Mitochondria are membrane-bound organelles that act as hubs of metabolism and of innate immune signalling in cells. Each mitochondrion contains several copies of the mitochondrial genome (mtDNA), which can be damaged by extrinsic environmental stressors or intrinsic genetic mutations. This can cause degradation of the mtDNA, reducing the total number of mtDNA copies in the organelle and so leading to mitochondrial dysfunction. In addition, healthy mitochondrial function relies heavily on crosstalk between mitochondria and the nucleus1,2. Writing in Nature, Tigano et al.3 uncover a mechanism by which cells sense toxic mtDNA damage to initiate an immune response in the nucleus.
Conditions of acute stress, such as viral infection or irradiation, can lead to the activation of pro-death (apoptotic) pathways in the cell. Mitochondria have a key role in these pathways. Pore-forming proteins called BAK and BAX accumulate on the mitochondrial membrane, leading to the release of cell-death factors from the organelle into the cell’s cytosolic fluid through a process called mitochondrial herniation4. In some instances, cell-death factors are not activated, in which case mitochondrial contents such as DNA and RNA are instead released into the cytosol4. The accumulation of cytosolic mtDNA and mtRNA initiates a potent antiviral response5–7. But precisely which aspects of mitochondrial dysfunction lead to the extrusion and accumulation of this mitochondrial material has been unclear.
Tigano et al. set out to examine one form of stress: cleavage of mtDNA. The group manipulated mammalian cells using ‘molecular scissors’ constructs called mitochondria-targeted TALENs (mTLNs), which generate double-strand breaks (DSBs) in mtDNA. They used RNA sequencing to analyse changes in gene expression in cells treated with mTLNs, and found increased transcription of nuclear genes involved in the innate immune response; these included interferon-response genes, which are typically involved in combating viral infections. The authors also found that the transcription factor STAT1 was modified by phosphate groups and relocated to the nucleus — a key part of the interferon response8.
Breaks in mtDNA that occur through other means, such as treatment with toxic, DNA-damaging agents or errors in replication, often lead to compromised organelle function9. But Tigano and colleagues found that the mTLN treatment reduced the number of mtDNAs by only around 60%, which did not seem to have an immediate impact on mitochondrial function. The group observed no changes in key indicators of normal mitochondrial function, such as morphology, the gradient of protons (H+ ions) across the membrane, and the generation of reactive oxygen species. These data indicate that mtDNA cleavage is a key trigger of antiviral responses.
Next, Tigano et al. set out to identify the signalling molecules that relay the message of mtDNA instability to the nucleus. Although the mTLN-treated cells had intact mitochondrial function and were not apoptotic, the group showed that BAK–BAX pores did form on the membrane, consistent with mitochondrial herniation. The authors found that mtRNA — but not mtDNA — accumulated in the cytosol of these cells. The mtRNA molecules were detected by an RNA-sensing protein called RIG-I, which is better known as a sensor of viral RNA in the cytosol10. Working with its adaptor protein on the mitochondrial outer membrane, dubbed mitochondrial antiviral signalling (MAVS), RIG-I triggers a signalling pathway that activates interferon-response genes in the nucleus10. These findings point to a framework by which cells engage mitochondrial signalling molecules in immune-surveillance mechanisms (Fig. 1).
DNA-damaging agents such as radiation, which is used to treat cancer, elicit a systemic immune response that is thought to be driven by DNA damage in the nucleus. Tigano and colleagues found that radiation depleted mtDNA numbers by 40% and elicited the same immune response as mTLNs, suggesting that DSBs occur in mtDNA as well as in nuclear DNA following irradiation. Strikingly, induction of the interferon response during irradiation was nearly completely abrogated in cells lacking mtDNA. This observation indicates that mtDNA damage caused by radiation can be a driver of interferon responses. Of note, the induction of several other innate immune responses still occurred in cells lacking mtDNA, suggesting that depletion of mtDNA specifically impairs the interferon response.
The study highlights an immunostimulatory role for mtRNA. However, questions remain. For instance, mtRNA molecules are highly unstable in nature11 — how are mtRNAs stabilized so that they accumulate in the cytosol, as was observed in the current study? Another avenue for further investigation is the factors that stimulate the formation of BAK–BAX pores following mtDNA breaks. It would be of broad interest to study whether drugs that inhibit this pore formation can suppress an inflammatory immune response. The discovery of a mechanism by which cells recognize self-RNAs from mitochondria to initiate an immune response also raises the question of whether this pathway might be involved in autoimmune disease. Finally, it would be exciting to explore whether artificially induced mtDNA damage could be used to increase the efficacy of targeted immunotherapies for cancer.
Nature 591, 372-373 (2021)