**3. Mitochondrial Dynamism in the Failing Heart**

The development of real-time methods involving the use of chemical dyes and fluorescent proteins or the use of transmission electron microscopy (TEM) has allowed scientists to see mitochondria and study their real morphology, which is highly dynamic and impacts cell function, physiology, and disease [68]. It is interesting to note how several lines of evidence show key roles for mitochondrial fitness in a broad range of cardiac and metabolic disorders, including the development and progression of HF and hypertrophy [69]. In cardiomyocytes, mitochondria are clustered among the myofibrils and the plasma membrane (subsarcolemmal mitochondria or SSM), across the sarcomere myofibrils (intermyofibrillar mitochondria or IMF), and tightly adjacent to the nucleus [70]. Together, these mitochondria are associated with the sarcoplasmic reticulum (SR), a subcellular compartment that participates in cardiac contraction by controlling the processes of Ca2+ storage, and release and reuptake in skeletal muscles [71,72].

Overall, the mitochondrial morphology is controlled by known phenomena of balanced fusion and fission that interconnect with each other and to events involving an increased mitochondrial mass (biogenesis) and "in excess" organelle removal (mitophagy) [73]. Whether changes in these pathways are a cause or consequence, or simply accompany the pathology of heart diseases, remains unclear [74].

#### *3.1. Ultrastructural Abnormalities*

Important ultrastructural abnormalities were detected from an analysis of mitochondria in HF tissues, and these abnormalities, which include the loss of mitochondrial granules, swelling with disorganization of the cristae, vacuolar degeneration, mitochondrial fragmentation, deep structural lesions (disruption of the OMM and IMM and loss of the electrodense matrix), and a strong reduction in the mitochondrial volume [52,75–78], become increasingly prominent from the initial to end stages of the disease. All of these abnormalities can often contribute to cytochrome c release, which ultimately leads to apoptosis, and were found with both IFM and nuclear mitochondria but not SSM mitochondria [79–82]. Moreover, the surrounding environment undergoes corresponding changes, including disorganization of sarcomeres and T-tubules.

For correct maintenance of the morphology and structure of the cristae, some proteins located in the IMM that result in being impaired in HF are important; these include mitofilin and cardiolipin [83,84]. Mitofilin is a structural protein, whereas cardiolipin is a phospholipid that is needed to support energy production and regulation of the mitochondrial structure, biogenesis, and dynamics [56–58]. Recent studies demonstrated that the downregulation of mitofilin in living cells induces the formation of mitochondria with disorganized inner membranes, loss of mitochondrial function, increased ROS production, and apoptosis [85]. Notably, low levels of mitofilin were also found in tissues of human hearts with HF [86]. However, the downregulation of cardiolipin in HF induces an increase in the production of ROS and mitochondrial dysfunction in cardiomyocytes until death [87]. The maintenance of mitochondrial function through the prevention of mitofilin and cardiolipin is thus important to limit the development of HF.

#### *3.2. Biogenesis*

As previously mentioned, several studies using animal models of HF showed that the expression of some key regulators of energy metabolism, such as PGC-1α and β, is strongly reduced [88–90]. Normally, PGC-1α is a protein with a key role in mitochondrial biogenesis and improves OXPHOS and FAO [91]. Its overexpression, which usually occurs in response to external stimuli (i.e., physical exercise and exposure to cold temperatures), triggers a significant increase in the volume of the whole mitochondrial network. Moreover, PGC-1α is able to bind several transcription factors, including nuclear respiratory factor (NRF) 1 and 2, to activate transcription factor A (TFAM), which is responsible for initiating the transcription of nuclear-encoded proteins with structural and mtDNA replication properties [92]. In response to PGC-1α deregulation, abnormal mitochondrial biogenesis rates were observed in animals with HF and in patients with different etiologies [86]. As opined by Karamanlidis and colleagues, the main mitochondrial dysfunctions (including their reduced turnover) in failing hearts are due to defects in the mtDNA amount and integrity instead of alterations in the gene expression profile [93]. Indeed, in a significant number of failing human hearts, a reduction in mtDNA-encoded proteins derived from a loss of mtDNA was also observed [93,94].

Nevertheless, several kinds of cardiovascular diseases including HF in advanced stages were observed accompanying the course of mitochondrial-based pathologies caused by mtDNA mutations [95,96]. Among those, MELAS (mitochondrial encephalomyopathy, lactic acidosis, and stroke-like episodes) syndrome was observed in concomitance to 3243 A>G mtDNA substitution [97,98].

In addition to its role in mitochondrial biogenesis and its relationship with energy metabolism, PGC-1α can also improve contractility and endothelial function in models of HF [99–102]. Recent studies performed using PGC-1α-KO mice showed an accelerated development of HF, which was linked to downregulation of OXPHOS genes [90] and deficiencies in the function and reserve of cardiac energy [103,104].

#### *3.3. Fusion and Fission Machinery*

The process of mitochondrial biogenesis is tightly interconnected with two additional pathways named fusion and fission, and these pathways allow the origin of an interconnected and fragmented mitochondrial network, respectively [105]. It was shown that proteins contributing to mitochondrial dynamics are greatly expressed in the mammalian myocardium, and their ablation is deadly [72,106–109]. This finding suggests that in cardiomyocytes, mitochondrial dynamism is essential for noncanonical functions governing mitochondrial quality control, Ca2+ signaling, cardiac development, and cell death [110].

Although an imbalance in mitochondrial dynamics usually accompanies several cardiovascular diseases, such as ischemia-reperfusion (I/R) and diabetic cardiomyopathy [111], very few studies have addressed these pathways in HF [112]. Mitochondrial dynamics represent continuous events of fission and fusion occurring at all developmental stages of a cell and contribute to cell division, cell death, mtDNA, and nutrient exchange to sustain metabolism in the heart properly [79–81,113]. Proteins involved in mitochondrial fusion and fission were highly preserved during evolution [114]; they utilize GTP energy to guide conformational changes and exhibit distinct mitochondrial sublocalization [105]. Mitochondrial fusion is a complex mechanism involving three steps: tethering, OMM fusion, and IMM fusion. Tethering occurs through one or two homologous proteins that belong to the superfamily of mitochondrial transmembrane GTPases. Proteins that are located on the OMM and involved in fusion are mitofusin 1 (Mfn1) and 2 (Mfn2) [115,116]. These mediators can form three different molecular complexes: (i) Mfn1 homotypic oligomers, (ii) Mfn1-Mfn2 heterotypic oligomers, and (iii) Mfn2 homotypic oligomers [117,118]. Fusion of the OMM is normally followed immediately by IMM fusion directed by another dynamin superfamily GTPase, optic atrophy 1 (Opa1), which leads to the final formation of filamentous organelles [119]. This protein is located on the IMM and is redistributed to the IMS. In

addition to controlling IMM fusion, Opa1, according to certain protein modifications, can also exert specific effects on the cristae structure [120].

Previous studies showed the essential role of mitofusins during adulthood, which involve conferring protection against long-term cardiac dysfunction, and in early embryonic cardiac development [121]. Unexpectedly, their absence induces mitochondrial fragmentation, and this effect is accompanied by respiratory dysfunction and progressive HF [108]. Additionally, a reduction in Opa1 is representative of mitochondrial variations affected by HF in humans. All these protein changes reflect a population of smaller mitochondria with a truly fragmented network controlled by posttranscriptional events and always associated with apoptosis and consequent loss of functional cardiomyocytes [122,123]. Among all established models of HF, Opa1 is the only protein that always shows a decrease in expression during HF [123]. Indeed, perhaps in an effort to compensate for this loss in OPA1, mitofusins were found to increase following HF.

A report study conducted by Menezes T. and colleagues found that Mfn1 is a substrate target for PKCβII (a Ca2+-dependent protein kinase). PKCβII accumulates in the OMM during HF and produces extensive mitochondrial fragmentation following the phosphorylation of Mfn1 at serine 86, which results in a loss of GTPase activity, as detected in both animal and human samples.

Contrary to what is known, Mfn1-dependent fusion in cardiomyocytes fails in response to dysregulation of Ca2+ cycling and inefficient cardiac contractility [124]. Being all features that characterize HF, further studies should be performed in in vivo samples to confirm the possibility that targeting Mfn1 expression can be fully considered as therapeutic opportunity.

Mitochondrial fission in mammalian cells is led by the cytoplasmic protein dynaminrelated protein-1 (Drp-1), which forms complexes at fission sites on the OMM [125], and fission protein-1 (Fis-1), which encircles the outer mitochondria and promotes the assembly of protein complexes on the OMM [126,127]. Indeed, Fis-1, mitochondrial fission factor (Mff), and mitochondrial dynamics proteins 49 and 51 (MiD49 and MiD51) act as receptors for the recruitment of Drp1 to the mitochondrial surface [128]. Drp1 performs multiple functions, often at the intersection between mitochondrial fission and mitophagy [129]. Drp1 cytoplasm-mitochondria shuttling is crucial for its roles and this is finely regulated by many post-translational modifications including phosphorylation, SUMOylation, palmitoylation, and ubiquitination [129].

In HF models, Fis-1 appears to play a minor role because its levels usually remain unchanged; otherwise, the overall amount of Drp1 tends to be higher in human samples of HF while these data are not confirmed in animal models [123]. In light of this finding, on one hand, Drp1 might also play minor roles in HF, including its crucial direct role in cell death when it is upregulated; on the other hand, post-translational modifications (PTM) rather than changes in the overall amount of the protein may occur in HF. For example, it was observed that Sentrin/SUMO-specific protease 5 (SENP5) is able to deSUMOylate and repress Drp1-dependent mitochondrial fission. SENP5 is upregulated in HF and induces a phenotype of apoptotic cardiomyocytes [130]. However, the most well described PTM for Drp1 is phosphorylation, especially at Serine 616 (S616), which ensures Drp1 translocation at mitochondria and fragmentation. In a mouse model subjected to pressure overload, mitophagy is transiently upregulated in the hypertrophic heart in the first week, in a manner dependent from the phosphorylation of Drp1 at S616. This pathway is not further maintained in later stages of hypertrophy where this repression determines irreversible dysfunctions and HF [131].

#### *3.4. Mitophagy*

Both under baseline conditions and in response to stress, cells activate a highly regulated mechanism, called mitophagy, to digest senescent and damaged mitochondria [74]. The organelles are engulfed by autophagosomes and are subsequently delivered to lysosomes for degradation. This mechanism is crucial for the maintenance of cellular home-

ostasis; defects in mitophagy trigger and amplify mitochondrial dysfunction (due to the accumulation of aberrant mitochondria), and this effect is accompanied by the development of cardiomyopathies and ultimately HF with severe contractile dysfunction [132]. This process could be regulated by mitochondrial PTEN-induced kinase 1 (PINK1) and the cytosolic ubiquitin ligase Parkin. Normally, when mitochondria are healthy, PINK1 is imported into the mitochondrial matrix through the translocase of the outer membrane (TOM) complex. In contrast, when a damaged mitochondrion exhibits a loss of mitochondrial membrane potential, PINK1 accumulates on the OMM [133], which results in the recruitment of Parkin from the cytosol to the mitochondrial membrane [134]. However, to initiate Parkin-mediated mitophagy, two events induced by PINK1 are needed. The first event is the phosphorylation of Mfn2, which causes attraction on the mitochondrial surface of Parkin, whereas the second event consists of the phosphorylation of Parkin at Ser65 (ubiquitin-like domain), which increases its E3 ligase activity [135,136]. These steps introduce the ubiquitination of mitochondrial proteins to promote phagosome recruitment and the successive degradation of mitochondrial proteins by the lysosome. Studies suggest that for degradation by the autophagosome, mitochondrial protein ubiquitination via the Lys63 linkage plays a signaling role in the recognition of damaged organelles. In this mechanism, some adaptor proteins, such as NBR1 and p62, bind to ubiquitinated mitochondrial proteins and interact directly with LC3 on the autophagosome, creating a link between mitochondria and autophagosomes [137,138]. In addition, the ubiquitination of mitofusins via Lys48 linkage induces their degradation by the ubiquitin/proteasome system [139] to prevent the fusion of both healthy and damaged mitochondria, which aids the final aim of mitophagy.

From these findings, it is clear that mitophagy is a critical mitochondrial quality control mechanism in myocytes [140–142]. Defects in mitophagy associated with the proteins PINK1 and Parkin have negative consequences for cardiomyocytes. Although cardiac function is normal in young Parkin-KO mice, as they age, they accumulate abnormal mitochondria, develop irreversible HF, and show increased cell death following myocardial infarction [141]. Similarly, the absence of PINK1 leads to cardiac mitochondrial dysfunction with a great burst of ROS and irreversible cardiac hypertrophy [140]. The loss of PINK1 in mice also increases susceptibility to pressure overload-mediated HF and I/R injury [140,143]. Similar findings were also documented in humans affected by advanced HF, and these can be characterized by inefficient mitophagy as a consequence of the reduction in the overall PINK1 levels. Although the knowledge of mitophagy as a cause or consequence of HF remains controversial, a recent report suggested that a shift between two isoforms of AMP-activated protein kinase (AMPK) alpha (from 2 to 1) occurs in failing hearts and is responsible for its pathogenesis [144]. Restoring AMPKα2 and thus increasing the phosphorylation of PINK1 at serine 495 increases mitophagy to efficient levels to prevent the progression of HF [144].

Increasing evidence in the last decade highlighted a role also for Parkin- and PINK1 independent pathways in the activation of mitophagy under conditions of stress, as those occurring under I/R [132,145]. These molecular routes would include additional players such as cardiolipin, Bcl2/adenovirus E1B 19 kDa protein-interacting protein 3 (Bnip3), a receptor for LC3-II binding at the OMM and FUNDC1, a protein that under hypoxia becomes phosphorylated at S17 by Unc-51 similar to autophagy activating kinase (Ulk1), and dephosphorylated by phosphoglycerate mutase family member 5 (PGAM5) at S13, triggering mitophagy [145,146]. Moreover, and of great interest, an in vivo model of cardiac ischemia showed the activation of a phosphorylations cascade involving S555 of Ulk1, S179 of Rab9, and S616 of Drp1. These actively participate in the formation of a multiprotein complex in which Ulk1 is phosphorylated by AMPK, indirectly inducing the phosphorylation of Drp1 (and thus mitochondrial fission and mitophagy) via assembling with the Rab9-Rip1-Drp1 axis [147].

The fine balance among all these pathways (biogenesis, fusion, fission, and mitophagy) is crucial either to prevent or overcome injuries following stressful conditions and can thus

be considered a promising and feasible therapeutic target [1,25,94]. To understand this fine balance better, increased mitophagy might cause excessive mitochondrial clearance, which would leave the myocytes with too few mitochondria to produce sufficient ATP. During acute cardiac injury, such as MI or I/R, a limited increase in mitophagy could be beneficial to clear damaged mitochondria, but in chronic cardiac diseases, such as HF, sustained upregulation of mitophagy might be harmful [132].

These studies clearly show that the dysregulation of mitophagy has the potential to lead to the accumulation of abnormal mitochondria, contractile dysfunction, and ultimately the progressive loss of myocytes.
