Cardiac Mitochondria: A Surprise about Size

Abstract

Mitochondria dynamics are regulated by cycles of fusion and fission. Increasingly, evidence suggests that alterations in mitochondrial size and appearance may be linked to cardiac dysfunction and disease. Now, a new report suggests that those proteins involved in regulating the fusion of cardiac mitochondria may have functions that extend beyond mitochondrial dynamics and that the size of cardiac mitochondria may not be as informative as previously believed.

The endosymbiotic theory posits that mitochondria can be traced back to an unlikely event some 1.5 billion years ago when a proteobacteria was engulfed by a primitive single cell organism. Although ancient in origin, it wasn’t until the late 19^th century, in large part due to the observations of the German physician Carl Benda that the term mitochondria was given to this intracellular organelle. The name itself derives from two ancient Greek words, mitos meaning thread and chondrion meaning granule. Benda peering through his microscope instantly realized the seeming dual nature of this organelle. On the one hand, they appeared small and round (i.e. chondrion), while also capable of arranging themselves in apparent interconnected threads (i.e. mitos). Since these very early observations, it has become increasingly clear that mitochondria are in fact highly dynamic organelles that, as Benda intuited, can constantly interconvert their morphology existing as either fragmented short mitochondrial rods or spheres or an elongated interconnected network [1, 2]. This morphological wizardry is achieved through the reciprocal processes of mitochondrial fusion and fission [3, 4]. Considerable evidence suggests that mitochondrial dynamics are not simply a morphological alteration but also intimately connected to mitochondrial function [1, 3]. Indeed, alterations in mitochondrial morphology have been implicated in a variety of different biological process including embryonic development, metabolism, autophagy, apoptosis and cell death [1–3, 5, 6]. However, while informative, many of these studies have simply correlated a specific mitochondrial morphology with a subsequent alteration in physiological function. As such, a causative role for the observed changes in mitochondrial shape and organization has not always been unambiguously established. Moreover, while mitochondrial dynamics are clearly robust in many cell types commonly used in the laboratory (often immortalized, transformed cell lines), the role of this process in more specialized cell types (e.g. cardiomyocytes) remains largely unexplored.

The machinery that allows for dynamic changes in mitochondrial morphology involves an expanding group of mitochondrial fusion and fission proteins. For fission, a process whereupon a single mitochondria pinches down on its inner and outer membrane to form two (not necessarily equal) daughter mitochondria, the dynamin-related peptide 1 (Drp1) appears to play a major role (Figure 1). This large GTPase is well conserved from lower species such as yeast and Drosophila to mammals. Of course, Drp1 does not work alone and a host of other proteins such as the mitochondrial fission protein 1 (Fis1) are necessary to recruit and regulate the overall fission process. The reciprocal process of mitochondrial fusion, wherein two mitochondria fuse to generate a single larger organelle, requires the joining of both the outer and inner mitochondrial membranes. After two mitochondria are reversibly tethered, the outer membrane fusion event is governed by outer membrane mitofusins (Mfn1 and Mfn2 in vertebrates and MARF in Drosophila), while the inner membrane fusion is mediated by Opa1. When fission predominates, the mitochondrial network is termed fragmented, while increased fusion leads to an interconnected, tubular network of mitochondria.

Figure 1.

The relationship between mitochondrial size and function is less clear than once believed. Manipulation of the genes involved in fission (Drp1) or fusion (MARF and Opa1) can produce similar alterations in mitochondrial size but very different effects on the function of Drosophila hearts. Similarly, knockdown of MARF and Opa1 produce a similar inhibition of mitochondrial fusion and result in a cardiomyopathy, however these functional defects are rescued by non-overlapping genetic strategies.

Cardiomyocytes require a large and constant energy supply in order to accomplish the work the circulatory system entails. Not surprisingly, mitochondria occupy 30% of cardiac cell volume and are essential for proper heart function [7, 8]. In contrast to cells routinely studies in the laboratory, adult cardiomyocytes possess what appears to be a limited range of mitochondrial dynamics [2, 9, 10]. This is perhaps due to the presence of densely packed and highly organized mitochondria that are often found between contractile filaments leading to a lattice-like cytoskeletal structure that may impede mitochondrial movement and hence fusion/fission cycles [11]. In contrast to these observations, numerous studies have implicated changes in mitochondrial morphology as impacting cardiac development, heart failure, ischemia–reperfusion injury and the threshold for cardiomyocyte apoptosis [2, 12–15]. Mitochondrial morphological changes have also been described in mice with cardiac specific Mfn2 ablation [16], reduced expression of Opa1 [17], as well as in cardiac-specific Mfn1-null mice [18]. Other studies have suggested that mitochondrial morphology is interconnected with cardiac differentiation and that mitochondrial dynamics and intracellular calcium can coordinately regulate important developmental pathways [19]. Similarly, there is a growing interest in the role of mitochondrial dynamics in the progression of heart failure and other cardiac pathologies [12].

While significant progress has been made, numerous questions remain regarding exactly how mitochondrial size and shape regulates cardiac function. Moreover, it remains unclear whether the proteins involved in fusion or fission have roles outside of their known function in regulating mitochondrial morphology. As such, the observations of Bhandari and colleagues provide some welcome clarity [20]. In this recent study, the Dorn lab continues their exploration of mitochondrial dyanamics and function in cardiac physiology using Drosophila hearts as a genetic model [20, 21]. In particular, they examined heart tube function in flies in which they used cardiomyocyte-specific expression of an RNAi to suppress expression of the Drosophila single outer membrane mitofusin MARF, or to suppress expression of the inner membrane fusion protein Opa1. Both strategies would be expected to inhibit mitochondrial fusion, and if this was the only thing these two proteins did, than one might expect that the overall phenotype of the MARF-deficient and Opa1-deficient flies should be similar, if not identical. In some ways, the two sets of flies were indeed similar. First, consistent with a fusion deficit, mitochondrial size decreased similarly (by about 50%) in both models. In addition, the two sets of hearts showed impaired cardiac function, there was evidence for increased levels of reactive oxygen species (ROS), and there was evidence for similar degrees of mitochondrial impairment. The later property was assessed by the percentage of mitochondria showing a marked decline in mitochondrial membrane potential, a signature of de-energized, and hence potentially damaged, mitochondria.

All the above data therefore pointed to the conclusion that disrupting MARF or Opa1 function led to a disruption of mitochondrial fusion. This was not completely unexpected given the requirements for inner or outer mitochondrial membrane joining in any fusion event. Furthermore, in the absence of successful fusion, highly fragmented mitochondria might be expected to spew out increased ROS. This oxidative stress in turn could fuel a positive feedback loop leading to a further redox-dependent decline in mitochondrial function, and ultimately leading to compromised cardiac energetics and function. However, when Bhandari and his colleagues went a bit further to explore whether or not disruption of MARF and Opa1 were in fact completely identical, they came to a somewhat surprising conclusion. As mentioned, their data suggested that both MARF-deficient and Opa1-deficient hearts showed evidence of increased ROS. They therefore asked whether or not this oxidative stress contributed to the impaired cardiac phenotype. To alter the redox balance they chose to either overexpress the antioxidant scavenging protein superoxide dismutase (SOD) or to suppress expression of the oxidant-generating mitochondrial enzyme ROS modulator 1 (ROMO). Both antioxidant strategies resulted in a marked suppression of the adverse remodeling and decreased heart function caused by deficiencies in the inner mitochondrial membrane protein Opa1, yet failed to ameliorate the same properties in the MARF-deficient animals. In contrast, based on their work and others [20, 22, 23], MARF-deficient flies were rescued by overexpression of the transcription factor Xbp1. This transcription factor protects cells from ER stress by up-regulating a subset of proteins involved in protein folding ultimately helping to accelerate the clearance of damaged and misfolded proteins. Remarkably, Xbp1 overexpression, while nearly completely rescuing the MARF-deficient animals, was of no benefit to flies lacking Opa1. Thus, flies lacking cardiac MARF or Opa1 expression have similar impairments in fusion, with mitochondria roughly 50% of the size of wild type animals. This defect is associated with a similar decline in cardiac function. Nonetheless, deficiency in the outer membrane protein MARF was rescued by augmenting ER stress tolerance, while deficiency in the inner mitochondrial membrane protein Opa1 was rescued by augmenting ROS scavenging.

There is one final observation of Bhandari and colleagues that is important to emphasize. For those focusing on mitochondrial size, it would seem that flies lacking either Opa1 or MARF both had small mitochondria, and it is possible that these genetically-induced diminutive organelles might by themselves be important. To test this point, the authors also overexpressed the fission-promoter Drp1. As expected, this drove fragmentation of the mitochondrial network, achieving the same, roughly 50% reduction, in mitochondrial size seen in either MARF- or Opa1-deficient flies. However, while the Drp1-overexpressing animals had tiny mitochondria, mitochondrial function and heart tube function appeared indistinguishable from wild type flies.

What are the lessons learned? For one thing, as we all know, looks can be deceiving. For instance, Drp1-overexpression and MARF- and Opa1-deficiency all produce markedly smaller mitochondria, yet these genetic manipulations led to dramatically different phenotypes with regard to cardiac function. Thus, the size of the mitochondria, although often discussed and commented on, is probably by itself not terribly meaningful. Another lesson learned is that experiments that disrupt fission or fusion need to be interpreted with caution. The observation that MARF-deficiency and Opa1-deficiency are rescued by non-overlapping genetic strategies suggests that these proteins most likely have specialized and unique functions that extend beyond fusion. Hints have already emerged that this may be the case for the fusion proteins Mnf1 and Mnf2 [24–26]. Similarly, Opa1has been reported to control cristae remodeling during apoptosis, a function that appears to be independent from its well characterized role in mitochondrial fusion [27].

We are by nature visual creatures that tend to believe what we see and intrinsically think that form and function are intertwined. Since the time of Carl Benda, we have been trying to intuit mitochondrial function by looking at mitochondrial shape. As the current results of Bhandari and colleagues demonstrate, those quaint 19^th strategies are undoubtedly due for a 21^st century update.