1 A break in mitochondrial endosymbiosis as a basis for inflammatory diseases Michael P. Murphy1,2 & Luke A.J. O’Neill3 1MRC Mitochondrial Biology Unit, University of Cambridge, Cambridge, UK. 2Department of Medicine, University of Cambridge, Cambridge, UK. 3School of Biochemistry and Immunology, Trinity Biomedical Sciences Institute, Trinity College Dublin, Dublin, Ireland. Summary Mitochondria retain bacterial traits due to their endosymbiotic origin, but host cells do not recognise them as foreign, because the organelles are sequestered. However, the regulated release of mitochondrial factors into the cytosol can trigger cell death, innate immunity and inflammation. This selective breakdown in the 2 billion year-old endosymbiotic relationship enables mitochondria to act as intracellular signalling hubs. Mitochondrial signals include proteins, nucleic acids, phospholipids, metabolites and reactive oxygen species, which have many modes of release from mitochondria, and of decoding in the cytosol and nucleus. Since these mitochondrial signals likely contribute to the homeostatic role of inflammation, dysregulation of these processes may lead to autoimmune and inflammatory diseases. A possible reason for the increased incidence of these diseases may be due to changes in mitochondrial function and signalling in response to such recent phenomena as obesity, changes in diet and other environmental factors. Focusing on the mixed heritage of mitochondria therefore leads to predictions for future insights, research paths, and therapeutic opportunities. Thus, while mitochondria can be considered “the enemy within” the cell, evolution has used this strained relationship in intriguing ways, with increasing evidence pointing to the recent failure of endosymbiosis being critical for the pathogenesis of inflammatory diseases. Introduction It is now well accepted that mitochondria play many central roles within the cell, far beyond their essential energetic functions in oxidative phosphorylation, the Krebs cycle and fatty acid oxidation 1-4. These non-canonical activities include signalling, biosynthesis and the regulation of cell fate 3. Particularly intriguing is a deluge of findings into how mitochondria act in the immune and cell death signalling pathways that enable cells to respond to infection or damage 5,6. A recent analysis revealed that since 2011, publications on mitochondria began to eclipse those on other organelles 3, reflecting the upsurge in interest in mitochondria beyond bioenergetics. This raises the question of why so many pathways that react to the challenge of infection or tissue injury utilise mitochondria as a central signalling hub to integrate and transduce the cell’s response? A likely factor is that the endosymbiotic origin of mitochondria marks them 2 apart from the rest of the cell in a way that can be co-opted to produce key messages pertaining to cell fate 2,7-10. Here we discuss how the evolutionary origin of mitochondria has been utilised by cells to facilitate their response to injury and infection. As well as providing an explanation for the critical role of mitochondria in cell fate, this hypothesis raises intriguing new research questions. One in particular that we raise here is whether the increase in inflammatory diseases such as systemic lupus erythematosus, multiple sclerosis, rheumatoid arthritis and inflammatory bowel disease 11-16 over the past few generations could be due to a break in the endosymbiotic relationship between mitochondria and the cell, driven by such relatively recent phenomena as the rise in obesity, chronic stress, changes in diet towards processed food, social changes such as lack of sleep, and synthetic chemicals in the environment. Might these insights provide new therapeutic options to treat chronic inflammatory diseases, many of which remain difficult to address and present a significant burden on humanity, contributing to over 50% of all deaths 17? Implications of endosymbiotic origin While the details are still vigorously debated 18-20, the broad outlines of the endosymbiotic origin of mitochondria and eukaryotic cells are now well accepted 21. Although a bacterial ancestry to mitochondria was first suggested late in the 19th century by Portier and Wallin 18,19, this idea did not gain widespread acceptance until it was compellingly revived in 1967 by Lynn Margulis 22. The consensus is that eukaryotic cells arose as a single event about 2 billion years ago when an Asgard archaeon entered into an endosymbiotic relationship with an α-proteobacterium 19,21,23-26. Both cells must have benefitted from this arrangement, although the mechanical basis of these advantages is still disputed 24,27,28. Over time, most of the α-proteobacterial DNA either relocated to the nascent nucleus or it was eliminated with its functions replaced by the host’s genome, leaving a small mitochondrial DNA molecule, 16.5 kb in mammals, that encodes only 37 genes required for the assembly of the oxidative phosphorylation machinery 29-31. The remaining 1,200 or so types of protein present within mammalian mitochondria are translated in the cytosol for subsequent import into the mitochondrion 32,33. Intriguingly, mitochondria retain a number of bacterial traits: metabolic pathways, unmethylated DNA, double stranded (ds)RNA, N-formylpeptides, elevated reactive oxygen species (ROS) production, and the phospholipid cardiolipin (See Box 1). Importantly, these bacterial aspects of mitochondria are all associated with the matrix, inner membrane and intermembrane space 25,34. These compartments are surrounded by the mitochondrial outer membrane which probably arose from both the outer membrane of the α- proteobacterium and the endocytic plasma membrane of the Archaeon host 25,34. This mixed origin is indicated by the mitochondrial outer membrane containing β-barrel proteins that evolved from bacterial outer membrane precursors 35, while the phospholipid composition of the mitochondrial outer membrane is similar to other cell membranes, but distinct from the mitochondrial inner membrane 25,34. Thus, while mitochondria are clearly now fully integrated into the eukaryotic cell, we can also consider mitochondria as a pseudo bacterium “bricked in” behind the mitochondrial outer membrane which allows exchange of metabolites but – under most conditions – retains macromolecules. Any breach in the protective barrier provided by the 3 mitochondrial outer membrane will allow these pseudobacterial factors to enter the cytosol which can potentially be recognised as “foreign” 9, activating immune signalling and cell death pathways. So, evolution has not erased the bacterial origins of mitochondria but has retained many of these traits because of their usefulness to the cell. Mitochondrial signals to the cytosol This retained “otherness” of mitochondria reveals itself to the rest of the cell by the release of molecules into the cytosol by a number of mechanisms (Fig. 1) that activate multiple immune signalling or programmed cell death pathways (Fig. 2). Proteins from the mitochondrial intermembrane space, such as cytochrome c, SMAC/Diablo and Omi, are released upon mitochondrial outer membrane permeabilisation (MOMP) to induce the intrinsic apoptosis pathway in response to cell stress 36-38. The formation of the MOMP pore is still incompletely characterised, but occurs in response to the interaction of pro-apoptotic members of the Bcl-2 family proteins such as Bax and Bak 38-40. MOMP exposes the inter membrane space to the cytosol without altering the permeability of the inner membrane allowing large proteins to exit 39,40. Many intermembrane space proteins such as cytochrome c are largely present within cristae, and are therefore not in equilibrium with the space between the mitochondrial inner boundary membrane and the outer membrane consequently there is extensive further remodelling of the cristae by OPA1 in concert with MOMP to enhance release of these proteins 38,41,42. The peroxidation of unsaturated fatty acids of cardiolipin in the inner membrane is also proposed to enhance cytochrome c release 43. Access and egress from the cristae to the space between the mitochondrial inner boundary and outer membranes is also modulated by the action of the Mitochondrial Contact Site and Cristae Organising System (MICOS) complex 44,45. The mitochondrial inner membrane is far greater in surface area than the outer membrane, consequently swelling of the mitochondrial matrix can rupture the outer membrane with the subsequent release of intermembrane space proteins 10,46,47. One mechanism of matrix swelling is by induction of the mitochondrial permeability transition pore (mPTP) which forms in the inner membrane, rendering it permeable to molecules up to ~1.5 kDa 46,48. While the composition of the mPTP is disputed 48, the mPTP is induced by a wide range of factors, notably by oxidative stress and mitochondrial calcium overload, and hence is often associated with necrotic cell death 46. However, there are many other redox and stress modifiers of its function including the mitochondrial membrane potential, while the pore itself has a number of different permeability states, although their (patho)physiological significance is not yet clear 48. Whether induction of the mPTP can be modulated to selectively release proteins from the inter membrane space, or whether it leads on to rupture of the inner membrane and release of matrix macromolecules is not certain. MtDNA is distinct from that in the nucleus, resembling bacterial DNA due to its lack of CpG methylation 49. So, if mtDNA appears in the cytosol it will be detected by sensors of bacterial DNA such as Toll-like receptor -9 (TLR9) and cGAS, which generates the second messenger cGAMP 50. These sensors will lead to the induction of multiple pro-inflammatory genes, notably cytokines including Type I interferons. There are a number of potential mechanisms by which mtDNA can be released from the mitochondrial matrix 51. One is 4 herniation of the mitochondrial inner membrane which occurs as the mitochondrial matrix swells, pushing sections of the inner membrane encapsulating mtDNA and its binding proteins through Bax/Bak pores in the mitochondrial outer membrane 52. Furthermore, an unbiased screen has revealed a possible role for cristae remodelling by OPA-1 in mtDNA release41. The mtDNA seems to be intact and remains encapsulated within vesicles and then can conjugate to endosomes enabling the mtDNA to be recognised by endosomal TLR9. There is also the direct release of fragmented mtDNA from the matrix into the cytosol which seems to be associated with induction of the mPTP and oligomerisation of the pore-forming protein Voltage-dependent Anion Channel (VDAC) in the mitochondrial outer membrane 53. Finally, there is also the release of newly synthesised and oxidised mtDNA molecules, which have been shown to interact with the NLRP3 inflammasome 54. Activation of NLRP3 leads to the stimulation of caspase-1, which processes the precursors of the inflammatory cytokines IL-1β and IL18, and the Gasdermin family of proteins to promote pyroptotic cell death 55. Regulation of NLRP3 by mitochondria however remains complex, since it has also been shown that mitochondrial ATP is required for NLRP3 activation, generating phosphocreatine, which in turn leads to cytosolic ATP production via Creatine Kinase B, the ATP being required for NLRP3 inflammasome assembly via direct binding to the NACHT domain of NLRP3 56. In all these cases, considerable uncertainty remains about whether the released mtDNA is intact, fragmented or oxidised, or indeed if the process for extruding the mtDNA is selective, occurs directly into the cytosol, or via vesicles that can be further processed, or even extruded from the cell 51. The release of mitochondrial RNA (mtRNA) into the cytosol also occurs in a range of scenarios 57. Mammalian mtDNA encodes 2 rRNAs, 22 tRNAs and 11 mRNAs, two of which are bicistronic thereby encoding 13 mitochondrial polypeptides. As mtDNA encodes genes on both strands that overlap this can lead to dsRNA molecules 58, that upon release can activate the viral sensing pathways through RIG-I like receptors (RLRs) including RIG-I itself and MDA-5, that interact with MAVS, which perhaps conveniently localises to the surface of the mitochondrial outer membrane 59. The mechanism(s) of dsRNA release are less certain but may be similar to those for mtDNA release. The 13 mitochondrially-translated polypeptides contain an N-terminal N-formylated methionine (fMet) residue which originates from the N-formylation of a proportion of the methionine-charged tRNA used to initiate translation on mitochondrial ribosomes 60. N- formylated Met is retained in 12 out of 13 of the mitochondrially-encoded proteins in mammals, the one exception being the Cox III subunit of cytochrome oxidase 61. Bacterial proteins also contain an N-terminal fMet and consequently mammalian cells contain a series of N-formyl peptide G-protein coupled receptors on cell membranes that act as chemotaxis receptors to respond to bacterial fMet peptides 62. In addition, a number of N-formyl peptides that originate from the N-termini of mitochondrially-translated proteins are released during various forms of inflammation 63 and seem to act via N-formyl peptide G-protein coupled receptors 64. How these peptides are processed and released from mitochondria is unclear but is yet another example of the bacterial origin of mitochondria generating potential signals. Cardiolipin (CL) has a slightly different structure from other phospholipids, with two phosphatidic acids linked by a third glycerol molecule 26,65. CL is widespread in bacteria but in eukaryotes is found primarily within the mitochondrial inner membrane where it is synthesized and the fatty acid composition remodelled 66,67. In both bacterial and mitochondrial membranes, 5 the properties of CL are utilised to adjust membrane curvature and are also closely associated with, and essential for, the function of many membrane proteins, indicating why it has been retained by mitochondria 65. In a number of situations CL is translocated to the mitochondrial outer membrane, probably at contact sites between the mitochondrial inner and outer membranes 67. The translocation of CL is often associated with the peroxidation of the unsaturated fatty acids that predominate in CL, and once exposed on the mitochondrial outer membrane peroxidised CL may contribute to the formation of the pore that enables MOMP 67. Furthermore, the exposure of CL on the cytosolic facing surface of the mitochondrial outer membrane can facilitate the formation of protein signalling assemblies, potentially including the recruitment and activation of NLRP3 68. Furthermore, exposure of CL on the mitochondrial outer membrane can also mark the organelle for mitophagy 67, which may have been retained since the origins of the endosymbiosis of the proto-mitochondrial α-proteobacterium 26,28. Interestingly, Polly Matzinger, originator of the Danger Hypothesis which theorises that the immune system recognises various threats in order to become activated, considered CL recognition to be an excellent example of a ‘danger’ signal 69. Current questions include how the peroxidation of unsaturated fatty acids on CL is regulated and how the remodelling and translocation of CL to the outer membrane are coordinated. The mitochondrial respiratory chain can produce superoxide from complexes I and III 70,71. The production of superoxide by complex I by reverse electron transport (RET) under conditions of high protonmotive force (∆p) and a reduced Coenzyme Q pool, is of particular interest because it can be regulated physiologically in response to changes in these key mitochondrial parameters 72,73 and is emerging as a potential mode of mitochondrial redox signalling 74,75. Thus, there is considerable interest in this process underlying a redox signal. For example, RET could be activated, by turning off the consumption of ∆p through ATP synthesis by the FoF1-ATP synthase or by enhancing glutaminolysis as a source of succinate to drive RET 76. The expectation is that superoxide produced within the mitochondrial matrix by either complex would be converted to H2O2 by the high local concentration of Mn-superoxide dismutase (SOD) 70. H2O2 can act as a potential redox signal, most probably by reversibly oxidising cysteine residues on key signalling relay proteins 77,78, such as is seen in the redox- sensitive OxyR transcription factor in bacteria 79 and similar redox sensitive cysteines are widespread in eukaryotic proteins 77. Even so, there are uncertainties about the circumstances under which H2O2 can diffuse from mitochondria within cells, or whether elevated H2O2 may act within mitochondria and thereby contribute to the release of other factors into the cytosol, for example by the activation of the mPTP 80. Similarly, the peroxidation of CL is a further potential mode of signalling, but whether CL is modulated directly by mitochondrial ROS is unclear. In some circumstances there are suggestions that elevated mitochondrial ROS can act as a means of killing invading bacteria within cells 81. The sequestration of much of central metabolism within the mitochondrial matrix, coupled with the selective transport of polar metabolites such as α-ketoglutarate, citrate, itaconate and succinate across the mitochondrial inner membrane, enables the release of mitochondrial metabolites to act as signals 6,82-84. In all cases, a signalling modality requires the regulated production and export of the metabolite to the cytosol/nucleus followed by a mode of responding to this signal. Elevated generation of citrate and its export to the cytosol 6 produces large amounts of acetylCoA, some of which drives fatty acid biosynthesis, but which also alters levels of histone acetylation 85. During inflammation succinate production is enhanced by increased glutaminolysis in conjunction with reduction of the CoQ pool, which leads to succinate export via the dicarboxylate carrier to the cytosol 75. Once in the cytosol, succinate can affect the activity of α-ketoglutarate-dependent dioxygenases (αKGDD) which act to introduce an oxygen atom into protein targets. This is important in the activation of hypoxia inducible factor (HIF)-1α, leading to the induction of such cytokines as interleukin (IL)-1β 76,86. Mitochondrial metabolites such as succinate and α-ketoglutarate can also alter DNA and histone methylation as potential epigenetic signals 87-89. Similarly, following inflammatory activation of macrophages there is upregulation of Aconitate Decarboxylase-1 (ACOD1, encoded by the gene IRG-1) that produces itaconate from aconitate that is then exported from the matrix to the cytosol 90. Itaconate is a metabolite gaining intense interest, and has been shown to have anti-bacterial effects but also to modulate various inflammatory proteins, leading to a net anti-inflammatory effect 91,92. In addition to individual metabolites signalling from mitochondria, there is also the well-established metabolic shift known as the Warburg effect 93. The rationale for the Warburg effect is generally thought to be that it enables increased flux through glycolysis and the pentose phosphate pathway and hence greater production of metabolic building blocks for cell growth 94. The far greater flux through glycolysis during immune cell activation and in cancer cells often leads to sufficient ATP production so that oxidative phosphorylation is no longer required. However, it is important to note that ATP production by oxidative phosphorylation is just one of many roles carried out by mitochondria. Many of these, notably the TCA cycle and the respiratory chain, are essential for biosynthesis and thus are still required for cells undergoing a shift to glycolytic metabolism 95-98, even if mitochondria are no longer the main suppliers of ATP. An underappreciated aspect of this shift from ATP supply by oxidative phosphorylation to glycolysis is that it can also be brought about by blocking the ability of mitochondrial oxidative phosphorylation to supply ATP to the cytosol, thus forcing the cell to switch to glycolytic ATP production. In summary, the shift to aerobic glycolysis is not an indicator of a loss of mitochondrial function, since most tumours and inflammatory require active mitochondria with a functioning respiratory chain and Krebs cycle that can generate oncometabolites or intermediates for biosynthesis 95. Thus, the sequestration of mitochondria from the rest of the cell in conjunction with their “otherness” enables mitochondria to be utilised as signalling hubs for multiple processes, notably in cell death and immunity. But this raises many questions and leaves numerous loose ends. Among these are were there any antecedents to these pathways in the bacteria that became mitochondria? And were there precursors of the receptors now active in the eukaryotic cells present in archaea, or did these pathways evolve after endosymbiosis? Inheritance or exaptation? Sequestered mitochondrial compartments within eukaryotic cells containing components of bacterial origin, therefore likely facilitated the evolution of signalling pathways centred on the organelle 26,28. Two complementary scenarios for this are probable. One is that mitochondrial 7 signalling pathways developed from those already utilised before endosymbiosis. These could be from the proto-mitochondrial α-proteobacterium that were transferred to the host genome, or from the archaeal host. The alternative is that endosymbiosis enabled many mitochondrial components to be repurposed for new functions that arose to take advantage of the new situation following endosymbiosis – a process known as exaptation 99 - defined as features acquiring new functions for which they were not originally selected. For example, the role of Bax/Bak in conjunction with OPA-1 in mtDNA release fits with the exaptation model, whereby that function might have evolved after endosymbiosis. Most likely, a combination of both processes contributes. Considering bacterial antecedents for mitochondrial RET, it is important to note that complex I from the α-proteobacterium Paracoccus denitrificans, which is often used as a model for the proto-mitochondrial endosymbiont 25,100 can readily undergo RET 101,102. While the role of RET in P. denitrificans is unclear, it indicates that RET was likely available as a potential mode of signalling in the early prokaryotic cell. Bacteria such as P. denitrificans can elevate ROS production in response to stressors such as DNA damage leading to cell death 103 , hence it is plausible that mitochondrial ROS production and lysis in the eukaryotic cell are derived from bacterial antecedents. The release of factors from mitochondria into the cytosol recapitulates a mechanism by which bacteria protect against phage infection by lysis of infected cells before the invader has a chance to fully replicate 8,104. Furthermore, bacteria can release vesicles encapsulating intracellular components 105. However, the mechanistic parallels between these bacterial processes and those of MOMP, vesicle release and mPTP induction in mitochondria are not fully clear. Furthermore, key components of eukaryotic cell-autonomous innate immune responses have evolved from bacterial precursors of the cGAS-STING pathway, Gasdermins and Toll-IL-1 receptor-Resistance (TIR) domain containing proteins (such as occur in TLRs in mammals), that all protected bacteria against phage infection 8. It is therefore possible that at least some of these pathways existed in the archaeon that originally hosted the α- proteobacterium, and were then co-opted into sensing the release of factors from the endosymbiont should the need arise 106-108. In that original endosymbiont, nuclear gene expression might have been initiated by these systems to ensure homeostasis should the endosymbiont become damaged and reveal itself. Mitophagy could clear the damaged endosymbiont, perhaps by recognising exposed CL 26, although quite how mitophagy evolved in still uncertain 109. In multi-cellular organisms, which took a further c1.4 billion years to evolve, these same pathways, along with additional ones, might therefore provoke local or systemic inflammation in order to restore homeostasis following tissue injury or infection. What therefore began as the capacity of a single cell to survive in response to mitochondrial damage, became an organismal inflammatory process, in both cases, the protective events being triggered by the recognition of components from the damaged mitochondria. Sterile and non-sterile inflammation overlap Pattern recognition receptors (PRR) respond to both sterile inflammation via Danger- Associated Molecule Patterns (DAMPS) and to infection via Pathogen-Associated Molecular Patterns (PAMPs), with considerable overlap in the PRRs that respond to endogenous DAMPS 8 from mitochondria with those that respond to external PAMPs from bacterial or viral infection 9,110. This raises the question as to which came first? Did the innate immune system arise to respond to cell damage that led to the release of DAMPS, and then subsequently respond to PAMPs, or was it originally designed to address infection and was then co-opted to respond to mitochondrial damage? Evidence is emerging of an interesting interplay between PAMP and DAMP sensing. RNA-containing viruses such as the influenza virus, dengue virus or norovirus can be detected by RNA sensors such as RIG-I- that then promote the release of mtDNA which amplifies the induction of Type I IFNs via cGAS-STING 111-113. DNA viruses such as the Herpes virus also drive the release of mitochondrial DNA to promote anti-viral innate immunity 114. It therefore appears that it is a PAMP/DAMP combination that is required for an optimal response, the DAMP originating in the mitochondria in the form of mtDNA (Fig. 3). Endosymbiosis and autoimmunity The similarity in the PRRs activated by DAMPs from mitochondria and PAMPs from infection raises the question of cross talk between DAMP/PAMP sensing contributing to autoimmunity. In this context it is important to note that in many autoimmune diseases antibodies against mitochondrial antigens are detected: these include against CL 115, mitochondrial proteins 116, mtDNA 117 and mtRNA 118. Furthermore, the overlap between the responses to mitochondrial and bacterial DNA and mitochondrial and viral dsRNA makes this another way in which an autoimmune response could arise. In addition, there are situations where succinate arises from mitochondrial metabolism 75,119, which may contribute to inflammatory signalling 86. Supporting this, succinate is found in inflamed joints and promotes IL-1β production via the succinate receptor SUCNR1 120. A recent study has also revealed the prominence of dsRNA in multiple autoimmune and inflammatory diseases 121. Strikingly, the authors report on how aberrant editing of dsRNA by Adenosine Deaminase Acting on RNA (ADAR) (which converts adenosine to inosine, rendering dsRNA more like ‘self’ RNA), generates immunogenic dsRNA which is sensed by MDA5 promoting inflammation. Since a large proportion of dsRNA in eukaryotic cells is likely to be of mitochondrial origin 58, this suggests that aberrant sensing of mitochondrial dsRNA might be pathologic in a number of autoimmune conditions. Finally, a link from fumarate hydratase (FH) to the release of mitochondrial dsRNA has been made 122. Activation of TLR4 with the gram-negative bacterial product LPS leads to a decrease in FH in macrophages, which is linked to dsRNA release via an unknown mechanism. The dsRNA then drives Type I IFN production via MDA5 and RIG-I. FH was also reported to be repressed in monocytes from patients with systemic lupus erythematosus, a known interferonopathy, again possibly linking mitochondrial dsRNA to autoimmunity. FH-deficient tumours have also been shown to release their mtDNA and promote Type I IFN production, which further supports a role for FH in mitochondrial nucleic acid release 123,124. Emerging aspects The insights gained from considering the bacterial origin of mitochondria suggests broadening our outlook to also consider whether other bacterial pathways might also be used by 9 mitochondria in eukaryotic cells. For example, bacteria use quorum sensing by which they self- produce extracellular chemical signals, which can accumulate in a local environment to levels that are required to activate transcription of specific genes 125. Mitochondria regularly release metabolic signals such as succinate, citrate and itaconate 82 and can also release ROS and other signals to modulate the action of adapter proteins such as Miro that regulate the ways in which mitochondria interact with and are moved by the cytoskeleton126,127. At the moment the focus has been on the reading of these signals in the cytosol and in the nucleus, but these messengers can also be easily taken up by mitochondria in the same cell and indeed pass from cell to cell through plasma membrane transporters, or act on cell surface receptors 75. Related to this is the generation of mitokines such as FGF21 and GDF15 which are generated in response to mitochondrial stress to signal to other tissues 128, but much still needs to be resolved about whether these are mitochondrial-selective or general stress responses. Other potential modes of signalling worth exploring include the generation of lipid peroxidation signals from mitochondria, potentially from the peroxidation of unsaturated fatty acids on CL 67, or of ROS induced ROS release 129. Whether these signals can affect local mitochondrial transcription/translation has not been explored, but may be a fertile area, particularly in cells with mitochondria at a distance from the nucleus, such as in neurons. The chemotactic mobility of bacteria in response to signals is well established. Mitochondria readily move, divide and recombine throughout the cell and this process is also linked to mitophagy 130-132. Mitochondrial dynamics involves the close interplay between the mitochondria, cytoskeleton and other organelles, notably the ER. In addition, mitochondrial location within the cell can respond dynamically to a number of stimuli, moving close to the plasma membrane under conditions of ATP demand, or in response to infection 130-132. The current assumption is that the mitochondria are moved around the cell by the cytoskeleton, but there may well be local signals generated by and acting on mitochondria intracellularly that contribute to these processes. Intriguingly, it is now clear that intact mitochondria can migrate from cell to cell within the body, including from one cell type to another 133-137. This can occur by tubular connections, transfer of vesicles or by cell-to-cell connections such as gap junction channels and by cell fusion 135,136. It may also be that damaged mitochondria can trigger the activation of mitochondrial transfer from healthy cells or from stem cells 135. Related to this, there is considerable interest in mitochondrial transfer/transplantation as a therapeutic strategy, for example following ischaemia-reperfusion injury to the heart 138, 139. External mitochondria are readily endocytosed within cells and the tacit assumption has been that the addition of new mitochondria enhances oxidative phosphorylation in the target cells, although this is seldom demonstrated explicitly and many other interpretations of the positive effects are possible. This ability of mitochondria to migrate from cell to cell is of course a further reflection of their endosymbiotic origin. Outlook A key question is when does endosymbiosis break down? Since inflammation is a normal physiological process designed to restore us to health following injury and infection why does this process go rogue in so many diseases? The rise in incidence of these diseases has led to 10 various culprits being implicated including obesity, chronic stress, poor diet (potentially with a contribution from processed foods), lifestyle changes such as disrupted sleep or environmental toxins 17, most likely in the context of specific genetic backgrounds. Mitochondria provide us with a unifying target for all of these processes 1,140-142 (Fig. 2). Might the pressure on mitochondria caused by obesity, chronic stress, certain foodstuffs, disrupted circadian rhythms or damage to mitochondria caused by various environmental toxins exacerbate a process whose homeostatic goal is to drive inflammation, but in a beneficial way? All of these triggers could well give rise to enhanced release of the mitochondrial factors discussed here, be they ROS, metabolites, peptides or nucleic acids, with the sensors designed to trigger beneficial inflammation being hyperactivated, leading to inflammatory diseases. It is intriguing to consider what the early pioneers of mitochondrial research, such as Hans Krebs or Peter Mitchell, might make of the exciting developments discussed here. The theory of endosymbiosis provided us with a huge insight into the evolution of eukaryotic life. Without this unlikely random event, we as a species would not have evolved to consider it. That the Krebs cycle gets repurposed for inflammation, or that mitochondrial nucleic acids might be key triggers for immunity via induction of cytokines gives us a whole new perspective on what mitochondria are doing in our cells. Might these new insights provide an explanation for the increased incidence in inflammatory and autoimmune diseases? And might these diseases be explained in part by a fracture in a 2 billion year relationship? This insight could well give rise to new therapies to treat a rapidly growing and troubling group of diseases. Future studies should help clarify these issues and allow us to target these most intriguing of organelles for therapeutic gain. Acknowledgements We thank Evanna L. Mills, Dylan G. Ryan and Hiran A.Prag for helpful discussions. Author contributions Both authors contributed equally to the writing of the manuscript. Competing interests None References 1 Murphy, M. P. & Hartley, R. C. Mitochondria as a therapeutic target for common pathologies. Nature reviews. Drug discovery, doi:10.1038/nrd.2018.174 (2018). The many pathogical roles of mitochondria are discussed 2 Chandel, N. S. Evolution of Mitochondria as Signaling Organelles. Cell metabolism 22, 204-206, doi:10.1016/j.cmet.2015.05.013 (2015). The key signalling roles of mitochondria are here discussed in an evolutionary context 3 Picard, M. & Shirihai, O. S. Mitochondrial signal transduction. Cell metabolism 34, 1620-1653, doi:10.1016/j.cmet.2022.10.008 (2022). 4 Monzel, A. S., Enriquez, J. A. & Picard, M. Multifaceted mitochondria: moving mitochondrial science beyond function and dysfunction. Nat Metab 5, 546-562, doi:10.1038/s42255-023-00783-1 (2023). 11 A recent review that highlights the many emerging facets of mitochondrial biology 5 Tait, S. W. & Green, D. R. Mitochondria and cell signalling. J Cell Sci 125, 807-815, doi:10.1242/jcs.099234 (2012). 6 Bahat, A., MacVicar, T. & Langer, T. Metabolism and Innate Immunity Meet at the Mitochondria. Front Cell Dev Biol 9, 720490, doi:10.3389/fcell.2021.720490 (2021). 7 Marchi, S., Guilbaud, E., Tait, S. W. G., Yamazaki, T. & Galluzzi, L. Mitochondrial control of inflammation. Nat Rev Immunol 23, 159-173, doi:10.1038/s41577-022- 00760-x (2023). 8 Wein, T. & Sorek, R. Bacterial origins of human cell-autonomous innate immune mechanisms. Nat Rev Immunol 22, 629-638, doi:10.1038/s41577-022-00705-4 (2022). Here the authors suggest how the origins of mitochondria can lead to innate imunity mechanisms 9 Krysko, D. V. et al. Emerging role of damage-associated molecular patterns derived from mitochondria in inflammation. Trends Immunol 32, 157-164, doi:10.1016/j.it.2011.01.005 (2011). 10 Galluzzi, L., Kepp, O. & Kroemer, G. Mitochondria: master regulators of danger signalling. Nature reviews. Molecular cell biology 13, 780-788, doi:10.1038/nrm3479 (2012). 11 Kaplan, G. G. & Windsor, J. W. The four epidemiological stages in the global evolution of inflammatory bowel disease. Nat Rev Gastroenterol Hepatol 18, 56-66, doi:10.1038/s41575-020-00360-x (2021). 12 Dinse, G. E. et al. Increasing Prevalence of Antinuclear Antibodies in the United States. Arthritis Rheumatol 72, 1026-1035, doi:10.1002/art.41214 (2020). 13 Wang, R., Li, Z., Liu, S. & Zhang, D. Global, regional and national burden of inflammatory bowel disease in 204 countries and territories from 1990 to 2019: a systematic analysis based on the Global Burden of Disease Study 2019. BMJ Open 13, e065186, doi:10.1136/bmjopen-2022-065186 (2023). 14 Duarte-Garcia, A. et al. Rising incidence and prevalence of systemic lupus erythematosus: a population-based study over four decades. Ann Rheum Dis, doi:10.1136/annrheumdis-2022-222276 (2022). 15 Walton, C. et al. Rising prevalence of multiple sclerosis worldwide: Insights from the Atlas of MS, third edition. Mult Scler 26, 1816-1821, doi:10.1177/1352458520970841 (2020). 16 Shi, G. et al. Estimation of the global prevalence, incidence, years lived with disability of rheumatoid arthritis in 2019 and forecasted incidence in 2040: results from the Global Burden of Disease Study 2019. Clin Rheumatol 42, 2297-2309, doi:10.1007/s10067-023-06628-2 (2023). 17 Furman, D. et al. Chronic inflammation in the etiology of disease across the life span. Nat Med 25, 1822-1832, doi:10.1038/s41591-019-0675-0 (2019). An excellent account of how chronic inflammation changes as we age and how environmental factors impact on inflammatory diseases 18 Gontier, N. in Encyclopedia of evolutionary biology Vol. 4 261-271 (Elsevier, 2016). 19 Garg, S., Zimorski, V. & Martin, W. F. in Encyclopedia of evolutionary biology Vol. 1 511-517 (Elsevier, 2016). 20 Dacks, J. B. et al. The changing view of eukaryogenesis - fossils, cells, lineages and how they all come together. J Cell Sci 129, 3695-3703, doi:10.1242/jcs.178566 (2016). 12 21 Roger, A. J., Munoz-Gomez, S. A. & Kamikawa, R. The Origin and Diversification of Mitochondria. Curr Biol 27, R1177-R1192, doi:10.1016/j.cub.2017.09.015 (2017). 22 Sagan, L. On the origin of mitosing cells. J Theor Biol 14, 255-274, doi:10.1016/0022-5193(67)90079-3 (1967). A classic paper that led to the acceptance of endosymbiosis as the origin of mitochondria 23 Zaremba-Niedzwiedzka, K. et al. Asgard archaea illuminate the origin of eukaryotic cellular complexity. Nature 541, 353-358, doi:10.1038/nature21031 (2017). 24 Martin, W. F. & Mentel, M. The origin of mitochondria. Nature Education 3, 58 (2010). 25 John, P. & Whatley, F. R. Paracoccus denitrificans and the evolutionary origin of the mitochondrion. Nature 254, 495-498, doi:10.1038/254495a0 (1975). 26 Geiger, O., Sanchez-Flores, A., Padilla-Gomez, J. & Degli Esposti, M. Multiple approaches of cellular metabolism define the bacterial ancestry of mitochondria. Sci Adv 9, eadh0066, doi:10.1126/sciadv.adh0066 (2023). 27 Martin, W. & Muller, M. The hydrogen hypothesis for the first eukaryote. Nature 392, 37-41, doi:10.1038/32096 (1998). 28 Raval, P. K., Martin, W. F. & Gould, S. B. Mitochondrial evolution: Gene shuffling, endosymbiosis, and signaling. Sci Adv 9, eadj4493, doi:10.1126/sciadv.adj4493 (2023). 29 Wallace, D. C. Mitochondrial diseases in man and mouse. Science (New York, N.Y 283, 1482-1488 (1999). 30 Gorman, G. S. et al. Mitochondrial diseases. Nat Rev Dis Primers 2, 16080, doi:10.1038/nrdp.2016.80 (2016). 31 Gustafsson, C. M., Falkenberg, M. & Larsson, N. G. Maintenance and Expression of Mammalian Mitochondrial DNA. Annual review of biochemistry 85, 133-160, doi:10.1146/annurev-biochem-060815-014402 (2016). 32 Rath, S. et al. MitoCarta3.0: an updated mitochondrial proteome now with sub- organelle localization and pathway annotations. Nucleic acids research 49, D1541- D1547, doi:10.1093/nar/gkaa1011 (2021). 33 Morgenstern, M. et al. Quantitative high-confidence human mitochondrial proteome and its dynamics in cellular context. Cell metabolism 33, 2464-2483 e2418, doi:10.1016/j.cmet.2021.11.001 (2021). 34 Gross, J. & Bhattacharya, D. Mitochondrial and plastid evolution in eukaryotes: an outsiders' perspective. Nat Rev Genet 10, 495-505, doi:10.1038/nrg2610 (2009). 35 Paschen, S. A., Neupert, W. & Rapaport, D. Biogenesis of beta-barrel membrane proteins of mitochondria. Trends Biochem Sci 30, 575-582, doi:10.1016/j.tibs.2005.08.009 (2005). 36 Gross, A. et al. Caspase cleaved BID targets mitochondria and is required for cytochrome c release, while Bcl-Xl prevents this release but not tumor necrosis factor- R1/Fas death. J. Biol Chem 274, 1156-1163 (1999). 37 Liu, X. S., Kim, C. N., Yang, J., Jemmerson, R. & Wang, X. D. Induction of apoptotic program in cell-free extracts - requirement for datp and cytochrome c. Cell 86, 147- 157 (1996). 38 Giacomello, M., Pyakurel, A., Glytsou, C. & Scorrano, L. The cell biology of mitochondrial membrane dynamics. Nat Rev Mol Cell Biol 21, 204-224, doi:10.1038/s41580-020-0210-7 (2020). 39 Kalkavan, H. & Green, D. R. MOMP, cell suicide as a BCL-2 family business. Cell Death Differ 25, 46-55, doi:10.1038/cdd.2017.179 (2018). 13 40 Suhaili, S. H., Karimian, H., Stellato, M., Lee, T. H. & Aguilar, M. I. Mitochondrial outer membrane permeabilization: a focus on the role of mitochondrial membrane structural organization. Biophys Rev 9, 443-457, doi:10.1007/s12551-017-0308-0 (2017). 41 He, B. et al. Mitochondrial cristae architecture protects against mtDNA release and inflammation. Cell Rep 41, 111774, doi:10.1016/j.celrep.2022.111774 (2022). 42 Frezza, C. et al. OPA1 controls apoptotic cristae remodeling independently from mitochondrial fusion. Cell 126, 177-189, doi:10.1016/j.cell.2006.06.025 (2006). 43 Ott, M., Zhivotovsky, B. & Orrenius, S. Role of cardiolipin in cytochrome c release from mitochondria. Cell Death Differ 14, 1243-1247, doi:10.1038/sj.cdd.4402135 (2007). 44 Munoz-Gomez, S. A., Slamovits, C. H., Dacks, J. B. & Wideman, J. G. The evolution of MICOS: Ancestral and derived functions and interactions. Commun Integr Biol 8, e1094593, doi:10.1080/19420889.2015.1094593 (2015). 45 Friedman, J. R., Mourier, A., Yamada, J., McCaffery, J. M. & Nunnari, J. MICOS coordinates with respiratory complexes and lipids to establish mitochondrial inner membrane architecture. Elife 4, doi:10.7554/eLife.07739 (2015). 46 Bernardi, P. & Di Lisa, F. The mitochondrial permeability transition pore: molecular nature and role as a target in cardioprotection. Journal of molecular and cellular cardiology 78, 100-106, doi:10.1016/j.yjmcc.2014.09.023 (2015). 47 Scarlett, J. L. & Murphy, M. P. Release of apoptogenic proteins from the mitochondrial intermembrane space during the mitochondrial permeability transition. FEBS Letters 418, 282-286 (1997). 48 Bernardi, P. et al. Identity, structure, and function of the mitochondrial permeability transition pore: controversies, consensus, recent advances, and future directions. Cell Death Differ 30, 1869-1885, doi:10.1038/s41418-023-01187-0 (2023). 49 Liu, B. et al. CpG methylation patterns of human mitochondrial DNA. Sci Rep 6, 23421, doi:10.1038/srep23421 (2016). 50 Riley, J. S. & Tait, S. W. Mitochondrial DNA in inflammation and immunity. EMBO Rep 21, e49799, doi:10.15252/embr.201949799 (2020). 51 Kim, J., Kim, H. S. & Chung, J. H. Molecular mechanisms of mitochondrial DNA release and activation of the cGAS-STING pathway. Exp Mol Med 55, 510-519, doi:10.1038/s12276-023-00965-7 (2023). 52 McArthur, K. et al. BAK/BAX macropores facilitate mitochondrial herniation and mtDNA efflux during apoptosis. Science (New York, N.Y 359, doi:10.1126/science.aao6047 (2018). First description of a role for BAK/BAX and mitochondrial herniation in the relaease of mtDNA 53 Kim, J. et al. VDAC oligomers form mitochondrial pores to release mtDNA fragments and promote lupus-like disease. Science (New York, N.Y 366, 1531-1536, doi:10.1126/science.aav4011 (2019). 54 Xian, H. et al. Oxidized DNA fragments exit mitochondria via mPTP- and VDAC- dependent channels to activate NLRP3 inflammasome and interferon signaling. Immunity 55, 1370-1385 e1378, doi:10.1016/j.immuni.2022.06.007 (2022). Evidence for oxidised mtDNA as an activator of the NLRP3 inflammasome 55 Mills, E. L., Kelly, B. & O'Neill, L. A. J. Mitochondria are the powerhouses of immunity. Nat Immunol 18, 488-498, doi:10.1038/ni.3704 (2017). 56 Billingham, L. K. et al. Mitochondrial electron transport chain is necessary for NLRP3 inflammasome activation. Nat Immunol 23, 692-704, doi:10.1038/s41590- 022-01185-3 (2022). 14 Evidence that mitochondrial phosphocreatine generated from ATP derived from oxidative phosphorylation is required for ATP production in the cytosol by Creatine Kinase B, for NLRP3 activation 57 Chowdhury, A., Witte, S. & Aich, A. Role of Mitochondrial Nucleic Acid Sensing Pathways in Health and Patho-Physiology. Front Cell Dev Biol 10, 796066, doi:10.3389/fcell.2022.796066 (2022). 58 Chen, Y. G. & Hur, S. Cellular origins of dsRNA, their recognition and consequences. Nat Rev Mol Cell Biol 23, 286-301, doi:10.1038/s41580-021-00430-1 (2022). 59 Seth, R. B., Sun, L., Ea, C. K. & Chen, Z. J. Identification and characterization of MAVS, a mitochondrial antiviral signaling protein that activates NF-kappaB and IRF 3. Cell 122, 669-682, doi:10.1016/j.cell.2005.08.012 (2005). 60 Wang, F., Zhang, D., Zhang, D., Li, P. & Gao, Y. Mitochondrial Protein Translation: Emerging Roles and Clinical Significance in Disease. Front Cell Dev Biol 9, 675465, doi:10.3389/fcell.2021.675465 (2021). 61 Walker, J. E., Carroll, J., Altman, M. C. & Fearnley, I. M. Chapter 6 Mass spectrometric characterization of the thirteen subunits of bovine respiratory complexes that are encoded in mitochondrial DNA. Methods in enzymology 456, 111- 131, doi:10.1016/S0076-6879(08)04406-6 (2009). 62 Le, Y., Murphy, P. M. & Wang, J. M. Formyl-peptide receptors revisited. Trends Immunol 23, 541-548, doi:10.1016/s1471-4906(02)02316-5 (2002). 63 Dorward, D. A. et al. Novel role for endogenous mitochondrial formylated peptide- driven formyl peptide receptor 1 signalling in acute respiratory distress syndrome. Thorax 72, 928-936, doi:10.1136/thoraxjnl-2017-210030 (2017). 64 Cai, N. et al. Mitochondrial DNA variants modulate N-formylmethionine, proteostasis and risk of late-onset human diseases. Nat Med 27, 1564-1575, doi:10.1038/s41591- 021-01441-3 (2021). A fascinating report linking mitochondrial N-formylmethionine formation and pathology 65 Paradies, G., Paradies, V., Ruggiero, F. M. & Petrosillo, G. Role of Cardiolipin in Mitochondrial Function and Dynamics in Health and Disease: Molecular and Pharmacological Aspects. Cells 8, doi:10.3390/cells8070728 (2019). 66 Pizzuto, M. & Pelegrin, P. Cardiolipin in Immune Signaling and Cell Death. Trends Cell Biol 30, 892-903, doi:10.1016/j.tcb.2020.09.004 (2020). 67 Dudek, J. Role of Cardiolipin in Mitochondrial Signaling Pathways. Front Cell Dev Biol 5, 90, doi:10.3389/fcell.2017.00090 (2017). 68 Iyer, S. S. et al. Mitochondrial cardiolipin is required for Nlrp3 inflammasome activation. Immunity 39, 311-323, doi:10.1016/j.immuni.2013.08.001 (2013). 69 Matzinger, P. Tolerance, danger, and the extended family. Annu Rev Immunol 12, 991- 1045, doi:10.1146/annurev.iy.12.040194.005015 (1994). 70 Murphy, M. P. How mitochondria produce reactive oxygen species. Biochem J 417, 1- 13, doi:BJ20081386 (2009). An overview of how mitochondrial redox signals may be generated 71 Wong, H. S., Dighe, P. A., Mezera, V., Monternier, P. A. & Brand, M. D. Production of superoxide and hydrogen peroxide from specific mitochondrial sites under different bioenergetic conditions. The Journal of biological chemistry 292, 16804-16809, doi:10.1074/jbc.R117.789271 (2017). 72 Robb, E. L. et al. Control of mitochondrial superoxide production by reverse electron transport at complex I. The Journal of biological chemistry 293, 9869-9879, doi:10.1074/jbc.RA118.003647 (2018). 15 73 Wright, J. J. et al. Reverse Electron Transfer by Respiratory Complex I Catalyzed in a Modular Proteoliposome System. Journal of the American Chemical Society 144, 6791-6801, doi:10.1021/jacs.2c00274 (2022). 74 Roca, F. J., Whitworth, L. J., Prag, H. A., Murphy, M. P. & Ramakrishnan, L. Tumor necrosis factor induces pathogenic mitochondrial ROS in tuberculosis through reverse electron transport. Science (New York, N.Y 376, eabh2841, doi:10.1126/science.abh2841 (2022). 75 Murphy, M. P. & Chouchani, E. T. Why succinate? Physiological regulation by a mitochondrial coenzyme Q sentinel. Nat Chem Biol 18, 461-469, doi:10.1038/s41589- 022-01004-8 (2022). 76 Mills, E. L. et al. Succinate Dehydrogenase Supports Metabolic Repurposing of Mitochondria to Drive Inflammatory Macrophages. Cell 167, 457-470 e413, doi:10.1016/j.cell.2016.08.064 (2016). This paper describes how mitochondrial metablism can be repurposed to generate succinate as a signal 77 Xiao, H. et al. A Quantitative Tissue-Specific Landscape of Protein Redox Regulation during Aging. Cell 180, 968-983 e924, doi:10.1016/j.cell.2020.02.012 (2020). 78 Holmstrom, K. M. & Finkel, T. Cellular mechanisms and physiological consequences of redox-dependent signalling. Nature reviews. Molecular cell biology 15, 411-421, doi:10.1038/nrm3801 (2014). 79 Christman, M. F., Storz, G. & Ames, B. N. OxyR, a positive regulator of hydrogen peroxide-inducible genes in Escherichia coli and Salmonella typhimurium, is homologous to a family of bacterial regulatory proteins. Proc Natl Acad Sci U S A 86, 3484-3488, doi:10.1073/pnas.86.10.3484 (1989). 80 Redza-Dutordoir, M. & Averill-Bates, D. A. Activation of apoptosis signalling pathways by reactive oxygen species. Biochim Biophys Acta 1863, 2977-2992, doi:10.1016/j.bbamcr.2016.09.012 (2016). 81 West, A. P. et al. TLR signalling augments macrophage bactericidal activity through mitochondrial ROS. Nature 472, 476-480, doi:10.1038/nature09973 (2011). 82 Ryan, D. G. et al. Coupling Krebs cycle metabolites to signalling in immunity and cancer. Nat Metab 1, 16-33, doi:10.1038/s42255-018-0014-7 (2019). 83 Murphy, M. P. & O'Neill, L. A. J. Krebs Cycle Reimagined: The Emerging Roles of Succinate and Itaconate as Signal Transducers. Cell 174, 780-784, doi:10.1016/j.cell.2018.07.030 (2018). 84 Martinez-Reyes, I. & Chandel, N. S. Mitochondrial TCA cycle metabolites control physiology and disease. Nat Commun 11, 102, doi:10.1038/s41467-019-13668-3 (2020). 85 Sivanand, S., Viney, I. & Wellen, K. E. Spatiotemporal Control of Acetyl-CoA Metabolism in Chromatin Regulation. Trends Biochem Sci 43, 61-74, doi:10.1016/j.tibs.2017.11.004 (2018). 86 Tannahill, G. M. et al. Succinate is an inflammatory signal that induces IL-1beta through HIF-1alpha. Nature 496, 238-242, doi:10.1038/nature11986 (2013). Evidence for macrophage-derived succinate being a pro-inflammatory signal 87 Selak, M. A. et al. Succinate links TCA cycle dysfunction to oncogenesis by inhibiting HIF-alpha prolyl hydroxylase. Cancer Cell 7, 77-85, doi:10.1016/j.ccr.2004.11.022 (2005). A key paper linking succinate to HIF1-alpha activation 88 Matilainen, O., Quiros, P. M. & Auwerx, J. Mitochondria and Epigenetics - Crosstalk in Homeostasis and Stress. Trends Cell Biol 27, 453-463, doi:10.1016/j.tcb.2017.02.004 (2017). 16 89 Santos, J. H. Mitochondria signaling to the epigenome: A novel role for an old organelle. Free Radic Biol Med 170, 59-69, doi:10.1016/j.freeradbiomed.2020.11.016 (2021). 90 Mills, E. L. et al. Itaconate is an anti-inflammatory metabolite that activates Nrf2 via alkylation of KEAP1. Nature 556, 113-117, doi:10.1038/nature25986 (2018). 91 Day, E. A. & O'Neill, L. A. J. Protein targeting by the itaconate family in immunity and inflammation. Biochem J 479, 2499-2510, doi:10.1042/BCJ20220364 (2022). 92 McGettrick, A. F. & O'Neill, L. A. Two for the price of one: itaconate and its derivatives as an anti-infective and anti-inflammatory immunometabolite. Curr Opin Immunol 80, 102268, doi:10.1016/j.coi.2022.102268 (2023). 93 DeBerardinis, R. J. & Chandel, N. S. Fundamentals of cancer metabolism. Sci Adv 2, e1600200, doi:10.1126/sciadv.1600200 (2016). 94 Liberti, M. V. & Locasale, J. W. The Warburg Effect: How Does it Benefit Cancer Cells? Trends Biochem Sci 41, 211-218, doi:10.1016/j.tibs.2015.12.001 (2016). 95 DeBerardinis, R. J. & Chandel, N. S. We need to talk about the Warburg effect. Nat Metab 2, 127-129, doi:10.1038/s42255-020-0172-2 (2020). 96 Weinberg, F. et al. Mitochondrial metabolism and ROS generation are essential for Kras-mediated tumorigenicity. Proceedings of the National Academy of Sciences of the United States of America 107, 8788-8793, doi:10.1073/pnas.1003428107 (2010). 97 Martinez-Reyes, I. et al. Mitochondrial ubiquinol oxidation is necessary for tumour growth. Nature 585, 288-292, doi:10.1038/s41586-020-2475-6 (2020). 98 Sena, L. A. et al. Mitochondria are required for antigen-specific T cell activation through reactive oxygen species signaling. Immunity 38, 225-236, doi:10.1016/j.immuni.2012.10.020 (2013). 99 Frenkel-Pinter, M. et al. Adaptation and Exaptation: From Small Molecules to Feathers. J Mol Evol 90, 166-175, doi:10.1007/s00239-022-10049-1 (2022). 100 Jarman, O. D., Biner, O., Wright, J. J. & Hirst, J. Paracoccus denitrificans: a genetically tractable model system for studying respiratory complex I. Sci Rep 11, 10143, doi:10.1038/s41598-021-89575-9 (2021). 101 Henry, M. F. & Vignais, P. M. Production of superoxide anions in Paracoccus denitrificans. Arch Biochem Biophys 203, 365-371, doi:10.1016/0003- 9861(80)90188-5 (1980). 102 Kotlyar, A. B. & Borovok, N. NADH oxidation and NAD+ reduction catalysed by tightly coupled inside-out vesicles from Paracoccus denitrificans. Eur J Biochem 269, 4020-4024, doi:10.1046/j.1432-1033.2002.03091.x (2002). 103 Hong, Y., Zeng, J., Wang, X., Drlica, K. & Zhao, X. Post-stress bacterial cell death mediated by reactive oxygen species. Proc Natl Acad Sci U S A 116, 10064-10071, doi:10.1073/pnas.1901730116 (2019). 104 Lopatina, A., Tal, N. & Sorek, R. Abortive Infection: Bacterial Suicide as an Antiviral Immune Strategy. Annu Rev Virol 7, 371-384, doi:10.1146/annurev-virology-011620- 040628 (2020). 105 Toyofuku, M., Schild, S., Kaparakis-Liaskos, M. & Eberl, L. Composition and functions of bacterial membrane vesicles. Nat Rev Microbiol 21, 415-430, doi:10.1038/s41579-023-00875-5 (2023). 106 Horvath, P. & Barrangou, R. CRISPR/Cas, the immune system of bacteria and archaea. Science (New York, N.Y 327, 167-170, doi:10.1126/science.1179555 (2010). 107 Georjon, H. & Bernheim, A. The highly diverse antiphage defence systems of bacteria. Nat Rev Microbiol, doi:10.1038/s41579-023-00934-x (2023). 108 Li, Y. et al. cGLRs are a diverse family of pattern recognition receptors in innate immunity. Cell, doi:10.1016/j.cell.2023.05.038 (2023). 17 109 Wu, X. et al. Phylogenetic and Molecular Evolutionary Analysis of Mitophagy Receptors under Hypoxic Conditions. Front Physiol 8, 539, doi:10.3389/fphys.2017.00539 (2017). 110 Zhang, Q. et al. Circulating mitochondrial DAMPs cause inflammatory responses to injury. Nature 464, 104-107, doi:10.1038/nature08780 (2010). 111 Moriyama, M., Koshiba, T. & Ichinohe, T. Influenza A virus M2 protein triggers mitochondrial DNA-mediated antiviral immune responses. Nat Commun 10, 4624, doi:10.1038/s41467-019-12632-5 (2019). A role for mitochondrial DNA in the induction of anti-viral immunity in response to an RNA virus (influenza) 112 Jahun, A. S. et al. Leaked genomic and mitochondrial DNA contribute to the host response to noroviruses in a STING-dependent manner. Cell Rep 42, 112179, doi:10.1016/j.celrep.2023.112179 (2023). 113 Sun, B. et al. Dengue virus activates cGAS through the release of mitochondrial DNA. Sci Rep 7, 3594, doi:10.1038/s41598-017-03932-1 (2017). 114 West, A. P. et al. Mitochondrial DNA stress primes the antiviral innate immune response. Nature 520, 553-557, doi:10.1038/nature14156 (2015). 115 Colaco, C. B., Scadding, G. K. & Lockhart, S. Anti-cardiolipin antibodies in neurological disorders: cross-reaction with anti-single stranded DNA activity. Clin Exp Immunol 68, 313-319 (1987). 116 Colapietro, F., Lleo, A. & Generali, E. Antimitochondrial Antibodies: from Bench to Bedside. Clin Rev Allergy Immunol 63, 166-177, doi:10.1007/s12016-021-08904-y (2022). 117 Chen, P. M. & Tsokos, G. C. Mitochondria in the Pathogenesis of Systemic Lupus Erythematosus. Curr Rheumatol Rep 24, 88-95, doi:10.1007/s11926-022-01063-9 (2022). 118 Becker, Y. et al. Autoantibodies in Systemic Lupus Erythematosus Target Mitochondrial RNA. Front Immunol 10, 1026, doi:10.3389/fimmu.2019.01026 (2019). 119 Chouchani, E. T. et al. Ischaemic accumulation of succinate controls reperfusion injury through mitochondrial ROS. Nature 515, 431-435, doi:10.1038/nature13909 (2014). 120 Littlewood-Evans, A. et al. GPR91 senses extracellular succinate released from inflammatory macrophages and exacerbates rheumatoid arthritis. The Journal of experimental medicine 213, 1655-1662, doi:10.1084/jem.20160061 (2016). 121 Li, Q. et al. RNA editing underlies genetic risk of common inflammatory diseases. Nature 608, 569-577, doi:10.1038/s41586-022-05052-x (2022). 122 Hooftman, A. et al. Macrophage fumarate hydratase restrains mtRNA-mediated interferon production. Nature 615, 490-498, doi:10.1038/s41586-023-05720-6 (2023). 123 Zecchini, V. et al. Fumarate induces vesicular release of mtDNA to drive innate immunity. Nature 615, 499-506, doi:10.1038/s41586-023-05770-w (2023). 124 Sciacovelli, M. et al. Fumarate is an epigenetic modifier that elicits epithelial-to- mesenchymal transition. Nature 537, 544-547, doi:10.1038/nature19353 (2016). 125 Whiteley, M., Diggle, S. P. & Greenberg, E. P. Progress in and promise of bacterial quorum sensing research. Nature 551, 313-320, doi:10.1038/nature24624 (2017). 126 Lopez-Domenech, G. et al. Miro proteins coordinate microtubule- and actin- dependent mitochondrial transport and distribution. EMBO J 37, 321-336, doi:10.15252/embj.201696380 (2018). 18 127 Debattisti, V., Gerencser, A. A., Saotome, M., Das, S. & Hajnoczky, G. ROS Control Mitochondrial Motility through p38 and the Motor Adaptor Miro/Trak. Cell Rep 21, 1667-1680, doi:10.1016/j.celrep.2017.10.060 (2017). 128 Croon, M. et al. FGF21 modulates mitochondrial stress response in cardiomyocytes only under mild mitochondrial dysfunction. Sci Adv 8, eabn7105, doi:10.1126/sciadv.abn7105 (2022). 129 Zorov, D. B., Juhaszova, M. & Sollott, S. J. Mitochondrial reactive oxygen species (ROS) and ROS-induced ROS release. Physiol Rev 94, 909-950, doi:10.1152/physrev.00026.2013 (2014). 130 Campello, S. et al. Orchestration of lymphocyte chemotaxis by mitochondrial dynamics. The Journal of experimental medicine 203, 2879-2886, doi:10.1084/jem.20061877 (2006). 131 Grafstein, B. & Forman, D. S. Intracellular transport in neurons. Physiol Rev 60, 1167-1283, doi:10.1152/physrev.1980.60.4.1167 (1980). 132 Eisner, V., Picard, M. & Hajnoczky, G. Mitochondrial dynamics in adaptive and maladaptive cellular stress responses. Nat Cell Biol 20, 755-765, doi:10.1038/s41556- 018-0133-0 (2018). 133 Spees, J. L., Olson, S. D., Whitney, M. J. & Prockop, D. J. Mitochondrial transfer between cells can rescue aerobic respiration. Proc Natl Acad Sci U S A 103, 1283- 1288, doi:10.1073/pnas.0510511103 (2006). 134 Rustom, A., Saffrich, R., Markovic, I., Walther, P. & Gerdes, H. H. Nanotubular highways for intercellular organelle transport. Science (New York, N.Y 303, 1007- 1010, doi:10.1126/science.1093133 (2004). 135 Liu, D. et al. Intercellular mitochondrial transfer as a means of tissue revitalization. Signal Transduct Target Ther 6, 65, doi:10.1038/s41392-020-00440-z (2021). 136 Liu, Z., Sun, Y., Qi, Z., Cao, L. & Ding, S. Mitochondrial transfer/transplantation: an emerging therapeutic approach for multiple diseases. Cell Biosci 12, 66, doi:10.1186/s13578-022-00805-7 (2022). 137 Dong, L. F. et al. Horizontal transfer of whole mitochondria restores tumorigenic potential in mitochondrial DNA-deficient cancer cells. Elife 6, doi:10.7554/eLife.22187 (2017). 138 McCully, J. D., Levitsky, S., Del Nido, P. J. & Cowan, D. B. Mitochondrial transplantation for therapeutic use. Clin Transl Med 5, 16, doi:10.1186/s40169-016- 0095-4 (2016). 139 Hayashida, K. et al. Mitochondrial transplantation therapy for ischemia reperfusion injury: a systematic review of animal and human studies. J Transl Med 19, 214, doi:10.1186/s12967-021-02878-3 (2021). 140 Sardon Puig, L., Valera-Alberni, M., Canto, C. & Pillon, N. J. Circadian Rhythms and Mitochondria: Connecting the Dots. Front Genet 9, 452, doi:10.3389/fgene.2018.00452 (2018). 141 Chang, E. M., Chao, C. C., Wang, M. T., Hsu, C. L. & Chen, P. C. PM(2.5) promotes pulmonary fibrosis by mitochondrial dysfunction. Environ Toxicol 38, 1905-1913, doi:10.1002/tox.23817 (2023). 142 Gioscia-Ryan, R. A. et al. Lifelong voluntary aerobic exercise prevents age- and Western diet- induced vascular dysfunction, mitochondrial oxidative stress and inflammation in mice. The Journal of physiology 599, 911-925, doi:10.1113/JP280607 (2021). 19 Box 1 Remnants of the endosymbiotic origin of mitochondria Mitochondria Bacteria/virus Eukaryotic cell Unmethylated DNA Y Y N N-formylated peptides Y Y N ROS production Y Y ? Cardiolipin Y Y N dsRNA Y Y Y (but less) The mitochondria within the eukaryotic cells retain many bacterial remnants that are more typical of bacteria. The mtDNA is present in many copies and is not methylated at CpG, in contrast to much of the nuclear DNA, the mtDNA is more susceptible to oxidative damage; slightly different genetic code; the 13 polypeptides translated within mitochondria are all initiated with an N-formylmethionine; ROS levels are elevated and can be driven by high membrane potential; bidirectional transcription of mitochondrial genes – hence overlapping mRNA with the potential for dsRNA. 20 Fig.1| Pathways of molecular signal release from mitochondria Molecular signals can be released from mitochondria to the cytosol by a number of mechanisms. a, proteins such as cytochrome c can be released from the mitochondrial inter membrane space (IMS) by Mitochondrial Outer Membrane Permeabilisation (MOMP) which is due to a pore formed in the mitochondrial outer membrane by Bcl2 family proteins. Upon 21 release into the cytosol these proteins activate apoptotic signalling pathways. b, induction of the mitochondrial permeability transition pore (mPTP) by oxidative stress and elevated Ca leads to the formation of a pore in the mitochondrial inner membrane that renders the inner membrane permeable to molecules up to ~1.5 kDa, but can also lead to swelling of the mitochondrial inner membrane, potentially rupturing the outer membrane and in some cases the inner membrane. c, mtDNA and dsRNA can be released from mitochondria to the cytosol by a number of incompletely characterised pathways. One may be by the oxidation of newly formed mtDNA fragments via the mPTP. Another is by the herniation of mitochondrial inner membrane vesicles containing mtDNA possibly through MOMP-like structures. The mtDNA containing vesicles can then be processed by the endosomal pathway to expose the mtDNA to TLR4TLR9, although the other fates of these vesicles are uncertain. d, cardiolipin (CL) is normally resident in the mitochondrial inner membrane, but can be transferred to the outer membrane at contact sites. In addition, oxidation of the polyunsaturated fatty acids in CL to oxCL and /or to lysoCL can affect release of cytochrome c and can also act as a marker for mitophagy on the surface of the mitochondria. e, Many polar mitochondrial metabolites are readily exchanged across the mitochondrial inner membrane by a series of carrier proteins and can thus be generated in the mitochondria and signal to the cytosol, including itaconate, succinate and citrate. f,g, The mitochondrial respiratory chain complexes I and III can produce superoxide. Superoxide production by complex I can be by RET while that by complex III comes from a ubisemiquinone radical at the Qo site. In both cases the superoxide formation is greatly enhanced by a high protonmotive force. Once formed within the matrix, the superoxide can act itself on aconitase to release Fe and form hydrogen peroxide, or can dismutate to hydrogen peroxide by the action of MnSOD. Once formed, the hydrogen peroxide can act as a redox signal by acting on thiols, or by inducing lipid peroxidation on CL. 22 Fig. 2| How breakdown in endosymbiosis can lead to inflammation A large number of factors as depicted, such as obesity and environmental pollutants impact on mitochondria, disrupting their integrity and driving the release of a range of factors which via specific sensors drive inflammation, notable examples being NLRP3 and nucleic acid sensors. Metabolites derived from mitochondria can also provoke inflammation, and one, fumarate, when disrupted can drive release of mitochondrial dsRNA which in turn will drive Type I IFN production. Might the increase in these various provoking factors be a reason for the rise in incidence of inflammation and autoimmune diseases? 23 Fig. 3| Did nucleic acid-sensing PRRs evolve to sense mitochondrial nucleic acids? Mitochondrial DNA has been shown to be sensed by the PRRs cGAS and TLR9, whilst mitochondrial dsRNA can be sensed by the PRRs MDA-5 and RIG-1. These PRRs were however discovered as sensors of microbially-derived nucleic acids so one question is which came first? Recently it has been shown that RNA viruses such as Influenza, Dengue and Norovirus are sensed by RIG-I but also provoke the release of mitochondrial nucleic acids which are sensed in turn by the relevant PRR. This PAMP/DAMP combination would appear to be necessary for an optimal response to these viruses. Ultimately might this apply to the sensing of all microbes, whereby the response is initiated by the sensing of the microbial PAMP, which in turn drives release of the mitochondrial DAMP for an optimum response?