Ischemic Protection and Myofibrillar Cardiomyopathy
Dose-Dependent Effects of In Vivo δPKC Inhibition
To delineate the in vivo cardiac functions requiring normal δ protein kinase C (PKC) activity, we pursued loss-of-function through transgenic expression of a δPKC-specific translocation inhibitor protein fragment, δV1, in mouse hearts. Initial results using the mouse α-myosin heavy chain (αMHC) promoter resulted in a lethal heart failure phenotype. Viable δV1 mice were therefore obtained using novel attenuated mutant αMHC promoters lacking one or the other thyroid response element (TRE-1 and -2). In transgenic mouse hearts, δV1 decorated cytoskeletal elements and inhibited ischemia-induced δPKC translocation. At high levels, δV1 expression was uniformly lethal, with depressed cardiac contractile function, increased expression of fetal cardiac genes, and formation of intracardiomyocyte protein aggregates. Ultrastructural and immunoconfocal analyses of these aggregates revealed focal cytoskeletal disruptions and localized concentrations of desmin and αB-crystallin. In individual cardiomyocytes, cytoskeletal abnormalities correlated with impaired contractile function. Whereas desmin and αB-crystallin protein were increased ≈4-fold in δV1 hearts, combined overexpression of these proteins at these levels was not sufficient to cause any detectable cardiac pathology. At low levels, δV1 expression conferred striking resistance to postischemic dysfunction, with no measurable effects on basal cardiac structure, function, or gene expression. Intermediate expression of δV1 conferred modest basal contractile depression with less ischemic protection, associated with abnormal cardiac gene expression, and a histological picture of infrequent cardiomyocyte cytoskeletal deformities. These results validate an approach of δPKC inhibition to protect against myocardial ischemia, but indicate that there is a threshold level of δPKC activation that is necessary to maintain normal cardiomyocyte cytoskeletal integrity.
The δ and ε isoforms of protein kinase C (PKC) are implicated in heart failure, myocardial hypertrophy, and ischemic preconditioning, based largely on patterns of PKC isoform translocation in human heart disease and experimental animal models.1–5⇓⇓⇓⇓ Accordingly, delineating the in vivo consequences of PKC isoform activation in myocardial tissue is a scientific priority. Studying these signaling molecules by modulating their expression can potentially be confounded by promiscuous interactions between overexpressed signaling proteins and their downstream effectors or substrates, or by opportunistic compensation of related signaling proteins in gene ablation models. These concerns are especially relevant to PKC, a group of ubiquitous enzymes composed of at least 11 different, but related, isoforms, many of which are coexpressed in the same tissue or cell type.6–8⇓⇓ To avoid alterations in enzyme-substrate stoichiometry that are inherent with PKC isoform overexpression or mutational ablation, we have utilized complementary gain- and loss-of function, achieved by modulating PKC isoform interactions with their respective membrane anchor proteins, or RACKs.9–12⇓⇓⇓ In this manner, in vivo δ and εPKC activation revealed opposing effects on myocardial ischemic protection, but identical cardiotrophic influences,12 which suggested a novel therapeutic approach for myocardial ischemic salvage by εPKC activation and δPKC inhibition, alone or in combination. However, growth-modifying effects of chronically activated δ and εPKC suggest caution in manipulating their myocardial activity. Indeed, chronic in vivo inhibition of myocardial εPKC has caused a dilated cardiomyopathy,10,11⇓ and initial attempts to generate a viable mouse model of chronic δPKC inhibition were unsuccessful due to early development of a lethal cardiomyopathy.
In the present study, we report the results of chronic myocardial δPKC inhibition through transgenic expression of δV1, which encodes the putative δPKC RACK binding domain.12 Viable δV1-expressing mice were obtained using mutant α-myosin heavy chain (αMHC) promoters lacking one or the other thyroid response element (TRE). Mice expressing δV1 at the lowest levels exhibited marked resistance to myocardial ischemia but were otherwise normal in terms of myocardial structure, function, and gene expression. In contrast, when expressed at higher levels, δV1 transgenic mice developed, in a transgene dose-dependent fashion, a cardiomyopathy characterized histologically and ultrastructurally by cytoskeletal disorganization and intracellular proteinaceous aggregates resembling those in human myofibrillar cardiomyopathies.13,14⇓ In the context of δV1 localization to cytoskeletal elements within cardiac myocytes, these results indicate a role for δPKC not only in the myocardial response to ischemia, but also in maintaining normal myofibrillar integrity.
Materials and Methods
Generation of Transgenic δV1 Mice
Analogous to the previously described εPKC selective inhibitory protein fragment, εV1,9,15⇓ δV1 was modeled after the first variable region of rat δPKC (amino acids 2 to 144), which encodes the putative δPKC RACK binding domain.12 For transgenic expression, the cDNA for the δV1 peptide, preceded by a FLAG epitope, was inserted into the SalI-HindIII site of the full-length αMHC promoter16 followed by the human growth hormone polyadenylation sequence (αMHC-δV1). For attenuated δV1 expression, a Bgl-II cassette of the αMHC promoter, with either the first or second thyroid responsive element individually ablated,17 was inserted into the analogous portion of the αMHC-δV1 construct, generating TRE-1 αMHC-δV1 and TRE-2 αMHC-δV1 (Figure 1). All transgene constructs were confirmed by double-stranded sequencing, microinjected into male pronuclei of fertilized mouse oocytes (FVB/N), and implanted into pseudopregnant dames by the University of Cincinnati Transgenic Core (directed by John Neumann). Founders were identified by genomic Southern analysis of tail clip DNA. Studies were performed in accordance with protocols approved by the University of Cincinnati Institutional Animal Care and Use Committee, and animals were supplied by the UC Transgenic Mouse Core facility.
SDS-PAGE and Immunoblot Analysis
Western blotting was performed as described.9 Briefly, myocardial homogenates were fractionated by differential centrifugation into cytosolic (100 000g supernatant) and Triton X-100–extracted particulate (100 000g pellet) fractions, size-separated on 10% SDS-PAGE gels, transferred to polyvinylidene membranes, blocked with 5% dry milk, and incubated with primary antibody and secondary antibodies. Blots were developed with enhanced chemifluorescence and quantitated using a Storm PhosphorImager. To accurately quantitate cytosolic PKCδ, it was necessary to load 75 μg of protein, compared with 25 μg in the membrane fraction. Antibodies were as follows: anti–αB-crystallin (Calbiochem), anti–FLAG-M2 and anti-desmin (Sigma), anti–δ and εPKC, and anti–heat shock protein 60 (Santa Cruz).
mRNA levels were quantified by dot blotting as described.18,19⇓ Total ventricular RNA was applied (2 μg/dot) to nylon membranes, hybridized with 32P-labeled antisense oligonucleotides specific for atrial natriuretic factor (ANF), αMHC, βMHC, sarcoplasmic reticular ATPase (SERCA), α skeletal actin, or glyceraldehyde-3-phosphate dehydrogenase (GAPDH), and quantified using a Storm PhosphorImager (Molecular Dynamics).
Ischemic Injury and Recovery
Isolated hearts from 12- to 20-week-old mice were Langendorff-perfused for 20 minutes for equilibration, followed by 40 minutes of no-flow ischemia and a 20-minute reperfusion period.12 Contractile function was measured as left ventricular developed pressure (LVP) and the peak rate of increase of pressure (+dP/dt). Creatine kinase (CK), a marker of myocyte injury, was assayed in the perfusate using a Sigma kit.
Polyethylene catheters were introduced into the femoral artery and vein and a 1.4F Millar catheter placed into the left ventricle through the right carotid artery to monitor real-time heart rate, arterial and left ventricular pressures, and +dP/dt and −dP/dt. Data were archived on a G4 computer (Apple Computers) using MacLab software and interface.19
Isolated adult ventricular myocytes were prepared20 and analyzed according to genotype and presence of intracellular aggregates determined by phase contrast microscopy. Myocyte +dL/dt (peak rate of unloaded shortening), −dL/dt (peak rate of relengthening), and percent shortening were recorded during field stimulation at 0.25 Hz.
Routine histological examination was performed on formalin-fixed Masson’s trichrome-stained sections. Samples were preserved in glutaraldehyde for ultrastructural analysis. Immunoconfocal studies used desmin, αB-crystallin, vimentin, and δPKC antibodies (the latter a gift from Dr Daria Mochly-Rosen, Stanford University, Stanford, Calif), all at a 1:200 dilution, with Texas red- or fluorescein-conjugated secondary antibodies (1:200). Imaging was performed with a dual laser Nikon PCM2000 system and a Nikon Eclipse E800 microscope using emission wavelengths of 515±30 nm (fluorescein) and 605±32 nm (Texas Red).
Results are presented as mean±SEM. Experimental groups were compared using Student’s t test or 1-way ANOVA, as appropriate. The Bonferroni test was applied to all significant ANOVA results using SigmaStat software. A value of P<0.05 was considered significant.
Premature Lethality and Myofibrillar Cardiomyopathy of αMHC-δV1 Transgenic Mice
Our prior studies of transgenic mice expressing a δPKC activating peptide revealed exaggerated postischemic injury.12 We therefore considered whether inhibition of δPKC could be cardioprotective. To test this possibility, transgenic mice were created expressing the δPKC inhibitory protein fragment, δV1, in the heart. However, the response of these δPKC-inhibited hearts to ischemia could not be assessed due to early lethality. Of three original αMHC-δV1 founders, two died before breeding, whereas the third sired only 9 first generation δV1 transgenic mice (out of 245 pups; 3.7%) before he also died. No αMHC-δV1 first generation mice survived long enough to breed; all were dead by the age of 41 days (mean 27±3 days). Before death, αMHC-δV1 transgenic mice appeared lethargic, with rapid respirations and tissue edema, suggesting heart failure. Indeed, although only one αMHC-δV1 mouse survived to an age and size sufficient for invasive hemodynamic evaluation, this mouse exhibited a 64% reduction in peak dobutamine-stimulated +dP/dt compared with its control (NTG=19 877 versus αMHC-δV1=6720 mm Hg/s), and pathological examination of all 9 αMHC-δV1 mice revealed pleural effusions, ascites, and enlarged, thick-walled hearts with intramural thrombi (Figure 2A). Compared with NTG siblings, whole-heart weights of αMHC-δV1 mice were increased by 89% (93±38 mg NTG versus 176±13 mg αMHC-δV1; P<0.001), and lung weights were increased, indicating pulmonary congestion (158±20 mg NTG versus 215±21 mg αMHC-δV1; P=0.024). The thick-walled δV1 phenotype contrasts with dilated cardiomyopathy in the analogous εPKC inhibitor (εV1) transgenic mice (Figure 2A) (which is, however, equally lethal10,11⇓). Immunoconfocal microscopy of δV1 using the incorporated FLAG-epitope (Figure 2G) revealed decoration of cytoskeletal elements, similar to anti-δPKC (Figure 2F).21 This pattern of subcellular localization contrasts with localization of εPKC-derived εV1 to cross-striated elements.10
A unique feature of the δV1 phenotype, and one that could explain the development of lethal heart failure without ventricular dilation, was frequent intramyocyte blue-staining areas on Masson’s trichrome-stained myocardial sections that did not specifically label for mucin, collagen, glycogen, or mucopolysaccharides (Figure 2B). On ultrastructural examination these areas corresponded to electron-dense regions of amorphous protein within focal subcellular cytoskeletal distortions (Figure 2C). Mitochondria, other cellular organelles, and streaming fragments of sarcomeres were visualized within these areas (see online Figure in the online data supplement available at http://www.circresaha.org).
The distortions of cardiomyocyte subcellular architecture in αMHC-δV1 hearts closely resemble those seen in human myofibrillar myopathies13,14,22⇓⇓ and their genetically modified mouse cardiomyopathy analogs.23,24⇓ In these myopathies, blue-staining intramyocyte protein aggregates contain high levels of desmin and αB-crystallin, due to mutations affecting one or the other protein.13,14⇓ We therefore examined δV1-expressing cardiomyocytes for similar immunohistological features. As shown in Figures 2D and 2E, the intracellular aggregates in δV1 myocardium stained positively for both desmin and αB-crystallin, the latter of which, as in human myofibrillar myopathies,14,22⇓ was preferentially localized to the periphery (Figure 2E). In contrast, δV1 staining was not more prominent within the aggregates (Figure 2G), and there was no qualitative change in desmin or αB-crystallin in normal-appearing δV1 cardiomyocytes (not shown). Not surprisingly, given the abnormal accumulation of these proteins into intramyocyte aggregates, immunoblot analysis showed that expression of both desmin and αB-crystallin was increased approximately 4-fold in αMHC-δV1 myocardium, compared with nontransgenic (Figure 3A). In contrast, expression of a related chaperone, HSP60, was not altered (Figure 3A). A direct phosphorylation effect of δPKC on desmin and αB-crystallin was excluded by 2-dimensional immunoblotting, which showed no differences in phosphorylation states between hearts from αMHC-δV1, nontransgenics, and the recently described δPKC activator mice12 (not shown).
Given the previously established roles of mutant desmin and αB-crystallin in human myofibrillar myopathies, their upregulation and localization within protein aggregates in cardiomyopathic δV1 hearts raised the possibility that increased expression of these two proteins in combination might contribute to the cardiomyopathy. To test this notion, we crossbred recently developed mouse models of desmin and αB-crystallin overexpression, neither of which are reported to exhibit myocardial pathology.23,24⇓ Immunoblot analysis demonstrated that the double transgenics maintained the ≈4-fold levels of desmin and αB-crystallin overexpression of the parent lines (Figure 3B). Confirming the previous characterization of these mice, single transgenic animals were normal in terms of myocardial pathology and cardiac contractile function (Figure 3C and data not shown). The desmin/αB-crystallin double transgenic mice also were normal in all respects, including heart weight, ventricular contractility at baseline and in response to dobutamine (Figure 3C and data not shown), and myocardial histological appearance (Figure 3D). From this, we conclude that simple upregulation of myocardial desmin and αB-crystallin is not sufficient to cause the δV1 myofibrillar cardiomyopathy phenotype, and infer that δPKC activation and targeting to the cardiomyocyte cytoskeleton is necessary for normal cardiomyocyte structure and contractile function.
To confirm that the observed cardiomyocyte cytoskeletal abnormalities impaired αMHC-δV1 myocardial performance, contractile function of isolated ventricular myocytes was assessed as a function of the presence or absence of cytoskeletal distortions. Using phase contrast optics, cardiomyocytes with normal cellular architecture were easily distinguished from those with intracellular aggregates. Accordingly, compared with normal-appearing αMHC-δV1 cells, the unloaded fractional shortening of abnormal αMHC-δV1 cardiomyocytes was depressed by 63%, and the peak rate of unloaded shortening was decreased by 59% (Table 1). Normal-appearing αMHC-δV1 myocytes had unloaded fractional shortening similar to NTG (Table 1).
Phenotypic δV1 Dose-Response With Attenuated αMHC Promoters and Cardiac Protection With δPKC Inhibition
The unexpected phenotype of myofibrillar cardiomyopathy and lethality in the αMHC-δV1 mice precluded an analysis of the effects of δPKC inhibition on the cardiac response to ischemia/reperfusion injury. To obtain viable animals suitable for testing our original hypothesis, two approaches were undertaken. First, the original αMHC-δV1 transgene construct was reinjected at lower concentrations of DNA (≈2 versus ≈6 ng/μL) to obtain additional founders with fewer integrated copies of the transgene. Second, the αMHC promoter was modified in order to attenuate its expression while retaining its characteristic cardiomyocyte-specificity. Toward this end, we mutated the δV1 construct to individually ablate the two thyroid responsive elements (TRE), which have previously been reported to be necessary for full activity of this promoter.16,17⇓ Herein, the mutant promoters are designated TRE-1 when the first thyroid response element is ablated, and TRE-2 when the second site is ablated (see Figure 1).
Multiple founders were identified for each transgene construct, αMHC-δV1, TRE-1, and TRE-2. Because results were similar between lines for each promoter, the data have been combined. Expression of δV1, measured by immunoblot blot analysis of the incorporated amino terminal FLAG epitope, corresponded with the anticipated activities of the promoters; TRE-1 mice expressed 72% and TRE-2 mice 40% of the level of the reinjected, lower copy number αMHC-δV1 mouse (Figure 4A). At levels of δV1 expression achieved in the αMHC-δV1 and TRE-1 mice, heart weight increased in proportion with δV1 expression, whereas TRE-2 hearts were normal in size (Table 2). Likewise, the increase in fetal cardiac genes that characterizes cardiomyopathic syndromes25 also showed a relationship to δV1 expression level (Figure 4A). Compared with NTG (and TRE-2) mice, αMHC-δV1 mouse ventricles had a 3-fold increase in βMHC and a 15-fold increase in ANF mRNA expression, whereas TRE-1 mice similarly demonstrated a 3-fold increase in βMHC, but an 8-fold increase in ANF mRNA. In addition, αMHC-δV1 mice had a 50% reduction in SERCA expression, compared with a 30% reduction for TRE-1; again, TRE-2 mice were similar to NTG. TRE-1 and TRE-2 mice exhibited no increase in mortality up to 12 months of age. However, like those of the original δV1 line, first generation mice from the reinjected αMHC-δV1 founder died prematurely from apparent heart failure, with a mean survival time of 18±2 weeks (n=44).
The effects of different levels of δV1 expression on myocardial δPKC activation, assessed as translocation to subcellular membrane particulates, were assessed in isolated perfused mouse hearts. As shown in Figure 4B, a 40-minute period of global ischemia followed by 20 minutes of reperfusion stimulated a marked redistribution of δPKC from the cytosolic to membrane subcellular fraction (compare control NTG to ischemia/reperfused NTG). In contrast, translocation was significantly inhibited in TRE-2 hearts (76±2.3% of NTG), and TRE-1 hearts (42±4.4% of NTG; P≤0.05 versus NTG and TRE-2). No effect of δV1 expression was seen on subcellular partitioning of related εPKC (Figure 4B). Thus, variable myocardial expression of δV1 has specific, dose-related inhibitory effects on ischemia/reperfusion-stimulated δPKC translocation.
Because a hallmark feature of the original δV1 cardiomyopathy was the presence of intracellular desmin/crystallin-containing aggregates, we performed histological examination to establish that the phenotypic continuum achieved with variable δV1 expression extended to these characteristic structural abnormalities. Although (as hoped) none of the new δV1 transgenic lines developed a phenotype as severe as the original, the proportion of cardiomyocytes (percent of total) with aggregates that stained blue with Masson’s Trichrome correlated with δV1 expression (7.7±0.6% αMHC-δV1, 1.1±0.4% TRE-1, 0% TRE-2) (Figure 5). Furthermore, cardiac functional abnormalities paralleled the alterations in cardiomyocyte/myocardial structure. Invasive hemodynamic analysis revealed that both the αMHC-δV1 and TRE-1 mice had significantly depressed basal and stimulated +dP/dt. At the maximal dobutamine dose, αMHC-δV1 and TRE-1 mice exhibited 43% and 28% reductions in +dP/dt, respectively (12265±740 mm Hg/s αMHC-δV1, 15418±69 TRE-1, and 21217±720 NTG; P≤0.05 versus NTG). As with histological appearance, TRE-2 mice were functionally normal (peak +dP/dt 18802±2102 mm Hg/s; P=NS).
The varying “dose” of δV1 expression afforded by the mutant promoters permitted a correlative analysis of biochemical δPKC inhibition and functional response to ischemia/reperfusion. TRE-2 mouse hearts, with only modest δPKC inhibition and no detectable abnormalities in structure, function, or gene expression were nevertheless markedly resistant to global ischemia, with rapid recovery of contractile function (Figures 6A and 6B) and strikingly less myocyte damage (assessed as CK release) compared with NTG (Figure 6C). TRE-1 mouse hearts, with significantly decreased δPKC translocation inhibition (see Figure 4B), appeared at first glance to have increased ischemic injury (Figure 6A). However, the basal function of TRE-1 hearts was depressed due to the modest myofibrillar cardiomyopathy (Figures 5 and 6⇓A), and when the postischemic data are corrected for the baseline functional abnormality, there is no difference in post-ischemic cardiac functional recovery compared with NTG controls (Figure 6B). However, as with TRE-2, a striking cytoprotective effect was evident (Figure 6C). It was not possible to assess the effects of ischemia/reperfusion on the αMHC-δV1 mice, due to the severity of the cardiomyopathy.
These studies establish dose-dependent effects of a δPKC translocation inhibitor in the in vivo mouse heart. Lower levels of δV1 expression, which inhibited ischemia-stimulated δPKC translocation by 24%, protected hearts from ischemia/reperfusion injury. Importantly, these benefits occurred in the absence of any detectable structural or functional alterations, including highly sensitive assays of myocardial gene expression. With progressively more δPKC inhibition however, cardiomyopathies developed exhibiting pathological and histological characteristics of myofibrillar myopathies. Cytoskeletal abnormalities in these δV1 cardiomyopathies, together with targeting of δV1 (and hence activated δPKC) to the cardiomyocyte cytoskeleton, identify a critical function for δPKC in maintaining myocyte cytoskeletal integrity.
An unexpected finding of these studies was the degree of cardiomyocyte protection (as assessed by CK release) afforded by δV1 expression in both the TRE-1 and TRE-2 lines (Figure 6C). We considered that the presence of protein aggregates and depressed basal contractile function of TRE-1 mice would compromise functional recovery after ischemia/reperfusion. Indeed, absolute recovery of contractile function was severely depressed in TRE-1 mice, although when normalized to basal function, TRE-1 mice had exactly the same degree of functional recovery as NTG controls (55.5±8.6% NTG versus 55.5±15.3% TRE-1). Nevertheless, a dramatic cytoprotective effect was observed, exceeding even that of the functionally superior TRE-2 mice. A possible explanation relates to increased expression of αB-crystallin in the TRE-1 mice, which was not seen in TRE-2. αB-crystallin, aka, heat shock protein 22, is a chaperone for the intermediate filament desmin and is known to exert a cardioprotective effect in ischemic preconditioning.26 Thus, δV1-mediated upregulation of αB-crystallin could contribute to cytoprotection in these hearts. It is therefore interesting that activation of εPKC, which also protects from postischemic cardiac dysfunction, is rather less cytoprotective than δPKC inhibition, and is not associated with regulated expression of αB-crystallin or desmin9 (unpublished results). Another difference between cardiac protection afforded by transgenically expressed ψεRACK versus δV1 was delayed development of mild, albeit normally functioning ventricular hypertrophy, with induction of βMHC gene expression in ψεRACK mice.9,10⇓ Absence of any such physical or molecular modifications by the chronically expressed δPKC inhibitor (at low expression levels) may be advantageous in terms of a potential therapeutic approach.
When δV1 was expressed at incrementally higher levels in the heart, a cardiomyopathy developed with severity in proportion to δV1 expression level. The most surprising aspect of higher-level δV1 expression was that this cardiomyopathy had the histopathological and functional characteristics of specific human genetic disorders, the myofibrillar myopathies. These skeletal and cardiac myopathies are a heterogeneous group of diseases that share a common histopathological picture of intramyocyte proteinaceous aggregates with sarcomere disorganization.13,14,22⇓⇓ The best characterized of this group are the desmin-related myopathies caused by mutant forms of desmin or αB-crystallin. Missense mutations in the carboxy-terminal part of the desmin rod domain cause myopathies by interfering with desmin assembly.13 The αB-crystallin R120G mutation leads to myopathy by impairing its ability to act as a molecular chaperone for desmin, and interfering with proper desmin integration into intermediate filaments.14,22⇓ Thus, pathological mutations of either of these proteins result in abnormal accumulation of protein into characteristic intracellular aggregates, accompanied by cytoskeletal and sarcomere disruption. Interestingly, a similar phenotype has been observed in the desmin knockout mouse model,27,28⇓ which implies that the pathology is associated with loss of desmin function, not increased desmin expression as observed in the δV1 mouse model and other hypertrophy conditions.29 Indeed, our present studies demonstrate that neither overexpression of normal desmin alone, nor its overexpression in combination with αB-crystallin, causes any measurable functional or structural myocardial abnormalities. Rather, the observation that activated δPKC in cardiac myocyte translocates to the cytoskeleton establishes a mechanistic link between δPKC function and myofibrils. Studies in cultured rat neonatal cardiomyocytes and in δV1 mice (data not shown) demonstrate that activated δPKC colocalizes with cytoskeletal vimentin,30,31⇓ and that vimentin and δPKC coimmunoprecipitate.30 Thus, δPKC substrates exist at the cytoskeleton, and our results indicate that attenuated δPKC activation causes loss of cytoskeletal organization. The identities of critical cytoskeletal δPKC substrates, and of the cytoskeletal δPKC binding protein, ie, δPKC RACK, are currently unknown.
The δV1 transgenic approach interrogates δPKC function in the in vivo mouse heart through targeted expression of a protein fragment designed to competitively inhibit δPKC binding with its membrane anchor protein, or RACK.12,32⇓ We have previously used the approach of translocation inhibition to characterize in vivo functions of myocardial εPKC.9–11⇓⇓ In those studies, εPKC inhibition by transgenic expression of its first variable region, εV1, had little effect at lower expression levels, but at the highest expression level caused a lethal dilated cardiomyopathy.10 In both the current δV1 model, and the previous εV1 model, expression of PKC V1 peptides at lower levels was well tolerated, but higher expression levels resulted in development of lethal cardiomyopathies. However, the cardiomyopathies themselves are strikingly different. Whereas εV1 expressors developed profound ventricular wall thinning and chamber dilation, δV1 expressors had thick ventricular walls and normal chamber dimensions, associated with the aforementioned histological and ultrastructural abnormalities. Likewise, the antithetic experiments, ie, facilitated translocation of δPKC and εPKC through transgenic expression of their respective ψRACK peptides, resulted in similar hypertrophic phenotypes, but opposite effects on myocardial recovery from ischemia.10,12⇓ Taken together, these observations indicate that δPKC and εPKC have distinct roles in myocardial cytoskeletal maintenance and the response to ischemic injury, but an overlapping function in myocardial growth. This latter functional redundancy may explain why mutational ablation of the εPKC gene has no overt cardiac phenotype,33 because δPKC may opportunistically compensate for the absent εPKC.
In the present studies, a great deal of effort went into generating mouse models with varying levels of transgene expression. The need was to generate δV1-expressing animals that survived long enough to study, and to assess the dose-dependent effects of the transgene while retaining the favorable characteristics of αMHC promoter, ie, robust, cardiomyocyte-specific ventricular transgene expression that largely begins after birth. Thus, we chose to mutationally attenuate the αMHC promoter. Attenuated versions of the αMHC promoter were obtained by individually mutating the thyroid response elements, as originally conceived by Rindt and Robbins,16,17⇓ and the resulting TRE-1 and TRE-2 δV1 phenotypes represent form frustes of the original αMHC-δV1 phenotype, permitting a detailed analysis. Importantly, multiple independent lines of TRE-2 mice did not develop a cardiomyopathy when followed up to one year, indicating that a threshold expression level exists for pathological effects of δPKC inhibition. We have observed similar threshold effects when overexpressing either intact signaling molecules, such as Gαq,19 or catalytically inactive peptide fragments, such as the εPKC inhibitor, εV1.10 Gαq overexpression at twice endogenous levels had no effect on cardiac structure, function, or gene expression, whereas 4- and 5-fold Gαq overexpressors developed the characteristic hypertrophy, contractile dysfunction, and abnormalities in gene expression.19 Likewise, in the lowest expressing line (≈8 copies) cardiac-specific expression of the εPKC inhibitory peptide, εV1, had no detectable basal phenotype (although a phenotype was revealed by superimposed expression with Gαq11), whereas the highest expressing line (>200 copies) died prematurely of dilated cardiomyopathy, and an intermediate expressing line (≈40 copies) exhibited modest but significant alterations in contractile function and increased fetal cardiac gene expression.10 In the context of the present and previous studies where transgenesis has been utilized as a drug delivery system for catalytically inactive peptides (Gq-mini gene,34 βARK-ct35), delineation of dose-dependent effects should be an essential component of the phenotypic characterization, and the TRE mutant αMHC promoter constructs have demonstrated utility for this purpose.
In summary, we have delineated essential roles for δPKC in the in vivo myocardial response to ischemic injury and for normal cardiomyocyte cytoskeletal integrity. A critical function for δPKC in maintenance of normal cardiomyocyte cytoskeletal structure was indicated not only by the histopathological/ultrastructural findings and associated contractile dysfunction in mice expressing δV1, but also by decoration of intermediate filaments by δV1, thus identifying the cardiomyocyte cytoskeleton as a subcellular target for activated δPKC. Despite the deleterious effects of high-level δV1 expression, the potential therapeutic utility of δPKC inhibition for postischemic myocardial protection is supported by identification of a therapeutic window, where low-level δV1 expression is protective, without adverse consequences.
This work is supported by P50 HL52318 and R01 HL58010. We wish to thank Gary Lin for his technical contributions to this project.
Original received July 1, 2002; revision received August 28, 2002; accepted September 3, 2002.
- ↵Gu X, Bishop SP. Increased protein kinase C and isozyme redistribution in pressure-overload cardiac hypertrophy in the rat. Circ Res. 1994; 75: 926–931.
- ↵Schunkert H, Sadoshima J, Cornelius T, Kagaya Y, Weinberg EO, Izumo S, Riegger G, Lorell BH. Angiotensin II–induced growth responses in isolated adult rat hearts: evidence for load-independent induction of cardiac protein synthesis by angiotensin II. Circ Res. 1995; 76: 489–497.
- ↵Bowling N, Walsh RA, Song G, Estridge T, Sandusky GE, Gouts RL, Mintze K, Pickard T, Roden R, Bristow MR, Sabbah HN, Mizrahi JL, Gromo G, King GL, Vlahos CJ. Increased protein kinase C activity and expression of Ca2+-sensitive isoforms in the failing human heart. Circulation. 1999; 99: 384–391.
- ↵Speechly-Dick ME, Mocanu MM, Yellon DM. Protein kinase C: its role in ischemic preconditioning in the rat. Circ Res. 1994; 75: 586–590.
- ↵Qiu Y, Ping P, Tang XL, Manchikalapudi S, Rizvi A, Zhang J, Takano H, Wu WJ, Teschner S, Bolli R. Direct evidence that protein kinase C plays an essential role in the development of late preconditioning against myocardial stunning in conscious rabbits and that epsilon is the isoform involved. J Clin Invest. 1998; 101: 2182–2198.
- ↵Nishizuka Y. Studies and perspectives of protein kinase C. Science. 1986; 233: 305–312.
- ↵Nishizuka Y. Protein kinase C and lipid signaling for sustained cellular responses. FASEB J. 1995; 9: 484–496.
- ↵Newton AC. Protein kinase C: structure, function and regulation. J Biol Chem. 1995; 270: 28495–28498.
- ↵Dorn GW II, Souroujon MC, Liron T, Chen CH, Gray MO, Zhou HZ, Csukai M, Wu G, Lorenz JN, Mochly-Rosen D. Sustained in vivo cardiac protection by a rationally designed peptide that causes ε protein kinase C translocation. Proc Natl Acad Sci U S A. 1999; 96: 12798–12803.
- ↵Mochly-Rosen D, Wu G, Hahn HS, Osinska H, Liron T, Lorenz JN, Yatani A, Robbins J, Dorn GW II. Cardiotrophic effects of protein kinase C ε: analysis by in vivo modulation of PKCε translocation. Circ Res. 2000; 86: 1173–1179.
- ↵Wu G, Toyokawa T, Hahn HS, Dorn GW II. εProtein kinase C in pathological myocardial hypertrophy. J Biol Chem. 2000; 39: 29927–29930.
- ↵Chen L, Hahn HS, Wu G, Chen CH, Liron T, Schechtman D, Cavallaro G, Banci L, Guo Y, Bolli R, Dorn GW II, Mochly-Rosen D. Opposing cardioprotective actions and parallel hypertrophic effects of δ PKC and ε PKC. Proc Natl Acad Sci U S A. 2001; 98: 11114–11119.
- ↵Johnson JA, Gray MO, Chen CH, Mochly-Rosen D. A Protein Kinase C Translocation Inhibitor as an Isozyme-selective Antagonist of Cardiac Function. J Biol Chem. 1996; 271: 24962–24966.
- ↵Subramaniam A, Gulick JU, Neumann J, Knotts S, Robbins J. Transgenic analysis of the thyroid-responsive elements in the α-cardiac myosin heavy chain gene promoter. J Biol Chem. 1993; 268: 4331–4336.
- ↵D’Angelo D, Sakata Y, Lorenz JN, Boivin GP, Walsh RA, Liggett SB, Dorn GW II. Transgenic Gαq overexpression induces cardiac contractile failure in mice. Proc Natl Acad Sci U S A. 1997; 94: 8121–8126.
- ↵Bova MP, Yaron O, Huang Q, Ding L, Haley DA, Stewart PL, Horwitz J. Mutation R120G in αB-crystallin, which is linked to a desmin-related myopathy, results in an irregular structure and defective chaperone-like function. Proc Natl Acad Sci U S A. 1999; 96: 6137–6142.
- ↵Wang X, Osinska H, Dorn GW II, Nieman M, Lorenz JN, Gerdes AM, Witt S, Kimball T, Gulick J, Robbins J. Mouse model of desmin-related cardiomyopathy. Circulation. 2001; 103: 2404–2407.
- ↵Wang X, Osinska H, Klevitsky R, Gerdes AM, Nieman M, Lorenz JN, Hewett T, Robbins J. Expression of R120G-αB-crystallin causes aberrant desmin and αB-crystallin aggregation and cardiomyopathy in mice. Circ Res. 2001; 89: 84–91.
- ↵Chien KR, Knowlton KU, Zhu H, Chien S. Regulation Of cardiac gene expression during myocardial growth and hypertrophy: molecular studies of an adaptive physiologic response. FASEB J. 1991; 5: 3037–3046.
- ↵Ray PS, Martin JL, Swanson EA, Otani H, Dillmann WH, Das DK. Transgene overexpression of αB-crystallin confers simultaneous protection against cardiomyocyte apoptosis and necrosis during myocardial ischemia and reperfusion. FASEB J. 2001; 15: 393–402.
- ↵Milner DJ, Weitzer G, Tran D, Bradley A, Capetanaki Y. Disruption of muscle architecture and myocardial degeneration in mice lacking desmin. J Cell Biol. 1996; 134: 1255–1270.
- ↵Heling A, Zimmermann R, Kostin S, Maeno Y, Hein S, Devaux B, Bauer E, Klövekorn WP, Schlepper M, Schaper W, Schaper J. Increased expression of cytoskeletal, linkage, and extracellular proteins in failing human myocardium. Circ Res. 2000; 86: 846–853.
- ↵Goodnight J, Mischak H, Kolch W, Mushinski JF. Immunocytochemical localization of eight protein kinase C isozymes overexpressed in NIH 3T3 fibroblasts: isoform-specific association with microfilaments, Golgi, endoplasmic reticulum, and nuclear and cell membranes. J Biol Chem. 1995; 270: 9991–10001.
- ↵Mochly-Rosen D. Localization of protein kinases by anchoring proteins: a theme in signal transduction. Science. 1995; 268: 247–251.
- ↵Akhter SA, Luttrell LM, Rockman HA, Iaccarino G, Lefkowitz RJ, Koch WJ. Targeting the receptor-Gq interface to inhibit in vivo pressure overload myocardial hypertrophy. Science. 1998; 280: 574–577.
- ↵Koch WJ, Rockman HA, Samama P, Hamilton RA, Bond RA, Milano CA, Lefkowitz RJ. Cardiac function in mice over expressing the β-adrenergic receptor kinase or a βARK inhibitor. Science. 1995; 268: 1350–1353.