Activation of CaMKIIδC Is a Common Intermediate of Diverse Death Stimuli-induced Heart Muscle Cell Apoptosis*

Ca2+-calmodulin-dependent protein kinase II (CaMKII) is expressed in many mammalian cells, with the δ isoform predominantly expressed in cardiomyocytes. Previous studies have shown that inhibition of CaMKII protects cardiomyocytes against β1-adrenergic receptor-mediated apoptosis. However, it is unclear whether activation of CaMKII is sufficient to cause cardiomyocyte apoptosis and whether CaMKII signaling is important in heart muscle cell apoptosis mediated by other stimuli. Here, we specifically enhanced or suppressed CaMKII activity using adenoviral gene transfer of constitutively active (CA-CaMKIIδC) or dominant negative (DN-CaMKIIδC) mutants of CaMKIIδC in cultured adult rat cardiomyocytes. Expression of CA-CaMKIIδC promoted cardiomyocyte apoptosis that was associated with increased mitochondrial cytochrome c release and attenuated by co-expression of Bcl-XL. Importantly, isoform-specific suppression of CaMKIIδC with the DN-CaMKIIδC mutant similar to nonselective CaMKII inhibition by the pharmacological inhibitors (KN-93 or AIP) not only prevented CA-CaMKIIδC-mediated apoptosis but also protected cells from multiple death-inducing stimuli. Thus, activation of CaMKIIδC constitutes a common intermediate by which various death-inducing stimuli trigger cardiomyocyte apoptosis via the primary mitochondrial death pathway.

CaMKII family encoded by four different genes (␣, ␤, ␦, and ␥). The ␦ isoform of CaMKII family is predominantly expressed in the heart of many mammalian species, including humans (2)(3)(4)(5). Two primary splicing variants of the ␦ isoform, CaMKII ␦B and CaMKII ␦C , have been cloned from rat heart (2). Compared with CaMKII ␦C , CaMKII ␦B has an additional 11-amino acid sequence that contains a nuclear targeting signal (6). Thus, CaMKII ␦B and CaMKII ␦C localize to the nuclear and the cytosolic compartments, respectively, in cardiac myocytes (6).
Over the past decade, a number of studies have been focused on the role of CaMKII (without reference to specific isoforms) in regulating cardiac excitation-contraction coupling. Most if not all of the characterized CaMKII target proteins in the heart are involved in the regulation of intracellular Ca 2ϩ and excitation-contraction coupling. These include sarcoplasmic reticulum (SR) Ca 2ϩ release channels known as the ryanodine receptor (RyR2) (7), SR Ca 2ϩ pump and its regulator phospholamban (PLB) (8,9), and sarcolemmal L-type Ca 2ϩ channels (10 -13). Phosphorylation of these substrates plays an essential role in regulating diverse and important cardiac functions such as the well established frequency-dependent accelerations in contractile relaxation and Ca 2ϩ transient decay (14,15) and cardiac pacemaker activity (16). However, CaMKII activation might be abnormally enhanced under certain pathological conditions that disrupt intracellular Ca 2ϩ mobilization, such as cardiac ischemia, acidosis, myocardial infarction, and excessive ␤-adrenergic receptor (␤AR) stimulation. Because of its positive feedback biochemical nature (i.e. autophosphorylation) (1), the exaggerated CaMKII activation or expression has been implicated in the pathogenesis of arrhythmia (17,18), cardiac hypertrophy (19 -22), and cardiomyopathy (5,(22)(23)(24).
It has been demonstrated that CaMKII is persistently activated during sustained ␤AR stimulation in cardiac myocytes both in cell culture and in vivo (25)(26)(27) and that this activation is part of the signaling pathway relaying ␤ 1 AR-evoked apoptotic signals in the heart (25). In this model, inhibition of CaMKII prevents ␤ 1 AR-induced cardiac myocyte apoptosis, whereas overexpression of CaMKII ␦C enhances the ␤ 1 AR pro-apoptotic effect (25). It is also noteworthy that the cytosolic isoform, CaMKII ␦C , is selectively up-regulated in a rabbit heart failure model (24), whereas there is little or no change in the expression or activity of the nuclear isoform (CaMKII ␦B ). Despite these studies implicating CaMKII ␦C in the ␤ 1 AR-induced cell death, it is currently unclear whether activation of the cytosolic isoform CaMKII ␦C is sufficient to cause heart muscle cell apoptosis and whether other death-inducing stimuli require CaMKII activity to cause cell death. Because apoptosis is a key cause factor of chronic heart failure, the identification of fundamental molecular players involved in heart muscle cell loss has became, in the past decade, an important research focus in the field of cardiovascular biology and medicine.
The goals of the present study were (a) to determine whether activation of the cardiac cytosolic isoform, CaMKII ␦C , is sufficient to trigger cardiac myocyte apoptosis and, if so, (b) to determine whether CaMKII ␦C is a necessary intermediate in the death signaling caused by stimuli other than chronic stimulation of ␤ 1 ARs. To address these questions, we used cultured adult rat cardiac myocytes in conjunction with adenoviral gene transfer of constitutively active (CA-CaMKII ␦C ) or dominant negative (DN-CaMKII ␦C ) mutants of CaMKII-␦C to elevate or suppress CaMKII ␦C activity, respectively. We demonstrate that enhanced activation of CaMKII ␦C is sufficient to trigger robust cardiac myocyte apoptosis via activating the mitochondrial apoptotic pathway and that inhibition of CaMKII protects cardiomyocytes not only from ␤ 1 AR-induced apoptosis, as previously reported (25), but from a number of other death-inducing stimuli as well. These findings demonstrate that activation of CaMKII ␦C is an important common intermediate in the death signaling pathway of many different stimuli that induce apoptosis in heart muscle cells.

EXPERIMENTAL PROCEDURES
Construction of Viral Vectors-HA-tagged constitutively active CaMKII␦ C (CA-CaMKII) was generated by replacing the residue Thr 287 with Asp (T287D) using the transformer site directed mutagenesis kit (Clontech), whereas dominant negative CaMKII␦ C (DN-CaMKII) was generated by replacing the residue Lys 43 with Ala (K43A), as described previously (27). Adenovirus expressing WT-CAMKII ␦C (25) was kindly provided by Dr. Joan Heller Brown at the Department of Pharmacology, University of California San Diego. The generation and amplification of adenoviruses harboring the target gene were performed in HEK293 cells (27). Recombinant replication-adenovirus expressing Bcl-X L was obtained from the University of Pittsburgh NHLBI, National Institutes of Health Pre-clinical Vector Core.
Cell Culture and Adenoviral Gene Transfer-Single cardiac myocytes were isolated from the hearts of 2-3-month-old Sprague-Dawley rats using a standard enzymatic technique, then cultured, and infected with adenoviral vectors at a multiplicity of infection (m.o.i.) indicated, as described previously (27,28). Briefly, myocytes were plated at a density of 0.5 to 1ϫ 10 4 /cm 2 on coverslips or in dishes precoated with 10 g/ml laminin. The culture medium was M199 (Sigma) plus 5 mmol/ liter creatine, 2 mmol/liter L-carnitine, 5 mmol/liter taurine, 0.1% insulin-transferrin-selenium-X, 1% penicillin and streptomycin, and 25 mmol/liter HEPES, pH 7.4, at 37°C. Adenovirusmediated gene transfer was implemented by adding adenoviral vectors encoding rat WT-CaMKII ␦C , DN-CaMKII ␦C , CA-CaMK ␦C , Bcl-X L , or ␤-gal into the culture dish. The experiments were done with cells cultured 24  Cell Apoptosis-Cell apoptosis was detected by terminal deoxynucleotidyltransferase-mediated dUTP nick end labeling (TUNEL) assay as previously described (25). The percentage TUNEL-positive cells was determined by randomly counting 500 -800 cardiac myocytes over 20 randomly chosen fields in each culture dish. DNA fragmentation was also visualized by DNA laddering assay, as previously described (25).
Western Blotting-CaMKII-dependent phosphorylation of PLB at Thr 17 (PLB-Thr 17 ) was detected by Western blot using a site-specific antibody (Badrilla, West Yorkville, UK) (15,27). To quantify the expression of WT and mutant CaMKII ␦C , cell lysate (30 -50 g of protein) were loaded in a Ca 2ϩ -free loading buffer containing 20 mM EDTA and immunoblotted using anti-HA (1:1000, Covance) or anti-CaMKII antibody (Santa Cruz Biotechnology) and horseradish peroxidase-conjugated secondary antibody (Bio-Rad). Following incubation with a peroxidase-conjugated antibody, the films were exposed to the chemiluminescence (ECL; Amersham Biosciences) reaction and quantified with a video documentation system (Bio-Rad). The actin amount used as protein loading control was detected with an anti-actin antibody (Santa Cruz).
Statistical Analysis-The results are presented as the means Ϯ S.E. Statistical significance was determined by oneway analysis of variance or unpaired Student's t test when appropriate. A p value of Ͻ0.05 was considered to be statistically significant.

Manipulation of CaMKII ␦C Activity in Cultured Adult Rat
Cardiomyocytes-Previous studies have shown that CaMKII activation has been implicated in ␤ 1 AR-induced apoptosis of adult mouse cardiomyocytes (25) and Ca 2ϩ influx-mediated apoptosis of feline cardiomyocytes (31). Here, we sought to determine whether there is a causal relation between the activation of CaMKII ␦C and heart muscle cell apoptosis. We specifically enhanced or inhibited CaMKII ␦C activity by adenoviral gene transfer of a constitutively active (CA-CaMKII ␦C ) or a dominant negative CaMKII ␦C mutant (DN-CaMKII ␦C ), respectively. The expression of adenovirally delivered HA-tagged WT-, CA-, and DN-CaMKII ␦C was determined by Western blotting with either an anti-HA or an anti-CaMKII antibody in cultured adult rat cardiomyocytes 24 h after infection (Fig. 1A). The autonomous activity of CaMKII, i.e. Ca 2ϩ -calmodulin-independent kinase activity, assayed by 32 P incorporation into a substrate peptide, was elevated by 5.8-fold in cardiomyocytes infected with Adv-CA-CaMKII ␦C , whereas it was suppressed by 70% in cardiomyocytes infected with Adv-DN-CaMKII ␦C (Fig.  1B). To further evaluate the functional consequence of manipulating CaMKII activity, we examined CaMKII-mediated PLB-Thr 17 phosphorylation. Consistent with the kinase activity profile, the expression of CA-CaMKII ␦C and DN-CaMKII ␦C caused a significant increase and decrease, respectively, in the phosphorylation of PLB-Thr 17 (32). These results indicate that we can specifically enhance or suppress CaMKII ␦C activity using adenoviral gene transfer techniques in cultured intact adult rat cardiac myocytes.
Increased CaMKII ␦C Activity Is Sufficient to Cause Heart Muscle Cell Apoptosis-We next explored the potential effect of CaMKII ␦C signaling on myocyte viability. Enforced expression of CA-CaMKII ␦C alone caused increased cardiomyocyte apoptosis, documented by increased TUNEL staining-positive cells (Fig. 2, A and B) and DNA fragmentation revealed by DNA laddering assay, whereas overexpression of WT-CaMKII ␦C did not induce apoptosis in these cells (Fig. 2, A and C), consistent with the kinase activity profile under the same experimental conditions (Fig. 1B). CA-CaMKII ␦C -mediated myocyte apoptosis occurred within 24 h after adenoviral gene transfer and increased in a time-dependent manner (Fig. 2B). Inhibition of CaMKII activity with the peptide inhibitor, AIP (5 M), or coexpression of the DN-CaMKII ␦C mutant effectively suppressed CA-CaMKII ␦C -induced apoptosis (Fig. 2B). Furthermore, CA-CaMKII ␦C protein abundance and the kinase activity were closely correlated with the amount of Adv-CA-CaMKII ␦C virus delivered to the cultured cells (Fig. 3, A and B), with the severity of myocyte apoptosis correlating with the level of CaMKII activity (Fig. 3, C and D).
Involvement of Mitochondrial Death Machinery in Cardiomyocyte Apoptosis Induced by CaMKII ␦C Activation-To investigate whether the mitochondrial death pathway is involved in CaMKII ␦C -mediated heart muscle cell apoptosis, cytochrome c release into the cytosol was monitored in subcellular fractions by Western blotting. We found that cytochrome c was markedly increased in the cytosolic fraction but decreased in mitochondrial fraction in cardiomyocytes expressing CA-CaMKII ␦C , but not in those expressing DN-CaMKII ␦C or WT-CaMKII ␦C (Fig. 4, A and B), suggesting that enhanced CaMKII ␦C activity leads to cytochrome c release from mitochondria into the cytoplasm. The involvement of mitochondrial death signaling was further substantiated by the fact that overexpression of the anti-apoptotic Bcl-2 family member, Bcl-X L , suppressed CA-CaMKII ␦C -mediated cardiac myocyte apoptosis (Fig. 4, C and D).
Activation of Endogenous CaMKII by Various Stimuli That Trigger Myocyte Apoptosis-We next explored the possibility that other cell death-inducing stimuli, such as increased intracellular Ca 2ϩ concentration, acidosis, and oxidative stress, activate endogenous CaMKII to induce myocyte cell death. Indeed, CaMKII-dependent phosphorylation of PLB, the key physiological regulator of SR Ca 2ϩ -ATPase whose phosphorylation at PLB-Thr 17 serves as an intracellular marker of increased CaMKII activity, was augmented by 2-3-fold in myocytes   (Fig. 5), demonstrating that endogenous CaMKII activity was elevated in response to those treatments.
The Ca 2ϩ -elevating agents markedly promoted myocyte apoptosis as assessed by increased TUNEL staining (Fig. 6A). On average, the percentage of apoptotic cells was increased ϳ3.5-fold. Similarly, intracellular acidosis or oxidative stress (H 2 O 2 , 12 M) markedly augmented the percentage of TUNELpositive cells (Fig. 6, B and C). Inhibition of CaMKII with KN-93 (2 M) or AIP (5 M) effectively protected cells not only from the Ca 2ϩelevating agents (Fig. 6A) (Figs. 7). Thus, increased CaMKII activity is associated with the administration of a number of different cell death-inducing stimuli. Expression of DN-CaMKII ␦C is able to inhibit the kinase activity and the associated cell death as well, indicating that activation of CaMKII ␦C is a common intermediate in the signaling pathways triggered by multiple stimuli that induce myocyte apoptotic death.

DISCUSSION
In the present study, there are two novel findings. First, isoformspecific activation of CaMKII ␦C by expression of a constitutively active mutant is sufficient to trigger myocyte apoptosis. Second, inhibition of CaMKII effectively protects myocytes not only from adenoviral gene transfer of CA-CaMKII ␦ but also from that caused by  increased intracellular Ca 2ϩ , intracellular acidosis, and oxidative stress. These results support the conclusion that enhanced activation of CaMKII ␦C in cardiac myocytes is a common intermediate in the death signaling pathways initiated by diverse cell death-inducing stimuli.

Activation of Cardiac CaMKII ␦C Functions as a Common Intermediate Converging a Multitude of Cardiac Apoptotic
Signaling Pathways-As is the case for most cells, upstream apoptotic signaling of cardiac myocytes can be classified into two general pathways: the death-ligand receptor-mediated pathway involving activation of caspase-8 and downstream executioner caspases and the mitochondrial or intrinsic pathway, the terminal steps of which involve the release of cytochrome c, recruitment of apoptotic protease activating factor-1 (Apaf-1), and the activation of caspase-9 and downstream executioner caspases (33,34). In contrast to the extrinsic pathway that transduces death signaling from a specialized set of death receptors, the intrinsic pathway integrates a broader spectrum of extracellular and intracellular stresses that activate signals converging on the mitochondria, leading to the release of a number of apoptogenic proteins that either directly or indirectly activate caspases. The present study has shown that activation of the cardiac cytosolic CaMKII isoform, CaMKII ␦C , is a common intermediate that integrates various apoptotic stimuli and relays the apoptotic signals to the mitochondrial death machinery, as evi-denced by the robust increase in mitochondrial cytochrome c release and the protective effect of Bcl-X L , an important anti-apoptotic member of Bcl-2 family (35). Altogether, the present findings have identified a potentially crucial molecular intermediate involved in heart muscle cell loss and may shed new light on our understanding of the pathogenesis of heart failure and lead to a potentially important therapeutic approach for reducing myocyte loss, thus preventing or retarding the progression of heart failure caused by various etiologies.
Intracellular Distribution Dictates CaMKII Isoform-specific Functions-Recent work has demonstrated that stimulation of G protein-coupled receptors such as ET-1 leads to cardiac myocyte hypertrophy via a signaling pathway sequentially involving G q , phosphatidylinositol 1,4,5-trisphosphatemediated nuclear Ca 2ϩ release, activation of nuclear CaMKII (CaMKII ␦B ), HDAC5 phosphorylation, and activation of MEF2dependent transcription (35). Remarkably, this Ca 2ϩ -dependent, nuclear CaMKII-mediated hypertrophy signaling pathway cannot be activated by the global Ca 2ϩ transients that cause myocyte contraction (36). Thus, it is reasonable to assume that the distinct intracellular localization of cardiac CaMKII ␦B and CaMKII ␦C may enable myocytes to distinguish simultaneous local and global Ca 2ϩ signals and exhibit different functional roles. This assumption is supported by studies in CaMKII transgenic mouse models. Specifically, overexpression of the cytoplasmic CaMKII ␦C isoform leads to hyperphosphorylation of substrates involved in cardiac excitation-contraction coupling, resulting in SR Ca 2ϩ leak and cardiomyopathy (37,38), whereas overexpression of the nuclear CaMKII ␦B isoform leads to cardiac hypertrophic gene expression profile (21). These findings indicate that the intracellular distribution of the kinase dictates its accessibility and sensitivity to physiological and pathological stimuli.
Our previous studies have shown that overexpression of CaMKII ␦C but not CaMKII ␦B enhances ␤ 1 AR-mediated cardiac myocyte apoptosis (25). The present study has further demonstrated that increased activation of CaMKII ␦C is sufficient to trigger cardiac myocyte apoptosis, whereas isoform-specific inhibition of CaMKII ␦C is able to significantly protect cardiomyocytes against multiple death-inducing stimuli-mediated death. These findings provide an explanation for the observation of more severe heart failure and premature death phenotype in transgenic mice overexpressing cardiac CaMKII ␦C (37, 38) compared with those overexpressing cardiac CaMKII ␦B (21). Future investigation is merited to better appreciate the distinct functional roles of these cardiac CaMKII isoforms and their regulation under a variety of physiological or pathological circumstances.
Potential Implications of CaMKII Deregulation in the Pathogenesis of Heart Failure-Multiple lines of evidence suggest that deregulation of CaMKII acts as an important pathogenic factor for heart failure, although this kinase plays a pivotal role in normal cardiac excitation-contraction coupling and pacemaker activity (10 -16). First, previous studies have demonstrated that sustained ␤ 1 AR stimulation, a characteristic of the failing heart, causes cardiomyocyte apoptosis via activation of CaMKII signaling, independently of the classic cAMP/cAMP-dependent protein kinase pathway (25). Second, CaMKII expression is increased in the failing hearts of humans and animals (5,(22)(23)(24). Third, transgenic overexpression of the cytosolic isoform CaMKII ␦C induces severe heart failure (37), which is associated with enhanced SR Ca 2ϩ leak, reduced SR Ca 2ϩ content and enhanced fractional SR Ca 2ϩ release (37). Moreover, ␤ 1 AR-induced fetal gene expression, a hallmark of cardiac hypertrophy, is mediated by a CaMKII-dependent mechanism in cultured rat neonatal cardiac myocytes (39). In contrast, inhibition of CaMKII activity prevents cardiac arrhythmias and suppresses after depolarizations, a crucial mechanism responsible for heart failure-associated arrhythmias (17,18). Finally, CaMKII inhibition in vivo improves cardiac functional performance and reduces catecholamine-and myocardial infarc-tion-induced maladaptive cardiac remodeling (26). These studies have demonstrated that deregulation of CaMKII signaling is essentially involved in many aspects of cardiac hypertrophy and heart failure, marking CaMKII as a promising therapeutic target for the treatment of heart failure.