A Time to Reap, a Time to Sow: Mitophagy and Biogenesis in Cardiac Pathophysiology

Abstract

Balancing mitophagy and mitochondrial biogenesis is essential for maintaining a healthy population of mitochondria and cellular homeostasis. Coordinated interplay between these two forces that govern mitochondrial turnover plays an important role as an adaptive response against various cellular stresses that can compromise cell survival. Failure to maintain the critical balance between mitophagy and mitochondrial biogenesis or homeostatic turnover of mitochondria results in a population of dysfunctional mitochondria that contribute to various disease processes. In this review we outline the mechanics and relationships between mitophagy and mitochondrial biogenesis, and discuss the implications of a disrupted balance between these two forces, with an emphasis on cardiac physiology.

Introduction

Mitochondria function as cellular power plants essential for meeting the energetic demands of eukaryotic cells. Their role extends to regulating fuel utilization, calcium stores, intracellular signaling and cell death. Because of the broad range of cellular functions they are involved in, mitochondria inherently occupy an important position as mediators of cellular homeostasis. Consequently, this crucial position associates the dysfunction of mitochondria to the development of various human diseases. Notably, studies to dissect the etiology of Parkinson Disease (PD) were among the first to highlight the physiological consequence of having poor mitochondrial quality control. Genetic models strongly implicate mitochondrial dysfunction as a common feature in development of this neurodegenerative disease that leads to the loss of dopaminergic neurons (reviewed in [1–5]). In support of this is the fact that aging increases the risk of developing PD, which correlates with higher incidence of mitochondrial DNA mutations in dopaminergic neurons [6]. Moreover, agents that induce mitochondrial toxicity have been shown to lead to PD-like symptoms in animal models [7].

The major chronic diseases we face today such as neurodegenerative diseases, cancer, aging, diabetes, and heart failure are accompanied by mitochondrial dysfunction, and in fact, many elements of these chronic diseases may be directly attributed to mitochondrial pathology [8]. Mitochondrial disorders may be inherited either through maternal transmission of an abnormal mitochondrial genome or through autosomal transmission of mutations in the nuclear-encoded mitochondrial genes. However, far more commonly, mitochondrial dysfunction is a consequence of derangements in the ordinarily robust systems that orchestrate and maintain the health and function of these vital organelles.

Mitochondrial quality control collectively describes the cellular systems used to maintain a population of optimally-functioning mitochondria. Mitochondria possess an internal protein quality control system to refold or eliminate misfolded proteins, comprising chaperones (Hps10, Hsp60 and others) and proteases (Lon and other AAA proteases). Import of nuclear-encoded proteins must be coordinated with expression of mitochondrial subunits for proper assembly of oxidative phosphorylation (OXPHOS) complexes. Homeostatic control of this is mediated through the mitochondrial unfolded protein response (UPRmt), which is activated by an imbalance of nuclear vs. mitochondrial OXPHOS subunits [9]. Mitochondrial turnover is another integral aspect of quality control in which dysfunctional mitochondria are selectively eliminated through autophagy (mitophagy) and replaced through expansion of preexisting mitochondria (biogenesis). Impaired mitochondrial quality control results in accumulation of damaged mitochondria that may generate more reactive oxygen species (ROS), produce ATP less efficiently, have a lower threshold for cytochrome c release (apoptosis) or mitochondrial permeability transition pore (MPTP) opening (necrosis), or may release mitochondrial components (mtHSP60, oxidized mitochondrial DNA) into cytosol where its recognition by receptors for damage-associated molecular patterns (DAMP) activates inflammation. In this way, impaired mitochondrial quality control gives rise to a myriad of disease states. Mitochondrial quality control is critically dependent on autophagy; factors that impair autophagy, such as advanced age or the metabolic syndrome (MetS), will impact mitochondrial quality control and accelerate the development of chronic disease phenotypes. In this review, we focus on the mechanics of mitophagy and mitochondrial biogenesis, and discuss the interplay between these two forces. We then discuss the pathophysiological consequences with an emphasis on the heart.

1. Mechanics of Mitophagy and Mitochondrial Biogenesis

1.1 Mechanics of mitophagy

Autophagy is a lysosome-dependent cellular degradation system in eukaryotic cells that allows for the bulk recycling of unwanted cytoplasmic aggregate proteins or dysfunctional organelles [10]. Along with the ubiquitin proteasome system (UPS), autophagy is important for maintaining proteostasis in the heart [11]. Mitophagy is the selective targeting and removal of mitochondria through autophagy. While some authors refer to the general process as mitochondrial autophagy and use the term mitophagy to mean Parkin-dependent elimination of mitochondria, in this review we will use ‘mitophagy’ to indicate autophagic removal of mitochondria, and where appropriate, will specify Parkin-dependent mitophagy. Mitophagy plays a critical role in protecting the heart during ischemia/reperfusion injury [12–14]. Depolarization of mitochondria is a prerequisite for Parkin-dependent mitophagy, but mitophagy mediated by Bnip3 and NIX may be triggered through other pathways including reactive oxygen species (ROS) [15], which promote dimerization of Bnip3 (and potentially NIX) on the mitochondrial outer membrane [16]. Nutrient stress (fasting) activates AMPK and general autophagy, which is associated with production of ROS from mitochondrial complex I [17]; however, fasting-induced mitophagy is impaired in cyclophilin D-deficient mice [18], which have hyperpolarized mitochondria. Thus there are hints that mitophagy initiated by nutrient stress may be initiated by mitochondrial depolarization and Parkin translocation, but a role for ROS and Bnip3 is not excluded.

Parkin-dependent (macro)mitophagy has been commonly studied using chemical uncouplers of mitochondria such as carbonyl cyanide 4-(trifluoromethoxy) phenylhydrazone (FCCP) or carbonyl cyanide m-chlorophenyl hydrazone (CCCP). Cellular stresses such as ischemia also trigger mitochondrial depolarization [13], resulting in stabilization of the serine/threonine kinase phosphatase and tensin homolog (PTEN)-induced kinase 1 (PINK1) on the outer mitochondrial membrane (OMM) and recruitment of the E3 ubiquitin ligase Parkin, key factors for mitophagy [19–22]. PINK1 and Parkin function as critical partners to mediate the clearance of dysfunctional mitochondria [23, 24]. Another Parkin-dependent mechanism for degrading mitochondrial components is through mitochondria-derived vesicles (MDV), which transit to multivesicular bodies and eventually the lysosome, or to the peroxisome [25].

Mitochondrial dynamics (fusion and fission) also play a critical role in mitochondrial quality control, and the process is closely linked to mitophagy, where fission is favored and fusion is suppressed, enabling engulfment by autophagosomes. Fission of reticulate mitochondria into smaller fragments is essential for mitophagy to occur [26, 27]. Key to this process is the dynamin-related protein 1 (Drp1), a GTPase in the dynamin super family of proteins, which is recruited to the mitochondria and facilitates the process of mitochondrial fragmentation [28]. Fission 1 (Fis1) is another key player in mitochondrial dynamics that interacts with Drp1 to facilitate mitochondria fragmentation [29]. Mfn1 and 2, which promote OMM fusion, are ubiquitinated and targeted for elimination by the UPS. Optic atrophy protein 1 (OPA1), important for fusion of the inner mitochondrial membrane, is degraded during mitophagy by the inner membrane zinc metalloprotease OMA1, which has overlapping activity with matrix AAA proteases [30–32].

PINK1 is constitutively made and continuously degraded by the mitochondria-specific proteases presenilin-associated rhomboid-like protein (PARL) and mitochondrial processing peptidase (MPP). Loss of membrane potential across the inner mitochondrial membrane inactivates PARL and MPP through an uncharacterized mechanism and permits the accumulation of PINK1 on the OMM. The kinase domain of PINK1 faces the cytosol and phosphorylates OMM proteins facilitating the recruitment of the E3-ubiquitin ligase Parkin [33–35]. PINK1 has been reported to phosphorylate a number of targets including Parkin itself [36, 37], mitofusin 2 (Mfn2) [15], and mitochondrial rho 1 (MIRO) [38], a component of the microtubule-associated motor complex that anchors kinesin to mitochondria. Mfn2, which functions in mitochondrial fusion events and links endoplasmic reticulum to mitochondria, functions as a Parkin receptor after phosphorylation by PINK1, thereby recruiting Parkin to the mitochondria, where it ubiquitinates a number of OMM targets. Voltage-dependent anion channel 1 (VDAC1) has been shown to be a Parkin target essential for mitophagy [19], although this finding has been contested [39]. Ubiquitination and proteasomal degradation of MIRO, Mfn2, and Mfn1 serve to immobilize the mitochondrion and prevent it from rejoining the mitochondrial network through fusion [15, 38, 40–42]. Ubiquitination of OMM proteins facilitates recruitment of autophagy adapter proteins such as neighbor of BRCA1 (NBR1) or sequestosome-1 (p62/SQSTM1). These bifunctional adaptor proteins have an ubiquitin binding domain (UBA) and microtubule-associated protein 1 light chain 3 (LC3) interacting region (LIR) to bring the developing autophagosomal membrane in proximity to the tagged mitochondrion in a zipper-like process [43, 44]. SMAD-specific E3 ubiquitin ligase 1 (SMURF1) has also been linked to Parkin-dependent mitophagy [45]. Surprisingly, its ability to facilitate mitophagy has been found to be independent of its E3 ubiquitin ligase function. Another Parkin-interacting autophagy promoter, activating molecule in Beclin 1-regulated autophagy (Ambra1) dissociates from mitochondrial Bcl-2 to bind Beclin1 to initiate autophagy [46, 47]. Ambra1 interacts with Parkin to promote mitophagy, but is not a substrate of Parkin [48].

Mitophagy that is independent of PINK1/Parkin/ubiquitin can be initiated through atypical members of the Bcl-2 homology domain 3 (BH3) family members such as BCL2/adenovirus E1B 19 kDa protein-interacting protein 3 (BNIP3) and BCL2/adenovirus E1B 19 kDa protein-interacting protein 3-like protein (BNIP3L aka NIX). These proteins insert into the OMM and facilitate engulfment by the autophagosome through a LIR domain that can interact with LC3 isoforms including gamma-aminobutyric acid receptor-associated protein (GABARAP) and GABARAP-like 1 (GABARAPL1) [49, 50]. One study demonstrated in cardiomyocytes that Bnip3 recruited Parkin and Drp1 to the mitochondria to promote fission and mitophagy [51]. In hypoxic conditions mitophagy has been reported to be mediated by the OMM protein FUN14 domain containing 1 (FUNDC1) which contains a LIR [52].

The process of mitophagy is rather complex and requires coordination of UPS-mediated degradation of outer membrane proteins with autophagic engulfment of the remainder of the organelle. The autophagosome, with its cargo, fuses with a lysosome, forming the acidic autophagolysosome. In cells in which the pH-sensitive fluorescent protein Keima is targeted to the mitochondria, one can monitor the delivery of mitochondria to the autophagolysosome by monitoring the pH-dependent shift in fluorescence of mitoKeima [53]. The cargo is degraded by lysosomal hydrolases, liberating amino acids and fatty acids which are exported to the cytosol via lysosomal permeases. A model representing our current knowledge of the process of mitophagy is shown in Figure 1. While mitophagy is responsible for bulk degradation of the organelle, turnover of individual components may proceed at asynchronous rates through redistribution of components via fusion events, selective degradation of proteins via mitochondrial proteases, and proteasomal elimination of some outer membrane proteins. Even in the case of Parkin-dependent mitophagy, some outer membrane proteins are “rescued” through transfer to the ER [54]. Proteomic studies using heavy isotope labeling [55] revealed that proteins of the IMM turn over with rates that are similar to mitochondrial turnover based on historic radiolabeling studies [56, 57], suggesting that IMM proteins (primarily OXPHOS constituents) may be primarily cleared via mitophagy. This also corresponds to studies which showed that matrix and OMM were readily redistributed across the mitochondrial network when fusion and fission were intact; however, IMM constituents redistributed much more slowly [27]. In contrast to the IMM proteins, OM proteins turned over faster in many cases [55], possibly because outer membrane proteins can be degraded by the UPS, by translocation to other sites (ER or peroxisomes), by the MDV pathway, or by mitophagy. Matrix and IMM proteins have fewer routes of degradation: while mitophagy predominates, matrix and IMM proteins can be found in MDVs, and are substrates for Lon and other AAA proteases in the matrix and intermembrane space. The rate of mitochondrial protein turnover in mice is much slower in heart (mean half-life 17d) than in liver (4d). A related study of mitochondrial protein synthesis in rat comparing heart and liver using isotopic labeling and mass spectrometry also showed a 6-fold slower rate of turnover in the heart. Turnover also varied according to subcellular location: protein turnover was ~15% faster in subsarcolemmal mitochondria than in interfibrillar mitochondria [58]. Mitochondrial protein import is inhomogeneous: the fluorescent MitoTimer protein revealed “hot spots” for synthesis and import of this mitochondrially-targeted protein [59].

Figure 1. Parkin-dependent Mitochondrial Autophagy (Mitophagy).

1. Cellular stress, such as ischemia/reperfusion, triggers fragmentation of the mitochondria mediated by Drp1, segregating low-membrane potential mitochondria from the rest of the network. Ischemia/reperfusion injury also leads to the collapse of mitochondrial membrane potential which deactivates PARL and MPP, allowing for PINK1 stabilization on the OMM.
2. Parkin is recruited to the OMM where it binds Mfn2 and ubiquitinates multiple OMM proteins, marking them for proteasomal degradation and targeted recognition of the ubiquitin-decorated mitochondrion.
3. Autophagy adapter proteins such as p62 are then recruited to the mitochondria which in turn bind the ubiquitinated mitochondrion to the phagophore through interaction with LC3 or homologs.
4. Once the autophagosome has fully engulfed the mitochondrion, it fuses with a lysosome to form the autophagolysosome where final degradation of bulk contents is completed.
5. Shaded area indicates atypical players that participate in recognition and targeting of mitochondria for autophagic clearance. These include Nix and Bnip3 which bind LC3 or homologs including GABARAPL1.

The importance of mitophagy in the heart was highlighted in our previous work demonstrating the requirement for mitophagy in cardioprotection conferred by ischemic preconditioning [13] and acute statin administration [12]. These results suggest that mitophagy is part of the final common pathway for various cardioprotective interventions, and indeed, may be the ultimate effector. Beyond cardioprotection, Parkin-dependent mitophagy plays a role in ischemia tolerance [60], myocardial aging [61], and pathologic remodeling in response to pressure overload [62]. The cardiac effects of Parkin deficiency are phenocopied by PINK1 and Atg5 deletions [63, 64]. Similarly, deletion of Mfn2, considered an essential mitochondrial docking partner for Parkin, leads to mitochondrial dysfunction and heart failure [65]. These findings highlight the importance of autophagy and mitophagy in cardiac function.

1.2 Overview of mitochondrial biogenesis

Mitochondrial biogenesis, which acts in concert with mitophagy to maintain homeostasis in cells, depends on coordination of nuclear and mitochondrial-encoded gene expression. The nuclear-encoded genes are primarily controlled by the transcription cofactor peroxisome proliferator-activated receptor gamma coactivator 1-alpha (PGC-1α) [66]. First identified as a binding partner of peroxisome proliferator-activated receptor γ (PPARγ) that increased its transcriptional activity during thermogenesis, PGC-1α is a member of the nuclear receptor superfamily of proteins that are responsible for assembling functional macrocomplexes of transcriptional machinery at specific DNA sequences [67, 68]. PGC-1α controls the expression of nuclear respiratory factors 1 and 2 (NRF-1 and NRF-2), which in turn control the expression of mitochondrial transcription factor A (Tfam) [69]. Tfam plays a key role in mitochondrial biogenesis by regulating the expression of mitochondrial genes (tRNAs, rRNAs and 13 subunits of the respiratory chain) from the heavy and light strands of the mitochondrial genome; it is also essential for replication of mitochondrial DNA (mtDNA) [70].

Several studies have demonstrated the physiological importance of PGC-1α in mitochondrial biogenesis in response to cold and exercise. These external stimuli increase expression of PGC-1α leading to increased expression of mitochondrial enzymes such as ATP synthase (β-subunit), COX (cytochrome c oxidase) subunits (COX II and COX IV) and δ-aminolevulinate synthase (δ-ALAS) [71, 72]. Increased mitochondria content allows for more efficient thermogenesis in response to cold and enhanced capacity to generate ATP to sustain exercise bouts. An isoform of PGC-1α, PGC-1β has also been identified, and although overexpression of this protein increased mitochondrial biogenesis and basal oxygen consumption much like PGC-1α, it was not induced by cold or exercise, suggesting alternate pathways for induction of mitochondrial biogenesis [73].

Aside from influencing the transcription of key players in the respiratory chain, PGC-1α also interacts with and upregulates the expression of other transcription factors such as peroxisome proliferator-activated receptors (PPARs) [74], hormone receptors for estrogen and thyroxine, as well as estrogen-related receptors (ERRs) α and γ [75]. ERRs are a particularly interesting set of nuclear receptors known as orphan receptors due to a lack of an associated ligand [76]. These proteins are involved in PGC-1α-dependent regulation of mitochondrial biogenesis. For example, Schreiber et al have demonstrated that over-expression of PGC-1α results in upregulation of 151 genes that encode mitochondrial proteins involved in many metabolic functions of mitochondria such as fatty acid β-oxidation (FAO), tricarboxylic acid cycle and oxidative phosphorylation, as well as mitochondrial ribosomal machinery and mitochondrial membrane transport proteins. This effect is inhibited by siRNA targeted to ERRα, and conversely mimicked by overexpression of ERRα [77]. Endonuclease G is regulated by ERRα and PGC-1α, and its deletion results in cardiac hypertrophy and mitochondrial dysfunction [78]. Additional factors implicated in mitochondrial biogenesis include Lon protease and Hsp78 [79].

The expression of PGC-1α is controlled primarily through signaling cascades. Calcineurin A-dependent (CnA) and Ca2+/calmodulin-dependent protein kinase IV (CaMKIV) regulation of PGC-1α has been well characterized. CnA interacts with and activates myocyte enhancer factors 2C and 2D (MEF2C and MEF2D), which regulate the transcription of PGC-1α directly [80, 81]. Furthermore, activation of PGC-1α results in upregulation of MEF2C and 2D, creating a feed forward loop that allows PGC-1α to increase its own expression [82]. CaMKIV activates PGC-1α by phosphorylating the transcription factor cAMP response element (CRE)-binding protein (CREB). Phosphorylated CREB binds to promoter elements of the PGC-1α gene to drive its expression [80]. Other players controlling PGC-1α expression include p38 mitogen-activated protein kinase (p38 MAPK) and AMP-activated protein kinase (AMPK). p38 MAPK activity is increased following exercise, and this leads to the activation of MEF2 and activating transcription factor 2 (ATF2), both of which drive the expression of PGC-1α [83, 84]. The activation of AMPK in response to glucose depletion results in direct phosphorylation of PGC-1α on threonine-177 and serine-538, which is crucial for activation of PGC-1PGC-1α-dependent transcription from the PGC-1α promoter [85].

Post-translational modifications such as phosphorylation and acetylation regulate PGC-1α activity. Kinases that have been implicated in controlling of PGC-1α activity include: AMPK and Akt during caloric restriction and p38 MAPK after endurance exercise [85–87]. Likewise, p38 MAPK increases the activity of PGC-1α by directly phosphorylating threonine-262, serine-265, and threonine-268, which stabilizes the protein and disrupts the interaction between PGC-1α and its inhibitor p160MBP [86, 88]. Conversely, insulin inhibits the activity of PGC-1α through Akt, directly through phosphorylation of the serine-570 residue on PGC-1α, and indirectly through phosphorylation of the Clk2 kinase which in turn phosphorylates the C-terminal serine and threonine-rich regions of PGC-1α, thereby decreasing its co-transcriptional activity [89, 90]. In an even more indirect manner, glycogen synthase kinase 3b (GSK3β) has also been shown to inhibit PGC-1α activity in response to acute oxidative stress by increasing its proteasomal degradation and inhibiting the activity of Sirt1, an NAD-dependent deacetylase thought to activate PGC-1α [91]. This deacetylation event is crucial for the activation of PGC-1α, as the protein is very heavily acetylated by the acetyltransferase GCN5, inhibiting its activity and sequestering it in nuclear foci distant from promoter regions of its target genes [92]. Sirt1 activity is dependent upon the coenzyme nicotinamide adenine dinucleotide (NAD⁺), and it is therefore highly sensitive to the changes in the energetic state of the cell. Increased NAD⁺/NADH ratio–which may occur in response to fasting, exercise or redox stress—activates Sirt1, leading to PGC-1α deacetylation [93, 94]. The result of this deacetylation is an increase in transcription of PGC-1α targets, leading to mitochondrial biogenesis [95– 97]. Interestingly, AMPK may once again play a role in activating PGC-1α, not only by directly phosphorylating the protein, but also indirectly by increasing NAD⁺ levels in the cell by fatty acid oxidation, thereby increasing the activity of Sirt1 [97]. Other posttranslational modifications such as ubiquitination or methylation also play a role in modulating the activity of PGC-1α in response to energy demands and oxidative stress, states that require mitochondrial biogenesis [98].

Other proteins that play a major part in mitochondrial biogenesis are vascular endothelial growth factor (VEGF) and hypoxia-inducible factor 1 (HIF-1). VEGF is a key regulator of angiogenesis which involves substantial cell proliferation and production and remodeling of extracellular matrix, processes which utilize large amounts of ATP. A study by Wright et al demonstrated that VEGF stimulates mitochondrial biogenesis by coordinated upregulation of OMM protein Tom70 and activation of PGC-1α [99]. Interestingly, PGC-1α can also activate VEGF, by coactivating ERR-α on conserved binding sites found in the promoter and in a cluster within the first intron of the VEGF gene, driving angiogenesis in response to oxygen deprivation independently of HIF-1 [100]. HIF-1 is a master regulator of the adaptive response to hypoxia, and as such is intimately linked to inducing mitochondrial biogenesis. Several studies have linked PGC-1α and PGC-1β with HIF-1 activity. O’Hagan et al reported that mitochondrial biogenesis driven by the expression of PGC-1α results in increased oxygen consumption and decreased intracellular oxygen tension, permitting stabilization of HIF-1 and activation of a gene expression program to increase oxygen delivery to the tissue [101]. Zhang et al demonstrated that HIF-1 represses mitochondrial biogenesis by controlling the transcription of PGC-1β through downregulation of c-MYC transcription factor [102]. Figure 2 highlights the central role of PGC-1α in regulating mitochondrial biogenesis.

Figure 2. PGC-1α Regulation of Mitochondrial Biogenesis.

PGC-1α is considered a master regulator of mitochondrial biogenesis. Transcriptional control of PGC-1α expression is closely linked to environmental cues of fuel availability, fuel type, and cellular energy requirements. PGC-1α transcription is governed by multiple transcription factors (trans) including PPAR/RXR, MEF2, C/EBP, FoxO, CREB/CRTC, ERRγ, and MyoD/E2A. These factors in turn are activated by specific signal pathways including free fatty acids, AMPK, calcineurin, p38 MAPK, CaMK IV, and PKA, and suppressed by other signals including GCN5, AKT and SHP. In addition to transcriptional control, PGC-1α activity is regulated by acetylation and phosphorylation by the factors illustrated here. Ultimately, PGC-1α increases mitochondrial biogenesis and the capacity to perform OXPHOS, in particular, fatty acid oxidation.

1.3 Interplay between mitophagy and mitochondrial biogenesis

Mitophagy and mitochondrial biogenesis are opposing forces that govern the rate of mitochondrial turnover. This dynamic tension allows for a readily adjustable population of mitochondria to match cellular demands. Here we discuss several players that participate in this regulatory cross-talk.

Sirt1, an NAD-dependent lysine deacetylase, stimulates autophagy directly by deacetylating various autophagy proteins including Atg7, Atg5, and Atg8 (LC3) [103]. Sirt1 also deacetylates and activates PGC-1α [104], thus positively regulating both mitophagy and biogenesis. Activation of PGC-1α is a key event in initiating mitochondrial biogenesis, but no less important are the repressors of the process. Parkin-interacting substrate (PARIS or ZNF746) has recently been identified as a direct transcriptional repressor of PGC-1α [105]. The accumulation of PARIS in the nucleus leads to direct inactivation of PGC-1α transcription and inhibition of expression of PGC-1α-dependent genes. Aside from its role in facilitating mitophagy, Parkin was shown to directly target PARIS for degradation through the UPS. Events leading to the upregulation of mitophagy also increase Parkin activity which then degrades PARIS, relieving repression of mitochondrial biogenesis. This relationship establishes an intricate system that links mitophagy with mitochondrial biogenesis. A model illustrating this mechanism linking mitophagy with mitochondrial biogenesis is shown in Figure 3. As the relationship between Parkin and PARIS has thus far been identified only in neuronal models, an important and exciting question remains as to whether this system exists in the heart. Moreover, what are its implications in the setting of ischemia/reperfusion injury?

Figure 3. The Parkin-PARIS Axis Coordinates Mitophagy with Mitochondrial Biogenesis.

Basal state cellular homeostasis is characterized by balanced mitophagy and mitochondrial biogenesis (mitochondrial turnover). This maintains a network of healthy mitochondria. Mitophagy is linked to a transcriptional program for mitochondrial biogenesis. One pathway in this tightly coordinated process involves Parkin and PARIS. Triggers of mitophagy increase Parkin expression and activity, leading to proteasomal degradation of PARIS. Diminished PARIS levels relieve the transcriptional repression of PGC-1α, priming mitochondrial biogenesis.

Other repressors of mitochondrial biogenesis operate in less direct manner. Nuclear co-repressor 1 (Ncor1) acts as a repressor of PPARγ, PPARδ and ERR activity by interacting with histone deacetylases such as HDAC3 and SIRT1 to maintain tonic repression of MEF2, PPARδ, and ERR, thereby suppressing their participation in transcription programs involving mitochondrial oxidative metabolism [106]. mTOR is a serine/threonine kinase involved in numerous cell functions and can directly activate PGC-1α to control mitochondrial biogenesis. During fasting, mTOR is inhibited and autophagy/mitophagy is active. However, as lysosomal degradation releases amino acids, mTOR is reactivated, suppressing autophagy and supporting lysosomal and mitochondrial biogenesis [107]. Eukaryotic translation initiation factor 4E (eIF4E)-binding proteins (4E-BP) prevent translation of targets including nuclear encoded mitochondrial protein mRNAs including TFAM (transcription factor A, mitochondrial) and subunits of complex V and complex I. This inhibition is lifted by the action of mTORC1 which inhibits 4E-BP proteins from binding their targets [108, 109].

Autophagy and mitochondrial biogenesis are linked in both directions: PGC-1α induces the expression of transcription factor EB (TFEB) [110], the master regulator of lysosome biogenesis and the autophagy pathway [111]. TFEB and PGC-1α regulate one another’s expression, and the nutrient sensing regulator GCN5L1 suppresses both TFEB and PGC-1α [112]. Thus like the mythical ouroboros (serpent devouring its own tail), autophagy and mitochondrial biogenesis, mitophagy and lysosomal biogenesis, are elements of a system whose stability derives from its dynamic regulation.

2. Mitochondrial Turnover in the Context of the Organ and the Organism

2.1 Circadian Rhythms and Metabolic Status

The circadian rhythm is an important regulator of cellular and whole-body homeostasis. Disruption of this oscillatory system can cause and aggravate several health issues. The connection between autophagy and circadian rhythm was first demonstrated in the 1970’s using electron microscopic analysis of different rat tissues to show that the number of autophagic vacuoles changed over the course of the day [113, 114]. More recently an important work established a clear connection between circadian rhythm and the induction of autophagy [115]. Ma and collaborators showed that autophagy was upregulated during the dark cycle, preceded by increased mRNA expression of autophagy initiators including Ulk1, and Bnip3, but not factors involved in later phases of autophagy such as LC3B, Atg4, and Atg7. These events were regulated by the transcription factor CCAAT-enhancer-binding protein β (C/EBPβ). C/EBPβ is a liver clock protein that is tissue autonomous and regulates whole body bioenergetics. The induction of autophagy usually promotes mitophagy as well by upregulating mitochondria-targeting machinery such as the autophagy adapter p62/SQSTM1 and the E3 ubiquitin-ligase Parkin [116, 117]. Mitochondrial biogenesis is also directly regulated by the circadian clock, ensuring coordination of mitophagy with biogenesis [118–120]. Conditions that disrupt the circadian clock will also impact autophagy and mitochondrial biogenesis, with consequences for cell and organ function; conversely, disruption of autophagy may affect the clock [121].

The circadian rhythm is closely linked to eating behaviors (and diet types), another important regulator of mitophagy in animals. Mitophagy can be regulated by several different energy sensing stimuli: the availability of AAs in a cell regulates mTOR activity; the glucose level in blood regulates release of insulin and glucagon which have opposing effects on autophagy; and conditions such as metabolic syndrome and caloric restriction regulate autophagy and mitophagy. Insulin suppresses macroautophagy by activating the phosphatidyl-inositol triphosphate (PIP₃) cascade leading to phosphorylation and activation of Akt and subsequently mTOR, which inhibits autophagy [122]. Usually high insulin is followed by an increase in AAs in the cytosol, which also activates mTOR, reinforcing the same pathway. Insulin also promotes mitochondrial fusion, whereas hypoglycemia and insulin resistance promote fission [123–125]; both processes are essential to mitochondrial quality control. The nutritional overload of a high fat diet suppresses autophagy and by extension, mitophagy [126–128]. More recently, impaired mitochondrial turnover in mice fed a high-fat diet was demonstrated using Timer protein targeted to mitochondria (MitoTimer). In this study, the authors electroporated a MitoTimer construct under a constitutive promoter into the skeletal muscle of mice which were then maintained on chow or a high-fat diet. They observed a shift in the MitoTimer ratio to red, indicating slower mitochondrial turnover [129]. Although they attributed the red shift to increased oxidative stress, previous work showed that MitoTimer maturation (color shift) was not sensitive to oxidants [130]. MitoTimer is a mitochondria-targeted fluorescent molecular clock that we developed as a tool to monitor mitochondrial turnover [59].

2.2 Cardiac Development, Cardioprotection and Cardiac Pathology

The development of stem cells into differentiated adult cells is a tightly regulated process. Autophagy is involved in the maturation of several different types of cells. Cardiomyocyte development is regulated by the fibroblast growth factor (FGF) signaling axis [131] and in skeletal muscle FoxO is the responsible factor [132]. In both cases, autophagy and mitophagy are tightly linked to differentiation and tissue plasticity [131, 133]. FGF suppresses autophagy and thereby prevents differentiation of cardiac progenitors [134], while FoxO signaling induces autophagy as part of the regeneration and growth of the muscle tissue. As discussed above, mitophagy and biogenesis are tightly linked. These examples highlight the importance of autophagy to tissue remodeling and repair beyond degradation.

Genetic deletions of mitofusin-2 and PINK1 illustrate the importance of mitophagy to heart development and homeostasis. Genetic deletion of mitofusin-2 is embryonic lethal [135] and is essential in the heart not only for mitochondrial dynamics [136], but also ER-mitochondrial calcium signaling [137], mitophagy [15], and autophagosome-lysosome fusion [138]. Absence of PINK1 has profound consequences for postnatal heart development [63] and exacerbates ischemia/reperfusion injury [139].

Among the most potent interventions to protect the heart from ischemia and reperfusion injury are ischemic pre- and post-conditioning [140, 141]. Pre and post conditioning require autophagy to deliver the protection [142–144], although this is controversial in the brain [145, 146]. Mitophagy is part of the autophagy response that is specifically required for protection [13]. Other interventions that protect the heart against ischemic injury, including chloramphenicol [147], caloric restriction [148] simvastatin [12], and SAHA [149] all act through the autophagy/mitophagy pathway, thus establishing autophagy/mitophagy as a hub for cardiac protection. There are few direct inducers of autophagy; rapamycin is an mTOR inhibitor widely used as a drug to induce autophagy. Rapamycin administration also decreased ischemia/reperfusion injury [149, 150] while upregulating autophagy. In chronic models of heart failure, rapamycin also helps to ameliorate the phenotype [151]. Taken together, these facts support the beneficial effects of inducing autophagy.

In conditions where autophagy/mitophagy is impaired the opposite is true: there is increased cardiac dysfunction and exacerbation of ischemia/reperfusion injury in the setting of Parkin and PINK1 deletion [60, 139], deletion of macrophage migration inhibitor factor, an inducer of homeostatic autophagy [152], and obesity [153]. Obesity [154] and diabetes [155] disrupt normal energy metabolism, changing basal activation of mTOR and other nutrient signaling cascades that regulate autophagy. High fat diets are known to increase ischemia/reperfusion damage in hearts [156, 157] and there is now a significant amount of work linking the high fat diet to impaired autophagy [154] and accumulation of dysfunctional mitochondria [158, 159], highlighting the importance of mitophagy to cardiac ischemia tolerance. The regulation of cardiac mitophagy and physiological importance of this process is illustrated in Figure 4.

Figure 4. Cardiac Mitophagy Regulation and Significance.

Mitophagy is essential during cardiomyocyte differentiation and for homeostatic mitochondrial turnover to maintain a healthy population of mitochondria. During cardiac stress such as ischemia/reperfusion, mitophagy functions to eliminate damaged mitochondria and reduce injury. Mitophagy is also critical for ischemic preconditioning. Circadian rhythm regulates basal levels of cardiac mitophagy. Nutrient overload, type 2 diabetes, obesity, and advanced age may compromise cardiac autophagy and mitophagy, disrupting this adaptive physiological response to stress.

Autophagy may not always be a good thing in the heart. Infection of juvenile mice with Coxsackievirus B virus can exacerbate stress-induced myocardial injury in adulthood [160]. Autophagy is needed for the spread and reproduction of the virus [161], and virus-induced autophagy triggers premature differentiation of cardiac-resident progenitor cells, contributing to heart failure later in life [160]. Excessive autophagy has been implicated in doxorubicin-mediated cardiac injury [162, 163], although other studies have reported a beneficial role for autophagy [164–167]. Interestingly, deletion of Nrf2, a transcriptional regulator of autophagy and mitochondrial biogenesis, exacerbated doxorubicin toxicity, but this was reversed by overexpression of Atg5 [168]. They did not examine whether restoring autophagy resulted in mitochondrial biogenesis independent of Nrf2. It seems likely that unless mitophagy is balanced by biogenesis, problems will ensue.

Mitochondrial biogenesis in the heart is tightly responsive to oxygen tension. This is manifest at the transition from the fetal hypoxic state to the postnatal aerobic environment, when HIF signaling is lost, thereby favoring mitochondrial fusion and mitochondrial biogenesis [169]. During cardiac hypertrophy in response to aortic banding, mitochondrial dysfunction and decreased biogenesis were noted [170]. Downregulation of PGC-1α is observed in animal models of heart failure, but attempts to restore mitochondrial biogenesis by overexpression of PGC-1α did not improve cardiac function despite a modest increase in mitochondrial content [171]. In fact, inducible overexpression of PGC-1α in the heart resulted in abnormal mitochondrial morphology and cardiomyopathy which was reversible upon normalization of PGC-1α levels [172]. These studies did not examine mitochondrial autophagy. However, in a porcine study of renovascular hypertension, hypertrophy was accompanied by upregulation of mTOR, increased abundance of markers of autophagy and mitophagy, and decreased mitochondrial protein content, all of which were reversed by the angiotensin II receptor blocker valsartan [173]. The authors concluded that hypertension increased autophagic clearance of mitochondria and valsartan suppressed autophagy and restored mitochondrial biogenesis. However, because they did not measure autophagic flux or p62 (a surrogate marker of flux), the data lend themselves to the opposite interpretation: that hypertension impaired autophagic flux, thereby limiting mitochondrial biogenesis possibly through the Parkin/Paris/PGC-1α network. Evidence in support of the latter interpretation is the finding that mTOR was strongly upregulated in the hypertensive hearts, which would suppress autophagy. The observed increase in Beclin 1 and LC3-II could reflect increased autophagy or impaired flux. Beclin 1 is known to interfere with autophagic flux [174, 175]. The perinuclear accumulation of Parkin-decorated mitochondria is also indicative of impaired lysosomal clearance of autophagosomes. This also illustrates the importance of determining autophagic flux and mitochondrial turnover before reaching a conclusion.

3. Prospects and Challenges for the Future

Mitochondrial quality control depends upon mitophagy, biogenesis, fusion, and fission, as well as selective protein quality control via AAA proteases and chaperones. To date, most studies have explored mitochondrial dynamics (fusion/fission) and mitophagy (Parkin-dependent and Parkin-independent mitochondrial autophagy). Mitophagy is increasingly recognized to play a significant role in the heart, yet in order to maintain homeostasis, biogenesis must keep pace. Therefore, approaches to monitoring mitochondrial turnover (the integrated outcome of these four processes) are needed. Recent advances include the analysis of the half-lives of mitochondrial proteins using mass spectrometry analysis and deuterium labeling [55], mito-Keima, which can report on mitochondria delivered to the lysosome [53], and MitoTimer, a fluorescent protein that can be used to monitor mitochondrial turnover [59, 129, 130]. What lies ahead is the application of these tools to study physiologic and pathologic processes in the heart.

A major challenge to overcome is imaging autophagy (or mitophagy) in humans. Relatively few studies have examined autophagy in the human heart, largely because of the challenges of accessing tissue, and none have examined mitophagy, although animal studies indicate that mitochondria are a frequent target of autophagy. There is a significant need to develop better tools for in vivo imaging of autophagy and mitophagy.

Still lacking is a thorough understanding of mitochondrial biogenesis: are all mitochondria equally capable of expanding and undergoing fission to give rise to daughter mitochondria enriched for newly-imported proteins and highly functional OXPHOS assemblies, or is there a subset of mitochondria that are specialized for mitochondrial regeneration? Studies of MitoTimer suggest that protein import preferentially takes place in mitochondria closest to the nucleus [59, 130]. This could be a trivial consequence of mRNA proximity, and import of MitoTimer may not necessarily reflect sites of biogenesis. It is exciting, however, to speculate that the subpopulation of mitochondria most actively engaged in importing newly-synthesized protein is indeed unique. Future studies may shed light on this.

Mitochondrial protein import is essential for biogenesis, but is also implicated in the regulation of mitophagy because PINK1 must transit through the intermembrane space in order to be degraded by PARL. Few studies have considered whether defective protein import is the red flag that signifies a mitochondrion due for autophagic elimination. It has not been demonstrated whether pre-amyloid oligomers might disrupt mitochondrial protein import, yet this might explain the deterioration of mitochondrial function (78) and impaired biogenesis (79) that often accompanies Alzheimer’s disease and potentially other protein folding disorders. We can expect that in the coming years investigators will integrate information exchange between mitochondria and cytosol/nucleus, for which the TOM/TIM complex and VDAC serve as important carriers.

A key to mitochondrial homeostasis is the ability to remove and replace components throughout the network: not only proteins, but also lipids and mtDNA copies. In the heart, where mitochondrial fusion and fission events seem to occur with a frequency approximately equal to the rate of turnover of the entire organelle, intra-mitochondrial degradation and protein import generalized across the network may play a larger role than regionally restricted biogenesis followed by redistribution via fusion events. The intriguing observation that proteins in subsarcolemmal mitochondria turn over faster than in interfibrillar mitochondria suggests that different turnover mechanisms may operate within the same cell. The thought-provoking discovery that individual mitochondrial proteins have widely differing half-lives raises questions about the mechanisms governing this process; regulation of these different mechanisms for degrading mitochondrial proteins may be quite complex, and their contribution to disease phenotypes will be equally so. The importance of mitochondrial protein import is emerging: recently it was reported that redirecting a mutant form of alanine:glyoxalate aminotransferase from mitochondria to peroxisomes corrects primary hyperoxaluria 1 (PH1), a lethal metabolic disease [176]. Enzymes that may traffic either to mitochondria or peroxisomes can have radically different consequences depending on their location; the potential significance of this process for heart disease is unknown at present.

Yet another emerging area is the role of miRNAs in regulating autophagy and mitochondrial biogenesis. miRNA-149 inhibits poly(ADP-ribose) polymerase-2 (PARP-2), thereby allowing an increase in cellular NAD⁺ and activation of sirtuin-1, leading to mitochondrial biogenesis [177]. miR-27a and miR-27b impair mitochondrial biogenesis [178]. miRNAs also regulate the Nrf2 pathway [179] and autophagy [180–184]. Elucidating the contribution of miRNAs to the dynamic regulation of mitophagy and biogenesis will require a systems biology approach.

Many open questions remain to be resolved, but technical advances continue to make new discoveries possible. The advent of novel gene therapy approaches, cell permeable proteins, and small molecule therapeutics targeting mitochondrial quality control mechanisms hold promise for treating a variety of diseases from the perspective of the underlying mitochondrial dysfunction. It is not too farfetched to envision mitochondrial medicine becoming a medical specialty as much as surgery, cardiology, or genetics.

Mitochondrial Mysteries.

Roberta A. Gottlieb

• We know so much yet understand so little

• About mitochondrial ox-phos and fusion and fission

• Mitochondrial autophagy and biogenesis

• MitoTimer and lenses have given us celluvision

• Though heart cells live years it’s quite different within

• Mitochondrial life is counted in weeks

• Outer and inner membrane proteins vary yet more

• In their lifespans revealed by mass spectrum peaks.

• Protein import must match what’s inside

• Lest proteins unfold and fall prey to Lon

• The peptides escape to the cytosol

• To trigger transcription of chaperones.

• Try we must to describe and define

• The complex nature of the proteome

• As mitochondria expand and divide

• Fragment and fall into autophagosomes

• Yet for all we know and all we learn

• The mysteries grow and questions expand

• Like Mandelbrot sets of fractal images

• We see the work of divinity’s hand

HIGHLIGHTS.

• The role of mitophagy and biogenesis in the heart are discussed

• Mitochondrial quality control depends on balanced mitophagy and biogenesis

• Factors regulating these processes are summarized

• Mitochondrial turnover is discussed in the context of heart disease

• Major open questions are enumerated