Why might mitochondria fuse? Mammalian mitochondria continuously undergo fusion in vivo, a process accomplished by an elaborate machinery of energy-dependant GTPase proteins. These proteins are extensively post-translationally regulated and the fusion process involves actively merging both inner and outer mitochondrial membranes to shape extensive networks. Executing this energetically costly process must hold an essential physiological function. Recent reports demonstrate that cells with fusion-incompetent mitochondria fail to thrive, especially under challenging conditions. Bacteria, the ancestors of mitochondria, also undergo life-promoting networking under challenging conditions. Here, we consider evolutionarily conserved behavioral similarities among mitochondria and their bacterial counterparts, as well as recent exciting discoveries in mitochondrial dynamics. We propose that the hunger of mitochondria to undergo fusion results in interconnected mitochondrial networks that establish a life-sustaining unity and coherence within cells, allowing them to thrive in times of energetic hardship.
Insight from bacteria: Communicating for survival
Insight into why mitochondria undergo fusion can be gained from observing the mitochondrion's ancestors: the bacteria. Under normal circumstances where energetic substrates are highly available, bacteria typically exhibit solitary behaviors. However, when bacterial population density increases and energy supply decreases, individual bacteria engage in cell-to-cell communication. This allows the synchronization of gene expression among members of the whole colony 1, 2, a process termed quorum sensing. Through quorum sensing, bacteria can act in unison by coordinating certain behaviors, which if performed in isolation would be in vain. One culminating result of coordinated bacterial behavior is the ability to form a larger complex multicellular structure called the biofilm 3. This intercellular communication involves the production of cytokines and elaborate signal transduction pathways, and is therefore an energetically costly process. Nevertheless, quorum sensing and biofilm formation enhance bacterial survival and promote colonization of host organisms 3. Thus, for the bacterium to behave as a coordinated unit, it must be starved (i.e. the population density increased). In short, conditions leading to nutrient deprivation initiate cell-to-cell communication, allowing a population of individual bacteria to reach quorum and “agree” on a communal direction to be taken. Once quorum is reached the bacteria behave as a united community working towards a common goal: survival. Could the purpose of fusion in the domesticated bacteria-derived mitochondria serve a similar goal?
Mitochondrial fusion: Physiological roles
The dynamic fusion of mammalian mitochondria is highly regulated by post-translational modifications of the nuclear-encoded proteins, the mitofusins Mfn1 and Mfn2, and optic atrophy 1 (OPA1) 4 (Fig. 1). OPA1 is cleaved within mitochondria at two different sites S1 and S2 by the matrix-processing peptidase OMA1. This gives rise to short and long forms that heterodimerize to induce inner membrane fusion. Mfn1 and Mfn2 are ubiquitinylated by the E3-ligase MARCH5, and probably also experience other function-defining post-translational modification (reviewed in ref. 4). That post-translational modifications are indeed the key element in the regulation of this process is also suggested by the time frame – minutes to hours – within which mitochondria can undergo global fusion.
The purpose of mitochondrial fusion can be gleamed from observing the consequences of the absence of fusion. Undeniably, cells with fusion-incompetent mitochondria fail to thrive 5. These cells have reduced growth rates and can only perform oxidative phosphorylation to limited levels 6. Mice completely lacking Mfn1 and Mfn2 fail to develop and die early in embryonic stage 7. Mice completely lacking OPA1 develop past embryonic stages and live to adulthood, but they die early. Mice lacking only Mfn2 in skeletal muscle have fusion-incompetent mitochondria. They do live to adulthood but exhibit a 50% reduced body growth, impaired thermogenesis, and reduced oxidative capacity; they also accumulate high levels of mitochondrial DNA (mtDNA) mutations 8. Excessive mitochondrial fragmentation in cells or organisms in which fusion is disabled impairs mitochondrial bioenergetics, renders host cells more susceptible to apoptotic death 9, 10, and promotes accumulation of respiratory chain deficiency secondary to mtDNA abnormalities 11. Impaired mitochondrial fusion thus leads to changes in function of the organelle itself, and of the cell as a whole.
It is a fundamental biological principle that survival of the whole depends on interactions between its parts. Braschi and McBride 12 have argued that, to act towards the greater good of the cell and the organism, mitochondria must act as a unified coordinated network. This is accomplished by active fusion of mitochondria into a relatively interconnected reticulum, or mitochondrial network 13. Microscopic evidence demonstrates that mitochondrial fusion is necessary to allow sharing of mtDNA and large proteins 8, 14. Some have argued that the effect of mitochondrial fusion-fission dynamics on mtDNA plays a determinant role in cellular aging 15. In addition, several reports indicate that cells with fused elongated mitochondria might exhibit greater bioenergetic efficiency and have pro-survival effects both in vivo 9, 16 and in vitro 17, 18. Thus, adequate mitochondrial fusion is required for the function of the organelle, of the cells that contain them, and of the organism as a whole 19.
From a global perspective, it therefore appears that both bacteria and mitochondria can adopt a networking behavior promoting survival under challenging conditions. In the case of bacteria, this is achieved through the releases of soluble factors that mediate communication between individual bacteria. In addition to these well-established mechanisms, recent evidence indicates that bacteria can also transfer information through secreted vesicles 20, 21, bacterial conjugation channels 22, and express proteins (in particular DynA) that can tether membranes and mediate nucleotide-independent membrane fusion in vitro 23. This suggests that bacterial quorum-sensing mechanisms may involve contact between membranes. In the case of mitochondria, quorum sensing has not been explicitly considered. But here we suggest that this phenomenon does occur, and that one of the primary mechanisms involves fusion of individual mitochondria into elaborated networks. Such community building behavior among mitochondria could have profound implications in determining how eukaryotic cells will fare during metabolic perturbations.
Implications for retrograde signaling and cellular coherence?
Mitochondria do indeed respond to changes in energetic substrate levels by undergoing substantial and rapid morphology changes. More specifically, low energy levels (i.e. starvation) triggers mitochondrial fusion into elongated tubules 16, 17, whereas exposure to high glucose and lipids levels has the opposite effect, fragmenting the mitochondrial network 24, 25. In addition to these changes in morphology, substrate levels and ATP/ADP levels also induce important allosteric control of oxidative phosphorylation components that regulate mitochondrial function 26. Due to their extreme sensitivity to energetic substrate levels, mitochondria are exquisite sensors of the cellular energy status. Details about this energy status represent crucial information for the nucleus where epigenetic mechanisms modulate gene expression and ultimately dictate cell function 27. So, does effective retrograde signaling from mitochondria to the nucleus necessarily require potent and coherent signals involving the coordinated function of most mitochondria behaving in unison?
As with bacteria, could mitochondrial fusion also represent a means by which total mitochondrial population size is monitored? Besides the increased bioenergetic efficiency of fused mitochondria, longer organelles exhibit resistance to depolarization and permeability transition pore (PTP) opening. They also produce fewer reactive oxygen species (ROS) than their fragmented counterparts 28, 29. ROS and other mitochondrial molecules released during reversible PTP opening (e.g. Ca2+, ATP, acetyl coenzyme a, NAD+, and Cyt c) act as cellular signaling molecules that can affect signal transduction pathways and ultimately modulate gene expression of mitochondrial genes 27, 30. This may partly explain why in most cell types mitochondria tend to cluster more densely around the nucleus. Therefore, morphology regulation – via its effect on mitochondrial outputs – could represent a potent mean of monitoring and transducing signals about mitochondrial mass and energetic status to the cytoplasm and nucleus of the cell.
Like bacteria, might mitochondria communicate to reach quorum? Does mitochondrial fusion constitute an essential community check that enables population-wide behavior among these organelles? Mitochondrial fusion is an energetically costly process but nevertheless appears essential to promote cell survival and adaptation to shortage of energy supplies. Mitochondria have preserved many fossil features of their ancestors, the bacteria, including a double membrane and circular DNA. Cell-cell or mitochondrion-mitochondrion communication might be one more evolutionarily conserved feature to be added to this list. Consideration of the functional consequences incurred by mitochondrial morphology transitions, as well as of their impact on cell function and survival, may enable the discovery of yet additional layers of regulation for this fantastically complex organelle.
The authors' work is supported by a scholarship from the National Sciences and Engineering Research Council of Canada (M.P.), fellowships in Psychosocial Oncology and in Systems Biology from the Canadian Institute of Health Research (M.P.), and a Career Award from the Fonds de Recherche en Santé du Québec (Y.B.).