|Home | About | Journals | Submit | Contact Us | Français|
The ubiquitin-proteasome system (UPS) and autophagy are two major intracellular protein degradation pathways. The UPS mediates the removal of soluble abnormal proteins as well as the targeted degradation of most normal proteins that are no longer needed. Autophagy is generally responsible for bulky removal of defective organelles and for sequestering portions of cytoplasm for lysosomal degradation during starvation. Impaired or inadequate protein degradation in the heart is associated with and may be a major pathogenic factor for a wide variety of cardiac dysfunctions, while enhanced protein degradation is also implicated in the development of cardiac pathology. It was generally assumed that the UPS and autophagy serve distinct functions. Therefore, the functional roles of the UPS and autophagy in the hearts have been largely investigated separately. However, recent advances in understanding the shared mechanisms contributing to UPS alteration and the induction of autophagy have helped reveal the link and interplay between the two proteolytic systems in the heart. These links are exemplified by scenarios in which inadequate UPS proteolytic function leads to activation of autophagy, helping alleviate proteotoxic stress. It is becoming increasingly clear that a coordinated and complementary relationship between the two systems is critical to protect cells against stress. Several proteins including p62, NBR1, HDAC6, and co-chaperones appear to play an important role in harmonizing and mobilizing the consortium formed by the UPS and autophagy.
The ubiquitin-proteasome system (UPS) and the autophagy-lysosome pathway are the two major pathways responsible for the clearance of proteins and organelles in eukaryotic cells.1 UPS-mediated proteolysis generally consists of two steps: ubiquitination and proteasome-mediated degradation. Macroautophagy (commonly referred to as autophagy) is the process engulfing a portion of cytoplasm, often including organelles, in a membrane bound compartment for degradation by the lysosome. Therefore, autophagy mediates bulk degradation of cytoplasmic proteins and defective or surplus organelles.2-3 Earlier studies defined proteins into two categories including “short-lived” and “long-lived” proteins based on their degradation kinetics.4 The UPS predominantly degrades short-lived normal protein molecules after they have fulfilled their duty in the cell.5 On the other hand, autophagy is primarily responsible for degrading long-lived proteins.6 Notably, the distinction of substrate preference between the two proteolytic systems is relative. Recent studies indicate that the UPS can participate in the degradation of long-lived proteins while the autophagy can also be involved in the degradation of short-lived proteins.4, 7-8
An alternative way to categorize proteolysis is based on the functional outcome rather than the kinetics of degradation. From the functional point of view, the UPS degrades two types of proteins: (1) the fully functional proteins which are degraded as a regulatory mechanism, such as proteins involved in regulation of signal transduction, gene transcription, endocytosis, and cell division; and (2) abnormal or non-functional proteins whose degradation serves a critical step of posttranslational protein quality control (PQC) in the cell.9-10 Autophagy also degrades two types of proteins: functional proteins which are degraded in a bulk fashion as a nutrient recycling mechanism, and the aggregated misfolded proteins. Hence, the two protein degradation pathways cooperate with each other in the removal of misfolded proteins.11 A recent study with cultured cells has revealed that misfolded cytosolic proteins are partitioned into two distinct PQC compartments. Soluble misfolded proteins tend to migrate to a juxtanuclear compartment preferentially for degradation by the UPS. In contrast, terminally aggregated proteins are partitioned in a perivacuolar inclusion and are mainly degraded by autophagy because they can not be efficiently degraded by the proteasome.11 This review will highlight the alterations of the UPS and autophagy and their potential roles in cardiac disease, with an emphasis on the interplay between the two proteolytic pathways.
Marked increases in ubiquitinated proteins and accumulation of preamyloid oligomers in cardiomyocytes of human failing hearts suggest inadequate PQC in heart failure.12-15 Moreover, increased level of E3 ligase MDM2, ubiquitin carboxyl-terminal hydrolase (UCH), and enhanced 20S-proteasome chymotrypsin-like activities were also observed in human hearts with dilated cardiomyopathy.14, 16 Alteration of the UPS has also been observed in experimental animal models with various forms of cardiac dysfunction, including desmin-related cardiomyopathy (DRC),3 myocardial ischemia or ischemia/reperfusion (I/R),17-18 and load-induced heart disease.19-20
Desmin-related myopathy (DRM) is a family of myopathy caused by genetic mutations of desmin, αB-crystallin (CryAB), and other desmin partner proteins. Its pathological feature is the presence of aberrant desmin-positive protein aggregates in myocytes. DRC is the cardiac manifestation of DRM. Several transgenic (tg) mouse models of DRC were created by cardiomyocyte-restricted tg expression of human DRC-linked mutant genes, including a 7 amino acid deletion (R172~E178) mutation of the desmin gene (D7-des) and a missense mutation (R120G) of CryAB (CryABR120G).15, 21-22 Cardiac UPS proteolytic function has been evaluated in both CryABR120G and D7-des tg mice by our laboratory.23-25 The tremendous advantage of these studies is the employment of a tg mouse model with ubiquitous overexpression of a previously validated UPS surrogate substrate, a green fluorescence protein (GFP) modified by carboxyl fusion of degron CL1, referred to as GFPdgn.26 In both cases, an increased level of ubiquitinated proteins and significant accumulation of GFPdgn indicates proteasome functional insufficiency (PFI) in the DRC hearts despite marked increases in 20S proteasome peptidase activities. The reduction of key subunits of the 19S proteasome was observed and suggests that a key defect resides in the delivery of the targeted protein substrates by the19S into the 20S proteolytic chamber. Increased proteasome activities here are likely a compensatory mechanism for the cardiomyocytes to cope with increased misfolded proteins. When misfolded proteins exceed the degradation capacity of the UPS or escape from degradation they accumulate and trigger aberrant protein aggregation, which in turn impairs proteasome functions, forming a vicious cycle. Indeed, in vitro experiments have shown that aberrant protein aggregation is required by the DRC mutants to impair UPS in cardiomyocytes.23-25
In animal models of ischemic cardiomyopathy, declined peptidase activities accompanied by accumulation of oxidatively modified and ubiquitinated proteins were observed.27-28 Additionally, the decreased proteasome peptidase activity is associated with oxidative modification of several proteasome subunits and contributes to accumulation of ubiquitinated proteins in animal models of myocardial ischemia and I/R injury.27, 29-30
Accumulation of ubiquitinated proteins was also observed in pressure-overloaded mouse hearts.12, 31 Depressed proteasome activities occurred in pressure-overloaded mouse hearts prior to discernible cardiac dysfunction.12 However, some other studies showed elevated abundance of proteasome subunits and E3 ligases, as well as increased proteasome peptidase activities in transverse aortic constriction (TAC)-induced cardiac hypertrophy.31-33 Additionally, the transcript levels of ubiquitin B (UbB), ubiquitin ligases including atrogin-1 and muscle-specific RING finger 1 (MuRF-1), as well as the proteasomal subunit PSMB4 were increased in rat hypertrophic hearts.34 It should be pointed out that the observed up-regulation of UPS components and proteasome peptidase activities do not necessarily indicate that UPS function has surpassed the cell’s increased needs to deal with proteotoxic stress under these conditions. On the contrary, the increases in peptidase activities and in the abundance of UPS components, coupled with the elevated levels of ubiquitinated proteins, might reflect severe PFI and PQC inadequacy in the diseased hearts.31-32 It will be interesting and important to test this proposition by introducing a surrogate substrate into these disease models and assessing a change in its clearance.
Recent studies involving genetic perturbation of specific components of the UPS provide valuable information about the role of the UPS in maintaining normal cardiac function and shine light on the poorly understood pathophysiological significance of UPS dysfunction in cardiac dysfunction.
Several observations suggest that some E3-ligases play a key role in myocardial ischemia. For instance, mice deficient in CHIP (carboxyl terminus of Hsc70-interacting protein), a co-chaperone with ubiquitin ligase properties, show accelerated age-related pathophysiological phenotypes and a heightened vulnerability to I/R injury.35-36 MDM2 has also been shown to protect cardiomyocytes in both cell cultures and intact mice. Isolated cardiomyocytes overexpressing MDM2 acquire resistance to hypoxia/reoxygenation-induced cell death and phenylephrine- and endothelin-1-induced hypertrophy.37 Conversely, inactivation of MDM2 by a peptide inhibitor promotes hypoxia/reoxygenation-induced apoptosis. Consistent with the in vitro findings, hypomorphic MDM2 expression increases mouse heart’s sensitivity to I/R injury.37 The ability of MDM2 to promote ubiquitination and proteasomal degradation of p53, an important effector of I/R, likely contributes to the protective effect.38
MuRFs are an important subfamily of RING-finger E3 ligases that are specifically expressed in striated muscle. MuRF1 is a key regulator of protein kinase C (PKC)-dependent hypertrophic responses.39 Overexpressing MuRF1 attenuated phenylephrine-induced hypertrophy in isolated cardiomyocytes.40 Compared with wild type (WT) mice, mice lacking MuRF1 were found to exacerbate cardiac hypertrophic responses to pressure overload.39 An important role of MuRF3 in maintaining cardiac integrity and function after acute myocardial infarction (AMI) is suggested by the finding that mice lacking MuRF3 display normal cardiac function but are prone to post-AMI cardiac rupture. MuRF3′s cardioprotection is likely attributed to the turnover of four and a half LIM domain protein (FHL2) and γ-filamin.41 Single knockout of MuRF1, MuRF2 or MuRF3 in mice does not cause any sign of cardiac disease at the baseline but sensitizes the hearts to pathologic stimuli. Double knockout of MuRF genes do yield phenotypes under basal conditions. For instance, MuRF1/MuRF3 double-knockout mice exhibit hypertrophic cardiomyopathy and decreased cardiac function.42 Similarly, MuRF1/MuRF2 double-knockout mice develop extreme cardiac hypertrophy.43 Taken together, these studies demonstrate that MuRF genes seem to function redundantly to protect the heart.
As an F-box protein in myocytes, atrogin-1 associates with Skp1, Cullin1, and Roc1 to assemble an SCFatrogin-1 complex with ubiquitin ligase activity.44 Overexpressing atrogin-1 attenuates, whereas down-regulation of atrogin-1 enhances, phenylephrine-induced hypertrophy in isolated cardiomyocytes.45 Consistent with the observations in cell cultures, tg overexpression of atrogin-1 in mouse hearts blunts TAC-induced cardiac hypertrophy by targeting calcineurin for ubiquitin-mediated degradation.45 Conversely, mice lacking atrogin-1 develop exaggerated cardiac hypertrophy in response to voluntary exercise through regulating FoxO1 ubiquitination.46
Both detrimental and beneficial effects were reported with pharmacologically induced PSMI on myocardial ischemia. Recent data from both in vitro and in vivo experiments support the concept that short-term PSMI under certain conditions may play a protective role in myocardial I/R but sustained PSMI may be detrimental.18 The cardioprotection of PSMI may be mediated by several mechanisms including inhibition of the nuclear factor κ-B (NF-κB) activity, inactivation of G-protein-coupled receptor kinase 2 (GRK2), and inhibition of apoptosis.18
Given that NF-κB is a critical factor in the hypertrophic growth of cardiomyocytes and that the activation of NF-κB relies on the proteasome-dependent degradation of the inhibitory κB (IκB),47 it follows that PSMI likely suppresses cardiac hypertrophy through repressing NF-κB activity.48 This is confirmed by several studies. For instance, in neonatal rat cardiomyocytes, partial inhibition of the proteasome by low doses of proteasome inhibitor effectively suppressed cardiomyocyte hypertrophy.49 Moreover, in hypertensive Dahl-salt sensitive rats, low doses of PSMI significantly reduced hypertrophic heart growth.49 Similar findings have been observed in aorta-banded mouse models.32, 48 After the onset of pressure overload, the proteasome inhibitor epoxomicin decreased NF-κB activity and prevented or even reversed cardiac remodeling without discernibly affecting cardiac function.48 Further supports for a hypertrophy inhibition role of proteasome inhibitor treatment come from the findings that in a mouse model of isoproterenol-induced cardiac hypertrophy, the proteasome inhibitor PS-519 effectively prevented the development of hypertrophy and more importantly, promoted the regression of existing left ventricular hypertrophy.50 These studies with rodents provide compelling evidence that proteasome function is required for cardiac hypertrophic growth.20
However, the long term outcome of PSMI in human cardiac diseases remains to be tested. Cardiac complications including arrhythmias and heart failure have been reported in clinical use of proteasome inhibitor bortezomib in cancer patients.51-52 Interestingly, aging-associated decreases in proteasome activities do not appear to inhibit hypertension-induced cardiac hypertrophy in the human elderly.
Ablation of autophagic or lysosomal function can, under some circumstances, bedetrimental to the heart. For instance, dilated cardiomyopathy is developed by genetic impairment of autophagy in cathepsin L- and LAMP2-deficient mice.53-54 Conditional deletion of Atg5 in the adult heart leads to cardiac hypertrophy, left ventricular dilatation, contractile dysfunction, and heart failure in the absence of exogenous stress.55 Thus, basal level of autophagy is required to maintain homeostasis in cardiomyocytes.
Although it is quite clear that deficiency in autophagy can disrupt normal cardiac function, a critical question concerning the role of autophagy in the development of the cardiac diseases initiated by mutations or environmental stress, remains to be addressed. To answer this question, the autophagic activity has been evaluated in a wide range of cardiac diseases and more importantly, the effect of genetic and pharmacological manipulations of autophagy on cardiac diseases has been assessed.
Aberrant protein aggregation, in the form of pre-amyloid oligomer formation, has been observed in failing human hearts resulting from hypertrophic (HCM) or dilated cardiomyopathies (DCM).15 Therefore, at least a subset of HCM and DCM arguably belong to proteinopathy. Increased autophagic activity was reported in human hearts with DCM.56-57 Recently, the hallmark of autophagy has been detected in the heart of several mouse models of cardiac proteinopathy. For instance, significantly increased autophagosomes and lysosomal markers were detected in mouse hearts with tg expression of an 83 residue–long polyglutamine repeats (polyQ83),58 which is sufficient to cause amyloid-based cardiac dysfunction in mice.59 Similarly, increases in autophagosomes were also observed in CryABR120G-expressing cultured cardiomyocytes and CryABR120G tg mouse hearts.60 Ubiquitinated proteins, oxidized proteins, and lipofuscin accumulate in cardiomyocytes as the mouse ages. Delaying cardiac ageing in mice by suppression of phosphoinositide 3-kinase (PI3K) or its downstream effector (mTOR) was associated with increased autophagy in the heart.61 Unloading the heart of patients with DCM using the left ventricular assisting device down-regulates autophagic markers in myocardial biopsies.57 These findings suggest that autophagy activation in diseased hearts is an adaptive response to inadequate PQC.
The protective role of autophagy in CryABR120G based DRC is demonstrated by blunting autophagy by belcin1 haploinsufficiency. The latter induces greater accumulation of high molecular weight polyubiquitinated proteins in cardiomyocytes, accelerates heart failure progression, and causes early mortality in the mice expressing a human CryABR120G.60 Consistently, suppressing autophagy by 3-methyladenine (3-MA) increases aggresome accumulation in CryABR120G-expressing cultured cardiomyocytes.60 Thus, autophagy is activated and plays a compensatory role in removing misfolded proteins in cardiomyocytes of CryABR120G-based DRC.
Autophagy is induced in the heart during myocardial ischemia.62-63 Enhanced autophagy was also observed in mouse hearts and isolated cardiomyocytes subjected to simulated I/R (sI/R).63-64 The functional role of autophagy in these circumstances was investigated. In studying autophagy in an ischemic mouse model, Matsui et al. showed that attenuated autophagy enhanced myocardial injury in acute myocardial ischemia.63 Another study reported that the autophagy marker in pig hearts with repetitive myocardial ischemia was much more prevalent in areas of subendocardium where apoptosis was absent, but areas with increased apoptosis exhibited much less cathepsin B immunostaining.62 Thus, autophagy may contribute to the survival of cardiomyocytes in response to both acute and chronic myocardial ischemia.
In contrast, several reports suggest that enhanced autophagy may contribute to cell death in response to I/R and prolonged hypoxia. Matsui et al. found that autophagy was detrimental during reperfusion although it was protective during ischemia. In their study, myocardial I/R injury was attenuated in a mouse model with diminished autophagy.63 Prolonged hypoxia also induced autophagic cell death, which was reduced by treatment with 3-MA or by knockdown of beclin-1 or Atg5.65
Similarly, both detrimental and protective roles of autophagy were reported in cell cultured models subjected to sI/R. In vitro experiments of the study by Matsui et al. showed that inhibition of autophagy by beclin1 knockdown increased cell viability in response to H2O2, which is consistent with their in vivo experiments.63 The findings were verified by another study.64 However, autophagy has also been shown to protect against cell injury in cultured cardiac cells in response to sI/R. It was found that cell injury was decreased by enhanced autophagy through overexpression of beclin1 and increased by inhibition of autophagy through expression of a dominant-negative Atg5, knockdown of Beclin1, or treatment of 3-MA.66-67 These inconsistencies may arise from variations in the timing, extent, and duration of ischemia and perhaps in the methods used for autophagy manipulation.
Ischemic preconditioning (IPC) induces tolerance to subsequent ischemic episode in the heart. Several studies indicate that autophagy is involved in cardioprotection of IPC. Enhanced autophagy and up-regulated BAG-1 (Bcl-2-associated athanogene-1) were observed during hypoxia adaption both in vitro and in vivo. Furthermore, inhibition of autophagy by treating rats with either wortmannin or BAG-1 siRNA through intramyocardial injection abolished IPC-induced LC3I to LC3-II conversion and cardioprotection.68-69 These studies indicate that myocardial protection by IPC is mediated at least in part by up-regulation of autophagy in association with BAG-1 protein. Similarly, 2-chloro-N(6)-cyclopentyladenosine (CCPA), an adenosine A1 receptor agonist, markedly induces autophagy in short term and confers a preconditioning protection against sI/R injury in HL-1 cells. Moreover, the cardioprotective effect of CCPA on cells subjected to sI/R injury is abolished when autophagy is blocked with a dominant negative Atg5.63 The role of autophagy in preconditioning can also be inferred from agents known to activate autophagy also promote cardioprotection. For instance, rapamycin, an autophagy inducer, induces potent preconditioning-like effects against AMI in rat hearts.70
During the compensatory hypertrophy stage of pressure overload, autophagic activity is suppressed; however, in failing hearts, autophagic activity is induced by accumulated protein aggregates or damaged organelles.31, 55, 71
Disrupting autophagy by cardiomyocyte-restricted inactivation of Atg5 deteriorated TAC- or β-adrenergic stimulation- induced cardiac dysfunction.55 However, a maladaptive role of load-induced autophagy was implicated in a study using mice with genetic inhibition of beclin1, a gene required for the early events of autophagy. In this study, heterozygous ablation of the beclin1 gene blunted the load-induced autophagy activation and pathological remodeling; whereas tg mice with cardiomyocyte-restricted overexpression of beclin1 had amplified autophagy and enhanced pathological remodeling in response to stress.71 These two sets of experiments had different levels of autophagy at work in the pathogenesis of pressure overload. In Atg5-deficient hearts, the basal level of constitutive autophagy is lost. In contrast, the belcin1+/− mice have only a 50% reduction in autophagic flux.72 Therefore, it is likely that the basal level of constitutive autophagy in the heart is a homeostatic mechanism and is protective;55 whereas, stress-related increases in autophagy can be maladaptive in the setting of load-induced hypertrophy.71-72 An alternative interpretation is that manipulating belcin1 is less specific than inhibiting Atg5 in terms of impact on autophagy in the pressure overload setting.
Furthermore, in vitro studies suggest that the absence of basal levels of autophagic activity in cardiomyocytes contributes to cellular hypertrophy. For instance, Nakai A et al. found that inhibition of autophagy by Atg7 knockdown induced morphological and biochemical features of cardiomyocyte hypertrophy.55 Moreover, compared to cells from WT mice, cardiomyocytes isolated from Atg5-deficient mice showed higher sensitivity to isoproterenol.55
The roles of autophagy in cardiac hypertrophic growth can also be inferred from studies on agents known to pharmacologically manipulate autophagy. Rapamycin prevents cardiac hypertrophy induced by TAC or thyroid hormone treatment, and regresses established cardiac hypertrophy. 73-74 It is very likely that the induction of autophagy contributes to the cardioprotection of rapamycin.
The alterations of the UPS and autophagy in cardiac dysfunction reviewed so far suggest that UPS proteolytic function often becomes inadequate or impaired by severe cellular stress and under this circumstance autophagy may be increasingly activated to help ameliorate the stress. Understanding how the upstream stimuli and signal mechanisms lead to UPS alterations and induce autophagy would shine light on the interplay between the two proteolytic pathways.
Perturbation of ER environment induces ER stress and leads to the accumulation of unfolded/misfolded proteins. Cardiomyocytes have a PQC system to deal with the challenge. Unfolded protein response (UPR) triggered by ER stress is an integrated part of intracellular PQC.75 Terminally misfolded ER proteins are retrogradely transported out of the ER and immediately subjected to ubiquitination and proteasomal degradation via ER-associated protein degradation (ERAD) in the cytosol.76 However, sustained ER stress causes accumulation of UPS reporter substrates, which indicates that sustained ER stress has an inhibitory effect on the UPS.77 When ERAD is overloaded by ER inhibitors or blocked by proteasome inhibitors, autophagy is mobilized to degrade terminally misfolded ER proteins via the ER-activated autophagy (ERAA) pathway.78-81
In a recent study, Hill’s group reported that accumulation and aggregation of ubiquitinated proteins upregulated the UPR regulator Bip and triggered activation of autophagy in a mouse model of load-induced heart failure.31 Additionally, accumulation of misfolded proteins, such as polyQ72 aggregates, in the ER stimulated LC3 conversion from LC3-I to LC3-II through phosphorylation of PERK (RNA-dependent protein kinase-like ER kinase) and eukaryotic initiation factor 2α (eIF2α).79 A related study with human tumor cells reveals that UPR protects against hypoxic tumor cells through inducing LC3 and Atg5 gene expression via the PARK/eIF2α signaling branch.82
Hypoxia or sI/R also leads to ER stress and activates UPR in cultured cardiomyocytes.83-84 Moreover, UPR is also activated in cardiomyocytes adjacent to, but not in those distal to, the infarct zone in mouse hearts subjected to AMI.85 Additionally, it has been suggested that δPKC activity mediates I/R-induced ER stress in the myocardium.86
Protein ubiquitination and unfolding during UPS-mediated proteolysis are ATP-dependent. ATP is also required to assemble the 26S proteasome complex.2 Therefore, it is likely that decreases in ATP levels contribute to diminished 26S proteasome activity and impaired UPS function in certain cardiac dysfunction.
When ATP levels in a cell decrease, various intracellular adaptive mechanisms are initiated, attempting to restore ATP levels. Among them, autophagy is one of the most important. In the study by Matsui et al.,63 glucose deprivation-induced autophagy in cardiomyocytes was mediated by activation of AMPK and inhibition of mTOR. Similarly, during myocardial ischemia, a rapid drop in ATP concentration also increases the AMP/ATP ratio and activates AMPK.87 Moreover, autophagosome formation in response to myocardial ischemia is decreased in tg mice with cardiac specific expression of a dominant negative AMPK, suggesting that endogenous AMPK is essential in mediating autophagy during myocardial ischemia.63 As a prominent energy sensor, AMPK is involved in activating energy generating pathways to maintain ATP levels and is required to activate autophagy.88
Bnip3 (Bcl2/E1B 19kDa interacting protein), a member of BH3-only proteins, is an important contributor to I/R injury and is induced in cultured ventricular myocytes and adult rat hearts subjected to hypoxia.89-90 Overexpression of Bnip3 caused significant induction of autophagy in HL-1 myocytes, while expression of a dominant-negative form of BNIP3 reduced hypoxia-induced autophagy.89 These studies suggest that Bnip3 likely contributes to the upregulation of autophagy during I/R and hypoxia.
Furthermore, the downstream players mediating Bnip3-induced autophagy have been identified. Bnip3 was shown to directly bind Rheb (Ras homolog enriched in brain) and inhibit its GTPase activity, thereby playing an important role in mTOR inactivation in response to hypoxia.91 This inactivation of mTOR might mediate Bnip3-induced autophagy. Alternatively, as an integral mitochondrial membrane protein, Bnip3 induces mitochondrial defects leading to the opening of the mitochondrial membrane permeability transition pore (mPTP).90 Given that mPTP opening induces autophagy in mammalian cells,92 it is possible that Bnip3 induces autophagy through damaging mitochondria. Reperfusion also triggers mPTP opening.93 Hence, mPTP opening may also contribute to autophagy activation in myocardial I/R.
Production of reactive oxygen species (ROS) is associated with a number of stimuli, inducing tumor necrosis factor (TNF) stimulation, ER stress, starvation, and inhibition of mitochondrial function.94 ROS are generated in the myocardium during I/R.95 Nitric oxide is released in heart failure and myocardial ischemia.96 It is known that the UPS is involved in removing mildly oxidized proteins. However, heavily oxidized proteins appear to form aggregates and then extensive covalent cross-linkage that make them highly resistant to UPS-mediated proteolysis.97 Moreover, some proteasome subunits may be suppressed or inactivated by oxidative stress. For instance, the 26S proteasome subunit S6 ATPase is very sensitive to oxidative inactivation.29 Oxidative modification of 20S proteasome subunits by the lipid peroxidation product 4-hydroxy-2-nonenal (HNE), which occurs in cardiac I/R, results in selective inactivation of 20S proteasome activity.30 Taken together, these findings explain the decreased peptidase activity of the proteasome in animal models of myocardial ischemia and I/R injury.27
On the other hand, ROS can activate autophagy in cardiomyocytes. For example, a study by Gottlieb’s group showed that lipopolysaccharide (LPS) induced autophagy through oxidative stress in both cardiomyocytes and mouse hearts. Nitric oxide synthase (NOS) inhibitor L-NMMA or potent antioxidant reagent N-acetyl-cysteine (NAC) reduced LPS-induced autophagy in cardiomyocytes.98 Furthermore, the ability of H2O2 and NO to directly activate autophagy was also tested in HL-1 cells in this study. It was observed that either H2O2 or SNP, a NO donor, was able to trigger autophagy in HL-1 cells.98 This suggests both reactive oxygen and nitrogen species induce autophagy directly.
Notably, a study by Scherz-Shouval and colleagues reveals a molecular underpinning of redox regulation of the autophagic process. They have shown ROS, specifically H2O2, are essential for starvation-induced autophagy because this oxidative signal leads to inactivation of Atg4. Atg4, a cysteine protease, is involved both in the initial lipidation processing of Atg8 to allow the conjugation of Atg8 to phosphatidylethanolamine (PE) and in the delipidation step by cleaving PE from Atg8-PE. This oxidative signal leads to inactivation of Atg4 at the site of autophagosome formation, thereby promoting lipidation of Atg8, an essential step in autophagosome formation.94, 99
Another consequence of I/R injury is calcium overload. It has been established that increased cytosolic calcium is a potent inducer of autophagy. Various Ca2+ mobilizing agents inhibit the activity of mTOR and induce massive formation of autophagosomes.100
Therefore, as summarized in Figure 1, multiple factors are capable of impairing UPS function and triggering autophagy activation. The latter participate in regulating multiple cellular processes.
As mentioned earlier, autophagy plays a dual role in the pathogenesis of cardiac diseases. Autophagy is activated to protect cells against cellular stress, while excessive autophagy promotes cell death. Therefore, the extent to which autophagy is activated may affect the pathophysiological outcome of autophagy. It has also been suggested that whether autophagy plays a protective or detrimental role is etiology-dependent. Thus, it is important to understand how autophagy contributes to cell survival or cell death in response to different stimuli in the heart.
The autophagic clearance of aggregate-prone proteins is well-established in neurodegenerative diseases. For instance, activation of autophagy by rapamycin has been shown to increase clearance of aggregate-prone protein both in vitro and in vivo.101-103 Intrigued by the common features shared by neurodegenerative diseases and cardiac proteinopathy, cardiac biologists also tested the autophagic degradation of aggregate-prone proteins in the heart. Pattison et al. observed that polyQ83 aggresomes in the heart are frequently engulfed by autophagosomes, associating autophagy with protein aggregation.58-59 More importantly, inhibition of autophagy has been shown to increase the prevalence of protein aggregates in cardiomyocytes in CryABR120G-based DRC models both in vivo and in vitro.60
In both neurodegenerative diseases and DRC, it is generally believed that the mutant proteins exert particular toxicity in the oligomer form and that microscopically visible protein aggregates may be less toxic. Increased autophagic activity not only directly removes aggregates but also clears aggregate precursors, shifting the equilibrium away from aggregate formation and thereby attenuating its toxicity in the cell.11, 104
Moreover, under the condition that proteasome is inhibited, autophagy is likely activated in response to ER stress and plays a pro-survival role.105 The activated autophagy serves to compensate for the reduced proteasome function so that proteins that fail to be degraded by the UPS can be cleared via autophagy. By doing so, autophagy ameliorates ER stress and suppresses proteasome inhibitor-induced cell death at an upstream site of the death pathway.105
During ischemia and nutrient limitation, autophagy is perhaps the main mechanism for adaptation and salvage of ATP. Autolysosomal degradation of membrane lipids and proteins generates free fatty acids and amino acids that can be used as fuel for ATP generation.106
The study by Matsui et al. showed that inhibition of autophagy by 3-MA further reduced ATP levels and increased cell death in the setting of glucose deprivation.63 Takemura et al. confirmed this result and further showed that rapamycin mitigated the reduction in ATP level and improved the survival of glucose-starved cells.107 These findings are consistent with the notion that autophagy promotes cell survival through preserving cellular ATP content during glucose-deprivation. Further supporting this notion is the finding that treatment with bafilomycin A, a lysosome inhibitor, did not affect cardiac function in normally fed mice; but instead, bafilomycin A severely depressed cardiac function and led to significant left ventricular dilatation with severely reduced myocardial amino acid and ATP contents in the starved mice.108
Since damaged mitochondria can release cytochrome c and other pro-apoptotic factors and generate ROS during I/R, their sequestration and degradation are cytoprotective. Autophagy is the only process known to degrade organelles in their entirety. Removal of ROS-producing mitochondria through autophagy would lower the oxidative stress experienced by a cell, facilitate mitochondrial biogenesis, and help with the mitochondrial quality control.
Several lines of evidence suggest that damaged mitochondria are selectively removed by autophagy. One example is that in rat hepatocytes, depolarized mitochondria are selectively sequestrated into autophagosomes.92 Similarly, the association between autophagosomes and damaged mitochondria is indicated by their colocalization in HL-1 cells where fragmentation of mitochondria is induced by overexpressing Bnip3.89 The damage in rat neonatal cardiomyocytes treated by LPS was limited by the rapid removal of damaged mitochondria triggered by the stimulation of autophagy.109 Likewise, during chronic hypoxia, Bnip3-mediated mitochondrial autophagy leads to reduction in mitochondrial DNA, mitochondrial mass, cell respiration and ROS production in response to hypoxia.110 Conversely, autophagy inhibition causes accumulation of damaged mitochondria. Ultrastructural analyses of Atg5-deficient hearts has revealed a disorganized sarcomere structure and the misalignment and aggregation of mitochondria.55 Moreover, suppression of autophagy by 3-MA in neonatal rat ventricular myocytes accumulates mitochondria with reduced membrane potential and mitochondria accumulation is partly reversed after 3-MA withdrawal.111 Taken together, upregulation of autophagy serves as a protective response likely by sequestering damaged mitochondria.
Previously, autophagic cell death was mainly a morphologic definition. Autophagic cell death is characterized by cell death with accumulation of vacuoles in physiological and pathological conditions.112 However, the question arises whether autophagic activity in dying cells is the cause of death or is actually an attempt to prevent it. Morphological definition can not prove a causative relationship between the autophagic process and cell death. The current definition of autophagic cell death includes an association with autophagosomes as well as dependence on autophagy proteins.113
Inhibition of apoptotic pathways can switch cellular response to stress from apoptosis to autophagic cell death. For instance, Bax/Bak-deficient mouse embryonic fibroblasts (MEFs), which fail to trigger apoptotic cell death, undergo non-apoptotic cell death when exposed to various cytotoxic agents, such as etoposide, staurosporine, and thapsigargin. This non-apoptotic form of cell death can be inhibited by suppressing autophagosome formation with autophagy inhibitors, such as 3-MA and wortmannin, or by silencing Atg5 and Atg6.114 Independently, other studies have also shown that with cytotoxic agents, autophagic cell death is induced with pan-caspase inhibitor, and this caspase-independent cell death can be inhibited by either chemical autophagy inhibitors or by the silencing of beclin1, Atg7 or Atg5.115-116
Excessive activation of autophagy itself can serve as a mechanism of cell death. Hyperactivation of autophagy, which destroys large proportions of the cytosol and organelles beyond a certain threshold, would lead to irreversible cellular atrophy with a consequential collapse of cell function.106 During reperfusion phase, low levels of ROS induce autophagy to degrade antioxidant catalase, thereby increasing the level of H2O2, which further induces autophagy. This positive feedback loop between autophagy and ROS accumulation resulting in a prolonged H2O2 signal is responsible for shifting the outcome from survival to death.117 Similarly, prolonged hypoxia induces autophagic cell death mediated through Bnip3.65 Therefore, autophagic cell death would explain the detrimental role of excessively activated autophagy under I/R or prolonged hypoxia, at least in some experimental conditions.
Recent studies suggest a cross-talk between two self-destructive processes: autophagy and apoptosis. As mentioned in 8.4.1, inhibition of apoptotic pathways can elicit autophagic cell death in cells subjected to cellular stress. On the other hand, autophagy constitutes a stress adaptation to avoid apoptosis in some scenarios. One example is that nutrient depletion triggers apoptosis under autophagy suppression.118 Notably, Atg5 serves as a point of cross-talk between autophagic and apoptotic pathways. While it is a key player in autophagy, Atg5 may also play a role in apoptosis. Under lethal stress, Atg5 is cleaved by calpains to generate a 24kDa truncated Atg5. The truncated Atg5 loses its autophagy-inducing activity and instead acquires a proapoptotic one. It translocates from cytosol to the mitochondria and associates with the anti-apoptotic molecule Bcl-XL, provoking apoptotic cell death.119 Therefore, Atg5 serves as a “double-agent” in both autophagy and apoptosis, which is regulated by the proteolysis of ATG5. Furthermore, the finding that Atg5 mediates interferon-γ-induced cell death by interacting with FADD (Fas-associated protein with death domain) also suggests that Atg5 has pro-apoptotic effects. This Atg5-mediated cell death is both autophagy- and caspase-dependent.120 Hence, Atg5 can trigger apoptosis through two different mechanisms.
There is growing evidence to suggest that lysosomal proteases are actively involved in apoptosis. For example, cathepsin D exerts its apoptosis-stimulating effect upstream of caspase-3-like activation in NRVMs.121 Cathepsin G also induces SHP2 (SH2 domain-containing tyrosine phosphatase 2) activation, which leads to focal adhesion kinase (FAK) tyrosine dephosphorylation and promotes apoptosis in NRVMs.122 A study using Hela cells shows that disruption of lysosomes by a lysosomotropic agent results in translocation of lysosomal proteases to the cytosol and induction of cell death through a caspase-dependent mechanism, involving Bid cleavage by papain-like cysteine proteases.123
Pharmacologically induced PSMI induces autophagy in multiple mammalian cell types.31, 105 Genetic impairment of the proteasome also induces autophagy in Drosophila.124 Importantly, autophagy induction in the setting of proteasome impairment is likely protective. Supportive evidence comes from the reports suggesting that the degenerative phenotypes caused by proteasome impairment are enhanced in an autophagy-deficient background, whereas the degeneration is significantly alleviated with autophagy induction in Drosophila.124 In cultured cells, it has been shown that the pretreatment with rapamycin attenuates proteasome inhibitor lactacystin-induced apoptosis and ubiquitinated protein aggregates;125 whereas, suppression of autophagy by 3-MA or Atg knockdown enhances proteasome inhibitor-induced cell death and accumulation of ubiquitinated proteins.31, 105
On the other hand, multiple lines of evidence suggest that reduced autophagy also results in enhancement of proteasome-dependent protein degradation. One example is that autophagy inhibition by 3-MA results in a robust increase in proteasome activities in NRVMs.31 Moreover, elevated polyubiquitinated protein levels and increased proteasome activities have been observed in Atg5-ablated hearts both in the basal state and in response to TAC.71 Taken together, the UPS and autophagy seem to help each other in PQC.
Ubiquitination used to be the hallmark of protein degradation by 26S proteasome. Modification of a protein with K48-linked polyubiquitin chains targets the substrate to the proteasomal degradation. However, recent findings reveal the involvement of K63-linked polyubiquitin chains in autophagy pathway.126 Adaptor molecules such as p62/SQSTM1, NBR1 simultaneously bind both ubiquitin and autophagy machinery. Identification of these adaptors provides a mechanistic link between the UPS and autophagy (Figure 2).
p62/SQSTM1(hereafter referred to as p62) appears to be an adaptor molecule linking ubiquitinated proteins to the autophagic machinery.127 The C-terminal portion of p62 binds polyubiquitinated substrates through its ubiquitin-associated (UBA) domain and directly binds to LC3 via the LC3 interacting region (LIR) motif.128 p62 can also polymerize via its N-terminal Phox/ Bem1p (PB1) domain and interact with proteasome via an N-terminal ubiquitin-like (UBL) domain. Interaction of p62 UBL domain with proteasome may be involved in shuttling substrates for proteasomal degradation.129
The increase in both transcript and protein levels of p62 in response to PSMI indicates that p62 may sense the proteolytic stress with PSMI and be involved in mediating the alternative degradation pathway to alleviate the proteolytic stress.130 It has been suggested that p62 provides a key link between autophagy and the UPS by facilitating autophagic degradation of ubiquitinated proteins.131
p62 also co-localizes with autophagosomes and can be degraded by autophagy.128 The degradation of p62 by autophagy is evidenced by the elevated level of p62 following autophagy inhibition and the decreased level in rapamycin treatment.1, 127, 132 Besides co-localizing with autophagosomes, p62 also localizes in a variety of ubiquitin-positive inclusion bodies in many neurodegenerative diseases, including Lewy bodies in Parkinson’s disease, neurofilbrillary tangles in tauopathies, huntingtin aggregates in Huntington’s disease, and aggregates seen in familial amyotrophic lateral sclerosis.133-136
Notably, p62 is required for the formation and autophagic degradation of polyubiquitin-positive inclusion bodies in response to stress.128 The aggregate-promoting property of p62 is suggested by the observation that depletion of p62 diminishes the formation of ubiquitin-positive inclusions upon puromycin treatment.128 Moreover, p62-null mice fail to form ubiquitin-positive inclusions in hepatocytes and neurons in response to misfolded protein stress in an autophagy-deficient background.137
Several lines of evidence also suggest the involvement of p62 in autophagic degradation of ubiquitinated proteins in response to misfolded protein stress. For instance, in neurons of p62-null mouse brains, there is a detectable increase in ubiquitin staining paralleled by accumulation of insoluble ubiquitinated proteins.138 Similarly, hyperphosphorylated tau, neurofibrillary tangles, and neurodegeneration are observed in p62-null mice.139 Reduction of p62 protein levels or interference with p62 function significantly increases cell death induced by the expression of mutant huntingtin.127 However, one study using mice with genetic inactivation of p62 and Atg7 found that loss of p62 markedly attenuated liver injury caused by autophagy deficiency; whereas p62 deficiency had little effect on neuronal degeneration. This study indicates that the pathologic process associated with p62 deficiency in autophagy-deficient tissues is cell-type specific.137 Taken together, p62 plays a critical role in the formation of ubiquitin-positive inclusions, degradation of ubiquitinated proteins under stress and is a perfect candidate for a messenger between the UPS and autophagy.
The relatively mild phenotypes as well as the cell-type specific phenotypes in p62 knockdown or knockout experiments raise the possibility of the existence of other p62-like proteins that might compensate for the lack of p62.133 NBR1 is the candidate for another autophagic receptor for ubiquitinated cargo.
NBR1 shares similar domain architecture with p62 and contains both a LIR motif interacting with ATG8 family proteins and a C-terminal UBA domain interacting with ubiquitin. NBR1 also has a PB1 domain and this region of NBR1 interacts directly with p62 in a canonical PB1-PB1 fashion. NBR1 can also homodimerise via the N-terminal two coiled-coil domains.140 The similar domain architecture between p62 and NBR1 raises the possibility for NBR1 to be another autophagic adaptor for ubiquitinated proteins.
On one hand, NBR1 co-localizes with p62 and Atg8-family proteins in the cell and can be degraded by autophagy in an LIR-dependent manner. This autophagic degradation is independent of p62.141 On the other hand, NBR1 co-localizes with p62 in the ubiquitin-positive aggregates in response to autophagy inhibition as well as in Mallory bodies in the liver of a patient with alcoholic steatohepatitis. More importantly, knockdown experiments suggest that both p62 and NBR1 are required for the formation of ubiquitin-positive aggregates upon autophagy inhibition and puromycin treatment. These findings lead to the conclusion that p62 and NBR1 cooperate in sequestration and degradation of ubiquitinated proteins.141
It is still unclear whether NBR1 can function as a substitute for p62 to perform some of the functions of p62. The recent finding that aggregation of NBR1 is dramatically increased under the condition that both p62 and Atg7 are deficient raises the possibility that the increased NBR1 can partially compensate for the loss of p62 to form aggregates.141 The involvement of autophagy receptor p62 and NBR1 in the formation of aggregates suggests a possible link between aggregate formation and autophagy.
Aggresome formation may help detoxify misfolded proteins through packaging more toxic and active oligomers of misfolded proteins into large inclusion bodies so that fewer vital cellular constituents would come into contact with them. The transportation of aggresome precursors via microtubules and their motor proteins to the vicinity of microtubule organizing center (MTOC) is critical to aggresome formation and subsequent removal by autophagy.142
HDAC6 is a microtubule- and dynein- associated deacetylase. It binds ubiquitin via the C-terminal BUZ domain and interacts directly with dynein motors, thereby serving as a linker coupling protein aggregates to the retrograde microtubule motor. HDAC6 has been found to localize to the experimentally induced ubiquitin-positive aggresomes.143-144 Moreover, HDAC6 deficiency results in defects in aggresome formation in response to experimentally induced misfolded protein stress.144 HDAC6 may therefore regulate the retrograde transport of misfolded proteins to MTOC to form aggresomes. More strikingly, HDAC6 is essential for retrograde transport of autophagosomes and lysosomes to MTOC where the aggresomes are mainly located, thereby facilitating the autophagic degradation of aggresome.143 The two functions of HDAC6 in promoting aggresome formation and autophagic degradation could be coupled and represent one integrated mechanism. From this point of view, efficient autophagic degradation of aggregated proteins may be promoted by HDAC6.11 This notion is supported by the finding that HDAC6 is required for autophagic clearance of aggregated mutant huntingtin in cultured cells and the mutant androgen receptor in a fly model of Kennedy’s disease.124, 143 Similarly, autophagy is induced by, and compensates for, UPS impairment in a fruitfly model of spinobulbar muscular atrophy in an HDAC6-depedent manner.124
p62, NBR1, and HDAC6 may work in synergy to pack the misfolded proteins together and facilitate their interactions with the phagophore, thus providing the specificity required for the degradation.11 It has also been suggested that K63-linked polyubiquitin chains interacts with p62, NBR1, and HDAC6, thereby providing a signal for selective autophagic degradation.145-146
CHIP can mediate α-synuclein degradation by both degradation pathways: the tetratricopeptide repeat domain is critical for proteasomal degradation, whereas the U-box domain is sufficient to direct α-synuclein toward the lysosomal degradation pathway. This study suggests that CHIP acts as a molecular switch between proteasomal and lysosomal degradation pathways.147
The Hsc/Hsp70 co-chaperones of the BAG protein family are modulators of PQC. BAG1 acts as a physical link between the Hsp70 and the proteasome. Moreover, BAG1 can accept substrates from Hsc/Hsp70 and direct them to proteasomal degradation through the direct interaction with CHIP.148 In contrast, BAG3-HSPB8 complex facilitates the degradation of Htt43Q through the activation of autophagy.149 In cell models of ageing, it has also been shown that BAG1 and BAG3 regulate proteasomal and autophagic pathways, respectively, for the degradation of polyubiquitinated proteins. More importantly, due to the increased ratio of BAG3/BAG1 in aged cells compared with young cells, the aged cells depend more intensively on autophagy to degrade polyubiquitinated proteins.150 Moreover, BAG3 interacts with p62 to promote p62-dependent autophagic degradation.150
Thus, as highlighted in this section and illustrated in Figure 2, a number of factors and signaling events orchestrate the UPS and autophagy to perform PQC in the heart. Each of the two proteolytic pathways has its own responsibilities at the baseline but they attempt to assist each other during cardiac dysfunction.
PQC and protein degradation are extremely important for maintaining normal cardiac function. PQC inadequacy occurs in the progression of a wide range of cardiac diseases, as reflected by elevated levels of ubiquitinated proteins and sustained UPR in failing human hearts and in many animal models of cardiac dysfunction. The accumulation of misfolded proteins resulting from inadequate PQC in turn impairs PQC in the heart, forming a vicious cycle. Very few studies have shed light on the feasibility and effect of enhancing UPS function. This is probably because enhancing UPS function may accelerate the degradation of critical short-lived intracellular regulators, thereby being detrimental to the cell.151
In contrast, autophagy seems to be more amenable to manipulation. In most scenarios of cardiac dysfunction, autophagy protects cells by: (1) removing misfolded proteins and alleviating ER stress; (2) maintaining the ATP level in the cell; and (3) removing damaged organelles such as damaged mitochondria. Growing evidence links autophagy with the pathogenesis of multiple forms of heart disease, suggesting that autophagy may be therapeutically effective in treating the heart disease.71, 152 However, the role of autophagy is dose- and context-dependent, which poses special challenges. Excessive autophagy would cause cell death through excessive self-digestion of essential organelles and proteins. Therefore, in the clinical setting, physiologically beneficial autophagy should be preserved, whereas excessive autophagy should be avoided.
A growing body of evidence indicates that pharmacological upregulation of autophagy is indeed beneficial in animal models of a large subset of cardiac dysfunction, such as myocardial ischemia, cardiac hypertrophy, cardiac ageing. Given that drugs mitigating autophagy are already in clinical use,104 substantial advances in attaining more effective treatment for human heart disease by manipulating autophagy are on the horizon.
Additionally, caloric restriction and exercise are known to activate autophagy.153-154 Moreover, the up-regulated autophagy mediated by caloric restriction and exercise shows cardioprotection. For instance, removal of ROS-producing mitochondria through autophagy is associated with less oxidative stress and results in high-quality mitochondria which exhibit efficient ATP production and better resistance to stress.153, 155 Therefore, lifestyle change may show great promise in alleviating cardiac disease.
Dr. X. Wang is an established investigator of the American Heart Association (AHA). Research in his laboratory is supported in part by grants R01HL072166, R01HL085629, and R01HL068936 from the NIH and grant 0740025N from the AHA (to X. W.) and by the MD/PhD Program of the University of South Dakota. Dr. Q. Zheng is a recipient of AHA Predoctoral Fellowship (Reference # 0815571G). We thank Ms. Emily McDowell for her assistance in preparing this manuscript.
Disclosure The authors declared no conflict of interest.