The innate immune response mounts a defence when immune cells recognize general hallmarks of infection, such as lipopolysaccharide (LPS) molecules, which are present in many types of bacterium. However, the inappropriate unleashing of an innate immune response can lead to autoimmune disorders. Gaining a better understanding of how innate immune responses are regulated might lead to improvements in clinical treatments for such disorders. Writing in Nature, Zhong et al.1 report that DNA synthesis in organelles called mitochondria has a key role in triggering an innate immune response.
Mitochondria can regulate how immune cells respond to infection and tissue damage. For example, these organelles can produce pro- or anti-inflammatory signals by altering the levels of metabolites produced in the Krebs cycle2,3, or by changing the level of production of reactive oxygen species (ROS)4,5. More and more examples are being found of mitochondrial functions being repurposed in unexpected ways to contribute to inflammatory signalling2–5.
The inflammasome is a multiprotein complex that assembles in immune cells during an innate immune response. It provides defensive functions when the inflammasome-associated enzyme caspase-1 cleaves and activates inflammatory proteins such as IL-1β. Inflammasomes that contain the protein NLRP3 can form in immune cells called macrophages, and the initial steps in the assembly or priming of this type of inflammasome are reasonably well understood: if LPS binds to the receptor protein TLR4 on the macrophage surface, there is an increase in signalling by the NF-κB pathway. This causes an increase in expression of NLRP3 and of the precursor form of IL-1β.
However, the process that triggers inflammasome activation, which occurs when the enzyme caspase-1 is recruited to the inflammasome and aids the production of inflammatory proteins, is not fully understood. It was puzzling that many highly diverse molecular cues can trigger this step. Yet hints from experimental studies have suggested that these cues might ultimately act through a mitochondrial pathway associated with high levels of mitochondrial ROS3,6,7 — which are required to oxidize mitochondrial DNA — and the release of oxidized mitochondrial DNA, which binds to the inflammasome8.
The binding of mitochondrial DNA to an NLRP3-containing inflammasome is essential for inflammasome activation9,10. Zhong and colleagues studied mice to assess whether the availability of this mitochondrial DNA might regulate inflammation. The authors engineered animals so that their immune cells lack the protein TFAM, which is required for mitochondrial DNA replication. This led to a loss of mitochondrial DNA, resulting in defective inflammasome activation. When the authors transferred synthetic oxidized mitochondrial DNA into macrophage cells grown in vitro from the animals lacking TFAM, this triggered inflammasome activation in response to an LPS signal.
The authors investigated how mitochondrial sensing of innate-immunity triggers might lead to mitochondrial-DNA synthesis. They report that LPS binding to TLR4 activates a pathway that drives expression of the enzyme CMPK2, which is required to produce the nucleotide cytidine triphosphate (CTP) (Fig. 1). Zhong and colleagues engineered mouse macrophage cells to lack CMPK2, and found that such cells were deficient in inflammasome activation. It is unknown how CMPK2 and the mitochondrial CTP pool operate as a control point for mitochondrial-DNA synthesis in macrophages.
To track newly made mitochondrial DNA, the authors introduced a labelled building block of DNA into macrophage cells grown in vitro. When these cells received an inflammasome-activating cue, such as LPS, newly made DNA was found to be associated with the inflammasome, and DNA-sequence analysis confirmed its mitochondrial origin. Intriguingly, the authors did not find evidence that the oxidized DNA had to be mitochondrial DNA to bind to the inflammasome. The introduction of oxidized nuclear DNA could do the job just as well, suggesting that oxidized DNA is the key signal.
Zhong and colleagues’ work fills in the gap between the priming and activation of the inflammasome by indicating that newly synthesized mitochondrial DNA can give rise to oxidized mitochondrial DNA fragments that exit the organelle to activate NLRP3-containing inflammasomes. Their core conclusions are convincing; however, the solidity of these findings inevitably focuses our attention on those points that are still uncertain. One intriguing issue is the nature of the newly synthesized mitochondrial DNA. The authors’ findings suggest that this is produced by the polymerase enzyme that normally replicates mitochondrial DNA, but it is unclear whether the entire mitochondrial DNA sequence is replicated or whether replication terminates prematurely once sufficient DNA is made to generate an inflammatory signal. And is newly formed mitochondrial DNA particularly susceptible to oxidative damage? Could it be that the newly synthesized DNA lacks protection from the nucleoid proteins that normally bind to mitochondrial DNA, thereby increasing its exposure to ROS?
The authors incorporated the oxidized nucleotide 8-hydroxy-2ʹ-deoxyguanosine into cells grown in vitro as a way to generate oxidized mitochondrial DNA. This type of nucleotide is frequently found in oxidized DNA, but there are many other types of oxidative DNA modification, and it would be interesting to explore which of these can activate inflammasomes.
How do the ROS needed for DNA oxidation arise? The tacit assumption is that non-specific organelle damage generates ROS. Yet this is debatable11. I suspect that the mitochondrial ROS production during NLRP3-inflammasome activation might be just as regulated as the process of mitochondrial DNA synthesis. Perhaps the succinate molecules that accumulate after LPS stimulation are oxidized to drive mitochondrial ROS production4.
Another area worthy of future investigation is how oxidized mitochondrial DNA is released into the cytoplasm. The authors make the plausible proposal that a large mitochondrial pore might provide an exit route. One candidate is the mitochondrial permeability transition pore, which forms in response to increased levels of ROS12. However, there are also other possibilities to consider: for example, mitochondria can release microvesicles containing oxidized DNA and protein13.
The authors’ insights into the activation of NLRP3-containing inflammasomes immediately suggest targets for the development of anti-inflammatory drugs. One area to explore is inhibition of CMPK2 during inflammation, and other parts of the pathway that the authors uncovered are worth considering as targets, too.
This finding of yet another fascinating link between mitochondria and inflammatory signalling in the innate immune system might reflect the organelle’s early evolutionary origins as a bacterial cell. This inherent otherness could give mitochondria a head start in being recognized as foreign by the innate immune system.
Also writing in Nature, Dhir et al.14 report that the release of double-stranded RNA from mitochondria acts as an antiviral signal. This provides an additional example that the release of mitochondrial nucleic acids to the cytoplasm can act as a signal that triggers a defence response.
Nature 560, 176-177 (2018)