A pro-inflammatory response, the senescence-associated secretory phenotype, can affect development, ageing and cancer. It emerges that one trigger for this response is the presence of DNA in the cytoplasm. See Letter p.402
Senescence is a state of usually irreversible cell-cycle arrest. It is often considered to be a protective response to cellular insults, such as high levels of DNA damage or stress associated with the expression of cancer-promoting genes1. Senescent cells also exhibit a pro-inflammatory response known as the senescence-associated secretory phenotype (SASP)1,2. However, the signals that trigger the SASP have remained elusive. Some light is now shed on this mystery in papers in Nature by Dou et al.3 (page 402), in Nature Cell Biology by Glück et al.4 and in Proceedings of the National Academy of Sciences by Yang and colleagues5.
Cellular senescence occurs in normal development, as well as in several pathological contexts, including cancer and some diseases associated with old age. During development, senescent cells can harness the SASP to promote the secretion of signalling molecules called cytokines. The cytokines recruit immune cells, which can remodel tissue through cell clearance6,7. Senescence is also beneficial if cells express cancer-promoting genes or have high levels of DNA damage, because the combination of cell-cycle arrest and SASP-mediated inflammation, which can help to recruit tumour-targeting immune cells, provides a double barrier against tumour formation. However, senescence can also have harmful effects, for example, when SASP-associated factors promote the ability of cancer cells to invade other locations in the body1.
The current papers show that the enzyme cGAS is a key player in the establishment and maintenance of the SASP (Fig. 1). This protein acts a 'first responder' in the DNA-sensing branch of the innate immune system8. Unlike an adaptive immune response, which is tailored to a specific pathogen, innate immunity recognizes infection by identifying general changes that are characteristic of pathogen infection, such as DNA being present in the wrong place in the cell — the cytoplasm. If cGAS binds to cytoplasmic DNA, the enzyme catalyses production of the molecule cGAMP and triggers a signalling cascade that launches a pro-inflammatory response8. That cGAS is required for a SASP therefore suggests that cytoplasmic DNA is one key trigger for SASP initiation.
Disruption of the outer layer of the nucleus (nuclear envelope) through a degradation process known as autophagy is a feature of senescence. This process enables genomic DNA and its associated histone proteins, together known as chromatin, to escape from the nucleus into the cytoplasm9. Dou et al. and Glück et al. show that such damaged chromosomal fragments ejected into the cytoplasm are directly detected by cGAS, evoking a SASP in human cells grown in vitro, and in vivo in mice. Their results also provide an explanation for why the SASP has a tumour-suppressing role: in genetically engineered mice, cGAS and the downstream signalling protein STING were required for immune-mediated clearance of cells that carry cancer-promoting mutant Ras proteins or that have high levels of DNA damage. Dou and colleagues also report that cells seem to retain at least a partial cGAS-dependent SASP response, even when they have managed to escape the cell-cycle block triggered by mutant Ras expression. This could potentially enable immune-mediated elimination of such senescence-evading cells.
It is possible that the degree of SASP activation is linked to different cellular triggers, which could explain other effects associated with the presence of cytoplasmic chromosomes. Cell-division errors or DNA breakage can result in chromosomes being partitioned into abnormal nuclear structures called micronuclei. These structures are fragile and prone to releasing their contents into the cytoplasm10. It has been proposed that chromosomes released from micronuclei activate cGAS and that, when this occurs, cells exhibit a pro-inflammatory transcriptional response11,12. This response would probably be protective, assuming that it causes growth arrest and promotes immune-mediated clearance of the cell, because micronuclear chromosomes undergo a mutational process termed chromothripsis, which can promote tumour formation13.
The SASP has a self-reinforcing nature, so it is appealing to speculate that a micronuclear-triggered pro-inflammatory response represents an initial stage of what would later become a 'full-blown' SASP. Consistent with this idea, ruptured micronuclei initially generate only a modest level of cGAS-pathway stimulation, perhaps because cGAS activity can be inhibited by chromatin14. However, if these cells then enter senescence and initiate autophagy, autophagic degradation of the nuclear envelope would generate more cytoplasmic DNA and also lead to cleavage of the histone proteins9. This might thereby relieve the chromatin-mediated inhibition of cGAS and promote a greater level of SASP activation.
Chromatin-independent activation of a SASP is a possibility. Senescence can be a programmed event during normal development, and this type of senescence occurs through direct activation of the protein p21 and without obvious DNA damage6,7. Consistent with the idea that developmental senescence is cGAS-independent, mice engineered to lack the protein develop normally15, although partial compensation by a programmed-cell-death process6 might mask some developmental defects in this context. However, it remains unclear whether autophagic degradation of nuclear-envelope proteins is suppressed during developmental senescence. If such degradation does occur, one might expect the release of chromosomes into the cytoplasm. Investigating whether cGAS and STING act in developmental senescence could be an interesting avenue for future research.
It is also possible that SASP activation is associated with layers of downstream signalling that undergo context-dependent regulation to produce diverse cellular consequences1. cGAS–STING activates two major transcriptional responses: the production of interferon proteins by the transcription factor IRF3, and the production of secreted cytokines that are transcribed through the action of the NF-κB protein8. However, cancer cells experience selective pressure to suppress interferon production to evade immune detection16. One way in which cancer cells can achieve this is by activation of the enzyme p38, which disrupts the IRF3-mediated transcriptional response; this strategy is also used by some pathogens to elude immune detection17,18. Senescent cells also secrete factors that can suppress expression of interferon genes, which provides another way to evade immune detection. Finally, there are several 'flavours' of NF-κB signalling19, and how the SASP might affect these in various cellular contexts remains to be determined. Thus, this complex, pro-inflammatory signalling network might be modified by signals from both within and outside the cell to generate a variety of cellular outcomes.
The current papers also have relevance to cancer immunotherapies. Most such therapies have manipulated the adaptive rather than the innate immune system. However, these studies will heighten interest in the idea that boosting the SASP response and pro-inflammatory signalling through the innate immune system might also promote a robust adaptive immune response against cancer cells if cytokines aid tumour-targeting immune cells. Drugs that activate STING have been developed and are in early-stage clinical trials for cancer treatment4,20.
However, activation of innate immune signalling comes with risks, including possibly accelerating the progression of diseases associated with ageing, so it would be useful if a molecular test could identify the patients most likely to benefit from activation of the innate immune system. Perhaps cells with a high level of cytoplasmic DNA could be preferentially targeted for elimination through treatment with STING-activating drugs. The progress that has been made in these studies towards understanding the mechanisms of SASP initiation might provide a foundation that ultimately leads to improvements in clinical treatments.
Campisi, J. Annu. Rev. Physiol. 75, 685–705 (2013).
Coppé, J.-P. et al. PLoS Biol. 6, e301 (2008).
Dou, Z. et al. Nature 550, 402–406 (2017).
Glück, S. et al. Nature Cell Biol. 19, 1061–1070 (2017).
Yang, H., Wang, H., Ren, J., Chen, Q. & Chen, Z. J. Proc. Natl Acad. Sci. USA 114, E4612–E4620 (2017).
Muñoz-Espín, D. et al. Cell 155, 1104–1118 (2013).
Storer, M. et al. Cell 155, 1119–1130 (2013).
Cai, X., Chiu, Y.-H. & Chen, Z. J. Mol. Cell 54, 289–296 (2014).
Dou, Z. et al. Nature 527, 105–109 (2015).
Hatch, E. M., Fischer, A. H., Deerinck, T. J. & Hetzer, M. W. Cell 154, 47–60 (2013).
Harding, S. M. et al. Nature 548, 466–470 (2017).
Mackenzie, K. J. et al. Nature 548, 461–465 (2017).
Zhang, C.-Z. et al. Nature 522, 179–184 (2015).
Zierhut, C. & Funabiki, H. Preprint at bioRxiv http://doi.org/cdkf (2017).
Li, X.-D. et al. Science 341, 1390–1394 (2013).
Shankaran, V. et al. Nature 410, 1107–1111 (2001).
Chen, Y. et al. J. Exp. Med. 214, 991–1010 (2017).
Freund, A., Patil, C. K. & Campisi, J. EMBO J. 30, 1536–1548 (2011).
Dejardin, E. Biochem. Pharmacol. 72, 1161–1179 (2006).
Fu, J. et al. Sci. Transl. Med. 7, 283ra252 (2015).
About this article
Cite this article
Umbreit, N., Pellman, D. Genome jail-break triggers lockdown. Nature 550, 340–341 (2017). https://doi.org/10.1038/nature24146
Journal of Cell Biology (2020)
Molecular Biology of the Cell (2020)
Nature Reviews Genetics (2019)