OUP user menu

Calcineurin regulates NFAT-dependent iNOS expression and protection of cardiomyocytes: Co-operation with Src tyrosine kinase

Kofo Obasanjo-Blackshire, Rui Mesquita, Rita I. Jabr, Jeffery D. Molkentin, Stephen L. Hart, Michael S. Marber, Yang Xia, Richard J. Heads
DOI: http://dx.doi.org/10.1016/j.cardiores.2006.05.026 672-683 First published online: 1 September 2006


Objective To determine the role of calcineurin and Src tyrosine kinase in the regulation of inducible nitric oxide synthase (iNOS) expression and protection in cardiomyocytes.

Methods iNOS expression was studied in isolated neonatal rat ventricular myocyte cultures in response to bacterial lipopolysaccharide (LPS) or following transfection with constitutively active calcineurin or Src and in hearts isolated from wild-type or calcineruin Aβ knockout mice. Cell injury in response to simulated ischemia–reperfusion was studied following overexpression of active calcineurin. Regulation of the iNOS gene promoter by calcineurin was studied using promoter-luciferase reporter and chromatin immunoprecipitation assays.

Results Overexpression of constitutively active Src co-operated with [Ca2+]c elevation to induce iNOS expression, and LPS-induced iNOS expression was abrogated by pharmacological inhibition of calcineurin or tyrosine kinase. LPS also induced tyrosine kinase-dependent but calcineurin-independent phosphorylation of Src Tyr418. LPS induced myocardial iNOS expression in wild-type but not calcineurin Aβ knockout mice. Overexpression of constitutively active calcinuerin in isolated cardiomyocytes caused deposphorylation and nuclear accumulation of the c1 isoform of nuclear factor of activated T-cells (NFATc1), induced strong iNOS expression, and induced NOS-dependent protection against simulated ischemia–reperfusion prior to cardiomyocyte hypertrophy. Co-transfection of a mouse iNOS promoter-luciferase reporter in combination with active calcineurin and wild-type or dominant negative Src confirmed that constitutive activation of calcineurin was sufficient for transactivation. Chromatin immunoprecipitation confirmed calcineurin-dependent in vivo binding of NFATc1 to consensus sites within the iNOS promoter.

Conclusions These results support a cardioprotective role for calcineurin mediated by NFAT-dependent induction of iNOS expression and co-operativity between calcineurin and Src.

  • Calcineurin
  • iNOS
  • Src
  • Heart
  • Hypertrophy
  • Cardioprotection

This article is referred to in the Editorial by J. Heger and G. Euler (pages 612–614) in this issue.

1. Introduction

Calcineurin (Cn), or Protein Phosphatase 2B (PP2B), is a calcium- and calmodulin-dependent protein serine/threonine phosphatase which plays a fundamental role in inflammation, cardiac myocyte hypertrophy, ventricular remodelling and heart failure, but may also contribute to cardioprotection. Calcineurin consists of a Ca2+ and calmodulin binding catalytic A subunit (CnA), for which there are two isoforms expressed in the heart (α and β) and a regulatory B subunit which also binds Ca2+. Calcineurin regulates the expression of cytokines in T lymphocytes during the T-cell mediated immune response [1,2] and in the heart, calcineurin has recently been shown to play a fundamental role in the regulation of cardiac myocyte hypertrophy [3,4].

The regulation of cardiac hypertrophy and of interleukin-2 (IL-2) synthesis in T-cells by calcinuerin are dependent on the Nuclear Factor of Activated T-cells (NFAT) family of transcription factors [1,3]. NFATs are activated by calcinuerin through a dephosphorylation mechanism which induces nuclear translocation and binding of NFATs to elements within the promoters of genes such as IL-2 [2]. To date, NFATs represent the best described substrates for calcinuerin. The induction of gene expression by calcineurin is complex and is dependent on crosstalk between different signalling pathways. For instance, in T cells, Ca2+ signalling through calcineurin-NFAT obligatorily co-operates with other signals such as the small GTPases Ras and Rac, protein kinase C (PKC) and the non-receptor tyrosine kinase Src resulting in the activation of the transcription factors AP-1, MEF, c-Maf, GATA-4 or NFκB which heterodimerise with NFAT and stabilise its DNA binding on the IL-2 promoter [5]. Recent evidence suggests that the calcinuerin-NFAT signalling axis co-operates with extracellular signal regulated kinase (ERK)-dependent induction of the activity of AP-1 at the level of complex formation between NFAT and AP-1 on the DNA which enhances the hypertrophic growth response in the heart [6]. This observation correlates directly with the concept of a co-stimulus involving co-operativity between calcineurin and other pathways as in T-cell activation.

Little is known regarding the role of calcinuerin in cardioprotection. However, a protective role for calcinuerin is inferred from the evidence that calcineurin A beta (CnAβ) gene-targeted mice show a greater propensity for loss of viable myocardium due to cardiomyocyte apoptosis following acute ischemia–reperfusion injury [7]. Contrary to this, CnAα overexpressing transgenic mice show a rapid progression to LV dilatation, contractile dysfunction and increased apoptosis at later time points following coronary artery ligation [8], suggesting that chronic activation of calcinuerin may be detrimental. Furthermore, cardiac myocyte-specific overexpression of the calcineurin inhibitory protein Modulatory Calcineurin Inhibitory Protein-1 (MCIP-1) in transgenic mice inhibits the cardiac hypertrophy induced by cardiac myocyte-specific transgenic overexpression of constitutively active CnAα [9] and decreases post-infarction remodelling [10]. These results suggest that calcineurin may be an important positive regulator of post-infarction remodelling. Thus, the precise role of calcineurin in supporting protective signalling pathways and stress-adaptation versus activation of pathways which increase the progression to heart failure requires further detailed characterisation.

Expression of the inducible isoform of nitric oxide synthase (iNOS: NOS2) is strongly upregulated by inflammatory stimuli such as LPS and cytokines and plays a role in the depression of cardiac function in response to cytokines such as Tumour Necrosis Factor α (TNFα) and sepsis [11]. In addition, iNOS is elevated following brief myocardial ischemia–reperfusion and has been linked to cardioprotection [12,13]. Calcineurin has been indirectly linked to iNOS induction in colon epithelial cells or macrophages since LPS-induced iNOS expression and NO production are inhibited by cyclosporin A and FK506 [14]. Thus, we hypothesised that calcineurin may also increase cardiomyocyte survival via the regulation of iNOS expression.

2. Materials and methods

2.1. Cardiomyocyte culture

Primary cultures of cardiomyocytes were prepared from neonatal rats as previously described [15] and cultured on 6-well plates (Greiner) at an initial density of 1 × 106 cells per well. Cardiomyocytes were treated with ionomycin (1μM in 0.1% methanol) or cyclopiazonic acid (30μM in 0.1% methanol) for 2h; with PMA (100nM in 0.1% DMSO) for 2h or with bacterial lipopolysaccharide (LPS, Sigma) at 10μg/ml for 3h. For inhibitor experiments, cardiomyocytes were treated with the PKC inhibitor GF109203X (Calbiochem, UK) at 2μM (0.1% DMSO); SrcTK inhibitor SU6656 (Calbiochem, UK) at 10μM (1:1000 dilution of 10mM stock in DMSO); geldanmycin (Calbiochem) at 10μM; PP2 (Calbiochem) at 10μM; JNK inhibitor SP600125 (Calbiochem) at 2μM (1:5000 dilution of 10mM stock in DMSO); calcineurin inhibitor cyclosporin A (CsA) at 5μM (in 1% ethanol) and NOS inhibitor amino guanidine (Calbiochem) at 200μM (in H2O). Vehicle treatments had no effect on iNOS expression in response to LPS. Cardiomyocytes were treated with inhibitors for 30min prior to LPS treatment. After 3h of LPS treatment cardiomyocytes were returned to normal maintenance media.

2.2. Calcineurin Aβ gene targeted mice.

Calcineurin Aβ knockout mice have been described previously [16,7]. Female − / − mice were mated with male +/ − mice to produce heterozygous (+/ −) progeny. Breeding pairs were set up which generated wild-type (+/+), knockout (− / −) and heterozygous (+/ −) progeny. Genotypes were confirmed by PCR and male +/+ and − / − mice were used for the experiments. Animals were kept in accordance with UK Home Office guidelines and the study conformed to the NIH Guide for the Care and Use of Laboratory Animals.

2.3. Plasmids and transfections

Transfections were performed using a non-viral integrin targeting peptide as previously described [17], which utilises an integrin targeting peptide ([K]16-GACRRETAWACG) (see online methods supplement for details). Transfection efficiency was approximately 20% (see online data supplement, Fig. 1). Cells were transfected with pCAGGS-GFP, pEF (null vector), pEF-c-SrcF527 (ca-Src); pSGT (null vector), pSGT-c-SrcA295/F527 (dn-Src), pSRα (null vector), pSRα-ca-CnAα (ca-CnA) or pEF-N17Ras (ca-Ras) as appropriate. pCAGGS-GFP was made in our laboratory; pEF-SrcF527 was from Dr. Richard Marais (Imperial College London, UK); pSGT-c-SrcA295/F527 was from Professor Jorge Martin-Perez (Madrid, Spain), pEXV3-N17Ras was from Dr. Stephen Fuller, (Imperial College London); pSRα-ca-CnA (δCamAI: ca-CnA) was from Dr. Stephen O'Keefe (Merck, New Jersey, USA) [18]. Adenoviral expression of ca-CnA has been described previously [8].

Fig. 1

Activation of NFAT and induction of iNOS expression by a Ca2+ and Src co-stimulus. Neonatal rat cardiomyocytes were treated with vehicle (methanol or DMSO), ionomycin (1μM for 2h) or PMA (100nM for 2h) or transfected with active Src (ca-Src); active calcineurin A (ca-CnA) or empty vectors and isolated for Western blotting with anti-NFATc1 (panel A), anti-iNOS (panels B and C) anti-pp60Src or anti-CaN antibodies (panel D). In panel C, cardiomyocytes were transfected with active Src (ca-Src) alone or then treated with ionomycin (IM: 1μM for 2h) or the SR Ca2+ uptake inhibitor cyclopiazonic acid (CPA: 30μM for 2h). Panel D shows ectopic expression of active CnA (ca-CnA) and active Src (ca-Src) compared to empty vector (null). The lane marker ‘C’ corresponds to control untransfected or untreated as appropriate to the experiment.

2.4. Luciferase assays

Cardiomyocytes were transfected with 5xNFkB-luc, 4xNFAT-luc (Stratagene) or an iNOS promoter-luc construct (Fig. 7, panel A), either alone or co-transfected with empty plasmid (control) or ca-CnA. In order to control for differences in transfection effeiciency 0.2μg Renilla luciferase plasmid was also included in each transfection. A dual luciferase assay was performed (Promega) using a luminometer (TD 20/20, Turner Designs) and activity was normalised to Renilla activity. The proximal iNOS promoter construct (full length) in pGL2-basic was from Dr. Charles Lowenstein (Johns Hopkins University, Baltimore, USA).

Fig. 7

Calcineurin activates the iNOS promoter through NFAT binding. The structure of the proximal iNOS promoter is represented schematically in panel A. The main regulatory sites bound by the transcription factors NFAT, NFκB, AP-1 are shown. In Panel B cardiomyocytes were transfected with the full length proximal promoter linked to a luciferase reporter gene and the response to expression of ca-CnA determined. Luciferase activity is expressed as fold activation normalised to a co-transfected Renilla luciferase plasmid under control of the TK-promoter. Values are expressed relative to control (normalised to 1) and are means+S.E.M. of three independent transfection experiments (**p≤0.005 vs control by ANOVA, n=3). Expression of a dominant negative (DN) Src mutant following plasmid transfection is shown in panel C. Expression of GFP is shown for comparison. Panel D shows activation of the iNOS promoter-luciferase reporter following co-transfection with null plasmid (pSRα) or ca-CnA together with null (pEF), wild-type (WT) or DN Src. Other conditions are as for panel B. The results represent a single batch of cardiomyocytes with each transfection performed in triplicate. In vivo binding of NFATc1 to the iNOS promoter is shown in panel E. Isolated neonatal rat cardiomyocytes were infected with adenovirus encoding active CnA (ca-CnA) or empty control virus (SR), harvested after 48h and subjected to chromatin immunoprecipitation assay. DNA complexes were immunoprecipitated either with an antibody against NFATc1 (N) or a control antibody against GAPDH (C). Immunoprecipitated complexes were subjected to RT-PCR using primers designed to amplify the regions of the mouse iNOS promoter containing the proximal NFAT sites between − 503 and − 350, − 1000 and − 877 or − 1588 and − 1417. PCR products were resolved on 1% agarose gels and gave rise to 153bp, 123bp and 171bp, respectively.

2.5. Simulated ischemia

Following transfection, simulated ischemia (SI) was induced for 3h as previously described [15] (and see online methods supplement). After 1.5h of ‘reperfusion’ the supernatant was collected to assess cell membrane damage by measuring creatine phosphokinase (CPK) efflux using a CPK assay kit (CK-20, Sigma). After 24h of ‘reperfusion’ cell viability was determined using an MTT bioreduction assay (see online methods supplement for details). Data were pooled from triplicate wells and from three independent experiments and statistical analysis performed using ANOVA. Results were expressed as mean±S.E.M.

2.6. Immunoblotting

Electrophoresis and Western blotting was performed essentially as described previously [15]. Membranes were probed with polyclonal anti-iNOS C-terminal (610333, BD Transduction Labs; polyclonal anti-c-Src (Santa Cruz); monoclonal anti-NFATc1 (7A6, Santa Cruz); polyclonal anti-GFP (Living Colors, Clontech) or monoclonal anti-CnA (CN-A1, Sigma). Antibody–antigen complexes were visualised by enhanced chemiluminescence (ECL, Amersham).

2.7. Immunostaining and fluorescence microscopy

Neonatal cardiomyocytes were cultured on permanox slides (Nalge Nunc International) coated with gelatin. Following treatment slides were washed, fixed, permeabilised and probed with monoclonal anti-NFATc1 or α-actinin (Sigma) and co-labelled with the nuclear dye To-Pro and analysed by confocal microscopy (see online methods supplement).

2.8. Chromatin immunoprecipitation (ChIP) assay

Potential NFAT binding sites were identified by analysis of the promoter using the Transfac MatInspector software [19]. ChIP was carried out essentially as described in [20] (and see online methods supplement). Neonatal rat cardiomyocytes were infected with adenovirus encoding constitutively active calcineurin (ca-CnA) or control virus (SR-) and immunoprecipitation performed using either rabbit polyclonal anti-NFATc1 (H-110: Santa Cruz, USA) or rabbit polyclonal anti-GAPDH (FL-355: Santa Cruz, USA). The primers used in the PCR to amplify the promoter regions containing the NFAT sites were set 1 (− 1588 to − 1417): forward 5′-GAC TTT GAT ATG CTG AAA TC-3′ reverse 5′-CAG CCT AGC CTA CTA GGC AGG T-3′; set 2 (− 1000 to − 877): forward 5′ − ATG AGT GGA CCC TGG CAG GA − 3′ reverse 5′-GCT TCC AAT AAA GCA TTC ACA-3′ and set 3 (− 503 to − 350): forward 5′-CTA TTC TGC CCA AGC TGA CTT-3′ reverse 5′-TTC TCT CAG TGA GGT TAG ATG-3′.

2.9. Statistical analysis

For experiments on isolated cardiomyocytes, experiments were performed on at least three cell preparations (unless otherwise stated) and treatments performed in triplicate wells for each preparation. Data were pooled and expressed as mean±S.E.M. (n=3). Groups were compared by one-way analysis of variance (ANOVA) with Tukey's post hoc test using Graphpad Prism software.

3. Results

Sustained elevation of [Ca2+]i activates calcineurin resulting in dephosphorylation of the transcription factor NFAT. Therefore, activation of endogenous calcineurin was examined by treatment of cardiomyocytes with the Ca2+ ionophore ionomycin and ectopic sustained calcineurin activation by transfection with a constitutively active calcineurin mutant (ca-CnA). Calcineurin activation was confirmed by an increase in the mobility of NFATc1 on Western blots which was consistent with NFAT dephosphorylation (Fig. 1, panel A). Three bands were observed for NFATc1 which correspond to the NFATc1/A, NFATc1/B and NFATc1/C variants which arise from alternate promoter usage/alternative mRNA splicing [21]. The constitutive activation of calcineurin was achieved by deletion of the calmodulin binding and auto-inhibitory domains [18] which removes the requirement for both Ca2+ and calmodulin. As shown in Fig. 1 panel B transfection of cardiomyocytes with ca-CnA resulted in strong iNOS induction. However, although calcineurin was activated by ionomycin as evidenced by NFATc1 dephosphorylation, there was no induction of iNOS 24h following ionomycin pretreatment alone (Fig. 1, panel B).

Therefore, we tested the possibility that iNOS induction by transient activation of endogenous calcineurin is dependent on co-operativity with other signaling pathways. As shown in Fig. 1 panel B, treatment of cardiomyocytes with the phorbol ester, phorbol myristoyl acetate (PMA) plus ionomycin (but not PMA or ionomycin alone) caused synergistic iNOS induction suggesting co-operativity between Ca2+ and a phorbol ester sensitive signaling pathway (i.e., PKC and/or Src/Ras/Raf). Whereas ca-CnA alone was sufficient for iNOS induction, in contrast ectopic expression of ca-Src by itself did not induce iNOS expression (Fig. 1, panel C). However, expression of ca-Src followed by treatment with ionomycin or the SR calcium uptake inhibitor cyclopiazonic acid resulted in a synergistic action on iNOS expression which was near maximal (Fig. 1, panel C). Co-expression of ca-Src with ca-CnA resulted in slightly potentiated iNOS expression compared to ca-CnA alone, suggesting that constitutive activation of CnA was sufficient for near maximal iNOS expression. Overexpression of the ca-CnA or constitutively active Src (ca-Src) mutant proteins following transfection was confirmed by Western blotting (panel D).

In order to determine whether both calcineurin, PKC or Src play a role in endogenous iNOS induction in response to inflammatory stimuli, we used LPS treatment, since LPS has been reported to induce potent iNOS expression in cardiomyocytes [11]. As shown in Fig. 2, LPS induced iNOS expression in neonatal cardiomyocyte cultures 24h following treatment. LPS-induced iNOS expression was inhibited by pretreatment with the selective calcineurin inhibitor cyclosporin A and abolished by the selective PKC inhibitor GF109203X (panels A and D). Furthermore, the non-receptor tyrosine kinase (NRTK) inhibitors geldanamycin (GA), PP2 (panels B and E) and the Src tyrosine kinase-selective inhibitor SU6656 (panels C and F) also inhibited iNOS expression in response to LPS, indicating an obligatory requirement for one or more Src family NRTKs in iNOS expression. PP2 inhibited iNOS by approximately 50% compared to SU6656, possibly indicating the preferential use of a particular isotype such as Lck (which is more sensitive to SU6656) in response to the agonist in vivo. In contrast to geldanamycin or SU6656, PP2 appeared to slightly decrease actin levels without any apparent affect on cell viability; therefore, for PP2, iNOS levels were compared to GAPDH which was invariant (panels B and E). Src activation by LPS was confirmed by analysis of Src Tyr418 phosphorylation (Fig. 2, panel G). Tyr418 in the activation loop of Src is autophosphorylated and is indicative of Src activation. Tyr418 phosphorylation was blocked by GA but not by CsA, indicating that Src is activated in parallel to calcineurin in response to LPS. iNOS expression was unaffected by pre-treatment with the JNK inhibitor SP600125 (panels C and F). Thus, activation of calcineurin, PKC and Src family NRTKs appear to contribute to iNOS expression in vivo.

Fig. 2

Endogenous iNOS expression is calcineurin, PKC and Src-dependent. Cardiomyocytes were treated with 10μg/ml LPS for 3h and iNOS expression determined 24h later. Where indicated, cells were pretreated with the calcineurin inhibitor cyclosporin A (CsA: 5μM) or the PKC inhibitor GF109203X (GF: 2μM) (panels A and D); the Src family tyrosine kinase inhibitors PP2 (10μM) or geldanamycin (GA: 10μM) (panels B and E) and the Src family tyrosine kinase inhibitor SU6656 (SU: 10μM) or the JNK inhibitor SP600125 (SP: 2μM) (panels C and F). iNOS expression was normalised to the actin (or GAPDH: panels B and E) level following densitometry and represented quantitatively in panels D–F as mean±S.E.M. (n=3). **p≤0.01; ***p≤0.001 vs vehicle and † †p≤0.01; † † †p≤0.001 vs LPS; ANOVA). Cardiomyocytes were harvested 2h after LPS treatment and analysed for Src Tyr418 phosphorylation by Western blotting (panel G). Total Src is shown below for comparison. Where indicated, cells were pre-treated with CsA (5μM) or GA (10μM). The lane marker ‘C’ corresponds to control untreated.

We next determined whether iNOS expression is regulated in vivo in response to a pro-inflammatory stimulus in hearts where CnA activity is altered. CnAβ has been reported to be the predominant isoform detected in myocardium [7]. Therefore, we analysed hearts from mice in which CnAβ was knocked out by homologous recombination for iNOS expression in response to intra-peritoneal LPS injection compared to wild-type littermate controls. As shown in Fig. 3 (panels A and B) iNOS levels were significantly increased by LPS in hearts from wild-type mice but not in hearts from CnAβ gene targeted mice.

Fig. 3

Knockout of calcineurin Aβ abrogates cardiac iNOS induction in vivo. Hearts were isolated from wild-type (WT) mice and mice in which CnAβ has been knocked out by homologous recombination (KO) following pre-treatment with LPS for 24h. Proteins were extracted and subjected to SDS-PAGE and Western blotting (panel A). WT (wild-type); KO (knockout). iNOS expression was normalised to the actin level following densitometry and represented quantitatively in panel B as mean±S.E.M. (n=3). **p≤0.01 vs vehicle; ANOVA. The lane marker ‘C’ corresponds to control untreated. ‘+ve’ corresponds to positive control for iNOS expression (LPS treated primary cardiac fibroblasts).

Transient upregulation of iNOS expression in cardiomyocytes following delayed ischaemic preconditioning is protective against both infarction and stunning [22]. However, it is not known whether iNOS induction following a systemic inflammatory response in vivo or in vitro can induce cross-tolerance against ischemia–reperfusion injury. Therefore, since the above results demonstrate that iNOS induction in isolated cardiomyocytes by LPS is calcineurin-dependent and that ectopic expression of ca-CnA induced strong iNOS expression (see Figs. 1–3) we tested whether the latter had any effect on survival of neonatal rat cardiomyocytes against simulated ischemia/reperfusion injury. Fig. 4 (panel A) shows that expression of ca-CnA but not the control GFP-expressing construct induced iNOS expression. When cardiomyocytes were subjected to simulated ischemia/reperfusion (sI/R), there was a reduction in cell injury as determined by a lower release of creatine phosphokinase (CPK) into the supernatant in cells transfected with ca-CnA but not with the GFP control. Furthermore, cell viability as determined by MTT bioreduction was also significantly increased by ca-CnA. This protective effect was reversed by the NOS inhibitor amino guanidine, suggesting that the calcineurin-induced protection was NO-dependent and that calcineurin activation protects against sI/R injury.

Fig. 4

Activated calcineurin induces iNOS expression and protection. Cardiomyocytes were transfected with green fluorescent protein (GFP), empty vector (null) or active calcineurin A (ca-CnA) and isolated for Western blotting (panel A) or 48h later subjected to 3h of simulated ischemia and ‘reperfusion’ (SI/R) (panels B and C) either alone or following pretreatment with the iNOS inhibitor amino guanidine (AG: 200μM) given 30min prior to and during SI/R. Following SI/R creatine phosphokinase (CPK) release was determined and is expressed as units per litre (U/l) (panel C). Cell viability was assessed as methyl thiazoyl tetrazolium (MTT) bioreduction and expressed as arbitrary units (absorbance at 570nM) (panel B).

Constitutive activation of calcineurin was able to activate NFAT and substitute for a co-stimulus and induced a high level of iNOS expression (as shown in Fig. 1). Therefore, we hypothesised that high sustained calcineurin activity prevents nuclear export of NFAT, thus maintaining NFAT in the nucleus in a DNA-binding-competent form. To test this possibility, we overexpressed the same ca-CnA mutant in both isolated neonatal and adult cardiomyocytes using adenovirally mediated expression and determined the subcellular location of NFATc1 using immunostaining and confocal microscopy. As shown in Fig. 5 NFATc1 expression in control cells (transduced with empty virus) was uniformly distributed. In contrast, in ca-CnA transduced neonatal cardiomyocytes NFATc1 strongly co-localised with the nuclear dye To-Pro. This was also the case for adult cardiomyocytes (see online data supplement, Fig. 2). This result confirms that constitutive activation of calcineurin is sufficient to maintain NFAT in the nucleus in the absence of a co-stimulus.

Fig. 5

Over-expression of active calcineurin results in constitutive activation of NFATc1. Neonatal cardiomyocytes were infected with empty (control) adenovirus (SR) or adenosvirus encoding active CnA (ca-CnA). Cells were then fixed and immuno-stained for NFATc1 (panel A) or the nuclear dye To-Pro (panel B) and analysed using confocal fluorescence microscopy.

In order to examine whether calcineurin regulates iNOS gene transcription through NFκB or NFAT, ca-CnA was co-transfected with luciferase reporter constructs containing consensus recognition sites for NFκB or NFAT. As shown in Fig. 6 (panel A), the NFκB reporter was strongly activated by MEKK1, an upstream kinase regulating JNK and NFκB kinase, but not by ca-CnA. Surprisingly, co-transfection of MEKK with ca-CnA resulted in strong inhibition of NFκB-luc activation. Since iNOS was induced by PMA+ionomycin and NFkB can be activated by the PKC-Ras-Raf signalling axis, we tested whether NFkB-luc was activated by co-transfection with a ca-Ras mutant (RasN17). As can be seen in Fig. 6 (panel B) NFκB-luc was strongly activated by ca-Ras, indicating that the Ras/Raf pathway can also act upstream of NFκB. Co-transfection of ca-Ras with ca-CnA also resulted in a significant reduction of the Ras-mediated NFκB-luc activation. Therefore, it seems unlikely that calcineurin mediates its effects on iNOS expression in cardiomyocytes through NFκB and that calcineurin may even be inhibitory with respect to NFκB-dependent effects.

Fig. 6

Calcineurin represses activation of an NFkB-dependent reporter gene and activates an NFAT-dependent reporter gene. Cardiomyocytes were transfected with luciferase reporter gene plasmids containing consensus NFκB (panels A and B) or NFAT (panel C) binding sites. These were co-transfected with plasmids expressing active calcineurin (ca-CnA), active MEKK1 or active Ras. Luciferase activity is expressed as fold activation normalised to a co-transfected Renilla luciferase plasmid under control of the TK-promoter. Values are expressed relative to control (NFkB-luc) and are means+S.E.M. of three independent transfection experiments (panels A and C: *p≤0.05 vs control; panel B: **p≤0.005 vs control and p≤0.05 vs ca-Ras; panel C: p≤0.05 vs ca-CnA and §p≤0.05 vs ca-CnA by ANOVA, n=3).

In contrast, ca-CnA strongly activated the NFAT-luc reporter (Fig. 6, panel C). In addition, ca-Ras also caused a 2.5-fold activation of this reporter, and the effects of ca-CnA and ca-Ras on NFAT-dependent transcription were additive. Since this reporter only contains NFAT binding sites and not NFκB or AP-1 sites, it is likely that the Ras pathway can activate NFAT, most likely through direct phosphorylation of the NFAT transactivation domain via PKC and/or ERK as previously described [6].

Since the previous experiments demonstrated that ca-CnA substituted for the requirement for a co-stimulus we analysed the ability of ca-CnA to activate the mouse iNOS promoter in co-transfection experiments using a 1.7kb proximal promoter spanning from +161 to − 1588 linked to a luciferase reporter gene. Analysis of this proximal region using a computer program to identify potential consensus transcription factor binding sites revealed two potential NFκB, two AP-1 and five NFAT DNA binding sites (represented schematically in Fig. 7, panel A).

The proximal iNOS promoter construct was co-transfected with empty vector or ca-CnA. As shown in Fig. 7 (panel B), the promoter was activated approximately 12-fold by ca-CnA. These results suggest that proximal NFAT sites contribute significantly to activation by calcineurin. Src appeared to be activated in parallel to calcineurin in response to LPS. To further confirm this we tested whether Src plays a role in the calcinuerin-dependent activation of the iNOS promoter. Cells were co-transfected with the iNOS-promoter-luc and plasmids expressing either wild-type (wt) or dominant negative (dn) mutant of Src in the presence or absence of ca-CnA (Fig. 7C and D). ca-CnA alone activated the promoter and was slightly potentiated by wt-Src. This modest potentiation was not observed in the presence of dn-Src, but the predominant effect of activation by ca-CnA was overall unaffected. These results indicate that Src-dependent activation is additive to the calcineurin-dependent activation of the iNOS promoter. Furthermore, when constitutively active, CnA does not require a Src co-stimulus to achieve high level transcriptional activation and that Src is not downstream of calcineurin.

To further delineate whether NFAT is associated with these sites within the iNOS promoter in intact cells, chromatin immunoprecipitation assays (ChIP) were performed. Cardiomyocytes were infected with ca-CnA expressing adenovirus and compared to control (empty) virus. ChIP was performed with antibodies directed against NFATc1 or GAPDH (control antibody). RT-PCR was performed on the immunoprecipitates using primers generated to the three regions spanning the putative NFAT binding sites (see materials and methods). The results shown in Fig. 7D demonstrate that in vivo binding of NFATc1 occurred in response to ca-CnA overexpression in the three regions containing NFAT sites corresponding to (− 530 to − 350), (− 1000 to − 877) and (− 1588 to − 1417). These results suggest that proximal NFAT binding sites are important for calcineurin-dependent activation of the iNOS promoter.

Over-expression of active calcineurin in transgenic mice or isolated cardiomyocytes induces cardiomyocyte hypertrophy [3]. Therefore, we examined the time-course of hypertrophy induction in relation to the observed induction of iNOS and cytoprotection. Cardiomyocytes were infected with empty virus (SR-) or Ad-CnA, fixed and stained for α-actinin at different time points post-infection, analysed by confocal microscopy and cell surface area determined. Fig. 8 shows that there was a sharp and dramatic increase in cell size at 3days post-infection. Since iNOS expression and cytoprotection were observed at 24h post-infection, the protective, adaptive response occurred prior to and in the absence of overt hypertrophy.

Fig. 8

Over-expression of active calcineurin induces a late hypertrophic response. Neonatal cardiomycoytes were infected with empty adenovirus (SR-) or adenovirus expressing ca-CnA (AdCnA). At different time points following infection cardiomyocytes were labelled with antibodies for alpha-actinin (panel A) or the nuclear dye To-Pro and analysed using confocal fluorescence microscopy. Cardiomyocyte surface area was determined by computer planimetry and expressed quantitatively (panel B).

4. Discussion

In this study we have demonstrated that calcineurin plays an important role in the regulation of iNOS gene expression in cardiomyocytes. A number of studies in various cell types have reported that the calcineurin inhibitor cyclosporine A (CsA) inhibits iNOS expression [14,23]. However, only one of these studies [23] demonstrates a direct involvement of calcineurin (in macrophages) and none demonstrates a direct involvement of NFAT in the regulation of the iNOS gene promoter. We have further shown that calcineurin functions to regulate iNOS expression through the NFATc1 isoform and that this pathway interacts with a Src-TK signalling axis which has a co-operative effect on iNOS expression. This co-operative effect observed in cardiomyocytes appears to parallel the co-operative effect of calcineurin and PKC in the regulation of IL-2 and IL-17 [24,25] or COX-2 [26] gene expression following T-cell activation. This can be mimicked by treatment of T-cells with PMA and a Ca2+ ionophore [27]. Furthermore, Src-family tyrosine kinases can co-operate with calcineurin in the induction of IL-2 gene transcription during T-cell activation via NFAT/AP-1 [28] and with calcineurin and PKC in angiotensin-II-dependent tyrosine phosphorylation of p130cas in cardiac muscle [29]. Our results demonstrate that a calcineurin/Src co-stimulus is also operative in cardiomyocytes with respect to iNOS expression.

In cells overexpressing constitutively active calcineurin, iNOS expression was at or near maximal, suggesting that this is sufficient to maintain and/or stabilise the calcineurin-dependent nuclear signalling. Calcineurin-dependent gene expression is mediated by the activation of the NFAT family of transcription factors. Calcineurin activation results in dephosphorylation of SP repeat sequences adjacent to the NFAT nuclear localisation sequence (NLS) within the regulatory domain which results in nuclear translocation of most NFATs [5]. These NFAT SP repeats can be rapidly re-phosphorylated by a number of kinases including JNK, p38-MAPK and GSK3β which results in rapid re-export of NFAT from the nucleus. Therefore, NFAT itself results in weak and transient DNA binding unless stabilised by interaction with other DNA binding factors. NFATs have been shown to interact with a variety of other transcription factors including AP-1, MEF2 and GATA-4. The interaction with MEF2 and GATA-4 are particularly important in the regulation of cardiomyocyte hypertrophy (reviewed in [30]). Thus, NFATs can act as signal integrators on the promoters of certain genes. Our results demonstrate that overexpression of active calcineurin results in an increase in the nuclear accumulation of NFATc1. Thus, the maintenance of this NFAT signal by sustained calcineurin activity may be sufficient to activate iNOS expression in the absence of a co-stimulus and independently of interaction with other transcription factors. The physiological relevance of sustained NFAT activity in response to constitutive activation of CnA may be that proteolytic cleavage of CnA under pathological conditions results in a similar removal of the C-terminal auto-inhibitory and CaM binding domains, for instance, in failing human myocardium [4]. Our results suggest that Src is activated in parallel to calcineurin, rather than in a canonical sequence, which under normal conditions may result in activation of another factor which interacts with and stabilises NFAT on the iNOS promoter. For instance, AP-1 has been demonstrated to form dimers with NFAT and to mediate the PKC/calcineurin co-stimulatory effects on the COX-2 promoter [26]. Alternatively, direct, C-terminal phosphorylation of NFATc3 has been shown to enhance its DNA binding activity [6]. Thus, activation of Src TK signaling may increase NFAT DNA binding activity either directly or indirectly.

We have demonstrated direct binding of NFATc1 to the proximal iNOS promoter following calcineurin activation and that calcineurin-dependent effects on the iNOS promoter appear to be mediated by NFAT. Furthermore, LPS-dependent induction of iNOS expression in vivo was abolished by knockout of CnAβ by homologous recombination. This demonstrates that calcineurin is necessary for iNOS expression in the heart in response to LPS.

Our results suggest that calcineurin-dependent iNOS induction occurs primarily via an NFAT-dependent mechanism and that calcineurin appeared to be inhibitory to MEKK-dependent activation of an NFκB-dependent reporter gene. The binding of NFATc1 to proximal sites in the iNOS promoter suggests that the NFAT binding sites centred around − 1475, − 1000 to − 902 and − 470 and are important for calcineurin-dependent activation. Furthermore, the κ3-like site situated at − 1443 may be important for binding NFAT dimers, as has been previously described for κ3 sites which can bind p50/Rel dimers or NFAT dimers [31].

Induction of iNOS expression by ectopic ca-CnA expression was associated with a NOS-dependent protection against simulated ischemia-reperfusion injury, further supporting a cardioprotective role for calcineurin in addition and in parallel to its role in the induction of cardiomyocyte hypertrophy. Following overexpression of caCnA the hypertrophic response occurred late (>72h) whereas iNOS induction and NO-dependent protection occurred prior to hypertrophy (>24h). Interestingly, NO-dependent cyclic guanosine monophosphate (cGMP) signalling has been shown to inhibit hypertrophy through the activity of protein kinase G I (PKGI) upstream of calcineurin-dependent NFAT activation (activated downstream of L-type Ca2+ channel activation) [32]. This suggests that iNOS expression may be anti-hypertrophic and promote functional depression [32,33]. In agreement with this we have recently shown iNOS–NO-dependent depression of Ca2+ transient amplitude in conjunction with increased cardiomyocyte survival in response to IL-6 [34].

Our results demonstrate that iNOS induction by LPS was calcineurin-dependent and that calcineurin-dependent iNOS induction induced cross-tolerance to ischemia–reperfusion injury. This may be physiologically relevant in the context of septic shock and myocardial infarction because microcirculatory collapse during sepsis results in tissue hypoxia, metabolic and energetic imbalances and organ (including myocardial) dysfunction [35]. NOS signalling (including iNOS-derived NO) feeds back negatively on β-adrenergic and Ca2+-dependent positive inotropic responses [33] and thus upregulation of iNOS under pathological conditions contributes to myocardial contractile depression, stunning and hibernation, but may act in a negative feedback loop to limit hypertrophy and increase cardiomyocyte survival.

In conclusion, our results suggest that calcineurin mediates an early adaptative protective response through iNOS–NO which occurs before cardiac hypertrophy. Whether or not calcineurin–dependent iNOS upregulation ultimately contributes to decompensation and heart failure following chronic activation remains to be determined.


This study was supported by Grant PG2001066 (KOB) from The British Heart Foundation and Fundação para a Ciênca e a Tecnologia (FCT), Portugal (RM).

Appendix A. Supplementary data

Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.cardiores.2006.05.026.


  • Time for primary review 26 days


  1. [1]
  2. [2]
  3. [3]
  4. [4]
  5. [5]
  6. [6]
  7. [7]
  8. [8]
  9. [9]
  10. [10]
  11. [11]
  12. [12]
  13. [13]
  14. [14]
  15. [15]
  16. [16]
  17. [17]
  18. [18]
  19. [19]
  20. [20]
  21. [21]
  22. [22]
  23. [23]
  24. [24]
  25. [25]
  26. [26]
  27. [27]
  28. [28]
  29. [29]
  30. [30]
  31. [31]
  32. [32]
  33. [33]
  34. [34]
  35. [35]
View Abstract