A lot of excitement has been generated by many reports demonstrating that inhibiting some aspects of the DNA damage response activate the cGAS-STING pathway and potentiate immunotherapy. The manuscript from Wayne and colleagues examines the effects of treating a panel of commonly used cancer cell lines with a number of inhibitors for the ATM/ATR/CHK1/Wee1 kinases in the presence or absence of different types of damage. One major trigger for the cGAS-STING response has been proposed to be cytoplasmic dsDNA arising from replication forks, telomeric recombination or micronuclei. Using high content imaging and several approaches, the demonstrate that ss and dsDNA is increased in some cell lines by some agents but there is not a universal response and there is significant variability in the results. They also use a BrdU labeling approach to demonstrate that some of this is of nuclear origin, as mtDNA can be a confounding factor in these analyses. They next find that while TBK1 is phosphorylated in response to many of the treatments, many expected downstream events, including the activation if IRF signaling, show no or very mild changes. This is further elaborated using a number of CHK1 inhbitors, showing a dose dependent effect on TBK1. They next examined a BRCA1 null cell line that was previously shown to activate interferon responses following DNA damage. These cells did not activate TBK1, but instead activated IKKe and showed some activation of NFKB at late time points after DDR inhibition. They next examined the interferon response using cytokine arrays and gene expression analysis and found that the TypeI interferon response was not activated by DDR inhibitors and that their effect on cytokine responses was distinct fron cGAMP. Treatment with CHK1 inhibitors potentiated DNA damage induced accumulation of dsDNA but impaired increased IRF1 and pSTAT1, reduced the nuclear accumulation of RelB and failed to activate NFKB, IRF3 or PD-L1.

The paper is timely and contains a large amount of well performed, quantitative analyses of DNA damage inhibitors. It has important data in that it demonstrates clearly that there is massive heterogeneity in the responses between different cell lines and in response to agents that target connected pathways. The overall message, that DDR inhibition may not be a universally effective way to activate the TypeI interferon response and augment immunotherapy, is one that will undoubtedly cause some controversy but the data is quite clear and the paper will serve as an important resource for the field to examine the different responses documented here. That being said, the manuscript is at times difficult to follow, as it is not always completely clear how the experiments connect sequentially, and the number of inhibitors used is confusing if one does not know all of the names (this is not the fault of the authors but I had to many times cross reference the inhibitor name to recall the target). Overall, I think that the results are important and highlight the complexity of activating the innate immune response in genetically diverse cancer cells.

Comments

1. The statement that TBK1 phosphorylation is STING independent, based on the low levels of STING in U2OS cells, is somewhat of an overstatement. First, while U2OS cells do express very low levels of STING, a point that has been previously documented, it is expressed. Making the conclusion the signaling is therefore STING independent in a number of other completely unrelated cell lines with their own genetic variability is not well founded in the absence of any experimental data, although I acknowledge it is a possibility.
2. Perhaps I missed something but the authors mention no correlation between activation of ATR based on S345 staining (line 217) but I did not find this data.
3. In Figure 3A, it appears panels of the western are spliced together. If this is not pdf artifact, it should be made more clear where panels are joined if that is the case.
4. The authors use numerous sophisticated assays but many are non-standard. Including some positive controls and example images would be helpful in many cases, particularly with the ss/dsDNA assays in Figure 1. Mean nuclear intensity is measured in many figures and no examples of actual data are shown. IRF3 is never induced- does the antibody work? I expect that some experiments were done to validate these things but in their absence, it is hard to interpret the data in some cases with no point of reference.
5. In line 517 they authors pose the question as to whether decreased membrane integrity suppresses innate immune responses- I did not follow the logic of this after reading this many times. It would be useful to expand this idea a bit more clearly, it seems counter intuitive to me given the preceding paragraphs.
6. A minor point- in line 66 the authors say that DNA breaks activate CHK1, 2 and ATR/ATM. DSBs do activate ATM/CHK2 directly but a large amount of evidence suggests ATR/CHK1 is activated by ssDNA-dsDNA transitions, not DSBs. These transitions can be generated by resection of DSBs so this gets a bit semantic but I would not make the statement the authors make in a paper without clarifying this because it is an important point when one gets into the mechanisms underlying these signaling pathways.

I want to thank the authors for their effort in clarifying details and adding more example data to support the graphed data. I think this improves the manuscript significantly and it will be a nice resource for the field moving forward.