNA preparation
Cells were treated as described, supernatants were removed and total cellular RNA was isolated using High Pure RNA Isolation Kit

| Protein detection
Cells were lysed with PBS containing 0.1% Triton X-100 (Sigma). IL-33 protein in cell lysates and sST2 protein in cell culture supernatants were measured by a specific enzyme-linked immunosorbent assay (ELISA) (both R&D Systems, Minneapolis, MN, USA), as described by us previously. 27

| Immunofluorescence analysis of IL-33 in explanted hearts
Paraffin-embedded sections were deparaffinised and then boiled in target retrieval solution (DAKO North America, Inc, CA) for 10 minutes. Staining with either mouse monoclonal anti-IL-33 (Alexis Biochemicals, Enzo Life Sciences AG, Lausen, Switzerland, dilution 1:200) or with rabbit polyclonal anti-Troponin I (abcam, Cambridge, UK, dilution 1:100) antibodies was performed as described. 27 The IL-33 fraction of the stained myocardial tissue was quantified using ImageJ software. Mean fluorescence intensity (MFI) was compared between patients with or without statin treatment in order to determine the myocardial IL-33 content.

| Statistics
Values are expressed as mean ± SD. Data were compared by twotailed student's t test or by ANOVA. Values of P ≤ 0.05 were con-  Figure 1A) and protein levels ( Figure 1B). Hydrophilic pravastatin did not change IL-33 expression levels ( Figure 1A,B).
This stimulatory effect on IL-33 expression was reproducible in cells derived from different donors' hearts when the cells were treated with fluvastatin at 5 μmol L −1 for 48 h (Table 1). IL-33 expression was up-regulated from 1.7-to 3.8-fold in HACM (n = 6 donors) and from 1.5-to 3.2-fold in HACF (n = 4 donors). In a series of experiments where the cells were treated with different concentrations of statins for different time points, we observed that the induction of IL-33 by statins was concentration-dependent in both cell types ( Figure 1C) and time-dependent in HACM ( Figure 1D). Fluvastatin was effective at 5 μmol L −1 and 2.5 μmol L −1 in both HACM and HACF, and even at 0.5 μmol L −1 in HACM ( Figure 1C). Treatment with simvastatin at 2.  Figure 1D). sST2 protein was detectable in cell culture supernatants of HACM in three from five tested donors and its secretion was reduced by fluvastatin at 5 μmol L −1 to 15% ± 2% after 24 h and to 29% ± 19% after 48 h of treatment compared to the respective controls (both P < 0.05). Fluvastatin also showed a trend towards downregulation of mRNA specific for sST2 in HACM in five donors tested (to 85% ± 24% after 24 h and to 62% ± 38% after 48 h as compared to the respective controls), but had no influence on the levels of ST2L mRNA (data not shown). To confirm that statin-induced IL-33 expression is driven by prevention of protein prenylation, we incubated HACM with perillic acid, which is a general inhibitor of prenylation. 28

| IL-33 is increased upon RhoA kinase inhibition and decreased upon RhoA activation
One of the most prominent geranylgeranylated protein is RhoA. 31 We tested if inhibition of ROCK, which is activated by active RhoA,  HACM and HACF isolated from different donors were incubated for 48 h in the absence or presence of fluvastatin at a concentration of 5 μmol L −1 . Cells were permeabilised with PBS containing 0.1% Triton X-100, and IL-33 protein in cell lysates was determined as described in "Materials and methods". Values are given in x-fold of control, which was set as 1.0.

| Increased IL-33 expression in myocardial tissue of statin-treated patients
In order to support our in vitro data, we additionally examined IL-33 protein level in myocardial tissue from patients undergoing heart transplantation for end-stage HF under statin treatment or not treated with statins. As shown in Figure 5B, myocardial tissue from statin-treated patients express higher level of IL-33 protein compared with patients without statin treatment ( Figure 5A, P = 0.048). Quantification of MFI of the staining is shown in Figure 5C.

| DISCUSSION
In our study, we report a novel aspect of action for two widely pravastatin had no effect on apoptosis or the expression of components of the plasminogen system and osteoprotegerin in human cardiac, endothelial, or smooth muscle cells in vitro. 7,8,25,36 In order to support our in vitro findings, we analysed IL- District 2) study concluded that higher-intensity statin use is associated with beneficial outcome in HF patients with ischaemic heart disease. 43 Interestingly, an overexpression of FFP synthase, the primary target of BPs, induced cardiac hypertrophy and HF in mice. 44 Furthermore, inhibition of FPP synthase attenuates hypertrophy in neonatal rat cardiomyocytes. 45  is released by necrotic cells during damage or by living cells during mechanical strain. 47,48 We showed previously the intracellular expression of IL-33 in human cardiac myocytes and fibroblasts, and the release of IL-33 during necrosis of these cells. 27 Moreover, the "decoy" receptor for IL-33, sST2, was shown to have predictive value in patients with coronary artery disease and HF 16,49-51 as well as in critically ill patients. 52 In our study presented here, we found that fluvastatin is able to reduce secretion of sST2 from human cardiac myocytes. Interestingly, a recent report showed that sST2 levels were attenuated after 12 months of atorvastatin treatment in patients with HIV. 53 However, rosuvastatin had no effect on the As an inflammatory mediator IL-33 plays a role in a variety of diseases such as bronchial asthma, inflammatory bowel diseases, and rheumatoid arthritis. 16 On the other hand, IL-33 antagonised prohypertrophic stimuli in rat neonatal cardiomyocytes and exhibited an anti-hypertrophic effect after transverse aortic constriction (TAC) in mice. 21 Furthermore, IL-33 protected rat neonatal cardiomyocytes from hypoxia-induced apoptosis and improved cardiac function in rats after experimental myocardial infarction. 22 IL-33 also attenuated ischaemia/reperfusion-induced myocardial injury in diabetic mice. 23 Recently, a cardioprotective role of the IL-33/ST2 system was shown in IL-33 knockout (IL-33 -/-) mice subjected to TAC, which exhibited exacerbated left ventricular hypertrophy, aggravated fibrosis, and impaired survival compared to wild-type littermates after TAC. 24 Therefore, IL-33 has been recognised as a cardioprotective cytokine.
In this context, it should be emphasised that the effects of statins on IL-33 expression in human cardiac cells were never studied before. Only one previous study showed that simvastatin induced IL-33 mRNA expression in human endothelial cells. 55 We further investigated, which mechanisms are responsible for the increase of IL-33. It was shown previously that the MVA pathway inhibitors simvastatin and ibandronate affect epigenetic regulation in cancer cells. 35 We demonstrate here that the promoter  RhoA is a prominent geranylgeranylated protein, which has also been associated with the diseases of the cardiovascular system and bone. 31 Statins block the isoprenylation and thus the membrane targeting and functional activation of the Rho family member. 31 It has been shown that fluvastatin reduces RhoA in the membrane fraction of rat cardiac myocytes. 29 In spontaneously hypertensive rats simvastatin attenuated RhoA activity and myocyte hypertrophy. 57 The most important direct target of RhoA is ROCK. We therefore suspected RhoA and subsequent ROCK inactivation of being responsible for the effect of statins and BPs observed by us in human cardiac cells. We confirmed our hypothesis by treating these cells with the ROCK inhibitor Y-27632. We found that Y-27632 at  Figure 6.
It should be noted that the concentrations of statins and BPs as well as the components of the MVA pathway and specific inhibitors used in this study were in the same range as concentrations of these substances used in numerous other tissue culture studies. 3,[7][8][9]14,25,29,36 In conclusion, our study showed for the first time that IL-33 could be an important mediator of the pleiotropic effects of statins and BPs in human cardiac tissue. Considering potential safety issues, F I G U R E 6 Schematic representation of the downstream targets of the MVA pathway, depicting its various intermediates (blue), end products (green) and functional implications (yellow). Prenylation of RhoA with GGPP anchors RhoA to the plasma membrane, where it can be activated by signalling factors. Active RhoA activates ROCK, which triggers activation and translocation of SRF into the nucleus, where it represses the transcription of IL-33. Inhibitors used in this study are depicted in red: statins inhibit HMG-CoA-reductase; BPs inhibit GPP synthase and FPP synthase; perillic acid inhibits protein prenylation; latrunculin B prevents SRF activation; Y-27632 is a ROCK inhibitor. The RhoA activator U-46619 is depicted in green the feasibility and possible implications of the therapeutic modulation of IL-33 signalling, which is known to act cardioprotective, [21][22][23][24] are currently discussed. 58 As statins and BPs represent widely used and standardised classes of drugs for primary and secondary prevention of cardiovascular disease, our results might have impact also in the clinical setting.