|Home | About | Journals | Submit | Contact Us | Français|
Protein kinase Cα (PKCα), PKCβ, and PKCγ comprise the conventional PKC isoform subfamily, which is thought to regulate cardiac disease responsiveness. Indeed, mice lacking the gene for PKCα show enhanced cardiac contractility and reduced susceptibility to heart failure. Recent data also suggest that inhibition of conventional PKC isoforms with Ro-32-0432 or Ro-31-8220 enhances heart function and antagonize failure, although the isoform responsible for these effects is unknown. Here we investigated mice lacking PKCα, PKCβ and PKCγ for effects on cardiac contractility and heart failure susceptibility. PKCα−/− mice, but not PKCβγ−/−, showed increased cardiac contractility, myocyte cellular contractility, Ca2+ transients, and sarcoplasmic reticulum Ca2+ load. PKCα−/− mice were less susceptible to heart failure following long-term pressure overload stimulation or 4 weeks after myocardial infarction injury, while PKCβγ−/− mice showed more severe failure. Infusion of ruboxistaurin (LY333531), an orally available PKCα/β/γ inhibitor, increased cardiac contractility in wildtype and PKCβγ−/− mice, but not in PKCα−/− mice. More importantly, ruboxistaurin prevented death in wildtype mice throughout 10 weeks of pressure overload stimulation, reduced ventricular dilation, enhanced ventricular performance, reduced fibrosis, and reduced pulmonary edema comparable to or better than metoprolol treatment. Ruboxistaurin was also administered to PKCβγ−/− mice subjected to pressure overload, resulting in less death and heart failure, implicating PKCα as the primary target of this drug in mitigating heart disease. As an aside, PKCαβγ triple null mice showed no defect in cardiac hypertrophy following pressure overload stimulation. In conclusion, PKCα functions distinct from PKCβ and PKCγ in regulating cardiac contractility and heart failure, and broad acting PKC inhibitors such as ruboxistaurin, could represent a novel therapeutic approach in treating human heart failure.
The protein kinase C family of Ca2+ and/or lipid-activated serine-threonine protein kinases function downstream of many signal transduction pathways.1,2 Approximately 10 different isozymes comprise the PKC family, which are broadly classified by their activation characteristics. The conventional PKC isozymes (PKCα, βI/II, and γ) are Ca2+- and lipid-activated, while the novel isozymes (ε, θ, η, and δ) and atypical isozymes (ζ, and λ) are Ca2+ independent but activated by distinct lipids.1,2 PKCα is the predominant conventional PKC isoform expressed in the mouse, rabbit, and human heart, while PKCβ and PKCγ are expressed at lower but detectable levels.3–5 Hemodynamic pressure overload in rodents promotes translocation and presumed activation of PKCα, β, γ, ε, and θ, during the hypertrophic phase or during later stages of heart failure.6–10 Human heart failure has also been associated with increased activation of conventional PKC isoforms.11–13
Overexpression of either wild-type or a constitutively active deletion mutant of PKCβ in a mouse heart was reported to induce cardiomyopathy,14,15 but more recent investigation has suggested that lower levels of expression or adult onset PKCβ activation benefits ischemic recovery,16,17 suggesting that PKCβ may be slightly protective when activated. In contrast, PKCβ−/− mice showed reduced infarct size following myocardial ischemia-reperfusion injury and greater functional recovery, similar to treatment with ruboxistaurin.18 Thus, the function of PKCβ as a disease modifier in the heart remains controversial and almost nothing is known of PKCγ’s function in the heart. By comparison, transgenic mice with greater PKCα activity showed decreased cardiac contractility, ventricular dilation, and secondary hypertrophy, suggesting that increased PKCα signaling is detrimental to the heart.19,20 Indeed, PKCα−/− mice are protected from insults or genetic mutations that would otherwise induce heart failure.19,21 Transgenic mice with inducible expression of a dominant negative PKCα mutant in cardiomyocytes of the heart also showed reduced failure progression after myocardial infarction (MI) injury.21
Inhibition of PKCα has been associated with protection from heart failure through at least 2 different mechanisms; one involving augmented cardiac Ca2+ handling and the other involving enhanced myofilament performance.19,22,23 PKCα phosphorylates inhibitor 1 (I-1) at a novel site, resulting in greater protein phosphatase 1 (PP1) activity, leading to greater phospholamban (PLN) dephosphorylation and less activity of the sarcoplasmic reticulum Ca2+ ATPase (SERCA2) pump.19,24 Less SERCA2 activity reduces sarcoplasmic reticulum (SR) Ca2+ load leading to reduced Ca2+ release during systole, hence reduced contractility. Alternatively, PKCα appears to directly phosphorylate key contractile proteins, leading to reduced force production in isolated, skinned myocytes.23 Here we show that PKCα functions distinct from PKCβ and PKCγ in the heart, and that the cardioprotective effect of the conventional PKC inhibitor ruboxistaurin is likely due to PKCα inhibition.
The PKCα−/−, PKCβ−/− and PKCγ−/− mice have been described previously.19,25,26 Transverse aortic constriction (TAC) and MI were performed as previously described by us.27 The β-adrenergic receptor antagonist metoprolol (Sigma-Aldrich) was given in the drinking water at 2 g/l as previously reported.28 Cardiac histological sections were analyzed by Metamorph as previously described.29 Ventricular myocytes were isolated as described previously.30 Intracellular Ca2+ was measured with Fluo-4 as described previously.31
Here we utilized genetically modified mouse models lacking PKCα, PKCβ, and PKCγ to gain insight into which isoform might underline propensity to heart failure upon stress stimulation. Fortunately, combinatorial deleted mice were viable as adults, although PKCαβγ triple nulls had reduced body weights and hence were not extensively studied. Western blotting from cardiac extracts with antibodies specific to PKCα, PKCβ, and PKCγ showed loss of each protein in the correctly targeted single, double and triple null mice (Figure 1A). Cardiac PKCε and PKCδ levels were also not changed in the PKCαβγ gene-deleted mice (data not shown).
Given our previous results that PKCα−/− mice are protected from developing heart failure following pressure overload stimulation, we were interested in examining if PKCβ and PKCγ functioned similarly. Given a concern of compensation we reasoned that PKCβγ double null mice would be best to compare against PKCα−/− mice. Measurement of cardiac contractility by invasive hemodynamic assessment with a Millar pressure transducing catheter showed that PKCα−/− mice, but not PKCβγ−/−, had enhanced contractility and relaxation at baseline (Figure 1B–D). We also generated individual myocytes from adult hearts of these mice for assessment of contractile function in isolation. Cardiac myocytes from PKCα−/− mice, but not PKCβγ−/− showed greater fractional shortening, greater contractility and faster relaxation (Figure 1E–H). Thus, loss of PKCα showed greater cardiac contractile performance in vivo and in isolated adult myocytes, while loss of both PKCβ and PKCγ had no effect on these parameters.
Consistent with greater cellular contractility, adult myocytes from PKCα−/− hearts showed greater Ca2+ transient amplitudes (Figure 2A). Interestingly, myocytes from PKCβγ null hearts had a subtle, albeit significant reduction in the amplitude of the Ca2+ transient (Figure 2A). We had previously shown that myocytes from PKCα−/− hearts had greater SR Ca2+ loads due to a mechanisms associated with altered PLN phosphorylation.19 We confirmed greater SR Ca2+ loads in PKCα−/− myocytes compared with wildtype myocytes, although PKCβγ deficient myocytes had a small but significant reduction in SR Ca2+ load (Figure 2B,C). Similarly, PKCα−/− mice showed enhanced PLN phosphorylation at serine-16, but not at threonine-17, which would increase SERCA2 activity and SR Ca2+ load (Figure 2E). PKCβγ−/− mice showed no change in PLN serine-16 phosphorylation, although we consistently observed a small increase in threonine-17 phosphorylation (Figure 2E). No other changes were observed in levels of PLN, SERCA2 or ryanodine receptor (RyR) proteins in the hearts of either genotype of mice. Increased PKCβ or PKCγ protein levels in cultured neonatal myocytes by adenoviral gene transfer also did not alter the phosphorylation status of I-1 (data not shown). Isolated cardiac myocytes from PKCβγ−/− mice showed no difference in myofilament Ca2+ sensitivity (Figure 2D). Together, these results suggest that PKCα functions distinct from PKCβ and PKCγ in regulating Ca2+ handling and contractility.
While not the main focus of the current study, conventional PKC isoforms have been previously implicated in regulating cardiomyocyte hypertrophy in culture.32 While we failed to observe any reduction in pressure overload-induced hypertrophy in mice lacking PKCα, and Buttrick and colleagues similarly failed to observe a reduction in PKCβ−/− mice, the possibility of compensation amongst the 3 conventional isoforms remained a concern.19,33 We performed 2 weeks of TAC stimulation in wildtype, PKCα−/− PKCβγ−/− and PKCαβγ−/− mice, which revealed no attenuation of heart growth in any group (Figure 3A). Thus, loss of all conventional PKC isoforms has no attenuating affect on pressure overload-induced hypertrophy.
Longer periods of TAC stimulation can produce heart failure and even greater increases in heart weights due to ventricular remodeling and chamber dilation. Here we subjected individual PKCα, PKCβ, PKCγ, and PKCβγ null mice to 8 weeks of TAC to analyze their susceptibility to heart failure. While all genotypes continued to show a robust hypertrophic response after 8 weeks of TAC, the PKCβγ−/− mice showed significantly heavier hearts, suggesting more severe failure (Figure 3B). As we have previously reported,19 assessment of fractional shortening after 8 weeks of TAC showed significant protection in PKCα−/− mice compared with wildtype controls (Figure 3C). By comparison, loss of PKCβ or PKCγ separately did not significantly improve or worsen fractional shortening compared with the reduction observed in wildtype controls (Figure 3C). However, PKCβγ−/− mice showed a significant worsening of ventricular performance after TAC compared with wildtype mice (Figure 3C). Indeed, PKCβγ mice were the only group that showed significant increases in lung weights, indicative of pulmonary edema and more severe heart failure (Figure 3D). These results indicate that loss of PKCα protects the heart from pressure overload associated decompensation, while loss of PKCβγ renders the heart more susceptible to decompensation.
The conclusion concerning the disparity in function between PKCα and PKCβγ in pressure overload stimulated mice was further confirmed in another independent model of heart failure due to MI injury. The left coronary artery was permanently ligated and function was followed by echocardiography every week for 4 weeks. No difference in infarction size or scar size was noted amongst the groups, suggesting that any alterations in performance was likely due to remodeling and/or heart failure progression differences between the groups (data not shown). Consistent with past results, PKCα−/− mice maintained function better at weeks 2, 3, and 4, compared to wildtype mice after MI (Figure 3E). The reduction in fractional shortening in PKCβ or PKCγ single nulls was not different from wildtype, although PKCβγ double nulls showed significantly worse function at all 4 time points (Figure 3E). Collectively, these data suggest that PKCα is normally cardiomyopathic when activated in the heart during stress stimulation, while PKCβγ function in a slightly protective manner when activated (see discussion).
Ruboxistaurin, a previously reported PKCβ selective antagonist, has been through late stage clinical trials for diabetic macular edema and shown to be well tolerated.34 While ruboxistaurin was reported to be PKCβ selective,35 we determined that it was equally selective for PKCα (IC50 of 14 nM for PKCα versus 19 nM for PKCβII). These values for ruboxistaurin were obtained from Upstate Biotechnology with their IC50Profiler Express™ system. Given that PKCα protein levels are also much higher than PKCβ in the human heart,5 it further suggests that ruboxistaurin might function predominantly through a PKCα-dependent mechanism. Indeed, we directly measured cardiac contractility upon acute ruboxistaurin infusion in mice lacking either PKCα or PKCβγ. We previously observed that ruboxistaurin administration to rats increased baseline contractility by 28%.5 Here, wildtype control mice showed a small but significant increase in cardiac contractility within minutes of ruboxistaurin infusion, as did PKCβγ−/− mice (Figure 4A). However, ruboxistaurin infusion had no ability to further augment contractility in PKCα−/− mice (Figure 4A). These results indicate that ruboxistaurin enhances cardiac function specifically through effects on PKCα, but not PKCβ or PKCγ.
To explore the potential therapeutic ramifications of our results we instituted a large study with nearly 100 mice (C57BL/6) treated with 2 doses of ruboxistaurin (40 and 120 mg/kg/day) administered in chow. Other groups included chow vehicle or a known agent used to treat heart failure, the β-adrenergic receptor antagonist metoprolol (2 g/l in drinking water) (Figure 4B). At 8 weeks of age mice were subjected to TAC to induce pressure overload on the heart, or a sham procedure as a control. After 1 week of TAC all mice showed comparable pressure gradients across the aortic constrictions (data not shown). Throughout the treatment period, low and high dosage ruboxistaurin groups had blood levels of drug-metabolite of approximately 50–200 ng/ml and 400–1400 ng/ml, respectively (measured every 2 weeks).
Within a few weeks of TAC stimulation vehicle treated mice began to show mortality, such that only 6 of 15 enrolled mice were alive at week 10 (Figure 4C). Metoprolol and low dose ruboxistaurin treated mice showed slightly less lethality, while high dose ruboxistaurin treated mice showed no lethality whatsoever (Figure 4C). None of the mice in any of the 4 sham groups died over the 10-week treatment period. These results suggest that ruboxistaurin protected mice from lethality associated with chronic cardiac pressure overload stimulation.
Metoprolol and low and high ruboxistaurin treatment groups all showed significantly better fractional shortening compared with vehicle treated mice after 10 weeks of TAC (Figure 5A). Moreover, all groups subjected to TAC showed dilation of the left ventricle in diastole. However, the high dosage ruboxistaurin group exhibited significantly less dilation than the vehicle group (Figure 5B). Direct measurement of heart weights normalized to body weight also showed significantly less secondary hypertrophy associated with the failing heart in both the metoprolol and high dose ruboxistaurin treatment groups (Figure 5C). However, assessment of myocyte cross-sectional areas from histological sections showed comparable hypertrophic increases, suggesting that the greater increase in organ hypertrophy in the vehicle and low dose ruboxistaurin groups was due to greater remodeling (Figure 5D). Assessment of lung weight normalized to body weight showed significant protection in the metoprolol and high ruboxistaurin groups compared with vehicle treated mice after TAC (Figure 5E). Evaluation of fibrosis from ventricular histological sections stained with Masson’s trichrome showed a significant degree of fibrotic material deposition in all 4 groups subjected to TAC, although the three drug treatment groups had significantly less (Figure 5F). These results collectively indicate that ruboxistaurin antagonizes the progression and severity of heart failure in mice subjected to chronic pressure overload stimulation.
To verify that the protective effects or ruboxistaurin are due to PKCα inhibition, but not PKCβγ, we conducted a second independent study involving treatment of PKCβγ−/− mice subjected to TAC for 4 weeks. PKCβγ−/− mice were more sensitive to TAC than wildtypes, such that only 3 of 11 PKCβγ−/− mice treated with vehicle food survived 4 weeks of TAC, while 8/9 ruboxistaurin treated (high dosage) PKCβγ−/− mice survived (Figure 6A). Separate wildtype controls were also performed, and once again, no wildtype mice died during TAC on ruboxistaurin, but 2 of 7 vehicle treated mice died after 4 weeks (Figure 6A). Other indexes of heart failure were also positively influenced by ruboxistaurin treatment in PKCβγ−/− mice, such as an improvement in fractional shortening, reductions in secondary increases in heart weight normalized to body weight, prevention of pulmonary edema, and less fibrosis (Figure 6B,C,D,E). These results indicate that ruboxistaurin influences heart failure susceptibility and death after TAC stimulation independent of PKCβγ, suggesting it benefits the heart by inhibition of PKCα or an entirely different kinase.
Ruboxistaurin was initially touted as a PKCβ selective inhibitor, although our analysis suggested equal selectively towards PKCα and PKCβ, and likely even PKCγ. While this observation does not singularly implicate PKCα as the primary biologic target for ruboxistaurin in mediating cardioprotection, a stronger case emerges when various points of datum are collectively analyzed. For example, we observed that another PKC inhibitor with selectivity for the conventional PKC isoforms, Ro-31-8220, enhanced cardiac contractility in mice and restored ventricular function in a mouse model of dilated cardiomyopathy.5 Similarly, a third conventional PKC isoform inhibitor, Ro-32-0432, also increased cardiac function in 2 different mouse models of heart failure, and more importantly, it did not increase contractility in PKCα−/− mice.5 While each of these pharmacologic inhibitors can have non-selective effects on other kinases, such as a known effect of ruboxistaurin on PDK1,36 that all three have an identical effect on enhancing cardiac contractility and antagonizing heart failure strongly suggests it is the commonality in blocking conventional PKC isoforms that is mechanistically important. Indeed, acute infusion of PMA, a broad acting PKC activator, reduced cardiac contractility in Wt but not PKCα−/− isolated hearts in an ex vivo preparation.5 Here we extended the case even further by showing that acute infusion of ruboxistaurin augmented cardiac contractility in wildtype and PKCβγ−/− mice, but not PKCα−/− mice. More importantly, ruboxistaurin protected PKCβγ−/− mice from heart failure and death following pressure overload stimulation, unequivocally demonstrating that this drug is not mediating cardioprotection by inhibition of PKCβ or γ. These results strongly suggest that the contractile and cardioprotective effects of ruboxstaurin are mediated through inhibition of PKCα.
The various pharmacologic studies discussed above are further buttressed by various genetic or gene therapy-based experiments. For example, deletion of PKCα in mice enhanced cardiac contractility and protected from heart failure.19 Transgene-mediated expression of dominant negative PKCα in a cardiac myocyte-specific manner similarly increased contractility and protected from heart failure.21 Moreover, adenoviral mediated gene transfer of dominant negative PKCα into a rat model of heart failure improved function.5 Finally, overexpression of wildtype PKCα reduced cardiac function and led to heart failure.19 By comparison, loss of PKCβ or PKCγ, or both together, did not enhance cardiac Ca2+ handling nor was it cardioprotective after injury. In fact, loss of PKCβγ resulted in slightly lower Ca2+ levels in the SR, which may be anti-arrhythmic. Collectively, these various lines of evidence further implicate PKCα as the main target of ruboxistaurin responsible for the cardioprotective effects observed here.
PKCα protein levels and activity are increased in end-stage heart failure.6,7,10,11–13,23 While PKCβ is also activated in heart failure, it appears to function completely different from PKCα in affecting cardiac disease states. For example, PKCβ−/− mice were not protected from pressure overload or MI-induced heart failure, similar to PKCγ−/− mice (Figure 3 and data not shown). PKCβ overexpressing transgenic mice were reported to be protected from remodeling post MI, suggesting that increased PKCβ activity could even be protective to the heart,16 although this interpretation is somewhat incongruent with the more recent observation that PKCβ−/− mice show less injury after acute ischemia-reperfusion.18 PKCβ overexpressing transgenic mice were also shown to have increased contractility and an increase in the amplitude of the Ca2+ transient,17 a phenotype that is opposite of PKCα, which negatively regulates contractility and the amplitude of the Ca2+ transient. We also observed that PKCβγ double null mice faired slightly worse after pressure overload or MI injury, suggesting that PKCβγ together might have a protective role in the heart. Thus, it is fairly certain that despite their sequence similarity and membership within the conventional subfamily, PKCα and PKCβ have different biologic functions in the heart.
While deletion of PKCβ and PKCγ appeared to negatively impact the heart, the inhibitory effect of ruboxistaurin towards PKCβ and PKCγ may not be overly concerning, as inhibiting PKCα clearly predominates in providing protection to the heart. The protective effects of ruboxistaurin are likely mediated through a mild enhancement in cardiac contractility in a myocyte autonomous manner given our previous work with transgenic mice, gene-deleted mice, and isolated heart and adult myocyte work.5,19,21 However, it remains possible that the cardioprotective effects of this drug involves non-myocytes and endocrine effects from outside the heart. Regardless of the mechanism, we believe that ruboxistaurin is an attractive agent to apply to the heart failure clinical setting, especially given its apparent safety in late phase clinical trials.34
One final issue is how PKCα inhibition mechanistically attenuates heart failure. We showed previously that PKCα directly phosphorylates I-1, resulting in altered PP1 activity, which in turn regulates PLN phosphorylation. Alterations in PLN phosphorylation regulate SERCA2 function in the heart, which controls Ca2+ loading and the magnitude of the Ca2+ transient.24 Here we observed a similar enhancement in PLN phosphorylation at serine-16 in PKCα−/−, but not PKCβγ−/− hearts. Thus, pharmacologic inhibition of PKCα activity would function at the level of SR Ca2+ handling to augment contractility. Such an enhancement in contractility through SR Ca2+ loading can be beneficial to the heart in the face of insults that promote heart failure.37 For example, deletion or inhibition of PLN dramatically increases ventricular and myocyte cellular performance and reduces or prevents cardiomyopathy in diverse models of disease.37 Adenoviral-mediated overexpression of SERCA2 in cardiac pressure overloaded rats also rescued heart failure and improved survival.37 Thus, PKCα inhibition represents an enzymatic approach towards achieving greater contractility through an SR-dependent mechanism, which could benefit heart failure susceptibility to select insults. Alternatively, it is also possible that inhibition of PKCα enhances cardiac contractility through a mechanism involving the phosphorylation status of select myofilament proteins.23 Such a mechanism could also be cardioprotective by simply enhancing the efficiency of myofilament function. Finally, it remains possible that inhibition of PKCα is cardioprotective through other unknown mechanisms. Regardless of the mechanism, our current data strongly suggest that PKC inhibitors, such as ruboxistaurin, should be evaluated in heart failure patients.
Sources of Funding
This work was supported by the National Institutes of Health (J.D.M., S.R.H., X.C.) a grant from the Fondation Leducq, and the Howard Hughes Medical Institute (J.D.M.). Q.L was supported by a fellowship from the Ohio Valley Local Affiliate American Heart Association. S.M.M was supported by a fellowship from the Pennsylvania Delaware Affiliate of the American Heart Association.
Disclosures: J.D.M. was an inventor on a filed patent application that details claims related to PKCα inhibitors as a treatment strategy for heart failure, although the application remains in litigation.