Preservation of base-line hemodynamic function and loss of inducible cardioprotection in adult mice lacking protein kinase C epsilon.

Signaling pathways involving protein kinase C isozymes are modulators of cardiovascular development and response to injury. Protein kinase C epsilon activation in cardiac myocytes reduces necrosis caused by coronary artery disease. However, it is unclear whether protein kinase C epsilon function is required for normal cardiac development or inducible protection against oxidative stress. Protein kinase C delta activation is also observed during cardiac preconditioning. However, its role as a promoter or inhibitor of injury is controversial. We examined hearts from protein kinase C epsilon knock-out mice under physiological conditions and during acute ischemia reperfusion. Null-mutant and wild-type mice displayed equivalent base-line morphology and hemodynamic function. Targeted disruption of the protein kinase C epsilon gene blocked cardioprotection caused by ischemic preconditioning and alpha(1)-adrenergic receptor stimulation. Protein kinase C delta activation increased in protein kinase C epsilon knock-out myocytes without altering resistance to injury. These observations support protein kinase C epsilon activation as an essential component of cardioprotective signaling. Our results favor protein kinase C delta activation as a mediator of normal growth. This study advances the understanding of cellular mechanisms responsible for preservation of myocardial integrity as potential targets for prevention and treatment of ischemic heart disease.

Coronary artery disease and chronic heart failure are important causes of death worldwide. Cellular signaling pathways that regulate cardiovascular development and responses to injury, particularly those involving protein kinase C (PKC) 1 isozymes, are under intense investigation. For example, Mochly-Rosen et al. (1) found that activation of PKC⑀ translocation in transgenic mouse hearts caused physiological hypertrophy and reduced the size of individual myocytes. In contrast, postnatal inhibition of PKC⑀ translocation produced lethal dilated cardiomyopathy and increased cardiac myocyte volumes (1). Wu et al. (2) later observed that activation of PKC⑀ translocation in G␣ q transgenic hearts improved contractile function. Conversely, inhibition of PKC⑀ translocation converted the G␣ q hypertrophic phenotype to a dilated cardiomyopathy (2). However, no previous investigations established whether PKC⑀ activation was an absolute requirement for normal cardiac growth or whether compensatory changes in the expression of other myocardial proteins developed in the absence of PKC⑀mediated signaling.
Two lines of evidence support the hypothesis that PKC⑀ activation promotes myocardial resistance to injury during periods of oxidative stress. First, both ischemic preconditioning and pharmacological approaches that induce cardioprotection increase PKC⑀ immunoreactivity and kinase function in cell particulate fractions (3). Second, transgenic expression of a constitutively active PKC⑀ (4) or a peptide agonist of PKC⑀ translocation (5) reduces cardiac myocyte necrosis during ischemia reperfusion. However, it is unclear whether PKC⑀ activation represents a final common pathway for cardioprotection versus one branch of a network of cellular mechanisms that preserve myocardial integrity. Distinction between alternative models would establish the importance of this signaling molecule as a therapeutic target for prevention of ischemic heart disease. Previous efforts to test PKC⑀ function as a requirement for cardioprotection were limited by toxicity and poor specificity of in vitro inhibitors of phosphorylation activity (6 -8).
Ischemic preconditioning and pharmacological approaches that induce cardioprotection activate both PKC␦ and PKC⑀ in rat heart and cardiac myocyte models (9 -12). The functional significance of PKC␦ activation during myocardial ischemia reperfusion is controversial. For example, Chen et al. (13) found that inhibition of PKC␦ translocation in rat hearts reduced myocyte injury measured as creatine kinase release during reperfusion. Hahn et al. (14) observed that transgenic expression of a peptide antagonist of PKC␦ translocation in mouse hearts improved contractile recovery after prolonged ischemia. Contradicting these reports are investigations of pharmacolog-ical preconditioning in which PKC␦ activation was shown to be essential for cardioprotection. For example, Kudo et al. (15) found that beneficial effects of adenosine A 1 receptor agonist pretreatment were associated with PKC␦ up-regulation and blocked by isozyme-selective antagonists such as rottlerin. Fryer et al. (12) observed that ␦ 1 -opioid receptor agonist treatment reduced infarction size during coronary artery occlusion in association with PKC␦ translocation to cardiac myocyte mitochondria, effects that were also blocked by rottlerin (12).
In this study, we examined hearts from PKC⑀ knock-out (KO) mice under physiological conditions and during ischemia reperfusion. PKC⑀ expression was clearly not a requirement for normal development because null-mutant mice could not be distinguished from their wild-type littermates by general appearance or cardiovascular function. However, acute cardioprotection induced by ischemic preconditioning or ␣ 1 -adrenergic receptor stimulation was blocked in PKC⑀ KO hearts. Targeted disruption of the PKC⑀ gene increased expression and activation of PKC␦ in cardiac myocytes. However, PKC␦ activation in the absence of PKC⑀ expression did not alter resistance to ischemia-reperfusion injury. These results support PKC⑀ activation as an essential component of cardioprotection and an important target for prevention of the pathophysiology associated with coronary artery disease. Our data favor PKC␦ function as a mediator of normal cardiovascular development, particularly in the context of compromised PKC⑀ signaling.

EXPERIMENTAL PROCEDURES
Animals-Mutant mice lacking PKC⑀ were originally derived by homologous recombination in J1 embryonic stem cells (16). F1 generation hybrid C57BL/6Jx129SvJae heterozygous progeny were then intercrossed to generate F2 generation hybrid wild-type and PKC⑀-null littermates for study. Male mice 12-16 weeks of age were used for experiments. Animal care and handling procedures were in accordance with established institutional and National Institutes of Health guidelines.
In Vivo Hemodynamic Measurements-Systolic blood pressure and heart rate were measured in conscious mice after training using a non-invasive computerized tail cuff system (VisiTech Systems, Apex, NC). Results are expressed as the means of three independent daily measurements with at least 15 of 20 successful analyses each (18). Echocardiographic studies were performed on conscious mice after training using a 15-MHz linear array transducer and commercially available imaging system (Acuson Sequoia c256, Mountain View, CA). Cardiac catheterization was performed in closed-chest mice after the administration of ketamine. A Mikrotip TM pressure-transducing catheter (Millar Instruments, Houston, TX) was inserted into the right carotid artery and advanced retrograde into the left ventricle. Heart rate, aortic pressure, and left ventricular pressures were monitored continuously using a chart recorder with signal conditioning units (Gould Electronics, Hayward, CA).
Wild-type and PKC⑀ KO hearts were pretreated with 2-min global ischemia (transient ischemic preconditioning, TIP) or 2-min phenylephrine (PHE) infusion (20 mol/liter) followed by a 5-min washout prior to index ischemia reperfusion. In separate experiments, wild-type and PKC⑀ KO hearts were preconditioned with four cycles of 4-min global ischemia and 6-min reperfusion (4-CYCLE IP). Hearts isolated from C57BL/6J mice were pretreated with PKC⑀ agonist peptide (YGRKKRRQRRR-HDAPIGYD), antagonist (YGRKKRRQRRR-EAVS-LKPT), or scrambled antagonist (YGRKKRRRQRRR-LSETKPAV) infusion (1 mol/liter) with no washout period before ischemia reperfusion. Peptides were synthesized at the University of California San Francisco Biomolecular Resource Center using N-(9-fluorenyl)methoxycarbonyl (Fmoc) chemistry. All of the peptides were purified (Ͼ95%) by preparative reversed-phase high performance liquid chromatography (19). Creatine kinase (CK) activity in coronary effluent collected during reperfusion was measured using a commercial kit (Sigma) and corrected for flow rate and wet heart weight. After reperfusion, hearts were perfused with 1% triphenyltetrazolium chloride solution, fixed in 10% neutral buffered formalin, and sectioned (20). Planimetry of viable (stained) and necrotic (unstained) tissue was performed using NIH Image software. Infarction size was corrected for the weight of each heart section. Protein Kinase C Assays-Western blot analysis of PKC isozyme expression was performed as described previously (19). Left ventricular tissue was homogenized, and samples of the 100,000 ϫ g supernatant and Triton X-100-extracted pellet fractions were subjected to SDS-PAGE and then transferred to nitrocellulose membrane. PKC distribution was determined using isozyme-selective primary antibodies (Transduction Laboratories, Lexington, KY) and enhanced chemiluminescence detection (Amersham Biosciences). In separate experiments, intact mitochondria were isolated from individual mouse hearts using differential centrifugation protocols (21) and subjected to Western analysis. Immunofluorescence microscopy was performed as described previously (5,11). Myocytes were fixed with 4% paraformaldehyde and permeabilized with 0.2% Triton X-100. Cells were incubated with PKC isozyme-selective primary antibodies (Santa Cruz Biotechnology, Santa Cruz, CA) and fluorescein isothiocyanate-conjugated secondary antibodies. In selected experiments, cells were stained with MitoTracker Red CMXRos (Molecular Probes, Eugene, OR) prior to fixation. Fluorescent images of labeled myocytes were acquired using a Leica TCS confocal laser-scanning microscope system (Leica Lasertechik, Heidelberg, Germany).
Statistical Analysis-Results are reported as the means Ϯ S.E. Comparisons between groups were made using one-way ANOVA or repeated measures ANOVA as indicated. Differences were confirmed using a Bonferroni post hoc test. p Ͻ 0.05 was considered significant.

Normal Base-line Morphology and in Vivo Cardiac Function in PKC⑀ KO Mice-Because PKC⑀ activation is important for
physiological myocardial development in other mouse models (1, 2), we were surprised to find that hearts from PKC⑀ KO mice were normal in weight and appearance (Table I). Furthermore, cellular volumes of myocytes isolated from PKC⑀ KO hearts were no different from wild-type cardiac myocytes (22,275 Ϯ 958 versus 24,565 Ϯ 918 m 3 ; n ϭ 4, p ϭ not significant) as measured electronically with a Coulter counter (22). Therefore, targeted disruption of the PKC⑀ gene had no obvious effect on overall heart morphology or individual cardiac myocyte size.
PKC activation also regulates myocardial contractility through direct phosphorylation of myofilament proteins (23). However, we found no difference in base-line hemodynamic function between wild-type and PKC⑀ KO mice during in vivo catheterization (Table I). Non-invasive techniques performed in conscious mice validated results obtained using invasive approaches. First, there were no differences in left ventricular mass or contractile function between groups measured using two-dimensional and M-mode echocardiography (Table I). Second, tail cuff measurements of systolic blood pressure and heart rate were normal in PKC⑀ KO mice ( Table I).
Loss of Cardioprotection in PKC⑀ KO Mice-PKC⑀ activation in cardiac myocytes induces resistance against myocardial ischemia-reperfusion injury (4,5). However, it is unclear whether PKC⑀ activation is required for the beneficial effects of acute ischemic or pharmacological preconditioning. Here we investigated cardioprotective signaling using buffer-perfused hearts from wild-type and PKC⑀ KO mice to eliminate potential confounding effects of PKC⑀ gene disruption on pulmonary, vascular (24), and central nervous system (25) functions. We observed no differences in LVDP between wild-type and PKC⑀ KO hearts (84 Ϯ 3 versus 86 Ϯ 3 mm Hg; n ϭ 18, p ϭ not significant), confirming measurements made during in vivo catheterization.
Our study next focused on whether improvement of LVDP recovery was associated with reduction of myocardial necrosis (29,30). We observed that 20-min ischemia and 30-min reperfusion produced extensive infarction in both wild-type and PKC⑀ KO hearts ( Fig. 2A). TIP reduced infarction size (Fig. 2B controls; n ϭ 6, p Ͻ 0.05). However, effects of TIP on infarction size were blocked in PKC⑀ KO hearts (Fig. 2B). Similarly, PHE reduced infarction in wild-type hearts to 29 Ϯ 3% LV mass (n ϭ 6, p Ͻ 0.05 versus WT controls) but did not induce protection in PKC⑀ KO hearts (Fig. 2C). Creatine kinase activity in coronary effluent validated measurements of infarction size (Fig. 2).
PKC⑀ Activation Is Required for Acute Cardioprotection in C57BL/6 Mice-Although the PKC⑀ KO mouse model is a powerful tool for investigation of cardioprotective signaling, we sought alternative approaches for inhibition of PKC⑀ function. Chen et al. (13) previously used modulators of PKC isozyme translocation in ex vivo rat hearts to examine the effects of PKC␦ and PKC⑀ activation on creatine kinase release during ischemia reperfusion. Gustafsson et al. (31) recently confirmed that trans-activating protein-mediated protein transduction delivered cardioprotective peptides into buffer-perfused rat hearts, including cytosolic and mitochondrial fractions.
In this study, we added a peptide agonist of PKC⑀ translocation (5, 32) linked to amino acids derived from the protein transduction domain of human immunodeficiency virus transactivating protein protein (33) to perfusate for 20 min before global ischemia. PKC⑀ agonist pretreatment did not affect the subcellular distributions of PKC␣ and PKC␦ in left ventricular lysates (Fig. 3A). However, agonist infusion increased the proportion of PKC⑀ localized to 100,000 ϫ g particulate fractions (0.63 Ϯ 0.02 versus 0.46 Ϯ 0.02 for controls; n ϭ 3, p Ͻ 0.01), indicating activation (3,34). In independent experiments, we found that TIP did not alter the subcellular distribution of PKC␣ but increased the proportion of PKC␦ (0.54 Ϯ 0.02 versus 0.40 Ϯ 0.02 for controls; n ϭ 3, p Ͻ 0.01) and PKC⑀ (0.57 Ϯ 0.06 versus 0.37 Ϯ 0.03 for controls; n ϭ 3, p Ͻ 0.05) in particulate fractions (Fig. 3B). PKC⑀ antagonist had no effect on TIPinduced translocation of PKC␦ (Fig. 3C). However, the antag-onist blocked translocation of PKC⑀ to particulate fractions (64 Ϯ 2 versus 117 Ϯ 12 density units for scrambled antagonist; n ϭ 3, p Ͻ 0.05), indicating degradation of activated isozyme not bound by its anchoring proteins or receptors for activated C-kinase (3). Scrambled PKC⑀ antagonist had no effect on the subcellular distributions of the PKC isozymes examined. Therefore, trans-activating protein-mediated protein transduction delivered selective activators and inhibitors of PKC⑀ function into mouse hearts.
Ping et al. (4) developed transgenic mice in which cardiac expression of constitutively active PKC⑀ induced resistance to ischemia-reperfusion injury. Dorn et al. (5) established that PKC⑀ activation caused by transgenic expression of PKC⑀ agonist peptide in mouse hearts also caused sustained cardioprotection. Here we found that pretreatment with PKC⑀ agonist improved LVDP recovery (Fig. 4A) in normal hearts from C57BL/6 mice (66 Ϯ 4 versus 28 Ϯ 4 mm Hg for controls; n ϭ 6, p Ͻ 0.05). PKC⑀ agonist pretreatment also reduced infarction and creatine kinase release caused by ischemia reperfusion (Fig. 4A). Thus, direct activation of PKC⑀ translocation through acute modulation of interactions with anchoring proteins mimicked the cardioprotective effects of ischemic and ␣ 1 -adrenergic receptor-mediated preconditioning.
We previously used inhibitors of PKC⑀ translocation in a neonatal rat cardiac myocyte model of hypoxic preconditioning to show that PKC⑀ activation was necessary for acute protection against hypoxia-induced cell death (11). Ping et al. (35) later observed that adenovirus-mediated expression of dominant-negative mutant PKC⑀ in isolated rabbit cardiac myo-

FIG. 2. Loss of inducible protection against myocardial infarction in PKC⑀ KO hearts. Representative sections showing viable (red-stained) and necrotic (unstained) LV tissue after ischemia reperfusion.
A, infarction size and creatine kinase (CK) release at 30-min reperfusion were substantial in hearts from WT and PKC⑀ KO mice that were not preconditioned. B, TIP reduced infarction and CK release in WT but not PKC⑀ KO hearts. C, PHE reduced infarction and CK release only in WT hearts; n ϭ 6/group; *, p Ͻ 0.05 versus WT control (CON). cytes also blocked protection during simulated ischemia. In this study, pretreatment of intact heart with PKC⑀ antagonist had no effect on base-line contractile function. However, inhibition of PKC⑀ translocation blocked improvement of LVDP recovery after preconditioning (Fig. 4, B and C). Selective PKC⑀ inhibition also prevented TIP-and PHE-mediated reduction of infarction size and creatine kinase release. Scrambled PKC⑀ antagonist had no effect on cardiac contractility, infarction size, or creatine kinase release. Thus, PKC⑀ activation was necessary for the acute protection of intact myocardium against ischemia-reperfusion injury in C57BL/6 mice.
Contractile Recovery and Infarction after Pretreatment with Four Cycles of Transient Ischemia-In an earlier investigation, Saurin et al. (36) observed that four cycles of 4-min ischemia and 6-min reperfusion prior to prolonged ischemia reperfusion also reduced infarction size in ex vivo hearts from mice expressing PKC⑀ but not in PKC⑀ KO hearts. In contrast to our current study, those investigators demonstrated that four-cycle ischemic preconditioning improved contractile recovery during reperfusion in both control and PKC⑀ KO hearts. Differences in results from the two laboratories raised the possibility that PKC⑀-independent myocardial signaling sufficient to reduce oxidant stress and stunning during reperfusion may develop after repeated exposure to brief ischemia (36). We tested this hypothesis by subjecting wild-type and PKC⑀ KO hearts to four cycles of 4-min ischemia and 6-min reperfusion prior to prolonged ischemia reperfusion. 4-CYCLE IP improved LVDP recovery at 30-min reperfusion in wild-type hearts (59 Ϯ 8 versus 20 Ϯ 4 mm Hg for controls; n ϭ 6, p Ͻ 0.05) but not in PKC⑀ KO hearts (Fig. 5A). Similarly, 4-CYCLE IP prevented pathological elevation of LVEDP after ischemia reperfusion only in wild-type hearts (Fig. 5A). In agreement with the results of Saurin et al. (36), we found that 4-CYCLE IP reduced infarction after prolonged ischemia reperfusion in wild-type hearts (28 Ϯ 3 versus 47 Ϯ 3% LV mass for controls; n ϭ 6, p Ͻ 0.05) but not in PKC⑀ KO hearts (Fig. 5B). Creatine kinase activity in coronary effluent validated measurements of infarction size. Therefore, improved contractile recovery caused by four-cycle preconditioning was associated with reduced tissue necrosis in wild-type hearts but was blocked in PKC⑀ KO hearts.
Compensatory Increases in PKC␦ Expression and Activation in PKC⑀ KO Hearts-Hypoxic preconditioning activates both PKC␦ and PKC⑀ in neonatal rat ventricular myocyte models (11). Ischemic preconditioning is also known to activate PKC␦ and PKC⑀ in ex vivo rat hearts (9,10). However, PKC␦-mediated signaling during ischemia reperfusion has not been fully explored. Here we confirmed that targeted disruption of the PKC⑀ gene blocked the expression of PKC⑀ protein (Fig. 6A). We also found that PKC␦ protein expression increased substantially in hearts from mice lacking PKC⑀ (166 Ϯ 14 versus 93 Ϯ 8 density units for wild type; n ϭ 4, p Ͻ 0.01).
Because Western analysis provides crude localization of signaling molecules, we adapted an adult mouse cardiac myocyte model suitable for confocal microscopy techniques (17). We next observed that PKC␦ localized to nuclei of all wild-type cardiac myocytes (Fig. 6B). In contrast, PKC␦ localized to perinuclear sites in all of the myocytes prepared from PKC⑀ KO hearts (Fig.  6B). Importantly, PKC␦ translocated acutely to identical perinuclear sites in wild-type cells after treatment with the PKC activator 4␤-phorbol 12-myristate 13-acetate (Fig. 6B). These results suggested that PKC␦ was chronically activated in cardiac myocytes from PKC⑀ KO hearts.
Mitochondrial localization of activated PKC␦ may be required for physiological function. For example, Caruso et al. (37) established that mitochondrial translocation of PKC␦ was necessary for insulin stimulation of pyruvate dehydrogenase complex activity in muscle and liver cells. Fryer et al. (12) found mitochondrial translocation of PKC␦ important for the cardioprotective effects of ␦ 1 -opioid receptor stimulation in a rat model of acute myocardial infarction. In this study, incubation of wild-type myocytes with fluorescent probe MitoTracker Red labeled mitochondria in longitudinal arrays between myofibrils (Fig. 6C). In contrast, staining with selective PKC␦ antibodies revealed predominant localization to transverse tubules (38).
Mitochondrion-selective labeling of PKC⑀ KO cardiac myocytes also revealed longitudinal arrays between myofibrils (Fig.  6C). Importantly, targeted disruption of the PKC⑀ gene did not increase localization of PKC␦ to mitochondria, demonstrated by confocal images simultaneously acquired from myocytes dual-labeled with PKC␦ antibodies and MitoTracker Red (Fig.  6C). In separate experiments, we subjected mitochondrial fractions from wild-type and PKC⑀ KO hearts to Western analysis and found no differences in PKC␦ content between groups (Fig.  6D). Thus, PKC␦ expression and perinuclear localization increased in PKC⑀ KO hearts without changes in distribution to mitochondria or base line resistance to ischemia-reperfusion injury.

DISCUSSION
Our echocardiographic analyses and miniaturized catheterization studies yielded the first published in vivo measurements of hemodynamic function in PKC⑀ KO mice and demonstrated normal cardiac morphology and contractility. In an early investigation of ex vivo mouse heart function, Saurin et al. (36) also observed that ischemic preconditioning reduced infarction size during ischemia reperfusion in hearts expressing PKC⑀ but not in PKC⑀ KO hearts (36). In contrast to our study, those investigators demonstrated that ischemic preconditioning improved contractile recovery during reperfusion in both control and PKC⑀ KO hearts. At least three differences in experimental conditions used by each laboratory may provide important insights regarding PKC function in cardioprotective signaling. First, although similar strategies were used to target the PKC⑀ gene, knock-out mice were independently derived in different genetic backgrounds (16,39). Second, Saurin et al. (36) paced hearts at 10 Hz to mimic rates observed in intact mice and generated LV developed pressures below physiological levels (36). We paced all of the hearts at 6 Hz to remain within the oxygen and substrate transfer limits of the buffer perfusion system and generated LV developed pressures equal to those measured during in vivo cardiac catheterization. Finally, the hearts in the previous study were preconditioned with four cycles of transient ischemia, whereas our work focused on cardioprotection conferred by a single cycle of transient ischemia.
We tested the hypothesis that PKC⑀-independent cardioprotective signaling develops after repeated exposure to brief ischemia by subjecting hearts from wild-type and PKC⑀ KO mice to four-cycle ischemic preconditioning. We found that 4-CYCLE IP reduced subsequent infarction in control hearts but not in PKC⑀ KO hearts. In contrast to the results of Saurin et al. (36), we found that 4-CYCLE IP improved contractile recovery in control hearts but not in PKC⑀ KO hearts. Thus, PKC⑀ activation was required for reduction of tissue necrosis and improvement of function caused by four-cycle ischemic preconditioning. There was no dichotomy of infarction size and contractile recovery (36) in any combination of animals and preconditioning strategies studied. Our investigation was not designed to identify cellular mediators of cardioprotection downstream of PKC⑀. However, the clear distinction of preconditioning effects on contractility between PKC⑀ KO mice derived in different backgrounds suggests that these animals will be useful for future exploration of myocardial stunning using genomic and proteomic approaches. Alternatively, more exhaustive examination of pacing rate, oxygenation, and substrate utilization effects on PKC⑀ KO heart function may yield novel insights regarding metabolic pathways that are responsible for preservation of myocardial integrity during periods of oxidative stress.
Results from experiments using null-mutant animals supported the hypothesis that PKC⑀ activation is required for reduction of infarction size during prolonged ischemia reperfusion. However, we recognized that the loss of inducible cardioprotection in PKC⑀ KO mice might be a consequence of polymorphism in the genetic background or potential confounding effects of linked genes (40,41). We were aware that few other approaches selectively block PKC⑀ function under physiological conditions. For example, pharmacological agents such as chelerythrine chloride exert biological effects besides PKC inhibition (7,8). Similarly, transgenic expression of the first variable (V1) region of PKC⑀ in adult mouse hearts prevents PKC⑀ translocation but progressively impairs contractility (1). Accordingly, we used protein transduction methods to introduce peptide modulators of PKC⑀ translocation acutely into hearts approaches of gene targeting and pharmacological modulation of interactions between PKC⑀ and its anchoring proteins independently established a requirement for this signaling pathway in acute cardioprotection.
In vivo myocardial ischemia causes release of norepinephrine from sympathetic efferent nerves and up-regulation of ␣ 1 -adrenergic receptors (42). Stimulation of ␣ 1 -adrenergic receptors activates numerous signaling pathways within ventricular myocytes including mitochondrial K ATP channels (28), glucose transporters (42), and PKC isozymes (9). However, it is unclear which of the pathways protect acutely against injury and which of them contribute chronically to myocardial pathology. In this study, phenylephrine pretreatment caused activation and translocation of PKC⑀, but not PKC␦, to myocyte particulate fractions (data not shown). Beneficial effects of ␣ 1 -adrenergic receptor stimulation were blocked in PKC⑀ KO hearts and in C57BL/6J hearts after perfusion with PKC⑀ antagonist. This is the first demonstration of a requirement for PKC⑀ signaling in cardioprotection mediated by ␣ 1 -adrenergic receptors. Our findings suggest that PKC⑀ KO mice will be useful for future investigation of exercise-induced protection, a strategy that reduces ischemia-reperfusion injury in association with norepinephrine release (43).
Changes in expression of one PKC isozyme have been shown to alter expression of other PKC isozymes within the same tissue. For example, Ways et al. (44) observed that transfection of human MCF-7 breast cancer cells with PKC␣ enhanced the neoplastic phenotype, increased endogenous expression of PKC␤, and reduced expression of PKC␦ and PKC. Targeted disruption of the PKC⑀ gene did not alter expression of other PKC isozymes in mouse dorsal root ganglia or central nervous system tissues (16). However, Western analysis of PKC⑀ nullmutant hearts in this study revealed up-regulation of PKC␦ protein expression. Using confocal microscopy techniques not employed previously, we found consistent localization of PKC␦ to perinuclear regions in PKC⑀ KO myocytes, sites identical to those in wild-type myocytes after activation by phorbol ester and in neonatal rat cardiac myocytes after hypoxic preconditioning or norepinephrine stimulation (1). These findings supported chronic activation of PKC␦ in hearts from mice lacking PKC⑀. Definitive evaluation of compensatory changes in cardiac signaling triggered by PKC⑀ gene ablation awaits microarray analysis of null-mutant hearts. Our data favor PKC␦ as one mediator of normal myocardial development, particularly in the context of compromised PKC⑀ function. A similar role for PKC␦ in maintenance of cytoskeletal integrity was proposed by Hahn et al. (14) based on a mouse model of myofibrillar cardiomyopathy.
Given the importance of PKC⑀ activation in inducible cardioprotection, we were surprised to find that wild-type and PKC⑀ KO hearts could not be distinguished by base-line susceptibility to ischemia-reperfusion injury. At least two models of cardiac PKC␦ function in the absence of PKC⑀ protein expression may explain these results. In the first model, PKC␦ signaling serves as one mediator of cardiotoxicity and isozyme-selective inhibition would improve contractile recovery and reduce myocardial necrosis during reperfusion. In the second model, PKC␦ signaling contributes to cardioprotection and isozyme-selective inhibition would further impair contractile recovery and increase infarction size. Although Western analysis and immunofluorescence microscopy data indicated chronic PKC␦ activation in PKC⑀ KO myocytes, full function of up-regulated isozyme may require translocation to cellular sites and substrates not available under control conditions. For example, Zhao et al. (6) demonstrated that transfection of neonatal rat cardiac myocytes with wild-type PKC␦ did not alter cell viabil-ity during simulated ischemia (6). These alternatives will be tested using adenosine A 1 receptor agonists, ␦ 1 -opioid receptor agonists, and direct modulators of PKC␦ translocation in experiments beyond the scope of the present investigation.
In summary, we developed complementary approaches for inhibition of PKC⑀ function in intact hearts. We demonstrated for the first time that PKC⑀ expression was not required for normal in vivo morphology and contractility under physiological conditions. We established that PKC⑀ activation was necessary for acute cardioprotection induced by ischemic preconditioning and by ␣ 1 -adrenergic receptor stimulation. Finally, we showed that increased PKC␦ expression in hearts lacking PKC⑀ might have sustained normal cardiac growth but did not alter base-line resistance to oxidative stress. This study advances the understanding of the molecular mechanisms responsible for myocardial integrity as potential targets for prevention of ischemic heart disease.