The multifaceted roles of mitochondria

Abstract

Much more than the “powerhouse” of the cell, mitochondria have emerged as critical hubs involved in metabolism, cell death, inflammation, signaling, and stress responses. To open our mitochondria focus issue, we asked several scientists to share the unanswered questions, emerging themes, and topics of investigation that excite them.

Membrane contacts, lipid flux, and fission

Mitochondria are double-membraned, 250–500-nm-wide organelles that form a tubular network. Under normal conditions, mitochondria appear more connected than fragmented, but fission allows cells to sense and respond to external cues and understanding how it is regulated becomes important.

Mitochondrial fission is executed by the mitochondrial dynamin (mDyn) with an apparent frequency of ~10 per minute. The evolutionarily related endocytic dynamin (eDyn) also manages fission but does so at a staggeringly high rate of releasing ~2,500 vesicles per minute. What is the mechanistic basis for this difference? During endocytosis, coats bend the PIP₂-rich plasma membrane into buds connected by 30-nm-wide necks, which are enriched in eDyn adaptors. eDyn also binds PIP₂ and reconstitution assays show that it can sever tubes of this size with physiological PIP₂ levels, even without adaptors. So, the presence of adaptors at the neck would only guarantee eDyn recruitment and fission is a highly likely outcome. On the other hand, mDyn adaptors are integral membrane proteins that are constitutively displayed on the organelle. Furthermore, recent reconstitution assays reveal that mDyn can sever tubes of up to 400 nm in size. Why then is mitochondrial fission such an infrequent event? Mitochondria undergo dramatic fission during mitophagy and apoptosis, wherein loss of membrane potential and externalization of cardiolipin from the inner to the outer membrane coincides with mDyn recruitment. mDyn binds cardiolipin and several other anionic lipids. So, under normal conditions, fission is most likely restricted by the limiting levels of cardiolipin on the outer membrane. Thus, while endocytic fission is an outcome of an orchestrated process involving membrane remodeling and adaptor recruitment, mitochondrial fission lies in waiting for specific lipids on the outer membrane.

Recent studies indicate that membrane contact sites (MCSs) between the endoplasmic reticulum (ER), lysosomes, and Golgi-derived vesicles and mitochondria mark sites of fission. MCSs facilitate lipid transfer and growing evidence indicates that they also mediate lipid conversion. Unlike mitophagy and apoptosis that signal a widespread stress response via the mitochondrial lipid cardiolipin, MCSs could elicit a local homeostatic response by transferring specific lipids from other organelles or by the conversion of mitochondrial lipids.

We know very little about dynamic changes in the lipid composition of organelles, and we know even less of such changes at MCSs. Acute perturbation in both the levels and saturation of anionic lipids have been shown to influence mitochondrial fission. Thus, methods for rapid lipid profiling, a systematic survey of lipids that impact mDyn functions and development of fluorescent sensors that report changes in lipid identity and saturation are needed. These could illuminate a novel signaling paradigm where the cell discriminates between homeostatic and stress responses depending on an outside-in or inside-out flux of lipids at the mitochondria.

Re-envisioning the powerhouse

The earliest explorations of mitochondria had humble beginnings. Scientists wielded rudimentary compound microscopes to witness subcellular structures eventually identified as the molecular machines for energy production. Decades later, the universally embraced phrase “powerhouse of the cell” was born. Despite a myriad of technological advances, I maintain that the most striking and relevant contributions of mitochondrial biology to human health and disease originate from three simple features of the mitochondrial network: shape, composition, and number. Together, these are a foundation for diverse signaling cascades, intra-/inter-organellar communications, and informing the extracellular microenvironment.

I have studied cell death signaling for almost 20 years, and I continually ask how mitochondria actively maintain a biochemical environment that is permissive for proapoptotic BCL-2 protein function. My laboratory revealed contributions of mitochondrial lipids, shape, and number toward whether or not a cell dies. However, we still cannot mechanistically appreciate how lipids directly control mitochondrial outer membrane permeabilization nor how structurally related BCL-2 proteins when touched by a BH3-only ligand either support survival or initiate death. I speculate that lipids inform these phenotypes adding both context and signaling diversity. I reckon these pathways are also controlled by intra-/inter-organellar communications via lipid metabolism and organelle quality control.

In my laboratory, we also observed that mitochondrial shape changes disrupt normal metabolism when oncogenes are present, and these mitochondrial changes predict cancer in premalignant lesions. Yet we have little insight into how cancer mechanisms and mitochondrial shape changes crosstalk to inform about disease and whether these same phenotypes and signals directly influence mitochondrial disorders, neuropathies, or aging. As links between mitochondrial shape and inter-organellar communication are evident, the roles for additional organelles (e.g., the ER) in these processes require distinct re-evaluation to parse “pure” mitochondrial contributions from biochemical miscommunications between organelle networks.

Another example of needing to know more about mitochondrial network composition centers on the mitochondrial unfolded protein response. Here, bioenergetic decay within a mitochondrion (or mitochondrial region) is detected by a transcription factor that is stabilized by loss of delta psi, leading to the transactivation of genes necessary to stabilize, repair, and strengthen the mitochondrial network. Despite the general acceptance of this pathway, there are few characterized physiological inducers or molecular details into how diverse stressors activate mitochondrial repair to yield specific outcomes. While most mitochondrial networks physically expand prior and/or during repair, and this is often observable through the microscope, we do not know whether all mitochondrial unfolded protein responses are equivalent. Finally, let’s conclude with a pivotal question: when do mitochondrial adjustments to physiological and/or pathological cues stop being helpful and begin negatively impacting tissue function to promote disease? Literature remains incomplete as to the influences of diverse mitochondrial metabolites and signaling on intracellular stress resolution, inflammation, and the immune landscape.

Indeed, the “humble beginnings” of mitochondrial research have rapidly advanced over the decades toward a “Hubble Telescope future” as recent data from experiments on Earth (by you and me) and in Space (by NASA) support that mitochondria serve as more than only the powerhouse of the cell. They are the basis for metazoan existence, subsequent cellular stress mitigation, and a multitude of intracellular and extracellular adaptations to the diverse scenarios and consequences of living on Earth (and beyond).

Mitochondrial stress responses and aging

The aging process both influences and is influenced by mitochondrial stress responses. During aging, a reduction in the ability to cope with mitochondrial stress and a decline in the ability to activate mitochondrial stress responses have been extensively reported in multiple model organisms. On the other hand, mutations that increase mitochondrial stress resistance have been shown to promote lifespan extension. Therefore, it will be exciting to explore the molecular mechanisms that govern age-related decline of mitochondrial stress responses and to examine the possibility of tuning the threshold/activity of mitochondrial stress responses to control tissue deterioration and aging.

Upon mitochondrial dysfunction, a mitochondrion-to-nucleus communication, known as the mitochondrial unfolded protein response (UPR^mt), activates the transcription of nuclear genes, which function to protect mitochondria. UPR^mt has a distinct temporal profile. In C. elegans, it can only be activated before worms reach the adult stage. Similarly, the decline of the mitochondrial stress response has also been reported in aged rats. These phenomena correlate with the observations that mitochondrial disorders occur during aging, and mitochondria play essential roles in age-related diseases such as neurodegenerations. Unraveling the molecular mechanisms that govern the age-related decline of UPR^mt will be of particular importance. Studies in simple model organisms, in which mitochondrial stress responses can be genetically manipulated, could provide essential insights. Based on the “disposable soma” theory, it will be interesting to test whether a germline-to-soma signal exists to relocate limited resources for investment in reproduction rather than maintenance of the soma through stress response pathways.

Multiple mitochondrial stress response pathways have evolved to protect cells against mitochondrial dysfunction, including UPR^mt, expression of antioxidant proteins, mitochondrial fusion and fission, and mitophagy. However, stress responses will ultimately activate programmed cell death if the damage is very severe. It has been well documented that the constitutive level of intrinsic insults results in a sustained chronic stress response in aged animals. Due to the reduced ability of aged cells to elicit acute stress responses, it is possible that exposure of cells to mitochondrial insults will compound and exacerbate the chronic stress to activate a programmed stress response and activate cell death. Therefore, it will be exciting to understand the threshold and the molecular mechanisms of mitochondrial stress response-dictated cell death and test whether tuning the equilibrium of mitochondrial stress responses will prevent cell death and tissue deterioration in aged animals.

A breakdown in endosymbiosis driving inflammation?

Many immunologists have become obsessed with the role of mitochondria in immune cell activation. This is perhaps one reason why the number of papers published on mitochondria outstrips papers on that now rather secondary organelle, the nucleus, at least from the point of view of this voices article. Starting with cytochrome c release driving apoptosis, it turns out that mitochondria are a source of molecules that drive specific signaling pathways. Metabolites made as part of the TCA cycle such as succinate have roles in promoting inflammation, for example by regulating production of the pro-inflammatory cytokine IL-1beta or acting as a chemokine for neutrophils. Another metabolite, itaconate, which is made from aconitate in the TCA cycle has the opposite effect, with evidence so far indicating that it can suppress inflammation in macrophages by inhibiting the function of inflammatory proteins such as the NLRP3 inflammasome. Mitochondrial DNA can leak into the cytosol of damaged cells, and is sensed by cGAS, which ultimately drives type I interferon production, which can be pathologic in some inflammatory diseases. More recently, evidence has emerged of mitochondrial double-stranded RNA being released and then being sensed by receptors such as MDA-5, which again promotes type I interferon production. The mitochondria can then be viewed as a signaling hub that links into inflammation and immunity. How could this have evolved? Mitochondrial damage is obviously an indicator that something bad is going on and so that needs to be sensed, promoting immunity and inflammation, which is designed to restore homeostasis. If this response is too strong or chronic, however, inflammatory diseases might well be the result. What is striking is the innate immune receptors such as cGAS and MDA-5 were first found as sensors of microbial nucleic acids. In fact, they are seen as frontline sensors of infection, with a critical role in host defense. Given that they can also sense mitochondrial nucleic acids, the question arises as to which came first. Might it be that when endosymbiosis happened 2.5 billion years ago, these receptors evolved to sense when the mitochondria, a relic of the aerobic bacteria that first went inside another cell, go rogue and attempt to break free from their shackled existence? And then later, the same receptors turned out to also sense microbial nucleic acids, possibly in multicellular organisms, that were prey to bacterial or viral infection? Recent evidence also indicates that infection might well be initially sensed by the nucleic acid receptors, but then mitochondrial sources take over to amplify the response against the invading pathogen. Whatever the case may be, we are learning more and more about the role mitochondria play in immunity and inflammation. If we can tame them, we might devise better therapies for immune and inflammatory diseases.

Healthy mitochondria for a healthier life

Mitochondria power the cell and control apoptosis. They also produce ROS as inevitable by-products during OXPHOS, which may cause oxidative stress. Thus, cells have evolved elaborate and sophisticated “policing systems” to monitor mitochondrial behavior and remove defective mitochondria via a mechanism known as mitophagy. To date, an expanding number of molecules, including NIX/BNIP3, FUNDC1, Bcl2-L13, and Mcl-1, have been found to contain the LIR motif that interacts with LC3 to engage mitophagy in response to mitochondrial stress and cellular cues. As a mitochondrion may be too bulky to be engulfed by a mitophagosome, there may exist a type of piecemeal mitophagy that constantly prunes damaged portions of mitochondria. If so, how are these mitophagic machineries activated by internal mitochondrial damage and how do mitophagosomes form at select sites on the mitochondrial surface? Certainly, such a process would require multiple membrane sources and extensive membrane remodeling. Advancements in super-resolution imaging combined with EM and AI techniques may enable us to quantitatively visualize such complex processes in real time. Further understanding the molecular and mechanistic details will be helpful to reconstitute the mitophagosome in vitro.

It is plausible that mitophagy is balanced with mitochondrial biogenesis to maintain mitochondrial homeostasis. It appears that mitophagy pathways, including PINK1/Parkin, NIX, and FUNDC1, crosstalk with PGC1 and/or NRF1 (master regulators of mitochondrial biogenesis). In particular, NRF1 can transcriptionally regulate FUNDC1 and the loss of FUNDC1 decreases the stability of PGC1, resulting in the downregulation of mitochondrial biogenesis. The direct coupling and feedback regulation between mitophagy and mitochondrial biogenesis ensures a delicate balance between the production of new mitochondria and the disposal of old ones. It is technically challenging to directly monitor mitophagy or biogenesis in a temporal-spatial, real-time, and quantitative fashion. Fluorescence-based reporters such as Mito-Kemia and mitoQC can detect mitophagy in vivo. These probes are useful to detect the endpoint of mitophagy but are often unreliable in measuring mitophagic flux. The development of imaging probes with greater sensitivity, combined with chemical biology and optogenetic approaches, are needed for real-time monitoring of mitophagy in vivo.

“To be or not to be” is the constant question for mitochondria and the cell. Mild stress may activate mitophagy to remove damaged mitochondria, protecting cells from oxidative stress, inflammasome activation, and apoptosis. However, severe stress may trigger mitochondria-dependent programmed cell death. Interestingly, several apoptosis regulators contain the LIR motif, such as Mcl-1 and NIX/BNIP3, to promote mitophagy. On the other hand, Bcl-xL directly interacts with parkin and NIX/BNIP3 to inhibit mitophagy. We observed that Bcl-xL, but not Bcl-2, inhibited PGAM5 phosphatase, which dephosphorylates FUNDC1 to activate mitophagy. Post-translational modifications of these molecules may act as molecular switches to determine interacting partners and guide decision making of cell fate. Emerging evidence suggests that, as mitochondria are the major source of iron and ROS, mitophagy may regulate ferroptosis. Further investigation is required to dissect the complex relationship between mitochondrial behaviors and cell fate.

As mitophagy regulates mitochondrial and metabolic reprogramming, and even cell fate, it is not surprising that mitophagy has been implicated in aging and a myriad of diseases. Enhanced mitophagy may benefit the cell, improving longevity and health span. A key issue for future investigation is the identification of small molecules that specifically target mitophagy machinery without affecting overall autophagy, or the identification of bioactive agents that activate mitophagy without disrupting mitochondrial function or dynamics. Alternatively, mitochondrial fission/fusion or homeostatic machineries could be targeted to accelerate the rate of mitochondrial turnover and keep the mitochondrial population young and healthy. The insights gained from these studies could have therapeutic implications in aging and aging-related diseases.

Biographies

Thomas J. Pucadyil

Indian Institute of Science Education and Research, India

Jerry Edward Chipuk

Icahn School of Medicine at Mount Sinai, USA

Ying Liu

Peking University, China

Luke O’Neill

Trinity College, Ireland

Quan Chen

College of Life Sciences, Nankai University, China