Orphan nuclear receptor Nur77 affects cardiomyocyte calcium homeostasis and adverse cardiac remodelling

Distinct stressors may induce heart failure. As compensation, β-adrenergic stimulation enhances myocardial contractility by elevating cardiomyocyte intracellular Ca2+ ([Ca2+]i). However, chronic β-adrenergic stimulation promotes adverse cardiac remodelling. Cardiac expression of nuclear receptor Nur77 is enhanced by β-adrenergic stimulation, but its role in cardiac remodelling is still unclear. We show high and rapid Nur77 upregulation in cardiomyocytes stimulated with β-adrenergic agonist isoproterenol. Nur77 knockdown in culture resulted in hypertrophic cardiomyocytes. Ventricular cardiomyocytes from Nur77-deficient (Nur77-KO) mice exhibited elevated diastolic and systolic [Ca2+]i and prolonged action potentials compared to wild type (WT). In vivo, these differences resulted in larger cardiomyocytes, increased expression of hypertrophic genes, and more cardiac fibrosis in Nur77-KO mice upon chronic isoproterenol stimulation. In line with the observed elevated [Ca2+]i, Ca2+-activated phosphatase calcineurin was more active in Nur77-KO mice compared to WT. In contrast, after cardiac pressure overload by aortic constriction, Nur77-KO mice exhibited attenuated remodelling compared to WT. Concluding, Nur77-deficiency results in significantly altered cardiac Ca2+ homeostasis and distinct remodelling outcome depending on the type of insult. Detailed knowledge on the role of Nur77 in maintaining cardiomyocyte Ca2+ homeostasis and the dual role Nur77 plays in cardiac remodelling will aid in developing personalized therapies against heart failure.

Scientific RepoRts | 5:15404 | DOi: 10.1038/srep15404 overstimulation is demonstrated by improved clinical outcome after inhibition of β -adrenergic receptor activation by β -blockers, currently the first-line therapy for heart failure 3 . With a prevalence of more than 23 million heart failure patients worldwide, and 5-and 10-year survival rates of respectively 50% and 10%, heart failure is one of the leading causes of death in the Western world 4 . Hence, unravelling the molecular pathophysiology in the early stages of cardiac remodelling, before the onset of functional heart failure, is of great importance for the development of novel therapeutic strategies.
Activation of β -adrenergic Gs-protein-coupled receptors induces activation of signalling cascades, leading to larger intracellular calcium ([Ca 2+ ] i ) transients, necessary for enhanced contractility. Larger increases in [Ca 2+ ] i will lead to stronger contractions 5 . In addition, [Ca 2+ ] i homeostasis is essential to regulate excitation-contraction coupling, maintain electrophysiological balance, and control transcriptional regulation by activating key enzymes such as Ca 2+ -calmodulin-activated phosphatase calcineurin 6 . Consequently, altered cardiomyocyte Ca 2+ homeostasis is an important characteristic of diseased myocardium and plays a central role in the pathophysiology of cardiac remodelling and heart failure 7 .
Orphan nuclear receptor Nur77, also known as NR4A1, NGFI-B or TR3, is an immediate-early gene which is involved in cellular stress responses, metabolism, proliferation, differentiation and apoptosis in various cell types. In vascular disease, we and others have shown that Nur77 inhibits smooth muscle proliferation 8,9 , macrophage accumulation 10,11 , pro-inflammatory macrophage polarization [10][11][12][13] , lipid loading/foam-cell formation 13 , and matrix metalloproteinase production 9 . Together with the promotion of a contractile phenotype in smooth muscle cells 8,9 , enhanced endothelial cell survival and endothelial cell-driven angiogenesis 14 , Nur77 has been attributed an atheroprotective role in the arterial vessel wall. At present, the role of Nur77 in the heart is less clear. Nur77 has been shown to mediate cardiomyocyte apoptosis upon oxidative stress 15 and in response to angiotensin II-induced pressure overload, Nur77-deficient (Nur77-KO) mice and rats with cardiac-specific Nur77 knock-down exhibit attenuated adverse cardiac remodelling 16 . However, after myocardial infarction, Nur77-KO mice were shown to present worsened outcome 17 . Together these reports imply a dual role for Nur77 in cardiac disease. Interestingly, β -adrenergic stimulation has been shown to significantly up-regulate Nur77 expression in the heart 18 implying a role for Nur77 in β -adrenergic overstimulation-induced cardiac remodelling. Very recently, it was shown that Nur77 overexpression in murine hearts could attenuate hypertrophy after β -adrenergic overstimulation 19 . Here, we assess the effect of Nur77 deficiency on calcium homeostasis in cardiomyocytes, in cardiac remodelling induced by chronic β -adrenergic stimulation, and in relation to cardiac pressure overload by transverse aortic constriction (TAC).

Results
Nur77 expression in the healthy heart and isoproterenol-challenged cardiomyocytes. Nur77 is rapidly and highly expressed in many tissues upon various stimuli. However, as shown on multi-tissue Northern blots, endogenous Nur77 expression is most pronounced in healthy human heart and skeletal muscle (Fig. 1a). Given that NR4A mRNA expression in whole hearts is increased by β -adrenergic stimulation 18,19 , we assessed expression of Nur77, Nurr1 and NOR-1 in neonatal rat cardiomyocytes (NRVMs) upon β -adrenergic stimulation. After 1 h of stimulation with β -adrenergic agonist isoproterenol, gene expression of all three NR4A family members was robustly increased compared to PBS-stimulated cells, however Nur77 was expressed to the highest extent amongst all NR4A family members (Fig. 1b). Expression of all NR4A mRNAs was reduced back to control levels after 3 h of isoproterenol stimulation ( Supplementary Fig. S1). Nuclear Nur77 protein is observed in healthy murine cardiomyocytes and its expression was markedly enhanced after 1 h and 3 h of isoproterenol stimulation (Fig. 1c). This high and rapid induction of Nur77 upon isoproterenol stimulation suggests a prominent role for Nur77 in the heart's response to adrenergic stimulation. To see whether this has pathological implications, we assessed isoproterenol-induced cardiomyocyte hypertrophy upon Nur77 knockdown. Transfection with Nur77-specific siRNA (siNur77) resulted in markedly down-regulated Nur77 mRNA expression, and it significantly inhibited the observed isoproterenol-induced Nur77 mRNA upregulation when compared to control-transfected (siCon) cells (Fig. 1d). Interestingly, Nur77 knockdown in unstimulated NRVMs already resulted in significantly larger cardiomyocytes than siCon-transfected cells (Fig. 1d). Upon isoproterenol stimulation, siNur77 and siCon-transfected NRVMs significantly increased in size. Again, the siNur77-transfected NRVMs were significantly larger than siCon NRVMs, showing that reduced Nur77 induces a hypertrophic response in isolated cardiomyocytes.

Nur77-KO cardiomyocytes exhibit elevated [Ca 2+ ] i and prolonged action potentials. As Ca 2+
is the key effector of adrenergic signalling, we analysed [Ca 2+ ] i transients in adult wild type (WT) and Nur77-KO cardiomyocytes in the absence or presence of the synthetic adrenergic agonist isoproterenol or the native adrenergic agonist norepinephrine. Already at baseline, Nur77-KO cardiomyocytes exhibited significantly larger [Ca 2+ ] i transient amplitudes compared to WT cells, a difference which was preserved upon stimulation with increasing concentrations of either isoproterenol or norepinephrine (Fig. 2a). The relative amplitude increase upon adrenergic stimulation did not significantly differ between WT and Nur77-KO cardiomyocytes (Fig. 2b). Interestingly, basal diastolic [Ca 2+ ] i concentrations were significantly elevated in Nur77-KO cardiomyocytes compared to WT cells, yet without differences in the time to peak and decay time of the [Ca 2+ ] i transient (Fig. 2c). Thus, Nur77-KO cardiomyocytes seem To further characterize Nur77-KO cardiomyocytes, we measured if the changes in [Ca 2+ ] i influence the electrophysiological properties of these cells. Figure 3a,b, show typical action potentials (APs) and  average AP properties, respectively, measured at 6 Hz. No differences in resting membrane potential (RMP) and AP amplitude (APA) were observed between WT and Nur77-KO cardiomyocytes. Also, maximal AP upstroke velocity (dV/dt max ), a measure of sodium channel availability, did not differ between the groups. This is in line with unaltered ex vivo conduction parameters observed between isolated, Langendorff-perfused, WT and Nur77-KO hearts ( Supplementary Fig. S2). While no significant differences in AP duration at 20% and 50% of repolarization (APD 20 and APD 50 , respectively) were observed, Nur77-KO APs display a significantly longer duration at 90% repolarization (APD 90 ) compared to WT. Specifically a 24% lengthening of the APD 90 was observed. Prolonged APs are in line with the trend towards prolonged ex vivo effective refractory period (ERP; p = 0.06) of Nur77-KO hearts. Significant  changes in APD 90 were evident at all measured stimulation frequencies (Fig. 3c). Interestingly, at a stimulus frequency of 1 and 2 Hz, early after-depolarisations (EADs; Fig. 3c inset; arrow) were observed in 22% of Nur77-KO cardiomyocytes, but never in WT cells (P ≤ 0.05, Fisher exact test). Taken together these data suggest a role for Nur77 in electrochemical Ca 2+ homeostasis maintenance in cardiomyocytes.
Expression of cardiac Ca 2+ -handling-related genes. As Nur77 is a transcriptional regulator, we wondered if the altered Ca 2+ homeostasis in Nur77-KO cardiomyocytes may be explained by differences on gene expression level. Thus, we analysed gene expression of adrenergic receptors and Ca 2+ -handling proteins in left ventricular lysates of healthy WT and Nur77-KO mouse hearts. Neither α -nor the major β -adrenergic receptor subtypes (adra1, arb1, adrb2) were expressed differentially in Nur77-KO left ventricles compared to WT (Fig. 4a), suggesting there is no obvious difference in endogenous cardiac adrenergic stimulation between the groups. Interestingly, only a significant down-regulation of serca2a gene expression was detected in Nur77-KO mice, while genes encoding for phospholamban (plb), the ryanodine receptor (ryr2), the L-type Ca 2+ channel (cacnb2), the Na + /Ca 2+ exchanger (ncx) and plasma membrane Ca 2+ ATP-ase (pmca) did not differ significantly (Fig. 4b).
Nur77 deficiency aggravates isoproterenol-induced cardiac remodelling. Alteration of cardiomyocyte [Ca 2+ ] i has long been implied as an important regulator of maladaptive cardiac remodelling and heart failure 6 , therefore we assessed the effect of Nur77 deficiency on the development of adverse cardiac remodelling in response to chronic β -adrenergic stimulation. We implanted osmotic minipumps, delivering a relatively high dose of isoproterenol for one week, in WT and Nur77-KO mice. During the experiment, 3 out of 16 isoproterenol-treated Nur77-KO mice died, two of which had no apparent cardiac tissue damage and one mouse exhibiting cardiac rupture (tamponade). All isoproterenol-treated WT mice (n = 16) and control Nur77-KO and WT mice (n = 9 per group) survived the experiment. Although   no significant changes in cardiac function were apparent after 1 week ( Supplementary Fig. S3), the effectiveness of this short-term hypertrophy model is demonstrated by a higher heart weight to tibia length (HW/TL) ratio in the isoproterenol-treated mice compared to their controls (Fig. 5a). Furthermore, the surviving Nur77-KO mice exhibited a trend (p = 0.06) towards a higher HW/TL ratio when compared to the WT mice. Validation of the cardiomyocyte surface area showed that cardiomyocytes were significantly larger in isoproterenol-treated mice compared to the controls. In line with Nur77 knockdown in cultured rat cardiomyocytes, isoproterenol-treated Nur77-KO cardiomyocytes were significantly larger than in isoproterenol-treated WT mice (Fig. 5b). Assessment of the hypertrophic foetal gene program 20 , revealed that all measured genes were up-regulated in ventricular lysates of isoproterenol-treated mice compared to the untreated controls (Fig. 5c). In addition, skeletal alpha actin (acta-1) and β -myosin heavy chain (myh7) were expressed significantly higher in isoproterenol-treated Nur77-KO mice, while there was no difference in atrial and brain natriuretic factor (anp and bnp) expression. It should be noted that both Acta-1 and Myh7 are involved in contractility, suggesting involvement of Nur77 in cardiomyocyte contractility.
Fibrosis is an important pathological feature associated with adverse cardiac remodelling 2 . In the isoproterenol-treated mice, already in the short time frame of one week, isoproterenol stimulation caused interstitial fibrosis in hearts of both WT and Nur77-KO mice, when compared to the control hearts (Fig. 5d). Upon quantification, a 3-fold induction in fibrosis was measured in the Nur77-KO mice (median 4.1%), as compared to WT mice (median 1.4%). As Nur77 has been shown to regulate apoptosis in a variety of cells, including cardiomyocytes 15 , we assessed apoptotic cells in the heart after isoproterenol-induced remodelling. However, apoptotic cardiac cells were only observed in a fraction of the mice within each group, showing no difference between WT and Nur77-KO ( Supplementary Fig. S3).
Protein phosphatase calcineurin is activated by sustained [Ca 2+ ] i elevation 21 and calcineurin activation has long been implied as a sufficient inducer of hypertrophic signalling in heart 22 . We hypothesized that the elevated [Ca 2+ ] i in Nur77-KO cardiomyocytes may lead to higher activation of calcineurin and therefore may explain the enhanced remodelling observed in Nur77-KO mice. Therefore, we assessed calcineurin activity in left ventricular lysates. Indeed, calcineurin activity is already 2-fold higher in control Nur77-KO mice compared to WT (Fig. 5e). While calcineurin activity is significantly enhanced in WT hearts after isoproterenol stimulation as expected, calcineurin was interestingly not further increased in Nur77-KO mice.
Taken together, these results demonstrate a cardioprotective role for Nur77 upon chronic β -adrenergic stimulation. However, how do these data relate to the existing Nur77 data on cardiac remodelling? Nur77 deficiency attenuates TAC-induced cardiac remodelling. Very recently, it was shown that local adenoviral overexpression of Nur77 in murine hearts could reduce hypertrophy after β -adrenergic stimulation with isoproterenol 19 , supporting our findings. In two different models of pressure overload, by chronic angiotensin-II infusion 16 or TAC 15 , up-regulation of cardiac Nur77 expression was observed. Interestingly, cardiac remodelling by angiotensin-II-induced pressure overload was attenuated in Nur77-KO mice 16 . As this is in contrast with the observations after β -adrenergic overstimulation, we subjected WT and Nur77-KO mice to TAC-induced cardiac pressure overload, to establish whether the effect of Nur77 deficiency is dependent on the type of cardiac stressor. In line with the angiotensin-II model, cardiac pressure overload in Nur77-KO mice led to attenuated adverse cardiac remodelling compared to WT after 28 days of TAC. Survival was equal between the two groups; 1 out of 13 WT and 1 out of 12 Nur77-KO mice died during the experiment. Nur77-KO mice showed a trend towards enhanced cardiac function (p = 0.10 ejection fraction, p = 0.08 fractional shortening) assessed by echocardiography ( Supplementary Fig. S4). In addition to cardiac function, left ventricular weight/TL ratio was lower in Nur77-KO mice compared to WT (Fig. 6a), suggesting reduced hypertrophy. Indeed, the mean cardiomyocyte surface area was significantly smaller in Nur77-KO mice (Fig. 6b). Moreover, expression of anp, bnp, acta-1 and myh7 were all significantly lower in Nur77-KO ventricular tissue (Fig. 6c). As TAC induces cardiac pressure overload, we assessed perivascular and interstitial fibrosis separately. Interestingly, perivascular fibrosis was significantly higher in WT mice compared to Nur77-KO mice, whereas no significant difference was observed in interstitial collagen deposition (Fig. 6d). As in the isoproterenol model, no difference in the number of apoptotic cells in WT and Nur77-KO hearts was found after TAC ( Supplementary Fig. S4).
In conclusion, our data demonstrate that Nur77 plays a detrimental role in the heart upon TAC-induced pressure overload, in contrast to its role in isoproterenol-induced cardiac remodelling. Taking together our data on Nur77 in the heart with existing data, we propose the following scheme on the function of Nur77 in different cardiac insult models, leading to heart failure (Fig. 6e).

Discussion
Essential roles for Nur77 have been demonstrated regarding atherosclerosis 10,11,23 , cancer 24 and metabolic disease 25 . However, knowledge on the functional role of Nur77 in the cardiac setting is limited. The abundant endogenous expression of the NR4A family of nuclear receptors, including Nur77, in healthy murine heart tissue was previously described by Myers et al. 18 . Here, we demonstrate that also in the healthy human heart, endogenous Nur77 expression is relatively high as compared to other tissues.  (a) Heart weight/tibia length (HW/TL) ratio was higher after isoproterenol stimulation and further increased in Nur77-KO mice compared to WT mice. Tibia length did not differ between WT and Nur77-KO mice. (b) Cardiomyocyte cross-sectional area was significantly larger in isoproterenol-treated Nur77-KO mice compared to treated WT mice, as assessed by fluorescent wheat germ agglutinin staining in > 75 cells per heart. Photomicrographs are shown at 630× magnification. (c) RT-PCR analysis showed significant up-regulation of foetal gene expression after isoproterenol treatment. Nur77-KO mice exhibited a significant increase in acta1 and myh7 after isoproterenol, when compared to WT. anp: atrial natriuretic peptide; bnp: brain natriuretic peptide; acta1: α 1 skeletal actin; myh7: β -myosin heavy chain. (d) Collagen deposition was assessed by Masson's Trichrome staining assessed by quantitative morphometry, photomicrographs are shown at 50× magnification. The left ventricle and septum of Nur77-KO mice was affected stronger by interstitial fibrosis (blue) after isoproterenol than WT mice. (e) Calcineurin phosphatase activity is significantly higher in myocardium of control Nur77-KO mice. After chronic isoproterenol stimulation, calcineurin activity is significantly enhanced in WT hearts, while it does not further increase in Nur77-KO mice compared to Nur77-KO control mice (n = 3-4 in each group). Boxplots represent median, inter-quartile range and minimum/ maximum values; bars represent mean+ SEM; *p < 0.05, **p < 0.01, ***p < 0.001. A number of studies describe enhanced expression of Nur77 in the diseased heart [15][16][17]19 . Transient up-regulation of cardiac Nur77 mRNA and protein has been shown in animal models of pressure-overload induced by TAC 15 , chronic angiotensin-II infusion 16 , as well as in ischemia following myocardial infarction 15 . Nur77-KO mice and rats with cardiac-specific Nur77 knock-down exhibited attenuated cardiac remodelling after chronic angiotensin-II infusion. Angiotensin-II induces cardiac hypertrophy by increasing systemic blood pressure in conjunction with increased signalling of the angiotensin-II type I receptor on cardiomyocytes 26 . Nur77 is considered detrimental in this model by activating mTORC1, leading to cardiac hypertrophy by enhancement of protein synthesis, cardiomyocyte growth and reactive oxygen species (ROS) production 16 . In cultured cardiomyocytes, a ROS challenge was shown to cause Nur77 translocation to mitochondria, causing cytochrome c release and subsequent apoptosis 15 . In the same study, mitochondrial translocation of Nur77 was observed after ischemia/reperfusion injury in the heart. In our murine studies, apoptosis did not seem to play a key role in the observed pathologies. In contrast, Hilgendorf et al. 17 observed after myocardial infarction, that Nur77-KO mice exhibit enhanced adverse cardiac remodelling with increased inflammation, fibrotic scar size and ventricular dysfunction 17 .
Rapid up-regulation of Nur77 in whole mouse hearts was previously reported after a single intraperitoneal injection of isoproterenol 18,19 . Here, we show that both Nur77 mRNA and nuclear protein are rapidly and highly up-regulated in isolated cardiomyocytes in response to isoproterenol stimulation. To assess Nur77 function, knockdown of Nur77 in cultured cardiomyocytes revealed that Nur77 deficiency caused a hypertrophic response, which was further enhanced upon isoproterenol stimulation. Very recently, Yan et al. 19 demonstrated, in accordance with our data, attenuated cardiomyocyte hypertrophy after isoproterenol stimulation when overexpressing Nur77. In addition, adverse cardiac remodelling in mice was reduced by cardiac-specific Nur77 overexpression. They reveal an interaction of Nur77 with the pro-hypertrophic transcription factors NFATc3 and GATA-4, of which NFATc3 is regulated by calcineurin. In our study, we show that the adverse cardiac remodelling we observe in Nur77- KO 27 . This diastolic Ca 2+ accumulation is usually associated with diminished sarcoplasmic sequestration, often by lowered Serca function 28 . Indeed we observed significantly lower Serca2a mRNA expression in Nur77-KO hearts. However, as transient decay times were not prolonged and amplitudes were not diminished, Ca 2+ sequestration does not seem to be impaired in Nur77-KO cardiomyocytes. To shed light on the mechanism behind the elevated diastolic and systolic [Ca 2+ ] i in Nur77-KO cardiomyocytes, the different mechanisms which facilitate cytosolic Ca 2+ influx and efflux must be assessed. Higher sarcoplasmic Ca 2+ load, increased sarcoplasmic Ca 2+ leak due to enhanced open probability of RyR channels, increased L-type Ca 2+ current, decreased expression or enhanced reverse mode activity of NCX and cytosolic Na + accumulation have all been shown to interplay with elevated diastolic and systolic Ca 2+ (as reviewed previously 29 ). Electrophysiological characterization revealed that cardiac conduction was unaffected in Nur77-KO hearts, but Nur77-KO cardiomyocytes displayed prolonged APs and EADs. Both prolonged APs and EADs may form pro-arrhythmic substrates 30 , possibly explaining the two dead Nur77-KO mice in our isoproterenol experiment, which did not reveal an obvious cause of death. Prolonged APs and EADs may occur upon reduction of repolarization reserve by either increased L-type Ca 2+ current or reduced potassium (K + ) current, or a combination of both 31,32 . Nur77 may regulate the K + channel, I K,slow2 , encoded by the Kv2.1 gene (kcnb1), since a mutation in a potential Nur77 DNA-binding site in the kcnb1 promoter reduced its activity 33 . Decreased kcnb1 levels have been reported in cardiac disease 34,35 . On the other hand, an increase in L-type Ca 2+ current in Nur77-KO cardiomyocytes would be in line with the larger [Ca 2+ ] i transient amplitudes. In neuronal cells, K + Cl − -induced membrane depolarization enhances [Ca 2+ ] i rises with subsequent activation of calcineurin 36,37 . Downstream of calcineurin, cAMP response element binding protein (CREB) subsequently stimulates expression of Nur77 38 . Expression of NR4A family member Nurr1 is also regulated by calcineurin and this induction was inhibited upon L-type Ca 2+ channel blockade 39 . Taken together, we hypothesize that Nur77 may exert a feedback mechanism on [Ca 2+ ] i elevations and calcineurin activity.
In line with our isoproterenol experiment, Nur77-KO mice exhibited a worsened outcome after myocardial infarction 17 . However, this effect was largely attributed to deficiency of reparative Ly6C-low monocytes in the Nur77-KO mice 17 , while potential changes in cardiomyocytes were not taken into account. Given that Nur77 is essential for the development of Ly6C-low monocytes 40 and that these cells are instrumental to repair cardiac damage and resolve fibrosis 41 , the enhanced cardiac damage observed in the Nur77-KO mice in our isoproterenol model may partially be explained by Ly6C-low monocyte deficiency. Interestingly, the absence of Ly6C-low monocytes does not seem to affect cardiac damage after TAC in a negative manner, supporting the notion that Nur77 is also important in cardiac cells.
The Nur77-KO cardiomyocyte phenotype appears unfavourable, yet, challenging Nur77-KO mice with different cardiac stressors resulted in opposite remodelling outcomes. In line with published data 16 , we also observed that Nur77 has an adverse function in cardiac remodelling caused by pressure overload, while we reveal a protective role for Nur77 in cardiac remodelling caused by β -adrenergic overstimulation. While activation of β -adrenergic signalling may be present as a compensatory mechanism upon pressure overload, the role of Nur77 seems to be dependent on the dominant cardiac stressor. We hypothesize that increased [Ca 2+ ] i availability in Nur77-KO cardiomyocytes preserves contractility, which may compensate for the increased pressure, leading to improved outcome after TAC. Chronic Scientific RepoRts | 5:15404 | DOi: 10.1038/srep15404 isoproterenol stimulation modulates Ca 2+ handling by increasing cardiomyocyte AP duration, enhancing reverse Na + /Ca 2+ -exchange and decreasing K + current density (reviewed in 42 ). In Nur77-KO cardiomyocytes, where basal [Ca 2+ ] i is already relatively high, we hypothesize that chronic β -adrenergic stimulation induces [Ca 2+ ] i abnormalities more easily.
Since only limited changes were observed in cardiac function in both models as assessed by echocardiography, our study contributes to the understanding of early stages of cardiac remodeling, when intervention to prevent the heart from eventual failure is still feasible.
In conclusion, we have shown for the first time that Nur77 deficiency leads to elevated intracellular Ca 2+ levels and prolonged action potentials in cardiomyocytes and enhanced calcineurin activity in Nur77-KO hearts, however resulting in opposite cardiac remodelling outcome after different types of cardiac insults. Modulation of Nur77 may be a novel therapeutic target in managing adverse myocardial remodelling and heart failure, discriminating between cardiac stressors. Our study supports the assumption that knowing the induction mechanism of cardiac remodelling in patients is crucially important before deciding on a therapeutic strategy.

Methods
Northern blot analysis. Human multiple tissue poly(A) + RNA blots were purchased from CLONTECH (Palo Alto, CA, USA) and hybridized according to manufacturer's protocol. Detection was performed using a fragment of human Nur77 cDNA (base pairs 864 to 1905) and the supplied human β -actin cDNA probe for normalisation.
Cell isolation and culture. Neonatal rat ventricular myocytes (NRVMs) were isolated as previously described 43 . In short, hearts from 1-3 day old Wistar rats were excised, atria were removed, and hearts were cut into 4-6 pieces. The pieces were left overnight to rotate at 4 °C in HBSS (Invitrogen, Carlsbad, CA, USA) containing 1 mg/ml trypsin (Affymetrix, Santa Clara, CA, USA). The next day, cells were dissociated using 1 mg/ml collagenase (Worthington, Lakewood, NJ, USA) in HBSS. Cells were collected and resuspended in M199 culture medium, supplemented with 1% HEPES, 1% NEAA, 2 mg/L vitamin B12, 3.5 g/L glucose, and antibiotics (all from Invitrogen), and pre-plated to separate myocytes from fibroblasts. After 2 hours, non-adherent cells were collected, counted and plated on fibronectin-coated plates. NRVMs were serum-starved for 24 h in Dulbecco's modified Eagle medium (DMEM) supplemented with penicillin/streptavidin (Invitrogen) before experiments.
Protein isolation and Western blotting. NRVMs were stimulated with 25 µm isoproterenol (Sigma-Aldrich, St. Louis, MO, USA) or equivalent volume of PBS for indicated time points. Nuclear protein was isolated from NRVMs using the NE-PER Nuclear and Cytoplasmic extraction kit (Thermo Scientific, Rockford, IL, USA) according to manufacturer's protocol. Lysates were boiled, separated by SDS-PAGE and transferred onto PVDF membranes. After transfer, membranes were blocked with Odyssey blocking buffer (LI-COR, Lincoln, NB, USA) and incubated with anti-Nur77 (Santa Cruz, Dallas, TX, USA) and anti-H3 (Abcam, Cambridge, UK) for normalisation. Anti-rabbit IRDye800 was used as secondary antibody (LI-COR). Blots were scanned and quantified using the LI-COR Odyssey Infrared Imaging System. Nur77 knockdown and NRVM size assessment. NRVMs were transfected with a final concentration of 175 nM Nur77 siRNA (SMARTpool, Dharmacon, Lafayette, CO, USA) or a non-targeting control siRNA for 6 h using Lipofectamine RNAiMAX (Invitrogen) according to manufacturer's instructions. For assessment of Nur77 knockdown, NRVMs were cultured for 48 h after transfection and were subsequently stimulated with 25 µm isoproterenol or equivalent volume of PBS for 1 h. For cell size experiments, NRVMs were stimulated with 25 µm isoproterenol or PBS for 48 h after transfection. Cells were fixed with 4% paraformaldehyde (Roth, Karlsruhe, Germany) and stained with anti-α -Actinin antibody (Sigma Aldrich), Hoechst and anti-mouse AlexaFluor488-conjugated secondary antibody (Molecular Probes, Eugene, OR, USA). Size was measured in ≥ 250 Actinin-positive cells per group using ImageJ software (U.S. National Institutes of Health, Bethesda, Maryland, USA).

Mice. All animal housing, care and procedures were approved by the institutional Animal Experimental
Committee for Animal Welfare in accordance with the Directive 2010/63/EU of the European Parliament. Mice used for cardiomyocyte isolations and experimental procedures are C57Bl/6 WT or Nur77-KO mice in C57Bl/6 background, all 10-12 week-old males.
[Ca 2+ ] i transient measurements. Left ventricular cardiomyocytes from WT and Nur77-KO mice were isolated by enzymatic dissociation 44 and loaded with fluorescent Ca 2+ probe Indo-1-AM (Molecular Probes, Eugene, OR, USA) as described previously 45 . Using 4 Hz stimulation frequency, baseline [Ca 2+ ] i was measured for 1 min, followed by measurements directly upon stimulation with increasing concentrations of isoproterenol (Sigma) or norepinephrine (Centrafarm, Etten-Leur, The Netherlands). Action potential measurements. Action potentials were recorded in unstimulated, single cardiomyocytes isolated from WT and Nur77-KO mice, by the amphotericin-perforated patch-clamp technique using an Axopatch 200B amplifier (Molecular Devices, USA) as described previously 46 . Ex vivo assessment of cardiac conduction parameters. Atrio-ventricular and ventricular conduction parameters were investigated in isolated, Langendorff-perfused WT and Nur77-KO mice hearts using optical mapping, as described previously 46 . Chronic isoproterenol stimulation in vivo. Osmotic minipumps (Alzet, Cupertino, CA, USA) were implanted subcutaneously in WT or Nur77-KO mice. The procedure was performed under isoflurane anaesthesia (4% isoflurane for induction, 2.5% isoflurane and O 2 for maintenance of anaesthesia; Baxter, Illinois, United States) with subcutaneously-administered Carprofen (4 mg/kg) as a local analgetic. The osmotic minipumps continuously released isoproterenol (Sigma) dissolved in saline, in a dose of 60 mg/ kg/day. Control mice received minipumps containing saline only. After 7 days the mice were euthanized through a lethal dose of intraperitoneal-injected ketamine (238 mg/kg)/xylazine (102 mg/kg). Prior to and 6 days after implantation of the osmotic minipumps, transthoracic echocardiography was performed in isoflurane-anaesthetized mice (4% isoflurane for induction, 2.5% isoflurane and O 2 for maintenance of anaesthesia). Using the Vevo770 imaging system equipped with a 30-MHz linear array transducer (VisualSonics Inc, Toronto, Canada), a two-dimensional short-axis view of the left ventricle was obtained at the level of the papillary muscles and 2D M-mode tracings were recorded. Left ventricular ejection fraction and fractional shortening were calculated using the following formulas: LVEF% = (LVvol;d-LVvol;s)/ (LVvol;d)*100; LVFS% = (LVID;d-LVID;s)/(LVID;d)*100.
Transverse aortic constriction (TAC). TAC surgeries and subsequent cardiac function assessment was performed as described previously 47 . In brief, TAC was performed in WT or Nur77-KO mice. The procedure was performed under isoflurane anaesthesia (4% isoflurane for induction, 2.5% isoflurane and O 2 for maintenance of anaesthesia) with subcutaneously-injected Buprenorphine (0.05 mg/kg) as local and post-operative analgetic. After 28 days, 2D M-mode echocardiography (Aloka SSD 4000 echo device, 12 MHz transducer; Aloka, Tokyo, Japan) was performed under isoflurane anaesthesia, where after the mice were euthanized by 100% CO 2 ventilation.
Histological analysis. Ventricles were fixed in 4% formaldehyde for 24 h, embedded in paraffin, sectioned (7 μ m) and mounted on Starfrost glass slides (Thermo Scientific). Nur77 protein expression was assessed by immunohistochemistry using anti-Nur77 antibody (Santa Cruz) and anti-rabbit-IgG-HRP polymer (BrightVision, ImmunoLogic, Duiven, The Netherlands). Antigen detection was performed by development with diaminobenzidine tetrachloride (DAB; Immunologic) and slides were imaged using Leica QWin V3 software (Leica Mircosystems, Cambridge, UK). To assess cardiomyocyte cross-sectional area, cardiac tissue was stained with AlexaFluor488-conjugated wheat germ agglutinin (Molecular Probes, Eugene, OR, USA) and DAPI. Cardiomyocytes were included in size measurement if they were transversely cross-sectioned and revealed thin, intact cell borders and visible nuclei located in proximity of the cell centre. Per mouse, ≥ 75 cardiomyocytes were measured using ImageJ software to acquire a proper representation of the tissue. Collagen quantification was performed on Masson's Trichrome-stained (Sigma-Aldrich) sections, using Leica QWin V3 software.
Apoptotic cells were assessed by TUNEL assay using the In Situ Cell Death Detection Kit (Roche, Indianapolis, IN, USA) according to manufacturer's instructions and quantified as number of TUNEL+ cells by fluorescence microscopy.
RNA isolation and real time-PCR. Total RNA was isolated from NRVMs using the Aurum Total RNA mini kit (BioRad Laboratories, Veenendaal, The Netherlands) and total RNA from cardiac tissue was isolated using the RNeasy Fibrous Tissue Mini kit (Qiagen, Frederick, MD, USA), all according to manufacturer's instructions. cDNA was synthesized with the iScript cDNA Synthesis kit and real-time PCR was performed using iQ SYBR Green Supermix and was measured with the MyIQ system (all from BioRad). Specific primer sequences are listed in Table 1. mRNA levels were normalized for the combined value of ribosomal protein P0 (p0) and ribosomal protein L13 (rpl13) housekeeping genes.
Calcineurin activity assay. Protein was isolated from snap-frozen murine left ventricles using lysis buffer supplied in the calcineurin activity assay kit (Enzo Life Sciences, Exeter, UK). Calcineurin phosphatase activity was measured in 5 ug protein from each lysate, according to the manufacturer's instructions.
Statistical analysis. All data analyses were performed using GraphPad Prism 5 (La Jolla, CA, USA).
Data distribution was tested with Kolmogorov-Smirnov normality test. Accordingly, data is presented as mean ± SEM and tested with Student's t-test, two-way repeated measures ANOVA or Fisher's exact test. In case of a failed normality test, data is presented as boxplots with whiskers for minimum/maximum values and tested with Mann-Whitney-U test. Holm-Bonferonni or Newman-Keuls test post-hoc correction was used when appropriate.